212 BARNETT COHEN 



determination of the velocity constant up to the point where the 

 numbers of organisms have been reduced by 50 per cent. To 

 avoid error from the former consideration as well, k should be 

 determined for the middle part of the disinfection curve, say, 

 between the points for 25 and 75 per cent reduction. 



The temperature coefficient (Qio) in bacterial viability 



We have already mentioned the broad empirical generaliza- 

 tion that the temperature coefficient is about 1 for the accelera- 

 tion of physical processes by a rise of 10° in temperature, and 2 

 or more for chemical processes. The rule is not a hard and fast 

 one, and its basis is obscure. It is therefore important to re- 

 cognize that the coefficient is at best a suggestive observation 

 until we possess a better knowledge of the mechanism of the 

 temperature effect. 



Obviously, in strictly controlled experiments where only one 

 reaction is allowed to take place, a temperature coefficient so 

 derived is of some value. On the other hand, in biological ex- 

 periments, the observed effect may be the resultant of an unknown 

 number of independent as well as interdependent reactions, 

 which may be both physical and chemical in nature. A tempera- 

 ture coefficient derived under such circumstances must be of 

 doubtful significance. Changes of temperature may affect 

 consecutive reactions in totally different ways. Mellor (1909) 

 and Osterhout (1917) discuss this aspect of the problem and cite 

 illuminating examples. 



With these considerations in mind we may note that the tem- 

 perature coefficients found in this investigation are in the main 

 those to be ascribed to chemical reactions. This would indicate 

 that somewhere in the series of consecutive processes ending in 

 death, the slowest or limiting reaction velocity was that ascrib- 

 able to the ordinary chemical reaction. Of more theoretical 

 interest is the finding in experiment 7 that the death rates for 

 Bad. coli increased for 10° intervals from 0° to 30°C. in an ex- 

 ponential ratio of the form: 2° : 2' : 2' : 2^ a regularity that 

 seems more than accidental. Its interpretation is difficult for 

 the present but we have here a suggestion for a possible method 



