42 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



The experiment in various forms has been repeated a great many times in the course of the last 

 few years, and I have succeeded in photographing all the light-phenomena appearing. The results of 

 these experiments will be fully described in a later part, and the light-phenomena illustrated by numerous 

 photographs. 



There are light phenomena produced by the rays that beat directly down upon the terrella, and 

 which, in my opinion, answer most nearly to the light-phenomena and certain magnetic storms in the 

 auroral zone on the earth. 



There are light-phenomena produced by rays made to fall upon a movable screen, for the purpose 

 of ascertaining how those rays behave that do not fall directly upon the terrella, but move about in its 

 immediate neighbourhood. I think that such rays can give a natural explanation of the cause of 

 certain universal magnetic disturbauces and sometimes to aurora polaris, if the ray-stream comes near 

 enough to the atmosphere of the earth. 



Finally, there is a flat, detached bright ring round the magnetic equator of the terrella, which 

 immediately recalls Saturn's ring. 



It seems as if this bright ring might bring us almost to the solution of other most important 

 terrestrial magnetic problems. 



In a lecture "On the Cause of Magnetic Storms, and the Origin of Terrestrial Magnetism" , given 

 before the Scientific Society in Christiania, on the 25th January, 1907, I gave a sketch of the results 

 of the terrestrial-magnetic investigations which will be produced in the present work. 



The conformity discovered by Sabine and others between sun-spots and magnetic perturbations, as 

 also aurora, has become apparent through observation and the summing up of a large number of single 

 phenomena. It must necessarily be supposed from this conformity, that also in single cases it must be 

 possible to prove a connection between these phenomena. This has often, especially in more recent 

 times, been observed in particularly marked cases. 



It will therefore be an important task to endeavour to discover the course of the process which 

 at times takes place in the neighbourhood of the sun-spots, and gives rise subsequently to aurora and 

 magnetic perturbations, and thus show that these terrestrial and solar phenomena are only different 

 phases in a continuous process. 



In order to solve this problem, one is naturally led to take one of two ways. The most rational, 

 if the necessary material were forthcoming, would be to start from the sun, where the process begins. 

 This is the way I have formerly taken. Starting with the hypothesis that the sun-spots are the source for 

 the emission of cathode rays, I have endeavoured to follow the process from the sun to the earth, and 

 by analogy with the above-mentioned experiment see how some of the rays strike the earth, and some 

 glance past it under the influence of terrestrial magnetism. This is moreover the way my friend, Pro- 

 fessor STSRMER, has taken in his mathematical investigations of the path of such rays from the sun to 

 the earth. He has published the complete results of his investigations in a special part of the present 

 work; but these results will already be to some extent known from his earlier papers. Here, for the 

 first time, a detailed mathematical treatment of the aurora problem and kindred problems will be found. 



The other way is to start with the conditions upon the earth, study a single perturbation, seek 

 for the terrestrial processes that might be able to influence them directly, and follow these up until, if 

 possible, we are stopped at the point when the cause can no longer be sought upon the earth, but in 

 the arrival of something from without; and here the two ways may meet. 



It is by going both ways, employing both methods, that we have thought we might have the best 

 prospect of solving our problem. 



That which, at a certain spot on the earth, and at a given moment, characterises a magnetic per- 

 turbation, is the strength and direction of the perturbing force. 



