474 Royal Society : — 



ing strata of the earth's surface, just as a current can only move 

 backwards or forwards along an insulated wire. Indeed a little re- 

 flection will show us that earth-currents are not local phenomena ; so 

 that if we endeavour to estimate quantitatively their influence on the 

 magnet according to the hypothesis of direct action, we must first ex- 

 tend our field of observation, and obtain their value in other countries 

 besides our own. 



But to return to our comparison of curves. We see that for the 

 greater part of the disturbance both the horizontal force and the 

 declination were very much above their normal lines, while on the 

 other hand the currents were frequently crossing their zero lines, — and 

 that both currents were simultaneously and for a long portion of 

 their time very near zero, although during this time the magnetic 

 disturbance was considerable. 



Next, with respect to a very abrupt disturbance which commenced 

 about 8 b 50 m , the corresponding earth-current curves are exceedingly 

 powerful, alternately passing and repassing the zero line to nearly 

 the same distance on both sides, while the absolute disturbance of the 

 horizontal force, and probably of the declination, was not very great. 



We have thus, in the first place, a very sluggish action of earth- 

 currents, while the magnetic disturbance was considerable, and in 

 the next place a very violent action of the former when the absolute 

 disturbing force was by no means excessive ; and we may add that 

 at 9 h 50 m both earth-currents were near zero, while both elements of 

 the earth's magnetism were much disturbed. 



For all these reasons this comparison of the curves is unfavourable 

 to the hypothesis of direct action. 



Let us now consider the other hypothesis, or that of induced action, 

 and we shall find the following points in its favour. 



1st. That in this disturbance, for at least one hour, both elements 

 of the earth's magnetic force remained at a considerable distance 

 from their normals, and that during this time the earth -currents 

 observed were exceedingly small. Now, on the hypothesis of induced 

 action, the earth- current effect depends not on the absolute value of 

 the disturbing force, but on its rate of change ; and if during this 

 period we examine the horizontal-force curve, we shall find the 

 abruptness of change not so great as in those cases where greater 

 earth-currents were produced, while in the declination-curve the 

 abruptness of change during this period is exceedingly small. 



2nd. A reference to the diagrams will show us that in general the 

 most abrupt magnetic disturbances are those which are accompanied 

 by the greatest earth-currents, and that in particular a very abrupt 

 disturbance, which took place about 8 h 50 m , was accompanied by very 

 strong earth-currents, alternately positive and negative, those of the 

 one name being nearly as powerful as those of the other, while on 

 the other hand the corresponding magnetic disturbances were on an 

 average decidedly on one side of the normal lines. 



On the other hand, the following fact seems at first to tell against 

 the theory of induction. An inspection of the curves will show that 

 we have currents remaining on one side of the zero-line for some 

 length of time, during which the magnetic disturbances have never- 



