RAPANUI SOURCES AND VARIETY. 43 



boldly the sunrise sea under the instinct of that heliotropism which 

 dominates the race. We discover that this swarming was made at a 

 time which is marked upon the calendar of speech, even if not upon a 

 tale of numbered years. When the Rapanui forefathers sailed out of 

 Samoa the mother tongue was still using its true aspirates, for there 

 were two in Proto-Samoan ; and it had not yet acquired the formative 

 elements which have availed in Nuclear Polynesia to maintain the final 

 consonants of closed stems; and in that mother tongue the accumula- 

 tion of a new and fashionable stock of speech material had not yet tucked 

 these ancestral words away into the nooks and corners of language to 

 live on obscurely as specific survivals. We shall find occasion in the 

 final summation to revert to the several points here established. 



Another element of a distinctive nature has been segregated in this 

 Rapanui vocabulary, the element which is to be credited to a Tongafiti 

 source and for which no Proto-Samoan affiliates are identifiable. We 

 now pass to the examination of the 119 items (839-957) so classed. Of 

 these items, those in which there is found such variety in sense between 

 the Rapanui and the Maori as to challenge attention are listed in the 

 following table : 



840 846 848 855 865 885 891 907 911 926 934 941 951 954 

 842 847 852 861 876 890 902 909 912 932 936 



It will be observed that there is a wide difference in the conditions of 

 this comparison when we come to deal with this specifically Tongafiti 

 contribution, of practically equal extent with that which we have segre- 

 gated as derived from a Proto-Samoan source. In dealing with that 

 material we enjoyed the opportunity of making a double comparison. 

 Thanks to my discovery and considerable reconstruction of the Proto- 

 Samoan mother speech we have been able to compare both Rapanui 

 and modern Samoan with that norm, and thus to compare them with 

 one another in the computation of the angle of divergence from the 

 norm, both in form and in sense. 



But in dealing with the Tongafiti contribution I set beside it for com- 

 parison another modern speech, the Maori. This inheres in the condi- 

 tions of the research. We have an excellent dictionary of that language ; 

 we soon shall have a better, when the Venerable Archdeacon Williams, 

 of Gisborne, brings to a conclusion, which may not fail of being brilliant, 

 the arduous toil of Maori lexicography upon which he has long been 

 engaged. We lack the tertium quid which should make Tongafiti 

 comparisons a matter of definitely ascertained and positively fixed 

 values, such as the discovery of the Proto-Samoan has given us for the 

 elder migration of the race. It has not yet occurred to the workers 

 upon the languages of the later migration to delve for the mother speech 

 of those more recent migrants. Although we recognize the inaccuracy 

 which must attend the comparison of two modern languages when we 



