856 ACROSS THE PAMIRS [Chap. XXV
most hamlets of Röshan were all excellent cragsmen and quite experts in building rafaks or ledges of brushwood and stones along otherwise impassable precipices. It was fully five hours before a tolerably safe track had been made higher up and we had crossed the worst of those treacherous scarps (Fig. 375) ; yet the direct distance was scarcely more than a mile. At last we reached the head of the inlet, lined with half-submerged thickets of birch-trees and juniper. Ascending the valley amidst fine groves of trees for a couple of miles and then crossing what looked like an old terminal moraine, we arrived at a widening stretch where cultivation had been resumed, since the earthquake, by six Röshani families. Their smiling fields of barley and oats lay at an elevation of about II ,000 feet and some 500 feet above the level of the lake as it stood then. Yet even here dread was felt of the continued rise of its waters.
During a day's halt at this pleasant spot our hillmen succeeded in improving the track above the Yerkh inlet sufficiently to bring, somehow or other, their sure-footed ponies across. Accordingly on August Igth we moved up the valley to the south, which contains at its bottom a succession of small lakes formed by glacier action between old moraines. Small hanging glaciers showed at the heads of the side valleys on either side. The route had never been surveyed and had come into use only since that across the Marjanai pass between Saréz and the Alichur Pamir had been blocked by the newly formed lake. As our route led continuously over old moraines and boulder-strewn fans, progress was troublesome. But fortunately on arrival at Ushinch, about II miles farther up, where the valley bottom widens in view of an amphitheatre of ice-crowned peaks to the south, we were met by fresh Kirghiz transport kindly sent by the Commandant of Pamirski Post. This opportune help made it possible to push up the valley, which now turned to SE. and widened into a Pamir-like expanse. After passing three more small lakes we camped at an elevation of about 14,40o feet.
Next morning, ascending first to SE. and then turning east, we reached after a march of 5 miles the Langar pass, forming an almost level talus-covered saddle at about 15,400 feet. A large hanging glacier to the NW. of the pass sends its drainage partly to the small lake of Emin-köl, which we had come to before reaching the pass, and partly to the Langar-köl on the other side. The descent into the Langar valley was easy and brought us mostly over gentle grassy slopes to the stone huts known as Langar. There we camped at an elevation of about 12,30o feet, after a total march of 20 miles.
SECTION IV.—BY THE ALICHUR AND GREAT PAMIR
Route along Our route from Langar turned eastwards to the Yeshil-köl and Alichur Pamir, and as these
Yeshil-köl. as well as the Great Pamir to the south have been often visited and described,1 the account of my rapid passage may be brief. Some points of special interest must, however, be noticed. On ascending from Langar the easy spur which separates the mouth of that valley from the western end of the Yeshil-köl, an excellent view offered over the head of the Ghund valley leading down through Shughnan. Looking across its wide floor and the grass-covered easy slopes which flank it, we were able fully to appreciate the advantages offered by it for direct communication from the Pamirs westwards to the Oxus. It is true that the modern Russian cart-road which leads from Pamirski Post down to Shughnan leaves the Alichur Pamir above the Yeshil-köl and does not enter the Ghund valley until more than thirty miles below the exit of the Ghund river from the lake. But the route which from the Alichur Pamir keeps to the Yeshil-köl and then enters the Ghund
1 For useful summaries of information concerning the Schultz, Forschungen im Pamir, pp. 61 sqq. ; on the Great
Alichur Pamir, cf. e. g. Geiger, Pamir-Gebiete, pp. 131 sq. ; Pamir, Geiger, ibid., pp. 128 sqq. ; Schultz, ibid., pp. 72 sqq.