Environmental Earth Sciences (2018) 77:577

https://doi.org/10.1007/s12665-018-7744-4

THEMATIC ISSUE

Petrographic and geochemical investigations on the volcanic rocks

used in the Punic-Roman archaeological site of Nora (Sardinia, Italy)
Stefano Columbu1

Received: 22 September 2017 / Accepted: 6 August 2018 / Published online: 21 August 2018
© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
The research focuses on the geochemical and petrographic characterisation of volcanics that were used in the Roman city
of Nora archaeological site. These rocks formed during the Late Eocene–Miocene orogenic Sardinian magmatic activity
(38–15 Ma), were used as construction materials (ashlars, decorative frame, etc.) and as aggregate in the ancient mortars of
some important buildings of Nora (e.g. Roman theatre, first century ad). Several samples coming from the Roman theatre
(taken as study cases) and other ancient buildings and artefacts of Nora village were studied and chemically petrographi-
cally analysed. To define their provenance, analytical data of major and trace elements by XRF method were compared with
those of several volcanics coming from the sector around the Nora archaeological site and from Sulcis area, frequented in
ancient times by Romans. To highlight the geochemical discriminant markers or affinities between the sample populations
of volcanics, factorial analysis was used, taking some significant variables: ­SiO2, ­Na2O, ­K2O, L.o.I, ­Al2O3, Rb, Ba, Sr. The
results of this research allow also defining the aspects of stone quarry extraction and construction technologies used in the
Roman period.

Keywords  Petrographic · Chemical analysis · XRF · Raw material · Roman technologies

Introduction surrender to Karales (today Cagliari, Fig. 1) who had sup-

ported the emperor Caesar and his faction in the struggles
The Nora archaeological site is located in the Gulf of for the conquest of supremacy in Rome. Nora, however,
Cagliari near the city of Pula (south western Sardinia, continued to consolidate her importance as a residential
Fig.  1). Nora was founded by the Phoenicians around centre. It became a “town hall” and had its own magis-
the mid-eighth century bc and it was Carthaginian from trates—extending its influence thanks to its strategic posi-
the sixth century bc . It became Roman in 238  bc (Ber- tion of crossroads of roads connecting the metallic and
nardini 2004; Mastino 2005). For centuries, Nora has agricultural hinterland with Karales. Its decline began in
been the main Roman administrative and political centre the fifth century ad, due to the several incursions of the
of Sardinia, a primacy that in the Republican age had to Vandals and the Arabs who forced the inhabitants to flee
towards Sardinia. The archaeological remains of Nora
This article is part of a Topical Collection in Environmental (on approximately 30,000 m 2 (Figs. 2, 3) were discov-
Earth Sciences on “Stone in the Architectural Heritage: from ered with the excavations that Gennaro Pesce conducted
quarry to monuments—environment, exploitation, properties and in 1952 (Pesce 1954; 1972) which brought to light part
durability”, guest edited by Siegfried Siegesmund, Luís Sousa, and
of the city (Tronchetti 2001). Emerged part reveals the
Rubén Alfonso López-Doncel.
remains of a first Punic settlement pre-existing to the
Electronic supplementary material  The online version of this Roman city. A necropolis with well-tombs, remains of
article (https​://doi.org/10.1007/s1266​5-018-7744-4) contains several multi-storey houses which have been demolished
supplementary material, which is available to authorized users.
in Roman times, and a rectangular temple that has been
* Stefano Columbu attributed to goddess Tanit, located on the culmination of
[email protected] the central relief of the promontory, have been identified.
From the top of this temple, the lines of the Roman urban
1
Dipartimento di Scienze Chimiche e Geologiche, Università structure can be clearly seen (Figs. 2, 3): imperial straight
degli Studi di Cagliari, Cagliari, Italy

13
Vol.:(0123456789)
577 
Page 2 of 19 Environmental Earth Sciences (2018) 77:577

the “sanctuary” of the salutary gods are important too,

in particular a restored and expanded at different times
Punic-Roman temple.
The remains of interesting ancient buildings are well
preserved (Fig. 2): “Terme a mare” (i.e. Baths at the Sea),
“Piccole Terme” (i.e. small baths), Eusculapio Temple,
while of other only traces remain: Christian Basilica (now
submerged under the sea, Fig. 2), Tanit Temple. The most
important building, the only one of its kind in Sardinia, is the
Roman Theatre (Figs. 2, 3), set in front of the sea. It surely
is the best preserved and it was the first monument of the
ancient city to be unearthed during the Pesce excavations
(Pesce 1954, 1972). It still shows the original construction
features, characterised by a semi-circular isodomic front
with a W-NW axis (Fig. 4). The theatre building was built
during the Augustan or Giulio-Claudio period (first century
ad; Bejor 1999) using a variety of geomaterials (sandstones,
conglomerates, volcanic rocks, marbles, bricks, mortars,
etc.) and different kind of mortars, as often used by the
Romans (Columbu et al. 2017; Columbu and Garau 2017).
After the abandonment of Roman Nora city (from fourth
century ad), over the centuries, the village facilities were
plundered and vandalised; more than half of the ashlars of
the tiers of cavea are missing or have been recently replaced
(Fig. 3).
This paper aims to define the provenance of the
Eocene–Miocene intermediate-acid volcanics that have
been used to construct some ancient buildings of Nora. Due
to their easy workability, these volcanic rocks were widely
used in Sardinia in different historical times (Antonelli et al.
2014; Columbu 2017; Columbu et al. 2018). The origin of
raw materials has always been a topic of great interest and
a heated debate both from the petrographic, archaeological
and archaeometric points of view, as the data can highlight
possible ways of communication and/or existing trade of
civilizations. Moreover, the geochemical and petrographic
investigations on the features and origin of raw materials
Fig. 1  Geological map of Sardinia with localization of the Roman
Nora village and other important Punic-Roman city from the south are also useful to understand the ancient routes the Romans
to the north of island, i.e. Karales (today Cagliari), Forum Traiani used for the stone transport, the construction technologies of
(Fordongianus), Turris Lybisonis (Porto Torres); from Columbu and ancient buildings that have been used in different historical
Garau (2017). Legend of patterns and colours refers to lithologies:
times and also to define the conservative problems (Alvarez
white, recent alluvial sediments; light gray, Oligo-Miocene volcanics;
dark gray, Plio-Pleistocene volcanics; gray stippled, Miocene marine et al. 2000; Bianchini et al. 2004; Columbu et al. 2015; De
sediments; gray crosses, Paleozoic crystalline basement and Meso- Luca et al. 2015; Franzini et al. 2000; Maravelaki-Kalaitzaki
zoic formations. Red continuous and dashed lines, faults et al. 2003; Moropoulou et al. 2000; Riccardi et al. 1998;
Verdiani and Columbu 2010; Vola et al. 2011).
To define the origin of raw volcanic materials: (1) geo-
and paved streets separate the neighbourhoods including chemical classification and petrographic characterisation of
public buildings, shops and houses. Two of these latter the main volcanic rocks used in the Roman construct of Nora
have the floors covered with a precious polychrome mosaic site (where the theatre was taken as study case); (2) compari-
with sophisticated and elegant motifs’ decoration. There is son of chemical data between the samples of the theatre and
also a melting workshop (foundry), laundry, tannery and other artefacts of Nora site with the samples coming from
the ruins of a likely covered market (Macellum horreum, probably supply areas: the volcanic district of Sarroch-Pula
Fig. 2). The remains of courtyards and chapels that formed and the Sulcis basin (Figs. 1, 5), more frequented by Romans

13
Environmental Earth Sciences (2018) 77:577 Page 3 of 19  577

Fig. 2  Sketch map of Roman

Nora archaeological site, which
highlights the main monument
sectors built in Roman age
from 238 bc to III sec. ad (from
Superintendence of Cultural
Heritage of Cagliari, modified)

in ancient times; (3) defining some aspects of stone quarry carried out according to the NORMAL Recommendations
extraction and construction technologies used in the Roman (3/80 1980). The sampling points of the materials have
period. been chosen taking into account the need not to disfigure
With the aim of achieving these objectives, 136 samples the monuments.
of rock samples from field outcrops and from Nora monu- 68 samples from Nora monuments (#43 from the theatre
ments were analysed by XRF methods. signed with TN in the tables ESM 1a, 1b (see Fig. 4) and
#25 from Nora site other ancient artefacts, Fig. 2, signed
with VN in the table ESM 2) were taken. However, due to
Materials and methods small size of specimen, six samples taken from the thea-
tre were studied but not chemically analysed. The rock
The analysed volcanic rocks were taken from the Roman samples of theatre come from the volcanic ashlars used
Theatre (i.e. cavea tiers, Fig. 3d, e) and other ancient arte- for the tiers of cavea sector (Fig. 4) that have a size about
facts of Nora archaeological site (i.e. houses, Macellum hor- 80 × 60 × H = 50 cm. The other samples belong to the other
reum masonry, inner street, thermal baths; Figs. 2, 3g, h). Roman artefacts located inside the archaeological Nora area
These volcanics show a clear autoclastic structure (Fig. 6c–f) (i.e. Roman paved street, ashlars of building walls, frames,
characterised by a chaotic presence of large lava clasts (up thresholds of house doors; Figs. 2, 3).
to 30 cm and generally with sharp edges) and smaller clasts To find the origin of monument samples, 68 samples of
(with size < 1 cm and with a rounded edges) immersed in a rock samples from the following field outcrops: Nora area
more glassy matrix characterised by a lower welding degree (within the archaeological site, signed with PN in the table
than the clasts (Fig. 6). ESM 3), Perdu Pranu (near to archaeological site, signed
Sampling activity from Nora monuments, subject to the with PN), Su Casteddu (volcanic apparatus, signed with
approval of the Archaeological Superintendence of Cagliari, CVR), St. Antioco area were taken (Figs. 1, 5).
was performed taking into account the most representative Mineralogical and petrographic analysis of volcanic rocks
and/or predominant lithotypes in the entire architectural was performed on thin sections under the polarising micro-
structure; the planning was done examining what analysis scope (Zeiss photomicroscope Pol II).
would be carried out on the samples, and knowing, there- Chemical composition of volcanic rock samples from
fore, the amount of material needed to achieve them. It was field outcrops and Nora site was determined with a

13
577 
Page 4 of 19 Environmental Earth Sciences (2018) 77:577

Fig. 3  Views of Roman Nora

site: a south-east overview of
Nora archaeological area; b
northwest overview of Foro,
the Roman public square with
part of original volcanic stone
paving; c completed overview
of the theatre (down on the left
the hyposcenium, space under
the stage, the palcoscenium)
with the two tribunalia sectors
on the left and right (see also
Fig. 4); d south view of cavea
tiers of theatre (and orches-
tra floor with mosaic, down)
realised with volcanic ashlars
(now partially changed by other
sandstone ashlars); e view detail
of central sector of cavea tiers;
g original Roman volcanic stone
paving between the Macellum
horreum and “Terme a mare”
Bath sectors (see Fig. 2); h wall
masonry within the Macellum
horreum sector (see Fig. 2) with
big squared volcanic ashlars
positioned at the crossing point
of walls

spectrometer in X-ray fluorescence (XRF) Philips PW1400 the elements. For trace elements, after repeated determi-
with a Rh tube to analyse the major elements and some nations, the values of detection limits were determined
trace elements (Rb, Sr, Pb, Zn, Y, Nb, Zr), and with a W according to 3σ, 2σ and 1σ; they are reported in Table 1
tube to analyse of Ni, Cr, Ba, V, La, and Ce. Data reduc- (Garau 2005). The accuracy of trace elements is ± 2–3% to
tion of major elements was performed by the Franzini 1000 ppm, ± 5–10% at 100 ppm and ± 10–20% to 10 ppm.
et al. method (1975). Data reduction of trace elements Weight loss for calcination (L.o.I., loss on ignition)
was performed by Criss method modified (Criss 1977). was determined by calculating the loss in wt% at 1100 °C,
Measurement accuracy is ± 1% for ­S iO 2, ­TiO 2, ­A l 2O 3, while the FeO was determined by volumetric titration with
­Fe2O3, CaO, K ­ 2O and MnO and ± 4% for MgO, N ­ a2O and ­KMnO4 10 N in acid solution.
­P2O5. Detection limits are about 3 ppm to 3σ for most of

13
Environmental Earth Sciences (2018) 77:577 Page 5 of 19  577

Fig. 4  Plan of the Nora Roman

theatre with the sampling points
(in red) of volcanic rocks. Leg-
end of sectors: a cavea (audi-
torium); b orchestra; c wall of
stage; d hyposcenium (space
under the palcoscenium); e
front scene; f portico behind the
scene; g west archway entrance;
h access ladder at the west
tribunalia; i west tribunalia; l
parasceni; m external niches;
n enclosure; o via (corridor
atop the cavea); p scalaria; q
euripus (underground channel
for water drainage); r dolia; s
small pillars under the palcosce-
nium (from Columbu and Garau
2017; Garau 2005, modified)

Following petrochemical parameters have been calcu- between 38 and 15 Ma, mainly along the Sardinian Oligo-
lated: Differentiation Index (given by the sum of norma- Miocene rift, a large tectonic pit, which structure crossing
tive sialic minerals except the anortite; D.I. = normative Sardinia (Cherchi and Montadert 1982a, b; Coulon 1977;
Q + Ab + Or + Ne + Kp + Lc) according to Thornton and Tut- Dostal et al. 1982).
tle (1960); Solidification Index according to Kuno (1968) as The volcanic activity shows a peak during the 22–18 Ma
time range (Beccaluva et al. 1985; Carminati and Doglioni
MgO
S.I. = (by weight %); 2012; Gattaceca et al. 2007; Speranza et al. 2002 and refer-
MgO + FeOtot + Na2 O + K2 O
ences therein). Since that time, in various areas of Sardinia,
Agpaitic Index according to Shand (1951) as prevalently along the western graben trending N–S, known
as the “Fossa tettonica sarda” (Vardabasso and Atzeni 1962;
Na2 O + K2 O
A.I. = (in molecular terms); Fig. 1), a highly explosive fissural activity, with simulta-
Al2 O3 neous and alternating emissions of basaltic and andesitic
SAL, given by the sum of all normative sialic miner- lavas, produced abundant pyroclastics and dacitic–rhyolitic
als; FEM, given by the sum of normative mafic minerals; products.
C.I.P.W. norm according to Cross et al. (1903). The volcanic rocks used in the theatre and other build-
ings of the Nora archaeological site belong to the volcanic
Sarroch-Pula district (Figs. 1, 5), which can be attributed
to this later volcanic activity. According to the radiomet-
Geological setting ric analysis of various authors, the volcanic rocks of the
Sarroch-Pula district have the following ages: 24–21.6 Ma
The volcanic rocks, which have been used as construction (Savelli et al. 1979), 24.7–22.2 Ma (Beccaluva et al. 1985)
materials in the archaeological Nora site, belong to the sig- according in part to the data for Arcuentu volcanic district
nificant Sardinian Late Eocene–Miocene orogenic magmatic determined by Coulon (1977; 22.3–17.5 Ma) and Assorgia
phase (Fig. 1; Beccaluva et al. 1985, 1989, 2005a, b, 2011; et al. (1984; 30.5–16 Ma).
Lustrino et al. 2011), that includes tholeiitic, calcalkaline, The Sarroch-Pula district (Sheet 234 of the “Carta
shoshonitic and ultrapotassic products. This phase occurred d’Italia”) is located at the southwestern Sardinia and it has

13
577 
Page 6 of 19 Environmental Earth Sciences (2018) 77:577

Fig. 5  Geological sketch map of Nora-Pula sector, showing the dif- and sandstone ashlars to realise the cavea tiers and structure walls of
ferent volcanic outcrops and the localization of the ancient quarries the theatre, respectively. NW and SW map vertices have the following
of “Su Casteddu” (northwest of Pula village) and “Fradis Minoris” geographical GPS coordinates (according to World Geodetic System,
(northwest of Nora) from which the Romans extracted the volcanic 1984): N 39.035800, E 8.949000; N 38.964400, E 8.949000

a main NW–SE orientation, occupying altogether an area eastern sector the massive rocks are prevalent. Western
of about 90 km2. It is separated from the rest of the island pyroclastic outcroppings in the Sarroch area, as well as
by the Campidano graben and displays abundant Miocene those from Nora-Pula, are often rich in volcanic clasts
calcalkaline volcanism (Sulcis and Sarroch areas, Fig. 1). (as cognate fragments or “inclusions”, from centimetric
In the Miocenic age, a series of fractures affected the area to decimetric) that are dispersed in an substantially cin-
comprising both the edges of the Campidano graben and eritic matrix. These clasts vary in shape, nature and size:
the nearby Paleozoic reliefs. By these fractures, consider- from sharp edges, predominant (usually in larger clasts),
able volcanic manifestations occurred, including those with to more rare rounded edges (in smaller clasts) and they
mainly andesitic character of the Sarroch-Pula district. are usually of sub-volcanic type. These rocks with clasts
Miocene volcanism in the Sulcis area is represented by an can be framed between volcanic breccias passing through
accumulation of up to 500 m of volcanic products, mainly conglomerate types as they move away from their emis-
extensive pyroclastic flows well exposed on San Pietro and sion volcanic centres. Such breccias and conglomerates are
St. Antioco islands, and on the mainland near Carbonia sometimes associated with particular massive lava, which
(Assorgia et al. 1990; Garbarino et al. 1990). also have a breccia appearance and they have no substan-
Sarroch volcanic area is characterised by the presence of tial differences between the clasts and the matrix. These
a main tectonic line N25E that allows distinguishing into are autobreccias, which are quite similar to those present
eastern and western sectors (Conte 1989). in the lower part of the Mount Arcuentu series (Assorgia
In the Sarroch western sector, the volcanic breccias et al. 1986).
and conglomerates are the dominant element while in the

13
Environmental Earth Sciences (2018) 77:577 Page 7 of 19  577

Fig. 6  Macro-photographs
of volcanic stone used in the
Roman theatre: a worked ash-
lars of staircase steps positioned
in the east cavea sector, near
tribunalia sector (see Fig. 4);
b photo detail of previous
photo (look at red rectangle
highlighted in the photo b);
c volcanic ashlars used for
the wall under the tribunalia
vault (Fig. 4) belonging to Su
Casteddu outcrop, where it is a
typical breccia structure of rock;
d, e volcanic ashlars used for
the wall Macellum horreum, in
front of the stone paving; f vol-
canic ashlar from the wall near
“Terme a mare” Baths (with
more welded lithoclasts and
cognate fragments immersed
within a less welded matrix)
coming probably from Perdu
Pranu outcrops

Table 1  Detection limit values Element 3σ 2σ 1σ Results

(at 3σ, 2σ and 1σ, respectively;
Garau 2005) established after Nb 3.2 2.1 1.1
repeated determinations for Petrographic features
Zr 3.0 2.0 1.0
trace elements of samples
from Roman theatre, Nora Y 3.0 2.0 1.0 Volcanics from probable origin areas
artefacts and Oligo-Miocenic Sr 4.5 3.0 1.5
volcanic rocks outcropping Rb 4.3 2.8 1.4
in the Nora and Perdu Pranu The volcanic rock outcroppings in the Sarroch-Pula vol-
Pb 12.2 8.2 4.1 canic district show different petrographic and volcanologi-
areas, Su Casteddu apparatus,
Sant’Antioco Zn 5.1 3.4 1.7 cal features (Table 2). In the area of Pula-Nora, the rocks
Ni 2.1 1.4 0.7 show a porphyritic index from 5 to 8% in more evolved
Cr 3.3 2.2 1.1 products (dacitic rocks of north Pula) to 14–28% in more
Ce 16.6 11.1 5.5 basic (andesitic) lava (Nora area). The groundmass varies
V 2.9 1.9 1.0 from hypocrystalline, to hypohyaline, to glassy. According
Ba 10.2 6.8 3.4 to Conte (1989), among these rocks there are both litho-
La 8.9 6.0 3.0 types that have the phenocrystals the amphibole and other
lithotypes in which it is absent. In paragenesis of the first
lithotype (most represented in the outcrops around the Nora
area), the amphibole phenocrystals, generally partially
opaque and sometimes pecilitic, coexist with plagioclases,
opaques and orthopyroxene (sometimes altered), while the

13

577 

13
Page 8 of 19

Table 2  Summary scheme of petrographic features defined by polarised microscope analysis on thin sections and geochemical characteristics by XRF analysis (with rock classification) of sam-
ples from Roman theatre, other Nora artefacts and Oligo-Miocenic volcanic rocks outcropping in the Nora and Perdu Pranu areas and Su Casteddu apparatus
Origin of Petrographic features by microscopic analysis Geochemical characteristics by XRF analysis
samples
Structure Porphy- Phenocrystsa Ground mass Texture Rock classification Affinity series D.I. range
ritic Index
(%) De La Roche Le Bas et al. Peccerillo and Taylor Middlemost Irvine and Thornton and
et al. (1980) (1986) (1976) (1975) Baragar Tuttle (1960)
(1971)

Theatre tiers Porphyric, 5–8 Pl, Hbl, Opq From hypo- From vitro- (An), Dc, An, Dc, An, Dc Calcalkaline, Na Subalkaline, 53–68
of cavea (glomero- (i.e. Mt, hyaline to clastic to RyDc, TrAn, (high-K (transi-
porphyric) Ti-Mt), hypocrys- eutaxitic (LtAn), (TrDc) calcalka- tional)
± Cpx, talline to massive (Lt) line)
± Opx vitrophyric
Su Casteddu Porphyric, 5–8 Pl, Hbl, Opq From hypo- From vitro- Dc, RyDc Dc Dc Calcalkaline Na Subalkaline 66–70
volcanic (glomero- (i.e. Mt, hyaline to clastic to
apparatus porphyric) Ti-Mt), hypocrys- eutaxitic
± Cpx, ± talline to massive
Opx vitrophyric
Perdu Pranu Porphyric, 12–22 Pl, Hbl, Opq Hypocrystal- From vitro- An, Dc An An Calcalkaline, Na, (K) Subalkaline 50–56
volcanic (glomero- (i.e. Mt, line, micro- clastic to (high-K
outcrops porphyric) Ti-Mt), ± granular eutaxitic to calcalka-
Cpx, massive line)
± Opx, (± Bt)
Nora volcanic Porphyric 16–28 Pl, Hbl, Opq From Tendentially An, Dc An An Calcalkaline, K, (Na) subalkaline 45–57
outcrops (i.e. Mt, hypocrys- isotrope (high-K
Ti-Mt, Il), talline to calcalka-
± Opx, holocrystal- line)
(± Cpx) line

Minerals abbreviations: Opq opaque minerals, Pl Plagioclase, Opx Orthopyroxene, Cpx Clinopyroxene, Bt Biotite, Hbl Horneblende, Mt magnetite, Ti-Mt titano-magnetite. Rock classification
abbreviations: RyDa rhyodacite, Dc dacite, TrAn trachi-andesite, TrDc trachi-dacite, Lt latite, LtAn lati-andesite, An andesite, AnBa andesitic-basalt, BaAn basaltic andesite, ( ) subordinate sam-
ple number
a
 Minerals according to their increasing abundance
Environmental Earth Sciences (2018) 77:577
Environmental Earth Sciences (2018) 77:577 Page 9 of 19  577

clinopyroxene is generally subordinated or almost absent. processes. Pyroxenes are smaller than plagioclases (about
Some samples show the presence of rare olivine microphe- 0.5 mm although some individuals reach the millimetre).
nocrystals, which are completely transformed in iddingsite, The classification of pyroxenes is doubtful due to advanced
sometimes included in opaque amphiboles together with alteration in bastite: sometimes the habit assumes that it
pyroxenes. Clinopyroxene phenocrystals often form mono- is orthopyroxene. Amphiboles (i.e. brown hornblende)
mineral aggregates, sometimes together with orthopyroxene, are about 1–1.5 mm in size and they are less altered than
plagioclase and opaque, thus conferring on these rocks a pyroxenes and plagioclases; they often crowned by edges
glomerular-porphyric structure. Orthopyroxene crystals are of opaque minerals, due to the disequilibrium with the liq-
generally isolated and idiomorphic, but quantitatively sub- uid. More rarely, the opacity phenomenon affects the entire
ordinate to those rich in calcium. Plagioclase phenocrystals crystal. Mesostasis consists predominantly of plagioclase
have an evident zonation and they are often characterised by microlites, secondarily of pyroxene and/or amphibole, and
the presence of corroded and partially reabsorbed nucleus more or less abundant glass. The crystals are oriented in the
surrounded by oscillatory zoning edges. Opaque minerals direction of the major axis, resulting in a strongly fluidal
are mostly at the edge of the orthopyroxene or included in hyalopilitic structure. In these rocks, there are also cognate
the clinopyroxenes. The groundmass is generally hypocrys- fragments, having a different degree of crystallinity but simi-
talline and it consists of plagioclase, clinopyroxene, orthopy- lar paragenesis. As already described before, such fragments
roxene and opaques. (clasts) give at macroscopic level to an autoclastic brecciate
The paragenesis of second lithotypes (without amphibole) structure, characteristic of these rocks.
consists of plagioclase, clinopyroxene, orthopyroxene, and Due to their different origin from several outcrops with
opaque, both as phenocrystals and in the groundmass. Pla- different compositions, the volcanic samples taken from
gioclase phenocrystals usually have normal zoning with a other Nora artefacts generally show a porphyritic structure
composition in the edges generally similar to that of the with a variable index (from 5% in the rhyodacitic and dacitic
microlites of the groundmass. However, plagioclases with rocks, to about 28% in andesitic rocks) for phenocrystals of
inverse zoning and crystals with glassy inclusions are also opaque, abundant plagioclase (from 0.4 to 2 mm in size),
present. Orthopyroxenes are hypidiomorphic and isolated, highly-altered clinopyroxenes (that make every determi-
while clinopyroxenes often form aggregates giving to the nation difficult), and brown hornblende (often completely
rock a glomerular-porphyric structure. Opaques are rare opacified) varying in size between 0.5 and 1 mm (though
and mostly confined to the edges of orthopyroxenes and often exceeding 3 mm). The groundmass consists of pla-
clinopyroxenes or nucleus of these latter. All of the above- gioclase, clinopyroxene, amphibole and subordinately glass.
mentioned phases also constitute the groundmass.
Geochemical features and rock classification
Volcanics of Roman Nora artefacts
Tables ESM 1, 2, and 3 show the chemical analysis of the
The samples taken from the theatre generally show a por- major elements (wt.%) and trace elements (ppm) of volcan-
phyritic structure (Fig. 7) and a hypocrystalline groundmass. ics belonging to Roman theatre, Nora artefacts and from
The porphyritic index generally varies between 5 and 8% for probable origin areas: Su Casteddu volcanic structure, Perdu
early opaque phenocrystals, dominant plagioclase, pyroxene Pranu area, Nora outcrops.
and brown hornblende. Due to their shape, the opaque min- Figure 8a, b shows the trend of the major and trace ele-
erals are presumably made up of magnetite or titanium–mag- ments against D.I.: as expected from normal magmatic evo-
netite. Plagioclases often are heuedral or subheuedral, zoned, lutionary series, a positive correlation is observed between
sometimes corroded and/or with reabsorption processes for D.I. and ­SiO2, ­Na2O, and ­K2O, while there is a negative
disequilibrium with the liquid, with cribrous appearance due correlation between D.I. and ­FeOTot, ­TiO2, MgO, and CaO.
to dust inclusions. They have a size between 0.5 and 1.5 mm Sr shows a weak negative correlation with increasing the
and they are geminated according to albite (and more rarely evolution degree of the series for the samples of Nora thea-
albite-Carlsbad). Meta-somatization processes (for circula- tre, Su Casteddu and St. Antioco outcrops, probably in con-
tion of late fluids) with pseudomorphic calcite on plagio- junction with the crystallisation of plagioclase. For some
clase and secondary alteration are frequent. elements (i.e. Rb, Sr, Ba and Nb), it is evident the trends
Another generation of plagioclase (less frequent) is rep- of the samples coming from theatre and Su Casteddu, other
resented by smaller microphenocrystals (of about one order artefacts and those of Perdu Pranu and Nora outcrops, while
of magnitude compared to the previous ones), with non- the samples from St. Antioco show an evolutionary trend
zoned heuedral crystals, geminated according to albite and almost always different and separate from other samples.
albite-Carlsbad, in equilibrium with the liquid, with more According to the Middlemost diagram (1975, Fig. 9), the
evolved chemical composition, and devoid of alteration Eocene–Miocene volcanics of all the Roman monuments

13
577 
Page 10 of 19 Environmental Earth Sciences (2018) 77:577

Fig. 7  Microphotographs of vol-
canic rocks coming from theatre
(sigle TN), from Su Casteddu
ancient quarry (CVR) and
Perdu Pranu outcrop (PN). a, b
Plain polars: TN49 and CVR3
samples with altered phenocrys-
tals (i.e. mainly plagioclase,
hornblende, pyroxene) and
cognate fragment immersed in
from hypohyaline to hypocrys-
talline groundmass; c–f crossed
Nicol: TN49, CVR3, TN41,
CVR5 samples with plagio-
clase, hornblende with opacified
border, pyroxene phenocrystals
with porphyritic index of 7%; g,
h plain and crossed Nicol: PN63
sample, with phenocrystals of
plagioclase and hornblende, and
higher porphyritic index (22%)

(including the theatre and other ancient artefacts) and field in potassium series (Fig. 9; Table 3). According to the AFM
samples from Nora area and St. Antioco are generally of diagram of Irvine and Baragar (1971; Fig. 10), the same
sodium series, more rarely of potassium series while the samples show the typical trends of the calcalkaline series.
similar volcanics of the nearby island of St. Antioco belong Within the TAS diagram of Le Maitre et  al. (2002)
to both potassium and sodium series and are rarely too high (Fig. 11), the samples of the theatre, other artefacts and Nora

13
Environmental Earth Sciences (2018) 77:577 Page 11 of 19  577

Fig. 8  a Variation diagrams: major elements (wt%) vs. differentia- Sant’Antioco. Dotted curves indicate less significant correlations.
tion index (D.I.) of Thornton and Tuttle (1960) (where D.I. = norma- Symbols as legend of Fig.  9. b Variation diagrams: trace elements
tive Q + Ab + Or + Ne + Kp + Lc) for the volcanic samples from the (ppm) vs. differentiation index (D.I.) of Thornton and Tuttle (1960)
theatre, other artefacts of Nora and field volcanic outcrops. Curves (where D.I. = normative Q + Ab + Or + Ne + Kp + Lc) for the volcanic
represented are the regression for each set of samples: in violet for samples from the theatre, other artefacts of Nora and field volcanic
the samples of the theatre; in red for the samples of Nora artefacts; outcrops. Colour regression curve notes as Fig. 14a. Symbols as leg-
in dark green for the samples of Perdu Pranu and Nora field; in end of Fig. 9
light green for the samples of Su Casteddu; in blue for samples from

outcrops fall into the field of subalkaline series, although Fig. 12), the analytical values are reported according to
numerous exception samples fall into the field of transi- the two parameters R 1 and R 2, which take into account
tional series. The theatre samples are classified as andesite, all the major elements (expressed in cations multiplied
dacite, trachy-andesite and trachy-dacite. The samples of by 1000) with the only exception of P and Mn. For these
other artefacts are classified as andesite and dacite (with reasons, this diagram has been finally selected as the most
some rare cases falling in the fields of basaltic andesite and suitable for the classification of all volcanics. The sam-
trachy-dacite). The volcanics of the countryside surrounding ples of the Roman theatre mainly fall in dacite field and,
Nora areas (Perdu Pranu and Nora areas, Fig. 5) fall into the subordinately, in lati-andesite, rhyodacite, andesite and
andesite and dacite. latite fields, in order of abundance (Fig. 12; Table 3). The
Volcanics from the St. Antioco island fall into subalka- samples of other Nora artefacts are classified as dacite,
line and transitional fields, ranging from basalt, to basal- andesite, rhyodacite and basaltic andesite. The rocks of
tic andesite, (to dacite), to rhyolite and were analysed. In Su Casteddu volcanic apparatus consist of rhyodacites
Fig. 12, the same data were projected on the diagram of and dacites. Perdu Pranu volcanics are mainly dacites,
De La Roche et al. (1980), where the segment that cuts in some cases with transitional products to the andesite.
the origin of the axes (with a slope of 45°) represents the The samples from Nora outcrops mainly show an andesitic
under-saturation critical plan. The representative points composition, with some dacitic samples. Samples from
of the analysis samples fall between that plane and the the island of St. Antioco are classified as alkali-rhyolite,
axis of the abscissas; so the samples are over-saturated basaltic andesite, rhyolite, basalt, rhyodacite and dacite
and belong to subalkaline series in accordance with TAS (Fig. 12; Table 3).
diagram (Fig. 11). In diagram of De La Roche et al. (1980;

13
577 
Page 12 of 19 Environmental Earth Sciences (2018) 77:577

Fig. 8  (continued)

Fig. 9  Na2O vs. ­K2O wt%

classification diagram of Mid-
dlemost (1975) between the
high-potassium, potassium and
sodium volcanic series, where
the volcanic samples from the
theatre, other artefacts of Nora
and field volcanic outcrops were
plotted

13
Environmental Earth Sciences (2018) 77:577 Page 13 of 19  577

Table 3  Distribution of all analysed volcanic samples within sodium The provenance study of stone materials, which have
(Na), potassium (K), and high-potassium (HK) series, according to been used in the construction of Nora monuments, has been
Middlemost (1975)
tackled both petrographically and chemically, comparing
Origin De La Roche et al. Middlemost the volcanics from the theatre and other artefacts of Nora
(1980) classification (1975) clas- archaeological site with the field samples from the near-
sification
est areas (Su Casteddu, Perdu Pranu and Nora volcanic
Na K HK outcrops, Fig.  5) and with those from St. Antioco area
Theatre tiers of cavea An 2 – – (more distant but frequented by Romans). However, this
Dc 24 – – comparison shows almost always that the volcanics of St.
RyDc 4 – – Antioco area have different trends in all the diagrams previ-
LaAn 5 – – ously considered. Observing Fig. 8a, b, these samples have
Lt 2 – – an evolutionary trend separated from the others, while the
Sub-total 37 – – samples of the theatre and the Su Casteddu area form over-
Nora artefacts An 3 2 – lapping sample populations. This is evident in all analysed
AnBa 2 – – diagrams (Figs. 8, 9, 10, 11, 12). By contrast, the samples
Dc 13 1 – of Nora artefacts, together with the samples of Perdu Pranu
RyDc 2 1 – and Nora volcanics (outcrop in the countryside around the
Sub-total 20 4 – Roman city) projected in all diagrams outline evolutionary
Su Casteddu volcanic apparatus An 1 – trends that overlap well.
Dc 2 3 – These reasons led to the exclusion of St. Antioco area as
RyDc 3 – – source of material supply for the construction of both the
Sub-total 6 3 – theatre cavea and other Roman artefacts of Nora city.
Other volcanic outcrops (Perdu An 3 2 – Thus, chemical data show similarities between the sam-
Pranu, Nora area) Dc 2 3 – ples of the theatre and the area of Su Casteddu (located in
Sub-total 5 5 – the northern suburbs of Pula; Figs. 5, 14), and among the
Sant’Antioco area Ba 4 – – volcanics of the other artefacts and those of Nora and Perdu
BaAn 10 – – Pranu outcrops. To better highlight such geochemical affin-
An 5 6 – ities, factorial analysis has been used. The variables that
Dc 1 1 – were most significant are diagrammed in the Fig. 13. In most
RyDc – 2 – of the binary diagrams, there are two distinct populations
Ry – 6 – formed, respectively, by the samples of the theatre and Su
AlRy – 20 2 Casteddu together, and by those of Nora artefacts and the
Sub-total 20 35 2 countryside surrounding the city together. The first (upper-
Total samples 88 47 2 left) diagram is the one that better separates the two sample
populations.
Rock classification according to De La Roche et al. (1980). Abbrevia- Based on both the diagrams (especially Ba vs. Ni, ­SiO2
tions as Table 2
vs. V, Sr vs. CaO, Rb vs. Zn, K ­ 2O vs. Cr; Fig. 13) and the
above-mentioned similarities, the volcanics of the Nora thea-
Discussion tre tiers come from the Su Casteddu outcrop and in no case
from the Perdu Pranu outcrops. Only some samples from the
Regarding the geochemical aspects of volcanic series, the theatre, most likely referable to lithoclasts or cognate frag-
results of chemical analysis highlight that all analysed sam- ments, have lower values of N ­ a2O (ranging in 3.63–3.65%,
ples from Nora ancient buildings and field outcrops are rhyodacites TN 36, 33, Table ESM 1b) with respect to the
subalkaline-like volcanics according to the De La Roche mean (normally > 4%), and another sample has a greater
et al. (1980) R1–R2 diagram (Fig. 12). The samples show an value of MgO (4.05%, andesite TN 50, Table ESM 1a).
affinity calcalkaline s.s. in agreement with Irvine and Bara- These assumptions made on the basis of chemical data
gar Alk-FeOT–MgO (AFM) diagram (Fig. 10). In addition, are supported by the results of the other investigations. The
most volcanics, particularly those from the theatre and Su macroscopic analysis of the petro-vulcanological character-
Casteddu outcrop, have sodium affinity according to Middle- istics (i.e. structures, welding degree, etc.) showed that the
most diagram (Fig. 9). The other volcanics are both sodium volcanic lithotypes of theatre and those of Su Casteddu are
and potassium series, and are rarely high in potassium series very similar. They are both purplish-red and with a charac-
(the latter being among the alkali-rhyolitic samples coming teristic autoclastic structure, with their cognate fragments,
from St. Antioco). which usually have a darker colour of embedded rock and

13
577 
Page 14 of 19 Environmental Earth Sciences (2018) 77:577

Fig. 10  AFM diagram of Irvine

and Baragar (1971) where the
volcanic samples from the
theatre, other artefacts of Nora
and field volcanic outcrops were
plotted. Symbols as legend of
Fig. 9

Fig. 11  Total alkali-silica (TAS)

diagram [(Na2O + K2O) vs. S ­ iO2
wt%] of Le Maitre et al. (2002;
after Le Bas et al. 1986), where
the volcanic samples from the
theatre, other artefacts of Nora
and field volcanic outcrops were
plotted. Symbols as legend of
Fig. 9. Data were not reported at
100% excluding H ­ 2O+ (accord-
ing to the IUGS recommenda-
tions; Streckeisen 1976, 1978).
Without the analytical values
of Zr, Pb, Nb Y, La and Ce, the
two series of the samples from
Nora and Perdu Pranu areas are
not included in the respective
variation diagrams

13
Environmental Earth Sciences (2018) 77:577 Page 15 of 19  577

Fig. 12  R1 vs. R2 classification diagram of De La Roche et al. (1980) where the volcanic samples from the theatre, other artefacts of Nora and
field volcanic outcrops were plotted. Symbols as legend of Fig. 9

a rounded shape; the dimensions can be very variable from because they show different geochemical characteristics, as
one millimetre to several centimetres. Even microscopically, highlighted by the various figures (Figs. 8, 9, 10, 11, 12) in
the rocks of Su Casteddu have many similarities with the the previous paragraphs where they form a separate popula-
theatre volcanics (Fig. 7). Both show from hyalopilitic to tion practically in every diagram. Observing chemical data
hypocrystalline structure, porphyric (with a porphyry index (Table ESM 3), the analysed samples from Perdu Pranu
variable between 3 and 10) for opaque phenocrystals of area (PN 61–64), that between the all outcrop samples show
magnetite or Ti-magnetite. Plagioclase is dominant, very more similar macroscopic features (i.e. sample PN 63) with
altered and with the substitution characteristics with calcite the volcanics of Su Casteddu, and have following values:
that has also been observed in the volcanics of the theatre. ­SiO2 < 62%, ­TiO2 > 0.5%, ­FeOtot > 4.92%, MgO > 1.71%,
The pyroxenes are very altered and there are relicts of brown CaO > 4.43%, ­N a 2 O < 3.75%, V > 88  ppm, while Su
hornblende with opacity edges. Mesostasis is characterised Casteddu samples (i.e. CVR 3–7) have: S ­ iO 2 > 63%,
by dominant plagioclase, pyroxene and/or brown hornblende ­TiO2 < 0.5%, ­FeOtot < 4.83%, MgO < 1.68%, CaO < 3.67%,
and more or less abundant glass. ­Na2O > 4.1%, V < 56 ppm.
Other important information can be derived from the Also microscopically they appear different (Fig. 7): the
extraction modes of raw material and the morphology of samples from Perdu Pranu and Nora generally have porphy-
the volcanic outcrop (Fig. 14): (1) the volume of extracted ritic structure with a higher index (> 15%) for phenocrystals
volcanic material is certainly compatible with the size of of opaque, abundant plagioclase, highly-altered clinopyrox-
Su Casteddu apparatus; (2) the steep slopes of Su Casteddu enes, opacified brown hornblende.
facilitated the extraction and transport of the ashlars to the However, despite the geochemical differences, the sam-
flat land. ples taken from the Roman street (e.g. VN 1, 12, 16, 18,
The volcanics, which were used to construct the other 19, 47) and other found within “Terme a mare” building
Roman buildings in Nora archaeological site (i.e. houses, (VN 21, 22, 23, probably coming from the paving of Roman
Macellum horreum masonry, inner street, thermal baths; street) show clear macroscopic and microscopic similarities
Figs. 2, 3), come mainly from the countryside of Nora and to those of the theatre and Su Casteddu volcanics. Never-
Perdu Pranu areas and probably not from Su Casteddu, theless, a more careful observation of the samples of such

13
577 
Page 16 of 19 Environmental Earth Sciences (2018) 77:577

Fig. 13  Variation diagrams of major (wt%) and trace (ppm) elements derived from the factorial analysis on the samples of theatre, other Nora
artefacts, Nora and Perdu Pranu outcrops and ancient Su Casteddu quarry (Garau 2005)

artefacts shows some petro-vulcanological differences: they (from irregular to squared) with decimetric and submetric
generally have a greater welding degree and a slightly higher size. Internally, in these walls there are occasionally large
of porphyritic index; the phenocrystals are less altered. metric volcanic blocks too (positioned in the wall crossings,
The wide use of the volcanics from Su Casteddu outcrop Fig. 3h) with very similar characteristics to those of Perdu
for the construction of the Nora theatre probably depends on Pranu and Su Casteddu areas.
the good workability of these rocks (due to their lower weld- These clear evidences indicate that the Romans chose
ing degree with respect to the other outcropping volcanics) the construction materials on the base of petro-physical fea-
and on the possibility of extracting large and regular ashlars tures of stones. Due to a good workability and easy quarry
to be used for the tiers of cavea, which in some cases even extractability and, at the same time, to their optimal physi-
have dimensions of 80 × 60 × 50 cm (Fig. 14b–d). Instead, cal–mechanical characteristics, the volcanic rocks from Su
the andesitic rocks of Nora outcrops (inside the archaeo- Casteddu were used to realise big and curved ashlars for the
logical site and, therefore, potentially exploitable) were cavea tiers of theatre without problems of fracturing during
not taken into account to realise the volcanic ashlars for the processing. Vice versa, to realise the Roman architectural
the theatre, because they show different petrographic and buildings of Nora (e.g. irregular and variable in size ashlars
physical features that also involve a lower workability of for masonry wall), where it is not necessarily particular fea-
the materials. tures of rocks, the Romans used more comfortably all dif-
The andesitic rocks, along with other sedimentary and ferent volcanic outcroppings inside the archaeological site
intrusive rocks, were used to make ashlars of various shapes or outside from other outcrops.

13
Environmental Earth Sciences (2018) 77:577 Page 17 of 19  577

Fig. 14  Macro-photographs of
Su Casteddu volcanic appara-
tus where located the Roman
ancient quarry: a north view of
volcanic neck; b, c ancient exca-
vation traces in two different
extraction points of the Roman
quarry, with the average dimen-
sions of the extracted ashlars/
blocks; d working trace in the
volcanic rock at the base of the
ashlars to be extracted

Conclusions Casteddu quarry and those used for the tiers of cavea, and
by possibility that the Romans (as they often used to do)
The results of the research have allowed defining the geo- used the near Rio Pula river and then the sea to transport the
chemical characteristics and the origin of the volcanics that material from the quarry to Nora site.
have been used in the construction of the Roman Nora city. In different way, the volcanics, which have been used
These rocks belong to subalkaline series of the Sardinian for the Nora artefacts, mainly come from the more nearly
Late Eocene–Miocene volcanic cycle, with a mainly sodic countryside areas. Due to their higher physical–mechani-
affinity and more rarely potassic. The samples from the cal resistance, for the stone paving of the main inner street,
Roman theatre were mainly classified (in order of abun- the Roman Foro and parts of the other buildings (i.e. base
dance) as dacites and, subordinately, as lati-andesites, rhyo- of public fountain), more welded andesitic–dacitic volcanic
dacites, andesites and latites, while the samples taken from rocks (as autobreccias) were used, mainly coming from
the other masonries of Nora site are classified as dacites, Perdu Pranu area.
andesites, rhyodacites and basaltic andesites. To construct the masonry of the houses and commercial
The andesitic and dacitic volcanics were mainly used as area to the west seaport (i.e. Macellum horreum), andesitic
construction material for artefacts of Nora (i.e. Roman stone rock outcroppings inside the Nora site were used, especially
paving of roads inside the village) but also as aggregate in as irregular ashlars of the walls.
the mortars (i.e. wall bedding mortars, concretes of vaults For frames and door thresholds of Nora buildings, the
and cavea foundation, plasters) together obsidian rocks, Romans used different volcanic materials, mainly repre-
quartz–feldspar sandy, Paleozoic rocks. sented by welded volcanic breccias and conglomerates
To define their provenance, the volcanics from Nora coming from Nora area and surrounding outcrops, together
archaeological site have been compared with those from the occasionally grey fine sandstones (i.e. thresholds of Macel-
surrounding areas and from Sulcis basin, more frequented lum horreum).
by the Romans, and in which the Nora territory is included. The supply of the raw materials from different areas
Chemical and petrographic data highlight that the dacitic around the Nora site highlights that the Romans chose the
rocks (represented by volcanic autobreccias), which have stones for the construction of ancient Nora village according
been used on the Nora theatre tiers, come from the Su to their petro-physical characteristics which clearly affect
Casteddu volcanic apparatus, located 2 km at north-east of the workability and quarry extractability. Thus, consider-
Nora site. This assumption is also confirmed by the com- ing only the sources around Nora site (or at a reasonable
parable material volumes of extracted stone from the Su distance from it) and only within them, the distance between

13
577 
Page 18 of 19 Environmental Earth Sciences (2018) 77:577

the quarry and the building site is not the main factor leading Carminati E, Doglioni C (2012) Alps vs. Apennines: the paradigm of a
material selection. tectonically asymmetric Earth. Earth Sci Rev 112:67–96
Cherchi A, Montadert L (1982a) Il sistema di rifting oligo-miocenico
del Mediterraneo occidentale e sue conseguenze paleogeografiche
Acknowledgements  Special thanks to the University of Cagliari for sul terziario sardo. Mem Soc Geol Ital 24:387–400
funding this research; the Superintendence for Archaeological Herit- Cherchi A, Montadert L (1982b) Oligo-Miocene rift of Sardinia
age for the Provinces of Cagliari and Oristano for authorization to and the early history of the western Medittanean basin. Nature
sample the materials from the monument; the staff and tourist guides 298:736–739
of the archaeological site of Nora for their willingness regarding the Columbu S (2017) Provenance and alteration of pyroclastic rocks from
study activities. the Romanesque Churches of Logudoro (north Sardinia, Italy)
using a petrographic and geochemical statistical approach. Appl
Phys A Mater Sci Process 123(3):165. https​://doi.org/10.1007/
s0033​9-017-0790-z
References Columbu S. Garau AM (2017) Mineralogical, petrographic and chemi-
cal analysis of geomaterials used in the mortars of Roman Nora
Alvarez JI, Navarro I, Martin A, Garcia Casado PJ (2000) A study of theatre (south Sardinia, Italy). Ital J Geosci 136:238–262
the ancient mortars in the north tower of Pamplona’s San Cernin Columbu S, Sitzia F, Verdiani G (2015) Contribution of petrophysical
Church. Cem Concr Res 30:1413–1419 analysis and 3D digital survey in the archaeometric investiga-
Antonelli F, Columbu S, De Vos Raaijmakers M, Andreoli M (2014) tions of the Emperor Hadrian’s Baths (Tivoli, Italy). Rend Fis
An archaeometric contribution to the study of ancient millstones Acc Lincei 26:455–474
from the Mulargia area (Sardinia, Italy) through new analytical Columbu S, Sitzia F, Ennas G (2017) The ancient pozzolanic mortars
data on volcanic raw material and archaeological items from Hel- and concretes of Heliocaminus baths in Hadrian’s Villa (Tivoli,
lenistic and Roman North Africa. J Archaeol Sci 50:243–261 Italy). Archaeol Anthropol Sci 9:523–553
Assorgia A, Brotzu P, Morbidelli L, Nicoletti M, Traversa G (1984) Columbu S, Garau AM, Lugliè C (2018) Geochemical characterisation
Successione e cronologia (K-Ar) degli eventi vulcanici del comp- of pozzolanic obsidian glasses used in the ancient mortars of Nora
lesso calco-alcalino Oligo-Miocenico dell’Arcuentu (Sardegna Roman theatre (Sardinia, Italy): provenance of raw materials and
centro-occidentale). Period Mineral 53:89–102 historical–archaeological implications. Archaeol Anthropol Sci.
Assorgia A, Barca S, Onnis G, Secchi FAG, Spano C (1986) Episodi https​://doi.org/10.1007/s1252​0-018-0658-y
sedimentari e vulcanici oligo-miocenici nel settore occidentale Conte A (1989) Studio petrografico, mineralogico e geochimico delle
dell’Arcuentu e loro contesto geodinamico (Sardegna SW). Mem rocce magmatiche calcoalcaline dei distretti di Sarroch (Sardegna
Soc Geol It 35(10):229–240 meridionale) e di S.Antioco (Sardegna Sud-Occidentale), Ph.D.
Assorgia A, Fadda A, Gimeno Torrente G, Morra V, Ottelli L, Sec- Thesis. Roma University, Italy
chi FA (1990) Le successioni ignimbritiche terziarie del Sulcis Coulon C (1977) Le volcanism calco-alcalin cenozoique de Sard-
(Sardegna sud-occidentale). Mem Soc Geol Ital 45:951–963 aigne, Italie. Unpublished master’s thesis. Univesité St. Jerome,
Beccaluva L, Civetta L, Macciotta G, Ricci CA (1985) Geochronology Marseille
in Sardinia: results and problems. Rend Soc Ital Mineral Petrol Criss JW (1977) NRLXRF: a fortran program for X-ray fluorescence
40:57–72 analysis. Cosmic, Athens
Beccaluva L, Brotzu P, Macciotta G, Morbidelli L, Serri G, Traversa G Cross W, Iddings JP, Pirsson LV, Washington HS (1903) Quantita-
(1989) Cainozoic tectono-magmatic evolution and inferred mantle tive classification of igneous rocks. University of Chicago Press,
sources in the Sardo-Tyrrenian area. In: Boriani A, Bonafede M, Chicago
Piccardo GB, Vai GB (eds) The lithosphere in Italy. Advances De Luca R, Miriello D, Pecci A, Domìnguez-Bella S, Bernal-Casasola
in Earth Science Research, vol 80. Atti Conv. Acc. Naz. Lincei, D, Cottica D, Bloise A, Crisci GM (2015) Archaeometric study
Rome, pp 229–248 of mortars from the Garum Shop at Pompeii, Campania, Italy.
Beccaluva L, Bianchini G, Coltorti M, Siena F, Verde M (2005a) Ceno- Geoarchaeology 30:330–351
zoic Tectono-magmatic evolution of the Central western Mediter- De La Roche H, Leterrier J, Grandclaude P, Marchal M (1980) A clas-
ranean: migration of an Arc-interarc Basin System and Variations sification of volcanic and plutonic rocks using ­R1–R2 diagram and
in the Mode of Subduction. In: Finetti I (ed) Elsevier Special major-element analyses—its relationships with current nomencla-
Volume, “Crop Project—Deep Seismic Exploration of the Central ture. Chem Geol 29:183–210
Mediterranean and Italy”, pp 623–640 Dostal J, Coulon C, Dupuy C (1982) Cainozoic andesitic rock of Sar-
Beccaluva L, Bianchini G, Bonadiman C, Coltorti M, Macciotta G, dinia (Italy). In: Thorpe RS (ed) Andesites: orogenic andesites
Siena F, Vaccaro C (2005b) Within-plate Cenozoic Volcanism and related rocks. Wiley, Chichester, pp 353–370
and Lithospheric Mantle Evolution in the Western-central Medi- Franzini M, Leoni L, Saitta M (1975) Revisione di una metologia
terranean Area. In: Finetti I (ed) Elsevier Special Volume, “Crop analitica per fluorescenza-X, basata sulla correzione completa
Project—Deep Seismic Exploration of the Central Mediterranean degli effetti di matrice. Soc Ital Mineral Petrol Rendiconti
and Italy”, pp 641–664 XXXI(2):365–378
Beccaluva L, Bianchini G, Natali C, Siena F (2011) Geodynamic Franzini M, Leoni L, Lezzerini M, Sartori F (2000) The mortar of the
control on orogenic and anorogenic magmatic phases in Sardinia “Leaning Tower” of Pisa: the product of a medieval technique
and Southern Spain: inferences for the Cenozoic evolution of the for preparing high-strenght mortars. Eur J Mineral 12:1151–1163
western Mediterranean. Lithos 180–181:128–137 Garau AM (2005) Le pietre e le malte del teatro di Nora (Sardegna
Bejor G (1999) L’area del teatro. In: Tronchetti C (ed) AA. VV. SW). Unpublished doctoral dissertation. University of Cagliari,
Ricerche su Nora-I (anni 1990–1998). Sainas Ed, Cagliari Cagliari
Bernardini P (2004) Cartagine e la Sardegna. Dalla conquista Garbarino C, Lirer L, Maccioni L, Salvadori I (1990) Carta vulcanolog-
all’integrazione (540–238 a.C.). RS-tFen 32:35–56 ica della’Isola di San Pietro (Sardegna). Scala 1/25000. SELCA
Bianchini G, Marrocchino E, Vaccaro C (2004) Chemical and mineral- Eds, Firenze
ogical characterisation of historic mortars in Ferrara (NE, Italy). Gattaceca J, Deino A, Rizzo R, Jones DS, Henry B, Beaudoin B, Vade-
Cem Concr Res 34(8):1471–1475 boin F (2007) Miocene rotation of Sardinia: new paleomagnetic

13
Environmental Earth Sciences (2018) 77:577 Page 19 of 19  577

and geochronological constraints and geodynamic implications. Riccardi MP, Duminuco P, Tomasi C, Ferloni P (1998) Thermal,
Earth Planet Sci Lett 258:359–377 microscopic and X-ray diffraction studies on ancient mortars.
Irvine TN, Baragar WRA (1971) A guide to the chemical classification Thermochim Acta 321:207–214
of the common volcanic rocks. Can J Earth Sci 8:523–548 Savelli C (1975) Datazioni preliminari col metodo K-Ar di vulcaniti
Kuno H (1968) Differentiation of basalt magma. In: Hess H, Polder- della Sardegna sud-occidentale. Rend Soc Ital Mineral Petrol
vaart A (eds) The Poldervaart treatise on rocks of basaltic com- 31:191–198
position, vol II. Wiley, New York, pp 623–688 Savelli C, Beccaluva L, Deriu M, Macciotta G, Maccioni L (1979)
Le Bas MJ, Le Maitre RW, Strekeisen AL, Zanettin B (1986) A chemi- K/Ar geochronology of the tertiary “Calc-Alcalik” volcanism of
cal classification of volcanic rocks based on the total alkali-silica Sardinia (Italy). J Volc Geotherm Res 5:257–269
diagram. J Petrol 27:745–750 Shand SJ (1951) Eruptive rocks; their genesis, composition, classifica-
Le Maitre RW, Streckeisen A, Zanettin B, Le Bas MJ, Bonin B, Bate- tion, and their relation to ore-deposits, with a chapter on meteor-
man P, Bellieni G, Dudek A, Efremova S, Keller J, Lamere J, ites, 4th edn. Wiley, New York
Sabine PA, Schmid R, Sorensen H, Woolley AR (2002) Igneous Speranza F, Villa IM, Sagnotti L, Florindo F, Cosentino D, Cipol-
rocks: a classification and glossary of terms. Recommendations lari P, Mattei M (2002) Age of the Corsica-Sardinia rotation and
of the International Union of Geological Sciences, Subcommis- Liguro-Provençal Basin spreading: New paleomagnetic and Ar/
sion of the Systematics of Igneous Rocks. Cambridge University Ar evidence. Tectonophysics 347(4):231–251
Press, Cambridge Streckeisen A (1976) Classification of the common igneous rocks
Lustrino M, Duggen S, Rosenberg CL (2011) The Central-Western by means of their chemical composition. Neues Jahrb Mineral
Mediterranean: anomalous igneous activity in an anomalous col- Monatsh, pp 1–15
lisional tectonic setting. Earth Sci Rev 104:1–40 Streckeisen A (1978) Classification and nomenclature of volcanic
Maravelaki-Kalaitzaki P, Bakolas A, Moropoulou A (2003) Phisico- rocks. I.U.G.S. (Int. Union Geol. Sci.), Subcomm. on the Sys-
chemical study of Cretan ancient mortars. Cem Concr Res tematics of Igneous Rocks. Inter. Rep. Contrib., p 70
33:651–661 Thornton CP, Tuttle OF (1960) Chemistry of igneous rocks--[Part] 1,
Mastino A (2005) Storia della Sardegna antica. Ed. Il Maestrale, Differentiation index. Am J Sci 258:664–684
Sassari Tronchetti C (2001) Nora. Collana Sardegna Archeologica. Carlo Del-
Middlemost EAK (1975) The basalt clan. Earth Sci Rev 11:337–364 fino Editore and C, Sassari
Moropoulou A, Bakolas A, Bisbikou K (2000) Investigation of the Vardabasso S, Atzeni A (1962) Il bacino Ooligocenico di Oschiri-Ber-
technology of historic mortars. J Cult Heritage 1:45–58 chidda nella Sardegna nord-occidentale. Mem Soc Geol Ital 3:717
Peccerillo A, Taylor SR (1976) Geochemistry of eocene calc-alkaline Verdiani G, Columbu S (2010) E.Stone, an archive for the Sardinia
volcanic rocks from the Kastamonu area, Northern Turkey. Con- monumental witnesses. Lect Notes Comput Sci 6436:356–372
trib Mineral Petrol 58(1):63–81 Vola G, Gotti E, Brandon C, Oleson JP, Hohlfelder RL (2011) Chemi-
Pesce G (1954) Scavi nell’area della città, 1952–1954. In: Fasti Arche- cal, mineralogical and petrographic characterization of roman
ologici 9. Editrice Belriguardo Libreria, Ferrara, pp 356–358 ancient hydraulic concretes cores from Santa Liberata, Italy, and
Pesce G (1972) Nora. Guida agli scavi, II. Fossataro Ed, Cagliari Caesarea Palestinae. Israel Period Mineral 80:317–338
Recommendations Nor.3/80 MaL (1980) Stone materials: sampling.
CNR-ICR, Roma

13