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

interior bare to the centre and showed it to be constructed throughout of bricks, 15″×8″×4″, the same size as noted elsewhere at the site. A low, flat mound about 25 yards to the south of the Stūpa showed no traces of structural remains, but may mark a position once occupied by less solidly built dwellings.

Remains of
defensive
position. Less than half a mile to the west of this ruined Stūpa, and about the same distance from the main ruins, the top of a very steep and almost completely isolated rocky spur (v in plan, Plate 54) shows clear signs of having once been occupied as a place of safety. The crest of the spur has been artificially enlarged here by terrace walls, partly in uncut stone and partly in brickwork, to a level space measuring about 105 feet from north to south and about 66 feet across. It is thickly covered with potsherds of ancient appearance, but bears no structural remains. A small bastion-like projection to the north seems intended to defend the only side from which the top of the spur can be approached without difficult climbing. On all other sides it is cut off from neighbouring ridges by very steep rocky ravines. A small knoll on the slope to the north, on a level about 60 feet lower down, is occupied by a badly decayed mound which may be the remains of a Stūpa. It evidently was taken for one by those who long ago made a cutting reaching to its centre. One side of what looked like a square base measured about 33 feet. The cutting showed the mound to be constructed partly of sun-dried bricks and partly of uncut stones. Its present height is about 15 feet above the natural rock of the knoll top.

Former
agricultural
settlement. No coins or other datable relics were found at the site. But judging from the structural and other remains, the conclusion seems justified that its shrines continued to receive worship as long as those of 'Ming-oi', and that they suffered destruction about the same time as the latter. It is not likely that a monastic establishment could have existed here without some larger agricultural settlement in its vicinity. For this the clayey steppe to the east and north might have afforded ample room, provided that in Buddhist times the water-supply from the mountains was more abundant and regular than it now is. Desiccation thus seems to have played its part here too. We left the picturesque site with regret, after our stay over Christmas Day had sufficed for clearing its ruins and for R. B. Lal Singh's survey work on the watershed of the range to the south; for in spite of the greater elevation and the vicinity of the snow-capped mountains there was warmth in the air and clear sunshine, doubly welcome after the Sarmatian winter days we had spent in the Kara-shahr plain.

Skirting the foot of the range where scrub and jungle-tree growth were abundant at the mouth of all ravines, we reached the snow-covered plain at Shīkchin by the evening of December 26. On the following day a long march over interesting ground brought us to the Korla oasis. After

Ancient
watch-
tower south
of Shīkchin. crossing for close on seven miles the low ridges and terraces into which the range coming from Khōra breaks up at its easternmost foot, the track led past a massive ruined watch-tower placed on a high and commanding terrace. It is the one to which I have had occasion, on p. 1199, to refer in connexion with the ancient tower above the caves of 'Ming-oi'. It is built very solidly, about 32 feet square at the base, and still rises to a height of about 19 feet. It is constructed throughout of well-made sun-dried bricks measuring 12″×8″×3·4″. The decay it has suffered in spite of its massive construction would alone attest its antiquity. But what attracted my attention at once were the regular layers of reeds inserted between the brickwork. At places where big fissures had formed in the latter I could clearly see that, while in the interior masonry the reed-layer appeared after every single course of bricks, a plainly distinguishable outer casing of brickwork, about 4 feet thick, showed it regularly after every three courses. This latter arrangement is usually found in the brick-built watch-towers of the Tun-huang Limes.

The similarity to these which the ruin showed in position, general appearance, and that