A procedure introduced by McCollum, Simmonds and Parsons (5) for the study of the relative values of different proteins, or for the supplementary values of proteins from different sources, involves feeding a series of groups of animals on a series of diets which are of uniformly excellent quality with respect to all factors except quality of protein. The diets are all made to contain 9 per cent, of total protein, which may be derived from a single food or from two or more sources as stated above. On such diets young rats are observed during the entire growth period and thereafter up to the time when pronounced development of senile characters are noted. The groups contain animals of both sexes, and the fertility and infant mortality are carefully observed. The behavior of the animals, especially with respect to a tendency to destroy their new-born young when the protein supply is unsatisfactory, is also carefully recorded.

This method was the outgrowth of considerable experience in searching for the most satisfactory procedure for making a comparison of the effectiveness of different foods as protein supplements for one another. McCollum and his co-workers had conducted numerous experiments in feeding animals diets containing different percentages of protein, and otherwise enhanced so as to make them of good quality, and had made an effort to study the relative values of purified proteins and of the mixtures of proteins contained in individual natural foods, both for growth and for the replacement of the nitrogen lost through endogenous or tissue metabolism (6). In the early studies too little was known about supplementing factors other than protein. None of these results are to be regarded as very satisfactory for the reason that factors other than protein in the experimental diets employed, were not always satisfactorily improved to make the results turn entirely upon the quality of the proteins of the food.