THE BATSA-SINGI AND THE DETSCHE-TSO. 229
it, leaving it on the north, and did not come into contact with it again. The open basin too, in which we encamped on the 9th November, and which is in great part, perhaps indeed throughout its entire length, traversed by the river, is in a certain respect a puzzle. For its floor, from the point where we emerged upon it out of the mountains, is to the eye perfectly level, and consists of fine yellow clay, cracked occasionally into polygonal patches, and presenting a shallow, level, convenient surface like a parquet floor. After passing two such expanses we came to a third, and on it we encamped. This alluvial clay evidently owes its origin in some way or other to the river. The impression I got was, that there had once been a lake there, and that it has been filled up with sediment. It may also have been laid bare through the issuing river having cut its channel deeper and deeper, so that the lake has been gradually emptied. And this is all the more likely when we call to mind, that a river flowing out of a lake higher up, and therefore with clear water, would hardly be able to fill up a basin with mud, unless the current were swollen very appreciably during the rainy season through tributaries which it picked up below the lake. This however brings us face to face with the difficulty of accounting for the absence of older erosion terraces beside the river, and such there no doubt would be had the river deepened its bed step by step. But this river had no terraces of that sort. There is also another reason why we might expect to find them. The lake into which the Ravur-tsangpo empties itself has, as its old strand-terraces prove, dropped at least 25 m., and this would naturally intensify the erosive activity of the river, causing it to advance up-stream. As however erosion terraces are wanting, we are in a dilemma as to where the explanation is to be found. It is not permissible to suppose that the river terraces have been obliterated in the soft graveland-shingle, because at some point or other their former existence is always traceable, in however slight a degree. It is more probable therefore that the big salt lake into which the Ravur-tsangpo formerly emptied, and which filled the entire valley that is now broken up into a number of depressions, sank so rapidly in consequence of changes in the climate that time did not admit of the formation of terraces alongside the river. Under these circumstances one could not of course expect, that the fall of the river would be increased in consequence of the subsidence of the lake. The difference of altitude between Camp CXXIII and Camp CXXIV amounts to 55 m. The highest strand-terrace above the lake runs at 25 m. Hence this lake formerly sent out a bay to meet the mouth of the river, which would then be somewhere in the middle of the existing transverse glen. At that time the bottom of the lake would have the same slope that the river has now. When the lake receded southwards the fall of the river, following it, would not therefore be increased, but would remain the same all through the steady desiccation of the lake. The fact of the water in the western basin of the lake, into which the river empties, changing from salt to fresh is an entirely secondary phenomenon, dependent solely upon the fact that its western half possesses at any rate a temporary outflow into the eastern half, and it being the actual terminal lake, this is salt.
As for the northern lake, which has disappeared, and in the south-eastern part of which our Camp CXXIV stood, let us in the first place consider the character of its bottom. The whole of its southern part is occupied by the clay expanses
IIIF Color HighRes
