Sec. iii] FROM YARKHUN HEAD-WATERS TO TAGH-DUMBASH PAMIR 51
and enforced a hurried farewell to Captain Stirling, who had insisted on sharing our toils as far as this.
Fortunately the descent to the east proved less trying ; for the huge névé-beds at the head of the glacier below afforded an easy slope, and when after some three miles' progress in soft snow the glacier proper was reached, we found most of its crevasses sufficiently covered with fresh snow. Farther down we could follow the line of the main lateral moraine on the north, and after a descent of five hours a dry spot was found under its shelter where it was possible to bivouac in safety. The bitter cold of that night at an elevation of over 15,o00 feet made me anxious about some of our Wakhi porters whose loads had caused them to lag behind ; but they too turned up safely in the morning, having wisely kept moving all through the night. It was a source of great satisfaction to me that my feet, in spite of loss of toes and impaired circulation, results of the frost-bite accident at the close of my second journey, were still equal to a thirteen hours' struggle over snow and ice at great elevations.
The snout of the glacier (Fig. 61) was not passed until after another four miles' steep descent in the morning. On arrival in the valley below at the high grazing ground of Buattar I was glad to find fresh transport from Hunza awaiting us. Once again the arrangements made weeks ahead through Humayûn Beg, the Wazir of Hunza, my old acquaintance of 1900, had not failed to assure rapid progress. After two miles' easy descent from Buattar, the path struck the side of the mighty ice-stream of Khûz, which comes down from the high range on the south and completely fills the bottom of the main valley as it trends eastwards. Passing the narrow side valley to the north-west, up which lies the route to the Irshad pass previously mentioned, we skirted the Khûz glacier for a distance of more than four miles before its snout was reached in the now widening valley bottom (Fig. 6o) at an elevation of about 12,000 feet. Only a little beyond, on the alluvial fan of Yarz-yarz, the first signs of former occupation were met with in the shape of abandoned fields and decayed huts. They were to prove a characteristic feature of the Chapursan valley at many points farther down. Some four miles beyond we halted at Baba-ghundi, marked by a much-frequented Ziarat, said to be the most famous throughout Hunza (Fig. 58). Here, too, at an elevation of 11,600 feet, there is no permanent occupation ; but there are some oat-fields, which, we were told, had been brought under cultivation again during the lifetime of the father of the present Mir of Hunza.
A long but easy march on September 4 down to Spandrinj allowed me to see the greater portion of the Chapursan valley and realize its peculiar character. In spite of the great height of the ranges to the north and south, rising to peaks of over 22,000 feet and as yet but very imperfectly explored, the bottom of this valley is more open and unobstructed than any other part of Hunza. Its gentle slope is sufficiently indicated by the fact that Spandrinj lies at an elevation of only 1,600 feet or so less than Baba-ghundi, though at a distance of about twenty-five miles from it by road. The Chapursan valley, notwithstanding the great moraines that glaciers on the south had once pushed down into it, contains a larger area of fairly level and easily irrigated ground than probably all the rest of Hunza put together. It was therefore of distinct geographical interest to note the extensive areas of abandoned cultivation that we passed between Baba-ghundi and Spandrinj. Neither want of water for irrigation nor present climatic conditions seem to furnish any adequate explanation for their abandonment. It is true that below the outflow of the big Ishkuk glacier there is ground where cultivation has, according to local tradition, been destroyed by glacier detritus brought down through a change in the flood beds. But this factor is absent elsewhere, and does not account by itself for the virtual abandonment of a valley that, with occupation such as is found on corresponding ground in Wakhan, might support a number of comparatively