146 HEREDITy AND SELECTION IN SOCIOLOGY 



D. — The Extinction of Species. 



We have seen that a species is a comflex of adaptations, and 

 that species result, in every case, from adaptation to the en- 

 vironment. It is not those general variations which De Vries 

 has called mutations that are the cause of the origin of new 

 species ; for such mutations, as De Vries has shown, are in them- 

 selves neither good nor bad. Their preservation cannot, there- 

 fore, be due to natural selection ; for the latter can only act in cases 

 of adaptation when two variations present themselves, one of 

 which is more advantageous to the species in the given condi- 

 tions. If mutation were the origin of those organic distiactions 

 which we call species, natural selection would be reduced to the 

 negative role of eUminating those variations which were useless 

 or harmful. The possibilities of transformation would obviously 

 be greatly reduced if mutation were the sole cause ; and it is 

 difficult to see what cause could act on the different members 



of bird. This rarity is due to the fact that Madeira is less than 200 nulea 

 distant from the African coast, and that, as many mainland birds are 

 blown over to the island every year, crossing between the island birds 

 and those from the mainland has been frequent, with the result that new 

 variations have always been kept in check by amphimixis. 



Amixis, or the prevention, through isolation, of crossing with the parent 

 species, does not, however, necessarily lead to variation. In Sardinia we 

 find two varieties of the species Vanessa urtica, one of which is endemic 

 and has varied distinctly from the parent species, while the other has not 

 varied. It has been suggested that the two stocks were not isolated at 

 the same time, but we can find no justification for this supposition and 

 its corollary, that the variation manifested by the endemic stock is due to 

 the direct action of the chmate. There are some butterflies which occur 

 both in the Alps and in the Polar regions. Some of these have varied, so 

 that we can distinguish an Alpine from a Polar variety, but others have 

 remained the same in the two regions ; and yet we know that all of them 

 were isolated at the same time. The explanation is that at the time of 

 geographical separation during the Glacial period one of the species was 

 in a period of variation, whereas the other was in a period of constancy. 

 Isolation has prevented the variations from being " swamped," while it 

 has had no effect on the other species, which remain to-day as they were 

 in the Glacial period before their separation from the parent species. No 



