Shan-shan identified with Lou-lan by Chien Han shu.
Importance of route north of Lop-nôr.
344 HISTORICAL NOTICES OF LOP, SHAN-SHAN, AND LOU-LAN [Chap. IX
the territory situated to the south of Lop-nor and the terminal Tarim the new official designation of Shan-shan took root.2"
On the other hand, the transfer of the capital southwards which M. Chavannes and after him Dr. Herrmann have suggested involves a difficulty with regard to the political change in connexion with which it is supposed to have taken place. There can be no doubt that the deeds recorded in 77 B.C. were planned to strengthen the Chinese hold over the Lop region. Now all geographical and historical considerations point to the fact that the part of this district which was of chief strategic importance to the Chinese during Han times was the tract north of the Lop-nor marshes and along the foot of the Kuruk-tagh, through which the most direct route between Tun-huang and the northern oases of the Tarim Basin passed. Had the seat of the Lou-lan chief stood here before 77 B. C., it would obviously have been an advantage for Chinese policy to let it remain there ; for it would have been far easier to provide the support and control which the new king needed on that great military and trade road than away to the south of Lop-nor where the military colony of I-hsün was actually established. And the whole story of China's relations with Central Asia shows that economy of effort has, here as elsewhere, always been a characteristic feature of its strategy and statecraft.
Nor does it seem to me safe to brush aside lightly the explicit statement of the Ch`ien Han shu which asserts that the original name of the kingdom of Shan-shan was Lou-lan '. We have seen above that all Chinese records, from Later Han down to Tang times, uniformly persist in correctly showing Shan-shan in the position of Lop, and in placing its chief settlement to the south of Lop-nor. We have also seen that this position of the ` capital ' of the territory was dictated by physical conditions which have continued to the present time in spite of all changes. In view of this consistency of later historical evidence during prolonged periods, the testimony of the Former Han Annals, which fully conforms with it, has an additional claim to credence. Moreover, the reliance which it deserves in this case is considerably strengthened by the fact that the events related of 77 B.C., with the change of Lou-lan into Shan-shan, are separated by only about a century and a half from the time when the extant text of the Clz`ien Han shu was composed by Pan Ku
A. D. 92).27
Summary of It will here be convenient briefly to summarize the main results of geographical and anti-
ancient Lop quarian interest which our examination of the earliest Chinese records concerning the Lop region
topography. has yielded. In the first place, it is certain that the name Lou-Ian, subsequently changed into Shan-shan, corresponded to the mediaeval and modern Lop in its widest sense, and was applied to a territory which comprised the whole of the depression between Kuruk-tagh and Altin-tagh, with the terminal courses of the Tarim, Charchan, and Konche Rivers and the Lop-nor marshes fed by them. The population of the territory depended mainly on pastoral pursuits. The chief agricultural settlements were confined to the present Charkhlik tract, where streams coming from the snowy range south assured permanent chances of irrigation. In it was situated, certainly from 77 B. c. onwards, but probably earlier also, the political centre of the territory, the capital Yii-ni
46 It would not be difficult to quote numerous instances in which Chinese nomenclature, old or modern, has utilized the local name of an earlier period for the express purpose of distinguishing the part of a territory, though there was no doubt that the original application of the term had been different. Thus when the present Kenya District (hsien) was separated in the eighties of the last century from the Khotan Prefecture (chou), it received the official designation of Yi1-
t `ien T. , though it is perfectly well known to all educated
Chinese in Eastern Turkestan that this is the ancient designation of Khotan, now officially called Ho-t'ien *1 Ng.
s7 This argument would, of course, carry even more weight if it is assumed with Dr. Herrmann that the ' Notes on the Western Regions ' embody in the main an official compilation dating from circ. 3o B.C. ; cf. Seidenstrassen, pp. 35 sq.