248 BACTERIOLOGY. 



incubator at 37 C. for twenty-four hours. From 

 this plate cultures are made. In our experience this 

 and other methods have not enabled us to detect the 

 typhoid bacillus where we have failed to find it by 

 making direct plate cultures. As a matter of fact, the 

 typhoid bacillus is found in such a small number of the 

 specimens where we actually know that it is or has 

 been present in the water from which they were ob- 

 tained, because of cases of typhoid fever which have 

 developed from drinking the water, that we must con- 

 sider our lack of finding the bacillus in any given 

 case as absolutely no reason for considering the water 

 to be free from danger. Another serious drawback 

 to the value of the examinations is that they are fre- 

 quently made at a time when the water is really free 

 from contamination, though both earlier and later the 

 bacillus was present; it is hardly worth while, there- 

 fore, except in careful experimental researches, to ex- 

 amine the water for the typhoid bacillus, but rather 

 study the location of the surrounding privies and sources 

 of contamination. The colon bacilli are far more easy 

 to detect, because they are apt to be more abundant, 

 and, also, because they grow more readily in artificial 

 culture media. A method suggested by Theobald 

 Smith is of value in both finding and excluding the 

 presence of bacilli of the colon group. He adds a few 

 drops of the suspected water to glucose nutrient bouillon 

 in fermentation tubes, and keeps it at 37 C. for from 

 thirty-six to forty-eight hours. If no fermentation 

 occurs no colon bacilli are present. If it does occur 

 plates are made and the bacteria isolated and tested. 

 In the bouillon the colon bacilli when present usually 

 increase in numbers and are then readily detected. The 



