374 



AN INTRODUCTION TO MODERN GENETICS 



in the same chromosome have a greater effect in reducing the number 

 of facets in Drosophila eyes than they have when they lie in the normal 

 position of one in each homologue. The phenomenon was a very 

 peculiar one; in a homozygous Bar stock, normal individuals appeared 

 at the same time as individuals with the extremely reduced eyes which 

 were attributed to the fact that both Bar genes had entered the same 

 chromosome. Sturtevant showed that the phenomenon involved an 



T V 



Bt> Ib 



i 4 ^m 



^ 



Back Mutation Bar 



BB 



■v.^-ir. 



j^S^I]-* h S^S - 



BB — ^ 



NormoL 



Ocuble 8ar 



Fig. 151. Mutations of the Bar Locus. — At the left are diagrams of the mutation 

 from homozygous Bar to heterozygous double-Bar (the latter reduces the size 

 of the eye still more than the former). Above is the old interpretation, when Bar 

 was thought of as a definite gene; the mutation involves a crossing-over and was 

 supposed to be due to this crossing-over being unequal. Below is the modern 

 scheme, according to which Bar is itself a duplication; the unequal crossing-over 

 causes three of the duplicated segments to come into the double-Bar chromosome. 

 At the right are the salivary gland chromosomes. 



(From Bridges.) 



unequal crossing-over, in which one chromosome broke to the right of 

 Bar while the other broke to the left, so that one of the cross-over 

 chromatids would apparently have no Bar gene while the other had 

 two. There was considerable discussion as to the possible bearings of 

 this process on the origin of new genes during evolution, but recent 

 work has somewhat reduced the importance of the case in this con- 

 nection. It turns out that the Bar "gene" is actually a small duplication, 

 the duplicated fragment lying immediately beside its homologue.^ The 

 unequal crossing-over is therefore not difficult to understand, and the 

 chromatid with "two Bar genes" really has three of the duphcated 

 segment, that with "no Bar gene" is perfectly normal with the segment 

 represented only once. 



^ Bridges 1936, Muller, Prokofieva, and Kossikov 1936. 



