TKRRESTKIAL A[AGNKTISM. 



129 



of magnetic observation, indicating that at some high level the 

 air is in a peculiar electrical condition, in which its resistance 

 to the flow of currents is largely broken down. The air is said 

 to be ionized, and another result of this is the production of 

 ozone, which is known to exist in the upper atmosphere. The 

 absorption of very blue (or, rather, ultra-violet) light from the 

 sun seems to be the cause of the ionization of the air, so that, as 

 we naight expect, the atmosphere in the sunlit or " day hemi- 

 sphere of the earth is much more affected than that over the 

 dark hemisphere. Thus the lunar tidal action produces most 

 result over the former, which accounts for the observed fact 

 that the aforesaid magnetic variations (and the electric currents 

 in the upper air of which these variations are a direct result), 

 while governed, as to type, by the position of the moon, are yet 

 of greater intensity during the hours of solar day than of night. 



The magnetic variations which depend solely on the position 

 of the sun are produced in a similar manner, except that the 

 atmospheric motions also are in this case governed by the sun, 

 as well as the ionization. The atmospheric motions are due 

 partly to thermal and partly to tidal action on the part of the 

 sun. The two sets of electric currents, due to moon and sun, 

 flow in the same ionized layer, and undergo similar modifica- 

 tions owing to the seasonal changes of ionization resulting from 

 the varying incidence of the solar rays throughout the year. 



A further connection between the earth's magnetism and the 

 sun is revealed by the increase or decrease in the amount of these 

 magnetic variations with the increase or decrease in the dis- 

 turbance of the solar surface, as indicated by the number of 

 spots and prominences visible upon it. This solar disturbance 

 shows a fairly regular cycle of variation, with a period of about 

 11 years ; the spots and prominences are probably not to be 

 regarded as the main form of disturbance, but as merely sympto- 

 matic of great changes affecting the whole of the upper strata 

 of the sun. The emission of heat and light from our luminary 

 possibly shows small variations corresponding to the 11-year 

 period, but these are negligible in comparison with the great 

 variations which appear to occur in the emission of the ultra- 

 violet rays which ionize the earth's upper atmosphere. At 

 times of sunspot maximum the conducting power of the ionized 

 layer seems to be about double that at sunspot minimum, so 

 that the same atmospheric motions produce much larger mag- 

 netic variations at the one epoch than at the other. 



k; 



