HALDANE ON SELECTION 



z 5 i 



characters unless they are extraordinarily- 

 valuable to their possessors. Such char- 

 acters appear far more frequently than 

 dominant mutations, but in their early 

 states are selected infinitely more slowly. 

 In Part II (5) of this series of papers, the 

 author considers the influence of partial 

 self-fertilization, inbreeding, assortative 

 mating, and selective fertilization on 

 the composition of Mendelian popula- 



enormously increase the number of reces- 

 sives when these recessives are sufficiently 

 rare . Assortative mating, which has been 

 shown by Pearson (iz) to exist in human 

 populations, will on the other hand have 

 but little effect. Hence, inbreeding or 

 self-fertilization appears to be necessary 

 in early stages of selection of a recessive 

 character if this process is to be fast 

 enough to be an effective cause of evolu- 



TABLE 3 



Effects of complete selection 



CHARACTER ELIMINATED 



TYPE OF SELECTION 



SEX 



PER CENT 

 AFTER 5 

 GENERATIONS 

 FROM 50 

 PER CENT 



PER CENT 

 AFTER 10 

 GENERATIONS 

 FROM 50 

 PER CENT 



Non-amphimictic 



Any 



Both 



O 



O 



Autosomal dominant 



Bisexual 



Both 



O 



O 



Autosomal recessive 



Bisexual 



Both 



2-44 



O.768 



Autosomal dominant 



Unisexual 



Both 



1.83 



O.O572. 



Autosomal recessive 



Unisexual 



Both 



8.88 



3-2.7 



Sex-linked dominant 



Bisexual 



JHomozygous 

 \Heterozygous 





 





 



Sex-linked recessive 



Bisexual 



JHomozygous 

 \ Heterozygous 





 1.56 







0.0484 



Sex-linked dominant 



In homozygous sex 



JHomozygous 

 \Heterozygous 





 





 



Sex-linked recessive 



In homozygous sex 



JHomozygous 

 (Heterozygous 



5-34 



18.5 



1.74 

 13.2.8 



Sex-linked dominant 



In heterozygous sex 



JHomozygous 

 [Heterozygous 



1.83 

 3.1Z5 



0.0572. 

 0.0977 



Sex-linked recessive 

 Autosomal dominant 



In heterozygous sex 

 Gametic unisexual 



JHomozygous 

 \Heterozygous 

 Both 







3.12.5 



1.83 







0.0977 

 0.0572, 



Autosomal recessive 



Gametic unisexual 



Both 











tions and on natural selection, and finds 

 that for partial self-fertilization and 

 partial inbreeding, the selection is rapid 

 when the recessives are few. Partial 

 assortative mating and selective fertiliza- 

 tion on the other hand have but slight 

 effect on the rapidity of selection. The 

 case of inbreeding is of importance since 

 the moderate degree of inbreeding which 

 must occur in human populations in such 

 instances as the mating of cousins will 



tion. Inbreeding and self-fertilization 

 cannot be replaced by moderate degrees of 

 selective mating or fertilization. 



In Part III (5), the author considers 

 the cases of a single but completely 

 dominant factor and of several interacting 

 factors. Mating is supposed to be at 

 random, the populations to be very 

 large, and the generations not to overlap. 

 He derives for these cases expressions for 

 the changes caused by slow selection in 



