DARWIN 



and those shorter; of mentally strong and weak, etc. 



While, at first sight, it seems possible, though hard- 

 ly probable, that such directed breeding might occur 

 in the natural state, two facts destroy the force of the 

 analogy even with this granted. New varieties and 

 races created by artificial selection revert to the orig- 

 inal type as soon as they are left to their own devices; 

 and, in no case, has mutual sterility been produced 

 between different varieties; the most different kinds 

 of dogs, pigeons, and other domesticated animals 

 breed freely together and their offspring is a mongrel ; 

 while successive general intercourse soon obliterates 

 all the special traits. It is evident that fixity of traits 

 and sterility are essential to establish a species, and 

 that in some way these must occur in a state of nature. 

 The importance of artificial selection was first crit- 

 icised by Huxley, and its value as a proof has stead- 

 ily declined until now many biologists admit that 

 there are fewer features in common between natural 

 and artificial selection than the Darwinians supposed. 



Let us now turn to natural selection and let us as- 

 sume that a useful variation has occurred in some in- 

 dividuals of a certain species; under these favourable 

 conditions we can discuss what the chances are that 

 this variation will continue to increase until the orig- 

 inal variants have produced an abundant offspring 

 so different from the original stock that a new species 

 has been formed. Again, to make our argument pre- 

 cise, let us consider the case of the long neck of the 



I 221 ] 



