EMORY L. ELLIS AND MAX DELBRUCK 379 



to conform to formula (1) (see Table II). In this experiment we 

 could also have inferred the average number of particles per sample, 

 using formula (2), from the fraction of the plates showing no plaques 

 (giving 0.93 per sample) instead of from the total number of plaques 

 (27/33 = 0.82 per sample). 



Experimental Measure of Efficiency of Plating 



If the samples are incubated until the bursts have occurred, and 

 then plated, the samples which had no particles will still show no 

 plaques, those with one or more particles will show a large number, 

 depending on the size of the burst, and on the efSciency of plating. In 

 any case, if we wait until all bursts have occurred, only those samples 

 which really contained no particle will show no plaques, quite inde- 

 pendent of any inefhciency of plating. From this fraction of plates 

 showing no plaques we can therefore evaluate the true number of 

 particles originally present in the solution, and by comparison with 

 the regular assay evaluate the efficiency of plating. In this way we 

 have determined our efhciency of plating to be about 0.4. For 

 instance, one such experiment gave no plaques on 23 out of 40 plates, 

 and many plaques on each of the remaining plates. This gives 



23 

 Pa = —ov 0.57 from which n = 0.56 particles per sample. A parallel 



assay of the stock phage used indicated 0.22 particles per sample; 



0.22 

 the plating efficiency was therefore tttz = 0.39. This plating effi- 

 ciency remains fairly constant under our standard conditions for 

 assay. The increase in probability of plaque formation which we 

 suppose to take place following the infection of a bacterium by the 

 phage particle, i.e. the initial rise, brings the plating efficiency up to 

 0.65. 



The Burst Size 



Single particle experiments such as that described above, revealed a 

 great fluctuation in the magnitude of individual bursts, far larger than 

 one would expect from the differences in size of the individual bacteria 

 in a culture; indeed, they vary from a few particles to two hundred 

 or more. Data from one such experiment are given in Table III. 



51 



