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

on the ceiling of the passage at the back. Two small caves adjoining to the south had apparently
served as monks' quarters.

Large
ruined
enclosure. On the opposite side of the gorge, the top of a very steep hillock is occupied by the badly decayed
remains of what appears to have been a large enclosure containing the chief monastic quarters.
Owing to the precipitous nature of the slopes, a great portion of the walls once enclosing an area
about 50 yards by 40 have slid down. No structural features could be made out in the interior,
which was completely overlaid by rubbish-heaps and pottery debris. There were indications of
burrowing by 'treasure-seekers' along the line of the walls and elsewhere, and systematic clearing,
which would have taken days, offered small promise of reward. I turned my attention instead to
two groups of small caves, IV, facing each other on a steep little spur to the south of this area. The
farther one was found to comprise two little shrines communicating by passages. Complete clearing
revealed only traces of painted panels, with small seated Buddha figures. The four caves of the
other group were undoubtedly used as quarters, a larger one being joined to a smaller one by a
passage, as the plan of T.A. IV. i (Pl. 43) shows. On removing the loose earth which filled these cave-
dwellings, we found in the one marked a (Pl. 43) two dozen Chinese coins, partly embedded in the
flooring. Of these, twenty-one are T'ang issues and three are uninscribed, probably of earlier date.
Thus the conclusion, already suggested by the style of the wall-paintings, that the site had continued
to be occupied during the T'ang period received definite confirmation before the end of our visit.

Return to
Kuchā. The necessity of making arrangements for Lāl Singh's survey work obliged me to start back
for Kuchā on April 28th, in the midst of a violent sand-storm. From Shōr-yailak I dispatched
Afrāz-gul to the south-east with orders to survey the ground between Yulduz-bāgh and Khanak-
atam, the southernmost settlement of Kuchā to the east of the Muz-art river (Map No. 17. c. 2).
He reached Khanak-atam, our rendezvous, in three marches ; his observations on the ruins that
he passed will be briefly noticed later. I may, however, mention here with advantage what I
subsequently ascertained, with his help, with regard to the alleged 'Tatis' beyond the western-
most limit of the present cultivated area on the west of the Muz-art river.

Antiques
collected
from
Dawān-
kum. From Aziz Palwān, our guide to Tonguz-bāsh and the minor sites described above, I obtained
a number of small ancient objects, mostly of metal, but some of glass and stone, which, as a profes-
sional antique-hunter, he had picked up on his visits to old sites. They are enumerated, together
with those bought from Mīr Sharīf, in the Descriptive List below. Azīz stated that he had collected
most of these small objects from an area of ancient occupation which he called Dawān-kum,
and which, from his description, was obviously a wind-eroded 'Tati'. The place was known to
others by that name, and was situated some distance beyond the westernmost cultivation of Yulduz-
bāgh, on the direct desert route to Ak-su. I accordingly arranged, at the close of my stay at Kuchā,
for Afrāz-gul to pay a special visit to this area when surveying this route as far as Kara-yulghun
(Map Nos. 17. A. 1, 2 ; 12. B, C, D. 2).

Search for
'Tatis' of
Dawān-
kum. Afrāz-gul's route report shows that when he left Torpak-bāzār for this purpose on May 10th
he had to content himself with a 'guide' who soon proved to be very imperfectly acquainted with
the track to be followed ; Aziz Palwān had refused to accompany him. At Lampe, some eight
miles from Torpak-bāzār, the last cultivated patches were left behind, and two miles farther on he
reached wind-eroded ground where, for a distance of about three-quarters of a mile, abundant pottery
debris indicated ancient occupation. The guide called this place Hajelik, and applied the same
name to a similar area about a mile and a half beyond it. A further march of about four miles
brought the little party to the tract known to the guide as Dawān-kum. No water was obtainable
at that camp ; nevertheless Afrāz-gul spent a day there, making an extensive search for the alleged
site, as shown by his devious route line marked on the map (No. 17. A. 2). The ground proved to