456 SCIENCE PROGRESS 



touches the hydroxyl group, we shall find that a perfect ferment 

 for glucose would be written thus : 



(1) X- 



(2) X- 



(4) X- 



•X(3) 



Here four "wards" come into play. In the case of mannose, 

 only three of these (i, 2, and 3) are able to act; in galactose, i, 

 3, and 4 would be able to attack the hydroxyl groups ; while in 

 talose only two (i and 3) could come into action. This analogy 

 of lock and key must not be pushed too far, but there certainly 

 appears to be some relation between configuration and the ease 

 of fermentation in certain compounds. 



Results of a somewhat similar nature were obtained by 

 Bertrand^ in his studies of the action of the sorbose bacterium 

 {Bacterium xylimmi) on certain polyhydric alcohols. He found 

 that the following alcohols were attacked by the organism : 



OH HO OH 



I I I 



CH^OH-C-CH.OH CHoOH-C-C-CHoOH 



H H H 



Glycerol Er>'thritoI 



OH OH H OH OH OH OH 



III I I I I 



CH.OH - C C C - CH,OH CH.OH - C C C C - CH,OH 



III' I I I I ■ 



H H OH H H H H 



Arabitol Sorbitol 



In each ease the heavily printed — CHOH— group is oxidised 



to —CO — . On the other hand, compounds such as those below 



are not attacked in this w^ay : 



Glycol H OH H 



CH,OH-CHoOH I I I 



CH.,OH - C C C - CH.OH 



H OH OH H 'III 



I I I I OH H OH 



CH, OH - C C C C C - CH,OH Xylitol 



I I I I 



OH H H OH Dulcitol 



^ Bertrand, Bull. Soc. ckim.^ [3], 19, 347, 947, 999 (1898) j Covtpt. rend., 122, 900 

 (1896); 126, 762(1898). 



