PART II. CHARACTERS AND DIAGNOSIS 73 



not be misled into the notion that it is the cause of the trouble. 

 In Egypt the largest infections of E. coli we have encountered, and 

 some of these very large indeed, have been in perfectly healthy men 

 with normal stools. Further, taking into consideration a large 

 series of hospital cases with diarrhoea the percentage of E. coli 

 infections is considerably below that of healthy men in the same 

 locality. The cases which have intestinal symptoms of diarrhoea 

 with an E. coli infection invariably retain the E. coli infections 

 when they recover, and as will be seen from the records of the 

 E. histolytica infections treated by us, the E. coli infection almost 

 invariably remains after the E. histolytica has disappeared. Of 

 greatest importance, however, from the point of view of the patho- 

 genicity of E. coli .axe the cases of infection which have been 

 followed for very long periods. It is impossible to state that these 

 individuals never have attacks of diarrhoea at any time, for who 

 does not at some time or another during the course of every year, 

 especially in a warm climate? But it can be definitely asserted 

 that they show no abnormal tendency to diarrhoea. There is, 

 therefore, in our opinion no justification for regarding E. coli as 

 pathogenic. 



A note of warning may perhaps be sounded for the benefit of 

 those medical men to whom the subject of the intestinal protozoa 

 of man is a new one. Those who have not been accustomed to 

 study these organisms in any detail, and have only recently begun 

 clearly to differentiate between them, must not be led astray by 

 their comparative attractiveness and large size. The great majority 

 who examine for intestinal organisms do so only in the case of sick 

 people who have some intestinal disorder, and they have no means 

 of comparing their findings with what occurs in healthy individuals. 

 Accordingly it seems quite natural to them to attribute to the 

 intestinal protozoa, both amoebae and flagellates, which are com- 

 paratively easily recognized, pathogenic powers which they may 

 not possess. It seems to us that there is only one possible plan of 

 arriving at a safe conclusion as regards the relative frequency of 

 protozoa in healthy and sick people, and it is one which is hardly 

 practicable on a large scale. It is to take a large series of healthy 

 individuals and to examine their stools for about one week while 

 they are being purged by means of salines and to compare the 

 results thus obtained with those from a similar series which are 

 suffering from diarrhoeic conditions acquired naturally. In this 

 manner the relative incidence of some of the amoebic and flagellate 

 infections in healthy and diarrhoeic subjects might be obtained. 



