374 ENTOMOLOGY 



from the Pacific to the Atlantic, exhibits many geographical 

 varieties, some of which are clearly due to temperature, as 

 experiments have shown. 



Geographical isolation is often followed by changes in the 

 specific characters of an organism, as witness the endemic 

 species and varieties of oceanic islands. Even in the same 

 archipelago, the different islands may be characterized by dif- 

 ferent varieties of one and the same species, or even by differ- 

 ent but closely allied species of the same genus. Thus Darwin 

 and Alexander Agassiz found that in the Galapagos Islands 

 each island had its own species of Tropidurus (a lizard) and 

 had only one species, with almost no exceptions. The same 

 phenomenon occurs in the two Galapagan species of Schisto- 

 cerca .S*. melanocera and 5*. literosa. In melanocera, as 

 Scudder discovered, " Three or four distinct types are becom- 

 ing gradually differentiated on the eight [now ten] islands 

 from which they are known." Snodgrass, who has recently 

 made important additions to Scudder's account, says, in regard 

 to the two species, " The specimens from the different islands 

 show striking, though, in most cases, slight differences distin- 

 guishing the individuals of each island as a race, from those 

 inhabiting any other island. There are two exceptions. 

 Abingdon and Bindloe have the same form, and Albemarle 

 supports at least two races." Each of these two species pre- 

 sents no less than five racial types, to which distinctive names 

 have been applied. Though the relationships and evolution of 

 these races have been ably discussed by Snodgrass, definite 

 conclusions upon these subjects are still needed. Isolation in 

 general we have considered briefly in Chapter VII. 



Faunal Realms. The general distribution of life is such 

 that naturalists divide the earth into several realms, each of 

 which has its characteristic fauna and flora. As to the precise 

 boundaries of these faunal realms, zoologists do not all agree, 

 owing chiefly to the fact that faunae overlap one another to 

 such an extent as to render their exact separation more or less 

 arbitrary. Five realms, at least, are generally recognized : 



