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

Kāšyar and Yārkänd by Kāšyarī. But there is also Hsüan-tsang's « great sand mountain », which
he passed, according to the text of the Memoirs, after crossing the Šitā or Yārkänd River and before
reaching Chê-chü-chia. I think that this « great sand mountain » was the « desert » which was res-
ponsible for the name « South of the Desert » being given to Chih-chih-man. Here again we are
faced with the alternative either of correcting Hsüan-tsang's text and assuming that, coming from
Kāšyar, he crossed the «great sand mountain» before reaching the Yārkänd River; or of looking
for the « great sand mountain » much farther to the south. But I doubt whether the second solution
would be acceptable, since Chia Tan's itinerary certainly does not follow a round-about course to
Kök-yar. On the whole, I am inclined rather to correct Hsüan-tsang's text and to identify with
Qaryalïq both *Čakuka and *Čukupa. The surprising fact remains, nevertheless, that Hsüan-tsang,
when speaking of *Čakuka, gives as its «ancient» name a form Chü-ch'ü which is neither the ancient
name of Yārkänd, nor any ancient form of the name of Qaryalïq, be it Tzü-ho or Chu-chü-po.
A last name, declared by HERRMANN, with Sven HEDIN's approval, to refer to Yārkänd, is that
of the kingdom of 烏 鎩 Wu-sha, through which Hsüan-tsang passed on his way from Taš-quryan
to Kāšyar (Southern Tibet, viii, 59-60, 451). I cannot agree with such a view. There could be no
point, when going from Taš-quryan to Kāšyar, in making first a long detour towards Yārkänd. In my
opinion, STEIN's commentary on that part of the Memoirs is conclusive : from Taš-quryan, Hsüan-
tsang went north-north-east, crossing the Chichiklik-davan, and arrived at Kāšyar via Igiz-yar and
Yangï-ḥiṣār; the kingdom of Wu-sha must be the region of Yangï-ḥiṣār, as was already proposed many
years ago by VIVIEN DE SAINT-MARTIN (cf. STEIN, Ancient Khotan, 42-44, 87-88). I am less willing
to accept STEIN's view that Hsüan-tsang's Wu-sha is the same name which was written 烏 扞 Wu-
ch'a in Han times; STEIN is of course aware that the Han kingdom of Wu-ch'a was located in the
region of Taš-quryan, but supposes that the name had a much wider extension, applying to
Taš-quryan, Yangï-ḥiṣār and Yārkänd. But Hsüan-tsang's Wu-sha is *·Uo-ṣat (cf. T'oung Pao,
1936, 279, and add T'ang shu shih-yin, 24, 2a). As to Wu-ch'a, it is probably an ancient
*·Uo-d''a, although the phonetic glosses on the name are desperately corrupt (cf. T'oung Pao, 1936,
276-279). There is no satisfactory correspondence between the two names, in form, in place, or in
time; the best thing to do is to keep them apart.
Hsüan-tsang's route from Taš-quryan to Kāšyar must be the one described the other way round
in Chia Tan's itineraries (Hsin T'ang shu, 43 b, 15 a) : «From Shu-lo (Kāšyar), going to the south-
west, one enters the valley (谷 ku) of 劍 末 Chien-mo (*Kïem-muât), the 青 山 嶺 Ch'ing-shan-
ling (' Pass of the Blue-green Mountains '), the 青 嶺 Ch'ing-ling (' Blue-green Pass ') and the
不 忍 嶺 Pu-jên-ling (' Unendurable Pass '), and after 600 li arrives at the frontier post (守 捉
show-cho) of the Ts'ung-ling, which is the ancient kingdom of Chieh-p'an-t'o (*Karbanda, Taš-
quryan)... » An abridged version of the same passage is given in Hsin T'ang shu, 221 a, 9 b, and
has been translated by CHAVANNES, Doc. sur les Tou-kiue, 124, 311. Although it is possible to
follow the itinerary on the map, none of the names mentioned has so far been met with elsewhere.
STEIN says that, during his stay in Yārkänd, no « ancient sites in its vicinity nor any antiqua-
rian objects of pre-Muhammedan origin » were brought to his notice. At the same time, he reminds
us that c. A. D. 1500 the king of Kāšyar, in the course of his treasure-seeking excavations in ancient
cities, is reported to have found « much treasure » in the ruins of old Yārkänd (ELIAS and Ross,