600 196. ÇANGHIBAR
of southern Indo-Chinese and Indonesian races. K'un-lun was not always used with the pejorative meaning attributed to it by Hui-lin; we hear for instance of Buddhist literature in the K'un-lun language. Nevertheless, the name soon became a designation of blackish people, with curly hair, who often came to China as slaves. The K'un-lun thus began to be mixed up with the Zängi. In other words, Indonesian Negritos may have been called Zängi without being African negroes, while African Zängi came also to be known in China as K'un-lun. In exactly the same terms which Hui-lin employed for the K'un-lun barbarians, Idrisi speaks of the natives of Jalus (? read Bàiùs = Baros), probably in Sumatra, who he says are Zänj (transi. JAUBERT, I, 77, 79; DEVIC, Le pays des Zendjs, 110; Fe, 182, 184).
Since I discussed the problem of the K'un-lun in 1904 (BEFEO, iv, 215-231), FERRAND has examined it in great detail in JA (1919), and I have added further texts in TP, 1923, 271-272, 290, and in Etudes asiatiques, II, 257-263. A T'ang Sino-Sanskrit vocabulary renders K'un-lun as Skr. Dvipântara (cf. Mémorial Sylvain Lévi, 392-397), meaning something like «the other continent ». Despite FERRAND'S researches, we cannot yet tell what the original was of the word transcribed as K'un-lun. FERRAND was not far from believing that it was Qomr, Qomor, and he would identify with that name the name of the Khmêr and the African Qomr (cf. JA, 1919, II, 212). The question is made more obscure by the earlier existence of a K'un-lun in Chinese legend (the original Qomr according to FERRAND !), and also by the deficiency of Middle Chinese which did not have a phonem like -om or -um. Moreover, the traditions relative to Kaundinya,
which is sometimes transcribed A. Chü-iun (*Kiu-ljuén) may have reacted on the Indo-Chinese
Ku-lun or K'un-lun; finally, at a later date, K'un-lun became the Chinese name of Pub Condor (see «Condur »). To try to unravel such an entangled skein would require a whole monograph, only to reach perhaps an indifferent result. Suffice it to say here that the name K'un-lun was applied by the Chinese to black curly-haired (or frizzy-haired) races at least as early as the 4th cent. I have dilated somewhat on the name K'un-lun because we find it in Sung times much more closely connected with Zängi than in Hui-lin's note of 817. Chou Ch'ü-fei, whose Ling-wai tai-ta was written in 1178, devoted a paragraph to a *ffl. pjjK'un-lun Ts'êng-ch'i (*Kuan-luan Dz'ang-g'ji), an island where there were giant birds which could mask the sun in their flight (see
« Ruc ») and the inhabitants of which, black as lacquer and with frizzled hair, were carried off as slaves to Arab masters. This text was copied almost word for word by Chao Ju-kua in his Chu fan chih of 1225 (HR, 149). HIRTH and ROCKHILL, interpreting K'un-lun Ts'êng-ch'i as «The Zanj (or Blacks) from K'un-iun », saw in K'un-lun a transcription of Kanbalu, the native name of our Pemba Island, north-north-east of Zanzibar. For FERRAND (JA, 1919, II, 208-210), K'un-lun represented Qàmrùn, in a series Qomr (in Central Asia) > Qàmrùn (in Indo-China and Indonesia) > Qomr (at Madagascar and on the oriental coast of Africa); as the form Qàmrùn was not known in Africa, the true original of K'un-lun Ts'êng-ch'i must have been «Zang of Qomr ». With all due regard for FERRAND'S industry, the whole theory seems to me ruinous. Particularly, in so far as K'un-lun Ts'êng-ch'i is concerned, I do not accept for one moment that the Chinese should, when they had heard « Zang of Qomr », feel that Qomr was the same as a Qàmrùn which they elsewhere transcribed K'un-lun. In the case of Chou Ch'ü-fei, we have to deal with a man who had never left China, and who, when collecting his information from merchants and sailors, could
IIIF Color HighRes
