ELIMINATION OF SPURIOUS TESTS FOR B. COLI 343 



been identified as B. proteus or B. fiuorescens-liquefaciens. In 

 all cases the original colonies were well separated on the plate, 

 and usually we have been at a loss to account for the association 

 of two separable species in a supposedly pure culture giving the 

 reactions ascribed to B. cloacae. In some other cases contami- 

 nations subsequent to the primary isolation were responsible; 

 this was assumed in those few instances when the gelatin lique- 

 fier proved to be a spore bearer. At any rate we have come to 

 question whether so-called B. cloacae is not often a mixture of a 

 gelatin liquefying non-lactose fermenter with B. coli. In our 

 studies all such organisms were planted on plain agar and set 

 aside for further study. Unfortunately in some of them we lost 

 one of the organisms, or must we say, the culture lost its lac- 

 tolytic properties? In the case first referred to above the cul- 

 ture would not ferment lactose when re-tested, though it had 

 formerly done so actively; it conformed therefore to the usual 

 description of B. proteus and we conceive that we lost B. coli. 

 In one other case of this series we clearly separated B. proteus 

 from B. coli by plating out in gelatin from a fermentation tube 

 of lactose broth inoculated with what appeared to be 5. cloacae. 

 An interesting point in this second series is that the propor- 

 tion between 1st and 2nd trials at isolation of B. coZz" is reversed, 

 as compared to the 1st series, over one-half of the successful 

 isolations coming from the second trial, indicating, we believe, 

 a loss in number and vigor of B. coli due to storage. There were 

 three samples in which B. coli was found the first tune but not 

 the second; the smallest amounts of water in which B. coli could 

 originally be found, according to the Bureau of Sanitary Engi- 

 neering were respectively 10 cc, 1 cc, and 0.1 cc. There were 

 two samples originally containing B. coli in 10 cc. and 1 cc. 

 respectively in which B. coli was found the second time and not 

 the first and one sample originally showing B. coli in 10 cc. in 

 which we never found it. As against these data showing the 

 disappearance of B. coli from standing samples of polluted 

 water we have only one case in which the presence of gas forming 

 anaerobes could not be corroborated in the repeated test. 



