334 AN EMBRYOLOGIST'S VIEW 



vergence and probably of considerable adaptational divergence 

 as well. Tims, in any zone of overlap the two groups may be 

 assumed to have somewhat different ecological niches. This may 

 take the form of differences in breeding time, habitat, food 

 preferences, and so on. The net effect will be to limit the pos- 

 sibility of forming hybrids. 



The previously evolved adaptative differences will tend to 

 restrict competition between the two populations in the zone of 

 overlap. Where competition does occur, however, there are some 

 conditions under which further divergence of the populations 

 will be promoted. Let us designate the two populations A and B 

 and assume two habitats alpha and beta. Let us further assume 

 that A was previously isolated In a zone where alpha was the 

 commonest habitat and that the course of A's evolution was for 

 adaptation in alpha although it also occupied beta. We will 

 assume that B was found in an area where beta was the prevail- 

 ing habitat and B became better adapted to beta in spite of the 

 fact that it occupied the available alpha to a certain extent. If 

 the two populations became sympatric in a region with alpha 

 and beta, we would anticipate that each species would become 

 restricted to a single habitat. A being better adapted to alpha 

 would displace B from this habitat. Similarly, we would antici- 

 pate that B would displace A from beta. In the zone of overlap, 

 any genetic difference would have a selective advantage if it 

 tended to restrict each species to the habitat to which it was 

 better adapted. Thus, an isolating mechanism (habitat prefer- 

 ence) would be enhanced, but this would be a consequence of 

 natural selection reducing competition and not of natural selec- 

 tion reducing hybridization. 



In the situation described, there seems no reason to believe 

 that the gene differences that promote habitat isolation in the 

 /one of overlap will spread throughout the population. The gene 

 differences will tend to be restricted to the /one where they 

 have an adaptive signifieance. 



Much the same line of reasoning would lead us to question 

 the ability of selection against hybridization to produce isolating 

 mechanisms that characterize the entire population. Let us again 



