Ellis held that the Polynesians had been conveyed from America to Oceania by winds and currents, but this hypothesis has had scarcely any adherents. It is in too direct contradiction with all the physical, philological, and social characters, which refer the Polynesians to the Malay races as strongly as they separate them from the Americans. Dumont d’Urville has proposed a theory which, at first sight, is more satisfactory, and still has a few supporters. In his opinion, Polynesia is the remains of a great continent which was originally connected with Asia. This land sank after some geological revolutions; the sea covered the plains and hills, the highest summits only being now visible and forming the present archipelago. The Polynesians are the descendants of those who survived the catastrophe. This hypothesis has the advantage of preserving those relations which were broken by that of Ellis. And, curious to relate, it agrees with the tradition of the deluge as preserved by the Tahitians. They say that the great inundation happened without either rain or tempest. It was the sea which rose and covered the whole earth with the exception of a flat rock where one man and a woman took refuge. We might say that there was nothing in this account but a mistake which is easily understood. The sea never rises, but the land may sink, and other people besides the Tahitians have been deceived. Nevertheless, we cannot accept the theory of Dumont d’Urville. It is in contradiction to the zoological facts so thoroughly investigated by Darwin and Dana. If some of the atolls of Oceania shew signs of subsidence, a great number of islands offer incontestable proofs of upheaval, and Tahiti itself is one of the latter. But the most serious argument which can be brought against d’Urville is derived from the inhabitants themselves. If travellers agree upon one point, it is that from the Sandwich Islands to New Zealand, from the Tonga Islands to Easter Island, all the Polynesians belong to the same race, and speak the same language with mere variations of dialect. Now the Polynesian area, the limits of which I have just pointed out, is of greater extent than the whole of Asia. What would an _Asiatic Polynesia_ be like, if that continent were to sink beneath the waters and leave only the summits of its mountains visible, where some representatives of the present inhabitants might take refuge? Is it not at once evident that each archipelago, and often each island, would have its own race and language? The considerations drawn from the identity of populations and languages in Polynesia are of themselves sufficient to justify the assertion that all the Islanders have a common origin; and consequently, that, starting from some unknown point, they have, in their advance from archipelago to archipelago, peopled by degrees the maritime world in which we find them.