154 SPECIAL FEATURES 



degree of specialization along their own lines, and have 

 evolved in many dijBferent directions. 



If we sum up the more important general points in 

 regard to island faunas, we may note first that Wallace's 

 distinction into oceanic and continental islands is not 

 of very great practical importance, because the pro- 

 gress of geology, since the classification was laid down, 

 has made it difficult to apply. It involves a conception 

 of the permanence of ocean basins which is not now 

 generally accepted. Thus, while Wallace names the 

 Galapagos group as a characteristic oceanic group, 

 others beheve that the presence of giant land tortoises 

 is in itself a proof of a former land connexion. For 

 our purpose then it is sufficient to recognize the follow- 

 ing three conditions : 



1. When an island has never had a connexion with 

 a continent, e. g. many coral and volcanic islands, or 

 has only had such a connexion at a geologically remote 

 period (Galapagos?) before the higher forms Hving at 

 the present day had originated, or before they had had 

 time to become widely distributed, its f aima contains no 

 mammals, no amphibia, (1) no reptiles, or (2) reptiles of 

 a very primitive type, or (3) such as could be accident- 

 ally carried by floating wood, &c. The most important 

 members of its fauna will be birds, insects, and land 

 shells. Owing to the isolation the animals will tend to 

 exhibit special pecuHarities, fitting them for their 

 habitat. Among flying forms the conditions favour 

 those in which the power of flight is moderately de- 

 veloped, or has disappeared. Examples of such islands 

 are the Azores, the Bermudas, St. Helena, the Sand- 

 wich Islands, &c. Where such islands have had a 

 continental connexion, this must have been prior to 

 the deposition of the lower Tertiary beds, for at this 



