INDONESIAN ANNOTATIONS ON THE VOCABULARY. 143 
Tangled as the problem may appear at the first and even at the 
second view, the clue may be threaded out. We find the kava cus- 
tom in the possession of the people at present the sole and slightly, 
if at all, contaminated inhabitants of Polynesia. We have been able 
to trace through linguistic survivals the tracks of their migration; we 
have established their ancient occupancy of Indonesia, their traverse 
through Melanesia, their present possession of Polynesia. In Poly- 
nesia we find the complete, in Melanesia the partial, sway of the kava 
custom; in Indonesia the complete absence thereof and the equally 
complete dominance of sirih-chewing. In the treatment of this 
material Dr. Rivers shows that it is impossible to account for kava as 
introduced from Polynesia into Melanesia, and in this decision I am 
wholly at one with him upon the grounds which he adduces and upon 
others which have not come within the scope of his investigation. In 
general he inclines to regard the origin of kava as Melanesian in some 
sort, as there assumed by the wandering Polynesians and transported 
along their eastward course. In this direction he is led most strongly 
by two observed factors—the commonly ritual and religious employ- 
ment of kava in Melanesia, its complete absence in Indonesia. The 
latter of these will be included in our further discussion of the problem. 
The former seems to me to clear itself up. If we look upon primal 
Melanesia as inhabited by peoples ignorant of the kava custom, if 
we introduce thereto in the migration movement the greatly superior 
Polynesian culture carrying the kava custom, we should look to find 
the Melanesians adopting somewhat timorously, on the established 
principle of omne ignotum pro magnifico, this distinctive custom of 
the strange and conquering race. It would then be natural to find 
this custom surrounded by evidences of the respect which the superior 
Polynesians had won for themselves and all that was theirs; we should 
find the foreign custom restricted to chiefs, to religious ceremonies, to 
the mysteries of the secret societies; and these are precisely the 
Melanesian restrictions upon kava. 
The assignment of a strictly Melanesian source of kava is not 
sufficient to account for the spread of that culture through the Pacific. 
I have set forth two migration swarms out of Indonesia into the 
Pacific—the Proto-Samoan the elder, the junior Tongafiti, separated 
within the mixing area of Nuclear Polynesia by a considerable lapse 
of time, and I have made it clear that while the Proto-Samoan course 
lay along the island chains of Melanesia in two somewhat distinct 
streams, the Samoa stream up the wind and the Viti stream to lee- 
ward, the linguistic record makes it plain that the Tongafiti course 
lay entirely outside of Melanesia. In Nuclear Polynesia we have 
had considerable success in distinguishing these two migrations, and 
in this distinction we have found each to be in possession of the kava 
custom. It would be impossible to assign a Melanesian origin to the 
