332 HISTORICAL NOTICES OF LOP, SHAN-SHAN, AND LOU-LAN [Chap. IX
of five hundred men, to organize a military colony at Lou-lan. On the West this officer will dominate the roads which lead to Yen-ch`i (Kara-shahr) and Ch`iu-tzti (Kuchâ). On the South he will fortify the courage of Shan-shan and Yü-tien (Khotan). On the North he will overawe the Hsiung-nu (Huns). On the East he will be a neighbour to Tun-huang. It is in this that there lies real advantage.'" The same record, in reproducing Pan Yung's answer to a question raised in council, makes him indicate still more clearly the purpose of the proposed military colony at Lou-lan. It was primarily intended to secure the Chinese hold over Shan-shah and the indispensable route leading through it. ` Now the king of Shan-shan, Yu-huan )G it, is a descendant of Chinese on the female side. If the Hsiung-nu carry out their projects, Yu-huan will certainly die. Now, though these people be like to birds and to wild beasts, they know how to avoid what is harmful. If we move out troops to establish a military colony at Lou-lan, this will suffice to secure us their hearts.' 18
In the following chapter I hope to show by archaeological proofs that the position of this
military colony at Lou-lan, then projected and afterwards actually established, is represented by the ruins of the fortified settlement north of Lop-nor which Dr. Hedin first discovered, and which I was able fully to survey and explore." There, too, will be the place to examine the reasons which may explain why the earlier designation of Lou-lan, originally borne by the whole of Shan-shan or Lop, was retained or revived in Chinese official use for this particular locality.
In any case it is easy to see how completely the position of that ruined settlement answered the
strategic objects set forth by Pan Yung for his proposed military colony. Situated on the most direct route from Tun-huang towards the terminal course of the Konche-daryd. or Kara-shahr River, it commanded the roads both to Kara-shahr and Kucha. A Chinese garrison placed there had the chief settlements of Shan-shan, about Mirân and Charkhlik, sufficiently near on the south to assert a reassuring and controlling influence over them. At the same tine it was particularly well situated to ward off any Hun raids which might be directed against the route from the Turfân side on the north and across the western Kuruk-tagh. Finally, as reference to the map shows, its distance from Tun-huang or Sha-chou was shorter than that of any other place capable of permanent occupation on the routes connecting the Tarim Basin with that true base for all Chinese enterprise westwards.
Pan Yung's recommendation as regards the military colony at Lou-lan was not immediately
acted upon. But after Hun inroads had again devastated Kan-su, an offensive move was at last decided upon in A. D. 123, and Pan Yung was appointed ` Chang-shih of the Western Countries' 2° In the first month of the following year (February, A. D. I 24) Pan Yung arrived at Lou-lan and rewarded the king of Shan-shan for his submission by new honours21 The kings of Kuchâ, Ak-su, and Uch-Turfân came to offer their allegiance. Taking the numerous force brought by them, Pan Yung then moved upon Turfân and, after inflicting a defeat on the Hsiunb nu or Huns, established a military colony at Lukchun, in the Turfan depression.22
In the following year Pan Yung, with a large force comprising also troops of Shan-shan, carried his campaign across the Tien-shan and gained a signal victory over the allies of the Huns in the territory of Posterior Chii-shih, corresponding to the present Guchen (Ku-ch`êng-tzû) district. The
17 See Chavannes, Toung pao, 1906, p. 248.
'$ See ibid., p. 249.
'0 See below, chap. xi. sec. i—iv, x.
20 Cf. Chavannes, Toung-pao, 1906, p. 252 ; 1907, p. 167.
a' This passage shows clearly that Lou-lan was included in the territory of Shan-shan. The period of the year chosen for the move to Lou-lan has its significance. It is certain that the difficulties presented by the desert route west of the
Tun-huang Limes, where some stages probably, just as at present, offered only brackish water, could best be faced in the very depth of winter. [My explorations of 1914 have proved that the ancient Chinese route from Tun-huang to the Lou-lan Site for a distance of close on 120 miles crossed desert ground which must have been wholly waterless throughout historical times.]
22 See Chavannes, T`oun pao, 1906, p. 252.