course of the year. Both these lakes, the Tschargut-tso and the Addan-tso, are no doubt covered with thick ice in winter, and after the cold season has set in in the circumjacent mountains the inflow into the two lakes will be reduced to the smallest possible amount. But on the other hand these two lakes are so extensive and so deep that the Jagju-rapga might well continue to flow for a considerable time without causing any appreciable drop in the ice-bound lakes. Add to this the probability that both lakes may be fed by a countless number of subterranean springs, all of which help of course to maintain them at the same constant level. In any case it is conceivable, that even during the long winter the drop which takes place in the river is extremely slight, or perhaps there is no noticeable drop at all. Marks on the eroded banks prove conclusively, that the river could not rise higher than the level at which it stood on 9th September, but it could of course be lower. In that case, the river would at the period of our visit be standing at its highest level, in consequence of the upper lakes having been filled with rain-water during the past summer. I think it most probable however, that the oscillations of level have a rather small range.
The volume of the Jagju-rapga was barely half as much as that of the Satschutsangpo a few days earlier (56.5 cub.m.). The latter was however dropping, and it would not be long before it dropped to the same dimensions as the former. But whereas the Jagju-rapga would remain constant during the winter, the Satschutsangpo would continue to shrink until it reached an exceedingly low ebb. In other words, the Jagju-rapga is active all the year through, while the Satschu-tsangpo confines its activity to the rainy season, but at all other times, and especially during the cold period of the year, leads a moribund existence. When you consider the enormous quantities of water that the Satschu-tsangpo pours into the Selling-tso during the rainy season, you are inevitably driven to the conclusion, that, taking the year as a whole, this river yields on the average a far greater tribute to the lake than does the constantly active Jagju-rapga. And a superficial estimate is enough to confirm this: if the Jagju-rapga contributes to the lake a constant volume of 26 cub.m. in the second, it will pour into it during the course of the year a total volume of 82o million cubic meters; if the Satschu-tsangpo pours into the lake 150 cub.m. in the second for the space of three months, that makes a total for the year of i Ioo million cub.m. _ And even though this figure, 15o cub.m. in the year, be too high, the amount which flows down the river during the remaining three-quarters of a year will be sufficient to bring up the total volume to 1200 million cub.m. The Jagju-rapga will therefore be only two-thirds as big as the Satschu-tsangpo.
Besides these two rivers I saw only one other entering the Selling-tso, namely the Alan-tsangpo, and it carried on i oth September a volume of 7 cub.m. So far as I touched the southern shore, the lake received no affluent on that side, except perhaps subterranean affluents, and Littledale found no river on the east side. All the sanie during the rainy season the lake will certainly be entered by a number of temporary tributaries from all sides. The drainage-area of the Selling-tso is particularly extensive; indeed of all the self-contained drainage-areas of the Tibetan highlands, there is only one, namely that of Kum-köl, that can compare with it. Since however the lake is now shrinking, it is fair to infer that the amount of
IIIF Color HighRes
