216 Miscellaneous Linkage Topics 



throughout the entire length of the autosomes of Drosophila 

 instead of being restricted to certain small, highly localized seg- 

 ments. Cooper also suggests that these conjunctive segments 

 are nongenic, chromosomal organelles comparable to the cen- 

 tromere, matrix, and nucleolus organizers. He has named them 

 collochores. 



Somatic Crossing Over 



Is crossing over purely a phenomenon of meiosis or does it also 

 occur in somatic cells? In most organisms, there is no evidence 

 that somatic crossing over occurs, and it is hardly to be expected 

 since the two chromosomes of a homologous pair do not attract 

 one another at prophase and since they line up on the equatorial 

 plate independently of one another. In Drosophila melanog aster, 

 however, Stern has produced data that can hardly be explained 

 except by assuming somatic crossing over. The opportunity for 

 such an occurrence w^ould be greater in Drosophila than in most 

 organisms because in the Diptera the homologous chromosomes 

 tend to lie alongside one another in ordinary somatic cells. 



Somatic crossing over has been found between the genes y 

 (yellow body) and sn (singed bristles). The locus of y is at the 

 distal end of the X chromosome about sixty-six map units from 

 the centromere, whereas sn is twenty-one map units from yel- 

 low, on the side nearer the centromere. Somatic crossing over 

 is increased by the presence of certain dominant genes known 

 as the minutes. If a fly is heterozygous for y and sn, having 

 received one chromosome with y and sn from one parent and 

 one chromosome with the two wild-type alleles from the other 

 parent, crossing over may occur between one chromatid of one 

 somatic chromosome and one chromatid of the homologue (Fig. 

 68). The two chromosomes may then line up on the metaphase 

 plate in one of two ways. If they are so oriented that a + + 

 chromatid and a y sn chromatid go to one pole and the y + and 

 + sn chromatids to the other pole, both cells will be wild-type 

 with respect to each pair of alleles. On the other hand, if a + + 

 and a + sn chromatid go to one pole and a ?/ + and a y sn chro- 

 matid to the other, the latter will form a cell which has no wild- 

 type allele for y. Such a cell would divide a number of times 

 to produce a group of such y cells. They would be observable 

 phenotypically as a small area of yellow-colored body in the 



