USE OF PHAGES IN EPIDEMIOLOGICAL STUDIES 411 



The structural formulae provide more information about the 

 lysogenically determined types than do the symbols in routine 

 use. The relations between many types become obvious and an 

 explanation is found for the ability of some adapted typing 

 preparations to lyse heterologous Vi-types. 



It so happens that the nonlysogenic precursors of all the 

 naturally occurring lysogenically determined types impress only 

 a host-induced modification on Vi-phage II. In contrast, all 

 lysogenic types hitherto examined are lysed only by host-range 

 mutants which may or may not have undergone additional host- 

 induced modification. Moreover, the identity of the mutants 

 able to carry out this lysis is decided, as might be expected, by 

 the determining phages. Closely related determining phages 

 such as d6 and f2 erect barriers to the multiplication of Vi- 

 phage II that are overcome by related groups of mutants. 

 Thus, in this system, the interesting phenomenon is found of the 

 selection of easily definable host-range mutants of one phage by 

 lysogenization of cells with other phages, some of which are re- 

 lated to each other, but none of which is demonstrably related 

 to the Vi phage. The mutant selection is so specific that it can 

 be used for the identification of the determining phages that 

 govern it. 



Type A seems to be the most primitive form of Salmonella 

 typhi in this scheme, because it is sensitive to all the adaptations 

 of Vi-phage II, because it is the ancestral type of many lyso- 

 genic Vi-types, and because nonlysogenic types often tend to 

 lose their specificity and to change into type A. A number of 

 Vi-types are known that can be lysed only by host-range mutants 

 of Vi-phage II, yet are not demonstrably lysogenic, and it is not 

 known what controls the specificity of such types. Nor is it 

 known why some nonlysogenic types produce host-induced 

 modification in Vi-phage II, while others select host-range 

 mutants. 



All phages studied hitherto show host-range mutations, and 

 phenotypic flexibility of host range is sufficiently frequent to 

 suggest that this also is a widespread phenomenon. The striking 

 feature of Vi-phage II is the latitude over which these changes, 



