________________
-
8
-
pieces of wood can be seen that seem to explain the V shaped part crowning the buddhist horseshoe" - to quote Jouveau-Dubreuil. However, nothing of the kind can be seen in Thurston's photographs in his Castes and Tribes of Southern India; and on the oldest detached stone building, the hall at Cezārla dating from the Gupta period (4th-5th cent. A.D.) the gable end has the shape of a supplementary triangle with a curving hypotenuse.
Previously we mentioned the gable end of the so-called #horseshoe" which often occurs in a V shape, though also as a kind of slim pointed vase, and, on the Gaņeśaratha at Mamallapuram, as a trident on a small human head. As is well known, anthropomorphic representations appear later than theriomorphic or symbolic ones. Thus from Bagh Gumphā to the NW of Bhuvanesvar (about 100 B.C.) one knows of a tiger's head with wide open muzzle over the cave's entrance; in later styles in the South it is a lion's head and is therefore called simha-mukham. Moreover, in Amaravati two centuries later a yakşi was carved in a "horseshoe" on the top of which, and partly covering it, a śrivatsa is found (representing prakrti) which from a distance resembles a face with its tongue stretched out: the roof end.
Now this fact reminded Held, in his thesis on the Mahābhārata,' of southeast Asian gable faces on which Rassers, an ethnologist, had worked. Neither of them, by the way, used Coomaraswamy's articles any more than the latter, apparently, knew of the former. In his treatise on the Javanese theatre Rassers had shown, by means of sacred men's houses such as those in the region of the river Sepik
in New Guinea, that their front side with the initiation demon's mask on the gable is identical with the stylised triangular kekayon or hand screen which the operating dalang waves at the beginning and end of