346 THE BACTERIOPHAOE 



in question (GUdemeister and Herzberg 1923, Kuttner 1923, Kline 1927, Hadley 1928, 

 Fukuda 1928, Idieneberger 1932) was at first regarded as a powerful argument against the 

 virus theory, and in favour of the view that the phage was a product, enzymic or other, 

 of the bacteria themselves. It was indeed the failure of many workers to confirm these 

 findings (Ogata 1924, Reichert 1924, Meissner 1924a, Arkwright 1924, Sonnenschein 1927), 

 rather than any reaUzation that they were quite compatible with the virus theory, that 

 deprived them of much of their force. It is quite probable (see Burnet 1934) that some, 

 at least, of the positive findings were due to technical errors. But in the light of our present 

 knowledge it would be by no means surprising if many of them were correct. 



Our present conception of the real nature of phage-bacterium parasitism has 

 been determined largely by the proved existence of what are known as " lysogenic 

 strains " of bacteria. 



The classical examples of these lysogenic strains are the Bad. coli strain, which 

 was first described by Lisbonne and Carre re (1922) and later studied by Bordet 

 and Renaux (1928), and another Bact. coli strain which was studied by Gildemeister 

 and Herzberg (1924a). Both these strains, although they themselves show no 

 evidence of phage lysis, regularly yield filtrates which lyse Sh. sliigce, i.e. they are 

 permanent carriers of a phage to which they are themselves resistant, but to which 

 Sh. shigcB is sensitive. 



Such strains were, for a time, regarded as bacteriological anomalies, but such 

 a view is no longer tenable. Here, again, our present concepts have been largely 

 influenced by the studies of Burnet (1932, 1934). 



In a careful study of 34 strains of Salm. enteritidis he found that 27 were lysogenic, in 

 the sense that they yielded a phage to which they were themselves resistant, but which 

 produced transmissible lysis in other, specially susceptible bacterial strains. From these 

 27 lysogenic strains three different types of phage were obtained. A, B, and D ; fourteen 

 strains yielded phage B, seven phage D, two A and B, and one A and D. Phage A was also 

 frequently isolated from lysogenic strains of Salm. paratyphi A, Salm. paratyphi B and 

 Salm. typhi-muritivi, while phages B and D, and another phage N were sometimes present 

 in strains belonging to these species. Such results as these, as Burnet points out, are 

 incompatible with the view that the lytic agent in these lysogenic strains is some com- 

 ponent of the genetic apparatus of the bacterial cell. 



In the case of Salm.. paratyphi C, on the other hand, all strains examined, whether 

 they had been isolated in Russia, South America, or the East Indies, were found to be 

 carrying a single antigenic type of phage. A similar state of affairs may, as Burnet points 

 out, exist in the case of C. diphtherice, since Smith and Jordan (1931) found that every 

 strain examined was lysogenic when tested against a single, sensitive indicator strain of 

 the same species. The occasional occurrence of " nibbled " colonies that have been noted 

 by many workers in certain bacterial cultures represent, in Burnet's view, lysogenic strains 

 in which the resistance of the carrying bacterium is slightly unstable. 



We must, then, if we are to accept the virus hypothesis, to which all other evidence 

 clearly points, also accept the view that symbiosis between phage and bacterium is 

 an exceedingly common event ; so common that it would at the moment be unwise 

 to assert that any bacterial strain was certainly not carrying phage. 



What place this symbiotic process will eventually take in our conceptions of 

 bacterial structure, bacterial variation, antigenic behaviour, and other similar 

 problems we cannot yet prophesy. It may be a relatively minor one ; but it 

 may not. 



(For some of the methods of unmasking lysogenic strains, see Flu 19386, Lominski 

 1938, Fisk 1942a). 



