want of a better name I will call these altars ayaka-platforms in describing them here. Until I discovered them they were apparently unknown to archaeologists although the Great Stupa at Amaravati possessed them. The position of the platforms and the pillars is shown in Plate XII. The outer black line shown on plan indicates the position of the railing and the open gateways of the Great Stupa at Nagarjunakonda. The railings and gateways were of carved wood and stood on brick foundations. It was only on very rare occasions that they were executed in stone. In later times as in Ceylon, a circular brick parapet wall sometimes took the place of the wooden rails of the earlier stupas. The object of the railing was to enclose the sacred processional path which encircled the base of the stupa. In important monuments the path was usually paved with stone and within this enclosure a number of small shrines and images, the gifts of pious devotees, invariably sprang up as time went on. Outside the gateways, one on each side of the entrance, were sometimes lofty stone or wooden pillars surmounted by Buddhist symbols carved in the round, such as a wheel, or a miniature stupa symbolising the Buddha's death, It was here, where vendors of flowers and garlands had their stalls and. sold their wares to the crowds of worshippers who passed through the gates and deposited their humble offerings on the Case | facing the gateways. It is clear from the numerous beautiful sculptural slabs recovered from the stupas and also from the actual remains of ayaka-platforms unearthed, that the ayaka-platforms were regarded as the most important features of the stupas and no trouble or expense was spared in making them as beautiful and attractive as possible, All the best sculptures recovered at Nagarjunakonda originally belonged to ayaka-platforms, including some long stone beams superbly carved which served as cornice stones to these ornate platforms.
Both the bas-reliefs portraying stupas and the actual remains of such structures unearthed, show that two kinds of stupas existed at Nagarjunakonda in the second and third centuries A, D. One was a plain simple structure built of brick and plaster, while the other’ was profusely decorated from top to bottom, the lower portion being faced with carved limestone slabs fixed in mortar to the brickwork of the stupa (Plate XI (a) and (b)). In studying these excellent representations of the two kinds of stupas discovered at Nagarjunakonda, the first thing to remember is that the correct ground level of the stupa is on the line A, B. Let us take the plain type of stupa first (Plate XI (a)). Standing on the line A, B, we have the front elevation of a simple brick and plaster structure of the usual kind. In all these representations of stupas the upper part of the dome is always shown encircled by a festoon ornament, a very ancient device, and presumably, usually executed in plaster. The drum and ayaka-platforms are decorated with the rail pattern and a few popular scenes from the Jataka. The ayaka-pillars are shown with rounded tops, octagonal shafts and square bases. This is correct and all the ayaka-pillars so far discovered are of this pattern and could not have supported any kind of capital or other ornament as sometimes wrongly portrayed in the bas-reliefs (Plate XIII (a)).
On top of the dome is a rectangular coffer-like object known to archaeologists as the fee, an anglicized form of the Burmese word Avi. The tees appear to have
