McCollum56 doubted whether the significant nitrogen content of milk could be so easily removed as would appear, and was inclined to attribute the success attained with protein-free milk to the supplementary value of the nitrogen still present.

Mitchell and Nelson57 endeavored to reduce the amount of residual nitrogen by the following procedure: Where the original method involves precipitation of the casein with slight excess of hydrochloric acid, filtration, heating the filtrate to boiling for one-half minute, filtering off the precipitate of lactalbumin, neutralizing the clear filtrate, and evaporating to dryness at a temperature not exceeding 70o C. upon neutralization of the filtrate after separation of the lactalbumin, a precipitate is always obtained, probably consisting largely of calcium phosphate. This was filtered off and the filtrate evaporated to dryness as usual. In this way preparations obtained contained on an average about 0.10 per cent less nitrogen than the preparations made after the original method of Osborne and Mendel. Presumably this reduction is largely if not entirely in the protein nitrogen.

These observers have found the preparation apparently as effective for periods of five or six months at least in covering the nutritive requirements of white mice for simple maintenance as rations containing the same amount of unfiltered production. They suggest as a better method of preparation of protein-free milk, precipitation by trichloracetic acid. The optimum proportions were found to be 12 cc. of 50 per cent trichloracetic acid per 10 g. of milk powder, the filtrate being boiled for thirty minutes to an hour and filtered again. Two separate preparations were made according to this method, and found to contain 0.401 and 0.42% N respectively, equivalent to about one twenty-third of the total nitrogen of the milk. Tests failed to indicate with any certainty the presence of protein material. Although the tests have not been continued for a sufficiently long time to be entirely satisfactory, the product thus prepared seems to be as efficient for maintenance at least, as the preparations of Osborne and Mendel. In all cases Mitchell and Nelson used the dried centrif-ugalized milk put on the market by the Merrill-Soule Co., of Syracuse. This was dissolved in about twenty times its weight of distilled water, and treated accordingly to one or other of the above methods.

56 McCollum, Am. J. Physiol. 25, 120; McCollum and Davis, J. Biol. Ghem. 19, 260; 20, 647; McCollum and Simmonds, Am. J. Physiol. 46, 275, 1918.

57 Mitchell and Nelson, J. Bi. Ch. 23, 459, 1915.

Kennedy53 states that "protein-free milk" as prepared by the method of Osborne and Mendel contains either unprecipitated protein or peptides of considerable size as shown by amino nitrogen determinations before and after acid hydrolysis and the nitrogen distribution as well as by the increase in amino nitrogen after tryptic digestion.

Later, when it was recognized that the "protein-free milk" carried vitamins as well as inorganic salts, Osborne and Mendel evolved the following mixture of pure salts.69

In the salt mixture used as reported in the Journal of Biological Chemistry 37, 317, 1013, the proportion of sodium fluoride ia given as 0248. The other constituents are as given above.

The chemicals used were analyzed and allowance was made for moisture, etc. The acids were mixed and the carbonates and ferric citrate added to them. The traces of potassium iodide, manganese sulphate, sodium fluoride, and potassium aluminum sulphate were added as solutions of known concentration. The final resulting mixture was evaporated to dryness in a current of air at 90-100° C, and ground to a fine powder. This mixture with the addition of purified lactose was intended to reproduce as exactly as possible the composition of the "protein-free milk" with the exception of the traces of nitrogenous compounds and the vitamins.

58 Kennedy, J. Am. Ch. Soc. 41, 388, 1919.

59 Osborne and Mendel, J. Biol. Chem. 25, 317, 1913.

McCollum and co-workers have experimented with a variety of salt mixtures, the following being noted: 60

Salt Mixture I

Salt Mixture II

2.4 gm. of the above with 1.3 gm. of calcium lactate was used with each 100 gm. of ration.61 Iodine was supplied in the drinking water once each week.62

Mixture I has been used satisfactorily by Stccnbock 64 who states however that it is more complex than necessary, as he has secured excellent growth, reproduction, and the rearing of young on rations free from sulphates. Mitchell and Nelson65 using mice as experimental animals were less successful. They say:

Our experiments with this salt mixture in experiments on mice have not been particularly encouraging. Rations containing all their mineral constituents in this artificial preparation have invariably produced unmistakable symptoms of malnutrition in the experimental animals, generally after two or three weeks time, though occasionally these symptoms have been deferred for a month or even two months. ... It is possible that it rests simply on the fact that different species of animals have been used. Another possible explanation is that the discrepancy is due to the use by McCollum and Davis of wood shavings in their experimental cages and of paper excelsior by ourselves. It is possible that in the former case the wood shavings may have furnished some indispensable mineral constituents absent from the artificial salt mixture which constituted a part of the rations. We cannot agree with the statement of McCollum and Davis66 that they do not look upon the consumption of a small amount of wood fibre as objectionable to any greater degree in this type of experiment than is the feeding of agar agar.

Also the availability and nutritive value of the nitrogenous and mineral substances of the wood cannot be as lightly disregarded, we believe, as they have been by these investigators.67

60 McCollum and Davis, J. Biol. Chem. 15, 167; 10, 250; McCollum and Kennedy, Ibid. 34, 494; McCollum and Simmonds, Ibid. 3a, 355; McCollum, quoted by McArthur and Luckett, Ibid, 20, 163.

61 McCollum, quoted by McArthur and Luckett, J. Biol. Chem. 20, 16B, 1015.

62 McCollum and Davis, J. Biol. Chem. 20, 646, 1015.

63 Steenbock, Kent, and Gross, J. Biol. Chem. 35, 67, 1018; Steenbock, Boutweli, and Kent, Ibid. 521; See also Hogan, Ibid. 27, 194, 1916.

64Mitchell and Nelson, J. Biol. Chem. 23, 459, 1915.