ESTIMATES OF GENETIC HAZARD OF RADIATION 437 



tation rate per gamete per generation, in so far as this is adjustable in 

 the course of evolution, or that the rate would be lower in man, where 

 there is less wastage of offspring between fertilization and maturity. 

 Even if the spontaneous rate per gamete and generation in man is taken 

 to be as high as that in Drosophila, this figure is still only one-hundredth 

 of that assumed by Evans, or, holding the other factors constant, 10~^ 

 per locus per generation. Wright points out that there is no incompati- 

 bility between rates of the order of 10~^ for hemophilia and epiloia and 

 a rate of 10"'^ for the average of all loci. 



If, as Wright claims is possible, the average spontaneous mutation 

 rate per locus in man is only one-hundredth of that assumed by Evans, 

 and if the other factors assumed by Evans are correct, then the dose 

 required to double the rate of mutation in man is also only one-hun- 

 dredth of that calculated by Evans, namely, 3 r. Both this figure and 

 the 300 r obtained by Evans are based on the rate of induced mutation 

 in irradiated sperm. Irradiation of early germ cell stages would, on the 

 basis of results in Drosophila, be perhaps only from one-half to one- 

 fifth as effective. 



Regardless of the evolutionary considerations raised by Wright, it 

 seems to the present author that the product of the figures assumed by 

 Evans for number of loci and spontaneous recessive lethal mutation rate 

 in man is too high. A consideration of the effect it would have on the 

 offspring of first-cousin matings, for example, will show that it can 

 hardly agree with the facts. The value taken by Evans for the average 

 spontaneous rate of mutation to recessive lethals per locus and genera- 

 tion in man has already been given and is a = 10~^. He accepts the 

 view that A^, the number of loci in man, probably lies between 10^ and 

 10^. His figure for a minimum estimate of the accumulation factor is 

 771 = 50. Accordingly, the average number of recessive lethals per gam- 

 ete would be aNm = between 5 and 50. Individuals would, therefore, 

 on the average, be heterozygous for from 10 to 100 lethals, and the two 

 common grandparents of first cousins would together be heterozygous 

 for from 20 to 200. The probability that a particular gene heterozygous 

 in one of the common grandparents will not become homozygous in an 

 offspring of first-cousin mating is 63/64. Ignoring linkage, and lethals 

 common to the two grandparents, the probability that the child of a 

 first-cousin mating will not be homozygous for any of the 20-200 lethals 

 heterozygous in the grandparents is, sufficiently accurate for our pur- 

 pose, (63/64)20 to (63/64)-oo, or 0.73-0.043. Thus, according to the 

 limiting values assumed by Evans, the percentage of death as a result 

 of consanguinity in the offspring of matings of first cousins who have 

 average grandparents would be from 27 to 96 per cent. 



