sion, in which is the basin that communicates with the Akâs Ganga. A short way higher up, the ridge again divides into two, inclosing a longish depression, across which two cut-stone walls were built, dividing and forming it into tanks; the walls or embankments are now in ruins and the tanks dry.
On the spur overlooking the lower tank was a small temple of coarsely cut-stone. Stone from it and from elsewhere has subsequently been used to build a modern shrine overlooking the upper tank (also dry now); it is a small, square, flat-roofed house of no interest; it is dedicated to Sivá, and a lingam stands in the sanctum.
The ascent is up the western or left-hand ridge of the two which inclose the depression just noticed; it goes up northwards, and at the first secondary peak, to the west of the great main one, it meets the other spur, which was noticed before as having, cut across it, a shallow double line to mark an intended line of walls and a gateway; from this peak the ascent goes eastwards up the slope of the great main peak.
The highest peak is crowned by a Jain or Sarawaki temple with a bulbous conical dome built of older materials. To the west of this and immediately under its walls, at the summit of the hill, is a deep chasm or hollow. I did not attempt to descend and ascertain its depth; it is overgrown with jangal which rendered a descent unsafe. The Jain temple stands to the east of and immediately on the brink of this chasm. On the rock is cut a figure of the Varaha incarnation, from which it appears probable that originally a Brahmanical temple stood on this spot.
Close to this Jain temple, which is of no interest, stands another small pyramidal-roofed temple and some other buildings, none of any interest.
I have omitted to notice a single line of inscription in characters of, perhaps, the 1st century of our era, which is cut on the rock on the ascent.
This closes my account of Mandar Hill. For a sketch of the peaks and lines of principal ridges, I refer to the plate; it is merely a sketch. I could not spare the time necessary to make a survey of the hill and of its offshoots. The hill is not encircled by a spiral mark of any kind; there is a little bit of a slanting line towards the base, and a little towards the summit, formed by the jagged end of a broken shell or outer layer of rocks, and imagination probably con-
