schung, and it issues, I was told, from a lake to the north; this would explain the absence of water-marks on its banks. In fact, we have here a repetition of the same relations as in the Jagju-rapga, in which the Addan-tso and the Tschargut-tso moderate the river's flood and distribute its volume over the seasons of the year. It is possible later on in the winter that the level of the river drops below that at which we saw it, and that it stood then, on the 9th November, at its maximum. On Deasy's map, on which this river is entered, one of its branches, coming from the north-west, is shown as issuing from a lake, the Bum Cho, and a second as emerging from one of the usual desiccated basins of the lake. But Deasy had no opportunity, any more than I had, to follow the river up to its source, for our routes intersected on its bank, and whereas I left the entire system on my right, Deasy's route crossed over it and proceeded northwards. He may therefore be mistaken in representing the last-named basin as destitute of water, and it may in fact be most probable that the Ravur-tsangpo really does issue out of a lake; otherwise it would hardly have carried such a big volume of water in the winter, and had it not passed through a clearing-basin, the water would have been much muddier than it actually was. We may also take it, that the west branch does issue out of a lake, Bum Cho, while the eastern originates in springs in the dry basin, and this would explain why the water was not perfectly clear. The surface of the stream was then only very slightly lower than the ground at the bottom of the valley; but of deeply eroded scarped terraces there was not a trace to be seen. In fact the actual bank was frequently nothing more than a slight swelling, with a scanty sprinkling of vegetation.
Meanwhile we kept along the left bank towards the north-north-west. Although the glen widens out a little, it still continues to be only narrow. Its bottom is remarkably level and rises uniformly; yet it is so far noticeable that the water often broke over small cascades. On both sides the glen is shut in by gravel-and-shingle hills, those on the west being the bigger. Countless gullies furrow them on both sides, though they were then all dry. At a distance of 5 or 6 km. to the west-south-west rose a more dominating protuberance, called Jagar; evidently it belongs to the mountainous region which I have already mentioned as bearing that name. "I'he river is generally divided into two or more branches, which embrace mud-banks between them. At a point where the entire volume was gathered into one stream we forded it, and then continued along the right bank. Just there the river was joined by a tributary coming from the west and bigger than any of the others; although it was dry, it evidently comes from a long way off. Higher up the river is broader and shallower, and is divided into several superficial undecided arms. We now left this river behind us, as also the range which it cuts its way through, and came out upon an open plain, one of the usual basins, with a dry clay bottom, flat and open, and stretching from north-west to south-east. This basin corresponds to Deasy's Kachon. He encamped in it in the same spot that I did, and his altitude, • 4451 m., agrees remarkably with my own, which was 4447 m. From that point I travelled south of that part of Tibet which Deasy explored in such an accurate and admirable manner. Meanwhile however no further attempt can be made to explain the hydrographical position of the Ravur-tsangpo; for I now travelled away from
IIIF Color HighRes
