A number of methods have been employed in attempts to induce encephalitis in guinea pigs with the Levaditi C strain of herpes virus. Some of these consisted of different modes of inoculation of the virus itself and others of different ways of combining it with vesicular stomatitis and neurovaccine viruses so as to obtain the concomitant effects of both. In still another test the Levaditi virus was combined with the neurovaccine in a manner calculated to bring about the maximum action of each at the same time. By all these methods, the Levaditi virus failed to evoke the characteristic encephalitis which this specimen is capable of inducing uniformly in rabbits. On the other hand, when the Levaditi herpes virus is inoculated into the brain of guinea pigs in conjunction with suitably timed corneal injections, it acquires active encephalitogenic properties. The results just noted suggest several considerations: 1. The possibility of increasing the virulence of a filtrable virus by animal passage in a special manner. It is not likely that the increase as observed was due to dosage, for after the virus acquired its encephalitogenic property for guinea pigs, the usual amounts of virus suspensions sufficed to induce, in a uniform way, typical encephalitis. 2. The opinion previously expressed by Flexner that the guinea pig serves merely to separate weak from strong strains of herpes virus is supported: for only according to the particular method described, could the encephalitogenic power of the Levaditi virus be developed and the weak be converted into a strong herpes strain. With the acquisition of this power, the Levaditi virus acted in precisely the same manner as strong herpes strains both in the guinea pig and the rabbit. Moreover, it was shown in guinea pigs that cross-immunity occurs between weak and strong strains. 3. The two samples of neurovaccine virus employed were incapable of inducing encephalitis in guinea pigs after intracutaneous, intratesticular, corneal, or intracerebral inoculation, although they were actively encephalitogenic in rabbits. In spite of the fact that the vaccine virus and herpes virus are different, as shown by the histopathology and absence of cross-immunity, they behave in the same way when injected into the brain of the guinea pig. The failure of the concomitant action of both viruses to induce encephalitis in the guinea pig suggests that the association of two viruses, under the experimental conditions outlined, is incapable of inducing encephalitis, if either, by itself, is non-encephalitogenic. 4. The serum from normal guinea pigs may neutralize a weak (Levaditi C) but not a strong (H.F.) strain of herpes virus; but the neutralizing action of the serum on Levaditi C virus is not uniform. 5. The Levaditi strain of virus can increase in quantity in the brain of the guinea pig to a degree which permits detection and yet fails to evoke any distinctive clinical picture or definite histopathological changes. 6. Repeated intraperitoneal injections of Levaditi virus in guinea pigs elicit no signs of infection, yet they induce a solid immunity to strong strains of herpes virus.