MANNER OF PRODUCTION OF MUTATIONS 503 



hypothesis, exen if it were admitted that such linearity was to be expected 

 on that hypothesis. So, for example, the results of Villee (1946) on the 

 irradiation of pre-imaginal stages with X rays showed a markedly lower 

 frequency of lethals when the irradiation was protracted over a period of 

 more than an hour than when it was given in l-l^i minutes. Since 

 gene mutations and primary breaks are known to be produced by single 

 ionizing particles and must therefore be independent of the time-intensity 

 characteristic of the dose, the difference found with change in timing 

 must reside in that class of lethals which is connected with structural 

 changes. For these are (except in certain special types of cells, such as 

 spermatozoa) produced in higher frecjuency at higher intensities, as 

 previously pointed out. However, it is only on the position-effect 

 interpretation of these lethals that the addition of them could result in an 

 increase of the over-all lethal frequency, since on that of mutation-by- 

 breakage they would simply be drawn off from a general reservoir of 

 l)reakage lethals that would otherwise appear in the point-mutational 

 category. It is also in accord with the position-effect interpretation that, 

 in recent results of J. I. and R. M. Valencia and Muller (unpublished 

 data), the mutations which are associated with structural changes appear 

 as a group which is added on to the point-mutational ones, as the dosage 

 increases, while the latter taken by themselves remain essentially propor- 

 tional in their frequency to the first power of the dose. As yet, however, 

 these last results are only preliminary. 



It had been thought by Kaufmann and Gay (1946, 1947a, b) that the 

 apparent lack of increase in sex-linked-lethal frequency observed by them 

 in X-ray experiments, in which the frecjuency of structural changes had 

 been markedly increased by application of infrared given as a supple- 

 mentary treatment after the X rays, showed that the additional struc- 

 tural changes did not give rise to additional lethals, caused by position 

 effects, and that the lethals must therefore have been associated with the 

 primary breaks. However, a study of their data indicates that, from 

 the standpoint of statistical error, the numbers dealt with are not suffi- 

 cient to prove this point. At the same time, the increase in over-all 

 lethal frefjuency to be expected in this case on the position-effect interpre- 

 tation is not as great as supposed and cannot at this time be calculated, 

 for the same reasons, already given, as apply when the dosage of ionizing 

 radiation is raised. 



Results different from those of Kaufmann and Gay were obtained in 

 experiments of Glass (1949). In this work, when X rays were applied 

 to Drosophila with and without an infrared treatment, the over-all 

 sex-linked-lethal frerjuency was in fact raised significantly by the infra- 

 red, along with the frequency of structural changes. Moreover, in 

 experiments of Swanson, Hollaender, and Kaufmann (1948) on 

 morphologically expressed mutations in Aspergillus terreus, infrared (as a 



