________________
178
THE INDIAN ANTIQUARY
( SEPTEMBER, 1932
shows it to be Vinnu. Similar traces of the once existing symbols are visible, but much more faintly, on the bricks behind the two other statues."7 The standing position, most commor. to late medieval Visnu images in India, the smooth and refined modelling of the slender body, the beautifully executed ornamente, and above all the physiognomy of the figures suffice to show that they are frankly Indian in character, belonging to a period not later than the eleventh century A.D.
Now we come to the images in the niches of the outer walls of the temple. The niche at the eastern end of the south wall is occupied by a representation of the Varâha avatára of Vişņu (fig. 5). The figure, like all the others, is badly mutilated. The boar-head has specially suffered, but the attitude of the legs and the position of the head turned towards the left shoulder, on which the defaced female figure of the seated Bhûdevî can be clearly noticed, are more than convincing. The heavy chignon of the goddess falling on her back and the hands clasped in adoration are represented with a thoroughness of detail. The hands are mostly broken off ; the attributes cannot, therefore, be recognised, but the mace (gadá) held at the middle by the left lower hand, as well as the petals of the lotus throne are clearly distinguishable. It may be mentioned here that the attitude of the two legs which is generally determined by the attitude of the head, has here been to some extent misunderstood. If the head is turned towards the left, it is natural for us to expect, from similar examples in India as well from artistic requirement, that the left leg should be bent and the right kept straight and strong, or vice versa. But, here, though the head is turned towards the left, the leg bent is the right one and not the left.
One of the niches is occupied by a representation of the Narasimha avatára of Vişnu (fig. 6). The attitude of the legs with their knee-joints bent forward as well as the lower hands holding something in the lap are interesting, and are the determining factors for the identification of the divinity. The figure on the lap is completely gone, traces of the stone are only left; but the lines of the lion face of the principal figure with at least six hands and the sharp nails of one of the hands that are used to rend the body of Hiranyakasipu that can yet be traced, leave no doubt as to its being the Narasimha avatára of Vişnu.
A third niche is occupied by a representation of a two-handed divinity standing erect on a lotus throne with the head slightly bent towards the left (fig. 7). The head is crowned with the usual but peculiar head-dress with flame designs on two sides ; but the face is mutilated and it is difficult to make out anything. The god is dressed to the knees, and the hands holding respectively the arrow and the bow at once show that it is a figure of the Ramachandra avatára of Vişnu. The iconographic texts would lay down that "Ramachandra should never have more than two arms; in the right hand the bana, or arrow, should be held, and in the left hand the dhanus, or the bow;"! and the present icon strictly follows this text. But some of the texts demand that an image of Sri Rama should be a standing one, with three bends in the body, in other words, it has to be a standing image of the tribharga variety, an injunction followed in most of the South Indian images of Ramachandra. This has here been disregarded ; nor is the divinity accompanied by Sita, Laksmana or Hanumana, as laid down in certain other, especially South Indian, texts.
A fourth niche is occupied by an image easily distinguishable as Parasurama (fig. 8), another avatára of Vişnu. The figure stands on a lotus throne flanked by two full blown lotuses; it is crowned by the usual head-dress and decorated with simple ornaments. The attitude of the body is erect, but the head is slightly bent towards the right. The hands,
6 T. A. Gopinatha Rao, Elements of Hindu Iconography, vol. 1, Part I, p. 189. » Ibid., p. 186.