one would tike to find elsewhere some confirmation of the fact that ghar was more or less known in the region towards the middle of the 18th cent. Whatever the case may be, neither Pers. kâshi, nor Mong. Or or Hindi ghar can actually be the source of the name « Kàsyar ».
VIVIEN DE SAINT-MARTIN'S remark (in JULIEN, Mémoires, II, 427) that « gar is the common ending of a great number of place-names in the dialects of northern India, with the meaning city '» has been simply repeated by FRANKE (SPA W, 1903, 186, who erroneously credits JULIEN with it), but is ignored by STEIN, and rightly, I believe; the name has no chance of being Indian. THOMAS (Tibetan Texts and Documents, I, 166) also speaks of « gar », which is « perhaps fundamentally the same as Tibetan sgar, gar, encampment'» and « was, no doubt, widely used in Tibeto-Burman dialects»; Kàsyar is one of the names adduced. I am afraid that much of THOMAS'S list must be left out, for instance « rGya-gar », India, « Pho-dkar », which I believe to be Bokhara, and « Kàsyar » itself.
RICHTOFEN's attempt at explaining the first part of « Kàsyar » from the Turk. qas, « jade », was opposed by STEIN (loc. cit. 50), who however, like RICHTHOFEN, felt inclined to connect with qas the name of the Kâaca 6pn. STEIN'S opinion was that the name of the qas stone may have come to the Greeks from the Great Yüeh-chih « who probably spoke a language of the TurkiMongolian family ». MARQUART, on the other hand, saw in qas the outcome of « an old vrddhiform *khâsa = stone from the Khasa or Kao-l'a land » ( Wehrôt and der Fluss Arang, 68; Ueber das Volkstum der Komanen, 201). HERRMANN (art. « Kasia » in PAULY-WISSOWA) repeats MARQUART's etymology Khàsa > qas, and considers as established that the KArza 61077 owed their name to qas, jade (cf. also my remarks in TP, 1930, 299, and HERRMANN, Das Land der Seide and Tibet, 141-142). I do not believe in the Turco-Mongolian appurtenance of the Great Yüehchih, nor do I think it probable that qas is derived from *khäsa. It would rather appeal to me
simply to connect Khasa and the Kaa-ca As to the word qas, it occurs for the first time in
1076 in Kàsyari (BROCKELMANN, 150); but its earlier history is still unknown (see also «Cotan »).
BURNOUF had the intuition of what I believe to be partly the true etymology of the name « Ka;') ar », when he suggested to HUMBOLDT *Khasagairi, « Hill of the Khasa » (cf. STEIN, Ancient Khotan, 50-51). Not that I think that kâs- has anything to do with Khasa, but the second half of the word seems to me to have the greatest probability of corresponding to Skr. girl, Avest. gairi, Khotan. Bari, Wakh. yar, Yayn. gor < *gar. This may even be what Hui-lin meant when he rendered *Kasgiri in Chinese as « Garrison of the Onion Range » instead of « Garrison of Shu-lo ». A natural inference would be to look also into the languages of the Pamir for a word meaning < onion » as the first component of *Kasgiri. It just happens that the word for « onion » is kasu in Shin'a, )'asu in Burusaski (cf. LORIMER, The Burushaski Language, III, 182). But these are not Iranian languages, and they have no word resembling gairi, girl, for « mountain ». Hui-sin's gloss is not bound to give us the literal sense of the whole name, and I prefer to consider the first element of *Kagiri as still undetermined.
But, alongside of *Kasgiri, a short form based only on the first element seems to have been current in Sogdian. This is at least the conclusion we must accept if, as appears very probable, Kàge is to be understood as < Kàsak and means « of Kàs; ar » , « Kâshgarian », while the Q's'nc = Kàsànc of Mahrnâmag, line 146, would represent its feminine form (cf. HENNING, in
IIIF Color HighRes
