NO. 6 SOLAR RADIATION AKD WEATHER CLAYTON 63 



In 1925 the forecast was changed to a prediction of the value of 

 solar radiation to be expected on the same day as that on which the 

 faculse and spots were observed. These forecasts are forwarded to 

 Washington about 24 hours before the observed values of solar 

 radiation are received at Canton. A plot of the observed values 

 (full line) and the predicted values (broken line) is given in Fig. 45 

 for March and April, 1925. The letter u indicates that the observa- 

 tion was uncertain. It is seen that the general trend of the solar 

 changes was predicted but not the detailed changes. In other words the 

 solar radiation was more variable than shown by visible phenomena. 



A THEORY OF THE METHOD BY WHICH SOLAR HEAT CHANGES 

 AFFECT ATMOSPHERIC CONDITIONS ON THE EARTH 



The rapidity with which the pressure rises in high latitudes of the 

 earth when the solar radiation increases and in turn falls when solar 

 radiation decreases, indicates that the influence is exerted in some 

 way directly on the atmosphere itself and not indirectly through 

 changes at the earth's surface. The atmospheric changes appear 

 to occur at the centers of action simultaneously with the solar changes 

 or at most there is a delay of only a few hours. It seems to me 

 probable that an increase of solar radiation heats the upper air, more 

 particularly in the equatorial belt where the sun is nearly vertical. 

 There results from this heating an expansion of and a movement of 

 the air from the equatorial belt which causes the pressure to fall 

 within that belt and to rise in high latitudes, determined by the lower 

 mean temperature of those latitudes and the influence of a rotating 

 earth on moving air. That such an influence can be exerted rapidly 

 at a great distance is shown by the rise of pressure in high latitudes 

 coincident with the diurnal fall of pressure in tropical and subtropical 

 regions. 



This is the primary effect of the increased solar radiation and 

 remains more or less the same throughout the year, because the 

 relation of pole to equator remains the same, modified to some extent 

 by the north and south movement of the sun. It may be called the 

 first variable. 



A second variable arises from the distribution of land and water. 

 The land is colder than the water in high latitudes in winter and 

 warmer in summer. This difference determines an increase of pres- 

 sure with an increase of solar radiation over the continents in winter 

 and a fall over the oceans, and a reversal of this efifect in summer. 

 This reversal of the efTect with the seasons in high latitudes becomes 

 most marked with slow and prolonged changes in the mean values 



