22 Darwinism and Other Essays. 



cations, and not by disconnected acts of creation. 

 The facts of classification alone are sufficiently 

 decisive. By the older naturalists, who sought to 

 arrange animals and plants in groups according to 

 their resemblances, attempts were often made to 

 construct a linear series in which each group 

 should be intermediate between those which pre- 

 ceded and those which followed it. All such at- 

 tempts proved futile, and after a half-century of 

 discussion and criticism it became evident that 

 the only possible classification which correctly 

 represents the facts is one in which organisms are 

 arranged in divergent groups and sub-groups, like 

 the branches and twigs of what is aptly termed a 

 family tree. Wherever different orders, families, 

 or genera show points of resemblance to each 

 other, the resemblances occur always at the bot- 

 tom, among their least highly developed species. 

 Apes, bats, and rabbits are sufficiently distinct in 

 type, but the lowest members of the orders to 

 which these animals respectively belong are strik- 

 'ngly like one another. At the bottom of the 

 mammalian class, the echidna and duck-bill have 

 many points in common with birds and reptiles; 

 while birds and reptiles not only draw together so 

 that it is hard to distinguish their most primitive 

 forms as clearly bird or clearly reptile, but these 



