National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

from a place on the Su-lo Ho which must have been in the vicinity of the present An-hsi and prob-
ably not far from the point where the high road of our times passes the river. Considering that this
road, except for a small detour between Ta-ch'tian and Sha-ch'tian-tzŭ due to necessities of water-
supply, leads in what practically is a straight line from An-hsi, or Kua-chou, to the Hāmi oasis,
it seems safe to assume that the ancient route which the pilgrim intended to follow—the story, as
recorded in the Life, clearly shows that after the first four marches he strayed from it—could not
have lain far away from the present route line.

But before we attempt to follow the traces of the pious traveller on his adventurous journey, Topo-
which threatened nearly to end with his dying of thirst in the desert, it will be well to indicate graphical
briefly certain main topographical facts concerning the ground traversed by the present high road. features on
As the Maps Nos. 80, 81 show, the first five marches from An-hsi lead across a succession of narrow An-hsi-
hill ranges, all striking approximately east to west and rising but little above the wide plateau-like Hāmi route.
valleys between them. The halting-places offering water are all situated close to the foot of these
ranges. The subsoil drainage comes to light in springs at the first three stages (Pei-tan-tzŭ, Hung-
liu-yüan, Ta-ch'tian) and is easily reached by wells not more than 6–8 feet deep at Ma-lien-ching-tzŭ
and Hsing-hsing-hsia. The water is fresh at all these stages, and some scanty grazing obtainable.
It is probably not without reason that the boundary between the provinces of Kan-su and Hsin-chiang
is fixed now close to Hsing-hsing-hsia ;⁹ for beyond the character of the ground changes, distinctly
for the worse. Much bare rocky ledge and detritus is passed on the next two marches to Sha-ch'tian-
tzŭ and K'u-shui, there being a steady descent of some two thousand feet from the average level of
the preceding stages. Vegetation becomes increasingly scanty and the water decidedly brackish,
as the name of K'u-shui, ¹ Bitter Water', rightly indicates.

But it is the next march, that to Yen-tun (Map No. 76. A. 4), which is most dreaded of all Trying
by Chinese wayfarers. For a distance of some 35 miles it leads down over absolutely bare gravel marches
slopes into a depression lying at its bottom some 1,500 feet below the level of K'u-shui. Totally beyond
devoid of water or shelter of any sort, this long march is attended with risks on account of the great K'u-shui.
summer heat here experienced and the icy north-east gales to which it is exposed in the winter and
spring. This great depression or trough of Yen-tun extends far away to the east, as our surveys of
1914 proved, and serves apparently as a main conductor for the bitterly cold winds which sweep
from Southern Mongolia across the eastern Kuruk-tāgh and down to the Lop basin. Carcases of
transport animals marked the route all the way from K'u-shui ; nor are losses in human lives unknown
here. From Yen-tun another march, over similar gravel wastes but much shorter, brings the
traveller to the springs of Chang-liu-shui (Map No. 73. D. 2) at the southern edge of a wide belt of loess
ground which receives subsoil water from the snows of the Karlik-tāgh and is covered with abundant
scrub and reed-beds. At Chang-liu-shui the first tiny patch of Hāmi cultivation is met, and after
two more easy marches the town of Hāmi, or Kumul, is reached in the central oasis.

With these topographical features of the route from An-hsi to Hāmi the essential points in the Record of
story of Hsüan-tsang's desert journey, as related in the Life, can be shown to be in close agreement. Hsüan-
This agreement is all the more remarkable because we are unable to control here the statements of tsang's
the Life by the text of the Hsi-yü-chi, since Hsüan-tsang's own account of his travels does not begin journey.
until after his departure from Kao-ch'ang or Turfān. An obvious lacuna in the story of the Life,
which we shall have occasion to notice presently, can cause no surprise, considering what we know