1 144 TO HAMI AND TURFÀN [Chap. XXVIII
about the vicissitudes which the text of Hui-li's biography had undergone before it was published.° This text is manifestly imperfect in places and awaits future critical treatment. But a variety of details and personal touches seem to me to support the impression that Hui-li gathered this graphic account of the desert adventures direct from the Master himself and has reproduced it with faithfulness. We know too much of Hsiian-tsang's pious ardour and naïve credulity to discredit the few references to supernatural incidents : they obviously reflect genuine subjective illusions such as conditions of intense strain and real peril were most likely to produce in a mind so devout and fervid.
The main facts about the journey to be gathered from I-Iui-li's narrative are as follows : Forsaken by the ` young barbarian who was to act as his guide, soon after the start from the Su-lo Ho bank, the pilgrim moves ahead alone, guiding himself along the track by the bones of dead animals and the droppings of horses.? Visions of armed hosts in the distance cause him alarm, only to disappear on closer approach, ` vain images created by demons'. Obviously mirages are meant here, such as I frequently observed on my first few marches beyond An-hsi in spite of the autumnal coolness which had set in by that time in the desert. After covering 8o li, Hsiian-tsang arrives at the first signal-tower. In order to pass it unobserved he hides himself until nightfall. On trying then to replenish his water-bottle from the water near the tower, he is shot at with arrows by the men on guard. On declaring himself a monk come from the capital, he is taken before the commandant of the post. This man, a native of Tun-huang, called Wang-hsiang, receives him kindly and, having failed to persuade him to return, directs him in the morning to proceed to the fourth tower, commanded by a relative of Wang-hsiang. Arriving there the same night he goes through a similar experience. After having been first shot at by the guards, he is taken before the commandant, who on learning of Wang-hsiang's message offers hospitable welcome, but warns him not to approach the fifth and last watch-tower as it is held by men of violent disposition. Instead, he is advised to go to a spring a hundred li off, called Yeh-ma-ch`üan, ' the Spring of the Wild Horses ',8 and to replenish there his water-supply.
` A short distance from there he entered the desert called Mo-ho yen#Ai 0 ,^ which has a length of 800 li and which in ancient times was called Sha-ho, or the " River of Sand ". One sees there neither birds nor quadrupeds nor water nor pasture.' In this desert he loses his way, after having been troubled again by demonic visions, i. e. mirages, and fails to find the ` Spring of the Wild Horses'. To add to his distress he drops the big water-skin he had been given at the fourth tower and loses its contents10 ` Besides, as the route made long detours, he no longer knew which direction to follow. He then meant to turn back to the east, towards the fourth signal tower.' 11 But after having thus proceeded for w li he thinks of his oath not to take his way eastwards again until he had reached India. He thereupon, fervently praying to Kuan-yin, directed himself to the north-west'. All round he sees only limitless plains without a trace of men or horses. Troubled at night by lights lit by wicked spirits and in day-time by terrible sand-storms, he suffers cruel torments from thirst. After having travelled thus for four nights and five days without water, 'he lay down exhausted. In the middle of the fifth night, after fervent prayers to Kuan-yin,
6 Cf. Julien, Vie, Preface, pp. lxxviii sq.
7 See Julien, Vie, pp. 23 sqq. Beal's translation, Life of Hiuen-lsiaug, pp. 17 sqq., appears to be a mere version of Julien's and offers no help, beyond correcting an obvious error in the figure given for the length of the Mo-ho-yen desert.
8 This name Yeh-nza-ch'ilan * itj Vt is still a frequent designation for desert localities beyond the Kan-su border. I heard it applied, e. g., to a spring on the route I followed in 1914 from Mao-mei towards the Karlik-tagh.
9 This is the correct transcript of the name ; cf. Julien, Mémoires, ii. 516 ; Chavannes, Pures occid., p. 74, note 3. Beal, Life of Hiuen-Isiang, p. 21, reproduces Julien's erroneous transcript Mo-kiayen, but corrects Julien's ' quatre-vingts li', one of the great Sinologue's not infrequent lapses in the matter of figures, rightly into 800 li.
'0 By Julien's ' grande outre d'eau ' evidently a water-skin or ' Mussuck ' is meant.
" See Julien, Vie, p. 29.