The protein content of the four diets was as follows: (a) pellets: a minimum of 21 per cent "crude" protein (according to the manufacturer); (b) diet 20 C: 20 per cent casein; (c) diet 8 C: 8 per cent casein; (d) diet 8 C + AA: 8 per cent casein supplemented with 12 per cent of a mixture of essential amino acids. All diets provided an adequate supply of minerals and vitamins. They were administered ad lib.

Three strains of pathogens virulent for mice were used for the infection tests, namely: Staphylococcus aureus, Mycobacterium fortuitum, and Mycobacterium tuberculosis bovis. The bacteria were injected by the intravenous route.

The experimental regimens were begun at different times before infection, and were continued until death of the animal, or until termination of the experiment.

It was found that mice on the 8 C diet exhibited much greater susceptibility to infection than did mice on the 20 C diet; mice receiving pellets were intermediate between these two groups. The infection-enhancing effect of the 8 C diet could be entirely corrected by amino acid supplementation (diet 8 C + AA). Indeed, mice fed diet 8 C + AA proved the most resistant to infection.

The fact that animals fed pellets (which contain a minimum of 21 per cent protein) consistently died faster following infection than did animals fed diets 20 C or 8 C + AA suggests that qualitative characteristics of the protein in the regimen are as important as the quantity of protein fed in determining susceptibility to infection.

The differences in susceptibility exhibited by the mice on the four experimental diets were the same whatever the species of bacterial pathogen used for the infection test, the size of the infective dose, and the duration of the disease.

There was no apparent relation between the effects of the diets on the weight curves of the animals, and on resistance to infection. Mice on diet 8 C (which were most susceptible) gained weight as rapidly as those on 20 C and more rapidly than those fed 8 C + AA (which were most resistant).

All the tests reported in the present paper were carried out with young mice, which were placed on experimental diets within 1 to 2 weeks after weaning. Preliminary experiments suggest that the relation between dietary factors and susceptibility to infection was more difficult to bring out in older animals. There was evidence also that this relation was most apparent during the first weeks that the animals were fed the experimental diets, and became less striking after several weeks.