In order to gain information concerning the relation between the character of the diet of the mother and the nutritive quality of her milk, McCollum and Simmonds (1) carried out a series of experiments with lactating rats, whose diets were faulty in known respects. They observed the effect on the growth of the young nursed by mothers receiving a highly satisfactory diet until the completion of their term of pregnancy. As soon as the young were born the litters were in all cases reduced to four in order that the nutritive undertaking of the mother should be in no case burdensome. The mother was at once placed upon a diet which would not induce any growth in a young rat separated at weaning time from the mother. The food of the mothers in the course of various experiments was made faulty in different respects. The faults included all the recognized factors which contribute to making a diet satisfactory, but the number of deficiencies in any single diet varied from one to three.

In one case a mother was fed upon a diet of purified protein, carbohydrate and a salt mixture supplemented with an alcoholic extract of wheat germ to furnish the factor, water-soluble B. This diet contained everything necessary for the nutrition of a young rat during growth, except fat-soluble A. The problem was to find whether the mother could, through the agency of the mammary gland, synthesize this substance from some other component of her diet. Experience has shown that the young animal after weaning cannot, for its own preservation, produce the fat-soluble A from other complexes in its food. The results of these experiments indicated clearly that the quantity of fat-soluble A in the milk of the mother confined to this deficient food was distinctly below the optimum, or below the average content of milks. The milk secreted under the circumstances did not furnish a sufficient amount of this substance to promote growth in the young at the maximum rate.