518 183. COTTON
paper, quoted in his turn the same story from the T'ung tien, and since Tu Yu, the author of the T'ung tien, quotes various passages of the lost account written by his kinsman Tu Huan after having returned from the Arab countries in 762, CHAVANNES supposed that Tu Huan had been the propagator of the legend in China; FERRAND (JA, 1932, I, 199) sided with CHAVANNES. In fact, both dates, 713 as well as 762, are too late. The passage is not included among those which Tu Yu quotes from Tu Huan's work, but, on the contrary, Tu Yu states most definitely that the story was told by the Arab envoys who came in yung-hui (650-655), in fact in 651; these envoys were those who said that their kingdom was then in its 34th (lunar) year, and was ruled by its third ruler. So the story of the wâgwdq tree in China must be dated 651 A.D., and is more ancient than any Arabic account.
It occurs in fact in other Chinese works than the three works quoted hitherto, but always
in the same form. The texts are : (1) , p , Shu-i chi (ch. I, 13 a), attributed to fl~ J Jên
Fang (460-508). But that a work of c. 500 should speak of the Ta-shih (Arabs), known to the Chinese only from the middle of the 7th cent., is of course a blatant anachronism. In fact, the Shu-i chi is a spurious work which is not earlier than the 8th or the 9th cent. (cf. Ssû-k'u ... , 142, 43-44). (2) The T'ung tien (193, 8 b), completed in 801. (3) The T'ang hui-yao of 804 (on which cf. JA, 1913, I, 262), as quoted in the Pei-hu lu of c. 875 (Shih-wan-chüan-lou ts'ungshu ed., I, 14 b). (4) The Chiu T'ang shu (198, 12 b), and the early quotations from it in T'aip'ing yii-lan (796, 14a, and 961, 3 b). (5) The T'ang hui-yao of 961 (100, 12 a; cf. JA, 1913, I, 262.) (6) The T'ai-p'ing huan yii chi (186, 13 b). (7) The T'ung-chih (196, 31 b). (8) The Wên-hsien t'ung-k'ao, 339, 7 a. All are derived from one and the same ,original. The most complete of the ancient texts is that of the T'ung tien : « [The envoys of 651] moreover said :
'Their king had sent ce ch'ang = ch'ang, given in the other texts) a man on a boat, with
clothing and provisions; [the man] went to sea, and sailed across it for eight years without reaching the western coast. In the middle of the sea, he saw a square rock; on the rock there was a tree, with red branches and blue-green (ch'ing) leaves. On the tree small children (4. W., hsiao-êrh) had been born all over, six or seven inches high. On seeing men, they did not speak, but all could laugh and wave their hands and feet (I think that CHAVANNES mispunctuated the text here, and that TERRIEN DE LACOUPERIE, 902, was right; but the author of the Tung-hsi yang-k'ao, 12, 14 b, read the text as CHAVANNES did). Their heads were attached to the branches of the tree. When people plucked [some of them] and took them into their hands, they immediately became dry and black. The envoy took a branch [of the tree] and went back; now [the branch] is kept at the residence of the king of the Ta-shih. » SCHLEGEL (in DE GOEJE, loc. cit. 303) added that «the name of the tree was ie-mie ». As CHAVANNES remarked (TP, 1904, 487), there is nothing of the kind either in the Wên-hsien t'ung-k'ao, or in the T'ung tien, or, I may add, in any of the parallel texts relating the story. The curious fact is that TERRIEN DE LACOUPERIE too (p. 902) gives the name, though as [ 74 yeh-mu, «coconut-tree »; but he knew SCHLEGEL'S note, and seems to have tacitly corrected it so as to fit his own identification of the tree. Although the tree is given no name in Chinese texts, it is clear that it is the same as the wâgwâq tree of Arabic texts; but it is no less certain that wàgwaq can be neither j si Yüeh-kuo (*Jjwet-kwak), «South China and northern Indo-China », as thought by TERRIEN DE LACOUPERIE,
IIIF Color HighRes
