l68 SUMMER DIARRHCT.A 



From this table it will be seen that 77 varieties of bacilli which do not ferment 

 lactose or liquefy gelatin were isolated. Of these, 45 were isolated only once, seven 

 twice and eight three times. 



The varieties isolated belong to 6 groups and 20 sub-groups. Bacilli belonging to 

 groups A and H were found with equal frequency (about 2 •*/()) in flies from diarrhoea 

 infected and non-infected houses. Bacilli belonging to groups B and C were met with 

 twice as frequently in flies from diarrhoea infected houses. These slight differences in 

 the degree of infection appear to be of little importance. On the other hand bacilli 

 belonging to group G, sub-group a (Morgan's bacillus), were found much more 

 frequently (nine times as often) in flies from diarrhcea infected houses than in flies fiom 

 non-infected houses. 



Up to the present a sufficient number of observations have not been made on 

 which to determine definitely the relative frequency of the various types of bacilli in 

 flies at different times during the season. The observations made on Birmingham 

 flies showed that from August 14th to October ist the weekly relative proportions of 

 bacilli belonging to groups B, C and H varied very slightly. Members of group A 

 were found relatively more frequently during the last three weeks, whereas members 

 of group G, sub-group a, were much more frequently present during the first five 

 weeks. No member of this group was isolated from this series after September i6th. 

 The observations on the seasonal occurrence of this group are given in the following 

 table : 



TABLE 22. Sliozui/ig the zocekly percentage of ui feet ion with 

 bacilli of the Ga type (^Morgan's bacillns) of flies from 

 diarrho'a infected Jionses. 



* Only a few flies could be obtained for examination. 



l-'rom this table it is evident that during the year 1911, both in Cambridge and in 

 Birmingham, a much higher proportion of flies were infected with Morgan's bacillus 

 (Ga) during the fourth week in August than at any other time. 



(3) TJie relative frequency of the occnmnce of varieties of bacilli 

 zvhich do not ferment lactose and do not liquefy gelatin in flies 

 and in the excreta of children. 



In order to show the relative frequency of the occurrence of varieties of non-lactose 

 fermenting bacilli in the excreta of children, healthy and suffering from diarrhoea, and 



