140 DISINFECTION . 



merely show that a logarithmic curve describes the death of bacteria under the 

 action of a disinfectant, just as it describes a chemical reaction whose rate is 

 governed by the concentration at any moment of one of the reacting substances. 

 Two alternative explanations have been offered to account for the form of the 

 curves which Chick has described. The first of these suggests that the varying 

 resistance of the bacteria in any given suspension can be described in the form 

 of a frequency curve, as is almost certainly the case, and that the survival curves 

 described by Chick are simply an expression of this difference in resistance. 

 The obvious objection to this hypothesis, as Chick has pointed out, is that the 

 form of the frequency curve describing the distribution of resistance must be 

 supposed to be of the extreme skew form, if it is to account for the experimental 

 results. Withell (1942a), however, has shown that the distribution of bacterial 

 resistance is normal if the survival times are plotted on a logarithmic instead of 

 an arithmetic scale. Such a logarithmic distribution of a characteristic has been 

 noted in pharmacological and zoological work (see Gaddum 1933, Hemmingsen, 

 1934), and its occurrence in bacteria need not therefore occasion surprise. 



An alternative explanation (Chick 1910) is that the death or survival of 

 any given bacterium during any interval of time is determined by a multi- 

 tude of small and independent causes — by "chance" in the statistical sense — 

 the presence of the disinfectant weighting the chance of survival against each 

 bacterium to a constant degree, for any given concentration of the disinfectant, 

 and with other controllable conditions held constant. If the chance of each 

 bacterium dying during any unit of time is x, and remains x over the whole period of 

 the experiment, then the death rate will be the same during each unit of time ; 

 the survivors at the end of any one time interval will suffer the same propor- 

 tionate decrease in their numbers during the time interval which follows, and a 

 logarithmic curve of decrease will result. This explanation does not, of course, 

 mean that variations in resistance of individual bacteria play no part in dis- 

 infection. With vegetative bacteria, the rate of death is often represented by a 

 sigmoid curve rather than by a straight line, suggesting that differences in resistance 

 dependent on the age of the individual organisms are responsible for the deviation. 

 The real question is whether, in the disinfection of spores that do not differ materially 

 in age, the exponential type of curve is due to chance in the statistical sense, or to 

 a frequency distribution of resistance of the logarithmic type. As Irwin (1942) 

 points out, it would require very accurate data to distinguish between the two. 

 (For a further discussion of this subject see Eijkraan 1908, Hewlett 1909, Reichel 

 1909, Reichenbach 1911, Loeb and Northrop 1917, Brooks 1919, Cohen 1922, 

 Knaysi 1930c/, Knaysi and Gordon 1930, Bancroft and Richter 1931, Jordan and 

 Jacobs, 1944.) 



Concentration of Disinfectant.— Chick (1908) found that the relationship between 

 the concentration of a disinfectant and the time taken for disinfection is not a 

 simple but an exponential one, the exponent of the concentration being a factor 

 varying with each disinfectant. That is to say, doubling the concentration of 

 phenol does not halve the time necessary for the completion of the reaction as 

 might be expected, but diminishes it to a far greater extent, Watson (1908), work- 

 ing on Chick's figures, found that the relation could be expressed by the formula 



C" ^ = a constant 



where C is the concentration, n a constant varying with each disinfectant, and t 



