p. 186 NAGARJUNAKONDA, or Nagarjuna's hill, is the name of a big rocky hat-topped hill on the right bank of the Krishna river in the Palnad taluk of the Guntur district of the Madras Presidency, and 15 miles west by north of Macherla railway station, the terminus of the new line from Guntur opened in 1931. The hill stands in a valley completely shut in by at ring of hills, an offshoot of the Nallamalais (Black Hills) of the adjoining Kurnool district, on three sides, with the Krishna, river on the fourth or north-western side, where it forms the boundary between this part of the Madras Presidency and the Nizam's Dominions. The annexed site plan (Plate I) shows the geographical features of the area and the positions of the monuments discovered. Nagarjunakonda is about 60 miles distant from Amaravati as the crow flies, but considerably further by river. It is a wild and desolate spot, and being shut in by the surrounding rocky hills is usually very hot during most months of the year. There is a rough cart track from Macherla to Nagulavaram, a distance of 10 miles, but the remaining 5 miles over the hills and through the valley to Nagarjunakonda has to be performed on foot, as no cart traffic is possible. The hill was once fortified, and remains of brick and stone fortifications still remain all along the rugged cliffs surrounding the plateau on its summit, showing that it was once used as a citadel; but no ruined buildings of interest were discovered on the hill. At the eastern foot of the hill and scattered throughout the valley are a number of ruined stupas of all sizes, from little structures 8 feet in diameter to large ones like the Great Stupa, 106 feet in diameter. There are also many ruined monasteries and apsidal Buddhist temples, showing that, at one time, there existed here a large and flourishing Buddhist settlement, far larger in fact than the one at Amaravati lower down the river. A number of important inscriptions in Prakrit and in Brahmi characters of about the second century A.D. were discovered in connection with the Great Stupa and two apsidal. temples. Professor Vogel of Leiden University has published an account of these old records in the Epigraphia Indica, volume XX, 1931. Besides a number of inscriptions end ruined buildings, many lead coins of the Andhra period, gold and silver reliquaries, pottery, statues and over four hundred magnifi cent bas-relief sculptures similar to those from Amaravati, were recovered curing the excavations which I conducted at Nagarjunakonda during the cold seasons of 1928 to 1931, when T completed the explorations, A brief account of these discoveries appears in the Annual Reports of the Archaological Survey of India for those years, but a fully illustrated account of the remarkable discoveries made would fill a large volume, and has yet to be written. The historical information furnished by the inscriptions is somewhat meagre, and the careless manner in which some of them were engraved adds to the difficulty of interpreting the precise meaning of certain words and sentences. The records belong to the Southern Ikhaku dynasty, who were ruling in this part of India between the second and third centuries A.D. It is clear from these inscriptions that they were kings of considerable importance, but also wih they formed matrimonial alliances not only with the rulers of Vanavasa (North Kanara), Ikhakus revealed by the inscriptions, is that while the rulers were followers of Brahmanism and performed Vedic sacrifices, their consorts were devotees of the Buddha and erected buildings for the Buddhists settled at Nagarjunakonda and made pious donations to the stupas. Most of these buildings owed their existence to the piety of certain queens and princesses belonging to the royal house of Ikhaku, the principal founder being a princess named Chamtisiri, who is praised for her munificence in many of the inscriptions belonging to the Great Stupa, or Mahachetiya, as it is called in the piller inscriptions belonging to it, and which was founded ____________________ Note.--The copyright of the photographs reproduced to illustrate this article gioal Survey of india. p. 187 or perhaps rebuilt, when the pillars were added, by the lady in question in the sixth year of the reign of king Siri-Virtapurisadata between the second and third centuries A.D. The same royal lady built a monastery and an apsidal temple close to the eastern gate of the Great Stupa, the ruins of which remain. Another important inscription was found engraved on the stone floor of an apsidal temple situated on a rocky hill about two furlongs to the east of the Great Stupa, and known locally as Naharallabodu. This temple and a monastery standing alongside of if were built by a lady named Bodhisiri and dedicated to the fraternities of Ceylonese monks settled at Nagarjunakonda. The inscription relates that these Ceylonese Buddhists had converted the people of Kashmir, Gandhara, China, Ceylon, Bengal, Kanara, and other places in India. The latter part of the inscription mentions other pious works by Bodhisiri, including a pillared hall or mandapa at Kantakasela, which, as Dr. Vogel points out in his account of these inscriptions, must be identical with "the emporium Kantikossula," mentioned by Ptolemy as being situated "after the mouths of the Maisolos (Krishna)." The Periplus speaks of "the region of Masalia," stretching a long way along the coast," and adds, "a great quantity of muslins is made here." The ancient name by which the Krishna delta was known to the Greeks is preserved in that of the seaport of Masulipatam. In the same inscription (F of Dr. Vogel's list), the name of the ancient city that once existed in the Nagarjunakonda valley is given as Vijayapuri, and the hill now known as Naharallabodu, on which Bodhisiri erected the temple and monastery for the Ceylonese monks, is called the Lesser Dhammagiri situated on Sriparvata. The hill in question is an offshoot of the surrounding Nallamalais of the adjoining Kurnool district. These hills extend in a south-westerly direction all along the river into the Kurnool district, where, on the top of a wooded hill some 50 miles south-west of Nagarjunakonda and facing the river, stands the famous Srisailam temple sacred to Siva and a great place of pilgrimage in the spring, when a big annual festival is held there. It thus seems from this inscription that in early times the Nallamalais were known as Sriparvata. This is an interesting point, because there is an ancient tradition preserved in Tibet that the famous Buddhist divine Nagarjuna ended his days in a monastery on Sriparvata in Southern India. If this monastery is the same as the ruined one on the Lesser Dhammagiri, if would follow that the association of Nagarjuna with this locality has been preserved up to the present day in the name Nagarjunakonda (Nagarjuna's Hill). The fact that a monastery and a temple were built specially for the benefit of Ceylonese, monks shows that very cordial relations must have existed between the Andhra, Buddhists and their co-religionists in Ceylon at that period. The existence of such relations can be readily accounted for by the sea-borne trade which was carried on between the ports of Ceylon and the great emporium Kantakasela of the Krishna delta. It was no doubt this trade which was mainly responsible for the flourishing state of Buddhism in this part of Southern India, which enabled the Buddhist merchants and their royal masters to raise monuments of such magnificence as those at Nagarjunakonda and Amaravati. As Dr. Vogel mentions, the decline of Buddhism in the lower Krishna valley may have had other causes besides the general wane of that religion all over India, there may have been economic factors at work, such as the decline of the sea-borne trade with the West, which had caused vast quantities of Roman gold to pour into Southern India. There was also the conquest of the south by the Gupta Emperor Samudra Gupta and the rise of powerful dynasties devoted to Brahmanism, like the Pallava dynasty in the South and the Chalukya in the West. The ruined buildings discovered, represent the remains of stupas, monasteries, apsidal temples and a palace. They were all built of large bricks measuring 20' x 10' x 3", the same dimensions as the bricks recently found at Bulandibagh near Patna in Bihar, the ancient site of Pataliputra. It is strange that at two sites so far distant both should yield large bricks of the same dimensions. The pillars, floors, statues and important sculptures were executed p. 188 in white or grey limestone resembling marble. No other stone was used, and it was brought to the site by means of the river and landed at a stone-built wharf that still remains (see Plate I, 12). The wharf is about 250 feet in length, 50 feet wide and 6 feet in height along the river front and at both ends. Three rows of broken stone pillars extending from end to end show that it was originally provided with a wooden roof, probably thatched. It seems to have served as a kind of Customs House, with a row of shops or godowns on either side. Here, the Krishna is more than half a mile wide, with numerous sandbanks and huge rocks in its bed, but during the rains it is a very large river and navigable for country craft right down to the sea. On plan and in construction, the Andhra stupas differ from those found in the North. They are built in the form of a wheel with hub, spokes and tire all complete and executed in brickwork (see plan of stupa On Plate III). The open spaces between the radiating walls were filled up with earth, and the dome or brick casing built over the structure. As no traces of structural stone tees have been discovered in Southern India, we may presume that they were built of brick and plaster and decorated with the rail ornament in the latter material. The stupas were covered with chunam, or fine shell-lime plaster, from top to bottom, anti the moulding and other ornamentation was usually executed in stucco or plaster. The dome rested on a circular platform or drum from 2 to 5 feet in height according to the size of the monument. On top of the drum was a narrow path encircling the foot of the dome, and on each of the four sides, facing the cardinal points, was a rectangular platform resembling an altar and the same height as the drum. In the inscriptions these platforms are described as ayaka-platforms, because they usually supported a group of five stone pillars, called ayaka-khambhas (ayaka-pillars). The precise meaning of the word ayaka is not known, but it is used much as we use the word 'altar.' From the has-relief representations of stupas recovered from the Nagarjunakonda and Amaravati stupas the ayaka-platform appears as an altar, on which pious -donors are portrayed depositing their offerings of fruit and flowers. All Andhra stupas had those platforms, but only those belonging to large and important monuments were provided with pillars. As each group consisted of five pillars, the total number of pillars for each stupa so decorated was twenty. The inscriptions show that these pillars represent gifts made to the stupa in honour of the Buddha and to the merit of the pious donors who provided the money for the work; hut no information is given as to the meaning or symbolism of the pillars. The chief scenes portrayed in the sculptures recovered from these Andhra stupas represent the five great 'miracles,' or chief events in the life of the Buddha, namely, the Nativity, Renunciation, Sambodhi, First Sermon, and the Buddha's Death. These five incidents are portrayed over and over again, either as beautifully executed bas-relief scenes, or else as mere conventional symbols, such as a tree, wheel and stupa. In this form they are found engraved on some of the bases of the ayaka-pillars belonging to the Amaravati Stupa now in the Madras Museum; and I discovered at Nagarjunakonda four bases of ayaka-pillars each ornamented with a bas-relief representation of the 'First Sermon.' The presence of these symbols carved on the bases of the pillars seems to indicate that they were set up to commemorate the five great miracles; just as we know Asoka erected pillars to mark the sacred spots where these events are said to have occurred in Nepal and Bihar. As it was impossible for those living in the Krishna district to erect the pillars an the aotual spots in Northern India, they seem to have hit upon the idea of conventionslising the pillars into groups of five for the sake of convenience, so that the events could be commemorated locally, and also, perhaps, with a view to adding to the splendour and importance of the stupas, as in the case of the Amaravati Stupa, where the stone casing to the dome, the ayaka-platforms and pillars, and the stone railing, were all added to the monument in the second or third century A.D. This we know from the inscriptions belonging to that monument. In earlier times the ayaka-pillars were unknown, and they only occur in the Andhra stupas of that period. p. 189 The platforms and pillars vary in size and height according to the dimensions of the stupa to which they belong. The pillars vary from 10 to 30 feet in height, with square bases and octagonal shafts. The tops are round, showing that they could not have supported capitals or any other kind of ornaments. In some of the bas-relief pictures of stupas, the pillars are shown crowned with trisula ornaments, the centre pillar often with a miniature stupa as capital. This is incorrect and purely decorative, as they never supported anything and could not do so as the tops were round, so that any ornament placed there would fall immediately to the ground. In this case the ornaments merely indicate that the pillars were dedicated to the Buddha, and the inscriptions confirm this. In the sculptures two kinds of stupas are depicted--one a plain brick and plaster structure like the stupas of the Asokan age; and the other is similar in all respects, except that the brick surface is faced with richly carved stone slabs embedded in mortar. This stone casing was applied only to the face of the drum, ayaka-platforms and lower portion of the dome. The upper portion of the domes of all Andhra stupas was executed in brick and plaster and decorated with a characteristic garland ornament encircling the dome. This ornament always appears in the has-relief representations of stupas, and is in the form of a broad festoon decorated with big lotus medallions executed in plaster. The stone casing was applied only to the base of the dome, as it is obvious that flat stone slabs could not be fixed to the curved surface of the upper portion of the dome. In order to do this, each stone would have to be specially cut with a convex front and a concave back, and even then it would be very difficult to keep the stones in position, so this part of the stupas was always in plaster. These decorated stupas were faced partly with stone slabs and partly with plaster ornamentation, the two materials being used together, and when the work was completed the stupa wits given a coating of shell-lime plaster from top to bottom, to hide any defects or inequalities in.the work. For this purpose, the white limestone used for this work was specially suitable, as it was of the right colour and takes whitewash or plaster readily, being very absorbent. It was no doubt these considerations and the fact that it is soft and easy to work when freshly quarried, that led to its general use in the Krishna valley. From the remains of slate-stone bas-reliefs and plaster ornament recovered from the ruined stupas of Gandhara, it seems that they were decorated in the same manner as those erected by the Andhras. The inscriptions show that there was considerable intercourse between the Buddhists of Gandhara and their co-religionists in the South, and in all probability the Andhras adopted the custom from the Gandhara builders in the second century A.D., or thereabouts. Gandhara influence is also strongly marked in many of the Andhra bas-reliefs and statues in the round. Traces of Roman influence are also manifest in a few of the sculptures and in two small gold medallions recovered from Nagarjunakonda. This is not surprising, as we know that in the second and third centuries of our era there was considerable sea-borne trade between Rome and this part of Southern India. When complete, the Great Stupa at Nagarjunakonda must have been a perfect example of a plain Andhra stupa (Plate II, fig. 2). It is built of large bricks measuring 20' x 10' x 3", and in the usual form of a wheel (Plate III, fig. 2). It was covered with plaster from top to bottom, the dome being decorated with the usual garland ornament, and the drum with a few simple mouldings executed in plaster. No stone was used in its construction, the ayaka-pillars alone being of that material, end, as at Amaravati, they probably represent a later addition to the stupa. They were gifts, as their inscriptions show, and were erected between the second and third centuries A.D. The diameter of the stupa including the drum is 106 feet. The drum is raised 5 feet above the ground level, and the total height of the monument, excluding the tee, must have been about 70 to 80 feet. On top of the drum is a narrow path, 7 feet wide, extending all round the base of the dome. No traces of steps p. 190 up to this path were found, but it is possible that they may have existed. No steps are depiotep redepio ed in the has-relief representations of stupas, so perhaps there were none to any of these monuments. The ayaka-platforms are 22 feet in length and 5 feet in width, and the bases of the five stone pillars were securely built into the brickwork. In the stone-faced stupas, the ayaka-platforms were the most highly decorated features of the stupa. Here the Andhra sculptor exhibited his best works of art, partly because these platforms were regarded as very holy structures resembling altars on which votive offerings were placed, and mainly perhaps, because they faced the four open gateways of the stupa, so that they were the first objects seen by anyone entering the sacred precinct around the stupa. The stupa was surrounded by a processional path 13 feet in width, and enclosed by a wooden railing standing on brick foundations, which still remain. The gateways were formed by extending the railing outwards so as to form a screen on each side of the entrance, but there were apparently no transoms spanning the entrance, like those of the Sanchi toranas. No traces of stone rails or toronas were found at Nagarjunakonda, and it is quite clear that none existed there. As a rule, the fails and gates were constructed of carved woodwork, no doubt resting on brick foundations, to protect them from damp and the ravages of white ante. It was only in very special cases that they were ever executed in stone, and then they were merely stone models of carved wooden originals. When first discovered, the Great Stupa at Nagarjunakonda was a large mound of earth and broken brick overgrown with grass and jungle, with two ayaka-pillars standing erect, the remaining eighteen pillars having fallen. As the whole of the dome of the stupa had been demolished, the ayaka-pillars and platforms thrown down and broken by treasure seekers, the chances of finding any relies in the edifice appeared very remote indeed. The first thing was to remove the debris and trace out the plan of the structure and recover the broken pillars. When this work was finished and the excavations completed, the appearance of the Great Stupa may be gathered from Plate II, fig. 2. Fortunately, instead of piecing the relies in the centre of the Great Stupa, they were deposited in one of the outer chambers on the north-western side of the stupa, where they escaped the notice of the treasure seekers who wrecked the monument (Plate III, fig. 1). As the slupa contained 40 chambers, all of which had to be excavated down to the natural ground level, the excavation of this monument was a very laborious task that took a month to complete. At last, when we had given up all hopes of finding anything of interest, one of the coolies noticed a small broken ]pot in the north-western corner of the chamber marked with a, cross on the plan (Plate III, fig. 2). The pot had been crushed when the chamber was filled with earth by the Buddhists, and all that remained is shown in Plate IV, fig. 1. On the surface were a few white crystal beads and a, tiny gold box. After carefully sifting the contents of the pot the following objects were found:-a fragment of bone placed in a small round gold reliquary three-quarters of an inch in diameter. This was placed in a little silver casket, shaped like a miniature stupa, 2 1/2 inches in height, together with a few gold flowers, pearls, garnets and crystals. The three large crystal beads and the round earornament were placed in the pot and not in the casket. The latter unfortunately was very corroded and broken, but a replica was made, which appears in the photograph showing the finds recovered from the tomb (Plate IV, fig. 2). The earthenware pot containing the casket and reliquary was placed originally in the corner of the chamber, which was filled up with earth as soon as the consecration ceremony was over. The brick dome was then built over the remains, and the plastering and decoration of the stupa completed. No traces of ornamental plaster were found in the debris round the monument, except portions of simple mouldings that once decorated the plinth and cornice of the drum. It must have been a perfectly plain structure like those of the Asokan age before the ayaka-pillars were added in the second century A.D. (Plate II, fig. 2). p. 191 In the inscriptions belonging to the Great Stupa, the monument is called the "Mahachetiya of the Lord, the Supreme Buddha," clearly showing that the tomb was consecrated to the Great Teacher and to nobody else. The discovery of the dhalu, or bone relic, proves that the monument was a dhatugarbha, or 'tomb containing a relic,' and that it was not a mere 'dedicatory' stupa. The latter were memorial stupas, which contained no relics, and, like Asoka's pillars, were erected on celebrated sites sacred to the Buddha, such as his birthplace, and so on. It is, therefore, obvious that the Great Stupa did not belong to this class of memorial monument. The inscriptions do not definitely state why the stupa was built; they merely state that the ayaka-pillars were dedicated to the Buddha, and that they were Bet up by the princess Chamtisiri and other royal ladies of the same house. Supposing the stupa to have been already in existence prior to the erection of the pillars, it would have been necessary first to enlarge the drum and build the ayaka-platforms to accommodate the pillars, and then replaster and decorate the stupa from top to bottom to complete the work. In fact, it would have meant rebuilding the whole of the exterior of the monument. Dr. Vogel is of opinion that the inscriptions show that the Mahachetiya was " founded " by Chamtisiri, but it is by no means clear whether she built, rebuilt, or merely contributed to the structure. If she did build the stupa, then it was she who enshrined the relic found in the chamber;but it is impossible to believe that so great an event as this could have occurred without the fact being recorded in at least one of the many inscriptions referring to the stupa. We know that the monument was consecrated to the Buddha, as the inscriptions are quite clear on this point. Therefore, it is reasonable to assume that the relic recovered from the tomb represents a dhatu, or corporeal relic of the Great Teacher, otherwise there could be no possible reason for calling the tomb the " Mahachetiya of the Lord, the Supreme Buddha." That the Mahachetiya was regarded as a particularly holy shrine is obvious from the tone and wording of the inscriptions found at the site. Again, the size of the tomb, the number of pious donations made by ladies of royal blood, and the fact that pilgrims came from ail over India and Ceylon to reverence it, afford testimony of this. Unfortunately, the meaning of some of the words and phrases met with in the inscriptions is very obscure. Commenting upon this, Dr. Vogel says-"A considerable difficulty in the way of interpreting the Nagarjunikonda inscriptions is the want of precision of which they show ample evidence. Considering that these inscriptions were meant to be perpetual records of pious donations made by ladies of royal blood, the careless manner in which they have been recorded is astonishing. Not only single syllables but whole words have been omitted," Dr. Hirananda Sastri, Epigraphist to the Government of India, who has also made a study of these inscriptions, found the same difficulty, and, as might be expected in the circumstances, his interpretation of the precise meaning of certain words differs from Dr. Vogel's. The records belonging to the Mahachetiya open with an invocation to the Buddha, who is extolled in a long string of laudatory epithets. Dr. Hirananda Sastri is of opinion that the style and wording of the invocation shows that the Mahachctiya has been specified in these inscriptions as " protected by the corporeal remains of the Buddha and that the genitive case is used here to discriminate this stupa from others not larly consecrated. Nine ruined stupas were discovered at Nagarjunakonda, four of them highly decorated with stone bas-reliefs similar to those recovered from Amaravati, but the Mahachetiya is the only one bearing inscriptions indicating that it was consecrated to the Buddha. The discovery of the relic and the fact that inscription B. 2 of Dr. Vogel's List, definitely gives the name of the monument as the Mahachetiya of the Buddha, seem conclusive evidence that the monument was originally built to enshrine some corporeal remains of the Buddha, as Dr. Hirananda Sastri maintains. The stupa was probably built long before Chamtisiri set up the pillars and rebuilt the structure in the second century A.D., or there abouts, which p. 192 would explain why the inscriptions give no information about the consecration or how the relic was obtained. If the Mahachetiya did exit prior to the second century A.D., the fact that it contained corporeal remains of the Great Teacher would have been known throughout India and Ceylon, thus making it unnecessary to record this information in inscriptions added to the monument in later times. We know from the inscriptions recovered from Sanchi, Sarnath and Amaravati that the great stupas that existed at these three famous sites were all rebuilt in later times. These inscriptions give the names of some of the pious donors who found the money for the additions to these monuments, but, like the Nagarjunakonda, inscriptions, they give no information concerning the purpose for which the stupas were built, or when they were erected, just the very points which we should so much like to know. The Amaravati inscriptions show that the stone casing, ayaka-pillars and stone railing were added to the Great Stupa at that place in the second or third century A.D., that is, at the same period as that in which Chamtisiri set up the pillars and rebuilt the Mahachetiya at Nagarjunakonda. Originally, the Amaravati Stupa seems to have been a,plain brick and plaster stupa similar to the Mahachetiya, and it must have been a particularly holy shrine, else it would never have been enlarged and decorated in so costly a fashion. Perhaps when Chamtisiri learned what was taking place at Amaravati, she felt it incumbent upon herself, as the leading devotee of the Buddha at Nagarjunakonda, to redecorate and improve the Mahachetiya. Personally, like Dr. Hirananda Sastri, I do not think there can be any doubt that the Mahachetiya was originally built to enshrine some corporeal remains of the Buddha, and that the fragment of bone found in the gold reliquary represents a genuine dhatu, or relic, of the Great Teacher. There is no reason why such a relic could not have been obtained from Northern India long before the days of Chamtisiri.