200 123. CASCAR
Prâkrit form of suvarna, « gold» (for -s- and 4-, cf. soma «moon », given as the original of a *lima° or *soma°, ibid. 1038; and for -m- and -y- cf. a *suman° interpreted as « gold » suvarna, ibid., 1039). We can establish from another source that *Khara-Syâmâka must be abandoned. As remarked by THOMAS (loc. cit. 93), the Tibetan translation, in the passage corresponding to the text of the Sûryagarbha which mentions Chia-lo-sha-mo, gives rDo-rcub « Hardstone », clearly representing an original Kharàsman. THOMAS adds that possibly there is some relation between the name Kharâsman and that of the sage Kha-ra-sva (Kharâéva) ; moreover, he is inclined to connect these names with that of Hwàrizm. I do not feel much inclined to accept the second hypothesis; and I have no doubt that the first one is correct. Kharàsva and Kharâsman (Kharâsma) are in fact one and the same name, and we have only here one more case of the confusion between -sva- and sma- to which we owe the double forms asvagarbha and asmagarbha, Asvaparânta and Asmàparânta. They may be due to the hesitation of Buddhist compilers when rendering into Sanskrit Prâkrit forms in -ssa- (cf. BEFEO, iv, 379; JASB, 1925, 48-51). Whatever the case may be, the various texts concerning *Kharostra (*Kharostrag) and *Kharâsman provide an interesting example of the way in which, in the first centuries of our era, traditional names which had perhaps quite a different value originally became connected with Chinese Turkestan, with occasional precise references to Khotan and even to Kâsyar.
The name of Shu-lo has been adduced by MARQUART (Erdnsahr, 282-284; cf. also STEIN, Ancient Khotan, i, 56) to explain the name of the convent of the kingdom of Kapi§I which is said in Hsüan-tsang's Life (JULIEN, Vie, 71) to have been built when a son of the Chinese Emperor was
there as a hostage, and which was known under the name of 7% 3 Sha-lo-chia (*iSa-lâk-ka).
But, at the same time, MARQUART saw in Sha-lo-chia, restored as Saraka, a name of the Chinese. The apparent contradiction seems to have been solved, in MARQUART'S mind, by his idea that the name, to be read as *Sâraka, originally meant « Prince of Kàsyar », but in Hsüan-tsang's time was misunderstood as meaning « Chinese ». As a matter of fact, the contradiction is in the names themselves. The name of Sarag, Skr. Saraga, is now well attested as that of Lo-yang, the eastern capital of the Han and the T'ang, but Sha-lo-chia can only be *Sâraka or *Sâlaka; to obtain the name
of China, or of Lo-yang, we ought to correct {p sha to so (*sud) or rather to %4. so (*sâ). If
we retain Sha-lo-chia and connect it with the form Sha-lo used by Buddhist pilgrims for Shu-lo, *Sâraka will hold good, but then the difficulty is that the Life speaks of a son of the Chinese Emperor, not of a son of the king of Kâsyar. I still hesitate between the two solutions (cf. my remarks in JA, 1927, II, 139-141, and also Mi, 225; MINORSKY's objection to the surprising transcription of g or gh as i by early Arab writers may perhaps be answered by supposing that the Arab transcription is based on *Sarak). I do not believe in MARQUART'S *Sârakavati.
I should perhaps feel more inclined to accept *Sarah as a form derived from Shu-lo or Shaio if we knew more about this name (MARQUART'S restoration as *Sàrak in Wehrôt and der Fluss Arang, 68, is possible, but not proved). It occurs in Tibetan texts of about the 10th cent. as Suiig (cf. THOMAS, in STEIN, Ancient Khotan, 581, 582), or sometimes Su-leg (THOMAS, Tibetan Texts and Documents, 118), but these appear to be forms merely borrowed from the Ch. Shu-lo. BAILEY (BSOS, VIII, 883) says that Sùlya « seems, in a document from the Khotan region, to mean the people of Kâsyar ». We must wait until we know more about it, since the initial is neither
IIIF Color HighRes
