22 Dr. J. Larinor on Electromagnetic Induction 



where R is equal to specific resistance divided by thickness, 

 so that p could be small, and therefore the screening nearly 

 complete with but a moderate amount of conducting-power in 

 the rarefied strata of the atmosphere. 



Although the sheet of air is very thick, yet it is still so thin 

 in comparison with the radius of the earth that the solution 

 applies to the problem by the principle of similarity, assuming 

 Ohm's law to hold for the current. Weighty evidence has 

 been brought together from different experimenters by Edlund 

 (Phil. Mag. 1880) in favour of the theory that much the 

 greater part of the resistance experienced by the electric dis- 

 charge in passing through vacuum tubes occurs at the sur- 

 faces of the electrodes, at which accordingly an opposing 

 electromotive force mav be supposed to act (vide C. F. Varley, 

 Proc. Roy. Soc. 1871; E. Wiedemann, Phil. Mag. 1880). That 

 being so, rarefied gases may well be fairly good conductors of 

 electricity, even when they refuse to allow a spark to pass 

 between electrodes, a view which is also adopted by Balfour 

 Stewart in his theory of the magnetic variations mentioned 

 below. 



It is well known that disturbances of the terrestrial mag- 

 netic force are largely connected with changes of the sun- 

 spots. There is also a daily variation in the magnetic elements 

 which follows the sun, but lags behind him by about 40°. 

 Lloyd and Chambers have shown that this is of such a cha- 

 racter that it cannot be accounted for by a direct inductive 

 magnetic action of the sun upon the earth, and in particular 

 that the lagging could not be so explained. Balfour Stewart 

 (Encyc. Brit., Art. " Meteorology ") gives reasons for assign- 

 ing the magnetic variations to the effect of currents actually 

 produced in the upper conducting strata of the atmosphere by 

 the circulation caused by the sun's heating and tidal action. 

 He also refers the large disturbances, amounting sometimes 

 to y^o of the total force, to the same cause; and as any cause 

 capable of such large effects in a short time may also be 

 assumed to be capable of producing the total magnetism of 

 the earth (cf. W. G. Adams, B. A. Report, 1880), he is inclined 

 to look upon the earth's magnetism as the accumulated effect 

 of such action. 



Now, even leaving out of account its high temperature, we 

 could not suppose the matter of the sun to be so strongly mag- 

 netized as to produce direct effects at the distance of the earth ; 

 but if the very considerable changes in the earth's magnetism 

 as well as in the long run its total amount, are to be referred 

 to the effect of convection-currents produced by the sun's 



