The method adopted was the following: Casein purified by twice repeated precipitation was washed, dried, and ground. It was then placed in a large jar having an outlet at the bottom which was closed with a plug of cheese-cloth loose enough to permit a slow passage of water through it. The jar was filled with water acidified with acetic acid. When it had nearly all drained off the jar was again filled. The casein was frequently stirred to prevent its forming a compact mass. This washing was continued during seven or eight days, the last twenty-four hours washing being with distilled water. The product thus obtained was dried and ground. It was very poor in ash, 10-gram samples yielding but a trace of calcium. By this treatment practically all the water soluble constituents were dialyzed out of the granules. With casein prepared in this way combined with dextrin, butter-fat, and salts no appreciable growth was obtained even during the first month.

Punk and McCallum 47 tested casein prepared by the method just given and compared it with that prepared by their own method, with identical results. They suggest that heated casein may possibly have its value reduced through loss of C, but this is easily made good by addition of orange juice to the diet.

Osborne and Mendel at one time 48 believed that casein prepared by ordinary methods was entirely satisfactory, without further purification, for use in experiments with vitamin A. Subsequent experience however has led them to modify their views somewhat. In a later publication49 they discuss the allegation that the ability of animals to grow for some time in the supposed absence of fat-soluble vitamins is due to "exceptional vitality" of the individuals or to reserve stores of the vitamin in the body. Unless one accepts some indefinite explanation of this sort, it seems necessary, according to Osborne and Mendel, to conclude in the light of their experience, that removal of the fat-soluble vitamin from even purified proteins and carbohydrates is accomplished with far greater difficulty than has been hitherto suspected. An entirely convincing crucial experiment, in which nutritive failure immediately ensues upon the administration of diets fully adequate in every respect except for the presence of fat-soluble vitamin, remains to be made. It is significant that older rats thrive for a longer time than do the younger ones on the same diets nearly if not entirely free from the fat-soluble vitamin.

This is in contrast with the well-established fact that at all periods the lack of water-soluble vitamin is speedily manifested. They found that rats were capable of appreciable growth for some time on a diet of casein or edestin, commercial corn starch and dried brewery-yeast (all of which had been purified by boiling three times with absolute alcohol under a reflux condenser for one hour), a salt mixture, and lard (either commercial lard, alcohol-extracted lard, or lard rendered at comparative low temperatures in their own laboratory).

43 Drummond and Coward, Bioch. J. 14, 661, 1020.

44 Funk and Macallum, Zeitschr. f. physiol. Chem. 02, 17; J. Biol. Chem. 23, 413, 1016.

45 McCollum and Davis, J. Biol. Chem. 23, 233, 1015. 46 Funk and Macallum, Zeitschr. f. physiol. Chem. 02, 17. 47 Funk and Macallum, J. Biol. Chem. 27, 60, 1017. 48 Osborne and Mendel, J. Biol. Chem. 34, 156, 1018.

Drummond and Coward50 state that commercial casein contains relatively large amounts of A and that animals may obtain sufficient of this vitamin for prolonged growth from impure casein incorporated in the diet. They recommend the use of casein which has been heated for 24 hours or more to a temperature of 102° C, in shallow dishes, and then subjected to prolonged and continuous extraction with alcohol and ether.

In all tests on vitamin A Drummond and Coward5l recommend the use of a hydrogenated vegetable oil instead of a natural fat, since even those vegetable oils which are supposed to be devoid of A are not entirely so. Hydrogenated cottonseed oil they regard as absolutely vitamin-free, however, so far as could be ascertained by numerous tests.

49 Osborne and Mendel, J. Biol. Chem. 45, 277, 1921. 50 Drummond and Coward, Bioch. J. 14, 661, 1020. 51 Drummond and Coward, 1. c.