III. 



The Origin and Classification of Islands. 



EVERY island has its history, and in the case of all but newly-formed 

 volcanic islands or coral islets, every island has a double history, 

 that of the island itself and that of its colonisation by the plants and 

 animals which live upon it. The rocks of which an island consists 

 will give us an insight into, though not always a complete knowledge 

 of, its geological history ; and a study of its living inhabitants will 

 generally enable us to decide whether it has been colonised as an 

 island or by direct former connection with a continent. Some biologists 

 maintain that the fauna of an island will show whether it has ever 

 been united to a continent or not, and this is the question which I 

 propose to discuss in the present paper, because the answer to it 

 involves some important inferences and conclusions. 



There are many ways in which an island can be formed ; it may 

 be but a portion of a continent severed from the mainland by the 

 erosive action of the sea ; it may be the mountainous part of a 

 country which has sunk beneath the ocean ; it may have been thrown 

 up from the floor of the ocean by volcanic action, or it may have been 

 built up by the growth of reef-making corals. There are, however, 

 only two ways in which an island can have been populated without 

 the intervention of man ; either it must once have been united to a 

 continent and its inhabitants must be the descendants of those that 

 then lived on that continent, or else its tenants must have been 

 transported across the sea by the help of drift-wood, or by birds, or 

 by winds and storms. 



It is evident, therefore, that in most cases there is likely to be a 

 certain relation between the geological structure of an island and the 

 nature of its fauna and flora. Islands formed in the ocean, whether 

 by direct upheaval, or by volcanic eruptions, or by coral growths, are 

 not likely to possess a large assemblage of plants or of animals ; they 

 may be covered with vegetation, but the animals found on them must 

 be the descendants of occasional waifs and strays. On the other hand, 

 an island which has once been part of a continent will, if it remain 

 large enough, continue to support a large nvmiber of animals, and 

 these will generally include a certain number of Mammalia and 

 Amphibia. 



Islands have consequently been (divided into two great classes — 



