TECHNICAL REPORT NO.

No. 33

MINISTRY OF ENVIRONMENT AND NATURAL RESOURCES

MINES AND GEOLOGICAL DEPARTMENT

GEOLOGY AND
GROUND WATER CONDITIONS
IN THE

NAKURU AREA

by
by

G.J.H. McCall, B.Sc., Ph.D., A.R.C.S., D.I.C.

Geologist

First print 1957

Reprint 2007
 GEOLOGY AND
GROUNDWATER CONDITIONS
IN THE

NAKURU AREA

By

G. J. H. McCALL, B.Sc., Ph.D., A.R.C.S., D.I.C.

Geologist
 FOREWORD
This report on the geology and groundwater conditions in the Nakuru/Elmenteita
region of the Rift Valley is the first of a series, which, in time, will cover the more
important agricultural areas of Kenya. Although this report makes use of all the
records of drilling operations to date, this has been supplemented by a considerable
amount of field work. Much of the new work done is geological, but it has long been
realized that the problems of groundwater cannot be understood without detailed
geological mapping and an assessment of the structural geology of the region.
The conclusions reached in this report are by no means complete. Two important
questions remain to be answered : —
(i) How much of the annual rainfall passes the capillary zone of the ground to
become available as groundwater? (X.)
(ii) How much groundwater is lost by evaporation or escapes from the ground-
water bodies into areas of less economic significance or to uneconomic
depths? (Y.)
In our present state of knowledge answers cannot be given to these questions.
However, the difference (X—Y) represents the maximum amount of groundwater
available for development, and is thus still indeterminate.
The present study indicates that there is probably a subterranean watershed at
the southern boundary of the area and there is little possibility of appreciable escape
from the east or west flanks. The subterranean escape of water to the north is possible,
but its extent cannot yet be assessed. It can, however, be said that the development
of the groundwater resources in the area can safely proceed without any immediate
risk of overwithdrawal.
As development progresses and more water is drawn from the groundwater
reservoir, the rest level of the water-table is likely to fall. A state of equilibrium may
be presumed at the present time, where the groundwater not being used, now escapes
in spring flows, by evaporation from the lakes or by subterranean flow. To enable
more groundwater to be used, there must be a readjustment of the water-table, small
though this may be, to decrease losses. The use of more groundwater may, for example,
lower the present lake levels. There could, however, come a stage when, without
increasing the utilization of groundwater, the water—table continues to recede. This is
the danger point, for then the groundwater body is being drawn upon in excess of
recharge, and disaster is not far away.
The effects of the future development of the groundwater resources of this area
must be kept-under observation and a halt called before a harmful recession of
the groundwater table commences. This situation is not likely to occur for some
considerable time at the present rate of development.
The author, at the time this work was undertaken, was a geologist working with
the Groundwater Section of the Hydraulic Branch of the Public Works Department,
under the Senior Geologist, Mr. S. Stock, A.R.S.M. Facilities provided by the Depart-
ment of Mines and Geology for some of the more detailed petrographical studies are
acknowledged.

Nairobi, H. J. SQUIRES,
2nd May, 1957. Chief Hydraulic Engineer.
 ' CONTENTS

INTRODUCTION
II
HUMAN SETTLEMENT AND AGRICULTURAL DEVELOPMENT
III
GENERAL SUMMARY OF THE GEOLOGICAL HISTORY OF THE RIFT VALLEY IN
THENAKURUAREA .. .. .. .. .. 3
IV GEOLOGICAL FORMATIONS—VOLCANIC ROCKS .. 5
THE MENENGAI VOLCANO 12
VI THE EEURRU AND ELMENTEITA VOLCANIC SERIES 21
VII ' THE SEDIMENTS 24
VIII TECTONICS 28
IX HYDROLOGY 30
HISTORY AND RESULTS OF GROUNDWATER DEVELOPMENT 34
XI DESCRIPTION OF AQUIFERS . .. .. 35
XII GROUNDWATER CONDITIONS. . 37
XIII GEOPHYSICAL METHODS OF PROSPECIING FOR WATER 40
XIV SODA DUST FROM LAKE NAKURU .. ‘ 41
XV THE NJORO RIVER 42
XVI THE POSSIBILITY OF UTILIZING STEAM AS A SOURCE OF POWER IN THE RIFT
VALLEY. . . . 43
APPENDIX—~ANALYSIS OF DRILLING RESULTS 45
NOTE.—Full details of all boreholes drilled In the area covered by this report appear at the
end of the Appendix.

ILLUSTRATIONS
PAGE
Fig. 1 .—Map and Serial Sections of a small fault block, 1% miles west of Kariandusi.
(After B. N. Temperley) 11
Fig. 2 .—Subsidence Calderas. The Mechanism of Formation. (After A. Holmes,
van Bemmelen and H. Williams.) . . .. .. . . 15
Fig. 3.—Menengai Crater View from the western rIm . . 17
Fig. 4.-——(a) Sketch Of Subsidence Structure on western side Of Caldera. 18
(b) Sketch showing the reversal of the normal outward slope of lava floWs
in a promontory projecting from the western wall of Menengai Caldera,
consequent on the foundering of the central part of the volcano 18
Fig. 5.——Sketch showing the renewed faulting Of basaltic tuffs Of “High” volcano. . 25
Fig. 6.—Sequence Of changes of the Rift Valley lakes. (After E Nilsson) . . . 26
Fig. 7.——AS above . . 27
Fig. 8.—ISOhyets Of average rainfall for the 10 year period 1943—1952 . . 31
Fig. 9.—Diagrammatic representation Of the hydrological cycle In the Rift Valley 33
Fig. 10.——Comparison of a confined aquifer producing artesian conditions and the
open system of aquifers typical of the floor Of the Rift Valley 38
Fig. 11 —Comparison of resistivity curves in boreholes drilled In the Nakuru Area 41
Plate l.—Mould of tree (completely vaporised) preserved In welded tufl‘ flow, Nakuru
Municipality Quarry, Ravine Road 12

MAPS
1. Geological map of Nakuru Area.
2. Map showing groundwater conditions in the Nakuru Area.
3. Geological Sketch Map of the Menengai Caldera.
4. Geological Sketch Map of the basaltic cones of Elmenteita.

SECTIONS
Geological Sections illustrating groundwater conditions in the Nakuru Area.
 ABSTRACT
An area of approximately 1,200 square miles in the Rift Valley near Nakuru has
been studied in detail in order to gain more information aboutgroundwater conditions.
The geology of the area, so far but briefly referred to by J. W. Gregory, has been
described in some detail. The nature of the aquifers and the hydrostatic conditions
have been described, and the effect on groundwater conditions of the Menengai and
Elmenteita volcanoes, still active to a very minor degree, has been assessed. The failure
of boreholes in the Kampi-ya-Moto area is explained, and the presence of ground-
water ridges and perched water-tables under influent steams has been described.
The geology and groundwater conditions have been illustrated by maps and
sections. A hydrological cycle in the internal drainage system of the Rift Valley has
been suggested.
Detailed geological maps have been drawn of the Menengai and Elmenteita volcanic
centres, both the source of lava eruptions within the last few hundred years.
Geophysical methods of prediction have been discussed, and are considered to be
of little value in the prevailing geological conditions. Brief notes on the ground-
water conditions in the Lake Nakuru basin and their hearing on the soda dust problem,
and the possibilities of natural steam as a source of power in the Rift Valley are
included.
 GEOLOGY AND GROUNDWATER CONDITIONS IN THE
NAKURU AREA
By G. J. H. McCall, PILD.

CHAPTER I—INTRODUCTION
This report covers the geology and the groundwater
conditions in an area of the
Rift Valley near Nakuru lying just south of the Equato
r and intersected by the
36° E. meridian.

The field survey was carried out in 1952—53 as part

of a detailed investigation
of the geological and»groundwater conditions in the
northern part of the settled
area within the Rift Valley. The area was chosen for the
following reasons:—
(i) It is an area in which geophysical methods of predicti
on have proved to be
of little value.
(ii) The extent and geological relationships of thermal
conditions in this part
of the Rift Valley, particularly those of the recent volcanic centre
of Menengai,
required investigation.
(iii) An extensive area near Kampi-ya-Moto had failed to produce
a single success-
ful borehole, the reasons for which were obscure.

The limits of the area are: the Man Escarpment to the west; the Eburru
Volcano
to the south; and the Kamasia Reserve to the north. The
eastern limit was taken
along the line of the Subukia—Bahati Forest Escarpment and
its southerly extension
in the Gilgil Escarpment. A part of the Njoro area has also
been included. For
convenience the area has been divided into six sub-sections:—

(1) Nakuru.
(2) Rongai.

(3) Solai—North Menengai.

(4) Kampi-ya-Moto—Lomolo.

(5) Elmenteita.

(6) Njoro (North).

The geological survey was carried out by means of detailed mapping

of surface
exposures, together with microscopic examination of all available
samples from com-
pleted boreholes. As more than 150 boreholes have been drilled ample
data is avail-
able for such a study of groundwater conditions in the extensively
faulted Tertiary
volcanic rocks.

Topographical detail was obtained from air photographs, military

1:25,000 maps
and preliminary plots issued by the Survey Department of Kenya.

Altitudes of boreholes were tied in to known heights by means

of aneroid measure-
ments, but owing to the large diurnal variation no very high
degree of accuracy
is claimed for these measurements. The accuracy is sufficien
t, however, to provide a
picture of the subsurface hydrostatic conditions throughout
the area. Many of the
boreholes in the vicinity of Nakuru town have subsequently
been accurately surveyed
by the Hydrology Section and these results are incorporated
in this report.
 2

survey has been completed,

This is the first area in which a detailed groundwater
carried out in other areas of the Colony.
but similar investigations are being

are to be found in the

References to the area in earlier geological literature
following publications : —
, 1921.
The Geology of the Rift Valley of East Africa, I. W. Gregory
L. Sikes, 1934.
The Underground Water Resources of Kenya Colony, H.
, 1931 (with geological
The Stone Age Cultures of Kenya, L. S. B. Leakey
summary in appendix by J. D. Solomon).
Nilsson, 1932.
Quaternary Glaciations, and Pluvial Lakes in East Africa, E.

An Outline of the Geology of Kenya, S. M. Cole, 1950.

East African Lakes, L. S. B. Leakey, 1932.

being made in the

This report was revised in March, 1957, certain amendments
geological investigations
light of new data obtained from further boring for water and
carried out near Gilgil by Dr. B. N. Temperley.

CHAPTER II—HUMAN SETTLEMENT AND AGRICULTURAL

DEVELOPMENT
and large areas
With the exception of the Kamasia Native Reserve to the north
entirely
of forest on the Mau and Bahati Escarpments, the Nakuru area is almost
area. Mixed farming
alienated for European settlement. Farming is the .mainstay of the
ates in the richer and more fertile districts of Solai and Rongai, while larger
predomin
grassy plains in the vicinity of Lake Nakuru and
ranching farms occupy the dry,
coffee flourishes
Elmenteita. Wheat and maize are the main cereal crops grown while
for feeding cattle is grown by irrigation from surface
in the Solai Valley. Lucerne
country on the
streams and boreholes. Sisal plantations occupy a large tract of dry
01 Punyata, and
edge of the Kamasia Reserve near Kampi—ya-Moto, Lower M010 and
another 8,000-acre plantation is situated on the west side of Lake Nakuru.

industries
Nakuru is the centre of this large farming area, having as yet few
tanning, wool
other than those related to the farming operations in the district (e.g.
centre, and the
treating, etc.). It is a flourishing and rapidly expanding commercial
administrative centre of the Rift Valley Province.

The main demand for water comes from the farming community. The rainfall
soil over
over the area is not heavy and is liable to fail over successive years. The
of the area is too porous to allow the storing of water in small surface
much
and in these
reservoirs. Certain parts of the area are well watered by surface streams,
such areas
localities there has been little recourse to groundwater resources. Amongst
and Rongai Rivers and the Bahati Plain. However, small
are the valleys of the M010
stream flows combined with stream losses have brought forward proposals for the
piping of water from these streams to the farms in order to increase the amount of
Stream and
water available for development. Schemes are in being from the Crater
A large scheme has recently been completed from the Rongai River,
the Njoro River.
above
the flow of which has been augmented by a tunnel driven a mile into the hills
pipe—line schemes are being constructe d on the Westacre, Vissoi
Elburgon. Further
and Olobanaita Rivers.
 3

Unfortunately these schemes, although important, cannot relieve the water shortage
in such areas where deep boring has so far been a complete failure. One of the
objects of this report is to explore the possibility of extending the areas which can
be supplied from groundwater reserves; Of the 146 boreholes completed in this
area at the time the field work for this report was carried out, 118 were for the purpose
of farm development. Although the first rush of post-war development has passed, the
demand for the further development of groundwater resources will continue for many
years to come, especially in those areas where groundwater is the only practical source
of increased water supplies.
Twenty-eight boreholes have been drilled for purposes other than the development
of the farming industry. These include such purposes as municipal supply, Nakuru
industries (wool treatment, tanning, etc.), cooling water for electricity power supply,
saw-mills, residential plots and road constructions. The Nakuru municipal water
supply uses both river and borehole waters. The fluorine content of the ground-
water alone is considered to be too high for a public water supply. Very little close
drilling has been carried out except in the Mereroni area near the municipal boundary
of Nakuru. The boreholes in this locality have high yields and are used by the
Municipality of Nakuru. Although there is a danger of overpumping in this locality
it has not yet taken place.

CHAPTER III—GENERAL SUMMARY OF THE GEOLOGICAL HISTORY

OF THE RIFT VALLEY IN THE NAKURU AREA V
The area surveyed lies within the main Rift Valley of Kenya, known to geologists
as the Gregory Rift Valley. This structure, which is part of the greater Rift System
extending from Syria to the Zambesi, is a complex fault trough with a general north—
south orientation. The rift valleys do not form unbroken north—south troughs, the
system being broken up by oblique structures and shorter fault troughs normal to the
main trend, of which the Gulf of Aden and the Kavirondo Rift Valley are among
the best examples. These lateral and oblique structures are in most cases associated
with a lateral offset of the main Rift Valley. They have been discussed by R. B.
McConnell [4] who attributed them to earlier shear zones in the Basement platform,
caused by a shearing couple resultant on ancient orogenic movements of great magni—
tude. McConnell considers that the Tertiary structures utilized the same lines of
weakness, although resultant on a much later orogenic adjustment of the earth’s crust.

Nakuru is situated at the intersection of the Kavirondo Rift Valley, a fault trough
extending westwards to the shores of Lake Victoria, with the Gregory Rift Valley. '
At this junction the Rift Valley is complicated by two factors. Firstly, a definite deflec-
tion in its course can be observed resulting in'more complex fault structures than are
normally seen in the straighter sections of its course. Secondly, as in the case of
the “Mbeya Angle” in Tanganyika and other intersections, the junction has been
the focus of intense vulcanism, evidenced by four major volcanoes—Loldiani, Kilombe,
Menengai and Eburru. Loldiani stands astride the junction of the two Rift Valleys,
while Menengai and Kilombe lie to the east,- rising from the floor of the main Rift
Valley. Eburru, a twin-peaked massif, is situated slightly to the south, forming a
barrier across the Rift Valley floor. Later eruptions have been on a smaller scale
and mainly confined to renewed activity in the earlier centres. However, a later group
of volcanoes developed along a zone of intense north—south faulting extending from
Lake Elmenteita over the eastern shoulder of Eburru to the west shore of Lake
Naivasha. Basaltic tutf and cinder cones have arisen, and basalt lava flows have been
poured out in successive eruptions extending to comparatively recent times.
The Tertiary history of the Rift Valley commenced with downwarping of narrow
troughs on the site of the present-day rift valleys. This downwarping interrupted the
drainage of that part of the African Shield, and a series of shallow lakes were formed;
 4

the deposits formed in these lakes are the Miocene Lake Beds of Rusinga Island,
Tambach and Turkana, which rest upon a peneplained surface of ancient metamorphic
rocks. Among the earliest lavas to be poured out were the plateau lavas, phonolite
flows extending over vast areas of peneplained Basement Complex, and the first Tertiary
nephelinite volcanoes. These are the Laikipia and Uasin Gishu phonolites which directly
overlie the Basement Complex and Miocene Lake Beds, on the eastern (Elgeyo) escarp-
ment of the Rift Valley to the north of Nakuru.

The plateau phonolites have no visible source of origin, but their extent and
uniformity of composition leads to the conclusion that they were erupted from fissures
which opened during the earliest tensional phase of the rifting. Plateau phonolites
must cover great areas of the down-faulted peneplained surface of the ancient Basement
rocks in the Rift Valley, but in the Nakuru area no comparable types are seen and
it is presumed that they have been completely obscured by later effusions of lava
and pyroclastic deposits. Near Lake Hannington, however (to the north of Solai), the
Losuguta phonolites, which closely resemble the plateau phonolites, emerge in a great
north—west—facing scarp, a feature supporting this assumption.

,The eruptions from fissure sources appear to have continued for a considerable
geological period on the Rift floor after the first great fault displacements occurred.
The oldest volcanic rocks exposed on the valley floor in the Nakuru area are phono-
lites, basalts and phonolitic trachytes, extensive in outcrop and uniform in composition,
having no central sources visible at the present day. It is thought that movements,
along faults were closely associated with a welling out of lava from fissures along the
margins and across the floor of the Rift. These flows are strongly faulted and form
the clear-cut scarps and grid structures. It is, however, the opinion of the writer that
these are not among the earliest fault structures; a view that is borne out by the fact
that these faults do not exactly coincide in direction with the main marginal scarps
of the Mau, and the Bahati—Subukia Horst, an outlying spur of the eastern wall which
both from the tectonic viewpoint and from the nature of its component rocks is con-
sidered to be part of the eastern wall of the Rift Valley.

After the faulting had diminished in magnitude the great trachyte volcanoes arose.
These are not affected by major faults though slight subsidences along earlier lines of
weakness have occurred. Loldiani and Eburru have no well-preserved craters, and
their slopes are appreciably dissected by erosion. Kilombe is equally dissected, but
appears to possess a comparatively well-marked, though eroded, crater. Menengai
has dissected slopes, deep gullies cutting into the loose pyroclastic mantle. The present-
day appearance of Menengai difiers from the other three in that it is the only one of
the volcanoes that has been affected by cauldron subsidence. No detailed description
of these volcanoes (other than Menengai) is given in this report, being beyond the
scope of this survey.

The later central eruptions are characterized by tongue-like lava flows of restricted
extent, the source and the termination of these flows being easily determined when
mapping in the field. Renewed vulcanism has occurred in Menengai, extending up
to the last few hundred years, and this may also be true Of some of the other
volcanoes. Subsequent to the great volcanoes being built up eruptions were much
more limited in extent—and only the small volcanic cones of Elmenteita and Honey-
moon Hill arose. These do not exceed a mile in diameter, and many a smaller. The
largest, named High, stands about 500 feet above the surrounding country, but the
majority do not stand up more than two to three hundred feet. The outer slopes,
composed of compacted tufl‘, are not appreciably gullied, but the cones have been
affected by minor faulting. Present—day volcanic activity is limited to minor steam vents
in the volcanic zones of Menengai and Elmenteita.
 5
I

The tectonic history and allied vulcanism comprises only part of the picture,
for, concurrently sedimentation took place. Since the time when the first Miocene
warp movements interrupted the drainage pattern of the African shield, the Rift
Valley has been an area of internal drainage. Successions of shallow lakes formed
only to be deformed by earth movements, broken up by faulting and dammed up by
lava flows. Into these ephemeral lakes volcanic detritus was carried by short and
intermittent streams flowing off the higher ground. Thus the volcanic rocks of the
Rift Valley are intercalated with sediments—volcanic grits, clays and reworked tuffs.
In the Pleistocene era a great lake extended over the Gilgil Escarpment from Menengai
to Lake Naivasha and since that time a series of separate lakes have occupied its
site, diminishing gradually to the present shallow lakes of Naivasha, Nakuru and
Elmenteita.

The continuation of earth movement in the Rift Valley up to the present day
is well demonstrated in Nilsson’s [5] work on the Pleistocene glaciations of East Africa.
Earthquakes and fissuring of the land surface have occurred during the last 30 years
and the Nakuru area must still be considered as a zone of crustal instability, though
of minor order compared with the world‘s great earthquake zones.

CHAPTER IV—GEOLOGICAL FORMATIONS—VOLCANIC ROCKS

The volcanic rocks exposed in the Mau and Bahati‘Subukia Escarpments, and
those of the Legisianan Escarpment are thought to be the oldest volcanic rocks exposed
within the area. The succession is as follows:— -

(a) The Older-faulted Volcanic Rocks

(3) Man and Bahati tufls.
(2) Solai Phonolite.
(l) Solai Basalt.

Reconnaissance of the Mau has revealed that the bulk, if not all, of the volcanic
rocks near the surface consist of a series of greenish-grey welded trachytic tufis,
together with yellow pumic tulfs, and sedimentary intercalations, reworked tuffs and
clay. On the Bahati Escarpment between Mbaruk and Gilgil similar welded trachytic
tulfs and yellow pumice tulf intercalations are the only formations exposed; while to
the north onbaruk and at Solai they overlie black porphyritic phonolites. To
the north of Milton’s Siding the welded tufis disappear altogther and extensive flows
of the porphyritic phonolites form the surfaCe of the fault blocks.

The phonolites underlying the Man and Bahati luff formation have been named
“Solai Phonolite”. They consist of a series of flows characterized by abundant ideo-
morphic nepheline phenocrysts easily identifiable in thin section, and often to the
naked eye. The phonolite exposed on the Subukia escarpment is black, massive and
characterized by prominent tabular felspar phenocrysts. Numerous major faults are
exposed along the escarpment and in the vicinity of these dislocations the phonolite
is either smashed or altered to a dove-grey to white closely jointed alteration product,
with the tabular felspar phenocrysts still preserved. .

The Legisianan Escarpment is capped by a thin layer of welded trachytic tuff

dying out to the north and underlain by a succession of phonolite flows. These include
porphyritic phonolite similar to the phonolite of the Subukia Escarpment, and finer
textured varieties devoid of felspar phenocrysts. Locally, streaky textures are developed.
A characteristic type of phonolite exposed on this escarpment shows abundant pheno-
crysts of nepheline surrounded by a ring of green aegirine au‘gite, felspar phenocrysts
being sparse or absent.
 6
ents are composed of
The fault blocks in the valley between the two escarpm
tabular felspar phenoc rysts matchi ng the phonol ite of the Subukia
black phonolite with is exposed
of the same phonolite
Escarpment; to the south a solitary fault scarp
farm. To the west of the Legisia nan Escarpment black phonolites out—
on Mascall’s
are freque ntly fissile and rarely porphyritic; the
crop as far West as Lomolo; these
surface is, howeve r, overlai n by patches of welded trachytic tufi, and there
phonolite
part of the same series of phonolitic flows which
can be little doubt that they are
the Legisia nan Escarp ment. The relation of the welded tufi to the phonolite
outcrop on
may have been already down-
(brought out in the mapping) suggests that the phonolite ed. For in the
warped or faulted at the time at which the welded tuffs were deposit
rapidly as one passes away from
Solai and Subukia areas the welded tufl dies out
is laid bare on the surface. The
the line of the now hidden older fault, and phonolite
appare ntly been banked against an existing escarpment. Pre—
welded tuff has thus
near Elburg on and Londiani suggest
liminary reconnaissances of the Mau Escarpment
out and disappears as
that here also the very deep Welded tuff formation thins
soon as the crest of the Mau is traversed.
lies east of McKenzie’s farm.
Streaky phonolites form the long fault block which
Legisianan Escarp ment. They were first mapped as phono-
These resemble those of the
rations , and their relation to the
litic trachyte, but it is clear from structural conside of the Solai
of older rocks
welded tuffs, that the fault block must be made up trachyte
Phonolite group, though there are probably a few small outliers of phonolitic
on the phonolite at the south end,
of the later Nakuru group lying unconformably
follows the line of the Nakuru—01 Punyata
near to the line of the older fault, which here
road.
d on the crest of the road
The basalt underlying the Solai Phonolite is expose
sion in the road cutting is:—
pass from Milton’s Sidings to Subukia. The succes
(2) Green massive phonolite.
(1) Buff basaltic ash.
ted vesicles infilled with
Aphanitic basalt; dove—grey in colour, with elonga
yellow and white secondary deposits.
named “Solai Basalt”. Basalts are
This basalt underlying the Solai Phonolite has been
boreholes Cl746, C1745 and C1954.
shown below the Solai Phonolite in sections of
but many of the lava specimens examined from the
These were easily recognisable,
extreme alteration and decomposition,
vicinity of the‘ Subukia Escarpment showed
tentativ e. It is possibl e that there are more intercalations
and the identification is only
of basaltic lava than are shown.
layer within the phonolite on
A fresh black basalt was noted forming a thin
Subukia. This has not been shown on
the road pass between Milton’s Sidings and
n the fresh state of this lava and the enclosing
the map. From the contrast betwee
much later age, probably contemporary
phonolite it‘is considered to be a small dyke of
eruptio ns. This occurre nce shows that small thicknesses
with the Elmenteita fissure
in this locality, quite unrelated to the
of basalt may be found in any borehole sample
succes sion of lava flows.
generally regular stratigraphic
lite outcrops on the Bahati
A phonolite in every way similar to the Solai Phono
k; 1t underl ies yellow pumice tufi and welded
Escarpment to the north of Mbaru
trachytic tufi.
rocks in the various localities
A summary of the successions of these oldest volcanic
is given in Table I.
 .mGOmuGHNOHOHGH}
ABM—98:5
2.3mm
summm :ammm “gum 0:320 :mmsm . . :Mmmm 50m
_ 833382”
bEmEEow .
‘ -93 .mBoc
90 3:8
ozmaonxov 8:98am 3:98am
new—ocean ofiononm 2:25am 2:395 8:98am oEbEEOQ outwanrom 3:893 30m
, , 93D bah
coxuokomv...
hush ”Sash 93H 8:55 Cash. 8:55
30:86... 30:85» 30:05...
use any fish use may ask , use
qoSH oumaofih 2%:a 0.523.? Omaha—08% 033088 033089
3295 @0203 In U020? 8203 @0203 3303 @0203 Mush augmeca g2
$32.0 $2.0 $5.0 . $3.0 EoEamm EoESaomm x232 EoEQROmm
2938M floaouom 20:23” ofionohom $5334 «335 32
<52 BSM<Z E 300M o~z<oqo> 5340 E? E zoammuobmlg “and.
 03:00: 5:05: :0...
.m0080::0:w
.wmwfiugem ~3m 33:0: .mmafiugoa 033%
£033: £03: .mmmfiugouw mama—m
mwmfim 3:03:00 3:03: 303: 6:: ”$900 £03: 6::
.05: ”mambofinn £003:
0:: ӣ35823 05:: 00202003:
.59: 02:00:92: 0:05 ”$960058: 820233
00.20292: 0:05:85 “:03a £03m
£80 £30m l -88: £20 .zmmmm I EoEEoESBO
#005 8 3:0 .zmmum 30>
55%: 05682
90:3 Gimmm 03.802
005052 :0:
$30050: -020 :05 0:00:30 :0 E0200: 000385
0::0aaoz 3:20.: 5:80».—
8:09,“ 0:08 :5 £55 on: .flmboofifia 5&3 392005;:
ham—0: 2:2:t .030. 03:82 :0 50200:
8:00 .58: E 03.08%: 0:083: 0&Q .00: 03082:: 0&Q
.wEBfiwoB 00020.5 03:80:23
0:0 520 050:0» 5:028: 05—2302 -EEoE b35588
$05: 33:38:20 5:20.: 0395?: gnaw
33:00:05: 00:20”: 0.5853 9:593:0n 0:20:32 .320
MERE? 000.35 00:9: 050E302 30:06:88
0:300:00 8:093:05 05:28:20 5 -388 30:5 3—0004
330%? ”03:23
E mEBfixoB 0000:; 9505003 £2005? .8868 00:3 @9u 3:230: Gum—Hanna
0:3: S0:w-:£:0000 .0336: @3050 6:3:
:23: $320? 30: mo Sufism 50:: 0.50:8 3022 6:3:
0003038 0:20:95 :0:w-:2:00:w 00 0—005
E08» Em: 3 >20 Brianna. 08:» BEA £00m-:m_:00:m 9 0—005
A600”: m=0m 0:00.893. .3300; 0:20:93
I .mBEoSP 0:30:23 £00305“. 03:80:: 0:30:29 03:30::
labia Z :0 300 500%
0:532 925 :5: p250 908339 08050 .00: EB A020 0830
0:00:085 0200585 33:02 00:5 35—02 00:01—
0:0fiasomm zuaam EoEamm zwnmm
mk>fl0<¢rfi UChAOZOm—nm DMDM‘Z
«004nm >03<> ER 0:: no $.q Bz<oqo> MES v: ZOHmmmuqxid 339
 9

The Solai succession has been barely outlined, and a fuller understanding will
only be obtained when the area between Subukia and the Sattima (Aberdare) massif
has been surveyed.

(b) The Younger-faulted Volcanic Rocks

To the south of Menengai a series of lava—flows outcrop in faulted blocks, emerging
from thick Pleistocene deposits of the Lake Nakuru—Elmenteita Basin. These lava
flows are considered to be of a much later age than the formations described above
and are named “Nakuru Phonolitic Trachyte’“ and “Mbaruk Basalt”.
The successions observed in the main localities in which these lavas outcrop are
summarized in Table II.
The Mbaruk Basalt is the most easily identified lava in the area, being characterized
by plagioclase phenocrysts up to a centimetre in length, a most unusual feature in
Tertiary basaltic lavas of the Rift Valley. It is similar in appearance and petro-
graphy to the Kijabe Basalt described by Shand [11 and 12] and Shackleton [14],
but lacks the pronounced alignment of phenocrysts of that basalt. It is clearly a flow.
later than the main boundary faults of the Rift Valley, and correlation with the
Kijabe Basalt is improbable in spite of petrographic similarity.

On the west side of Lake Nakuru and at Soysambu an old land surface separates
the basalt from the overlying phonolitic trachyte. In the first locality, a few feet of
brick-red material and a thin layer of buff-stratified volcanic sediments separates the
two lavas; at Soysambu the old land surface is less well marked. The two lava
formations appear to be almost entirely concordant, little erosion and no tectonic
disturbance having taken place between the extrusion of the basalt and the phonolitic
trachyte.
Evidence from borehole samples (C869) suggests that the Mbaruk Basalt is under-
lain by a porphyritic phonolite. A dark lava of such composition appears to underlie
the basalt about tWo miles north of Mbaruk. Exposures are, however, poor in this
locality, and owing to the situation of the outcrop on a major fault zone, it is not
possible to say definitely whether the relationships are normal. At the Nakuru Lake
Syndicate quarry a cliff of massive black lava is being worked for road storre; this
lava apparently underlies the Mbaruk Basalt. It has the appearance of a phonolite,
but the texture is too fine for its composition to be determined without analysis.
The phonolitic trachytes are extensively developed on the west side of Lake
Nakuru and on the east side (Sirkon). In both localities there appear to be several
flows piled on top of one another. At Soysambu and to the north of Mbaruk there
is only one flow of trachytic lava overlying the Mbaruk Basalt; in many localities
its thickness does not exceed to 30 feet. Near Mbaruk Station the phonolitic trachyte
is absent altogether.
To the east of Lake Elmenteita the phonolitic trachyte has a much greater
development, and the fault scarps of the Gilgil series are composed of a single flow
locally more than 200 feet thick. The origin of these flows is not known. They do
not appear to emanate from any central volcano that can be recognized at the present
day. Their extent and uniformity over a wide area, particularly striking in the case
of the Mbaruk Basalt which has been identified over an area of about 130 square
miles suggests that eruption from fissures by a process of quiet welling out during

‘The Nakuru Phonolitic Trachyte was given this name on account of the frequent discovery
of nepheline in sections from the western half of the area. At the time of revision (March, 1957)
examination of the lavas of this group from near Gilgil showed them to be mainly quartz trachyte
(verbal communication, B. N. Temperley). Comendites have also been collected from the fault
blocks near Elmenteita. The conclusion is inescapable that there is a wide composition range,
possibly even within single flows and the name Nakuru Trachyte Group is suggested as a preferable
alternative.
 10

a period of tension is the most likely mode of origin. This is supported by the fact
that there are no pyroclastics other than insignificant wisps of welded tutf associated
with these lavas.
The relationship of these lavas to the tufls which are exposed on the Mau and
Bahati Escarpments was not easily determined. From a careful examination of the
.Bahati Escarpment between the Nakuru to Thomson’s Falls road—Mbaruk—Kariandusi
and Gilgil the following facts were discovered during the original geological survey
in 1952—53 :—
(i) On the escarpment north of Mbaruk there is a marked structural discordance
between the tutfs forming the upper part of the escarpment and the Mbaruk
Basalt which outcrops in a steep fault scarp along the foot. The tufls form
a comparatively gentle, strongly dissected slope, and outcrop in a series of
fault blocks tilted strongly towards the south. The Mbaruk Basalt forms a
continuous scarp between Mbaruk and the Nakuru—Thomson’s Falls road,
apparently unafiected by the tilting seen in the tuif above it and to the
east.
(ii) No exposures of Mbaruk Basalt or Nakuru Phonolitic Trachyte are seen
to the east on the high ground of Bahati; nor are they seen in the high
ground of the Mau. '
(iii) The Bahati Tutf and the Nakuru Phonolitic Trachyte are seen in contact a mile
west of Kariandusi diatomite quarry. They are separated by an unconformity.
The margin of the tuffs of the Bahati Escarpment makes a continuous line
running south from 'the Nakuru—Thomson’s rFalls road to Gilgil. It 'is
repeatedly offset by north—south fault dislocations.
(iv) An outcrop of phonolite, similar in appearance and texture to the Solai Phone:
lite, emerges from beneath the yellow pumice tufis and welded tuffs that
form the hill called Ngorika, three miles north of Mbaruk. This phonolite
separates the yellow pumice tuif from the Mbaruk Basalt. Further south the
two formations are in contact, and to the north phonolitic trachyte is inter-
posed between the tufis forming the upper part of the escarpment and the
Mbaruk Basalt.
The only satisfactory explanation for these very confused relationships is that
the Bahati scarp of the Rift Valley was already formed when the Mbaruk Basalt
was extruded. The tuff formations and Solai Phonolite exposed in the fault scarps were
derived from much earlier eruptions. These fault scarps were strongly eroded, and
all trace of the original boundary faults of the Rift Valley has been obscured, but
the rounded escarpments of the Man and Bahati remain, representing the original
marginal scarps of the Rift Valley.

The Mbaruk Basalt and the Nakuru Phonolitic Trachyte are envisaged as having
been erupted from fissures during a subsequent renewal of tension. They were
restricted in extent to the floor of the earlier valley which they covered in flat sheets
of lava. Subsequently large-scale faulting disrupted these flat sheets into fault blocks,
at the same time forming the secondary steep set of fault scarps at the foot of
the Bahati Escarpment, north of Mbaruk and at the foot of the Mau on the west side
of Lake Nakuru.

In 1957 Dr. B. N. Temperley re-examined exposures on and to the west of the

Gilgil Escarpment. His conclusions were in agreement with the field relations described
by the writer, and he was able to locate the unconformity between the Nakuru
trachyte and the Welded tufi and tuff formation of the Bahati Escarpment. He interprets
the slivers of later lava up to a few hundred yards thick seen throughout the length
of the Bahati Escarpment at the foot of the slope of older rocks as due to refaulting—
along the original boundary fracture of the rift graben, after erosion had cut back the
 ll

fault scarp some hundred ur so yards from the line of the fault and later trachyte
had flowed over it and up against the eroded scarp. The actual unconformable contact
can be seen in the fault block one and a half miles west of Kariandusi diatomite quarry
(see Fig. 1 drawn by Dr. Temperley) and possible on the Gilgil Escarpment, where
it is nearly vertical and might be a fault contact, though there is no evidence in the
topography of any displacement.

ELMENTEITA

0 Va
Mile a“— '4 ‘X l Hill
Szallx— I:50000 (Approx)

se'cnou o- D'

LEGEND
F
SI3 Talus mm "can! sure / Foam «finite

F/
m Gelatin Lit! Mde ’4 I Falls lulu"!
A . , . .
Kantian Lake id! A/ A-A 9-3 C-C I 0-!) sullen lines

Helm 9mm“: Trachyle

Tahs (rem surly scarp

'II‘CI Trashy“ To"

Fig. 1.—Map and serial sections of a small fault block 1} miles west of Kariandusi.
(After B. N. Temperley.)
 13

Norm of 3111111111»; the contrast b 11.6611 111: 1.1.111 1'1» 111c11g the ~11vcr of 3111111111
Bus-11h at the foot of the 81111311 itarp 111111 the 1:123:19 .iopc 01 1116 m1 cc of 1116 01d:
1111f\ Lind 11.11.1311 111115 behind 13 A211 52cm. D1, Tanipcrley 11;; 311 0 1: sted the 16 11
"Faced 1.1111: “mm." for Itf‘. 1.11111 gaps 211011111; :1 S11V€f of 1.1: I‘ 411.1 facing the
01dc't‘ 011111n 141.111 line stump. There is no 1101111 1 that 11111111: to recognize 2171.
1);): 01‘ 18.111l. \khich must bc commtm 1111011151103: 111:: R11: \"11‘11C§.. 111.13' 11.1w 1 {“11
111 the 171151 :0 considcmble (301111131011 1n cor: 1.11103 of the volcanic rocks.
P1111 1.»The tree mouldxin01615111111111.1116 flow 01‘ \1e11cr1g111

To [111: $11111: 1011711111 111.1 81.115111: .11‘8 cxpt‘xct‘l mo" 11 “ids .xi'C‘LL.
puxxint UNUCF me 11' ‘1c recent \0'1.‘.1111.\ 01 .1 ‘ 1~ 1111 Added {ruchytic {.111
merlying mew 1.1\':1< md 11 >cc111~ .1NQ1_\ 111111 11:11 - ' ‘ .1..1 «11111511411 the amic
 l9

s The products of these eruptions may be divided into:—

.(a) Pumice and Ash Deposits: mantling the outer slopes, the Rongai and Bahati
Plains, and banked up against the Mau Escarpment, together with loosely
consolidated boulder tufls associated with pumice at Lanet, Kariandusi and
south of Nakuru.
(b)Trac/1yte Lavas of the Caldera 'Floor: erupted from secondary centres within
' the caldera, and from a few very limited fissure zones on the outer slopes;
small cinder cones within the caldera.
(c) Active Steam Vents: within the caldera, with small associated conelets of red
siliceous tufa. Very restricted yellow incrustations of sulphur ('2) around the
actual steam vents.
Pumice and Ash Deposits
The outer slopes of the volcano are mantled by extensive deposits of unconsoli-
dated pumice. The pumice has a very crude stratification, and individual layers are
composed of fragments of remarkably uniform size. The fragments rarely exceed an
inch in diameter. These pumice deposits haVe a much greater development on the
western side of the crater, where, as noted by Gregory, they have mantled the under-
lying formations and the finer ashy material has banked up against the Mau scarp
on the west side of the Rift Valley. This may indicate an easterly prevailing wind at
the time of the emission of the showers of finely divided material that gave rise to
these pyroclasts. They form a smooth and level surface mantling subsidence flexures
in the underlying lavas of the crater wall, and form deep drifts in the hollow
formed by down warping in the course of the subsidence. They are probably con-
temporary with the caldera formation.
These deposits are of earlier date than many of the secondary flows of the
caldera floor, since a flow on the outer slopes north of the Eldama Ravine road,
one of the earlier caldera floor eruptions, covered this mantle of pumice. In a small
subsidence crater at the caldera rim, this flow is clearly seen resting on a surface of
unconsolidated pumice overlying the uppermost glass trachyte lava in the wall. At the
edge of this flow, blocks of lava have spilled olf, and a loose layer of volcanic
boulders rests on the pumice. It is supposed that the eruption which initiated‘the
caldera ejected vast quantities of cinders and ash which covered an enormous area.
The finest material is probably represented by red ash with pumice that is banked
up against the Mau Escarpment and against the slopes of Kilombe and Eburru.
Some coarse fragmental material was also ejected by these explosions since sections
in the railway cutting near Lanet show pumice interspersed with tufis containing
rounded blocks ,of glassy trachyte. Some of this pumice near Mereroni has been
partially cemented, probably due to the cementing actior1_,.v.of water- of the earlier
extensiVe Nakuru Lake. The showers of pumice and ash were succeeded by a renewal
of effusive activity within the caldera, giving rise to the extensive trachyte flOWS that
now cover the caldera floor.
Trachyte Lava Flaws of the Caldera Floor
The floor of the caldera is completely covered by recent lava flows erupted
from secondary centres during periods of renewed activity There is a antral eminence
marking the main centre from which these flows streamed down’until‘ they ‘Were
arrested by the caldera wall. There were, however, other eruptive centres since some
of the flews haVe no connexion with the main body of lava that radiates from this
centre. For example, the large flow in the “clover leaf” bay at the'north-west corner
of the caldera is bounded by a terminal cliff facing the centre of the crater. From the
form of the pressure ridges a centre near to the caldera wall is deduced.
The flows in the" caldera are strikingly diflerent from those forming the walls,
although the’difference in composition is. not great. Most are characteristiCally b'locky
flows with‘surface and terminal cliffs composed of'a jumble of 'looselblo’cks. Other
 20

flows are glassy in part, composed of obsidian, and show a crude, ropy texture.
BOuldery flows with cliff—like, block-strewn edges rising 100 feet above the surface
over which they have moved, are widespread. The eruption of lava in a partially
solidified state with a very high gaseous content is the accepted mode of origin of
such lava flows.
The relative age of the flows within the caldera can be judged from the degree
of vegetation cover. The earlier flows are covered by a wellLestablished grass and bush
vegetation, from which the pressure ridges stand out as bare crests. The more recent
flows have little or no grass cover, the vegetation being restricted to small bushes
established in crevices. One flow is completely devoid of cover by vegetation and,
from the fresh state of‘ the surface, it seems that the last eruptions of. lava took
place within the last century. The fissuring of the lava surfaces is so pronounced
that walking across them is laborious in the extreme. The writer has taken an hour
to traverse a quarter of a mile. The whole surface is thrown up into a remarkable
pattern of pressure ridges similar to those found on glaciers. Seen in air photographs
these clearly reveal the direction of flow. Near the eruptive centres from which the '
flows qioriginated this flow orientation is often lost. One such centre consists of a
broken field of lava, transected by open fissures more than 100 feet deep. These
fissures are mostly oriented north—south which may indicate a relation between the
Menengai eruption and the older dislocations of the Rift Valley. Some of the last
flows to be erupted emanate in or near definite small cinder cones within the caldera,
but eruption appears mainly to have taken place from fissures and small sinks in
the surface of the flows rather than from distinct craters.
Lavas of the secondary eruptions are not entirely restricted to the caldera. In two
localities eruptions from fissure zones have spilled out trachyte flows on to the outer
slopes. A flow on the southern slope, north-west of Nakuru is spectacular, a tongue
of ropy lava with a block-strewn surface having spilled down the steep slope, terminat-
ing in a. low boulder clifi. This clitf, situated on the north side of the Nakuru-Eldama
Ravine road, has been extensively quarried, and the nature of the flow is seen in
section. The main body of the flow consists of massive black trachyte lava, traversed
by numerous shrinkage cracks. The upper surface is ropy and scoriacious, having
a glassy texture, and on this surface a jumble of loose blocks has been carried along
with the flow and spilled down the lateral and terminal cliffs. From the nature of
the vegetation cover it is deduced that this flow, and a similar flow that has escaped
on to the north slope, of equivalent age to the earlier caldera flows. This flow has
originated in a north-south zone of instability. At the crater rim the older lavas form-
ing the wall Can be seen to have subsided in a narrow zone; the subsidence has pro-
duced extensive north-trending fissures, and close inspection of the wall reveals small
feeders of recent trachyte penetrating the older lava of the wall. On the surface of
the flow there is a line of small, perfectly formed craters, oriented on a north—south
line of weakness. In one of these, a cone-like body of recent lava can be seen, penetrat-
ing the older streaky trachyte exposed in the wall. The walls of these craters are
vertical and they are attributed to abrupt subsidences, forming orifices through
which lava has risen to feed the overlying flow.

Some cinder conelets occur within the caldera surrounded by the recent flows of
lavas. They‘ are in the form of conical heaps of scoriae, up to a few hundred yards in
diameter. The grass-covered mound that forms the secondary summit (6,858 feet)
has no central depression but the crude cone to the east, from which a large lava
flow emanates, has a well-marked central hollow

Active Steam Vents

Within the crater, steam vents are still active in restricted localities. The steam
issues from small fumaroles, about two inches in diameter, emerging only under
very slight pressure. Mr Thornhill, who has farmed on the edge of the crater for
 2L-

30 years, has observed a definite variation in activity according to the hour of the
day. A similar hourly variation is seen in the activity of numerous fissures blowing
cold air in the neighbourhood. It seems probable that changes in atmospheric pressure
are responsible for this regularity of behaviour. '
These steam vents have built up mounds of a light-red siliceous tufa. In one
case the tufa mound has the form of a conelet. The largest of these tufa mounds
forms an elongated ridge several hundred yards long, with numerous fumaroles situated
in the tufa along the crest of the ridge, but there is no central depression. The
actual vents are often encrusted with a yellow deposit, probably sulphur, but there
is no evidence of sulphur deposits of any magnitude. Boreholes two miles north of
the caldera have encountered steam under low pressure at 200 feet.

CHAPTER VI—THE EBURRU AND; ELMENTEITA VOLCANIC SERIES

The Eburru Volcanic Series
The great twin-peaked volcano of Eburru. is outside the area covered in this
survey. Lava flows from the volcano have, however, been. mapped on the extreme
southern margin of the area. These can be divided into two groups:—
(i) Older porphyritic trachyte lavas.
(ii) Newer glassy trachyte and trachy—obsidian flows.

The Eburru volcano is later than the grid faulting, but the first group is affected
by substantial posthumous fault dislocatiOns, whilst the second group is only afiected.
by minor fissuring. The first group of lavas show a striking similarity to the earliest
porphyritic trachyte lavas of Menengai. ‘

The Elmenteita Volcanic Series

The cluster of comparatively small volcanoes south of Lake Elmenteita have
been referred to by Gregory as a group of broken basalt craters rising from the
Elmenteita alluvium. He also mentions the connexion of the Eburru, Station steam
vents with north—south faulting. The whole complex of volcanic cones and basaltic
flows is illustrated in Map No. 4. It is evident that there are several series of volcanoes
in this locality. The Eburru peaks are two comparatively old and dissected volcanoes.
The ridge formed by these peaks traverses the Rift Valley in an east—west direction,
whilst the more recent Elmenteita volcanoes are restricted to a north—south zone
about five miles wide cutting across the eastern shoulder, of the Eburru ridge. This
line of volcanoescontinues southwards towards Lake Naivasha, but only the area
north of the old railway alignment is described here.
The volcanic eruptions may be divided as follows:— ’
(i) Small Tuff Volcanoes, broken by north—south faults of restricted throw.
Basaltic Tufl's t(B). Buff and grey in colour, crudely stratified, often with
included angular blocks or boulders of phonolitic trachyte, porphyritic basalt
(Mbaruk type) and vesicular bas‘alt of similar type to that described below.
Basalt Flows affected by north and south faulting of restricted throw
(and included boulders in t-nfls). Vesicular basalt—(Bel) rough and compara-
tively coarse textured in hand specimen, numerous large individual crystals
set in a glass ground mass, prominent rounded vesicles.
(ii) Basaltic cones with distinct and well-preserved craters composed dominantly
' of red scoriacious 'cindery basalts, cindery flows sometimes exposed in the
crater walls, outer slopes mantled, in loose Cinders derived from the waning
stages of eruption. '
 22

Basalt Flows, rough textured, black-in colour, with numerous round

vesicles. Several successive flows superimposed on one another, Be", Be“, BeC.
Small broken volcanoes, probably of similar date to the tuff cones
composed of whitened basaltic pumice.

(iii)A recent-basalt cone, a mile north of Eburru Station, similar to the other
cinder cones but the almost perfectly preserved form suggests that it is the
-most recent volcano of any size in the vicinity. This is borne out by the
fact that it is situated on the edge of a zone of thermal activity.
Very recent flows of grey or black basalt with poorly established vegeta-
tion eover'; glassy in texture and characterized by elongated vesicles,
porphyritic in part. '
Flows of glassy black basalt with prominent ideomorphic felspar pheno-
Icrysts, in dike-like bodies, filling irregularly orientated fissures in the crater
of the volcano named “High”.

The Tufi‘ Cones and Associated Basalt Lava

Tuff cones make up the greater part of the earliest phase of renewed vulcanism
near Lake Elmenteita. The associated lavas have mostly been obscured by subsequent
eruptions and a cover of lake beds, but no doubt they covered a considerable area.
Basalt flows which outcrop in the vicinity of Lord Delamere’s house at Soysambu
probably represent these lava eruptions, since. unlike the remainder of the Elmenteita
basalts, they are transected by minor north-south faults. The presence of basalt
of similar type in the boulder beds intercalated in the tufis forming the cones, shows
that the earliest effusions of a vesicular basalt were at least simultaneous with, if not
earlier in time than, the formation of the cones.
The volcanoes of this group are all very steep sided and formed of crudely
stratified tufis, dipping> steeply outwards at inclinations of 30 to 40 degrees. The
tufls are horizontally disposed on the crest of the crater rim and dip inwards at an
equivalent angle of inclination on the inner wall of the crater.
Boulder'tufi’s Containing fragments of phonolitic trachyte and Mbaruk Basalt
are often seen in these cones indicating that the formation of the craters was accom-
panied by the explosive eruption of fragmental material. Some cones, however, are
entirely composed of finer tufi. The majority of the cones in this group are transected
by north-south faults which downthrow the median sector of the cone, leaving the
lateral sectors standing above a median trough. The occurrence of these sharply down-
thr‘ownfault troughs is indicative of the subsidence subsequent to volcanic eruption.
The craters'represent vents situated on north—south fissures, and the explanation
may be that these vents enlarged longitudinally as the explosive eruptions continued
along the line of the fissure. The structure is comparable in origin to subsidence
calderas. It may well be that these structures have a similar origin to some of the
major fault troughs of the Rift Valley, the process being here seen in microcosm.
This has been'suggested'by Shand [11] in a brief note on the Rift Valley of Kenya.
The absence of lithic material in these cones, except for angular blocks of the under-
lying‘ country,_rock suggests that these cones may well represent explosive phreatic
eruptions caused by water from the lakes seeping down reopened, deep-seated fissures
and coming into Contact with very hot lava. It is noticeable that these small explosion
vents are clustered in lake basins and are not found elsewhere.
’The cone atu'HoneymOOn Hill, south of NakuruTown, and the small fragment
called Crescent Hill, immediately to the east, are very similar to the EImenteita cones
and must be considered as part of the same closely related eruptive series. The median
fault trough iS'not so-‘distinct on Honeymoon Hill, but the cone is transected by small
northesouth ffaults of small throw. The'blocks included in the tuffs are composed
of vesicular basalt and porphyritic trachyte. , '
 23,,
These cones show lake-shore lines identified as Gamblian midway up their
slopes, and are believed to be of Kanjeran age.

The problem of the time relation between these cones and Menengai is not
easily solved. However, the Eburru volcano is definitely older than the Elmenteita
craters, and since porphyritic trachytes, identical with those of the first Menengai
eruption, outcrop on the slopes ‘bf Eburru. it is concluded that the first Menengai
eruptions may poss1bly have preceded the first Elmenteita basalt eruptions. '

The Basalt Cinder Cones and Associated Vesiculal Lava Flows I V V r .

These have a much smoother and more rounded profile than the hilt cones.
No faults disrupt theirslopes, which show outcrops of -red basalt cind: rs instead
of dip slopes of tufi. One of the most westerly and the largest of these cones appears
to have a double crater; a later and perfectly formed crater lying within an earlier,
larger and more elognated orifice. It is possible, however, that this nested appearance
is due to subsidence of the median sector as in the case of the earlier tuff cones.
No fault planes can, however, be made out and the walls of the outer hollow have
a distinctly curved trace. Several lava flows are exposed in the wall of the central
crater. These are of grey scoracious basalt with prominent, clear, felspar insets.

Microscope slides indicate a considerable degree of cataclasis. Two of these

cinder cones are isolated within a very recent lava flow. They can be seen to be-
cones of similar type to the one referred to above, but owing to ditficultyrof access
were not examined in detail. Smaller cones near the shore of Lake Elmenteita show
no distinct crater, consisting simply of conical piles of red cindery basalt.

The connexion between the unfaulted vesicular basalt flows. and individual
volcanoes is obscure, but it is clear from the map that there is an areal relation-
ship. Thus it seems likely that the earliest group of unfaulted flows emanated from
these craters A vast tract on the south and west sides of Lake Elmenteita has been
overridden by these flows. They cover the earlier deposits of the Nakuru—Elmenteita
basin but have to some extent been subsequently covered by a thin and discontinuous
mantle of lake deposits at a recent date when the lake waters covered a larger, area.

The Basalt Scoria Volcano near Eburru Station and A tsociated Vesicular Flaws
This 15 a perfect cone, undisturbed by faulting, having cinder-covered slopes much
steeper and more regular than any of the other basalt cinder cones. It has a deep
crater some 200 yards in diameter. Red cindery basalt is exposed on the crater wall.
There is a smaller parasitic vent immediately to the north of the volcano suggesting
an even more recent eruption on the fissure line after the main vent had been plugged.

The most recently erupted basalt lavas differ from the earlier flows in the
presence of very characteristic elongated vesicles and a glassy texture. In colour they
are generally lighter than the earlier erupted vesicular baSalt of recent age. From the.
slope of the surface of this latest flow, the source would appear to be the Eburru
Station Volcano. Another separate flow of this lava flows down to the south from a
small cone of red cindery basalt. These flows are covered only by a coarse Euphorbia
scrub and, from the degree to which vegetation hasestablished itself, would appear
to be of similar age to the earlier secondary eruptions of the'Menengai centre.

The Eburru Steam Zone .

This lies within the Elmenteita zone of posthumous faulting and Vulcanism.
Numerous active steam vents are seen immediately north of the railway. No evidence
of steamvactivity has been observed in the craters or in the surface of the lava flows
further north.‘The steam vents occur along well-defined north—south surface fissures.
 24

The Origin-of the Elmenteita Volcanic Zone

From a study of the fault structures near Gilgil and Soysambu it can be deduced
that Elmenteita lies in a graben formed in the Lower Pleistocene or Lower Middle
Pleistocene epoch. Similar fault troughs are those of Solai and Lake Nakuru. In these
troughs fissuring is widespread and there is strong evidence of present-day crustal
instability. In the Nakuru fault trough only Honeymoon Hill provides evidence of
a renewal of volcanic eruption, though the Decent- eruptions and steam zone of
Menengai are probably related to the northward continuation of the same trough.
The Solai fault troughs show evidence of very recent crustal instability but no
recurrence of vulcanism. The Elmenteita trough apears to be one of the least stable
in the Rift, and has been the focus of repeated recurrences of volcanic eruption.
Present-day steam conditions testify to the continued activity. This line of instability
and thermal conditions probably continues southwards over the shoulder of Eburru,
and through Lake Naivasha to the vicinity of Hell’s Gate (Njorowa Gorge).

CHAPTER VII—THE SEDIMENTS

The sediments may be subdivided as follows .
(i) Reworked tuffs and clays, intercalated in the older tuff formations of the
Mau and Bahati Escarpments.
Yellow to bufi sediments; some fragments showing rounding by water
abrasion; clayey matrix.
(ii) Sediments intercalated in the older-faulted lava flows of the Rift Valley floor.
Similar to the above, but water rounding more marked.
(iii) Lake sediments of the Pleistocene Lake of the Nakuru—Elmenteita Basin.
Yellow-buff volcanic grits, reworked tut’fs, clay, diatomite.
The first two groups, intercalated in strongly faulted volcanic rocks bear little
relation to the present-day topographic features or drainage basins. They are, however,
important in that they they provide the aquifers within the faulted rocks. The third
group of sediments were deposited in a drainage basin closely resembling that of
Lakes Nakuru and Elmenteita at the present time, but modified by minor faulting
and warping movements.

Kanjeran Sediments ,
The conclusion has been reached that there were two major faulting episodes
in the Rift Valley. The second of these disrupted the extensive basalt and phonolitic
trachyte lava flows covering the Valley floor, forming a series of horst and graben
structures. It is upon this serried surface that the earliest Pleistocene lake sediments
were deposited. These sediments, the Kanjeran Lake Beds, are exposed at Kariandusi,
on ”Cole’s Farm further north, and at Soysambu (according to Solomon, Leakey and
Nilsson). Solomon identified tufis on Cole’s Farm as Lower Kamasian;* a point of
great significance, since these reworked tuffs appear to be continuous with the coarse,
crudely stratified basaltic tufts that built High volcano and the other yellow tul’f
cones of Elmenteita. (The distinction .between tuffs and lake beds in this locality is quite
arbitrary.) The tuff‘cones of Elmenteita are thus thought to be of Kanjeran age.
The Kanjeran Sediments of Kariandusi, and the tutfs of these cones are affected
by minor faults; posthumous dislocations on the earlier lines of weakness. The fact
that Kanjerandiatomitic lake beds are displaced by a fault of more than 100 feet
throw at Kariandusi has been taken by Solomon and Leakey to indicate that the
Gilgil fault scarps are among the latest in the Rift Valley. However, a study of High
volcano on Cole’s Farm south of Kariandusi reveals a major fault’scarp of trachytic
*At the time of the revision of this report in 1957, before publication, the sediments here
referred‘to as Kamasian had been ascribed to the Kanjeran stage by Leakey . . ., i.e. Upper Middle
Pleistocene, slightly later than the Kamasian. They are referred to as Kanjeran in‘the subsequent
text.
 25

lava over 200 feet high, obscured by the tufts of this volcano whichhaVe banked
up over the fault scarp. These tufis have been faulted, owing to a major posthumous
movement along the same lines, but they are not displaced more than 30 feet. The
downthrow of 100 feet at Kariandusi is extreme for one of these minor posthumous
faults, but it must be noted that the Elmenteita Zone has been the site of repeated
vulcanism over a considerable period during the Pleistocene and it is logical to
expect late subsidence movements to be at their greatest development in this zone.
Shackleton (1955) in a paper published after the production of the original limited
edition of this report states that the throw of “fifty metres” on the fault displacing
the diatomitic sediments at Kariandusi is much less than the displacements atfecting
the trachyte flows which form the lines of lava ridges, and he concludes (like the
writer) that the diatomite was deposited on a surface of faulted lava, and was later
subjected to minor faulting.
FAULT soup ABOUT 30 F7? may
AFcr/NG rurrs or 'man’
VOLCANO (emu/u or mun/m;
0N rm: EARL/£2 LINE OF WEAKNESS)

fAll/ER FAULT SCARP 3£60M55

03.960p BY TUFFJ‘ 0F 'NIGH'
Voted/‘10 PILED 0V5! TOP DFIT.
2x?

/ — may]: LAVA:
~;:~Ti-
ozmz FAULT SCARP zooiaoo' may 39 V
AFFECTING TRACIIYTE LAVA.
Fig. 5.—Sketch showing the renewed faulting of basaltic tufi's of “High” volcano—a
posthumous adjustment on an older major fault zone.

The Gamblian and Makalian lake beds have been differentiated by Nilsson into
seven distinct‘lakes,-from correlation of shoreline terraces. The first Gamblian lake
extended over the shoulder of the Gilgil Escarpment, the Naivasha and Nak‘uru basins
being joined in one great‘lake. *’ a ' '
This ‘Gamblian lake left a shoreline terrace on the Elmenteita tutf cones and on
the slopes ,of Menengai. The early phase of eruptionfrom Menengai must certainly
be pre-Gamblian and possibly pre-Kamasian. Mr. Lepersonne of the Belgian Congo
recently‘visited Menengai and considered that the state of erosion of the ‘inner cliffs
indicated an alihost historic age. Such a dating is, hOWever, in conflict with all the
other available evidence. ‘ '
 26
Subsequently a succession of lakes were formed in the wetter periods, the trend
‘
being towards less and less extensive lakes until the very restricted lakes of the-present
dayzwere left. (See Figs. 6 and'7.) ' ‘ ‘
36f: ‘ ' ' to” \ ~20‘~

5nAKunu
, me

so’
d u.)-
. (mm

40 LMEuTUTA

MIX. DEVELOPMENT
U? LIKE e. 75,000 LC.

sol . 1“/,
‘9
[if c
4?
IO

tan;
VNAIVASH

. \\>\\ x " T/=- § -.

~ , “\m vex \\ t \t -
’ ”h

Hull 0 l ‘2 5 A 5 I0 ljllis

Fig. 6.—Sequence of changes of the Rift Valley lakes. (After E. Nilsson.)' '
 27

5_7'A_§5 El. MEN 75/54 NA not SIM APPROX. DA 75

75, 000 5. C.
/

4 m, 000 5. C

6 Heist/W DA Y

Fig. 7.—Sequence of changes of the Rift Valley lakes. (After E. Nilsson.)

The Makalian lake beds are overlain by a deposit of consolidated grey tut’f known
locally as building stone. This tuft extends as far as Njoro, Elburgon and Rongai,
and is concentrated in the present—day valleys having apparently been washed in by
rain after the existing drainage pattern has been initiated. This tut’t closely resembles
the glassy tut‘fs that form scattered outcrops on the east side of Menengai. They are
the uppermost consolidated volcanic rocks, on the slopes of Menengai, and have
been taken to represent the products of the explosive eruptions "coincident with the
first stages in the formation of the caldera. They are never seen within the caldera.
lf this assumption is correct the formation of the caldera occurred after the Makalian
subdivision of the Pleistocene not more than l0,000 years ago, and the process
was completed considerably later, though how much later cannot be estimated.

The Makalian ash deposits are affected by minor faults posthumous on the earlier
lines of weakness. These may well be part of the same system of subsidences which
formed the caldera.
The glassy tufts east of Menengai (correlated with the Makalian ash) are overlain
by the extensive unconsolidated pumice deposits. These form a mantle over the
western slopes of Menengai, and are banked up against the Mau, Kilombe Hill and
Eburru. The recent shoreline terrace 145 feet above Lake Nakuru dated as Nakuran
is composed of coarse pumice clearly derived from these pyroclastic deposits.
The deposits of the present-day lakes are mainly composed of fine clay, silt
and volcanic sand.
 28 __‘ ‘
CHAPTER VIII—TECTONICS
It must be accepted that the Rift Valley is essentially a tectonic structure. The
imprint of such a regular pattern of linear troughs by erosional processes on the
centre of theAfrican continent is inconceivable. In the Nakuru area the following
tectonic episodes can be recognized:—

(i) Major faulting, originating the Rift Valley. Fault zones of enormous dis-
placement along the line of the Man Escarpment and the Bahati—Subukia
Escarpment. The actual faults are mostly obscured by later eruptives and
the faults are now represented by fault line scarps. (Miocene?)

(ii) Major faulting further breaking up the scarps formed during the original
faulting episode; disrupting the lavas covering the valley floor into a compli-
cated series of horsts and grabens. Renewed movement on older bounding
faults. (Pliocene or earliest Pleistocene?)

(iii) Minor renewals of faulting along the earlier lines of weakness. (Pleistocene?)

(iv) Fissuring and very minor adjustments accompanied by earth tremors continuing
at the present day. (Recent)

The evidence for the two major faulting episodes has largely been dealt with in
the discussion of the relationships of the various volcanic formations. The earlier
faults are not exposed, and the majority of the major faults seen in the area are
attributed to the second episode of faulting. There is a slight variation in trend of the
faults of the second series. The Mau Escarpment shows a west-north-west—east-south—
east deviation between M010 and Njoro—while the Bahati-Subukia Escarpment trends
north-west to south-east. This latter escarpment is cut up into a series ’of north—south
troughs by later faulting. This is particularly well seen between Mbaruk, Kariandusi
and Gilgil where the older fault scarp is repeatedly offset by north—south faults, bring-
ingrthe lavas of the Rift Valley floor into contact with the older tuffs of the escarpment.

Posthumous faults affect the glassy tuffs of Menengai near the Crater Stream and
west of Mbaruk. The Makalian ash east of Njoro is also similarly disturbed. The
Kanjeran lake beds and the tulf cones of Elmenteita are displaced by faults of similar
magnitude. Some of these later fault movements apparently coincided with the large-
scale subsidence of Menengai within the last 10,000 years.

' ‘The concentric pattern of faults and monoclines to the north of Menengai is
believed to have been imprinted on the older north—south fault pattern concurrently
with the caldera formation. The trough faults that cut across the median line of
the Elmenteita tuff cones (Kanjeran?) are thought to be earlier structures due to
subsidence immediately after the explosive eruptions ceased in these volcanoes.

Flssures and very minor faults cut recent pyroclastics exposed in a railway cutting
at Mereroni (Lanet). Further evidence of recent movement is seen in the form of
surface fissures most. frequently found in the centre of - fault. troughs. 0n Milton’s
’farm at Solai these are most strongly developed. On this farm there are some small
vertical fault planes exposed, the soft SOlai tuff showing no degradation by erosion.
A zone of recent fissuring croSses the Njoro River east of Glanjoro Farm. Earth-
quakes are not unknown in the Naku’ru area and tremors are frequent in the Solai
Valley. Though earth movements are' at present of a very small order, it is evident
that Nakuru is situated in a zone of crustal'instability and there‘is a possibility of
the recurrence of catastrophic earthquakes.
 mux< :mzvzz an... E
440.0040uw Lo 2.0.2.54. m}... Ouknmwwan, .
0:5?“
.okruinuaau s.
(qua: o: u a)

Lilli
BNIL'InVi
1006”?

.....o<zno<..m ”“40“ ._ 32:23.. 92 .213...»

E... , KNEW . .5220 55.. .0 (>3
92 2...... 2. cut...
:3...
HIVN

3:1... 30.330 1.3.3.5: 0.53.2.0:

u_r.r.xu<..... 0:93;.»
. Awurflozoxn .. V .m
. 100.... . 953.... no m
nu zrflfimdunpwoonWwafi n<r<rhfi an» .M:.W<raznz.w 5:. no .25.. a:uw.wwroa._.u.: :
>34; N
#323 3:53; .3293 :36 ..o (>3 0.2.4:: 1.005....
, 22! 0100.: 35,133... 33:05.... l<u.n.:.: 0.53:.a 25:25.13; n
_ 01.1.9.0”. 3.33:1;
fizauOPO.dJA
[dz/OJ
“
_ .uaooodou H
<turzasfi ..o 4 . 9...: 3
92.53: . 3243.3. l5
uaoxnxrao. 250.... 5.3: :35: w Dd1103ud. our:
9...... zoniruzoa a
102.: M |Ior<6 2530 9.55. 2.00060. L
3:3 1...... 5553:... n on 0.3.... o
n. 3.505.923... 1439249 H
I 6...... nnxuoz...‘
29

_ £95 .32.. ounce? N

_ a. .513
" uaoood. 1
.10 91:04... .2910: ,
$1.53.. 0.02. ..o :n< 23.: . .
3023150.. n33.» 13.5.0 20 26.5 ntnonun 32.. 2:: 1.1.3.2:
101.: 92.0....3u23332 .o :2,» 2:0...3
:2. u...rxu<ur .3239 nxa.‘ 12341429 uaovoé , ,
9.313....“ :3:
-
H1... .5303... u
.2013. 2 n>on< Ea... £31.... .2532: 1
..o 3:... 1nn.=oa.x,< nurzuxuu 31.3.. .5 V . a
03.35.... 233.3: 35...»... 3.330..» noon» 2.
_ , . . , n
. L
<turzn<3u h 3.3.5.
032.. ”>355... .
. nouzOu .3026 534$ «00.... <1u0..<u...o ado: 5...:950 x. 3.232..
«5433 ¢.._nrz..:..m our>1u<¢r 2252.3. 02.50.15 a3 5.5.... .50:
, 5...) .5 332.0...» 32.. I
—
ganja; 2 .43.... 5...; I... .3 «12.. . _ 2.. 293:»: “Hrlflflflndd 43H
L
. , , . HG
(
kl . 50 ‘
<
OXUOK 0.1(UJ07
mfl Tm<H
 30
The Nature of Fault Movements

The apparent relation of the Rift Valley system to previous tectonic structures.
has been discussed by Shackleton [3] and McConnell, [4]. Considering the Tertiary
structures, for which many theories have been advanced, the only type of dislocation
s
exposed on the surface in the Rift Valley are steep normal faults. The actual
fault
planes are not very frequently exposed, but where théy are there can be no doubt
as
to their nature. The fractures are open and usually filled with rubble. The
whole
pattern of the'faulting suggests simple gravity collapse. It is suggested
that the
most recent minor faults may be caused by subsidence of the crust
subsequent to
extrusion of great volumes of lava and tuff.

The tilting of the fault blocks downwards away from the-scarps

is a marked
feature of the grid structure of the Rift Valley floor. It is identical
with the structure
described by Le Conte [10] in the Basin Range of North America,
and; suggests
that magma rose near to the surface, the fault blocks, already
opened up by tension,
adjusting themselves to a position of equilibrium while floating
in a liquid magma.
The grid faults must penetrate to great depths if this condition
existed.
The oldest faults, boundary faults to the original fault trough, are
obscured. The
nature of these earliest faults is not evident. from a study of this
area. ’
The trapezoidal shape of the Menengai caldera clearly reflects
a control exerted
by the old fault line scarp running north-westwards towards
Ol Punyata, and the
later (Lower Pleistocene (7)) fault trough which run north-no
rth-eastwards towards
Laikipia from Menengai. The interesection of these two fault
lines may indeed be
causally related to the presence of the caldera at this point. The
pattern of the grid
faulting and later faulting north of Menengai shows the dominan
t north—south trend
and also an oblique north-westerly trend which can be interpreted as
posthumous move-
ment following the line of the oldest known fracture system.

Conclusion ~ -
It is the conclusion of the writer that the whole complex structure
of horst and
graben, grids, etc., is superimposed upon an older fault trough.
The cause of the
formation of the original fault trough and the nature of the faults
is obscure, but
there is no evidence of compressional structures and the probabil
ity is that they are
normal tension faults. The second (grid faulting) episode was apparentl
y caused by a
renewal ‘of tension: the several minor recurrences in more recent time are attributed
to gravity adjustments, consequent on volcanic eruption.

CHAPTER IX—HYDROLOGY
Rainfall

Typical annual rainfall figures for the Nakuru area are given below.
It is seen
that the rainfall on the forested high ground of the Man and Bahati
rises to 50
inches and more, while on the floor of the Rift Valley average readings vary
between
25 and 35 inches per annum. -

y The rainfall is mainly inkthe form of heavy thunderstorms in the wet months,
March [April and ”July/August. ‘
Fig. 8 shows the Isohyetsfor the Nakuru drainage basin.
 /
,
/
oo-ss/z; / ‘
35' — ’ ’
-
sass/Io: ‘ /
o , ,
’ 1
/

3%.“; \

“\J o.
/ . / N/ \y
{fig-337531;
/\
‘ 5
/" s/ue S

\ \\.
ease/as
seas/297
\d/
\
SO'SCé 25

>/
\

’0'35 92

3055/": 0-.35/41
, A
33

. fi

\ \. w
'O~56/l47
\o

E HA l "1' mu
\45 6165 \
09E

I 59 as~ as
3‘

mu: 4 5 v 0 5 IO mus

Fig. 8.—Isohyeis of average rainfall for the 10-year period 1943—1952.

 as as , / 3 7
41 , ’ 49
, ,/ 3
' ll 13 As ’ / ”.5 ,5,
, ’ , 1 o
.
/ /
' /
/
teas/2g / I \1
35' _ , ’ ’
3035/“?! / (
-
O
/
, ‘
y
, nae/94 SO'Sfi/W /
/ O '
\ . ‘o' 39 v

‘4‘“ , " mum u MOIO ,

‘° 1 { swam»
so 35 :25 O .
/ \ ‘4
90- lo -\
sou/loo 5.} o 1
O 5/ 30' ‘l
80-3 06 I
‘2 t

I
‘
9.-
\ ° kg. 9' sows/72
\ J
x
» 7 7
’J
I

/ \ so-ae/Izo
/“‘\/ \ 0!
//\/ \3
:0»: 20’35'
{59532.1 53' /s‘u I:
// r"\

\
\
x
\ sou/:07
o.\ ‘ ‘
/ 90.16/50 \\ V
‘ V

/ h\W ‘
.
W ‘
3:

“
.

O~5‘/l47 <
\o

£LM£u1n\u
\43 92345 \
903m.
\ . \ \/

31 33 3' 29 17 n 15

Mill! 4 3 l O 5 IO Mlllfi

Fig. 8.—Isohyeis of average rainfall for the 10-year 'period 1943—1952.

 32

TABLE [V

Period over Average

STATION which Average Rainfall
taken in inches

A.-—MAU—-
Molo (Railway Station) . . . . . . . . 46 years 50-43
Elburgon (Mariashoni) . . . . . . . . 2 ,, 48-63
Elmenteita (Mau Narok) . . . . . . . . 27 ,, 41 ~51
Njoro (Plant Breeding Station) . . . . . . 23 ,, 36-24
B.—BAHATI—
Nakuru (Bahati Forest) .. .. .. .. 20 ,, 48-11
CZ—RIFT VALLEY FLOOR—
Rongai (Gogar). . ' . , . . . . 21 ,, 34-38
Kampi-ya-Moto (Rongai Bdg') . . . . . . 38 ,, 34-95
Solai (Rhodora) . . . . . . . l4 ,, 1 37-09
Solai (01 Punyata) .. .. . . . . . . 34 ,, 40-45
Nakuru (Met. Station) .. . . . . . . 34 ,, 34-59
Elmenteita (Soysambu) .. . . . . . . 15 ,, 26-37

The Hydrological Cycle

The Rift Valley contains a system of internal drainage basins, with no outlet
to the sea. The hydrological cycle as envisaged in one of these basins is illustrated
in Fig 9.
 $215 «a 2: 5 9.93 Esme—2P3 a... *9 525585».— 323.835! a a;
2341 «353. .«u
nkivthfiu): Luflau Q! < Manx» ink ‘k N»:
#6,.3.‘ Hun-3.37380 Kfi hkkwtwkus E
ugutsnu at £53». 33$:VQ :3 it!
xtkkuuugt cg 3 99. «St... ‘33::
I "214:1...“ .
_:
.45: :38. t3 ufi «4.3q
mwunfizqw.‘ u: . kuxw no uwxwv um: meeting
wtnfiwq 38:33 =3 3333 2.313»: 3.52%
t figfl $0333 2
/
$u§$§ :4...
IQC ‘Qfik '2 V
MHIWK x:
in... tax Zn: ‘9‘ §u nu»
29: Eiuz‘wn 3.
3:13 «at: E
in wwwsiufli‘ .3: 8 «$3 .33 ‘Q «£8223
awkwiww 2 mg wk 9: kg HE SEQ 333.. .3 hunks:
-
S 1E3; 3.2 «a 53 3» .§Q\L1!~QK / ‘5 Q V“). QWL gkxu 6N0
NU‘kuSh wfifi‘flnfi 2x
hflx‘Uix MM.MN Qk‘ “Kan—Q g6»! “.t
VV‘KcQN §k 3‘3 uukik hi-‘mflkh .
N§QN ku‘ux‘ x: NWENNM. 3h Hung:
b§k€u {‘NMKM WU‘Kfihh
Sun tuwk §§§§x§w «u»
,, zutgimk‘fl 3
«3% 33‘33 wank:
wusax \E :36» zSiuuSkS E \\
.. Sh sfixukwfi quikxfiVfi
KO gttal§tk NWN‘Q V d :d d
Why‘Ux‘x Gm .QV
fl Q‘Kx‘x‘fi
“fig. 3. (“km . Hulk
3 33; .mwjmw m
 34
The main basins in the Kenya Rift Valley are the Naivasha Basin, the Nakuru—
Elmenteita Basin and Hannington—Baringo—Rudolf Basin. The watersheds are the
Eburru Ridge and the Menengai—Njoro Ridge. The possibility of subterranean outlets
between these basins must be considered. It is possible that there is an underground
flow between Naivasha and Elmenteita, but not enough is known of the area between
the two lakes to provide any evidence at the present date. From the form of the piezo-
metric contours at the northern end of the Nakuru Basin it seems unlikely that
there is any outlet under Menengai. The water in Borehole C.1246 at Nakuru Show-
ground stands 4 feet above the level of the lake floor, indicating a very gentle hydraulic
gradient southwards towards the lake.

_ Owing to the extremely variable nature of the geological formations, and the
lack of data, no more detailed account of the hydrological conditions can be attempted
at the time of writing.

CHAPTER X—HISTORY AND RESULTS OF GROUNDWATER

DEVELOPMENT
Post History
The conditions governing the occurrence of groundwater in the Colony were
described by a former Director of Public Works, H. L. Sikes, in a memoir published
in 1934 and entitled The Underground Water Resources of Kenya Colony. Knowledge
of the subject was extremely limited at the time with only some 200 boreholes drilled
in the Colony. These were restricted to the few areas in which the needfor water
was extremely pressing. Between 1934 and 1953 2,000 more boreholes have been drilled,
with a considerable extension of our knowledge of the potentialities and mode of
occurrence of groundwater over many parts of the Colony.
At the time of Sikes’s survey only 23 boreholes had been drilled in the Rift Valley.
Of these, only eight were successful, four of which were in the Kamasia Reserve.
These very poor early results were due to a concentration of boring operations in the
Kedong sector, where steam conditions and very deep water table conditions pre-
dominate. Failures were also caused by the fissures encountered during drilling
operations.
Sikes based his general conclusions regarding groundwater’potentialities of the
Rift Valley on the very meagre evidence provided by these 23 boreholes. His conclu—
sions were affected by two misconceptions. Firstly, he overestimated the thermal condi-
tions. It can now be demonstrated that the steam areas are limited to the immediate
vicinity of recent centres of vulcanicity, though high temperature conditions are wide-
spread throughout the Rift Valley. The second, and far more significant, misconception
was the idea that the groundwater of the Rift Valley is an entirely separate system
from that of the high country flanking it on either side. It is one of the facts most
clearly demonstrated in the present survey that the Rift Valley in the Nakuru area,
though an area of internal surface drainage, obtains its main recharge of groundwater
not from the rain that falls within it, but from the rain that waters its high flanks,
the Bahati Forest and the Mau. Sikes appears to have underestimated the volume of
recharge available from these forested flanks, and he considered that any water reach-
ing the valley floor would at once sink to great depths to be later returned to the
atmosphere as steam from invisible fumaroles.
It was not until. after 1940 that drilling on a large scale commenced in the Colony,
and then the potentialities of the Rift Valley were revealed. B'oreholes drilled by J. A.
Macdonald and H. H. Kjaertinge proved the existence of groundwater in quantity
and at reasonable depth in the Nakuru basin and on the Rongai plain. Drilling
at Elmenteita proved the existence of large supplies in the thick sedimentary deposits
of the lake basin. Interest in boring for water was aroused, and widespread boring
operations in the succeeding years have revealed that favourable aquifers in the form
 35
flows are
of sedimentary intercalations and old land surfaces separating the lava
cause of boring
ubiquitous throughout the vo‘léanic rocks in the area. The main
tic_(or
failures in volcanic or sedimentary rocks is a condition where the hydrosta
depth limits,
piezometric) level is too deep for Water to be obtained within practical
in no way due to the absence of favourabl e geologica l formation s.
and
the area.
Drilling operations in the Solai Valley lagged behind the remainder of
farm of A. J. Simpson near 01 Punyata produced
An early attempt in 1930 on the
Keppe drilled
steam under low pressure, and though ten years later Manuel and
River near Solai Police Post, this was apparently
successfully close to the Tindaress
was
regarded as a rather fortuitous tapping of a river seepage. The Solai Valley
the proximity of
considered an area of little potentiality for many years because of
curves completely
Menengai and the fact that resistivity depth probes yielded low-value
in the valley
devoid of any indications. Up to 1952 only-five boreholes had been drilled
Successful
(excluding the vicinity of Lake Solai) of which only one was sucessful.
then obtained on subsidized sites at Tindaress and on the farms of
results were
was carried out by an enterprisi ng drilling
G. B. Ireland and B, McKenzie. Drilling
r on farms where subsidy had not been granted for C. M. Brown, H. H.
contracto
exceeded
Feet and Mrs. Hawkins and the tested yields of each of these seven boreholes
being an
20,000 gallons per day, demonstrating that the Solai Valley, far frOm
ntly
unfavourable area, is, in fact, an exceptionally good area for drilling. Subseque
E. and R. H.
boreholes drilled on farms of W. B. Prentice, J. Seys, J. Eames, K.
revealed that on the eastern and south-east side of
Adams and P. E. L. Howard have
the caldera.
Menengai water-tables at reasonable depth exist close to the rim of
block of land
The borehole on the farm of J. A. Seys is suitated on a collapsed
the 300—
virtually within the caldera while Prentice’s borehole is but 1,500 yards from
feet cliff bounding the crater.

Drilling in the Kampi-ya—Moto area has always been attended by total failure,
and
though success has been achieved further north in the Kamasia Reserve (C711)
at Lomolo. The reasons for this failure are considered in another section of 'this report
that
No great hope of success in this area can be entertained. It is just possible
water might be obtained in some parts by drilling to nearly 1,000 feet, but the hydro-
static level is so deep that pumping would be both difficult and expensive.
In Elmenteita drilling has been virtually 100 per cent successful. The thick series
of Pleistocene lake beds provide excellent aquifers. As would be expected, however,
many of the boreholes near the lake yield water of a highly mineralized character.

CHAPTER Xl—DESCRIPTION 0F AQUIFERS

In the area under consideration the geological formations are limited to the
Kainozoic Volcanic rocks (more correctly defined as Tertiary-Quatenary, since erup-
tions of lava have continued up to the last few hundred years) and sedimentary
deposits, mainly'of 'lacustrine origin, extending in time of deposition from the Tertiary
era to the present day.
The volcanic formations consist of basalt, phonolite, phonolitic trachyte and
trachyte flows, together with intercalated tufis and reworked tufts containing abundant
water-rounded fragments. There is a further form of volcanic rock, widespread through-
out the area, which may best be described as welded tuff. These show a grey or green
glassy matrix choked with pyroclastic fragments and are trachytic in composition.
They differ from loosely consolidated tufis in being impervious and their effect on
groundwater conditions is similar to that of trachyte lava flows.
The sediments, which can invariably [be detected in borehole samples, from the
presence of water-rounded fragments, grade from the reworked tufis mentioned above
to lacustnne deposits, which show more complete rounding, of. the particles in the
 36

coarser layers. It is impossible to distinguish, in samples from percussion boreholes,

between reworked tutfs and the lacustrine sediments as both show characteristic buff
or yellow colouration and evidence of water transport in the rounding of the detrital
fragments. In the sections such light-coloured sedimentary deposits are considered
together and given symbols L, Ln, Le, etc., for the reason that during the early
part of the survey in the Lake Nakuru basin all such formations were identified
as lake beds.
Sikes considered that old land surfaces between the lava flows provided the main
aquifers within the volcanic formations. Whilst this type of aquifer undoubtedly
occurs, it is now known that the principal aquifers are sediments extending to several
hundred feet in thickness, rather than very limited weathered zones forming old
land surfaces between an otherwise unrelieved series of lava flows. Sedimentary forma-
tions deposited in the lake basins since the cessation of the major faulting in the
Rift Valley similarly provide ideal aquifers. They are mainly composed of well—
rounded volcanic grits and diatomites, together with red clays.
The types of aquifer are :—
(i) Sedimentary intercalations in the faulted lava flows.
(ii) Old land surfaces separating lava flows.
(iii) Fissures within the body of the lava flows and welded tuff formations.
(iv) Lacustrine sediments deposited subsequent to the major faulting.
These will now be described in some detail.
Sedimentary Intervalations
The analysis of drilling results given in Appendix A reveals that these comprise
about 90 per cent of the aquifers. The volcanic grit horizons form ideal media through
which water can flow, whilst the clay horizons provide impervious aquicludes pre-
venting the water sinking to deeper levels. It is not uncommon to find several water-
tables within a single thickness of sediments separated by impervious clay layers.
The base of a succession of sediments immediately above the top surface of a lava
flow is the most common position of a water-table, though the water is frequently
struck a few feet down into the body of the flow, since the tops of the trachytic flows
are usually scoriacious, broken up and weathered.
Old Land Surfaces
Water—tables in restricted weathered zones separating flows are not uncommon.
Reddened samples devoid of water-rounded fragments, and restricted to a very few
feet, usually indicate such aquifers, but such reddening is also found in fissures.
Fissure Supplies
The lava flows are essentially impervious and no water can pass through the main
body of the rock. Though the top lay is often scoracious and broken up, allowing
water to pass through, completely permeable flows are unknown, the lower part of the
flow being invariably massive and impervious in texture. The extensive faulting and
fissuring in the Rift Valley provides fissured aquifers within the main body of the
flows, but such aquifers are of limited occurrence since the fissures are usually widely
spaced and steeply inclined. Thus deliberately siting a borehole to penetrate a
fissure is not a practical possibility.
It is a general condition that fissures are more abundant on the downthrow side
of the faults, since the downthrown block has moved relative to the upthrow block
and is more likely to be fractured. Sections exposed in the recently constructed
Rongai tunnel on the Man scarp reveal that extensive fissuring also occurs on the
upthrow side of the faults but the fissures (cutting yellow tufl'), though yielding
adequate supplies, are widely spaced and the difficulty of locating them by any means
 37

from the surface is manifest. In this case the fault planes and fissures immediately
adjacent to the fault zone are infilled with clay and yield no water at all. This condi-
tion may account for a few surprising failures of boreholes immediately on fault
planes, of which C.1001 on Ronda Sisal Estate is an example.
Lacustrine Sediments of More Recent Age than the Main Faulting
There are many boreholes in the Elmenteita and Nakuru areas obtaining very
abundant supplies from these recent lake beds, which‘ have been the subject of exten-
sive investigations beeakey and Nilsson. The conditions of occurrence are exactly
similar to the earlier sedimentary formations. The thick sedimentary formations
of the Rongai area are apparently lake beds of comparatively recent age, since they
contain intercalations of diatomite. The Rongai plain was probably a lake basin;
the lake beds are now completely obscured by a mantle of pyroclastics, black ash
and pumice from Menengai.

CHAPTER XII—GROUNDWATER CONDITIONS

The Sections A-A, B-B and C-C illustrate the groundwater conditions in the
area. It is clear from these sections that there is a hydrostatic connexion between
the underground water bodies throughout the greater part of the area. The hydro-
static level does not usually vary when a second aquifer is struck in a borehole. This
indicates a connexion between successive aquifers in depth. In thick sequences of
sedimentary rocks, water is usually encountered a .few feet 'above the junction
with underlying impervious lava. This suggests that the water bodies consist of shallow
layers of water flowing slowly over the impermeable lava surface. Conditions of
complete saturation of the permeable acquifer are uncommon.
Perched Water—tables
Perched water—tables are mostly encountered on the high ground above or upon
steep scarp slopes. At Kabazi a perched water-table near the surface was encountered
in two boreholes, while in others the rest level was much lower, the first water being
absent. At Njoro a perched water-table yielding very considerable supplies occurs in
the vicinity of the Njoro River. The supply of this aquifer is believed to be derived
from seepages down faults in the Njoro River bed. Measurements of losses in
the Njoro River reveal that the river loses great quantities of water in the‘stretches
in which very high yields from the perched water—table have been obtained. For
instance the loss between Egerton school and the Plant Breeding Station are negligible,
and boreholes in the vicinity have very low yields. Between the Plant Breeding
Station and the Ngata intake pump, where the river turns along a course parallel
to the major fault zone a loss of one cusec has been recorded. Two boreholes
here yielded more than 2,500 gallons per hour from the perched water-table.
The yields in the vicinity of the river are exceedingly variable, and it has been
found that the best yields are obtained close to fault zones crossing the river.
One borehole (C1362) failed to encounter this perched water—table—though sited but
300 yards from the river. Water was not struck until a depth of 600 feet had been
reached and the rest level at 575 feet represents the main hydrostatic surface in this
area. The failure of C.1089 is attributed to the fact that there is no perched water-
table here and the main piezometric surface is too deep for water to be encountered
even in an SOD-ft. deep borehole.
The conditions along the Elburgon—Njoro scarp are extremely variable and there
appears to be a series of perched water-tables, fed by seepages along faults crossing
the river beds. In places where the perched water—table is absent, boreholes may
have to be drilled to 600 feet and more (in one case over 900) to strike the main
water body which is in hydrostatic connexion with the groundwater underlying the
Rongai plain.
 38
Groundwater ridges are commonly built up beneath influent streams. At Nakuru
Industries a very abrupt rise in the hydrostatic contours is attributed to a large
addition to the groundwater from the Ngosur River which sinks underground imme-
diately to the north. This forms a groundwater ridge raising the water-table above
the level of Lake Nakuru in the north-east corner. This feature is responsible for
the permanent supply to the hippo pools. Another groundwater ridge is present
near the Njoro River at Ronda Sisal Estate where C307 has a rest level 30 feet
higher than C306 drilled at the same time.
A perched water-table north of Elmenteita Station is probably due to seepage
from the Nderit River. Elsewhere in the lake basins there is a close correspondence
between the hydrostatic level and the levels of Lakes Nakuru and Elmenteita.
The flattening of the hydraulic gradient near Rongai is very marked. In one
borehole (c.1461) the water rises to within 20 feet of the surface. There is a ground-
water ridge in the vinicity of the Rongai and Molo Rivers, apparently due to seepage
from these rivers into a layer of coarse gravel.

The Relation of Faulting to Groundwater Conditions

That the area is intensely faulted is seen from a glance at the geological map
(Map No. 1). In the Rongai district and the Nakuru—Elmenteita basin the faults are
obscured by later pyroclastics or lake sediments, but still exist beneath this cover.
The faults are open fracture zones, normally devoid of impervious clay filling. They
thus act principally as conduits rather than as barriers. The aquifers within the
older faulted formations—sedimentary intercalations and the old land surfaces—are
displaced by these faults. This forms an irregular series of enclosed flat-lying aquifers,
' coupled by faults which form the connecting links joining them up into one hydro-
static system. The fault which forms the outlet of one aquifer itself provides the
inlet to the next aquifer. This system does not favour artesian conditions, the aquicludes
being disrupted at intervals in the fault zones and the water body being rarely com-
pletely confined. Sub-artesian conditions are, however, widespread. Artesian water has
only been obtained in one borehole in this part of the Rift Valley at Elburgon. This
is situated on a major escarpment and can be attributed to the land surface having
been locally eroded below the piezometric surface. (See Fig. 10.)

_fl0fl’-EEEA‘&'4_
ill...“
ABT£5IAN

l coma/710”: LIA
/D!I IVfl/C/I
[Effildfl SUPPLY
F 51/61/755
(FAULT WIT/I [1555750 ””05” 14"! NAM/Ia.
lMPfl wot/5
cur H: l)

MY 14x5
SMALL ARTESIAN R/JE
DUE TO IM‘EMEAME Cl
CQVER. -

ma “3'5“, ’ ,/ - 2M: Z-Cfl/VDIT/OMS‘ TYPICAL or

'- ’ ' F 771: LAKE NARI/RU 545w.
Fig. 10.——Comparison of a confined aquifer producing artesian conditions and the
open
system of aquifers typiml of the floor of the Rift Valley.
 39

The Cause of Borehole Failures

The most common cause of unproductive boreholes in the area is the great depth
of the piezometric surface. Besides the borehole (C.1089) already referred to, a
large area near Kampi-ya-Moto and McCall’s Sidings has failed to produce one drop
of water from a borehole. It is considered that the area bounded by Menengai Station,
Kampi-ya-Moto Station, McCall’s Sidings, Ol Punyata Station and P. E.’ L. Howard’s
borehole (C2006) is one of very deep hydrostatic conditions. In a few cases it is
possible that drilling to 1,000 feet might produce water, but in the greater part even
that would be of no avail. Further, no appreciable sub-artesian rise can be expected
from any very deep supply obtained. The vicinity of the Menengai crater has long
-been considered to be hopeless as far as groundwater is concerned. But recently
several boreholes have been drilled with success close to the rim. C.1847 but 1,500
yards from the rim was observed to yield more than 3,000 gallons per hour and
C2006 yielding 1,800 gallons per hour is scarcely more than a mile away. C.1807
was drilled on a subsided block of the caldera wall, virtually within the limits of
the caldera, and produced a large yield at less than 350 feet. It is the opinion of
the writer that the greater part of the caldera is underlain by a water-table, there
being no evidence of any discontinuity near to the crater, but due to fissuring and
broken topography drilling within the crater is not likely to be practicable. A large
area of high ground on the northern and western slopes of the volcano is unfavourable
for drilling since the predicted hydrostatic level is too deep for water to be struck
within reasonably economic depth limits.
In the immediate vicinity of the caldera rim, drilling is rendered difficult or
impossible by the presence of numerous fissures—c.1917, A and B, were abandoned
for this reason, though it is believed that had it been mechanically feasible to drill
to a greater depth water would eventually have been struck, possibly at a depth
beyond the practical limits. The conclusions reached as to the effect of the thermal
zone of Menengai volcano on the groundwater conditions is discussed below in a
separate section dealing with steam conditions in general.

High Temperature Conditions and Groundwater

High temperature conditions are of very common occurrence in the Nakuru area.
Boreholes situated along the major fault scarps in particular often yield hot water
and hot springs exist at several places along the Subukia—Bahati scarp. The relation
between thermal conditions and faulting is explained by the fact that hot juvenile
gases rise from the deeper zone of the crust through the openings provided by fault
zones and mingle with the groundwater heating it up to considerable temperatures
C991 and C501 near the Man scarp, C.855, C.1674 and C785 are examples of very
hot boreholes sited close to major scarps.
Steam conditions occur in Menengai crater itself where many rather weak stream
fumaroles issue from the surface. Boreholes c.1066 and 131 encountered steam at
200 feet, south of 01 Punyata Station. The recent Menengai volcanic activity appears
to be related to a north—south zone of instability, and these two boreholes are situated
on the direct continuation of this line. The extent of the steam conditions near 01
Punyata is a matter of conjecture. It is probable that they are localized to the
‘ immediate vicinity of faults. The extreme localization of thermal conditions is illustrated
by the fact that at Tindaress one borehole (C.1674) produced hot water whilst another
(c.1675) less than half a mile away yielded cold water. The low pressure at which
steam issues suggests that it emanates from heated groundwater rather than directly
from juvenile sources. The absence of any mineral content in Rift Valley steam
vents is also in favour of a source in heated groundwater.
A second zone of steam conditions occurs at Eburru Station where numerous
steam vents are condensed for watering cattle. This occurrence is related to the north—
south trending zone of recent basaltic eruption, and it is coincident with north—south
 40

faults on which adjustment has continued up to compartively recent times. Steam

. w..-
conditions are confined to these localities and are unknown elsewhere in the area.
It is clear that such conditions are confined to the vicinity of very recent volcanoes
and not widespread throughout the Rift Valley as suggested by Sikes. The Menengai
caldera, being a relatively early structure, does not define the limits of present-day
steam conditions, and steam conditions are only to be expected near to the recent
secondary centres of eruption of this volcano.

CHAPTER XIII—GEOPHYSICAL METHODS OF PROSPECTING FOR WATER

Detailed geological investigations in connexion with groundwater problems were
only possible to a limited extent prior to 1951. Sikes [2] made a preliminary survey
of the problems based on a geological approach. Between 1940 and 1950, owing to
shortage of available staff and a great increase in the demand for the services of
geologists of the Hydraulic Branch, a largely geophysical approach was adopted.
Attempts were made in the Rongai and Nakuru area to show a relationship between
the average apparent resistivity values obtained in depth probes and the degree of
saturation of the underlying formation. (Unpublished reports by W. H. Reeve and
A. 0. Thompson filed at the Hydraulic Branch, P.W.D., Nairobi.) A direct relation-
ship between the average resistivity value and the tested maximum yield of bore-
holes was also postulated. However, the results of further drilling in the Rongai
area failed to bear out those conclusions. Before leaving the Hydraulic Branch Reeve
concluded that a geological approach to groundwater problems rather than a purely
geophysical one should be attempted as soon as a sufficient staff of geologists was
available.

The present study has revealed that the water bodies are of very limited thick-
ness, conditions of general staturation over considerable thicknesses of rock being
unknown in the area. Such water bodies are likely to give a negligible direct reflection
on a resistivity curve, and might very easily give no reflection at all. A relationship
between yields and the average resistivity values is clearly not to be expected under
such conditions, or, in fact, under any conditions. The presence of a series of aquifers
in hydrostatic connexion throughout the greater part of the area has now been
revealed. Failures of boreholes are almost entirely due to a great disparity between
the ground level and the piezometric surface.

Direct identification of water bodies by resistivity methods being impossible, the

question as to whether indirect methods of prediction are of any value must be con-
sidered. These involve the correlation of changes in slope of the resistivity curve with
breaks in the underlying formations, likely to provide aquifers. A series of resistivity
curves taken on the sites of boreholes in the area are shown in Fig. 11. The great
variety in shape and average apparent resistivity values is at once apparent. It is seen
that similar curves have produced abundant yields in one case, and not a drop of water
in another. A comparison with detailed geological sections reveals that there is little
or no agreement between changes on the curves and breaks in the geological formation.
It is the opinion of the writer that indirect methods of prediction are of no practical
use in volcanic formations, disrupted by a complex series of faults of varying age. To.
use indirect geophysical methods something must be known of the succession before
predictions can be made. In relatively unfaulted stratified rocks extrapolations between
known successions can be made with confidence. In the Rift Valley it is not possible
to do so.

It has never been claimed by geophysicists that resistivity methods are applicable
to such complex structures and discontinuous formations as are seen in the Rift Valley.
It must be concluded that the use of resistivity methods in this area will never yield
information of any value.
 4]

Kit 0 - DHM- 6M5. Kilo -0/IM‘CM5. KILO '01!” - CM5

s an l? m 15'!m 666566”: 5 u Clucma ' I0 I: :0
T.

V 9":
,

’6' A“
ENr/IELYLAKE JED/Miflri‘im; . ' . ’ .' .

,4
‘/
man [out ‘
5;

Nagy—t-..“
3—2
200;! 2w}!

:3 xi?

\g/M’o‘
rt \ i __ _. 3001'! : l 3001'!

tum fr.- If
2365‘1 / _ a a 4mm . man i
E . :‘
”WY

a” l
| _-_._.
\ '7
IOOFZDIK

\ ll

gen-i / : mm

K -0I/M‘C
canny, . 19.1: ”2? :5 ””535 4a a 57
1‘5 W.
5. It
{’3 y "/
z: /. Q \l"
é“ '1'! Ri
U
A
.
//
l ‘Ufirl.
El NIH/Cf. m CAN/C one/7115.
LP.- 1y.
( fl BLACK TIIFF~
-:“J335

?
é/MY 120m 5/; more macwr: "rum
E] EEO/MEWS. ’bwufla sueczsf'
I Pnouaz/r/c mam/m. ire.
/
72.000 Y J 300”

I
/ 400::
Fig. ll.—Comparison of resistivity curves in boreholes drilled in the Nalmru area showing
lack of correlation between depth probes of similar type and results obtained by
drilling.

CHAPTER XIV—SODA DUST FROM LAKE NAKURU

Of economic significance to the district is the nuisance at present being caused by
the removal of soda dust from the dry surface of Lake Nakuru—«to be blown in murky
clouds over the town and surrounding country. The incidence of fluorosis in children
is well known and the high fluoride content of this dust suggests that it could have a
considerable toxic effect on human beings, livestock and growing crops, although this
has not been convincingly established.

The hydrological aspect of this problem has been under investigation and a report
issued (F. Grundy [13]). A brief description of the groundwater aspects is given here.
 42

Lake Nakuru is a shallow pan in the centre of a drainage basin of some 450 square
miles. The recharge of the lake and underlying aquifers is partly from intermittent
surface streams—the Njoro, Lamudiac, Makalia and Nderit rivers—together with
directly absorbed rainfall; and partly by subterranean recharge from the water—table.
The hydrostatic con-tours fall towards the lake, indicating a steady flow of groundwater
into the centre of the basin, and the absence of any subterranean outlet north or south
of the lake.
The lake is at present in the ultimate stage of its existence. In the present climatic
cycle the lakes of East Africa are gradually receding and drying up. Lake Rudolf is
regressing to a marked degree, and Lake Solai has completely dried up. Lake Nakuru
itself has reached a state of frequent alternation between complete dryness and a shallow
expanse of water.
The most recent sediments deposited in the lake consist of a fine compact, silty
clay which is comparatively impervious allowing water to pass through but slowly. The
recharge from the surrounding country reaches the lake area and passes under this
clay layer through earlier deposited and more pervious sediments. It rises slowly through
the clay by capillarity and is evaporated from the surface, leaving behind an encrusta-
tion of soda crystals derived from the saline constituents of the groundwater. It is this
crust of soda which is caught up by the wind and blown up into dust clouds. The
artesian pressure under the lake is small. Not more than 4 ft. of artesian head has
been recorded, It is attributed to the relatively impervious clay layer which confines
the water under the lake in an extensive enclosed aquifer. Owing to the flat nature
of the terrain and the absence of any enclosing acquiclude beyond the limits of the lake,
the artesian head is negligible. .
The periodical filling of the lake after rain seldom extends to more than a few
inches above the clay surface and probably depends entirely upon increments of surface
water. In very wet years, however, the lake fills up to a depth of 5 ft. In 1951—1952
it remained full for a period of 18 months. The water cannot sink through the saturated
clay, consequently, it remains until evaporated, leaving once again a dry, dusty surface.
Local opinion favours the idea of keeping part of the lake full by artificial means. In
this choice it relies on the local belief that “great quantities of artesian water must
exist beneath the lake”. This is a fallacy based on an inability to distinguish between
confined aquifers and open systems. ‘
The conditions whereby such an artesian supply could be found are illustrated in
a
Fig. 10. The first drawing shows a continuous confined aquifer inclined in such
way as to produce an artesian head. The second drawing shows the condition of open
aquifers and confined aquifers of limited extent broken up by open fault fractures—
the condition typical of the Nakuru area. Schemes to concrete the faults in river beds,
and other undertakings might produce a little increase in the water supply to the lake,
but the chance of obtaining the estimated requirements of about 30 cusecs from
artesian supply is infinitesimal. The writer is of the opinion that the ultimate solution
to the problem will be found in speeding-up the natural process of reclamation by
vegetation rather than by artificial rejuvenation of the lake.
CHAPTER XV—THE NJORO RIVER
Geological Note
The Njoro River follows a zig-zag course from the Egerton School to Glanjoro
Farm, alternately following the line of north-south fault zones and cutting eatswards
across the area between the fault zones. In the Njoro area the older faulted lavas and
tuffs are covered by compact black ash from Menengai, usually about 30 to 50 ft. thick.
This black ash, known as building stone, is usually concentrated in the river valleys,
and the Njoro River is no exception. The river is cut down through overlying uncon-
solidated pumice and ash and runs over the hard upper surface of the black ash. This
ash has been disrupted by minor faults on the older north—south lines and probably
about 8,000 to 10,000 years old.
 43

A study of the groundwater conditions indicates the presence of perched water-

tables in the vicinity of the Njoro River. This is attributed to losses down the fault
zones. Lower down the Njoro River movements are still proceeding in a fault trough
extending from Butleigh Stone Quarry through Glanjoro to Ronda Sisal Estate.
Evidence of the present-day adjustments in the Rift Valley is invariably found in fault
troughs. In this case surface fissures on Ronda Sisal Estate and the open fissures of
Proctor’s and David’s Faults, down which the Njoro River loses a great deal of the
water remaining in it at this point, clearly indicate that tectonic adjustments on a small '
scale are still proceeding. .
The Njoro River tapers completely on reaching the Lake Nakum plain, firstly,
because of these losses in the fissure zones, and secondly because its bed runs over
porous pumice and lake sediments and the water sinks to join the water-table which
feeds Lake Nakuru.

CHAPTER XVI—THE POSSIBILITY OF UTILIZING STEAM AS A SOURCE

OF POWER IN THE RIFI' VALLEY
This possibility has recently been suggested by the late Mr. James Scott [8]. The
paper referred to is based on very sweeping conclusions about the geology of the Rift
Valley, which are not supported by detailed investigations. For instance, Mr. Scott
claims that the proportion of pyroclastic rocks to lavas in the Rift Valley exceeds
80 per cent. This is based on a computation of surface exposures. In bulk, lavas greatly
predominate on the floor of the Rift Valley, though pyroclastics form a superficial
mantle over wide areas. Again, Mr. Scott considers that Menengai volcano is almost
entirely composed of pyroclastic material, a conclusion with which the writer, who
has examined the caldera in detail, completely disagrees. Mr. Scott’s theory of the
origin of the Rift Valley due to the weight of the ice on the polar regions during the
Pleistocene era is refuted by the geological evidence—the Rifit Vally is far older than
the Pleistocene glaciation.

Steam under pressure possibly does exist at depth in the Rift Valley, though the
broken and faulted nature of the overlying rock would tend to allow a continuous
release, and prevent any great pressure being built up. The idea of a fold of Kamasian
sediments acting as a cap retaining the steam is not accepted by the writer. These sedi-
ments are faulted and broken to a considerable extent, and it is extremely doubtful if
such a perfect, continuous structure as visualized by Mr. Scott exists anywhere in the
Kenya Rift Valley.
That steam under some pressure may be found by deep drilling is not completely
impossible, but three questions should be borne in mind:—
(i) Whether there is any likelihood of steam being encountered in sufficient volume
and under sufficient pressure.‘
(ii) Whether there is a demand for electricity in the vicinity to warrant such an
expensive scheme.
(iii) Whether the engineering problems are surmountable. In the Lardarello Valley
(Italy) where a similar scheme has been put into operation, many problems
due mainly to corrosion of pipes have been encountered, and considerable
funds have to be spent yearly on replacement.
Such problems would be more critical in Kenya, a country in which engineering
equipment has to be transported thousands of miles at prohibitive cost.
Three localities in the Nakuru area are pOSsible sources of steam for pOWer
projectsz— .
(i) Eburru—Elmenteita. (ii) Menengai.‘(iii) Kamasia.
* A paper by the writer dealing with this problem was read at the XXth International Geological
Congress, in 1956, before revision of this report. The writer concludes that the discovery of
steam in any quantity and under sufficient pressure is extremely improbable.
 44

REFERENCES
[1] J. W. Gregory, The Rift Valleys and Geology of East A frica, 1921.
[2] H. L. Sikes, The Underground Water Resources of Kenya Colony, 1934.
[3] R. M. Shackleton, The Kavirondo Rift Valley, 1951. ‘
[4] R. B. McConnell, Rift and Shield Structure in East Africa,'1948.
[5] E. Nilsson, Quarternary Glaciations and Pluvial Lakes in East Africa, 1932.

[6] L. S. B. Leakey and J. D. Solomon, The Stone Age Cultures of Kenya, 1931.
[7] L. s. B. Leakey, East African Lakes, 1932. ’
[8] J. Scott, The Great Rift Valley and its Economic Possibilities, 1953.
[9] S. M. Cole, An Outline of the Geology of East Africa, 1950. (Contains a useful list
of references relating to the geology of East Africa.)
[10] J. 1e Conte, 0n the Origin of Normal li‘aults and the Structure of the Basin Range
-
Region, 1889.
[11] S. J. Shand, Rift Valley Impressions, 1936.
[12] S. J. Shand, The Rocks of the Kedong Scarp, Kenya Rift Valley, 1937.
[13] F. Grundy, Lake Nakuru, 1953.
[14] R. M. Shackleton, Geology of the Nyeri Area, 1945.
[15] R. M. Shackleton, Pleistocene Movements in the Gregory Rift Valley, 1955.

ACKNOWLEDGMENTS
The author is indebted to Mr. S. Stock and Mr. M. W. Jowett for final revision
of text and illuStrations before submission to the Government Printer.
 45
APPENDIX
ANALYSIS OF DRILLING RESULTS
N.B.———The figures contained in this appendix include the drilling results up to the
end of April, 1956.
1. NAKURU _
Number of boreholes drilled . . . . . . . . .. 56
Successful boreholes (over 208 gallons/hour) .. .. 54
Average yield in gallons/hour .. .. .. .. .. 2,285
Average depth in feet . . . . . . . .' . . . .. 439
Unsuccessful boreholes . . . . . . . . .. . . 2

(a) C.822——S. Greensted

Not drilled deep enough to strike the water-table (probably stopped 50 ft.
short).
(b) C.1001——Ronda Sisal
Sited on a major fault, fissuring caused loss of drilling water and caving.

2. RONGAI
Number of boreholes drilled .. . . . . .. . . 34
Successful boreholes .. . . . . . . .. . . 28
Average yield in gallons /hour . . . . . . . . . . 1,212
Average depth in feet .. . . .. . . . . . . 425
Unsuccessful boreholes . . . . . . . . .. . . 6

(a) C.62——I. W. Whitmore

a
This borehole, an early attempt at obtaining groundwater, was stopped only
few feet above the hydrostatic level. It is clear that deepening to about
600 ft. would produce a satisfactory supply.
(b) C.l47——-Lord Francis Scott (Deloraine)
(c) C.148—L0rd Francis Scott (Deloraine)
Shallow boreholes which obtained only a limited supply. A later borehole
drilled much deeper in the vicinity (0.1647) obtained a satisfactory supply.
(d) C.322—-Lord Francis Scott (Deloraine)
This borehole was drilled to 600 ft., below the probable hydrostatic level. It is
situated on the lowest spurs of the Mau scarp, in what is undoubtedly a
highly faulted zone, and it may be that faulting and fissuring have some
bearing on the complete absence of any aquifer.
(e) C.826—Captain H. Barclay
The hydrostatic level in the vicinity of this borehole is estimated at about
5,750 ft. above sea level. As the altitude of the borehole is 6,420 ft., water
can only be expected below 770 ft. Drilling beyond the existing depth of
800 ft. would probably have produced a satisfactory supply but the rest
level of about 770 ft. would have made pumping very costly.
(f)C.2152—Mrs. H. McKenzie

3. SOLAI—NORTH MENENGAI
Number of boreholes drilled . . .. .. .. . . 44
Successful boreholes .. . . .. . . . . .. 36
Average yield in gallons/ hour .. .. .. .. 1,677
Average depth in feet .. .. .. .. .. .. 436
Unsuccessful boreholes . . .. . . .. .. . . 8 .
 46

(a) 122—A. J. Simpson

Shallow. Further drilling would probably have encountered steam conditions.
. g (b) 131—A. J. Simpson
Steam conditions encountered.
(c) C.257—Manuel and Keppe
Tested yield only 2,926 gallons per day. The cause of this failure is obscure.
‘ It is suggested that the aquifer might have been sealed off by the puddling
effect of the drill, in which case surging might have produce a much higher
yield. Alternatively, the low yield compared with that of C.270 some
' 300 yards downstream, may be due to the fact that the latter may have
derived its greater supply by seepage down a well-exposed fault, which
traverses the river between the two boreholes. _
(d) C.341—H. F. Ward, Kabazi
(e) C.534—H. F. Ward, Kabazi
These boreholes were drilled in a deep fault trough largely infilled with clay;
a little water was obtained in C341, some 900 gallons per day. It is
probable that the aquifers encountered in the high yielding boreholes, a
few hundred yards to the east, are sealed off by this impermeable clay.
(f) C.lO66—A. J. Simpson
Steam conditions encountered.
(g) 0.1917 (A and B)—J. A. Seys
Sited within a quarter of a mile of Menengai caldera. Intense fissuring related
to the caldera subsidence rendered drilling beyond 300 ft. impossible.
(h) C.2162—S. 0. V. Hodge

4. KAMfI-YA-MO’I‘O—LOMOLO
Number of boreholes drilled~ .. .. .. .. .. 13
Successful boreholes .. . .. .. .. .. 5‘
Average yield in gallons/hour .. .. .. 1,685
Average depth in feet . . .. .. .. .. .. 525
'Unsuccessful boreholes . . .. .. .. .. .. 8
V(a)H. l—E. G. Thomhill
(b) C.557——Land Bank
' (c)C.591—E. G. Thomhill
(d) C.615—E. G. Thornhill
(e) C.642—J. Stanning
(f) C.680—W. E. McLean
, (g) C.1036—-W. Evans
(h) 01809—11; cf. 1. Hissey
All due to great depth of the hydrostatic level in the highly fissured and
faulted Kampi-ya-Moto—McCall’s Sidings area.

Drilling to 800 ft. on the site of 0642 would probably obtain a supply. In
the case of the other abortive sites drilling to 1,000 ft. might fail to obtain
a supply, and 1n the event of any very deep aquifer being struck, the rest
level would show little or no sub-artesian rise, making pumping costly
and difficult'in the extreme.
 47

5. ELMEN’IEITA _ c
Number of boreholes drilled .. .. .. / .. .. 33
Successful boreholes .. .. .. .. .. .. 29
Average yield in gallons / hour . . .. .. .. . . 1,460
Average depth in feet .. .. .. .. .. .. 489
Unsuccessful boreholes .. .. .. .. .. .. 4

(a) C.22/88——Cresswell Ltd.

Abandoned at 910 ft.
(b) C.254l—Block Estates Ltd.
(c) C.2542——Block Estates Ltd.
(d) C.2543—Block Estates Ltd.
Abandoned owing to drilling difficulties.
6. NJORO (NORTH)
Number of boreholes drilled .. .. .. . . .. 8
Successful boreholes . . . . .. . . .. .. 7
Average yield in gallons/hour .. .. .. .. .. 1,687
Average depth in feet . . .. .. .. .. .. 521
Unsuccessful boreholes . . ‘. . .. .. .. .. 1

(a) C.1089——G. B. Farms ' , g

Drilled to 800 ft. Dry. The higher perched water-table was not encountered
(? restricted to the vicinity of the Njoro River). The topographical
position of this borehole indicates that the main lower water-table might
be encountered by drilling on to 900 or 1,000 ft., but there would be
little or no sub-artesian rise.
 .3055: 3.0 05308505: 9::
LEE 3t 3:05:38 03: 533 E 20:33 03:. £06 0Nm6 new . . 02-0mm .03 0:. 0.5.: .. .. . . .03 33w 00:3: “3:0
. .533 wEEE 00 32
08:30 3:53: 00300001003 :30 3 :0 005.5 026 .| l 000 :2 . .. I .05 3m 30:3: 80:0
.030: «>3: 3.303.: 03:20:.“— 030 .5035:
0033053: .838: 30863 03: 533 :30. 00.0.6 036 0; . . mov .0mm 6? 000 . . . 35.23.00 52 .D .0 800
.0530 30:5t 5.5q 003.: .
50.: 030000 ”308000 0:535 003: 308 0:000: 03.6 356 m . . . . n 5: 00m; . . _ 030586 003..— 3332 030
.30: 33— 320: 3:05:50 20— .533 03.6 $06 6mm . . . . 0mm man 00%: . . . . .555... .n .0 ~30
33: 0.63 3.58600 93— E 3&6 N006 03. . . . 2;. 03. 0mm , . . . . 5339 .0 .m 030
@300 03:02..
00:98.3 mE33000 003.050. 0:0. 20 E 53>? v26 v36 00m . . 00¢ 63. v3. 00m.~ . . E5020 cow—0:3: 0:20 3:3.— 025
.533 8:
330: «>3: E 00330505 3:02:08 22 533 0Nv6 «~06 mom . . . . .00 man 000.: . . . . :ofifl. 0o :otomm 0.8.: mmmo
6:58:08 03: 533 00 85:03 05:; 2:6 036 Ru A . . mmn Sm 000.: . . . . . 0003 .m .2 030
0305: 002 mmu6 l .l 1! mm». :2 . . . . . . 0320:0000 .m wo
0305— 82 mmu6 $.06 mom . . A 0? gm 00:.N . . 0333095 .m mSO
do: 35> 533 ”030: «>3: 2.303: .
0::oaonn 5 0033052: 95.503 0:: 53>? 2:6 0:6 in . . Zn KS. .02 own 000.: . 5:023:00 003: .93."— 300
.Hohscd 320: acute: 3.0 30350::
5:3 3:08:03. 0:: 533 00 85:000. 5:: 926 306 00: A . 03 .02 006 0006 . . 5:332:52 2.5032 N30
0302.550 0:: 533 00 00:03.00 0—0:: N36 0006 2N 03 6mm Now 000.N . . . . Bacon—02 .< .n 02.0
.300 03:03: 0:50:05: 0 E .03
300 «>3: 0 05:50:: acute: baa—58:08 < .62 0006 0006 0: . . 03.02 .03 0mm +0006 . . . . . . 3032 .Q 050
.28 35> 53>? acute: EU HEN
.0300
48

3.300.: 29:28.3 5 0033053: 3:08:00»

03— .533 .00 50.0 005.5: 35.50 00 000m .3: 056 $06 03 .. .. 000 6% 00m 03.: .. : .. 3:202 .3 .Q 030
£305: 87: 3:6 .. . . . . SN 03 03.: . . A . . .5326”: 3.5v 0000
6:55:08 03: 533 00 0255000 05:; 026 0006 0- . . . . N: 6NN «mm 0006 . . . A . 20.8002 :4. A :NmU
.5305. SZ 3—6 :06 v3. 0n». .0mN do: .02 :3. 0E; 03.8002 .< .n 963.0
0::8 533 3.5q 33.: 033530
000E 0.3 5:30 .3 00:00.50 300E600, 2.2 53>? 036 036 3 . . . . me 0m~ 00m.~ . . . . . .05 30mm 30:3: >000
50:3 .533 ”3332 003.: 032530 _
20E 0.3 5:30 .3 0038.50 3.5.5000 03. .E. “Maw—>9
. 056 056 N6 . . mmN .0N~.06 own 0006 . . . . .03 306 3003: 6000
69: Mm .030: «>0: 8.300.: u::o=o:n war: . .
.505.— ::3 53:00 3 35.5000 03— 533 00 03m 036 n56 non . . . . 0N». 02. 0mm: . . 88.9360 23D .m .n .0 SNU
.macofimuom 2.2 53>? 0: . . RN08 mmu com.” . . . . . 03.5305 5:532 ”and
5:58:50 $3 53>? 306 :06 0w . . 00m 60: 0m~ 0666 . . . . . .505n 3332 03.0
0:58:50. 03: 53>.» 3—6 9&6 02 . . . mnu-mm_ MN 000.». . . . 30:09.65 3332 «2.0
.0305: 82 3:6 www6 0mm . . SN 62 mum 00w; . . 300.5265 3332 wm.<m
_0>0.— 030:6 ..=._
33:5: 0.3 5:50< 00 0.532 003:3. 00050040 320: :uahm 53>» 500C \30 5:30 .02
E 20;
. .55.: .53: 085E
Dana—<21."
 . . 898m con 83 . . . . . . . . .U.._..<.m oSNU
5.8%.. 352.5 9 3325033 35.5 .1 1 8: 330
we: . . 2465 3;. 89m . . . .U.P.<.m
.tono: 353.8 8 b53333 35:9 I I
.33: Ema—0:02 05 . . 83:95 .3 5.30
8:6 com . . 8m 98 8:3 . .
3313:: 32m 258:9 ES :21...» mo 8:2:Muwm m :: owed .
. .9. Eu news
.. SYN—m pm». 83. .. .. 3:33:52 5.5a
3:» 2.2 3:25am b35858 x25.
.633 :36 :wnw. com
“Eu .
wéflm 638,32 8.30m:
.Ed. .. 23:68:: 2332 $20
83:5 2:: Bo 29: 8 2.0 086 8% 8m .. Sn 3. Sow
02:22: 5:88
.65.: hbflm date: .020 o:nuoE._a_ .. .. .. 3302 ._. cow—U
«2.0 owaw SN .. n3 .mem 2m coca ..
ca u>onu $5568.22 :8?» .8 8:8 :25 <
.950: «>3 $20
Ga . . mmm Nmm coo; . . . 33:0». .3
wig—38 55.5: at» 0:30—03 b55868 s 5 wood Ema
.30: «>2 8.30:: .< .H wan—U
23 . . mm». 34 ooo.~ . 320002
03:95:: a mix—5:5: 3:05:q Eu. :8“? find «Ga
Am .wou 3: .533 8:.— . .553. A. .0 22.0
.5s cwmd coma awn . . mum one
8: 30> .36: :22 5950: 3.58%?“ 22.u:::m
.633 .330 go a?» 0: mm":
:EEE .Edd
43:3 .3923a .533 .95v 331— 032.038 330
no cm n2 8:. . . . . . . .2): :35 5:3—
29: 5:58 :u E 3:83: $5E€3 22 .533 E 03% whim
.84: mm“: 6::2
b93023: 55>? 5.5a 8:5 37:25 3:0
mm , . . . on. o: co?” . . 33m .22
82.: 3:53 :u E 3:83: 3:0EEuw 2.2 .533 E 03% «2.6 . . 3:30 .m Gone—U
036 on o; .3.“ .3: new ooofi .
53:5. Mi 036
. 3: 320
2m .n .<
49

n56 mmmd ow: , a? den .02 0m». 8m; . . .

2:833 35863 Eu— ..oEB .«o 3:053. :25 < own 08.. . . Bacon—o2 .< A . $20
.358v 22 :855 .«0 8:258 :25 < 036 wood 3: . .. . no». dn. .
£055: .220 32:09:: . . 3302 .n :20
mnmfi com . . man 63 men coed . .H
:a 26:: 3:05:32: Es: .533 .«0 3:258 .E. gownw<
. m2.» V
cmnm 6::3 b22083.» .533 .26: «>2 $20
wEH awn .. 8m om». 8o_~ .. .. 0:580 Beam .¥.m.<.m
222...: "5:293 : 320: 3:08:q ES .533 510
30:? .83: 0238.31— :_ $3: . . . . 65 32m 3:0: 0520
2.06 aP.m mam . . cmm .02 mon ooo.~ . .
33:30 35868 22 :25? :0 8:258 :35 E
:235553: :oNEo: .320 m .< .m «mo—O
3— 0mm 6mm 63 68 mom 3m . .00 i— Q
32335: ”$5563 32 .853 .«0 3:253.950: :25. 926 Saw
. . mm». .03 gm o3; . . 0:370 :oEEm: @520 .53 96:0
@3838 ooflsm :5: 20 3—6 Vmwfi Em . .
5 32%.: cum—0:93
_ .533 ”cm .558: ESEEEom .532 .
a 5 W5 .950: 2.38.: 03:20:: 3:333. .3 .m 330
and own : 92. 6% o? 03..— .. .. .326t
8:0: 8 b85838 25:35 Em :0 om“: 2: 2 ommd
:33 83:6 .5
583 £900 :50 5:30 .02 .
2:: :o:5< :0 23:2 225:? coon—gum 320m xosbm
fiEEom
5 Eu;
4:35 .55: coach
Q§UV|DM DM<ZIA
 33:09:: wEbuouas 5:5 83:00 33$
$2: 5 3.293%
.933 EB 3:» £529, £5568 .8 3:253. £022. 036 mu fl 6 N H N . . . EN 8m ooo.u . . 532.5300 vmom .Q.3.A ~80
maomiomfim unite—E:
35 8 0335.3 ;
5.:5 83:8 309$
.330 can at» £529, Jase-£380 ooaoawow x35. ommd ”mud mm . . . manna En can; . cocuabmaou 30M 635$ mmho
.mmfio was £5 2529, £58883 3533. x029 ofld oohn omv . . . . 82:? mg 83 . . . . . 303t .22 «.80
.nofiuon >30 038036
m=o_>._oqE_ 5 952a bBaBoEEm Sham“
28 “£20 was at» £529» £20833. 22 533 256 Soa Rm . . . . 3m 63 wow 8: . . . . . £08a .3 A “$0
.mom .
.5505— 32 026 >26 m- .Nmm in den 6% cow cum . . . . zoom £255 98}— 3mo
. acute: Em Emynofiaom
.532 a E Em Kong—am 9—2 30 .5 Emma .
233.5 3:893 E van ”8:323 mama—noun;
53 88:8 “a mucoEEom E E ”flown—um 3:3. od Nmnd wmm 2o .03 d; o3 ooa .. .. CawoOv neowunn .m .< Emu
Aw“ .mov no: “can? SE .933 was
at» 858—3 icofimuow 22 .533 .«o 3:8 oh o26 aid 3N . . . . now own 03 . . . Camomuv noomunfl .m .< Emu
z oomé I I I cow :2. . Scum £83m v.84 NNmO
.w 20
can warm 8520.» .mEoEEom mo canon—won Nutty 02,0 mixn pm .. .. can m; com . . . . nomkom .m .0 mmNU
.m Eu
“an at» £52? Econ—ES .3 8553» «SEE ooud 236 now A . Own 2:. 8o; . . . . . . 9::a .1 .3 was
.5505 82 03% ~36 mm . . a v3 ca . . 3582o 38m £85m “ES 36
0 .938 can 3:» £520>I£58€8 23 H33$ 036 2nd 3 . . mmu .02 Em mm. . . 955209 308 £88m v.5..— 5:0
5 .5583 82 cued I N. . . . . 3m omm o3 . . 853209 38m ”Emmi q ago
23% I I I new :2 . . . . . . 22:a .3 ._ «no
:UZOZI."
€32 3535 8 $53033 855 men .. : m3. .3 N: 8°.” .. : .. 328 .m .: 8&0
.983“ 359.8 3 baaozcomnsw 35.5 wmn . . . . of. Km». 8m coo . “ounoQ .2 .E .Q 330
6368 A. 2%
€88 :2:o 8 bacon—magnum “EEG NNm .N. 3v 6% don can con.” . . . Sham aunnmuomam D8q
.282 $538 8 328333 855 cum : : 8m .8” 2m 83 . .. .. 228902 .< ._. ~88
£32 3538 B b33333 3:19 can . . . . 8v 6: Sn coca . unsohéotum .M 320
€82 35min. 2 339.5033 22q SN . . . . emm mmm coca . . uafiohéotam .M vmmuo
.258 3538 2 328333 3:5 o: .. .. 81.: 8.. 83 . .. :E .m .3 .34 880
€32 3528 3 22.3833 3:5 “2 .. : 3m 3. 83 . . . . a .36 2.3; Emu
£82 3538 8 323333. 855 c8 3. a: 52 6% «a. 02.” . . e231: .4 a. .m.<.m Sumo
€82 3538 8 3.82333 855 m: .. : own o: Stu : : .. $52.38 P a 0:8
$34 8&Sm .5
3.380% 2.8 Hubnw< we 2532 2.552 305q 320m Mosbm 53>? Egon :80 8:30 .02
. . 5 Eu;
#55 and coach
EEDVIDMDviIA
 @5556“ Samoa N: mm 5505— bfiafico 5.5 , . .
be . . . . of #3 ooohm . . . . . . . . m2o .0 330
2320 93 Bin 2522, .oonoacom #820833 ad 226
.565— uoz l l we . . 9: .9; .2 Sn 8o; . . . . 85x83 Emaom $30
. l .l. :I I cow :2 . . . . 323—02 .3 .EE «END
SN w in . . . . 03.09. 0:. of. . . . . . . wmtsi .< .m $30
6.3—0 23 Btu £529 cocoagwbfimfimxvom onvd
.8: no: :5 $30
enmfi ~86 3m .. 1 omnénm m3 804 2 .. .. EEO .m .0
83:00 268" .3533 385863 Mo owns 2‘.
:85 3313:: £15 58:8 3 mam—o can Etw
Owed ammfi Gm .. cmmémm .2." Sn 8oz” .. .. .. .r 23m a 820
own—«20> Eco—Evan 22 .1385 mo gjwflcnwm «wwwh
259829 5953 @3352q flaw 335530.".
5 En 2S ofioaonn msomiomfi_ «Ea—“3::
5:» 88:8 30%. 33338:: can ”Eco—=63
5&3: 8:: 83 83 2: . one :3 .52 5:. coo; . 3532o :oum $25.5 0.54 :20
22 53; Mo 85:3“ x25 :_ E .Edd
flannm é>£ 3313:: 5:5 68:8 “a 93—9
we cocoa—com oP 036 m:.v fl. .. 0mm .02 do mom 8: .. .. : nEmckoHRmaoM SEQ
can at» 3:83.» £5838
, .Edd wank £22
9.5325 ~33 88cc”. 3 .Edd waum .320
£5833 0 85:38 x35. 026 ~26 mm .. .. cm: .3. SN octm .. .. . Eman— .< do! «0.30
was man 258—?
.E. d wank .mwflo
ENE «26 an .. .. cm: .02 9: Sea .. .. : gown—.5 .m .< 320
can 31m 252? .mfioEEom We 8:033. :25.
.E. d Ynflm .23....
25. Ewe 026 on .. .. 31% EN ooefi : : : :oowuna .h .< 830
can at» ciao—9' £5583 .«0 8:033,
d>3 mama—.59.: 5:» 65:8 3 m 20
find em 10 oh .. .. com .03 o2. oodm : .. .. .85 non—3m Sum—U
was 31»352?» Jana—=53 ho 3:033 M35.
51

.m .20
8533 ”EFF 2% Raw 8m 33:. .313 8m 83 : : wackggdd :20
E; at» 2529 waning be
.Edd ~5o €52 wfibuovua 5:» 83:8
23 83 8 .. own .5 .8 o: 85 .. : .. "55505 .m 3S0
3 $20 2; 2E 2&2? 555.com m5 .533
.E. . Sum N2. 83 ._ : .. 208255 .2: 820
.56 2; at» 2522 .2353 22 535 23 Ed Na .. .. 21.8
.Edd manm dam—ocean 323 93—0 9.5
Eu: .8 8:3 x05. 83 23 n: .. : ea 52 am So : .. eamoov 8&25 .m .< 38
at» 2528 £585..“ 22
dam—ocean E 63285::
EM? 2% Raw n5 : : a». ma. So .. .. .. 29533 .3; 38
930 23 at» 252o>|§oafi.m a So 22
new at
in... 8a... 33 a: .. : 832 on 83 : .. .. 52:3 .32 E6
2529 £5563 333 ho 8:268
32 o8 :2 .. .
.. .. 333 .m .66 38
on? . I I |
5:» 32:33:5 8
8am
2:3u>on <v
maom>uom8_ unite—E:
and 2 3 2 m . . . . Rm 9% . 8m; . . 532:;q 98M .93.; 88
.936 2.3 at» 3523 $58.8”? 8533 x03...
#054 oofiBm .E
2:532 coon—gum 320m gosbm .633 £30 :«0 .5550 .02
Eta—com Ea 3:53.... we EBNZ
, E 20;
, .533 Sum uBmoH
G§DvI—<OZOMI.N
 .EE «.5a I I I I
.3 2:: .m 553 cam 60.82385 mat—Emu 8:35 SN :2 3v
026 I I I Km :2 AB
.8383 oz . . I . . whom .4 H :30
03% Bow wmv . . . . now 02. con _ . . 23$ steamy 50$ .0
.32“ 830
«>3 50339 vows—«225 3553“ 3537.3 5 Ema 036 W; . A 2m
. :83 can 8o 80:". 333‘ .m .M a. .m .M Ema
.
0:20:23 S 322 «5:233 8&3” 82 2o E 02% So.“ 2.. : .. «mm «a. 8m 8%
.=3oE_ 82 Sofi envfi
.. $.m same 50.: .o «86
Ev . A . . com coo occur. . . .mofimm .3 .H 220
$32 Maw—.252 3BonImEoEE8 0:823 5 236 55% Em A A . 0mm 63
.G 0mm 3 widow o.— m2 0% fl whom < H 586
058 @0323 86 .Ed as soon moi—cum .33 am
voSmmc no ouutsm can 20 39.383 3.0a «>3 5 mNm.m 31m .
.3333 U22 20 .5 3 95m gamma ewv «>2 5 Em cam A A . 98 On... Sea o_N:oMouE 2m m 320
”35858 39.58 .833 E 95:: a “EN A A . .
USN an“ Sud ovod o3 .. mNm .ONN 6% 3m 8—.N
.25863. 3958 .623 E . .. .m.~=_am..m.n~ 320
026 exam of . A . . 03. con coo.” .
.835 “com I m 320
began .20: E 30— Bow. .m3oc «>3 AC u===€
3.30.9:
«
3.228% E 383285 25569. 32:8 .6235 code owed 8v . . . . 8m
.38 «>3 5 353.35 35568 @0258 5:35 awn ova . . . . . . “com .1 .I 320
cove o-6 2: . . . . n3” don Em SYN 2:8:29 333a: .22
.933. «>2 5 “yoga—«835 ScoEEom vogue 533 owmfi mvww mam . A . . 3v an” 820
$5833 2533 E can “Econ—63. m: 8N; . . . . “Ego: .m .0 320
we 8:8 £02» « flue—.3 «>2 355983 E E 036 . A A
d>£ 323 satanic b85888
306 com A . 02. .03 m5. com; . . . . 28.0.: m 0 3:0
coves?— ..3a3 E o5.» 026 o: . . A A own 03 8o." . . “Eu—9: Am .0 3.20
.3 95 25E 5 :52: «>2 5 “an 8% m: : A. com in 83 A $65 $55: no
52

.3q%8 3958 .533 onnd M26 3 . . $20

.BcoEEom @358 can 3n coed . cue—coon: m < :20
.533 W36 5 1w Q: can £2 won Sod A
A8335 30: 82:38 88% 856 I I I . . ABA 323:2. $20
.035 .633 3:23 .~>£ mom :2 . . . 5355 H A< 80—0
6:3: .
:33 o5 ouzonofi 5 8333235 EEBEBm on: :3. on we a: .3 am
.
.053 >253 venouom .32 an 82 £35: “53 m m 50
.mnn dun
:23 was ouzoaonn E 30:23..a EmanoEEom onnK EmK mm .Eméa—
. .
. 432 6: .3 mon 81m . _ €3a v.33 m I omno
:33 23 8:829 5 8233835 @255 om 3% 8:. 8m : : can .8
.3353 W2 33. I I
8N 8% . €33: “53 .m .m 88
A .33: I 9% =2 . €3q Fag .m .: Sumo
:23 Ea 0329.93 E 32382a bqEEom nwmfi mm; o% .
A i=9: :5 5 . mom in com.” . . snaps: 2&5 .m .E 2.8
E 3% 35 83 o3.» 08 . . . . EN 3“ 9.
.3353 02 23 . , €35: 33 .m .m 2.8
432
3v.“ 3 . . . . 2 o: on; . . . 030M a 3232 28
E 333385 35568 wows—>2 “25$ 03% Nana ma . . 32.2 new
A23o 305 32:28 Exam ommfi I
02 . . . 930M a. 3:52 END
I I v5 :2 . . . . dawns—mm A .< 2:
82 l I I w: :z . £855 a .< NS
:33 uoflBw .E
wigs—om 2.8 5.8”? we BENZ 3332 30381 323 u—onbm .833 finen— 350 5:30 .02
.335
E 20;
.53 853.
:52n EMOZIHASOd
 @o .3850 W W W W
.Edd
gas—glzfieaow 22 52.3 83 8 83 _ m2 1 I . ma. .2: .2: v2. 82 . I . 253m 0232 _ :8
60 Em aim
<5<2<M .
.maou 2:888 88:; 22 Bo “in“ 96”.. 3“ . an an on". . I I I £536 a $8
I .533 32 as m g... E. I I 8“ 5 8% . . I I 8386 a $8
.cfimmhmz 83 5 v w? o: ~2 8~ ~ . I . 8:55 .n 38
238L888 5 8:28? *8.“ ‘ 83 m3. . mm mm 83 . . . I . I 2355 In :8
us: 2° 5 2&3 . . I . . 8256 .H ~80
82 I I I I an :2
30204
as.“ I I I 8 =2 . I . ~38: a .o I: .88
8% I I I as :2 I . I . I. 2:5 .3 £20
83 I I I ~8 =2 . I 52002 .m .3 88
$3 I I I o~m =2 . . I . . =E=§F .o .m :8
8”.“ I I I 2~ :z . . I : E 59F d .m :8
RE I I I OS :2 I I 85 .2332 < 25 ES :8
an.“ I I I o: :2 . I I I I I =EESF .0 Im E
0H02.<>-E2<V~
SOEOAIOHOE.§-EE§I.V
53

€32 352.5 2 3325033 355 I I n: . I . I an 88 8 I I . . £555 338

£38 3528 3 362.333 855 I I 8_ . . I I To: $~ 8~.~ : 8&8: IH In ~88
.58.. 3525 2 858%.; 3:5 I I 3~ . I I 83: 8». 8m.— . I . I 538 In 888
€82 3538 2 85:33”: 855 I I o: . . I 8n .2: 8m 83 . . I 238 .o .o $~8
€32 8528 3 232.333 855 I I 8.2 08.8
. 8m 83 . I. 2380 .o .w ~w~8
8 858053 855 I I 3 I . a: 5: a: 83 . . . :82? .0 .< .3 ~88
882 3538 I I 336 I: In 2~8
€82 3238 8 8:883 855 $8 83 8~ 83$ 8“ 8“.
8 283333 855 ~86 an.“ 8“ . 9% Sn 8o m . I I >5 .m m .3 a 2:8
.282 3:38 . I I . . Beam .0 .3 8.8
€82 3598 2 2325033 855 I I 3 . $28 an 8wI~
3538 2 833833 355 I I 8“ . n: 8“ 8 I I . . . 030m .> .o .m ~28
£82 . . I . . I 32°: .0 58
I282 3538 2 32.2833 855 I | 3 I . m? .88 S». 83
855 I I an ”an .2“ 88 8» .. . I 3E“: .< .3 .38 ~28
€82 3538 8 858883 I .. 233m .4 .m .m 888
5883.22 20 .8 £5588 22 as; E 88..» 33 -~ 5 8» 8mm . .
886 8% 2w P~ 8_ E on.“ . I I 1395 @5302 I0 «36
.92 E 2:35 E~ .883 25 2o E In: an 8C . . . H magmm £55 320
I .885” 9:: 30 ~86 8n n :3 3.20
.8388 oz 8} 23 8~ .I I. one 3m 8m 88 : I =5: .0
—o>o\~ 008q :5
u—uP—am .533 590n— \.~a0 .5550 .02
since—0M van uOfiSU< .«O 0.532 oflflfix< @005q BO—om
5 so;
.525 Sam vouch.
G=eIH<OZmZm2 IbMOZl~<QOmI6
 €82 353.8 8 3303033 35.5 I I I I
€a 3538 8 333333 35.5 com :2 . . . . .BA 333$ :83
I I I I 8n = Z . . . .Eq 336m 385
.tonou 3535 2 bacuscumnnm 3:75 I I I $30
5.8%.. 3528 8 3205333 3:75 I m3 :2 . . . . .65 333m #85 $30
I I a!“ conucoo 62 .vom ooh o2; . . . . . 323mm ouofiflofl
€092 3538 8 3.55333 35.5 I I o2 . . . . m : 330
.toqfi 353.5 8 3285333 3:10 «2 com; . . . . . 3.53 .< .0 830
4..q I I can . . 94 6% dun vow coo; . . . . . 65 558.6
REES 2 33333.; 35.5 I I I I So «ammo
4.6%.h :2 _ . . . . 65 =95320 mmuuo
=2:q B 33333:.“ 35.5 I I mm” mwv .Nov :3 m3
.toafl REES 8 b33333 35.5 I I 0% ooo.~ . . . 03:30 .0 £2 VmNNU
. . m5 mow cow . , . . 323mm. Buss—on— «END
.moEEMMmoZ Sod nvmd S . . . . M. mam o2. . 35mm 32989 «ozm .m .2 $80
um .83.; .235 £33 ozifisn 5 233m 2.? 83 m2 : . . w: an
6:23 can 28 .633 ovod
8: : 33:38 8.3%. 80825 2.88
vmwfi con . . . . o2 3% Se”— . 2383.89 mogfimm 205200
.338: 05:95:...— E 93%; came nmwd new . . . 820
.28 ..3«BI£=oE€om 8:5 8m Sm SC . . . 32:95.? was 2.602 3.20
wwmd $3. in . . Se .oG .mmm 2e 8m.— . . . . . goo—m .4 v.3 .<
.03.»:25 3.95m .535 “oz 926 coax. 03 : NS com 63 :30
A... 2&3 9:5c «>2 2.2%.: 03:28: «no 85% .. 52m .A “a“ .< $20
53$ 35% mm . . . . .03 SN con N SEE“: 9.50 33
.Edd $m Egg—vow 8:3 3—6 ouww 8m . . omm o2
cum 220
.Edd m$~Hm .2303. moo 8m.— . $3329 zoom 4 320
0:29—23 E E Sumac .Em .2329: 0:20:93
30% £5828 83 6:” .Scofiwvum 83 J2 ,
«Ex. wohm o2 . . can 65 62 oum com”. . . 333m— o>2mcoov 2.35:3 .M
£5863 8—3 max mow m v2 . . . . 820
.8388 02 .565— 32 cm. 2». man u . AnaEammomv 823mm 806200 $20
036 coma oov . . . . .. .mov nmn o3..— . . Begum 832030 .
.8t :2 BO $06 03% 3N m 3 m3. Nov mom 820
.333 03.. :Bon 2m 80 _ . . Aspfimiomv mfifiwm BoEEoQ m2 6
54

E9... 2.338 oE—ononn wuunuwnom «85.6:

Eon—Eon Eu. .633 55 v5 8&5» was 20 cmod on; com . . .EN 03 8A" . . 9:375 558m “520 33 530
.535“ “oz I I. 2m . . . poov M? a? to." . .
.mEoEEom 9—3 036 . . . . muffin .2 .3 $8
com m owv . m2. mm» n3. mm“. B— _ com—apex .2
do: .533 .22 .833 2&0
ImaaoEEom “EN ME: 5 8:23..a ecoESow .2 0:.» ~36 2v : . . com .2».
.0528 one ooo.~ . 31.8: .2 .m 350
can 28 535 .53: 03.38: E ban—am 053E Rte unwfi who .. .. o: .mS
, .3556»... 3.3 Rs 95; .. .. v.85 .4 23 .< 9.8
026 Saw w: . . . . .mw. 02 8». 33:30 >65 08t .m
6:23 was “on .833 .3383 o: .535— 82 5&6 Ema omn . . . . can o3 030
.23 222:. 2238.3 2,8“ 35563 o9 8o; . . . . x85 .4 Ba .< "mvo
£5 83 can an . . . . mm 9% RE . 998339 ”swam 80835
.a>E 03.383 96% EEEEom 8—2 5 3—6 2a 3m . _ . . new con
add
.3353 02 .565— 32 m2”— . . . . . “—005 .A can .< 030
08K in 0 RN . own 0% cow 2 . . 2—0585 .3 .3 END
.EEBQEaSU
530868 8:3 >26 ooww an . 3m mam 83 .m .< .m 33m >65 923020 EU
.flaofimuum 8:3 .Emconfiaao
~26 ~86 9% r of. 8v; .m .< .m 38m 305 939020 m8
334 32.5w .3
3q0¢ 25 Hobac< mo 0.332 3322 @85q 320m #035 33>? finen— EMO .5550
. 6.4
#53 .53— E 30; W .
uBmoh ,_
.
<H~EZMEAHIfi
 nflwIOOmlmnom .Md .0
aha 0:33:
M .3a Setafi 96% E3 @3832 .zaueeom cs.» 23 8v . : o9 v8 83 : 82% 2685 22E 220
I I w: . So 58 as o; . . Stem «£83m 22E $20
_ I .Edd a Twanmwfium
82:, 32:3E 26% muqrmmfioww .wcwwwmom 8% 56 2a . .. we. a: mm 83 : : .8225 a .3 3:0
8203 “soEEE 26% £3 3.832 .maofiuom an» 23 E . .. 8» “3 So; .. caste Stemm— BS $20
.33 3203
EOEEE 5 8.2855 $323832 .nEoEEow n3.» ”:6 SN . . . 82mm 9:. 83 . . . 559:3q :8, 2:0
83. I I I 8» =2. .. 2.5m .m .0 $80
«3 o8 n . .. as £532 36
55

I. I. 2N . : n:
68:3 2.2 20 3v w SN w o2 . .. SN 8». on» . . :25 mo £5o 23 38
. H934 ooatam .E
I £99 a .6850 .02
_ 315.com ucu homaa< mo 833A 33:? woo—60m 320m gushm .533
I 22>
5 coach
495A HEM .
:HMOZIOKOHZI6
 13

time as the volcanic rocks referred to above. They are often scoriacious and fissile.
Nepheline has never been identified in thin sections of these lavas. There is a
considerable resemblance to the faulted Nakuru phonolitic trachytes which cover
the floor of the Rift Valley south of Nakuru.
Samples from boreholes C1837 and C2006 show that a succession of flows of
phonolitic trachyte which underlie the Menengai volcanics on the south-west edge
of the Rongai plain to the west of the volcano are underlain by flows of aphanitic
phonolite with abundant interstitial nepheline. The paucity of surface exposures make
it an impossibility to determine what the relationships of these lavas are to the remainder
of the faulted lavas in the area.

CHAPTER V—THE MENENGAI VOLCANO

The Menengai volcanics differ from the older volcanic rocks in that they can be
related to a single central volcano, Menengai, which lies immediately north of Lake
Nakuru. Menengai is the largest volcanic caldera in Kenya, the crater, roughly
trapezoidal in shape, covering more than 35 square miles and having an E.N.E.—W.N.W.
long axis of six to seven miles. It is a subsidence caldera, complicated by renewed
eruptions of lava, almost entirely confined within the caldera.
Physical Features
Menengai has a gently rounded outer profile, smooth grassy slopes rising at first
steeply from the surrounding country, but easing off to a very gentle slope near the
crater rim. This type of profile, contrasting as it does with the steepening profile
of most volcanic cones, is in itself suggestive that the central section has been dropped
by subsidence. The caldera walls are formed of abrupt cliffs, exposing tiers of successive
lava flows. The height of the walls is very variable in different sections of the periphery
of the caldera. In the north-west section the wall is completely absent, while on the
south side it is over 1,000 feet high.
The crater is almost completely waterless; only one permanent stream, the Crater
Stream, flows into it, to vanish underground on encountering the fissured and broken
flows of lava covering the floor. A few damp patches occur between the bounding
wall and the terminal cliffs of recent flows that have been arrested by contact with
the wall. These are due to small accumulations of surface water in the'hollows so
formed. The fauna in the crater is very limited, the only permanent inhabitants being
small antelopes and hyraxes. Though reputedly inhabited by numerous snakes, not
one was seen within the crater by the writer. ‘
Geological History of the Volcano
The First Phase
Eruption commenced with the effusion of lava from a vent situated approximately
in the centre of the present caldera. Of this vent no vestige now remains. The lava
was extruded over a considerable period with intermittent pauses in activity, represented
by brick-red coloured old land surface deposits exposed in the wall of the caldera.
The nature of the eruption was at first quiet, pryoclastic material being completely
absent from lower flows exposed in the walls. The process is envisaged as a welling
out of lava from the vent; the flows extending further and further from the vent
as the volcano built itself up above the surrounding country. The uppermost flow
of the series extends to a distance of more than six miles from the centre of the
caldera, whilst the lower flows appear to have terminated much nearer the vent.
The earliest flows are composed of porphyritic trachyte with a spherulitic micro-
texture. The spherulitic clusters of felspars are commonly interspersed with small
cavities, devoid of any secondary infilling. These lavas as normally black in colour,
but a brown coloration is seen in the vicinity of old land surfaces separating the
flows. Subsequently more glassy lava was poured out and the topmost flow exposed in
the wall consists of grey trachytic glass, passing up into green devitrified and scoriacious
 14

glass. This uppermost flow has the appearance of a single lava flow but shows a streaky
banding and inclusions of pyroclastic material, and there is one well-marked horizon
containing boulders of trachytic lava in a devitrified glassy matrix; the boulders are
angular and some exceed four feet in diameter. The uppermost flow is thought to be
a welded tuff or trachyte ignimbrite. The uppermost horizon of all consists of green
crystal trachytic tuff composed of clear felspar crystals and small pyroclastic fragments
in a glassy matrix. This tuff is known locally as “building stone”. It varies cOn-
siderably in appearance and texture, forming discontinuous outcrops on the outer
slopes. On the south and east slopes coarse dominantly crystal tutfs outcrop, while on
the northern slopes finer green glassy ash overlies the uppermost ignimbrites.
In March, 1957, workings in a quarry on the Eldama Ravine road, one mile
west of Nakuru, exposed moulds of trees, within the body of the uppermost ignimbrite
flow of Menengai. These moulds (Plate 1) have the form of cylinders, of a foot
diameter and about six feet in height. They taper upwards. Only two vertically situated
moulds were seen, but there were many more in the floor of the quarry, lying hori-
zontally with a parallel arrangement away from the centre of Menengai. Some of these
horizontal moulds are very small (branches?) and a bifurcation suggesting the branch-
ing of a tree was seen in one case. The vertical cylinders show a crude spiral ribbing
pattern. The ignimbrite is discoloured and oxidised, where in contact with the moulds,
which contain a little loose gravel but no charcoal. The banding of the ignimbrite
is bent upwards around the tree moulds. The absence of any charcoal is explained by
the theory that the trees were completely vaporized by the enveloping mass of incan-
descent ash and gas, which rolled down the mountain side in a heavy cloud or nue’e
ardente. The cloud coalesced into a mass of very hot glass which subsequently moved
a little way down the slope, in the manner of a lava flow, thus producing flow structures.
It is noteworthy that the trees are situated near the termination of the flow where
coalescing would be advanced and movement sluggish. The preservation of these
tree moulds in situ is strong evidence for an origin in nuéer ardentes for these welded
tufl‘s; a normal trachyte lava near its termination would almost certainly uproot the
trees and would leave some remnant in the form of charcoal.
Within the coarse glassy tufis on the eastern slopes are included small fragments
of aegirine syenite, a coarse holo-crystalline rock thought to be derived from a central
intrusive plug of syenite underlying the original vent. This may be compared with the
well-known syenite plugs of Mount Kenya and Mawenzi. Many boulders of this
syenite are scattered on the slopes of the volcano above Menenhill Farm. This syenite
is completely devoid of nepheline, compatible with the fact that nepheline has not
been identified in any of the Menengai lavas. It is probable that syenite represents
the hyperbyssal magma from which the trachyte effusives were derived. Similar syenite
boulders are found on Longonot (verbal communication, Mr. R. G. Dodson).
The condition of the volcano at the termination of the first phase of eruption
is envisaged as a cone, with a diameter of about 12 miles; the slopes were smooth,
rising to a comparatively small central crater between 8,000 and 9,000 feet above
sea level.
The Formation of the Caldera
Calderas are defined as “extensive craters out of all proportion to the size of the
volcano”. Two mechanisms are generally accepted for their formation: (i) catastrophic
evisceration by explosion; (ii) piecemeal foundering of the superstructure of the cone
into the underlying magma chamber. The fundamental difference between the two
processes is that in the first type of caldera the material from the obliterated super-
structure is scattered in the form of pyroclastic deposits over a large area, outside
the limits of the caldera, whilst in the second process the bulk of the lost material
is engulfed within the limits of the caldera. A caldera that shows extensive evidence
of inward foundering is considered to be a subsidence caldera even though explosive
eruptions played a large part in its formation. Subsidence calderas are characterized
 15

by large deposits of airborne pumice around the volcano, but little or no fragments
recognizable as being derived from the original superstructure. Menengai is a caldera
showing such features. Subsidence was undoubtedly the dominant mechanism, though
this subsidence appears to have been the direct result of explosive eruptions.
The sequence of events in the formation of subsidence caldera is illustrated
in Fig. 2 which is based on the work of van Bemrnelen and Howell Williams and
appears in Physical Geology by A. Holmes, 1944. The magma chamber underlying
the central vent is supposed to be stoped out until it has a much greater diameter
than the vent; the level of the magma can be dropped, either as a result of explosive
eruptions rapidly ejecting a large part of it through the central vent, or through the
opening of a parasitic vent on the lower slopes of the volcano, tapping the central
magma reservoir. The sudden, recession of the magma level leaves the superstructure
of the cone wholly unsupported, fractures develop, and blocks of the superstructure
founder int-o the depths, where they are probably to a certain degree assimilated. There
is no evidence of any parasitic vent near Menengai, and there is evidence that the
formation of the caldera coincided with a period of explosive eruption scattering
pyroclastic material over a wide area surrounding the volcano.

r,
A

,f; 0/)
(r33) (“J
“in
L

Fig. 2.—Subsideuce ealderas. The mechanism of formation. (After A. Holmes, van Bemmelen
and B. Williams.)
 16

This evidence of explosive eruptions, initiating the foundering process is seen

in the increased pyroclastic content, and the highly scoracious nature of the ignimbrite
that. forms the uppermost layer of the original volcano. This ignimbrite is sharply
truncated by the present-day bounding fractures that circumscribe the caldera, but
it is probably contemporaneous with the actual explosions that started the subsidence.
It is indeed possible that sections of the cone nearer to the vent had already com—
menced to founder simultaneously with the eruption of these lavas and pyroclasts, but
the major part of the caldera formation post-dated the deposition of these ignimbrites.
In the surrounding country the explosive eruptions which initiated the caldera are
thought to be represented by the extensive deposits of tufl and ash which overlie the
older-faulted lavas. For example at:—

Makalia . . .. .. Makalia Ash.

Njoro . . . . . . W
Rongai . . .. . . . . Black Ash.
Elburgon . . . .
Solai . . .. .. .. Solai Tufi.
Mbaruk .. .. .. Building Stone.
These deposits vary greatly in colour and texture but all are trachyte tufts. with
abundant felspar crystal fragments. They can easily be cut into rectangular blocks
for building purposes. All these formations outcrop in the bottoms of the present-
dayrvalleys, possibly having been washed in by rainwater. They are thus of subse-
quent age to the formation of the present drainage pattern. These tuffs often contain
rounded fragments of green pumice, and of black obsidian with prominent felspar
insets, a rather unusual lava type also observed in the highest glassy trachyte flow
of Menengai caldera wall. These tuffs are afl’ected by minor faults believed to be
coincident with the subsidence movements which formed the caldera. In a subse-
quent section dealing with the sedimentary formations it will be seen that there
is evidence that these tutfs are about 10,000 years old. It is likely, therefore, that the
caldera of Menengai was formed somewhat less than 10,000 years ago.

That subsidence has taken place cannot seriously be doubted. Subsidence structures
are preserved at several points along the rim. These structures are characterized by a
reversal of the normal outward dip of the lava flows in the wall, a marked tilt inwards
having been effected. Where only partially developed the structures are monoclinal,
but the more strongly developed structures are bounded by steep or vertical fault
dislocations. For example, on the south wall, between the Menengai Forest Reserve
and the caldera, a block of the wall is preserved in a partially foundered state. This
block is bounded on either side by monoclinal flexures, but these pass into a fault
dislocation in the centre section, duplicating the caldera wall over a distance of about
a mile. The tongue-like projection of the wall on the western margin shows all the
features typical of these structures. The outward slope of the lavas in the wall is
reversed by a gentle monoclinal flexure which passes inwards into several abrupt
dislocations hading very steeply towards the crater, and throwing down the peninsular
segment of the wall 300 feet into the crater. This is illustrated in Figs. 3 and 4, drawn
from photographs, which clearly show the marked tilt of the subsided block towards
the centre of the crater.

Subsidence in the north-west sector has dropped the rim of the crater to such
an extent that later lava flows erupted within the crater have piled up higher than
the actual rim, and the terminal cliff of the later lava flow rises 100 feet above
the rim, the caldera wall being completely obscured by this later flow. This structure
and the two trough-like structures near Rhodora Farm and Prentice’s Farm appear
to have developed along earlier structural lines disrupting the underlying faulted lavas
over which the volcano erupted. In the Rhodora structure a central flap, bounded
 .5... E33? 2: E9: to; £8.20 Eanofiznld Eh
.355 rad: .0 030...
arrxo<uk rnn<aa oz< war
2.2542. 0:32:10;
dur<du ax». agar—go auto-3 35,3 :35 .53.". 32:
Jahao 3:. or so «axon udaan: 10.: .83.
«ac—nadpsu (>14 .3004; €39.30 3 395.3: 292:3», £2313: 3.3.2... _
n4002>n
30—2: aur(n_.—OoxO¥5
r... —<4da>o or 3:251“; r3440
manhuadpnunfifi o 20:30 nunu<$ou
7:; 30.:
3.594.: 523... .35
52.1.5; 5; |<dun.:u so x: zlurnnil
.r ..¢|..:.......
17

\\ nx<ann<qo =3 u>oa<
.00. 3:»... :30 < 10.: ‘30:
.l/ pzauadurzana< 4.13 4.50.2.»
_ :...h:.;....2..3 ,
~ \.
lll I = .\ ll .
, I
£2327?...§=§..§3§§:
.
.rl. r
33:3 9.3 So Go 3:,
3.. :5: S .5 .5 ya 355...: 5:2
95:»..22. u:_..>..nu2io._. .2; 528.. ._ .
.3553:
.3233 3552.: A: v :51: :2...
xgflaflawufi: 1:18 3.0.: 3.333..
C5 “193.5... 2:433». at: 2...; :3;
.25..
3:35: 523.3: .5: .2 u A 3. $3.”
.2 5.5
wars... 9.2
33°: 2...;
:5:
5.3.: 5.52 5:: .5»:
SE... 3:033:33 , n 323 it.»
59.3 :5...» a 931.:
5.; 2:5 .25: c.3295 n.333
.o .5523 25:5: 5.3.3» :3. 3::
.53 u.r:.u
.3015 12.3.5 uZVnacu 5.2:...-
 18

PARTLY' FOUNDEEED 556l OF

SUPERSTRl/EE. TIL TED ””1480:-
MP GLAfiY T£ACIIYTE 1(0o
“N7211- omuam av Pumas. sour” mu.
emu-2 ca”: .
J ;
«awe
umomumrta

Fig. 4a.—Sketch of subsidence structure on western side of caldera.

STEEF DELDG‘TIONS
HAD/N6 TOWARDS CALDiBA

Wl M
‘1
"li
l iii
r H
y“
[I

Fig. 4b.—Sketch showing the reversal of the normal outward slope of lava flows in a
promontory projecting from the western wall of Menengai caldera, consequent
on the foundering of the central part of the volcano.
(Erratum: for top lava read trachyte ignimbritc.)

by faults has dropped inwards along a hinge-like monoclinal flexure, parallel to the
caldera wall. The curved facets in the crater rim are revealed most strikingly on
the map (Map No. 3) and are in themselves a strong indication of the piecemeal
nature of the subsidence process.

Secondary Eruptions Contemporary and Subsequent to the Caldera Formation

Though these secondary eruptions are certainly to some extent later in date
than the Elmenteita Basalt eruptions they are for convenience described here together
with the earlier Menengai Volcano eruptions.