300 DISTRIBUTION OF BACTERIA 



logical water analysis the number of bacteria per cc. is 

 determined by plate cultures. It is impossible to judge the 

 character of a water by the number of bacteria it contains 

 alone, but the bacterial content is frequently a matter of im- 

 portance taken in connection with other factors. The number 

 of different kinds of bacteria present in water is also important, 

 but, as in the quantitative analysis, does not give information 

 which can be used alone. It is very rare that B. typhosus or 

 Msp. comma can be detected in the water. The reason for 

 this is, among others, that the analysis is usually not at- 

 tempted until considerable time has elapsed since the introduc- 

 tion of these bacteria into the water. These bacteria live for 

 only a short time in water under ordinary conditions and, 

 therefore, have probably disappeared or become extremely 

 few in number at the time of an analysis, and again it is a 

 very difficult matter to separate B. typhosus from other closely 

 related bacteria. What is sought for in a water analysis is the 

 presence of B. coli. This is usually done by introducing the 

 water to be examined in varying amounts from one tenth of a 

 cc. to ten cc. into special culture media. The culture medium 

 most generally used is dextrose bouillon in fermentation tubes. 

 The presence of this bacterium is determined in what is fre- 

 quently called the presumptive test by its ability to ferment the 

 sugar in the dextrose broth of the fermentation tubes, with the 

 production of certain amounts of gas of a definite formula, 

 the production of red colonies on lactose litmus agar, the 

 nonliquefaction of gelatin, and the production of indol. 



