GEOGRAPHY OF THE MIGRATIONS. 150 
suggest must essentially inhere in the problem that I have restricted 
my cultural chart to a very few elements and to such only as appear 
before us with any wealth of data. 
In the cultural chart (Chart I) the division by tint distinguishes a 
culture division which beyond any doubt represents two alien cultures 
with the least possible admixture of a subsisting Melanesian culture. 
Here we have the areas of the kava culture of the Polynesians and the 
sirih culture of the Indonesians. It is botanically possible that the 
area of the culture may represent an admixture of Polynesian upon a 
Melanesian base. In a few instances noted in the linguistic discus- 
sion of this theme we may admit the persistence of Melanesian kava 
culture, but in the main this kava culture is distinctively Polynesian. 
In the area of the sirih culture we see at present nothing which might 
suggest a subsisting Melanesian culture. The Indonesian source is 
beyond doubt, even though we find the culture extended far beyond 
the direct reach of Malayan voyages. ‘The sharpness of the boundary 
between the two cultures is more graphic than real. It will be seen 
that in the geography of Melanesia we are not dealing with land- 
masses of any considerable size, but with islands which fall loosely 
into groups where canoe-sailing is possible from any point of departure 
to the next landfall immediately in view, and these groups are gener- 
ally separated by such intervening spaces of empty sea that voyaging 
must be fortuitous to navigators lacking compass and chart. 
It is only at a very few points, therefore, that this sharpness of 
boundary as set forth on the chart needs consideration. In a nega- 
tive sense two of these considerable points are in the Admiralty 
Islands north of New Guinea and the region of Vanikoro-Ticopia 
north of the New Hebrides complex, these being the sole instances 
in which is found any concurrency of the sirih and kava customs. 
As already explained, the apparent lacuna between the two cultures 
in the region of the islands of the Polynesian Verge to the east of the 
Solomon Islands is based upon mere paucity of vocabulary material; 
the probability is that these islands fall within the kava culture area. 
The only important point of a sharply drawn boundary lies in the 
southern Solomons. Here is a narrow and readily voyaged strait; on 
the windward side of it the sirih culture is fully established, to leeward 
Guadalcanar and San Cristoval lie outside of both cultures. We 
have some grounds for looking upon this leeward group as having at 
one time had the kava culture which has now passed out of memory, 
that in the advance of the sirih culture it has been passed by. 
This particular area of the southern Solomons attracts our attention 
in the consideration of another specific problem. The two problems 
do not appear associable, except in geographical community, for one 
to a certain extent appears to postulate the absence of Polynesian 
influence and the other its presence. This second problem has to do 
