CiiAr. XII. INHABITANTS OF OCEANIC ISLANDS. 357 



and we know, from Mr. J. M. Jones's admirable account of Ber- 

 muda, that very many North American birds occasional!}' or 

 even frequently visit this island. Almost every year, as I am 

 informed l)y !Mr. E. V. Harcourt, many EuropeaTi and African 

 birds arc blown to Madeira ; this island is inhaljited by 99 

 kinds, of which one alone is peculiar, though very closely 

 related to a European form ; and three or four other species 

 are confined to this island and to the Canaries. So that the 

 islands of 13ermuda and Madeira have been stocked from the 

 neighboring continents with birds, whicli for long ages have 

 struggled together, and become mutually adapted ; hence, 

 when settled in their new homes, each kind wduld be kept by 

 the others to its proper place and habits, and would conse- 

 quently be but little liable to modification. Any tendency to 

 modification would also be checked by intercrossing with the 

 unmodified immigrants from the mother-country. Madeira, 

 again, is inhabited by a wonderful lumiber of peculiar land- 

 sliells, whereas not one species of sea-shell is peculiar to its 

 shores : now, though we do not know how sea-shells are dis- 

 persed, yet we can sec that their eggs or larva?, perhaps at- 

 tached to seaweed or floating timber, or to the feet of wading- 

 birds, might be transported across three or four hundi'cd miles 

 of open sea far more easily than land-shells. Tlie dilTerent 

 orders of insects inhabiting Madeira present nearly similar 

 cases. 



Oceanic islands are sometimes deficient in animals of cer- 

 tain Avhole classes, and their places are occupied by other 

 classes: thus in the Galapagos Islands reptiles, and in New 

 Zealand gigantic wingless birds, take, or recently took, the 

 j)lace of mammals. Altliough New Zealand is here spoken of 

 as an oceanic island, it is in some degree doubtful whether it 

 should 1)0 so ranked ; it is of large size, and is not separated 

 from Australia by a profoundly deep sea; from its geological 

 character and tlie direction of its mountain-ranges, the Rev. 

 W. 15. Clarke has lately maintained that this island, as well 

 as New Caledonia, should be considered as appurtenances of 

 Australia. Turning to plants. Dr. Hooker has shown that in 

 the Galapagos Islands tlie ])roportional numbers of the difl'er- 

 ent orders are very different from what they are elsewhere. 

 All such diflerences in luimber, and the absence of certain 

 whole grou]\s of animals and jilants on islands, are gtMierally 

 account(>d for by su]iposed ditVerences in their ])hysical condi- 

 tions; but this explanation is not a little doubtful. Facility 



