Sec. i] SIMHAPURA REVISITED 49
General Cunningham has given a general account of the position of Ketas, and Mr. Talbot's Gazetteer usefully supplements it with details regarding the sacred pool supposed to have been formed by the tears which rained from Siva's eyes at the death of his consort.' It is the scene of a great annual pilgrimage attracting many thousands of Hindus in the first days of the month of Vaigâkha,8 and its sanctity accounts for the numerous shrines and Dharmas5.15.s erected around it in modern times. General Cunningham has briefly described such remains of old temples as were seen by him on the terraced slope of the flat-topped hill overlooking the sacred pool from the west. They had all suffered badly by decay and clumsy repairs, and several of the smaller shrines mentioned by him as corn-prised in the Sat-ghara group of temples have disappeared since his time. The rough sketch plan ( Plan 5) , prepared by Dr. FA bri, will help to convey some idea of the present state of these ruins.
As already correctly pointed out by Cunningham, the general style of these temples in their constructive features and such decorative details as are still recognizable shows close similarity to that of the temples of Kashmir dating from the eighth to the tenth century A. n. Hence they can safely be assigned, like some other and better-preserved structures of that type in the Salt Range (Amb, Malot; also Tila Käfirkot, &c., on the Indus) , approximately to the same period. But there can be no doubt that the successive terraces on which these temples are built are of much greater antiquity. They are all very massively built with a facing of large dressed slabs of sandstone. The north side of the terrace on which the three topmost temples, A—C, stand, rises at its north-east corner to 19 feet above the present ground-level and is divided into two stories by a boldly projecting moulding. The upper one still shows traces of seven flat pilasters.
A similar decorative scheme, reminiscent of that common at the bases of Buddhist stûpas in Gandhâra and Swat, is to be seen somewhat better preserved on the north face of the basement on which the ruined structure, D, is built. The westernmost of the pilasters which divide this face at a distance of 4 feet 3 inches from each other, still retains the outlines of a surmounting double bracket once stuccoed. This structure D, which was obviously not a temple but a hall, or mandapa, has undergone great alterations, and its original dimensions ( about 40 by 43 feet) are uncertain. But the superior masonry of the older walls and the `Buddhist rail' pattern of the open-work filling in the two
7 See Gazetteer of the Jhelum District, pp. 41 sq.
8 The great religious fair visited by Muhammadans almost simultaneously at the tomb ,of Saidân Shah Shirazi, at Chôa Saidan Shah, is another instance of that `continuity of local worship' so
frequently observed at old sacred sites in India and elsewhere in Asia; cf. my paper in J.R.A.S., 1910, pp. 839 sqq.; also Râjatararigini transi., ii. p. 340.