Published online by Cambridge University Press: 15 May 2009
It is shown that bits of allantois-on-shell cut from any one egg do not differ either in susceptibility to infection by influenza viruses or in yield of haemagglutinin. For this reason the dose-response curve within any egg is Poissonian.
Bits cut from different eggs vary both in susceptibility and in haemagglutinin yield. Unlike in whole eggs however, the scatter is so small that it can be demonstrated only by refined tests done on a large number of subjects. For practical purposes the dose-response curve may still be taken as Poissonian.
Susceptibility and yield vary independently of one another.
The data are used to define the accuracy of infectivity tests in bits of allantois-on-shell; the expected errors are given both for straight replication and for tests done with different reagents, by different operators, or at different times.