192 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1931 



DAILY MARCH OF SOLAR VARIATION 



As regards the daily variation of the solar radiation, our results 

 are less satisfactory. Many days are wholly lacking from the 

 record, owing to unfavorable weather conditions, and many others 

 are unsatisfactory. The three widely separated observing stations 

 show a similarity, it is true, in their record of the daily march of 

 solar variation, but are not in close enough agreement to fix it 

 definitely. We have in mind, however, several improvements of 

 apparatus which we hope will bring closer agreement. 



WEATHER CHANGES ASSOCIATED WITH SEQUENCES OF SOLAR 



VARIATION 



The station at Montezuma is far more satisfactory than either of 

 the others. I show in Figure 15 the record of its daily observations 

 since 1924. Sequences of rising and falling solar-constant values, 

 respectively, are indicated in the figure by full and dotted curves. 

 I prefer to segregate these by months, including in a separate 

 group all the sequences of a given month for the seven years 1924 

 to 1930. I have computed the average march of departures of tem- 

 perature and barometric pressure at Washington, and of temperature 

 at Williston, associated with these instances of rising or falling 

 solar radiation. These results for March, April, September, and 

 October are shown in Figures 16 and 17. About 10 cases contribute 

 to each mean curve shown. Full and dotted curves correspond, 

 respectively, to rising and falling solar-radiation sequences. The 

 several months differ in detail, but agree in this, that the march 

 of weather shows opposite trends following the incidence of solar 

 sequences of opposite directions. 



It is curious and interesting to note that, though concurring with 

 Washington in displaying this phenomenon of opposition just de- 

 scribed, Williston temperature curves generally run in opposite sense 

 to those of Washington. I mean that the full curve corresponding 

 with rising solar radiation being found generally above the dotted 

 one for Washington, we find it generally below the dotted one for 

 Williston. 



We note for both stations certain critical dates when the separation 

 of the full and dotted curves reaches maxima. At Washington these 

 critical dates are approximately 5, 11, and 17 days after the full 

 development of the solar sequence. At Williston corresponding 

 critical dates seem to occur after about 2, 7, and 13 days, respectively. 



If I am right in regarding these contrasting weather phenomena 

 as really depending on rising and falling sequences of solar radia- 

 tion, I must make the assumption that the solar changes produce 



