666 RADIATION BIOLOGY 



Some of the breaks primarily produced are eliminated by restitution, or 

 restoration of the original sequence of parts, so that the sites of the 

 original lesions are not recognizable in the repaired chromosome (see, for 

 example, Carlson, 1941a; Faberge, 1940; Sax, 1940). Calculations by 

 Lea and Catcheside (1945), based on a limited number of observations, 

 indicate that about a third of the breaks induced in the X chromosomes 

 of spermatozoa of Drosophila by a 3000-r X-ray treatment undergo 

 restitution. It has been suggested that all recessive lethal mutations not 

 associated with cytologically detectable chromosomal aberrations repre- 

 sent restitutional breaks (Lea and Catcheside, 1945 ; Lea, 1946) ; but 

 Fano (1947) and Muller (1950a) have noted the difficulty of reconciling 

 this theory with the total data available for the production of recessive 

 and dominant lethals at different dosage levels. 



Breaks that do not undergo restitution are detectable, either because 

 they remain "open" or "unhealed," or because the breakage ends join 

 inter se or participate in structural rearrangement. Several lines of 

 experimental evidence indicate that the detectable breaks are scattered 

 at random within and among cells, and therefore presumably represent 

 a random sample of those originally induced. 



Single breaks, such as those that separate fragments from the centro- 

 meric portions of the chromosomes, are produced with frequencies pro- 

 portional to the dose, at least when the dose is sufficiently low that the 

 mean number of breaks per cell is less than one (Fano, 1941). The dose- 

 frequency relation is expressed by the formula N/N = kD, in which N 

 is the number of breaks occurring in iVo cells, D the dose, and k a propor- 

 tionality constant. When the frequency N/N approaches 100 per cent, 

 the deviations from linearity that occur can be attributed to the produc- 

 tion of more than one break per cell. The linear relation of breaks to 

 dose suggests that breakage is due to a single "hit," use of this term being 

 subject to the various qualifications that have been outlined by Muller, 

 Fano, and others (summarized in Fano, Caspari, and Demerec, 1950). 



The experimental results are well illustrated by Carlson's study (1941a) 

 of fragment production in neuroblast cells of the grasshopper, Chortophaga 

 viridifasciata, by X-ray treatment. The distribution of the fragments 

 among nuclei was found to conform to a Poisson distribution. The 

 number of breaks was proportional to the amount of the dose within 

 the range tested, which extended from 7.8 to 125 r (Fig. 9-13a). Appar- 

 ently there is no threshold below which breaks are not produced; it thus 

 appears to be an all-or-none effect. 



In Drosophila the increase in frequency of dominant lethals induced by 

 X rays, as determined from the proportions of eggs that fail to hatch, is 

 essentially linear at lower dosage levels, indicating that the single-break 

 type of aberration is primarily involved (Demerec and Fano, in Demerec, 

 Kaufmann, Sutton, and Hinton, 1940; Demerec and Fano, 1944; Catche- 





