72 MOLECULES, VmUSES, AND BACTERIA 



the double cross-over classes show clearly the origins of the recom- 

 binant structures. When the outside markers are as in parent 1, re- 

 combination has taken place by a double cross-over in which a parent 1 

 chromatid has included a short region of parent 2 sequence. When, on 

 the other hand, the outside markers are like those of parent 2, it is a 

 parent 2 chromatid which has included a short segment of parent 1 

 sequence. In other words, when the outside markers are like those of 

 parent 1, it is a parent 1 chromatid which is the "recipient," and when 

 the outside markers are like those of parent 2, it is a parent 2 chroma- 

 tid which is the "recipient." 



In cross A, the frequencies of these two events are not the same 

 (see Figure 10). The double cross-overs in which parent 2 is the "re- 

 cipient" are twice as successful in producing a wild-type adenine locus 

 as are the double cross-overs in which parent 1 is the "recipient." Now 

 the closest outside marker is the paba marker, and it is a thousand 

 times farther away from the adenine locus than are adis and adi? from 

 each other. Therefore, we would not expect the positions of the outside 

 markers to have much influence on the quantitative results. If there 

 were such an influence, we would expect class 2 recombinants in cross 

 A to be the less frequent, for the switch to the right of adis might be ex- 

 pected to fall to the right of paba some of the time, rather than between 

 adi5 and paba. Class 2 is, however, the more frequent of the double 

 cross-over classes. 



If we now examine cross B, it is evident that the outside markers 

 indeed do not influence the frequencies of classes 1 and 2, for here the 

 adi5 and adi7 sites have been interchanged. We see that there is a two- 

 fold dijfference in the relative frequencies of the two classes, but it is 

 now the class in which the outside markers are from parent 2 in which 

 the frequency is low. hi other words, when the adn site is in the "re- 

 cipient" chromatid, recombination frequency is low. In terms of the 

 transformation model, this would mean that the adu mutant site is a 

 more extended site than the adis site. In other words, adn comprises 

 more minimum recombination units than does adis- 



Many examples of the sort just described can be found, in data ob- 

 tained with other loci in Aspergillus, with Neurospora, and with Sac- 

 charomyces. Thus the crosses discussed are far from being an isolated 

 case. The analogies between such data and transformation data are 

 too striking not to be significant. To the extent that they indicate iden- 

 tity of recombination mechanism in transformation as well as in sys- 

 tems of equal parental participation, one can conclude that recombina- 

 tion events observed in the fine-structure analysis of chromosomes in- 

 volve crossing-over between segments of chromosomes which are of the 

 same order of magnitude as the markers themselves, and that the di- 

 mensions of the very markers employed to analyze chromosomes at the 

 molecular level can no longer be neglected, as they have been in the 



