NATURE OF THE GENETIC EFFECTS 361 



already present in just one chromosome) lead to the induction of a tumor 

 by radiation. However, it is probable that in most organisms other than 

 Diptera (flies) crossing over would be much less readily induced, if at all, 

 in somatic and gonial cells, than has been found to be the case in Drosoph- 

 ila. For Diptera are peculiar in having, even normally, an exception- 

 ally strong tendency to synaptic association of chromosomes in the 

 ordinary somatic cells and in the nonmeiotic germ cells, and this evi- 

 dently makes crossing over much more readily possible in this material. 

 The regions of chromosomes near the centromere, and, to a lesser 

 extent, in some material, those near the tips and some other small inter- 

 stitial regions, are distinguished by their mode of staining and by certain 

 other properties (vide infra). They are designated as heterochromatic, in 

 distinction from the rest of the chromosome, which is termed euchromatic. 

 The promotion of crossing over by radiation, which occurs most markedly 

 in the vicinity of the centromere, extends not only to the heterochromatic 

 region in that location but considerably beyond it, to euchromatic regions 

 for some distance on either side of it, only gradually fading away. More- 

 over, it is to be found also in some heterochromatic regions derived from 

 the region originally near the centromere when they have by special 

 means been removed so as to be far from the latter. 



The influence of radiation on the centromeric and other heterochro- 

 matic regions leads to the occurrence of recombination between the 

 homologous or partially homologous heterochromatic portions of the X 

 and Y chromosome even in the Drosophila male, despite the fact that in 

 the Drosophila male (unlike the male of most organisms) crossing over 

 does not ordinarily occur at all. Thus the radiation in this case results 

 in chromosomes composed partly of X and partly of Y and, by a second 

 step, to combinations of two X chromosomes attached together. Such 

 combinations, which also occur without radiation, as shown by Philip 

 (1935), but with far lower freciuency, had previously been ascribed to 

 mere chromosome breakage, to fusion (Stern, 1926), to translocation 

 (Stern, 1927), or to uncompleted division (L. V. Morgan, 1922). It is 

 probable that the line of demarcation between crossing over and structural 

 change of chromosomes, caused by their breakage and recombination 

 at nonhomologous points (seep. 362), is not a sharp one where hetero- 

 chromatic regions are concerned, since in such regions the genes at differ- 

 ent loci behave more nearly as homologues than they do in euchromatic 

 regions. At any rate, there is probably a good deal of leeway in the 

 positions of breakage of the two participating chromosomes, relative to 

 one another, when the "crossing over" is located in a heterochromatic 

 region. Moreover, even the union of pieces is in such regions less orderly, 

 being frecjuently reverse in arrangement, so that from A BCD and ahcd 

 the combinations ABba and dcCD can be formed. Such cases therefore 

 may be regarded as transitional to those next to be discussed. 



