I02 A Factor in Evohition 



sweeping change in the environment, of such a kind that 

 only the variations lying near the extreme n can accommo- 

 date to it and live to reproduce. The next generation would 

 then show variations about the mean ;/. And the chances 

 of fossils from this generation, and the subsequent ones, 

 would be of creatures approximating ;/. Here would be a 

 great discontinuity in the chain of descent and also a wide- 

 spread prevalence of variations seeming to be in a single 

 direction. This seems especially likely when we consider 

 that the paleontologist does not deal with successive gen- 

 erations, but with widely remote periods, and the smallest 

 lapse of time which he can take cognizance of is long 

 enough to give the new mean of variation, ;?, a lot of gen- 

 erations in which to multiply and deposit its representative 

 fossils. Of course this would be only the action of natural 

 selection upon ' preformed ' variations in those cases which 

 did not involve positive changes, in structure and function, 

 acquired in ontogenesis; but in so far as such ontogenetic 

 accommodations were actually at hand, the extent of 

 difference of the //-mean from the ,r-mean would be 

 greater, and hence the resources of explanation, both of 

 the sudden prevalence of the new type and of its dis- 

 continuity from the earlier, would be much increased. 

 This additional resource is due to the organic selection 

 factor.^ 



We seem to be able also to utilize all the evidence 

 usually cited for the functional origin of specific characters 

 and groupings of characters. So far as the Lamarckians 

 have a strong case here, it remains as strong if organic 

 selection be substituted for the * inheritance of acquired 



^ A synopsis of the ajjplications of this principle is given below, in Chap. 

 XIII. 



