NO. 



SOLAR RADIATION AND WEATHER— CLAYTON 



41 



peratiire in high latitudes of the earth in response to changes in solar 

 radiation, there could not be a high direct correlation between the day- 

 to-day weather changes at any one station and the day-to-day changes 

 in solar radiation. But in order to determine approximately how 

 large such a correlation may be for stations in the United States 

 during an interval of a few months when the mean level of the 

 solar output is low and nearly stationary, the departures from normal 

 temperature at Williston, N. Dak., and New York, N. Y., were cor- 

 related with solar radiation for the interval January to April, 1924. 



Table 11. — Correlation between Solar Radiation and Maximum Temperatures of 



Williston, Dak., and New York, N. Y., for Same Day and 32 Days 



Following Observations of Solar Radiation, 



January-April, 1924- 



Note: The correlation coefficient for the day preceding the solar observation was for Williston 

 0.24. 



The mean correlations for 32 days following the observed values of 

 solar radiation and for one day preceding at Williston are given in 

 table II. These are plotted in figure 34, which shows that during 

 this interval there was a distinct maximum correlation at Williston 

 on zero day, that is, on the same day that the solar radiation was 

 measured. Secondary maxima occur 11, 16, and 28 days later, the 

 latter no doubt being due to a return of the same influence by a 

 rotation of the sun on its axis. These correlations are not high, 

 the maximum being only 0.32 on zero day. At New York the maxima 

 of the correlations all come three to four days later than at Williston, 

 but are not so large, the maximum being 0.23 on the third day follow- 

 ing the observation of solar radiation. 



