________________
131
phasimilar in rear parsvan
Description and Chronology of Caves century CE. This panel is crowded with figures, carved in two vertical rows on each side of Parávanātha. They represent the attackers in various postures. It is quite similar to the scene depicting Mara's attack on Buddha in Cave XXVI at Ajanta. The differences are the seated posture of Buddha and presence of Mara's daughters. Most of the other panels are of the later period and are from south India (Sivaramamurti 1983: Plates 33. 37, 80, 94, 121, 128). All are rock cut and pre-date Ellora counterparts, though a few steles are known from post-Ellora phase also. The examples from Badami and Aihole are the earliest. All these panels are similar in representation. The figures of Padmāvati holding the chhatra, Kamatha in añjali mudrā near Parsvanātha's feet and a stone thrower above Parsvanatha are common to all. The panel from Karaikoyil has figures of tuo more attackers. There are some minor variations. But all the panels have minimum figures. Compared to these, the panels at Ellora are very elaborate with large number of attackers in various attitudes, carved in great details. Thus, though the concept of such panel was borrowed from south India, it was experimented and evolved by Ellora artists. No other site, earlier or later than Ellora, has such wonderful depiction of the scene. Some of the later representations display clear influence of Ellora. Thus, the representations from Mulgunda, Gadag district, of 902-903 CE and Hombuja, Shimoga district, Karnataka retain the figures of Padmāvati holding chhatra above the head of Parsvanatha, the stone thrower and Kamatha seated in añjali mudrā near Parsvanātha's feet, much in the same fashion as at Ellora (Nagarajaiah. 2000: Plates 13, 76, 78). One of the steles from Hombuja has a figure of a fierce looking female with pendulous breasts, holding a dagger in the right hand and a sarpa in the left hand (Nagarajaiah 2000: Plate 77). This figure is identical to the female figure in J20 at Ellora. Apart from these similarities, these later representations show a number of differences also as a few small figures, including some animals like elephant, are introduced. A clear difference can be seen in the representation of Dharanendra. While in all the representations from Ellora he is shown in the snake form above the head of Parsvanatha, in the cases of aforementioned two steles from Hombuja, he is shown in both snake and human forms.
The Bahubali panels of Ellora are more akin to the representations in south India than north India. Though, the panels are similar to Badami-Aihole examples, as Maruti Nandan Tiwari and Kamal Giri have shown, it was at Ellora that the panel was depicted in most details and the process of raising the status of Bahubali to a Jina was begun (1988: 335-344).
Intrusive icons
Another important feature to be noticed is that in some of the caves, a number of icons are intrusive in nature, as suggested by their placements. They are carved mostly on pillars and pilasters or whatever space available. Most of such icons are found in J10 and J25. The icons on pilasters and the pillars of J10 and in J25, the icons on the pilasters, pillars, on the front wall of the verandah, above the door to J26 and near shrine doorway are intrusions. The icons of a Jina and a goddess, carved outside the cave are also intrusions. In J18, the figures of Padmāvati near the door to J20 and a standing figure of Jina near the door leading to J19 are carved later on. The image of seated Parsvanatha between J23 and J24 is also an independent donation, which does not have any relation
