Hsüan- tsang's crossing of Su-lo Ho.
Local data in Hsüantsang's L.
Topographical data verified.
1098 THE MARCHES OF OLD KUA-CHOU [Chap. XXVI
In the third.watch they came to the river, and sighted the Yü-mên barrier from a distance. At Jo li above the point where the barrier stood, the river banks were not much more than a chang (10 feet) apart.' 13 Here a crossing was effected by a rough footway which the young barbarian improvised with cut tree-branches, etc. After resting by the water's edge and having discharged the young ` barbarian ', who was unwilling to venture further, the pilgrim bravely set out alone on the track leading to the first watch-tower. We shall follow him on this venture in a subsequent chapter, and show that there is good reason for believing that his track could not•have lain far away from the present caravan route connecting An-hsi with Hami.13$
From this short account and the local information previously given to Hsüan-tsang we gather the following topographical data. Starting from the town of Kua-chou, the route towards Hâmi first led north for 5o li to the river Hu-lu, where the watch-station of Yü-mên kuan was placed. From this point the route towards Hâmi turned to the north-west and passed into the desert. Hsüan-tsang, having to avoid the watch-station where his unauthorized move beyond the border would have been stopped, set out from Kua-chou at night and reached the river in the third watch at a point some 10 li above the Yü-mên barrier, and there managed to effect a crossing unperceived. Thence he picked up the track leading to the nearest of the watch-towers, at which, as we are told in the Life, he arrived after covering 8o li."
It is easy to see that these indications are in full accord with the topographical facts as the map shows them. By the river Hu-lu no other but the Su-lo Ho can be meant.16 From the ruined town of Kua-chou-ch'êng, which in view of its central position and surviving local tradition may safely be assumed to mark the approximate site of the Kua-chou of Tang times,16 it is 8 miles almost exactly due north in a straight line to the point where the present road to Hâmi crosses the Su-lo Ho. If we assume that its course lay in Hsüan-tsang's days about two miles further to the north, where Rai Ram Singh's reconnaissance survey marks an old river-bed,17 the agreement in distance with the 5o li of the Life becomes still closer ; for the equation of 5 li to the mile is the one which my experience of Hsüan-tsang's distance reckonings on Central-Asian ground has proved to be the generally correct average.18 That the road to I-Hâmi, after leaving the Su-lo Ho, leads steadily in a north-west direction is a fact which the map clearly demonstrates. Finally it is noteworthy that the 8o li which Hsüan-tsang is said to have covered from the river crossing to the first watch-tower agree remarkably well with the 16 miles or so which Map No. 81 shows between
tion in favour of the substantial veracity of the account as handed down by his biographers. We shall see that the story attributes a considerable share in the pilgrim's ultimate escape from dying of thirst and exhaustion in the desert to the local sense of his mount. Of this local instinct of horses accustomed to desert travel I have had ample personal experience, and I am therefore inclined to believe that the specific reference made to this much-travelled animal is based on a matter-of-fact feature of Hsüan-tsang's adventure.
At the same time, the way in which the Life connects the acquisition of this useful mount, of russet colour and lean', with a prognostic received long before shows the same intermingling of sense of reality and nave credulity which characterizes Hstian-tsang's personality throughout his own Mémoires.
is I reproduce here the passage according to the interpretation which Dr. L. Giles has indicated to me as the correct one. Julien's version would imply that this point where the cussing was made was at the barrier itself. But this obviously
cannot be the sense intended, since the passage had to be effected secretly.
lsa See below, pp. 1 143 sqq.
" Cf. Julien, Vie, p. 24.
m The identification with the Su-lo Ho, the Bulungir of the Mongols, had already been made correctly by V. de Saint-Martin ; cf. Julien, M. moires, ii. p. 262.
16 In the same way the Tun-huang district has retained its administrative centre in the immediate vicinity of the Shachou of T'ang times. See above, p. 6o8 ; also p. 1091.
17 See Map No. 81. D. z. I regret that delay in starting from An-hsi prevented my examining this old bed when I passed it on Oct. 8, 1907. I regret still more that I did not make a reference to the account of the Life during my busy stay at An-hsi.
7s Cf. above, p. 649, n. 12 ; p. 716, n. ro. It must be remembered also that Hsüan-tsang's reckonings relate to actual marching distances, not to direct distances measured on the map.