My purpose is to offer some constructive comments on the sampling methods and treatment of data used in climatological‐statistical analyses described in various scientific papers concerned with sunspot‐weather relationships and allied research into calendaricities or singularities. It is contended that much time can be saved and more profitable results obtained if, wherever possible, investigations include analyses based on data samples which are broadly homogenous with respect to time of year or season, topography (elevation, distance from sea or ocean, etc.), and geographical latitude and if account is taken of the main measured characteristics of the sunspot cycle such as the phase and amplitude, or the amplitudes of adjacent cycles.
It is reasonably possible that some changes in solar radiation could result in a more continental climate with lower winter temperatures and higher summer temperatures but that more extreme changes of sunspots and ‘continentality’ could result in lower temperatures in both summer and winter, perhaps because of more vigorous development of heat (thermal) lows over continents in summer [Lawrence, 1965a]. Such a process seems not to have been considered, for example, by Shaw [1965] or by Wexler [1956] in his work on the variation in insolation, general circulation, and climate. Shaw's assumption that sunspot changes must have similar influences (if any) on both summer and winter temperatures cannot be justified, and his negative results are of limited value. Wexler may have been deflected from pursuing the possibility of a sunspot‐continentality connection because his earlier work [Wexler, 1953] showed little evidence to suggest a clear‐cut continentality factor in sunspot‐climate relationships.