When mice were injected intracerebrally with doses of Bordetella pertussis vaccine greater than 5 ImD50 and challenged intracerebrally 14 days later with virulent B. pertussis there was an immediate reduction in the numbers of organisms. An analysis of this in vivo bactericidal effect has shown that large doses of an unrelated vaccine, Salmonella typhosa, equivalent in cell mass to about 50 ImD 50 of B. pertussis vaccine can achieve this effect, so for such doses the effect must be partly non-specific. This action is not maintained and so is not ultimately protective. Local immunoglobulin was also demonstrable 14 days after 300 ImD 50 of B. pertussis vaccine but following smaller doses of 10–20 ImD 50 it could not be found until after the mice had been infected and the blood–brain barrier impaired.

A similar immediate reduction in the numbers of infecting organisms inoculated 1 day after vaccination has been shown to follow very small, non-protective doses of vaccines unrelated to B. pertussis and to be achieved with lipopolysaccharide and endotoxin isolated from B. pertussis. Brains were not sterilized and only in mice receiving protective B. pertussis vaccine was the lowering of infection maintained beyond 2 days and the brains eventually sterilized.

The antibody passively protecting mice against intracerebral infection was found in the 19 S and 11 S globulin fractions of the serum of once-vaccinated mice and in the 11 S and 7 S fractions of the serum of rabbits and ascitic fluid of mice receiving repeated doses of vaccine. The IgM probably eliminated infections by immediate sterilization but had to be present locally to do so since it was unable to pass from the circulation into the brain, and was therefore inactive when injected intraperitoneally. The IgA and IgG were not so restricted and both the 11 S and 7 S globulins were capable of exerting an immediate suppressive effect on infecting organisms. The 7 S globulin was also capable of a maintained or delayed suppressive effect.

Lymphocytes from fully protected once-vaccinated mice, transferred 2–3 weeks after intraperitoneal vaccination, were able to confer some protection when injected intraperitoneally or intracerebrally into recipient mice infected 2 weeks after transfer. Homologous, non-concentrated antiserum from once-vaccinated mice, injected intraperitoneally 1 hr. before infection sometimes augmented the transferred immunity, whereas alone it was inactive.