ReportSNoFflS

GOVERNMENT OF KENYA

MINISTRY OF COMMERCE AND INDUSTRY / GEOLOGICAL SURVEY OF KENYA

GEOLOGY oF‘ THE IKUTHA AREA

DEGREE SHEET 60, NW. QUARTER

(with coloured geological map) by J. WALSH, B.Sc. V Geologist UNWERSETK ()1: NAIROBI GEOLU-SY DEPARTMENT LAEC‘RATCRY REFERENCE BOOK HGT TO BE REMOVED!

Ten Shillings - 1963 ~ l GEOLOGY ‘ OF THE IKUTHA AREA

DEGREE SHEET 60, N.W. QUARTER

(with coloured geological map) byby. J. WALSH, B.Sc. Geologist FOREWORDFOREWORD

TheThe IthukaIthuka areawis area-is oneone'that ,that isis littlelittle visitedvisited asas aa whole;who1e;forfor thoughthough the the railwayrailway andand thethe Mombasa-NairobiMombasa-Narobi roadroad passpass throughthrough its its south-westsouth-west corner,corner, muchmuch ofof itit isis notnot easilyeasily accessibleaccessible except except onoa ffdotjotz PartPart ofof itit lies,lies, however, however, inin thethe north-westnortfi-west cornercorner ofof thethe TsavoTsavo RoyalRoyal NationalNational Park, Park, andand maymay become become betterbetter knOWn known inin the the future. future.

TheThe areaarea isis noteworthynoteworthy inin soso farfar asas itit containscontains aa graphitegraphite minemine thatthat has has operated operated forfor aa numbernumber ofof years,years, andand has has produced produced somesome 3,0003,000 tonstons ofof thethe mineral. mineral. AA generalgeneral accountaccount ofof thethe graphitegraphite deposits deposits ofof thisthis partpart ofof KenyaKenya waswas givengiven inin aa previousprevious reportreport (No.(No. 37,37, 1957) 1957) onon thethe SouthSouth Kitui Kitui area,area, area, whichwhich adjoinsadjoins thethe Ithuka Ithuka area area onon thethe north. north. Mr.Mr. Walsh, Walsh, in in thethe the presentpresent present report,report, report, continues continues continues the thethe account, account, account, givinggiving detailsdetails ofof thethe processprocess thatthat isis usedused forfor extraction,extraction, andand a a proposedproposed processprocess including including winnowing,winnowing, which which hashas beenbeen suggestedsuggested inin viewview ofof thethe acuteacute shortage shortage ofof waterwater in in thethe minemine area. area. TheThe area area isis crossedcrossed byby aa partpart ofof thethe YattaYatta plateau,plateau, whichwhich isis oneone ofof thethe uniqueunique featuresfeatures ofof KenyaKenya geology.geology. MuchMuch argumentargument hashas centred centred roundround the the emplacementemdacement ofof itsits thinthin capping capping ofof phonolite,lava.phonolite,lava. Mr.Mr. WalshWalsh adheresadheres torto thethe valley-valley-flowflow hypothesishypothesis andand providesprovides anan interestinginteresting calculation calculation toto showshow thatthat suchsuch aa lavalava couldcould havehave flflowedowed thethe lengthlength ofof the,the plateauplateau inin aa periodperiod ofof aboutabout sevenseven months.months.

Nairobi,Nairobi, WILLIAMWILLIAM PULFREY,PULFREY, . , 30th30th January,January, 1959.1959. ’' Ag.Ag. CommissionerCommissioner (Mines(Mines && Geolbgy)Geology)

Imus}, ”“ G“0Lo~ EOF NAIROBI LABORATORY PAR TMEN T T EFISREENCE 500K CONTENTS PAGE.PAGE Abstract > ' ’3 *mz,\Mg <*.+(“u-x.” .. a I-IntroductionI—Introduction and General Information ...... 1 ..< . c 11-~reVi"ousII—Prexiiou‘s Geological 165:1:irk ...... 3 111-PhysiographyIII—Physiographs' ...... 4

IV-SummaqIV—Summary _ofof Geology ...... 10 V—Details 1110;016n .. .. €- 11 11.. ~abmentBitumen! System ...... 11 (1) MetamorphasedMetamorphosed psa-ticpsammitic &limed&sediments . . . 13 (2) Metamorphosed semi-peliticscmi—pelitic sediments . . . . 15 (3) Metamorphosed pelitic sediments ...... 18 (4) M-orphdMetamorphosed cal-calcareous sediments . . . 19 (5) M~gmatitegMigmatites . . . 20 (6) Intrus~vesIntrusivas ~nto'iheinto the ~&meniBasement system'System . : . . 21 22. . Tertiary-YattaTertiary—Yatta Plateau PhonolikPhonolite ...... 23 33.. Pleistocene and Recent ...... 23 (1)(1)1.avas Lavas ...... 23 (2) Alluvial deposits ...... 24 (3) SuwcialSuperficial deposits ...... 24 VI-MetamorphismVI—Metamorphism and Granitization ...... 25 VII—Structure ...... 1115, 25». VIII—Mineral Deposits _29 (1) Graphite ...... 29 (2) Limestone ...... 34 (3) Vermiculite ...... 35 (4) siiteSillimanite ...... 35 (5) Water-supplies ...... 35

IX—Referenoes 36

LIST OF ILLUSTRATIONS

FigFig.. 1.—Physiographical1.- Physiographical map ... . -.. .1411-.-...-...... 11:- 4...... - A ...... 5 FigFig.. Z-Structural2.—Structural map ...... 26 FigFig. . 3.--Contoured3.—Contoured stereographic diagrams of poles to foliation planes and of linea- tions ...... 27 FigFig.4. 4.-Frequencyx—Frequency diagPamdiagram of joints ...... 29 FigFig.. 5..-Graphite5.—Graphite flow.sheet,flow-sheet, Ganesh Claims, Kamstoi ...... 31 FigFig.. 6.-Experimental6.—Experimental graphite flow.sheet,flow-sheet, Ganesh Claims, Kamstoi . . . .. 32

MAPS

Geological map of the Ikutha Area (degree sheetsheet 60.60, N.W.NW. quarter)quarter).. Scale 1 : 125.125,000000 at end 3:1!:“13892 A ' 5.) bin. miteuixmni «J

' magma swivm‘i.~ “it

qdmpizw‘l—wl ii ABSTRACT m) 3070 wr~nrrti?,.-mggrtmg..-:y1VI The report describes an area of approximately 1,200 squa Machakos Distriats 'omedj%;r'b6~iby parallels 2" and 2" and 38"38° 3(Y30’ E. Physiographically the area is divided into fifiveve uni"a drainage,drainage, of gently undulating countrycountry in the west and sosouth-es Plateau, (3) the fairly fairly deeply dissected catchment areaarea of the Tiv fir Yatta Plateau,Plateau, (4) the Yamala-Kimathena-IthumbaYamala-Kimathena-Ithumba rangerange, and ( * ,, plains in the east,east, grading to thethe .end-Tertiary.end-Tertiaryend-Tertiary ,peneplain, with erosmn; residual;. The solid rocks rocks of the areaarea fall into threethree groupsgroups (1) BasementBasend System {3‘-@&- \ cambrian), consisting mainlymainly of para-gneisses with minormino intrusions, all-«strongly W and metamorphosed,metamorphosed, (2) Tertiary volcanics, represented by a single extensive W , lava flflow,ow, and (3) Recent volcanics,volcanics, comprisingcomprising two flofl floows of olivine‘m M . _. deposits of Pleistocene and Recent age include redred and black soils, Sioondarydoondary lame-l&+ monesstones and alluviumalluvium. , 1 The petrography tot,of the various rock rock typestypes is described, and thethe structure 11}?— morphism and granitization of the Basement System rocks are disemdisc&&E?~~~~~dfimm" -T ‘ Occurrences of economic minerals are described, with specialspecial attentm‘wwattenthsir* which is being workedworked in the area.area. . tc,:?.J!

«A

ww- . x “.ugfia

, 1... -r“ I-INTRODUCTIONI—INTRODUCTION AND 1 GENERAL INFORMATION . ‘ General.-TheGeneral.—-The area described in this report is is approximately lr,2001',200 square miles in extent, and is boundedbounded byby'parallels parallels 2"2° and 2"2° 30'30’ S. and meridians 38" and 38"38° 30'30’ E.E It comprises the north-west quarter of degree area No. 60 60 (Kenya), (Kenya), and is the area covered by the flai#carratauctuate of Overseas Surveys Sheet No.No. 175. It lies inin the Southern Province, the partpart north north and east of the Athi river being administered administered from Kitui and the remainderremainder from Machakos. More thaq;c&e-thirdthawohe—third of sthethe area, in the east and south- east, lies in in the Tsavo Royal National Paik,Park, and of the remainder, most most of the land land north-west of the National Park Park formsforms part of the Kamba Native Land Unit, the rest being CrownCrown Land. There are three blocks of land in the south—westsouth-west leased leased from the the Crown, two two of which are planted with with sisal, sisal, by by Dwa Plantations Ltd.Ltd. near Kibwezi Kibwezi (LR 917),,and917), and by Masongaleni Sisal Sisal Estate Ltd.Ltd; at Masongaleni (LR 5903). The third block, JLR5LR 3683'at3685) at Manoni, produces sugar-~nesugarcane and paw-paw under a a system system of irrigafionirrigation ~hkhwhkh taps the Kibwezi riverriver and a larg

The Athi valley, whichwhich has a perennial perennial river and is in many placesplaces bordered bordered by fertile alluvial flats,flats,flats, is devoid of settlement except except at the c~ossingcrossing of the Kibwezi-Kitui road. Various Various explanations were given to hehe writer for this—abundancethis-abunbce of big game, liability to flooding,flooding, and seasonal infestati6, by mosquitoes mosquitoes and tsetse Ilks.flies. The latter would appear to be the true reason, sincesince the lo&local" Wakamba have little fear'fear" ‘ofof elephant and rhinoceros, and serious floodingflooding appears to be aa rare occurrence. 6 . The ~fricadAfrican Land Development Board Board has instituted the Athi-Tiva Reclamation Scheme, whereby thousands of acres of bush country between the two riversrivers have been cleared by tractors andand bull-dozersbulldozers and,and, by a planned and controlled system of burning, rich pastures pasturespastures, freebefrom tsetse flyfly have been developed for stock grazing.grazing. Many earth dams have been built and wells sunk for watering.watering. It is hoped that by levying levying a small annual grazing charge on each animalanimal the scheme will be made self-supporting. Grazing will bebe strictly strictly controlled controlled to prevent soil erosion, whichwhich is a serious serious problem problem in the extreme north of the the area.area. There There tracts tracts of of landland many many square miles in extent extent have been denuded of grass cover by cattle,cattle, goatsgoats and sheep, which has led to-to, serious gullykggullying and the loss ofof’ much of the topsoil.top—soil. Serious efforts efforts to prevent fufiherfurther erosion and to reclaim bdlost grazing land are are being made by compulsory purchaseofpurchase of cattle to keep numbers down to a safe figure,figure, and“by‘“scratch‘-plonghing”,and 'W6kratch'-phghing", the cutting of shallow furrows acrossacress the gullied traotstracts at intervals of a few feet to trap rainwaterrain-water run-off and to stirnulatestimulate the growth of new grass.grass. The gullies themselves are dammedmedat short intervals with with rocks and woven sticks.sticks. Such a ploughedploughed area isis usually regenerated regenerated to a suficientsuflicient extent to allow lightlight grazing grazing again again within two or three years. Game is abundant in the less settled regions regions and in the Tsavo Royal Royal National Park, particularly elephant and rhinoceros. In the National Park, particularly nearnear the Tiva river,;theriver,,the number of elephant is so great that vast areas have been flattenedfl flattenedattened and . trampled and the ground isis litteredlittered with fallen trees. BuffdoBuffalo are common in the thick bush of the Yatta Plateau. In the more open grasslandsgrasslands around IthumbaIthumba there are large herds of zebra, gazelle and impala, withwith lesser numbers of giragiraffe,fle, oryx, wildwild pig,pig, wart-hog and ostrich. 2

Climate and Vegetation.-RainfallVegetation—Rainfall statistics, which are recorded only at Ikutha.lkutha, Dwa and Masongaleni are tabulated below, together with readings made at Kanziku. two miles outside the northern border of the area, east-north-east of Ikutha.lkutha.

‘ Total Rainfall (inches) No. of No. of Locality Rainy Yearly Years i 1954 i 1955 l 1956 Days, 1956 Average 'Reoorded I I1 ‘ . I Ikutha . . .. 3 17-67 1‘ a 23-51 32 18.2718-27 5 Dwa 1:. I' 21.5321-53 14-19 27.7127-71 52 24-30 37 ~Uon~aIkhiMasongaleni . . 1| 20.4820-48 . t1 t1‘ 93-9723-97 47 Kanzik~Kanziku ... } 11-22 ,._~16,-a4-46.84 . . -25.78.25-78 51 23-1123-71 14 .. I I p- ‘Records*Records incomplete. ?NotTNot recorad.recorded.

The rainfall is bi-annual, with a marked maximum in November-December, and a lesser peak in April.

The main agricultural crops are maize, beans and cassava, with castor-oilcastor—oil beans as a cash-crop. A small amount of tobacco is grown for local consumption and in a few favourable locations there are plantations of bananas and plantains. Two crops are planted each year, to coincide with the rainfall maxima, but it is not uncommon for rains (and crops) to fail completely, as they did inid early 1956, wbenwhen the rainfall for March to May was confinedconfined to,@to.-:-a few short showers. In such times the Wakamba must rely to a great extent on their livestock fqrfor food, supplemented by wild honey.honey It is common to find,find, even in apparently trackless bush, a native beehive in almost every large tree. Much of the area is covered by acacia thorn Scrubscrub with occasional baobab trees, and where game tracks are few penetration is a slow and didifficultflicult business. Natural water is almost completely lacking over most of thethe area, since only the Athi and Kibwezi rivers are perennial, though water can be found along the whole length of the Tiva river only a foot or two below the surface. All the springs mapped are very small with the exception of that in the Kibwezi lava near Manoni, which has a measured flowflow of 25,000 gallons per day. The water-holes mapped are either shallow pans, dry for most of the year, or hollows and crevices in rock platforms which are regularly cleaned and enlarged to provide domestic water. The wells at Kitubua, in the north- east, are dug through sandy soil down to the water-table a foot above bed-rock, to a depth of up to 25 ft.,ft, and ramps for cattle are made down to the water. There are here upwards of 20 wells in an area dof about one-sixthonssixth of a square milemile, each belong- ing to a particular family. As already;already-mentionedmentioned a large number of earth damsdims and two concrete dams have been constructed,constrtlcted, many of which hold water for-thefor.the grerttrgreater part of the year. Bore-holes have been sunk on the Dwa and Masongaleni sisal estates, and a public bore-hole is in operation at Ndovoini, north of Dwa.

In the past few years only, small-holdings have been cleared on the Yatta Plateau at Kabete and Mwala, and the red soil derived&rived from the underlying phonolite has proved particularly fertilefertile. However, since tbethe soil is nowhere above two feet in thick- ness and frequently less than one foot, such small-holdings.small-holdings are?re pa~rf$ularlyparticularly dependent on a good and well-spaced rainfall. +-

Communications.—TheCornn~~micarions.-~henications.-The main Nairobi-Mombasa road and the East African railway cross the south-west cornercomer of the area, and two railway stations lie within it, Masonga- 3 leni and Kikumbuliu. The latter is a new station built as part of a scheme to increase the traffic-carryingtraffic-carrying capacity of the single-track line by providing a passing point for . trains travelling between Masongaleni and Kibwezi. The main road fgamJawezifibwezi to Kitui Kitui passes through Ikutha,Ikutha, crossing the Athi river by a low concrete bridge and running across the sandy bed of the Tiva.Tiva. It follows the route of an old footpa*footpath across the Yatta -MU,Phidu, where a fault has given rise to a col some 200 ft. below the plateau level. There is saidsaid to be a Stone-age livingliving site near this colcol but neither ”thethe writer nor Dr. B. N. Temperley of the MinistryMinistry of Works were able to to findfind it. From IkuthaIkutha aa fairlyfairly good road runsruns east and north-east to Kanziku, with motor- able tracks braichingbranching south to to Kasaala and Kimathena. An excellent road hashas been made throtlghthrough thelTsavothetavo Royal National Park northwards from Voi to meet the latter track at Kimathena.~Kimathena.=: From just south-west of IkutbaIku-tha a new motor-track runs north-north- west to Kalebu and Kitui Kitui as an access road for the Athi-Tiva Reclamation Scheme. Two roads reach the Athi river in the south-west, south-west, the most northerly, primarilyprimarily a fifire-re- break for the Dwa sisal estate, reaching the river at Masalani, and the otherother aa few miles to the the south-east running running to Kithiulu from Manoni, which is reached reached from eithereither Kibwezi, via Dwa, or Masongaleni. South A.37 Maps.-Two maps covering the area have been published, sheet sheet Maps.—-Two maps covering the area have been published, shay—$70 on a scale of l:250,000,1 :250,000, dated 1912, and Voi (E.A.F. No. 1714) on a scale of 1:500,000,1 :500,000, dated 1946. The latter mapmap appearsappears toto be based largely on the 1912 sheet which, while good in parts, containscontains glaring inaccuraciesinaccuracies which have been perpetuated on the 1946 sheet. In view of the unreliability of these maps the wholewhole of the topographytopography for the geological map was taken from air photographs made by the R.A.F. in 1948, controlled by a plane-table survey based on surveyed trigonometrimltrigonometrical stations. The form-lines are based on corrected spotspot heights obtained by a single aneroid-barometeraneroid barometer and must there- fore be considered as only approximate.approximate. The area toto the north was mapppdma- in 1952 by E. P. Saggerson of the Mines and Geological Department (Saggerson, 1957)‘1957)* since whichwhich time a new triangulation survey has been made and several new stations fifixedxed near the commoncommon border,border, enabling points onon the border to be fifixedxed with greatergreakr accuracy. It will be seenseen that some commoncommon points on the border differ by up toto half a mile on the two maps.

ACKNOWLEDGEMENTS

Thanks are due to the Administrative OfficersOfiicers of the Kitui area for their assistance during the survey, survey, and to the Manager and staff of Dwa Plantations Ltd. for help and hospitality.

. .--- - .-. , -, lI—PREVIOUSII--PREVIOUS GEbhOGIéALmGEOLOGICAL wormWORK The fifirstrst travellertraveller who left aa record of his journeyings in the area was Ludwig Krapf (Richards, 1950, pp. 54-83)54-83) who passed through thethe north-western part of the area in 1849, but mademade no mention of the geology. In 1877, J. M. Hildebrandt (1879, pp.pp. 334-342) followed the Athi river for several miles north-westnorth-west from Mikomani (possibly Ngomano) and crossed the Yatta Plateau to reach the TivaTiva somewhere west of Ikutha. He mentions variousvarious soilsoil types, and of 3ethe Yatta PlateauPlateau he says “It''It seems to be an outoutflowflow (auslaufer) of the snow-mountainsnow-mountain Kenya".enya".

‘ References are quoted on page 36. 4

3.'I. W:W. Gregory (1896, pp. 76-76-79)79) in 1893 followedfollow$ the route of thetbe' presentaaypresent-day railway between Masongaleni and =%we@;Kr‘bwezi, widewhere he,stajedhe stayed ,at at the East AfAfridfican RottiahScottish Mission whose station was situated near Wd'bcli,Md’kock,' .s4at the ageedge of the KibwGKibwezi lava;lav‘a;to to which he refers, commenting that its $911.ace1sq$aps fpwegfeatures hdjqte111(1t that it must be of recent

age.age He a@nagain visited East Africa .inin 1912,19%. and1n.+&' .">"; uent book (Gregory, 1921, pp..p._p 184-191) discussed the Yatta PlateaqPlateaa 11111111111;*&!HBylll15 1 1111111)u)- volcanoes. 1 . f, $. ! &..i&*, Captain L. AylmeiAylnier 1nin 1907 1907 trave$d,;frv.k&iqtravelledtfr Wife}! ipz3' utha, where he began a traverse dowddown the Tiva which took him eastwhdseastwards $511113ybnd e'elifnits ''. -ts of the present area (Aylmer,(Aylmer,1908). 1908). The geology is tersely summed up.up in the sentence "Quartz“Quartz aboundedaboundcd everywhere.everywhere but no shale formation was seenwu.whi&.~,ye,pr..ofwhich gaye 91mm of coal."coal. ” . Brief repor6reports on the geologygew,wewere ,1 . E E. W pp. 4-5) and H. B. Muff.Muff (Maufe).'(Maufe) J908,1908, p.p 21). 21).21) Both mention gneisses, and 1111:q 1e 1019113: of the Yatta Plateau and KibpeziKiWezi valley. . :... . ~.*. ',. ' .:~ .. .* . , .. '.t <.& E.E Krenkel~renkelvisited the area in 1908'$iid!$r1908 anti '1her (Krenkel,(Krenkel 1911,p192 lf2.'.2!57).beefly2‘57)brie fly yeferredperened to crystalline schists, strongly folded onqn N:-S.N.—S. axes, with vpla~~@~s~t'ti"~.~.mamas 11111.11. In'a'la't~In a Iate; book (Krenkel, 1925, p. 241) he iden-identi fied the Yatta ~late&jBd'~::~~~~&Plateah‘;$1: an "&1 Kapiti phonolites, probably quoting Gregory, and his map xX~XXI &1%1'0111111 the(mm. broad.broad detail. 1

.,,;J. : $ritzFritz Behrend'sBehrends map (Behrend, 1918) shows the area as crystallinecrystalhe sihists'overlqinschists overIsin by younger volcanics of the Chyulus and the Yatta Plateau, the latter being 511% an iwedisolated s&t.sheet. 1 fame ._ 1 ._ The existence of grapygraphite in the area was firstfirst recorded ib111 1940, and a &.cDgood ' of prospecting has beenbeen- doned6ne since then.then This is discussed at greater length 011‘ori. i . .114.“ 1%:~* Much of,the area to the south was mapped by 3. Parkinson (1947), but Much of the area to the south was mapped by J Parkinson (1947), but than“.. . gap of a few.few miles between the limits of the two areas.areas 11151101,.!- -J1 d B.B N.N Temperley (1956, Vol.V01. II,11, pp.pp 64-122) studied the lavas of the KibKibw Masongaleni.Masongaleni valleys as part of a wider survey concerned with water-supplies..wateI-supplies {“111

E. P. SaggersbtiSaggerson‘ (1957) mapped the adjoining area to the north, and thkjmefifi differences between the nomenclature of rock types used for.hisfor’his map and that~bfthat bf 1 present map. Saggersan'sSaggerson’s group Xh'Xh’ (biotite-hornblende(biotiteharnblende gneisses including hornbl&hornblende granulites) has been subdivided min the presentpreserrt -areaarea into Xh'Xh’ (hornblende-bi&(hornblende-biotite gneisses) and Xhh (hornblende gneisses), the latter symbol being being used used for rocks in which biotitebiotk does Mt'not ex&exceed 2020- pefper cent dof .thethe total hornblende-biotite conttdt,content. addition SaggersonsSaggersbn's group XsX: wdtic(psammitic bandsbad in.M'k:kemi-peliticu‘ semi-pelitic series) is reprreé fl by Xggn (quartzo-felspathic(quartzo-f-elspathic gneisses. .and granulites) and in one cyecase by Xn gneisses). 1 .1 . .1 ) . t9.1, .r ,. : it; 1;. .. . . ~ . ... - if; *.‘ * .. . . . III-PWSIaGRApHYIII-PHYSIOGRAPI-IY ,- .m1'12 x The most noticeable feature of the.pm.the are: is the Yatta Plateau, which is roerolfliyv paralleled by thethe, Athi river on .thethe westw,&,aridand thethe Tiva river on the east, each of which has its own distinctivedistinctive drainagedraina$e2.s5 *' To'To the east of t4lethe Tiva system i5is a gag: sloping bush-coveredbush-covered plain with p~&inedpo defined drainagedrainage, broken by a major rawmingle;w $ his and other isolated hills and ridges. .d , hills and other isolated hills and ridges . 1 , 1- _, 1.5%, 11 . .:.,, .+*+ : 2lI: The Yatta Plateau 1sis capped by lava which is considered to have flflowedowed cd~dM‘dh the sub-Miocenesub-Miocene peneplain (Schoemann, 1948, p. 3), and is here taken as the &&&datum level for the erosionerbsion surfaces 1nin the area (Fig. I),1), there being no fossil evidence of the .-. Knolotwa ' M5

‘ .lthuvpbl ‘ ‘ 3405

u’sa'e}

''' , Wmnma‘éud-Ctmuflm Macon.moco (magnumns " og . -L Janus mlromiSOOtoMOOfeotOD) , ‘e ' ' , . - - Shannen» 6W ' -—.- - 1...... “ . ’ y Wind fiee‘ieneldqp’e .(rumfmn m IMMOD)

l-+%ydographiudsketch-map (De area *. ' .. &$'fig. 1—,”msketeh-map of the lkntbnllmtlm area , ,, age ofat the,the w~-MvLio~sub-Miocene wor any other of the surfaces. The spot-heights marked on Fig.Fig.1 1 ,werewere $gadmud on the presmt-daypresent-day surface of tbethe lava, which maintains a fairly con- stant thicknesaethickpea offi aboutabaut 30 ft.ft.-_The. The plateau thus falls from about 3,100 ft. in thethe north- west to about 2,59025590, ft. in‘thein the south-east,southeast, a3 steadystudy fall of 12 ft. perpenmile, mile, excludingexcluding'the the mtionportion of the plateau inip the north-west, which is ,tilted tilted by post-Miocene faulting;dfaujting FR~ locallylocally; slopes dmdown toto the fi;8orth.orth, FurtherFurther remnants ofof» the sub-Miocenesub-Miocene, suffice are represeatedrepgesehted by weU4efiqedwellbdefiged ishelvesshelves on the hillshi& of the Ngunumu-MbatuNgunumu—Mbatu ridge in the c~n!nfrecgntre of the area,areafand and in the south-west by the isolated hills of Chae,Chane, Dwa, KafimaKalima 6

Koii and Muliluni, which, while high in relation to the Yatta Plateau (perhaps a result of post-Miocene tilting), show a good accord of summit levels. Poorly marked shelves and ridges on Kimathena Kimathena andand neighbouring hills hills may also markmark remnants remnants of this surface, but this could not be determined with any degree of certainty. East of the Tiva drainage system the ground slopes gently to the east-south-east, and the extreme north-east is considered to be part of the end-Tertiary peneplain, which is well developed over much of Kenya. It lies between about 1,700 and 1,900 ft. here, and gradesgrades down to a base-level of erosion of 400 ft. in the .KiliKi-Mazerasfi-Mazeras area (Caswell, 1956, p. 6). This surface is still developing by cutting back to the westward, across country with a gradient of about 25 ft. per mile. In the southeastsouth-east an average gradient of 55 ft. per mile along the line of the Tiva river is an indication that the TimTiva is still cutting down to this WA11:190e few miksmiles further east the Tiva opens out into a flat, swampy bed in the end-Tertiary peneplain proper (Sanders, 1963). No erosion surface older than the sub-Miocene was proved to be present, though both Kimathena and Ithumba, and possibly other hills on the ridge continuing north- wards from Kimathena, probably mark muchdegradedmuch-degraded remnants of the endCretaceousend-Cretaceous peneplain, which would be expected to have lain more than'than a thousand feet higher than the sub-Miocene surface here, i.e. aboveabove" 4,000 ft. at Kimathena. At three points points on the eastern eastern slope slope of the YamYatta Plateau between Ikugi and Wathoni shelves up to 150 ft. wide were noted, at 2,260 ft. at Ikugi, 2,345 ft. just north of Wathoni and at 2,230 ft. at W&oai.Wathoni. However, no corresponding bevels were found on the western slope, and it seems probable that these shelves are due to differentialdifierential rates of erosion in an area of rapidly alternating rock types rather than than remnants of old erosion surfaces. Residual hills are not uncommon, particularlypalticularly on the end-Tertiary surface and on the pediment between it and the sub-Miocene surface. They take the form of tors and inselbergs, and usually mark the outcrop of steeply-dipping stmta.strata. Although the rock types in such hills vary from granitoid gneisses to hornblende gneisses and migmatites they are always of a leucocratic, and therefore fairly competent, facies.facie. The whale-back rock of Katolotwa, seven miles north-west of Kimathena, is an excellent example of a miniature inselberg. It is oval in plan, about 1,200 ft. by 600 fL,ft., elongated along the strike, its summit being over 200 ft. above the surrounding plain. Its slopes are convex, but vertical or near vertical at the foot, where the ground is littered with exfoliation debris resulting from "onion-skin”"onion-skin" weathering, the splitting-offsplitting-OE of successive sheets of rock parallel to its outer surface, with little regard to the foliation of the rock. The finalfinal stage of such weathering which still remains visible above the soil cover of the plains is a flat flatflat rock rock platform, bare of vegetation except where stunted bushes havehave gained a aroot-hold root-hold in pintsjoints and crevices. Such platforms are seen at Chae and Siulungu, parts of a discontinuous elongatedelongated outcrop seven miles south-east of Kimathena.Kimathena; ,I The Yatta Plateau has long been an object of interest to geologists. ,It is a lava- capped feature beginning at 01 Doinyo Sabuk, near Thika, as a broad plateau, soon narrowing to an average width of one to two miles and running south-south-east and finallyfinally east to a point east of Tsavo, a total length of about 180180.miles. miles. Gregory (1921, pp. 184- 184-184-189)189) 189) summed up the salient facts as followsz—thefollows: -the lavalava“ caps the summit of a long, high ridge, and nowhere shows evidence of having overovertlawedfloWed the ridge or flowedflawed over the plain at the foot of the ridge:.ridgerv‘there there is no evidence that the lava flowflow ever covered an appreciably wider area thanthan that that of the present-daypr~nt;da). plateau:platau: abethe lava surface falls gradually along itsits whole whole length, and never ataf'a a steepersaper gradient than 30 ft. in a mile. He made errors in some minor details, e.g. "Its“Its width E. of Masonga-Mamma- leni is betweenbetWeen three and four miles . . ."..”. LaterLater work has disclosed that pebbles of the lava are found up to 15 miles from the plateau, across a watershed,watershed, in the area west of Kitui township (Schoemann, 1948, p. 6) and pebbles have also been found at 7

Kinyiki hill, south-west of the present area, and 13 miles from the plaieauplateau and separated from it by the Athi river. East of Tsavo (Sanders, 1%3)1963) records a a gradieotgradient of 35 ft. per mile. Some Some variations of gradient also occur due to post-Miocene faulting, as in the north-w&w%f-thenorth-wett'pafi‘hf‘the present area. Gregory concludedconcluded that the lavalava flowedflowed down aa valley, subsequent erosion having wovworn. down the the flanking flankingflanking gneisses at a much greater rate than the the lava itself, which prpteEtedprptiected the underlying rocks rocks to leave the plateau as it is today, in this area up to 1,000 ft. higher than the bed of the TivaTiva and up to 600 ft. higher -than~than the Athi. Dodson (1953, pp. 4-5) has put forward forward several arguments against this hypothesis. He writes that it it seems highly improbable that a single lava flowflow could continue to flowflow over such a great distance, distance, especially since thethe thickness of-the flowflow seldom exceeds 50 ft., andand’that that the the characteristics of the lava suggestsuggestwthat .$at it must have been fairly viscous. He adds that nowhere in the south-east Machakos ar&area were underlying underlying valley deposits found. He therefore advances the alternative theodtheory‘ that the lava lava was extruded alongalong fault-line fissuresfissures extending dongalong its whole length and that the surface flowflow nowhere spread for a distance greater than 500 yards from its its point of extrusion. One obvious weak point in his argument is Bishis tacit acceptance of the original thickness of the lava as 50 ft. HeI-Ie accepts the fact that the lavalava was extruded on thethe subMiocenesub-Miocene peneplain, endand his map shows dissection in the gneisses to a depth of about 600 ft. below this surface. Much Much of this erosion is due to pediplanation, and the main process of erosion of the pbonolitephonolite must be by Wallateral spalling, but surface surface erosion of the lava lava is taking place, proved by altitudes taken across the plateau which showed the present surface of the lava to be as much as 15 ft. . higher in the centre than on the flanks.flanks. It must therefore be concluded concluded that the the flowflow might well have been twice its present thickness. ' The writer made extensive traverses along andand across across thethe summit of the plateau, and looked for basal contacts at 38 different localities.loealities. The base of the lava lava could seldom be fixedfixed with any degree of accuraqaccuracy owing to the heavy bush and thick soil cover and the large large amountamount of lava scree that masks theleslopes. Nothing suggesting valley deposits was found, and often it appeared &atthat the lavaEava rests onon. the the rodred soil of the original peneplain,peneplain, which is indistinguishableindistinguishable from more recent recent soils. (The phmo-phono- lite itself weathers to a bright red soil on the plateau, not to blackbkxk soil such as so often occurs on phonolites phonolites inin Kenya.)Kenya) Fairly reliable fixesfixes of the lavalava base gave thicknesses varying between 10 and 55 ft, but thethe average is about 30 ft. The width of the plateau (excluding spurs) variesvaries from two and a half miles to three-quarters of a mile, and at two places the lava is breached along fault-lines. Both of these gaps andand the area between the plateau and the large large outlier of 'NdoyaniNdoyani in the centre centre of the area were carefullycarefufly traversed in aa searchsearch for feeder dykesdykes as envisaged by by Dodson. In the two fault gaps the country rock isis greatlygreatly obscuredobscured by scree and huge boulders fallenfallen fmmfrom the summit, but south south of Ndoyani lava flfloatoat is almost completely lacking over much of the country,cowtry, and clearly no dyke existsexists here. Similarly no dykegdykeslof of phonolites wereWere found flflankinganking the plateau, and it seems certain that none exist, since by the evidence of the the rock of the plateau itself such aa‘dyke dyke woultl-hmwould-form a prominent feature. The S-bend at Kabaa is difficultdifficult to explain on the the dyke hypothesis, but is a very very character-character; istic riverriver feature.feature. Frequently the large phenocrysts of the phonolite show a marked preferred align- medtment over fairly large areas. Sixty-fiveSixty-five such alignments were recorded and almost with; out exception they lie parallel or sub-parallel to the margins of the lava at those points. It hiis considered that thethe phenocrystphaocryst alignments here indicate the direction of flowflow o£of the lava. TheThe lava is visibly vesicular,vesiyu;lar, and must have had a highhigh gas content when molten,molten; and would therefore have been very fluid.fl fluid.uid. Gregory considered that local .vesicularvesicular structure may have have been due to steam generated as the lava flowedflowed over “rampss$vwps and lakes, and if thethe volume of water was not enough to have “frozen""frozen" the lava the steam steam caught up may have have increased its fluidity.fluidity. 8

Dodson’sDodson's main point, that “It"It seemsseems highly improbableimprobable that that a single:sinde lava-lav~llowflow could continuecontinue toto flflowow forfor a distance of about 170 miles”miles" will bear further investigation,in~e9tigatbo. In much ofsf thethe present area, thethe gradient of the surface on which thethe lavalava flfl0q.dowed is nowww as lowlow as 12 ft.£t.per mile, .butbut itit has been proved from bore-hole loglogs kept by the,the Ministry of Works thatthat thethe Kapiti phonolite,phonolibe, the parent mass of thethe YattaYath phonolite,phon&e. . has been tiltedtilted since itsits deposition, and it is clear thatthat thethe Wholewhole sub-Miocene surface must have been tilted toto thethe south-east. The overall fall fromfrom the northernmostnorthermt point of thethe flflowow tota itsits end isis 3,3,000000 ft. over 180 miles,&$, a gradient of 16.619.6 ft. per mile, but thethe gradient of thethe mature peneplain would have been only about half thatthat fifigure,gure, or eight feetfeet per mile. Kempe’sKempe's EngineersEnginahs YearYezc Book (1954, Vol. 1, p. 146) gives a formula for thethe flflowow of water inin openopeti channels.6hanntls. V=V = CVRC S,wherewhere V isis velocityvekity inin feet per second, C isis a commco~ttsuf&fi&pS end toto be 100) R is thethe ratio of area of cross-sectioncross-sectiart of the channel to thethe Wetted»%kt& periméter@&hdter of the channel inin thatthat cross-section, and.mcf S thethe slope of thealie channelchannel. Considering€!ofisidfri&'& '2' ihowBpw of water down thethe valley inIn which thethe Yatta lava flowed,dowed, assumingasshfng a depth'depe ofef'76060 ft,#.; andknd a V-shaped channel several milesdl+ wide (since thethe width of6f thetBe charchariacl'fiiel isf$ very largele~gecomparedcompad to thethe depth thethe wetted perimeter is virttlalfyvim* the samewhte as the&A width):aiblh);'&'K bunniesk~es60/26012 = 30, and S isis 8/5280,815280, and V isis calculatedcalctdated as 21.3 11g!ff, second}dec&id!* “Thus ael We!*t flft6wow comparable inin shape with what thethe lavalava isi$ thought «'126 lk!i? Va&! hadhid originaliymt~rjghl&f~'Wd.t.ak~flddake only 12.4 hours30ats toto 'covercover thethe 180 miles.lhaes. IfIf, forfor thethe purpose of argument;ar-c 86th8W.w-wales and moltenmo4f& lavalava , .a:. . , &!*, far both are considered as idealideal Newtonian liquids, liquids, in .thethe formule,vforrq&,, 9 _, ., liquidsliquids g (gravity)(gravity) h (depth)(depth) andaad sinasinor (gradient)(gradient) remain& 4fie:the £111&&,:;&i*tht'1 as that thethe.aws'twe’ d density . T..' velocities vary according toto g- = ---.density The density of the congealedibn&&d lava Was*as 11q viscosityvlsCos1ty foundf~undto be 2.2-49,49-, indicating a maximumarraximum fifieregure for the moltenm lava Of}.4, No yisdoslty fifiguresgures of phonolites are available, but DalyDab (1933, p. 72) quotes measured:viscgs “T flflowingowing Hawaiian (basaltic) lavas as low as 11 times that of Water. lavasIavas have a relatively low viscosity is indicated by fifigures quoted in; the 19", _ f: Physical Constants (Geol. Soc. America, Special Paper NoNo. 36, 1942, pp. 134—135)134-135)-

poise:j7Oi~eS : ' . Laboratory melts of Japanese lavasIavas » (at 1200" 1(1)4.), Andesine basalt, Motomoura ...... 31,200.31,206 , Olivine basalt,basaft, Gembudo ...... 3,180 Olivine basalt, Konoura ...... 732 Nepheline basalt, Nagahama ...... 190 (The viscosity of water at about 20°20" C. is .O10.1 poises)poises), Since theSethese are laboratorylabomtoyy~melts'melts of congealed lavas it is certain that most or all of the volatilevolatiIe flfluxesuxes had been lost, andahd the true viscosity of the original lav must have been many times lower than thOSethose quoted. TheThe large phenocrystspbenocrysts of nepht". linehqe andhdanortho- anartho- clasectase in the Yatta phonolite clearlyclehrly famed before extrusion, .butbut bothb?th would MstalliSef$sta!lise at above 1100“1100" C,C., so that the aboveahve'&uresfigures can be taken as some guideguae totethethe viscosity of the lavalava. Assuming that the original viscosity of the Yatta lava was of the order of 1,000lwtimes that of water, if the same viscosity were maintained throughout the flflowow . and assuming the lava to be an ideal Newtonian liquid of densityden& 24,2.4, the flflowow would have required 2149214.9 days to cover the 180 miles. The movement of lava flflowows at Panchtin&~&tinhas been described by Krauskopf (1948, pp. 1267-1283). Soon after extrusion the surface of a ?a:?lava ,hcarcjpshardens to scoriaceous blocks which are carried along with the flflow,ow, building marginalmarnnal iiihraines%raines which dam the lava to a great extent, and also forward over the snout to be overrun by the still molten lava. Krauskopf likens the mechanism to that of a caterpillar tractor. Tempera- tures in the lava fall only very slowly along~along its length, and a lava comparable with the - Yatta phonolite above 100 ft. thick (before congealing the thickness in places may have 9 been severalSeveral times this fifigure,gure, as the depthdepth of lava Wouldwould fall£all as the flflowow moved on) would hold its heatheat by its very bulk, aided by the insulatingInsulating effect&ect of the cooled blocksblacks on its surface and latent heat of crystallization given alloff by individual minerals as the lava congealed.cohgealed.

The actual time takentaken for emplacement of the flflowow maymay have been several times the estimate for an ideal flflow.ow. There is nothing inin the microscopic detail of the rock to suggest that a timetime of eveneven seven or eight hundred hundred days would bebe tOotoo great.groat. The groundmass is fifinene but hole-crystalline,hulo-crystalline, and never glassy glassy as it would be if very quickly chilled.

Undoubted flflowsows of lava of lengths roughly comparable with that of the Yatta Plateau have been described.described, BarthBad (1952, p. 148) quotes a lava flflowow along a water- grade in the westernwtofthewetitera+& of &heGrand CanyonGanyon which fl&wedowed for 135 km. with a drop of only 210 m.—am.-a gradient of little more than eight feet per mile. It therefore «emsaeem certainstriain that Gregory’sGregory's hypothesishypothKsi$ is the truetrue one,om, and&ad that the YattaYatts Plateau marks the course of an old valley.vrdky. In the pmentpr-t areaadea the S-hcndS-bend in the plateauplafeau is very characteristicchtacfsri#ti&of a& rivertire feature,feltllrs, thOttheugh the reasonfeP8oh forfot the double bend is not now clear. The fault thatth8f runsmSnorth-northeastfiotth-ilorlh-eMt fromftom Kabaa may have ininfluencedfluenced the coursecoutsc of thetlw river,riydfI thoughtMi@ it Wasum notnet traced toa0 the south—Westsouth-west of the bend. GregoryOregoo [allowedfollowd hisbls theoryCheury to theWe logical conclusionconclasien that the emplacementempldeement of the lavalava must have diVerteddiverted the original riverrivtr into Notwo rivet-s,fiV~fs,one'onOW ~fieach side of the flflow,ow, nownow representedrepranted by the present-day Athi and Tiva rivers.dvets. Sanders (1963) has prevedproved that thethe' the Tiva formerlyformerly paralleledparalleled the the Yatta PlateauPlateau far to the southsouth of the sharp easterly bend near Wathoni, which he attributes to river capture. The Athi and Tiya rivers show very dissimilar characteristics.chatacteristics. The Athi,Afhi, which has a perennial~erennialflAow ow compared to the very intermittent Surface6urface flflowow of the Tiva, shows few of the features of a mature rivérriver system.system. LongLong stretches are perfectly straight, and shallow rapids are not uncommon. Its average gradient is fairly low, however, about 14 ft. perper mile in the present area,area, and the river isis frequentlyfrequently bordered bordered by alluvium which stretches up to 500 ft.ft. fromfrom either bank, but isis generally generally less.less. ItsIts locallocal catch-catch- ment area isis gently undulating country, and even itsits majormajor tributariesiributasies seldom Cutcut down toto solid rock.rock. The Masongaleni riverriver inin factfact meanders sharplysharply forfor Severalseveral miles. InIn contrast thethe Tiva’sTiva's catchment isis fairlyfairly deeply dissected, and thethe river’sriver's course isis frequentlyfrequently determined by steep boundingbounding rockrock fades,faces, usually usually steeply steeply dipping ridgesridges of quartzo—felspathicquartzo-fekpathic or granitoidgrani.toid gneisses.gneissn. ItIt has nevertheless cut a flfloodood plain ab0utabout a mile wide across which itit meanders. The gradient of thethe flfloodood plain is 21 ft.ft. per mile, but thethe meandering reduces thethe gradient gradient of of thethe riverriver bed toto 124}12+ ft.ft. per mile. TheTRe eastward portion of thethe riverriver fromfrom thethe bend of capture near Wathoni isis not taken intointo account inin theseltgese calculations since, as mentioned earlier, the gradientgradient shows a marked steepening herehere toto 55 ft.ft. per mile. = The dissection of thethe TiVaTiva catchment area isis a resultresnlt of thethe riverfiriverlter capture. Altitudes readread alongdong thethe old course of thethe TivaTiva southsonth of Wathoni showshm thattMt thethe capturecaptwe ledled toto a cutting-downcutfing-down ofuf thethe river-bedriver-bed toto 200 ft.ft. below itsits original level.level. The marked meanders meanders of thethe riverrivet are an indicationindic&tioathatthat thethe riVer'river itselfitself isis is approaching grade, but the shortness of itsits tributariesttibutaties and thethe lowlow rainfallrainfall have notnut yet allowedallowed themthem toto cut dowudown thethe catch-catch- ment area toto keep pace with the_fl‘iva’sthe Tiva's Tiva's downcutting. The fifinalnal episodeepissde inin thethe river’sriver's history isis an accelerated down-cutting,dowipcutting, which has incisedincised thethe rivertivet bed toto a8 depthdeptli of 15 fL,ft, sweeping awayaway thethe 'alluviumalluvium which thethe flood-plainfl flood-plainood-plain must must have contained. North-west of Ikuthalkutha twotwo cut-OEcutaff meandersmeandee can still'be’st~Wr&agnizkdrecognized whose beds are now above the Tiva’sTiva's bed, showing thatthat theythey were were cut off before the meanders completedoompleted theirtheir incision.incision. The reasonreason forfor thisthis lastlast dawncuttingd6wncutting was not found, butbut isis probablyprobaMy a resultresult of Pleistocene rejuvenation. - .- 10

The third river river system of the area is the poorly poorly defineddefined drainage of the end- Tertiary peneplain and the pediment pediment above above it in the north-east. north-east. The rivers here, here, the largest of which are the Mavuko and Kalobato,Kalobnto, are choked with with sand and flowflow only after prolonged rains, and then only for a few hours. They They rarely rarely show rock exposures, exposures, and the nature of the ~~n$t~s~ggests€9,9RUXfi9SgCStS that down-cutting is not keeping pace with the weathering of the unaerlyingunderlying rocks. These rivers, rivers, like like the the Tiva, Tiva, drain towards the the central reachesreaches of the Tana, but it it is not knownrif-theyknowngiffithey reach that river or are are lost in the plains plains before before then. *‘7‘ "‘5 ‘ '

“IV—SUMMARY,IV-SUMMARY OF GEOLOGY The solid rocks underlying the whole area are ofof, t@ethe Basement System, considered as Archaean in age, withwith. very minor intrusions. For the most part they are covered with superficialsuperficial deposits pf PleistocenePleistocene-and and Recent age. Two.seriesTWo=series of lavas are present, one of Miocene and.and the @erother of Recent age. * - Basement System.-TheSystem.——The Basement System rocks consist of a succession of para- gneisses, mainly hornblendic hornblendic in character, with with biotite biotite appearing appearing as an important constituent towards the east. Relatively minor intercalations intercalations in this series are diopsidic, quartzo-felspathic and granitoid gneisses, with crystalline limestones, quartzites, and one small small band band of sillimanite gneiss. All are strongly metamorphosed metamorphosed and meta- somatized, and contain no no fossils by which they can can be dated. They They are assigned to the Basement System from their their close resemblance resemblance to rocks of that system in in many many other parts of Kenya, some of which have have been dated by chemical means and proved to be Precambrian in age. The granitoid gneisses and quartzo-felspathic gneisses, both of which consist of little more more than quartz quartz and and felspar, form the most conspicuous features of the area, other than the the Yatta Plateau. They They are usually steeply steeply dipping and form long sharp ridges,ridges. The granitoid gneisses which form the hiphesthighest point in the area, the ridge ridge of Kimathentt,Kimathena, are tightly folded into the almost almost vertical vertical core core of a syncline which has remained as an erosion residual. The hornblendic gneisses are' aré""generally generally poorly exposed owing ,toto their ease of erosion, and the most maficmafic facies ~fof these gneisses, the plagio- clase amphibolites, generally form negative features. The crystalline limestones, while while sometimes forming distinctdistinct positive features, are often planed planed off to the level of the surrounding gneisses and, where superficial superficialsuperficial covercover is thick, are often impossible to trace for any distance along along thethe strike. strike. They They usually carry small small flakes flakesflakes of graphite, and locally includeinclude or overlie workable depositsdeposits of graphite. Intrusives into the Basement System rocks are few and small. They include a dyke of pyroxenite, a plug plug of olivine norite and a few small pegmatites consisting of quartz, felsparfeISpar and mica. The Basement System rocks are all of sedimentary origin. The hornblende-biotite gneisses are thought to to represent a difference of composition in the original original sediments,sediments, perhaps an abundance of chlorite, as compared with those which gave rise to the horn- horn- blende gneisses, gneisses, rather than a differencedifierence in in metamorphic'grade.metamorphi’c‘grade. The granitoid and quartzo-felspathic gneisses are considered to be of sedimentary origin,origin, and originally were sandy strata and lenses in rather muddy sediments. sediments. Similarly Similarly the crystallinecrystalline lime- stones and diopsidic gneisses reflectreflect local increases in the calcareous content content of the original sediments. The plagioclase amphibolites probably developed from fine-grainedfine-grained mudstones with a low low sand sand content. content. They They are always conformable with the strike strike of the neighbouring rocks. Tertiary Volcanics.—~TheVo1canics.-The lava capping capping the Yatta Plateau is of the type that covers the Kapiti Plains Plains southsouth south and south-west of Nairobi, where it is known as Kapitian Phonolite. It has a fifine-grainedne-grained groundmass, but carries carries largelarge phenocrysts of anortho- clase and nepheline. It is considered considered to be Miocene Miocene in age, since it lies on on the sub- Miocene peneplain.penepIain. Over most of the Kapiti Kapiti PIainsPlains the phonolite phonolite weathersweathers toto a a black black 11

clay soil, but here here the the soil, soil, where it does not exceed two feet in depth, is is freefree from from waterlogging and weathers bright bright red. Pleistocene and Recent.-MuchRecent.--Much of the solid solid rockrock of the area is mantled withwith red- brown sandy sandy soils, Jpilainlv.-x,qidualmainly-residual in character and and derived by weathering and mechanical breakdown breakdown of the underlying rocks. Where drainage is poor black-cotton soils have developed. The Athi river has a 11amnav,bndband of sandy alluvium along most of its length, sometimes interstratifiedinterstratified with thinthin' 'greygrey layers of hefine silt deposited during flfloods.oods. Most of the larger rivers havehave sandy beds, but none have alluvial flats.flats. Small deposits of surface limestone (kunkar) have developed developed on and near most of the limestones and over the more maficmafic hornblende rocks. Nodules of kunkar areare also developed at anlland near the surface surface of the black-cotton'soils.black-cotton‘soils. Two tbd@estongues bf olivine basalt invade invade the southiwestsouth-west comercorner of the area along the valleys of the KibweziKibweki and Masongaleni rivers, having their origin in the Chyulu hills. They havehave been intensivelyintensively studied by Temperley (1956, Vol. II,11, pp. 64-122), who has has recognizedreoognized three superimposed flflows,ows, all of olivine basalt, in the part of the Kibwezi Kibwezi valley in this this area. area. There is no evidence of any considerable lapse of time between successive flflows.ows. They havehave aa coarse, coarse, blocky and sometimes ropyr-opy surface, surface, andand though they are well wooded all the trees and bushes are rooted in soil which has been blown or washed into crevices in the lava, and the upper surface is virtually unweatheredunweathered and free from soil. These are characteristics of lavas of a recentrecent origin,origin, and the age of the Kibwezi and Masongaleni lavas is probably in the order of 1,000 years.

In the northern part of the area the succession appears to be:—be:- 3. Yatta Plateau Phonolite. Average Average thickness 30 ft. Long periodperiod ifbf erosion. 2. HornblendeHomblende gneisses withwith granitoid and ,quartzo-felspathic gneissesgneisses, and minor intercalations of crystalline limestones, biotite -->gneisses{gneisses and plagioclase . . amphibolites. Present thickness about 70,000 ft. .—. ‘ l.1. Homblende—biotiteHornblende-biotite and hornblendehornblehde gneisses with granitoid and quartzo—quartzo-felspathicf—elspathic gneisses, plagioclase amphibolitesarnpbibolita and well—developedwell-developed crystalline limestones. Thickness about 10,000 ftft.

V—DETAILSV-DETAILS OF GEOLOGY l.—The1.--The Basement System The rocks of the Basement System System in this area are classifiedclassified as follows:—follows:- (1) MetamorphdMetamorphosed psammitic sediments-sediments— * (a) Quartzites. (b) Granitoid gneisses. (c) Quartzo-felspathic gneisses and granulites.granulites; '‘w ---- -‘ (d)(6) Biotite gneisses. (e) Biotite-garnet gneisses.

(2) Metamorphosed,Metamorphosed

(3) Metamorphosed pelitic sediments-sediments— (a) Plagioclase amphibolites. (b) Sillimanite Sillimanite gneisses. 12

(4) Metamorphosed calcareous sediments-sediments~ (a) Crystalline limestones.

(5) Migmatites. (a) HornblendeHomblende rnigmatites.migmatites. (b) Biotite migmatites. , -

. , IntrusivcsIntrusives into the Basement System comprisecomprise— 'v‘ 3-4’ (a) Pyroxeaites.Pyroxenites. (6)(b) Olivine writes.norites. , , (c) Minor acid acid pegmatites.pegmatitee. "II.

.- -, a .- - ---' ThcThe lowest part of thethe” successtonsuccession that can be determined with any txHahvcertainty ieis the Yamdu-KimathenaYamalu-Kimathena syncline, where it is as folIows:-follows: —— Approximate averéxieaver& rhickrieathickness ' .4 (feet)(full ' Biotite migmatitesrnigmatites ...... :_ ...... - ...... ml%500’ HornblendeHomblende gneisseagneisses ...... 400 ' Hornblende-biotiteHomblende-biotite gneisses, with minor bands of quartzo-felspatbilquartzo-felspathic . gneiss ...... 1,400 ' Granitoid gneisses with thin biotite migmatites ...... 1,200 ' Plagioclase amphibditesamphibolites ...... 1,m1,000 Granitoid gneisses . . . . .- ,...... 300 4300 l

This group is separated from adjoining exposures to the west by a fault which probably downthrows to tbethe west, though this was not proved. The next highesthighmt rocks are the hornblende-biotite gneisses about Ndawemwe, which enclose a sin&single band of crystalline limestone near the base, refolded to outcrop three times, and another near the top, together with a band of hornblendediopsidehornblende/diopside gneiss, and discontinuous bands of granitoid gneiss, qua--felspathicquartzn-felspathic gneiss, plagioclase amphibolite and and biotite migmatite, the broad outlines of the succession being as folIows:-follows:— Approximate average thickness (feet)Veet) Hornblende-biotite gneisses with crystalline limestone and plagioclnseplagiocIase amphibolites ...... - . . ti . . 1,500 Quartzo-felspathic gneisses ...... t..n .. ,.&b, 150 Hornblende-diopside gneisses ' ...... a! ...fi. ; 350 Hornblende-biotite gneisses withbith ‘quartzo-felspathicquartzo-felspathic gneisses . 1,800 Crystalline limestone ...... 400 - - Hornblende-biotiteHomblende~biotite gneisses/granitoidgneisseslgranitoid gneissesgoeisses .. - ...... 1,OOO1,000

:'§' «‘ 5,200

The remainder of the Basement System rocks fall‘ihl'o'onéfall ihPu&em'@oup,fiifoup, characterizedcharacterimd by hornblende gneisses with intercalations of granitoidgranitoId gneiss,gneiss,‘ qmrtzo-felspathicquar-fekptibic new and crystalline limestone, and lesser bands of hornblende-biditehornblendebiotite gneis,gneiss, biotiteKiatite gneiss, horn- blendediopsideblende-diopside gneiss, plagioclase amphibolite, hornblendehornblende migmatitem-te and quartzite. - 13

The present thickness of this group is about 70,000 ft.fi but, as discus&discussed later, the true thickness must be much less, the ‘widthwidth of outcrop beingbung exaggerated by repetition of beds by strike-faultingstrikefaulting and folding. AveragingAveraging out across the whole area of the out- crops, the 70,000 ft.ft.~-W8idW&wnwas“ as follows (not in stratigraphical order):order):-— Approximate thicknesses , =-' (feet) HomblendeHornblende gneissesgneiss& ...... , 59,000 Granitoid gneisses ...... 3,700 Quartzo-felspathicQuartzo-fc$spathic gneissesgneisse~ ...... , 1. .. .. 3,300-3,300 ' Crystalline limeStbneslimktbnes ...... 1,4801AW Horriblende—biotite~ornbiende-biotitegueisses gneisses .. , .. ‘ ...... 740 Biotitegneisses‘Biotite gdeisses' ...... 740 Homblende—diopsideHornblende-diopside gneisses ...... 440 Plagioclase amphibolites ...... 370 Hornblende migmatites ...... 150 Quartzites ...... 80

70,000

(l) METAMORPHOSED PSAMMH'IC Smmms The rocks included in this section had their origin in sandstones and arkoses con- taining onlyonly minorminor amounts of ferruginousfermginous or calcareous minerals. The distinction between granitoid gheisses and quartzo-felspathicquartm-fdspathic gneisses and granulites was made on their aspect in the fifield,eld, the granitoid gneissesgneisscs weatheringweatbering into craggy tors and broken and rugged topography, against the much more even outlines and smooth rock pave- ments typicallytypically developed in the quartzo-felspathicquarbmfelspathic gneissesgnaieses and granulites. Later , examination of microscope sections showed thatthat the granitoid gneisses almost invariably contain a largelarge proportion of microcline, and-whereand where fifieldeld identiidenti6cationfication was doubtful thisthis criterion has been used for classiclassification.fication.

(a)(a) Quartzites True quartzitesquartzite.$ are few'few 1nin thethe area, the felspar content of most ‘ofbf thethe psammitic rocks beingbeing high enough toto put themthem intointo the quartzo-felspathic or granitoid gneiss classes.»clgsm. A true quartzite which occurs at Utundani on thethe Tiva river,rivet, specimen 60/202‘,60/202*, isis a dark grey-brown rock with numerous small cavities coated with brown ironiron oxide. It'has onlypdy a vague foliation, which isis marked by thethe alignment of thethe cavities. In thin section&on itit shows a mosaic of coarse anhedral interlockinginterlocking unstrained quartz grains which enclose numerous small euhedral or subhedral pale pinky-brpinky-bhwnfiwn garnets, and small grains of augite surrounded by and intergr‘ownhwigh‘ppaqueinxagr-own, withopaque ironiron ore. ItIt contains 92 per cent quartzquartz, withwitb 4 per cent garnet, 1 per cent augite and 3 per cent ironiron ore.T0re.t Specimen 60/187‘from601187 from ImiwaImiwa seven miles north-east of Ikutha, is lightlight red in colour, colour. due toto staining by ironiron oxide. It contains 80 per cent quartz and 20 per cent microcline, with ironiron oxide and tinytipy prisms of apatite as accessories.

(b)(6) Granitoid gneisses These are generally coarse-grained rocks,rocks, varying in colour from cream to pink, and often carrying very small amounts of mamaficfic minerals, usually magnetite, hornblende and biotite. Foliation isis usually only poorly marked, and shows itself by alignment of

‘ Numbers preprefixedfixed 60/601 referrefer tofo specimen inin thetheregional regional collectioncollation for Degree Sheet 60 inin thethe Mines andand Geological Department, NairobiNanobt. 1*t All modes quoted areare volumetricvolumeQricand estimated. l4 mamaficfic crystals and sometimes by the flatteningfl flatteningatteningand and elongation of quartz and felspar grains. In thin sectionsection this flflatteningattening is seen to be duedue to recrystallizationrecrystdbttion of quartz, and toto a lesser extent felsparfdspar (usually microcline), individual grains@grains showingshowbg as large, clear blebs with roundedrounded outlines, interlayered with fifiner-grainedner-grained angular aggregates of quartz and felspar, the latter usually cloudy and altered, and all showing fracture and undulose extinction. TheseThese rocksrocks always always contain plagioclase felspar, of the composition of albite-oligoclase or oligoclase, which occurs in thethe older generation of felspars. The re-crystallization of quartz andand felspar isis due toto preferential attack by alkali meta- somatizing fluidsflfluidsuids which have aaffectedflected the original quartz and felspars at a late stagestage in the metamorphicmetamorphic historyhistory of thethe area andand permitted them to re-crystallize in an un.un- strained or only slightly strained condition. Accessories include magnetite, hornblende, hornblende, biotite, muscovite, garnet, epidote,.- sphene- --- and zircon. SomeSome estimated modes are:—are:-

‘ 60/167 ‘ 60/132 ‘ 60/165 | 60/144 60/191 60/139 60/148 % ”A% % % % % % Quartz . . . . 55 40 40 27 25 23 12 Orthoclase ...... —- 32 , —- 24 6 10 28 Microcline ...... 36 —- ' - 40 40 58 60 55 Plagioclase ...... 8 25 18 6 10 7 4 Accessories ... . .,. .. 1 3 2 3 1 + l1

60/167—Dwa601167-Dwa Rock. 60/1601132—Ngunumu. 32-Ngun~mu. ‘-60/165—~Iviani60/165--Iviani Ithi‘eti.Ithieti. 60/144—Kimathena.60/144-Kimathena. 60/60/191-Chamwalanya.l9l—Chamwalanya. 60/139—4-Kimakimwe.60/139-~Kimakiimwe. 60/148—Dudini.60/148-Dudini.

(c)(c) Quartzo-felspathic gneisses andand granulites

These vary inin colour fromfrom cream toto bubuff,fl, or moremore rarelyrarely pink or red,red, and areare., usually of fifinerner grain than thethe granitoid granitoid gneisses, and areare sometimes granulitic. TheyThey usually show good foliation marked by alignment of marnaiicfic accessory minerals.minerals. Inter-Inter- stitial recrystallizationrecrystallization of quartz and felsparsfelspars as described inin thethe granitoid gneisses sometimes occurs, as inin specimen 60/151 fromfrom Ndoyani, Ndoyani, butbut thethe the newer generationgenerxtion of felspars here here are orthoclase, not microcline, onlyonly slightly microperthitic against thethe strongly microperthitic microperthitic orthoclaseorthoclase of thethe older older generation. generation. Plagioclase felsparsfelspars are all of thethe composition of oligoclase. AccessoryAccessory minerals includeinclude hornblende, hornblende, biotite, muscovite ironiron ore and apatite.

Some estimated modesmoded are:—-are: - , 1t t*,*u“ 60/211 60/151 "60/13'6' , % % y % Quartz ...... 50 27 15 Orthoclase ...... » ... 12 65 , 45 , .. .,*,I Plagioclase ...... f . 34 6 35 Accessories ...... 4 “ 2"" ‘V 5

60/211601211 Quartzo-felspathic granulite, Ithangathi.Itbangathi. 60/.151601.151 Quartzo-felspathic gneiss, Ndoyani.Ndoyani. ,r 60/130601 130 Quartzo-felspathic gneiss, Kwambleti. l4 maficmafic crystals and sometimes by the flatteningflatteningvand and dongationelongation of quartz and felspar grains. In thin section section this flatteningflattening is seen to be due to recrystallizationrecrystalkation of quartz, and to to a lesserlesser extent felspar (usually (usually microcline), individual grains showing as large, clear blebs with rounded outlines, interlayered with finer-grainedfiner-grained angular aggregates of quartz and felspar, the latter usually cloudy and altered, and all showing fracture and undulose extinction. These rocks always contain plagioclase felspar, of the composition of albite-oligoclase or oligoclase, which occurs in the older generationgeneration of felspars. The re-crystallization of quartz and felspar is due to to preferential attack by alkali meta- somatizing fluidsfluids which have affected affectedaflected the originaloriginal quartz and felspars at a late stage in the metamorphic metamorphic historyhistory of the area and permitted them to re-crystallize in an un- strained or only slightly strained condition. Accessories include magnetite, hornblende, biotite, muscovite, garnet, .-,epidote, . sphene-__- and zircon. Some estimated modes are:- are:-are:—

i 60/167 ‘ 60/132 i 60/165 60/144 ‘ 60/191 l 60/139 ‘ 60/148 % % % % % % % Quartz . . . . 55 40 40 27 25 23 12 Orthoclase . . . .. —- 32 . --—- 24 6 10 28 Microcline . .A . . 36 -—- ' 40 40 58 60 55 Plagioclase . . . .. 8 25 18 6 10 7 4

Accessories . . . . 1 3 2 3 1l + 1

+ +

+ I 60/167-Dwa60/167—Dwa Rock. 601160/1 32-Ng~n~mu.32—Ngunumu. '- 601165-Iviani60/165—Iviani ZhW.IthiBti. 601160/144—Kimathena. 44-Kimathena. ' 601160/1914amwalanya. 9l—Chamwalanya. 601160/139—LKimakimwe. 39-~Kimakimwe. 60/148-Dudini.60/148—Dudini .

(c) Quartzo-felspathic gneisses and and granulites These vary in colour fromfrom cream to bubuff,fi, or or more rarely pink or red, and are . usually of finerfiner grain thanthan the granitoid gneisses, and are sometimes granulitic. They usually show good foliation marked by alignment of maficmafic accessory minerals. Inter- stitial recrystallization recrystallization of quartz and felspars as described in in the granitoid granitoid gneisses sometimes occurs, as in specimen 60/151 from Ndoyani, but the newer generation of felspars here are orthoclase, not not microcline, microcline, only slightly microperthitic against the strongly microperthitic orthoclase of the older generation. generation. Plagioclase Plagioclase felspars felspars are all of the compositioncomposition of oligoclase. Accessory Accessory minerals include include hornblende, biotite, muscovite iron ore and apatite. t Some estimated mode$modes are:are:-- - _ t te,f 601211 l 60/151 1 60/1i060/156' --60/211 I. % °/o% % Quartz ...... 50 27 15 Orthoclase ...... 12 65 u. 45 Plagioclase ...... > " .' 34 6 35 Accessories ...... i 4 ’ 2*" 5' 5 -

6012160/211 1 Quartzo-felspathic granulite, Ithangathi. 60/.151 Quartzo-felspathic gneiss, ,Ndoyani.Ndoyani. ,— 60160/130 130 Quartzo-felspathic gneiss, Kwambleti. Kwambleti. 15

(d_)B1'otite(d) Biotite gneisses

Biotite gneisses are fewfew and small in outcrop, occurring mainly as thinthin discontinu- Ousous bands NM&&ha..bgmblendefimblende gneisses. A typical example, example, specimen 60/131601131 from Chunguni, 1sis a light butt,buff, medium-grainedmedium-grained-gneissgneiss containing up to 10 per cent of biotite inin fifinene flflakes,akes, whose alignment gives grati?.$&j#rked~foliationfied foliation toto thethe rock. GranitizationGranitization has has causedtaused a small-small-scalescale development ofof-porphyrbblasticporphyr‘oblastic felspars.

(e) Biotite garnet gneissesgneisses

The onlyoqly* occurrence of thisthis rock typetype\hrge:~13rge enough toto be shown on thethe map isis the;1the, isolatedisolated.outcropoutcrop of Nduni, south of IthumbaItbumba, which forms a hill over 100 ftft. high elongatedeIon@ted along the strike. In hand-specimen (60/E49)(601M9) it isis cream inin colour, speckled with black ’biotitebiotite flflakesakes which are aligned inin bands along thethe foliation. In thinthin section thethe biotite is seen toto be dichroic from yellow toto dark brownish-green, and the garnets bright pink, much fractured, and with subhedral outlines. BothBoth orthoclaseorthoclase and plagio- clase felsparsfelspars occur, thethe latterlatter of thethe composition of sodic oligoclase. A second repre- sentative, specimen 60/195,601195, isis a locallocal variationvariation of thethe quartzo-felspathic gneiss at Mkongweni. In hand-specimen iti

Estimated modes of thesethese two two rocks are:—are :-

60/149 60/195 , % 7., Quartz . . . .' . . 32 35 Orthoclase ...... 44 12 ’ Microcline . ,. _ + ’ Plagioclase ...... 17 44 Biotite ...... 5 6 Muscovite _ + Garnet ~ g , 2 3 Iron oreore .. .. " .. + +

60/601 149 Nduni. 60601/ 195 Mkongweni.

(2) METAMORPHOSED SEMI-WC SEDIMENrs

The metamorphosed semi-pelitic sedimentssedimen\ts are divided into four groups, viz.:—-viz.:- (a) Homblende-biotiteHornblende-biotite gneissesgneisses. (b) HomblendeHornblende gneisses. (c) Homblende-garnetHornblende-garnet gneisses. (d) Homblende-diopsideHornblende-diopside gneisses. The divisions tend toto overlap, and to grade towards the plagioclase amphibolites. The distinction between hornblende-biotite gneisses and hornblende gneisses has been made where biotite amounts to 20 per cent of-theof #thetotal hornblende/biotite content. All these rock types are thoughttbouglit to have deVielopeddeVelow by metamorphism of relatively fifine-grainedne-grained sediments with marked calcareouscalcareous content.content. 16

(a) Hornblenderbiotita-Hornblende-biotite gneisses These occupy a large region in the north of thetbe area and pass westward into horn- blende gneisses, the transition being gradual over a width of outcrop of a quarter of a mile. They are typically buiTbuff or grey in colour, with the madema6c minerals arranged arranged in layers interspaced with more leucocratic layers, giving a poorly dedefinedfined banding usually on a fifinene scale, though locally alternating bands up to halfbalf an inch thick thick were noted. In thin section the hornblende is found to be pleochroic from light to very dark olive- green, and the biotite dichroic from yellow to dark greenish-brown or black. The biotite is usually associated with grains of Opaqueopaque iron ore and is often marginal to or intergrown with the hornblende, to which it sometimes shows a clear replacive relation- ship. Plagioclase felspar present has the compositioncom~sitionof oligoelasesligoelase or oligoclase- andesine. In specimen 60/1341mm60/134fmm ,KasasimuchUsadmuch of the original rock material has re- crystallized under pressure.presswe. Most of the hornblendehornblade and biotite is in wisps and grains aligned along thekfoliation,the'foliation, and much of the microcline of the original rock has been replaced by later quartz and orthoclase. Garnet, zircon, apatite and iron ore occur as accessories.

Four estimated modes, listed in order of decreasingdecrersing hornblende content, are:—are: -

60/175 60/143 60/169 60/134 7 % ' 7 , ° QuartzQuaitz .,...... , 22° 25 . 13' (3 .OrthoclaseOrthoclase ...... 13 13 22 19 . . Microcline . . 29:". . — 44 — 55 Plagioclase ...... 52 8 40 6 Hornblende ...... 7 5 4 3 Biotite ...... 5 4 l4 1 Accessories ...... 1 l 5 1

60/175——Masalani. 60/143—Twani1i. 60/169—Dwa. 60/134—Kasasi.

(b) HHornblendeornblende gneisses The hornblende gneisses are exposed in a broad band flflankinganking the TivaTivq river and a narrower band along thethe AliAthiriver. rivq, thoughthough it is fairly certain that they underlieundpdie the superficialsuperficial soils covering the whole of the south-western part of the.area. Their resist- ance to erasionerosion is small, and thoythey outcrop only in river sections and in the dissected catchment area of the Tiva. They vary in colour from bullbuff to almost black, accordingaccarding to their content of mamaficfic minerals. Where exposures are continuous, as in stretches of the Tiva and Diliu rivers, the colour index changes rapidly overeshortover-&ort distances, and in places the rocks are clearly banded on all scalesvales from a fractionfractiqq of antinchan inch to tens of feet. Along the Tiva river north-west oflftuthad Butha the gneisses contain bands of reddish quartzo-felspathic gneiss, perfectly conformable with the strike and which, while often as8s little as two feet thick, can be traced for several miles as‘thtiyfehtas tbbidll9Wcessivefifiecessive meandersmeanders of the river. Sometimes these bands reach hundreds of feet in thickness, when they are shown on the map as quartzo-felspathiequartzo-felspatbic or.or granitoid gneisses. Similarly other bandsbade and lenses of crystalline limestone, homblende-diopsidehornblwde-diopside gneiss, biotite gneiss etc. are intercalated with the hornblende gneisses. InXa alldl thesthe microscopemicrosop slides examined thetho 17

hornblende is strongly pleochroicpleoehroic pleochroic in shades of olive-green and freQuentlyfrquently has small amounts of biotite in contact with it, sometimes as a replacement,replacement, the latterlatter being dichroic fromfrom straw-yellow to dark grey-brown or black,black. In specimen 60112160/121 the biotite occurs occurs as slmsds~h~mblendesheds Whornblende at wide anglesangles to the foliationfoliation andand occasionally as cores to hornblende crystal aggregates, suggesting suggesting here aa replacementreplacement of biotite by hombleqde.hornblende. though'though in1n ogtono part of the slide biplti?bipthe ISis seen to be followed by hornblende without replacement. Quartz occurs in in all .thethe slides examined, and in all but one plagioclase felsparfelspar is,is an important constituent, constituent, ranging inin composition composition from acid oligoclase to andesine. Orthoclase felspar felspar occurs occurs in varying amounts in most of the sections, but microcline occurs only in the the slide slide in which plagioclase is lacking. lacking. This slide, specimen specimen 601194,60/194, from Kaandui,Kaandui, is much fifinerner in grain thanthan normal normal and is almost schistoseschistose. Except where quartz and felsparfelspar have been been re—crystallizedre-crystallized the consti-consti- tuent mineralsdn"es allhll show shattering and strain effects, as a result of strong shearing.

Minerals present present as accessories in the the hornblende hornblende gneisses are biotite, garnet, garnet, zircon, apatite and iron ores, ores, and in specimen 60110860/ 108 hypersthene and diopside. Some estimated estimated modes are:are:— -

1 60/120 i 60/153 i 60/121 60/210 ' 60/108 J 60/194 ‘ 60/209 , ° % ° ° % % % QuattzQuartz ,...... '7 20 5 46 8 4o 53 OrthoclaseOrthodnse . . . .. 15 ~— 35 8 22 9 10 - - MicroclineMicrochne . . ., ... . ~— —— —- -— -—- 35 - PhgioclascPlngloclase ... . '‘ . . 55 62 45 35 , 55 —— 25 HornblendeHornblonde ...... 22 16 15 12 10 8 8 Biotite.Biotin. ,. . . . . 1 + '1 2 2 2 2 Other mner&'minerals ... . + 2 1 3,1, 3 5t . 2,

‘Mainly*Mainly iron ore. 60/120~—-Kisou.601 120-Kiu. 60/153—Diliu.6011 53-Diliu. 60/121—Ngwate.60/121-Ngwate. 60/210—Mbatu.60/210--Mbatu. 60/108—Muthungwoi.60/108-Muthungwoi. 60/194—Kaandui.6011 94--Kaandui. 60/209—Kazi.‘601209-Kazi. (c) HornblendeHornblende-garnet---fiarnet- gneisses--. The aspect in thethe fieldfield of the hornblende-garnet gneisses points to their being local variations of the hornblende gneissesgueisses in which two two of the three exposures were found, lying conforrqablyconfomiably to their strike.strike. In no case were contactscontaots between the two rock types types found, however, and the sedimentary originorigin of the hornblende-garnet gneisses mnnotcannot be regarded as proven. The two specimens examined examined in thin section, 60112760/127 from Nzauni and 60118860/188 from Makaiye,Makaiye, are from outcrops outcrops separated in in the fieldfield by by six miles, and apparentlyapparently are not parts of one continuouscontinuous outcrop. In hand-specimen both are fine-grained,fine-grained, dark grey-brown in colour, with foliation poorly markedmarked by a fine-fine- scale alternation of leucocratic and-melanocraticand .melanocratic bands. Under thethe microscope both show the hornblende to hebe pleochroic in olive-green, associated with opaque iron ore, ore, and pale pinkpink garnets,garnets, euhedral or subhedral and fractured,fractured, butbut without embayed out- lines. The only felspar is plagiodase of composition An..+An,, + (labradorite). Hypersthene is a major constituent in 60/601 127 and an accessory in 60160/ 188, and is strongly pleochroic, 18

X = pink, Y = yellow-green and Z = blue. Diopside occurs only in 60/188601 188 where it is a major major constituent. constituent. The modes of these twotwo rocksrocks were were estimated estimated as:—as:-

60/127 60/188 .. % % Qua?.Quartz ...... — 5 PlagmclasePlagioclase ...... 50 41 HornblendeHomblende ...... 15 8 Biotite.Biotite ...... — + Diopside v... -.-. . . — 18 Hypersthene ...... 25 2 Garnet...... 8 23 Iron ore ...... 2 3

601127-Hornblendehypersthene-garnet60/127—Hornblende-hypersthene-garnet gneiss, Nzauni. 60/188—Hornblende-diopside-garnet601188-Hornblende-diopsidegarnet gneiss, Makaiye. (d)(4Hornblende-diopside gneisses The hornblende-diopside gneisses are dark dark grey in colour, of medium or coarse grain, and usually foliated on a hefine scale by the concentration of the melanocratic minerals in marked layers. In thin section section the hornblende is seenseen to be pleochroic from light toto dark olive green, and intergrown with diopside, faintly pleochroic in light blue- greens. Where hypersthene hypersthene occurs it is strongly pleo‘chroicfleo'chroic from pink to green. In spec+enspecimen 60160] 157 from Ndoyani ..*erethere is clear evidence of replacement of both diopside and hypersthene by hornblende, a good deal ,ofof Ironiron ore having been ,thrownthrown out in thet$e process. In the same specimen are a few flflakesakes of red-brown biotite derived from horn- blende.blende.Plagioc1ase Plagioclase occurs in all the the sections examined and is andesine except in specimen 60/118,601 118, from Katakolo, where it is oligoclase-andesine, and is accompanied by orthoclase. Estimated modes of the hornblende-diopside gneisses are:—are :-

60/161 50/157 60/180 60/118 % % % % Quartz .... —— —- 20 5 Orthoclase — — —— 15 Plagioclase 55 45 61 47 Biotite .... — _ 1 + + HornblendeHomblende 28 ‘ 23 12 12 3 Diopside 16 ‘ 15 4 8’ r HyperstheneHypemthene —.-_ . 13 — . .8 . Garnet .... I —— + —— “’ Apatite ... . + 1 1 2 Iron ore. . + 2 2 3

60/161—Koa1a 60ll$7-—'I!H‘d}*ani. 60/180—Uthainga. ., V 60lll8—Katakolo. ‘ ' “-2" “‘5’? (3) METAMORPHOSEDMBTAMORPHOSBD PELITICPatinc SEDIMENTS 9. This section comprises comprises only two rock types,types, of very varying aspects—plagioolaseaspects-plagiodase amphibolites and sillimanite gneisses. 19

(a) Plagioclase amphibolites The fifieldeld relationships of the plagioclase amphibolites could usually.usually be determineddetermined with ssaesome degree ofof, - ertainty, and proved that they are conformable with the trend of the rocksrocks in tl%&jfciMan %t tp,fly, or are found in conformable contact with other rocks, and all are consideredconsidered to be of sedimentq,,origin.sedimentaryaorigin. As already mentioned mentioned there is a grmationgraduation of types~betweentypes~ibetween the plagiodaskp1agio‘ei‘5iS‘e'amphibolites -amphibolites and the hornblende-bearing gneisses. Rocks included included in this section section " are those with with a a high amphibolelpyroxeneamphibole/ pyroxene content and a fairly calcic plagioclase. All All are of fairly coarsecoarse grain,grain, darkdark grey-green to black in colour, speckled with white or pale pink felspars. Foliation is usually fairly well marked by alignments of individual crystalscrystals and and the tendency for the felsparsfelspars toto occur in"dismtgin‘tdiscrete layers. In thin thin section thethe hbrnblendehornblende is found to to be pleochroic in olivs,olive, green,greeq, and sometimes shows clearclear evidence.evidence of replacing diopside and and /or hypersthene. !hethe hornblende itself sometimes showing replacement by biotite. No No quartz was found in any of the slidesslides and the only felspar is plagioclase, varying from andesine to labradorite. labradorite.Iabradorite. Calcite occurs in specimen 60/159601159 from Maindunduni as lobate patches' intergrown intergrown with with hornblende. hornblende. Other calcic minerals are are rare,rare, being represented only by sphene and apatite, bothboth in trace amounts.amounts. Estimated modes modes are as follows:-follows:—

60/113 60/159 60/176 60/201 60/145 % % % % % Plagioclase . . . . 25 31 25 52 44 Biotite ...... —- 2 —- + — Hornblende . . . .. 67 65 55 38 37 Diopside ...... — — 20 8 —— Hypersthene . . . . 7 — — —— 18 Calcite.Calcite ...... —— l — — V—. A~atiteApatite ...... — + — + — ~iheneSphene ...... — —- ' + — A— Iron ore . .. . . 1 1 — 2 1

I

601160/l 13-Mavia.13—Mavia. 601160/159—Maindunduni. 59-Maindunduni. 601176-N60l176.—N omano. 601201-&oni.60/201—‘ athoni. 601145-Kimathena.60/145—Kimathena. (6)(b) SillimaniteSillinianite gneisses Only one representative representative of this rock type was found, in the nose of a plunging anticline alongalong thethe railway at Kabingo. It is poorly exposed in drainage trenches and is not thought to exceed 30 ft. in thickpefs,thickness, its o- outcrop; havinghaving been necessarily exagge-exagge- rated on the map. In hand-specimen (601172)(60/172) it is of medium grain, fairly friable, white and purple in colour, the latter due to powdery iron oxides. Under the microscope the sillimanite is seen to occuroccur in felted clusters of fifinene fibresfi fibresbres enclosed enclosed in a quartz matrix. AssociatedAssociated minerals are rutile, in stubby prismsprisms and anhedral anhedral grains, anhedral grains of opaque iron oreore andand lesser amounts of translucent red red iron oxide. Its mode was estimated as:-as:— QuartzQuartz 47 per cent, sillimanitesillimani& 50 per cent, rutile 1 per cent, iron iron minerals 2 2 per per cent. (4) METAMORPHOSEDCALCAREOUS SEDIMENTS The only representativesrepresentatives of this class of rock in the classificationclassification adopted for the area areare thethe-crystalline crystalline limestones. Other rocksrucks with with a a lesser calcareous content have been classiclassifiedfied --asas hornblendic gneisses and migmatites, and included with with the pelitic and semi-pelitic rocks, rocks, or the migmatites. migmatites.migmatites, 20

(a) Crystalline limes-toneslimestones These occur as bands and lenses sometimes reaching a thousand feet in width of outcrop. All are very coarse in grain, that at Otekilawa having grains of calcite up to 2 cm. in diameter. Colours range from pure white to light red and shades of grey and blue-grey. The colour often varies widely in didifferentfierent parts of the same outcrop, and could not be used as a criterion for identifying discontinuous outcrops of the same hand. Specimens from all the larger outcrops were tested bybyia a colorimetriccolorimetric method and found to contain a small percentage of dolomite except 60/124,’60/124, from Kasaala.

Graphite occurs in the majority of the exposures examined, usuallyutndy as single flakesflakes disseminated throughout the rock, but occasionally in large lenses enclosed in and marginal to the limestone.lislestone. At"KmWui,.-onAt"Kamstoi,'"on the northern boundary of the area, such Lnseslenses are being worked as graphite ore. A microscope section of a hand-picked speci-speci— men of this ore, 60160/190, 190, shows graphite in flakesflakes up to 5 mm. in diameter intergrown with quartz and felspar and associated with small amounts of apatite, phlogopite and a pale green mica. The graphite content of the slide was estimated as‘as3030 per cent.

Other accessory minerals in the marbles occur both as disseminated grains and as knots and lenses which seldom reach a diameter of more than 10 cm. Such knots tend to stand out sharply on weathered surfaces. One such knot, specimen 60111760/117 from a point two miles south-east of the Kamstoi graphite workings, was sectioned and found to consist mainly of cloudy felspqfelspars, chiechietlyfly microcline but with small amounts of orthoclase and plagioclase (oligoclase). Other minerals present in signisignificantficant amounts are carbonate relics, quartz in clear, elongated belbs and pale browbbrown to cdourlesscolourless phlogopite. Apatite, sphene andandnopaque opaque iron ore occur in trace amounts.

Quartz is a fairly common constituent of the limestones, and chalcedony occurs in specimen 60160/106 106 from Muthungwoi. Scapolite occurs in large grains in specimens 60/141601141 from Kasasi and 60160/ 164 from Mutula, and in a slide of the latter is associated with phlogopite. Forsterite occurs widely, sometimes altering to talc. In specimen 60/601208208 from Waamata on the Tiva river in the extreme south-east scapolite is seen in large rounded grains. This last specimen is interesting as beingbeing-particularly particularly impure, knots of greenish black mineral aggregates making up one third of the bulk of the rock, the calcite groundmass being pink in colour. In addition to scapolite the follow- ing minerals were identifiedidentified in this section:-pinkishsection:—pinkish brown sphene,Sphene, apatite, quartz. diopside and actinolite.

'(5) MIGMATITES In the east of the area and locally in the hornblende gneisses in the centre there are outcrops of rocks that are characterised by the intermixing of mineralmiqeral aggregates of very marked contrast, taking the form of leucocratic host-rocks enclosiggenclosing fen??reuse; and scblierenschlieren of melanocratic character. They are characterised by acute c‘ontortionscontortions on a small scale, indicative of tight folding.whilstfolding whilst in1n a semi-plastic state. mereWhere such rocks are weathered to smooth pavements or low rounded erosion residuals, as is ofteh the case, it is usually possible to determine their overall strike and dip, which always prove to be in good accord with the regional pattern.pa-.ttern The melanocratic components of these rocks are always hornblendic, but the host-rock sometimes contains biotite and some- times hornblende, and it is on this crited~~~thatciiteri’qrfithat they are diviW.divided.

(a) Hornblende migmatites Hug—.-;-A .-.*-:-me:- TheThe' hornblende migmatitesmigrnatites occur as narrow bands near thethe centre of the area and in larger exposures in the south-east.southeast. Along the Tiva river east of Wathoni there is a gradual transition westwards into hornblende gneisses, and the divisioadivision marked between the two types on the map is onlyapproximate.only approximate. A typical specimen, 60/182601182 21 fromfrom Kwambagi, isis coarse-grained, and cream inin colour with black inclusionsinclusions which frequentlyfrequently taketake thethe formform of boudinage bodies. InIn thinthin section section thethethe leucocraticreucomatic faciesfacies isis foundfound toto consist predominantly of felspar,felspa~,includingincluding both slightlysaightly microperthitic ortho- clase and plagioclase of thethe composition of andesine. Other constituents are quartz and hornblende, thethe but:hbx{s-g fishing 3 per centcent of thethe whole, with tracetrace amounts of biotite and ironiron oreore. The melanocraticmelanocrdc facies,facies, specimen 60/183,601183, isis black inin colour, banded with bullbpff layersiayers of leucocraticleuqocratic material tha-1;that, occasionallyoccaslmally reachreach 2 cm. inin thicknessthickness and give thethe rockrock a wavy foliationfdiation 1nin good accord with thethe regionalregional foliation.foliation. A thinthin section was cut across such a leucocraticleucocratic layer, layer, and shows itit to to consist of quartz and ortho-ortho- clase felspar,felspar, sometimes microperthitic, withwith occasionaloccasional ill-deill-definedfined lobatelobate patches of myrmekite. The melanocratic layerslayers inin thisthis section section are virtually plagioclase amphi- bolites, with.with an almost equal percentage of hornblende and andesine. Other minerals, inin order of theirtheir importance,importance, importance, are hypersthenehypersthene, opaque ironiron ore, ore, orthoclase, apatite, biotitebiotiteUahdvand gal-net.gahet. Specimen 60/197601 197 fromfrom Chumanenze shows an unusual texture inin thethe melanocratic parts. ItsIts black groundmass is dotted with bull-colouredbuff-cobed blebs consisting of aggregates of garnet grainsgrains rimmedrimmed by felspar.felspar. Such blebsblebs sometimes sometimes reachreach 2 cm.crn. inin diameter,diameter. and often show a good crystal outline,ontline, with prism edges and dome or pyramid termina-termina- tions.tions. Their outline suggests thatthat theythey may be pseudomorphs after staurolite or possibly andalusite,anddusite, thoughthough thesethae minerals were nowhere foundfound inin thethe the area.area. InIn thinthin section thethe garnets are seen toto be pale pink inin colour with frequentfrequent inclusionsinclusions of opaque iron ore. They generally have a sutured and embayed outline, but occasionally show good crystal facesfaces inin part. part. The felsparfelspar isis andesine. Homblende,Hmblende, pleochroicpleochruic from lightlight toto dark olive-green, isis thethe predominantpredominant mamaficfic mineral, and hypersthene alsoalso occurs. occurs. The estimated mode of the mamaficfic parts of the rock is plagioclase 28 per cent, hornblendehurriblende 47 per cent, hypersthenehypysthene 12 per cent, garnet 10 per cent, iron ore 3 per cent and a trace of apatite. (b) Biotite migmatiteshigmatites A typical example of the biotite migmatites is profdedprovi'ded by specimen 60/116601116 tramfrom Ndnlukye.Ndolukye. The leucocraticleucomatic component is mineralogicallyminerdogically a fairly typical biotitebi&e gneiss,gmiss, except that the plagioclase is rather basic, being of the composition of‘andesine.of andesine. It usually shows albite twinning. Biotite, dichroic from yellow-brown to black, makes up 12 per centant of the rock and quartz and orthoclase are important constituents. In the melanocraticmdanocratic components biotite makes up only 2 per cent against 65 per cent of hom-horn- blende.blen&. PlagioclasePlagiochse felsparfdspar here is mostly untwinned, again andesine.andefine. Small amounts of orthoclase and quartz also occur. In specimenspecimen 60/204601204 from Nza the plagiociaseplagiodase fdsparMsppr is oligoclase in the leucocratic parts but again andesine in the dark components. In the latter biotite occurs in minor amounts partly replacing hornblendehornMeode and pyroxene; the hornblende itself seems to have been derived from diopside and hypersthene,hypersthe, both of which are still present in significantsigni significantficant amounts. No quartz or orthoclase were seen, and the mode was estimated as plagioclase 33 per cent,cmt, biotite 3 per@r cent, hornblende 40 per cent, diopside 7 per cent, hypersthe1_1_ehypersthene--- l_515-per- per cent- - and apatite and iron ore lI per cent each. (6) INTRUSIVES mo THE BASEMENT SYSTEM Intrusives that have penetrated into the Basement System rocks are few and small, those recognized beingbehg a pyroxenite at.at Makaiye, westwest of Ikutha, a biotite pyroxenite at Wathoni,Watbd, an olivine norite at Yumba, two miles south-south-west of Ikutha,Ikutba, and minor acid pegmatitesatpegmatites at various localities,localities, all too small to map. SomeSome of the pegrnatitespegmatites are not true intrusives, but are considered here for convenience. (a) Pyroxenite The pyroxenite at Makaiye (specimen 60/126)601126) takestakes thethe form of a slightslight ridge ridge elongated along the strikestrike of the neighbouringneighbring hornblende gneisees,gneisses, the slurp&up junction between the two being exposed only in one small gully. It is considered to be an ultra- 22

basic dyke or sill emplaced in the Basement rocks before metamorphism. In hand- specimen it is non-foliated,non—foliated, of medium grain and dark blue-grey inin' colour, speckled with white felspars. In thin section the predominant mineral is a hedenbergitic diopside, pleochroic from yellow to medium blue-green. The felspar is all plagioclase (labradorite), with strong albite twinning, and is usually somewhat saussuritized margin- ally. Epidote and sphene occur in small irregular grains, and opaque iron ore in fairly large anhedral grains, always in contact with or enclosed by diopside. The mode was estimated as plagioclase 27 .perper cent, diopside 70 per cent, epidote 1l per cent, iron ores 2 per cent and sphene in trace amounts.amounts In the biotite pyroxenite at Wathoni, specimen 60147,60/47, the predominant mineral is enstatite, plwchroicpleochroic from pale green to pink, intergrown with lesser amounts of horn- blende and pale green diap*,diopsuie, The _bi@,bi_oti_t_e, which is dichroic from straw-yellow to deep red-brown occurs in large ragged llakes.01igoclaseflakes. Oligoclase is the only felspar present, totdingtotalling less thamthan 10 per cent of the slide.

(b) Olivine norite The outcrop of this rock is slightly exaggerated on the map. It is seen as large rounded boulders exposed in red soil over a roughly elliptical area 300 by 150 ft.,ft, the long axis being aligned along the regional strike. No other solid rock is exposed nearby, ear by, and it was concluded that the outcrop represents an erodederoded, volcanic plug. In hand-specimen (60/(60/154)154) the rock is very coarse, dark grey and black, with no trace of foliation The most conspicuous mineral is a black mica in plates up toto 2 cm.crn. in diameter. It isis, not vermiculitic. In thin section the mica is pleochroic from very pale straw Jellowyellow to light brown or green, and is considered to be phlogopite rather than biotite. Olivine occurs in large anhedral crystals, much fractured and frequently rimmed ‘ -by‘by very pale green hypersthene. This hypersthene does not appear to be an alterationalteration product after olivine but rather a reaction product between the early-crystallized olivine and the residual magma. Almost colourless hypersthene, with ex-solution lamellae, and sometimes small inclusions of olivine, also occurs as large d$cretediscrete crystals. The felsparsfeISpars sometimes show very poor albite twinning and more frequently a vague cross- hatching reminiscent of microcline.rnicrocline. Refractive index determinations show it to be a sodic oligoclase.Oligoclase. It includes many prisms of apatite and blebs and streaks of calcite. That such a sodic felspar should occur in a rock basic enough to contain a large propor~propor- tion of olivine can be explained by the separating out of calcium from originally basic felspars in the course of metamorphism thus raising them to a more sodic status (the present Oligoclase),oligoclase), the calciummlcium being recrystallized as calcium carbonate. By this criterion alone it is considered that the rock was emplaced before the conclusion of the regional metamorphism and granigranitizationfization of the Basement System rocks. The mode of the rock,rock estimated from examination of two thin sections, is plagioclase 18 per cent, phlogophite 22 per cent, hypersthene 20 per cent, olivine 38 per cent, apatite 1 per cent, calcite 1 per cent, iron ore a trace.

a a , k* (c)(0) Minor acid pegmatitespegmatiter t A Two occurrences of coarse pégmatitespkgmatites were noted, but in neither iasecase could their relationship to the surrounding gneissesgaeisses be established. Specimen 60/178601 178 was taken from large floatfloat blocks on the slopes of the Yatta Plateau near Masalani. It consists of white quartz, pale cream felspar (determined microscopicallymicroscop~yas albite) and lesser amounts of biotite and muscovite. 'I;heThe ..biotitebiotite occurs in fq-cturedfractured plates up to 5 cm. in diameter, and is slightly vermicditk.vermiculit The muscovite flakesflakes never exceed a few millimetres No other minerals were seen. - .* --.-m‘. , -.*,

Specimen 60112360/123 from Mongua was taken from a pegmatitepegmatite seen cross-cutting quartzo-felspathic gneisses. It is cream and pink in colour, and consists only of quartz, felspar and biotite, flakesflakes of the latter averaging 2 cm. in diameter. It is one of many stringers that reach a width of up to 30 cm. growing outwards from knots up to 1 m. in diameter the =&fhgt53~narrowingfingers-‘narrowing with increasing distance from thethe centres. They show no displacement of the country country rock yrnor any-any chilled margins, and areare replace- ment pegmatites formed by re-crystallization'ihre-crystallizatioij‘ii: s‘itu'Lofsitu of the minerals of the country rock under thethe influenceinfluence of rmetasomatizingmetasomatizing agencies.

In the bed of the Tiva river at Munyuni where the hornblende migmatite is cut by minor shears, replacement veins in the form of stringers follow the shears. TheyThey seldom exceed 20 em. in thickness, but are are unusual in that they consist of a groundmassgroundmass of magnetite whifhwhich supports phenocrysts of felspar and a little quartz. In specimen 60/205601205 about 75 per ceqtcent of the rock is magnetite, with white translucent felspar phenocrysts which average about 1 cm. in diameter making up 24 per cent, the remaining 1 per cent being quartz.

2.—Tertiary—Yatln2.-Tertiary-Yatta Plateau Phonolite

In hand-specimen the phonolite of the Yatta Plateau, which corresponds to that of the Kapiti Plains south of Nairobi, is dark grey in colour with a grain so fifinene as to be barely distinguishabledistinguishable to the naked eye, supporting numerous numerous phenocrysts phenocrysts of white anorthoclase in elongated plates plates and and a a lesser amount of nepheline in shortshort stubby stubby prisms, almost black in colour and with a resinous resinous lustre.lustre. Anorthoclase phenocrysts were seen that measured as much as 8 cm. in length, but the average is about 2.5 cm. The nepheline phenocrystspifenocrysts seldom exceed 1 cm.crn. No difference in texture or phenocryst content was noted along the whole length of the flflowow in this area. Hollow vesicles are rare but areare sometimes seenseen with with a thin thin c’natinghating of zeolites, tentatively identified identifiedidentified as.as natrolite. Smaller Smaller vesicles up to 3 cm. in diameter completelycon~pletelyfifilled lled with analcime,'areanalcime, are fairly common. A typical specimen, 60160/129 129 from Nzauni, was sectioned and, besides anorthoclase and nepheline, anhedral'anhedd and euhedral microphenocrysts of olivine,olivide, usually rimmed by cossyrite, were were identified. identified.identified. The groundmass consistsconsists of a feltedfelted mass of laths of anorthoclase, small crystals of nepheline, cossyrite pleochroic from brown to black, kataphoritekataphori-te pleochroic from very pale green to purple, and lathslaths andand angular angular grains of deep green aegirine. Augite also occurs,occurs, palepale green green or'or brownish-green in colour, often often simply twinned twinned on 100. Other minerals identifiedidentified werewere a few smallsmall anhedral grains of apatite, small grains of opaque iron ore, and analcime which forms an abundant base, and is particularly noticeable in irregular patches scattered through- out the rock in in which there are deficienciesdeficiencies of dark minerals.

3.—-Pleistocene3.-Pleistocene and Recent .-.--i-._- _- , . . ’».«.* -.- (1) LAVASLAVM

Specimens of the Chyulu lavas that flflowedowed down the Kibwezi and Masongaleni valleys were sectionedsectioned (60(60// 168 and 60/601170)170) and identiidentifiedfied as olivine basalts. In the fifieldeld thethc lavas are darkdark blue-grey in colour, very scoriaceous, with a fifinene granular texture andand no visible phenocrysts.phenocrysts. Zeolites are rare,rare, and where seen lining vesicles were identiidentifiedfied as natrolite. In thin sectionsection the rocks consistconsist of a fine-grainedfi fine-grainedne-grained holo- crystalline groundmass supporting microphenocrysts of euhedral olivine and sometimes prisms of plagioclase. Minerals identifiedidenti identifiedfied in in the groundmass werewere plagioclase ofof composition An50An,, (andesine-labradorite), olivine, very pale greengreen augite and octahedra of magnetite. In section 60/601170170 the olivine phenocrystsphenocrysts are marginally altered to iddingsite. 24

The chemical composition of the Kibwezi lava has been determined as follows:follows:——- -- NormNam Per cent Per catcent Si02..$102 ...... 43.7343-73 or ...... 10.5710-57 N203A1203 ...... 13-40 ab .,~ . . . . . 4.734-73 *Fe203l"F82()3 .. . . 15-0715“” an ...... 14.7514‘75 MgO..M30“ ...1 . ., 10.2110-21 ne ...... 14.2014-20 CaO ...... 9.589-58 di ...... 26.5926-59 Na20 . . . . 3-70 01 ...... 18.5218-52 K2O..K20...... 1.801-80 mt ..1 ...... 5.335-33 TiOz..T102...... 3-01 il1.1 ... . ,... . . 56785‘78 Ma0MnO . . . . -.U)-20 - ... ,.1.... 100-10-,“lao-zo---. 160.47100-47 Anal. East African LndustrialIndustrial Research Baard,Board, Feb.,Feb, 1951.

‘ Total iron as Fe,O,.Pep... From more recent, as yet unpublished dys&analyses of Chyulu lavas the ratio FeO:FeO:Fe,O, Fe,O, is assumed to be 7:7:1 1 No felspathoids wereWere recugnizedrecognized in the thin section and the presence of so much nepheline in the norm is remarkable. Presumably there is an alkaline mesostasis that is not easily detectable.

(2) Auuvm. Drposns 4 . Along the course of the Athi river are alluvial deposits con-consisting mainly of fifinene sand with occasional thin layers of grey silt harkingmarking seasonal floods,floodsi ThiclurefsesThickness-es of alluvium up to 12 ft.ft wen:wep measured, but the base was seldom seen a4and it mrrymay be much thicker in places. The alluvium strqtchesstretches for up to 500 ft. on either sideside, of the river in the north-west, narrowing southwards to less than 109100 ft. Very thin mcomolidatedunconsolidated gravel sheets sometimes occur, consisting of poorly-roundedpoorly-round@ pebbles of Basement rocks and phonolite. Many of the other rivers, notably the Tiva; Mavuko adand Woboto,Kaloboto, have sandy beds but no Aankingflanking alluviumalluvium. Panned concentrates from the#the.Tiva river showed (after removal:removal of quartz, felspaxfelspar and biotite) magnetite, ilmenite, gafllet,garnet, horn-born- blende, diopside, rutile, zircon, apatite and sphene. In addition to all these mineralsminerals. a concentrate from the Athi river (specimen 60/165a) showed a high proportion 06of epidote, which is very rare in the rocks of the present area.

(3) Surrarrcm. Dru-05m Over much of the area a red or red-brownred-browtr soil, usually sad&,sandy, has developed by weathering of the underlyingunderlyhg rocks. Where the vegdationvegetation covercayer has been removed by overgrazing or poor husbandry gullying quickly develops and leads to barren sandy areas. The phonolite of the YattaYatb Plateau weathers to a non-nun-sandysandy red soil very similar in appearance to that derived from Basement System rocks.rodks. Crops planted in the few places where the bushbus]: has been cleared on the PlakauPlateau show that this soil is more fertile than those derived from Basement System rocks. In areas of poor drainage typical "black-cottan"“black-cotton” soils are developed, hard and fissured when dry but tenacious clays when wet. In most of the black soil areas examined a discontinuous layer of nodules of Secondarysecondary limestone occurs at or a few inches below the sdrface.surface. Unlike'$&Unlike31%“ red soils, which support a thorn-scrubthorn-scfub vegetation, &tenoften dense, the black soil areas are always open grassland...... ':a, .. Over 'andand near outcrops of the more calcareous'focksx$econdarycalcareous rocksKecondary limestones are widely developed. They are Usuallymually buff or pink in colour due to iron-staining, andaad often include angular fragments of quartz and felspar. They seldom exceed a few inches in thickness. - .- 25..

VI-METAMORPHZSMVl—METAMORPHISM AND GRAM'FIZATLONGRANITIZATION - Like many other areas of Basement System System rocks inin Kenya Kenya the the sequence sequence of events appears to have b fimfifig deposition of sediments, secondly wides$readwidespread folding with with accdrnpanylngaccompanying r6gire %% aiaifl~iphistn,fiorphisin, and lastly granitization.granitimtiOn That the granitization was later than the regional regional metamorphismmetamorphism islpiovedis 11110ved by the formation of fine-scalefine-scale layering of re-crystallized quartz and felspa~;felsparf skidomseldom showing strain strain shadow,shadow, between layers of cloudy, fractured and strained minerals. minerals. The abundance of orthoclase or microcline felspar in these later layers, compared withwith thethe felspars in the older older portions of the rock, show that thethe metasomatizingmetasomatizing agentsagents werewere potassic.potassic. WhereWhere granitization granitization has been locally locallylocallyrmore more intense felsparfelspar porphyroblastsporphyroblasts developeddeveloped and grow laterallylaterally intointo the minerals of the surroundingsurrounding rock without noticeablynoticeably displacingdisplacing them,them, showingshowing that the porphyroblasts grewgrew by assimilation of the existifigexisting minerals,miner@, without significantsigni significantficant addi-addi- tion to the-the rockrock ; mass. At thethe highest highest degree degree of granitization the rocks invadedinvaded attained a degreedegree of plasticity that led to theucontortionsthe contortions contortions and swirlings seen in thethe migrnatitesmigmatites and finallyfinally to the obliteration of most of the originaloriginal sedimentarysedimentary characterscharacters of the rocks, rocks, andand. giving granitoid gneisses. The metamorphic grade grade of the area is high, and is characteristic of the almandine- diopside-hornblende sub-facies sub-facies of the amphibolite amphibolite faeiesfacies (Turner, 1948, pp. 87-88),87-88), a typical mineralmineral assemblageassemblage being hornblende-plagioclase-hypersthene-diopside with or without quartz and orthoclase. Plagioclase Plagioclase is commonly of the compositioncomposition of oligo- clase or oligoclase-andesine. HornblendeHomblende is sometimes clearlyclearly secondary secondary to(to the pyroxenes but but sometimessometimes apparently apparently in in equilibrium equilibrium with with them. Where biotite biotite occursoccurs it is usually secondary after hornblende. Further evidence as to the high grade of meta-meta-meta— morphism is given-bygiven.by the occurrence of sillimanite in one exposure, the alumina content of the original sedimentssediments being too small to allow thisthis mineral mineral to develop elsewhere. elsewhere.elsewhere IhIn the crystalline limestones actinolite, forsterjteforsterite and diopside are commonly developed.developed The appearance of these minerals is in good accordaccord withwith metamorphismmetamorphism to the stage of the amphibolite facies.facies. - r '

VII—STRUCTURE The main structural structural features of the area area are shown in in Fig. Fig. 2. The general strikestrike of the rocks is west-north-west to east-south-east, and is in good good accordaccord with many of the other other BasemeatBasement System areas of Kenya so far mapped. In the extreme south-west the strikes strikes trendtrend east east of north. The lack of exposures between this comer andand the remainder of the area makes correlationcorrelation of ”the'the two parts difdifficult,ficult, but for reasons stated later thethe writer considers considers that there is a gradual transition between the two strike directions and not an abrupt change suchsuch as might be caused by a major fault. Foliations are readily determined in mostmost of the exposures,exposures, and are marked mainlymainly by the preferred orientation orientation and alignment of mineral grains grains and the tendency for maficmafic mineralsminerals to occur in well-marked hands.bands. ,,The.£oliationThefiation invariably parallels parallels the the junctions of the different rock types, and marks the the bedding planesplanes of of the original sediments. Lineations are marked by fifine-scalene-scale puckering on foliation planesplanes and the align- ment of elongated minerals minerals and mineral grains. grains. Six hundred poles to to observed foliation planesplanes werewere recordedrecorded in the area (excluding the south-west corner,corner, which which is considered in in detail detail later) later) and were plotted plotted on a Schmidt equal-area equal-area stereographic stereographic net net and contoured (Fig. 3a). The concentration of poles along the greatgreat circlecircle z1: indicates a single systemsystem of folding withwith p[3 as fold axis. The mean strike,strike, shown shown by the maximum maximum point point of concentration, is‘is is 340".340°. with aa westerly dip of 55".55°. It will be be noted that the mean mean strike strike differs slightly from the direction of the fold fold axis, axis, which is is 334",334°, owing to the plunging of the fold fold axis axis to the north at at 10°10" from the horizontal.horizontal ~ 26

In the same part of the area 173 lineationslineations were recorded, and the contoured plot isis shown inin Fig. 3b. The maximum of this diagram correspondscornsponds to a mean lineationLineation direction ofof 330°330" with a northerly plunge plunge of 12°.12". The regularngulrv extensionextenaim of contours intointo thethe southernsouthern part of the plot plot showsshows that thethe southerly-plungingsoutherly-pluoging IineationsLineations belong to thethe same system, and indicatesindicates thatthat the variations in plunge result fromfrom minor minor flflexures.exures. z oo‘5

Ndnvoini rm

."‘ 7% 3 [51? h sent a z’Jo'r s 0 S :OHIIS

* , . t , -vm,.d,i I,’ mithdvectmddwnd

.mduit‘wd~~"d"&a~d& "What. & & W.cmnmmmenI dong fbdl v 1‘};‘.:r {If‘ti 1““lim at.“ m' of d"? Felt-tinder : a. W o; 7’\ W \ Vertical rm s—GmnlhedMnlsketeh-nlpofthelkflhm 27

$3.54.»; a: ’4"... .x ,

(e) -W ~ (I). Fla.Fib Honoured3-Conbmm.d sterognphicetcro~rpphiediagramdiagrams ofof polespdcs toto follationfoliation plane.pllwr andand of {hectic- InIn(kIhthpm:-lie [kill areas-

(#)Pub to hlidon pl.~s,IQIQ areaarts excludingexchniiog thetbe south-westsod-wwt cor-er.corner. SixSI. hundredL& (I) Pole- to blhfion planes, lknlla ' mdhgeoniematl,5,10and15pereentperoaepereentam"dhp, coaQlln at 1, 5.10 MQ 15 per cent per one per cent ucs. z (b)(b)Lht.fhm~,[heath-e, lknILntb.flna areasm enablingex- thetbe south-west&-west corner.corn. OleOk hundredL.ndrrd andand seventy-clcmtp ,Ihreenadhp,»n!omatl,lo,20aa¢lMpreentperoae‘pereentam.th#md~coatmJl,l%~d30praatperonepercentrven.

é1 (c)Q IdeaPdea toto Iollationfolbtion pines,pha, mane-tsoqe-west cornerwmer of thetbe lkathaIlrmfha aria.ubL Seventy-onumdhm-Sevmty-oaca rrrdbpi contomatl,5,lcon~.t1,510ad15pcrmtpcroecp~~ua.flndl§peraalperoaepereeatarea. 4. .

(d)fd) LlneaLinenfionr,flona, southwestsodb-weat cornercorner ofof thethe MaIblh. area.sm.Twenty-three Twenty-tha readings,mdints, eintonra&torn at-at 5,10,152nd20pereentperonepereemam.S,lO,UnadZ@pcr~tpcroec~cent.rt.. 28

7”“ _ . -_ ___a- - The good accord of the axis of folding .ofof Fig. 3a with the lineation maximum of Fig. 3b is a clear indication of monoclinic SYmmetryshetry in the rock structure, and proves thatthatthe the lineation is a true b-lineation marking the axis of folding. Thus the structure of the area is the re&~&Mpgleresalgofigagsingle major tectonic phase in which pressure was applied along west-south-west to east-north-east directions, the overturning of folds to the east-north-east indicating that- the pressure was~,3<&liedwasgia’ppliedfromfrom the west-south-west. Figs. .3c3c and 3d are qentouredcontoured stereographic plots of poles to foliation planes and lineationslineatiOns respectively 06servedobserved in the south-west corner of the area, which is delimited roughly by a line from Ndovoini south-eastsouth-east to Muliluni. Fig Fig. 3c, while showing a wider spread of points than Fig. 3a, shows a marked conceptrationconcentration of poles along the great c$clecircle x1:, , and a-gainagain shows a single system of folding with P(3 as fold axis.axis The axis of foldingfelding here is 0Uo,0112“, with a plunge of 12°12" to the north.'north.‘ Although only 23 readings wereWere made of lineations,Iineations, the plot nevertheless shows a good maximum, again coinciding with the fold axis and proving homogeneity in the rock structure. The contours of lmeationslineations are elongated towards the west of the diagram, suggesting that the lineations swing over gradually between the south-west corner and the remainder of the area, without a sharp break.

The majormajorfold fold in the area is the Kimathena syncline, with a near-verticalnear—vertical core of granitoid gneiss and biotite migmatite which has, by its compacted nature, resisted erosion to a much greater extent than the flankingflanking gneisses and now forms a steep-steep— sided ridge rising to 2,000 ft. above 'thethe plain. In diriectdirect contrast the Kitui anticline to the west has been eroded at the same rate as the flankingflanking rocks and makes an ill- defineddefined feature. The Kitui anticline was so named by Saggerson (1957, Fig. 2.), who traced it northwar+:throughnorthwards gthrough the South Kitui area and correlated it with an anticline terminating a€zit-Kitui Kitui township (Sanders, 1954, Fig. 3). Other lesser synclines and anti- clines occuroccm between these two major elements»elemenbandand in poor exposures east of Kima-Kima— thena ridgeridge. Severalseveral minor folds were mapped in the extremeeaeme south-west. Between . these two areas of folding is a broad belt of predominantly hornblende gneisses in which only one fold was recognized, the Yumbuni syncline.sync1ine. The regular westerly dip of this band of hornblende gneisses would,Would, if undisturbed, represent an originaloriginal thick- qessness of deposition of upwards of 70,000 ft. The writer considers that the width of this outcrop is rather due to repetition of the beds by strikefaultingstrike-faulting such as the fault followed by the Ngunumu river in the centre of the area. . Faults seen cutting the Yatta ghonolitephonolite on either side of the Ngunumu fault may indicate late movement along Basement System strike-faults. Further evidence bfof such faulting is shown in the rocks of the line of ridges from NgunurnuNgunumu to Mbatu. AUAll the thin sections of rocks from thesetheSe ridges showed characteristics of shearing in the plane of the foliation, though only cross-faultscross-fanlts were identifiedidentified in the field.field.

E. P. Saggerson, who at the time of writing this report was mapping the Simba- KiWiziKitiw‘ezi area to the west of the present arearjnformedthawriterarea,inlbm *writer that plunging folds parallel to thosethose' in the present area reach to within a few miles of the common bound- ary, the meridian 38"38° E.

The evidence therefore suggests that some of the crustal shofieningshortening in the horn- blende gneisses was accomplished by reversed strike-faulting. However the existence of one big fold (the Yumbuni syncline) and the folding on either flankflank of these gneisses suggests that other folds do occur, but have not been detected. In the folded areas minor folding is common, iwith'with amplitudes of 20 ft. or more down to a few inches, alwaysalways.p,arallelparallel or subparallel to the major -,foldsfolds, and marking puckers imposed on the major folds. ' Prominent joints were recorded at 68 localities, all vertical or near vertical,vqtical, and were plotted as a frequency diagram, FigFig. 4. This shows a marked maximum at about 29

My. 5‘33“ o ..

i' I Fig.Fig Faueney4--Fmgoency dismaldiagramof jointsWits in the lkiithaIkdm area. Sixty+&btSixty-eat readings 60°, at right-naglaright-angles to tbbthe tread,trend b06 fold60M a&,axes, representing ac or cross-joints,,cross-joints, producedproduced" by relief of meme&esscs,Set'd$'set up during deformation.dtf6nzlation. Such joints would only oCcur occur upon relief of conco&(Iihgfining pressuresgpemd'sucb'such asa9,bouldinould be produced by unloading, when the rocksr0cks would have lost myany teadcncytendency oto plasticity and would be able to adjust themselves only by fractm.fracture. 'Ufhtfe'Where such*jaintssuch‘3101nts are large and relative movement has taken place along them, wkktwhether at the time of their formation or subscguently,subsequently, true faults are produced. Most of the lesser faults of Fig. 2 are parallel to the upac joint system and are genetically related-mimeacrelated W. ,the uc jointsjoints. VIII-MINERALVIII—MINERAL DEPOSI'ISDEPOSITS The only minerals of economic importance which were found in the area are graphite, linieatone,limestone, vermiculite and sillimanite, thoughth0ugh only graphite is known to occur in sacientsufficient quantities or of a suitable grade to warrant.cxpk&&ion.warrant exploitation.

l.—-Graphite1.--Graphite GrapWGraphitic gneisses occur as lenses and impersistent bands associated with the crystalline limestone:limestones cromirigcroseidg the Ikutha—KanzikuIlrutha-Kanziku road and extending into the area to the north.north.-'- In his repoFtreport on that area Saggerson (1957, pp. 35-43)35—43) has given an exhaustive review of the prospecting and mining done up to 1953. BrieBriefly,fly, the firstfirst record of the existence of graphite was by E. V. KinlochKinloch, then of Masongaleni, who does not appear to 'havehave pegged any claims. In 1942 and 1943 E. RR. Wright pegged claims over one of the marble bands, these claims being abandoned in 1944, after their transfer to Raw Materials Development Ltd. C. R. Stokes-Fair held a single claim between 1945 and 1948. No commercial production was made from any of these claims. In 1950 R. O.0.Johnston erected a protection notice which was allowed&owed to lapse, and in 30

1951 D. K HadtonHamilton (E.C.A. Geologist) and D. HobdmHobden (Prospector) prospected the area for the Mines and Geological Department and proved the existence of economic deposits. In the same year claims were pegged by Bewick Mo&gMoreing & Co. and by Shah VenhiVershi Devshi & Co., both of whom carried out extensive prospecting. Bewick Moreing & Co.'sCo.’s claims finallyfinally lapsed in 1955, and at the time of writing (January, 1957) only Shah Vershi Devshi & Co. have workings in the area, on claims at Kamstoi. In 1956 this company pegged most of the known graphite area under an ExdusiveExclusive Prospecting Licence which is bounded by the triangle of roads joining Ikutha, Mutomo (to the north of Ikutha) and Kanziku. Shah Vershi Devshi & Co., unlike other Kenya producers, have not concentrated on the production of graphite for crucibles, for which the rather exacting specificationsspecifications are based on flakeflake size, but on the production of graphite for surfacing moulds and for lubricants, where tbs-wainthe-main qualiqualificationfication is high carbon content. Owing to the acute shortage of water near the workings for mastmost of the year (bore-(bore— holes have been sunk but their yield is small) the flotationflotation plant has been re-sitedre.sited at the Tiva river near the roadcrossingroad-crossing south-west of Ikutha, and the milled ore is carried by lorries the ten miles from the mine, the dried flotationflotation product being returned to the mill for finalfinal processing, details of which are as yet confidential.confidential. In August, 1956, the flflow-sheetow-sheet for the mill was as given on Fig. 5. A new milling and separation technique, still in the experimental stage, the flow-sheetflow-sheet of which is stmwnshown in Fig. 6, has been devised by the company. The important feature of this new system is the winnow, developed in its present form in the laboratories of the Mines and Geological Depart- ment, Nairobi. After preliminary milling the graphite is blown from the gangue, , giving a product which approaches 50 per cent graphite. ,With With further screening and winnowing the graphite content can be raised to 80 per cent if necessary. The flotationflotation unit is to remain at Ikutha, but much less transport will be needednwded to carry the enriched ore from the mine. 'At‘At a later date it is intended to move all the equipment for finalfinal processing to the flotationflotation p1ant;soplant; so that the finishedfinished graphite can be cartEedcarried straight on to Kibwez~Kibwezi for railage.’railage.* Graphite from Kamstoi commands a ready sale to lubricant manufacturers. The run-of-mill graphite reaches 80 to 82 per cent carbon content, but aftetafter finalfinal processing and screening three different products are obtained, 93 to 95 per cent carbon, 80 to 85 per cent carbon and 65 to 70 per cent carbon. The 93 to 95 per cent fraction has been further treated chemically, on a laboratory scale, to obtain a carbon percentage of 99.5 per cent, but a good deal of work requires to be donedune before such treatment can be regarded as an economic proposition. At the moment only a limited market exists for such a high grade product. Production began in 1952, and records in the Mines and Geological Department show that the output since then has been:—been:-

Year Ore Production Peccenta&sPercentage Value‘Value* Milled 1 R-v . ' Tons Tons {1 ' £ 1952 ...... 400 31 | 7-75 1,406 1953 ...... 1,870 127 6-27 5,765 1954 ...... No production due tato reorganization of millmiU 1955 ...... _ 1,671 116868 10-05 6,653 1956 ...... _ ,_.,<. .~' 6,551 492491-~ 1 7-117-51 1 19,680 1957 ...... ’ 10,791 | 942 8-73 39,847

‘ Estimated value base;based on average bricev&~&~&Idprioe'of‘g‘ra’phitéilsold in tktthat year.

'Final‘Fin’al procasingprocessing has been camedcarried out at the flotationflotation plant since late 1957. .. 31

QUARRY JAw CRUSHER \‘ / auto-M"

v 4 g \ \ 1—»omi- CRUSHING TROMMEL ‘ . v ROLLS set SCREENS —10+50+-1OrS01 -50¢-SO+ t°to 16*16+ The 50"=SO* screens are Hosted concentric around thethe way ‘ 4, ., 3104*:loafscreens ' ' . . -""' ROUGHER FLOIATIONFLOTATION CONDIIIONERCONDITIONER ®@ CELLS ir I V —__, Tailings A toto dump CONDITIONER

/ 6 CROSSFLOW " FLOIATIONFLOTATION CELLS

80-82ZC. DRIER FILTER

f D i 3/

a;

I / 93.95293-952 c.C. Flake- - "‘* _ + FINAL PROCESSING ~-85XC. F~ flake t _ AND SCREENING 89 351 C. Fme flake - < L f f '535-70265- 70% C. Fines f ri-.“(‘J‘W'i‘.*>%-i *Can be brought up toto 96-97196-97% C. TCant Can be brought up toto 907.’90%C. Fig.Fig. S—ai%pibite flle low-sheet—GanshBowaheet-Canesh Claims,C¶aims, KamstoiK~mrtoi 32 QUARRY

to 1/2 " ' - \ / \ /JAWJAW CRUSHERS set5-44.“, set EI—gL—Ea SCREENS

{40* +10? I I

HAMMER MILLS V .4 HAMMER MILLS TROMMEL W SCREENS WINNew- - \ ,T ‘11!) dump CONDITIONER WEIR FLQTAHQN , ——> TailingsTsil~ngs I ‘1 to dump 7 g ‘ 5' I to dumP

Rob MILL d ROD MILL 6CONDITIONER

12 CROSSFLOW FLOTATIONFLOTATION CELLS ‘t i ) i , (4, . - - DRIER - FMER .. 95 901C-

-

3;?!“ ? '' 9697296-97.. c_C. u FINAL PROCESSIN0PROCESSING“I23 *-.?- ‘imx AND SCREENING \ 8080-85%~.as /c, - - Chins, Kamstoi - .,_..,.-.. Fig.Fig +Experimental6—Experimental graphite flflowaheet--Ganeshow-sheet—Ganesh Claims, Kamtoi 33

In 1956 the graphite was shipped in roughly equal quantities of three grades,grades, 93 to 95 per centcent carbon, 90 to 92 per cent cent carbon and under 80 per cent cent carbon. carbon.

Messrs. Bewick Moreing & Co., in the course of their prospecting, cut trenches across the marble outcropsoutcrops southsouth of Kasaala, and uncovered bands of graphite gneiss up to four feet in thickness on the western flankfl flankank of the more easterly of the two bands. These gneisses are nowhere seen at surface due to soil and scree covering covering from the limestone ridge, but it is probable that discontinuous bands or lenses of graphite gneiss occur along the length length of the limestone outcrop, outcrop, and may'may‘ repay further further prospecting.

Appendix on Graphite Reprinted from "Kenya“Kenya Trade and and Supplies", October, 1958, pp. 233-234.233—234. , I - -___

KENYA GRAPHITE

Until fairly recently, the world'sworld’s graphitegraphite requirements requirements were met met chiefly chieflychie fly by Madagascar and Ceylon, but it it isis—cthe the hope hope of Messrs. Shah Vershi Devshi and Com- pany Limited, of Thika, the the company which has developed the.the production of local graphite, that some of these requirements requirements may in futurefuture be met by Kenya. Kenya. The Kenya graphitegraphite mine mine is situated in the Southern Kitui District, District, 175 miles from Nairobi on the road to Mombasa.Mombasa Many difficultiesdifficulties were experienced by by the company when the mine was started, started, chief among these being being lack of water. Three boreholes were sunk at the site of the mine, without success, and water had to be brought in drums from the Athi riyer, 25 miles away.away. Communications were also a major problem and when breakdowns inTfi machinery occurredoccurred it was necessary for messages to be sent sent to Thika, 165 miles away, by by carcar'overover very bad roads. Two oror three days usually elapsed before information regarding the breakdown reachedreached Thika and it was thenthen necessary for spare parts to be procured from Nairobi beforebeforeLrepairs‘repairs could bebe carried out. ConditionsConditions improved improved when the company foundfound water was was available from the Tiva river, about 12 miles south of the mine, and moved their rerefineryfinery to the riverside,riverside. A road was constructedconstructed by the company company which reduced the distance between the mine and the refineryrefinery from 12 to eight miles. Each )rear,year, nowadays, approximately 15,000 tons of ore are transported overover thisthis road to the refinery.refinery. The company has also installed radio-telephone communicationcommunication between the mine and Thika and breakdowns are now usually dealt with in a day or two.

A further problem arose in that during the rainy seasons the levels of Athi river and Tiva river rose to such an extent that the'the road from the rerefineryfinery to the railway station at Kibwezi became floodedfl floodedooded and impassable. This difdifficultyficulty was metmet byby the building byby Kenya Government of a bridge over thethe Tiva rivérrivkr and by the improve- ment of the roads in thetHe district. These didifficultiesfliculties having been overcomeovkrcomc,-+togetherflogether with those of obtaining labour and.and administrative stastafffl? for such an isolatediqolated district, the company continued to develop the minemine withwith an unshakeableunshakable faith in thethe futurefuture of Kenya graphite, which can bearbear comparisoncomparison with with graphite produced in most other parts of the world. Electric plant has been installed in replacement of the diesel engines previously used and testing apparatus has recently beenbeen htroduced.ihtroduced.

Production began on a smallsmaU.5qde1c 1nin 1953, but the necessitynicessity for training the labour and the “trial"trial and error”error" methods which had to be adopted, rendered unprofitable unprofitableunprofitable thethe firstfirst years of operation. By dogged perseverance, perseverance, pmpm;dracSt~flucfion‘ on a commercial scale began inin 1956 and small exports of graphite were achieved in that year. An overseas tour by the Principal of the company resulted 1nin good business contacts and production was stepped-stepped stepped- up to 100 tons a month. Exports rose in 1957, the principal purchasers - -- being as follows: -— 34 Short tons United Kingdom ...... 44.8 Australia ...... 11.2 Western Germany ...... 196.0 United States of America ...... 168.0 Japan ...... 290.0 India ...... 30.2 Canada and Newfoundland ...... 1.1 The future looked bright when, at the end of 19571957, there was a sudden slump in world demand for graphite. This state of affairs continued into 1958 and MrMr. C. DD. Malde, the youngest Director of the company, decided to set out on a.a world tour to study the trend of world markets, to make new business contacts and, thereby, to estimate the futureaffuture._o£ ,the..the... gra&&egraphite industry in Kenya. His four-monthfour~monthtour took him to most European countries, Canada, the United States of America, Japan and India and he reported on his return that he had received the utmost co-operation and assistance from the Governments of all the countries he visited. The demand for good quality graphite in the United States of America is great and there is every hope that a part of this market, at least, will be available to KenyaKenya. Germany is also an important market for graphite and it is hoped that Kenya‘sKenya's sales to that country, also, will 1ncrease.increase. The company employs 250 Africans and eight Asians andand now.now produces approxi- mately 1,500 tons a year of a wide variety of graphite of from 50 per cent to 98 per cent carbon and ranging from fifinesnes through small fl flakesakes to large flflakes.akes. The graphite produced is used for lubricants and paints and in foundries, glass-works,glass-warks, etc.; orders to ’ specificationspecification are also ..metmet. Kenya'sKenya’s ownow^ annual requirements are low at around 100 tons. If the demand from overseas increased, production could could. be stepped up to 2,000 tons a year. ," Every effort is1s being made ;toho reduce the cost of production whilstwhist maintaininglhaintaining the quality of the product and there 1sis every chance that the future of Kenya graphite will justify the faith and determination shown by the promoters of the KenyaKenya,minemine in the early stages of its developmentt.development.

2.-Limestone2.—Limwtone All the larger outcrops of limestone were tested to determine their calcitezdolomite calcite: dolomite ratios, the method usedused‘being- being colorimetric, the ratios obtained being to the nearest fivefive per cent. Results are as follows:-follows:— Ratio Specimen No. Location Calcite, CaCQ,CaCO, : Dolomite, CaMg (C0,)z(CO,), 60/124601124 Kasaala ...... 100.la:: 0 60160/ 138 Kimakimwe ...... 95.95:: 5 60160/142 142 Otekilawa ...... 9595:: 53 60/601 164 ‘ MutulaMvtula ; ...... 959515,:4 5 ,, . 60160/ 184 Kasaala~Kasaala. : . ,...... 95:,9$ : 5 60/20360 / 201 Utundahi.Utundahi ...... 80 :20: 20 0Analyst—Mrs.Analyst-Mrs. R. Inamdar. A ratio of 95.95: 5 is equivalent to a magnesia (M30)(MgO) content of about 1 per cent, which is within the maximum allOwable allowable content in raw limestone used for Portland CementCement. Only random samples;same@, however, were. were taken for these determinations, determinations; and before the limestones as a whole could be proved suitable for cement-making systematic sampling along and across the strike would be necessasyatqtprovenvy&+prove that the low magnesia content persists throughout the outcrop. All the occurrences are near to existing motor tracks orar can be easily reached by new tracks. For small scale cement-making or for burning for quicklime brushwood fuel is readily available at all the outcrops. 35

3.-—Vemficulite Vermiculitic biotite occurs in a massive pegmatite exposedexposed in an old prospecting pit about 15 ft.ft- square and 8 ft. deep at Manoni. The pegmatite lieslies under about three feet of red soil,soil. and no indication was found as to its extent. It was discovered in 1944, and reported on by the Mines and Geological Department. The bulk density of the raw unprocessed material was found to be 56 lb. per cubic foot, and that of the best quality expand*expanded material between seven and nine pounds per cubic foot. While this indicates a fairly good quality the quantityquanticty availableavailable’seems seems to be far too small to warrant exploitation. Flakes of vermiculite up to six inches in diameter were seen, but always in single single crystals,crystals, and the percentage of vermiculite in the pegmatite where exposed does not exceed 5 per cent in volume. Thus even if .thethe pegmatite proved to be ofOf large extent the 9933..“cost of extracting the vermiculite would far exceed its market value. I. -, I .. - --- 4.—Siliimanite The only occurrence noted in the area was in a quartz-sillimanite gneissgneiss exposed exposed in drainage trenches along the railway about a mile south-east of Kikumbuliu Station. The thickness of the outcrop, which is exposed around the nose of a plunging anticline, is apparently no more than 30 ft. From a single microscope section the percentage of sillimanite was estimated as 60 per cent, but the small extent of the ore reserves so far indicated makes makes the occurrenceoccurrence quitequite uneconomic despite its its ideal situation on the railway.

5.—Water-supplies The Athi and Kibwezi rivers areare perennial, the former being fed from the Nairobi arcaarea and thethe latter 'fpomfmm springs in the Kibwezi lava, lava, which which carries a considerable considerable volume of underground water frqmfrom the Chyulu hills (Temperley, 1956, Vol. 11,II, pp. 64- 11111).1). There are two major springs, one at Chae, a mile west of Dwa rock, and the other at Manoni near the snout of the lava, both of which are extensively utilized for irrigation and stock-watering, and the former for the sisal factory at Dwa. The smaller lava tongue at Masongaleni also formerly gave a perennial flowflow to the Masongaleni river (T (Temperley,emperley, op. cit.,cit, pp. 112-122), but this flowflow ceased almost overnight in 1918 for reasons which have not been determined,determined, but which Temperley considers most probably due to diversion of the ground-water flowflow by faulting. (An earth tremor was recorded at Masongaleni on 17th April, 1918.) Water can be obtained by digging a foot or two into the sandy bed of the Tiva river. 'A‘A few minor springs and water-holes occur in the area, but all have only a small supply and are liable to dry up within a few weeksweeks after thethe rains rains have ceased.ceased. Bore-holes Bore-holes have have been been sunk onon Dwa and and Masongaleni sisal estates, and at Ndovoini, four miles north of Dwa Rock. Records of these bore-holes supplied by the Ministry of Works, Nairobi, are as follows:—follows:-

; |

+ + NO.No. I ' ~;ocalityLocality- mthDepth 1 Water*Water- +’ 1 mfer“Pater Yield per . StrutkStruck Pest-levelRest-level 24 hours

Feet Feet Feet Gallons C.33 Masongaleni . . . . 400 ? 42 8,640 C.34 Masongaleni .. .. 254 : ’ ? 4 11,520 073 Masongaleni ...... 313 40 200-230,200—230, 37 24,000 . ,> ' ~77-i'. - $00-313500—313 C.640C.640 / Ndovoini . . ‘ 335 315 45 64,800 c.1004C. 1004 Dwa Plantation (2'rkles(2 miles N 567 .«‘17.-*i*-r 2&28k: 250 12,000 of Masongaleni). ‘ _ C.1005c.1005 Dwa Plantation (2 miles N. 190 150 150 7,200 of Masongfeni).Masongaleni). 2 ‘ (bitter) | . ----,e.. 36

In recent years many dams have been built to impound water in seasonal rivers and streams. They are mostly of earth, built by hand labour or with the aid of a single tractor, and the majority of them hold water for a considerable part of the year. At Ngali, south-east of the peak of Kimathena, and on the Katakolo river fivefive miles south-south-east of Ikutha,Iku-tha, concrete dams have been built across steepsidedsteep-sided channels. Five and a half miles south-west of Ikutha a very large stone-faced earth dam has been built across the Diliu river. At the time of the survey this damdarn was not completed, but when filledfilled it is expected to impound many millions of gallons of water.

The building of numbers of small dams is clearlyclearly‘the the right solution to the water problem in the area, both from the aspect of low cost and high proportion of success so far attained. It has often been proposed that dams should be built across the Tiva river, but at present this would appear to be rather unnecessary, since ample water for domestic use~andstv~kuse ’ahd‘stoc'k watering'k'uratermg is available in1n the sandy bed with a minimum of digging. A dam would very soon silt up, and digging would again be necessary to reach the water held in the sand behind the dam, unless the retaining wall was of sufsufficientficient height to allow sub-surface water to be piped to troughs or tanks. .j . IX—REFERENCES

Aylmer, L., 1908.-"Captainl908.—“Captain Aylmer'sAylmer’s Journey in the Country South of the Tana River, East Africa."Africa.” Geog. lourn.,Journ., Vol. XXXII, pp. 55-59. Barth, T. F. W., 1952.-"Theoreticall952.—“Theoretical Petrology."Petrology.” 1' Behrend, F., 1918.-"Beitrage1918.,—“Beitrage zur geologischen Erforschung der Deutschen Schutzgebeite."Schutzgebeite.” ' . . < Caswell, P. V., 19%-"Geology1956.—“Geology of the Kilifi-MazerasKilifi-Mazeras Area."Area” Report No. 34, Geol. Surv., Kenya. Daly, R. A., 1933.-"Igneousl933.—“Igneous Rocks and the Depths of the'the: Earth."Earth.” Dodson, R. G., l953.—“Geology1953.-"Geology of the South-east Machakos Area."Area.” Report No. 25, Geol. Surv., Kenya. Gregory, J. W., 1896.-"Thel896.—“The Great Rift Valley."Valley.” , 1921.—“The1921.-"The Rift Valleys and Geology of East Africa."Africa.” Hildebrandt, J. M., l879.——“Von1879.-"Von Mombassa nach Kitui."Kitui.” Zeitschr. d. Gesellsch. f. Erdk., Vol. XIV, pp. 321—350.321-350. Krauskopf.Krauskopf, K. B., 1948.—“Lava1948.-"Lava Movement at Paricutin Volcano, Mexico."Mexico.” Bull. Geol. Soc. America, Vol. LIX, pp. 1267-1283.1267- 1283. Krenke1,E.,Krenkel, E., I91911L—“Geologische 4.-"Gsologische Beobachtungen in ~rbis~h-~st@ika."Br1t1sch-Ost31fr1ka ” N. Jahrb. Min. Geol. Pal.,Pal, 31, pp.243-267.pp 243-267. ‘ “‘ . , 1925.-"~eolo~ie1925.—“Geologie Afrikas."Afrikas.” Vol. I.

MuMufffi (Maufe), H. B., 1908.-"Reportl908.——“Report relating to the Geology of the East African Protectorate.”Protectorate." E.A. Protectorate No. 45, Cd.Cd 3.828,3&B, H.M.S.O.H.M..SO Parkinson, J., 1947.-''Outlipas1947 —“Outlines .dof the Geology of the-MtitotheéMtito Andei-TsmoAndei-Tsavo Area."Area.” Report No.13,GeolNo. 13, Geol. Surv., Keb)ra.Keflya. Richards, C. G., 19501950.-"Krapf,—“Krapf, Missionary aii&EYi8f%%r."and‘ififilfier” East African Literature Bureau. Saggerson, E. P., l957.—“The1957.-"The Geology of the South KituiKi-tui Area."Area.” Report No. 37, Geol. Surv., Kenya. , - - 37

Sanders, L. D., l954.—-“The1954.-"The Geology of thethe Kitui Area.”Area." ReportReport No.No. 30,30, Geol.Geol. Surv.,Surv., Kenya. ——,, l963.—“The1963.-"The Geology of thethe Voi-South Yatta Area.”Area." Report No. 54, Geol.Gwl. Surv., Kenya. Schoeman, J. 1.,J., 1948.—“A1948.-"A Geological ReconnaiReconnaissancefiance of thethe Area West of Kitui Town- ship.”ship." Report No. 14, Geol. Surv.,SUN., Kenya.Kenya, Temperley, B. N., 1956.——“Report1956.-"Report on the Kibwezi-Chyulu 1954-1955 Groundwater Investigation.”Investigation." Ministry of Works, Nairobi, departmental report.report. Turner, F. J., 1948.—“Minera.logical1948.-"Mineralogical and Structural Evolution of thethe Metamorphic Rocks.”Rocks." Memoir 30, Geol. Soc., America. . -0, , v- a- .g.-.-*- Walker, E. E., l902f—f—“Report1902t-"Report on the Geology of thethe East Africa Protectorate}?Protectorate." Africa No. 11, Cd. 1769;1769: H.M.S.O.

.., min} um;

-- G.P.K. 1278 (60)—Irn—3/63(60)-lm-3/63