the well-preserved condition of the rampart. A few days later on this same narrow isthmus between the Selling-tso and the Naktsong-tso I observed no fewer than seven similar beach-lines, and with one of them the rampart of which I have been speaking is connected.
From the isthmus we had a magnificent view towards the east — the Naktsong-tso with its picturesque girdle of mountains and its expanse of bright blue limpid water. At the first glance I could hardly believe but that it was connected with the Selling-tso, but as soon as I reached its shore I saw clearly that it was another lake. Its water was as fresh as spring-water and contained Algæ. The question that now faced us was, whether we should go round it by the western shore or by the northern. It was evidently the same lake that Bower places south of his Caring Cho. But with regard to its shape and its situation Bower's map was not of the slightest use; for, in the first place, Bower only touched one part of it, and, in the second, his map is drawn on far too small a scale, so that from it it is impossible to draw any even the broadest inferences. He also was compelled, as I was, to turn back on the southern shore of the Selling-tso, and to go back by the way he had come, namely to the Tschargut-tso. The only thing that his map showed as beyond a doubt was that he had travelled between the Selling-tso and the Naktsongtso, notwithstanding that the shape of the latter is distorted to such an extent as to be unrecognisable.
A hurried reconnaissance southwards along the western shore of the lake showed that it would not be possible to travel that way with camels, at least not without great difficulty and making wide detours. Accordingly we turned back and travelled along the northern shore, passing round the beautifully curving bay, which swells out northwards towards the Selling-tso, and then kept principally towards the
IIIF Color HighRes
