The Fermentation Tube 205 



fers with rare exceptions, quite markedly from that upon glu- 

 cose or lactose. The examination of cultures from different 

 sources has revealed two distinct varieties, one of which pro- 

 duces a considerable quantity of gas, the other little or none. 

 With the former variety the type of gas production varies 

 somewhat from culture to culture. In general the fluid is 

 driven out very slowly and the gas production may last sev- 

 eral weeks. These statements are well illustrated in table IV. 



When the gas production goes on very slowly the growth 

 in the open bulb becomes exceedingly abundant. This is 

 most probabl}^ due to the slow neutralization of the bacterial 

 alkali, formed in the open bulb, by the acid resulting from 

 the slow fermentation in the closed branch. The gradual en- 

 trance of this acid fluid into the bulb acts as a continuous 

 stimulant to the multiplication of bacteria there. When the 

 gas production is rapid the fluid in the bulb remains acid and 

 the growth speedily subsides. 



What the true significance of the varying behavior of the 

 B. coli group towards cane sugar is, can be determined only by 

 more extended investigations. I venture to suggest however, 

 that the saccharose fermentation may require in the slow fer- 

 mentation the presence of an inverting ferment while that of 

 lactose and glucose goes on without it. This ferment is ap- 

 parently no longer formed by some bacteria otherwise not dis- 

 tinguishable from B. coli. The whole subject is very interest- 

 ing and seems to indicate either that this species may readily 

 lose the capacity to act on cane sugar or else that it is in a 

 transition stage towards the more pathogenic species of this 

 large group of bacteria. The peculiarity of the saccharose 

 fermentation suggests the thought that the presumable ferment 

 is formed only in the fluid in contact with oxygen and that it 

 very slowly diffuses thence into the closed branch. A layer 

 of sterile oil on the fluid of the bulb would perhaps answer 

 this question. But I have had no opportunity to try this ex- 

 pedient. We may summarize the facts concerning the gas- 

 producing power of B. coli comimuiis briefly as follows : — 



In feebly alkaline peptone bouillon containing 2 per cent, of 

 glucose or lactose, about 50 to 60 per cent, of the closed branch 

 of the fermentation tube will be occupied by gas in 3 or 4 



