2. Rather incomplete pathological studies indicated that typhus in the cotton rat is an acute infection primarily involving mesothelial tissue, and tissues of mesothelial origin. Depending somewhat upon the route of infection, rickettsiae were most easily demonstrated in smears from liver, brain, and pericardial and peritoneal exudates. By subinoculation into the yolk sacs of fertile eggs, rickettsiae were readily isolated from these sources and, with rather less facility, from blood.

3. Although relatively large numbers of rickettsiae were found necessary to produce lethal infections, the susceptibility of the cotton rat as measured by the development of an immunizing infection proved to be very high. The supporting data suggest that cotton rats might be suitable for the isolation of new strains.

4. Contrary to the earlier report (1), age was not found to influence markedly the susceptibility of the cotton rat to lethal infection with any of the strains studied.

5. Comparative studies indicated that cotton rats were just as susceptible to murine infection as white mice and more susceptible to Breinl infection than the latter animal. The susceptibility of cotton rats and of guinea pigs to Breinl infection was found to be about the same. Comparative studies in these animals were not made with other strains.

6. No toxic activity of rickettsial suspensions could be demonstrated for cotton rats, although suspensions of infected cotton rat livers were shown to be toxic for mice.

7. A method has been described by which three strains of Rickettsia prowazeki were carried in serial intracardial passage as fatal infections for cotton rats. No evidence was obtained that prolongation of such passage increased the virulence of the strains for the rats.

8. Cotton rats surviving typhus infection were shown to be solidly immune to reinfection with homologous virus and to possess nearly complete immunity to rickettsiae of heterologous immunologic character. Sera from recovered rats, furthermore, were shown to possess neutralizing, complement-fixing, and antitoxic antibodies but no agglutinins for proteus OX19.

9. The technique of a serum neutralization test in cotton rats has been described in which the test antigen consists of suspensions of infected cotton rat livers. By this technique neutralizing antibody was demonstrated in human, rabbit, guinea pig, and cotton rat sera. Although some degree of cross-neutralization could be shown, serum titers against homologous antigen were uniformly greater than against comparable doses of antigen of heterologous immune nature.