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. iii]   SHAN-SHAN IN THE LATER HAN ANNALS   333

placing of a garrison at ` Lou-lan is not specifically recorded. But the report contained in Pan Yung's biography makes it clear how important a part Shan-shan, to which Lou-lan belonged, had played as the advanced base at the beginning of his successful campaign. It is worth notice that

the operations against Turfan and Guchen presuppose the use of the routes which lead from the Lop region due north across the Western Kuruk-tagh, as Yen-ch`i or Kara-shahr still barred the use of the main road from the Tarim Basin to Turfan and did not submit to Chinese power until A.D. 127.23 Owing to advanced desiccation all these routes across the Kuruk-tagh [as our surveys of 1914-15 have proved] would now offer very serious obstacles to the advance of any large force from the south, and some would be wholly impracticable.

From the period corresponding to A. D. 132—I 34 onwards the Later Han Annals record a gradual decay of Imperial prestige in ' the Western kingdoms', which ' oppressed and attacked each other in turn'. The notices we receive of them for the remaining century of the dynasty's reign are distinctly meagre and do not contain any further references to Shan-shan.24 But in view of the evidence already given for the epoch immediately succeeding, there is no reason to doubt that Chinese control was maintained in one form or another at this outer gate of the Empire's Marches.

SECTION IV.—EARLIEST RECORDS OF LOU-LAN UNDER THE FORMER HAN.

It now only remains for us to trace back the history of the Lop region to the oldest of our extant records, mainly furnished by the ` Notes on the Western Regions ' of the Former Han Annals (Chapter xcvi).' The account which these give of the kingdom of Shan-shan is exceptionally ample, a proof by itself of the importance with which geographical and historical factors had invested that territory during the initial period of Chinese expansion westwards. ` The

original name of the kingdom of Shan-shan was Lou-lan   1. The capital is the city of Yü-ni
(Wylie : Woo-ne), which is distant from the Yang barrier 1,600 li and from Ch`ang-an (i.e. Hsi-an-fu) 6, loo li. The kingdom contains 1,57o families, comprising a population of 14, Ioo, with 2,912 trained troops, a Guardian Marquis, a Chio-hu Marquis, a Protector-General of Shan-shan, a Protector-General for repelling the Chii-shih, a Right Chii-ch`ii, a Left Ciü-ch`ii, a Prince for repelling the Chii-shih, and two interpreters-in-chief. The seat of government of the Chinese Governor-General lies to the

north-west 1,785 li. The kingdom of Shan w gi is distant 1,365 li ; and Chü-shih   OI j lies to
the north-west 1,890 li.'

In these opening passages of the account it is of special interest to note that the capital of the

kingdom is placed at the city of Vu-ni   Yii-ni, as we have already shown in discussing the
data contained in Li Tao-yüan's commentary on the Shui thing, lay not far from the southern shores of the Lop-nor marshes, and its position is probably marked by the early remains of the Miran Site.2 The reference which the commentator makes to Yii-ni as the place ` ordinarily called the old eastern town' renders this identification certain. With it the various reckonings of distance to other localities fall into easy agreement. The estimate of 1,600 li from the Yang barrier appears very reasonable for the distance of about 310 miles which the map shows between Miran and the oasis of Nan-hu, to the south-west of Tun-huang, where the gate-station of Yang kuan must probably be located.3 The reckoning from the Yang barrier suggests that the route meant was the one which

Turkestan at close of Han period.

Former

Han Annals' notice of Shan-shan.

Position of town of Yii-ni.

23 Cf. Chavannes, T`oung pao, 1906, p. 254.

" See T`oung-pao, 1907, pp. i67 sq.

' See Wylie's translation in J. Anthrop. Inst., x. pp. 23 sqq. The need of a new critical translation, with names transcribed according to a recognized system and also reproduced in the Chinese characters, must be badly felt by

every non-Sinologist student who wishes to use this important historical and geographical record. In the extracts given below, Mr. Wylie's transcription has been adapted to the Wade system with the kind assistance of Mr. A. D. Waley.

See above, pp. 326 sq.

' Cf. below, chap. xvt.,sec. iv.