Autopsy of the mice revealed widespread necrosis of liver and spleen, hemorrhages into the intestines, pleural exudation, and other occasional findings. Survivors frequently developed pulmonary consolidation or jaundice. The dominant type of lesion depended on the strain of virus used.

All attempts to demonstrate propagation of the influenza viruses outside of the respiratory tract failed. It was concluded that the early lesions were the result of toxic activities of the virus and not of virus multiplication in the affected tissues.

Injection into chick embryos of highly diluted inocula produced higher titers of virus, hemagglutmin, and toxicity in the allantoic fluids than the use of more concentrated seed culture. Serial passage of various strains in high dilution frequently increased the toxic activity.

The infectivity often reached its peak in 24 hours when tests for toxicity were still negative. Maximal toxicity was usually not attained before 48 hours.

The toxic activity could not be separated from the infective property by such means as differential centrifugation and adsorption onto and elution from chicken red cells. However, upon heating, formalinization, and irradiation with ultraviolet light, the ability of the agents to propagate was lost at a faster rate than the toxic property.

The toxic property remained stable for 2 to 3 months at 4°C. This stability was comparable to that of the infectivity for chick embryos.

Specific immune sera neutralized in high dilution the toxic activity of the homologous virus. Non-specific neutralization occurred in low dilutions of normal and heterologous immune sera. Strain differences were indicated by this method of testing.

Vaccination of mice by the subcutaneous or intra-abdominal routes protected mice specifically against the toxic effects of intra-abdominally or intravenously injected preparations of virus.