502 EADIATION BIOLOGY 



being associated with lethals, is too great to be reconciled with the seem- 

 ingly linear relation shown by the data (Fano, 1947; MuUer, 1950). 



If, however, the position-effect interpretation is accepted for the great 

 majority of the lethals which are associated with nondeficient structural 

 changes, the question still remains, why does the over-all frequency of 

 lethals not rise, at higher doses, increasingly above the frequency expected 

 on a linear relation? A part at least of the answer to this lies in the 

 considerations set forth in the preceding paragraph, which show that 

 there must be an increasing reduction at higher doses in the frequency of 

 recoverable two-break deletions, relatively to the total number of them 

 produced, because of their becoming increasingly drawn into aneucentric 

 combinations. This influence would continue to operate even if lethals 

 were seldom or never associated with primary breakage. Whether it 

 would be of sufficient strength to counteract the entire rise above linearity 

 at higher doses, which the addition of the position-effect lethals would 

 tend to produce, so as to result in the observed approximately linear rela- 

 tion, is a matter that cannot at present be calculated, owing to the imt 

 that not enough is known of the relative frequencies of different types of 

 changes. Moreover, the answer might well be different in different types 

 of material. 



For the purposes of such a calculation much would depend on how fre- 

 quent very small deletions are, including even those deletions which are 

 too small to be definitely visible cytologically. There are certain results 

 on mutations at "visible" loci (Valencia and Muller, 1949; unpublished 

 data) indicating that these ultraminute deletions may form a high propor- 

 tion of lethal point mutations induced in Drosophila. In addition to 

 these results is the fact that many of the lethals produced by ionizing 

 radiation in the single X chromosome of spermatogonia act detrimentally 

 on the proliferation of the individual cells containing them, even when 

 these are supported by a normal organism, as evidenced by the strong 

 "germinal selection" which has been found to operate against the pro- 

 liferation of germ cells containing them (see pp. 408-409) . It would be 

 rather surprising for such a high proportion of lethal gene mutations to 

 have so pronounced an effect on the vigor of individual cells. If, how- 

 ever, ultraminute two-break deficiencies do form a considerable propor- 

 tion of the lethals produced by ionizing radiation, then their influence, 

 acting in the manner which has been described, might well go far, when 

 taken in connection with the opposite influence exerted by position-effect 

 lethals, in explaining the apparent linearity of the frequency-dosage 

 curve. 



In the opinion of the author, the exact linearity of the relation between 

 lethal frequency and dose at high doses of ionizing radiation has not yet 

 been sufficiently well establisljed in Drosophila to serve as the basis of 

 far-reaching conclusions such as those of the mutation-by-breakage 



