HISTORICAL ZOOGEOGRAPHY 117 



groups of animals, and under the weight of his prestige, the delimita- 

 tion and subdivision of the faunal regions have long constituted one 

 of the principal branches of zoogeographic inquiry. 



The belief that there is a division into faunal regions of general 

 validity for all the classes of animals cannot be maintained in the 

 light of modern knowledge. For the Mantidae, there is no sharp divi- 

 sion between the Ethiopian and Oriental regions, which have numerous 

 types in common. The few palaearctic forms are directly derived from 

 these, so that in general, this family can be divided only into two 

 regional groups, a palaeotropical and a neotropical. 58 Nor can the 

 distribution of the non-European genera of mites be brought into 

 harmony with the usual regional divisions. 59 With reference to the 

 relations of its earthworms, New Guinea would be grouped with south- 

 eastern Asia, while its other faunal affinities are primarily Australian. 60 

 Ceylon, by contrast, agrees closely with Australia in its earthworm 

 fauna. 60 Chile differs from the rest of South America in its mollusks, 

 its fresh-water fishes, and its earthworms, but its mammalian and 

 avian fauna is typically South American. 



An exact delimitation of the regions is also impossible, and opinions 

 as to their proper limits have accordingly varied exceedingly. The 

 boundary between the Oriental and Australian regions has been a 

 special bone of contention, and Celebes has been placed now with one 

 and now with the other, according to the group of animals employed 

 as a criterion. The creation of transition areas, equally allied to the 

 regions between which they lie, shows that the supposed regions are 

 not objectively defined areas. They are abstractions, combinations of 

 more or less allied faunal elements, and they will be of varying extent 

 and limits according to the animals emphasized. 



It is true that there are common features in the distribution of 

 many different terrestrial animals. The conditions of dispersal and the 

 barriers and highways available must save been the same for many 

 groups. The differences are probably chiefly due to the relative ages of 

 the different groups of animals. Classes which arose at the end of the 

 Triassic could not spread by means of land connections which were 

 available for the early Mesozoic scorpions, insects, and mollusks. The 

 older a group of animals, the more manifold will have been its oppor- 

 tunities for dispersal. The differences in vagility must be taken into 

 consideration. Flying animals, such as birds and bats and some 

 insects, have been able to spread by routes closed to other land 

 animals, for, even though they do not fly across great stretches of 

 ocean of their own free will, chains of islands and archipelagoes have 

 facilitated their dispersal to islands which were permanently inacces- 



