PART I. ON MAGNETIC STORMS. CHAP. II. 



99 



electric currents. Starting from physical considerations, we are thus naturally led to seek to explain the 

 field by a system, which, in its average effects, has the character of two vertical currents in opposite 

 directions, connected by a horizontal part. 



In their main features, the conditions for such ray-currents can approximately be settled, as there 

 is a long series of experimental and theoretical investigations on the course of cathode rays in a magne- 

 tic field. It will be sufficient for our purpose to refer to papers by POINCARE( I ), myself( 2 ), 

 and ViLLARD( 4 ). 



In accordance with the facts learned from the above-men- 

 tioned papers, I have here put forward a hypothesis regarding 

 the course of the rays in the vicinity of the earth, by which, 

 as it will be seen, the magnetic fields of force observed during 

 magnetic storms are explained in a simple manner. 



Figure 48 illustrates by diagram this hypothesis, which 

 is to the effect that the rays which are drawn in towards 

 the earth in the sharply wedge-shaped space in the polar regions, 

 always whirling around the magnetic lines of force, (fig. 48 a) 

 either, as generally happens, pass the earth with an average 

 curvature such as is shown by the curve b, or, less frequently, 

 with a loop such as curve c shows. 



In those regions of the earth in the auroral zone, that 

 lie close beneath the rays, the rays in the lowest bend of the 

 curves b and c will mainly condition the magnetic disturbances; 

 and the perturbing forces produced will be in reverse direc- 

 tions in the two cases. This will mean that the current-arrows 

 for this area will generally point from east to west along the 

 auroral zone (answering to the form of curve b), while less 

 frequently the reverse direction may occur (corresponding to the 

 form of curve c). 



In the equatorial perturbation of the gth December, 1902, 

 it is mentioned that the direction of the polar storm that finally 

 supervenes, is the reverse of our ordinary polar night storms. 

 We thus have before us a field that can be explained by a 

 current-system, the effect of which is the same as that produced 

 by a linear current of about the same form as the loop in fig. 48 c. 



We shall farther on meet again and again with these reversed polar storms. Fields similar to that 

 of the gth December will often be formed, principally on the noon and afternoon side, frequently breaking 

 suddenly in upon an ordinary polar storm, only to disappear again as suddenly, when the first storm 

 once more resumes its course. 



In reality, the violent deflections that are found in nearly all magnetograms from the polar regions 

 during a storm, are probably due to "loops" appearing locally, and repeatedly coming and going nearly 

 over the place of observation. 



Fig. 4 8. 



0) POINCARE, Remarque sur une experience de M. BIRKELAND. Comptes Rendus 123, p. 930, 1896. 



(a) KR. BIRKELAND. Archives des Sciences Phys. et Nat. Geneva (4) p. 497, 1896; and September, 1898. 



(3) C. ST0RMER, Snr le Mouvement d'un Point, etc. Videnskabsselsk. skrifter i Mathem.. Naturvidensk. Cl. No. 3. 1904. 



(4) Comptes Rendus, June 11 & July 9, 1906. 



