28 



proximity to brickfields in which were deposited daily some 

 30 tons of fresh house refuse, which bred incalculable 

 numbers of flies " directly the meteorological or seasonal 

 conditions became favourable for their development. 



Dr. J. Niven (1910, p. 45) is of opinion that the close corre- 

 spondence between flies and cases of fatal diarrhoea receives 

 a general support from the diarrhoea history of sanitary sub- 

 divisions of Manchester, and that the few facts available 

 for the study of correspondence of flies and fatal cases in 

 different subdivisions (Manchester) in the course of the same 

 year also lend support. No other explanation, in his view, 

 even approximately meets the case. 



Peters (1910), in his very exhaustive study of the epidemic 

 in Mansfield in 1908, paid especial attention to the possible 

 part played by flies. From an analysis of his observations 

 he, I think justly, came to the conclusion that the evidence 

 pointed to the fact that, whatever part they might play as 

 carriers from an infected household to a neighbour, flies did 

 not bring infection with them from the manure heaps where 

 they had been bred. Peters constantly noted that the 

 individual seemed to be a source of infection to the neigh- 

 bours adjacent to the house, within the precincts of which 

 his excreta, if infectious, would most likely be deposited. 

 This he points out could be most readily explained by the 

 supposition that flies were the agents of dissemination of 

 the infection, but from the frequent want of correspondence 

 in different localities between numbers of flies and amount 

 of diarrhoea, he finds no support for this hypothesis from the 

 distribution in space of flies and cases. 



I doubt very much whether any evidence of great value 

 could be obtained upon this point. Even supposing it to be 

 true that fly carriage is of first importance, I should expect 

 that if all the facts were known a much higher correlation 

 would be discovered between diarrhoea and careless- 

 ness with regard to disposal of excreta and protection of 

 food from the visitation of flies than between diarrhoea and 

 fly prevalence. 



Many of the facts which I have brought forward merely 

 indicate some form of infective agent, and do not necessitate 

 recourse to the hypothesis that carriage by flies dominates 

 the situation. I would, however, point out : (1) that the 

 fly-carrier hypothesis is the only one which offers a satis- 

 factory interpretation of the extraordinary dependence of the 

 epidemic upon the accumulated effect of temperature ; 

 (2) that it offers a ready explanation of the spread of infec- 

 tion to neighbouring children who have no direct personal 

 contact with the patient ; (3) that the peculiarities of the 

 relation in time between fly prevalence and the epidemic in 

 different localities are not inconsistent with the view that 

 fly carriage is essential to epidemicity. 



No other interpretation which is, so far, forthcoming is 

 nearly so satisfactory, and it is at least worthy to guide in 

 the meantime our efforts at prevention. 



