THE PHYSICAL STATE OF THE PHAGE 239 



bacterium will in many cases be susceptible to only one of the phages contained 

 in a crude mixed filtrate, repeated transfer in growing cultures of this bacterium 

 will often eliminate all phages except the one to which it is sensitive. This method 

 is not, however, always reliable. A crude filtrate will often be found to contain 

 more than one phage acting on the test bacterium selected. In such a case each 

 of these phages may be propagated in successive transfers. It often happens that 

 one or more are reproduced more rapidly than the others, and so have a better 

 chance of transfer ; but such phage cultures may remain mixed throughout a 

 number of generations. An alternative, and better, method is to pick from isolated 

 plaques on agar plates, just as we pick isolated colonies in attempting to purify 

 a mixed bacterial culture. By a combination of these two methods it is usually 

 possible to obtain phage strains of undoubted purity ; but the procedure may be 

 laborious, and contamination is very liable to occur. For these reasons strict 

 attention to details of technique is required in any work of this kind (see Asheshov 

 et al. 1933a, Rakieten and Rakieten 1937). 



The ways in which various strains of phage can be differentiated from one 

 another will be considered in later sections of this chapter. At the moment we 

 are concerned only with the point that such strains exist, and are identifiable. 



The Physical State of the Phage. Is it Particulate or in Solution ? 



Given a bacteria-free filtrate containing active phage, we may determine 

 whether it behaves as a homogeneous solution or whether the activity is dis- 

 tributed in a discontinuous fashion throughout the solution. The answer will 

 depend entirely on the sensitivity of the method of examination. The question, 

 therefore, " Is phage particulate ?," must be understood in the restricted sense, 

 " Is the lytic activity in the solution continuous or discontinuous ? " Should it 

 prove to be continuous, the fact is not evidence that phage is not particulate in 

 the general sense of the term, but merely that the technique was too insensitive 

 to detect the separate units in action. 



As d'Herelle showed, isolated plaques form on agar cultures over which a high 

 dilution of phage filtrate has been spread. As further evidence that the phage 

 activity was discontinuous, d'Herelle noted that when a very high dilution of 

 phage filtrate was distributed in equal volumes among a large number of tubes, 

 each containing a young growing culture of a sensitive bacterium, lysis occurred 

 in some tubes but not in others, indicating that the active agent was not uniformly 

 dispersed in solution. This observation has been confirmed by a number of sub- 

 sequent observers (Bronfenbrenner and Korb 19256, McKinley and Holden 1926. 

 Bronfenbreuner 1927) ; and both Feemster and Wells (1933) and Luria (1940) 

 have shown that the proportion of tubes showing lysis among large numbers 

 inoculated with varying amounts of phage filtrate is in accord with statistical 

 expectations based on the assumption that the agent is dispersed in lytic particles. 



A possible escape from the obvious impUcation of these observations was provided 

 by the assumption that only a few of the bacteria in any culture were susceptible to phage 

 action. This assumption had little inherent probabihty and could hardly account for 

 the test-tube experiments referred to above, but it was accepted by some observers, mainly 

 on account of other observations that seemed to tell against a grossly particulate nature. 

 Thus Bordet (1925) stated that it was impossible to concentrate a phage filtrate by centri- 

 fugation at high speeds. We now know that this can readily be done if the speed is 

 adequate. Olsen and Yasaki (1923) stated that the phage was volatile. This would 

 have finally disposed of the view that it was particulate, or a living agent of any nature ; 



