My contribution to this discussion is an appeal for a more careful and consistent recognition of the effects and operations of chance in the study of data which may be employed in investigations of solar and terrestrial relations, periodicities, etc. Weather conditions, atmospheric transmission of radiation, magnetic phenomena, sunspots, observed intensities of radiation, and values of like phenomena are subject to large and irregular accidental variations, due account of which must be taken in reaching conclusions.My point of view is illustrated in an admirable manner by figure 1, which serves to show how weak the observational basis still is to justify the claim that there are important irregular changes from day to day in the intensity of solar radiation.