Abstract
This is the last of a series of 3 papers dealing with methods for analyzing biological responses which are recorded as times. The author discusses the problems arising from incomplete data which are the result of intentional or accidental failure of the experimenter to observe the response in some individuals or which result when individuals are either immune to treatment or recover from the effects of treatment A mathematical model for survival is presented and the problems of immunity and recovery are elaborated. Finally, applications to biological assay are discussed.