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

understanding of « d'Ely »; the confusion with [Rāman]tali accepted in Pe, 251, is most unlikely.
R has already « Dely » in Polo, perhaps under the influence of the « Deli » he gives in Conti's text;
cf. Hallberg, 196-197.

The Mussulman texts write هلي Hīlī (= Hēli); there is no basis for the transcription Haïli,
frequently adopted. We find in the early 15th cent. 昔 立 Hsieh-li, Hēli, in the Chinese map of
Mussulman origin reproduced by Phillips in JNCB, xix, 223; it is correctly placed south of
Mangalore. Rockhill (TP, 1915, 453) has also seen Eli, Heli, in the 下 里 Hsia-li, *Hali, of
the Tao-i chih-lio of 1349-1350; he may be right, but then the position given in the text, between
Quilon and Calicut, is wrong; this mention would be interesting as showing an initial h- independently
from Mussulman cartography. Cf. also TP, 1933, 289, but taking into account that, geographically,
Quilon may be said to border on Cochin, not on Eli; Fei Hsin repeats here the error of Tao-i chih-lio;
for other mentions of Hsia-li in Fei Hsin's work, cf. TP, 1915, 448, 462.

Hīlī is mentioned in Fe, 281, 526, 540; but p. 281, a text from Ibn Al-Baytār († 1248) refers
to « Sindāpūr, in the territory of Hīlī ». Ferrand makes no comment, but this localization is
impossible, as « Sindāpūr » is the old name of Goa (cf. Hobson-Jobson², 837-838). Incidentally
I may remark that there is no ground for transcribing سنداور or سنداور as « Sindāpūr ». The real
form is Šindābūr (and Šindābūr), or even Šandabūr (and Šandabūr). The Catalan Map has « Chin-
tabor », the Medici « Cintabor ». Yule has already supposed an initial element « Chandāo ».
This is confirmed by the Chinese map of the 15th cent. where we find a 瞻 打 兒 兒 Ch'an-ta-
wa-êrh, *Candawar, not identified by Phillips (JNCB, xix, 63), but certainly identical with the
so-called « Sindāpūr ».

219. ERCOLIN

archoline, arcoline L arculini VL ercolin, erculin F
archolini VB choccholini TA¹, TA³ ercolini, erculini Z
arcolini L, VA; R cucullini G herculini LT, P

Three times, in his notices on the Kingdom of «Concis, on the Land of Darkness, and on Russia
(cf. Vol. 1, 472, 473, 474), Polo gives an identical list of furs, twice in the order «sables, ermines,
squirrels, ercolins and [black] foxes », and once in the order « sables, ermines, ercolins, squirrels,
and foxes ». In the original French text, the word here rendered « squirrel » is vair; although
more or less obsolete in French and English, it might have been retained, as it has been by Yule,
Ricci-Ross, and, under the Italian form vaio, by Benedetto. As to the « ercolins », their name
has not been traced elsewhere hitherto.

« Vair » (< Lat. varius, « variegated », > Fr. and Engl. vair, It. vaio, pl. vai) was used in
mediaeval French as the designation of a particular colour of eyes, of the coat of various animals,
but, above all, of certain kinds of squirrels and their furs. There are three main names of squirrels
n mediaeval texts : « vair », « gris », and « écureuil » (ordinary squirrel, Sciurus Vulgaris). The