THE ORIGIN OF SPECIES 



Population Size and the Effectiveness of Selection. All of the above 

 is based upon the assumption that mutation and selection are proceeding 

 in a population of indefinitely large size, but Sewall Wright showed that 

 the effectiveness of selection is greatly affected by the size of the popula- 

 tion concerned. The mathematical basis for this proposition is complex, 

 but the results are simple. It appears that mild selection pressures are 

 relatively ineffective both in very small and in very large populations. Se- 

 lection has its maximum effectiveness in moderate sized populations. What 

 these relative population sizes mean in numerical terms is much less clear 

 than might be desired. It should be pointed out, however, that it is the 

 actual breeding population rather than the total population which is 

 important here. Opponents of this idea point out that populations of 

 natural species are, in most cases, immense, though good estimates are 

 available only for a few endemic species. Its advocates, however, point 

 out that such immense natural species are divided into more or less iso- 

 lated subspecies; that these are further subdivided into local populations; 

 and that these local populations are the significant breeding units. Migra- 

 tion from one unit to another will tend to blur the lines which separate 

 them, but, generally speaking, the factors which tend to differentiate local 

 populations and subspecies will be stronger than the migration pressure. 

 This type of breeding population structure seems to be rather well estab- 

 lished: the basic question which remains is, how large is large, and how 

 small is small, from the viewpoint of population dynamics? 



Wright believes that evolutionary changes might occur with "explosive" 

 rapidity if favored by a combination of mutation, selection for a character 

 (or combination of characters), and optimum population structure, as 

 described above. This may be the answer to the question posed above, 

 whether some new principle, in addition to mutation and selection, is 

 necessary to account for the observed results of evolution. 



Genetic Drift, the Sewall Wright Effect. Yet another factor which 

 tends to upset the Hardy-Weinberg equilibrium is one which has been 

 referred to under several names, including genetic drift ( perhaps the best 

 name), scattering of variability, and the Sewall Wright effect. The phe- 

 nomenon was originally described by Fisher, but he has since disavowed 

 its value completely, and it is Wright who gets ( and deserves ) the major 

 credit for the development of this concept. Genetic drift refers to the 

 accidental fluctuations in the proportion of a particular allele which de- 

 pend upon the fact that the assortment of genes into gametes and the 

 combination of gametes to form zygotes are random processes. It is well 

 known that such accidental deviations from the theoretically expected 

 assortment of alleles, and from their expected recombinations at fertiliza- 

 tion, are responsible for the fact that Mendelian ratios are rarely obtained 

 exactly. But because the deviations in each experiment are random, they 

 tend to cancel out, and hence the validity of Mendelian principles is de- 

 monstrable by adding the results of many experiments, or by performing 

 large-scale experiments. 



In nature, the large-scale experiment is already provided. Such acci- 

 dents of sampling do not have an important effect on large breeding 



258 



