408 THE LOU-LAN SITE [Chap. XI
that we have special reason to feel grateful for the number of exactly dated documents, which enable
us to fix with sufficient accuracy the lower chronological limit of the occupation of the site. Among the records analysed by M. Chavannes there are not less than fifteen dated from years corresponding to A. D. 263-70, and all of them found at L.B. ii or in the refuse-heaps adjoining this ` Ya-mên'. Only two, Doc. Nos. 738 and 721, belong to the years A.D. 263 and 264 respectively, bearing `nien-haos' of the closing reign of the Wei dynasty, and the rest to the years A.D. 265-70, which are comprised in the initial regnal period, t`ai-shih, of the Emperor Wu-ti, the founder of the Western Chin dynasty.4 Of him we know from the Chin slzu that he reasserted Chinese supremacy in the ` Western Countries ' after the period of internal disruption in the empire known as the ' Epoch of the Three Kingdoms' (A.D. 221-65).5 As the dates on the documents found by Dr. Hedin, according to Herr Himly's preliminary notice, appear also to be confined to the years A.D. 264-7o,6 we may safely conclude that this was a period when the desert route and the station guarding its western end saw exceptional traffic and activity.
Throughout the reign of the Emperor Wu-ti (A.D. 265-89) the continued political relations
with the Western Countries are attested by notices of the Chin Annals, which, between A.D. 270 and 287, repeatedly mention embassies with tribute from Central-Asian territories even as distant as Ta-yuan (Farghâna) and Kang-chü (Samarkand). Thus in A. D. 283 we learn that the young son of a Shan-shan or Lop chief was sent to the imperial court for service.' After his reign these
*relations are declared to have completely ceased.' But that Chinese occupation at the Lou-lan Site did not come to an end with that reign is proved by two of our documents, Nos. 910 and 886, which record dates corresponding to A. D. 312 and 330. The last named is a slip found in L.A. r, recording a payment of money to a certain Inc or barbarian.
It is interesting to note that the very way in which this latest date is recorded conveys a clear
indication of the final abandonment of the site then being near. The year is stated as the eighteenth of the ' nien-hao ' Chien-hsing, which commenced in A. D. 313. But as this ended in A.D. 316 together with the reign of the last emperor of the Chin dynasty whose sway extended over Northern China, it is clear, as M. Chavannes has rightly recognized,' that the little station must by that time have been completely cut off from official intercourse with the central authorities of the empire and
abandoned to its own resources. Only thus is it possible to explain that, in its isolation, the post continued using the obsolete ' nien-hao', the abrogation of which could not be notified to it, for fully sixteen years longer.'"
Yet we should not be justified in assuming that this condition of administrative isolation was merely local. That it must also have affected other Chinese garrisons surviving in the Tdrim Basin is rendered highly probable by a number of significant observations. The ruined site could not have been a mere outlying post ; it must have lain on an important line of communication for the
refuse-heap adjoining a single ' Ya-mên' to furnish four fragmentary documents which directly emanate from, or are addressed to, the ` Chang-shih of the Western Countries ', i. e. the chief representative of Chinese authority in the Tarim Basin. No. 752 communicates an order received and forwarded by the Chance shih for an officer to start by a specified date and proceed to certain
4 See Chavannes, Doc. Nos. 722-27, 730, 739 (A.n. 265-
69); Nos. 733, 735, 736, 748, 896 (all A.D. 270).
° Cf. M. Chavannes' note on China's political relations with Central Asia under the Emperor Wu-ti, Ancient Khotan,
i• P. 537.
° Cf. Herr Himly's note, quoted from Petermann's Miltheilungen, 1902, pp. 288-90, in Hedin, Central Asia
and Tibet, ii. p. 144.
I See M. Chavannes' note in Ancient Khotan, i. p. 537.
Cf. M. Chavannes' remark, evidently based on the Chin shu, loc. cil.
9 See Chavannes, Documents, pp. 182 sq.
30 For similar instances of the use 2f obsolete `nien-haos' due to similar circumstances, cf. Ancient Khglan, i. pp. 275, 533.