24 MUTATION AND PLANT BREEDING 



nomenon in Drosophila (15, 16), and (b) to explain the occurrence 

 of noncrossover alpha cases, the sister strand exchanges would have 

 to be unequal. More recently, evidence has been presented (21) in 

 support of the occurrence of sister strand crossing over in maize, but 

 it is apparent that the question of the existence of the phenomenon 

 is still debated among geneticists. In any case, if it is argued that 

 crossing over is a function of the reduplication process or that it 

 occurs at the time of reduplication of strands, it is difficult to visualize 

 how old and new member elements of the A h complex are, as 

 required, in precise juxtaposition at the time of and immediately 

 following replication, yet at the same time are in oblique association 

 to facilitate the required unequal sister event. Moreover, according 

 to the sister strand hypothesis, there is no apparent basis on which 

 to explain the enhancing effect of the deficient homologue on the 

 frequency of occurrence of the noncrossover alpha derivative. 



We are now searching for independent cytological evidence 

 bearing on the validity of the A. A. scheme proposed here and have 

 chosen the bar locus in Drosophila for this purpose because it too 

 represents a tandem, serial duplication and moreover, presents the 

 opportunity for cytological analysis. To be sure, we cannot expect to 

 observe the double loop in meiotic tissues of either Drosophila or 

 maize, nor would such an observation be decisive. But we note 

 (Figure 4) that the A. A. mechanism calls for the removal (or addition) 

 of one complete member of the duplication, no more, no less. Thus, 

 in the case of the bar duplication, the hypothesis calls for the occur- 

 rence from bar of both noncrossover normal and noncrossover double- 

 bar individuals, corresponding to the loss and gain, respectively, of 

 one complete member of the duplication. Since the duplicated seg- 

 ment in bar is clearly defined as carrying the seven bands of the 16A 

 region of the salivary map, cytological analysis of noncrossover deriv- 

 atives should constitute a critical test of the A. A. hypothesis. 



Studies carried on in this laboratory over the past year (17) indi- 

 cate that noncrossover wild-type reversions from bar do occur, and 

 with a sufficiently high frequency to permit their statistical analysis. 

 Thus far, in addition to the expected crossover cases, reversions not 

 associated with recombination for marker genes have been obtained 

 from bar individuals heterozygous for the deficiency B-263-20, from 

 homozygous bar individuals heterozygous for the C1B chromosome, 



