202 BACTERIOLOGICAL EXAMINATION OF WATER. 



Parietti's fluid), and following Hankin's technique if the next 

 tube or next but one to this be taken it will contain ten to 

 eleven drops of Parietti, or 0'15 per cent, of carbolic, in which 

 B. typhosus often does not grow in twenty-four hours. Con- 

 sequently by Hankin's method B. typhosus may not be detected 

 when it is accompanied by B. coli. Hilbert carefully investi- 

 gated Hankin's method by adding a twenty-four hours broth 

 culture of B. typhosus to water from different sources. He 

 first estimated the number of bacilli which would be found in 

 varying dilutions of the twenty- four hours broth culture, and 

 came to the conclusion that with a dilution of 1-100,000,000, 

 three drops of the fluid would contain a typhoid bacillus. The 

 following table gives the results of his experiments with three 

 different samples of water : 



In pure filtered water the B. typhosus was detected when only 

 three drops of a dilution of 1-100,000,000 was used for the 

 experiment. In water from a pump in the Drummstrasse the 

 same result was obtained with a dilution of 1-10,000,000. In 

 Pregel water, which contained the B. coli, the B. typhosus 

 could not be detected. Hilbert then endeavoured to tind out 



