634 AT RUINED SITES OF TURFAN [Chap. XVIII
Frescoes set Repacked like the rest of my collection with systematic care at Kâshgar, the contents of all
up lh at New these cases reached their Indian destination without loss or damage. The setting up of the Bezeklik
frescoes in the building erected for their accommodation at New Delhi has since 1921 taken up most
of the time and labour which Mr. Andrews has been able to devote to the arrangement of the antiquities collected during my third expedition. But at the time of writing the work is still far from being completed.3a The reproduction and full description of these wall paintings must therefore be reserved for a separate future publication. Meanwhile I have to content myself with embodying in the present record a plan of the site (Pl. 30) which, based on a regular survey, supplements the sketch furnished by Professor Grünwedel,4 as well as some photographs illustrating features of
its remains (Figs. 314, 315, 318, 319).
Reasons While this work of fresco removal was proceeding, I was obliged to absent myself from Murtuk
for visit to between December 14, 1914, and January 7, 1915, mainly for the purpose of a visit to Urumchi, Urumclù.
the provincial capital. I was drawn there by a great desire to see again Pan Ta-jên, the kind
friend and patron of my first two journeys. His administrative merits and his rare and widely recognized reputation for perfect clean-handedness had, after his retirement during the troubles of the ` revolutionary ' period, brought him well-deserved promotion to the post of Financial Commissioner of the Province. But even without this personal motive I should have felt obliged to undertake this journey for quasi-diplomatic reasons. Notwithstanding the helpful intercession from Peking secured in the spring by the British Minister, I had reason to apprehend that the spirit prompting the official obstruction, which in January had seriously threatened to bring both my archaeological and geographical work to a standstill, had by no means disappeared from provincial head-quarters. As the sequel showed, this apprehension was only too justified. In order to guard against this risk, or at least to delay the resumption of obstructive tactics, it seemed clearly advisable to endeavour by a personal visit to secure a more favourable attitude of those in power at Urumchi, and in any case to assure myself of that friendly support of Pan Ta-jên which had proved so helpful in the course of my first two journeys. I might, moreover, hope to obtain advice as regards my injured leg from the medical officer attached to the Russian Consulate at Urumchi ; though slowly improving, its condition still continued to be a cause of anxiety and impediment.
Climatic The journey along the high road from Turfân town to Urumchi and back, together with a
conditions week's stay at the latter, occupied my time between December 18th and January 3rd. The rapidity on 'Men-
shan water- of the marches by which the distance of some 115 miles between the two places had to be covered,
shed. the difficulty I still experienced in walking, and the wintry season, owing to which the country
traversed from near the Tien-shan watershed northward to Urumchi was snow-covered, all combined to prevent survey work. But even so the journey furnished a variety of interesting observations. It showed me the utter barrenness of the slopes of gravel and decayed rock over which the ascent is made from the Turfân depression to the watershed ; the remarkably low elevation of the latter, which on the plateau of Ta-fan-ch`êng scarcely rises above 3,000 feet ; and the ease of communication secured by this route between the Turfân basin and Dzungaria. At the Chinese village of Ta-fan-ch`êng close to the watershed irrigation is still required for the fields, and a lively stream coming from the high Bogdo-ula mountains to the north-east serves this purpose. But on the low hills over which Urumchi is approached cultivation depends on rain and snowfall only. This marked change in climatic conditions made its effects strikingly felt when we reached Ta-fan-
3s Since done. to individual shrines (in arabic figures) besides the ` site-
4 For the sake of facilitating reference to Professor marks ' (in small roman figures) indicating those from which
Grünwedel's detailed account of the site it has appeared fresco panels were removed by us.
advisable to insert in the plan the numbers given by him