A Dhyāna-Yoga Maheśamūrti, and Some Reflections on the Iconography of the Maheśamūrti-

Images
Author(s): Krishna Kumar
Source: Artibus Asiae, Vol. 37, No. 1/2 (1975), pp. 105-120
Published by: Artibus Asiae Publishers
Stable URL: http://www.jstor.org/stable/3250215 .
Accessed: 10/06/2014 16:47

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 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
 KRISHNA KUMAR

A DHYANA-YOGA MAHESAMURTI,
AND SOME REFLECTIONS ON THE ICONOGRAPHY
OF THE MAHESAMURTI-IMAGES

he Mahesamurti"is symbol and image, linga and murti,in one; the concrete form of the
SupremeSiva fully manifested"., Having relativelyearlierbeginnings,it graduallybecame
a popular cult-icon of the Lakullsa-Pasupatasect in the mediaevalperiod,2by which time quite
a few varietiesof thisimagecameinto existence.Butamongthemno imagein dhyjdna-yoga-
attitudehaspreviously beennoticed.

DHYANA-YOGA
MURTI
As is gathered, the large sculpture carved with the Mahesamurti-imagein dyadna-yoga-pose
illustrated and discussed here was found during a site-clearancedone by the Department of
Archaeology and Museums, Governmentof Rajasthanin the surroundingsof the MataTemple
at Nimaj in District Pali of the state: (Plate I).3 This buff sandstone sculpture seems to have
onceembellished
themula-prasdda
of thatshrine.
The Mahesamurti-imagewas once flankedby attendantsplaced in separateniches. While the
figure on the right is gone, the female attendanton the left in tribhanga-posture
holds a hanging
endof drapery
in herlefthand:(PlateI).
The three-headedand six-armedfigure of Mahesais shown seated in padmasanaover a full-
blown lotus: (Plate II). There are indications that the middle head representing Sadyojata-
Nandivaktra
wasoriginally
decorated
withstrandsof hairformingloopsin tieredfashion,but
thejata-jta together with the face is now almost completely destroyed. The deity wears be-
a pairof pearl-necklaces,
jewelledear-rings, a sacredthread,a plainwaist-band,
two pairsof
armlets,anda pairof bangles.Mahesa'spalmsrestin dhyana-mudrd.
Theleft headof Aghora-Bhairava,
hasa skull-ornament
in the loopedskeinsof hair.A
ribbonof skullsaroundthe forehead,the makara-kduyala
earring,the emaciated
face,the
curved strands of the beard, the gaping mouth, and the protruding eyes of the deity express
fury. The terribleaspect is furtheremphasizedby the skull-bowl and the dagger held in left and
righthands.It spellsthedoomof Creation.
The right head of Vamadeva-Umavaktrashows the beautiful, tender face of a woman
wearingan elegantcoiffuredecorated
witha coronet,a patra-kutydala
or circularearring,and
I Stella Kramrisch, "The image of Mahadeva in the cave-temple on Elephanta island", AncientIndia, 2, 946, p. 7.
2 I have dealt with the association of Mahesamurti-imageswith the Lakulisa-Pasupatasect in my article, "The Mahesa-
murti-images of Ellora-Their typology and cult-affiliation", awiting publication in VishveshvaranandIndologicalJournal.
3 I am indebted to Mr. Michael W. Meister for this information, who wishes to thank the personnel of the Department of
Archaeology, Rajasthan,for their assistance during his visit to the site.

I05

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
several wristlets and rings. She is absorbed in tenderly looking in the mirrorheld in her left
handandapplying,like a DarpanaNayika,vermilionon the forehead(or puttingkohl in her
eyes)with a stickin the rightone.4
Notwithstandingits damagedcondition, enough is left to attemptan evaluationof the artistic
meritsof this work.The hieraticfrontalityof the centralfigureis relievedby the profilesof the
two lateralheads, and its static dispositionis balancedby an easy, but gracefulmovement of the
additionalarmsand the pairof Vidyadhara-angels hoveringat the top. Despiteeffortson the
partof the artistto renderthefeaturesof the variousfacesexpressiveof theirrespectivemoods,
the stolidity in modelling and treatmentof the ornamentsis symptomaticof mediaevalism.The
sculptureis executed in high relief, attaininga sort of tri-dimensionality,which is deepened by
the colonnets,producinglight and shade.Besidesservingiconographicneeds, the beautiful
attendantfigure provides a pleasing contrastto the composition as a whole: (Plate I). Her calm
and tenderface as well as the easy movement of the limbs convey a distantecho of Gupta art.
While Sadyojatais in deep contemplation,the Vamadeva and Aghora aspects remain en-
gaged in their eternal function of creation and destruction. The first is passive, the other two
can be said to combinethe featuresof
are active and dynamic.As the Mahesamurti-image
Ardhanarlsvara-Siva,as well, it symbolically represents the union of the two fundamental
entities, namely, Prakrtiand Purusa,which embody Creation.sIn a sense the Mahesamurtiis a
Trinity.It representsa synthesisof the threestatesof feeling
Sivaiteversionof the Brahmanical
or rasas,namely,santa,srngara,and raudra,or peace,love and terrordenoting the threefundamen-
tal aspectsgunasof the realityof Nature:sattva,rajas,andtamas.
In respectof the position of the Vamadevaand Aghora heads, the sculptureis comparableto
two Senaperiod images-one preservedin the Museumof BangiyaSahityaParisad(ca.eleventh
century)6 and the other in the PennsylvaniaUniversity Museum (ca.twelfth century).7Our
sculptureon the otherhandis in theClhamanaidiomof the Maha-Maru styleof thelaterhalfof
the tenth century.
Although other asana-typeimages of Mahesaare known, they differfrom our sculpturein
one respect.There,Tatpurusais in the centre,the left headrepresentsVamadeva,andthe right
one, Aghora.One such three-facedand four-armedfigureof Mahesaof the laterpartof the
eighth centuryis executed in low relief on the plinth of the northerncorridorof KailasaTemple
at Ellora: (Plate III).8 One three-faced and four-armed sculpture from Pandrethan, Kashmir
4
Epigraphic records point to some ointment and vermilion as indispensable toilet articles for a married woman. The
sculptures as well as the paintings reveal that vermilion (sindbura) was applied by the damsel with the right hand, while
she held the mirror in the left hand. Cf. K. Oberoi, "Toilet in Ancient India", Studiesin IndianHistoryand Culture,Dhar-
war, I97I, p. 507-
s T.A.G.Rao was the first to identify the so-called Trimuirtiof Elephanta with the Mahesamurti-image of Siva. Cf.
Elementsof HinduIconography, Madras, I9I6, II, ii, pp. 382-3 85. Subsequently E. B. Havell contradicted the view of Rao
and asserted that it is the Hindu Trinity comprising Parvati (taking the part of Brahma), Visnu and Siva. Cf. A Hand-
book of Indian Art, reprinted Varanasi, 1972, p.i88. While there is no doubt about the correctness of Rao's view, the
credit, however, goes to Havell, who had for the first time recognised the Parvati aspect, which James Burgess earlier
took to be Visnu. Cf. James Fergussion and Burgess, The Cave Temples of India, reprinted Delhi, 1969, p.468. In her
masterly treatment of the sculpture, Stella Kramrisch has identified it as Mahadeva with Uma on the left. Cf. Ancient
India, 2, p. 7.
6 M.
Ganguly, Handbook to thesculpturesin Bangiya Sahitya Parisad, Calcutta, I922, pp. 70-71, pl. XV; Museum no. GCT(a)/284.
7 Ananda Coomaraswamy, "Some Indian sculptures in American Museums", Rupam, 18, I924, p.66, fig. 4.
8 Ellora is distinguished for a unique group of ninteen Mahesamurti-imageslocated in various cave-temples. I have
studied them separately. See note 2.

io6

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
(ca. ninth century) is now housed in Shri Pratap Singh Museum at Shrinagar:(Plate IV).s
Another four-faced and eight-armed Sena image datable to about the eleventh century is pre-
servedin the BangiyaSahityaParisad.Io
Two four-headed and multi-armedseated icons of Mahesa (ca. eighth century) have been
located-one at Melcheri near KaveripakkamIIand the other at Gokak Falls in District Bel-
gaum.,2 Although the published illustrationsof the two images help us neither in the proper
appreciationof the physiognomic features of the two lateralfaces nor in the identificationof
theirattributes,the middlehead with a terrorinstillingface is undoubtedlythat of Aghora.
Most probably the lateralfaces representTatpurusaon the left and Sadyojataon the right side.
From the foregoing discussion, it is evident that the extant seated-typeMahesamurti-images
are divisible into three groups: first, with Tatpurusain the middle, Vamadeva on the proper
left and Aghora on the proper right; second, with Sadyojatain the centre, Aghora and Vama-
deva occupying the properleft and right sides; and third,with Aghora in the middle, Tatpurusa
and Sadyojataareperhapscarvedon the properleft andrightsides.Whereasthe imagesof the
firsttwo groupsarefromvariouspartsof northernIndia,thosebelongingto the thirdtype are
confinedto the south.
While the identity of the symbols of the centralhead in almost all the three types of Mahesa-
murti-images,namely citron and rosary,complicatesthe identification,a carefulobservation^of
the lateralheadsin relationto the orientationof the fourfacesof a caturmukhalinga,
theirphys-
iognomy, and attributes solves the problem. A caturmukhalinga with its heads in the cardinal
directions has the following four combinations,of three faces each:

View Proper Centre Proper Illustration

right left

East Aghora Tatpurusa Vamadeva Plate V

West Vamadeva Sadyojata Aghora Plate VI

South Sadyojata Aghora Tatpurusa Plate VII

North Tatpurusa Vamadeva Sadyojata Plate VIII

Thus we arriveat the conclusion that the three types of the seatedMahesamurti-images enumer-
ated above are in consonance with the hfirstthree views of a caturmukhalingafacing east, west,

9 Together with the other Mahesamurti-images, H. Goetz has identified this sculpture also as Sadasiva. Cf. "A Kashmir!
Lingam of the ioth century", Artibus Asiae, XXVII, 3, I964, p. 276, fig. 6. In fact it is a Mahesamurti-image showing the
river-goddess Ganga descending from heaven. Thus together with the Uma-Mahesvara terracotta figurine in the Bikaner
Museum it represents a new type in the range of extant Ganigadhara-sculptures.
10
Ganguly, Handbook., pp. 68-69; Museum no. G(a)I/379. He has identified the two Mahesamurti-images in this museum
as Sadasiva, but in view of the fact the fifth head representing the Sadasiva-Isana aspect has not been depicted in either
sculpture, his identification does not hold ground.
I" Rao, Elements.,II, ii, pp. 380-382, pl. CXVI.
12 Ibid., p.
382, pl. CXIV, I.

107

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
and south, respectively. No Mahegamurti-image,however, showing the north view of the
caturmukhalinga(namely,the fourthtype):withVamadevain thecentre,Sadyojata andTatpurusa
occupying the proper left and right sides has so far been discovered. This may be due to the
reasonthat the Sivaitesprobablydid not like the idea of giving supremacyto Devi over Siva-
theircult-deity.Typologicallyourimagerepresentsthe secondtype.
The bust-typeMahesamurti-images,which occur in larger number, are divisible into two
types, showing either Tatpurusaor Sadyojatain the middle. While in respect of the position of
lateralheads, they correspondto the first two seated varieties noted above, no such sculpture
representingthe third type, i.e. showing Aghora in the centre, is known to me. The famous
ElephantaMahesamurti,I3 andthosein KangraFort,14ParaSurameSvara Templeat Bhubanesh-
vara,IsMrtyunjaya Templeat Jageshvara,I6 TrimurtiTempleat Badoli,GanesalenaCaveat
Ellora,I7the two images from and
GupkarI8 Pandrethan,Is now preservedin the ShriPratap
SinghMuseum,Shrinagar, andthatfromAnadra,20, etc. representthe firsttype. The sculptures
in LakulilaTempleat Jageshvara,21 SivaTempleat Devangana,22 Sammiddhesvara Templeat
Chittorgarh,23 Siva Templeat Pathari,24 Templeat Ellora:(PlateIX),2 the one
LafikeSvara
fromPadhavaliin GwaliorMuseum,26 thatin AmberMuseum,27 andthosefoundat Baro28and
Lamba,29 etc., pertainto the second one.
The standing variety of the Mahesamurti-images, which are confined to north-western India
arealsoof two typesonly.The firsttype showingTatpurusa in the centreis represented
by two
imagesof Kashmirianstyle-one fromAkhnoor(ca.ninthcentury)30 the otherfrom Pandre-
than(ca. ninthcentury):3I(PlateX), now preservedin the DograArt Gallery,Jammuandthe
ShriPratapSinghMuseum,Shrinagar, respectively.The solitarysculptureof the secondvariety
giving Sadyojata the middle position comes from Akhun Dheri near Charsada in Afghanistan.32
This sculpture in the Indianised Gandhara style of the third century33 is perhaps the earliest
discoveredas yet.
Mahesamarti-image
I3 Kramrisch, AncientIndia, 2, pp. 4ff.
I 961-62 - A Review,p. 103, pl. CLII A.
I4 IndianArchaeology,
I5 Remainsat Bhubanesvara,
K.C.Panigrahi, Archaeological Bombay, 1961, p. 73; Jose Pereira,"A MaheSamurtiat Bhubanes-
vara", Artibus Asiae, XXX, i, I968, pp. 85-87, fig. I.
i6 IndianArchaeology, I962-63 - A Review,p. o02z,pl. CXII B.
17Together with other Mahesamurti-imagesat Ellora, it is
awaiting publication elsewhere. Cf. note 2.
18 Museum no. 2496.
9gMuseum no. 242 /20.
20 Natesa Ayer, "TrimurtiImage in PeshawarMuseum", ArchaeologicalSurvey of India,AnnualReport,1913-I4, p. 278, n. 2, pl.
XXII b.
21 K. P. Nautiyal, Archaeologyof Kumaon,Varanasi, i969, p. I28, fig. 42.
22
ProgressReport,Archaeological Surveyof India, WesternCircle,1906-7, p. 30.
23 Rao, Elements,II, ii, p. 385, pl. CXVIII.
24 D.R. Patil, The
Descriptiveand ClassifiedList of Archaeological
Monuments in MadhyaPradesh,Gwalior, 1952, pp. I09- 10.
25 Fergussion and Burgess, CaveTemples,p.459.
26
J.N. Banerjea,TheDevelopment of HinduIconography,
Calcutta,1956, p. 477; pl. XXXIX, fig. 3.
27 Museum no. 280/60.
28
A.S.I., A.R., p. 278, n. 2, pl. XXII, d.
29 M.A.Dhaky,"The old templeat Lambaand KamesvaraTempleat Auwa",Journalof theAsiaticSociety,
VIII, 3, 1966,
P. I44.
30 Museum no. 2.
31 Museum no. 242I/3.
32 A.S. I., A.R., pp. 276-280.
33 A.K.Coomaraswamy, History of IndianandIndonesian
Art, New York, I965, p. 55.

Io8

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
 The Uma-Mahesvaratype Mahesamurti-images34 constitute a single group showing Tat-
purusain the centre.35It is representedby two standing and one seated image of Kashmirian
style; one from Baramula (ca.sixth century) in the Shri Pratap Singh Museum, Shrinagar:
(PlateXI),36the otherfrom Gandhara(ca.seventhcentury),37 andthe thirdfromKashmir(of
about the late seventhto earlyeighth century)both in the BritishMuseum.38 Theseimages
have certaininterestingfeatures:first,in all of them Sivais shownas urdhvaretas
(ithyphallic)
secondly, in two images Siva carriessun and moon,39respectively in his upper left and right
hands. The Uma-Mahesvarasculpturesof this variety, essentiallyconfined to Kashmirand the
Gandhararegion,belongto the Shahiperiod,when Kashmirian
arthada deepimpactboth in
Gandharaand easternAfghanistan.40In this context a referenceto a unique terracottafigurine
from Rangmahal(ca.fourthcentury)now preservedin the GangaGolden
of Uma-Mahesvara
Jubilee Museum, Bikaner, may be useful: (Plate XII).41 In this relief the three-headedSiva
Unlike the sculpturesfrom Kashmir
shows his Tatpurusaaspect in the centre, with urdhvalinga.
and Gandhara,sun and moon are held in the right and left hands,respectivelyby a figure
emerging from the bust upwards above the three heads of Siva. Moreover, a damagedfemale
figure is shown hovering horizontally above the emerging bust. A similarflying female figure
is also seen in the Mahesamurti-imagefrom Pandrethan:(Plate IV). V.S.Agrawala has iden-
I, however,stronglyfeel
tifiedthe two figuresover Siva'stripleheadas a pairof Vidyadharas.42
that while the bust shows the fourthface of Siva, the flyingfemalemay be identifiedwith
goddess Ganga43descendingfrom heaven at the requestof king Bhagiratha.No Kalyansundara
aspect as has been pointed out by V.S.Srivastava44is implied in this terracottatablet which
representsa proto-typeof the Kashmiriansculptures.
In the absenceof anytextualreferenceit is difficultto explainthe reasonassigningthecentral
positionto one or the otheraspectof Siva in a Mahesamarti-image. It is probably,due either
to some ritualisticconsideration,or the sectarianor personal leaning to one or another form
of Siva;or to caste-biasenjoinedby the Vi.s.nudharmmottara.45
So far as the physiognomy and attributesof the centralhead of the Mahesamurti-imageare
concerned,the situationis more or less anomalous.While the physiognomicfeaturesofAghora
are quite distinctive, no such differenceis discerniblein the facial expressionsof Tatpurusaand
Sadyojata.Moreover,as enjoinedby certainiconographictexts,46the citronandrosaryarethe
attributesof Tatpurusaonly, but at the sametime thereis some literaryas well as sculptural
in a Mahesamurti, Siva is accompanied by Devi. Cf. Rao, Elements,II, ii, p. 379.
34 According to the Suprabhedigama,
35 In almost all the MaheSamuirti-imagesfrom Kashmir, Tatpurusa occupies the central position, which may be due to
sectarian bias.
36 Museum no. 1I6-37.
37 Douglas Barrett, "Sculptures of the Shahi Period", OrientalArt, III, I, I957, PP. 58 and 59, fig. Io.
38 Ibid., p. 59, fig. I2.
39 It is not a cakraand a kapalaas held by Barrett. Cf. OrientalArt, p. 59, but a sun and a moon as proposed by P. Banerjee
and R.C.Agrawala. Cf. "Hindu Sculptures in Ancient Afghanistan", India's Contribution to WorldThoughtand Culture,
Madras, 1970, p. 222.
40 Ibid.,p. 222.
41 V. S. Agrawala, "The Religious Significance of the Gupta Terracottasfrom Rangmahal", Lalit Kala, 8, 960, pp. 66-67.
42 Ibid., p. 67.
43 V.S. Srivastava, Cataloguee Guideto GangaGoldenJubileeMusetm,Bikaner,Jaipur, I96o-6I, p. 5.
44 Ibid., p. 5.
4s Rao, Elements,II, ii, p. 376.
46 Ianafiva-gurudeva-paddhati, IV, I 2-I3.
IIII,xii, 355;Rupama.ndana,

Io9

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
 andAghora,
evidence,whichprovebeyonddoubtthatthe two objectsaresharedby Sadyojata
also.47This fact is amply illustratedby the central head in all the three varieties of the seated
Mahesamurti-imagesreferredto above. The miniaturerelief of Mahesaat Ellora displays the
middleheadof Tatpurusacarryinga citronin the left handandprobablya rosary,in the right
Universitysculpture,the
(PlateIII). Similarlyin the Pennsylvania
one displayedin abhaya-mudrd
centralhead representingSadyojata,carriesa citron as well as a rosaryin the lower and upper
left hands, while the lower and upper right arms are displayed in varadaand abhaya-pose,
respectively.48 Likewise, in the six-armed image of Mahesa near Gokak Falls, the middle face
representingAghora holds a citron in the left hand and a rosaryin the right.49This identity in
in both the varietiesof the bust-typeas well as
attributesof the centralheadis also maintained
of the standing and Uma-Mahesvaratype Mahesamurti-imagesnoted above. In the Mahesa-
murti-imagesof Kahsmirand Gandhara,which perhaps constitute a separateclass (with Tat-
This featureis
purusain the centre),the itronin the left handis replacedby a water-vessel.50
also observedin the Mahesamurti-image fromPadhvaliin MadhyaPradeshandthe terracotta
fromRangmahal in Rajasthan.
Although the similarityin attributesof the middle head is a little confusing, most probably
it is due to the fact that differenttexts were followed by the sculptors in time and space, all of
which are not availableto us. In this context StellaKramrischhas rightlyobserved:"The
in the various texts are not identicalin detail. There are as many, if not necessarily
dhdyana-slokas
the same variations,there as in the images. The vision held by the signs and words is the same:
differentages and differentschools see it under particularmodifications."51Thus the identity in
physiognomyof the centralhead(in caseof Tatpurusaand Sadyojata)
as well as theirsymbols
(in all the three types) proves that the iconographicfeaturesof the middle face alone cannot be
the guidingfactorsin the properidentificationof thetriuneheadsin generalandthe centralface
in particular.In thistasktherelativepositionof the lateralheads;theirphysiognomyandattrib-
utes; andlastly,in casewherethe sculpturesarein situ,theirorientations2 areof greathelp in
settlingthe issue.

MAHESAMURTIAS A BUDDHISTDEITY
becameso popularin ancientIndiathatit no longer
The conceptof the Mahesamurti-image
could remainconfined within the folds of Saivismalone. With the emergenceof the Vajrayana,
like many other Brahmanicalcult-icons, this too was soon incorporatedin the ever expanding
pantheon of Northern Buddhism, and through its well-known process of addition to and
substitutionof the attributesand other changes,the Buddhisticonographershad skilfully
transformedit into one of theirdivinities.A three-headedand six-armedmediaevalimageof
gilt copper was noticed by E.B.Havell in Nepal, which he had identified as the Buddhist
47 Rao, Elements.,II, ii, p. 379; Vis.nudharmmottara,
III, 48, I-20.
48 Coomaraswamy, Rfipam,p. 66, fig. 4.
49 Rao, Elements.,II, ii, p. 383, pl. CXIV, fig. I.
Viudharmmottara,III, xlviii, 9.
50so
s5 AncientIndia, p.6, n.
52 It is observed that the Mahesamurti-imagesat Ellora with Tatpurusain the centre (excepting one looking west) face south
or east,whilethoseshowingSadyojatain the middlelook towardsnorthor west.

IIO

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
Plate I Mahesamurti- Nimaj - ca. Ioth century. (Courtesy Michael W. Meister)

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
 *-sfi 59 J i, ,i

S IaY1seqDZ1\
(la.sya/~Lh[ dSaXlno3) *XJnXua,yloI *pJ~- leur1 - Iz~nusasayeX II sU14

I_ _ ~~~~~~~~~~~~~~~~~~

|D Xl__ D~~~~~~~~~~~~~~~

_E*_UiD

_~~~
I" _-~ I

_ -f;- _
_
'-~~~~~~~~~~~) -

S'u;Ftt~
s_N_

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
Plate III Mahesamurti- Ellora (Kailasa) - ca. 8th century. (Photo Krishna Kumar) Plate IV Mahesamuirti - Pandrethan - ca. 9
(Courtesy American Institute

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
Plate V Caturmukhalifnga- (East Face) - Kannauj Plate VI Caturmukhalifiga(West Face) - Kannauj
ca. 8th century. J. N. Kapoor Collection, Kannauj. ca. 8th century - J. N. Kapoor Collection, Kannaluj.
(Courtesy Prof. Pramod Chandra) (Courtesy Pramod Chandra)

Plate VII Caturmukhalifiga(South Face) - Kannauj Plate VIII Caturmukhalifnga(North Face) - Kannauj
- ca. 8th century - J. N. KapoorCollection,Kannauj. ca. 8th century - J. N. KapoorCollection,Kannauj.
(Courtesy Pramod Chandra) (Courtesy Pramod Chandra)

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
Plate IX Mahesamurti- Ellora
(Lafnkesvara)- ca. gth century. -
(Courtesy Director General, Archaeological
Survey of India, New Delhi)

Plate X AMahesamurti - Pandrethan-

ca. gth century -
Shri Pratap SinghMuseum,Shrinagar
(Courtesy American Institute of Indian
Studies, Varanasi)

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
 El A
^
Eck -
-.:""^-E -: AI- .
- .AI,

''.ci~ ~ ~ ~ ~ ~~ e :"
'.i ' .~t .. ,I.cC....

Plate XI Uma-Mahesvara- Baramula - ca. 6th century - Shri Pratap Singh Museum, Shrinagar. Plate XII Uma-Mahesvara - Rangmahal - ca.
(Courtesy American Institute of Indian Studies, Varanasi) (Courtesy American Institut

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
Trimarti.53The spreadof Buddhismbeyondthefrontiersled to the transportation
of thisimage
evento the distantlands
ands CentralandEasternAsia,andthusit assumeda pan-Asian-Asian
of Central character.
TheTantrikaformof thethree-headed deitywasfirstnoticedby AurelSteinon threepainted
wooden panels: (Nos. D.VII. 6, D.X. 5 and D.X. 8) from Dandan Uiliq in Khotan.s54 A wall-
thethree-headedgod.ss While
also depicts the
painting discoveredby him at Balawaste:(Bal.0200ozoo)
in all the examplesnoted above the three-headedand four-armeddivinityshows the pacific
the
aspectin the ceterrtre,ific
centre, theand
beauterriful the displayedon
facesareinvariably proper
left andrightside,respectively.The god wearsa tiger-skinandis shownas urdhvamedhra,
while
he carriesa pomegranate and a vajrain the lowerlef ahands.
it Whereasthe attributes,
which he held in the upper hands are not clear in other examples,in two cases at least (wooden
panel D. VII. 6 and fresco Bal. 0200) he is depicted as carryinga sun and moon56respectivelyin
the upper left and right hands. In a wooden panel: (D. X. 8) the deity is also accompaniedby his
Sakti,who kneelsovertherightthighof herconsortandoffershimacupof somedrink(seemingly
wine) held in her right hand. As noticed by Stein this arrangement is common in Tibetan represen-
tations of the specialtutelarydivinitiesor Yi-dam.57To some extentthesepanelsmaybe compared
with the three-headedithyphallicUma-Mahesvarasculpturesfrom Baramula(ca.sixthcentury)in
the ShriPratap Singh Museum: (Plate XI) and from Gandhara (ca. late seventh to eighth
century)in the British Museum,58 and the Uma-Mahesvaraterracottafigurinefrom Rangmahal
(ca. fourth century) in the Bikaner Museum: (Plate XII), which also display the association of
sun and moon with Siva and assign the centralposition to the pacific(Tatpurusa)head.
The wooden panels together with the frescoes from Central Asia, as held by A. K.Coomaras-
wamy, "date back from the seventh or eighth century and afford illustrations of Indian painting
of that period, subject to Persianand Chineseinfluences,the Indian element predominating".s59
He further observes: "The type, which alreadyoccurs in Gandharaand on the coins of Vasu-
deva, could easily have reached Khotan through Kashmir in the second century; it extended
laterto China(Yun Kang) and appearsin Japanas Dai Itoku, usuallyrenderedas Yamantaka."60
Here it may be pointed out that since no representationof this deity belonging to a period earlier
than the seventh-eighth century has yet come to light from CentralAsia, it is likely to have
reachedthere not before the late Gupta period.
On the basis of the occurrence of the sun and moon motifs as the emblems of Siva in a
Kusana relief at Musanagar,P. Banerjeehas identifiedthe three-headedfiguresin CentralAsian
panels as those of Trimurti-Siva.61In this connection it may, however, be observed that the
sun and moon find parallel depiction on Sivaite as well as Buddhist sculptures from a very
early period; they were most probably borrowed from the common art heritage of the proto-

53 Havell, Handbook,p. 189, pi. LXVIII B.

54 Aurel Stein, AncientKhotan,Oxford, I907, I, pp. 261 and 279 (pl. LX), 300 and 30I (pl. LXII).
55 F.H.Andrews, Catalogueof Wall-paintingsfromancientshrinesin CentralAsia andSeistan,Delhi, 1933, p. 87.
56 Banerjee and Agrawala, I. C. T. W.T. C., p.222; P.Banrjee, "Hindu Deities in Central Asia", I. C. T. W. C. C., p.285,
pi. 50.
57 Ancient Khotan,I, p. 26I.
58 Barrett, Oriental Art, p. 59, fig. I2.
59 Coomaraswamy, History, p. I49.
60 Ibid.,
p. I49.
61 LC.T. V. T. C., p. 285, pl. 50.

Ix7

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
historic civilization of the Old World. This is testified by an impressionfrom an Assyrianseal,
where sun, moon and Venus representedby seven starsare shown.62A fragmentaryfrieze from
Amaravatiof the firstcenturyB.C., is carvedwith a stupasurroundedby the astralfiguresof
the sun on the properrightand the moon on the properleft side as well as sevenpersonified
stars.63 According to G.Tucci, there are some references in Buddlist texts, which say that,
"Buddhacould touch with his hands sun and moon".64This provided the necessarytheoretical
back-groundfor the artiststo conceive the representationsof Buddhaholding sun and moon in
his hands.In fact, thereis an image of a standing Buddhain ShansiMuseum,where he is shown
in the act of receiving a disc bearinga representationof the moon in his right hand, while in the
palm of the left hand is anotherdisc containingan image of the sun.65The sun and moon were
hypostatisedalso as the two Bodhisattvas:S-uryaprabha
and Candraprabha.66
Thus it is clear
thatsunandmoonwerenot theattributesof Siva,alone,buttheywereequallysharedby Buddha,
also. Therefore,merelyon the basis of sun and moon, the three-headedfiguresin wooden panels
and the wall-paintings from Central Asia, cannot be identified with Trimurti-Siva.This is,
however, not meant to deny a deep iconographicimpact of the Mahesamuiirti-images on these
paintings,as has alreadybeen observedby Stein, who identifiedthem as Avalokitesvara.67
Sivaiteinfluenceis no doubtpresentin thesefigures,but Sivais depictedherein the role of a
Buddhistdeity.
In additionto the presenceof various Sivaitetraits,e. g. the physiognomy of the three-heads
resembling that of Mahesa,jata-jAta(piled up skeins of hair) with crescent, tiger-skin,urdbva-
linga,and a pair of bulls in panel D. VII. 6, the figuresin CentralAsian paintings displaycertain
namelya vajraheldin the lowerleft hand,andin panelD. X. 8, Saktiis shownas
characteristics,
offering a skull-cup of a drink to her consort, which distinguish them from the Mahesamurti-
Siva, and at the same time point towards their affiliationto the TantrikaBuddhist pantheon.
Whilein the absenceof any textualevidenceit is difficultto identifythesedivinitieswith cer-
tainty,it may be pointedout that they possess some qualities,which recallVajradhara, the
highest god of the Vajrayanapantheon,also called Adi-Buddha; e.g.: the three faces, the blue
colour of the centralface, a vajrain left hand, and the paryankdsana-pose. Moreover, in panel
D.VII. 6 the depiction of a pair of bulls below the seat of the deity similarto that of the five
Dhyani Buddhas,who originatefrom Vajradharaand sit on a pair of animals.In panel D. X. 8
the deityis in Yab-yum,as he is embracinghis Sakti,Prajnaparamita,
who occupieshis right
thigh and holds a skull-cup (kapala) in her right hand. This position of Saktiis not in keeping
with Sivaite practice. In the Uma-Mahesvaratype of the Mahesamurti-imagesfrom western
India referredto above, Uma always occupies the left thigh of Siva. This change in the tradi-
tional position of Saktiis presumablydue to the influenceof TantrikBuddhism.Here it may be
observed that in an image of UchchusmaJambhala(ca. eleventh century)in the Archaeological

62 R. Sengupta, "On the identity of the Mlechchas", Prof.K. A. Nilakanta Sastri Felicitation Volume, Madras, I97I, p. I86,
fig. 3b.
63 Ibid., p. I 86, fig. 3 a.
64 "The Tibetan <<white-sun-moon> and cognate deities", East and West, XIV,
3-4, I963, p. I37.
65 Ibid., p. I
37.
66
Ibid., p. I 37.
67 Ancient Khotan, p. 26I.

II8

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
Museum, Sarnath, the god is accompanied by his consort Vasudhara on his right side.68 As I
have already noticed elsewhere this image in the Sarnath Museum betrays Tantrik influence.69
The presentstateof uncertaintyin the properidentificationof the three-headedBuddhist
divinity depicted in Central Asian paintings was also experienced by Stein.70The main reason for
our difficultyis thatthe studyof the iconographyof the NorthernBuddhistdeitiesis stillin its
infancy.Alice Getty has rightlyremarked:"Unfortunately, of all the gods of the
sadhana[s]
Mahayana Pantheonhavenot yet been discoveredandthereremainsa numberof deitiesabout
whom very little is known."71Moreoveras observedby Tucci, "... the religiousstudy of
Central Asia is certainly one of the most complex ones, we know; the impact of great religions,
or of the most vital sects, which met along the caravan routes that crossed it in all senses caused
and intensified a great interchange of ideas and forms, borrowings and contaminations".72
The photographsillustratingthis articlehave been obtainedthroughthe courtesyof Mr.
MichaelW.Meister;Prof.P dra ; the AmericanInstituteof IndianStudies,Varanasi;
and the Director General, Archaeological Survey of India, New Delhi. I am grateful to the
Curator, Shri Pratap Singh Museum, Shrinagar; and the Director, Department of Archaeology
and Museums, Government of Rajasthan, Jaipur for permission to reproduce their museum
exhibits, and also to Prof.M.A.Dhaky and Mr.M.W.Meister, who have very kindly gone
through the article. My grateful thanks are also due to Dr. S.C. Ray, who has been kind enough
to draw my attention to the Mahesamurti-image at Badoli.

68 D.R.Sahni, Catalogue of the Museum of Archaeologyat Sarnath, Calcutta, 19I4, p. I 35.

69 "The Buddhist origin of some cave-temples at Ellora", awaiting publication in East and West.
70 Ancient Khotan, p. 279.
7, The Gods Northern Buddhism, Tokyo, I962, p. viii.
of
72 East and West, XIV, 3-4, I963, P. I44.

POSTSCRIPT
In the courseof my recentvisit to the NationalMuseum,New Delhi, I had the chanceto
see somelinenbannersfromChien-fo-tung(ThousandBuddhaCaves)at Tun Huangin China.
Six of them, datable to ca. the seventh-eighth century, variously depict Bodhisattva Avalokites-
vara with either one, three or eleven heads. The deity besides carrying other attributes also
supportstwo discsrepresentingthe sun andthe moonin his upperhands.Whilethe sun bears
a bird, the moon is painted with a tree and a rabbit; in one case the moon shows a skull, too.
Althoughin most casesthe triuneheadsinvariablydisplayinganangrymoodcannotbe distin-
guished from one another, in one example at least: (CH xxiioo 3) the left hand holding a moon
and the right one a sun representrespectivelythe femaleand the maleaspectsof the divinity.
On one hand the Chinesebannersbetraya deep impactof the iconographicfeaturesof the
Mahesamurti-image in the make-up of Avalokitesvara, and on the other they prove the Buddhist
character of the deity painted on the wooden panels from Khotan. It is likely that the Central
Asian panelsbeing the earlycontemporaries of the Chinesebanners,the formermight have
inspiredthe latter.This probabilityis furtherstrengthenedby the fact thatsituatedas it is on
the ancientsilk-routeTun Huangwas the meeting-ground of ChineseandCentralAsianart.

ii9

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions
 Mario Bussagli,who also noticed a Sivaiteaspect in the DandanUiliq image, has pertinently
observed: the "luminous and astral values connected with the sun and moon... were partic-
ularly important in Khotanese Buddhism".IMoreover, he has pointed out the possibility of
relating these divinities "in some way to the four-armedfemale figures on Sassanianand Sog-
diancups,whichin additionto sunandmoon,carryothersymbolsnot alwaysthe sameas those
of the maledivinitiesof Khotanbut moreor less equivalentin meaning".2
The sun and moon in the CentralAsian panels as well as the Chinesebannersmay be iden-
tified respectivelywith snyatad(the void) which isprajna(the gnosis) andkarunda
(the compassion)
whichis upaya(the means)in the Kalacakrayana. Accordingto the Tibetantexts this systemof Bud-
dhistphilosophyoriginatedin thecountryof Sambhala,whichLauferhasattemptedto locateto
the north of riverTarimin EastTurkistan.As the conceptionof Kalacakra
andVajradhara is
substantiallythe same4andthe Lamaism of Tibet from
originated the the
Kalacakrayana, iden-
tification of the so-called 'Trimurti-Siva' on the Central Asian panels with Adi-Buddha as pro-
posed in this article gets further corroboration.The reason that lies behind the iconographic
influence of the Mahesamurti-imageon the Buddhist pantheon is explained by the fact that
from the Tantric texts, viz., Sri-kala-cakra-tantrasand Lagh-kala-cakra-tantra-ra 6-tika6 we
come to know that by the introduction of this cult an attempt was made by the Buddhists to
make themselves united with the Hindus under the common banner of the god Kalacakra
against the culturalpenetrationof the Mlecchas,i. e., the Muslims. Later history has, however,
shown that this effort proved totally futile.7

I Painting of Central Asia, Geneva, 1963, p. 6 5.

2 Ibid., p. 66.
3 Toung Pao, I907, p. 404.
4 S.B.Dasgupta, Introductionto Tantrik Buddhism, Calcutta, I958, p.67.
s Cambridge, MS. Add. No. 1364.
6 Asiatic
Society Bengal, Calcutta, MS. No. 4272.
7 B. Bhattacharya, Indian Buddhist Iconography,Calcutta, 1958, pp. I87 and 389.

I20

This content downloaded from 195.78.109.164 on Tue, 10 Jun 2014 16:47:55 PM

All use subject to JSTOR Terms and Conditions