When 5 per cent of butter fat or cod liver oil is added to a bread and milk diet, in itself adequate to promote the breeding and rearing of a healthy stock of mice through a period of years, the resistance of these mice to per os infection with the paratyphoid mouse typhoid bacillus (B. pestis caviæ), as compared to that of mice on the unmodified diet, is definitely increased. A similar effect may be obtained with a McCollum "complete" diet, even when the butter fat is omitted, and with a bread and milk diet in which the milk used has been rayed with a mercury vapor lamp. When an inactive fat like Crisco is added to the bread and milk diet, the results obtained are not very clear-cut. While seasonal fluctuations in resistance to mouse typhoid were not completely eliminated by the various modified diets, they were nevertheless reduced, the modified diets tending to stabilize the death rate at a point lower than that usually reached by the mice on the control diet.