DETERMINATION OF THE PLACE OF RAPANUI. 



161 



net stock by subtraction from the gross of the several sums of the extra- 

 Rapanui element. We have no means of knowing in what proportions 

 this unidentifiable element may be associated with either of the ele- 

 ments of the several languages under comparison, but to include first 

 with one element and then with the other this irreducible element tends 

 to bring the computations of the two series into harmony. We set 

 forth in the following table, therefore, this second reckoning of net 

 stock for comparison with Table 3 1 . 



Table 36. 



In comparison with Table 31 we see at once that while the extra- 

 Rapanui component enters into the other languages of the province 

 in the order Paumotu-Tahiti-Marquesas-Mangareva, the order for the 

 Rapanui component is Paumotu-Marquesas-Tahiti-Mangareva, the 

 two last standing together. If, temporarily and solely now by way of 

 illustration, we treat the Rapanui speech as a unit intrusive to South- 

 east Polynesia, these figures will show that on its way to Easter Island 

 Hotu Matua's fleet made its largest landfall in the Paumotu, only much 

 smaller and fairly equal straggling squadrons scattered to the other 

 archipelagoes. 



It will be advantageous to subjoin to this memorandum of the pro- 

 portion to which Rapanui enters into sister speech the complement of 

 the picture, the relative diffusion computed upon the Rapanui stock: 



Table 37- 



Here we see that there is a marked difference in the diffusion when 

 referred to the Rapanui or to the several exterior bases. Mangareva 

 and the Marquesas are equally potent, Tahiti is considerably better 

 discoverable in Rapanui than is Rapanui in Tahiti, and the Paumotu 

 records are most remarkable in the passage from the top of one list to 

 the bottom of the other. We shall have to revert to this comparison of 

 the two tables when we shall have subjected the material to more 



