250 



L. A. EOGERS, W. M. CLARK AND H. A. LUBS 



we have studied in the present series all but one ferment both 

 lactose, a sugar of animal origin, and melibiose, a plant sugar 

 closely related to lactose. The one culture which failed to 

 ferment lactose failed to ferment melibiose. 



In this scheme of classification the positions assigned to the 

 lower groups are determined in large measure by the gas forma- 

 tion. While the data found in the literature are for the most part 

 unreliable and the results which we have obtained are meagre, 

 the following determinations are significant. 



Cultures of B. paratyphi, B. enteritidis and B. dysenteriaeirom. 



» the American Museum were grown at 37°C. on the medium used 



in the gas determinations with colon cultures and the gas volume 



and ratios were determined by the accurate method previously 



described. 



The results are as follows. 



Keyes and Gillespie (1912) have reported hydrogen and carbon 

 dioxide as products of the activity of B. typhi. This might be 

 considered to invalidate the position given to B. typhi in figure 5. 

 However, Keyes and Gillespie obtained only about 0.2 cc. total 

 gas per 10 cc. with B. typhi and the gas ratio CO2/H2 was large 

 and inconstant. 



Theoretically it is of course unnecessary to assume that carbon 

 dioxide is a constant product of the metabolism of bacteria, yet 

 experimentally there are evidences that small amounts of this 

 gas are formed by certain ''non-gas producing" bacteria. This 

 is neglected in figure 5 as an unessential consideration. 



