232 AIDS TO BACTERIOLOGY 



unanimously regarded as the most important datum 

 obtained in the ordinary bacteriological examination of 

 water. If the colon bacillus is found, it is usual to report 

 it as absent in such a quantity, but present in so many 

 cubic centimetres. Thus, if typical B. coll be found in 

 the tube to which 10 c.c. of the water had been added, 

 but that receiving 5 c.c. remained unchanged, the organ- 

 ism is present in 10 c.c., but absent in 5 c.c. 



Savage came to the following conclusions (1902), the 

 justification of which is very generally conceded: Waters 

 which show no B. coli in 50 c.c. are of a high degree of 

 purity, and therefore the proved absence of this organism 

 in this amount, and still better in larger quantities, is 

 of great value. B. coli should be absent from at least 

 50 c.c. of spring water, possibly from greater amounts. 

 In upland surface waters the presence of B. coli in 40, 

 10, or even 2 or 1 c.c., means contamination, but not 

 necessarily a contamination which it is essential to pre- 

 vent. It may be from contamination with the excreta 

 of animals grazing on the gathering areas, and is by no 

 means necessarily from sewage or other material contain- 

 ing specific organisms of infection. If B. coli are present 

 in numbers greater than, say, 500 per litre (or even in 

 that amount), such a water is suspicious, as it is rare to 

 get so many B. coli in a water from the kind of animal 

 contamination indicated, and further investigation is 

 desirable. In filtered samples the number of B. coli is 

 as a rule considerably reduced. In surface wells B. coli 

 in large numbers indicate surface or other contamination, 

 generally very undesirable, if not actually dangerous. 



Deep-well or spring water should not usually contain 

 B. coli in 100 c.c. 



As regards atypical B. coli, we cannot do better than 

 again quote Savage: ' The nearer these . . . approach 

 typical B. coli in their characters, the more nearly are our 

 numerical standards for that organism applicable to them, 

 while if they lack essential characters, a proportionately 

 greater number must be present to justify an adverse 

 opinion.' 



Enumeration of Streptococci. Quantities similar to 

 those used in the estimation of B. coli are inoculated into 

 glucose broth or glucose neutral red broth, and incubated 

 at 37 C. for forty-eight hours, when hanging- drop or 



