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

THE DRAGON LAKE ACCORDING TO RAWLINSON, CURZON AND STEIN.   55

than it is now. Sor-köl was the Dragon Lake of SUNG YÜN. The stream flowing out on the west is the Pamir river. Regarding the hydrography to the east Hsüan-chuang made simply a mistake committed by later and better-equipped travellers.'

That Victoria, and not Chakmak, Lake is alluded to is further demonstrated by the direction, viz. southeast, taken by the pilgrim, after passing the lake on his way to Chieh-pan-t`o (assuming the latter, with Yule, to be Sarikol with its capital Tashkurghan). If he had been marching by the Little Pamir route, he would have had to proceed north-east from Chakmak to Aktash.2

In my opinion the size does not prove much, as the exaggeration is enormous whichever of the Pamir lakes is chosen. Nor is anything proved by the direction, which in both cases is wrong, for T ash-kurg an is situated to the E. N. E., and not S. E. of Sor-köl, and N. E. of Gliakmakden-kill. These arguments have no greater value than those which could be used in favour of Kara-köl, where I, in the beginning of March 1894, marching over the frozen lake, was told by my native servants from Ferghāna that the peculiar sounds accompanying the steps of our horses over the ice, were caused by big fishes (perhaps dragons !) knocking their heads against the ice. Nor could the depth of down to 756 feet be an argument for Hsüan-chuang's: «it cannot be fathomed».

The most serious and erudite of all attemps that have been made to reconstruct the itinerary of Hsüan-chuang and bring it in accordance with our present knowledge of Chinese geographical literature and the results of modern exploration, is that of STEIN in his work » Ancient Khotan» .3

M. A. STEIN regards the great Chinese pilgrim as our main authority concerning ancient Sarikol. In the summer of 642 A. D. he returned from India, passing, on his homeward way to China, through Badakhshan and into the kingdom of Ta-mo-Iasi=t`ie-ta, which beyond doubt has been identified with Wakhan. Seven marches to the N. E. brought him to the valley of Pa-mi-lo. Stein believes he travelled through Great Pamir and past Wood's lake, for the pilgrim's description of his great Dragon Lake in a higher degree resembles the Sor-kul than any other lake in the Pamirs. Five hundred li to the S. E. took him through uninhabited regions and over mountains covered with ice and snow to the kingdom of Chieti p`an-ro or Sarikol. Had he travelled S. E. from Little Pamir and Chakmakdenköl, he should not have arrived in Sarikol, a view in which Stein agrees with Curzon. From Great Pamir he had to make a choice between two different roads, either eastward via the Naiza-task Pass, or into the Aksu valley at the issue of the river from Little Pamir, and thence by Bayik and other passes into the upper part of Tab hdumbash Pamir. The general direction would here be S. E.4 And this road had been taken by Hsüan-chuang.

As a third alternative Stein regards the track across the Wakhjir Pass, where an important caravan road passed in ancient times. In 190! Stein even regarded this route as the most probable :

ictt   Notwithstanding its great elevation 5, the Wakhjir Pass and its approaches both from west and

east are comparatively easy. Comparing the topographical facts with Hiuen-Tsian's account in the

d,   Hsi-yü-chi, I am led to conclude that the route followed by the great Chinese pilgrim, when travelling

= This is quite true! We have only to think of the endless struggle about an outflow from Manasarovar.

2 Op. cit., p. 70.

3 P. 3o et seq. et passim.

4 This can hardly be said to be the case, as even the upper portion of the Taghdumbaslì valley is E. S. E. of Sor-köl.

5 Stein gives 16 Zoo feet. I found Aug. i 5 th, 1895 16 ego feet.