298 FROM CHARCHAN TO CHARKHLIK [Chap. VIII
the figure of barbarians (hu) ; the old men questioned said that it was Lu Kuang who had had them made on the occasion of his expedition against the barbarians.'
Sung Yün's account is of interest because it makes it quite clear that the Charchan of his time had become an oasis of very modest extent. His description of the primitive conditions in which agriculture was practised points to a population occupying a far lower cultural plane than that indicated by the remains of the Niya Site more than two centuries earlier. This finds its probable explanation in Sung Yün's previous statement that the neighbouring Shan-shan or Lop-nor tract had been conquered and actually held by the Tu-yü-hun.25 We know that these were nomad tribes of uncertain origin who in Sung Yün's time and for centuries later occupied the high plateaus stretching westwards from the Koko-nor region. It is very probable that this Tu-yü-hun conquest at the time of Sung Yün's visit already extended to Charchan ; for the Pd shih distinctly mentions both Shan-shan and Chü-mo as territories held by the Tu-yü-hun during the reign of their king ICua-lü, who is mentioned for the first time in A. D. 540.2°
Sung Yün's reference to the sacred images supposed to date from Lu Kuang's invasion is also of historical interest. It shows that the expedition undertaken by this Chinese general into the Târim Basin in A. D. 384 was not restricted solely to the temporary subjection of Kara-shahr and Kuchâ.27 It further supplies us with a definite instance of the early influence which specimens of Chinese art must have exercised on Buddhist sculptural work in the Tarim Basin, in exchange as it were for the far stronger influence carried by Central-Asian Buddhist art eastwards. It only remains to add that the distance of sixteen hundred and forty li which Sung Yün's itinerary gives between Shan-shan and Tso-mo is greatly over-estimated, while the subsequent reckoning of twelve hundred and seventy-five li from Tso-mo to Han-mo, which corresponds to Hsiian-tsang's P1-mo,26 is distinctly too low. But, as M. Chavannes justly emphasizes, these and other serious errors in Sung Yün's distance estimates can cause no surprise, if account is taken of the critical defects of the text in which his itinerary has reached us.2°
Hsiian-tsang, following the same route just a century and a quarter after Sung Yiin, but in the reverse direction, has also left us a record of Charchan. After leaving the ruins of the old Tu-huo-lo kingdom, which, as we have seen already, must be located at Endere, ` he travelled
about six hundred li eastward and arrived at the old Ché-mo-to-na $jj- 115 kingdom, which
is precisely the territory of Chii-mo. The city walls are very lofty, but there are no inhabitants.' 30
That the form Nieh-mo, which the present text of the Hsi yii-chi shows here for Chii-mo .11*,
is but a graphic error, is clearly proved by the correct form found in the Life of the pilgrim. That his Chê-mo-t`o-na is probably meant as a reproduction of the current indigenous form of the name which in the Prâkrit of Kharosthi tablets from the Niya Site figures as Calmadana, has been previously mentioned.3'
Hsiian-tsang's testimony as to the deserted condition of Charchan at the time of his visit is of particular interest ; for when Chinese control had been re-established, some fifteen years after his passage, Charchan figures once more in the Tang dynasty's Annals as a place duly garrisoned. In an itinerary which is given by the Tang shu for the route from Shan-shan to Khotan, and of which M. Chavannes has translated an abstract,32 we are informed that after leaving
27 Cf. regarding this expedition, Chavannes, Voyage de Song Yun, p. 13, note 2 ; Ancient Kholan, i. p. 543, note 2.
25 See Chavannes, Voyage de Song Yun, p. i 2.
26 Cf. ibid., p. r 2, notes 2 and 7. As regards Shan-shan, cf. below, p. 323. [For the Tu-yü-hun, cf. now M. Pelliot in J. Asiat., 1916, janvier—février, pp. r r 7, 122.]
28 Cf. Ancient Kholan, i. p. 456.
29 Cf. Chavannes, Voyage de Song Yun, p. 14, note 4.
S0 See Julien, 31/moires, i. p. 247 ; also Vie de H.-Th., p. 290 ; Watters, Yuan Chwang, i. p. 304 ; cf. above, p. 288, and Ancient Kholan, i. p. 435, note 9.
91 See above, p. 297.
S2 Cf. Chavannes, Voyage de Song Yun, p. r 2, note 9. 11Iy extracts of this itinerary are here and elsewhere supple-