PERSIAN TEXTILES RUGS 493 



as a product of India. 1 In the Sui Annals it appears as a product of 

 Persia. 2 CHAVANNES has justly rejected the fantastic explanation given 

 in the dictionary Si min, which merely rests on an attempt at punning. 

 The term, in fact, represents a transcription that corresponds to a 

 Middle-Persian word connected with the root Vtab ("to spin")* 

 cf. Persian taftan ("to twist, to spin"), tabad ("he spins"), tdfta or tqfte 

 ("garment woven of linen, kind of silken cloth, taffeta"). Greek Tcnrrjs 

 and TCLTrrjTlov (frequent in the Papyri; TairLdvQol, "rug-weavers") are 

 derived from Iranian. 3 There is a later Attic form dams. The Middle- 

 Persian form on which the Chinese transcription is based was perhaps 

 *taptaii, tapetaii, -an being the termination of the plural. The Persian 

 word resulted in our taffeta (med. Latin taffata, Italian taffeta, Spanish 

 tafetan). 



71. To the same type as the preceding one belongs another Chinese 

 transcription, ffi ffl fo(t*o)-pi, ftj 8? tso-p*i, or $5 '& tso-pi, dance- 

 rugs sent to China in A.D. 718 and 719 from Maimargh and Bukhara 

 respectively. 4 These forms correspond to an ancient *ta-bik (2 or i$) 

 or *ta-bi5 (&), and apparently go back to two Middle-Persian forms 

 *tabi% and *tabe5 or *tabiS (or possibly with medial p). & 



72. More particularly we hear in the relations of China with 

 Persia about a class of textiles styled yue no pu fif 'ft?. 6 As far as I 

 know, this term occurs for the first time in the Annals of the Sui Dy- 

 nasty (A.D. 590-617), in the notice on Po-se (Persia). 7 This indicates 

 that the object in question, and the term denoting it, hailed from Sasa- 

 nian Persia. 



1 E. CHAVANNES, Les Pays d'occident d'apres le Heou Han Chou (T'oung Pao, 

 1907, p. 193). Likewise Jin the Nan $i (Ch. 78, p. 5 b) and in Cao Zu-kwa (trans- 

 lation of HIRTH and ROCKHILL, p. in). 



2 Sui $11, Ch. 83, p. 7 b. 



3 P. HORN, Grundriss iran. Phil., Vol. I, pt. 2, p. 137. NOLDEKE'S notion 

 (Persische Studien, II, p. 40) that Persian tanbasa ("rug, carpet") should be derived 

 from the Greek word, in my opinion, is erroneous. 



4 CHAVANNES, T'oung Pao, 1904, p. 34. 



5 These two parallels possibly are apt to shed light on the Old High-German 

 duplicates tepplh and teppid. The latter has been traced directly to Italian tappeto 

 (Latin tapete, tapetum), but the origin of the spirant x i n tepplh has not yet been 

 explained, and can hardly be derived from the final /. Would derivation from an 

 Iranian source, direct or indirect, be possible? 



6 According to HIRTH (Chau Ju-kua, p. 220), "a light cotton gauze or muslin, 

 of two kinds, pure white, and spangled with gold"; but this is a doubtful explana- 

 tion. 



7 Sui su, Ch. 83, p. 7 b. This first citation of the term has escaped all previous 

 writers on the subject, Hirth, Chavannes, and Pelliot. From the Sui su the text 

 passed into the T'ai p'in hwan yil ki (Ch. 185, p. 18 b). 



