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

gives « Quian » in both cases (there is no name for the river at Ch'êng-tu in Z, still abridged in
that portion), and Z uses « Quiansui » for the Yang-tzŭ at Yang-chou. « Quian » occurs twice on
Fra Mauro's map (Zu, 37, 38; Hallberg, 424); and I think that the « Ch'ianfuy » of the Catalan Map,
which baffled Cordier's attempt at an identification (L'Extrême-Orient, 23), is to be read « Ch'iansuy »,
although it has become the name of a city on the river.
While such discrepancies are somewhat astonishing, it really seems as though Polo used two
forms of the name. « Quian » is of course Ch. 江 Chiang (> Kiang), « River », and is in itself a
common and sufficient designation of the Yang-tzŭ-chiang or Ta-chiang (« Great River »). On the
faith of Pauthier (Pa, 368) and Charignon (Ch, II, 194; III, 61), « Quiansui », restored as 江 水
Chiang-shui, has also been accepted as a normal and current name of the river (cf. Y, II, 40; RR,
425; B¹, 440; also Bl, II, 329). Chiang-shui literally means « river-water », and is listed as such in
Giles's Chinese-Engl. Dictionary; it may also be construed as meaning « water of the River »,
but it is far from evident that it was ever used as a real name and as a popular equivalent of Chiang
alone. Charignon says that Chiang-shui and 河 水 Ho-shui appear on ancient Chinese maps as
designations of the Yang-tzŭ and of the Huang-ho respectively, but he does not quote any authority.
Even if he were right on that point, such written mentions would not imply the use of the same names
in spoken Chinese, the only source from which Polo or his informants would have drawn. Never-
theless, I feel reluctant to look for a foreign final element in « Quiansui » like Turk. su, Mong. usu,
both meaning « water » and « river », notwithstanding the partly foreign nomenclature Polo uses in his
description of Western China. It may be after all that Chiang-shui was then a term of the dialect
of Ssŭ-ch'uan.
Odoric calls the Yang-tzŭ « Talay », which is Mong. talai, dalai, now used for « sea », « ocean »
(Y, II, 206; Wy, 468; Ch, III, 61, sees in Odoric's form a Ch. 大 海 ta-hai, « great sea », which is
out of the question); but talai, dalai, like the corresponding Uigh. talut, must also have meant
a « great river ». Once, Rašidu'd-Dîn (Bl, II, 329) uses a form which seems to be كنك جاى, rather
surprising, as it is apparently a Mongolo-Chinese hybrid « Qaān-Kāng », with kāng = Ch. chiang,
and the whole name would mean « Emperor river », or « Imperial river »; but the text may be corrupt.
Blochet's comment on a would-be confusion of the 揚 yang of Yang-tzŭ-chiang with 皇 huang,
« Emperor », « Imperial », is of no value. Elsewhere, Rašid writes كنك مران, Kāng-mōrān, « Kāng
river », or simply « Kāng » (Bl, II, 331, 333, 334); « Kāng-mōrān » perhaps refers in another passage
to the river of Canton (Bl, II, 450), but the case is not so clear as would appear from Blochet's
note. In two other passages, Blochet's edition of Rašid gives Pers. daryā, « sea » and « river »,
as another designation of the Yang-tzŭ, and there would be nothing abnormal in such a name;
but in both cases (Bl, II, 491, 536), the word daryā is an addition due to Blochet himself.
Fra Mauro has a curious notice (omitted by Zurla and Hallberg) on the river Phison which,
owing to the diversity of languages, was called « Scierno », « Gāges », « Phison », « Quinanfu »,
« Thalay ». « Quinanfu » is for « Quiansui », and « Thalay » is the name used by Odoric for the
Yang-tzŭ river. « Scierno » is Nicolò Conti's « Cernove », probably Lakhnaoti (or Gaur) in Bengal
cf. Hallberg, 463; Yule, Hobson-Jobson², 796; TP, 1933, 422-423), and the « Gāges » is of course
the Ganges. But I wonder whether there did not occur in Persian-speaking circles a contamination
between Kāng = Ch. Chiang (< Kiang), and the name Gāng of the Ganges, identical with Kāng