INDISCRIMINATE ELIMINATION. 209 



selection which is controlled by activities belonging to nature outside of the 

 species. It is, therefore, clearly distinguished from active (or endonomic) selec- 

 tion, which is controlled by differences of aptitudes or of habitudes in the different 

 groups for dealing with the environment, and not by exposure of the different 

 groups to different environments. 



9. Indiscriminate Eliminalional Intension. 



Eliminational intension is segregation and divergence produced by 

 the indiscriminate destruction or failure to propagate of a part of the 

 individuals of an intergenerating section of a species. Though indis- 

 criminate destruction can not be classed as a form of natural selection, 

 it may nevertheless be the cause of transformation ; and when a species 

 is distributed in sections that are prevented from intergenerating, 

 divergent evolution will often be hastened by the indiscriminate 

 destruction of part of the members of one or more sections. If a species 

 inhabiting a large island is divided by geological subsidence into two 

 large sections, there may be a very close resemblance in the average 

 character of the two sections ; but if a subsequent eruption of hot ashes 

 destroys a large portion of the individuals of one section, or of both, the 

 probability of a close correspondence in the average character of the 

 two sections will be very much less than before the eruption. 



Again, when the area occupied by a species is divided into two or 

 more large districts, the occupants of which can have little or no 

 opportunity for crossing, divergent evolution will arise in the different 

 districts unless there is some constantly operating cause that insures 

 that all the varieties surviving and propagating in anyone district shall 

 survive and propagate in all the districts. No such cause has ever been 

 pointed out, but, on the contrary, it can easily be shown that the prob- 

 ability is very small that such a correspondence would occur, even if 

 at the time of the division of the areas every individual in each district 

 was represented by a completely similar individual in each of the 

 other districts. Let us suppose a case: 



(1) Suppose the creatures under consideration to be a species of 

 mollusk, the sexual instincts of which act without any segregative 

 tendency between the varieties of the same species, there being no 

 aversion or other impediment that interferes with the free crossing 

 of all the variations occurring within the limits of one district. 



(2) vSuppose that the number of individuals in each district is 

 10,000,000. 



(3) Suppose that one in a thousand of these has a tongue strong 

 enough to feed on the bark of the tree the leaves of which are the ordi- 

 nary food of the species, and that one in a thousand is capable of 



