472 BACTERIOPHAGES 



In the spreading technique the agar plates should be dried by 

 pre-incubation overnight so that the sample will be rapidly 

 adsorbed and not remain as fluid pools on the agar surface. 



If the adsorption period is long and the rate of adsorption 

 slow the initial bacterial concentration cannot be used in the 

 calculations since the bacterial count is increasing exponentially, 

 doubling in 1 generation time, which is about 20 min for E. 

 coli in broth at 37 °C., so that the count of surviving bacteria 

 will be too high with respect to the initial bacterial assay. The 

 simplest way to avoid this problem with slowly adsorbing phages 

 is to increase phage input so that a reasonable proportion of the 

 bacterial population, about 50 per cent, will be infected in 

 5-10 min. If less than 20 per cent of the bacteria are infected 

 during Vo generation time, the method is not usable. The 

 method cannot be used with a multiplicity of infection higher 

 than 2 because the Poisson distribution is no longer applicable 

 without correction. The reason for this is that the derivation 

 of the Poisson distribution assumes complete uniformity in the 

 bacterial population with respect to the probability of adsorption. 

 Actually the probability of adsorption of a phage particle to a 

 bacterium is a function of the bacterial surface area, and this 

 is certainly not uniform in bacterial populations. This in- 

 troduces little error at low multiplicities of infection, but at 

 multiplicities higher than 2 the error begins to exceed the 

 experimental error of plating and leads to low values for n. 

 A method has been worked out by Dulbecco (1949a) for cor- 

 recting this discrepancy by measuring the size distribution of 

 bacteria in the population and appropriately altering the for- 

 mula for the distribution. The method is not included here 

 since it is unlikely that one would have to use multiplicities 

 above 2 to determine adsorption rates. 



Use of the proportion of surviving bacteria to calculate ad- 

 sorption rates of phage to bacteria is possible in certain instances 

 in which the first 2 methods given are inapplicable. Phage 

 treated with ultraviolet irradiation is "killed," i.e., unable to 

 reproduce itself, and hence no longer produces plaques when 



