202 CELL HEREDITY 



second locus showed none, giving only 4:4 segregations. What the re- 

 lationship may be between 5:3 ratios and the other types of aberrant 

 tetrads discussed above is not at all clear at present. It does seem likely 

 that the frequency of aberrant allelic ratios at some loci may be much 

 higher than was ever suspected in the past, when many such examples 

 were discarded as technical errors. 



Re- Evaluation 



The data of classical cytogenetics are fairly consistent with the chias- 

 matype theory of crossing over, which holds that genetic recombination 

 is the consequence of reciprocal exchanges between nonsister chromatids 

 occurring after chromosome duplication and pairing, when the four 

 chromatids lie close together, constituting the bivalent of meiotic 

 prophase. Clearly, the new genetic data from studies of recombination 

 in very short intervals cannot be interpreted on this theory. 



The great difference in findings between the two systems results from 

 the size of the region under observation. In studying recombination 

 between markers which are several map units apart, one overlooks the 

 events which are detected by selecting recombinants in regions smaller 

 than 0.1 map unit. Intragenic recombination at a particular locus is 

 generally too rare an event to be found in an unselected sample of progeny. 

 A comparison is given in Table 7.6 of the observations made in conven- 

 tional and in fine-structure mapping. 



In evaluating these findings, two alternative views have been pro- 

 posed: (1) a unified hypothesis, that all recombination is of the same 

 type, based upon a copy-choice mechanism which operates during DNA 

 replication and chromosome duplication accompanied by sister-strand 

 exchange; and (2) a two-mechanism hypothesis, that the aberrant events 

 of recombination within loci occur by copy-choice at the time of DNA 

 replication and, in addition, reciprocal exchanges occur between any two 

 of the four non-sister chromatids. 



Preliminary attempts to distinguish experimentally between these 

 alternatives have been based on the view that, if the mechanisms are 

 different, it should be possible to alter the frequency of one process 

 independently of the other. Treatments with UV and with heat have in- 

 dicated that the approach is a promising one, but no conclusive results 

 have been obtained as yet. 



In the absence of conclusive experimental data, we may consider these 

 two hypotheses in the light of existing information. If all recombina- 

 tion is of the same type, there is a timing problem. Replication of DNA 

 seems to be completed before zygotene pairing of homologous chromo- 



