4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 1 10 



Specifically corrected for absorption of atmospheric ozone, as ex- 

 plained in volume 5 of Annals of the Smithsonian Astrophysical 

 Observatory, pages 124 to 131. Hence this suggestion of critics 

 applies neither to the v^ork of 1920, nor to the work subsequent to 

 1923. 



In conversation with Dr. Nicholson of Mount Wilson Observa- 

 tory, in September 1947, I asked him if he knew of other occasions 

 when great sunspot groups passed centrally through the sun's disk. 

 If so, I proposed to see if a similar depression of solar-constant 

 values occurred. In reply he suggested that sunspots were "like shot- 

 gims, rather than like rifles," when they pepper space with electric 

 ions. Hence it might well be that whenever a severe magnetic storm 

 occurs there will be generated a shower of ions embracing our line 

 of sight and introducing Rayleigh scattering through the 93 million 

 miles of space between the earth and the sun. 



MAGNETIC STORMS, 1923 TO 1946 



I undertook to test this hypothesis. The phenomena of March 1920, 

 are so exceptional that I omitted them in a general tabulation. From 

 the journal "Terrestrial Magnetism" I found over 70 occasions in 

 the years 1923 to 1946 when very severe magnetic storms were re- 

 ported. Not being very familiar with the terms used by the observers 

 at the magnetic stations, I am not sure that I found all the dates of 

 severe magnetic storms during this interval. Moreover, the magnetic 

 observers, if they see this paper, may not regard all the storms I 

 selected as severe. There was, indeed, some discrepancy between the 

 estimates of severity from different magnetic stations reporting in 

 "Terrestrial Magnetism." Whatever may be the incompleteness or 

 inexactness of my selection, I feel sure that experts will agree that all 

 the storms included in table i, which follows, were strong, if not 

 always deserving the description severe. 



The magnetic storms continued from 2 to 10 days. It was often 

 uncertain which day to take as representing the height of the storm, 

 that is, the day most likely to be the day when the shower of ions was 

 densest. All the several tabulations of the data which I made showed 

 clearly a depression of the solar constant at or near the height of the 

 magnetic storm. Hence I thought it fair to select as zero day that day 

 during the height of the storm when the solar constant was most 

 depressed. 



SOLAR-CONSTANT OBSERVATIONS 



Unfortunately I could not use all the storm dates selected. Our 

 observations of the solar constant made at Montezuma and at Mount 



