________________
204
THE INDIAN ANTIQUARY
Interesting and fruitful as this search was, I felt a strong longing for a chance of resuming exploratory tasks in the open air of the desert. But my leg had not yet recovered from the accident in the summer, and could not face long tramps such as a return to the wastes of the Lop Desert would have necessitated. So I had to be content with what satisfaction Lal Singh's safe return towards the close of January from his expedition into the Dry Mountains' brought me. In the face of great physical difficulties and risks he had accomplished important survey work. After reaching Singer, the only permanent homestead in that vast area of barren plateaus and hills, he had started triangulation; and in accordance with my instructions carried it south-east to the vicinity of the Lou-lan ruins in the wind-eroded desert. His patient wait there for a week, amidst icy gales and with temperatures falling well below zero Fahrenheit, was rewarded when the dust-laden atmosphere cleared at last and allowed him to connect his triangles with previously fixed' high peaks of the snowy K'un-lun range from 150 miles south. Thus it became possible later on to realize my hope of getting the Indian triangulation system extended by this link to the Tien-shan range in the north.
[SEPTEMBER, 1917
With Abdur Rahim, the experienced hunter of wild camels from Singer, whose help had proved so valuable to us a year before, Lal Singh had then pushed into the unexplored and absolutely sterile region to the north-east of Altmish-bulak. His fuel supply had given out for several days, and he had to brave the severe cold of the nights without a fire before he decided to turn again westwards from beyond 91° long. He then picked up an old desert track once used by hunters of wild camels from Hami, before certain salt springs had dried up, and followed it down to the salt marsh that forms the deepest part of the Turfan Basin. There he took numerous observations with the mercurial barometer which, I hope, will make it possible to determine its depression below sea-level with greater accuracy. In spite of all he had gone through, Lal Singh allowed himself but a brief rest at our base, and by the first week of February set out afresh for the Kuruk-tagh.
The packing of our plentiful "archæological proceeds" from Turfan had cost great efforts. But at last, on February 6, I could start my big convoy of antiques, making up fifty camel-loads, under Ibrahim Beg's care for its two months' journey to Kashgar. On the same day, I sent off Afrazgul Khan to the Lop Desert for a supplementary survey of the Lou-lan region and the dried-up ancient sea-bed to the east and south. I myself proceeded to Yar-khoto for a detailed survey of this curious site, where a maze of ruin dwellings and shrines, carved out of the loess soil of an isolated and naturally strong plateau, represents the remains of the earlier Turfan capital during Han times. Some days more were taken up by arrangements for the completion of the large-scale map of the Turfan depression in six sheets and by the collection of supplementary data bearing on its extant irrigation resources. Their comparison with those which must be assumed to have existed in Buddhist times is made particularly instructive by the fact that now the greater portion of the culti vated area is irrigated from karezes or springs tapped by underground canals, a system which is known to have been introduced into Turfan only during the eighteenth century. My last days at Turfan were made somewhat anxious by a renewed attempt at Chinese obstruction, now directed against my archæological activity. Fortunately this time, too, I was on the point of starting into desert parts where no interference with my plans was practicable, and the safe transit of my antique collection, about which I had reason to feel apprehensive, was secured by the prompt steps my old friend, Sir Ceorge Macartney, took to parry the stroke from Provincial headquarters.
(To be continued.)