APPLICATIONS OF STATISTICAL ANALYSIS 



15 



Dilute starting culture 

 of bacteria 



Dilute starting culture is 

 distributed 1 ml per tube 

 into many tubes 



III 



Growth period of 

 several hours 



I 



Concentrated final cultures 

 of bacteria, each tube 

 grown separately from the 

 dilute starting culture 



Fig. 6. This experiment differs from that of Fig. 5 in that the bacteria are 

 distributed into separate tubes before the growth occurs. The final tubes are 

 then sampled to determine the number of mutant bacteria in each tube. 



biotic. One asks whether the testing with the drug induced the resistance 

 in a small fraction of the cells or whether the property had occurred 

 previously, by chance (spontaneously) , in that small fraction of the cells. 

 Thus the question was posed as to whether mutations are spontaneous or 

 induced. The answer to the question involved an anlysis of variance. 



Consider, as in Fig. 5, a flask containing 100 ml of medium, inoculated 

 with some cells which then grow up to some final concentration. If 

 the 100 ml are distributed in 100 tubes and analyzed for the number of 

 mutants per tube, we will find an average value for the mutant concen- 

 tration and a variance calculated from the deviations from the average. 



Next, consider a parallel experiment in which the 100 ml are distributed 

 into the 100 tubes as soon as the flask is inoculated. The bacteria then 

 grow up to the same final concentration as before (see Fig. 6). These 

 100 tubes are now analyzed for the number of mutants per tube and, 

 as before, an average and a variance are obtained. 



If the mutation mechanism is by induction of drug resistance, then 

 the results of the two experiments should be entirely similar, since the 

 drug is acting on the same number of cells in the same number of tubes. 



