ENCEPHALOMYELITIS OF MICE II. A METHOD FOR Tm~ MEASUREMENT OI~ VIRUS AcTrvITY

The activity of a given virus-containing material is usually expressed in terms of minimal infective (X~.Ln.), or lethal, doses. Groups of test animals inoculated with material containing 1 ~r.I.D. per inoculum will show, on the average, 50 per cent positive results. A titration of material of unknown activity is simply an a t tempt to find this 50 per cent morbidity rate end point. In most instances the value has to be calculated through interpolation from two observations, preferably by means of the method of Reed and Muench (1). The result is usually given as log ~.r.D. per quantity of inoculum, x In titrating a specimen containing an unknown amount of a potent virus, one usually has to try at least six different dilutions. In the calculation of the titer, however, only two, or at most three, of the test groups are of great value. The accuracy of the method varies with the number of test animals used. If each test group consists of eight animals the theoretical error of the titer value obtained is from 0.2 to 0.25 log ~.r.D. The result hardly justifies the use of such large numbers of animals. Horsfall (2), applying the method of Reed and Muench to a series of titrations of influenza virus, found a mean deviation of 0.16 to 0.20 log virus units.