518 RADIATION BIOLOGY 



work the important finding was made that not only is the frequency of 

 single breaks of chromosomes and chromatids proportional to the first 

 power of the dose, as it had already been found to be with X and 7 rays, 

 but that even gross exchanges and small deletions produced by neutron 

 treatment follow this rule, instead of varying with some higher power of 

 the dose as they do with X and 7 rays. Moreover, in the production of 

 these changes by neutrons, unlike what had been found for X and 7 

 rays, the timing or intensity of the exposure proved to be a matter of 

 indifference. 



As Giles pointed out, both the linear relation and the time independence 

 show that with neutron treatment the two breaks of a structural 

 rearrangement are produced by ionizations or activations belonging to 

 the track of the same projected particle. It could be calculated, how- 

 ever, that the doses used were sufficiently large and the exposure time in 

 much of the work was sufficiently short to result in several or many of 

 these tracks falling within the same nucleus in the critical interval before 

 much union of broken ends could have taken place. It follows therefore 

 that the broken ends derived from the same track must have joined 

 selectively with one another rather than with those from different tracks. 

 This showed, first, that the breaks must have been produced within or close 

 to those tracks which caused them and, second, that, in this material at 

 least, broken ends join almost exclusively with others in their immediate 

 vicinity. That is, there is a narrow spatial limitation both in breakage 

 and in joining. 



The linear frequency-dosage relation for structural changes involving 

 two or more breaks would not necessarily be expected to hold when 

 Drosophila spermatozoa were irradiated with neutrons; for, as has been 

 noted previously, the breaks produced in Drosophila spermatozoa are 

 retained until after fertilization, at which time the movements of chromo- 

 some parts should give better opportunity than in ordinary interphase 

 cells for broken ends which had arisen at relatively distant points to 

 come into contact. This would tend to make the frequency of struc- 

 tural changes in this material depend on a power of the dose higher than 

 1 with neutrons, much as it does with X rays, even if there was as rigorous 

 a spatial limitation on the process of breakage in this material as in 

 Tradescantia. However, evidence at this time, to be cited, if taken at its 

 face value, appears contrary to such freedom of movement before union 

 since the relation has been found to be approximately linear for neutrons, 

 even when they are applied to Drosophila spermatozoa. 



That the frequency-dosage relation for structural changes produced by 

 treating Drosophila spermatozoa with neutrons is different from that 

 obtained with X rays might already have been suspected from the con- 

 trast between the findings of Dempster (1941b) and those of Demerec, 

 Kaufmann, and Sutton (1941); for whereas Dempster had reported that 



