ABSTRACT
Between these long (1()4-105 years) and short (1 day and 1 year) periods no such regularity is apparent in observed atmospheric behaviour. Hints of recurrence tendencies at intervals not strictly constant but of the order of 5, 15, 30 and 45 days, which appear for a time and then are lost again, may correspond to the life-cycles (maturing and decay) of an individual anticyclone, of one mode of poleward transport of heat, or of the migration of some tropospheric or stratospheric feature around the Arctic or Antarctic respectively. Such evanescent rhythms have, at best, a very limited predictive value: even this can only be attained by monitoring the progress of some flow-system in the atmosphere, the 'mechanism' of which is either reasonably well understood or at least familiar in performance. The nearest approach to regular, and continuous, operation may be that of the 'Southern Oscillation', to be described later in this chapter, and the attendant (approximately) 2-year fluctuations in surface weather some of which have already been mentioned (Table 5.1, p.209).
