6 SISSANO. 
adjust the speech of the people under his charge to the establishment of 
his speculative theory that all the people of the island area, denominated 
by him the Oceanic race, derive from some pre-Mosaic Semitic stock. 
Prebendary Codrington is our standard authority for southern 
Melanesia. His work falls within the first class by reason of the fact 
that he provides more or less extended notes upon the grammar of 
34 languages between the Loyalties and the southern Solomons. In 
his vocabularies, however, he sets his work distinctly within the third 
class; he presents in tabular form for convenience in comparison a 
series of 70 vocables in 43 languages. Such comparison is interesting 
as far as it goes, but it is clear that the material is far too scanty to 
serve as the foundation for a valid theory of the interrelations of the 
languages thus briefly noted. 
In the linguistic report of the Cambridge Torres Straits Expedi- 
tion Mr. Ray has pursued the same method. For the Mabuiag and 
the Miriam he records vocabularies of several hundred items; for the 
languages of Cape York Peninsula his record falls into a table of 35 
vocables in 12 languages; from the coasts of British New Guinea, 
now Officially styled Papua, he extracts languages of two classes 
which he designates Papuan and Melanesian. He subjects the 
material in each case to tabulation of a list of the same 154 vocables, 
which he treats with varying fullness of record— 46 languages in the 
Papuan class, 39 in the Melanesian. 
Captain Friederici also combines exploratory and comparative 
work in his study of the languages of the Bismarck Archipelago 
and adjacent coast of New Guinea. He has recognized that the 
tabular method has the advantage of facilitating ocular examination 
of material, but that the very mechanical system which produces this 
advantage sacrifices the intimate detail which is necessary to a proper 
comprehension of the material involved. The three authorities ex- 
hibit a growth in method as this difficulty was recognized. Codring- 
ton provides a rigid table, Ray a series of tables with footnotes of the 
utmost terseness. Friederici has hit upon a method of synoptical 
tables which adds greatly to the logical value of the information 
therein presented. Three columns he assigns to Indonesia, generically 
to the Bahasa Tanah and Alfuros of Ceram, of Buru, and of north- 
eastern Celebes respectively; next, and centrally situated, comes the 
column dealing with the speech of his particular study, that of the 
Barriai group in western New Pomerania; three similarly generic 
columns are used to carry the investigation still further—western 
Papuo-Melanesians, Solomons, and New Hebrides. I employ the 
genus suggestion in describing this extremely flexible system because 
in each column it is possible to include material from specific lan- 
guages pertinent thereto. The notes succeeding these tables are as 
complete as it has been possible for the author to make them, in many 
