they pass around the capital, irrigate the fields and gardens and work mills'; for this description exactly applies to the stream which issues from the gorge of Sengim and by means of canals taking off from its branching beds irrigates the whole of the oasis of Kara-kheja (Map No. 28. c. 3).
The Kao-ch`ang people's love of music, on which Wang Yen-tê lays stress, their habit of never undertaking promenades or excursions without taking with them musical instruments (pp. 57 sq.), are well illustrated by archaeological finds ; see below, Chap. xix. His references to details of the Chinese calendar, to seasonal sacrifices, to great collections of Chinese texts, Buddhist and others, to an archive of imperial edicts, &c. (pp. 57 sqq.), all go to prove how deep and lasting were the effects upon the Turfan people of prolonged contact with Chinese rule and civilization. The general prosperity there prevailing under Uigur rule is indicated by the mention of the practice of feeding such poor as there were at the public expense, and of the great age generally attained by the inhabitants.
It is quite certain that Wang Yen-tê, when proceeding to the northern court ' of Arslàn Kagan, travelled by the route leading across the pass above Pa-no-p`a. But it is difficult to fix exactly the stages he mentions. Travelling obviously in the leisurely fashion befitting an imperial envoy, he took six days to cross the district of Chiao-ho (YAr-khoto) and to reach ` the entrance of the passage through the Chin-ling mountains ', a locality by which perhaps the halting-place of Shaftulluk, the ` Dragon spring ', may be meant (see above, p. 563).
The next two marches brought him to Han chia chai, ` the camp of the Chinese ', which possibly corresponds to Yoghan-terek. [In the Sung History the character is not
chia, ` family', but §g chung, ` mound '.—Dr. L. Giles.] Five days more were occupied in crossing the mountain range, a not unreasonable allowance for the passage of so great a dignitary, if reckoned to a point at the northern foot of the Tien-shan whence he could reach Pei-t`ing in one day (p. 62).
As Wang Yen-tê arrived at Kao-ch`ang in the fourth month (May) of the Chinese year corresponding to 982, and by the
seventh month (August) of the same year (pp. 65 sq.) was invited at Pei-t`ing to prepare for the return journey to China, his passage north must have been effected in June or July. Yet he found heavy snow on the range and was assailed by torrents of rain and snow while crossing it. Where the ` hall of the dragon ' (Lung-t`ang), situated on the pass and apparently a cave, is to be looked for I am unable to say. An inscribed stone there recorded the name of the pass as Hsiao hsiieh-span, ` the little snowy mountain '.
Of the ` valley of Pei-t`ing' the narrative states that it was several thousand li in length and breadth. This shows that the Uigur dominion at the time must have extended for a great distance along the northern slopes of the Tien-shan (cf. Grenard, J. As., 190o, janv.-févr., p. 29). The lake near the capital, on which the envoy was entertained by the king to a musical festival, must be the marshy lake that the Russian Trans-frontier Map marks to the north-west of Guchen with the name of Ulan-nôr. I now regret not to have visited it. The abundance of horses in the region of the old Besh-balik of Uigur times is brought out by Wang Yen-Ws mention that a piece of silk three yards long was the customary price for an inferior horse such as sold for meat.
7 See Julien, J. As., 1847, ix. pp. 56, 64. The statement, p. 56, that the frontiers to the south-west touched the country of the Ta-shih or Arabs and Po-ssû or Persians is clearly an exaggeration.
8 Julien, loc. cit., p. 56, wrongly indicates a doubt as to the five inches of rain which, Wang Yen-tê states, had fallen in the year 970 and had destroyed a multitude of huts and houses. It is certain that even a lesser quantity of rain, falling within a short period, would nowadays work excessive havoc among the mud-brick dwellings of the Turfan oases.
9 See Julien, ibid., p. 57 ; also p. 64 relating to the
Uigurs of Pei-t`ing. 10 Cf. ibid., pp. 58 sq.