PHILOLOGY, 391 



this to the work of Dr. W. H. R. Rivers, of Cambridge University, 

 whose two volumes, ''The History of Melanesian Society," have but 

 lately issued from the press. This independent evidence confirms me 

 in my opinion that the Melanesian people present to us the picture of a 

 race who have but recently, and as yet but imperfectly, acquired the 

 employment of the consonants in the distinguishing of human speech 

 from the animal cry. The development of this hypothesis has been pro- 

 gressive through my work as already presented in the foregoing tabula- 

 tions. In my paper on "Root Reducibility in Polynesian" I set forth, 

 in preliminary statement, as among the objects of these researches, the 

 establishment of the principle that human consonantal speech evolves 

 from the non-human animal cry by the employment of consonantal mod- 

 ulants with coefficient value. This principle becomes more and more 

 clear as the work progresses and it is quite justifiable to express the 

 assurance that the completion of these studies will establish it beyond 

 peradventure. It therefore appears that this group of languages in the 

 South Pacific is to yield us the chapter of the evolution of human speech 

 seated in its proper place in the history of the evolution of man. 



At this point it is eminently proper to restate the hypothesis to the 

 determination of which the remainder of this work will, with more and 

 more singleness of attention, be directed. I repeat the statement as 

 made in "Root Reducibility in Polynesian," page 386: 



Now we shall sum up our notes upon this group of roots : 



ta the non-ego and the not-here reached by action outward, and prob- 

 ably downward. 



va refers to that which intervenes between the ego and the not-here. 



ga gives a limit of the extent of the not-here, a reduction of its distance 

 or degree. 



pa the beginning in the ego of action in the direction of the not-here. 



ka makes plain that the not-here is not the ego but something external 

 and therefore adversative. 



ma joins the ego and the not-here with a link. 



sa a general statement of the non-ego and the not-here. 



na a particular statement of the non-ego and the not-here. 



la a highly particularized statement of the non-ego and the not-here. 

 If we could master our problems of philology as we do those of algebra we 

 should see a common factor in each member of this table. On the one side a 

 is the greatest common divisor; on the other is that factor which we have 

 uniformly traced to be that which is non-ego, not-here, not-now, three which 

 are in essence one, the distal as contrasted with the proximal, the peripheral 

 in contradistinction to the central. We should further see that as this con- 

 sistent primary intonation of the voice was modulated by introductory 

 closures of the organs of speech we obtained certain limitations of definitions 

 of the peripheral sense of the primary vowel, and we might be led to regard the 

 initial consonants as in some sort coefficients and to make to each one the 

 provisional assignment of some germ of speech. Further to deal with this 

 series of consonantal meanings as coefficients of this and other vowels, having 

 the value of determinants of space and in such other senses as we may prove 

 them to possess, would require us to pass in review the whole of Polynesian 

 speech, the Samoan and its near kin and its more remote congeners in the 



