The Biological Assay of Carcinogens

Many experiments have been described in which a batch of animals are painted or injected with a carcinogen, and the dates recorded when warts or malignant growths are detected. The object of such experiments is usually to compare the potency of different carcinogens, or of different doses or methods of application, or else to compare the sensitivity of different batches of animals to a given carcinogen. It is convenient to have some means of summarizing the results of each experiment in a single number, together with an indication of its precision. The purpose of the present paper is to describe suitable procedures for arriving at such a number. The methods described are not essentially new; in particular the procedures advocated by C. I. Bliss (4) in a paper on the calculation of the time-mortality curve are in the main applicable to the present problem, and the method given below for dealing with experiments in which some animals die prematurely without tumor is an adaptation of a method described by W. L. Stevens (17) in an appendix to the Bliss paper.