GENETIC RF.COiMBINATION IN BACTERIAL VIRUSES 7 



rion appear. One way of explaining this result would be to suppose 

 that small frequent duplications of the linkage structures occur: small 

 to account for the lack of correlation between distant factors, fre- 

 quent to account for the 2 per cent heterozygous at any given locus. 

 Practically every phage should carry some of these duplications. 



THE LINKAGE SYSTEM 



A linkage map in the usual sense of the term cannot be drawn 

 directly on the basis of recombination data in phage crosses. The 

 reason for this is that the phage cross is very different from a cross 

 between two organisms in the classical sense of Mendelian genet- 

 ics. An attempt can be made, nevertheless, to see if linkage exists, and 

 to demonstrate linear arrangement of the loci. The first attempt of 

 this kind was made by Hershey and Rotman with different mutants 

 of T2. In two-factor crosses the frequency of recombinants in the 

 total yield varied from 40 per cent downward. Three linkage groups 

 were found, each of which showed 40 per cent recombination with 

 the other two. As no higher value of recombination was obtained, 

 these groups were considered independent or unlinked. Lower values 

 of recombination between two loci indicate that they belong to the 

 same linkage group. Inside the linkage groups all kinds of recombina- 

 tion values can be found. If these characters are chosen rather closely 

 linked to each other by different combinations of three-factor crosses, 

 which one is in the middle can be determined. If the order is A, B, 

 C, the frequency of the ABC type will be much smaller in the cross 

 AC X B than in the crosses AB x C or A x BC. More complete and 

 elaborate data have been collected recently by Doermann and Hill 

 (1952) working with phage T4. From these data full evidence can 

 be obtained for the linearity of the markers on the different linkage 

 groups. 



THE MECHANISM OF GENETIC RECOMBINATION 

 IN PHAGE 



The interpretation of the data on phage recombination requires 

 some generalization of the idea of a cross. That the mixed infection of 

 a bacterium is not a straightforward analogue of a simple genetic 

 cross can be easily demonstrated by the following facts: 



