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

while I was near Korla, I had learned from a letter of Raī Rām Singh, written after he had reached
Khotan via Charkhlik, that the search set on foot under my instructions of the preceding spring and
carried out by some of my old 'treasure-seeking' guides on the Khotan side had resulted in the
tracing of several unexplored sites in the desert below Keriya and Khotan. A letter from my old
friend and factotum Badruddin Khān, the Aksakāl of Indian and Afghān traders at Khotan, from
a trader just arrived delivered to me at Kuchā, confirmed this information. The details it gave of
the great number of 'old houses', i.e. ruined structures, traced at several of these sites, and about
the position of the latter, furnished ground for believing that I should need all the time that could
possibly be secured in order to explore them before the heat and the sand-storms of the spring would
make work on that ground impracticable.

Decision
to cross
Taklama-
kān. This consideration compelled a move south as early as possible, and as Kara-dong near the
terminal course of the Keriya River was reported as a site where more ruins had come to light
since my visit of 1901, I decided to strike due south from Kuchā to where the Keriya River dies
away in the sand, and thus to save time by a 'short cut'. I knew well that this desert crossing was
beset with serious difficulties and risks also. But Dr. Hedin's pioneer journey of 1896 showed that
it was practicable, if prepared and carried through with due care and precautions. The saving in
time which this plan promised to effect was a great attraction—and so, too, to me personally,
I confess, the chance of once more crossing the very heart of the desert. For the safe transport to
Khotan of my heavy convoy of antiques, making up twenty-four camel-loads, there offered itself
conveniently the well-known trade route leading up the dry bed of the Khotan River, and on this
it was accordingly started under the care of Chiang Ssŭ-yeh.

Visits to
Kuchā
sites. The considerations here briefly indicated will explain why during my week's stay at Kuchā
I could make no attempt to supplement in any way the protracted labours which had been devoted
to the plentiful ancient remains of this great and important oasis during the preceding five years
by Japanese, Russian, German, and French archaeological expeditions. I managed, however, to
visit practically all the chief ruined sites within a day's ride from Kuchā town, the interesting cave-
temples near Kum-tura, Kizil-kāghe, and Kirish, and the large temple ruins of Duldur-ākhur and
Su-bāshi (Map No. 34. B—D. 1, 2; Fig. 292). To notice at the last two sites the systematic
thoroughness and care with which their final clearing had been effected by the French Mission
under Professor P. Pelliot was a source of special satisfaction. Useful information was collected
also from local Afghān traders as to the necessarily rather different operations by which so many
important acquisitions of ancient manuscripts from Kuchā, including the famous 'Bower Manuscript',
now mainly in the British collection formed under Dr. Hoernle's care, had been secured since 1890.

Importance
of Kuchā
in anti-
quity. Kuchā has at all periods been one of the most important territories in the Tārim Basin, in many
ways a worthy pendant of Khotan owing to its geographical position and the rôle it has played in
Buddhist art and civilization. It is hence particularly fortunate that Professor Sylvain Lévi, in the
same masterly paper in which he has proved the identity of the remarkable Indo-European language
previously designated as 'Tokhāri B' with the tongue once spoken at Kuchā, has also furnished us
with a lucid and exhaustive analysis of all historical data about Kuchā found in the Chinese Annals
and other sources.¹ This makes it easier for me to leave whatever observations I might have to
offer on the historical topography of Kuchā, and on the part which geographical conditions have
played in determining the importance of this great and flourishing oasis in ancient times, for the
report on my third journey. Then I was able to devote a number of weeks to a close survey both
of the actually cultivated area and of that which, as is proved by the evidence of numerous ancient
sites found scattered in the scrubby desert to the east, south, and west, must once have formed part