4,48 183. COTTON
yang commentary on the Ch'un-ch'iu (cf. Txu-t'ung, 18, 9). The Chinese surname of the royal family of Kuéâ is given as CJ Po in Chin shu, 97, 6 a, Pei shih, 97, 6 a, Sui shu, 83, 5 b, Chiu T'ang shu,198, 7 a, and Hsin T'ang shu, 221 A, 8 a, but as , Po in Liang shu, 54, 14 a, and Tzii-chip t'ung-chien (cf. CHAVANNES, in Ancient Khotan, I, 544). In Buddhist works, the surname is usually written ,M Po for people who either are of Kuchean origin, or have taken over the surname of a Kuchean master (cf. NANJI6, Catalogue, App. II, Nos. 28, 36; BEFEO, vi, 350; x, 224; BAGCHI, Le canon bouddhique, II, 129, 319). In his Hsing-shih hsün yiian (43, 4 b), CHANG Chu gives only , Po as the surname of the king of Ku&i. In the case of po-tieh, the first form, seemingly meaning « silky tieh », occurs in the Hua-yang kuo-chih and the Hou-Han shu, perhaps both drawing from the Wei-lio, and I know only one other example of it (the verse of Chou Pang-yen, 1057-1121, cited in the P'ei-wên yün fu is merely copied from the Hou-Han shu), in a siitra translated c. A. D. 300 (NANn6, No. 165, ch. 4; Tokyo Tripit. of Meiji, $, x, 41 a, reading confirmed ibid. ,, VIII, 186 a); here a true cotton fabric is certainly meant, and is described as such in Hui-sin's gloss. The second form, meaning apparently «white tieh », is much more common, and has finally prevailed. We find it first, in the second quarter of the 3rd cent., in Mêng K'ang's gloss on « ta cloth ». Almost at the same date, it occurs in the
fit Wu shih wai-kuo chuan, that is to say in one the accounts of the mission of K'ang
T'ai and Chu Ying to Fu-nan (Cambodia) c. 245-250 (cf. TP, 1923, 121-122; Etudes asiatiques.. . de l'EFEO, II, 245). This account is lost, but many quotations from it have come down to us, including the following one, which has been preserved in the commentary on the Hou-Han shu, 116, 8 a, and in the T'ai-p'ing yti-lan, 820, 20 b : « The women of the country of 4 fkijChu-po
weave po-tieh flower cloth » ?E ;f ti po-tieh hua pu, i. e. « cloth made with po-tieh flowers »;
on this meaning of hua pu, cf. infra, p. 475; my old rendering « cotonnades à ramages » in BEFEO, iv, 270, 275, may have to be abandoned; the same may perhaps be said of my translation « étoffes à ramages » for hua-pu in BEFEO, II, 146, 147, 176, and of the analogous translations
in HR, 48 [where I am not certain that Q / .fi chi-pei hua pu is not the designation of
only one product], 152, 155, 171; although the meaning of hua pu may have changed between Chin and Sui or post-Sui times, I am struck by the fact that, as late as 1618, the Tung-hsi yang k'ao [6, 14 b] employs hua pu in the general sense of «cotton goods », lit. « cloth [made] from flowers »). Chu-po (*Tigiwo-b'âk) certainly renders a derived form *Javag (<1- *Javaka) of the name of Java, although it may designate Sumatra as well (see « Java »; cf. BEFEO, III, 264; Iv, 270, 275-276; FERRAND, in JA, 1919, II, 5-6; 1922, II, 209). The transcription is not satisfactory,
although a later Chê-po (*Zia-b'âk) provides an interesting parallel form for the second
part of the name. As to 2 chu (*Cgiwo), we may here again suppose a labialization of the first vowel under the influence of the following labial consonant, but the surd nature of the initial consonant, which ought to render c~ , not j-, remains unexplained. In Kuo P'o's (276-324) commentary on the Shan-hai thing (Ching-hsiin-t'ung ts'ung-shu ed., 16, 2 b; from there it has passed into the sub-commentary on the Êrh ya, Shih-san-ching chu-shu ed., 9, 24 a), we find a quotation which is copied verbatim from the Wu-shih wai-kuo-chuan (it is the one on the Huoshan, the Volcano, and my note in BEFEO, III, 276, has to be changed accordingly), and there
the name is written Ch'i-po (*G'ji-b'âk), clearly corrupt for ; {j Chê-po; but it would
IIIF Color HighRes
