The ability of various phages to propagate on Escherichia coli strain W was investigated. Phages P1 , T2, T3, T4, T5, T6 and T7 could not be shown to form plaques on this strain. Phages T2, T3, T5 and T6 prevented the development of a bacterial lawn when added to a plate at an input ratio of about three phage per bacterium; it appears these phages exerted a killing effect on strain W. Phage P1 and phage T4 did not exhibit this killing effect. Phage T1 formed atypical plaques on strain W with an efficiency of plating of 10
Ò
; it appears these plaques are due to mutants occuring in the T1 population able to propagate on strain W. All of the above phages adsorbed efficiently to strain W with the exception of T4, explaining its inability to either propagate on or kill strain W. The infection of strain W by P1 was similar in most respects to that of λ but in order to establish the occurence of conventional restriction, DHA degradation would need to be demonstrated. Phages were isolated which propagate on strain W; they are similar in morphology to phages T5 and λ and do not readily adsorb to E coli strains B, C or K. The supernatant from broth cultures of strain W was shown to contain two closely related phages, one plating on E coli C, and the other on E coli K. Each possesses a characteristic pattern of plating efficiencies on strains C and K when propagated alternately in these two hosts but the two phages were shown to be co-immune and identical with respect to heat sensitivity, morphology and serology. Both tended to lose the ability to exclude phage P1 on lysogenising strain C once having mutated to plate on strain K. This may be due to the integration of the mutated phage at alternate "non-restricting" sites on the E coli C chromosome. A series of conjugal crosses was employed to determine the sites of integration of the phages on the chromosomes of E coli strains W and C. The phage present in the W supernatant which plated on K was found to integrate close to the proline loci on the chromosome of E coli W but the phage plating on C appeared to have more than one locus, one of which may map close to the 85 minute mark on the linkage map of E coli W (36, Figure 22). No information has so far been obtained concerning the sites of integration of the w phages in restrictive and non-restrictive lysogens of E coli C. The failure to obtain a 'cured' strain of E coli W by elimination of the prophage integrated at the two mapped sites leaves open the possibility of the existence of more than one integration site of phage W.C on the E coli W chromosome.