NATURE OF THE GENETIC EFFECTS 419 



ceptible rate, although probably not at as high a rate per unit of elapsed 

 time as during the other two, namely, that of the mature spermatozoa. 

 This is demonstrated by aging the spermatozoa, for when this is done it is 

 found that the older spermatozoa carry more mutant genes. 



The above striking correspondence between the periods of highest 

 mutation rate and those of highest mitotic activity strongly suggests 

 that either the process of gene reproduction itself, or at any rate some 

 feature or features of the heightened metabolism associated with pro- 

 liferation, are somehow conducive to the occurrence of mutations. Some 

 evidence of this correspondence exists also in results from human material, 

 if we may accept Haldane's (1947) calculation (based on certain Danish 

 data on the incidence of a human mutation) indicating that the frequency 

 of newly arisen mutant genes is much higher among human spermatozoa 

 than among human eggs, inasmuch as there has been much more active 

 proliferation in the production of the former than of the latter. On the 

 other hand, the accumulation of mutations in aging spermatozoa is an 

 example of a contrary kind, indicating that spontaneous mutations can 

 also occur, at least under given conditions, when the genes are (as shown 

 by Muller and Settles, 1927) in a state of dormancy. 



One of the purposes of the experiment on aging in Drosophila was to 

 attempt to throw light on the question, raised by the present writer 

 (1928c), whether gene mutations consist in "mistakes" made in the 

 synthesis of daughter genes by their mother genes, or in alterations of the 

 already completed genes, or whether mutations of both types occur. 

 The positive correlation bet\veen mutation frequency and mitotic fre- 

 quency, both in Drosophila and apparently in man, might be interpreted 

 as an indication of the mutations in that case being mainly of the first 

 type, although it might alternatively be supposed that the greater 

 metaboHc activity of growing cells was in some way conducive to muta- 

 tions in the completed genes. On the other hand, the apparent accumula- 

 tion of mutations in mature spermatozoa would seem to indicate that in 

 this case already formed genes had been changed. Yet it might, alterna- 

 tively, be supposed in this case that the mutations occurred only after 

 fertilization, in the process of construction of the daughter genes, under 

 the influence of mutagenic substances accumulated during aging. It is 

 true that on that supposition the mutant and normal genes would be 

 mosaically distributed in the resulting individual, but the evidence as to 

 whether or not this is the case is still lacking. Hence none of these experi- 

 ments are conclusive so far as this problem is concerned. 



More direct light on the question has been thrown by experiments with 

 bacteria. In 1944 Zamenhof reported a correlation of mutation fre- 

 ciuency in bacteria with frequency of reproduction. On the other hand, 

 Novick and Szilard (1950), by means of highly refined methods involving 

 the use of their "chemostat," succeeded in showing that, when reproduc- 



