40o to 500 m., expanding trumpet-fashion until it was more like a bay of the Panggong-tso. The ice in the middle of the current was however thin and would nowhere bear right across. Just before it enters the Panggong-tso, two tongues of alluvium jut out into the river, both from the right. They point towards the east, that is to say up-stream, against the current. This direction is of course imparted to them not only by the prevailing wind, but also by the waves to which it gives rise, and which run counter to the current and affect especially the surface water. Near the lower of these two capes we found a small round hole in the ice, made either by an eddy or by a spring at the bottom. The mouth of the river put me strongly in mind of the embouchure of the Satschu-tsangpo into the Selling-tso. In both cases there is a regular channel, a broad bed widening out towards the lake and terminating in an open estuary.
All day the wind blew hard from the west, so that we were all the more surprised to find a considerable expanse of ice stretching from the mouth of the river for a long way out into the lake. In fact we could only just discern its greenish blue water as a narrow ribbon beyond the white edge of the ice. This ice-sheet at the eastern end of the Panggong-tso was triangular in shape, its blunted apex entering the river-mouth, where it was directly connected with the river-ice. This was the only ice we saw on this lake, except for one or two narrow strips which we passed that same day close to the northern shore. The ice-sheet did not however present a smooth level surface, but was rugged, consisting for the most part of pounded ice, which the waves had gradually driven into this bay from the west, and there packed up in circular belts. How this ice happened to originate in that part of the lake which is most exposed to the wind is not difficult to explain. It is fresh water out of the Tso-ngombo which has spread out over the salt water of the Panggong-tso and then frozen during the still cold nights. We found that the water which was being blown into the river had a temperature of 1 '/2° to 2°, and it did not therefore freeze until it arrived at the middle reaches of the river, where it was cooled in the shallow bed. In the lowest part of the channel, where the stream is twenty to thirty times broader than it is higher up, and where the depth increases until in the river-mouth it amounts to 5 or 6 m., the current was so little noticeable that it offered no impediment to the frost. The temperature in the eastern part of the Panggong-tso was 0.9°, and consequently it too presented no hindrance to the freezing of the layers of fresh water resting upon it. Accordingly the latter freezes in narrow belts, which however are broken up the following day and the pieces driven against the edges of the nearest resistant ice. In this way the triangular ice-sheet will go on increasing during the course of the winter, spreading out farther over the lake, though I do not suppose that it advances to any very great distance unless the wind drops for a sufficiently long period to allow of the ice acquiring a certain power of resistance.
From the lower cape we rode towards the north-west, keeping along the shore. The ground next the lake consisted of barren schor, on which older, curved beach-lines were discernible, formed by the wave-beat of the Panggong-tso. A little distance from the shore is a stretch of dunes, about a meter high and overgrown with grass. The beach is crossed at that point by a dry river-bed coming from the
IIIF Color HighRes
