Precisely similar results were subsequently obtained by Buchner and Antoni [1905, 2] by the dialysis of yeast-juice. One hundred c.c. of juice were dialysed for twenty-four hours at 0° against 1300 c.c. of distilled water, and the dialysate was then evaporated at 40° to 50° to 20 c.c. The fermenting power of 20 c.c. of the dialysed juice was then determined with the following additions:-- (1) 20 c.c. of dialysed juice + 10 c.c. of water gave 0·02 gram CO{2}. (2) 20 c.c. of dialysed juice + 10 c.c. of evaporated dialysate gave 0·52 gram CO{2}. (3) 20 c.c. of dialysed juice + 10 c.c. of boiled juice gave 0·89 gram CO{2}. It was shown in the previous chapter that phosphates are essential to fermentation, and hence it becomes necessary to inquire whether the effect of dialysis is simply to remove these. Experiment shows that this is not the case. Soluble phosphates do not confer the power of producing fermentation on the inactive residue obtained by filtration. Moreover, when yeast-juice is digested for some time before being boiled, it is found, as will be subsequently described, that the boiled autolysed juice is quite incapable of setting up fermentation in the inactive residue, although free phosphates are abundantly present [Harden and Young, 1906, 2]. The filtration residue is never obtained quite free from combined phosphorus, but the production from this of the phosphate necessary for fermentation to proceed, may be so slow as to render the test for co-enzyme uncertain, owing to the absence of sufficient phosphate. When a filtration residue is being tested it is therefore necessary to secure the presence of sufficient phosphate to enable the characteristic reaction to proceed, and at the same time to avoid adding phosphate in too great concentration, as this may, in the presence of only small amounts of enzyme or co-enzyme, inhibit the fermentation (p. 71). The proof that a filtration residue or dialysed juice is quite free from co-enzyme is therefore a somewhat complicated matter, and not only involves the experimental demonstration that the material will not ferment sugar, but also that this power is not imparted to it by the addition of a small concentration of phosphate. As it has been found (p. 73) that the fermentation of fructose is less affected than that of glucose by the presence of excess of phosphate, the practical method of examining a filtration residue for co-enzyme is to test its action on a solution of fructose (1) alone and (2) in presence of a small concentration of phosphate. If the residue produces no action [p063] in either case, but produces fermentation when a solution of co-enzyme is added in the presence of the same concentration of phosphate as was previously employed, it may be concluded that this sample was free from co-enzyme but contained enzyme; such an experiment also affords a definite proof that the co-enzyme does not consist of phosphate.