SOME QUANTITATIVE ASPECTS OF EVOLUTION 



populations, because an accidental increase of A in one part of a popula- 

 tion will generally be counterbalanced by an accidental increase of a in 

 another part of the same population. Or the genetic drift of one season 

 will be reversed in the next season. But in small populations, the situation 

 is very different. If there are only 100 individuals in a particular popula- 

 tion ( and smaller breeding populations do exist, for example, the whoop- 

 ing crane), and if a particular allele is present only once, then an accident 

 of sampling might easily in a single generation remove it irrevocably from 

 the population, or increase it many fold, say to 10 per cent. The result is 

 that, in small, isolated populations, genes may be completely lost or com- 

 pletely fixed by genetic drift, without reference to selective value. Genetic 

 drift is thus a force working in opposition to selection, for it tends to 

 preserve or to destroy genes without distinction, whether favorable, neu- 

 tral, or unfavorable. But selection tends to preserve those which confer 

 some adaptive value and to destroy those which impair the adaptive 

 value of a species. Severe selective forces will, of course, destroy disad- 

 vantageous genes irrespective of the population size. 



Wright's viewpoint on the significance of genetic drift has been much 



Figure 99. Allometry in Baboon Skulls, 

 showing the great increase of facial length 

 with respect to cranial length with increasing 

 total size. Actually, k = 4.25, a very high 

 figure. ( From Huxley, "Problems of Relative 

 Growth," Methuen & Co., Ltd., London, 

 1932.) 



259 



