ORIGIN OF SPECIES. 139 



animals have the power of expressing and exercising a pre- 

 ference in the manner above indicated, then it is easy to 

 understand how variations might be transmitted or intensi- 

 fied in this way. The male who is most .attractive to the 

 female, will, other things being equal, have the best chance 

 of propagating his species, and is likely to leave the largest 

 number of descendants. His male offspring will inherit the 

 peculiarities by which their sire was rendered pre-eminently 

 attractive in the eyes of their mother, and thus a well-marked 

 breed might be produced, by the preservation or intensifi- 

 cation of characters of this nature. Mr Darwin is disposed 

 to believe that colour and song in most, if not in all, animals 

 are thus to be ascribed to the action of Sexual Selection, 

 through numerous successive generations ; but other com- 

 petent authorities are unable to concur in this view. 



Numerous objections have been brought forward to prove 

 the insufficiency of the view that the Evolution of species 

 has been effected by Natural Selection. The student de- 

 sirous of making himself acquainted with this subject should 

 consult Mr Mivart's ' Genesis of Species ; ' but the follow- 

 ing are the chief difficulties which the advocate of Natural 

 Selection has to meet : 



i. Natural Selection, whilst doubtless capable of preserv- 

 ing favourable variations, cannot initiate changes of any 

 kind. The origin, therefore, of variations is not elucidated 

 in any way by the doctrine of Natural Selection, and we are 

 compelled to believe that the variability of the individuals 

 of a species depends upon some internal law with which we 

 are not as yet acquainted. It thus remains open for us to 

 believe that the law which gives rise to variations is in 

 every way a more important one than that under which 

 they are simply preserved. Unfavourable variations must 

 be at least as common as those which are advantageous, 

 and whilst Natural Selection can produce neither, it can at 

 best but preserve the latter. It seems clear also that many 

 variations which, when fully developed, are very useful to 



