INTRODUCTION. XV 



snails, why not their eggs? This I believe explains the wide 

 distribution of closely related Leptachatinas. This genus ; 

 alone, in the AcJuitinellidce, is oviparous ;^and in it alone the j 

 minor groups are distributed widely, while in the other \ 

 (viviparous) genera, the minor groups are special to the j 

 several islands. The young snails at birth are many times 

 heavier than the eggs of snails of like size, and probably are 

 not carried far by the wind, else they would become more 

 widely spread locally. 



If we have no logical ground for the belief that the vivipar- 

 ous Acliatinellidce have been spread over sea from island to 

 island by such means as we have just 'Considered, how has 

 their spread been effected? Only by the traveling of the 

 snails themselves over land and through forests now sub- 

 merged. No other hypothesis is adequate to explain the facts 

 of distribution, and the mutual affinities of the several island 

 faunas. 



IV. 



CENTERS OF DISPERSAL. While none but palaeontological 

 evidence can be considered entirely conclusive in determin- 

 ing the area of origin and original dispersal of a group, yet 

 in dealing with groups of sedentary, closely related, and 

 strictly localized species, in a limited area, there cannot be 

 much chance of error in holding that the region of greatest 

 variety and abundance of such forms has been their center 

 of evolution. Thus, in Oahu the eastern half of the main or 

 Koolau range has fifteen species and many varieties of the 

 'group Metamastra, while the Waianae range has but three 

 species, of which two are specifically identical with main 

 range forms, the other closely related to them. The evidence 

 is therefore very strong that the center of radiation of this 

 group of species was in the eastern half of the main range, 

 where also fossil species have been found, the few Waianae 

 species being recent emigrants. By similar reasoning I con- 

 clude that the radiation-center of the group Paramastra was 

 in the Waianae (Kaala or Western) range. By the same 

 criteria, the typical group of Amastra, Heteramastra and 





