SOURCES OF MELANESIAN MATERIAL. 3 
many, France, and Great Britain have interrupted the tedium of guard 
service with the accurate determination of geographic positions. 
In the second line of knowledge, the intensive study of individual 
communities in their social life, Melanesia falls far below the other 
divisions. A beginning has been made; we have Prebendary Cod- 
rington’s somewhat diffuse work, Parkinson’s careful account of thirty 
years in St. George’s Channel of the Bismarck Archipelago, Rivers’s 
brilliant study of the social units of the New Hebrides, a few records 
of missionary endeavor in one of the straitest sects which may yield 
to wearisome search scanty glimpses of the life of Melanesian folk. 
From these sources we derive important information on the cultural 
history of the several peoples, but there is by no means sufficient 
information on which to base more than the most cursory com- 
parative study. We have still less information in the domain of 
anthropometry; the records are few and so scattered over the area 
that we are far from a conspectus; the most that we can derive from 
their comparison is the recognition of the possibility that more than 
one race is included within the designation of Melanesians. 
Until our knowledge of Melanesia has been brought to a higher 
stage, the chief reliance in our studies must rest upon the linguistic 
record. Of course this is not to be considered final in the determina- 
tion of race and affinity, but it is so much the best material available 
that we are justified in utilizing its data for the establishment of com- 
parative investigation. Yet even here our knowledge of Melanesia is 
of very unequal advancement. We find three stages clearly marked. 
In the first we have discovery records—imore or less scanty collections 
of words gathered by explorers; in the second stage we have works 
which purport to be dictionaries of a few languages; in the third are 
the essays which assume to state the problems of Melanesian philology 
and in some sort to solve them. 
Since our study of Melanesians must, as already set forth, rest 
most largely upon the speech record and the use which we make of it, 
it is essential that we pass under more detailed review these three 
several stages, in order that we may evaluate the records and estimate 
the importance which each may possess in our work. 
The first, the discovery record, need not detain us here. In this 
monograph we are to subject to intimate examination the discovery 
record of one new-found language, and we shall find it advantageous 
to include the general consideration with the particular examination 
of the discovery of the Sissano. We note, however, as an essential 
preliminary, that such works as Codrington’s “ Melanesian Languages”’ 
and Ray’s report on the languages of Torres Straits are at bottom 
discovery records. 
In the second, the vocabulary or dictionary stage, we are to find 
our best material, for each of these works purports to be such a gram- 
