Now we were already far to the west; the force of circumstances had forced us to leave behind us step by step ever larger areas of unknown country to the north. I was vexed, but I would, at any rate, endeavour to do all that was possible in my hampered condition. At Shamsang, Ryder’s Lahtsang, we were at the place where the actual source streams of the Brahmaputra converged from various directions. I had long determined to push on to the unknown source, unless the Tibetans placed unsurmountable obstacles in my way.
The learned and clear-sighted Colonel Montgomerie had sent Nain Sing in the year 1865 up the valley of the upper Brahmaputra (Illust. 380). From our Shamsang the Pundit crossed the Marium-la, and said in his report that the sources of the river were certainly in the huge chain seen in the south, and were fed by its glaciers. He did not, however, go to look for the actual sources, but continued his journey westwards.
The next year, 1866, Thomas Webber made an excursion into Tibetan territory, and his route lay a little to the south of Nain Sing’s. On his sketch-map it may be seen that he crossed some of the source streams of the Tsangpo, but of the tract in which the sources are situated he gives no further indication than “Snowy ranges unexplored.” And when he says in his text that here are the sources of the great Brahmaputra, which have their origin in the Gurla glaciers, the confusion is hopeless; for the sources of the river lie 60 miles from Gurla, a mountain which has nothing whatever to do with the Brahmaputra.
The political expedition which, under the command of Rawling in the close of the year 1904, had Gartok for its destination, and the chief result of which was the admirable map of the upper Brahmaputra valley surveyed by Ryder and his assistants (Map 7), travelled from Shamsang over the Marium-la and north of the Gunchu-tso to Manasarowar. It was therefore of the greatest importance to me to travel to the south of their route through country they had not touched on. They travelled by the same road as Nain Sing, and left the source of the river at a distance of 40 miles to the south. From Ryder’s report it might be supposed that he considered the Marium-la to be the cradle of the Brahmaputra; but in a letter I have recently received from him, he states that such is not the case, but that he always recognized that the actual source must lie among the mountains in the south-west, which he has set down on his map from bearings taken of their peaks. Ryder also remarks in his report that the principal headwaters come from there.
Instead of entering into a diffuse discussion of the problem, I introduce in this book small sketches of the maps of my three predecessors, Nain Sing, Webber, and Ryder. No other traveller had ever been in this region, and I would on no account miss the opportunity of penetrating to the actual source of the Brahmaputra and fixing its position definitely.
How was this to be done? At Shamsang the source streams meet, and below this point the united river bears the name Martsang-tsangpo. First of all, I must, of course, gauge the quantities of water in the source streams, and, if they were nearly equal, we must be content to say that the Brahmaputra has several sources.
With ten men, the boat, and the necessary measuring apparatus, I betook myself first, on July 8, to the point on the southern side of the valley where two streams run together, the Kubi-tsangpo from the south-west and the Chema-yundung from the west. A short day’s march farther west the Chema-yundung receives the Marium-chu, which comes from the Marium-la. First the united stream was gauged, and found to discharge 1554 cubic feet of water per second, and immediately after the Chema-yundung, which discharged almost 353 cubic feet. Subtracting this from the volume of the united river, we get 1201 cubic feet as the discharge of the Kubi-tsangpo. This river is then three and a half times as large as the Chema, and it should be remembered that the Chema also receives the water of the Marium-chu, so that its 353 cubic feet represent the united volumes of two tributaries.
When we encamped in the evening with the main caravan in the Umbo district (15,427 feet), where the Chema-yundung and the Marium-chu unite, the rivers were very considerably swollen, and the water, which had been clear in the morning, had become turbid. Therefore only the two measurements taken at the same time were directly comparable, and I will pass over all the subsequent measurements. To arrive at the source we had only to know that the Kubi-tsangpo is far larger than the two others, so we had to follow its course up into the mountains, which none of my predecessors had done. The Tibetans also said that the Kubi was the upper course of the Martsang-tsangpo.
On July 9 we parted from Guffaru and the main caravan, which was to keep to the great high-road and cross the Marium-la to Tokchen, while Robert and I with three Ladakis and three armed Tibetans followed the Kubi-tsangpo up to its source. Our way ran west-south-west. Where we crossed the Chema-yundung, a good distance above the last delta arms of the Marium-chu, the river carried little more than 140 cubic feet of water, and therefore the Kubi-tsangpo, flowing to the south-east of it, is here fully eight times as large. At the ford our Tibetans drove a peg with a white rag into the edge of the bank, and when I asked why, they answered: “That the river may not become tired of carrying its water down the valleys.”
At Tok-jonsung, where we bivouacked among some black tents, the Chema looked very large, but its water ran very slowly. The nomads of the district go up to the Chang-tang in winter. Here also we heard, as on many former occasions, that smallpox was raging frightfully in Purang, and that all the roads leading thither were closed. No country lies so high that the angel of death cannot reach it.