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

Sec. v]   . THE HAN LIMES FROM YÜ-MÊN-HSIEN TO AN-HSI   1139
Bulungir was certainly well placed for watching the several routes through the Pei-shan by which a nomadic enemy could make his way from Hâmi to the Su-lo Ho and the Kan-su marches. It

could also conveniently guard the narrow defile through which the river and the high road along it pass some twelve miles west of Bulungir (Map No. 83. B. 2).

The low but steep and rugged spur of Wan-shan-tzû,s a north-eastern continuation of the hill Defile of range we have passed north of Ch`iao-tzû, there juts out towards the ` thalweg ' of the river, and Wan-shan-

tzû.

is closely approached on the opposite side by a small and almost isolated ridge representing a last offshoot from the southernmost Pei-shan range. The river passes between the two in a winding defile scarcely half a mile wide at its bottom. As the end of the Wan-shan-tzû spur falls off precipitously to the river, which actually washes ' its foot, the road must necessarily ascend the terminal ridge, leading over it at a height of about 200 feet above the river bed. It is this point, particularly easy to watch and defend, that the engineers of the ancient Limes, with their unfailing eye for the military advantages of the ground, had chosen for bringing their border-line from the right to the left bank. The tower shown in the map as crowning the top of the ridge facing Wanshan-tzû from the north proved on its examination in 1914 undoubtedly to date from Han times. The clearing of the débris at its foot then yielded one or two Chinese documents on wood of that period. The wall itself, here built mainly of rough stones, was clearly traceable along the southern and south-eastern foot of the same ridge. On the low ground further west, where it must have run down to the river bank, its traces had, of course, disappeared.

On the southern bank of the river and along the north-west end of the Wan. shan-tzû spur I was Towers unable to discover any clearly recognizable remains of the Limes line. But that end of the spur and shrines

bears two large watch-towers close to the road, and beside the western one remains of   small at Wan-

bears   four   shan-tzû.
structures of brick, which appear to have been shrines. I may add that by the left bank of the river, where it is first approached on the east by the foot of the Wan-shan-tzû spur, there are the ruins of a well-built temple with several Stupas, all destroyed during the Tungan rebellion and known as Lao-chün-miao. I have had before occasion to suggest that the defile at the foot of the Wan-shan-tzû spur might well have served for a ` Gate ' station of the Han Limes before its line had been extended to Tun-huang and beyond.9 If that had ever been the case, we might, perhaps, in view of what has been previously explained, recognize in these shrines a lingering trace of the local worship which appears to cling invariably to such ` Gate ' sites.

(MapOf the wall which must have continued the Limes line on the left bank of the river beyond Limes wall Wan-shan-tzü I failed to trace any remains until after passing the village area of Hsiao-wan traced Ma No. 83. A. 2). The low-lying ground here between the river and the foot of the hill range beyond 3•   )•   Y• g g   g Hsiao-wan.
south is covered with scrub and jungle, where not actually cultivated, and on ground of this

kind remains of the ancient a gxrer could scarcely be expected to survive. Some five miles below the main village of Hsiao-wan, where the road emerges from the area of jungle and abandoned cultivation to more open ground, I came upon the ruins of a walled enclosure, about 208 yards square, known as Po-ch`ha-tzai, ` the old town '. Its walls still rise to 10-12 feet in height and do not bear a very ancient look. Crossing beyond it the canal which carries water to the southernmost portion of Kua-chou cultivation, I was now approaching the stretch of bare ground where on my way to Ch`iao-tzû three months earlier I had sighted two towers of the Limes and the line of wall connecting them.'Ô About a mile to the west of P`o-ch`êng-tzû I found a much-decayed mound of clay (marked by a on map), about 12 feet high and io feet in diameter, which, as it falls exactly into line

8 This appears to be the correct form of the name as

recorded by Chiang-Ssii-yeh. Wan-shan-!si   WT.,the
winding hill [range]', seems a very appropriate designation,

looking at the curving shape of the spur shown by the map. ° Cf. above, pp. 727, 1099, note 20. 10 See above, p. 1099.

7 F 2