PART I. ON MAGNETIC STORMS. CHAP. II. 



in the southern hemisphere as could practically be reached without too great expense, by accompanying 

 hunting-expeditions, would without doubt raise the veil that obscures the great question of the origin 

 of terrestrial magnetism, which has hitherto been one of the greatest mysteries of Nature. 



In order to illustrate clearly the course of the rays in the case illustrated in fig. 66, Stermer has 

 calculated the trajectories of cathode corpuscles answering to those in this experiment, and has shown 

 the result in a wire model, which is photographed in three positions in fig. 71. 



Stermer has added some remarks upon this model, which he kindly allows me to quote. 



"This wire model (fig. 71) represents a number of trajectories of negatively -charged corpuscles, 

 moving under the influence of an elementary magnet. 



"The trajectories are constructed on a graphic method of integration, worked out for the occasion, 

 which will be more fully described in the second part of this work ( 1 ). 



"The model was specially made for Birkeland's experiments, and the sphere therefore repre- 

 sents the terrella, and the plate on the right the cathode. The elementary magnet is placed in the 

 centre of the sphere, with its axis parallel to the black rods, and the south pole upwards, the latter 

 being marked with a cross. The sphere is fitted with a rod representing the earth's axis of rotation. 



"The lowest layer of rays consists of plane curves lying in the magnetic equatorial plane; they 

 are calculated exactly, and are a good check upon the others, which are constructed graphically. Above 

 this lowest layer of trajectories lie four other layers, so that the model shows more than 50 different 

 paths. To each path in the model, there is also a corresponding one that is symmetrical with the first 

 with reference to the magnetic equatorial plane; but all the trajectories thus produced are omitted so as 

 not to make the model too intricate. 



"The ring is clearly seen that answers to the luminous ring round the terrella in Birkeland's ex- 

 periment. If we call the moment of the elementary magnet M, and express the characteristic constant ('-) 



of the corpuscles by H (t o u , then the radius of the ring equals 



cm. 



"On the third photograph are marked the points in which the trajectories intersect a sphere con- 

 centric with that in the model, and with a radius rather less than that of the ring. At the points of 

 intersection, the tangents to the trajectories have also been drawn. 

 It will be seen how the directions of the tangents form a vortex; 

 and symmetrical with this, there is a vortex on the other hemi- 

 sphere, below the magnetic equatorial plane. If arrows are marked 

 all over the sphere in directions the reverse of those of the above- 

 mentioned tangents, we obtain the accompanying figure 72 in which 

 the sphere is seen from without. The figure is only diagrammatic. 

 We see that the part upon which the corpuscles impinge has the 

 same form as that visible in the experiment; and above this there 

 are two contrary cyclonic current-vortices in the direction of the 

 arrows, situated symmetrically with reference to the magnetic 

 equatorial plane, and answering to the positive currents that might 

 produce cyclo-median perturbations. 



"The trajectories that have been chosen in the wire model 

 are especially those that approach the elementary magnet, and 

 then once more recede to an infinite distance, and not such as Fig. 72. 



(') Cf. "On the Graphic Solution of Dynamical Problems", by Carl Stormer. Videnskabsselskabets Skrifter; Christiania, 1908. 

 ( 2 > Cf .Carl Stermer's "Sur les Trajectoires des Corpuscules Electrises dans 1'Espace, etc." Archives de Geneve, July October, 1907. 



