154 SUMMER DIARRHCEA 



and 1906 have been constructed from his figures. These curves 

 very clearly show the close relationship between the number of 

 flies captured and the diarrhoea deaths. It will be noticed that 

 in 1904 the epidemic was nearly a month earlier than in 1906. 



" It is evident that if flies are responsible for the ascent of the 

 diarrhoeal curve, it requires a goodly number to start it on its 

 upward course ; in fact, we must imagine to ourselves an infant's 

 food visited by a very large number of flies, some only of which 

 carry infection from the excreta of a previous case. The 

 quantity of infection in a given case will thus at first be 

 small, and a number of slight cases will be produced, which 

 scarcely attract notice ; but as flies multiply, and cases multiply, 

 the amount of infection conveyed to foods will increase... .As 

 flies and cases continue to multiply, however, the fly-borne 

 infection predominates more and more, and we get a massive 

 and fatal infection which quite overshadows the direct process." 



In Manchester flies were observed "to cluster especially 

 about the nose and mouth of infants suffering from diarrhoea, 

 and no doubt frequently visit napkins and diarrhoeal excreta of 

 older children" (see p. 144). The ways in which flies so infected 

 can contaminate foods such as raw and condensed milk have 

 already been fully dealt with. Flies are known to move from 

 house to house within restricted areas (see p. 112), and no doubt 

 those coming from infected houses often contaminate syrup, 

 milk, fruit, etc., in houses and shops. 



3. — Towards the end of summer many flies may be captured 

 in houses, which are unlikely to convey infection either because 

 they are numbed with cold or are attacked by Enipusa mnsccE 

 (see Chapter XXlli). Hence at this period the numbers captured 

 may not bear a close relation to the diarrhoea deaths. 



4. — Even breast-fed infants ma}' become infected by articles 

 such as comforters soiled by flies, or by flies settling on their 

 lips or noses, or directly by infected adults. 



5 and 6. — Nivcn clearly shows that rainfall only has an 

 indirect relationship to diarrhoeal fatality. Heavy and continuous 

 rainfall may reduce the temperature and consequently the 

 number of flies and deaths, but sometimes heavy rain may be 

 accompanied by a rise in the atmospheric temperature, and then 



