MANNER OF PRODUCTION OF MUTATIONS 521 



ment of Drosophila spermatozoa, it was shown by Timofeeff-Ressovsky 

 and his co-workers (Timofeeff-Ressovsky, 1938; Timofeeff-Ressovsky and 

 Zimmer, 1938; Zimmer and Timofeeff-Ressovsky, 1938; Zimmer, 1938) 

 that neutrons have only about two-thirds the efficiency of ordinary X rays 

 in producing these changes, when doses of the same ionizing capacity in 

 tissue are compared. At the same time, the frequency relation of the 

 lethals to neutron dose, like that to X-ray dose, was found to be linear, 

 up to the highest dose used, which gave about 3 per cent of lethals. The 

 lesser efficiency was confirmed by Dempster (1941b), by Demerec, 

 Kaufmann, and Sutton (1941), by Giles (1943), and by Fano (1942), and 

 the linear relation to dose has recently been extended to a frequency of 

 9 per cent by MuUer and Valencia (unpublished data). Similarly, with 

 a rays applied to spores of Aspergillus, Stapleton, Hollaender, and Martin 

 (1952) have found for visible mutations a distinctly lower efficiency than 

 for X rays and a linear relation to dose. 



Timofeeff-Ressovsky and Zimmer, and later Lea and Catcheside, inter- 

 preted the lesser efficiency of more densely ionizing radiation to mean 

 (to paraphrase their discussion) that the mutagenic pathway of an 

 ionization or activation is so short that, with neutrons, about two out of 

 every three ionizations (or activations or natural clusters) which would by 

 themselves have caused a mutation, fell within mutagenic range of a gene 

 already being caused to mutate by one or more other of these activating 

 agents, one-third of them thereby becoming superfluous, whereas with 

 X and 7 rays the predominantly wider spacing of the active particles or 

 groups allowed very little of this overlapping of their effects. It may be 

 observed that a possible alternative explanation of at least a part of the 

 lowered effectiveness might have been postulated to lie in a speedier 

 neutralization of negative ions by positive ions where, as with neutrons, 

 they were more crowded, with the consequent removal of some of them 

 before they had a chance to exert a mutagenic influence (as postulated by 

 Lea, 1947, in a different connection). However, a more detailed analysis 

 of neutron mutagenesis (MuUer and J. I. Valencia, 1951 and unpub- 

 lished data) has indicated that neither of these two possible interpreta- 

 tions is the correct one but that, nevertheless, the conclusion must be 

 drawn that the ionizations or activations produce the mutagenic effects 

 observed within an ultramicroscopic distance of the points at which they 

 originated. 



In the work last referred to, mutations involving specific loci known to 

 affect visible characters were obtained by the method, already used for 

 other radiation, of crossing irradiated males having normal genes at those 

 loci to females having viable fertile mutant alleles at the same loci. 

 Thus the mutations found and bred which showed visible changes involv- 

 ing the given loci included not only the viable fertile types but also any 

 cases in which the new mutant would, by itself, have been lethal or 



