Abstract
Parker 1 has presented statistical evidence for the concept that the frequency distribution of lesions with vaccinia virus is related to the chance distribution of virus particles in the inocula. This was based on the resemblance of the observed frequency to the theoretical Poisson frequency to be expected if a single virus particle causes infection. Similar studies on the purified papilloma protein 2 have shown that results with this virus follow this theoretical expectation as well as did those with vaccinia virus. However, consistent deviations of observed from theoretical results at the extremes of the frequency distributions, and the finding of about 100,000,000 papilloma protein molecules in the dilution giving the 50% point (where only 0.69 particles should have been according to this concept) have led to critical examination of Parker's data and interpretations. Using the procedure recommended by Haldane 3 for determining x2, the fit of observed to theoretical frequencies in Parker's data was good in 2 experiments and improbable in 2.
On the basis of these inconsistencies other relations were sought to explain the character of results with vaccinia virus and the papilloma protein. It has seemed that a more logical interpretation and one more consistent with experimental results is that the observed frequency is an S shaped dose-response curve determined by variations in host resistance. If host variation follows a normal distribution, response to logarithmic doses will be S shaped and will be constant for a given material and population of test animals. Such a curve would represent an integrated frequency curve which can be placed in straight line form by plotting normal equivalent deviation against logarithm of dose. 4
Get full access to this article
View all access options for this article.