DELAYED EXPRESSION OF MUTATIONS 461 



developing on these plates represent the numbers of resistant clones present 

 at the time of spraying, that is, Ri and R 2 , respectively. The remaining plate 

 from each incubation period is washed with ten cc of normal saline and the 

 numbers of bacteria present (Ni and N 2 , respectively) determined by colony 

 counts. 



Where time 1 is the time of plating the bacteria, and no divisions can have 

 taken place, Ni is determined in a more direct manner. Instead of plating and 

 then washing off the bacteria plated, an equivalent quantity of the culture 

 from which the inoculum was taken is diluted and colony counts made. 



An estimate of mutation rate is thus obtained from four plates. In all ex- 

 periments these four plates were replicated several times, and a corresponding 

 number of independent estimates of rate obtained. These independent esti- 

 mates have been averaged, and the standard deviations calculated. 



In all experiments in which growth was determined by the use of duplicate 

 plates, care was taken to ensure the same amount of growth on both plates. 

 All plates were warmed in the incubator before plating the bacteria, and when 

 removed for plating were kept warm on a thermostatically controlled warm 

 table until returned to the incubator. All platings and removals from the 

 incubator followed an accurately timed schedule. The temperature in the incu- 

 bator was kept as uniform as possible by circulating the air rapidly with fans. 

 At the end of incubation, growth was stopped abruptly by chilling plates in 

 contact with the cold metal of a refrigerator freezing unit. This chilling did 

 not affect the survival of the bacteria or the phenotypic expression of the 

 mutants. 



Additional precautions were required with respect to (1) the choice of cul- 

 tures from which to inoculate the plates, (2) the number of bacteria plated, 

 and (3) the amount of phage applied by spraying. 



(1) When relatively large numbers (of the order of 10 8 ) of bacteria are 

 plated, there will be a certain number of resistant cells in the inoculum. These 

 resistant cells have occurred by mutation during the growth of the culture 

 from which the inoculum is taken. Cultures vary widely in the number of 

 mutants present at the end of growth, and in these experiments the number 

 present in the inoculum (Ri) was determined by spraying with phage immedi- 

 ately after the bacteria had been plated. Where the number is excessive it is 

 apt to obscure the increase in resistant clones resulting from growth, or to ren- 

 der the determination of the increase less accurate. For this reason, cultures 

 having an excessive number of resistant cells were not used asinocula. The cul- 

 tures that were used contained from 5 to 50 resistant bacteria per 10 8 sensitive. 



There is no evidence that this selection biased the results, since mutation 

 rates obtained using these inocula were the same regardless of the number of 

 resistant bacteria present. 



(2) The size of the inoculum was adjusted so that the end number of bac- 

 teria on the plate would be approximately 2X10 9 . With end numbers of less 

 than this the number of resistant colonies was reduced, and at the same time 

 the accuracy of the method. With excessively large end numbers of bacteria 

 there is a reduction in the apparent mutation rate. The precise interpretation 



63 



