177 APPENDIX A 



distributed than ever before the amount of diarrhoea, 

 instead of showing early and rapid decline, would still be 

 increasing. It would almost appear that the advocate of the 

 'fly-borne diarrhoea hypothesis' must necessarily fall back 

 in support of his theory upon the hypothetical organism, 

 conveyed by the fly, which he may claim is affected by 

 temperature in such a way as to bring about correspondence 

 between the diarrhoea curve and the fly curve. The very 

 closeness of the correspondence between these two curves 

 may indeed from this point of view be thought of as 

 constituting a difficulty rather than a point in favour of the 

 hypothesis that summer diarrhoea is caused by flies." [The 

 italics are mine. C. G. EL] No one, I think, would argue 

 this. What is argued is that there are excellent grounds 

 for believing that flies carry the infective organism or 

 organisms of summer diarrhoea. 



Niven (1904) suggested that the explanation of the falling 

 off of the diarrhoea curve while the number of flies still 

 remain large might be due to the exhaustion of susceptible 

 material. 



The one point which does not appear to be considered and 

 which, I think, will explain this seeming objection is the 

 great susceptibility of flies to changes of temperature, which 

 fact all who have studied closely the habits of M. domestica 

 and its allies will admit. When the temperature falls, flies 

 become more sluggish and retire more into the shelter of 

 houses and other buildings, although their numbers may 

 still be considerable. It is necessary, therefore, to study the 

 temperature curve in addition to the fly and diarrhoea curves. 

 If this is done, it is usually found that the fall in the 

 number of flies is preceded by a fall in the temperature and 

 that these two curves are associated closely, that is, the 

 numerical activity of the flies since the numbers caught are 

 more indicative of their numerical activity than of their 

 numerical occurrence is dependent upon the temperature, 



