shore which was more directly exposed to the westerly or south-westerly wind and to the action of the waves. But the lake was never especially big, even when its level was 2 5 ni. higher than it is now; but it formed at all events a sufficiently extensive sheet of water to give rise to strand-terraces, though their formation was not possible on the western or southern shore. How matters stand on the northern shore, I was unable to ascertain, owing to the great distance, but in all probability there are beach-lines all the way along that side. From that point of view, it would have been more instructive to have travelled on the north side of the two lakes, as well as shorter, but the grazing there was said to be inferior. The strand-terraces are gapped by a great number of energetically excavated watercourses, looking like crevices and fissures slashed across the mountain's flanks.
At length we left this lake region definitively behind us and climbed up to an unimportant threshold in the northern hills, whence we obtained our last look across the western part of the lake. To the S. 40° E. appeared the extreme little promontory belonging to the range that borders the lake on the north. At its foot is a grassy plain, sloping gently down towards the lake, and then follows a chaos of alternate sheets of water and gypsum formations. In the shallow parts the water, owing to the white gypsum underneath, assumed a shade of light green.
From the top of the pass we beheld an unexpected sight, namely right down below us and east of the threshold a narrow glen, some 4o m. deep, carved through the beds of gravel-and-shingle, out of which the ridge is composed that possesses the four terraces mentioned above, and at the bottom of the glen was a river, bigger than any that we had seen for a long time, in fact considering the season it was unexpectedly large for the western part of Tibet. Since the gravel-and-shingle ridge forms, as it were, a connecting link or bridge between the higher parts of the mountains on the north of the latitudinal valley, this glen ought to be regarded as a transverse glen. Not very far from the point where we first caught sight of it, probably one kilometer, the river, after making a sharp bend, emptied into the northernmost bay of the lake. The existence of this river, which was then pouring into the lake a volume of some 3 cub. m in the second, accounts effectually for the freshness of its water, a thing which had at first puzzled me. On the mountain slope, immediately east of the river's mouth, three terraces were distinctly discernible, corresponding to three out of the four which I have mentioned above. That only three were visible at this spot may have been due simply to local peculiarities. These were even more energetically defined than the preceding, but the slope on which they were is more exposed to the beat of the waves from the west.
We went down to the river by a short steep slope, and then turned up the stream on its east or left bank. At the ford where we crossed over it the breadth amounted to 16 m., the mean depth to 0.30 m., the mean velocity to 0.75 m., and the volume to 3.4o cub.m. in the second. The stream was flowing in a lively and noisy fashion, carrying with it a great number of ice-sheets. It was only the relatively quiet bays that were frozen. The water was not perfectly clear, but its slight turbidity may be mostly accounted for by the collision of the drift-ice against the banks. Although all the conditions were favourable, there were said to be no fish in this river. The name given to it by the Tibetans is Ravur-tsangpo or Ravur-
IIIF Color HighRes
