RADIATION AND THE STUDY OF MUTATION IN ANIMALS 1239 



data are somewhat off, but their numbers are lower. The observations 

 of Gowcn and Gay (53) and of Demerec (22), however, are distinctly 

 different from the others. Demerec has paid particular attention to 

 the measurement of intensity, having recently calibrated his dosimeter. 

 When, however, his corrected dosage values are changed back to the 

 uncorrected readings, his data agree quite well with those of Timofeeff- 



Demerec (0.72 A) 

 EfroimsonC0.22A) 

 [froimson (1.75 A) , 



Gowen and Gay (1.537 A) 

 Gowen and Gay (2.285 A) 

 Hanson, Heys and Stanton 

 Oliver^ 

 J,. Timofeeff-Ressovsky 



1 



I.OOO 



3.000 



9,000 



12,000 



. 6,000 

 Intensity in r Units 



Fig. 6. — Percentage of lethals in the X-chromosome of Droaophila melanogaster in 

 relation to the intensity of radiation. The data are taken from Tablea 19 and 19a; for 

 Demerec's values, the dosages of the table have been divided by 0.61. In the case of 

 Gowen and Gay, the intensities have been divided by 2.44. The controls are not indicated ; 

 they may be found in Table 9. 



Ressovsky and Oliver. One may suspect some systematic error on the 

 part of the manufacturers of dosimeters, or in the calibration value which 

 Demerec obtained from the Bureau of Standards. No such simple 

 explanation can be used for the data of Gowen and Gay, which, however 

 (see Fig. 6), when their dosage values are divided by the factor 2.44, fall 

 into the same group as the other data. 



The interpretation of the curve is not entirely obvious. If a single 

 gene only were considered, the explanation generally used in the work 

 with short-wave radiation (Crowther, 12a; Glocker, 48a), based on the 

 change in the probability of a quantum hit at different dosages, would be 

 quite simple. But in a population of genes, the problem becomes more 

 complex. The evidence presented previously indicates that different 



