108 ANNUAL. REPORT SMITHSONIAN INSTITUTION, 1923 



the matter could be put, for the first time, to a rigorous test. And 

 now we have made this test. It is conclusive and proves the sub- 

 stantial character of solar variation. Hereafter we walk by sight, 

 where hitherto we walked by faith. 



In the publication cited (Monthly Weather Beview, February, 

 1923) we show that in over 100 days, when results were obtained at 

 both Arizona and Chile, the average deviation of one station from 

 the other is 0.68 per cent, thus indicating a probable error of one 

 day's observation at one station of 0.41 per cent. The average devia- 

 tion for two years of monthly mean values between the two stations 

 is 0.3 per cent, and this small value would doubtless be considerably 

 smaller if the individual days of the several months had always been 

 coincident. Although in opposite hemispheres, where winter at one 

 station falls in summer at the other station, there is no evidence of 

 seasonal divergence between the two stations. They unite in show- 

 ing solar variation. Indeed the march of the monthly mean values 

 from November, 1921, to September, 1Q22, in which they agree closely, 

 gives the most conspicuous instance of long-continued solar change 

 in a given direction which we have ever noted. The reader may 

 compare it with the four years of monthly means curve given at the 

 top of the illustration. 



In short, there can not be, we think, any longer a reasonable ques- 

 tion that the sun varies, or that our observations can reveal these 

 variations satisfactorily. It is now a question for meteorologists 

 whether these variations are of importance in weather forecasting. 

 As we report these satisfying conclusions, it ought to be reported at 

 the same time that the financial support furnished the Astrophysical 

 Observatory by the Government would not have sufficed to obtain 

 these results without the aid of the Hodgkins fund of the Smith- 

 sonian Institution, and the generous financial support and wise 

 counsel of Mr. John A. Koebling. 



Although done a few days after the close of the period covered by 

 this report, it will be fitting in this connection to mention some pre- 

 liminary observations on changes of the appearance of the sun accom- 

 panying changes in its output of radiation. Being at Pasadena in 

 July, 1923, the director availed himself of permission given by 

 Director Adams to examine two years of record prints from direct 

 photographs and hydrogen (Ha) spectroheliograms of the sun made 

 at the Mount Wilson Solar Observatory. Four general rules or 

 principles seemed to be well established by the comparisons made. 



1. When increased sun-spot activity appears, either by new spot 

 groups forming on the visible solar disk, by the growth of spots 

 already present, or by the coming on of a new group due to the 

 solar rotation, then on that very day the solar constant value in- 

 creases. 



