448 BIRKELAND. THE NORWEGIAN AURORA I'OLARIS EXPEDITION, 1QO2 1903. 



We find the perturbation-areas in which the negative equatorial storm is most distinctly apparent, 

 during the perturbation of the 313! October, 1902, and the 8th February, 1903 (see figs. 107 116, with 

 description in Art. 66 & 67; and figs. 87 & 88, with description on p. 189). In the latter case, however, 

 we have suggested, at the foot of p. 189, another possible interpretation of the field. 



The last cyclo-median type of perturbation, we have supposed would answer to effects of rays of a 

 degree of stiffness answering to the experiments shown in figs. 66 & 68, i 6. How the rays in these 

 triangular figures move, is indicated in the lowest of the three figures 71, and in fig. 72. These should 

 be rays that came comparatively near the earth in lower latitudes, and which formed fields similar in 

 form to these figures, that is to say spirals in which the direction of the current-arrows was anti-clock- 

 wise. We find similar spiral fields in the areas of convergence in the negative polar storms. In the 

 cyclo-median storms, however, the forces in low latitudes must be more powerful in comparison with the 

 forces in the polar regions, than in the negative polar storms. We have only a few instances of such 

 perturbation-fields that can be characterised as rather well defined. We believe the perturbation of the 

 6th October is a storm of which the field of force should be explained as the effect of such a cyclo- 

 median system. In our discussion of the compound perturbations, we have also several times come across 

 fields that would naturally be due to cyclo-median systems, but in which nothing certain could be decided, 

 owing to the complicated character of the storm. Fields of this kind are to be found in figs. 78 and 79, 

 for the 25th December, 1902, and figs. 87 and 88, for the 8th February, 1903. 



These five types of perturbation, however, as the above shows, must not be considered as com- 

 pletely separate phenomena. There will be a genetic connection between them, and this frequently finds 

 expression in the fact that when there are simultaneous effects of several systems, a change in one system 

 will be accompanied by a change in the other. This is especially distinct in simultaneous positive and 

 negative polar storms, but is also very prominent in simultaneous positive equatorial and polar storms. 



For all the perturbation-areas we have studied, a natural and simple explanation of the main features 

 has been found by the aid of these five types of perturbation. In addition to the direct magnetic effect 

 of these corpuscular systems, there will also be effects of simultaneously-occurring earth-currents, and 

 possible atmospheric ionic currents and secondary cathode rays. There seems to be no doubt that the 

 first of these exert a considerable influence, and we shall study them more closely in a later chapter; 

 but what effect the atmospheric currents might have is a rather more doubtful question. 



In Chapter IV of Part I, an estimate is made of the intensity of the corpuscular currents that appear 

 in the polar storms, and the amount of energy they carry. The making of such an estimate has been 

 made possible by the fact that we have two stations, Axeloen and Kaafjord situated one on each side of the 

 auroral zone, and that, as already mentioned, the current-systems form in the auroral zone, that is, between 

 the two stations. We have assumed that in the simplest perturbations the conditions up there can be 

 regarded approximately as effects of an infintely long, horizontal rectilinear current, situated between 

 these two stations. We can then, by the aid of the observations at the two stations, determine both 

 the strength of such a current, and its height. The question will indeed be over-determined, and we can 

 thus obtain a kind of idea of the approximation with which an assumption such as this can be employed. 

 In the simple cases that we have studied, the approximation, as a rule, must be considered as quite satis- 

 factory. We found the average strength of the current in the storms we investigated to be about io (i 

 amperes, and the average height about 400 kilometres. If we also used Dyrafjord and Matotchkin Schar, 

 we sometimes arrived at greater heights, up to more than 1500 kilometres; but we believe these are 

 probably due to the fact that our assumption in this case does not hold good. 



We have further examined into the amount of energy which these current-systems must represent, 

 and have come, by estimating, to figures such as about 2 X IC)7 h- P-i if we assume that the systems are 



