THE SOUTHERN GATEWAY. 491 



for voyagers from the Gulf of Papua to some Melanesian island, for colonists 

 coming out of Melanesia, to bring this Polynesian content of the Melanesian 

 speech. But the language record militates against this. As between dif- 

 ferent languages on the New Guinea coast we have found differences which 

 indicate secondary borrowing, but these readily reduce to well-marked foci 

 on the same coast; as to the material segregated at these foci, their char- 

 acter is clear: they are primary loans of the Polynesian. 



On the strength of such material as was then available from the Motu, 

 I had no hesitation in adopting that language as establishing a proof point 

 of the swarm of Polynesian migration through Torres Straits, the first point 

 fixing the course to which, from its ultimate destination, I have assigned 

 the designation of the Viti Stream. This newly accessible material but 

 confirms the former conclusion; instead of a single early station on that 

 course we now have a number and all confirmatory. 



The Torres Straits stations, with two exceptions, are all along the coast. 

 The two exceptions are Mabuiag and Miriam, which Dr. Ray assigns respec- 

 tively to the provinces of Australian and Papuan speech. Although the 

 vocabularies are the largest in our possession from the region, the material 

 which they have contributed to these studies is the least ; six vocables from 

 Mabuiag and three from Miriam are all that may be associated with the 

 Proto-Samoan. On the map these islands seem to lie in the fairway for 

 any voyage through the straits; it would seem that there the Polynesian 

 influence should be at its maximum. But in the conditions of Proto- 

 Samoan sailing these islands were remote. The Polynesians were sailors, 

 but they were not bigoted in their navigation. Hardy to risk the unknown 

 expanses of open sea when the sky lay empty before them, they preferred 

 the greater certainty of a coast to follow; any shore, even when it lay alee, 

 served to deflect their course. This is a part of their seamanship which we 

 have already had to consider when examining the courses through the 

 Solomons and the New Hebrides. It is this facility of coastwise voyaging 

 which has established for us so many points of Polynesian influence along 

 the New Guinea shore from the great gulf to the remote eastern islands. 



In conclusion we judge that these languages of New Guinea that have 

 been called Melanesian are susceptible of correlation with languages situ- 

 ated geographically in Melanesia ; that such correlation rests almost wholly 

 on the vocables here examined ; that this element common to Torres Straits 

 and Melanesia is common to Melanesia and to Polynesia. Therefore the 

 result of the investigation is not the establishment of a distinctively Mela- 

 nesian content in the languages limited by Motu and Nada, but the asso- 

 ciation of that content through the Melanesian with the earliest type of 

 the Polynesian. As the result of this added study we now find the southern 

 gateway out of Indonesia most satisfactorily established. 



