7'4 



BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 19021903. 



On a previous occasion we have in reality shown under somewhat different circumstances that th 

 rays of corpuscles ejected from the cathode are reflected from the walls. In a paper in C. R., March n 

 3913, I have said that the long pencils of rays emitted from the cathode and carrying a quantity of dis- 

 integrated matter from the cathode, are reflected from a wall like rays of light from a mirror. 



If, therefore, it is desired to study the course of these bundles of rays, or their ability to pass 

 through, for instance, thin aluminium-foil, employment should be made of a layer of fat to intercept the 

 corpuscles after their passage, as the ray-phenomena are then more easily demonstrated. 



After the treatment with grease mentioned above, the 

 cathode is well fitted for patch-experiments, and intensely 

 powerful disruptive discharges are formed even without 

 additional external capacity. 



In a high vacuum the patches generally consist of 

 groups, but may also consist of a single patch at each place. 

 They may often remain in one place for a measurable length 

 of time. They are surrounded by the previously-mentioned 

 vortices (see fig. 261), rotating in opposite directions on the 

 two magnetic hemispheres. Round the single patches more 

 particularly, these vortices attain a surprising clearness and 

 regularity. 



It appears on comparison of the above with HAL 

 photographs of sun-spots with vortex-formations, that I have 

 been guilty of a misunderstanding. 



The experimental vortices are in the reverse direction 

 to HALE'S, supposing the magnetic north pole to be on 

 the top of the cathode-globe. In my descriptions I have 

 reckoned the vortex from the centre outwards, contrary to 

 Hale, who has considered them in the more usual way. 



But the consideration of the experimental whirls and the solar vortices as analogous phenomena 

 does not seem to involve any contradiction. 



In my experiment the magnetic power in the spot is determined by the magnetisation of the cathode- 

 globe. The current-strengths carried by the discharges are too small to produce any marked local field. 

 In a sun-spot, on the contrary, the local magnetic field predominates, and it may very well be due to 

 the enormous conditions on the sun. In some way or other with which we are not now acquainted, 

 vortices may arise from the discharges. The current-strengths are so great that the magnetic forces 

 formed by them will be able to entirely reverse the original magnetic field which was due to the general 

 magnetisation of the sun. 



Here it should possibly be considered that the current-paths in the photosphere around a spot are 

 "selected", so to speak, at the first moment, before the current-strength in the discharge has attained to 

 any magnitude worth mentioning. Later, when it becomes perhaps millions of times greater, the current- 

 paths retain to some extent their orientation, and produce a corresponding magnetic field. 



We have repeatedly pointed out the resemblance that exists between the light-phenomena about a 

 magnetic cathode-globe and corresponding solar phenomena, such as the corona with the radiating oft- 

 shoots in the polar-regions, and the sun-spots. 



The light-phenomena about a magnetic anode-globe, on the other hand, are quite different, except 

 that the radiation in the polar regions is sometimes nearly like that from a cat/iodf-globe, and resembles 

 the polar radiation of the sun. 



