44 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, IQO2 1903. 



by several simultaneously operating current-systems; it may moreover be the effect of far distant systems 

 that are not even always horizontal. The current-arrow is simply and solely a geometrical representation 

 of the perturbing force. 



With regard to the number of charts that should be worked out for each perturbation, it will be 

 a matter of opinion how many should be taken in each case. We have, however, throughout made it 

 a rule that for perturbations in which the perturbing force undergoes slow changes, the time between 

 each chart shall be longer than for those in which the perturbing force oscillates. 



By a comparison of the charts, a clear idea of the development of the perturbation will be obtained. 



In this way we can, however, represent the perturbation only for certain separate points. In order 

 to obtain a representation of the connected course of the perturbation, a plate will be drawn of 

 each perturbation, reproducing on a somewhat reduced scale the actual registered curves for variations 

 in H, D, and V. 



These copies of curves from all the observatories will be found all together, arranged according 

 to date, in front of the special treatment of the separate perturbations. 



To ensure the best possible result being obtained from this method, material should be collected 

 from a large number of stations distributed over all parts of the world. The best material for the 

 purpose would include registerings of all three magnetic elements from a ring of stations round both 

 poles of the earth, and a number of other stations more or less evenly distributed over the rest of the 

 world --as many as possible. 



We have no such material at our disposal. Our simultaneous observations of 1902 and 1903 are 

 all, with the exception of the registerings from Batavia and Christchureh, New Zealand, confined to the 

 northern hemisphere. In the arctic regions, moreover, we have observations only from our own four 

 stations; and although we think that these four stations were admirably situated for their object, yet the 

 material has not proved quite sufficient for a comprehension and elucidation of the perturbation-conditions 

 in the regions around the so-called auroral zone. 



In order to throw more light upon these conditions in the auroral zone itself, we have made a 

 special investigation of the conditions in these regions, and for this purpose have made use of the 

 material from the polar year, 1882 83. 



Our study of the universal character of the magnetic perturbations thus divides into two sections. 



The first section comprises the working-up of the material from 1902 and 1903. In the course of 

 this, an attempt is made, by the employment of the previously-cited method, to throw light both upon 

 the conditions in lower latitudes, and upon the possible connection of these conditions with the storms 

 occurring at the same time at our four stations near the auroral zone. 



The second section comprises an investigation by the same method, which is more especially 

 directed to the conditions in the arctic regions in and about the auroral zone. We have moreover, for 

 the sake of completeness, and in order to be better able to compare the results of these two sections, 

 also included in our investigations of the polar observations from 1882 83, observations from a few 

 stations that have a more southerly situation, namely, Christiania, Gottingen and Pawjowsk. 



14. On the Calculation of the Perturbing Force. For the calculation of the perturbing force, 

 there are registerings of the variations in horizontal intensity and declination, and for some stations in 

 vertical intensity also. When there are only the first two, only the horizontal component of the per- 

 turbing force can be determined. 



When no perturbations occur, the curves will have an even course, having only a slight bend 

 owing to the daily variation. If the curve has a marked divergence from this line, which must be 

 ascribed to the alteration in the magnetic constants, we then have a perturbation. 



