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

about 45 feet; it is occupied on its top by a mosque and the habitation of a faqīr.
The mound, which is thickly covered with potsherds, consists wholly of debris
accumulations from houses built with sun-dried bricks. The remains of walls
from such dwellings are clearly traceable at different points in the deep ravines
which erosion has cut into the slopes of the mound on the east and south. At
the foot of one wall thus disclosed a row of burnt bricks, measuring 15 × 10 × 2
inches, formed a foundation. In a hollow just below this there was found a terra-
cotta mould (B.A. 46), apparently meant to represent an enthroned figure.

Two small stones carved in relief were brought to us by the faqīr as having
been similarly found where rain had laid them bare on the mound. One of them
(B.A. 1; Pl. II) shows a well-modelled small figure floating in the air and
evidently meant for a Gandharvī, as often represented in Graeco-Buddhist
sculpture. The fragment appears to have adjoined the halo of a larger relief.
The other stone is a slab (B.A. 047; Pl. II) about 6 inches square, divided into
nine squares. Within these are shown in relief two birds and two open lotuses,
while each of the four corners is decorated with a four-petalled clematis-like
flower, a motif very common in Gandhāra reliefs. The subjects and style of
these carvings leave no doubt about the mound having been occupied down to
late Buddhist or Hindu times. This conclusion was supported by the evidence
afforded by the relief-decorated pottery (B.A. 14, 15, 23, 25, 28, 33; Pl. II) on
which were frequently to be noted patterns formed by rosettes within pearl
borders and concentric circles and concentric oblongs. But there appears also
reason to believe that occupation of the site started from an earlier period; for
a considerable proportion of the painted pottery picked up (B.A. 3, 4, 6, 7, 12,
16, 20; Pl. II), with varied patterns in black over red and pink bands, showed a
distinct resemblance to the painted ware which I had found on the mounds of
Dera Ismail Khān surveyed in 1927 along the foot of the Takht-i-Sulaimān
Range and in Northern Balūchistān, and which may safely be ascribed to late
prehistoric or early historic times.⁹ There were picked up also fragments of
shell bangles and a piece of glazed ware (B.A. 34) which may be ascribed to
the eighth to tenth century A.D.

At a distance of about 240 yards to the east there is the still larger mound,
B, extending for more than 700 yards in length and about 300 yards across
where widest. Owing to repeated digging for manuring earth the height of
the mound has been reduced to a maximum of 25 to 27 feet in the west, while
elsewhere it is much less. The excavations have in many places laid bare walls
of dwellings built with sun-dried bricks. The plentiful relief-decorated and
painted potsherds showed close agreement in types with those found on mound