国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

many pieces of silk, quantities of fallen pieces of stucco decoration, painted and

gilded, and the painted fragment reproduced on plate ix.

The next move was to Kan-chou, where an unfortunate riding accident occurred.

The Badakhshi stallion Stein was riding became restive, reared, and fell back,

crushing Stein's leg and causing extensive muscular lesions which gave him very

serious inconvenience and pain for many weeks after. Fortunately, although he

was completely incapacitated for walking or riding for some time, no bones were

broken.

Turning back from Kan-chou, the route lay north-west through Hámi and

Barkul, then west to Guchen, and south to Turfán, which was reached on 25

October 1914.

Turfán, being of considerable historical importance as the scene of repeated con-

flicts between the Chinese and the barbarian Huns (or Hsiung-nu), was carefully

surveyed and studied, and a very full account of the region is given in Innermost Asia.

The aspect that has immediate connexion with the present work is the existence

of Buddhist shrines and their painted walls. The first of the sites examined was

Kara-khója, where, besides great numbers of miscellaneous objects, many small

fragments of painted plaster were gathered, a few of which are shown on plate xi.

After returning from a short tour to certain sites north-east of the Turfán depres-

sion the next to be examined was Toyuk, `most picturesque of all Turfán locali-

ties'. Among the many cave shrines and temples along the gorge that runs from

Toyuk, excavation was started in the debris thrown down from the upper slopes,

covering shrines below; and in this, fragments of manuscripts and miscellaneous

objects were found. Several shrines on an upper terrace were then investigated

with varying results. In one of these was a small room with a vaulted ceiling,

boldly painted; reproduced on plate ix. In a rock-cut shrine a considerable part

remained of the domed ceiling, beautifully painted. With great skill and patience

this was successfully removed in small sections and subsequently reconstructed by

me in New Delhi (plate x).

The next site to be examined was that of Bezeklik, where there is a group of

cave shrines,' previously visited by Professors Grünwedel, who stayed there for

two months, and von Lecoq, who removed numbers of the paintings to Berlin.

Now Stein, observing the extent of the progressive damage the still-remaining

paintings were suffering from local vandalism, decided to remove as many as

I See Sketch Plan, plate c.

,OE

xi