CHAP. III. BRAHMANICAL ROCK-CUT TEMPLES. These dated caves and buildings have also rendered another service to the science of archaeology, inasmuch as they enabled us to state with confidence, even before the inscriptions were properly translated, that the principal caves at Mamallapuram must be circumscribed within the same limits. The architec- ture there being so lean and poor, is most misleading, but, as hinted above, I believe it arose from the fact that it was Dravidian, and copied literally from structural buildings, by people who had not the long experience of the Buddhists in cave architecture to guide them. But be that as it may, a comparison of the Hindu sculptures at Badami with those of Elura on the one hand, and Mamallapuram on the other, renders it certain that they were practically contem- porary. The famous bas-relief of Durga, on her lion, slaying Mahishasura, the Minotaur, 1 is earlier than one very similar to it at Elura ; and one, the Viratarupa or Vamana, is later by probably a century than the sculpture of the same subject in cave 3 at Badami. 2 Some of the other bas-reliefs are later, some earlier, than those representing similar subjects in the three series, but it seems now impossible to get over the fact that they are practically synchronous. Even the great bas- relief, which I was inclined to' assign to a more modern period, probably belongs to the /th or 8th century. The great Naga king, whom all the world are there worshipping, is represented as a man whose head is shaded by a seven-headed serpent- hood, but also with a serpent-body from the waist downwards. That form was not known in the older Buddhist sculptures, but has now been found on all the Orissan temples (for instance Woodcut No. 318), and frequently at Badami. 3 This difficulty being removed, there seems no reason why this gigantic sculpture should not take the place, which its state of execu- tion would otherwise assign to it say A.D. 700 as a mean date, subject to a subsequent adjustment. 4 In a general work like the present it is of course impossible to illustrate so extensive a group as that of the Brahmanical caves to such an extent as to render their history or affinities intelligible to those who have not by any other means become familiar with the subject. Fortunately, however, in this in- stance, sufficient literature on the subject is available by which any one may readily attain the desired information. 5 1 'Transactions of the Royal Asiatic Society,' vol. ii. plate 4. 2 Loc. cit. plate 6 ; and Burgess, ' Re- port on Belgam,' etc., plate 31. 3 Loc. cit. plates 20, 23, 40. 4 There is a second bas-relief, almost similar but in worse preservation, about 30 yards south from this. 5 Apart from the older works, refer- ence may be made to 'The Cave Temples of India' (1880), pp. 165 et seqq. ; the ' Reports of the Archaeological Survey of Western India,' vols. i. iii. v. and vi. ; and ' The Rock Temples of Elephanta' (Bombay, 1871).
