When 0·1 gram of dihydroxyacetone in 5 c.c. of water was brought in contact with 1 gram of living yeast, about half was fermented, 17 c.c. of carbon dioxide (at 20° and 600 mm.) being evolved in excess of the autofermentation of the yeast (13 c.c.). A much greater effect was obtained by the aid of yeast-juice, and the remarkable observation was made that whilst yeast-juice alone produced comparatively little action a mixture of yeast-juice and boiled yeast-juice was much more effective, quantities of 20 to 50 c.c. of yeast-juice mixed with an equal volume of boiled juice, which in some experiments was concentrated, yielding with 0·4, 1, and 2 grams of dihydroxyacetone almost the theoretical amount of carbon dioxide and alcohol in excess of that evolved in the absence of this substance. It was further observed that the fermentation of this substance commenced much more slowly than that of glucose. No explanation of either of these facts has at present been offered. The conclusion drawn from their experiments by Buchner and Meisenheimer that dihydroxyacetone is readily fermentable, was confirmed by Lebedeff [1911, 1], who further made the important observation that during the fermentation of dihydroxyacetone the same hexosephosphoric acid is produced as is formed during the fermentation of the hexoses. Lebedeff accordingly propounded a scheme of alcoholic fermentation according to which the hexose [p106] was first converted into two molecules of triose, the latter being first esterified to triosephosphoric acid and then condensed to hexosediphosphoric acid, which then underwent fermentation, after being hydrolysed to phosphoric acid, and some unidentified substance, probably an unstable modification of a hexose, much more readily attacked by an appropriate enzyme than the original glucose or fructose [1911, 1, pp. 2941-2]. The idea that the sugar is first converted into triose and this into triosemonophosphoric acid had been previously suggested by Iwanoff who postulated the agency of a special enzyme termed /synthease/ [1909, 1], and supposed that this triosemonophosphoric acid was then directly fermented to alcohol, carbon dioxide and phosphoric acid. According both to Iwanoff and Lebedeff the phosphoric ester is an intermediate product and its decomposition provides this sole source of carbon dioxide and alcohol. This is quite inconsistent with the facts recounted above (Chap. III), which prove that the formation of the hexosephosphate is /accompanied/ by an amount of alcoholic fermentation exactly equivalent to the quantity of hexosephosphate produced, and that the rate of fermentation rapidly falls as soon as the free phosphate has disappeared, in spite of the fact that at that moment the concentration of the hexosephosphate is at its highest, whereas according to Iwanoff's theory it is precisely under these conditions that the maximum rate of fermentation should be maintained.