50 HISTORICAL NOTICES OF KASHGAR [Chap. III
We cannot attach greater value to a second etymology of Su-lé, found in a gloss on Hsüantsang's account of Kashgar, which represents that old name as a corruption of Shih-li-chi-li-to-ti 15. Here, too, a Sanskrit derivation is evidently intended, but the meaning of the alleged original is not vouchsafed to us. Nor need we regret this reticence in view of the manifest artificiality of the explanation 16.
Before leaving, however, this old name Su-lé, it may be well to point out that Dr. Franke suggests Sulek or Surak as an earlier pronunciation of the Chinese name, and that such a reading possibly finds support in the form Shulik, by which Kashgar is designated in early Tibetan records 17.
The etymological speculations of Western scholars have naturally turned upon the name Kâshgar, by which alone the territory and its chief town have been known since the earliest Muhammadan conquest. Resemblance in sound has tempted a series of distinguished scholars, extending from De Guignes and D'Anville to Lassen and V. de Saint-Martin, to connect the name of Kashgar with the Kc na ôpq of Ptolemy 18. But Baron Richthofen, in his masterly analysis of this portion of Ptolemy's geography, has conclusively demonstrated that the Kasia mountains represent the great Kun-lun range, and that, in view of the considerable distance which separates this from Kâshgar, the assumption of a direct connexion between the two names must fall to the ground 19.
It was no less a scholar than Eugène Burnouf who first coupled the equation Kâvta op' : Kashgar, with another equally problematical conjecture deriving both names from that of the Khaga race frequently mentioned in Sanskrit texts 20. The weakness of this identification, from the geographical point of view, did not escape the critical judgement of Baron Richthofen,
1S See Julien, Mémoires, ii. p. 219 : ' Anciennement ce royaume s'appelait Sou-le; c'était le nom de sa capitale. L'orthographe correcte est Chi.la-ki-li-to-ti (Çri-kritati). Le mot Sou-le est corrompu.' As regards the authenticity of this gloss, compare Franke, Sb.P.A.W., 1903, p. 738.
16 Julien's transcription *hri-kritati permits of no appropriate,explanation as far as the second part is concerned ; Skr. Sri can safely be recognized in the first part, since the
two initial characters 14j are regularly used by
Hsüan-tsang to render that term ; see Julien, Mémoires, ii. p. 504, Dr. Franke proposes to transcribe *Srikiritadhi holding the diadem of Fortune,' and recognizes in this an attempt to provide an auspicious alternative for the somewhat derogatory etymology previously discussed; comp. Sb.P.A. W.,
1903, p. 738.
17 Compare Franke, Sb.P.A. W., 1903, p. 187 ; and concerning the Tibetan name the authorities there quoted Wassiljew, .Buddhismus, p. 55 ; Rockhill, Life of the Buddha, p. 24o n.
18 For references to the works of De Guignes, D'Anville, Humboldt, Lassen, compare Richthofen, China, i. p. 485, note 2 ; see also Ritter, Asien, v. p. 413, quoting Pater Georgi. It is from Richthofen, China, i. p. 494 n., that I take the reference to V. de Saint-Martin's view as to the location of tfie Kasia mountains near Kashgar, set forth in his Étude sur la géographie grecque et latine de l'Inde ' (Mémoires présentés d l'Académie des Inscriptions, vol. vi. r86o), in the chapter on ' la Sérique de Ptolémée' (pp. 258-286), which is at present not accessible to me.
" See Richthofen, China, i. pp. 484 sqq. In note 4, p. 485, the possibility of a derivation of Kashgar from Kash, the Turki term for jade, is learnedly discussed. The evidence in support of such a derivation appears to me very scanty, seeing that Kâshgar lies far away from the region which has been the source of the jade supply of Turkestan from ancient times, Khotan and the valleys of the Upper Yarkand river. Nor can Kâshgar be rightly assumed to have ever been the chief centre for the once flourishing jade trade. In ancient times, when a great commercial route led along the southern edge of the Tarim Basin direct from Khotan to Lop-Nor and Sha-chou, Khotan itself was, no doubt, this centre. When that route fell into disuse during the Middle Ages, Yarkand became naturally the main mart, as clearly indicated in Benedict Goéz' narrative ; see Yule, Cathay, ii. p. 564.
It would be far more tempting to seek, with Baron Richthofen, a similar relation between the names of the Kash stone and of the Kasia Mountains. Traders like the agents of the Macedonian Mats, upon whose information the account of Serica given by Marinus and reproduced by Ptolemy mainly rests, may well have recorded the local names of that distant region as they heard them from the Great Yüeh-chih' people in the more accessible Oxus region, who probably spoke a language of the Turki-Mongolian family (compare for the linguistic relationship of the ' Great Yüeh-chih' my remarks in Indian Antiquary, xxxiv. p. 84).
20 Burnouf's suggestion was contained in a note contributed to Humboldt, Asie Centrale, i. p. 115. His derivation of Kashgar from a supposed form *Khafagairi
Western etymologies of the name.