The part of the lake with which I am now dealing forms a clearly marked and separate basin. Its western boundary is, as we shall soon find, very sharply defined, while the eastern boundary coincides with the narrow sound between the two projecting pier-like promontories. At its widest part the basin probably measures 5 km. across, unless this estimate is exaggerated by reason of the atmospheric refraction. Towards the west however the basin rapidly contracts to a breadth of I km., and finally is not more than 35o m. across.
The wide insular country which we had all the way on our right hand is traversed, as I have said, by three medium-sized ranges, rising to a few hundred meters in height. We have already seen how they terminate eastwards in three headlands jutting out into the lake; the continuation of our trip revealed their western terminations. The first range, the one farthest south, is fairly short, and at its termination a low tongue of sediment juts out from a slight incurving of the shore. A precisely similar, but somewhat narrower, headland projects from the same shore about midway along the southern range. Both promontories point towards the east-north-east, parallel with the shore-line. Thus the former one, that at the sound, bends towards the east. This cannot be due to pure chance, but must be an effect — as the sedimentary capes are in the north Tibetan lakes — of the prevailing wind ; which thus appears to come from the west or possibly the south-west, and gives rise to a current along the southern shore of the great island, forcing the solid sediment to deposit itself in the manner described above. Where this sediment comes from was at first inexplicable, for the water of the lake was as bright as crystal, as if it issued out of the purest spring; but it does not require much reflection to trace the source of its origin. Consider first the westernmost of the
IIIF Color HighRes
