I'AKT II. POLAR MAGNKT1C P1IKNOMKXA AND TKRRKI.l.A KXPFRIMK.N TS. CHAP. III. 521 



THE SEAT OF THE RADIANT SOURCE. 



97. The eruptive character of the occurrence of magnetic storms, indicates that the period might be 

 explained by a periodic change in the intensity of the source, just as certain periods have been found 

 for the eruption of geysers. 



But such an explanation cannot be maintained; for, owing to the rotation of the sun, the radiation 

 would have to issue from a number of sources, and it is hardly conceivable that a large number of 

 sources would vary with the same period and be in the same phase. 



The only possible explanation left, seems to be that the period of storminess is the synodic period 

 of revolution of some layer of the sun. From this view it follows, that if the storminess is to show only 

 :>ne distinct monthly maximum, there must be a fairly limited region of the sun, the activity of which 

 as a radiant source, is predominant, and we see from the curves fig. 190 that it must maintain its pre- 

 dominance during several revolutions of the layer to which the source is attached. 



If such a source on the surface of the sun gave out electric radiation from a surface element 

 according to the same law as for the radiation of light, the storminess due to such a source ought to vary 

 approximately according to a sine or a cosine law, or 



S T = A sin "in -_ 



in which T is the time of revolution of the source ; and in which it is to be remembered that only positive 

 values have a physical interpretation. We should get a number of separate waves according to this 

 sine law, the effect of the predominant source would be felt during half the period, or 14.6 days. I he 

 effect would increase somewhat rapidly, but in the neighborhood of the maximum it would keep nearly 

 constant for several days. 



The curves of storminess, however, show a very marked difference from the sine form. The effect 

 of the predominant source, far from being felt during half a period, is generally only felt for a few 

 days at the time of the maximum, which occurs suddenly, and assumes a very pointed form. 



How can this discrepancy be explained? There are only two possible explanations, 



1 1 1 That the suddenness is due to an eruptive character of the source, or 



(2) That the radiation is greatly predominant in certain directions. 



In view of the violent changes observed in the upper layers of the sun, great and sudden changes 

 in the ray-emission will probably take place and influence the character of the phenomenon; but such 

 changes alone are insufficient to explain the character of the variation in storminess. Above all, it can 

 hardly account for the comparative regularity with which the maxima occur. 



On an average., the source must be quite as active when it is turned away from the earth as when 

 it is turned towards it. If the maxima were solely determined by the eruptive changes, there would be 

 far greater changes in the length of the period of storminess than are actually observed. We see, from 

 the curves, that the periods within each group of maxima only show comparatively small differences. It 

 is, of course, conceivable that there might be a period of variation of the source, which could produce 

 the observed effect; but from a physical point of view it is scarcely probable that a period of eruption 

 would be so regular, and farther that it would coincide with another quite independent period - the 

 period of the sun's rotation. 



On the other hand, the second assumption, namely that the radiation is greatly predominant in 

 certain directions, gives at once a simple explanation of the variation of storminess. 



According to this view, the radiation would be mainly restricted to certain narrow pencils starting 

 from the source. 



When the earth comes sufficiently near to such a pencil, there will be a perturbation. 



