QUALITATIVE CATABOLIC REACTIONS OF BACTERIA 77 
took place in two steps: (a) the removal of hydrogen from ethyl alcohol 
to form acetaldehyde, and (b) the subsequent oxidization of acetalde- 
hyde to acetic acid. Xeuberg and Windisch^ have studied this 
reaction more in detail and find it takes place as follows: 
C2H5OH C2H6OH C2H6OH 
\ / \ /- \ 
CH3.CHO2 CH3.CHO2 CH3.CHO etc. 
\ \ 
CH3.COOH CH3.COOH 
It will be seen that e^-entually most of the alcohol will be transformed 
through the acetaldehyde stage, to acetic acid. 
The unequivocal importance of aldehydes, demonstrated interme- 
diary products in about all the fermentation reactions thus far studied 
carefully is extremely illuminating, and suggestive of the character of 
fermentative processes in general. 
The sugars containing sLx carbon atoms appear to be somewhat 
more utilizable than their corresponding alcohols: thus, the Shiga 
bacillus (B. dysenteriae) cannot ferment mannitol; it can, however, 
readily ferment glucose. This would suggest that the aldehyde 
group— CHO— is somewhat more readily attacked than the alcohol 
group— CH2OH—, for mannitol has no aldehyde group and glucose 
has an aldehyde group. The alcohols in general appear to be less 
readily acted upon by bacteria than are the corresponding aldehydes 
or even organic acids, provided the latter are not too greatly dissociable. 
The products of fermentation of higher alcohols, as mannitol, by 
bacteria are somewhat different from those of the corresponding sugars 
(aldoses). The chief point of difference, according to our present 
knowledge, consists principally in the production of more alcohol 
when the higher alcohols are utilized than when the corresponding 
aldoses are concerned. This has been worked out satisfactorily for 
certain bacteria, notably the colon and the typhoid bacilli, by Harden,^ 
and by Frankland and Frew,^ and Frankland and Lumsden.^ It is not 
definitely known for many other organisms. The gas-forming bac- 
teria, as a rule, produce more gas and more alcohol from the alcohols 
of the Ce series than from their corresponding aldoses. This gas 
formation appears to result from the decomposition of formic acid 
by the activitv of a specific enzyme, formiase, according to the equa- 
tion HC00H'= CO2 + H2.« Thus, B. coli and related gas-forming 
bacteria, according to this theory, produce the ferment, /or ///7V/.s'e, while 
B. typhosus, which also produces formic acid from the decomposition 
of glucose does not possess this ferment and consequently forms no 
gas in sugar solutions. Formic acid is, therefore, somewhat promi- 
1 Biochem. Ztschr., 1925, 146, 454. ^ Cannizzaro reaction. 
•^ Jour. Hyg., 1904, 5, 488. 
•• Jour. Chem. Soc, 1892, 61, 254. - Ibid., p. 432. 
« Franzen and Steppuhn: Ztschr. f. physiol. Chem., 1912, 77, 129. 
