PART I. ON MAGNETIC STORMS. CHAP. II. 



79 



perturbation might not also be explained by a polar current-system. Considering that the perturbation 

 may be due to currents of a cosmic nature that approach the earth under the influence of terrestrial 

 magnetism, there would be a possibility of the existence of current-systems that consisted of current- 

 spirals, which stretched down at the poles, and in this way acted as though magnet poles were put down. 

 Poles such as these, however, though they might explain the principal features in the form of the field, 

 would not be reconcilable with the fact that the force increases towards the equator. 



We are therefore of necessity led to seek the explanation in currents that have their greatest den- 

 sity in low latitudes near the magnetic equator. We thus naturally come to consider the two possi- 

 bilities -- the perturbation either has its direct cause in currents at the surface of the earth, or in cur- 

 rents above the earth. 



It seems hardly likely that the phenomenon is due to earth-currents. These currents, it is true, 

 would explain the small vertical intensity as regards magnitude, as it might be assumed that the current 

 was distributed over a large portion of the earth's surface; but a wide-spread system of earth-currents 

 such as this would hardly explain the other properties of the perturbation. The direction of the earth- 

 currents must, in such a case, be from east to west, the reverse of the direction of the current-arrows 

 marked; and it would then be difficult to explain how the force P has a component directed upwards 

 north of the equator, and downwards south of the equator. Such earth-currents, if, as independent pheno- 

 mena, they are to be able to explain the perturbations, cannot be induced currents, but must depend 

 upon conditions in the earth itself. As, however, the direct cause must be sought in processes in the 

 earth itself, it is incomprehensible how these currents can have so universal a character, and main- 

 tain so constant a direction with so singular a form. It seems especially impossible to explain the 

 simultaneous serrations; for the perturbing force would then at each place principally be determined 

 by that part of the current that passed beneath the place. From a physical point of view there are 

 greater difficulties in assuming that different parts of a wide-spread current-system such as this, which 

 should have its direct cause in the earth itself, should act rhythmically, and that the alteration of current- 

 density with the latitude at each point of time should take place so regularly and connectedly. The 

 question might, indeed, be settled, if they were surface-currents, by looking at the registerings for the 

 earth-current. If the perturbation were conditioned by surface-currents on the earth, the curve of the 

 earth-current should exhibit a course similar to that of the curve on the magnetograms. If, on the other 

 hand, the perturbation is due to currents lying outside the earth, the curve for the earth-current will 

 look like vibrations about the normal line, as the rapid changing in the perturbing force would produce 

 corresponding induced alternating impulses. 



We have no complete set of earth-current registerings, however, for any station except Kaafjord. 



Here, indeed, we do find that the earth-current curves are of the character described. They are 

 undoubtedly for the most part induced currents, but their direction is mainly determined by the local 

 conditions, as for instance the conductivity of the soil in the various directions. 



When the great perturbations show maximal deviation, the earth-currents usually pass a value. 



As we shall see later on, it is easy to reconcile the existence of such conditions in the polar 

 regions with the fact that certain magnetic disturbances in southern latitudes, far away from the storm- 

 centre, may often in great part be caused by earth-currents. 



The earth-currents will be treated in a subsequent part of this work. 



We have already mentioned that this equatorial perturbation often comes as a precursor of polar 

 storms; and indeed, we have really never met with an entire perturbation of this kind with which there 

 have not, within the same period, been polar storms. The necessary consequence of this must be that 

 these two kinds of perturbations should be closely connected with one another. Now there is no doubt, 



