164 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1913. 



The corona during the same period always undergoes a definite 

 evolution. Toward the epoch of sun-spot minimum the polar rays 

 are fine and vertical like the bristles of a brush. The jets in the mid- 

 dle and mean latitudes are much longer and bent toward the Equator. 

 At the maximum period there is little difference with the latitude. 

 During the transition years the poles and the Equator are almost 

 clear and the rays are developed only in the middle latitudes, giving 

 the whole a rectangular appearance. 



The more we reflect upon these facts the less are we led to regard 

 the sun as a monarch, inaccessible, and shut up in a tower of ivory. 

 It, like the earth, must have seasons connected with the revolution of 

 the planets and tides connected with its own rotation. To sift out at 

 least the more active of these external influences is a legitimate task, 

 even though it is not an easy one. 



First, do we find one or several bodies which could be held responsi- 

 ble for a cycle of 11 years? The stars seem to be beyond considera- 

 tion, since in that period there is no appreciable change in their 

 linear or angular distances. 



We could, as did John Herschel, blame one or several swarms of 

 meteors, imagined for the purpose. Describing very eccentric orbits, 

 they might graze the surface of the sun, causing the spots. Suitably 

 choosing their revolution periods, inclinations, eccentricities, and the 

 distribution of the matter in their orbits, we could explain the 

 phenomenon in all its details. We must confess that the permanence 

 of swarms of meteors put every 11 years to such a violent test does not 

 seem probable. There is no doubt that meteors fall into the sun in 

 great numbers. But we have no direct proof that this happens 

 periodically and so as to produce visible effects. Such proof we feel 

 that we must demand for this very supple and convenient hypothesis. 

 As these swarms have not been detected, we must leave them and 

 direct our investigations to the planets. 



The most important of these planets brings a coincidence at first 

 sight very seductive. Nearly every 11 years Jupiter, in a determinate 

 sense, crosses the plane of the solar equator; also in every 11 years 

 the numerical predominance of the spots passes from the northern 

 to the southern hemisphere of the sun. The same interval separates 

 the return of Jupiter to its least distance from the sun and the return 

 of the sun-spot numbers to their extreme value. 



We must not hurry, though, to sing our victory. It is not an 

 approximate concordance but a precise one which we should demand. 

 The periods in years are 11.86 for the revolution of Jupiter and 

 11.13 for the sun-spot cycle. For the second period, w^hich is less 

 well defined, the incertitude is in the hundredths. For more than a 

 century we have careful records of spot numbers which reappear 

 regularly. Now, in the course of a century the difference of eight 



