Such being the case, the primitive home of those speakers of the "ocean language" may be supposed to have been somewhere in the Indo-Malayan or Austro-Malayan regions, or, perhaps, to speak more correctly, in the ancient extensions of the Asiatic and Austral continents which they represent.
At any rate, the dispersal of the primitive speakers of the ocean stock-language must have been long after the migration of the Australians, and still longer after that of the Tasmanians.
It seems, however, not a little remarkable that the migrations of the offshoot, which is now represented by the Melanesians, should have extended from New Guinea, or perhaps from a point further west, around and beyond but not touching Australia.
Compared with the sea distances which must have been passed over (since the common stock-language proves that they were acquainted with canoes) before reaching Fiji, the stretches of sea between Timor and Australia, and New Guinea and Australia, must have been comparatively insignificant.
At any rate it would seem that Torres Strait separated the Papuans from the Australians almost as effectually as Bass Strait separated the latter from the Tasmanians.
The Melanesians occupy a vast insular extent, touching New Guinea at the one end and Fiji at the other, and probably represent the older race on which the Papuans have intruded.
It seems not unreasonable to consider these facts as indicating migrations of three branches of mankind in successive stages of ethnical development and culture.
If the Australians migrated from the north-west by way of New Guinea, as I have suggested, it may be that they brought with them some elements of language common to the ancient oceanic stock-language, to crop out here and there in the Australian speech as words having a resemblance to Malay.
But, it appears to me, there must be very grave doubts as to the Malayan element in the Australian aborigines as formulated by Mr. Mathew.
