The resistance of mice to tuberculosis appeared to be independent—within wide limits—of the protein content of the diet. It is true that mice fed a diet very low in protein and high in carbohydrate proved highly susceptible, but resistance was normal if part of the carbohydrate was replaced by fat (peanut oil)—without any change in the protein content of the food.

Resistance to tuberculosis could be consistently and markedly decreased by adding sodium citrate (or glutarate) to a variety of diets.

The survival time following infection was greatly shortened if dinitrophenol or thyroxine were administered per os in amounts sufficient to limit the weight gains of non-infected controls. There was usually a lag period of several days before the infection-enhancing effect of these metabolic stimulants became manifest.

The procedures which increased the susceptibility of mice to infection with virulent tubercle bacilli also made it possible to establish in these animals a fatal infection with BCG.

There was no constant relation between weight gains of uninfected mice on the various regimens, and the effect of the latter on susceptibility to tuberculosis.

These findings appear compatible with, but do not prove, the hypothesis that a decrease in resistance to infection can be brought about by metabolic disturbances which cause either a depletion of the glycogen reserves of the body, or a reduction in the glycolytic activity of inflammatory cells, or an increase in the concentration of certain polycarboxylic acids and ketones in the tissues.