Applied macroeconomists commonly regard the seasonal variations of the economy as a hindrance rather than a help to the understanding of behaviour. Thus both in commenting informally on economic developments and in the more formal tasks of modelbuilding and forecasting seasonally-adjusted data are almost invariably used in preference to raw data when both are published. The nature of the patterns displayed by seasonal variation is very little discussed. One purpose of this note is simply to describe seasonal variation as it is estimated in some of the official data series, providing some tables which may be useful for general reference. But the aim is not just descriptive. It will be argued that seasonal variations may throw useful light on some controversial issues in macroeconomics.