Löb has adduced many striking facts in favour of this view, and has shown that very dilute alkalis produce no lactic acid but formaldehyde and a pentose as primary products. These substances represent the first stage of depolymerisation and are also formed by the electrolysis of glucose. Löb has himself been unable to detect definite intermediate products of fermentation by adding reagents, such as aniline, ammonia, and phloroglucinol, which would combine with such substances and prevent their further decomposition [1906]. The occurrence of traces of formaldehyde as a product of alcoholic fermentation by yeast-juice [Lebedeff, 1908] is at least consistent with this theory, but no decisive evidence has so far been obtained either for or against it. In all the foregoing attempts to indicate the probable stages in the production of alcohol and carbon dioxide from sugar, a single molecule of the sugar forms the starting-point. The facts recounted in Chapter III as to the function of phosphates in alcoholic fermentation, which are summed up in the equation:-- 2C{6}H{12}O{6} + 2R{2}HPO{4} = 2CO{2} + 2C{2}H{6}O + 2H{2}O + C{6}H{10}O{4}(PO{4}R{2}){2}, render it in the highest degree probable that two molecules of the sugar are concerned. The most reasonable interpretation of this equation appears to be that in the presence of phosphate and of the complicated machinery of enzyme and co-enzyme two molecules of the hexose, or possibly of the enolic form, are each decomposed primarily into two groups. Of the four groups thus produced, two go to form alcohol and carbon dioxide and the other two are synthesised to a new chain of six carbon atoms, which forms the carbohydrate residue of the hexosephosphate. The introduction of the phosphoric acid groups may possibly occur before the rupture of the original molecules, and may even be the determining factor of this rupture, or again this introduction may take place during or after the formation of the new carbon [p118] chain. Sufficient information is not yet available for the exact formulation of a scheme for this reaction. Such a scheme, it may be noted, would not necessarily be inconsistent with the views of Wohl and of Buchner as to the way in which the carbon chain of a hexose is broken in the process of fermentation, but would interpret differently the subsequent changes which are undergone by the simpler groups which are the result of this rupture. The reaction might thus proceed without the formation of definite intermediate products, whilst opportunity would be afforded for the production of a small quantity of by-products such as formaldehyde, glycerol, lactic acid, acetic acid, etc., by secondary reactions. A symmetrical scheme can readily be constructed for such a change, but much further information is required before any decisive conclusion can be drawn as to the precise course of the reaction which actually occurs in alcoholic fermentation. [p119] CHAPTER IX.