. THE COLON GROUP OF BACILLI 127 



Of course it must be remembered that the advantages 

 of lactose-bile over dextrose broth are partly due to 

 the inhibiting effect of the bile salts and partly to the 

 use of lactose instead of dextrose which cuts out the 

 dextrose-positive lactose-negative group to which allusion 

 has been made earlier in the chapter. The relative 

 importance of these two factors, lactose and bile, is 

 well brought out in a study by Stokes and S toner (1909). 

 These authors have compared a considerable series 

 of preliminary enrichment tests followed by final 

 isolation in dextrose broth, lactose broth and lactose 

 bile. Of 567 colonies from positive dextrose broth 

 tubes only 52 per cent were colon bacilli; of 3752 

 colonies from positive lactose broth and lactose bile 

 tubes, 88 per cent of the lactose broth colonies and 95 

 per cent of the lactose bile colonies were B. coli. 



With sewages and heavily polluted waters in par- 

 ticular the lactose-bile medium has proved of the 

 greatest value. When a large proportion of sewage 

 is present the colon bacilli are fresh from the intestine 

 and apparently able to resist the antiseptic salts. On 

 the other hand, the high numbers of other bacteria 

 present make the danger of overgrowths particularly 

 great. With waters of fair quality, such as those 

 with which we ordinarily deal in sanitary water analysis, 

 lactose bile is open to the same objection as phenol 

 broth and the Eijkman test though in less degree. 

 It inhibits not only the overgrowing forms but the 

 weaker representatives of the B. coli group itself, and 

 the net effect is to diminish positive results. 



Hale and Melia (1910) inoculated unsterilized water 



