Section 7 — Cytogenetics 



mal) X chromosome is the affected chromosome, 

 the large autosomes representing different types 

 of inversion heterozygosity. The X was divided 

 into nine regions, each of which was tested se- 

 parately in four parallel experiments: the stand- 

 ard autosomal homozygote as a control, the 

 In(2L + 2R)Cy heterozygote, the In(3L + 3R)P 

 heterozygote, and the combination of these two 

 inversions. It turned out that in all regions the 

 Curly + Payne combination is very much more 

 effective in increasing the crossover value than 

 these two inversions used separately. When the 

 effect curves of Curly-, Payne-, and combined 

 experiments are computed, all three prove to be 

 nicely parallel, showing conspicuous peaks at 

 both ends of the X and in the very middle of this 

 one-armed chromosome (more or less around 

 vermilion). This result suggests that all these 

 types of inversion heterozygosity affect the X in 

 basically the same way: they simply enable the 

 X to produce more crossover points than nor- 

 mally. As there are now -instead of one or two 

 crossovers -quite often three in the same tetrad, 

 it is only natural that, owing to interference, they 

 should be situated as far from each other as 

 possible at regular intervals, i.e. at both ends and 

 in the middle. This state of affairs is reflected in 

 the three-peaked curves obtained. 



Research supported by PHS Grant RG-6780. 



7.27. The Centromere Effect on the Distribution of 

 Exchanges in Drosophila Females. Peter E. 

 Thompson (Ames, U.S.A.). 



Tests of crossing-over with homozygous trans- 

 locations have shown that a distal region attached 

 near a centromere undergoes a reduction in ex- 

 changes. Conversely, a region normally lying 

 near a centromere shows an increase in crossing- 

 over when translocated to a distal position. This 

 phenomenon is widely known as "the centromere 

 effect on crossing-over". 



In studies with these same translocations made 

 heterozygous, the finding has generally been that 

 crossing-over is reduced adjacent to the break, 

 possibly as a result of synaptic ambiguities. The 

 present investigation has shown, however, that 

 marked regions very near the centromere pro- 

 duce more crossing-over if one of the homologues 

 has a translocation break between the centro- 

 mere and the markers. This has been strikingly 

 true for chromosome 4, which normally does not 

 undergo exchange in diploids, but which has 

 produced appreciable frequencies of crossing- 



over in the presence of a translocation having a 

 break near the centromere of 4. 



This increased crossing-over in translocation 

 heterozygotes applies only to those centric inter- 

 vals which show a great disparity in cytological 

 and genetic (crossover) length. Since the trans- 

 location has in effect removed only one of the 

 adjacent centromeres, it is suggested that the 

 normal distribution of exchanges is strongly in- 

 fluenced by some condition arising out of actual 

 centromere pairing. The credibility of a repulsion 

 of homologous centromeres prior to exchange is 

 being examined. 



These findings suggest a new interpretation of 

 the "interchromosomal effect", based on switch 

 pairing of homologous and non-homologous 

 elements when rearrangements are present 



7.28. Distal Crossing-over in Free-X and in Reversed 

 Acrocentric Attached- X Triploid Drosophila 

 melanogaster. J. D. Mohler and John 

 C. Neeley (Corvallis, U.S.A.). 



Free-X triploids having normal chromosome 

 structure give an estimated 6.0 per cent crossing- 

 over for all X-chromosomes in the y-w region. 

 The increase over the diploid frequency of 

 1 per cent accounts for nearly all of the distal 

 increase reported for the longer y-rb region 

 (Bridges and Anderson, 1925). In free-X triploids 

 having crossing-over suppressed in one chromo- 

 some by Ins (1) sc 8 , dl-49, the frequency of 

 crossing-over in the y-w region between the 

 normal chromosomes is 9.3 per cent, which is 

 equivalent to a mean recombination frequency of 

 6.1 per cent for all X-chromosomes. Apparently 

 two of the three chromosomes pair distally and 

 yield cross-overs in roughly 9 per cent of the 

 disjunction products. In uninverted triploids the 

 non-pairing, non-crossover chromosome may 

 be any of the three randomly; but in the inversion 

 heterozygote the inverted X is usually the ex- 

 cluded chromosome. 



The possibility of new investigation of this 

 distal increase is opened by studies involving a 

 reversed acrocentric attached-X triploid strain. 

 The distal normal member of the attached-X 

 probably pairs with the normal, free homologue, 

 leaving the inverted (Ins (1) sc s , dl-49), proximal 

 chromosome unpaired. In this case the recom- 

 bination frequency in the y-w region is 16.5 per 

 cent. This secondary increase may be causally 

 related to the initial increase over the diploid; 

 but since our attached-X is deficient for some 

 large part of the X-heterochromatin we are not 

 yet able to distinguish between the various 

 mechanical and physiological hypotheses that 



126 



