The Circle of Ancient Iranian Studies
IRANIAN ART: ARSACID DYNASTY
The Indo-Parthian Beginnings of Gandhara Sculpture
By: Chantal Fabrégues
1941, Bachhofer wrote: "What makes a piece of sculpture a Gandhara work is,
of course, not the subject-matter, but the treatment of forms. Gandhara
sculpture is easily recognized when the style is fully developed; the difficulty
arises when one comes to the beginnings.”
Following Foucher, Bachhofer contended that since Gandharan art was certainly
derived from Hellenistic art, its earliest products must be those displaying
most strongly the influence of this predecessor.
However, the presumed Hellenistic origin of Gandharan art has itself been a
cause of controversy, as proponents of a Graeco-Roman origin have come forward.
The recent Butkara excavations have put both theories in question. Bachhofer's
definition, which applies only to a welldefined but limited portion of Gandharan
output, is no longer adequate. Again, Wheeler, in 1963, maintained that the
representation of the Buddha was an essential feature of Gandharan art.
Thus, on this basis, the problem of dating the appearance of Gandharan art was
identical with that of dating the Buddha image. All the sculptures which display
a style similar to that of the objects found in the Indo-Greek and Saka levels
at Sirkap, or which can be shown for other reasons to be early, and which do not
include representations of the Buddha, have to be designated, as they were by
present, however putting aside a fragmentary stucco at Guldara),
there is no evidence for the representation of the Buddha before the time of
Kaniska, whose coins show the Enlightened One on their reverses. It is likely,
nonetheless, that these coins reproduce sculptural images of a Gandhara style
that was already well established in the reign of Kaniska, and that they show
the actual appearance of these images.
Yet this only means that they give a terminus ante quern for a certain type of
image, which depends on the date of Kaniska's accession to the throne. The dates
most commonly proposed for his accession are CE 78 and CE 128, but the
representations on coins consequently do not provide a firm date for the
sculptures depicted on them. Moreover, it is difficult to say for how long a
time before Kaniska similar images may have been in existence. According to
Cribb (whose opinion is widely shared), "If the invention of the Buddha
image does predate his reign, it is only by a matter of a few years"
(cited, n. 8, p. 243). Cribb considered the Gandhara image to have evolved from
Mathura prototypes around CE 150, that is, by the latest possible accession year
excavations by Faccenna at Butkara in Pakistan have now brought to light new
evidence for an opening date for the appearance of sculptures with definite
They illustrate not only a particular moment in the development of the style,
but also raise the broader question of its origins. Numismatic evidence tends to
show that these Butkara sculptures are to be placed around the beginning of the
period of Parthian domination in the area. This conclusion is supported by the
comparison of the sculptures in question with some from Taxila displaying
similar subjects and style. Since all these sculptures are undoubtedly
proto-Gandharan in Marshall's sense, they raise the question of the length of
time between the first proto-Gandharan sculptures and the first reliefs which
show the Buddha image. A review of a series of reliefs, and in particular the
examination of one ornament depicted on them, indicates that the gap was not
long, and that the Buddha image was likely to have been created towards the end
of Gondophares' reign.
Evidence From Butkara
the center of the Sacred Precinct at Butkara stands the Great stūpa, rebuilt
five times, with each new construction superimposed on the previous one.
In the dome of the first Great stūpa was found a die-struck "Tribal"
copper coin (ascribed in BMC Ancient India, 232/144, to Taxila) bearing
obv. "crescent on hill" and rev. "taurine." Göbl, following
Durga Prasad, was inclined to attribute this issue to Chandragupta
Maurya, though recent opinion tends to regard the "crescent on hill"
as the rājānka of Aśoka.
In either case, the earliest building would be dated to the third century BCE A
painted inscription on an earthenware pot found on the north side of the Sacred
Precinct called that stūpa Dharmarājika.
It is thus likely that it owes its origin to Aśoka,
who is called in the Divyavadana the Dharmaraja. For this reason, among others,
Sir John Marshall attributed the Dharmarājika stūpa at Taxila to Aśoka
Techniques used in both constructions, in which the hemispheres of the domes are
cut by walls radiating from the central axes, are consistent with the hypothesis
of construction during the oldest period, that is to say, contemporary with the
Shapur Tope near Shah-Dheri,
and the Jain stūpa at Kankali Tila.
The final phase (numbered 11) of the Great stūpa at Butkara is dated by an
associated square copper coin of Menander (bearing obv. elephant's head, rev.
club) to the end of the second or beginning of the first century BCE.
begin with phase III of the Great stūpa. This stūpa is circular, decorated
with Corinthian pilasters, and has four stairways disposed at the cardinal
points. The pradaksinā path around the stūpa is surrounded by a circle
of bases for freestanding columns. On the northern side, along the east-west
tangent to that circle, there are two more columns (33 and 135), and four richly
decorated subsidiary stūpas (14, 17, 27, 133). On the basis of stratigraphy the
excavators deduced that column 135 and phase III of the Great stūpa were
contemporary. Furthermore, stūpas 14 and 17 either belong to the same period or
are immediately successive to it. A date for the construction of these several
associated monuments is provided by one of the finds in a reliquary recess in
column 135: two bowl-shaped relic caskets of green schist, full of objects. A
coin (cat. no. 5229)
in one bowl was at first attributed to the Indo-Scythian ruler Azes I (fig. 1);
later, however, (probably in the light of the discussion by Jenkins)
attribution to Azes II was proposed, suggesting a terminus post quem for the
construction near the end of the first century BCE
The importance of this coin as a key to the absolute dating is obvious. Jenkins has shown that the silver issues of Azes II (fig. 2) can be distinguished from those of Azes I by the fact that their obverse displays the king mounted with a whip (hereafter KMW), rather than the king mounted with a spear (KMS) as for the earlier ruler. The silver issues in turn are linked to the corresponding bronze issues by their monograms. The piece in question here, a bronze with obv. type "Tyche" and rev. type "Hermes to left," belongs to the series of Azes II. With regard to its position within the reign, Gdbl observed that the specimen seemed to be overstruck on the bronze issue of Azes II showing obv. "bull" and rev. "elephant," the overstrike being in fact obverse on reverse. This overstrike confirms the sequence proposed by Jenkins, according to which the coin found at Butkara represents the last issue but one in the series of Azes II at the mint in question.
It is difficult to establish the precise dates of Azes II. The founder of the Indo-Scythian dynasty, Azes I, established the so-called Vikrama Era of 57 BCE The coin-series makes clear that Azes I was succeeded by Azilises, and the latter in turn by Azes II at a date for which there is no direct evidence. Allowance for generations suggests that the last ruler's accession might fall between 15 and 5 BCE A closer dating of Azes II is provided by the evidence relating to Indravarma.
stone relic-casket of uncertain provenance, known as the Avaca or Indravarma
casket, bears a Kharosthi inscription dated in the year 63 of the Maharaja Azes
deceased (maharayasa ayasa atidasa).
The inscription alludes to one Ramaka, therefore relating it to a further
dedication by the same Ramaka dated to the year 74 of the "Maharaja of
ancient times" (maharajasa purvakalisa).
Doubtless the era is that of 57 BCE and the resulting dates consequently CE 6
and CE 17. Since in CE 6 it is stated that there had been a former king Azes,
who was deceased, the implication could well be that the younger king (Azes II)
was still alive and reigning, not only at the first of the two dates, but
possibly also at the second. The reign, at any rate, presumably ended before CE
20, the year in which, on the evidence of the Takht-i Bahi inscription, the
Indo-Parthian emperor came to the throne.
can therefore conclude that column 135, phase III of the Great Stūpa, and stūpas
14 and 17 were erected towards the first decades of the first century CE, that
is during the late Indo-Scythian or the beginning of the Parthian periods. The
entablatures have been preserved on stūpas 14 and 17, which are built with
"small soapstone blocks." They consist of an architrave with two
fasciae (horizontal panels) and a row of ovoli, surmounted by a frieze
ornamented with lion protomae, above which are dentils and a cyma reversa
decorated with heart-shaped leafand-dart, and reversed ovolo patterns. The top
of the entablature of stūpa 14 has three courses of plain slabs, while that of
stūpa 17 has another frieze with triglyphs and metopes.
friezes with animal decoration provide additional support for the early dating
of these monuments. The frieze with lions' heads on stūpa 17 (fig. 3) is
punctuated by small pilasters and horizontal scrolls. On stūpa 14 (fig 2), the
lion protomae alternate with the so-called honeysuckle palmettes, eagles, or
winged cupids on lotus flowers. The lions on both stūpas have protruding eyes
with incised pupils. The alternation of floral ornament with lions' masks is
undoubtedly a motif of Greek origin, appearing on late Classical buildings where
the lions are used as spouts (cf. the Heraeum at Argos, the temple of Athena at
Priene, the Mausoleum at Hallicarnassus), and on such Hellenistic monuments as
the stoa of the agora at Priene and the Heroon at Belevi.
In the latter example, as on the stūpas 14 and 17 at Butkara, the cyma reversa
is adorned with a leaf-and-dart motif, also a common Greek ornament. Palmettes
and leaf-and-dart motifs, as well as lions' masks, similarly appear at Āi-Khānum
Although no example of lions' masks combined with honeysuckle palmettes has been
discovered in Bactria, it is probable that this combination also was brought to
Gandhara by the Graeco-Bactrians. The manner in which the manes of the Butkara
lions are composed of separated tufts with curled ends is similar to the
treatment of a lion's head used as a spout at Āi-Khānum.
Nevertheless, the lions' heads on the Butkara friezes differ from the Greek
examples and from those at Āi-Khānum in having pointed ears, resembling some
of the griffons represented on the gateways of Bharhut and Sanchi. There are
particularly close connections between the frieze on stūpa 14 at Butkara and
certain of the sculptures on the second crossbeam of the southern gateway at
On the latter the winged lions have manes similar to those of the lions at
Butkara, and are associated with honeysuckle palmettes. One feature that is
noticeable on both the Butkara and the Sanchi honeysuckle palmettes is the
depiction of the ribs. Earlier at Bharhut, the brassards in the shape of
honeysuckle palmettes worn by male figures have the same peculiarity. All these
similarities with Sanchi suggest artistic connections. Since the Sanchi gateways
are dated to the second half of the first century BCE, and Sanchi is no doubt
the earlier site, it may be regarded as providing a terminus post quem for the
The Evidence From Taxila
more precise dating, however, for the Butkara examples can be drawn in comparing
them with two friezes from the Dharmarājika stūpa at Taxila, showing alternate
lion protomae and honeysuckle palmettes. Frieze I (Taxila Museum 9631937-38; fig. 5)
is said to have been found near building L to the southwest of the Great stūpa.
Marshall claimed it belonged to building L, which, being built of large diaper
masonry, is dated to the end of the first century CE
The large diaper masonry succeeded the small diaper masonry used during the
third quarter of the first century CE The latter date derives from the discovery
(chamber G 5 in the area of the Dharmarājika) of the Taxila Silver Scroll dated
to the year 136 of Azes (corresponding to CE 78).
Confirmation is found in shrine A at Kalawan, also built of small diaper
masonry: a copper plate inscription dated to the year 134 of Azes (CE 76).
The similarities between sculptures from the Kalawan shrine and others found
near building L confirm their contemporaneity. Frieze I therefore can be dated
towards the end of the first century, and hence much later than the Butkara
friezes. Nevertheless, its examination is helpful in relation to the second
Dharmarājika frieze, our frieze II (Taxila Museum 962-193738; fig. 6), which
bears below the decorative band a Kharosthi inscription, of which only six
characters remain: sa ba se . . . dre to na. Marshall does not indicate
the find-spot. However, when listing inscriptions discovered in the vicinity of
building L, he mentions one beneath a frieze with alternating lions' heads and
honeysuckle palmettes, which he reads samvatsa . . . trena, which
suggests that this frieze formed part of our frieze II, and that it should have
come from the proximity of building L. Yet if frieze II came from building L
itself, it should have been identical in detail with frieze I from that
structure. However, the lions' heads and honeysuckle palmettes have very
upshot is that frieze II could belong to stūpa D 3, which stands at the
northeast corner of building L, in the ring of small stūpa s surrounding the
Great stūpa. (The monuments are so close together that Marshall attributes the
same sculptures sometimes to building L and sometimes to D 3.) Most of these
small stūpa s are made of rubble masonry with a facing of squared kanjur,
as is stūpa D 3. Coins of Maues and of Azes I in some of the stūpas built with
that type of masonry date their construction to the second half of the first
century BCE, and accordingly also that of D 3 and our frieze II. The
palaeography of the inscriptions belongs in this early dating. Indeed, the
aksaras sa, ba, and na have a shape similar to those on the oldest
inscriptions, such as the Shahbazgarhi and Mansehra edicts of Aśoka
(third century BCE), or on the Indo-Greek coins (second century BCE). Moreover,
frieze II has another early connection. The mane of the lion's head is there
rendered with parallel lines and a median parting as is the ceramic lion's head
from Āi-Khānum, and the honeysuckle palmettes have their extremities turned
backwards, a feature often met in the Hellenistic world.
the other hand, there are indications that stūpa D 3 is somewhat later than the
second half of the first century BCE Unlike the other small stūpas, its base is
divided into panels by eight slender pilasters. A similar division of the base
appears at Sirkap on a shrine in block F and on a stūpa in block G, both built
in rubble masonry faced with kanjur. Marshall assigns these structures to
the late Śaka
period because a coin of Azes II ("mounted king and Pallas type") and
one of Kujula Kadphises ("Roman bust and seated king") were found in
block F, and eight coins of Azes II ("king seated cross-legged and
Hermes" type) in the relic casket of a stūpa in block G. Now the intervals
between the pilasters of the shrine in block F (the famous shrine of the
double-headed eagle) form niches containing architectural facades, on top of
which are eagles.
representation of birds upon a roof is attested in early Indian art, and though
birds often appear in Indian iconography, the eagle is not prominent. Marshall
suggests the eagle was introduced in Gandhara by the Śakas.
Many representations of eagles with spread wings come from the Scythian lands
around the Black Sea, and a bronze plaque showing frontally a double-headed
eagle was discovered in the kurgan of Popovka near Kiev.
In southern Siberia also, the eagle seems to have been a favourite subject: gold
plaques from the Altai display fights between eagles and various animals; and
from the Touva region, north of the Altai, come gold pieces in the shape of
eagles with outspread wings.
In Central Asia the motif is quite rare,
but it is very common among the Parthians, who even knew the double-headed
eagle. Thus a tessera from Palmyra shows on its obverse the bust of Aglibol, and
on its reverse a double-headed eagle.
The motif may have originated in Mesopotamia, where a Kassite cylinder of the
second millennium BCE shows a similar subject.
may plausibly assume that the Parthians introduced this subject into Gandhara,
and accept the traditional dating of this shrine to the Indo-Parthian period
beginning in CE 20. Stūpa D 3 of the Taxila Dharmarājika, because of its
similarities to the shrine and the character of its frieze, should therefore be
assigned to the same period. The palaeography of the accompanying Kharosthi
inscription need not conflict with such a solution. Ramaka's dedication of 74
v.E./ CE 17 shows similar Indo-Greek letter-forms still in use at that time, and
they may even have persisted later. The Butkara stūpas 14 and 17, their bases
decorated with slender Corinthian pilasters surmounted by a frieze of lion
protomae and honeysuckle palmettes, certainly have much in common with the
Taxila stūpa D 3. All may be assigned to the early years of the Parthian
period. I have suggested that the eagles on the frieze of stūpa 14 could have
had a Parthian inspiration. The representation of Eros standing on a lotus
flower also could have been introduced during the Parthian period, since a
copper relief featuring a similar theme was found in the Parthian levels of
block C, together with a coin of Azes II, one of Aspavarma, one of Gondophares,
and five of Kujula Kadphises.
Faccenna pointed out, the friezes of stūpas 14 and 17 can be connected with two
others, MAI 2587 and MAI 6841 (fig. 7), showing below a row of ovoli a similar
succession of lion protomae and eagles on lotus flowers, and decorated on a
lower register with a row of busts. The full implication of the dating of the preceding
friezes now appears, since it enables us to place a sculptural style in the same
period. The busts are seen frontally, and each figure holds in the right hand a
bunch of lotus flowers. Between each pair of heads, filling the flat surface, is
set an eight-petalled flower, and the sequence ends with a large acanthus leaf.
The characteristic features of the portraiture are the wig-like hairstyle,
either straight or curled; the large eyes with incised pupils (a rendering very
similar to that of the lions); the long, pointed moustaches; the linear markings
on the necks; and the flat rendering of the draperies with the folds indicated
by parallel incised lines. All these characteristics are reminiscent of a toilet
tray found in the Greek strata at Sirkap (Taxila, vol. 3, pl. 144, no.
63). Similar features appear on a Hellenistic (i.e. pre-Roman) toilet tray from
and on others from the Śaka-Parthian
levels at Sirkap. They are typical also of the so-called archaic sculptures from
these Palmyrene analogies are some relief fragments from the so-called
T-foundation of the sanctuary of Bel.
They belong to the older temple, and are therefore earlier than CE 32, the date
of the consecration of the new sanctuary.
Their frontality and the treatment of the drapery are similar to that on the
Butkara fragments. Other features common to the two sites are the use of two
lines to mark the neck and the wideopen eyes. The plinth of an ex-voto, from the
early temple at Palmyra, offers resemblances not only of style but also of
composition and of decorative motifs.
It is composed in two registers, each of panels framed between Corinthian
columns and bearing a carved epistyle. One of the epistyles is ornamented with
acanthus leaves, and the other with a row of ovoli. Outside the columns are
carved long palm leaves. The upper register contains a frontal male bust with
curled hair, and, on his right and left, a griffon and an eight-petalled
the lower register appears the bust of a god with radiate nimbus flanked by
eagles with spread wings and sixteen-petalled rosettes containing rosettes with
eight petals. The busts have been identified as Malakbel. Collart and Vicari
have linked this plinth with other monuments from Palmyra which are all dated
before the middle of the first century CE One of these (Palmyra Museum BC 449) shows a draped
personage and a pine-cone, with the extremity of a wing visible to the
personage's left, in a frame of palms surmounted by a row of ovoli. Another
(Palmyra Museum B 450) shows within the frame of palms, above which is a vine
scroll, an eagle sheltering under each of its spread wings a four-petalled
rosette. Further, in the temple of Baalshamin are three niches framed with a
vine scroll, and a lintel on which are similarly depicted eagles and rosettes
that may represent Baalshamin.
Palmyrene pieces contain elements borrowed from eastern Hellenism, from the
ancient Semitic repertory, and from the newer Orient of the Parthians.
The eagle with spread wings and the rosette placed in the field should originate
in the ancient Orient; the naturalistic treatment of the eagle, the Corinthian
columns, the acanthus, the ovoli, and the vine scrolls from Greek Mesopotamia;
while the frontality is probably Parthian.
the Butkara fragments share with the Palmyrene monuments a common style, make
similar use of frontality and low relief, and also depict eagles, rosettes,
ovoli and acanthus. It is probable that the Hellenic elements existed in
Gandhara before the arrival of the Parthians. However, the manner of their
arrangement, no less than other features, resembling those at Palmyra with their
Parthian connections, suggest that they arrived via Parthia in Gandhara. The
Palmyrene evidence, with its dating horizon of CE 32 provided by the restoration
of the temple of Bel, leads to the conclusion that the Gandhara monuments
examined here can be estimated to date between CE 20 and 30. Other friezes from
Butkara displaying rows of busts executed in similar style can be assigned to
the same period (fig. 8). A fragment of a frieze carved from micaceous schist,
found between building L and stūpa D 3 at the Dharmarājika, is certainly to be
dated as early as the Butkara friezes. On its upper edge is a vine scroll with contiguous
arches containing frontal busts beneath. The triangular space between each pair
of arches holds an eagle with wings outspread. The motifs once more remind us of
the Palmyrene monuments. A similar motif of figures under arches, with spread
eagles in the spandrels, is also met in the Bimaran casket.
dating here suggested for the Gandhara stūpa elements under consideration is
corroborated by some of the brackets from Sirkap in the shape of winged male
figures with hair treated in the manner of wigs and globular eyes with incised
pupils. The folds of their draperies are indicated by multiple incised lines.
These brackets were found in strata I and II which, according to the
stratigraphic chronology, belong to the first century CE The characteristic
letter-forms (notably that of the sa) on the bottom of one bracket are
appropriate to the early first century CE
A number of similar brackets from various parts of Taxila and Butkara (e.g.,
figs. 7, 8, 9), despite some conflicting local excavation evidence, may also be
assigned to the Parthian period.
indication of the link between these brackets and the Butkara and Taxila
sculptures can be found in the jewellery represented on two of them. Two of the
brackets from Taxila, and another of cruder workmanship from Saidu Sharif (fig. 10),
wear torques represented as made of plain wire. The extremities are flattened
and twisted round the wire itself. A torque of similar form is worn by one of
the male busts on the Butkara facade (fig. 7) and by two standing male donors
also from Butkara (fig. 12). A golden torque of that type was found in the
Parthian stratum at Sirkap.
Its Parthian origin is evident in the fact that the Arsacid king Phraates (c.
3/2 BCE) wears one on his coins. Again the two donors from Butkara are dressed
in what appears to be a Parthian costume, a short jacket without a collar,
crossing at the front and secured by the belt, and the full trousers covered by
horseman's chaps. The costume is identical with that worn by the bronze Parthian
noble from Shami (denoted to General Surena), tentatively dated to the second
century BCE but possibly later (fig. 12). Similar costume can be seen on coins of Jihonika (Zeionises),
which may indicate that it was used also by the Śakas.
However, there is little doubt that the two donor figures from Butkara represent
Parthians. One of the two has his trousers, sleeves, and hem of his jacket
ornamented with an embroidered band (fig. 14). This piece, stylistically
superior to the other examples considered, might be taken as later in date, but
its donor stands on a pedestal inscribed with the Kharosthi letter ka in script
not greatly different from that of the Shahbazgarhi inscriptions. Hence this
sculpture need not be considered significantly later than those discussed
might be cast on the value of a single coin of Azes II for dating the Butkara
monuments. It is true that the coin may have been obsolete when placed in the
deposit, and the monuments therefore later, or that the coin had been inserted
into a pre-existing structure. However, similar coins were found at Taxila
associated with monuments whose decoration is comparable, adding weight to the
evidence of the coin for dating. The Azes II coin from Butkara was from the
penultimate Arachosian issue. At Taxila, the Azes II issue was associated with
specimens of the Kujula Kadphises "Roman bust" type, probably struck
after CE 10 and related to events after the demise of the younger Azes. This
association suggests that the structures were being erected towards the end of
the latter's reign, and that the work continued into the subsequent interregnum.
The Parthian features and motifs of the sculptures probably belonging to these
monuments strongly suggest that they were completed at the beginning of Parthian
domination in Gandhara. At this time, coins of Azes II were still, or had
recently been, in circulation, and so help along with the evidence of the
earlier sculptures from Palmyra to fix more precisely the appearance of
proto-Gandharan sculptures close to the moment of the inauguration of
evidence emerges for the existence of the Buddha image in association with these
structures. Thus a lower bracket for the appearance of the Buddha image can be
set around CE 30, while the reign of Kaniska, either CE 78 or CE 128, must be
taken as the upper bracket. It is not unlikely that the inaugural Buddha image
presented features similar to those seen on Kaniska coins, yet closer in style
to that of the proto-Gandharan sculptures. In fact, the sculptures from Loriyan
Tangai (fig. 14) and Butkara accord with this assumption.
One may note that on these examples the Buddha wears a moustache and his hair is
dressed in a top-knot tied with a band, as on the Buddha coins of Kaniska. Yet
the wide-open eyes, the manner of representing the hair, and the very fine
pleating of the draperies with close parallel lines are characteristic of the
work which I assign to the Parthian period. According to van Lohuizen de Leeuw,
who assumes that the Gandhara Buddha was created towards the end of the first
century BCE on the model of a Mathura Buddha, the first two features would be
due to influence from the Mathura sculptures and the latter to influence from
the early Indian sculptures. However, even though these Loriyan Tangai and
Butkara reliefs studied by van Lohuizen undoubtedly present many links with the
ancient art of India, it is clear that their style continues an earlier one
already in existence in Gandhara and represented by the proto-Gandharan
sculptures. As van Lohuizen points out in the only reference that she makes to
our protoGandharan sculptures, the latter resemble the reliefs discussed by her
in being carved out of chloritised mica schist. This material, according to
Marshall, was mainly used when the school was in its infancy, and was afterwards
given up in favour of phyllite.
A further link between the two phases of sculptures is that an Eros figured on
one of the Loriyan Tangai reliefs (fig. 14, upper left) wears a torque with the
extremities coiled around the wire, a type almost exclusively represented on the
proto-Gandharan sculptures. Both observations argue for placing the reliefs
noted by van Lohuizen in an intermediate position between the proto-Gandharan
sculptures and the representations of the Buddha on Kaniska coins. Although time
must be allowed for such pieces as the Bimaran casket and the fragmentary stucco
from Guldara, it may well emerge that these reliefs are not separated by a long
interval from the phase of the proto-Gandharan sculptures here examined.
 I would like to acknowledge the help I have received in the preparation of this article from my supervisor Dr. A. D. H. Bivar, and also the financial assistance given by the School of Oriental and African Studies, University of London, from funds made available by the Neil Kreitman Trust.
 L. Bachhofer, "On Greeks and Sakas in India," JAOS 61 /1941, 226.
 A. Foucher, L'art grew-bouddhique du Gandhara, vol. 1 (Paris, 1905), 258.
 H. Buchthal, "The Western Aspects of Gandhara Sculpture," Proceedings of the British Academy 31 (1945), 4; A. Soper, "The Roman Style in Gandhara," AJA 55 / 1951), 301-19; B. Rowland, The Art and Architecture of India (London, 1953), 79.
 M. Wheeler, "Gandhara Art: A Note on the Present Situation," in Le rayonnement des civilisations grecques et romaines sur les cultures peripheriques: 8eme Congres d'archeologie classique (Paris, 1965), 556.
 J. Marshall, Taxila (London, 1951), vol. 2, p. 693.
 G. Fussman and M. Le Berre, Monuments bouddhiques de la region de Caboul. I. Le monastere de Guldara, MDAFA, vol. 22 /Paris, 1976), p. 20, figs. 23, 24.
 Cf. J. Cribb, "The Origin of the Buddha Imagethe Numismatic Evidence," South Asian Archaeology 1981 (Cambridge, 1984), 231-44, esp. 231.
 D. Faccenna, "The Excavations of the Italian Archaeological Mission (ISMEO) in Pakistan: Some Problems of Gandhara Art and Architecture," in Proceedings of the International Conference on the History, Archaeology and Culture of Central Asia in the Kushan Period, Dushanbe 1968 (Moscow, 1973), vol. 1, pp. 161-76.
 D. Faccenna, Butkara 1 (Swat, Pakistan) 19561962, pt. 1, Istituto Italiano per il Medio ed Estremo Oriente, Centro Studi a Scavi Archeologici in Asia, Reports and Memoirs, vol. 3, pt. 1 (Rome, 1980), 21.
 R. Gobl, A Catalogue of Coins from Butkara I (Swat, ParkistanJ, Istituto Italiano per il Medio ed Estremo Oriente, Centro Studi a Scavi Archeologici in Asia, Reports and Memoirs, vol. 4 (Rome, 1976), p. 11, nos. 3, 4 (Inv. nos. 7786 and 7798), and pl. X-3; Durga Prasad, "Observations on Different Types of Silver Punch-marked Coins, Their Period and Locale," Journal of the Royal Asiatic Society of Bengal, Numismatic Supplement for 1937-38, 47/3 (1937), pp. 6162, pl. IX-5.
 . F. R. Allchin, "Upon the Contextual Significance of Certain Ancient Indian Signs," BSOAS 22 )1959), 548; P. L. Gupta, The Amaravati Hoard of Silver Punch-marked Coins (Hyderabad, 1963), 152; A. D. H. Bivar, "Bent Bars and Straight Bars: An Appendix to the Mir Zakah Hoard," Studia Iranica 11 (1982), 52.
 L. Petech, "A Kharosthi Inscription from Butkara I," E W 16 / 1966 ), 45-49.
 Marshall, Taxila, vol. 1, p. 24.
 Foucher, L'art greco-bouddhique du Gandhara, vol. 1, p. 87.
 V. A. Smith, The lain Stupa and other Antiquities of Mathura /Allahabad, 1901), 4.
 Gobl, A Catalogue of Coins from Butkara 1, p. 15, no. 15 (Inv. no. 5229) and pl. I.
 G. K. Jenkins, "Indo-Scythic Mints, " Journal of the Numismatic Society of India 17 /1955), 3.
 H. W. Bailey, "Two Kharosthi Casket Inscriptions from Avaca," JRAS /1978), 3; A. D. H. Bivar, "The Azes Era and the Indravarma Casket," in H. Hartel, ed., South Asian Archaeology 1979 /Berlin, 1981), 372; R. Salomon, "The Avaca Inscription and the Origin of the Vikrama Era," JAOS 102 (1982), 60.
Fussman, "Nouvelles inscriptions Saka: Ere d'Eucratide, ere d'Azes, ere
Vikrama, ere de Kaniska," BEFEO 67 (1980), 6.
 J. Charbonneaux, La Grece hellenistique Paris, 1970), figs. 29, 34, 35.
 P. Bernard, Fouilles d'AY Khanoum 1: Campagnes de 1965, 1966, 1967-68, MDAFA, vol. 21 (Paris, 1973), pls. 59, 101-4.
 P. Bernard, "Fouilles d'Ai Khanoum (Afghanistan)-Campagnes de 1972 et 1973," CRAI (1974), 305.
 J. Marshall and A. Foucher, The Monuments of Sanchi, vol. 2 (Calcutta, 1947), pl. XIV.
 Marshall, Taxila, vol. 1, p. 251.
 Ibid., 256.
 Ibid., 327.
 Ibid., 164.
 E. H. Minns, Scythians and Greeks /Cambridge, 1913), 178.
 M. Griaznov, Siberie du Sud (Geneva, 1969), figs. 45, 70.
 L. Hambis, "Recherches sur les Saces et leurs rapports avec les Wou-Souen," Annuaire du College de France, 1977-78 /Paris), 591.
 R. Du Mesnil du Buisson, Les tesseres et les monnaies de Palmyre (Paris, 1962), fig. 215.
 Ibid., fig. 216.
 Marshall, Taxila, vol. 1, p. 193; Taxila, vol. 3, pl. 186, no. 419.
"Excavations," 173 (cf. n. 9 above); for the reproduction of MAI
2587, idem, Sculptures from the Sacred Area of Butkara 1 (Swat, Pakistan),
pt. 3, Istituto Italiano per il Medio ed Estremo Oriente, Centro Studi a
Scavi Archeologici in Asia, Reports and Memoirs, vol. 2, pt. 3 (Rome, 1964),
 H.-P. Francfort, Les palettes du Gandhdra, MDAFA, vol. 23 (Paris, 1979), pl. L-b.
 H. Seyrig, "Antiquites syriennes: Sculptures archaiques de Palmyre," Syria 22 (1941), figs. 10, 11; Syria 24 (1945), pl. I.
 The old sanctuary is said in an inscription dated A.D. 25 to be called: "the house of the Palmyrene gods." Cf. J. Cantineau, Inventaire des inscriptions de Palmyre (Beirut, 1933), fasc. ix, p. 21, ins. no. 12.
 Seyrig, "Antiquites syriennes," 39-44; D. Schlumberger, L'Orient hellenise (Paris, 1970), fig. 13.
 P. Collart and J. Vicari, Le sanctuaire de Baalshamin a Palmyre: Topographie et architecture, vol. 1, Texte (Rome, 1969), pp. 154-77, pls. XCV-CI.
 D. Schlumberger, "Descendants non-m6diterran6ens de Part grec," Syria 37 /1960), 267.
 On the origin of frontality, cf. E. Will, Le relief cultuel greco-romain /Paris, 1955), 249; R. Ghirshman, Parthes et Sassanides /Paris, 1962/, 6-9; M. A. R. Colledge, The Art of Palmyra London, 1976/, 125-26.
 Marshall, Taxila, vol. 3, pl. 217, no. 78; vol. 1, p. 251.
 S. Konow, Corpus Inscriptionum Indicarum, vol. 2, pt. 1: Kharosthi Inscriptions, with the Exception of Those of Asoka, /Calcutta, 1929/, p. 99, pl. XIX-6 (= Marshall, Taxila, vol. 3, pl. 213, no. 11 /. Cf. other brackets in Marshall, Taxila, vol. 3, pl. 212, no. 8, and pl. 213, no. 12.
 Second Butkara donor, MAI 3514: Faccenna, Sculptures from the Sacred Area of Butkara 1 (Swat, Pakistan), pt. 3, pl. CDLXXXVI. For the Sirkap golden torque, Marshall, Taxila, vol. 1, p. 191; vol. 3, pl. 195, no. 147.
 Ghirshman, Parthes et Sassanides, fig. 99.
 Marshall, Taxila, vol. 2, p. 839; vol. 3, pl. 241, no. 200.
 J. E. van Lohuizen de Leeuw, "New Evidence with Regard to the Origin of the Buddha Image," in H. Hartel, ed., South Asian Archaeology 1979, (Berlin, 1981 /, pp. 380-81 and figs. 6, 7; pp. 382-99 and figs. 9, 10, 12, 14-17, 19, 21, 32-34.
 Marshall, Taxila, vol. 2, p. 639.
Copyright © 1998-2015 The Circle of Ancient Iranian Studies (CAIS)