SELECTION. 179 



ively isolated from admixture with individuals which do not 

 belong to them, one can be isolated as the result of an actual 

 geographic separation from individuals which could cross with 

 its members, or as the result of some geno-typic quality which 

 makes the individuals succeed only in situations where individ- 

 uals not of its species are not or seldom present. Another 

 group may be isolated wholly by the wishes of breeders, who 

 control the breeding of the group to make its members con- 

 form to a certain standard. Both these groups are species, the 

 first a natural species and the second a domestic one. Both 

 have this in common, that the situation and constitution of 

 the group is such, that the potential variability cannot but 

 decrease. 



Either the potential variability of such a group is zero or as 

 near zero as is consistent with the differentiation into individ- 

 uals of different sex, or the potential variability of such a 

 group attains a certain magnitude. If the latter is true, further 

 reduction of the Total potential variability of the group is still 

 possible, and as a result of this fact, the eventual type, the 

 eventual geno-typic constitution of the group is not yet rigidly 

 set. It is conceivable, that in such a case there are no pheno- 

 typic alternatives given in the constitution of the groups, but 

 it is clear, that very often the eventual t37pe for which a group 

 can become pure is not as yet rigidly laid down. In such groups, 

 selection may affect the outcome, just as much where the 

 reason for the inclusion of certain individuals among the num- 

 ber which produce off-spring rather than others, lays in their 

 greater chance to develop up to reproducing age (natural 

 selection), or in their greater comformity to the breeder's ideal 

 (artificial selection). 



But we must always bear in mind, what we considered in the 

 chapter on the reduction of variability, namely, that minorities 

 have no chance. Natural selection cannot do other than affect 

 the average chance of survival of certain individuals. We can 

 take a simple illustration, a certain variation in size in a group 

 of organisms, which partly depends upon the geno-typic con- 



