500 BACTERIOPHAGES 



ror, which in this case is about 10 per cent. The plaque count 

 on the mixed indicator plate confirms this, since at the most only 

 3 of a total of 95 infected bacteria could have liberated both 

 kinds of phage, since only 3 clear plaques were found on this plate. 

 These 3 clear plaques might also have been due to accidental 

 overlaps or to a bacterium caught in the act of dividing, 1 mem- 

 ber of the pair liberating Tl and the other T7. 



In general, when bacteria are mixedly infected with 2 immu- 

 nologically unrelated phages, they liberate either 1 or the other 

 phage type but not both. This phenomenon has been termed by 

 Delbriick the mutual exclusion effect. 



The burst size of the mixedly infected bacteria is calculated 

 from plaque counts on the separate indicator plates made from 

 S.G.T. after all bursts have occurred. Burst size is of course cal- 

 culated on the basis of the number of bacteria which have actu- 

 ally been found to liberate each kind of phage rather than from 

 the total number of infected bacteria; i.e., the total yield of T7 

 phage divided by the number of infected bacteria which liberate 

 T7 phage gives average burst size of those bacteria which liber- 

 ate T7. The burst size of mixedly infected bacteria is markedly 

 less than that of bacteria infected with only 1 phage type. This 

 has been termed the depressor effect (Delbriick, 1945c). 



The average multiplicity of infection for Tl and T7 can be cal- 

 culated by any of the methods described for the one-step growth 

 experiment (pp. 473 ff.). If the average multiplicity of infec- 

 tion is known for each phage, the proportion of mixedly infected 

 bacteria can be calculated by application of the Poisson distribu- 

 tion. This assumes, of course, that the adsorption of 1 phage par- 

 ticle by a bacterium does not affect the rate of adsorption of an- 

 other phage particle by that bacterium. This can be readily 

 tested by permitting the bacteria to adsorb Tl phage alone for 5 

 min. and then adding T7 phage and determining if the rate of 

 adsorption of T7 phage is normal. This is most simply done by 

 centrifuging samples at intervals and plating on B/1,5 so that 

 only the unadsorbed T7 phage in the supernatant is counted. 



