In the meantime, G. FERRAND had made a scathing criticism of GRANDIDIER's views first in Les Musulmans à Madagascar (Paris, 1893, II, 83-90), and again in Trois étymologies arabicomalgaches (Mém. Soc. Ling., mil, 413-422) ; it is still repeated by him in 1928 in EI (s. v. Madagascar). FERRAND concluded that Madagascar must be l,;a. Madagas-bar, « the Land of the Madagas (or Matakasi, Madegase) »; this was more or less accepted by CORDIER in Y, II, 414. In Bull. Acad. malgache, 1932, 15-18, and 1933, xxvi, G. MONDAIN, without putting too much confidence in a form « Madangaskara » found by Govr JULIEN in an « old » Arabico-Malagasy Ms., starts from Polo's «Madagascar » and explains it as Malagasy « *Madagasikari», «riche en nacre sacrée » or « béni par la possession de la nacre sacrée ».
And yet I have no doubt that GRANDIDIER was right, and that, in Marco Polo, neither the description nor the name applies to Madagascar. The question of distance from Socotra, in spite of BENEDETTO, has no weight. Polo speaks here by hearsay, narrating what he was told when traveiling from India to Ormuz, and his mentions of distances are often fanciful even when he speaks of places he has travelled to. More important is the fact that he places the would-be Madagascar between Socotra and Zanzibar, which only fits Mogadiso. It is true that he says the would-be Madagascar is an island, but then he says the same of Zanzibar, and all agree that his text does not apply to the small island of Zanzibar, but to the mainland, when he gives it a compass of some 2 000 miles; I side against FERRAND with GRANDIDIER and see here the result of the loose use of j"azirah in Arabic, meaning « island » and « peninsula » (the same ambiguity prevails in Skr. with dvipa, and in Chinese with chou). Moreover Polo does not seem to have had any clear idea about Eastern Africa, and has probably supposed that the kingdoms of the Eastern African coast were so many islands. Then there is the fact that Arabic texts know of Madagascar in Polo's time, but give it quite different names, and that neither « Madaga"s-bar », nor even « Madagas », has even been met with in one of them; Madagas-bar is FERRAND'S arbitrary creation, on the type of Zanzibar and Malabar. Above all, there is the fact, acknowledged by all commentators, that Polo's description applies correctly to Mogadiso, and cannot fit Madagascar. Finally, it is owing to the idea of Polo mentioning Madagascar, coupled with the identification of the ruh with the fossil Æpiornis (see « ruc »), that Madagascar has been given as the equivalent of the K'un-lun-ts'eng-ch'i of Chou Ch'ü-fei (1178) and of Chao Ju-kua (1225).
But there is something more. Most former commentators have discussed the matter as if we had two names in Polo's texts, when we have always only one. The text in F has only « Madeigascar », the preliminary list of chapters has only « Mogelasio » (for the value of this preliminary list, see « Sumatra »). Even with this last single mention, FERRAND admitted that « Mogelasio » was undoubtedly Mogadiso, while keeping « Madeigascar » alongside it as meaning Madagascar. But the different stages of alteration are shown by « Magastar » of VB and R, « Macdesca star » of K, «Machdoscastar» of KI, and now that we have «Mogdaxo» in Z, we can be sure that «Madeigascar», a name which never existed, is simply, as GRANDIDIER said, altered from the form which Polo adopted for Mogadixo.
This last name is spelt,. in Arabic, and this is generally read Magdasau, but the Swahili form is L;, ..U. or ,x. Mogadiso (FERRAND, Trois étymologies, 414), and FERRAND himself has
said that the traditional vocalization of the Arabic form is wrong for Mogadiso (cf. Fe, 321; also, for
IIIFカラー高解像度
