The sedimentation behavior of influenza virus in dilute solutions of electrolyte was found to be quite variable. At times the virus activity appeared to sediment at a rate comparable with that of particles about 80 to 120 mµ in diameter, at other times at a rate comparable with that of particles about 10 mµ in diameter, and at still other times the bulk of the activity appeared to sediment at a rate comparable with that of the larger particles and the residual activity at a rate comparable with that of the smaller particles. However, in the presence of a sucrose density gradient, the virus activity was always found to sediment with a rate comparable to that of particles about 80 to 120 mµ in diameter; hence it appeared that the variable sedimentation behavior in dilute electrolyte solution was due to convection or mechanical disturbances during centrifugation.
About 30 per cent of the high molecular weight protein present in the allantoic fluid of chick embryos infected with the F 12 strain of influenza virus was found to consist of a component having a sedimentation constant of about 30 S, and hence a probable particle diameter of about 10 mµ. The residual protein of high molecular weight was present in the form of a component having a sedimentation constant of about 600 S, and hence a probable particle diameter of about 70 mµ. The proportion of the 30 S component in allantoic fluid of chick embryos infected with the PR8 strain of influenza virus was found to be considerably less. The 600 S and 30 S components of F 12 allantoic fluid were purified and separated by differential centrifugation. The purified preparations of the 600 S component were found to possess a specific virus activity from 100 to over 10,000 times that of the purified preparations of the 30 S component, the difference in activity apparently depending only on the degree of fractionation of the two components. The purified 30 S component was found to sediment normally in the presence of 12 per cent sucrose, whereas the small residual virus activity of such preparations was found to sediment in the presence of a sucrose density gradient with a rate comparable to that of much heavier particles. It is concluded that influenza virus activity is not associated with material having a particle diameter of about 10 mµ, but is associated solely with material having a sedimentation constant of about 600 S and hence a probable particle diameter of about 70 mµ.
A STUDY OF CONDITIONS FOR THE OPTIMUM PRODUCTION OF PR8 INFLUENZA VIRUS IN CHICK EMBRYOS