158 SISSANO. 
antiquity as is possible to delicate materials in lands where humidity 
and mildew work together to limit durability, for in these islands we 
can infer the loom only from the web. But in Nukumanu and 
Iiuaniua, as well as in other islands of the Polynesian Verge, we 
encounter a mass of legend whose interpretation looks toward some 
degree of communication with the equatorial islands by way of Kapi- 
ngamarangi. Upon this point attention is directed to the investiga- 
tions of Thilenius (Ethnographische Ergebnisse aus Melanesien, I 
Theil, Die Polynesischen Inseln an der Ostgrenze Melanesiens, Nova 
Acta, Halle, 1902). 
Before we enter upon the detailed examination of the general track 
or tracks of speech movement out of Indonesia through Melanesia 
into Polynesia we must devote particular attention to a critical area 
represented by New Guinea. ‘This great island interposes a mass of 
land in the general sweep of folk migration out of the Malay seas 
which is to be considered not so much an obstacle as a conduit. In 
the modern navigation of vessels well found and equipped with instru- 
ments of mathematical precision for their direction, New Guinea is 
an obstacle to the seaman, a region of channels tortuous through a 
tangle of hidden-dangers, a land-mass to be shunned. ‘To the primitive 
voyager in his crazy canoe, following the line of coast from headland 
to headland, completely lost when driven offshore, such a land-mass 
is a welcome aid to his navigation. Ignorant of any haven to which 
his wanderings may bring him, he is content to coast in the stiller 
waters and to follow the leadings of the shore always within his view. 
To the Proto-Polynesians under the impulse of flight out of Indonesia 
before the better-armed Malayans advancing from the west, New 
Guinea has projected itself as a wedge deep into the channel of their 
escape. In the nature of this flight there is naught which might pre- 
dispose to a course on one side of New Guinea or the other. Those 
who put to sea from the northern islands of the Malay Archipelago 
would reach the New Guinea coast at various points on its northern 
shore and would be led generally eastward under the orientation with 
which the whole course of their migration is instinct. Other bodies 
of wanderers coasting along the southern tier of Malayan islands 
would reach the New Guinea coast along the Arafura Sea and would 
be led generally eastward on their flight. Those whose flight took 
its departure from islands centrally situated in Indonesia, e. g., Celebes, 
on reaching the nearest New Guinea coast might be led in either 
direction, as determined by some condition quite narrowly local. 
W hen first encountered in the line of flight the New Guinea wedge 
produces but a slight deflection, yet in the middle course the diver- 
gence of the two streams is great. In my judgment they did not 
reu nite until many weary leagues of sea had been passed and the 
severed fleets reassembled first in Nuclear Polynesia. It is upon this 
