360 RADIATION BIOLOGY 



in conformity with the principle that, in the regions near the centromeres 

 of all the major chromosomes, crossing over is considerably promoted by 

 irradiation, as others have also shown it to be by extremes of heat and 

 cold, chemical mutagens, and some other environmental as well as physi- 

 ological and genetic influences, while in regions farther from the centro- 

 meres, except perhaps near the very ends, a less marked decrease in 

 crossing-over frequency is produced. The decrease is in part, at least, 

 an indirect consequence, resulting from interference, of the promotion of 

 crossing over elsewhere. It may be that the decrease in the inhibiting 

 influence of the centromere on crossing over, caused by radiation, is 

 related to that decrease in the accuracy of transportation of chromosomes 

 which is expressed in their lagging and nondisjunction. 



The promotion of crossing over by ionizing radiation (as well as by 

 some other influences) is so strong that it even leads, in Drosophila, to 

 occasional crossing over in spermatogonial cells, as shown by Friesen 

 (1933, 1936) and in somatic cells, and in these cases, too, the induced 

 crossing over is mainly, although not exclusively, near the centromere, 

 just where it is of least frequent occurrence normally during meiosis. 

 There is evidence that, as in normal crossing over, this induced crossing 

 over occurring in gonial and somatic cells usually involves only two of 

 four chromatids of a tetrad, and that the two participating chromatids 

 undergo breakage at exactly corresponding points. Since it is not fol- 

 lowed by meiotic divisions, the daughter and descendant cells are still 

 diploid. However, they may by this process of crossing over become 

 alike with respect to genes of which the homologous chromosomes origi- 

 nally carried different alleles. Thus if genes in the original paternal and 

 maternal chromosomes are represented as A BCD and abed, respectively, 

 the centromere being at or near D, a somatic cell may at the mitosis fol- 

 lowing crossing over receive a noncrossover chromatid with abed and a 

 crossover chromatid of composition abCD. In that case the genes a and 

 b, which previously had been unable to express themselves effectively 

 because of the simultaneous presence of the dominant alleles A and B, 

 will now be able to produce their characteristic effects, when in the 

 appropriate region of the body and stage of development. 



If, then, the crossing over has occurred in an embryonic somatic cell, 

 or in any cell subject to further proliferation, a portion of the body or 

 patch of tissue may thereby come to exhibit recessive characteristics 

 not shown by the body as a whole. This turns out to be, in Drosophila, 

 the chief mechanism for the appearance of such patches following irradi- 

 ation, although it had earlier been thought that they were usually pro- 

 duced by loss of chromosome parts. This is one method by which irradi- 

 ation can bring preexisting but hidden mutations to light, in a patch of 

 somatic tissue or portion of the body. It is evident that this mechanism 

 might sometimes (when a recessive gene of the appropriate kind is 



