FERMENTATION 93 periments for supplying these data, which, in fact, could not reasonably have been attempted by anybody before him, as it is he to whom we owe our knowledge of the qualitative elementary composition of the substances con cerned, and indeed of organic substances generally. Before giving his numbers, it may be stated that he re garded acetic acid (a small quantity of which is present in most wines) as, like alcohol and carbonic acid, a constant product of vinous fermentation. According to Lavoisier, 95 9 parts of cane sugar in fermenting yield Alcohol 57 7 Carbonic acid 35 3 Acetic acid . , 2 5 95 5 And according to his elementary analysis of these sub stances, the proportions by weight are Carbon. Hydrogen. Oxygen. In 35-3 of carbonic acid .... 9 9 nil 25 4 In 577 of alcohol 167 9 6 31 4 In 2-5 of acetic acid .. , 6 2 17 9-8 77 58-5 61-4 27-2 In 95 9 of sugar 26 8 From these numbers Lavoisier concluded (and he was quite justified in doing so, considering the imperfections of his methods of analysis) that sugar in fermenting simply breaks up into these three substances, without any access of matter from without. But if he thus managed to arrive at what we now know to be a substantially correct conclusion, this can be credited to him (if at all) only as a happy stroke of divinatory genius, as his numbers are all of them mon strously incorrect, the errors going far beyond what even, with his necessarily imperfect method, could be tolerated as " observational errors." Lavoisier s numbers were sub sequently corrected by Gay-Lussac according to his own analyses of sugar, alcohol, and carbonic acid. His results, which have remained unimpugned to the present day, may be stated, with substantial correctness, to have been as follows : In vinous fermentation very nearly one -third of the carbon goes off as carbonic acid, while the rest passes into the al cohol; and reducing to 1, 2, and 3 parts of carbon, we have Carbon. Hydrogen. Oxygen. Sum. In carbonic acid 1 + nil + 2 667 - 3 667 In alcohol 2 + 5 + 1 333 - 3 833 Sums 3 Found in cane sugar .. 3 0-5 0-458 4-0 3-667 7-125 The agreement being by no means satisfactory, Gay- Lussac suspected that his analyses of sugar were infected with unobserved errors, and he corrected his figures so as to make them agree with those given above opposite to Sums." These values, when measured by the customary units (namely C for twelve parts of carbon, II for one part of hydrogen, O for sixteen parts of oxygen), assign to sugar the very simple formula CjHoOj leading to an equally simple equation for the reaction, which is : 6x(CH 2 0) = C 6 H 12 6 = 2C a H 6 + 2C0 2 ; i.e., 180 of sugar gives 2 x 46 of alcohol + 2 x 44 of carbonic acid ; or 45 ,, ,, 23 ,, + 22 ,, ,, This equation is still looked upon as substantially correct, though not in Gay-Lussac s sense. It is so, if by sugar we understand either of the two kinds of "glucose" which form the bulk of the sweetening principles in fruit juices, and which are composed according to the formula C H 12 O . Cane sugar, as Dumas and Boullay showed, really has the com position following from Gay-Lussac s analysis, which, as is easily seen, corresponds to the formula C 12 H 22 O n = 2C fi H 19 O 6 - H 2 O, where H 2 O means the elements of 18 parts of water ; and these 18 parts of water, as Dumas and Boullay showed, actually are taken up in the fermentation of C 12 H. )0 O U = 324 parts of cane sugar. Gay-Lussac s equation being, as we said, only substan tially correct, we must now state the qualifications implied. Schmidt of Dorpat found in 1847 that vinous fermentation always results in the formation of small quantities of suc- cinic acid. Guerin Vary showed, by quantitative experi ments, that in the fermentation of glucose the alcohol and carbonic acid produced account only for about 96 3 per cent, of the glucose. And the present writer happens to know that a certain German apothecary made the interesting discovery that wines, beside the unfermented remnant of glucose that may be left, may contain an unfermentable sweet principle which he recognized as glycerine. These observations, however, were little heeded until Pasteur, in a now classical memoir, proved that glycerine and succinic acid are constant products of normal vinous fermentation, the correct balance sheet of the reaction, according to him, being as follows : 100 parts of cane sugar, in fermenting, pass into 105 4 parts of glucose, which then break up, yielding (approximately) of Alcohol 51-1 Carbonic acid 49 - 4 Succinic acid 07 Glycerine 3 - 2 Hatter passing to the yeast 1 -0 Sum 105-4 But even this is not quite an exhaustive statement, a small portion of the sugar always passing into the form of higher alcohols ("fusel-oil") and compound ethers. Vinous fermentation, then, is a far more complex reaction than Gay-Lussac imagined ; but it still remains true that all the products formed are derived from the dissociation of the sugar. What is it that brings about this singular de composition! We call it a singular reaction, because it is one which sugar has never been seen to undergo when sub jected by itself or as an aqueous solution to the action of heat or electricity or any ordinary reagent. And we have theoretical grounds for presuming that the reaction is not likely ever to be realized by some " reagent" that has not yet been tried. According to many experiences, an arithmetically possible reaction is the more likely to be realized the greater the heat evolution which, supposing it were realized, it would involve. Now, the reaction formu lated in Gay-Lussac s equation C 6 H 12 O 6 = 2CO 2 + 2C 2 H 6 O, as Professor De war pointed out some years ago, supposing dry sugar could be made thus to split up, would yield only an insignificant amount of heat, if any. Actual fermentation does involve a liberation of heat, as we know, but the quan tity of heat per unit weight of sugar destroyed, according to Dewar s experiments, amounts only to about 83 heat-units, which can be accounted for as being produced by the hydra- tion of the alcohol formed, and, at any rate, is too small to characterize the decomposition of sugar into carbonic acid and alcohol as being at all of itself a probable reaction. Even the somewhat higher result previously arrived at by Dubrunfaut, namely, 135 heat-units per unit of sugar, can not affect this conclusion. Before going further let iis take an exact survey, from the chemical standpoint, of the con ditions which are known to favour or impede the actual process. (1.) Pure solutions of cane sugar or glucose do not ferment under any circumstances. (2.) Many kinds of impure sugar solutions, such as grape juice, brewers wort, &c., do ferment. The range of temperatures most favourable to this process lies between about 20 and 24* C. (or 68 and 75 F.). But even grape juice does not ferment at temperatures lying too close to the freezing-point, nor does it ferment at tempera tures exceeding a certain limit, which lies at about 60 C. (140 F.). The most lively fermentation comes to a stop when the liquid is boiled, and, after cooling, it takes a longer or shorter time before it resumes. (3.) Grape juice which has been strengthened by evaporation or addition of sugar from without, does not ferment, when the ratio of water to sugar falls below a certain limit-value.
