The apparatus employed for this purpose consists of a brass tube provided with a flange in which the gelatinised candle is held by a compressed india-rubber ring, and is shown in section in Fig. 5. Two such apparatus are used, each capable of holding about 70 c.c. of the liquid to be filtered. The tubes, after being filled with the yeast-juice, are connected by means of a screw joint with a cylinder of compressed air and the filtration carried out under a pressure of 50 atmospheres, [p060] the arrangement employed being shown in Fig. 6. In the earlier experiments 25 to 50 c.c. of yeast-juice were placed in each tube and the filtration continued until no more liquid passed through. The residue was then washed several times /in situ/ by adding water and forcing it through the candle. The time occupied in this process varied from six to twelve hours with different preparations of yeast-juice. The candle was then removed from the brass casing and the thick, brown-coloured residue scraped off, dissolved in water, and at once examined. It was subsequently found to be possible to dry this residue /in vacuo/ over sulphuric acid without seriously altering the fermenting power, and this led to a slight modification of the method, which is now conducted as follows. Two quantities of 50 c.c. each of yeast-juice are filtered, without washing, and the residues spread on watch-glasses and dried /in vacuo/. Two fresh quantities of 50 c.c. are then filtered through the same candles and the residues also dried. The 200 c.c. of juice treated in this way give a dry residue of 17 to 24 grams. The residue is then dissolved in 100 c.c. of water and filtered in quantities of 50 c.c. through two fresh gelatinised candles and the residue again dried. A considerable diminution in weight occurs, partly owing to incomplete removal from the candle and brass casing, and the final residue only amounts to about 9 to 12 grams. Occasionally it is necessary to repeat the processes of dissolving in water, filtering, and drying, but a considerable loss both of material and fermenting power attends each such operation. [Illustration FIG. 5. (Filter Apparatus, Tube/Flange Detail--Ed.)] The sticky residue dries up very rapidly /in vacuo/ to a brittle, scaly mass, which is converted by grinding into a light yellow powder. The filtrate was invariably found to be quite devoid of fermenting power, none of the enzyme passing through the gelatin. [Illustration: FIG. 6. (Filter Apparatus--Ed.)]