NOTES ON ZOOGEOGRAPHY. xliii 



united. In addition to these primitive Amastridce, almost 

 every non-Achatinellid genus of land and fresh water shell + 

 found in the archipelago occurs on all the islands, excepting 

 the low and arid Niihau and Kahoolawe, where part are want- 

 ing. It appears highly improbable that so many genera 

 should have become so generally diffused by any ' ' accidental ' ' 

 means of over-sea transportation. That Kauai has ever con- 

 tributed faunally to the other islands by means of drifting 

 trees or the like is the more improbable because such drift 

 would be against what current there is, and directly across 

 the course of the trades. The snail fauna of Kauai is 

 "primitive" only in lacking representatives of the Achatinel- 

 lid<z, and the arboreal genus Auriculella. The fauna of 

 Amastridce and other families is on a par with the faunas of 

 the other islands. The absence of Achatinellidce from Kauai 

 seems to be a rather strong argument, albeit negative, against 

 the view that the fauna of the entire group emanated from 

 Kauai. 



Mr. Perkins, in discussing the dispersal of Achatinellidce 

 states that ' ' once, on Molokai, a young living Achatinella was 

 found attached to the feathers of the Drepanid bird Chloro- 

 drepanis. Frequently they become adherent to one's clothes , 

 when passing through the brush. Doubtless in high winds 

 very young shells are sometimes carried to a distance in curled \ 

 up leaves in which they hide" (Fauna Hawaiiensis, Intro- 

 duction, p. Ixvi). While such means of travel must have 

 had some part in the distribution of Achatinella, we are in- 

 clined to believe it a small part, for the reason that 

 anomalies in distribution are very rare. If birds had carried I 

 the snails about, there would be colonies of "erratics," like 

 granite boulders in a glaciated limestone region, instead of 

 the thoroughly consistent distribution which is the rule on 

 every island. 



It is now known that the family Achatinellidce differs in 

 important structural characters from the Amastridce. While 

 the two are allied, their separation must, if we judge by the 

 analogy of other groups of known age, date back to the early 

 Tertiary at least. Indeed I believe that no existing genus of 



