Nothing puzzles me more than time and space; and yet nothing troubles me less, as 1 never think about them." (Charles Lamb, 1810) WE CAN HARDLY ACCUSE the climatologist of not thinking about time but he must certainly admit to being deeply puzzled by it. Many of the changes within the atmosphere happen so quickly or so slowly that he can neither perceive them nor measure them and he is inclined to dismiss these events as "long-term departures from the normal" or "irregular short-term fluctuations."l These attitudes tempt the regional climatologist to retire into a world in which all that is really siyificant hsts no more than 30 years and no less than a few hours. His instruments, observational procedures, and descriptive models are usually designed to detect only the changes occurring within these limits. Time is, however, a continuum and any limits which wc impose upon it are arbitrary. This paper presents a descriptive time-scale which looks beyond these arbitrary limits to emphasize the continuity of climatological events and relationships in time. Such disparate topics as the amplitude, perception, measurement, or estimation and cause of change are accommodated and interrelated within the scale. Although for the sake of simplification the scale is developed only for temperature, it can be readily adapted to incorporate other elements.
THE SCALEThe upper limit of the time-scale is fixed by the age of the earth and the lower limit is arbitrarily fixed at one second ( Figure 1 ) . Between these limits the time axis is calibrated in powers of 10 seconds. Each point along this axis represents the period of a temperature change. Certain of these points are related to very familiar changes such as those due to the axial rotation and orbital revolution of the earth. The amplitudes of temperature change vary along the time axis. Figure 1 identifies schematically high amplitudes in five parts of the time axis: (1) period one day, due to axial rotation; (2) period one year, due to orbital revolution; ( 3 ) period approximately lo4 to lo5 years, due to the alternation of glacial and interglacial ages within the Quarternary period; (4) period of the order of los years, the approximate interval between glacial periods in geological time; (5) period 4.5 billion years, due to the aging of the earth.There has been no attempt to evaluate amplitudes of change in Figure 1 because their magnitudes are subject to considerable spatial variation. However, methods of power spectrum analysis are available for evaluating amplitudes of change.2 Within the period of availability of climatological data these methods might detect 365 CANADIAN GEOGRAPHER, XVI, 4, 1972
