1320 BIOLOGICAL EFFECTS OF RADIATION 



lengths. Between the curve for chromium and the curve of Demerec for 

 tungsten He the results of the other investigators. This broad variation 

 from a curve of a slope of g-°»"*-^*- ^q one of g-'^" ^•*"- naay ^n part be 

 accounted for by inaccuracies in measuring the ionization dose. This 

 explanation does not seem likely for the data of Hanson on gamma rays, 

 of Demerec with tungsten, or of those for the copper and chromium. 

 Two other possibilities of accounting for this variation suggest themselves : 

 (o) The absorption of energy within the fly tissues before the sperm are 

 reached may be rather high in the long wave-lengths. This fact could 

 bring this material nearer the observations on the heterogeneous rays of 

 tungsten, although in view of the direction of the differences between the 

 copper and chromium results, this interpretation would appear unlikely. 

 It will furthermore not account for the gamma-ray data. (6) The second 

 possibility is that chromatin may have a band specific absorption for 



o 



X-rays in the tungsten region, 0.7 A. This, too, seems unlikely. The 

 explanation of these variations needs more extended data from the same 

 laboratory where under like conditions radiant energy over a wide range 

 of wave-lengths may have its effects determined. 



The curves of Fig. 3 are all of a form which suggest that a single 

 absorption in a gene is sufficient to produce a mutation. This fact is 

 further borne out by genetic evidence. The mutation of one gene, 

 although it is itself small, generally has no effect on other genes so far as 

 the writer's evidence has shown (25). The occurrence of other mutations 

 within a chromosome is the result of other random absorptions. The 

 development of the sperm (although it is generally haploid) sometimes 

 leads to the growth of the genes and their splitting into two new ones. 

 Evidence is available to show that despite the close proximity of these two 

 genes one may be affected by X-ray while the other remains unchanged 

 (25, 59, 61). The average sphere of action of the rays is therefore small, 

 as it should be by physical theory. Different genes within the chromo-" 

 some are, of necessity, close together. Muller (66) has stated, although 

 without presenting data, that in his opinion a gene next to one which 

 mutated also mutated more frequently than it should by chance. From 

 this fact he has argued that mutation may be in the nature of a chain 

 reaction where the energy only starts a process capable of spreading to 

 other close genes. He considers that the ions, even taking the chance 

 straight path which they sometimes do, would be unable to reach both 

 genes because the distance is thought to be too great. The data on 

 which this hypothesis is built are quite naturally scanty at this stage of 

 the work. In the writer's own research, where the foregoing possibility 

 was clearly in mind during the experiment, no evidence in its support 

 was obtained. For the present at least the direct-action theory as the 

 method by which the X-rays produce their effects seems adequate. 



