1. In September, 1925, explosive and highly fatal epidemics of a Friedländer-like bacillus infection arose in three of four experimentally controlled mouse populations. In the populations to which daily additions were made, characteristic secondary waves ensued. These waves appeared at definite population levels, lasted a definite number of days, and reduced the population to a definite low census. Subsequently, endemic periods intervened, followed by the disappearance of the disease during March, 1926.
2. In August, 1926, renewed epidemics of the infection arose in all four populations. These outbreaks resembled one another but were more enduring than the preceding ones. Moreover, fewer secondary waves followed, and the disease disappeared in February, 1927.
3. The virulence or disease-producing power of the Friedländer-like organisms, as determined by the ability to spread and induce typical epidemics among populations of previously unexposed mice, and by direct inoculation titrations remained constant during pre-epidemic, epidemic, post-, and inter-epidemic phases of the infection.
4. The Friedländer-like organisms were not recovered in cultures from bedding, food, or feces, but they were obtained from the nasal passages. Mice harboring the organisms usually died within a few days of the test; a few, however, survived and ceased to be carriers. The carrier rate in a given community was found to increase prior to epidemic outbreaks and to decrease shortly before the time of peak mortality.
5. The substitution of relatively susceptible for "standard" immigrants in a population was followed by an increase in severity and spread of the disease, whereas a substitution of "standard" for resistant immigrants was followed by a fall in severity and extent of disease.