506 RADIATION BIOLOGY 



Still clearer in this respect are the results on visible mutations produced 

 by X-ray treatment of spores of Aspergillus by Stapleton, Hollaender, and 

 Martin (1952) . For in this case the frequency of the mutations, although 

 showing a linear relation to dose at low doses, unmistakably rises with 

 ever-increasing slope at high doses. This result, especially when con- 

 sidered in connection with the continued linearity found in the same 

 material with a rays, can hardly be interpreted otherwise than as being 

 due to the combined effect of point mutations, which preserve a linear 

 relation with dose, plus position effects of structural changes which, 

 increasing as a power of the dose higher than 1, form an ever larger propor- 

 tion of the total effects at higher doses. In other words, this material 

 gives just the curve which had been depicted by sponsors of the mutation- 

 by-breakage view as that to be expected on the view which they opposed: 

 that of position effect. It would seem reasonable to refer the difference 

 between these results and those obtained in Drosophila to variations in the 

 relative frequencies of different classes and to selection rather than to the 

 existence of fundamentally different mechanisms in the two cases. 



Despite the arguments which have been given, it is still entirely 

 possible, as first suggested by Muller (1941a) and recently emphasized 

 by Herskowitz (1951), that some primary breaks of chromosomes which 

 later restitute are associated with permanent mutational changes in genes 

 located at or near the point of breakage. Moreover, if primary breaks 

 were so very numerous as on one of the views suggested by the ring- 

 * chromosome data, the association of only a small proportion of them with 

 gene mutations could, nevertheless, result in a fairly high frequency of the 

 latter. Yet this would not result in a strong tendency to depression of 

 the lethal frequency below Hnearity at high doses because only a small 

 proportion of these numerous primary breaks would ever become lost to 

 view through aneucentric structural changes. Hence the loss in lethal 

 frequency at high doses, thereby brought about, would be far less than 

 the increase in this frequency brought about by the large added class of 

 position-effect lethals ; for the evidence of the reality of this class is con- 

 vincing. Additional factors such as those concerned with the fate of small 

 deletions would then have to be invoked after all in the explanation of 

 an over-all linear relation at high doses. 



Still other facts would, however, remain to be brought in Hne mth even 



sity curve appear to be a straight line. At low doses the intensity factor makes very 

 little difference, and it may therefore be assumed that the results for the high intensity 

 would at the low doses have been about the same as for the low intensity. If the 

 low-dose results for low intensity are corrected in such a manner as to make them fit in 

 with the rest of the low-intensity curve and these values are used to represent also 

 the low-dose results for high intensity, a curve is obtained for high intensity which 

 ascends with increasing slope, just as would be expected from the fact that it shows 

 time dependence. 



