PART 111. EARTH CURRENTS AND EARTH MAGNETISM. CHAP. I. 785 



If we compare the curves for the two earth-current components for November i 2, we find 

 :hroughout the most striking resemblance between the two curves. Every single jag and deflection in 

 :he one curve is accompanied by so exactly corresponding a deflection in the other curve, that by 

 altering the sensitiveness we should be able to get all the briefer deflections to become very nearly 

 dentical. [The pait of the curves just after 20'' (Gr. M. T.) answers to a time-mark, the earth-current 

 icre being interrupted for about 5 minutes in both lines (not exactly simultaneous)]. 



In other words we here find again the same peculiarity in the earth-currents that we found 

 U Kaafjord. 



We shall also find the same thing on looking at the perturbation on the 5th November, 1883; 

 )ut in this case the curves are not photographically reproduced, but are drawn with Indian ink, as 

 he originals were too faint and rubbed out. 



If we look at the remaining curves from which we have simultaneous registerings in the two 

 >arth-current components -- which we have not reproduced here --we make everywhere the same 

 ibservation. 



From this we may conclude, as at Kaafjord, that the earth-currents in the district here observed, 

 ollow very nearly the same direction in the earth. As a consequence, however, the one component in 

 lie brief variations -- with which we are principally concerned -- will be sufficient to characterise the 

 nurse of the earth-currents. The want of the second component is therefore not of great importance. 



We have at the end of the present volume reproduced a number of examples of various typical 

 Magnetic storms with their attendant earth-currents, from 1883. 



On looking at these, several things are at once apparent. In the first place there is always a great 

 esemblance between the course of the earth-current curve and the Z>-curve in nearly all details, which 

 eems to indicate that the latter component is strongly influenced by earth-currents; but on the other 

 land we very often meet with conditions that indicate induction-currents. This is most noticeable in the 

 implest polar storms. The following are some examples where the conditions are especially distinct: 



1883, Nov. 5, Dec. 9, March i, Oct. u, Nov. 28 29. 



While in these cases the deflections in the D-curve as a whole increase comparatively evenly to 

 maximum, only to decrease once more to o, a change takes place in the earth-current. 



At the beginning of these perturbations the current flows in one direction, then turns, and during 

 he last half we find the direction to be the reverse. 



If, however, we examine the time of the change in the earth-current curves, we find, that it does 

 >ot as a rule coincide with the time of the maximum of :he deflection in D. 



The reason of this, however, is easy to demonstrate. We need only look at the curves for the 

 torm of the 5th November. 



It is easy to prove the presence of a number of small serrations both on that day and on other 

 ays on which we see without doubt the effects of almost exclusively earth-currents. With their assi- 

 tance we can now, as before, eliminate the effect of the earth-currents, leaving only the direct effect of 

 ic outer system. 



In the case of the D-curve the agreement is so distinct that it presents no special difficulties. It 

 ; often difficult, however, to measure the small serrations in the magnetic curves, as we have only 

 lue copies, in which, as a rule, the small details are not at all sharp. We have determined the rela- 

 on between the deflections in the D-curve and the two earth-current curves. 



In one day the figures found exhibit a (fairly) satisfactory constancy; but from one day to another 

 ic conditions vary somewhat. In this way we found the following figures: 



