IN ORGANIC EVOLUTION 81 



tion the destruction of the less fit is more drastic 

 than in nature and the segregation of the garden 

 or the flock is more perfect than is ever found in 

 field or forest. There are no natural barriers so 

 effective as those which may be reared in field or 

 garden. 



The existence of cognate or " geminate spe- 

 cies," as I have elsewhere called them, the one 

 representing the other on opposite sides of some 

 barrier, has been long recognized by naturalists. 

 In a general way such species agree with each 

 other in all the respects which usually distinguish 

 species within the genus. Their differences ap- 

 pear in minor regards, characters of degree, or 

 proportion; traits which we may safely suppose 

 to be of more recent origin than the ordinary 

 characters marking off species within the group. 



Illustrations of geminate species of birds, 

 mammals, fishes, reptiles, snails, crustaceans, in- 

 sects, trees, flowers, are well known to students 

 of these groups. 



To take familiar examples, each well separated 

 island in the West Indies has its own form of 

 golden warbler. Each island in the East Indies 

 has its own forms of reptileSj monkeys, snails, and 

 fresh water fishes. Each island in Hawaii has 

 its own species of each genus of Drepanine birds ; 

 each forest its own type of land snails. Each of 

 the three groups of rookeries in Bering Sea has 

 its own species of fur seal. Each section of the 

 Isthmus of Panama has its geminate species of 



