488 BACTERIOPHAGES 



and the contents immediately poured over the surface of an agar 

 plate. The plate is rocked gently for a few seconds to mix the 

 sample with the agar and to insure uniform layering of the agar. 

 Several 0.5 ml. samples from the sample flask containing the 

 remainder of the diluted suspension are also plated to determine 

 average burst size in the suspension. 



c. Results. Of the 50 plates, 35 had no plaques, and 1 5 plates 

 had a total of 1750 plaques, individual plate counts being 



Average plaque count on the plates made from the sample flask 

 after the rise period was 39. 



d. Calculations. Thirty-five of 50 plates had no placjues, 

 hence no infected bacteria. P(0) = 35/50 = 0.7 = e~" = 

 ^-0.35^ so the average number of infected bacteria/sample was 

 0.35 and the average number/ml. of the sample flask was 0.7. 

 The dilution factor in going from the adsorption tube to the 

 sample flask was 7 X 10", so the number of infected bacteria in 

 the adsorption tube was 4.9 X lO'', thus agreeing closely with 

 the bacterial assay before infection. This could also be checked 

 by plating 0.1 ml. of dilution tube #4 by the agar layer technique 

 before the end of the latent period to determine the actual num- 

 ber of infected bacteria present. 



If 50 samples contain an average of 0.35 infected bacteria/ 

 sample, the total number of infected bacteria in the 50 samples 

 should have been 50 X 0.35 = 17.5 bacteria. However, only 15 

 plates had plaques. This must mean that 2 or 3 of the plates had 

 phage yields coming from 2 infected bacteria. The expected 

 distribution calculated from the Poisson formula is 



P(0) = ^-f-^a = 0.7, or 35 plates out of 50 with no infected fjactcria 



0.351 X e~'^-^^ 

 P{\) = = 0.245, or 12 plates out of 50 with 7 infected 



bacterium 



