224 Theobald Smith 



to the presence of carbo-hydrates. Some years ago, Petru- 

 schky^° examined the acid and alkali-producing functions 

 of bacteria by using as a culture medium specially pre- 

 pared whey from milk. I called attention to the fact that 

 such classification had only a limited value since it depended 

 entirely on the composition of the culture fluid'. The whey, 

 having as an important ingredient, lactose, would prove only 

 such bacteria acid-producing which were able to cause fermen- 

 tation of the milk sugar while those which could not do this 

 would show themselves as alkali producers. Bearing on this 

 subject are the statements made by bacteriologists in the early 

 days of this branch of biology that cultures of many bacteria 

 are at first slightly acid before becoming alkaline. I suggest- 

 ed that this was probably due to traces of sugar in the culture 

 fluid and I was able to prove this by causing an oscillation 

 from an acid to an alkaline reaction and back again by adding 

 at intervals small quantities of glucose to the bouillon. The 

 alkali formed during the multiplication of bacteria was neu- 

 tralized by the acid derived from the fermentation of the glu- 

 cose. If this was small in quantity the acid or acids were 

 formed in correspondingly small quantities and the alkaline 

 reaction soon reappeared. I was able to show furthermore 

 that the addition of small quantities of fermentescible sugar 

 greatly favored the multiplication of bacteria by keeping 

 down the alkaline reaction. After I began testing peptone 

 bouillon for muscle glucose with gas-producing bacteria, I 

 found that in bouillon free from sugar the multiplication of 

 various acid producing bacteria such as streptococci, staphylo- 

 cocci, B. typhosus, B. diphtherics, B. coli, andi?. cholera: suisisnot 

 attended with any acid reaction, either temporary or permanent. 

 So far as my observations have gone they show that all bac- 

 teria are alkali producers in bouillon free from carbo-hydrates, 

 and that when one or the other of this group is present a very 

 large number of the most easily cultivated bacteria are acid 

 producers. This two fold activity probably serves a useful 

 purpose in keeping the medium in which they live more or 



tures of the bacillus of Friedlander ethyl alcohol, acetic acid and a little 

 formic and succinic acid. Grimbert^-' detected among the products of 

 B. orthobutylicus normal butyric alcohol, butyric and acetic acid. 



