168 SISSANO. 
of Papua. In “The Polynesian Wanderings,’ working over the 
material afresh and with a broader scope, I felt justified in proposing 
a general eastward migration stream from southern Indonesia and 
along the south coast of New Guinea through Torres Strait. We now 
proceed to the general examination of this critical region relative to 
the two migration streams on either side of New Guinea. 
It has been pointed out that just at the spot where we should expect 
to find the most complete evidence of reversal of language movement, 
at the extreme tip of New Guinea, we find the curves of affiliation at 
their lowest degree. North of that point on the east coast, west thereof 
on the south coast, we find considerable areas in each of which affilia- 
tion is conspicuous, yet we lack the evidence of connection between these 
diverse areas. Furthermore, assuming the rounding of the cape, 
there is nothing on these charts which might establish such rounding 
as directed westward from the east coast rather than northward from 
the south coast; there is quite as much argument on one side as on the 
other. Each extremity of this critical area is disjunct from the migra- 
tion stream with which it is sought to connect it. But there is observ- 
able here a difference in the conditions of the problem. ‘The Colling- 
wood Bay region of affiliation is separated from the Vitiaz Strait by 
some 3 degrees of latitude; the Motu region is parted from the nearest 
point of Indonesia in which we have record of Polynesian language 
material by all of 10 degrees of longitude. Yet here a difference 
exists. In the 3 degrees between Vitiaz Strait and the Kiviri speech 
the land has been visited; at several points the languages have been 
collected at least to the same extent as in the region for which these 
lines have been traced, and at every such datum point the Polynesian 
element is wanting. In the greater distance between the Arafura 
Sea and Motu the land has been almost unvisited; there is a long 
stretch of coast for which we have no authority to postulate that it 
does not contain Polynesian survivals. ‘Thus far we have done no 
more than advance the argument that it is quite as possible for the 
folk movement to have progressed easterly around the southeastern 
cape as westerly, and we must add at this point our knowledge that 
all Polynesian migration is instinct with motion away from the west. 
Now consider the quality of the affiliation of these Polynesian 
survivals. On the whole of the east coast the quality is low; it calls 
for no little art to discover in common speech the persistence of a Poly- 
nesian stem; and of the survivals along the north coast of New Guinea, 
with which it is sought to link this movement, the quality is equally 
low. On the other hand, the quality reaches its highest to the west. 
In Motu and in Sinaugoro and Galoma exist many words which would 
pass muster for Samoan current to-day. It seems impossible to con- 
sider a speech affiliation of low quality arising from a contaminated 
source as capable of recovering its purity the farther it travels from 
