460 



NATURE 



[Sei'TEmi;er 6. 



1900 



LA TITUDE- VARIA TION, EAR TH- 

 MAGNETISM AND SOLAR ACTIVITY. 



TN the Astronoinische Nachrichten (No. 3619) I have published 

 -'■ the results of an investigation dealing with the effects of 

 periodic changes in solar activity on the motion of our planet. 

 It is there shown that these changes, as indicated by the fre- 

 quency of sun-spots, exert a subtle bui pregnant influence on 

 the secular variations of the earth's elements ; and, moreover, 

 that disturbances precisely similar to those which appear in 

 the observations of the obliquity and of the sun's longitude are 

 distinctly exhibited in the variation of terrestrial latitude. 



In the further pursuit of these researches I have been led to 



Fig. I. 



(Fnm 'Ast Nacfi" 3649 ) 



A, frequency of aurorse (Loomis) ; bi, frequency of magnetic disturbances at Greenwich (Ellis); 

 B2, frequency of magnetic disturbances at Greenwich (Airy) ; c, frequency of aurorse (Europe 

 south of Arctic Circle) from 1812-1871 (Fritz), and in Scotland from 1865-1899. The "great" 

 period has been eliminated m both curves in order to exhibit more clearly the eleven-years' 

 fluctuations, i, semi-amplitudes of latitude-variation (Chandler, Nyr^n and Albrecht) ; 2, resi- 

 duals of obliquity (Fig. c in A.N. 3619) ; 3, Greenwich corrections to the R.A. of stars derived 

 from observations of the sun (Thackeray). 



magnetic strain in iron rods, it seems reasonable to believe that 

 something similar to the' molecular displacement in the rod may 

 take place in a magnetic body like our earth with respect to its 

 magnetic axis. As has long been known, this axis is by nt> 

 means coincident with the earth's axis of figure, but is inclined 

 to it, according to Gauss, at an angle of about 12°. It is there- 

 fore not unnatural to conclude that a molecular strain in the 

 direction of the magnetic axis will occasion an asymmetric 

 change of the earth's figure, and will thereby produce a displace- 

 ment of the axis of figure relative to the instantaneous axis of 

 rotation. Such a displacement could remain constant only so- 

 long as the total magnetic potential of the earth was not subject 

 to alteration, in which case the pole of rotation would describe 

 a circle of constant radius round the pole 

 ■4< tin/)': of figure as centre. But, as already stated, 



various facts compel us to believe that the 

 magnetic potential of the earth does alter, 

 and, indeed, changes synchronously with 

 the state of solar activity — the most strik- 

 ing instance being the regular increase of 

 auror?e and of magnetic disturbances with 

 the increasing frequency of sun-spots. 

 Now, if we consider aurorae as discharges- 

 of the earth's electric force, it follows that 

 the strain in the direction of the magnetic 

 axis should abate after an auroral display,, 

 and the pole of figure should therefore 

 approach the pole of instantaneous rotation.. 

 The outcome of this hypothesis would be 

 that changes in the state of solar activity,, 

 since they produce a measurable effect on- 

 the terrestrial magnetic forces, should also 

 be accompanied by corresponding changes 

 in the motion of the earth's axis. In A.N", 

 3649 I have endeavoured to test this 

 theory by such facts as have up to the pre- 

 sent been discovered regarding the com- 

 plex phenomenon of latitude-variation, 

 and I .shall here briefly describe the results- 

 there obtained. 



In the first place, from the material 

 afforded by the investigations of Loomis- 

 and Fritz, and from observations made in 

 Scotland during the past thirty years, I 

 constructed curves exhibiting the frequency 

 of auroral displays from 1S12 to 18991 

 (Fig. I, curves A and c.)' In these the 

 nfluence of the eleven-years solar period 

 is most conspicuous, but there is also ex- 

 hibited an intimate connection with the 

 "great" cycle of sun-spots. In contrast 

 with the spot-curve, minima of aurorre are 

 found to be of unequal heights. Accord- 

 ingly, if a second curve ss's" be drawn 

 through the aurora-curve A in such a way 

 as to bisect as nearly as possible every 

 wave appearing in it, this new curve will' 

 be seen to attain its maxima and minima 

 simultaneously with the "great" spot 

 period. It is therefore beyond dispute, 

 as already proved by Wolf and Fritz, that 

 the existence of this "great" period of 

 solar activity is an established fact, and 

 that it is clearly brought out by the curve 

 of aurorge. 



In the next place, curves were formed to- 

 show the variations in the frequency of 

 terrestrial magnetic disturbances. For this 



conclude that the anomalies existing in the observations of the I purpose I took the annual number of days on which such dis 



sun's right-ascensions and declinations are to be attributed 

 exclusively to changes in the position of the earth's axis of 

 rotation with regard to the axis of maximum moment of inertia, 

 and that these changes in their turn are intimately connected 

 with the varying display of forces on the solar surface. In a 

 subsequent article which appeared in Nature (No. 1584, 

 March 8) I made a suggestion as to the nature of this connection, 

 and advanced the hypothesis that the magnetism of the earth is 

 probably the medium through which the changes of solar energy 

 react upon the motion of the earth's pole. 



In view of the results of Joule's experiments regarding 



NO. 1 610. VOL. 62] 



lurbances were observed at Greenwich, as recorded by Mr. Ellis- 

 in Monthly Notices 60, December 18991 (Fig. i, curve Bj. ) 



Mr. Ellis subdivides his records according to the intensity of 

 the disturbance, but only three of his subdivisions were taken into- 

 account ; those, namely, showing the frequency of moderate, 

 active, and great disturbances respectively. For the satisfactory 

 combination of these different sets, the principle was adopted' 

 that the weight assigned to each set should be inversely pro- 

 portional to the mean frequency — unit weight being given to- 

 the "active" disturbances— on the ground that the small fre- 

 quency of "great" disturbances is probably compensated for byr 



