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

346. SUMATRA   839

gives ;~L . Sumutrah (var. ~h; w Sumutrah). As the name of the whole island, umutrah occurs in Ibn Maid (1462 and 1490) and in Sidi `Ali Celebi (1554); cf. Fe, 734, and JA, 1917, i, 334-335. In 1595, the Ain-i-Akbari have the abnormal spelling 51.L sdmatrdi or sdmutrdi, «of Sàmatrà (or Sàmutrà ?) »; cf. Hobson-Jobson2, 867; Fe, 545.

The first dated native mention is in the form « Samudra », a state named along with «Lamuri» in the Neigarakrëtdgama of 1365 (cf. Fe, 652).

A king of San-fo-ch'i (Palembang-Jambi) who sent an embassy to China in 1017 is called

iTibut   `` 'jfs Hsia-ch'ih-su-mu-ch'a-p'u-mi, which FERRAND restored first into Haji

Sumutabhûmi = Ha)i Sumutrabhûmi, and, later on, into Haji Sumatrabhûmi, interpreting it as

« King of the Land of Sumatra ». This shows, according to FERRAND, that, already in 1017, when

there was no question of the state of « Samudra » east of Achin, the name « Sumatra » was used as a designation of the whole island (JA, 1917, I, 331-335; 1919, I, 277-278; 1922, II, 19-20). But

~ 1 wu is *mivat; in spite of all of FERRAND'S later argumentation, the normal restitution is Ha)i

Sumutabhûmi, or, with the cerebral value of Malay dentals, Ha)i Sumutrabhûmi. There is still nevertheless a certain element of uncertainty, as we have here to accept p'u with its old value as

beginning with an aspirate sonant be-, while the same character p'u, in 1225, represents, according to FERRAND, the Indonesian title pu or mpu, Cam po, and, if so, would seem to be then used, as it was a little later under the Mongol dynasty, to represent a surd initial (cf. JA, 1922, II, 9; but the same p'u transcribes bu in other names of the same work).

Granting the near certainty of FERRAND's restoration of the title as Ha)i Sumutrabhûmi, his conclusion that it implies the use of the name of Sumutra or Sumatra as a designation of the whole

island of Sumatra does not necessarily follow. The current opinion is that Sumutra, Sumatra,

represents etymologically Skr. samudra, « ocean ». Starting from that etymology, ROUFFAER proposed to see in the king of 1017 a king of Tumasik or Témasik, i. e. Singapore, as Témasik

or Tumasik is an infixed form of Malay tasik, « sea » (cf. Bijdragen... van Ned.-Indië, Vol. 77 [1921],

75; Fe, JA, 1922, II, 19-20). I am not prepared to accept ROUFFAER'S view. A king of San-fo-ch'i must be a king of San-fo-ch'i (Palembang-Jambi), not of Singapore. But it is enough that the

suzerainty of that king should extend, or claim to extend, over the whole island and a great part of the Malay Peninsula to permit him to take the high-sounding title of « King of the Ocean Lands »; the title does not necessarily imply a general name for the island of Sumatra, nor has it anything to do with the particular state of « Sumatra » in the north-western part of the island (but cf. « Samudradvipa » in a Skr. text not later than the 11th cent., quoted by S. Levi in JA, 1918, I, 85).

It is true that FERRAND denies any connection between Skr. samudra and the name of the kingdom or of the island of « Sumatra ». His reasons are that all islands being in the ocean, « ocean »

cannot become the specific name of any of them, and also that all Chinese transcriptions show su

for the first syllable and ma for the second (JA, 1919, I, 278; 1922, II, 19-20). But he does not offer any explanation of « Sumatra ». COEDÉS has already refuted the first argument (BEFEO,

xxiii, 469-470). As to the second, FERRAND forgets that the Chinese transcriptions are not alone, and that, for a kingdom in Malay territory, the form Samudra of the Net garakrëtdgama of 1365 cannot be tacitly thrown overboard. Moreover, in the West, we seem to find an -a- in the first syllable in Conti, in Fra Mauro and in Girolamo da Santo Stefano, and we have it certainly in the

15.