Mr. J. A. Broun on Terrestrial Magnetism. 95 



but it appeared to him that the coincidence then remarked had 

 not always existed. 



Having found that the magnetic disturbance was greatest not 

 only when the sun was on the equator, but also when the moon 

 was on the equator, I was induced to compare the times when 

 the lines of the moon's nodes and of the apsides would nearly 

 agree with each other, and the position of the sun at the equi- 

 noxes. I found that the longitude of the moon's ascending node 

 was 0° in 1838 and 1857, while it was 180° in 1848; and that 

 the moon was in perigee when its longitude was 0° in the same 

 years, which were those of maximum disturbance. The times of 

 revolution of the nodes and perigee are well known to be about 

 19 and 9 years respectively. 



If the coincidence of the decennial laws of frequency of the 

 solar spots and of magnetic disturbance be not accidental, we 

 should have a correspondence between the laws of the aurora 

 and of the solar spots. Do M. Schwabe's observations show 

 maxima of frequency at the equinoxes? If so, we must conclude 

 that the earth has something to do in the formation of the solar 

 spots, and therefore probably other planets, which would intro- 

 duce other periods. Jupiter, whose revolution occupies about 

 twelve years, was in longitudes of about 180° in the years 1826, 

 1838, and 1850. 



It is not the case that the magnetic distm'bance coexists 

 always with the spots ; but it is not improbable that during the 

 formation of the spots the disturbance is produced, that is to say, 

 at the period when the supposed discharge of the sun's electrical 

 atmosphere occurs. 



If the sun acts as a magnet, how does it produce the magnetic 

 variations? It is probable that it acts directly on the suspended 

 magnet, but that this action is quite secondanj. The diurnal 

 variations would be produced by a shifting of all the lines of 

 equal declination, inclination, and force usually represented as 

 due to the action of the terrestrial magnet. Are not the mag- 

 netic variations, then, due to the shifting of these lines, by the 

 inducing action of the solar magnet, the direction of the shift and 

 of the lines determining the epochs of maximum and minimum ? 



That the sun might act as a magnet is of course not a new 

 idea. It had occurred to myself whilst investigating the lunar 

 magnetic influence, twelve years ago, that some of the results 

 obtained might be due to the sun's rotation on his axis; a simi- 

 lar idea occurred to M. Kreil, but an examination of the results 

 seemed, in both cases, to prove the inaccuracy of the supposi- 

 tion. The results at which I have arrived lately, however, have 

 induced me to examine the question more carefully. If the sun 

 acts as a magnet, it is probable, from the analogy of our earth, 



