8o 



BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



as will be shown later on, that the polar storms are due to currents above the earth ; if so, this should 

 also be the case with the equatorial perturbations now under consideration. 



According to this, we must necessarily seek the cause of the perturbations in currents above the 

 surface of the earth. If the current is to be sought at a distance from the surface of the earth that is 

 small in comparison to the earth's dimensions, we must, in order to explain the field, have a wide- 

 spread plane current circulating round the earth. Our being obliged to have a wide-spread plane 

 system is a consequence of the fact that otherwise the fields would be limited more rapidly. If, as is 

 the case, the effect is extended to all parts of the earth, there must also be currents in those regions. 

 A system of this kind, however, if it is to satisfy the actual conditions, is inadmissible; for we meet 

 here with difficulties similar to some of those in the way of the acceptance of the earth-current theory. 

 The first of these is that the relative strength of the current in the various districts of the earth 

 should remain fairly constant throughout long periods, notwithstanding that the field, as already men- 

 tioned, is remarkable for great variableness in strength : the variations take place in all districts in about 

 the same proportion. It seems, moreover, impossible, if we are not to have recourse to the mysterious, 

 but keep to the well-known possibilities of physics for the production of cosmic currents, to have the 



stability of the current explained; for the 

 current, as we know, if composed of free 

 portions, is deflected by terrestrial magne- 

 tism, the separate bearers of the electric 

 charge whatever the physical nature of the 

 latter moving in spirals about the magnetic 

 lines of force, or being carried out into 

 space, if the corpuscular current-rays arc 

 stiffer. 



The only possibility then left is that the 

 positive equatorial perturbations are due to 

 the effect of a current-system, whose distance 

 from the earth is of the same order as the 

 dimensions of the earth. Owing to the distri- 

 bution of force in the field, and the symmetry 

 that is found, as a rule, with regard to the equator, this current, as already mentioned, must have its 

 greatest effect about the plane of the equator; and on account of the direction of the perturbing force, 

 the current-lines, at any rate in the region nearest the earth, must lie in planes that are approximately 

 parallel with the plane of the magnetic equator. 



There are still two essentially different cases possible here, 



(1) that the current passes round the earth, and 



(2) that the earth is quite outside the system that in the main conditions the perturbation. 

 When, on account of the field, the currents must be sought at so considerable a distance from 



the earth, we are compelled, with the knowledge we at present have of the physical qualities, to assume 

 that these currents are corpuscular in constitution. The systems that may then be formed must be such 

 as may arise when a magnet is subjected to corpuscular electric radiation of some kind or other. 



In order to become better acquainted with the systems that may arise under these conditions, a 

 little attention should be given to the experiments I have made, in which a magnetic terrella is ex- 

 posed to cathode rays. These will be fully treated in Volume II, and illustrated by numerous photographs; 

 but even here we will draw attention to a few important circumstances. 



In addition to the polar precipitations there are still in particular two characteristic phenomena. 



Fig- 37- 



