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

814   320. QUENLINFU

under a form which YULE writes « Kansan ». Although the name mentioned by Odoric very probably refers to Hsi-an-fu and to the province of Shàn-hsi, it is difficult to see why Odoric did not use either a Chinese name, or the ready made Kinjanfu of the Persians; I do not believe he was learned enough to leave out fu as being an administrative appellation and not really a part of the name. On the other hand, the Mss. differ widely here; but I do not attach much weight to the form « Casairon » given by Ms. C of Odoric and adopted by the most recent editor only in the title of the chapter (IFy, 483, 484).

Since KLAPROTH, Kinjanfu has been explained by 14 Jf f{f Ching-chao-fu. Ching-chao-fu

had already been in use in the T'ang dynasty; the Chin called it the   lu of Ching-chao-fu, and the
name was retained in the beginning of the Mongol dynasty. In 1262, the Mongols created a single province of Shàn-hsi and Ssû-ch'uan, the provincial seat being at Ching-chao; this name of Ching-

chao was changed to the lu (department) of   An-hsi in 1277. Ssû-ch'uan became a separate
province in 1286, and Shàn-hsi proper became the province of Shàn-hsi; in 1312, the An-hsi-lu became the ,4 it $ Fêng-yüan-lu. The names of the lu do not seem to have been in common use, at least among foreigners, and Polo, as well as Rasidu-'d-Din, always speaks of fu and chou. No better explanation than KLAPROTH'S has been proposed; in any case, the Ching-ch'êng[ ih]-fu of Bl, II, 496, never existed. Very likely, the name of Kinjanfu, derived from Ching-chao-fu, superseded in Central Asian use the old name of Ijumdan or Ijumudan during the Chin dynasty (on Iumdan, see the bibliography in Mi, 225 and 229). The only difficulty is about the n of -Ian-instead of the u which one would have expected; I do not believe that VON RICHTHOFEN can have heard it correctly at the end of the last century from a Chinese boy he met on the spot (cf. BEFEO, Iv, 771). And it is still more puzzling to find an -n also in Odoric's « Cansan » or whatever it be, without the final fu, if it means really the same name. As to the use of Kinjanfu in the Ming vocabularies to designate the province of Shàn-hsi, it is in agreement with Polo's use of the names of provincial cities for the provinces themselves. Even in Chinese, we occasionally find the province named after its metropolis; for instance Ching-chao-hsing[ ii]-shêng, An-hsi-hsing-shêng and Shàn-hsi-hsing-shêng all appear in Yung-lo ta-tien, 19417, 13 a, 15 b; 19418, 9 b; this may be due to a lax use of the terms in Mongolian.

Already identified with   fir_ J Chien-ning-fu by MARTINI in the 17th cent. (Novus Atlas

Sinensis, French ed., 153). The name of Chien-ning-fu goes back to the Sung; under the Mongol dynasty, Chien-ning was a lu . On the importance of Chien-ning at that time for the postal commu-