You are on page 1of 37

Bodhisattvas and Buddhas: Early Buddhist Images from Mathur

Author(s): Prudence R. Myer


Source: Artibus Asiae, Vol. 47, No. 2 (1986), pp. 107-142
Published by: Artibus Asiae Publishers
Stable URL: http://www.jstor.org/stable/3249969 .
Accessed: 17/10/2011 13:23
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 support@jstor.org.

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

http://www.jstor.org

PRUDENCE R. MYER
BODHISATTVAS AND BUDDHAS:
EARLY BUDDHIST IMAGES FROM MATHURA

is well known, the first images of the Buddha are said to have been produced either
in the ancient province of Gandhara, in the northwest of the Indian subcontinent,
or at Mathura, a flourishing commercial and religious center some I25 kilometers south
of Delhi.1 This city, which once formed the focus of the southern part of the I(ushana
As

empireand, before that, had been ruled by satraps(ksatrapa)belonging to a branchof the


Sakas(a CentralAsianpeople betterknown in Europeas Scythians),is the scene of a story

telling how Mara, the Lord of Illusion, was vanquished and converted by the renowned
monk and preacher Upagupta. Lamenting that he had been born too late to see the Blessed
One, Upagupta commanded his erstwhile enemy to show the appearance of the Buddha,
and Mara assented on condition the monk not prostrate himself before it. But when the
radiant apparition appeared, the monk promptly fell to his knees before it. Reproached
by the god for having broken his word, Upagupta justified his spontaneous action by

saying:
Of course, I know that the Best of Speakers
has gone altogether to extinction,
like a fire swamped by water.
Even so, when I see his figure,
I bow down before that Sage.
But I do not revere you! . . .

Note onterminology
andtransliteration:
Indian words are italicized the first time they appear. Sanskrit words are transliterated in accordance with the system
employed by Monier-Williams' Sanskrit-EnglishDictionaryexcept for the unaspirated palatal c, which is transliterated

as chin orderto makeit more consonantwith English pronunciation.

1
Specialists will recognize that the present author is profoundly indebted to the late Johanna E. van Lohuizen-de
Leeuw, whose doctoral dissertation brilliantly resolved the fundamental problems of chronology and style associated
with the Buddhist images from Mathura (The "Scythian"Period: An Approach to the History, Art, Epigraphy, and
Paleographyof North Indiafrom the ist Centuryto the 3rd CenturyA.D. [Leiden, 19491, hereafter to be referred to as

van Lohuizen-deLeeuw).

The most comprehensive study of Buddhist art here is R. C. Sharma,BuddhistArt of Mathura(Agam Kala Prakashan,
Delhi, I984). This is particularly valuable for its account of the excavations that have taken place in and around the
city, especially the finds from the ancient mounds known as Govindnagar, which lie westward from the Katra mounds
and on the other side of the railroad. It was in I969-70 that laborers leveling the mounds in preparation for the
construction of an extensive housing colony began to turn up quantities of sculptures. Most of these were purchased

by antiquedealers,but in 1976Mr. Sharma(then Curatorof the MathuraMuseum)and his staffattemptedto rescue


andrecordthe piecesthatwerebeing found,andthe U. P. governmentdeclaredthatareaa protectedsite. The material
so recovered, although only a fraction of what had been dispersed, showed that the site went back to the pre-Maurya
period and remained an important Buddhist site until the Huna invasion of the early 6th century.
107

Just as men bow down


to clay images of the gods,
knowing that what they worship
is the god and not the clay,
so I, seeing you here,
wearing the form of the Lord of the World,
bowed down to you,
conscious of the Sugata,
but not conscious of Mara.2

In summarizing this story as evidence of the growing practice of Buddbha-pjd,the cult


of devotion to the Buddha as Lord, Sukumar Dutt points out that the description of the
figure or illusion created or presented by Mara is couched in such vague and general terms
as to indicate that the text antedates the first man-made images of the Buddha.3The present
paper reconsiders the early Buddhist images from Mathura, starting with the seated and
standing figures identified by inscription as Bodhisattvas. A suggested explanation for this
nomenclature is followed by comparisons with images of the "bejewelled Bodhisattva"
type, and a concluding study of the early Buddha images considered to show Gandharan
influence.
All the Mathura sculptures are carved of red or rose-hued sandstone, usually mottled
or dappled with cream color, but as the passage quoted from the Asokadvaddna
suggests,
wood and clay must also have been used for images here as in other parts of India. That
some sculptural types were originally conceived in these materials is indicated by the
presence of such features as the upraised right hand, which is so ill-adapted to the medium
of stone that it obviously caused early stone-carvers some difficulty. Bronze may also have
been used, for two small cast-bronze images were found in Kushana levels of Sonkh, an
archaeological site 221/2kilometers southwest of the city;4 but no early Buddhist bronzes
have yet been identified.
Discussions

of the Buddhist

sculpture from Mathura often


figure dedicated at Sarnath by the monk (bhikSu) Bala in the
and with the smaller and well preserved stele found at the
Because of its material and inscriptions (L.925, L.g27), which

start with the monumental


year 3 of the Kanishka era,
IKatra Tila (Figs. I, 10, 13).
are the earliest indisputable

JohnS. Strong, TheLegendof KingAsoka: A Studyand Translation


(PrincetonUniversityPress,
of theAsokdvaddna
Princeton,N.J., 1983), PP. 195-196.
3 Sukumar
Dutt, TheBuddhaandFive After-Centuries
(London, 1957),P. 236.
4 HerbertHartel,"SomeResultsof the Excavationsat Sonkh,"German
onIndia,Contributions
to IndianStudies,
Scholars
ed. by CulturalDepartmentof the FederalRepublicof Germany(Bombay,n. d.), II, pp. 70-99.
The Sonkhexcavations,which have revealedan uninterruptedsequenceof occupationsfrom the the late Kushana
perioddown to pre-Mauryalevels, constitutethe first reliablearchaeologicalevidencefor the historyof the Mathura
B.C. Mathuraseemsto have brokenawayfromthe Sungaempire,and coins foundin levels 28-25
region.About 00oo
areinscribedwith the namesof Gomitraandthreesuccessorshavingnamesendingwith -mitra.Followingthesecame
a king namedRamadattawhose reign,at leastat Sonkh,seemsto have been interruptedor overlappedby that of the
KshatrapasHagamasa,Rajuvula,and Rajuvula'sson Sodasa(levels 24 and 23); and these were immediatelyfollowed
by the KushanasWimaKadphisesand Kanishkain levels 22 and 21. Becausethe beginningof the eraestablishedby
Kanishka(whichcontinuedto be used at Mathurafor more than I50 years)has been datedas earlyas 78 A.D. and
as late as 144 or even later,particularinterestattachesto Hartel'sassertion,"Notwithstandingthe theory one may
follow in datingKanishkaI, after Sonkh there is no justificationof placinghim in the second or even in the third
centuryA.D."
2

io8

post-Asokanones,5it is sometimesassumedthat Bala'simage must be among the earliest


Buddhist images carved at Mathura.6 Yet as early as I9IO the catalogue of the Mathura
Museum, in discussing the Katra stele, noted that both its very dark red color and the
language of its inscription (L.I 2 a), identifying its donor as one Amohaasi (Amoghadasi),

indicatethatit pre-datesthe Kanishkaera.7Both imagesareidentifiedby theirinscriptions


as Bodhisattva,
and both obviously come from the sameiconographicand stylistictradition,
even though one is standing and the other seated. The head of the Sarnath image has been
damaged and its right arm and hand broken off, but it is apparent the lost hand must have
the gesture signifying "no fear".
been raisedin front of the shoulderin abhaya-nmudra,

This gesture, familiarto all studentsof Buddhistart, may well have been introduced
into India by the Sakas(more accuratelySaka-Pahlavas),who reigned at Mathurabefore
the advent of the Kushanas.8 It was not restricted to male figures but appears in a number
of representations of a goddess usually identified as Trisala, the mother of the Jina
Mahavira. It is presumably she who is depicted (although the inscription names her

AryavatI)on a tablet found at the site of KankaliTila, which was dedicatedby the lady

Amohini in the year 42 or 72 during the reign of the Mahakshatrapa(Great Satrap) Sodasa,

the second Sakarulerof Mathura.(Fig. 2).9Her pose is virtuallyidenticalwith that of the


images dedicated by Bala except that her right hand is not directly in front of her shoulder

but off to the side, the palm facing outward. Similarfigures appearboth in reliefs and
independentimages during the Kushanaperiod,10and the same gesturewas also used for
early representationsof Hindu gods.11
It seems that this gesture is probably the same one that l'Orange calls "the Gesture of

Power" or "Saving Right Hand," which can be traced back to Babylonianseals and
reliefs.12There it perhapssignified the divine power to protect the god's servant from
malign influences and evil spirits; and it appears frequently in Achaemenid reliefs of
Ahuramazda and the kings and satraps who served as his earthly counterparts. Its meaning

thereforeseems to be similarto that of abhaya-mudra,


althoughin West Asia the armwas
fully extended. The differencecan be explainedby the artistic conventions of the two
regions, for in West Asian reliefs figures were depicted in profile, while in India divine

Sarnath Museum, B(a) I. D. R. Sahni and J. P. Vogel, Catalogueof the Museumof Archaeologyat Sdrndth(Calcutta, I 9 I 2),
pp. 33-37; H. Liiders, "A List of Brahmi Inscriptions from the Earliest Times to about 400 A.D.," Appendix to
EpigraphiaIndica, X (I909), pp. 93-94.
6
Benjamin Rowland refers to it as "one of the very first images of Buddha to be carved at Mathura" (The Art and
Architectureof India, Buddhist,Hindu, Jain, The Pelican History of Art, ist paperback ed. [Baltimore, Md., 1970],
P I 53)
7 Mathura Museum, oo.A I.
J.P. Vogel, Catalogueof the ArchaeologicalMuseumat Mathura(Allahabad, 1910), pp. 47-48,
pl. 7; also, H. Liiders, MathuraInscriptions,ed. Klaus Janert (G6ttingen, I961), pp. 30-3I. These two will hereafter
be referred to as Vogel, Cat. A.M.M., and Liiders-Janert, respectively.
8 B. N. Mukherjee, Mathuraandits Society:TheSaka-PahlavaPhase(Calcutta, I98 I), especially Chapter VI. See also supra,
note 4.
9 Lucknow State Mus., J. . Vincent Smith, TheJain Stupa and OtherAntiquitiesof Mathura(Allahabad, i901), p. 2zi, pl.
XIV; Vasudeva S. Agrawala, A Short Guide-Bookto the ArchaeologicalSection of the Provincial Museum Lucknow
(Allahabad, 1940), p. 5; Debala Mitra in Jaina Art and ArchitecturePublishedon the Occasionof the 2aooth Nirvana
Anniversaryof TirthankaraMahavira,ed. A. Ghosh (New Delhi), I, p. 67, pl. 19.
10Lucknow State Mus., J. 623, and Mathura Mus., oo. F.6. See Jain Art and Architecture,pl. 3, and Ludwig Bachhofer,
Early Indian Sculpture(New York, 1929), I, pl. 75.
11Mathura Mus., 00.2520. Vasudeva S. Agrawala, Studiesin IndianArt (Varanasi, I965), pp. 191-93, fig. Io6.
12H.P. l'Orange, Studieson the Iconography
of CosmicKingshipin the Ancient World (Oslo, 1953), pp. 39-58.
5

IO9

figures are traditionally presented frontally. Since it is impossible to carve the outstretched

hand of a frontalfigure without riskingdamageto it, Indianartistsseem to have adapted


the older position of the raised hand holding a flower or flywhisk (chauri),familiar from

earlierrepresentationsofyaksas andyaksinzs.13
Both van Lohuizen-de Leeuw and the Japanese scholar Takata Osamu14have compared
the earliest Buddhist images with the tiny figures of Jinas (Tirthankaras)depicted on some

of the earlyayaga-patas
found at KanikallTila, which musthave been a majorcenterof Jaina
None
of
these squareslabs are dated, but they have been generallyattributed
worship.15
to the Saka period. The earliest one may be that here illustrated, whose encircling
lotus-creeperis carvedin a flat relief style very close to that of a fragmentarydoor-jamb,
found at Mora,whose inscription(L.82a)shows it was executedduringthe reignof Sodasa
(Fig. 3).16 The Jina in the central medallion, identifiable by the multiple cobra-hoods of
the serpent king as Parsvanatha, is seated on a low throne and venerated by a pair of naked
Jain monks.17 The four large nandipada shapes occupying

the space between

the medallion

and lotus-creeper are framed by double cords; their surfaces are carved with rows of small
leaf-shapes alternating with pearl bands, a motif recalling both the garlands hanging from

the upperborderof Amohini'stabletand the surfaceof the cushion-likemassof stone that


supports the raised hand of the Katra Bodhisattva. The similar ayaga-pata dedicated by
Simha-nadika is distinguished by a pair of pillars supporting a wheel (chakra) and an
elephant.18 Here the nandipadas are bordered by pearl bands and their forms are more fully
modelled, as are those of the faceted shafts of the framing pillars and the figure in the
central medallion, while the Jina's torso is slimmer, with the diagonal lines of his upper

arms reinforcingthe triangularityof his proportions.


Van Lohuizen-deLeeuw and Takatapoint to close similaritiesbetween these Jaina
images and the small Buddha figures seen in a couple of architectural reliefs, and have
concluded that such miniatures antedated the development of independent images. A
notable example appears on the center of a crossbar from a torana or gateway illustrating
the story of Indra's visit while Buddha meditated in the Indrasaila (or Indra-sala) cave
(Fig. 4).19 Here the king of the gods approaches from the right attended by a pair of
13 E.g., Ananda K. Coomaraswamy, History of IndianandIndonesian
Art New York, I927), pls. IV-I7,
figs. 25, 39, 45.
14 Van Lohuizen,
Takata
no
theBuddha

XI-37; Rowland,

Osamu, Butsugo kigen(The Originof


pp. 145-80;
Image) (Tokyo, I967), PP. 320-97.
(I am indebted to Michiko Grube-Sato for having translated relevant portions of this book for me.)
15 These
square slabs are generally referred to as votive plaques, but Mitra argues that they were mounted on oblong
bases or platforms near the stupa and may derive from the sacred seats (redi),often set up at the foot of a sacred tree
to signify "the physical presence of the invisible divinities," which were therefore themselves objects of worship (Jaina
Art and Architecture,pp. 63-65).
16Mathura Mus., oo.367. Takata, 345,
p.
fig. 142; Liiders-Janert, p. I55.
17 Lucknow State Mus.,
J. 253. Smith, p. 17, pl. X; Takata, pl. 45. fig. 53
18 Lucknow State Mus., J. 249. Smith, p. I4,
pl. VII; J.P. Vogel, "La Sculpture de Mathura", Ars Asiatica, XV (Paris
et Bruxelles, I930), pp. 122-23, pl. LIV-a.
19Mathura Mus., M.3.
Vogel, Cat. A.M.M., pp. i63-64 and "Sculpture de M.," p. 94, pl. VII. Rowland observes that
the blank background of this relief and "nimbus-like" arrangements of the scarves relate it to the stair-riser reliefs
from Buner (Gandhara), which show strong influence from ist century Roman art (Benjamin Rowland, "Gandhara,
Rome and Mathura: The Early Relief Style," Archivesof the ChineseArt Society,X [I956], I 5-I6.) Precisely the same
style can be seen in the crossbar (Mathura Mus. SOIV-36) found near the Sonkh Apsidal Temple No. 2, which can
be dated in the period of Kanishka (Hartel, pp. 95, 99, fig. 44).
I IO

celestial females and three elephants, while his emissary, the divine musician Panchasikha,
and six other females approach from the left. The small figure seated within the cave is
heavy and resembles the Jina of the earlier ayaga-pataexcept that its left hand rests on the

left thigh and the right is raisedin front of the shoulderin clearanticipationof the Katra
type. Although it has no halo, the uppergarmentof this figure also anticipatesthat of the
Katrastele, clingingcloselyto the torso as if so transparentthatthe form of the body shines
through it. This appears to be a variation of the arrangement observable in the foreground
figures of a pillar relief from Bharhut (Fig. 5), where the sculptor carefully indicated how
the pleated length of cloth forming the upper garment was wrapped about the torso and

the free end thrown over the left shoulder.20The folds over the upperarmsof the Mathura
figures show how its other end, which normallyhung down in front, could be brought
back and wrapped around the upper arm and then tucked back behind the body.
Another one, which is carved on a railing post or stambhafound at Isapur on the other

bankof the Jumna,depictsa similarbut clumsierfigure seatedon a tall throne supported


by couchantlions and surroundedby four figures identifiedas the guardiansof the four
quarters(Fig. 6).21The centralfigure,like that of the Indrasailareliefand Jain ayaga-patas,
has no halo but the position of the raisedhand resemblesthat of Aryavatiof Amohini's
dedication,and folds seem to fall under the left arm, suggesting that it belongs to the
period when artistswere still working out the conventions for the type.
Takata believes these representations of Buddha receiving the homage of gods were

but the Buddhist figures are


inspired by and modelled after those of the ayaga-patas,22
distinguished from Jain ones in several ways. Although their torsos have deep navels, the
lines passing under their right breasts and up over their left shoulders show they are
conceived as swathed in clinging garments. Moreover the feet of Jina figures are merely

crossed at the ankles, while Buddhist ones are locked in the full lotus position (padmaasana),the soles being turnedoutwardtowardthe observer.Nor does theirscalenecessarily
indicatethat they were inspiredby the Jainatablets. Indeed the Indrasailarelief must be
datedto the period of Kanishkabecauseof its stylisticresemblanceto the archaeologically
databletoranafound at Sonkh.
Whether the small Buddhist depictions derive from Jain prototypes or from an
independentBuddhisttraditionpeculiarto the Mathuraregion, therecan be no doubt that
larger images were also made and installed in Buddhist monasteries (yihdras)of this region

during the Kshatrapaperiod, for Liidersidentifiesat least five inscriptionsof this period
as coming from image bases.23To these may be added the fragmentaryimage dedicated
by a Kshatrapalady namedNamda,a seatedimage found at Sravasti,and two uninscribed
ones whose style also argue for a date antedatingthe establishmentof the Kanishkaera.24
20 Ananda

K. Coomaraswamy, La sculpturede Bharhut(Paris, I956), Fig. 34.


21Mathura Mus., H. 12. Vogel, Cat. A.M.M., pp. I 3I-I 32; van Lohuizen, pp. 157-158.
22 Takata,

p. 378.

23Liiders-Janert, pp. 30-3I, 105-o6, II5-I6, I2I, I90. Three of these (L.I25a, L.88, L.97b) identify the image as a
Bodhisattva while the two others (L.I43a, L.I 3 5c) lack explicit identifications.
24Liiders seems to have reserved the term Kushana for inscriptions datable to the reigns of Kanishka and his successors.
Most scholars now agree that Mathura had been earlier conquered by the Kushana Wima Kadphises, as confirmed
by the Sonkh excavations. (See note 4, supra, and B.N. Puri, India Under the Kushdnas[Bombay, I965], pp. I9-28.)
III I

Of the Buddhist inscriptions attributed by Luders to the Kshatrapaor (more accurately)


pre-Kanishka era, the only one associated with a fully preserved image is that dedicated
by Amohaasi and found at the Katra (Figs. I, IO). But the others are interesting because
of the light they cast on the life of the Buddhist community at Mathura even when the

images themselvesare missing or badly mutilated.Thus one (L.97b) refers to an image


of a Bodhisattvaset up "for the acceptanceof the Samitiyateachers,"and another(L.97d)
says that something was dedicated "at the Alanakaconvent for the acceptanceof the
for the worship of all Buddhas."25
Mahasaghiyas(Mahasdnghikas)
The broken image van Lohuizen-de Leeuw considered the earliest example of this type
is distinguished by the representation on the base of two figures, the gods Brahma and
Indra (Fig. 7).26 All that remains of the major figure are folded legs and one foot, but it
can be seen that the soles of the feet were turned outward toward the observer and incised
with wheels. The hem of the undergarment, which is treated as two cord-like ridges,
ripples over the calf and hangs down to overlap the folds spread out over the throne.
Brahma and Indra wear long stole-like uttarzyaswhose folds are indicated by paired ridges,
that of Brahma rising stiffly from the shoulders to enframe the head, much like celestial
figures of the Indrasaila crossbar.
A different treatment of drapery appears on the similarly damaged figure executed by
a Mathura sculptor which was dedicated by two Ksatriyabrothers at the Jetavana vihara
at Sravasti.27Here the folds over the throne are treated as crisply defined pleats radiating
like a fan across the throne, the ends of the girdle lie neatly over the central pleats, and
the hems of the garment flare slightly over the calves, describing smooth curves that
emphasize their rounded volumes. The soles of the feet are almost horizontal and the left
hand seems to have been clenched on the thigh.
Much the same style is to be seen in an uninscribed and similarly damaged image in

the MathuraMuseum(Fig. 8).28 Its most significantfeaturesare the clenchedfist and the
rampantlions supportingthe throne.Those at the cornersarelithe and gracefulwith long
thin tails terminating in tight spirals and long tassels or plumes, and the one enface at the

center, although somewhatclumsy, is equally slender.


The last of the brokenBodhisattvaswas found in a shrinenear the Katraand appears
to have been very similarto that dedicatedby Amohaasi(Fig. 9)29. The survivingportion

of the inscription (L.I25C) identifies the donor as a Ksatrapa lady named Namda, who
dedicated a Bodhisattva "for the welfare and happiness of all sentient beings for the
As in the Katra stele, the throne projects
acceptance of the savasthidyas (Sarvdstivddas)."30

25 Luders-Janert, pp. I I 5-I6, 121.


I
26Lucknow State Mus., B.I8. Van Lohuizen-de Leeuw, pp. i74-75,
pl. XXI-34.
27Lucknow State Mus., 66.48. G.R. Sahni, "A Buddhist Image
Inscription from Sravasti," A.S.I. Ann. Report 90o8-o9,
PP. I33-38.
28Mathura Mus., 00oo.2073.V. S. Agrawala, "Buddha and Bodhisattva
images in Mathura Museum," Journal of U.P.
Historical Society,XXI (1978), p. 73.
29Mathura Mus., A.66.
Vogel, Cat. A.M.M., pp. 63-64.
30Liiders noted that the letters of the inscription look very archaic, but concluded it should be dated to the Kushana
period because of its faulty spelling (Liiders-Janert, pp. 31-32).
112

from a backslab on which can be seen the lower portion of a standing attenldant;the fingers

of the Bodhisattva'sleft hand are similarlyextendedto rest on his calf, and the sole of the
foot is incised with a wheel and the toes with tiny swastikas,but the outwardturn of the
sole of the foot and the undulating hem of the garment indicate it antedates that piece.

Moreoverthe rampantwinged lion of the base,althoughsomewhatless fluid and energetic


than those of the precedingimage, is very differentfrom those squattingon the bases of
the Katra stele and its successors.31
Despite the fragmentary state of these images, enough remains to show that the Katra

stele was not only more complex but more carefully integrated, and deserves to be
considered the classical statement of the type (Fig. IO).The Bodhisattva's smooth, swelling
torso rises vigorously from the broad, horizontal base formed by his folded legs and feet

and culminatesin the highly articulatedvolumes of his neck, egg-shapedhead, and coiled
top-knot of hair;the inwardslant of his right forearmis balancedby that of his left upper
arm; the strong three-dimensionalaccentformed by his projectingright hand is balanced
by the deep folds over his left arm, and these continue downwardand are continued by
the lines of his girdle, unifying the upper and lower parts of the figure. Comparison with

the surviving portions of the earlierdedicationby Namda show that there the backslab
was relativelywider, the throne projectedless, and the Bodhisattva'slegs and soles tilted
outward (Fig. 9). These, together with the wavering line described by the hem of the
garment and the soft shape of the hand, contrast greatly with the strongly architectonic

forms of the Katra figure. By setting the left hand fartherout, closer to the bent knee,
and straighteningthe fingers so they continueand emphasizethe verticalline of the lower
arm, the later sculptor suggested a vigorous downward pressure that both counterbalances
and emphasizes the significance of the gesture made by the right hand. The Bodhisattva's

suave but vigorously modelledvolumes, unobscuredby any folds, are set off by the lively
rhythmsof the pipalatree (presumablythat underwhich SakyamuniachievedEnlightenment), the hovering figures above him, and the tilted heads and swaying postures of the
two standing figures. Here everything seems to be in movement, effectively contrasting

with and emphasizingthe stabilityof the centralfigure and strengtheningits expression


of power and authority.
Significant comparisons can be made between this and two well preserved steles found
some two hundred kilometers north of Mathura at Ramnagar (District Bareli, U.P.).32 One
appears to derive from the pre-Kanishka period, for the soles of the Bodhisattva's feet turn

outwardand his left hand, although placed just above the knee, is clenchedin a fist like
that of the central figure on the Isapur railing-post, and the lower portions of the attendant

on the Bodhisattva'sleft are almostidenticalwith those of Namda'sdedication.While the


general clumsinessof the piece may be partiallyattributedto incompetenceon the part
of the sculptor (as seen in the misunderstanding

of the lateral branches of the tree, the

These rampant lions link these two Bodhisattva images with a number of architectural pieces that seem to reveal the
influence of Iran or the mixed cultures of old Bactria.
32 Debala Mitra, "Three Kushan Sculptures from Ahichchhatra,"Journalof the Asiatic Society,XXI, I (I95 5), pp. 65-67,
Pls. II-IV, VI.

31

II3

inconsistent proportions of the figures, and the asymmetry of the flying figures' positions),

he probablyfollowed his model in the rectangularshapeof the upperportion of the stele,


for the outer legs of the flying figure are bent at an acute angle to fill in the cornersof
it. This, togetherwith the rigid stanceof the attendants,suggest that the subtletiesof the
Katra stele were the result of conscious refinements perfected over a period of time.

The otherRamnagarstele, which is datedin the year32 (presumablyof Kanishka'sera),


is clearlymore developed, both stylisticallyand iconographically(Fig. I I). The swaying
postures of the standing attendants and the attitudes of the flying ones, whose gazes seem
to be directed toward the Bodhisattva, together with the horse-shoe shape of the backslab
and the rigidity of the central figure, show that we here have to do with a later work. The
standing figures do not carry flywhisks but are clearly differentiated by their costumes and
attributes. The cluster of lotuses carried by that to the central figure's left indicate that

he maypossibly representPadmapani("lotus-hand"or "holdinga lotus in the hand");but


although this title is often applied to images of Avalokitesvara, the most popular Bod-

hisattvaof the Mahayanapantheon,in this casethe figuremaybe only a divine worshipper


or attendant. The other one, whose short skirt animal-skin mantle and thunderbolt (yajra)
attribute suggest he may possibly represent Indra in the guise of Hercules,33 is more
probably Vajrapani, the yaksa who, according to Mahayana traditions, accompanied the

Buddhaon his (apocryphal)journeyto Gandhara.34


In summary,we have seen that the type of the seated Bodhisattva seems to have
developed during the Sakaor Kshatrapaperiod and reachedits classicalstatementprobably shortlybeforethe beginningof Kanishka'sera,in such works as the image dedicated
by two Ksatriyabrothersat Sravastiand the Katra stele dedicatedby Amohaasi. They
continued to be made well into the reign of Huvishka, as indicated by the inscription

(L.4Ib) of an image from Pallkheradated in the year 39 (Fig. I2),35 but later examples
reveal a distinct loss of artistic refinement, especally in the treatment of the folds of the
garments. Their dedicatory inscriptions show they were dedicated by both monks and lay

people, especiallywomen, at viharasbelonging to the Mahasamghikaand Samityaschools


or sects; but duringthe Kushanperiod the popularityof the type was rivaledby standing
images and the type of fully-drapedimage to be discussedat the end of this paper.
The earliest surviving images of the standing type are known to have been dedicated

by membersof the Sarvastivadaschool at sites hallowed by their associationswith the


Buddha's ministry. The first was unearthed by Cunningham in I862 in the ruins of the

am indebted to A.C. Soper for this suggestion.


It may be noted that J.E. van Lohuizen-de Leeuw believed that this type of image inspired the first Buddhist images
from the Gandhara region ("New Evidence with regard to the Origin of the Buddha Image," SouthAsian Archaeology
179 ...jth InternationalConferenceof South Asian Archaeologists,ed. Herbert Hartel [Berlin, n.d.], pp. 377-400). In
support of this thesis she presented a number of reliefs which closely resemble the seated images from Mathura, some
of which come from the level of the great stupa at Butkara in Swat datable to the pre-Kushana period. While it cannot
be doubted they represent a hitherto unrecognized extension of the North India style, there are significant iconographic and stylistic differences between them and the early Buddhist images from Mathura, which suggest they may
represent parallel developments, unrelated except for their common roots in the thought and practice of Buddhists
in the two regions.
35 Indian Mus., Calcutta N. S. 4145. Bachhofer, pl. 83-2, Liiders-Janert, I65-66.
33 I
34

II4

Fig. 2 Tablet dedicated by Amohinl in year 72 (42?), LucknowState Museum..


(Photo Lucknow Mus.)

Fig. i

Bodhisattva dedicated by Amohaasi ("Katra stele") MathuraMuseum.


(Photo J. Huntington)

Fig. 4 Indrasaila legend (detail), LucknowState Museum.


(Photo A. Peres after Takata)

Fig. 5 Buddha preaching at Safikasya, from Bharhut stupa railing,


IndianMuseum,Calcutta.
(Photo Indian Mus.)

Fig. 6 Railing post from Tsapur,MathurdMuseum.


(Photo Mathura Mus.)

Fig. 7 Bodhisattva with Brahma and Indra, LucknowState Museum.


(Photo Lucknow Mus.)

Fig. 8 Uninscribed Bodhisattva, MatburdMuseum.


(Photo P. Myer)

Fig. 9 Bodhisattva dedicted by Namda, MathuraMuseum.


(Photo Mathura Mus.)

Fig. Io

Katra stele, MathureaMuseum.


(Photo Mathura Mus.)

Fig. I

Stele from Ramnagar dated year 32z,National Museum,New Delhi.


(Photo P. Myer)

Fig. I2

Bodhisattva from Pllkhera dated year 39, IndianMuseum,Calcutta.


(Photo Indian Mus.)

Fig. x3 Bodhisattva dedicated by


Bala at Sarnath in year 3, SarndthMuseum.
(Photo Sarnath Mus.)

A
Fig. 14 Indrasaila legend, Indian Museum,(Calcutta.
(Photo Indian Mus.)

Fig. IS5 Buddha addressing a king, LucknowState Museum.


(Photo A. Peres after Takata)

Fig. I6

Bodhisattva (?) from Ganeshra, LucknowState Museum.


(Photo Lucknow State Mus.)

Fig. 17 Bodhisattva (?) from Ganeshra, right rear view,


LucknowState Museum.
(Photo P. Myer)

Fig. 20 Image from Anyor dated year 5i, MatburaMuseum.


Mathura
(Photo
Mus.)
(Photo Mathurd

Fig. 2z

Mu
Standing image,Mathura

Mus.)

Figs. 22 a, b Buddha from Maholi, MathuraMuseum.


(Photo Mathura Mus.)

seum.

It
O'P
u3

O'
41
00

Jetavanaramaat Sravastiand is now in the IndianMuseumin Calcutta,while its inscribed


umbrellastaff is in the Lucknow Museum.36Although the dates of its inscriptions(L.9I8,
L.9I9) are obliterated, they tell us that the image was a Bodhisattva dedicated by the bhiksu

of the Kosamba
Bala,masterof the Tripitaka(trepitaka),at the Buddha'swalk (chankrama)
vihara for the acceptance of the Sarvastivadin teachers. Its head is badly damaged and the
right arm and hand are lost, but in size, iconography and style it is virtually identical with

that dedicated by Bala at Sarnath except that a cluster of lotus-flowers occupy the space
between the legs where that one has a lion.
The often-illustrated Sarnath Bodhisattva, discovered in I904-o05 near the Asoka
column, bears three inscriptions, one on the front of the base, one on the back of the
undergarment, and a long one on the shaft of the great stone umbrella or chattrathat once

shelteredthe figure (Fig. I3).37 Like the Sravastiimage it has lost its right arm and hand,
and the badly-abraded head was once broken off and has been re-affixed. A barely
recognizable lion squats between the feet and a cluster of lotus flowers and leaves are
carved next to the left leg to support the weight of the falling folds of the robe, while traces
of color on the breast and thorax show that it was once fully polychromed.
A smaller image of identical type, found in the ruins of the venerable Ghositarama at
IKausambi (Kosam) and dated in year (22 ?), names the nun Buddhamitra trepitikaye as

donor.38Presumablythis is the same learned lady referredto in a Mathurainscription


(L.38) dated year 33, which identifies her as a pupil of the bhiksu Bala.39While it is
conceivableBuddhamitramay have been the first to dedicatea standingBuddhistimage,
it seems more probablethat Balawas the innovatorand she was emulatinghim, in which
case her image probably should be dated year 22, as suggested by Ghosh. Some confirma-

tion of the later date appearsin the carvingof the folds over its arm, which seem slightly
less finely cut than those of Bala's Sarnathand Sravastiimages.
The upper portions of these images are very similarto those of the seated ones. It is
apparentthat their right hands performedthe abhaya-mudraand their torsos are fully
revealedby theirtransparentgarments,but theirleft handsare clenchedand the folds over
their arms extend from shoulder to wrist. Their skirt-likeunder-garmentsresemblethe
undercloth worn in Thailand by monks of the HYnayanabranch of Buddhism, being
brought round the hips so that the long pleatedends hang down evenly in front and the
upper portion is turned down all around and secured by a ribbon-like belt or girdle
(kayabandhana)tied on the right hip.40 It has generally been assumed that they wear a single

Indian Museum, Calcutta, Si.B. John Anderson, Catalogueand Hand-Bookof the IndianMuseum.Part I (Calcutta, 1883,
reprinted New Delhi, i977), pp. i94-95. For inscriptions, see H. Liiders, "Set-Mahet Image Inscription of the Time
of Kanishka or Huvishka," Epigraphia Indica, VIII, 80o-8I; T. Bloch, "Inscription on the Umbrella Staff of the
Buddhist Image from Sahet Mahet," EpigraphicaIndica, IX, 290-91; Liders, EpigraphicaIndica, X, Appendix, pp.
92-93?.
37Supra, n.6.
38Allahabad Mus., AM 69. Pramod Chandra, Stone Sculpturein the Allahabad Museum(American Institute of Indian
Studies, n.p., n.d.) pp. 61-62, pl. XXXVII.
39Liiders-Janert, pp. 54-5 5.
40 A.B. Griswold,
"Prolegomena to the Study of the Buddha's Dress in Chinese Sculpture," Artibus Asiae, XXVI, i
(1963), 87-88, fig. i.

36

I27

voluminous overgarment, but careful examination shows that the sculptors depicted two
separate garments which must be identified as the uttardsangaand sanghati, the second and
third of the Three Garments (trichbvara)comprising the monastic habit specified by the
Vinaya, the rule of life for the bhikshus and their female counterparts.41The uttarasamga
or robe, like the seated Bodhisattvas' uttariyas, is worn in the "open mode," the cloth

passing around the back (where its upper edge is concealedby the images' haloes) and
under the right arm and up and over the left shoulder and upper arm, the free end or
overthrow being brought forwardand held in place by the left hand. The samghatior
overrobeis loosely drapedabout the legs, its upperedge encirclingthe hip while the hem
is gatheredup into a thick roll.42Both these garmentsappearto have been madeof a fine,
stretchyfabricwhich has been dried in pleats but is drawnsmoothly over the body. The
folds of the uttarasamghaand the hem of the samghatiare indicatedby ridges, each one
with a lightlyincisedline, while the morewidely spacedfolds radiatingfrom or converging
on the left hip are indicatedby shallow grooves which are separatedfrom each other by
pairs of shallow incisions.
The inscriptionsof these images tell us that they (and presumablythe Sravastione as
well) were installedat chankramasor promenadeswhere the bhiksuswere accustomedto
practice walking meditation. This almost certainly explains their stance, since standing
figures are obviously more appropriate than seated ones for such a location. Since the
monasteries where they were erected were ancient ones hallowed by their associations with
events of the Buddha's life, they attracted pilgrims from every quarter of the rapidly-

one may suppose that these figures, with their radiant


Buddhistandi
expanding
haloes, powerful forms and impressivesize, must have appearedto the pilgrims almost
as impressive and awe-inspiring as the illusion created by Mara at the behest of Upagupta.

It can scarcelybe doubted that the type of these images originatedat Mathura.Not

only are their materials and style unmistakably Mathuran and their forms adapted from

the earlier Bodhisattvaimages executed here, but we know that both Bala and Buddhamitra were natives (or at least residents) of this city. Judging from the number
discovered, standing images of this type must have been very popular during the first

half-centuryor so of the Kanishkaera,but none of them appearto predatethe datedones.


An unusuallywell preservedone, brokenat the waist but retainingboth arms,head, and
most of the halo, was acquired for the Mathura Museum from the Govindnagar mounds.43
Its proportions and carving are very close to those of the fine one from Lakhnau (Aligarh

Dt.), which is dated in the year 35 in the reign of Huvishka.44Both show a similar
reductionof the thick folds, which have become simple roundedridges, and the folds of
the overrobe, which are treated as widely spaced incisions. Most of the others are more

or less clumsy variationson the type.

41 Ibid.,

pp. 88-90, figs. 2-a,3.


AM 71 clearly shows the distinction between these two garments (Chandra, P1. XXXVI and pp.
60o-6 ).
43 Mathura Mus., 71.I05. Sharma, p. 183, fig. 92.
44 Mathura Mus., oo.A63. Vogel, Cat. A.M.M., p. 62; Bijutsu Kenkyuj,
234 (May i964), pl. V.

42 Allahabad Mus.,

28

Interestinglyenough, a stylisticallypre-Kanishkarelief found at KankalilTila offers


evidence that Mathuraknew an earlier type of standing Sakyamuniimage (Fig. I5).45
Although little Buddhist material has been identified here, van Lohuizen-de Leeuw
recognized this as the depiction of an unidentified episode from the Buddha's life, for it
shows a kingly figure, accompanied by three attendants, confronting a taller man whose

halo, gesture and head crowned by a low twist of hair unmistakablymark him as
SakyamuniBuddha.While the slim-hippedbody of this figure resemblesthat of the Bala
Bodhisattvas and contrasts with the more obese forms of the king and his minister, all three
are similarly dressed except that the Buddha wears no jewelry. Their legs are swathed in
dhoti-like paridhanaswhose ends hang down in fine pleats and long uttariyas are loosely
slung about them, one end hanging down along the left side and the other falling over
the left wrist. It would require only slight readjustments - tightening the cloth around the

torso and throwing its end over the shoulder,bringing the other end around the upper
armand tucking it back behindthe figure- to achievepreciselythe effect depictedin the
seated Bodhisattvaimages.
Some writers assert that all of these images are recognizable because they display the

one born to become


of a GreatBeing (Mahbpurusa),
thirty-twoGreatMarks(Mahalak.sana)
either a World Ruler (Chakravartin)or a Buddha, which were recognized by the seers who
predicted the future of the infant Siddhartha.46Study of the traditional lists show that the
lakshanasseem to have come from several sources. A few of them seem to refer to physical
peculiarities or even abnormalities that may well have been considered auspicious.47Many

seem to reflect traditionalIndian ideals of beauty,48while a few can be best understood


as metaphorsexpressiveof qualitiesappropriateto a heroic ruler.49A numberof them are
incapable of depiction in sculpture,50and some are so variously or ambiguously stated that

it is impossibleto know preciselywhat was intended.51


The most puzzling and most discussedof the Mahalakshanasis usn-sa-s'iraskata
(Pali,
In most periods and regions of the Buddhistworld this has been understood
nhzisasisa).
to refer to a protruberance on the top of the head. Its literal meaning seems to be that the

head has the shape of a turbanor cap, but it has been variouslyinterpreted.The learned
commentatorBuddhaghosa(fifth centuryA.D.), thought that it signifieda broadand full

45 Lucknow State Mus., J.531. Van Lohuizen-de Leeuw, pp. I58-I6I; Takata, Plate 48, fig. 56.
46 The Pali and Sanskrit lists vary slightly, especially in sequence. We have used the Sanskrit list from Lalita-vistara as

presented by Har Dayal, The BodhisattvaDoctrine in BuddhistSanskrit Literature(reprint, Delhi, I970), PP. 300-305.
474, between his eyebrows a white [tuft of?] hair; 7, forty teeth of even size; I8, when standing erect and not bending,
hands reach down to the knees; 23, penis concealed in a sheath; 3 , wheels on the soles of the feet.
48 2, hair turning towards the right in dark blue locks; 3, forehead even and broad; 5, eyelashes like a cow; 6, very dark
pupils; 8, no gaps or interstices between the teeth; 9, white teeth; I4, evenly-rounded shoulders; 15, seven convexitites
or prominences (i.e., backs of arms, legs, shoulders and trunk); i6, space between shoulders "heaped up"; 17, fine
skin of the color of gold; 20, body of the symmetrical proportions of a banyan tree; 24, well-rounded thighs; 26, long
fingers; 27, long heels.
49 I3, jaw like a lion's; 19, front part of body like a lion; 25, legs like an antelope's; 32, feet well-set or well-planted.
50 I0, excellent voice; I I, acute and keen sense of taste; 12, large and slender tongue; 21
i; each hair on the body rising
feet.
and
hands
and
delicate
soft
straight upward; 29,
51 28, ankles prominent (or the reverse?); 30, hands and feet webbed or netted.

129

forehead and well-rounded head,52 and several modern scholars have advanced other

explanationsfor it. Thus Coomaraswamyconcluded that it should be understood as


while van Lphuizen-deLeeuw arguedthat it refers
meaning"destinedto weara turban,"53
in
to the spiralcoil (or, Gandhara,bun or chignon) of hair aroundwhich a turbancould
be wound and that its connotationis "having a head fit for a turban."54While all three
of these explanationsare visibly consistent with the appearanceof the early Mathuran
images (and early GandharanBuddha images as well), it must be noted that Buddha images
from the Andhra region of South India display modest but distinct cranial protuberances

covered, like the rest of the head, with spiralcurls. Although the earliestof these, which
are laterthan those from Mathura,they indicatethat an alternative
come from AmaravatT,
tradition regardingthe Buddha's head was known there. Both southern and northern
accountsof the youth of the futureBuddhaagreethat afterleaving his father'spalaceand
divesting himself of his jewels the young prince cut off his hair and his chuda (the hair
together with the turban wound around it) was transported to the Trasyastrimsa Heaven,

there to be venerated by the gods.55 The Sanskrit Mahdvastu,a Sarvastivadin text, says
nothing more about the hair, but the Nidanakatha, the Pali introduction to the Jataka
commentary, goes on to say that the "hair was reduced to two inches in length, and curling

lhe
from the right, lay close to the head, remaining of that length as long
as
precisely

as seen in the images

lived,t"56

and reliefs from Amaravati.

Finally it may be observed that Mathura images do not display the disproportionately
long arms specified in all Mahalakshana lists, which say "when he is standing erect and
not bending, his arms reach down to the knees."57 Indeed such very long arms are rare

in Buddha images, although they appear much later in Thailand and occasionally in Japan,
and in medieval

Indian

two conimages of the Jain Tirthankaras as well. Moreover,


spicuous features of the Mathuran images, namely the wheel-marks on their upraised hands
and the swastikas on their toes, are not listed among the Mahalaksanas, even though they

undoubtedly come from the ancient repertory of auspicious signs. In short, the Bodhisattvas from Mathura cannot be said to display the Mahalaksanas distinguishing a Buddha.
While Buddhist literature contains several references to images of Sakyamuni produced
during his life or shortly after the Parinirvana, most scholars agree that the "invention of
the Buddha

image"

took

place

shortly

before

the advent

of the Kushanas

and was

stimulated by the rise of new attitudes and practices which called for images to receive
the devotion of the faithful.58 S. Dutt has thus summarized this view:
J.N. Banerjea, "Usnzsa-siraskata (a mahapurusa-laksana) in the early Buddha images of India," Indian Historical
I
Quarterly, VII (I 9 3 1), 5oo00-0o
53 Ananda K.
Coomaraswamy, "The Buddha's cidd, Hair, usnisa and Crown," Journal of the Royal Asiatic Society of Great
Britain (1928), 829-35.
54 Van Lohuizen-de Leeuw,
pp. I63-67.
55
Coomaraswamy, "Buddha's ciuda,"p. 821.
56 Ibid., p. 827.
57
Dayal, p. 32; see supra, note 41.
58
Huntington has argued that the literary evidence for very early seated images of the Buddha is corroborated by the
discovery of a small plaque depicting an image of the "seated Bodhisattva" type, which he dates to the Maurya period.
Unfortunately the plaque is lost, but even if it proves to date so early it would merely prove that such images were
known some 200 years after Buddha's death. See, John C. Huntington, "The Origin of the Buddha Image: Early
Image Traditions and the Concept of Buddhadarsanapunya,"Studies in Buddhist Art and Archaeology, ed. A. K. Narain
(New Delhi, I986).
52

130

It was the urge of this yearning that resulted in the invention of an image of the Lord in human
likeness which appears in the early sculpture of Mathura,-a seated image in stone in the form
of a Yogin and Superman with the physiognomical marks on him for recognition. This was
probably in the early part of the first century A.D.59

Two objections must be raised to this statement.First, as demonstrated,the early


Mathura images do not in fact display the marks or Mahalaksanas of a Superman.

Moreover,the survivingones and their possible lost predecessorswere probablymodeled


after wooden or clay (i.e., terracotta)prototypes. Thereforethe "invention"of the type
should probablybe put somewhat earlier,perhapsin the precedingcentury.
More significant are the dress and attitudes of the seated images. Unlike the standing

ones, they do not wear the monastic habit or Three Garmentsbut are garbed in the
paridhanaand uttariyainvariablyworn by representationsof gods and laymen. But they
dispensewith the jewelryand turbanscommonlyworn by those, in this respectresembling
the asceticsand hermitsdepictedin reliefsat Bharhutand SanchT.Although their legs are
locked in the lotus position (padma-asana)appropriatefor meditation, their wide-open
eyes and energetic gestures (which became more vigorous as the type was perfected) seem

more appropriatefor a god or hero offering protectionto his votaries. As we have seen,
although it was to become standard in later Buddha images, the gesture of abhaya-mudra

was by no meansrestrictedto Buddhistimagesbut appearsin manyimagesof deities,male


and female, Jaina and Hindu. In combination with the jutting elbow and down-pressed

left hand, it seems to be expressiveof inner power barely restrainedby the exertion of
perfectcontrol. In short, their characteristicsare so remotefrom those observablein most
Buddhaimagesthatthereis little to identifythem as representationsof SakyamuniBuddha
except, of course, for the pippala tree under which he achieved Enlightenment.
And finally, why are these images identified in their inscriptions as Bodhisattva? In

early Buddhistliteraturethis term is appliedeither to the youthful Siddharthabefore his


Enlightenmentor to his previous lives since he vowed to become a Buddha.While it was
once thought that these images might representSakyamuniimmediatelybefore his Enlightenment,this theory seems to have been generallyrejected.Van Lohuizen-deLeeuw
sought to resolve the problem by pointing out that "the literal translationof the term
Bodhisattva,'He whose essence (or object) is perfect knowledge,' by no means restricts
this denominationto creaturesbeforethe Enlightenment,"and found Pali authorityfor
Yet this does not
and disciplesof Buddhas."60
applyingit to "Buddhas,Pratyekabuddhas
solve the problems of the images's dress and attitude.
Given these puzzling problems, we may turn back to the period when the first images
of this type probably originated, the last centuries preceding our era when several religions

and cults are known to have flourishedat Mathura.The yaksacult was perhapsthe first

to make use of monumental images, and a number of impressively large standing images
have been found within twenty-five kilometers of Mathura at the villages of Parkham,

59 Dutt,
60 Van

p. 240.
Lohuizen-de Leeuw, p. I79.

I3

All have been generally


Nagla Jhinga, Baroda and Noh (BharatpurDt., Rajasthan).61
thought to be productsof the second centuryB.C. and all show corpulentfiguresadorned
with heavy earringsand necklaces;but their faces and hands,togetherwith any attributes
they carried, have been lost. A very similar figure, except for its seven-fold cobra-hood,

has been identified by Joshi as a Naga, one of the deities associatedwith water,62but
indistincttracesof the attributeheld in the left hand suggests that this may be a plough
or digging stick, which would indicatethat it representsBalaramawho, togetherwith his
of the Vrishni
youngerbrotherVasudeva-Krisna,were among the five heroes(panchavzras)
clan.63 The unmistakable, better preserved image from Junsuti, complete with plough,

club and cobra-hoodswhich project from the top of the rectangularbackslab,shows a


differentphysique,with broadhips but surprisinglyslenderwaist.64The discoveryat Mora
of a slabinscribedduringthe reign of Mahaksatrapa
Rajula(fatherof Sodasa)and referring
to a shrine with images of the panchavTrasled to the discovery of two life-sized statues.65

Although these have lost their heads and armsand lower legs, the carvingof their lavish
ornaments and torsos bring them closer to the style of the Katra stele, although their

protruberantbellies still resemblethose of the yaksas.Finally,a word may be said about


two terracottaheadsfound at Sonkh,66which cast some light on Upagupta'sreferenceto
"clayimages of the gods." One comes from the pre-Mauryaor EarlyMauryalevel of the
mound and the other, found near the apsidal temple, is datable to the early Kushana period.
The older, depicting a moustachioed male, is somewhat summarily modelled but carefully

detailed,but the other, which is hollow, is quite as accomplishedas any contemporary


stone image.
If, therefore, there is evidence that the people of Mathura were accustomed to seeing

images, the Buddhists of this region had precedentsfor their first images; it does not
follow, however, that these necessarilyrepresentedthe person of the BuddhaSakyamuni.
Despite the traditions that have come down to us regarding marvellous images of the
Buddha produced even during his lifetime, there were excellent reasons why his followers
for centuries eschewed the depiction of the Blessed One. Unlike the gods, he was no longer

presentin any visible or even imaginableform but had broken the fetters of the cycle of
births and enteredthe state of Nirvana,which is totally inconceivableto men and gods.
During his last life he had inhabited a physical body (Rupakaya),which had been reverently
cremated by pious laymen and its ashes dispersed; the living body he left on earth was

61 Mathura Mus., C 5, C 23, 72. 5. Sharma, pp. 13 I-32; Vasudeva S. Agrawala, A Catalogue of the Brahmanical Images

in MathuraArt (U.P. HistoricalSociety,Lucknow, 195I), pp. 75-78.

62 Mathura Mus., 17. I 303. N.P. Joshi, Mathura Sculptures (Archaeological

Museum, Mathura, n.d.), p. 79, fig. i. While

this corpulentfigure differsmarkedlyfrom the laterKushanaimagesof nagas,it may be noted that it corresponds
to that of the seatednaga king depictedon the crossbeamfound at Sonkh referredto in note I9.

63 For the early cult of the Vrishni pancavzras, see Doris

Srinavasan, "Early Krishna Icons: The Case at Mathura,"


Kaladarsana: American Studies in the Art of India, ed. Joanna G. Williams (E.J. Brill, Leiden, I98I), p. I29.

64LucknowStateMus., G.z 5. Agrawala,ShortGuide-Book..


.Lucknow,p. 41, fig. i; Sharma,fig. 2.

65 Mathura Mus., E 21, E 22. Vasudeva S. Agrawala, Mathura Museum Catalogue. Part III. Jaina Tirthankaras and other

Miscellaneous
Figures(U.P. HistoricalSociety,Lucknow, 1952),pp. 48-49; JohnM. Rosenfield,TheDynasticArts of
theKushans
(Universityof CaliforniaPress, Berkeleyand Los Angeles, I967), pp. I 5I-5 2, fig. 5I.

66Hartel, pp. 86, 96, Figs. 22, 41.

32

the corpus of his teachings(Dharmakdyd),


which had been gatheredand preservedby his
disciples.67If the story of Upagupta and Mara attests to the stubbornly human yearning
to see the sacred being (which persists in India in the practice of darsan),68it equally shows
that only the Lord of Illusion himself could hope to represent the tupakaya.
But what else could these images be except representations of the Buddha? A possible

clue may be found in Har Dayal's discussion of the meaning of the term Bodhisattva.69
Its firstpart,Bodhi-,offersno problem,signifyingas it does full Enlightenmentor supreme
knowledge, but its second part, -sattva, has been variously translated. Dayal lists no less
than seven interpretations for it, concluding that it is related to the Vedic satvan,which
signified a warrior or hero, and should therefore be interpreted as signifying a "heroic
being, spiritual warrior." This so closely corresponds to the expressive values embodied

in the seatedBodhisattvaimages that it suggests a possible explanationfor their original


significance and for the historic process by which the inhibition against the representation

of the person of the Buddhawas overcome.

It is my suggestion that the first Bodhisattva images were intended as anthropomorphic


symbols of the Three Refuges (Trisarana)which Buddhists still invoke with the ancient

formula:
I take refuge in the Buddha;
I take refuge in the Dharma;
I take refuge in the Samgha.70

This simple statement of faith, often referred to as the Three Jewels (Triratna),is said

to have been prescribedby the Buddha himself when the disciple Mahanamansought
admission to the Order.71Dutt, however, observes that it is inconsistent with the Buddha's
teaching, very shortly before his death, that after his Nirvana the only refuge (sarana)for
his followers would be the Dharma he had taught, and must therefore reflect an early

manifestationof the practiceof bhakti,the cult of devotion to the Buddhaas Lord.72But


since the recitation of the Trisaranais enjoined and practiced by Buddhists of every school,
it must date back to a very early period.
With the passage of time, however, it must have become increasingly difficult for

devotees to rememberthat such images were only symbols. Their stabilityand vigor may

67

Strong, pp. 105-109.


68Diana D. Eck, Daran, Seeingthe Divine Imagein India (Anima Books, Chambersburg, PA, 198I).
69
Dayal, pp. 4-9.
70

Dr. Soper has kindly reminded me of a passage in the Sarvastivadin Vinaya, translated into Chinese in 404 A. D., which
seems very apposite: the Elder Anathapindada, seeking permission to erect and adorn stupas, says, "Lord of the World,
since it is not permitted to make a likeness of the Buddha's body, I pray that the Buddha will grant that I make

likenessesof his attendantBodhisattvas"(AlexanderCoburnSoper, "EarlyBuddhistAttitudesTowardthe Art of


Painting,"TheArt Bulletin,XXXII [195o], I48).
This suggests that the Sarvastivadin tradition retained a meniory of a period when (anthropomorphic?) symbols
were flanked and attended by bejewelled figures that later generations identified as Bodhisattvas (in the Mahayana
71
72

use), just as they interpretedthe symbolsof Trisaranaas depictionsof the personof the Buddha.
EtienneLamotte,HistoireduBouddhisme
a I'ereSaka,Bibliothequedu Museon,vol. 43 (Louvain,1958),
indiendesorigines
pp. 74-76.
Dutt, pp.

200-202.

I33

well have been understood as expressive of the qualities practiced through countless lives
by that being who ultimately achieved Enlightenment beneath the Bodhi-tree at Uruvela,
and gradually they came to be understood as depictions of him.73This is almost certainly
true of the Katra stele because of the presence of the Wisdom Tree, and probably of the
similarfragmentarystele dedicatedby Namda as well (Figs. 10, 9). But even in these it is
possible that the seated images were "correctly" understood as symbols even though
worshippers interpreted them as representations of the Buddha.
Similarly, it is difficult to be entirely sure whether the seated figures in the depictions

of episodes of the Buddha's life, as on the railing-post and gate-bar discussed, were
intended as representationsor symbols. Elsewherethe Buddha'spresenceis invariably
symbolized, as by the throne, Wheel or Tree, and Mathura artists may have adapted the
seated Bodhisattva to signify that even the gods take refuge in the Trisarana. Some support
for this interpretation comes from a probably slightly later representation of the Indrasaila
Cave episode (Fig. I4).74 Here the figure seated within the cave is much smaller than either
Indra or his attendant and, moreover, is partially obscured by the mouth of the cave. One
can scarcely doubt that had it been intended to depict the person of the Buddha himself
the seated figure would have been given a scale and compositional prominence commen-

suratewith his importance,but in this case it looks very much as though the artist was
thinking of a relativelysmall image of the "seatedBodhisattva"type.
The relief depicting Sakyamuniconversing with an unidentifiedking, on the other
hand, clearly shows that the seated Bodhisattva type had become so fully identified with
the historical Buddha that the sculptor envisions such a figure standing, his upper garment

loosened and hanging about him in the customarymode (Fig. 15). This contrastssignificantly with the standing Bodhisattvasdedicated by Bala and his pupil, the nun Buddhamitra,which representa differentadaptationof the seated Bodhisattvatype. But the
treatment of the lower portions of the standing Bodhisattva images are formally derived
from another early type represented by the splendidly bejewelled figure from Ganeshra
(Figs. I 6, I7).75

This is a monumentalbut headlessimage whose stance, gestures and draperyclearly


3). The dressof this figure resemblesthat
anticipatethose of the BalaBodhisattvas(Figs. 13).
his garmentsare extravagantlylong.
of the figures
the of the
T lat
Ka
Tfigures
His uttariya hangs down to the calf and its free end is disposed in a lavish loop that falls
over the left arm, while one of the pleated ends of his transparent paridhana loops down
to the knee and is gathered up and grasped by the left hand. The great swag formed by

It may be remembered that early Christians also eschewed the representation of their Lord, and it was not until the
fourth century (when Christianity had become the official religion of the Roman Empire) artists began to represent
him, modelling their images after the symbolic figures of Orpheus and the Good Shepherd.
74 Indian
Mus., Calcutta, M 7. Vogel, "Sculpture de M.", pp. 59, 122, P1. LIII-b.
75 Lucknow State Mus., B izb. Smith, p. 43, PI. LXXXVII; Agrawala, Short Guide,p. 15, P1. 7. Ganeshra, which lies
a few miles west of the city, appears to have been an important early Buddhist center. The three mounds outside the
village have not been archaeologically investigated but fragmentary carvings and inscriptions datable to the Kushan
and Ksatrapa periods have been found here, and inscribed bricks show that some kind of a building was built here
as early as the beginning of the last century B.C. (Liiders-Janert, pp. I 56-60).
73

I34

the upper garment resembles that of the Bala-type samghatis, and its folds or pleats are
similarly carved as close-set ridges incised with fine grooves. His jewelry, which is of

unprecedentedsplendor,consists of cuff-likebracelets,a long necklacecomposed of pearl


strandsconnectedby flat metallic(?) placquesand tied at the back, a torqueencirclingthe
neck, and a magnificent girdle. This appears to be made up of long fibers (or possibly very
fine pleated muslin) bound at intervals with rings, and it is tied in a temporary square knot,
its longer end terminating in a knot and the other finished off with an ornament resembling
a compound flower from which dangle strings of seed pearls. Even the cushion-like mass
of stone that reinforces and supports the raised hand and arm is adorned with bands of

miniaturepearlsand leaves, like that of the Katrastele (Fig. i).


Stylistically the anatomy of this figure is clearly earlier than that of the dated images.

The awkwardposition of the right hand and arm, which are virtuallyidenticalwith that
seen in Amohini's tablet (Fig. 2), reveals the tendency toward flatness characteristic of the
earliest Indian reliefs and suggests that it may possibly have been adapted from such a
work. The torso, however, is very close to that of the Bala dedications except for the
sloping shoulders and slightly misplaced nipples, and the sculptor has shown how the flesh

of the hips is constrictedby the girdle and the swelling shape of the thighs. Its identity
remains a puzzle for it lacks the obesity characterizing the monumental yaksas. It has been

generallyreferredto as a Bodhisattvabecauseits splendiddress and adornmentsresemble


those of GandharanBodhisattvas,but these belong
theto
Mahayanabranchof Buddhism,
which did not develop until the Kushana period. If, then, it is unclear whether this
represents an otherwise lost type of Bodhisattva or a yaksa or belonged to some other cult,
it or very similar images were adapted for later Buddhist images, not only for those
dedicated by Bala but also for those representing Maitreya, the Buddha to come.
The earliest unmistakable example of the Maitreya type is an uninscribed figure now
in the National Museum in New Delhi (Fig. 8).76 It bears no inscriptionand its head and
arms are missing, but its stance, costume and jewelry are so very similar to a smaller one

found at Ramnagarand explicitlyidentifiedby inscriptionas Maitreyathat there can be


little doubt that it also depicts the Future Buddha.77A thick necklace of twisted strands
of pearls encircles the throat and a longer one, made up of chains and pearls fastened by
a metallic placque, falls over the chest where it turns to form a V-shape. The paridhana
is tied with a simple kayabandhaand is so sheer that it reveals the genitals, its uneven pleats
falling down to the ground in a conical mass, while the uttariya (whose

are broken
ends

off) falls from the left shoulder as a narrow roll of pleats and is brought around the right
leg to be held in place by the now-broken left hand. The right hand was presumably raised
in abhaya-mudra while the other probably held a water vessel (kamandalu)of the type
traditionally carried by Brahmans, the usual attribute of Maitreya in Kushana art.78The
broken head probably resembled that of the Maitreya carved on a railing-post found at

76

Mathura Mus., A 40, now on loan to the National Museum, New Delhi. Vogel,

pp. 23 1-32, fig. 54.


77
Mitra, "Three Kushan Sculptures," pp. 63-64, Pl. I.
78 Rosenfield,

pp. 229-33.

I35

Cat. A.M.M.,

p. 56; Rosenfield,

Jamalpur,79 which shows much the same smooth cap-like hair as the Sarnath figure

dedicatedby the TrepitakaBala, except that the latter undoubtedlywore a snail-shaped


top-knotwhile the crown of the former'sheadis smooth. The factthatthe Jamalpurfigure
is also provided with a sunshade or parasol, like that of Bala's Bodhisattvas, suggests that

a similarstructuremay possibly have shelteredthe broken Maitreya,although this must


remainin the realmof speculation.The tiny brokenfigures kneelingon either side of the
base probably represent two of the gods of the Tusita heaven where Maitreya now dwells.
Like the earlier one from Ganeshra, this image had no halo, while both the Ramnagar
and Jamalpur Maitreyas display very large aureoles with inverted scallops forming star-like
points, precisely like those of the Katra stele and Bala's two dedications. This suggests
that it was probably commissioned by an adherent of the Sarvastivadin school, for it is
a tenet of this sect that Maitreya is still a layman (prthag'ana) because he has not yet cast

off the bonds (samyojana)


that hold him fast to the cycle of rebirths.80
It is well known that
the Sarvastivadin school, whose name (literally, "all exists") refers to its doctrine that the
past, present and future are all real, was deeply concerned with questions regarding the

successivestagesof the Bodhisattvacareerand thereforewith the natureof Maitreya,who


according to the tradition accepted by Buddhists of every school is only awaiting the time
for his last birth.81 It is also possible that the Sarvastivadins of Mathura were willing to

acceptimages of the future Buddhaat a time when they still avoided the representation
of Sakyamunihimself.
Given the strong similarities between these Maitreya images and the one found at
Ganeshra, it seems possible to suggest that the older one may have also depicted the Future

Buddha, even though it lacks the identifying attribute of the water-bottle. What is
unmistakableis the artistic relationshipbetween the headless Maitreyaand the early
standing Bodhisattvas, for their stances, proportions and surface modelling are se close
that it seems certain they were produced in the same atelier and probably even by the same

sculptor.
A broken image unearthed in Mathura City is more unusual, both for its style and
subject(Fig. I9).82 Its inscriptionidentifiesit as an image of KasyapaBuddha,the last of

the Buddhas preceding Sakyamuni,but its proportions, stance and costume are much
closer to those of a yaksathan any of the early Bodhisattva-Buddhatypes. It is close to
the monumental image of Manibhadra found at Pawaya,83 at least to judge from the

surviving portions. It standseasily with the weight resting on the right leg and the left
knee slightly flexed, the left handhanging down to graspthe folds of the pleateduttariya,
image except that it is diagonally bound and both ends terminate in what appear to be

79 Lucknow State Mus.,


B83. Coomaraswamy, History, pp. 63, 233, P1. XXI-79.

du Petit Vehicule,Publications de l'Ecole franSaisd'Extreme-Orient, t. XXXVIII


Bareau, Les Sectesbouddhiques
(Saigon, 1955), P. 144.
81
Lamotte, pp. 666-67, 693-94.
82 Mathura
Mus., oo.2739. Agrawala, "Buddha and Bodhisattva Images," pp. 75-76; idem.,Studies,pp. 152-54 (reprinted
fromJUPHS, X, Pt. II [Dec. 1937], PP. 35-8).
83 Gwalior Mus. Coomaraswamy, History,
p. 34, P1. XVIII-63.
80 Andre

I36

beaded tassels,84and the pleats of the uttariya are very close to those over the arm of the
Katra Bodhisattva. This, together with the general fluency of the modelling, suggests that

it was executed early during the Kushanaera, probablyin the first years of Kanishka's
reign.
The belief in previous Buddhas was well established by the third century B.C., for
Asoka's Nigali Sagar edict says that the emperor visited and subsequently enlarged the

stupa of the Buddhapreceding Konakamana(the BuddhaprecedingKasyapa),and the


sculpturesof the Bharhutstupa railing depict the Bodhi trees of five of the six Buddhas
precedingSakyamuni(Fig. 5).85It seemstheremayhave been a cult of the formerBuddhas
by the first century, for a number of dedicatoryinscriptionsfrom Mathurainclude the
phrase "for the worship of all Buddhas". Two of these (L.97b and L.97d) are attributed
to the Ksatrapa period and allude to the Samitiya and Mahasanghika schools; another
(L.79b), found in the same place and at the same time as the Kasyapa image and dated in

the year I6 of the Kanishkaera, also refersto the Mahasanghikas;while a fourth (L.77a)

undoubtedly pre-dates Kanishka and is dated in the year 270 of an unknown era.86 It seems

reasonableto surmisethat the image of KasyapaBuddhawas made for an establishment


belonging to the Mahasanghikas, but in the absence of other depictions of former Buddhas

its resemblanceto the older yakshatypes remainspuzzling. However its dissimilarityto


the seated Bodhisattva images tends to confirm our contention that these were not
originally understood as representations of Sakyamuni.
Two images found at the village of Anyor, some eighteen kilometers west of Mathura,
are significant because they represent both the older and a new type of seated image. One

is a partially preserved stele of the familiar type. The treatment of its folds indicate it is
to be dated no earlier than the fourth decade of the Kanishka era, but its significance is
that its inscription (L. I 3) explicitly refers to it as a Buddha image (Budhaprati,8
even

though it is visually indistinguishable from earlier ones identified by inscription as


Bodhisattvas. The other is a well-preserved if crude example of a new type bearing a
striking, albeit superficial, resemblance to the usual Gandharan type of Buddha image
(Fig.

20).88

The characteristic feature of these images is that they are fully enveloped by

their mantles or overrobes so that only the head, necks and hands (and, in the case of
standing figures, lower legs and feet) are exposed. Hence they may be referred to as
samghati images.
The inscription (L. I 2a) of the Anyor samghati image is dated in the year 5I and makes
reference to the Mahasamghika sect. It also seems to refer to the image as a Bodhisattva,
although the loss of several letters makes this less than absolutely certain.89 The face
girdle is even closer to those of the famous Patna yaksas (Coomaraswamy, History, P1. XVIII-67 and Rowland,
fig. 25). It may be noted that these, which had been generally attributed to the Maurya period, have recently been
re-attributed to the period "about the beginning of the Christian era" (Niharanjan Ray, Mauryaand Post MauryaArt
[Indian Council of Historical Research, I975,], pp. 37-38).
85 Coomaraswamy, Sculpturede Bharhut,pp. 65-66, Pls. XXII-XXIII.
86 Liiders-Janert, pp. II5-I6,
izi,
162-64.
87 Mathura Mus., A 2. Vogel, Cat. A.M.M., pp. 48-49, P1. 8; Bachhofer, P1. 83-I;
Liiders-Janert, p. I71.
88 Mathura Mus., 65. Vogel, Cat. A.M.M., p. 63; van Lohuizen-de Leeuw, pp. I80-8I,
188-97,
fig. 39.
89 Luders-Janert, pp. 170-71.

84 The

I37

resembles that of the Jamalpur Maitreya, with strongly marked eyeballs and smiling lips,
but the hair is incised with short lines as if it had been cut in layers even though traces
remain of a topknot or chignon. Its general proportions are heavy, with very sloping

shoulders,but the folded legs are relativelythin and seem inadequateas a supportfor the
upperportions. The folds of the samghatiare treatedas incised lines, much like those of
the loosely drapedgarmentsworn by the standingimages from Govindnagarand Lakhnau, and theirwide spacingseemsto indicatethey depicta more substantialkind of fabric,
even though this is somewhat inconsistent with the clear indication of swelling breasts.
Close observation shows that the sculptor has carefully indicated how the garment was

worn. It appearsto have been at least as wide as the wearer'sheight and twice as long,
for the overthrowis long enough to be broughtforwardand graspedby the left handand
its lower portion envelops both knees and leaves enough to pass up over the right arm
before falling over the lap. The excess of both front and back edges is gatheredtogether
and grasped by the left hand, the ends falling vertically over the left leg.90 But these
descriptive elements are obscured and even camouflaged by the symmetry of the positions
of the hands, the folds descending from the shoulders in a series of U-shapes, and the
curiously similar shapes formed by the edge passing over the right wrist and the folds
falling from the left hand. The crude carvings on the pedestal or throne are even more
symmetrical, showing two lions presented enface, two worshippers, one of whom seems
to wear a monastic robe, and a meditating Buddha whose heavy body is swathed in a

samghatiwhose folds are depictedby a series of unvaryingarcs.


Among images of this type, van Lohuizen-de Leeuw has identified two which, because

of their stylistic similaritiesto the Katrastele and Bala dedications,undoubtedlypredate


the Anyor one. One is a standing figure, very well-preserved except for the lower legs,
feet and base (Fig. 2i).91 Its halo and right hand, which is raised shoulder-high and
supported by a decoratively-carved cushion of stone, resemble those of the Katra image,

and the volumes of the body and legs are revealedthrough the robe, much as in images
of the standing Bodhisattva type. As in the Anyor figure, the mantle falls over the right

arm to describea deep U-shapemarkedby incised folds, its left edge is graspedand held
shoulder-high by the left hand, and vertical folds fall from the hand. The treatment of the
hair, however, whose concave hairline echoes the upturned arcs of the smiling mouth and

lines of the neck, and whose topknot has been reducedto a smallconicalmass, arguesfor
a date several decades after the beginning of the Kanishka era. The other is an uninscribed

stele found at the Jamalpurmounds,which seems to be basedon the familiarBodhisattva


triad type.92Unlike those it is enframedand the attendantsturn inward, their outer feet
and knees projecting as far as the plane defined by the front of the throne although their
inner sides are overlapped by the elbows and knees of the central figure. As in the standing
image, the Buddha's hand is raised shoulder-high, but the folds of the samghati are more

90
Compare with Griswold, Fig. z-a, c, d.
91Mathura Mus., oo.A.4.
Vogel, Cat. A.M.M., p. 49, P1. XV-a; Bachhofer, P1. 86-I; van Lohuizen-de Leeuw, pp.
I87-87, fig. 36.
92 Lucknow State Mus., BI4. Van Lohuizen-de Leeuw,
pp. I84-88, fig. 33.

I38

deeply gouged. Moreover, while many of the fold-lines of the standing Buddha'srobe
swing down but stop short just above their lower point, thus preserving the even
distributionof lines over the surface,in this relief the incomplete U-shapesare confined
to the areaof the mantlecovering the Buddha'suppertorso, while the hanging folds fall
in an undulating or serpentine mass that flares out as it touches the throne. The squatting
lions supporting the throne are notably clumsier than those of the I(atra stele, with massive

heads and straightnecks that rise at right angles from their haunches,like those of the
Palkhera stele of the year 39 (Fig. 12).
Two other seatedimages can be identifiedas representingearlierexperimentswith the
samighatitype. One is a statuette depicting a figure seated under a pipala tree and escorted
by a single attendant(Fig. 22).93The trunk and foliage and hanging garlandsof the tree,

like the back of the majorfigure and his attendant,are carvedin low relief and the trunk
of the tree is not directlybehind the seated figure but set off to one side to balancethe
attendant. The upper part of the tree, part of the standing figure, and the Buddha's lower

left armand face have been lost, but the resemblanceto the earlierseatedBodhisattvatype
is striking. The shapes of the legs and body, with its broad shoulders and deep navel, are
fully revealed, the right hand is raised shoulder-high, and the pleats of the undergarment

fan out over the throne.The presenceof the robe is indicatedonly by the edge that curves
up to the left hand, the fish-tailshapeformedby its hanging folds, and a systemof incised
horizontal and vertical lines that apparently depict the kind of monastic garment (pdmsukula) stitched together from castaway rags.94 The lions supporting its throne and the

bulging form of the left upperarm suggest this is not very much earlierthan the stele of
the year 39 from the village of Palikhera(Fig. i2).
Palikherais also the sourceof the earliestdatedsamghatiimage yet identified,a broken
one whose inscription (L.2IC)

says it was dedicated in the year 8 by Simhaka (Fig. 23).95

The deeply indented navel of the broken torso resemblesthat of the Katra Bodhisattva
and its incised folds are also very much like those over the shoulder of that figure (Fig. IO).
The positions of the broken arms and hands can be reconstructed because the hand that

graspedthe hangingfolds must have been held waist-highand the curving shapethat rises

behind the right knee must represent the edge of the robe, while the area between it and
the body shows how the elbow was broken off. The drapery concealing the feet shows
an effort to depict a relatively substantial textile - perhaps thick cotton or raw silk or even
possibly wool-and its U-shaped folds are carved as undulating ridges, each marked by

a pair of shallowly incised lines, as if the sculptor had adaptedthe convention used for
the overrobes of Bala's Bodhisattvas (Fig. 13). The zigzag folds hanging down over the

lap are simplifiedversions of the delicateones that fall between the legs of the headless
93 Mathura Mus.,

5I4. Agrawala, "Buddha and Bodhisattva," p. 68; Annual Bibliographyof IndianArchaeology,IX (Leiden,

1934), P. I4, PI. IV-a.


94 An earlier depiction of such patchwork robes appears on the reverse of a pre-Kushana Mathura tympanum in the

Boston Museum of Fine Arts (Vogel, "Sculpture de M.", P1. LV). Such robes were to become a spectacular feature
of monastic dress in China and Japan, where contrasting colors and textures were often used for the squares and the
bands separating them.
95 Mathura Mus., 664. Agrawala, "Buddha and Bodhisattva," p. 68; Luders-Janert, pp. I67-68.

I39

Maitreya (Fig. 18), and their double or mirror-image form seems to be intended to show
they are the edges of the front and back of the garment. The base, which shows bhiksus
worshipping the Dharmachakraor Wheel of the Law, appears to be a crude copy of an
older work. Rampant winged lions support the outer corners of the throne, but their legs
and wings are clumsy compared with earlier ones (Figs. 8, 9), and their sway-backed bodies

and thick-tasseledtails resemblethose of the Katrastele. The curioushook-shapedobjects


carried by two of the worshippers defy identification unless, perhaps, they represent

lengths of cloth which are being waved as an act of adoration.96


Both these early samghati images resemble Bodhisattvas of the classical type both in
their modelling and proportions, but they may be based on different kinds of models. The
figure robed in a garment made up of stitched-together rags seems to be conceived in
essentially linear terms, as if adapted from a painting or drawing, while the broken one
was clearly envisioned in sculptural terms, as is seen in the treatment of the swag over the
lap. Like all the other Buddhist images here discussed, they display the "no fear" mudra,
but the positions of their left hands alter their expressive characters. Instead of pressing
downward and thereby suggesting heroic energy and resolution, they are fully occupied
with holding together the edges of their robes, which would otherwise tend to fall off their
left shoulders. As a result the emphasis shifts to the gesture of abhaya-mudra, to such a
degree that the sculptor of the pamsukula image balanced his composition by moving the
trunk of the tree to one side and introducing a single attendant to the right of the Buddha's

upraisedhand.

It has been generally assumed that these Mathura images depicting the Buddha swathed

in his monasticrobe betraythe influenceof Gandhara,where most Buddhaimages wear


heavily draped robes reminiscent of Roman togas. Yet it requires no such foreign influence
to explain the invention of the samghati type, for all Buddhists must have been accustomed
to seeing the bhiksus begging their daily food, as they still do in parts of Southeast Asia,
their shoulders modestly covered by their overrobes in according with the regulations laid
down by the Vinaya.97 Once people began to identify the older Bodhisattva images as

representationsof Sakyamuni,it must only have been a matterof time before a demand
developed for new types which showed him clad in the Three Garments that distinguished
Buddhist monks from layfolk and those of other sects, a demand that was satisfied, as has
been shown, both by the seated samghati images and by those of the type dedicated by

Bala only five years before that of the Pallkherastatue of the year 8.

A puzzling problem is the apparent loss of artistic quality in later images of this type.
While later images of the seated Bodhisattva and standing types lose relatively little of their

refinementof proportionsand surface,laterimagesof the samghatitype tend to be clumsy


andcrudelycarved.Moreover,the earlierones retainthe gracefulasymmetryof the mature
Bodhisattva images, while the later ones are rigidly symmetrical both in the positions of
96For
comparable example of worshippers waving cloths, see the Bharhut relief of the worship of Sakyamuni's
Bodhi-tree (Coomaraswamy, History, P1. XII-4I).
97 Under certain circumstances bhiksus are
required to adjust the overrobes to bare their right arms and it is also
permissable at times to fold them into a narrow strip or to doff them entirely (Griswold, p. 88).

140

their hands and the shapes formed by their drapery. One hesitates to attribute this to a
general stylistic change, for same images of the period, such as the great Naga from
Chhargaon dated in the year 40,98 are as fine as any Bodhisattva and show even more

compositionalfreedom. It may be possible that after Kanishka'sdeath, in the year 23 of


his era, Buddhismsufferedfrom a loss of officialand popularpatronagethat was reflected
in the quality of Buddhistimages producedunder his successor,the emperorHuvishka.
The development of iconographic types, however, generally tends to reinforce and clarify

their expressivevalues, as observed in the pre-KanishkaBodhisattvas.It seems therefore


more probablethat the changes observed in the sahmaghati
images have at least as much
to do with their meaning as with the skills of the sculptors who carved them.

It is perhaps significant and certainly appropriate that the earliest known example of
the samghati type was found at Palikhera, which is known to have been the site of a vihara
belonging to the Mahasanghika school.99 According to its own tradition this school

separated from Buddha's other followers at the time of the First Council, which tooli place
immediately following the Parinirvana, although alternative traditions said it separated
some I 37 or I 60 years later.100Whatever the reliability of these several traditions, it appears
the Mahasanghikas were always more open to the spiritual needs of lay members of the
community than were the Theravadins (the school represented by the Pali canon), who
were primarily concerned with the ordained members of the community, the therasand
their disciples.101Few Mahasanghika texts survive, but the school seems to have early
developed strongly docetic views of the nature of the Buddha. While the Sarvastivadins,
who belonged to the Theravadin or Hinayana branch of Buddhism, held that the historical
Buddha had been truly born and aged and died, the Mahasanghikas held that he was a
completely pure being who only appeared to have been born and lived and died like other
living beings.102 Thus in the Mahivastu it is said:
The conduct of the Exalted One is transcendental, his root of virtue is transcendental. The Seer's
walking, standing, sitting and lying down are transcendental. The Sugata's body, which brings
about the destruction of the fetters of existence, is also transcendental ... It is true that Buddhas
eat food, but hunger never distresses them. It is in order to provide men with the opportunity
to give alms that in this respect they conform to the world. It is true that they drink, but thirst
never torments them-this is a wondrous attribute of the great seers. Their drinking is mere
conformity with the world. They put on robes, and yet a Conqueror would also be covered
without them and have the same appearance as devas. This wearing of robes is mere conformity
with the world. They keep their dark and glossy hair close cropped, although no razor ever cuts
it. This is mere conformity with the world.103

We suggest that the development of the samghati images reflect a search for forms
expressive of the Buddha's true nature as a transcendental or supramundane (lokottara)

Mathura Mus., c I3. Vogel, Cat. A.M.M., pp. 88-89.


99 A
large stone bowl (Mathura Mus., oo.662) found at this village bears an inscription (L. I43d) saying that it was given
"for the acceptance of the (Maha)samghlyas (Mahasanghikas).. ." (Liiders-Janert, p. I65).
100 Lamotte,
pp. 3I2-I6.
101 Dutt,
I44-47.
102
Lamotte, pp. 690-92.
103 Mahavastu, tr.
I, I32-33.
by J. J. Jones (Sacred Books of the Buddhists, vol. XVI, London I949),
98

I4I

being who seemsto be a man of "mereconformitywith the world."The datedimage from


Pallkhera, with its sensitively modelled torso and ingenious depiction of heavy cloth falling
in vertical and curving folds, and the small one clad in a patchwork robe both retain much
of the sensuousness and liveliness of the classical seated Bodhisattva images (Fig. 22,23).
But those datable to Huvishka's reign, such as the standing image and that from Anyor,
seem to reject all appeals to the senses and all implications of dynamism (Fig. 20,21).
Their
surfaces do not suggest cloth and flesh and hair any more than their stances and gestures
evoke kinaesthetic responses, and their rigid symmetry seems to symbolize the immutability of one no longer chained to the Wheel of Becoming (Sansdra).*

* The catalogue of the I986 traveling exhibition KushanScurlpture:


ImagesfromEarly India (Cleveland Museum of Art,
I985, Stanislaw J. Czuma, with the assistance of Rekha Morris), includes a color plate no. 15, of the stele dated in
the 32zndyear found at modern Ramnagar (ancient Ahicchatra), Myer's Fig. II.

142

You might also like