国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

Sec. iv]   THE ' WHITE DRAGON MOUNDS'

305

Chinese records of route across Salt Marsh.

Li Taoyuan's account of Salt Marsh.

Accuracy of Li Tao-yuan's description,

overlooks the valley between Achchik-bulak and Besh-toghrak, that the ancient route from Lou-lan would find its natural and easy continuation towards the terminal basin of the Su-lo-ho and the adjoining end of the Tun-huang Limes. The most direct line, it is true, from near our Camp civ towards Besh-toghrak would have led through the southernmost hill range of the Pei-shan. But there, in an absolute desert of stone and gravel, neither grazing nor water could have been found. Nor would the saving in distance, being comparatively slight, have compensated for the needlessly prolonged hardships.

Before, however, I proceed to give an account of our experiences along the former, the obvious, route, it will be convenient for us to turn back once more to the salt-encrusted wastes crossed on the preceding three marches. We may examine, in the light of the observations there made, the notices that the early Chinese records have preserved for is of the Lou-lan route where it passed the P`uch`ang lake or the ` Salt Marsh '. The most instructive of these notices, though not the earliest, is furnished by what Li Tao-yuan's commentary on the Shui thing tells us in continuation of the passage concerning the `Town of the Dragon'. The definite topographical indications furnished in that passage have enabled us to locate this ` site ' with confidence at the great belt of Mesas past which our march to the north-east of the Lou-lan area took us on February 28th.1 The graphic account which Li Tao-yuan proceeds to give us of the region comprising the ` Town of the Dragon' may easily be proved to be similarly derived from an authority possessed of accurate local knowledge, and this justifies our dealing here with it first. The account, as rendered in M. Chavannes' extracts from the commentary, runs as follows : 2

` This region has an extent of a thousand li ; it is entirely formed of salt, but of salt in a hard and solid state. The travellers who pass through it spread pieces of felt for all their domestic animals in order to make them sleep on these. If one digs beneath the surface one finds blocks of salt, big as large cushions, which are piled up one above the other in regular fashion. [In this region there are] as it were mists which rise and clouds which float, and rarely does one make out there the stars and the sun. Little is found there of living animals and plenty of demons and strange beings.'

` [That region] touches, on the western side, Shan-shan   f5 , and connects, on the eastern

side, with the Three Sands =   I t constitutes the northern limit of the lake. This is why the

Pu-ch`ang [lake] A pi also bears the name of the " Marsh of Salt ", Yen-tsê   .'

When previously discussing this account among the Chinese records concerning the Lou-lan Site,3 I have indicated briefly how closely the general description here given of the region near the ` Town of the Dragon ' agrees with the result of our surveys of 1914-15 of the vast area comprising the ancient sea-bed and the wastes immediately adjacent. The statement as to the ground ` entirely formed of salt, hard and solid ' is completely borne out by the observations recorded above of the bottom of the dried-up sea where we crossed it. Làl Singh, whose route lay farther north, encountered the same terrible surface of crumpled-up salt over a wider stretch of its northern extension. Of the trials encountered on this vast expanse of hard salt crust, when traversing it over a still greater distance from south to north, Professor Huntington has given a very graphic record.4 The reference in the Chinese account to the piled-up ` blocks of salt, big as large cushions ', must appeal vividly by its truthfulness to those who have had to pick their way between and across those endless hummocks and huge buckled-up cakes of hard salt and have also seen the salt blocks, pressed one above the other like pack-ice, deep down in those innumerable fissures and hollows above mentioned (Fig. 18o).

1 See above, pp. 292 sqq.   S Cf. Serindia, i. pp. 424 sq.

' Cf, T`oung-pao, 1905, p, 571.   6 See Huntington, Pulse of Asia, pp. 251 sqq.

Rr