SECTION I.—THE GEOGRAPHICAL POSITION OF TURFAN AND THE
EARLIEST HISTORICAL NOTICES
MY arrival, on October 25th, 1914, close to the town of Turfân marked the beginning of the season that I proposed to devote to archaeological and geographical labours in the Turfân basin. The extent and variety of the interests presented by this region both to the antiquarian and geographical student are so great that the time, just over three months, which I was able to spend there would certainly have appeared altogether inadequate if a systematic survey of all its ancient remains and its physical features had been called for. In neither direction was there room or need for so wide a programme.
As regards the ancient remains of the Turfân basin, Dr. Klementz's reconnaissance visit, carried out in 1897 under the auspices of the Russian Academy of Sciences, first showed not only how great was the number of pre-Muhammadan ruins but also how easy they were of access ; since then the search among them for antiques had proceeded for a considerable number of years and on an extensive scale. Between the years 1902 and 1907 fully equipped parties of German archaeologists under the leadership of such distinguished scholars as Professors A. Grünwedel and Von Lecoq had been at work at Turfân sites for periods nearly ten times as long in the aggregate as that which I was able to devote to this area. Their extensive operations, rewarded by rich results, and also the facilities for the ready sale of antiques offered by the vicinity of Urumchi and by the position on a great trade route, had stimulated the ruthless exploitation of the ruins by the local population. Much destruction had been caused for a long time past through the constant digging for manuring earth among the ruins of the ancient capitals, Yâr-khoto and Kara-khôja, and this process was being accelerated by the increasing extent of cultivation around those two important sites. Mr. Tachibana, the Japanese traveller, too, had spent several months for archaeological purposes in the Turfân district in 1910 and again in 191 I. Hence it was clear to me from the start that it would be necessary, as an essential preliminary to useful work, to ascertain by reconnaissances which sites were still comparatively undisturbed or contained remains, such as wall-paintings, that deserved to be saved from further risks by careful removal.
The reconnaissances were also to familiarize me with the typical features of the Turfân depression, and thus to enable me to direct and check more closely that detailed survey of its topography which the geographical interest attaching to it made me anxious to secure. This survey, on the comparatively large scale of one inch to the mile, was to be carried out by Surveyor Muhammad Yâqûb ; he had been carrying a plane-table traverse across the waterless. desert between the terminal basin of the Hâmi drainage and the eastern border of the Turfân district, and I expected him shortly to rejoin me. The experience I had gained in the course of previous travel of the young surveyor's qualifications made me wish to keep within easy reach of him in order to exercise supervision and control.
R. B. Lâl Singh, with his much greater experience and energy, could be trusted to make the fullest use of such opportunities for fresh surveys on unexplored or as yet very imperfectly