13^ THE DISPERSAL OF SHELLS. 



and river margins, might possibly be carried with 

 floating islands or rafts, but it is obvious, of course, 

 that the landing of objects of this kind on foreign shores 

 in such a manner as to enable slow-moving terrestrial 

 creatures to disembark must be very rare indeed, yet 

 such occurrences, as Lyell maintained, notwithstanding 

 their extreme rarity, may possibly account " in tropical 

 countries for the extension of some species of mammalia, 

 birds, insects, la7id-sJiells, and plants to lands which 

 without such aid they could never have reached." * 

 Professor Semper has remarked that large land-shells, 

 and such as live in the highest branches of trees and 

 lay their eggs there, like all the species of Cochlostyla, 

 will obviously be far more difficult to transport than 

 small species which can creep into rifts in trees or be- 

 tween the roots ; and species belonging to groups, which 

 (like Helix similaris) live on the ground among stones 

 and earth, will, he observes, be almost as well protected 

 during a sea voyage as the operculate kinds. This he 

 shows is in harmony with certain known facts in dis- 

 tribution, most of the small species and of the operculate 

 species having a much wider range than the large 

 inoperculate forms. In the Philippines the typical 

 genera (or sub-genera) Cochlostyla, Rhysota, Chlorcea^ 

 and Obbina, which principally live on trees, are 

 almost confined to that group of islands, while the 

 small genera, as Siibtilina, TrochomorpJia^ and Ennea^ 

 among the Helicidae, and the operculate genera Cyclo- 



^ •' Principles," ii. p. 367. 



