406 BACTERIOPHAGES 



of its complementary mutant by the type on which phage prop- 

 agation was carried out. It was shown by Anderson and FeHx 

 (1952, 1953a, c) that this suggestion was pardy true. However, 

 the latter workers demonstrated that another phenomenon, 

 host-induced modification, played an important role in the 

 elaboration of the typing phages. Host-induced modification 

 in phage was first described by Luria and Human (1952) in 

 phage T2. The subject was examined in some detail by Bertani 

 and Weigle (1953) and Weigle and Bertani (1953) in phages 

 P2 and lambda. A general description of the phenomenon was 

 given by Luria (1953b). As it appears to affect only the phage 

 phenotype, the present author has usually referred to host- 

 induced modification as "phenotypic" modification. A given 

 phage Xa, in which the subscript letter indicates the host range, 

 which will yield a plaque for every particle plated on a particular 

 indicator strain A, may have a low eflficiency of plating (EOP), 

 of, say, 10~^, on another strain B. The selection and propaga- 

 tion of plaques appearing on strain B may yield a phage X^b 

 having an EOP of 1 on B. We will assume for the present that, 

 whatever changes it may undergo, the phage always retains 

 an EOP of 1 on strain A. Clearly, the phage newly grown on B 

 could be descended from a host-range mutant able to multiply 

 in B, or from a particle which has been able to adapt itself to 

 B without genetic change. It is possible to distinguish between 

 the two types of change by propagating the phage X^b for a 

 single cycle in strain A, when every particle of a phage changed 

 in phenotype only will revert to phage X^, whereas a host-range 

 mutant will multiply unchanged as phage X^b- Moreover, if the 

 change to X^b is phenotypic only, the particles destined to under- 

 go it, having no genetic continuity (in terms of the ability to mul- 

 tiply in organism B) in the original phage stock, will exhibit a 

 Poisson distribution in a series of small samples grown for a single 

 cycle in strain A, while host-range mutants will show a clonal dis- 

 tribution (Luria and Delbriick, 1943; Luria, 1951). Bertani and 

 Weigle (1953) observed that phenotypically modifiable particles 

 of phage P2 conformed to a Poisson distribution, and Anderson 



