22 



every two or three days. This estimate of the prevalence of 

 flies, together with the deaths from diarrhoea, were then 

 plotted as ordinates against weeks as abscissae. Dr. Niven's 

 observations extend over five years, and refer to several dis- 

 tricts in Manchester. Dr. Hamer's organisation for the 

 capture of flies was on a larger scale, and his observations 

 were made at various parts of London during 1907, 1908, and 

 1909, mostly in the neighbourhood of refuse and manure 

 depdts where flies are bred. Another excellent series of 

 observations was made by Dr. E. Dudfield (1912) in 

 Paddington during the summer of 1911. It concerns a 

 smaller number of cases, but this enabled a more intensive 

 study of many factors contributing to the epidemic to be 

 made. 



The first point brought out by all these observations is the 

 dependence of both the number of flies and the epidemic 

 upon the cumulative effect of previous warm weather as, 

 for instance, is indicated by the earth temperature 4 feet 

 below the surface a fact to which attention was first drawn 

 by Ballard (1889). The curve for flies rises first, followed 

 soon by that for cases. The latter is longer in getting going ; 

 this is particularly marked in the cool summer of 1907. Both 

 flies and cases reach their maxima about the same time. In 

 1907 the number of cases is diminishing, whilst flies are 

 still on the increase. In 1908 and 1909 both curves decline 

 together, but that for flies more gradually. 



A notable feature of the curve is that the same fly popula- 

 tion which is accompanied by a rise in the number of cases 

 in early August corresponds with a fall in early September. 

 This fact has been quite properly emphasised by the critics 

 of those advocating the importance of flies. 



If we propose to regard the time relations of fly pre- 

 valence and diarrhoea cases as more than an interesting 

 coincidence, we must find a satisfactory explanation for the 

 above facts. 



Possible Explanations of Dependence of Diarrhoea Epidemics 

 upon the Accumulative Effect of Temperature. 



The reason why the number of flies is dependent upon 

 this factor is obviously because the generation time (cycle 

 from egg to egg) varies with temperature and requires 

 three weeks or upwards in our climate. Months of warm 

 weather are, therefore, required to produce any multitude 

 of flies from the few surviving the winter. 



Why the epidemic should exhibit this dependence is not 

 obvious. Not knowing the etiology of diarrhoea, we must 

 assume some sort of infective agent. From analogy 

 with typhoid fever, cholera, acute food poisoning, and 

 dysentery we will suppose that we have to deal with 

 some form of bacterium growing in the outside world 



