Xv. the Kinetics of Alcoholic Fermentation of Sugars

ARMSTRONG [1904] advanced the hypothesis that the specificity exhibited by yeast towards glucose, mannose and fructose was due to the existence of an enol form common to the three hexoses. Hopkins [1931] developed a modification of this hypothesis to show certain features of configuration common to ac-glucopyranose, ac-mannopyranose and fructofuranose using the configurations of Haworth. In the case of glucose, Willstiitter and Sobotka [1922] and Hopkins [1931] had shown that, when a solution ofglucose is fermented rapidlyby brewer's yeast and the fermentation is interrupted, the residual solution exhibits mutarotation in the positive sense. Thiswas interpreted tomean that the yeast selectively preferred the ac-form of glucose. Hopkins [1931] had further shown that fructose, partly fermented by brewer's yeast, also exhibited mutarotation in the positive sense. The inference was drawn that brewer's yeast specifically ferments a form of fructose possessing a less negative specific rotation than that of fructose at equilibrium; this was quite possibly fructofuranose which shows a certain configurational relationship with glucopyranose and mannopyranose. The simplest explanation of these phenomena is to suppose that brewer's yeast specifically ferments ac-glucose, ac-mannose (for which the evidence will be furnished in a later communication) and some form of fructose. However, there is the alternative explanation that, whilst both acand fl-forms of the sugars are directly fermentable, the ac-form is preferred just as glucose is preferred to fructose in mixtures of the two. Certain experiments of Willstatter and Sobotka [1922] seem to support this, in that ocand fl-glucose separately fermented at the same rate, just as glucose and fructose do. However, these fermentations were conducted at concentrations high enough to prevent concentration of sugar from being the limiting factor. Lastly there remains the possibility that intermediate forms of the sugars are specifically attacked. In recent years evidence has been forthcoming to strengthen this last hypothesis. Lippich [1932] measured the HCN-binding powers ("Blausaurezahl ") of sugar solutions and concluded that open-chain forms were present in appreciable amounts. One or, more probably, two such "N-forms" may exist. Thus for glucose he suggests either