104 DISINFECTANTS 



strain" of the typhoid bacterium is the only strain used 

 for standard experiments, but even with this strain death 

 may occasionally occur more rapidly or more slowly than 

 usual. In order to exclude such fluctuations, the F.D.A. 

 method considers phenol coefficients as being correctly 

 established only when the control cultures are killed by 

 phenol within the following ranges of time and concen- 

 tration : 



Dilution Concentration 5 min. 10 min. 15 min. 



This means that the culture may be so resistant that it 

 can tolerate 1.00% phenol for 15 minutes, but must 

 be killed by l.ll%o in 10 minutes, or it may be so sensi- 

 tive that it is killed by 1.11% in 5 minutes, but it must 

 survive for at least 10 minutes at 1.007^. 



It may seem strange that the death rates fluctuate so 

 greatly while phenol coefficients are uniform enough to 

 be universally accepted. However, the fact is that we 

 are really fooling ourselves about the accuracy of the 

 phenol coefficient. Both methods are based on the same 

 principle, and their accuracy should therefore be ap- 

 proximately the same. The apparently greater accur- 

 acy of the phenol coefficients is due to the enforcen-ent 

 of very specified conditions, and to the discarding of all 

 those results where the controls did not remain within 

 the narrow limits permitted by the rules of the method. 

 If such severe restrictions had been applied to the de- 

 termination of the concentration exponent, its accuracy 

 would have been equal to that of the phenol coefficient. 



But even the range allowed by the F.D.A. method 

 permits a great fluctuation of results; an experiment is 



