308 TROPICAL NATURE, AND OTHER ESSAYS. 



Fernandez are thus deficient. They all of them, however, 

 possess birds, and most of them bats ; and whenever 

 small mammalia, such as goats, pigs, rabbits, and mice 

 have been introduced they have run wild and often in- 

 creased enormously^ proving that the only reason v/hy 

 such animals were not originally found there was the 

 impossibility of them crossing the sea ; while such as 

 could fly over — birds, bats, and insects — existed in 

 greater or less abundance. If, on the other hand, they 

 had once formed part of the continent, it is impossible 

 to believe that some of the smaller mammalia, as well 

 as frogs, would not have continued to exist in the 

 islands to the present day. 



If we compare the productions of different islands, we 

 meet with peculiarities which throw much light on the 

 subject of distribution. In the Galapagos islands, 

 between 500 and 600 miles from the west coast of South 

 America, there are thirty-two species of land-birds, all 

 but two or three being peculiar to the group. In 

 Madeira, about 400 miles from the coast of Morocco, 

 there are nearly twice as many land-birds as in the 

 Galapagos, but only two of these are peculiar to the 

 island, the rest being South European or N. African 

 species. The Azores are 1,000 miles west of Portugal, 

 and they contain twenty-two species of land-birds, every 

 one of which is European except one bullfinch which is 

 slightly different and forms a peculiar species. This 

 remarkable difference in the proportion of peculiar 

 species between the Galapagos and the Atlantic islands, 

 is well explained by the theory that land-birds rarely 

 fly directly out to sea, except when carried against 

 their will by storms and gales of wind. Now the 



