154 The Realm of Nature CHAP. 



ward, he will see how the weather varies in all parts of the 

 islands as the cyclone passes along its path. By shifting 

 the centre to north or south, and changing the direction 

 of passing (but always moving the tracing as the arrow 

 flies), the effect on the student's own locality of cyclones 

 passing in any direction may be followed. Remembering 

 that as isobars of successively lower value are passing the 

 barometer is falling, and that as isobars of higher value 

 are passing the barometer is rising, it will be found possible 

 to identify the actual movements of a cyclone by watching 

 the barometer and the .changes of wind and weather. 

 In order to predict on Monday the kind of weather and 

 direction of wind on Tuesday when a cyclone is passing, 

 it is necessary to know where the centre is, at what 

 speed, and in what direction it is moving, so that a map 

 of the conditions expected on Tuesday can be drawn up 

 from the data supplied by Monday's observations. But 

 in order to predict the intensity of the weather and the 

 force of the wind, it is necessary to know whether the 

 cyclone is "deepening" or "rilling up," that is, whether 

 the gradient of pressure from circumference to centre is 

 growing greater or less. Only experience and practice can 

 guide a forecaster in these matters, and the success of the 

 predictions issued daily by all civilised governments depends 

 on the knowledge and skill of the men who make them. 

 It often happens that a cyclone does not follow the usual 

 path, or that the pressure at the centre increases when the 

 forecaster thought it would diminish, or that a secondary 

 depression suddenly forms in an unexpected place, and of 

 course in all such cases the forecast made is a failure. Yet on 

 the whole more than 80 per cent of the predictions issued 

 in Britain and America are successful. The British Islands 

 are divided for purposes of forecasting the weather into 

 eleven districts. At 10 A.M. and 8.30 P.M. forecasts for 

 the next 24 hours of the weather in each of these districts 

 are published at the Meteorological Office in London from 

 observations made all over the country at 8 A.M. and 6 P.M. 

 The weather-charts, reports, and forecasts in a daily and 

 weekly form are sent out to subscribers by the Meteoro- 



