262 THE RELATION OF FLIES TO SUMMER DIARRHOEA 



One of the most interesting and highly suggestive results of the 

 research was the discovery of Morgan's bacillus in flies. " Batches 

 of flies came for examination from infected and uninfected houses 

 in Paddington and from a country house situated many miles from 

 London, where no cases of diaiThoea had occurred, at any rate, 

 within a radius of two miles. The flies were killed with ether 

 vapour and crushed Avith a sterile rod in peptone broth. The 

 result was that Morgan's bacillus was isolated from nine of the 

 thirty-twQ batches from infected houses and from one of the 

 thirty-two batches from uninfected houses^ It was also got in 

 five out of twenty-four batches from the country house." Dr 

 Morgan in the course of a letter to me says: "I certainly think 

 they are carriers of summer diarrhoea, and the variety I especially 

 suspect of doing this is the Musca domestica." 



In a study of the micro-organisms occurring in flies caught in 

 normal surroundings and in diarrhoea-infected locations Graham- 

 Smith (1912) found that a greater proportion of flies is infected 

 with the non -lactose-fermenting bacteria during August and the 

 early part of September than at other times. Of the groups of 

 bacilli into which these organisms can be divided the Morgan or 

 Ga gi'oup is the only one which occurs frequently in flies from 

 diarrhoea-infected houses and rarely in non-diarrhoea-infected 

 houses. He considers it certain that flies infected with Morgan's 

 bacillus can contaminate materials on which they feed or over 

 which they walk. 



The epidemiological evidence for and against the hypothesis 

 that the house-fly is an agent in the dissemination of summer 

 diarrhoea has been subjected to a most careful and detailed 

 examination and analysis by Martin (1913), whose paper is in 

 my opinion the most judicious criticism of the problem as it now 

 stands. 



After pointing out that the observations of Niven and Hamer 

 indicate the dependence of both the number of flies and the 

 epidemic upon the cumulative effect of previous warm weather, 

 Martin calls attention to the fact that a notable feature of the 



1 Morgan's bacillus has also been isolated from naturally infected dies by 

 Nicoll (1911) and Cox, Lewis and Glynn (1912). See pp. 245 and 296. 



