MEANS AND METHODS OF EVOLUTIONARY CHANGE 349 



the case with blue and brown eyes in many parts of the world. If, on the 

 other hand, one gene, for example the recessive one, is rare, most mem- 

 bers of the population will exhibit the characteristic produced by the domi- 

 nant gene and only occasional individuals will show the recessive charac- 

 teristic. Thus among sheep, although the predominant color is white, black 

 individuals appear now and then. Occasional appearance of albinos in 

 almost all species of higher animals affords another example. Other ex- 

 amples will be found in the discussion of polymorphic species (pp. 375- 

 376). In this connection we should stress the fact that there will be no 

 tendency for the recessive gene, though rare, to "die out." This fact is 

 particularly pertinent to our discussion in view of the fact that all "new 

 mutations" are rare at first. 



The question under discussion is. What will happen to new mutations 

 when they occur? We may now conclude that they will be incorporated 

 into the genetic structure, the "gene pool," of the population, and that 

 there will be a tendency for establishment of an equilibrium between the 

 number of "new" genes and the number of "old" ones. 



Genetic Drift 



The genetic equilibrium just described is most effectively maintained 

 when the size of the population is large. When the population is small, 

 chance may cause radical deviations from the expected equilibrium. In a 

 small population, confined, for example, to a small island, or to one moun- 

 tain valley, or to one pond (in the case of an aquatic form), one of two 

 things may happen: By the action of the laws of chance the gene a may 

 be lost entirely, or, contrariwise, the gene A may be lost entirely, all mem- 

 bers of the population coming eventually to have the new characteristic, 

 i.e., to be aa. This phenomenon is known as "scattering of the variabiHty" 

 or genetic drift. Our knowledge of it is based largely upon the mathemati- 

 cal studies of Professor Sewall Wright and others in the field of population 

 genetics. Without recourse to mathematics we may cite a simple example 

 illustrative of the fundamental idea involved. We noted previously that a 

 heterozygous individual (Aa) will produce two kinds of germ cells, some 

 containing the dominant gene (A), some the recessive gene (a). These 

 two types of germ cells will be produced in about equal numbers; hence 

 on the average about half of the individual's offspring will be expected to 

 receive gene A from him, about half will be expected to receive gene a 

 from him (cf. Fig. 15.1 ). But that statement is true only if the individual 

 contributes genes to large numbers of offspring. Suppose, by contrast, 



