1. A design of experiments and technical procedure is described for the study of interference with transplantable tissue growth by test agents. The methods are guided by the principle of randomization and are based on the pairing of test and control in each host.
2. A system of analysis, employing well known statistical measures, is applied to results obtained with these methods and shows that they lead to consistently reproducible results.
3. A simple analytical method is proposed for correlating results obtained with different kinds of transplantable tissue in different strains of animals.
This method is based on the condensation of an extended frequency distribution of events in control series of cases into a three category array of a normal frequency distribution. The three categories of test action are the inhibitory, the ineffectual, and the accelerative.