298 BACTERIOPHAGES 



genes from one strain to another for comparison and mapping, 

 just as in higher organisms. This approach to descriptive phage 

 genetics has been exploited particularly by Adams (1951b) and 

 by Streisinger (1956b) with benefits of both expected and un- 

 expected kinds. The information gained in this way will be 

 utilized below when necessary to supplement information 

 obtained by direct isolation of mutants. 



a. Mutations Affecting Specificity of Adsorption 



A given phage is likely to adsorb only to certain members of a 

 group of closely related bacterial strains. The phage may 

 acquire by mutation the ability to adsorb to a bacterium non- 

 receptive toward the original type, or lose by mutation an 

 adsorptive capacity it formerly possessed. Such mutations are 

 called, by common consent, host range mutations, and the locus 

 of such mutations is called an h locus (Hershey, 1946b). The 

 term host range in this connection is unsatisfactory because host 

 range can vary in ways not related to adsorption, but the usage is 

 too firmly entrenched to change. The term h mutation, how- 

 ever, should be restricted to its present exact meaning. 



The typical procedure for the isolation of h mutants begins 

 with the isolation of a series of bacterial mutants resistant to the 

 wild type phage. The bacterial mutants are then used to 

 select phage mutants capable of infecting them. In general 

 some but not others of the bacterial mutants will prove sensitive 

 to h mutants present in the original stock of phage. The typical 

 result of this procedure is the isolation of a mutant, h, differing 

 by a single mutation from the parental stock, called h "*", as shown 

 by recombination tests. The effect of the mutation is usually 

 an extended host range; the mutant can infect two bacterial 

 strains of which the wild type phage can infect only one. 



The extent to which such variation can be carried by successive 

 mutational steps has never been investigated, though the ques- 

 tion is of obvious importance to both phage and bacterial taxon- 

 omy (Stocker, 1955). As one remarkable instance, mutants of 

 Pasteurella pestis phages are known that can infect various Shigella 



