BARRIERS AND MEANS OF DISPERSAL 69 



A boa constrictor was washed up on the beach of the island of St. 

 Vincent coiled around the trunk of a cedar; it at once attracted atten- 

 tion and was killed. 60 



It must be recognized, however, that the actual establishment of 

 a species of mammal or reptile in a new territory or island, in con- 

 sequence of this sort of transport, could take place only under espe- 

 cially fortunate circumstances. If the sea journey lasts more than a 

 few days, mammals, and especially small ones, will die of starvation; 

 reptiles might fast for longer periods. In a strong wind, which is 

 necessary for rapid flotation, the waves would break over the floating 

 object, and endanger the travelers; and driftwood often is pounded for 

 days by the surf, and only rarely carried directly ashore. Even if the 

 traveler has successfully reached the land, colonization can take place 

 only if both sexes of the species in question or at least a fertilized or 

 pregnant female have completed the journey. The ability to continue 

 will even then depend on the habitat conditions afforded by the new 

 locality. Thus the chances for the colonization of new territory in this 

 way, especially by mammals, seem slight. 



Matthew 61 and Barbour 62 have critically examined the possibilities 

 of colonization by rafting. Matthew estimates that 10,000,000 rafts 

 may have been carried to sea during the 3,000,000 years (Walcott's 

 estimate) of Cenozoic time, and that the chance of a mammalian 

 species obtaining a foothold would be such that we might expect this 

 to have happened about 100 times, which is sufficient to account for 

 the dozen or two cases of mammals on the larger oceanic islands. 

 Barbour does not believe that rafting furnishes a feasible explanation 

 of such distribution, while Matthew thinks that it does. 



The chances are much better for forms with a dormant stage or 

 with eggs which may be carried on these natural rafts. The eggs of 

 reptiles might come ashore unharmed in the roots of trees, or in holes 

 or crevices in their trunks. The geckoes lay hard-shelled eggs attached 

 to the bark or trunks of trees, which require five months for develop- 

 ment, a period ample for transport to great distances. 63 Insects may be 

 distributed in their pupal stages, especially wood-eating forms which 

 pupate within the trunks of trees. This may explain the preponderance 

 of snout beetles on St. Helena, where they compose more than half 

 of the native beetle fauna. Snails, with and without opercula, are 

 good subjects for this form of transport. The experiments of Darwin 

 and Aucapitaine have shown how resistant they are to sea water; of 

 100 snails kept under sea water in a perforated box, 27 were still 

 alive after 14 days, among them 11 out of 12 species of Ericia. In 



