0 evaluări0% au considerat acest document util (0 voturi)
69 vizualizări15 pagini
Early Activity at Amaravati by vidya dehejia. Evidence leads us to postulate the existence of a small stupa during this period to which we would attach the term post-asokan. The Early Phase also included work on the casing of the drum.
Early Activity at Amaravati by vidya dehejia. Evidence leads us to postulate the existence of a small stupa during this period to which we would attach the term post-asokan. The Early Phase also included work on the casing of the drum.
Early Activity at Amaravati by vidya dehejia. Evidence leads us to postulate the existence of a small stupa during this period to which we would attach the term post-asokan. The Early Phase also included work on the casing of the drum.
Reviewed work(s): Source: Archives of Asian Art, Vol. 23 (1969/1970), pp. 41-54 Published by: University of Hawai'i Press for the Asia Society Stable URL: http://www.jstor.org/stable/20111012 . Accessed: 05/03/2013 16:26 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. . University of Hawai'i Press and Asia Society are collaborating with JSTOR to digitize, preserve and extend access to Archives of Asian Art. http://www.jstor.org This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions Early Activity at Amaravati Vidya Dehejia History Department, University of Sydney, Australia The discoveries in recent years at Amaravati of various sculptured and inscribed frag ments executed in a rudimentary style seem to make it imperative to recognize the existence of an early stage of activity, prior to the First Period or Early Phase of scholars. The evidence leads us to postulate the existence of a small stupa during this period to which we would attach the term Post-Asokan. The terms Early Phase or First Period would be unsuitable in this context. These terms refer to the main stupa at Amaravati, and describe the period of the early simple uprights carved on one side only and simple cross-bars with one face plain, topped by a coping stone carved with animals-and-herdsmen or the dwarf-and-roll theme. Apart from work on this railing, the Early Phase also included work on the casing of the drum: the quadrant slabs were carved with pilasters and crowned with a frieze, while the ?yaka-plztiorm slabs included sculptured scenes between the pilasters. There has been a general tendency to include in this First Period certain pieces which to us represent a much earlier stage of activity and which we would place in a Post-Asokan period. The term Post-Asokan, which not many years ago would have been somewhat irrelevant as applied to Amaravati, is today an appropriate one. Recently a sandstone slab, with traces of an original polish and the fragmentary remains of an inscription, was recov ered at Amaravati. D. C. Sircar points out that the language of the inscriptions is quite similar to that of the Girnar version of Asoka's edicts,1 and it seems likely that the inscrip tion is part of an Asokan edict. In view of the difference of opinion regarding the date of the earliest sculptures and inscriptions at Amaravati, we shall first treat these on an entirely relative scheme. The resulting sequence should be mostly acceptable to all scholars, whether they place the earliest carvings and in scriptions around 200 B.C. (Srinivasan, Lalit Kal?), in the first century B.C. (Sivaramamurti, "Andhra", Encyclopaedia of World Art, p. 412; Srinivasan, Lalit Kal?), in the earlier half of the first century A.D. (A. H. Dani, Indian Palae ography, Oxford 1963, p. 72), or around 80 A.D. (Barrett, B.M.Q., p. 47). We shall later give our own opinion on the absolute dates to be attached to this early stage of activity. POST-ASOKAN STAGE A The Post-Asokan period comprises several in scribed fragments of plain pillars, cross-bars and copings, and a few sculptured pieces, and may be subdivided into two stages on the basis of palaeog raphy (Fig. 13). Stage A consists of over fifty inscribed fragments, unsculptured save for one piece. It includes Chanda 1 and 3-20; and some thirty or so inscribed fragments recovered re cently in the course of excavations at Amaravati. These inscriptions display a ha in which the hori zontal and vertical are of equal length. There is no attempt at the equalization of the verticals of any of the letters, and the forms of ga and ta reveal a distinct angularity. Early forms of bha are to be found side by side with some later ones, and similarly, angular flat-based forms of ma may be found beside some of an earlier variety. The Chanda group is carved on plain unsculp tured pillars and cross-bars, and Sivaramamurti suggests that as these pillars appear to be much smaller than those comprising the railing of the main st?pa, they belonged either to an older and more simple rail, or to the harmik? o? the main st?pa. He favours the idea of the harmik? and other scholars have since followed him on this point. In view of the growing amount of evi dence of extensive activity during a Post-Asokan period, we feel there is little doubt that these pieces formed part of a smaller early st?pa at Amaravati. A large number of fragments of cross-bars and pillars were recovered in recent excavations and 41 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions some of these have been noticed in the Annual Report on Indian Epigraphy 1959/60, although the list given there does not appear to be com plete. The inscriptions on these pieces are very similar to those of the Chanda group and ob viously belonged to the same stage. We may note that these records include two donations from the P?k?takas, who are the donors also in Chanda 8 : Chanda suggests that these are the people who later came to be known as the V?k?takas. In scriptions of this stage include the record of a senagopa Mudukutala (Sivaramamurti, No. 18) and a fragmentary record, of which there is a plate only in Burgess (PI. LVI, 7) and which we have otherwise been unable to identify or locate. A coping piece with an inscription recording a donation from the attendants of Princess Sam mali has been noticed by P. R. Srinivasan in his note on some of the early inscriptions in Lalit Kal?, 10. The coping is about 21 inches high, and while some three to six inches smaller than the general size of the coping of the main st?pa, it is not so small as to be assigned to the harmik? and surely proves the existence of an early rail. Included among the pieces recently uncovered is a cross-bar with a roughly incised st?pa and a tree within a railing, with an equally roughly engraved inscription below (Fig. 1). It may be noted that the perspective shown in the depiction of the railing around the tree seems to indicate an advance on the depictions at Bh?rh?t. Stage A includes also the inscriptions on three enormous pillars lying on the site itself. The records on two of these are contained in the Annual Report on Indian Epigraphy 1959/60, where one (No. 25 of the Report) is assigned to the 2nd or 3rd century B.C., and the other (No. 61 of the Report) to the 1st century B.C. It appears to us that all three records belong together and that it would be stretching the point to separate them thus in date. The pillars are of varying widths but are all massive?one measures 37 inches across, the other 38 inches, and the third 52 inches?and certainly prove the existence of an early railing of some sort. Belonging also to Post-Asokan Stage A are the donatory inscriptions on two slender octagonal pillars in the Amaravati Museum. The two rec ords, so far unpublished, are identical and read as follows: Acinakaputa(t?})naUtarasaGul?tasa ca thabho (Fig. 2), or "the pillar of Utara and Gul?ta, sons of Acinaka". One of the columns rises out of a rather clumsy ghat a (water-vessel) base (Fig. 3). Two pillar capitals placed above an inverted ghata with part of the octagonal column intact may also be seen at the museum. The size of the inscribed base columns and of the capital-topped columns is identical, and it is tempting to associate the inscribed columns with these capitals. The capitals depict addorsed winged animals, the palmette and rosette motifs, and are of shallow carving (Fig. 4). If indeed they belong to the inscribed columns, they are among the earliest sculptures at Amaravati. A consideration of the capitals of the Bh?rh?t torana reveals a similar shallow carving of the addorsed animals, as well as the palmette and rosette motifs. The Amaravati Museum pillars may be compared also to the pillar at Madras, but the Madras elephant-and-rider capital, by con trast, is differently conceived, more deeply cut and the carving is highly advanced (Fig. 5). On the basis of sculptural style there is little doubt that the Madras column is a later advanced ver sion of these two pillars. POST-ASOKAN STAGE B Stage B of the Post-Asokan period consists of records that display a rounded later form of ta, and a distinct tendency towards the equalization of the verticals of pa. The vertical and horizontal of ka are still of the same size. This stage includes Chanda 2; Chanda 37, 38, 43; two sculptured and inscribed pillars; and the records on a set of flat rectangular pillars at present in the court yard of the Amaravati Museum. Chanda 2 has long been attributed wrongly to Amaravati. Douglas Barrett points out that it is to be found instead on a sculptured fragment from Jaggay yapeta (Burgess, PI. LIV, 4).2 This inscription indicates that the Jaggayyapeta st?pa belonged to the Post-Asokan period, and raises certain questions which we shall consider later along with those regarding the cutting of the drum slabs of the main st?pa at Amaravati. Chanda 37, 38, 43 are all inscribed on a single slab with carvings on both sides, now in the Madras Museum. One face of the slab includes a depiction of the War of the Relics (Fig. 6). One panel depicts three elephants with relic caskets on their heads and a rider and attendant. Another panel shows a man with drawn bow and arrow, while a panel above shows men with arrows. A glance at the S?nchi gateway (torana) depictions of the scene re veals that here we have an earlier style. The relief is so shallow that it may be described as 42 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions Fig. 1. Cross-bar from Post-Asokan st?pa, Amaravati. (Archaeological Survey of India, Government of India.) Fig. 2. Inscriptions on tivo octagonal pillars, Post-Asokan st?pa, Amaravati. (Archaeological Survey of India, Government of India. ) Fig. 3. Base of inscribed octagonal column, Post-Asokan st?pa, Amaravati. Fig. 4. Addorsed animal capital, Post-Asokan st?pa, Amaravati. Fig. 5. Madras Museum pillar, Main st?pa, Amaravati. 43 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions Fig. 6. Slab carved on both sides, Post-Asokan st?pa, Amaravati. Fig. 7. Vais?li panel of sculptured pillar, Post-Asokan st?pa, Amaravati. (Archaeological Survey of India, Govern in en t of India.) Fig. 8. Finial of ?yaka pillar of the main st?pa, Amaravati. WWI^nP^EigBWWBWii^BWBBWWMUaTWCTWWIIl.l IIIIIWWWW?ww F/?. 9. N er an jar a pillar, Post-Asokan st?pa, Amaravati. (By courtesy of the India Office Library, London.) 44 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions incised or engraved. Both faces of the slab in clude representations of very simple chaitya arches, and both sides are framed with a border of the bead-and-reel motif. Belonging to Stage B of the Post-Asokan period are the labelling inscriptions on the three sides of the important sculptured pillar published recently in Ancient India 20/21. It is interesting to compare the Jetavana scene engraved on one face of the pillar with the depictions of that story at Bh?rh?t, BodhGay? and Sanch?. At Bh?rh?t the details of the story are vividly depicted, with the ox-cart in the foreground, men covering the ground with the square coins, the onlookers on one side and the merchant An?thapindaka in the centre. At BodhGay? the story has been abbre viated?the merchant himself is absent, as is the ox-cart and the onlookers. However, men may be seen covering the ground with coins, and a labourer carries a container of coins on his shoul der. In the panel on the Sanch? tor anas the scene is barely recognizable except for two layers of square coins along the front left of the panel. At Amaravati, as at Bh?rh?t, the colourful de tails of the episode are depicted. The stone is much damaged and many further details must have disappeared. The unyoked ox and the laden cart are clearly visible in the foreground. The entire ground of the relief seems to have been covered with the coins and to the right men can be seen squatting on the ground, laying out the coins. The scene depicts, however, a technical advance on Bh?rh?t in the easy manipulation of the laden cart, the casual resting ox, and the domed building to the left of the relief. The skilful depiction of the stairs to the lower left of the Jetavana scene may be noticed. It is of in terest to note that the depiction of the chaitya arch and doorway is almost exactly similar to that at Bh?rh?t. The outward slope of the door jambs is clearly visible in the representation of the shrine in the Vais?li panel on the same pillar (Fig. 7), as it is also in the medallions at Bh?rh?t. The arch itself is filled in with the criss-crossing seen at Bh?rh?t. This is characteristic of a phase prior to that on the Sanch? foranas where the chaitya arches are represented with the tie-rod clearly distinguished. The painstaking care with which the tie-rod is depicted at Sanch? would seem to indicate that it was a very recent innova tion and the fact that one solitary panel depicts a chaitya arch without a tie-rod is further proof of this.3 The invention of the tie-rod removed the need for slanting door-jambs in the wooden construction of the time as the tie-rod now took the weight and supported the roof. This little architectural detail supports the sculptural style in assigning the pillar to a phase prior to that of the Sanch? foranas?prior at any rate, to the gen eral use of the tie-rod in wooden construction in the Amaravati area. It may be noted that the tie-rod is clearly depicted in the chaitya arches on the octagonal blocks that formed the finial of the ?yaka pillars of the main st?pa at Amaravati (Fig. 8). The sculptured pillar reveals the bead and-reel motif framing the panels and we have seen that this motif is present also on the slab in the Madras Museum with shallow incised carv ings on both sides. It may be noted that a similar framing by this motif occurs on an early Bud dhap?da piece (Stern and B?nisti, PI. VII b). The shallow carving of the worshippers on either side is quite similar to the treatment on the sculp tured pillar just considered, and it appears possi ble that the Buddhap?da fragment belongs to Post-Asokan Stage B. It seems likely that the inscriptions on the much damaged pillar, found by Burgess in the vicinity of the east gate, belongs to this stage (Fig. 9). The inscriptions neranjara and gama nam are certainly labels to sculptured scenes. The word gamanam carved above the horse and below the torana refers undoubtedly to the Great De parture, and the groom in front of the horse carries an umbrella in his hand. To the lower right of this episode is another scene of which the identification is not quite certain. Sivara mamurti feels that it answers best to the Tempta tion of the bodhisattva by M?ra and his daughters ?an event that took place in the Neranjara region. The fact that the words neranjara are inscribed above the scene are to Sivaramamurti a definite indication of this episode, and so far this is the most satisfactory identification. Towards the left of the pillar is another scene identified as the archery contest in which Siddh?rtha dis plays his skill and dexterity. The trees in the foreground of the scene are depicted in a manner very similar to that on the sculptured pillar previously considered. The Vais?li panel, for example, reveals a depiction of foliage in this same manner. The attitude of the figures also suggests a similar style in general, and we would have little hesitation in assigning the two pillars to the same stage. The scene of the gamanam may be compared with the Great Departure scene 45 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions Fig. 10. Pradakshina pat ha slab from Kesanapalli. (Depart ment of Archaeology, Government of Andhra Pradesh.) Fig. 11. Slab from Kesanapalli. (Department of Archaeol ogy, Government of Andhra Pradesh.) Fig. 12. Inscription of Stage II on Early Phase pillar. The British Museum, London. ?a loose fragment?from the early Buddhist caves at Pitalkhor?. The carving there is some what more precise, but the horse is depicted in a very similar attitude, and the torana, while it has three architraves, is similar to that in the gama nam scene and in the paintings in chaitya X at Ajant?. Inscribed labels to sculptured scenes appear elsewhere only at Bh?rh?t, and it has generally been agreed that the reason for this is that, by the time of the BodhGay? and Sanch? torana sculp tures, the stories from the various lives of the Buddha were well enough known for the in formative labels to be omitted. The fact that such labels appear at Amaravati on these sculp tures belonging to Post-Asokan Stage B, adds perhaps to the argument based on palaeography and sculptural style, that these two sculptured pillars belong to a date following soon after Bh?rh?t. However, while the argument based on inscribed labels may be considered after all other evidence, it does not appear that one is justified in basing a chronology on it. Douglas Barrett has pointed out that P. R. Srinivasan's reversal of Chanda's sequence?dedicatory inscriptions being placed in the first century B.C., and labels being taken back to the second century B.C.?does not appear valid.4 Palaeographic evidence too does not permit such a reversal of the sequence: we have seen that the inscriptions on the Sammali coping and the Mudukutala pillar, as well as other donatory records such as Chanda 3-20, be long to Stage A of the Post-Asokan period, while the labelling inscriptions are to be assigned to Stage B. It appears too that an inscribed label does not invariably and without exception indi cate the early date of a sculpture. The yaksha Candamukha inscription, which we shall consider later at length, is apparently a label providing us with the name of the figure carved below, but it appears to belong to a date somewhat in ad vance of the Post-Asokan period. Having defined the Post-Asokan period at Amaravati and divided it into stages A and B, we must now consider the probable absolute dates of these records. The only comparison possible is with the records on the st?pas at Bh?rh?t and Sanch?, and with the inscriptions in the caves of the western Deccan. The records of Post-Asokan Stage A reveal a style distinctly earlier than in scriptions on the foranas of st?pa I at Sanch?, and somewhat earlier than the record on the Bh?rh?t torana (Fig. 13). Later characteristics of the Bh?rh?t record include a rounded ga and a defi 46 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions nite tendency towards the equalization of the verticals of pa. Stage A records appear to be slightly earlier than those at N?n?gh?t, where too a tendency is to be seen towards the equalization of the verticals of pa. The contents of the Bh?rh?t torana inscription give us a definite in dication of the absolute date of the record. The inscription informs us that the torana and some further stone-work were carved during the reign of the Sungas, and we would place the record towards the end of Sunga rule?around 80 B.C. It may not be placed at a much earlier date since it reveals palaeographic features somewhat in advance of the Besnagar Heliodorus record that is securely dated to c. 120-100 B.C.5 The Stage A records from Amaravati display features earlier than the Bh?rh?t torana record, and akin to the inscriptions on the railings of st?pa I and st?pa II at Sanch? and with the railing inscriptions at Bh?rh?t. On this basis we would suggest a date of approximately 90-60 B.C. to cover these rec ords. The script of Post-Asokan Stage B is similar in many ways to the inscriptions at N?n? gh?t. Both reveal a tendency towards the equal ization of the verticals of pa. At N?n?gh?t the beginnings of the tendency towards such equal ization may be seen in the case of sa, ha and la also. In this Amaravati group it may be seen in the case of ha and la. It would seem that on this basis we must place Stage B as belonging some where within the period of c. 60-25 B.C. These inscriptions of the Post-Asokan period and the sculpture on associated fragments indi cate the existence of a st?pa, of some nature, on or beside the site of the main st?pa at Amaravati, but at a considerably earlier date. This Post Asokan st?pa seems to have been surrounded by a largely unsculptured railing and perhaps a small harmik?, and to these must have belonged the various inscribed pillar fragments, coping stones and cross-bars. The two sculptured pillars seem to indicate that columns with scenes carved on them were perhaps set up at the entrance gate way. The position in this scheme of pieces, such as the slab with shallow carvings on both sides, of the flat rectangular pillars with Stage B in scriptions on them, and of the slender octagonal columns apparently topped with addorsed animal capitals, remains uncertain. In this context we may note that while large numbers of sculptured slabs from Amaravati have been destroyed, sev eral more may still be waiting to be uncovered. This is particularly to be borne in mind in view of the fact that over half the fragments enabling us to define a Post-Asokan period have been un covered as recently as during the last ten to fifteen years. The Post-Asokan inscriptions con tain references to a general, a royal scribe and a princess, and indicate that this small early st?pa arose under the patronage of some unknown local dynasty. According to our chronology, on the evidence of both sculpture and palaeography, this st?pa was commenced around 80 B.C., soon after the erection of the railings of the Bh?rh?t st?pa, and the railings of s t? pas I and II at Sanch?, and fairly soon after the cutting of some of the earliest caves in western India, such as the Kon divte chaitya and the Bh?j? chaitya. The evidence from various other sites in the lower Krishna basin indicates that there was a considerable amount of activity in the area dur ing the Post-Asokan period, and that the con struction of the early Amaravati st?pa was by no means an isolated phenomenon. At the nearby site of Guntupalli a small rock-cut chaitya and a few vih?ras appear to have been cut prior to 80 B.C., and were presumably in occupation when the Post-Asokan Amaravati st?pa was construct ed. Few scholars would dispute the analysis that the Guntupalli chaitya displays features reminis cent of the Bar?bar chaityas and distinctively prior to the typical early western chaitya such as Bh?j?, and we would assign the cave to a date prior to 100 B.C. West of Amaravati, at the village of Kesana palli?which together with Amaravati and Jag gayyapeta forms the three points of a roughly equilateral triangle?Waheed Khan recently un covered evidence of activity at an early date.6 A small st?pa, roughly twelve feet in diameter, was surrounded by a pradakshin? patha, and situated on a raised brick platform, and beside this were the remains of at least one rectangular structure. The pradakshin? patha was paved with limestone slabs, several of which have donatory inscriptions engraved on them, and shallow carvings of pendant lotuses and the conventional full lotus (Fig. 10). Waheed Khan suggests that an un usual sculptured slab with a semicircular end depicts a pool with lotuses, judging from the presence of the fishes carved at the right (Fig. 11). The inscriptions, some fifteen in number, are all donatory and belong to Post-Asokan Stage B. Features indicating this include the ka with horizontal and vertical of equal size, the pa in which there is a tendency towards the equaliza tion of the verticals, and a somewhat rounded 47 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions form of ta. The present excavations have not yielded any remnants of a railing that might have surrounded the st?pa. Later activity at the site is indicated by an inscription reporting the dedi cation of an ?yaka pillar in the reign of Vasisht? putra Sr? Chamtam?la, the founder of the Ikshv?ku dynasty. The Post-Asokan st?pa at Kesanapalli was apparently enlarged in the early part of the third century A.D. South of Amaravati at the site of Bhattiprolu also there was fairly extensive activity at an early date, with the construction of a st?pa. Two of the three caskets found at the site are said to be for the relics of the Buddha himself, and this may be regarded as some indication of the antiquity of the st?pa. The inscriptions mention a Ku beraka raja under whose auspices the st?pa and the relic caskets seem to have been prepared. The inscriptions on the caskets display certain fea tures that appear to be a local peculiarity for which we have no explanation. The forms of the letters, considered together with the inscriptions on the crystal, suggest contemporaneity with the records of Post-Asokan Stage A at Amaravati. The sculptured pieces recovered from Bhatti prolu include two fragments of drum slabs similar in style to those from the st?pa at Jaggayyapeta. On palaeographic evidence, the Jaggayyapeta st?pa too appears to have been constructed during the Post-Asokan period. Chanda inscription 2 engraved on a sculptured fragment from Jaggayyapeta, indicates that the st?pa there was completed by the end of the first century B.C. At the site of Amaravati, the end of the Post Asokan period seems to have been followed by a general lull in activity over a span of fifty years or so, after which construction began again when work on the main st?pa was commenced. On the evidence of the neranjara pillar it would seem that the existing Post-Asokan st?pa was enlarged and its railing dismantled and discarded, some pieces being re-used. Bernet Kempers describes the process thus. "Part of the carved work was used for the new one, not only for the decoration, but also for building materials. Reverence for the past did not prevail. Thus a square block (or part of it), which was decorated on two sides at least with series of superposed bas-reliefs (pos sibly it once formed a corner of the entrance) was transformed into an octagonal pillar. The four corners of the block were cut off, only the top and some parts of the base were preserved. The shaft of the pillar was smoothed and is con siderably smaller than the base. . . The latter was only crudely worked?it was meant to be underground?the corners were cut off without care, the planes were not smoothed and partly preserved their carving, as far as it had not been broken during the work. Thus the two frag ments of the decoration of the old block were preserved by chance. Originally they formed part of the greater panels of the block."7 The pillar was re-carved perhaps to take a seated lion on top, in which form it may now be seen in the Madras Museum. It would seem to us that cir cumstances not only permit, but actually call for a gap of some decades between the completion of the Post-Asokan st?pa and the re-use of its ma terials for the construction of the main st?pa. Several palimpsest pieces and split pillars indicate that there was no gap in the re-use of pieces on the main st?pa itself, but in this instance it in volved the complete dismantling of one st?pa for use on another. It would seem to us that there is a fundamental difference between the two. The inscriptions on certain sculptured pillars, cross-bars, coping pieces and drum slabs that dec orated the main st?pa at Amaravati, will be our concern now. We shall examine records inscribed on pieces belonging primarily to the Early Phase or First Period. The evidence afforded by these inscriptions has never been made the subject of a study, and a palaeographic analysis indicates that these records may be subdivided into Stage I (Pre-Nahap?na), Stage II (similar to records of Nahap?na and Gautam?putra S?takarni), and Stage III (those displaying features characteristic of the inscriptions of Pulum?vi). We shall see that in certain instances the palaeographic de velopment of the records does not coincide exact ly with the sculptural development. In pointing this out we do not intend to imply that in such instances the order of the sculptures is to be reversed. Rather, we shall attempt to find an explanation for the slight inconsistency or dis crepancy between palaeographic and sculptural evidence. STAGE I Inscriptions of Stage I, or of Pre-Nahap?na style, display in general a tendency towards an angularity of the letters. There is a lengthening of the verticals of ka, ra and a, all of which main tain a straight lower end, and an equalization of the verticals of sa. We have an early form of da, 48 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions s V < s: ?< z: <5 M? 2 cet -j -* CD <Txs> C 3. T7 C -D < 3*V (A CM 3 4^ G^?j J |D ?C ?*~ W-3 50 ~? *> -9 ?--9 r* < SE uu V? <c z: UJ > I o il i i I! S? h ?J4 PC3 T3 uu po: o rP* H 3 x 7C 4 u 4>ig_ te* S?* , "0 ' Q H*t ?JC v-' fC-V -o ^ 3 8 o?*-? ^ * ^HP? ?T?, ^^l-<^> r^ -< Aoh j? 1 ?, ""-o j? "^ Oo ?? -p -o .-5C?. 5C 3C a c- < -D-6?J ? ^ c * -) - *-> -^ ?Jfc PC c? a < 4- c- < 3 3 ^ J> Si .S? Ci J5 < o U) I \ o a. I? X*. ol? 1 MJ ?? > - c n>3 ^ S!' __5_ vu H -^ X .-) p 1! ^ J 4 va - *? -J +1 5 H -a H: >a -) ? -? -o-i. j> ? < X -J + ?- k* ^ vo ^ v> ^ o ?l o -4 i >c ? S* - ~> *? r< 4^ < o < ?^ X 3 4 C *^?euj ? ? 3 ?- i ! \ ?, I ^ ?6^ ~H4-i l'?. * i * i i * l i 4 5 49 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions the form of ta is h-shaped, and there is a ten dency towards the notched variety of bha. The inscriptions are in general similar to those in the Karle chaitya recording the donation of various portions of the cave (Fig. 13). 1-1. The yaksha Candamukha inscription is the first record to be considered in this group (Sivaramamurti, PL LXV 9; Chanda 36). The inscription is very brief and unfortunately does not include key letters such as ga, pa and ta: it is, however, very precisely engraved as compared with inscriptions of the Post-Asokan period. One feature that may be regarded as an indication of its later date is the definite lengthening of the vertical of ha. The forms of da, ma, sa and na offer few clues. The form of va may again be taken as an indica tion of the later date of the record. The letter has a flattish base, but this in itself is no indica tion of an advanced age as the va with such a base occurs in inscriptions of various ages. Here, however, the va has no vertical at all, and this feature?an unusual one?usually occurs at a late date. The form of kha is a very strange one and may in fact be described as unique: this is apparent from the fact that so reliable and ac curate an authority as Chanda should read it as ga. However, the letter occurs twice in the in scription and Sivaramamurti's reading appears in little doubt. The yaksha piece has been classified as an ?yaka slab with pilaster (Barrett, B.M.Q. p. 43 ). Presumably the yaksha comprises the shaft of the pilaster which then terminated without the usual inverted ghata and addorsed animal capital. The piece is an unusual one, the only similar handling of the pilaster figure being on the Chakravartin slab from Jaggayyapeta where the remnants of more than one chaitya arch are visible above the pilaster figure. A noteworthy feature is the large heavy square earrings seen on the yaksha slab: these are strikingly similar to those on a yakshi fragment recently recovered at Amaravati (An cient India 20/21, PL XLIV A), and to those worn by the Queen on the Jaggayyapeta Chak ravartin slab. It is similar also to those seen on the British Museum palimpsest drum slab (Bar rett, PL V). The later age of the yaksha slab is indicated also by the carving of the chaitya arch which is depicted with a distinct inward curve at the lower end. It is highly advanced when com pared with those depicted on pieces belonging to Post-Asokan Stage B, and is definitely closer in style to the arches carved on the block forming the finial of an ?yaka pillar of the main st?pa (Fig. 8). This yaksha ?yaka-platioicm slab, to gether with the British Museum ?yaka slab (Bar rett, PL V), the slab in Madras depicting a vriksha-chaitya (Sivaramamurti, PL XV, 1), and the slab representing the worship of a domed chaitya (A.R.A.S.I. 1908/09, PL XXIX d) were among the very first pieces to be carved at the site when work began on the main st?pa. The ?yaka-platiorm slabs certainly comprised the first sculptural work on the st?pa, prior probably to work on the railing. Other records of Stage I include: 1-2. The donatory inscription on the "Sri" coping (Sivaramamurti, PL XV, 3). 1-3. A fragmentary record on a similar coping with the dwarf-and-roll theme (Burgess, PL XXXI, 3). 1-4. The record on the garment of a life-size figure of a worshipper (Chanda 39; Siva ramamurti, PL XVIII, 2, 3). This record at present has so few letters intact that it is difficult to be very specific on its palaeographic position and it could, in fact, have belonged to an earlier stage. 1-5. A much blurred inscription on a slab de picting the facade of a building (Stern and B?nisti, PL X a). We have been unable to locate any translation or photograph of an estampage of this record. 1-6. Gift of a coping stone (Chanda 29). 1-7. Donation of a cross-bar (Chanda 32). 1-8. Donation of three cross-bars, engraved on the corner of a disc with a big lotus (Chanda 33). 1-9. A fragmentary donation (Chanda 35). I-10. Donation mentioning the navakarmika, engraved on a fragment of a rail pillar (Chanda 40). 1-11. Gift of a coping stone (Chanda 42). 1-12. Gift of a coping stone and a Buddhap?da (Chanda 44). 1-13. Gift of a cross-bar (Sivaramamurti No. 85,P1.LXV, 16). 1-14. Gift of a coping slab, engraved on a split pillar (Sivaramamurti, No. 94 and PL LXV, 11). 1-15. A fragmentary donation engraved on a slab showing the lower portion of the feet of a man and three women (Burgess, PL LVIII, 33). 50 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions I-16. Donation of a pillar engraved on the lower portion of an octagonal pillar, on the re verse of which is a list of Pallava kings (Burgess, PL LXI, 51). 1-17. Fragmentary donatory record engraved on a cross-bar with a lotus medallion (Bur gess, PL L VI, 14). 1-18. A fragmentary inscription on a broken pillar at Amaravati (An. Rep. Ind. Ep. 1959/60, No. 43). STAGE II Belonging to a slightly more advanced stage in which the script shows affinities with the in scriptions of Nahap?na and Gautam?putra S?ta karni, are a group of records on various sculp tured pieces. These inscriptions display a light but distinct curve of the lower ends of the verti cals of a, ka and ra, and also of medial u. The form of da is occasionally of the early variety, but more often of a new form open to the right. II-1. The donation of a chaitya pillar, engraved on an octagonal sculptured column (Fig. 12. Also Barrett, Fig. IX, b, c, d). II-2. The fragmentary inscription on the re verse of the Madras Museum palimpsest (Douglas Barrett, "Two Unpublished Sculptures from Amaravati", B.M.Q. XX, 1956, PL XXIV, a, b). II-3. The inscription on a slab representing the Great Departure (Barrett, Catalogue No. 73, Middle Phase). The inscription is badly damaged but the letters may be discerned reasonably clearly in Fergusson's reproduction (Fergusson, PL XCVI, 3). II-4. The record on a pillar carved with st?pas and standing Buddha figures (Stern and B?nisti, PL LXVII a). This pillar indicates that plain unsculptured columns were oc casionally set up with donatory inscrip tions on them, and that sculptures were added at a later date: the carving in this instance belongs undoubtedly to the Late Phase. II-5. The donation of a pillar, inscribed on a fragmentary slab built into a temple (Bur gess, PL LIX, 39). II-6 & 7. The fragmentary inscriptions on two drum slabs classified as Middle Phase pieces (Barrett, PL XVI, XVII), appear to be long to the transition to Stage III. Once again Fergusson's reproductions are valua ble for the inscriptions (Fergusson, PL XCV, 3,4). II-8. The inscription on the top of the rail pillar seen in Barrett, PL XXIV also has an in scription belonging to the transition to Stage III (Fergusson, PL LXI, 1). II-9. The inscription at the top of the rail pillar on Barrett, PL XXI b belongs undoubtedly to the transition to Stage III. STAGE III The inscriptions of Stage III, of the time of Pulum?vi, reveal definite advances in script. The lower ends of the verticals of ka, ra and a are invariably curved, and in the case of jha and ?a, optionally so. Elaborate flourishes of medial i and u may be seen. The left arm of pa, ha and ba usually curve inward, and the form of da is always the latter one, open to the right. A new form of ta now appears: the letter is formed in a single stroke and results in a hooked variety. The earlier form may be seen to persist, however, in several inscriptions of this stage. III-l & 2. Prominent among Stage III pieces are two donatory records on Early Phase rail pillars (Barrett, PL XX b; Barrett, Cata logue No. 1 ). The inscriptions display all the advanced features characteristic of the records of Pulum?vi, with the elaborate curves of the lower ends of the verticals and the flourishes of the medi?is. These rail pillars with one face plain and the other carved with one full lotus and two half-lotuses, belong undoubtedly to the earliest stage in the decoration of the rail. The inscription on Barrett, PL XX b is intact and refers to the gift of two p? dukas by Sivaka and various members of his family, while the record on Barrett, Catalogue No. 1 is fragmentary. Ill-3. Belonging to the Early Phase is a slab that formed part of the final projection of a gateway, with mortise holes in the top to hold the lion figure, and in the base to fix it to a pillar (Barrett, PL XIV b). The inscription on the slab refers specifically to the donation of a lion-seat (sihath?na). The script displays, even more than the two rail pillars considered above, the elab orate flourishes characteristic of the rec ords of the time of Pulum?vi. It is apparent that there is some discrepancy here between the palaeographic and sculptural evidence?a discrepancy which must be recog nized and discussed. Any scholar comparing the n This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions pieces discussed as III-3 and II-9 will agree that the inscription on II-9 undoubtedly displays (whatever the reason for this may be) an earlier form of script. Yet sculpturally the situation is reversed as the carving of II-9 is of a later date than III-3. The main discrepancies between palaeographic and sculptural style will be exam ined below and at the same time we shall con sider the slight anomalies presented by some of the inscriptions. In the course of examining the contents of various records, we have come across a number of instances in which the donations are apparently inscribed on a piece other than the one actually donated. The possible implications of this will be apparent. While this explanation does not provide all the answers, it nevertheless raises interesting possibilities. The inscriptions on two Early Phase rail pillars belong to the time of Pulum?vi (III-l and 2), while the records on two coping stones that must have crowned pillars of this type belong to the Pre-Nahap?na stage (1-2 and 3). It would ap pear possible that the rail pillars were also set up in the Pre-Nahap?na stage with the inscriptions being added in later days. The record on one of the rail pillars certainly refers to an unconnected gift of footprints of the Buddha. The record on the other rail pillar is fragmentary. This explana tion, while possibly valid in this particular in stance, may not be extended to all such problem records. The Pulum?vi style inscription on the slab for a lion mount (an Early Phase piece) reports the gift of that very slab (III-3). It in dicates either that the piece itself must be as signed to the Middle Phase, or that the Early Phase which commenced at an earlier date, per sisted into the time of Pulum?vi. The fragmentary inscription on the top of the magnificent Middle Phase rail pillar (H-9) refers to the gift of two p?dukas by the mother of ?nada, The portion of the inscription referring to the actual donation is quite intact. While the record belongs palaeographically to the Pre Pulum?vi stage, the pillar itself is one of the finest examples of the Middle Phase rail pillars, and must have belonged to the time of Pulum?vi. It appears probable that an unsculptured pillar was set up at an earlier date and the donation of the p?dukas (which one assumes were located nearby) was engraved on it. This inscription was left intact when the pillar was carved at a later stage. That this in fact did happen is indicated by the piece discussed earlier as II-4. This pillar has a record in the characters typical of the in scriptions of Nahap?na and Gautam?putra S?ta karni, while the carving depicts standing Buddha figures and belongs undoubtedly to the Late Phase. It is apparent in this instance that an un sculptured pillar was donated at an early date and the gift inscribed on it. When it was later carved, the earlier record was allowed to remain. This explanation seems most probable for the magnificent Middle Phase rail pillar under con sideration. There is the possibility that the earlier script persisted into the Middle Phase, but it seems unlikely that once the pillar with its splen did carving was completed (presumably through the munificence of an un-named donor) an un connected record would then have been inscribed on it. The third possibility?an unlikely one? is that the rail pillar actually belongs to an earlier period. One other anomalous inscription to which we may draw attention, although it does not pose any problems of sculptural or palaeographic dis crepancy, is the donatory inscription engraved on a Late Phase drum frieze (Barrett, PL X & XI). The record mentions various gifts to the "Great Chaitya" at Amaravati including that of two chaitya slabs, three p?dukas, a coping, and a slab with a flower vase. It also mentions the erection of some structure at the st?pa at R?jagiri (Luders, No. 1225). It is interesting that apart from the various gifts at Amaravati (engraved on an unconnected slab) donations to other stupas are even recorded. Other examples of such anomalous inscriptions exist.8 Two drum slabs of the Middle Phase contain inscriptions that illustrate the transition to the days of Pulum?vi (II-6 and 7). Either these pieces belong sculpturally also to the transition from the Early to the Middle Phase, or the earlier style of script persisted side by side with that of Pulum?vi. Another fragment, apparently from a drum slab, displays an even earlier style of script (1-15). We must consider also the questions raised by a comparison of the slab II-2 with Chanda in scription 2 that belongs to Post-Asokan Stage B. The latter inscription, engraved on a fragment from Jaggayyapeta, indicates that the decoration of the drum of the Jaggayyapeta st?pa took place in the Post-Asokan period. The inscription on the reverse of the Madras Museum palimpsest, which is a quadrant slab (II-2), belongs to the stage illustrating the transition to the days of Pulum?vi, and indicates that the similar decora tion of the drum at Amaravati belongs to a later 52 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions date. It is certainly true that apart from the treatment of the shaft of the pilasters "no stu dent . . . would be able to distinguish stylisti cally between the pilaster forms on either type of slab [?yaka-platiorm or quadrant] or between those of Amaravati and Jaggayyapeta."9 On this basis, it might be expected that Post-Asokan Stage B and Stage II of the main st?pa belong fairly closely together. However, the appearance of the male and female figures on the shafts of all pilasters at Jaggayyapeta, and on the shafts of the ?yaka-plztiorm slabs only at Amaravati (though not necessarily on all such) may defi nitely be considered as an earlier feature, which was followed on the quadrant slabs of the Amara vati st?pa, by shafts decorated in imitation of a rail pillar with one full and two half-lotuses. Construction of the main st?pa at Amaravati was commenced some time after work at Jaggayya peta was completed, and the ?yaka-phtforms were carved first with the shafts of the pilasters following the earlier style of Jaggayyapeta. The quadrant slabs were cut later, probably contem porary with the carving of the rail, and the sculpting of human figures on the pilaster shafts was abandoned. The inscription on the Amara vati quadrant slab (II-2) is then understandably later than any on the Jaggayyapeta st?pa. We may also note that there is undoubtedly a certain difference in the treatment of the sculptured scenes on the ?yaka slabs at Jaggayyapeta and Amaravati. The scenes at Jaggayyapeta?the Chakravartin or the Punyas?la?reveal a com pletely flat, entirely incised treatment. The Amaravati ?yaka slabs by contrast reveal a deeper cutting, and the British Museum palimpsest for example (Barrett, PL V) could no longer be described as incised, although the cutting is shal low and flat as compared to later sculptures. We must now consider the absolute dating of the records on the main st?pa at Amaravati. We have termed the Stage I records Pre-Nahap?na, by which we mean prior to the first known rec ord of Nahap?na of the year 41. We do not believe that his inscriptions refer to the Saka era of A.D. 78, and largely following the argument of the editorial article "Date of the Karle Chait ya", in Lalit Kala 3/4, 1956/57, we would iden tify the year 41 with the year A.D. 95. On this basis we would place the Stage I records roughly between A.D. 50-95. Stage II then belongs be tween A.D. 95-110, and Stage III to A.D. 110 138. Those who assign Nahap?na to the Saka era may adjust the dates accordingly. We have seen earlier that pieces belonging sculpturally to the Early Phase have engraved on them inscriptions of all three stages: these pieces could then be placed anywhere between A.D. 50 and A.D. 138. This period may, however, be lessened to around sixty years by assigning the Stage III inscriptions to the first years of Pulu m?vi?to around A.D. 110. The first appearance of his new style of script may justifiably be placed towards the start of his reign. That the sculptural decoration of the monument in Early Phase style extended over such a period is cer tainly feasible. It would seem to us that the Early Phase had a slow and somewhat uncertain start, possibly under the influence of one of the branch lines of the S?tav?hana dynasty, while their central territories were being lost to the Kshahar?tas. With the re-establishment of S?ta v?hana power and the extension of their domains, work on the monument was subject to accelera tion. While a few pieces of Early Phase style may still have been carved in the time of Pulum?vi, stylistic development was rapid and there was a swift transition into the style of the Middle Phase. The one or two pieces of Middle Phase style, which bear Stage II inscriptions, indicate that Early and Middle Phase styles undoubtedly co-existed for some time, just as Stage II and Stage III styles of script must have been in vogue side by side for a certain length of time. Among the artisans involved in work at Amaravati there would have been the older master craftsmen and the younger pupils, and there must undoubtedly and understandably have been stylistic distinc tions between their work. In sculptural style, as in palaeographic style, the transition from one phase to another could not have been so abrupt that, at any certain time, two stages in the de velopment of a style may not be found together. / must acknowledge here the help and advice received from Mr. Douglas Barrett of the British Museum, who was kind enough to read through the typescript of this article and offer me sug gestions for the improvements. Ultimate responsibility for the views on the dating must, however, rest with me.?V. Dehejia. 53 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions NOTES 1. D. C. Sircar, "Fragmentary Pillar Inscription from Amaravati", Epigraphia ?ndica, XXXV, 1%2, pp. 40-43. 2. Barrett, British Museum Quarterly, XXXII, 1967, p. 45. 3. The chaitya arch is depicted without the tie-rod on the top most panel of the left pillar on the south torana. 4. Barrett, ibid., p. 46 f. 5. The inscription mentions the erection of a garuda-dhvaja by the ambassador from king Antialkidas, who is known to have ruled between 120-100 B.C. 6. Mohd. Abdul Waheed Khan, A Buddhist Stupa at Kesanapalli (District Guntur, Andhra Pradesh), Hyderabad, 1969. 7. A.J.Bernet Kempers, "Note on an ancient sculpture from Amaravati", Acta Orientalia, X, 1932, p. 365. 8. Consider, for example, the inscription on a drum pilaster belonging to the Late Phase (Barrett, PI. Villa) which reports the gift of a p?duka slab (Luders 1217). 9. Barrett, ibid., p. 44. ABBREVIATIONS Ancient India 20/21?A. Ghosh and H. Sarkar, "Beginnings of Sculptural Art in south-east India: a stele from Amaravati", Ancient India 20/21, (1964 & 65), pp. 168-177. A.R.A.S.I.?Annual Report of the Archaeological Survey of India. An. Rep. Ind. Ep.?Annual Report on Indian Epigraphy. Barrett?Douglas Barrett, Sculptures from Amaravati in the British Museum, (London, 1954). Barrett, B.M.Q.?Douglas Barrett, "The Early Phase at Amaravati", British Museum Quarterly XXXII, 1967, pp. 35-48. Burgess?James Burgess, The Buddhist Stupas of Amaravati and Jaggayyapeta, (London, 1887). Chanda?R. P. Chanda, "Some Unpublished Amaravati Inscrip tions", Epigraphia Indica, XV, 1919/20, pp. 258-275. Fergusson?James Fergusson, Tree and Serpent Worship, (London, 1873). Luders?H. Luders, "A List of Brahmi Inscriptions", Appendix to Epigraphia Indica, X, 1909/10. Sivaramamurti?C. Sivaramamurti, Amaravati Sculptures in the Madras Government Museum, (Madras, 1942). Srinivasan, Lalit Kal??P. R. Srinivasan, "Recently discovered in scriptions from Amaravati and their significance", Lalit Kola 10, 1961, p. 59 f. Stern and B?nisti?Philippe Stern and Mireille B?nisti, ?volution du style Indien d'Amaravati, (Paris, 1961). 54 This content downloaded on Tue, 5 Mar 2013 16:26:22 PM All use subject to JSTOR Terms and Conditions