WHAT IS A GENE? 165 



few wild-type phages attributable to baek mutation. Quite a number of 

 rll mutational changes have been located on the phage chromosome in 

 this way (Figure 6.5). 



Some mutants such as rl64 seem to involve changes that extend for 

 some distance, overlapping a number of others. These will not produce 

 recombinants with any of the mutants whose positions they overlap; for 

 example, mutant 164 will give wild-type recombinants only with mutant 

 173 among all those shown in Figure 6.5c. Because they do not revert, 

 thev are judged to be deletions. They proved of practical value in allow- 

 ing a short-cut in mapping almost 1000 independent mutants. This was 

 necessary because pairwise matings of 1000 mutants would require 

 1000 X 999/2 crosses, a task of heroic dimensions. The short cut con- 

 sisted of crossing the mutant to be located with a series of broad dele- 

 tions (Figure 6.6). If the cross did not yield wild-type recombinants, the 

 defect could be assigned to one of the segments. For example, if a 

 mutant gives wild-type recombinants with all but rH 88, rA105, and 

 rl64, its defect must be located in segment 4. The one which gives 

 recombinants with all must have its defect in segment 1. Segments 7, 

 9, and 11 are not distinguishable by the set of mutants shown in Figure 

 6.6. In this way, then, many unnecessary crosses can be avoided. 



If strain K12 is mixedly infected with wild-type phages and an rll 



mutant, , the wild type supplies the function lacking in the mutant 



and the bacteria lyse, liberating both types of phages. The trans- 

 configuration is simulated by infecting strain K12 with a pair of rll 

 mutants. If both mutants belong to the same segment, A or B, e.g.. 



Segment number 



FIGURE 6.6. Arrangement of a number of extensive defects (deletions) at the HI 

 locus of pfiage 74 (after Benzer and Freese, 1958, Proc. Nafl. Acad. Sci., 44:112). 



