158 



ANNUAL KEPORT SMITHSONIAN INSTITUTION, 19 3 8 



being examined. If additional proof were needed for the explanation 

 of the changed frequencies, it may be stated that the double and triple 

 lines found in the spectrum of sunspots indicated that the light was 

 polarized just as in the case of the polarized light waves coming from 

 the laboratory source upon which the magnetic field is impressed. 

 Thus came the startling revelation that sunspots were not only terrific 

 hurricanes but every center was in itself a powerful magnet. Since a 

 magnetic field may exert a repulsing effect upon swiftly moving elec- 

 trons, we see some reason that charged electric particles can be actu- 

 ally hurled from sunspot centers at velocities which may carry them 

 through space into the earth's atmosphere, thus ionizing the upper 

 regions of the air in a way that would produce auroral displays. In 

 the light of such a mechanism, therefore, we see a possible reason why 

 aurorae occur in greater numbers and at greater brilliance at times 

 when these solar storms occur most frequently. 



Curiously enough, for nearly 300 years it has been known that the 

 direction of the compass needle and the intensity of the earth's mag- 

 netic field appear also to be affected by sunspots, for in the year when 



1840. 1850 1860 1870 1880 1890 1Q0O 1910 1920 1930 

 Figure 2.— Sunspots and variations in earth's magnetism fluctuate together through the years. 



sunspots are most numerous, magnetic disturbances are most fre- 

 quent and appear with marked intensity. 



If we examine a graph representing the sunspot numbers for the 

 last 100 years, we become aware of the close parallelism between these 

 magnetic variations on the earth and the variations in the numbers of 

 sunspots. Furthermore, it will be evident that the years when sun- 

 spots are most numerous follow with more or less regularity an interval 

 of somewhat over a decade between the times of maximum sunspot 

 activity. This solar cycle, or sunspot period as we sometimes call it, 

 is usually conceded to be on the average of about 11.3 years duration. 

 But an examination of the graph will show that sometimes the inter- 

 val between maxima is as little as 9 years and on one occasion was as 

 great as 17 years. We appear now to be in the midst of the period of 

 maximum activity of the present cycle. 



The source and nature of the earth's magnetism is still one of the 

 great mysteries of science. Although the close parallelism between 

 magnetic changes on the earth and the appearance and disappearance 

 of sunspots has been recognized for many years, it was not until the 



