The virus is not inactivated by the process of adsorption or precipitation : guinea pigs and mice inoculated in the brain with these materials develop lethal encephalomyelitis in the same manner as when chemically untreated mouse passage virus has been used. Moreover, there is no difference in the rate of absorption in vivoof the chemically treated and untreated virus preparations. After storage of the three immunizing preparations—the longest periods thus far studied being 2 to 3 months for mouse passage and for precipitated suspensions, and 6 months for adsorbed material—each was found to contain an amount of virus sufficient to produce immunity in animals against the usual intracerebral test inoculation. Finally, the protection afforded by the three preparations is apparently durable, as is true of many active viruses utilized in preventive treatments.

The amount of the virus necessary to confer protection may be defined as that which immunizes (a) with the least number of antigenic units and (b) with the minimum of febrile reaction and blood infection. In proportion as this amount is exceeded, the incidence of fever and of circulating virus increases and, on the other hand, as this amount is decreased, the degree of induced immunity is diminished.

We have thus shown that for this particular virus and in the guinea pig, one or two subcutaneous doses of I cc. of any of the different virus preparations, each containing 3 x 10³ to 3 x 10⁴ mouse infective units, bring about protection regularly against experimental infection by way of the nose or subcutis. The results are irregular when the test is made by way of the brain. By three injections, resistance is invariably obtained against as many as 10³ to 10⁴ lethal doses, given intracerebrally.

No matter in what form the virus is given, as mouse passage, or adsorbed, or precipitated material, in certain instances fever occurs and virus circulates. With the amount of virus adequate for immunization (3,000 to 30,000 m.i.u.) a mild or subclinical infection may occur in the guinea pig without other manifestation of disease. Lesser quantities of virus apparently fail to gain a foothold in the animal and thus fail to bring about resistance.

To conclude, a quantitative basis has been established for the comparison of the immunizing capacities of preparations employed in experimental equine encephalomyelitis in guinea pigs.