Now that the importance of rhythmicity in biological behavioural functions has been recognized, the traditional distinctions between psychology and physiology have had to be reappraised. For example, the regular alternation between mania and depression in manic-depressive psychosis suggests the possibility of underlying rhythmic phsyiological mechanisms1. Similarly, the temporal structure of social interaction has been investigated and regular patterns have been found2. Chapple3 and others4 have suggested a loose analogy between these patterns and biological rhythms. Nobody, however, has analysed the temporal structure of interaction with the perspective and statistical techniques of chronobiologists. © 1970 Nature Publishing Group.