XVII.] THE PROBLEM OF THE OKIGIN OF SPECIES. 213 



These may be compared to trees whicli have been left 

 alone by the gi'adual destruction of their neighbours, and 

 have at the same time lost many of their own branches. 



A different fact, pointing in the same direction, is the 

 following : — In the chapters on Habit and Variation we 

 have seen that if a character begins to vary, it will con- 

 tinue to vary for an indefinite time. It is a case of this 

 law, that when variations have arisen under domestication, 

 and consequently a very short time ago, geologically speak- 

 ing, the part that has varied, and is characteristic of the 

 variety, continues to be the most variable part of the 

 organism. The varieties of the domestic pigeon are the 

 best instance of this.^ But what I wish to remark is, that 

 what I have just stated of the varieties of a species is 

 equally true of the species of a genus. "When any part in 

 one species of a genus is unlike the corresponding part Cliaracters 

 in the other species, whether in size or in shape, or in any as between 

 other way, the same part is also usually variable as between so within^'' 

 the different varieties, and between the different individuals the species. 

 of the species ; or, what is saying the same thing in more 

 concise language, specific characters are more varial)le than 

 generic ones.- For instance, in some groups of beetles the Wings of 

 presence or absence of wings is of no importance in classi- ^^^ ^^" 

 fication : there are genera in which some species have 

 wings, and others have none ; and sometimes this cha- 

 racter is variable even between individuals of the same 

 species. This fact is what might be expected on the sup- 

 position, that species are only " well-marked and permanent 

 varieties : " on the supposition of the separate creation of 

 every species, it is utterly inexplicable. 



Eeversion to ancestral characters is closely connected Reversion 

 with variation, and is sometimes due to the same cause : ^jgg 

 namely, mixture of races. In aU the breeds of pigeons, 

 individuals are sometimes found that have reverted to the 

 plumao-e of the rock-pigeon, from which the domestic 

 breeds* are descended ; but these are most numerous when 

 two breeds have been crossed. Crossing of the breeds is 

 also a cause of general variation. 



1 Darwin's Origin of Species, p. 180. ^ Ibid. pp. 177, 1S2. 



