1. A kind of mouse village was set up into which was introduced a small number of mice fed on a culture of so called mouse typhoid (Bacillus pestis cavice of the Bacillus paratyphosus B group) bacillus. The spread of the infection so induced to the cages, or "homes," of the other mice was left to accident through the attendant who fed the animals and cleaned the cages. That this means was likely to be sufficient was deduced from the epidemic reported by Lynch. A spot map was kept throughout the experiments which extended from 1919 into 1922.
2. The first effect of the exposure of normal mice to a much smaller number of mice fed on the culture is to set up a sporadic, not an epidemic outbreak of mouse typhoid. This is the regularly recurring incident of the experiment as shown by low mortality and low cage attack rate. Such a sporadic prevalence is self-limited in time.
3. The introduction of fresh, normal mice into a community in which sporadic deaths are occurring leads regularly, not to the further extension of the sporadic deaths, but to an epidemic spread, as shown by high mortality and high cage attack rate. The epidemic begins with deaths among the new mice, but extends to the old mice which succumb later. The spread ceases and the wave subsides before all the mice have succumbed. A state of equilibrium between the infecting bacillus and the surviving mice is reached; no more deaths occur. The epidemic outbreak, therefore, is also self-limited in time.
4. If, now, another new addition of normal mice is brought into the potentially infected community, the events are reenacted; deaths occur among the new, another epidemic wave sweeps through the population, again claiming victims among the previous survivors. Through the replacement of the destroyed mice with fresh, normal mice, epidemic wave after wave is produced, until certain groups of old survivors are entirely wiped out. There seems to be no limit to this process, as there will always be survivors at least of the later groups added.
5. The dying down of the epidemics and the attaining of the equilibrium do not mean the elimination of all the bacilli. Potential infection still lurks in mouse "carriers" and on the hands of the attendant. No successful method of completely removing the bacilli from the hands of the attendant was found.
6. The maintaining of the epidemic waves is dependent on the presence of new lots of mice, whether supplied from without or produced within through new births. It is the latter sources which provide the consumable material in such natural epidemics as that described by Lynch. Hence their slower movements as compared with the epidemic spreads and rise and fall of the epidemic waves in the artificially propagated instances in which new batches of mice are brought into the village in bimonthly intervals.
7. The evidence at hand is to the effect that the degree of infectivity of "mouse typhoid" bacilli is highly fluctuating, and it appears that all the bacilli which are included under that name, classed variously as Bacillus enteritidis, Gärtner's bacillus, Bacillus paratyphosus B, Bacillus suipestifer, and Bacillus pestis caviœ, infect mice in a similar, possibly indistinguishable manner, inducing self-limited outbreaks of disease reaching at times epidemic proportions. This quality of infectivity, or virulence, is one factor in the process but alone does not suffice to account for the observed facts. A second influence is not improbably quantity, or dosage, of the inciting microorganism. Attention to the manner of spread of an epidemc, as revealed graphically by the spot map, shows that it is never uniform but patchy. The supposition is, therefore, that among the new mice are certain individuals so highly susceptible as to react to small numbers of bacilli of average infectivity. Within these animals the multiplication is rapid, so that a wider spread of much larger amounts of these average, or even temporarily enhanced bacilli takes place with the inevitable effect of communicating, through greater dosage, etc., the infection to other and less susceptible individuals among the new and also the older lots. This process continues so long as rapid multiplication can occur. This latter is, in turn, determined by the innate tendency of the bacilli to return to an average of infectivity and by mouse individuals of greater resistance to restrict free growth and multiplication. In this manner, possibly, the epidemic spread is checked and the curve representing it made to fall more or less quickly to a base-line.