On cycles of pressure especially in the neighbourhood of symmetry points

Schuster's criterion and the theory of probabilities have established extraordinarily few meteorological periods. The reason for the failure of periodogram analysis is that it is not applicable to damped oscillations. On the other hand, symmetry points in pressure variations appear to indicate cycles of some persistency and some short series have been analysed by Fuhrich's self-correlating method, of which a short summary is given. The importance of Brunt's criterion for utility has been shown by applying it to the periods of the motion of the earth's pole. Fuhrich's criterion for reality is then treated and several features are discussed—independence of the analysis interval, the representation of artificial functions and the case of equal amplitudes. Sixteen analyses of the pressure variations near symmetry points have been summarised, and it is found that a duration including two or three cycles when analysed by Fuhrich's method yields a good representation of the variations with a correlation coefficient exceeding o.8 between the observed and calculated data. The method appears capable of giving at times a forecast of the pressure for 10 to 14 days in winter in central Europe.