Their study of the role of appetite in regulating food consumption has led them to state: "The experience which we have gained in measuring the food eaten by many rats at all periods of their growth has given us the conviction that the intake of the individual is determined in large measure by the energy requirement at any given period" (8) (p. 358). To quote from page 359 of the same paper, "At any rate the animals do not consume proportionately more of the artificial ration because it happens to be decidedly poor in protein; but roughly speaking, they appear to limit their feeding to the amount of food yielding approximately the requisite energy." This view is in accord with that of Rubner (9), who found that food which a dog will eagerly devour when in a room at the temperature of 0° C, he will in part refuse when the temperature is raised to 33°, under which conditions, his energy requirements are much lower.

In a later publication Osborne and Mendel (10) showed the food consumption of rats fed diets containing varied amounts of fat. Since fat has about double the energy value of protein or carbohydrate, much smaller amounts of the fat-rich diets would furnish the required amount of energy needed than would be necessary with the fat-poor diets. In all cases food consumption was found to be regulated by energy needs and was comparable in animals on diets of similar calorific value.

It is only necessary to point out that McCollum and his coworkers have throughout their studies employed diets low in fats, and very much alike in their energy values. Food intake should have been, therefore, according to Osborne and Mendel's experience, entirely comparable. Numerous records obtained in recent years in the laboratory at Johns Hopkins University have shown clearly that this is the case. Quoting again from the paper of Osborne and Mendel from which extracts have already been presented (p. 523), "McCollum does not state the amount of food which his experimental rats ate; but if we assume that they ate quantities of the type of food which he employed in amounts approximately equal to those eaten by rats in our experiments on foods of corresponding calorific value, the difference between the results of his experiments and ours disappear." The force of their repeated criticisms of the work of McCollum and his co-workers largely vanish with the recording of this comment.

Owing to faults in the technic employed in earlier experiments, we find ourselves to-day with extremely little information concerning the manner and extent to which any purified proteins supplement one another, and to the manner and extent to which they are by combination mutually enhanced in biological value. On the other hand we have a very considerable amount of accurate knowledge concerning the supplementary dietary relations among certain of our most important food-stuffs. To these supplementary relations we will now turn our attention.