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

shepherds from Yangi-hissār and other small oases by the high road west of Korla regularly graze
their flocks along the Inchike and Charchak river-beds. .

Desert From our next camp at Mukhtār-chöl (Map No. 45. c. 3) we subsequently pushed a net of sys-
north of tematic reconnaissances into the desert north of the present jungle belt of the Charchak-daryā in
Charchak-
daryā. order to test the equally positive assertions about an extensive old town which two other Korla
hunters, who had joined later, declared that they had seen in that direction some seven years earlier.
Here two old river-beds were met with, lined by the usual rows of Toghraks and tamarisk-cones
mostly dead ; beyond extended a bare clay steppe where the drift-sand lay only in small dunes and
where the view was quite open and ranged far towards the foot of the mountains. In the end I was
able to establish with certainty that those elaborate reports, too, had no more substantial foundation
than the existence of a rude log-built hut almost buried between tamarisk-cones, evidently an old
'Öghil' of herdsmen, and the occurrence of a few fragments of coarse pottery and slag found on
ground slightly eroded some 10 miles north of Mukhtār-chöl. There was no indication of great age
about any of these remains, while, on the other hand, I found ocular evidence that this ground was
being still regularly visited by men in search of fuel from Charchi, a small oasis on the high road.
Close examination of the physical conditions convinced me that no permanent settlement on any
scale could have existed on this ground within historical times.

What still continued to puzzle me was how to account for the psychology and conduct displayed
by my would-be guides. Their assurance about what they believed that they had seen appeared to
be quite genuine, and they all stuck stoutly to their story, undisturbed by the fact that the location
of their respective 'old towns' avowedly differed. There was no obvious reason for conscious
imposition, since no reward for their guidance had been offered or asked for in advance. At last the
clue to the mystery revealed itself in a folk-lore belief ascertained after careful sounding. All the
Korla folk were under the spell of the old tradition, which elsewhere too haunts the outskirts of the
great desert, about 'old towns' buried by the sands and full of hidden treasure. With this tradition
is firmly coupled a belief that the demons guarding such ruins prevent those who chance upon them
from ever seeing them a second time. Mūsā Hājī and his fellow-hunters had manifestly tried their
luck in searching for the ruins of their local kötek-shahri at different times and in different areas of
the desert. In offering their guidance they had acted quite bona fide, prompted by the hope that
my supposed 'Wilāyati' magic would overcome the evil genii hiding the walled town of which their
own imaginative eyes had before caught merely a glimpse—always during, or immediately before,
a dust-storm ! That I, too, failed to secure them a chance of discovering all those hidden treasures
was a source of genuine disappointment for them. The tradition accounting for all this is obviously
but a survival of old popular legends like that which Hsüan-tsang had heard at Pi-mo, west of
Khotan, and the localized form of which he records in his story about the sand-buried town of
Ho-lao-lo-chia.¹¹

Section II.—FROM THE INCHIKE RIVER TO KUCHĀ

Ruins of Meanwhile the efforts of a sober-minded Korla Yüz-bāshi had secured a shepherd guide who knew
Gumbaz. the exact position of the previously reported 'Gumbaz' near the Inchike river. For this I set out
on January 9, and, after crossing again the dune belt south of the Charchak bed, reached the ruins on
the following day. They proved to be situated amidst dunes with plentiful remains of kötek, or dead
forest, at a point about 2 miles north of the Inchike river (Map No. 45. A. 4). It soon became clear