________________
Description and Chronology of Caves
125 background of the period, he concludes that these caves were excavated in the reign of Rashtrakuta monarch Amoghvarsa, between 815 to 878 CE (1977: 25-28).
Thus, these are dated roughly to gth-10th century CE. It is an accepted fact that the Jaina caves post-date the Kailasa monolith and are stylistically similar to Lankesvara. Though some scholars believe that Kailasa was excavated over a long period. lasting from 735 CE to 12th-13th century CE (Goetz 1952: 107), Pereira has rightly pointed out that the Kailasa is a weil conceived, meticulously planned unit, and was started as well as finished by the Rashtrakuta king Krisna I during 750-780 CE (1977: 95-101). The Lankeshvara cave, excavated in the northern rock face of the pit in which the Kailasa stands, soon followed.
The composite type of the pillar, comprising of both pürnaghata and kunibha or amalaka, occur in Lankesvara for the first time and are used extensively in the Jaina caves. The other two types of pillars in the Jaina caves are also stylistically closer to the pillars in Lankesvara. Even the minute details of the pillar ornamentation such as the band of festoons, the band of lozenges alternating with rounds or flowers, the beaded rim of kalaša, the alpanāsis on the body of the kalasa, the taţi in double unit, a band of beads in slanted rows between two tatis, the floral or creeper design on the median band of the kumbha are common to Lankeshvara and the Jaina group. The only difference is that while the pillars in Lankesvara cave are short and stout, those in the Jaina caves are longer and slender.
The ground plans of these two groups have some similarities and a few variations. Lankesvara, with its sāndhāra layout and the navaranga hall is more close to Kailasa, though the pradaksiņā at Kailasa is from outside. On the other hand, the Jaina caves, with verandah, hall and shrine as basic units and hall with pillars arranged in a square, are akin to the Buddhist Caves II and III. Secondly, Lankesvara is open from the side and not the front, as the Jaina caves are. However, unlike Kailasa as well as the Buddhist caves, the hall in the Lankesvara and the Jaina caves, is open from the front. The parapet wall, dividing the hall from the verandah, carved with mithuna figures and with a chandraśila in the centre, is almost identical in both the cases. The Jaina caves also share some of the features with Kailasa. The kaksāsanas in the verandah with dwarf pillars and decorated with purnaghatas alternating with a pair of round pilasters, mattavarana carved with elephant heads, kaksāsanas carved with pürnaghatas, kapota depicted with flying figures and vallimandala and stone rafters on the underside of the kapota are some of the common features. The depiction of the lion figures on the mattavārana is the new feature introduced in Jaina caves. Another unique feature of the Jaina caves is the treatment of the facade as the sikhāra of the vimāna, depicting sala and kuta śikharas. Though it is not unprecedented to have the frontage of the cave treated as a vimāna, right from the adhisthana to śikhāra, as it is found in a number of Pallava examples, it is a new feaiure at Ellora. Though the Buddhist Cave X displays the façade with a large pañjaranāsi flanked by smaller pañjaranāsis, all the features of a vimāna, right from adhisthana to sikhāra appear for the first time in Jaina caves at the site. The paintings in the antechamber ceiling of Kailasa are very similar in theme, style and the colours with the painting in the Jaina caves (Plate 83).
Thus, it is clear that the Jaina caves are stylistically and conceptually close to Lankesvara and also Kailasa to some extent. Therefore, these must have been initiated in
