490 183. COTTON
(=Abû Hanifa) that, in their place, cotton grew so as to reach the height of an apricot tree, and lasted twenty years »; Abû Hanifa continues by saying that «the best (cotton) was the one which was recent and had been sowed in the course of the year » (cf. LECLERC, Traité des simples, III, 9293). Polo says the same.
So we know from the Wu lu that a true cotton, the Gossypium arboreum, was cultivated in Yung-ch'ang in the last quarter of the 3rd cent. Now, it is precisely in the region of Yungch'ang that the Hua-yang kuo-chih and the Hou-Han shu mention the po-tieh cloth of the Ai-lao. The Hua-yang kuo-chih is slightly later in date than the Wu lu, so that the interpretation of potieh as « white [true] cotton cloth » would make no difficulty. As to the Hou-Han shu, it is true that, in principle, it deals with a period earlier than the Wu lu; but we have seen that, in the present case, it seems in fact to be indebted to the Hua-yang kuo-chih. Even if it were not so and if both works had copied from a lost and unknown source of the Han period, we should have to go back only to the 2nd cent. A. D., one century before the Wu lu, and there is no reason to suppose that the cultivation of cotton had just begun in western Yün-nan when the Wu lu was written. I have long thought of devoting a special memoir to the route between Burma and Yunnan, a route which was already known to Greek geographers and which has been of much greater moment in the history of civilization than is generally believed. The early introduction of cotton cultivation into western Yün-nan is but a link in a long chain.
From the Wu lu again, we know that, in the 3rd cent., true cotton, represented by the Gossypium arboreum, was cultivated not only in western Yün-nan, but also in North Annam and Tongking. Both regions formed part of the Chinese Empire, but they were border provinces, and cotton was still unknown in China proper, except as an imported ware. The same location of mu-mien in western Yün-nan and Tongking occurs in the extant fragments of the Kuang chih (cf. supra, p. 462). In the 4th cent., the Lo-fou shan chi, devoted in theory to what was to be found on that mountain of Kuang-tung, speaks of mu-mien, which, in the first month, produced flowers similar to those of the fu-jung (« nelumbium », or « hibiscus »; cf. supra, p. 461), and a floss out of which Southerners made cotton wool. The early flowering and the fact that the text speaks of the floss as being used not for making cloth, but for stuffing, would favour the hypothesis that we find already in this early text the confusion, well attested in Ming times, which transferred the designation mu-mien from the Gossypium to the Bombax. The same doubt may be entertained as to the value of mu-mien in a passage of the Nan-yüeh chih (cf. supra, p. 462) and in an anonymous Kuang-chou chi, probably pre-T'ang in date, which says that the mu-mien occurs in Tongking and Kuang-tung (cf. supra, p. 462). But the fact remains that, already before the T'ang dynasty, there was some mu-mien growing in Kuang-tung; and we must not forget that Hai-nan was part of Kuang-tung province.
That Gossypium arboreum, and not Gossypium herbaceum, was the species mainly grown in eastern Indo-China in the 6th cent. seems to be established by the description given in Liang shu, 54, i b (copied also into Nan shih, 78, z b). The text speaks of the products of Lin-i (= Champa), which included chi-pei (ku-pei in Nan shih), and adds : « Chi pei is the name of a tree (shu; not ts'ao, ' plant'). When its flowers are completely formed, they are like goose down. [The people] pull out the filaments (11 hsii) and spin them so as to make a cloth which
IIIF Color HighRes
