From the data given it will be seen that under the same experimental conditions and on the same diet the variation in gain of body weight amounted to as much as 75 per cent and the gains per gram of protein ingested varied as much as 10 to 30 per cent in different animals. The most plausible explanation of the lack of uniformity in the rate of growth in their animals was that their stock included animals of very low vigor as well as those of greater vitality. There was thus introduced into their interpretation an averaging of growth in animals not able to make efficient use of food for growth with growth in others in which growth impetus and synthetic powers were greater. If we desire to learn the extent to which a food protein during growth can be converted into body proteins it would be more logical to accept only the greatest gain observed in any animal, discarding all other results, rather than to average the gains made by good and poor ones. From their data Osborne and Mendel estimate that the gain of weight per gram of protein eaten during a period of eight weeks was for lactalbumin 2.34 and for casein 1.70 grams. In three experiments with casein fed at 12.0 per cent of the food mixture, the animals made an average gain of 2.25 grams per gram of casein consumed. One animal attained a gain of 2.39 grams and another of 1.99 grams. Their results are not very convincing as proof of the superiority of this method as a measure of the relative biological value of proteins. In interpreting the value of the data in the tables in which lactal-bumin and casein are compared, it should be borne in mind that neither protein was fed as the sole source of nitrogen, but that it was supplemented with the nitrogen of 28.2 parts of protein-free milk. It has already been pointed out that lactalbumin is actually an incomplete protein and is incapable of inducing any growth whatever unless its deficiencies are made good by some other source of amino-acids. The extent to which the protein-free milk nitrogen supplemented the proteins in these experiments was very great. Thus in the diets where 16 per cent of lactalbumin was included, 8.3 per cent of the total nitrogen of the diet came from protein-free milk. In the diets which contained 10.0 per cent, 6.2 per cent and 3.3 per cent of lactalbumin, the per cent of the total nitrogen derived from protein-free milk was 13.4 per cent, 22.4 per cent and 41.0 per cent, respectively. The recorded data of Osborne and Mendel do not in any sense represent a comparison of the biological values of casein and lactalbumin, and actually give a wholly false impression. It has already been pointed out that they were forced to revise because of the faulty technique in their experiments their views concerning the values of several of the proteins which they studied most extensively (4).