CHROMOSOME ABERRATIONS IN ANIMALS 665 



that more convincing evidence is needed if the close correlation between 

 break frequency and mitotic chromosome length is to be regarded as 

 merely coincidental, especially since cytogenetic studies have shown that 

 the presumably indivisible blocks in the proximal region of the X may be 

 broken in a number of different places by X-ray treatment (Kaufmann, 

 1944). The general conclusion drawn from these various lines of inves- 

 tigation is that breaks induced in the spermatozoa of Drosophila are dis- 

 tributed at random among the chromosomes and along their lengths, and 

 represent a random sample of the number originally induced by the 

 ionizing radiations. 



In the study of the distribution of breaks along the X chromosome of 

 D. melanogaster it was found that certain intercalary subdivisions, such 

 as 11 A, 12D, and 12E (Table 9-9; see also Kaufmann, 1944, 1946a), have 

 high coefficients of breakage (determined by comparing the number 

 observed with that expected on the basis of length represented in the 

 salivary-gland chromosome). These regions thereby simulate the 

 behavior of the proximal heterochromatin, and it is suggested, in the 

 light of the foregoing considerations, that they also contain hetero- 

 chromatin. The probability that other intercalary heterochromatic 

 regions are scattered along the chromosome is indicated by the essen- 

 tially normal frequency distribution curve obtained by plotting the 

 coefficients of breakage. 



Breaks detected in salivary-gland chromosomes of Sciara after irradia- 

 tion of oocytes are not distributed at random among chromosomes 

 (Bozeman and Metz, 1949). The rearrangements, however, are almost 

 exclusively intrachromosomal (Crouse, 1950), suggesting a recombination 

 pattern different from that of breaks induced in chromosomes of sperma- 

 tozoa of Drosophila. The absence, in the chromosomes of most organ- 

 isms, of the clear pattern of linear differentiation that characterizes the 

 salivary-gland structure makes precise determination of break position 

 much more difficult. There are, however, several reports of nonrandom 

 break distribution. In the ( Jrthoptera some members of the chromosome 

 set appear to be more susceptible to fragmentation than others (White, 

 1935b; Bishop, 1942), and some regions of the chromosomes also appear 

 to be especially fragile (Helwig, 1933, 1938), although they are not 

 necessarily characterized by any special morphological features (Bishop, 

 1942). If it is assumed that breaks are distributed at random among and 

 along these chromosomes, the observed departures from randomness are 

 to be attributed to the subsequent behavior of the breakage ends with 

 respect to restitution or recombination, or to the difficulty of detecting all 

 breaks in later stages of the same mitotic cycle in which they are pro- 

 duced. These aspects of the problem will be discussed later. 



3-ld. Dose-Frequency Relations. Breaks that are detected cyto- 

 logically represent only a part of those induced by ionizing radiations. 



