Early Buddhist Caves at Junnar

Author(s): Vidya Dehejia

Source: Artibus Asiae, Vol. 31, No. 2/3 (1969), pp. 147-166
Published by: Artibus Asiae Publishers
Stable URL: http://www.jstor.org/stable/3249428 .
Accessed: 18/06/2014 12:31

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Artibus Asiae Publishers is collaborating with JSTOR to digitize, preserve and extend access to Artibus Asiae.

http://www.jstor.org

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 VIDYA DEHEJIA
History Department,Universityof Sydney

EARLY BUDDHIST CAVES AT JUNNAR

into the hills and valleysaroundthe smallandratherisolatedtown of Junnarareseven

Cut sets of earlyBuddhistcaveswith a total of some hundredand forty excavations.These
consistof a groupin the Tulji hill, an extensiveseriesin the Shivnerihill,thosein the Lenyddri
range,threesetsintheManmodi
hill,andthecavesattheheadoftheNinighitpasssomeeighteen
milesaway.Describedby Burgessin a monographas earlyas I874, theywerethenincludedin
the Burgessand Fergussonclassicwork TheCaveTemples of India.The earlycavesof western
Indiabelong to the phaseof Sitav~ihana ascendancy,and were all locatedalong the principal
traderoutesof the time,speciallyalongthoseconnectingportsto markettownsof the interior.
Junnarmustat one timehavebeenan importantandprosperousmartalongthe routelinking
the coastaltown of Kalyin throughthe Ninighit pass to the Sitav~ihana capitalof Paithan.
Unlikethe better-knowncavesof westernIndiasuch as those at Bhiji or Kirle, the cavesat
Junnararerarelyvisitedandhaveremainedrelativelyunknown.
We shallexaminethe Junnarcavessite by site, discussingin some detailthose excavations
that appearsignificantbecauseof theirarchitectural
or sculpturaldetails,or becauseof the in-
scriptionsengravedin them.It was thoughtbest to adhereto the modernlocal namesof the
caves-many of whichreflecttheiradaptation to Hinduusage-althoughtheseoccasionally differ
fromthetermsusedbyBurgess.Weshallthroughoutcomparetheseexamplesof rock-architecture
withthegeneraldevelopmentof theseriesof earlyBuddhistcavesin westernIndia.In thecourse
of our analysisof the variousarchitectural
componentsof thesecaveswe often findourselves
facedwithwhatPanofskytermsan "organicsituation",in whichthebeginningof ourinvestiga-
tions seems to presupposethe conclusions. "It is true that the individualmonumentsand docu-
mentscanonlybe examined,interpreted andclassifiedin thelightof a generalhistoricalconcept,
whileat the sametimethis generalhistoricalconceptcanonly be builtup on individualmonu-
mentsanddocuments."Thisis not,however,a hopelessviciouscircleandis indeedtheonlybasis
on whichourtheoryof architecturaldevelopmentcanbe built."Everydiscoveryof anunknown
historicalfact,andeverynewinterpretationof a knownone, will either'fitin' withtheprevalent
general conception, and thereby corroborateand enrich it, or else it will entail a subtle, or even
fundamentalchange in the prevalent general conception, and thereby throw new light on all
that has been known
before."'
The Junnarexcavationsreveala numberof variationsfrom the generalpatternof development
in western India. There are some apsidalvaulted chaityassuch as are standardat most other cave
sites, but equally significantare the flat-roofedrectangularones and the single unique circular
I
Erwin Panofsky, Meaningin the VisualArts, New York 1955, p. 9 f.

147

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
chaitya(fig. I). Some of the Junnar complexes contain more than one chaityaand several of the
caves, including a few chaityas,remain unfinished. The 'blind' chaiya window to be seen on certain
excavations at Junnaris a feature otherwise unknown in the western caves. There are a few
quadrangular of the kindfoundcommonlyelsewhere,but mostof the vihdras
vihdras do not con-
formto anyspecifictype.Thesemanydepartures fromthe typicalpatterncouldbe interpreted
to implyan earlyandexperimental in
stage rock-cutting,whena standardformhadnot yet been
evolved.We shallsee, however,thata numberof the cavesbelongto an advanceddate,when,
presumably, rock-architecturewaswell enoughestablishedfor cuttersto departfromthe norm.

NANAGHAT

A very roughtrack,used normallyonly by mountainponiesand the occasionalbullockcart,

leadsfromJunnarto thesteepdescentof thepassat Nindghdt.At theheadof thepassis situated
an importantinscribedcavewith a seriesof cisternsbesideit. A smallcaveandcisternmaybe
seenon the rock-faceopposite,but the caveis so roughlyhewnthatit couldeasilybe takenfor
a naturalone.It seemspossiblealsoto tracetheexistenceof a few otherexcavations, now sharply
undercutandscarcelyrecognisable, alongthe samescarp.None of thesecaveshasanytraceof
architecturalembellishment. Theinscribedcaveoncecontainedfull-lengthbas-reliefportraitsof
the earlyrulersof the Sitavihanadynasty.Now only tracesof the feet of two of the figuresre-
main,buttheintactinscribedlabelsabovetheirheadsleavelittledoubtof theirone-timeexistence
or theiridentity.The figuresrepresented King Simuka,the founderof the S~itav~ihana dynasty,
King S~itakarni,Queen Nayanikd, a general and
(mahdrathi), threeprinces.The entiresidewalls
of the cavearecoveredwith the inscriptionsof QueenNayanik~i recountingthe generosityof
the royalfamilyat the performance of Brihmanicalsacrifice.We areleft in no doubtas to the
wealthof the Sitav~ihana rulers:asmanyas i i ooocowsandiooo horses,amongothergifts,were
offeredat a singlesacrifice,anda numberof sacrificeswith similardonationsarereported.
Ninigh~itof itselfis aninsignificantplace:a smallset of cavesprobablyusedas a resthouse
atthetopof thesteepandlengthyascent,withno settlementandapparently no monasticestablish-
mentin theimmediatevicinity.It acquiresa positionof consequenceonlywhenconsideredfrom
the pointof view of its strategiclocationat the headof a passalongthe routelinkingthe ports
to Junnarandfromthereto the markettowns of the interior.Yet, it was at this spot thatthe
earlySdtavihanarulersdecidedto engravenotablerecordsandto carveout of solid rockfull-
lengthreliefsculpturesof variousmembersof the royalfamily.This evidenceemphasizesthe
importance placed by the Sitavihanas on the coastal trade and would seem to indicate some de-
finite connection between this trade and their rise to power, and certainly with the wealth of the
dynasty.
The Ninighit records undoubtedly mention the first and third rulers of the Sltavihana
The date of the establishment of the dynasty is a matter of dispute, some scholars
dynasty*.
2 All Purinic accounts agree on the names of the first three rulers of the dynasty - Simuka, Kanha, and Sdtakarni. Kanha,
the brother of Simuka, is mentioned in a record at Nisik. His omission at Ninaghit is probably due to the record having
been engraved by the queen of Sdtakarni I, who was the son of Simuka, and to whom the throne should have passed
directly after Simuka.

148

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
placingthis event around200oo B.C., and othersat c. 28 B.C.3We would like, very briefly,to
examinethe palaeographic evidencewhichthrowsimportantlight on the matter.It appearsto
us thatthe Nin~ighSit records(nos. 1-8) displaypalaeographic featuresthatplacethemslightly
laterthanthe SdfichiII railinginscriptionsandthe Bhdrhiittoranarecord(fig.17).Theyarecer-
tainlypriorto therecordson theeast,southandwesttoranas at Sfi chistapaI, andseemto accord
with the inscriptionson the northtorana.Featuresindicatingthis arethe definitelyroundedga;
the ta whichis not quiteangularbut showsa tendencytowardscurvature;the equalizationof
the verticalsof pa; andthe slighttendencytowardssuchequalization in the caseof sa,haandla.
Theformofchais stillthatof anovalwitha verticalthroughit andhasnotyetreachedthe double-
loopedSafichitorana variety.Theproblemof actualdatesfor theseinscriptionsis a complexone
thatwe havetreatedat lengthelsewhere4.Herewe will stateonly thatwe considerthe Bhirhat
toranainscriptionwhichstartswith the wordssuganam raje,or "in the reignof the Sufigas",to
provideus witha datesomewheretowardthe endof Sufigarule--perhapsc. 80 B.C.It cannotbe
placedearlyin the dynastysince the inscriptiondisplayspalaeographic featuressomewhatin
advanceof the BesnagarHeliodorusrecordthat is securelydatedto c. 120zo-oo B.C.s If it is
acceptedthatthe Bhdrhiittoranainscriptionbelongsto c. 80 B.C., then the N~inghdtrecords
thatbelongto a phasesoon aftermaybe placedat c. 70/6oB.C. Workingbackwardsfromthis
date,andallowingapproximately fiftyyearsfor the rulesof Simuka,Kanha,andSitakarniI, the
establishment of the Sdtavihana dynastymaybe placedaround120/zox B.C.
Theactualcuttingof theN~inghit cavesprobablyoccurredsometimearound70 B.C.,when
presumablyQueenNayanikdcommissionedthe royalportaitsandthe engravingof the records
on the wallsof the maincave.The continuingimportanceof the passsometwo hundredor so
yearslateris confirmedby two inscriptions:one tellsof the donationof a cistern(no. 9, fig.17)
andbelongspalaeographically to the timeof the laterSdtavihana rulers,to eitherGautamiputra
or
SdtakarnmiPulumivi;the otherreportsthe gift of a well duringthe reign of VdSishtiputra
Chatarapana who maybe identifiedas Pulumivi'ssuccessor.
Sdtakamrni,

TULJA LENA

Situateda few mileswestof Junnar,the hill containingthe Tulji cavesis approachedby a cross-
country track,from which a gentle climb leads to the site a
comprising single circular
chaitya,
some ten vihdras,
anda few cisterns.The cavesareall in a badlyruinedcondition,none of the
facadesremainsintact, and the rock is sharplyundercut.
The unique circularchaityaconsists of a chambersome thirty feet in diameterwith twelve
plainoctagonal columns eleven feet high surroundinga centralstipa, above which rises a domed
a There are many advocates of both theories of dating. Perhaps the most outstanding and recent re-stating of these two
points of view are: Prof. G.Venket Rao, "The Pre-Satavahanaand SatavahanaPeriods", chapter II in G.Yazdani (ed.),
TheEarly Historyof theDeccan,Oxford p. for the earlierdating; and D. C. Sircar,"The Satavahanasandthe Chedis",
196o, go
chapter XIII in R.C.Majumdar (ed.), TheAge of Imperial Unity, 3rd ed., Bombay i96o, p. 195 for the late start of the
dynasty.
* Vidya Dehejia, "The Chronology and Development of the Cave Architectureof Western India", unpublished Ph.D. thesis,
CambridgeUniversity, I967, chaptersII and III.
s The inscription tells of the erection of a garu4adbvaja
by Heliodorus, ambassadorfrom Antialkidas to king Bhagabhadra.
The reign of Antialkidasis placed with some certainty between i2zo-i oo B.C.

I49

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
roof(figs.I, z). Therock-cutchaitya thatappearsto be closestin planto Tuljiis thatatGuntupalli
in the Krishndbasin(fig. 3). We findtherea circularchambercontaininga stma, with a domed
rooffilledin withrock-cutribsmeetingat the apex.At Guntupalli,however,thereareno pillars
to markout apradaksfitapatha. The entranceto the cavehasa simplechaitya archaroundit. The
Guntupallichaitya wouldseemto occupya chronological positionsomewherebetweenthe Bard-
barcavesandthetypicalwesternchaitya exemplified by Bhiji. TheBardbar chaityas consistof two
chambers,one circularandthe otherrectangular, excavatedparallelto the rock-face.The small
doorwayopeninginto the rectangularchamberhas a chaityaarchcarvedjust above it. The
Kondivtechaitya (fig. 3) seemsto providethe nextstagein the developmentof the chaitya form.
Here, the two-chamber plan of the Baribar caves is preserved, but the is
cutting perpendicular
to therock-face,withtheentrancealongtheshorteraxisof therectangular chamberasis universal
in the westerngroup.Whenthe superfluityof the innerchamberwas realised,its shapewas
retainedbut its wallsremoved,leadingto a small,pillarless,apsidalcavewith a rock-cutstipa
withinthe apse.The ideaof settingup pillarsto markout a definitepradaksipna pathamusthave
arisensoon after,resultingin the typicalearlychaitya formseenat Bhlaii.The Guntupallichaitya
with its archjust abovethe doorwayratherthanraisedhigh aboveit as at Bhiji, belongsto a
distinctlyearlyphase.The absenceof pillarsis alsosuggestiveof an earlydate.At Tulji,by com-
parison,the occurrenceof pillarsto definethepradaksina pathasuggestsa datenot earlierthan
thatof the Bhj*ichaitya. We cansayverylittleaboutthe completionof the facadeat Tulji. The
lower portionis now completelyopen, while the rock-faceaboveis undercutand no longer
retainsits originalsurface.It wouldappearthatthis circularpillaredchaitya constitutesan early
variationon the apsidalplan6.
ThedomedroofovertheTuljistipais accompanied by a half-arched roof overthepradaksina
patha.This half-arched roof is a featureof somesignificance anda studyof the seriesof apsidal
chaityasin thewesterncavesrevealsthatthehalf-arched side-aisleceilingsaregenerallyindicative
of an earlydate,withflatceilingsreplacingthemat a laterstage.An examination of the Bhlrhit
sculpturesprovides evidence to confirm the earlierdate of the half-archedceilings,as the three
undoubtedchaitya representations at Bhtrhit displaythis feature.The occurrenceof this factor
at Tulji is indicativeof its earlydate.
The Tulji chaityahassimpleoctagonalcolumnswithoutbaseor capital.Althoughit is rather
difficultto judgethe slantof the columnsin this circularcave,theredoes appearto be a slight
inwardrake.Whenwe examinethecolumnsin thewesterncaveswe findthattheearlyoctagonal
columnusuallyslantedinwardstowardsthetop. It hasbeensuggestedthatthiswas in imitation
of wooden forms where such a slant would counter the thrust of the bent wood arches. The
slanting stone column, it is furtherimplied, continued until the rock-cuttersrealisedthat such
strictimitationof wooden forms was both unnecessaryand ungainly.A study of the carvingson
the Bhtrhat and Safichistipasgives us an interestingclue in this context. It is on the Safichitorana
reliefs that we first come across representationsof the chaityaarch with the tie-rod (fig. 4). This
comparativelyminor detail is depicted so preciselyand prominentlythat we may conclude that

6 The Bairit excavations (Stuart Pigott, "The Earliest Buddhist Shrines", Antiquity,XVII, 1943, PP. I-Io) seem to prove
the existence of a circularpillared chaityaas early as the Mauryanperiod. Another indication of an early date for this form
is the Bhdrhutrepresentationof the Sudhamma Sabbawhich depicts a pillared,domed circularchaitya.

150

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
it wasa recentinnovation7.Withtheintroductionof the tie-rodtherewasno longeranyneedto
have slantingwoodencolumnsas the tie-rodcounteredthe outwardthrust.The firstcavesin
whichentirelyperpendicular columnsappearmaythenbe consideredto be roughlycontemporary
or to follow soon after.On this basisthe Tulji chaitya
with the Sfinchitoranas, wouldprobably
belongto a phasejustpriorto or perhapscontemporary with the of I.
torahas Sinchistfpa
The Tuljdstfpa consistsof a simple drum and anda,the harmiki, and chhatrabeing
yas.ti
absent.It hasoftenbeenstatedthatthe formof the stfpais a featurethatmaybe regardedas a
chronologicaldeterminant,"the most ancientbeing a simplehemisphereand the latesta tall
tower surmountedby a dome"8. This is undoubtedlytrue when applied over a wide range of
time,butwithintheperiodof theearlyHinayina-phase excavationsthisfactorprovesunreliable.
A consideration of theproportionsof thestfpaswithinthechaityas
revealsthatin boththeearliest
andthelatest(Kirle),thedrumis low, whilein theinterveningperiod(Nisik)a tendency
(Bh•iji)
towardselongationappears.Wewouldbe on rathershakygroundsin tryingto obtaina sequence
on thisbasis.
Tracesof paintingmay be seen on the pillarsof the Tulji chaityaand one relativelywell-
preservedfragmentdepictsthe figureof a woman.The eleganceandgraceof the pose is remi-
niscentof the Chhaddantajitaka scenesin AjantdchaityaX, but we arereluctantto expressany
positivejudgementon the basisof a singlefragment.No sculpturalembellishment remainson
Remnantsof suchdecorationmaybe seenon two of the vihdras.
the chaitya. Bandsof rail-pattern
connectthe archesaroundthe doorways;andthe decorationabovethisrevealsminiaturearches
with slenderbracketson one vihdra, andstfpasandhumanfigureson the other.It is unfortunate
thatno inscriptionsremainin thissetof caves,asthesewouldhavegivenus somefurtherchrono-
logicalindications.Presentevidence,however,indicatesdefinitelythatthe Tulji chaitya
belongs
to an earlyphaseof rock-cuttinganddatespriorto, or perhapscontemporary with the
torahas
of S-fichistfpaI.

BUDH LENA - MANMODIHILL

Lyingabouta milesouthof the townof Junnar,the Manmodi

hill containsthreegroupsof
cavesthatarenow connectedby a pathwayskirtingthe hill. The BudhLeni groupconsistsof
a singleunfinishedchaityaanda smallset of vihdras.
The exteriorof the chaitya
appearsto have
been completedand to have been of stone, but muchof it is now brokenaway(fig. 5). The
interior,some thirtyfeet deep and measuringabouttwelve feet betweenthe pillars,remains
incompletewitha roughlyblockedout stfpaat thefarend(fig.A).Thechaitya followsthe apsidal
plan and the right side pillarshave been roughly hewn, although we have no indicationof how
they were to have been treated. The right aisle has also been partly completed and the half-
archedceiling has taken definiteshape--a featureindicative of an early date. We have seen that
the typical early western chaityawas an apsidal cave with pillars. In some of the earliestcaves,
while the chaityaarchwas of stone, the rest of the facadewas carvedalmost entirelyfrom wood.
7 It may be noted that even at Sifichi we find one isolated panel depicting the chaityaarch without the tie-rod. This is on the
west pillar of the south torapa- top panel, east face. It seems apparentthat the innovation was a recent one. It is of interest
to note that in the representation of the IndraiYlaguhab on the north torana(east pillar, upper panel, west face) the cave
shown is semi-artificialwith a chaityaarch embodying the tie-rod.
8 A.Cunningham, Bhilsa Topes,London
1854, P. 178.

II

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
The nextstageseesthe woodenfrontagebeingreplacedentirelyby stoneas at the Ndsikchaitya
andat AjantdchaitaIX, andit wouldappearthatthe BudhLendchaitya belongedto thisphase.
The chaitya archon the facaderetainsthe classicshapeseenon otherwesternchaityas,but is
no longera functionalchaiyawindowas it hasnot beencut throughandis of solidstone.This
'blind'arch,seenherefor the firsttime,appearsalso on two otherJunnarchaityas thatbelong
to a slightlylaterphase.This featureis completelyunknownelsewherein the westerncaves,
andseemsto be a regionalvariationof uncertainsignificance. The windowhaslost its original
purposein whichthelightfilteringthroughfell directlyon theobjectof veneration-thestfpa-
andhasbecomea decorativefeatureinstead.TheuniqueBudhLendfacadecontainsa halflotus
with sevencarvedpetalscontainedwithinthe 'blind'arch(fig. 6). The themeappearsto be of
Gajalaksmi, the figurein the centralpetalwith her left armakimboandthe rightraisedin the
gestureof assurance-aswascommonin earlyIndianiconography--being Theelephants
Laksrini.
on eitherside of her standon halflotusesandhold water-potsin theirraisedtrunks.The re-
mainingpetalscontainthe dancingfiguresof Yaksasand Yaksinis,with their handsjoined
togetherabovetheirheads.The conceptionof the themeis ratherunusual,and the assurance
with whichthe artisthashandledhis spacewouldimplyan advanceddate.The figuresremind
one stronglyof the couplesembellishing theverandaof the Kdrlechaitya. Evenmoreimpressive
arethetwo heraldicfigures(Ndgas?)flankingtheapexof thechaitya window.Bothrevealsimilar
proportionsandthe samepowerfulmodellingas the malefigureson the Kdrleverandapanel.
Thechaitya containsa singleinscription,engravedon the centralflatsurfaceof the halflotus
and recordingthe donationof the facadeby a Yavana,a foreigner.The inscriptiondisplays
palaeographic featuresthatappearto placeit withthe Kdrlespecimens,andaredistinctlyearlier
thanthose of the timeof Nahapina(alldatedbetweenthe years40 and46). We wouldthusbe
inclinedto assigntheKdrlerecordsandtheBudhLendinscriptionto aroundA.D. 50-70. Those
who considerNahapina'sdatesas referringto the Sakaeraof A.D. 78 ratherthanto a regnal
year,as we do, mayaccordinglyadjustthesedatesby aroundthirtyyears9.
The sculpturaldecorationof the vihdraadjoiningthe chaitya is mostlyintact.A rail-pattern
bandconnectsthe archesover the doorways,and above this is a row of stipasand miniature
arches(fig. 7). The BudhLendvihdras appearcomplete:an indicationthatthey wereprobably
excavatedpriorto thechaita,whichseemsto havebeenabandoneddueto a badstratumof rock
andan extensiveleakageof water.Thissmallset of caveswouldappearto havebeenexcavated
roughlycontemporary withthe Kdrlechaita,andprobablybelongedto the periodA.D. 50-70.

AMBA/AMBIKA - MANMODI HILL

The Ambi/AmbikBgroup containsone unfinishedchaityaand six vihdras.The apsidalchaityahas
a small verandain front with two full and two engaged columns (figs. I, 8). This cave furnishes
9 It is our opinion that Nahapina's dates do not referto the Sakaera, but to a regnal year as is universally accepted to be the
case with the later Sdtavdhanadates. We would basically follow the arguments put forward in the editorial article "The
Date of the Karle Chaitya", Lalit Kala, 3/4, 1956-57, and place Nahapdna'sreign from A.D. 54-Ioo. GautamiputraSdta-
karni,who was in power for some ten yearsafterdefeating Nahapdna,would then have ruled until A. D. 1lO.His successor
Pulumdvi, who ruled for twenty-eight years, accordingly belongs to A.D. 10o-138.According to the theory that assigns
Nahapdna 's dates to the Saka era of A.D. 78, his last known year of 46 results in a date of A.D. 124 for the end of his
reign. Gautamiputrawould then have ruled until A.D. 134 and Pulumdvi would date between A.D. 134-162.

152

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
us withournext stagein the developmentof thewesternchaitya, in whichthe simpleapsidalcave
witha stonefacadewaselaborated throughtheadditionof a veranda.Thetypeof verandavaried:
at somesitesit was a smallneatstructurewith no sculpturaldecoration;whileat othersites,as
at Kdrle,it was deepandwideandlavishlydecoratedwith carvings.
The Ambi/Ambikdverandapillarsrise out of a ghata(water-vessel)on a steppedbaseand
terminatein aninvertedghata, a steppedabacus,anda rectangular blocksupportingthe entabla-
ture.This is an advancedandhighlydevelopedformof column.The straightoctagonalpillar
of the westerncavesfirstacquireda baseconsistingof aghataon a steppedplatform,anda rudi-
mentarycapitalin the formof a rectangular blocksuchas we see in the Ndsikchaitya.
The fully
developedpillar in the was
chaityas achievedwhen an elaborate capitalwas addedconsistingof
a "bell",an enclosedimalakaro, a steppedabacus,andcrowninganimals.The chronologicalde-
velopmentof thesestagesis apparent.A numberof variationsmay,however,be seenamongthe
fullydevelopedcolumns,andan analysisrevealsfour significantcategoriesbasedon changesin
the depictionof the "bell"capital(fig. 9).
I In the firstphasethe "bell'is noticeablyincurving,almostwaisted,andhasdistinctlydepicted
petalsas at Bedsd.
2 In the next phase the petals are no longer clearlydistinguished,but are depictedas ridges on
the "bell".Thisphaseis evidenton the pillarswithinthe Kdrlechaitya.
3 The thirdphasesees the disappearance of the petalmarkingsthoughthe slightlyelongated
"bell"shape is retained. Due to the absence of petalmarkingswe now describethis formas
an inverted This is evidenton the pillarsof vihiraX at Ndsik.
gha.ta. phase
4 The finalphaseseesthe losingall hintof its original"bell"formandacquiring
invertedgha.ta
a distinctlyroundedshapewith the neckof the It is as if the base
gha.taveryclearlydepicted.
wereinvertedandreduplicated at the top of the column.Pillarsof this categorymaybe
gha.ta
seenin Ndsikvihdra III.
The Ambd/Ambikipillarshave a distinctlyroundedinverted and to the final
gha.ta belong
phaseof this development.
The Ambd/Ambikd windowis locatedon the innerwall of the verandaandis a true
chaitya
one.It is of thetraditional
shapealthoughthe actualopeningis in theshapeof a narrowhorseshoe
ratherthantheusualsemicircle(fig.Io). Theinteriorhasbeenblockedout into apsidalshapeand
the stfipahas been roughlyhewn. The pillarshave not yet been commencedand the ceiling,
apparently meantto be vaulted,is incomplete.The cavemeasuresapproximately twenty-sixfeet
to the frontof the stipa,andthirteenfeet across.
The AmbB/AmbikBchaityaappearsto belong to an advancedstage of development.The cave
has acquireda verandawhich is to be seen for the first time in the western series at Bedsl. The
pillarsreveal that the cave belongs to an even more advancedstage, definitelylater than Klrle.
The inscriptionsconfirmthis as all the recordson the chaitya(nos. 2o-30, fig. 17) belong palaeo-
graphicallyto the time of Pulumivi, or roughly between A.D. I ". These recordsare en-
zo-138
graved on the main wall of the verandaand even on the chaityaarchitself. Record 17, engraved
1o The imalakarefers to the flat ribbed cushion-like member that appearsbetween the "bell" and the stepped abacus(see figs.
13,I4). It usually appearsto be enclosed within a frame or sometimes merely between two slabs of stone.
x Those who follow the Gakaera dating of Nahapina may accordingly adjust the dates of the later Sdtavahanarulers.

153

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
on a vihdra,alsobelongsto thispalaeographicphase.TheotherAmbi/Ambikirecords--nos.18,
19, 3x--are to be found on vihdrasand displayfeaturesthat assignthem to an earlierphase,
probablyto the timeof Gautamiputra Sdtakarnmi(A.D. 0oo-i o). It wouldappearthatmostof
thevihdras werecutbeforethechaitya,andthisis indeedasmaybe expectedin view of thelatter's
unfinishedcondition.Workon thevihdras certainlywouldnot havebeencontinuedonceexcava-
tion of the chaitya
wasabandoned.The one vihdra thathasa recordbelongingto the samephase
as the chaitawas presumably begunat the sametimeas the chaitya.Beingsmallerandarchitec-
turallylessadventurous,it wasno doubtcompletedbeforeexcavationof the chaitya hadreached
the stagewherethe naturalfaultin the rocktowardsthe backof the cavewas discovered,and
it wasrealisedthatcuttingwouldhaveto be abandoned.Amongthe vihdras is one thatconsists
of sevencells,placedsideby sideandopeninginto a commonveranda.

BHIMA SHANKAR- MANMODIHILL

TheBhimaShankargroupconsistsof a singlechaitya andsomesixto eightvihdras.The veranda

of the chaitya
containstwo full andtwo engagedcolumnsrestingon a balustrade.The capitals
of thesepillarsincludea roundedinvertedghatathatbelongsto the finalphasein the develop-
mentof the "bell"capitalandis indicativeof an advanceddate.Abovethe invertedghatais the
stepped abacus, with a small rectangularblock supporting the entablaturewhich is carved in
imitationof a rail.The facadeof the chaitya
onceagaincontainsa 'blind'arch(fig.i1). The arch
haslost its functionaloriginandappearsto havebecomea meretraditionalmotif.Buthereeven
the classicshapeof the archhas been abandonedandit consistsof a simplesemicirclewith a
horseshoeexcavatedwithinit in fairlydeeprelief.Thearchis locatedhighup on the rockabove
the exteriorwall of the verandaso thatto be functional,an excessivethicknessof rockwould
havehadto be cut away.
Whatwe haveseenof verandaandfacadewouldleadus to expecttheinteriorto be anapsidal,
vaultedstructure, similarto theAmba/Ambikd Butthisdoesnotappearto be so. Although
chaitya.
partlyincomplete,excavationhasreacheda stagefromwhichit is possibleto statethatthe cave
was apparentlymeantto follow a differentplan (fig. ). The interiorwas not designedto be
apsidaland pillared,but as a long, narrowrectanglethat measuressome thirty-threefeet by
twelvefeet. Theroofwasnot to be vaultedbutflat,andBurgesstellsus thatthereweretracesof
plasteron the ceiling-tracesno longerin evidence.The stzpawasleft at the farendas an enor-
mous massof rock,and this has recentlybeen carvedroughlyto representa squattingimage
worshippedlocally as BhimaShankar.A well at the far end of the cave presumablyexplainswhy
the excavationwas abandoned.Although the interior is treated quite differently,we would be
inclined to state, on the basis of the form of the "bell" capital,that this cave probablybelonged
to the samephaseas the Ambi/AmbikBchaitya,andhenceto the time of Pulumivi. A veryfragmen-
tary recordengravedon the rock-faceto the immediateright of the Bhima Shankarchaitya(no.
34, fig. 17) seems to confirmthis, as it revealspalaeographicfeaturessimilarto those in the re-
cords of Pulumivi.
The vihdranext to the chaityais of ratherunusual plan. A verandawith two full and two
engaged columns leads not into a hall as one might expect from a study of the vihdrasin the

I54

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 0
I.SHIVNERI XLVIII 2. SHIVNERI I

0000902

-0
O0 ?0
o
0o F
o
OKI0q
3. LENYADRI VI 4. LENYADRI XV

o o -
---- -- -
/ ,
o
oO
0
o
o
/" -

5. TULJA 6. MANMODI- BUDH LENA

I (
;

7. MANMODI - AMBA / AMBIKA 8. MANMODI - BHIMA SHANKAR

Fig. I Junnar ChaityaFloor Plans (to relative scale)

(All photographs by author)

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 Fig. 2 Tuljd Lend,circularchaitya

Fig. 7 Budh Lend, viharas

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 Fig. 6 Budh Lend, chaityafagade, detail

Fig. 5 Budh Lend, chailyafagade

GUNTUPA L LI
KONDIVTE

Fig. 3 ChailyaFloor Plans: Guntupalli and Kondivte (to same scale as fig. i) Fig. 4 Sdfichitie-rod

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 Fig. 8 Ambd/Ambikd,chaitya

Fig. 9 Stages in the development of the "bell" capital

Fig. io Detail of Ambd/Ambikdchaityawindow

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 Fig. i i Bhima Shankar,chaitya

Fig. I2 Bhima Shankar,vihara

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 Fig. 13 Lenyddri,chaityaVI fagade

Fig. 14 Lenyddri,chaityaVI: interior pillars

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 Fig. I5 Lenyddri,Ganesh Lend, vihira

Fig. 16 Shivneri, chaityaXLVIII: painted ceiling

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 NASix NANC4 SAI mix NA HAPANA Gau M
E A - *.AC w . . KA. ...,AR *, ss4..
JUM,.an4 J1,nr,
,,.jAsMr
•,MIPAA s*,ses.
SA•u,
RAmuc* s SseARv. '
j~as **AN e.,c
I?1?
...
S',.=,,
14wNAsaa .6 KAILs Junrra* NAo. KmALs PUL.U K(
NAsen
,=A
f I SwU o t"
SL L L L L LL
C o
e
o O0
k. + + + + + + t + t++-Ytt
1 1 3 a
,- s I Al In .,,..
lnA n
u n A n 0
u u2
~,..., nn~/
C A 9(1 W I A 1 A1 I A U) AAAAAA j
d LJ cl 0wI ca. C0 U) 18 C U i3W~
? id ddd d dd d d- d d dd d dd dd d
jE EE EE E6 E6IE/ E E

CaC & AC C
bk jO O O o

I
i
I I
Ie ple I IV, L
AAA ; AA A AA
Ib l h bbiA
h l Ad z88 .
$tA A d
i I
rr 1
L
? I LI r Z
--? I i JI I I J
t~~rjO I I I I lL s L-) (1 0
rZ

UAu ob u
up a tou uu
uuo Io
po A
P

0 O 0 OO0 cOO

r. Ob
Idoo :
1
1 -L
I
J-01
I
V) I'
d I I I)e 2 1r8 e J

b b6 bblb
6 bb~b 6
b 6abbb
-

bA ba6
46 & 6 Od

table.Numberingof inscriptionsas in J.Burgess," Reporton the BuddhistCaveT

palaeographic
Fig. I7 Comparative
plesandtheirInscriptions",Archaeological ofWe~stetrn
Surry~ India,IV, Londonr883-

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
westerncaves,but directlythroughthreedoorwaysinto threecellsplacedsideby side(fig. 2).
The pillarshavean inverted similarto those of the adjoiningchaitya, andwe mayassume
gha.ta
thatworkon the vihdra andchaitya was commencedsimultaneously, with the smallercavebeing
completedbeforethe discoveryof the wellwithinthe chaitya led to its abandonment. TheBhima
Shankargroupcontainstwo otherinscriptionsof whichone, mentioningthe gift of a cistern
(no. 33),belongswiththechaitya recordto thetimeof Pulumavi.Theotherinscription,engraved
besidea cisternto the far eastof the chaitya,recountsthe gift of a hall and cisternby Ayama,
ministerof mahdksatrapa Nahap~ina in the year46. This recordwouldindicatethatworkat the
site commencedsometimepriorto excavationof the chaitya.

LENYADRI

Situatedaboutthreemilesnorthof Junnar,the Lenyidricavesarereachedby a seriesof built-

up stepsfromthe roughtrackthatleadsto the foot of the hills.Therearesomethirtycaveshere
includingan apsidal-ended a rectangular
chaitya, flat-roofedchaitya,
a largequadrangular vihdra,
anda numberof cisterns.
ChaityaVI is an apsidal-ended cave measuringapproximately forty-onefeet by twenty-two
feet, with a narrowverandain frontcontainingtwo full andtwo engagedcolumns(figs. 1, 13).
The verandapillarsas well as thosein the interior(fig. 14)risefroma on a
gha.ta steppedbase,
andterminatein aninvertedghata, anencloseddmalaka, a steppedabacus,andcrowninganimals.
The is distinctly"bell"-shaped andwouldbelongto the thirdphasein the developmentof
gha.ta
the "bell"capital,andon this basisto a datesomewhatlaterthanKirle.
The aislesof the chaityahavehalf-arched ceilingswhichsuggeststhatthe cavebelongsto an
earlydateandis perhapsearlierthanthe Kirle chaitya wherethe side-aisleceilingsareflat.The
ceilings of both nave and aislescontainstone ribbing.This featureis often consideredan indi-
cationof anadvanceddate.In accordance withthe view thatthe cavesreveala gradual"emanci-
pation"froman imitationof woodenconstruction 1, it mightbe expectedthatthe use of wood
in the caveswould decrease,andthatthe earliercaveswouldhavewoodenribbingwith stone
ribscomingin at a laterdate.However,whilethe Kdrlechaitya haswoodenribsin the vaultof
its nave,the smallanddistinctlyearlychaitya at Guntupallihasstoneribs.Moreover,withinthe
samechaitya, as at PitalkhordIII andAjantdX, bothtypesof ribsmayoccur-with stonein the
sideaislesandwood in the nave.Thenagain,whilethe Kirle stipahasa woodenchhatra, earlier
stipas such as that at BedsHhave stone chbatras.Except in the treatmentof the facade,the actual
use of wood in the constructionof the caves is not a factor of chronologicalsignificanceand the
stone ribbing at Lenyidri is not necessarilyan indicationof its later excavation.
The Lenyidri facade has a 'blind' chaityaarch with a rail-patternband engraved just below
it. The fronton of the arch is barely visible due to the extremely worn condition of the rock
surface.It is not possible to explainaway this arch as unfinished,as the surfacewithin the semi-
circle appearsto have been smoothed, and this would not have been done if deeper cutting was

1z S.K. Saraswati,"Art", chapterXX in R.C. Majumdar,(ed.) TheAge of ImperialUnity, speaks of"the gradualemancipation
from wooden conventions", p. 499.

163

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
intended.In anycase,to be functional,the thicknessof the rockto be hewnawaywouldhave
to be equalto the entirewidthof the verandaandmore.Thisis no longera truechaitya window.
A consideration of thevariousarchitectural featuresof chaitya VI seemsto implythatthecave
was excavatedeithercontemporary with or soon afterthe K~irlechaitya. It is not possibleto be
more categoricalon architectural groundsalone,since the half-archedaisle ceilingsimplyan
earlierdatewhilethe"bell"capitalbelongsto a laterphase.Inscriptionno. xx (fig. 7), engraved
abovethedoorwayleadingintothechaitaandreportingits donationby aninhabitantof Kalydn,
belongspalaeographically to a phasedefinitelymoreadvancedthanthe Kirle records.A consi-
derationof this evidence,togetherwith the architectural, would lead us to suggestthat the
Leny~idri chaityawasexcavatedsoon afterK~irle--perhaps aroundA. D.9go/xoo.
Chaitya XV consists of a veranda with two full and two attachedcolumnsleadinginto a
rectangularhall measuringapproximately twenty-twofeet by thirteenfeet (fig. I). The stfipa
withinhasits chhatra engravedon the flatroof above.At somestagein the developmentof the
chaifya formwe comeacrossthe introductionof the rectangular usuallyconsistingof a
chaitya,
flat-roofedhallcontaininga stfpa,with a verandain front.Thesecavesno longerhavethe tra-
ditionalchaitya windowon theirfacade.Thechronologicalpositionof suchaninnovationis not
immediately apparenton a consideration of planandfacadealone.It is unfortunate thatthepillars
of chaitya XV areno longerintact:onlythe steppedplatformof the baseandthe steppedabacus
of the capitalremain.Inscriptionno. I2z(fig.17) engravedon the wall just outsidethe chaitya
mentionsits donationby a goldsmithfromKalyin.Palaeographically, the recordbelongsto the
timeof Pulumdviandis indicativeof the lateexcavationof chaitya XV.
VihdraVII, known as the Ganesh Lend, has a large veranda with six full andtwo engaged
columnsrisingfrom a balustrade(fig. 15). Thesepillarsare similarto those of the adjoining
chaitya VI, with an inverted"bell"-shaped gha.ta belongingto the thirdphase.Threedoorsand
two windowsopeninto a largehall,approximately fiftyfeet square,thathasa totalof nineteen
cellsopeningout fromit. The caveis similarto vihdras X andIII at Ndsikandis one of the few
quadrangular to be
vihdras found at It is
Junnar. uncertainwhethertherewas once a bas-relief
stfpa on its back wall such as existsat Nisik: this mayperhapshavebeencontainedwithinthe
centralcellthathasnow beenappropriated for theworshipof Gane'a.Mostof the othervihdras
at Lenyadribreakawayfromthe quadrangular planandconsistof rowsof interconnected cells
with stone beds in each,these cells being furtherlinkedby windows.Occasionally,the cells
containmorethanone stonebed,andone vihdra revealsas manyas fivein eachchamber.
Besidesthe inscriptionson the two chaityas, the Lenyidrigroupcontainsfourotherrecords,
all of whichtell of the gift of vihdras andcisterns.Inscriptionsio and14 (fig. 17)areof the same
palaeographicphaseas record xx on chaityaVI and are distinctlypre-Pulumivi,belonging prob-
ably to aroundA.D. 90/Ioo. Records 13 and I5 displayfeaturesthat assign them to the time of
Pulum~ivi.

SHIVNERI

The upperpartof the Shivnerihill seems to consist of a sheerrock-face,andseen from a distance

the caves appearinaccessible.There is, however, a steep and difficulttrack leading up to this
group of some sixty caves that includes two chaityas,several vihdras,and numerous cisterns.

164

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
Chaitya XLVIIIis a rectangular structure measuring thirty-one feetby twentyfeetandconsists
of avestibule separated fromthemainhallbytwofullandtwoengagedcolumns(fig.i). Towards
thebackof thehallis thestRpa withits chhatra engraved on theflatceilingthatrisesto a height
of eighteenfeet.Onedoorandtwo windowsopeninto thevestibuleof the cave.Thepillars
areof thefullydevelopedvarietyandhavea roundedinverted ghataof thefinalphase,witha
rectangular blockabovethestepped abacus to support theroof.Theceilingstillretains itspainted
decoration in theformof circleswithinsquares (fig.16)in orange,yellow,red,whiteandblack,
andtherearetracesof painton thestfpaaswell.Anintriguingfeatureis the rowof notchesin
thefloorbetweenthestypaandthevestibule pillars:perhaps thesewereintended to holdwooden
postsfor some purpose now unknown. On the basis of theform of the it is evident
pillarcapitals,
thattheexcavation of thechaitya belongs to an advanced date. no.
Inscription 4 (fig.17),cutin
therockto therightof thechaitya doorwayandrecording its donationby a merchant, belongs
palaeographically to the timeof Pulumiviandconfirmsour conclusionregarding the later
excavation of thecave.
CaveI is a smallchaitya consisting of a narrowveranda withtwofullandtwoengagedcol-
umnsfromwhicha singledoorwayleadsinto a chamber,approximately twentyfeetsquare,
thatcontainsonlythedrumof a stfia (fig.i). It wouldseemthatthestgpawaspartlystructural,
asthesmoothsurfaceof thetopof thedrumrulesoutthepossibility of breakage.Thecolumns
havetherounded, inverted ghataof thefinalphase,andthesewouldassigntheexcavation of the
caveto thesamephaseasthatof chaitya XLVIII.
CaveXLVIIis a vihdra withfivecellsopeningoutfromits quadrangular hall.Againstthe
backwallof thecaveis a simplebas-relief stRpa. The front of the caveis now completely open.
Theceilingandwallsof thecaveseemto havebeenpaintedin a stylesimilarto thatof thead-
joiningchaitya XLVIII.A greatvarietyof vihdra plansmaybe seenin thisgroup.Apartfrom
the quadrangular we findseverallargeopencells,a numberof smallchambers,
vihdras, a few
hallswithbenchesrunningalongthesidewalls,andsomelongverandas withcellscutsideby
sideintothebackwall.Thereis alsoa two-storeyed vihdrain whichthelowerstoreyconsistsof
an openverandaleadingintoa hall,approximately twenty-four feetby twenty-two feet,with
ten cellscutintoitsbackandsidewalls.Theupperstoreyis an emptyhall,twenty-one feetby
elevenfeet,whichaccording to epigraphic evidencewasintendedasa reception room.
TheShivneri groupcontains nineinscriptions including theoneonchaitya XLVIII(nos.I-9,
fig.17),allof which belongpalaeographically to the time of Pulumavi. Theentiresetof cavesat
Shivneri may thusbe assigned to thisadvanced phase.

RtSUME~

The excavations at Junnar appear to range over a considerable period of time. Among the earliest,
following immediately after the reign of Sitakarni I, the third ruler of the Sitavihana dynasty,
were the caves at the head of the Ninighit pass. On paleographic grounds we have placed the
Ninighit records at around 70 B. C., and, accordingly, the excavation of the caves and cisterns
must belong to the same date. The next group, consisting of those in the TuljHhill, were excavated
soon after 70 B. C., and certainly well before A.D. o; more likely between c. 70-40 B. C.
5

165

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions
 Thenextphaseof activityis apparent in theBudhLenigroupin theManmodi hillwhere
theevidence pointsto thegroup'scontemporaneity withKirlearoundA.D. 5o-70.Wearethus
confronted witha gapof almosta hundred years,betweenTulj!andBudhLen!,duringwhich
no excavation tookplace.A similarpauseis alsoevidentatothercavesitesin westernIndiaand
appearsto coincidewiththetemporary declineof theSitavihana power,whenthedynastysplit
intoseveralbranches rulingoversmallareas.Thisabsenceof architectural activitywouldseem
to implythatexcavating andvihdras
chaityas depended to someextenton politicalstability. It is
clearfromtheinscriptions thata number of cavecomplexes weredueentirely to themunificence
of wealthytraders andartisans. Thisis especially trueof Junnarwherewe finddonations from
severalhouseholders andmerchants, froma goldsmith andfromguildsof corndealers, bamboo
workers, and braziers.Bycomparison there is onlyone donation from the the
nobility; giftof
a vihdraandcisternby a ministerof kingNahapina. Presumably anunstable politicalsituation
wasnot conduciveto prosperous trade,nor,consequently, to the givingof generousgrants
towards theexcavation ofmonastic establishments. Infact,thisseemstobeageneral phenomenon
withBuddhist establishmentsthroughout historyin India,principally becausethemonasteries
depended heavilyon the patronage of traders. This was also trueof Central Asiawheretrade
flourished onlyunderstablepoliticalconditions, andthe fortunesof monasticestablishments
waxedandwanedaccording to thevagariesof localpolitics.Indeed,the studyof therelation
betweenthehistoryof Buddhist monasteries andof commerce wouldbe a fascinating subject.
Presumably, whenpoliticalconditionswereonceagainfavourable, buildingactivitywas
resumed in westernIndia.AtJunnarworkshiftednextto theLenyidrihill,wherechaiyaVI,the
GaneshLenivihdra, andcertainotherswereexcavated sometimeafterKirle andBudhLeni
aroundA.D. ioo. At aboutthe sametimesomeof theearlyvihdras in theBhimaShankar and
Ambi/Ambikigroupsin theManmodi hillwereexecuted. Evidencefordatingthisactivityis
entirely
palaeographic, withtheinscriptions of Nahap~ina's ministerbesidea cisternintheBhima
Shankar group,andtherecordsbelongingto thetimeof Gautamiputra Sitakarni in theAmbi/
Ambikigroup.Thefinalandthemostprolificphaseof rock-cutting atJunnar is attestedpalae-
ographically by records belonging to the time of Pulumivi. This activity,which woulddate
roughlybetweenA.D. i io-138,includestheexcavation of theBhimaShankar andtheAmbi/
Ambikichaityas, withsomevihdras inbothgroups;chaitya XVandsomeothervihdras atLenyidri;
andtheentiregroupof somesixtycavesintheShivneri hill.TheJunnar cavesbelongentirely to
theearlyphaseof Buddhism in westernIndiaandshowno signsof havingbeenreoccupied in
laterMahiyinadays.

I66

This content downloaded from 185.2.32.152 on Wed, 18 Jun 2014 12:31:06 PM

All use subject to JSTOR Terms and Conditions