THE NAKTSONG-TSO - EAST AND SOUTH. 53
three promontories; we see that it lies immediately east of the point where the western half of the latitudinal valley between the southern and the middle range reaches the lake. During the violent rains which fall in this region in the summer there flows clown this valley a not inconsiderable stream, formed by the confluence of the numerous small torrents which issue out of the short, steep side-glens that seam the mountain-sides. Vast quantities of fine, powdery, disintegrated material are in this way carried down to the lake, and are mostly swept eastwards by the current along the shore, and this settles gradually against one or other of the pier-like headlands, thus constantly lengthening them. Now let us consider the middle promontory; it is formed principally in the same way, the only difference being that in this case there is no main valley to gather up the torrents from the small side-glens, but each torrent pours its sediment-laden water directly into the lake. There the sediment is caught by the current spoken of above and deposited against the middle headland, which is smaller, and narrow, in consequence of the drainage-area by which it is fed being less extensive. But it is more difficult to explain why the eastern promontory should be the largest, because the drainage-area upon which it draws is the smallest. Possibly the cause is to be sought partly in the configuration of the lake-bottom: that is to say the lake is just in this part rather shallower than it is - farther west. At all events it is very unusual to find flat mud peninsulas at the foot of a range that plunges almost vertically down into a lake. The primary condition of their origin is of course the shallowness of the lake. There appeared to be small promontories also along the southern shore.
After passing the third promontory and casting a glance up the latitudinal valley between the middle range and the western range, we continued our trip towards the west, skirting the southern foot of the middle range. The strip of shore was somewhat broader than before, and the grass on it was so excellent and so untouched, that I concluded it was inaccessible to the flocks of the Tibetans, especially as these were now grazing in other parts of the great island. Very likely impassable mountains also cut off access to it from the landside. There were, it is true, animal droppings, proving that that quarter of the island is indeed visited; but they will be left in the winter, when the flocks can easily be driven across the ice from the south.
The middle range, which is in general higher and wilder than the southern range, terminates at its western end in a tapering, perpendicular, and rather large rocky gateway, and below it, on the actual shore-line, rise a couple of small isolated rocky pinnacles. By this the lake had shrunk to a sound only 35o m. broad, and over on its opposite or western side the middle range is continued in another range, which however we only saw foreshortened. Its eastern end forms a great rocky mass, wild and rugged, with steep, bare, fantastic flanks, which plunge down sheer into the lake without the smallest strip of shore at their foot. From that point the southern shore was seen to extend towards the S. 37° E. At the southern foot of the rocky mass was a solitary stone hut, then uninhabited. Somewhat nearer towards the south rises a small detached butte. The wide, marshy plain on the south of the lake appeared to continue a fairly long way towards the west, merging into a latitudinal valley. In this we should, I have no doubt, have found a con-
IIIF Color HighRes
