FROM CHARCHAN 'I'O CHARKHLIK
SECTION I.—EARLY ACCOUNTS OF CHARCHAN
THE journey from the Endere River to Charchan was covered in six marches, between November 16 and 20, the same number as in Hsüan-tsang's itinerary. The route, described already in my personal narrative,1 can have changed but little since his time, lying in an almost straight line to the north-east. It carefully hugs the line where the glacis of the sterile ` Sai' of gravel, sloping down from the foot of the Ktun-lun and overrun in parts by high ridges of sand, is fringed northward by a zone of desert vegetation, varying in width and as yet unsurveyed (see Maps Nos. 43, 46).
For direct archaeological observations there was no scope here. But I was able to convince
myself here also of the truthfulness of the record which Hsüan-tsaug has left of the impressions gathered on his journey eastwards to Lop-nor.2 On leaving- the territory of Ni fang or Niya ` he took the direction to the east and entered a great desert of moving sands. These sands have an immense extent ; they are piled up or scattered according to the wind. As there are no track for wayfarers, many go astray. On every side there extends a vast space, with nothing to go by. So travellers collect the bones of animals left behind to serve as road-marks. There is neither water nor grazing, and hot winds frequently blow. When these winds rise, men and animals lose their senses and become unwell. Often one hears singing and whistling, and sometimes wailing. While looking and listening, one becomes stupified and unable to direct oneself. Hence travellers frequently lose their lives here. The phenomena are caused by demons and sprites.' We shall see how curiously the facts and superstitions here described reappear in Marco Polo's account of the great desert crossed between Lop-nor and Sha-chou.3 No doubt, the pilgrim's remarks were meant to apply generally to the desert route as he saw it on his way from Niya to the Lop tract and hence to Sha-chou or Tun-huang.
For the period immediately following Hsüan-tsang's journey, a succinct account of this route is
contained in the itinerary which the Tang Annals furnish from Sha-chou to Khotan, and of which M. Chavannes has translated an abstract.4 We are told there that after leaving ` the garrison of Po-hsien, which is the ancient town of Chü-mo ' and, as we shall presently see, identical with the modern Charchan, ` one passes the Hsi-li-claih wells, the Yao wells, the river Wu-chê, and, after 500 li, arrives at the military post of the town of Lan, which is east of Yii-t`ien'. By the latter expression we may assume that the eastern frontier of Khotan territory is meant, and in this case the dis-
tance indicated would justify us in identifying ' the military post of the town of Lan' with
the Tang fort of the Endere Site first explored in 1901. The mention of wells as stages west
' See Desert Cathay, i. pp. 317 sqq.
2 See Julien, Mhnoires, ii. p. 246 ; Watters, Yuan Chwang, ii. p. 304. I reproduce the latter translation, except where Julien's version seems to give a better context.
3 Cf. Yule, Marco Polo, i. pp. 196 sq., with the illuminat
ing notes of Sir Henry Yule on the widespread belief in goblins haunting deserts, ibid., pp. 201 sq.
4 See Chavannes, Voyage de Song Yun, p. 12, note 9 ; also cf. Ancient Khotan, i. p. 436, note 14.