590 PROCEEDINGS OF SECTION I, 



inoculation of the infective material, or else when there are pos- 

 sibilities of repeated small doses, does infection occur. The oc- 

 currence of cases where a single act of pollution of food or milk 

 results in successful infection of one or more human beings may 

 therefore probably be considered as having presented opportunity 

 for incubation in the chain leading from excreta to case. 



From certain considerations the opportunities for successful 

 spread by food may be even more narrowly circumscribed, for, 

 from bacteriological data, I have a strong feeling that it is sterile 

 or relatively sterile food, milk, &c., which, accidentally inoculated 

 with B. typhosus, is the most likely medium for causing a suc- 

 cessful spread. 



We know that it is sterilized food such as brawn, tinned food, 

 sausages, cooked cold meats, &c., that are the usual vehicles of 

 spread of food poisoning infection, and the natural presumption 

 is that they form more suitable media for growth and multiplica- 

 tion of accidentally inoculated organisms, and it seems probable 

 that in the spread of the typhoid bacillus a similar favouring in- 

 fluence will occur. There is no doubt that side by side with 

 other organisms B. typhosus is a relative weakling under ordinary 

 artificial cultural conditions, and the degree of its multiplication 

 is almost undoubtedly facilitated by tlie absences of most of the 

 organisms likely to be present in, say, milk; whether the inhibi- 

 tion of the growth of B. typhosus is due to the presence of the 

 bacilli, or simply to the acidity produced, is immaterial here; it 

 can be taken, I feel sure, that the more the milk or food ap- 

 proaches sterility the more suitable the opportunity for multi- 

 plication. Relative sterility is found in quite fresh milk, boiled 

 milk, boiled puddings, cold meat, &c. 



To return to the question of our investigation — ^the fixation 

 of the probable date of infection of the cases. This can be as- 

 sumed to be about fourteen days before the definite onset of the 

 illness, but unfortunately longer or shorter incubation periods are 

 known (8). 



Also very important is the consideration of the relation borne 

 by the cases to each other. Sometimes there are " explosive " 

 outbreaks where a large number of cases develop the sickness on 

 or about the same date, strongly pointing to an infection occurring 

 on one particular date. This may occur in water infection, but 

 it is likely to be found that the first explosion will be followed 

 by more cases for some time afterwards, depending on the nature 

 of the pollution of the water, whether it be by some temporary 



