1246 BIOLOGICAL EFFECTS OF RADIATION 



indicate a difference in stability, similar to that noted by Gowen and 

 Gay (53a) for the so-called "ever-sporting" eye colors. 



Muller (112) has carried out experiments designed to test the effects 

 of different physiological states on the induced-mutation rate. He 

 found no significant differences between fed and starved, or virgin and 

 impregnated females. More recently, Hanson and Heys (65, 66, 68, 69) 

 have published a series of papers on this and related problems, which 

 claim positive differences under various conditions. Their experi- 

 ments await confirmation. One may surmise that the field is in pretty 

 much the state of the general field of X-ray mutation before Muller's 

 work. 



THE QUESTION OF " DIRECT" EFFECTS 



With the foregoing data in mind, the consideration of the mode of 

 action of radiation in producing mutations becomes possible. The 

 apparent linear relation between dosage and effect, the absence of evi- 

 dence of any time factor, and the failure of an obvious effect of tem- 

 perature during radiation, have in general led to the conclusion that the 

 effect of radiation was directly upon the genes, causing intragenic changes. 

 Moreover, specific experiments purporting to answer the question have 

 been carried out by Muller (112) and Timof^eff-Ressovsky (177, 178). 



They have tested the effect of treated cytoplasm on untreated 

 chromosomes, by treating females, and testing the untreated paternal 

 chromosome of the progeny for mutations. No significant differences 

 were found in the large-scale experiments. Timofeeff-Ressovsky 

 (177, 178) has also tested the behavior in later generations of treated 

 chromosomes which had no immediate mutations. Were an indirect 

 effect present, this might be delayed to later generations, so that succes- 

 sive generations from X-rayed progenies would show higher mutation 

 rates. This experiment likewise gave a negative result. 



Another test of a delayed effect consists of the search for small mutant 

 patches in the progeny of treated individuals. Were the mutation 

 delayed, this kind of mosaic should be found very frequently, since only 

 a small number of cells of an individual would show a mutation which had 

 arisen in late stages. Muller (112) reports negative results in this regard, 

 which are confirmed by the later data of Moore (96). Yet in Moore's 

 data there are indications that the problem is by no means solved. 



Moore studied the "fractional" mutations, found as mosaics, of 

 Muller (105). He has determined frequencies of fractional and complete 

 mutations in the progenies of males and females treated at different 

 stages in the life cycle. It is noteworthy, in contrast to the results of 

 Patterson (139) on chromosome breakage, that even in the earliest stages, 

 there is a frequency of fractional mutations comparable to that found 

 when adults are treated. But the histological data of Kerkis (81) show 



