XIX POLYNESIANS AND THEIR MIGRATIONS 401 



knowledge of the Polynesian languages, is well qualified to 

 give an opinion on this matter, says it is quite the contrary. 

 In his Polynesicm Reminiscences, p. 402, he observes : " It 

 is, however, remarkable that in all these many instances 

 of authenticated driftings, the course of the drifted canoes 

 has been from east to west before the prevailing trade 

 winds, and not from west to east before the westerly 

 winds ; " during the prevalence of which he tells us the 

 natives do not usually venture out on fishing or travelling 

 expeditions.^ 



In this case, too, the corroborative proof by language 

 completely fails, for though there is an undoubted Malay 

 element in the Polynesian language, it is an element 

 derived from the civilized Malay and Javanese tongues, 

 not from those of the Moluccas, which are totally distinct. 

 It is to be noted also that this Malay element in the 

 language has all the character of a recent introduction, 

 since the Malay words are hardly changed, except by the 

 phonetic character of the language which has received them. 



But even if the Malay formed a much larger portion of 

 the Polynesian language than it does, this would not 

 prove a community of race, unless the physical characters 

 also in some degree corresponded. It is here that we find 

 an absolute defect of all evidence bearing upon the point 

 in question — the similarity of the Polynesians to any race 

 speaking the Malay language. Almost the only evidence 

 adduced by M. Quatrefages is as to the similarity of the 

 brown race of Timor to those of Polynesia. But the 

 Timorese are not Malays at all ; they belong to that 

 curious race which has close affinities to the Papuan in all 



^ Later and more extensive inquiries have shown that accidental 

 drifting of canoes from north or north-west to south-east does often 

 occur, especially in the region of the northern tropic, where violent 

 storms and hurricanes are not unfrequent, and it is when unexpectedly- 

 caught by such storms that canoes get carried far away from their course, 

 and occasionally drifted to distant and unknown islands. Such was 

 the immense double canoe with men, women, and children, which over 

 a hundred and fifty years ago reached Rotuma from Tarawah in the 

 Gilbert Islands. This was a journey of over a thousand miles in a 

 direction south by east. Numerous other cases have been recorded in 

 various Pacific Islands of canoes carried by storms from the north-west ; 

 and most anthropologists now believe tliat the Mahori race originally 

 came from some part of the Asiatic continent or its adjacent islands. 



VOL. I. D D 



