NATURAL SELECTION 81 



importance in the production of new species."* It 

 is evident, however, that if a species be divided 

 into two parts by barriers, this cannot decrease the 

 difficulties already mentioned — this cannot prevent 

 the merging of variations, nor produce cross-sterility 

 between parent and offspring. The same difficulties 

 that arise in forming a new species from a whole 

 species, apply with equal force to the formation of a 

 new species from a part of a species. The dividing 

 of a species by barriers cannot therefore account for 

 the formation of a new species. 



When we consider that many hundreds of thous- 

 ands of species have existed, we cannot for a moment 

 suppose that sufficient barriers could have isolated 

 them, or even any considerable per cent of them, 

 while being formed. 



Besides, closely related species occupy the same ter- 

 ritory, which could not generally be the case if they 

 had been isolated from each other by barriers while 

 being formed, for if such barriers had existed they 

 would have kept the species permanently isolated. 



Besides, if the supposed barriers were sufficient to 

 isolate the parts of a species, they would be sufficient 

 to prevent the species from separating into parts — 

 for the barriers are, by supposition, impassable. 



Mr. Darwin, however, as already stated, denies that 

 " migration and isolation are necessary for the forma- 

 tion of new species." When we consider the largo 

 numbers and relative positions of species, it seems 

 necessary to assume that closely related species have 

 generally been formed without being insolated by 

 barriers. 



With regard to the difficulty of preserving varia- 

 tions in nature, Le Conte says: " But how can the 

 process of progressive divergence begin, when slight 



6 * Origin of Species, p. 106. 



