The oat kernel, according to McCollum's investigations, contains protein of poorer quality than either the maize or wheat kernel. When all other dietary factors are properly adjusted, nine per cent of oat protein in the diet serves to induce slow growth for a time, but never for more than about a month (experiments with rats). Casein, which serves as such an efficient adjunct to the wheat and maize proteins, does not seem to supplement oat protein in a very satisfactory manner; a diet with 9 per cent of protein from the oat kernel and 10 per cent purified casein did not induce growth at a maximum rate as did similar combinations of casein with wheat and maize proteins. In this connection McCollum reports the unexpected finding that gelatin supplements the protein of the oat kernel more effectively than does casein.

The ash constituents of the oat kernel must always be supplemented in order to induce growth. Fat soluble A is present in the oat kernel in very small amounts. The amount of water soluble B is adequate. Growth at more than half the normal rate may be obtained when the oat diet is supplemented by the addition of a suitable salt mixture and either butter fat or a suitable protein. When all three of these supplements are added, growth is normal but somewhat slow. McCollum believes that excessive feeding of the oat kernel causes some injury to the animal.

The white bean, when fed as the chief component of the diet, gave results indicating that its proteins are of lower nutritive efficiency than those of the cereal grains. The bean protein can be supplemented by the addition of 9 per cent of casein to the diet. The inorganic content of the white bean is not such as to induce growth, but must be supplemented by a suitable salt mixture (or by food of a different ash content from that of the bean alone). The white bean seems to contain less of fat soluble A than do the cereal grains. It contains water soluble B in abundance. The bean diet appeared to exert an unfavorable effect in that animals fed on a diet containing a smaller proportion of beans (25 per cent of the total food) seemed better nourished than those whose diet contained a larger proportion. It is suggested that beans may contain some unknown substance which is harmful when taken in too large an amount; or that the pressure of the intestinal gases resulting from fermentation of the hemicelluloses for which the higher animals have no digestive enzyme may result in a somewhat asphyxial condition of the intestinal wall, thus interfering with the normal processes of absorption and unfavorably affecting the general condition of nutrition.

Seeds in general are held by McCollum to require supplementing in order to make a diet which will support normal growth and reproduction. As supplement to a diet consisting largely of the products of cereal grains or other seeds, milk is found to be especially effective. It is also found that while seeds are not effectively supplemented by other seeds, they may be supplemented by the leaves and probably also by the roots and tubers of plants so that it is feasible, if desired, to draw a balanced diet, adequate for all the requirements of growth and reproduction in an omnivorous animal, entirely from the products of plants. Thus McCollum kept rats through four generations upon a carefully adjusted ration of maize, alfalfa, and cooked peas. Growth and reproduction were normal. The mothers successfully suckled young up to the normal age of weaning, after which they took the same food mixture as the adults. In this connection it is interesting to note that rats which were free to make their own selection from a much greater variety of vegetable foods never grew beyond half the normal adult size.

In practice milk is found to be most highly efficient as a supplement to diets consisting largely of seeds or their products: "The dietary should be built around bread and milk." The chemical constitution of its proteins and its high calcium and vitamine contents are all factors in the unique nutritive efficiency of milk, and make it possible for a moderate addition of milk to render adequate a diet otherwise composed entirely of seeds.

Cotton-seed meal or flour * constitutes an abundant and concentrated source of protein and energy which as yet has been but little utilized in human nutrition. This is doubtless largely because bad results have sometimes followed its use in stock feeding, leading to the general belief that it is somewhat toxic, at least when used in considerable quantities. Withers and Carruth succeeded in extracting from the kernels of the cottonseed a substance, gossypol, which shows deleterious action when fed and to which the toxicity of raw cotton seed and of some cotton-seed meals was attributed. This substance, however, is thermo-labile, and apparently is more or less completely destroyed by the heating to which cotton-seed meal or flour is ordinarily subjected in connection with the processes of crushing and pressing. Feeding experiments to determine whether the well-prepared cotton-seed meal or flour now available for human food has any appreciable toxicity, and to what extent it meets the nutritive requirements of normal growth and reproduction, have recently been reported by Richardson and Green and by Osborne and Mendel. Richardson and Green, feeding a high-grade commercial cotton-seed flour, found that no evidence of toxicity appeared although this flour constituted 45 to 50 per cent of the ration of albino rats through four successive generations or during 565 days of the life of an individual (about two thirds the entire normal life span); that the cotton-seed flour met all protein requirements of maintenance and growth, and when supplemented with protein-free milk and butter fat was able to support normal growth and reproduction. They found that no better growth was induced, but more frequent reproduction with lower mortality and more general well-being of animals were obtained when 5 per cent of casein was added to a diet containing 50 per cent cotton-seed flour with butter fat, protein-free milk, lard, and starch. Normal growth and reproduction did not result from diets containing 50 per cent cotton-seed flour in which there was a lack of butter fat, protein-free milk, or both. On a diet containing fifty per cent cotton-seed flour with the addition of casein and butter fat, but with no mineral matter other than that from the cotton seed, rats grew normally and reproduced, but the second generation did not make quite normal growth.

* Cotton-seed flour is prepared by finely grinding, sifting, and perhaps also aspirating the meal so that particles of lint, hulls, etc., are removed more completely than from the ordinary cotton-seed meal used in stock feeding.

Osborne and Mendel also found the proteins of cotton-seed flour to be efficient in nutrition, not only when fed alone in relatively abundant amounts but also when used as supplements to maize protein. They obtained toxic effects from the feeding of cotton-seed kernels but not from the cotton-seed flour. Like Withers and Carruth they demonstrated that the harmful substance could be removed from the kernels by extraction with ether; but the kernels can also be rendered harmless by steaming, which is a step in the usual commercial process of extracting the oil. The results of heating were, however, not altogether uniform and Osborne and Mendel suggest that undue heating may render the meal unpalatable or otherwise unsuitable for nutrition, in addition to destroying the original deleterious substance, and that these facts may help to explain the conflicting evidence regarding the alleged suitability of different samples of commercial meals.

These recent investigations upon cotton-seed flour are worthy of careful study both because of the great economic importance of this material and because they illustrate well the application of modern methods of nutrition research to the solution of a long-standing problem regarding the utility of an abundant but relatively neglected food material.