13G Rev. J. T. Gulick on the 



from generation to generation by a law that is quite distinct 

 from natural selection. It was also shown that endowments 

 only partially segregative (as, for example, somewhat diver- 

 gent habits of feeding), when not concurrent with any forms 

 of cross incompatibility, are liable to be obliterated by 

 crossing ; but, when associated with segregate fertility and 

 cross infertility, will increase from generation to generation, 

 even if the mongrels are as well adapted to the environment 

 as the pure forms. I at the same time called attention to the 

 fact that, when associated with some form of partial positive 

 segregation (as divergent habits of feeding or segregative 

 sexual and social instincts), greater vigour of pure forms, as 

 contrasted with the mongrels, would have the same effect as 

 their greater fertility. In other words, Segregate Vigour 

 would preserve a partially segregated variety as effectually as 

 Segregate Fecundity. 



Incompatihilities will disappear unless preserved by 

 Positive Segregation. 



Mr. Wallace has given a very instructive computation on 

 pages 181-184 ; but it does not seem to me to prove, as he 

 supposes, that infertility between the individuals of a species 

 cannot increase "unless correlated with some useful variation," 

 but that it cannot arise, except as a transitory variation, 

 unless associated with some positively segregative principle, 

 causing those to pair together which are fertile with each 

 other. My contention is that, without some positive form of 

 segregation, fecundity and cross sterility can never arise, and 

 that, after it has arisen under segregation, no amount of corre- 

 lation with useful variation will preserve it if the positive 

 segregation is removed. If, for example, all the species of 

 humming-birds were brought together in one country, and 

 were deprived of all segregative habits and instincts, it cer- 

 tainly would not require many generations to reduce them to 

 one species. If equally adapted to the environment, the 

 species that would succeed in perpetuating itself would be the 

 one represented by the largest number of individuals ; or, if 

 several species were entirely cross fertile and were in the 

 aggregate represented by a larger number of individuals than 

 any other similar group of species or than any single species, 

 then the resulting species would be the hybrid descendants of 

 this most numerous group. All the other species would be- 

 come extinct through failing to mate with " physiological 

 complements." 



