2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 1 10 



Rayleigh proved that the scattering by particles (such as molecules 

 and very small dust particles) which are small compared to the wave 

 length of light is proportional inversely to the fourth power of the 

 wave length. Thus it happens that the sky is blue, because the blue 

 rays, being of shorter wave length than the red or yellow rays, are 

 much more scattered out of the direct sun beam by the molecules of 

 the atmospheric gases. 



If now, as stated above, the earth is being showered at times of 

 magnetic storms by multitudes of electric ions, which certainly are 

 small compared to the wave length of light, the direct sun beam, 

 shining 93 million miles through these showers, must be weakened 

 by Rayleigh scattering. The only question is how much. This paper 

 gives the results of an investigation of that question. 



Our first experience of such a phenomenon came to us in the year 

 1920. About March 20 to 23, 1920, there was an enormous sunspot 

 group central on the sun's disk, as shown in plate i. There was also 

 a severe magnetic storm on the earth, accompanied by fine auroral 

 displays. The storm was most severe on March 22 and March 23. 

 Smithsonian observations of solar radiation made at Calama, Chile, 

 followed the course shown in the upper curve of figure i. The 

 phenomena of central passage of the great sunspot group included a 

 diminution of the observed values of the solar constant of radiation 

 of the order of 5 percent, reaching the minimum value on March 23. 

 Possibly the very low value of March 23 may have been made un- 

 duly low by experimental error, but the value of March 24, nearly 

 as low, is of quite as high a grade as most of the Montezuma values 

 of that year. 



Critics may suggest that these low values of the solar constant 

 were caused, not by Rayleigh scattering from electric ions along the 

 93-million-mile path of sun rays through space, but rather by a hazing 

 of the earth's atmosphere, produced by the adherence of water-vapor 

 molecules to the ions, after they entered the atmosphere ; in other 

 words, that the solar-constant values were erroneous. This sugges- 

 tion, however, runs counter to the observations. For though the 

 lower curve in figure i, which traces the march of values of atmos- 

 pheric transmission coefificients for green light at wave length 0.5 11 

 micron, does show that the atmosphere became less transparent ^ 

 during the passage of the sunspot through the central position on the 

 sun's disk, that change alone, if it were not countered by other factors 

 in determining the solar constant — factors also affected by atmos- 



1 This change will be accounted for when we consider figure 3 below. 



