New YorK AGRICULTURAL EXPERIMENT STATION. 5 
lactose with evident gas formation, -but in all of the later 
studies it failed to produce gas even after having been re- 
vivified. Although this germ apparently lost its ability to 
produce gas after the beginning of this study it should not 
be concluded that this was the rule, as it was really an ex- 
ception. Owing to the stimulating effects of revivification 
which was practiced only, during the latter part of the study 
a very considerable number of germs formed gas during the 
later tests when they had not done so during the earlier ones. 
The fermentation tube has been largely used in the past to 
differentiate closely related species. It might be concluded 
in the light of the contradictory resuits which have been ob- 
tained in this study that the fermentation test is really of 
little value for this purpose. - Such a general conclusion does 
not logically follow from this data because the particular 
group of organisms which have been studied here chance to 
lie on the very border line of visible gas formation and ac- 
cordingly are not typical of germ life in general. 
The fact that this group does lie on the border line of visible 
gas formation is shown by the fact that the amount of gas is 
small in most cases, in some cases only a small bubble appear- 
ing in the fermentation tube, and also by the failure of even 
this small amount of gas to appear in a considerablé propor- 
tion of the tests with some of the strains. 
The crudeness of the fermentation tube as a measure of the 
total gas formation is well brought out by Keyes.** He found 
in his study of B. coli that the total gas produced under com- 
parable conditions in the fermentation tube and in a vacuum 
were in the ratio of 12.4 to 99.9. 
In considering the importance which should be attached to 
the results obtained with the fermentation tube it should be 
remembered that all forms of living protoplasm respire gases 
and with the bacteria the carbon dioxide of respiration is an 
easily measurable quantity even with those forms which show 
no evidence of gas formation in the fermentation tube. The 
*Keyes, F.G. The gas production of Bacillus coli. Jour. Med. Research, 
N. S., 16: 69-82. 1909. . 
