NATURE OF THE GENETIC EFFECTS 427 



of the amount of detriment it produces, as measured by the average risk 

 of genetic death which it gives rise to in individuals carrying it. For this 

 reason the total "load" or disability imparted to the population, in the 

 course of a succession of generations, is in the end as much for a mutant 

 gene which has a slightly detrimental effect as for one which has a highly 

 detrimental or fully lethal action on the individuals possessing it. The 

 average amount of this total load may be said to add up to unity, i.e., 

 the complete disability of one individual, for each detrimental mutant 

 gene which is received by one offspring of the first generation after its 

 origination.^ This unit total load is in the case of slightly detrimental 

 genes distributed in smaller fractions per individual but over correspond- 

 ingly more individuals, than it is in the case of more markedly detrimental 

 genes. 



In the above treatment no consideration was given to whether the 

 mutant gene was dominant or recessive ; hence the value for impairment, 

 i, as used, had to be taken as an average for the effect in all individuals 

 having the gene, including both those heterozygous and those homozygous 

 for it. The value was therefore in part dependent on the numbers of 

 these types relative to each other. Actually, however, it is usually the 

 amount of impairment in the heterozygote which plays the decisive role. 

 It can be shown that most mutant genes very seldom become homozygous, 

 since they exert enough detrimental effect (even though this is very 

 slight) when heterozygous to become eliminated before they have a 

 chance to exist in homozygous condition. In this sense they are said to 

 be effectively dominant (MuUer, 1950b). 



In Drosophila, genes that are lethal or nearly lethal in homozygotes 

 commonly produce an impairment of some 2 to 7 per cent, or roughly 

 3.^5, in heterozygotes (Stern et al, 1948, 1951, 1952; MuUer, 1950b, c). 

 Thus for the lethals in heterozygotes p — 25; i.e., each such gene tends 

 to pass down through some 25 individuals, on the average, before being 

 eliminated. As the frequency, among the germ cells of the general 

 population, of already existing lethals occupying any given locus is 

 usually far less than 1 in 25 (as will be shown in what follows), the 

 chance is small that before its elimination any given lethal will meet 

 another of the same kind at fertilization and thereby become homozygous. 

 Much the greater part of the damage done by such genes in the popula- 

 tion is therefore made up of the collective impairment exerted by them 

 in heterozygotes. There is reason to infer that those mutant genes, 

 much more numerous than lethals, which are less detrimental than 

 lethals to individuals homozygous for them, have in general relatively 

 more expression in heterozygotes, as compared with that in homozygotes, 

 i.e., that they have a greater tendency toward dominance than the lethals. 

 Moreover, this tendency is probably more marked in man than in Dro- 



^See opposite page for footnote. 



