792 B1RKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



Examples of the other type, besides those already mentioned, are found in the storms of 



February 26 27, 1893, 

 January 27 28, 1893, an ^ 

 August 2526, 1895. 



In several of these, however, the phenomena seem to be of a very mixed character, and still 

 more so in a number of other storms, e. g. of 



November 3 4, 1889, 

 March 5 6, 1893, 



January 7 8. 1895. 

 We thus see that the earth-current conditions in these districts exhibit throughout exactly the same 

 chief peculiarities as in Germany and France; but at the same time the cases that we have characterised 

 as exceptional may possibly occur somewhat more frequently here. 



This, however, only agrees with what, according to the above, we should consider probable. 

 The districts in which the observations were made here, have of course a somewhat more northerly 

 position magnetically than the two corresponding districts in Germany and France, and a removal in a 

 direction N S in relation to the points of convergence of the perturbation-systems, the respective 

 vortex-centres of the earth-current systems, must be assumed to bring about just such deviations as we 

 have here observed. 



In reality, these earth-current conditions at Greenwich may be regarded as an indication of a 

 change from the conditions in Germany to the current-conditions in the auroral zone. In these last 

 districts we have seen that the conditions are practically always as in the above-mentioned excep- 

 tional cases. 



EARTH-CURRENTS AT PAWLOWSK. 



160. Some examples of earth-current registerings and simultaneous magnetic registerings of tun 

 magnetic storms were also sent us, as already mentioned, from Pawlowsk; but as we have had no 

 opportunity of going through the greater number of these, we have been unable to form any well- 

 grounded opinion as to the nature of the conditions here. Two of the sets of curves that have been 

 sent us show conditions during rather powerful storms, and the curves are of so jagged and disturbed 

 a character that it is very difficult to follow them. Local disturbances also seem to exert a great in- 

 fluence. It was our intention, however, to give a reproduction of the third storm sent, namely, that of 

 March 17 18, 1889. Here too there are great local disturbances, but nevertheless the principal course 

 of the curves can be clearly followed. Unfortunately it appears at the last moment that the original 

 curves are missing. 



If these curves are considered from the same points of view as before, it will be easily discovered, 

 on looking at the course of the curve as a whole, that there exists an approximate proportion between 

 the N S and E W curves, and the rate of change in the D and H magnetic curves respectively. 

 Changes in the direction of the brief deflections in the earth-current curves correspond in time with tin- 

 extremes of the magnetic curves; and the most powerful deflections in the earth-current curves take 

 place simultaneously with the greatest variations in the magnetic curves. Thus the induction-phenomenon 

 comes out clearly on consideration of both components. 



The two components of the earth-current exhibit a fairly strong resemblance; but the direction 

 appears to be a little more variable than in Germany. 



Here too, the same- remark may be made as before, namely, that the resemblance between the 

 two sets of curves becomes quite striking if the earth-current curve is moved slightly along the time- 



