EARLY ARABIA AND OCEANIA. 465 



if at all, on linguistic grounds, that is, by proving that the primitive 

 Oceanic was related, or belonged to, some known Asiatic family. The 

 piesent paper is to state the view that they carried it into the island 

 world from Early Arabia; and that the proof, the only available con- 

 vincing proof of this is the linguistic proof that it was related, or 

 belonged to, the Semitic family, being, not derived from, but a sister 

 »<> tlie Himyaritic, Arabic, Ethiopic, Assyrian, Aramaic, Hebrew, and 

 Phoenician, the existing Oceanic dialects, as the Malagasy, Malaysian, 

 Poljmesian, and Melanesian, being thus cousins to, not derived from, 

 the existing Semitic dialects, as the Modern Himyaritic, Modern 

 Syriac, and Modern Arabic; so that ''the 50,000,000 islanders must 

 be added, to the number hitherto classified as Semitic speakers." 



On looking at a map embracing the southern coast of Asia and 

 tliis Indo-Paciiic island region, it is not surprising that the south- 

 eastern, or Indo-Chinese, peninsula, first suggested itself, rather than 

 Arabia, the south-western peninsula of Asia, as the starting point from 

 which the Oceanic entered into and was diffused over the whole island 

 world. Aud there are ^ome whose minds have become so preoccupied 

 with this idea that they reject the Arabian view because it is opposed 

 to it. This, however, is not scientific. For, that the Indo-Chinese 

 peninsula was the starting point of the island speech is an unproved 

 li'V'liothesis. And it has the whole vast central width of the Indian 

 Ocean, lying between Malaysia and Madagascar, against it. Sir 

 Joseph Banks, writing in 1771 (see his Journal, published in 1896, pp. 

 424-26^, says:— 



■' From this similitude of language between the inhabitants of 

 the Eastern Indies (Malaysia) and the islands in the South Sea, I 

 should have ventured to conjecture much did not Madagascar inter- 

 fei-e : and how any communication can ever have been carried between 

 Madagascar and Java to make the bi-own, long-haired people of the 

 latter speak a language similar to that of the black, woolly-headed 

 natives of the other, is, I confess, far beyond my comprehension : 

 unless the Egyptian learning running in tAvo courses — one throughi 

 Africa, the other through Asia — might introduce the same words, and,, 

 what is still more probalile, numerical terms into the language of 

 people who had never had communication with each other. But this 

 point, requiring a depth of knowledge of antiquities, I must leave to 

 antiquarians to discuss." 



The Indo-Chinese hypothesis lias failed to account for the facts, 

 and to justify its existence, by not being able to show how, oi- why, 

 or that, the Malagasy crossed the Indian Ocean from Malaysia ; or 

 that the Malagasy is derived from the Malay or Malaysian, though all 

 are admittedly of the same origin. The opposite view, that Mada- 

 gascar was the starting point, would, so far as the inter-island 

 language diffusion is concerned, be just as tenable; and, so far as the 

 negro element of blood in the Oceanic speakers is concerned, even 

 more tenable: for thus at least some account would be given of the 

 fact of the negroid element being in Melanesia as in Madagascar. (See 

 various opinions in "Man, Past and Present," Cambridge, 1899, pp. 

 248-256). In the suggestion of Banks, considering the scanty data he 

 had before him, there is something of the intuition of genius. It 

 points to the same quarter for the starting point of the Oceanic, in 

 considei-ation of the facts, as the Arabian view. The latter view has 



2e 



