94 BACTERIOLOGICAL EXAMINATION OF WATER. 



With regard to the organisms which slowly liquefied gelatine, 

 Gordon found that sub-cultures varied in the date of com- 

 mencement of the liquefaction ; some gaining, others losing, 

 power of producing liquefaction. 



During the present year I have carefully studied the cultural 

 reactions of 150 varieties of B. coli isolated partly from 

 normal and partly from typhoid stools. Between 50 and 

 60 per cent of these cultures produced gas in sugar media, 

 soured milk, formed indol after seven days incubation at 

 37 C. in peptone, and gave a coloured growth on potato. 

 About 30 per cent produced gas in sugar media, but failed to 

 coagulate milk and form indol. The remaining cultures failed 

 in only one of the typical reactions ; the coagulation of milk was 

 the reaction most frequently absent. Some of the varieties 

 of B. coli isolated from typhoid stools were very peculiar. The 

 colonies up to forty-eight hours incubation were typical ; but 

 later on wedge-shaped processes grew out from the margin of 

 the colony so that it slowly acquired a rosette-like appearance, 

 and the gelatine slowly liquefied. When stabbed into gelatine 

 tuft-like processes grew out from the growth along the line of 

 inoculation. All these colonies produced gas in sugar media, 

 soured milk, and gave very marked indol reaction ; in many 

 respects they resembled the Proteus cloacinus, described by Laws 

 and Andrewes as one of the commonest organisms in sewage. 



A consideration of Gordon's table suggests that possibly the 

 varieties may be derived from the typical B. coli as a result of 

 unfavourable surroundings, want of food, &c. Some support 

 to this view is derived from observing the changes produced in 

 B. coli after prolonged immersion in sterile sewage. I found 

 that a typical B. coli, freshly isolated from a normal stool, after 

 forty-two days immersion in sewage, gradually lost its power 

 of producing indol, and became much less resistant to the 

 action of carbolic acid. But agglutination experiments made 

 with the serum of animals immunised against B. coli, point 

 strongly to each race being quite distinct. I have found that 

 cultures of B. coli, derived from the same source and quite 

 indistinguishable by cultural tests, produced sera which were 

 quite distinct, each serum reacting to its own bacillus, but 

 failing to influence in any way the other cultures. Jatta's 



