PHYSIOLOGY OF THE MICROBES 79 



There exist also, it is true, "fermentations by force of 

 example," where a fermentation started on a certain sugar may 

 extend its attack to another sugar which at first was not 

 fermentable : galactose, for example, can be fermented when 

 it has been " baited " by glucose. Similar affinities, capable 

 of modification by custom, come into play without doubt in the 

 sensitiveness or resistance, natural and acquired, of an animal 

 body towards a pathogenic ferment. Beneath the biological 

 specificity there lies a chemical specificity. 



One of the best examples that can be quoted is that of the 

 bacterium of sorbose studied by G. Bertrand. 



This makes a selection among the polyatomic alcohols, 

 attacking only in their molecule a link of the formula CH'OH, 

 transforming it into CO and consequently producing always a 

 ketone body. Further, for this link to be attacked its hydroxyl 

 OH must not be on the side of the H atom of the neighbour- 

 ing link CHOH. Finally the secondary group attacked is 

 always next to one of the primary group CH 2 OH, which 

 terminates the chain, at least in the formulae below C 7 . The 

 stereochemical structure of the sugars thus plays an important 

 part in the matter : it is on it and on it alone that the possibility 

 of attack by the bacterium depends. Its action is so narrowly 

 specific that it only transforms certain chemical groups, taking 

 no interest in the rest of the molecule, whatever may be its 

 mass and structure. A fermentation of this kind has all the 

 value of a chemical reaction. 



Food-stuffs, culture media, infected organisms, represent 

 for the bacteria their soil, " If we consider," says Bertrand, 

 "on the one hand the differences in chemical composition, 

 even qualitative, which may exist between two closely related 

 species, and on the other hand the extraordinary variety of 

 proteins which it is possible to conceive of nowadays, it will 

 hardly appear unreasonable to compare animal species, or 

 physiological variations of the same species, to culture 

 media varied like those which I used in the study of the 

 sorbose bacterium, nor to account for their immunity or 

 susceptibility towards a given microbe by a chemical or even 



