436 183. COTTON
thought that the intermediary form was the Malay kapas ; LAUFER (Sino-Iranica, 491), taking into consideration only the form kupei, stated that the Chinese must have received the term from an Indo-Chinese language, and that Bahnar köpaila provided « the nearest approach » to ku-pei. I cannot entirely agree with either opinion. The correspondence Indonesian (and Cham) -a- > Bahnar -5- is well known, and the -ö- of the modern Bahnar form köpai ja must be a late development, as is shown by the various Indo-Chinese and Indonesian forms, conveniently collected by CABATON, Dictionnaire Cam-Français, 57 : Jay., Dayak, Malay, Sund. kâpas, Makass. kapasa, Battak hapas [and kapas], Bis. gapas, Cham kapah, Bahnar köpaiïi (add Radè kapas, Kuoi kabas, K6o kopas, Sedang köpè), Khmer krâbas (add Khmer kâbbas , kâppas, as a secondary form derived from Pali kappâsa). The word seems to be represented in Bima as kafa, « thread », and even to have passed into Melanesian and Polynesian languages : Fijian kava, Samoan 'afa, Maori kafa, all meaning « string » (cf. O. DEMPWOLFF, Die Lautentsprechungen der indones. Lippenlaute ... , Berlin, 1920, 8vo, p. 13, and BLAGDEN, in BSOS, II, 152). But, on the other hand, the Malay Opus is also out of the question. Just as we have Skr. karpâsa and Pali kappâsa, we find in Indian the modern forms Beng. keirpeis and kapâs, Hindi kapâs, kapâs, (for other forms in modern India dialects, cf. J. BLOCH, La formation de la langue marathe, 309; for Anglo-Indian, cf. YULE, Hobson-Jobson2, s. v. « capass »). In the Indo-Chinese forms, only the Khmer form krâbas shows an r (with metathesis) as in the Skr. karpâsa (it may go back to the period of Sanskrit influence in Cambodia). There is no etymology of karpâsa and the word stands isolated in Indo-European languages (the connection suggested by UHLENBECK, Kurzgefasstes etymolog. Wörterbuch, 46-47, with Skr. kûrpâsa, «woman's jacket », is highly problematical, and leads nowhere, since the origin of kûrpâsa is also unknown) ; in all likelihood, the word is pre-Aryan in India (cf. the very similar Skr. karpûra, Pali kappura, modern Indian dialects kapiür and keipar, Jay. kapur, Mal. kâpur). So far I agree with PRZYLUSKI's views as expressed in Bull. Soc. Ling., xxv, 1924, 69-70; but here he and I must part. PRZYLUSKI, starting from the Indo-Chinese languages, supposed a root *bas, meaning «to use a bow », both for shooting, and for carding cotton (I do not know why PRZYLUSKI does not adduce in favour of his thesis Stiengpahi, Annamite vcti, and Siamese fài, all meaning «cotton »). The forms of the kâpas type would represent this root with a ka- prefix, enlarged by a liquid in the Skr. karpâsa; to the same root, with the dropping of the k- and a nasal infix, would also belong the Khmer ambas, ambôlt., meaning « cotton » according to PRZYLUSKI , who, on the other hand, does not cite Khmer krâbas. None of these hypotheses is in itself impossible, since parallel cases may easily be adduced for every one of the supposed transformations of the root. But it is krâbas which is the word used in Khmer for « cotton » in general, and ambas, or ambôh, means «spun cotton », «cotton thread », and I see no sufficient reason to believe that the two words are connected. If there are in Indo-China and Indonesia other words probably connected
with but not directly traceable to karpâsa, I would gather look for them in the Malay keipuq,
kâpoq, « floss of the silk-cotton tree », our « kapok (cf. LOKOTSCH, No. 1067), a word which has passed into the Amoy dialect as ka pék-mt", « kapok cotton » (C. DOUGLAS, loc. cit., 329), and in the Malay kâba or kabukâbic «cotton », which is also the general name for cotton among the primitive tribes of the Malay Peninsula (cf. SKEAT and BLAGDEN, Pagan Races of the Malay
IIIF Color HighRes
