PART 1. ON MAGNETIC STORMS. CHAP. II. 



105 



If we look specially at the perturbation under discussion, we see, true enough, that the vertical 

 components at the Norwegian stations have about the same magnitude as the horizontal component. 



The conditions at these stations at i a.m. have already been mentioned. From these it appears 

 that the total perturbing force at Kaafjord, Axeleen and Matotchkin Schar may be explained as the effect 

 of a galvanic current, which drops at a certain angle towards the earth in a direction from Axeleen towards 

 Dyrafjord. The current here is so near the stations, that the nearest part will be the important part. 

 We make use of the law that when we approach an infinitely thin conductor, in which a stationary 

 current is flowing, the effect will be approximately that which would be obtained if the system were 

 replaced by an infinitely long current of the same strength, which passed through the nearest point on 

 the conductor. 



The conditions which we have educed from our current-system for the vertical components in more 

 southerly latitudes, are corroborated in a striking manner by comparing the conditions at the few other 

 stations from which we have received the vertical curves for this perturbation. In accordance with our 

 hypothesis, the vertical components in these latitudes are very small in comparison with the horizontal. 

 For instance, at i 1 ' and i h 15, P t for Pawlowsk and Ekaterinburg is imperceptible, whereas at the 

 same time, in the case of Val Joyeux, which is situated nearer the point of convergence, the oscil- 

 lation in the vertical curve is distinct, although faint, and answers to a perturbing force directed 

 upwards. 



Now when the current-system moves towards ENE, we should expect that the vertical intensity 

 would also become noticeable at the two first-named stations, since, by the movement, they would be 

 brought into the area in which vertical components might be expected. This is confirmed by the actual 

 circumstances. 



In the following charts, we find a noticeable vertical component for Pawlowsk and Ekaterinburg, 

 while at the same time it diminishes in the case of Val Joyeux, but is directed upwards in all three. 

 As the effect is so limited, and the vertical components so great, the width of the current must 

 be small in proportion to, for instance, 1000 kilometres. I have supposed a maximal width of 500 km. 

 in my report, "Expedition Norvegienne", etc., 1. c., p. 26, although it is probable that the boundary is 

 not sharply defined. It must therefore be understood that it is the main body of the current that has 

 this narrow width. 



It follows from the cosmic constitution of the whole current, that the form we have assumed for 

 the current-system that shall be able to explain the field, is only an ideal form, which in its main 

 features characterises the system; and further it is to be understood that it is the total effect outwards 

 that in its principal features is explained by a system such as this. It does not follow from this, 

 of course, that the trajectories of the separate corpuscles must coincide with the direction of the as- 

 sumed system. 



The field at each place is in reality the sum of 

 the magnetic effect of the separate corpuscles at each 

 moment. 



It is evident, both from my experiments and 

 Stermer's calculations, that a drawing-in of rays generally 

 takes place over areas of greater or less extent; and 

 we will here only suggest that the effect of a bundle 

 of rays in which the course of the rays is, as shown 

 in fig. 50 a & b, very near, will be the same as that of 

 a linear current consisting of two vertical currents con- 

 nected by a horizontal one. Fig. 50. 



Birkeland. The Norwegian Aurora Polaris Expedition, 1902 1903. 14 



