BACTERIOPHAGE GENETICS 307 



for the probability of extracting the single segregant in the next generation). 

 And the value of ^c for closely linked markers becomes: 



where e^ for distant markers is as before, the burst size divided by twice N, 

 the pool size. This calculation, which gives ^c = 1-50 for T4, is in better 

 agreement with the results obtained by Stahl (1956); however, in this case 

 the markers used were very closely linked, and the results may be affected by 

 the negative interference which occurs over very short distances of the 

 genetic map. 



VII. Fine Structure Genetics 



A. The rll System 



In 1955 Benzer made a discovery which opened up a completely new class 

 of investigations in phage genetics. This discovery was of a purely technical 

 nature, but its ingenious exploitation made possible a kind of genetic analysis 

 which had not previously been possible either with phage or with any other 

 organism. Benzer found that a strain oiE. coli K12, lysogenic ^ for the phage 

 A, would not produce plaques when used as an indicator bacterium for a 

 certain class of r mutants of the phages T2 and T4, whereas other variants 

 of these phages were able to plate on K12 (A) with an efficiency which was 

 almost as high as on the usual indicator, E. coli B. Although this system 

 shows a superficial similarity to the host-range mutants discussed earlier, it 

 has many differences which make it very much more satisfactory for detailed 

 genetic analysis; r mutants arise spontaneously with a frequency which is of 

 the order of 1 in 10,000, and Benzer found that about two-thirds of these 

 mutants fell into the rll class, that is, they did not plate on K12 (A). The 

 isolation of a large number of rll mutants of independent origin is thus not 

 excessively difficult. But the more important point is that when two of 

 these mutants are crossed with each other, using B as the host cell of the 

 cross, wild-type recombinants are produced as in any cross except that the 

 recombination frequencies are aU small. By plating the progeny of the cross 

 on K12 (A), one can selectively comit these recombinants, since only they or 

 spontaneous reverse mutants will produce plaques on this indicator. In those 

 cases where the number of recombinants is sufficiently large so that the 

 comparison can be made accurately, the number of wild-type recombinants, 

 as determined by direct observation on the ceU B and by selection of plaques 

 on K12 (A), is the same. This means that one has a simple and direct method 



^ See Chapter 9 for a definition and discussion of lysogeny. From the point of view of 

 the present discussion, K12 (A) is simply another strain of E. coli. 



