446 183. COTTON
together»; fig na-i, as a designation of the Buddhist garb, anciently occurs as *03 na-i, for
instance in A, x, 16 a, and in , viii, 166 b; cf. also SOOTHILL and HODOUS, A Dictionary of
Chinese Buddhist terms, 336, 339; under the Yüan, the tribute from Fu-chien included 200 na
robes [ Fg na-ao], according to WANG Ch'i's Hsü Irén-hsien t'ung-k'ao, 32, 16 a). In his
lost commentary on the Ch'ien-Han shu entitled Han shu yin-i, Mêng K'ang, who lived in the first half of the 3rd cent., had a note saying that « to pu is J ai po-tieh » (cf. also T'ai-p'ing yü-lan, 691, 1 a; in spite of the Chêng-tzû t'ung, s. v. .A mien, Mêng K'ang of course does not yet speak of « chi-pei »). In the first half of the 7th cent., Yen Shih-ku opposed this view, saying (Ch'ien-Han shu, 91, 3 b) : « [Ta-pu] was a coarse cloth; it was cheap, and that is why it had to be delivered [in paying taxes] in the same weight as leather; it is not po-tieh. As to ta, [it expresses] the idea of ' thick and heavy' (J_ hou-chung), but readers have foolishly given it the sound t'a, which is not correct. » Later authors have quoted Yen Shih-ku's text, more or less accurately, in particular Chang Shou-chieh, who, in his commentary on Ssû-ma Ch'ien, dated 736 (not « 737 », as said by CHAVANNES, Mém. histor., I, ccxvi, and TP, 1907, 183, followed by LAUFER, The Story of the Pinna, 115), added on his own account (TAKIGAWA ed., 129, 35) : «Po-tieh is woven from cotton (mu-mien), which does not exist in China. »
Yet Yen Shih-ku's dogmatic assertion is not conclusive. The Tz'û yüan, under to pu, maintains that the to pu of Ma Yüan's biographical notice is the same as the hempen light
material known in Kuang-tung as Alf to pu or Ai, fx to-lo, which is worn in summer, and,
at the same time, under = ta, says that ta in to pu is the same as tieh, « to duplicate », and expresses the idea of « thickness ». But, in view of the alternation of to pu and to pu in the Hou-Han shu and the Tung-kuan Han chi, it seems probable that both designate the same textile and are doublets of the same term. On the other hand, this double form would almost imply that both represent a foreign southern term borrowed twice at different stages of phonetic evolution; the relation between ta (*tâp) and to (*tuo) would be practically the same as between chieh (*kivp) and ku (*kuo) in the transcriptions chieh-pei and ku-pei of *kappeii < kappa's( (cf. above, p. 441). Yü Chêng-hsieh (Kuei-ssü lei-kao, 14, 4 a) accepts Mêng K'ang's view that the to pu was po-tieh, « cotton ». FUJITA has also hypothetically connected with to pu the puzzling term tT ,ffi to pu which occurs several times in the Tao-i chih-lio, written in the middle of the 14th cent. (Tao-i chih-lio chiao-chu, Hsüeh-t'ang ts'ung-shu ed., 15 a); on this late to pu, cf. ROCKHILL, in TP, 1915, 86, 237, where to pu is supposed to be some cotton fabric (but the reference to a term it :(ri to pu in Chao Ju-kua must be suppressed, since Chao Ju-kua
actually speaks of f ;4i yii-ta pu; cf. HR, 160). I feel very sceptical about FUJITA'S
hypothesis. The interval in time is so great that the proposed connection remains en l'air unless the expression can be traced in some southern dialect. Moreover it may not be necessary. In his Chen-la fêng-t'u chi, « Description of the customs of Cambodia », Chou Ta-kuan, also a writer of the Yüan period, speaks of the rules which decided the kind of « cloth » ( iii pu) which everyone was entitled to « put on » (JT ta), and also of the brahmans, who, besides the « cloth put on » (ta pu) by everybody, wore a band (Ku-chin shuo hai ed., 6 a, b, 8 a; BEFEO, II, 146148). It may be that the to pu of the Tao-i chih-lio was a term of the spoken language, designating a cheap cotton fabric, a « cloth to wear ». One might be also tempted to see the same
IIIF Color HighRes
