a bay of the Tschargut-tso, the rudiment of which still survives in the form of a little bay. However that may be, the two lower terraces at all events are a proof that the level of the Addan-tso was formerly higher than it is now. Nobody, I suppose, would maintain, that these beach-lines have been made by fluctuations of level at the present time; that would be tantamount to asserting that the Addan-tso rises several meters in the spring when the snows begin to melt on the mountains to the south, or later in the summer and in the autumn during the rainy season. For in proportion as the inflow into the lake increases so will its outflow through the Tsangmo-rapga, and the resultant effect at any rate will be, that the lake will maintain itself at the same level. Such fluctuations as do take place in its niveau are certainly so insignificant that they cannot under any circumstances be brought into connection with the terraces in question. Around the Selling-tso I found terraces the altitude of which above the existing level of the lake I estimated at about 5o m. ; and the shrinkage that is going on in that lake I attribute to a diminution in the rainfall. To the same cause we must also ascribe the drop in the two freshwater lakes, though in them the subsidence takes place much more slowly, for in their case another factor steps in as a modifying agent, namely fluvial erosion. For if you conceive these two lakes to have been ten meters higher than they are now, that is to say as running together to form one long lake, with a narrow sound between them, the shape of which can be outlined thanks to the terraces on the peninsula, then the level of the Jagju-rapga would also have been ten meters higher than it is now, otherwise a position of equilibrium could not possibly be maintained. In order to bring about a drop of ten meters in the lakes, that is to bring them down to the level at which they now stand, an erosion to that same amount must be posited in the Jagju-rapga. This erosive activity is still going on, and is accompanied by a parallel fall in the two freshwater lakes. But if we suppose the diminution in the rainfall to take place so rapidly that the erosive activity of the river is unable to keep pace with it, then we should sooner or later have this result, that the Tschargut-tso would cease to emit any water into the Jagju-rapga, but on the contrary would become cut off from it and would gradually turn salt. And the same fate would happen to the Addan-tso. Later on we shall come across lakes in which this process actually has taken place, and amongst these are instances of lakes which are on the point of totally disappearing off the face of the earth.
As for the fourth member in this peculiar lacustrine family, namely the Naktsong-tso, we have ascertained that it occupies a more independent position than the other two freshwater lakes; for, while it does receive affluents from the south and west, it has, on the other hand, no visible outflow.
Leaving behind us the north-west bay of the Addan-tso, we travelled for some distance westwards along its northern shore, then struck off to the northwest, making for a not particularly high pass in the range which separates this lacustrine valley from the next latitudinal valley to the north, and which, I imagine, is the one along which Littledale travelled. Here we found various small freshwater pools and numerous nomad tents. In the district in which we pitched Camp LXXXVIII, at an altitude of 4,686 m. the grazing was not particularly good. Meanwhile the caravan had made its way from Camp LXXXIV on the east side of the Tschargut-
IIIF Color HighRes
