Sec. vi] A RECORD IN RUNIC TURKISH SCRIPT
In the first place, I think, importance must be claimed for the entry which mentions a number of yarigs arrived from the town of Sugchu '.8 This town, as Professor Thomsen has duly recognized, is undoubtedly the present Su-chou, the well-known and important town of Kan-su within the westernmost bend of the later Great Wall. Marco Polo's Succiu preserves, just as Sug'chu does, the older pronunciation of the first syllable su, which was suk.7 Now the Chinese historical records, as shown by M. Chavannes' extracts and analysis, make it perfectly clear that from about A.D. 756-8 the Tibetans gradually overran the whole of Kan-su, and that after A. D. 766, when they had finally established themselves there, they completely stopped all intercourse between China and those portions of Eastern Turkestan where Chinese garrisons were still holding out against Tibetan invasion.8 From that date onwards no Turkish soldiers are likely to have come from Su-chou to Lop.
As to the tribal or national affinity of these men we receive valuable evidence from the statement immediately following in the list that of those yarigs from Sugchu ` there were [given] six 9'arigs to the Bayirqus '. The Bayirqus, as pointed out in Professor Thomsen's note, ' were a Turkish tribe nearly related to the Uigurs and living north of the great desert '. The extracts concerning the Uigurs or Hui-ho, which M. Chavannes has translated from the Tan; shu and fully annotated, in fact enumerate the Bayirqus or Pa yeh-ku among the different tribes composing the Uigur nation.9 They are also mentioned in certain of the ` Runic' Turkish inscriptions from Orkhon which Professor Thomsen first deciphered.10 Evidently the men in question were detached to a party of Uigurs.
Now from the interesting Chinese records which relate the events immediately preceding the final downfall of the Tang dominion in Eastern Turkestan, and which M. Chavannes has rendered accessible to research," we see clearly that throughout the struggles by which the Chinese political officers and commandants of the garrisons north and south of the Tien-shan, though cut off from the empire since about A. D. 766, maintained themselves for close on twenty-five years longer against the invading Tibetans, the Uigurs fought with them as allies. Their tribal settlements then reached close to Pei-t`ing,. in the vicinity of the present Guchen and north of the Turfan depression. In this connexion it should be noticed here that among the entries in sheet b of the Miran record there is one concerning a yarlig given to a certain Ktiräbir Urungu Sangun for going to the town of Qochu. This corresponds, as Professor Thomsen points out, to Kao-ch`ang, the capital of Turfan in Tang times, and now represented by the ruins of Kara-khbja.'Z
It is to the period immediately preceding that isolation of the Chinese garrisons in the protectorates of An-hsi (Kucha) and Pei-t`ing that I should be inclined to assign the probable date of our Old-Turkish record from Miran. We know that the Tibetans, who had temporarily made themselves masters of the Tarim Basin between A.D. 67o-92, continued from about A. D. 717 onwards to threaten the ' Four Garrisons ' by repeated aggressions from the south.ts Throughout
See the text of a, 1. 14, J.R.A.S., 19x2, pp. x86, x88.
Cf. Yule, Marco Polo", i. pp. 217 sq. As Sir Henry Yule points out, the name was still recorded as Suk-c11û by Rashid-ud-din and Shah Rukh's ambassador. For an account of Su-chou see below, chap. xxvii. sec. iii.
a Cf. M. Chavannes' remarks, Ancien! Kholan, i. pp. 534 sqq., especially p. 534, note 2.
9 Cf. Chavannes, Turcs occid., pp. 88 sq.
10 See Thomsen, Inscriptions de POrkhon d&chiffrées,
p. 109 (I take this reference from Chavannes, loc. cil.)
l' Cf. M. Chavannes' extracts, Ancien! Kholan, i. pp.
12 Cf. J.R.A.S., 1912, p. 187, 1. r i of text b, recto. Li the translation the mention of the town of Qochu has been omitted by an oversight ; but see note, ibid., p. 189.
13 For a summary account of this period in the history of Eastern Turkestan, Ancient Kholan, i. pp. 61 sqq. may conveniently be consulted. There full references have been given to M. Chavannes' extracts from the Tang Annals and other Chinese texts upon which our knowledge of those events is mainly based.
Mention of Sugchu, i. e. Su-chou.
Tribe of Bayirqus.
Uigurs as allies of Chinese
Probable date of Old Turkish record.