EVIDENCE FROM DISTRIBUTION 143 



derstand why a large continental island should be 

 so poor in mammals. The contrast between the 

 large islands of the East and West Indies is very re- 

 markable; Borneo, Sumatra and Java are extremely 

 rich in mammals, while Cuba, Haiti and Porto Rico 

 have very few. The oceanic islands have peculiarly 

 limited faunas; on the theory of special creation, we 

 should expect to find in them as large a variety of 

 animals as their area, their vegetation and their 

 climate would enable them to support. On the 

 theory of evolution, they must have received only 

 such plants and animals as could be carried to them 

 by the wind, or currents of the sea, carried on drift- 

 wood or floating trees, or transported by birds. 

 When an oceanic island is very far from the nearest 

 land, the accession of a new animal or plant must be 

 a very rare event. 



That islands are actually, not merely hypothetic- 

 ally, stocked in this manner is clearly shown by the 

 case of Krakatoa, a volcanic island not far from 

 Java. The great eruption of 1883 blew a great part 

 of the island out of existence and buried the re- 

 mainder so deeply under volcanic debris, that every 

 living thing perished. In a surprisingly short time 

 the island was re-vegetated and abundant insect 

 life arrived, all borne by the winds and the sea from 

 Java and Sumatra. Professor Selenka writes of a 

 visit to Krakatoa: "Under the shade of a Casuarina, 

 amidst cocoa-nut palms and thickets as high as my 

 head, I found here, to my astonishment, an active 



