194 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 193 5 



years, and about 6 years afterwards no spots may be seen for many 

 days running. Later still the spots will become more frequent and 

 in another 5 years or so will have reached their maximum. Sun 

 spots are carried across the sun's disk as the sun rotates on its axis, 

 so that a spot which lasts sufficiently long will reappear every 26 

 days, this being the time taken by the equatorial part of the sun 

 to rotate, as seen from the earth. 



The connection between the irregular variations of the earth's 

 magnetism and sun spots is seen in the fact that magnetic disturb- 

 ances tend to recur at intervals of 26 days. Also the number of 

 magnetic disturbances increases during those years when there are 

 many sun spots. Certain individual magnetic disturbances seem to 

 be associated with definite spots. 



The regular changes in the earth's magnetic field also show a con- 

 nection with sun spots. Thus, the average amplitude of the regular 



1840 1850 J860 1870 1880 



FiGUKB 7. — Relation between terrestrial magnetism and sun spots. 



The continuous line shows the changes In the mean annual sun-spot number, while the 

 dotted line shows the mean annual range of the daily swing of the compass needle. The 

 latter may be taken to indicate also the conductivity of the upper atmosphere. Note 

 that the 11-year period In both cases is marked, but not quite regular. Note also the 

 close connection between the two phenomena. 



daily swing of the compass needle for any year is most closely asso- 

 ciated with the average number of sun spots for that year. We 

 believe that the regular diurnal magnetic variations are connected 

 with changes in the conductivity of the upper air and that this 

 conductivity is due to the action of the sun's ultraviolet light on the 

 upper air, whereby free electrons are produced. There is a small 

 lunar diurnal variation of the earth's magnetic field which is appar- 

 ently due to tides set up by the moon in the upper air. These tides 

 should have no effect on the magnetic field unless the air is a con- 

 ductor, and it has been shown that during the night when sunlight is 

 cut off, the lunar tides have no effect, their presence being seen during 

 the hours of daylight. 



Results from radio Tneasurenients. — While the study of terrestrial 

 magnetism gave the fir^t indication that the upper air was an elec- 

 trical conductor, our knowledge has been rapidly extended in recent 



