National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

> > > >
Color New!IIIF Color HighRes Gray HighRes PDF   Japanese English
0205 Serindia : vol.2
Serindia : vol.2 / Page 205 (Color Image)

New!Citation Information

doi: 10.20676/00000183
Citation Format: Chicago | APA | Harvard | IEEE

OCR Text

 

Sec. i] THE EXTENSION OF THE `GREAT WALL' BEYOND TUN-HUANG 727

If the Limes really had for some years its temporary head at this place, the marshy ground to the south and along the right bank of the Su-lo Ho near its bend would have provided an effective Ranking defence. I may add that the route surveyed in 1914, which passes from Shih-êrh-tun to Su-chou, keeping north of the rugged hill chain represented in Maps Nos. 86, 88, is quite as practicable and about as long as the present high road from Yii-mên-hsien to Su-chou."

The other point which may come into consideration here as a likely position for the Jade Gate of 103 B.c. is the defile between Bulungir (also Bulunjir, Pu-lung-chi) and An-hsi (Map No. 83. B. 2), where the Su-lo Ho passes the foot of the low barren hill-chain of Wan-shan-tzû on its left bank and an outlying ridge on its right.lza It is the only defile along the whole Su-lo Ho course after the river's debouchure from the mountains, and it is particularly easy to watch and defend. It was, no doubt, for this very reason that the point was selected by the military engineers of the Emperor Wu-ti for bringing the line of the Limes across the Su-lo Ho. On the right, or northern, bank I succeeded in 1914 in tracing the line of the Limes wall, badly decayed as it is here, close to the foot of the outlying ridge above mentioned, which still bears a watch-tower of Han times. On the steep slopes of the opposite southern bank, where the line of the Limes must have had its continuation, I found no surviving traces of the ancient wall. But a series of watch-towers of later date, crowning the successive small spurs where the present high road from Bulungir to An-hsi passes this much-broken ground (see Map No. 83. B. 2), proves that the defile was specially guarded long after the ancient Limes was abandoned. About i 2 miles further on, to the west of the small oasis of Hsiao-wan, the remains of the Han wall and towers could be followed again quite clearly over the bare gravel glacis running in the direction of An-hsi.

1 consider that the defile just described offered a position eminently suitable for a main frontier station guarding the highway from the Tarim Basin until the westward extension of the Limes to the terminal basin of the Su-lo Ho was accomplished. In support of this belief I may point to two local observations. Only about i 2 miles to the east of the Wan-shan-tzû defile, and thus within striking distance, we find the large but now almost wholly abandoned circumvallation of Bulungir, which during Manchu times, even after the Chinese reconquest of Eastern Turkestan, was occupied by a considerable frontier garrison.18 Then again, immediately below the point where the road coming from Bulungir first ascends the Wan-shan-tzü spurs, and close to the left riverbank, there is a group of ruined Chinese temples, known as Lao- chün-miao, all wrecked since the great Tungan rebellion, but still objects of local worship. Their position, far away from inhabited ground, points to the early sanctity of the site, and, in view of what I had occasion to explain above about the special tenacity with which local worship clings to all places where routes passed outside the ancient line of the Limes,18a the surmise readily suggests itself that the site originally derived its sanctity from the vicinity of an ancient ' Gate'.

Su-lo Ho defile at Wan-shanIzü.

Likely posi-
tion of Yii-
m€n, r o3 B.C.

17 Whether the name Yil-mên-hsien borne by the small town and district headquarters south of Shih-êrh-tun (Map 85. A. 3) is of great age, and whether it is based on some kind of local tradition or mere antiquarian conjecture, are questions into which I am unable to inquire at present. I did not hear of any old remains within or close to the oasis. It is, however, well placed for cultivation, canals from the debouchure of the Su-lo Ho affording irrigation. Hence the occupation of the locality by a military colony may have recommended itself from an early date. Local antiquarian opinion, as communicated to me by Chiang Ssû-yeh, seemed to identify Yii-mên-hsien

with the old Chin-ch'ang   for the latter cf. Chavannes,
Dix Inscriptions, p. 67, note 2, and below, chap. xxvt. sec. ii.

17a Concerning this defile and its tactical importance, see below, chap. xxvii. sec. v.

18 Bulungir must have been a place of importance to have given its name (apparently of Mongol origin) to the Su-lo Ho, which often figures as the' Bulungir River' in European maps.

The small fortified town of Ch'iao-wan-ch'ing (Map 83. D. 2), situated on the right bank of the river about ro miles to the east-north-east of Bulungir and just within the ancient Limes wall, must have served a similar purpose as a frontier garrison. Its solid walls and structures, said to have been built by the Emperor Chien-lung, were reduced to ruins during the last great Tungan rebellion.

18a Cf. above, pp. 602, 696 ; also below, chap. xxvi. sec. ii.