The Fermentation Tube 229 



probabl}^ a putrefactive organism, pure and simple. The up- 

 per limit excludes B. cloaccB, which, in spite of its name, I can- 

 not range among fecal bacteria. Between the limits of forty 

 and seventy per cent, of gas are included all varteties of B. 

 coli, the hog cholera group, B. lactis aerogenes and Friedland- 

 er's bacillus. 



There are several objections which may be urged against 

 this as against any approximative method. In the first place 

 it does not include a large number of pathogenic species, 

 among them B. typhosus and Sp. cholera Asiatics. But, it may 

 be answered, the object of the method is not to reveal all pos- 

 sible disease germs but to use the colon group as an index of 

 pollution because, as I maintain, they must come directly 

 from the digestive tract. The presence of B. coli even in 

 small numbers is amply sufficient to make anj^ water sus- 

 pected. 



In the second place it may be claimed that the evolution of 

 gas may be either checked or augmented in the presence of a 

 number of species in the same tube. This objection involves 

 the rather broad subject of antagonism among bacteria, which 

 cannot be discussed here. There are, however, a few facts 

 which show the objection to be in the main pointless. In the 

 thermostat only very few bacteria from water develop owing 

 to the high temperature, so that rarely more than one species 

 survive and multiply in the fermentation tube if the quantity 

 of water added be not too great. Again, the presence of two 

 gas-producing bacteria in the same tube is not likely to occur 

 owing to their relative scarcity. To test their mutual behav- 

 ior, however, I inoculated a number of tubes simultaneously 

 with two different gas-producing species. In general B. coli 

 produced the quantity of gas peculiar to it, unless B. cloaccs 

 was inoculated with it. In one out of three trials B. clo- 

 accs triumphed and drove out all fluid from the closed branch, 

 in the other two B. coli conquered. There may, therefore, be 

 an occasional masking of the presence oi B. colihy B. cloaccs. 

 This error will not occur if small quantities of water be used 

 or if the experiment be repeated in the event that more than 

 half the tubes inoculated show gas production. 



