western Kuruk-tagh is traceable to Roborovskij's mistake as to the real significance of the Chara-teken-ula, and in the light of Prschevalskij's information this must be regarded as a step in the wrong direction. The fact of both ranges appearing on Grum-Grschimajlo's map is attributable to the fact that, although his journey took place a year before Roborovskij's, his book did not appear until four years after the publication of the accounts of the Tibetan Expedition.
Returning to the passages quoted from Roborovskij, his statement, that the actual ranges rest upon a flat swelling, is in accord with the descriptions of other travellers and with my own experience. It is very interesting to learn of the existence of an isolated sandy desert, Ak-bel-kum, on the southern shore of the lake, and to read the description of its gigantic dunes. South of this desert the Chara-teken-ula (i. e. the Kuruk-tagh) was only faintly visible in places by reason of the great distance, which is greater than • it appears from Roborovskij's map. On the other hand, to judge from Kosloff's and Grum-Grschimajlo's accounts, Roborovskij's idea, that this range unites with the Tschöl-tagh five days south-east from Toksun, is absolutely incorrect. Both of them lay stress upon the general parallelism which prevails in this mountainous region.
My own views with regard to the structure of the Kuruk-tagh system are embodied in the accompanying little sketch-map. On each side of the big Juldus valley, the valley of the Chajdik-gol, there are two imposing main ranges of the Tienschan, lying parallel and close to one another. They widen out however in one place sufficiently to allow of the existence of an important cauldron-valley, in the midst of which lies the Baghrasch-köl. The northern chain is continued eastwards under the name of Tschöl-tagh and the southern under the name of the Kuruk-tagh, the divergence between them gradually increasing as they proceed. The former of these two ranges marks the northern edge, and the latter the southern edge, of an east-west swelling, which, although not very high, is nevertheless distinctly marked and definitively bounded. On the north this swelling is bordered by several depressions, e. g. Hami and Luktschin, the latter lying t 30 m. below the level of the sea. On the south it is bordered by the Desert of Lop and the valley of the Bulundsirgol. On the crown of the swelling between the two border-ranges there are a number of smaller parallel ranges, crests, ridges, and chains of hills.
Farther east, where the great highway from Hami to An-si-tschou and Sutschou crosses the Gobi, several travellers have traversed the eastward continuation of the ranges which we are now discussing, e. g. Sosnovskij and Piasetskij, Mandl, Carey, Prschevalskij, Grum-Grschimajlo, Roborovskij, Obrutscheff, Futterer and Holderer, and others. Roborovskij's route was farthest to the west, so that my own excursion, which I shall describe lower down, comes nearest to his route. Prschevalskij's third journey (1879-8o) was taken between Hami and Sa-tschou, Carey's journey between Hami and An-si-tschou. Futterer's line coincides in part with Grum-Grschimajlo's and Obrutscheff's. His account of these regions is the best, but before considering it, I must first give a few brief extracts from the descriptions of Prschevalskij and Roborovskij.
Prschevalskij's journey was begun on the 1st June (0. S.) 1879, and consequently fell in the most unfavourable season of the year; in fact, he often travelled
IIIF Color HighRes
