________________
198
THE INDIAN ANTIQUARY
[OCTOBER, 1932
up in a knot; and there should also be a pair of kundalas in the ears, a deer-skin worn in the upavita fashion, the sacred thread, a waist-zone and a kaupina... He should also carry with him a book. All these are intended to show that the image is that of a Vedic student or Brahmanical brahmacharin." Some authorities hold that the image should be represented as a deformed dwarf, and they, therefore, require that "the image should be worked in the form of an ill-shaped man with hunchback, protruding joints of bones and a big belly."13 The image under consideration, deformed and dwarfish as it is, seems roughly to agree with the latter description.
The other one, we have already surmised, is Kalki (fig. 11). It can be described as a standing image with two hands, holding in the right a khadga, and in the left an attribute that can be distinguished. It has the usual head-dress, heavy ornaments, and a loincloth folds of which can easily be distinguished. According to the Agni-purana, Kalki should ride on a horse and carry the dhanus and the bara, but the Vaikhanasagama states that he should have the face of a horse and the body of a man with four hands carrying respectively the samkha, the chakra, the khadga and the khelaka. But, in the present example the image neither rides a horse nor has the face of a horse. The only attribute that is distinguishable in one of the two hands, and on the strength of which we make the identification, is the khadga. It is not impossible that the left hand carried a khetaka or shield. And once we have made sure of our identifications of six of the avatáras of Visņu-namely, Buddha, Varâha, Narasimha, Ramachandra, Parasurama and the Trivikrama, and when the present one is not any of the remaining three avataras, namely, Matsya, Karma or Krsna, we are led by a process of elimination to identify the present icon as the Kalki avatara of Viṣṇu, and assert, in the same breath, that the three niches that are now empty once sheltered the images of the Matsya, Kurma and Kṛṣṇa avatáras of Visņu, the most important god of the Hindu Triad.
III. Art and Historical Background.
The Nât-hlâung images belong undoubtedly to the late medieval period. It is also evident that they were executed by Indian artists, probably imported for this purpose. We have already discussed the South Indian Tamil inscription palæographically dated in the thirteenth century A.D. We have tried in that connection to show that the epigraph refers not to the erection of the temple itself, but to a mandapa and a door, which might well have been added later on. The temple cannot, in our opinion, if we are to judge by the sculptures in its niches as well as by the architectural style, date later than the middle of the eleventh century A.D.
In view of the fact that a Tamil inscription has been discovered in the debris of the temple purporting to say that the temple had been founded and resorted to by Vaisnavas from various parts of the Peninsula, and that the iconography of some of the images (e.g., the Surya) are distinctly South Indian, it is only natural to jump at the conclusion that the Nat-hldung images owe their artistic inspiration to a contemporary South Indian school of art. Had it been so our problem would have been as easy as one could expect in such cir. cumstances. Unfortunately, such an assumption is not borne out by a careful consideration of the sculptures themselves.
The images are all very badly mutilated, and it is difficult to make a systematic survey of each individual image; but the general features and characteristics are easily recognisable from what remains of some of them. The forms and attitudes are mostly conventional, nor
13 T. A. Gopinatha Rao, Hindu Iconography, vol. I, Part I, pp. 163-64.