MUTATION AND PHENOTYPIC VARIATION IN PHAGES 313 



At this point several facts could be coordinated by making- 

 some radical assumptions. The study of lysogenic bacteria had 

 just suggested that prophages occupy a phage-specific site on the 

 bacterial chromosome, with the further implication, perhaps, of 

 genetic homology between phage and bacterium (Lederberg 

 and Lederberg, 1953; Lwoff, 1953; Bertani, 1953b). The 

 experiments of Weigle and of Fraser and Dulbecco suggested an 

 extension of this idea to explain the anomalous aspects of muta- 

 genesis in lambda and T3. The proposal was made that the 

 observed changes are not mutations at all, but effects of genetic 

 recombination between phage and bacterium (Delbriick, 1954). 

 This idea has been elaborated very skillfully by Stent (1958), 

 who shows that it has undeniable merits. These are also illus- 

 trated by the following work more or less directly prompted by 

 the idea. 



Fraser (in Hershey, Garen, Fraser, and Hudis, 1954) con- 

 tinued work on T3 mutagenesis with the following results. 

 Typically, phage T3 brings about prompt lysis of infected host 

 cells without the production of mutant phage particles. Under 

 certain physiological conditions phage T3 may form a relatively 

 stable complex with the host cell that persists for hours and pos- 

 sibly multiplies. These complexes may be induced to lyse by 

 dilution into broth at 37 ° C, when they liberate a yield of phage 

 containing about one per cent of host range mutants. Ultra- 

 violet irradiation of the bacteria before infection also produces 

 many mutants among the phage progeny. The mutants pro- 

 duced in this way are of the same phenotype as those liberated 

 from late-lysing unirradiated bacteria. Fraser suggested that 

 these "mutants" might arise as the result of recombination be- 

 tween the phage genetic material and a homologous region in 

 the host cell genetic material. If this were so it might be antici- 

 pated that the "mutational patterns" would differ when the 

 same phage strain was used to infect different bacterial strains. 

 This was tested and found to be the case. 



Jacob (1954c) studied mutations affecting virulence in phage 

 lambda with results somewhat different from those reported by 



