RAPANUI SOURCES AND VARIETY. 35 



these words lies in the a-i collocation; it is this vowel skeleton which 

 holds the meaning. 



Look now at our tables of English and French source and see what 

 the Rapanui men under their own instinct of speech have done with 

 their borrowings. An excellent illustration is pent, interesting because 

 we find it duplicated by penetuli of the same sense from the French 

 peinture. The word paint is on two accounts impracticable for Rapanui 

 enunciation ; it ends in a consonant ; it carries concurrent consonants ; if 

 dealt with by the foreigner intent upon fitting the English word for 

 island use by the method of parting the concurrence by a vowel of light 

 shade, the word would assume some such form as peniti. This would 

 come under the regimen of another rule of Polynesian speech, that of the 

 penult accent, and we should find that peniti is unrecognizably remote 

 from original paint. Governed by his own comprehension of that which 

 is permanent and dominant in every vocable, the Rapanui man seizes 

 upon the vowel which meets his ear; of the succeeding consonants 

 adopts that which is most lasting in his consonantal scheme, the nasal, 

 and rejects the mute. Therefore peni pictures to his eye the distinc- 

 tive determinant sound which paint makes upon his ear. 



Thus we are introduced to an important detail of the use of the con- 

 sonants. Not only are the Polynesians masters of far fewer consonants 

 than our needs require, but of those consonants which they do possess 

 the mastery is varied in degree. The tier of consonants which lies 

 nearest the vowels is that which alone can be said to be universally in 

 Polynesian possession, the palatal nasal ng, the lingual nasal n, the labial 

 nasal m. These three are almost constant; mutation inter se is rare, 

 and mutation in series (that is to say, m to other labials, and the like) 

 is almost wholly restricted to the possibility of the l-n and n-l mutation. 

 This exception, again, is genetically valuable, for it points the way to a 

 line whereby the vowel in evolution through the channel of the liquid 

 may attain to consonant figure. Our studies of Polynesian etymologies 

 show us in fact the tangi illustration shortly heretofore employed 

 offers a full exposition that t is impermanent, it may become k by an 

 extraordinary shift to the palatal series, and in its own series it may 

 become /, s, h, or vanish entirely. Therefore we are led to the conclu- 

 sion that in dealing with concurrent consonants in its borrowings the 

 Polynesian selects that of each two which is the older and better estab- 

 lished in his own speech. 



This we find again instanced in nira, a selection of the liquid over the 

 mute in the dl of needle. 



In a considerable group of these borrowed words we have to do with 5 

 concurrent with some other consonant, either in the preceding or the 

 succeeding position. Here the resultant is conditioned by the fact 

 that the sibilant is impossible to the Polynesian in general, the Samoan 

 being the chief exception, and commonly is represented by an aspiration 



