J863.] 667 



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 sero-line for some 

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

 theless changed in both directions. When, however, we reflect on 

 this circumstance, we are led to see that since we have two sets of 

 disturbances taking place simultaneously, so we must also have two 

 sets of earth-currents. Now one of these disturbances, which we may 

 perhaps call the soft-iron one, reproduces those small and rapid 

 changes which take place in the primary force, while on the other 

 hand the hard-iron disturbance averages these small changes and 

 presents us with a disturbance- wave of long period. Precisely, then, 

 as in the magnetic curves we have waves of short period superim- 

 posed upon waves of long period, so will it be in the earth-current 

 curves. Those currents due to the soft-iron disturbances will be 

 superimposed upon those due to the hard-iron ones, with this 

 difference, that we are not entitled to assume that the proportion in 

 intensity between the two simultaneous earth- currents must be pre- 

 cisely that which exists between the rates of change of the two cor- 

 responding simultaneous disturbances. It will be apparent that this 

 feature of duality ought also to be presented by the aurora ; and here 

 it is well known that we have at least two phenomena, one of a more 

 fitful and the other of a more permanent character, namely, the 

 streamers and the auroral arch. We may suppose the first of these 

 phenomena to correspond to the soft-iron, and the second to the hard- 

 iron disturbances. Indeed it is questionable whether the different 



