MANNER OF PRODUCTION OF MUTATIONS 519 



his one neutron dose, which by measurement or as calculated from the 

 lethal rate was equivalent to about 2500 r, had an efficiency, when 

 applied to spermatozoa, about 1.25 times as great as X rays in producing 

 translocations, the latter investigators reported that their two neutron 

 doses, of about 3750 and 5000 reps, respectively, had approximately the 

 same efficiency as X rays in causing structural changes. This con- 

 trast in results would indicate a falling off in the efficiency of neutrons as 

 compared with X rays in producing such changes when the dose was 

 increased. Although the latter investigators believed their own results 

 to indicate that neutrons and X rays behave similarly in causing struc- 

 tural changes in Drosophila, their data actually show an even lower fre- 

 quency at their higher dose than would be calculated on a linearity rela- 

 tion from the results for their lower dose, contrary to what X rays 

 would be expected to give. However, the two doses which they used, 

 bearing the ratio 3:4, were too much alike, considering the high statis- 

 tical error, to make a decision concerning the frequency-dosage relation 

 for neutrons possible from their data alone. The first definitive data on 

 the frequency-dosage relation for the production of structural changes by 

 neutrons applied to Drosophila spermatozoa were those reported by 

 Catsch, Peter, and Welt (1944). The doses which they used had a range 

 which overlapped the range of the doses employed by the earlier investi- 

 gators of this matter. The range in the work of Catsch, Peter, and Welt 

 was, however, much wider, the doses tested having the approximate rela- 

 tion 1:2:4. The frequencies of autosomal translocations found at these 

 doses turned out to be very nearly 1, 2, and 4 per cent, respectively, and 

 thus were related in a remarkably linear way to the doses. Proceeding 

 to a higher range of neutron doses, yielding from about 5 to 9 per cent of 

 autosomal translocations, Muller and J. I. Valencia (1951; unpublished 

 data) found that an approximately linear frequency-dosage relation, or 

 one which even seemed to fall somewhat below the linear, holds for this 

 higher range also. 



Since it is readily calculated that, with the doses used and more 

 especially with the higher doses, several or many tracks traverse the 

 same sperm nucleus, which is about 7 n long, these results, like those of 

 Giles, appear to mean both that the breaks are produced within or in the 

 immediate submicroscopic neighborhood of the tracks and that the 

 majority of broken ends, even of chromosomes derived from spermatozoa, 

 eventually undergo union with broken ends that had at the time of 

 breakage been in their immediate vicinity. That is, the pieces do not 

 ordinarily become moved about extensively before they undergo union. 



That there are some exceptions to this rule is shown by the observation 

 of Sidky (1940) that broken ends produced by irradiation of spermatozoa 

 may, on occasion, join even with those arising in chromosomes derived 

 from the mother. In addition, it should be noted that such narrow 



