Although it is somewhat doubtful whether the enzymes contained in dog dung are of glandular origin,[100] it is quite certain that other enzymes are secreted by bacteria developing in the dung while it is kept prior to being used for puering. These enzymes may be separated by the following method. About 150 c.c. of puer is well mixed with an equal quantity of glycerin, and allowed to stand for seven days. It is then filtered through paper by means of a pump, and yields a clear filtrate of a deep golden-brown colour; the filtrate is poured in a thin stream into a tall vessel containing about 1500 c.c. of 98 per cent. alcohol. A copious flocculent precipitate of the albuminous matter and enzymes is thrown down, the solution is filtered and the precipitate washed on the filter with absolute alcohol, and then dried over sulphuric acid in vacuo. The resulting powder is, of course, a mixture of all the albumins in solution, and probably only a small portion of it consists of the pure enzymes. We have merely succeeded by this method in concentrating them. The property of albuminous bodies in the act of coagulation to carry down soluble matter is well known, and this also renders the preparation of any pure proteid extremely difficult. It may be mentioned here that recent evidence goes to show that enzymes are not of a proteid nature, since, by repeated purification, the proteid matter may be almost entirely got rid of, while the activity of the residue containing the enzyme becomes considerably greater. [100] See, however, the recent paper by Eberle and Krall, Ueber den Nachweis des Trypsins im Hundekot, Collegium 1911, p. 201, in which the authors endeavour to show the presence of unchanged trypsin in dog dung. Their proof depends upon the action of an antipancreatic serum on infusions of dog dung based on the work of Achalme, Ann. de l’Inst. Pasteur, 1901, p. 737. See also the criticism on this paper by Dr. Otto Röhm and Dr. Max Goldman, Collegium, 1911, p. 265. Hammarsten (Physiol. Chemistry, 1911, p. 494) states that “among the secretions which undergo putrefaction in the intestine, the pancreatic juice, which putrefies most readily, takes place first.” Krawkov’s method of preparing diastase from saliva (Green, p. 46) may also be mentioned, as it is of general application. It consists in salting out the enzymes, by saturating the clear solution (in this case saliva diluted with an equal volume of water) with neutral ammonium sulphate. The precipitate which is caused by the saturation is collected on a filter, and washed for a short time with strong alcohol. It is then allowed to stand under absolute alcohol for one or two days, and finally dried at 30° C. On extraction with water it yields a solution which is strongly diastatic, and which gives no proteid reactions.