On the other hand, in a later set of experiments 24 in which the ration used was purified protein, 18 per cent, "artificial protein-flree milk,"25 29.5 per cent, starch, 16.5 to 26 per cent, butter-fat, 18 per cent, lard, 7 to 16 per cent, and yeast, 1.5 to 2 per cent, while the rats grew normally when the protein was casein they usually failed when certain vegetable proteins were given, and almost invariably when lactalbumin was fed. It would seem, therefore, that the natural " protein-free milk " used in the earlier experiments provided something necessary to supplement the lactalbumin which was lacking in the later experiments. Osborne and Mendel say: " Whatever the nutritive value of the unknown constituents of milk may be, our experience manifestly leads to the conclusion that the supplementary value of " protein-free milk " in the diet is, as a general rule, decidedly greater than that of yeast".

McCollum, Simmonds, and Parsons 26 obtained somewhat similar results when using a basal ration of 45 per cent dried peas, 9 per cent purified protein, 38.2 per cent dextrin, 5.0 per cent butter-fat, and 1.8 per cent salts. Much better growth was secured with casein than with lactalbumin. They stated therefore: "We are forced to the conclusion that lactalbumin is a poorly constituted or an incomplete protein and that the excellent results of Osborne and Mendel were due to the high proportion of nitrogen derived from the 'protein-free milk' which was present in their food mixtures and served to supplement the lactalbumin with respect to some as yet undetermined cleavage product which is essential for growth." Emmctt and Luros27 found indications of a peculiar relation between lactose and lactalbumin, since a diet made up of lactalbumin (10 per cent), butter-fat, lard, starch, salt mixture, and purified lactose promoted excellent growth, whereas on a diet similar in all respects except that the lactose was replaced by additional starch and alcoholic extract of either wheat germ or yeast, growth ceased.

They suggest that lactose, when added to a lactalbumin diet, may either overcome some toxicity of the lactalbumin molecule or carry a vitamin other than that found in wheat germ or yeast.

23 Osborne and Mendel, J. Biol. Chem. 20, 351, 1915; 26, 1, 1916; 37, 223, 1919.

24 Osborne and Mendel, J. Biol. Chem. 31, 155, 1918.

25 See p. 62.

26 McCollum, Simmonds, and Parsons, J. Biol. Chem. 37, 287, 1919.

Sure 28 obtained evidence that lactalbumin is an incomplete protein which must be supplemented by the addition of nitrogen compounds which are present in the "protein-free milk " before it is adequate for either growth or maintenance. When lactalbumin was fed as 12 and 18 per cent of a ration of which the other components were dextrin, butter-fat, agar-agar, salts and extract of wheat embryo, no growth was obtained unless the lactalbumin was supplemented with one per cent of its weight of the amino acid cystine. With this supplement 12 per cent of lactalbumin gave excellent growth, but if it was reduced to 9 per cent of the diet (the amount represented as adequate by Osborne and Mendel) nutritive failure ensued. If, however, the lactalbumin was still further supplemented by the addition of five per cent of its own weight of a second amino acid, tyrosine, the growth was excellent at this low level of lactalbumin intake. Sure believes that the efficiency of a protein may largely depend on its constitution as well as composition in amino acids, and that chemical analysis of amino acid content may at times be inadequate to explain the nutritive failure or success obtained with certain proteins.

Since lactalbumin is reported as having the average percentage of cystine present in other proteins, he suggests that cystine is so oriented in the complex polypeptide chain of lactalbumin that when it is hydrolysed by the enzymes in the digestive tract it is split into simpler peptides of such form that a great part of them escape further cleavage, and are therefore de-aminized and converted into urea. In view of the results of Osborne and Mendel above referred to, he is of the opinion that protein-free milk supplies the necessary tyrosine (a hypothesis which is confirmed by qualitative tests on "protein-free milk ") and also either carries cystine as part of its nitrogen of unknown source, or contains other forms of organically bound sulphur which the animal readily transforms into cystine. An analysis of protein-free milk showed a sulphur content of 02 per cent, of which the greater part was in organic combination.

27 Emmett and Lurot, J. Bid. Chem. 38, 147, 1919. 28 Sure, J. Biol. Chem. 43, 457,1920.

Logical as this hypothesis seems, however, it offers no explanation of the results of Emmett and Luros in which the efficiency of the lac-talbumin was apparently dependent upon the presence of lactose.

According to Osborne and Mendel,29 lactalbumin is much more efficacious than casein as a supplement to corn gluten. When as little as one-quarter of the protein in the ration is furnished by lactalbumin and the remainder by corn gluten, a rat is able to make perfectly normal growth, while a similar ration in which as much as one-half of the protein was casein and the remainder corn gluten proved inadequate.

Inasmuch as mother's milk serves as the sole food of young and rapidly growing animals it is generally assumed that it must contain all essential inorganic elements for the optimum growth of that species. Lane-Claypon30 quotes experiments carried on by Hittcher31 which would appear to indicate that the nutritive value of cow's milk for calves can be increased by the addition of certain salts. Boiled milk with suitable salt additions was apparently more nutritious than raw milk. Commenting upon this, Lane-Claypon remarks: "If the addition of salts to mother's milk can bring about such striking differences, then that difference in salt content, which is known to exist between the milks of different species of animals may reasonably be expected to play an important part in the nutritive value of the milk of a foreign species".

29 Osborne and Mendel, J. Biol. Chem. 18, 1, 1914; 20, 69, 1917.

30 Lane-Claypon, 1. c, p. 166.

31 Kalberfutterungsversuche, Landwirtsch. Jahrb. 1909.