INDUCED CHROMOSOMAL ABERRATIONS IN ANIMALS 1187 



genetic map. Thus, the distance between the third chromosome genes 

 sf and cu (Figs. 4 to 6) is equal to six map units or about one-eighteenth 

 of the total length of the genetic map. The st-cu interval, however, is 

 no less than one-fifth of the chromosome cytologically. The genetic 

 distances between the second chromosome loci It, rl, and tk are very small, 

 being equal to fractions of one map unit, or less than one-hundredth 

 of the genetic map. Cytologically this region amounts to about one- 

 fourth of the chromosome. The h-lt interval is shorter than one-fifteenth 

 of the length of the genetic chromosome, but longer than one-fourth 

 of its cytological map. 



The distances between the genes located near the middle of either 

 limb of the second and the third chromosomes are relatively much longer 

 genetically than cytologically. For instance, the cn-px and the dp-h 

 intervals in the second chromosome constitute each about one-third 

 of the whole genetic map, but cytologically the length of these intervals 

 is almost negligibly small (Figs. 4 and 5). Finally, the distances betw^een 

 the genes lying close to the right end of the second chromosome, and 

 probably also close to its left end, are again somewhat longer cytologically 

 than expected on the basis of the genetic map (the px-sp and bw-sp 

 intervals. Fig. 5). Whether the same relation holds true for the third 

 chromosome is as yet not clear, because the number of known breaks in 

 the end regions of this chromosome is too small. 



The cytological location of genes in the small fourth chromosome was 

 studied by Bolen (11). In a translocation this chromosome was broken, 

 the fiberless fragment became attached to the X-chromosome, and a 

 fragment of the X became attached to the remaining part of the 

 fourth which preserved its spindle fiber. Duplications for either part of 

 the fourth chromosome are viable, and this enabled Bolen to show that 

 the gene bent is located closer to the spindle fiber than the gene eyeless 

 (Fig. 4). 



As suggested above, the discrepancies between the genetic and the 

 cytological determinations of the relative distances between genes are 

 due to the variable frequency of crossing over per unit length in the 

 different parts of the chromosomes. In spite of the fact that the present 

 knowledge of the cytological maps leaves much to be desired, a comparison 

 of the genetic and cytological maps of the respective chromosomes even 

 now suggests certain interesting generalizations. In the vicinity of the 

 spindle fiber attachments the frequency of crossing over per unit of 

 the absolute distance is low, and consequently the genetic distances are 

 much less than the cytological ones (the inert region of the X, the b-cn 

 interval in the second, and the D-cu interval in the third chromosome). 

 In the regions more remote from the spindle attachments, but not too 

 close to the ends of the chromosomes, the frequency of crossing over is 

 high, hence the genetic distances are relatively greater than the cytologi- 



