RAPANUI SOURCES AND VARIETY. 45 



Following the system employed in the study of the Proto-Samoan 

 variants- we may assemble these items into classes. 



We remove from consideration those items in which our dictionary 

 material is insufficient to form a proper basis for comparison. These 

 are: 



garara gorigori guha 



Similarly we must strike out those items in which dictionary error 

 is recognizable. In this series are: 



hakura henua 2 hope puapua 



Of the residue after this elimination we find two interesting groups. 

 In the former the Rapanui offers a more primitive sense than is encount- 

 ered in the Maori. It seems closer to the sources of distribution whence 

 the two languages have moved. This is the list: 



eva ragaraga reva titiri tua 2 



kauihaga reke 



On the other hand we find a slightly longer list of items in which the 

 Rapanui is employed in a specific sense where the Maori has the closer 

 approximation to the primitive signification so far as we feel justified in 

 establishing such sense. These are: 



ariga maki reva tika uga 



kopikopi reherehe teitei titaa umiumi 



Last of all we note two vocables, huhu 6 and tnahara, in which the 

 Rapanui carries a sense that can only be distinguished as secondary in 

 evolution, so great a deviation does it show from the Maori compara- 

 tive material. 



For the reasons already set forth we must refrain from the more 

 general comment which in summing up might serve to explain these 

 variations. 



Two distinct items yet remain for consideration, one of sense and one 

 of phonetics, each applicable to all the languages of this province. 



The former item, dealing with a certain characteristic of word mean- 

 ing, the inversion of sense, must be postponed for later study, because 

 in certain psychological peculiarities of Polynesian speech we may best 

 find an explanation. 



In the phonetic treatment of Rapanui I have postponed detailed dis- 

 cussion of the aspirates. They are very irregular, at least very irreg- 

 ularly recorded, in this province, and I shall have to revert to them in 

 dealing separately with each language. To facilitate the examination 

 I subjoin a series of lists of all the items in which one or other of the 

 aspirates appears, or should appear, in Rapanui. Because our com- 

 parative apparatus varies widely in its incidence I have assembled these 



