152 SISSANO. 
culture elements, and it may prove that this extreme southern region 
will fall into a group of its own. It is equally clear that in southeast 
New Guinea and in the western Bismarck Archipelago we have a 
culture which differs from that of the Buka. On the north coast of 
New Guinea we find an interlacing of cultures, some of which may 
be assignable to Melanesian. In the southeast Ray has set forth a 
dual system of Melanesian and Papuan. ‘This is good as far as it 
goes, but his Melanesian of British New Guinea will need further 
study before we may feel sure of its association with the Melanesian 
of other regions. 
Thus it will be seen that we are not in position to postulate any- 
thing like a consistent Melanesian culture upon which has been over- 
laid the alien cultures of Polynesia and Indonesia respectively. We 
look rather for several somewhat discrete cultures in Melanesia as 
persisting under the foreign veneer and at times protruding through 
the overlying mass. 
In like manner, as there is variety in the subsisting base of Melane- 
sian culture, so is there diversity in the application of foreign influence. 
In general it is probably safe to characterize the application of 
Polynesian culture as direct, the product of personal contact during 
the period of the migrations; and to characterize the movement of 
the advance of Indonesian culture as a movement of convection 
where new cultural phenomena have indirectly extended in advance 
of the zone of personal contact of Melanesians with Indonesians and 
have reached new ground by the gradual spread of new material and 
moral objects. We should, then, expect to find in the Melanesian 
area at least these possibilities: one or more regions in which a Mela- 
nesian culture remains practically unmodified by alien accretion; 
regions in which Melanesian cultures have been modified by Polyne- 
sian culture. Such would be the condition throughout Melanesia 
after the Polynesian migration era and before the succeeding era of 
Indonesian influence. When the Indonesian culture followed more 
slowly over a Melanesia thus conditioned we should look for three 
possibilities—a Melanesian culture still unmodified by either culture, 
a Melanesia modified by one or the other culture, a Melanesia first 
modified by Polynesian culture and then by the Indonesian and show- 
ing various degrees of intermixture and assimilation of the three 
elements. 
Such a subdivision is natural in any hypothesis. It is really a 
statement of the problem far more simple than would arise in actual 
study if we had all the data upon which to rest such an investigation. 
On this plan we should be dealing with at least five secondary stages 
of no less than three Melanesian bases, sufficient to show that the 
subdivision of Melanesia by culture areas would be a problem far 
from simple. It is by reason of this intricacy which I have sought to 
