844 ACROSS THE PAMI RS [Chap. XXV
to the great Pamir-like valley of Môji west of it, which borders the watershed towards the Russian Pamirs on the lakes of the Rang-kul and Great Kara-ku1.3 By descending this valley he connected his survey with the work done in 1900 on my first expedition from the side of Muz-tâgh-atà. Then crossing the Ulûgh-art pass he rejoined me at Bostan-arche, situated in a high valley to the north-east of a conspicuous ice-clad massif of that range (Map No. 2. C. 3).
Dispatch of There, on a small fir-clad alp 4 situated above a favourite Kirghiz camping ground at an
antiques to elevation of over io,000 feet (Figs. 357, 364), I was able to devote ten days of delightful coolness Kashmir.
and seclusion to much hard work on reports, accounts, proposals about the collection of antiques, &c. There, too, the final arrangements were made for the long journey which by the middle of October brought the large convoy of ` archaeological proceeds ', under the personal supervision of R. B. Lai Singh, safely to their temporary place of deposit at Srinagar. He was assisted in this task by Naik Shamsuddin and Surveyor Muhammad Yaqûb, who also accompanied the collection to India. He was thus able to supplement our previous surveys by useful topographical work along the caravan route followed across the Yangi-dawan and by the uppermost Yarkand river to the Indian frontier on the Kara-koram pass.s Of my assistants there remained with me only young Afràz-gul, who, I knew, would always prove useful, even where, as in Russian territory, neither topographical work nor excavations would be possible.
SECTION II.—ALONG THE ALAI VALLEY
On July 19th, with a sense of freedom regained after weeks of clerical toil, I was able to start from Bostan-arche for the high meridional range before us and the Pamirs beyond. It was a great encouragement to feel that across them the road now lay open for me to those mountain regions north of the Oxus which, by reason of their varied geographical interest and their ethnic and historical associations, had attracted me since my youth. The route led up the main valley descending from the Ulûgh-art, and on the following day we crossed this high pass, about 16,600 feet above sea-level. From the narrow saddle, gained after a very steep ascent, the clouds lifting at intervals revealed a grand view across the wide valley of Môji, flanked by the ice-crowned range on which we stood and by the mighty eastern rampart of the Pamirs. Below the pass there we could see the middle and lower reaches of a magnificent glacier about io miles long, descending from a high spur to the south.
The descent to where this glacier turns sharply to the west was extremely precipitous and impracticable for laden animals. No passage is left farther down between the flank of the glacier and the spurs on its north side against the foot of which it presses. So the track, difficult throughout, has to ascend these spurs in succession close to where small glaciers overhang and divide them. Fine ` hanging glaciers ' flank the valley also on the south. By the time we had reached easier ground on plateaus overlooking the snout of the large glacier at the valley bottom, I felt duly impressed with the fact that I had passed the great mountain barrier of ancient Imaos, which divided Ptolemy's ` Inner ' and ` Outer Scythia '. Pushing down past a small Kirghiz camp at Sara to the main valley below Môji, I arrived the same night, after a walk and ride of 33 miles, at Kun-tigmaz. There I met Sir Percy Sykes returning to Kashgar from the Pamirs with his
3 See Map No. 2. A—C. 2.
4 The conifers of Bostan-arche were, apart from wild poplars met lower down at Yolchi-moinak (Map No. z. c. r) and some white poplars at Daraut-kurghân, the last trees of any sort seen until we approached the head of the R6shân
valley near Saunâb. Their presence in this high valley, as also in others farther north of the same range (Map No. 2. B. 2), must obviously be attributed to the influence of the moister climate prevailing on the Tien-shan, here approached by the
meridional range. 5 See Maps Nos. 6, 9, ro.