Stream Gate

The Environmental History of Silver Refining in New Spain and Mexico, 16c to 19c:
A Shift of Paradigm
Saúl José Guerrero Quintero
Department of History and Classical Studies

McGill University, Montreal
March 2015
A thesis submitted to McGill University in partial fulfilment of the requirements of
the degree of Doctor in Philosophy
© Saúl José Guerrero Quintero, 2015

1
Table of Contents
Table of Contents ........................................................................................................... 1

List of Figures ................................................................................................................ 6
List of Tables ............................................................................................................... 26
Abstract ........................................................................................................................ 38
Résumé......................................................................................................................... 39
Acknowledgements ...................................................................................................... 40
Guide to the text ........................................................................................................... 44
Introduction .................................................................................................................. 46
1 The genesis and nature of silver ores. ............................................................ 59
1.1 Introduction .................................................................................................... 59
1.2 Subduction and the New World ..................................................................... 62
1.3 The geological difference in New World silver deposits ............................... 71
1.4 The chemistry of the sources of silver in New Spain / Mexico ..................... 72
1.4.1 Silver sulphide and silver chloride ............................................................ 74
1.4.2 Argentiferous galena ................................................................................. 79
1.4.3 The silver belt in New Spain / Mexico ...................................................... 79
1.5 The chemistry of the sources of silver in Europe ........................................... 80
1.6 The silver content of New World ores ........................................................... 87
1.7 Mercury .......................................................................................................... 97
1.8 Salt, sodium chloride (NaCl) .......................................................................... 99
1.9 Lead .............................................................................................................. 100
1.10 Final remarks ................................................................................................ 101
2 The dry refining process: smelting of silver ores......................................... 103
2.1 Introduction .................................................................................................. 103
2.2 Smelting of silver ores : the human context. ................................................ 108
2.3 The chemistry of smelting and the nature of the ore .................................... 116
2.4 The infrastructure of smelting in New Spain ............................................... 123
2.5 The architecture of smelting in New Spain .................................................. 135
2.6 The environmental impact vectors from smelting........................................ 142
2.6.1 Lead: its nature and magnitude................................................................ 143
2.6.2 Lead: its directionality ............................................................................. 156
2.6.3 Lead: its source ........................................................................................ 163
2
2.6.4 Charcoal and wood: its magnitude .......................................................... 165

2.6.5 Charcoal and wood: its directionality and source.................................... 166
2.6.6 Regional and local environmental impact of smelting ............................ 169
2.7 Concluding remarks ..................................................................................... 173
3 The wet process, amalgamation of silver ores ............................................. 178
3.1 Introduction .................................................................................................. 178
3.2 The alchemy of Mercury .............................................................................. 180
3.3 The gold connection ..................................................................................... 186
3.4 The chemistry of amalgamation ................................................................... 193
3.5 The implementation by stages of the amalgamation process ....................... 199
3.6 The infrastructure of amalgamation in New Spain ...................................... 210
3.6.1 Milling the ore ......................................................................................... 211
3.6.2 The amalgamation reactor, or patio. ........................................................ 216
3.6.3 The handling of mercury within the hacienda. ....................................... 219
3.6.4 Planillas ................................................................................................... 221
3.6.5 Desazogaderas or capellinas and the recycling of mercury. .................... 222
3.7 Amalgamation: the human factor ................................................................. 231
3.8 The architecture of patio amalgamation....................................................... 234
3.9 The mass balance of amalgamation: the role of the correspondencia ......... 251
3.10 The environmental impact vectors of amalgamation ................................... 262
3.11 Final remarks ................................................................................................ 275
4 Hacienda Santa María de Regla. ................................................................. 280
4.1 Introduction .................................................................................................. 280
4.2 The Adventurers in the Mines of Real del Monte ........................................ 287
4.3 The Hacienda de Regla. ............................................................................... 293
4.3.1 Ore reception and stamp mill .................................................................. 298
4.3.2 Arrastres and water power in the Hacienda ............................................ 300
4.3.3 The patio reactor...................................................................................... 305
4.3.4 Furnace Area A: Capellinas and the recycling of mercury ..................... 308
4.3.5 Furnace Area B ........................................................................................ 315
4.3.6 Final comments on the architectural layout of the Hacienda de Regla ... 326
4.4 The mass balance of the amalgamation process at Regla, 1872 to 1888 ..... 327
4.4.1 Silver ore received for amalgamation...................................................... 328
4.4.2 Milling of silver ore destined for amalgamation. .................................... 330
4.4.3 The amalgamation tortas (cakes). ........................................................... 333
3
4.4.4 The silver content of ores used for amalgamation................................... 338

4.4.5 Interlude: pizzas or juggling acts. ............................................................ 340
4.4.6 The mass balance for salt. ....................................................................... 342
4.4.7 The mass balance for copper sulphate. .................................................... 344
4.4.8 The mass balance for mercury used during amalgamation ..................... 345
4.4.9 The mass balance for silver. .................................................................... 346
4.4.10 The mass balance for mercury consumed ............................................... 349
4.4.11 The overall mass balance for the amalgamation process at Regla .......... 352
4.4.12 The energy balance of the amalgamation process ................................... 353
4.5 The mass balance of the smelting process at Regla, 1875 to 1886. ............. 355
4.5.1 The mass balance for ore ......................................................................... 355
4.5.2 The mass balance for litharge as input .................................................... 358
4.5.3 The mass balance for silver: output ......................................................... 358
4.5.4 The mass balance for litharge: output ..................................................... 360
4.5.5 The mass balance for smelting at Regla .................................................. 361
4.6 The environmental loss vectors in the period 1872 to 1888. ....................... 365
5 The economies of refining silver. ................................................................ 376
5.1 The most valuable jewel of the Crown......................................................... 376
5.1.1 The Spanish Crown ‘ganando indulgencias con escapulario ajeno’. ...... 382
5.1.2 The financing of Almadén and the price of mercury in New Spain. ....... 386
5.1.3 The environmental impact of the Crown monopoly on mercury. ........... 389
5.2 The historiography on the economies of refining. ....................................... 395
5.3 The accounting records for Regla ................................................................ 416
5.4 The macroeconomic context in the nineteenth century. ............................... 417
5.4.1 Silver........................................................................................................ 418
5.4.2 Maize ....................................................................................................... 421
5.4.3 Salt ........................................................................................................... 423
5.4.4 Copper sulphate ....................................................................................... 425
5.4.5 Mercury ................................................................................................... 425
5.4.6 Charcoal ................................................................................................... 430
5.4.7 Litharge.................................................................................................... 431
5.5 The partial variable production costs of amalgamation at Regla, 1872-1888.
...................................................................................................................... 432
5.6 The cost of ore: the missing link in variable costs at Regla ......................... 436
4
5.7 The capital cost of amalgamation at Regla. ................................................. 439

5.8 The partial variable production cost of amalgamation as a function of silver
content of the ore. ........................................................................................ 441
5.9 The partial variable production costs of smelting at Regla, 1872-1888....... 445
5.10 The variable production cost of smelting as a function of the silver content of
the ore........................................................................................................... 447
5.11 Patio amalgamation vs smelting at Regla .................................................... 451
5.12 Extrapolating the results from Regla to other historical scenarios............... 453
5.13 The structure of labour costs at Regla. ......................................................... 463
5.14 Patio vs Barrel Amalgamation ..................................................................... 471
6 The environmental impact of silver refining: a shift of paradigm. .............. 488
6.1 The magnitude of the impact: the method employed ................................... 488
6.1.1 Caja of Zacatecas .................................................................................... 495
6.1.2 Caja of Guanajuato.................................................................................. 501
6.1.3 Caja of Mexico ........................................................................................ 505
6.1.4 Caja of Durango ...................................................................................... 507
6.1.5 Caja of San Luis Potosí ........................................................................... 511
6.1.6 Caja of Guadalajara ................................................................................. 516
6.1.7 Caja of Pachuca ....................................................................................... 520
6.1.8 Caja of Sombrerete.................................................................................. 524
6.1.9 Caja of Bolaños ....................................................................................... 528
6.1.10 Caja of Rosario....................................................................................... 530
6.1.11 Caja of Zimapán. .................................................................................... 532
6.1.12 Caja of Chihuahua. ................................................................................. 535
6.1.13 Aggregate totals for New Spain .............................................................. 537
6.1.14 Aggregate totals for Mexico, 1820 to 1900 ............................................. 547
6.1.15 Environmental impact vectors, sixteenth to nineteenth century: conclusions.
................................................................................................................. 549
6.2 A change of paradigm. ................................................................................. 557
6.3 The human choices. ...................................................................................... 562
6.3.1 What did they know and when did they know it? ................................... 563
6.3.2 Historicity and the reality of the refining process ................................... 568
6.3.3 The option of smelting all the ores in New Spain ................................... 575
Epilogue ..................................................................................................................... 580
APPENDICES ........................................................................................................... 585
5
A. Silver production in smelting haciendas...................................................... 585

B. The accounting books of Regla.................................................................... 587
C. Inventory areas at the Hacienda de Regla ................................................... 596
D. Report of the costs of refining by cazo amalgamation and smelting, 1801. 604
E. Sensitivity matrix for refining costs ............................................................. 609
F. Smelting costs in Europe. ............................................................................ 611
Glossary of technical terms ........................................................................................ 614
Archival sources......................................................................................................... 621
Bibliography .............................................................................................................. 622
6
List of Figures
Figure 1-1. Top ten silver deposits ranked by aggregate production and main mercury
sources up to the late twentieth century. Data from Table 1-I. ............ 64
Figure 1-2. The spreading ocean floor subducts under the continental crust of the
Americas (illustration based on Sillitoe in footnote 51). ...................... 68
Figure 1-3. Subduction along the Andes, Cordillera and Japan (based on illustration
in footnote 55). ..................................................................................... 70
Figure 1-4. A simplified representation of Sillitoe’s interpretation of the effect of
weathering on an original (hypogene) non-weathered sulphidic silver
deposit. Drawing based on interpretation of Sillitoe in footnote 74..... 76
Figure 1-5. The main historical silver, lead and salt deposits of New Spain / Mexico
(based on footnote 86). ......................................................................... 81
Figure 1-6. Location of main historic silver mining regions in Europe by mid
sixteenth century. Adapted from Blanchard in footnote 98. Map and
inset not to scale. .................................................................................. 84
Figure 1-7. Altitude of the main historical silver ore deposits found in the Hispanic
New World and in Europe. For sources see footnote 100. ................... 86
Figure 1-8. Plot of average silver content in ores of New Spain / Mexico grouped by
century, data from Table 1-II................................................................ 90
Figure 1-9. Plot of a selected range of data for the silver content in ores of the Vice-
Royalty of Peru / Bolivia, as per Table 1-III, grouped by century. ...... 91
Figure 1-10. Plot of published data for the silver content in ores of Europe, as per
Table 1-IV, grouped by century. .......................................................... 91

7
Figure 2-1. Schematic diagram of two stage refining of silver ores using lead, adapted
from Craddock, footnote 179. ............................................................ 121
Figure 2-2. a) Molino mill stones, Monte Caldera, San Luis Potosí. The diameter can
reach 2 m b) a modern reconstruction of a molino in Zacatecas, at the
exit of the El Edén mine c) drawing of a molino, reproduced from
footnote 192. ....................................................................................... 125
Figure 2-3. Exterior of smelting furnace at the ruins of the Hacienda Santa María in
Monte Caldera. The arched port would have been used to feed ore or fuel
to the furnace. ..................................................................................... 129
Figure 2-4. Ruins of the Hacienda de Aranzazu in Guadalcazar. a) front view
showing archways under two smelting furnace chimneys, height around
7 m b) back view of chimneys, showing possible aperture for drive shaft
of bellows c) section of hacienda wall d) fields of grasas e) image from
Google Earth © 2014 DigitalGlobe, 22°37’21” N 100°24’9” W. ...... 130
Figure 2-5. Section of a blast furnace as found at Regla. Adapted from footnote 213.
............................................................................................................ 131
Figure 2-6. a) Ruins of smelting hacienda HMC2, with chimney b) alquibris (tuyere)
c) one of two bowl-shaped depressions with caved-in roof, connected via
the tubular section observed in d). ...................................................... 133
Figure 2-7. Mounds of grasas in Monte Caldera a) Hacienda Santa Maria, chimney
of Figure 2-3 in the background b) Hacienda HMC2. ....................... 134
Figure 2-8. The solid white lines encapsulate the minimum area that can be clearly
identified with each smelting hacienda, the dotted line the minimum area
of the extant dumps of grasas. All satellite images from Google Earth ©
2014 DigitalGlobe. The Hacienda Santa María lies at 22°12’10” N

8
100°44’27” W, and the Hacienda HMC1 at 22°12’31” N 100°44’47”W.
............................................................................................................ 139
Figure 2-9. Copy (1986) drawn by Carlos Morán de la Rosa of the original map by
Captain Manuel Pascal de Burgoa, 1794, showing the division of the city
of San Luis Potosí into barrios (quarters) by the Viceroy Marques de
Branciforte. Digital copy courtesy of AHSLP, Colección Mapas y
Planos. ................................................................................................ 140
Figure 2-10. Scheme of the mass balance for lead during the smelting of silver ores.
Letters in bold indicate mass input, letters in capitals indicate mass
output. ................................................................................................. 147
Figure 2-11. Main areas of lead deposition within and around the reconstruction of
the Hacienda Santa Maria in Monte Caldera, adapted from original
drawing with permission of Prof. Guadalupe Salazar González, footnote
267. ..................................................................................................... 158
Figure 2-12. Digital copy of photograph by Charles Waite titled ‘Mexican Adobe
Smelter Taxco Guerrero’, 1905, number 13 in the series Tema y
Tecnología (CIG-AGN)...................................................................... 160
Figure 2-13. Relative alignment of arrays of furnaces from three different smelting
haciendas in the area of Monte Caldera, adapted from the architectural
reconstructions in footnote 275. ......................................................... 162
Figure 2-14. Location of the main mines, smelting haciendas, charcoal production,
agricultural and cattle rearing areas around the town of San Luis Potosí,
adapted from the original map in footnote 300. ................................. 171
Figure 3-1. Timeline of main stages in the implementation of amalgamation in the
New World. ........................................................................................ 210

9
Figure 3-2. The main stages of the patio amalgamation process as practised in New
Spain / Mexico. Dashed lines indicate optional stages. ...................... 212
Figure 3-3. a) a horse-powered Molino in Mexico, late nineteenth century reproduced
from footnote 418 b) example of the pit of a molino, from where the finer
grains that have been shovelled through the mesh in a) are withdrawn
via the arched tunnel and taken to the tahonas/arrastres. Photo taken in
the ruins of the Hacienda San Juan Nepomuceno, Marfil, Guanajuato c)
drawing of a tahona /arrastre, showing its four voladoras, reproduced
from MacDonald, footnote 420 d) photo of a wasted voladora stone, over
1.6 m in length, taken at the home of the Morrill family, previously the
Hacienda Bustos, Guanajuato. ........................................................... 215
Figure 3-4 a) patio of the Hacienda de San Xavier, Guanajuato, photo reproduced
from Antúnez Echegaray, footnote 423 b) patio of the Hacienda de La
Luz, Guanajuato, photo reproduced from Hernandez and Rangel,
footnote 425 c) patio of unidentified hacienda in Guanajuato, photo
reproduced from Rickard, footnote 426. ............................................ 217
Figure 3-5. Amalgamation in ‘fixed casks’, Born’s variation on Barba’s cazo process.
Illustration reproduced from footnote 427. ........................................ 218
Figure 3-6. Photo of an azoguería (mercury room), showing mercury flasks, scales
and vertical white manga held by chains from a beam to the right of the
background. Original photo from footnote 435. ................................ 221
Figure 3-7. a) cross section of a planilla, reproduced from Laur, footnote 439 b) photo
of planilleros at work, showing the cárcamo running with the waste
water from the washings, the inclined planillas and the scoop made from
a bull’s horn in their right hand. Reproduction of illustration in footnote

10
438. It seems to correspond to a larger image from where a more reduced
area has been reproduced in a photograph by Charles Waite titled
‘Washing the Tailings, Guanajuato’, 1907, number 16 in the series Tema
y Tecnología (CIG-AGN). .................................................................. 223
Figure 3-8. Barba’s illustration of the apparatus used at the end of the sixteenth
century to recycle mercury from the amalgam. A is the iron vessel where
the amalgam is placed, B is the caperuza (clay or metal) that fits on top
of A, with a nozzle (C) that ends below the surface of the water placed
in tank E. The whole assembly sits on the ring D on top of the brick
furnace. Figure from Barba, reproduced from footnote 444. ............. 226
Figure 3-9. a) Cross-section of capellina used by nineteenth century in Mexico. The
stack of amalgam disks (A) is covered by the capellina. A temporary
wall made of bricks (B) surrounds the capellina, and embers are placed
in the space between the two. Mercury condenses in the water basin
below the capellina assembly (V) and is collected via the main water
channel (C). Adapted from illustration in Laur, footnote 452 b) indirect
heating of capellina and use of pulleys to manoeuvre capellina into
place, original lithographs by Tillmann, as reproduced in Macdonald,
footnote 452. ....................................................................................... 229
Figure 3-10. a) schematic cross section of a capellina building in Guanajuato, original
illustration from footnote 455 with annotations added b) and c) capellina
building, option 2, within the historic Hacienda La Escalera,
Guanajuato, at present part of a private dwelling, used as a storage shed
d) capellina building, option 2, renovated by the Morrill family and
converted into a studio, previously the Hacienda Bustos, Guanajuato e)

11
reconstruction of a capellina building, option 1, at a museum on mining
on the grounds of the Real de Minas hotel, Guanajuato. The museum was
not functioning as of February 2013. ................................................. 230
Figure 3-11. Simple recycling assembly of mercury from amalgam, based on inverted
clay water bottles. Original illustration from footnote 456, with labels
added................................................................................................... 231
Figure 3-12. Perimeter walls a) Hacienda La Escalera, Guanajuato, scale bar 1.6 m b)
Hacienda San Juan Nepomuceno, Marfil, scale bar 1.6 m c) Hacienda
Santa Ana, Marfil, scale bar 1.6 m d) Hacienda Las Mercedes,
Zacatecas, scale bar 1.7 m e) Hacienda, name unknown, Pacumo,
Zacatecas, scale bars 1.7 m................................................................. 237
Figure 3-13. Monthly silver production for four amalgamation haciendas in
Guanajuato, calculated from data in footnote 477.............................. 240
Figure 3-14. Plan of the Hacienda Casas Blancas, Marfil, 1885, AHUG, Mapoteca,
Hda. Casas Blancas, 5p3. Digital image supplied by AHUG. ........... 241
Figure 3-15. Main process areas identified in Figure 3-14. .................................... 242
Figure 3-16. Historical plan of the Hacienda de Rocha, Guanajuato, reproduction in
MacDonald, footnote 481, from an original lithograph in a report by E.
Tillmann of 1866. ............................................................................... 243
Figure 3-17. Main process areas of the hacienda de Rocha, according to Figure 3-16.
............................................................................................................ 244
Figure 3-18. Digital image of the original hand drawn plan, ink on paper, of the
Hacienda Las Mercedes, AHEZ, Fondo Mapas e Ilustraciones, Serie V:
Planos Siglos XVIII al XX, number 16, 3 July 1850. ...................... 245
12
Figure 3-19. Schematic plan of the main process-related areas of the Hacienda Las
Mercedes, adapted from Figure 3-18.................................................. 246
Figure 3-20. Plan of the Hacienda de Proaños, Fresnillo, Zacatecas, annotations
added to the digital image supplied by MMOB, Colección General,
Estado de Zacatecas, Varilla CGZAC03, Numero de control 12780-
CGE-7241. No scale is supplied in the drawing................................. 248
Figure 3-21. A perspective of the Hacienda Nueva de Fresnillo, annotations added to
the digital image supplied by MMOB, Colección General, Estado de
Zacatecas, Varilla CGZAC03, Numero de control 12780-CGE-7241.
249
Figure 3-22. Simplified mass balance of amalgamation process ............................ 257
Figure 3-23. Sensitivity of Hg/Ag weight ratio to the fraction of silver chloride and
sulphide of the total silver present in the ore (fa) and on the fraction of
physical loss (fb). Reproduced from footnote 495. ............................ 260
Figure 3-24. Sensitivity of the Hg/Ag weight ratio to fb values, for the range of fa
values expected from ores refined by amalgamation. Reproduced from
footnote 495. ....................................................................................... 261
Figure 3-25. The Observatorio Metereológico of Guanajuato (Conagua and
Universidad de Guanajuato), sited on a landfill of waste from refining
haciendas. Photo taken from the grounds of the historic Hacienda de
Rocha (at present the Hotel Reales de Minas).................................... 272
Figure 3-26. Main loss vectors of calomel and mercury ......................................... 275
Figure 4-1. Production of silver in metric tons, from 1493 to 1900, adapted from
footnote 543. ....................................................................................... 283

13
Figure 4-2. The Hacienda de Regla in relation to Real del Monte, Pachuca and Ciudad
de México. Map adapted from footnote 551. ..................................... 289
Figure 4-3. The locations of three of the historical silver refining haciendas operated
by the Compañia Real del Monte in the second half of the nineteenth
century, on the outskirts of present day Huasca de Ocampo, Hidalgo
State, Mexico. Adapted from Google Earth © 2013 DigitalGlobe,
reservoir location 20° 13’ 36” N 98° 33’ 46” W. ............................... 292
Figure 4-4. Section of the lithograph that shows the location of the three refining
haciendas (San Miguel, San Antonio and Regla) with respect to the Ojo
de Agua that provided the guaranteed water supply for their processes.
Reproduced and adapted from footnote 566. ..................................... 294
Figure 4-5. Satellite image of Regla. The lake is the most evident of modern
additions. The photograph is aligned on a South to North axis from left
to right. Google Earth © 2013 DigitalGlobe, 20° 14’ 15” N 98° 33’ 42”
W. ....................................................................................................... 296
Figure 4-6. Assignment of general process areas of Regla. Satellite image from
Google Earth © 2013 DigitalGlobe. ................................................... 296
Figure 4-7. Reconstruction of main functional areas at Regla (see text for details).
............................................................................................................ 297
Figure 4-8. ‘Entrada a los trapiches de la hacienda Santa Maria de Regla’, by Johan
Moritz Rugendas, oil on cardboard, 24 August 1832. Image reproduced
with permission from the web portal of the Universidad de las Artes,
Aguascalientes, Mexico: http://www.aguascalientes.gob.mx/
temas/cultura/ webua/catalogo/ johanmoritz.html (accessed 9 May
2013). .................................................................................................. 299

14
Figure 4-9. View of the sixteen circular vats of the arrastres. The remains of two
molino tracks are hidden by the cactus leaves on the lower right-hand
corner. ................................................................................................. 301
Figure 4-10. Photograph of water-driven arrastre, from unnamed hacienda in
Mexico, photograph reproduced from footnote 579........................... 302
Figure 4-11. Water distribution channels and overflow outlet in the southern area of
the Hacienda. The four pillars on the grassy area may be the remnants
of the structure that sustained a roof over the patio area. ................... 303
Figure 4-12. The pool at the foot of ‘El Salto’. ...................................................... 304
Figure 4-13. ‘Patio de la Hacienda de Regla’ by Eugenio Landesia, 1857, oil on
canvas, 46 by 64 cm. Image reproduced from footnote 588. ............. 307
Figure 4-14. The Hacienda de Regla, from an unidentified illustration in footnote
590. ..................................................................................................... 307
Figure 4-15. Proposed assignment of dry process areas related to the amalgamation
process in the southern part of Regla. ................................................ 309
Figure 4-16. Tentative location for a capellina, showing the channel for cooling
water. This picture was taken in March 2013. Further excavation around
this location has been carried out by November 2014. ...................... 310
Figure 4-17. a) Drawing with the engineering specifications for the construction of
iron capellinas in the workshop of the Compañia de Real del Monte (date
unknown). The left-hand capellina has a height of ‘5 pies 6 pulgadas
ynglesas’, 5 feet 6 english inches, an internal diameter of 21 inches and
an external diameter of 24 inches. The capellina on the right has the same
width but a height of 4 feet and 6 ½ inches. Digital image of undated
drawing courtesy of AHCRMyP, Sección: Administración Interna,

15
Serie: Departamento de Ingenieros, Subserie: Croquis y Planos, Vol 204
Carpeta 1. b) Photo of capellina in the entrance to the Hacienda Santa
María de Regla, photo courtesy of Mr. Josue Soto Samperio, November
2014. Original height impossible to ascertain since capellina has been
embedded, but at least over 1 m. ........................................................ 311
Figure 4-18. Stone water channels (just to the right of the pyramidal chimney stack)
on the roof of the Furnace Area A corridor, that distribute water supplied
by the external aqueduct for the condensation of mercury. The shrubs in
the foreground hide the connection to the furnace topped by the
rectangular chimney stack. ................................................................. 313
Figure 4-19. a) Ordenes de Christo b) Nuestra Señora de los Dolores, inscriptions
over arches where hornos castellanos are deemed to have been located
to the left of c) Furnace Area A corridor, to the right of which d) the
vaults of the proposed storage areas for salt and copper sulphate. Photos
4-19 a) and b) courtesy Mr. Josue Soto Samperio.............................. 314
Figure 4-20. Extant chimney stacks, date unknown, but similar in location and shape
to two of the five depicted in Figure 4-14. ......................................... 315
Figure 4-21. Plan of Furnace Area B, floor level including courtyards. ................. 318
Figure 4-22. Plan of loading floor of Furnace Area B, at present without a roof. .. 319
Figure 4-23. a) Blast furnaces B2 and B3, h = 4m and w = 6m; distance from hearth
front to front furnace wall: ~2 m, distance to exposed back furnace wall:
~2.7 m. b) trough in front of furnaces, possibly for water. The photo was
taken in March 2013. By November 2014 this trough had been
demolished as part of alterations underway in this area by the hotel c)
furnace stacks, with opening to charge the furnace; the left stack is taller,
16
notebook in lower corner served as scale for photo measurements, strong
sunlight precluded laser measurements d) right hand stack was built in
1854. ................................................................................................... 322
Figure 4-24. a) Detail of Landesia’s painting of the Hacienda de Regla in 1857,
showing the presence of three stacks in Furnace Area B. The single
chimney in the foreground is assigned to reverberatory furnaces b)
probable location of same chimney stack in modern Regla. .............. 324
Figure 4-25. a) Courtyard B-I. This photo was taken in March 2013. By November
2014 this crenelated wall no longer exists, having been half demolished
in alterations being carried out by the hotel b) Original view in March
2013 of Courtyard B-II. By November 2014 the area has been converted
to a set of pools and toilets c) View of courtyard B-II from vaulted
furnace area. Photo taken in March 2013. In November 2014 this arch
was being filled in using stones from the demolished interior walls . 325
Figure 4-26. Monthly deliveries of raw ore to Regla. Data adapted from Informe
Mensual. ............................................................................................. 329
Figure 4-27. Average monthly values and standard deviation of ground silver ore
destined for amalgamation, in the period mid 1872 to mid 1888. Data
calculated from Table 4-I. .................................................................. 331
Figure 4-28. The monthly amounts of silver ore ground for and processed by
amalgamation. Data from Table 4-I and the Informe Mensual. ......... 333
Figure 4-29. Histogram of the days required for amalgamation, as recorded over the
period October 1872 to December 1873. The average amalgamation run
lasted 13 days. Raw data from Informe Mensual. .............................. 334

17
Figure 4-30. Scatter graph and plot of averages as measured at various amalgamation
periods that show an evident correlation between the percentage of silver
extracted and the number of days for an amalgamation run. Raw data
from Informe Mensual. ....................................................................... 335
Figure 4-31. Histogram of the number of cargas per torta as practised at Regla (1872
to 1888). Source data from Informe Mensual. .................................... 337
Figure 4-32. Histogram of ore quality refined by amalgamation. Raw data from
Informe Mensual. ................................................................................ 339
Figure 4-33. Decrease in percentage of refined silver from the ores processed by
amalgamation. Percentage calculated from raw data in Informe Mensual.
340
Figure 4-34. Consumption of salt per kg of silver refined by amalgamation (1872-
1888). Raw data from Informe Mensual............................................. 343
Figure 4-35. Consumption of copper sulphate per kg of silver refined by
amalgamation (1872-1888). Raw data from Informe Mensual. ......... 344
Figure 4-36. Monthly production in kg of silver by amalgamation (1872-1888). Raw
data from Informe Mensual. ............................................................... 347
Figure 4-37. Losses of silver, expressed as percentage, during amalgamation (1872-
1888), calculated from raw data in the Informe Mensual. All negative
values reported in the ledger have been excluded. ............................. 348
Figure 4-38. Weight ratio of mercury consumed to silver refined by amalgamation
(1872-1888). Raw data from Informe Mensual. ................................. 350
Figure 4-39. Monthly deliveries of ore for smelting to Regla (1875-1886). Raw data
from the Informe Mensual. ................................................................. 356

18
Figure 4-40. Monthly values (1875-1886) of silver ore ground prior to smelting, and
the quantities of ore smelted. Raw data from the Informe Mensual. .. 357
Figure 4-41. Histogram of the silver content of ores destined for smelting at Regla.
Raw data from the Informe Mensual. ................................................. 357
Figure 4-42. Monthly production of silver by smelting (1875-1886). Raw data from
the Informe Mensual. .......................................................................... 359
Figure 4-43. Comparison of silver losses incurred during smelting and amalgamation
(1872-1888). Raw data from the Informe Mensual. ........................... 360
Figure 4-44. Weight of litharge lost per 1 kg of silver smelted (1875-1886). Raw data
from the Informe Mensual. ................................................................. 360
Figure 4-45. Weight of charcoal required to produce on average 1 kg of silver by
smelting (1875-1876). Raw data from the Informe Mensual. ............ 365
Figure 4-46. Main mass transit corridors at Regla, average monthly quantities in the
period 1872/73 and 1875/88 (amalgamation) and Jun 1875 to Jan 1886
(smelting). ........................................................................................... 366
Figure 4-47. Main loss vectors of waste material, monthly average at Regla in the
period 1872/73 and 1875/88 (amalgamation) and Jun 1875 to Jan 1886
(smelting). ........................................................................................... 367
Figure 4-48. The Valley of the Metztitlan River, Hidalgo State, Mexico. Satellite
image from Google Earth Image Landsat. The Lago de Metztitlan is
situated at 20° 41’ 04” N 98° 51’ 01” W. The depth of the valley floor
relative to Regla is best appreciated in Figure 4-4. ............................ 368
Figure 5-1. Production cost and price of mercury in New Spain, plotted from data in
footnote 693. ....................................................................................... 381

19
Figure 5-2. The evolution of the gold to silver ratio from 1690 to 1900. Data from
footnote 781. ....................................................................................... 418
Figure 5-3. The evolution of the price of silver in the London market. Data from
footnote 781. ....................................................................................... 419
Figure 5-4. Time series for the expenditure on maize as fodder for animals at Regla
(1872-1888). The source data used to calculate the values of monthly
unit costs of maize in pesos per carga are from Contabilidad Mensual.
............................................................................................................ 423
Figure 5-5. Monthly expenses of salt consumed at Regla (1872-1888), in pesos per
arroba. Values calculated from data in Contabilidad Mensual. ........ 423
Figure 5-6. Yearly average expense on salt, in pesos per arroba (1853-1888). Values
calculated from data in Contabilidad Mensual correspond to salt prices
for Regla. Values calculated from data in Estados Comparativos
correspond to the average price of salt registered for all the haciendas of
the Compañia Real del Monte carrying out amalgamation in any given
year. .................................................................................................... 424
Figure 5-7. Monthly expenses of copper sulphate consumed at Regla, in pesos per
pound. Values calculated from data in Contabilidad Mensual. ......... 425
Figure 5-8. Monthly expenses of mercury consumed at Regla, in pesos per pound.
Values calculated from data in Contabilidad Mensual. ..................... 426
Figure 5-9. Yearly average expense on mercury, in pesos per pound. Values
calculated from data in Contabilidad Mensual correspond to mercury
cost at Regla. Values calculated from data in Estados Comparativos
correspond to the average cost of mercury registered for all the

20
haciendas of the Compañia Real del Monte carrying out amalgamation
in any given year. ............................................................................... 427
Figure 5-10. Nineteenth century mercury prices from footnote 799. The scale on the
right applies to US and London prices, the scale on the left to prices from
Italy. .................................................................................................... 429
Figure 5-11. Monthly expenses on charcoal for smelting consumed at Regla (1875-
1886), in pesos per carga. Values calculated from data in Contabilidad
Mensual. ............................................................................................. 431
Figure 5-12. Monthly expenses on litharge consumed at Regla (1875-1886), in pesos
per kg. Values calculated from data in Contabilidad Mensual. ......... 432
Figure 5-13. Monthly production costs of silver refined by amalgamation at Regla
(1872-1888). Values calculated from data in Contabilidad Mensual.433
Figure 5-14. Percentage breakdown of the amalgamation cost of silver at Regla in the
period 1872 to 1888, excluding the cost of silver ore at the plant gate.
436
Figure 5-15. Percentage breakdown of the total variable cost of production by
amalgamation at Regla, based on the production cost of ore from the
mines of Real del Monte in 1853-1854. Other data as in Figure 5-14.
439
Figure 5-16. Cross-over point between the average variable cost of processing one
montón by patio amalgamation at Regla, and a deemed maximum value
for the silver extracted from the ore. Data sourced from Table 5-VIII.
443
21
Figure 5-17. The projected cost of production using patio amalgamation at Regla
(1875-1888) as a function of the gross silver content of the ore. Data
from Table 5-VIII. .............................................................................. 445
Figure 5-18. Monthly production costs of silver refined by smelting at Regla (1875-
1886). Values calculated from data in Contabilidad Mensual. .......... 446
Figure 5-19. Percentage breakdown of variable smelting costs at Regla (1875-1886),
(a) without and (b) with the deemed variable cost of the ore. ............ 448
Figure 5-20. Cross-over point between the average variable cost of processing one
carga by smelting and the value of silver extracted from the ore. Data
from Table 5-X. .................................................................................. 450
Figure 5-21. The variable production cost by smelting at Regla as a function of the
silver content of the ore. Data from Table 5-X. ................................. 450
Figure 5-22. Total variable refining cost at Regla (1875-1888), as a function of the
silver content of the ore. Data from Tables 5-VIII and 5-X. .............. 456
Figure 5-23. Comparative production costs of amalgamation and smelting in the
context of the second half of the sixteenth century as a function of the
silver content of the ore. See Appendix E for source data. ................ 458
Figure 5-24. Total production cost as a function of the silver content of the ore, within
the context of the period from the mid-seventeenth to mid-eighteenth
century. See text for details on the generation of the values plotted. . 462
Figure 5-25. Annual average of cargas of silver ore processed at the refining
haciendas of the Compañia Real del Monte in the period 1853 to 1873.
Source data from Estados Comparativos. .......................................... 475
Figure 6-1. Map of regional Cajas of main mining districts in New Spain, snapshot
of the 1760s, adapted from original map in footnote 874. ................. 491
22
Figure 6-2. The fraction of total silver refined by amalgamation in Zacatecas in the
period 1670 to 1820. Prior to 1700 the time intervals are exactly one year
only for the Lacueva data. Fractions calculated by the author on the basis
of the raw data from TK set and footnote 894.................................... 496
Figure 6-3. Registry of silver at the Caja of Zacatecas according to refining process.
Data from Table 6-II. .......................................................................... 497
Figure 6-4. The fraction of total silver refined by amalgamation as registered in the
Caja of Guanajuato in the period 1679 to 1816. Prior to 1720 the time
intervals of the raw data in the TK data have been approximated to the
calendar years. .................................................................................... 502
Figure 6-5. Registry of silver at the Caja of Guanajuato according to refining process.
Data from Table 6-V. ......................................................................... 502
Caja of México in the period 1786 to 1816. The raw data are from the
TK data set. ......................................................................................... 506
Caja of Durango in the period 1696 to 1813. Between 1737 and 1765 no
distinction was made between amalgamated and smelted silver in the tax
register. Prior to 1713 I have approximated the irregular time series of
the raw data in the TK data to their nearest calendar years. ............... 508
Figure 6-8. Registry of silver at the Caja of Durango according to refining process.
Data from Table 6-XI. ........................................................................ 509
Caja of San Luis Potosí in the period 1713 to 1806. Source of raw data
is the TK set. ....................................................................................... 512

23
Figure 6-10. Registry of silver at the Caja of San Luis Potosí according to refining
process. Data from Table 6-XIV. ....................................................... 513
Caja of Guadalajara in the period 1691 to 1804. Prior to 1699 I have
approximated the irregular time series of the raw data in the TK set to
their nearest calendar years. ............................................................... 517
Figure 6-12. Registry of silver at the Caja of Guadalajara according to refining
process. Data from Table 6-XVII. ...................................................... 517
Caja of Pachuca in the period 1667 to 1820. Prior to 1706 I have
Figure 6-14. Silver registered at the Caja of Pachuca. Data from Table 6-XX. ..... 522
Caja of Sombrerete in the period 1680 to 1820. Prior to 1760 I have
Figure 6-16. Silver registered at the Caja of Sombrerete. Data from Table 6-XXIII.
............................................................................................................ 525
Caja of Bolaños in the period 1753 to 1804. Raw data from TK set. 529
Caja of Rosario in the period 1770 to 1813. Raw data from TK set. 531
24
Caja of Zimapán in the period 1729 to 1806. Raw data from TK set.
533
Caja of Chihuahua in the period 1788 to 1813. Raw data from TK set.
536
Figure 6-21. Registry of silver by process, as projected for New Spain. ................ 540
Figure 6-22. Approximate geographical distribution of main refining processes
applied in the Cajas of New Spain. Black: smelting, Grey:
amalgamation. Adapted from Figure 6.1. ........................................... 543
Figure 6-23. Listing of Cajas by the magnitude of the vector corresponding to lead
and lead compounds. Data from table 6-XXXVIII ............................ 544
Figure 6-24. Woodland consumed by smelting and amalgamation according to Caja.
Data from Table 6- XXXVIII. ............................................................ 545
Figure 6-25. Ranking of Cajas according to the magnitude of the environmental
impact vector of mineral waste voided into waterways. Data from Table
6-XXXVIII. ........................................................................................ 545
impact vector of salt voided into waterways. Data from Table 6-
XXXVIII............................................................................................. 546
impact vector of copper sulphate consumed and voided into waterways.
Data from Table 6-XXXVIII. ............................................................. 546

25
impact vector of calomel voided into waterways. Data from Table 6-
XXXVIII............................................................................................. 547
Figure 6-29. Amalgamation and smelting fraction of silver presented at the Mexican
mints, 1876 to 1892. Raw data from footnote 900. ............................ 548
Figure C-1. Inventory of raw ore destined for amalgamation. Raw data from Informe
Mensual. ............................................................................................. 596
Figure C-2. Inventory levels of ground silver ore destined for amalgamation. Data
Figure C-3. Monthly inventory levels of salt. Raw data from Informe Mensual. ... 597
Figure C-4. Monthly inventory levels of copper sulphate. Raw data from Informe
Mensual……. ..................................................................................... 598
Figure C-5. Monthly inventory levels of mercury. Raw data from Informe
Mensual…… ...................................................................................... 599
Figure C-6. Inventory levels of raw silver ore destined for smelting. Raw data from
the Informe Mensual Regla. ............................................................... 599
Figure C-7. Inventory of ground silver ore ready for smelting. Raw data from the
Informe Mensual. ................................................................................ 600
Figure C-8. Inventory of litharge. Raw data from the Informe Mensual. .............. 601
Figure C-9. Inventory of charcoal for smelting. Raw data from the Informe
Mensual…… ....................................................................................... 602

26
List of Tables
Table 1-I. Major silver deposits ranked by aggregate production. For sources see
footnote 39. ........................................................................................... 65
Table 1-II. Silver content of ores reported for New Spain / Mexico. Sources are listed
in footnote 113. ..................................................................................... 94
Table 1-III. Silver content of ores in the Vice-Royalty of Peru. Sources are listed in
footnote 113. ......................................................................................... 95
Table 1-IV. Silver content of ores in Europe. Sources are listed in footnote 114..... 96
Table 1-V. Major deposits of mercury through history. Sources indicated in footnote
116. ....................................................................................................... 98
Table 2-I. Published weight ratios of lead to silver used in the smelting of silver ores.
Sources are given in footnote 248. ..................................................... 148
Table 2-II. Range of lead to silver weight ratios from individual smelting runs carried
out in the region of Veta Grande, Zacatecas, in 1718. The source data
are from footnote 250. ........................................................................ 150
Table 2-III. Range of percentage values for lead losses during smelting of lead ores.
Sources in footnote 251. ..................................................................... 151
Table 2-IV. Lead content in slags from different smelting sites and periods. Sources
in footnote 255. ................................................................................... 153
Table 2-V. Lead, arsenic and sulphur content of two samples of grasas from Monte
Caldera. ............................................................................................... 154
Table 2-VI. Assumptions applied to the mineral and lead mass balance for the
Hacienda Santa Maria, Monte Caldera. For details and sources see text.
............................................................................................................ 155
27
Table 3-I. List of debtors on supplies provided by the Hacienda La Escalera,
Guanajuato, with their estimate of selected assets offered as collateral to
the debt, as of 1791. Raw data from footnote 473. Numbers in italic have
been calculated by author from original data. .................................... 239
Table 3-II. A comparison of some of the main spatial and operational features of
amalgamation haciendas in Mexico, nineteenth century. Data in italics
have been estimated, other data are derived from plans or from sources
in text. ................................................................................................. 250
Table 3-III. Historical values of Hg/Ag weight ratios calculated from values of
correspondencia reported in the historiography. Sources from footnote
498. ..................................................................................................... 256
Table 3-IV. Percentage breakdown of mercury losses to the environment as a result
of the refining of silver ores with mercury. ........................................ 262
Table 4-I. Monthly amounts, in cargas, of ore ground for amalgamation. Raw data
Table 4-II. Partial transcription of data that appear for the five weeks ending on
November 28th, 1885, as registered in the Informe Mensual. Numbers in
italics represent numbers shown in red in original document. The units
of silver content are registered in marks and hundredths of a mark. .. 341
Table 4-III. Additional consumables at Regla, from Memorias de Gastos for the four
weeks ending on May 26th 1877. ....................................................... 351
Table 4-IV. Overall mass balance for the amalgamation of silver ores as practised at
Regla between 1872 and 1888. Data compounded from different sections
of this chapter. The numbers in italic denotes a calculated number, not
directly derived from the monthly accounting ledger. ....................... 353

28
Table 4-V. Total consumption of firewood and charcoal registered on a weekly basis
at Regla, according to Memorias No. 19 - 21 de los Gastos de la Hcda
de Regla, for the four weeks ending on May 26th 1877. Data in italics
correspond to inventory make-up. ...................................................... 354
Table 4-VI. Overall mass balance for the smelting of silver ores as practised at Regla
between June 1875 and January 1886. Data was compounded from the
sections on smelting of this chapter. The numbers in italic were
calculated and not obtained directly from the accounting data. ......... 363
Table 5-I. Interpretation of Villaseñor’s working examples and method that sustained
his argument against decreasing the price of mercury. Data adapted from
footnote 730. ....................................................................................... 398
Table 5-II. Summary of amalgamation and smelting costs in New Spain/Mexico from
the historiography up to the nineteenth century. Figures in italics are
calculated from source data. Sources are indicated in footnote 737. . 402
Table 5-III. Production costs as reported in 1802 for the cazo, patio and smelting
refining processes carried out at Catorce (San Luis Potosí), adapted from
data in footnote 743 and Appendix D................................................. 405
Table 5-IV. Generic profile of patio amalgamation costs (excluding ore cost) in
Mexico as practised around 1840, expressed in terms of g of fine silver.
Specific export costs are highlighted in bold and italics. Data from
footnote 749. ....................................................................................... 408
Table 5-V. The percentage breakdown of the main variable amalgamation costs at
Regla, excluding the cost of ore at the plant gate. The percentage values
were calculated from the individual headings within the monthly account
29
data, and then averaged for the year. A total of 153 data sets are
represented in the table. Source data from Contabilidad Mensual. ... 435
Table 5-VI. The percentage contribution to the total variable amalgamation refining
cost of the total fuel required by the amalgamation process. Source data
from Estados Comparativos. .............................................................. 435
Table 5-VII. Mining and other costs for Real del Monte mines in the period 1849 to
1854, raw data adapted from footnote 809. ........................................ 437
Table 5-VIII. Matrix to determine the variation of total production cost by patio
amalgamation at Regla as a function of silver content in the ore. All data
derived from the accounting books of Regla, except for the cost of ore
which has been derived from Section 5.6. .......................................... 443
Table 5-IX. The percentage contribution to the partial variable refining cost of the
main cost elements of the process. The percentage values were
calculated on a monthly basis, and then averaged for the year. A total of
103 data sets are represented in the table. Source data from Contabilidad
Mensual. ............................................................................................. 446
Table 5-X. The derivation of the total cost of smelting per carga and of the variable
production cost per kg of silver as a function of silver content in the ore.
For source of data see text. ................................................................. 449
Table 5-XI. Selection of historical costs from the sixteenth to nineteenth century of
the main factors that determine the production cost of amalgamation and
smelting. Sources in footnote 830. ..................................................... 455
Table 5-XII. The profile of costs registered at Regla in the third quarter of the
nineteenth century within the historical context of New Spain / Mexico.
For sources see Table 5-XI and Sections 5.8 and 5.11. ...................... 456
30
Table 5-XIII. Sensitivity values for a cost approximation to the context of refining of
silver in the second half of the sixteenth century. .............................. 458
Table 5-XIV. A theoretical context of production costs viable for the period 17c to 18c,
and within the limits of data provided in Table 5-XII. ....................... 461
Table 5-XV. Breakdown of labour man-days and costs for the various refining stages
carried out at at Regla based on data for the four week period ending on
May 29th 1877 (Memorias Numero 18 - 22)....................................... 465
Table 5-XVI. Labour costs at Regla, according to the Memorias 18 – 22, and the
deemed distribution of labour costs between patio amalgamation and
smelting. ............................................................................................. 468
Table 5-XVII. Labour man-days at Regla, based on data from the Memorias 18 – 22,
and the deemed distribution of man-days (manual and non-manual
labour) between patio amalgamation and smelting. ........................... 469
Table 5-XVIII. Labour force at Regla, based on the Memorias 18 – 22, and the
deemed distribution of the workforce (manual and non-manual labour)
between patio amalgamation and smelting. ....................................... 469
Table 5-XIX. Variable production costs in pesos per kg of refined silver for the various
refining haciendas of the Compañia Real del Monte, in the period 1853
to 1873. The gaps in grey indicate the haciendas were not in use at the
time. The gaps in white indicate a lack of primary sources for the period.
Table 5-XX. Silver content of ore before processing. Raw data from Estados
Comparativos. .................................................................................... 477
Table 5-XXI. Losses of silver, expressed as weight percentage, registered at the
refining haciendas of the Compania Real del Monte, in the period 1853
31
to 1873. The gaps in grey indicate the haciendas were not in use at the
time. The gaps in white indicate a lack of primary sources for the period.
Table 5-XXII. Mercury to silver weight ratio registered at the refining haciendas of the
Compania Real del Monte, in the period 1853 to 1873. The gaps in grey
indicate the haciendas were not in use at the time. The gaps in white
indicate a lack of primary sources for the period. Source data from
Estados Comparativos. ....................................................................... 479
Table 5-XXIII. The average amount of pesos required to refine 1 kg of silver using
the two amalgamation processes. The haciendas in italics used the barrel
process, and the haciendas in normal script used the traditional patio
amalgamation. The data has been calculated for the period 1853 to 1873
using as source the Estados Comparativos. ....................................... 480
Table 6-I. Silver production in Mexico, nineteenth century. Sources from footnotes
875 and 876. The data after 1875 corresponds to fiscal years beginning
in the year indicated............................................................................ 492
Table 6-II. Magnitude of the environmental impact vectors for calomel, liquid
mercury and volatile mercury, corresponding to the silver obtained by
amalgamation and registered in the Caja of Zacatecas. In this and
following tables A: Amalgamation, S: Smelting, For the method and
sources see text. .................................................................................. 499
Table 6-III. Projected magnitude of other main environmental impact vectors
corresponding to the silver obtained by amalgamation and registered in
the Caja of Zacatecas. ........................................................................ 500

32
Table 6-IV. Projected magnitude of the main environmental impact vectors
corresponding to the silver obtained by smelting and registered in the
Caja of Zacatecas. .............................................................................. 501
Table 6-V. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Guanajuato. .................. 503
Table 6-VI. Projected magnitude of other main environmental impact vectors
the Caja of Guanajuato. ...................................................................... 504
Table 6-VII. Projected magnitude of the main environmental impact vectors
Caja of Guanajuato............................................................................. 505
Table 6-VIII. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of México. Values in bold from
footnote 894. ....................................................................................... 506
Table 6-IX. Projected magnitude of other main environmental impact vectors
the Caja of México. ............................................................................ 507
Table 6-X. Projected magnitude of the main environmental impact vectors
Caja of México. .................................................................................. 507
Table 6-XI. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Durango. Silver production

33
figure in bold from footnote 896, ceiling for calomel estimates indicated
in bold italic figures. ........................................................................... 510
Table 6-XII. Projected magnitude of other main environmental impact vectors
the Caja of Durango. .......................................................................... 511
Table 6-XIII. Projected magnitude of the main environmental impact vectors
Caja of Durango. ................................................................................ 511
Table 6-XIV. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of San Luis Potosí............. 515
Table 6-XV. Projected magnitude of other main environmental impact vectors
the Caja of San Luis Potosí. ............................................................... 516
Table 6-XVI. Projected magnitude of the main environmental impact vectors
Caja of San Luis Potosí. ..................................................................... 516
Table 6-XVII. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Guadalajara. ................. 519
Table 6-XVIII. Projected magnitude of other main environmental impact vectors
the Caja of Guadalajara. ..................................................................... 520

34
Table 6-XIX. Projected magnitude of the main environmental impact vectors
Caja of Guadalajara. ........................................................................... 520
Table 6-XX. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Pachuca. ....................... 523
Table 6-XXI. Projected magnitude of other main environmental impact vectors
the Caja of Pachuca. ........................................................................... 523
Table 6-XXII. Projected magnitude of the main environmental impact vectors
Caja of Pachuca. ................................................................................. 524
Table 6-XXIII. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Sombrerete. .................. 527
Table 6-XXIV. Projected magnitude of other main environmental impact vectors
the Caja of Sombrerete. ...................................................................... 528
Table 6-XXV. Projected magnitude of the main environmental impact vectors
Caja of Sombrerete............................................................................. 528
Table 6-XXVI. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Bolaños. ........................ 529

35
Table 6-XXVII. Projected magnitude of other main environmental impact vectors
the Caja of Bolaños. ........................................................................... 530
Table 6-XXVIII. Projected magnitude of the main environmental impact vectors
Caja of Bolaños. ................................................................................. 530
Table 6-XXIX. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Rosario. ........................ 531
Table 6-XXX. Projected magnitude of other main environmental impact vectors
the Caja of Rosario. ............................................................................ 532
Table 6-XXXI. Projected magnitude of the main environmental impact vectors
Caja of Bolaños. ................................................................................. 532
Table 6-XXXII. Production of silver by smelting as registered in the Caja of
Zimapán. ............................................................................................. 534
Table 6-XXXIII. Projected magnitude of the main environmental impact vectors
Caja of Zimapán. ................................................................................ 534
Table 6-XXXIV. Magnitude of the environmental impact vectors for calomel, liquid
amalgamation and registered in the Caja of Chihuaha… .................. 536

36
Table 6-XXXV. Projected magnitude of other main environmental impact vectors
the Caja of Chihuahua. ....................................................................... 537
Table 6-XXXVI. Projected magnitude of the main environmental impact vectors
Caja of Chihuahua. ............................................................................. 537
Table 6-XXXVII. Summary of main magnitudes projected for each of the main
mining Cajas of New Spain. Woodland figures expressed in units of a
thousand ha, all the others in units of a thousand t. ............................ 538
Table 6-XXXVIII. Amalgamation and smelting by Caja over the whole colonial
period in New Spain. Source data from Table 6-XXXVII. ................ 542
Table 6-XXXIX. Projected magnitudes of main environmental impact vectors for
amalgamation and smelting in Mexico, 1821 to 1899. ...................... 549
Table 6-XL. Total magnitude of environmental impact vectors from amalgamation and
smelting as projected for New Spain and Mexico. All numbers have been
rounded off, and expressed in thousand t except for woodland which is
in thousand ha. .................................................................................... 550
Table A-I. Account book prepared by Lopez de la Madriz, Valle de Pozos, AHSLP,
Fondo Alcaldía Mayor 1650.3, expediente 8. ................................... 585
Table A-II. Weekly accounts of the Hacienda de Nuestra Señora de los Dolores,
corresponding to the year 1773, signed by Lorenzo Mata. AHSLP,
Fondo Alcaldía Mayor 1773.2. ........................................................... 586
Table B-I. Assignment of account headings in Contabilidad Mensual into subsets
(labour, mercury, salt, copper sulphate and other costs) used in the
analysis of production costs at Regla. ................................................ 593

37
Table B-II. Comparison of amounts accounted for in the Contabilidad Mensual and
Estados Comparativos for the year 1873 for production costs at Regla.
See text for analysis. ........................................................................... 595
Table C-I. Areas required by the average inventory of the main reagents, fuel and ore,
as calculated from the raw data in the Informe Mensual Regla. ........ 603
Table E-I. Amalgamation sixteenth century context ............................................... 609
Table E-II. Smelting sixteenth century context ....................................................... 609
Table E-III. Amalgamation seventeenth and eighteenth century context ................ 610
Table E-IV. Smelting seventeenth and eighteenth century context. ....................... 610
38
Abstract
The environmental history of silver refining in New Spain is the aggregate result of two
refining processes, amalgamation and smelting, that emit two completely different sets of
chemicals and impose two distinct levels of woodland depletion upon the environment. Over
60% of its silver was produced by amalgamation, a physical term that hides the complex and
concatenated chemical reactions that transform silver sulphides into amalgamated silver and
calomel. The chemical ratios and the historical levels of the correspondencia (mercury to silver
weight ratio) are shown to mathematically restrict the possible level of physical losses of
mercury during amalgamation to less than 15% on average, with mercury in calomel
constituting the balance. Just under 40% of silver was refined by smelting in the presence of
lead, with high emissions of lead fume and high energy requirements. Waterways would be the
waste disposal channel for amalgamation, the air for lead fumes, and woodland depleted for
smelting. A mass balance analysis is applied to each process to arrive at ratios of by-product
emissions and energy input per kilogram of silver extracted. The raw data is derived both from
historical sources and also from the nineteenth century accounts of the Hacienda de Regla, one
of the major silver refining centres in nineteenth century Mexico. The choice of refining
method in New Spain was determined by the chemical nature of the silver ores. The analysis
of production costs at Regla and their projection back to the sixteenth century indicate that
changes in the historical context could alter substantially the relative economies of both. The
strategic decision by Spain to favour amalgamation over smelting was influenced by the fiscal
importance of its mercury revenues. The main conclusion of this thesis is that a paradigm based
on lead and calomel determined the material impact of the environmental history of silver
refining in the New World.

39
Résumé
L’histoire environnementale du raffinage de l’argent à la Nouvelle Espagne est le
résultat de l’agrégat de deux procédés, l’amalgamation et par la fonte, qui dégagent deux
ensembles complètement différents de produits chimiques et produisent deux niveaux de
réduction des bois. Presque 60% de l’argent a été produit par l’amalgamation, un terme
physique qui cache les réactions chimiques complexes qui transformaient les sulfures d’argent
en argent amalgamé et calomel. Les rapports chimiques et les niveaux historiques de la
correspondencia réduisent mathématiquement le niveau possible des pertes physiques du
mercure à moins du 15%, et calomel compris le 85% restante. À peu près de l’autre 40% de
l’argent était produit par la fonte avec du plomb, avec grands émissions de la fumée au plomb
et un très grand besoin d’énergie. Les cours d’eau seraient les voies pour jeter les déchets de
l’amalgamation, l’atmosphère pour les fumées au plomb, et les bois épuisés par la fonte. Une
analyse du bilan de masse de chaque procédé donne les rapports des émissions et de l’énergie
requis par chaque kilogramme de l’argent produit. Les chiffres pour le calcul sont retrouvée
des sources historiques et aussi des cahiers de comptabilité du dix-neuvième siècle provenant
des opérations à la Hacienda de Regla, une des usines de raffinage de l’argent plus important
au Mexique. Le choix du procédé de raffinage à la Nouvelle Espagne était déterminé para la
nature chimique du minerai. L’étude des couts de la production à Regla et leur projections
jusqu’au seizième siècle indique que les changes du contexte historique pouvaient changer le
bilan entre les deux. La décision stratégique de l’Espagne qui a favorisé l’amalgamation sur le
procédé para la fonte a été influencée par des raisons fiscales à cause des revenues de la vente
du mercure. La conclusion principale de cette thèse propose qu’un paradigme fondé sur plomb
et calomel a déterminé l’impact matérielle de l’histoire environnementale du raffinage de
l’argent à la Nouvelle Espagne.

40
Acknowledgements
My gratitude first of all to the History Department of McGill for having taken the risk
of accepting a grandfather in its graduate program. Prof. Gershon Hundert in my first meeting
told me the consequences of that decision were firmly in my hands, and I hope I have not let
the Department down. Without the fellowship from the Cundill Foundation I would not have
been able to come to McGill, again a leap of faith on their part for which I am extremely
grateful. It is not an easy task to supervise and guide a mature student. My warmest thanks to
Prof. Catherine Desbarats who had the patience to manouevre me through the hectic reading
of the first year, and provided support throughout my stay at McGill. Prof. Daviken Studnicki-
Gizbert then accepted the challenge to supervise what turned out to be a multidisciplinary
sourcing of material with strong chemical overtones that had to be marshalled within the
discipline of History. I have tried to learn from his incisive questioning and detailed editing
along the way, and I am grateful for his contributions that have strengthened my thesis. I could
not have hoped for a better external advisor than Prof. Pamela Welbourn of Queen’s University,
who supported my work, sharpened my writing, pointed out glaring mistakes and helped me
over the more lonely stretches. The Faculty of Arts of McGill and the Department of History
have provided financial support for my travels to Mexico and my attendance at conferences,
for which I am most grateful. The staff at McGill Library chased down many a source for me.
I have benefited throughout my work from the generosity of total strangers, though the
final responsibility for any errors in my work is solely mine. By order of chapter: Dr. Richard
H. Sillitoe (U.K) for answering my questions and providing his latest paper on silver deposits;
Dr. Alexandre Desbarats (Natural Resources Canada) who provided very useful comments on
my attempt to lay out a firm foundation of geology in Chapter 1; Prof. Guadalupe Salazar
González, Universidad Autónoma de San Luis Potosí (Mexico) for organizing and providing
41
all the support necessary, including sacrificing a long and dusty Saturday, to arrange my field
trip to visit the ruins of the haciendas in Monte Caldera and providing an invaluable guide to
the interpretation of these ruins; Don Rafael Morales Bocardo, Director of the Archivo
Histórico de San Luis Potosí, and his staff, who welcomed me with the warmth, politeness and
hospitality I would find everywhere in Mexico. He shared information on his own research and
details on the history of San Luis Potosí; Doña Maria Esther Méndez Tobías of the Museo de
Arte Sacro in Guadalcazar, who spent a Sunday after mass to take me to the ruins of the
Hacienda de Aranzazu, and waited patiently in the shade while I walked around and took
photographs and measurements; Dr. Eduardo Manzanares Acuña, of the Unidad de Estudios
Nucleares, Universidad Autónoma de Zacatecas, for discussing with me his research on the
abnormally high levels of lead measured in the blood of children of modern Mexico who live
in houses built over historic mining and refining fills; Prof. Raynald Gauvin and Mr. Nicholas
Brodusch of the Department of Materials Engineering of McGill University for carrying out
pro bono the electron microscopy work on slag samples from Monte Caldera; Dr. David
Johnston (U.K.) for guiding me through his paper and for digging up further information from
his old laboratory notes; Dr. Thomas Hillerkuss of the Universidad Autónoma de Zacatecas,
who drove me to various mining and refining sites around Zacatecas and Pánuco, shared his
library, provided me with documents, set up meetings and overall was a generous guide to the
area; the Director of the Archivo Histórico de Zacatecas, Ing. María Auxilio Maldonado
Moreno, Ms. Genoveva Raquel Andrade Haro and her colleagues; Maestra Maria del Socorro
Cardoso Girón, the State appointed local historian of Pánuco for welcoming me to her house
to talk about the refining haciendas of the area; the Morrill family and Dr. Virgilio Fernández
del Real who opened their homes in historic refining haciendas of Guanajuato and Marfil and
shared their stories; the historian Jaime Medina Martínez who photocopied his thesis on
refining haciendas for my benefit; Licenciada en Historia Eréndira María Guadalupe Guzmán
42
Segoviano, Coordinator of the Historical Section of the Archivo de la Universidad de
Guanajuato, who made it possible for me to review the maximum number of relevant
documents; the historian Ada Marina Lopez Meza, who provided insight on haciendas in
Guanajuato; the present owner of the Hacienda Las Mercedes, who with his wife received me
on a Sunday to show me around the reconstructed hacienda until the evening chill forced us to
retreat; Prof. Carmen Giunta, the Editor of the Bulletin for the History of Chemistry, who took
the risk of publishing my hybrid paper and whose constructive editing enriched it; the Director
of the Museo de Minería y Archivo Histórico de Pachuca, Asociación Civil, Lic. Belem
Oviedo Gámez, and Maestra Aracelys Monroy Pérez, who from the first email strongly
encouraged me to visit their archives in Pachuca, thanks to which the whole course of my thesis
changed. They provided me with all the support, guidance and kindness I could hope for; Prof.
Herbert S. Klein (Columbia University) for generously providing his Excel files with data from
the Cajas of New Spain; Dr. Daniel Engstrom (St. Croix Watershed Research Station) and Dr.
Colin Cooke (Dept. of Environment and Sustainable Resource Development, Alberta) for
sharing their research results, for their words of encouragement and for providing a very wide
audience for part of my work. Their use of the term paradigm emboldened me to apply it to the
title of this thesis.
Finally, to my grown-up family (Cristina and Juan Cristobal, Saul Ignacio and Claire,
Mariana and Juan Carlos, Carlos Pedro and the sky), who have had to wait for the grandfather
to come back to the joy of all our lives, my grandchildren Manuela, Olivia and Leo. As to my
wife Adriana, she played a major role in the transcription of many a data point into my
spreadsheets. I cannot love her more, she has had to endure an unexpected late twist to our life,
yet has never complained nor failed in her support during our many recent wanderings.
43
To Adriana
44
Guide to the text
Chemical symbols
Ag: silver
Cl: chloride
Cu: copper
Hg: mercury
Na: sodium
O: oxygen
Pb: lead
S: sulphur
Units of measure
kg = kilogram
m = metre
m2 = square metre
m3 = cubic metre
Ma = million years
oz = ounce
t = ton (metric)
t/d = tons per day
t/m = tons per month
t/y = tons per year
lb = pound
y = year
Equivalence of units of measure
1 quintal = 46 kg
1 mark = 8 oz = 0.23 kg
1 arroba = 11.5 kg
1 carga = 12 arrobas = 138 kg
45
1 troy oz = 0.031 kg
1 lb = 0.454 kg
Tons (t) are metric tons = 1,000 kg
1 fanega maize = 1/2 carga of cereal = 52 kg
1 mark = 8.5 to 8.75 pesos
1 vara = 0.84 m
Monetary units
All calculations involving monetary units of pesos, reales and tomines have been
rounded off on the basis of pesos only.
Translations
All translations are by the author unless otherwise indicated. Non-English words in the
text and footnotes are inserted in italics, except for proper names and institutions.
Photographs, illustrations and drawings
All plots, tables, photographs and drawings by the author unless otherwise indicated.
Digital images were taken by the author of original prints made public prior to 1920. Digital
copies of other material were provided by the AHSLP, MMOB, AHUG and AHCRMyP.
Satellite images from Google Earth have been reproduced for academic and non-commercial
purposes, as well as single images from texts under ‘fair dealing’. Where applicable,
permission to reproduce is indicated in the captions to the figures.
Geopolitical terms
The geopolitical designations of New Spain and Mexico are used either separately or
together, depending on the historical period being covered. New Spain covered a larger
territory than republican Mexico, including territory in Central America, except for present day
Panama that was assigned to New Granada.
46
Introduction
From the mid sixteenth century to the third quarter of the nineteenth century, Spain
dominated the world production of silver. According to the data collated by TePaske, silver
from the Hispanic New World constituted 40 % (sixteenth century), 74% (seventeenth century)
and 71% (eighteenth century) of world production totals. The silver refining industry of New
Spain came to produce on average nearly half of the world’s silver during the eighteenth
century.1 Only the United States of America as of the 1870s overtook republican Mexico to
become the world’s leading producer of silver.2 The historical impact of this wave of bullion
from the New World on global trade and the future course of the European and Asian
economies has been amply analysed within a global perspective. 3 The same silver is credited
by Flynn and Giraldez with giving rise in 1571 to the first instance of truly global trade, the
year Spain set up its entrepôt in Manila.4 There is consensus that the mining and refining of
silver in the New World changed the course of world history as of the mid sixteenth century.
The reverse side of the silver coin was the environmental legacy in the New World that
was left in the shadow of this pivotal moment in history. Up to the end of the nineteenth century
1
John Jay TePaske and Kendall W. Brown, A New World of Gold and Silver (Leiden, Netherlands; Boston: Brill,
2010), 140.
2
Charles White Merrill, Summarized Data of Silver Production (US Government Printing Office, 1930), 18.
3
Modern historians of note such as the Chaunus, Braudel, Carla Rahn Phillips, de Vries and others have been
cited in this regard, as in for example the analysis of the concatanated economic effects of New World silver flows
as it made its way through Europe, the Ottoman Empire, Persia, India and ultimately to China, in Ronald Findlay
and Kevin H. O’Rourke, Power and Plenty. Trade, War and the World Economy in the Second Millenium
(Princeton: Princeton University Press, 2007), 212-26. John Kenneth Galbraith argues that it was not the silver
per se that strongly assisted the birth of European capitalism, but its inflationary effect on prices of goods coupled
with a decrease in wages, that led to increasing profits and capital accumulation, in John Kenneth Galbraith,
Money, Whence It Came, Where It Went (London: Penguin Books Ltd., 1995), 10-11. Silver as a ‘gift’ to Europe,
product of Spanish coercion, and its effects on Indian and Chinese economies is analyzed in Kenneth Pomeranz,
The Great Divergence. China, Europe, and the Making of the Modern World Economy (Princeton, New Jersey:
Princeton University Press, 2000), 269-74.
4
Dennis O Flynn and Arturo Giraldez, "Cycles of Silver: Global Economic Unity through the Mid-Eighteenth
Century," Journal of World History 13 (2002).
47
there is little in the way of published primary documents that comment on the collateral impact
on communities and their habitat as a result of refining silver. One notable and very early
exception was the following text by Inca Garcilaso de la Vega (1539-1616), the son of a
Spanish conquistador and an Inca princess, which refers to the refining of mercury from its
ore:
‘The Inca Kings felt that [mercury] was harmful to the welfare of those who extract it and
refine it, for they saw it caused tremors and the loss of consciousness. In view of which (as
Kings that care so much for the welfare of their subjects, according to their name “Lover of
the Poor”) they prohibited by law its extraction, or any memory of it. So much did the Indians
abhor it that they erased its name from their memory and language’.5
The early European sources from the Viceroyalty of Peru adopt a more business as
usual approach to the workplace dangers of refining silver. Viceroy Toledo imposed larger
heights of chimneys from smelting furnaces than their counterparts in amalgamation units.6
Alonso Barba, author of the only major metallurgical work in the early seventeenth century to
come from the New World, recommended that workers do not stand downwind during the
heating of the amalgams to avoid the danger of mercurialism (becoming azogado) should an
accident occur.7 In the eighteenth century, de Gamboa would comment on the poisonous nature
of smelters and amalgamation haciendas.8 At least one lurid account of the deemed mortal
effect of amalgamation on workers was penned by Sir William Rawson in England in the early
5
‘Los Reyes Inca … sintieron [que el azogue] era dañoso para la vida de los que lo sacan, y tratan, porque vieron
que les causaba el temblar y perder los sentidos. Por lo cual (como reyes que tanto cuidaban de la salud de sus
vasallos, conforme al apellido “amador de pobres”), vedaron por ley que no lo sacasen, ni se acordasen de él.
Y asi lo aborrecieron los indios de tal manera, que aun el nombre borraron de la memoria y de su lenguaje’.
Garcilaso de la Vega and Aurelio Miró Quesada S, Comentarios reales de los Incas (Caracas: Biblioteca
Ayacucho, 1976), 555.
6
As quoted in Peter J. Bakewell, Miners of the Red Mountain. Indian Labor in Potosí, 1545-1650 (Albuquerque:
University of New Mexico Press, 1984), 150.
7
Alvaro Alonso Barba, Arte de los metales (Barcelona: Editorial Labor, 1977 ), 170.
8
Francisco Xavier de Gamboa, Comentarios a las ordenanzas de minas dedicados al católico rey nuestro señor
Carlos 111, Madrid, Oficina de Joaquín Ibarra (1761), 462.
48
nineteenth century.9 However the detailed accounts by European technical observers of this
period do not mention any fatal environmental or workplace conditions as a consequence of
the refining processes.10 Concern over the depletion of woodlands was expressed throughout
the colonial period within the context of the impact of mining in general, perhaps not so much
in the modern sense of safeguarding the environment for future generations but more as a
problem of sourcing new supplies of fuel so as not to impair the course of silver production.11
A generalized interest in the negative impact of human actions on the environment only
arises after the middle of the twentieth century. As more information became available on the
toxicity of chemicals, greater attention began to be paid to the environmental effect of industrial
activity, with Rachel Carson’s publication of Silent Spring in 1962 a significant milestone.12
Her concern was triggered by the indiscriminate use of man-made pesticides, but in that same
timeframe and context the devastating effects of mercury and its by-products on humans would
become apparent to a world audience observing its effects on the fishing communities around
Minamata Bay in Japan.13 It cannot be said that the tragedy at Minamata immediately sensitized
the historiography to the possible environmental impact from the use of mercury in the New
World. The emphasis remained on the amounts of mercury produced or imported in relation to
the refining of silver, not their environmental impact. Prior to public reaction to events at
Minamata, Lohmann Villena had published in 1949 his detailed study on mercury production
9
William Rawson, "The Present Operations and Future Prospects of the Mexican Mine Associations Analysed,
by the Evidence of Official Documents, English and Mexican, and the National Advantages Expected from Joint
Stock Companies, Considered, in a Letter to the Right Hon. George Canning,"(London: J. Hatchard & Son, 1825),
19.
10
Saint Clair Duport, De la production des métaux précieux au Mexique, considérée dans ses rapports avec la
géologie, la métallurgie et l'économie politique (Paris: Firmin Didot Frères, 1843).;M.P. Laur, "De la metallurgie
de l'argent au Mexique," Annales des Mines, 6th series, 20 (1871).
11
For a recent review of historical texts on depletion of woodlands in New Spain and the Andes see D. Studnicki-
Gizbert and D. Schecter, "The Environmental Dynamics of a Colonial Fuel-Rush: Silver Mining and Deforestation
in New Spain, 1522 to 1810," Environmental History 15, no. 1 (2010).
12
Rachel Carson, Silent Spring (Greenwich, Conn.: Fawcett, 1962).
13
Shigeo Ekino et al., "Minamata Disease Revisited: An Update on the Acute and Chronic Manifestations of
Methyl Mercury Poisoning," Journal of the Neurological Sciences 262, no. 1-2 (2007).
49
at the mine of Huancavelica, in present day Peru. Using his data, it is possible to calculate a
yearly consumption of mercury for the refining of silver around Potosí of approximately 300 t
per year for at least 74 continuous years.14 The aggregate total of mercury consumed in the
New World from 1556 to 1700, based in part of Lohmann Villena’s data, was then estimated
by Pierre Chaunu in 1959 at approximately 49,000 t (1,069,494 quintales), a yearly average for
the refining centres of the New World of 340 t over 144 continuous years, that would increase
in tandem with silver production in the eighteenth century.15 One estimate on the total amount
of mercury (and its compounds) discharged into Minamata Bay is from 260 t to 600 t of
mercury over a period of 36 years, at an average of 7 to 17 t/y, part of which found its way into
the human food chain as very toxic methyl mercury. 16
Further refinements in these totals were published, but none of these magnitudes
triggered a warning on the potential for major negative health and environmental issues for the
communities of the New World arising from the amalgamation of silver ores.17 The historical
data represented an unprecedented major source of anthropogenic waste chemicals based on
mercury voided into the communities and ecology of the New World. Quite the contrary, the
technique of amalgamation, which needed to be supplied with such major quantities of mercury
in order to function, was hailed in much of the modern historiography in quite positive terms.
The importance of New World silver on world history was reflected onto the technique of
14
Saúl Guerrero, "Chemistry as a Tool for Historical Research: Estimating the Contraband of Silver from Potosí
and Oruro, 1576-1650," Bulletin for the History of Chemistry 37, no. 2 (2012): 75.
15
Pierre Chaunu, Seville et l’Atlantique (1504-1650). La Conjoncture (Paris: SEVPEN, 1959), Tome VIII (2,2)
1975.
16
Ronald Eisler, Handbook of Chemical Risk Assesment: Health Hazards to Humans, Plants, and Animals.
Volume 1: Metals (Boca Raton, FL.: Lewis Publishers, 2000), 322.
17
Richard L. Garner, "Long-Term Silver Mining Trends in Spanish America: A Comparative Analysis of Peru
and Mexico," The American Historical Review 93, no. 4 (1988): 916-23.; D. A. Brading and Harry E. Cross,
"Colonial Silver Mining: Mexico and Peru," The Hispanic American Historical Review 52, no. 4 (1972): 563. In
the industrial world of the twenty-first century, average mercury emissions from global coal burning in 2013 are
estimated at 474 t/y. Data from UNEP, "Global Mercury Assesment 2013 "(Geneva: United Nations
Environmental Program), 9.
50
amalgamation.18 The pioneering historians of silver refining would mention the danger posed
by the use of mercury, but were limited by the fact there was no mention of widespread
mercurialism in the primary documents spanning the complete span of amalgamation in the
Americas.19 Examples of the dangers of mercury in the refining of mercury ores were cited in
Kendall Brown’s study on Huancavelica (Peru).20 In general, the absence of a strong
commentary on this potentially serious issue as a direct consequence of amalgamation possibly
led the environmental historian John F. Richards to state as late as the year 2003: ‘as yet, the
true environmental costs of silver [refining in the New World] have not been fully explored or
acknowledged by scholars’.21 Nicholas Robins, in his recent monograph on the ravages of
mercury in the Andes of Huancavelica and Potosí, drew attention to the dangers of mercury
but his strongest examples derive from the effects of refining mercury from cinnabar at
Huancavelica, or from the dangers of mining and the attrition on local communities caused by
the mita system of forced labour in the Vice-Royalty of Peru. His case is much weaker on the
18
‘[amalgamation] stands out as a unique event in the annals of Spanish-American technology, up to the present
time’, Marcel Roche, "Early History of Science in Spanish America," Science 194, no. 4267 (1976): 807.;
‘[amalgamation] may rank ... as high as any other technical innovation made in the Americas since Europeans
first went there’, Peter J. Bakewell, A History of Latin America, 2nd ed.(Oxford: Blackwell Publishing, 2004),
185.;‘in the refining of metals ... the americans contributed through amalgamation with perhaps the most important
innovation of the period’- ‘En el beneficio de los metales ... los americanos aportaban con la amalgamación la
innovación tal vez más importante de la época’. Bernd Hausberger, "El universalismo científico del Barón Ignaz
von Born y la transferencia de tecnología minera entre Hispano américa y Alemania a finales del siglo XVIII,"
Historia Mexicana 59, no. 2 (2009): 614. ‘Revolution’, ‘epoch-making’ and ‘the most transcendental event in the
history of world metallurgy’ are some of the other epithets that have been used to describe the implementation of
amalgamation of silver in New Spain in the 1550s. For the respective quotes see Manuel Castillo Martos,
Bartolomé de Medina y el siglo XVI (Santander: Servicio de Publicaciones de la Universidad de Cantabria, 2006),
75. ; Mervyn F. Lang, "Silver Refining Technology in Spanish America (patio y fundición) " in 5th International
Mining History Congress, ed. James E. Fell, P. D. Nicolaou, and G. D. Xydous (Milos Island: Milos Conference
Center-George Eliopoulos, 2001), 139. ; Manuel Castillo Martos and M. F. Lang, Metales preciosos--unión de
dos mundos : tecnología, comercio y política de la minería y metarlurgia iberoamericana (Sevilla: Muñoz Moya
y Montraveta Editores, 1995), 99. Not all historians contribute to this panegyric on amalgamation. Portuondo
dedicates just two lines to this refining process in her 20 page paper, María Portuondo, "Constructing a Narrative:
The History of Science and Technology in Latin America," History Compass 7, no. 2 (2009): 505.
19
In one of the most recent examples, in the second edition of the textbook on Latin American history, published
in 2004, Bakewell dedicates two lines of text to the toxicity of mercury in relation to the use of amalgamation to
refine silver in the New World. See Bakewell, A History of Latin America, 186.
20
Kendall W. Brown, "Workers Health and Colonial Mercury Mining at Huancavelica, Peru," The Americas 57
(2001).
21
John F. Richards, The Unending Frontier: An Environmental History of the Early Modern World (Berkeley:
University of California Press, 2005), 366.
51
environmental impact of amalgamation per se, both on chemical grounds as will be argued in
Chapter 3 and because he can only cite one primary source (Pedro de Oñate, a Jesuit priest) on
the danger of mercury from the amalgamation process itself.22
When attention was finally focused on the environmental impact of silver refining, it
was conditioned by a narrative that had been constructed to explain the uniqueness of a mercury
amalgamation process that was only applied massively in the New World and not in Europe.
This narrative posited that the richer superficial silver ores of the New World had been quickly
exhausted by mid sixteenth century, and that smelting was incapable of refining at a profit the
deeper ores with a low silver content. The implementation of mercury amalgamation was
therefore the only viable technical and economic choice that allowed Spain to reap the wealth
of silver from the ‘other’ poorer type of ore prevalent in the New World. No strategic choice
was involved between two refining options, but rather a decision imposed by necessity. As a
corollary to this narrative, smelting had to be relegated to a minor role in the production of
silver, since ‘rich’ ores were not the average norm in the American continent. Thus
amalgamation and mercury came to dominate the historical narrative on colonial silver
refining, a modern yet unfiltered echo of the voices of colonial miners complaining of lower
silver yields and clamouring for more and cheaper mercury as the key to silver production in
the New World.23
22
N.A. Robins, Mercury, Mining, and Empire: The Human and Ecological Cost of Colonial Silver Mining in the
Andes (Indiana University Press, 2011), 140.
23
It is difficult to single out any specific historian as to the origin of this narrative, since it merges well with some
of the primary documents from the colonial era. The Spanish historian of mining and silver refining Modesto
Bargalló is a very useful source for extracts from early historical sources for mining and refining in New Spain
and Peru. He presents in his book published in 1955 one of the first modern versions of this narrative, in Modesto
Bargalló, La minería y la metalurgia en la América española durante la época colonial (Mexico: Fondo de
Cultura Económica, 1955), 240-45. The same narrative is voiced by the most quoted of English language
historians of colonial silver refining, as for example in: Bakewell, A History of Latin America, 184-89.; Brading
and Cross, "Colonial Silver Mining: Mexico and Peru," 552-56.; Richard L. Garner and Spiro E. Stefanou,
Economic Growth and Change in Bourbon Mexico (Gainesville: University Press of Florida, 1993), 111. An
indication of their role as references to the history of silver refining is the fact the they are the historians cited for
52
The next step in the same direction came not from historians but from the discipline of
environmental science. In attempting to interpret modern levels of mercury depositions world-
wide, the residual contribution of historical silver refining activities needed to be assessed. In
1993 Jerome O. Nriagu published a one-page note in the journal Nature proposing that all the
mercury used to amalgamate silver in the New World had been lost during the process through
physical means, up to 65% during a heating stage of the amalgam, and the rest by spills or in
waterways. Nriagu explicitly tied his line of reasoning to modern observations of artisanal gold
mining in the Amazon.24 His argument was taken up by successive authors, both environmental
scientists and historians, with the latest proposal by Robins in 2012 stating that up to 85% of
the mercury used in Potosí was ultimately volatilized.25 In all these studies no mention is made
of any chemical transformation of mercury during the refining process, nor is any other heavy
metal mentioned as a source of environmental impact from the historical refining of silver ores.
This mainstream environmental narrative based on ‘poor silver ores / amalgamation /
volatile mercury loss’ has faced major challenges in recent years on both fronts, by historians
and indirectly by environmental scientists. The Spanish historian Lacueva, in 2010, questioned
many of the tenets of this narrative. He begins by stressing that in the five year period prior to
the implementation of amalgamation in New Spain, the shipments of precious metals to Spain
(mainly silver) from the port of Veracruz had risen 70%. Thus the use of amalgamation was
not a response to a crisis in silver production. He then argues that certain texts of the period
from Zacatecas contradict the notion of ‘poor’ silver ores. He posits that in fact smelting was a
more profitable refining method than amalgamation, though without a quantitative base to his
the sections on silver refining in John Huxtable Elliott, Empires of the Atlantic World : Britain and Spain in
America, 1492-1830 (New Haven: Yale University Press, 2006).
24
J. Nriagu, "Legacy of Mercury Pollution," Nature 363 (1993): 589.
25
Robins, Mercury, Mining and Empire, 109.
53
argument. Finally, he recognizes that the increase observed in smelting in Zacatecas as of the
mid seventeenth century cannot be readily explained by the ‘poor ore / amalgamation’
narrative.26
Bakewell had earlier recognized the potential contradiction of events in the seventeenth
century to his narrative based on ‘poor ores’. If smelting was held to be only viable for very
rich silver ores, then only new sources of such an ore could justify the surge in smelted silver.
He therefore proposed that gunpowder was helping miners find new pockets of ore with high
silver content. The argument is weak but inevitable for Bakewell to remain consistent with his
narrative. The weakness has been pointed out by Lacueva, who attempts to circumvent the
contradiction trap by claiming that in fact smelting represented the lowest cost option for
refining silver ores.27
The second challenge faced by the mainstream narrative has come indirectly from the
results of recent studies (2011) from the field of environmental science. Cooke et al took
measurements of mercury in deposits at the bottom of Lake Lobato, 6 km east of Potosí. When
core samples of the sediments were measured for mercury, it resulted in a curve of mercury
concentration that decreases during all the historical period amalgamation was practised at
Potosí. In all, 70% of all the mercury deposited over Lake Lobato corresponded to periods
before amalgamation was introduced by Spain. Pre-conquest levels of mercury in the sediment
were higher than those of colonial Potosí.28 This result pointed to a major flaw in the current
26
Jaime J. Lacueva Muñoz, La plata del rey y de sus vasallos : minería y metalurgia en México (siglos XVI y
XVII) (Sevilla: Consejo Superior de Investigaciones Científicas, Escuela Superior de Estudios Hispano-
Americanos; Universidad de Sevilla; Diputación de Sevilla, 2010), 93-107,147-228,259-68.
27
Ibid., 133-34.
28
C.A. Cooke et al., "Pre-Colombian Mercury Pollution Associated with the Smelting of Argentiferous Ores in
the Bolivian Andes," AMBIO: A Journal of the Human Environment 40, no. 1 (2011).
54
narrative on colonial silver refining, with direct consequences for the way its environmental
history should be studied.
The doubts raised by the current direction of the mainstream narrative, of which the
above are just some of the salient objections, required a complete revision of the subject,
starting with the most basic of questions. Why was Spain the only European power to come
across the only major deposits of silver in the world? Why was amalgamation only used
massively in the New World and not in Europe? Was amalgamation the only option available
to Spain to refine the ores it found in the New World? How much silver was produced by
amalgamation and how much by smelting, that is consistent with registered sale revenues of
mercury? The balance of production by each refining method is the key to estimating two quite
distinct sets of environmental impacts. What was the nature and amount of the chemicals
voided to the environment from either refining process? Was mercury the only chemical of
note to have an environmental impact in the New World as a result of silver refining?
The answers to these questions lie in a series of concatenated events that defined the
environmental history of silver refining in the New World. It begins with the geology of silver
deposits, which determine the different chemical nature of silver ores. The chemistry of the
silver compounds then define their response to the two possible refining methods known in this
period, smelting and amalgamation. While smelting can be applied to any silver ore with
enough patience and lead flux, amalgamation of silver ores is in reality a misnomer for a
complex set of chemical reactions that require the transformation of silver sulphides into silver
chlorides (via the action of roasting with salt or through the reaction with copper sulphate),
which are then reduced by mercury into elemental silver, which will only then amalgamate
with the available mercury that has not been transformed into calomel (mercurous chloride).
The presence of lead would impair the efficient use of mercury to amalgamate silver ores. Thus
55
the choice of refining process for silver ores was determined in the first instance by the
chemical nature of the ores.
What was the role of production costs and economics if chemistry was the gatekeeper
to the two quite different paths of environmental impact? Under what economic scenarios of
cost production could smelters derive a profit if the nature of their ores precluded the use of
amalgamation? If lead fluxes were available to use with lead-poor ores, could refiners have
turned to smelting and still made a profit during periods when mercury became too scarce or
payment terms too onerous? Did smelting in fact compete on production costs with
amalgamation? On the other hand, what exactly were the economic reasons behind the use of
mercury, first in New Spain and then in the nineteenth century in republican Mexico?29 For all
that has been written on mercury and amalgamation, there is not a single detailed analysis of
comparative production costs between amalgamation and smelting that can serve to answer
this question, for lack of a suitable historical data base of production costs. What influence did
the fiscal opportunity cost for the Spanish Treasury between amalgamation and smelting have
as a result of Spain’s ownership of the mines at Almadén and Huancavelica? Did the obvious
value of mercury as a revenue stream to the Treasury tilt the balance towards the use of
amalgamation in New Spain, and thus alter the course of its environmental history?
What follows is based on a chemical history of silver refining in the New World, using
the tools from chemistry to interpret the historical documents, much as an economic history
relies on applying principles of economy to decipher historical data. Following Occam’s
dictum I have sought the minimum number of variables that could explain the overall patterns
29
While independence in the nineteenth century marks a clear historical boundary between New Spain and
republican Mexico, the practice and environmental impact of amalgamation and smelting form a more continuous
narrative that starts in mid sixteenth century and terminates with the introduction of the cyanide process at the end
of the nineteenth century.
56
as reported in the historiography of silver refining. I have thus based my analysis on the
chemical behaviour of the two main groups of silver compounds that were the source of silver
production in New Spain and then Mexico: silver chloride with silver sulphides, and
argentiferous galena. The following chapter explains how geology determined Spain’s
predominance in the silver market, and uses current theories on the evolution of silver deposits
to interpret their composition and the manner in which this affected the course of silver refining
in the New World. Geology also explains the fundamental chemical difference, and similarities,
between silver ore deposits on both sides of the Atlantic. The conclusions reached in this
chapter set the groundwork for the behaviour of the ores when subjected to either smelting or
amalgamation in its various forms.
In the following two chapters I will weave between New Spain and the Vice-Royalty
of Peru, because for the latter the extant documents provide a more continuous narrative on the
state of smelting as inherited from Europe and a more detailed source on the critical evolution
of amalgamation to an industrial-scale technique capable of refining sulphidic silver ores. For
each process I establish its generic environmental footprint, which is intimately bound to the
architectural legacy of every refining unit in the New World. Using the chemistry of the
reactions I identify and quantify the separate set of environmental impact vectors generated by
smelting and amalgamation.
I will then focus only on New Spain, where smelting competed more closely with
amalgamation. An important key to interpret the historical trail of silver refining was found in
the impressive architecture of the Hacienda de Santa Maria de Regla on the outskirts of
Pachuca. It is very rare to find at present both the well-preserved physical remains of a major
refining hacienda together with more than a decade of very detailed accounting records. What
made Regla unique was the fact that detailed records for both amalgamation and refining
processes were available. Chapter 4 provides a close-up view of the historical operation of a
57
silver refining hacienda as seen through the mass balance of materials entering and leaving its
walls. These accounts also serve to make up for the absence of historical measures of chemicals
voided to the environment, since mass balances can be used to calculate the magnitude of the
environmental impact of its operations. The simple tenet of the conservation of matter meant
that what entered a refining compound had to be accounted for in what exited the compound,
regardless of its chemical form. As a result it is possible to arrive at ratios of chemicals voided
into the environment per kg of silver refined by amalgamation or smelting. In Chapter 5, the
economic data from Regla give a very rare opportunity to analyse in depth the comparative
production costs and structure of amalgamation and smelting. In turn these insights are used to
explain why refiners in New Spain could have applied in a profitable way both amalgamation
and smelting, subject only to the nature of the silver ore being refined.
The mass ratios for the main environmental impact vectors, calculated for each of the
processes, complemented by the data on silver produced and mercury sold obtained from
primary and secondary sources, are then used in Chapter 6 to project over the main mining
districts of New Spain, or in total over Mexico, a gross estimate for the mass balances of the
major waste products issued to the environment as a result of the historical refining of silver.
It will quantify how the different chemical compositions of the ores being refined lead to totally
different environmental impact vectors, thus highlighting the importance of the geochemistry
of the silver ore deposits to the environmental history of silver refining in the New World. The
results presented at the end of this research propose a major departure from the narrative that
until now has dominated the studies on this topic. The quantitative data indicate it was lead and
its compounds from smelting, not mercury from amalgamation, that was the only heavy metal
to be issued to the air in major quantities. Calomel that exited as solid waste was the end result
of the consumption of mercury during amalgamation. I will argue that the chemical and
physical nature of the refining reactions mitigated the impact of mercury on the workers and
58
communities, while exacerbating the environmental impact of lead and its compounds. Human
choices were made in full knowledge of the toxicity of both lead and mercury, and the final
environmental impact owed less to human foresight than to a fortunate attenuation of effects
from a balance of refining processes dictated by fiscal reasons.
One final word of caution. The conclusions reached in this work only apply to the silver
refining activity as carried out in New Spain and Mexico, and in no way represent the
environmental impact of the mining and processing of mercury ores in Almadén or
Huancavelica, nor to the overall environmental consequences of mining.

59
1 The genesis and nature of silver ores.
‘It seemeth to me a thinge undecent to reade so much of golde and sylver and to know
so lyttle or nothinge of the naturall generation thereof’’ Richard Eden (1555)
‘wherever the earth moves, metals are concentrated ... silver … a gift from the
underworld sealed in cracks’ Fortey, Earth: An Intimate History (2005)
1.1 Introduction
There is no Costa da Prata marked on maps of Africa alongside the Gold Coast and
other shorelines that eponymously signposted each new class of trade goods for the expanding
Europeans. From the fifteenth century to the seventeenth century, the Portuguese would hug
the coasts from Africa to India and all the way to China with the fortified port-cities of the
Estado da Índia, without having a single opportunity to become masters of a world-scale
deposit of silver, and not for lack of searching for one.30 The most they would achieve was to
become middlemen in the Sino-Japanese trade of the sixteenth century, when Japan was known
as the silver islands, and ‘the profits made there by the Portuguese were the envy of sailors,
merchants and adventurers from Amsterdam to Manila’.31
Unbeknownst to them, the Treaty of Tordesillas had effectively quarantined the
Portuguese from ever adding to their wealth and possessions from new silver mines. In fact,
with the exception of Spain and France, no other European power engaged in global expansion
during the Early Modern Era would have as strong an option of coming across world-scale
30
As for example the military campaign by Portugal in search for silver mines in Southern East Africa. A. R.
Disney, A History of Portugal and the Portuguese Empire : From Beginnings to 1807 vol. II (New York:
Cambridge University Press, 2009), 145-71.; or the search for the ‘islands east of Japan “rricas de plata”’ as cited
by Paul W. Mapp, The Elusive West and the Contest for Empire, 1713-1763 (Chapel Hill; Williamsburg, Va.:
University of North Carolina Press, 2011), 109.
31
William S. Atwell, "International Bullion Flows and the Chinese Economy circa 1530-1650," Past and Present
95 (1982): 69,71.
60
deposits of silver ore.32 The dearth of any major primary silver ore deposits in India or China
raises the question as to why huge areas of the Earth remain barren to the present of any
important ‘criaderos’ (nurseries) of silver, to use the organic phrasing found in Spanish
metallurgical tracts of the Early Modern Period. This lack of silver would drive part of the
Asian transcontinental trade towards a Europe incapable as yet of producing the textiles,
porcelain and spices that India, China and the Spice Islands could provide.33 And yet it was no
mean advantage for Europeans to have access to their own silver and gold mines and to have
developed the necessary technology to extract the ores and refine these precious metals,
supplemented by the gold bartered from Africa. By conquering the mountainous spine of the
New World, Spain would have access to a source of silver of a magnitude beyond the powers
of imagination of the European miners of the fifteenth century.
There is a sense of wonder at the unique triangulation of geological gifts that resulted
from the trans-oceanic voyages due west of the Spanish mainland. The chemical context that
defined the technical options for refining silver ores in the New World, and consequently its
environmental history, belongs to the longest of the Braudelian cycles, that of geological time.
Geology did not determine European imperial policy, but it did favour the Imperial ambitions
of the Spaniards, who stumbled across the vast silver deposits found in Peru and New Spain to
complement their own vast deposits of mercury of Almadén in Spain and Huancavelica in the
Vice-Royalty of Peru. During the Early Modern Era, Portugal’s name would be mostly
associated with African and Brazilian gold, while Spain would fund its political aims to a great
32
‘ore: any naturally occurring material from which a mineral or aggregate of value can be extracted at a profit’.
L. J. Robb, Introduction to Ore-Forming Processes (Malden, MA: Blackwell Pub., 2005), 6.
33
Dennis Flynn and Arturo Giráldez, "Latin American Silver and the Early Globalization of World Trade," in
National Identities and Sociopolitical Changes in Latin America (London and New York: Routledge, 2001).;
Andre Gunder Frank, ReOrient: Global Economy in the Asian Age (Berkeley: University of California Press,
1998).
61
extent based on New World silver from the highlands.34 The same geological forces that also
explain the dearth of silver deposits in India and China, or the brief silver bounty in Japan,
configured the nature and direction of the first global flows of trade. Geology is the only path
to provide an answer to Richard Eden’s complaint quoted in the epigraph to this chapter.35
What geology cannot explain is the luck of the draw that kept France from joining Spain as a
major producer of silver of the Early Modern Era. Its colonists of New France were
tantalizingly within reach of a source of silver in the Cobalt area of modern Ontario that
equalled over 80% of the total silver extracted from Potosí by Spain.36
The environmental history of silver refining in the Americas is thus the end result of a
chain of concatenated events that begins with the geological birth of each silver deposit. The
geological genetic imprint defines the nature of the silver compounds contained in each ore,
34
Gold would be found on both sides of the Tordesillas Line, though geologists have remarked on the fact that
silver and gold deposits tend to be dominated by one or the other (Frederick T. Graybeal, Douglas M. Smith, and
Peter G. Vikre, "The Geology of Silver Deposits," in Handbook of Strata-Bound and Stratiform Ore Deposits.
Part IV, ed. K. H. Wolf (Amsterdam: Elsevier 1986), 159. ; Walter Pohl, Economic Geology : Principles and
Practice : Metals, Minerals,Coal and Hydrocarbons - Introduction to Formation and Sustainable Exploitation of
Mineral Deposits (Chichester, West Sussex; Hoboken, NJ: Wiley-Blackwell, 2011), 227. This observation is made
even by the non-specialist observer: ‘one never has at the same time a high content in silver and gold’ -‘on n’a
jamais à la fois haute teneur en argent et en or’, in Albert Bordeaux, Le Mexique et ses mines d'argent (Plon-
Nourrit et cie., 1910), 291. In South America Colombia, Brazil and Venezuela are known more for their gold
than for their silver.
35
Richard Eden (1555) as quoted by Cyril Stanley Smith in his introduction to Vannoccio Biringuccio, The
Pirotechnia, trans. Cyril Stanley Smith and Martha Teach Gnudi (Cambridge, Mass.: The M.I.T. Press, 1966),
xxii. Richard Eden was the translator to English of Biringuccio’s metallurgical text.
36
Over 600 million ounces of silver [over 18,000 t] was produced from the Cobalt area in modern Ontario,
according to Graybeal, Smith, and Vikre, "Geology Silver Deposits," 15. This amount represents 84% of the silver
mined from Potosí up to the end of the eighteenth century (22,170 t, as reported in TePaske and Brown, Gold and
Silver, 184. In monetary terms this would have provided a windfall of over 700 million pesos within the French
colonial economy (at 40 pesos per kg of silver) or over 4,200 million livres (assuming 6 livres to the peso in the
eighteenth century, according to Carlos Marichal, Bankruptcy of Empire : Mexican Silver and the Wars Between
Spain, Britain, and France, 1760-1810 (New York: Cambridge University Press, 2007), 54.). Annual values of
sugar arriving in France from its Caribbean plantations went from 15 to 75 million livres per year from 1730 to
1790 (Robert Louis Stein, The French Sugar Business in the Eighteenth Century (Baton Rouge: Louisiana State
University Press, 1988), 103.), so just the silver value of these deposits represented around 60 years of sugar
imports from the French Caribbean to France at the highest range of sugar prices. If the French mining and refining
establishment of the eighteenth century had the expertise to mine and smelt the Cobalt silver ores, the impact of
silver from the area around modern day Cobalt throughout the French colonial economy of the New World, and
the ripple effect on the overall French strategy towards its empire in the Americas, would have been substantial.
The deposits were discovered in 1903.
62
and of any collateral metallic species. The chemistry of the silver compounds and the other
metals present in the ore in turn defines the optimal refining process that can be applied to
profitably extract the silver content of the ore, either smelting with lead or amalgamation with
mercury. Each refining process finally produced a completely different set of environmental
impact vectors that would create the environmental history of the local communities around
refining centres and their landscape. The Spanish historian Castillo Martos has stated that when
determining the reasons why mercury amalgamation became the method of choice in the New
World and not in Europe, ‘the difference in the nature of American and European [silver] ores
is a fact not to be ignored; however other factors were more influential’.37 I will argue
throughout this thesis precisely the opposite, that the difference in the chemistry of the silver
ores on both sides of the Atlantic is what defined at every step of the way the environmental
consequences of silver refining in the Americas.
To understand what gave rise to the difference in the chemical composition of silver
ores I will briefly review in the following sections the current geological explanations for the
appearance of major silver deposits in the Americas and in Europe, based on present theories
of metallogeny and plate tectonics.38
1.2 Subduction and the New World
Nine of the ten largest sources of primary silver ore known to humankind (those where
at least 50% of their revenue is or was derived from the production of silver) lie on the
37
'La diferente calidad de las menas americanas y europeas es un dato que no habría que despreciar; sin
embargo, influyen más otras consideraciones’. Castillo Martos, Bartolomé de Medina, 106.
38
‘metallogeny [is] the study of the genesis of ore deposits in relation to the global tectonic paradigm ... a range
of processes responsible for the formation of the enormously diverse ore deposit types found on Earth’ Robb,
Ore-Forming, vii.
63
mountainous spine of the New World (Figure 1 and Table I).39 The only source of silver outside
the New World to figure in this exclusive list are the Erzgebirge silver deposits in Europe. The
combined silver endowment of Mexico (smaller now than New Spain) and Peru / Bolivia (the
Vice-Royalty of Peru) continues to be the greatest known deposit of silver on Earth, with a
total amount of silver (produced to date and remaining) to the year 1994 estimated at 441,405
t.40
It is a further measure of the uniqueness of the geological landscape that was conquered
by the Spanish Crown that more than half of the mining locations listed in Table I would have
been recognized by a miner in the New World of the late sixteenth century onwards. 41 Little
did Charles V suspect there was no hyperbole in the motto granted under his reign to Potosí: ‘I
am the rich Potosí, treasure of the world, Lord of all the mountains and the envy of the Kings’.42
At present Potosí remains what Laznicka terms the world’s ‘largest silver supergiant’ ore
reserve, a distinction all the more remarkable since the search for new deposits has continued
since the sixteenth century, with the support of ever increasing technical sophistication. 43 The
Spanish employed the word bonanza to denote a ‘spectacularly rich precious metal zone’, but
39
Graybeal, Smith, and Vikre, "Geology Silver Deposits," 2-32. Their listing published in 1989 includes only
deposits that had yielded over 28 t of silver (one million ounces) to date, and does not include the Imiter silver
deposit of Morocco, a deposit with an estimated 8,000 t of silver, as detailed in Alain Cheilletz et al., "The Giant
Imiter Silver Deposit: Neoproterozoic Epithermal Mineralization in the Anti-Atlas, Morocco," Mineralium
Deposita 37, no. 8 (2002). As of the twentieth century the majority of silver produced is a by-product from the
mining and refining of other metals (for example, copper), which explains why the Lubin Kupferschiefer district
in Poland is now listed as the single largest deposit of silver, even greater than Potosí, even though the silver
content of the ores is a paltry 4 grams per ton, Peter Laznicka, Giant Metallic Deposits : Future Sources of
Industrial Metals (Berlin: Springer, 2006), 54. Sources for Table I: a) Graybeal, Smith, and Vikre, "Geology
Silver Deposits." b) Laznicka, Giant Metallic Deposits. c) TePaske and Brown, Gold and Silver.
40
D.A. Singer, "World Class Base and Precious Metal Deposits; A Quantitative Analysis," Economic Geology
90, no. 1 (1995): 91.Table 2.
41
Graybeal, Smith, and Vikre, "Geology Silver Deposits," 2-32.
42
‘Soy el rico Potosí, del mundo soy el tesoro, el rey de todos los montes y la envidia de los reyes’ quoted in Pedro
Cunill Grau, "El paisaje andino: Punas, Salares y Cerros," in Potosí: plata para Europa ed. José Villa
Rodríguez(Sevilla: Universidad de Sevilla, 2000), 81.
43
Laznicka, Giant Metallic Deposits, 108. The terms giant and supergiant are based on a relative scale determined
by the scarcity of each metal and the size of the endowment in each deposit, thus the more scarce a metal the
lower the threshold in endowment size required to classify it as a giant or supergiant deposit. See ibid., 38-54.
64
it could as well describe the whole geological panorama that opened up to them in the New
World.44 It has been estimated that of the total 337 ‘giant metal deposits’ discovered until
now, only 13 were known prior to 1492, and only an additional 12 would be found up to 1800.45
Cobalt Erzgebirge
Coeur d’Alene
Virginia City 4 6 10
9 10
Idrija
New Almaden Toponah Almaden
Parral 7
Zacatecas 5
3
Guanajuato
1
Pachuca
Huancavelica
Oruro 8
The Treaty of
Tordesillas line 1494
2
Potosí Primary silver deposit zones Major mercury deposits
Figure 1-1. Top ten silver deposits ranked by aggregate production and main mercury
sources up to the late twentieth century. Data from Table 1-I.
44
Graybeal, Smith, and Vikre, "Geology Silver Deposits," 1. According to the Diccionario de la lengua española
(http://www.rae.es), bonanza was a Spanish term originally applied to sailing, and denoted a calm (favourable)
sea.
45
Peter Laznicka, "Discovery of Giant Metal Deposits and Districts," in Proceedings of the 30th International
Geology Congress: Energy and Mineral Resources for 21th Century: Geology of Mineral Economics ed. Pei
Rongfu(VSP, 1997), 356-357.
65
aggregate production production dates endowment/reserves+production/contained

ranking by
aggregate country location ore
silver metal
production metric tons source years source million source source
tons
tons
42,452
1 Mexico Pachuca 36,400 a, 17 1528-1976 a, 17 35 a, 17 (production plus b, 137
reserves)
2 Bolivia Potosí > 28,000 a, 9 1545-1950 a, 9 100 a, 9 86,000 (endowment) b, 131
3 Mexico Guanajuato 31,700 b, 137 to 1984 b, 137 8 a, 17
4 United States Coeur d'Alene 26,180 a, 22 1884-1980 a, 22 50 a, 22 34,000 (contained) b, 364
5 Mexico Zacatecas 23,236 b, 138
6 Canada Cobalt, Ontario ca. 18,000 a, 44 1903-1967 a, 15
7 Mexico Parral 13,496 a, 16 1631-1982 a, 16 50 a, 16
12,000
8 Bolivia Oruro 7,588 a, 8 1595-1962 a, 8 27 a, 8 (production plus b, 146
reserves)
9 United States Virginia City 5,376 a, 28 1859-1950 a, 28 19.1 a, 28
United States Tonopah 4,878 a, 28 1900-1950 a, 28 8.8 a, 28
10
Erzegirbe
7,000 (Freiberg,
Germany (including 4,788 a, 12 1163-1910 a, 12 20 a, 12 b, 364
contained)
Freiberg)
Table 1-I. Major silver deposits ranked by aggregate production. For sources see
footnote 39.
Spain would find three of the dozen in the first fifty years after starting their conquest
of the New World, including the two major epithermal deposits of primary silver ore. Six of
the ten ore deposits that appear in Table I would be under control of the Spanish Crown for a
space of some 250 years. The scale of the bonanza found in the Spanish silver mines in the
New World is highlighted by the fact that ‘Pachuca [in New Spain, present day Mexico, is] the
66
world’s second largest epithermal accumulation [of silver known up to the year 2006] after
Potosí’.46
In its westward thrust of conquest and expansion Spain had stumbled without knowing
across ‘the Andean and Cordilleran orogens of the western Americas [that] contain the greatest
concentration of metals on Earth, and are pervasively mineralized from one end to another’.47
According to Evans, the mineralization observed in the continental margin arcs of the Andes
and Cordillera is unique in the variety and magnitude of its metal endowment.48 The
explanation for the mixture of bounty and scarcity of silver deposits in Figure 1-1 had to wait
for recognition of the dynamic nature of global plate tectonics, the new paradigm in the field
of geology that as of the late 1960s would radically change the view held until then of Earth as
a ‘static, rigid sphere’.49
Subduction is a word that does not appear in the historiography of Spanish America
and New World silver, and yet subduction is the geological script that determined the history
of Spain’s role in the New World.50 In 1972 Richard Sillitoe published a paper in which he
46
Giant Metallic Deposits, 137. The difference between the top tier deposits in Table I and the remainder of
known silver deposits in the world is very large. Over 115 primary silver deposits in the world (excluding the ex-
USSR) were reported in 1980 which had produced over their lifetime over 1 million ounces (oz), equivalent to 28
t. Graybeal, Smith, and Vikre, "Geology Silver Deposits," 6-30. To place this threshold in perspective, 28 tons of
lifetime production corresponds to approximately a quarter of just one year’s output in Potosí averaged over 250
years continuous production during the Early Modern Era. World-class silver deposits that have accounted for
79% of all silver discovered and produced are those that contain over 2,400 t of silver (the top ten percent of all
known deposits), while supergiant deposits of silver contain over 22,000 t of silver, and account for 37% of total
silver endowment. Singer, "Precious Metal Deposits," 10.
47
Robb, Ore-Forming, 338.
48
Anthony M. Evans, Ore Geology and Industrial Minerals : An Introduction (Oxford; Boston: Blackwell
Scientific Publications, 1992), 331.
49
R. W. Carlson, "Introduction," in Treatise on Geochemistry : The Mantle and Core., ed. R. W. Carlson (Oxford:
Elsevier-Pergamon, 2004), 15. There is an interesting account of the paradigm change brought about by plate
tectonics, and a discussion in terms accessible to the non-geologist as to why deposits of silver arise from
movements within the Earth, in Richard A. Fortey, Earth : An Intimate History (London: Harper Perennial, 2005),
202, 253.
50
The theory of subduction as the explanation for the silver wealth exploited nearly exclusively by Spain has not
yet received attention in recent histories of the region to explain the uniqueness of its silver deposits, for example:
Kendall W. Brown, A History of Mining in Latin America : From the Colonial Era to the Present (Albuquerque:
University of New Mexico Press, 2012).; Robins, Mercury, Mining and Empire.; Elliott, Empires Atlantic World.;
Manuel Castillo Martos, "Plata y revolución tecnológica en la América virreinal," in Historia de las Ciencias y
67
proposed an explanation as to why the western area of the New World has been privileged with
metallic richness.51 According to Sillitoe, the metallogeny of silver ore deposits of the
American continents begins at the mid-ocean ridges within the Pacific Ocean, extended gashes
on the sea-floor where ocean crust is being continually issued forth from material that rises
from the mantle, and where metal rich solutions are vented and precipitated minerals fall on
the newly created and spreading ocean floor. The new ocean crust spreads on either side of
each ridge until the eastern portion reaches first the continental crust of the Americas (Figure
1-2). Instead of a head-on-collision, a subduction of the oceanic crust takes place as it slides
under the thicker continental crust, a very slow motion version of the last step of a mechanical
escalator sliding continuously under the more static landing stage. In so doing it entrains with
it all surface metallic deposits vented from the mid-ocean ridge and also accumulated during
its travel under the Pacific Ocean, together with quantities of sea-water. As the subducting
ocean crust slides deeper under the continental crust it enters a zone where, in Forley’s powerful
and succinct summation, whenever the Earth moves metals are indeed concentrated. 52 In the
case of silver, hydrothermal processes will play a major role: ‘The most efficient metal
transport in ore formation is by aqueous fluids [in which] metals dissolve … through simple
de las Técnicas, ed. Luis Español González , José Javier Escribano Benito, and María Ángeles Martínez García
(Spain: Universidad de la Rioja, 2004).; Bakewell, A History of Latin America.
51
R.H. Sillitoe, "Relation of Metal Provinces in Western America to Subduction of Oceanic Lithosphere,"
Geological Society of America Bulletin 83, no. 3 (1972): 815. For a more recent discussion on subduction and the
American continent see Suzanne Mahlburg Kay, Víctor A. Ramos, and William R. Dickinson, Backbone of the
Americas : Shallow Subduction, Plateau Uplift, and Ridge and Terrane Collision (Boulder, Colo.: Geological
Society of America, 2009).Other possible processes that have been proposed involve the scraping off of certain
metal content from the base of the overriding plate during subduction; inhomogeneous distribution of metals
below the solid crust of the Earth, scavenging of pre-existing concentrations at depth, and accretion from large
meteorites that fell at an early stage of Earth’s history. For a more detailed discussion see A. H. G. Mitchell and
M. S. Garson, Mineral Deposits and Global Tectonic Settings (London; New York: Academic Press, 1981),
191,324-326. For an approach based on melt generated by subduction then zoned magma chambers leading to
volatile stripping of copper and silver via magmatic vapour phases followed by extended fractional crystallization
of metals and hydrothermal activity see B. Lehmann, A. Dietrich, and A. Wallianos, "From Rocks to Ore,"
International Journal of Earth Sciences 89, no. 2 (2000): 287-292.
52
Fortey, Earth : An Intimate History 253.
68
ions ... or complexes’.53 Hydrothermal processes involve the action of water to transport metals
under the Earth in concentrated solutions akin to the strong metal-rich soup of a hot mineral
spring. Epithermal deposits is the term used for the resulting hydrothermal ore deposits formed
at shallow depths (less than 1500 meters) and fairly low temperatures (50-200oC).54
PacificOcean
Pacific Ocean Andes
Andesand
andcordilleras
Cordillera
Ocean trench
Gold- Silver-Lead- Tin
Iron Copper Zinc
ascent of
magmas and
included
metals
Oceanic
crust
East Pacific Rise

(section)
Partial melting on subduction,
sequential release of metals
entrained on ocean crust
Figure 1-2. The spreading ocean floor subducts under the continental crust of the Americas
(illustration based on Sillitoe in footnote 51).
Silver and its compounds are among the metals that have been deposited in major
amounts via subduction processes along the mountainous spine of the Americas. Figure 1-3 is
of great help in visualizing the role of subduction in defining the location of major primary
silver deposits in the New World. It also shows why the history of silver refining in Japan has
53
Lehmann, Dietrich, and Wallianos, "From Rocks to Ore," 285.
54
Robb, Ore-Forming, 7.
69
the same geological roots as that for New Spain.55 In the Pacific coast of the Americas
subduction started in the Mesozoic [201 to 66 Ma] and early and middle Cenozoic [66 to 2.5
Ma] and is still very much active beneath Central and South America. 56 I have already
mentioned that the Andes have been described by modern geologists as the single most
important concentration of metallic ore deposits to be found in the world.57 According to plate
theory the reason for this distinction is the ‘singular longevity of the convergent plate boundary
of the eastern Pacific rim’, so that even as these words are written a section of oceanic crust
that was formed some 55 million years ago at the mid-ocean ridge is now sliding continuously
under the South American plate in an on-going process of subduction.58 Nowhere else on earth
has such persistent and active subduction occurred so close to a large and inhabitable landmass.
This proximity is crucial for historical events of the Early Modern Era to have been influenced
by this geological process. To quote a strangely disturbing Freudian turn of phrase: ‘Latin
America was, in comparison with other regions in which mineral wealth was potentially to be
extracted, eminently penetrable’.59 Long- term and active subduction is what distinguishes the
Spanish silver mines of the New World from the blank spaces in Figure 1-1.
Not all subduction zones are created equal as far as silver is concerned. The cornucopia
of New World metals, including silver, ‘may’ be related to its proximity to the East Pacific
Rise, as metals keep being added to the spreading oceanic crust. ‘The whole Eastern Pacific
55
William R. Dickinson, "Anatomy and Global Context of the North American Cordillera," in Backbone of the
Americas : Shallow Subduction, Plateau Uplift, and Ridge and Terrane Collision, ed. Suzanne Mahlburg Kay,
Víctor A. Ramos, and William R. Dickinson (Boulder, Colo.: Geological Society of America, 2009), 2.
56
Sillitoe, "Metal Provinces in Western America," 815. Ages in parenthesis throughout are expressed in million
years (Ma) and correspond to the classification as published in the 2011 version of the International Stratigraphic
Chart (www.stratigraphy.org).
57
One important exception relevant to the history of silver refining in the Americas is iron.
58
Pohl, Economic Geology, 222.; Mitchell and Garson, Global Tectonic Settings, 186.; F. T. Graybeal and D.M.
Smith Jr, "Regional Distribution of Silver Deposits on the Pacific Rim," in Silver - Exploration, Mining and
Treatment(Mexico City: Institution of Mining and Metallurgy, 1989), 7.
59
Peter J. Bakewell, "Introduction," in Mines of Silver and Gold in the Americas ed. Peter J. Bakewell (Aldershot
(UK): Variorum, 1996), xxii.
70
continental margin arcs from British Columbia to Chile are well endowed with major silver
deposits while the Western Pacific continental margin has virtually none’.60 According to
Graybeal and Smith the shorter the time interval to subduction, as applies to the eastern side of
the Pacific Ocean crust, the more evident the effect of ‘enrichment of silver in young ocean
crust at spreading centres’, which is why the western areas are silver-poor in comparison.61
Asia
North America South America

Mexico
Japan
NAZCA PLATE
PACIFIC PLATE
ANTARTIC
PLATE
COCOS PLATE
circum Pacific arrows indicate

orogenic belt subduction of plate
Figure 1-3. Subduction along the Andes, Cordillera and Japan (based on illustration in
footnote 55).
Even if subduction is the fundamental geological process that gave rise during different
geological eras to silver deposits both in the Americas and Europe, the resulting chemical
nature of the silver compounds and accompanying metals varied substantially in each location,
as I will discuss in the following sections.
60
Evans, Ore Geology, 330-331.
61
Graybeal and Smith Jr, "Regional Distribution of Silver Deposits on the Pacific Rim," 3.
71
1.3 The geological difference in New World silver deposits
New World silver did not arise from a single type of metallogeny, since there is a
distinct genesis to silver ores in the Andes compared to those of New Spain, even if both are
the ultimate product of subduction processes. It has long been evident that the two silver
refining industries proceeded along different historical lines, and historians have proposed this
arose from differences in the silver content of the ores and from the tax burden imposed upon
refiners.62 In geographical terms the silver in the Andes has been concentrated in few and large
deposits at very high altitude, such as Potosí. In the case of New Spain, the geographical
dispersion of silver deposits is much more pronounced, and none has reached the magnitude of
Potosí. As to the geological difference between the two, according to Sillitoe:
“It should be noted that the Bolivian silver deposits are parts of tin systems formed in a
back-arc setting by chemically reduced magmas dominated by a crustal source. In contrast, the
main Mexican silver deposits are accompanied by lead, zinc and subordinate gold and are
associated with chemically oxidized magmas of combined mantle plus crustal origin that were
formed in an arc setting above a shallowly inclined subduction zone. Hence, the two regions
differ substantially in both their tectonic and magmatic settings [emphasis added]’.63
There is as yet no work that relates directly the geological difference of both areas to
the geographical concentration, dispersal and size of silver deposits. However, the greater
probability of finding both lead and gold together with silver in the deposits of New Spain
(Mexico) compared to the Andean locations is extremely relevant in the light of the importance
of smelting in New Spain compared to the Andes. The presence of lead is a necessary condition
62
Brading and Cross, "Colonial Silver Mining: Mexico and Peru."
63
Dr. Richard Sillitoe, private communication. The mention of crustal or mantle sources of magma indicate the
origin of the magma from where metals and their compounds are ultimately leached from by hydrothermal
processes. Chemical reduction or oxidation will determine the final chemical profile of these metals and
compounds. Subduction processes of a different nature are indicated, back-arc setting (inland side of the Andes),
or shallowly inclined subduction zones in the case of Mexico.
72
for smelting, and the added cash flow from gold would assist in making this process
economically viable. I will return to both issues in the chapters that follow.
1.4 The chemistry of the sources of silver in New Spain / Mexico
It is virtually impossible to obtain a wide sampling of the ores found by Europeans at
each of the mining sites developed during the first centuries of silver mining in the New
World.64 In one location even the historical mines have been physically erased from the face
of the earth by modern day mining techniques in search of parts per million of gold. 65 The
problem is compounded by the fact that silver is extracted from a greater number of different
ores than any other metal: over 200 varieties of silver ore have been reported.66 The complexity
of the multiple sources of silver is captured in its bewildering visual array by the early historical
observers:
‘it is common to find some [silver] clean and purified, that does not need to be refined … it has
waxed sometimes as glitter; other times, wrapped around a stone like a thin piece of string
made of fine silver ... silver that arises in minerals encrusted in stones ... is a marvellous thing
to see in how many ores it is nurtured ... because some are black; others, yellow, grey-brown,
light-brown, light-coloured or of many shades of colour; some, extremely hard and thus
stubborn, and others soft, tender ... some ores are earthy, others leaden, others are laced with
iron pyrites; and others are mixed with gold, copper, tin, lead, caparrosa [ferrous or copper
sulphate], to sum it up, there is hardly any [silver] that can be found that is not mixed in various
ways’.67
64
‘No specimens of silver minerals survive from the original mines [of Potosi]’. T.C. Wallace, M. Barton, and
W.E. Wilson, "Silver & Silver-Bearing Minerals," Rocks & Minerals 69, no. 1 (1994): 35. While there are reports
of samples collected and sent back to Spain from the New World, the fate of these samples is unknown.
65
The whole Cerro San Pedro on the outskirts of San Luis Potosí is at present being levelled to the ground through
open-pit mining operations by the Canadian mining company New Gold. The historical mines dating from the
late sixteenth century onwards have disappeared literally into thin air.
66
Claudia Gasparrini, "The Mineralogy of Silver and its Significance in Metal Extraction," CIM Bulletin 77, no.
866 (1984): 99.
67
‘se suele hallar alguna limpia y acendrada, que no tiene necesidad de beneficiarse … cuajase algunas veces
como escarcha; otras, revuelta a una piedra como un delgado hilo de plata fina … la plata que nace en minerales
incorporada en piedras … es cosa de maravilla ver cuán diferentes son los metales en que se cría. Porque unos
son negros; otros, amarillos, pardos, de color castaño, rubio y de todos colores; unos, durísimos y por extremo
empedernidos, y otros blandos, tiernos … unos metales hay terrosos, otros plomizos, otros margajitosos; y otros
tienen mezcla de oro, cobre, estaño, plomo, caparrosa; y, en suma, casi no se halla ninguno que no tenga varias
mixturas’. Bernabe (S.J.) Cobo, Historia del Nuevo Mundo, ed. Francisco (S.J.) Mateos, vol. 91, Biblioteca de
Autores Espanoles (Madrid: Atlas, 1964), 141.
73
In the absence of physical historical evidence it is necessary therefore to turn to modern
research on the main chemical compounds from which the major part of silver production has
been obtained, and the current state of knowledge on the chemical transformations that are
known to take place in deposits of silver ore. I have expressly avoided addressing each of the
historical silver deposits of New Spain in the formal terminology of geology that categorizes
ore deposits of different genetic types according to all the metals present, the host rocks that
ultimately make up the gangue or waste mineral material, and the mechanisms by which these
metal compounds were deposited in the host rocks.68 This would require detailed historical
knowledge of the geological genetic type of silver ore deposit for each mining district (Real de
Minas) of New Spain, which according to Humboldt numbered around 500 in the 1800s.69 On
the contrary, I will base my analysis of the environmental impact of historic silver refining on
the one factor common to all the Reales, that the silver produced in New Spain came from
basically just two chemical groups of silver compounds distributed among these deposits: 1)
silver sulphide compounds (simple to complex), that via weathering as described below could
give rise to surface concentrations of silver halides (mainly silver chloride) and also metallic
silver and 2) argentiferous galena (lead sulphide, PbS, that contained silver compounds).70
68
As one example, these descriptive models have been published by the United States Geological Survey for
deposits such as currently found in Pachuca (Model 25 b, Dan L. Mosier et al., "Descriptive Model of Creede
Epithermal Veins," in U.S. Geological Survey Bulletin 1693, ed. Dennis P. Cox and Donald A. Singer
(Washington: U.S. Geological Survey, 1987), 145-49. ) There is an updated level of description that now also
takes into account the environmental consequences of the various chemical compounds that make up the deposit,
under the term ‘geoenvironmental models’. For model 25 b see Geoffrey S Plumlee et al., "Creede, Comstock,
and Sado Epithermal Vein Deposits," in Preliminary Compilation of Descriptive Geoenvironmental Mineral
Deposit Models, Open File Report OFR-95-0831 ed. E. du Bray (Denver, CO: US Geological Survey, 1995), 152-
61.
69
Alexander von Humboldt, Essai politique sur le royaume de la Nouvelle-Espagne (Paris: Chez F. Schoell,
1811), Tome III, 310.
70
In contrast to gold, the main primary source of silver was not the native metal but the chemical compound silver
sulphide (Ag2S). V. M. Goldschmidt, Geochemistry (Oxford: Clarendon Press, 1954), 189. Sulphur is not one of
the most common elements on Earth (around 0.05% of the crust) but its role in the movement of metals
underground might explain its unmistakeable presence around active volcanoes as a yellow efflorescence of
elemental sulphur on the crater walls or via the rotten-egg smell of hydrogen sulphide (H2S). Silver is classified
as a calcophile metal, one that prefers to react with sulphur, as anyone eating a boiled egg with a silver spoon can
quickly ascertain. Pohl, Economic Geology, 208,356. Silver sulphide presents itself at ambient temperature as the
74
1.4.1 Silver sulphide and silver chloride
Nearly one hundred years ago, Emmons explained that a silver ore deposit is not an
inert mountain of material but a constantly evolving chemical reactor in which over time the
composition of the original (hypogene) silver compounds is constantly changing. 71 Because
the surface layer of the deposit is a zone where oxygen plays an important role, the term
‘oxidized zone’ has led to the erroneous statement in some of the historiography that silver
oxide is generated in this surface layer.72 It is not silver that is oxidized but other chemical
species present in the ore deposit: the ferrous ions to ferric ions, or the sulphide ions to sulphate.
Quite the opposite, any silver ion will be reduced to elemental silver, or be solubilized by the
sulphate ions and later precipitated as chlorides, and other halides.73
mineral acanthite, the most common of all silver ores, so soft that it can be cut with a knife, with a colour range
‘from black to mirror white silver’. Acanthite is the low temperature (< 173 ° C) form of argentite, and in some
of the literature both terms appear to be readily interchanged. Wallace, Barton, and Wilson, "Silver-Bearing
Minerals," 25,27. Acanthite would be found in most mines of New Spain and Peru in huge quantities, as the main
hypogene (original) silver mineral. It can also be found as part of the silver content in galena, the sulphide of lead.
71
William Harvey Emmons, The Enrichment of Ore Deposits vol. 625 (Washington: United States Geological
Survey, 1917), 264.
72
‘the interaction between mercury and the surface of gold is easier than with the surface of silver, and this may
be the reason why the Romans did not amalgamate silver. The latter [silver] oxidizes when exposed to air and,
sometimes, appears with a film of silver oxide (Ag2O) that impedes contact between the metals [silver and
mercury]’- ‘La interacción entre el azoque y la superficie de oro nativo es más fácil que con la de plata y ello
puede ser motivo por el cual los romanos no amalgamaron plata. Esta última se oxida en el medio ambiente y, a
veces, aparece recubierta de una capa de óxido de plata (Ag 2O) … que impide el contacto entre los metales.’
Castillo Martos, Bartolomé de Medina, 74.; mention of finely ground silver ores (oxides, chlorides and sulphides)
in Bakewell, "Introduction," xvi. The chemical texts are quite clear on this topic: ‘silver oxide could not
accumulate in oxidizing zones, because it is soluble in acid and also [to a limited extent] in water ... it is unknown
as a natural mineral [of silver]’ S.A. Cotton, Chemistry of Precious Metals (Bristol: Blackie Academic &
Professional, 1997), 282.; Emmons, The Enrichment of Ore Deposits 625 255.; ‘other naturally occurring silver
compounds, including oxides … are either rare or unknown’, in Richard H. Sillitoe, "Supergene Silver
Enrichment Reassessed," in Supergene Environments, Processes and Products, ed. Spencer R. Titley (Society
Economic Geology, 2009), 22.; ‘neither metal [silver and gold] is attacked by oxygen, but silver reacts with H2S
in town air forming a black tarnish of Ag2S.’ Cotton, Chemistry of Precious Metals , 275.
73
Although it is an older paper, Emmons provides detailed chemical reactions that take place in the oxidation
zone, and serves as a guide to a correct interpretation of the chemical changes in this region. Emmons, The
Enrichment of Ore Deposits 625 157-62,252-74.
75
Sillitoe has published an extensive review of the research up to the year 2009 on the
changing nature of the chemical composition of silver ore deposits.74 He states there is a
fundamental difference between copper deposits, where enrichment has been studied and
confirmed even down to levels below the water table, and the behaviour of deposits of silver
sulphide, which are the ones that predominate in the New World. The revisionist proposal by
Sillitoe that ‘supergene sulphide enrichment is an economically unimportant process’, being
absent in most examples of major extant silver deposits he reviews in his paper, leads to a
conclusion of great relevance in the interpretation of the early subjective reports by miners in
the New World. The first generation of Spanish refiners in the New World found a text-book
example of an undisturbed deposit of silver sulphide, in which superficial levels had undergone
weathering to silver chloride and silver in the zone above the water table. According to Sillitoe,
the absence of supergene enrichment means that on average the silver content found at the more
superficial levels is indicative of the silver content as a whole for the deposit. In other words
the major change as extraction proceeded within most mines in the New World was not so
much in total silver content as in the nature of the silver compounds within which it was to be
found (Figure 1-4).75
Not all the silver sulphide deposits or even all the veins of a major deposit necessarily
undergo any process of chemical change at all. When it does take place, the first segment a
74
Sillitoe, "Supergene Silver Enrichment," 21-23. In this paper Sillitoe provides a more detailed geological
description of major silver deposits in the world, including an estimate of their reserves of silver. An earlier
overview of silver sources based on the type of deposit in the Americas is provided in D.M. Smith Jr, "Geology
of Silver Deposits along the Western Cordillera," in Silver - Exploration, Mining and Treatment (Mexico City:
Institution of Mining and Metallurgy, 1989).
75
There is no doubt that miners came across pockets of extremely rich silver content, as in this description: ‘I
have seen in the mine of Zacatecas such a rich vein of silver that, on placing it in the fire, it spit out pieces of
silver the size of a broad bean’ - ‘yo he visto en la mina de Zacatecas una vena de metal tan rico que, quemando
en el fuego, escupía pedaços de plata como habas’ Agustin de Sotomayor, as quoted in Julio Sánchez Gómez, De
minería, metalúrgica y comercio de metales : la minería no férrica en el Reino de Castilla, 1450-1610 (Salamanca:
Universidad de Salamanca : Instituto tecnológico geominero de España, 1989), 34. According to Sillitoe’s
argument these bonanza zones are not the end product of weathering but the high concentration level of silver was
present from the origin of the deposit.
76
Spanish miner would have found close or at the surface would have provided him initially with
an ore mainly in the form of native silver and silver chloride (chlorargyrite), for the most part
generated by oxidation-reduction reactions above the water table.76 As the Spanish miners
continued to extract silver ores at deeper levels down to and below the water table, these would
revert to the original and primary (hypogene) silver sulphide, native silver, and more complex
sulfosalts such as pyrargyrite that contain antimony, an element that interferes in all silver
refining processes.77
surface
oxidation zone
water table
silver sulphide
silver sulfo-salts
elemental silver
silver chloride
NO WEATHERING AFTER WEATHERING
Figure 1-4. A simplified representation of Sillitoe’s interpretation of the effect of

weathering on an original (hypogene) non-weathered sulphidic silver deposit. Drawing based
on interpretation of Sillitoe in footnote 74.
The earliest historical texts that reflect the mining practices of the Spanish conquerors
of the sixteenth century reflect from the first moment the main classes of silver compounds
76
Blanchard has identified silver chloride as the main silver compound that differentiates silver ores of the New
World from those known in Europe up to the sixteenth century. Ian Blanchard, Russia's "Age of Silver". Precious-
metal Production and Economic Growth in the Eighteenth Century (London; New York: Routledge, 1989), 3.
77
Sillitoe, "Supergene Silver Enrichment," 22,30.
77
indicated in the discussion based on Sillitoe’s review. A useful guide is found in the dictionary
by De Llanos from 1609, which describes three main visually distinct classes of silver ores:
a. Very rich silver ore or native metal that could be worked directly with a hammer was
called respectively tacana or machacado.78
b. Ores found from the surface down to the water table were called colorados in New
Spain and pacos in Peru. In some areas the presence of reddish iron oxide-hydroxide minerals
such as limonite-haematite in this zone (gossan is the term used in geology texts) or the mixture
of oxidized iron pyrites and enriched silver minerals found near the surface gave rise to the
term of ‘coloured’ ores in Spanish. By inference these ores would be made up primarily by
native silver and silver chloride, other silver halides, and some silver sulphide.79 The depth of
the oxidized zone will vary according to each deposit and its climatic conditions. According to
Burkart the colorados in certain mines of New Spain reached down to 150 metres, in others
the negros (see below) reached the surface, testimony to the fact that the degree of weathering
78
Garcia de Llanos, Diccionario y maneras de hablar que se usan en las minas y sus labores en los Ingenios y
beneficios de los metales (1609) (La Paz, Bolivia: MUSEF, 1983), 79-80. Even if ‘it is noteworthy that in many
well-known mining districts there is very little of the native metal’, native silver can constitute either an important
hypogene or secondary source of the metal. Emmons, The Enrichment of Ore Deposits 625 263. With the
exception of Cobalt, Ontario, it is claimed that it was more common to find native silver in Europe than in the
New World. For example, at Schneeberg (Erzgebirge area) a solid mass of native silver measuring 1x2x4 metres
was found in 1477 from which a table was made for the Duke Albrecht of Saxony, which weighed around 20 tons.
Wallace, Barton, and Wilson, "Silver-Bearing Minerals," 20-22.
79
Garcia de Llanos, Diccionario, 79-80. For a description of how gossan is formed see Pohl, Economic Geology,
85. and Robb, Ore-Forming, 239. Silver chloride, AgCl, is found as the mineral cerargyrite (chlorargyrite, horn
silver). Cerargyrite is very soft, white or transparent when fresh; on exposure to light, it immediately darkens and
becomes opaque. Typically, specimens are brown. Wallace, Barton, and Wilson, "Silver-Bearing Minerals," 28.
‘Cerargyryte [in modern texts written as cerargyrite] (horn silver, AgCl) is probably unknown as a primary
constituent of ores deposited by ascending hot waters but is commonly developed by weathering, alteration … at
or near outcrops of silver-bearing sulphide lodes... its occurrences include nearly all sulphide deposits ... in arid
undrained areas it is an important ore mineral, so important that the term “chloriding” is generally used in such
regions for pocket hunting near the surface. It is fairly stable at the surface, particularly in arid countries’.
Emmons, The Enrichment of Ore Deposits 625 272-73. Silver halides in general would give the mines of Catorce
in the province of San Luis Potosí a special distinction with regards to the refining process they would adopt at
the end of the eighteenth century.
78
is not necessarily the same in every silver ore deposit.80 The greater the aridity, the greater the
amount of silver chloride (cerargyrite or chlorargyrite) to be expected in a deposit.81
c. Darker and deeper ores found above and below the water table called negros or
negrillos, the silver sulphide and sulfo-salt ores.82
As the knowledge of the nature of the different silver ores advanced in tandem with
chemical theory, the information on silver ores in the historiography converges on the average
chemical profile of a silver sulphide ore deposit as described in Sillitoe’s summary above. 83
The lack of scientific knowledge of the indigenous workers at the mine was more than
compensated by the skill to pursue a vein underground based on the visual and tactile evidence
under the tenuous flicker of a yellow flame. However, what would have posed a greater
challenge to the Spanish miner-refiner aboveground was the ability to recognize that ores could
contain the same overall silver content but require completely different levels of refining skills
to extract the same amount of silver from them. The historical consequence was the waste of
80
J. Burkart, "Du filon et des mines de Veta-Grande, près de la ville de Zacatecas, dans l'état du même nom, au
Mexique," Annales des Mines, 3eme série 8 (1835): 66-67.
81
Emmons, The Enrichment of Ore Deposits 625 , 256; Pohl, Economic Geology, 222.
82
Garcia de Llanos, Diccionario, 86. Another sulphur containing hypogene silver mineral of commercial
importance is pyrargyrite, Ag3SbS3, a sulphide containing both silver and antimony. Pyrargyrite is an important
silver mineral in the mines of Zacatecas, Guanajuato and Pachuca of New Spain. Emmons, The Enrichment of
Ore Deposits 625 276. ‘The name comes from Greek, fire and silver ...dark ruby silver, is more common than
proustite [see below] and is an important ore of silver; a deeper red than proustite and less sensitive to light; in
mining lore, high grade pyrargyrite ore is known as “blood mining” in reference to the color and texture of a
freshly excavated face’. The presence of compound silver sulphide salts together with other metals would lead
them to be branded as rebellious, since they did not respond in a straightforward manner to amalgamation or
smelting. Another sulphide hypogene silver mineral is proustite, Ag3AsS3, which contains both silver and arsenic.
‘Known as ruby silvers due to their translucent red colour when fresh, lighter than pyrargyrite. perhaps the most
vivid color in all the mineral kingdom is the scarlet-vermilion of proustite ... [though] exposure to light darkens
it’. Wallace, Barton, and Wilson, "Silver-Bearing Minerals," 29.
83
Humboldt identifies by name silver sulphides, horn silver (silver chloride) and antimony/arsenic compounds of
silver among the main silver ores of Mexico, in Humboldt, Essai politique, Tome III, 354-61.; by mid nineteenth
century silver chloride is identified as one of the main components in pacos or colorados, together with
descriptions of silver and antimony sulphides, among others in Edward Pique, A Practical Treatise on the
Chemistry of Gold, Silver, Quicksilver and Lead, Tracing the Crude Ores from the Mines Through the Various
Mechanical and Metallurgic Elaborations, Until the Pure Metal is Obtained (San Francisco: Towne & Bacon,
printers, 1860), 81-84.; by the turn of the century Emmons was publishing his research on the chemistry of silver
ore deposits.
79
good silver ores from silver sulphide deposits, since the absence of adequate refining skills
resulted in the complaint by Spanish miners of the apparent sudden onset of poor silver content
relatively soon after mines started to be exploited.84 The miners and authorities did not have
the knowledge base at the time to recognize that the real poverty lay not in the silver content
of the deeper ores but in the range of refining skills they could bring to bear on silver sulphide
ores during the early years of silver production.
1.4.2 Argentiferous galena
The other major source of silver in New Spain was argentiferous galena, which
contributed an important fraction of the close to 40% of silver that would be refined by smelting
(Chapter 6). These deposits were found in the region of San Luis Potosí, Durango, Sombrerete,
Chihuaha and Zimapán in New Spain. This was the silver compound familiar to silver refiners
in Europe up to the conquest of the New World that will be described in section 1.5. The
economic consequences of the weathering of deposits where galena is the main silver ore has
not received as much attention as the case for silver sulphides presented above.85
1.4.3 The silver belt in New Spain / Mexico
The silver belt of New Spain lies at the edge of the North American continental shelf,
and is traversed by the Volcanic Belt of Mexico (Figure 1-5). The Mexican metallurgist
84
In Chapter 3 I will comment in detail on this issue, since the early amalgamation process cut its teeth on the ore
recovered from mountains of tailings. In the mid eighteenth century women would still be combing through 200
year-old tailings in search of useful silver ore. Manuel Jose Dominguez de la Fuente, Leal Informe Politico-Legal
(Observaciones sobre la maniobra de las minas, hechas en el Real de Guanajuato en 1774) (Guanajuato, Mexico:
Ediciones La Rana, 1999), 79.
85
The literature on weathering of galena deposits is much more limited than for deposits of hypogene silver
sulphide. An early discussion that questioned the current opinion at the time that no great supergene enrichment
of galena deposits is observed can be followed in John Stafford Brown, "Supergene Sphalerite, Galena, and
Willemite at Balmat, New York," Economic Geology 31, no. 4 (1936). Formation of supergene silver chloride in
deposits such as Slocan, British Columbia, are generated from the fraction of silver sulphides also present in a
deposit that also contains galena. See the description of the genetic type model 22c, in Dennis P. Cox, "Descriptive
Model of Polymetallic Veins," in U.S. Geological Survey Bulletin 1693, ed. Dennis P. Cox and Donald A. Singer
(Washington: U.S. Geological Survey, 1987).
80
Guillermo Salas has classified the locations of deposits of Mexico into six metallogenic
provinces, of which two are of special interest to the history of silver refining. The first is the
Sierra Madre Oriental, an extension of the North American Rocky Mountains that is the link
between the Cordilleras of the north to the Andes of the south, where the main deposits contain
lead, silver (in galena or silver sulphide ores), zinc and copper. It encompasses the historic
mining districts of San Luis Potosí, Guanajuato, and Zacatecas. According to Salas this metal
province is rich in argentiferous lead, lead-zinc and lead deposits. The second is the Provincia
del Eje Neovolcanico Mexicano (also known as the Volcanic Belt of Mexico) which contains
the mining districts of Pachuca and Real del Monte and Taxco, also producing both silver (in
galena or silver sulphide deposits) and lead.86 The more recent monograph by Coll-Hurtado et
al mentions that gold-silver and argentiferous lead deposits are to be found in Zacualpan,
Sultepec and Taxco, as well as in Pachuca and Real del Monte, all areas historically known as
pioneers in the mining and refining of silver. Other argentiferous lead deposits are found in
Guanajuato, Zacatecas and Fresnillo.87
1.5 The chemistry of the sources of silver in Europe
The silver ore of the New World was to be refined on the basis of the accumulated
experience of Europe and Asia of mining and metallurgy dating back at least two thousand
years.88 To understand the challenge faced by the miners of the sixteenth century when
confronted for the first time with the type of silver ores found in the New World, it is necessary
86
Guillermo P. Salas, Carta y Provincias Metalogenéticas de la República Mexicana (Mexico: Consejo de
Recursos Minerales, 1980), 69-73. The work includes a detailed fold-out map of the regions.
87
Atlántida Coll-Hurtado, María Teresa Sánchez-Salazar, and Josefina Morales, La minería en México: geografía,
historia, economía y medio ambiente (Mexico: UNAM, 2002), 16-22. A metal province has been defined as an
area ‘characterized by an abnormal concentration of large deposits of a particular metal or metals, by numerous
occurrences of a metal, or both’ as distinct from the term metallogenic belt which ‘shows the types of deposit of
a metal or metals formed within a narrow time range, ideally not more than 10 to 20 million years ... or less’.
Mitchell and Garson, Global Tectonic Settings, 5-6.
88
An overview is presented in Ian Blanchard, Mining, Metallurgy and Minting in the Middle Ages. Continuing
Afro-American Supremacy 1250-1450. , vol. 3 (Munich: Franz Steiner Verlag Stuttgart, 2005).
81
to understand how the different geological roots on both sides of the Atlantic determined the
nature of the silver ore in the deposits. Subduction processes had also given rise to the Variscan
Sta Eu ESM
Pa
Pacific Ocean D
S F C Gulf of Mexico
Zac
C : Catorce SLP
D : Durango CP G Z
F : Fresnillo
P
G : Guanajuato
P: Pachuca
T MVB
Pa : Parral Lead, Galena
Sta Eu : Santa Eulalia
Peñon Blanco, salt
SLP : San Luis Potosí
T : Taxco
Zac : Zacatecas Boundary of Eastern Sierra Madre
(ESM) and Central Plateau (CP)
Z : Zimapán
Boundary of Mexican Volcanic Belt (MVB)
Figure 1-5. The main historical silver, lead and salt deposits of New Spain / Mexico (based
on footnote 86).
orogen, the geological process of rock deformation and mountain building in Europe that took
place during the Carboniferous to Permian, 359 to 252 Ma. It gave rise to silver-bearing
polymetallic ore deposits stretching from Devon/Cornwall to Spain and the modern day Czech
Republic.89 The first major difference on both sides of the Atlantic is that the European silver
ores were deposited over 250 Ma ago while the silver ores in the mines to be worked by the
89
‘Variscan and Hercyninan orogenies are essentially synonymous terms’ Robb, Ore-Forming, 110, 334-38.
82
Spaniards are as young as 12 Ma.90 The different metallogenic epochs led to a fundamental
change in the way silver ore deposits were generated, thus in their chemical make-up. This is
a critical difference that sustains the main arguments of this thesis, so I will quote at length on
this topic:
‘The principal historic source of silver in Precambrian and Paleozoic mineral deposits
[European ores] has been as a by-product or co-product of base metal ores. Most of this silver
was concentrated by syngenetic [concurrent] or diagenetic [transformation in time of existing
deposits] processes ... the principal historic source of silver in Mesozoic and Cenozoic mineral
deposits [New World] has been as a coproduct or major economic component of ores. Most or
all of this silver was epigenetically concentrated [ores deposited in cracks formed after the host
rocks were created] ... the increased abundance of relatively young deposits in which silver
is the principal economic component reflects a fundamental evolutionary change in the
abundance of silver in the crust and/or in the processes which concentrate silver into ore
[emphasis added]’.91
In other words, in Europe the main silver source known to generations of miners up to
the early sixteenth century were ores in which silver was secondary to metals such as lead or
copper. The first silver deposits to be exploited in Europe were argentiferous galena. The silver
within this lead ore has been reported as consisting of ‘minute crystals of the silver sulphide
… rather evenly spaced through massive lead sulphide, even when only a little silver is
present’.92 The silver content can range from 0.01% to over 1%. 93 These are the lead-silver
sources that funded the Athenian Empire, which provided the silver and lead of Rome, and
from the Middle Ages the silver of the Harz and some of the Erzgebirge mines of Germany, of
Kutna-Hora and the lead and silver of England.94 Lead would gift to European silver refiners
90
Laznicka, Giant Metallic Deposits, 150.; There is an interesting graph that shows the clear division in time
between the formation of the silver deposits at Erzgebirge and those of the Americas in Graybeal, Smith, and
Vikre, "Geology Silver Deposits," 39.
91
"Geology Silver Deposits," 163-64.
92
Emmons, The Enrichment of Ore Deposits 625 367.
93
Pohl, Economic Geology, 195-97.
94
For example, in the sixteenth century Agricola wrote: ‘In the same region is found Goslar, where one finds so
much galena from which lead is extracted that one could say the whole mountain is made of lead’ - ‘Dans la même
région se trouve Goslar, ou l’on rencontre tant de galène dont on tire le plomb que l’on peut dire que toute cette
83
an in-built key for the extraction of metallic silver via smelting, but lead could also interfere
by amalgamating with mercury. These are the lead based silver ores that had served to generate
European mining and refining expertise on silver over many centuries.
When from the fifteenth century the argentiferous lead ores of Europe began to be
exhausted or required deeper mines subject to flooding, it was the expertise in a metallurgy
based on lead smelting that was now adapted to extract secondary silver from copper ores found
at Erzgebirge (Joachimsthal), in the Tyrol (Schwaz) and in Hungary (Neusohl). This was the
new generation of silver bearing ores that would make the Fuggers an extremely rich banking
family based on their large scale approach to refining operations.95 The silver in the copper ore
is inserted in complex copper minerals such as tetrahedrite. It cannot be discerned as solid
discrete particles as in the case of argentiferous galena.96 In both cases silver is not the primary
economic target in the ore, in total contrast to the silver-bearing ores found in the Spanish mines
of the New World. This secondary role of silver, and the fact metallurgy of silver in Europe
cut its teeth on argentiferous lead ores, is what separates the average European silver ore from
its counterpart in the New World.
Figure 1-6 situates the main silver historic mining areas within Europe. The earliest
mining in Central Europe took place in the Harz region around Goslar, including Rammelsberg
and Freiberg, known for their argentiferous galena as source of its silver. The Erzgebirge, the
Ore Mountain region straddling Germany and the modern day Czech republic, rose to
prominence as a major source of silver rich copper ores such as made Joachimstahl (present
montagne n’est que galène.’ Georgius Agricola, Bermannus, trans. Robert Halleux and Albert Yans (Paris: Belles
Lettres, 1990), 18.
95
J.U. Nef, "Silver Production in Central Europe, 1450-1618," The Journal of Political Economy (1941): 578-85.
The evolution of silver refining techniques in Europe will be treated at greater depth in Chapter 2. The pivotal
role of the Fuggers in the history of silver refining in the New World will be discussed in Chapter 5.
96
Gasparrini, "The Mineralogy of Silver," 99-100.
84
day Jachymov) famous; it is the only European silver deposit to figure in Table I. The English
deposits of Devon and Cornwall have been more important for tin and lead than for silver but
played a historic role in supplying lead to the silver smelters of Europe and New Spain.97 The
copper ores that contained silver of the Hungarian mines at Neusohl would supply the major
silver and copper refining centres units set up by the Fuggers both at Neusohl (Hungary) and
Rammelsberg
Freiberg Goslar
(Ag/Pb) Pb
Thuringia Pb
Tarnowitz
Joachimsthal
Devon and
Cornwall Neusohl
(Ag/Pb)
Tyrol Saigerhütten
Pb Villach (Ag/Cu)
Venice Pb Lead production
Rio Tinto
(Ag/Pb)
Figure 1-6. Location of main historic silver mining regions in Europe by mid sixteenth
century. Adapted from Blanchard in footnote 98. Map and inset not to scale.
also at Vilbach to process the ores from the Tyrol region.98 Finally the Rio Tinto and
Guadalcanal mines in Spain were the sources of jarositic and argentiferous galena ores.99
97
Mitchell and Garson, Global Tectonic Settings, 280.; Stephen Rippon, Peter Claughton, and Chris Smart, Mining
in a Medieval Landscape: The Royal Silver Mines of the Tamar Valley (Exeter: Univ of Exeter Press, 2009), 13-
52.
98
Ian Blanchard, "England and the International Bullion Crisis of the 1550s," in Precious Metals in the Age of
Expansion: Papers of the XIVth International Congress of the Historical Sciences ed. Hermann Kellenbenz
(Stuttgart: Klett-Cotta, 1981), 87-93.; Nef, "Silver in Europe," 584.
99
On the composition of silver ores at Rio Tinto: ‘Jarosite is the lead-silver equivalent to fahl, by degradation of
pyrites, formed a potassium iron sulphate rich … and is found in the Rio Tinto silver deposits at Huelva, Spain,
formed at junction between weathered pyrites and primary pyrites’. P. T. Craddock, Early Metal Mining and
Production (Washington, D.C.: Smithsonian Institution Press, 1995), 29. South-west Spain is the only place where
silver has been extracted from jarositic earths according to Leonard U. Salkield, "Ancient Slags in the South West
85
The map is important because the history of silver refining in New Spain and then
Mexico is woven from threads of technical experience spooled from each of these areas.
German smelting know-how based on lead would migrate from the mines of MittelEuropa to
New Spain in the early sixteenth century, as I will narrate in Chapter 2. Venice will play a
leading role in the development of the amalgamation refining process in Europe, as I will
explain in Chapter 3. Cornish miners expert in dressing tin ores and English smelters raised on
lead ores will immigrate to Mexico in the nineteenth century, where I will catch up with them
in Chapters 4 and 5. The Fuggers will lend unwillingly their considerable wealth obtained from
smelting European argentiferous copper ores to the initial supply of mercury to New Spain, as
I will argue in Chapter 5.
Figure 1-7 visually summarizes the differences in the nature of ore deposits on both
sides of the Atlantic. The geological youth of the New World silver deposits is reflected in the
altitude at which they are found, compared to the historical silver deposits of Europe. 100 In
of the Iberian Peninsula," in La minería hispana e iberoamericana. Ponencias del I coloquio internacional sobre
historia de la minería (León: Cátedra de San Isidro, 1970), 94.
100
Figure 1-7 is a simplified picture of complex geological forces at work. For example, the silver deposits at
Cobalt (Ontario) were unrelated to subduction processes operating on the western seaboard of the continent.
Altitude is a fickle guide, since the Himalayas are quite barren of silver deposits. Aridity is a condition that can
change over millions of years, so current conditions may be quite different from those existing when the deposits
were formed (see Laznicka, Giant Metallic Deposits, 172.) Sources: treeline range, Christian Körner, "A Re-
Assessment of High Elevation Treeline Positions and Their Explanation," Oecologia 115, no. 4 (1998): 446-47.;
maximum altitude Potosí (Cunill Grau, "Paisaje andino," 79.); Porco (C.G. Cunningham et al., "Relationship
between the Porco, Bolivia, Ag-Zn-Pb-Sn deposit and the Porco caldera," Economic Geology 89, no. 8 (1994):
1833.); Pasco (Laznicka, Giant Metallic Deposits, 127.); Oruro (José De Mesa and Teresa Gisbert, "Oruro. Origen
de una villa minera.," in La minería hispana e iberoamericana. Ponencias del I coloquio internacional sobre
historia de la minería.(León: Cátedra de San Isidro, 1970).); Zacatecas (Burkart, "Mines de Veta-Grande," 60.);
Real del Monte (José J. Galindo y R, El distrito minero Pachuca-Real del Monte ([Pachuca?]: Cia. de Real del
Monte y Pachuca, 1957), 2.); Catorce (Rafael Montejano y Aguiñaga, El Real de Minas de la Purísima
Concepción de los Catorce, SLP (San Luis Potosi: Editorial Universitaria Potosina, 1993), 169.); Cerro San Pedro
(Alvaro Sánchez-Crispín, Eurosia Carrascal, and Alejandrina de Sicilia Muñoz, "De la minería al turismo: Real
de Catorce y Cerro de San Pedro, México. Una interpretatión geográfico-económica," Revista Geográfica, no.
119 (1994): 85.); Guanajuato (Yann René Ramos-Arroyo, Rosa María Prol-Ledesma, and Christina Siebe-
Grabach, "Características geológicas y mineralógicas e historia de extracción del Distrito de Guanajuato, México.
Posibles escenarios geoquímicos para los residuos mineros," Revista Mexicana de Ciencias Geológicas 21, no. 2
(2004): 273.); Parral (Robert C. West, The Mining Community in Northern New Spain: the Parral Mining District
(Berkeley and Los Angeles: University of California Press, 1949), 9.); central European mines (Humboldt, Essai
politique, Tome III, 333.).
86
practical terms this meant that erosion had not worn down and dispersed the topmost layers of
silver compounds, a task that would be accomplished by the Spaniards, who on the back of
indigenous labour in a few centuries achieved what would otherwise have taken place over
millions of years of geological
Arid to semi-arid or moderate to moderate

5,000
Local energy source limited
4,500 to shrubs and animal dung
4,000
3,500
approx. altitude (m)
expected tree line

3,000
2,500
2,000 Subject to rainfall levels

local forests are energy
source until depletion
1,500
1,000
500
Silver primary metal mainly as silver sulphide with Silver secondary metal
weathered zone of native silver and silver chloride within galena or copper
ores
Also found in argentiferous galena
On-going to recent geological activity (subduction) on Pangea-era subduction
average < 65 Ma on average > 250 Ma
Figure 1-7. Altitude of the main historical silver ore deposits found in the Hispanic New
World and in Europe. For sources see footnote 100.
87
time. It also meant that in the case of the Andean deposits, most were at or above the treeline,
which would have immediate consequences on the sourcing and pricing of fuel:
‘[Silver] is generated usually in sparse and sterile lands, in paramos and punas [Andean
highlands with very scarce vegetation, usually above the tree line] of great cold, hills and snowy
ranges ... the most highly regarded are the mines in high mountains and places, the mountains
with mines are bare, treeless, with no vegetation’.101
1.6 The silver content of New World ores
The current narrative on the history of colonial silver refining has been built upon the
notion that the average silver content of the ores in the New World was very poor, which in
turn is claimed to have limited their refining to just one viable technique, amalgamation.102 The
roots of this line of thought can be traced to the earliest years of silver refining in the New
World, when for example in the 1550s in New Spain it was claimed that: ‘the ore that only had
three marks [1.5 % by weight] was considered poor’.103 On the other hand claims can also be
found for a high level of silver in these ores, as stated in the early 1600s:
101
‘criase [la plata] de ordinario en tierras ásperas y estériles, en paramos y punas de riguroso frio, en cerros,
lomas y sierras nevadas … estímanse mas las minas de cerros y lugares altos.. son los cerros de minas rasos y
pelados, sin arboleda’. Cobo, Historia del Nuevo Mundo, 91, 141.
102
Among the most recent examples in the modern historiography are the following: ‘the vast quantities of ore
with a low to medium silver content’ - ‘las inmensas cantidades de mineral de ley baja y mediana’ in Hausberger,
"El universalismo científico del Barón Ignaz von Born y la transferencia de tecnología minera entre Hispano
américa y Alemania a finales del siglo XVIII," 607.; ‘low-yield ore, characteristic of South America’ in Lang,
"Silver Refining Technology in Spanish America (patio y fundición) " 140.; ‘the low [silver] content of the silver
deposits’-‘la baja ley de las menas de plata’, Peter Bakewell, "La transferencia de la tecnología y la minería
hispanoamericana, siglos XVI y XVII: algunas observaciones," in Hombres, técnica, plata : minería y sociedad
en Europa y América, siglos XVI-XIX, ed. Julio Sánchez Gómez, Guillermo Mira Delli-Zotti, and Francisco A.
Rubio Durán (Sevilla: Aconcagua Libros, 2000), 365. The major exception is Blanchard, who bases his analysis
on the different types of ore found on either side of the Atlantic, and not on any deemed difference in silver
content. He explains the adoption of refining methods not on a deemed poverty of the ores but rather on their lead
content. Blanchard, Russia's "Age of Silver". Precious-metal Production and Economic Growth in the Eighteenth
Century 3-31.
103
‘el que era de tres lo tenían por pobre’. Juan Suárez de Peralta, Tratado del descubrimiento de las Indias :
(noticias históricas de la Nueva España) (México, D.F.: Consejo Nacional para la Cultura y las Artes, 1990), 164.
88
‘we understand, that minerals are mined in the provinces of Germany and metals are refined
from them though with little benefit, and we have been told … that ours have much more silver
content’.104
In the early nineteenth century Humboldt stated that Mexican silver ores were deemed
much richer than European ores. However ‘it is not thus … by the intrinsic richness of the ores,
but rather by the great abundance in which they are found in the ground ... that distinguishes
the mines of America’.105 In the late nineteenth century Burkart would claim from his own
first-hand experience: ‘Mexican silver ores are not in general inferior in silver content to those
of other mines: on the contrary, they are richer than those from the majority of other mines and
locations in Europe’.106 Historical judgements of this nature need to be judged with care. Even
by the nineteenth century the assaying of silver ores in Mexico prior to refining was the
exception and not the rule.107 This problem of the absence of analytical information does not
seem confined to the New World, since Burkart complains of the difficulty of finding sufficient
data to calculate the silver content of European ores.108 In addition, sampling of large ore
masses was a major challenge, and silver content when reported was based on the silver that
could be extracted, which was never necessarily the silver actually present in the ore.109
104
‘habemos … entendido, de que en las provincias de Alemania se labran minerales y se benefician los metales
dellos que son de muy poco aprovechamiento, y nos han afirmado ser de la misma suerte que los negrillos que
tratamos y que estos tienen mucha mas ley y plata’ Marcos Jimenez de la Espada, Relaciones Geograficas de
Indias - Peru II, ed. Jose Urbano Martinez Carrera vol. 184, Biblioteca de Autores Espanoles (Madrid: Atlas, 1965
(1588)), 126.
105
‘ce n’est donc pas ... par la richesse intrinseque des minerais, c’est plutôt par la grande abondance dans
laquelle ils se trouvent au sein de la terre .. que les mines de l’Amerique se distinguent’ Humboldt, Essai politique,
Tome III, 371.
106
‘los minerales de plata mexicanos no son inferiores por cuanto a su ley tomada en general, a los de otras minas:
por el contrario, son más ricos que los de la mayor parte de las demás minas o distritos [de Europa]’. Johann
Burkart, "Memoria sobre la explotación de minas de los distritos de Pachuca y Real del Monte de México," Anales
de la Minería Mexicana (Revista de Minas) I (1861): 97.
107
Duport, Métaux précieux au Mexique, 138-39. By mid nineteenth century no assaying of ores was carried out
in Catorce, Sombrerete, Fresnillo, Zacatecas and Guanajuato, according to Laur, "De la metallurgie de l'argent au
Mexique," 182.
108
‘the majority [of European silver mines] do not possess the necessary data to calculate the silver content of the
ores’- ‘las mas de ellas no contienen los datos necesarios para calcular la ley de los frutos’ Burkart, "Memoria
Real del Monte," 92.
109
Laur, "De la metallurgie de l'argent au Mexique," 49. See Chapter 4 for examples of problems in sampling
with respect to data of silver content even by the late nineteenth century.
89
There are three ways to address this issue. Brading and Cross question the statements
on operational grounds:
‘various problems confront the historian who attempts to come to a closer view of colonial
refining... on many occasions both miners and royal officials claim that ore levels had fallen;
they then provide an average figure, let us say, of one ounce silver per hundredweight of
mineral. The historian then has to decide: did these refiners know how much silver their ores
really contained?’110
To their observation I would add the following question: could the inability to extract
silver from an ore that has suddenly changed its chemical structure so as to place it outside the
scope of the refining method being used have been misinterpreted as a sudden impoverishment
of the silver content of the ore?
Finally, what exactly is meant by a ‘low’ silver content? There is no atemporal and
absolute threshold that determines when an ore is low or high in silver.111 What exists is a
location-specific set of human skills and a total production cost in each historical period that
determine how much silver is extracted from a given ore, which in turn is a function of silver
content, chemical nature of the ore, variable and fixed costs of production, and the option to
market any metallic co-products from the refining process. Therefore the only number that is
relevant is the minimum value of extracted silver that is required to meet the cost of production,
a value that changes with technology, location and historical period. An orphaned number
designating a silver content, devoid of context, can mislead rather than assist in deciphering
the events that determined the environmental history of silver refining in the New World.
110
Brading and Cross, "Colonial Silver Mining: Mexico and Peru," 555.
111
To highlight how relative the terms ‘poor’ or ‘rich’ are in relation to the silver content of silver ores, in modern
times the typical range reported for silver ore deposits is between 10 to 1,000 grams of silver per ton, 0.001% to
0.1%, the latter being now considered a ‘rich’ silver ore. Wallace, Barton, and Wilson, "Silver-Bearing Minerals,"
18,30.
90
In view of the weight placed upon the notion of ‘poor ores’ in the historiography
(see Introduction) I have nevertheless reviewed a selection of the historical numerical data on
silver content as reported in primary and secondary sources, though the exercise has to be
interpreted with caution. First of all the source of much of this information is sometimes
unclear, whether it was hearsay, assumptions or actual assays. Second, there is no histogram
available up to the end of the nineteenth century that segments total quantities of ore in New
Spain or Mexico according to their silver content. At the most the data reported in the
historiography can be grouped approximately by period, and a strong aggregation of values
may be representative of the average of silver content during a certain period. Since they are
reported in a variety of units, I have calculated where necessary their value as a weight
percentage.112 The data plotted in Figures 1-8, 1-9 and 1-10, represent a non-exhaustive
sampling of the published values of silver content in ores from New Spain/Mexico, Peru and
Europe, excluding bonanza values (those above 2.5% silver content by weight).
2.5
silver content (%)
1.5
0.5
0
15 16 17 18 19
century
Figure 1-8. Plot of average silver content in ores of New Spain / Mexico grouped by
century, data from Table 1-II.
112
In modern geological texts the grade of an ore (silver content) is expressed as grams per ton.
91
2.5
2
silver content (%)
1.5
0.5
0
15 16 17 18 19
century
Figure 1-9. Plot of a selected range of data for the silver content in ores of the Vice-Royalty
of Peru / Bolivia, as per Table 1-III, grouped by century.
2.5
2
silver content (%)
1.5
0.5
0
16 17 18 19
century
Figure 1-10. Plot of published data for the silver content in ores of Europe, as per Table 1-
IV, grouped by century.
92
Tables 1-II and 1-III include the sources of the data for the New World.113 Table 1-IV
corresponds to the data for some mines of Europe.114 The statistical significance of each data
113
Sources: a) Miguel Othon Mendizábal, La minería y la metalurgia mexicanas (1520-1943) (Mexico: Centro
de Estudios Históricos del Movimiento Obrero, 1980). b) Jaime García Mendoza, "La administración de las minas
de plata y haciendas de beneficio de la familia Sandoval en Taxco (1562-1564)," in La plata en Iberoamérica:
Siglos XVI al XIX, ed. Jesus Paniagua Pérez and Nuria Salazar Simarro (Leon: Universidad de León, 2008). c)
Gonzalo Gómez de Cervantes, La vida económica y social de Nueva España al finalizar el siglo XVI (Mexico:
Antigua Librería Robredo de José Porrúa e hijos, 1944). d) Linda A. Newson, "Silver Mining in Colonial
Honduras," Revista de Historia de América, no. 97 (1984). e) Ciriaco Pérez Bustamante, "Las minas en los grandes
geógrafos del periodo hispánico," in La minería hispana e iberoamericana. Ponencias del I coloquio
internacional sobre historia de la minería (León: Cátedra de San Isidro, 1970). f) José Antonio Fabry,
Compendiosa demostracion de los crecidos adelantamientos, que pudiera lograr la real haciencda de su Magestad
mediante la rebaja en el precio del azogue que se consume para el laborio de las minas de este reyno ... con una
previa impugnacion à las reflexiones del contador Joseph de Villa-señor y Sanchez ... Añadese un breve modo de
reducir, ligar, y alear el oro, y la plata à la ley de 22. quilates (Mexico: Impressa por la viuda de J.B. de Hogal,
1743). g) Dominguez de la Fuente, Leal Informe Politico-Legal. h) Garner, "Long-Term Silver Mining."i) D. A.
Brading, "Mexican Silver-Mining in the Eighteenth Century: The Revival of Zacatecas," The Hispanic American
Historical Review 50, no. 4 (1970). j) Enrique Tandeter, "Forced and Free Labour in Late Colonial Potosi," Past
& Present, no. 93 (1981). k) D.A. Brading, Miners and Merchants in Bourbon Mexico 1763-1810 (London:
Cambridge University Press, 1971). l) Bargalló, Minería y metalurgia colonial. m) Joseph Garcés y Eguía, Nueva
teórica y práctica del beneficio de los metales de oro y plata por fundición y amalgamación que de órden del rey
nuestro Señor Don Carlos Quarto (Mexico: D. Mariano de Zúñiga y Ontiveros, 1802). n) Humboldt, Essai
politique. o) Clara Elena Suarez Arguello and Brígida Von Mentz, Epístolas y cuentas de la negociación minera
de Vetagrande, Zacatecas, 1791-1794, 1806-1809 (Mexico: Centro de Investigaciones y Estudios Superiores en
Antropologia Social (CIESAS), 2008). p) Burkart, "Mines de Veta-Grande." q) Duport, Métaux précieux au
Mexique. r) Laur, "De la metallurgie de l'argent au Mexique." s) T. Flores, Étude minier du district de Zacatecas
vol. 17 Guía de excursiones del X Congreso Internacional de Geología México (Xalapa: Institut Geologique
National, 1905). t) Figures 4-32 and 4-41, Chapter Five of this work u) John Arthur Phillips, The Mining and
Metallurgy of Gold and Silver (London: E. and F.N. Spon, 1867). v) Claude T. Rice, "The Silver-Lead Mines of
Santa Barbara, Mexico," The Engineering and Mining Journal LXXXVI, no. 5 (1908). w) Bordeaux, Mexique
mines d'argent. x) Benjamin Ponce and Kenneth F. Clark, "The Zacatecas Mining District; A Tertiary Caldera
Complex Associated with Precious and Base Metal Mineralization," Economic Geology 83 no. 8 (1988). y)
Graybeal, Smith, and Vikre, "Geology Silver Deposits." A) Bargalló, Minería y metalurgia colonial. B)
Gwendolyn Ballantine Cobb, "Supply and Transportation for the Potosí Mines.," The Hispanic American
Historical Review 29, no. 1 (1949). C) S.E. Ramírez, "La minería y la metalurgia nativa en el norte peruano (siglos
XVI-XVII)," Anuario de estudios americanos 64, no. 1 (2007). D) W. E. Wilson and A. Petrov, "Famous Mineral
Localities: Cerro Rico de Potosi, Bolivia," Mineralogical Record 30 (1999).
E) Carlos Serrano Bravo, "Historia de la minería andina boliviana (siglos XVI al XX) "
http://www.unesco.org.uy/phi/biblioteca/archive/files/370d6afed30afdca14156f9b55e6a15e.pdf. F) Wallace,
Barton, and Wilson, "Silver-Bearing Minerals." G) Arthur F. Wendt, "The Potosi, Bolivia, Silver District,"
Transactions of the American Institute of Mining Engineers XIX (1891). H) María del Carmen Martínez Martínez,
"Plata y mineria en la correspondencia privada de las Indias," in Ophir en las Indias. Estudios sobre la plata
americana. Siglos XVI-XIX, ed. Jesus Paniagua Perez and Nuria Salazar Simarro (Leon: Universidad de Leon,
2010), 30. I) Luis Capoche, "Relación General de la Villa Imperial de Potosí," in Biblioteca de Autores Españoles,
ed. Lewis Hanke (Madrid: Atlas, 1959). J) José de Acosta, Historia natural y moral de las Indias, ed. Fermín del
Pino (Madrid: Consejo Superior de Investigaciones Científicas, 2008). K) G. Arduz Eguía, Ensayos sobre la
historia de la minería altoperuana (Madrid: Editorial Paraninfo, 1985). L) Jimenez de la Espada, Relaciones
Geograficas de Indias - Peru II, 184, 126. M) CG Cunningham et al., "The Age and Thermal History of Cerro
Rico de Potosi, Bolivia," Mineralium Deposita 31 no. 5 (1996). N) De Mesa and Gisbert, "Oruro." O) Robins,
Mercury, Mining and Empire. P) John Fisher, "Silver Production in the Viceroyalty of Peru, 1776-1824," The
Hispanic American Historical Review 55, no. 1 (1975). Q) Newson, "Silver Mining Honduras." R) Rose Marie
Buechler, "Technical Aid to Upper Peru: The Nordenflicht Expedition," Journal of Latin American Studies 5, no.
1 (1973). S) Tandeter, "Labour in Potosi." T) Garner, "Long-Term Silver Mining." U) Mervyn F. Lang,
"Azoguería y amalgamación," Llull 22, no. 45 (1999). V) Castillo Martos, "Plata y revolución tecnológica."
93
point varies and cannot be quantified. The visual impression from the three sets of data points
to similar ranges of silver content on ores on both sides of the Atlantic, within the limitations
of the comparison, grouped around a value of 0.25% from the seventeenth century onwards.115
This is as much as can be said on this matter, without incurring the mistake that the
silver content of ores of the New World could be compared simply on absolute value with those
of the ores of Europe. As will become more evident in the subsequent chapters, a silver ore
containing 0.25% of silver in the presence of major amounts of lead or copper in Europe cannot
be compared on the same basis as a silver ore containing 0.25% of silver in the form of silver
sulphide in New Spain. The only straightforward conclusion from this analysis is that the
concept of a ‘poor’ silver ore is too simple and devoid of context to be of use in the analysis of
114
Sources: a) George Papadimitriou, "Mining and Metallurgical Activities in Ancient Laurium and its Impact on
the Golden Era of Athens " in 5th International Mining History Congress, ed. James E. Fell, P. D. Nicolaou, and
G. D. Xydous (Milos Island: Milos Conference Center-George Eliopoulos, 2001). b) Wallace, Barton, and Wilson,
"Silver-Bearing Minerals." c) R. F. Tylecote, "Roman Lead Working in Britain," The British Journal for the
History of Science 2, no. 1 (1964). d) Blanchard, Mining, Metallurgy and Minting in the Middle Ages, 3 e) Lynn
Willies, "Introduction" (paper presented at the Boles and Smeltmills: Report of a seminar on the History and
Archaeology of Lead Smelting, Reeth, Yorkshire, United Kingdom, 15 to 17 May 1992). f) P. Braunstein,
"Innovations in Mining and Metal Production in Europe in the Late Middle Ages," Journal of European Economic
History 12 (1983). g) William Jacob, An Historical Inquiry into the Production and Consumption of the Precious
Metals, vol. 1 (New York: Augustus M. Kelly, 1968; repr., 1831). h) Georgius Agricola, De re metallica, trans.
Herbert Clark Hoover and Lou Henry Hoover (New York: Dover Publications, 1950). i) Hermann Kellenbenz,
"Final Remarks: Production and Trade of Gold, Silver, Copper and Lead from 1450 to 1750," in Precious Metals
in the Age of Expansion: Papers of the XIVth International Congress of the Historical Sciences, ed. Hermann
Kellenbenz (Stuttgart: Klett-Cotta, 1981). j) Christoph Bartels, "The Production of Silver, Copper and Lead in the
Harz Mountains from Late Medieval Times to the Onset of Industrialization," in Materials and Expertise in Early
Modern Europe : Between Market and Laboratory, ed. Ursula Klein and E. C. Spary (Chicago: University of
Chicago Press, 2010). k) Arduz Eguía, Minería altoperuana. l) Mikuláš Teich, "Born's Amalgamation Process
and the International Metallurgic Gathering at Skleno in 1786," Annals of Science 32, no. 4 (1975). m) Antoine-
Marie Héron de Villefosse, De la richesse minérale considérations sur les mines, usines et salines des différens
états, et particulièrement du Royaume de Westphalie, pris pour terme de comparaison (Paris: Levrault, 1810). n)
Louis Edouard Rivot, Description des gites métallifères, de la preparation mecanique et du traitement
metallurgique des minerais de plomb argentiferes de Pontgibaud (Paris1851). o) Principes généraux du
traitement des minerais métalliques traité de métallurgie théorique et pratique (Paris: Dalmont et Dunod, 1859).
115
According to Blanchard, by the sixteenth century the silver content in European ores was below 0.2%: ‘During
the first [silver] long-cycle [1125-1255] output had peaked on the basis of argentiferous lead ores containing in
excess of 100 oz of silver/ton [this is equivalent to 0.28% silver]. During the second long-cycle [1250-1392/1412]
the best quality ores contained no more than 40 oz [0.11%] of silver per ton. Finally during the third long-cycle
[1425-1560] whilst in the Balkans ores of 20-40 oz silver content were exploited, elsewhere 12 ½ oz [0.03%] ores,
the minimal amount which could be processed by the prevailing cupellation method were the norm’. Blanchard,
Mining, Metallurgy and Minting in the Middle Ages, 3 xvi.
94
arrobas de
oz / marks / mark / plata por pesos/ oz /
original units g/t kg/ton
quintal quintal arroba quintal de ton carga
New Spain tierra
conversion source
factor to % 0.0001 0.063 0.50 2.00 0.1 25.0 0.003 0.02
weight
% weight
mine location period
calculated
Zacatecas 1550s 5 to 15 10 to 30 a, 40
Zacatecas 1550s 50 2 a, 40
New Spain 1564 1 to 1.5 2 to 3 b, 48-49
New Spain 16c 0.5 1 c, 150
New Spain 16c 0.06 1 c, 151
Honduras 16c 0.5 to 0.6 8 to 10 d, 48

Honduras 16c 0.7 to 0.9 12 to 14 d, 48
Honduras 16c 2 to 3 4 to 6 d, 48
Honduras 16c 1.5 to 2 3 to 4 d, 48
Honduras 16c 0.24 4 d, 48
New Spain 1620s 8 to 10 16 to 20 e, 301
Zacatecas 17c and 18c 0.09 to 0.12 1.5 to 2 d, 48
New Spain 1740s 0.06 to 0.5 1 to 8 f, 1-5
Guanajuato 1774 over 0.08% g, 90-91
Honduras 18c 2 to 16 1 to 8 d,48
Honduras 18c 0.50 1 d,48
New Spain 18c 0.12 to 0.18 2 to 3 h,914
New Spain 18c 0.18 3 i, 668
New Spain mid to late 18c 0.15 0.3 j, 100
Zacatecas 1775 0.12 2 k, 672
Zacatecas 1801 0.19 l, 153-154
New Spain 1800s 0.18 to 0.24 3 to 4 m, 241
New Spain 1802 0.25 n, 144-146
Guanajuato 1800s 0.24 4 o, 413
Zacatecas 1808 0.4 20 p, 699
Zacatecas 1835 3 6 q, 81
Zacatecas 1835 1.5 3 1/6 q, 81
Zacatecas 1835 0.9 1 7/8 q, 81
Zacatecas 1835 0.21 3 1/2 q, 81
r, 83, 85, 236, 279-

Mexico 1840s 0.2 to 0.3 280, 351
Mexico 0.1 to 0.5 s, 57-62, 254
Zacatecas 1854-1868 0.15 1.48 t, 4
Zacatecas 1876 0.07 0.73 u, 4
Pachuca 1870-1888 0.19 v
Pachuca 1870-1888 1.8 v
Mexico 1860s 0.09 w, 471
Santa Barbara,
1900s 0.025 to 0.05 250-500 x, 208-209
near Parral
Zacatecas ~1910 0.21 70 y, 278
Real del Monte ~1910 0.12 1200 y, 275
others near 1000 to

~1910 0.10 to 0.15 y, 275
Pachuca 1500
Zacatecas 1980s 0.012 120 z, 1680
Pachuca 20c 0.018 180 aa, 137
Table 1-II. Silver content of ores reported for New Spain / Mexico. Sources are listed in
footnote 113.
95
pounds / ducats /
oz / marks / pesos / mark /
original unit g/t oz / ton weight % cajón (5000 marks / cajón hundred
quintal quintal quintal pound
Vice-Royalty Peru lbs) weight
source,
conversion factor
0.0001 0.063 0.50 0.003 1 0.06 51 0.02 0.01 0.05 page
to %
% weight
calculated
PotosÍ 1545 40 to 45 80 to 90 A, 241
PotosÍ early 1550s 49 100 B, 124
PotosÍ 1550s 49 1 C, 175
PotosÍ 1550s 24 480 D, 11
PotosÍ 1560s 2 E, 44
PotosÍ mid 16c 20 20 E, 43
PotosÍ mid 16c 25 25 F, 36
PotosÍ 1545-1572 25 25 G, 75
PotosÍ 1568 1 2 B, 124
PotosÍ 1573 2 4 H, 30
PotosÍ 1574 4 to 4.5 8 to 9 A, 241
PotosÍ 1580s 0.2 to 0.4 2 to 3 3a6 I, 122
200 to
PotosÍ late 16c 12 to 15 J, 109
250
PotosÍ late 16c 2 to 3 30 to 50 J, 109
PotosÍ late 16c 0.1 to 0.4 2 to 6 J, 109
PotosÍ late 16c 0.50 50 K, 50
PotosÍ 1588 1 2 L, 126

PotosÍ 16c 30 to 40 30 to 40 M, 374
PotosÍ 16c 5 500 M, 40
Oruro 1605 0.2 3 to 4 N, 563
Oruro 1605 0.4 6 to 8 N, 563
Oruro 1605 0.7 1.4 N, 564
PotosÍ 1600s 0.4 to 0.5 20 to 25 O, 81
PotosÍ 1600s 0.2 10 to 12.5 O, 81
PotosÍ 1600s 0.1 2.5 O, 81
PotosÍ 1607 0.1 1.5 A, 241
PotosÍ 1618 0.20 52 G, 76
PotosÍ 1636 0.10 50 G, 77
PotosÍ 1600 to 1630s 0.1 to 0.4 50 to 150 D, 12
PotosÍ 1600 to 1650 0.1 12 to 13 M, 51
PotosÍ 1660 0.40 150 G, 77
Cerro Pasco 1700s 0.1 10 to 12 P, 33
PotosÍ 1740 0.1 5 O, 81
PotosÍ 1700 to 1750 0.04 4 M, 51
Cerro Pasco 1700 to 1750 0.1 8 M, 110

800 to
Peru mid 17c 0.1 to 3 E, 59
30,000
Cerro Pasco 1790s 0.1 8 P, 33
PotosÍ late 18c 0.10 0.75 Q, 48

3 marks 1 ounce to 6
Peru late 18c 0.03 to 0.06 marks 2 ounces
R, 62
Cerro Pasco late 18c 0.06 to 0.25 6 to 25 P, 31

PotosÍ late 18c 0.04 to 0.06 0.08 to 0.12 S, 100
PotosÍ 18c 0.06 to 0.12 1 to 2 T, 914
New World 16c to 18c 0.12 to 0.24 2 to 4 U, 661
PotosÍ 16c to 18c 0.005 to 0.02 50 to 200 V, 83
Cerro Pasco 1820s 4 400 Q, 36
PotosÍ 1886 0.20 70 D, 13
PotosÍ late 19c 0.04 to 0.06 15 to 20 G, 89
PotosÍ late 19c 1.96 700 G, 91
PotosÍ late 19c 0.21 to 0.24 75 to 80 G, 102
Table 1-III. Silver content of ores in the Vice-Royalty of Peru. Sources are listed in footnote
113.
96
the environmental history of silver refining in the New World.
Silver deposits by themselves do not explain the dominance and choices exercised by
Spain over silver production. Geology would also provide the Spanish Crown with the
ownership of vast deposits of the three chemical substances that subsequent chapters will show
were indispensable for the refining of the silver ores: mercury, salt and lead.
pound
troy oz /
weight g / 100 silver / marks marks /
original units ounce / oz / ton 100lb
% kg talent of /cajon hundredweight
ton ore
Europe ore
conversion page
factor to % 0.0031 0.0029 1 0.001 0.06 1.76 0.01 0.5
weight
% weight
calculated
Laurion BCE 0.04 0.04 a, 34
Laurion BCE 0.12 to 0.26 40 to 90 b, 18
Laurion BCE 1.9 600 c, 26
average European
1125-1225 0.29 100 d, xv
silver ores
England 12 c 0.07-0.28 25-100 e, 1
average European 1250-

0.12 40 d, xv
silver ores 1392/1412
minimum quality
1425-1566 0.036 12.5 d, xv
European
Balkans 1425-1566 0.058 to 0.12 20 to 40 d, xv
Eisleben, Schwaz 1480 0.79 790 f, 589
Eisleben, Schwaz 1494 1.38 1380 f, 589
Austrian mines 1460s 0.12 to 1.32 2 to 22 g, 242
Goslar?
1520s 0.88 to 1.76 0.5 to 1 h, 36
Erzgebirge?
Europe 1526 0.52 1.053 i, 316
0.03 to
Rammelsberg 16c 0.03 to 0.07 j, 80
0.07
Rammelsberg 16c 10 10 j, 80
Cornwall/Devon <18c 0.4 130 c, 26
16 to
Freiberg, Saxony 1750s 0.16 to 0.625 k, 110
62.5
Slovakia 18c 0.09 to 0.12 1.5 to 2 l, 312, 314
0.01 to
Harz 1805 0.01 to 0.04 m, 102
0.04
Pontgibaud,
1840s 0.1 0.1 n, 193-197
France
o, 317-319,
Flintshire 1850s 0.03 0.03 386-388
Cornwall/Devon 19c 0.08 to 0.12 25 to 40 c, 26
Table 1-IV. Silver content of ores in Europe. Sources are listed in footnote 114.
97
1.7 Mercury
During the Early Modern Era there were only three sources of mercury in the world for
all practical purposes, of which the two most important ones would be under the direct control
of the Spanish Crown, one on each side of the Atlantic (see Figure 1-1). Nearly 80% of all the
mercury produced in this period was owned by Spain. The mercury mine in Almadén, Spain,
remains the doyen of the group even after over one thousand years of production. In Europe
only Idrija achieved important and sustained levels of exploitation, but always inferior to those
of Almadén. Spain would find in the New World the second most important source of mercury
of the world of that time at Huancavelica, in present day Peru. As Table 1-V shows, there were
no other known major deposits of mercury during this period, since China never materialized
as a viable supplier of mercury to the New World in spite of repeated attempts by the Spanish
authorities to find Chinese supplies.116
There is a sense of incredulity at the scale and variety of geological deposits at the
service of the Spanish Crown during this period. Not only had Spain conquered exclusive
access to the main silver depository on Earth, it was already in possession of a mercury deposit
that even modern geologists address in mystified terms:
‘Almadén ... is a most enigmatic mineral system ... [with an endowment of ca. 271 thousand
metric tons] it is the largest Hg [mercury] “supergiant” that stores close to 30-40% of mercury
of the world’s endowment. It is also the number 1 deposit in terms of geochemical magnitude
of accumulation, of all metals. Despite this, there is no satisfactory explanation where this
mercury had come from, and why it had accumulated in this geologically almost “normal”
setting’.117
116
Sources: a) S.M. Cargill, D.H. Root, and E.H. Bailey, "Resource Estimation from Historical Data: Mercury, A
Test Case," Mathematical Geology 12, no. 5 (1980). b) Lars D. Hylander and Markus Meili, "500 Years of
Mercury Production: Global Annual Inventory by Region until 2000 and Associated Emissions," Science of The
Total Environment 304, no. 1–3 (2003). c) Laznicka, Giant Metallic Deposits.
117
Giant Metallic Deposits, 356-58. A similar sentiment, ‘possibly the largest geochemical anomaly on the
planet’, is expressed in Cris M. Hall et al., "Dating of Alteration Episodes Related to Mercury Mineralization in
the Almadén district, Spain," Earth and Planetary Science Letters 148 no. 1-2 (1997): 287.
98
aggregate
production to 1800
production
production
ranking country location source
period
metric tons metric tons
Almadén 271,000 a, 492 1499-1975 a, 493

1 Spain
country total 304,666 b, 16 67,966 b, 15 1501-2000 b, 16
104,000 a, 492 1850-1975 a, 493
2 Italy Monte Amiata n/a
104,473 b, 16 1851-1982 b, 15
102,200 a, 492 1495-1975 a, 493
3 Slovenia Idrija 35,609 b, 15
107,311 b, 16 1501-1997 b, 16
Huancavelica 52,000 a, 492 50,321 b, 15 1566-1810? a, 493
4 Peru incl. Yanacocha
57,962 b, 16 n/a 1551-2000 b, 15
after 1993
New Almaden 38,000 a, 492 1847-1975 a, 493-94
5 United States New Idrija 20,000 a, 492 n/a 1854-1975 a, 493
country total 131,106 b, 16 1801-2000 b, 15
Ukraine Nikitovka 33,700 c, 359 n/a not indicated
Nikitovka,
6
ex- USSR Kyrgyzstan and 77,788 b, 16 n/a 1881-2000 b, 15
Tajikystan
7 China country total 45,978 b, 16 >5 b, 15 1601-2000 b, 15
8 Mexico country total 32,250 b, 16 n/a 1891-2000 b, 15
9 Africa continent total 17,212 b, 16 n/a 1921-2000 b, 15
Table 1-V. Major deposits of mercury through history. Sources indicated in footnote 116.
Up to 1977 the Almadén mine has produced about one-third of the world’s mercury,
with extraction commencing prior to the arrival of the Romans in Spain.118 Spain would be
further awarded in the New World with the mercury mine at Huancavelica, which would
provide the majority of the supply of mercury to the amalgamation processes carried out in the
Vice-Royalty of Peru until its exhaustion at the beginning of the nineteenth century.
118
Cargill, Root, and Bailey, "Resource Estimation from Historical Data: Mercury, A Test Case," 492.
99
1.8 Salt, sodium chloride (NaCl)
‘Without salt there was no silver. Without silver another would have been the history
of New Spain’.119 Salt was as critical an ingredient for the refining of the silver ores in the New
World as mercury. Yet again the size of the natural deposits of salt that were gifted by geology
to Spain at locations close to the main silver deposits could only have strengthened their belief
in Divine Intervention. The largest salt deposit in the Americas, Uyuni, lies in the Andes, and
the other nearby salt deposit of Yocalla was harvested by amalgamation refiners of Potosí, who
were thus spared the cost and logistics of bringing salt from the sea.120 While New Spain did
not have a single deposit of salt of the scale of the Andean Uyuni, Peñon Blanco (Figure 1-4)
among others satisfied the industrial demand for salt:
‘from the 16c to the 19c Southern and Central Mexico relied mainly on their own inland
resources, besides importing salt from either coast. Colima salt came to the altiplano via
Guadalajara and was sold as far away as Guanajuato... salt came from Campeche via Veracruz
and then packtrain to the capital and neighbouring mines. The salinas del Peñon Blanco
supplied all the mines from northern central Mexico up to Durango … Peñon Blanco would
also sell as south as Pachuca ... [in contrast to the monopoly on mercury] free market economy
regulated trade’.121
119
‘Sin sal no había plata. Sin plata, la historia de la Nueva España habría sido otra.’ Juan Carlos Reyes,
"Introducción," in La sal en México ed. Juan Carlos Reyes (Colima: Universidad de Colima,Consejo Nacional
para la Cultura y las Artes, 1995), vii.
120
Cunill Grau, "Paisaje andino," 77. The Uyuni salt deposits are clearly seen as a white patch on satellite images
taken from a height of 64,000 km as appear on Google Earth© images of the central Andes.
121
Ursula Ewald, The Mexican Salt Industry, 1560-1980 : A Study in Change (Stuttgart; New York: G. Fischer,
1985), 202-203.; Coll-Hurtado, Sánchez-Salazar, and Morales, La minería en México, 26. Apart from Peñon
Blanco, other salt mines are mentioned as belonging to Ocotlan, Piaxtla, Chila, Tehuacan, Cuzcatlan, Omitlan,
Chiautla, Acatlan, Jasco and Sinaloa, in Ewald, Mexican Salt Industry, 20.; Mendizábal, La mineria mexicana,
80.; Jaime J. Lacueva Muñoz, "Nueva Vizacaya y sus yacimientos minerales hasta el descubrimiento de San José
del Parral," in Ophir en las Indias. Estudios sobre la plata americana. Siglos XVI-XIX, ed. Jesús Paniagua Pérez
and Nuria Salazar Simarro (León: Universidad de León, 2010), 106. Salt came from many more smaller locations
than Peñon Blanco, and a good case study of local multiple sources of supply for the silver mines in the Taxco
area is given in Margarita Menegus Borneman, "Las comunidades productoras de sal y los mercados mineros: los
casos de Taxco y Temascaltepec," in Minería regional mexicana. Primera reunión de historiadores de la minería
latinoamericana, ed. Dolores Avila Herrera and Rina Ortiz (Mexico: Instituto Nacional de Antropología e
Historia, 1994).
100
1.9 Lead
It is important to insist on the fact that lead was present in important quantities in New
Spain, for geological reasons previously pointed out by Sillitoe. The argentiferous lead mining
districts are (among others) Sombrerete, Veta Grande in Zacatecas; Catorce, Matahuela in San
Luis Potosí; Taxco in Guerrero; Santa Eulalia, Hidalgo del Parral in Chihuahua; Temascaltepec
and Sultepec in Mexico, Zimapán close to Pachuca. Lead deposits are found in Chihuahua
(Santa Eulalia among others), Coahuila (Sierra Mojada), Nuevo León (Sombrerete, Ojo
Caliente) and Zacatecas.122 Mendizábal draws attention to the important fact that even though
it was known that lead deposits existed in New Spain no effort was made to develop them:
‘Lead ... was present in great abundance ... as well as in argentiferous galena … it seems that
lead was never the aim of any special exploitation, except in the mines of El Cardinal and
Zimapán (Lomo del Toro) in the state of Hidalgo, which produced some fifteen thousand cargas
annually (4,140 tons), sufficient quantity to meet the industrial demand for lead ... at the end
of the eighteenth century lead mines are exploited in Sultepec’.123
The location of the principal lead resources in the northern provinces that took longer
to pacify may have delayed their discovery (for example the deposits at Santa Eulalia were
only discovered in 1704) but Mendizábal points to an absence of clear directives from the
Spanish Crown to search for lead mines as diligently as for silver or mercury sources.
The lead endowment of Mexico as calculated to 1994 was of 11,062,988 t, just below
that of Germany at 12,150,180 t.124 Up to the Early Modern Era both Germany and England
122
Teodoro Flores, Yacimientos minerales de la República Mexicana, con algunos datos relativos a su producción
(México: Instituto Geológico de México, 1933), 18,28. Also in Coll-Hurtado, Sánchez-Salazar, and Morales, La
minería en México, 16.
123
‘[El] plomo … existía en gran abundancia … como en la composición de la galena argentifera y [aunque se
aplicaban los impuestos del rey desde el siglo XVI] parece ser que el plomo no fue objeto de explotación especial,
sino en las minas de El Cardonal y Zimapán (Lomo del To) en el estado de Hidalgo, que producían unas quince
mil cargas anuales (4140 toneladas), cantidad suficiente para las necesidades industriales del plomo … a finales
del siglo XVIII se explotan minas de plomo en Sultepec’. Mendizábal, La mineria mexicana, 71.
124
Singer, "Precious Metal Deposits," 94.
101
had based their refining of silver ores on the use of their endogenous lead deposits and the
application of smelting technology. From a geological point of view there was no lack of local
deposits of lead available to refiners in New Spain.
1.10 Final remarks
Portugal had enforced a no-sail zone down the Atlantic seaboard of Africa by the end
of the fifteenth century, all the while whetting the appetite of Europe by returning with African
gold loaded onto its ships. The only unopposed expansion route by sea for other European
powers lay to the West, and Spain was the first to back the leap into the unknown. As a result
of this decision by the Reyes Católicos, Spain stumbled onto a continent with an active and
extended subduction zone along all its western coastline. It colonized first the narrowest portion
of what is now North America, endowed with a very rich metallogenic zone, and would thread
the rest of its conquest and colonization along the metal-rich spine of the Andes. Spain thus
came to control a unique monopoly of mineral resources that would allow it to corner the
market of silver production for nearly three centuries, a geological-political conjunction that
has not been repeated for any other metal up to the present. It would only be in the nineteenth
century that the new United States of America would join Spain in profiting from the silver
and gold bonanza of its own politically controlled subduction area.
Spain would control in just 50 years the two major deposits of primary silver ore known
to man, together with the major mercury deposit of Huancavelica to complement its Spanish
mine of Almadén, the salt deposits of Uyuni and the polymetallic deposits that contained lead
and gold in New Spain. The diversity of this geological bonanza meant that Spain had at its
disposal all the raw materials needed to make possible the refining of the silver ores of the New
World. The term ‘silver ore’ however is too generic, and implies a uniformity that does not
exist. The silver ores of Europe, the Andes of the Vice-Royalty of Peru and the Silver Belt of
102
New Spain do not share an identical geological genesis, though all are the product of subduction
processes. The different roots in geological time may be the reason why silver ores found in
Europe were extracted in order to profit from their content of lead or copper, with silver as a
collateral benefit. In contrast, in the New World it was only the silver content that sustained all
mining and refining production costs, in the chemical form of silver sulphides or argentiferous
galena. This was the fundamental and relevant difference on both sides of the Atlantic, not the
silver content in each ore.
Once geology determined the embarrassment of mineral riches of the New World, it
was up to the Spanish miners and authorities to make a conscious choice as to which technical
route to follow in their pursuit of silver. The environmental history of the refining of silver in
the New World was not the unavoidable consequence of ores poor in silver content. Sillitoe’s
proposal that there is no major supergene enrichment above the water table in sulphidic silver
deposits, shores up the strong suspicion that the early decrease in the production of silver, as
mines clawed deeper into the mineral veins of the New World, was a technical issue unrelated
to silver content. I have argued in this chapter that it was the consequence of the change in the
chemical profile of these deposits, from superficial elemental silver and silver halides to a
deeper and much more intractable silver sulphide ore. The reasons why silver halides and
argentiferous lead were easy to refine for the early Spanish miners, yet silver sulphides posed
initially an insurmountable challenge, and the very distinct environmental consequences of the
refining method applied to each type of ore, now lead this narrative into the next two chapters.
103
2 The dry refining process: smelting of silver ores.
‘The most general and proper way, better adapted to the nature of metals, to separate
them from the earth and stones where they are raised so as to reduce them to their purity and
perfection … is through the fire of the furnaces, which to this end are called smelting ovens’.
Alvaro Alonso Barba, Arte de los metales (1637)125
‘in those times there had come from Castille and the islands many Spaniards poor and
greedy, curs hungry for wealth and slaves’. Bernal Díaz, Historia verdadera de la conquista
de Nueva España (ca 1568)126
2.1 Introduction
The Spanish priest Alvaro Alonso Barba (born 1569) is a singularity amongst the early
Spanish refiners of silver ores in the New World. He wrote the only extant metallurgical text
of the early period that is sourced in the practice of the New World, which provides a first-
hand guide to the mind-set and skill level of the time. He also proposed the last original refining
method for silver ores based on mercury, the cazo process that will be described in the
following chapter. As a historical figure he is firmly entrenched in the historiography on
amalgamation of silver ores in the Americas. Thus at first sight it might seem odd that he would
exalt smelting as the ‘most general and proper way’ to extract silver from its ores.127
In fact, the major part of Barba’s much cited book, Arte de los Metales, is dedicated to
the smelting of ores (approximately 68 pages of Books IV and V, out of a total of 198 pages),
longer than his discussion on what he terms ordinary amalgamation (Book II, 36 pages) or even
125
‘El modo más general, más propio, y más conforme a la naturaleza de los metales, para apartarlos de la tierra,
y piedras con que se crían, y reducirlos a la pureza y perfección … es mediante el fuego en los hornos, que para
este efecto se llaman de fundición’
126
‘en aquel tiempo vinieron de Castilla y de las islas muchos españoles pobres y de gran codiçia e caninos y
hambientos por aver riquezas y esclavos’.
127
Barba, Arte de los metales, 130. The metallurgical term ‘smelting’ is confusingly close to the word ‘melting’.
The latter implies the use of heat to change the physical state of a crystalline substance from solid to liquid. The
former is only applied to metallic ores, and is an operation that requires both heat and chemical reactions to bring
about the extraction of a metal from its ore. For a definition of smelting see Manuel Eissler, The Metallurgy of
Argentiferous Lead (London: Crosby Lockwood & Son, 1891), 33.
104
the description of his new cocimiento (cooking) process (25 pages of Book III). And yet forty
years before this manuscript was sent to Madrid for printing (1637), the practice of smelting
silver ores in New Spain had suffered a fate similar to Mark Twain’s news of his early demise:
‘and with respect to smelting, I say that it is very much forgotten since 35 years have passed
since it was last used, so if new ores with sufficient silver tenor for smelting were to be
discovered, no man would still be alive who knew how to smelt them, nor would there be a
smith to make the tools required for smelting; and there may come a time when this work may
be held of some interest and of great curiosity by the fact that it [knowledge on smelting] should
have been preserved in writing’.128
The author of this premature obituary for smelting was Gómez de Cervantes, a
commentator at the end of the sixteenth century on events unfolding in the New World, and
not a minero.129 He was claiming that smelting had long since disappeared in New Spain in the
face of the success of mercury amalgamation. The notion that amalgamation swept its way past
smelting to become virtually the sole refining process in the New World still percolates its way
through the modern historiography.130 Yet the view expressed at the time by Gomez de
Cervantes contains two important errors of fact. First, towards the end of the sixteenth century
not enough mercury was being imported into New Spain to amalgamate all its silver ores.
128
‘y en cuanto a la fundición, digo que esta tan olvidada por haber más de 35 años que no se usa, que entiendo
que si se descubriesen metales de ley, que se hubiesen fundir, no habría hombre que los supiese fundir, y aun
dudo si habría herrero que supiese hacer las herramientas de fundición; y quizás, vendrá tiempo que se tenga en
algo esta relación y por muy gran curiosidad esta prevención haberlo escrito’ in Gómez de Cervantes, Nueva
España siglo XVI, 156-157.
129
The term minero is applied in the early documents both to the person that owned and/or operated the mines
and the person who owned and/or operated the refining haciendas. It was only by the eighteenth century that the
business of refining could be regarded as separate from that of mining, with the introduction of the maquila or toll
(Chapter Five). The use of the masculine does not mean women were excluded from the business of mining and
refining, but they do not appear in documents as much as the men. The following extract shows an interesting
exception since women are involved on both sides of the business: ‘Doña Francisca de Paz minera of this village
[San Luis Potosí] declares before Your Eminence one hundred and fifty cargas of ore from my mines to be
processed in the refining hacienda of Doña Ysabel de Adriansen which she owns in Los Pozos’ - ‘Doña Francisca
de Paz minera en este pueblo manifiesto ante Vmd [Vuestra Merced] ciento y sinquenta cargas de metal de mis
minas para beneficiarlas en la hacienda de minas de doña Ysabel de adriansen que la susodicha tiene en los
posos’, AHSLP, Fondo de Alcaldía Mayor 1635.3, expediente 19, 11 July 1635. Many widows undersign
documents relating to sales or rental agreements of refining haciendas in New Spain.
130
For example, the recent claim that more than 95% of all silver was produced by amalgamation in Manuel
Castillo Martos, "Alquimia en la metalurgia de plata y oro en Europa y America " in Informes para obtener plata
y azoque en el mundo hispánico (Granada: Universidad de Granada, 2008), xxiv.
105
Second, not all of New Spain’s silver ores could be refined via the amalgamation process
applied at that time, since as indicated in the previous chapter argentiferous lead also
constituted an important part of the silver sources in New Spain. Very relevant in this regard
were the deposits of silver rich galena discovered towards the end of the sixteenth century in
the mines of the Cerro San Pedro close to the town of San Luis Potosí, towards the north of
Ciudad de México.
Barba is definitely one of the strong and credible voices of the late sixteenth century on
silver refining in the New World, and his evidence gives a contemporary lie to the report by
Gómez de Cervantes. The overall impression in Barba’s text is that amalgamation is never
taken for granted as being a better refining method than smelting. Time and again he cautions
his contemporary readers against being led astray by the limitations of mercury amalgamation,
which could not even be used to provide a true assay of silver in an ore. ‘Whoever deals with
ores without knowing how to assay them by fire to learn correctly their silver content ... runs a
great risk ... do not trust the assay by mercury, which is very deceitful’.131 Deceitful mercury
is certainly not a part of the mainstream historiography on colonial silver refining. Barba knew
first-hand that mercury amalgamation was not a process that would extract silver efficiently
from every ore. In his words:
‘a certain minero … extracted a lot of very rich ore, but did not realize this; he assayed it by
mercury [amalgamation] and measured four or five pesos per quintal [0.24 to 0.3%]… he
abandoned the mine, because he deemed it without profit ... [later I found the ore and] assayed
it by fire [smelting with lead] and it had nine hundred pesos per quintal [approx. 54%]’.132
131
‘Muy a riesgo esta … el que tratando en metales no supiere ensayarlos por fuego para enterarse con
certidumbre de la ley que tienen … no se fien del ensayo de Azogue, que es muy engañoso’, Barba, Arte de los
metales, 151.
132
‘cierto Minero ... saco cantidad de metal riquísimo, aunque no lo conoció; ensayolo por Azogue, a cuatro o
cinco pesos por quintal ... desamparo la Mina, porque no le era de provecho … [luego yo halle del metal y]
ensayelo por fuego, y tenía a novecientos pesos por quintal’. Ibid., 71. Metal is the word used in Spanish texts to
denote both the metallic element as well as the ore. Halleux provides the etymology of the Latin word metallum,
derived from the Greek metallon, which initially referred to the mine, the ‘underground cavity of extraction’.
After the first century CE it came to designate the minerals in the mine, the ores, whether metallic or not. It was
only at the end of Antiquity that it was applied to seven specific bodies: gold, silver, copper, iron, tin, lead, and
106
Barba’s windfall profits were the result of a confusion still found to the present day that
amalgamation of silver in ores was of the same nature as the amalgamation of gold. Mercury
amalgamation was one of the tested methods to assay gold ores in Europe in the sixteenth
century, and amalgamation did not require the skills of assaying with lead using a cupel and a
furnace.133 The clue to explain the great discrepancy between assaying with mercury or by
smelting can be deduced from the similar experience of another priest that was a friend of his:
‘In the Cerro de Santa Juana … ores like Soroche [galena, lead sulphide] were extracted, that
when assayed with mercury showed little or no silver; they were thrown away by the mineros
[until I assayed them by smelting] and found they had sixty or more pesos per quintal [3.6%
silver] ... on my advice he collected many [and] extracted much wealth from them’.134
The chemical explanation to both priests’ profit-taking is that lead is the main metal
present in galena or soroche [as it was called in the Andes] and it forms an amalgam with
mercury, thus competing and interfering with the extraction of silver. The silver extracted from
a soroche ore by amalgamation can therefore be a minor fraction of the real amount of silver
present. This is the reason why argentiferous lead was never refined efficiently by
amalgamation. The limitations of simple amalgamation also extend to non-lead silver ores rich
in silver sulphide or sulfo-salts, the negrillos. Mercury alone will not reduce silver sulphide to
silver metal, as I will explain in greater detail in the next chapter, so simple assaying of
either mercury or electrum. The concurrent use of metallum to designate both these seven bodies and also the ore
of a mine is the reason why the Spanish texts use the word ‘metal’ for both purposes as well. Robert Halleux, "La
nature et la formation des métaux selon Agricola et ses contemporains," Revue d'histoire des sciences (1974):
212.
133
L. Ercker, Treatise on Ores and Assaying, trans. Anneliese Grünhaldt Sisco and Cyril Stanley Smith (Chicago:
University of Chicago Press, 1951), 57-60, 96-97. Any of the metallurgical texts of this period provides details
on how to make cupels from crushed bones, hollow receptacles in which ores and lead could be melted in a
furnace.
134
‘En el Cerro de Santa Juana … se sacaban unos metales como Soroches, que por el ensaye ordinario de
Azogue mostraban ninguna o muy poca plata: echabanlos por ahí los Mineros [hasta que los ensaye al fuego] y
halle que tenían a sesenta, y mas pesos por quintal … recogió con mi aviso cantidad de ellos [y] mucha riqueza
de ellos saco’. Barba, Arte de los metales, 71. The fact there is an earlier tale in the historiography of a skilled
priest buying cheaply an ore wrongly classified as poor in silver and then refining its true worth to his profit
underlines the ubiquity of technically proficient Spanish priests with a good eye for wealth in the mining
landscape of the New World. See Juan de Peralta, late sixteenth century New Spain, as quoted in Henry R.
Wagner, "Early Silver Mining in New Spain," Revista de Historia de América, no. 14 (1942): 61-62.
107
negrillos by mercury amalgamation without the required chemical pre-treatment would again
have given false low values.
Smelting never disappeared as a major contributor to the production of silver in the
New World, regardless of reports to the contrary.135 The attention paid by Barba in his
metallurgical text on the technical merits of smelting indicates this was a topic of current
interest to the mining community from whose experience this text was sourced. The handful of
German, Siennese or Spanish authors who wrote texts on mining and refining in the sixteenth
and seventeenth century reflected the most relevant practices of the mining district from where
the authors drew their experience.136 A case in point is the fact that Agricola does not address
the technique of refining silver ores by amalgamation with mercury. He was writing about
techniques applicable to the silver ores found in the Erzgebirge and Harz mountains,
argentiferous copper or lead for which amalgamation was not an option.137 In contrast, both
Agricola and Ercker mention in detail the amalgamation of gold ores, a reflection on the
135
This perception is shared by the English historian Mervyn F. Lang: ‘All descriptions and accounts of South
American mining emphasise the amalgamation technique giving the impression that smelting was totally
discarded. This was not true’ in Lang, "Silver Refining Technology in Spanish America (patio y fundición) " 140.
Two further examples in-between Gómez de Cervantes and Castillo Martos of the notion that amalgamation
dominated over smelting are cited as follows. In his monograph promoting the use of a mineral additive
(tequesquite) to facilitate the smelting of silver ore, Garcés y Eguía pays the following compliment to
amalgamation: ‘the master key that has made possible the extraction of the prodigious sums of silver with which
the Americas have astounded the world’ - ‘la llave maestra que ha facilitado la extraccion de las prodigiosas
sumas de plata conque las Américas han asombrado al mundo’, in Garcés y Eguía, Nueva teórica del beneficio
de plata, 76. Sonneschmidt, a firm believer in amalgamation as practised in New Spain, said around the same turn
of the eighteenth century: ‘this refining by mercury … has produced the greatest part of the enormous quantity of
silver that is circulating in the world’ - ‘este beneficio por azogue … ha producido la mayor parte de la enorme
cantidad de plata que esta girando por el mundo’ in F.T. Sonneschmidt and J.M. de Fagoaga, Tratado de la
amalgamación de Nueva España (Galería de Bossange (padre), 1825), 160. In Chapter Six I provide quantitative
evidence as to the historical importance of smelting in New Spain, where it produced 40% of all the silver.
136
‘Agricola’s ... outlook is severely local to Germany and the topics considered are almost completely restricted
to the activities current within regions in and around the Harz mountains and the Erzgebirge... De Re Metallica
gives a picture of the best practices of the age, practices that had made Germany lead Europe in non-ferrous
metallurgy and which caused the services of German metallists to be sought by the rulers of many other countries’.
Leslie Aitchison, History of Metals (London: Macdonald & Evans, 1960), 373.
137
An alternate interpretation advanced in the historiography is that Agricola’s silence on mercury amalgamation
of silver ores proves that Biringuccio’s instructions for the amalgamation of silver ore were quickly forgotten and
never put in practice. Bargalló, Minería y metalurgia colonial, 111.
108
importance of this new process as of the fifteenth century in the Rhennish workings of
Europe.138 Biringuccio, a Siennese and not a German, is the only metallurgist and author of the
sixteenth century to mention in detail the use of mercury to refine silver ores by amalgamation,
in a book that was published in Venice, because he was addressing a Venetian mining audience
already conversant with the amalgamation of silver ores, as I will explain in the following
chapter.
2.2 Smelting of silver ores : the human context.
The early history of smelting silver in the New World is defined by the profile of the
Spanish actors who arrived in the newly conquered lands in search of a wealth they could never
attain back in their homeland. When Spain reached the New World it turned its attention very
early to mining activities. It is claimed that some ten percent of the 1,500 strong contingent that
came on the second voyage of Columbus was made up of ‘workers ... to take gold out of the
mines’ on Hispaniola.139 Precious metals were not the only target:
‘On September 18, 1505, Fernando, having heard good reports about the possibility of finding
copper in La Española, dispatched three caravels from Seville with all equipment needed for
such an enterprise. He sent not only equipment, but a hundred African slaves’.
Five years later the King would add another 250 slaves destined to mining. It was
mining, and not sugar, that marked the start of African slavery to the New World.140
138
‘Agricola was the first writer to give a comprehensive account of the metallurgy of gold and his most extensive
recordings deal with that subject [including amalgamating gold with mercury]’. Aitchison, History of Metals, 385.
According to Cyril Stanley Smith ‘there are no books on metallurgy among the incunabula’. The most prominent
mining texts published in the sixteenth century and early seventeenth centuries are: the anonymous German
Probienbuchs (early sixteenth century), the Siennese Biringuccio’s De Pirotechnia (1540), the German Agricola’s
De Re Metallica (1556) and Ercker’s Treatise on Ores and Assaying (1574) and the Spaniard Barba’s Arte de los
Metales (1637). See the introduction to Biringuccio, The Pirotechnia, x-xix.
139
‘gente trabajadora ... para sacar oro de las minas’, as quoted in J.E. Pérez Sáenz de Urturi, "La minería
colonial americana bajo la dominación española," Boletín Millares Carlo, no. 7 (1985): 55.
140
Hugh Thomas, Rivers of Gold : The Rise of the Spanish Empire (New York: Random House, 2005), 256-257.
291.
109
The Spanish contingent that swarmed over New Spain on the heels of the conquest was
described by Bernal Diaz, one of the original band of conquistadores of New Spain under
Hernán Cortés, as ‘poor and greedy, curs hungry for wealth and slaves’.141 Oviedo describes
well the predatory, rags-to-riches mentality of many of these mineros:
‘And in particular, those in these parts that have no intention of remaining nor wish anything
from this land other than to ravish it and return to their homes, turn to trading or to the mines
... or any other activity that will allow them to get rich quickly and leave ... for most who are
here treat this land as a step-mother, even though many have fared much better here than in
their own motherland’.142
Part of the problem faced by Spain was the very small pool of home-grown talent from
where the early miners could be sourced, with no generational experience on any type of silver
ore. Two historical texts that refer to the early period, ca 1550, attest to the very low level of
skills prevalent at the time:
‘this way of extracting silver [smelting] was not learnt from the indians, nor did men go from
here who knew about it, because they did not know how to smelt, and they were also ignorant
of refining over a bed of ash … previously they used to disinter the dead and burn their bones,
so as to benefit from the ash alone to make the cupels in which they refined, and in a similar
manner there were other primitive actions that point to the ignorance of that time’.143
‘the Spaniards did not have the experience of the ancient asturianos [from Asturias], or
Portuguese, or Gallegos [from Galicia] gained in antiquity working the mines of the provinces
... of Spain, from where the Romans had taken great treasures’.144
141
‘pobres y de gran codiçia e caninos y hambientos por aver riquezas y esclavos’. Bernal Díaz del Castillo,
Historia verdadera de la conquista de la Nueva España : Manuscrito "Guatemala" (Mexico: Colegio de México
: Universidad Nacional Autónoma de México, 2005), 834.
142
‘Y en especial, los que en estas partes no tienen pensamiento de permanecer ni quieren desta tierra sino
desfructalla e volverse a sus patrias, danse a la mercaduria o a las minas … e a otras cosas con que presto
alleguen hacienda con que se vayan … porque los mas que por aca andan, tienen esta tierra por madrastra,
aunque a muchos hales ido muy mejor que en su propia madre’. Gonzalo Fernández de Oviedo y Valdés, Historia
general y natural de las Indias (Madrid: Ediciones Atlas, 1959), 80.
143
‘este género de sacar plata no se aprendió de los indios, ni de acá fueron hombres que lo supieran, porque no
sabían fundir, y también ignoraban el afinar sobre cendrada de ceniza … antes solían desenterrar a los muertos
y quemar los huesos, cuya ceniza sola decían que aprovechaban para hacer la capella en que afinaban, y
asimismo había otras rusticidades en que se conoce bien la ignorancia de aquel tiempo’ from a letter from Agustín
de Sotomayor to the King, dated 20 April 1573, as quoted in Tomas González, Noticia histórica documentada de
las célebres minas de Guadalcanal, desde su descubrimiento en el año 1555, hasta que dejaron de labrarse por
cuenta de la Real Hacienda, vol. II (Madrid 1831), 409.
144
‘los españoles no tenian aquella esperiencia de los antiguos asturianos, e lusitanos, e Gallegos tuvieron
antiguamente en este ejercicio de las minas en las provincias ... de España, de donde los romanos tan grandes
tesoros hobieron’ Fernández de Oviedo y Valdés, Historia natural de Indias, 47.
110
Even by the early seventeenth century Barba was sufficiently worried by the overall
lack of skills he observed in Peru to propose that only those who passed an official exam to
demonstrate their ability to assay ores by smelting should be allowed to refine silver ores, such
had been the waste incurred by ignorant refiners in the past.145 Modern historiography shares
the judgement on the poverty of metallurgical skills:
‘the number of expert miners that were available was very small and their knowledge very
rudimentary, since except for the mines of Vizcaya, and those of mercury at Almadén,
exploited by the German bankers the Fuggers, there was no exploitation of important mines at
that time in Spain, given that the famous silver mines of Guadalcanal, in the Sierra Morena
[some 110 km north of Seville] were not discovered until 1555’.146
The Spanish historian of mining, Julio Sánchez Gómez, argues that from Roman times
up to the sixteenth century it became common practice in mining deposits of argentiferous lead
(galena) in Spain to only work the most easily accessible surface deposits, abandoning the mine
as soon as it required deeper levels of mining.147 By the middle of the sixteenth century the
contribution of Spanish mines, with the exception of Almadén, to the revenues of the Crown
was on average less than 0.4%, which gives an indication of how little mining and refining
know-how would have figured among its subjects.148 From a metallurgical point of view the
working of iron ores in Asturias and the Basque country (part of the Kingdom of Castille at the
145
Barba, Arte de los metales, 70-71.
146
‘el número de mineros expertos de que se disponía era muy pequeño, y sus conocimientos muy rudimentarios,
pues salvo las minas de Vizcaya, y las de mercurio de Almadén, explotadas por los banqueros alemanes Fugger,
no se trabajaban entonces minas importantes en España, dado que las famosas minas de plata de Guadalcanal,
en la Sierra Morena, se descubrieron hasta el año de 1555’. Mendizábal, La mineria mexicana, 19. Silver ore
from the mines at Guadalcanal was smelted, since it was argentiferous galena as is evidenced in the following
extract: ‘in this month [January 1559] three thousand six hundred and eighty eight marks of silver were extracted;
refining goes well, with the care that has been taken in assaying the lead’ - ‘en este mes [enero 1559] se saco tres
mil e seiscientos e ochenta e ocho marcos de plata; las afinaciones andan buenas, y con el cuidado que se ha
tenido de ensayarles el plomo’. González, Noticia histórica minas de Guadalcanal, II 17.
147
Sánchez Gómez, Minería no férrica en el Reino de Castilla, 47. The Mexican historian Mendizábal shares the
view that ‘due to the imperfection in mining techniques and metallurgy the miners abandoned their works at the
first sign of impoverishment [of the ore]’ - ‘debido a la imperfección de la técnica de laboreo y la metalurgias
los mineros abandonaban los trabajos al primer indicio de empobrecimiento’ in Mendizábal, La mineria
mexicana, 21.
148
Sánchez Gómez, Minería no férrica en el Reino de Castilla, 269.
111
time of the conquest of the New World) would have provided some background on furnaces
and smelting but little in the way of providing prior experience for the majority of silver ores
to be found in the New World.149 The technology employed in the iron works of the Basque
country is claimed to have been too simple to be of use in the refining of silver in the New
World.150
The Spanish colonial authorities recognized the technical deficiency in the field of
mining and refining silver ores and turned to German technical expertise for New Spain from
the 1520s and for Peru in the 1570s.151 ‘Tradition holds that the Germans or <alamans> had a
[mining and refining] know-how that was much sought after’. The term German encompassed
those originally from Germany but also included Swiss, Bohemians, Slovaks, Austrians and
other Middle Europeans.152
‘at the forefront of [metal refining] were the German master miners. Having gained experience
in the Harz and Saxony, these men travelled to the new mineral fields across Europe,
transferring knowledge and skills ... their skills in operating smelters was much sought after’.153
149
The role of Basque families as owners of mining investments in Zacatecas has been described in E. Fernández
de Pinedo y Fernández, "Influencias recíprocas de las técnicas extractivas entre la minería vasca y la americana
en la Edad Moderna," Areas, no. 16 (1994): 36. The Basque (Vascongados) fraction in Potosí has been made
famous thanks to the description of their ongoing fights in Bartolomé Arzáns de Orsúa y Vela, Historia de la villa
imperial de Potosí, ed. Lewis Hanke and Gunnar Mendoza L (Providence: Brown University Press, 1965).
150
Fernández de Pinedo y Fernández, "La minería vasca y la americana," 38. English language historians are
equally dismissive, see Peter J. Bakewell, "Mining in Colonial Spanish America," in The Cambridge History of
Latin America, ed. Leslie Bethell (Cambridge: Cambridge University Press, 1984), 111.;R.C. West, "Early Silver
Mining in New Spain, 1531-1555," in In Quest of Mineral Wealth. Aboriginal and Colonial Mining and
Metallurgy in Spanish America. , ed. Alan Craig and Robert C. West (Baton Rouge: Geoscience Publishers, 1994),
122.
151
Instructions to Toledo as cited in C. Salazar-Soler, "Innovaciones técnicas, mestizajes y formas de trabajo en
Potosí de los siglos XVI y XVII," in O trabahlo mestigo. Maneiras de pensar e formas de viver séculos XVI a
XIX, ed. Eduardo Franca Paiva and Carla Maria Junho Anastasia (Sao Paulo: UFMG, 2002), 149. The authorities
in Madrid did not hesitate in turning to whichever foreign expertise they considered could assist the mining and
refining effort in the New World, since in the next chapter I will refer to similar instructions being given in the
case of amalgamation.
152
‘La tradition veut que les Allemands ou <alamans> aient des compétences très recherchées. Sous ce terme
générique, on désigne des personnages originaires d’Allemagne, mais aussi de la grande province germanique
incluant la Suisse, la Bohème, la Slovaquie, l’Autriche, etc.’ Marie-Christine Bailly-Maître, L'argent : du minerai
au pouvoir dans la France médiévale (Paris: Picard, 2002), 143.
153
Martin Lynch, Mining in World History (London: Reaktion Books, 2002), 17. For similar references to the
migration of German skilled miners within Europe see John U. Nef, The Conquest of the Material World (Chicago:
University of Chicago Press, 1964), 12.
112
From the thirteenth to the seventeenth century German miners would be employed in
England, France, Sardinia, Norway, and in the Serbian Brskvo mines, among others.154 The
presence of German miners in the New World came about either through individuals
circumventing the initial restrictions on emigration for non-Castillians, or later in the sixteenth
century as part of technical teams sent by either the Welsers (1529) to assist in the extraction
of gold, or by the Fuggers at the request of the Spanish authorities, once travel was allowed
after November 15, 1526 to subjects of the Holy Roman Empire under Charles V (Germans,
Flemish).155
The need for technical assistance in smelting the ores of New Spain arose quite soon
after mines started to be exploited. According to a document dated June 6th 1571 by the Cabildo
de México to the Concejo de Indias, less than ten years had passed between the start of mining
in 1532 and the observation that by 1542 ores had declined in silver content and in ease of
smelting, ‘minas comenzaron a perder la ley y la Buena fundición’. This is a very significant
pairing of concurrent events in a sixteenth century document that strengthens the line of
argument that the difficulty in smelting the deeper silver sulphides triggered the conclusion
that silver content had suddenly decreased substantially. Then, according to this account, a
German dubbed with the generic name of Juan Alemán advised the Vice-Roy on German
154
German miners in Sardinia and Serbia, Lynch, Mining in World History 17.; in England, Jacob, An Historical
Inquiry into the Production and Consumption of the Precious Metals, 1 292.; in Norway, Wallace, Barton, and
Wilson, "Silver-Bearing Minerals," 32.
155
Demetrio Ramos, "Ordenación de la minería en Hispanoamérica durante la época provincial (siglos XVI, XVII
y XVIII)," in La minería hispana e iberoamericana. Ponencias del I coloquio internacional sobre historia de la
minería. Cátedra de San Isidro (León: Cátedra de San Isidro 1970), 381-382.; Hugh Thomas, The Golden Empire.
Spain, Charles V, and the creation of America (New York: Random House, 2010), 113. The ill-fated expedition
of eighty miners sent by the Welsers came mainly from the Erzegirbe area, where the main mining activity was
focused on argentiferous copper ores that were being refined by the liquation process with lead in the early
sixteenth century. All except a few would die in the New World, with little support for their endeavours. Juan
Friede, "La introducción de mineros alemanes en América por la compañía Welser de Augsburgo," Revista de
Historia de América, no. 51 (1961): 99-104.
113
smelting techniques using lead and litharge with rich ores.156 Events seem to have taken place
even earlier. Bargalló places the arrival in New Spain of the German smelters Juan Enchel and
others in 1536, sent by the factors of the Fuggers in Seville,
‘with tools and techniques to smelt metals from silver mines that until then had not been
understood, and they set up grinding and refining facilities [most likely in Sultepec, according
to Icazbalceta], from where came great benefit to the republic and great service to your
Majesty’.157
The Germans were not the only source of smelting skills to aid the initial wave of
Spaniards. The technical role of African slaves within a smelting hacienda is an intriguing facet
to the history of smelting in the New World. They are mentioned as of the mid-sixteenth
century by Bartolomé de Medina: ‘And so I have seen how such ores are processed in many
places using greta and cendrada and with great cost to the owners of the mines and with great
risk to the life and health of those involved in their processing, both of indians and negroes’.158
The seventeenth century records of Zacatecas make frequent mention of the African slaves that
are sold or rented together with the physical assets of a refining hacienda. The level of skills
of some of these slaves is indicated by the following notations: ‘a black called domingo smelter
from the land of angola; another black called juanito … smelter from the land of angola’.159
Since all the slaves in the inventory are identified according to their place of origin in Africa,
most probably as an indicator of their behavioural and physical nature, the fact that the smelters
156
Henry R. Wagner, "Early Silver Mining in New Spain," ibid., no. 14 (1942).; Bargalló, Minería y metalurgia
colonial, 91. For a review on German miners in the New World see West, "Early Silver Mining," 16.
157
‘con aparejos e yndustria para fundir los metales de las mynas de plata que hasta entonces no se entendían, e
hyzieron yngenios de moler i fundir los metales (seguramente en Sultepec, según Icazbalceta) de donde se siguió
mucho provecho para la rrepublica y gran servizio a Su Magestad’. At least one Spaniard, Alonso Carreño, is
said to have established an ‘ingenio de fundicion’, a smelting refinery, in Sultepec by 1543. Bargalló, Minería y
metalurgia colonial, 58, 95.
158
‘Y así he visto como se benefician los dichos metales en muchas partes con greta y cendrada y la muy grande
costa de los dueños de las minas y el mucho riesgo de las vidas y salud de los que en el beneficio de elas entienden,
así de indios como de negros’ as quoted in Castillo Martos, Bartolomé de Medina, 112.
159
‘otro negro llamado domingo afinador de tierra de angola; otro negro llamado juanillo … afinador de tierra
de angola’ in a rental agreement for a smelting and refining hacienda between Pedro de Medina and Andrés
Pereira, 20 March 1608, AHEZ, Notaria-Colonia, Numero 01 (Pedro Venegas, 1608), expediente 1.
114
came from Angola may not necessarily indicate a premium being placed on the level of skills
originally brought by the slave. Slaves would persist in silver mining duties, and Mendizábal
has pointed out the great number of slaves that worked in refining haciendas of Zacatecas,
though according to the quote provided from the Bishop Alonso de la Mota y Escobar (1602)
the indigenous workforce was better skilled than the African slaves or the Spaniards for both
smelting and amalgamation.160
A picture emerges of a stage from early to mid sixteenth century with a very limited
resource of refining skills in the New World, given the possible exception of some literate
clergy who refined silver ores, together with pockets of German, Andean and African know-
how of smelting methods. Bargalló also refers to this period as a proto-smelting period in New
Spain where the lack of knowledge of German smelting techniques led to many trial and error
processes.161 In Chapter 3 I will return to this scenario, since it may have been an important
factor in the fast adoption of an alternate refining process, one more amenable to such a large
group of unskilled refiners of silver ores.162 The contrast with the evolution of the mining and
refining work-force of Europe remains to be studied. The technical complexity of the smelting
operations carried out on a large scale has been used to argue that it led in Europe to the
160
Mendizábal, La mineria mexicana, 32, 36.
161
‘el desconocimiento de dichas técnicas [las europeas alemanas de minería del siglo 16] hubieran evitado la
serie de tanteos de los primeros mineros que por cierto han merecido severas críticas … de cuantos han estudiado
ese periodo inicial de la metalurgia hispanoamericana’ Bargalló, Minería y metalurgia colonial, 23. It is
interesting to note the assertion that at the other end of the historical period studied in this thesis : ‘The dependence
on Spanish mining methods was so great on the American [United States of America] frontier that by the 1880s
mining men were literally relying on Spanish techniques hooked up to a steam engine’. The author argues that the
Spaniards had a strong mining and refining experience before arriving in the New World. Otis E. Young, "The
Spanish Tradition in Gold and Silver Mining," Arizona and the West 7, no. 4 (1965): 299. It can be counter argued
that the rich experience applied north of the Mexican border in the late nineteenth century was the product of three
centuries of experience in the New World, not from Europe.
162
Even by the nineteenth century in Europe the high level of skill required for an efficient use of smelting to
refine silver ores is remarked upon in the preface to Schlutter’s classic textbook by M. Hellot. He points out that
among the lead mines in France are those that give six ounces of silver per quintal [under 0.4%], ‘their smelting
and refining very amenable … that have been abandoned due to a lack of intelligence and [management]’ – ‘sa
fonte et l’affinage très-asaisés … qui ont été abandonnées faute d’intelligence & de conduit’. Christophe-Andre
Schlutter, De la fonte des mines, des fonderies, etc. , trans. M. Konig, vol. 2 (Paris: Jean-Thomas Herissant et
Jacques-Noel Pissot, 1853), xiii.
115
development of a high degree of organization and literacy within the mining work-force,
beginning as early as the eleventh century in Goslar (Harz Mountains). Bartels presents a
persuasive argument that since smelting involved careful and consistent control of many
process variables (temperature, exact knowledge of the silver content to be able to monitor the
extraction process, consistent batches prepared for each furnace charge) therefore:
‘assaying, precision measurement, data collection, the use of mathematics, attempts at

standardization, the writing of technical instructions, the writing of technical books to be
published – all this contributed to the practice of mining and metallurgy’.163
The fact that profit levels could not depend on contingent bonanza ores but on the
careful monitoring of all the process variables, required:
‘extensive written communication ... in mines, ore processing, and smelting works not only the
experts ... but even every foreman had to be able to write and to do basic calculations from the
mid-sixteenth century onwards. By the seventeenth century the vast majority of Harz miners
were literate, and it was no longer possible to enjoy respected status in the community without
being able to read and write’.164
Even when Spain found an alternative refining route in the New World, it still applied
smelting in a major way in many mining districts, as will be detailed in Chapter 6. The level of
smelting skills of the indigenous workforce or the African slaves, and their technical hierarchy
within a smelting hacienda of New Spain is as yet an unexplored topic, and Bartel’s
conclusions raise the question if a similar path of development was ever observed at any level
for the smelting labour force of the silver refining industry of the New World.
One final topic is the size of the work-force required for smelting, and whether this was
a factor that was adversely affected by the epidemics that ravaged the indigenous communities
after the arrival of the European diseases. Berthe has argued that a major reason why the Crown
was looking so enthusiastically for an alternative to the tried and true smelting method was the
163
Bartels, "Production of Silver in Harz Mountains," 73, 98.
164
Ibid., 98-99.
116
shortage of manpower after the epidemics among the indigenous population in 1545.165
Medina in his promotional report on the advantages of amalgamation does focus expressly in
the mid-sixteenth century in New Spain precisely on the size of the workforce of the rival
process to amalgamation, claiming that :
‘since [for] a horse-powered ingenio [smelting hacienda] working with one sound furnace …
it is necessary to have four smelters and four carriers and two Spaniards that will need rooms
and people to handle the horses of the ingenio and their rooms and two to refine and to grind
the greta and cendrada another two persons and to make the furnaces and work the stones
another two and to work the cendradas for every time they refine six persons are required
because at the end of two days per week it will come to two persons each day and at night
twelve negroes, and more to cover and take out such charcoal’.166
The impression left by both texts is that amalgamation required a smaller workforce
than smelting, though no quantitative figures are proposed to establish man-hours per kg of
silver refined by both processes. This exercise will be carried out in Chapter 5.
2.3 The chemistry of smelting and the nature of the ore
Smelting of silver ores is a complex operation for two reasons: in most cases it takes
place at temperatures that require a furnace and not a simple cooking fire, and second, it
requires a high degree of empirical skill to guide the silver in the ore through at least two major
changes before isolating it in a pure state. However, the early history of silver refining in the
New World hinges on the fact that the only exception to this statement is for an ore rich in
silver chloride, like those found in the weathered section of a silver sulphide deposit. The
165
Jean Pierre Berthe, "Le mercure et l'industrie mexicaine au XVIéme siècle," in Ciencia, vida y espacio en
Iberoamérica: trabajos del programa movilizador del C.S.I.C., "Relaciones científicas y culturales entre España
y América". ed. Jose Luis Peset Reig (Madrid: Consejo Superior de Investigaciones Científicas, 1989), 144.
166
‘porque un ingenio de caballos que trae un horno andando bueno … ha menester cuatro fundidores y cuatro
cargadores y dos españoles que se muden por sus cuartos, y por personas que anden con los caballos del ingenio
por sus cuartos, y más dos afinadores, y para moler la greta y cendrada otras dos personas, y para hacer los
hornos y labrar las piedras otras dos, y para hollar las cendradas cada una que afinan, son menester seis
personas, porque al final de dos días a la semana que vendrán a ser dos personas cada día y de noche doce
negros, y más para cubrir y sacar dicho carbón’ as quoted in Castillo Martos, Bartolomé de Medina, 112.
117
highest accessible temperature range to a person with no special training is that of an ordinary
camp fire, in the range of 600 to 650o C.167 Native silver melts at 962o C, silver sulphide at 825
o
C, so this temperature would not be high enough to cause these compounds trapped in a
mineral matrix to flow and be recovered as a puddle in the ashes of the fire. However silver
chloride, with a melting point as low as 455o C, does soften and liquefy under these
conditions.168 Furthermore, in the presence of charcoal (carbon) lying at the base of the fire it
will be reduced to elemental silver at these relatively low temperatures.169
The weathering that chemically transformed the more superficial veins of silver
sulphide deposits in the New World therefore gifted the early mass of untrained Spanish
mineros with silver chloride, the easiest of silver compounds to refine by smelting, with
virtually no skills required. All they needed was to avoid overheating the silver ores so as to
avoid losing silver via volatilization, and even this level of care may have been beyond them.
The downside to such an easy operation was that no effort was made to refine the silver to its
highest possible level.170 It is no surprise that as soon as these silver halides were depleted, and
more silver sulphide was present in the ore, the gaggle of Spanish dilettante refiners would find
less and less silver coming out of their primitive operations, regardless of the true silver content
of their ores.
167
Aitchison, History of Metals, 34.
168
Cotton, Chemistry of Precious Metals 277.
169
Wallace, Barton, and Wilson, "Silver-Bearing Minerals," 17. Reduction in this thesis, except when it appears
within a quotation, refers to the modern chemical term as applied to redox equations, whereby the metal in its
ionic state gains electrons from another element that in turn is being oxidized (losing electrons), so that the metal
is transformed to its elemental state. When the word appears in historical quotations prior to the nineteenth century
it is used in in accordance with the Latin original of the word, reducere – “to lead back” to an original state.
‘Reduction had the more specific sense ... of the isolation or extraction of a metal from ... an ore’. William Royall
Newman, Atoms and Alchemy : Chymistry and the Experimental Origins of the Scientific Revolution (Chicago:
University of Chicago Press, 2006), xiii.
170
‘it is not likely that for a long time anything but very impure silver was shipped’ in Wagner, "Early Silver
Mining in New Spain," 61.
118
For all other silver compounds contained in ores the process requires a much greater
level of skill, such as was supplied through the German artisans in the New World. Europe by
mid fifteenth century had been refining its silver mainly from argentiferous galena, lead
sulphide ores.171 In contrast to most metal ores, silver will not form silver oxide on heating in
air, and heating silver ores to high temperatures over 1000o C can simply lead to major physical
losses of silver in the smoke of a furnace. It is estimated that at some point prior to 2500 BCE
a major breakthrough in smelting occurred, possibly invented by tribal people on the southern
shore of the Black Sea: silver could be processed as a by-product of smelting lead from
galena.172 The lead obtained by heating galena absorbs in its melted state any silver present in
the ore, either as native silver or silver that has also been reduced at high temperature from any
silver compounds present (silver sulphide, silver sulfo-salts, argentojarosite). The process from
a chemical point of view is the same whether lead is present in the ore (as in argentiferous
galena) or whether lead is added to the ‘dry’ silver ore.173 The difference only lies in the
economics of production, as analysed in Chapter 5.
Figure 2-1 represents the two main stages involved in the refining of silver from either
of these two starting minerals. The diagram is based on the description by Craddock of a two-
step refining process, where the first step is a smelting process under high temperatures and
reducing conditions that creates a silver rich lead. In Spanish historical texts this is called
fundición. The second step is sometimes called the cupellation stage, or afinación in Spanish
texts. The silver-enriched lead, plomo rico in Spanish, is heated in a cupel to around 1,000° C
171
Craddock, Early Metal Production, 211.
172
Galena is one of the easier ores to smelt, since at temperatures as low as 800° C lead oxide (litharge) is formed
from the lead sulphide at the upper, more oxidizing zone, while the reduction of the oxide to metallic lead (which
melts at 327° C) takes place in the presence of unburnt charcoal or carbon monoxide. In addition a further quantity
of lead is produced by the reaction between the lead oxide and the lead sulphide. N. H. Gale and Z. A. Stos-Gale,
"Cycladic Lead and Silver Metallurgy," The Annual of the British School at Athens 76(1981): 178.
173
It was known by the sixteenth century that to avoid losses via volatilization even native silver or silver chloride
benefit from the addition of lead prior to smelting. Agricola, De re metallica, 400.
119
under oxidizing conditions, achieved by blowing air (oxygen) onto the surface of the molten
lead so that litharge (lead oxide) is formed as a surface scum that entrains with it the oxides of
the majority of other metals present except gold.174 The litharge (greta in Spanish texts) is
skimmed off to be recycled. The litharge is also absorbed by the material of the cupel (porous
crushed bone in many cases) which is then also recycled (cendrada in Spanish texts). Recycling
did not require reducing lead from litharge, and had the added advantage of recovering any
silver entrained in the litharge.175
Simply heating efficiently and uniformly to the right temperature (in itself a major
technological breakthrough) is not enough: in the first step it must be done under conditions
that reduce the silver compounds (chlorides, sulphides or more complex sulphosalts) to metallic
silver, to assure maximum recovery of the silver content of the ore. Silver has to be absorbed
by the lead and not lost in the slag or volatilized if the temperature is too high. The second step
requires quite the opposite, a careful oxidation of the surface of the molten lead avoiding spatter
or other physical loss of silver into the litharge that is being continually scraped off, within a
furnace that has to be kept at or over one thousand degrees centigrade. 176 As a historian of
silver refining in the New World and mining engineer has written, after describing smelting of
silver ores:
‘litharge fluxed everything ... smelting furnaces fell apart in a week ... refining furnaces ...
lasted a little longer ...This [his description of the process] is so oversimplified that it may
appear anyone could do it. If so, blame this narration, because it is a very complicated and
difficult process which taxed the skill of experienced furnacemen’.177
174
Craddock, Early Metal Production, 210-223.; Gale and Stos-Gale, "Cycladic Lead and Silver Metallurgy,"
180. To observe the other possible inputs to the two stages of smelting see F. Ströbele et al., "Mineralogical and
Geochemical Characterization of High-medieval Lead–Silver Smelting Slags from Wiesloch near Heidelberg
(Germany)—An Approach to Process Reconstruction," Archaeological and Anthropological Sciences 2, no. 3
(2010): 212.
175
Biringuccio, The Pirotechnia, 172.
176
Ibid., 163.
177
Alan Probert, "Bartolome de Medina: the Patio Process and the Sixteenth Century Silver Crisis," Journal of
the West 8, no. 1 (1969): 96.
120
Smelting is not ‘a relatively simple affair’, ‘much more simple ... than
amalgamation’.178 On the other hand, smelting can be applied to any silver ore, though the
operation becomes more technically demanding according to the nature of the ore. When dry,
polymetallic ores are smelted, the two-stage process of Figure 2-1 can become an iterative
sequence of multiple stages of smelting, as has been reconstructed for the way jarositic earths
containing silver of the Rio Tinto mines in Spain were refined by smelting in Phoenician times,
using lead imported from Carthage.179 The silver sulphide and complex silver sulfo-salts of
the silver ores in the New World represent the intermediate spectrum of smelting complexity
between the beginner’s luck of silver chloride and the unique challenge of the jarositic silver
compounds of Rio Tinto. Smelting was always the refining workhorse for ores that could not
be refined by amalgamation in the New World, such as those containing lead or antimony.180
In Europe as of the mid fifteenth century, silver refining techniques had been forced to
adapt to a new type of silver source, the argentiferous ores rich in copper.181 The technical
challenge these new ores posed and the impact on silver production once the new refining
process had been mastered has been considered a pivotal moment in the history of mining in
Europe:
178
Brading, Miners Bourbon Mexico, 137.; ‘el proceso de fundición era mucho más simple ... que el de
amalgamación’, Lacueva Muñoz, La plata del Rey, 89.
179
180
Antimonial ores were being efficiently smelted together with iron or copper pyrites by 1604. West, The Parral
Mining District, 30.; ‘when smelting ores from hard sulphidic veins, or with antimony, it would have been enough
to mix them with lead ores, or greta, or cendrada’ – ‘Al tratar de fundir las menas de sulfuros duros, o
antimoniosos, hubiera bastado con mezclarlos con menas plomizas, o greta o cendrada’ in Bargalló, Minería y
metalurgia colonial, 91.; argentiferous lead, copper or zinc ores rebellious to amalgamation required smelting
according to Thomas Egleston, The Metallurgy of Silver, Gold, and Mercury in the United States (London; New
York: J. Wiley & Sons, 1887), 40-41. For detailed technical descriptions of how smelting was carried out up to
the nineteenth century see for example Laur, "De la metallurgie de l'argent au Mexique," 240-255.; Bargalló,
Minería y metalurgia colonial, 92-98, 249-251.; La química inorgánica y el beneficio de los metales en el México
pre-hispánico y colonial (Mexico: Universidad Nacional Autónoma de México, 1966), 109-110.; West, "Early
Silver Mining," 26.
181
Blanchard, Mining, Metallurgy and Minting in the Middle Ages, 3 1071.
121
‘An invention of the mid-fifteenth century ... was of even greater importance ... it was
discovered that the separation of silver from the argentiferous copper ores, which abounded in
Central Europe, could be effectively accomplished with the help of lead ... the rich copper ores
had been little exploited before this time because of the difficulty of extracting silver from them
... no other invention had so stimulating an effect … upon the development of the mining and
metallurgical industries in Central Europe on the eve of the Reformation’.182
Argentiferous ‘dry’ silver ore (AgS or

Galena (PbS + AgS or AgCl or Ag) + Lead or
+Ag + other metals) Litharge
loss of lead
solid slag R
Argentiferous Lead (Pb + Ag + other metals)
loss of lead
O solid litharge
Litharge (PbO) + oxides

Silver of other metals
Heating stage in furnace

R: reduction
O: oxidation Litharge (greta) cendrada
Figure 2-1. Schematic diagram of two stage refining of silver ores using lead, adapted from
Craddock, footnote 179.
The new copper liquation process (Saigerprozess) also made use of the capacity of
liquid lead to absorb silver. Silver was then recovered from the lead by the traditional
cupellation method.183 The increase in percentage terms of silver production in Europe was the
same order of magnitude as that which would be experienced in the New World with the
182
Nef, Conquest of the Material World, 36.
183
Blanchard, Mining, Metallurgy and Minting in the Middle Ages, 3 973.; Aitchison, History of Metals, 385.
122
introduction of amalgamation.184 ‘More than any other invention, this accounts for the great
prosperity of mining and metallurgy during the century which followed’.185
Even though this new refining breakthrough was not applied in the New World, I cite
it for two reasons. First, to emphasize that mining and refining are human activities that
continually respond to new challenges, so that events in the New World are not exceptional in
the sense that refining technology was already on a wave of large-scale innovation even before
Columbus sailed to his West. Second, because the banking dynasty of the Fuggers, who will
become a vital part of the early history of amalgamation in New Spain, owed their wealth in
part to this new technology. The Central European mining industry evolved into ‘a major
industrial complex’ thanks to the symbiotic production and sale of copper and silver, a pairing
essential for the commercial success of the new technology. It also incorporated into the pricing
of silver the costs associated with external sourcing of lead, since the argentiferous copper was
in general lead-poor. That it resulted in a very profitable venture can be judged from the success
of the Fuggers, who built their fortune on being able to coordinate ‘a complex chain of raw
materials, technical expertise and parallel marketing of two and even three metals, one noble
and two base’.186
The level of smelting skill developed in Europe allowed its refiners to extract silver
from ores with as low as 0.04% silver content.187 This technical threshold should not be
184
From 1470 to 1490 production of silver with the new process increased over five times to 22,794 kg and in a
further 20 years to 34,563 kg, .Bruce T. Moran, Patronage and Institutions : Science, Technology, and Medicine
at the European Court, 1500-1750 (Rochester, NY, USA: Boydell Press, 1991), 11.
185
Nef, "Silver in Europe," 576.
186
Blanchard, Mining, Metallurgy and Minting in the Middle Ages, 3 974.; Aitchison, History of Metals, 317.
187
Approximately 12 oz of silver per short ton of ore, Agricola, De re metallica, 388. A threshold of 0.02% for
smelting as practised in Saxony in the late eighteenth century is mentioned in Francisco Javier de Sarria, Ensayo
de metalurgia (Mexico: D. Felipe de Zuniga y Ontiveros, 1784), 105. In Poland argentiferous lead ores containing
just 0.02% were smelted, according to D. Molenda, "Silver Production in Poland, XVIth to XVIIIth Century " in
Hombres, Técnica, Plata. Minería y sociedad en Europa y América, Siglos XVI-XIX ed. Julio Sánchez Gómez and
Guillermo Mira Delli-Zotti (Sevilla: Aconcagua Libros, 2000), 23. A minimum value of 0.03% is given in in
Phillips, Metallurgy Silver, 433.
123
confused with the profit threshold determined by the balance between production costs and
total revenues obtained from refined silver and all other major metals present in the ore (gold,
lead or copper).
2.4 The infrastructure of smelting in New Spain
Though smelting was carried out in many locations and different periods within New
Spain, the region around the town of San Luis Potosí became a major producer of silver during
the seventeenth century on the strength of its smelting of silver ores. The silver deposits in the
vicinity of San Luis Potosí began to be exploited by the Spanish conquerors of the northern
Chichimeca territory by the end of the sixteenth century (1592). The most prominent were the
deposits of the Cerro San Pedro, but due to the lack of sufficient water at the site some of the
smelting haciendas were built in Monte Caldera, some 7 km from the mines and 25 km from
the town of San Luis Potosí.188 Others were located in the town itself. The presence of lead in
most of the ores found in the vicinity of the town of San Luis Potosí made smelting the only
viable refining process right from the start.189 In the absence of technical documents written by
refiners of this region, it is the historical legal documents of San Luis Potosí that provide an
indirect guide to the way smelting was practised by its major exponent in New Spain. 190 The
188
For the location of the main sites around San Luis Potosí mentioned in this chapter see Figure 2-14.
189
Even by 1744, by which time the use of mercury in some haciendas of the region has been documented, a
committee of miners including the local mayor (Alcalde Mayor) set up to test a new amalgamation recipe at the
request of the Crown made the observation that ‘what is generally processed are ores by fire [smelting], and at
this time none by mercury … the ores for mercury are of low silver content, with no stability [of supply] … but
to demonstrate their obedience to the Crown they will set up the test’ - ‘que lo que generalmente se benefician en
este mineral, son metales de fuego, y en estos tiempos ninguno de azogue … los de azogue, de cortissimas leyes,
de ninguna estabilidad, … pero para mostrar su obediencia a la Corona montaran el ensayo’ AHSLP, Fondo
Alcaldía Mayor 1744.1, expediente 35, dated 13 December 1744. Not all the ore was rich in lead, at least by 1622,
when there are references both to metal plomosso (lead ores) and to metal seco or sequillo (dry or dryish ores) in
AHSLP, Fondo Alcaldía Mayor 1622.6 expediente 15, receipts for ores dated 1622 and 1623.
190
The Spanish refiners did not find in New Spain the same advanced level of metallurgy their counterparts would
find in the Andes, where the indigenous craftsmen had been smelting silver ores in huayras, small tube furnaces
riddled with holes so that the blast was provided by the force of the wind on hillsides, well before the arrival of
the Spaniards. For studies on pre-hispanic smelting practice in the Andean region see Colleen M Zori and Peter
Tropper, "Late Pre-Hispanic and Early Colonial Silver Production in the Quebrada de Tarapaca, Northern Chile.,"
Boletín del Museo Chileno de Arte Precolombino 15(2010): 68,82,85.; Pablo Jose Cruz and Jean-Joinville Vacher,
124
operational stages of the smelting process have been described at length in the
historiography.191 I will only focus on those areas where some further light can be shed:
a) The introduction of molinos: a common sight at present around San Luis Potosí are the large
circular stones from the molinos (called Chilean mills in the English literature), Figure 2-2,
used to crush the raw ore placed in the path of the stone within the circular trough. 192 In this
example the stone was powered by a mule tethered to the arm leading to the axle of the
wheel.193 Smelting does not require the flour-like consistency demanded by amalgamation (see
next chapter), so breaking down the ore by hand was a viable option, especially so for small
refining haciendas set up during the early years of refining.194 It is not clear when and to what
eds., Mina y metalurgia en los Andes del sur: desde la época prehispánica hasta el siglo XVII (La Paz: Institut de
Recherche pour le Development / Instituto Francés de Estudios Andinos, 2008).; Warwick Bray, "Ancient
American Metal-Smiths," Proceedings of the Royal Anthropological Institute of Great Britain and Ireland, no.
1971 (1971): 29-32.; Heather Lechtman, "A Metallurgical Site Survey in the Peruvian Andes," Journal of Field
Archaeology 3, no. 1 (1976): 36-38.; Heather Lechtman et al., "Procesamiento de metales durante el Horizonte
Medio en el Altiplano Surandino (Escaramayu, Pulacayo, Potosi, Bolivia)." Boletín del Museo Chileno de Arte
Precolombino 15 (2010): 9-27. For a questioning whether cupellation was practised before the arrival of the
Spaniards see Mary Van Buren and Claire R Cohen, "Technological Changes in Silver Production after the
Spanich Conquest in Porco, Bolivia.," ibid.15: 33.
191
For sixteenth century practices in Spain and the New World, Barba, Arte de los metales, 130-170.; González,
Noticia histórica minas de Guadalcanal, II 410-411.; Carlos Sempat Assadourian, Zacatecas, conquista y
transformacion de la frontera en el siglo XVI : minas de plata, guerra y evangelización (México, D.F.: Colegio
de México, Centro de Estudios Históricos, 2008), 151-152. For later smelting practices see Francisco de Paula
Hermosa, Manual de Laboreo de Minas (Besanzon: Libreria de Rosa Bouret y Cia, 1857), 250-260. In English
the nineteenth century texts are the most useful, for example Phillips, Metallurgy Silver, 497.; Henry F. Collins,
The Metallurgy of Lead & Silver (London: Griffin & Co., 1899), 273-352.
192
Manuel Amador, Tratado práctico y completo de trabajos de minas y haciendas de beneficio (México: E.
Sánchez, Editor, 1901), 66-68, Lamina 5a Figura 1.
193
For a general description see Egleston, The Metallurgy of Silver, 269-270.
194
The final size should range between a grain of rice and that of a ‘Grueso de garvanzo’, the thickness of a chick
pea, and the more docile the ore, the bigger the particle, according to Garcés y Eguía, Nueva teórica del beneficio
de plata, 63-64. The bean size had already been pointed out in Biringuccio, The Pirotechnia, 153.
125
a) b)
c)
Figure 2-2. a) Molino mill stones, Monte Caldera, San Luis Potosí. The diameter can reach
2 m b) a modern reconstruction of a molino in Zacatecas, at the exit of the El Edén mine c)
drawing of a molino, reproduced from footnote 192.
extent these molinos were used in the seventeenth century smelting haciendas of San Luis
Potosí. The legal documents of the period are ambiguous, since they mention a wheel (rueda),
but in such a way that it could also apply to the assembly required by a bellows and not only
the rueda of a molino.195 According to West, Chilean mills were not introduced in New Spain
195
‘wheel, pinion, shafts, cross-beam and bellows’ - ‘la rueda y lanternilla peon exe y gualdra y fuelles’, as in the
sales contract between Nicolas de Peralta and Cristobal Zapata, AHSLP, Fondo Alcaldía Mayor 1667.3,
expediente 6, 18 July 1667. I have used the English translation for the Spanish terms as reported in Appendix I of
Peter J. Bakewell, Silver Mining and Society in Colonial Mexico: Zacatecas, 1546-1700 (Cambridge, U.K.:
University Press, 1971), 267. A guide in Spanish appears in de Gamboa, Comentarios Ordenanzas de Minas, 399-
401. Another example is the listing of ‘gualdra, lanternilla, peon y rueda y tablon y canones’ which includes the
cañones or nozzles of the bellows, in the sale contract dawn up by Juan Dominguez de Sequera, AHSLP, Fondo
Alcaldía Mayor 1653.1, expediente 8, 14 March 1653. The drive wheel is a fundamental element that transforms
the horizontal circular motion imparted by mule or water power via a crankshaft into the reciprocating movement
that drives the bellows for the furnace.
126
until the nineteenth century.196 For the largest of the smelting establishments Salazar Gonzalez
identifies a mortero (stamp mill) by 1628, and a stamp-head for a mortero by 1703 as part of
their infrastructure.197
b) The role of dressing ores: the importance given to the requirement of water for the
smelting haciendas at first sight is surprising, since water seems a more critical issue for
amalgamation than for smelting:
“Antonio de Espinoza … townsman and miner of the mines of San Luis Potosi states that as is
notorious and public [knowledge] the ores that are extracted from this hill [Cerro San Pedro]
cannot be processed or smelted without washing. And due to the great lack of water that has
been and still is … some haciendas have ceased to process [these ores].’198
The need to have water for the workers and animals is evident.199 What is interesting is
the use of water for the washing of the ores. This meant that silver ores in San Luis Potosí were
dressed, concentrated for silver and lead content prior to smelting based on a differential
sedimentation rate in water depending on the density of the mineral particles, in the process
explained by Agricola as of the sixteenth century.200 According to Barba, washing ores prior
to smelting was not normal practice in the New World.201 The dressing of ores that was an
important part of the initial strategy of the English investors of the nineteenth century in
Pachuca was in fact never implemented (Chapter 4). It is important therefore that the
196
West, The Parral Mining District, 113. In Chapter 3 I cite evidence that places them in Zacatecas earlier.
197
Guadalupe Salazar González, Las haciendas en el siglo XVII en la región minera de San Luis Potosí: Su
espacio, forma, función, material, significado y la estructuración regional San Luis Potosí (Universidad
Autonoma de San Luis de Potosí, 2000), 84, 90.
198
‘Antonio de Espinoza … vezino y minero de las minas de San Luis de Potosi digo que como es notorio y publico
los metales que deste serro se sacan no se pueden beneficiar ni fundir sin labar. Y por la gran falta que a avido
y ay de agua … a sesado el beneficio en algunas haciendas’ AHSLP, Fondo Alcaldía Mayor 1635.1, expediente
6, 16 January 1635.
199
Where water power was deficient, animal power was required to drive mills and bellows, as well as to haul ore
from the mines to the haciendas. As late as 1778 a new invention was being touted in San Luis Potosí that would
not require ‘abundant water for the beasts’ – ‘ni abundancia de Agua para Bestias’. Claim by Don Juan Martin
de Irrazu, AHSLP, Fondo Alcaldía Mayor 1778.2, 30 October 1778.
200
Agricola, De re metallica, 300-310.
201
‘at great length Agricola teaches how to wash ores prior to smelting; it is not much used in these Kingdoms’-
‘muy dilatadamente ensena el Agricola a lavar los metales antes de fundirlos; poco se usa en estos Reynos’ in
Barba, Arte de los metales, 148.
127
documents show that for San Luis Potosí dressing was part of the common practice in smelting
haciendas. Lavadores, workers who carried out the washing of ores, are mentioned amongst
the labour force of a smelting refining hacienda in this area as late as 1773.202
c) The chimneys of the smelting furnaces: smelting was carried out in an horno
castellano.203 Barba defines them as simply the same type of furnace as currently in use in
most of his known world to smelt all sorts of ores, as described by Agricola.204 They were
initially very simple structures built from stone and lime mortar (‘piedra y cal’) in the form of
a square pillar up to 2 metres tall and with an internal square cross section of under one square
metre, or built in the shape of an inverted sectioned cone.205 It was an inexpensive construction
made from local materials that could be easily rebuilt as required by the wear and tear of
smelting. An inventory of a rented smelting hacienda in Zacatecas, dated 1608, describes well
the precarious nature of the early structures: ‘two chimneys of stone and lime one good and the
other with openings in three or four places … four furnaces of stone and lime with chimneys
made from adobe, all open in many places and propped up’.206 Their presence in the ruins of
these haciendas is inferred by the more evident remains of a tuyere, an alcribis in Spanish texts,
202
Weekly accounts of the Hacienda de Nuestra Señora de los Dolores, signed by Lorenzo de Mata for the year
1773, AHSLP, Fondo Alcaldía Mayor 1773.2.
203
This is translated to English as a castillian furnace, though the term ‘Spanish furnace’ might be more
appropriate.
204
‘Llaman en este Reyno Hornos Castellanos a los que en las otras tres primeras partes del mundo, han sido
usados, y comunes para la fundición de toda suerte de metales. De ellos solos trata el Agricola para este efecto’.
205
‘erect these furnaces plumb to the ground, in the shape of a square pillar somewhat taller than wider at the
cavity. Their height is one vara [ approx.. 80 cm], some nearly two and some less … at the back they have a small
aperture … the [alcribis] where the nozzles of the bellows are placed … others make these furnaces round, wider
at the top than at the bottom’ … ‘Levantase estos hornos a perpendículo, en forma de un pilar cuadrado algo mas
largos, que anchos por lo hueco. Tienen de alto algunas una vara, otros casi dos, y otros menos … por la parte
de atrás en una ventanilla … el alchrebiz en que han de estar los cañones del fuelle … otros hacen estos hornos
redondos, mas anchos de arriba que de abaxo’ ibid.
206
‘dos chimeneas de piedra y cal la una buena y la otra abierta por tres o quatro partes … quatro hornos de
piedra y cal con las chimeneas de adobe todas abiertas por muchas partes y apuntaladas’ in the inventory drawn
up by Pedro de Medina and Andres Pereira for an hacienda ‘de refinar y afinar’ (to smelt and refine) rented from
Doña Margarita de Cobarrubia in Fresnillo, 20 March 1608, AHEZ, Notaria-Colonia, Numero 01 (Pedro Venegas
pendiente), expediente 1, 4r.
128
the element of the furnace wall fitted with an orifice that allowed the bellows to pump air
through a nozzle (cañon) into these furnaces.207
Low chimney heights would leave the workers and immediate surroundings exposed
to the lead fumes issued from the furnace.208 The surviving pyramidal chimney that
characterizes the ruins of the smelting haciendas around San Luis Potosí (Figure 2-3) is not the
original from the early period but acquired its present shape and height following the mining
laws of the eighteenth century.209 By the nineteenth century ‘ordinarily two of these furnaces
are placed one beside the other, under the same pyramidal chimney without a roof on its top’.210
The only example I found of two standing chimneys (Figure 2-4) was in the ruins of the
Hacienda de Aranzazu situated in Guadalcazar, a historical mining and refining district some
100 km north-east from the town of San Luis Potosí.211 The ruins of this hacienda have not yet
been analysed or reconstructed from an architectural standpoint. The castillian furnaces
207
This would have been the most critical part of the furnace, judging from the manner in which the weight of the
alcribis is singled out in the rental agreement between Fraga Gorbaran and Rodrigo de Aldana, in which it is
stated that the new alcribis that weighs 37 lbs must be returned at the end of the rental period. AHSLP, Fondo
Alcaldía Mayor,1631.3, expediente 38, 27 December 1631. In a similar vein the cost of three sets of bellows (at
70 pesos each) and the set of an alcribis said to weigh 48 lbs and a nozzle of 13 lbs costing together 60 pesos are
among the few fixed costs listed by Juan Lopez de la Madriz in his book of accounts, AHSLP, Fondo Alcaldía
Mayor 1650.3, expediente 8. The importance given to the weight of the alcribis in both documents is noteworthy.
208
The chimneys of these early furnaces were sometimes raised in height not because of the fumes but to capture
any silver entrained in the flue gas. Barba, Arte de los metales, 167. Barba’s dimensions and the height of the
extant pyramidal chimneys measured in Monte Caldera and Guadalcazar are significantly lower than Bakewell’s
account of a mining edict implemented by Vice-Roy Toledo in 1574 in Peru whereby lead smelting had to be
carried out an enclosed building with chimneys some 7 meters tall (4 estados). Bakewell, Miners of the Red
Mountain, 150.
209
Prof. Gonzalez Salazar, private communication. Hermosa describes a ‘German furnace’ in the nineteenth
century with a height of 6 varas, under 5 m, which is more in line with extant chimney heights. Hermosa, Manual
de Laboreo de Minas, 254-255.
210
‘ordinairement deux de ces fourneaux sont places a cote l’un de l’autre , sous une même pyramide sans toiture
dans le haut’ in Duport, Métaux précieux au Mexique, 70. There is an illustration of the side by side arrangement
of furnaces under a common chimney in the paper on silver smelting by Václav Vaněk and Dalibor Velebil, "Staré
hutnictví stříbra," in Stříbrná Jihlava 2007. Studie k dějinám hornictví a důlních prací (Jihlava: Archaia Brno /
Muzeum Vysočiny Jihlava, 2007). The illustration can be accessed via http://www.velebil.net/clanky/hutnictvi-
stribra/stribrna-hut-4.
211
I visited Guadalcazar at the urging from the Director of the Archivo Histórico San Luis Potosí, Dr. Rafael
Morales Bocardo. In Guadalcazar I was led to the ruins of the Hacienda de Aranzazu by Doña Maria Esther, who
is also in charge of the colonial museum of the church. For further general information on the area see Alejandro
Galvan Arellano, Arquitectura y urbanismo de la ciudad de San Luis Potosí en el siglo XVII (San Luis Potosí:
Editorial Universitaria Potosina, 1999), 70.
129
described in eighteenth and nineteenth-century texts correspond to more elaborate affairs than
Barba’s ‘horno Castellano’.212 By the nineteenth century, blast furnaces were installed in
Mexican smelting haciendas, with the blast of air driven either by water or steam engines.
These furnaces required top-loading of the smelting charge, via an opening to the furnace
situated at an upper floor to the level of the furnace hearth, as indicated in Figure 2-5. 213
Figure 2-3. Exterior of smelting furnace at the ruins of the Hacienda Santa María in Monte
Caldera. The arched port would have been used to feed ore or fuel to the furnace.
212
For example, in de Sarria, Ensayo de metalurgia, 109-110.; Garcés y Eguía, Nueva teórica del beneficio de
plata, 67-68.; Duport, Métaux précieux au Mexique, 69.; Pique, A Practical Treatise on Silver, 67-69.; Phillips,
Metallurgy Silver, 477.
213
Robert H. Lamborn, The Metallurgy of Silver and Lead : A Description of the Ores; their Assay and Treatment,
and Valuable Constituents (London: C. Lockwood, 1878), 125.
130
a b
furnaces grasas
hacienda
wall
69 m
d e
Figure 2-4. Ruins of the Hacienda de Aranzazu in Guadalcazar. a) front view showing
archways under two smelting furnace chimneys, height around 7 m b) back view of chimneys,
showing possible aperture for drive shaft of bellows c) section of hacienda wall d) fields of
grasas e) image from Google Earth © 2014 DigitalGlobe, 22°37’21” N 100°24’9” W.
Photographs of these furnaces as installed at the Hacienda de Regla, near Pachuca, are provided
in Chapter 4.
Because a smelting furnace required a set of adjacent bellows, its architectural footprint
required an additional area contiguous to the furnace to fit both the bellows, the mules turning
131
round in a circle, or a waterwheel fed by hydraulic power. Thus for example: ‘there are two
joined mules that move the wheel of said ingenio [machinery] that makes the blast provided by
load floor
not to scale
~5 m
Furnace front
facing north at Regla
tuyere
exact type of hearth unknown
Figure 2-5. Section of a blast furnace as found at Regla. Adapted from footnote 213.
a set of bellows that are placed on the side of one of two furnaces that are found in this
hacienda’.214 In the nineteenth century the description had hardly changed : ‘the blast of air is
given by two bellows … one mule for each furnace’.215 The space required by this additional
area shared by up to a pair of furnaces, where the effective power to drive the bellows was
generated, is not always possible to identify in the reconstruction of smelting haciendas.
214
‘estan dos bestias mulares uncidas que mueven la rueda de dicho ingenio y hacen dar soplo a una parada de
fuelles que están puestos hacia la parte de un horno de dos que hay en esta hacienda’ from the report of 1593
written by Juan Lopez de Riego on the smelting hacienda of Captain Miguel Caldera, as cited in Galvan Arellano,
Arquitectura de San Luis Potosí, 60-61.
215
‘le vent est donne par deux soufflets … une mule pour chaque fourneau’ in Duport, Métaux précieux au
Mexique, 70.
132
Cupellation of the silver-enriched lead bars (barras in Spanish texts, or ‘pigs’ in the
English literature) usually took place in a separate reverberatory oven, where heating is by
indirect reflection from a curved roof and the ore is not in direct contact with the carbon fuel.216
The whole set, furnace and prepared bed, is named a vazo, baso or vaso (literally a vessel, a
tumbler) in the legal documents of San Luis Potosí and Zacatecas.217 A single vaso could serve
to refine the enriched lead bars from up to four smelting furnaces.218
Because of the attrition of the processes on these structures it is not usual to find whole
vasos from the first two centuries of refining activity. In the ruins of Hacienda Number 2 (HMC
2) in Monte Caldera are two caved-in depressions in the shape of bowls connected via a tubular
structure (Figure 2-6). A complete archeological study would be required to establish if these
are the remains of a vaso from the original smelting hacienda.
d) The fields of grasas: the smelting furnaces produced a solid waste that was dumped close to
the refining hacienda. Historically, smelting refineries have had no compunction about soiling
their own nests, and major slag heaps still abound around the husks of smelting haciendas in
San Luis Potosí, giving the landscape the desolate look and crunchy step of an old lava field
(Figure 2-7).219
216
The low domed roof of these ovens reflects the heat from the wood fire onto a separate chamber where the
material to be heated is placed. For details of early sixteenth century reverberatory ovens used in the New World
see Barba, Arte de los metales, 136-38.; for later periods there are more sources, for example Phillips, Metallurgy
Silver, 449.
217
By the nineteenth century they are still called vasos: ‘le vaso ou fourneau du coupelle’, the biggest with of 1.2
and 1.4 m, and a depth of 15 cm; these ovens do not have an enclosed vent to the roof. Duport, Métaux précieux
au Mexique, 70-72.
218
The concentration of silver in the lead bars refined in the vaso is said to be between 8 and 10%. Pique, A
Practical Treatise on Silver, 59.; Phillips, Metallurgy Silver, 488. This explains why these enriched bars were
stolen from haciendas, as in the claim by Rodrigo de Noriega against Juan Rodríguez for allowing two indigenous
workers to refine three stolen bars of lead using his set of bellows. AHSLP, Fondo Alcaldía Mayor 1650.3,
expediente 1, 13 June 1650.
219
The heaps of slag at Laurion in Greece were of the order of several million tons, of which 1.5 million inland
and the same dumped into the sea. Hans-Gert Bachmann, "Archäometallurgische Untersuchungen zur antiken
Silbergewinnung in Laurion. II. Charakterisierung von Bleiverhüttungsschlacken aus Laurion=
133
a b
c d
Figure 2-6. a) Ruins of smelting hacienda HMC2, with chimney b) alquibris (tuyere) c) one
of two bowl-shaped depressions with caved-in roof, connected via the tubular section observed
in d).
The lead-rich slag from smelters are termed grasas (literally ‘greases’) in Spanish texts of the
period.220 Other solid waste products are called granzas and mazamorras (discarded broken-
down ore with very low silver content) or lamas (fine waste from washing the ores during
Archaeometallurgical Investigation on Ancient Silver Smelting at Laurion. II. Characterisation of Lead Smelting
Slags from Laurion," Erzmetall 35, no. 5 (1982): 246.
220
I have not found an etymology for this use of the word ‘grasas’. A clue may lie in the description by Barba of
molten slag from lead ores: ‘when the slag is well melted, and liquid like oil’-‘quando la escoria esta muy
derretida, y liquida como azeyte’. The visual similarity to oil from grease may have led to the word ‘grasas’.
134
dressing).221 None of these solid wastes were considered truly final and worthless products,
and the sale contracts for smelting haciendas carried a stock phrase along the lines of explicitly
including ‘grasas granzas lamas mazamorras y desechaderos’ among the tangible assets of the
hacienda being sold.222. Furthermore, idle land that was suspected of containing waste from
previous refining activity was sought after and dug up to recover these potential sources of
recoverable traces of silver and lead.223
a) b)
Figure 2-7. Mounds of grasas in Monte Caldera a) Hacienda Santa Maria, chimney of
Figure 2-3 in the background b) Hacienda HMC2.
221
Ricardo N. Alonso, Diccionario Minero. Glosario de voces utilizadas por los mineros de Iberoamérica
(Madrid: Consejo Superior de Investigaciones Científicas, 1995).
222
Contract for sale of smelting hacienda by Juan Dominguez de Sequera to Cristobal del Castillo in Monte
Caldera, AHSLP, Fondo Alcaldía Mayor 1653.1, expediente 8, 14 March 1653; by Nicolas de Peralta Pimentel to
Cristobal Zapata, AHSLP, Fondo Alcaldía Mayor 1667.3, expediente 6, 18 July 1667; by Pedro de la Perna to
Captain Juan Manuel Rendon, AHSLP, Fondo Alcaldía Mayor, 1696.3, 6 November 1696.
223
Request by Juan Lopez de Meza to dig up a site previously occupied by a smelting hacienda near the Convent
of San Agustin in San Luis Potosí in order to process its content of grasas and other solid waste. AHSLP, Fondo
Alcaldía Mayor 1672.1, 10 February 1672.
135
2.5 The architecture of smelting in New Spain
The region around the town of San Luis Potosí epitomizes the use of smelting to refine
silver ores in New Spain, so its haciendas are a prime example of the physical structure of this
genre. In the historical legal documents, these haciendas are not characterized by their
nameplate production capacity, expressed as their capacity to handle a certain throughput of
ore per month or year. Rather they are described by the number of smelting furnaces (ornos or
hornos) and cupelling furnaces (vasos) they possessed, which would indicate that furnace size
was uniform and their capacity so well known it did not merit a special mention in the legal
documents pertaining to the sale or rent of these haciendas. A non-exhaustive survey of textual
sources points to a range of smelting furnaces per hacienda between one and sixteen, while
the number of cupelling furnaces is in general just the one, very rarely two.224 The only textual
clue as to the capacity of these smelting furnaces is found in a document dated 1620 which
states that a total of 22 smelting haciendas produced in a year 150,000 marks of silver from
100 furnaces.225 If the size of these furnaces is as standard as I have assumed, on average one
smelting furnace could produce at least 1,500 marks of silver (345 kg) per year. The data can
also be interpreted to mean that an average smelting hacienda had between 4 to 5 furnaces, and
produced in that period approximately 1.5 tons of silver per year.
There is an account book of smelting operations in the Valle de Pozos (see Figure 2-14
below) that gives credence to this average. It dates from a later period, 1660 to 1661, and was
224
Sale by Juan Dominguez de Sequera to Cristobal del Castillo of an hacienda in Monte Caldera with one
smelting furnace, AHSLP, Fondo Alcaldía Mayor 1653.1, expediente 8, 14 March 1653; sale by Nicolas Peralta
de Pimentel to Cristobal Zapata of an hacienda within the town of San Luis Potosí with two smelting furnaces,
AHSLP, Fondo Alcaldía Mayor 1667.3, expediente 6, 18 July 1667; rental by Mathias Pardo to Sanchez and
Rodriguez of an hacienda with four smelting furnaces in the valley of Pozos, AHSLP, Fondo Alcaldía Mayor
1629.3, expediente 24, 31 March 1629; the Hacienda de Briones in Monte Caldera is reported as having eight
smelting furnaces in 1628, and up to sixteen furnaces have been reported according to Salazar González, Las
haciendas de San Luis Potosí, 90, 97.
225
Ibid., 24, 80.
136
presented as evidence for the state of business dealings (‘trato’) between Juan Lopez de la
Madriz and the deceased Miguel de Santibañez. As such it is not an account book based on
operational data, but a set of prepared accounts of expenditures and revenues from sales of
silver and gold produced between May 1660 and December 1661 that are presented in defence
of its author, de la Madriz. The entries cannot be used to establish an operational cost of
production, since they deal mainly with payments made to individuals over this time period.
The account book however includes a series of entries at the end that detail the amount of silver
and gold produced. The pages are still stitched together within their original leather bindings,
and the folio numbers are consecutive, but there is no guarantee that it reproduces all the
production data for the period covered. The aggregate amount of silver reported from May to
November 1660 was approximately 500 kg, which prorated to the whole year corresponds to
approximately 860 kg, ignoring production fluctuations according to the season. A similar
aggregate amount (approximately 880 kg) is reported for the whole year of 1661. If these
records correspond to the production of a single hacienda, a fact that cannot be established
from the documents, then in both years the maximum monthly production was approximately
150 kg of silver, within the range (20% above) of the yearly average of 1.5 tons calculated for
1620. The other finding of note to come out of this account book is the role played by the
revenues from gold refined from these silver ores. In 1660 and 1661 gold contributed around
25 to 30% of the total revenues reported by its author, Lopez de la Madriz.226 Thus the presence
of gold would have played a major role in meeting the production costs of smelting.
The whole state of San Luis Potosí has at present 36 ruins of refining haciendas that
correspond to the sixteenth and seventeenth centuries, all either in an advanced state of decay
226
AHSLP, Fondo Alcaldía Mayor 1650.3, expediente 8, selected accounts of the years 1660 and 1661. For the
detailed calculation see Appendix A.
137
or with substantial changes to their original architecture. Galvan Arellano commenced and
Salazar Gonzalez has continued in greater detail and scope the architectural study and
reconstruction of some of these haciendas.227 Without a detailed knowledge of the spatial size
and distribution of the production elements within a smelting hacienda it is impossible to
determine the spread of the environmental impact around each smelting location. An estimate
of the architectural footprint, and of the area outside the boundaries of the hacienda proper that
can be deemed industrial and thus off-limits to agricultural activities or to dwellings is also
required. Finally, the local wind patterns and proximity to waterways that can extend the
geographical area impacted by the production process also has to be established for each
refining unit. This is a very detailed agenda for every refining district and in this section I will
only carry out an approximate exercise for a few examples in the area of Monte Caldera.
The area occupied by a smelting hacienda is an important factor, since a large area
would tend to contain within its boundaries a larger share of the wind-borne lead from low
furnace chimneys, thus attenuating its impact on neighbouring habitations. A document from
1772 details the sale of a smelting hacienda in Monte Caldera between Cristobal Pardo and
Juan Nieto for 640 pesos, with a dimension of ‘140 varas de oriente a poniente, 85 varas de
norte a sur’, approximately 110 m east to west by 70 m north to south, for an area of 7,700 m2.
It included a small reservoir (‘tanquesito’) to collect water in the rainy season.228 An idea of
the balance between functional areas and waste areas can be gained from modern satellite
images. Figure 2-8 shows satellite images of the area around the village of Monte Caldera, and
the location of ruins of smelting haciendas. The approximate area for the Hacienda HMC1 is
227
Galvan Arellano, Arquitectura de San Luis Potosí, 211-13, 271-72.; Salazar González, Las haciendas de San
Luis Potosí, 83-119, 428-34. Due to the generous arrangements provided by Prof. Salazar I was able to visit four
of the eight haciendas that according to Galvan existed in Monte Caldera. Galvan Arellano, Arquitectura de San
Luis Potosí, 56.
228
AHSLP, Fondo Alcaldía Mayor 1772.2 expediente 18, 7 November 1772.
138
3,000 m2, of which approximately 2,000 m2 are taken up by the waste grasas. In the case of the
Hacienda Santa María, approximately 3,500 m2 correspond to operational areas including the
water reservoir, and approximately the same area for the grasas. In each of these satellite
images the area occupied by the waste grasas either matches or surpasses the extant footprint
of the historical hacienda, thus doubling the size of the industrial plot required overall for silver
production.229 The environmental implications are important, since these wastelands of grasas
inadvertently helped to isolate the hacienda from agricultural or cattle grazing activities and
from human dwellings, and by desolating their stretch of land they acted as a sink of further
chemical depositions issued from the smelting chimneys.230
Some of the smelting haciendas of the state of San Luis Potosí were set up within
town limits.231 It would not have been possible for each smelting hacienda within the town to
have occupied a space or to have spread its mineral waste in the same manner as in the
countryside, but I have no information on sizes, areas or how grasas were legislated by the
town authorities. Mounds of grasas from previous or existing smelting haciendas inside the
city were considered to be sufficiently part of the urban landscape that they merited being
included in a map of San Luis Potosí from as late as 1789 (Figure 2-9). When the streets of
present day San Luis Potosí are dug up for major road works or to lay pipes, workers come
across an underground layer of grasas, as if a historic volcano had at one time covered the area
in ash.232
229
Satellite images do not necessarily show the whole extent of the hacienda or the grasas, the former due to a
blurring of physical perimeter walls with time, the latter due to cover from trees or displacement with time by
agricultural land, disposal as gravel to line tracks in the countryside around Monte Caldera, or even by the stream
when in flow. Even with these limitations the images are a useful tool to appreciate the relative dimensions of
each area.
230
The large area taken up by grasas and other solid wastes is another reason to include them in sale contracts
that involved the ownership of land.
231
For the location of haciendas in the region close to the city see Salazar González, Las haciendas de San Luis
Potosí, 396.
232
Personal communication from Dr. Rafael Morales Bocardo.
139
HMC 2
La Luz
Church and
main square
HMC 1
SANTA
38 m MARIA
131 m
Figure 2-8. The solid white lines encapsulate the minimum area that can be clearly
identified with each smelting hacienda, the dotted line the minimum area of the extant dumps
of grasas. All satellite images from Google Earth © 2014 DigitalGlobe. The Hacienda Santa
María lies at 22°12’10” N 100°44’27” W, and the Hacienda HMC1 at 22°12’31” N
100°44’47”W.
140
Figure 2-9. Copy (1986) drawn by Carlos Morán de la Rosa of the original map by Captain
Manuel Pascal de Burgoa, 1794, showing the division of the city of San Luis Potosí into barrios
(quarters) by the Viceroy Marques de Branciforte. Digital copy courtesy of AHSLP, Colección
Mapas y Planos.
It would be useful if either the historical sale price or the rent of these haciendas
were a guide to their refining capacity, and thus their size. However the data are not conclusive,
since the assets sold or rented do not only comprise furnaces but at times mines, charcoal
making facilities (carboneras), slaves, indigenous work squads, livestock and mounds of solids
of varying magnitude. The limited evidence for now points to the year of the sale, rather than
the number of furnaces or other assets, that determines the market value of the hacienda. This
would be consistent with a valuation based on expected future cash-flows rather than on the
cost of construction. Thus an hacienda with four smelting furnaces is to be sold for 20,000
pesos in 1628, while another with just one smelting furnace is sold for 1,700 pesos in 1653,
and another hacienda with just one furnace is sold for 700 pesos in 1667. The near halving in
price per furnace as the century advanced correlates well with a decrease in the silver refining
141
activity in the area.233 I only have one reference to the actual cost of building a smelter within
an hacienda in the Valle de Bledos being rented by Juan de Sandoval in 1607. It cost 525 pesos
to build, but the cost excluded both wood and certain basic equipment (bellows and other
accessories) furnished by Sandoval. The contract calls for construction to be completed within
budget in 40 days, which reflects a simple structure. It does not state how many furnaces are
covered by the contract.234
Some, but not all, rental contracts of smelting haciendas also specify the number of
furnaces being rented, as well as a wide menu of additional services, such as workers, or access
to mines or mounds of grasas, which in turn determine a range in rents between 250 and 1,000
pesos per year.235 In the case of rents the correlation with the year is less evident. The most
intriguing contract is for the use of a smelting furnace at 5 pesos per day, and for mines and
two indigenous workers ‘that belong to me’ at 50 pesos/y, since it shows a significant degree
of uninhibited entrepreneurship within the private silver refining business in this region.236
233
Proposed sale of the hacienda of the deceased Juan Perez Basurto by Antonio de Arismendi Gogorron for the
proposed sum of twenty thousand pesos, photocopy of original document dated 30 May 1628, folio 52 r, AHSLP,
Colección Miguel Iwadare; sale by Juan Dominguez de Sequera to Cristobal del Castillo of an hacienda with one
smelting and one refining furnace in Monte Caldera for 1,700 pesos, AHSLP, Fondo Alcaldía Mayor 1653.1,
expediente 8, 14 March 1653; sale by Nicolas de Peralta Pimentel to Cristobal Zapata of an hacienda behind the
convent of St. Francis in the town of San Luis Potosí with one smelting and one refining furnace for 700 pesos,
AHSLP, Fondo Alcaldía Mayor 1667.3, expediente 6, 18 July 1667.
234
The builder, Juan de Vargas, is asking for the pending amount of 232 ‘pesos de oro común’ that have not yet
been paid since Sandoval is not satisfied with the work carried out. AHSLP, Fondo Alcaldía Mayor, 1607,
expediente 1, 9 October 1604 and 24 April 1608.
235
Two refining haciendas, furnaces not given, and their mounds of grasas, at 500 pesos/y each, between
Geronimo de León and Palomo y María de Mendoza. AHSLP, Fondo Alcaldía Mayor 1627.5, expediente 27, 24
December 1627; one hacienda with four smelting furnaces at 500 pesos/y between Mathias Pardo and Sanchez
and Rodriguez (document damaged). AHSLP, Fondo Alcaldía Mayor 1629.3, expediente 24, 31 March 1629; one
of four furnaces within a smelting hacienda, plus use of refining furnace once a month, and partial use of existing
workforce, plus supply from existing mounds of grasas, at 1,000 pesos/y between de Fraga and Rodrigo de
Aldana Chavez. AHSLP, Fondo Alcaldía Mayor 1631.3, expediente 38, 27 December 1631; one smelting and one
refining furnace but without fuel at 250 pesos/y between Gaspar de Villanueva and Fernando de Mesa Godines.
AHSLP, Fondo Alcaldía Mayor 1641.1, expediente 9, 25 June 1641; an hacienda with two smelting and one
refining furnace for 400 pesos/y between Geronimo Dias and Alonso de Borja. AHSLP, Fondo Alcaldía Mayor
1658.1 expediente 4, 7 January 1658; an hacienda with three smelting furnaces for 750 pesos/y between Francisco
Dias del Campo y Diego Sanchez. AHSLP, Fondo Alcaldía Mayor 1658.1, expediente 18, 18 March 1658.
236
Rental agreement between Francisco Dias del Campo and Hernan Vasquez, AHSLP Fondo Alcaldía Mayor
1635.5, expediente 28.
142
Overall these contracts show a cross-section of an active market for the outsourcing of refining
services. The level of rent in this market was the equivalent of approximately 30 to 120 marks
of silver per year, at the most some 1% of the average silver production per hacienda registered
in 1620. It is interesting that the rental contract was not tied in some manner to the total silver
produced by the renter. One strong incentive to rent must have been the provision of ores and
skilled workers more than the relatively simple infrastructure that could be built in five weeks.
Towards the end of the eighteenth century, smelting would be overshadowed by cazo
amalgamation in the refining of silver presented to the Caja (Treasury) of San Luis Potosí. The
boom in silver production shifted to the mines of Catorce in the northern part of the state, where
the nature of ore (mainly silver halides, poor in lead) made the cazo process the best refining
option.237 The difficulties faced by some smelting haciendas during this period is reflected in
the nineteen surviving weekly accounts from the Hacienda de Nuestra Señora de los Dolores,
corresponding to the year 1773. An analysis of the records that tracked the silver refined per
week indicate that on average 2.4 kg of silver were produced on a weekly basis, approximately
0.12 tons per year. This is one third of the value reported for just one furnace in 1620. If the
number of cargas of ore set down in these weekly accounts were the source of the silver refined,
the values indicate a silver content for the ore being smelted around 0.6%. The data show a
persistent operational loss in the accounts being rendered, based only on silver revenues.238
2.6 The environmental impact vectors from smelting
The smelting process presents two environmental impact vectors of special importance:
the emission of lead products and the consumption of wood for charcoal. I treat these
237
See Chapter 6.
238
Weekly accounts signed by Lorenzo Mata that cover, with major gaps, the period from 27 December 1772 to
28 November 1773, AHSLP, Fondo Alcaldía Mayor 1773.2. The details of the calculation based on the primary
data are given in Appendix A.
143
production variables as vectors, since they not only have a magnitude but also a directionality
that is relevant to any environmental analysis. In modern environmental impact studies the
level of emissions is determined by in situ field measurements of the chemicals being studied.
For historical estimates of chemical emissions I will use the principle of the conservation of
matter in this and the following chapters. The mass of all the chemical reagents and ore that
entered an hacienda has to equal the silver produced, any by-products sold, and the total
emission of chemicals and inert mineral matter to the environment.239 The method will provide
the quantitative ratio of each emission or amount of energy consumed per kg of silver refined.
This ratio in turn will allow me to project over a whole region the quantitative environmental
impact of each major emission or energy source simply by knowing the amount of silver
produced either by smelting or amalgamation. While the macro-impact lends itself to this
method, the local impact is location specific, and thus can only be established on a case by case
basis.
2.6.1 Lead: its nature and magnitude
In the case of lead the only sources that supply it to the smelting process are the ore
itself and / or any fresh lead compound (poor lead, greta or cendrada) that is added to
compensate for the dryness of the ore or for losses of lead during the smelting process.
Recycling of lead, greta, cendrada or the recovered accretions from the furnace walls does not
enter the gross mass balance equation. The amounts of lead that must be replenished during
the smelting process are due to four causes (Figure 2-10):
239
I have not come across any similar application to estimate historical environmental impacts, but cannot affirm
the method is original across all disciplines. I would argue it provides better order of magnitude estimates than to
measure modern soil concentrations of the offending chemical and then projecting back in time a quantitative
emission factor for said chemical.
144
1. Loss to the atmosphere of lead and its compounds, spread via the chimney flue gas
or within the area around the smelting and refining furnaces. The heating of lead or lead ores
creates was is known as a lead fume, an aerosol of particles composed mainly of lead sulphide,
lead sulphate, lead oxide, lead carbonate and metallic lead.240 The exact chemical composition
and the size distribution of the particles varies according to the temperature in the furnace, the
presence of oxygen, and the other compounds present in the furnace. This fume is so rich in
lead and lead products that in Europe it was recovered in horizontal flue traps to extract its lead
products as of the end of the eighteenth century.241
The toxicity to humans of these lead compounds varies substantially, both with
chemical nature, particle size and type of exposure (inhalation, ingestion, contact with the skin).
Because of the many combinations of factors possible, a very approximate scale of high to low
toxicity spans lead oxide, lead carbonate, small particles of lead inhaled, to the least toxic lead
sulphide.242 The toxic effects of the lead fume that constituted the accretions to the smelting
furnace walls on the workers who were assigned to dislodge them at the end of a smelting run
240
Pure lead melts at 328° C and its boiling point is 1750° C. Modern studies on the composition of lead aerosols
from smelting lead indicate the presence of Pb, PbS, PbSO4, PbO, PbCO3 and others. USEPA, Air Quality
Criteria for Lead EPA/600/R-5/144aF, vol. I (2006), 2_4, 2_8. Lead fumes analysed in the nineteenth century,
when furnace conditions corresponded more closely to the period of this study, conform to this profile, but the
source of the fume (ie furnace conditions and presence of other compounds) determine which specific lead
compound predominates. See examples in John Percy, The Metallurgy of Lead Including Desilverization and
Cupellation (London: J. Murray, 1870), 451-58. Why metallic lead should be present in the aerosol is not evident
due to the high boiling point of lead. Measurable volatile lead is reported from 1,200 °C, in Katsunori Homma,
"Experimental Study for Preparating Metal Fumes," Industrial Health 4, no. 3 (1966): 132. Volatile lead is
reported from approximately 1,100 °C and volatile lead oxide from 550 °C, but the presence of sulphur and
chloride will shift the lead compounds to lead sulphate and very volatile lead chlorides, in Anders Ljung and
Anders Nordin, "Theoretical Feasibility for Ecological Biomass Ash Recirculation: Chemical Equilibrium
Behavior of Nutrient Elements and Heavy Metals During Combustion," Environmental Science & Technology 31,
no. 9 (1997): 2502. This is a range of temperature that could be attained in the furnaces of the period in question.
In the discussions that follow I will refer to ‘lead and its compounds’ to encompass all the lead speciation present
in lead fumes that are produced during smelting.
241
Percy, Metallurgy of Lead, 434-51.
242
The estate of research on the toxicity of lead and its compounds is extensively covered in USEPA, Air Quality
Criteria for Lead EPA/600/R-5/144aF, vol. I,II (2006). A more accessible guide as to the approximate order of
toxicity is the early article by Lawrence T. Fairhall, "Inorganic Industrial Hazards," Physiological Reviews 25, no.
1 (1945): 184-85.
145
is well documented by de Gamboa in his description of the smelting of silver ores at the end of
the nineteenth century in New Spain.243 The phrase ‘lead poisoning’ or ‘lead toxicity’ has to
be taken to refer not only to the metal itself but to its many toxic compounds.
2. Loss via solid particles spread by the wind from stockpiles of ore, greta or cendrada
within the hacienda compound (fugitive lead). Again the size, chemical nature and type of
exposure of the particles would determine their toxicity. Ingested greta or cendrada (containing
lead oxide) would be the most toxic, gross particles of the ore (lead sulphide) the least. Oral
ingestion by children of the workers of this fugitive lead would be a major problem.244
3. Loss of lead contained in the solid grasas dumped alongside the haciendas. This lead
(metallic lead, lead sulphide, other lead compounds) is encased in a solid matrix, either porous
like a lava stone or glassy. Leaching of the lead into the soil would be expected under mostly
acidic conditions.245
4. Sales of poor lead and/or greta as a by-product of the process. The last group is
contingent on the manner in which each individual smelting hacienda managed its business.
Lead-poor ores would require a total recycling of any greta or poor lead to minimize the
purchase of fresh additions of lead to reach the required lead to silver ratios in the smelting
recipe. Lead-rich ores would create a surplus of greta that would be thrown away as waste or
else offered to other mining localities that were deficient in lead. In San Luis Potosí legislation
243
de Gamboa, Comentarios Ordenanzas de Minas, 286.
244
The topic is addressed for modern cases of workers acting as transport vectors for lead compounds from the
workplace to the home, or of children poisoned by ingestion of lead contaminated soil or dust, in USEPA, Air
Quality Lead, I 3_17, 3_27, 3_28.
245
One study of the portioning of lead in the soil of historical lead smelting sites and the mobility of lead with
decreasing pH is J. E. Maskall and I. Thornton, "Chemical Partitioning of Heavy Metals in Soils, Clays and Rocks
at Historical Lead Smelting Sites," Water, Air, and Soil Pollution 108, no. 3-4 (1998). During the visit to the ruins
of the Hacienda Santa María in Monte Caldera, one of the family members who uses the remains of the hacienda
as an animal pen remarked that when they last held a barbeque and used grasas to line the fire pit, these exploded
with the heat and some of the family fell ill afterwards. It is only one anecdotal instance, but the heating of the
slag will liberate the lead content, among other toxic substances.
146
was in place that regulated this export of greta, forcing potential sellers to offer it locally during
nine consecutive public offerings (pregones) at a set price before it could be sold out of the
jurisdiction.246 Penalties for contraband export set in 1678 were high: confiscation of the greta,
the cart, fines, jail for the Spaniards, and 200 lashes for any indigenous workers or African
slaves caught participating in the act.247 Overall, the fact that it is documented that greta was
offered with no takers locally, or that contraband was attempted, would confirm indirectly that
lead was not an issue for the haciendas of San Luis Potosí. As for poor lead, it will figure in
my mass balance of Chapter 4 for the case study in Pachuca, but I did not find any evidence
for a market of poor lead in San Luis Potosí. My mass balance analysis will therefore focus on
the first three groups.
To calculate the emissions of lead and its compounds into the environment per kg of
silver refined by smelting I will base my mass balance calculation on the weight of lead coming
in and out of a smelting operation, regardless of the nature of the chemical compound of lead
involved. This allows me to arrive at a mass ratio without having to know exactly the profile
of lead compounds involved for each specific furnace condition. A more detailed study on the
toxic effects of this loss of lead will require a more detailed knowledge of the speciation of
lead products issued during smelting.
246
Request by Antonio Maldonado Zapata to sell in Sombrerete and Guanajuato 200 quintales of greta, AHSLP,
Fondo Alcaldía Mayor 1674.3, expediente 11, 31 August 1674; request by Dionissio de Rojas y Valdez to export
30 quintales of greta, AHSLP, Fondo Alcaldía Mayor 1674.4, expediente s/n, 18 September 1674; request by
Fernando de Vaca y Castro to offer locally or export 600 quintales of greta or of lead obtained therefrom, Fondo
Alcaldía Mayor 1680.2, expediente 10, 5 October 1680.
247
Capture of contraband of 27 cargas of greta destined for Guanajuato by Antonio Veles de la Torre, Alcalde
Ordinario of San Luis Potosí. An added incentive was the share of the proceeds from the sale of the impounded
contraband greta between the judge and the person who informs / captures the contraband. AHSLP, Fondo
Alcaldía Mayor 1686.1, expediente 11, 14 March 1686.
147
ATMOSPHERIC LOSS OF ATMOSPHERIC LOSS OF

LEAD + LEAD LEAD + LEAD
COMPOUNDS COMPOUNDS
Lead in silver ore
Lead in new greta
Reduction Oxidation in Optional: reduction of
silver- greta or Pattinson’s
in horno enriched lead reverberatory
process to produce poor
castellano furnace lead
Recycled lead,
greta, grasas,
cendradas and
furnace
accretions
FUGITIVE LEAD IN
SILVER POOR LEAD
LEAD GRASAS GRETA
Figure 2-10. Scheme of the mass balance for lead during the smelting of silver ores. Letters
in bold indicate mass input, letters in capitals indicate mass output.
Since both lead rich and lead poor silver ores were smelted, using added greta or
recycled lead to make up any deficiency in lead, the key is to know what is the average total
lead to silver ratio in the smelting recipe. The smelting recipes that appear in the historiography
do not provide sufficient information on the lead content of the ores or the quality of the greta
to allow a calculation of this ratio. There are however direct and indirect indications of this
ratio, as set out in Table 2-I.248 The indirect values are given as proportions of lead and lead
248
a) Barba proposes a ratio between 2:1 and 5:2 for ores rich in silver, and for smelting silver sulphides he
recommends a ratio approaching 4:1. Barba, Arte de los metales. b) A ratio of 1.5:1, subject to the lead and silver
content of the ores. Gómez de Cervantes, Nueva España siglo XVI. c) West quotes from a 1539 document that in
Taxco (New Spain), 25 hundredweight of litharge were required to refine 75 to 126 ounces of silver. Robert West,
"Aboriginal metallurgy and metalworking in Spanish America: a brief overview," in In Quest of Mineral Wealth.
Aboriginal and Colonial Mining and Metallurgy in Spanish America. , ed. Alan Craig and Robert West,
Geoscience and Man (Baton Rouge: Geoscience Publications, 1994). d) The ratios could range from just over 1:1
to 4:1 according to J. de Oñate, Nuevas leyes de las minas de España: 1625 edición de Juan de Oñate: con tratado
de re Metalica de Juan de Oñate (Santa Fe, New Mexico: Sunstone Press, 1998). e) de Sarria, Ensayo de
metalurgia. f) Hermosa, Manual de Laboreo de Minas. g) Bruno Kerl, William Crookes, and Ernst Otto Röhrig,
148
containing fluxes (greta, cendrada) to the amount of ore, without providing the total lead
content of the flux or the silver content of the ore. Thus the values in italic in the table have
been calculated based on a 100% content of lead in the flux and a 2% content of silver in the
ore. The former is an approximation,
weight lead per

location source
unit weight silver
100 to 200 New Spain, 16c a, 152-161

> 75 New Spain, 16c b, 158-159
250-450 New Spain, 16c c, 127
50 to 200 New Spain, 17c d, 81-85
200 New Spain, 18c e, 105
100 New Spain, 19c f, 254

200 to 300 Halsbrucke, 19c g, 219
100 to 1 Utah, 19c h, 355
Table 2-I. Published weight ratios of lead to silver used in the smelting of silver ores. Sources
are given in footnote 248.
based on the fact that greta with an 82% content of lead would have been the main substitute
to pure lead in the recipe. The latter assumption is based on data presented in Chapter 4. They
fall within the same order of magnitude, 100 to 300, and apply on both sides of the Atlantic.
A Practical Treatise on Metallurgy Adapted from the Last German Edition of Prof. Kerl's Metallurgy (London:
Longmans, Green, and Co., 1868). h) smelted ore had a silver content of 0.16% and a lead content of 16%, in
Eissler, The Metallurgy of Argentiferous Lead. Garcés y Eguía states that he uses one quintal of greta for every
six marks of silver in an ore, but he does not provide the lead content of the ore. Garcés y Eguía, Nueva teórica
del beneficio de plata. There are other higher published ratios but they correspond to recipes for assaying small
quantities of ore by smelting.
149
There is an operational record dated 1718 that provides confirmation of the ranges
reported in Table 2-I. It forms part of a bundle of accounts rendered by Andres de Soliz related
to his management of a ‘fuelle’ (a bellows) in Veta Grande (Zacatecas), ‘en que me ocupó de
mayordomo en un fuelle que tenía en Beta Grande’, belonging to Captain Don Salvador de
Inostrosa.249 ‘Fuelle’ in this context is a smelting furnace situated close to the minehead. The
specific account that is of interest is signed by Marcos Alcay and begins with an invocation to
‘Jesus, María y Joseph’, followed by the title: ‘Book of charges and discharges of lead and ore
that were received in this smelter of captain Don Salvador de Inostrosa … August 16 1718
year’. It contains a record of individual smelting runs carried out from August 16 to September
15 in the year 1718, registering the amount of lead added to a specified quantity of silver ore,
and the total amount of silver obtained from the operation.250 I have included in Table 2-II the
information provided in the document, except for two runs where the data are not clear, and
my calculations of the minimum silver content of the ore being smelted, and the resulting lead
to silver weight ratio for each run. The results show that the richer silver ores required less lead
for smelting, and that a ratio of 100 to 1 can represent the operational range of the lead to silver
weight ratio used to smelt silver ores with approximately 2% silver content.
249
AHEZ, Serie Civil C15-E08.
250
‘Libro de cargo y descargo del plomo y metal que resivo en este fuelle del capitán Don Salbador de Inostrosa,
Agosto 16 de 1718 años’, AHEZ, Serie Civil C15-E08, 13r, 17 r,v.
150
lead added ore silver smelted silver in ore lead to silver

year 1718
quintales arrobas quintales arrobas marks oz % minimum ratio
8 2 4 2 9 2 1.03% 184
16 August
2 3 5 4 3 1.75% 126
21 August 5 1 3 8 1.33% 131
2 3 0 6 10 3.33% 55
2 2 6 7 3 2.46% 68
26 August
3 3 9 4 0.89% 188
2 1 5 3 1.20% 150
01 September 5 2 3 1 10 2 1.58% 107
7 1 4 19 2.38% 76
05 September
5 9 19 4.22% 53
15 September 8 4 1 14 2 1.68% 112
19 September 4 2 1 3 4 0.78% 229
Table 2-II. Range of lead to silver weight ratios from individual smelting runs carried out
in the region of Veta Grande, Zacatecas, in 1718. The source data are from footnote 250.
With respect to the losses of lead, they have been reported either as total losses, losses
in slags or losses as lead and lead compounds to the atmosphere. Table 2-III sets out the range
of published values.251 According to Biringuccio it was preferable to lose a certain amount of
lead through losses to the atmosphere as lead fume in the last step of cupellation rather than to
lose silver entrained during the removal of the last traces of litharge.252 The sources derive their
values from European lead smelters where there was an economic incentive to avoid as much
as possible losses of lead. Thus these values are minimum values, since this incentive did not
251
a) Ian Blanchard, "Technical Implications of the Transition from Silver to Lead Smelting in Twelfth Century
Britain," in Boles and Smeltmills Seminar, ed. Lynn Willies and David Cranstone (Reeth, Yorkshire: Historical
Society, Ltd., 1992). b) Michael C. Gill, "Analysis of Lead Slags,"ibid. c) Jerome O. Nriagu, "Tales Told in
Lead," Science 281, no. 5383 (1998). d) Eissler, The Metallurgy of Argentiferous Lead. e) Lynn Willies,
"Derbyshire Lead Smelting in the Eighteenth and Nineteenth Centuries," Bulletin of the Peak District Mines
Historical Society 11(1990). f) Pique, A Practical Treatise on Silver. g) Phillips, Metallurgy Silver. h) Rivot,
Description des gites métallifères. i) Danuta Molenda, "La metallurgie du plomb en Pologne au moyen age et aux
XVIe - XVIIIe siecles. ," in Mines et métallurgie ed. Paul Benoit, Les chemins de la recherche (Villeurbanne:
Programme Rhône-Alpes recherches en sciences humaines, 1994).
252
151
total as slag in lead fume process source
12 lead smelting 14c a, 10

6 5.4 lead smelting 19c
one to four, ratio of loss as slag to slag in castillian furnace b, 53

volatile lead 19c
10 to 25 cuppellation, early
c, 1623
2 to 5 cuppellation, 19c
Nevada mid 19c silver

up to 40 d, vii
smelting
5 to 20 other smelting furnaces
75 Spanish slag hearth e, 13
over 10 other smelting furnaces
silver ores with one step

7 f, 59
smelting
14 cupellation 19c g, 488
17 5 12 blast furnace19c g, 480
10 6 to 10 cupellation, 19c h, 187
cupellation Poland, 13 -
30 i, 53
18c
Table 2-III. Range of percentage values for lead losses during smelting of lead ores. Sources
in footnote 251.
necessarily apply to silver refiners in New Spain, where in some localities the lead content in
the ore would have sufficed to smelt without added greta, and thus any greta produced would
have had to be disposed of or sold. At the limit, even a total loss of lead during refining could
152
have been accommodated without impacting the silver refining profit. 253 I will therefore
assume as a conservative scenario that 5 to 10 % of the total lead mass input was lost as lead
and lead compounds to the atmosphere in both heating stages of the smelting process.254
This implies as a minimum a range of 5 to 10 kg of lead and lead compounds issued to
the atmosphere per kg of silver refined. With respect to the lead content that remained trapped
by the mounds of grasas around each smelting hacienda, Table 2-IV summarizes what has
been measured at silver smelting sites from various historical periods.255 The lead content of
two samples taken from the surface of the mounds of grasas that line the southwest side of the
Hacienda Santa María in Monte Caldera was measured, and gave the results shown in Table
2-V.256 The values for lead fall within the expected average range for an efficient smelting
253
A similar assumption has been applied to other historical silver smelting sites: ‘the principal objective of most
ancient smelting operations seems to have been the recovery of silver from argentiferous lead minerals and not
the production of lead metal. Under such circumstances significant lead losses are unlikely to have been
considered disadvantageous’. Paul Budd et al., "The Possible Fractionation of Lead Isotopes in Ancient
Metallurgical Processes," Archaeometry 37, no. 1 (1995): 148.
254
This analysis concurs with the conclusions reached by Collins in his textbook that : ‘there are very few figures
obtainable on this point [loss of lead fume during smelting of silver rich ores]. The average loss of lead … on the
whole refining process is supposed to vary between 3 and 8 per cent, by far the largest part of which is in the
cupellation’. Collins, Metallurgy of Lead & Silver, Vol. I, 347. In the early centuries I would expect the loss in
New Spain to have been substantially higher. For example, in the nineteenth century a “Spanish slag hearth’ was
introduced in England to smelt lead ores. The diagram of this furnace and dimensions are similar to that of an
horno castellano. Up to 75% of lead was lost when using this furnace, according to Willies, "Derbyshire Lead
Smelting," 13. A report dated 1802 on smelting tests carried out at Catorce (reproduced in full in Appendix C)
lists a total loss of 18 arrobas (207 kg) of ‘perdida de liga’, loss of added lead flux, incurred in refining 7.5 marks
(1.7 kg) of silver. This is a ratio of over 100 to 1, of which according to the working assumption of this section 5
to 10% was lost as lead and lead compounds to the atmosphere and over 90% was lost mainly as slag.
255
a) Gill, "Analysis of Lead Slags." b) John Percy, Metallurgy. The Art of Extracting Metals from their Ores.
Silver and Gold, vol. I (London: J. Murray, 1880). c) Willies, "Derbyshire Lead Smelting." d) Phillips, Metallurgy
Silver. e) I C Freestone et al., "Role of Materials Analysis in the Reconstruction of Early Metal Extraction
Technology: Zinc and Silver-Lead Smelting at Zawar, Rajasthan," Materials Research Society Symposia
Proceedings 185(1990). f) Bachmann, "Archäometallurgische Untersuchungen zur antiken Silbergewinnung in
Laurion. II. Charakterisierung von Bleiverhüttungsschlacken aus Laurion= Archaeometallurgical Investigation on
Ancient Silver Smelting at Laurion. II. Characterisation of Lead Smelting Slags from Laurion."
256
The analysis was carried out by a commercial laboratory (Actlabs, Ontario, Canada). Samples were first
prepared by milling and then digested with sodium peroxide. Sulphur was measured by Infrared (IR) analysis and
lead and sulphur by Instrumental Neutron Activation Analysis (INAA). No procedural blanks were run. A certified
sample was run by Actlabs for every analytical procedure used. The results of the certified samples indicate that
lead values are within 5.5%, sulphur within 2.5% and arsenic within 1%. A third sample was measured by Prof.
Salazar González and gave 3.35% lead (personal communication), method unknown. Electron microscopy of
sections of these samples show the presence of galena particles within the matrix of the fused slag (courtesy of
Prof. Raynald Gauvin and Mr. Nicholas Brodusch of the Department of Materials Engineering of McGill
University).
153
operation, if no leaching has taken place over the centuries. Arsenic levels are high, just
approximately 30 to 40 times less on average than the lead content.257 Two samples taken from
the surface of the mounds, thus from the most recent historical period, are not statistically
significant, and the analysis was undertaken solely as an initial probe into the probable
chemical make-up of the slags. Even with these limitations in mind, the limited results
definitely point to the need to carry out a formal study with historically relevant sampling areas
and sample sizes, together with the required chemical analysis and leaching measurements to
determine their long-term effect on the groundwater. During smelting arsenic would have been
average lead in
range (%) location source
slag (%)
3 0.15 to 11 England, 19c a, 53
8 to 25 general b, 280
10 to 12.5 c, 10
England, 19c
2 to 8 c, 12
most favourable
2 conditions silver d, 478
smelting 19c
India, first
4 to 13 e, 619
millenium
15 9 to 25 Laurion, antiquity f, 248
Table 2-IV. Lead content in slags from different smelting sites and periods. Sources in
footnote 255.
257
1% is equivalent to 10,000 ppm.
154
another toxic element present in the lead fume, though primary sources even up to the
nineteenth century do not provide much guidance in this regard. The limited measurements of
Table 2-V indicate that lead is not the only toxic element present in these mounds of waste.258
Though arsenic will not figure in the subsequent discussion in this work on the environmental
impact of historical silver refining, it has already received attention in current research in
Mexico.259
Lead Arsenic Sulphur

sample
(%) ( ppm) (%)
1 2.67 832 0.8
2 4.27 964 0.55
Table 2-V. Lead, arsenic and sulphur content of two samples of grasas from Monte
Caldera.
Sulphur levels are inconclusive, since in sample 1 they would be consistent with high
levels of galena, which was identified via electron microscopy, while in sample 2 they are much
lower than expected if all the lead was in the form of galena.260
258
I would like to thank Prof. Pamela Welbourne for arranging a presentation on my on-going research at Queen’s
University, Ontario, where the presence of arsenic in silver ores was brought to my attention by Dr. Geme Olivo.
259
See for example Yolanda Jasso-Pineda et al., "An Integrated Health Risk Assessment Approach to the Study
of Mining Sites Contaminated With Arsenic and Lead," Integrated Environmental Assessment and Management
3, no. 3 (2007).
260
Electron microscopy of sections of these samples show the presence of galena particles within the matrix of
the fused slag. Particles of fused lead sulphide are said to indicate lower furnace temperatures, but the results
obtained for these samples is too limited in scope at present to attempt to recreate the smelting conditions that
gave rise to them. See Gill, "Analysis of Lead Slags," 51.
155
Table 2-VI is an estimate of the magnitude of lead losses from an average smelting
hacienda fitted with four smelting furnaces (hornos castellanos) and refining an ore with an
average of 2% silver. The ratios are those derived in the previous sections, and I further assume
smelting furnaces 4
silver per furnace 345
total 1,380
incoming ore at 2% 69,000
solid waste generated 67,600
lead + compounds to kg/y
6,900 to 13,800
atmosphere
weight of slag 67,600
3% lead in slag 2,000
fugitive lead unknown
Table 2-VI. Assumptions applied to the mineral and lead mass balance for the Hacienda
Santa Maria, Monte Caldera. For details and sources see text.
that all the non-silver solids in the incoming ore is incorporated into the waste slag. 261 Losses
via fugitive lead are impossible to estimate, so I acknowledge them in the table but cannot
quantify them. Fugitive losses would be located all around the soil and walls of the hacienda
compound, and would also include lead taken out of the compound entrained on the skin and
clothes of workers and their families, or on pack animals.262 The outstanding figure is that for
261
‘slag consisted of … unburnt ore, partially oxidized or reduced ore, gangue (non-metallic materials), and
metallic lead’ in Willies, "Derbyshire Lead Smelting," 2.
262
Modern environmental impact assessments around lead smelting facilities have concluded that ‘fugitive
emissions, or those from non-point sources, such as transportation routes and smelter floors, are major sources of
particulate matter, even exceeding stack emissions, but only within the smelter confines’. Measurements at a
modern smelter site in Trail, British Columbia, Canada, show that from 41% to 87% of total lead losses are
attributable to secondary sources. Fariborz Goodarzi et al., "Sources of Lead and Zinc Associated with Metal
156
losses of lead and lead compounds to the atmosphere, with a minimum range from
approximately 7 to 14 tons of lead per year from an average sized smelting hacienda.
2.6.2 Lead: its directionality
In 1761 a refiner by the name of Manuel Correa started digging the foundations for a
new smelting hacienda in the Real de Pánuco, Zacatecas. In what might be one of the shortest
colonial civil suits on record in New Spain (it was initiated on the 28th March and an agreement
reached between the parties on the 8th of April 1761) a group of townspeople objected to the
new hacienda on the grounds of the toxicity of its smoke and its effect on the neighbours and
nearby church.263 The wording of the complaint and of the defence made by Correa are worth
quoting extensively since they make very clear that for the communities the most evident
environmental impact of a smelting hacienda was through the smoke of its smelting operation:
‘Juan Estevan y Francisco Messa and all the other residents of the Real de Panuco … that were
named in the previous writ … [on] the opposition to the construction of an Hacienda for
extracting silver by fire in the proximity of said Real by Manuel Correa … since with its smoke
… it will add to the imminent damage to all the neighborhood, since not a single expert in
Medicine exists that does not disapprove of the noxious, diffuse and extended turbulence of
the air over the Real of the pestiferous smoke from diverse ores, and ingredients, that will be
poison to the health of the inhabitants … no smelting haciendas are found in centres of
population knowing that … it [the smoke] harms so that it kills all the animals it comes
across’264
Smelting Activities in the Trail Area, British Columbia, Canada " Journal of Environmental Monitoring 4, no. 3
(2002): 403. However, I have no way of quantifying historical fugitive losses, since the modern percentage figures
from Trail cannot be extrapolated to colonial smelting haciendas. The modern smelter at Trail fitted with a 100 m
chimney stack and filters to maintain a low level of lead in the flue gas increases the relative importance of fugitive
emissions as a percentage of the total losses of lead. In historical smelting haciendas, with a much higher expected
amount of lead in the flue gas since no attempt was made to recover or contain the flume, the relative percentage
of the fugitive emission to total emissions would be substantially lower.
263
An earlier but similar case of complaint against the construction of a smelting hacienda has been documented
in Bernd Hausberger, "Una iniciativa ecológica contra la industria minera en Chihuahua (1732)," Estudios de
Historia Novohispana 13, no. 13 (1992). It is interesting that in both cases the parish priest took the side of the
smelter being questioned.
264
‘Juan Estevan y Francisco Messa, y todos los demás vecinos del Real de Panuco … que se nombraron en el
anterior escrito … [de] la oposición a que se construya Hazienda de sacar plata por fuego próxima a dicho Real
por Manuel Correa … pues con sus humos … se agregara el daño inminente a todo el vecindario, pues no abra
perito en Medicina que no desapruebe la nosiva turbulencia difusa, y estendida por los aires sobre el Real de los
humos pestíferos de diversos metales, e ingredientes, que serán veneno de la salud de los moradores … no se
157
Even the irony in the argument put forward by Correa to defend his decision to build
the smelting hacienda is built on the toxic nature of the smoke:
‘Unhappy would be the city of Zacatecas if it consented that in its centre be built similar
haciendas, that all would live either sick, or bothered by their smoke: but what am I saying?
Can my adversaries deny, that they have right in the middle of the city four haciendas
surrounded by many houses?265
In the concluding writ, where Correa desists from his intentions to build the hacienda
after agreeing that both parties should share the costs of the litigation, the negative view on the
nature of the smoke is repeated: ‘ because of the harm to the neighborhood of said Real from
the sulphurous particles in the smoke from said ores’.266
The problem with the smoke was not so much sulphur as lead, as evidenced in the levels
of lead and lead compounds issued to the air as a result of smelting silver ores presented in the
previous section. In Figure 2-11 I use the reconstruction of the Hacienda Santa María (Monte
Caldera) proposed by Salazar Gonzalez to illustrate the directionality of these loss vectors of
lead and lead compounds issued to the atmosphere.267 The first area within the compound of
high concentrations of lead in the air is the working space around and in front of the smelting
furnaces. The lead barras, or pigs, were handled in the open area in front of the furnace where
the molten lead flowed into its moulds, so there is no doubt ambient lead levels were much
descubre introducir haciendas de fuego en las poblasones sabiéndose que … perjudica de modo que es homicida
de quantos animales encuentra’ AHEZ, Serie Civil C37-005, 28 March 1761
265
‘Infeliz fuera la ciudad de Zacatecas si consintiera que en su centro se fabricasen semejantes haziendas que
todos vibirian , o enfermos, o fastidiados de los humos de ellas: mas que dije? Podran negarme los adversarios ,
que en su puro medio tiene esta ciudad quatro haziendas rodeadas de muchas casas?’ arguments by Manuel
Correa, 31 March 1761, AHEZ, Serie Civil C37-005.
266
‘por lo perjudicial que pueden ser al besindario de dicho Real las partículas asufrozas que enbuelben en si
los humos de dichos metales’, agreement to suspend litigation between the parties, 8 April 1761, AHEZ, Serie
Civil C37-005.
267
Salazar González, Las haciendas de San Luis Potosí, 432.
158
higher than any modern industrial and occupational guideline.268 The second area of high risk
to the hacienda workers was the refining furnace, graphically described in the nineteenth
century in terms that leave no doubt as to the toxicity of this space for the workforce:
‘these ovens do not have chimneys, and the smoke exits the furnace at the place where the
circumference of the cupel ends … the smoke and the lead vapours that coat the walls in lead
oxide rise in a thick column under a pyramid similar to those of a castillian furnace’269
HIGH LEVELS OF LEAD AROUND FURNACE WORKING AREAS
LEAD IN FUME
FUGITIVE Mineral
LEAD waste,
including
lead
compounds,
LEAD IN in waterways
SLAGS
Figure 2-11. Main areas of lead deposition within and around the reconstruction of the
Hacienda Santa Maria in Monte Caldera, adapted from original drawing with permission of
Prof. Guadalupe Salazar González, footnote 267.
Whatever fraction of lead fume escaped through each chimney from the hornos and
vasos was then deposited around the surroundings of the hacienda leaving a footprint that
268
There is a very illustrative drawing of a blast furnace at work, showing pigs being formed in the area in front
of the furnace hearth. While part of the hearth may have been under a hood, the ambient levels of lead fumes
would have been substantial in all this work area. Mark Bowden, Furness Iron : the Physical Remains of the Iron
Industry and Related Woodland Industries of Furness and Southern Lakeland (Swindon: English Heritage, 2000),
4, 52.
269
‘ces fours n’ont pas de cheminées, et la fumée sort du fourneau à l’endroit où se termine la circonférence de
la coupelle, … la fumée et les vapeurs plombeuses qui tapissent les paroirs en oxyde de plomb s’élève en colonne
épaisse sous une pyramide semblable à celle des fours castillans’ in Duport, Métaux précieux au Mexique, 71-
72.
159
mirrors the shape of the wind rose at each locality. The thick column of smoke described by
Duport is well evident in the photograph in Figure 2-12 of a nineteenth century smelting
hacienda. Two smelting furnaces and possibly one refining furnace are at work, and the thick
smoke has to be pictured issuing from an horno castellano at just above the eye-level of the
workers in the early centuries of smelting in New Spain, not at over 10 m as in the photograph.
Contrary to Correa’s argument that ‘as every idiot knows, the centre of smoke are the aerial
regions’, in the period under study much of this lead rich smoke would have been closer to the
ground on which lead would ultimately deposit.270 The exact distribution of the lead deposited
between hacienda compound, immediate vicinity (i.e. over fields of grasas) and long-distance
spatial spread remains to be established. For the workers involved in scraping the inside of
furnaces and chimneys between smelting runs, the dust would represent an additional source
of ingested lead and lead compounds.271
What is reported is that the concentration of lead deposited in the areas surrounding a
smelter could be so high as to cause the deaths of livestock and other animals, a fact
acknowledged in the extracts quoted above from the legal writ against Correa. This
phenomenon was well known in England in the eighteenth and nineteenth century: ‘fumes
emitted from Cupolas or low arched reverberatory furnaces … poisoned herbage for ¼ mile or
more around the Cupolas, and owners were obliged to pay a rent to the farmers for the damage
caused to their land’. The poisoning of cattle and humans in Derbyshire went by the local term
of Belland or bellanding, and hens, sheep and even dogs were also poisoned by the lead
deposited from the smoke of the smelters. No statistics were recorded for any of these
270
‘Siendo sabido a el mas idiota, que el centro del humo son las regiones aereas’ writ by Manuel Correa, 31
March 1761, AHEZ, Serie Civil C37-005.
271
de Gamboa, Comentarios Ordenanzas de Minas, 286.
160
incidents.272 In England lead smelters started to recover fume as of 1792, and especially as of
mid-nineteenth century, by installing long flues attached to the exit of the chimneys.273 No such
measures of control have been detected in the Hacienda de Regla run by English management
and investment in the first half of the nineteenth century (see Chapter 4).
Figure 2-12. Digital copy of photograph by Charles Waite titled ‘Mexican Adobe Smelter
Taxco Guerrero’, 1905, number 13 in the series Tema y Tecnología (CIG-AGN).
272
I. Thornton and P. Abrahams, "Historical Records of Metal Pollution in the Environment," in Changing Metal
Cycles and Human Health, ed. J. O. Nriagu (Heidelberg, New York, Tokyo: Springer-Verlag, 1984), 12. For
studies on the lead concentration around historical lead smelting sites in England and the incidence of bellanding
even in recent times from historical lead deposits is discussed in Mike Wild and Ian Eastwood, "Soil
Contamination and Smelting Sites," in Boles and Smeltmills Seminar, ed. Lynn Willies and David Cranstone
(Reeth, Yorkshire: Historical Society, Ltd., 1992). See also Willies, "Derbyshire Lead Smelting," 3, 13. Similar
cases of compensation are not reported in the historiography in New Spain nor have I found any in my non-
exhaustive search of the records from San Luis Potosí, Zacatecas, Guanajuato and Pachuca. One reason may be
the common ownership of refining haciendas and cattle/agricultural haciendas as documented by Salazar
González, Las haciendas de San Luis Potosí, 153-210. and Eugenio Martín Torres, El beneficio de la plata en
Guanajuato, 1686-1740 (Guanajuato: Presidencia Municipal de Guanajuato, 2001), 162-166.
273
‘fume is the general name given to the usually greyish white, feathery , partially crystalline , partially dusty
deposit [lead sulphate and oxides, arsenic, silver] which sublimates or adheres onto the sides of chimneys and
other flueways along which gaseous material from the furnaces pass’ in Willies, "Derbyshire Lead Smelting," 13.
For the installation of flues see Gill, "Analysis of Lead Slags," 53.; Willies, "Derbyshire Lead Smelting," 3, 14.
The use of chambers to condense and extract metal fumes from the furnace appears as early as Agricola, De re
metallica, 394. See also Pique, A Practical Treatise on Silver, 72.
161
Owners and builders of the smelting haciendas were conscious of the need to site these
facilities taking into account the prevailing winds of the region. Again the civil suit against
Correa provides interesting insights:
‘there being more than two hundred paces between my church and home and the place Manuel
Correa intends to build his hacienda, and being this location at a greater height than the church,
and my home, without the winds from the South and West (as it seems to me) being able to
push the smoke so that it can harm the church or me … since it seems to me only the winds
from the Southwest could cause harm according to what was established by the Architects for
the construction of this Hacienda’274
Indirect evidence of the care taken to integrate the wind direction with the orientation of the
smelting furnaces within the haciendas comes from the reconstruction of the smelting
haciendas in the area of Monte Caldera, San Luis Potosí. Figure 2-13 aligns the architectural
footprint of three haciendas published by Salazar González with the direction of due North.275
The three arrays of smelting furnaces appear remarkably parallel to each other, and the only
common factor to all would be the predominant wind direction in the area of Monte Caldera.
Even without specific knowledge on the wind rose of the area during the colonial period, the
parallelism observed in Figure 2-13 indicates that local builders took it into account when
locating the smelting furnaces within the hacienda compound, so as to minimize the impact of
the plume on the compound itself.
In addition to the main loss via lead fume, three other directions for the loss of lead
need to be considered. One is within the compound, in the form of fugitive lead loss, from the
274
‘haviendo de la iglesia y de mi casa mas de doscientos pasos a el lugar adonde intenta Manuel Correa poner
su hazienda, y hayarse este sitio en gran altura respecto de la Iglesia, y de mi casa, sin que los Aires de Sur y
Poniente (según me pareze) puedan aviolentar a el humo para que perjudique o a la Iglesia o a mi … por
parecerme ser solo el aire Suroeste el que hubiese de dar algún perjuicio según lo determinado por los Arquitectos
para la favrica de dicha Hazienda’, sworn statement by the parish priest, Don Joseph de Siloa, in support of
Manuel Correa, 28 March 1761, AHEZ, Serie Civil C37-005.
275
Salazar González, Las haciendas de San Luis Potosí, 428, 431, 436.
162
Hacienda Buena Vista
Hacienda La Luz
Smelting furnace
Hacienda Santa Maria
Cupelling furnace
note: each plan is drawn to a
different scale
Figure 2-13. Relative alignment of arrays of furnaces from three different smelting haciendas
in the area of Monte Caldera, adapted from the architectural reconstructions in footnote 275.
physical loss of lead or litharge as dust. The second direction is through leaching of lead and
lead compounds from the slag heaps into the soil and water table. It is important to know the
nature of the lead compound in the grasas since it will determine in part its extraction rate to
the environment under long- term atmospheric exposure. A more systematic study of the
163
environmental impact of these grasas is required, that includes establishing the leaching
behaviour of the grasas and its potential consequences to the local population, livestock, crops
and the water sources of each area.
The remaining vector of lead emissions from each hacienda was the waste water from
the washing of the ores. ‘The process of Buddling – separating lead ore from gangue materials
using water- used to poison the streams’, made rivers suddenly turn yellow with the amount of
mineral sediment voided into them, and fish died and cattle were poisoned from drinking this
industrial waste in the water.276 The magnitude of this vector can only be established comparing
the amount of lead in the mined ore with the amount of lead in the dressed ore, but these data
are not yet available for the period in question.
2.6.3 Lead: its source
The absence of lead in silver ores was never an obstacle to smelting them. In Europe
since at least the mid-fourteenth century there had been a traffic of lead to supplement the so-
called ‘dry’ ores: ‘by the 1330s Polish lead was utilised for smelting and refining the “dry”
gold and silver ores processed in the metallurgical enterprises of the Hungarian-Transylvanian
Carpathians and the Bohemian-Moravian and Saxon Erzegirbe.’277 In fact, the lead industry of
Europe was forced to respond to the increase in demand once the copper liquation process was
applied to argentiferous copper ores. By the 1540s England was already an important supplier
of lead to the silver/copper refining centres of Europe, which required around 4,800 tons of
lead per year and absorbed 60-85% of an English lead production of 3,200 tons.278 Then, from
1543 to 1549, the traditional European market chain for lead destined to silver-copper refining
276
Thornton and Abrahams, "Metal Pollution in the Environment," 12.
277
278
Kellenbenz, "Final Remarks Silver Production," 332.
164
was disrupted by the appearance of English lead stripped from Church holdings that halved the
prices as of 1538-1539.279
Local sources of galena at Taxco and Sultepec had been sufficient initially to provide
the required quantities of lead for smelting, but soon the whole dynamic changed and imports
of lead to New Spain were initiated.
‘From 1536 the trade, based now almost entirely on monastic lead, passed directly to Seville
[and] shipped to the New World ... abundant and cheap supplies of lead benefited the bullion
producers of central America ... via Seville English lead, carried as ballast by the Indies fleets,
passed to Veracruz and thence northwards to satisfy the demands of the producers in the
booming centre of Zacatecas, where the deposits, though rich in silver, were poor in lead ...
[this would create] a dependence which left the central and south American smelter
dangerously exposed ... by 1554 prices rose again on the Seville market and [New World]
producers were drawn into the general bullion crisis’ [emphasis added].280
As demand rose from a European industry requiring more lead per unit of silver refined
and a New World providing further new silver deposits to Spain refined on the basis of smelting
with lead, a brake on English lead exports brought about strong pricing fluctuations that saw a
threefold rise in prices in the early 1560s (compared to prices in the 1540s) until they dropped
to a third above 1540s level. Between the 1560s and late 1570s lead supply would be re-
established from various European sources, but by that time mercury was displacing lead as
the main refining agent of choice by Spain to the point that by 1572 an English visitor in Central
America could say
‘as for this charge of quicksilver it is a new invention, which they find more profitable then to
find their use with lead ... wherefore they shall not need any of our lead, as they have neede
thereof in times past.’281
279
Blanchard, "England and the International Bullion Crisis of the 1550s," 87.
280
Ibid., 89, 90, 107.
281
Ibid., 95-108., citing R. Hakluyt, The Principall Navigations Voiages and Discoveries of the English Nation
(1589), II, p. 549
165
According to Blanchard English lead would continue to be exported to New Spain from
1580 to 1590, from 1610 to 1614 and from 1625 to its final phase in the 1640s, after which
point it was displaced by local lead sources in New Spain.282 From the mid-seventeenth century
onwards the movement of lead, litharge and silver-rich ores between mines and refining
haciendas in the same or different Caja districts is not documented in the historiography. This
intra- and inter- regional transit of silver-rich lead ores and lead fluxes required for smelting
would determine the balance between amalgamation and smelting, as will be seen in Chapter
6.
2.6.4 Charcoal and wood: its magnitude
Charcoal was the main fuel requirement for smelting in New Spain, while wood was
the minor fuel for cupelling in the reverberatory ovens.283 Charcoal was necessary not only to
provide the heat required to reach the necessary furnace temperatures, but also to act as
reducing agent during smelting. Charcoal was used to smelt all metal ores, but the quantities
required varied according to the metal. To smelt copper during this period the weight ratio of
charcoal to copper was in the range of 20:1 to 50:1.284 In the case of iron, the ratio was 30:1
approximately.285 Lead required less charcoal, with values as low as 3:1 or 6:1 being reported
for ores containing 45% lead.286 In sharp contrast, for the smelting of silver ores the metal to
charcoal weight ratios reported in the historiography are considerably higher. Between 975 and
1,145 kg of charcoal are reported as being required by the smelting operations at Wissenbach
282
Russia's "Age of Silver". Precious-metal Production and Economic Growth in the Eighteenth Century 19. In
Chapter Six I will come back to the important question as to whether there were sufficient endogenous lead
deposits in New Spain to have met the requirements for smelting, and the attitude of the Spanish Crown with
respect to prospecting for lead.
283
Duport, Métaux précieux au Mexique, 81. In Chapter 4 the large difference in fuel requirements for charcoal
and wood in smelting will be fully documented, so for the rest of this chapter I will only focus on charcoal.
284
Craddock, Early Metal Production, 193.; Salkield, "Ancient Slags in the South West of the Iberian Peninsula,"
93.
285
Nef, Conquest of the Material World, 174.
286
166
(Europe, Vosges, late sixteenth century) to produce 1 kg of silver from ores containing between
0.07 and 0.13% silver. Rech also cites data from smelting of silver ores in the eighteenth
century (nature of ore and silver content not given), that required between 800 and 960 kg of
charcoal to produce 1 kg silver.287 Studnicki-Gizbert and Schecter have used the information
in the accounting books of two smelting haciendas in New Spain to calculate a ratio of 1,185
kg of charcoal per kg of silver in San Luis Potosi (1611-1612) and 1,168 kg of charcoal per kg
of silver in Pachuca (1782-1783), though the silver content of these ores is not given.288 I will
thus use as a working figure an average weight ratio for charcoal of 1,000:1 up to the nineteenth
century, when the use of more efficient blast furnaces will decrease substantially this ratio
(Chapter 4).
2.6.5 Charcoal and wood: its directionality and source
The directionality of this vector is determined by the location of the woodlands that
ultimately provided the charcoal to the smelting furnaces. These vectors were not distributed
radially around each refining nucleus, but could extend like tendrils for many kilometres in a
single direction. In the case of the region around San Luis Potosí, charcoal was sourced as far
away as Peotillos (Figure 2-14). In order to estimate the amount of natural resources required
to meet this high consumption of charcoal it is necessary to proceed in stages. Charcoal is
obtained from wood, wood is obtained from forests, and forests are organic systems that can
be regenerated given sufficient time. To estimate the amount of charcoal generated from a
hectare (ha) of woodland, the following factors have to be taken into account: a) individual tree
species can produce quite distinct amounts of charcoal per cubic metre of wood b) the moisture
287
Georges Rech, "La fonderie de Wisenbach (Vosges)," in Mines et métallurgie ed. Paul Benoit, Les chemins de
la recherche (Villeurbanne: Programme Rhône-Alpes recherches en sciences humaines, 1994), 178-81.
288
Studnicki-Gizbert and Schecter, "Colonial Fuel-Rush," 112.
167
content also varies according to the state of the wood and c) the type of charcoal making process
will also determine the efficiency of the operation. The average amount (expressed as volume
in cubic metres) of growing stock (trees above a certain height and diameter) per ha of forest
‘is an estimation of how well or how poorly stocked the forests are’. For the year 2010 the
Food and Agricultural Organisation of the United Nations (FAO) has calculated an average of
44 m3/ha for Mexico.289
As to the efficiency of the transformation from wood to charcoal, the FAO establishes
a range for developing countries of 10 to 27 m3 of wood required to produce 1 ton of
charcoal.290 Using the average density of wood reported by FAO (650 to 750 kg per m3), this
range of conversion of 5 to 15% of the weight of wood to weight of charcoal overlaps fairly
well with the estimation in the historiography that in general the amount of charcoal
corresponds to 10 to 20% of the weight of wood.291 Thus using an average of 18.5 m3 of wood
per ton of charcoal (approximately 10% conversion), and a charcoal to silver weight ratio of
1,000 to 1, the amount of forest area required to produce enough charcoal for the refining of 1
kg of silver in New Spain (Mexico) would have been 0.4 hectares in New Spain (Mexico).292
Studnicki-Gizbert and Schecter estimate that one kilogram of silver consumes 6,332.8 m2 (0.63
ha) of woodland, based on reported data by other authors from mesquite growth and charcoal
production in the arid regions of Arizona.293 Their projection is within the order of magnitude
predicted using FAO data.
289
Global Forest Resources Assessment 2010, (Rome: Food and Agriculture Organization of the United Nations,
2010), 270-271. http://www.fao.org/docrep/013/i1757e/i1757e.pdf , accessed 15 July 2012. For the other silver
producing regions the values are 74 m3/ha for Bolivia and 120 m3/ha for Peru
290
http://www.fao.org/docrep/Q1085E/q1085e0c.htm.
291
Table 11 in http://www.fao.org/docrep/007/j4504E/j4504e08.htm; Craddock, Early Metal Production, 193.
292
0.2 hectares in colonial Peru (average of modern day Bolivia and Peru).
293
Studnicki-Gizbert and Schecter, "Colonial Fuel-Rush," 112. In my limited discussions with local historians
and residents it seems encino (evergreen oak) and not mesquite was the tree of choice for mining and charcoal
manufacture around San Luis Potosí, Zacatecas, Guanajuato and Pachuca.
168
A projection can also take into account the capacity of forests to regenerate themselves
in the space of decades, even without the assistance of modern forestry methods. Values vary
in the published literature, but for forests in the tropical zone the above ground biomass has
been reported to be recoverable in a span of 20 to 30 years, though this type of estimate is again
very site-specific.294 Jones and Salkield in their studies of the slag heaps from Roman mining
of silver at Rio Tinto in Spain use a 40 year cycle of forest generation in South-West Spain to
estimate the amount of forest required to supply 40,000 tons of charcoal per year. 295 On the
other hand, Studnicki-Gizbert and Schecter do not take into account any natural reforestation
and multiply the forest depletion per unit of charcoal by the total charcoal required by the
aggregate of historical production of silver in New Spain.296 Since during the period covered
in the present study there were no strong demographic pressures acting on the forests, a 30 year
cycle of regeneration to estimate total deforestation requirements would take into account the
natural resiliency of this resource. This would reduce substantially any long-term projection on
forest depletion due to mining activities.
A supporting infrastructure was set up to supply with charcoal the smelting operations
of ores from some of the largest silver deposits in New Spain, as can be read in an extract from
a report on the hacienda of Gonzalez de Mera in 1584:
‘a large quantity of silver has been and is being produced ... and to improve this production, he
founded and peopled a settlement of charcoal makers in the Serrania of Santa Catalina, where
294
M.V.N. d'Oliveira et al., "Forest Natural Regeneration and Biomass Production after Slash and Burn in a
Seasonally Dry Forest in the Southern Brazilian Amazon," Forest Ecology and Management (2011): 1496.
295
Salkield, "Ancient Slags in the South West of the Iberian Peninsula," 94.; G. D. B. Jones, "The Roman Mines
at Riotinto," The Journal of Roman Studies 70 (1980): 161.
296
Studnicki-Gizbert and Schecter, "Colonial Fuel-Rush," 99. As I will explain in greater detail in Chapter 6, their
method overestimates the depletion of forest cover by a factor of at least 4, or greater if natural cycles of forest
renovation are taken into account. Smelting consumed much more charcoal than amalgamation, thus the
calculation must be adjusted to take into account the consumption ratio for each refining process, and the split
between silver produced by smelting and by amalgamation. In addition the specific fuel requirements of the cazo
amalgamation process as practised in the area of Catorce must be determined, since it accounted for the majority
of silver production in the last quarter of the eighteenth century.
169
they produce a large quantity of charcoal for the smelting of said ores and those of the mines
of Mazapil, Nieves, Sombrerete and Fresnillo, which became of great importance’297
These carboneras could be owned by smelting haciendas or the charcoal could be
produced by third parties, at times indigenous groups, who then supplied the haciendas. In the
latter case the conflict of interests is shown in a document whereby the suppliers of charcoal
complain to the authorities that the hacienda owners try to cheat them by hitting too many
times the bags of charcoal on delivery, thus decreasing artificially the volume occupied by the
solid fuel. The purchase price corresponded to the volume of charcoal occupied within a bag
that could be hit up to three times on delivery to insure its contents were well packed.298
2.6.6 Regional and local environmental impact of smelting
The pollution of air and waters due to smelting of ores is as old as metallurgy, as
Agricola was well aware: ‘the miners violate Nature: they make the air pestilential with their
smoke and the waters with their waste: they destroy their health’.299 Once the nature of these
pestilences has been established, the next step is to quantify an order of magnitude of their
historical amounts. The method employed in the previous sections has been based on a review
of numerical data from the historiography and primary sources. In Chapter 4 a more specific
mass balance for all chemicals and fuel consumed in the smelting process will be calculated
297
‘se ha sacado y se saca gran cantidad de plata … y para que esto se pudiese mejor hacer , fundo y pobló un
asiento de carboneras en la Serrania de santa Catalina.. donde se hace mucha cantidad de carbón para la
fundición de dichos metales y de los de las minas de Mazapil, y de las Nieves, Sombrerete y Fresnillo, que fue de
mucha importancia’ Lacueva Muñoz, "Nueva Vizacaya y sus yacimientos minerales hasta el descubrimiento de
San José del Parral," 99.
298
Complaint by a group of charcoal makers to the authorities, AHSLP, Fondo Alcaldía Mayor 1767.2, expediente
3. The original document is so damaged it is not available to researchers, so I have taken the description of its
content from the summary in the file by Mrs. Carmen Cordero; in 1616 a caxon (box) of charcoal had to contain
at least 30 sacas (bags) of charcoal which have been hit twice. AHSLP. Fondo Alcaldía Mayor 1616.5, expediente
33, 20 December 1616.
299
‘les mineurs violent la nature … ils empestent l’air de fumées et l’eau de déchets ; ils se détruisent la sante’
Agricola, Bermannus. p. xviii.
170
based on operational data from the Hacienda de Regla. This will allow a crosscheck to be
effected between the factors estimated in this chapter and the reality of industrial operations.
The aim of establishing a self-consistent set of weight ratios of environmental impact
vectors (lead, charcoal) per kg of silver refined by smelting is to configure a macro-scenario of
its environmental impact by mining region. This calculation will be carried out in Chapter 6
based on the registers of smelted silver reported or estimated for each regional Caja (Treasury).
On the micro-level the exercise is much more complex, but would be required in order to better
estimate the consequences of lead emissions or deforestation on the local population and
ecology. Each hacienda is an environmental world on its own, location and period specific: the
architecture (height of walls and chimneys, type of furnace efficiency, size of courtyards,
internal disposition of spaces), extent and location of the deposits of grasas around the
hacienda, the wind rose, rainfall and temperature, the water basin corresponding to nearby
streams, rivers or wells. In addition, the location of each hacienda with respect to population
and agricultural centres will play a critical role in the consequences of the chemical emissions
to the environment. I will briefly discuss the case of the smelting activity around San Luis
Potosí in the seventeenth century to illustrate the scale of the challenge facing a detailed
environmental impact assessment from historical silver refining.
First of all the main historical centres of human and economic activity in the refining
region have to be identified, as well as all the mines and refining haciendas for each period.
An example is the research carried out by Salazar Gonzalez for the area around San Luis Potosí,
which I reproduce in a simplified way in Figure 2-14.300 The next step requires identifying the
refining capacity of each hacienda and determining the area of deposition of lead from its
300
Salazar González, Las haciendas de San Luis Potosí, 396.
171
10 km
Guadalcazar
Agua Hedionda
Mines
Smelting Haciendas
Carboneras (charcoal making sites)
Cattle
Peotillos
Approximate higher terrain
Sierra de Pinos
Cerro San Pedro
Armadillo
Monte Caldera
SLP
San Francisco Sauceda

de los Pozos
Pardo
Hacienda de Gogorron
Bledos
Valle de San Francisco
Figure 2-14. Location of the main mines, smelting haciendas, charcoal production,
agricultural and cattle rearing areas around the town of San Luis Potosí, adapted from the
original map in footnote 300.
furnaces according to the wind rose at each location, subject to the architectural reconstruction
of each hacienda or by using an average size hacienda as a generic point source. At first sight
it seems that refining centres in the area of San Luis Potosí were sufficiently apart from cattle
rearing and agricultural centres to lower substantially the risk of bellanding (see above, Section
172
2.6.2), but the air dispersion of lead should be mapped, as well as the diffusion channels along
the water basins in the area.
The direct risk to human population centres seems more pronounced than the risk to
agriculture, though again the exercise requires a quantitative accounting of smelting carried out
within city limits in San Luis Potosí and Guadalcazar, to name two population centres with
known refining activity. Major refining haciendas are reported by Salazar Gonzalez well
outside the city limits. Thus the Hacienda de Gogorron, with its sixteen furnaces could have
produced on average around 50 tons of lead and lead compounds issued to the atmosphere per
year, yet its relative isolation (Figure 2-14) would have restricted its impact to the local
population, its workers and their families within its compound. A smaller hacienda behind the
convent of St. Francis within the town of San Luis Potosí would have caused much greater
problems on a per capita basis to the general population around it.
In all cases the brunt of the environmental impact of lead smelting would have been
borne by the workers labouring close to the furnaces, handling the bars of molten lead and
transporting greta within the compound.301 The families of the workers would be the second
most exposed segment of the population to lead. Wives working as ore crushers and /or
washing contaminated clothes would have been subject to higher lead levels than normal.
Children playing in the dirt of these compounds would have shown at a minimum the same
high level of lead in their blood as has been measured in modern times for children living in
houses built upon old slag heaps near to Zacatecas.302
301
For details on the social role of living spaces for the indigenous workers of a smelting hacienda in the area of
San Luis Potosí see the Ph. D. thesis by Laurent Corbeil, "Identities in Motion. The Formation of a Plural Indio
Society in Early San Luis Potosi, New Spain, 1591-1630 " (McGill, 2014), 171.
302
The impact of lead from old slag heaps has been researched in Zacatecas, where dwellings have been
constructed on top of the jales. Abnormally high levels of lead have been detected in children who live in these
areas. Blood samples of children living over old mining dumps and in the vicinity of current mining operations in
173
To be studied in detail each refining region requires a mapping of activities such as
shown in Figure 2-14 as a function of time, each temporal snapshot as specific as a fingerprint,
with its ridges of hills and climates, and its whorls of different human activities and
architectural variations. The sequence of snapshots would see pollution loci appear and vanish
as mines became depleted or new ones sprung up, total shifts in pollution patterns as refining
shifted from smelting to amalgamation and back again, variation in pollution levels as chimney
heights increased and/or furnace efficiency improved, population density and agriculture
pulsing closer or retreating from refining centres as a function of local mining activity,
woodlands receding or advancing as a function of refining and mining intensity. No gross
extrapolations can be applied from one mining region to the other, or within a mining region
across the centuries. Each matrix of smelting haciendas, main urban centres, wood sources for
charcoal, agricultural and cattle rearing areas requires a degree of quantitative information that
may be available but awaits to be deciphered.
2.7 Concluding remarks
Little is known of the day-to-day working environment in a smelting hacienda of New
Spain, as of the sixteenth to the seventeenth century. No texts written by practising
metallurgists or working owners of smelting haciendas have as yet come to light, in contrast
to what I will review in the following chapter for the Peruvian Andes of the same period. It is
fortunate therefore that smelting of ores to extract metals from their mineral matrix is one of
Vetagrande, Zacatecas, have shown very high levels of lead in the blood, with levels of 15 to 24 µg/dL in one
third of the children. Children under the age of 11 are the most affected by lead as a result of playing closest to
the soil. Both the floors of the houses and the streets where the children live are just compacted earth, so conditions
would be similar to those of the historical period of interest in this chapter. Eduardo Manzanares Acuña et al.,
"Evaluación de riesgos ambientales por plomo en la población de Vetagrande, Zacatecas,"(Zacatecas: Universidad
Autónoma de Zacatecas, Unidad de Estudios Nucleares, 2005).; Eduardo González Valdez et al., "Niveles de
plomo en sangre y factores de riesgo por envenenamiento de plomo en niños mexicanos " Revista de la Facultad
de Ingeniería de la Universidad de Antioquia, no. 43 (2008): 116-18. Hausberger’s study on the complaints by
the community of Chihuahua against lead smelters singles out the greatest toxic effect of smoke on young children.
Hausberger, "Una iniciativa ecológica contra la industria minera en Chihuahua (1732)," 120.
174
the oldest technical processes applied by mankind to convert nature into mass man-made
objects. The data from historical smelting practices together with Barba’s testimony from the
Andes can be used to flesh out the sparse information from documentary sources in the
historical record for New Spain. Inventories of assets sold or rented, requests to the local
authorities on issues related to smelting, criminal cases for contraband, accounts after the fact
rendered as a result of legal wrangles between owners, their inheritors, and the administrators
that run the smelting haciendas, all have provided pieces of the puzzle. The major missing link
remains to this day a detailed operational account book that can provide a day to day diagnostic
of the inner workings of a smelting hacienda, but as I will further explore in Chapter 5, this is
a failing that applies both to smelting and amalgamation in New Spain up to the nineteenth
century.
Thanks to the work carried out by Prof. Guadalupe Salazar González it is possible to
drape the information derived from the texts around the three dimensions of the infrastructure
of smelting. The deterioration of the surviving ruins of smelting haciendas around San Luis
Potosí adds a level of urgency to her research. The historical mines within the Cerro San Pedro
have already vanished under the grinding machinery of modern open-pit mining, but it was
their deposits of lead-rich silver ores that made this region a showcase for smelting in New
Spain for nearly two centuries, as of 1592. San Luis Potosí and its mines do not seem to have
suffered the ignorance and wasteful predatory efforts of the first cohorts of dilettante refiners
in New Spain. Its silver deposits did not offer the easy-to-refine silver chlorides that spoilt, and
were spoilt by, the initial primitive Spanish refiners. The Cerro San Pedro offered the
traditional European silver ores, rich in lead, which required the standard European approach
based on a smelting process. A picture emerges of a community around San Luis Potosí geared
to provide the necessary context within which the smelting process could function, sufficiently
pragmatic as to comprise women with leading roles on both sides of the refining business.
175
The mines were fixed by the geology of the region, while the smelting haciendas
sprouted close to water, to pastures and at times close to the safety of numbers provided by the
towns. Ores were dressed by washing away the lighter non-productive fraction in water. The
same gold that at present has caused the annihilation of the Cerro San Pedro, contributed to the
revenues of these haciendas. The presence of lead was the sine qua non condition to implement
smelting in the first place, but gold and lead combined to guarantee its permanence and success.
In New Spain neither the sale of copper nor lead served to meet the cost of smelting of silver
ores, as in Europe (see Chapter 5), but additional revenues from gold did. How much lead was
present in these ores is not evident, though at least some of them were dry ores, in the parlance
of the period. Enough lead flux was generated in some districts to permit its export to other
silver smelting regions in New Spain, either officially or by contraband. Charcoal, the other
prime necessity of the smelting process, was made wherever wood was available, sometimes
by owners of haciendas who diversified upstream. Distance in procuring fuel does not seem to
have been an obstacle. An energetic private enterprise sector manifested itself through rental
contracts that offered a varied menu of services for hire to those who wished to pursue the
illusion of become refining magnates.
Casting its pall over this hive of entrepreneurial activity were the emissions of lead and
lead compounds from the smelting and refining furnaces. There can be no doubt that lead
emission levels within the workspaces of the haciendas exceeded by far any modern standard
set by legislation on occupational safety and health standards. When Medina predicated the
merits of amalgamation with mercury he was correct in highlighting the dangers that smelting
posed to the welfare of the indigenous and slave labour, ironic as this may seem in the light of
modern knowledge on mercury. For every kg of silver smelted, five to ten kg of lead or lead
compounds was disseminated from the furnace area into the surrounding air, to be deposited
around each working area close to the furnaces or entrained by the flue gas out of the chimney
176
stacks to ultimately settle on the soil, streams and surfaces within and without the hacienda. It
remains to be determined how much of this lead ended up by chance within the desolate no-
man’s land of the mounds of grasas that established a non-sanitary cordon around each
smelting hacienda. It is an intriguing question whether the extensive fields of grasas were ever
expressly located downwind of the arrays of smelting furnaces, to act as lead sinks on a ground
already condemned to a barren future.
Additional lead would be trapped within the fused shards of slag, or as the insidious
dust of fugitive emissions within each compound, which at times harboured the indigenous
workers, slaves and their families, and a Spaniard or two who could not yet afford his own
independence. The location of the fields of slags, their extension, the architecture and wind
patterns around each hacienda convert every refining unit into a unique footprint of lead
depositions that can only be studied on a case by case basis. The mass balance of lead smelting
determined in this chapter will serve to guide the regional calculations in Chapter 6. Without
the geographical fingerprint of the centres of human activity within each mining region, as set
out by Salazar Gonzalez in her integrated study of San Luis Potosí, together with a detailed
breakdown of the size, architecture and wind rose at each locality, it is not yet possible to
establish in detail the environmental impact of the individual lead emission footprints from the
scattered clusters of smelting haciendas.
The Spaniards and Germans of the sixteenth and seventeenth centuries who smelted
silver ores with lead knew of the environmental impact of the operation. Even Agricola
decorated his textbook with bare hillsides dotted with stumps of trees, while heavy smoke
billows out of furnaces, to be captured at times in chambers that have not appeared as yet in
the infrastructure of smelting haciendas in New Spain or Mexico. Up to the sixteenth century
they did not even have a choice, since heat and the reducing power of carbon were the only
means available to extract metals from their ores. In the second half of the sixteenth century
177
an alternative to smelting would be implemented in the New World, a reduction of silver
compounds to silver metal without the need to heat to high temperatures with charcoal. It was
first applied simply as an extension of deceptively similar practices used for gold, but mercury
amalgamation of silver ores was a chemical reaction in disguise that would not be fully
deciphered until late in the nineteenth century. With no theory to guide them, its self-taught
practitioners would stumble through a unique process of trial and error in the scientific
wasteland of the sixteenth century Andean highlands until they configured in just under sixty
years a mature industrial-scale process capable of refining even the silver sulphide ores.
178
3 The wet process, amalgamation of silver ores
‘I am a spagyric philosopher, alchemist … I make gold from herbs, from egg-shells, from hair,
from blood, from urine … if the princes knew of this they would stuff me in a jail so as to save
on the trips to the Indies’. Francisco de Quevedo, La Fortuna y el Seso y la hora de todos
(1635).303
‘I was then and there convinced that metallurgy in settled and civilized countries was one thing,
and metallurgy in the wilderness another … in the latter he has to adapt himself as best he can
to circumstances’. Eisler, The Metallurgy of Argentiferous Lead (1891)
3.1 Introduction
Seville had become the conduit that was feeding new supplies of bullion from another
continent to Europe. The application of a new technology based on amalgamation with mercury
had increased substantially the production of bullion to meet the ever increasing demands from
a Europe that had little else to offer the merchants of Asia. The new process was based on
mixing the mineral ore with an excess of mercury until an amalgam was formed, placing the
liquid amalgam in a cloth and squeezing out the excess mercury. The solid amalgam was then
carefully heated to separate the mercury, while the solid precious metal left behind could be
further purified by smelting. For Spain this new outlet for mercury from its mine at Almadén
represented therefore a novel marketing opportunity, and foreign bankers and merchants were
involved at many stages of this new activity. A description on how to refine the precious metal
using mercury amalgamation had already been long in print. Europe did not know a New World
full of silver even existed.
The period in question was 1460 to 1485, and according to Blanchard the bullion was
African gold, eight tons of which had been produced via amalgamation in Egypt and North
303
‘soy filosofo espagírico, alquimista ... hago oro de yerbas, de las cascaras de huevo, de cabellos, de sangre
humana, de la orina ... si lo supiesen los príncipes me engullirían en una cárcel para ahorrarse los viajes de la
Indias’
179
Africa, with an additional three tons produced in Europe. A total of forty five tons of mercury
from the Almadén mine in Spain was consumed in the process.304 The description of
amalgamating gold is from Theophilus, ca. twelfth century, as quoted by Dorothy Wyckoff in
a footnote to her translation of Albertus Magnus’ Book of Minerals.305 Bartolomé de Medina,
the cloth salesman of Seville, had not yet been born. As the French historian Jacques Heers has
explained:
‘From mid fifteenth century, before the discovery of gold from America, Castille was already
the major redistribution centre of precious metals: it is there that came the Genoese and
Florentines ... Cadiz, Seville are for certain, towards 1460, the <capitals> of gold and of the
white metal … before America the pattern was already established’.306
The history of amalgamation of silver in the New World is one of continuity, rather
than of a magnificent Ibero-American singularity devoid of a relevant past and with no negative
repercussions as to the future. Amalgamation as a large-scale method to refine silver ores did
not suddenly erupt on the world scene in Pachuca during the 1550s. The sequence of events
that led to the industrial use of mercury in the New World is a much more interesting history
of blind alleys and pragmatic refiners that is best deciphered through the chemistry of the
process. Silver and quicksilver (mercury) had been firmly wedded in the chymical mind-set of
the age much earlier than the discovery of silver by Spain in the New World.307 These are the
roots of the mentalité of a period that would lead to the longest continuous anthropogenic
disposal of mercury products to the environment in the history of mankind, from the adoption
304
305
Albertus Magnus, Book of minerals, trans. Dorothy Wyckoff (Oxford: Clarendon Press, 1967), 180.
306
‘Dès le milieu du XVe siècle, avant la découverte de l’or d’Amérique, la Castille est déjà le grand centre de
redistribution du métal précieux ; c’est là que viennent Génois et Florentins … Cadix, Séville sont bien, vers 1460,
les <capitales> de l’or et du métal blanc … avant l’Amérique les habitudes sont déjà prises’ Jacques Heers,
Gênes au XVe siècle, activité économique et problèmes sociaux (Paris: S.E.V.P.E.N., 1961), 71.
307
‘“Chymystry” [is the ] archaic word [that refers] to early modern alchemy-chemistry, a discipline that still
viewed the transmutation of base metals into gold (chrysopoeia) as viable and yet contained much in addition that
is identifiable to us moderns as chemistry’. Newman, Atoms and Alchemy : Chymistry and the Experimental
Origins of the Scientific Revolution xi.
180
of amalgamation in the mid-sixteenth century until its displacement by the cyanide process at
the end of the nineteenth century.308 In contrast to smelting, it was a wet process, due not only
to the use of liquid mercury, but because water was such a critical part of the process that as
Thierry Saignes, the French historian of Bolivia, quoted: ‘when it rained they say it rains
silver’.309
3.2 The alchemy of Mercury
For many centuries the only substance known to physically interact without altering the
two most precious of metals, gold and silver, was mercury, the semen of Shiva.310 This alone
would have been enough to anoint it as a special substance of the material world. Together with
the visual allure and the mystery of a metal that was liquid at room temperature, it became an
ontological force within the Arab and European theories on matter and metals, summed up in
the phrase ‘Mercurie without which nothing being is’.311 The line of intellectual stepping
stones that led to this sweeping judgement can be traced back to Plato, for whom matter was a
passive recipient on which properties can be imposed. The one and only primordial matter
would be the prima materiae. Aristotle then postulated all metals had a common origin and
were produced from two vapours which rise through the earth; so subtle they pass through
stones but are capable of condensing into forms of metals. Under this concept transmutation is
simply the readjustment of the quantities of these two vapours.312
308
In 1891 MacArthur and Forrest would request formal permission to the Mexican government to apply their
new cyanide process to refine first gold then silver ores. Mendizábal, La mineria mexicana, 107.
309
Thierry Saignes, "Las técnicas mineras de Potosí según una relación inédita de 1600," Arte y arqueología 8,
no. 9 (1982/1983): 171.
310
John Read and F. H. Sawyer, Prelude to Chemistry; An Outline of Alchemy, Its Literature and Relationships
(London: G. Bell and Sons, ltd., 1936), 19.
311
George Ripley, The compound of Alchemy (London: Thomas Orwin, 1591), following B3.
312
For a more in-depth discussion on the sequence of ideas from Plato to Aristotle see Aitchison, History of
Metals, 260-302.
181
Arab thinkers then associated mercury with Aristotle’s moist vapour and sulphur with
the dry or smoky vapour.313 ‘The view that sulphur and mercury were the basic components
of all metals first entered Europe textually in the work of Jabir Ibn Hayyan ... and continued to
be the cornerstone of alchemical theory up through the seventeenth century’.314 Mercury
evolved into ‘the prima materia of which all metals were made’, and only the relative quantities
of sulphur and mercury defined each type of metal.315 Raymond Lull, one of the most famous
alchemists from Spain in the thirteenth century, held that mercury was the first matter of
Genesis and that it was found in everything that had been created by God, and that mercury
served as the channel whereby the heavenly bodies could induce changes in the sublunary
world.316 Arnaldo de Vilanova would call it the ‘sperm of metals’.317 In some alchemical texts
the Sophic Mercury (the essence of mercury, not to be confused with the mercury of the
laboratory) was referred to as Azoth, showing the same Arabic roots to the word used in
Spanish texts for mercury, azogue.318
The alchemical symbol of mercury was the only one to fuse the symbols of the moon
(silver) with that of the sun (gold).319 ‘Beyond a doubt [mercury was] the metal most commonly
used as a reagent [in alchemy]... it is ... so nearly a precious metal that one recipe tells how to
313
Ibid., 268.
314
Pamela H. Smith, "Vermilion, Mercury, Blood, and Lizards: Matter and Meaning in Metallurgy," in Materials
and Expertise in Early Modern Europe : Between Market and Laboratory, ed. Ursula Klein and E. C.
Spary(Chicago: University of Chicago Press, 2010), 39.
315
L. Fabbrizzi, "Communicating about Matter with Symbols: Evolving from Alchemy to Chemistry," Journal of
Chemical Education 85, no. 11 (2008): 1506.
316
Bruce T. Moran, Distilling Knowledge : Alchemy, Chemistry, and the Scientific Revolution (Cambridge, Mass.:
Harvard University Press, 2005), 20.
317
Juan Eslava Galán, Cinco tratados españoles de alquimia (Madrid: Tecnos, 1987), 31.
318
Bruce T. Moran, Andreas Libavius and the Transformation of Alchemy : Separating Chemical Cultures with
Polemical Fire (Sagamore Beach, MA: Science History Publications/Watson Pub. International, 2007), 63.
319
Smith, "Vermilion, Mercury, Blood, and Lizards: Matter and Meaning in Metallurgy," 40-41. Eslava Galán,
Cinco tratados españoles de alquimia, 30, 43, 44, 46. Also Read and Sawyer, Prelude to Chemistry; An Outline
of Alchemy, Its Literature and Relationships 20.; John Read, Through Alchemy to Chemistry; A Procession of
Ideas & Personalities (London: G. Bell, 1957), 44.
182
make mercury alchemically out of lead’.320 It was used either as an aid to achieve transmutation
of base metals into silver or gold, or as the object of transmutation itself. According to Flamel
‘the first time that I made proiection ... was upon Mercurie, whereof I turned halfe a pound ...
into pure Silver, better than that of the Mine ... this was upon a Munday, the 17. of January ...
in the yeere ... 1382.’321 According to William Newman, one of the historians of alchemy, in
the fourteenth and fifteenth centuries ‘it was to be a favourite and prevailing theory of
transmutation ... that gold and silver could be made artificially from mercury alone, using
perhaps a little gold or silver to initiate the process’.322
Mercury generated a very rich alchemical symbolism, and became one of the popular
images of the art to the laymen of the time. The following extract is from a masque performed
for King James II of England in early 1616 titled ‘Mercury vindicated from the alchemists at
court,’ by Ben Jonson, the dramatist contemporary with William Shakespeare:
‘I am ... their Hermaphrodite ... I am corroded, and exalted, and sublim’d and reduc’d and
fetch’d over, and filtred and wash’d and wip’d ... now a sous’d Mercury, now a salted Mercury,
now a smoak’d and dri’d Mercury ... now a pickl’d Mercury ... my whole life with them hath
bene an exercise of torture ... they eat nor smell no rost-meate but in my name ... [the alchemists
promise] mountains for their meat, and all upon Mercuries securities.’323
In this same vein Quevedo in Spain would mock alchemy as presenting a better
alternative to the trips to the New World in search of precious metals.324 Framed by the lore of
320
W. J. Wilson, "An Alchemical Manuscript by Arnaldus de Bruxella," Osiris 2 (1936): 251.
321
Read, Through Alchemy to Chemistry; A Procession of Ideas & Personalities 52. Though my examples on
mercury are all taken from the European experience, from the second to the fifth century CE, Buddhist texts are
said to report the attempted transmutation of metals into gold using mercury together with other substances, and
this craft was called raseśvaradarśana (the science of mercury). Arab and Hindu texts recognize the capacity of
mercury to feed (amalgamate) on certain metals. Wilson, "An Alchemical Manuscript by Arnaldus de Bruxella,"
326.
322
William .R. Newman, The 'Summa perfectionis' of Pseudo-Geber: A Critical Edition, Translation and Study,
vol. 35 (Leiden: Brill, 1991), 205.
323
S.J. Linden, "Jonson and Sendivogius: Some New Light on Mercury Vindicated from the Alchemists at Court,"
Ambix 24, no. 1 (1977): 39, 46, 49.
324
The quote from Quevedo in the epigraph to this chapter is taken from ‘La Fortuna y el Seso y la hora de todos’
in Eslava Galán, Cinco tratados españoles de alquimia, 112-13.
183
alchemy, the practical knowledge on the amalgamation of mercury with various metals began
to appear in texts. Gerber in the early thirteenth century classified the perfection of bodies
according to how well they amalgamated with mercury, citing the ‘easy amalgamation of
quicksilver by a body of solar [gold] or lunar [silver] perfection’.325 From trials of
amalgamation to its application in refining silver ores was a small step. Biringuccio (1540) is
the most widely cited source on the use of mercury on silver ores, but there is another
contemporary text among the works of Paracelsus that is not widely quoted:
‘volatile and fugitive metals, such as gold and silver, if they are to be separated from their
minerals, since they can neither be treated by fire nor with strong waters, should be
amalgamated, separated and extracted by means of Mercurius vivus. Afterwards the Mercurius
vivus must be abstracted and separated from the calx of the gold, or silver by the grade of
distillation ... for this is the nature of Mercurius vivus that it is amalgamated with metals and
wholly united with them, but more quickly or more slowly with one than another, according as
the metal is more or less akin to its nature’.326
In this period there is an overlap between alchemical practice and the pragmatic world
of mining and refining. For example, silver is reported as having been obtained by
transmutation not less than fourteen times in London in 1578, in which copper was converted
into silver.327 There is an obvious link between these exercises in transmutation and the fact
already established at the time that lead and copper ores contained silver that could be extracted
without the need to recur to additional effects of legerdemain. As mining became more
important in Europe starting from the eleventh century, it also relied for the refining of ores on
the practical knowledge generated in the alchemical workshops, and vice-versa, so that as Tara
Nummedal states :
325
Newman, 'Summa perfectionis' of Pseudo-Geber, 35 783.
326
Biringuccio, The Pirotechnia, 47-48.; Paracelsus, The Hermetic and Alchemical Writings of Aureolus Philippus
Theophrastus Bombast, of Hohenheim, Called Paracelsus the Great : Now for the First Time Faithfully Translated
into English trans. Arthur Waite (1976), 164.
327
Eslava Galán, Cinco tratados españoles de alquimia, 132, 141.
184
‘distinguishing alchemy from metallurgy ... in the early modern period is quite difficult
…Princes and wealthy investors ... thought about alchemy as an extension of their long-
standing interest in mining technology. Patrons hired alchemists and mine experts to address
the same kinds of technical problems ... patrons frequently responded to alchemical proposals
with the same kind of investor mentality that framed their responses to mining proposals ...
alchemists were expected to produce not merely ideas but also increased profits’.328
Mercury, silver and alchemy were therefore already very much present in the mentalité
of the Spanish Crown by the time the first silver mines were being exploited in the New World.
When King Philip II of Spain turned to alchemy and the transmutation of mercury to silver in
order to pay off the numerous debts of the Crown he was acting in perfect consonance with the
context of the age. He is known to have sought the services of alchemists in 1557, 1559 and
1567, dates that overlap the introduction of amalgamation in the New World. The first attempt
took place in Flanders just one year into his reign, when he declared his first default on
payments by the Crown.329 From a report filed by the Venetian Ambassadors to the Spanish
Court it is stated that a German called Pedro Sternberg successfully transmuted for the King
six parts of mercury into six parts of silver, a one to one equivalence of mercury to silver
produced, with the aid of one ounce of unidentified ‘powders’.330 This type of transmuted silver
was considered an option to pay the troops in Flanders.331 By 1567 Philip II was still pursuing
transmutation of mercury so as to be able to pay his creditors.332
328
Tara E. Nummedal, Alchemy and Authority in the Holy Roman Empire (Chicago: University of Chicago Press,
2007), 33,86. For an example of the historiography on the evolution of ideas of alchemy and metallogeny through
the early Modern Era, and its relation to the development of science and mining in general, and also to mining
practice in the New World see Newman, Atoms and Alchemy : Chymistry and the Experimental Origins of the
Scientific Revolution and J. Norris, "The Mineral Exhalation Theory of Metallogenesis in Pre-Modern Mineral
Science," Ambix 53, no. 1 (2006).
329
Francisco Javier Puerto Sarmiento, "La panacea aúrea. Alquimia y destilación en la corte de Felipe II (1527-
1598)," Dynamis (Granada, Spain) 17(1997): 114.
330
Francisco Rodriguez Marin, Felipe y la Alquimia (Madrid1927), 16-17.
331
Ibid.
332
Ibid., 18-26. Philip II’s attitude in this episode, a mixture of pragmatic scepticism and blind faith, is well
summarized in the Spanish saying ‘no creo en brujas, pero de que vuelan, vuelan’ – ‘I don’t believe in witches,
but that they can fly, they can fly’.
185
Philip II’s interest in alchemy went beyond the practicalities of transmutation, as his
design of El Escorial has shown.333 For the Spanish historian Puerto Sarmiento the King’s
interest in the transmutation of mercury into silver was triggered by the success of Medina’s
amalgamation process in New Spain. This however does not match well with the dates or the
general alchemical context of the period. I would argue quite the contrary, that one major
reason why mercury would always receive such prompt backing from the Spanish Crown and
its officials was the aura of its alchemical and ontological role in the transmutation to silver.
Philip II would keep to his death a gift of a silver tray claimed to have been made in Brussels
from transmuted silver.334
Did the mineros in the field think that mercury was transformed into silver during the
treatment of the ores by amalgamation? At least one contemporary voice certainly thought they
did. Juan de Cardenas, a medical doctor who published in 1591 the earliest extant technical
description of the amalgamation process as practised in New Spain, states that ‘some say that
since mercury is so similar to silver ... those ... pounds ... that are missing of mercury were
converted into silver’.335 As Barba tells the story, the discovery of his cazo process started as
an attempt to ‘coagulate’ (cuajar) mercury in an iron vessel; since he did not have much iron
to build the vessel, he tried out a copper one, and added some silver ore powder to assist in the
coagulation.336 Coagulation of mercury into Luna, set on its way by the addition of some silver,
was a common alchemical attempt at transmutation. A later historian concurs, stating that in
333
Eslava Galán, Cinco tratados españoles de alquimia, 110-11. His nephew, the Emperor Rudolph II, would also
become a patron of alchemy in Prague. David Goodman, "Philip II's Patronage of Science and Engineering," The
British Journal for the History of Science 16, no. 1 (1983): 55. The role of alchemy and royal patronage in Europe
has been studied by Nummedal, Alchemy in the Holy Roman Empire.
334
Puerto Sarmiento, "La panacea aúrea. Alquimia y destilación en la corte de Felipe II (1527-1598)," 115.
335
‘algunos dicen que como el azogue es tan semejante a la plata ... aquellas ... libras ... que faltan de azogue se
convirtieron en plata’ in Juan de Cárdenas, Primera parte de los problemas y secretos maravillosos de las Indias
(México: Academia Nacional de Medicina, 1980), 161.
336
186
the sixteenth century amalgamation was confused with transmutation.337 I would posit it was
deeper than a simple confusion, that there had never been a strong conceptual divide between
the two. As I follow the path of amalgamation to its industrial apotheosis in the New World, I
will point out the many traces of its alchemical past that crop up in the recipes, the words and
the lore of the amalgamation refiners.
3.3 The gold connection
According to Brading and Cross, ‘at first the industry [New World silver mining]
formed little more than an overseas extension of the great central European mining boom of
the years 1451 -1540’.338 On mining techniques this statement may apply up to a point, but on
refining methods it merits a more critical assessment. The European boom to which it refers
was based mainly on the Saigerprocess or copper liquation, a more complex smelting technique
than cupellation, but still based on using lead as the primary refining agent. It is thus difficult
to understand how European silver refiners, who had nurtured all their technical skill from the
smelting of argentiferous copper or lead, would have any expertise outside of smelting with
lead to offer the New World. Amalgamation was never applied to argentiferous copper or
galena in Europe because it did not work on this type of silver ore, as will be made clear in the
following sections of this chapter.
The picture changes if we focus on the introduction of a different technology that
appears on a large scale in Europe around the middle of the fifteenth century. Blanchard
speculates on the role played in its transmission either via Alpine gold workings of the twelfth
337
‘en el siglo XVI se confunde la amalgama con la transmutación’ Rodriguez Marin, Felipe y la Alquimia 62-
69. Other historians have remarked upon the alchemical context of the sixteenth century, for example Bargalló,
La química inorgánica 101-104.; Castillo Martos, "Plata y revolución tecnológica," 91.;"Alquimia en la
metalurgia de plata y oro en Europa y America ".; Bartolomé de Medina, 54-55, 59-79. I have not come across a
detailed tracing of the roots of mercury and amalgamation as presented in this section.
338
187
and thirteenth centuries or by Genoese spies in the Maghreb in the 1430s. Whichever the
channel:
‘in the 1440s Genoese merchantmen may be seen transporting mercury from Almadén to
Bruges for transhipment to the Rhineland where knowledge of the intricacies of mercury
amalgamation, previously confined in the main to the gold refineries of Asia and Africa
[Sudan], was now current’.339
The skills developed during the amalgamation of tons of gold ore in Europe are very
relevant to the development of silver amalgamation in the New World. Between 1460 and
1485 Blanchard states that approximately eight tons of gold were produced via amalgamation
in Egypt and North Africa with an additional three tons produced in Europe, requiring a total
use of forty five tons of mercury from the Almadén mine in Spain. 340 Blanchard’s data
challenge the notion that amalgamation before the New World was only practised using minor
quantities of material. An industrial scaling-up of gold amalgamation had already taken place
in Europe half a century before Columbus sailed to the New World.
As soon as Columbus found samples of placer gold on Hispaniola he returned on his
second expedition with a contingent of experienced miners from Almadén for mining and from
other sites for the extraction of placer gold from washings. The finding of one kg of mercury
at the site of La Isabela, the first colony to be founded on Hispaniola in 1494, together with
339
The exact roots of amalgamation applied to ores is unknown but according to Blanchard were Afro-Asiatic in
origin. Blanchard, Mining, Metallurgy and Minting in the Middle Ages, 3 1029. A thriving market for mercury
exports from Spain existed well before the fifteenth century: ‘Geniza [Old Cairo, Egypt] records show ... mercury
... carried eastward ...after copper, mercury and its derivative, mercuric sulphide (cinnabar) were the most noted
Andalusi metallic exports. Amongst other uses, mercury was important in the refining of gold [this is a statement
applying to a period before 1212] ... mercury was traded through Seville to markets in the Mediterranean, England
and Flanders during the thirteenth and fourteenth centuries ... references to the export of Andalusi mercury are
earlier than those for copper, with Mas’udi describing Andalusi mercury “exported to the entire Islamic and non-
Islamic world” by the middle of the twelfth century’, in Olivia Remie Constable, Trade and Traders in Muslim
Spain : the Commercial Realignment of the Iberian Peninsula, 900-1500 (Cambridge; New York: Cambridge
University Press, 1994), 20, 165, 186-87, 217.
340
Blanchard, Mining, Metallurgy and Minting in the Middle Ages, 3 1030, 1034.
188
galena and implements required for the assaying by smelting of ore samples, is strong evidence
for a technical approach to gold refining from the start.341 As Born recognized in his 1791 text,
‘quicksilver … after the arrival of the Spaniards … may have been used even in America for
the extraction of gold-dust from the sands’.342 Mercury would have been used to assay or refine
by amalgamation placer gold, using methods similar to those that Biringuccio and Agricola
would include in their works published some 50 years later.343
The production and export of gold by the Spaniards in the Caribbean islands have been
interpreted as being somehow deficient in the historiography:
‘The searches for sources of both metals [gold and silver] carried the Spaniards far and wide
across the Americas ... on the promise of gold they first settled the Caribbean; finding little in
the islands they were lured on by golden visions to the Isthmus’.344
However the hard data indicate otherwise. Gold exports just from Hispaniola between
1492 and 1555 reached a total of 23.4 t, and in the peak decade of 1501 to 1510 they averaged
over 1.3 t per year.345 This represents roughly half the total annual gold produced by
amalgamation from Rhineland gold workings in Europe from 1460 to 1485.346 Columbus had
not exaggerated on the wealth of gold he found on Hispaniola, equivalent to any European
source of gold known up to that time.347 This major amount of new gold production attracted
341
A. M. Thibodeau et al., "The Strange Case of the Earliest Silver Extraction by European Colonists in the New
World," Proceedings of the National Academy of Sciences of the United States of America 104, no. 9 (2007).
Sanchez Gomez refers to a decree dated 1495 in which 15 quintales (approx.. 700 kg) of mercury are requested
from Almadén to send to the Indies. It is further evidence of the technical approach adopted by Spain in its search
for bullion in the New World. The mercury would have been used for assaying gold and also for production by
amalgamation. Sánchez Gómez, Minería no férrica en el Reino de Castilla, 277.
342
Inigo Born, Baron Inigo Born's New Process of Amalgamation of Gold and Silver Ores, and Other Metallic
Mixtures: As by His Late Imperial Majesty's Commands Introduced in Hungary and Bohemia, trans. R.E. Raspe
(T. Cadell, 1791), 8.
343
Biringuccio, The Pirotechnia, 30.; Agricola, De re metallica, 243. A modern description of field assaying of
gold ores with mercury is given in Young, "The Spanish Tradition in Gold and Silver Mining," 300-301.
344
Bakewell, "Colonial Mining," 108.; ‘Columbus first established bases on the islands ... though they yielded ...
but little gold’ in Aitchison, History of Metals, 360.
345
TePaske and Brown, Gold and Silver, 33, 56.
346
347
The same geological deposits of gold found by Columbus in Hispaniola have placed modern day Dominican
Republic seventh among the top ten gold sources of the world, with over 33 million ounces of gold reserves at the
189
the attention of European miners and artisans with experience in gold amalgamation as already
practised on an industrial scale in the Rhineland and Austrian mines. Did some of these miners
have knowledge of the advances already made in Europe on the amalgamation of silver ores?
The Republic of Venice had also been working its silver mines on the mainland. Its
geographical location (see Figure 1.7) made it an obvious choice for the export of silver and
gold to Asia. According to Day it exported the equivalent of up to one metric ton of gold per
year to Asia during the first half of the fourteenth century, equal in value to around 25% of all
new silver mined in Europe in that period.348
‘Italian prospectors searching for ores in Bohemian or German mountains ... known as “Vlach”
in Czech ... were often sent out by glassworks in Venice to collect ores for the dyeing of glass.
At the same time they gained recognition as able metallurgists and ... employed as such by the
nobility and ... King ... [thus] the Royal Mint in Kutná Hora [is the] “Vlachian court” ... there
is a Vlachian Street in Prague. Later these Italian prospectors reached the Harz Mountains in
Germany ... they also experimented in ... transmutations’.349
It is more than probable that Venetian and German miners exchanged their know-how
on gold amalgamation and silver refining during this very active production period. What was
good for gold was good for silver as well, and the illustration on gold amalgamation in Ercker
could well represent the later process on silver ores, down to the detail of mercury being
squeezed from a leather pouch onto the ore being amalgamated.350
Vergani has described the Venetian silver mines at Schio in the Haut Vicentin area,
which saw their highest level of activity in the period 1490 to 1530. He argues that though
amalgamation of silver had been used for a long time in the gilding of metals and glass, it had
not been used to refine silver ores prior to the sixteenth century. He presents strong evidence
Pueblo Viejo deposit. Data from Global Gold Mines & Deposits 2012 Ranking, published by Natural Resource
Holdings, http://www.nrh.co.il/i/pdf/NRH_Research_2012%20World_Gold_Deposits.pdf.
348
John Day, "The Great Bullion Famine of the Fifteenth Century," Past & Present, no. 79 (1978): 11, 38-39.
349
V.Í. Karpenko, "The Oldest Alchemical Manuscript in the Czech Language," Ambix 37, no. 2 (1990): 64.
350
Ercker, Treatise on Ores and Assaying, 113.
190
that the first attempt took place in Venice in 1506, since the following year two ex-prospectors
of gold, Tommaso Cusano and Giovanni Antonio Mauro of Verona applied for a patent from
the Republic of Venice to ‘extrahere argentums sine igne’, to extract silver without fire.
According to the filing they were owners of a foundry and of mines of silver and lead in the
Schio. Subsequent paperwork indicates ‘without the shadow of a doubt that the invention
consisted in treating the silver ore without fire and with water and quicksilver’. Wars intervened
and it was only in 1526 that Mauro applied himself again to the process, and to defend the
patent against unauthorized use by other refiners.
‘all the information concerning these patents emphasize on the one hand the dearth of wood …
and on the other on the fact that the new process can be carried out "without costs of charcoal
and of wood" … according to an estimation made by Mauro in 1530 … the refining of one
mark of silver from local ores cost 16 L. using traditional methods, against a cost of 5 L. using
the new process, which in addition was four times faster’.351
Berenguccio stayed in Venice from 1507 to 1509. He mentions that he paid with a
diamond ring worth 25 ducats to ‘one who taught it to me’, but does not provide either name
or nationality.352 The circumstantial evidence is strong that the origin of his amalgamation
recipe came from refining practice in the Venetian mines of the Schio region.353 Other reports
in the historiography correlate well with Vergani’s scenario. Mercury is being exported from
Spain to the Schio area by mid fifteenth century.354 According to Blanchard amalgamation of
gold and silver ores was being practised in the mining regions of Bohemia and Austria as of
the fourteenth century, and the Venetian mines were in the mainland territory contested
351
‘toute la documentation relative a ces brevets met l’accent d’une part sur la penurie des bois ... et d’autre part
sur le fait que le nouveau procede s’effectue "sans frais de charbon et de bois" … selon une estimation de Mauro
… de 1530 … l’extraction d’une marque d’argent des minerais locaux coutait alors 16 L. par les methodes
traditionelles contre 5 L. par le nouveau procede, lequel était de surcroit quatre fois plus rapide’. Raffaello
Vergani, "La métallurgie des non-ferreux dans la république du Venise (XVe-XVIIIe siècles)," in Mines et
métallurgie ed. Paul Benoit, Les chemins de la recherche (Villeurbanne: Programme Rhône-Alpes recherches en
sciences humaines, 1994), 210.
352
353
P. Braunstein, "Les entreprises minières en Vénétie au XVe siècle," Melanges d'archéologie et d'histoire 77,
no. 2 (1965): 529-607.
354
Heers, Gênes au XVe siècle, activité économique et problèmes sociaux 490.
191
between the Republic of Venice and Austria.355 The most telling corroboration is the fact that
in 1557 Philip II is advised to seek the help of amalgamation experts from Venice:
‘the refining of [silver] ores is held to be the right choice in New Spain … in Spain we believe
it will also be so [referring to the new mines at Guadalcanal] ... it is in the interests of Your
Majesty that we look in Venice for a good official that knows about this mercury [process]. It
will be of benefit to all the mines and to your Majesty’.356
The good official may have been a Garci Hernandez, who received in 1565 the
following letter:
‘His Majesty understands that in that city [Venice] there are some masters who know the art to
extract silver without fire … I have been entrusted to write to Your Excellency … [so that] you
may send a true and correct report to His Majesty ... being satisfied that these are not fictions
nor mockery as sometimes happens, because if this is true His Majesty will wish to bring them
to Spain so as to gain from their art’.357
The contents of these letters show that even by 1565 there was still some hesitation at
the highest levels in Madrid about the amalgamation process being tested in the New World,
so that the more technical support at hand, the better.
Was the Venetian know-how of amalgamation the recipe that was originally transmitted
to New Spain in the mid-1550s? Vergani argues that it is, and proposed a route originating in
Haut Vicentin, then Germany, Spain and finally New Spain. He argues that the transmission
agent was not Biringuccio’s book, since in this period techniques were made known not so
355
Blanchard, Russia's "Age of Silver". Precious-metal Production and Economic Growth in the Eighteenth
Century Footnote 5, 356.
356
‘el sacar el metal con açogue se tiene en la Nueva España por muy açertado … en Espana creo que será lo
mismo … conviene al seruicio de V.M. que se enuie a Veneçia por algún buen ofiçial desto del açogue.
Aprouechara para todas las minas y redundara en prouecho de V.M.’ Sánchez Gómez, Minería no férrica en el
Reino de Castilla, 326.
357
‘Su Magd. a entendido que en esa cibdad ay alguns maestros que tienen arte para sacar plata sin fuego …
ame mandado que yo escriba a V.M. … [para que] pueda embiar muy verdadera y çierta relación a Su Magd. …
estando satisfecho que no son fiçiones ni burlas de las que suele auer, porque siendo verdad querria su Magd.
lleuarlos a España y fauoreçer su buen arte’. Ibid., 315.
192
much through texts but by the migration of skilled artisans.358 However the chemical signature
of the Schio process, the addition of a water soluble copper compound to the amalgamation
recipe, is not observed in the New World until the 1590s (see the following section). This is
the only reason to doubt this interpretation, but it is a strong one. Had the experience from
Schio been transmitted verbatim to the New World, it would have cut some 50 years from the
learning curve of the Spanish amalgamators.
Independently of events at Schio, European and German artisans would have
transmitted know-how on large-scale gold amalgamation to Hispaniola, from where it would
arrive to New Spain. There are two examples of this technical migration attracted by the gold
of the New World. In 1525 the Italian mining specialist Paolo Belvio was sent to Hispaniola
with a batch of mercury to reactivate the placer gold workings.359 A more intriguing trail is the
mention by the German historian Schafer of a Fleming named Gaspar Loman, sent initially by
the Welsers to extract gold in Venezuela, who then moves to Cuba in 1540 to revive copper
mines in the Sierra Madre.360 In 1550 a Gaspar Lohman is awarded a merced (a royal grant)
for proposing a method to dress ores prior to smelting, and together with Bartolome de Medina
another merced in 1556, in recognition for his proposal of an amalgamation process of silver
ores, based on drawings he had brought from Germany of machinery used for the refining of
gold and silver ores with mercury.361 For the refiners of the day, if smelting was running into
trouble, the use of mercury would have been a natural option to try next. Whether Loman and
Lohman are one and the same remains to be established, but their lives illustrate well how
358
Vergani, "La métallurgie des non-ferreux dans la république du Venise (XVe-XVIIIe siècles)," 207-211.;
359
Sánchez Gómez, Minería no férrica en el Reino de Castilla, 277.
360
as quoted by West, The Parral Mining District, 106.
361
‘recibe orden del virrey, como persona experta … había mostrado al virrey ciertos dibujos y trazas que el se
trajo de Germania a Nueva España de unos ingenios en que se beneficiaban los metales de oro y plata con
azogue’. S. Zavala, "La amalgama en la minería de Nueva España," Historia Mexicana XI, no. 3 (1962): 418-19.
193
mining and refining in this period was a continuum of experience applied in the field, not
restricted to any particular metal. When New Spain was found to be richer in silver than in
gold, the Lomans of the time would have responded without a break in their stride.362
The large-scale mercury amalgamation of gold in Africa and then the Rhineland of
Europe set the stage for the application of amalgamation in the New World. It is precisely
because the initial recipe came from a much simpler process for amalgamating gold that its
implementation in the New World for silver ores would require a learning curve spanning some
fifty years. As the nature of the main silver ore being amalgamated changed from native silver
and silver chlorides to the deeper silver sulphides, so too would the amalgamation recipe have
to be adapted to each new challenge.
Peru would lag some fifteen years behind New Spain in applying a silver amalgamation
process, but in less than 30 years would produce fundamental changes to the simple gold
amalgamation recipe that would transform it into an efficient industrial refining process
specifically tailored to silver ores. To follow these historical events it is first necessary to
review the seven basic chemical principles that defined the history of amalgamation of silver
ores in the New World.
3.4 The chemistry of amalgamation
1. Mercury only amalgamates elemental silver. It does not amalgamate silver

compounds
This is the fundamental but often overlooked tenet of amalgamation with mercury. In
the strict use of the term, amalgamation does not involve a chemical transformation of either
362
The report to the Crown that silver was proving to be more important than gold in New Spain took place in
1533, according to Mendizábal, La mineria mexicana, 19-20.
194
mercury or of elemental silver. An amalgam does not have a fixed amount of mercury. It is a
physical mixture that does not change the original chemical nature of its constituents. It can be
either a liquid or a solid depending on the amount of excess mercury that it can retain like a
sponge retains water.
The confusion arises when amalgamation is the term applied to the whole refining
process using mercury, during which chemical reactions involving mercury takes place.
2. Mercury reduces silver chloride to silver, and in turn oxidizes to form calomel,
mercurous chloride (Hg2Cl2)
This is the chemical reaction that allowed the first miners of weathered silver deposits
in the New World to extract silver using the simple amalgamation recipe derived from the
refining of gold. Mercury plays a double role in the amalgamation of silver ores: it is an
amalgamating agent, and it is also an active chemical reagent that reduces any silver chloride
it comes into contact with according to the following equation:363
2Hg(l) + 2AgCl(s)  2Ag(s) + Hg2Cl2(s) Reaction 1
3. Iron and copper can also reduce silver chloride to silver

Iron and copper can also act as reducing agents to silver chloride, as for example: 364
AgCl (s) + Fe (s)  Ag (s) + FeCl2 (aq) Reaction 2
363
In all the chemical reactions the letters in parenthesis indicate the physical state of the reactant, thus (s) solid,
(l) liquid, (aq) in aqueous solution.
364
Percy, Metallurgy Silver, 94. In the case of copper, this is the reason why the cazo process invented by Barba
cuts down significantly the consumption of mercury during amalgamation. Iron would be part of the recipe in
later versions of the cazo and barrel processes. Copper was added to amalgamation recipes to cut down loss of
mercury at Guadalupe and Calvo according to Bakewell, Silver Mining in Zacatecas, 314.
195
4. Mercury does not reduce silver sulphide to silver
Laboratory runs carried out with mercury, salt and silver sulphide have failed to yield an
amalgam of silver and mercury under conditions that replicate the amalgamation process of
silver ores.365
5. Silver sulphide reacts with copper salts and sodium chloride in the presence of
oxygen to form silver chloride
Silver sulphide reacts with cupric ions in saline solution to produce silver chloride:366
Ag2S(s) + 2Cu2+ (aq) + 8Cl- (aq)  2AgCl(s) + 2[CuCl3]2- (aq) + S(s) Reaction 3
6. Heating a silver sulphide ore with salt will transform the silver sulphide into silver
chloride
This is an alternative route to convert silver sulphides into silver chlorides prior to
amalgamation with mercury. It is known by the term ‘roasting’ and was used intermittently in
the New World.367
7. Lead renders a silver ore unfit for amalgamation
Lead amalgamates with mercury, and would be present in much larger quantities than
silver in the galena ore. This competition between a much larger content of lead than silver for
the same amount of mercury being added would lead to a serious depletion in the amount of
365
Dr. David Johnson, private communication.
366
According to the laboratory experiments it is oxygen that acts as oxidizing agent and the copper in magistral
ultimately does not change its oxidation state, thus acting only as a catalyst. D.A. Johnson and K. Whittle, "The
Chemistry of the Hispanic-American Amalgamation Process," J. Chem. Soc., Dalton Trans., no. 23 (1999): 4242.
The amalgamation cakes were subjected to periodic turnovers and treading, all of which would have incorporated
fresh oxygen into the slurry mix. Thus it is reported that continuous treading by mules over twenty four hours
accelerated the amalgamation process, but costs were deemed too high to implement this on a regular basis, in
Duport, Métaux précieux au Mexique, 266.
367
Biringuccio, The Pirotechnia, 137,142.; Agricola, De re metallica, 273-276. A discussion on roasting as
practised in the nineteenth century can be found in Egleston, The Metallurgy of Silver, 226.; C.H. Aaron, Leaching
Gold and Silver Ores: The Plattner and Kiss Processes (Barry & Baird, Printers, 1881), 15-20.
196
mercury available to either amalgamate any silver or to reduce the silver chloride, leading to a
very low extraction rate and efficiency of the process. Other metals such as copper and bismuth
can also interfere with this process.368
In summary, silver ores without lead respond to amalgamation by three parallel routes:
i) by the direct amalgamation of any native silver with mercury ii) by the reduction of silver
chlorides with mercury, iron and/or copper to elemental silver, which then amalgamates with
mercury iii) by the formation of silver chloride from silver sulphide in the presence of copper
and chlorine ions in aqueous solution, which then reacts as in ii). Silver chloride can also be
generated from the sulphide by roasting the ore with salt prior to amalgamation. The only
scenarios where calomel is not formed is when native silver is being amalgamated, in which
case consumption of mercury could drop to zero. If silver chloride is also reduced to silver by
copper or iron, then the consumption of mercury during amalgamation will also be lower.
368
The historiography up to the end of the nineteenth century is very clear that lead was a major impediment to
the use of amalgamation. Though I will cite only the main authors, there are many other documents of the whole
period that underline that this was common knowledge in the field. For example: ‘the lead content of an ore is a
formal impediment to refining by amalgamation’-‘por quanto es impedimento formal lo plomoso de las piedras
para el beneficio de azogue’ in José Antonio de Villaseñor y Sánchez, Theatro americano : descripcion general
de los reynos y provincias de la Nueva España y sus jurisdiccions (Mexico: Imprenta de la Viuda de D. Joseph
Bernardo de Hogal, 1746), 132. Sonneschmidt warns that amalgamation is not apt for ores that contain copper,
lead or antimony. Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 55. ‘moreover
the ores of the Harz forest are very rich in lead, and cannot therefore be treated with quicksilver in so easy and
cheap a manner as other ores’ in Born, Born's New Process of Amalgamation, 239. Kerl, Crookes, and Röhrig,
Prof. Kerl's Metallurgy, 319.; Eissler, The Metallurgy of Argentiferous Lead, 163-172.; G. Küstel, Roasting of
Gold and Silver Ores: and the Extraction of Their Respective Metals Without Quicksilver (AJ Leary, stationer and
printer, 1880), 14, 39. It is in the modern historiography that this knowledge becomes somewhat blurred. A case
in point is the interpretation by historians of the failure of the amalgamation trials held at the newly discovered
mines of Guadalcanal in Spain right after its implementation in New Spain. One of the few historians to have
correctly interpreted the failure of the trials as caused by the lead content of their ores is Joaquín Fernández Pérez,
"La amalgamación de los minerales de plata," in El oro y la plata de las Indias en la época de los Austrias, ed.
Concepción Lopezosa Aparicio (Madrid: Fundacion ICO, 1999), 149. In contrast, Castillo Martos blames the
failure on Boteller for attempting to pass as an expert on amalgamation, becoming unmasked by his inability to
amalgamate the unamalgamateable, the argentiferous lead ores of Guadalcanal. Manuel Castillo Martos,
"Primeros beneficios de la plata por amalgamación en la América colonial (1565-1600)," in Minería y metalurgia.
Intercambio tecnológico y cultural entre América y Europa durante el período colonial español, ed. Manuel
Castillo Martos (Sevilla, Bogota: Muñoz Montoya y Montraveta Editores, 1994), 383.
197
Towards the end of the eighteenth century German metallurgists were quite correct in
pointing out that mercury only amalgamated with elemental silver. Born’s original statement
is unimpeachable: ‘[since] it is an axiom in chemistry … that quicksilver never unites with
metallic calxes, it naturally followed that … silver cannot be extracted from … ores by
quicksilver … but that the … silver calxes must be reduced into their metallic form’.369 Thus
the conclusion posited by the English historian Tristan Platt that ‘German metallurgical
discourse was based on a theory of matter that denied the possibility of refining most silver
ores by amalgamation with mercury’ or that there was a European ‘monopoly of truth’ or a
‘system of ideas’ that can be interpreted as opposing European science to American tradition,
cannot be based on the assumption that the German scientists were mistaken in their concept
of amalgamation.370 It remained a fact of chemistry that amalgamation could not be applied to
the majority of silver ores of Europe, which consisted either of silver-rich lead or copper ores.
In the early nineteenth century Humboldt was the first observer to attempt to understand
the chemistry behind amalgamation by replicating its process under controlled laboratory
conditions.371 The lack of an established chemical theoretical framework hampered most
efforts of the early nineteenth century to explain how mercury was able to extract silver from
its ores.372 The challenge was to recognize that mercury had two functions, one to create
elemental silver from the fraction of silver chloride, the other to amalgamate all the metallic
369
To find a way around the fact that mercury only amalgamates silver metal, Born argued that silver particles,
like larvae, were ‘wrapt up in sulphur and arsenic … or other metals and semi-metals, so as to become wholly
invisible to the naked eye … thus enveloped in ore and stony matter, should be freed of their external disguise …
either mechanically or chemically … assisted by calcination, which expels and destroys these heterogeneous
bodies’. Born, Born's New Process of Amalgamation, 5, 73-74. Born argues that metallic silver silver was always
present, but hidden.
370
T. Platt, "The Alchemy of Modernity. Alonso Barba's Copper Cauldrons and the Independence of Bolivian
Metallurgy (1790-1890)," Journal of Latin American Studies (2000): 2, 3, 30.
371
Humboldt, Essai politique, Tome IV, 76-80.
372
Amalgamation of silver ores reached its industrial pre-eminence in the New World well before Mendeleyev
published the first version of the periodic table of the chemical elements in 1869. Early theories on the mechanism
of amalgamation can be found in Duport, Métaux précieux au Mexique, 120-44.; Laur, "De la metallurgie de
l'argent au Mexique," 258-70.
198
silver present. It very soon became evident that a salt of mercury was being formed, which was
variously described as ‘sweet mercury’ (which may reflect the sweet taste of calomel),
chlorinated mercury, a subchloride or chloride of mercury, even if the exact formula was not
yet known.373 By 1868 a German textbook on metallurgy could state unequivocally that ‘the
large loss [of mercury] is chiefly caused by the formation of chloride of mercury … the
presence of mercury being necessary, not merely as a means of collecting the particles of silver
… but also as a chemical reagent’. In addition it also recognized the fact that other metals could
reduce silver chloride in place of mercury: ‘if iron is not present in sufficient quantity [in the
amalgamation recipe]… mercury will be wasted by its conversion into calomel’.374
The work by Johnson and Whittle was a long-overdue modern laboratory study in
which they put to the test the main amalgamation theories proposed in the historiography.375
Their results confirm the transformation of mercury into calomel as an inherent part of the
process, and yet the historiography in general has been slow to point out that mercury was not
‘lost’ during the process but consumed as a chemical reagent. The following example
underlines the problem in not recognizing the chemical nature of the amalgamation process:
‘losses of mercury, an expensive substance, tended to be great when the amalgamation
373
‘part of mercury is converted into a [chloride]’ -‘une partie du mercure se convertit en muriate’ in Humboldt,
Essai politique, Tome IV, 77. ; Mercurio dulce in Sonneschmidt and de Fagoaga, Tratado de la amalgamación
de Nueva España, 125.; ‘du mercure chloriné’ in Duport, Métaux précieux au Mexique, 124.; ‘the chloride of
silver … reacts on the metallic mercury, a portion of which is converted into subchloride, whilst the remainder
combines with the silver thus liberated’ in Pique, A Practical Treatise on Silver, 115.
374
Kerl, Crookes, and Röhrig, Prof. Kerl's Metallurgy, 325, 338-40. Cloruro de azogue (chloride of mercury) is
identified in Hermosa, Manual de Laboreo de Minas, 239.
375
Platt advocated that ‘future research, requiring chemical knowledge and laboratory resources, must attempt to
repeat all the procedures experimentally … only then will be able to evaluate properly … American … practice’
of the amalgamation process. Platt, "The Alchemy of Modernity. Alonso Barba's Copper Cauldrons and the
Independence of Bolivian Metallurgy (1790-1890)," 15-16. In the recent historiography the only attempt to
correlate the laboratory with historical practice has been Johnson and Whittle, "The Chemistry of the Hispanic-
American Amalgamation Process." Sets of chemical equations have been proposed that include the formation of
calomel (for example Bargalló, Minería y metalurgia colonial, 194.; Castillo Martos, Bartolomé de Medina, 223.
None have the correlation with experimental results presented by Johnson and Whittle.
199
processes of the day were applied to ores of very high silver content’. 376 The statement is
misleading on two fundamental issues. First, mercury is mainly consumed via a chemical
reaction, not lost. An industry that buys a reagent for its processes is not incurring a loss but a
production cost. Ulloa makes reference to the disdain of European metallurgists on the
amalgamation process as practised in the New World due to the great waste of mercury.377
There was no such great waste, but simply the required consumption of mercury in a chemical
reaction. Second, from a chemical point of view mercury is only consumed if there is silver
chloride in the ore to react with, forming calomel. If the silver content is of native silver only,
whether high or low, the consumption of mercury could in theory be nil, since it will not react
chemically with native silver. This lingering misconception on the fundamental nature of the
amalgamation process of silver ores affects the manner in which the environmental impact
vector of mercury has been interpreted in most of the literature from the twentieth century to
the present, so I will return to this critical issue in Sections 3.9 and 3.10.
3.5 The implementation by stages of the amalgamation process
Chronologies of the practice of amalgamation in the New World have been published
in the historiography, but none follow the chemical pas de deux between the chemistry of the
recipe ingredients and that of the silver ores being treated in the New World.378 To follow the
trail I will be switching from New Spain to the Andean experience around Potosí, since the
historical records of the latter provide a wealth of technical detail not as yet found for the
former, which correspond to the critical last quarter of the sixteenth century. It is a period of
376
Bakewell, Miners of the Red Mountain, 18.
377
Antonio de Ulloa, Noticias americanas: Entretenimientos phisicos-historicos, sobre la America Meridional, y
la Septentrional Oriental (Madrid: Imprenta de Don Francisco Manuel de Mena, 1772), 267.
378
Bargalló, Minería y metalurgia colonial, 112-133.; Carlos Sempat Assadourian, "Base técnica y relaciones de
producción en la minería de Potosí," in Ciencia colonial en América, ed. Antonio Lafuente and José Sala Catalá
(Madrid: Alianza, 1992).; Castillo Martos, "Primeros beneficios amalgamacion."
200
fifty years in which human ingenuity aided by the mineral riches of the subducted Andes guided
the theoretically blind but very pragmatic Spanish refiners through the complexity of the
concatenated chemical reactions needed to extract the most silver out of a suitable ore. The
profile of the tongue-tied, uncouth yet skilled refiner that implemented the amalgamation
process in the sixteenth century is well described by the Viceroy Toledo when he refers to the
manner in which the amalgamation method was brought to Potosí, thanks to the arrival at
Cuzco:
‘of a poor Spaniard from Mexico who calls himself Pero [sic] Hernández whose knowledge
cannot be questioned even if it only covers what he does nor can he barely express himself
were it not for a sharper companion who understood what lay within the simplicity of the
other’.379
The end result was a process simple to operate at ambient temperature, requiring less skills
than smelting, whose longevity and lack of change is not a sign of backwardness but rather of
a chemical maturity achieved as early as the end of the sixteenth century.
In spite of the attention given to Bartolomé de Medina in the historiography as the man
responsible for implementing amalgamation in the New World, there is at present not a single
historical document that provides any detailed information of his refining activity carried out
in Pachuca, New Spain, in the mid-1550s.380 The first extant account with a minimum of
379
‘un español de Méjico pobre que se dice Pero Hernández y no lo es menos de entendimiento porque no le tiene
para más de lo que hace ni aun apenas lo sabe declarar con otro compañero agudo que entendió lo que avía
dentro de la simplicidad destotro’ as quoted in Sempat Assadourian, "Base técnica y relaciones de producción,"
128. The burst of empirical genius shown in the altiplano of the Andes in the sixteenth century has not received
the same attention as Medina’s efforts in New Spain, even though the technical documents are much stronger for
the Andes.
380
The lack of due diligence on the claim that Medina invented amalgamation of silver ores, and the uncritical
repetition of the claim in the historiography, raises the question whether a similar laxity would be shown by
historians on a claim of similar importance in the history of technology in Europe. The Spanish historian who first
brought to light of what little was left by Medina wrote: ’the documents [left by Medina] … do not deal with the
discovery of Bartolome de Medina in its scientific aspect, but of its history … so, then, I will not deal with the
technical part, only the historic one’ - ‘los documentos … no tratan del descubrimiento de Bartolome de Medina
en su parte científica, sino en la historia … asi, pues, no voy a tratar la parte técnica del descubrimiento, sino de
la historica’ in Francisco Fernandez del Castillo, "Algunos documentos nuevos sobre Bartolome de Medina,"
201
operational details of amalgamation in New Spain is from a report published in 1591 by Juan
de Cardenas.381 For the initial period that saw the introduction of amalgamation I therefore turn
first to the description by Montesinos of Fernando de Velasco’s pioneering attempts to use
amalgamation in Peru:
‘sometimes he added the rocks with metal to the mercury, sometimes he broke them down in
pieces, and still in neither manner would the metal take the mercury, until tired of these trials
he broke the rocks into smaller pieces and leaving them in mercury forgot about them for many
days and went back to his smelting operations ... [after a few days he] separated the mercury
... squeezed it through a cloth and saw a bit of pella ... if the metal were to be more finely
ground, the mercury would act more quickly’.382
The simplicity of the method mirrors the amalgamation of gold as practised in the Western
Sudan in the twelfth century: ‘garnering their fragments of gold (quartz), mixing them with
mercury and mixing the whole over a charcoal fire so that the mercury evaporated leaving a
mass of founded pure gold’.383
In 1585 Luis Capoche published his first-hand account of how amalgamation was
carried out in Potosí during the period that had seen a nine-fold increase in silver refined, over
eighty amalgamation units built, and seven of the system of lakes that would guarantee a
Memorias de la Sociedad Alzate 47(1927): 231-251. In the absence of any technical document, the claim in favour
of Medina rests on two facts: that he was awarded a merced (a royal grant) that allowed him to receive payments
from any refiner applying his amalgamation process, and that refiners using his method were failing to pay him.
A merced by itself is not proof of a technical innovation that truly works, and during the same period a merced
was also awarded to the use of solimán, mercuric chloride, a useless recipe that was soon dropped. A joint merced
for amalgamation of silver ores was awarded both to Medina and Lohman, in 1557 leaving open the equally valid
interpretation that both of them were adapting a known process to the silver ores of New Spain, not inventing one.
For an example of a recent hagiography on Medina that includes all these facts, including the use of mercury
amalgamation of silver ores in Venice prior to Medina, yet manages to exclude them from its overall conclusions,
see Castillo Martos, Bartolomé de Medina. For a fictionalized reconstructions of events in Pachuca see Probert,
"Bartolome de Medina: the Patio Process and the Sixteenth Century Silver Crisis."
381
Cárdenas, Primera parte de los problemas y secretos maravillosos de las Indias 156-65.
382
As quoted in Bargalló, Minería y metalurgia colonial, 136.
383
As quoted from al-Idrisi in Blanchard, Mining, Metallurgy and Minting in the Middle Ages, 3 1157.
202
constant supply of water power to drive the stamp mills.384 According to Capoche : ‘having
ground the ore … to a fine flour ... the Indians place it in …boxes … where they mortify it
with brine ... and ... add mercury ... there is sufficient brine for the ore to make a slurry’.385 This
implies that up to 1585 the main ores being refined at Potosí were still mainly native silver and
silver chloride, the only two sources of silver that would amalgamate under such a simple recipe
that only required water, salt and mercury.386 Silver sulphides would not have responded to the
recipe described by Capoche, except under the spurious presence of soluble copper salts in
certain batches of ore.
The huge amounts of weathered silver ore discarded by the initial wave of ignorant
refiners were the source material for the simple amalgamation recipe initially tried out in New
Spain. The chemical impotence of this primitive recipe derived from gold workings became
evident once these weathered ores containing silver chloride and native silver run out, and
refiners were faced with increasing amounts of silver sulphide ores (the negrillos). This in turn
triggered in the following twenty years a burst of innovation that would not be repeated in the
history of amalgamation in the New World. It would lead to the birth of an amalgamation
process tailored specifically to silver ores, capable of refining not only native silver and silver
halides but also the more abundant silver sulphides as well. It is very well documented thanks
to the Spanish Viceroy of Peru, Don Fernando de Torres y Portugal, who in 1588 sent Juan
Ortiz de Zárate to investigate the state of technical innovation in Potosí. He was not driven by
384
Capoche, "Villa Imperial de Potosi," 117. Arzans writing in the 1700s states that already by March 1577 a total
of 10 million ‘reales a ocho’ had been invested in building some 100 ‘cabezas de Ingenio’, mill heads, in Potosí.
Arzáns de Orsúa y Vela, Villa imperial de Potosí, 169.
385
‘molido el metal ... hacen la harina … delgada … la pasan los indios a los cajones …, donde la mortifican con
salmuera … y … echan el azogue … echándole tanta salmuera que se hace el metal un barro ’ Capoche, "Villa
Imperial de Potosi," 123-24.
386
It is important to note that Henry Wagner, one of the few modern authors who had also been a miner in the
field, made a similar point when he drew attention to the fact that mercury and salt only work on the weathered
section of silver deposits. Wagner, "Early Silver Mining in New Spain," 64.
203
scientific interest but by the troubling rumours that more efficient recipes were being applied
that would reduce the revenues from the sale of mercury by the Crown. Ortiz was instructed
‘to find out with great care and diligence which new invention or inventions have been
proposed or are being proposed to refine those ores and who began first and in which manner
and who has continued them and wishes to continue in the future and which trials have been
made on a small or large scale and which have been positive and which not’.387
The wide scope of the instructions resulted in a detailed technical snapshot of the state
of amalgamation in Potosí precisely during the period when the refining community was
empirically finding its own solution to the challenge of refining silver sulphide ores. Two
innovations would persist until the end of amalgamation in the nineteenth century. First, finely
ground iron was included in the recipe, which reduced the consumption of mercury. 388 The
second innovation was the result of developments that are best seen as a weave of two
synergistic and concurrent improvements rather than a planned sequence of events. The nature
of the ores was changing: more and more silver sulphides (negrillos) were coming to the
surface but were not responding to the simple amalgamation recipe. That the silver sulphides
represented the major challenge to the amalgamation recipe being used at the end of the
sixteenth century in Peru is confirmed by a letter from 1600 that requests the Spanish Crown
387
‘que con particular cuidado y diligencia sepa y averigue que invencion nueva o invenciones se han pretendido
y pretenden hacer para el beneficio de los dichos metales y quien las comenzó primero y de que manera y quien
las ha proseguido e quiere proseguir e que experiencias se han hecho por menor y por mayor e cuales han salido
ciertas o inciertas’ as quoted in Jimenez de la Espada, Relaciones Geograficas de Indias - Peru II, 184, 121.
388
For the trials that led to the use of iron, see ibid., 124-28. It was clear in the 1600s that iron only decreased the
amount of mercury consumed, but did not increase the amount of silver extracted: ‘even though iron helped to
avoid loss of mercury, the silver extracted was so little that it was of little profit’- ‘aunque el hierro ayudaba para
que no se perdiese el azogue, era tan poca la plata que se sacaba que no era de ningun aprovechamiento’, in the
anonymous document titled Descripción de la Villa y Minas de Potosí, Año 1603, as reproduced in Anonymous,
"Descripción de la Villa y Minas de Potosí, año de 1603 " in Relaciones Geográficas de Indias - Perú I ed. José
Urbano Martínez Carrera Biblioteca de Autores Españoles (Madrid: Atlas, 1965 (1588)), 375.
204
to send German silver refiners to Potosí to assist in finding the right refining method for the
negrillos.389
Two new approaches were therefore adopted. The deeper and more abundant negrillo
ores started to be roasted with salt prior to being mixed with the weathered surface ores (pacos)
and then amalgamated.390 This increased the pool of silver chlorides in the ore that would be
reduced either by iron filings or mercury, leading to more silver extracted by amalgamation.
At some point it was decided to use a mineral mix using copper rich mineral instead of the
more expensive roasted negrillos to mix with the pacos, with even better results.391 The idea
may have come from the use of copapiri, a blue stone from Lipes (a location in the surroundings
of Potosí), that was added to the amalgamation recipe, as proposed by Juan Fernandez Montaño
around 1588. Copapiri was a mineral rich in copper sulphates, the first step to the critical
incorporation of copper ions to the recipe.392 The use of roasted copper pyrites would displace
the use of copapiri. The success achieved by employing all available materials at hand
underscores Eissler’s dictum that isolated refiners in primitive settings need to adapt to the
circumstances of their surroundings.393
The last major innovation to the amalgamation recipe would be introduced by Alonso
Barba in the early 1600s. By heating the amalgamation recipe in copper pots, the copper surface
389
Copy of a letter from the Audiencia of Charcas to the King of Spain, dated 6th march 1600, as quoted in Jimenez
de la Espada, Relaciones Geograficas de Indias - Peru II, 184, 132.
390
‘el pueblo esta contentísimo, porque es el beneficio que han menester, y de mayor importancia que beneficiar
los metales negrillos de por si, que por ser pocos, se acabarían en buen tiempo, y ayudando a los pacos duraran’
excerpt from a letter written by Don Pedro de Cordoba Messia to the Viceroy of Peru, 1st novermber 1602, as
quoted in ibid.
391
The greater cost of roasted negrillos is claimed to be due to deeper and flooded mines, plus the treatment of
the negrillos. See ibid., 375.
392
Ibid., 122, 128.
393
The quotation from Eissler at the beginning of the chapter, where he refers to metallurgists working in the
wilderness, in fact referred to conditions in Nevada, U.S.A, in mid nineteenth century, which makes the
innovations in the Andes of the late sixteenth century even more remarkable. Eissler, The Metallurgy of
Argentiferous Lead, vi.
205
would serve (inadvertently) as a reducing agent to the silver chloride present, thus decreasing
substantially the amount of mercury transformed into calomel.394 Barba’s process was never
applied on a similar scale as the other amalgamation processes, since it suffered from the same
problems that it passed on to its progeny, the barrel process promoted by Baron Born at the end
of the eighteenth century. Namely, it only worked on ores rich in native silver and silver
halides, such as those mined in Catorce (San Luis Potosi) towards the end of the eighteenth
century.395 For other types of silver ores it was necessary to roast them prior to amalgamation,
in order to increase their content of silver chloride. Since the cazos and barrels were also heated,
this led to very high consumption rates of fuel, as will be seen in Chapter 5.396 The advantages
of the cazo and barrel techniques was an important reduction in the reaction time to a matter of
hours, not weeks, and a significant decrease in the consumption of mercury.
The azogueros of Potosí would remember in 1617 the despair and joy that led to each
of the main recipe stages described in the previous paragraphs:
‘it seems convenient to remember that the metals initially found in Potosí ... were so rich that
they were refined by smelting and this lasted some twenty years until as they run out and
became less rich ... in time it was necessary to introduce amalgamation ... however ... due to
the great losses of mercury that occurred ... due to not having found the best way to amalgamate
them ... when they [the refiners] were starting to weaken in their resolve, amalgamation adding
iron appeared ... and some twelve years ago [1605] amalgamation with added copper, that was
such that it caused a redemption of that city [Potosí]’.397
394
Barba, Arte de los metales, 105-29.
395
For a nineteenth century description of the cazo process see Duport, Métaux précieux au Mexique, 144, 284.;
on ores for the cazo process, Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 130.
396
The best source for the barrel process is the work by its promoter, in Born, Born's New Process of
Amalgamation.; on the presentation of the Born process to his refining peers in Europe see Teich, "Born's
amalgamation process," 320-329.; Francisco Omar. Escamilla Gonzalez, "Ilustración alemana y ciencia
novohispana: la biblioteca de Fausto de Elhuyar," in Alemania y México: Percepciones mutuas en impresos, siglos
XVI-XVIII, ed. Horst Pietschmann, et al.(Ciudad de Mexico: Universidad Iberoamericana, 2005), 403-406.; on the
historical timing of the Born proposal see Arthur P. Whitaker, "The Elhuyar Mining Missions and the
Enlightenment," The Hispanic American Historical Review 31, no. 4 (1951): 578-579. Towards the end of the
nineteenth century a pan amalgamation process was implemented in the silver refining mills of the United States
of America, where it is claimed there was resistance to adopting the barrel process. Again the changes were in the
physical infrastructure of the process, not in its fundamental chemistry. See Egleston, The Metallurgy of Silver,
348.; Eissler, The Metallurgy of Argentiferous Lead, 34-40.
397
‘parece conveniente traer a la memoria, que los metales que dio el cerro de Potosí a los principios .. fueron
tan ricos que se beneficiaron por fundición, y esto duro cerca de veinte años hasta que por irse acabando y
206
The semblance of a chronological succession of three amalgamation recipes should not
mislead the reader. The use of iron filings accomplished a completely different result from the
effect of roasting or adding copper sulphate. Iron (or a copper surface) decreased the
consumption of mercury, while roasting and the addition of copper sulphate converted the
silver sulphides into silver chloride, which could then be reduced either with iron, copper or
mercury.
The evolution that took place in the amalgamation process is a very good example of
how to be right for all the wrong reasons. The use of salt in the recipe was justified at the time
by the need to remove any physical impediment between the mercury and the silver metal
particles believed to be embedded in the ores: ‘to clean (castrate or peel away from) silver of
the films or coverings within which it is found’.398 Acosta states that the ore was ‘mortified
with a concentrated salt solution … and this is done so that the salt can degrease the ground
metal, from the mud or inert content, so that mercury can receive better the silver’.399 The use
of the term ‘mortified’ is one of the signs of the alchemical context of the process: ‘nothing
can be Animated and born again, unless it first suffer Mortification … by which dissolution …
a most secret and noble change is brought about’.400 Mercury was thought to be consumed
because it tended to split up into tiny drops (lis) when subject to mechanical agitation. Lis
would then be lost in the interstices of the locations where amalgamation was carried out, or
teniendo ya menos riqueza...obligo el tiempo a que se introdujese el beneficio de los azogues… sin embargo …
por las grandes pérdidas que tuvieron de azogue … por no haber dado en el punto de su beneficio… cuando ya
comenzaban a enflaquecer, se dio el beneficio del hierro… y habrá doce años que se dio en el beneficio del cobre,
que fue tal que causo una redempcion de aquella villa’ Juan de Ibarra, "Suma de lo que el licenciado J de Ybarra
como procurador general de la villa de Potosi pide [14 Agosto 1617],"(Madrid, 1617), BL General Reference
Collection C.62.i.18.(37).
398
‘aplicar la sal para que limpie (castre o desenzurrone) la plata de las telillas o capuzes con que se halla’.
Bargalló, Minería y metalurgia colonial, 180.
399
‘Mortifican el metal con salmuera … y esto se hace para que la sal desengrase la harina de metal, del barro
o lama que tiene, con lo cual el azogue recibe mejor la plata.’ In the same alchemical-religious vein Acosta would
claim that the torment and purges suffered by silver metal were akin to those suffered by souls in their path to
God. Acosta, Historia de las Indias, 113-14.
400
S.J. Linden, "Alchemy and Eschatology in Seventeenth-Century Poetry," Ambix 31, no. 3 (1984): 119.
207
washed away with the water. Iron filings were claimed to help coalesce the lis into larger
particles that were easier to recover.401 Barba argued that his process avoided mercury losses
because:
‘the loss of mercury ... that is caused by its diminution and division into very small parts with
each mixing, due to which it is entrained by water and in the lamas [fine silt]. This
inconvenience does not happen in this [Barba’s] refining process, because mercury lies as a
whole on the bottom ... without any movement that will break it down, and so no lis is ever
seen in this process’.402
Roasting an ore was seen to disaggregate the particles of an ore, to eliminate malezas
(weeds, in the sense of an unproductive and interfering component) as well as unwanted
sulphur, and thus paved the way for a more intimate mix with mercury. Adding roasted copper
pyrites gave heat to mercury and so hastened its embrace of silver, while lime was added to
cool down an overheated process.403
For lack of primary sources it is impossible at this time to reconstruct in a similar
manner the evolution of the amalgamation recipe in New Spain. If the recipe applied by
Medina, Lohman and others in New Spain from the start had been taken from Biringuccio’s
work, it would have included not only salt but also verdigris.404 The Viceroy of New Spain
mentions in a letter addressed to the King dated 20th November 1554 that Medina has added ‘a
certain’ magistral as part of his amalgamation recipe.405 The term magistral in the context of
401
Garcia de Llanos, Diccionario, 39.
402
‘la perdida del Azogue ... que se causa por subtilizarse, y dividirse en pequeñísimas partes con los repasos, a
cuya causa se sale con el agua, y con las lamas. Inconveniente, que en todo cesa en este modo de beneficiar;
porque se esta en el fondo el Azogue unido ... sin movimiento que lo desmenuce, y assi nunca se ve lis en este
beneficio’. Barba, Arte de los metales, 117.
403
Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 96. A good guide to how
historical observations on the amalgamation process can be interpreted with modern chemical theory can be found
in Johnson and Whittle, "The Chemistry of the Hispanic-American Amalgamation Process," 4242.
404
Biringuccio, The Pirotechnia, 384. Verdigris is a complex mixture of copper salts, some of them soluble in
water, and thus a potential source of copper (II) ions for the conversion of silver sulphide to silver chloride.
405
Berthe, "Le mercure et l'industrie mexicaine au XVIéme siècle," 142.
208
New World silver has become synonymous in much of the historiography with a copper
sulphate additive, but to interpret the Viceroy’s letter it is necessary to place the word
‘magistral’ in the alchemical context of the period. The alchemical art was known in Spanish
as the Magisterio, and magistral referred to any powdered substance used to assist in
transformations.406 It is not possible to conclude therefore from one isolated reference whether
Medina used copper sulphate as early as 1554. We do not even know if he actually made the
comment to the Viceroy, since the term magistral does not appear in the few extant papers
from Medina. It could well have been the Viceroy giving an alchemical turn to his report to a
King known to be steeped in alchemical lore.
The sparse reports on amalgamation in New Spain prior to the seventeenth century do
not mention the addition of a copper salt to the recipe. The reports on expenses incurred in
materials required for amalgamation of ores from Taxco in New Spain in an amalgamation
hacienda from 1562 to 1564 do not include any copper additive. 407 Neither Gomez de
Cervantes nor Juan Cardenas mention copper magistral in their descriptions of amalgamation
in New Spain dating from the last decade of the sixteenth century. The earliest mention of
magistral in New Spain is from 1602, when the Bishop Alonso y Mota does refer specifically
to the need to add it in order to be able to refine negrillos.408 If the need for copper magistral
406
‘“Magisterium” is a word which even in classical times had developed the meaning of “method”, and the
alchemists often use it of their processes ... from this sense the word slipped over rather easily into a designation
of the effective agent ... i.e. “salt alemboch is the magistery of all magisteries ... of itself it congeals and holds
mercury, and converts silver into purest gold” ... here magistery is a substance’ in Wilson, "An Alchemical
Manuscript by Arnaldus de Bruxella," 303-304. It is in this general context that for example the term magistral
was applied to solimán, mercuric chloride, as stated in Castillo Martos, Bartolomé de Medina, 210.
407
García Mendoza, "Minas de plata en Taxco," 54.
408
Gómez de Cervantes, Nueva España siglo XVI, 143-153.; Bargalló, Minería y metalurgia colonial, 125-27.;
209
to complete the amalgamation recipe was not established by trial and error in Peru until the
1590s, it is most probable that it only became known in New Spain until after that date.409
To argue otherwise would open the door to a very intriguing scenario whereby the Vice-
Royalty of New Spain held back for nearly 50 years critical information from the main producer
of silver at the time, to the detriment of the Spanish Treasury. Since the authorities of New
Spain had awarded at least two mercedes on amalgamation they could not have pled ignorance
as to the technical details of the processes involved.410 Taking into account the surprising level
of technical interest shown from Madrid into developments taking place with mercury and
silver in the Americas during this period, this is a scenario hard to justify.411 If the refiners of
Peru were the first to identify the usefulness of copper magistral it remains to be seen how this
409
Even in Spain the use of a copper salt magistral does not figure in the recipe for amalgamation ‘as applied in
New Spain’ for the ores of Guadalcanal as of the 1560s. González, Noticia histórica minas de Guadalcanal, II
408-12. The majority report in the historiography is that a copper salt magistral was introduced in New Spain
only towards the turn of the sixteenth century. See for example Bargalló, La química inorgánica 90.; Wagner,
"Early Silver Mining in New Spain," 66.; Bakewell, Silver Mining in Zacatecas, 203.;Pérez Sáenz de Urturi, "La
minería colonial americana bajo la dominación española," 67.
410
Castillo Martos, Bartolomé de Medina, 115-19.
411
The chronology of events shows a central bureaucracy moving at colonial light speed. We have seen how
Viceroy Velasco kept Charles V informed on Medina’s testing of amalgamation during 1554. The Viceroy states
he urged Medina to send for the German (identified in the July letter as Lorenzo) in view of the importance of this
process for the Spanish Treasury (Berthe, "Le mercure et l'industrie mexicaine au XVIéme siècle," 142-43.). The
King had also received from another channel the report of the arrival of Medina and of his claims to be able to
refine silver ores with mercury. The report suggests his German advisor be allowed to travel to New Spain and to
bring major amounts of mercury with him (Castillo Martos, Bartolomé de Medina, 86.) In early September 1555
the Crown is already instructing Viceroy Luis de Velasco to start searching for mercury mines. On the 29
December 1555 Medina set down on paper in Xilotepec, New Spain, how he managed to refine silver ores with
mercury, though managing to leave out all practical information on the process itself (Fernandez del Castillo,
"Algunos documentos nuevos sobre Bartolome de Medina," 231.). On the 31 December 1555 the Princess Regent
of Spain would request the administrator of the Guadalcanal mines to check with the German Johann Schürren,
the delegate from the Fuggers in charge of mining affairs in Spain, to determine if they had used mercury to refine
silver ores, in view of the news coming out from New Spain (Sánchez Gómez, Minería no férrica en el Reino de
Castilla, 325.). The ores of Guadalcanal were mainly argentiferous galena, so I would expect the Germans to have
shrugged their shoulders in genuine ignorance of this application of mercury to silver and not gold. It is the
chemical nature of the silver ores in Guadalcanal that explains the lack of success of the amalgamation trials in
1557 by Rivas, then by Mosen Boteller in 1562, and not any technical ignorance on the part of the frustrated
amalgamators as has been suggested in the historiography (Bargalló, Minería y metalurgia colonial, 121-24.;
Manuel Castillo Martos, "Introduccion," in Minería y metalurgia: intercambio tecnológico y cultural entre
América y Europa durante el período colonial español, ed. Manuel Castillo Martos (Sevilla, Bogota: Muñoz
Montoya y Montraveta Editores, 1994), 383.). By 1559 Spain restricts the supply and sale of mercury to the
monopoly of the Crown (Bartolomé de Medina, 141.)
210
information was transmitted back to New Spain. It also implies that as in Peru, New Spain
would be refining mainly native silver and silver chlorides until the turn of the century.
A timeline of the main events in the implementation of amalgamation of silver ores in
the New World is provided in Figure 3.1.
amalgamation
Africa
gold ores
amalgamation
Alps
gold ores
start of gold
production,
Hispaniola major production of gold
use of
amalgamation
patent for
amalgamation
Venice method to
refine silver
ores
smelting continues to be practised (ie main refining method San Luis Potosí starting 17c)
arrival
start of German Medina/ Lohman
New Spain silver ore smelters granted Definite
mining for silver mercedes on amalgamation
ores simple ? recipe with
amalgamation magistral
method (mercury, implemented
salt, water)
copper salt
simple
(magistral )
amalgamation
added to
method Barba
start of iron added to what
transferred implements
major recipe becomes
Vice-Royalty from New cazo process
definite
silver ore smelting Spain and
Peru amalgamation
mining implemented
(Potosí) recipe
smelting continues to be practised
1536-
dates 12c 13-14c ? ca 1494 1507 ca 1532 1545 1556-1557 ca 1570 ca 1585 by 1600
1542
Figure 3-1. Timeline of main stages in the implementation of amalgamation in the New
World.
3.6 The infrastructure of amalgamation in New Spain
Once the critical breakthrough was made on the amalgamation recipe to allow its use
on silver sulphides, it would remain to all practical purposes unaltered in its chemical principles
until its substitution at the end of the nineteenth century by another chemical process based on
cyanide. Amalgamation has been described at length in the historiography, and in the case of
211
New Spain it is referred to in the later historiography as the patio process, since the
amalgamation slurry was spread out in tortas (cakes) placed in a courtyard (patio).412 The best
sources to consult are those written by first-hand observers from the nineteenth century, who
present a better picture of the tight mosaic of multiple concatenated operations that is
sometimes lost in the more summarized versions in the modern historiography. 413 Figure 3-2
is the basic outline of the operation, once the silver ores have been sorted either for
amalgamation or smelting. Its virtual immutability means that data from the nineteenth century,
including a pool of technical drawings and even photographs, provide a much needed depth to
the sequence set out in Figure 3-2, as detailed in the following sections.
3.6.1 Milling the ore
If the number of smelting furnaces was the defining feature in the legal description of
smelting haciendas in San Luis Potosí, it is the number and nature of the milling equipment
that is the heart of the description of ‘haciendas de minas del beneficio de sacar plata por
azogue’ in legal documents of two of the main mining districts of the period, Zacatecas and
Guanajuato.414 Detailed descriptions of the various components of morteros (stamp mills) and
molinos (Chilean mills) figure prominently in seventeenth century documents for Zacatecas.415
412
It has been suggested that the term patio amalgamation was only applied in New Spain as of the eighteenth
century. Bartolomé de Medina, 185-86.
413
The following accounts are good guides to the historical process: Sonneschmidt and de Fagoaga, Tratado de
la amalgamación de Nueva España, 38-47.; Duport, Métaux précieux au Mexique, 86-120.; Laur, "De la
metallurgie de l'argent au Mexique," 137-215.; Hermosa, Manual de Laboreo de Minas, 192-242.; Collins,
Metallurgy of Lead & Silver, 34-62.; Amador, Tratado práctico de haciendas de beneficio, 65-91.
414
For Zacatecas I have only found the term ‘Hacienda de Patio’ in a document dating from the late eighteenth
century, an incomplete valuation of inventory of an ‘Haz. de Patio’, 2 July 1788, AHEZ, Poder Judicial Civil,
C45-E18. It is also the only document of Zacatecas in which I found the term taona.
415
For Zacatecas, seventeenth century: Rental agreement, inventory of Hacienda Santa Catalina Mártir, 7 June
1651, AHEZ, Notarías/Colonia, Number 4 (Manuel Rodriguez), 73r to 74r; inventory hacienda of Doña Isabel
Saldivar Mendoza, 28 July 1659, AHEZ, Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 - 1680),
expediente 3, 142 r to 143 r; rental agreement by the Jesuits for an hacienda, 20 December 1664, AHEZ,
Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 - 1680), expediente 4, 141 r 141 v; sale agreement by
Captain Joseph de Monrreal, 7 December 1671, AHEZ, Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 -
1680), expediente 5, 199r to 200v; inventory of the Hacienda de La Sienaguilla, 9 May 1673, AHEZ
212
silver ore
milling:
molino, mortero, roasting + salt
arrastre/tahona
copper roasting copper

salt + water amalgamation sulphate pyrites
+ mercury slurry metal iron
entrained amalgam
excess mercury amalgam and mercury
removed from separated in captured in
amalgam in manga washing vats planillas and
recycled to process
mercury recycled waste fine mineral silt

from amalgam calomel
SILVER
by heating in mercury and silver loss
capellina other additives
Figure 3-2. The main stages of the patio amalgamation process as practised in New Spain /
Mexico. Dashed lines indicate optional stages.
Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 - 1680), expediente 6, 46 v to 48v; rental agreement by
Doña Cathelina, Doña Maria and Doña Agustina Hurtado (?), 9 October 1673, AHEZ Notarías/Colonia, Number
5 (Felipe de Espinosa 1653 - 1680), expediente 6, 125v to 127r; rental agreement by the Jesuits, 4 June 1674,
AHEZ Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 - 1680), expediente 7, 50r to 52r; rental agreement
by Ramón Guerero, 27 June 1674, AHEZ Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 - 1680),
expediente 7, 59v to 61r; rental agreement by Don Fernando de Aranda , 9 July 1674, AHEZ Notarías/Colonia,
Number 5 (Felipe de Espinosa 1653 - 1680), expediente 7, 72r, 72v; rental agreement by the Jesuits , 1678?,
AHEZ Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 - 1680), expediente 8, 3r to 4r; rental agreement by
Doña Elvira Perez de Bocanegra , 30 June 1678, AHEZ Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 -
1680), expediente 8, 93r to 94r; rental agreement by Doña Magdalena Guerrero, 8 April 1690, AHEZ
Notarías/Colonia, Number 11 (Lucas Fernandez Pardo 1690-1700), expediente 2, 86v to 88v; lawsuit between
Ramón de Mendoza y Juan Miguel de Bouzo (?), 21 September 1696, AHEZ Notarías/Colonia, Number 11
(Lucas Fernandez Pardo 1690-1700), expediente 2, 168r to 172r; rental agreement by Francisco de Aragón, 29
September 1694, AHEZ Notarías/Colonia, Number 11 (Lucas Fernandez Pardo 1690-1700), expediente 5, 201r
to 203r. A very useful guide to colonial stamp-mills using the data from Potosí is Peter J. Bakewell, "The First
Refining Mills in Potosi: Design and Construction.," in In Quest of Mineral Wealth: Aboriginal and Colonial
Mining and Metallurgy in Spanish America., ed. Alan K. Craig and Robert C. West (Baton Rouge: Geoscience
Publishers, 1994).
213
According to Rankine the molino was first introduced in Zacatecas, and only later in
Guanajuato, as a better option than the morteros, processing up to seven times more ore and
milling it to a finer size.416
The importance of the mills to the success of the process is reflected in the Christian
names given to each mill within the hacienda: Jesus, Santa Ana, San Pedro, San Pablo, San
Jose, Santa Isabel, among others.417 A photograph of a molino at work at the end of the
nineteenth century in Mexico is shown in Figure 3-3a, and its date does not detract from the
fact this was the same ingenio [machinery] that had been used since the seventeenth century
throughout New Spain.418
The key to amalgamation lay in achieving the maximum surface area of ore in contact
with the chemicals of the recipe.419 The finest mesh sizes were only achieved using arrastres,
also called tahonas, after first passing the raw ore through a mortero or molino.420 Such was
the attrition suffered by the arrastre/tahona stones (voladoras) that they could add up to 10%
of the weight of the material obtained after grinding (Figure 3-3 b and c).421 Humboldt declared
416
Margaret E Rankine, "The Mexican Mining Industry in the Nineteenth Century with Special Reference to
Guanajuato," Bulletin of Latin American Research (1992): 43.
417
Inventory of an hacienda owned by the Conde de Santa Rosa, 22 June 1706, AHEZ Poder Judicial-Civil C05-
E17, 9r,v; auction of the inventory of the Hacienda San Jose de las Perlas, 9 November 1761, AHEZ Serie Civil
C38-E02, 18r.
418
T. A. Rickard, Journeys of Observation (San Francisco: Dewey Pub. Co., 1907), 123 (facing).
419
The importance of milling for the success of amalgamation is so high that it has been argued that amalgamation
began in Sultepec because that is the location of the first water-powered stamp mills in New Spain built by
Germans. West, The Parral Mining District, 16. Not all mining districts aimed for the same mesh size, and in the
nineteenth century it was Guanajuato who claimed the finest milled ores, according to Duport, Métaux précieux
au Mexique, 250.
420
The drawing of an arrastre has been taken from Bernard MacDonald, "Old Mexican Methods " Mining and
Scientific Press (1907): 125. For the equivalence of the terms tahona and arrastre see Hermosa, Manual de
Laboreo de Minas, 196. The word taona or tahona does not figure in the documents of Zacatecas reviewed for
this chapter of the seventeenth century. The term arrastre appears in documents of the eighteenth century of
Guanajuato, for example in the sale of the Hacienda San Pedro, 7 December 1739, AHUG, Protocolo de Minas
libro 1732-1739, 315v; rental agreement for the Hacienda San Gerónimo (de Capetillo), 10 May 1747, AHUG,
Protocolo de Minas libro 1744-1747, 220r to 221r; rental agreement for the Hacienda de Nuestra Señora de
Guadalupe, 22 Octubre 1760, AHUG, Protocolo de Minas libro 1757-1761, 43r,to 44v.
421
Egleston, The Metallurgy of Silver, 279.
214
(a)
(b)
215
(c)
(d)
Figure 3-3. a) a horse-powered Molino in Mexico, late nineteenth century reproduced from
footnote 418 b) example of the pit of a molino, from where the finer grains that have been
shovelled through the mesh in a) are withdrawn via the arched tunnel and taken to the
tahonas/arrastres. Photo taken in the ruins of the Hacienda San Juan Nepomuceno, Marfil,
Guanajuato c) drawing of a tahona /arrastre, showing its four voladoras, reproduced from
MacDonald, footnote 420 d) photo of a wasted voladora stone, over 1.6 m in length, taken at
the home of the Morrill family, previously the Hacienda Bustos, Guanajuato.
216
he had never seen in Europe silver ores ground so fine as those he observed in an hacienda of
Guanajuato in New Spain.422 Power was supplied by water or animals.
3.6.2 The amalgamation reactor, or patio.
This was the operational area of greatest extension in an amalgamation hacienda of
New Spain. It was a courtyard (patio) covered in tightly fitting planks of wood or paving stones
to minimize loss of reagents through seepage to the soil (Figure 3-4a).423 Humboldt suggested
that the patio floor should be lined instead with iron or copper, an idea that was never taken up
but that represented a unique opportunity to have introduced the first major change in the
process since Barba.424 His idea would have decreased substantially the consumption of
mercury during the process, both on chemical grounds and by interposing a better barrier to
seepage to the soil. According to local weather or custom it would have a roof or remain open
to the elements. It comes as a surprise to observe that at least in the nineteenth century the
amalgamation slurries were not necessarily laid out in arrays of neat circular tortas (cakes), but
could be set out in square areas or simply take up all the available space in the patio (Figure 3-
b and c).425 The use of animals instead of workers to tread over the slurries reduced but did not
eliminate the human exposure to the chemicals in the slurry.426 The primitive nature of the
scenes in Figure 3-4 should not obscure the fact that the patio was an open-air industrial
422
Humboldt, Essai politique, Tome IV, 57.
423
Hermosa, Manual de Laboreo de Minas, 207. The paving stones can be observed in the photo of a patio área
that appears in Francisco Antúnez Echegaray, Monografía histórica y minera sobre el Distrito de Guanajuato
(Mexico: Consejo de Recursos Naturales No-Renovables, 1964), 389. The paving stones are also clearly depicted
in Piedro Gualdi’s painting (1846) of the patio reactor of the Hacienda Nueva de Fresnillo.
http://www.sothebys.com/en/auctions/ecatalogue/2012/latin-american-art-n08907/lot.15.html.
424
425
Berenice Pardo Hernández and Oscar Sánchez Rangel, Mineral de la Luz. La obra fotográfica de John Horgan
Jr. en México (Guanajuato: Editorial La Rana, 2010), 210.; Rickard, Journeys of Observation, 145.
426
Horses are said to have been introduced in Guanajuato haciendas by 1770. Ada Marina Lara Meza, Haciendas
de beneficio de Guanajuato. Tecnología y usos del suelo 1770-1780 (Guanajuato: Presidencia Municipal de
Guanajuato; Dirección de Cultura y Educación, 2001), 82. Their legs had to be carefully washed to avoid the
generation of cracks in the skin due to the corrosive action of the amalgamation chemicals. No mention is made
on the fate of the workers feet. Amador, Tratado práctico de haciendas de beneficio, 71. It is reported that
217
Figure 3-4 a) patio of the Hacienda de San Xavier, Guanajuato, photo reproduced from
Antúnez Echegaray, footnote 423 b) patio of the Hacienda de La Luz, Guanajuato, photo
reproduced from Hernandez and Rangel, footnote 425 c) patio of unidentified hacienda in
Guanajuato, photo reproduced from Rickard, footnote 426.
amalgam was recovered from the stomach of these horses. Horse manure was used as joint sealer within the patio
area. Rickard, Journeys of Observation, 136-37. What has not yet been studied is the effect of horse manure and
urine on the chemistry of amalgamation.
218
Figure 3-5. Amalgamation in ‘fixed casks’, Born’s variation on Barba’s cazo process.
Illustration reproduced from footnote 427.
chemical reactor of uncommon flexibility. In an amalgamation hacienda the operator enjoyed
great freedom in fixing the size of his batches, thus eliminating the usual handicap faced by
modern batch industrial processes based on a fixed reactor size. While Figure 3-4 does not
relay the same impression of scientific progress as the ordered sets of barrels driven by gears
depicted in Born’s work (Figure 3-5), the real contrast lies between the simplicity and
flexibility of the former compared to the rigidity and complexity of the latter, a complexity that
needs to justify itself on costs and efficiency and not simply on the basis it visually indicates
change, which can be confused with progress.427 The range of operational flexibility offered
by a patio reactor at a minimal capital and maintenance cost can only be underestimated at the
cost of altogether missing the point of the industrial process that took place for over three
hundred years in New Spain / Mexico.
427
Plate XXI, Born, Born's New Process of Amalgamation, following 178.
219
With regards to the amalgamation recipes, water, salt, mercury, a soluble copper
additive, iron or copper, and lime make up the usual list of ingredients. Exact quantities vary
with location and period and are well reported in the historiography on amalgamation. In
Chapter 4 I will be addressing a case study with details on the ingredients and quantities
employed. I will only point out that little if any assaying of ores took place, so that reagents
were added on the basis of the total amount of ore in a torta, and not on its silver content, and
even the size of the tortas varied according to each mining district. While mercury is always
added on the basis of a ratio to the silver in the ore, the absence of assaying made this an
exercise either in guesswork or an example of very uniform silver content in most ores for
amalgamation. Salt was usually added in excess, while copper sulphates were not, and water
was always present to avoid the cakes drying out.428
3.6.3 The handling of mercury within the hacienda.
There is one reference to mercury stored in barrels in the mercury room (aposento de
azogue or azoguería) in a Zacatecas hacienda.429 By the nineteenth century iron flasks became
the norm, so in transit and storage losses would have been reduced substantially. 430 There
remained two major steps that required direct contact with mercury by the workers. The first
was the manual addition of mercury to each amalgamation torta. This was done by squeezing
mercury carried in a leather or cloth pouch, weighing between 5 to 9 kg. 431 The second direct
manipulation of mercury took place whilst the amalgam was being separated from the mineral
428
Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 30-32.; Amador, Tratado
práctico de haciendas de beneficio, 73-75.; Hermosa, Manual de Laboreo de Minas, 209-29.
429
Rental agreement by Antonio de Larrinilla, 13 December 1682, AHEZ, Notarías/Colonia, Number 8 (Blas
Nunez Hurtado 1682 - 1683), 5r to 8r.
430
A document from Zacatecas dated 17 July 1807 refers to 104 iron flasks storing 78 quintales of mercury, as
cited in Suarez Arguello and Von Mentz, Epistolas y cuentas Vetagrande, 606.
431
Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 24.; Egleston, The Metallurgy
of Silver, 286.
220
slurry, washed, and then placed within a cylindrical cloth (the manga) where it was left to stand,
then squeezed or hit with paddles until gravity and brute force had forced the excess of mercury
out of the amalgam.432 The white, twisted cylinder observed to the right in the background of
Figure 3-6 is the cloth manga used to strain mercury from the amalgam.
Note also the presence of the scale, a vital piece of equipment that figures in every
historical itemized description of an amalgamation hacienda, an indication of the careful
accounting of mercury throughout every stage.433 A typical description from the seventeenth
to early eighteenth century in Zacatecas would read thus: ‘a room for mercury and within it
seven quintales of liquid mercury = a large table = weighing scales [various sizes of bronze
and iron weights] = a wooden pine box with keys to store mercury = a manga to separate
mercury with its iron hoop, [support?] and cloth’.434 Two hundred years or more separate
similar descriptions from the view in Figure 3-6, but apart from the iron flasks now used to
store mercury lining the floor of the azoguería, the infrastructure remains the same.435
432
Hermosa, Manual de Laboreo de Minas, 235-37.
433
According to West, refiners had to keep books with mass balances of mercury being used that were checked
by officials. West, The Parral Mining District, 113.
434
‘un aposento que sirve de azogue y en el siete quintales de azogue en caldo = una mesa grande = un peso de
cruz con sus balanzas [pesas varias de hierro y bronce] = una caxa de pino con su llave que sirve para guardar
azogue … una manga para desazogar con su aro de fiero, contramanga y paño de azogue’ in the inventory of an
hacienda owned by the Conde de Santa Rosa, 22 June 1706, AHEZ Poder Judicial-Civil C05-E17, 10v to 11r.
435
Plate 11, Percy F. Martin, Mexico's Treasure House (Guanajuato) (New York: The Cheltenham Press, 1906),
52-54.
221
Figure 3-6. Photo of an azoguería (mercury room), showing mercury flasks, scales and
vertical white manga held by chains from a beam to the right of the background. Original photo
from footnote 435.
3.6.4 Planillas
Once the amalgamation process within a torta was deemed by the master amalgamator
(azoguero) to have run its course, the next stage in the process required separating the heavier
amalgam from the lighter fraction of the mineral matrix (gangue) in the slurry, minimizing the
potential collateral loss of amalgam, excess mercury and unreacted silver ore. All these
valuable components in the slurry of the torta could be entrained with the water used to wash
away the mineral matrix that made up over 99.6 % of the solid content of the slurry. The
historiography describes in detail the vats and stirring paddles used to wash away the mineral
matrix (gangue) of the slurry.436 The description of the labadero is also given in great detail in
436
222
legal documents that describe the infrastructure of amalgamation haciendas.437 The washings
were voided into a canal (cárcamo) that run through the hacienda until it voided into a nearby
stream. Along the way a sediment was deposited within this cárcamo which was regularly
scooped out and washed in the planillas, an inclined plane on which the planilleros, women
and men, patiently spooned water and sediment over the planillas (Figure 3-7).438 The aim was
to separate by gravity the heavier fraction of mercury, amalgam and silver ore entrained in the
washing water, allowing the finer silt to be washed away.439 The visual evidence of the physical
reality of the process is important for the later discussion on the stages of the amalgamation
process where physical losses of mercury were most likely to occur.
3.6.5 Desazogaderas or capellinas and the recycling of mercury.
The description and understanding of the recovery of mercury from the amalgam is
critical for the analysis of the historical environmental impact of silver refining by
amalgamation. The previous sections have shown via photographs the conditions under which
mercury was handled in the patio area. After days or weeks of treading and shovelling a slurry
made up of over 99.6 % of waste material, the process had now separated the two most valuable
components, mercury and silver. The equal importance of both to the refiner is reflected in
inventories of haciendas that single out the wooden chests with locks where both were stored,
side by side.440 The first time that mercury is again handled in a closed environment is within
437
See documents in footnote 416.
438
See photograph in Bordeaux, Mexique mines d'argent, facing p. 160.
439
Amador, Tratado práctico de haciendas de beneficio, 82-84, 89-91. For a drawing of a cross-section of a
planilla see Laur, "De la metallurgie de l'argent au Mexique." Plate II, Figures 5 and 6.
440
For example, ‘dos caxas grandes con sus llaves una en que se guarda el azogue otra en que se guarda la plata’
in the rental agreement by the Jesuits for an hacienda, 20 December 1664, AHEZ, Notarías/Colonia, Number 5
(Felipe de Espinosa 1653 - 1680), expediente 4, 141 r, v; similar mention in the rental agreement by Mateo de
Herrera, 15 July 1645, AHEZ Poder Judicial-Civil C01-E40.
223
(a)
(b)
Figure 3-7. a) cross section of a planilla, reproduced from Laur, footnote 439 b) photo of
planilleros at work, showing the cárcamo running with the waste water from the washings, the
inclined planillas and the scoop made from a bull’s horn in their right hand. Reproduction of
illustration in footnote 438. It seems to correspond to a larger image from where a more reduced
area has been reproduced in a photograph by Charles Waite titled ‘Washing the Tailings,
Guanajuato’, 1907, number 16 in the series Tema y Tecnología (CIG-AGN).
the azoguería (mercury room), and a comparison between Figures 3-6 and 3-7 indicates the
quantum leap regarding the degree of control over operational losses. Now the amalgam,
having been carefully washed free of any remaining sediment, is strained through a manga so
as to separate as much as possible the excess of liquid mercury. In New Spain the pliable
224
amalgam was removed from the manga and moulded into flat disks or small spheres that could
be stacked, separated by layers of ashes, for the most important step of separating mercury
from silver, the heating of the amalgam and the recovery of mercury.
The amalgam was never heated over an open fire, contrary to what is observed in
modern artisanal gold amalgamation practice. The recycling of mercury always took place
under controlled conditions within a closed recipient whose only outlet was to a water trap that
cooled and condensed the mercury vapour. The twentieth-century historiography has employed
very loose terminology in describing this critical stage. For example, even though there is never
a direct contact between flame and mercury, the stage has been described as ‘burning’ of
amalgam.441 The efficiency of the method to recycle mercury is questioned without providing
any evidence, as for example: ‘to distill off the mercury ... which could in part be recovered’
(emphasis added).442 A claim has been made that ‘the resulting pina [amalgam] … was then
smelted to vaporize any remaining quicksilver. It was at this point in the production process
that tons of mercury were released in … Potosí and breathed in by its inhabitants each year’.443
It is not only a question that refining terms are being employed incorrectly (it is incorrect to
state that the amalgam was smelted), but an operational stage is being misrepresented in the
modern historiography and major conclusions on the environmental history of silver refining
are derived from the unproven assumption that the major mercury loss vector was volatilized
mercury.
441
Brading, Miners Bourbon Mexico, 138.
442
Bakewell, Silver Mining in Zacatecas, 22.
443
Robins, Mercury, Mining and Empire, 11.
225
To understand the care taken to operate an efficient recovery of volatile mercury from
the amalgam it is necessary to review in detail the four types of heating/condensing assemblies
as described in the historiography up to the nineteenth century:
Caperuzas: the early versions of the recycling apparatus used to recover mercury from
the amalgam was described and illustrated by Barba (Figure 3-8). The lower part was an iron
vessel in which the amalgams, called a piña due to their cone shape, were placed on an iron
tray. The upper part was a caperuza or lid resembling a Venetian plague mask, whose nozzle
ended in the water trap. Both parts had to fit snugly to avoid losses of mercury vapour, and the
joint was sealed with a mixture of ash and other components. The assembly was placed on top
of a furnace, much like a kettle on a wood-fired stove. Barba complained that caperuzas made
from ordinary clay in the late sixteenth century were too porous, and that ideally either a special
clay as for crucibles, or better still iron or copper should be used instead in order to lessen the
loss of mercury in this stage.444
Desazogadera: in New Spain it appears that this term was more common than caperuza,
and the use of metal ones appear cited as early as the 1600s. 445. In a non-exhaustive survey of
contracts drawn up in Zacatecas during the seventeenth century, desazogaderas (literally
removers of mercury) or as equipment for desazogar are the terms utilized, as for example:
‘and a room that is used to remove mercury … an iron tray and small plate to remove
444
Barba, Arte de los metales, 99-104, 169-70. The dictionary on mining and refining terms compiled in Potosí in
1609 by Garcia de Llanos also mentions the use of caperuzas made from fired clay. Garcia de Llanos, Diccionario,
38. The complaint by Barba that mercury vapour could traverse these primitive and porous desazogaderas may
be the source of a similar claim made by Robins to justify his proposal that 85% of mercury was lost as volatile
mercury during the historical heating of amalgams. Otherwise Robins does not state the experimental or
historiographical basis for his proposal that: ‘the porous nature of the clay facilitated the escape of mercury, as
did the generally inefficient and artisanal nature of the entire process’ Robins, Mercury, Mining and Empire, 88.
445
‘the apparatus to remove mercury of metal, copper or bronze’- ‘desazogadera de metal, de cobre o bronce’
Gómez de Cervantes, Nueva España siglo XVI, 153.
226
Figure 3-8. Barba’s illustration of the apparatus used at the end of the sixteenth century to
recycle mercury from the amalgam. A is the iron vessel where the amalgam is placed, B is the
caperuza (clay or metal) that fits on top of A, with a nozzle (C) that ends below the surface of
the water placed in tank E. The whole assembly sits on the ring D on top of the brick furnace.
Figure from Barba, reproduced from footnote 444.
mercury’.446 The number of desazogaderas could reach four or more, and some are described
as embedded in the ground.447 The latter implies a heating arrangement as applied to capellinas
(see below). The fact they do not figure in every inventory of the seventeenth century in
Zacatecas, or are listed as incomplete, especially of those haciendas that have not been
446
‘y un aposento que sirve de desogadera … yten un candelero de fiero para desazogar y un platillo’ Rental
agreement by Mateo de Herrera, 15 July 1645, AHEZ Poder Judicial-Civil C01-E40. Also in the, inventory of
Hacienda Santa Catalina Mártir, 7 June 1651, AHEZ, Notarias/Colonia, Number 4 (Manuel Rodriguez), 73r to
74r.
447
‘el aposento de desazogar que sigue al dicho de azogue con quatro desazogaderas’ in the inventory of an
hacienda owned by the Conde de Santa Rosa, 22 June 1706, AHEZ Poder Judicial-Civil C05-E17, 10v to 11r;
‘Dicho aposento … de desazogadera … con su puerta y un candado grande con llave … y en el dicho aposento
dos ollas enterradas para desazogar’ in Rental agreement by Antonio de Larrinilla, 13 December 1682, AHEZ,
Notarías/Colonia, Number 8 (Blas Nunez Hurtado 1682 - 1683), 5r to 8r.
227
operating, is unexpected due to their important role in the process.448 Either they were
inexpensive to make (for example if they were made of clay) and so did not merit a special
mention as a capital asset and/or when made from metal they were readily pilfered from
abandoned haciendas.
Capellina: in New Spain the term capellina appears regularly in documents from the
eighteenth century of Zacatecas and Guanajuato. It is made up of two halves, a lower vessel
with a rounded base made of metal, in which is placed an iron tray that will hold the stack of
flat amalgams separated by layers of ashes. A metal bell (hence the name capellina) fits snugly
onto this base, with the joint sealed by a combination of ashes and the weight of the
capellina.449 A tube channelled the mercury vapours through the bottom of the assembly onto
a water trap where mercury condensed and could be recovered. Copper, bronze and iron
capellinas figure prominently in the inventories of amalgamation haciendas in Zacatecas and
Guanajuato at least as early as the 1720s.450 By the nineteenth centuries large capellinas appear,
weighing as much as 500 kg or more, requiring a pulley to be manoeuvred into place.451 Thus
448
A good example of incomplete equipment listed in an inventory is the following: ‘the equipment to remove
mercury with two vats and without the receivers’ - ‘la desazogadera con dos piletas y sin recividoras’ in the
auction of the inventory of the Hacienda San Jose de las Perlas, 9 November 1761, AHEZ Serie Civil C38-E02,
18r.
449
According to Hermosa, 23 to 34 amalgam disks (marquetas), weighing 1 arroba each (11.5 kg) were placed in
rows of six under a capellina. Hermosa, Manual de Laboreo de Minas, 238-39.
450
‘una desazogadera de cobre’ in rental agreement for the Hacienda San Antonio, 1 January 1726, AHUG,
Protocolo Cabildo 1726, Libro 30, 331r; ‘una desazogadera de cobre’, inventory of Hacienda San Antonio, 15
March 1737, AHUG, Protocolo de Minas libro 1732-1739, 200r; ‘una dessasogadera de cobre’ in inventory
(Memoria de los apegos) attached to rental agreement for the Hacienda San Antonio, 17 March 1741, AHUG,
Protocolo de Minas libro 1740-1741, 83r; ‘una desasogadera de cobre’ in the rental agreement for the Hacienda
San Antonio, 5 Junio 1755, AHUG, Protocolo de Minas libro 1754-1756, 152r; ‘the cavity for two capellinas with
their corresponding vat’-‘el hueco de dos capellinas con su pila correspondiente’ in the inventory for the Hacienda
la Sangre de Christo, 4 October 1784, AHUG, Minería, Caja 15 expediente 547; ‘3 copper capellinas with 938
lbs [of copper]’- ‘3 capellinas de cobre con 938 libras’ in the valuation of inventory of an hacienda (incomplete),
2 July 1788,AHEZ, Poder Judicial Civil, C45-E18. Other examples of metal capellinas are given in Table I.
451
500 kg in Fresnillo, according to Duport, Métaux précieux au Mexique, 269. Over one ton in Regla, Chapter
Four; ‘Capellinas: 1 iron bell weighing 900 lbs [approx. 400 kg]’ –‘Capellinas: 1 campana de fierro con peso de
900 lb’ in inventory of the Hacienda de Dolores de Granadita, Primer semestre de 1884, Protocolo de Minas,
Tomo 1884, 157r. For a detailed description of how the stack is set up within a capellina see Amador, Tratado
práctico de haciendas de beneficio, 84-88.
228
metal capellinas are in use during the two centuries that produced nearly 90% of the total silver
coming from New Spain / Mexico from the sixteenth to the nineteenth century. Figure 3-9a is
a cross-section of a capellina assembly from the nineteenth century.452 Heating was always
indirect, piling hot embers around the side of the capellina, and according to one account the
cycle was monitored by the sounds emanating from within or by the quantity of mercury
recovered.453 After up to 30 hours, depending on the size of the amalgam stack, the recovery
of mercury was complete.454
In Guanajuato capellina was the name also given to a two-storey building that housed
the capellina assembly in its upper part, and a tube for the condensing mercury vapours reached
to the lower part of the structure where it was cooled by water and mercury collected. The
lower floor could be either below or above ground level (Figure 3-10).455
Finally, Collins describes at the end of the nineteenth century a very simple
arrangement used by small refiners in Mexico and South America, based on clay water bottles,
though empty metal mercury flasks fitted with a screwed-on pipe were used as well (Figure 3-
11).456 The amalgam is rammed into the bottom of each bottle, up to 35 lb at a time, and the
bottle inverted so that the open end now lies below the water level in the tank below. Heating
is applied indirectly through hot embers heaped around the inverted bottles.
452
Laur, "De la metallurgie de l'argent au Mexique," Plate II, Fig. 11.; MacDonald, "Old Mexican Methods," 126.
The drawings that appear in MacDonald’s article are reproduced from a report drawn up in 1886 by E. Tillmann,
Royal Commissioner of Mines in Prussia, who visited Guanajuato. I have not been able to locate the original
report which would provide very valuable lithographs on the patio process as practised in the late nineteenth
century. The same drawings are reproduced in Rickard, Journeys of Observation.
453
454
Descriptions of the capellina and the recycling of mercury can be found in Sonneschmidt and de Fagoaga,
Tratado de la amalgamación de Nueva España, 49-51.; Hermosa, Manual de Laboreo de Minas, 237-39. Amador,
Tratado práctico de haciendas de beneficio, 84-89.; Percy, Metallurgy, I 627.; Kerl, Crookes, and Röhrig, Prof.
Kerl's Metallurgy, 327.
455
For a detailed cross-section of a two-story structure for a capellina see Egleston, The Metallurgy of Silver, 328.
456
Collins, Metallurgy of Lead & Silver, Vol. II, 136-38.
229
B
A
V C
(a)
(b)
Figure 3-9. a) Cross-section of capellina used by nineteenth century in Mexico. The stack
of amalgam disks (A) is covered by the capellina. A temporary wall made of bricks (B)
surrounds the capellina, and embers are placed in the space between the two. Mercury
condenses in the water basin below the capellina assembly (V) and is collected via the main
water channel (C). Adapted from illustration in Laur, footnote 452 b) indirect heating of
capellina and use of pulleys to manoeuvre capellina into place, original lithographs by
Tillmann, as reproduced in Macdonald, footnote 452.
230
Whether the pioneer retort described by Barba, the humble assembly mentioned by
Collins or the elaborate capellina illustrated by Laur, every one indicates the care taken to
constrain the mercury vapour produced by an indirect heating of the amalgam within an
enclosed space, so that it would only condense on contact with water. Of all the stages in the
process, this was the most easy to control with regards to losses of mercury. The simplicity of
an arrangement that requires no special furnace, no direct contact of the flame with the
amalgam, and minimal operational oversight within an isolated working area, made this stage
extremely efficient with regard to the recovery of mercury from the amalgam at a minimal loss
during normal operational practice (see section 3.10 below).
capellina
Ground level,
option 2
(b) (c)
water
trap
Ground level,
option 1
(a)
(d) (e)
Figure 3-10. a) schematic cross section of a capellina building in Guanajuato, original

illustration from footnote 455 with annotations added b) and c) capellina building, option 2,
within the historic Hacienda La Escalera, Guanajuato, at present part of a private dwelling,
used as a storage shed d) capellina building, option 2, renovated by the Morrill family and
converted into a studio, previously the Hacienda Bustos, Guanajuato e) reconstruction of a
capellina building, option 1, at a museum on mining on the grounds of the Real de Minas hotel,
Guanajuato. The museum was not functioning as of February 2013.
231
indirect heating via inverted clay water

embers placed around bottle or metal
inverted bottles mercury flask with
screwed-n pipe
water tank
Figure 3-11. Simple recycling assembly of mercury from amalgam, based on inverted clay
water bottles. Original illustration from footnote 456, with labels added.
3.7 Amalgamation: the human factor
The patio amalgamation process presented the advantage of hiding the complexity of
its chemical reactions behind an easy-to-follow operation carried out for the most part at
ambient temperature. The labour was either chokingly dusty or cloggingly wet, always back-
breaking, and for the planilleros as mind-numbing as for a modern factory worker staring at a
never ending stream of parts on a conveyor belt.457 The heating stages (roasting of ores in some
locations, heating of the capellinas and the final casting of silver bars) were very simple
compared to the requirements of smelting. The responsibility for the operation lay in the hands
of the azoguero, literally the mercury man, who in the absence of assaying controlled the
amalgamation process using organoleptic triggers.458 All the rest was the repetitive application
of a well-known recipe that had the seal of approval of centuries of production, ‘neither does
457
‘planilleras, that is how we call those that wash the fine silt, since only women, with their patience and capacity
for waiting, can do such distressing work’ - ‘planilleras, que asi llamamos a las que lavan los polvillos, porque
solo las mujeres, con su paciencia y espera, pueden hacer trabajo tan penoso’ in Juan Moreno y Castro, Arte o
nuevo modo de beneficiar los metales de oro y plata (Mexico: Imprenta Biblioteca Mexicana, 1758), 24.
458
Azoguero has a more technical, operational, connotation in New Spain compared to Peru, where it was also
applied to the rich entrepreneurs involved in the silver mining and refining business.
232
it require … trained and experienced workers; they can be easily and readily trained in all that
is necessary’.459
Little is known of the azogueros of New Spain, the human lynchpin for the patio
amalgamation process, if they were only Spaniards, or if they could also be indigenous artisans
or even African slaves entrusted with the operation.460 Subject to the size of the hacienda there
could be more than one azoguero in charge of the patio operations. As all pragmatic artisans,
they were wary of novelty:
‘those that call themselves Azogueros … are men … so ordinary as their common birth and
customs since they never know how to discourse, nor wish to treat any important subject …
[the] Owners continue to use this same method … telling us : Sir. Here have come many Artists
with great enthusiasm, but useless’.461
Barba’s desire that an exam filter out the least trained of refiners had still not
materialized by the early nineteenth century, though not all azogueros could be as flawed as
this one:
‘Vetagrande [Zacatecas], September 12 of [18]06, Señor Don Antonio de Bassoco. My dear

sir: In the hacienda del Buen Sincero the administrator and azoguero is Don Lorenzo de Ovalle,
who not only does not have the capacity and skill required to carry out both jobs, but also has
the vice of an incurable gambler’.462
459
‘tampoco exige … peones prácticos y enseñados; pues en un instante se adiestran para todo lo necesario, con
facilidad’. in Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 92.
460
Salazar Soler states that the azoguero was either a Spaniard or a mestizo, of inter-racial heritage, but does not
furnish further background information on this issue. Salazar-Soler, "Innovaciones técnicas, mestizajes y formas
de trabajo en Potosí de los siglos XVI y XVII," 147. They were housed within the hacienda compound, for
example ‘vivienda para el azoguero’ in the auction of the inventory of the Hacienda San Jose de las Perlas, 9
November 1761, AHEZ Serie Civil C38-E02, 17v. I have found no mention of slaves or indigenous workers as
azogueros.
461
‘los que se llaman Azogueros … son unos hombres …tan ordinaria, como la plebeya de su nacimiento y
costumbres pues nunca saben discurrir, ni quieren materia alguna importante… [los] Duenos siguen este mismo
método … diciendo como nos han dicho: Senor, Aquí han venido varios Artistas con muchos entuciasmos, que
de nada sirven’ in Dominguez de la Fuente, Leal Informe Politico-Legal, 133. See also Mendizábal, La mineria
mexicana, 306.
462
‘Vetagrande, septiembre 12 del [18]06, Señor Don Antonio de Bassoco. Muy señor nuestro: En la hacienda
del Buen Sincero se halla de administrador y azoguero Don Lorenzo de Ovalle, quien además de no tener la
extensión y habilidad necesaria para cumplir con ambos empleos, tiene el vicio de tahúr irremediable’ as quoted
in Suarez Arguello and Von Mentz, Epistolas y cuentas Vetagrande, 530.
233
The role of the human actors of patio amalgamation is still virgin territory in the
historiography. The learning curve for each type of amalgamation recipe seems to have been
fairly brief, though the modes of propagation of the amalgamation recipe are as yet
undetermined. How the technical recipe spread remains to be studied. Secrecy had marked the
initial development of refining techniques in Europe: ‘there are no written accounts of the
technique [of assaying] before 1500, probably as a consequence of the traditions of secrecy
operating in medieval metal production’.463 Why refiners in the New World were able to obtain
relatively quickly details of innovations that represented huge market advantages is still an
open question. Did the Spanish authorities play any role in guaranteeing the spread of skills in
the use of mercury, or was this left to private enterprise that by nature would be reticent to
sharing any knowledge that would help the competition? The role of the indigenous workers
in this transmission process needs clarification, and the absence of written instructions would
have made even more attractive a process with easy to follow manual steps, to avoid aggregate
errors of interpretation of an orally transmitted technology.
Amalgamation also revealed the ethical context of the silver refining industry in the
New World. Bribery up to the highest levels was a normal part of the process of procuring
sufficient mercury quotas from the monopoly exercised by the Crown authorities, but even
goodwill payments could not guarantee its supply.464
‘the royal officials were given a gift of 1000 pesos each one for the assignment we were given
of 600 quintales of mercury; and maybe we will be favourably served in other instances in
relation to the last distribution of the remaining ingredient. It is our impression that they have
remained grateful’465
463
Bartels, "Production of Silver in Harz Mountains," 87-89. The spread of the new copper liquation process was
hampered initially because it was treated as an entrepreneurial secret by Thurzo Fugger, according to Kellenbenz,
"Final Remarks Silver Production," 310.
464
Bribery of the Viceroy to obtain mercury supplies is mentioned in Brading, Miners Bourbon Mexico, 141.
465
Extract from letter dated 3 October 1806 sent by Isidoro Sarachaga and Manuel de Libron to Don Antonio de
Bassoco, one of the rich and powerful mine and hacienda owners of Zacatecas though he lived in Mexico City:
‘los señores oficiales reales quedaron obsequiados de 1000 pesos cada uno por la asignación que nos hicieron
234
The glimpse offered by this correspondence extends beyond mercury supply into the
overall problem of mercury and silver contraband which in turn places a healthy measure of
doubt on the reliability of official silver and mercury data for the whole colonial period. I will
return to this issue in Chapter 6, when I present an overall balance of minimum levels of
chemicals issued to the environment, a base line that in reality would have been higher than
that which is projected from the official statistics.
3.8 The architecture of patio amalgamation
Three architectural features stand out when visiting the amalgamation haciendas that
still survive in Mexico: the imposing height of their perimeter walls, the industrial magnitude
of some of these haciendas, and their clinging like limpets to the banks of streams. The first is
the consequence of pioneer mining characterized by ‘isolation and insecurity … for many years
and even centuries at the mercy of warrior hordes’.466 The imposing outer shell of the haciendas
have been ascribed to the fortress mentality of the ‘arquitectura del temor’, the ‘architecture
of fear’, the legacy of mining and refining carried out within a ‘territorio de Guerra’, the
frontier lands where indigenous groups still battled the Spanish conquerors, even as late as the
nineteenth century.467
de 600 quintales de azogue; y acaso nos sirvan gustosos en las demás ocurrencias relativas a el ultimo
repartimiento que hagan de dicho ingredient que existe y demás. Nos parece han quedado agradecidos’, quoted
in Suarez Arguello and Von Mentz, Epistolas y cuentas Vetagrande, 535. They were wrong, in February of 1807
no mercury was allocated to them, which forced them to offer up to 12 pesos more than the official price if 600
quintales were available. Ibid., 571-72.
466
‘aislamiento e inseguridad …estuvieron por muchos años y aun siglos a merced de las hordas guerreras’, in
Mendizábal, La mineria mexicana, 45.
467
Aurelio de los Reyes, Los caminos de la plata (Mexico: Universidad Iberoamericana, 1991), 13. In the north
‘the Spanish miner continually on guard against attack … the reales de minas … were virtual military garrisons’
in West, The Parral Mining District, 4-7. ‘in general all the advanced [mines] in territories still held by non-
conquered tribes … were protected by small forts or reales; from which the generic name Real de Minas’- ‘en
general todas las de avanzada en territorio poblado por tribus no sometidas … eran protegidas por fortines o
reales; de donde tomaron el nombre genérico de Real de Minas’ in Bargalló, Minería y metalurgia colonial, 63.
For an overview of the Chichimeca wars and mining see Bakewell, Silver Mining in Zacatecas, 4-40. ‘Incursiones
bárbaras’, barbaric incursions, are still cited as taking place in the nineteenth century, cited in Cuahtemoc Velasco
Avila et al., Estado y minería en México (1767-1910) (Mexico: Fondo de Cultura Económica, 1988), 234.
235
‘the dearth of towns made the haciendas into tiny human centres scattered within the boundless
countryside, requiring protection against the sudden attacks not only by chichimecas, but also
by bandits and guerrilleros .. a state of permanent insecurity is reflected in the old hulks of the
haciendas’468
Protection from those without and protection for the wealth within that provided the
resources to push up stone over stone well beyond five metres in height, massive bastions at
times crenelated like the walls of castles (Figure 3-12). The region around Pachuca was rife
with brigands even in the late nineteenth century, as is clear from any of the travel diaries cited
in Chapter 4. Not only travellers but refining haciendas were constantly under threat: ‘Attached
to these works [Hacienda Velasco] is a handsome house, deserted. No officer dare live in it.
Not long since its walls were scaled by a robber band, though they could find but little booty.’469
The height of these walls not only isolated the hacienda from a hostile exterior, in some cases
they also helped to encapsulate some of the environmental impact vectors within the confines
of each hacienda.
With regard to the scale of production of amalgamation haciendas, I will use as a guide
the case of Guanajuato at the end of the eighteenth century. Guanajuato was the region of New
Spain that by the nineteenth century was most renowned for the quality of its amalgamation
process.470 Its refining units have been divided by historians into zangarros or haciendas, the
former defined as a smaller and less permanent production unit than the hacienda.471 The
Mexican historians Martin Torres and Lara Meza have analysed the nature and location of these
468
‘la escasez de villas hacía de las haciendas diminutos núcleos humanos esparcidos en la inmensidad del
paisaje, necesitadas de protección contra los ataques relámpagos no solo de los chichimecas, sino de los bandidos
y de los guerrilleros … inseguridad perenne reflejan los viejos cascos de las haciendas’ in Reyes, Los caminos
de la plata, 19.
469
Gilbert Haven, Our Next-Door Neighbor: a Winter in Mexico (New York: Harper & Brothers, 1875), 150.
470
For example see Duport, Métaux précieux au Mexique, 263.
471
According to Salazar González zangarros was a term also applied to small sugar refining units. Salazar
González, Las haciendas de San Luis Potosí, 83.
236
units in and around Guanajuato and Marfil, including an extensive review of ownership, sale
and rental values, and the web of interrelated business dealings and circulation of capital that
fuelled the sector in the seventeenth and eighteenth century. The average size of each class of
refining unit is not established, though they make clear that the term hacienda encompasses a
very wide range of production capacities.472
There are two documents in the historical archives of Guanajuato that shed more
quantitative data on this general classification. The first is a list of haciendas and zangarros
that had been supplied (aviado) by the Hacienda La Escalera in Guanajuato, and now had to
prove they had the assets with which to repay a considerable aggregate debt of 556,349 pesos
up to the 31st October 1788.473 The documents first of all highlight the existing hierarchy within
the business model of refining haciendas in Guanajuato, with La Escalera serving as
distribution and credit centre to a constellation of third-party refining units held together by a
relationship of supply and credit dependency.474 In each deposition the debtors under oath set
out a list of assets to demonstrate their capacity to repay their portion of the debt, but without
validation by an independent evaluator. It is not an inventory of each refining unit, but a
selection of assets made by each owner. Some items (silver content of ores, silver in patio) are
prone to over-estimations, while the inventory of mercury, number of amalgamation tortas
(cakes) in a patio, number of arrastres or mules are more difficult to manipulate. I have listed
472
Lara Meza, Haciendas de beneficio de Guanajuato, 53-104.; Martín Torres, El beneficio en Guanajuato. The
pioneering study on the ownership and business structure of the colonial refining sector in Guanajuato is reported
in Brading, Miners Bourbon Mexico, 261-302.
473
AHUG, Bienes Difuntos, Caja 8, Expediente 57, dated 18 mayo 1791 to 7 julio 1791, 6r to 29v. I have only
excluded from Table 3-I the data for the Hacienda Nuestra Señora de Guadalupe that appear in 23r,v since the
owner declares he will extract from 80 montones more than 4,000 marks of silver, which would imply an ore with
50 marks per montón, five times higher than the average of the other haciendas in the document.
474
The full name is the Hacienda de San Antonio de Escalera, situated within the city of Guanajuato, a photo of
its impressive perimeter walls appears Figures 3-11 a) and a surviving capellina house in Figure 3-9 d). The
grounds of La Escalera belong at present to a private housing condominium.
237
in Table 3-I some of the assets as reported, though the significance of a missing item may only
mean the owner decided not to include it as proof of collateral.
(a) (b) (c)
(d) (e)
Figure 3-12. Perimeter walls a) Hacienda La Escalera, Guanajuato, scale bar 1.6 m b)
Hacienda San Juan Nepomuceno, Marfil, scale bar 1.6 m c) Hacienda Santa Ana, Marfil, scale
bar 1.6 m d) Hacienda Las Mercedes, Zacatecas, scale bar 1.7 m e) Hacienda, name unknown,
Pacumo, Zacatecas, scale bars 1.7 m.
From the data in Table 3-I, I conclude that as a rule of thumb zangarros on average
handled in their reactor patio 20 montones of 10 cargas each, less than 30 t of ore, about one
238
third smaller than the average of the units listed as haciendas in the table.475 The haciendas
being supplied by La Escalera had on average in their patios around 650 marks of silver value
(approx.. 150 kg silver), in just three tortas of 20 montones each, for a total of 600 cargas of
ore, or approximately 83 t. The data on mercury incorporated into each patio, around 800 kg
on average, is indicative of the amount of liquid mercury to which the skin of the indigenous
workers was exposed every month, starting with the initial addition of mercury to the tortas.
The information on the number of molinos and arrastres is more limited, but an average of 12
arrastres and one molino is given by the data in Table 3-I. If these haciendas worked on the
basis of a four week amalgamation period, the annual revenues would be in the range of 60,000
pesos, and the annual throughput of ore around 1,000 t, based on a silver content of not more
than 0.2%.476
With respect to La Escalera there is a second document that places it among the
industrial-size haciendas of New Spain of this period. It is a short accounting book that has
registered the monthly amounts of silver produced at four haciendas in Guanajuato, between
May 1791 and March 1792.477 Figure 3-13 tracks the monthly amounts in marks of silver,
resulting in an average monthly production of 4,039 marks (929 kg) of silver for La Escalera.
This is a scale of production approximately half of that by Regla (Chapter 4) and one fifth of
the Hacienda Nueva de Fresnillo (Zacatecas) in the nineteenth century (see below). Two other
haciendas show a level of silver production higher than those of Table 3-I: Hacienda San
Antonio, with a monthly average of 1447 marks of silver (333 kg) and La Purísima with 1218
475
This would make a zangarro more formal than the ‘occasional sheds’ described by Brading. He estimates
between 200 to 300 zangarros and 50 to 75 haciendas in the Guanajuato/Marfil area in the period from 1780 to
1803. Brading, Miners Bourbon Mexico, 282. Martin also treats zangarros as very primitive and non-permanent
units, but he does not go further into their analysis. He provides a list of 64 haciendas in the Guanajuato area that
existed between 1686 and 1740. Martín Torres, El beneficio en Guanajuato, XII-XIII, 154.
476
Lara Meza, Haciendas de beneficio de Guanajuato.; Martín Torres, El beneficio en Guanajuato.
477
AHUG, Minería, Caja 15, Doc 551, 1r to 5v, initial date 7 May 1791.
239
marks of silver (280 kg). Finally the Hacienda San Juan Nepomuceno presents a monthly
average within the range of Table 3-I, with 544 marks of silver (125 kg). It is interesting that
silver content inventory

montones tortas mercury in value of silver in capellinas
Hacienda marks / mercury, arrastres mill mules
in patio (cakes) patio , lbs patio , marks number, weight
monton quintales
Santísima Trinidad 66 3 2,000 10 660 35 2 96
Nuestra Señora de los
42 3 1,218 8 336 7 6 150
Dolores del Presidio
San Juan 78 2,028 9 702 17 1, 380 kg 15 1 113
San José 67 1,004 10 670 12 1 63
Nuestra Señora de
45 1,904 405 32 2, cost 500 pesos 72
Guadalupe de Rocha
3, copper, 506 kg,
de Cuevas 2,500 800 15 265
288 kg and 161 kg
San Nicolás 48 15 720 36 1, copper, 357 kg 65
Nuestra Señora de
46 1,100 15 690 19 32
Guadalupe
San Antonio 3 600
San Ignacio 15 3 2,400 800 36 124
San José 120 1,100 30 14 1 124
de Mota 110 1,100 550 30 65
Durán 50 7 350 40
average 62 3 1,695 11 645 27 12 1 106
silver content inventory
montones tortas mercury in value of silver in
Zángarro marks / mercury, capellinas arrastres mill mules
in patio (cakes) patio , lbs patio , marks
monton quintales
San Buenaventura 16 432 1, copper, 276 kg 24
San Miguel 50 10 500 6 1, copper, 288 kg 8 1 60
El Cantador 5 200 80
San José 9 16 144 4 12
Nuestra Señora de
21 500 220 6 20
Guadalupe
Señora de los Dolores 12 4 12
average 20 377 13 236 7 6 1 26
Table 3-I. List of debtors on supplies provided by the Hacienda La Escalera, Guanajuato,
with their estimate of selected assets offered as collateral to the debt, as of 1791. Raw data from
footnote 473. Numbers in italic have been calculated by author from original data.
the accounting of silver states the amount in marks but also in number of pinas, and in one case
for La Escalera in number of capellinas (2 capellinas producing 551 marks, or approximately
64 kg of silver per capellina).

240
What would have been the architectural footprint of an average amalgamation hacienda
such as in Table 3- I? The number of historical plans of amalgamation haciendas of New Spain
is nil, as far as I have found.478 In the case of Guanajuato Lara Meza confirms there are no
8,000
7,000
6,000
marks silver
5,000
4,000
3,000
2,000
1,000
0
May June July Aug Sept Oct Nov Dec Jan Feb Mar
1791 1792
San Antonio San Juan Nepomuceno La Purísima La Escalera
Figure 3-13. Monthly silver production for four amalgamation haciendas in Guanajuato,
calculated from data in footnote 477.
extant plans dating from the eighteenth century.479 For the nineteenth century I have found only
four historical plans of amalgamation haciendas, two of which were located in Guanajuato and
two in Zacatecas. The first is a drawing in ink on paper dated 19th May 1885 of the Hacienda
Casas Blancas situated in Marfil, close to Guanajuato, where many refining haciendas were
located on the banks of the stream named the Quebrada de Marfil (Figure 3-14; the plan has
been redrawn in Figure 3-15 to aid its interpretation). The plan was drawn up by José M. Lira,
478
The earliest detailed plans of the complete layout of an amalgamation hacienda dates from Arzans’ description
of life in Potosí around mid seventeenth century. His drawings correspond to amalgamation carried out in
buitrones (rectangular vats) and provide a very useful guide to operating conditions (some of the Spanish overseers
seem to be carrying whips) and spatial distribution of the process stages, even though they are not drawn to scale.
See for example Arzáns de Orsúa y Vela, Villa imperial de Potosí, 91.
479
Lara Meza, Haciendas de beneficio de Guanajuato, 124. She presents in her work a reconstruction of the
Hacienda de Salgado as of 1775. A generic version of a plan for an amalgamation hacienda in Mexico was
published in the nineteenth century in Amador, Tratado práctico de haciendas de beneficio, no page number.
241
Civil Engineer, and identifies the work spaces together with their main dimensions.480 Even
though it indicates smelting areas, it is an amalgamation hacienda and these may correspond
to the casting of silver bars or an occasional smelting run. I have no information on the
background to this hacienda, and a visit to the existing structure, which cannot be accessed,
did not add to the information in the plan.
Figure 3-14. Plan of the Hacienda Casas Blancas, Marfil, 1885, AHUG, Mapoteca, Hda.
Casas Blancas, 5p3. Digital image supplied by AHUG.
The second is a reproduction of the 1866 lithograph depicting the layout of the
Hacienda de Rocha that occupied the grounds of what is at present the Hotel Reales de Minas
480
AHUG, Mapoteca, Hda. Casas Blancas, 5p3.
242
in Guanajuato (Figure 3-16). The original lithograph was part of a technical report on
amalgamation haciendas in Guanajuato by the German mining expert E. Tilmann.481 In 1850
the Hacienda de Rocha was purchased by a Scot, Alexander Cumming Langton, whose son
Carlos Ignacio Cumming would establish Cummings y Jimenez, a long-running partnership
that in addition to the Hacienda de Rocha would be involved in other mining projects. The
hacienda is described as ‘the best equipped of its time’.482 The perimeter walls depicted in the
inset are as usual imposing. For the sake of clarity I schematize the main process areas in
Figure 3-17.
W patio reactor s
W
s
capellina
?
s arrastres
D
ore
? AP
chapel
e
D
mill e AP : animal power
AP W : water well
AP s : smelting
AP
D : dwellings
e : assay room
Figure 3-15. Main process areas identified in Figure 3-14.
481
The plan of the Hacienda de Rocha is reproduced in MacDonald, "Old Mexican Methods," 123. The original
report by Tilman, titled Der Bergbau und das amalgamations. Distrikte von Guanajuato in Mexico, was printed
in 1866 and commissioned by Guillermo Brockmann as part of a fund-raising effort in Germany to interest
investors in the refining industry in Guanajuato, which ultimately failed, as cited in Rankine, "The Mexican
Mining Industry in the Nineteenth Century with Special Reference to Guanajuato," 40.
482
Amor Mildred Escalante, "Redes familiares empresariales en la ciudad de Guanajuato, México, 1877-1911,"
in XXI Jornadas de Historia Económica (Caseros, Argentina: Asociación Argentina de Historia Económica,
2008).
243
capellina
Figure 3-16. Historical plan of the Hacienda de Rocha, Guanajuato, reproduction in

MacDonald, footnote 481, from an original lithograph in a report by E. Tillmann of 1866.
The third example is the plan of the Hacienda de Las Mercedes, on the outskirts of
Zacatecas, from the mid nineteenth century.483 The hacienda was to be sold by auction in 1850,
483
I found the original drawing in the AHEZ, Fondo Mapas e Ilustraciones, Serie V: Planos Siglos XVIII al XX,
number 16, 3 July 1850. The drawing was originally part of the report on the evaluation of the assets of the
Hacienda that is located under Notarias: Fernandez y Ferniza, Juan; 6 January 1850 to 22 December 1850, foja
158. I afterwards learnt that the existence of the plan had been made public by a Mexican historian, Antonio
Ramirez Ramos, during the V Reunion de Historiadores de la Minería Latinoamericana, held in San Luis Potosí
in 1997. His presentation was titled ‘Aplicación y vigencia del procedimiento de amalgamación en la ciudad de
Nuestra Señora de Zacatecas’. Since I have not been able to access his presentation paper I have opted to carry
out my own analysis. According to Lara Meza he described the plan and the activities carried out at the hacienda.
See Lara Meza, Haciendas de beneficio de Guanajuato, 47.
244
stores
50 m
AP AP treatment of fines
stores
arrastres
azoguería mill
mill
AP patio reactor
casting furnace
planillas
capellina
water tank
washing
AP
Figure 3-17. Main process areas of the hacienda de Rocha, according to Figure 3-16.
so a survey of its value was carried out, including the drawing of its architectural footprint,
together with dimensions and the identification of its functional areas. It is a unique document
in the historiography since it contains more detailed information (dimensions, building
materials, equipment, values) than the previous plans. Figure 3-18 is a digital image of the
original map, which I have re-drawn in more schematic form in Figure 3-19. The three areas
where mercury and silver are being isolated and refined (see grey circles in Figure 3-19) are
close to each other, as expected from Amador’s dictum that these areas should always be placed
245
Figure 3-18. Digital image of the original hand drawn plan, ink on paper, of the Hacienda
Las Mercedes, AHEZ, Fondo Mapas e Ilustraciones, Serie V: Planos Siglos XVIII al XX,
number 16, 3 July 1850.
246
O
point of entry T T M
T W
O
O M T
T
T
A T
A
T
C G
N
A
A
M stamp-mill
separation
W amalgam
P T tahona
ore mercury
O M room with
handling
manga
patio
P reactor capellina
C
G magistral
A animal power
50 varas
Figure 3-19. Schematic plan of the main process-related areas of the Hacienda Las
Mercedes, adapted from Figure 3-18.
so as to facilitate their vigilance by the hacienda overseers.484 These three areas, separation of
the amalgam from the slurry (W), mercury room where the excess mercury is squeezed in the
manga (M) and the space where the capellinas are heated (C) require just 3% of the total area
of the hacienda. The space required just by the capellinas is only 0.5% of the total area. No
mention is made of a furnace to cast the final silver bars, or of any chimney structure for the
heating area of the capellinas. The valuation report rendered by Felipe Semeria arrives at a
484
Amador, Tratado práctico de haciendas de beneficio, 70.
247
total value of just under 6,800 pesos for the hacienda proper (which at the time was not active),
plus another 700 pesos for nearby dwellings and orchard.
The largest of these industrial amalgamation haciendas of the nineteenth century is the
Hacienda Nueva de Fresnillo, Zacatecas, whose plan is the only one more readily available in
the historiography (Figure 3-20), built to process the ores from the nearby mines of Proaño.
The plan printed by the Escuela Práctica del Colegio de Minería in the mid-nineteenth century
shows a geometrical array of spaces and equipment reminiscent of the colonnades of a Greek
or Roman temple.485 Even without the benefit of a scale it is evident that the three areas
dedicated for the capellinas and the one area for the smelting of silver bars are a minor fraction
of the total operational area of the hacienda.
The capital cost of the Hacienda Nueva de Fresnillo, including steam engines, by 1844
was just over one million pesos.486 This flagship of amalgamation haciendas in Mexico could
accommodate 64 tortas in its patio, each one of 120,000 lb, for a total of 3.5 t of ore at any one
time. This is approximately 40 times the average capacity of the haciendas in Table I. Duport
states there were 314 arrastres (tahonas in the plan by the Escuela de Minería), though only
200 were in use at any time. The hacienda also had 12 molinos. Animal power was furnished
by 1,500 horses and mules. The diameter of the tortas was approximately 15 m, with a height
485
There are two sources for the plan of the Hacienda Nueva de Fresnillos. One is the plan drawn by Luis Pozos
Rosas Fito, Escuela Práctica del Colegio de Ingeniería, 1857 (MMOB, Colección General, Estado de Zacatecas,
Varilla CGZAC03, Numero de control 12780-CGE-7241) and printed by Litografía Salazar, prepared for the
Escuela Práctica del Colegio de Minería. On the top right hand corner it states ‘Lam. [Lámina] 9’, which implies
it is one of series, but there is no other information on the content of the other engravings. There is another
reference to a plan of the Hacienda Nueva in the plates at the end of Duport’s book on silver refining in Mexico,
but unfortunately the plates have been removed from the copies of Duport’s work available for consultation.
According to Duport he obtained his copy of the plans from a French engineer, M. Doy, who worked at the
hacienda. Duport, Métaux précieux au Mexique, 261. Silliman reproduces the plan published by Duport, in
Benjamin Silliman, Sketch of the Great Historic Mines of the Cerro de Proaño at Fresnillo, State of Zacatecas
(New Haven, Conn.: Tuttle, Morehouse & Taylor, 1883), 33. The Escuela Práctica of the Colegio de Minas to be
situated in Fresnillo was created by decree in 1853, according to http://www.palaciomineria.unam.mx/
recorrido/dir_jose_maria_tornel.htm. Thus it is most probable the plan in Duport’s book is the original version.
486
Great Historic Mines of the Cerro de Proaño, xvii.
248
of 28 cm. The copper capellinas measured approximately 90 cm high and 45 cm in diameter,
and weighed 500 kg. In 1841 approximately 32,500 t of ore were amalgamated, producing over
51 t of silver.487 It indicates a refined silver content with an average over 0.16%. Figure 3-21
shows an internal view in perspective of the patio reactors, an interesting mix of traditional
technology together with nineteenth-century innovations, such as steam powered tahonas and
furnaces to
prepare planillas
copper
magistral
stables
mills
steam-powered
tahonas
tahonas
separation of
amalgam
from slurry capellinas
P
P : patio
storage area
reactor areas P P
smelting of
P silver bars
administrative
offices
Figure 3-20. Plan of the Hacienda de Proaños, Fresnillo, Zacatecas, annotations added to the
digital image supplied by MMOB, Colección General, Estado de Zacatecas, Varilla CGZAC03,
Numero de control 12780-CGE-7241. No scale is supplied in the drawing.
487
Duport, Métaux précieux au Mexique, 260-83.
249
railtracks and canals acting as conveyor belts for the fines. Though all these additions to the
traditional process would have eased some of the back-breaking manual and animal labour, the
heart of the operation remained the patio reactor.
separation amalgam from

slurry
steam boilers
tahonas
railtracks for steam-powered
milled ore tahonas
canal to transport fines
Figure 3-21. A perspective of the Hacienda Nueva de Fresnillo, annotations added to the
digital image supplied by MMOB, Colección General, Estado de Zacatecas, Varilla CGZAC03,
Numero de control 12780-CGE-7241.
To compare the main features of these four haciendas I have placed in Table 3-II the
main spatial and operational characteristics that can be measured from the plans, together with
additional information on processing capacity or infrastructure reported in the historiography
or in the primary sources. I have complemented this information with my estimate of their
production capacity (when not provided by other sources), using a formula based on their
reactor patio area (see Section 4.4.3 of the next chapter). Of the four, the Hacienda Las
250
Mercedes is closest to the haciendas listed in Table 3-I, which probably represent the average
amalgamation hacienda of the colonial period, though not enough data has been reported to
establish this for certain. At the other extreme the Hacienda Nueva de Fresnillo represents a
scale of industrial silver refining output that I have not found matched elsewhere in Mexico up
to the mid-nineteenth century.488 The disproportion in operational areas required at both
extremes of the process (patio and milling vs the tiny area reserved for the capellinas) is
indicative of the extremely reductive nature of the whole refining process in terms of volume
and mass, a characteristic that will become more apparent in the next chapter which analyzes
in detail the Hacienda de Regla, of a magnitude similar to that of Rocha.
approximate
Approximate Ore Silver Approximate distribution of operational areas
mills arrastres patio reactor
area processed production (% of total)
area
Hacienda
patio mills, animal
m2 number number m2 t/m t/m storage capellinas
reactor arrastres power
Casas Blancas 10,500 1 to 2 37 1,900 < 300 0.3 to 0.6 18 16 10 ~ 0.2
Rocha 23,000 2 84 7,000 <1,000 1 to 2 30 18 19 9 ~ 0.1
0.15 to
Las Mercedes 2,700 1 8 1,000 < 150 36 29 11 6 ~ 0.5
0.3
Nueva de Fresnillo 12 314 2,700 4.3 no scale in plan
Table 3-II. A comparison of some of the main spatial and operational features of
amalgamation haciendas in Mexico, nineteenth century. Data in italics have been estimated,
other data are derived from plans or from sources in text.
488
The recognition that amalgamation haciendas in general were industrial units has been made in the
historiography: for example, West, The Parral Mining District, 26.; Lara Meza, Haciendas de beneficio de
Guanajuato, 31. Other examples of industrial amalgamation haciendas are: at Sombrerete the Fanoaga family mill
had 84 arrastres and 14 furnaces, according to Brading, Miners Bourbon Mexico, 140.; at Sauceda, Zacatecas, in
the 1770s, 3000 quintales of ore per week (approx.. 600 t per month) were processed in the Hacienda La Sagrada
Familia, built by Jose de la Borda, with 70 arrastres and 10 stamp mills, as reported in Brading, "Mexican Silver
Mining," 672-73. In the 1780s the haciendas (number unknown) of the Marques de Valle Ameno refined 3200
quintales of ore every two weeks, approximately 300 t per month. Moreno y Castro, Arte de beneficiar los metales,
12.
251
From an environmental point of view the economy of scale in the largest industrial
haciendas did not translate into an economy of waste. Regardless of size, from the zangarro to
the Hacienda Nueva de Fresnillo, over 99.6% of all the solids that entered each unit were
eliminated as waste via the nearest available stream. Whatever efficiency was translated into a
lower consumption of reagents per kg of silver produced paled in comparison to the magnitude
of the mineral matrix, the gangue, that held no economic interest for the refiner. In contrast to
the case of smelting haciendas, this waste from the process did not form a belt of lunar
landscape around them but was flushed constantly downstream, converting the streams that
nurtured these haciendas into their waste-disposal unit. If natural water streams were not
available or sufficient, water pumped from mine shafts was diverted to these haciendas, at
times even through other haciendas without permission, via aqueducts whose remains in the
more arid Zacatecas dot the hills like dismembered vertebrae of ancient monsters.489 Water
played a vital role in keeping the surrounding areas of the amalgamation haciendas free of
mounds of debris, so that as well as providing power to the mills, keeping the slurries wet so
the chemical reactions could take place, washing the ore slurries, slaking the thirst of the
animals that ploughed blindfolded through the tortas, and for the workers as well, it was water
that washed away the mounds of silt and the calomel, mercury and excess salt and other
chemicals generated by these industrial units.
3.9 The mass balance of amalgamation: the role of the correspondencia
There is one mathematical constant to the history of amalgamation in the New World.
Barely five years after this process was being applied in New Spain, it became an established
489
Lawsuit brought by the Hacienda San Tadeo against the Hacienda La Sauceda for constructing an aqueduct
without permission across their land to bring water from the mines in Vetagrande (Zacatecas) to the Hacienda La
Sauceda, 11 February 1808, EHEZ, Poder Judicial Civil C56-E07.
252
rule of thumb that two weights of mercury were consumed for every weight of silver
extracted.490 What is remarkable is that Duport in the mid nineteenth century commented that
the ‘loss’ of mercury per mark of silver reported from 1570 to 1585 was the same he observed
on average in Mexico some three hundred years later.491 The authorities quickly took advantage
of this inherent ratio of the process and used it as the benchmark in their apportioning of
mercury to the refiners, to attempt to control the production of contraband silver. This
benchmark was called the correspondencia and it is usually expressed as marks of silver
produced per quintal of mercury consumed. The rule of thumb value for the correspondencia
was usually 100 marks of silver to one quintal of mercury, a weight ratio of 2:1 (mercury to
silver), but in practice it varied according to location. This consumption of mercury was broken
down in amalgamation lore into one part of mercury consumed for one part of silver produced,
and the remainder was deemed to be the physical loss of the process. ‘The principle that the
loss of mercury was at least equal to the weight of silver obtained, is a prejudice so embedded
in most of the azogueros, that it is a waste of time to discuss with them on this issue’.492 The
490
‘with a quintal of mercury they only extract half [a quintal] of silver’ - ‘con un quintal de azogue no sacan mas
que medio de plata’, letter from the Viceroy of New Spain to the King, 30 July 1561, as quoted in Castillo Martos,
Bartolomé de Medina, 145. In the discussion that follows the point of reference is patio amalgamation. Barba’s
cazo process shows very low levels of mercury consumption because the underlying chemistry during refining is
completely different.
491
12 oz of mercury per mark of silver. Duport, Métaux précieux au Mexique, 144. At the close of the eighteenth
century Garcés y Eguía had already commented upon the theoretical implications of a common value of mercury
consumption in relation to the aleatory nature of human operational skills: ‘the particular luck or misfortune of
this or that miner [refiner] cannot question the result of calculations, since these arise from a standard behaviour
of all the body [of refiners]; so that the consumption of mercury does not depend on this or that individual, but on
all together … unless it can be proved that there is a problem common to all the body’ – ‘la fortuna o desgracia
particular de uno o de otro Minero, no puede hacer falible el resultado de los cálculos, porque estos proceden
según la regularidad de todo el cuerpo; al modo que los del consumo de azogue no penden de este o aquel
individuo, sino de todos juntos … mientras no se verifique un mal que comprehenda a todo el cuerpo’, in Garcés
y Eguía, Nueva teórica del beneficio de plata, 3. What in fact was common to all the body was the chemical
underpinning to the correspondencia.
492
‘Le principe d’une perte de mercure égale au moins au poids de l’argent obtenu, est un préjugé tellement
enraciné chez la plupart des azogueros, que c’est peine perdue de discuter avec eux sur ce point’. Duport, Métaux
précieux au Mexique, 119.
253
obstinacy of this myth of a one to one conversion betrays its alchemical roots in the assumption
that mercury is transmuted into silver during amalgamation.
For the majority of silver ores refined in the New World the need to constantly replenish
mercury stocks destined for the amalgamation process was mainly due to the chemical reaction
that transformed it into calomel, the chemistry of which was indicated in Section 3.4 above. By
the nineteenth century chemical knowledge had matured to the point it became possible to
quantify this chemical consumption of mercury. In 1872 Manuel Maria Contreras calculated
the weight of mercury transformed into calomel as 1.85 times the amount of silver recovered,
and conditioned this ratio to the absence of competing reduction routes for silver chloride.
Though the author recognized that the physical losses during washing would alter this value,
he did not proceed further into the consequences of his quantitative analysis.493 The more recent
historiography has not echoed his analysis, and has treated the correspondencia as a useful
number to estimate silver production by amalgamation, for the lack of a better alternative. 494
And yet the correspondencia factor is not just a number whose only discernible connection to
493
Manuel Maria Contreras, "Empleo de los ensayes de pella y de residuos para determinar los adelantos y fin de
la amalgamacion de la plata en el beneficio de patio.," in Historia de la ciencia en México: Estudios y textos., ed.
Elias Trabulse (Mexico: Conacyt y Fondo de Cultura Economica, 2003), 721-38. The original paper was published
in Anales del Ministerio de Fomento de la Republica Mexicana, Imprenta de Francisco Diaz de Leon, tomo X,
México, D.F., 1872. Contreras fought against the North American invasion of Mexico when he was 14, then
graduated as a mining engineer, was named assayer of the Casa de la Moneda, became a politician (member of
Congress and Senator), a noted mathematician and author of textbooks on mathematics and geometry. There is a
small town in Mexico named after him, in the State of Veracruz, and also a street in Ciudad de Mexico.
http://biblioweb.tic.unam.mx/diccionario/htm/biografias/bio_c/ contreras_manuel.htm. At the turn of the century
Humboldt had written that ‘if, in the patio process, all the silver extracted was due to a decomposition of silver
chloride by mercury, the ratio of mercury lost to that of silver in the chloride would be approximately 4:7,6, since
that is the respective oxidation [value] of the two metals’ - ‘Si, dans le procédé por patio, tout l’argent retiré étoit
dù à une décomposition de muriate d’argent par le mercure, il se perdroit une quantité de mercure que seroit à
celle de l’argent dans le muriate, à peu près comme 4 :7,6, car cette proportion est celle des oxidation respectives
des deux métaux.’. Humboldt, Essai politique, Tome IV, 80-81. Humboldt was therefore the pioneer in this correct
chemical approach to the problem of mercury consumption during amalgamation, but reverses the weight ratio so
that it stands at 1 to 1.9, mercury to silver. In another part of his work he states that the mercury loss was between
1.4 to 1.7 per kg silver in patio, and around 0.2 in the barrel process. Ibid., Tome IV, 68. Whether the inversion
in the first ratio was an editing mistake, an ambiguous reading of the original in German or an error in the
‘oxidation’ values he adopts for silver and mercury is not clear, but Humboldt may have predicted the theoretical
basis for the mercury to silver weight ratio as early as the 1800s.
494
One of the most recent examples is in Lacueva Muñoz, La plata del Rey, 204.
254
the amalgamation process is a spurious stability in value throughout three centuries of refining
history in the New World. On the contrary, it is a number with a sound theoretical base, the
mathematical consequence of the chemical correlations inherent to the amalgamation process,
and of the physical losses of mercury incurred, as I will argue in the following paragraphs.495
The observation that a relatively constant weight ratio applied to the consumption of
mercury per kg of silver refined is strongly indicative that the underlying reason is the
stoichiometry of the chemical reaction that involves both mercury and silver.496 Returning to
the sequence of reactions specified in Section 3.4, the two steps of the basic amalgamation
reaction, once the silver sulphides present in the ore have been converted into silver chloride,
can be condensed into the following equation:
(n + 2)Hg(l) + 2AgCl(s)  Ag2Hgn(l) + Hg2Cl2(s) Reaction 4
This chemical equation tells us that in the absence of native silver in the ore, of iron or
copper metal, and excluding other side reactions of mercury or physical losses, for every mole
of silver (107.87 g) refined by amalgamation, one mole of mercury (200.59 g) will be consumed
in its transformation into solid calomel.497 In this scenario, the theoretical weight ratio of
mercury consumed to silver produced using amalgamation would be 1.86. This is equivalent
to a correspondencia of just over 110 marks of silver per quintal of mercury under the
conditions cited above. A selection of historical correspondencia values reported for Peru and
495
The main results presented in this section have been published in Saúl Guerrero, "Chemistry as a Tool for
Historical Research: Identifying Paths of Historical Mercury Pollution in the Hispanic New World," Bulletin for
the History of Chemistry 37, no. 2 (2012).
496
Stoichiometry: in chemistry, the determination of the proportions in which elements or compounds react with
one another. Encyclopaedia Britannica, 15th ed.(Chicago: Encyclopaedia Britannica, Inc., 1992), Micropaedia,
11, 279.
497
The presence of native silver, which does not react chemically with mercury, will reduce the consumption of
mercury. This fact is remarked upon by Egleston, The Metallurgy of Silver, 307.
255
New Spain during the period of interest is summarized in Table 3- III.498 The mercury to silver
weight ratios have been calculated for each correspondencia value, and have all been given
equal weight to arrive at an average value of 1.8 ± 0.3. This historical average is very close to
the theoretical value of 1.86 calculated above, taking into account that an additional physical
loss of mercury was also taking place.
To incorporate the effect of physical losses on the correspondencia value I start with a
very simple mass balance for an ore that contains no native silver and where no physical losses
are involved (Figure 3-22). Mercury was added to the ore in a range between 5 to 10 times the
deemed amount of silver estimated to be extracted in the ore. I will use as a starting point a
proportion of seven to one in weight.499 Thus if I begin my amalgamation cycle using 100kg
of mercury, that means I am treating a total quantity of ore that holds 14 kg (to the nearest
integer) of silver in the form of chlorides or sulphides. It is not necessary to know the total
quantity of ore treated or the exact proportion of either silver compound. According to Reaction
4, for each kilogram of silver extracted, the amalgamation reaction converts 1.86 kg of mercury
into calomel. It is irrelevant for this exercise whether the silver chloride that is reduced by
mercury was present originally in the ore or is the result of the conversion of silver sulphides
498
The sources for the data in Table 3-III are the following: a) G. Cubillo Moreno, "Los dominios de la plata: el
precio del auge, el peso del poder. Empresarios y trabajadores en las minas de Pachuca y Zimapán, 1552-
1620,"(Col. Divulgación, Serie Historia, Instituto Nacional de Antropología e Historia/Consejo Nacional para la
Cultura y las Artes, México, 1991). b) Gómez de Cervantes, Nueva España siglo XVI. c) Peter J. Bakewell, "Notes
on the Mexican Silver Mining Industry in the 1590's," in Mines of Silver and Gold in the Americas, ed. Peter J.
Bakewell (Aldershot (GB); Brookfield (Vt.): Variorum, 1996). d) "Registered Silver Production in the Potosi
District 1550-1735," Jahrbuch für Geschichte Lateinamerikas 12 (1975). e) complaint by miners on the price of
mercury, 26 April 1679, AHEZ Notarías/Colonia, Number 5 (Felipe de Espinosa 1653 - 1680), expediente 9. f)
lawsuit against the Count of Santa Rosa, for mercury debt to the Royal Treasury, 6 December 1692, AHEZ, Real
Hacienda- Judicial 1690. g) Silver and Entrepeneurship in Seventeenth-Century Potosí. The Life and Times of
Antonio López de Quiroga (Dallas: Southern Methodist University Press, 1995). h) Newson, "Silver Mining
Honduras." i) Brading and Cross, "Colonial Silver Mining: Mexico and Peru." j) Arduz Eguía, Minería
altoperuana. k) Mendizábal, La mineria mexicana. l) Burkart, "Mines de Veta-Grande." m) Duport, Métaux
précieux au Mexique.
499
Five to six parts of mercury to one of deemed silver content according to Amador, Tratado práctico de
haciendas de beneficio, 75. ; Hermosa, Manual de Laboreo de Minas, 216.; ten parts of mercury to one of silver
in Duport, Métaux précieux au Mexique, 269.
256
according to Reaction 3. The critical assumption is that only mercury reduces the silver chloride
to elemental silver.
Mercury Silver Location Period Hg/Ag Source
1 quintal 100 mark Pachuca, New Spain 16 c 2.1 a, 165
1 quintal 115 mark Pachuca, New Spain end 16 c 1.8 a, 184

1 lb l mark New Spain 1580s 2 b, 154-155
1 quintal 115 mark New Spain 1590s 1.8 c, 175
1 quintal 150 mark Potosí, Peru 1588 1.4

d, 82
1 quintal 160 mark Potosí, Peru 1635 1.3
300 quintales 36,000 marks Zacatecas, New Spain 1679 1.7 e, 63r
165 quintales 16,500 marks Zacatecas, New Spain 1690 2 f, 33r

13,000 lb 7800 lb Potosí, Peru 1.7 g, 59-60
1 quintal 80 mark New Spain 2.6
c, 175
1 lb 1 mark Honduras 2 h, 53
1 quintal 100 mark New World best practice colonial era 2.1
1 quintal 85 mark Bolaños, New Spain 2.4

i, 556
1 quintal 125 mark Guanajuato, New Spain 1.6
1 quintal 112-126 mark Zacatecas, New Spain 1.6-1.8
Potosí, Peru 1750s 1.5 j, 105
1 quintal 80 mark San Luis Potosí, Sultepec, others, New Spain 1770s 2.6 k, 75
1 quintal 125 mark Guanajuato, New Spain 1.6
Zacatecas, New Spain 1835 1.5 l, 80
12 oz 8 oz Guanajuato, New Spain 1.5 m, 118

mid 19 c
10-24 oz 8 oz Guanajuato, New Spain 1.2-3 m, 119
1 lb 1 mark New Spain 1836 2 m, 134
12 oz 8 oz Catorce, New Spain 1.5 m, 143
15 oz 8 oz Zacatecas, New Spain 1.9 m, 251

mid 19 c
12-14 oz 8 oz Fresnillo, New Spain 1.5-1.8 m, 275, 279
8-13 oz 8 oz Guadalupe, New Spain 1-1.6 m, 319, 328
Table 3-III. Historical values of Hg/Ag weight ratios calculated from values of
correspondencia reported in the historiography. Sources from footnote 498.
257
14 kg Ag in ore 100 kg Hg
loss of Hg as solid
calomel (Hg 2Cl2)
during washing
with water 4 kg of Hg
recovered
26 kg of Hg
Liquid excess Hg squeezed
lost as solid
from amalgam
calomel
70 kg of Hg recovered
by condensing of
vapours during firing
Solid amalgam (pella) of pella
containing 70 kg Hg
and 14 kg Ag
solid
liquid 14 kg refined as Ag
Figure 3-22. Simplified mass balance of amalgamation process
I further assume that in a single cycle I extract the totality of the 14kg of silver, which
then implies a total chemical conversion of 26 kg of mercury into calomel. All throughout this
exercise I only refer to the weight of mercury lost as calomel, not to a weight of calomel
produced. The composition of the amalgam after the excess mercury was extracted was
approximately 5 parts mercury to one part silver.500 The amalgam is assumed to contain 14 kg
of silver and 70 kg of mercury, while 4 kg of mercury were extracted in the manga.
500
This is an average proportion, that is found at both ends of the historical period under study. For example see
Capoche, "Villa Imperial de Potosi," 124.; Amador, Tratado práctico de haciendas de beneficio, 75.
258
In summary, this simple mass balance shows that after extracting 14 kg of silver there
are 74 kg of mercury recovered in two different batches, one by a simple operation of squeezing
a liquid amalgam through a cloth, and the remainder by heating the amalgam in the mercury
recycling assemblage (caperuza, desazogadera, capellina) and condensing the mercury
vapours. In addition 26 kg of mercury are converted to solid calomel. I will now designate by
fa the fraction of silver present as silver chloride or silver sulphide in the ore, with values
between 0 and 1. It is important to underline that the variable fa is not the silver content in the
ore. In fact, the same fa value can apply to two silver ores with quite different silver content.
Thus the weight ratio of mercury lost as calomel to silver refined for ores containing native
silver as well as silver chlorides and sulphides can be expressed in the following manner:
[Hg/Ag]calomel loss = 1.86 fa
If the silver content in the ore is made up of pure native silver, fa is equal to zero and no mercury
is lost as calomel. If the silver content is made up only of silver chloride and sulphide, then the
weight ratio of mercury lost to silver extracted will be 1.86. I am ignoring further chemical
losses in the form of secondary reactions between mercury and excess copper magistral that
can produce calomel or mercury with sulphur to produce cinnabar, but they could be factored
into the equation in an analogous manner.
The most simple way to include the effect of physical losses on the weight ratio is to
assume a single physical loss factor (fb) between zero and unity that encompasses both mercury
lost through volatilization and mercury lost through washings and spills. By taking this
approach the exact amount of mercury remaining in the amalgam does not need to be known.
Based on the amalgamation recipe of seven parts mercury to one part extracted silver from the
ore, the weight ratio of mercury to silver due to physical losses of mercury will then be:
[Hg/Ag]physical losses = [ 7 – 1.86 fa] fb

259
The term between brackets on the right-hand side of the equation corresponds to the
weight ratio after eliminating the chemical loss. It would be possible to separate ‘cold’ losses
of mercury (spills, washings) from ‘hot’ losses (volatilization) but for the purpose of this
exercise I will continue using a single loss factor, fb.
The total weight ratio, taking into account both chemical and physical losses, can now
be expressed as:
Hg/Ag = 1.86 fa + [ 7 – 1.86 fa] fb
where
Hg/Ag = weight ratio of mercury lost to silver extracted
fa : weight fraction of silver chloride and sulphide in the silver present in the ore,
between 0 and 1
fb : total physical loss of mercury during amalgamation, expressed as a weight fraction
between 0 and 1
The ratio Hg/Ag, as well as the correspondencia, is independent of the total silver
content of an ore.
The relevance of the equation is that it demonstrates that the values of the Hg/Ag ratio
(as representing the correspondencia), silver content (as chloride and sulphide) and physical
losses are not three independent variables. Thus each of the three constrains the values the other
two can adopt. It is easier to visualize this interdependence in Figure 3-23. Each value of the
Hg/Ag ratio can be calculated from different pairings of fa and fb values. Three values of this
ratio are plotted as curves in Figure 3-23: 1.5 (lower limit of standard deviation of historical
average), 1.8 (historical average) and 2.1 (upper limit of standard deviation of historical
260
average) respectively. The only pairings of fa and fb values that are relevant to the present
discussion are those that fall between the bottom and top curves representing the two extremes
of the historical Hg/Ag ratio. In addition, the silver ores treated by amalgamation were the
negrillo ores, rich in silver sulphide and low in native silver, which would therefore have fa
values that tend to unity. I will assume as a working figure that the most likely range of f a for
these ores lies between 0.75 and 1. As can be seen in Figure 3-23, this in turn would limit the
values of fb to below 0.15. The rectangle in grey visualizes the ‘boxing in’ of fb imposed by the
value of fa and the range of observed Hg/Ag ratios.
Figure 3-23. Sensitivity of Hg/Ag weight ratio to the fraction of silver chloride and sulphide
of the total silver present in the ore (fa) and on the fraction of physical loss (fb). Reproduced
from footnote 495.
If for the treatment of negrillos (assumed range of fa values between 0.75 and 1) higher
physical losses are assumed, i.e. values of fb well over 0.15, the mercury to silver weight ratio
would quickly reach values much greater than 3, well above the historical range reported in
Table 3-II. This is illustrated by the plots in Figure 3-24, where the grey area denotes the
average range of observed Hg/Ag ratios. Only ores very rich in native silver (low to zero fa
261
values, not shown) would mathematically allow a high physical loss of mercury, but since these
were smelted rather than amalgamated, they do not represent a historically representative case.
4.0
Hg/Ag weight ratio
3.5
fa=1
3.0
fa=0.75
2.5
2.0
average range of historical values
1.5
1.0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
fb
Figure 3-24. Sensitivity of the Hg/Ag weight ratio to fb values, for the range of fa values
expected from ores refined by amalgamation. Reproduced from footnote 495.
To calculate the breakdown between mercury consumed in its transformation to
calomel and mercury lost through physical causes the following equations apply:
1.86 𝑓𝑎
% Hg consumed as calomel = (1.86 𝑓𝑎+(7−1.86 𝑓𝑎)𝑓𝑏 x 100%
(7− 1.86 𝑓𝑎)𝑓𝑏

% Hg physical loss = (1.86 𝑓𝑎+(7−1.86 𝑓𝑎)𝑓𝑏 x 100%
The expected range for total physical losses of mercury during amalgamation is
calculated using these equations from approximately 10 to 30%, subject to the value of the
correspondencia (Table 3-III). The values in Table 3-IV have been obtained by fixing the value
of the mercury to silver weight ratio, and choosing a range for fa consistent with the historical
262
reality that it was negrillo ores (silver sulphide) which were destined primarily for
amalgamation.
fa 0.75 0.75 0.85
fb 0.13 0.02 0.05
Hg/Ag 2.1 1.5 1.86
% mercury converted to
69% 91% 85%
calomel
% mercury physical loss 32% 10% 15%
Table 3-IV. Percentage breakdown of mercury losses to the environment as a result of the
refining of silver ores with mercury.
Since the values of correspondencia are not the result of an aggregate of contingent
human errors of operation but are dictated mainly by the chemistry of the reactions during
amalgamation and the nature of the ore, it converts this index from a passive mirror of
operational empiricism into a theoretically solid tool of historical analysis.
3.10 The environmental impact vectors of amalgamation
Of all the components of an amalgamation recipe applied to silver ores, the usual
environmental suspect that springs to the forefront for a modern reader is mercury. The
historiography of the period also focused most of its attention on the consumption of mercury,
but for other reasons. Within the lore of amalgamation it was accepted that there was a fixed
one to one ratio between the weight of mercury ‘destroyed’ during the process and the weight
of silver refined. This was called ‘el consumido’, or the consumption of mercury. Any
263
additional need for mercury was due to operational losses (‘el perdido’) during the process.501
Thus from the very beginning the total consumption of mercury was interpreted as the sum of
two processes, one chemical (the consumption as a transformation of matter), and one physical
(the loss of liquid mercury due to mechanical causes).
The historiography was also well aware of the dangers posed by mercury to the workers
exposed to its vapours, and Chapter Six will review the historiography on the risk to the health
of workers posed by mercury during this historical period. Barba criticizes the careless use of
equipment to heat the amalgam, so that if joints are not properly sealed and a poor quality clay
is used to make the recipients there is a danger of volatile mercury escaping to the
environment.502 Barba also goes on to strongly suggest the use of iron or beaten copper
recipients as the most secure means to recover the mercury from the amalgam.503 The warning
on the accidental exposure of amalgamation workers to heated and volatile mercury continues
throughout the period in question.504 It was not only the health of the workers that spurred this
concern. As Gomez de Cervantes recognized very early, the economic importance for a careful
husbandry of mercury stocks from the point of view of the refiner merited a strict supervision
of the efficiency of the recycling stage:
501
Laur, "De la metallurgie de l'argent au Mexique," 163.
502
503
Ibid., 101.
504
‘In all these works he who is present should always place himself upwind of the furnaces, so that if a vessel
should break... the smoke of the mercury will not cause harm ... which is very great.’-‘En todas estas obras se
ponga siempre, el que a ellas asistiere, a barlovento de los hornos, por el riesgo de que quebrandose algun vaso
... no cause el humo del açogue los daños ... que son muy grandes’ in ibid., 170. According to Sonneschmidt the
older type of clay vessels could break easily, so that workers who came to dampen the embers were at risk of
mercury poisoning: ‘I have found various individuals that in such circumstances have been poisoned by mercury,
and fell to the floor senseless’ -‘He encontrado a varios sugetos que en tales circunstancias se han azogado, y
cayeron en el suelo privados de sentidos’. Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva
España, 51.
264
‘If mercury is lost and the loss cannot be remedied … this is the greatest loss that can be had,
and it can happen to a miner during the refining process; and when it happens, not even the
value of the silver [refined] can match the value of the mercury that is lost’.505
The successful control of the loss of volatile mercury during the heating stage of the
amalgam is amply commented upon in the technical descriptions that are available for New
Spain. Thus Born states that by the end of the eighteenth century iron vessels [capellinas] were
used to recycle mercury using a water channel such that ‘all the quicksilver is recovered without
loss’.506 Humboldt does not even mention volatile losses of mercury during the heating stage
of the amalgam in his analysis of the possible causes of the physical losses of mercury. 507
Sonneschmidt states that when carried out correctly the losses during the capellina stage are
minimal.508 Duport cites total mercury losses during the recycling stage of mercury of just
0.06%.509 According to Laur ‘except in case of accidents, or negligence, the losses of mercury
caused by the operation [the capellina stage] are of little importance [around 0.001%]’.510
Losses of volatile mercury could occur through accidents, but the available historiography up
to the end of the nineteenth century concurs that normal practice did not incur a regular loss of
volatile mercury to the air of any relevant magnitude.
If mercury was not lost to the air, how did contemporary observers interpret the fate of
the consumido and the perdido? For the former, I have conjectured on the alchemical roots for
the notion of an equivalent weight of mercury required to be transmuted into silver. However,
beyond the lore of the azogueros that exasperated Duport, by the nineteenth century it was
505
‘Si se pierde el azogue y no se puede remediar ... es la mayor pérdida que puede haber y sucede al minero en
el dicho beneficio; y cuando sucede, no llega el valor de la plata al azogue que se pierde.’ Quoted from Gomez
de Cervantes (1599) in Bargalló, Minería y metalurgia colonial, 245-46.
506
Born, Born's New Process of Amalgamation, xxi, 133.
507
Humboldt, Essai politique, Tome IV, 80-81.
508
Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 53.
509
One ounce (28 g) per quintal (46 kg), in Duport, Métaux précieux au Mexique, 117.
510
‘a moins d’accidents, d’ailleurs fort rares, ou de négligence, les pertes de mercure causes par l’opération sont
peu importantes … [1.3 to 1.75 per thousand]’ in Laur, "De la metallurgie de l'argent au Mexique," 176.
265
clear that the formation of a salt of mercury was responsible for the chemical consumption of
mercury, as I reviewed in Section 3.9 above. As to the perdido, its ultimate fate was the
waterways and the soil, according to the historiography up to the nineteenth century. The
physical loss that was caused by entrainment in the water used to wash the slurries or ore in
order to separate the amalgam extended also to unextracted silver and amalgam. On the heels
of the implementation of amalgamation, the observation was made in 1571 on ‘the loss or at
least the greater part of quintales of mercury and silver that is taken by the river … in this New
Spain’.511 In the 1740s Dominguez de la Fuente would write:
‘the drawbacks of the present refining process [patio amalgamation] run manifesting
themselves in the creeks and streams close to the haciendas, from where the women regularly
extract silver, and who are known by the title of ‘Plomilleras’ … the first is from the same
silver that comes from the washing vats of the haciendas, the second is potential silver … that
dust … from which these women extract it [silver] by fire’.512
In 1802, the same concern persists in New Spain, where Jose Antonio de Ortega
proposes to use ‘buzos’ (divers) or dams to collect the sediment and recover the silver lost in
this manner :
‘[I] denounce to Your Excellency the Gold, and Silver, that incorporated in Mercury, exists, in
the depths and natural deposits, exists, in the Rivers, of this Kingdom [New Spain], and in that
of Peru, proceeding from the washing vats of the refining of said Metals in Haciendas that
existed and exist’.513
511
‘la perdida o alomenos gran parte de los quintales de azogue y plata que se lleva el rrio … en esta Nueva
España’, extract from a merced awarded in New Spain in 1571 by the Viceroy of New Spain to Fernando de
Portugal y Leonardo Fragoso for their method to increase the efficiency of recovering silver and mercury from
the washings, AHSLP, Colección Powell, catálogo p. 94, Patronato 18 2 ramo 42, rollo 166-7-50 21; AGN,
Instituciones Coloniales, Mineria, 28304, Volumen 17.
512
‘la insuficiencia del presente beneficio, corre manifestándose por las canadas y Arroyos, vecinos de las
Haziendas, de donde sacan regularmente las Mujeres plata, que se conocen por el titulo de Plomilleras … la
primera es de la misma plata que arrojan de los lavaderos de las Haziendas, la segunda, es de la plata potencial
… aquel polvillo ..[que] estas mujeres la sacan de el, por fuego’ in Dominguez de la Fuente, Leal Informe Politico-
Legal, 136.
513
‘denuncio a la Superioridad de Vuestra Excelencia el Oro, y Plata, que incorporado en Azogue, existe, en las
profundidades y depositos naturales , existe, en los Rios, de este Reyno, y en el de Peru, procedentes de los
labaderos de veneficio de dichos Metales de la Haciendas que existieron y existen’, AGN, Instituciones
Coloniales, Minería, 28304, Volumen 17. Ortega was a member of the Royal Basque Society and collector of
266
Humboldt concluded that the consumption of mercury during amalgamation was due
in greatest part to its entrainment in solids carried away in the water washings. 514 Laur sums
up the views held by the end of the nineteenth century that the causes of the consumption of
mercury are two, chemical reactions inherent to amalgamation and mechanical losses during
the washing of the slurry due to the extreme fineness of the particles of ore.515
With respect to losses of liquid mercury to the soil, there are references to the soil of
abandoned amalgamation haciendas being excavated in the 1670s to recover the mercury
contained within them:
‘Nicolas de Villareal, of this city [Zacatecas] stated that when digging into the foundations of
the old houses that belonged to General Agustin de Zavala … within the soil that he dug from
the place where the hacienda of the General used to be, he found samples of mercury and
because if he washes more soil he may recover more quantity of said substance … [he requests]
that your lordships decide how best to serve [the interests] of His Majesty’.516
Hermosa, another first-hand observer of the process, proposes that the main routes for
the physical loss of mercury is seepage to the soil and entrainment in the water washings. 517
Referring to amalgamation mills in the United States of America of the nineteenth century,
Egleston writes: ‘By far the greater loss [of mercury] is mechanical. The ground under some
of the old-fashioned [silver refining] mills was richer in mercury than a quicksilver mine’.518
Sonneschmidt is one of the few in the historiography to doubt that much mercury was lost in
taxes in Oaxaca, ‘socio Benemerito de la Real Sociedad Bascongada, Administrador de las Reales Rentas del
Partido de Villa Alta, Oaxaca’.
514
515
516
‘Nicolas de Villareal vzo de esta ciudad dijo que abriendo unos cimientos en las cassas viejas que fureon del
Gral Agn de Zavala … entre la ttierra que se a sacado de ellas en el cittio donde hera la hazienda que el dicho
Gral tuvo parece apinttado algunas muestras de azogue y porque puede ser que lavando otras tierras se recoga
alguna canttida de dicho genero … vms dispongan lo que mas sea del servicio de su Magestad’, dated 29 March
1673, AHEZ, Real Hacienda 1673, document with lower right hand corner missing.
517
Hermosa, Manual de Laboreo de Minas, 239.
518
267
this manner, stating that on lifting the lining of the patio he could not find evidence of mercury
droplets, though he does not state how deep in the soil he searched.519
The mention of mercury losses in the historiography up to the nineteenth century does
not mean there was an interest on the impact these losses would have on the communities or
on nature around the amalgamation haciendas. As a case in point, Duport’s excellent review
of silver refining in Mexico does not address any environmental issue. Though the main
historical concern was to explain the disappearance (or waste for some observers) of an
expensive and limited reagent critical to refine silver, some degree of care for strict operational
guidelines that would lower the risk of inhaling mercury vapours is also evidenced. The other
component of the amalgamation recipe that received attention was the amount of solids voided
into the waterways. In the case of Guanajuato, where the amalgamation haciendas were
clustered close to the town and around the main waterway that ran through it, the consequences
of dumping the fine mineral silt from the constant washing of the amalgamation tortas caused
problems that had to be addressed on a yearly basis:
‘the water from the river that run through the city was used by the refining haciendas … a city
whose mining activity regulated the economic life of its inhabitants generated waste inherent
to the refining process, these were thrown into the river causing it to flood’.520
These wastes would be dug out of the river bed and sent to landfills around the city, as
is clear from the instructions sent by the Viceroy of New Spain to the Cabildo of Guanajuato
requiring it to estimate a budget and plan for:
‘the cleaning of the River … the solid fill and filth that will have to be removed … the land
where they are to be deposited, without affecting the public, as well as the retaining walls that
519
Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, 128.
520
‘el agua del río que atravesaba la ciudad era utilizada por las haciendas de beneficio de metales … una ciudad
cuya actividad minera regía la vida económica de la población generaba desperdicios propios del beneficio de
metales, estos eran arrojados al río causando su desbordamiento’ in Alma Linda Reza, Guanajuato y sus
miasmas. Higiene urbana y salud pública, 1792-1804 (Guanajuato: Presidencia Municipal de Guanajuato, 2001),
40.
268
will be required to contain the landfills so that they do not cause harm to the city or return to
the causeway of the River’.521
The historiography of the twentieth century introduced a major change in the portrayal
and analysis of the environmental legacy of historical silver refining in the New World. 522 It
was the field of environmental science that provided the first novel interpretation of the
consumption of mercury during the historical amalgamation process practised in the Americas.
Nriagu’s landmark one-page note to the journal Nature in 1993 based its analysis on the
premise that all the consumption of mercury during the period of historical silver refining by
amalgamation was due to a 100% physical loss of mercury, of which 60 to 65% were deemed
to be due specifically to volatile losses of mercury. No mention was made of calomel. In the
same note Nriagu mentions that the consumption of mercury in the colonial amalgamation of
silver is very similar to that observed for the modern amalgamation of gold in the Brazilian
Amazon region. Further, he states that the fraction of mercury lost to the air during colonial
refining of silver ores is comparable to the 65 to 83% range of mercury losses during gold
amalgamation measured in modern times in the Amazon.523 The mainstream historiography
has followed Nriagu’s paper to propose similar ranges of loss for volatile mercury, up to 85%
521
‘en la limpia del Rio … los atierres e immundizias que habran de extraerse … los parajes en que harán de
depositarse, sin perjuicio del publico, como también los pretiles o calicantos que a caso sea necesario hacer para
contener los citados atierres que no hagan daño a la ciudad ni vuelvan a la Caja de dicho Rio ’ in AHUG, Actas
de Cabildo, 5 December 1782, 153v to 156r.
522
One of first modern works to bring attention to the harm caused by mercury to the indigenous communities as
a consequence of mining and refining of silver was Eduardo Galeano, Las venas abiertas de América Latina
(Buenos Aires: Siglo Veintiuno, 1982), 62. The value of this work as an objective appraisal of historical events
should be judged in the light of the author’s recent comments during the II Bienal del Libro in Brasilia, Brasil,
held in April 2014. Galeano confessed that ‘I would not be able to read the book again. That traditional left-wing
prose is too boring … I did not have the necessary knowledge [to write it]. I do not repent having written it, but it
was a stage that, for me, has long been over.’ - ‘yo no seria capaz de leer el libro de nuevo. Para mi esa prosa
de izquierda tradicional es pesadísima … yo no tenia la formación necesaria. No estoy arrepentido de haberlo
escrito pero fue una etapa que, para mi, fue superada’ as reported in
http://es.brasil247.com/es/247/sociedad/1199/Galeano-la-realidad-cambi%C3%B3-no-leer%C3%ADa-
m%C3%A1s-Las-venas-abiertas.htm
523
Nriagu, "Legacy of Mercury Pollution," 589.
269
in the latest proposal by Hagan et al and Robins in 2011.524 The formation of calomel continued
to be excluded from these analyses. Though mercury salts as by-products of amalgamation are
mentioned in some modern historical texts on silver refining in the New World, the majority
view is that overall it was volatile mercury that constituted the main cause for the consumption
of mercury in the New World.525
The omission of calomel as a principal factor in the consumption of mercury not only
ignored the chemistry of the process and the historical texts of the nineteenth century, but it
implied a magnitude of historical air deposition of mercury around refining sites that could not
be corroborated by new results from research on historical mercury depositions from the
atmosphere. In 2011 Cooke, Balcom, Kerfoot, Abbott and Wolfe published the levels of
mercury deposited in the sediment of Laguna Lobato, some 6 km west of Cerro Potosí, from
600 to 2000 CE. High initial levels of mercury up to ca. 1300 CE are explained as coming from
pre-Columbine smelting of native silver found on the surface of the Cerro Potosí. From that
date to the present the levels of airborne mercury deposits decrease continuously throughout
the historical period that saw amalgamation used on a massive scale in Potosí and surrounding
areas. In contrast, levels of airborne lead are seen to spike around the period large-scale
smelting was introduced by Spain in Potosí.526 This result contradicted the assumption that
major losses of airborne mercury took place in Potosí as a consequence of amalgamation. In
524
Julio A. Camargo, "Contribution of Spanish–American Silver Mines (1570–1820) to the Present High Mercury
Concentrations in the Global Environment : A Review," Chemosphere 48(2002).; S. Strode, L. Jaeglé, and N.E.
Selin, "Impact of mercury emissions from historic gold and silver mining: Global modeling," Atmospheric
Environment 43, no. 12 (2009).; Nicole Hagan et al., "Estimating historical atmospheric mercury concentrations
from silver mining and their legacies in present-day surface soil in Potosí, Bolivia," ibid.45, no. 40 (2011).;
Robins, Mercury, Mining and Empire.
525
An example is the following discussion on historical mercury losses from amalgamation: ‘other losses resulted
from binding of mercury in insoluble compounds to sulfides, chloride or other salts in the ore. Some, perhaps
most, of the losses occurred in vaporization of mercury’, from Richards, The Unending Frontier: An
Environmental History of the Early Modern World 370.
526
Cooke et al., "Pre-Colombian Mercury Pollution Associated with the Smelting of Argentiferous Ores in the
Bolivian Andes."
270
2012 I published the results presented above in Section 3.9.527 In 2014, a review paper by
Engstrom et al. argued that the evidence from a world-wide survey of historical mercury
airborne depositions is not consistent with the assumption of large scale volatile mercury losses
from historical periods of silver refining by amalgamation, and they proposed a shift in the
paradigm applied to the interpretation of historical mercury emissions, in line with the
mathematical results of the model I proposed based on calomel (Section 3.9).528
With regard to the amount of liquid mercury that seeped into the soil as a result of
historical silver refining, the research is still at an early stage. In present day Potosí, disturbing
the topsoil by excavating down to 50 cm in locations where amalgamation units used to located
will liberate very significant amounts of mercury to the air. In contrast, undisturbed sites show
very low levels of mercury in the air. The authors propose that this is due to mercury being
present in different chemical forms according to depth, thus making it less susceptible to be
airborne when in the topsoil.529 The oral history of San Luis Potosi, Zacatecas and Guanajuato
confirms in a qualitative way the presence of mercury in the soil. The Director of the Historical
Archive in San Luis Potosi, Don Rafael Morales Bocardo, recounted to me that when
excavations were being carried out on the floor of the assaying room of the Casa de la Moneda,
he saw the newly exposed soil weeping globules of mercury. Mrs. Maria del Socorro Cardoso
Girón, the local Historiadora [designated resident historian by the provincial government] of
the village of Pánuco, close to Zacatecas, told me that many of the historical haciendas lay in
ruins, with only remnants of some perimeter walls standing, because in the past these haciendas
527
Guerrero, "Historical Mercury Pollution in the Hispanic New World."
528
Daniel R. Engstrom et al., "Atmospheric Hg Emissions from Preindustrial Gold and Silver Extraction in the
Americas: A Reevaluation from Lake-Sediment Archives," Environmental Science & Technology 48, no. 12
(2014).
529
P. Higueras et al., "Mercury vapor emissions from the Ingenios in Potosí (Bolivia)," Journal of Geochemical
Exploration 116(2012): 6.; P. Higueras et al., "Multielemental pollution of soils at the Ingenios, decommissioned
mineralurgical sites in Potosí (Bolivia)," (2010).
271
had been bought just for the value of the mercury in the soil. Once all the impregnated soil had
been scraped from the grounds of the hacienda, and carted off to be processed elsewhere, it
was left as an empty shell. Finally in Guanajuato, both Mr. Morrill, owner of one of the few
standing capellina buildings which he has now converted to a studio, and Doctor Virgilio
Fernandez del Real, the venerable Spanish owner of the Hacienda Santa Ana in Marfil,
currently the Museo Gene Byron, gave me independent reports of trees that grow normally
until one day their roots strike what they believe to be a mercury rich section of the earth under
the patios and suddenly wither away and die.
The fate of calomel washed away together with the mineral silt from the washed tortas
awaits future research. In the case of Guanajuato the modern city retains physical evidence of
historical landfills from the waste dredged from its waterways where both mercury and calomel
may reside. Figure 3-25 shows the Observatorio Metereológico, of the Comisión Nacional del
Agua and University of Guanajuato, built on a hill made up from a landfill of the waste
recovered from the river.530 How much of the total waste voided initially into rivers was
returned as landfill to the land remains to be established. In other locations such waste would
have been deposited along the bed of each river, to an extent that remains to be studied.
530
Personal communication, Lic. Silvano Pozos Suarez, in charge of the Observatory. The hill of mineral waste
on which the Observatory was built lies over the road in front of the grounds of the historic Hacienda de Rocha. I
have not been able to determine if the present road was built over the historic trace of the stream that run along
the hacienda.
272
Observatory
Figure 3-25. The Observatorio Metereológico of Guanajuato (Conagua and Universidad de

Guanajuato), sited on a landfill of waste from refining haciendas. Photo taken from the grounds
of the historic Hacienda de Rocha (at present the Hotel Reales de Minas).
In summary, the relevant environmental impact vectors for patio amalgamation can be
grouped into four sets: calomel and mercury, the water-soluble reagents (salt and copper
magistral), the waste solids from the ore and finally fuel. Waterways, not air, are the main
conduit of the environmental impact vectors for patio amalgamation. Within the hacienda,
mercury was consumed and transformed into calomel during the patio amalgamation process,
and the solid and insoluble calomel was washed away into the adjoining waterways together
with the portion of the mineral ore that had no economic value to the refiner. In this waste water
were present the excess salt, iron and copper compounds used in the recipe. A huge amount of
solid waste, over 99.6% of all the ore that was processed in each hacienda, was suspended in
the washing water voided into the streams.531 Soil was the second major conduit for the
531
For the haciendas producing 600 marks of silver per month this represents approx. 75,000 kg of solids voided
per month into a waterway. For an hacienda of the size of La Escalera, it would have reached approximately
450,000 kg in the 1790s.
273
physical loss of mercury and water-soluble components of the amalgamation recipe, whether
from spills during transport or storage, or more important from seepage during the wet slurry
phase of amalgamation.
The only stage of the amalgamation process where air-borne particles of any nature
would have been a daily significant environmental issue would have been close to the molinos
and tahonas / arrastres. This area would have been the source of very high background levels
of dust and noise, with a direct impact on the workers handling the ore. 532 As to losses of
volatile mercury, it is only during the casting of silver bars that any residual mercury not
recycled during the capellina stage would have been issued to the air.533 A loss of one percent
of the total weight of silver cast into bars was judged to be a standard of good practice, and the
figure is validated by the operational accounting data I present in the next chapter.534 Losses of
mercury by direct volatilization from the tortas are in theory possible but in practice would
have been negligible.535 Any losses of volatile mercury during the heating of the amalgam
would be the result of accidents and not a consequence of normal operating conditions. Even
in such a case, the mercury that escaped would quickly deposit itself on any surface in the
immediate area, as can be deduced from the following description of artisanal gold refining
practices in the Amazon:
532
According to Egleston the use of stamp mills imparted a radial, upward movement to the particles of crushed
ore. Egleston, The Metallurgy of Silver, 176.
533
30 kg silver bars have been described as cast from moulds the Hacienda de Loreto (Pachuca) into which melted
silver has been poured using an iron ladle which was dipped into a vessel full of molten silver. Collins, Metallurgy
of Lead & Silver, 141.
534
535
A very approximate idea can be gained using the results published by Winter on the loss in mass of a 0.2 g
spherical drop of mercury exposed to a ventilated room, away from direct sunlight and winds, at an ambient
temperature around 20⁰ C. He measured a maximum weight loss of 7 µg per hour. For the sake of argument, if 6
million such drops could be assumed to exist within an average torta, it would lose 14 kg of mercury in two weeks.
Since only a minority of mercury droplets would be exposed to sun and wind, I will assume that 5% of this
theoretical total was the effective loss, less than 1 kg per month per torta. Thomas G. Winter, "The Evaporation
of a Drop of Mercury," American Journal of Physics 71 (2003).
274
‘the hot flame would burn off the mercury. It would dissolve and rise as a vapor – looking
indeed like white water vapor- to a height of about two to three metres, before condensing
again and settling back down. In this process of first rising as vapor and then settling as
droplets, the mercury would settle on any object at hand- including on men’s eyebrows, and
the hair on their heads, their moustaches, even on their forearms, as a form of eery looking
white mist.’536
A more important area of constant human contact with mercury would be during the
addition of liquid mercury to the tortas, since the workers did not use any gloves or other
implements to avoid prolonged mercury to skin contact. The same applies to the treading of
the slurry with or without the aid of horses and mules.
Figure 3-26 represents a schematic pathway for the main losses of calomel and mercury.
In the absence of iron or copper, and with a correspondencia of 1.8, the transformation of
mercury into calomel would have accounted for around 85% of the total consumption of
mercury during the patio amalgamation. The majority of the remaining 15 % would have been
physically entrained by water or seeped to the soil. A minor amount, not exceeding 1 %, would
have been lost as volatile mercury, in a regular manner each time silver bars were cast, or in
isolated accidents during the heating of the capellinas. The addition of iron to the recipe could
lower the correspondencia to 1.3, but the percentage breakdown of the consumption of mercury
remains similar. In Chapter 4 I will be analysing in greater detail the specific mass balance of
the process as carried out in a major amalgamation hacienda, including the directionality of all
these vectors.
536
Helmut Waszkis, Mining in the Americas: Stories and History (Cambridge: Woodhead Publishing Limited,
1993), 202-203.
275
Condensation
make-up of and recycling Mercury in air
Mercury in air
of mercury emissions
mercury loss emissions
Storage & Amalgamation Washing and Heating of Casting of silver

handling of torta separation of amalgam in bars
mercury amalgam capellina
Liquid mercury Liquid mercury Calomel Liquid Liquid mercury

loss loss mercury loss silver
entrained loss
in water
Figure 3-26. Main loss vectors of calomel and mercury
3.11 Final remarks
I have argued in this chapter that the chemical reactions that are an inherent part of the
amalgamation process of silver ores lead to the formation of calomel in the majority of
historical scenarios, as was pointed out by all nineteenth century sources. Thus the
environmental legacy of the patio amalgamation process in New Spain / Mexico lies dispersed
and deposited along the riverbeds and water basins where millions of tons of waste solids from
the washings of the amalgamation haciendas were voided, or in any landfills that were formed
from these wastes. It was the chemistry of amalgamation, and not its physical facet, that
consumed the greater fraction of the mercury required by the process, converting liquid
mercury into solid and insoluble calomel. Much less important was the physical loss of
mercury, a lower fraction lost in the washings or seeped into the soil where it lies dormant until
disturbed, within the high walls that still contain the spoor of past chemical processes. The
minor part that was volatilized was lost mainly during the final casting of the silver bars, while
276
the loss during the heating of the amalgam was negligible under normal conditions. The best
evidence for the control over the consumption of mercury during the process is the impressive
immutability of the correspondencia ratio over 300 years. The total consumption of mercury
did not simply depend on the skill and experience of each azoguero but was determined to a
great extent by a chemical reaction common to all amalgamation locations and historical
periods, together with the extreme care taken in conserving the scarce and expensive mercury.
I have complemented the chemical and mathematical analysis of the amalgamation
process with a much-needed visual reminder of the conditions under which amalgamation was
carried out. Patio amalgamation was for most of the time a large-scale wet and messy process
where it would have been impossible to completely control a small but constant physical loss
of mercury and other components of the amalgamation recipe by seepage to the ground or
entrainment in the waste water. No amount of planilleros or paving stones could control the
inherent leakage of this watery matrix, though minimize it they did. Each photo of a patio
conjures up the sucking and slurping noises of an ever-present mud tugging at the feet of men
and animals, a watery medium constantly seeking cracks or trickling into waste channels, its
puddles ever-present in paintings and photographs. The only stages of the whole process where
the operators could exercise the highest control over physical losses took place outside the
patio, once all the water had been separated and they could work with kilograms and not tons
of material. The careful handling of mercury and silver took place in the controlled
environment of the mercury room that housed the manga for squeezing dry the amalgam,
during the heating of the desazogueras or capellina ensembles that could be as heavy and solid
as a compact car, and finally in the casting of the silver bars. To propose that up to 85% of the
consumption of mercury took place during the heating cycle of a capellina , as do Hagan and
Collins, is to ignore the reality of the nature and conditions under which patio amalgamation
took place.
277
Any photo of an amalgamation patio is also a Rorschach test for the biases of the
viewer. For many it represented a long and primitive process that did not evolve over centuries,
proper to a country and people untouched by progress. On the other hand Humboldt grasped
the true nature of this space when he suggested covering its floor with iron or copper. An
amalgamation patio is no less than an open-air chemical reactor, symbol of an impressive
solution to the challenge of refining silver sulphide ores that were poor in lead, a unique
example of a batch industrial process carried out in a reactor that could extended laterally or
contract as needed, a pioneer vision of industry perfectly suited to the available level of human
skills and materials. It is a sign of strength, not of weakness, that the chemical maturity of the
process was reached by the end of the sixteenth century, and that it lasted for three hundred
years until a new chemical process based on cyanide displaced it.
Amalgamation required from the start a careful planning of inventory of reagents,
starting with salt and mercury, the two chemical linchpins of the process. Even copper sulphate
could have been supplanted if necessary by roasting the sulphides with salt prior to
amalgamation. The amalgamation haciendas came in a wide range of sizes, from the small
zangarros of Guanajuato to the industrial behemoth at Fresnillo, but their capacity to pollute
the environment did not offer an economy of scale. The bigger haciendas may have been more
efficient in some aspects of the operation, but both they and their smaller brethren voided into
streams over 99.6% of all the solid ore they received, accompanied by smaller fractions of
calomel, salt and copper compounds. The imbalance of these different components of the
amalgamation process is reflected in their architectural footprint, the patio and milling areas
greedy for space, while the final stages of the separation of the amalgam, mercury and silver
appear huddled together to control pilfering, until the areas of the capellinas are all but
forgotten spaces, dwarfed by the other spatial needs of the process. Dry ore and water, not fire,
set the tone for the arrangement of working areas. Where furnaces merit a mention in the legal
278
documents it is in the context of preparing magistral amd roasting rebellious ores, never in the
context of recycling mercury.
The narrative of this chapter has paid significant attention to clearing up issues that may
not appear relevant to the environmental history of amalgamation of silver ores. Apart from a
desire to tie up evident loose threads, how critical is it to seek the roots of amalgamation in
European alchemy or to question the credibility of a fully-fledged amalgamation recipe
sprouting from the metallurgical virgin forehead of Medina? The reason lies in human choice,
or its absence. If amalgamation had no roots in previous alchemical or refining experience in
Europe, then it would be possible to argue that neither the Crown nor the refiners knew a priori
what its effects on the workers and their communities would be if applied on a major scale.
Furthermore, if 1555 is anointed as the annus mirabilis of amalgamation in the New World,
after which little of technical importance took place, it casts a shadow of inevitability over the
whole history of amalgamation. If amalgamation was indeed the only technical key to unlock
silver from its New World ores, then there is little point in searching for the role of human
agency in the charting of its historical course.
The alternative is to reconstruct the roots of amalgamation via alchemy and European
metallurgy, two areas that overlapped with pragmatic ease in the sixteenth century. What was
good for gold was equally good to extract silver, and though the chemistry of the two processes
is totally different, the simple amalgamation formula happened to work on the mounds of
discarded ore left by the first waves of ignorant miners. The same geological luck of the draw
that allowed ignorant conquerors to easily smelt ores rich in native silver and silver chloride,
also allowed a simple gold amalgamation recipe of water, salt and mercury to extract silver
from the huge stockpiles of a similar type of ore that had been discarded by the impatience
born of that same ignorance. The success of the group of refiners of the Andes in the 1580s in
creating an amalgamation recipe specifically tailored for silver sulphides is in fact the real
279
starting point of the history of amalgamation in the New World. As to the contention that
without mercury there would have been no silver from the New World, the evidence so far
presented is that from a technical point of view smelting with added lead would have been able
to refine the type of ore found in New Spain. In Chapters 5 and 6 it will be argued that even
though the choice of refining method was first and foremost determined by the chemical nature
of the ores, under certain historical scenarios both refining processes would have offered the
same economic incentive to refiners.
One of the major challenges facing any historian in attempting to reconstruct the
historical environmental impact of silver refining in the New World is the dearth of hard data.
If the original architectural plans are not available (and I have set out the very limited selection
of historic plans that can be found for now, of which none correspond to smelting haciendas),
then the ruins of the existing haciendas are too dilapidated or altered by modern conversions
to be of much assistance. Similarly, as yet no surviving account books have been found from
New Spain to establish the mass flow within the haciendas with an existing architectural plan,
without which it is impossible to project the environmental impact of their operations. In the
following chapter I will present an important exception to this state of affairs. These are the
very detailed accounting records of the commercial operations carried out in the Hacienda
Santa Maria de Regla, which practised in parallel both smelting and amalgamation. Though
again no historical plan has been found for this hacienda, its imposing historical remains allow
for a straightforward reconstruction of its operational areas. The wealth of quantitative
information for this hacienda will provide in the next two chapters a much needed window into
the detailed operational and economic structure of a silver refining hacienda, and its impact on
the environment.
280
4 Hacienda Santa María de Regla.
‘Over there – one argued to oneself – was Chichen Itza and Mitla and Palenque, the enormous
tombstones of history, the archaeologist’s Mexico ... and for the businessman the silver mines
of Pachuca ... for the priest prison, and for the politician a bullet’. Graham Greene, The Lawless
Roads (1939)
‘One should view it as we did, in a thunderstorm, for it has an air of vastness and desolation,
and at the same time of grandeur ... down in a steep barranca, encircled by basaltic cliffs, it
lies : a mighty pile of building ... all is on a gigantic scale: the immense vaulted storehouses for
the silver ore; the great smelting furnaces and covered buildings where we saw the process of
amalgamation going on; the water wheels- in short, all the necessary machinery for the smelting
and amalgamation of the metal’. Fanny Calderón de la Barca, Life in Mexico (1843)
4.1 Introduction
The chemistry of the refining of silver ores remained constant from the sixteenth to the
end of the nineteenth century, when amalgamation with mercury is replaced by the use of
cyanide to extract silver from ores using a wet process. Though the chemical recipes did not
change, the physical trappings of each process did evolve with time, as seen in the substitution
of clay by metal desazogueras or capellinas, or the increasing height of smelting chimneys,
and the switch from hornos castellanos to more efficient blast furnaces. What did change
substantially was the amount of silver produced in each of the centuries that span this period.
Though the environmental history of silver refining by amalgamation and smelting in the
Americas covers nearly four centuries of continuous production, the greater part would take
place in the nineteenth century when Mexico had already become a republic.
Mexico began its struggle for independence from Spain in 1810, the year that as the
Vice-Royalty of New Spain, it was providing ‘three-quarters of all profits from Spanish
281
American holdings’.537 Eleven years of fighting would pass before the first Republic, the
Estados Unidos Mexicanos was created in 1821. It would survive an invasion by Spain in 1829
after which ‘Mexican history teetered between simple chaos and unmitigated anarchy’.538
Between 1821 and 1876 seventy five different Presidents would hold that office in Mexico. In
the north, the other Estados Unidos would annex half the territory of Mexico by 1848 (present
day Texas, New Mexico, Arizona, California, among others) and in the process access the
remaining subduction deposits of silver and gold in the American Cordillera. This would
catapult the U.S.A. from being a nonentity in silver production to being the leading producer
of silver in the world by 1872.539 Faced with a dwindling treasury, Mexico would declare a
moratorium on all repayments of foreign debt in 1861, leading to a blockade of Mexican ports
by France, Great Britain and Spain. France would then proceed to invade Mexico and install
an Emperor, Maximillian Joseph. The Republic would be restored by 1867, and the French-
backed Emperor shot by a firing squad. The toll on Mexico’s development of such a long period
of unrest, violence and foreign intervention is obvious. In 1860 Mexico had only 150 miles of
laid railway track, compared to over 30,000 miles in the U.S.A. When Porfirio Diaz embarked
on his drive to modernize Mexico (a period known as the Porfiriato), ‘in 1876, except in a few
of the larger cities, the country had scarcely been touched by the scientific, technological and
industrial revolution, or the material conquests of the nineteenth century’.540
The impact of these events in the maintenance of mines and refining haciendas has
been commented upon in the historiography, highlighted by the fact that many of the main
mining districts to the north of Mexico (San Luis Potosi, Zacatecas, Guanajuato) figure
537
Michael C. Sherman William L. Meyer, The Course of Mexican History (New York: Oxford University Press,
1979), 196.
538
ibid., 205-49.
539
Merrill, Summarized Data of Silver Production 7.
540
Meyer, Mexican History, 262-323.
282
prominently in the struggle for Independence and in the political turmoil of the nineteenth
century.541 And yet the overall data as reported in Figure 4-1 shows a remarkable resilience of
silver production in the midst of a society being violently uprooted every few years during this
century. Wars did not end with the declaration of independence or when the invading army of
the United States of America left Ciudad de México. The Hacienda de Regla whose data
sustains this chapter continued to produce silver even during the French occupation of Mexico
in the 1860s, during which it was forced to loan money to the invading French army quartered
in the region.542 This is a significant measure of the sturdy nature of the refining processes
involved. A silver refiner of the seventeenth century transported to the year 1870 could not
have understood the political reality around him. The expulsion of the Spaniards from Mexico,
an Emperor imposed by France and shot by a Mexican government firing squad, a member of
the Zapotec indigenous group elected as President of the Republic, a Catholic Church with
much reduced power, Mexican miners taking industrial action, all these issues involved
concepts that were not conceivable or did not even exist in the age where he came from.
In sharp contrast, this same individual would have had no problem putting his technical
skills to good use at any of the silver refining haciendas of Mexico in 1870. The amalgamation
process had remained virtually unaltered, and did not require imported machinery or foreign
know-how to process the silver ores. Production would not otherwise have survived the
sequence of foreign intervention into Mexico or the negative balance of payments that
characterize most of this century. Even the loss of Spanish subsidized mercury did not strangle
541
ibid., 252-62.
542
The period in question is 1863 to 1866, and loans were repaid. Fighting did not end with the retreat of the
French, between 1876 and 1877 the revolution of Tuxtepec took place. Rocio Ruiz de la Barrera, "La Empresa
de Minas del Real del Monte (1849 - 1906)" (Colegio de Mexico, 1995), 200-201, 231.
283
production. The major boom in silver production observed all over North America (Mexico
and U.S.A.) in the nineteenth century up to the 1900s, was based on the same chemical
reactions stumbled upon by trial and error in the South American Andes in the 1590s.
70,000
second half
60,000
first half
metric tons silver
50,000
complete period
40,000
30,000
20,000
10,000
0
1493-1600 1601-1700 1701-1800 1801-1900
Figure 4-1. Production of silver in metric tons, from 1493 to 1900, adapted from footnote
543.
The curve of silver production in Mexico for the nineteenth century reflects the short-
term influence of the political turbulence. The entry of foreign capital into the mining sector
took place from the mid-1820s, while by the 1850s many of these ventures had collapsed.
Overall Mexico would produce in the nineteenth century more silver (1.9 x 109 fine ounces,
nearly 58,000 t) than the sum total produced in the three previous centuries by the Vice-Royalty
of New Spain (1.4 x 109 fine ounces, nearly 45,000 t). Nearly two thirds of this production
284
took place in the second half of the nineteenth century, making the period of 1850 to 1900
nearly as productive in silver as the whole colonial period.543
The arguments presented in the previous paragraphs lead to the two propositions that
sustain the analyses made both in this and the following chapter. First, the immutability of the
amalgamation and smelting processes from a chemical point of view makes it possible to profit
from the wealth of operational and accounting data of a major silver refining hacienda
operating in the late nineteenth century.544 Since the chemistry did not change, then the period
of time between the seventeenth and nineteenth centuries would only have increased the
efficiency of the various operational steps, thus decreasing the quantity but not the nature of
the chemicals generated per kg of refined silver that had an impact on the environment.
Therefore the nineteenth century values can be interpreted as a historical minimum, a baseline
with regards to emissions per unit of silver produced and other operational variables. Second,
the environmental impact of silver refining in Mexico from 1850 to 1900 would have been
more intense on a yearly basis than at any other period since the arrival of the Spanish silver
refiners. The total emissions from silver refining from the last fifty years of the nineteenth
century would have been chemically identical and quantitatively close to the total emissions
from the sixteenth to the eighteenth century.
543
Merrill, Summarized Data of Silver Production 8, 10. Merrill’s data gives a total from 1493 to 1799 of 1,441
million fine ounces, 44,833 t, which correlates well with the total of 44,211 t for an equivalent period reported by
TePaske and Brown, Gold and Silver, 113. For the purposes of this chapter Merrill has the advantage of reporting
production for the nineteenth century which coincides with other data sets in the historiography, as reviewed in
Chapter 6, section 6.3.14. He reports the data for the first and second half of the nineteenth century, which
therefore smooths out over 50 years both the fall in silver production during Independence and the boom of the
latter part of the century.
544
The realm of chemistry follows Einstein’s observation that ‘the distinction between past, present, and future is
only an illusion, however persistent’. The quote is from a letter of condolence sent to the family of a friend,
Michele Basso, recently deceased, dated 15th March 1955. Einstein would die just over one month later, on the
18th April 1955. As cited in The Oxford Dictionary of Quotations (Oxford; New York: Oxford University Press,
1979).
285
Any conclusion on the environmental impact from silver refining as a whole in New
Spain / Mexico obtained from a case study from the latter half of the nineteenth century is as
relevant as any other example within the whole 350 year period. This chapter will therefore
reconstruct the material balances for amalgamation and smelting as practised in the second half
of the nineteenth century at the Hacienda de Regla, situated near Pachuca in the modern state
of Hidalgo. The mass balance of these processes is a very powerful tool to estimate the
emissions of heavy metals and other substances to the environment as a result of the refining
of silver. In modern environmental impact studies it is possible to measure in-situ and in real
time the amounts of chemicals voided into the environment, but this option is obviously not
possible in the study of historical industrial pollution sites. The existence of detailed accounting
books that track the consumption of all materials including energy, concurrently with the
production of silver, provide the only option to arrive at an order of magnitude of historical
emissions to the environment that can be calculated with great precision down to the daily,
weekly or monthly level at any time from the sixteenth to the nineteenth century, if the extant
records can be found. This level of detail of specific historical periods is impossible to achieve
with current methods that determine the historical deposition of elements in the soil through
the dating and analysis of cores of soil samples.545
The method is based on the principle of the conservation of matter: the weight of all
materials consumed in the hacienda is equal to the weight of all materials produced at the
hacienda, silver and waste products. Though none of the historic accounting books used in this
chapter track waste produced (in modern industry all waste has to be accounted for and
monitored), they do register in detail all materials consumed and the silver produced, which
545
For a very technical discussion on the advantages and limitations of the current methods of dating samples see
for example Merritt R. Turetsky, Sturt W. Manning, and R. Kelman Wieder, "Dating Recent Peat Deposits,"
Wetlands 24, no. 2 (2004).
286
leaves a simple subtraction to arrive at figures on waste. It is of critical importance to correctly
identify the chemical reactions taking place and the physical causes of loss, which explains the
role of the previous two chapters in the development of my argument, so as to be able to allocate
with confidence the weight of waste according to each chemical species voided to the
environment.546
In addition, the detailed knowledge of mass balances provide a quantitative fleshing out
of the physical structures observed in the architectural remains of silver refining haciendas.
The data give credibility to spaces designated as storage areas, establish the importance of clear
transit corridors, and connect milling infrastructure with required monthly throughput
capacities. In aggregate they allow a more informed analysis of the internal distribution of areas
as observed in the plans of the previous chapter and of the reconstruction carried out in this
chapter for Regla. Correlations can be established between the areas of amalgamation patios
and the refining output of an hacienda, a very valuable indicator since patio areas are usually
the only remaining physical evidence of these historical production units. The accounting
information will also provide needed insights on the average silver content of the ores, the
composition of amalgamation tortas and the technical reasons for the amalgamation periods
adopted. Finally, the following sections leave no doubt as to the efficiency of the workforce
and managers without which these historical units would not have been capable of producing
such major amounts of silver on a continuous basis for over three centuries.
546
I have not come across a similar study for other historical polluting industries based on accounting records, but
I cannot affirm this is the first example of the method.
287
4.2 The Adventurers in the Mines of Real del Monte
At first reading it is not evident why Graham Greene would have chosen to single out
the silver mines in Pachuca in his book on travels in early twentieth-century Mexico.547
Guanajuato would have been a better example from a business point of view. 548 It seems that
like the Cornish-inspired pasties that are now staple fare for Mexicans in Pachuca, a snippet of
its past held on and managed to flourish in Greene’s mind, whispers of ‘that second South Sea
delusion, the Anglo-Spanish American mining fever’ that some one hundred years earlier had
surged, then burst, among mining investors in England.549 After all, the mines of Pachuca
would have lost their English investors some $5 million by mid-nineteenth century.550
Unfortunately for those caught up in the investment fever, the munificence of the Pachuca
mines occurred both before the English arrived with their steam engines, Cornish miners, and
substantial overhead costs, and after they left all the infrastructure to the gain of the Mexican
owners who took over.
The impressions set down by previous English travellers to the Hacienda de Santa
Maria de Regla, with its imposing walls and vaults reflecting roots of great opulence, may also
have left their imprint on the mind of the writer. The proximity to the Mexican capital (Figure
4-2) and the lure of silver mining were by themselves sufficient reason to attract foreigners
547
Graham Greene, The Lawless Roads (London: Longmans, 1939), 23.
548
‘The two greatest British concerns had their offices in Guanajuato-the Anglo-Mexican Company and the United
Mexican Company’ in Percy F. Martin, Mexico's Treasure House (Guanajuato) An Illustrated and Descriptive
Account of the Mines and their Operations in 1906 (New York: The Cheltenham press, 1906), 211.
549
Robert Anderson Wilson, Mexico and Its Religion; with Incidents of Travel in that Country During Parts of
Years 1851-52-53-54 (New York: Harper & Bros., 1855), 354. Cornwall and Pachuca would cross paths because
the General Manager of the English Company that would invest in Real del Monte believed firmly in the
advantages of dressing an ore prior to refining, and in his opinion the experts in dressing ores in England were
only found amongst the tin miners of Cornwall. John Taylor, Selections from the Works of the Baron de Humboldt,
Relating to the Climate, Inhabitants, Productions, and Mines of Mexico with Notes by John Taylor (London:
Printed for Longman, Hurst, Rees, Orme, Brown, and Green, 1824), xxv.
550
John H. Buchan, Report of the Real del Monte Company, Mexico (London: Taylor, Printer, 1855), 2. For a
history of the English investment in the Real del Monte Company see Robert W. Randall, Real del Monte: a
British Mining Venture in Mexico (Austin: The University of Texas Press, 1972).
288
travelling in Mexico to visit the Pachuca area.551 The unique architecture of the hacienda built
by the first Conde de Regla in the latter half of the eighteenth century, nestled amongst unique
scenery, left its mark not only in the diary of Fanny Calderón de la Barca: 552
‘the Hacienda de Regla ... forms the most extraordinary mass of buildings I ever saw in my life
... the prison-like castle, with its mining works – a gigantic, strong, irregular pile of household
building, over dungeons, vaults, and tunnels, with magazines, spires or turrets, courts, back
yards, furnaces, smelting and amalgamation works’.553
‘nothing can exceed the fairy-like vision which the situation of this place presents. It is entirely
surrounded, except at its entrance, by the most superb amphitheatre of perpendicular basaltic
columns that I ever beheld; far superior to anything that I saw two years ago at the Giant’s
Causeway in Ireland ’554
‘on the following morning we rode to Regla ... it is now an immense ruin, crowded with
monstrous arches of masonry, which appear as if they had been constructed as if to support the
world... preparations of a more useful nature were now in forwardness amongst the mighty
ruins but nothing could relieve the air of desolation, which gave the Hacienda the appearance
of a battered fortress. It lies deep in a precipitous Barranca, fenced in by fine basaltic cliffs, of
which so much has been said; and close above it is the celebrated Fall of Regla [El Salto] ...
the ravine is one of the most beautiful and perfect basins of basalt in the world’ 555
The massive walls and arches of the Hacienda de Regla that so impressed the visitors
of the nineteenth century had in fact been built at the end of the eighteenth century by Pedro
de Terreros, first Conde de Regla.556 It was named after the Virgen de Regla, the ‘sanctuary
551
Charles Bunker Dahlgren, Historic Mines of Mexico. A Review of the Mines of that Republic for the Past Three
Centuries (New York: Printed for the author, 1883), 184.
552
Fanny Calderón de la Barca, ed. Life in Mexico; the Letters of Fanny Calderón de la Barca, with New Material
from the Author's Private Journals (Garden City, N.Y.: Doubleday, 1966), 240-41.
553
William Parish Robertson, A Visit to Mexico, by the West India Islands, Yucatan and United States, with
Observations and Adventures on the Way vol. II (London: Simpkin, Marshall & Co., 1853), 221.
554
Henry Tudor, Narrative of a Tour in North America Comprising Mexico, the Mines of Real del Monte, the
United States, and the British Colonies : with an Excursion to the Island of Cuba : in a Series of Letters, Written
in the Years 1831-2 (London: J. Duncan, 1834), 309-10.
555
Capt. Lyon visited Regla in 1826 when English investment was rehabilitating its processing infrastructure, G.
F. Lyon, Journal of a Residence and Tour in the Republic of Mexico in the Year 1826 with Some Account of the
Mines of that Country (London: J. Murray, 1828), 153-55. For other glowing descriptions of Regla see Wilson,
Mexico and Its Religion.p. 363.; Thomas Unett Brocklehurst, Mexico To-day: a Country with a Great Future, and
a Glance at the Prehistoric Remains and Antiquities of the Montezumas (London: J. Murray, 1883), 144-45.
556
Two biographies of the first Conde de Regla are: Francisco Canterla y Martín de Tovar, Vida y obra del primer
Conde de Regla (Sevilla: Escuela de estudios hispano-americanos, 1975).; Edith Boorstein Couturier, The Silver
King : the Remarkable Life of the Count of Regla in Colonial Mexico (Albuquerque: University of New Mexico
Press, 2003). By coincidence terreros was the mining term for tailings in Spain.
289
...[that] was located on the last spit of land visible to ships leaving Cadiz towards America’.557
Having placed his destiny in the hands of the Virgin on leaving Spain, he made up for the many
favours received by erecting in her honour a massive temple of stone to refine his ores. Peak
production was reached towards 1770, using ores from the mines at Real del Monte, and also
lead-rich ores fit for smelting from the mines at Zimapán.558
Atotonilco el Grande
Hacienda
Atotonilco el Chico Santa María
~ 30 km de Regla
Real del Monte
Pachuca
to Ciudad de México, ~ 100 km
Figure 4-2. The Hacienda de Regla in relation to Real del Monte, Pachuca and Ciudad de
México. Map adapted from footnote 551.
From Canterla’s data it is possible to calculate that from 1761 to 1768 the annual
average reached 22 t of silver while from 1776 to 1781 it decreased to 12 t/y. In both periods
85% was obtained by amalgamation and 15% by smelting. Couturier’s figures give a yearly
average of 14 t and a peak production in 1770 of 30 t of silver.559 Either set of data place Regla
among the major silver refining haciendas of the whole period (see Chapter 3, Section 3.8). By
557
The Silver King, 20.
558
On the silver-rich lead ores of Zimapan see Miguel Othon Mendizábal, "Los minerales de Pachuca y Real del
Monte en la época colonial," El Trimestre Económico, no. 2 (1941): 263.
559
Canterla y Martín de Tovar, Vida y obra del primer Conde de Regla 41.; Couturier, The Silver King, 157.
290
the time the first English group arrived at Pachuca in the mid-1820s to rehabilitate the mines
and haciendas, this first stage of silver bonanza had long been over.
This functionally unwarranted but visually impressive pile of buildings, surrounded by
the most beautiful of natural settings one can choose for any industrial production site, would
be rehabilitated by the Adventurers in the Mines of Real del Monte as of 1824. By the early
nineteenth century both water and political turmoil had taken their toll on the infrastructure that
had proved so profitable to the first Conde de Regla. The mines had become flooded and the
refining haciendas stripped of contents and in ruin, so that production of silver had plummeted.
The new republic made it possible for foreign investors to enter the mining industry in Mexico,
and the allure of this historically lucrative sector attracted English capital that until then had
been barred from the Spanish colonial empire in the New World.560
The English investment would not be profitable, with revenues of $10.5 million against
expenditures spiralling to $15.4 million up to 1847.561 The new Mexican owners took
possession of the existing assets of the failed English Adventurers for less than 1% of the total
expenditures incurred by the Adventurers.562 They named as General Manager the same
560
Taylor, Selections from Humboldt, i-iv.; H. G. Ward, Mexico, vol. I (London: H. Colburn, 1829), 418.; Henry
English, A General Guide to the Companies Formed for Working Foreign Mines with their Prospectuses, Amount
of Capital, Number of Shares, Names of Directors, &c (London: Boosey & Sons, 1825), 55.; Wilson, Mexico and
Its Religion, 355.; Randall, Real del Monte, 34.
561
Buchan, Report Real del Monte, 41. An anonymous letter signed ‘Visitor’ printed in England and translated in
Mexico placed the blame of the losses on the ‘notorious extravagance … incapacity and madness of the persons
on which trust was placed, and not on any fault of the deposits [of silver]’- ‘a la notoria … extravagancia,
incapacidad y locura de las personas a quienes fueron confiados sus intereses, y no a la falta de mérito y bondad
de los Criaderos en cuestión’. According to the writer the fault lay on the very costly and bad management in
London, bad choice of employees, bad agents sent to Mexico, and even on their lack of Spanish. Viajero, "Las
Minas de Mexico," Revista de Minas Tomo I (1861): 177. All these faults apply equally well to modern
transnationals and the high costs incurred by expatriates who at times rarely speak or integrate themselves into
the local communities.
562
Buchan states in his report that the new capital invested by the year 1854, six years after the Adventurers had
dissolved their company, comes to $538,484, but does not include payments for transfer of rights or liquidation
of debts previously incurred. According to Robertson, the Adventurer’s Company was liquidated for just
$130,000, of which three-quarters were used to liquidate existing debts. Thus the shareholders in London would
receive approximately $30,000 in cash, and the new investors a very significant plant infrastructure in mines and
291
Englishman who as representative of the Adventurers had been responsible for obtaining the
best price for their assets in Mexico, John H. Buchan.563 Buchan’s penchant for accounting set
a precedent within the new Company that has made the present chapter possible:
‘I commenced by arranging the entire system of accounts on such a plan, that every week’s
result, in each mine and reduction-work, might be clearly shown, and the economy of the
different departments thus fairly compared against each other’.564
Thanks to Buchan the Hacienda Santa María de Regla, or simply Regla as it is referred
to in the accounting books, offers unique advantages to a historian reconstructing the
environmental history of historical silver refining. Without a detailed operational mass balance
of the amalgamation and smelting process it is impossible to arrive at a quantitative impact of
haciendas for just $130,000. Buchan, Report Real del Monte, 32,42.; Robertson, A Visit to Mexico, II 172. For a
more detailed breakdown, see Randall, Real del Monte, 200-209.
563
The conflict of interest for Buchan in selling the English assets of the Company he represented at such a low
price to a new Company where he held an evident interest is just one part of a chapter that may be more complex
than a straightforward change of ownership. Velasco Avila et al have been very critical of the business practices
of the new Mexican owners of the Compañia de Real del Monte in the 1850s. The new owners, amongst which
figured Nicanor Beistegui, Alejandro Bellange and Manuel Escandon, were also at the time leasing the Casa de
la Moneda (Minting House) in Ciudad de México and the monopoly on tobacco sales (Estanco del Tabaco). Their
speculative business practices, which included buying but leaving idle mines that did not immediately provide a
bonanza, and their strategy of expanding upstream (agriculture, salt mines, water sources, real estate) so as to
control the prices of consumables in the refining process, together with a constant pressure on the government to
provide them with special exceptions, have led Velasco et al to brand them ‘vampires of the Treasury’. Velasco
Avila et al., Estado y minería en México, 48, 141-143,246- 248. The new owners were very successful even under
conditions that included the French invasion of Mexico. Though they would soon lose the lease to the Casa de la
Moneda in Ciudad de México, which had given them the advantage of faster cash flows on silver remitted, the
Compañia Real del Monte would produce between 30 to 60% of all silver rendered to the Casa de la Moneda in
Ciudad de México. The Company is said to have produced around 8% of all the silver in Mexico between 1849
and 1861. The combination of business practices, government contacts, the bonanza of the Rosario mine and the
decrease in mercury prices (see Chapter 5) made the Compañia Real del Monte one of the leaders of the Mexican
silver refining industry. Ruiz de la Barrera, "La Empresa de Minas del Real del Monte," 3, 14, 37, 73-77, 284-85.;
Elvira Eva Saavedra Silva and María Teresa Sánchez Salazar, "Minería y espacio en el distrito minero Pachuca-
Real del Monte en el siglo XIX.," Investigaciones Geográficas, Boletín del Instituto de Geografía, no. 65 (2008):
91-92.;Velasco Avila et al., Estado y minería en México, 49.;José Alfredo Uribe Salas and Rubén Darío Nuñez
Altamirano, "Depreciación de la plata, políticas públicas y desarrollo empresarial. Las pequeñas y medianas
empresas mineras mexicanas de Pachuca y Real del Monte," Revista de Indias LXXI (2011): 463.
564
Buchan, Report Real del Monte, 43. John Buchan states he was sent to Pachuca in 1848 to wind up the affairs
of the Adventurers. Robertson and Randall associate Buchan with the Adventurer’s Company as of 1825.
Robertson speaks highly of Buchan, explaining he was came from a ‘good Scotch family but an Englishman by
birth and education’. Buchan was accompanied by his wife, ‘a lady of beauty and animation .. lively disposition
... warm heart .. good sense ... cultivated mind’. His brother-in-law Thomas R Auld will appear in managing
positions of the Mexican company in the following sections. Ibid., 39.; Robertson, A Visit to Mexico, II 163-65.;
Randall, Real del Monte, 201.
292
each process on the environment. Regla provides a rich source of accounting data for the period
from mid-1872 to mid-1888 which allows very detailed mass and energy balances to be
established on a weekly and monthly level for most of this period, both for patio amalgamation
and for smelting. The fact that the same Company and hacienda carried out in parallel both
processes in the same location establishes a level playing field when comparing the process
variables of both refining processes. The long time series available for the data even out
spurious correlations and eliminates the risk present when single years are chosen to determine
the long-term mass and energy balance of a process.
Hacienda Santa Maria de Regla

Amalgamation using patio process
N
Smelting of silver ores.
It is at present a hotel open to the public,
but has conserved most of its historical
structure.
El Salto
Hacienda San Antonio
Amalgamation using barrels
Roasting with salt of silver ores
It now lies submerged by a water
reservoir .
Hacienda San Miguel de Regla

Amalgamation using barrels
Roasting with salt of silver ores
It is at present a hotel with man-made
lakes stocked for trout that partly
submerges some of the historical
buildings.
1000 m
Figure 4-3. The locations of three of the historical silver refining haciendas operated by the
Compañia Real del Monte in the second half of the nineteenth century, on the outskirts of
present day Huasca de Ocampo, Hidalgo State, Mexico. Adapted from Google Earth © 2013
DigitalGlobe, reservoir location 20° 13’ 36” N 98° 33’ 46” W.
293
While the neighbouring Hacienda San Antonio de Regla now lies beneath a man-made
reservoir lake with only a chimney stack signalling its presence beneath very murky waters,
and San Miguel has many of its historic buildings knee-deep in a new recreational lake (Figure
4-3), Regla can be visited, measured and reconstructed on paper, thanks to the solid pile of
buildings left by the first Conde de Regla.
4.3 The Hacienda de Regla.
Regla stands at an elevation around 2000m, some 20 km from the present town of
Huasca de Ocampo, in the modern state of Hidalgo. The first refining hacienda to be built in
the basalt-lined gorge was called “El Salto”, and was bought by Terreros in 1753 ‘from the son
of another local miner, Isidro Escorcia’.565 The location was assured of an ample supply of
water from an Ojo de Agua (water spring) that compensated for the transport of silver ores
some 30 km by mule trains over rough tracks from Real del Monte. As amply commented upon
by travellers, it is situated in a barranca (gorge), in whose southernmost point a waterfall (El
Salto) provides the water required by the hacienda. The construction of the modern reservoir
makes it impossible to judge what the original volume of water cascading down El Salto may
have been. Three haciendas were fed by the Ojo de Agua in the second half of the nineteenth
century: San Miguel de Regla, San Antonio and Regla (Figure 4-4). The gorge of Regla pales
in comparison with the cleft in the landscape that nestles the town of San Sebastián.566
565
Couturier, The Silver King, 67.
566
The reproduction of the lithograph of the terrain and haciendas in the area of Pachuca and Real Monte from
which I have reproduced the section in Figure 4-4 was provided in digital form by the MMOB where it is classified
under: ‘Colección Orozco y Berra, Hidalgo, Varilla OYBHG001, Numero Clasificador:1233-OYB-7246-A-002,
Litografía a Color, Perfiles 1 y 2 de Pachuca a Real del Monte (sobre las vetas del Xacal y Vizcaina, y por la
Cañada, del Real del Monte hacia Omitlan y la Hacienda de Regla, Autor desconocido, Ano: 1700, Escala: en
varas, medidas: 24x126 cm’. The year indicated (1700) is wrong for three reasons: there was no Compañia Real
del Monte in 1700, the refining haciendas that appear did not exist, and the artist, Hesiquio Iriarte, was signing
his work as Litografia de Iriarte y Compañia as of the 1850s as reported in Manuel Toussaint, La litografía en
México en el siglo XIX : sesenta y ocho reproducciones en facsímil (México: Estudios Neolitho, 1934), 8. This
lithograph accompanies the text of Burkart, "Memoria Real del Monte."
294
Ojo de Agua San Miguel San Antonio Regla

Town of San Sebastián
Figure 4-4. Section of the lithograph that shows the location of the three refining haciendas
(San Miguel, San Antonio and Regla) with respect to the Ojo de Agua that provided the
guaranteed water supply for their processes. Reproduced and adapted from footnote 566.
Regla had been constructed in the late eighteenth century to refine ores both by patio
amalgamation and by smelting, and this dual functionality would persist under its new English
and then Mexican owners. In fact it would be singled out as the only hacienda of the Company
Real del Monte to refine ores by smelting. It does not surprise then that the Englishman in
charge of Regla in 1851 was the chief smelter of the Company, a Mr. Bell of Durham. 567 The
extravagance of the first Conde de Regla and the relative isolation of the site has made it
possible for most of the Hacienda de Regla to have withstood the erosion of time, in contrast
to the case of most historical haciendas in San Luis Potosi, Zacatecas or Guanajuato. Since no
historical or modern architectural plan could be found of Regla, it has been necessary to
reconstruct the various operational areas within the hacienda. I have used as my initial guide
567
Robertson, A Visit to Mexico, II 223. In 1828 the chief smelter at Regla was of German nationality. Letter from
Charles Tindal to Wllima Dollar, dated 4th October 1828. AHCRMyP, Sección: Correspondencia, Subserie:
Correspondencia General, 12-2: 20 Julio 1827- 6 Julio 1832
295
both the nature of the processes, the extant visual evidence (paintings, watercolours,
engravings, on-site photographs, historical photographs and satellite images), on site
measurements and travellers’ accounts.
Figure 4-5 is a Google Earth satellite image of the grounds of Regla, as they stood in
the year 2004. The outline of the hacienda resembles the shape of a Space Shuttle, and the
distance between the nose cone on the right and the top of the tail on the left is around 300 m.
At its widest point the Hacienda spans approximately 75 m. Figure 4-6 is the general
assignment of main process areas of Regla, which are given in greater detail in the plan of
Figure 4-7. The main operational areas are straightforward to identify when they retain furnaces
or circular milling areas. The spaces identified for storage comply with three conditions:
proximity to the process area where the stored materials will be used, a degree of protection
commensurate with the nature of the material, and the area assigned must suffice to store the
levels of inventory carried at Regla.568
The first impression of Regla on one of the first English members of the group sent by
the Adventurers to visit the haciendas of Regla, San Antonio and San Miguel, in June 1824
was not flattering. ‘The buildings upon them must have cost great sums, but they are now in a
state of decay. They are ill planned, and appear placed at random. The architect, whoever he
was, was a sworn enemy of right lines and angles’.569 This was a most sweeping and dismissive
indictment that came from an observer who was looking at his first industrial site designed
expressly for a joint patio amalgamation / smelting process. The unknown artificer-architect of
568
The detailed calculations are set out in Appendix C.
569
Anonymous, "Journal Descriptive of the Route from New York to Real del Monte by Way of Tampico," The
London Magazine 1826, 167.
296
Figure 4-5. Satellite image of Regla. The lake is the most evident of modern additions. The
photograph is aligned on a South to North axis from left to right. Google Earth © 2013
DigitalGlobe, 20° 14’ 15” N 98° 33’ 42” W.
67 m
Perimeter walls Patio amalgamation Furnace areas Approximate stream bed

area A and B
Wet area:
arrastres, washing Other buildings
Undetermined
Figure 4-6. Assignment of general process areas of Regla. Satellite image from Google
Earth © 2013 DigitalGlobe.
Figure 4-7.
Figure 4.7 Reconstruction of main functional areas in the Hacienda de Regla.
Perimeter walls Main Gate

FA SM
S1
SA1 S2
aqueduct Guest quarters
B B B CBI
Arr CBII
SA2 B1 FB1
B1
W P ST
FB2
67 m
Approximate river bed
N
Amalgamation area Smelting area Other
FB: Furnace Area B Turret structures,
SM: Stamp mill P: Patio (Blast furnaces, FB1 and function unkown
S2: Storage ground ores W: Washing tank cupellation , FB2). Church
Arr : Arrastres CBI: Courtyard I
FA : Furnace Area A, capellinas, casting silver bars CBII: Courtyard II S1 : Raw ore storage B: Buildings
Reconstruction of main functional areas at Regla (see text for details).

SA1 and SA2: Storage and working areas B1: Storage ST: Animal pens, fodder
297
298
Regla had managed to snuggle his massive walls and buildings between the basalt prisms and
the waterway, as close to the waterfall as possible in order to retain the full potential of its
hydraulic power.570 The merits of his scheme will become apparent as I analyze each of the
process areas in detail.
4.3.1 Ore reception and stamp mill
There was only one main entrance to Regla though which entered all the raw ore,
reagents and fuel required for the refining of the silver ores. Raw ore, both for amalgamation
and for smelting, would have been stored close to the main gate (S1: all abbreviations refer to
Figure 4.7). This would isolate the transit corridor for 600 mules (or the requisite number of
carts) periodically delivering raw ore, from the internal workings of the hacienda. From this
common storage area the flow of materials would separate into two distinct transit and work
corridors. Materials and silver ores for amalgamation would have been transported through the
arches straight ahead from the Main Gate to the amalgamation work areas, including Furnace
Area A. Materials and ores for smelting from the mid 1830s would have been sent to be
processed at Furnace Area B. The painting by Johan Moritz Rugendas shown in Figure 4-8
gives the artist’s impression of the scene on entering by the Main Gate in the year 1832. The
tall building observed on the right with the prominent corridor houses the accommodation
quarters for guests. The loaded mules are being taken though the massive arches to the
amalgamation areas in the southern part of the Hacienda.571 Blueish smoke eddies from the
hidden chimney stacks of Furnace Area A still in use in the background of the scene.
570
According to Couturier ‘contemporaries asserted that Pedro Terreros himself had planned this great hacienda’
but provides no primary source. Couturier, The Silver King, 69.
571
The use of the word trapiche for the title of the painting is an aid to identifying the work spaces. ‘Entrada a
los trapiches’ in the plural signals the entrance to an area of mills, in the plural. The trapiche in a sugar refining
hacienda is the place where the sugar cane is crushed Another painting by Rugendas titled ‘Trapiche del ingenio
de Tuzamapa’ shows the grinding of sugar cane in a crusher powered by the circular motion of horses. Renate
299
With regards to the mills, in 1771, due to the workers strike, it is claimed that ‘the five
magnificent waterwheels at Regla had nothing to grind’.572 An inventory from 1824 states:
‘Grinding establishment.- One wheel in working order. Five stamp heads’.573 One year later
the British Chargé d’Affaires in Mexico, H.G. Ward, would observe that:
‘the whole [Regla] is in a tolerable state of repair, with the exception of the stampers, for
braying the ore, which are now in ruins. These are to be replaced by a water-wheel constructed
by the Company, which is thirty–six feet in diameter, and is to put in motion forty-eight
stamps’.574
Figure 4-8. ‘Entrada a los trapiches de la hacienda Santa Maria de Regla’, by Johan Moritz
Rugendas, oil on cardboard, 24 August 1832. Image reproduced with permission from the web
portal of the Universidad de las Artes, Aguascalientes, Mexico:
http://www.aguascalientes.gob.mx/ temas/cultura/ webua/catalogo/ johanmoritz.html
(accessed 9 May 2013).
Moyssén Löschner, Xavier Echeverría, and Mario de la Torre, El México luminoso de Rugendas (Mexico City:
Cartón y Papel de México, 1985), 49. Trapiche has also been used to refer to the molino. Egleston, The Metallurgy
of Silver, 269.
572
Doris M. Ladd, The Making of a Strike : Mexican Silver Workers' Struggles in Real del Monte, 1766-1775
(Lincoln: University of Nebraska Press, 1988), 147.
573
574
H. G. Ward, Mexico vol. II (London: H. Colburn, 1829), 140.
300
By 1855 the stamping equipment has again changed. Buchan reports that the new
Mexican company has installed two water-wheels (of unknown dimension) providing power
to thirty stamps (of unknown size).575 The logistics of an efficient amalgamation process
requires that the stamping mill should be located as close as possible to the arrastres (area
marked Arr), powered by water supplied by the internal system of water channels. There is at
present a rectangular structure in the area designated SM that would have accommodated one
large vertical waterwheel. The proximity of the proposed stamping area SM to the living
quarters may explain the following extracts: ‘a poor woman [the wife of the English manager
at Regla], living all alone ... with no other sound in her ears from morning till night but the roar
of thunder or the clang of machinery’.576
Or as expressed more explicitly:
‘The works are on the patio system ... the noisy stamping-mills working night and day. To me
they had a peculiarly musical charm. It happened that in the middle of the night two or three of
the nearest stamping-mills stopped on account of something being wrong with the water-
wheels, and Bishop and I, who slept in adjoining rooms, both awoke and called to each other,
wondering what had happened’.577
No mention is made of clouds of dust emanating from the stamping operation, which
may imply that water was added during stamping in the 1880s.578
4.3.2 Arrastres and water power in the Hacienda
The circular vats in Figure 4-9 correspond to where the arrastres used to be located at
Regla, a total of sixteen units arrayed in two sections of eight each. Another two circular vats
575
Buchan, Report Real del Monte, 13.
576
Calderón de la Barca, Life in Mexico, 243.
577
Brocklehurst, Mexico To-day, 145.
578
301
are observed in the topmost terrace which correspond to the traces of two molinos. The
arrastres were powered by water-wheels, possibly two arrastres to one water-wheel. The
diameter of the well of each arrastre at present is approximately 3 m.
Figure 4-9. View of the sixteen circular vats of the arrastres. The remains of two molino
tracks are hidden by the cactus leaves on the lower right-hand corner.
Figure 4-10 is a photograph of a water-powered arrastre from an unnamed hacienda in
Mexico in the nineteenth century.579 Fanny de Calderon may have been referring to all these
water-wheels in her description of the Hacienda (see quote at the beginning of the chapter).
Ward mentions that ‘eight of the old arrastres, (worked by water) had been repaired’ and also
states that in the mid-1820s there were ‘twenty-four Arrastres, worked by horizontal water-
wheels’.580 That only eight required repair after a decade of neglect prior to the 1820s implies
579
Rickard, Journeys of Observation, 242.
580
Ward, Mexico, I 424.;Mexico, II 140. Humboldt makes the unexpected commentary that arrastres were
unknown at Regla, where ores were ground by stamp mills only. ‘Dans quelques grandes usines de la Nouvelle-
Espagne, par exemple a Regla, on ne connoit point encore les arastras’. Humboldt, Essai politique, Tome IV,
59. Castelazo’s account of the mining around Real del Monte, first published in 1820, states that the second Conde
302
a robustness of structure for these wet grinding units. The upper terrace in Figure 4-9 which at
present holds two rectangular tanks may have originally contained another eight circular vats.
Figure 4-10. Photograph of water-driven arrastre, from unnamed hacienda in Mexico,

photograph reproduced from footnote 579.
According to Randall, an inventory from 1849 includes ‘forty water-powered arrastres
capable of reducing 120 tons per week’.581 The grinding capacity is consistent with the
de Regla built masonry arrastres of Regla at a cost of more than 100,000 pesos, but does not provide a date. Mining
and refining in the area would progressively halt after the death of the second Conde in 1809. ‘el Segundo Conde
... en atención a la necesidad que tenia de habilitar las haciendas para beneficiar los metales que produjeran,
juzgando suficiente las de S. Antonio, S. Miguel y Regla, determinó repararlas … construyendo en ellas nuevas y
costosas oficinas, particularmente en la primera y última, cuyos arrastres de solida mampostería le tuvieron de
costo mas de cien mil pesos’. Jose Rodrigo Castelazo, Manifiesto de las riquezas que han producido y actualmente
contienen las celebradas minas de las vetas Vizcaina y Santa Brigida ubicadas en el Real del Monte, jurisdicción
de Pachuca, de las grandes obras que en ellas se hicieron y del estado en que actualmente se hallan (Mexico:
Imprenta de D. Mariano Ontiveros, 1823), 15. The third Conde visited Regla, whom he referred to as ‘Babylonia’,
the 31st January 1810. According to his journal for that day, ‘[it] is the only one still functioning, even if partially,
since only three [stamp] mills are working, eight smelting furnaces, twelve of the twenty four [ar]rastras that it
possesses’ ‘[es] la única que esta en ejercicio, aunque incompleto, porque solo andan tres morteros, ocho hornos
de fundición, doce rastras de las veinte y cuatro que tiene’. Manuel Romero de Terreros, "El condado de Regla
en 1810," Historia Mexicana 4, no. 1 (1954): 110. If Humboldt was not mistaken this would confirm that arrastres
are not installed in many haciendas of the late eighteenth century.
581
303
amalgamation activity at Regla, but it is not evident where the additional arrastres according
to Randall would have been located within the compound. By 1855 there are only 16 arrastres
included in the description by Buchan of the infrastructure at Regla, plus one molino, which
corresponds well to the modern image in Figure 4-8.
The guarantee of a constant large flow of water was necessary to provide power, both
to the stamp mill and to the arrastres. The main aqueduct enters the Hacienda on the south
side (Figure 4-11), fed by water retained in a small pool at the bottom of the waterfall (Figure
4-12).
Water-wheel / stamping mill
aqueduct
overflow
Figure 4-11. Water distribution channels and overflow outlet in the southern area of the
Hacienda. The four pillars on the grassy area may be the remnants of the structure that
sustained a roof over the patio area.
304
The presence of this reservoir would have cushioned the Hacienda from an irregular
flow from the spring. The opening in the perimeter wall pointed out in Figure 4-11 acts as an
overflow outlet to regulate the total volume of water supplied to the various process areas of
the Hacienda. Large quantities of water would also be required to prepare the amalgamation
slurry, to replenish the water lost by evaporation from the tortas, to wash away the mineral
matrix of the slurry and separate the amalgam and for the consumption of the workers and
animals (from 100 to 300 mules and horses kept within the Hacienda towards the latter part of
the nineteenth century, see below).
Figure 4-12. The pool at the foot of ‘El Salto’.

305
4.3.3 The patio reactor.
The ore, now milled to a very fine powder in the arrastres, could have been transported
in barrels or discharged directly from the arrastres to the patio reactor.582 The patio reactor at
Regla went through two quite distinct stages. Initially the inventory from 1824 describes a
‘Patio for quicksilver reduction, with forty-five masonry arches to support roofs, now in ruin.
Wooden floor partly serviceable, partly in ruin. Pillars supporting arches partially worn away
at base, due to salt’.583 The English Company opted to retain the roof, since by 1829 Ward
mentions ‘two covered Patios, each about 200 feet in length, in which the process of
amalgamation is carried on’.584 Figure 4.11 shows four pillars still standing at present, which
may correspond to the type of column mentioned in the 1824 inventory. However, neither of
the paintings shown below include these pillars. The practice of using wooden planks to floor
the patio area has been mentioned in Chapter 3. By 1835, according to Randall, Commissioner
John Rule decided to adopt the “Guanajuato” method, and decided to keep the patio area
uncovered and unheated except by the sun.585 By 1855 Regla has an open patio ‘all carefully
floored in beams’, and the area available for amalgamation is claimed to cover one and a half
acres (approx. 6,000 m2).586 The impact of rain on a patio reactor without a roof is well
illustrated in a letter from R. Bell at Regla to Mr. Auld, General Manager of the Real del Monte
Company, dated September 30th 1865, reporting that:
582
It is claimed that transport in barrels was more efficient than discharging each arrastre through an outlet, with
two workers discharging via shovelling into barrels the contents of 4 arrastres in half an hour. Hermosa, Manual
de Laboreo de Minas, 207-208.
583
584
Ward, Mexico, II 140.
585
A technical report drawn up for the Directors of the Company in London refers to the amalgamation process
carried out in Guanajuato as reaching “the highest state of perfection”. Randall, Real del Monte, 115-16.
586
Buchan, Report Real del Monte, 9. The patio reactor area in Figure 4.7 corresponds to over 5,000 m2.
306
‘the weather is dreadfull here, it began raining on Wednesday night & it has rained incessantly
ever since, our Patio is a complete lake, we cannot work it. The salt and sulphate must be all
washed out by this time, & I have not a grain of salt about the place to replace the loss’.587
It is this new open patio area that is prominent in two paintings of the Hacienda de
Regla. The collection at the Soumaya Museum in Mexico City houses a painting by Eugenio
Landesia that depicts the Hacienda in almost photographic detail (Figure 4-13). It was
commissioned to be painted from the location of the arrastres when Thomas R. Auld was
managing Regla in 1854.588 It is a view to the north of the hacienda, showing some fifteen
circular tortas being trod upon by teams of horses and men. ‘[the mixture] is ... pounded and
trampled upon, by both bipeds and quadrupeds ... men and mules are seen, in most singular
combination, performing this dirty but essential operation’.589 Some of the tortas are depicted
as being fenced in using provisional planks to the sides. What can only be appreciated in the
original painting are the small mounds of copper sulphate with their tell-tale intense blue
colour, waiting to be added to each torta.
There is also an anonymous, unidentified and undated painting reproduced in black and
white, in the textbook on the history of Mexico by Meyer, Sherman and Deeds.590 It is again a
view facing north, but takes in the whole Hacienda, having been painted presumably from the
heights of the barranca. The uncovered patio area places the date after the mid-1830s. As can
be seen in Figure 4-14 the arrastres appear covered with small roofs and some ten tortas are
587
AHCRMyP, Sección: Correspondencia, not classified, 1865. Auld was the family name of the wife of John
Buchan, the first General Manager under the new Mexican-owned company established in 1848.
588
Alfonso E. Pérez Sánchez, Benito Navarrete Prieto, and Gustavo Curiel, Tesoros del Museo Soumaya de
México : siglos XV-XIX (Madrid: BBVA, 2004), 213-15.
589
Tudor, Narrative of a Tour, 298.
590
Meyer, Mexican History, 150.
307
Figure 4-13. ‘Patio de la Hacienda de Regla’ by Eugenio Landesia, 1857, oil on canvas, 46
by 64 cm. Image reproduced from footnote 588.
Figure 4-14. The Hacienda de Regla, from an unidentified illustration in footnote 590.
308
being worked on the patio. The only smoke seen in the painting is coming from Furnace Area
A, in the most southern area of the Hacienda. Should any doubt remain as to the identity of
this Hacienda, the facade of the Church and the rendering of the basaltic prisms in the cliffs of
the background complete its identification. The width of the stream indicates a significant
volume of water at the time of the painting.
4.3.4 Furnace Area A: Capellinas and the recycling of mercury
Furnace Area A is currently designated as the site of the ‘Spanish Furnaces’, where the
original smelting furnaces built in the eighteenth century were located. Though the English
investors built the new blast furnaces to the north of the hacienda (Furnace Area B), Furnace
Area A remained the location where approximately 85% of all the silver of Regla was finally
separated from the amalgam. The initial separation of the mercury amalgam from the ore slurry
in the washing tanks (WT) was a milestone in the process. For the first time a product with a
very high tangible monetary value per unit weight has now been isolated into a relatively small
mass of mercury and silver. From this moment on, each stage of the process would require an
increasing amount of vigilance and control.591 Away from the open patio area there are three
process areas I have designated in Figure 4-15 as SA1, SA2 and FA (Furnace Area A). SA1 I
have assigned to a covered work and bulk storage area for two water-soluble high bulk
additives, salt and copper sulphate. SA2 I assign to a secure storage area for premium materials
such as the amalgam, mercury and intermediate storage of spongy silver from the capellinas.
591
‘in the [washing] vat the men handle the pella [amalgam] just a short time, at the most two hours; during this
time the guards are overseeing [them], and yet many thefts are detected … those who know of the tendency to
steal by the workers in mines and haciendas, will be able to appreciate this reasoning’ - ‘en la tina los hombres
manejan la pella un corto rato, que quando mucho llega a dos horas; en este tiempo están los Mandones de
guardas de vista, y sin embargo se averiguan muchos robos … quienes tengan conocimiento de lo propensas que
son al robo las gentes operarias de minas y haciendas, sabrá graduar el peso de esta razón’ in Garcés y Eguía,
Nueva teórica del beneficio de plata, 137. The overall level of tension between the workforce in the mines and
haciendas, their overseers and the first Conde de Regla is the subject of Ladd, The Making of a Strike.
309
It would also include the azoguería, the mercury room with manga and scales such as depicted
in Chapter 3.
From the 1830s, in Furnace Area A at least two heating operations took place: the
separation of mercury from silver in the capellina, and the casting of silver bars. Buchan
includes one distilling furnace (capellina) and one bar casting furnace in his description of the
amalgamation infrastructure at Regla in the year 1855.592 I have tentatively assigned the
circular bed observed in Figure 4.16 as housing the lower part of a capellina, with its channel
for cooling water that runs to the back wall. It was not possible to measure directly the diameter
of the bed since the area is at present off-limits to the public.
SA1 : covered working and SA2 : secure working and

storage areas for salt and copper storage areas for mercury,
sulphate amalgam and silver
FA : furnace area A
Figure 4-15. Proposed assignment of dry process areas related to the amalgamation process
in the southern part of Regla.
592
310
Figure 4-16. Tentative location for a capellina, showing the channel for cooling water. This
picture was taken in March 2013. Further excavation around this location has been carried out
by November 2014.
I estimate a diameter well over 1 m. The external diameter of the capellinas used at
Regla in the second half of the nineteenth century was 0.61 m, according to engineering
drawings of the Compañia Real del Monte from this period (Figure 4-17).593 This dimension
would be sufficient to fit the rounded base of an assembled capellina and leave space for the
fuel used for heating. Once assembled, the capellina loaded with stacked amalgam pellas inside
593
At the beginning of the nineteenth century the most common copper capellinas measured approximately 40
cm (half a vara) in width and up to 80 cm (one vara) in height, according to Sonneschmidt and de Fagoaga,
Tratado de la amalgamación de Nueva España, 48.
311
b
Figure 4-17. a) Drawing with the engineering specifications for the construction of iron
capellinas in the workshop of the Compañia de Real del Monte (date unknown). The left-hand
capellina has a height of ‘5 pies 6 pulgadas ynglesas’, 5 feet 6 english inches, an internal
diameter of 21 inches and an external diameter of 24 inches. The capellina on the right has the
same width but a height of 4 feet and 6 ½ inches. Digital image of undated drawing courtesy
of AHCRMyP, Sección: Administración Interna, Serie: Departamento de Ingenieros, Subserie:
Croquis y Planos, Vol 204 Carpeta 1. b) Photo of capellina in the entrance to the Hacienda
Santa María de Regla, photo courtesy of Mr. Josue Soto Samperio, November 2014. Original
height impossible to ascertain since capellina has been embedded, but at least over 1 m.
312
would have weighed approximately two tons. I arrive at this figure because just the weight of
the iron hood, based on the smaller model in Figure 4-17 a, would be approximately 700kg.594
As to the contents, I have used as a guide the fact that in the four weeks ending May 26th 1878,
a total of 11 capellina runs were carried out and 66 silver bars cast, equivalent to approximately
7 bars cast from the silver refined from each capellina run.595 This means that at least 210 kg
of silver and 840 kg of mercury (both contained in the amalgam) were loaded into each
capellina, making a total weight approaching two tons.596
Since the capellina cycle took approximately one day, a single capellina installation
would have been enough, as also observed in the plan of the hacienda de Rocha in Chapter 3.
From an architectural point of view the method required sufficient overhead space to allow the
capellinas to be lifted by pulley and manoeuvred into place, and the channelling of a guaranteed
flow of cold water at floor height. The water channels that supplied the constant stream of cold
water required to condense the mercury vapours during the heating of the amalgam can also be
clearly identified running the length of the roof of this area in Figure 4-18.
The placing of religious inscriptions over the original placing of these furnaces conform
to their deemed Spanish roots: Las ordenes de Christo, The Orders of Christ (Figure 4-19 a)
and Nuestra Señora de los Dolores, Our Lady of the Sorrows (Figure 4-19 b). At least one
other placement for a furnace can be observed at the end of the corridor of Furnace Area A
594
Based on estimating the volume of iron used for the hood (diameter, length and thickness as given in Figure
4.16) and using the density of iron to calculate the weight.
595
AHMP, Memorias No. 19 - 21 de los Gastos de la Hcda de Regla.
596
A 4:1 mercury to silver weight ratio in the amalgam after straining in the manga is indicated in Hermosa,
Manual de Laboreo de Minas, 236. My calculation for Regla coincides with the following generic description:
‘bricks of amalgam ... were pressed together ... one ton of these was piled on iron supports, over a stone tank
filled with water to nearly the top of a copper or iron bell, capellina, which is 0.90 m high and 0.45 m diameter ...
the yield of silver was about 200 kilos., and the charcoal used about 250 kilos. per charge.’ Egleston, The
Metallurgy of Silver, 304. The capellina run of 1761 cited in Chapter 3, Section 3.8, corresponded to a silver
content of around 60 kg, which confirms the increase in the size of the capellinas from the eighteenth to the
nineteenth century.
313
(Figure 4-19 c), to the right of which lie a succession of vaulted chambers which I have assigned
tentatively to the storage of salt and copper sulphate, both of which needed protection from the
rain. It is significant that the anonymous painting of Regla (Figure 4-14) shows at least four
heavily smoking chimneys in this area, three with pyramidal stacks and one with a rectangular
stack. At present one pyramidal and one rectangular stack can still be observed (Figure 4-20).
Other chimney flues can be found in this area, where reverberatory furnaces used to cast silver
bars may have been located:597
‘The retort silver [from the capellina], plata pasta, is refined in a small reverberatory furnace
built of adobes, and heated with wood, which receives a charge of 300 kilos. of the crude
bullion. This charge is refined for four hours. A little litharge and lead are added to remove the
impurities’.598
Figure 4-18. Stone water channels (just to the right of the pyramidal chimney stack) on the
roof of the Furnace Area A corridor, that distribute water supplied by the external aqueduct for
the condensation of mercury. The shrubs in the foreground hide the connection to the furnace
topped by the rectangular chimney stack.
597
‘reverberatory furnaces to cast silver into bars’-‘hornos de reverbero para fundir plata en barras’. Hermosa,
Manual de Laboreo de Minas, 250.
598
314
a b
passage to
possible
capellina bay
Figure 4-19. a) Ordenes de Christo b) Nuestra Señora de los Dolores, inscriptions over
arches where hornos castellanos are deemed to have been located to the left of c) Furnace Area
A corridor, to the right of which d) the vaults of the proposed storage areas for salt and copper
sulphate. Photos 4-19 a) and b) courtesy Mr. Josue Soto Samperio.
315
Figure 4-20. Extant chimney stacks, date unknown, but similar in location and shape to two
of the five depicted in Figure 4-14.
I have also found engineering drawings for kettles used at Regla and designated as
‘ollas para dissolver la plata’, kettles to dissolve silver, dated the year 1880. These would have
been used to melt the spongy silver and then to ladle it into moulds.
I have not found a detailed description of areas FA, SA1 and SA2 from a visitor to
Regla. In particular, area SA2 would have stored or manipulated much of the wealth generated
at Regla, in the shape of mercury, silver amalgams and silver from the capellinas, so it may
have been off-limits in any case to the visits by the tourists of the nineteenth century.
4.3.5 Furnace Area B
Smelting of lead-rich silver ores from Zimapán had been carried out since the 1770s at
Regla, most probably using hornos castellanos in the location described in the previous section,
Furnace Area A. The third Conde de Regla stated in the journal entry for his visit in 1810 that
at the time there were eight smelting furnaces still in place, though unfortunately he does not
316
state where.599 This is important to point out, to avoid the impression that smelting was
introduced at Regla by the English investors in the 1820s. By the time they arrived in the area
in 1824 ‘all the machinery in the large reduction works, formerly employed for extracting silver
from its ores, was gone’.600 It is also true that the Europeans came with smelting firmly in their
plans. By 1826, when Ward visited Regla, he was able to report ‘a number of furnaces for
smelting’ already functioning, though it is a pity he was not more specific on the technicalities
of the same.601 In the notes written in 1824 by the person who would be the General Manager
of the Adventurers in the Mines of Real del Monte until its dissolution, John Taylor, he lays
out his ideas regarding the furnaces to be used in Mexico for smelting: ‘by a better application
of fuel either in good reverberatory furnaces, or in the blast hearths used in the north of
England, the silver ores may be smelted with a certain quantity of lead’.602 I have not been able
to establish the exact year of installation of the blast furnaces in Furnace Area B. Events on the
ground soon changed the plans of the Company, for which I quote at length from Randall:
‘Once it was clear ... they could not make a rapid and dramatic change from amalgamation to
smelting as the sole method for ore reduction at Real del Monte ... [the company decided to]
operate patio amalgamation plants at both the Regla and Sanchez haciendas ... [and] pressed
ahead with efforts to construct a superior smelting establishment at the Regla mill ... John
Taylor ... reporting to the June 1831 stockholder’s meeting : “I believe that this important
branch of economy is more advanced [at the Regla mill] than in any other part of Mexico” ...
accordingly, blast machinery, intended to put more smelting furnaces into operation at the
Regla mill, were purchased in England and dispatched [to Mexico in 1833, 1834]’.603
599
Terreros, "El condado de Regla en 1810," 110.
600
601
Ward, Mexico, II 140.
602
Taylor, Selections from Humboldt, xiv.
603
Randall, Real del Monte, 111-12. The text quoted from Randall correlates well with a letter dated 21 st March
1834 where R. Mackenzie at the headquarters of the company in Real del Monte, after consultation with John
Rule, disregards the doubts raised by A. Mackintosh at Regla on the wisdom of extending the smelting facilities
at Regla, since equipment had already been purchased and labour force assigned. AHCRMyP, Sección:
Correspondencia, Serie: Compañía a Varios, Subserie: Correspondencia General, 33-11: 20 Marzo 1834 – 20
Abril 1835.
317
Just one year later, in 1835, Commissioner John Rule would begin the change in
strategy from smelting to amalgamation, arguing for a rehabilitation of the San Antonio
hacienda in order to achieve sufficient refining capacity, or “hacienda power”. ‘Smelting
would not again be considered the technical key to success for the Real del Monte Company’.604
According to the 1849 inventory, at the time Regla was returning to Mexican
ownership, the contents included a ‘foundry with ten furnaces, iron-cylinder blast powered by
water; capable of smelting from 27 to 30 t per week ... [and a] four-cylinder engine to provide
blast for furnaces’.605 It would seem that both water and steam engines had been used to provide
the blast for the furnaces. By 1851 Regla was running its own Mexican version of the dark
satanic mills of the English Industrial Revolution:
‘Mr. Bell conducted us to the great smelting works of Regla, which were at the time in actual
operation. The vault or cavity into which a row of huge furnaces disgorged their contents, was
about two hundred feet in length, and eighty to a hundred in width, smoke-dried, black and
rough. The scoria from the ores came out of the furnaces, in soft, ductile cakes; and as they
gradually cooled, they were thrown on heaps of now vitrified masses, misshapen and strange
in their conformations. To look into the heated furnaces, would have been something terrific
to such as had not visited our own manufacturing districts, more particularly those of
Staffordshire.; and, adding the blackened visages of the workmen, with their long iron
implements for handling the glowing materials of their Vulcan-like home; the scene was one
of almost satanic grandeur’.606
In the 1855 report by Buchan, he makes reference to ‘the smelting works of Regla, with
a powerful cylinder blast, and eight high furnaces’. In a table he adds to these eight furnaces
another two refining furnaces in the list of infrastructure for smelting.607 At present only six
604
Ibid., 112-14.
605
Ibid., 219.
606
Robertson, A Visit to Mexico, II 224-225. The dimensions correspond to the areas designated as FB and CBII
in Fig. 4.7.
607
Buchan, Report Real del Monte, 9, 13.
SMELTING FURNACES (B1 to B6) open courtyards bounded by very
high walls (6 to 9 m height)
Stairs to roof
Tunnel passage
L= 9m (laser)
height vault 6.5 m (laser)
AIR
B5 VENT
Courtyard B-I w1= 11m (laser)

B6
B4
Entry to inner furnace
h vault 4 m
h= 6m (laser)
(laser)
≈
B3 height vault 6.5 m

L= 34m (laser)
AIR
VENT
B2
w2 = 15m (laser)
h vault 4 m
Courtyard B-II
Area for water-

driven bellows
Stairs to inner chambers
height vault 6.5 m
Figure 4-21. Plan of Furnace Area B, floor level including courtyards.

B1 h= 9m h= 7.5 m
(photo) (photo)
Tunnel passage
≈
approx. 18 m (laser)
318
≈
Figure 4-22.
stairs
Chimney vent Air vent
(closed)
Chimney vent
L= 11m
(closed) Chimney vent
(closed) There are other air
circulation openings
giving on to the roof
(open arcs) not shown
Chimney stack, in this diagram
≈
base built < 1857

Air vent
Chimney stack The roof has different

base built in 1854 levels, the chimneys
are all at the lowest L= 15m
level.
Chimney vent closed at present, but

chimney stack base existed from 1850s to
at least 1962.
≈
≈
approx. 18 m
Plan of loading floor of Furnace Area B, at present without a roof.

319
320
remains of high furnaces are found in the area I have designated as Furnace Area B.608 At
present this area is designated by the term ‘hornos ingleses’, English Furnaces, but the
documents detailing the reasons for installing the new blast furnaces at this location in the
hacienda have not yet come to light. Figure 4-21 and 4-22 are a plan of Furnace Area B, floor
level and loading floor level, as measured by the author on-site.609 It comprises a lower level
vaulted area where the furnaces, cylinder air blasting equipment and storage rooms would have
been located. In contrast to the other five, furnace B6 lies within a closed internal chamber, its
dimensions as yet undetermined.610 The other five furnaces (B1 to B5) give onto walled-in
open courtyards, and are all covered by a roof that served as an upper-level loading area as well
as providing an outlet for air vents and skylights. Since the present roof was the loading floor
for the charge on each furnace, it is very probable there was originally a provisional roof
structure over the load floor to provide a round-the-clock working environment.
The spacing of the furnaces complies with the observation by Iles that blast furnaces
should be set not less than 1 m apart to avoid crowding the operations around them.611 The five
accessible furnaces present extensive damage to the lower portion of their structure, most
probably due to the extraction of all components made from iron once Regla was taken out of
commission in the early twentieth century.612 Furnaces B1 to B3 were fitted under the new
608
The possible location of the two other blast furnaces mentioned by Buchan is unknown.
609
Measurements were carried out wherever possible using a Bosch Laser GLR225. Where sunlight made this
impossible, photographs were taken and a notebook used as scale. If none of these options was available,
approximate measurements have been used either using Google Earth images or by visual evaluation of relative
heights in photographs.
610
The inner chamber where furnace B6 is located is unlit and no measurements could be taken. A photograph
taken with the aid of a flash shows a furnace hearth very different from the other five.
611
Malvern Wells Iles, Lead-Smelting The Construction, Equipment, and Operation of Lead Blast-Furnaces, and
Observations on the Influence of Metallic Elements on Slags and the Scientific Handling of Smoke (New York;
London: J. Wiley & Sons; Chapman & Hall, 1904), 80.
612
According to the description by Iles of nineteenth century blast furnaces, iron plates or jackets covered a major
portion of the lower surfaces of blast furnaces, in the form of water jackets that reached up to nearly 2 m in height,
crucible plates, iron floor plates around the outside of the furnaces and cast iron binding rods. Ibid., 39, 54-55, 60,
80. Residual silver may also have been sought in the materials extracted from the furnaces.
321
Mexican ownership with a pyramidal chimney stack and loading aperture. These brick chimney
stacks for furnaces B2 and B3 still stand at present, though it is unknown if their height was
adjusted during the second half of the nineteenth century (Figure 4-22). I can state that furnace
B1 had a similar configuration as B2 and B3 because the visual evidence supports the presence
of a third stack as early as mid 1850s and still existing in 1962.613 From Landesia’s painting it
is clear that the Mexican company did not install similar stacks for furnaces B4, B5 and B6.614
This may reflect a decision not to invest in such structures if the level of smelting output after
the 1850s did not justify it. I have seen no evidence for flue-ducts to capture the lead fumes
coming from the smelters, as was already the custom in England since the eighteenth century,
or of bag-houses as were installed in similar smelting facilities in the U.S. by late nineteenth
century.615 Finally, close to Furnace Area B there should be an area reserved for the waterwheel
that drove the blast iron cylinders, but this space I have not been able to identify.
The second major furnace operation carried out at Regla was the cupelling of the silver-
rich lead obtained from the blast furnaces. The observations from a visitor to Regla in the 1850s
confirms the use of a reverberatory furnace for this stage:
‘The mass of molten lead and silver is drawn off, and placed in a large oven with a rotary
bottom, into which tongues of flame are continually driven until the lead in the compound has
become once more oxydized, forming litharge, and the silver is left in a pure state. A little
beyond the furnace is a series of tubs, built of blocks from broken columns of basalt [the
arrastres]’.616
613
A photograph of Regla taken in 1962 shows the third stack, Calderón de la Barca, Life in Mexico, 578.
614
This assumes the date 1854 refers to a completely new structure, and not the rebuilding of stacks built by the
English.
615
As discussed in Chapter 2, the presence of lead in the fumes was well known in England in the nineteenth
century, and flue systems attached to lead smelting furnaces existed in Derbyshire as of the late eighteenth century,
as reported in Willies, "Derbyshire Lead Smelting," 13-14.
616
Wilson, Mexico and Its Religion, 366.
322
c d
Figure 4-23. a) Blast furnaces B2 and B3, h = 4m and w = 6m; distance from hearth front to
front furnace wall: ~2 m, distance to exposed back furnace wall: ~2.7 m. b) trough in front of
furnaces, possibly for water. The photo was taken in March 2013. By November 2014 this
trough had been demolished as part of alterations underway in this area by the hotel c) furnace
stacks, with opening to charge the furnace; the left stack is taller, notebook in lower corner
served as scale for photo measurements, strong sunlight precluded laser measurements d) right
hand stack was built in 1854.
323
This would be the cupelling furnace for afinación, distinct from the blast furnaces,
mentioned by Buchan in his listing of 1855. The statement that ‘a little beyond the furnace’ lie
the arrastres can be interpreted in two ways: either the cupelling furnace for the silver-rich
smelted lead was situated in Furnace Area A, or it was placed within the building with a
smoking chimney to the east of the Church as shown in Landesia’s painting (Figure 4-24). It
communicates with the area of the blast furnaces through indoor passages and is isolated from
the open patio area by a high wall. The high lintels would indicate either increased air
circulation or the manipulation of equipment requiring a high overhead clearance. Finally, the
fact that ‘poor’ lead is mentioned as being sold at Regla during the period when smelting was
carried out raises the question as to how, and where, this silver-poor lead was produced. The
Pattinson process had been introduced in England in the 1830s as a means to enrich the lead
from the smelter prior to cupellation, while at the same time producing a ‘poor’ lead that could
be sold.617 If the Pattinson process was not used, then silver-poor litharge had to be reduced in
a reverberatory furnace. The English historian David Cranstone makes the distinction in the
refining of silver in England between ‘a cupellation furnace [and] the more sophisticated
processes such as Pattinson’s or Parkes’, which would surely be used in any new mid 19 th
century silver refinery’.618 The exact location for either of these two processes within Regla
remains to be established.
To the north of the row of blast furnaces of Furnace Area B were located two
courtyards, one small and enclosed, B-I, the second one much larger and on two levels, B-II.
Both were surrounded by walls that measured up to 9 m in height (see Figure 4-25). The height
617
For details on the process, see Percy, Metallurgy of Lead, 121-148.
618
David Cranstone, "Excavations at Old Gang Smeltmills: an Interim Report," in Boles and Smeltmills: Report
of a seminar on the History and Archaeology of Lead Smelting, ed. Lynn Willies and David Cranstone(Reeth,
Yorkshire: Historical Metallurgical Society, 1992), 29. Since none of the account books mention the purchase of
zinc, I have ruled out the option that Parke’s process was used at Regla.
324
Figure 4-24. a) Detail of Landesia’s painting of the Hacienda de Regla in 1857, showing the
presence of three stacks in Furnace Area B. The single chimney in the foreground is assigned
to reverberatory furnaces b) probable location of same chimney stack in modern Regla.
of these walls means that fugitive emissions of lead in this area would tend to be trapped in
these courtyards. The apparent defensive nature of these walls may reflect that the ore with the
highest silver content was kept in this area. Between courtyard B-II and the outer perimeter
wall lies an area that I have very tentatively assigned to animal pens or to store fodder and other
general materials in use at Regla (see Figure 4-7). By November 2014 the space corresponding
to Courtyard B-II had been transformed into modern bathing pools and toilets, and the
325
crenelated wall measuring 6 m in height shown in Figure 4-25 had been half demolished, in a
process of further adaptation of this historic hacienda into a hotel.
7.5 m
9m
6m
Figure 4-25. a) Courtyard B-I. This photo was taken in March 2013. By November 2014 this
crenelated wall no longer exists, having been half demolished in alterations being carried out
by the hotel b) Original view in March 2013 of Courtyard B-II. By November 2014 the area
has been converted to a set of pools and toilets c) View of courtyard B-II from vaulted furnace
area. Photo taken in March 2013. In November 2014 this arch was being filled in using stones
from the demolished interior walls
326
4.3.6 Final comments on the architectural layout of the Hacienda de Regla
The Church at Regla remains one of its iconic features, its buttressed facade hopefully
having provided some blessing to the hard labour of the workforce. Just to the west of the
Church the housing of guests took place in the tall building with a corridor (see painting by
Rugendas, Figure 4-8). The height at which the living quarters are perched gives it a modern
look. The corridor facing the rooms provided an excellent viewpoint from which to oversee the
operations in the patio area:
‘We lean over the balcony of our hospitable quarters, awaiting breakfast, and see the horses
tread out the silver. A yard eighty rods square, poco mas y menos, is laid down to this work ...
beds of black mud are located over it ... two hundred horses are engaged in tramping out the
silver. Their tails are shaven, the mud has splashed up to their heads and backs ... they look ...
as if their labor were degrading ... eighty of these march round one circle, five abreast ... over
three hundred and fifty are owned by the company, and sometimes all of them are put into
service at once’.619
If the work was deemed degrading for the horses, no similar sentiment is expressed as
regards the local workforce. According to Ward the English investment of the 1820s in Regla
covered ‘stabling completed for 500 mules and horses’.620 By the late nineteenth century the
number of horses and mules dropped to around one hundred, with a monthly mortality rate of
around 5%.621 Animals, fodder and secondary materials would all be kept within the perimeter
of the hacienda, as well as areas for iron-working and carpenters, but their exact locations have
not been identified. It is possible that orchards also occupy some internal spaces, judging from
the anonymous painting in Figure 4-14. A fresh-water source, uncontaminated by the
619
Haven, Winter in Mexico, 155.
620
Ward, Mexico, I 424.
621
Informe Mensual Regla, 29 Jun 1872 – 27 Oct 1888.
327
operations of the Hacienda, would also have been required. The cluster observed at the most
northern tip of the hacienda comprises at least five circular structures capped by cupolas, but
their function and date of construction have not been identified.
4.4 The mass balance of the amalgamation process at Regla, 1872 to 1888
The main archival source for the data on consumption of materials and production of
silver at Regla for this chapter is the single tome accounting ledger that registers the mass
consumption of materials and production of silver by month, the Informe Mensual Hda. Regla,
29 Jun 1872 – 27 Oct 1888 (hereon referred to as Informe Mensual). The ledger identifies each
torta with a number, states the origin of the silver ore according to the mine it was extracted
from, registers the quantity of each type of ore that is treated in each torta, the initial silver
content of the ore that makes up each torta, the amount of silver extracted from each torta or
from each of the ore components of the torta, the corresponding percentage loss of silver, and
the consumption of mercury per torta or per ore component of the torta, and at times the
consumption of copper sulphate. The amount of data cover a fifteen year interval, though the
year 1874 was found to have been subject to major interruptions in production, to the point of
making it non-representative of the period.622 In total this represents a sample of 180 monthly
data sets from which a representative average can be derived for the most critical parameters
of the amalgamation process. I have not come across any similar set of published results in the
historiography of silver refining in Mexico or Peru up to the nineteenth century.
The raw accounting data, once checked for evident mathematical errors at source
wherever possible, are used to construe a mass transit profile on a monthly timeline for the
amalgamation process, both on an absolute and a relative basis using as common denominator
622
see Appendix B. The reasons for the atypical behaviour in 1874 is discussed in Chapter 5.
328
the production of one kilogram of silver. This profile is required to estimate the environmental
impact of the process as a consequence of solid, liquid and air emissions.
The time series on the inventory carried by Regla of the main materials required for the
process are reported in Appendix C, together with the assignment of spaces based on inventory
levels. The way the inventory was handled also provides insights on how the operational
management coped with the challenges of supply and forward planning of operations. The
numbers in the following sections define a very busy hive of physical activity, sounds and
emanations, day and night, converting mountains of ore into small silver bars whose final
volume paled into insignificance in the light of the tonnage of waste that every month was
dispersed into the environment.
4.4.1 Silver ore received for amalgamation.
The most important solids handling operation at Regla revolved around the
amalgamation process. Figure 4-26 shows the monthly amounts of ore received at Regla from
the mines in the Real del Monte area and earmarked for amalgamation. It is virtually identical
to the total sum of ores received at Regla. Among the mines that supplied silver ore to Regla
during this period were Guatimaztin, Dificultad, Aguichote, Jesus Maria, Moran, Rosario,
Porvenir, Sta. Ines, Sta. Brigida, Viscaina, Perro, San Genaro and Milanesa.623 Most of the ore
came from mines owned by the Compañia Real del Monte, though at times silver ore was
classified as ‘ajeno’ [belonging to another], indicating that tolling was carried out. Ores from
the same mine show a wide range of silver content. No information is provided in the
accounting ledger as to the nature of the silver minerals in the ore.
623
No attempt has been made in this analysis to calculate the amount and quality provided by each mine, and the
silver produced, though this information can be gleaned from the data in the ledger and from other primary sources
at AHCRMyP.
329
The average supply of total ore to Regla in this period was of 7,304 cargas per month
(1,008 t/m), of which an average 7,175 cargas per month (990 t/m) were milled for
amalgamation. The profile of deliveries over time shows a marked negative correlation in the
time series, a reflection of the contingent nature of a total mine production distributed over
different refining haciendas at the discretion of a central management.624 This average and the
whole time series in Figure 4.26 lies below the level indicated by Ward for Regla at the end of
the eighteenth century, when he claims ‘in 1795 five thousand cargoes of ore were received
there weekly’, though it is probable that he may have just been repeating an exaggerated
claim.625 The average inventory of raw ore for the period is 5,091 cargas (703 t) per month.
16,000
total incoming ore
for amalgamation
14,000
for smelting
12,000
10,000
cargas
8,000
6,000
4,000
2,000
Figure 4-26. Monthly deliveries of raw ore to Regla. Data adapted from Informe Mensual.
624
Negative correlation is the term applied to a sequence of high and low consecutive values within a time series,
or in more colloquial terms, the spikiness of a plot.
625
Ward, Mexico, II 140-41. Average silver production from Regla was higher over 1872 to 1888 than the average
amount of silver registered by the first Conde de Regla prior to 1795.
330
4.4.2 Milling of silver ore destined for amalgamation.
The raw silver ore destined for amalgamation was processed through the stamp mill
and the arrastres at an average monthly rate during this period of 7,239 cargas (999 t), as
reported in Table 4-I.626 In an era without electric power it is interesting to determine if water
supply was assured year-round or if the dry season (November to February) had an operational
impact on Regla, which did not use animal power to drive its arrastres.627 ‘Water in all new
mining countries is an exceedingly uncertain reliance, and a mill depending on it alone will
generally lie idle from a quarter to half a year’.628 This was not the case at Regla. There is no
evident correlation between monthly grinding output and the month of the year over the period
under analysis, as seen in Figure 4-27. The data show that the location of Regla was well
justified by offering a reliable source of hydraulic energy year round. The data in Table 4-I also
underline the operational flexibility shown by the milling units, since the values range from a
minimum of 1,714 cargas (237 t) in June 1875 to a maximum of 11,650 cargas (1,608 t)
registered in November 1872. In modern industrial terms, a turn-down capacity of one seventh
the maximum output for a batch processing plant is not common and represents a great
operational advantage to any production unit.
626
11% is the additional solids content from the erosion of the voladora stones in the arrastres of Regla according
to Laur, "De la metallurgie de l'argent au Mexique," 112. I have not made any corrections to the accounting data
since it does not include information on the erosion rate of the stones with which to cross-check Laur’s figure. An
error of at least plus or minus 5% should be assumed for all mass data.
627
The impact of climate on monthly production of silver in New Spain is also explored in Bernd Hausberger, La
Nueva España y sus metales preciosos (Frankfurt am Main: Vervuert, 1997), 113-25. Since his analysis is centred
on the amounts of silver registered at the regional Cajas (Treasuries), where factors such as distance from refining
centres, travel conditions, working hours of the Caja and other factors come into play, there is no direct
comparison between his data and the behaviour presented in this section for Regla. It is significant that in his
opinion ‘for our purposes it would be more useful to have analyzed the accounts of a refining hacienda.
Unfortunately I do not have at hand such an account’ - ‘para nuestro propósito seria de mayor utilidad el análisis
de las cuentas de una hacienda de beneficio. Desafortunadamente una cuenta de ese tipo no tengo a mi
disposición’. Ibid., 135.
628
331
1872 1873 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888
Jan 10,228 6,546 5,482 5,214 7,388 7,096 9,039 7,759 6,552 7,916 8,000 7,157 8,715 6,117
Feb 6,751 4,971 5,780 6,616 7,579 7,392 8,315 7,409 4,529 6,755 5,755 6,582 6,033
Mar 6,990 6,292 6,583 8,749 9,253 4,439 7,179 8,387 9,355 5,882 7,626 8,032 6,907 7,633
Apr 4,919 3,276 6,160 7,713 7,486 6,468 9,970 10,295 8,094 4,556 7,627 9,012 8,410 6,265
May 7,664 4,440 5,602 8,076 7,348 10,061 8,417 7,915 7,872 8,196 9,225 9,538 9,445 6,108 4,922
Jun 8,244 6,862 1,714 6,238 10,265 8,950 7,327 6,400 7,918 7,705 10,027 6,842 7,569 7,365 6,256 5,934
Jul 5,129 2,739 3,123 6,432 7,115 7,202 7,483 8,331 8,753 10,036 7,894 6,695 7,880 9,382 7,905 4,114
Aug 7,337 4,860 4,234 6,700 7,587 9,960 8,507 6,885 5,269 8,022 6,647 8,820 9,570 7,741 6,286 3,830
Sep 8,241 6,022 3,734 7,681 8,283 5,599 6,508 6,640 6,942 9,102 9,599 6,234 7,907 7,421 6,330
Oct 8,438 6,493 6,390 6,554 7,577 5,162 7,325 7,677 10,085 6,977 6,919 5,957 8,921 9,082 7,935
Nov 11,650 8,020 5,040 5,579 7,023 10,049 9,017 6,236 7,759 5,535 7,672 6,028 7,014 7,154 6,273
Dec 9,631 2,643 4,463 7,566 7,712 6,752 6,295 6,639 9,940 9,698 9,633 5,927 7,156 6,809 7,553
max 11,650 10,228 6,390 7,681 10,265 10,049 10,061 8,417 10,085 10,295 10,027 9,225 9,570 9,445 8,715 7,633
min 5,129 2,643 1,714 4,971 5,482 5,162 6,295 4,439 5,269 5,535 6,552 4,529 6,755 5,755 6,108 3,830
monthly average over period = 7,190 cargas
Table 4-I. Monthly amounts, in cargas, of ore ground for amalgamation. Raw data from
Informe Mensual.
10000
9000
8000
cargas
7000
6000
5000
4000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Figure 4-27. Average monthly values and standard deviation of ground silver ore destined
for amalgamation, in the period mid 1872 to mid 1888. Data calculated from Table 4-I.
Once the raw silver ore was ground in the stamp mill and in the arrastres, the average
monthly inventory over the whole period was 6,405 cargas (884 t) of ground ore. There is no
indication of major interruptions after 1874 up to 1888, and overall it shows a regular operation.
The level of inventory covers nearly one month’s milling production (see Table 4.I), which
332
would cover for short-term maintenance shut-downs or interruptions in the reception of raw
ore.
What do these levels of output say about the stamping capacity at Regla? The daily
milling capacity per stamp at a stamp-mill in the U.S. in the nineteenth century is said to
‘generally be from one to four tons per stamp in twenty-four hours, depending upon the
character of the rock and the weight and velocity of the stamp [and whether it is dry or wet
stamping]’.629 Assuming that the same 30 stamps reported in 1855 were still the number
operating in the 1870s (a major assumption), they would have needed to process over one ton
per stamp per day. Taking into account maintenance down times, an installed capacity of 2 t/d
would have allowed them to work at around 60% of total installed capacity, which seems a
more viable level. Based on these figures, each arrastre at Regla would have needed to process
a daily average of up to 10 cargas of ground ore (1.34 t) discharged from the stamp-mills. This
is a higher output from arrastres than the generic 9 quintales per day (0.4 t) mentioned by
Hermosa, but it is difficult to compare without knowing the mesh size of the ore being fed to
the arrastres, the hydraulic energy employed and the weight and nature of the voladora stones
employed.630 Arrastres with a 10 ft diameter (~ 3m) as in Regla were said to be able to grind 1
t/d if operated on a continuous 24 h basis.631
629
Ibid., 184, 185, 187.
630
Hermosa, Manual de Laboreo de Minas, 199. 1.5 to 3 m is the diameter given in Laur, "De la metallurgie de
l'argent au Mexique," 111. Hermosa’s output coincides with that of Humboldt, who in addition states that
arrastres had a diameter between 9 to 12 m, though his dimensions for arrastres do not match most other sources.
Humboldt, Essai politique, Tome IV, 58. A discussion on milling equipment in use at the end of the nineteenth
century for silver refining in Nevada can be found in Donald L. Hardesty, Mining Archaeology in the American
West: A View from the Silver State (Lincoln: University of Nebraska Press, 2010), 64-69. It focuses on how the
colder weather of Nevada required heating of the amalgam in pans, and shows how the physical structures of
milling and amalgamation had evolved prior to the introduction of the new cyanide process.
631
333
4.4.3 The amalgamation tortas (cakes).
Once milled in the arrastres, the next step was the preparation of the tortas, the cakes
which required the mixing of the ground silver ore with water to make a slurry, together with
added salt, copper sulphate and mercury. Figure 4-28 indicates a very coordinated process, with
a match within 1% between the monthly amounts of ore ground for amalgamation, and the
amount processed by amalgamation each month. The monthly average for the period is 7,166
cargas (989 t) processed by amalgamation.
With the exception of the year 1874, when production was seriously interrupted for
much of the year, it shows from 1877 to 1887 a period of relatively steady throughput during
production, the exceptions being the periods at both extremes. In 1855 the newly formed
Mexican company was aiming for an amalgamation capacity at Regla capable of processing
50,000 cargas in a year.632 By the 1880s the company was amalgamating over 60% more than
that initial target. If the size of the patio area did not change, nor the quality of the ore, what
can explain this major jump in output?
14,000
12,000
10,000
8,000
cargas
6,000
4,000
ground ore
2,000
amalgamated ore
0
Jun-72
Dec-72
Jun-73
Dec-73
Jun-74
Dec-74
Jun-75
Dec-75
Jun-76
Dec-76
Jun-77
Dec-77
Jun-78
Dec-78
Jun-79
Dec-79
Jun-80
Dec-80
Jun-81
Dec-81
Jun-82
Dec-82
Jun-83
Dec-83
Jun-84
Dec-84
Jun-85
Dec-85
Jun-86
Dec-86
Jun-87
Dec-87
Jun-88
Figure 4-28. The monthly amounts of silver ore ground for and processed by amalgamation.
Data from Table 4-I and the Informe Mensual.
632
334
The output in silver of the hacienda is a function not only of its milling capacity but of
the time it took to complete an amalgamation run for each torta on its patio reactor so as to
free space for the next batch. It is fortunate that the accounting data from October 1872 to
December 1873 also included the amalgamation period for each torta.633 Figure 4-29 is a
histogram of the days it took to complete the amalgamation runs during this period. On average
it required 13 days, with some runs as short as eight days and the longest at 18 days.634
20
18
number of amalgamation runs
16
14
12
10
8
6
4
2
0
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
number of days
Figure 4-29. Histogram of the days required for amalgamation, as recorded over the period
October 1872 to December 1873. The average amalgamation run lasted 13 days. Raw data from
Informe Mensual.
Thanks to the accounting data it is also possible to follow the extraction of silver as a
function of the time allowed for amalgamation. According to Figure 4-30 on average
633
From 1874 to 1888 no further data are registered regarding the amalgamation days required for each torta.
634
One run was recorded at 24 days in December 1873, but I have not included it in the data set, one datum point
out of a total set of 118 data points.
335
approximately 93% of silver had been extracted after two weeks. I have plotted the raw data as
a scatter graph and also as averages. The linear regression analysis gives nearly identical results
100
98
% silver extracted
96
94
92
90
y = 0.437x + 87.379
88 R² = 0.2743
86
5 10 15 20
amalgamation days
100
98
% silver extracted
96
94
92
90
y = 0.4384x + 87.456
88 R² = 0.8396
86
5 10 15 20
amalgamation days
Figure 4-30. Scatter graph and plot of averages as measured at various amalgamation periods
that show an evident correlation between the percentage of silver extracted and the number of
days for an amalgamation run. Raw data from Informe Mensual.
in both cases, though the plot of averages makes it easier to visualize the correlation. If I assume
that the extraction process was linear, 100% extraction of silver would occur on average after
an amalgamation period of 30 days. For the moment I will ignore the real possibility that the
amalgamation process is not linear throughout but becomes asymptotic towards its final
336
stage.635 The question is whether it made business sense to aim at 100% extraction of of silver
by waiting another two weeks, or to optimize the use of the patio reactor by amalgamating two
batches of ore within the same period of time. 93% of the silver of two batches of ore is better
than 100% of silver from just one batch. The law of diminishing returns applied after a two
week amalgamation period. Under a scenario of sufficient and continuous ore supply, there
was little incentive to carry out the amalgamation reaction to its final conclusion.
The profile of the histogram in Figure 4-29 is a reflection of this operating strategy as
carried out in the 1870s, but it was not always the case at Regla. In 1855 Buchan reported that
‘amalgamation lasts from thirty to fifty days’.636 At some point in the intervening years the
operators at Regla had realized it made little economic sense to wait beyond the two week
period, and this explains the increase of over 60% in the amount of monthly amalgamated ore,
assuming all other operating variables were kept equal.637
Among the advantages of the patio process was that instead of using a batch reactor of
fixed capacity, like a barrel, the slurry was spread out into one or more tortas of whatever size
was needed. The amount of silver ore processed in each torta thus varied widely, and
constituted the operational flexibility of this process. Figure 4-31 shows the histogram of the
635
Data already reported in the nineteenth century point to an asymptotic extraction rate for silver during patio
amalgamation, see Laur, "De la metallurgie de l'argent au Mexique," 157-58.
636
Buchan, Report Real del Monte, 9. The patio process has been critiziced both by its contemporaries and by
later historians for the long periods required to complete the extraction of silver from the ore. The data from first-
hand observers indicate a process taking weeks, in some special cases up to two months, and then only subject to
the difficulty of the ore to be amalgamated. For example, 8 days to more than 2 months, in Dominguez de la
Fuente, Leal Informe Politico-Legal, 93.; 8 days to 2 months, Sonneschmidt and de Fagoaga, Tratado de la
amalgamación de Nueva España, 32.;18 to 60 days in Duport, Métaux précieux au Mexique, 267.;10 to 25 days,
according to the nature of the ore, ambient temperature and degree of mixing during amalgamation, in Hermosa,
Manual de Laboreo de Minas, 233. Periods longer than 2 months cited in the secondary literature should be treated
with caution, such as the five months cited in Humboldt, Essai politique, Tome IV, 66.
637
Ruiz states that the frequency of mixing each torta at Regla was increased, which led to a shorter amalgamation
period. There is not enough information to determine the order of factors that led to the use of a shorter turn-
around time for the tortas in the patio reactor. Ruiz de la Barrera, "La Empresa de Minas del Real del Monte,"
309.
337
size of tortas according to their charge of silver ore. The minimum number of cargas per torta
in this period was 18 (2.5 t of silver ore), the maximum 2,059 (284 t), with an average of 753
cargas (104 t) per torta.
How many tortas on average could be amalgamated at Regla during one month? It
ranged from one to sixteen, of different sizes, with an average of 9.4 tortas, indicating this patio
reactor had an average capacity of one thousand tons of ore at any one time. The different
amalgamation periods dictated by the azoguero in charge of the process, even within the limits
observed in the histogram of Figure 4.28, would in time introduce a natural and unpredictable
staggering of the runs.
450
400
350
number of tortas
300
250
200
150
100
50
0
50
1050
1150
1250
1350
1450
1550
1650
1750
1850
1950
2050
150
250
350
450
550
650
750
850
950
number of cargas per torta
Figure 4-31. Histogram of the number of cargas per torta as practised at Regla (1872 to
1888). Source data from Informe Mensual.
Is it possible to relate the area of the patio reactor to the refining capacity of an
amalgamation hacienda? The majority of the surviving silver refining haciendas in Mexico are
reduced to ruins, with no or limited extant written records to provide a clue to their production.
What does sometime survive to the extent of being measureable is the area of the patio reactor,
338
or a close approximation to the original space. A rule of thumb can be established using the
data from Regla. Taking Buchan’s figure of the area of the patio reactor at Regla of 6,000 m2,
we have seen it process on average 1000 t of ground ore per month. Thus the processing of one
ton required a patio reactor area of 6 m2. Assuming a 93% extraction rate for silver in two
weeks, total production can be estimated from the area of a patio reactor according to the
following equation:
Monthly production silver in tons = 0.93 × (area of patio reactor in m2 ÷ 6) × (% silver
in ore by weight) × (0.5 ÷ average amalgamation period in months)
4.4.4 The silver content of ores used for amalgamation.
The accounts provide two sources for the silver content of the ores processed by
amalgamation. As an example, Table 4-II is a transcription of the raw data as reported for the
five weeks ending the 28th November 1885.638 Over the whole period from 1875 to 1888 the
silver content per monton of incoming ore can be used to prepare a histogram such as Figure
4-32, which indicates a bimodal distribution due to the overlap of two different distribution
curves over the period. The average of 11.2 marks per monton of 10 cargas corresponds to 0.19
% of silver by weight.639 The other value of interest is provided by dividing the total silver
refined during these years (303,334 kg) by the total weight of ore processed by amalgamation
(178,287,306 kg), in which case an average value of 0.17% is obtained. The difference
638
Monthly runs were accounted for in periods of either four or five weeks, and the date registered in the ledger
is the last day of any such period.
639
According to Humboldt, the silver content of the lowest quality ore from the Vizcayna vein at the start of the
nineteenth century had been 4 marks per monton, just 0.06% silver by weight. The data from Regla place the ore
quality for amalgamation in the latter part of the century above what Humboldt had classified as the second tier
silver ores (0.16%): ‘dans le district des mines de Pachuca ...les minerais de la seconde classe, 7 a 10 marks …
les plus pauvres, qui forment la troisième classe, ne sont évaluées qu’a 4 marks d’argent par monton.’ Humboldt,
Essai politique, Tome III, 370. In 1882, silver ores with less silver content were being amalgamated in the United
States. For example, ores with 41 ounces of silver to the ton (0.12%) were refined with mercury at Tombstone
Mill, in Arizona. Egleston, The Metallurgy of Silver, 422.
339
corresponds to the amount of silver left unextracted (approximately 10%), if the values for the
raw ore can be trusted (see Section 4.5.3). Furthermore, if these values of refined silver are
now plotted by month over the same period, it can be observed that the percentage of refined
silver decreased over the period, as seen in Figure 4-33, pointing to a lowering of the quality
of the ore over time. Table 4-II also shows how varied was the mix of ores used to constitute
an amalgamation torta, with silver contents per monton of some of the constituent ores reaching
up to 61 marks (1 % silver by weight) in this example.
800,000
weighted average ley
700,000
for amalgamation 11.2 marks / monton
600,000
500,000
400,000
cargas
300,000
200,000
100,000
0
4-6 6-8 8-10 10-12 12-14 14-16 16-18 18-20
marks per monton
Figure 4-32. Histogram of ore quality refined by amalgamation. Raw data from Informe
Mensual.
340
0.30%
0.25%
% silver recovered
0.20%
0.15%
0.10%
0.05%
6-72 6-73 9-75 9-76 9-77 9-78 9-79 9-80 9-81 9-82 9-83 9-84 9-85 9-86 9-87
June 1872 to June 1888
Figure 4-33. Decrease in percentage of refined silver from the ores processed by
amalgamation. Percentage calculated from raw data in Informe Mensual.
4.4.5 Interlude: pizzas or juggling acts.
As discussed in Chapter 3, a torta once formed from all its ingredients would be
expected to be a whole entity, an amorphous slurry subject to the constant mixing of its
components thanks to the tread of humans or animals. Thus, once the seven different silver
ores that comprise torta number 3397 in Table 4-II have been mixed with the ingredients of
the amalgamation recipe, it seems an impossible task to be able to measure the specific amount
of silver recovered from each ore fraction, or the specific amount of mercury losses that can be
assigned to each type of ore present in the torta.640 To argue otherwise would require the
concept that not all tortas were equal, that some tortas could somehow be sectioned like a pizza
into different ore flavours, each slice kept meticulously apart during the whole process, from
wet area to refined silver, and its consumption and production data carefully accounted for at
640
The same statement applies to detailed accounting of salt and copper sulphate consumption according to ore
type in hybrid tortas.
341
Account of refined ores [Cuenta de metales beneficiados ]
SILVER CONTENT SILVER PRODUCED LOSSES

PER MONTON [PLATA PRODUCIDA ] [PERDIDAS ]
REFINED ORES [LEY POR MONTON ]
CARGAS
[METALES BENEFICIADOS ] Crude ore Total Per Monton Of silver Of mercury
[Metal crudo ] [Total ] [Por Monton ] [De plata ] [De azogue ]
Marks [Marcos ] Marks [Marcos ] Marks [Marcos ] Percent [Por ciento ] Total [Total ] Per mark [Por marco ]
3397 Jesus Maria 149 8 52 117 7 85 7 87 90 14 31

San Miguel 44 7 5 28 6 34 9 68 21 12 -
Sta Brigida 76 31 97 223 29 34 8 23 170 12 20
Corteza 13 10 77 13 10 - 7 14 101 12 31
Aviadero 20 40 - 73 36 50 8 75 56 12 27
S. Pedro y S. Pablo 16 61 87 91 56 88 8 8 70 12 31
S. Antonio 1 50 - 5 50 - - - 4 12 80
3398 Dificultad 480 11 75 790 11 62 1 13 497 10 7
3399 id 827 8 - 674 8 15 1 81 437 10 37
3400 Aguichoto 730 12 25 840 11 50 4 4 569 10 84
1 Dificultad 1185 11 - 1180 9 96 9 51 821 11 13
2 Rosario 463 17 25 1018 15 35 11 11 656 10 31
3403 Guatemozin 701 15 50 1050 14 98 3 40 739 11 26
3405 Dificultad 118 10 - 122 10 34 3 39 95 12 46
Aguichoto 41 10 - 42 10 24 2 44 33 12 37
Guatemozin 17 10 - 17 10 - - - 13 12 23
San Gregorio 5 10 - 5 10 - - - 4 12 80
3407 Id id 449 7 25 200 8 3 10 50 149 11 92
5535 12 38 6488 11 72 5 30 4434 10 93
Smelting
[Fundicion ]
Dificultad 86 112 33 959 111 51 0 72

Aguichoto 48 112 29 535 111 46 0 74
Perro 16 137 50 219 134 88 0 45
Jesus Maria 4 85 - 34 85 - 0 0
San Antonio 6 100 - 40 100 - 0 0
Rosario 6 131 67 78 130 - 1 27
Guatemozin 3 123 33 37 123 33 0 0
Jacal 3 110 - 33 110 - 0 0
Porvenir 2 85 - 17 85 - 0 0
174 114 8 1972 113 33 0 65
Table 4-II. Partial transcription of data that appear for the five weeks ending on November
28th, 1885, as registered in the Informe Mensual. Numbers in italics represent numbers shown
in red in original document. The units of silver content are registered in marks and hundredths
of a mark.
every stage according to ore components. Either this represents a degree of operational
complexity unsuspected in the historiography, or a degree of accounting creativity of the data.
I have placed a grey background on the data in Table 4-II where there is no question it
would be obtained from direct measurements: the silver content of each component ore, the
342
amount of each ore weighed and added to the torta, and the total amount of silver refined from
the whole torta. This last quantity is reported in Table 4-II as a segregated quantity according
to ore type for the hybrid tortas, but it could have been calculated by the accountants.641 Other
notations are more difficult to justify. The ore fraction of torta 3397 with a silver content of 50
marks per monton is reported as not incurring any silver loss during refining, while the values
of silver loss for each of the other ore fractions is different and reported to a high degree of
precision, for an average of 8.17%.642 The data on mercury losses defie a similar calculation to
arrive at individual values for each ore fraction. One probable explanation is that for those cases
where hybrid tortas are registered, an accounting juggling act took place whereby certain boxes
were filled with creativity just to arrive at a known average value.643 The hybrid torta of Table
4-II is the occasional exception, since the great majority of tortas reported in the Informe
Mensual Regla are made from a single type of ore.
4.4.6 The mass balance for salt.
After the ore, salt was the major solid being stored and used at Regla. Figure 4-34 plots
the monthly consumption of salt required to produce 1 kg of silver, giving an average of 29.9
kg of salt per kg of silver. The sudden increase in salt consumption that is evident at the tail
end of this period shows a departure from what had been until then a relatively stable pattern.
The cause is not known, and is not repeated for the other components of the recipe. Since the
quality of the salt is not registered, I can only surmise that a poorer salt may have been used at
641
In the case of torta 3405 there is in fact no loss, but a gain in silver after refining. This type of result figures
predominantly in the ledger, and is due to errors in determining the average silver content of tons of ore from a
very small assaying sample of the ore.
642
A range of silver losses during patio amalgamation between 12% and 36% is given in Laur, "De la metallurgie
de l'argent au Mexique," 186-87.
643
It also challenges the reader as to why the ore fraction from Corteza should show such an abnormal
consumption of mercury registered in pounds, while the corresponding value of consumption of mercury per mark
of silver for this ore is clearly wrong. See footnote 125 on Laur’s mistrust of nineteenth century accounting
practices in haciendas.
343
the end. On a monthly basis an average of 4,310 arrobas (49.6 t) of salt was consumed during
the amalgamation process, approximately 5 t per average torta. Overall the ratio at Regla is
within the range of 2 to 5 % per torta suggested in a Mexican operations manual dated 1857.644
The average inventory value in this period was 18,459 arrobas (212 t).
60.0
50.0
40.0
kg
30.0
20.0
10.0
0.0
Jun-72 Jun-73 Jun-74 Jun-75 Jun-76 Jun-77 Jun-78 Jun-79 Jun-80 Jun-81 Jun-82 Jun-83 Jun-84 Jun-85 Jun-86 Jun-87 Jun-88
Figure 4-34. Consumption of salt per kg of silver refined by amalgamation (1872-1888). Raw
data from Informe Mensual.
Based on the chemical reactions set out in Chapter 3, if all the silver refined was present
originally as silver sulphide (Ag2S), then the stoichiometric requirement of salt per kg of silver
refined is 2.2 kg.645 Thus on average over 10 times more salt was being added than was required
by the basic chemistry of the conversion of silver sulphide to silver chloride. Since the purity
of the salt used at Regla would have been consistently under 100%, this explains in part the
higher ratio used. The excess of salt also covers the amount of salt lost as the saline water of
the slurry seeped into the soil during the amalgamation period, when water was being
constantly added to avoid the drying-out of the cakes.
644
645
According to Reaction 3 in Section 3.4, 2 moles of silver chloride are produced from the reaction of 8 moles
of sodium chloride with 1 mole of silver sulphide. Thus 2 moles of silver (215.72 g) require 8 moles of sodium
chloride (467.6 g), a weight ratio of 2.2 kg pure salt to 1 kg of silver.
344
4.4.7 The mass balance for copper sulphate.
The next major additive, as measured in quantity used and stored, was copper sulphate.
Figure 4-35 shows the profile of its monthly consumption per kg of silver refined by
amalgamation. On average 9,774 lbs (4.4 t) of copper sulphate were used per month, for an
average ratio of 2.6 kg of copper sulphate per kg of silver refined. In contrast to the use of salt,
it is not only the purity of the copper sulphate that can influence the amount used. Copper
sulphate or copper magistral was added according to the subjective judgement of the azoguero
in charge of the amalgamation process. Recommended ranges went from 0.2 to 8 %, reflecting
the variety of criteria in its use.646 At Regla an average of 0.4% was added, based on the weight
of the ground ore.
7.00
6.00
5.00
4.00
kg
3.00
2.00
1.00
0.00
Figure 4-35. Consumption of copper sulphate per kg of silver refined by amalgamation

(1872-1888). Raw data from Informe Mensual.
If all the silver refined was present originally as silver sulphide (Ag2S), then the
stoichiometric requirement of copper sulphate per kg of silver refined is 2.1 kg.647 In contrast
to salt, operators at Regla were using quantities much closer to the theoretical requirement,
646
647
According to Chapter 3, 2 moles of silver chloride are produced from the reaction of 2 moles of copper ions
with 1 mole of silver sulphide. Thus 2 moles of silver (215.72) require 2 moles of copper sulphate (446.32), a
weight ratio of 2.07 kg pure copper sulphate to 1 kg of silver.
345
taking into account the expected variations in its purity and the losses through seepage to the
soil. The presence of native silver or silver chloride in the ore would have lowered the need for
copper sulphate, and so may have provided some respite to their recipe. Approximately 0.5 t
would have been added to every average torta of 700 cargas.648 The level of monthly inventory
over this period was 15,109 lbs (6.9 t),
4.4.8 The mass balance for mercury used during amalgamation
The monthly ledger for this period did not register the amounts of mercury added to
each torta, only the amounts consumed, so they do not provide the total amount of mercury in
circulation on a monthly basis at Regla. Buchan stated that 18,000 lbs (8.2 t) of mercury were
in use on a monthly basis in 1855.649 This corresponds to a weight ratio of five parts of added
mercury per one part of assumed silver content in the ore.650 It also indicates the monthly
amount squeezed through the fingers of the workers at Regla. The profile of its monthly
inventory level provides an insight into how an amalgamation operation was run. Throughout
this whole period Regla maintained an average monthly inventory level of mercury of 29,830
lbs (13.6 t). This represents an inventory equivalent to 6.8 months of average mercury
consumption at Regla. John Taylor, the General Manager of the Real del Monte Company, was
prescient enough to write in 1824 from London:
‘I know that it will be necessary for each establishment to keep at all times a large stock of
this article [mercury] at the mines; and that thus, besides the risk of plunder and waste, a greater
648
Humboldt states that 1 to 7 pounds of magistral were added for every pound of mercury. In Regla this ratio
was just over 2 during the period. Humboldt, Essai politique, Tome IV, 64-65.
649
650
Assuming an average production of 1.6 t of silver produced by amalgamation per month and a recipe of 5 to 6
parts mercury to one part silver in the ore, then approximately 8 to 10 t of mercury would have been added on a
monthly basis to the ores under amalgamation. For an average torta of 740 cargas, this requires adding near 1 t of
mercury during amalgamation. This is the amount of mercury squeezed through the fingers of the workers for
each torta in the patio. Approximately 80 one-litre flasks of mercury would have been required per torta. For the
proportion of mercury to silver in the ore, see Hermosa, Manual de Laboreo de Minas, 216. Humboldt reports
that the weight of mercury added was six times the deemed content of silver in the torta. Humboldt, Essai
politique, Tome IV, 64.
346
capital than would otherwise be required must be provided and locked up in a distant
country’.651
Whether by coincidence or intent, Regla always had enough reserves of mercury to
jump-start the process if necessary from scratch.
4.4.9 The mass balance for silver.
The only commercial end product at Regla was silver.652 The average monthly
production of silver at Regla during this period via amalgamation was of 7,327 marks (1.7 t).
As observed in Figure 4-36, Regla managed a steady production rate from mid 1876 to mid
1887, with just one singular peak in production in early 1882. Regla would cast silver bars
from both amalgamation and smelting at an average rate of 64 bars per month, with most bars
registering a weight of 140 marks (30 kg). This reflects the account given by Countess
Kollonitz in 1864 : ‘[silver] is cast in very heavy bars ... every fifteen days 28 bars of silver are
produced ... twice per month the company guards escort the silver bars to the sea ports, from
where they are sent mainly to England’.653
How efficient was the patio amalgamation process at Regla, how much silver would
have been left in these ores irrespective of the amalgamation period adopted? There is no
straightforward answer to this question. First of all, the chemical nature of the batches of ore
amalgamated at Regla is not reported in the account books and is expected to have changed
651
Taylor, Selections from Humboldt, xxii.
652
Very minor amounts of lead were sold occasionally, as well as materials drawn from the storehouse, but none
represent amounts of note.
653
‘se funde en pesadísimas barras … cada quince días salen 28 barras de plata … dos veces al mes la guardia
de la compañía escolta los lingotes de plata hasta los puertos de mar, de donde son enviados principalmente a
Inglaterra’ in Paula Kollonitz, Un viaje a México en 1864 (México: Secretaría de Educación Pública, 1976), 146.
Silver bars weighed 31.3 to 35 kg according to Duport, Métaux précieux au Mexique, 117.; Amador, Tratado
práctico de haciendas de beneficio, 88.
347
from batch to batch.654 Second, by curtailing the amalgamation period to approximately two
weeks, a sacrifice of at least 5% is made on the maximum operational level of extraction of
silver. Third, even though the silver content of the raw ore was measured and registered in the
accounts, in many cases it is obviously incorrect, since it is registered as lower than the final
4,000
3,500
3,000
2,500
kg
2,000
1,500
1,000
500
0
Figure 4-36. Monthly production in kg of silver by amalgamation (1872-1888). Raw data

from Informe Mensual.
percentage of silver extracted from the ore. The operational challenge of sampling tons of ore
so as to obtain a small representative amount that can be analyzed in a cupel for silver content
is obvious, and it seems it was not overcome at Regla until the end of the period in question.
The ledger for Regla is full of data on silver ‘loss’ inked in red to indicate this negative territory
where more silver is registered as extracted than was measured in the raw ore.655 When the
non-negative values for the loss of silver are plotted, the average ‘loss’ of silver during patio
amalgamation at Regla is just 5.1% with a standard deviation of 2.8%, a reflection on the large
654
The silver ores of Real del Monte containing manganese, antimony and lead were known to be difficult to
amalgamate using the patio process, according to Phillips, Metallurgy Silver, 327.
655
Laur argues that the issue of silver refined by amalgamation being greater than expected from the range of
silver assayed in the ores is a result of large-scale manipulation of accounts by the administrators of haciendas to
cover for inefficiencies in their operation, including pilfering. Laur, "De la metallurgie de l'argent au Mexique,"
183-84.
348
variations observed in the data. The large gaps in data from 1875 to 1881 in Figure 4-37
correspond mainly to negative values that have been ignored, not to a lack of diligence on the
part of the assayers. The average corresponds well to what was expected from an amalgamation
interruptus.
Finally I have already pointed out the fact that the total amount of extracted silver is
approximately 10% less than would be expected from the average silver content of the ore
being amalgamated. Of this shortfall, around 5% corresponds to the curtailed amalgamation
process, and the remainder could well be the dilution of the ground ore with inert solids from
the wear of the voladora stones. Overall there is a good match between the data, and any
pilfering that took place was hidden within the range of error of the measurements taking place
and/or the registers in the accounts.656
14.0%
12.0%
Unextracted silver
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
6/72 6/73 6/74 6/75 6/76 6/77 6/78 6/79 6/80 6/81 6/82 6/83 6/84 6/85 6/86 6/87
Figure 4-37. Losses of silver, expressed as percentage, during amalgamation (1872-1888),

calculated from raw data in the Informe Mensual. All negative values reported in the ledger
have been excluded.
656
‘criminals against the industry’ were thieves of mercury, silver or tools, ‘who could be punished on the spot’.
Ladd, The Making of a Strike, 69.
349
4.4.10 The mass balance for mercury consumed
The ledger registers the amount of mercury consumed during the amalgamation
process. This detailed bookkeeping requires keeping track of any mercury recovered a) during
the separation and washing of the amalgam from the rest of the torta b) during the mechanical
extraction of mercury from the amalgam in the manga and c) during the recovery of the
mercury from the capellinas. The average monthly consumption of mercury at Regla was 4,638
lbs (2.1 t). When plotted as a function of kg mercury consumed per kg of silver refined, the
plot shown in Figure 4-38 shows two distinct sections. The first, at approximately 1.1-1.2 kg
of mercury for every kg of silver, is from 1875 to 1878 and at times is surprisingly smooth, not
only compared to the other plots of consumption of reagents, but also compared to the time
series on either side.657 After 1878 the plot averages around 1.4 kg of mercury per kg of silver.
The average value for this ratio from 1872 to 1888 at Regla comes to 1.25.658 I have commented
upon the fact that lower ranges are a sign of a) ores rich in native silver or b) the presence of
iron or copper in the recipe, which are reducing agents of silver chloride that lower the
consumption of mercury. The historiography confirms both factors are at work here.
Humboldt mentions that at the beginning of the nineteenth century the ores from the
Vizcayna vein of Real del Monte contained silver sulphides mixed with native silver.659 In his
review of historic mines of Mexico, published in 1883, Dahlgren states that as to the ores of
Real del Monte mines :
657
It is tempting to consider some accounting sleight of hand to explain the smoothness of data that one expects
by nature to be more erratic in behaviour. Some data could have been extrapolated, based on the expected values
of the ratio together with some coordinated adjustments to inventory and purchase orders. This would only be
explained if they were making up for non-existent field measurements.
658
The amalgamation recipe used in Potosí in 1603 shows a mercury to silver ratio of 1.3 with iron included as
an important part of the recipe. Anonymous, "Relaciones Geograficas de Indias - Peru I," 376.
659
‘Le filon de la Biscaina renferme ….de l’argent sulfuré mêlé d’argent natif’. Humboldt, Essai politique, Tome
IV, 16.
350
‘near the surface the usual “Colorados” appeared while as depth was attained Silver Sulphides,
native Silver (in quantities which have made this district famous), Dark Ruby and Prismatic
black Silver (the “Negros”) took the place of the “Colorados”. There was some Galena.’660
A modern treatise on this mining district also draws attention to earlier ores such that
‘native silver … in previous times was very frequent’.661
2.00
mercury to silver weight ratio
1.80
1.60
1.40
1.20
1.00
0.80
Figure 4-38. Weight ratio of mercury consumed to silver refined by amalgamation (1872-
1888). Raw data from Informe Mensual.
As to iron, the other haciendas of the Compañia Real del Monte were routinely adding
it to the amalgamation recipe so as to decrease mercury consumption during the barrel
process.662 Iron appears among the materials consumed at Regla, though it is not possible to
discriminate in the accounts the amounts destined for amalgamation, as an additive in smelting
and iron for general use in the hacienda (Table 4-III).663 However the very high wear and
660
Dahlgren, Historic Mines of Mexico, 196.
661
‘la plata nativa … en épocas anteriores fue bastante frecuente’ Galindo y R, Distrito Pachuca-Real del Monte,
5.
662
Iron was a standard additive in the recipe for the barrel process used in Mexico in the nineteenth century. Laur,
"De la metallurgie de l'argent au Mexique," 98.
663
Table V is an extract from the Memoria No. 19 to Memoria No. 21 de los Gastos de la Hcda de Regla, ,
corresponding to the four weeks ending on the 26 th May 1877. It shows that in addition to salt, copper sulphate
and mercury, the following reagents were also consumed: tequezquite (also spelt tequesquite), a naturally
occurring salt used as flux for the smelting process; lime, used to control the amalgamation process if the azoguero
deemed he had added too much magistral, and different types of iron. It is widely reported that iron was a
351
consumption of horseshoes made of iron at Regla raises the possibility that part of the iron that
reduced the silver chloride in the slurry came from horseshoes, as the following calculation
shows. The number of horseshoes replaced during the period covered in Table 4-III is
equivalent to up to 600kg of iron, at 2 kg per heavy duty shoe. The wear on the horseshoes
would have been an unwitting but extremely effective source of iron for the reduction of silver
chloride. The constant erosion of its surface on a very abrasive ore slurry would have spread
finely ground iron into the slurry every time each horse or mule tread his circular path through
a torta. Even if only 50% of the mass of each horseshoe found its way into the amalgamation
slurry, it would have been a very useful if unintended contribution to the observed lower
mercury to silver weight ratio. Whole mule and horse shoes were added to the barrel
amalgamation recipe mix in the United States in the nineteenth century.664
consumption
year 1877, tequezquite lime iron slag english iron mixed iron horseshoes
week ending
cargas cargas cargas lbs lbs number
05-May 2.3 0 43.5 0 0 114
12-May 2.8 4 27.8 0 0 43
19-May 3.0 0 27.3 993 0 87
26-May 2.8 0 0 0 1,053 78
total 10.8 4 98.5 993 1,053 322
Table 4-III. Additional consumables at Regla, from Memorias de Gastos for the four weeks
ending on May 26th 1877.
necessary component in the smelting of silver ores, for example Pique, A Practical Treatise on Silver, 69.; Phillips,
Metallurgy Silver, 433.; Willies, "Derbyshire Lead Smelting," 3.
664
Egleston, The Metallurgy of Silver, 38. In Potosí in 1603 iron was added in a proportion of 0.5 to 1 with respect
to the silver refined, and this was enough to bring down the mercury to silver ratio to 1.3. This would require
adding approximately between 1 to 2 t of iron per average torta at Regla. At the lower ratio horsehoes could have
added inadvertently around 30% of the required iron additive to the recipe.
352
4.4.11 The overall mass balance for the amalgamation process at Regla
Table 4-IV summarizes the magnitude of the main environmental impact vectors from
the amalgamation process as carried out at Regla within the period covered by the accounting
ledger. Based on the analysis presented in Chapter 3, Section 3.10, I propose that on average
85% of the total mercury consumption was in the form of solid calomel washed away after
amalgamation, and the remaining 15% corresponded to a physical loss of mercury. The
chemical transformation of mercury into calomel would consume on average 1.8 t/m of
mercury into an insoluble solid that was entrained in the water used to wash the tortas, and
deposited over an undetermined area in the water basins downstream from Regla. A total of
0.3 t/m of mercury was lost due to physical losses, of which the largest share would be by
entrainment in water used for washing tortas or in the condensation channel underneath the
capellina, together with seepage into the soil in the patio reactor area, for a total estimated
average of 0.28 t/m. Air emissions of residual mercury in silver being cast into bars, 1% of the
total weight of silver, represented an average of 0.02 t/m. A loss via evaporation from the tortas
that will not be taken into consideration (estimated as 0.002 t/m).665
According to this scenario, for every kg of silver refined by amalgamation at Regla,
approximately 1.1 kg of mercury was lost as calomel, 0.2 kg was lost as liquid mercury in the
waste water or by seepage to the soil at Regla, and just over 10 g were lost as volatile mercury
from the casting of silver bars. In terms of mass these losses pale in comparison to the 620 kg
of solid mineral waste and salts in solution that were washed downstream from Regla for every
kg of silver refined by amalgamation.
665
If all the mercury was exposed to sun and wind the deemed loss via evaporation would amount to 14 kg of
mercury per torta in two weeks, or a total of 250 kg for 9 tortas every month. Since only a minority of mercury
droplets would be exposed to sun and wind, I will assume that 1% of this theoretical total was the effective loss,
2 kg per month. Calculations based on Winter, "The Evaporation of a Drop of Mercury."
353
The yearly average amounts to approximately 12,000 t of mineral waste and salts, 20t
of mercury in the form of calomel, 4 t as losses of liquid mercury and 0.2 t of mercury via air
emissions.
amalgamation
copper
raw ore salt net mercury to be consumed total
average sulphate
input
t/m 989 49 4 2.1 1,045
chemical loss physical loss
ground and water air emissions

average silver waste ore waste additives
output mercury loss mercury loss as
mercury from casting silver bars
as calomel liquid mercury
t/m 1.7 987 53 1.8 0.28 0.02 1,045
average input copper

raw ore salt net mercury consumed
(kg) per kg sulphate
refined silver 587 30.0 2.6 1.3 621
waste additives and

average output mercury loss mercury loss as
silver waste ore byproduct salts mercury from casting silver bars
(kg) per kg as calomel liquid mercury
(except calomel)
refined silver
1 587 32 1.1 0.2 0.01 621
Table 4-IV. Overall mass balance for the amalgamation of silver ores as practised at Regla
between 1872 and 1888. Data compounded from different sections of this chapter. The numbers
in italic denotes a calculated number, not directly derived from the monthly accounting ledger.
4.4.12 The energy balance of the amalgamation process
Patio amalgamation as practised at Regla in this period was not a major consumer of
fuel. The only main process requirements for heat came from the capellinas and from the
casting of silver bars. A guide to the consumption of fuel is provided in the Memorias de Gastos
of weekly consumption of wood and charcoal used at Regla (process, cooking, heating), and
one example is reproduced in Table 4-V. These Memorias de Gastos register both consumption
and purchases to re-stock the central warehouse (data in italics), and covers both amalgamation
354
and smelting. Firewood in some Memorias de Gastos is explicitly designated as ‘leña para
afinaciones’, firewood for refining by cupellation. To arrive at a ratio of wood and charcoal
per kg of silver refined by amalgamation, the first column in Table 4-V is assigned as fuel for
the cupellation stage of the smelting process, while the second I will tentatively assign for the
heating of capellinas. The charcoal from ocote I assign to the casting of silver bars. The final
column of charcoal corresponds to the smelting of silver ore. In these four weeks 1,999 kg of
silver were produced by amalgamation.666 It should be noted that cooking and other domestic
requirements for fuel would have also been included within these figures, but I have ignored
these cosnumptions for want of information.
consumed or purchased (re-stocked) , central warehouse

year 1877, charcoal
firewood ocote charcoal
week ending (ocote)
cargas arrobas cargas cargas
05-May 366.5 206 27 189.5
50
12-May 146 24.5 150.5
81
50
19-May 118 18.5 209
1,183
49 850
26-May 164 19.5
993 187.5
total 515.5 634 89.5 850
Table 4-V. Total consumption of firewood and charcoal registered on a weekly basis at
Regla, according to Memorias No. 19 - 21 de los Gastos de la Hcda de Regla, for the four
weeks ending on May 26th 1877. Data in italics correspond to inventory make-up.
666
Informe Mensual Regla.
355
My working figures thus err on the high side for the fuel required for amalgamation.
According to my assignments, the total energy requirement for amalgamation during the month
of May 1877 was 634 arrobas (7,291 kg) of ocote and up to 89.5 cargas (12,351 kg) of charcoal
(ocote). To arrive at a ratio based on the energy content of charcoal, using a 10% conversion
rate of wood to charcoal (see Chapter 2), a maximum of 7 kg of charcoal was required for 1 kg
silver produced by amalgamation.
4.5 The mass balance of the smelting process at Regla, 1875 to 1886.
Compared to amalgamation, the solids handling logistics required for the smelting
operations at Regla were on a much more modest scale. Ore, greta and charcoal sum up the
input, while silver, slag and ‘plomo pobre’ [desilverized lead] comprise the output. Though
Regla was the only designated site to carry out smelting of silver ores for the Compañia Real
del Monte, for many periods no smelting was carried out. Thus the smelting data that are
analysed in the following sections start at June 1875 and end by January 1886, which still
represents approximately one decade of nearly continuous operating data registered on a
monthly basis. I follow the same method as for the amalgamation process.
4.5.1 The mass balance for ore
The average quantity of ore received at Regla destined for smelting was 172 cargas per
month (23.7 t) during this decade. As seen in Figure 4-39 there was a peak reception period,
mostly from the Rosario mine, around mid-1877. The gaps in the data reflect the lack of
sufficient information in the monthly account to separate ore for smelting from ore for
amalgamation. On average a monthly inventory of 216 cargas (29.8 t) was carried at Regla,
just over one month of smelting throughput.

356
1,000
900
800
700
600
cargas
500
400
300
200
100
0
Figure 4-39. Monthly deliveries of ore for smelting to Regla (1875-1886). Raw data from the
Informe Mensual.
The close correspondence between the monthly amounts of milled ore and then smelted,
as shown in Figure 4-40, indicates that once it was ground, the over-design of the blast furnaces
would have guaranteed a fast smelting operation.667 The average throughput for grinding and
smelting in both cases is approximately 179 cargas (24.7 t) per month.
The ore smelted at Regla had on average a silver content of 116.6 marks per monton
(1.9% silver) (Figure 4-41). The highest values are of the order of 3.6 % silver. In contrast to
the data presented for amalgamation, the average of silver extracted by smelting as calculated
from the accounting data (net of silver extracted from recycled slags) is also 1.9%. Losses of
silver during refining from the scarce data available are found to be roughly equivalent for
amalgamation and smelting (Section 4.5.3 below). One possible explanation is the diluting
effect of the erosion of the voladora stones for the ore milled prior to amalgamation.
667
The smelting capacity of Regla in mid nineteenth century was up to 30 t of ore per week, which implies that
the whole Furnace Area B was working in the 1870s and 1880s at one fourth of its capacity. As a comparison, in
the roughly same period as analyzed at Regla (1877 to 1891), the Newland blast furnaces in Cumbria (England)
run iron ore smelting campaigns from four to twenty-eight months, at the end of which the furnace was ‘blown
out’, as reported in Bowden, Furness Iron, 50. In the case of Regla the weekly accounts indicate much shorter
runs of days, not months.
357
900
800 Ground ore
700 Smelted ore
600
500
cargas
400
300
200
100
0
2/76 2/77 2/78 2/79 2/80 2/81 2/82 2/83 2/84 2/85
Figure 4-40. Monthly values (1875-1886) of silver ore ground prior to smelting, and the
quantities of ore smelted. Raw data from the Informe Mensual.
7,000
6,000
Weighted average ley
5,000 for smelting 116.6
4,000
cargas
3,000
2,000
1,000
0
60 80 100 120 140 160 180 200 220
ley (marks per monton)
Figure 4-41. Histogram of the silver content of ores destined for smelting at Regla. Raw data
from the Informe Mensual.
358
4.5.2 The mass balance for litharge as input
The ores sent to Regla for smelting did not contain sufficient lead to be smelted on their
own, so litharge had to be added.668 Exactly how much litharge was added per smelting batch
is not reported in the monthly accounting ledger. One visitor to Regla noted that ‘[the ore] is
mixed up [with] a quantity of lead, bearing an almost equal proportion to the other, and a certain
admixture of what is called slag’, where the lead is probably litharge, and the slag he refers to
may have been grasas, or lead slag that still contained silver.669 As in the case of mercury, what
is registered in the ledger is the amount of litharge lost per month.
Regla did not produce lead as a regular commercial product, though it did sell
occasionally what the weekly accounts of the Memorias term ‘poor’ lead, de-silverized lead.
However it did have an incentive to control as much as possible its losses of litharge, which
had to be purchased.670 After the first major charge to start a smelting cycle, new purchases of
litharge would only be required to compensate for losses during the process. The operator
would be expected to control losses of lead in the slag, since it would entrain losses of silver
as well. In case too much silver and lead was retained in the slag, it was re-smelted occasionally
as grasas.
4.5.3 The mass balance for silver: output
In the period June 1875 to January 1886, Regla produced on average 1,717 marks (395
kg) of silver by smelting on a monthly basis. When both types of process were being operated
668
According to Dahlgren, in 1883 the Real del Monte Co. was the owner of lead and iron mines in Zimapan.
Dahlgren, Historic Mines of Mexico, 20. According to Laur they supplied the litharge. Laur, "De la metallurgie
de l'argent au Mexique," 253. Zimapan does not figure as source of the silver ores in the monthly ledger for the
period 1872 to 1888.
669
Tudor, Narrative of a Tour, 298.
670
Even if the Company owned the source of litharge (for example from Zimapan), it still needed to account for
litharge as an expense. In modern terminology the Company was treating each operation as a stand-alone profit
centre.
359
in parallel, smelting would account for 18% of the total average silver production from Regla.
When the whole period from 1872 to 1888 is considered, smelting accounted for approximately
14% of total silver production. Compared to the steadier profile of amalgamation refining, the
smelting output reflects the uneven sourcing of ore for smelting over the period, as can be seen
in Figure 4-42.
6,000
5,000
4,000
marks
3,000
2,000
1,000
0
Dec-78
Dec-85
Jan-76
Jan-83
Oct-77
Jul-79
Oct-84
Nov-81
Jun-75
Aug-76
Aug-83
Apr-81
May-78
Jun-82
May-85
Mar-77
Feb-80
Sep-80
Mar-84
Figure 4-42. Monthly production of silver by smelting (1875-1886). Raw data from the
Informe Mensual.
The fact that Regla oversaw both amalgamation and smelting provides a singular
opportunity to compare the efficiency of each method to extract silver under the operating
conditions of a commercial hacienda. Figure 4-43 compares on the same graph the data on
silver loss from smelting and amalgamation as registered in the monthly ledger at Regla. As
was the case for amalgamation, many silver losses for smelted ores appear in red in the monthly
ledger, indicating the same lack of sufficient skill to sample and assay correctly some of these
ores. Smelting shows a degree of silver losses of 4.7 ± 3.1 %, commensurate with the result for
amalgamation already indicated above, of 5.1 ± 2.8 %. The production data from Regla does
360
not indicate that patio amalgamation was more efficient than smelting in extracting silver from
its ores, for the type of silver ore processed.671
16.0%
14.0%
amalgamation
Unextracted silver
12.0%
smelting
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
Figure 4-43. Comparison of silver losses incurred during smelting and amalgamation (1872-
1888). Raw data from the Informe Mensual.
4.5.4 The mass balance for litharge: output
The monthly average consumption of litharge from June 1875 to January 1886 was of
5.4 t (the ledger reports the data for this period first in pounds and then in arrobas, so I have
converted both to tons). An average of 13.7 kg of litharge was therefore lost for every kilogram
of silver obtained by smelting (Figure 4-44).
50
45
40
35
30
25
kg
20
15
10
5
0
May-75 May-76 May-77 May-78 May-79 May-80 May-81 May-82 May-83 May-84 May-85
Figure 4-44. Weight of litharge lost per 1 kg of silver smelted (1875-1886). Raw data from
the Informe Mensual.
671
In the historiography the contrary is affirmed, as for example: ‘amalgamation was a more efficient method of
separation than smelting’ in Brading, Miners Bourbon Mexico, 156. Since no production data are used to sustain
this conclusion, which in any case depends on the nature of the ore, such generic statements should be treated
with caution.
361
4.5.5 The mass balance for smelting at Regla
The average monthly mass balance for lead can be expressed as:
Lead in silver ore + lead in litharge consumed = total lead in air emissions (la) + lead in
slag (ls) + poor lead sold (p),
Where l denotes lead, subscript a indicates emissions to the atmosphere, subscript s
indicates solid slag and p is poor lead.
The amount of historical lead lost as fugitive losses cannot be quantified so I have
excluded it from the right hand side of the equation. The weight of lead in pure litharge is
92.8%, so if 5.4 t/m of litharge are consumed on average, this amounts to approximately 5 t/m
of lead contributed by litharge. As a working figure I assume that the occasional sales of poor
lead were of the order of 340 lb/m (150 kg/m).672. The equation can now be expressed in t/m
as:
[24.7 × % (lead in silver ore)] + 5 = la + ls + 0.15
la = [24.7 × % (lead in silver ore)] - ls + 4.85
The lead content of the silver ores is not reported in the monthly ledger accounts, nor
the lead content in the slags. I will therefore deduce a working value for these two variables
from a) the amount of non-lead and non-silver content in the ore that I will assume was totally
eliminated in the form of slags and b) ranges of lead content reported for historical slags from
smelting operations of lead ores. This will allow me to establish a first approximation to the
672
The weekly Memorias de Gastos provide two values for the production of poor lead, 280 lbs (127 kg) for the
monthly period that ended the 26th May 1877, and 400 lbs (182 kg) for the monthly period that ended 30 th March
1878.
362
range of monthly losses of lead in air emissions at Regla during the years when smelting was
carried out.
A working figure for the average content of lead in historical slag was deemed to be
3% (Chapter 2, Section 2.6.1). The remaining, non-lead, content of the slag can be
approximated by calculating the monthly amount of mineral in the ore that was not silver,
assuming no lead was present. This is equivalent to:
24.7 t/m of ore x 0.981 = 24.2 t/m
Ignoring losses of the non-lead fraction via volatilization, then the estimated value for
ls is:
ls = 0.03 × 24.2 t/m = 0.7 t/m
This calculation ignores, for want of data, the amount of iron used in the smelting
charge, and it includes by default the unknown amount of lead lost via fugitive losses. 673 The
average monthly loss of lead in air, for ores with no lead content, is given by the following
mass balance:
la = net lead in litharge consumed - deemed loss of lead in slag – sale of poor lead =
4.1 t/m
which corresponds to a 10 to 1 ratio with respect to the average monthly production
rate of silver by smelting of 0.4 t/m. This correlates with the upper level of the range deduced
673
At Regla the source of fugitive emissions would be particulates blown from the inventory piles of litharge and
from work areas around the furnaces. The high walls around the courtyards of Furnace Area B (CB-I and CB-II)
would serve to contain and concentrate these fugitive emissions over time. These fugitive losses would add to the
monthly consumption of litharge registered at Regla. They would reduce both the level of lead lost as slag and the
loss of lead and lead compounds to the atmosphere, but to an extent I cannot at this moment estimate.
363
in Chapter 2, Section 2.6.1, for the ratio of losses of lead and lead compounds to the atmosphere
for every kg of silver refined by smelting.
On this basis Table 4-VI summarizes the overall mass balance for smelting as practised
at Regla from early 1875 to early 1886. For every kg of silver, 59 kg of solid waste would be
generated, which for my simplified analysis I have converted all to waste slag with a deemed
average lead content of 3%. Under this scenario, an ore with zero lead content would produce
a loss of lead and lead compounds in the fumes of 10 kg of total lead per kg of silver refined
by smelting. Any additional lead in the ore (which in Table 4-VI is set as zero) or any loss of
waste ore by volatilization in the blast furnace instead of as slag would increase this baseline
level. On a yearly average, at Regla some 50 t of lead would have been issued to the atmosphere
during this period, and some 300 t of solid waste in the form of slags.
smelting
raw ore net lead in litharge consumed lead in ore total
average input
t/m 24.7 5.0 0 29.7
ground and water air emissions

waste ore as
silver sale of poor lead
slag lead in slag and fugitive
average output lead in fumes
emissions
t/m 0.4 24.3 0.7 4.1 0.15 29.7
raw ore net lead in litharge consumed lead in ore

average input (kg) per kg
refined silver
59 11.8 0 70.4
waste ore as lead in slag and fugitive

average output (kg) per kg silver lead in fumes sale of poor lead
slag emissions
refined silver
1 57.5 1.7 9.8 0.4 70.4
Table 4-VI. Overall mass balance for the smelting of silver ores as practised at Regla
between June 1875 and January 1886. Data was compounded from the sections on smelting of
this chapter. The numbers in italic were calculated and not obtained directly from the
accounting data.
364
The numbers from Regla can be used to derive a factor whereby the existing mounds
of the fields of grasas around a smelting hacienda, which in theory have a mass that can be
estimated, become an indirect guide to the amount of historic amounts of lead and lead
compounds that was issued to the air. According to the numbers in Table 4-VI, an order of
magnitude of the total mass of lead issued from the refining of silver ores can be approximated
by multiplying by a factor of six the estimated weight of extant grasas around the ruins of a
smelting hacienda.
4.5.6 The energy balance of the smelting process at Regla
According to notations made in some Memorias de Gastos, the charcoal employed is
made from encino [oak], the wood of choice both for charcoal and for timbers for the mines.
Ocote is a Mexican pine tree, used both for firewood and to make charcoal. It is important to
note that charcoal for blast furnaces requires certain strength to maintain the physical integrity
of a load of ore and charcoal during smelting. Thus charcoal for a reverberatory furnace need
not have the same strength as charcoal for a blast furnace, and both could come from different
types of wood.
The average ratio over this decade was 204 kg of charcoal per kg of silver (Figure 4-45). A
minor amount of additional fuel would be required to cast the smelted silver into bars, but I
will consider this included in the discussion on amalgamation fuel requirements. This ratio is
one fifth the ratio of approximately 1,000 kg of charcoal per kg of silver reported in the
historiography (see Chapter 2).674
674
A very low ratio of one arroba (11.5 kg) of wood to smelt one quintal (46 kg) of ore is reported for silver
refining in Honduras, but it requires validation. See Newson, "Silver Mining Honduras," 51.
365
600
500
400
300
kg
200
100
0
Jun-75 Jun-76 Jun-77 Jun-78 Jun-79 Jun-80 Jun-81 Jun-82 Jun-83 Jun-84 Jun-85
Figure 4-45. Weight of charcoal required to produce on average 1 kg of silver by smelting

(1875-1876). Raw data from the Informe Mensual.
4.6 The environmental loss vectors in the period 1872 to 1888.
Regla merged the requirements of two completely different silver refining operations
into an integrated whole. The challenge faced by whoever designed the Hacienda de Regla is
illustrated by the mass transit corridors needed to service the two parallel organizations shown
in Figure 4-46. These two completely different refining operations shared a somewhat
constrained physical location, between a stream and the sheer basalt sides of the gorge. There
were no conveyor belts, automatic hoppers or rotary mixers run on steam engines at Regla.
Man, horse and mule power received their only help from the endless supply of water coursing
through the stone channels, but it was still back-breaking work.675
On average every day at Regla over 35 t of materials were entering the compound
through a single gate, and another 35 t were in constant motion between the various process
areas of the Hacienda, some along transit corridors that had to handle two-way traffic. Since
both amalgamation and smelting were batch processes, the monthly averages do not fully
675
Any reader who has had to mix with a shovel just one bag of cement (50 kg) with sand and water will appreciate
the severity of the workload at Regla in preparing and handling one average torta without the aid of a cement
mixer.
366
convey the peaks of internal mass transport or the intense physical activity whenever a torta
was prepared or broken up for washing. The constant backdrop to these bursts of activity was
the periodic arrival and storage of raw ore and materials, and the continuous feeding of ore to
the stamps and arrastres, the jaws of this industrial organism. This animal however had a most
wasteful metabolism, expelling as useless to its well-being over 99.6 % of what it consumed.
~ 1,000 t raw ore

~ 150 t salt, copper sulphate,
mercury, litharge, charcoal, wood
1.7 t silver ~ 1,000 t ore milled

processed in 24 t ore for
for amalgamation smelting
capellinas and smelting
0.4 t silver 5 t litharge

by smelting 80 t charcoal
989 t ore for
amalgamation
2 to 3 tortas,
>200 t, washed 9.4 tortas of
each week ~ 100 t each
direction of flow
Figure 4-46. Main mass transit corridors at Regla, average monthly quantities in the period
1872/73 and 1875/88 (amalgamation) and Jun 1875 to Jan 1886 (smelting).
The mass balances presented in Tables 4-IV and 4-VI are visually summarized in Figure
4-47. Four areas of the environment were impacted by the refining activities at Regla:
a. The stream flowing past the south and eastern perimeter walls of Regla.
The stream to the side of Regla becomes a tributary of the Rio Metztitlán, which in turn
flows until it reaches the natural dam of the Laguna de Metztitlán, approximately 60 km
downstream from Regla (Figure 4-48). The mean annual run-off of the river into the lake in
the twentieth century is estimated at 1.6 × 108 m3. Drainage from the bottom of the lake exits
367
to form the Almolón River, at a level 250m below the lake.676 In terms of mass, the majority
of the waste from Regla would be discharged into this stream. In terms of heavy metals, it
would be the conduit to dispose of around 90% of all mercury losses, mainly in the form of
calomel. The loss of lead via the stream would be minor since there is no textual evidence for
the dressing of ores at Regla.
volatile loss of mercury

from casting of silver
bars : 0.02 t direction of prevailing winds undetermined loss of lead and lead
compounds to the
atmosphere from
smelting : 4 t
Lead in slag and

fugitive emissions: 0.7 t
liquid mercury in
soil and entrained
in wash water: 0.3 t
waste slag, to landfill or

waterways: 24 t
mercury as calomel: 1.8 t
salt, copper compounds: 53 t
direction of flow
mineral silt: 987 t
Figure 4-47. Main loss vectors of waste material, monthly average at Regla in the period
1872/73 and 1875/88 (amalgamation) and Jun 1875 to Jan 1886 (smelting).
676
Max Suter, "A neotectonic-geomorphologic investigation of the prehistoric rock avalanche damming Laguna
de Metztitlán (Hidalgo State, east-central Mexico)," Revista Mexicana de Ciencias Geológicas 21, no. 3 (2004):
397-411.
368
Figure 4-48. The Valley of the Metztitlan River, Hidalgo State, Mexico. Satellite image from
Google Earth Image Landsat. The Lago de Metztitlan is situated at 20° 41’ 04” N 98° 51’ 01”
W. The depth of the valley floor relative to Regla is best appreciated in Figure 4-4.
Every year some 12,000 t of solids would be discharged into the stream, to join the
waste generated by the haciendas upstream of Regla (San Antonio and San Miguel). This was
enough material to have covered to a height of 2.5 m the whole patio area at Regla. In terms of
concentration, just the mercury in the calomel from Regla would represent 0.17% of mercury
(1700 ppm) encapsulated within the fine solid mineral silt from the milled ore. Taking into
account that the river Metztitlán discharges 160 million tons of water into the Laguna
Metztitlán every year, the impact of all these loss vectors would be dissipated as the flow of
water approached the lake. For example, even if all the mercury in calomel reached the lake
(an improbable scenario), it would be present as 0.1 ppm of the yearly flow of water into the
lake.677 Accumulated quantities over an 80 year production span in the nineteenth century
677
Modern studies on the presence of mercury downstream from artisanal centres of gold extraction using
mercury are detecting it as far away as 600 km from the site of mining. Sarah Diringer et al., "Mercury
Biogeochemistry and Artisanal and Small-Scale Gold Mining in Madre de Dios, Peru,"(Poster, 11th International
Conference on Mercury as a Global Pollutant, Edinburgh 2013).
369
would obviously be more significant, but I have not found any studies to date of core samples
either from the river beds or the lake sediments in this area.
b. The soil within the perimeter walls of Regla.
The soil of the patio area, and of the storage and handling areas for mercury would have
been impregnated with mercury that percolated through the tortas and through the wood slabs
that covered the patio. In contrast to the waste continually removed from Regla by the stream,
the soil within Regla would become a depository of accumulated liquid mercury, down to an
unknown depth. How much of this mercury would have been entrained in the skin and clothes
of the workforce and transported to their homes and families is unknown.
Fugitive losses of lead and litharge took place and would also have accumulated within
the perimeter of Regla, but I am not able to quantify this amount.
c. Airborne emissions and deposition of airborne particles in and around Regla.
Air was the domain of lead emissions, not of mercury. Up to 170 times more lead (as
lead and lead compounds) was issued to the air than mercury during the whole period covered
by this chapter. Taking into account that smelting was not carried out during certain years, or
that smelted silver only corresponds to 14% of all the silver produced during the period, this is
a significant ratio.678 From the point of view of the workers, lead fume posed a much more
concentrated threat in time than the constant but much lower level of mercury emissions.
678
A total of 301 t of silver was produced at Regla between 1872 and 1888 by amalgamation and 50 t by smelting.
According to my calculations set out in Tables 4-IV and 4-VI, for every 1 kg of silver refined by amalgamation,
10 g of mercury is lost as volatile matter, while for every kg of silver refined by smelting, a conservative estimate
is that 10 kg of total lead in lead fume is lost to the atmosphere. Over the whole period I project that approximately
3 t of mercury and 500 t of lead were lost as emissions to the atmosphere during the process.
370
Within Regla, high ambient levels of lead and lead compounds would permeate the
areas around the hearths of the blast furnaces and in the load floor during the charging of the
stack. Other areas within and without Regla would be subject to airborne diffusion of lead and
lead compounds in particles, and to the deposition of these airborne particles from the furnace
stacks. The exact footprint of these lead depositions can only be established on the basis of the
prevailing wind direction in the gorge. The isolation of the site and its location within a gorge
would have concentrated the impact of its air emissions to the workforce and local dwellings,
while minimizing it on the surrounding area.
d. Loss of woodland
Smelting was a much greater cause of deforestation than patio amalgamation at Regla.
Even with much more efficient blast furnaces at the end of the nineteenth century, charcoal
consumption was still thirty times greater for smelting than for patio amalgamation, based on
the heating energy requirements to produce 1 kg of silver.679 If the whole period is taken into
account, amalgamation at Regla would have required a total of just over 2,000 t of charcoal,
compared to approximately 10,000 t for smelting. In Chapter 2 I calculated that approximately
0.04 to 0.08 hectares of woodland could supply 200 kg of an average quality charcoal. The
total amount of woodland that would have been deforested in this period in order to supply the
heating requirements for Regla would have been in the range of 2,500 to 5,000 hectares, of
which more than 80% was due to the smelting process.
In 1855 Buchan reported on the woodlands available to the Company at that time:
679
Barrel amalgamation as practised at Regla was much more energy intensive than patio amalgamation, and the
conclusions drawn in this paragraph do not apply to it (see Chapter 5). For patio amalgamation I have calculated
7 kg of charcoal equivalent for every kg of silver refined. In the case of smelting it is approximately 200 kg of
charcoal per kg of silver refined.
371
‘our consumption of wood is not less than 60,000 tons ... per annum ... for the supply of this
fuel we hold ... some 25,000 acres, for the most part forest ... the nearest of these woods have
already, during the last twenty-five years, been much diminished; but we have lately acquired
others ... and with due care of the young trees which are reproducing in those portions already
many years cut, even our nearest forests are not likely soon to fail; while from a distance ... the
supply is inexhaustible’.680
Buchan was referring to the total requirement for timber and wood for charcoal for all
the operations of the Company, and steam engines and barrel amalgamation were major
consumers of energy. Even a deforestation at the lower estimate of 6,000 acres in total over
this period just to supply the needs of Regla represents an important fraction of the total forest
holdings owned by the company in 1855, even assuming reforestation over 20 year cycles.
Regla must have smelled not of riches but of rich animal manure and the sweat of
overworked men and animals, overlaid with the earthy overtones of the dark mud of the silver
ore slurry spread out over the patio. The smelting runs punctuated with their acrid sulphur the
softer background levels of wood smoke from the reverberatory furnaces. Gritty dust must have
coated all surfaces, dust from the stamp mills, from the piles of ore and charcoal and litharge,
dust whirling impotently, imprisoned within those imposing perimeter walls. The basalt
columns of the gorge resonated with the cacophony from the daily pounding of the stamps, the
whirring and scraping of the arrastres, the braying from the mule-trains, the neighing of the
horses and the shouts of the workmen. In the background water gurgled continuously through
channels, splashed from spouts, shoved against water-wheels, lubricated the ground ore, held
together slurries and wetted the tortas, washed away mountains of unwanted waste, out of sight
and out of mind, leaving the grounds of the Hacienda free from the eye-sore of hills of slag or
680
372
useless ore. It is no surprise that the first Conde de Regla chose the less intense Hacienda de
San Miguel as his residence, instead of the fortress at Regla. Over this ‘Babylonia’ of back-
breaking hard labour hung the mistrust and eternal vigilance of its management, overseeing the
production of the small but heavy silver bars with the same obsessive attention to detail that a
miser pays to his hoard. A cat and mouse game of accounting and production versus pilferage
of ore, mercury, amalgam and silver, played every day of the year under the gaze of guards and
overseers, regardless of barred gates and underground storage chambers.
‘Reduction power’ or ‘hacienda power’ are two powerful phrases from the English
business jargon of the mid-nineteenth century that explain the appetite of this industrial beast.
To increase the reduction power of the company was to ramp up the capacity to refine the
greatest quantity of the type of ore that could be produced by the mines. Since the ore came
with so little lead the original intention to rely on smelting became impractical once the plan
to dress the ores was discarded. The only remaining option was to increase the role of
amalgamation, either by the traditional patio process or by importing the Freiberg barrel
process to attempt to process the more difficult ores. The economy of scale behind ‘hacienda
power’ is the reason Regla was operated to its limits throughout the 1872 to 1888 period, with
a brief hiatus in 1874. The environmental impact of this ‘hacienda power’ at Regla can be
quantified thanks to the detailed accounting records that span over ten years continuous
production at the end of the nineteenth century.
The mercury legacy of silver refining at Regla lies dormant underground, as calomel
entombed along river beds or as mercury impregnating the soil within the hacienda, but only a
minor fraction as mercury dispersed long ago in the air. The air at Regla was the domain of
dust from the morteros or the lead fume from the Satanic furnaces. The first would tend to be
confined within the compound of the hacienda, the second would heavily contaminate the work
areas adjacent to the furnaces, or come to ground after having been spewed from chimneys
373
with no long flues or bag houses to protect their immediate surroundings. The English
managers and workers had brought pasties and football from England to Pachuca, but there is
no evidence of an attempt to collect the lead set free in the furnaces. The history of the English
investment in the mining and refining of silver at Pachuca is marked by the omission of the
techniques known at the time to control the loss of lead fumes to the environment.
In terms of sheer weight, it was the volume of solid waste that would have had a major
impact on the environment had it not been washed downstream, away from Regla. The impact
of large amounts of fine mineral silt, calomel and increased salt and copper levels in the water
used by other communities far from the three refining haciendas polluting this stream of water
remains to be studied and quantified. As to the impact on woodlands, all metallurgical activity
created the need for firewood and charcoal. The data from the end of the nineteenth century
show that improvements in furnace efficiency had brought down the consumption of charcoal
for smelting to five times less than the level reported for Pachuca one hundred years earlier.
Even then the forests around the haciendas were being cleared for wood, at a rate that would
have become a major obstacle had more ores been destined for smelting.
How representative is Regla of the overall environmental impact of silver refining in
Mexico? Regla produced a mix of silver from amalgamation and smelting that mirrors the
nineteenth century more than the colonial period, when smelting was more prevalent in New
Spain (Chapter 6). On average the quality of the ore amalgamated was 0.19% silver by weight,
and 1.9% silver by weight for smelting, if anything within the higher range of silver ores in
Mexico. The ores sent for patio amalgamation at Regla were mainly silver sulphide ores, as
were most of the ores of Mexico, with the presence of native silver. There was no help from
silver chlorides in these ores. The consumption of mercury, salt and copper sulphate at Regla
lies on the lower range of the historical scale, but not outside the expected parameters. The
total silver output from Regla during the period covered in this chapter reached a total of 350
374
t, which represents 0.7 % of all the silver refined in New Spain from the sixteenth to the end of
the eighteenth century. The operations at Regla at the end of the nineteenth century were as
chemically representative as any other from the end of the sixteenth century. Even iron, the
additive that first appears in the Vice-Royalty of Peru in the late sixteenth century, continues
to make its presence felt at Regla. Only the efficiency of the process could have improved
during this whole period, as shown by the marked reduction in the use of charcoal for smelting,
or the late switch to faster turn-around times for the tortas in the patio area, but not its chemical
underpinning.
Regla is thus a microcosm that faithfully embodies the practices of historical silver
refining in the New World. This whole thesis is predicated on ignoring the distracting
multiplicity of what is undoubtedly a complex historical scenario, while trying to use basic
chemistry and physics to focus on the metaphorical forest instead of so many trees. Wind
patterns are site-specific, the architecture will change the relative height differential between
chimneys and the perimeter walls, the efficiency of the processes will improve with experience
and design, refining haciendas will be isolated units in the countryside or embedded within
city limits, silver ores come in one hundred different flavours, but the picture that emerges from
Regla transcends these inevitable accidents of time and location. Lead in the lead fume was the
only heavy metal that was discharged to the air in great quantities as a consequence of the
totality of historical silver refining activities in New Spain / Mexico, right up to the end of the
nineteenth century. Calomel, the solid and insoluble chlorine salt of mercury (I), trapped nearly
all the consumption of mercury and was washed downstream from all the amalgamation
haciendas. The immediate danger of mercury lay in its constant contact with the skin, and in
the accumulation of liquid mercury in the soil within the perimeter walls or entrained during
the washing of the slurries.

375
The value of the account books kept for Regla and the other refining haciendas of the
Compañia Real del Monte is not limited to the insights on the mass balance and environmental
impact of the operations. By providing information on the economies of each refining process
they also become the key to understanding the choices open to the refiners of silver in the New
World, choices determined not only by the chemistry but also by the production cost of each
process, the subject of the next chapter.

376
5 The economies of refining silver.
‘nor, unfortunately, is it possible to calculate the costs of amalgamation as against those of

smelting. Even the price of labour is unknowable … costs of upkeep of plant, raw materials …
are similarly an unknown quantity … it is at this point … that lack of knowledge of the costs
of mining becomes a true hindrance to explanation’ Peter J. Bakewell, Silver Mining and
Society in Colonial Mexico: Zacatecas, 1546-1700 (1971)
‘unfortunately at present we possess little information as to the price trends of such

commodities as salt, lead, firewood and copper pyrites …in the last resort until long-term
account books are discovered and studied, any explanation must remain at best, hypothetical’
D. A. Brading, Miners and Merchants in Bourbon Mexico 1763-1810 (1971)
‘precise cost figures rarely show up in archival documents because producers did not keep
accounts that showed all capital and operational costs’ Richard L. Garner and Spiro E.
Stefanou, Economic Growth and Change in Bourbon Mexico (1990)
‘our degree of concrete knowledge on the costs of production is still very limited, since of all
the necessary items required by mines and silver refining haciendas, only mercury has been
the subject of detailed study … for the time being one can only tackle this topic from an
eminently theoretical perspective, devoid of sufficient quantitative fundamentals and extract
just a few provisional conclusions that may serve … as a guide to future research to be realized
from sources that are still virtually unexplored’ Jaime J. Lacueva Muñoz, La plata del Rey y
de sus vasallos : minería y metalurgia en México (Siglos XVI Y XVII) (2010)681
5.1 The most valuable jewel of the Crown
The environmental history of silver refining in New Spain was not only defined by the
chemical nature of its ores, but also by the search for profits by the two groups of human actors
that exercised a choice between amalgamation and smelting. At ground level lay the private
sector involved in mining and refining, driven by the search for short-term substantial profits
to offset the high level of risk of their business. On a higher supervisory and legislative plane
681
‘en cualquier caso, nuestro nivel de conocimientos concretos sobre los costos de explotación es todavía muy
limitado, pues de todos los insumos necesarios para las minas y haciendas de beneficio, solo el mercurio ha sido
objeto de estudio detallado … por el momento solo es posible afrontar el tema desde una perspectiva
eminentemente teórica, desprovista de los suficientes fundamentos cuantitativos y extraer tan solo algunas
conclusiones provisionales que puedan ser … guía para futuras investigaciones realizadas a partir de fuentes
prácticamente inexploradas’
377
lay the Crown officials, having to respond to the pressure of providing cash flows to the Royal
Treasury by seeking to maximize the total fiscal rent from all activities related to silver refining
in New Spain. As a result, macroeconomic fiscal imperatives and microeconomic refining
production costs have to be analysed together with the chemical profile of the silver ores, so as
to understand how the final balance between amalgamation and smelting was arrived at, which
in turn determined the net environmental impact of silver refining in New Spain.
The role of the private sector in the production of silver in the New World is certainly
worthy of note, and is a reflection of the agency they exercised in the final choice between the
two refining options. When the first Conde de Regla died on the 27th November 1781, he left
what has been termed as ‘probably the largest estate of any noble in the colony [New Spain]’,
a sum that may have reached up to 5 million pesos. This included a hoard of 200,000 pesos in
silver coins and bars, an order of magnitude more typically associated with the content of the
royal treasury of New Spain.682 Though the Spanish Crown was the owner of all the minerals
to be found in its territories, it was private individuals who would retain around 80 percent of
the value of all the silver extracted and refined within the Spanish Empire as of the sixteenth
century, a very significant proportion by any standard.683 For nearly 300 years Spain placed the
whole financial risk of refining the silver ores of New Spain (and elsewhere in the New World)
on the pockets of private individuals. This primary risk was not attenuated in a major way by
any secondary support the Crown may have offered, as reflected in the survival rate within the
682
Couturier, The Silver King, 172.
683
Humboldt estimated that 13 to 19% of the value of silver was retained by the Crown. Humboldt, Essai politique,
Tome IV, 144-46.; 78.8% remained to the private silver refiner, according to Brown, History of Mining, 23.;
‘silver taxes accounted for at least 12 percent and mintage fees for another 6 percent of silver’s total value …
transportation and miscellaneous expenses another 3 to 4 percent’ in Garner and Stefanou, Economic Growth
Bourbon Mexico, 115. This large share of revenues left in the private sector is cited as one of the characteristics
of a ‘pro-business institutional framework’ in Rafael Dobado and Gustavo A. Marrero, "The Role of the Spanish
Imperial State in the Mining-led Growth of Bourbon Mexico's Economy," The Economic History Review 64, no.
3 (2011): 862.
378
mining and refining business. The first Conde de Regla was part of the very small group of
hugely successful investors and entrepreneurs who profited from this scheme. A far greater
anonymous mass of Spanish mineros and hacendados participated in the silver lottery of the
New World, where failure and bankruptcy was more common.684 Whatever their size or
success, the mining and refining operations created a novel and dangerous workplace context
for the indigenous workers, their communities and their environment. The extent of this
outstanding social and cultural debt on which the ultimate profitability of the Spanish silver
extraction industry was built remains invisible in the accounts to be analyzed in this chapter.
The Spanish Crown made one significant exception to its policy of allowing the
usufruct of its minerals in return for the payment of a royalty. The mercury mine in Spain at
Almadén would be operated by private parties until the mid-seventeenth century under lease to
the Crown, but throughout the colonial period the Crown retained a strict monopoly (estanco)
on the distribution and sale of mercury in the New World.685 No other raw material required
by the refining processes was subject to this attempt by Spain at a rigid control of both supply
and price, not even the silver ore itself. Why did mercury merit this singular attention? Mercury
was never simply another reagent, on par with salt or lead. Even before it was used to
amalgamate silver ores, the Reyes Católicos had considered the mines of Almadén to be the
684
In the Zacatecas of 1626, after 15 years of ‘unprecedented amounts of silver’, only 4 out of 95 owners of
haciendas and mines were truly rich, according to Bakewell, Silver Mining in Zacatecas, 207. See also "Colonial
Mining," 131. According to Brading, 8 out of 10 miners lost their money. Brading, Miners Bourbon Mexico, 169-
70.
685
The main reference work on Almadén are the two volumes by Antonio Matilla Tascón, Historia de las minas
de Almadén, 1 (desde la época romana hasta el año 1645) (Madrid: Gráficas Osca, 1958).; Historia de las minas
de Almadén, 2 (desde 1646 a 1709) (Madrid: Instituto de Estudios Fiscales, 1987). The other world-scale mercury
mine that belonged to Spain during colonial times was Huancavelica, in the Vice-Royalty of Peru. Huancavelica
would be operated directly by the Crown, but since it supplied a minor fraction of the mercury consumed in New
Spain it will not be included in this chapter. One of the standard works on Huancavelica is by Guillermo Lohmann
Villena, Las minas de Huancavelica en los siglos XVI y XVII (Sevilla: Impr. de la Escuela de Estudios Hispano-
Americanos, 1949). The mine at Idria (see Chapter 1) would also supply mercury to New Spain, mainly in the
eighteenth century. For studies on the mercury monopoly in New Spain see M.F. Lang, El monopolio estatal del
mercurio en el México colonial (1550-1710) trans. R.C. Gómez (Mexico: Fondo de Cultura Económica, 1977).
379
jewel in their crown.686 In 1609 the Viceroy of New Spain, Luis de Velasco, claimed that ‘the
most important business that exists today in the Indies is the trade in mercury, because it is its
main pillar’, evidence of its major role for the trade in silver.687 The concern for mercury
throughout the colonial period was twofold. It was evidently the technical lynchpin of the
amalgamation process, but it was also in its own right a source of additional revenues to the
Treasury, so that its role in the refining of silver had to be protected.
There is some evidence that the Spanish Crown consciously sought to obtain the highest
possible price for mercury that the refiners in New Spain could bear, a kind of pragmatic
sounding out of the elasticity of this novel market. Still, it is difficult to ascertain how the initial
price levels were arrived at. There is an instruction from Joanna of Austria, the Princess
Governess of Spain during the temporal absence of Philip II, who around 1559 issued a Royal
Decree (Real Cédula) stating that ‘we would benefit [on the sale of mercury in the New World]
by earning double of what it costs over here [production cost at Almadén]’.688 Whether this
was simply a rule-of-thumb commercial advice of a timeless nature or an informed decision
thanks to unknown advisors is not clear. The initial pricing levels per quintal in New Spain
were certainly more than a doubling of the cost of production: 215 pesos in 1560, 310 pesos
in 1565 and 1568, dropping to 180 pesos between 1572 and 1591, then to 165 pesos in 1591,
110 pesos in in 1597 until finally settling in for some 150 years at 82 pesos from 1617
onwards.689
686
‘los Reyes Católicos estimaban estas minas [Almadén] “como la joya mas apreciable de la monarquía”’.
Alfredo Menéndez Navarro, Un mundo sin sol, Biblioteca Chronica Nova de Estudios Historicos (Granada:
Universidad de Granada / Universidad de Castilla - La Mancha, 1996), 158.
687
‘el negocio mas importante que existe hoy en las Indias es el comercio del azogue, porque es su principal
sustento’, as quoted in Castillo Martos and Lang, Metales preciosos - union de dos mundos, 139.
688
‘Nos seriamos muy aprovechados y en dicho azogue se ganaría el doble de lo que acá costase’ in Castillo
Martos, "Primeros beneficios amalgamacion," 378. It is not clear if she is referring to a sale price or net profit
equal to double the production cost.
689
There are many sources of mercury prices in New Spain during the colonial period in the historiography. I
have chosen the values reported in Bakewell, Silver Mining in Zacatecas, 172.
380
The highest prices for mercury are registered during the initial years of the introduction
of the amalgamation method to the New World. Comparing the level of New World prices to
the known data sets of production costs incurred at Almadén (Figure 5-1), there is a comfortable
margin between production costs (dots) and the price of mercury (solid black line) in New
Spain until at least 1767, when the first decrease in price from 82 to 61 pesos is implemented,
with the final decrease to 41 pesos decreed in 1776.690 This conclusion remains valid even
taking into account freight costs to the New World (2 to 3 pesos per quintal between 1568 and
1620s) and packing (under 3 pesos per quintal in 1619).691 Prior to the second half of the
eighteenth century, it was only around the 1660s that the margin had shrunk to levels that might
reflect some financial sacrifice on the part of the supplier. Garner considers that the profit on
mercury gained by the Spanish Crown ranged from 100 to 300 %.692 Only in the last 30 years
of this 300 year period could it be argued that mercury was by State design sold at or below its
production cost at Almadén.693
690
Production costs at Huancavelica were much higher, in the 58 to 73 pesos per quintal range. Brown, History
of Mining, 32. However, mercury at Potosi was priced only slightly higher than in New Spain, as shown by the
following values per quintal: 105 pesos in the sixteenth century, 97 pesos in 1645, 79 pesos in 1779, 71 pesos in
1787 and 50 pesos in 1809. Bakewell, "Colonial Mining," 122.
691
Matilla Tascón, Minas de Almadén (to 1645), 98, 221-222. Inland freight in New Spain was 3-4 pesos per
quintal in the 1620s. Bakewell, Silver Mining in Zacatecas, 171. I have not found longer time series for these
costs.
692
Garner and Stefanou, Economic Growth Bourbon Mexico, 138.
693
Matilla Tascón, Minas de Almadén (to 1645), 37-183.; Minas de Almadén (1646-1709), 97-113.
Complimentary information from Bakewell, Silver Mining in Zacatecas, 172.; Lang, Monopolio Estatal, 50, 64.
The end of the eighteenth century also saw the import of mercury from Idria, the ‘German mercury’ that appears
in the accounts of the various Cajas Reales of New Spain, which was priced at the higher level of 62 pesos per
quintal. Garner and Stefanou, Economic Growth Bourbon Mexico, 139. According to Humboldt, Idria mercury
was priced at 52 pesos per quintal in 1784, but still provided a profit of 23% to the Spanish Crown. Humboldt,
Essai politique, Tome IV, 87, 89. This change in Crown policy is one of the consequences of ‘early Bourbon
reformism’ and signalled a decisive change from mercantilist policies on mercury to the proactive fostering of
silver production, according to Dobado and Marrero, "The Role of the Spanish Imperial State in the Mining-led
Growth of Bourbon Mexico's Economy," 869.
381
200
180
production cost
price New Spain
160
140
pesos per quintal
120
100
80
60
40
20
0
1550 1600 1650 1700 1750 1800
Figure 5-1. Production cost and price of mercury in New Spain, plotted from data in footnote
693.
Direct revenues from the sale of mercury were as prized by the Spanish Treasury as
those from taxes on silver: According to Herbert Klein ‘there is little doubt that mercury sales
were the single largest generator of income [from monopoly taxes] in ... Mexico’.694 The level
of that net income, once production costs at Almadén or payments for mercury from Idria have
been accounted for, has not been calculated, but the gross figures are impressive. From 1680
to 1809, the gross revenue from the mercury monopoly in New Spain was approximately 30%
of the total revenues obtained by the Crown from mining activities.695 A similar breakdown is
694
Herbert S. Klein, The American Finances of the Spanish Empire : Royal Income and Expenditures in Colonial
Mexico, Peru, and Bolivia, 1680-1809 (Albuquerque: University of New Mexico Press, 1998), 18.
695
Based on the data in Table 5.2 (Estimated Average Annual Income from Mining Taxes, Vice-Royalty of New
Spain, 1680-1809) and Table 5.3 (Estimated Average Annual Monopoly Tax Incomes, Vice-Royalty of New
Spain, 1680-1809), assuming 30% of the average values in table 5.3 correspond to the sale of mercury (the
percentage of total monopoly revenues corresponding to mercury related revenues is provided by Klein). Ibid.,
19, 80, 86. It is not clear if Klein derives his estimate of 30% from primary sources. An estimate based on an
average price of mercury at 72 pesos during this period, and government duties at 20% on silver indicates that
mercury revenues should have been at least 40% of silver revenues. The difference with Klein’s estimate may be
due to a combination of incomplete records. It has been estimated that even at the peak periods of silver
production, tax revenues from silver of the New World contributed one third of the revenues of the Spanish
Treasury. Mauricio Drelichman and Hans-Joachim Voth, "Institutions and the Resource Curse in Early Modern
Spain," in Institutions and Economic Performance ed. Elhanan Helpman (Cambridge, Mass.: Harvard University
382
observed in the data reported by Mendizábal for Pachuca in the early seventeenth century, with
50,000 pesos in revenues from silver taxes, and 30.000 pesos from the sale of mercury.696 Even
if adjusted by half to take into account production costs and freight to New Spain (inland freight
was charged and accounted for separately within New Spain), the data confirm that mercury
was an important source of revenue to the Crown in its own right.
In the light of the above, what can be deduced on the policy of the Spanish Crown in
setting the prices for mercury in New Spain? The question is most pertinent since mercury is
the economic parameter that has attracted the most attention in the historiography on
production costs of amalgamation (see Section 5.2 below). As argued above, there is no
evidence of mercury being sold consistently below its production cost before the third quarter
of the eighteenth century. On pricing alone, the Crown was able to maintain a positive and
substantial level of revenue thanks to its monopoly on mercury. This took place against two
different operating models at Almadén. During the first century of amalgamation in New Spain
the mercury mines of Almadén mines were leased to private concerns, with production targets
and price set by the Crown in a public tender. After 1646 the mines would be run directly under
the management of the Crown. The following two sections will explore how the Crown
exploited its mercury from Almadén under each regime.
5.1.1 The Spanish Crown ‘ganando indulgencias con escapulario ajeno’.697
Jakob Fugger, the founder of the banking dynasty, had managed to translate the very
risky business of lending money to powerful, unpredictable and profligate royal debtors into a
Press, 2008), 137. This would have made mercury sales the contributor of up to 10% of total revenues for the
Treasury in Spain.
696
Mendizábal, "Minerales de Pachuca," 274. His data are more in line with the expected ratio.
697
‘earning indulgences with another’s scapular’. A Spanish saying with strong Catholic overtones, it refers to
gaining credit (religious indulgence) thanks to the sacrifices made by someone else (i.e. the symbolic commitment
of having to wear a scapular under one’s clothing).
383
huge personal fortune by claiming as collateral two valuable royal assets: land and minerals.
Well before the first Conde de Regla bought his entry to Spanish nobility thanks to the refining
of silver, Jakob Fugger had made a personal fortune from the mining and refining of silver and
copper in Mitteleuropa. Columbus had not yet sailed to the New World when it was already an
accepted practice of the Hapsburgs to raise loans by allowing the lender to take over the mining
and refining of whatever commercial minerals happened to lie within their lands. Unpaid loans
to the Crown came to be considered by the Fuggers as strategic sunk costs in return for net
overall gains, whether as interest payments, minting rights or royal patronage in turbulent times
of defaults by the Crown. As Charles V and then his son Phillip II managed to extract even
greater sums from the heirs of Jakob Fugger, so was the Fugger family drawn into the sphere
of royal assets in Spain, claiming as collateral the revenues from the Maestrazgos and
ultimately the running of the silver mine at Guadalcanal. One of the prime assets of the
Maestrazgos was the mercury mine of Almadén. The Fugger banking house would manage the
mine in a series of periods that started in 1525 and would continue, with some lapses, until
1645, by which time the financially depleted banking family had ceded the production contract
of Almadén back to the Crown.698
Matilla Tascón comments that the first contract to operate Almadén was offered as early
as 1525 ‘in order to compensate him [Anton Fugger], according to what is said, for all that was
owed to him by Charles V’.699 The trail of debts owed by the Spanish sovereigns to the Fuggers
up to mid seventeenth century as reported in the historiography does not differentiate what was
owed due to mercury supplied from Almadén and what corresponded to unpaid capital and
accrued interest. The following snapshots however provide enough information to suggest a
698
Matilla Tascón, Minas de Almadén (to 1645), 37-326.
699
‘a fin de compensarle, según se afirma, de lo mucho que Carlos V le adeudaba’ ibid., 37.
384
financial crash in slow motion. By 1560 the total debt owed to them by the Spanish Crown
reached the sum of 2,975,797 ducats of 375 maravedies (over 4 million pesos).700 In 1582 the
Fuggers failed to produce mercury, and Matilla Tascón conjectures that this may have been
linked to the fact that the Crown could not be trusted to pay them on time. By the end of 1633
the Spanish Crown’s debts to the Fuggers reached over 4.7 million pesos. The representative
of the Fuggers at Almadén complained to the Crown officials that no payments had been
received for shipments of mercury from 1631 to 1633, and only partial ones had been made for
1630. In at least 1639 and 1645 remittances from New Spain derived from the sale of mercury
were diverted to pay pressing obligations of the Crown, not to pay outstanding obligations on
mercury to the Fuggers.701 In the end, the Fuggers could not meet the delivery targets of the
contract that run from 1636 to 1645.702
The leasing of Almadén to the Fuggers by the Crown took place well before silver
changed the order of magnitude of the market for mercury, yet I suggest it would lead to a
singular solution for its needs.703 On the one hand Almadén would serve as a collateral whose
value grew apace both with the need for mercury in the New World and with the increasing
amount of debt owed to the Fuggers. At the same time, part or all of the debt outstanding to the
Fuggers could be considered by the Crown as accounts receivable of the Fuggers, to be borne
in increasing amounts by Almadén, repayable by the Crown at some point in the indeterminate
future. By 1633 the total debt owed by the Crown to the Fuggers was equivalent to roughly
50,000 quintales of mercury at market price, capable of producing 5 million marks (1.2 million
kg) of silver by amalgamation at a correspondencia of 100 marks per quintal. From 1560 to
700
Ibid., 87.
701
Lang, Monopolio Estatal, 79.
702
Bakewell, Silver Mining in Zacatecas, 166-167.
703
There is no similar reading in the historiography nor a reference to documents of the period stating an explicit
policy of the Crown along the lines of my interpretation.
385
1639 a total of approximately 8 million kg of silver were refined by both amalgamation and
smelting in New Spain.704 Assuming that at least one quarter of this total was smelted in this
period, this means that up to 20% of the total silver produced by amalgamation in New Spain
by 1640 could have been due to mercury seemingly gifted by the Crown but ultimately paid
for (as irrecoverable debt) by the Fuggers.705 Since the values of mercury debt reported in the
historiography for New Spain do not reach the level of 4 million pesos, it is probable the Crown
did not even pass on the benefit of the total Fugger debt to the refiners in New Spain.706
The role of the Fuggers throughout this period raises at least two questions: why did
they persevere until they went bankrupt, and why did they not reap the benefit from a textbook
example of a new technology opening up a completely new market for a product only produced
by them under licence from the Crown? To the first there is an article by the German historian
of the Fuggers, Hermann Kellenbenz, which asks through its title whether the rent of the
704
Based on a total of approximately 8 million kg of silver produced in New Spain from 1560 to 1639, according
to Table 3-1 in TePaske and Brown, Gold and Silver, 112.
705
Almadén was the sole source of mercury in New Spain during most of the colonial period.
706
According to Lang, the debt run up on mercury by silver refiners in New Spain had already reached 1,080,000
pesos as early as 1580. Lang, Monopolio Estatal, 361. According to Bakewell the debt in New Spain by 1590 had
increased to 1,828,787 pesos, and would still remain at 1.1 million pesos by 1597. To place these sums in
perspective, new sales of mercury to the provinces of New Spain and Pachuca-Pánuco from 1590 to 1597 totaled
some 2,500,000 pesos. Bakewell, "Notes Mexican Mining," 175-177. By the end of the sixteenth century a
significant portion of the mercury that had been supplied had not been paid for. ‘another of the correct appraisals
by Calderon was to recognize that the [mercury] debt of the miners was irreversible’ – ‘otro de los aciertos de
Calderon fue reparar que la deuda de los mineros era irreversible’, cited from a report from Alfonso Calderon
to the Concejo de Indias, 15 April 1582 as quoted in Cubillo Moreno, "Dominios de la plata," 166. While there is
no study of the average longevity of a mining-refining venture in New Spain during the last decades of the
sixteenth century, it seems probable that not many of the original debtors were still in operation after seven years
or more even if the Crown had decided to forcefully claim the sums owed on mercury: ‘in Mexico, a mining
success or bonanza rarely lasted more than a decade without the need arising for major capital investment’.
Brading, "Mexican Silver Mining," 674. For mining/refining in the absence of bonanzas this period would be
expected to be shortened. Bakewell and Lang report a persistent debt on mercury throughout New Spain during
all the seventeenth century, with. Bakewell, Silver Mining in Zacatecas, 207.; Lang, Monopolio Estatal, 361.
Castillo Martos states that the debt continued to be rolled over, never repaid, and a century later, in 1763, was
approximately 1.12 million pesos for New Spain. Castillo Martos and Lang, Metales preciosos - union de dos
mundos, 54, 145. A similar situation was also observed in the Vice-Royalty of Peru, with a mercury debt of 2.5
million pesos run up in Potosí as reported for the year 1608. Brading, "Mexican Silver Mining," 48. Nearly a
century later the rolling debt remained high: ‘some azogueros never paid for the mercury advanced to them,
creating huge fiscal deficits. In the late seventeenth century, the Potosí refiners owed the treasury more than a
million pesos in mercury debts, most of which the Crown never collected’ in Brown, History of Mining, 23.
386
Maestrazgos was good business for the Fuggers in Spain after 1562. His short answer is yes,
due to payments of interest rates that offered a way of slowly amortizing the outstanding debt
and from collateral business opportunities, such as profits from arbitrage with Spanish silver
minted at Hall (Tyrol) and Venice. Events not analyzed by Hellenbenz however indicate that
ultimately by the mid-1640s the inclusion of Almadén had only aggravated their problem
instead of solving it. The Fuggers had tried to extract themselves from the Spanish Monarchy’s
version of a Ponzi scheme as early as the 1550s. However the lure of fresh future cash flows
offered as bait on each new demand for loans extended their line of credit to the breaking
point.707 As to the brave new world for mercury opened up by amalgamation, there is a certain
irony in the fact that by inadvertently financing the introduction of amalgamation, the Fuggers
became too weak to benefit from it.
5.1.2 The financing of Almadén and the price of mercury in New Spain.
The mercury debt owed by refiners to the Crown from the introduction of amalgamation
in New Spain had been continually rolled over by successive Viceroys until Madrid issued
instructions in 1634 to demand prompt payment from refiners for any new supplies of silver.
This in turn has been interpreted as a watershed in the history of refining in New Spain, since
the new and tougher attitude of the Crown with respect to mercury sales is claimed to have
triggered a major switch from amalgamation to smelting after the 1650s.708 This in turn
introduced a major change in the profile of the environmental footprint left by the refining of
silver. Lacueva places the context of this decision in 1634 to the new financial strategy of the
Duque de Olivares, who since the ascension to the throne of the young King Philip IV in 1621
707
Hermann Kellenbenz, "Los Fugger en España en la época de Felipe II Fue un buen negocio el arrendamiento
de los Maestrazgos después de 1562?," in Dinero y crédito (Siglos XVI al XIX), ed. Alfonso Otazu(Madrid-
Villalba-Segovia: Banco Urquijo 1977), 19-36.
708
Lacueva Muñoz, La plata del Rey, 84.
387
had been implementing new ways to strengthen the collection of royal revenues. 709 However,
the evidence presented in the previous section could also point to the realization by the Crown
that the loans by the Fuggers, that had kept the mercury debt of New Spain rolling over for
nearly one hundred years, were coming to an end.
The return of the mines to the direct control of the Crown did not offer any stability
with regards their operational budget:
‘[in the seventeenth century] the revenues from mercury sales to the Royal Treasuries … were
diverted repeatedly to satisfying urgent needs of the Crown … deprived of a stable budget,
Almadén and Huancavelica survived thanks to financial improvisations’.710
Matilla Tascon’s detailed history of this period highlights persistent problems meeting
the wages of its workforce or new capital investment in infrastructure, and a Royal Treasury
using alternate regional taxes to fund Almadén. Even if the Crown had intended to manage the
mines on the basis of their revenues from New Spain, this revenue never reached them in full
or in a regular manner.711 Variations in this eclectic manner of financing Almadén continue
until 1708, when the oversight of the mine passes from the Real Hacienda to a Junta de
709
Ibid., 79-87.
710
‘[en el siglo 17] los ingresos a las Cajas Reales por la venta de azogue … se desviaron repetidamente hacia
compromisos urgentes de la Corona … privados de su dotación estable, Almadén y Huancavelica sobrevivían
gracias a la improvisación financiera’ Castillo Martos and Lang, Metales preciosos - union de dos mundos, 146.
711
In 1648 its workforce had not received their wages in nearly two years, the total labour debt at the time reaching
21.3 million maravedises (approximately 78,000 pesos). There was no money either to pay for the costs of new
furnaces. The solution was to offer at least food to the workers (wheat, barley, oil) and to request the transfer of
funds from revenues to the Crown (servicio de millones) due from surrounding towns (Alcázar, Villanueva de los
Infantes, Ciudad Real). Ten years later the Kin g was still instructing the Viceroy in New Spain, the Conde de
Alba de Aliste, to send to Spain as a separate remittance the revenues accrued from the sale of mercury in New
Spain, the sum of 25 million maravedies, to cover investment in the mines. From 1650 to 1657, 3,165,814 pesos
worth of mercury was delivered to Veracruz in New Spain, but only 678,887 pesos were remitted to Spain. It
cannot be a surprise that from May 1658 to May 1659 production ceased at Almadén for lack of funds. By 1669
New Spain had still not sent to Spain the sum of 2,265,093 pesos still owed for mercury supplied in previous
years. The consistent failure of this type of instructions, which according to Matilla Tascón were obeyed but never
executed in the time honoured tradition of the authorities of the Hispanic New World, led the president of the Real
Hacienda in Madrid in 1672 to assign specific amounts to be destined to the mines of Almadén from the revenues
owed from the towns and regions of Córdoba, Villanueva de la Serena, Extremadura, Llerena and Trujillo. Matilla
Tascón, Minas de Almadén (1646-1709), 37-39, 117, 121.; Lang, Monopolio Estatal, 80-81.
388
Azogues, which is described by Matilla Tascón as giving ‘palos de ciego’ (‘swings by a blind
man with a stick’, a Spanish phrase that denotes any action taken in desperate ignorance of
what is going on), as they tried to resolve the financing of operations at Almadén. The mine
accounts had been inexistent for the past 5 years.712
The administrators of the operations at Almadén had no direct access to the sales
revenues obtained from its ultimate customers in New Spain. In contrast, the monopoly sales
from tobacco in New Spain were treated as a separate revenue channel that was sent directly
back to Spain, bypassing the option for the local Vice-Royalty to spend it.713 If the budget
destined to maintain Almadén in operation was not sourced from the revenues obtained from
its sales in New Spain, then the pricing of mercury was not determined by the normal sequence
of production costs, capital investment, financing costs, expected return on capital and market
reality (or elasticity). In such circumstances the application of any price level of mercury to its
main market in the New World becomes more a question of maintaining traditional levels
known to be acceptable, both as gross revenue streams to the Royal Treasury and by the needs
of the refiners, to be adjusted only by the inspiration from the authority in place. The previous
sections have underlined the discretionary nature of the price level of mercury for
amalgamation, for which no historical evidence has been found that it was an informed decision
based on the concatenated economies on the supply and production side of the silver refining
chain.
712
Matilla Tascón, Minas de Almadén (1646-1709), 125. Lang argues that the fortunes of Almadén prospered
during the eighteenth century once the Junta de Azogues could follow the interests of the Consejo de Indias and
not of the Royal Treasury. Lang, Monopolio Estatal, 63-96.
713
Klein, American Finances of the Spanish Empire, 85, 94.
389
5.1.3 The environmental impact of the Crown monopoly on mercury.
A geological quirk of fate had placed the world’s major source of mercury within the
mainland of Spain, with enough capacity to supply nearly all the needs of New Spain. The
Spanish Crown always regarded its Almadén mine as a special asset, and industrial scale
amalgamation applied to the huge deposits of silver ore in New Spain raised its role as a
revenue source to the Royal Treasury to unexpected levels. Since the mercury consumed in
New Spain was to all practical purposes only sourced from Almadén, the strict control that
resulted from the Crown monopoly of sales of mercury in New Spain also added the attraction
of attenuating the production of contraband silver. Thanks to the correspondencia benchmark,
the Crown officials could apportion mercury based on the production levels of silver, and not
simply on the word of the refiner. The combination of a centralized source and an easy to apply
operational benchmark made mercury a better option than lead or salt to keep track of the
official production of silver and the corresponding payment of taxes.714
The question is therefore whether its obvious value as a revenue stream to the Treasury
influenced the use of amalgamation in New Spain, and thus altered the course of its
environmental history. Did the Crown take the conscious decision to manipulate the price of
mercury so as to tilt the balance in favour of amalgamation versus smelting, thus guaranteeing
an additional stream of revenues from its monopoly on mercury? The question recognizes that
714
Smelting was a refining operation on the household scale which for certain ores would not even have required
a lead flux, so its control would be limited during colonial times. It could be argued that a rigid State control of
salt would have been easier to implement, simply because its high bulk density would have made it much harder
to traffic by contraband the large amounts required. The Spanish Crown only established a royalty on the
production of salt from the major salt deposits in New Spain. The policy of the Crown in relation to salt in New
Spain is similar to its policy on the usufruct of silver ores: ‘it became obvious the Crown was avoiding all
expenditure, even when, by so doing, some larger income was foregone’. The sources of salt in New Spain were
multiple and ‘on the periphery [where] it proved difficult to impose any strict control’. Local indigenous
production was allowed in certain locations, as well as the farming out of production contracts to Spaniards.
Ewald, Mexican Salt Industry, 19,21. As of the eighteenth century it established the Royal Monopoly of Salt Pans
(Real Monopolio de Salinas). Mendizábal, La mineria mexicana, 79-81.
390
other important factors came into play in determining the balance between amalgamation and
smelting, such as the chemical nature of the ore and the limits on the availability of all the
resources required for each process.715 Even so, the Crown could still have a deciding role, by
actions such as withholding incentives in the search for lead deposits in New Spain, or
guaranteeing the supply of cheaper mercury rather than cheaper charcoal. With no restrictions
on raw materials and attractive pricing, smelting could have processed the widest range of
silver ores in New Spain. The same could not be said of amalgamation.
Two parallel lines relevant to this question run through the primary sources of New
Spain. One is the constant attempt throughout the colonial period in the New World to
safeguard the level of mercury revenues to the Crown.716 The other is the express intention of
high officials to ‘succour’ the refiners, recognizing that without the production of silver there
was no major source of mercury revenues for the Crown.717 The tension between these two
opposing drivers is evident: how far was the Crown willing to sacrifice its mercury revenues
in order to assist refiners? The texts of the eighteenth century (Section 5.3) explicitly state the
choice in mathematical terms, balancing the expected increase in silver taxes to the Crown
against the expected loss of mercury revenues incurred by decreasing its price.
715
Chapter 6 will analyze the historical breakdown between smelting and amalgamation in New Spain and the
influence of these factors.
716
In Chapter 3, Section 3.5, reference was made to the zeal in prosecuting innovators in Potosí who in the late
sixteenth century were searching for a more efficient use of mercury, which was feared to lower the volume of
mercury revenues to the Crown. During the period of the Bourbon revision of mining and refining practices there
was relief that smelting would never come to displace amalgamation, as examined in Francisco Pelayo López,
"Las actividades mineras de J.C. Mutis y Juan José Elhuyar en Nueva Granada," Revista de Indias 50, no. 189
(1990): 462..
717
‘[it] succours the whole body of miners with the necessary mercury, taking note of those who need it and
practising the due diligence to maintain the integrity of this activity’ - ‘se socorre el cuerpo todo de la minería
con el azogue necesario, tomando cuenta a los que deben darla, y practicando las diligencias convenientes al
seguro del ramo’. This is a description of the role of the General Accountant for the Royal Mercury as quoted in
Alejandro Espinosa Pitman, José Antonio de Villaseñor y Sánchez, 1703-1759 (San Luis Potosí, S.L.P., México:
Universidad Autónoma de San Luis Potosí, 2003), 56.
391
Modern historiography has interpreted the pricing policy on mercury by the Spanish
Crown as a subsidy from the sixteenth century onwards. Herbert Klein is one of the several
authors who insists in the importance of the Crown subsidy to mercury. As he phrases it: ‘on
the question of subsidized mercury prices … the role of the crown [was] a major influence in
mine production’.718 Lacueva refers the decision by the Crown to demand prompt payment of
mercury supplied to refiners in New Spain as a key event that ‘signalled the end … of the
indiscriminate subsidy by the Crown … it became necessary to seek in the private sector the
capital required to maintain afloat this sector’.719
It is not clear what is meant by the above use of the term subsidy, a word and concept
that I have not come across in the documents of the period. The historical term subsidy in
England in the seventeenth century meant ‘a tax levied at a given rate (specified in the statute)
on the value of an individual’s movable goods or their income from land (whichever
greater)’.720 This is obviously not the case in New Spain, where the Crown ‘succoured’ the
miners via the supply of mercury at a set price or through the pardoning of outstanding loans.
In its modern economic sense, subsidy is an informed decision, based on a detailed knowledge
of the impact on the production cost structure of the process to be favoured by the subsidy. In
principle some greater national good (strategic, financial) must justify the channelling of
resources to compensate what by itself would be a loss-making or non-competitive business.
‘A subsidy is a form of government support extended to an economic sector … with the aim of
promoting an activity that the government considers beneficial to the economy overall and to
718
Klein, American Finances of the Spanish Empire, 82.
719
‘supuso el fin … de la subvención indiscriminada de la Corona … fue necesario acudir a la iniciativa privada
con el fin de allegar el capital necesario para mantener a flote el sector’. Lacueva Muñoz, La plata del Rey, 84.
720
http://www.nationalarchives.gov.uk/records/research-guides/taxation-before-1689.htm.
392
society at large’.721 It is therefore not evident how the policy of the Spanish Crown with respect
to mercury supplied to its colonial silver refiners meets the modern definition of a subsidy.
First of all, according to the history of mercury prices in New Spain (Figure 5-1), the
highest values of all its market history are registered precisely in the first decades when the
amalgamation process was being implemented in New Spain. This is the stage one expects a
subsidy to be applied, to assist the new technology in gaining a foothold against the much more
traditional process of smelting. Once amalgamation had taken hold, a major incentive to
sacrifice revenues to the Crown from mercury sales disappeared.
Second, as a consequence of its policy of allowing the usufruct of its silver ores, the
Spanish State had no first-hand experience during this period on the day-to-day running of a
silver refining industry by amalgamation. It had not even been able to administer directly the
Guadalcanal silver mine discovered in 1555, transferring its operation to the German financing
family of the Fuggers. Even if it had, the ore produced was fit for smelting, not amalgamation.
Amalgamation processes would only be implemented in Spain during the nineteenth century,
at the silver mine of Hiendelaencina, and even then under the instance of English investors.722
Its knowledge on the economic needs of the amalgamation process was at best second-hand,
based on information supplied by the parties most interested in obtaining cheap
consumables.723 The empirical fact that even at the high end of mercury prices the volume of
721
Norman Kent Jennifer Myers, Perverse Subsidies : How Tax Dollars Can Undercut the Environment and the
Economy (Washington, DC: Island Press, 2000), 5-8. Current definitions of a subsidy can be as varied and
complex as a reader may wish, since modern international trade and tax systems have now made subsidies a
complex legal battleground.
722
In his obituary, a John Taylor born at Holwell, near Tavistock, Cornwall in 1808, son of a Mr. John Taylor, is
mentioned as having been involved in the setting up of the Bella Raquel mining company that used the Freiberg
barrel amalgamation process to process the silver ores from the Hiendelaencina mines. It is most probable the
experience of John Taylor, father, at Real del Monte was being recycled now in Spain. Anonymous, "Memoirs of
members deceased in 1881," Proceedings - Institution of Mechanical Engineers 36 (1882): 14.
723
‘The Crown …did not intervene … in the day to day technical decisions of production, nor in their financing’
-‘La Corona … no intervino … en las decisiones técnicas que marcaban el día a día de la explotación, ni en su
financiación’. Julio Sánchez Gómez and Guillermo Mira Delli-Zotti, "Minería americana y minería europea,
393
New World silver flooded over world markets, overwhelming at the beginning the European
silver industry, gave the pragmatic answer to the Spanish officials that the mercury price being
applied was enough to insure continued production of silver, an argument that would hold until
the middle of the eighteenth century. The only quantitative production parameter that can be
claimed on historical evidence to have been available from the sixteenth century to the Spanish
authorities was the correspondencia. At a price of around 120 pesos per quintal of mercury in
the late sixteenth century, the cost of mercury would correspond initially to approximately 15%
of the value of silver produced by amalgamation.724 Had a subsidy been intended, a much lower
pricing range for mercury would have been adopted.
Third, accepting the premise that a flexible interpretation of subsidy would include not
only a price drop but also the non-payment of outstanding loans on supply, up to 1646 events
have revealed that if on the one hand the Crown in Spain was not receiving all the revenue
from the sale of mercury, neither was the Crown paying the Fuggers for all the mercury that
was being distributed in New Spain. German private capital had provided what Lacueva termed
as the ‘indiscriminate subsidy’ up to 1646. A forced largesse of this nature cannot be interpreted
as a subsidy to the refiners by the Spanish Crown, but rather the result of an opportunistic
financial play that evolved in time, contingent on unfolding events on either side of the
Atlantic.725
1750-1820: una perspectiva comparada," in Mundialización de la ciencia y cultura nacional: Actas del Congreso
Internacional "Ciencia, Descubrimiento y Mundo Colonial", ed. A. Lafuente, et al.(Aranjuez (Madrid): Doce
Calles, 1993), 107.
724
One kg of silver is equivalent to 4.4 marks, or approximately 35 pesos in value. Two kg of mercury are 0.043
quintales, which at 120 pesos per quintal represent 5 pesos in value. This represents some 15% of the total value
of silver refined. As the price of mercury dropped to 82 pesos, the cost of mercury as a percentage of the value of
silver refined would decrease to around 10%.
725
The Spanish Crown did offer its indirect support to miners and refiners in the New World, though the local
elites would have been expected to reap the most from this assistance (preferential supply of mercury to major
refining concerns, dispensations from the payment of royalties, and other major tweaks to the system). The fact
that the English investors in the region around Pachuca as late as the nineteenth century had to invest major
capital simply building roads is an indication of the limited extent of the Crown’s involvement in the general
394
Fourth, it is only after 1776 that the price of mercury in New Spain is seen to drop to
the level of its production cost, or even below, the first clear sign that one of the conditions of
a subsidy has been met. After 1776 the sacrifice of mercury revenues to the Crown had to be
justified not only by a significant increase in silver production but also by an efficient use of
the cheaper mercury, the latter a condition that will be further explored in Chapter 6.
Underlying the question as to whether a subsidy ever existed with regards to mercury
pricing set by the Crown in New Spain are two more fundamental issues. First, a subsidy is
applied by decreasing the cost to the producer of the input that has the greater impact in
determining the final price of the product, in order to allow it to compete in the market. The
price of silver in the period of interest was independent of the price of mercury, or of any of its
manufacturing inputs by either process, since it had remained fixed over centuries by other
economic conventions (Section 5.4). A decrease in the price of mercury by the Crown sought
completely different objectives. It increased the incentives for private individuals to assume
the business risk of producing silver, it increased the production of silver by increasing the
amount of ores that could now be profitable to refine by amalgamation, and it affected the
overall balance between amalgamation and smelting so as to keep the former in the
infrastructure required by its silver industry. The extent to which the ordinary miner-refiner benefitted from the
assistance of the Spanish Crown can be better judged in relation to what the German miners still received from
their King in the region of the High Harz at the beginning of the nineteenth century. The King had the right to
retain one tenth of the value of the minerals extracted (silver, lead, copper); to have a share in each mine, around
3%; to build all the common infrastructure required by mining and refining (water reservoirs, galleries to discharge
water from mines, for crushing the ore, and smelters) and to operate the washing and smelting of ores exclusively,
all for a fee payable by each mine. Finally, the Crown bought exclusively the lead and copper at a price lower
than market prices, and profited from the difference (silver was coined at the official rate). On the other hand, the
King had the obligation to provide all the wood required by mining and refining at no cost except that of cutting
and freight; all other industrial consumables were sold at a regulated price, at times below market; the chief
officials for mining, smelting and forests were paid by the Crown, all other wages by each mine; the Crown would
provide a fixed amount of cereal (rye) at a low price to each worker; the Crown would partially compensate the
mine for unexpected increases in certain materials if it could prove unable to meet the new prices; it would also
compensate for certain increases in the price of oats required for the animals used in mining; injured or sick
workers were cared for, and a fund established for widows. There is a major difference between the above and
providing a legal framework for the claims of mines, selling mercury above its cost of production, and providing
escorts for the shipments of silver. Héron de Villefosse, De la richesse minérale du Royaume de Westphalie, 89-
94.
395
ascendancy.726 Had the Crown decided to supply lead or charcoal below their production costs,
the balance would have tilted towards smelting.727 At the same time, the only reason that could
convince the Crown to sacrifice its revenues on mercury was to compensate with an increase
in the royalties it received from silver production and the concurrent increase in the monopoly
sale of mercury.
It has been an implicit assumption in all the previous discussion in the historiography
that the price of mercury was the defining economic factor that determined the competition
between smelting and amalgamation, or even the final production levels of silver in New Spain.
However, there is a dearth of hard economic data on production costs for amalgamation and
smelting in the historiography to sustain this assumption, as will be made evident by the review
presented in the following section.
5.2 The historiography on the economies of refining.
The historians of silver refining in the New World have been in the same quandary as
the Spanish Crown. They know how much silver was registered, how much mercury was sold
and at what price, but are quite in the dark as to the exact production costs involved in the
amalgamation or smelting of silver ores in the New World. Even mercury, the most studied of
all consumables of this period, was left at the end of the previous section with major questions
pending regarding its exact economic impact on the production costs of amalgamation and its
role in determining production levels of silver in New Spain. The quotations cited at the
beginning of this chapter show how the foremost English and Spanish historians of colonial
726
Dobado and Marrero argue that ‘it was in the interest of the Crown that amalgamation should be the preferred
technique for silver refining’, in Dobado and Marrero, "The Role of the Spanish Imperial State in the Mining-led
Growth of Bourbon Mexico's Economy," 866.
727
The balance in refining processes will be treated in Chapter 6. See also the analysis of the shifting balance
between amalgamation and smelting and the influence of mercury and lead costs and availability in Blanchard,
Russia's "Age of Silver". Precious-metal Production and Economic Growth in the Eighteenth Century 3-31.
396
silver refining have found a landscape quite devoid of hard economic data on the production
of silver from ores in the New World.728
The first historical attempt at an economic analysis of the production costs of
amalgamation for New Spain was written in the mid-eighteenth century.729 It is a discussion in
print between the Accountant General of the Royal Mercury (Contador General de los Reales
Azogues) José Antonio de Villaseñor y Sanchez and the Overseer of the Royal Mint
(Guardavista de la Casa de la Moneda), Jose Antonio Fabry.730 It takes place between 1741
and 1743, and contains as its central theme the two main arguments used at this time in favour
and against a decrease in the rice of mercury, which had been held at 82 pesos per quintal for
nearly 150 years. Villaseñor is defending the revenues to the Crown from the sale of mercury.
He grounds his analysis on the fact that it costs the same to amalgamate an ore irrespective of
its content of silver, except for the variable cost of mercury. On this basis he argues that even
at the price of 82 pesos per quintal, the refiner could still make a profit with ores of 2 oz of
silver per quintal (0.13% silver content by weight). Since the expense for mercury is
728
Bakewell, Silver Mining in Zacatecas, 187, 207.; Brading, Miners Bourbon Mexico, 154, 158.; Garner and
Stefanou, Economic Growth Bourbon Mexico, 118.; Lacueva Muñoz, La plata del Rey, 58. Other historians have
voiced similar concern over the dearth of hard economic data on refining by amalgamation or smelting in the New
World: ‘only minimal documentary evidence has been found’ - ‘no se han encontrado mas que minimas
referencias documentales’ Cubillo Moreno, "Dominios de la plata," 30.; ‘the documents on haciendas do not
mention the workforce they employed … and it seems the account books of these production units have
disappeared … no account books or other documents are known that registered specifically the levels of
production’ - ‘los documentos sobre haciendas no hablan de la mano de obra empleada en las mismas … y al
parecer los libros de cuentas de estas unidades de producción han desaparecido … no se conocen libros de cuenta
u otros documentos que anoten específicamente los niveles de producción.’. Lara Meza, Haciendas de beneficio
de Guanajuato, 102, 105. I have not found in my own more modest searches within the archives of San Luis
Potosí, Zacatecas or Guanajuato any account books that allow a reconstruction of production costs over time. The
snippet of data available do not match the amount of detailed quantitative information on prices, consumption,
costs, labour and ore quality as I have found for Regla in the second half of the nineteenth century.
729
In Chapter Three I have made reference to the anonymous document that includes annual costs incurred for
amalgamation in Potosí in 1601, but there is no comparable source at present for New Spain during the first two
hundred years.
730
Fabry, Impugnacion a reflexiones de Villaseñor. Jose Antonio de Villaseñor y Sanchez (1703-1759) is much
better known as the author of a detailed description of the main cities and provinces of New Spain, published in
two volumes as the Teatro Americano. For a biography see Espinosa Pitman, Villaseñor y Sánchez. He was the
senior official in New Spain in charge of accounting for the distribution of mercury under the Crown monopoly.
397
proportional to the silver content, ores with lower silver content would only incur a small cost
for mercury, and yet have to meet the much greater relative cost of all other expenses out of an
increasingly smaller monetary value of silver content. He concludes it would not make sense
to refine these ores even if mercury was given away for free. In case the cost of extraction could
be circumvented by using ores found discarded in tailings, he calculates that even at 82 pesos
for mercury the refiner could make a profit working with ores that had as low as 0.06% silver
content (see Table 5-I). By placing the emphasis on the influence of the extraction cost of the
ore on the final profit level of the refiner, he makes an extremely pertinent observation:
‘what makes mines [and refining] unprofitable is the labour-intensive drainage, the tough
nature of the ore wall, the timbering of the supports, the cutting and removal of the ores, the
wages of the workers, the need of food, forage for the animals powering the drainage … even
if they gave away lakes of mercury, these costs would never be met [for the low silver ores] …
what limits silver mining is the lack of this ingredient [mercury] not its price’.731
What is impressive about Villaseñor’s line of argument is not so much its mathematical
conclusions, which as always depend on the validity of the starting values, but because he
applies an elegant and concise method for analysing the sensitivity of refining production costs
to the silver content of the ore. He freezes as a virtual fixed cost all the variable costs of
production so they remain at the same level irrespective of the silver content of the ore. The
only exception is the cost of mercury, which he varies as a function of silver content (see Table
5-I). He includes the extraction cost of the ore in order to calculate the level of profit for the
refiner, an approach rarely seen in later exercises probably for lack of data. He applies the same
extraction costs regardless of the richness of the ore, which is correct for the range of ores he
includes in his working examples. He differentiates the ores from tailings by assuming the
731
‘lo que hace incosteable las Minas son los laboriosos desagües, la dureza de sus frontones, el echado de sus
respaldos, corta-saca de sus metales, carestía de operarios, necesidad de alimentos, y pastos para las bestias de
los desagües … aunque se diera en lagos el Azogue, nunca se costearían … hace escasa la minería de plata la
falta de ingrediente, no su precio’. Fabry, Impugnacion a reflexiones de Villaseñor, 5.
398
extraction cost of the latter is by now a sunk cost, thus nil. He recognizes that the higher expense
of mercury for ores with higher silver content is compensated by the increase in value of the
silver refined. All that is missing from a perfect score is the fixed capital cost to the refiner and
its impact on the break-even pricing for mercury.732
ore source mine tailings (terreros)
% silver (weight) % (weight) 0.5% 0.25% 0.13% 0.06% 0.06%
silver content per quintal oz / quintal 8 4 2 1 1
amount of silver in 100

marks 100 50 25 13 13
quintales
value of silver in 100

quintales @ 8 pesos 6 pesos 873 437 218 109 109
tomines per mark
mercury consumed pesos 82 41 21 10 10
other production costs,

including extraction, @ 1.5 pesos 150 150 150 150
pesos / quintal
only other amalgamation

costs @ 0.5 pesos per pesos 50
quintal
total variable production

pesos 232 191 171 160 60
costs
profit pesos 641 246 48 -51 49
Table 5-I. Interpretation of Villaseñor’s working examples and method that sustained his
argument against decreasing the price of mercury. Data adapted from footnote 730.
Fabry countered by arguing that in reality a greater portion of ores lie below the
breakeven point calculated by Villaseñor at a mercury price of 82 pesos per quintal. By
questioning the values adopted by Villaseñor he concludes that at 41 pesos per quintal it would
be possible to make profitable to the refiners the amalgamation of the abundant ores with a low
732
He also provides the historian with a guide as to an order of magnitude of extraction costs in mid eighteenth
century (1 peso per quintal), as well as that of amalgamation costs net of mercury (0.5 pesos per quintal). Ibid., 1-
5.
399
silver content. He then calculates that the increase in tax revenues from the additional
production of silver would more than compensate the Crown for the decrease in the revenues
from mercury. Fabry’s line of reasoning highlights the distinction made at the end of Section
5.1.3 that the aim of reducing the price of mercury was not to reduce the price of silver but to
increase its total output so as to compensate for the loss of mercury revenues.733
A very similar argument to that of Fabry is used in a 1774 report addressed to the
Spanish King, Charles IV, by the Guanajuato medical doctor and miner Manuel Jose
Dominguez de la Fuente. The reduction in the pricing of mercury from 82 to 62 pesos per
quintal had not pacified the miners and refiners, who sensed that the Bourbon King was willing
to decrease the price even further, since he was asking for their official view on what should
be the ultimate lowest price of mercury. It is in this light that any economic breakdown
provided by Dominguez de la Fuente should be read: it is an exercise in lobbying the Crown in
favour of lower mercury prices, not an accounting book.734 He argues for a lower price for
mercury since this would allow miners to market the majority of the ore extracted, that at the
time was being rejected as unprofitable by the refiners.735 Villaseñor would have countered that
733
Fabry questions both the monetary value of one mark silver (8 pesos 6 tomines instead of 7 pesos 5 tomines )
and the cost of extraction employed by Villaseñor. Ibid., 6-36.
734
As a miner of modest means he faced the separation implemented in the mining/refining model by the
eighteenth century, whereby those who only refined the ore by tolling (maquila) in haciendas took most of the
profit and left the risk to those who only extracted the ore: ‘the Miner is he who works; and the Refiner he who
benefits’ - ‘el Minero es el de el trabajo; y el Haziendero el beneficiado’,a pun on the refining (beneficio) of ores.
He claimed that the owners of refining haciendas purchased the ores from miners ‘under very questionable
assumptions [as to deemed silver content, in the absence of assaying] … the Miner leaves with his doubt and
mistrust and the Buyer keeps his reserve [the level of underestimation of the silver content] and his cautious
[approach]’- ‘bajo de mui falibles conjeturas … el Minero se va con su duda y desconfianza, y el Comprador
queda con su Reserva y cautela’. The value of silver extracted was paid after two and a half months in Guanajuato,
and the cost of the maquila was considered ‘very burdensome and because of that very disheartening’ - ‘mui
gravosissima y por esso de grande desconzuelo’. Since the refiners could only pass on to the miners the official
cost of mercury as part of the toll charge, the benefits in the decrease in price would reach them in two ways, by
increasing the amount of ore they could sell and by decreasing the cost of maquila. It is not known if it ever
reached the Royal Court. The paleography of the original manuscript by Dominguez de la Fuente was carried out
by its present owner, Salvador Covarrubias Alcocer, and published as Dominguez de la Fuente, Leal Informe
Politico-Legal, 102, 104, 199.
735
According to the author the most common ores had a silver content of 3 to 5 marks per montón of 32 quintales
I have inferred he refers to marks per monton, since marks per carga would indicate too rich an ore not to be
400
the price of ore proposed by Dominguez de la Fuente at 4 pesos per montón does not reflect
his real extraction cost estimated at 32 pesos per montón, in which case it still remained a
question of throwing good money after bad. Dominguez de la Fuente could have replied that
the lower quality ore was part of the normal output of a mine, so any price for his ore of 4
marks per montón silver content was better than throwing it away to the tailings heap, and the
Crown would still compensate the drop in mercury price by selling more mercury to
amalgamate the new quantities of ore coming to the refiners instead of being discarded.736
The next group of extracts from the historiography of the late eighteenth and early
nineteenth century correspond to the period of the lowest historical prices for mercury in New
Spain, thus the texts no longer discuss the issue of mercury revenues vs price. Production costs
are reported but with no details on how they are calculated. I have summarized the main
purchased. At the time refiners in Guanajuato were not buying ores for amalgamation with less than 5 marks of
silver (0.083%) per montón. According to his data the custom of refiners was to purchase ore (recibir de los
Mineros a la Ley) at a price of 11.6 pesos per kg of silver content. He proposes that refining centres (Haziendas
Refaccionarias) be set up to work expressly with the large amounts of ore available with just 4 marks per
Guanajuato montón (0.06 % silver), to be sold at a price from 3 to 8 reales per carga. The refiner would still
obtain a profit, since the lower price of mercury would compensate for the purchase of the ore with a lower silver
content His quantitative argument is not easy to follow, and the following interpretation is tentative: he uses
virtually the same variable cost of amalgamation (without mercury) as Villaseñor, at 16 pesos 1 real per monton;
he proposes that refiners can purchase ore at 3 reales per carga with a silver content of 4 marks per monton, or 4
pesos per monton. This adds up to 20 pesos for the cost of amalgamation of 1 monton, without mercury. He then
arrives at a total cost of 22 pesos 2 reales per monton, from which I deduce he is using a cost of 2 pesos 2 reales
for the cost of mercury consumed, which at a correspondencia (not given) of 100 marks per quintal of mercury,
implies a new lower price for mercury of approximately 53 pesos per quintal. Ibid., 90-91.
736
The compensation in direct revenues to the Crown comes from the opportunity cost of both additional silver
production and mercury sales that would not otherwise have been possible at the higher price for mercury, if
Villaseñor’s arguments are discounted for now. Thus assuming a correspondencia of 100 marks per quintal, a
deemed value of 8 pesos per mark of silver and direct government revenues of 20% of the silver produced, a drop
from 62 pesos to 41 pesos per quintal requires that total revenues of 222 pesos be maintained from a new level of
(1+x) quintales of mercury at 41 pesos per quintal and 0.2*100(1+x) marks of silver at 8 pesos per mark, which
gives a value of x of 0.1. This means that at least a 10% increase in mercury sales was required just for the Crown
to breakeven in direct revenues. Based on an ore with 4 marks per monton, this would require amalgamating 2.5
new montones for every 10 montones of ore previously amalgamated (assuming an average of 10 marks of silver
per monton, 0.16%), approximately a 25% increase in ore purchased by the refining haciendas, an obvious
windfall for miners. Since the alternative was to increase the amount of value-less tailings, from the view-point
of an opportunity cost Dominguez de la Fuente had a very valid point, finding the weakness in Villaseñor’s
arguments.
401
economic data that can be abstracted from all the texts included in this section in Table 5-II.737
The background of each source is relevant. Garcés y Eguía is promoting his smelting recipe
based on the claimed benefits of adding tequesquite, a naturally occurring sodium carbonate
salt.738 Humboldt provides one table on ore extraction cost, sundry information on mercury
prices, and no details on production costs.739 An unexpected ratio comes out of his figures on
labour for these mines. On the basis of the number of workers employed underground in the
mines, at Valenciana the ratio is 400 quintales of ore extracted, 200 marcs of silver produced
per worker, while in Germany it is 25 quintales of ore extracted, 18 marcs of silver per worker.
The German mine is less deep than the Valenciana mine (330 m to 514 m according to
Humboldt) and its ore nearly twice as rich in silver, yet Humboldt does not comment on the
disparity in the production output per underground worker between these two mines. It
indicates a manpower efficiency an order of magnitude greater in Guanajuato than for a mine
chosen from the most traditional mining area of the Erzgebirge in Europe.740
737
a) Fabry, Impugnacion a reflexiones de Villaseñor.b) Dominguez de la Fuente, Leal Informe Politico-Legal.
c) de Sarria, Ensayo de metalurgia. d) Garcés y Eguía, Nueva teórica del beneficio de plata. e) Humboldt, Essai
politique. f) Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España. g) Duport, Métaux
précieux au Mexique. h) Laur, "De la metallurgie de l'argent au Mexique." i) Collins, Metallurgy of Lead & Silver.
738
In his examples he works on the premise that it is possible to partition by washing on the planillones 100
quintales of an ore mixture with an average silver content of 0.25% into 90 quintales with a silver content of
0.14% (my calculation) that is treated by amalgamation, and 10 quintales with a silver content of 2.7% that is
smelted. It is on this unexpected premise that he argues for the benefits of his smelting process in saving expenses
on mercury, to the deemed benefit of the refiner. The higher the silver content prior to partition, the greater the
saving on mercury, the higher the benefit to the refiner. Garcés y Eguía, Nueva teórica del beneficio de plata, 144-
146. Sodium carbonate was a known agent of flux for smelting much prior to this publication, for example the
use of soda as a flux in Agricola, De re metallica. Footnote pages 233, 558. His offhand reference to Sarria’s
earlier text which already mentions the use of tequesquite may be an effort to distract attention from the fact this
additive was already known. Garcés y Eguía, Nueva teórica del beneficio de plata, 81. At Regla tequesquite would
be tried for a few months and then discarded.
739
Humboldt compared the extraction costs for the mines of Valenciana (Guanajuato) and Himmelfürst (near
Freiberg, Germany) in an average year at the end of the eighteenth century. At Valenciana his data leads to an
extraction cost of 6.9 livres tournois (1.3 pesos) per quintal, or 60.4 livres tournois (11.5 pesos) per kg of silver
refined. For the European mine, it comes to an extraction cost of 17.1 livres tournois (3.3 pesos) per quintal of
ore, 104.3 livres tournois (19.9 pesos) per kg of silver refined. Humboldt, Essai politique, Tome III, 413.
740
For Valenciana he reports an ore that had on average 4 oz (0.5 marks) of silver per quintal, so that 720,000
quintales of ore from said mine produced 360,000 marks of silver by smelting or amalgamation; for Freiberg the
ore has an average of 6 to 7 oz per quintal of silver, and from a production of 14,000 quintales a total of 10,000
marks of silver are refined. Ibid. The exchange rate of pesos (piastres) per livre tournois is calculated from
Humboldt’s assertion that 805 million livre tournois are equal to 153,333,000 piastres. Ibid., Tome IV, 255.
402
amalgamation costs smelting costs

extraction
silver content
location cost period
excluding cost of mercury,
including mercury, excluding
excluding cost of
source
cost of extraction of ore, extraction of ore,
extraction cost of ore
capital cost capital costs
peso /quintal
pesos / quintal ore pesos / mark silver pesos / quintal ore %
ore
generic 1 mid 18c 0.5 0.125 to 0.5 a, 1-4

Guanajuato ? 0.5 0.08 - 0.13 b, 90-91
generic ? 0.5 - 0.6 3-5 ? c, 131
generic ? 0.625 0.25
d, 144-146
generic ? end 18c 4 2.7
Valenciana mine,
1.3 0.25 e, 255
Guanajuato
generic ? 0.5 - 0.75 low to medium f, 92

Zacatecas ? 5.3 - 6.7 'rich minerals' g, 83-84
Sombrerete ? 2.7 - 3.3 'rich minerals' g, 85
Nieves ? 1.3 - 2.7 ? g, 85
Zacualpan ? 5 ? g, 86
? 0.63 ? g, 232
? 2.15 (maquila ) 0.23 g, 233

Guanajuato
? 3.3 (maquila ) 0.19 g, 233-234
? 4.8 (maquila ) 0.08 g, 235
4.1 2.55 (maquila ) 0.46 g, 252

Zacatecas mid 19c
? 3.7 (maquila ) 0.14 g, 275
Tasco 1.08 3.3 (maquila ) 0.15 g, 340-341
generic < 1.68 2.3 (maquila ) 0.2 g, 370-374
Cerro San Pedro,

? 1 up to 0.3 (35% lead) h, 246-248
San Luis Potosi
up to 1.75, medium to
Sombrerete ? 2.4 h, 248-250
low lead
0.15 to 0.9; max 10%
Catorce ? 7 h, 251-252
galena
Regla ? 6.1 1.7; poor in lead h, 253
Durango ? 0.90 1.15 0.18 i, 60
Zacatecas ? 0.32 0.39 0.07 i, 60
Zacatecas ? 0.33 0.39 0.05 i, 61

end 19c
Fresnillo ? 0.46 0.51 0.06 i, 61
Tasco ? 0.41 0.72 0.32 i, 61
Pachuca ? 0.41 0.73 0.30 i, 61
Table 5-II. Summary of amalgamation and smelting costs in New Spain/Mexico from the
historiography up to the nineteenth century. Figures in italics are calculated from source data.
Sources are indicated in footnote 737.
403
Sonneschmidt’s monograph was published late, in 1825, though he should be assigned
to the group of authors publishing around the turn of the century. He became an avowed
disciple of the Mexican patio amalgamation method, but provides no support to his estimates:
‘the amalgamation of New Spain … will subsist as long as the world subsists … the greatest
advantage [to amalgamation] … is its low cost … ores with low or medium silver content, can
be refined at the moderate cost of four to six reales per quintal … there is no cheaper refining
cost even in Europe’.741
He considers production costs to be the major obstacle for smelting to compete with
amalgamation, an important if qualitative conclusion from a first-hand observer and
practitioner of refining in New Spain. He also recognizes the critical distinction between total
mercury cost and the cost of mercury per unit of silver extracted: ‘only in the refining [by
amalgamation] of rich minerals does the cost rise, due to the greater consumption and loss of
mercury, which is always proportional to the amount of silver extracted’.742
Finally, to complement the authors of the turn of the century, there is a primary
document from 1802 that due to the scarcity of sources of this nature I have reproduced in full
in Appendix D.743 It is a narrative that throws further light into the world of Dominguez de la
Fuente. Middlemen now appear who made a living without necessarily investing in fixed
capital: the rescatadores who bought the ores at the mine mouth, refined it themselves and/or
741
‘la amalgamación de Nueva España subsistirá mientras que subsista el mundo … la major ventaja … es el
poco costo … minerales pobres y de mediana ley, se benefician con los moderados costos de cuatro o seis reales
por cada un quintal … beneficio más barato no lo hay ni en Europa’. Sonneschmidt and de Fagoaga, Tratado de
la amalgamación de Nueva España, 92.
742
‘solo en el beneficio de minerales ricos suben los costos, por motivo del mayor consumo y perdida de azogue,
que esta siempre proporcionada con la ley de plata que se extrae’, ibid.
743
It is preceded in the archive by a letter dated 21 April 1802 from a delegation of Mining Deputies from Catorce
(San Luis Potosí) informing that they have carried out the request by the Viceroy of New Spain dated 31 March
1802 to carry out a trial comparing the production costs of refining via the cazo process with smelting, using the
same quality of ore for each process. Since the results reported in the letter do not correspond to the details in the
report dated 21 April 1802 in Appendix D, with different signatories to each document, it would seem at least
two trials were carried out, of which only one set of detailed results has survived in the archival records.
404
by maquila, and sold the refined silver at a discount to the aviadores (suppliers of goods on
credit, also applies to the leasing of mines). The document is not an account book and its data
are limited to single trial runs on small quantities. Even with these limitations it still reveals
the nuances that apply to silver refining in New Spain. For example, even though amalgamated
silver is said to achieve a higher price than smelted silver (Chapter 2), in this report it is the
other way around. As to the processes involved, the document reveals the reality of the cazo
process as practised in Catorce in the 1800s. Since the cazo process only extracted two-thirds
of the silver content, in order to make it more profitable it was then necessary to treat the fine
silt that came out of the vessels (lamas cocidas) by the conventional patio amalgamation to
complete the extraction of the other third of silver.744
In this document the distribution of production costs for smelting is skewed by the very
high cost of the ore at the plant gate and the need to lease the smelting facilities (Table 5-III).
Both these factors would only apply to the case of middlemen involved in the refining of silver
at Catorce. Smelting is shown to be unprofitable for these ores with 3 marks of silver per carga
(0.5%) at a price of 12 pesos 4 reales per carga delivered to the refining unit. However, the
authors point out that ores with 10 marks of silver per carga (1.7 %) can be smelted at a profit,
subject to the pricing established by the mine owners.
From the mid-1820s to the end of the century much more information on refining
comes out of the new independent Mexico, concurrent with the opening up of the silver mines
to foreign investment and involvement. The most detailed economic data on amalgamation and
smelting in Mexico during the nineteenth century come from three sources, St Clair Duport
(1842), Buchan (1856) and Laur (1871). These authors provide an oasis of detailed information
744
I do not know if this two stage refining process applied to all the ores refined at Catorce.
405
cost for 3 pesos/kg

production variables summary pesos
cargas, in pesos silver
ore+freight 37.5 37.5 % of
miller 0.4 subtotal 1
tahonas workers 0.6
cazo workers 1.5
salaries 0.5 labour 2.9 33%
impure salt 1.0 impure salt 1.0 11%
cazo mercury 1.0 mercury 1.0 11%
palma 1.5 others 3.9 44%
straw for mules 1.3
tahona 0.4
deposits 0.4
heating and refining 0.4
subtotal 1 for 6 marks silver 8.9 6.4
mixing 1.5 % subtotal
washing vats and labour 0.3 2
amalgamator 0.4 labour 2.2 43%
patio
impure salt 1.1 impure salt 1.1 22%
mercury 1.6 mercury 1.6 32%
magistral and lime 0.2 others 0.2 4%
subtotal 2 for 3 marks silver 5.1 7.4
total exc. ore for 9 marks silver 14.0 6.8
total inc. ore for 9 marks silver 51.5 24.9
value of silver at 7
63
pesos per mark
profit (loss) 11.5
miller 1.6 % total inc.
mixer 0.8 ore
fluxing ore 1.9 labour 2.3 3%
litharge 22.5 litharge 24.4 30%
smelting
charcoal 7.3 fuel 7.3 9%
lease smelting furnace 9.0 others 10.8 13%
cupell 1.3
palma 0.5
total exc. ore 44.8
total inc. ore for 7.5 marks silver 82.3 47.7
value of silver at 7.5
56.25
pesos per mark
profit (loss) -26.0
Table 5-III. Production costs as reported in 1802 for the cazo, patio and smelting refining
processes carried out at Catorce (San Luis Potosí), adapted from data in footnote 743 and
Appendix D.
406
on production costs and practices based on a wide selection of Mexican refining operations,
within an otherwise barren landscape on either side of the nineteenth century. Duport’s
monograph on Mexican silver production is a work that well merits the recognition given by
his peers of the time.745 Based on his extensive experience in Mexico he was another staunch
European supporter of the patio amalgamation process. His breakdown of its production costs,
excluding the cost of ore, before steam engines were installed at the amalgamation hacienda in
Fresnillo is very perceptive. Instead of using the usual process stages such as milling, patio and
washing, he apportions the cost according to the following headings: animal power 12%, labour
22%, fuel 1%, mercury 34%, salt 17%, magistral 9%, maintenance and others, 5%. He thus
captures in a single set of numbers the essence of the colonial patio amalgamation process:
pure animal and human energy mixed with chemical reagents, with no delicate machinery in-
between.746 In the case of amalgamation he considers the lack of power options to lower the
cost of milling the ore a major challenge, since this stage absorbs more than half the total
production cost and around 19% of the final value of silver.747 Duport includes the variant of
roasting the ores prior to patio amalgamation, as practised in Tasco, where the threshold for
amalgamation to be profitable was a silver content of 0.125%.748
745
The treatise by St. Clair Duport was judged by Percy, himself a major figure in the nineteenth century on
metallurgy, to be ‘one of the best on the subject’ Percy also mentions that he could not add the details on operations
at Pachuca because a Mr. Buchan of the Real del Monte Company died suddenly a few days after promising to
add to the text to be published on patio amalgamation. Percy, Metallurgy, I 576. Randall has pointed out that
Duport did not include the major operations at Real del Monte and Regla, apparently due to conflicting
commercial interests. Randall, Real del Monte, 236.
746
His data from the Hacienda of Fresnillo (Zacatecas) is a very useful reference point for future sections of this
chapter, due to the industrial magnitude of its operations and the fact it did not possess sources of hydraulic power
or inexpensive sources of fuel. From 1st February 1838 to 31st January 1839, it processed 28,407 montones (of
920 kg) of an ore with an average 0.2% silver content, producing 229,035 marks of silver at a total cost of 645,370
pesos. Duport, Métaux précieux au Mexique, 279-80.
747
Ibid., 400.
748
Roasting according to Duport consumes 2% of the final refining cost of 2.08 pesos per quintal of an ore with
0.15% silver content. The extraction cost of ore accounts for 52%, at 1.08 pesos per quintal, 19% to mercury, 6%
to salt and 4% to magistral. The cost translates to 30.1 pesos per kg of refined silver, which is getting close to the
deemed silver value within the ore, but it already includes a 3.5% margin for the operator of the refining hacienda.
According to Duport, under these conditions (mercury at 130 pesos per quintal, salt at 3 pesos per quintal and
magistral at 3.5 pesos per quintal) the operator will not be able to refine at a profit ores with less than 0.125%
407
Based on his data set that covers most of the main mining districts of Mexico, Duport
is the first to propose a detailed profile for the generic production costs of patio amalgamation
in Mexico, based on an ore with 0.2% silver content [average], mercury at a price of 130 pesos
per quintal [high], and a consumption of mercury of 13 oz per mark of refined silver [average].
He takes as a reference point the costs of placing one kg of silver on board a vessel for export
from Mexico [free on board], expressing the costs in their equivalent in silver value, so as to
drive home the point that costs have to be lower than the value of silver in the ore. Table 5-IV
reproduces his estimates. The cash-in-hand margin available to the refiner to meet extraction
costs and profit places a ceiling on extraction costs of 5.04 pesos per carga of ore of 0.2%
silver content.749
He considered smelting as practised in Mexico up to the early 1840s to be inefficient
as to fuel consumption, with no effort made to improve furnaces or the fluxes used, losing 15%
of silver in the process. For Duport there was not sufficient incentive to improve current
practices since he considered it was not an option to treat the majority of ores with a silver
content between 0.15 and 0.2 % of silver.750 He stated that it was ‘impossible to establish the
separate production cost items … for smelting’ (Table 5-II).751
silver content, since already at this level the cost of production rises to 8 pesos per mark of silver (35 pesos per
kg of silver). Ibid., 340-41. The cut-off value corresponds to Villaseñor’s limit for amalgamation at 2 oz of silver
per quintal of ore.
749
Ibid., 370-374. The reasoning employed by Duport to treat export costs as discretionary implies the refiner
could sell with equal opportunity to the domestic market if he manages to meet the cash costs of extraction,
refining and government duties. It seems to ignore that since the majority of silver was exported as coin, these
costs would ultimately have to be factored back in by any party purchasing silver in Mexico for export.
750
Ibid., 398.
751
‘impossible d’établir des prix séparés pour les parties de la production … de la fonte’. Ibid., x, 83, 85, 351.
408
g fine silver
Government duties
export 35
entry to port 20 145
coinage 45
casting bars 45
Essay costs, smelting coinage
10
costs
Freight to port 25
Mercury 112
Patio costs
rent 17.1
general administration 20.52
other reagents 61.56 342
labour 47.88
milling 171
other costs 23.94
total 634
margin to cover extraction
366
cost and profit
446
margin without costs of export
Table 5-IV. Generic profile of patio amalgamation costs (excluding ore cost) in Mexico as
practised around 1840, expressed in terms of g of fine silver. Specific export costs are
highlighted in bold and italics. Data from footnote 749.
In 1855 John Buchan made public his report to the Directors of the Real del Monte
Company on the workings of the mines and the refining of silver in their haciendas, including
Regla. It is a source of generous information on the cost structure of a working commercial
concern, rare for the degree of detail provided to the public. 752 Due to the relevance of its
information to this chapter, I will address its content within the context of my own analysis of
the accounting data from Regla in the following sections. There is an additional source of
production economic data from the previous Adventurers Company of Real del Monte: ‘John
752
Buchan, Report Real del Monte. I am not clear what the intention was at the time to share with the public the
production costs of the company in such detail.
409
Phillips found that for the year 1840 the cost of smelting was only 34 per cent of the value of
silver produced, while that of patio amalgamation was 46.25 per cent’.753
The third author of note is Laur. He estimates that the range of silver content of the
majority of ores in Mexico lies between 0.1 and 0.27%, which in his view are sufficient to
cover the variable mining and amalgamation costs of the ore.754 In the case of smelting, he
reports that in general its costs are very high due to a lack of fuel, unless the ore is rich in lead
(around 25 %). For lead-poor silver ores he states that the threshold silver content for smelting
to be profitable in Mexico is 0.5 %. Of the partial variable cost, excluding the extraction cost
of the ore, labour can make up from 19 to 55%, fuel 25 to 41%, and litharge when required up
to 33%.755
By the end of the nineteenth century all major metallurgical texts included technical
details on the Mexican patio amalgamation process. I will single out Collins who provides an
interesting table of comparative costs for different amalgamation haciendas in Mexico that is
not derived from the previous three authors. His values of production costs without including
mercury are within the range of the other values in Table 5-II. Collins’ data include the sourcing
753
Quoted from a report to the Directors dated 29 June 1841, Real del Monte Proceedings, in Randall, Real del
Monte, 114. Burkart’s lengthy essay on Real del Monte, translated by Miguel Velazquez de Leon and published
in 1861, draws too much on Phillip’s statement of 1840 and Buchan’s report of 1855 to be considered a source of
new data for Mexican refining operations. Burkart, "Memoria Real del Monte."
754
By choosing a silver content of 100g per 100 kg of ore as the threshold value (0.1%), he is assuming that the
value contained in around 70 g of silver per 100kg of ore is enough to cover all costs and leave sufficient profit
for the refiner. In addition to the economic data on refining costs, Laur offers an interesting focus on the issue of
the sourcing of power, since steam and animal sources incur costs, while water was nominally free. In the case of
animal power it was subject to oscillations in the pricing of animal feed, and he cites the example of a major
refining hacienda in Fresnillo that switched to steam engines around 1850 after the yearly expenditure on animal
feed rose from 468,000 francs to 2,464,000 francs. He argues that silver production in Guanajuato decreased as
the price of maize increased from 1862 to 1864. He employs two examples of what he terms average refining
haciendas that use amalgamation to refine suitable ores, where the variable production cost (without including the
cost of the ore or fixed capital costs) for ores between 0.09 and 0.2% silver content lies in the range of 30 to 42 g
of fine silver per 100 kg of ore. Laur, "De la metallurgie de l'argent au Mexique," 51-62, 198-201.
755
He mentions a range between 165 francs (15 pesos) per ton of ore to 800 francs per ton (72 pesos). The peso
to franc conversion is based on Laur’s equivalence of 13 pesos per carga to 50 francs per quintal, 0.09 pesos to 1
franc. For ores with more than 20% lead he provides data on the basis of an implicit value of 1 kg silver at 39.1
pesos. Ibid., 106, 243- 254.
410
of power for each hacienda, though it is the cost of the inputs that determines the major
differences in his range of production costs.756
In the historiography since the 1970s I have only found three new examples of primary
sources on production economics.757 Brading in 1971 reproduces the data from one weekly
account for an unnamed hacienda in the region of Zacatecas in 1801. My calculation of his
breakdown of variable operational costs, net of the cost of ore, yields 17% for labour, 25% for
raw materials, 26% for mercury and 33% for others, for an ore with an average of 0.19% of
silver. The extraction cost of ore is 1.9 pesos per mark of silver.758 He concurs with Villaseñor
that ‘refining profits depended upon what must remain, for the historian, an unknown factor –
the price paid for the mineral in the auctions held at the pit-head’.759 Where he differs from
Villaseñor and Duport is in his view that with increasing silver content the expenditure on
mercury was also higher, so that he concludes ‘a point was then soon reached when it became
more profitable to smelt than to amalgamate’.760 He is not alone in the historiography in
focusing on the total cost of mercury incurred instead of on the cost per kg of silver obtained.
756
Collins, Metallurgy of Lead & Silver, 60-61.
757
The Mexican historian Rina Ortiz Peralta published two works on the Compañia Real del Monte. Rina Ortiz
Peralta, "El beneficio de minerales en el siglo XIX: el caso de la Compañía Real del Monte y Pachuca," Historias
30(1993).;"Algunos aspectos del beneficio de minerales en el siglo XIX: el caso de la Compañia de Real del
Monte y Pachuca," in Hombres, Tecnica, Plata. Minería y Sociedad en Europa y América, Siglos XVI - XIX., ed.
Julio Sánchez Gómez and Guillermo Mira Delli-Zotti (Sevilla: Aconcagua Libros, S.L., 2000). In the more recent
work she includes a table with production costs for the barrel and patio amalgamation processes, but based on
data from secondary sources already commented upon in this section (Burkart and Collins). Since she does not
add new primary sources to the discussion I do not include her in my selection.
758
His data includes a heading of a minor cost item under ‘consumption of lead and litharge in arrastre’, which
may imply that smelting was carried out on a fraction of ore with higher silver content separated during milling.
Brading, Miners Bourbon Mexico, 153-154. Table 10.
759
Ibid., 155.
760
Brading uses the following example: 1.875 pesos is the cost of mercury consumed in refining one montón (30
cargas) of an ore with 1 oz of silver, and 4.5 pesos for the mercury consumed refining the same amount of an ore
with 2.5 oz of silver. Ibid. Table 11. The latter corresponds to 1.8 pesos of mercury per oz of silver, a 4% decrease
in the mercury ratio for a 150% increase in silver amalgamated, in other words the mercury to silver ratio can be
considered constant, and it is the margin of profit that is important, not the absolute amount spent on mercury per
montón.
411
In 2008 Garcia Mendoza published his study of total mining and refining costs in two
refining haciendas of Tasco, as registered in the period from the 2nd August 1562 to the 26th
June 1564. This is an exciting time of innovation within the refining community of New Spain,
barely eight years after the first use of mercury to process silver ores, when smelting had to be
challenged successfully on costs and efficiency by this upstart technology if it was to be
displaced. Unfortunately his attempt to determine how much silver was produced by either
process is seriously flawed by his interpretation of the raw data.761 The document provides
useful information on the costs of consumables and labour in this early period, though not the
amount of ore processed, the silver content of the same, or other important production
parameters.762 The analysis does not answer the question as to why amalgamation displaced
smelting in New Spain.
In the same year Suarez Arguello and von Mentz published a compilation of
correspondence and an extensive transcription of weekly summaries of partial production data
of the mines and refining of silver carried out at Vetagrande on the outskirts of the city of
Zacatecas, for two time series at the turn of the eighteenth century (1791-94, 1806-09). The
quantitative information, such as the amount of ore treated and silver refined, debts, and other
sundry cost items are not analyzed from an accounting point of view and by their limited nature
761
When any archival document of the period states that ‘the yield for each quintal of lead was between two and
three marks of silver’- ‘el rendimiento por cada quintal de plomo era entre dos y tres marcos de plata’ it is
referring to the yield from silver-rich lead ores, not to the amount of lead flux (greta) required to refine a quantity
of silver. This misinterpretation invalidates his estimations of silver refined by smelting. In the case of mercury,
he uses a mercury to silver weight ratio of 6.83, the number of significant figures belying the fundamental error
of his value. This is an incorrect ratio, totally at odds with the chemistry and all the historical data on mercury to
silver ratios reported in Chapter Three. His estimate for silver obtained by amalgamation is also wrong. The author
then proceeds not to use his flawed breakdown between smelting and amalgamation, and analyzes costs (including
labour) based on the combined activities of mining and total refining. García Mendoza, "Minas de plata en Taxco,"
48-49. See also Table 2, p. 54.
762
It is very significant that his primary sources do not include magistral or copper sulphate amongst the expenses
incurred during amalgamation, which confirms the line of thought that the use of magistral originates first in Peru
and then is incorporated in New Spain as of the early seventeenth century (Chapter Three).
412
cannot be used to establish production costs.763 In one of the letters transcribed in their work
there is an interesting comment regarding the guidelines applied at Vetagrande (Zacatecas) to
decide on the refining option according to the content of silver in the ore:
‘in our smelting furnaces at Sauceda … we refine by fire … and as long as their silver content
is 20 ounces per carga [0.42%], they offer profit to whoever has their own smelting facility,
for the only other option is to throw them out together with the tailings, because using
amalgamation would only mean wasting money on them’.764
The last sentence is intriguing, since one interpretation is that he has fallen into the trap
of thinking that ores with a high silver content are more expensive to amalgamate since they
consume more mercury. Otherwise his conclusion is a negation of all that has been reviewed
so far.
The remainder of the historiography since the 1970s to the present is best grouped
according to three strands of narrative since no new primary sources are added to what has
been presented in the preceding paragraphs:
Amalgamation is the lowest cost option to refine ores with ‘low’ silver content. Subject
to the price of mercury, the lower limit of profitability is set at 1.5 to 2.5 oz of silver per quintal
of ore, 0.09 to 0.15% silver by weight.765 Mercury is considered ‘the single most expensive
item in the refining operations [by amalgamation]’.766 It is proposed that mercury constituted
18 to 41% of the total cost of refining in minerals with 2 to 4 oz of silver per quintal (0.13 to
763
Suarez Arguello and Von Mentz, Epistolas y cuentas Vetagrande.
764
‘en los hornos de fundición que tenemos en Sauceda … se les saca por fuego la que tienen, y como no baje su
ley de 20 onzas por carga, ofrece utilidad a quien tiene fundición propia, pues de lo contrario sería necesario
tirarlos al terrero, porque por el beneficio de azogue se perdería dinero en ellos'. Extract from a letter sent to
Don Antonio de Bassoco, on the 8th July 1808, as transcribed in ibid., 699.
765
Bakewell, Silver Mining in Zacatecas, 138.; Castillo Martos and Lang, Metales preciosos - union de dos
mundos, 147.; Brading, Miners Bourbon Mexico, 120-121.; Garner, "Long-Term Silver Mining," 242.
766
"Long-Term Silver Mining," 249.; Lacueva Muñoz, La plata del Rey, 72.
413
0.26%).767 It is also proposed that refining costs ranged from 18 to 24 percent of the value of
silver produced.768 No detailed breakdown of production costs is submitted to support these
ranges.
Smelting is the lowest cost option to refine ores with ‘high’ silver content, which would
otherwise incur a high cost of mercury. Lang does point out that this conclusion is only valid
if the cost for both processes is compared based on a montón of ore.769 The threshold cited to
pass from amalgamation to smelting is given between 4 and 8 oz of silver per quintal of ore,
0.26 to 0.52%.770 No detailed breakdown of production costs is submitted to support this range.
Smelting is the lowest cost option to refine silver ores in general. ‘The smelting process
was simpler, shorter and less costly than amalgamation’. This is certainly a minority opinion,
but has been stated forcefully by Lacueva in one of the most recent books on colonial refining
of silver in New Spain. In the absence of hard data on production costs, of which he is explicitly
conscious, he bases his arguments on indirect indicators such as number of process stages.771
It also seems to mirror John Phillips’ assessment in his report to the Directors of the
Adventurers Company of Real del Monte cited above.
Production choices over three centuries were made by a motley crew of refiners guided
only by the amount of money in their pocket at the end of the day, and this in turn influenced
the environmental history of silver refining in the New World. Amalgamation has
767
Brading, "Mexican Silver Mining," 668, 673.; Castillo Martos and Lang, Metales preciosos - union de dos
mundos, 140.
768
Brading, Miners Bourbon Mexico, 153.; Garner and Stefanou, Economic Growth Bourbon Mexico, 119.
769
Lang, Monopolio Estatal, 50-51.
770
4 oz in Castillo Martos and Lang, Metales preciosos - union de dos mundos, 140.; 8 oz in Lang, Monopolio
Estatal, 50-51.; 4 to 8 oz as a grey zone in Lang, "Silver Refining Technology in Spanish America (patio y
fundición) " 141.
771
‘el proceso de fundición era mucho más simple, más corto y menos costoso que el de amalgamación’ in
Lacueva Muñoz, La plata del Rey, 89. It is the basis for his argument that amalgamation had been imposed by the
sources of capital on the refiners, since thanks to its inherent drawbacks it provided a greater degree of
subordination of the production factors to the owners of capital.
414
overshadowed much of the discussion, and it is common to find sweeping but unsupported
statements such as ‘mercury was essential to colonial silver mining because without it most of
the silver ore could not be profitably refined’.772 Dissenting voices have started to question the
monothematic insistence on mercury:
‘the frequent correlation that is proposed between the lack of mercury with the mining crises
[in New Spain], underestimates how the miners could compensate using smelting according to
the silver content of the ores being extracted … it is necessary to investigate the regional
differences to establish the vulnerability of the economy to any restriction in the supply of
mercury’.773
Back in 1943, the Mexican historian Mendizabal had already proposed factors totally
unrelated to mercury to explain the unexpected increase in silver production registered in the
district of Rosario at the end of the eighteenth century.774 A necessary starting point, though
not as interesting as repressed sexual drives, is the quantitative analysis of the cost of refining
silver. Bakewell’s common sense dictum that ‘mining would obviously not have survived for
long unless someone were making a profit from it’ was correct in stating the obvious: if both
amalgamation and smelting had prevailed over centuries, and made some people very rich, it
772
Garner and Stefanou, Economic Growth Bourbon Mexico, 132.
773
‘la relación frecuente que se hace de la falta de azogue con las crisis mineras, menosprecia el complemento
que los mineros podían tener con el proceso de fundición de acuerdo a la ley del mineral que se estaba extrayendo
… hay que investigar las diferencias regionales para saber la vulnerabilidad de la economía ante la limitación
en el abasto de azogue’ in Rosa Alicia Pérez Luque and Rafael Tovar Rangel, La contabilidad de la Caja Real de
Guanajuato. Una aproximacion a su historia economica 1665-1816 (Guanajuato, Mexico: Universidad de
Guanajuato, 2006), 88-89.
774
The rich and various levels at which explanations can be found to account for variations in silver output in
each mining region is well illustrated by the case of silver, Rosario and sex. In order to explain why the mining
district of Rosario, in the period 1785 to 1789, contributed greater silver revenues than the more renowned Cajas
of Guadalajara, Pachuca, Bolaños and Zimapan, Mendizabal proposed that this was due to the proletarianization
of the indigenous workforce following the expulsion of the Jesuits from New Spain. As soon as the indigenous
males of Rosario were freed from the Jesuit restrictions of working in mining, an occupation that was held to
expose them to alcohol consumption, gambling and sexual relations, they flocked to the mines and haciendas of
Rosario to make up for lost time, and thus created the peak in silver production. Mendizábal, La mineria mexicana,
62.
415
obviously made economic sense to have used them.775 The rest of this chapter will attempt
however to arrive at a more quantitative solution to the following questions:
Was mercury the major variable production cost in the amalgamation of silver or is the
prominence of mercury costs in the historiography to a large extent the consequence of a lack
of information on other cost items? Villaseñor and Duport have provided enough information
in the historiography to render this narrative suspect. It is necessary to evaluate the influence
of the total cost structure for amalgamation, such as salt, labour, the source of power for milling
and the extraction cost of the ore in determining the final profit.
Under what conditions can the economies of smelting and amalgamation be compared
in a meaningful way? The indications are already present in the historiography up to the
nineteenth century that this comparison cannot be attempted by simply stating production costs
divorced from the silver content of the ore they refer to. The sensitivity of such costs to the
silver content of the ore was pointed out by Villaseñor over 250 years ago and yet remains to
be fully reflected in the discussion. A more quantitative analysis is required on the boundary
conditions as to silver content that would have determined the viability of one process over the
other. In particular it is necessary to revise the long-repeated notion that a higher consumption
of mercury made unprofitable the amalgamation of ores with a high silver content.
Why could smelting compete with amalgamation? As Chapter 6 will make evident,
smelting accounted for around 40% of all the silver refined in New Spain, which implies it
consistently provided profits to refiners throughout the colonial period. Could it be that,
contrary to a narrative along Orwellian lines of ‘amalgamation good, smelting bad’, either
process was profitable under certain conditions?
775
Bakewell, Silver Mining in Zacatecas, 187.
416
To reach an answer to these questions it is necessary to have access to a historical
accounting data base that does not incorporate anachronistic elements such as steam or
electricity, obtained from a commercially functioning industrial patio amalgamation and
smelting operation, from which the cost structure of production can be calculated. There is no
substitute for long time series of prices of consumables for each process, an operational
accounting of their consumption, a detailed breakdown of labour costs and structure, and the
amount and price of the product. The accounting books of Regla offer this resource to the
historian.
5.3 The accounting records for Regla
The account books of Regla and other haciendas of the Compañia Real del Monte
report in great detail the production costs for both amalgamation and smelting over the course
of the second half of the nineteenth century, and their analysis will take up the rest of this
chapter. 776 The value of the accounting records of production costs kept by the Compañia Real
del Monte lies not only in their detail over decades, but also because this company was one of
the very few mining and refining conglomerates in Mexico that ran three distinct refining
methods concurrently on an industrial scale (patio amalgamation, barrel amalgamation and
smelting). In the period 1853 to 1888, the Compañia de Real del Monte had spread its refining
activity of silver ores from the mines of Real del Monte over five different refining haciendas,
each with a specific remit as to the refining process it used to produce silver. Thus Regla applied
patio amalgamation and was the only refining unit to smelt ores; Sanchez, Velasco and San
Miguel only processed ores using amalgamation in toneles (barrels), along the lines of the
776
Initially the new Mexican owners of the company decided that its accounting should be carried out by the
Treasury of the Casa de la Moneda, and then by the Compañia de Tabaco, both enterprises that the owners also
managed during this period. It is thanks to the insistence of one of the owners, Nicanor Beistegui, that the company
was allowed to keep its own accounts as of 1852, which in turn made possible the contents that have served so
well this chapter. Ruiz de la Barrera, "La Empresa de Minas del Real del Monte," 76, 81, 83.
417
Barba/Born/Freiburg process pointed out in Chapter 3; Loreto used mainly patio amalgamation
though in at least one year it is listed as also applying amalgamation in toneles.777 The
production costs for each refining unit were also compared to each other, in the way modern
business compares the profitability of discrete production units within an overall corporate cost
structure.778
The analytical structure of the following sections is as follows. First I proceed to
establish the macroeconomic scenario of the nineteenth century, its impact on operating costs
and the viability of extrapolating economic data to the previous centuries. I then calculate the
production cost of both processes as practised at Regla in the third quarter of the nineteenth
century, averaged over at least a decade.779 I determine for each process the input economic
variables with greater impact on the final production cost, as well as a general breakdown of
production costs by main areas. Finally, based on the cost structure for each process, I evaluate
sensitivity scenarios to estimate how economic and technical conditions prevalent in the
previous centuries would have impacted the relative production costs of both processes.
5.4 The macroeconomic context in the nineteenth century.
In the following sections I will be calculating and comparing costs across many decades
without deflacting them. I will therefore briefly discuss the general economic context of these
periods, both with regards to the pricing of silver and to inflation, and whether there was any
impact from these factors on the time series of pricing for the main consumables at Regla.
777
The fame of Regla as a smelting centre transcended Real del Monte: ‘the most important factory in Mexico for
the refining of mineral ores by smelting is Regla’ - ‘L’usine la plus importante du Mexique, pour le traitement
des minerais par la fonte, est celle de Regla’. Laur, "De la metallurgie de l'argent au Mexique," 252.
778
In Appendix B, I present the structure of the accounting books and explain how their information at times
overlap or complement each other.
779
According to Laur accounting records from Mexican refining haciendas may be subject to an under-accounting
of received ores, so as to hide losses or pilfering or to compensate for inefficiencies of the refining processes. In
the case of Regla I am analyzing the data at face value. Laur, "De la metallurgie de l'argent au Mexique," 91.
418
5.4.1 Silver
The difference in the valuation of silver with respect to gold in Europe and China, and
the resulting effects of arbitrage on the world economy since the mid-sixteenth century have
been amply commented upon in the historiography.780 From a ratio of around 12 to 1 with
respect to gold, the value of silver depreciated over the centuries in Europe to reach the 15 to
1 range by the mid nineteenth century, a devaluation of just 25% over 300 years. 781 When the
Mexican investors stepped into the shoes of the Real del Monte Company they could have been
35
30
25
20
15
10
0
1690 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1820 1830 1840 1850 1860 1870 1880 1890 1900
Figure 5-2. The evolution of the gold to silver ratio from 1690 to 1900. Data from footnote
781.
780
For example see Flynn and Giraldez, "Cycles of Silver: Global Economic Unity through the Mid-Eighteenth
Century.";Dennis O. Flynn and Arturo Giraldez, "Arbitrage, China, and World Trade in the early modern period,"
Journal of the Economic and Social History of the Orient 38(1995).
781
‘Average commercial ratio of silver to gold each calendar year since 1687’ in the Report of the Director of the
Mint contained within the Annual Report of the Secretary of the Treasury on the State of the Finances for the
Fiscal Year ended June 30 1921, p. 654. http://fraser.stlouisfed.org/docs/publications/treasar/AR_TREASURY
_1921.pdf
419
70
60
pence per ounce of fine silver

50
40
30
20
10
0
1833
1836
1839
1842
1845
1848
1863
1866
1869
1872
1875
1878
1881
1896
1899
1902
1851
1854
1857
1860
1884
1887
1890
1893
Figure 5-3. The evolution of the price of silver in the London market. Data from footnote
781.
forgiven for thinking that past performance of silver pricing was a very good guarantee to
project future results. For nearly two hundred years the gold to silver ratio had been remarkably
stable (Figure 5-2).782 For any capital investor, the degree of confidence on future revenue
streams is a major element in determining the expected economic viability of an industrial
project. In the late 1840s the future price of silver was not a factor of concern to the silver
refiners of Mexico, judging by the most recent stability of its international price at around 60
pence per ounce of fine silver in the London market (Figure 5-3). And yet Duport ends his
extensive monograph on silver in Mexico published in 1846 with a phrase that captured both
its technical longevity and an intimation of its mortality: ‘a time will come, give or take a
782
The same cannot be said of the silver content in the coinage minted by the Crown. The initial silver content
was set at a value of 11 deniers 4 grains for 1 mark, which was made equal to 65 reales (8 pesos and 1 real). Then
came a succession of devaluations via the reduction of deniers to a mark, or reales to a mark, reaching a cumulative
18% from colonial times to 1826. Joaquin D. Casasus, La Question de l'argent au Mexique (Paris: impr. de Chaix,
1892), 28-30.
420
century, when the only limit to the production of silver will be imposed by the accelerated
decrease of its value’.783
As of the 1870s the floor shifted in a major way from under the silver market.784 By
1902 the price of silver in London, the benchmark for silver sales in the world and the
destination for much of the silver exported from Mexico in the nineteenth century, had
plummeted to less than half its 1870 value, a near mirror-image of the devaluation of silver
with respect to gold.785 In 1873 H.R. Linderman, the Director of the U.S. Mint, included in his
annual report an analysis of the developing weakness in the value of silver and its impact on
the main Western economies. His commentary on the world-wide concatenation of events and
on the self-fulfilling prophesy of the drop in the value of silver merits an extended quote:
‘the steady value of the money-unit [national currency] can only be maintained by making one
of the precious metals the standard or measure of value and assigning a subordinate position as
to coinage for the other.. gold, being less variable than silver, and of superior value, has been
adopted by ... Japan, Germany, the United States of America, Denmark, Sweden and Norway
[Great Britain in 1816] ... this system [single standard]... enhances the value of the one, and
depreciates that of the other ... large quantities of silver hitherto in circulation as standard
money ... will... be thrown on the market as bullion, and aid in its further depreciation ...India
has for many years past been the principal market for silver ... the decline [in demand is] due
783
- ‘le temps viendra, un siècle plutôt, un siècle plus tard, ou la production de l’argent n’aura d’autres limites
que celles qui lui seront imposées par la baisse toujours croissante de sa valeur’. Duport, Métaux précieux au
Mexique, 426.
784
This is the period when the major silver mines of the United States of America came into the market. Viollet
links the new production from the U.S.A. to the upheavals (‘bouleversements’) of the silver market. Eugène
Viollet, "Le problème de l’Argent et l’Etalon d’Or au Mexique" (Université de Paris, 1907), 6. It leads to the
question not analyzed in this thesis as to the comparative production costs of pan amalgamation, patio
amalgamation, barrel amalgamation and smelting between the new North American works, the Mexican ones and
the traditional European production sites, and the impact on silver pricing of new volumes coming to the market
under a new set of production costs.
785
‘Highest, lowest and average price of bar silver in London, per ounce British standard (0.925), since 1833; and
the equivalent in United States gold coin, of an ounce 1.000 fine, taken at the average price and par of exchange’
in the Report of the Director if the Mint contained within the Annual Report of the Secretary of the Treasury on
the State of the Finances for the Fiscal Year ended June 30 1921, p. 653.
http://fraser.stlouisfed.org/docs/publications/treasar/AR_TREASURY_1921.pdf. Part of this Mexican silver was
re-exported to Asia. For data on the amounts of Mexican silver sent from London to China and the Federated
Malay States Silver from 1864 to 1902 see Eduardo Flores Clair, Cuauhtémoc Velasco Avila, and Elia Ramírez
Bautista, Estadísticas mineras de México en el siglo XIX, vol. II(México: Instituto Nacional de Antropología e
Historia, 1985), 140-141.
421
principally to the fall in the price of cotton, soon after the close of the Civil War in this
country’.786
In spite of these major changes in the international valuation of silver, Mexican silver
production increased rather than decreased. The reasons that have been given state that most
production costs were paid in local currency and local inflation did not react in proportion to
the international decrease in the value of silver. In addition the Mexican State took specific
actions to protect its most valued industry, such as keeping the dual standard until 1905, so
refiners could exchange silver for gold at an attractive rate, and allowing the export of silver in
bullion, foregoing the previous restriction of only allowing the export of coin after payment for
coinage. Finally, the industry responded in the time honoured manner of compensating lower
prices with a higher output that included the sale of other metals, such as lead.787
5.4.2 Maize
With regards to the data from the accounting books of Regla, for the period between
1853 and 1873 the macroeconomic scenario is stable with regards to the value of silver and the
foreign exchange rate. The more critical period is from 1873 and 1888, when both the price of
silver and the exchange rate of the Mexican peso to the U.S. dollar had continuously decreased
by up to 25% with respect to their previous levels.788 Is there an impact in this period on the
786
H.R. Linderman, ‘Report of the Director of the Mint, November 1 st 1873’ in the Annual Report of the State of
the Finances to the Forty-third Congress, First Session, December 1, 1873, Washington, Government Printing
Office, 1873, pp. 476-477. http://fraser.stlouisfed.org/docs/publications/treasar/AR_TREASURY_1873.pdf
787
Velasco Avila et al., Estado y minería en México, 271-275.
788
Uribe Salas and Nuñez Altamirano have published a study focused on how the local small and medium mining
enterprises of Real del Monte and Pachuca responded to the drastic decrease in the value of silver, and provides
a detailed picture of how the Government of Mexico moved rapidly to shield its foremost industry. The authors
argue that the smaller enterprises in fact profited after 1873 and that the overall increase in silver production was
in fact due to this segment of the silver industry. Because the paper centres on the strategies of enterprises usually
overshadowed by the Compañia Real del Monte (the authors note wryly that all local enterprises other than that
company were by default medium to small), it does not include data on inflation or production costs. Uribe Salas
and Nuñez Altamirano, "Depreciacion de la plata.."
422
microeconomics of the production process at Regla that needs to be identified and isolated
from other factors that influence my comparative production cost analysis? The majority of
production costs are for domestic expenses charged in Mexican pesos, as will be demonstrated
in the following sections. An idea of the local inflation can be obtained from the long time
series of maize prices at Regla for the period 1872 to 1888. Expenses on maize were not a
major factor in determining the final production cost at Regla, since it was water and not animal
power that drove the machinery, but they are useful as secondary indicators of local inflation
around Regla. I will ignore the possibility that since maize at Regla was destined for animal
feed, it responded to market pricing and dynamics different from those that acted on maize sold
for human consumption.
The plot of monthly costs for maize purchased at Regla (in pesos per cargas) as
calculated from the data in the Contabilidad Mensual is presented in Figure 5-4.789 The overall
profile corresponds to long trends of values between 4 and 5 pesos per carga, with no clear
indication of any long-term increase in the same. Garner reports an average price of maize, for
the valley of Mexico from 1700 to 1800 as 13.2 reales per fanega, 3.3 pesos per carga.790 This
average correlates with the first plateau of values in Figure 5-4. Overall there is no clear
indication that the macroeconomic context influenced in a major way over this period the local
inflation rate at Regla.791
789
The raw data indicate the total monthly expenditure on maize at Regla, expressed in pesos, and total amount
in cargas of maize purchased. These data are then used to calculate the cost of maize in pesos per carga.
790
Richard L Garner, "Price trends in eighteenth-century Mexico," The Hispanic American Historical Review 65,
no. 2 (1985). p. 290. Fanegas have been used as units of volume, area and mass in nineteenth century Mexico. A
carga is equal to 2 fanegas. Manuel Carrera Stampa, "The Evolution of Weights and Measures in New Spain,"
ibid.29, no. 1 (1949): 15.
791
Garner’s extensive study on maize prices does not indicate any regulation on maize prices that would condition
any conclusion derived from their long term moverments.
423
pesos per carga 6
0
6/72 6/73 6/74 6/75 6/76 6/77 6/78 6/79 6/80 6/81 6/82 6/83 6/84 6/85 6/86 6/87 6/88
Figure 5-4. Time series for the expenditure on maize as fodder for animals at Regla (1872-
1888). The source data used to calculate the values of monthly unit costs of maize in pesos per
carga are from Contabilidad Mensual.
5.4.3 Salt
During the period of silver devaluation, salt prices show a step decrease between 1873
and 1875, then a very stable profile between 1875 and 1881, followed by a period of instability
that combines an initial raise in prices followed by a long term decrease (Figure 5-5). The
average over this period was 0.72 pesos per arroba.
1.40
1.20
1.00
pesos per arroba
0.80
0.60
0.40
0.20
0.00
6/72 6/73 6/74 6/75 6/76 6/77 6/78 6/79 6/80 6/81 6/82 6/83 6/84 6/85 6/86 6/87 6/88
Figure 5-5. Monthly expenses of salt consumed at Regla (1872-1888), in pesos per arroba.
Values calculated from data in Contabilidad Mensual.
424
The overall downward trend in prices is more evident on a much longer horizon, from
1852 to 1888 (Figure 5-6). The cost of freight played a major role in determining the cost of
salt at the plant gate. In 1855 freight cost 42 dollars per ton to bring it ‘from the State of San
Luis, being a distance of some 300 miles’.792 This would be equivalent to approximately 0.48
pesos per arroba, or around half the cost of salt at the plant gate. In the case of salt other factors
more critical than the macro-economy were at play during the whole period.793 In his report to
the Directors of the Compañia de Real del Monte in 1855, Buchan explained:
‘Amongst the materials required for the reduction of the ores, Salt is one of the most costly and
difficult to obtain ... to secure the supply ... and also with the hope of reducing its cost, we have
commenced .. the formation of large salt-works on the Lake of Tezcoco ... and we hope in time
to render them adequate to all our needs’.794
1.4
1.2
pesos per arroba of salt
0.8
0.6
0.4 Estados Comparativos
0.2 Contabilidad Mensual
Figure 5-6. Yearly average expense on salt, in pesos per arroba (1853-1888). Values
calculated from data in Contabilidad Mensual correspond to salt prices for Regla. Values
calculated from data in Estados Comparativos correspond to the average price of salt registered
for all the haciendas of the Compañia Real del Monte carrying out amalgamation in any given
year.
792
Salt was also brought in from salt pans in Campeche, using the ports of Tampico and Tuspan. Buchan, Report
Real del Monte, 18-19.
793
In the absence of information on the sodium chloride content in the salt being purchased over these decades, I
can only speculate that another factor to decrease cost is a worsening quality, since the consumption of salt per kg
of silver amalgamated increases substantially for the last years of this period (Chapter Four).
794
Buchan, Report Real del Monte, 19. Lake Tezcoco was situated less than 100 km from Regla, so freight costs
would have decreased substantially. It is not known if the project succeeded. At the present time all the lakes of
this area have virtually dried up due to human agency.
425
5.4.4 Copper sulphate
Copper sulphate is reported as having been one of the items imported by the company,
though it is not stated if all or part, and whether this was a regular practice or not. 795 If so, it
shows a remarkable resiliency to macroeconomic events in the 1873 to 1888 timeline, as
evidenced in the nearly flat array of data points in Figure 5-7, showing a large step decrease in
1875 and a minor one in 1878, and then a constant price to the end of the period.796 The average
cost over the period was 0.13 pesos per lb.
0.50
0.45
0.40
0.35
pesos per pound
0.30
0.25
0.20
0.15
0.10
0.05
0.00
6/72 6/73 6/74 6/75 6/76 6/77 6/78 6/79 6/80 6/81 6/82 6/83 6/84 6/85 6/86 6/87 6/88
Figure 5-7. Monthly expenses of copper sulphate consumed at Regla, in pesos per pound.
5.4.5 Mercury
The behaviour of mercury prices is also unexpected at first sight, being the one major
reagent that had to be imported in its totality and thus most subject to the impact of the
795
Ruiz de la Barrera, "La Empresa de Minas del Real del Monte," 293.
796
The Estados Comparativos report both the amount and expense of the purchases of copper sulphate and the
cheaper magistral, but do not add to the present discussion except to indicate that copper sulphate was also added
to amalgamation in barrels, not only to patio amalgamation. I have no information as to whether long-term
contracts would be responsible for the profile in Figure 5-7.
426
devaluation of the peso that begins around 1873. Figure 5-8 shows a sudden eruption of prices
around the year 1874, precisely the year during which refining operations at Regla became very
erratic. Once the aftershocks of the explosion subside, mercury prices are unexpectedly stable,
and in fact slowly decrease, from 1879 onwards, as the depreciation of the Mexican peso was
gathering steam, only picking up again after 1887. The average price from 1878 to 1888 was
0.6 pesos per lb, 60 pesos per quintal.
2
1.8
1.6
pesos per pound mercury
1.4
1.2
1
0.8
0.6
0.4
0.2
0
6/72 6/73 6/74 6/75 6/76 6/77 6/78 6/79 6/80 6/81 6/82 6/83 6/84 6/85 6/86 6/87 6/88
Figure 5-8. Monthly expenses of mercury consumed at Regla, in pesos per pound. Values
calculated from data in Contabilidad Mensual.
The uniqueness of the peak in mercury pricing observed around 1874 becomes more
evident when the average yearly cost for mercury per pound calculated from the Estados
Comparativos completes the landscape of prices in the period prior to 1874, as observed in
Figure 5-9. Not since the late sixteenth century had prices of mercury reached over 1.6 pesos
per pound (160 pesos per quintal) in the New World. Contrary to salt, the control of mercury
prices lay outside the scope of the management of the Compañia Real del Monte. In a scenario
that seems very familiar, the Spanish authorities in the nineteenth century in their search for
revenues, entered into negotiations with the Rothschild family that would lead the banking
427
concern to provide a series of loans of increasing uncertainty regarding payback, in return for
the concession of the mercury mine at Almadén. The Rothschilds then proceeded to profit from
their near monopoly on mercury after 1832 by increasing the price of mercury to more than
twice its previous level.797
1.6
Estados Comparativos
1.4
Contabilidad Mensual
pesos per pound
1.2
0.8
0.6
0.4
Figure 5-9. Yearly average expense on mercury, in pesos per pound. Values calculated from
data in Contabilidad Mensual correspond to mercury cost at Regla. Values calculated from
data in Estados Comparativos correspond to the average cost of mercury registered for all the
haciendas of the Compañia Real del Monte carrying out amalgamation in any given year.
This would be followed from 1846 by the production of mercury from New Almaden
and others in the western cordillera of the United States that would provide a new source of
mercury in competition to Almadén and Idria (see Chapter 1). The demand for mercury created
by the California Gold rush, the initial play for a monopoly position by the Rothschilds in the
period 1830 to mid-1850s, and then the entry of unexpected new major sources in the Americas
and the resulting glut of mercury in the market created the major pricing oscillations of the
797
An analysis of the involvement of the Rothschilds in the mercury market of the nineteenth century is given in
Miguel A. Lopez Morell and Jose M. O'Kean, "Seeking out and building monopolies, Rothschild strategies in non
ferrous metals international markets (1830-1940)," in 14th Conference of the European Business History
Association (Glasgow2010).
428
nineteenth century that were reflected in all mercury markets. In its final paroxysm it would
create the price spike that for a time must have contributed to derailing amalgamation
operations at Regla in 1874. The resulting overproduction of mercury starting in 1874 and
lasting to 1884 led to mercury prices from New Almaden dropping ‘from $126.22 per flask [of
76 lbs] in 1874 to $49.75 in 1875, and thereafter until 1883 the average price per flask was
about $30.00 and, for a time, $25.00’.798 The roller coaster pricing over the nineteenth century
can be seen in Figure 5-10, where I have used the data reported by Schmitz on the historical
pricing of mercury from three sources in the nineteenth century: Italy (Idria), London (mainly
Almadén) and U.S. (mainly U.S. mines).799
Assuming that prior to 1850 the Idria prices remain a faithful reflection of world
mercury market prices, the business fortunes of both the English Adventurers and their
Mexican successors in the Compañia Real del Monte need to be judged within the context of
the oscillations of this mercury market. The former were hit with a near tripling of mercury
prices after 1836 that lasted to the end of their business venture in Mexico.
‘Phillips estimated that the increase in the price of mercury between 1827 and 1840 had been
so drastic that the cost of amalgamation in the company’s mills was some $40,000 higher in
1840 than it would have been had the price of mercury remained at the 1827 level’.800
Herrera Canales has proposed that for Mexican silver refiners in the nineteenth century
it was more critical to guarantee mercury supply than to worry about mercury pricing, since
mercury ‘only had affected moderately their costs and profit’.801 The scenario as described
798
Henry Winfred Splitter, "Quicksilver at New Almaden," Pacific Historical Review 26, no. 1 (1957). p. 36.
799
Christopher Schmitz, World Non-ferrous Metal Production and Prices, 1700-1976 (London; Totowa, N.J.:
Cass ; Biblio Distribution Centre, 1979), 282-84.
800
Randall, Real del Monte, 116-117.
801
Inés Herrera Canales, "Mercurio para refinar la plata mexicana en el siglo XIX," Historia Mexicana 40, no. 1
(1990): 27-29.
429
above did affect major refiners such as Regla, and not only small producers. The new Mexican
owners of Regla were blessed at the beginning of their venture when mercury prices dropped
nearly to a level of the early 1820s. They would enjoy during the first critical years of
production, when capital investment needs to be repaid as promptly as possible, a certain
stability of mercury prices. When the new pricing hikes hit them in the early 1870s, it coincides
with a period of financial crisis of the company. The drastic increase in the pricing of mercury
alone could explain the fact that from May to July, and December 1874 no amalgamation was
carried out at Regla, while from January to March 1875 the amounts amalgamated were close
to nil or irrelevant. During these periods the only production of silver came from the smelting
of slags (grasas), approximately 100 to 200 kg per month.802 The combination of a low
production of suitable ores and the spike in mercury prices knocked operations at Regla in 1874
and the first trimester of 1875 out of kilter.
12,000 3500
Italy (marks/t)
10,000 US ($/t) 3000
UK (£/t) 2500
8,000
2000
6,000
1500
4,000
1000
2,000 500
0 0
1801 1806 1811 1816 1821 1826 1831 1836 1841 1846 1851 1856 1861 1866 1871 1876 1881 1886 1891 1896 1901
Figure 5-10. Nineteenth century mercury prices from footnote 799. The scale on the right
applies to US and London prices, the scale on the left to prices from Italy.
802
Data for 1874 and 1875 from Informe Mensual Regla.
430
The sole purpose of this chapter is to calculate the production costs at Regla for
amalgamation and smelting, not the study of the finances of the whole Compañia Real del
Monte or the effects of the macroeconomic context on its profitability to shareholders, but some
background information is in order. The Company went through an initial very profitable
period, when it was buoyed by a bonanza from the 1850s to 1860s, then suffered a financial
crisis from 1873 to 1875 that began when it failed to report profits in 1872. The Company
nearly went bankrupt and a cash injection of 1.5 million pesos was required by 1875. This was
eventually repaid between 1877 and 1884. The explanation proposed for the origin of the crisis
is the lack of investment to find new deposits leading to a drop in ore supply to its refining
haciendas.803 Other authors cite as the cause for the financial crisis the decrease of silver prices
in the international market. It is also stated that the level of profits descended but remained
steady in the period 1876 to 1892, which is the period whose operations are covered in this
chapter. The Mexican Company was sold to United Smelting, Refining and Mining Company,
a U.S. consortium, in 1906, by which time all the refining haciendas except Loreto had been
closed down, and the cyanide process had displaced amalgamation with mercury. 804 No
mention is made in either work of the spike in mercury prices precisely in the period of the
greatest financial crisis of the Mexican-owned company, and its impact on production costs for
amalgamation.
5.4.6 Charcoal
In the case of charcoal (Figure 5-11) the pricing profile shows very stable pricing until
1881, after which there is an approximate 40% increase in cost, which could be explained by
803
Ruiz de la Barrera, "La Empresa de Minas del Real del Monte," 110, 132-135.
804
Ines Herrera Canales, Cuahtemoc Velasco Avila, and Eduardo Flores Clair, Etnia y clase, los trabajadores
ingleses de la Compañia del Monte y Pachuca, 1824-1906 (Mexico: INAH, 1981), 3-4.
431
fuel demands finally outstripping supply in this period of continuous smelting, coupled with
the high demand for fuel from amalgamation in barrels, or by a change in sourcing. 805 The
average over the whole period is 1.3 pesos per carga of charcoal.
3.0
2.5
2.0
pesos per carga
1.5
1.0
0.5
0.0
6/75 12/75 6/76 12/76 6/77 12/77 6/78 12/78 6/79 12/79 6/80 12/80 6/81 12/81 6/82 12/82 6/83 12/83 6/84 12/84 6/85 12/85
Figure 5-11. Monthly expenses on charcoal for smelting consumed at Regla (1875-1886), in
pesos per carga. Values calculated from data in Contabilidad Mensual.
5.4.7 Litharge
In the opposite direction, the historic pricing of litharge (Figure 5-12) shows step
decreases punctuating periods of stable pricing, with an average of 0.08 pesos per kg.806 There
is no evident influence of the macroeconomic scenarios on this profile.
805
There is one reference in the literature to charcoal being imported from England and Germany in the nineteenth
century, and brought to the Compañia Real del Monte by rail from the port of Veracruz. Saavedra Silva and
Sánchez Salazar, "Espacio Pachuca-Real del Monte," 93. It does not specify either period, quantities or pricing,
and I have no other source to confirm this. A switch to imported charcoal could explain the increase in price.
806
As explained in Chapter 4, litharge accounts are reported both in pounds and arrobas, so to avoid confusion I
have converted all to kg.
432
0.12
0.10
pesos per kg 0.08
0.06
0.04
0.02
0.00
6/75 12/75 6/76 12/76 6/77 12/77 6/78 12/78 6/79 12/79 6/80 12/80 6/81 12/81 6/82 12/82 6/83 12/83 6/84 12/84 6/85 12/85
Figure 5-12. Monthly expenses on litharge consumed at Regla (1875-1886), in pesos per kg.
In general it can be concluded that prices of consumables for Regla were not seriously
affected by the macroeconomic scenario, a valid reflection of Mexico as a whole.807 This lack
of impact between the macroeconomic scenario and the input costs of production at Regla in
the nineteenth century, and the stability of the silver to gold ratio from the 1870s to at least the
1700s, will provide a relatively stable context within which I can compare and project
production costs without having to deflate my data. Prior to the eighteenth century a more exact
exercise would require deflation, but for the purposes of comparative analysis set out in the
following sections, the same margin of error for ignoring inflation applies to both sets of mixed
data from various centuries, and the conclusions on relative behaviour are compromised, if at
all, equally for both processes.
5.5 The partial variable production costs of amalgamation at Regla, 1872-1888.
In Figure 5-13 I have plotted the monthly variable amalgamation production costs as
calculated for Regla, except for the cost of extracting the ore. From mid 1872 to early 1875 the
variable amalgamation cost reflects the sudden increase and then decrease of mercury prices
807
Casasus, La Question de l'argent, 11.
433
around the 1874 peak (Figure 5-9). Because of the monthly interruptions of amalgamation
observed in the monthly accounts for 1874 I prefer to work with the data from mid 1875 to mid
1888 as being much more representative of average operating conditions. The average variable
cost for amalgamation, excluding the cost of ore, during this period is 7.8 pesos per kg of
refined silver.
16
14
12
pesos per kg silver
10
0
6/72 6/73 6/74 6/75 6/76 6/77 6/78 6/79 6/80 6/81 6/82 6/83 6/84 6/85 6/86 6/87 6/88
Figure 5-13. Monthly production costs of silver refined by amalgamation at Regla (1872-
1888). Values calculated from data in Contabilidad Mensual.
The annual percentage breakdown of these amalgamation costs into their main
components over the period is shown in Table 5-V. The details on how the accounting data are
grouped into these headings is provided in Appendix B. The main reagents make up on average
56% of the partial variable amalgamation cost, of which the share of mercury costs (24%) is
very similar to that of salt (23%). Labour contributes 17%, while the remaining 27%
corresponds to other expenses, as detailed in Appendix B. The cost of fuel is hidden within the
overall heading of ‘other’ since the Contabilidad Mensual does not provide a more detailed
picture. It is however possible to arrive at an approximate estimation using the information
both in the Memorias and in the Estados Comparativos. For the four weeks ending on May 29th
434
1877, this totalled 515.5 cargas of firewood, 634 arrobas of ocote and 89.5 cargas of charcoal
from ocote. According to the same Memorias No. 19 - 21 that served as the source for this
information, the total cost of this fuel came to 373.24 pesos, which represents 2.5% of the total
for that month. This order of magnitude for fuel required by amalgamation is confirmed by
using the accounting data in the Estados Comparativos for the years 1872 and 1873, the only
period during which there is an overlap in the available information. The fuel component in the
total cost of amalgamation in this period is 2.4%, as obtained from the data summarized in
Table 5-VI. The value as calculated errs on the high side, since it also includes fuel for cooking
and domestic heating. As a working figure I assume this range also applies over the period
1875-1888.
Combining the data from Tables 5-V and 5-VI, I obtain the average percentage
distribution of variable amalgamation costs at Regla in the period 1875 to 1888, set out in a pie
chart in Figure 5-14. The only variable cost not present in this calculation is the cost assigned
to the silver ore as delivered to the plant gate. The pie chart of Fig. 5-14 and the data in Table
5-V illustrate the major impact of salt on production costs. When consulting the source data I
was struck by the fact that in many months more money was spent on salt at Regla than on
mercury, and this can be observed at a yearly level in Table 5-V. It is understood that the period
from 1876 to 1888 corresponds to low levels of mercury pricing, but a similar context would
have arisen in colonial times after the price of mercury was lowered to 62 pesos a quintal in
mid-eighteenth century. Much emphasis is rightly placed in the historiography on the influence
of the cost of mercury within amalgamation, but the equivalent order of magnitude of the
expenditure on salt needs also to be taken into account. Salt was a major consumable in the
435
amalgamation
year labour mercury salt copper sulphate other total
mid 1872 24% 24% 24% 7% 21% 100%
1873 23% 27% 20% 7% 25% 100%
1874
mid 1875 15% 43% 14% 7% 21% 100%
1876 15% 36% 23% 7% 19% 100%
1877 17% 28% 25% 11% 20% 100%
1878 16% 23% 24% 13% 23% 100%
1879 18% 19% 24% 16% 23% 100%
1880 19% 19% 22% 13% 27% 100%
1881 16% 21% 23% 11% 29% 100%
1882 16% 22% 24% 10% 28% 100%
1883 14% 22% 23% 11% 29% 100%
1884 15% 20% 27% 9% 29% 100%
1885 16% 22% 24% 7% 31% 100%
1886 18% 23% 23% 5% 32% 100%
1887 18% 24% 20% 5% 33% 100%
mid 1888 19% 21% 22% 4% 34% 100%
whole period 17% 25% 23% 9% 26% 100%
1875-1888 17% 24% 23% 9% 27% 100%
Table 5-V. The percentage breakdown of the main variable amalgamation costs at Regla,
excluding the cost of ore at the plant gate. The percentage values were calculated from the
individual headings within the monthly account data, and then averaged for the year. A total of
153 data sets are represented in the table. Source data from Contabilidad Mensual.
June 1872 -
1873
December 1872
pesos
firewood (leña) 1,615.04 939.97
charcoal and wood
2,224.52 3,758.13
(from ocote )
total fuel costs 3,839.56 4,698.10
total amalgamation
159,410.60 196,719.61
costs
% fuel costs 2.4% 2.4%
Table 5-VI. The percentage contribution to the total variable amalgamation refining cost of
the total fuel required by the amalgamation process. Source data from Estados Comparativos.
436
labour
other 17%
25%
fuel
2%
mercury
copper sulphate 24%
9%
salt
23%
Figure 5-14. Percentage breakdown of the amalgamation cost of silver at Regla in the period
1872 to 1888, excluding the cost of silver ore at the plant gate.
process, and its impact on the economies of the process were as important as those of
mercury.808 Fortunately for amalgamation, salt was a domestic product, both in New Spain /
Mexico and Peru, and as shown by the management at Regla, steps could be taken to minimize
its cost.
5.6 The cost of ore: the missing link in variable costs at Regla
The accounting books do not include the cost of ore delivered to Regla. I have
calculated an approximate value in Table 5-VII based on data published by Buchan on the costs
for the ore produced in the period May 1849 to December 1852, and the years 1853 and 1854.809
808
‘the amount of salt needed for nutrition … and various other industries [in New Spain] appear to be of trifling
significance compared with the demands of the silver industry … the price of salt was of considerable importance
to the silver-ore processing plants’. Ewald, Mexican Salt Industry, 12, 211.; the fact that expenses on salt were
greater than expenses on mercury at Real del Monte has already been commented upon by Ortiz Peralta,
"Beneficio Minerales Real del Monte," 55.; "La Compañia de Real del Monte y Pachuca," 206. In the latter work
she analyzes what she considers to be the novel incursion of a mining and refining concern into the business of
salt production.
809
Buchan, Report Real del Monte, 26-31.
437
The unit cost per carga of mined ore, including freight to the refining haciendas, shows a
marked decrease over this short period. I use as a working figure the average of the years 1853
and 1854, at 2.6 pesos per carga of ore at the plant gate. I will assume as a working number
that this average cost remained the same over the second half of the nineteenth century.810 The
average percentage of silver extracted from the ores processed by amalgamation at Regla was
0.17% by weight (Chapter 4), so the average extraction cost for the ore was 11.1 pesos per kg
amalgamated silver.811
May 1849 to
1853 1854
December 1852
General expense [overheads] 92,456 30,152 36,410
Drainage of mines 215,541 69,334 83,707
Extraction cost 758,906 294,874 322,812
Freight to refining haciendas 105,283 63,768 82,640
total 1,172,186 458,128 525,569
cargas of ore 311,765 181,151 192,982
production cost per carga 3.76 2.53 2.72
all costs in dollars (equivalent to pesos )
Table 5-VII. Mining and other costs for Real del Monte mines in the period 1849 to 1854,
raw data adapted from footnote 809.
810
This is a major assumption, induced by the need to have a working figure on ore costs at Regla. Though labour
was responsible for a major part of mining costs, other factors such as flooding, depletion of the ore deposit, cost
of wood for timbers, investment in machinery could have created major variations in this cost in the second half
of the nineteenth century. On the other hand the strategy of this Company was to lease many mine holdings so as
to switch from one source of ore rapidly to another. This would have kept mining costs down if an alternative
deposit to a deeper or depleting mine could be found. Mendizabal pointed out that it was economic, not technical
causes that led to mines being abandoned. Mendizábal, La mineria mexicana, 26.
811
At a silver extraction rate of 0.17%, 0.23 kg of silver are refined from 138 kg (1 carga) of ore, so that at 2.6
pesos per carga it is equivalent to 11.1 pesos per kg refined silver.
438
This introduces a major new component in the breakdown of variable costs, a single
input greater in cost than the sum of mercury, salt and copper sulphate. The total average
amalgamation variable cost is now 18.9 pesos per 1 kg of refined silver. The pie chart of
variable production costs now changes to the breakdown shown in Figure 5-15. The cost of
extracting ores was the major influence on the total variable cost of amalgamating silver ores,
and not mercury, as pointed out by Villaseñor in 1741.
The pie chart of Figure 5-15 mirrors well the distribution of silver production costs
between mining and amalgamation for the Compañia de Real del Monte, which assigned to
refining 40% of the total costs of the company. 812 The chart as presented in a standard
breakdown of major production cost headings camouflages however a very significant statistic.
It has been estimated that labour costs made up 85% of the total extraction cost of mining ores
in Mexico in the nineteenth century.813 If this estimate is correct, then 55% of the cost of
amalgamating silver at Regla was due to the cost of labour in mining and refining, with up to
an additional 10% in cost of local labour hidden under the headings of salt, fuel and others.
812
Ortiz Peralta, "La Compañia de Real del Monte y Pachuca," 202.
813
Viollet, "Le problème de l’Argent," 121.
439
mercury
10%
fuel
1% salt
9%
copper sulphate
4%
labour
ore 7%
59%
others
10%
Figure 5-15. Percentage breakdown of the total variable cost of production by amalgamation
at Regla, based on the production cost of ore from the mines of Real del Monte in 1853-1854.
Other data as in Figure 5-14.
5.7 The capital cost of amalgamation at Regla.
The total cost of production by amalgamation at Regla would need to incorporate as
well the fixed capital cost due to the investment in infrastructure. Since in 1846 most of the
infrastructure at Regla was bought at a pittance from its previous English owners, from an
accounting point of view the servicing of this fixed capital cost is not representative of an
amalgamation hacienda built from scratch. The report by John Buchan includes a listing for
‘reforming and enlarging reduction works at Regla’ of $20,000, compared to the capital cost
of ‘erecting the new reduction work of Velasco’ at $209,750, based on the barrel process.814
The original cost of construction of Regla in the late eighteenth century has been reported at
814
Buchan, Report Real del Monte, 26. Another reference point is the cost of construction of the major
amalgamation facility of Proaño in Fresnillos, Zacatecas, with a larger amalgamation processing capacity than
Regla (see Chapter Three). It is reported as having cost 300,000 pesos in the first half of the nineteenth century,
before steam engines were installed. The fixed annual cost of servicing that investment is set at 5%. Since
depreciation for tax purposes was probably not an established accounting practice at this time, its fixed capital
cost is probably treated as a nominal interest rate on a loan. Duport, Métaux précieux au Mexique, 262-63, 278.
440
levels over $500,000, but there are strong reasons not to take even the original investment as a
guide.815
Visitors to Regla wondered ‘why his walls were built so thick, or why so many massive
arches should have been constructed, is an enigma to the present generation, as they could by
no means have been intended for a fortress down in a barranca’.816 As Humboldt pointed out,
one of the advantages of the amalgamation process was precisely its low capital expenditure in
plant infrastructure.817 Regla was overdesigned for the needs of both processes, so its capital
cost is not a guide to an average cost in the silver refining business.
According to Duport the rental in Guanajuato of an amalgamation hacienda would
amount to 50 pesos per year per arrastre.818 If this benchmark were applied to Regla, with 24
arrastres, it would have induced a rent of 1,200 pesos per year. With an average production of
nearly 20,000 kg of silver by amalgamation during this period at Regla, even a ten-fold increase
in this level of rent would only correspond to half a peso for every kg of silver produced. In the
light of all these considerations, I will therefore ignore the fixed cost of capital for
amalgamation at Regla and focus only on the total variable cost. Since I will require an estimate
on a generic fixed capital cost of amalgamation for my sensitivity runs later on in this chapter,
I will use Duport’s information that the rental cost of an amalgamation hacienda in the State
815
According to Terrero 2 million pesos were spent to construct the haciendas of Regla, San Francisco Javier,
San Miguel and San Antonio. Manuel Romero de Terreros, Antiguas haciendas de México (Editorial Patria, 1956),
300. Other figures are 425,708 pesos in Ladd, The Making of a Strike, 144.; £1 million in H. G. Ward, Mexico,
Second ed.(London: H. Colburn, 1829), 140., and £500,000 in Lyon, Tour of Mexico, 153.
816
Wilson, Mexico and Its Religion, 366.
817
Humboldt, Essai politique, 84.
818
‘The rent of an hacienda de beneficio is fixed according to the number of arrastras, at a rate of 50 piastres per
year’ - ‘Le loyer d’un hacienda de beneficio se règle d’après le nombre d’arrastras, a raison de 50 piastres par
an’. Duport, Métaux précieux au Mexique, 237. The wording is ambiguous, but a rental of just 50 pesos per year
for the hacienda would be trivial and it does not correlate with his breakdown of costs in page 232 of his book,
where rental came out to nearly 60 pesos for each arrastre.
441
of Guanajuato capable of processing 2,565 montones of ore a year represents 4 - 5% of the
variable refining cost of silver.819
5.8 The partial variable production cost of amalgamation as a function of silver content
of the ore.
The cost of refining is a function of the silver content of an ore, and without a
knowledge of this function it is not possible to compare the economies of amalgamation with
smelting. Table 5-VIII summarizes the method I have applied that allows me to calculate the
total variable production cost as a function of the silver content in the ore, and to establish the
breakeven point between silver value and amalgamation costs.820 The key column that
functions as the fixed axis of the matrix is shaded in grey. It corresponds to an ore with an
average silver content of 0.19% by weight, to which all costs reported so far refer to. From top
to bottom the first two values indicate the amount of silver in a montón (assuming a total
extraction rate of silver of 90%), then the deemed value of the refined silver in pesos (adopting
an equivalence of 38 pesos per kg of silver).821 The second tranche of values correspond to the
average variable costs calculated in Sections 5.5 and 5.6 adjusted to the total amount of silver
in a montón. The matrix is then assembled from this column, by choosing different silver
contents and adjusting the cost of mercury (at a mercury to silver weight ratio of 1.3) and fuel,
maintaining constant all the other values. The fuel costs vary as they reflect the amount of
amalgam fired in a capellina and the bars cast. The cost of salt and copper sulphate do not
vary with silver content since they were added proportional to the size of the montón, not to its
silver content. Labour and other costs are deemed constant throughout the range.
819
Ibid., 232.
820
A similar result would be obtained applying Villaseñor’s matrix based only on costs per montón. I developed
my method before coming across the work by Villaseñor, which helped me to understand his approach.
821
The value of 38 pesos per kg of silver is calculated from the data from 1849 to 1854 in Buchan, Report Real
del Monte, 26-31.
442
Before I plot the function I draw attention to the fact that the cost of mercury consumed
during patio amalgamation at Regla increases with the silver content of the ore only when
calculated on the basis of a montón. On the basis of variable production costs the whole
operation decreases in cost at higher silver content when calculated on the basis of one kg of
silver refined. The data point to the fallacy of judging the viability of amalgamating ores with
high silver content only on the increase in cost of the total amount of mercury consumed.
Based on the matrix in Table 5-VIII, the function of variable costs versus silver content
in Figure 5-16 indicates that it would not have been profitable to amalgamate ores at Regla
with a gross (unextracted) silver content below 0.09%. This correlates well with the histogram
in Figure 4.32, where the lowest tranche registered in the accounts of Regla corresponds to ores
with 0.7 to 0.10% silver content. Obviously a greater refinement is possible, since this simple
calculation does not include the fixed costs, yet it still remains a valid indicator of the limits to
amalgamation at Regla based on the quality of the ore. By processing ores of an average 0.19%
silver content, the total variable costs corresponded to approximately 50% of the silver value
in the ore, thus providing a healthy margin for the operators of Regla from which to retain a
profit from the operation.
With respect to the second parameter, the cost of amalgamation per kg of silver as a
function of silver content in the ore, Figure 5-17 shows the expected decrease in unit cost as
the silver content increases. I have chosen as a cut-off point all production costs over 50 pesos,
since by this time the cost of production will have exceeded the value of the silver being
produced. To emphasize that only one point of the curve has been calculated from the actual
443
% silver in ore 0.00% 0.02% 0.04% 0.06% 0.08% 0.12% 0.19% 0.60% 1.00% 1.90% 3.00%
kg of silver in
Amalgamation 0.00 0.25 0.50 0.75 0.99 1.49 2.36 7.45 12.42 23.60 37.26
monton
value silver
0 9 19 28 38 57 90 283 472 897 1416
pesos
variable
production cost
pesos per kg
variable production costs in pesos per monton
silver
Fuel 0.19 0.00 0.05 0.09 0.14 0.19 0.28 0.44 1.40 2.33 4.43 7.00
Mercury 1.91 0.00 0.48 0.95 1.43 1.90 2.85 4.52 14.27 23.78 45.18 71.33
Salt 1.78 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21
Copper Sulphate 0.72 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69
Labour 1.30 3.07 3.07 3.07 3.07 3.07 3.07 3.07 3.07 3.07 3.07 3.07
others 1.93 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54
ore 11.1 26.2 26.2 26.2 26.2 26.2 26.2 26.2 26.2 26.2 26.2 26.2
total 18.93 39.71 40.23 40.75 41.27 41.80 42.84 44.67 55.37 65.82 89.32 118.04
variable
production cost
161.96 82.03 55.39 42.07 28.74 18.93 7.43 5.30 3.78 3.17
pesos per kg
silver
Table 5-VIII. Matrix to determine the variation of total production cost by patio
amalgamation at Regla as a function of silver content in the ore. All data derived from the
accounting books of Regla, except for the cost of ore which has been derived from Section 5.6.
100
production cost
silver value
80
pesos per monton
60
40
20
0
0.00% 0.05% 0.10% 0.15% 0.20%
gross silver content in ore
Figure 5-16. Cross-over point between the average variable cost of processing one montón
by patio amalgamation at Regla, and a deemed maximum value for the silver extracted from
the ore. Data sourced from Table 5-VIII.
444
accounting records of the period, I have used a large circular marker for the sole accounting
value, with all the other points in the curve projected according to the methodology set out in
Table 5-VIII. The profile of the curve begs the question why patio amalgamation is never
indicated as being suitable for refining ores with a high silver content since after all, the richer
the ore, the lower the production cost. I have already commented in Chapter 4 that the
accounting records show amalgamation tortas being prepared with mixtures that included ores
of a silver content usually associated with smelting, though they never represented the majority
content of a torta. Three practical issues may have constrained the use of amalgamation for
ores with a high content of silver. First is the possibility that ores with a higher content of silver
also contained lead, which would make them unsuitable on chemical grounds for
amalgamation. Second, a greater use of mercury would have required purchasing and
maintaining an increased inventory. Over the whole historical period of amalgamation market
availability of mercury and cash flow constraints on working capital may have limited this
option. Third, there could be operational problems in the milling to a very fine powder the ores
with a high content of native silver. Native silver is not easy to amalgamate when present in
large sizes within an ore.822 What is important to point out is that on the basis of these graphs
there is nothing per se in the economies of amalgamation that would limit the application of
the process to ores with a high silver content.
822
Sarria and Sonneschmidt comment on the operational problems encountered milling and amalgamating ores
with a high content of native silver. de Sarria, Ensayo de metalurgia, 146.; Sonneschmidt and de Fagoaga, Tratado
de la amalgamación de Nueva España, 54, 56. In Peru in the sixteenth and seventeenth century they employed the
tintin method for very rich silver ores, which required pounding the rich silver mass in an ore immersed in mercury
repeatedly with an iron bar. ‘the ore in which pure Silver can be observed, mixed with stone … cannot be ground
well, nor is Mercury able to embrace such large [pieces of] Silver’ – ‘el metal en que se ven en su forma [la] Plata
puros, mezclados con la piedra … ni puede molerse bien, ni el Azogue abrazar [la] Plata tan gruessa’. For a
description of the tintin process see Barba, Arte de los metales, 127-29.
445
60
50
projected production cost

40
value from accounts
pesos per kg silver
30
20
10
0
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0%
gross silver content of ore
Figure 5-17. The projected cost of production using patio amalgamation at Regla (1875-
1888) as a function of the gross silver content of the ore. Data from Table 5-VIII.
5.9 The partial variable production costs of smelting at Regla, 1872-1888
The variable production cost per kg of refined silver obtained via smelting at Regla
averaged 5.2 pesos in the period June 1875 to January 1886, excluding the extraction cost of
the ore (Figure 5-18). To better illustrate the grouping of values in spite of outlying data points
I have opted to represent the scatter of the data. The percentage breakdown of the variable
refining costs of smelting, without the variable cost of the ore, is shown in Table 5-IX. The
deemed production cost of ore for smelting has been assumed to be exactly the same as that for
amalgamation, 2.6 pesos per carga or 1.04 pesos per kg of silver.823
823
At 2.6 pesos per carga production cost (1853-1854), for an ore with 1.9% silver this is equivalent to a cost of
ore of 1 peso per kg of silver refined by smelting. From another context Duport states : ‘one would need to be
able to separate the costs of extracting the ore for smelting from those of the ore for amalgamation, and this is
impossible; since, in all mines the selection is made on the mass of mineral that comes out of the mine, from
where the richest ore is selected for smelting’ - ‘il faudrait pouvoir séparer les frais d’extraction du minéral fondu
de celui destiné à l’amalgamation, et c’est une chose impossible ; car, dans toutes les exploitations on choisit sur
446
16
14
pesos per kg silver 12
10
0
6/72 6/73 6/74 6/75 6/76 6/77 6/78 6/79 6/80 6/81 6/82 6/83 6/84 6/85 6/86 6/87 6/88
Figure 5-18. Monthly production costs of silver refined by smelting at Regla (1875-1886).
smelting
year labour litharge charcoal others total
mid 1875 27% 26% 35% 12% 100%
1876 33% 20% 34% 14% 100%
1877 26% 24% 34% 16% 100%
1878 27% 26% 32% 15% 100%
1879 21% 21% 32% 25% 100%
1880 21% 21% 34% 25% 100%
1881 26% 14% 37% 23% 100%
1882 25% 19% 34% 22% 100%
1883 26% 12% 50% 12% 100%
1884 29% 20% 45% 6% 100%
1885 23% 21% 46% 10% 100%
average 26% 20% 38% 16% 100%
Table 5-IX. The percentage contribution to the partial variable refining cost of the main cost
elements of the process. The percentage values were calculated on a monthly basis, and then
averaged for the year. A total of 103 data sets are represented in the table. Source data from
Contabilidad Mensual.
la masse du minerai, tel qu’il sort de la mine les parties riches qu’on destine a la fonte’. Duport, Métaux précieux
au Mexique, 369. It could also be assumed to be zero, since the proportion of ores for smelting to ores for
amalgamation at Regla was 2:100 (Chapter 4), and it could be argued that the mining business was structured
around amalgamation, with rich ores being only a windfall profit. I do not adopt a zero value because I wish to
extrapolate my projections to cases where the ore for smelting was mined for its own sake.
447
Figure 5-19 shows how the inclusion of the deemed variable cost of the ore changes the
percentage profile, though to a much lesser extent than was observed for the case of
amalgamation. Fuel (31%), then labour (21%), are the main cost components of the process.
The influence of the cost of ore and litharge on the final production cost of smelting are equal
(17%). The cost of labour in mining and refining would amount to around 35% of the final
production cost by smelting, but the local labour fraction of fuel costs and others would
probably add another 15%. Though less than the 65% contribution of labour costs to the total
estimated for amalgamation, it still represents half the total variable cost for smelting at Regla
in this period.
5.10 The variable production cost of smelting as a function of the silver content of the
ore.
As in the case of amalgamation I will construe a matrix that will allow the calculation
of the function of the variable cost of smelting function with respect to the silver content present
in the ore. The column in grey in Table 5-X is the reference column for the subsequent
calculations. This column represents the average costs registered at Regla to smelt ores with an
average silver content of 1.9% (Chapter 4), plus the deemed extraction cost at 1.04 pesos per
kg of silver refined, for a total of 6.2 pesos per kg of silver refined, or 14.65 pesos per carga
of ore smelted.824 The account books list side by side costs in pesos per monton for
amalgamation runs and pesos per carga for smelting runs, so care must be taken in comparing
values. Table 5-X shows this set of values derived from the accounting registers as my
reference point in the column highlighted in grey. I then generate in the rest of Table 5-X the
824
In contrast to ores for amalgamation, I pointed out in Section 4.5.1 that the gross silver content calculated from
the accounts in ores for smelting is found to be equal to the silver extracted by smelting, within the error of the
data.
448
other greta
16% 20%
labour
26%
charcoal
38%
(a)
other ore
14% 17%
greta
17%
labour
21%
charcoal
31%
(b)
Figure 5-19. Percentage breakdown of variable smelting costs at Regla (1875-1886), (a)
without and (b) with the deemed variable cost of the ore.
variation in production cost per carga of ore and per kg of silver as a function of the silver
content. I assume that the only cost that varies in a manner directly proportional to the silver
content of the ore is the cost of litharge, and that all the other costs remain at the level that
corresponds to the smelting of the 1.9% ore. Since the whole mass of ore is heated, the cost of
charcoal will not vary with silver content; if it had, then results would have been more
favourable to smelting. Figure 5-20 shows that smelting at Regla during this period was not
an economic option for ores that had a silver content lower than 0.3%. According to the
449
histogram in Figure 4.41, the lowest tranche of silver ores refined by smelting had a silver
content of 1%.
With regards to the cost of production as a function of silver content, Figure 5-21 shows
both the single value derived directly from the accounting registers (large circular marker) and
the other data points projected according to the method set out in Table 5-X. Even at a silver
content of around 2% the costs of amalgamation at Regla (3.7 pesos per kg silver) are
significantly lower than those for smelting (5.9 pesos per kg silver). Even if the cost of ore for
smelting was set at zero, assuming that it was a windfall product from the normal extraction of
ores for amalgamation, the same advantage for amalgamation applies. This reinforces my
assumption that it was operational reasons, and not economic ones, that limited the use of
amalgamation at high silver contents at Regla.
% silver in ore 0.0% 0.04% 0.06% 0.08% 0.10% 0.12% 0.16% 0.19% 0.40% 0.60% 1.00% 1.90% 3.00%
Smelting kg of silver in a carga 0.00 0.0497 0.075 0.099 0.124 0.149 0.199 0.236 0.497 0.745 1.242 2.360 3.726
value of silver, pesos 0 2 3 4 5 6 7.6 9 19 28 47 90 142
variable production
variable production cost in pesos per carga
cost pesos per kg silver
Fuel 1.94 4.58 4.58 4.58 4.58 4.58 4.58 4.58 4.58 4.58 4.58 4.58 4.58 4.58
Litharge 1.05 0.00 0.05 0.08 0.10 0.13 0.16 0.21 0.25 0.52 0.79 1.31 2.49 3.93
Labour 1.32 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13
others 0.85 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00
ore 1.04 2.45 2.45 2.45 2.45 2.45 2.45 2.45 2.45 2.45 2.45 2.45 2.45 2.45
total 6.21 12.16 12.21 12.24 12.26 12.29 12.32 12.37 12.41 12.68 12.94 13.47 14.65 16.09
variable production
n/a 245.78 164.20 123.42 98.94 82.63 62.24 52.58 25.53 17.37 10.84 6.21 4.32
cost pesos per kg silver
Table 5-X. The derivation of the total cost of smelting per carga and of the variable
production cost per kg of silver as a function of silver content in the ore. For source of data see
text.
450
120
100 production cost
80 silver value
pesos per monton
60
40
20
0
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0%
silver content in ore
Figure 5-20. Cross-over point between the average variable cost of processing one carga by
smelting and the value of silver extracted from the ore. Data from Table 5-X.
60
50
projected production cost

40
value from accounts
pesos per kg silver
30
20
10
0
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0%
silver content of ore
Figure 5-21. The variable production cost by smelting at Regla as a function of the silver
content of the ore. Data from Table 5-X.
451
5.11 Patio amalgamation vs smelting at Regla
The results of the exercise shown in Tables 5-VIII and 5-X highlight the care that must
be taken in comparing the production cost of amalgamation versus smelting. At first sight the
results seem to correlate with the report that ‘John Phillips found that for the year 1840 the cost
of smelting was only 34% of the value of silver produced, while that of patio amalgamation
was 46.25 percent’.825 According to my calculations the corresponding values for the third
quarter of the nineteenth century are 16% and 49% respectively, but in any case both sets of
values as reported by Phillips give the impression that smelting is a better option than
amalgamation. However my set of figures (and I strongly suspect the same applies to those
stated by Phillips) have to be provided with the caveat that they are comparing apples and pears,
since they apply to an ore with 1.9 % silver in the case of smelting and to an ore with 0.19%
silver in the case of amalgamation.826
There is little point in stating a refining cost for a silver ore without indicating the silver
content it applies to. A meaningful comparison between the refining costs of amalgamation
and smelting can only be carried out at the same level of silver content of the ore being refined.
This is the academic equivalent of what a refiner in the field holding a piece of ore in his hand
will ask himself, which of the two processes shall I apply to this rock? Smelting could not
have been applied at Regla in the third quarter of the nineteenth century for an ore with just
0.19% of silver content, since the variable costs for smelting exceeded the value of silver that
could be extracted from it. In contrast, amalgamation could have been applied in theory at a
825
Quoted from a report to the Directors dated 29 June 1841, Real del Monte Proceedings, in Randall, Real del
Monte, 114.
826
Phillips, as many of the English managers of the company of the period, was strongly in favour of replacing
the Mexican patio amalgamation process with the English tradition of smelting, and offered his data as support to
his argument. Whether he was being disingenuous at the time or believed he was comparing costs on the same
basis is open to question.
452
lower production cost to ores with 1.9 % silver. The fact it wasn’t applied means that other
constraints (technical, supply) came into play, but not the theoretical margin of profit. It is
worthwhile at this point of the analysis to bear in mind the following advice from a nineteenth
century textbook on the metallurgy of silver ores:
‘the question as to the most economical treatment of an ore will be determined .. [by] all
conditions … the cost of the same … being only one of many … and it may happen that a
wasteful process is the best, or that a costly process is the cheapest … the proper treatment [is
that] which, however wasteful, costly, or even unscientific, enables the owner to make the most
money out of his ore’827
The results for Regla in the second half of the nineteenth century are unequivocal:
amalgamation was a more cost effective refining process than smelting for ores with a silver
content of 0.19%, and could break-even on ores with 0.09% silver. None of these ranges of ore
could be smelted at a profit, under the set of conditions prevalent at Regla in this period.
Smelting would be used until 1882 on ores that averaged 1.9% silver, and the minimum silver
content in the ore needed to be above 0.3%. Nevertheless the economic data point to a
production cost above that of amalgamation even for the ores with a higher content of silver. It
does not come as a surprise that as of the late nineteenth century the Company decided to export
the ores with a higher silver content to be smelted in England.828
One other important caveat applies to the comparison of production costs. In 1855
Buchan reported that: ‘As far, however, as I have been able to judge by a constant attention to
all these circumstances, the silver left unextracted by the several processes employed is nearly
as follows: by smelting 6 per cent; by patio amalgamation 15 percent.’829 In Section 4.5.3 the
data for Regla after 1874 do not show any major difference in the percentage of unextracted
827
Eissler, The Metallurgy of Argentiferous Lead, 349.
828
Ruiz de la Barrera, "La Empresa de Minas del Real del Monte," 285, 298.
829
453
silver between the two processes, admittedly within a range of limited values showing a very
high average deviation. I have therefore preferred to apply in my calculations a correction of
10% between the gross and the extracted silver content in ores for amalgamation, while no
correction was applied for the ores destined for smelting.
How do the data from Regla compare with the conclusions stated in the historiography
(Section 5.2)? On the minimum content of silver required by each process to compensate
extraction cost with the value of silver in the ore the correlation is good. Only 3 out of 17
examples in Table 5-II show ores below 0.09% silver content being amalgamated, and only
one example of smelting applied to ores below 0.3% silver. On generic statements the
divergence is notable. Smelting is not automatically the best choice on economic grounds for
ores with higher silver content, unless lead content or restrictions on working capital and supply
are at play. Mercury does not represent the highest production cost fraction in amalgamation,
it is the cost of ore (only Villaseñor and Brading have pointed this out). Salt could be as
important a cost factor as mercury. At Regla in the third quarter of the nineteenth century,
smelting was not a process with a lower cost than amalgamation. The next section will address
the question whether during the whole course of silver refining in New Spain / Mexico,
smelting ever came to be able to compete with patio amalgamation on production costs.
5.12 Extrapolating the results from Regla to other historical scenarios
Production costs are location and time specific. The former because the costs of
extracting ore, labour, fuel, other consumables, inland freight, availability of hydraulic power,
the degree of ingratiation with local authorities to obtain exemptions of duties, mercury at cost
or even a sufficient mercury quota at the official price in colonial times, all varied according to
the region. The latter because the pricing and availability of mercury and charcoal, the nature
and demands of the labour force, the fuel efficiency of the smelting furnaces, the degree of skill
of the smelters and the depth of the mines depend to a high degree on the period being studied.
454
Therefore instead of trying to cover all these eventualities, I will reproduce conditions that
would have been found during two important historical periods that can complement the picture
provided by Regla: the pioneering stage of the sixteenth century and the shifting sands of the
seventeenth to eighteenth centuries. In order to reconstruct the production costs for the
historical periods I first establish the spread of historical values of the main cost factors of
refining silver in Table 5-XI as can be found in the historiography.830 Table 5-XII then indicates
that the values registered at Regla do not all correspond to the same position in all the historical
ranges. This gives a strong indication that the relation between production costs of
amalgamation and smelting was not fixed but could show significant variations in different
historical periods from the results observed in the previous section.
Of the two periods that served as bookends to the history of silver refining with mercury
in the New World, Regla is the example that helps to explain the success of amalgamation in
Mexico towards the end of the nineteenth century. Figure 5-22 is a powerful advertisement for
the patio process, showing that at Regla it could even undercut the economies of smelting of
ores with a high silver content. These results confirm that Regla was well configured for
amalgamation. Its ore shows low to medium extraction costs, the hacienda had an extremely
low fixed capital cost, it enjoyed free and constant hydraulic power, medium to low cost for
mercury after 1875, employed one of the lowest ratios of mercury to silver for patio
amalgamation, and during the period 1876 to 1888 its patio reactor operated at full capacity,
thus optimizing labour costs. The same cannot be said of smelting, where the smelting capacity
830
a) García Mendoza, "Minas de plata en Taxco."; b) Menegus Borneman, "Las comunidades productoras de sal
y los mercados mineros: los casos de Taxco y Temascaltepec."; c) Fabry, Impugnacion a reflexiones de
Villaseñor.; d) Dominguez de la Fuente, Leal Informe Politico-Legal.; e) Brading, Miners Bourbon Mexico.; f)
Duport, Métaux précieux au Mexique.; g) Laur, "De la metallurgie de l'argent au Mexique." ; h) West, The Parral
Mining District.
455
location period cost unit reported notes calculated to a common unit source, page
Mercury generic 16c to 19c 30 - 190 pesos /quintal this chapter; a, 54
Tasco 16c 1.16 pesos /fanega 0.1 a, 54
New Spain 1717 18 to 20 real /fanega plus freight >0.3 b, 82
Pachuca 1743 20-22 pesos /carga 1.7-1.8 c
Guanajuato 1763 3 pesos /fanega 0.4 d, 207-212
New Spain late 18c 1 pesos /fanega 0.1 e, 153-154
Salt Regla 1801 3.75 pesos/ 75 lbs 1.3 pesos /arroba e, 155
Mexico mid 19c 2 to 12 pesos /carga 0.2 to 1 f, 87
Fresnillo mid 19c 11 to 26 real /fanega fanega of 200 lbs 0.2 to 0.4 f, 278
Mexico 1860s 5 pesos /carga 0.40 g, 63

Mexico second half 19c 4 to 12 pesos /carga 0.3 to 1 g, 65-66
Zacatecas late 18c 1.66 pesos /arroba 0.07 e, 153-154
Copper sulphate Mexico mid 19c 3 to 5 pesos /carga 0.12 to 0.2 pesos /pound f, 98
Fresnillo mid 19c 3.5 to 5.75 pesos /carga 0.14 to 0.23 f, 278
Tasco 16c 2.70 pesos /quintal 0.06 a, 54
pesos /2
Guanajuato 1760s to 1774 1.25 0.01 d,270
quintales
Litharge (greta ) pesos /kg
Mexico mid 19c 2.5 to 4 pesos /quintal 0.05 to 0.09 f, 75
Mexico second half 19c 0.07 pesos /arroba 0.01 g, 83
Catorce 1870 12 pesos /carga 0.09 g, 251-252
Tasco 16c 0.3 pesos /carga 0.3 a, 54
Fresnillo mid 19c 1.5 reales /arroba 2.3 f, 278
Charcoal Fresnillo second half 19c 0.19 pesos /arroba 2.3 pesos /carga g, 83
Zacatecas second half 19c 0.19 pesos /arroba 2.3 g, 83
Guanajuato second half 19c 0.21 pesos /arroba 2.5 g, 83
generic all nil tailings 0 f, 375
Parral 1649 0.25 to 0.3 pesos /quintal check 0.7 to 0.9 h, 96
Regla 1801 0.3 pesos /quintal 1 oz/qtal silver 0.9 e, 155
Regla 1801 0.8 pesos /quintal 2.5 oz/qtal silver 2.4 e, 155
pesos /14
Zacatecas 1839-1840 8.33 0.20% 7.1 f, 223
arrobas
pesos /14
buscones 1840s 8.1 ? 6.9 f, 223
ores arrobas pesos /carga
Zacatecas mid 19c 5.5 pesos /carga 0.17 to 0.25% 5.5 f, 246
pesos /mark
Zacatecas mid 19c 4.39 0.46% f, 252
silver
3.43 to
Fresnillo mid 19c pesos /carga 0.20% 3.4 to 3.8 f, 259
3.75
Guadalupe y
mid 19c 5.25 pesos /carga 0.25 to 0.3% 5.3 f, 308
Calvo
16c to 18c 1000
Furnace efficiency generic kg/kg silver this chapter
second half 19c 200
small scale
nil
smelting works
Fixed capital cost generic all
5% of large
variable amalgamation f, 279-280
cost works
nil water power
Source of power generic all 12% of
variable animal power f, 279-280
cost
Table 5-XI. Selection of historical costs from the sixteenth to nineteenth century of the main
factors that determine the production cost of amalgamation and smelting. Sources in footnote
830.
456
% variable costs at Regla

Regla historical values
amalgamation smelting
cost of ore pesos /carga 2.6 0 to 7.1 58 16
cost of power for approx. 12% of
free 0
mills, furnaces variable costs
approx. 5% of assumed assumed

fixed capital cost very low
variable costs negligible negligible
mercury to silver
1.3 1.3 to 2.1
ratio 10
mercury pesos /quintal 60 30 to 190
salt pesos /arroba 0.7 0.1 to 1.8 9
copper sulphate pesos /pound 0.13 0.07 to 0.23 4
efficiency furnaces 200 kg fuel/ kg silver 1000 kg fuel/kg silver

32
charcoal pesos /carga 1.3 0.3 to 2.5
litharge pesos /kg 0.08 0 to 0.9 17
labour high indirect; low to high 7 22
Table 5-XII. The profile of costs registered at Regla in the third quarter of the nineteenth
century within the historical context of New Spain / Mexico. For sources see Table 5-XI and
Sections 5.8 and 5.11.
60
50
amalgamation
pesos / kg silver
40 smelting
30
20
10
0
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0%
silver content
Figure 5-22. Total variable refining cost at Regla (1875-1888), as a function of the silver
content of the ore. Data from Tables 5-VIII and 5-X.
457
had been over-designed (six blast furnaces had been built, exceeding by far the amount of
available ore for smelting), labour for smelting had therefore become a fixed cost as it was
underutilized yet had to be retained due to its specialist skills, and the ore was poor in lead,
requiring the purchase of litharge. The one redeeming feature for smelting at Regla was the
very high fuel efficiency of its blast furnaces.
At the other bookend to this history, the second half of the sixteenth century, when
refiners were still grappling with smelting and amalgamation was beginning to offer some of
them solace, the context was completely different. For smelting it was characterized by
inexpensive infrastructure, negligible extraction costs for the initial rich surface deposits,
abundant and nearby fuel sources, inefficient furnaces and ores rich in lead in some locations.
As to amalgamation, it was initially applied to ores obtained from tailings, thus with zero
extraction cost, mercury was sold at the highest levels of historical prices, the novelty of the
process resulted in some very high mercury to silver ratios, investment in new stamp mills was
the major capital cost involved, and water was not always available to drive them. This new
context can be translated into numbers that will allow me to calculate a new set of comparative
refining costs for both processes. The aim is not to arrive at absolute production economic data
but to estimate changes in the relative efficiency of both processes to refine at a profit ores with
different levels of silver content. The method is analogous to that applied in Tables 5-VIII and
5-X, except that now I will change certain key values as set out in Table 5-XIII. Once the new
values are substituted in a calculation matrix (reproduced in Appendix E), a new plot of
comparative production costs as a function of silver content is generated (Figure 5-23).
Smelting is now the clear choice, on the basis of cost, for ores with a silver content above 0.5%,
while amalgamation remains the process of choice for ores with a silver content below this
threshold.
458
original Regla projected value for

data 16c context
explanation
pesos per kg of silver
amalgamation
mercury 1.94 9.93 a factor of 5.12 = (180/60)*(2.1/1.3)
labour 1.31 0.13 10% of 19c value
ore 11.1 0 tailings
fuel 0.19 0.04 (0.3/1.3)
power 0 1.2 approx. 12% variable cost
fixed capital 0 0.5 approx. 5% variable cost
smelting
fuel 1.94 2.23 a factor of 1.15 =(1000/200)*(0.3/1.3)
litharge 1.02 0 lead rich ores
ore 1.00 0 rich surface deposits
power 0 0 low cost
fixed capital 0 0 inexpensive infrastructure
Table 5-XIII. Sensitivity values for a cost approximation to the context of refining of silver in
the second half of the sixteenth century.
60
50
amalgamation
40 smelting
pesos / kg silver
30
20
10
0
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0%
silver content
Figure 5-23. Comparative production costs of amalgamation and smelting in the context of
the second half of the sixteenth century as a function of the silver content of the ore. See
Appendix E for source data.
459
The period between the bookends is too complex to reduce to a single graph. The
decades from mid seventeenth century to the end of the eighteenth century were a period of
alternating fortunes for amalgamation and smelting. Infrastructure costs become important as
major industrial concerns develop that cover mining and refining of their own ores as well as
the ores from smaller miners via the lucrative maquila, the toll charged for refining ores.831 In
contrast, smelting continued to require a much lower capital fixed cost, and could always be
carried out at a much smaller scale. Mercury prices would drop to their lowest historical levels
at the end of this period, though mercury to silver ratios would remain high except where iron
was used as an additive or the nature of the ore dictated otherwise. Furnace efficiency remained
low.
Two comments in the historiography point to a fine economic balancing act during this
period between amalgamation and smelting. The appraisal by Humboldt that ‘usually it is only
the abundance of mercury and the ease of procuring it that determine the choice of the miner
[refiner] on the [refining] method he will choose’, by its silence on refining costs is pregnant
with the implicit assumption that smelting was only displaced by an opportunistic supply of
mercury, not by an inherent economic shackling of its feet.832 Garner comments that in the
years between 1798 and 1803 in Zacatecas, cutting the diezmo to one half triggered an increase
831
The view from the other side was quite stark. Small miners bore all the risk of mining and had to sell their ores
at the mercy of big refining haciendas who never informed the miner of the real silver content of their product,
but simply returned to him an amount of silver corresponding to its deemed content on which the tolling charge
was calculated, with the refiner doing all the deeming. An extensive litany of complaints can be found in
Dominguez de la Fuente, Leal Informe Politico-Legal. The potential for fraud was ever present, regardless of the
legal context that developed around the maquila. The law recognized that refining was a legitimate activity
separate from mining. Haciendas by law had to publish their toll charges, such that the cost of mercury was at the
same price it was sold to refiners, while a 12% premium could be applied to other consumables. Mining judges
could revise the charges applied by the tolling haciendas. In principle the owner of ore could assist and intervene
during process. Federico Kunz, "Evolucion historica del regimen legal del beneficio de minerales en Mexico " in
Mineria Regional Mexicana. Primera reunion de historiadores de la mineria latinoamericana, ed. Dolores Avila
Herrera and Rina Ortiz (Mexico: Instituto Nacional de Antropologia e Historia, 1994), 65-77.
832
‘souvent ce n’est que l’abondance du mercure et la facilite de s’en procurer qui decident le mineur dans le
choix de la methode qu’il emploie’. Humboldt, Essai politique, Tome IV, 56.
460
of 2.5 times the previous amount of silver smelted, which again is mostly (but not all) lost in
1803 when on re-establishing the previous diezmo production of silver decreases by one
third.833 The decrease in the diezmo is equivalent to approximately 2 pesos in the value of one
kg of silver, just 5% of the value of the silver being refined. It was a very fine line that at times
kept smelting and refiners apart.
The method applied to the sixteenth century can be used to determine whether
conditions could have existed in parts of New Spain in the seventeenth and eighteenth centuries
that would cause the two functions of production costs versus silver content to overlap
substantially onto a single curve. This would have the practical consequence that a refiner could
apply either amalgamation to smelting, subject to the nature of the ore, without unduly
sacrificing his profit margin. Table 5-XIV and Figure 5-24 have been drawn up in this light.
They do not attempt to represent specific locations or years within this very complex period,
they only represent a context that is feasible in the light of what is known at the present time.
The plots that correspond to this new data set prove that it is possible to find a historical context
where the curves of production cost versus silver content in the ore for amalgamation and
smelting are superimposed on each other, as seen in Figure 5-25. It provides the first
quantitative explanation to the empirical observation in the historiography that after mid
seventeenth century major shifts are observed within silver refining in New Spain, with
smelting increasing its presence with respect to amalgamation (see Chapter 6). It proves that
under a certain set of conditions, refiners of this period could have been able to choose
amalgamation or smelting without having to sacrifice their profit levels.
833
Garner and Stefanou, Economic Growth Bourbon Mexico, 135. The diezmo was the royalty of 10% imposed
by the Crown on silver produced.
461
original Regla projected value for

data 17-18c context explanation
pesos per kg of silver
amalgamation
mercury 1.94 2.7 a factor of 1.4 = (60/60)*(1.8/1.3)
ore 11.1 16.7 a factor of 1.5, less efficient deep mining
fuel 0.19 0.04 (0.3/1.3)
salt 1.83 3.7 a factor of 2, salt sources not optimized
power 0 2.8 12% variable cost

fixed capital 0 1.3 5% variable cost
smelting
fuel 1.94 2.2 a factor of 1.15 =(1000/200)*(0.3/1.3)
litharge 1.02 0 lead rich ores

ore 1.00 0 cost borne by ore for amalgamation
power 0 0 low cost
fixed capital 0 0 inexpensive infrastructure
Table 5-XIV. A theoretical context of production costs viable for the period 17c to 18c, and
within the limits of data provided in Table 5-XII.
Charcoal would have been needed at low prices to offset the inefficient furnaces, and
lead fluxes not required or available at a very low price. The extraction cost of ores for smelting
would have had to be low, either in practice or at an accounting level.834 Salt needs to be
expensive, water power absent from the amalgamation hacienda, capital cost of infrastructure
834
ores with as little as 0.0125% (4 oz per ton) silver broke even on production costs if the ore was priced at zero
value at the smelting work in Konigsberg, Norway, in the nineteenth century. Percy, Metallurgy, I 513. However
see Appendix F for the context to European production costs and threshold silver values in the ores.
462
and inefficient mining costs high.835 None of these conditions can be described as unrealistic
for the period.
60
50
amalgamation
40 smelting
pesos / kg silver
30
20
10
0
0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0%
silver content
Figure 5-24. Total production cost as a function of the silver content of the ore, within the
context of the period from the mid-seventeenth to mid-eighteenth century. See text for details
on the generation of the values plotted.
There are still too many gaps in the historiography on economic data of refining for this
period to take this analysis any further, but it has demonstrated that the production costs of
smelting could have been equivalent to those of amalgamation for some refiners in New Spain.
835
‘already by the late eighteenth century …. mining had reached very deep levels’ - ‘ya desde las postrimerías
del siglo XVIII … las explotaciones habían alcanzado gran profundidad’. Ortiz Peralta, "La Compañia de Real
del Monte y Pachuca," 201.
463
This is very important since it means that refiners could choose between amalgamation and
smelting, subject to the nature of the ores, without sacrificing the margin of profit.836
5.13 The structure of labour costs at Regla.
Smelting has been reported in the historiography as requiring major amounts of labour.
Medina promoted the amalgamation process by drawing attention to the many stages of labour
required by the smelting process as being applied in his time.837 The deemed intensive need for
labour in New Spain during the initial period when only smelting was used has been set against
the ravages caused by epidemics within the indigenous labour pool to indicate a deficiency of
manpower that contributed to the introduction of amalgamation.838 On the other hand, a recent
compilation of the various classes of labour required in an amalgamation hacienda has
concluded that amalgamation required more labour than smelting.839 There is no equivalent
work on the structure of the labour force of a refining hacienda comparable to the study of the
Zacatecas mining community of the nineteenth century.840 In the case of Regla two studies
appear in the historiography, one related to the English component of the workforce, and the
836
In Chapter 6 I will come back to the problem of explaining the changes observed in the balance between
amalgamation and smelting in most of the Cajas (Regional Treasuries) throughout the colonial period.
837
‘Y así he visto como se benefician los dichos metales en muchas partes con greta y cendrada y la muy grande
costa de los dueños de las minas … de indios como de negros, porque un ingenio de caballos que trae un horno
andando bueno … así que … ha menester cuatro fundidores y cuatro cargadores y dos españoles que se muden
por sus cuartos, y por personas que anden con los caballos del ingenio por sus cuartos, y más dos afinadores, y
para moler la greta y cendrada otras dos personas, y para hacer los hornos y labrar las piedras otras dos, y para
hollar las cendradas cada una que afinan, son menester seis personas, porque al final de dos días a la semana
que vendrán a ser dos personas cada día y de noche doce negros, y más para cubrir y sacar dicho carbón’ in
Castillo Martos, Bartolomé de Medina, 112.
838
Berthe, "Le mercure et l'industrie mexicaine au XVIéme siècle," 145-46.
839
Lara Meza, Haciendas de beneficio de Guanajuato, 101.
840
Guadalupe Navas, "Zacatecas a fin del siglo XIX," in Mineria Regional Mexicana. Primera reunion de
historiadores de la mineria latinoamericana. , ed. Dolores Avila Herrera and Rina Ortiz (Mexico: Instituto
Nacional de Antropologia e Historia, 1994). It should be pointed out that prisoners housed in a Presidio (jail) were
part of the general labour used by the Compañia Real del Monte until the end of 1874, but only worked in the
mines. The Presidio was built in 1850. Inés Herrera Canales and Rina Ortiz Peralta, "Mexican Mining Heritage-
The Real Del Monte Site," CRM (Washington) 21, no. 7 (1998): 20.
464
other states there were eighteen different labour functions within Regla but does not include
any economic data.841
The degree of detail on labour included in some of the Regla account books shed further
light into the structure of labour costs in both processes. The Memorias Semanales provide a
very detailed breakdown of the labour structure and costs at Regla. Table 5-XV summarizes
one such example using the Memorias Number 18 to 21, accounting records that cover the four
weeks beginning the 5th May 1877 to the week starting the 26th May 1877, that spell out the
itemized labour costs incurred by both patio amalgamation and smelting at Regla.842 An initial
overview brings to light some interesting facets on the labour structure at Regla in the later part
of the nineteenth century.
a. Not all the workforce is on the payroll as full time employees of Regla. Those on
part-time wages are mostly craftsmen involved in the ongoing repairs and maintenance
required by the hacienda. Unskilled manual workers (peones) are paid according to different
wage scales on a per diem basis, but there is no guide to determine the reason for the pay
differentials. Part of the work is outsourced: the stamp mill is run by two work teams, which
are paid on the basis of cargas of ore processed per week; the cupellation of silver and its
casting into silver bars is also carried out by work teams and the cost charged per bars cast.
841
Herrera Canales, Velasco Avila and Flores Clair have focused their detailed analysis on the fortunes of the
English workforce at the Compañia de Real del Monte under its new Mexican ownership. They detect that
Englishmen were kept in key management posts (‘puestos de confianza’) for some twenty years after the Company
was bought from its original English owners. Part of their duties were those of supervision over manual labourers.
The post of Director from 1848 to 1868 was held by just one extended English family: John Buchan (1848 to
1856), followed by his brother-in-law Thomas R. Auld (until 1862) and then by his brother-in-law’s brother,
Edward Auld (until 1868). Cornish miners emigrated to Real del Monte and other parts of Mexico during the
nineteenth century, leaving evidence of their Cornish culture transplanted to Real del Monte. Two Mexicans were
responsible as Directors during the period of interest in this chapter, Jose Maria Camargo from 1872 to 1873, and
Jose de Landero, until 1899. Herrera Canales, Velasco Avila, and Flores Clair, Etnia y clase, 2, 7, 9-27, 47-64.
On the different labour functions within Regla see Ladd, The Making of a Strike, 7.
842
The wages at Regla were slightly below those reported by Laur for Fresnillo, which had a larger workforce
(415) and throughput (48,000 tons) than Regla. Laur, "De la metallurgie de l'argent au Mexique," 85-86.
465
by day or wage cost of labour

by task total man-days
week pesos pesos
Stamp mill
Epitacio Flores & team unknown
per cargas of ore 623 a
Geronimo Marquez & team unknown
subtotal 1 623 324
Arrastres
Vicente Osorio Captain (Capitan ) 1 7 / wk 24 b
Ysidro Rios Captain 1 7 / wk 24 b
Jesus Osorio Overseer (Velador ) 1 4 / wk 24 b
Workers (Peones ) 20 0.38, 0.44 / d 551 c
subtotal 2 23 623 274
Patio Amalgamation
Cayetano Ortiz Captain 1 9 / wk 24 b
Rafael Perez Captain 1 7 / wk 24 b
Luz Acevedo Assistant (Ayudante ) 1 6 / wk 24 b
Nestor Hernandez Assistant 1 6 / wk 24 b
Workers (Peones ) 44 0.19;0.38;0.5 / d 978 c
subtotal 3 48 1074 688
Capellina & casting of silver bars
Mauro Vargas for minding capellinas 1 1 per capellina unknown
Jesus Hernandez & team for casting 26 silver bars unknown 0.25 per bar cast unknown
subtotal 4 28
General Administration
Juan Cuataparo Administrator 1 40 / wk 24 b
Jose Adrian Palomo Paymaster ore (Rayador) 1 18 / wk 24 b
Jesus Guerrero Stock person (Recibidor) 1 12 / wk 24 b
Eduardo Fuentes Assistant 1 7 / wk 24 b
Jorge Castro Assistant 1 7 / wk 24 b
Jose Salazar Overseer 1 7 / wk 24 b
Ramon Hurtado Gatekeeper 1 7 / wk 24 b
Lauro Perez Gatekeeper 1 7 / wk 24 b
Justo Yslas Gatekeeper 1 4 / wk 24 b
subtotal 4 9 216 436
Repairs
Bricklayer 4 0.75 / d 90 c
Workers (Peones ) 21 0.38 / d 474 c
Master Carpenter 1 10 / wk 24 b
Carpenters 6 1;0.75;0.63 /d 135 c
Carpenters 3 1.5 / wk 72 b
Master Smith 1 8 / wk 24 b
Smiths 4 0.75;0.5;0,38/d 93.3 c
Sonadores? 2 0.19/d 46 c
Roofers 2 0.75;0.38/d 34 c
Leathersmith 1 0.75/d 14 c
subtotal 5 45 1007 625
Recovering amalgam from fine ore residue
Rafael Romero Captain 1 7 / wk 24 b
Ignacio Castillo Assistant 1 4 / wk 24 b
33 0.25;0.19;0.12 / d 695 c
subtotal 6 35 743 161
Stables
Joaquin Perez Head of stable 1 7 / wk 24 b
Workers (Peones ) 4 0.38 / d 112 c
subtotal 7 5 136 70
Smelting
Antioco Perez Captain 1 14 / wk 24 b
Panfilo Perez Captain 1 7 / wk 24 b
Lucas Guzman Assistant 1 5 / wk 24 b
Jose Mugica Assistant 1 5 / wk 24 b
Furnacemen 8 0.62 / d 222 c
Carriers 8 0.5 / d 221 c
Charcoal-loaders 8 0.25 / d 221 c
Workers (Peones ) 24 0.38;0.25;0.12 / d 356 c
Jesus Hernandez & team refining unknown per cupellation unknown
Pasenal Palacios various 1 per shift unknown
Eduardo Lopez Assistant 1 per task unknown
Ventilation team 1 per task unknown
subtotal 8 52 3 1116 550
Extraordinary expenses
Andres Yenteno Medical doctor 1 8 / wk 24 d
help with capellinas unknown
subtotal 9 1 24 40
notes: a. deemed equivalent to arrastres b. based on 6 day week c. as reported in days in Memorias. Miscellaneous work paid by
unit or varas such as horseshoe fitting, stone laying and wood sawing has not been included since no information is provided except total
cost of labour, and it represents a minor fraction of the total wage cost..
Table 5-XV. Breakdown of labour man-days and costs for the various refining stages carried
out at at Regla based on data for the four week period ending on May 29th 1877 (Memorias
Numero 18 - 22).
466
b. There is no designation of an azoguero by such name in the accounts. Most probably
the captains that are accounted for under patio amalgamation are the equivalent of an
azoguero.843
c. As expected, the work of minding the separation of mercury from silver in a capellina
is so simple that the minder of the capellina was paid just 1 peso per capellina run. The actual
man-handling of the capellina is done as a side activity by peones who are given a tip
(gratificacion) between that is registered under Various Expenses. The tip is equivalent in
value to the payment for minding a capellina.
d. The group of skilled workers (bricklayers, carpenters, etc) in charge of repairs and
maintenance are not named, not even the master carpenter or the master smith, both of which
had weekly wages equivalent to those of named captains or assistants. The peones are not listed
by individual names either. Thus it is not possible to confirm the large presence of women
expected within the planillero squads. On the other hand the junior gatekeeper merits being
identified as Justo Yslas earning 4 pesos a week, less than what a peon made.
e. The senior captain in charge of smelting earns double the wage of the captains in
charge of amalgamation, an indication of the greater skill required for the smelting process.
The wage structure of the smelting section indicates that the group of very skilled workers and
their assistants was kept on the weekly payroll irrespective of whether smelting was carried out
during the week.
f. There is no job description that can be readily identified for custodians of the integrity
of the refined silver or of the mercury inventory. This would place primary responsibility on
843
According to Duport the azogueros in the main haciendas of Guanajuato in the 1840s were earning 25 pesos
per week. This is considerably more than the 7 to 9 pesos being paid to the captains in charge of amalgamation at
Regla. Duport, Métaux précieux au Mexique, 237.
467
avoiding pilfering of materials or of silver on the captains of each work team, with central staff
completing the oversight. Just three gatekeepers covering a weekly 24 hour supervision of the
gate traffic seems a bare minimum for an hacienda the size of Regla. In general it seems a very
lean operation for providing oversight over each stage of the process.
There is sufficient information in Table 5-XV to carry out a quantitative analysis of the
manpower requirements of each process, though the following assumptions are still required:
i) Data on total man-days and man-weeks, and the cost of wages and labour by day or
task, while ample are not complete. Man-days spent on the stamp mills are not specified,
though the total cost is accounted for. Since this cost is higher than the total cost of labour for
the arrastres, I have assumed that the man-days spent on the stamp-mills are at least as high as
those spent on the arrastres.
ii) The length of a shift in a working day is unknown. According to Duport, commenting
upon labour in the haciendas of Guanajuato in the 1840s:
‘The wages of the workers are 0.50 piastres [pesos] per day. They come to work on Sunday
evening, and do not leave until the next Sunday in the morning, or at the earliest on the Saturday
evening. There is no shift of workers for the day and night [work]’.844
His report coincides with the continuity in the monthly and weekly accounts that
indicate that Regla functioned year round. According to Mendizabal, the work-day of a peon
earning 0.5 pesos a day was 12 hours.845 The accounting figures in the Memorias are provided
as either man-weeks or man-days. I have decided to convert where necessary the figures to
844
‘Le salaire des ouvriers est de piastres 0.5 par jour. Ils entrent le dimanche á la nuit, et ne sortent plus jusqu’au
dimanche matin, ou au plus tôt le samedi soir. Il n’y a pas de relais d’ouvriers pour le jour et la nuit’. Ibid.
845
Mendizábal, "Minerales de Pachuca," 295. See also Brading, Miners Bourbon Mexico, 153-154.
468
man-days by assuming a six day week. I did not use man-hours since I have no information on
the length of the working day for manual and non-manual labour at Regla.
I partition the known costs and man-days according to the nature of each refining
process. In those cases where both processes would have shared the same activity, I apply a
factor based on the percentage of silver refined by each process during the month of May 1877
(see Chapter 4). Common units such as General Administration would have had a higher wage
bill if Regla had functioned solely as either a smelting or a patio amalgamation hacienda, than
that shown by the breakdown method I have adopted. The approximation is however sufficient
for the purpose of the exercise.
Tables 5-XVI to 5-XVIII show the labour structure (cost, man-days and total number
of workers, respectively) for Regla as a whole and then for each refining process according to
the method detailed above.
patio amalgamation smelting

Regla as a whole
(77% of silver refined) (23% of silver refined)
pesos pesos pesos
Stamp mill 324 10% 249 11% 75 8%
Arrastres 274 9% 274 12%
Patio Amalgamation 688 22% 688 30%
Capellina & casting of silver bars 28 1% 28 1%
General Administration 436 14% 335 15% 101 11%
Maintenance 625 20% 480 21% 145 16%
Planilleros / recovery amalgam 161 5% 161 7%
Stables 70 2% 54 2% 16 2%
Smelting 550 17% 550 61%
Extraordinary expenses 40 1% 30 1% 9 1%
total 3,195 100% 2,298 100% 898 100%
Table 5-XVI. Labour costs at Regla, according to the Memorias 18 – 22, and the deemed
distribution of labour costs between patio amalgamation and smelting.
469

Regla as a whole
man-days man-days man-days
Stamp mill 623 11% 478 12% 145 9%
Arrastres 623 11% 623 16%
Capellina & casting of silver bars unknown unknown
Maintenance 1007 18% 773 19% 234 15%
Stables 136 2% 104 3% 32 2%
Smelting 1116 20% 1116 71%
Extraordinary expenses 24 0% 18 0% 6 0%
total 5561 100% 3980 100% 1582 100%
Table 5-XVII. Labour man-days at Regla, based on data from the Memorias 18 – 22, and the
deemed distribution of man-days (manual and non-manual labour) between patio
amalgamation and smelting.

Regla as a whole
total workers total workers total workers
Stamp mill unknown unknown unknown
Arrastres 23 11% 18 12% 5 8%
Capellina & casting of silver bars unknown unknown
Maintenance 45 21% 35 24% 10 15%
Stables 5 2% 4 3% 1 2%
Smelting 52 24% 0 0% 52 73%
Extraordinary expenses 1 0% 1 1%
total 217 100% 146 100% 71 100%
Table 5-XVIII. Labour force at Regla, based on the Memorias 18 – 22, and the deemed
distribution of the workforce (manual and non-manual labour) between patio amalgamation
and smelting.
The main conclusions derived from these tables are:
1. Repairs make up 20% of total labour costs, an important amount second only to patio
amalgamation, and above the combined cost of milling and arrastres, or the labour cost for the
470
smelting section. Though the machinery at Regla was not complex, it involved many moving
parts, which explains the high labour cost of maintenance.
2. The labour cost of the planilleros belies the amount of man-days and personnel
expended in the search for amalgam and mercury that was entrained in the slurry from the
washing of the tortas. Nearly one fifth of the total man-days required for the patio
amalgamation was used in this recovery process. In number of workers it matched the
workforce on maintenance duties and visually it would have represented a group of people
three-quarters the size of the whole patio amalgamation work-force. This underlines the
thoroughness applied to the recovery cycle of amalgam and mercury lost into the water streams.
3. In the smelting process, the labour related to the furnace smelting of the ores
represent 60% of costs, and around 70% of man-days and workforce. There is no equivalent
stage in the patio amalgamation that absorbs so much labour power. Regla may be an anomaly
in this respect, since it possessed six blast furnaces at one time, and was clearly overdesigned
for the amount of ore it received for smelting. In addition, the need to maintain a trained group
of smelting craftsmen and support workers regardless of the actual amount of ore being
delivered converted their wages into a virtual fixed cost. 846 For this reason the value of 2.7
man-days per kg of smelted silver at Regla (1,582 man-days / 590 kg silver) is higher than the
value of 2 .0 man-days per kg of amalgamated silver (3980 man-days / 1947 kg silver).847 These
values must be treated as applicable only to Regla during the late 1870s.
846
Since during the late 1870s Regla was operating its patio amalgamation process to its fullest extent, the same
consideration did not apply to the workforce assigned to amalgamation.
847
Laur states that an average of 4.5 man-days per ton of ore (average silver content of 0.11%) are required for
the amalgamation process practised in Mexico by the second half of the nineteenth century, based on his analysis
of two one-year accounts for two refining haciendas (Zacatecas and Guanajuato). Laur, "De la metallurgie de
l'argent au Mexique," 202. This is approximately twice the value I have estimated to apply at Regla, and the
average days of the amalgamation runs may account for this.
471
4. Labour constituted a higher percentage of variable costs in smelting (22 to 26%) than
in amalgamation (7 to 17%). The difference in labour cost per kg of silver produced is a
reflection of the fact that the skills of a smelting crew converted it into a fixed cost, so that in
scenarios of irregular or limited smelting runs it would have been a drawback compared to
amalgamation. Can the introduction of amalgamation be in part explained as a result of labour
shortages? The results from Regla do not reflect a smelting operation run at the optimal level,
so do not provide a clear-cut answer to this hypothesis.
5.14 Patio vs Barrel Amalgamation
Barrel amalgamation was the adaptation of Barba’s cazo process carried out by Baron
Borg in Europe at the end of the eighteenth century. At the time it was implemented it led to
strong but very contrasting opinions on the utility of the process in New Spain. Humboldt was
one of the earliest supporters of this ‘Freiberg’ process, but pointed out that the sheer scale of
the ores waiting to be refined in New Spain dwarfed the extent of its application in Europe.848
Sonneschmidt on the other hand concluded that after ten years of trials in New Spain it had
failed to convince its users since it was more costly, extracted less silver than the patio process,
consumed more mercury [in contrast to the original cazo process on which it was based] and
produced impure silver.849 It was firmly in the mind of the English managers of the Adventurers
Company of Real del Monte from the very beginning of their project implementation stage, as
evidenced in the following extract from a letter dated 23rd August 1825 to Roger Morgan, Esq.,
at Regla from James Vetch at Mineral del Monte (Real del Monte): ‘proceed to state the plan
we have agreed upon for applying the Freyberg method to existing circumstances’. 850 Not all
848
According to Humboldt 60 thousand quintales were refined using barrel amalgamation in Europe, while in
New Spain 10 million quintales would have been the required quantity. Humboldt, Essai politique, Tome IV, 85.
849
Sonneschmidt and de Fagoaga, Tratado de la amalgamación de Nueva España, iii, x.
850
AHCRMyP, Fondo Siglo XIX, Sección: Correspondencia, Serie: Compañía a Varios, Subserie:
Correspondencia General, 8-1: 20 Abril 1825 – 1 Noviembre 1825. According to Duport, the barrel process had
472
the English managers would feel it represented a better process. In a letter from Rule to London
in 1842 he admits that none of the attempts to improve on the traditional patio process had
worked better than the original process itself when properly executed.851 Nevertheless John
Buchan proceeded to convince the new Mexican owners to implement it on a much greater
scale.852 The results reported were not positive. In the initial years the barrel process only
extracted 80% of silver in the ores.853 Prior roasting of the ore had to be applied, at least until
the 1860s.854
Though it never threatened to displace the patio process, its appearance at the time of
the Bourbon initiatives adopted to increase the production of silver, and then during the phase
of foreign capital investment in Mexico, has made it an obligatory reference in discussions on
the technical relevance of both historical events.855 The historiography does not include a
been tried out in Oaxaca and Bolaños prior to the 1840s. Duport, Métaux précieux au Mexique, 51. According to
Laur the Real del Monte Company was the first to apply in any major scale the barrel process in Mexico. Laur,
"De la metallurgie de l'argent au Mexique," 225. It was only applied in Oaxaca, Bolaños and Real del Monte,
according to John Phillips, Descriptive Notice of the Silver Mines and Amalgamation Process of Mexico.
Extracted from the Railway Register (London: Pelham Richardson, 1846), 16.
851
Randall, Real del Monte, 115-118.
852
853
854
855
From 1877 to 1892 barrel amalgamation contributed just 4% of total silver refined in Mexico. Flores Clair,
Velasco Avila, and Ramírez Bautista, Estadísticas mineras, II, 161-62. For a positive review of the Freiberg
process as a more modern alternative to the patio process see Sempat Assadourian, "Base técnica y relaciones de
producción," 429. For the view that the more effective Bourbon reforms were administrative rather than technical,
see Castillo Martos and Lang, Metales preciosos - union de dos mundos, 157, 183. In the case of foreign capital
investment in the nineteenth century, Velasco et al conclude that: [as a consequence of English investment] there
was no fundamental change … the introduction of the steam engine to pump water from the mines and many
attempts to substitute the patio amalgamation process that resulted in minor modifications in the methods to
reduce ores’. They add: ‘in the relations of production very few fundamental modifications took place … in the
refining processes … no notable change was implemented in the form of organization of the labour process’ - ‘no
hubo un cambio fundamental … la introducción de la maquina de vapor para el desague y muchos intentos por
sustituir el beneficio de amalgamación en patio que redundaron en modificaciones menores de los métodos de
reducción de minerales’ and ‘en las relaciones de producción, hubo muy pocas modificaciones fundamentales …
en los procesos de beneficio … no hubo un cambio notable en la forma de organización del proceso de trabajo’.
Velasco Avila et al., Estado y minería en México, 106-107, 250-251.
473
detailed economic comparison between the production costs of barrel amalgamation and the
traditional patio process, so the reasons for the failure of the barrel method to displace patio
amalgamation have not been quantified from an economic point of view.
An understanding of the economic basis to the permanence of the patio process against
the forces of change is needed to understand the course taken by the environmental history of
Mexico in the nineteenth century. I will quote at length from Randall, though at times it is
difficult to untangle views reported from views endorsed by this historian of the Adventurers
Company of Real del Monte:
‘those of them who knew anything at all about silver reduction methods, and in particular John
Taylor, were aware that the traditional patio amalgamation process of Mexico was woefully
inefficient and should be improved … they continued to believe that they could devise a
method of reducing silver ore better than the one employed for centuries at Real del Monte and
throughout colonial Mexico. They were right in both instances, but the company went under
…those company officials who had to cope with the larger problem of extracting silver from
ore in an economical manner were never comfortable with the patio process. They found it to
be slow and increasingly expensive (owing largely to the rising cost of quicksilver) when
dealing with common types of low-grade ore – and almost entirely useless when dealing with
those types that were called “rebellious” ... in a sense the English expended huge amounts of
time, energy and money in a fruitless effort to learn a Mexican trick they did not even like
[emphasis added]’856
It is difficult to know how much on silver refining was really known by these English
managers. John Taylor never set foot in Mexico, and England by 1854 was producing a paltry
70,000 pounds troy of smelted silver compared to Mexico’s output of 1,750,000 pounds troy
coming mainly from patio amalgamation.857. He became the editor of a journal that lasted all
of one volume, in which he published his new design for retorts to separate more efficiently
mercury from the silver amalgam at the haciendas of the company in Mexico, so as to replace
856
Randall, Real del Monte, 86-87, 115.
857
Production data from J. D. Whitney, The metallic wealth of the United States, described and compared with
that of other countries (Philadelphia: Lippincott, Grambo & Co., 1854), 506.
474
the capellinas then in use. The phrasing of the results of trials of the retort at Regla encapsulates
the stubborn bias against all local skills: ‘although … the loss of mercury was rather greater
than by the usual method of the country, yet … they may be rendered perfect’.858 The
capellinas were never replaced at Regla, or anywhere else in Mexico, for the retorts designed
by Taylor. A Mexican viewpoint on the level of English expertise is given in the following
extract: ‘the recently arrived English miners … in general do not know any other minerals than
copper, tin, iron and coal , and they are completely ignorant of the patio refining process which
is so important [in Mexico]’.859
Randall does not back up with facts his modern assessment that ‘they were right’ in
branding the patio process as ‘woefully inefficient’ and that the new English technicians were
predestined to devise a better refining alternative to the patio process. It is fortunate therefore
that the Estados Comparativos kept by the Compañia Real del Monte compare over a period
that spans twenty years, though with significant gaps in the currently available time series, the
variable production cost profile for the two types of amalgamation process with which the
silver ores from the Real del Monte mines were refined. The patio amalgamation process was
858
John Taylor, "Description of Retorts for the Distillation of the Mercury from Amalgamated Metals " Records
of mining I(1829): 142.
859
‘los mineros ingleses recién llegados … por lo regular no conocen otros minerales que los de cobre, estaño,
fierro y carbón, y que ignoran por completo el beneficio de patio que es tan importante’ in Viajero, "Las Minas
de Mexico," 182. An anonymous Englishman returning from Mexico published in England in 1856, under the
pseudonym ‘Traveller’ (Viajero), a scathing criticism of the way English capital had been extravagantly wasted
on Mexican mining ventures. In a footnote, the Mexican translator appended his opinion on the skills of English
miners. A modern Mexican historian writes: ‘the arrogant English investors and administrators … believed they
possessed a vast knowledge …. much greater than the aggregate practical experience of the Spanish and Criollo
owners [of refining haciendas] in Mexico. The economic failure of the English company can be explained to a
large degree on the blind faith placed on that assumption.’ - ‘los altivos inversionistas y administradores ingleses
… creían poseer un vasto conocimiento … mucho mayor del que habían acumulado en la práctica los propietarios
españoles y criollos en México. El fracaso económico de la sociedad inglesa se explica en gran parte por la fe
ciega en tales principios’. Herrera Canales, Velasco Avila, and Flores Clair, Etnia y clase, 7.
475
applied at Regla and Loreto, and Sanchez, Velasco and San Miguel were based exclusively on
the barrel (toneles) process.
The relative importance of each hacienda in the whole scheme of production for the
company during the period can be judged from Figure 5-25. Velasco is the main refining unit
for the silver ores of Real del Monte in the first years of operation of the newly revived
company under Mexican ownership. Velasco came to process using barrel amalgamation over
twice the amount of ore than any other of the refining haciendas. However, by the early 1870s
it suffers a marked decrease in the amount of ore processed. As seen in the previous chapter,
Regla would maintain an average of around 84,000 cargas per year until 1888.860
120,000
100,000
cargas of silver ore
80,000
Velasco (B)
Regla (P)
60,000
Sanchez (B)
40,000 San Miguel
Loreto (P)
20,000
0
1856
1857
1858
1859
1860
1861
1862
1867
1868
1869
1870
1871
1872
1873
1853
1854
1855
1863
1864
1865
1866
Figure 5-25. Annual average of cargas of silver ore processed at the refining haciendas of
the Compañia Real del Monte in the period 1853 to 1873. Source data from Estados
Comparativos.
860
It has been reported that 80% of all ores at Real del Monte were processed using barrel amalgamation between
1849 and 1862, but by 1877 this fraction had decreased to just 24%. It is claimed to increase again to around half
by the end of the century, mainly on the back of tolling of third party ores. In her first article Ortiz Peralta includes
a table that assigns quantities to the different refining haciendas, but it incorrectly classifies the production at
Regla as under toneles, barrels. This places some doubt on her percentages. Rina Ortiz Peralta, "El beneficio de
minerales en el siglo XIX: El caso de la Compañia Real del Monte," Tzintzun : Revista de Estudios Historicos 14,
no. Julio-Diciembre (1991): 77-79. In a second article that is basically a slightly edited version of the first, she
omits the suspect table but maintains the percentages. "Beneficio Minerales Real del Monte," 54.
476
What made both Velasco and the barrel process the main refining venue for the
company? The variable production cost for the barrel process averages from 10 to 13 pesos per
kg of refined silver during this period, as shown in Table 5-XIX. Velasco is the leaner of the
barrel operations, while San Miguel shows the highest unit costs. All require more pesos to
refine silver than the average registered at Regla using the traditional patio method. The lower
values observed for smelting correspond to ores with a higher silver content than those refined
by amalgamation, so they are not directly comparable to the others in the table. Otherwise, in
the case of patio and barrel amalgamation the comparison is being made on the basis of very
similar ranges of silver content in the raw ore during this period, as shown in Table 5- XX.
Sanchez (B) Velasco (B) San Miguel (B) Regla (P) Regla (S) Loreto (B) Loreto (P)
1853 11.53 11.34 15.75 12.52 8.03
1854 10.84 9.36 12.73 9.91 9.60
1855 11.34 9.36 11.93 8.20
1856
1857
1858
1859 11.51 9.27 11.92 9.56 5.91
1860 7.58 8.25 10.76 9.06 8.50
1861 8.16 9.04 7.93 9.41 7.96
1862 6.82 9.56 11.34 8.56 9.34 10.90 6.99
1863 9.20 11.53 14.02 8.91 8.91 15.25
1864 10.81 10.57 12.43 8.70 8.97 11.29
1865 10.76 10.81 15.28 9.25 11.15 11.85
1866
1867
1868
1869 9.40 8.86 14.59 10.76 14.48
1870 6.97 10.75 9.23 8.42
1871 12.41 7.81 14.19 9.35 7.77
1872 12.84 11.73 19.14 10.12 11.84
1873 17.10 11.25 15.26
average 10.25 10.10 13.06 9.65 8.71 10.90 11.46
Table 5-XIX. Variable production costs in pesos per kg of refined silver for the various
refining haciendas of the Compañia Real del Monte, in the period 1853 to 1873. The gaps in
grey indicate the haciendas were not in use at the time. The gaps in white indicate a lack of
primary sources for the period. Source data from Estados Comparativos.
477
Sanchez (B) Velasco (B) San Miguel (B) Regla (P) Loreto (P)
1853 0.20% 0.16% 0.15% 0.14%
1854 0.26% 0.24% 0.17% 0.15%
1855 0.27% 0.24% 0.19% 0.21%
1859 0.29% 0.28% 0.19% 0.22%

1860 0.36% 0.32% 0.19% 0.24%
1861 0.34% 0.26% 0.27% 0.23%
1862 0.28% 0.24% 0.22% 0.24% 0.38%
1863 0.25% 0.22% 0.17% 0.22% 0.19%
1864 0.22% 0.21% 0.21% 0.23% 0.26%
1865 0.23% 0.21% 0.18% 0.25% 0.26%
1869 0.13% 0.25% 0.14% 0.16% 0.17%

1870 0.34% 0.19% 0.19% 0.28%
1871 0.17% 0.33% 0.16% 0.29% 0.36%
1872 0.17% 0.21% 0.09% 0.24% 0.24%
1873 0.13% 0.18% 0.18%
average 0.24% 0.24% 0.18% 0.21% 0.26%
Table 5-XX. Silver content of ore before processing. Raw data from Estados Comparativos.
If we now compare the losses of silver incurred by each hacienda, the amount of silver
left unextracted in the processed ore, they confirm the initial judgment made in 1854 by John
Buchan that barrel amalgamation was the least efficient of the refining processes in extracting
silver from the ore. As the data in Table 5-XX indicate, during the first three years of operation,
on average up to one quarter of the silver could be left unextracted in the ores processed at
Sanchez and San Miguel, and only slightly lower at Velasco. Velasco, the more efficient of the
barrel process haciendas, still lost 40% more silver than at Regla using the patio process, and
double the silver losses via smelting at Regla.
Where did the advantage of the barrel amalgamation process lie with respect to the
traditional patio method? First of all, it is the claimed lower consumption of mercury, due to
the fact other reagents (copper, iron) reduce the silver chloride in the ore to silver. The
operational data from the haciendas both confirm and question this assumption. The average
478
values of the Hg to Ag weight ratio reported in Table 5-XXI show that barrel amalgamation on
the whole consumed about half the amount of mercury with respect to the traditional patio
process. However, the yearly data for the latter part of this period, when Velasco loses its
predominance as the refining flagship of the company, show mercury to silver ratios
commensurate with those registered for the patio process. Ortiz assigns this fact to the
elimination of the prior roasting of the ore implemented between 1868 and 1872.861 But this
begs the question whether in monetary terms the magic had ever existed, since we have seen
that variable unit production costs for barrel amalgamation were always higher than for patio
amalgamation. In principle there was a saving to be made in mercury in certain years (though
evidently not in all), but the saving had been lost through additional costs incurred by the
process.
A more detailed breakdown of the variable production costs for each hacienda during
this period shows what happened. In the case of the barrel process, the savings on mercury
were overshadowed by the greater spending on the fuel required to roast the silver ores with
salt, and then to heat the barrels. Thus the column in Table 5-XXII indicating the total pesos
spent on both mercury and fuel to produce 1 kg of silver is higher for the barrel process than
for the traditional patio process. Since the cost of salt is roughly equivalent for both processes,
it turns out that the barrel process offered no net advantages on cost even when it has saved at
times on mercury consumption. Why then was the barrel process ever implemented by the
Compañia de Real del Monte?
861
"Beneficio Minerales Real del Monte," 55.
479
Sanchez (B) Velasco (B) San Miguel (B) Regla (P) Regla (S) Loreto (B) Loreto (P)
1853 19.3 18.6 23.3 11.8 9.9
1854 24.8 17.0 25.2 11.3 7.0
1855 24.8 19.1 26.5 10.8
1856
1857
1858
1859 20.4 15.3 19.4 12.9 6.5
1860 19.6 16.0 16.8 13.9 8.0
1861 19.5 15.6 13.1 12.2 5.8
1862 20.8 12.1 19.4 11.4 5.6 5.8 3.1
1863 19.9 10.7 14.4 3.7 6.8 10.2
1864 12.1 13.7 24.4 4.7 4.9 0.0 8.1
1865 12.5 9.6 21.9 7.6 2.6 0.0 1.4
1866
1867
1868
1869 9.1 9.1 12.6 6.0
1870 11.5 13.7 6.2 11.5
1871 12.4 12.8 12.0 10.0 7.2
1872 7.2 11.3 12.0 11.9 8.9
1873 13.4 7.0 6.5
average 16.7 14.0 17.6 9.9 6.9 5.2 6.1
Table 5-XXI. Losses of silver, expressed as weight percentage, registered at the refining
haciendas of the Compania Real del Monte, in the period 1853 to 1873. The gaps in grey
indicate the haciendas were not in use at the time. The gaps in white indicate a lack of primary
sources for the period. Source data from Estados Comparativos.
Sanchez (B) Velasco (B) San Miguel (B) Regla (P) Loreto (P)
1853 0.71 0.72 0.59 1.89
1854 0.61 0.63 0.67 1.61
1855 0.88 0.70 0.95 1.57
1856
1857
1858
1859 0.83 0.66 0.37 1.69
1860 0.51 0.57 0.56 1.65
1861 0.58 0.62 1.01 1.83
1862 0.61 0.62 0.40 1.74 1.16
1863 0.58 0.62 0.73 1.72 1.56
1864 0.71 0.62 0.63 1.42 1.53
1865 0.63 0.88 0.45 1.44 1.42
1866
1867
1868
1869 0.26 1.21 1.15 1.50 1.81
1870 1.48 0.83 1.40 1.33
1871 1.52 1.47 1.46 1.61 1.83
1872 1.52 1.52 1.26 1.49 1.79
1873 0.67 1.38 1.63
average 0.77 0.87 0.79 1.59 1.56
Table 5-XXII. Mercury to silver weight ratio registered at the refining haciendas of the
Compania Real del Monte, in the period 1853 to 1873. The gaps in grey indicate the haciendas
were not in use at the time. The gaps in white indicate a lack of primary sources for the period.
Source data from Estados Comparativos.
480
pesos spent to produce 1 kg silver by amalgamation
mercury salt copper sulphate fuel others+labour total mercury+ fuel
Sanchez 1.0 2.0 0.0 2.7 4.5 10.25 3.7
Velasco 1.4 2.2 0.0 2.4 4.1 10.10 3.9

San Miguel 1.2 3.2 0.0 2.7 6.0 13.06 3.9
Regla 2.5 2.2 0.9 0.2 3.8 9.65 2.7

Loreto 2.8 2.3 0.8 0.2 5.3 11.46 3.0
Table 5-XXIII. The average amount of pesos required to refine 1 kg of silver using the
two amalgamation processes. The haciendas in italics used the barrel process, and the
haciendas in normal script used the traditional patio amalgamation. The data has been
calculated for the period 1853 to 1873 using as source the Estados Comparativos.
Buchan justified the choice not on economic grounds but by assuming it was the only
way to treat major amounts of the recalcitrant silver ores being produced by the mines at Real
del Monte, in spite of the evident economic drawbacks that are clearly spelt out in his report of
1854.862 This is a more realistic appraisal of a necessary evil than an endorsement of the barrel
process. In addition, by cutting down on the amalgamation period it could compensate its low
rate of extraction with a higher output of silver. It may be significant that Velasco retained its
pre-eminence when it was able to process 2.5 times the amount of ore compared to Regla. Once
their throughput became very similar, for example towards the end of the 1853-1873 period,
Regla would have been a much better economic choice to process ores, under equal chemical
conditions.
862
Buchan, Report Real del Monte, 16-18.pp. The table with the comparative costs of production appears on page
17 of his report. Prior roasting followed by patio amalgamation may have been another option. According to
Duport, the silver ores that did not respond well to patio amalgamation in Zacatecas were the black or red sulphides
of antimony and silver, which could retain up to 40% of their silver after amalgamation. Duport, Métaux précieux
au Mexique, 246.
481
I would seriously question that it was ‘the most important innovation left by the English
who exploited the mines of Pachuca and Real del Monte’.863 I would argue that the reduction
of fuel requirements by a factor of five in the smelting furnaces merits that distinction. I
question even more strongly Randall's conclusion that ‘the firm succeeded in making lasting
technical advances … in the treatment of low-grade silver ore’ and that ‘it definitely improved
on Mexico’s traditional ore-reduction method … to supplant altogether the traditional Mexican
method of treating silver ore ... from the patio ... to barrels, where it was quick and less
destructive of that costly commodity [mercury]’, when the accounts of Regla offer such a clear
refutation.864
The barrel process as practised at Regla did not always reduce the consumption of
mercury, it never decreased the amount of unextracted silver left in the ore after treatment
(quite the contrary) and did not provide an ample margin of profit based on its variable
production cost, as has been claimed.865 Its capacity to treat difficult ores owed more to the
prior roasting with salt than to the barrel amalgamation itself. 866 It definitely never displaced
patio amalgamation as the most cost-effective option to refine the majority of Mexico’s silver
ores.867 The irony remains that even some who have recognized its failure still appeal to the
‘timelessness’ of the Mexican scenario. ‘In Europe where speed was important and labour
863
‘fue la innovación mas importante legada por los ingleses que explotaron los distritos mineros de Pachuca y
Real del Monte’ Ortiz Peralta, "Beneficio Minerales Real del Monte," 52-53.
864
Randall, Real del Monte, 87, 109, 118.
865
Ortiz Peralta, "Beneficio Minerales Real del Monte," 53.
866
Barba had recognized nearly three centuries before that the Cazo process (on which the barrel process is based)
worked better with ores high in silver content, either elemental silver or silver halides, as was the case in Catorce
at the end of the eighteenth century. Barba, Arte de los metales, 111-12. Roasting with salt converts silver
compounds in the ore to silver chloride (a silver halide), as indicated in Chapter 3.
867
It never achieved a major penetration of the refining market in Europe. In Spain ‘the most extensive
amalgamation (60 barrels) at La Bella Raquel, span ores from Hiendelaencina [Guadalajara]’.Kerl, Crookes, and
Röhrig, Prof. Kerl's Metallurgy, 331.
482
expensive, the much faster … more wasteful Born process was preferable’ while in Mexico
‘time was not important and labour cheap … [so that] it was more profitable to use the slower,
less wasteful patio process’.868 The barrel process did not fail because ‘time was not important’
in Mexico. It failed for the simple reason the patio amalgamation process was never a ‘Mexican
trick’ but an industrial process whose production costs per kg of silver refined from the typical
ores of Mexico were lower than those of the European barrel process. It is fitting that John
Phillips in 1846 acknowledged that ‘with respect to the loss sustained by English Companies
in the prosecution of mining undertakings in Mexico, that much of it has arisen from the
circumstance of their not having given due credit to the Mexicans for skill in the application of
the means they possessed’.869 In general the comparison of production costs on both sides of
the Atlantic needs to take into account that the refining processes in New Spain / Mexico had
to overcome the challenge that only the value of silver could cover the production costs, without
the additional revenues from lead or copper as in Europe (see Appendix F). The important
exception was in those cases where additional revenues from gold found together with the
silver aided substantially the refiner in covering his costs (see example in Section 2.5).
This chapter has been a search for a long overdue quantitative answer to the question
as to whether economic drivers determined a clear choice between amalgamation and smelting
in the New World. The environmental impact from each is so different that one would expect
that the answer can explain to a great extent the direction taken by the environmental history
of silver refining in the Americas. Thanks to the detailed accounting records kept at Regla it is
868
Clement G. Motten, Mexican silver and the enlightenment (Philadelphia: University of Pennsylvania Press,
1950), 53. The stereotype on time and Mexico is repeated by other authors: ‘[amalgamation] costs little … can be
worked on a larger scale as well as a small one… cheapness of plant compensates for the time … in Mexico …
time has no value’. Egleston, The Metallurgy of Silver, 311.
869
Phillips, Descriptive Notice, 20.
483
now possible for the first time to base an answer on a very long time series of operational data
corresponding both to patio amalgamation and smelting, as practised in one of the largest
refining haciendas in Mexico of the second half of the nineteenth century. I had no choice on
the span of years covered by the extant documentation, but was fortunate to find it included not
only a sudden break in operations that forced a smelting of discarded slags, but also a span of
more than ten years of the hacienda operating at full capacity as far as amalgamation was
concerned. In addition, from 1872 to 1888 the mercury market seemed to parody the whole
span of Spain’s pricing during the estanco, in a manner showing that in colonial New Spain
prices were not so different from what market drivers could determine in the nineteenth
century. With regards to smelting, Regla was the pièce de résistance of the whole stable of
refining haciendas, the only one designated to receive ores fit for smelting. The most important
technological innovation of the English investment and know-how in refining was the
installation of blast furnaces run with such efficiency they cut the consumption of charcoal by
a factor of five.
In the previous chapter I argued that in its operating essence, the nineteenth century
hacienda at Regla is a twin sister to any other hacienda of the previous 300 years, and not a
distant and unrecognizable relative. Thus the diagnostic carried out on its profile of production
costs could be used to recreate the competition over the centuries of amalgamation and
smelting. Thanks to the initial stability over two centuries of the silver to gold ratio, and the
immutability up to 1888 of production costs to the collapse of silver prices, I was able to ignore
deflaction over three centuries without jeopardizing the comparisons. During this extended
period amalgamation and smelting jostled to win not the hearts but the pockets of the mixed
patchwork of miners and refiners of New Spain, a technical jousting unique in its longevity to
the history of technology. For some 350 years neither technology (the traditional smelting and
the newcomer amalgamation) changed in a substantive manner, and neither managed to

484
displace the other. The balance observed between the two refining processes up to the
nineteenth century (Chapter 6) was a clear indication that many times these hard-headed, self-
made but very pragmatic individuals had found little to choose between them, at least on
economic terms.
The answer that the accounts from Regla have provided will bolster the argument to be
rounded off in the next chapter that the relation between amalgamation and smelting was not a
one-dimensional, zero-sum game. In other words, increases in production by amalgamation did
not come necessarily at the expense of smelting, or vice versa. For much of the time and in
many locations, amalgamation did offer the best cost alternative to extract silver from ores with
a minimum silver content in the range of 0.1 to 0.2%. The only period mercury was the major
influence on the costs of amalgamation in New Spain was in the second half of the sixteenth
century, when tailings were being raided and the Crown sold mercury for the highest price it
could get, without even having to pay for part of it. For the next three hundred years it was the
extraction cost of the ore that determined the profit margin of a refiner, as had been pointed out
by Villaseñor. This remains a valid statement even if the implementation of the tolling business
allowed refiners to divorce part of their profits from the mining cost involved. Reducing the
price of mercury increased the impact of salt prices on the total cost structure until they became
equal in importance, around 10% of total cost each at Regla, and yet salt was never the intense
focus of the lobby by refiners as was mercury. The cost of power for amalgamation haciendas
completes the trilogy of production costs that most influenced the margin of profit, yet it is
even more absent than the cost of salt from many a modern analysis. Copper sulphate, or its
less refined variety by the name of magistral, was the most cost effective of all ingredients for
the process. Fuel was a very minor contributor to the cost of amalgamation, thus isolating the
process from the vicissitudes of searching for sufficient woodland resources, as well as making
its impact on wood resources negligible. As to the fixed cost of capital, it would seem that once
485
the major hurdle of raising capital was cleared, its service was not a major burden on production
costs. As always, regardless of the breakdown of production costs, the need to continuously
provide working capital can make or break a business, thus the importance given in the
historiography to the allocation and sources of capital in the history of silver refining in the
New World.
Smelting remained throughout a viable option, first of all for the obvious reason that
for some ores rich in lead it was the only effective way to refine them. The minimum silver
content that made smelting profitable lay in the range of 0.3 to 0.5%, though in practice at
Regla no ores below 1% silver were smelted. The next chapter will underline the important
role played by smelting in New Spain, a fact that gives credibility to the proximity of
production costs for both processes proposed in the calculations of this chapter. Thanks to the
data from Regla I have been able to show quantitatively that the amalgamation and smelting
cost curves, as a function of silver content of the ore, virtually merge into each other under
certain conditions approximating those likely to have occurred in the period between the
seventeenth to eighteenth centuries. The capacity to scale down smelting operations and labour
requirements, compared to the more ponderous nature of amalgamation, made smelting a very
versatile option, subject mostly to the ability to access cheap and sufficient fuel.
More data on other haciendas are necessary to establish whether Regla was an isolated
case of a very efficient patio amalgamation operation, or whether patio amalgamation was a
more potent refining process than has been credited. Taylor had identified back in the 1820s
the two major challenges the English faced at Regla in order to implant smelting as the
preferred option: dressing the ore to increase its silver content, and finding enough fuel to feed
the blast furnaces. Smelting of silver ores made sense in the England of the Cornishmen and
the Erzgebirge of the Germans because the sale of lead and copper made profitable their
economies of production. In New Spain / Mexico the whole weight of meeting the costs of
486
smelting fell only on silver and whatever gold was present in the ore. This was the challenge
faced by smelting, not the scale of available ore that loomed so large an obstacle in the
imagination of Humboldt. Argentiferous copper was never a major component of Mexico’s
silver ore deposits, in contrast to the European scenario, so the production of copper was not
an option to defray the costs of refining silver in New Spain or Mexico. The production of lead
for the market would only become important in the twentieth century.
I cannot find any other example of a commodity that from the sixteenth century was
produced both in Europe and the New World, under an industrial context where the market
price was fixed at basically the same level for all producers on both sides of the Atlantic during
three hundred years. During this time refiners had to fit their local variations in wages, fuel,
reagents, infrastructure costs and government duties into a box of just one size, the valuation
of silver that had remained nearly unchanged for over twelve generations of refiners. Under
these conditions it makes no sense to analyse the historical scenario of the cost of production
from the point of view of the price elasticity of silver in the market, and this reality certainly
tied the hands of producers on both sides of the Atlantic. It would take the appearance of a huge
wave of silver suddenly coming to the market from the refining mills of Nevada in the 1870s
to shake the whole silver price structure to its roots.
At any given time the volume of silver in the market depends on what percentage in
known deposits can be profitably mined and refined given the production costs and efficiency
of the available technology. The production cost of smelted silver prior to the sixteenth century,
the technical limits to the smelting process applied, and the market for lead and copper played
a major role in fixing the volume of silver historically available to Europe, and thus of the final
silver to gold ratio. Amalgamation was not the process that historically set the price of silver.
It never progressed beyond a fringe operation in Europe as of the end of the eighteenth century,
and smelting was always the preferred choice.

487
It would be logical to argue that a lower wage scale in the New World compared to
Europe would have been a significant help to keep competitive the variable costs of
amalgamation, of which around two thirds are direct labour costs, or of smelting, at one half.
This fact should not overshadow recognition of the efficiency of the mining workforce in the
New World, clearly quantified in Humboldt’s data at being some ten times more efficient than
their German counterparts, though this factor was silenced in his commentary. It turns out that
time was just as important in Mexico as in Europe. To their efficiency must be added their
attrition, and those of their communities, in human lives, occupational diseases and
environmental damage from mining and refining, a hidden but very real contribution to
maintaining a competitive production cost of silver.
By the nineteenth century European silver production had recovered, and now even
offered better smelting economies for ores imported from the Mexican mines. In the tradition
of the hare and the tortoise, smelting would outlive the amalgamation of silver in New Spain.
Looking back, it is tempting to conclude that the environmental history of the New World
veered in a totally new direction as soon as some of the first miners of New Spain found they
were left with more coin at the end of the day if they used cold mercury instead of a hot and
fussy furnace. In this case, however, chemistry trumps economics. Lead would continue to be
the main environmental hazard for silver refiners, a heavy metal issued to the air in New Spain
in greater quantities than any volatile emissions of mercury. How much, when and where is the
subject of the next and final chapter.

488
6 The environmental impact of silver refining: a shift of paradigm.
‘There’s the story, then there’s the real story, then there’s the story of how the story came to
be told. Then there’s what you leave out of the story. Which is part of the story too’. Margaret
Atwood, MaddAddam (2013)
6.1 The magnitude of the impact: the method employed
Of all the chemical substances that shape the environmental history of silver refining in
in New Spain, only the production of silver and the consumption of mercury constitute a fairly
continuous data set in the official tax records. Both data sets can be used to reconstruct the
quantities of other chemicals and matter voided into the environment, using the ratios built up
in the preceding chapters and the theoretical base of the correspondencia. The more
fundamental problem lies in that the whole framework within which silver production and
mercury sales were registered was tinged with corruption:
‘The Royal Treasury in New Spain was a very efficient organization, carefully controlled and
with very precise working guidelines, but at the same time it was a centre of corruption and
traffic of influences. Then as now, these two aspects coexisted without interfering with each
other, so that as long as the accounting figures matched precisely, other parallel practices were
accepted so as to privilege certain miners in the distribution of mercury or in the evasion of
taxes’.870
Rampant corruption and contraband have been pointed out in the historiography as
distorting the historical production data, so that up to an estimated two thirds of the total silver
870
‘La Real Hacienda en Nueva España era una organización eficiente, controlada cuidadosamente y con normas
de trabajo muy precisas, pero al mismo tiempo era foco de corrupción y tráfico de influencias. Entonces como
ahora, estos dos aspectos coexistían sin estorbar el uno al otro, así que mientras los datos contables cuadraban
en forma precisa, por otro lado se aceptaban cohechos para privilegios a algunos mineros en el reparto de
azogues o para evadir impuestos’ in Pérez Luque and Tovar Rangel, Caja Real Guanajuato, 13. The phrase ‘then
as now’ is a warning that accounting coherence on paper does not have the same implications even at present in
different parts of the world.
489
registered has been deemed to have been excluded from the official record of New Spain.871 I
have proposed in a separate paper that any major disparity between the installed refining
capacity and official registered silver production can be used to estimate a probable degree of
contraband silver, since private capital, contrary to State finances, cannot sustain a fixed capital
investment frozen into an unproductive facility. In the case of Potosí, I estimated the installed
production capacity could have produced twice the official registered figure in the period 1576
to 1650.872 I have no such option to cover the whole period of silver refining in New Spain /
Mexico. I have therefore relied on the official registers for silver production available in the
historiography. These values should be interpreted as the base line, with real levels expected
to have been higher, so that the absolute magnitude of the environmental impact vectors could
have been substantially greater than any estimates to be calculated in this chapter. The relative
magnitudes of these vectors to one another however remain a valid guide to their relative
impact on the environment.
TePaske authored one of the most complete reviews of published data sets and provided
a compilation of the production of silver in New Spain by Caja (regional Treasury, usually but
not always a mining district), from the sixteenth century up to independence. 873 Refining
haciendas sent their silver to be taxed and stamped at the Cajas. Between 1521 and 1810 the
Cajas of interest are, in descending order, Zacatecas (20.6% of total colonial silver production),
Guanajuato and Mexico (17.3% each), Durango (12.1%), San Luis Potosí (8.8%), Guadalajara
871
Flynn and Giraldez, "Cycles of Silver: Global Economic Unity through the Mid-Eighteenth Century," 435.;
Hausberger, Metales preciosos, 41-44.; Bakewell, "Registered Silver Production in the Potosi District 1550-
1735," 80.
872
Guerrero, "Contraband of Silver from Potosí and Oruro," 76-77.
873
TePaske and Brown, Gold and Silver, 82, 115-16. Care must be taken with the pie chart of Figure 3.5 in
TePaske’s work that breaks down the total silver production expressed in kg for all the main Cajas of New Spain,
since for the Caja of Mexico the figure used is 3,703 thousand kg, which does not correspond to the total silver
production according to Table 3-4 (8,703 thousand kg). It is most probable the 8 became a 3 at some point of the
posthumous editing process. This skews all the percentages values reported in the pie chart. The data in his Table
3-4 should be used instead.
490
(7.8%), Pachuca (5.4%), Sombrerete (3.7%), Bolaños and Rosario (2.3 and 2.2 % respectively),
Zimapán (1.7%) and Chihuahua (0.5%). The Cajas however were not fiscal districts set in
stone. Their hierarchy and the extent of the regional spheres of each Caja changed over time,
as the Reales de Minas even within a region.
The environmental history of silver refining in New Spain / Mexico is thus a fluid
mosaic, where the content and nature of each element constantly change over time, reflecting
the temporal idiosyncrasy of each mining district reporting to the Cajas. Hausberger, from
whose work I have adapted the map shown in Figure 6-1, anchored his analysis of the silver
industry in New Spain using a comparative regional analysis of the data, an approach that will
also be adapted in this chapter.874 Figure 6-1 shows very clearly the disparate nature of these
regional environmental pressure points. For example, the magnitude of the impact vectors to
be calculated from the data in the Cajas of Guanajuato and Pachuca impacted a much more
reduced geographical area than the case of Durango and Guadalajara. Furthermore, while the
haciendas reporting to the Caja of Guanajuato were within or close to the city of Guanajuato,
in the case of Pachuca only the Hacienda de Loreto was within the city limits, and all the other
refining activity was well away from the main urban centres. To this geographical diversity
must be added a technical diversity, since amalgamation and smelting were not applied
uniformly across all the regional Cajas, or even within the historical period of interest for a
single Caja. This temporal diversity will become evident in the following sections.
The data on silver production in the nineteenth century are not as detailed as the data
for the colonial period. Its records suffered on par with the political situation in Mexico, and
the most detailed statistics are found only for the latter part of the century. I have set out in
874
Hausberger, Metales preciosos, 64, 82-87.
491
Table 6-I the main sources in the historiography on the production of silver in this period. The
figures calculated by Soetbeer up to 1875 are the source quoted for other estimates published
up to this year.875 For 1876 up to 1899 I include the data from Gonzalez Reyna and from Flores
Clair, Velasco Avila and Ramirez Bautista.876 I have not found an equivalent breakdown of
data by region as is available for New Spain, so only a gross national approximation can be
attempted on the basis of the total silver production from 1820 to 1899.
II
VIII
IX
VII
I
IV X
I Bolaños
II Durango III VI
III Guadalajara
IV Guanajuato V
V Mexico
VI Pachuca
VII San Luis Potosí
VIII Sombrerete
IX Zacatecas
X Zimapan
Figure 6-1. Map of regional Cajas of main mining districts in New Spain, snapshot of the
1760s, adapted from original map in footnote 874.
875
Adolf Soetbeer, Edelmetall-produktion und werthverhältniss zwischen gold und silber seit der entdeckung
Amerika's bis zur gegenwart (Gotha J. Perthes, 1879), 49-60. The other works that rely on Soetbeer’s data are
Merrill, Summarized Data of Silver Production ; Jenaro González Reyna, Minería y Riqueza de México,
Monografías Industriales (México: Banco de México, 1944).; Flores Clair, Velasco Avila, and Ramírez Bautista,
Estadísticas mineras, II.
876
González Reyna, Minería y Riqueza de México.; Flores Clair, Velasco Avila, and Ramírez Bautista,
Estadísticas mineras, II.. The data from the latter are presented also in Velasco Avila et al., Estado y minería en
México.
492
silver production (kg)

Period González Flores Clair
Soetbeer
Reyna et al
1801 - 1810 5,538,000
1811 - 1820 3,120,000
1821 - 1830 2,648,400
1831 - 1840 3,309,900
1841 - 1850 4,203,100
1851 - 1855 2,330,500

1856 - 1860 2,239,000
1861 - 1865 2,365,000
1866 - 1870 2,604,500
1871 - 1875 3,009,000
1821-1875 22,709,400
1876 546,410 570,000
1877 588,518 635,572
1878 610,683 641,502
1879 643,907 704,783
1880 694,000 756,505
1881 714,573 743,372
1882 718,658 756,345
1883 748,681 810,448
1884 793,377 849,579
1885 824,079 873,996
1886 876,724 959,215
1887 939,779 1,005,080
1888 986,382 1,051,995
1889 983,799 998,742
1890 990,237 1,068,088
1891-1899 13,250,485 13,796,861
1876 - 1899 24,910,292 26,222,083
1821-1899 48,931,483
Table 6-I. Silver production in Mexico, nineteenth century. Sources from footnotes 875
and 876. The data after 1875 corresponds to fiscal years beginning in the year indicated.
493
Partial breakdowns are available in the historiography that separate production as a
function of amalgamation or smelting, but none that spans the whole colonial period and all
the Cajas. The following sections present an estimate of the silver produced by each refining
process based on the raw primary tax and mercury sale data of each Caja as has been gathered
and transcribed by Tepaske, Klein and other collaborators in Mexico and Spain.877 Their data
set (to which I will refer henceforth as the TK set) includes the duties and senoreaje (duty on
coinage) identified as coming from silver refined by amalgamation (plata de azogue) and those
coming from smelted silver (plata de fuego). It does not distinguish between patio
amalgamation and amalgamation using the cazo process (as in Catorce, Caja of San Luis
Potosí). In addition, the records also show the amount of revenues from the sale of mercury by
each Caja. These can be converted into an approximate weight of mercury using the reported
price ranges for mercury for each period in question.878 Freight does not have to be subtracted
from these values since this was a separate cost and included as such in the accounts of the
Caja.879 The fact that silver taxes and mercury sales are registered for a same period does not
guarantee these amounts are correlated, due to credits, late payments by refiners, inventory
build-ups, or even contraband of mercury. By aggregating the mercury data into a decade
whenever possible, the potential for mismatches will tend to be minimized but not eliminated.
With this limitation in mind I report approximate mercury to silver ratios in the following
sections.
877
Prof. Herbert S. Klein of Columbia University very kindly sent me his Excel files with the raw information
collated from primary sources on tax revenues and mercury revenues during the colonial period for the provincial
Cajas. Any error in the sorting and calculations based on the raw data transcribed in his files is solely this author’s
responsibility.
878
For the calculation of the total weight of mercury I have used the Almadén price even for the small amounts
of mercury brought from Idria and Peru.
879
‘when selling to miners mercury and freight were charged separately’ – ‘al vender a mineros se cobra aparte
azogue y fletes’ in Pérez Luque and Tovar Rangel, Caja Real Guanajuato, 37.
494
The separation of silver royalties based on refining process starts to be reported in the
late seventeenth century. For the earlier periods where no such distinction was made in the
Cajas, I project where necessary a probable split using the information available on sales of
mercury or extrapolating from known ranges of the mercury to silver ratio. Once I have
estimated the periodic regional production of silver by amalgamation and smelting, I proceed
to estimate the magnitude of the main environmental impact vectors based on the ratios per kg
of refined silver calculated in the previous chapters.
The interpretation given to the historical split between amalgamation and smelting
observed for each Caja will first take into account the nature of the silver ores (rich in lead or
dry). Changes over time in the nature of the ore, or the introduction of lead fluxes from another
region, can be a leading cause for a change in the balance between refining processes.880 The
change of registry of the output of mines from one Caja to another will also impact the
amalgamation/smelting split by changing the nature of the ores being processed.
The impact of the pricing of mercury is easier to determine as the main cause of the
observed refining split in those Cajas where amalgamation historically predominated, since the
nature of the ores is seen historically to be amenable to amalgamation and the infrastructure
was already in place to immediately take advantage of the lower prices. 881 In general mercury
pricing was not the only factor nor was smelting the second-best option for refiners of the silver
880
For example, refiners in Zacatecas used lead flux sourced from Zimapán, as cited in Mendizábal, La mineria
mexicana, 72.
881
More aggresive aid than a simple general price decrease could keep amalgamation as the refining option over
smelting. In the late 1760s Jose de la Borda in Zacatecas was granted mercury at 30 pesos / quintal, no silver tax
was levied during renovation work on his mines, and the silver tax on his production was decreased by 50% for
20 years. Brading, "Mexican Silver Mining," 671. Even as the price of mercury was being decreased policy
makers were still debating whether to implement a complete switch from amalgamation to smelting in New Spain,
though problems in the supply of mercury may have been the real issue. Arthur P. Whitaker, "The Elhuyar Mining
Missions and the Enlightenment," ibid.31(1951): 573-576.
495
ores of New Spain or Mexico.882 The production cost curves derived in Chapter 5 show a
sufficiently fine line between amalgamation and smelting to allow them to compete on profit-
margins subject to location and time-specific costs, availability of ores and reagents. Major
capital requirements would not be required if there was no actual switch from one refining
process to the other, if only the quantities being processed by each type of hacienda varied with
time.
6.1.1 Caja of Zacatecas
The tax revenues on silver produced are reported in pesos in the TK set for the Caja of
Zacatecas under ‘1% y Diezmos de Plata de Azogue’ and ‘1% y Diezmos de Plata de Fuego’.
The fraction of silver refined by amalgamation and smelting is thus calculated directly from
the peso amounts registered under each heading in the same time period. These appear in Figure
6-2 and all subsequent plots of the following Cajas as the data points under the heading ‘duties’.
In addition, in certain periods a tax on the minting of coins (señoreage or señoreaje) is included
under the same rubric, as ‘1% Diezmos Señoreage Plata de Azogue’ or ‘1% Diezmos Señoreage
Plata de Fuego’.883 On the assumption that these amounts have not distorted the original split
of production between amalgamation and smelting, these data points are plotted under ‘duties
& coin’ in Figure 6-2 and subsequent plots of this nature. The smooth merging observed
between both sources of data confirms empirically my assumption. For the case of Zacatecas I
have also included the data reported by Lacueva, who has recently published annual data sets
of silver production in Zacatecas, since it covers a period prior to 1700 when the time periods
882
As an example, at the end of the nineteenth century the switch to smelting with coal as of 1893 increased in a
few years silver production from 39 to 74 million kg, according to Mendizábal, La mineria mexicana, 106.
883
‘señoreage : a tax on the minting of coins by private individuals’ - ‘señoreage : impuesto sobre la acuñación
de monedas por particulares’ in Pérez Luque and Tovar Rangel, Caja Real Guanajuato, 61. Also defined as a
mintage tax in Bakewell, Silver Mining in Zacatecas, 245.
496
for the TK set do not always correspond exactly to twelve month intervals.884 There is no
evident misfit between all sources of data for this period.885
1.00
0.90
0.80
amalgamation fraction
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1660 1680 1700 1720 1740 1760 1780 1800 1820
duties duties & mintage lacueva
Figure 6-2. The fraction of total silver refined by amalgamation in Zacatecas in the period
1670 to 1820. Prior to 1700 the time intervals are exactly one year only for the Lacueva data.
Fractions calculated by the author on the basis of the raw data from TK set and footnote 894.
The profile of Figure 6-2 indicates a predominance of smelting in the second half of the
seventeenth century shifting to a complete reversal between smelting and amalgamation by the
end of the eighteenth century. The interpretation of the alternations between amalgamation and
smelting in Zacatecas based on mercury alone has been questioned already. 886 The reason for
884
885
Bakewell published a table with percentage values of silver produced by smelting for some of the main Cajas.
He does not provide the source of the two values he provides for each Caja, for the 1720s or 1730s and for the
1760s of 1770s. I will be citing his values in the footnotes of the following sections, and overall they confirm my
own findings. The data presented in these sections span a much wider period than Bakewell’s data, which allows
a more general picture to emerge. For Zacatecas he states that smelting produced around 30% of its silver both in
the 1720s and 1760s. Bakewell, "Colonial Mining," 145.
886
Lacueva has argued that Bakewell, the leading proponent of the mercury argument, ignored the contribution
of smelting per se in Zacatecas irrespective of the supply of mercury. See Lacueva Muñoz, La plata del Rey, 187-
210. For other proponents of the mercury argument see Brading and Cross, "Colonial Silver Mining: Mexico and
Peru," 573-78. ‘It is not a surprise then that in the eighteenth century, when production in New Spain reached its
highest ever historical levels, the two processes had been working side by side, and the switch between one and
the other subject mostly to the price of mercury.’ In Blanchard, Russia's "Age of Silver". Precious-metal
Production and Economic Growth in the Eighteenth Century 3-31.
497
the change in refining process is quite straightforward. As of the 1680s the lead rich ores of
Sombrerete were no longer registered at the Zacatecas Caja, and the effect of this reaasignment
after 1680 of ores that had until then been smelted and registered at Zacatecas is evident in
Figure 6.3. Amalgamation is seen to peak in the 1730s and then responds to the price decrease
of mercury in the 1780s. There is no evidence to suggest a zero sum game, where refiners
switched from smelting to amalgamation when mercury prices dropped, at the expense of the
amount of ore smelted. The nature of the ore prevailed: ores that could be smelted continued
to be smelted, and a lower mercury price made it profitable to process more of the ore that
could be amalgamated. On average over this period amalgamation accounted for two thirds
and smelting for one third of total silver produced, but during the time the ores from Sombrerete
were included in this Caja, the fraction of smelted silver registered in the Caja represented
over half the total of silver produced.
600,000
500,000
400,000
kg silver
300,000
200,000
100,000
smelting amalgamation
Figure 6-3. Registry of silver at the Caja of Zacatecas according to refining process. Data
from Table 6-II.
In order to correlate the production of silver with the available data on mercury sales
broken down by period, I have repeated with the TK set one of the classical calculations carried
498
out in the historiography of silver refining, the conversion of silver tax revenues into a weight
of silver production.887 In parallel I have estimated the weight of mercury (in kg) consumed in
the amalgamation of silver ores by dividing each registered sale revenue by the price of a
quintal of mercury.
Table 6-II is the first example of how all these projections are combined. The data
plotted in Figure 6-2 are used to generate average values of the mercury to silver weight
according to each time period. The average of this ratio for the period 1670 to 1810 is 2.05,
and this value is extrapolated for the period prior to 1670.888 The weight of silver produced by
amalgamation is calculated using this ratio and the amount of mercury calculated from the sale
figures. The balance from the total silver produced according to the tax registries provides the
weight of silver obtained by smelting. Prior to 1651, the TK set does not segregate data on
silver according to refining process. For the period 1611 to 1650, I make use of two sets of data
in the historiography: total silver produced as reported by TePaske, and the amount of mercury
distributed to Zacatecas from late 1608 to 1649, as reported by Bakewell (these figures are in
bold in Table 6-II).889 For the first period from 1590 to 1610, for which I have no information
on mercury sales or distribution, I assume the same amalgamation fraction of 0.7 applies, and
887
I have followed Bakewell’s path in applying factors of 10.9 and 20.8 to reverse calculate from the tax data
under 1% and diezmos and quintos the value of silver produced in pesos (of 272 maravedies). I have used his
value of 8 pesos 1 real for a mark of silver up to the year 1700, and 8 pesos 6 reales after that date. Bakewell,
Silver Mining in Zacatecas, 245.
888
Brading assumed that prior to 1632 an average correspondencia value was 100, equivalent to a weight ratio of
2. Brading, Miners Bourbon Mexico, 11.
889
At least three sets of silver production data for this period have been published for the Caja of Zacatecas.
Bakewell published his production data in marks, TePaske in pesos (of 272 maravedies) and kg of fine silver and
Lacueva likewise in pesos. Their totals for this period differ slightly, according to my calculations based on their
data: 3,218,152 kg for Bakewell, 3,153,180 kg for TePaske and 3,336,221 kg for Lacueva, a spread of not more
than 6% over the lowest value. Bakewell, Silver Mining in Zacatecas, 242-46.; TePaske and Brown, Gold and
Silver, 115-16.; Lacueva Muñoz, La plata del Rey, 388-90. For the data on mercury see Bakewell, Silver Mining
in Zacatecas, 251.
499
from TePaske’s figure on silver production (in bold in Table 6-II) I work back to calculate the
deemed amount of mercury consumed.
breakdown (kg)
total silver mercury
process
fraction
mercury to
period silver amalgamated produced consumed
silver silver ratio
kg kg kg
as liquid as volatilized
as calomel
mercury mercury
A 1,050,000 2,152,500 1,829,625 301,350 21,525 2.05

1591-1610 1,500,000 0.7
S 450,000
A 1,185,402 2,430,075 2,065,564 340,211 24,301 2.05
1611-1650 1,650,000 0.7
S 464,598
A 144,558 296,345 251,893 41,488 2,963 2.05
2/1651 - 4/1661 315,388 0.5
S 170,829
5/1661 - 4/1663, A 68,658 140,750 119,637 19,705 1,407 2.05
5/1664 - 6/1670
242,674 0.3
S 174,015
A 268,230 468,508 398,232 65,591 4,685 1.7
7/1670 to 6/1681 670,575 0.4
S 402,345
A 120,259 218,731 185,921 30,622 2,187 1.8
1/1686 to 2/1692 200,432 0.6
S 80,173
A 125,105 281,986 239,688 39,478 2,820 2.3
3/1692 to 12/1700 250,211 0.5
S 125,105
A 158,044 310,918 264,280 43,529 3,109 2.0
1/1701 to 12/1710 316,089 0.5
S 158,044
A 322,025 649,013 551,661 90,862 6,490 2.0
1/1711 to 12/1720 536,709 0.6
S 214,683
A 435,909 742,401 631,041 103,936 7,424 1.7
1/1721 to 12/1730 544,886 0.8
S 108,977
A 290,728 616,362 523,908 86,291 6,164 2.1
1/1731 to 12/1740 415,326 0.7
S 124,598
1/1741 to 12/1750, A 279,839 620,106 527,090 86,815 6,201 2.2
exc 1749
349,798 0.8
S 69,960
A 188,148 490,145 349,955 135,288 4,901 2.6
1/1751 to 12/1760 313,580 0.6
S 125,432
A 144,790 290,293 246,749 40,641 2,903 2.0
1/1761 to 12/1770 241,316 0.6
S 96,526
A 346,139 587,454 499,336 82,244 5,875 1.7
1/1771 to 12/1780 494,484 0.7
S 148,345
A 447,820 1,185,986 832,945 341,182 11,860 2.6
1/1781 to 12/1790 559,775 0.8
S 111,955
A 485,298 790,299 671,754 110,642 7,903 1.6
1/1791 to 12/1800 606,623 0.8
S 121,325
1/1801 to 12/1810 ex A 460,578 1,287,572 856,675 418,021 12,876 2.8
1802
575,722 0.8
S 115,144
total 9,783,589 13,559,445 11,045,956 2,377,895 135,594
Table 6-II. Magnitude of the environmental impact vectors for calomel, liquid mercury and
volatile mercury, corresponding to the silver obtained by amalgamation and registered in the
Caja of Zacatecas. In this and following tables A: Amalgamation, S: Smelting, For the method
and sources see text.
The estimated breakdown in production by refining process is then used to project a
base-line scenario of the magnitude of the main environmental impact vectors that correspond
500
to the refined silver registered at the Caja of Zacatecas from 1590 to 1820. For amalgamation
the main vectors are calomel washed away in water streams, liquid mercury in the soil or in the
water, and volatilized mercury during the casting of the silver bars, as shown in Table 6-II,
using the breakdown established in Chapter 3 (85%, 14% and 1%, respectively, of the total
amount of mercury consumed). The amount of calomel cannot be greater than 1.86 times the
amount of silver refined, so when the mercury to silver ratio is above 2.3 for any period, the
ceiling value of calomel loss is applied and the liquid mercury loss is adjusted accordingly
(values in bold italics). The magnitudes for salt and copper sulphate consumed and washed
away from the treated ore, of solid waste washed away in waterways and woodland required
for firewood as a result of amalgamation are reported in Table 6-III, according to the ratios per
kg of silver calculated in Chapter 4 at Regla.
For smelting the vectors correspond to lead and lead compounds lost in flue gases,
woodland required to produce charcoal and solid waste as slag, and their magnitude established
according to the ratios reported in Chapters 2 and 4 per kg of silver smelted. I have chosen to
report for lead and lead in compounds issued to the atmosphere a range of 5 to 10 kg per kg of
smelted silver, rather than a single ratio. The values for woodland consumed in both Tables are
overestimated, since I have not factored in the natural cycle of regeneration, which would
reduce the projections by at least 50%. The results are reported in Table 6-IV. All figures in
Tables 6-III and 6-IV have been rounded off to the nearest significant number, to reflect the
copper equivalent
total silver by salt woodland
sulphate charcoal mineral waste
produced amalgamation consumed consumed
consumed consumed
1591 to 1810
thousand t thousand ha
10 7 200 17 50 4,000 20
Table 6-III. Projected magnitude of other main environmental impact vectors corresponding
to the silver obtained by amalgamation and registered in the Caja of Zacatecas.
501
total silver volatile charcoal woodland

by smelting slag waste
produced lead loss consumed consumed
1591 to 1810
10 3 15 to 30 160 3,300 1,300
Table 6-IV. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of Zacatecas.
limitations of the exercise. Finally, to check on my calculations on silver production, I compare
my results with the totals reported by Tepaske.890
6.1.2 Caja of Guanajuato
Prior to 1650 refiners of the Guanajuato area registered their silver in the Caja of
Mexico (Ciudad de México). After that date the level of silver ore extracted in the mines around
the city of Guanajuato made it necessary to set up a separate Caja. According to TePaske the
majority of the silver registered at the Caja of Guanajuato came from the ore extracted from
the mines in its near vicinity, in other words from refining haciendas close to Guanajuato.891
Figure 6-4 shows how the fraction of amalgamated silver varied over the period 1679 to 1816.
The impact of lower mercury pricing as of the 1780s is reflected in the steady increase of the
fraction of amalgamation that already dominated production of silver in this area, starting from
a market share of approximately 0.65.892 Again, the decrease in mercury pricing altered the
890
In Table 6-II the total is 9.8 million kg, compared to 10.1 million kg reported in TePaske and Brown, Gold and
Silver, 121-23.
891
Ibid., 95.
892
Bakewell reported 35% for smelting in the 1730s and 27% in the 1770s. Bakewell, "Colonial Mining," 145.
502
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1665 1685 1705 1725 1745 1765 1785 1805 1825
duties duties & mintage
Figure 6-4. The fraction of total silver refined by amalgamation as registered in the Caja of
Guanajuato in the period 1679 to 1816. Prior to 1720 the time intervals of the raw data in the
TK data have been approximated to the calendar years.
1,200,000
1,000,000
800,000
kg silver
600,000
400,000
200,000
Figure 6-5. Registry of silver at the Caja of Guanajuato according to refining process. Data
from Table 6-V.
503
breakdown (kg)
fraction mercury
process
silver produced mercury as
period total silver kg amalgamated as liquid to silver
kg consumed kg as calomel volatilized
silver mercury ratio
mercury
5/1665 to A 42,609 96,080 81,668 13,451 961 2.3
60,870 0.70
2/1671 S 18,261
3/1671 to A 117,859 191,991 163,192 26,879 1,920 1.6
168,370 0.70
3/1681 S 50,511
4/1681 to A 36,477 44,231 37,596 6,192 442 1.2
7/1684
49,969 0.73
S 13,492
6/1690 to A 138,010 171,413 145,701 23,998 1,714 1.2
1/1701
212,323 0.65
S 74,313
2/1701 to A 141,095 177,727 151,068 24,882 1,777 1.3
217,069 0.65
2/1710 S 75,974
3/1711 to A 155,588 270,894 230,260 37,925 2,709 1.7
263,708 0.59
12/1720 S 108,120
1/1721 to A 270,254 448,296 381,051 62,761 4,483 1.7
422,272 0.64
12/1730 S 152,018
1/1731 to A 361,802 530,439 450,873 74,261 5,304 1.5
565,316 0.64
12/1740 S 203,514
1/1741 to A 459,422 830,679 706,077 116,295 8,307 1.8
792,108 0.58
12/1750 S 332,685
1/1751 to A 409,902 647,787 550,619 90,690 6,478 1.6
650,639 0.63
12/1760 S 240,736
1/1761 to A 439,844 828,293 704,049 115,961 8,283 1.9
637,455 0.69
12/1770 S 197,611
1/1771 to A 772,686 1,414,740 1,202,529 198,064 14,147 1.8
1,058,474 0.73
12/1780 S 285,788
1/1781 to A 726,541 1,610,538 1,368,957 225,475 16,105 2.2
931,462 0.78
12/1790 S 204,922
1/1791 to A 1,138,029 1,588,932 1,350,592 222,450 15,889 1.4
1,371,119 0.83
12/1800 S 233,090
1/1801 to A 685,135 1,493,699 1,269,644 209,118 14,937 2.2
815,637 0.84
12/1806 S 130,502
total 8,216,791 10,345,738 8,793,877 1,448,403 103,457
Table 6-V. Magnitude of the environmental impact vectors for calomel, liquid mercury and
Caja of Guanajuato.
504
balance between amalgamation and smelting, but the baseline of smelting had been reached
before the first of the decreases in mercury pricing, as shown in Figure 6-5, so the increase in
amalgamation after the 1770s cannot be due from a poaching of ores that would otherwise have
been smelted. Though Guanajuato is a Caja where amalgamation always dominated, with an
average over the whole period based on total silver produced of 71%, the two refining processes
were more evenly balanced for nearly a century, before the price of mercury was cut by 50%.
Between 1690 and 1710 no distinction was made between amalgamated and smelted
silver. For lack of sufficient data I have extrapolated the value of the amalgamation fraction for
the period 1665 to 1681, and interpolated it for the period 1690 to 1710.893 Tables 6-V, 6-VI
and 6-VII provide the projections on the magnitude of the main environmental impact vectors
from 1665 to 1806.
copper equivalent
total silver by woodland
salt consumed sulphate charcoal mineral waste
produced amalgamation consumed
consumed consumed
1665 to 1806
8 6 170 15 40 3,600 17
Table 6-VI. Projected magnitude of other main environmental impact vectors corresponding
to the silver obtained by amalgamation and registered in the Caja of Guanajuato.
893
The total of 8.2 million kg of silver I have calculated as produced in this period corresponds well with TePaske’s
figure of 8.5 million kg in TePaske and Brown, Gold and Silver, 130-31. The information in the TK data set for
Guanajuato has also been published in detail in Pérez Luque and Tovar Rangel, Caja Real Guanajuato, 99-246.
505
total silver charcoal woodland

by smelting volatile lead loss slag waste
produced consumed consumed
1665 to 1806 thousand t thousand ha
8 2 12 to 24 140 2,300 900
Table 6-VII. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of Guanajuato.
6.1.3 Caja of Mexico
The Caja at México was the first established by Spain to channel the silver product
refined in New Spain. As production grew other regional Cajas sprung up, which means that
the tax and revenue records in the Mexico Caja not only reflect the refining activity in the
vicinity of Ciudad de México (58 Reales de Minas by mid 1760s, including the major
production centre at Taxco) but at different times have also included the silver and mercury
that were later reported by the new Cajas of Guanajuato, San Luis Potosí, Zimapán and
Pachuca.894 This creates major problems for the analysis of the TK set for this Caja. First of
all, the distinction between amalgamated and smelted silver only appears very late in the
records, as can be seen in the very limited results reported in Figure 6-6. Second, even though
the silver tax records are identified for other regions, which helps to avoid double accounting
of their totals, mercury revenues are reported as an aggregate, with no such distinction. I have
therefore opted to use Tepaske’s data on silver production for the Caja during the period 1521
to 1810, and have assumed an amalgamation fraction of 0.8, based on Figure 6-4, and a mercury
to silver ratio of 1.8 to apply over the whole period.
894
TePaske and Brown, Gold and Silver, 84-87. Mexicans refer to Ciudad de México simply as México, so the
Caja de México does not refer to the whole country, but to the Caja situated in its capital, Ciudad de México.
506
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1780 1785 1790 1795 1800 1805 1810 1815 1820
duties
México in the period 1786 to 1816. The raw data are from the TK data set.
breakdown (kg)
fraction silver mercury
process
total silver as
period amalgamated produced consumed as liquid
kg as calomel volatilized
silver kg kg mercury
mercury
1521 to A 0 0 0 0 0
770,410 0.00
1560 S 770,410
1561 to A 1,801,032 3,241,858 2,755,579 453,860 32,419

2,251,290 0.80
1600 S 450,258
1601 to A 2,227,072 4,008,730 3,407,420 561,222 40,087

2,783,840 0.80
1700 S 556,768
1701 to A 2,318,016 4,172,429 3,546,564 584,140 41,724

2,897,520 0.80
1810 S 579,504
total 8,703,060 11,423,016 9,709,564 1,599,222 114,230
Table 6-VIII. Magnitude of the environmental impact vectors for calomel, liquid mercury and
Caja of México. Values in bold from footnote 894.
507
The analysis for the Caja of Mexico is therefore less detailed and subject to a much
greater uncertainty than for all the other Cajas. It is also biased in favour of amalgamation.
Tables 6-VIII, 6-IX and 6-X present the projections of the relevant magnitudes of the main
environmental impact vectors.
copper equivalent
consumed consumed
9 6 190 16 40 3,900 19
Table 6-IX. Projected magnitude of other main environmental impact vectors corresponding
to the silver obtained by amalgamation and registered in the Caja of México.
total silver volatile lead charcoal woodland

produced loss consumed consumed
9 2 12 to 24 140 2,400 950
Table 6-X. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of México.
6.1.4 Caja of Durango
Established in 1599, the Caja of Durango represents the production of refining sites (42
Reales de Minas in 1761-1767) around the capital of the colonial province of Nueva Vizcaya,
some two hundred miles to the northwest of Zacatecas. Among the contributors were Parral
508
and Chihuahua, the latter becoming a Caja in its own right as of 1785.895 Figure 6-7 shows
how the fraction of amalgamated silver varied over the period 1679 to 1816. The registries of
tax for the two decades between 1740 and 1760 do not discriminate revenues according to
refining process. but the data on mercury sales in the TK set for this hidden period indicate an
increasing use of amalgamation.
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1680 1700 1720 1740 1760 1780 1800 1820
Durango in the period 1696 to 1813. Between 1737 and 1765 no distinction was made between
amalgamated and smelted silver in the tax register. Prior to 1713 I have approximated the
irregular time series of the raw data in the TK data to their nearest calendar years.
The price decrease of mercury coincides with a tilt towards amalgamation after the
1780s, together with a concurrent drop in the amount of ore being smelted, as seen in Figure
6-8. . This profile is an exception to that which was observed for the previous Cajas. It can
reflect either a poaching of ores by cheaper mercury from smelting to amalgamation or simply
the exhaustion of lead rich silver ores in the mines of the region. A historic import of lead flux
from other regions to smelt dry ores prior to the 1770s would favour rhe former explanation,
895
Ibid., 91-93, 119.
509
otherwise the technical difficulties of amalgamating lead rich ores would discount it. On
average smelting accounted for 61% of the total silver registered at the Caja of Durango.
450,000
400,000
350,000
300,000
kg silver
250,000
200,000
150,000
100,000
50,000
0
to 1600
to 1610
to 1620
to 1810
to 1590
to 1630
to 1640
to 1650
to 1660
to 1670
to 1680
to 1690
to 1700
to 1710
to 1720
to 1730
to 1740
to 1750
to 1760
to 1770
to 1780
to 1790
to 1800
Figure 6-8. Registry of silver at the Caja of Durango according to refining process. Data
from Table 6-XI.
In Table 6-XI, from 1622 to 1696 the average historical mercury to silver ratio
calculated for Durango (2.1) is applied to the TK data on mercury so as to project how much
silver would have been refined by amalgamation during this period, and from that projected
value I calculate the fraction of amalgamated silver. For the period prior to 1622, I complement
the data on silver production from the TK set with Lacueva’s data for the years 1578 to 1598.896
I then apply the amalgamation ratio projected for 1622 (0.97) to all this period, at the risk of
overestimating the use of amalgamation. The deemed amounts of silver obtained by
amalgamation allow me to estimate the amount of mercury consumed, again based on a ratio
of 2.1. The results are given in Tables 6-XI, 6-XII and 6-XIII.897
896
897
The total for silver produced and registered at Durango in the table from 1599 onwards corresponds nearly
exactly with the figure of 5.9 million kg in TePaske and Brown, Gold and Silver, 125-27.
510
breakdown (kg)
process
total as mercury to
silver kg as calomel volatilized silver ratio
mercury
A 106,231 223,085 189,622 31,232 2,231 2.1
1578-1598 109,516 0.97
S 3,285
1/1599 to A 102,739 215,752 183,389 30,205 2,158 2.1
105,916 0.97
6/1611 S 3,177
7/1611 to A 36,299 76,228 64,794 10,672 762 2.1
37,422 0.97
4/1615 S 1,123
5/1622 to A 28,845 60,574 51,488 8,480 606 2.1
29,828 0.97
4/1625 S 984
6/1632 to A 80,938 169,969 144,474 23,796 1,700 2.1
223,591 0.36
6/1641 S 142,653
7/1641 to A 69,274 145,476 123,655 20,367 1,455 2.1
234,813 0.30
12/1650 S 165,539
1/1651 to A 70,368 147,773 125,607 20,688 1,478 2.1
12/1659 171,920 0.41
exc 1654 S 101,552
1/1664 to A 73,541 154,437 131,272 21,621 1,544 2.1
204,639 0.36
5/1673 S 131,097
6/1673 to A 18,221 38,265 32,525 5,357 383 2.1
100,278 0.18
7/1677 S 82,056
1/1685 to A 33,010 69,321 58,923 9,705 693 2.1
75,456 0.44
12/1688 S 42,445
1/1689 to A 26,970 91,629 50,165 40,548 916 3.4
168,564 0.16
6/1700 S 141,593
7/1700 to A 42,704 97,567 82,932 13,659 976 2.3
203,352 0.21
6/1711 S 160,648
7/1711 to A 81,699 107,857 91,679 15,100 1,079 1.3
355,211 0.23
12/1720 S 273,513
1/1721 to A 44,461 51,731 43,972 7,242 517 1.2
404,195 0.11
12/1730 S 359,734
1/1731 to A 41,598 44,639 37,943 6,249 446 1.1
12/1740 415,980 0.1
exc. 1734 S 374,382
1/1741 to A 106,536 85,549 72,717 11,977 855 0.8

532,680 0.2
12/1750 S 426,144
1/1752 to A 128,477 159,861 135,882 22,381 1,599 1.2
428,258 0.3
12/1760 S 299,780
1/1761 to A 212,440 410,587 348,999 57,482 4,106 1.9
518,147 0.41
12/1770 S 305,707
1/1771 to A 253,483 677,749 471,478 199,493 6,777 2.7
478,270 0.53
12/1780 S 224,787
1/1781 to A 208,696 509,953 388,174 116,680 5,100 2.4
12/1790 342,124 0.61
exc 1787 S 133,428
1/1791 to A 313,419 735,234 624,949 102,933 7,352 2.3

474,878 0.66
12/1800 S 161,458
1/1801 to A 302,346 705,805 599,934 98,813 7,058 2.3
12/1810 444,626 0.68
exc 1806 S 142,280
total 6,059,664 4,979,042 4,054,571 874,681 49,790
Table 6-XI. Magnitude of the environmental impact vectors for calomel, liquid mercury and
Caja of Durango. Silver production figure in bold from footnote 896, ceiling for calomel
estimates indicated in bold italic figures.
511
copper equivalent
total silver by salt mineral woodland
sulphate charcoal
produced amalgamation consumed waste consumed
consumed consumed
1578 to 1810
6 2 70 6 17 1,500 7
Table 6-XII. Projected magnitude of other main environmental impact vectors corresponding
to the silver obtained by amalgamation and registered in the Caja of Durango.

6 4 18 to 36 220 3,700 1,500
Table 6-XIII. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of Durango.
6.1.5 Caja of San Luis Potosí
The split between smelting and amalgamation reflected in the tax records of the Caja
of San Luis Potosí is the best example why these curves must be interpreted first of all based
on the nature of the ore being processed. At a first reading of Figure 6.9 the change from
smelting to amalgamation coincides so well with the decrease in the price of mercury in the
1770s that a causal link seems the explanation. However it is the change of the type of ore, and
not the price of mercury, that determines the profile in Figure 6.9. The initial period
corresponds predominantly to the smelting of lead rich ores, first found in the mines of the
Cerro San Pedro on the hills that surround the town, and then from other locations such as
Charcas and Guadalcazar. In the early eighteenth century when mention of amalgamation
512
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1700 1720 1740 1760 1780 1800 1820
San Luis Potosí in the period 1713 to 1806. Source of raw data is the TK set.
haciendas start to appear in archival records.898 In the 1770s silver was discovered at Catorce,
250 km to the north of the town of San Luis Potosí.899 The ore, rich in native silver and silver
halides, was refined using Barba’s cazo amalgamation process and complemented by an
extraction using the patio process (Chapter 5).900 It is the production from Catorce that explains
the predominance of amalgamation in the production of silver registered at San Luis Potosí as
of the 1780s observed in Figure 6-9, not the decrease in the price of mercury. The impact of
898
During this period Bakewell reports 86% of silver produced by smelting in the 1730s dropping to 54% by the
1760s. Bakewell, "Colonial Mining," 145. Smelting is reported as accounting for 92% of production in 1718, and
then decreasing to 48.6% in 1761-1767 and then virtually disappearing at 1.6% by 1785-89 and later years, in Inés
Herrera Canales, "El método de refinación con azogue en la minería potosina colonial: del fuego al cazo " in La
plata en Iberoamérica: Siglos XVI al XIX, ed. Jesus Paniagua Pérez and Nuria Salazar Simarro (Leon: Universidad
de León, 2008), 68.
899
Ines Herrera Canales, "El auge de la plateria potosina en la segunda mitad del siglo XVIII, al argento vivo," in
Ophir en las Indias. Estudios sobre la plata americana. Siglos XVI-XIX, ed. Jesus Paniagua Perez and Nuria
Salazar Simarro (Leon: Universidad de Leon, 2010), 115-21.
900
The average mercury to silver weight ratio calculated from the TK data set for San Luis Potosí averages 1.7
between 1710 and 1780, and 0.9 from 1781 to 1806. The use of the cazo method cut average mercury consumption
by half, though this decrease is very dependent on the nature of the ore.
513
the mines of Catorce is reflected in the profile of the register of silver for this Caja (Figure 6-
10). It is interesting to observe that the baseline of smelting remained fairly constant over the
whole period covered in Figure 6-10, only decreasing substantially as of the 1790s.
900,000
800,000
700,000
600,000
kg silver
500,000
400,000
300,000
200,000
100,000
0
to 1630
to 1640
to 1650
to 1660
to 1670
to 1680
to 1690
to 1700
to 1710
to 1720
to 1730
to 1740
to 1750
to 1760
to 1770
to 1780
to 1790
to 1800
Figure 6-10. Registry of silver at the Caja of San Luis Potosí according to refining process.
Data from Table 6-XIV.
The hard data on the ratio of amalgamation to smelting begin to be recorded in the TK
set as of 1712. Prior to that date, the TK set for San Luis Potosí contains data on taxes paid on
silver, beginning with a figure said to correspond with the taxes paid from March 1528 to
March 1629. Since mining and refining only started in earnest at the end of the sixteenth
century, I have preferred to report that initial value converted into kg in Table 6-XIV as ‘up to
March 1629’. Revenues from mercury sales only appear from 1672 onwards, and I am
assuming smelting completely dominated production in the early years (see Chapter 3). From
1672 to 1710 I divide the weight of mercury sold by 1.7 (the average mercury to silver ratio
from 1710 to 1770 according to the TK data that does not involve the cazo process) so as to
514
arrive at a deemed weight of silver obtained by amalgamation, and from there I project an
amalgamation fraction for each period.901
Because of the geographical split between Catorce and the rest of the haciendas
registering their silver at the Caja of San Luis Potosí, it is prudent to divide the percentage of
silver refined by each process into two historical periods. The first, from the early 1600s to
1780, assigns 82% of all silver produced to smelting. The second, as of 1780, assigns 96% of
the silver produced to amalgamation, most of which was produced at the mines of Catorce. The
environmental impact of silver refining for this Caja is thus also divided along geographical
lines. Smelting would impact the region around and within the town of San Luis Potosí for
some 160 years, while the more isolated area around the mines of Catorce would be spared the
consequences of smelting. The results in Tables 6-XIV to 6-XVI must be interpreted bearing
this division in mind.
The area of Catorce would see in just 30 years most of the impact from amalgamation:
1,400,000 t of mineral waste, subject to the silver content of the Catorce ores, while a level of
2,000 t of calomel is consistent with the level of mercury consumed. No copper sulphate
consumption is reported because the cazo process did not use this reagent. The estimate of
woodland consumed for amalgamation is below the level expected to have occurred, since the
cazo process requires more fuel than the patio process.
The total amount of silver registered at the Caja according to Table 6-XIV is 3.9 million kg. The total reported
901
byTepaske up to the end of 1806 is just over 4 million kg. TePaske and Brown, Gold and Silver, 127-29.
515
breakdown (kg)
process
total silver as mercury to
kg as calomel volatilized silver ratio
mercury
A 0 0 0 0 0
up to 3/1629 27,519 0
S 27,519
A 0 0 0 0 0
2/1630 to 4/1640 172,664 0
S 172,664
A 0 0 0 0 0
12/1640 to 2/1651 155,489 0
S 155,489
A 0 0 0 0 0
12/1653 to 6/1661 82,141 0
S 82,141
A 0 0 0 0 0
7/1661 to 10/1671 112,388 0
S 112,388
A 3,396 5,774 4,908 808 58 1.7
12/1672 to 2/1675,
75,525 0.04
11/1677 to 4/1681
S 72,129
5/1681 to 7/1684,
A 6,844 11,635 9,890 1,629 116 1.7
70,374 0.10
3/1686 to 3/1688 S 63,529
A 4,800 8,159 6,935 1,142 82 1.7
1/1690 to 4/1701 128,682 0.04
S 123,882
A 1,235 2,099 1,784 294 21 1.7
1/1706 to 12/1710 41,832 0.03
S 40,597
A 5,451 8,165 6,940 1,143 82 1.5
1/1712 to 12/1720 68,138 0.08
S 62,687
A 6,373 7,250 6,163 1,015 73 1.1
1/1721 to 12/1730 91,046 0.07
S 84,673
A 15,472 34,540 29,359 4,836 345 2.2
1/1731 to 12/1740 110,512 0.14
S 95,040
A 5,751 11,788 10,020 1,650 118 2.0
1/1741 to 12/1748 71,885 0.08
S 66,134
A 93,576 188,859 160,530 26,440 1,889 2.0
1/1752 to 12/1760 252,907 0.37
S 159,332
1/1761 to 12/1770 A 85,937 133,103 113,138 18,634 1,331 1.5
186,820 0.46
exc 1765 S 100,883
A 124,271 138,863 118,033 19,441 1,389 1.1
1/1771 to 12/1780 318,643 0.39
S 194,372
A 679,743 659,313 560,416 92,304 6,593 1.0
1/1781 to 12/1790 738,851 0.92
S 59,108
A 791,417 571,467 485,747 80,005 5,715 0.7
1/1791 to 12/1800 815,894 0.97
S 24,477
A 424,969 406,692 345,688 56,937 4,067 1.0
1/1801 to 12/1806 424,969 1
S 0
total 3,946,278 2,187,709 1,859,553 306,279 21,877
Table 6-XIV. Magnitude of the environmental impact vectors for calomel, liquid mercury and
Caja of San Luis Potosí.
516
copper equivalent
sulphate charcoal
consumed consumed
1600s to 1806
4 2 70 - 16 1,400 7
Table 6-XV. Projected magnitude of other main environmental impact vectors corresponding
to the silver obtained by amalgamation and registered in the Caja of San Luis Potosí.

1600s to 1806
4 2 8 to 16 100 1,700 700
Table 6-XVI. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of San Luis Potosí.
6.1.6 Caja of Guadalajara
The silver registered at the Caja of Guadalajara came from multiple small and medium
refiners, with 46 Reales de Minas operating in the mid-1760s.902 Figure 6-11 shows how the
fraction of amalgamated silver varied over the period 1679 to 1816. Amalgamation was the
main refining process used, and the tendency to increase its share from approximately 60 %
until it became the predominant route to silver is observed even before the price reduction of
902
Ibid., 90-91.
517
mercury in the 1770s.903 The evidence that the price of mercury was not the only factor that
influenced the split between amalgamation and smelting also comes from Figure 6-12.
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1680 1700 1720 1740 1760 1780 1800 1820
Guadalajara in the period 1691 to 1804. Prior to 1699 I have approximated the irregular time
series of the raw data in the TK set to their nearest calendar years.
300,000
250,000
200,000
kg silver
150,000
100,000
50,000
0
to 1590
to 1600
to 1610
to 1800
<1580
to 1620
to 1630
to 1640
to 1650
to 1660
to 1670
to 1680
to 1690
to 1700
to 1710
to 1720
to 1730
to 1740
to 1750
to 1760
to 1770
to 1780
to 1790
Figure 6-12. Registry of silver at the Caja of Guadalajara according to refining process. Data
from Table 6-XVII.
903
Bakewell reported 26% smelting in the 1730s dropping to 8% by the 1770s. Bakewell, "Colonial Mining," 145.
518
Prior to the 1760s the balance between the processes had shifted from a close pairing
to an evident dominace of amalgamation. This shift was accentuated once inexpensive mercury
became available, but as Villaseñor had pointed out, not even free mercury could work
miracles. Amalgamation returned to the level it would have reached even without the decrease
in the price of mercury. The Caja of Guadalajara registered in total 73% of amalgamated silver
and 27% of smelted silver.904
The amalgamation fraction is calculated directly from the TK data set from 1690
onwards. The average ratio of mercury to silver calculated for this period is 2.1, as shown in
Table 6-XVII. To estimate the amalgamation fraction for the earlier periods I apply this ratio
to the data from the TK set on sales of mercury from 1611 to 1690. The average projected
amalgamation fraction for this period is 0.7. I then apply this fraction to the period where I
have no data on mercury sales, 1568 to 1611, to estimate a total amount of silver refined by
amalgamation, from where I obtain the deemed quantities of mercury consumed using the
average mercury to silver ratio of 2.1.
904
The total for silver produced and registered at the Caja of Guadalajara according to Table 6-XVI is 3.7 million
kg, similar to the total of 3.8 million kg in TePaske and Brown, Gold and Silver, 116.
519
breakdown (kg)
fraction mercury
process
total silver silver as mercury to
period amalgamated consumed as liquid
kg produced kg as calomel volatilized silver ratio
silver kg mercury
mercury
A 40,353 84,390 71,731 11,815 844 2.1
1568, 1578,1579 57,647 0.7
S 17,294
1/1581 to 12/1590, A 25,970 54,311 46,165 7,604 543 2.1
exc. 1585, 1588, 37,100 0.7
1589 S 11,130
A 49,320 103,143 87,671 14,440 1,031 2.1
1/1591 to 3/1601 70,457 0.7
S 21,137
4/1601 to 3/1611, A 45,508 95,171 80,895 13,324 952 2.1
65,012 0.7
exc. 1605 S 19,504
A 55,468 115,999 98,599 16,240 1,160 2.1
4/1611 to 3/1621 80,698 0.7
S 25,231
A 72,329 151,261 128,572 21,177 1,513 2.1
4/1621 to 4/1631 114,861 0.6
S 42,532
A 77,670 162,430 138,066 22,740 1,624 2.1
5/1631 to 4/1641 88,119 0.9
S 10,450
A 49,962 104,486 88,813 14,628 1,045 2.1
5/1641 to 5/1651 85,969 0.6
S 36,006
A 102,294 213,927 181,838 29,950 2,139 2.1
6/1651 to 12/1660 136,504 0.7
S 34,209
A 78,208 163,555 139,022 22,898 1,636 2.1
1/1661 to 2/1671 175,092 0.4
S 96,885
A 106,923 223,607 190,066 31,305 2,236 2.1
3/1671 to 3/1681 187,842 0.6
S 80,919
A 120,816 252,662 214,762 35,373 2,527 2.1
4/1681 to 6/1690 181,195 0.7
S 60,379
7/1690 to 6/1701 ex A 76,919 144,062 122,453 20,169 1,441 1.9
mid 1693 to mid 146,283 0.53
1696 S 69,364
A 101,471 177,326 150,727 24,826 1,773 1.7
1/1701 to 12/1710 154,299 0.66
S 52,828
A 99,955 195,663 166,313 27,393 1,957 2.0
1/1711 to 12/1720 170,740 0.59
S 70,785
A 111,278 249,515 212,087 34,932 2,495 2.2
1/1721 to 12/1730 172,303 0.65
S 61,025
A 150,680 298,825 254,001 41,835 2,988 2.0
1/1731 to 12/1740 199,184 0.76
S 48,504
A 159,427 307,543 261,412 43,056 3,075 1.9
1/1741 to 12/1750 204,671 0.78
S 45,244
A 147,564 356,895 274,469 78,857 3,569 2.4
1/1751 to 12/1760 184,246 0.80
S 36,682
A 279,318 572,694 486,790 80,177 5,727 2.1
1/1761 to 12/1770 296,484 0.94
S 17,166
A 268,537 527,232 448,147 73,812 5,272 2.0
1/1771 to 12/1780 304,919 0.88
S 36,382
A 186,589 483,742 347,055 131,850 4,837 2.6
1/1781 to 12/1790 252,230 0.74
S 65,641
A 183,414 330,124 280,606 46,217 3,301 1.8
1/1791 to 12/1800 215,008 0.85
S 31,594
A 66,009 167,555 122,777 43,103 1,676 2.5
1801 to 1804 78,527 0.84
S 12,518
total 3,659,392 5,536,120 4,593,040 887,719 55,361
Table 6-XVII. Magnitude of the environmental impact vectors for calomel, liquid mercury and
Caja of Guadalajara.
520
copper equivalent
sulphate charcoal
consumed consumed
4 3 80 7 20 1,600 8
Table 6-XVIII. Projected magnitude of other main environmental impact vectors

corresponding to the silver obtained by amalgamation and registered in the Caja of
Guadalajara.

4 1 5 to 10 60 1,000 400
Table 6-XIX. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of Guadalajara.
6.1.7 Caja of Pachuca
From the start of mining in the region of Pachuca in 1552 until 1667, all its silver
production was registered at the nearby Caja de Mexico. Its registry corresponds to refining
activities concentrated around two main sites, Pachuca and Real del Monte.905 Figure 6-13
905
Ibid., 96-98.
521
shows the evolution in the fraction of amalgamated silver over the period 1679 to 1816. At first
sight it indicates an unexpected change to smelting after the price of mercury had decreased.906
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1660 1680 1700 1720 1740 1760 1780 1800 1820
duties duties & mintage mintage
Pachuca in the period 1667 to 1820. Prior to 1706 I have approximated the irregular time series
of the raw data in the TK set to their nearest calendar years.
The interpretation however is not quite straightforward. As the fraction of amalgamated
silver decreases towards the end of the eighteenth century, so did the level of silver production
overall (Figure 6-14). In Chapter 4 I mentioned that lead-rich ores were supplied to the first
Count of Regla from the mines of Zimapán. One interpretation for Figure 6-13 is that the
smelting fraction increased as the availability of the ores for amalgamation decreased, while
the overall production of silver declined. This is another example of how the nature of the
available ore is more important than the price of mercury. In spite of the impression given by
Figure 6-13, the haciendas that registered their silver at the Caja of Pachuca were mainly
906
Bakewell reports 27% by of silver produced by smelting in the 1720s, decreasing slightly to 23% by the 1760s.
Bakewell, "Colonial Mining," 145. The migration of lead rich ores to the new Caja of Zimapan as of the 1730s
(see below) may have contributed to the low spike in the amalgamation fraction observed around the 1760s.
522
amalgamation haciendas, supplying 73% of the total silver produced. Smelting would
contribute with 27% of the production.
350,000
300,000
250,000
kg silver
200,000
150,000
100,000
50,000
smelting amalgamation total
Figure 6-14. Silver registered at the Caja of Pachuca. Data from Table 6-XX.
Tables 6-XX to 6-XXII summarize the data and results for the Caja of Pachuca. The
mercury to silver ratio for the period 1667 to 1806 is 2.907
907
The total of 2.5 million kg of silver produced from 1667 to 1806 corresponds well with the figure of 2.6 million
kg in TePaske and Brown, Gold and Silver, 132-33.
523
breakdown (kg)
fraction mercury
process
total silver as mercury to
period amalgamated consumed as liquid
silver kg produced kg as calomel volatilized silver ratio
silver kg mercury
mercury
9/1667 to A 30,553 52,305 44,460 7,323 523 1.7
34,341 0.89
2/1671 S 3,788
3/1671 to A 77,996 154,706 131,500 21,659 1,547 2.0
87,016 0.90
1/1680 S 9,020
11/1680 to A 165,695 376,266 319,826 52,677 3,763 2.3
5/1693
184,106 0.90
S 18,411
1/1706 to A 52,859 85,314 72,517 11,944 853 1.6
60,077 0.88
12/1710 S 7,218
1/1711 to A 128,283 255,624 217,281 35,787 2,556 2.0
189,289 0.68
12/1720 S 61,007
1/1721 to A 236,640 479,629 407,685 67,148 4,796 2.0
327,528 0.72
12/1730 S 90,888
1/1731 to A 152,517 295,395 251,086 41,355 2,954 1.9
12/1740
191,368 0.80
S 38,851
1/1741 to A 101,799 209,256 177,868 29,296 2,093 2.1
12/1749
134,187 0.76
S 32,388
1/1751 to A 210,418 341,768 290,503 47,847 3,418 1.6
12/1760
253,687 0.83
S 43,269
1/1761 to A 223,331 341,426 290,212 47,800 3,414 1.5
12/1770
299,372 0.75
S 76,040
1/1771 to A 158,277 299,692 254,738 41,957 2,997 1.9
12/1780
228,999 0.69
S 70,722
1/1781 to A 82,008 279,349 152,535 124,021 2,793 3.4
12/1790
153,629 0.53
S 71,621
1/1791 to A 108,221 246,137 209,217 34,459 2,461 2.3
12/1800
193,246 0.56
S 85,025
1801 to 1804, A 52,269 171,709 97,221 72,771 1,717 3.3
1806
111,808 0.47
S 59,539
total 2,448,652 3,588,578 2,916,647 636,045 35,886
Table 6-XX. Magnitude of the environmental impact vectors for calomel, liquid mercury and
Caja of Pachuca.
copper equivalent
consumed consumed
1667 to 1806
2 2 50 5 12 1,100 5
Table 6-XXI. Projected magnitude of other main environmental impact vectors corresponding
to the silver obtained by amalgamation and registered in the Caja of Pachuca.
524

1667 to 1806
2 1 3 to 6 40 700 270
Table 6-XXII. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of Pachuca.
6.1.8 Caja of Sombrerete
The Caja at Sombrerete was established in 1683, and the silver from its ores had been
registered until then at the Caja of Zacatecas. The bell-shaped profile of the plot in Figure 6-
11 of the fraction of silver amalgamated shows an impressive back and forth between smelting
and amalgamation.908 It is reported that from the 1760s more than 90 % of the registered silver
came from local mines of Sombrerete, and most from ‘the rich vein of El Pabellón … [which
had] a high lead content’.909 A high lead content in ores rules out amalgamation as the refining
method of choice, so it explains the return of smelting after the 1760s irrespective of the price
of mercury.
908
Bakewell reported 68% for silver by smelting in the 1720s, decreasing to 33% by the 1760s, but he did not
extend his data to the end of the century and so could not remark on the increase again in smelting fraction.
Bakewell, "Colonial Mining," 145. Lacueva indicates that at the end of the seventeenth century smelting was used
in the new mines of Sombrerete, where in the period from 1688 to 1699 up to 88% of silver would be produced
by smelting of ores. Lacueva Muñoz, La plata del Rey, 401.
909
TePaske and Brown, Gold and Silver, 98-99.
525
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1660 1680 1700 1720 1740 1760 1780 1800 1820
Sombrerete in the period 1680 to 1820. Prior to 1760 I have approximated the irregular time
series of the raw data in the TK set to their nearest calendar years.
350,000
300,000
250,000
kg silver
200,000
150,000
100,000
50,000
0
Figure 6-16. Silver registered at the Caja of Sombrerete. Data from Table 6-XXIII.
The peak in the amalgamation fraction is observed before the decrease in mercury
prices is implemented. The level of smelting on either side is fairly constant (Figure 6-16), so
no poaching of ores is taking place. It coincides with a peak in production as well so new ores
526
fit for amalgamation were being fed to the haciendas. Existing amalgamation haciendas would
have profited the most, since no new capital expenditures would have been required. By the
end of the eighteenth century production peaked again, but this time on the back of lead-rich
ores. TePaske proposed that refiners at Sombrerete switched from amalgamation to smelting
when mercury became scarce ‘due to disruptions in trans-Atlantic shipping’.910 It would not
have been profitable to attempt to amalgamate ores with a high lead content, even less before
the decrease in mercury prices, so I prefer to argue that the bell-shaped profile in Figure 6-11
is caused by major changes in the nature of the ore being refined.
The breakdown of my calculations based on the data in the TK set is presented in Table
6-XXIII.911 The mercury to silver ratio during this period shows an interesting behaviour. On
average it has a value of 2.1, which falls within the expected historical range. However, when
calculated by decade it shows an abnormal range after 1780, reaching the value of 9.2 from
1791 to 1800. The average from 1683 to 1780 is 1.6, but from 1781 to 1816 it increases to 4.8,
at the same time the amount of silver produced by amalgamation reaches a minimum. This
can be interpreted as very inexpensive mercury being used in an inefficient manner to
amalgamate lead-rich ores, thus leading to its waste in a non-productive manner from 1780 to
1810. By requiring on average twice the amount of mercury at half the traditional price to
produce the same amount of silver, it undercuts the argument that the decrease in mercury
prices was beneficial to the Crown revenues as a whole, at least for the Caja of Sombrerete
during this period.912 In addition came a concurrent environmental impact, with more liquid
mercury being lost while calomel amounts reach their chemical ceiling value (figures in bold).
910
Ibid., 99.
911
The aggregate total of silver produced (1.8 million kg) corresponds well to the total of 1.6 million kg in ibid.,
133-34.
912
It could also be argued that an abnormally high mercury to silver ratio indicates a combination of bad practice
and contraband of mercury. Why refiners would suddenly become bad operators after the experience shown in
527
The Caja of Sombrerete was in smelting territory. 68% of the total silver registered at
the Caja came from smelting haciendas, 32% from amalgamation. The magnitudes of the main
environmental impact vectors are estimated in Tables XXIII to XXV.
breakdown (kg)
silver
process
total silver fraction mercury as mercury to
period produced as liquid
kg amalgamated silver consumed kg as calomel volatilized silver ratio
kg mercury
mercury
5/1683 to 5/1690, A 23,823 32,071 27,261 4,490 321 1.3
exc mid 1684 -mid 119,115 0.2
1688 S 95,292
A 35,916 74,296 63,151 10,401 743 2.1
5/1690 to 5/1701 189,032 0.19
S 153,116
A 20,641 26,411 22,450 3,698 264 1.3
6/1701 to 3/1711 66,583 0.31
S 45,942
A 11,841 23,069 19,609 3,230 231 1.9
4/1711 to 12/1720 40,832 0.29
S 28,991
A 11,885 21,349 18,147 2,989 213 1.8
1/1721 to 12/1730 32,122 0.37
S 20,237
A 83,579 130,547 110,965 18,277 1,305 1.6
1/1731 to 12/1740 128,583 0.65
S 45,004
1/1741 to 12/1750 A 118,915 184,837 157,111 25,877 1,848 1.6
exc 1747,1748
152,455 0.78
S 33,540
A 30,868 65,535 55,705 9,175 655 2.1
1/1753 to 12/1760 40,616 0.76
S 9,748
A 32,464 33,103 28,137 4,634 331 1.0
1/1761 to 12/1770 45,723 0.71
S 13,260
A 68,647 61,405 52,194 8,597 614 0.9
1/1771 to 12/1780 118,356 0.58
S 49,710
A 42,052 158,534 78,217 78,731 1,585 3.8
1/1781 to 12/1790 107,826 0.39
S 65,774
A 16,524 151,716 30,735 119,463 1,517 9.2
1/1791 to 12/1800 183,605 0.09
S 167,081
1/1801 to 12/1809 A 33,926 133,794 63,103 69,354 1,338 3.9
exc 1807
339,262 0.10
S 305,336
A 5,226 12,567 10,682 1,759 126 2.4
1/1811 to 12/1816 87,100 0.06
S 81,874
total 1,651,212 1,109,234 737,467 360,675 11,092
Table 6-XXIII. Magnitude of the environmental impact vectors for calomel, liquid
mercury and volatile mercury, corresponding to the silver obtained by amalgamation and
registered in the Caja of Sombrerete.
the 1760s with mercury to silver ratios well below 2, or why the least expensive mercury in the history of New
Spain would make sense to contraband weaken these alternative explanations.
528
copper equivalent
consumed consumed
1683 to 1816
2 1 16 1 4 330 2
Table 6-XXIV. Projected magnitude of other main environmental impact vectors

corresponding to the silver obtained by amalgamation and registered in the Caja of Sombrerete.

1683 to 1816
2 1 6 to 12 70 1,100 450
Table 6-XXV. Projected magnitude of the main environmental impact vectors corresponding
to the silver obtained by smelting and registered in the Caja of Sombrerete.
6.1.9 Caja of Bolaños
Major mining in the region around Bolaños started in 1747, and the Caja was
established some six years later. Records are affected by a fifteen year tax exemption on the
diezmo granted to one of the principal miners of this region as of 1789. 913 The data from the
TK sets as plotted in Figure 6-18 show that Bolaños was amalgamation territory even before
the price decrease of mercury in the 1760s. Over 94% of all the silver registered in this Caja
came from amalgamation (Table 6-XXVI). The mercury to silver ratio again shows a step
increase that coincides with the decrease in the price of mercury, from 1.9 prior to 1760 to an
average of 3.2 from 1761 to 1804, reaching a value of 5.2 in the following decade. A non-
913
TePaske and Brown, Gold and Silver, 101.
529
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1750 1760 1770 1780 1790 1800 1810
Bolaños in the period 1753 to 1804. Raw data from TK set.
breakdown (kg)
fraction mercury
process
total silver mercury as

period amalgamated as liquid to silver
silver kg produced kg consumed kg as calomel volatilized
mercury
1/1753 to A 455,770 864,622 734,929 121,047 8,646 1.9
460,551 0.99
12/1760 S 4,781
1/1761 to A 155,080 376,216 288,449 84,005 3,762 2.4
166,019 0.93
12/1770 S 10,939
1/1771 to A 190,762 478,827 354,817 119,222 4,788 2.5
214,718 0.89
12/1780 S 23,957
1/1781 to A 201,123 637,264 374,089 256,802 6,373 3.2
218,996 0.92
12/1790 S 17,873
1/1791 to A 100,453 519,384 186,843 327,347 5,194 5.2
108,061 0.93
12/1800 S 7,607
1/1801 to A 8,598 23,103 15,993 6,879 231 2.7
10,479 0.82
12/1804 S 1,880
total 1,178,824 2,899,416 1,955,120 915,302 28,994
Table 6-XXVI. Magnitude of the environmental impact vectors for calomel, liquid
registered in the Caja of Bolaños.
530
efficient use of inexpensive mercury again implicates a change in the environmental impact of
amalgamation (figures in bold).914
copper equivalent
total silver by woodland
salt consumed sulphate charcoal mineral waste
produced amalgamation consumed
consumed consumed
1753 to 1804
1 1 30 3 8 700 3
Table 6-XXVII. Projected magnitude of other main environmental impact vectors

corresponding to the silver obtained by amalgamation and registered in the Caja of Bolaños.

1 0.1 0.5 to 1 4 70 30
Table 6-XXVIII. Projected magnitude of the main environmental impact vectors

corresponding to the silver obtained by smelting and registered in the Caja of Bolaños.
6.1.10 Caja of Rosario
According to Tepaske the Caja of Rosario was set up in 1770, moved to Alamos in
1783 and then to Cosalá around 1807.915 The records of the TK set are simply identified as
Rosario. Figure 6-19 shows smelting maintaining a relatively constant fraction under one third
even after the decrease in mercury pricing. Of the total silver registered in this Caja, 71%
corresponds to amalgamation and 29% to smelting (Table 6-XXIX).916 Again the mercury to
914
The total of 1.2 million kg of silver produced from 1753 to 1804, corresponds well with the figure of 1.1 million
kg in ibid., 135-36.
915
Ibid., 103.
916
The total of 1.1 million kg of silver registered from 1770 to 1809, is virtually the same as the total reported in
ibid., 135-36.
531
silver ratio is higher than the historical range for New Spain in the first two decades just after
the price decrease of mercury, but not to the extent of potential waste observed in the previous
two Cajas. The calomel projection is fixed at its ceiling value (figures in bold) during these
decades.
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1770 1775 1780 1785 1790 1795 1800 1805 1810 1815
Rosario in the period 1770 to 1813. Raw data from TK set.
mercury losses
fraction mercury to
process
total silver silver mercury as

period amalgamated as liquid silver weight
kg produced kg consumed kg as calomel volatilized
mercury
A 57,654 149,415 107,236 40,685 1,494 2.6
1/1770 to 12/1780 87,354 0.66
S 29,700
A 196,231 559,577 364,990 188,991 5,596 2.9
1/1781 to 12/1790 268,810 0.73
S 72,579
1/1791 to 12/1800 exc A 291,119 585,950 498,057 82,033 5,859 2.0
1794
373,230 0.78
S 82,111
1/1801 to 12/1809 A 231,756 490,704 417,099 68,699 4,907 2.1
360,896 0.64
exc 1806 S 129,141
total 1,090,290 1,785,646 1,387,382 380,408 17,856
Table 6-XXIX. Magnitude of the environmental impact vectors for calomel, liquid
registered in the Caja of Rosario.
532
copper equivalent
consumed consumed
1770 to 1809
1 1 25 2 5 500 2
Table 6-XXX. Projected magnitude of other main environmental impact vectors corresponding
to the silver obtained by amalgamation and registered in the Caja of Rosario.

1770 to 1809
1 0.3 2 to 4 20 300 130
Table 6-XXXI. Projected magnitude of the main environmental impact vectors

corresponding to the silver obtained by smelting and registered in the Caja of Bolaños.
Tables 6-XXX and 6-XXXI provide the ranges of magnitude of the remaining
environmental vectors for this Caja, reflecting the joint importance of both amalgamation and
smelting.
6.1.11 Caja of Zimapán.
The silver from Zimapán was initially registered as of the sixteenth century first in the
Caja of México and then after 1667 in the Caja of Pachuca, but in 1729 it was awarded its own
Caja. Though various mines produced its registered silver, by the 1760 Zimapán contributed
533
86% of the total with an ore rich in lead.917 It is the nature of the ore that determines the profile
seen in Figure 6-20, a near total absence of amalgamation in the refining of silver at Zimapán.918
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1720 1730 1740 1750 1760 1770 1780 1790 1800 1810
duties
Zimapán in the period 1729 to 1806. Raw data from TK set.
For all practical purposes the total production of silver came from smelting (Table 6-
XXXII).919 The only environmental impact vectors correspond to smelting, and their
magnitudes are projected in Table 6-XXXIII.
917
Ibid., 100. There were more than 100 smelting furnaces in Zimapán in 1795 according to Sonneschmidt and
de Fagoaga, Tratado de la amalgamación de Nueva España, 62.
918
Bakewell reported that smelting reached 90+% in the 1720s and 94% in the 1760s. Bakewell, "Colonial
Mining," 145.
919
The total of 0.8 million kg of silver produced from 1729 to 1806, corresponds well with TePaske’s figure of
0.9 million kg. TePaske and Brown, Gold and Silver, 134-35.
534
fraction
process
total silver silver
period amalgamated
kg produced kg
silver
A 0
1/1729 to 12/1740 67,661 0
S 67,661
A 0
1/1741 to 12/1748 60,545 0
S 60,545
A 0
1/1752 to 12/1759 71,005 0
S 71,005
A 0
1/1761 to 12/1770 106,210 0
S 106,210
A 0
1/1771 to 12/1780 150,355 0
S 150,355
A 0
1/1781 to 12/1790 125,489 0
S 125,489
A 0
1/1791 to 12/1800 143,017 0
S 143,017
A 1,498
1/1801 to 12/1806 74,912 0.02
S 73,413
total 799,194
Table 6-XXXII. Production of silver by smelting as registered in the Caja of Zimapán.

1729 to 1806
0.8 0.8 4 to 8 50 800 320
Table 6-XXXIII. Projected magnitude of the main environmental impact vectors

corresponding to the silver obtained by smelting and registered in the Caja of Zimapán.
535
6.1.12 Caja of Chihuahua.
According to TePaske this was the ‘last mining Caja created in New Spain’.920 The split
between smelting and amalgamation seems impervious to the decrease in mercury prices, most
probably due to the lead content of the ores. Santa Eulalia and Santa Barbara, the principal
mines feeding the refining haciendas that reported to the Caja in Chihuahua, are linked to the
few known lead bearing deposits of New Spain.921 It was only in the nineteenth century that
the amalgamation fraction increased (Figure 6-21). Overall smelting provide 60% of the silver
registered at this Caja, and amalgamation the remaining 40% (Table 6-XXXIV).922 The
abnormal range of mercury to silver ratios up to the end of the eighteenth century would
indicate a waste of inexpensive mercury due to amalgamating lead-rich ores.923 As commented
for other Cajas showing similar ranges, this would have an environmental impact, increasing
the amount of mercury physically lost to the environment (numbers in bold).
Tables 6-XXXV and 6-XXXVI complete the estimate of the magnitudes of the main
environmental impact vectors derived from the colonial refining activities of silver ores.
920
Ibid., 104.
921
Rice, "Silver-Lead Mines Santa Barbara," 208-209.
922
There is a gap in the TK set on silver production from 1791 to 1796 which I fill in Table 6-XXXIIV with
TePaske’s figure of 55,029 kg (figure in italic bold). Including this number, the total for silver in the table is
virtually the same as TePaske’s total for Chihuahua of 0.24 million kg. TePaske and Brown, Gold and Silver, 137.
923
The data in the TK set have an entry for the calendar year 1790 that indicates the purchase of 114,000 pesos of
mercury, approximately 128,000 kg of mercury. This amount of mercury does not correlate with the production
level of silver in the previous years, so I have placed it as mercury consumed in the following period.
536
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1785 1790 1795 1800 1805 1810 1815
duties
Chihuahua in the period 1788 to 1813. Raw data from TK set.
breakdown (kg)
fraction
process
total silver silver mercury mercury to

period amalgamated as liquid as volatilized
kg produced kg consumed kg as calomel silver ratio
silver mercury mercury
1/1785 to 12/1790 A 11,837 76,709 22,018 53,925 767 6.5

exc 1787
32,882 0.36
S 21,044
A 18,710 114,000 34,800 78,060 1,140 6.1
1/1791 to 12/1796 55,029 0.34
S 36,319
A 12,639 34,084 23,508 10,235 341 2.7
1/1797 to 12/1800 39,496 0.32
S 26,857
A 40,926 129,563 76,122 52,145 1,296 3.2
1/1801 to 12/1810 93,014 0.44
S 52,088
A 12,762 20,599 17,509 2,884 206 1.6
1/1811 to 12/1814 25,023 0.51
S 12,261
total 245,443 374,955 173,957 197,248 3,750
Table 6-XXXIV. Magnitude of the environmental impact vectors for calomel, liquid
registered in the Caja of Chihuaha.
537
copper equivalent
consumed consumed
1785 to 1814
0.2 0.10 3 0 1 60 0
Table 6-XXXV. Projected magnitude of other main environmental impact vectors

corresponding to the silver obtained by amalgamation and registered in the Caja of Chihuahua.

0.2 0.15 1 to 2 10 150 60
Table 6-XXXVI. Projected magnitude of the main environmental impact vectors

corresponding to the silver obtained by smelting and registered in the Caja of Chihuahua.
6.1.13 Aggregate totals for New Spain
Table 6-XXXVII gives a summary of all the magnitudes calculated for the Cajas of
New Spain analyzed in the previous sections. The first important conclusion is that
amalgamation accounted for approximately 64% of the silver produced in New Spain, and
smelting 36%. It coincides with the report in the historiography that in 1777 the Administrador
General de Minas reported that 40% of all production was by smelting.924 It also shows a much
924
As quoted in Mervyn F. Lang, "Amalgamacion y fundicion en la mineria americana.," in Estudios de historia
de las técnicas, la arqueología industrial y las Ciencias, ed. Juan Luis García Hourcade, Juan M. Moreno Yuste,
and Gloria Ruiz Hernández(Salamanca: Juanta de Castilla y Leon, 1998), 674. Up to the eighteenth century at
least half of the silver ores mined in Honduras were smelted. Newson, "Silver Mining Honduras," 52.
538
more balanced distribution between amalgamation and smelting output than what at times has
been reported in the
total total total

total silver total liquid total volatile total total volatile total slag total
total calomel total salt copper mineral silver
amalgamation mercury mercury woodland lead waste woodland
sulphate waste smelted
Zacatecas 7 11 2 0.1 200 17 20 4,000 3 15 to 30 160 1,300
Guanajuato 6 9 1 0.1 170 15 17 3,600 2 12 to 24 140 900
Mexico 6 10 2 0.1 190 16 19 3,900 2 12 to 24 140 950
Durango 2 4 1 0.05 70 6 7 1,500 4 18 to 36 220 1,500
San Luis Potosí 2 2 0.3 0.02 70 - 7 1,400 2 8 to 16 100 700

Guadalajara 3 5 1 0.06 80 7 8 1,600 1 5 to 10 60 400
Pachuca 2 3 1 0.04 50 5 5 1,100 1 3 to 6 40 270
Sombrerete 0.5 1 0.4 0.01 16 1 2 330 1 6 to 12 70 450
Bolaños 1 2 1 0.03 30 3 3 700 0.1 0.5 to 1 4 30
Rosario 1 1 0.4 0.02 25 2 2 500 0.3 2 to 4 20 130
Zimapán 0 0 0 0 0 0 0 0 0.8 4 to 8 50 320
Chihuahua 0.1 0.2 0.2 0 3 0 0 60 0.1 1 to 2 10 60
total New Spain 31 47 10 1 904 72 90 18,690 17 90 to 180 1,014 7,010
Table 6-XXXVII. Summary of main magnitudes projected for each of the main mining
Cajas of New Spain. Woodland figures expressed in units of a thousand ha, all the others in
units of a thousand t.
historiography.925 Three stages can be discerned in Figure 6-21. The data for the initial period
up to the 1640s are the one most subject to extrapolations of all, and probably overestimates
the importance of amalgamation (see discussion for each Caja above). In any case it was
925
An extreme case is the claim that 95% of all silver was produced by amalgamation. Castillo Martos, "Alquimia
en la metalurgia de plata y oro en Europa y America " xxiv. Humboldt estimated that the overall ratio of
amalgamated silver to smelted silver was 3.5 to 1. In the period around the beginning of the nineteenth century,
he estimated that smelting produced 10% of the total silver, and amalgamation (patio, cazo and barrel) the
remaining 90%. Humboldt, Essai politique, Tome IV, 49-50, 106. Modern historians have based their estimate of
the split between amalgamation and smelting on calculations using a deemed correspondencia value and the total
amount of mercury imported into New Spain. Brading and Cross used a correspondencia value of 110 marks per
quintal to arrive at a range between 70% and 87% for amalgamation. Brading and Cross, "Colonial Silver Mining:
Mexico and Peru," 570, 579. Garner estimated a range between 80% to 90 % amalgamation, based on a
correspondencia of 100 marks per quintal. Garner, "Long-Term Silver Mining," 918. Mendizabal proposed that
amalgamation accounted for 75% of all the silver produced in New Spain, without specifying his source.
Mendizábal, La mineria mexicana, 71-73.
539
characterized by a very lenient policy on the payment of mercury supplied to the refiners,
together with a source of tailings with a sunken cost that could be amalgamated. Both factors
can explain the peak in the amalgamated fraction of total silver production. From mid
seventeenth to mid eighteenth century the stricter policy on the supply of mercury levels the
playing field between amalgamation and smelting to the point both share equally the production
of silver in New Spain. It reflects the more natural balance between the two refining processes
based on the nature of ores available, with the bias on the part of the Crown towards
amalgamation somewhat muted. During this period ores with a silver content below 0.08%
continue to be discarded but make up an important part of the ore extracted from the mines (see
discussion in section 5.2, above). Starting in the 1740s according to Dobado and Marrero the
supply of mercury from Almadén increases, which would allow refiners to process an increase
in the ore being extracted.926
Amalgamation now shows a tendency to account for more than half of silver
production, a divergence of curves that becomes more significant once the step decreases in
the price of mercury are implemented in 1767 and 1776. The steep rise in silver production by
amalgamation in the latter part of the eighteenth century brings to mind the same drastic
increase seen in Potosí in the 1570s, and one possible cause is not only a sudden increase in
ore extraction but also the incorporation of tailings that could now be refined at a profit, due to
the low price of mercury and because their sunken cost of extraction was now nil, cutting their
amalgamation cost by more than half. Under these conditions it would have been impossible
for smelting to compete for the refining of these ores.
926
Dobado and Marrero, "The Role of the Spanish Imperial State in the Mining-led Growth of Bourbon Mexico's
Economy," 867. There is an element of chicken and egg in the arguments presented. Supply volume of mercury
by itself without a decrease in price does not induce an increase in silver production unless an increase has also
taken place in the amount of ore suitable for amalgamation at the prevalent pricing of mercury.
540
The downward price movements of mercury influenced the slope of the amalgamation
profile but not that of smelting. This reinforces the argument that it was the nature of the ore,
and not the price of mercury, that determined the choice of refining method. The reduction in
price of mercury did increase the amount of ore that could be processed at a profit in an
amalgamation hacienda, thus the increase observed after the 1760s. Thus each refining process
continued to be profitable with the ores available, and only the quantity available to each
process varied. No switching of ores or infrastructure is involved in a major way. There are
signs though that cheaper mercury was wasted in refining lead-rich ores in some Cajas.
4,000,000
3,500,000
3,000,000
2,500,000
kg silver
2,000,000
1,500,000
1,000,000
500,000
0
to 1620
to 1660
to 1700
to 1740
to 1780
to 1590
to 1600
to 1610
to 1630
to 1640
to 1650
to 1670
to 1680
to 1690
to 1710
to 1720
to 1730
to 1750
to 1760
to 1770
to 1790
to 1800
to 1810
Figure 6-21. Registry of silver by process, as projected for New Spain.
With regard to the magnitude of the main environmental impact vectors, I will report
total amounts over the colonial period rather than a single yearly average, since as seen for
each Caja, silver production is irregular and weighed towards the last 100 years of the colonial
period. Because the amalgamation/smelting split also varied with time the total amount cannot
be pro-rated according to the silver production by century. The total mineral waste voided into
waterways is the dominant value, with nearly 19 million t. In terms of weight, the major
541
chemical that was washed into waterways was salt, with an order of magnitude of 1 million t.
The remaining environmental impact vectors on water basins are an order of magnitude
smaller. The copper sulphate consumed, which would be washed away as copper ions or copper
salts, is projected as being greater (72 thousand t) than the water-insoluble calomel (47
thousand t). Liquid mercury, whether in the soil or washed away, is projected at 10 thousand t.
In the air it is lead as metallic lead or in lead compounds that dominates, with a range
between 90 and 180 thousand t. Volatile mercury would have accounted for around one
thousand tons over this whole period. Finally the impact on woodland is greatest for smelting,
with a maximum of 7 million ha of woodland consumed if no natural regeneration is factored
in, some 80 times greater than the requirements for amalgamation during the same period.
In Table 6-XXXVIII I present the amalgamation / smelting breakdown based on silver
produced for each Caja.927 Since the haciendas reporting to each Caja varied with time this is
only a snapshot of the average over the whole period, with overlap of regions over the whole
colonial period as new Cajas were hived off from existing ones. In order to visualize the
geographical distribution of the two refining processes I adopt Figure 6-1 as an approximation
to the whole colonial period and have shaded in Figure 6-22 the different territories of the Cajas
according to which refining process dominated their production: black for smelting (San Luis
Potosí (except for Catorce), Durango, Sombrerete and Zimapán, and Chihuahua is not shown
in the map), and grey where amalgamation predominated (Zacatecas, Guanjuato, Mexico,
Pachuca, Guadalajara, Bolaños and the district of Catorce in San Luis Potosi). The magnitude
of the environmental impact vectors due to smelting could however be greater even in regions
927
As a crosscheck, the total in Table XXXVII is rounded off to 48 thousand t, compared to 49.1 thousand t
reported in TePaske and Brown, Gold and Silver, 116.
542
nominally dominated by amalgamation, subject to the total amount of silver being produced by
both methods.
silver by silver by
%
Caja amalgamation smelting
amalgamation
% smelting
t t
Zacatecas 6,522 3,262 67% 33%
Guanajuato 5,895 2,322 72% 28%
Mexico 6,346 2,357 73% 27%
Durango 2,382 3,677 39% 61%
San Luis Potosí 2,249 1,697 57% 43%
Guadalajara 2,656 1,003 73% 27%
Pachuca 1,781 668 73% 27%
Sombrerete 536 1,115 32% 68%
Bolaños 1,112 67 94% 6%
Rosario 777 314 71% 29%
Zimapán 0 799 0% 100%
Chihuahua 97 149 39% 61%
total New Spain 30,353 17,429 64% 36%
Table 6-XXXVIII. Amalgamation and smelting by Caja over the whole colonial period in
New Spain. Source data from Table 6-XXXVII.
This is clear in the series shown in Figures 6-23 and 6-24 where I plot the magnitude
of the environmental impact vectors corresponding to smelting. While Durango remains the
area that would have been most affected by the environmental impact of smelting, it is closely
followed by areas where amalgamation predominated, such as Zacatecas, Mexico and
Guanajuato. With regard to the impact of amalgamation by region, it follows more closely the
ranking of Cajas where amalgamation dominated, as illustrated in Figures 6-25 to 6-28. All
543
these are average representations over the whole period, and according to the Caja could have
changed substantially during smaller historical spans.
II
VIII
IX
VII
I
IV X
I Bolaños
II Durango III VI
III Guadalajara
IV Guanajuato V
V Mexico
VI Pachuca
VII San Luis Potosí
VIII Sombrerete
IX Zacatecas
X Zimapan
Figure 6-22. Approximate geographical distribution of main refining processes applied in the
Cajas of New Spain. Black: smelting, Grey: amalgamation. Adapted from Figure 6.1.
To carry the regional environmental analysis forward on a more detailed basis will
require incorporating data on waterways and water basins, historical landfills, wind roses and
geographical contours, relative distribution of refining haciendas by location, size and refining
process, population centres, agricultural centres and cattle rearing, woodlands used to source
fuel, regeneration rates for the local woodland, transit routes to and within each region. In
addition a detailed knowledge on the changes in the architectural details of the refining
544
haciendas of each region over time is required, so as to be able to model the deposition of lead
around each establishment.928
Bolaños
Chihuahua
Rosario
Pachuca
Zimapán
Guadalajara
Sombrerete
San Luis Potosí
Mexico
Guanajuato
Zacatecas
Durango
0 5 10 15 20 25 30 35 40
range magnitude volatile lead vector (thousand t)
higher range lower range
Figure 6-23. Listing of Cajas by the magnitude of the vector corresponding to lead and lead
compounds. Data from table 6-XXXVIII
928
In chapter 2 I made reference to the regional studies identifying primary economic functions carried out by
Prof. Salazar in the region around the town of San Luis Potosí. Another study has centred on determining the role
of ‘mining as a creator of economic spaces due to its great organizing power’ - ‘mineria como creador de espacios
económicos por su gran poder organizador’. The authors draw up maps of geographical networks that grew
around the mining and refining activities in Pachuca and Real del Monte in the nineteenth century. Saavedra Silva
and Sánchez Salazar, "Espacio Pachuca-Real del Monte," 83-97. Each cluster of refining haciendas creates by
their sole presence a source of an economic force field, attracting by its needs the other economic activities that
grew out of refining. Overlaid on this economic ‘gravitational’ contour map would lie the vectors of environmental
impact due initially to the original refining of silver and then to the growth of other economic activity.
545
Bolaños
Chihuahua
Rosario
Pachuca
Zimapán
Guadalajara
Sombrerete
San Luis Potosí
Guanajuato
Mexico
Zacatecas
Durango
0 200 400 600 800 1,000 1,200 1,400 1,600
thousand ha
Figure 6-24. Woodland consumed by smelting and amalgamation according to Caja. Data
from Table 6- XXXVIII.
Zimapán
Chihuahua
Sombrerete
Rosario
Bolaños
Pachuca
San Luis Potosí
Durango
Guadalajara
Guanajuato
Mexico
Zacatecas
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500
thousand t
Figure 6-25. Ranking of Cajas according to the magnitude of the environmental impact
vector of mineral waste voided into waterways. Data from Table 6-XXXVIII.
546
Zimapán
Chihuahua
Sombrerete
Rosario
Bolaños
Pachuca
San Luis Potosí
Durango
Guadalajara
Guanajuato
Mexico
Zacatecas
0 50 100 150 200 250
thousand t
vector of salt voided into waterways. Data from Table 6-XXXVIII.
Zimapán
San Luis Potosí
Chihuahua
Sombrerete
Rosario
Bolaños
Pachuca
Durango
Guadalajara
Guanajuato
Mexico
Zacatecas
0 2 4 6 8 10 12 14 16 18
thousand t
vector of copper sulphate consumed and voided into waterways. Data from Table 6-XXXVIII.
547
Zimapán
Chihuahua
Sombrerete
Rosario
San Luis Potosí
Bolaños
Pachuca
Durango
Guadalajara
Guanajuato
Mexico
Zacatecas
0 2 4 6 8 10 12
thousand t
vector of calomel voided into waterways. Data from Table 6-XXXVIII.
6.1.14 Aggregate totals for Mexico, 1820 to 1900
Unfortunately the data that quantify silver production according to refining process is
limited in the period it covers (fiscal years beginning 1877 to 1896), contains evident errors in
its figures for smelting as of 1893, and the yearly totals calculated from its data are between
20% to 40% less than the yearly data in Table 6-XXXIX. Bearing this in mind, Figure 6-29
plots the percentage of amalgamation and smelting registered for a period when barrel
amalgamation and leaching (lixiviación) were also being applied.929 The latter reaches a peak
fraction of 0.18 in the fiscal years 1891 and 1892, but by 1896 the total silver registered has
decreased to 36,995 kg from its peak at 203,932 kg. The average fraction of silver produced by
barrel amalgamation in this period is 0.04.930
929
The leaching process mentioned in the statistics does not refer to cyanide leaching but to the earlier processes,
such as the hyposulphite process included in Collins, Metallurgy of Lead & Silver, 186-241.
930
Flores Clair, Velasco Avila, and Ramírez Bautista, Estadísticas mineras, II, 161-62. Laur and Duport estimated
that smelting produced 10% of silver in Mexico, though Laur cautioned that ‘it is however not possible to establish
548
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1876 1878 1880 1882 1884 1886 1888 1890 1892 1894
amalgamation smelting
Figure 6-29. Amalgamation and smelting fraction of silver presented at the Mexican mints,
1876 to 1892. Raw data from footnote 900.
In Table XXXIV I present a projection of the magnitudes of the main environmental
impact vectors for the whole of Mexico during the nineteenth century, subject to the limitations
in each of the data sets signalled in the previous paragraphs. In the absence of mercury
consumption data I have assumed an average mercury to silver ratio of 1.8. The other ratios
per kg of silver are the same as applied to colonial New Spain. I present the data in such a
manner the projected breakdown of the magnitudes of the environmental impact vectors can
be adjusted in the future as better figures comes to light on a regional basis.
with certitude their relative importance’ - ‘il n’est cependant pas possible d’établir, avec certitude, leur
importance relative’.Laur, "De la metallurgie de l'argent au Mexique," 106. ; Duport, Métaux précieux au
Mexique, 86.
549
mercury liquid volatile copper mineral

amalgamation calomel salt woodland
total silver consumed mercury mercury sulphate waste
period
produced
fraction kg thousand t thousand ha
1821 - 1830 2,648,400 0.75 1,986,300
1831 - 1840 3,309,900 0.75 2,482,425
1841 - 1850 4,203,100 0.75 3,152,325
1851 - 1860 4,569,500 0.75 3,427,125
1861 - 1870 4,969,500 0.75 3,727,125
1871 - 1880 6,317,362 0.75 4,738,022
1881 - 1890 9,116,860 0.7 6,381,802
1891 - 1899 13,796,861 0.7 9,657,803
1821-1899 35,552,926 64 30 5 0.4 1,100 90 22,000 100
smelting volatile lead slag woodland

total silver
period thousand
produced fraction kg thousand t
ha
1821 - 1830 2,648,400 0.25 662,100
1831 - 1840 3,309,900 0.25 827,475
1841 - 1850 4,203,100 0.25 1,050,775
1851 - 1860 4,569,500 0.25 1,142,375
1861 - 1870 4,969,500 0.25 1,242,375
1871 - 1880 6,317,362 0.2 1,263,472
1881 - 1890 9,116,860 0.2 1,823,372
1891 - 1899 13,796,861 0.2 2,759,372
1821 - 1899 10,771,317 50 to 100 700 4,300
Table 6-XXXIX. Projected magnitudes of main environmental impact vectors for

amalgamation and smelting in Mexico, 1821 to 1899.
6.1.15 Environmental impact vectors, sixteenth to nineteenth century: conclusions.
Since at present there is no regional breakdown for the data corresponding to republican
Mexico in the nineteenth century I will limit the following discussion to a comparison of the
projected aggregate totals for this period compared to those for New Spain (Table 6-XL). The
amount of silver produced in Mexico during the nineteenth century was approximately equal
to all the silver produced in the colonial period, and most of this production took place between
1850 and 1900. Thus the yearly environmental impact imposed upon the refining regions in
Mexico from amalgamation and smelting processes was mostly compressed into a period of
550
total total total total total total

total silver total total mineral total slag total
mercury in liquid volatile total salt copper silver volatile
amalgamation woodland waste waste woodland
calomel mercury mercury sulphate smelted lead
New Spain 31 47 10 1 900 70 90 19,000 20 90 to 180 1,000 7,000
Mexico 36 54 9 1 1,000 90 100 22,000 11 50 to 100 700 4,300
Table 6-XL. Total magnitude of environmental impact vectors from amalgamation and
smelting as projected for New Spain and Mexico. All numbers have been rounded off, and
expressed in thousand t except for woodland which is in thousand ha.
approximately 50 years. In the case of New Spain the span is greater on average but can vary
substantially with each Caja since production was never homogeneous over time nor, as
discussed, limited to one major process during the whole period. Bearing these conditions in
mind, the following observations can be made:
a, Due to the lower incidence of smelting and an increased efficiency of the furnaces,
60% less woodland was consumed as a result of smelting and about half the levels of emissions
of lead and lead compounds are projected to have been reached in the nineteenth century
compared to the whole colonial period, even though the total silver production levels in both
cases were similar. These lower levels of emissions would have counterbalanced in part the
effect of the shorter time period in which they took place, to produce lower yearly quantities
than expected had the same context persisted from the colonial era.
b. The levels of salt, copper sulphate, calomel, liquid mercury, waste minerals voided
into the waterways and woodland consumed for amalgamation are higher for Mexico than New
Spain, reflecting the higher level of amalgamated silver. The stress on the environment of
Mexico from these vectors during the nineteenth century would have been on average up to
five times greater than in colonial New Spain.

551
c. Lead (and its compounds) is projected as having been the main heavy metal issued
to the air as a result of the refining of silver in both periods, with a ratio of lead to volatilized
mercury estimated at an average of approximately 100 to 1. Even if all the consumption of
mercury had been by the short-term physical loss to the air (an assumption negated by the
arguments presented in previous sections), lead would still have been as important as mercury
in any analysis of the environmental impact of heavy metals as a result of historical silver
refining. Since calomel is the main cause for the consumption of mercury during patio
amalgamation, emissions to the atmosphere of lead and its compounds constitute the main
source of heavy metals issued to the air during the historical refining of silver ores in New
Spain. This conclusion holds valid even if amalgamation were to have produced 90% of all the
silver of the colonial period, a percentage negated by the data from the individual Cajas.
d. To what extent the losses of calomel and liquid mercury would ultimately end up
contributing to air emissions of mercury over decades and centuries remains to be determined.
Until the whole life cycle of calomel in waterways, river beds and landfills is established, its
effect as mercury source to the environment in the long term is unknown.931
e. Overall, smelting created a much higher level of pressure on the human communities
of New Spain than amalgamation, due to woodland depletion and to the large amount of heavy
metal (lead and its compounds) deposited in the areas around each smelting hacienda. To this
should be added the pollution due to the dressing of the ores using water, and the fact that its
potential as a household industry would have been a major source of toxic fumes to the local
community. Amalgamation on the other hand shows a more attenuated environmental impact
931
Calomel is a mercury compound that has not been studied regarding its life-cycle in aquatic environments, or
its impact on organisms at different levels of concentration. What is reported in the historiography on calomel is
very limited: it has been used as a diuretic, a laxative, a means to increase the rosiness of the cheeks of babies and
as a topical disinfectant.
552
profile. Calomel in the short term would have trapped most of the mercury consumed into a
non-soluble solid entombed within tons of fine mineral silt. In the long-term no conclusion can
be made as yet. The remaining 15% lost as liquid mercury to the ground and water is still an
important fraction, but in absolute terms it was lower than the loss of lead as metallic lead or
in lead compounds for the same amount of silver refined. Amalgamation required much less
fuel (except for the minority barrel or cazo processes) than smelting, and its impact on
woodland some 60 times lower per kg of silver produced. Fine mineral silt was its main waste
in terms of weight to the environment, then salt, both intrusions on the environment but of
another class to losses of lead and lead compounds or to the decimation of woodlands.
Amalgamation was the lesser of the two evils.
In relative terms, was the overall level of environmental impact similar to other
historical industrial processes up to the eighteenth century? There is no historic benchmark for
amalgamation, since it was only used at an industrial scale in the New World. In the case of
lead from smelting the most relevant comparison is with the level of lead emissions estimated
to have taken place during Roman times, when the lead industry reached levels not to be seen
again until the modern industrial era. It has been estimated that 5 to 10,000 t/y of lead were lost
as air emissions at the peak of its use during Roman times, 10 to 20% of the total peak
production of 50,000 t/y.932 In the case of my projection for New Spain, a conservative range
of 90 to 180,000 t of lead has been proposed as being lost over the whole colonial period as air
emissions, approximately 5 to 10% of the total lead required for smelting silver ores of 2%
silver content (Chapter 2).
932
Nriagu, "Tales Told in Lead," 1622-23.
553
Over the whole of New Spain this would correspond to a range up to 900 t/y of lead.
The Cajas whose refiners would have exhibited the highest peaks in yearly averages of lead
emissions are Durango (1750s) and Zacatecas (1680s) with over 400 t/y, then Guanajuato
(1750s) and Sombrerete (1800s) with over 300 t/y, followed by San Luis Potosi with under 200
t/y (1650s and 1790s). For nineteenth-century Mexico it is not possible with the data available
to calculate regional peaks. Overall, an average of at least 1,200 t/y of lead would have been
reached in republican Mexico, especially towards the end of the nineteenth century. These
levels have two important implications. First, taking into account the disproportion between
the levels of Roman lead production and the production of silver in New Spain, the levels of
historical lead emissions associated with silver smelting in New Spain and Mexico reveal a
strongly polluting industry per kg of silver obtained.
Second, studies on historic levels of lead deposition in Europe detect the spike in
ambient levels of lead that correspond to the Roman period and to silver refining in Germany.933
Now the average production of silver by smelting and amalgamation was approximately 190
t/y in New Spain and 580 t/y in nineteenth century Mexico. During the colonial period these
are production levels over a continuous period of some 250 years that were not matched
elsewhere in the world. China produced on average 7 t/y between 1401 and 1440, then on
average less than 2 t/y between 1441 and 1520, but by 1636, bullion production in China was
negligible.934 Japan between the sixteenth and seventeenth century is cited as exporting up to
200 t/y to China but was unable to maintain this level of production beyond a few decades.935
933
For example Martinez Cortiza A et al., "Atmospheric Pb Deposition in Spain During the Last 4600 Years
Recorded by Two Ombrotrophic Peat Bogs and Implications for the Use of Peat as Archive," The Science of the
Total Environment 292(2002).; Ingemar Renberg, Richard Bindler, and Maja-Lena Brännvall, "Using the
Historical Atmospheric Lead-Deposition Record as a Chronological Marker in Sediment Deposits in Europe,"
The Holocene 11, no. 5 (2001).
934
Atwell, "International Bullion Flows," 76,78.
935
Atsushi Kobata, "The Production and Uses of Gold and Silver in Sixteenth- and Seventeenth-Century Japan,"
The Economic History Review 18, no. 2 (1965): 248.
554
Africa and the sub-continent of India do not figure as historical global sources of silver. Europe
was the only alternate source for silver to the Hispanic New World until mid-nineteenth
century, but the production figures are an order of magnitude lower with respect to the New
World. The peak figure available for the mines of Freiberg indicate nearly 12 t/y by the end of
the eighteenth century, while at nearby Hartz the highest annual average between 1718 and
1724 was 14 t/y. The mines at Konigsberg, Norway, produced on average 3 t/y between 1624
and 1805.936 According to Humboldt just the silver sulphide in the ore from Valenciana
(Guanajuato) produced in one month 7 t of silver, half of what all the mines of Saxony were
producing in one year.937 In the case of the Slovakian mines, the highest peak of production
was reached at around 25 t/y.938 This comparative review of silver production levels indicates
that it should be possible to detect pre-twentieth century spikes in deposited lead in those
provinces of New Spain / Mexico where smelting of silver ores was the main historic refining
technique.
With respect to the level of woodland depletion, a high ratio of 1,000 to 1 of charcoal
to silver was observed both in Europe and New Spain up to the end of the eighteenth century.
A total of 7 million ha of woodland in New Spain are projected to have been required for
refining of silver, ignoring natural cycles of regeneration. Had Western Europe been forced to
supply an equivalent amount of silver from hypothetical silver deposits within its territory, its
forest cover would have been depleted by this amount up to the year 1800. This is another
instance of the concept of ‘ghost acreage’, now applied to silver refining. 939 What would have
936
Burkart, "Memoria Real del Monte," 97, 98, 100.
937
Humboldt, Essai politique, Tome III, 363.
938
Teich, "Born's amalgamation process," 310.
939
The concept of ‘ghost acreage … the computed, non-visible acreage which a country would require as a
supplement to its visible agricultural acreage … in order to be able to feed itself’ was first introduced by Georg
Borgstrom in 1965 to denote how trade or fishing allowed a country a virtual expansion of its food production
capacity without having to sacrifice its own land. Georg Borgstrom, The Hungry Planet; the Modern World at the
Edge of Famine (New York: Macmillan, 1965), 71. The same concept can be applied to silver refining, substituting
555
been the impact on European economies from this additional depletion of their woodlands? It
is not easy to find an estimate of residual woodlands for Europe in the Early Modern Period,
since even the definition of the historical term for forests is under scrutiny. The latest estimation
by Kaplan et al, which in the words of the authors is the lowest of recent research, concludes
that the forest cover in the area defined as Western and Central Europe dropped from 24.9% in
the year 1400 to 5.8% by 1850.940 Using the modern land areas for the countries in their table,
I arrive at a decrease of 47 million ha in this period, leaving this part of Europe by 1850 with
just 14 million ha of woodlands. Another estimate is by Williams, who proposed that between
18 and 25 million ha of land were cleared for agriculture in Europe between 1650 and 1749.941
Under both scenarios, had another 7 million ha of woodland been cleared during this period to
refine silver, the economic impact on Europe would have been major.
How would this level of forest depletion have impacted the forest cover of colonial
New Spain or Mexico in the nineteenth century? In the absence of data on forest cover in the
region during the historical periods of interest, I will use data from the present century. The
forest cover of Mexico in the year 2010 was 65 million ha. The forest depletion rate between
2005 and 2010 was measured at 0.24% per year.942 A loss of 7 million ha represents 10% over
250 years, without the attenuation of natural recovery or forest husbandry. It could be argued
that modern population pressures on these natural resources are greater than what would have
been the pressure from the colonial silver refining processes.
food for the amount of silver imported and the agricultural acres for the amount of woodland required to produce
the fuel for refining. Analogous calculations can be made on the basis of ‘ghost emissions’ of lead, calomel,
mercury, salt and copper salts. It can be extended to the effects on health of workers and communities if the
appropriate quantitative factors can be estimated.
940
Jed O. Kaplan, Kristen M. Krumhardt, and Niklaus Zimmermann, "The prehistoric and preindustrial
deforestation of Europe," Quaternary Science Reviews 28, no. 27–28 (2009): 3023.
941
Michael Williams, Deforesting the Earth from Prehistory to Global Crisis : an Abridgment (Chicago:
University of Chicago Press, 2006), 172.
942
Global Forest Resources Assessment 2010, 228, 233.
556
The environmental impact from smelting, plus the historically unprecedented
anthropogenic loss of liquid mercury to the rivers and soil at an average of 40 t/y continuously
over 250 years of the colonial period (over 100 t/y in Mexico during the nineteenth century),
is high considering the modest level of metallurgical activity in New Spain compared to
metallurgical output in England during a similar period. In New Spain the refining of metals
was basically limited to silver, and went from an average annual production of approximately
120 t in the decade 1601-1610 to an annual average of 480 t in the decade 1791-1800.943 In
England annual production from 1600 to 1800 via smelting went from 4,000 to 50,000 t of
lead, 600 to 2,000 t of tin, 50 to 6,000 t of copper and from around 15,000 to 25,000 t of bar
iron.944 These amounts exceed by far in raw ore processed and final metal refined the statistics
for the silver industry of New Spain. The problem lies in how to judge the relative
environmental impact of both activities (silver refining and the effects of the metallurgical
activity in England) in terms of the net economic benefit that ultimately derived to the various
players involved: the local economy of New Spain, the economies of Spain and of England.
The main challenge is that silver did not remain within the country that produced it
(Spain or Mexico) and that its production ultimately benefited a global economy reaching from
Europe to India and China. Thus the ultimate benefactors of the ‘refining ghost acreage’ (and
lead and mercury ‘ghost emissions’) are many, not all at the core of European based empires,
some at the non-European periphery, and even a residual value benefited the single region that
bore the environmental consequences. Defining the total distribution of benefits and
quantifying the environmental cost of a process that impacted in unknown ways the health and
943
TePaske and Brown, Gold and Silver, 115-16.
944
Roger Burt, "The Transformation of the Non-Ferrous Metals Industries in the Seventeenth and Eighteenth
Centuries," The Economic History Review 48, no. 1 (1995): 28.; Peter King, "The Production and Consumption
of Bar Iron in Early Modern England and Wales," ibid.58 (2005): 6.
557
culture of the local communities is not an evident task, both as a unique historical event or in
comparison to other non-European industrial processes that have been game-changers in the
world economy, i.e. the porcelain industry of China or the textile industry of India.
6.2 A change of paradigm.
The problem that has been faced in each of the chapters of this thesis is identifying what
exactly is the current paradigm that guides the discussion on the environmental history of silver
refining in the New World. As Kuhn argued, ‘in the absence of paradigms all the facts that
could possibly pertain to a development … are likely to seem equally relevant’.945 This would
explain the constant contradiction in views that introduce each chapter: silver ores in the New
World are poor/rich in silver; smelting is more complex / more simple than amalgamation;
volatile mercury / calomel are the main reason for the consumption of mercury during
amalgamation; lead is / is not a major factor in air pollution from silver refining; the same /
more woodland was consumed for smelting than for amalgamation; mercury was subsidized
by the Spanish Crown / by the Fuggers; the production cost of amalgamation is less / more than
smelting; mercury was the only / just another option for Spain to extract silver from the New
World. Each of these dichotomies, which are definitely not incommensurable, have to be
resolved in order to understand the human actions behind the historical events. Again citing
Kuhn, the advantages of an explicitly accepted paradigm that can be verified by hard data is
that it is ‘not necessary to build a field anew once a paradigm is taken for granted’.946 Many of
the branches in the path of this dissertation have been the result of an absence of a clear
paradigm that applies to the quantitative facets of this history.
945
Thomas S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1965), 15.
946
Ibid., 19.
558
Environmental history is a prime example of interdisciplinary research, since it is based
on the interpretation of historical texts and architectural studies, and on the disciplines of
chemistry, physics and economics. It certainly seems feasible to establish a model for the
interpretation of historical events based on quantitative data.947 In Kuhn’s terminology, the
paradigm under whose temporal aegis the quantitative study of the environmental history of
silver refining in the New World is carried out should be clearly identified and its evidence laid
out for critical analysis by all parties of the relevant community. The transparency of the
method becomes paramount over the numerical data itself, since the latter is always open to
new sources of better data. I have used as a guide two of the most intuitive definitions of
paradigm used by Kuhn: the map agreed upon or the set of rules to guide the activity of the
interested community in their common field of research.948
The only explicit paradigm that has guided most of the recent research in various fields
on the environmental legacy of historic silver refining is based on the assumption that all the
mercury consumed during amalgamation was in the form of physical losses, of which the
majority (65 to 85%) is posited as having taken place during the heating stage of the amalgam.
This paradigm was initially adopted in order to estimate projections of global deposition of
mercury over time. The origins of this paradigm were not experimental, born from field trials
of clay caperuzas or metal capellinas under controlled conditions to measure the escape of
mercury vapour and the ambient concentrations in the workplace. The only research carried
947
As Thommen has stated in his studies on the environmental legacy of ancient Greece and Rome: ‘for a more
adequate reconstruction of ancient environmental conditions, [an attempt is made] to include research from other
disciplines, even if no comprehensive interdisciplinary approach can as yet be realized’. Lukas Thommen, An
Environmental History of Ancient Greece and Rome (Cambridge; New York: Cambridge University Press, 2012),
1.
948
A more formal definition given by Kuhn is that paradigms are ‘achievements that are sufficiently
unprecedented to attract an enduring group of adherents away from competing modes of scientific activity’ and
‘also sufficiently open-ended for the refined group to continue’ working in the area so as to resolve further issues
in the field of the paradigm. Groups ‘whose research is based on shared paradigms are committed to the same
rules and standards’, in Kuhn, Structure of Scientific Revolutions, 10, 11, 90.
559
out in a laboratory at the end of the twentieth century was the paper published by Johnson and
Whittle which confirmed the formation of calomel during amalgamation, but it was ignored.949
The current paradigm also ignored the observations made in the historical texts up to the
nineteenth century and only retained the gross figures of mercury consumed as recorded by
historians.950 I can conjecture that the track record of modern artisanal gold amalgamation
determined this paradigm, but I cannot explain why calomel disappeared from every model on
global deposition of mercury, in spite of the wealth of data in the historiography of the
nineteenth century.
According to Kuhn a paradigm begins to be questioned when it is unable to explain
anomalies in its field.951 Three major anomalies persisted that were not be explained by the
paradigm based on 100% physical loss of mercury and no mention of lead:
a) the silence by first-hand observers of the nineteenth century on the assumed losses
of volatile mercury during the heating of the amalgam,
b) the historical silence on mercurialism, incompatible with the amounts of mercury
deemed to have been issued to the air, and
c) the absence of historic high levels of mercury deposited from the air in areas close
to amalgamation centres of silver ores, as has been evidenced by Cooke et al and Engstrom et
al.952
949
Johnson and Whittle, "The Chemistry of the Hispanic-American Amalgamation Process."
950
Nriagu, "Legacy of Mercury Pollution."; Robins, Mercury, Mining and Empire.
951
Kuhn defines an anomaly as an observation or issue ‘that cannot be aligned with professional expectation’
Kuhn, Structure of Scientific Revolutions, 6.
952
Colin A. Cooke et al., "Over three millennia of mercury pollution in the Peruvian Andes," Proceedings of the
National Academy of Sciences 106, no. 22 (2009).; Engstrom et al., "Atmospheric Hg Emissions from
Preindustrial Gold and Silver Extraction in the Americas: A Reevaluation from Lake-Sediment Archives."
560
Yet by the end of the nineteenth century all the relevant empirical facts required to
establish the road map for future research in any field related to historical silver refining had
already been reported:
 The dangers of mercurialism from accidents during the heating of the amalgam,
but not under normal operational practice
 the toxic nature of the smoke from smelting haciendas
 deforestation
 the effect of solid wastes, including mercury, on water sources
 the formation of a chloride of mercury as a by-product of amalgamation
None of these conclusions can be faulted even with the benefit of hindsight. The
absence of detailed theoretical or analytical tools and software programs to model air emissions
did not impair the empirical notations of these first-hand observers of the refining processes
being applied, even if they could not have known the structure of all the chemical compounds,
or measure their concentration in the water, soil or air. Most of these historical testimonies
became invisible in the second half of the twentieth century, victims to a scientific version of
historicity. This was not a shift of paradigm, since no paradigm had previously existed. This
was the creation of a paradigm devoid of hard data, which decided to focus solely on the one
vector of environmental impact that was considered irrelevant in the historiography of the
nineteenth century: copious emissions of volatile mercury during the heating of the amalgam.
An alternative paradigm can be proposed based on the historiography up to the end of
the nineteenth century, complemented by modern chemical and economic analysis. It is not
novel in the sense that every vector included had already been identified in the historical
sources, but I have accompanied it with an explicit quantitative base on which it can be judged
561
and improved. This paradigm is based on the following general scenario, defined by the
numerical ratios ratios which have been presented in the previous sections and chapters:
1. Amalgamation accounted for just over 60% of the silver produced in New Spain,
smelting just under 40%.
2. Lead (and its compounds) is the main heavy metal issued to the atmosphere as a
consequence of the refining of silver ores in New Spain / Mexico.
3. The main consumption of mercury is via the formation of calomel. The loss of liquid
mercury in the water used to separate the amalgam, or that percolated to the soil within the
haciendas, plays a secondary role. The loss of volatile mercury in the heating stage of the
amalgam was only due to accidents and overall is considered to be very low.
4. The correspondecia is not an empirical number but the reflection of the
stoichiometric relation between mercury and silver derived from the chemistry of the reactions
that take place during amalgamation.
5. Fine mineral silt washed away from haciendas into streams and landfills constituted
the main source by weight of solid waste from amalgamation haciendas.
6. Salt was the second major component in weight voided in major amounts into the
water basins downstream of the haciendas.
7. Copper compounds were voided into waterways in at least equal amounts as calomel.
8. The consumption of woodland was due primarily to smelting, with amalgamation
playing a minor role.

562
9. The relative costs of production between amalgamation and smelting varied
according to the historical period, and at times overlapped. Neither one nor the other was more
profitable across all regions and across all time periods.
10. It is the chemical nature of the silver ores that determined in the first instance the
choice of refining process.
6.3 The human choices.
The main conclusion of this thesis is that a paradigm based on lead and calomel
determined the material impact of the environmental history of silver refining in the New
World. The previous sections have proposed the chemistry that sustains this paradigm, and
have calculated the resulting quantitative ratios that define the mass balances of each process
and their economic consequences. This paradigm is only the initial stage for a revision of the
technical, scientific and historical narrative to date:
 on the technical side, the patio process needs to be replicated to be studied in
depth. How the very dense mercury droplets managed to interact with silver
compounds embedded within the matrix of the milled ore just some 25 cm thick
without the majority percolating by gravity through the planks or paving stones
into the soil below over a period of weeks is still an open question
 the search for calomel in river beds and landfills, and the study of its
environmental life-cycle is needed
 the mapping of historical airborne lead depositions up to the end of the
nineteenth century within the territory of Mexico will confirm or question the
paradigm proposed
563
 the detailed reconstruction of the architecture of the various phases of refining
haciendas is a work in progress, but is of vital importance in the modelling of
dispersion of airborne lead fume
 once detailed information on refining, agricultural and other economic activities
in relation to population centres and water basins can be expanded over all
mining and refining regions within New Spain, it will provide the framework
for a more detailed analysis on the impact of the chemicals on the health of the
workers and the communities, both at the time and their residual impact up to
modern times.
Beyond the technical issues conscious choices on policy were also made at each stage
of the process, not always guided by full knowledge of the science involved, but certainly
within a continuum of the technical know-how available in each period. I will briefly review
some of the human choices that were exercised and that also played a major role in the
environmental history of silver refining as a whole.
6.3.1 What did they know and when did they know it?
The wide use of lead during Roman times has been widely documented, as well as the
relative silence in the texts up to the nineteenth century on the effect of lead exposure on the
workers involved.953 When smelting was applied in the New World, the dangers and
precautions of working with lead were sufficiently recognized in practice, if not in medical
theory. Viceroy Toledo in Potosí specified that smelting furnaces had to have higher chimneys
than those used within amalgamation Ingenios due to the greater hazard from lead fumes:
953
Jerome O. Nriagu, "Occupational Exposure to Lead in Ancient Times," Science of the Total Environment 31,
no. 2 (1983).; Sven Hernberg, "Lead Poisoning in a Historical Perspective," American Journal of Industrial
Medicine 38, no. 3 (2000).
564
‘The danger of lead poisoning from vapors given off during the preparation of lead flux for
smelting silver, or in the recovery of lead after smelting, was recognized in Toledo’s mining
ordinances of 1574. To recover lead, refiners should use an enclosed building with chimneys
some 7 meters tall (4 estados) … the furnaces used to drive off mercury as vapor should be set
apart from the refinery itself, and equipped with chimneys some 5 meters in height (3 estados)
“so that the Indians shall not receive the smoke in any fashion”’.954
Equally, Garcilaso de la Vega’s (1539-1616) views on the responsibility of a monarch
to any of his subjects handling mercury quoted in the Introduction reflects a perfect
understanding of the risks posed by mercury. Amalgamation was much more recent than
smelting, but mercury had been known to be a poison for workers involved in its handling
many centuries before the Spaniards used it indiscriminately for silver refining in the 16c. Pliny
in Roman times had commented on its toxicity.955 Biringuccio wrote: ‘[Mercury] is numbered
among the poisons. It has the property of contracting the nerves of those workers who extract
it from ore if they are not very careful, and it makes the limbs of those who continually handle
it weak and paralyzed’.956 Martin del Rio, s.j. (1551-1608) argued against the use of
alchemically transmuted gold for medicinal purposes due to the ‘noxious qualities’ it retained
of mercury used in the alchemical process.957 The dangers of using mercury were published
throughout the period of silver refining in the New World.958
954
Bakewell, Miners of the Red Mountain, 150.
955
Pliny, The Elder Pliny's Chapters on Chemical Subjects (London: E. Arnold & Co., 1929), 111.
956
957
Quoted in M. Baldwin, "Alchemy and the Society of Jesus in the Seventeenth Century: Strange Bedfellows?,"
Ambix 40, no. 2 (1993): 44.
958
Rosen has listed the following authors on mercury poisoning: Gabriele Fallopius (1523-62) in his treatise De
Meteallis et Fossilibus noted the poisoning by mercury of workers at the mines and mentioned they could not
work more than three years as a result; Andrea Mattioli, a contemporary of Fallopius, comments on the chronic
mercurialism of the workers at the mercury mines at Idria (modern day Slovenia); Pieter van Foreest (1522-1597)
of Delft; Paracelsus dedicates the third book of his monograph on miners’ diseases to mercury poisoning (1567);
in 1665 Walter Pope comments with detail on the symptoms of mercury poisoning in workers; Bernard de Jusein
in 1719 presented a memoir on the situation of workers at the Almadén mercury mine in Spain to the Academy
of Sciences; some fifty years later Giovanni Scopoli described the effects of mercury poisoning on miners of Alto
Isonzo. As to the growing awareness of the age on the health concerns of miners, in 1700 Bernardo Ramazzini
published in his treatise on occupational medicine (as quoted in Rosen): ‘We must own that some arts entail no
small mischief upon the respective artisans, and that the same means by which they support life, and maintain
their families, are oftentimes the cause of grievous distempers which hurry them out of this world’. George Rosen,
The History of Miners’ Diseases. A Medical and Social Interpretation (New York: Schuman’s, 1943), 39-133. A
565
Even more important, Spanish workers had been producing mercury from Almadén
before the New World was conquered, making Spain one of the two places in Europe with first-
hand experience on the occupational hazard of working with large quantities of mercury. By
the 1550s the workers at Almadén had lobbied for special dispensations such as tax exemptions
since their work ‘damaged their mouths’ and led to a doctor ‘to cure the azogados [the victims
of mercurialism]… since there are no people on Earth … so subject to illness as the azogados’.
An apothecary was assigned, with free medicine supplied for the poorer workers. Such were
the perceived dangers of that workplace that in the absence of free workers first slaves, then
prisoners sent initially to the galleys, were required to help at Almadén.959 In the 1560s the
Fuggers had requested additional manpower for the mercury mines at Almadén in order to meet
the increasing demand from the New World. Some 40 convicts destined to serve their sentence
as galley-slaves were now sent to work at Almadén, a punishment that must have been deemed
not much better than the death sentence of rowing on a Mediterranean galley. In 1593 a special
judge, Mateo Alemán, drew up a report for the Spanish King on the conditions under which
these slaves and other workers were exposed to the effects of mercury while working in the
mine managed by the Fuggers. Among his detailed conclusions:
‘it is harmful to the men to assist in the vats used to cook the ores from where mercury is
obtained and to rake the ashes because they get into the eyes, mouth and nose … from that the
men are azogado [stricken with mercurialism] and become dim-witted and lose their mind and
become gravely ill’.960
useful though older review on the reporting of occupatiotanl helath isses through history is by H.E. Sigerist, "The
Wesley M. Carpenter Lecture:" Historical Background of Industrial and Occupational Diseases"," Bulletin of the
New York Academy of Medicine 12, no. 11 (1936). Articles on specific cases are: K. V. Fox, "Pedro Muñiz, Dean
of Lima, and the Indian Labor Question (1603)," The Hispanic American Historical Review 42, no. 1 (1962).;
Z.Z. SIavec, "Occupational Medicine in Idria Mercury Mine in 18th century," Vesalius IV, no. 2 (1998).; C.
Serrano, "Minería salud en el Potosí colonial; Mining health in the colonial Potosi," Arch. boliv. hist. med 11, no.
1/2 (2005).; Alfred Bogomir Kobal and Darja Kobal Grum, "Scopoli's Work in the Field of Mercurialism in Light
of Today's Knowledge: Past and Present Perspectives " American Journal of Industrial Medicine 53, no. 5 (2010).
959
‘se azogaban y se les “dañaban las bocas”….medico … para curar los azogados … no hay en el mundo gente
… tan sujeta a enfermedades como los azogados’ Matilla Tascón, Minas de Almadén (to 1645), 79.
960
‘dañoso a la salud de los hombres es el asistir en los buitrones a el cocimiento de los metales de que se saca
el azoque y el cerner las cenizas porque se les entran por los ojos y boca y narizes … dello se azogan los hombres
y quedan tontos y fuera de juyzio y uienen a enfermar grauemente’. There are more quotes on the dangers to
566
Alemán would then emigrate and die in New Spain ca. 1615, though little is known of
his activity in the New World.961 He represents at least one direct channel for the transmission
of knowledge on the toxicity of mercury to the authorities and mining community in New Spain
during the period that saw the introduction of major quantities of mercury to refine silver. As
far as I can ascertain, no similar study was commissioned by the Crown on the effect of mercury
on the workers in the New World that refined silver by amalgamation.
The answer to the question that heads this section is therefore straightforward. The
historiographical evidence points to the fact that the impact on worker’s health of smelting with
lead or working with mercury were well known in the sixteenth century at the highest levels of
policy makers in Spain, as they were even better known to the English investors who flocked
to Mexico in the nineteenth century. There is no evidence that this knowledge at any point in
time influenced any major decision related to the refining of silver ores in the New Spain /
Mexico up to the end of the nineteenth century. Overall the production of silver remained
unfettered by strict controls on the use of lead and mercury. To judge the mentalité of the period
on occupational health, it should be borne in mind that as late as nineteenth century Germany,
one of the important reasons to protect workers from the toxic smoke of the smelting furnaces
was because it tended to kill the best and most skilled of the smelters.962
The collateral damage on the health of the smelters or amalgamation workers caused
by both lead and mercury was judged by the norms of the period to be an acceptable cost of
each process. The fact that even in the nineteenth century the English managers and investors
health from working with mercury from interviews held by Alemán with workers at Almadén in German Bleiberg,
El Informe Secreto de Mateo Aleman sobre el trabajo forzoso en la Mina de Almaden (London: Tamesis Books
Ltd, 1985), 81.
961
Alemán is better known as the author of a novel, Guzman de Alfarache, that together with the now more famous
Don Quixote by Cervantes became a widely quoted example of the Spanish picaresque style. Irving A. Leonard,
"Mateo Alemán in Mexico: A Document," Hispanic Review 17, no. 4 (1949).
962
Schlutter, De la fonte des mines, 2 2.
567
at Regla did not comment on this topic indicates that it never became a major issue of concern
within its workforce.963 In the case of mercury it has been argued that its transformation into
calomel attenuated to a high degree the impact on the human environment, though a loss of
15% as elemental mercury is still a major historical health concern to all the communities
exposed to its effects. What would have been the response of refiners to a much higher level
of toxicity?
On the 7th February 1561 a petition was forwarded to the Viceroy of New Spain, Luis
de Velasco, requesting a merced for a new amalgamation recipe that included the addition of
solimán, mercuric chloride.964 To prove that its use was not injurious to health, the promoter
of the method, Pedro Martín from Tasco, swilled his mouth with the water used to wash the
ore after treatment, though the only ones forced to actually swallow the liquid were a cockerel
from Castille and a horse, to no apparent ill effect. On the 10th March 1561 the authorities in
Seville were quickly appraised of the benefits of this new recipe and requests made for new
shipments of solimán at an attractive price. By July of that year the same authorities in New
Spain would start to point out the hazard posed by the use of solimán: ‘it is dangerous for the
blacks [slaves], we give notice to Your Majesty of this fact’ (24 July 1561). By the time a year
had gone by it was clear that ‘there are great dangers to using solimán, for which reason the
miners do not make use of it’ (2 April 1562).965 Such was the risk involved, and its lack of
963
In the case of amalgamation Sonneschmidt takes the trouble of addressing this issue in a separate section of
his work to conclude that in spite of negative reports in Europe on the effects of amalgamation on the workforce,
he had met in his visits to Reales de Minas amalgamation workers with up to 40 years working in amalgamation
haciendas who did not show any health problems as a result. Sonneschmidt and de Fagoaga, Tratado de la
amalgamación de Nueva España, 94-95.
964
Almadén converted part of its mercury production into solimán. Sánchez Gómez, Minería no férrica en el
Reino de Castilla. pp. 275-276 check. What a difference a state of oxidation makes in life. Solimán and calomel
are both chloride salts of mercury, the former a mercuric salt (Hg+2), the latter a mercurous salt (Hg+1). Solimán
is an extremely toxic mercury compound, while calomel is not.
965
‘es peligroso para los negros, damos a vuestra majestad noticia de ello’; ‘hay grandes peligros de beneficiarse
con soliman, de cuya causa los mineros no se aprovechan de ello’ as quoted in Castillo Martos, Bartolomé de
Medina, 210-13.
568
technical success, that it soon disappeared from the repertoire of refining methods in New
Spain.
The toxicity of mercuric chloride, which is soluble in water, is much higher than that
of mercurous chloride (calomel), which is not.966 Its effects on the human body are much more
immediate and evident than the longer term and more insidious effects of either mercury or
lead poisoning. Retention of urine, vomiting and bloody diarrhea leads to death unless
treated.967 Even if solimán had been an effective amalgamation reagent, the very high level of
toxicity would have precluded its widespread use in the New World. Even if the tolerance to
occupational risks remained high until the twentieth century, the natural reticence of humans
to work with evident poisons was the ultimate threshold. Had the short-term toxicity of mercury
been equivalent to that of solimán, amalgamation would not have been used to refine silver
ores.
6.3.2 Historicity and the reality of the refining process
One of the anomalies to the paradigm based on copious emissions of volatile mercury
to the air from the heating of the amalgams is the silence in the historiography on widespread
mercurialism, more so in historical accounts of large population centres such as Potosí in the
immediate vicinity of a major concentration of amalgamation units.968 More mercury was
966
Solubility in water is critical in aiding the diffusion of any chemical into the body via skin absorption, through
the digestive system or even via the respiratory tract, thus hastening its presence in the blood stream and tissues.
This in turn will aid in promoting any toxicity inherent to the chemical. For an example of pathways into the
human body see Jaroslava Svarc-Gajic, General Toxicology (New York: Nova Science Publishers, 2009), 17-47.
967
See for example Laszlo Magos and Thomas W. Clarkson, "Overview of the Clinical Toxicity of Mercury,"
Annals of Clinical Biochemistry 43(2006).
968
New Spain did not have a similar example, since its silver production via amalgamation was more widely
disseminated. Guanajuato was one of the few cases of major amalgamation haciendas clustered close to the town.
569
handled by workers at these haciendas than was used by hat makers in England, yet no Mad
Amalgamator appeared to displace the better known Mad Hatter of the literature.969
The silence on any sign of widespread mercurialism extends to all textual sources.
There is no mention in the selections of letters that have been published from Spanish
immigrants to New Spain to their families in Spain.970 The more militant and vocal workers of
the eighteenth century, when the yearly average of mercury consumption would have reached
its peak in New Spain, did not include mercurialism among their list of workplace
grievances.971 I have not come across any reports on mercurialism from the English expatriates
supervising the refining of silver in the haciendas of the Real del Monte Company.972 Nor have
I come across reports of complaints against toxic smoke from amalgamation haciendas. The
969
The only exception being the the strange story of Fulgencio Orozco, recounted at the end of Chapter 6 in
Timothy Brooks, Vermeer’s Hat (London: Profile, 2008), 181-84. Brooks however does not make an explicit
connection between mercury and the derangement of the Spaniard who worked in the amalgamation Ingenio. De
Gamboa makes one of the few extant comments on the poisonous nature of amalgamation haciendas in de
Gamboa, Comentarios Ordenanzas de Minas, 462.
970
Enrique Otte and Guadalupe Albi Romero, Cartas privadas de emigrantes a Indias, 1540-1616 (México: Fondo
de Cultura Económica, 1993).; Marta Fernández Alcaide, Cartas de particulares en Indias del siglo XVI : edición
y estudio discursivo (Madrid; Frankfurt am Main: Iberoamericana ; Vervuert, 2009).
971
Ladd, The Making of a Strike, 19 ff.
972
In 1825 a very graphic portrayal of the deemed toxicity of the amalgamation process was published in England:
‘the men employed barefooted [in mixing the amalgamation tortas] soon became salivated and paralytic, by the
absorption of the mercury, and ultimately died the most painful deaths’, from Rawson, "The Present Operations
and Future Prospects of the Mexican Mine Associations Analysed, by the Evidence of Official Documents,
English and Mexican, and the National Advantages Expected from Joint Stock Companies, Considered, in a Letter
to the Right Hon. George Canning," 19. The text is very relevant, not for the improbable scenario it depicts, but
for being an early example of a strand of thought that still survives in the modern historiography, that has judged
the indigenous workforce as being incapable of running an efficient operation with regards to the recovery of
mercury and of a supine docility in the face of such claimed immediate mortal dangers. Indigenous workers
suffered great hardships in the mining and refining industries of New Spain and Mexico in order to feed their
families, but there is no evidence they were a completely inefficient and passive workforce. Not even an industry
based on a slave workforce could bear the economic impact of workers sickening and dropping like flies soon
after trampling amalgamation tortas. The gist of the first part of the letter is that English know-how will allow the
wealth of the silver mines in Mexico to be properly extracted. It was written before operations at Regla under its
new English investors were underway. There is no evidence that Sir William Rawson ever travelled to Mexico,
his source is not identified, and by providing a lurid caricature of events he diverts attention from the real problems
of historical refining of silver ores. His letter seems to be a later version of the type of negative European reports
mentioned by Sonneschmidt, see footnote 963.
570
only reports of mercurialism come from accidents during the heating stage of the capellinas,
as for example:
‘it is not too many years since capellinas have begun to be used, previously having used in
general clay pots. These broke frequently with greater risk to those who approached to quench
the fire. I have found various individuals that in such circumstances were poisoned by mercury,
and they fell to the floor senseless, and in spite of this they recovered nearly completely, only
left with a minor tremor of limbs that is triggered after much exercise.’973
Was historicity at work, with the Spanish and English sources hiding the truth on the
ravages of mercurialism behind the nested stories within a story so aptly described by Atwood
at the beginning of this chapter?974 The initial answer is that there was no mention of
widespread mercurialism either in Peru, New Spain or Mexico simply because the aggregate
amounts of volatile mercury issued to the air over time were the results of isolated accidents,
not part of standard amalgamation practice, and were confined to the immediate surroundings
of the heating area of the capellinas.975 When volatile mercury did escape, there is no doubt its
effects on the workers was immediate and toxic, but it was never a case of major losses of
volatile mercury affecting both the hacienda and the surrounding communities.976 It was solid,
973
‘no hace muchos años que las capellinas están en uso, habiéndose servido antes casi generalmente de ollas
de barro. Estas se reventaban con freceuencia en el mayor riesgo a los que acudieron para apagar la lumbre. He
encontrado a varios sujectos que en tales circunstancias se han azogado, y cayeron en el suelo privados de
sentidos, y sin embargo de esto se restablecieron casi enteramente, restándoles solo un leve temblor de miembros
que les acomete después de mucho ejercicio’ in Sonneschmidt and de Fagoaga, Tratado de la amalgamación de
Nueva España, 51. An exaggerated version of the immediate effect of mercury fumes upon life is provided in the
accounts of a visit to the refining hacienda of Fresnillo in Albert M Gilliam, Travels Over the Table Lands and
Cordilleras of Mexico, During the Years 1843 and 44 (Philadelphia: John W Moore, 1846), 260. As argued in
Chapter Three, metal capellinas were widely used in the eighteenth century, and Barba suggested strongly the use
of metal caperuzas as of the 1600s, to avoid the operational problems mentioned by Sonneschmidt.
974
Margaret Atwood, MaddAddam : a Novel (New York: Doubleday, 2013), 56.
975
Robins provides historical texts on the dangers of processing cinnabar at Huancavelica and from the mills
crushing ores, but no historical quotes on wide-spread mercurialism either in the refining workforce or among the
inhabitants of Potosí. Robins, Mercury, Mining and Empire.
976
This conclusion only applies to amalgamation refining haciendas. The issue of mercurialism was very much
present in the mercury mines of Huancavelica and Almadén. For a historical analysis of the hazards of mercury
at Huancavelica see Brown, "Workers Health and Colonial Mercury Mining at Huancavelica, Peru."
571
water-insoluble calomel and liquid mercury that were handled and spread in major amounts
during the amalgamation process.
The only problem with this answer is that while mercury is the usual suspect in any
environmental issue related to historical silver refining, so too should be lead. This was a heavy
metal that together with lead compounds was issued to the atmosphere in major amounts within
and around each smelting hacienda, under workplace concentrations that exceeded any modern
guideline. The greatest harm would fall on the smelting teams closest to the furnaces and then
on the rest of the workforce within the smelting hacienda, but as reported in Chapter 2 even
the communities around smelters were aware of the toxic nature of the smoke from the smelting
furnaces. The visual evidence from historical photographs and drawings leave no doubt as to
the degree of air pollution from these smelting haciendas. Yet again specific references in the
historiography to lead poisoning as a consequence of the smelting of silver ores are few and
far between.977
The general silence on occupational hazards began at home, with the way Spain ignored
its own workers at Almadén: ‘During the firing of the ores, the vapours from the furnaces
...extend over the town [of Almadén] so that its inhabitants are under the pernicious effect of a
mercurial atmosphere’.978 It would be hard to argue that the Spanish Crown would take greater
care of its new subjects in the New World than it took of its old ones in Spain. The prejudiced
view on the inhabitants of the New World and their role in relation to the wealth of silver and
gold is well described by the contents of the Parecer de Yucay. This was an anonymous
977
One of the few exceptions is the following observation ‘poisonous are the Smelters and amalgamation
haciendas’ - ‘venenosas las Fundiciones y las Azoguerias’, in de Gamboa, Comentarios Ordenanzas de Minas,
462.
978
‘en la época de fundición, los vapores de los hornos ... se extienden por el pueblo [de Almadén] con lo que
todos sus moradores quedan sometidos a la perniciosa influencia de una atmosfera mercurial’ José María Pontes
y Fernández, Historia de la antigua ciudad de Sisapón, hoy Almadén del Azogue (Madrid: Enrique Rojas, 1900),
12.
572
pamphlet against the arguments put forward by La Casas in defence of the inhabitants of the
New World and published in 1571, the same year the ViceRoy Toledo implemented the mita
to provide indigenous labour by forced mass migrations to the silver industry of Peru. Its
authorship has been attributed to a Dominican author, fray Garcia de Toledo. It also contains
an appendix on mining that has been fiercely criticized by the Spanish historian Isacio Perez
Fernandez, O.P. (Dominican Order), who has termed it ‘cheap theology … an irritating literary
sanctimoniousness, an insipid, boring and cheap theological focus’, assumed to be authored by
Polo de Ondegardo.979 The appendix merits quoting at length, since even as a caricature of a
theological tract it still serves as a guide to the manner in which some in Spain regarded their
right to the treasures of the New World and their views on the indigenous population. It begins
by stating that the Indies were the prize given by God to the Spanish Crown for the reconquest
of Spain from the enemies of Christianity. The author then goes on to criticize the stand taken
by Bartolomé de Las Casa in favour of the indigenous inhabitants of the New World, by way
of the following argument:
‘what was meant when God placed these indians with such miserable souls and so bereft of
God, brutes so lacking in skills … in kingdoms … so full … of gold and silver … what does
this mean except that God found himself, with these miserable people and with us, like a father
that has two daughters, one very white, very discreet and graceful … the other very ugly,
bleary-eyed, stupid and bestial? If he is to marry … the ugly, clumsy, stupid, wretched one [he
will have to] provide a large dowry: many jewels, rich clothes … it is with all this that God
comes to aid’.980
979
‘teología barata … una gazmoñería literaria irritante, un enfoque teológico insulso, ñoño y barato’. Isacio
Pérez Fernández, El anónimo de Yucay frente a Bartolomé de Las Casas : estudio y edición crítica del Parecer
de Yucay, anónimo (Valle de Yucay, 16 de marzo de 1571) (Cuzco, Perú: Centro de Estudios Regionales Andinos
"Bartolomé de Las Casas," 1995), 10, 103.
980
‘que quiere decir el haber puesto Dios a estos indios tan miserables en las almas y tan desamparados de Dios,
tan inhábiles y bestias .. en unos reinos … tan llenas .. de oro y plata .. que significa esto sino que se hubo Dios
con estos gentiles miserables y con nosotros, como sea un padre que tiene dos hijas, una muy blanca, muy discreta
y llena de gracias .. la otra, muy fea, laganosa, tonta y bestial? Si ha de casar ... a la fea, torpe, necia, desgraciada
[habrá que] darle gran dote: muchas joyas, ropas ricas … con todo esto Dios y ayuda’ excerpt from the Appendix
titled ‘On the working of the mines’ - ‘Sobre el beneficio de las minas’ in the Parecer de Yucay, as reproduced in
ibid., 157-62.
573
The analogy becomes clearer when the author argues that Europeans were so graceful
it was easy to be saved (married, without the need of an expensive dowry) by the Christian
religion, while for the brutish and ugly inhabitants of the New World their only path to salvation
lay in the size of their dowry as given by God, in the wealth of the mines that attracted both
soldiers and the spread of the Catholic faith, since ‘in the lands where there is no dowry of gold
and silver, there is no soldier or captain that wants to go, not even an evangelizing
missionary’.981 To oppose mining as Las Casas was doing was to take sides with the Devil, the
same Satan who convinced the locals to hide the location of mines from the Spaniards. In the
end, ‘the mines, from a moral point of view, are as necessary as the King himself, since without
them it will not be possible to conserve … the Gospel. Therefore, saintly and good they are,
and blind are the men who deny this, a sign of evil from the devil and his work’.982 Under this
line of reasoning any negative consequences from the use of mercury or lead on the health of
the local communities was no greater than the risks already borne in Europe, and were all for
the greater good carried out by the Crown.983
Rosen in another context has argued that while mining was the activity of slaves and
criminals it impeded any interest in documenting systematically their diseases, and that only
the appearance of free men as miners, with an intrinsic value now placed on their welfare and
the need to safeguard their increased technical knowledge, transformed them into capital assets
that justified the creation of a corpus of detailed medical literature related to their occupational
981
‘a tierra donde no hay este dote de oro y plata, ni hay soldado ni capitan que quiera ir, ni aun Ministro del
Evangelio’, in ibid., 160.
982
‘las minas, moralmente, tan necesarias son como es haber rey, pues sin ellas no se conservara … el Evangelio.
Luego, santas y buenas son, y gran ceguedad en los hombres negarlo, y malicia en el demonio, y obra suya’.Ibid.,
161.
983
This disdain of the local workforce is not a phenomenon restrained to the Spanish Crown, authorities and
miners of the sixteenth century. The silence of the English managers, investors and technical staff on mercurialism
can be explained by calomel, but the absence of measures to control the lead in the smoke from the smelting
furnaces signals that the workforce in the Mexico of the nineteenth century may not have yet been considered an
asset in the eyes of the English investors or their Mexican successors.
574
diseases.984 To complement Rosen it can also be argued that in the context of the perils of
mining and in the light of the immediate harm caused by silicosis due to the fine milling of ores
or the incidence of hernias on the workforce, or mortal accidents within the mines, long-term
lead poisoning or mercurialism was the least of the worries in the minds of the miners and
refiners.985 It is thus much more common to find references to the hazards of mining or of
milling than to the perceived hazards of lead and mercury within the labour-force. This normal
response to the clearest and most present dangers lowered the visibility of the potent but longer
term threat of lead emissions in those areas where smelting was carried out, though it did not
quite erase it from the historiography.
Was there a conscious decision to downplay the occupational, safety and health aspects
of the refining haciendas up to the end of the nineteenth century? Smelting would have created
more problems among the workers and surrounding communities than amalgamation, since
mercury in calomel would have posed a negligible short-term health hazard compared to lead
fume. The fact it did not receive any major attention in the historical texts points to a more
complex answer to this question. Mining and refining drove the economy of New Spain and
Mexico during the period of interest. The overriding interest of the Crown to maximize the
extraction of silver, the merging of refining and farming interests in the hands of strong owners
of capital within New Spain, and finally the entry of overseas investment allied to newly
emerging capital and industrial Mexican groups, created overall a very strong reason to accept
and downplay the negative effects of refining within official texts, accepting them as a
necessary consequence of capital and industrial growth. In private letters or journals written by
984
Rosen, History of Miners’ Diseases, 8-38.
985
During a panel discussion on modern artisanal gold mining at the 11 th International conference on Mercury as
a Global Pollutant (Edinburgh, 2013), members of the audience pointed out that the concerns of researchers did
not necessarily match the concerns of the artisanal refiners on the subject of workplace safety, for whom the
dangers of inhaling mercury fumes were the least of their daily problems.
575
Europeans I would argue that the paradox of Father Brown’s postman was also at work, in the
sense that the social biases of observers can render certain groups of people mentally invisible
to them.986 The indigenous population of refining workers was for the most part invisible to
most European observers of the nineteenth century, where they barely figure in the
commentaries. Whatever their illnesses may have been, they were not part of the mental
landscape covered by these texts. I have already cited the observation on the toxicity of copper
sulphate on patio horses that avoids any mention of its toxicity on the local workers. 987 The
environmental history of silver refining through the eyes and words of the indigenous workers
of the New World remains to be written.
6.3.3 The option of smelting all the ores in New Spain
In sharp contrast with later events in Peru, the statistics on silver production show no
sign of a silver crisis in New Spain at the time amalgamation was being implemented for the
first time. The first dip in production is only observed in the decade of the 1580s, well after
amalgamation was first implemented.988 Before mercury arrived on the scene total production
levels of silver in New Spain in the decade 1541 to 1550 corresponded to a yearly average
around 27 metric tons of silver. This was a level two to three times the yearly average for the
same period of all the main German/Central European mines, the main source of silver up to
that point.989 In the longer term smelting would produce just under 40% of the total silver from
986
From G.K. Chesterton’s short story “The Invisible Man” in G.K. Chesterton, Father Brown, Selected Stories
(London: Oxford University Press, 1961), 74-94.
987
See footnote 427.
988
Lacueva has argued that the absence of a production crisis of silver in New Spain at the time amalgamation
was implemented, together with the level of new capital investment that was required to switch from smelting to
amalgamation, are two very strong arguments against the accepted narrative that places amalgamation as a
necessary and natural next stage for the evolution of silver production in New Spain in the second half of the
sixteenth century. He suggests amalgamation represented an opportunity for the owners of capital to subordinate
the needs of the cash-strapped class of small refiners and miners. Lacueva Muñoz, La plata del Rey, 148-167.
989
Data for New Spain from TePaske and Brown, Gold and Silver, 113.; for Europe from John H. Munro, "The
Monetary Origins of the'price Revolution': South German Silver Mining, Merchant Banking, and Venetian
Commerce, 1470-1540,"(Department of Economics, University of Toronto, 2003), 43. Nef cites a maximum
576
New Spain, even after all the assistance provided to amalgamation through the periodic
decreases in the pricing of mercury over the period. Yet Juan Suarez de Peralta, one of the
earliest voices on the history of New Spain, was one of the first of many in the historiography
to claim that without mercury there would have been no silver from the New World.990
Without recourse to Almadén and Huancavelica it is probable that Spain would have
proceeded to smelt the remaining 60% of silver in New Spain that was ultimately refined by
amalgamation. Smelting was the traditional European refining method for silver ores in the
sixteenth century. In contrast to the case of England, where by the year 1400 already only 17%
of England and Wales’ usable land was covered by forests, Spain found in the New World a
far vaster virgin resource of wood for making charcoal.991 Even nearly five hundred years later,
the modern states of Mexico, Bolivia and Peru figure among the world’s top ten countries with
the largest areas of primary forest cover in the world.992 Even so, dressing of the ore would
have been a necessary pre-requisite to smelting, otherwise the need to heat the totality of an
ore containing on average 0.2% silver would have required an estimated additional 105 million
ha of woodland for charcoal, more than twice the area of woodland existing in Mexico in the
year 2010.993 Dressing the ore to a final silver content greater than 0.6% would have brought
down the additional requirement of woodland to at least one third of the previous estimate, and
production of silver from all central European mines in the decade 1526 to 1535 of nearly 3 million ounces
(approximately 94 tons). Nef, Conquest of the Material World, 42.
990
As quoted in Bargalló, Minería y metalurgia colonial, 240.
991
Kaplan, Krumhardt, and Zimmermann, "The prehistoric and preindustrial deforestation of Europe." p. 3023.
992
Global Forest Resources Assessment 2010, 13.
993
A split of 60:40 in silver produced by amalgamation and smelting, based on an average silver content of 0.2%
for amalgamation and 2% for smelting, corresponds to a 15 to 1 ratio in the weight of ore processed by each route.
Thus if 7 million ha were required to heat all the ore used for smelting in New Spain, 15 times that amount would
be needed to heat the ore that would have otherwise been amalgamated. The data on the forest cover in Mexico
in 2010 (65 million ha) are from ibid.
577
at the same time offered a comfortable degree of profit to the smelters, net of dressing costs
(see Figure 5.23).994
Lead would not have been a material bottleneck either had Spain chosen to smelt all
the ore. Up to an additional 300,000 t over 250 years would have had to be sourced locally or
imported from Europe (and serving as needed ballast to the outgoing convoys from Seville) in
order to make up for losses during the smelting of the additional ore. Mexico produced over 2
million t of lead just between 1931 and 1941, so if the Spanish Crown had promoted the mining
and production of lead the local deposits would have been sufficient to sustain the requirements
of smelting.995
When the initial wave of complaint rose from the ranks of amateur refiners in New
Spain that they were running into problems refining silver, the Crown had the option of
improving the smelting skills of this population, promoting the search for lead within New
Spain or even as market leader adjusting the price of silver to compensate the need for charcoal
or imported lead flux. Smelting was the universal refining process, not limited like
amalgamation to a certain types of ore. The rapid response of the Crown to the potential use of
mercury to refine silver ores, commented upon in Chapter 3, raises the unsurprising issue that
the ownership of the Almadén mine played a critical role in the refining policy, explicit or tacit,
made by the Crown. There is no doubt that amalgamation was a technique better suited to the
majority of ignorant refiners set loose in New Spain after its conquest. However, as the history
of San Luis Potosí shows, from the late sixteenth century as refining became more technical,
994
I am not adopting a defence of a policy that would have decimated the forests of Mexico along European lines,
but simply pointing out that from a material point of view a policy that was acceptable at the time could have been
adopted, even more so if forest regeneration cycles operated over a 250 year span, as would be expected. The
future use of coal, local or imported, to substitute charcoal in smelters cannot be ruled out in this counterfactual
scenario.
995
González Reyna, Minería y Riqueza de México, 3.
578
smelting was the natural choice when refiners were faced with silver-rich lead ores. With
sufficient lead and after dressing ores, the experience at San Luis Potosí could have been the
model for the expansion of silver production in New Spain as of the seventeenth century, and
for 100 years smelting refined silver ores on par with amalgamation in New Spain.
The use of smelting would however have posed an opportunity cost to the Crown. If
mercury was used, it opened up a revenue stream to the Treasury, in parallel to the revenues
from the royalty and other taxes on silver, as discussed in Chapter 5. This opportunity cost
could have reached some 10% of the total revenues to the Crown in the early seventeenth
century, a major sacrifice in earnings. Thus to the operational advantages of patio
amalgamation must be added the fiscal advantage over smelting it represented to the Royal
Treasury, made even more attractive during the first century of amalgamation in New Spain by
the role of the Fuggers.
I thus approach the end of my arguments by adding to the technical factors the human
choices, from the policy of the Spanish Crown to the historiography of the modern era, that
have played a decisive role in determining and then reporting the environmental history of
silver refining in New Spain and Mexico. Amalgamation was not the only technical key
available to unlock the silver from its ores in the New World, it was not the sole option to
process ores claimed to be too low in silver content for smelting. In the case of New Spain
smelting remained until the end of the colonial period a very viable and competing alternative,
as shown by its contribution of under 40% of the total silver produced even after efforts by the
Crown to sustain amalgamation. By choosing for fiscal reasons to favour the path of
amalgamation over smelting, a decision evident from the actions of the Crown and its colonial
authorities, the population and the environment of New Spain was spared a worst-case scenario
of even more lead issued to the air around each of the smelting haciendas and a wide-spread
consumption of woodland. Instead they would be subject to millions of tons of silt suspended
579
in water and settling along river beds, entombing mercury transformed into calomel, together
with liquid mercury entrained with dissolved salt and copper compounds. Within the haciendas
the environmental legacy would be residues of fugitive lead and soil soaked in mercury.
Finally, but significantly in terms of quantity and impact, air emissions of mercury would be
two orders of magnitude lower than lead and its compounds exiting the chimney stacks, a
devil’s alternative for the indigenous population

580
Epilogue
During 250 years Spain would control both the world’s largest reserves of silver and of
mercury, a historically unique geological and geopolitical triangulation that would make it the
master of the silver market for nearly three centuries. By conquering the western chain of
mountains from the Andes to the Cordillera of North America it conjoined its own vast reserves
of mercury at Almadén with the major deposits of silver that had been generated by a process
of subduction of tectonic plates all along the eastern Pacific Rim. Nowhere else in the world is
subduction as active, nor such a chain of silver deposits to be found, much younger and of
different geological origin than their more modest counterparts in MittelEuropa. The only
common denominator was the presence on both sides of the Atlantic of lead ores rich in silver.
In general the ore deposits of Europe were mined for copper or lead rather than silver, while
silver was the only economic reason to refine the ores of the New World. The absence of major
pre-Conquest extractive industries kept intact the composition of their weathered surface,
native silver and silver chlorides.
Due to the scavenging instincts of the initial swarm of Spanish colonists, avid for
wealth and lacking skills, they left behind generous mounds of discarded weathered ores, all
the mineral that did not immediately promise silver to their untrained eye or yield it to their
primitive smelting efforts. To the authorities they complained of decreasing silver content,
when in reality what was changing was the nature of the mineral, from weathered silver ore to
deeper and more intractable silver sulphides. The initial technical challenge was solved thanks
to imported German know-how in smelting. The environment of New Spain would feel the
first wave of volatile lead products from the smoke of smelting furnaces and of charcoal burners
smoldering their way through woodlands.

581
As the mountains of discarded ore grew, so did the frustration of the first generation of
self-taught refiners. In New Spain amalgamation did not arrive on the back of a sudden decrease
in silver production. It arrived from Europe as a process that was known to work with gold,
and was a much simpler method that suited untrained refiners better than smelting. Spain saw
a quadruple opportunity open up: more silver could be produced by its untrained colonists
foiled by smelting, it could gain much more revenues from its mercury mine at Almadén, it
could use the Fuggers to provide the mercury on credit as well as collateral for future loans,
and by controlling the mercury as a State monopoly it could gain a measure of control over the
contraband of silver. The hazard posed by mercury was well known, but in practice was not
seen to be worse than the toxic fumes from smelting, and certainly no greater than the burden
imposed upon its own people at Almadén.
Mercury, aided by its aura as the alchemical precursor to silver and gold, was thus
applied to the mounds of cheap discarded ores and to existing superficial deposits. These
responded well to the primitive amalgamation recipe that was known to work with gold. Had
the first generation of refiners been thorough in their triage and smelting, only lead-rich slags
would have been left behind. In such a case the primitive amalgamation recipe would not have
worked either on these or on the deeper silver sulphides or negrillos. As the mounds of
discarded but easy to amalgamate ores were run down, so did silver production suffer. Then
events in the altiplano of the Andes would radically change the industrial potential of the basic
gold amalgamation recipe. In a short burst of impressive technical creativity within the Spanish
refining haciendas of the Andean altiplano, the amalgamation recipe was converted into a
powerful chemical tool that was able to reduce the silver sulphide present in negrillos into
amalgamated metallic silver. Though an earlier incarnation had been applied previously in the
Schio mines by Venetians in the early sixteenth century, it was recreated independently through
the stubborn efforts of the self-made refiners around Potosí.

582
As amalgamation gained momentum, mercury was consumed in quantities never
observed before in any human industry. The workers of the amalgamation haciendas and the
population of New Spain were spared the ravages of mercurialism on a major scale both by the
safeguards adopted during the heating cycle of the amalgam and by the chemistry of the
amalgamation process that consumed mercury by converting it to solid, insoluble calomel.
Nevertheless tons of liquid mercury were squeezed through the fingers of workers onto the
slurries of ore, and liquid mercury was still washed away in water or lost by seepage to the soil.
Very little mercury escaped to the air during the regular heating cycle of the amalgams. Streams
and water basins around the clusters of amalgamation haciendas became their waste disposal
units. Salt seeped into the ground of the patios and was washed away together with copper
compounds and the millions of tons of fine mineral silt that were useless to the refiner once all
possible silver was extracted. All these materials would contaminate the water downstream
from each refining unit, and compromised its use for consumption and irrigation.
Patio amalgamation represented a technology best adapted to the medium where it was
implemented, competitive with the more traditional route of smelting. The longevity of its
recipe was never a sign of backwardness or technical stagnation, nor was the chemical
immutability of the smelting process over thousands of years. Amalgamation embodied all the
elements of a modern industrial process: planning of inventories, carefully concatenated stages
of physical treatment and chemical reactions, avoidance of operational bottlenecks to achieve
a smooth production output. The patio reactor that evolved in New Spain was the most efficient
answer to the amounts and nature of the ore that had to be treated, and to the materials at hand.
Nevertheless, the environmental history of silver refining in New Spain was determined
as much by smelting as by amalgamation. Smelting contributed with just under 40% of all the
silver produced during the colonial period, and possibly one quarter to one fifth during the most
of the nineteenth century. Lead and its compounds represented the main source of heavy metals
583
issued to the air from the historical silver refining in New Spain, on average two orders of
magnitude greater than the total air emissions of mercury. The historical impact of emissions
of lead and lead compounds from the smelting and refining furnaces was the aggregate of
multiple but singular depositional footprints determined by the local wind rose, furnace
efficiency, skill of the smelter, lead content of each ore, and by the architectural trace of the
hacienda and the location of its mounds of grasas. Some mining regions in New Spain would
never be exposed to much lead, others would have known no other airborne heavy metal, but
those who did live in the vicinity of smelters had no doubts on the toxicity of its smoke on
animals and humans. The other major environmental impact of smelting was on woodlands,
consuming them at a rate over 50 times higher than amalgamation would. This depredation of
woodlands was only attenuated by the increased efficiency of blast furnaces in the nineteenth
century, which decreased by an order of magnitude the rate of consumption of charcoal per kg
of silver smelted.
The environmental cost from silver refining as imposed on the indigenous communities
and the new settlers was never addressed in a significant manner in the texts of the period. The
silence is a reflection of many realities: the greater, more immediate dangers of mining silver
ore or processing cinnabar; the acceptance of occupational health risks from a workforce whose
pressing issue was to earn a living for their families within a context of limited labour options;
the cloak of social invisibility that covered most of the indigenous population in the eyes of
others. The silence encompassed both the Spanish and English contingents who came to extract
silver from New Spain and then Mexico, as well as the new local owners and operators of the
later republican period.
The silver of New Spain and Mexico could have been extracted only with smelting, had
this been necessary. Enough lead and woodland existed in its vast territory to have covered the
needs of this refining process. Without recourse to mercury, refiners would have sought a
584
greater efficiency from the furnaces, recovered lead from the fumes, spent more manpower in
dressing the ores and would have developed a secondary market for its smelted non-precious
metals, much as republican Mexico finally did as of the late nineteenth century. Had smelting
prevailed, the total environmental impact of its lead emissions and destruction of woodland
would have been more severe than from the historical mix of amalgamation and smelting.
Mercury as the lesser evil: this was the ultimate paradox of the environmental history of silver
refining in New Spain and Mexico.

585
APPENDICES
A. Silver production in smelting haciendas
Table A-I. Account book prepared by Lopez de la Madriz, Valle de Pozos, AHSLP, Fondo
Alcaldía Mayor 1650.3, expediente 8.
year month marks pesos silver pesos gold kg silver

1660 may 307 2,540 1,205
july 648 5,354 2,334 149
august 220
september 350 2,895 1,100
october 250 2,134 1,058
november 134 1,106 654
december 308 2,446 1,169
total 2,217 16,475 7,520 510
1661 january 455 3,757 1,685
february 400 3,304 1,532
march 159 1,319 640
april 271 2,128 857
may 315 2,604 852
june 625 5,082 1,454 144
july
august 439 4,233 1,644
september 295 2,454 808
october 373 3,080 403
november 359 2,962 610
december 152 1,258 247
total 3,843 32,181 10,732 884
586
Table A-II. Weekly accounts of the Hacienda de Nuestra Señora de los Dolores,
corresponding to the year 1773, signed by Lorenzo Mata. AHSLP, Fondo Alcaldía Mayor
1773.2.
silver per
cargas marks % silver silver per
dates week,
ore silver in ore week, kg
marks
12/10 to 17/10 29 4 0.07% 4.0 0.9
15/8 to 22/8 160 10.0 0.03% 10.0 2.3
30/5 to 6/6 40 8.3 0.10% 8.3 1.9
23/5 to 30/5 76 9.4 0.06% 9.4 2.2
16/5 to 23/5 60 8.6 0.07% 8.6 2.0
9/5 to 16/5 59 11.0 0.09% 11.0 2.5
3/5 to 9/5 60 8.6 0.07% 8.6 2.0
28/3 to 4/4 64 7.9 0.06% 7.9 1.8
1773 21/3 to 28/3 96 8.3 0.04% 8.3 1.9
14/3 to 21/3 11.5 11.5 2.6
8/3 to 14/3 160 13.1 0.04% 13.1 3.0
23/2 to 28/2 96 12.1 0.06% 12.1 2.8
14/2 to 21/2 270 14.8 0.03% 14.8 3.4
31/1 to 7/2 68 8.0 0.06% 8.0 1.8
24/1 to 31/1 68 12.5 0.09% 12.5 2.9
17/1 to 24/1 76 11.6 0.08% 11.6 2.7
10/1 to 17/1 165 10.5 0.03% 10.5 2.4
1772 27/12 to 3/1 14.5 14.5 3.3
average 0.06% 10.3 2.4
note: the % of silver in the ore is calculated on the assumption all the silver refined
in the week has been extracted from the total cargas of ore reported
587
B. The accounting books of Regla
The account books that were consulted in the Archivo Histórico de la Compañia Real
del Monte y Pachuca (AHCRMyP) as the source of the economic production data correspond
to the Fondo Siglo XIX. They comprise the following sections, series and sub-series:
1. Sección: Explotación y Beneficio, Serie: Informes de Haciendas de Beneficio,
Subserie: Informes Mensuales Hacienda de Regla Vol 225, Exp. 3: 29 Jun 1872 – 27 Oct 1888
This has been the main source of the monthly data on production and costs for Regla. I
have named this source as the Contabilidad Mensual de la Hacienda de Regla, June 1872 to
August 1888 or as Contabilidad Mensual for short. The months correspond to either four or
five week periods, and the accounts are dated according to the final day of each period. The
data only correspond to refining operations at Regla.
2. Sección: Negociaciones, Serie: Haciendas de Beneficio, Subserie: Hacienda de Regla
I, Vol 22: 1875-1878.
I have referred to this source as the Extracto de Memorias, 1875 – 1878. Each Memoria
is numbered and corresponds to a weekly account of certain expenditures (gastos) at Regla.
3. Sección: Contabilidad de la Dirección, Serie: Producción y Gastos, Subserie:
Estados Comparativos, 1853-1855, 1859-1865, 1869-1873.
This source is referred to as Estados Comparativos, a monthly and yearly comparative
summary of selected expenditures and costs per montón or carga of refined ore, for all the
refining haciendas active in any given period.
There is minimal overlap between these three sources for the period between 1853 and
1888, though the monthly and weekly accounts for Regla do coincide for at least a period of
588
three years (1875 to 1878). The comparative accounts for all the haciendas only coincide with
the monthly accounts for Regla in the year 1873. Even these limited overlaps have been very
useful in establishing what the accountants chose to include under any of the main headings.
By cross-checking the data from overlapping primary sources I have been able to arrive at the
following guides.
Contabilidad Mensual de la Hacienda de Regla, June 1872 to August 1888.
The Contabilidad Mensual de la Hacienda de Regla, June 1872 to August 1888
(referred to as Contabilidad Mensual) is the only source of accounting data used in this thesis
to have signatures that officially (and thus legally) validate each of the monthly accounts
presented. The office held by the signatories at times can be identified, such as the Managers
of Regla (Mr. Rule in 1872 and 1873) or the Administrator (Mr. Cuatáparo, from November
1876 to June 1877). From January 1881 until August 1888 most of the monthly accounts show
two signatures, that of Mr. Torres (who signs all the monthly accounts from March 1878 to
August 1888) and a higher ranking official of the Company, Mr. Landaro, who signals his
hierarchy by attaching a Visto Bueno (a sign of approval) to his signature. As in Chapter Four
I have decided to omit the data from January 1874 to March 1875 since this is a non-
representative period for Regla where an irregular refining of grasas (slag and tailings) and
limited amalgamation took place. Each monthly account sheet contains the following
information on the costs incurred:

589
1. A report on the monthly consumption by weight and total cost (in pesos) incurred of
the following major consumables, under the heading Almacén (virtual central warehouse at
Regla):996
Reagents: salt (sal), mercury (azogue), copper sulphate (sulfate de cobre), litharge
(greta)997
Fuel: charcoal (carbon); firewood is only reported in the period 1872 to 1873, and is
incomplete (see Chapter Four). A more complete and detailed breakdown of both charcoal and
firewood is provided both in the Memorias and Estados Comparativos.
Animal fodder: barley (cebada), straw (paja), corn (maiz)998
Animals: number in stock, losses by death
2. The monthly production costs (Costo de Beneficio) are reported within a separate
boxed-in area of each monthly account sheet. They are presented under some fifteen different
headings, some of which change during the 1872 to 1888 period. However it is fairly
straightforward to group these costs under the following sub-sets:
Labour costs: all the costs grouped under the following headings correspond only to
labour: stamp mills (Molienda de morteros), ore breakers (Almadaneta), Chilean mill
(Arrastre), amalgamation cake workers (Torteros), general labour (Peones), carriers
(Cargadores), fine ground ore workers (Lamadores), recovery of amalgam from the washings
996
I refer to it as a virtual warehouse since horses, mercury and litharge would not have been stored in a physical
central warehouse to be supplied as required, but the accountants kept track on paper of the different supplies and
purchases required by the operational needs of the hacienda.
997
Copper sulphate is a different material from magistral , the copper containing ore that needed roasting prior to
use.
998
‘straw and corn, the consumption of which is very important to sustain an also large number of mules and
horses’ - ‘la paille et le mais, dont la consommation est fort importante pour l’entretien d’un aussi grand nombre
de de mulets et des chevaux’ Duport, Métaux précieux au Mexique, 231. I am citing Duport since the corn could
also have been destined for human consumption.
590
of ore residues in a planilla (Planillones), horseshoe fitters (Herradores), repairs
(Composturas), bricklayers (Albañiles), sawyers (Aserradores), carpenters (Carpinteros),
roofers (Techadores), smiths (Herreros), general tasks (Faenas), final separation of silver from
amalgam, cupellation and casting of silver bars (Capellinas, afinación y fundición de barras),
amalgamation costs (Amalgamación en patio), Stables (Caballerizas) and other office and
general staff payroll costs (Gastos Generales). Not all these headings appear together in any
given year, since changes in accounting practice took place during this period, but they all refer
exclusively to labour costs. The identification of the nature of these headings was made
possible by crosschecking the data in the monthly and weekly accounts for the four week period
ending the 29th May 1877 (see below).
During the years (mid 1875 to early 1886) that smelting was also carried out within this
period, the heading named ‘Smelting’ (Fundición) corresponds to the labour costs of the
smelting process (see below).
Mercury, Salt, Copper Sulphate, Litharge and Charcoal : I use the accounting data
of the Almacén (virtual central warehouse) section to track the monthly cost of the main
consumables for amalgamation (mercury, salt and copper sulphate) and for smelting (litharge,
charcoal) and also to calculate the variations in their unit cost over time. Unfortunately these
monthly accounts do not provide data on the consumption of fuel for amalgamation, and this
information must be derived from the other sources (see below).
Other costs : this heading covers all the remaining production costs registered in the
accounts at Regla. For the years in which both amalgamation and smelting are being carried
out I take as my starting point the total cost of smelting that is reported as a separate entry by
the accountants of Regla. To arrive at ‘other costs’ for amalgamation I then subtract the total
smelting cost from the sum of all production costs, which gives me a total monthly production
591
cost for amalgamation only. I then subtract the total cost of labour and the costs of the three
main consumables to arrive at ‘other costs’ for amalgamation. As shown in more detail below
these ‘other costs’ include all other consumables (from the Almacén, or the sundry items from
nails to lard listed as Otros Efectos, or even supplied by other haciendas of the group) plus
minor non-operational costs that range from weekly masses to covering the costs to receive
visitors at Regla (some of which are grouped under the dire heading of Gastos Muertos, Dead
Costs).999 I do not subtract from the total production costs contingent monthly additional
revenues such as the rent of a store or the sale of poor lead. The aim of my calculations is not
the final profitability of the company but the quantitative breakdown of production costs in
each process.
In the case of smelting, I use the accountants’ figure on total monthly smelting costs,
subtract the costs for litharge, charcoal and labour (reported as fundición), and the net amount
I register as ‘other costs’ for smelting. Table B-I summarizes my overall approach. If the total
cost of smelting is not available (missing data) it can be arrived at by subtracting from the total
cost of production in a month the product of total montones amalgamated times the reported
production cost per monton by amalgamation, including mercury.1000
The cost of the ore at the plant gate is not included, nor is the fixed capital cost.
Extracto de Memorias, 1875 – 1878.
The weekly accounts have served to establish what is included under many of the
headings used in the monthly ledger that were not self-evident in their description. For
999
Under modern accounting practices the inclusion of masses as a production cost may be open to fiscal
questioning, but in the nineteenth century in Mexico it would have been as necessary for the labour force as the
corn they were given.
1000
A montón at Regla is defined as containing 30 quintales (10 cargas, 1.38 t). A carga corresponds to 138 kg
or 3 quintales or 12 arrobas.
592
example, does ‘Mill Grinding’, Molienda de Morteros, encountered in all three accounting
sources include all the costs expected from this stage of the process, or did it only cover labour
costs? It depends on the source. In the case of the Contabilidad Mensual, a comparison of the
numbers reported during those periods where the monthly and weekly accounts overlap
indicates clearly that the data only correspond to labour costs. The same headings when they
appear in the Estados Comparativos include both labour and other, unspecified costs (see
below).
The Extracto de Memorias provide a wealth of detail on the labour component of the
two refining processes used at Regla: the name of the worker in most cases, hours worked per
week, wage per hour and approximate description of work carried out or skills. The other
window into the processes carried out at Regla provided in these weekly accounts concerns the
consumption of fuel for amalgamation. As already pointed out in Chapter 4, the Contabilidad
Mensual only provides the consumption of charcoal for smelting, but apart from some
incomplete records for the years 1872 and 1873 does not include any information on fuel used
during amalgamation. This can be remedied in part for the period 1875-1878 by the information
on total firewood and charcoal consumption that appears in the weekly accounts, as detailed
under the heading Varios Efectos (Miscellaneous Materials). For the earlier years, 1853-1873,
the Estados Comparativos provide more information, as set out below.

593
June 1872- April 1875 - November 1876 - July 1877 - December March 1878 - January 1879 - January 1881-
Monthly Ledger November 1877-
December1873 October 1876 June 1877 December 1878 December 1881 August 1888
1877 February 1878
Miguel Solorzano for
Signed by R. Rule E. Benoit N. Cuataparo R. Torres Landero, Torres
Bustamante Bustamante
headings for present
analysis
salt, mercury,copper sulphate, greta, charcoal, firewood (only 1872-1873), animal feed (cebada, paja ), corn, animals (including deaths)
Materials supplied by the Regla
warehouse (Almacen )
Amount by weight (arrobas, pounds, cargas) and total cost (pesos)
Stamp mills (molienda de morteros )

Ore breakers (almadonetas )
Chilean mills (arrastres )
Amalgamation cake workers
(torteros )
General labourers (peones )
Carriers (cargadores )
Fine ground ore workers (lamadores )
Horseshoe fitters/ironsmiths
(herradores )
Treatment of ground ore residues
(planillones )
Stables (Caballerizas ) Labour costs
Amalgamation costs (Amalgamacion amalgamation
en Patio )
Capellina, final refining and casting of
fine silver bars (Capellina, afinacion
y fundicion de barras )
Repairs (composturas )
Bricklayer (Albañil )
Sawyer (Aserradores )
Breakdown of monthly variable costs Carpenters (Carpinteros )
Roofers (Techadores )
Smiths (Herreros )
General tasks (Faenas )
General costs (gastos generales )
Salt
Materials from warehouse
Mercury
(materiales consumidos del
almacen ) Copper Sulphate
Other materialscosts
Non-recurring (varios efectos )
(gastos
muertos )
Other costs
Contingent costs (Gastos

Extraordinarios / Gastos Diversos )
total amalgamation
= total production cost - total smelting cost
Total amalgamation cost cost
Total smelting costs, or smelting cost

total smelting cost
per carga
Smelting (Fundicion ) labour costs smelting
Litharge (greta ) litharge
Charcoal fuel
Other costs = total smelting cost - labour costs - litharge - fuel other costs
Table B-I. Assignment of account headings in Contabilidad Mensual into subsets (labour,
mercury, salt, copper sulphate and other costs) used in the analysis of production costs at Regla.
Estados Comparativos 1853-1855, 1859-1865, 1869-1873
The comparative tables of production costs incurred at each hacienda contain the
following information reported both on a monthly, quarterly and yearly basis:
Variable production costs: all production costs are listed as pesos per montón in the
case of amalgamation, and pesos per carga in the case of smelting. These are grouped under
many of the same headings encountered in the first two accounting sources, such as Molienda
594
de Morteros, Arrastres, Amalgamación patio, Fundición, etc. However in this case the
accountants decided to include both labour and other, unspecified, costs under these headings.
This became clear on comparing the data from the Contabilidad Mensual for the year 1873
with the data from the Estados Comparativos reported for Regla for the same year (Table B-
II). The amounts under unambiguous headings such as mercury, salt and copper sulphate only
show the very small deviations to be expected from two separate accounting sources averaged
over a year. In contrast, the amounts for grinding, capellinas and repairs are consistently and
significantly higher in the Estados Comparativos. This indicates that the accountants are
including other (unidentified) costs for each process stage apart from the labour costs already
identified for the Contabilidad Mensual. This interpretation is strengthened by the content of
the 1855 report presented by John Buchan, where in his table of comparative production costs
between all the haciendas of the company, he explains some of the accounting headings such
as stamping and grinding, by adding the words: ‘mostly labour’.1001
The case of the accounting of the costs of firewood and charcoal also shows how the
content of each heading changes from source to source. The Contabilidad Mensual to all effects
and purposes simply ignores them in the case of amalgamation, only reporting charcoal for
smelting.1002 The Memorias provide a very detailed breakdown of the weekly costs of all the
types of firewood and charcoal under the heading Varios Efectos. The Estados Comparativos
reports separately and distinctly the costs incurred for two specific but generic headings,
firewood and charcoal. In contrast to the Extractos de Memorias, it excludes them from the
accounting class of Varios Efectos.
1001
1002
This can be explained by the relative small contribution of this consumable to the total costs of amalgamation,
while it is a major cost component in smelting, as will be evident in a later section of this chapter.
595
selection of account headings

total production
accounting source year grinding costs (molienda cost per kg
capellina , repairs copper
de morteros, mercury salt silver
bars (reparos ) sulphate
almacenetas, arrastres )
Contabilidad Mensual 0.94 0.06 0.56 3.02 2.22 0.78 11.57
1873
Estados Comparativos 1.64 0.12 1.19 3.2 2.25 0.76 11.25
note: all production costs in pesos per kg of refined silver, calculated from data in primary source, average for the
year indicated
Table B-II. Comparison of amounts accounted for in the Contabilidad Mensual and Estados
Comparativos for the year 1873 for production costs at Regla. See text for analysis.
Information on the ore processed at each hacienda: number of cargas; the average
total silver content (ley, expressed in marks per montón) of the ore being processed, and the
silver actually extracted, reported also as a ley; the percentage loss of silver at different stages
of the process, the total percentage loss of silver and loss of mercury expressed as ounces of
mercury per mark of silver. Some of this information also appears in the monthly accounts of
Regla within the Contabilidad Mensual.
Information on the consumption of main reagents: total weight consumed of salt,
mercury, copper sulphate, magistral (only in some years), litharge, firewood, charcoal. Total
cost of these consumables is also listed, so that the average yearly cost per unit of weight for
each consumable can be calculated.
Information on the consumption of sundry items: the same yearly average (total
weight and total expense) is reported for all the sundry materials used at the haciendas.1003
1003
Efectos Diversos is mostly made up of : tools (herramientas), machinery (maquinarias), bricks (ladrillos),
lime (cal), planks (tablón), wood (madera), leather (cueros), lead (plomo), iron (hierro) and steel (acero),nails
(clavos), refractory stones (piedra refractaria), limestone (piedra de cal),bone ash (ceniza de hueso),stones for
arrastres (piedras voladoras), capellinas, and dead animals (animales muertos). The expense on fodder (maize,
barley, straw) is absent.
596
C. Inventory areas at the Hacienda de Regla
Raw ore for amalgamation: the inventory of raw ore ranged from zero to nearly
24,000 cargas (3,312 t) over the period (Figure C-1). Under conditions of guaranteed
availability, a very low inventory level lowers the amount of capital tied up in storage, and in
general up to 1885 Regla maintained low levels of raw ore awaiting amalgamation. The
mathematical average for the period is 5,091 cargas (703 t) per month of inventory, but it is
more realistic to calculate the required storage area on the basis of demands for a peak monthly
storage of 15,000 cargas (approx. 2,100 t) that repeats over the period. The ore could be stored
outdoors but with protection from pilfering behind the high perimeter walls of Regla.
30,000
25,000
20,000
cargas
15,000
10,000
5,000
Figure C-1. Inventory of raw ore destined for amalgamation. Raw data from Informe
Mensual.
Ground ore for amalgamation: the estimate of inventory storage requirements is
based on peak levels of 10,000 cargas (Figure C-2).

597
14,000
12,000
10,000
8,000
cargas
6,000
4,000
2,000
Figure C-2. Inventory levels of ground silver ore destined for amalgamation. Data from
Informe Mensual.
Salt: The inventory levels show some major oscillations but overall their median is
relatively constant compared to what I will show for other ingredients of the recipe at Regla.
The average inventory value in this period was 18,459 arrobas (212 t), but with peaks of up to
45,000 arrobas (520 t) as seen in Figure C-3. The importance of salt for the process is shown
by the fact that the average level of inventory covered four months consumption. For the
calculation of storage area I will use a peak value of 40,000 arrobas.
50,000
45,000
40,000
35,000
30,000
arrobas
25,000
20,000
15,000
10,000
5,000
0
Figure C-3. Monthly inventory levels of salt. Raw data from Informe Mensual.
598
Copper sulphate: The way its inventory was managed throughout the period reflects
a reaction to uncertain supply (Figure C-4). The median shows a positive slope after 1877, up
to the moment the decision was taken to draw down the inventory after mid 1886, possibly in
anticipation of a major decrease in operations at Regla. The threat of losing revenues due to
the lack of copper sulphate was real enough, though the alternative option had they run out of
copper sulphate would have been to roast the ores with salt prior to amalgamation, a routine
operation at San Miguel de Regla and other haciendas of the company. On average, with the
caveat that the baseline of the data shows a marked positive slope with time, the level of
monthly inventory over this period was 15,109 lb (6.9 t), reaching a maximum of 45,672 lb (21
t) towards the end of the period. With an average monthly consumption of 9,774 lb (4.4 t), the
inventory represented more than 1.6 months average consumption. I will use 40,000 lb as the
peak level of inventory to estimate the storage area requirements.
50,000
45,000
40,000
35,000
30,000
pounds
25,000
20,000
15,000
10,000
5,000
0
Figure C-4. Monthly inventory levels of copper sulphate. Raw data from Informe Mensual.
Mercury: Throughout this whole period Regla maintained an average monthly
inventory level of mercury of 29,830 lb (13.6 t), Figure C-5. This represents an average
599
inventory equivalent to 6.8 months of average mercury consumption at Regla. I will use 45,000
lb to calculate peak inventory storage requirements.
50,000
40,000
30,000
pounds
20,000
10,000
0
6/72 6/73 9/75 9/76 9/77 9/78 9/79 9/80 9/81 9/82 9/83 9/84 9/85 9/86 9/87 9/88
Figure C-5. Monthly inventory levels of mercury. Raw data from Informe Mensual.
Raw ore for smelting: the inventory profile shows three stages during this period: first,
a period of supply overtaking the capacity of Regla to grind and smelt the ores, leaving a
median with a positive slope during the first half of the decade; then a period when smelting
output was more balanced with ore supply, leading to very rapid draw-downs of accumulated
inventory. It then reverted to the behaviour observed at the beginning, suddenly cut-off in early
1886 when smelting operations ended, most probably with no prior warning, as can be deduced
from the behaviour of the other smelting inventories discussed below (Figure C-6). On average
700
600
500
400
cargas
300
200
100
Figure C-6. Inventory levels of raw silver ore destined for smelting. Raw data from the
Informe Mensual Regla.
600
a monthly inventory of 216 cargas (29.8 t) was carried at Regla, just over one month of smelting
throughput. I will use a peak level of 600 cargas to estimate the area required for storage of
inventory.
Milled ore for smelting: there were also periods when it would have been necessary
to store the milled ore that could not be smelted. It occurred more often at the beginning of the
decade, tailing off to discrete monthly peaks piercing plateaus of sustained zero inventory
(Figure C-7). This is another case where the mathematical average for the decade, 32 cargas
per month (4.4 t), says little from a storage point of view. Each peak would have required
sufficient storage area to have been available at short notice, capable of storing up to 300 cargas
(approx. 42 t) of valuable milled silver ore under secure conditions. Overall Figure C-7
reinforces the image of a smelting infrastructure capable of processing all the available silver
ore, and only running up important inventories of unsmelted ore during very few months over
the whole decade.
350
300
250
200
cargas
150
100
50
0
Feb-75 Feb-76 Feb-77 Feb-78 Feb-79 Feb-80 Feb-81 Feb-82 Feb-83 Feb-84 Feb-85
Figure C-7. Inventory of ground silver ore ready for smelting. Raw data from the Informe
Mensual.
601
Litharge: the behaviour of the inventory of litharge also shows how Regla had to plan
without having any guarantee as to how much ore for smelting was to be provided by the mines
(Figure C-8). The inventory built up at the beginning of the year 1886 reflected already a lower
expectation on the ores to be smelted, prevalent since the 1882’s, but still planned for smelting
to take place. The unexpected hiatus on smelting as of February 1886 would see the inventory
of litharge slowly erode on other uses, probably refinement by cupellation of silver obtained
by amalgamation. The final small jump in inventory belongs to the history of Regla after 1888.
Again, the mathematical average over the period of 69,100 lbs does not provide guidance on
the space requirements for litharge in this decade. Up to 200,000 lbs (approximately 100 t)
would have required storage space, at least up to the early 1880s
250,000
200,000
pounds
150,000
100,000
50,000
0
4/75 4/76 4/77 4/78 4/79 4/80 4/81 4/82 4/83 4/84 4/85 4/86 4/87 4/88
Figure C-8. Inventory of litharge. Raw data from the Informe Mensual.
Charcoal for smelting: the flat-lining of the inventory registered from early 1886 to
the end of the period in question shows a smelting heart that had ceased to beat (Figure B-9).
It would seem that management at Regla was caught unawares at the end of 1885 that smelting
would not continue in the months ahead. Why this charcoal was not sold or used for the heating
of capellinas or casting of silver bars is not clear, though it may indicate that it was kept for
future potential smelting runs, at least until towards the end of 1888.
602
3,500
3,000
2,500
2,000
cargas
1,500
1,000
500
Figure C-9. Inventory of charcoal for smelting. Raw data from the Informe Mensual.
Though the mathematical average of the inventory over the period is equal to 1,094
cargas (151 t) per month of charcoal, storage would have been required to cope with the peak
at 3,000 cargas (414 t) of charcoal inventory.
Estimating storage areas within Regla: The areas required for the stockpiling of
solids stored at Regla have been estimated based on the analysis of the time series of inventory
levels for the period 1872 to 1888 presented above. The calculation of a stockpile area is a
well-established procedure based on the bulk density of the solid, the angle of inclination of
the stockpile unless the solid is being constrained by walls or partitions, the height restrictions
of the area available and the shape of the stockpile. The base area of the stockpile can be
calculated via the geometry of the stockpile or the web pages of industries in the field of solids
handling can be used to obtain the total area requirements for a given quantity of solid material.
Table C-I sets out the relevant figures for the case of Regla. The source of the peak inventory
levels will be found in the detailed discussions of each process stage in the following sections
603
of this chapter, and this accounting data was used to calculate the required storage area. 1004
There is a good match between these calculated values and the approximate areas of the storage
spaces assigned in Figure 4.7.1005
deemed values calculated

operational data calculated approximate area
per stockpile per stockpile assigned
type of space security level space assigned
specific gravity angle of repose length width height calculated area amount Maximum inventory total area required
(m2 )
(m) (m) (m) (m2 ) (t) (t) (m2 )
2.4
raw ore for amalgamation (eq. to rock hard 38.6 10 6 5 300 1,125 2,100 560 open low S1 560
blasted)
2.0
ground ore (stamp mill) for open, with protection
(eq. to rock/stone 38.6 10 6 5 300 900 1,400 467 low S2 560
amalgamation from wind and rain
crushed)
1.9
salt 38.6 5 5 4 165 390 460 195 under roof medium SA1
(eq. to dry gravel)
210
copper sulphate 1.9 38.6 5 5 1.6 26 25 18 19 under roof medium SA1
The liquid would take

mercury 13.56 n/a n/a 20 up a volume of under under roof high SA2 140
1.5 cubic metres
charcoal for amalgamation 0.21 n/a 5 5 4 25 21 12 14 under roof medium FA 50
raw ore for smelting 2.4 38.6 5 3 2.4 75 140 80 43 open high S1 see above
ground ore (stamp mill) for

2 38.6 5 3 2.4 75 120 40 25 open high CBI or CBII
smelting 840
litharge 2 38.6 5 5 2.4 75 120 100 63 open medium CBI or CBII
charcoal for smelting 0.21 n/a 5 5 4 25 21 400 476 under roof medium B1 500
The bars would take

up a volume of
silver 10.49 n/a n/a 2 under roof very high ? n/a
approximately 0.2
cubic metres
Table 4.II Space requirements for storage of inventory within the perimeter walls of Regla.
Table C-I. Areas required by the average inventory of the main reagents, fuel and ore, as
calculated from the raw data in the Informe Mensual Regla.
1004
I have used the calculator on the website http://www.arthon.com/calculators/stockpile.shtml (18 June 2013).
1005
In extraordinary circumstances of unexpected peaks of inventory, all available spaces within Regla would
have been used to store materials.
604
D. Report of the costs of refining by cazo amalgamation and

smelting, 1801.
Razon de la Compra y Fletes de tres Cargas de Metal y el de su beneficio por Cazo y
Patio de ley de tres marcos cada una por ambos beneficios, dividiendo los costos para la mayor
claridad en el concepto de que en la compra del Metal no pase de diez pesos cada carga pues
de lo contrario disminuirá la utilidad del rescatador proporcionalmente de el mas o menos costo
que erogue en la compra de las dichas tres cargas de Metal … A saver
Primeramente por tres cargas de Metal a razón de diez pesos cada una ….030 [p] 0 [rs]
Por el flete de dichas tres cargas a razón de veinte reales carga 007 [p] 4 [rs]
Gastos que se erogan en el beneficio de dichas tres Cargas por el Cazo … A saber
Quebrador …………………………………………………………………… 000 [p] 3 [rs]
Taonero …………………………………………………………………… 000 [p] 4 ½ [rs]
Paja para las Mulas ………………………………………………………. 001 [p] 2 ½ [rs]
Caceadores ………………………………………………………………… 001 [p] 4 [rs]
Palma ……………………………………………………………………. 001 [p] 4 [rs]
Saltierra …………………………………………………………………… 001 [p] 0 [rs]
Taona …………………………………………………………………… 000 [p] 3 [rs]
Fondos …………………………………………………………………… 000 [p] 3 [rs]
Quema y Afinacion………………………………………………………… 000 [p] 3 [rs]
Salarios …………………………………………………………………… 000 [p] 4 [rs]
Perdida de azogue………………………………………………………… 001 [p] 0 [rs]
[total] 8 [p] 7 [rs]
d
Con las dichas tres carg lama que llamamos cocida se pone una pieza o Monton que
produzga tres Marcos de plata q.e para venderlo se erogan los gastos …….. A Saber
Saltierra …………………………………………………………………… 01 [p] 1 [rs]

Maxistral y Cal ….………………………………………………………… 00 [p] 1 ½ [rs]
Repasos …………………………………………………………………… 01 [p] 4 [rs]
Tina y peones …………………………………………………………… 00 [p] 2 ½ [rs]
Azoguero …………………………………………………………………… 00 [p] 3 [rs]
Perdida y Consumido de Azogue …………………………………………. 01 [p] 5 [rs]
Costos 51 [p] 4 [rs]
Han producidos por ambos procesos 9 marcos de plata que entregados al Aviador a 7
ps valen ……………………………………………………………………………. 63 [p]
Resulta de utilidad al Rescatador ……………………………………… 11 [p] 4 [rs]

605
Para beneficiar por Fuego las dichas tres cargas de Metal han de erogar los costos … A saber
Por su compra …………………………………………………………… 030 [p] 0 [rs]
Por su flete ….………………………………………………………… 007 [p] 4 [rs]
Por quebrarlas ……………………………………………………………… 001 [p] 4½ [rs]
Plomillos para ayuda ……………………………………………………… 001 [p] 7 [rs]
Rebolturero ………………………………………………………………… 000 [p] 6 [rs]
Maquila de Horno…………………………………………………………… 009 [p] 0 [rs]
De Carbon …….………………………………………………………… 007 [p] 2 ½ [rs]
Bazo …………………………………………………………………… 001 [p] 2 [rs]
Palma …………………………………………………………………… 000 [p] 4 [rs]
Perdida de Liga 18 @ [arrobas] a 15 p carga ………………………… 022 [p] 4 [rs]
Tienen de Costo 081 [p] 2 [rs]
Produciran de Plata por este beneficio quando bien valla siete marcos quatro onzas que a 7 p
4 valen ………………………………………………………………………… 056 [p] 2 [rs]
Resulta de perdida al Rescatadr 025 [p] 0 [rs]
Por lo que no es beneficiable esta clase de Metales por fuego que son los que abundan
en el Real de Catorce y si lo será el Metal de ley de diez marcos por carga que tenga algún jugo
porque plomoso no lo producen aquellas Minas siempre que los Duenos de ellas vendan a un
precio racional.
Resumen de los Costos de las tres cargas por beneficio de Azogue y los mismos por el
de Fuego
Ymporta el de las tres cargas por el beneficio de Azogue ………… 014 [p] 0 [rs]
Ydm las mismas por Fuego …………………………………………… 043 [p] 6 [rs]
Exceso por el de Fuego 029 [p] 6 [rs]
Se ha hecho este calculo por los infrascriptos actuales Diputados de Mineria de esta
Ciudad en presencia de los Ministros de Real Hacienda de esta Tesoreria, San Luis Potosi Abril
21 de 1801.
[fdo.] Matteo Garcia, Jose Ygnacio de Escalante
A nuestra vista y presencia

606
[fdo.] Cristobal Convalan? Francisco ¿ de Arce
Translation:
Account of the purchase and freight of three cargas of ore and their refining by Cazo
and Patio, with a silver content of three marks each, by both processes, dividing the costs for
more clarity under the condition that the price of the ore will not exceed ten pesos per carga
otherwise the profit of the buyer will decrease proportional to the greater or lesser cost of the
purchase of said three cargas or ore. … As follows:
First for three cargas of ore at ten pesos each one ………………….. 030 [p] 0 [rs]
For the freight of said three cargas at twenty reales per carga …….. 007 [p] 4 [rs]
Costs incurred in the refining of said three cargas by cazo ………. As follows
Bulk miller of ores ……………………………………………………… 000 [p] 3 [rs]
Operator of tahona ………………………………………………………… 000 [p] 4 ½
[rs]
Straw for mules ………………………………………………………. 001 [p] 2 ½
[rs]
Operators of cazos ………………………………………………………… 001[p] 4 [rs]
Palma [?] ……………………………………………………………. 001 [p] 4 [rs]
Impure Salt…………………………………………………………………… 001 [p] 0 [rs]
Tahona …………………………………………………………………… 000 [p] 3 [rs]
Deposits …………………………………………………………………… 000 [p] 3 [rs]
Firing [amalgam] and refining …………………………………………… 000 [p] 3 [rs]
Salaries …………………………………………………………………… 000 [p] 4 [rs]
Loss of mercury ………………………………………………………… 001 [p] 0 [rs]
With these said three cargas of silt we call cooked we prepare a monton that produces
three marks of silver so that to sell we incur the costs ……………….. As
follows
Impure salt …………………………………………………………………… 01 [p] 1 [rs]

Magistral and lime ….………………………………………………………… 00 [p] 1 ½ [rs]
Mixing the monton…………………………………………………………… 01 [p] 4 [rs]
Washing vat and workers …………………………………………………… 00 [p] 2 ½ [rs]
Master amalgamator ………………………………………………………….. 00 [p] 3 [rs]
Loss and consumption of mercury ……………………………………… 01 [p] 5 [rs]
Costs 51 [p] 4 [rs]
607
Both processes produced 9 marks of silver that sold to the Supplier at 7 pesos are worth
………………………………………………………………………………. 63 [p]
Result of profit to the Buyer [of the ore] ……………………………………… 11 [p]
4 [rs]
To refine by smelting said three cargas of ore the costs incurred are ………… As follows
For their purchase …………………………………………………………… 030 [p] 0 [rs]
For their freight ….………………………………………………………… 007 [p] 4 [rs]
For milling ……………………………………………………………… 001 [p] 4½ [rs]
Lead rich ore as flux ……………………………………………………… 001 [p] 7 [rs]
Mixer ………………………………………………………………… 000 [p] 6 [rs]
Rental of furnace …………………………………………………………… 009 [p] 0 [rs]
Charcoal …….………………………………………………………… 007 [p] 2 ½ [rs]
Cupell …………………………………………………………………… 001 [p] 2 [rs]
Palma [?] …………………………………………………………………… 000 [p] 4 [rs]
Loss of lead flux 18 @ [arrobas] at 15 p carga ………………………… 022 [p] 4 [rs]
Total cost 081 [p] 2 [rs]
Will produce silver by smelting when all proceeds well seven marks four ounces that at 7 pesos
4 [reales] have a value of ………………………………………………………… 056 [p] 2
[rs]
Resulting loss for the Buyer 025 [p] 0
[rs]
So that this type of ore cannot be refined by smelting, those that are abundant in the
Real de Catorce and can be refined is the ore with ten marks per carga that contains some
[lead] flux because lead-rich ores are not produced by those Mines, as long as the mine owners
sell them at a reasonable price.
Summary of the costs for the three cargas refined by mercury and the same by smelting
From the costs for the refining of three cargas by mercury………..………… 014
[p] 0 [rs]
608
Idem for the same by smelting …………………………………………… 043
[p] 6 [rs]
Excess [costs] by smelting 029 [p] 6 [rs]
This calculation has been made by the undersigned, at present Mining Deputies of this
City, in the presence of the Ministers of the Royal Treasury, San Luis Potosi, April 21 of 1801.
[signed] Matteo Garcia, Jose Ygnacio de Escalante
Witnessed in our presence
[fdo.] Cristobal Convalan [?] Francisco --- de Arce
Source: AGN, Instituciones Coloniales / Minería / 28368 / Volumen 82, folio 86 r,v
609
E. Sensitivity matrix for refining costs
Table E-I. Amalgamation sixteenth century context
% silver in ore 0.00% 0.04% 0.05% 0.06% 0.08% 0.12% 0.19% 0.60% 1.00% 1.90% 3.00%
kg of silver in
0.00 0.43 0.62 0.75 0.99 1.49 2.36 7.45 12.42 23.60 37.26
Amalgamation monton
value silver pesos 0 17 24 28 38 57 90 283 472 897 1416
variable production costs

pesos per kg silver
Fuel 0.04 0.00 0.02 0.03 0.03 0.04 0.06 0.10 0.32 0.54 1.02 1.61
Mercury 9.80 0.00 4.26 6.09 7.30 9.74 14.61 23.13 73.04 121.74 231.30 365.22
Salt 1.78 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21 4.21
Copper Sulphate 0.72 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69
Labour 0.13 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31
others 1.93 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54
ore 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
cost power 1.20 2.82 2.82 2.82 2.82 2.82 2.82 2.82 2.82 2.82 2.82 2.82
capital cost 0.50 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.18
total 16.10 14.76 19.04 20.87 22.09 24.54 29.43 37.99 88.12 137.03 247.08 381.58
variable production costs in pesos per kg silver
n/a 43.79 33.61 29.65 24.70 19.75 16.10 11.83 11.03 10.47 10.24
Table E-II. Smelting sixteenth century context
% silver in ore 0.00% 0.04% 0.05% 0.06% 0.08% 0.12% 0.20% 0.60% 1.00% 0.80% 1.20% 1.90% 3.00%
Smelting kg of silver in a
0 0.0435 0.0621 0.0745 0.0994 0.149 0.2422 0.7452 1.242 0.9936 1.4904 2.3598 3.726
carga
value of silver,
0 2 2 3 4 6 9.2 28 47 38 57 90 142
pesos
variable production cost
pesos per kg silver
Fuel 2.23 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27
Litharge 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Labour 0.13 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31
others 0.85 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00
ore 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
total 3.21 7.58 7.58 7.58 7.58 7.58 7.58 7.58 7.58 7.58 7.58 7.58 7.58 7.58
n/a 174.33 122.03 101.69 76.27 50.85 31.29 10.17 6.10 7.63 5.08 3.21 2.03
610
Table E-III. Amalgamation seventeenth and eighteenth century context
% silver in ore 0.00% 0.04% 0.05% 0.06% 0.08% 0.12% 0.19% 0.60% 1.00% 1.90% 3.00%
Amalgamation kg of silver in monton 0.00 0.43 0.62 0.75 0.99 1.49 2.36 7.45 12.42 23.60 37.26
value silver pesos 0 17 24 28 38 57 90 283 472 897 1416
variable production costs pesos per

kg silver
Fuel 0.04 0.00 0.02 0.03 0.03 0.04 0.06 0.10 0.32 0.54 1.02 1.62
Mercury 2.68 0.00 1.17 1.66 2.00 2.66 3.99 6.32 19.97 33.29 63.25 99.86
Salt 3.66 8.64 8.64 8.64 8.64 8.64 8.64 8.64 8.64 8.64 8.64 8.64
Copper Sulphate 0.72 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69 1.69
Labour 0.67 1.58 1.58 1.58 1.58 1.58 1.58 1.58 1.58 1.58 1.58 1.58
others 1.93 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54 4.54
ore 16.7 39.3 39.3 39.3 39.3 39.3 39.3 39.3 39.3 39.3 39.3 39.3
cost power 2.84 6.70 6.70 6.70 6.70 6.70 6.70 6.70 6.70 6.70 6.70 6.70
capital cost 1.32 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13 3.13
total 30.51 65.57 66.76 67.26 67.60 68.28 69.63 72.00 85.87 99.40 129.84 167.05
n/a 153.57 108.31 90.71 68.72 46.72 30.51 11.52 8.00 5.50 4.48
Table E-IV. Smelting seventeenth and eighteenth century context.
% silver in ore 0.00% 0.04% 0.05% 0.06% 0.08% 0.12% 0.20% 0.60% 1.00% 0.80% 1.20% 1.90% 3.00%
Smelting kg of silver in a carga 0 0.0435 0.0621 0.0745 0.0994 0.149 0.2422 0.7452 1.242 0.9936 1.4904 2.3598 3.726
value of silver, pesos 0 2 2 3 4 6 9.2 28 47 38 57 90 142
variable production cost

pesos per kg silver
Fuel 2.23 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27 5.27
Litharge 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Labour 0.66 1.56 1.56 1.56 1.56 1.56 1.56 1.56 1.56 1.56 1.56 1.56 1.56 1.56
others 0.85 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00
ore 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
total 3.74 8.83 8.83 8.83 8.83 8.83 8.83 8.83 8.83 8.83 8.83 8.83 8.83 8.83
n/a 203.09 142.16 118.47 88.85 59.24 36.45 11.85 7.11 8.89 5.92 3.74 2.37
611
F. Smelting costs in Europe.
It is stated that ‘German mines could not compete with the lower cost of American
silver and stagnated in the sixteenth century’, but there is no study that compares specific silver
production costs on both sides of the Atlantic.1006 The economics of smelting the majority of
silver ores in Europe followed a completely different tsructure than the one applied to silver
ores smelted in the New World, at least until the end of the nineteenth century. Three cases of
European smelting costs from the nineteenth century plucked from a non-exhaustive search in
the literature represent Germany, France and England. They show that it is not possible to carry
out a straightforward comparison of smelting costs on both sides of the Atlantic. In the case of
Europe it is the nature of the ores being treated and the existence of markets for non-precious
metals that determines the profitability of refining, not their silver content as in the analysis for
Regla.
Under Napoleon the French oversaw silver production in the German mines of the Harz
region, and as a result Villefosse drew up a report that contains partial production costs both
for mining of lead-silver ores and for their smelting for the years around 1805. On the basis of
his data it is possible to calculate that the ores smelted at a profit had between 0.01 and 0.04%
silver content. This would not have been possible at Regla using smelting, not for technical
reasons, but simply because the value of the silver content alone could not have covered the
cost of smelting. However, what allowed the German smelters to obtain a profit was the amount
of revenues from the sale of litharge, lead and copper, which in total amounted to 75%
1006
Brown, History of Mining, 42.
612
(Lautenthal), 94% (Frankenscharer), 104% (Altenau) and 45% (Andreasberg) of the smelting
costs.1007
From the data on the smelting of ores at Pontgibaud (France) from 1838 to 1849
reported by Rivot, it is possible to calculate that the average cost of refining 100 kg of ore with
0.1% silver was 23.45 francs. The silver content is similar to that of the ores smelted at Regla,
but approximately 200 times more lead and litharge was produced than silver, so production
costs cannot be apportioned easily to one or the other. The average value of the silver obtained
was only 24 francs, thus barely exceeding the refining costs, which approximates the scenario
at Regla where the minimum was 0.3% silver to cover costs. Again, the only reason the French
smelters were able to operate at a profit were the average additional revenues of 8.2 francs from
the sale of lead products per 100 kg of ore.1008
Rivot also refers to English smelters in Flintshire in the nineteenth century working
with ores that contain mainly lead (ca 75%) and little silver (25 to 35 gr of silver per 100 kg of
lead, less than 0.03% silver in the ore). He could not include real production data from these
works in his very extensive book on metal smelting since ‘the directors do not like to provide
strangers with details on their commercial affairs’, a sentiment that has dogged most archival
research for this topic. His case study for a generic smelter of Flintshire is based on a process
using crystallization (Patterson’s process) to enrich the lead prior to cupellation (Chapter Four).
He assumes a yearly refining of 21,000 tons of ore, producing 15,005 tons of lead and 3,740
kg of silver. Silver represents just 0.03% in weight of the metal sold from this generic plant,
but provided 10% of its sales revenues. There is little point in estimating which part of the total
1007
Héron de Villefosse, De la richesse minérale du Royaume de Westphalie. Raw data for my calculations taken
from Table facing p. 102.
1008
Rivot, Description des gites métallifères, 193-197.
613
production cost of 18.6 francs per ton of ore smelted can be apportioned to silver.1009 This
example encapsulates very well the symbiotic relationship that characterized the silver smelting
business in Europe. Lead and copper would bear most if not all of the variable production cost,
and silver would bring an important contribution to the revenue stream out of proportion to its
output in weight. It is no wonder that Percy reports that in England ‘foreign silver ores, chiefly
from South America [with a silver content upwards of 0.8%], have been largely imported and
smelted for the past 40 years … the business appears to have been highly profitable … not
more than 2s 6d was paid per oz of silver in the ore!’.1010
1009
‘les directeurs n’aiment pas à donner aux étrangers des renseignements exacts sur leurs affaires
commerciales’ Principes généraux Vol II, 317-19, 386-88.
1010
Percy, Metallurgy, I 524-525. If Collins is correct that 1 peso was equivalent to approximately 2s 6d in the
19c, the price being paid for the ore was around 1 peso per oz of silver, or 8 pesos per mark. I am not sure whether
the celebration in the exclamation mark comes from the seller, the buyer or both. Collins, Metallurgy of Lead &
Silver, Vol. II, 61.
614
Glossary of technical terms
Afinación: second stage of refining process where silver is separated from lead and litharge
(greta) is produced (cupellation)
Amalgamation: the original term refers to a physical process whereby mercury can absorb gold,
silver, lead and other metals and form a liquid or solid solution which does not alter the
chemical characteristics of either mercury or the metals. In the terminology of refining silver
ores in the New World it has been applied to represent a method of refining silver or gold using
mercury to extract silver in the form of an amalgam
Arrastre: circular grinding equipment using horizontal stones, powered by water or animal
power
Assay: analysis or test of an ore to determine the presence and amount of metal
Aviador: person who supplies material on credit to miners and refiners
Azogado: person intoxicated with mercury
Azogar : to intoxicate with mercury
Azogue : mercury
Azogueria: room where mercury was handled (stored, weighed, extracted by squeezing from
amalgam)
Azoguero : in New Spain, applied to the master in charge of the amalgamation process
Barra: bar of silver-enriched lead from first stage of smelting (see refinación)
615
Beneficio: term applied to refining, ie beneficio de plata por azogue meant refining of silver
ores by amalgamation, beneficio de plata por fuego, refining by smelting
Blast furnace: a more efficient furnace than the Horno Castellano, ore was loaded from the
top, a greater current of air was fed by force into the furnace and chimney height was increased
Bonanza: a very rich zone of precious metal in a deposit
Caja: regional Treasury, usually but not always a mining district
Calomel: mercurous chloride (Hg2Cl2)
Capellina: 1. Equipment to recover mercury from amalgam by heating, consisting of a metallic
top cover and base placed on water channel that condenses mercury 2. In Guanajuato, also
applied to building that houses the capellina ensemble.
Caperuza: upper part of early version (16c) of equipment to recover mercury from amalgam,
made from clay or metal
Carbón: see charcoal
Carcámo: channel to drain waste water that run through amalgamation hacienda.
Cazo: a pot or vessel, used by Barba for his cocimiento (cooking) process using mercury
Cendrada: bone ash impregnated with litharge, material used to form a cupel (vaso) that holds
the barras for the afinación
Charcoal: mainly carbon, is used both to supply heat upon combustion in a furnace and to act
as reducing agent to the metal compounds being smelted.
Chilean mill: see Molino

616
Correspondencia: amount of silver, in marks which could be produced by amalgamation that
consumed 100 quintales of mercury
Cupellation: see Afinación.
Deflaction : adjustment of time series of prices by the rate of inflation
Desazogadera: equipment to recover mercury from amalgam by heating, see capellina
Diezmo: tax of a tenth applied to silver registered at each local Treasury (Caja)
Dressing: concentration of silver content in ores by decantation in water.
Dry ore: silver ore with little or no lead destined for smelting
Fundición: first step of the refining of silver ores, it involves smelting of silver compounds in
the presence of lead to elemental silver. Silver is then absorbed by the molten lead. The silver-
enriched lead is then cast into barras (pigs)
Flue gas: gas generated in a furnace and channeled via a chimney to the atmosphere
Flux : additive used to facilitate the smelting of an ore
Fume: an aerosol of particles of lead and lead compounds (PbO, PbS, PbCO3, others) that are
lost to the atmosphere when lead and lead ores are heated above a threshold temperature
Fundición: smelting of silver compounds in the presence of lead to elemental silver. Silver is
then absorbed by the molten lead. The silver-enriched lead is then cast into barras (pigs)
Galena / Galena: lead sulphide (PbS), which can contain silver that can be extracted by
smelting
Gangue: the inorganic matrix with no commercial value that is extracted together with the
mineral of commercial interest from a deposit

617
Grasas: slag from smelting furnace
Greta: litharge, lead oxide (PbO)
Hacienda: original term referred to the creation of wealth, was then applied initially to silver
refining units in New Spain (called ingenios in Peru) and afterwards to agricultural and
livestock economic units
Horno Castellano: initially very simple smelting furnaces, in the form of a pillar with a square
or circular cross-section, built from mortar and stones and with a low chimney outlet
Ingenio: originally refers to a machine, and then in Peru was used to denote a silver refining
facility (an hacienda in New Spain).
Lavado de metales: see Dressing
Litharge: lead oxide, PbO (greta)
Maestrazgo: land and mining rents to the Spanish Crown from territories that historically were
under the control of Spanish military orders
Manga: vertical cloth filter used to squeeze excess mercury from amalgam
Maquila: business model whereby a refining hacienda accepts to process silver ores that belong
to third parties and extracts the silver for a fee that covers its operational costs plus a profit
margin
Mark: unit of weight of silver
Metal : ore in Spanish
Metallogeny: the study of how ore deposits are created

618
Mole: in chemistry, is a standard weight for each chemical element. The number of moles of a
reagent or product in a reaction is determined by the chemical equation of the reaction.
Molino: circular stone set on its edge and driven by water or animal power, used to crush ore
Montón: literally mound, was a unit of measure in the amalgamation patio, thus a torta at Regla
was composed of 20 montones, and each montón represented 30 cargas (see Units of Measure).
These are not universal values and vary according to local custom.
Mortero: mill that uses stamp-heads made of stone or metal to crush ore, driven by human,
animal or water power
Negrillos: term used to denote deeper silver sulphides in a deposit
Opportunity cost: the economic consequence of choosing an investment between two
competing business options
Ore : a naturally occuring material from which a mineral can be extracted at a profit
Oxidation: in modern chemistry denotes the increase of the oxidation state of an element, due
to the loss of electrons to another element that in turn is being reduced (see Reduction)
Patio: the courtyard where amalgamation cakes (tortas) were spread out until the silver
amalgamation process was deemed completed
Pig : English term in smelting for a bar (barra) of smelted metal
Placer gold: gold found in alluvial or eluvial deposits
Planillas: inclined planes to separate entrained amalgam, mercury or silver ore from the
washings of the amalgamation torta
Planilleros: workers stationed at the planillas

619
Plata de azogue: silver refined by amalgamation
Plata de fuego: silver refined by smelting
Real de Minas: legally established mining interests and community recognized by the Spanish
Crown and subject to its legislation
Reduction: in modern chemistry denotes a decrease in the oxidation state of an element, by
accepting electrons from an element that in turn undergoes oxidation (see Oxidation), for
example carbon in charcoal. The ultimate aim of all refining techniques for metals is to
chemically reduce the metal from its oxidized state in the ore, for example silver sulphide or
silver chloride, down to its elemental state as pure metal
Reverberatory oven: indirect heating reflected from curved roof, fuel is not in contact with
barras in the vaso
Señoreaje : duty paid on coinage of silver
Silver chloride: AgCl, when found as a mineral is called cerargyrite or chlorargyrite
Silver halides: silver chloride, silver bromide, silver iodide
Silver sulphide: AgS, when found as a mineral is called acanthite or in older texts argentite
Slag: fused waste minerals from smelting furnace, may contain lead and lead compounds, iron
compounds, traces of arsenic
Slurry: viscous suspension of fine solids in water
Smelting: metallurgical process based on the chemical reduction of metal compounds in ores
at high temperatures (see Fundición)
Solimán: mercuric chloride (HgCl2)

620
Stamp mill: see Mortero
Stoichiometry: fixed quantitative relation between reactants and products in a chemical
reaction
Subduction: geological process whereby an expanding ocean floor slides under continental
masses
Sunk cost: losses incurred in past operations that from an accounting point of view are written
off from future accounting operations involving the same product
Tahona or taona: see arrastre
Toll / tolling: see Maquila
Torta: amalgamation cake
Tuyere: element in back furnace wall with orifice to hold the nozzle (cañon) of the bellows
Vaso: term used for the ensemble of reverberatory oven and cupel used during the afinación
Weathered, weathering: as applied to ore deposits means the action of oxygen, water and
salinity over hundreds or thousands of years on the chemistry of the surface veins of the deposit
Wind rose: circular plot over 360 degrees of average speeds and direction of winds at a given
locality
Zangarro: amalgamation unit smaller than the hacienda

621
Archival sources
Archivo General de la Nación, Ciudad de México (AGN)
Instituciones Coloniales / Minería

Centro de Información Gráfica (CIG-AGN)
Serie Tema y Tecnología
Archivo Histórico del Estado de Zacatecas (AHEZ)

Notarías/Colonia
Poder Judicial-Civil
Real Hacienda
Real Hacienda- Judicial
Serie Civil
Fondo Mapas e Ilustraciones, Serie V: Planos Siglos XVIII al XX
Archivo Histórico de San Luis Potosí (AHSLP)
Fondo Alcaldía Mayor

Colección Miguel Iwadare
Colección Mapas y Planos
Archivo Histórico de la Universidad de Guanajuato (AHUG)
Fondo Ayuntamiento de Guanajuato

Protocolo de Minas
Minería
Actas de Cabildo
Bienes Difuntos
Ramo Tierras
Fondo Agencia de Minería
Mapoteca
Mapoteca Manuel Orozco y Berra, Ciudad de México (MMOB)

Colección General
Archivo Histórico de la Compañia Real del Monte y Pachuca (AHCRMyP)

Fondo Siglo XIX:
Sección: Explotación y Beneficio

Serie: Informes de Haciendas de Beneficio
Subserie: Informes Mensuales Hacienda de Regla Vol 225, Exp. 3: 29 Jun 1872 – 27 Oct 1888
Sección: Negociaciones
Serie: Haciendas de Beneficio
Subserie: Hacienda de Regla I, Vol 22: 1875-1878, Vol 23: 1878-1881, Vol 27ª: 29 Dicicembre
1877-11 Febrero 1888, Vol 28: Abril 1885 – Oct 1888
622
Sección: Correspondencia, Serie: Compañía a Varios, Subserie: Correspondencia General

8-1: 20 Abril 1825 – 1 Noviembre 1825
12-2: 20 Julio 1827- 6 Julio 1832
33-11: 20 Marzo 1834 – 20 Abril 1835
50-25: 28 Junio 1844- 3 Junio 1849
45: 14 Diciembre 1839 – 3 Marzo 1849
30-61: 1 – 30 Mayo 1856
Sección: Contabilidad de la Dirección

Serie: Producción y Gastos
Subserie:
Estados Comparativos, 1853-1855, 1859-1865, 1869-1873.
Seccion: Contabilidad de la Direccion

Serie: Produccion y Gastos
Subserie:
Gastos de Haciendas de Beneficio, Marzo 1871 – Nov 1877
Gastos de Haciendas de Beneficio, Sept. 1896 – Sept 1902
Sección: Administración Interna

Serie: Departamento de Ingenieros
Subserie: Croquis y Planos Vol 204 Carpeta 1
British Library, London, U.K. (BL)
Bibliography
Aaron, C.H. Leaching Gold and Silver Ores: The Plattner and Kiss Processes Barry & Baird,
Printers, 1881.
Acosta, José de. Historia natural y moral de las Indias. edited by Fermín del Pino Madrid:
Consejo Superior de Investigaciones Científicas, 2008.
Agricola, Georgius. Bermannus. Translated by Robert Halleux and Albert Yans Paris: Belles
Lettres, 1990.
———. De re metallica. Translated by Herbert Clark Hoover and Lou Henry Hoover New
York: Dover Publications, 1950.
Aitchison, Leslie. History of Metals London: Macdonald & Evans, 1960.
Alonso, Ricardo N. Diccionario Minero. Glosario de voces utilizadas por los mineros de
Iberoamérica Madrid: Consejo Superior de Investigaciones Científicas, 1995.
Amador, Manuel. Tratado práctico y completo de trabajos de minas y haciendas de beneficio
México: E. Sánchez, Editor, 1901.
Anonymous. "Descripción de la Villa y Minas de Potosí, año de 1603 ". In Relaciones
Geográficas de Indias - Perú I edited by José Urbano Martínez Carrera Biblioteca de
Autores Españoles, 372-385. Madrid: Atlas, 1965 (1588).
———. "Journal Descriptive of the Route from New York to Real del Monte by Way of
Tampico." The London Magazine 1826, 145-171.
623
———. "Memoirs of members deceased in 1881." Proceedings - Institution of Mechanical

Engineers 36 (1882): 13-14.
Antúnez Echegaray, Francisco. Monografía histórica y minera sobre el Distrito de Guanajuato
Mexico: Consejo de Recursos Naturales No-Renovables, 1964.
Arduz Eguía, G. Ensayos sobre la historia de la minería altoperuana Madrid: Editorial
Paraninfo, 1985.
Arzáns de Orsúa y Vela, Bartolomé. Historia de la villa imperial de Potosí. edited by Lewis
Hanke and Gunnar Mendoza L Providence: Brown University Press, 1965.
Atwell, William S. "International Bullion Flows and the Chinese Economy circa 1530-1650."
Past and Present 95 (1982): 68-90.
Atwood, Margaret. MaddAddam : a Novel New York: Doubleday, 2013.
Bachmann, Hans-Gert. "Archäometallurgische Untersuchungen zur antiken Silbergewinnung
in Laurion. II. Charakterisierung von Bleiverhüttungsschlacken aus Laurion=
Archaeometallurgical Investigation on Ancient Silver Smelting at Laurion. II.
Characterisation of Lead Smelting Slags from Laurion." Erzmetall 35, no. 5 (1982):
246-251.
Bailly-Maître, Marie-Christine. L'argent : du minerai au pouvoir dans la France médiévale
Paris: Picard, 2002.
Bakewell, Peter. "La transferencia de la tecnología y la minería hispanoamericana, siglos XVI
y XVII: algunas observaciones." In Hombres, técnica, plata : minería y sociedad en
Europa y América, siglos XVI-XIX, edited by Julio Sánchez Gómez, Guillermo Mira
Delli-Zotti and Francisco A. Rubio Durán 363-375. Sevilla: Aconcagua Libros, 2000.
Bakewell, Peter J. "The First Refining Mills in Potosi: Design and Construction.". In In Quest
of Mineral Wealth: Aboriginal and Colonial Mining and Metallurgy in Spanish
America., edited by Alan K. Craig and Robert C. West 287-302. Baton Rouge:
Geoscience Publishers, 1994.
———. A History of Latin America. 2nd ed. Oxford: Blackwell Publishing, 2004.
———. "Introduction." In Mines of Silver and Gold in the Americas edited by Peter J.
Bakewell xiii-xxiv. Aldershot (UK): Variorum, 1996.
———. Miners of the Red Mountain. Indian Labor in Potosí, 1545-1650 Albuquerque:
University of New Mexico Press, 1984.
———. "Mining in Colonial Spanish America." In The Cambridge History of Latin America,
edited by Leslie Bethell 105-151. Cambridge: Cambridge University Press, 1984.
———. "Notes on the Mexican Silver Mining Industry in the 1590's." In Mines of Silver and
Gold in the Americas, edited by Peter J. Bakewell 383-409. Aldershot (GB); Brookfield
(Vt.): Variorum, 1996.
———. "Registered Silver Production in the Potosi District 1550-1735." Jahrbuch für
Geschichte Lateinamerikas 12 (1975): 67-103.
———. Silver and Entrepeneurship in Seventeenth-Century Potosí. The Life and Times of
Antonio López de Quiroga Dallas: Southern Methodist University Press, 1995.
———. Silver Mining and Society in Colonial Mexico: Zacatecas, 1546-1700 Cambridge,
U.K.: University Press, 1971.
Baldwin, M. "Alchemy and the Society of Jesus in the Seventeenth Century: Strange
Bedfellows?". Ambix 40, no. 2 (1993): 41-64.
Barba, Alvaro Alonso. Arte de los metales Barcelona: Editorial Labor, 1977 (1770).
Bargalló, Modesto. La minería y la metalurgia en la América española durante la época
colonial Mexico: Fondo de Cultura Económica, 1955.
———. La química inorgánica y el beneficio de los metales en el México pre-hispánico y
colonial Mexico: Universidad Nacional Autónoma de México, 1966.
624
Bartels, Christoph. "The Production of Silver, Copper and Lead in the Harz Mountains from
Late Medieval Times to the Onset of Industrialization." In Materials and Expertise in
Early Modern Europe : Between Market and Laboratory, edited by Ursula Klein and
E. C. Spary 71-100. Chicago: University of Chicago Press, 2010.
Berthe, Jean Pierre. "Le mercure et l'industrie mexicaine au XVIéme siècle." In Ciencia, vida
y espacio en Iberoamérica: trabajos del programa movilizador del C.S.I.C.,
"Relaciones científicas y culturales entre España y América". edited by Jose Luis Peset
Reig 141-149. Madrid: Consejo Superior de Investigaciones Científicas, 1989.
Biringuccio, Vannoccio. The Pirotechnia. Translated by Cyril Stanley Smith and Martha Teach
Gnudi Cambridge, Mass.: The M.I.T. Press, 1966. Venice, 1540.
Blanchard, Ian. "England and the International Bullion Crisis of the 1550s." In Precious Metals
in the Age of Expansion: Papers of the XIVth International Congress of the Historical
Sciences edited by Hermann Kellenbenz 87-118. Stuttgart: Klett-Cotta, 1981.
———. Mining, Metallurgy and Minting in the Middle Ages. Continuing Afro-American
Supremacy 1250-1450. . Vol. 3 Munich: Franz Steiner Verlag Stuttgart, 2005.
———. Russia's "Age of Silver". Precious-metal Production and Economic Growth in the
Eighteenth Century London; New York: Routledge, 1989.
———. "Technical Implications of the Transition from Silver to Lead Smelting in Twelfth
Century Britain." In Boles and Smeltmills Seminar, edited by Lynn Willies and David
Cranstone 9-11. Reeth, Yorkshire: Historical Society, Ltd., 1992.
Bleiberg, German. El Informe Secreto de Mateo Aleman sobre el trabajo forzoso en la Mina
de Almaden London: Tamesis Books Ltd, 1985.
Bordeaux, Albert. Le Mexique et ses mines d'argent Plon-Nourrit et cie., 1910.
Borgstrom, Georg. The Hungry Planet; the Modern World at the Edge of Famine New York:
Macmillan, 1965.
Born, Inigo. Baron Inigo Born's New Process of Amalgamation of Gold and Silver Ores, and
Other Metallic Mixtures: As by His Late Imperial Majesty's Commands Introduced in
Hungary and Bohemia. Translated by R.E. Raspe T. Cadell, 1791.
Bowden, Mark. Furness Iron : the Physical Remains of the Iron Industry and Related
Woodland Industries of Furness and Southern Lakeland Swindon: English Heritage,
2000.
Brading, D. A. "Mexican Silver-Mining in the Eighteenth Century: The Revival of Zacatecas."
The Hispanic American Historical Review 50, no. 4 (1970): 665-681.
Brading, D. A., and Harry E. Cross. "Colonial Silver Mining: Mexico and Peru." The Hispanic
American Historical Review 52, no. 4 (1972): 545-579.
Brading, D.A. Miners and Merchants in Bourbon Mexico 1763-1810 London: Cambridge
University Press, 1971.
Braunstein, P. "Innovations in Mining and Metal Production in Europe in the Late Middle
Ages." Journal of European Economic History 12 (1983): 573-591.
———. "Les entreprises minières en Vénétie au XVe siècle." Melanges d'archéologie et
d'histoire 77, no. 2 (1965): 529-607.
Bray, Warwick. "Ancient American Metal-Smiths." Proceedings of the Royal Anthropological
Institute of Great Britain and Ireland, no. 1971 (1971): 25-43.
Brocklehurst, Thomas Unett. Mexico To-day: a Country with a Great Future, and a Glance at
the Prehistoric Remains and Antiquities of the Montezumas London: J. Murray, 1883.
Brooks, Timothy. Vermeer’s Hat London: Profile, 2008.
Brown, John Stafford. "Supergene Sphalerite, Galena, and Willemite at Balmat, New York."
Economic Geology 31, no. 4 (June 1, 1936 1936): 331-354.
Brown, Kendall W. A History of Mining in Latin America : From the Colonial Era to the
Present Albuquerque: University of New Mexico Press, 2012.
625
———. "Workers Health and Colonial Mercury Mining at Huancavelica, Peru." The Americas
57 (2001): 467-496.
Buchan, John H. Report of the Real del Monte Company, Mexico London: Taylor, Printer,
1855.
Budd, Paul, A. Mark Pollard, Brett Scaife, and Richard G. Thomas. "The Possible Fractionation
of Lead Isotopes in Ancient Metallurgical Processes." Archaeometry 37, no. 1 (1995):
143-150.
Buechler, Rose Marie. "Technical Aid to Upper Peru: The Nordenflicht Expedition." Journal
of Latin American Studies 5, no. 1 (1973): 37-77.
Burkart, J. "Du filon et des mines de Veta-Grande, près de la ville de Zacatecas, dans l'état du
même nom, au Mexique." Annales des Mines, 3eme série 8 (1835): 55-87.
Burkart, Johann. "Memoria sobre la explotación de minas de los distritos de Pachuca y Real
del Monte de México." Anales de la Minería Mexicana (Revista de Minas) I (1861): 3-
25,41-65,81-113.
Burt, Roger. "The Transformation of the Non-Ferrous Metals Industries in the Seventeenth and
Eighteenth Centuries." The Economic History Review 48, no. 1 (1995): 23-45.
Calderón de la Barca, Fanny, ed. Life in Mexico; the Letters of Fanny Calderón de la Barca,
with New Material from the Author's Private Journals edited by Howard T. Fisher and
Marion Hall Fisher Garden City, N.Y.: Doubleday, 1966.
Camargo, Julio A. "Contribution of Spanish–American Silver Mines (1570–1820) to the
Present High Mercury Concentrations in the Global Environment : A Review."
Chemosphere 48 (2002): 51-57.
Canterla y Martín de Tovar, Francisco. Vida y obra del primer Conde de Regla Sevilla: Escuela
de estudios hispano-americanos, 1975.
Capoche, Luis. "Relación General de la Villa Imperial de Potosí." In Biblioteca de Autores
Españoles, edited by Lewis Hanke 115-161? Madrid: Atlas, 1959.
Cárdenas, Juan de. Primera parte de los problemas y secretos maravillosos de las Indias [in
Spanish]. México: Academia Nacional de Medicina, 1980.
Cargill, S.M., D.H. Root, and E.H. Bailey. "Resource Estimation from Historical Data:
Mercury, A Test Case." Mathematical Geology 12, no. 5 (1980): 489-522.
Carlson, R. W. "Introduction." In Treatise on Geochemistry : The Mantle and Core., edited by
R. W. Carlson xv-xxi. Oxford: Elsevier-Pergamon, 2004.
Carson, Rachel. Silent Spring Greenwich, Conn.: Fawcett, 1962.
Casasus, Joaquin D. La Question de l'argent au Mexique [in French]. Paris: impr. de Chaix,
1892.
Castelazo, Jose Rodrigo. Manifiesto de las riquezas que han producido y actualmente
contienen las celebradas minas de las vetas Vizcaina y Santa Brigida ubicadas en el
Real del Monte, jurisdicción de Pachuca, de las grandes obras que en ellas se hicieron
y del estado en que actualmente se hallan Mexico: Imprenta de D. Mariano Ontiveros,
1823.
Castillo Martos, Manuel. "Alquimia en la metalurgia de plata y oro en Europa y America ". In
Informes para obtener plata y azoque en el mundo hispánico XIII-LXIX. Granada:
Universidad de Granada, 2008.
———. Bartolomé de Medina y el siglo XVI Santander: Servicio de Publicaciones de la
Universidad de Cantabria, 2006.
———. "Introduccion." In Minería y metalurgia: intercambio tecnológico y cultural entre
América y Europa durante el período colonial español, edited by Manuel Castillo
Martos 31-43. Sevilla, Bogota: Muñoz Montoya y Montraveta Editores, 1994.
626
———. "Plata y revolución tecnológica en la América virreinal." In Historia de las Ciencias

y de las Técnicas, edited by Luis Español González , José Javier Escribano Benito and
María Ángeles Martínez García 79-104. Spain: Universidad de la Rioja, 2004.
———. "Primeros beneficios de la plata por amalgamación en la América colonial (1565-
1600)." In Minería y metalurgia. Intercambio tecnológico y cultural entre América y
Europa durante el período colonial español, edited by Manuel Castillo Martos 375-
406. Sevilla, Bogota: Muñoz Montoya y Montraveta Editores, 1994.
Castillo Martos, Manuel, and M. F. Lang. Metales preciosos--unión de dos mundos :
tecnología, comercio y política de la minería y metarlurgia iberoamericana [in
Spanish]. Sevilla: Muñoz Moya y Montraveta Editores, 1995.
Chaunu, Pierre. Seville et l’Atlantique (1504-1650). La Conjoncture Paris: SEVPEN, 1959.
Cheilletz, Alain, Gilles Levresse, Dominique Gasquet, Moulay Azizi-Samir, Rachid Zyadi,
Douglas A Archibald, and Edward Farrar. "The Giant Imiter Silver Deposit:
Neoproterozoic Epithermal Mineralization in the Anti-Atlas, Morocco." Mineralium
Deposita 37, no. 8 (2002): 772-781.
Chesterton, G.K. Father Brown, Selected Stories London: Oxford University Press, 1961.
Cobb, Gwendolyn Ballantine. "Supply and Transportation for the Potosí Mines.". The Hispanic
Cobo, Bernabe (S.J.) Historia del Nuevo Mundo. Biblioteca de Autores Espanoles. edited by
Francisco (S.J.) Mateos. Vol. 91, Madrid: Atlas, 1964.
Coll-Hurtado, Atlántida, María Teresa Sánchez-Salazar, and Josefina Morales. La minería en
México: geografía, historia, economía y medio ambiente Mexico: UNAM, 2002.
Collins, Henry F. The Metallurgy of Lead & Silver London: Griffin & Co., 1899.
Constable, Olivia Remie. Trade and Traders in Muslim Spain : the Commercial Realignment
of the Iberian Peninsula, 900-1500 [in English]. Cambridge; New York: Cambridge
Contreras, Manuel Maria. "Empleo de los ensayes de pella y de residuos para determinar los
adelantos y fin de la amalgamacion de la plata en el beneficio de patio.". In Historia de
la ciencia en México: Estudios y textos., edited by Elias Trabulse 184-196. Mexico:
Conacyt y Fondo de Cultura Economica, 2003.
Cooke, C.A., P.H. Balcom, C. Kerfoot, M.B. Abbott, and A.P. Wolfe. "Pre-Colombian
Mercury Pollution Associated with the Smelting of Argentiferous Ores in the Bolivian
Andes." AMBIO: A Journal of the Human Environment 40, no. 1 (2011): 18-25.
Cooke, Colin A., Prentiss H. Balcom, Harald Biester, and Alexander P. Wolfe. "Over three
millennia of mercury pollution in the Peruvian Andes." Proceedings of the National
Academy of Sciences 106, no. 22 (June 2, 2009 2009): 8830-8834.
Corbeil, Laurent. "Identities in Motion. The Formation of a Plural Indio Society in Early San
Luis Potosi, New Spain, 1591-1630 ", McGill, 2014.
Cotton, S.A. Chemistry of Precious Metals Bristol: Blackie Academic & Professional, 1997.
Couturier, Edith Boorstein. The Silver King : the Remarkable Life of the Count of Regla in
Colonial Mexico Albuquerque: University of New Mexico Press, 2003.
Cox, Dennis P. "Descriptive Model of Polymetallic Veins." In U.S. Geological Survey Bulletin
1693, edited by Dennis P. Cox and Donald A. Singer 125-129. Washington: U.S.
Geological Survey, 1987.
Craddock, P. T. Early Metal Mining and Production Washington, D.C.: Smithsonian
Institution Press, 1995.
Cranstone, David. "Excavations at Old Gang Smeltmills: an Interim Report." In Boles and
Smeltmills: Report of a seminar on the History and Archaeology of Lead Smelting,
edited by Lynn Willies and David Cranstone, 28-31. Reeth, Yorkshire: Historical
Metallurgical Society, 1992.
627
Cruz, Pablo Jose, and Jean-Joinville Vacher, eds. Mina y metalurgia en los Andes del sur:
desde la época prehispánica hasta el siglo XVII La Paz: Institut de Recherche pour le
Development / Instituto Francés de Estudios Andinos, 2008.
Cubillo Moreno, G. "Los dominios de la plata: el precio del auge, el peso del poder.
Empresarios y trabajadores en las minas de Pachuca y Zimapán, 1552-1620." Col.
Divulgación, Serie Historia, Instituto Nacional de Antropología e Historia/Consejo
Nacional para la Cultura y las Artes, México, 1991.
Cunill Grau, Pedro. "El paisaje andino: Punas, Salares y Cerros." In Potosí: plata para Europa
edited by José Villa Rodríguez, 75-103. Sevilla: Universidad de Sevilla, 2000.
Cunningham, C.G., H. Aparicio, F. Murillo, N. Jimenez, E.H. McKee, G.E. Ericksen, and F.
Tavera. "Relationship between the Porco, Bolivia, Ag-Zn-Pb-Sn deposit and the Porco
caldera." Economic Geology 89, no. 8 (1994): 1833-1841.
Cunningham, CG, RE Zartman, EH McKee, RO Rye, CW Naeser, O. Sanjinés V, GE Ericksen,
and F. Tavera V. "The Age and Thermal History of Cerro Rico de Potosi, Bolivia."
Mineralium Deposita 31 no. 5 (1996): 374-385.
d'Oliveira, M.V.N., E.C. Alvarado, J.C. Santos, and J.A. Carvalho Jr. "Forest Natural
Regeneration and Biomass Production after Slash and Burn in a Seasonally Dry Forest
in the Southern Brazilian Amazon." Forest Ecology and Management (2011).
Dahlgren, Charles Bunker. Historic Mines of Mexico. A Review of the Mines of that Republic
for the Past Three Centuries New York: Printed for the author, 1883.
Day, John. "The Great Bullion Famine of the Fifteenth Century." Past & Present, no. 79
(1978): 3-54.
de Gamboa, Francisco Xavier. Comentarios a las ordenanzas de minas dedicados al católico
rey nuestro señor Carlos 111. Madrid, Oficina de Joaquín Ibarra. 1761.
De Mesa, José, and Teresa Gisbert. "Oruro. Origen de una villa minera.". In La minería hispana
e iberoamericana. Ponencias del I coloquio internacional sobre historia de la minería.,
559-589. León: Cátedra de San Isidro, 1970.
de Oñate, J. Nuevas leyes de las minas de España: 1625 edición de Juan de Oñate: con tratado
de re Metalica de Juan de Oñate Santa Fe, New Mexico: Sunstone Press, 1998.
de Sarria, Francisco Javier. Ensayo de metalurgia Mexico: D. Felipe de Zuniga y Ontiveros,
1784.
Díaz del Castillo, Bernal. Historia verdadera de la conquista de la Nueva España : Manuscrito
"Guatemala" Mexico: Colegio de México : Universidad Nacional Autónoma de
México, 2005.
Dickinson, William R. "Anatomy and Global Context of the North American Cordillera." In
Backbone of the Americas : Shallow Subduction, Plateau Uplift, and Ridge and Terrane
Collision, edited by Suzanne Mahlburg Kay, Víctor A. Ramos and William R.
Dickinson 1-29. Boulder, Colo.: Geological Society of America, 2009.
Diringer, Sarah, William Pan, Beth Feingold, Ernesto Ortiz, and Heileen Hsu-Kim. "Mercury
Biogeochemistry and Artisanal and Small-Scale Gold Mining in Madre de Dios, Peru."
Poster, 11th International Conference on Mercury as a Global Pollutant, Edinburgh
2013.
Disney, A. R. A History of Portugal and the Portuguese Empire : From Beginnings to 1807
Vol. II New York: Cambridge University Press, 2009.
Dobado, Rafael , and Gustavo A. Marrero. "The Role of the Spanish Imperial State in the
Mining-led Growth of Bourbon Mexico's Economy." [In English]. The Economic
History Review 64, no. 3 (2011): 855-884.
Dominguez de la Fuente, Manuel Jose. Leal Informe Politico-Legal (Observaciones sobre la
maniobra de las minas, hechas en el Real de Guanajuato en 1774) Guanajuato, Mexico:
Ediciones La Rana, 1999.
628
Drelichman, Mauricio, and Hans-Joachim Voth. "Institutions and the Resource Curse in Early
Modern Spain." In Institutions and Economic Performance edited by Elhanan Helpman
120-147. Cambridge, Mass.: Harvard University Press, 2008.
Duport, Saint Clair. De la production des métaux précieux au Mexique, considérée dans ses
rapports avec la géologie, la métallurgie et l'économie politique Paris: Firmin Didot
Frères, 1843.
Egleston, Thomas. The Metallurgy of Silver, Gold, and Mercury in the United States London;
New York: J. Wiley & Sons, 1887.
Eisler, Ronald. Handbook of Chemical Risk Assesment: Health Hazards to Humans, Plants,
and Animals. Volume 1: Metals Boca Raton, FL.: Lewis Publishers, 2000.
Eissler, Manuel. The Metallurgy of Argentiferous Lead London: Crosby Lockwood & Son,
1891.
Ekino, Shigeo, Mari Susa, Tadashi Ninomiya, Keiko Imamura, and Toshinori Kitamura.
"Minamata Disease Revisited: An Update on the Acute and Chronic Manifestations of
Methyl Mercury Poisoning." Journal of the Neurological Sciences 262, no. 1-2 (2007):
131-144.
Elliott, John Huxtable. Empires of the Atlantic World : Britain and Spain in America, 1492-
1830 New Haven: Yale University Press, 2006.
Emmons, William Harvey. The Enrichment of Ore Deposits Vol. 625 Washington: United
States Geological Survey, 1917.
Encyclopaedia Britannica. 15th ed. Chicago: Encyclopaedia Britannica, Inc., 1992.
English, Henry. A General Guide to the Companies Formed for Working Foreign Mines with
their Prospectuses, Amount of Capital, Number of Shares, Names of Directors, &c
London: Boosey & Sons, 1825.
Engstrom, Daniel R., William F. Fitzgerald, Colin A. Cooke, Carl H. Lamborg, Paul E.
Drevnick, Edward B. Swain, Steven J. Balogh, and Prentiss H. Balcom. "Atmospheric
Hg Emissions from Preindustrial Gold and Silver Extraction in the Americas: A
Reevaluation from Lake-Sediment Archives." Environmental Science & Technology
48, no. 12 (2014): 6533-6543.
Ercker, L. Treatise on Ores and Assaying. Translated by Anneliese Grünhaldt Sisco and Cyril
Stanley Smith Chicago: University of Chicago Press, 1951.
Escalante, Amor Mildred. "Redes familiares empresariales en la ciudad de Guanajuato,
México, 1877-1911." In XXI Jornadas de Historia Económica Caseros, Argentina:
Asociación Argentina de Historia Económica, 2008.
Escamilla Gonzalez, Francisco Omar. "Ilustración alemana y ciencia novohispana: la biblioteca
de Fausto de Elhuyar." In Alemania y México: Percepciones mutuas en impresos, siglos
XVI-XVIII, edited by Horst Pietschmann, Manuel Ramos Medina, Maria Cristina
Torales Pacheco and Karl Kohut, 399-451. Ciudad de Mexico: Universidad
Iberoamericana, 2005.
Eslava Galán, Juan. Cinco tratados españoles de alquimia Madrid: Tecnos, 1987.
Espinosa Pitman, Alejandro. José Antonio de Villaseñor y Sánchez, 1703-1759 San Luis Potosí,
S.L.P., México: Universidad Autónoma de San Luis Potosí, 2003.
Evans, Anthony M. Ore Geology and Industrial Minerals : An Introduction Oxford; Boston:
Blackwell Scientific Publications, 1992.
Ewald, Ursula. The Mexican Salt Industry, 1560-1980 : A Study in Change Stuttgart; New
York: G. Fischer, 1985.
Fabbrizzi, L. "Communicating about Matter with Symbols: Evolving from Alchemy to
Chemistry." Journal of Chemical Education 85, no. 11 (2008): 1501.
Fabry, José Antonio. Compendiosa demostracion de los crecidos adelantamientos, que pudiera
lograr la real haciencda de su Magestad mediante la rebaja en el precio del azogue
629
que se consume para el laborio de las minas de este reyno ... con una previa
impugnacion à las reflexiones del contador Joseph de Villa-señor y Sanchez ... Añadese
un breve modo de reducir, ligar, y alear el oro, y la plata à la ley de 22. quilates
Mexico: Impressa por la viuda de J.B. de Hogal, 1743.
Fairhall, Lawrence T. "Inorganic Industrial Hazards." Physiological Reviews 25, no. 1 (1945):
182-202.
Fernández Alcaide, Marta. Cartas de particulares en Indias del siglo XVI : edición y estudio
discursivo Madrid; Frankfurt am Main: Iberoamericana ; Vervuert, 2009.
Fernández de Oviedo y Valdés, Gonzalo. Historia general y natural de las Indias Madrid:
Ediciones Atlas, 1959.
Fernández de Pinedo y Fernández, E. "Influencias recíprocas de las técnicas extractivas entre
la minería vasca y la americana en la Edad Moderna." Areas, no. 16 (1994): 35-46.
Fernandez del Castillo, Francisco. "Algunos documentos nuevos sobre Bartolome de Medina."
Memorias de la Sociedad Alzate 47 (1927).
Fernández Pérez, Joaquín. "La amalgamación de los minerales de plata." In El oro y la plata
de las Indias en la época de los Austrias, edited by Concepción Lopezosa Aparicio 141-
155. Madrid: Fundacion ICO, 1999.
Findlay, Ronald, and Kevin H. O’Rourke. Power and Plenty. Trade, War and the World
Economy in the Second Millenium Princeton: Princeton University Press, 2007.
Fisher, John. "Silver Production in the Viceroyalty of Peru, 1776-1824." The Hispanic
Flores Clair, Eduardo, Cuauhtémoc Velasco Avila, and Elia Ramírez Bautista. Estadísticas
mineras de México en el siglo XIX. Vol. II, México: Instituto Nacional de Antropología
e Historia, 1985.
Flores, T. Étude minier du district de Zacatecas Guía de excursiones del X Congreso
Internacional de Geología México. Vol. 17 Xalapa: Institut Geologique National, 1905.
Flores, Teodoro. Yacimientos minerales de la República Mexicana, con algunos datos relativos
a su producción México: Instituto Geológico de México, 1933.
Flynn, Dennis, and Arturo Giráldez. "Latin American Silver and the Early Globalization of
World Trade." In National Identities and Sociopolitical Changes in Latin America 140–
159. London and New York: Routledge, 2001.
Flynn, Dennis O, and Arturo Giraldez. "Cycles of Silver: Global Economic Unity through the
Mid-Eighteenth Century." Journal of World History 13 (2002): 391-427.
Flynn, Dennis O., and Arturo Giraldez. "Arbitrage, China, and World Trade in the early modern
period." Journal of the Economic and Social History of the Orient 38 (1995): 429-448.
Fortey, Richard A. Earth : An Intimate History London: Harper Perennial, 2005.
Fox, K. V. "Pedro Muñiz, Dean of Lima, and the Indian Labor Question (1603)." The Hispanic
Frank, Andre Gunder. ReOrient: Global Economy in the Asian Age Berkeley: University of
California Press, 1998.
Freestone, I C, A P Middleton, P T Craddock, L K Gurjar, and D R Hook. "Role of Materials
Analysis in the Reconstruction of Early Metal Extraction Technology: Zinc and Silver-
Lead Smelting at Zawar, Rajasthan." Materials Research Society Symposia
Proceedings 185 (1990): 617-626.
Friede, Juan. "La introducción de mineros alemanes en América por la compañía Welser de
Augsburgo." Revista de Historia de América, no. 51 (1961): 99-104.
Galbraith, John Kenneth. Money, Whence It Came, Where It Went London: Penguin Books
Ltd., 1995.
Gale, N. H., and Z. A. Stos-Gale. "Cycladic Lead and Silver Metallurgy." The Annual of the
British School at Athens 76 (1981): 169-224.
630
Galeano, Eduardo. Las venas abiertas de América Latina Buenos Aires: Siglo Veintiuno, 1982.
Galindo y R, José J. El distrito minero Pachuca-Real del Monte [Pachuca?]: Cia. de Real del
Monte y Pachuca, 1957.
Galvan Arellano, Alejandro. Arquitectura y urbanismo de la ciudad de San Luis Potosí en el
siglo XVII San Luis Potosí: Editorial Universitaria Potosina, 1999.
Garcés y Eguía, Joseph. Nueva teórica y práctica del beneficio de los metales de oro y plata
por fundición y amalgamación que de órden del rey nuestro Señor Don Carlos Quarto
Mexico: D. Mariano de Zúñiga y Ontiveros, 1802.
Garcia de Llanos. Diccionario y maneras de hablar que se usan en las minas y sus labores en
los Ingenios y beneficios de los metales (1609) La Paz, Bolivia: MUSEF, 1983.
García Mendoza, Jaime. "La administración de las minas de plata y haciendas de beneficio de
la familia Sandoval en Taxco (1562-1564)." In La plata en Iberoamérica: Siglos XVI
al XIX, edited by Jesus Paniagua Pérez and Nuria Salazar Simarro 39-59. Leon:
Universidad de León, 2008.
Garner, Richard L. "Price trends in eighteenth-century Mexico." The Hispanic American
Historical Review 65, no. 2 (1985): 279-325.
Garner, Richard L. "Long-Term Silver Mining Trends in Spanish America: A Comparative
Analysis of Peru and Mexico." The American Historical Review 93, no. 4 (1988): 898-
935.
Garner, Richard L., and Spiro E. Stefanou. Economic Growth and Change in Bourbon Mexico
Gainesville: University Press of Florida, 1993.
Gasparrini, Claudia. "The Mineralogy of Silver and its Significance in Metal Extraction." CIM
Bulletin 77, no. 866 (1984): 99-110.
Gill, Michael C. . "Analysis of Lead Slags." In Boles and Smeltmills Seminar, edited by Lynn
Willies and David Cranstone 51-53. Reeth, Yorkshire: Historical Society, Ltd., 1992.
Gilliam, Albert M. Travels Over the Table Lands and Cordilleras of Mexico, During the Years
1843 and 44 Philadelphia: John W Moore, 1846.
Global Forest Resources Assessment 2010. Rome: Food and Agriculture Organization of the
United Nations, 2010.
Goldschmidt, V. M. Geochemistry Oxford: Clarendon Press, 1954.
Gómez de Cervantes, Gonzalo. La vida económica y social de Nueva España al finalizar el
siglo XVI Mexico: Antigua Librería Robredo de José Porrúa e hijos, 1944.
González Reyna, Jenaro. Minería y Riqueza de México. Monografías Industriales México:
Banco de México, 1944.
González, Tomas. Noticia histórica documentada de las célebres minas de Guadalcanal, desde
su descubrimiento en el año 1555, hasta que dejaron de labrarse por cuenta de la Real
Hacienda. Vol. II Madrid 1831.
González Valdez, Eduardo, Ezequiel González Reyes, Carlos Bedolla Cedeño, Edith Lorena
Arrollo Ordaz, and Eduardo Manzanares Acuña "Niveles de plomo en sangre y factores
de riesgo por envenenamiento de plomo en niños mexicanos ". Revista de la Facultad
de Ingeniería de la Universidad de Antioquia, no. 43 (2008): 114-119.
Goodarzi, Fariborz, Hamed Sanei, Marcel Labonté, and William F. Duncan. "Sources of Lead
and Zinc Associated with Metal Smelting Activities in the Trail Area, British Columbia,
Canada ". Journal of Environmental Monitoring 4, no. 3 (2002): 400-407.
Goodman, David. "Philip II's Patronage of Science and Engineering." The British Journal for
the History of Science 16, no. 1 (1983): 49-66.
Graybeal, F. T., and D.M. Smith Jr. "Regional Distribution of Silver Deposits on the Pacific
Rim." In Silver - Exploration, Mining and Treatment, 3-10. Mexico City: Institution of
Mining and Metallurgy, 1989.
631
Graybeal, Frederick T., Douglas M. Smith, and Peter G. Vikre. "The Geology of Silver
Deposits." In Handbook of Strata-Bound and Stratiform Ore Deposits. Part IV, edited
by K. H. Wolf 1-184. Amsterdam: Elsevier 1986.
Greene, Graham. The Lawless Roads London: Longmans, 1939.
Guerrero, Saúl. "Chemistry as a Tool for Historical Research: Estimating the Contraband of
Silver from Potosí and Oruro, 1576-1650." Bulletin for the History of Chemistry 37, no.
2 (2012): 72-81.
———. "Chemistry as a Tool for Historical Research: Identifying Paths of Historical Mercury
Pollution in the Hispanic New World." Bulletin for the History of Chemistry 37, no. 2
(2012): 61-71.
Hagan, Nicole, Nicholas Robins, Heileen Hsu-Kim, Susan Halabi, Mark Morris, George
Woodall, Tong Zhang, et al. "Estimating historical atmospheric mercury concentrations
from silver mining and their legacies in present-day surface soil in Potosí, Bolivia."
Atmospheric Environment 45, no. 40 (2011): 7619-7626.
Hall, Cris M., Pablo L. Higueras, Stephen E. Kesler, Rosario Lunar, Hailing Dong, and Alex
N. Halliday. "Dating of Alteration Episodes Related to Mercury Mineralization in the
Almadén district, Spain." Earth and Planetary Science Letters 148 no. 1-2 (1997): 287-
298.
Halleux, Robert. "La nature et la formation des métaux selon Agricola et ses contemporains."
Revue d'histoire des sciences (1974): 211-222.
Hardesty, Donald L. Mining Archaeology in the American West: A View from the Silver State
Lincoln: University of Nebraska Press, 2010.
Hausberger, Bernd. "El universalismo científico del Barón Ignaz von Born y la transferencia
de tecnología minera entre Hispano américa y Alemania a finales del siglo XVIII."
Historia Mexicana 59, no. 2 (2009): 605-668.
———. La Nueva España y sus metales preciosos Frankfurt am Main: Vervuert, 1997.
———. "Una iniciativa ecológica contra la industria minera en Chihuahua (1732)." Estudios
de Historia Novohispana 13, no. 13 (1992): 115-34.
Haven, Gilbert. Our Next-Door Neighbor: a Winter in Mexico New York: Harper & Brothers,
1875.
Heers, Jacques. Gênes au XVe siècle, activité économique et problèmes sociaux Paris:
S.E.V.P.E.N., 1961.
Hermosa, Francisco de Paula. Manual de Laboreo de Minas Besanzon: Libreria de Rosa Bouret
y Cia, 1857.
Hernberg, Sven. "Lead Poisoning in a Historical Perspective." American Journal of Industrial
Medicine 38, no. 3 (2000): 244-254.
Héron de Villefosse, Antoine-Marie. De la richesse minérale considérations sur les mines,
usines et salines des différens états, et particulièrement du Royaume de Westphalie,
pris pour terme de comparaison Paris: Levrault, 1810.
Herrera Canales, Ines. "El auge de la plateria potosina en la segunda mitad del siglo XVIII, al
argento vivo." In Ophir en las Indias. Estudios sobre la plata americana. Siglos XVI-
XIX, edited by Jesus Paniagua Perez and Nuria Salazar Simarro 109-123. Leon:
Universidad de Leon, 2010.
Herrera Canales, Inés. "El método de refinación con azogue en la minería potosina colonial:
del fuego al cazo ". In La plata en Iberoamérica: Siglos XVI al XIX, edited by Jesus
Paniagua Pérez and Nuria Salazar Simarro 61-68. Leon: Universidad de León, 2008.
Herrera Canales, Inés "Mercurio para refinar la plata mexicana en el siglo XIX." Historia
Mexicana 40, no. 1 (1990): 27-51.
Herrera Canales, Inés, and Rina Ortiz Peralta. "Mexican Mining Heritage-The Real Del Monte
Site." CRM (Washington) 21, no. 7 (1998): 19-22.
632
Herrera Canales, Ines, Cuahtemoc Velasco Avila, and Eduardo Flores Clair. Etnia y clase, los
trabajadores ingleses de la Compañia del Monte y Pachuca, 1824-1906 Mexico:
INAH, 1981.
Higueras, P., W. Llanos, M. E. García, R. Millán, and C. Serrano. "Mercury vapor emissions
from the Ingenios in Potosí (Bolivia)." Journal of Geochemical Exploration 116 (2012):
1-7.
Higueras, P., W. Llanos, M.E. García, R. Millán, C. Serrano, and G.N. María. "Multielemental
pollution of soils at the Ingenios, decommissioned mineralurgical sites in Potosí
(Bolivia)." (2010).
Homma, Katsunori. "Experimental Study for Preparating Metal Fumes." Industrial Health 4,
no. 3 (1966): 129-137.
Humboldt, Alexander von. Essai politique sur le royaume de la Nouvelle-Espagne Paris: Chez
F. Schoell, 1811.
Hylander, Lars D., and Markus Meili. "500 Years of Mercury Production: Global Annual
Inventory by Region until 2000 and Associated Emissions." Science of The Total
Environment 304, no. 1–3 (2003): 13-27.
Ibarra, Juan de. "Suma de lo que el licenciado J de Ybarra como procurador general de la villa
de Potosi pide [14 Agosto 1617]." Madrid, , 1617.
Iles, Malvern Wells. Lead-Smelting The Construction, Equipment, and Operation of Lead
Blast-Furnaces, and Observations on the Influence of Metallic Elements on Slags and
the Scientific Handling of Smoke New York; London: J. Wiley & Sons; Chapman &
Hall, 1904.
Jacob, William. An Historical Inquiry into the Production and Consumption of the Precious
Metals. Vol. 1 New York: Augustus M. Kelly, 1968. 1831.
Jasso-Pineda, Yolanda, Guillermo Espinosa-Reyes, Donají González-Mille, Israel Razo-Soto,
Leticia Carrizales, Arturo Torres-Dosal, Jesús Mejía-Saavedra, et al. "An Integrated
Health Risk Assessment Approach to the Study of Mining Sites Contaminated With
Arsenic and Lead." Integrated Environmental Assessment and Management 3, no. 3
(2007): 344-350.
Jimenez de la Espada, Marcos Relaciones Geograficas de Indias - Peru II. Biblioteca de
Autores Espanoles. edited by Jose Urbano Martinez Carrera Vol. 184, Madrid: Atlas,
1965 (1588).
Johnson, D.A., and K. Whittle. "The Chemistry of the Hispanic-American Amalgamation
Process." J. Chem. Soc., Dalton Trans., no. 23 (1999): 4239-4243.
Jones, G. D. B. "The Roman Mines at Riotinto." The Journal of Roman Studies 70 (1980): 146-
165.
Kaplan, Jed O., Kristen M. Krumhardt, and Niklaus Zimmermann. "The prehistoric and
preindustrial deforestation of Europe." Quaternary Science Reviews 28, no. 27–28
(2009): 3016-3034.
Karpenko, V.Í. "The Oldest Alchemical Manuscript in the Czech Language." Ambix 37, no. 2
(1990): 61-73.
Kay, Suzanne Mahlburg, Víctor A. Ramos, and William R. Dickinson. Backbone of the
Americas : Shallow Subduction, Plateau Uplift, and Ridge and Terrane Collision [in
English]. Boulder, Colo.: Geological Society of America, 2009.
Kellenbenz, Hermann. "Final Remarks: Production and Trade of Gold, Silver, Copper and Lead
from 1450 to 1750." In Precious Metals in the Age of Expansion: Papers of the XIVth
International Congress of the Historical Sciences, edited by Hermann Kellenbenz 307-
361. Stuttgart: Klett-Cotta, 1981.
633
———. "Los Fugger en España en la época de Felipe II Fue un buen negocio el arrendamiento
de los Maestrazgos después de 1562?". In Dinero y crédito (Siglos XVI al XIX), edited
by Alfonso Otazu, 19-36. Madrid-Villalba-Segovia: Banco Urquijo 1977.
Kerl, Bruno, William Crookes, and Ernst Otto Röhrig. A Practical Treatise on Metallurgy
Adapted from the Last German Edition of Prof. Kerl's Metallurgy London: Longmans,
Green, and Co., 1868.
King, Peter. "The Production and Consumption of Bar Iron in Early Modern England and
Wales." The Economic History Review 58 no. 1 (2005): 1-33.
Klein, Herbert S. The American Finances of the Spanish Empire : Royal Income and
Expenditures in Colonial Mexico, Peru, and Bolivia, 1680-1809 Albuquerque:
University of New Mexico Press, 1998.
Kobal, Alfred Bogomir, and Darja Kobal Grum. "Scopoli's Work in the Field of Mercurialism
in Light of Today's Knowledge: Past and Present Perspectives ". American Journal of
Industrial Medicine 53, no. 5 (2010): 535-547.
Kobata, Atsushi. "The Production and Uses of Gold and Silver in Sixteenth- and Seventeenth-
Century Japan." The Economic History Review 18, no. 2 (1965): 245-266.
Kollonitz, Paula. Un viaje a México en 1864 México: Secretaría de Educación Pública, 1976.
Körner, Christian. "A Re-Assessment of High Elevation Treeline Positions and Their
Explanation." Oecologia 115, no. 4 (1998): 445-459.
Kuhn, Thomas S. The Structure of Scientific Revolutions Chicago: University of Chicago Press,
1965.
Kunz, Federico. "Evolucion historica del regimen legal del beneficio de minerales en Mexico
". In Mineria Regional Mexicana. Primera reunion de historiadores de la mineria
latinoamericana, edited by Dolores Avila Herrera and Rina Ortiz 65-77. Mexico:
Instituto Nacional de Antropologia e Historia, 1994.
Küstel, G. Roasting of Gold and Silver Ores: and the Extraction of Their Respective Metals
Without Quicksilver AJ Leary, stationer and printer, 1880.
Lacueva Muñoz, Jaime J. La plata del rey y de sus vasallos : minería y metalurgia en México
(siglos XVI y XVII) Sevilla: Consejo Superior de Investigaciones Científicas, Escuela
Superior de Estudios Hispano-Americanos; Universidad de Sevilla; Diputación de
Sevilla, 2010.
———. "Nueva Vizacaya y sus yacimientos minerales hasta el descubrimiento de San José del
Parral." In Ophir en las Indias. Estudios sobre la plata americana. Siglos XVI-XIX,
edited by Jesús Paniagua Pérez and Nuria Salazar Simarro 89-108. León: Universidad
de León, 2010.
Ladd, Doris M. The Making of a Strike : Mexican Silver Workers' Struggles in Real del Monte,
1766-1775 Lincoln: University of Nebraska Press, 1988.
Lamborn, Robert H. The Metallurgy of Silver and Lead : A Description of the Ores; their Assay
and Treatment, and Valuable Constituents London: C. Lockwood, 1878.
Lang, M.F. El monopolio estatal del mercurio en el México colonial (1550-1710) Translated
by R.C. Gómez Mexico: Fondo de Cultura Económica, 1977.
Lang, Mervyn F. "Amalgamacion y fundicion en la mineria americana.". In Estudios de
historia de las técnicas, la arqueología industrial y las Ciencias, edited by Juan Luis
García Hourcade, Juan M. Moreno Yuste and Gloria Ruiz Hernández, 671-677.
Salamanca: Juanta de Castilla y Leon, 1998.
———. "Azoguería y amalgamación." Llull 22, no. 45 (1999): 655-674.
———. "Silver Refining Technology in Spanish America (patio y fundición) " In 5th
International Mining History Congress, edited by James E. Fell, P. D. Nicolaou and G.
D. Xydous 136-145. Milos Island: Milos Conference Center-George Eliopoulos, 2001.
634
Lara Meza, Ada Marina. Haciendas de beneficio de Guanajuato. Tecnología y usos del suelo
1770-1780 Guanajuato: Presidencia Municipal de Guanajuato; Dirección de Cultura y
Educación, 2001.
Laur, M.P. "De la metallurgie de l'argent au Mexique." Annales des Mines, 6th series, 20
(1871): 38-317.
Laznicka, Peter. "Discovery of Giant Metal Deposits and Districts." In Proceedings of the 30th
International Geology Congress: Energy and Mineral Resources for 21th Century:
Geology of Mineral Economics edited by Pei Rongfu, 355-366: VSP, 1997.
———. Giant Metallic Deposits : Future Sources of Industrial Metals Berlin: Springer, 2006.
Lechtman, Heather. "A Metallurgical Site Survey in the Peruvian Andes." Journal of Field
Archaeology 3, no. 1 (1976): 1-42.
Lechtman, Heather, Pablo Cruz, Andrew Macfarlane, and Sidney Carter. "Procesamiento de
metales durante el Horizonte Medio en el Altiplano Surandino (Escaramayu, Pulacayo,
Potosi, Bolivia).". Boletín del Museo Chileno de Arte Precolombino 15 (2010): 9-27.
Lehmann, B., A. Dietrich, and A. Wallianos. "From Rocks to Ore." International Journal of
Earth Sciences 89, no. 2 (2000): 284-294.
Leonard, Irving A. "Mateo Alemán in Mexico: A Document." Hispanic Review 17, no. 4
(1949): 316-330.
Linden, S.J. "Alchemy and Eschatology in Seventeenth-Century Poetry." Ambix 31, no. 3
(1984): 102-124.
———. "Jonson and Sendivogius: Some New Light on Mercury Vindicated from the
Alchemists at Court." Ambix 24, no. 1 (1977): 39-54.
Ljung, Anders, and Anders Nordin. "Theoretical Feasibility for Ecological Biomass Ash
Recirculation: Chemical Equilibrium Behavior of Nutrient Elements and Heavy Metals
During Combustion." Environmental Science & Technology 31, no. 9 (1997): 2499-
2503.
Lohmann Villena, Guillermo. Las minas de Huancavelica en los siglos XVI y XVII Sevilla:
Impr. de la Escuela de Estudios Hispano-Americanos, 1949.
Lopez Morell, Miguel A., and Jose M. O'Kean. "Seeking out and building monopolies,
Rothschild strategies in non ferrous metals international markets (1830-1940)." In 14th
Conference of the European Business History Association Glasgow, 2010.
Löschner, Renate Moyssén , Xavier Echeverría, and Mario de la Torre. El México luminoso
de Rugendas Mexico City: Cartón y Papel de México, 1985.
Lynch, Martin. Mining in World History London: Reaktion Books, 2002.
Lyon, G. F. Journal of a Residence and Tour in the Republic of Mexico in the Year 1826 with
Some Account of the Mines of that Country London: J. Murray, 1828.
MacDonald, Bernard. "Old Mexican Methods ". Mining and Scientific Press (1907): 123-126.
Magnus, Albertus. Book of minerals. Translated by Dorothy Wyckoff Oxford: Clarendon
Press, 1967.
Magos, Laszlo, and Thomas W. Clarkson. "Overview of the Clinical Toxicity of Mercury."
Annals of Clinical Biochemistry 43 (2006): 257-68.
Manzanares Acuña , Eduardo, Héctor Rene Vega Carrillo, Martha Celia Escobar León,
Consuelo Letechipía de León, Laura J. Guzmán Enríquez, Víctor Martin Hernández
Dávila, and Miguel Ángel Salas Luévano. "Evaluación de riesgos ambientales por
plomo en la población de Vetagrande, Zacatecas." 1-50. Zacatecas: Universidad
Autónoma de Zacatecas, Unidad de Estudios Nucleares, 2005.
Mapp, Paul W. . The Elusive West and the Contest for Empire, 1713-1763 Chapel Hill;
Williamsburg, Va.: University of North Carolina Press, 2011.
Marichal, Carlos. Bankruptcy of Empire : Mexican Silver and the Wars Between Spain, Britain,
and France, 1760-1810 [in English]. New York: Cambridge University Press, 2007.
635
Martin, Percy F. Mexico's Treasure House (Guanajuato) New York: The Cheltenham Press,
1906.
———. Mexico's Treasure House (Guanajuato) An Illustrated and Descriptive Account of the
Mines and their Operations in 1906 New York: The Cheltenham press, 1906.
Martín Torres, Eugenio. El beneficio de la plata en Guanajuato, 1686-1740 Guanajuato:
Presidencia Municipal de Guanajuato, 2001.
Martinez Cortiza A, García-Rodeja E., Pontevedra Pombal X., Nóvoa Muñoz J. C., Weiss D.,
and Cheburkin A. "Atmospheric Pb Deposition in Spain During the Last 4600 Years
Recorded by Two Ombrotrophic Peat Bogs and Implications for the Use of Peat as
Archive." The Science of the Total Environment 292 (2002): 33-44.
Martínez Martínez, María del Carmen. "Plata y mineria en la correspondencia privada de las
Indias." In Ophir en las Indias. Estudios sobre la plata americana. Siglos XVI-XIX,
edited by Jesus Paniagua Perez and Nuria Salazar Simarro 25-41. Leon: Universidad
de Leon, 2010.
Maskall, J. E., and I. Thornton. "Chemical Partitioning of Heavy Metals in Soils, Clays and
Rocks at Historical Lead Smelting Sites." Water, Air, and Soil Pollution 108, no. 3-4
(1998): 391-409.
Matilla Tascón, Antonio. Historia de las minas de Almadén, 1 (desde la época romana hasta
el año 1645) Madrid: Gráficas Osca, 1958.
———. Historia de las minas de Almadén, 2 (desde 1646 a 1709) Madrid: Instituto de Estudios
Fiscales, 1987.
Mendizábal, Miguel Othon. La minería y la metalurgia mexicanas (1520-1943) Mexico:
Centro de Estudios Históricos del Movimiento Obrero, 1980.
———. "Los minerales de Pachuca y Real del Monte en la época colonial." El Trimestre
Económico, no. 2 (1941): 253-309.
Menegus Borneman, Margarita. "Las comunidades productoras de sal y los mercados mineros:
los casos de Taxco y Temascaltepec." In Minería regional mexicana. Primera reunión
de historiadores de la minería latinoamericana, edited by Dolores Avila Herrera and
Rina Ortiz 21-31. Mexico: Instituto Nacional de Antropología e Historia, 1994.
Menéndez Navarro, Alfredo. Un mundo sin sol. Biblioteca Chronica Nova de Estudios
Historicos Granada: Universidad de Granada / Universidad de Castilla - La Mancha,
1996.
Merrill, Charles White. Summarized Data of Silver Production US Government Printing
Office, 1930.
Meyer, Michael C. Sherman William L. The Course of Mexican History New York: Oxford
Mitchell, A. H. G., and M. S. Garson. Mineral Deposits and Global Tectonic Settings London;
New York: Academic Press, 1981.
Molenda, D. "Silver Production in Poland, XVIth to XVIIIth Century ". In Hombres, Técnica,
Plata. Minería y sociedad en Europa y América, Siglos XVI-XIX edited by Julio
Sánchez Gómez and Guillermo Mira Delli-Zotti 23-31. Sevilla: Aconcagua Libros,
2000.
Molenda, Danuta. "La metallurgie du plomb en Pologne au moyen age et aux XVIe - XVIIIe
siecles. ." In Mines et métallurgie edited by Paul Benoit. Les chemins de la recherche,
45-57. Villeurbanne: Programme Rhône-Alpes recherches en sciences humaines, 1994.
Montejano y Aguiñaga, Rafael. El Real de Minas de la Purísima Concepción de los Catorce,
SLP San Luis Potosi: Editorial Universitaria Potosina, 1993.
Moran, Bruce T. Andreas Libavius and the Transformation of Alchemy : Separating Chemical
Cultures with Polemical Fire Sagamore Beach, MA: Science History
Publications/Watson Pub. International, 2007.
636
———. Distilling Knowledge : Alchemy, Chemistry, and the Scientific Revolution Cambridge,
Mass.: Harvard University Press, 2005.
———. Patronage and Institutions : Science, Technology, and Medicine at the European
Court, 1500-1750 [in English]. Rochester, NY, USA: Boydell Press, 1991.
Moreno y Castro, Juan. Arte o nuevo modo de beneficiar los metales de oro y plata Mexico:
Imprenta Biblioteca Mexicana, 1758.
Mosier, Dan L., Takeo Sate, Norman J. Page, Donald A. Singer, and Byron R. Berger.
"Descriptive Model of Creede Epithermal Veins." In U.S. Geological Survey Bulletin
1693, edited by Dennis P. Cox and Donald A. Singer 145-149. Washington: U.S.
Geological Survey, 1987.
Motten, Clement G. Mexican silver and the enlightenment Philadelphia: University of
Pennsylvania Press, 1950.
Munro, John H. "The Monetary Origins of the'price Revolution': South German Silver Mining,
Merchant Banking, and Venetian Commerce, 1470-1540." Department of Economics,
University of Toronto, 2003.
Myers, Norman Kent Jennifer. Perverse Subsidies : How Tax Dollars Can Undercut the
Environment and the Economy Washington, DC: Island Press, 2000.
Navas, Guadalupe. "Zacatecas a fin del siglo XIX." In Mineria Regional Mexicana. Primera
reunion de historiadores de la mineria latinoamericana. , edited by Dolores Avila
Herrera and Rina Ortiz 71-90. Mexico: Instituto Nacional de Antropologia e Historia,
1994.
Nef, J.U. "Silver Production in Central Europe, 1450-1618." The Journal of Political Economy
(1941): 575-591.
Nef, John U. The Conquest of the Material World Chicago: University of Chicago Press, 1964.
Newman, William .R. The 'Summa perfectionis' of Pseudo-Geber: A Critical Edition,
Translation and Study. Vol. 35 Leiden: Brill, 1991.
Newman, William Royall. Atoms and Alchemy : Chymistry and the Experimental Origins of
the Scientific Revolution Chicago: University of Chicago Press, 2006.
Newson, Linda A. "Silver Mining in Colonial Honduras." Revista de Historia de América, no.
97 (1984): 45-76.
Norris, J. "The Mineral Exhalation Theory of Metallogenesis in Pre-Modern Mineral Science."
Ambix 53, no. 1 (2006): 43-65.
Nriagu, J. "Legacy of Mercury Pollution." Nature 363 (1993): 589.
Nriagu, Jerome O. "Occupational Exposure to Lead in Ancient Times." Science of the Total
Environment 31, no. 2 (1983): 105-116.
———. "Tales Told in Lead." Science 281, no. 5383 (1998): 1622-1623.
Nummedal, Tara E. Alchemy and Authority in the Holy Roman Empire Chicago: University of
Chicago Press, 2007.
Ortiz Peralta, Rina. "Algunos aspectos del beneficio de minerales en el siglo XIX: el caso de
la Compañia de Real del Monte y Pachuca." In Hombres, Tecnica, Plata. Minería y
Sociedad en Europa y América, Siglos XVI - XIX., edited by Julio Sánchez Gómez and
Guillermo Mira Delli-Zotti 201-214. Sevilla: Aconcagua Libros, S.L., 2000.
———. "El beneficio de minerales en el siglo XIX: El caso de la Compañia Real del Monte."
Tzintzun : Revista de Estudios Historicos 14, no. Julio-Diciembre (1991): 72-85.
———. "El beneficio de minerales en el siglo XIX: el caso de la Compañía Real del Monte y
Pachuca." Historias 30 (1993): 51-60.
Otte, Enrique, and Guadalupe Albi Romero. Cartas privadas de emigrantes a Indias, 1540-
1616 México: Fondo de Cultura Económica, 1993.
The Oxford Dictionary of Quotations Oxford; New York: Oxford University Press, 1979.
637
Papadimitriou, George. "Mining and Metallurgical Activities in Ancient Laurium and its
Impact on the Golden Era of Athens " In 5th International Mining History Congress,
edited by James E. Fell, P. D. Nicolaou and G. D. Xydous 33-45. Milos Island: Milos
Conference Center-George Eliopoulos, 2001.
Paracelsus. The Hermetic and Alchemical Writings of Aureolus Philippus Theophrastus
Bombast, of Hohenheim, Called Paracelsus the Great : Now for the First Time
Faithfully Translated into English Translated by Arthur Waite 1976.
Pardo Hernández, Berenice, and Oscar Sánchez Rangel. Mineral de la Luz. La obra fotográfica
de John Horgan Jr. en México Guanajuato: Editorial La Rana, 2010.
Pelayo López, Francisco. "Las actividades mineras de J.C. Mutis y Juan José Elhuyar en Nueva
Granada." Revista de Indias 50, no. 189 (1990): 455-472.
Percy, John. The Metallurgy of Lead Including Desilverization and Cupellation London: J.
Murray, 1870.
———. Metallurgy. The Art of Extracting Metals from their Ores. Silver and Gold. Vol. I
London: J. Murray, 1880.
Pérez Bustamante, Ciriaco. "Las minas en los grandes geógrafos del periodo hispánico." In La
minería hispana e iberoamericana. Ponencias del I coloquio internacional sobre
historia de la minería 295-306. León: Cátedra de San Isidro, 1970.
Pérez Fernández, Isacio. El anónimo de Yucay frente a Bartolomé de Las Casas : estudio y
edición crítica del Parecer de Yucay, anónimo (Valle de Yucay, 16 de marzo de 1571)
Cuzco, Perú: Centro de Estudios Regionales Andinos "Bartolomé de Las Casas," 1995.
Pérez Luque, Rosa Alicia, and Rafael Tovar Rangel. La contabilidad de la Caja Real de
Guanajuato. Una aproximacion a su historia economica 1665-1816 Guanajuato,
Mexico: Universidad de Guanajuato, 2006.
Pérez Sáenz de Urturi, J.E. "La minería colonial americana bajo la dominación española."
Boletín Millares Carlo, no. 7 (1985): 53-120.
Pérez Sánchez, Alfonso E. , Benito Navarrete Prieto, and Gustavo Curiel. Tesoros del Museo
Soumaya de México : siglos XV-XIX Madrid: BBVA, 2004.
Phillips, John. Descriptive Notice of the Silver Mines and Amalgamation Process of Mexico.
Extracted from the Railway Register London: Pelham Richardson, 1846.
Phillips, John Arthur. The Mining and Metallurgy of Gold and Silver London: E. and F.N.
Spon, 1867.
Pique, Edward. A Practical Treatise on the Chemistry of Gold, Silver, Quicksilver and Lead,
Tracing the Crude Ores from the Mines Through the Various Mechanical and
Metallurgic Elaborations, Until the Pure Metal is Obtained San Francisco: Towne &
Bacon, printers, 1860.
Platt, T. "The Alchemy of Modernity. Alonso Barba's Copper Cauldrons and the Independence
of Bolivian Metallurgy (1790-1890)." Journal of Latin American Studies (2000): 1-54.
Pliny. The Elder Pliny's Chapters on Chemical Subjects [in English]. London: E. Arnold &
Co., 1929.
Plumlee, Geoffrey S, Kathleen S Smith, Byron R Berger, Nora Foley-Ayuso, and Douglas P
Klein. "Creede, Comstock, and Sado Epithermal Vein Deposits." In Preliminary
Compilation of Descriptive Geoenvironmental Mineral Deposit Models, Open File
Report OFR-95-0831 edited by E. du Bray 152-161. Denver, CO: US Geological
Survey, 1995.
Pohl, Walter. Economic Geology : Principles and Practice : Metals, Minerals,Coal and
Hydrocarbons - Introduction to Formation and Sustainable Exploitation of Mineral
Deposits Chichester, West Sussex; Hoboken, NJ: Wiley-Blackwell, 2011.
Pomeranz, Kenneth. The Great Divergence. China, Europe, and the Making of the Modern
World Economy Princeton, New Jersey: Princeton University Press, 2000.
638
Ponce, Benjamin, and Kenneth F. Clark. "The Zacatecas Mining District; A Tertiary Caldera
Complex Associated with Precious and Base Metal Mineralization." Economic
Geology 83 no. 8 (1988): 1668-1682.
Pontes y Fernández, José María Historia de la antigua ciudad de Sisapón, hoy Almadén del
Azogue Madrid: Enrique Rojas, 1900.
Portuondo, María. "Constructing a Narrative: The History of Science and Technology in Latin
America." History Compass 7, no. 2 (2009): 500-522.
Probert, Alan. "Bartolome de Medina: the Patio Process and the Sixteenth Century Silver
Crisis." Journal of the West 8, no. 1 (1969): 90-124.
Puerto Sarmiento, Francisco Javier. "La panacea aúrea. Alquimia y destilación en la corte de
Felipe II (1527-1598)." Dynamis (Granada, Spain) 17 (1997): 107-40.
Ramírez, S.E. "La minería y la metalurgia nativa en el norte peruano (siglos XVI-XVII)."
Anuario de estudios americanos 64, no. 1 (2007): 175-208.
Ramos-Arroyo, Yann René, Rosa María Prol-Ledesma, and Christina Siebe-Grabach.
"Características geológicas y mineralógicas e historia de extracción del Distrito de
Guanajuato, México. Posibles escenarios geoquímicos para los residuos mineros."
Revista Mexicana de Ciencias Geológicas 21, no. 2 (2004): 268-284.
Ramos, Demetrio. "Ordenación de la minería en Hispanoamérica durante la época provincial
(siglos XVI, XVII y XVIII)." In La minería hispana e iberoamericana. Ponencias del
I coloquio internacional sobre historia de la minería. Cátedra de San Isidro 373-397.
León: Cátedra de San Isidro 1970.
Randall, Robert W. Real del Monte: a British Mining Venture in Mexico Austin: The University
of Texas Press, 1972.
Rankine, Margaret E. "The Mexican Mining Industry in the Nineteenth Century with Special
Reference to Guanajuato." Bulletin of Latin American Research (1992): 29-48.
Rawson, William. "The Present Operations and Future Prospects of the Mexican Mine
Associations Analysed, by the Evidence of Official Documents, English and Mexican,
and the National Advantages Expected from Joint Stock Companies, Considered, in a
Letter to the Right Hon. George Canning." London: J. Hatchard & Son, 1825.
Read, John. Through Alchemy to Chemistry; A Procession of Ideas & Personalities [in
English]. London: G. Bell, 1957.
Read, John, and F. H. Sawyer. Prelude to Chemistry; An Outline of Alchemy, Its Literature and
Relationships London: G. Bell and Sons, ltd., 1936.
Rech, Georges. "La fonderie de Wisenbach (Vosges)." In Mines et métallurgie edited by Paul
Benoit. Les chemins de la recherche, 173-188. Villeurbanne: Programme Rhône-Alpes
recherches en sciences humaines, 1994.
Renberg, Ingemar, Richard Bindler, and Maja-Lena Brännvall. "Using the Historical
Atmospheric Lead-Deposition Record as a Chronological Marker in Sediment Deposits
in Europe." The Holocene 11, no. 5 (2001): 511-516.
Reyes, Aurelio de los. Los caminos de la plata Mexico: Universidad Iberoamericana, 1991.
Reyes, Juan Carlos. "Introducción." In La sal en México edited by Juan Carlos Reyes vii-ix.
Colima: Universidad de Colima,Consejo Nacional para la Cultura y las Artes, 1995.
Reza, Alma Linda. Guanajuato y sus miasmas. Higiene urbana y salud pública, 1792-1804
Guanajuato: Presidencia Municipal de Guanajuato, 2001.
Rice, Claude T. "The Silver-Lead Mines of Santa Barbara, Mexico." The Engineering and
Mining Journal LXXXVI, no. 5 (1908): 207-211.
Richards, John F. The Unending Frontier: An Environmental History of the Early Modern
World Berkeley: University of California Press, 2005.
Rickard, T. A. Journeys of Observation San Francisco: Dewey Pub. Co., 1907.
Ripley, George. The compound of Alchemy London: Thomas Orwin, 1591.
639
Rippon, Stephen, Peter Claughton, and Chris Smart. Mining in a Medieval Landscape: The
Royal Silver Mines of the Tamar Valley Exeter: Univ of Exeter Press, 2009.
Rivot, Louis Edouard. Description des gites métallifères, de la preparation mecanique et du
traitement metallurgique des minerais de plomb argentiferes de Pontgibaud Paris1851.
———. Principes généraux du traitement des minerais métalliques traité de métallurgie
théorique et pratique Paris: Dalmont et Dunod, 1859.
Robb, L. J. Introduction to Ore-Forming Processes Malden, MA: Blackwell Pub., 2005.
Robertson, William Parish. A Visit to Mexico, by the West India Islands, Yucatan and United
States, with Observations and Adventures on the Way Vol. II London: Simpkin,
Marshall & Co., 1853.
Robins, N.A. Mercury, Mining, and Empire: The Human and Ecological Cost of Colonial
Silver Mining in the Andes Indiana University Press, 2011.
Roche, Marcel. "Early History of Science in Spanish America." Science 194, no. 4267 (1976):
806-810.
Rodriguez Marin, Francisco. Felipe y la Alquimia Madrid1927.
Rosen, George. The History of Miners’ Diseases. A Medical and Social Interpretation New
York: Schuman’s, 1943.
Ruiz de la Barrera, Rocio. "La Empresa de Minas del Real del Monte (1849 - 1906)." Colegio
de Mexico, 1995.
Saavedra Silva, Elvira Eva, and María Teresa Sánchez Salazar. "Minería y espacio en el distrito
minero Pachuca-Real del Monte en el siglo XIX.". Investigaciones Geográficas,
Boletín del Instituto de Geografía, no. 65 (2008): 82-101.
Saignes, Thierry. "Las técnicas mineras de Potosí según una relación inédita de 1600." Arte y
arqueología 8, no. 9 (1982/1983): 171-176.
Salas, Guillermo P. Carta y Provincias Metalogenéticas de la República Mexicana Mexico:
Consejo de Recursos Minerales, 1980.
Salazar-Soler, C. "Innovaciones técnicas, mestizajes y formas de trabajo en Potosí de los siglos
XVI y XVII." In O trabahlo mestigo. Maneiras de pensar e formas de viver séculos
XVI a XIX, edited by Eduardo Franca Paiva and Carla Maria Junho Anastasia 141-163.
Sao Paulo: UFMG, 2002.
Salazar González, Guadalupe. Las haciendas en el siglo XVII en la región minera de San Luis
Potosí: Su espacio, forma, función, material, significado y la estructuración regional
San Luis Potosí Universidad Autonoma de San Luis de Potosí, 2000.
Salkield, Leonard U. "Ancient Slags in the South West of the Iberian Peninsula." In La minería
hispana e iberoamericana. Ponencias del I coloquio internacional sobre historia de la
minería 85-98. León: Cátedra de San Isidro, 1970.
Sánchez-Crispín, Alvaro, Eurosia Carrascal, and Alejandrina de Sicilia Muñoz. "De la minería
al turismo: Real de Catorce y Cerro de San Pedro, México. Una interpretatión
geográfico-económica." Revista Geográfica, no. 119 (1994): 81-106.
Sánchez Gómez, Julio. De minería, metalúrgica y comercio de metales : la minería no férrica
en el Reino de Castilla, 1450-1610 Salamanca: Universidad de Salamanca : Instituto
tecnológico geominero de España, 1989.
Sánchez Gómez, Julio, and Guillermo Mira Delli-Zotti. "Minería americana y minería europea,
1750-1820: una perspectiva comparada." In Mundialización de la ciencia y cultura
nacional: Actas del Congreso Internacional "Ciencia, Descubrimiento y Mundo
Colonial", edited by A. Lafuente, A. Elena, M.L. Ortega, Universidad Autónoma de
Madrid, Comité Conjunto Hispano-Norteamericano para la Cooperación Cultural y
Educativa and E.C. y Comunicación, 105-110. Aranjuez (Madrid): Doce Calles, 1993.
Schlutter, Christophe-Andre. De la fonte des mines, des fonderies, etc. . Translated by M.
Konig. Vol. 2 Paris: Jean-Thomas Herissant et Jacques-Noel Pissot, 1853.
640
Schmitz, Christopher. World Non-ferrous Metal Production and Prices, 1700-1976 [in
English]. London; Totowa, N.J.: Cass ; Biblio Distribution Centre, 1979.
Sempat Assadourian, Carlos. "Base técnica y relaciones de producción en la minería de Potosí."
In Ciencia colonial en América, edited by Antonio Lafuente and José Sala Catalá 121-
142. Madrid: Alianza, 1992.
———. Zacatecas, conquista y transformacion de la frontera en el siglo XVI : minas de plata,
guerra y evangelización México, D.F.: Colegio de México, Centro de Estudios
Históricos, 2008.
Serrano Bravo, Carlos. "Historia de la minería andina boliviana (siglos XVI al XX) "
http://www.unesco.org.uy/phi/biblioteca/archive/files/370d6afed30afdca14156f9b55e
6a15e.pdf.
Serrano, C. "Minería salud en el Potosí colonial; Mining health in the colonial Potosi." Arch.
boliv. hist. med 11, no. 1/2 (2005): 21-29.
SIavec, Z.Z. "Occupational Medicine in Idria Mercury Mine in 18th century." Vesalius IV, no.
2 (1998): 51-59.
Sigerist, H.E. "The Wesley M. Carpenter Lecture:" Historical Background of Industrial and
Occupational Diseases"." Bulletin of the New York Academy of Medicine 12, no. 11
(1936): 597.
Silliman, Benjamin. Sketch of the Great Historic Mines of the Cerro de Proaño at Fresnillo,
State of Zacatecas New Haven, Conn.: Tuttle, Morehouse & Taylor, 1883.
Sillitoe, R.H. "Relation of Metal Provinces in Western America to Subduction of Oceanic
Lithosphere." Geological Society of America Bulletin 83, no. 3 (1972): 813-818.
Sillitoe, Richard H. "Supergene Silver Enrichment Reassessed." Chap. 2 In Supergene
Environments, Processes and Products, edited by Spencer R. Titley 15-32: Society
Economic Geology, 2009.
Singer, D.A. "World Class Base and Precious Metal Deposits; A Quantitative Analysis."
Economic Geology 90, no. 1 (1995): 88-104.
Smith Jr, D.M. "Geology of Silver Deposits along the Western Cordillera." In Silver -
Exploration, Mining and Treatment 11-24. Mexico City: Institution of Mining and
Metallurgy, 1989.
Smith, Pamela H. "Vermilion, Mercury, Blood, and Lizards: Matter and Meaning in
Metallurgy." In Materials and Expertise in Early Modern Europe : Between Market
and Laboratory, edited by Ursula Klein and E. C. Spary, 29-49. Chicago: University of
Chicago Press, 2010.
Soetbeer, Adolf. Edelmetall-produktion und werthverhältniss zwischen gold und silber seit der
entdeckung Amerika's bis zur gegenwart Gotha J. Perthes, 1879.
Sonneschmidt, F.T., and J.M. de Fagoaga. Tratado de la amalgamación de Nueva España
Galería de Bossange (padre), 1825.
Splitter, Henry Winfred. "Quicksilver at New Almaden." Pacific Historical Review 26, no. 1
(1957): 33-49.
Stampa, Manuel Carrera. "The Evolution of Weights and Measures in New Spain." The
Hispanic American Historical Review 29, no. 1 (1949): 2-24.
Stein, Robert Louis. The French Sugar Business in the Eighteenth Century [in English]. Baton
Rouge: Louisiana State University Press, 1988.
Ströbele, F., T. Wenzel, A. Kronz, L.H. Hildebrandt, and G. Markl. "Mineralogical and
Geochemical Characterization of High-medieval Lead–Silver Smelting Slags from
Wiesloch near Heidelberg (Germany)—An Approach to Process Reconstruction."
Archaeological and Anthropological Sciences 2, no. 3 (2010): 191-215.
641
Strode, S., L. Jaeglé, and N.E. Selin. "Impact of mercury emissions from historic gold and
silver mining: Global modeling." Atmospheric Environment 43, no. 12 (2009): 2012-
2017.
Studnicki-Gizbert, D., and D. Schecter. "The Environmental Dynamics of a Colonial Fuel-
Rush: Silver Mining and Deforestation in New Spain, 1522 to 1810." Environmental
History 15, no. 1 (2010): 94.
Suarez Arguello, Clara Elena, and Brígida Von Mentz. Epístolas y cuentas de la negociación
minera de Vetagrande, Zacatecas, 1791-1794, 1806-1809 Mexico: Centro de
Investigaciones y Estudios Superiores en Antropologia Social (CIESAS), 2008.
Suárez de Peralta, Juan. Tratado del descubrimiento de las Indias : (noticias históricas de la
Nueva España) México, D.F.: Consejo Nacional para la Cultura y las Artes, 1990.
Suter, Max. "A neotectonic-geomorphologic investigation of the prehistoric rock avalanche
damming Laguna de Metztitlán (Hidalgo State, east-central Mexico)." [In English].
Revista Mexicana de Ciencias Geológicas 21, no. 3 (2004): 397-411.
Svarc-Gajic, Jaroslava. General Toxicology New York: Nova Science Publishers, 2009.
Tandeter, Enrique. "Forced and Free Labour in Late Colonial Potosi." Past & Present, no. 93
(1981): 98-136.
Taylor, John. "Description of Retorts for the Distillation of the Mercury from Amalgamated
Metals ". Records of mining I (1829): 141-144.
———. Selections from the Works of the Baron de Humboldt, Relating to the Climate,
Inhabitants, Productions, and Mines of Mexico with Notes by John Taylor London:
Printed for Longman, Hurst, Rees, Orme, Brown, and Green, 1824.
Teich, Mikuláš. "Born's Amalgamation Process and the International Metallurgic Gathering at
Skleno in 1786." Annals of Science 32, no. 4 (1975/07/01 1975): 305-340.
TePaske, John Jay, and Kendall W. Brown. A New World of Gold and Silver Leiden,
Netherlands; Boston: Brill, 2010.
Terreros, Manuel Romero de. Antiguas haciendas de México Editorial Patria, 1956.
———. "El condado de Regla en 1810." Historia Mexicana 4, no. 1 (1954): 107-114.
Thibodeau, A. M., D. J. Killick, J. Ruiz, J. T. Chesley, K. Deagan, J. M. Cruxent, and W.
Lyman. "The Strange Case of the Earliest Silver Extraction by European Colonists in
the New World." Proceedings of the National Academy of Sciences of the United States
of America 104, no. 9 (2007): 3663-3666.
Thomas, Hugh. The Golden Empire. Spain, Charles V, and the creation of America New York:
Random House, 2010.
———. Rivers of Gold : The Rise of the Spanish Empire New York: Random House, 2005.
Thommen, Lukas. An Environmental History of Ancient Greece and Rome Cambridge; New
York: Cambridge University Press, 2012.
Thornton, I., and P. Abrahams. "Historical Records of Metal Pollution in the Environment." In
Changing Metal Cycles and Human Health, edited by J. O. Nriagu 7-25. Heidelberg,
New York, Tokyo: Springer-Verlag, 1984.
Toussaint, Manuel. La litografía en México en el siglo XIX : sesenta y ocho reproducciones en
facsímil México: Estudios Neolitho, 1934.
Tudor, Henry. Narrative of a Tour in North America Comprising Mexico, the Mines of Real
del Monte, the United States, and the British Colonies : with an Excursion to the Island
of Cuba : in a Series of Letters, Written in the Years 1831-2 London: J. Duncan, 1834.
Turetsky, Merritt R., Sturt W. Manning, and R. Kelman Wieder. "Dating Recent Peat
Deposits." Wetlands 24, no. 2 (2004): 324-356.
Tylecote, R. F. "Roman Lead Working in Britain." The British Journal for the History of
Science 2, no. 1 (1964): 25-43.
642
Ulloa, Antonio de. Noticias americanas: Entretenimientos phisicos-historicos, sobre la

America Meridional, y la Septentrional Oriental Madrid: Imprenta de Don Francisco
Manuel de Mena, 1772.
UNEP. "Global Mercury Assesment 2013 ". Geneva: United Nations Environmental Program.
Uribe Salas, José Alfredo, and Rubén Darío Nuñez Altamirano. "Depreciación de la plata,
políticas públicas y desarrollo empresarial. Las pequeñas y medianas empresas mineras
mexicanas de Pachuca y Real del Monte." Revista de Indias LXXI (2011): 449-480.
USEPA. Air Quality Criteria for Lead EPA/600/R-5/144aF. Vol. I 2006.
———. Air Quality Criteria for Lead EPA/600/R-5/144aF. Vol. I,II 2006.
Van Buren, Mary, and Claire R Cohen. "Technological Changes in Silver Production after the
Spanich Conquest in Porco, Bolivia.". Boletín del Museo Chileno de Arte Precolombino
15 (2010): 29-46.
Vaněk, Václav, and Dalibor Velebil. "Staré hutnictví stříbra." In Stříbrná Jihlava 2007. Studie
k dějinám hornictví a důlních prací 188-205. Jihlava: Archaia Brno / Muzeum
Vysočiny Jihlava, 2007.
Vega, Garcilaso de la, and Aurelio Miró Quesada S. Comentarios reales de los Incas Caracas:
Biblioteca Ayacucho, 1976.
Velasco Avila, Cuahtemoc, Eduardo Flores Clair, Alma Laura Parra Campos, and Edgar Omar
Gutierrez Lopez. Estado y minería en México (1767-1910) Mexico: Fondo de Cultura
Económica, 1988.
Vergani, Raffaello. "La métallurgie des non-ferreux dans la république du Venise (XVe-
XVIIIe siècles)." In Mines et métallurgie edited by Paul Benoit. Les chemins de la
recherche, 207-215. Villeurbanne: Programme Rhône-Alpes recherches en sciences
humaines, 1994.
Viajero. "Las Minas de Mexico." Revista de Minas Tomo I (1861): 176-183.
Villaseñor y Sánchez, José Antonio de. Theatro americano : descripcion general de los reynos
y provincias de la Nueva España y sus jurisdiccions [in Spanish]. Mexico: Imprenta
de la Viuda de D. Joseph Bernardo de Hogal, 1746.
Viollet, Eugène. "Le problème de l’Argent et l’Etalon d’Or au Mexique." Université de Paris,
1907.
Wagner, Henry R. "Early Silver Mining in New Spain." Revista de Historia de América, no.
14 (1942): 49-71.
Wallace, T.C., M. Barton, and W.E. Wilson. "Silver & Silver-Bearing Minerals." Rocks &
Minerals 69, no. 1 (1994): 16-38.
Ward, H. G. Mexico [in English]. Second ed. London: H. Colburn, 1829.
———. Mexico. Vol. I London: H. Colburn, 1829.
———. Mexico Vol. II London: H. Colburn, 1829.
Waszkis, Helmut Mining in the Americas: Stories and History Cambridge: Woodhead
Publishing Limited, 1993.
Wendt, Arthur F. "The Potosi, Bolivia, Silver District." Transactions of the American Institute
of Mining Engineers XIX (1891): 74-107.
West, R.C. "Early Silver Mining in New Spain, 1531-1555." In In Quest of Mineral Wealth.
Aboriginal and Colonial Mining and Metallurgy in Spanish America. , edited by Alan
Craig and Robert C. West 119-135. Baton Rouge: Geoscience Publishers, 1994.
West, Robert. "Aboriginal metallurgy and metalworking in Spanish America: a brief
overview." In In Quest of Mineral Wealth. Aboriginal and Colonial Mining and
Metallurgy in Spanish America. , edited by Alan Craig and Robert West. Geoscience
and Man, 5-20. Baton Rouge: Geoscience Publications, 1994.
West, Robert C. The Mining Community in Northern New Spain: the Parral Mining District
Berkeley and Los Angeles: University of California Press, 1949.
643
Whitaker, Arthur P. "The Elhuyar Mining Missions and the Enlightenment." The Hispanic
Whitney, J. D. The metallic wealth of the United States, described and compared with that of
other countries Philadelphia: Lippincott, Grambo & Co., 1854.
Wild, Mike, and Ian Eastwood. "Soil Contamination and Smelting Sites." In Boles and
Smeltmills Seminar, edited by Lynn Willies and David Cranstone 54-57. Reeth,
Yorkshire: Historical Society, Ltd., 1992.
Williams, Michael. Deforesting the Earth from Prehistory to Global Crisis : an Abridgment
Chicago: University of Chicago Press, 2006.
Willies, Lynn. "Derbyshire Lead Smelting in the Eighteenth and Nineteenth Centuries."
Bulletin of the Peak District Mines Historical Society 11 (1990): 1-19.
———. "Introduction." Paper presented at the Boles and Smeltmills: Report of a seminar on
the History and Archaeology of Lead Smelting, Reeth, Yorkshire, United Kingdom, 15
to 17 May 1992.
Wilson, Robert Anderson. Mexico and Its Religion; with Incidents of Travel in that Country
During Parts of Years 1851-52-53-54 New York: Harper & Bros., 1855.
Wilson, W. E., and A. Petrov. "Famous Mineral Localities: Cerro Rico de Potosi, Bolivia."
Mineralogical Record 30 (1999): 9-36.
Wilson, W. J. "An Alchemical Manuscript by Arnaldus de Bruxella." Osiris 2 (1936): 220-
405.
Winter, Thomas G. "The Evaporation of a Drop of Mercury." American Journal of Physics 71
(2003): 783-786.
Young, Otis E. "The Spanish Tradition in Gold and Silver Mining." Arizona and the West 7,
no. 4 (1965): 299-314.
Zavala, S. "La amalgama en la minería de Nueva España." Historia Mexicana XI, no. 3 (1962).
Zori, Colleen M, and Peter Tropper. "Late Pre-Hispanic and Early Colonial Silver Production
in the Quebrada de Tarapaca, Northern Chile.". Boletín del Museo Chileno de Arte
Precolombino 15 (2010): 65-87.

Stream Gate

Uploaded by

Copyright:

Available Formats

Stream Gate

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Stream Gate

Uploaded by

Copyright:

Available Formats

The Environmental History of Silver Refining in New Spain and Mexico, 16c to 19c:

Saúl José Guerrero Quintero

Department of History and Classical Studies

© Saúl José Guerrero Quintero, 2015

Table of Contents ........................................................................................................... 1

2.6.4 Charcoal and wood: its magnitude .......................................................... 165

4.4.4 The silver content of ores used for amalgamation................................... 338

5.7 The capital cost of amalgamation at Regla. ................................................. 439

A. Silver production in smelting haciendas...................................................... 585

Americas (illustration based on Sillitoe in footnote 51). ...................... 68

in footnote 55). ..................................................................................... 70

Figure 1-4. A simplified representation of Sillitoe’s interpretation of the effect of

weathering on an original (hypogene) non-weathered sulphidic silver

deposit. Drawing based on interpretation of Sillitoe in footnote 74..... 76

(based on footnote 86). ......................................................................... 81

sixteenth century. Adapted from Blanchard in footnote 98. Map and

inset not to scale. .................................................................................. 84

century, data from Table 1-II................................................................ 90

Royalty of Peru / Bolivia, as per Table 1-III, grouped by century. ...... 91

Table 1-IV, grouped by century. .......................................................... 91

from Craddock, footnote 179. ............................................................ 121

reach 2 m b) a modern reconstruction of a molino in Zacatecas, at the

exit of the El Edén mine c) drawing of a molino, reproduced from

footnote 192. ....................................................................................... 125

to the furnace. ..................................................................................... 129

Figure 2-4. Ruins of the Hacienda de Aranzazu in Guadalcazar. a) front view

showing archways under two smelting furnace chimneys, height around

7 m b) back view of chimneys, showing possible aperture for drive shaft

of bellows c) section of hacienda wall d) fields of grasas e) image from

Google Earth © 2014 DigitalGlobe, 22°37’21” N 100°24’9” W. ...... 130

c) one of two bowl-shaped depressions with caved-in roof, connected via

the tubular section observed in d). ...................................................... 133

of Figure 2-3 in the background b) Hacienda HMC2. ....................... 134

2014 DigitalGlobe. The Hacienda Santa María lies at 22°12’10” N

100°44’27” W, and the Hacienda HMC1 at 22°12’31” N 100°44’47”W.

of San Luis Potosí into barrios (quarters) by the Viceroy Marques de

Branciforte. Digital copy courtesy of AHSLP, Colección Mapas y

Planos. ................................................................................................ 140

Letters in bold indicate mass input, letters in capitals indicate mass

output. ................................................................................................. 147

the Hacienda Santa Maria in Monte Caldera, adapted from original

drawing with permission of Prof. Guadalupe Salazar González, footnote

267. ..................................................................................................... 158

Smelter Taxco Guerrero’, 1905, number 13 in the series Tema y

Tecnología (CIG-AGN)...................................................................... 160

haciendas in the area of Monte Caldera, adapted from the architectural

reconstructions in footnote 275. ......................................................... 162

adapted from the original map in footnote 300. ................................. 171

Figure 3-1. Timeline of main stages in the implementation of amalgamation in the

New World. ........................................................................................ 210

Spain / Mexico. Dashed lines indicate optional stages. ...................... 212

Figure 3-3. a) a horse-powered Molino in Mexico, late nineteenth century reproduced

the ruins of the Hacienda San Juan Nepomuceno, Marfil, Guanajuato c)

drawing of a tahona /arrastre, showing its four voladoras, reproduced

from MacDonald, footnote 420 d) photo of a wasted voladora stone, over

Hacienda Bustos, Guanajuato. ........................................................... 215

from Antúnez Echegaray, footnote 423 b) patio of the Hacienda de La

Luz, Guanajuato, photo reproduced from Hernandez and Rangel,

footnote 425 c) patio of unidentified hacienda in Guanajuato, photo