BACTERIOPHAGE MUTATION 31 



There is no necessity that phage C should be added subsequent to the action 

 of C. Figures 2-4 show plates spread with a 1 : 100 dilution of a broth culture of 

 SF and, subsequently, spread over a central area with (Fig. 2) phage C (Fig. 3) 

 phage C and (Fig. 4) a mixture of the two. It will be seen that there are about 

 100 times as many colonies arising on (4) as on (3). Even if phage C is added 15 

 minutes after C there is a large increase in the number of resistant colonies over 

 those produced by C alone. 



In considering the time relationship in these experiments, it must be remem- 

 bered that neither phage is very actively adsorbed to the susceptible organism, 

 either when killed by heat (see Table 1) or in the living state. Even in the pre- 

 sence, as in all these experiments, of excess of bacteria it may be an hour or longer 

 before any given phage particle makes effective contact with a susceptible organ- 

 ism. During this period the bacterial population is also of course increasing in 

 density, so that the conditions are too complex for any strictly quantitative inter- 

 pretation of the results. 



Certain points are, however, clearly established ; the most important is that 

 the appearance of SF/C resistant colonies is due to the direct positive effect of 

 phage C on the coccus SF and not to a selection of pre-existent variants. This 

 latter alternative is in most examples of resistant production by phage very difficult 

 to exclude, and is often apparently the most important factor (Burnet 1929). A 

 glance at Figs. 3 and 4 show at once that, while the colonies on 4 (mixed phages) 

 are obviously resistant to C, 99 p.c. of the organisms from which they arose were 

 not initially resistant to C, since they fail to produce colonies on the plate spread 

 with this phage only. 



The change from SF to SF/C under the influence of phage C must in some 

 cases at least occur quite rapidly, probably within an hour, if the descendants are 

 to be protected against the action of undiluted C phage. Further, the experi- 

 ments indicate that a remarkably high proportion of effective contacts between 

 phage and bacteria must result in the induction of resistance rather than lysis. 

 The difficulties associated with the slow adsorption of phage to bacteria made it 

 impossible to give any numerical estimate of this proportion, but it seems very 

 unlikely that it is less than 5 p.c. On the evidence that when SF is plated on agar 

 covered with strong phage C just as many resistant colonies appear as susceptible 

 ones on ordinary agar, we were at first inclined to consider the possibility that 100 

 p.c. of contacts resulted in the appearance of resistant forms, phage multiplication 

 taking place by some other process than the usual one. Probably a majority of the 

 authors who have studied the question consider that in the initial stages of the 

 action of phage on a susceptible culture, multiplication of the lytic agent proceeds 

 in the absence of any bacterial lysis. Bronfenbrenner (1928) in fact claims that 

 there is a distinct growth-stimulating effect in the early stages. One of us (Burnet 

 1934) has argued that there is no experimental basis for this point of view, and 



214 



