PART III. EARTH CURRENTS AND EARTH MAGNETISM. CHAP. I. 749 



This, as already pointed out, is also apparent from the vector-diagram for the 3ist March (fig. 270). 



It we endeavour to find corresponding serrations in the earth-current curves and the magnetic 

 arves, we can, as regards the declination, show a number in which the correspondence is quite 

 Uisfactory. 



In the curves for March 30 31, in Series II, f. i., we find a number of serrations in which the 

 irrespondence is comparatively good. 



The resemblance here is striking if D and the N S curve are compared during the time from 

 :)out 2 h onwards. 



The serrations in the E W curve, on the other hand, have no distinct counterparts in the mag- 

 nic ; but unfortunately we have no observations of H at this time. 



As regards D, we have compared with the earth-currents, in all, 107 serrations, which showed the 

 ratest similarity; and in this way we found more or less constant values for the relation P^jJe SN 

 > that here an elimination of the effect of the earth-currents could be made. 



With regard to //, we have only succeeded, during the same period of time, in identifying to 

 rrations with an}' certainty. 



THE INFLUENCE OF THE EARTH-CURRENT UPON THE VERTICAL INTENSITY. 



151. We have hitherto only considered the connection between the earth-currents and the varia- 

 ms of the horizontal magnetic elements. 



On looking at the vertical curves, however, we also frequently find very characteristic points of 

 -H mblance between them and the earth-current curves. 



In the case of Kaafjord, where the direction of the current is constant, it is easy to form there- 

 1 mi an idea as to the quarter in which the main mass of the current is to be found. 



Identification is very much more difficult here than in the horizontal elements, but if we look at 

 te curves for the I5th February in Series II, a close examination will reveal a number of small simul- 

 tieous deflections in the P-curve and the earth-current curves. An upward deflection in the K-curve 

 ; swers in every case to a downward deflection in the earth-current curves, and vice versa. 



As the sensitiveness for the vertical intensity is comparatively small, the resemblance in the small 

 (flections will be difficult to demonstrate, especially in the copied curves. In the original photographs 

 tt- identification is easier. 



In the stronger deflections the resemblance is more striking; and if we compare the course of the 

 i rtical curve at the times when the earth-current curves show considerable deflections, very characteristic 

 1'ints of resemblance will as a rule be found between the two systems of curves. 



In these powerful deflections, however, external current-systems will always exert a considerable 

 cect influence, so that the phenomenon becomes less perfect. We may here point to a number of the 

 i >re powerful deflections, which give a distinct impression of this resemblance. 



From Series I 



Jan. 13, time about 18'' 18'' 30'". 



18, 15'' 30 i6 h 3o m . 



Feb. n, i8 h 20 m . 



From Series II. 



Dec. 24, time about 15'' 45 & i8 h 30. 

 Jan. 5, i6 l '3o m 17'' 30. 



Birkeland. The Norwegian Aurora Polaris Expedition, 1902 1903. 95 



