of the north branch. Beyond the river the ground rises slightly, but remains of the same gravelly nature as to the north of it. After about six miles we came to the top of a sharp but short descent, at the foot of which were two lakelets and a few patches of grass. This is the Elesu-nor or »Sand lake», which has an outlet into another stream emptying into the Ch'u-mar.»
During this portion of Rockhill's journey, from which I have extracted above the most important geographical observations, he came into more or less intimate contact with the routes of several other travellers, e. g. Prschevalskij's and Krishna's, both of whom crossed the Marco Polo range by the Angir-takschia pass, while Rockhill crossed it by the pass of Sharakui-daban, situated somewhat farther west. Carey and Dalgleish, in the course of their short trip on the Tibetan highlands, crossed the same range twice, first by the same pass that I used between my Camps XXXIV and XXXV (1896) and the second time by the pass of Angir-takschia just mentioned. Thus Rockhill intersected Carey and Dalgleish's route immediately east of the lake which I only saw at a distance from Camp XXXIV (25th September 1896). With my Tibetan journey of 1896 Rockhill did not come into contact, or more correctly I did not intersect his route in Tibet. The points at which we approached each other nearest are my Camp XXXVIII at Mössuto and his camp of 28th May at Umeké-ula; the distance between the two is only a short day's march along the valley of the Naichi-gol (Nadschi-muren). The Marco Polo range forms the eastward continuation of the main crest of the Arka-tagh. The Bokalik-tagh is certainly only a parallel range north of the Arka-tagh. The lofty snowy mass which Rockhill saw bearing west-north-west from his camp of 3 I st May can hardly be the Schapka Monomakha, for this lies 170 miles from that camp. Rockhill's belief, that the Koko-schili comes to an end towards the west, admits of easy explanation: the range is very irregular and in some places very low, especially in that locality. Farther west it swells up again to a very big mountain-chain, for I had it south of my itinerary for two months in 1896. The vast latitudinal valley which I then followed, and which is broken up into a great number of self-contained basins, is continued eastwards by the open country which Rockhill crossed over diagonally during the last days that I have cited, the country which is traversed from west to east by the river Tschumar or Namtschutu-ulan muren. Rockhill's route was intersected immediately south of his camp of I st June by Wellby and Malcolm. When I proceed to deal with the principal results of Wellby's journey, we shall have to speak again of the river Tschumar.
From the Tschumar and for a long distance to the south-west the only traveller who has journeyed there is Rockhill, and what he has to say about it is therefore of the greatest importance. His statements supplement also the observations which I made in the region immediately west of it. In his Diary for the 4th June Rockhill says: -
DA very gentle ascent of seven miles, the latter half over grass-covered hummocks, brought us to the foot of the Koko-shili-éken-k'utul, or »Upper-koko-schilipass». A good-sized brook flows down from the pass, the hills rising not over eight hundred or nine hundred feet above the level of the Elesu-nor. The point at which we are crossing the Koko-schilis appears to be very near the western extre-
Hedin, Tourney in Central Asia. IV. 5S
IIIF Color HighRes
