23 



with which our ailments might become infected. In this 

 case, would the dose of infection be dependent upon the 

 accumulated effect of temperature during the previous three 

 months ? Assuredly not. The generation time of bacteria 

 is so short in the case of bacillus coli it is 1 hour at 

 20 C. and hour at 30 that the number present at any 

 moment is to all intents and purposes dependent only 

 upon the mean temperature during the last few days. 

 To take a homely illustration, whether one's milk sours 

 does not depend upon the temperature last month, but last 

 night. 



In the accompanying table (II.) I nave sefc out the rate of 

 multiplication of B. coli at 20 C. , and that of flies inside a 



TABLE II. Showing the Rate of Multiplication (Maximum 

 Observed) of House Fly and B. Coli. 



house during a summer in England. You will observe that 

 one female fly might under these conditions possess 150,000 

 progeny in 60 days, but the bacteria derived from one 

 organism become innumerable in two days. Unless, there- 

 fore, the effect of two months' summer weather be to change 

 the nature of the organism so as to enhance its virulence, it 

 is quite impossible to account for this striking characteristic 

 of the epidemic by chance contamination of food-supply by 

 pathogenic bacteria. In a state of innocence regarding 

 the etiology of the disease we have no means of knowing 

 whether any change occurs in virulence under the 

 influence of continued hot weather, but it is just con- 

 ceivable. 



