国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

«there are a number of roaming vagabond people... They seldom show themselves, but yet they are
given to trade. Their wives and children, as frightful goblins as themselves, they carry about upon
donkeys » (Y¹, III, 245; this passage is omitted in Wy, 545); and further : « There are also wild men,
naked and hairy, who have wives and children, but abide in the woods. They do not show them-
selves among men, and I was seldom able to catch sight of one; for they hide themselves in the forest
when they perceive anyone coming. Yet they do a great deal of work, sowing and reaping corn and
other things; and when traders go to them, as I have myself witnessed, they put out what they have
to sell in the middle of the path, and run and hide. Then the purchasers go forward and deposit the
price, and take what has been set down » (Wy, 548; Y¹, III, 259). The first passage certainly refers
to the Veddah, the second one either to the Veddah or to the continental Poliar or southern India,
but more probably to the former.
The whole evidence seems quite consistent; yet I entertain doubt on some important points.
TENNENT (I, 570) takes Fa-hsien's account to refer to the conditions which prevailed in the traveller's
time, while Fa-hsien's very words show that he merely repeats what he heard about a legendary past.
About 400 A. D., Ceylon was a prosperous island, with a well advanced civilization and a prosperous
foreign trade. As to Al-Bīrūnī's text, it is true that Laṅkā was an ancient name of Ceylon, but the
conditions obtaining in Ceylon c. 1030 preclude that trade should have been done in the island on
the basis of the dumb trade described by the Mussulman writer. He himself remarks that the ac-
counts of Laṅkā given by travellers do not in the least tally with those of the Hindus. In my opi-
nion, this Laṅkā cannot be Ceylon, and he gives an indirect proof of it when he tries to derive the
name of Laṅkā from Skr. lavaṅga, « clove », and speaks of cloves as the article which the native of
Laṅkā delivered when bartering with foreign traders : the clove is not a produce of Ceylon, but of
Indonesia. The impossibility of locating Al-Bīrūnī's Laṅkā in Ceylon had already struck FERRAND,
who, on the strength of another passage (SACHAU, I, 310), thought that this Laṅkā, was Laṅga-bālūs,
the Nicobar Islands (Fe, 166; see « Necuveran »). ROUFFAER went further, and tried to establish
(Bijdragen tot de Taal-, Land en Volkenkunde van Nederlandsch-Indië, 77 [1921], 97-103, 360-365)
that, even in Indian texts, Laṅkā referred in most cases not to Ceylon, but to Ujung Tanah, alias
Johore, i. e. the south-eastern end of the Malay Peninsula. Although I would not suscribe to all
details in ROUFFAER's argument, and particularly do not agree with his identification of Lêngasuka
(in my opinion, the old name of Patani) with Laṅkā = Ujung Tanah (see « Lochac »), I concur with
him that Al-Bīrūnī's Laṅkā lay in Further India (cf. also Mi, 188-189; J. L. MOENS, Çrīvijaya, Yāva en
Kaṭāha, in Tijdschrift v. Ind. Taal-, Land- en Volkenkunde, Vol. LXXVII [1937], 397, 402, although,
as a rule, I do not share the views of the author).
The second in date of the mentions of dumb trade in Chinese texts refers to these same regions;
it is the one concerning the « kingdom » of 羅 刹 Lo-ch'a. Lo-ch'a (* Lâ-ṭṣ'at) is a transcription
of rākṣasa, « demon », and the name was given to the inhabitants, on account of their savage character,
either by Hinduized populations of Further India or by Chinese. The Chinese heard of Lo-ch'a in
connection with the mission of Ch'ang Chün to Ch'ih-t'u («Red Earth») in 607. In spite of GROENE-
VELDT and SCHLEGEL, who said that for a long time this was the designation of the Nicobar Islands, it
is out of the question to look for Lo-ch'a further west than the Malay Peninsula (cf. BEFEO, IV, 281,
400; LAUFER, in TP, 1915, 211; Y, II, 308; and see « Lochac »). The Lo-ch'a people had red (? red-