This study on pneumonia virus of mice (PVM) was carried out in order to obtain as accurate data as possible on the degree of variation which may be expected in titrations of the virus or of antibodies against it in vivo. It is believed that the knowledge gained will facilitate further investigations on this latent pneumotropic virus and make possible a more exact assessment of the significance of experimental results obtained with the agent.
The reproducibility of 50 per cent maximum score titration end points with PVM in mice is such that the chances are 19 out of 20 that a difference of 1.084 log units (i.e. a twelvefold difference) in the end points obtained in two separate titrations is significant. The reproducibility of 50 per cent maximum score serum dilution end points in neutralization tests with PVM in mice is such that the chances are 19 out of 20 that a difference of 0.626 log units (i.e. a fourfold difference) in the end points obtained with two sera against similar amounts of virus is significant.
It was found that there is a linear exponential relationship between the serum dilution end point and the quantity of PVM used in a neutralization test. This relationship appears to be identical with immune serum obtained from different animal species, and appears also to be identical to the linear relationship described previously in similar studies with influenza A virus.