and Nature of the Cosmic Electric Rays. 65 



which make the air luminous can be seen to penetrate 

 towards a magnetized terella. 



Also some experiments of Villard * show that bundles of 

 cathode rays made visible through collisions can move in 

 very definite orbits in a magnetic field. 



We are tints able to draw the conclusion that any knoivn 

 type of electric rays — under certain conditions — give a varia- 

 tion of luminosity per unit length of the streamer with the 

 change of the angle between the rays and the magnetic lines 

 of force. 



Now the great differences found with regard to distri- 

 bution of luminosity are explained as follows : — 



If the electric rays when they pass the upper strata of the 

 atmosphere are all nearly parallel to the lines of force they 

 produce very little light per unit length of the streamer. 



When they come down to a height of about 100 km. the 

 density of the air increases rapidly and the luminosity 

 suddenly increases. The increase of luminosity may also be 

 partly due to an increase of the number of turnings per 

 unit length, for as the magnetic lines of force are not 

 parallel, but converge as we pass towards the earth, a ray- 

 bundle which originally was nearly parallel to the lines of 

 force will gradually begin to turn round the lines of force, 

 and if the rays are not absorbed they may finally become 

 perpendicular to the lines of force and then turn back into 

 space. 



We might e.g. assume, that the field was that of a single 

 magnetic pole, then Poincare f has shown that the orbits 

 are geodetic lines on a cone with its top at the magneiic 

 pole. St^rmer % has studied the case when the field is 

 produced by an elementary magnet, and he has given a 

 number of curves illustrating the form of orbits of electric 

 rays, when they are moving in a field with converging lines 

 of force. 



If the electric rays may turn back into space without 

 being absorbed in the atmosphere, the height to which the 

 rays come dowm is not determined by the penetrating power 

 of the rays, but they are simply drawn away as a result of 

 the deflexion produced by the magnetic field of the earth. 



In this way we see that there is a possibility of explaining 

 variations of the height of the aurorae without assuming 



* Villard, C. R. 1906. 

 t H. PoincarS, C. E. cxxiii. (1896). 



% C. Sttfrmer, Vid. Selsk. Skr. Christiania, 1904; Archives d. Sri. 

 Phys. et Nut. 1907. 



Phil. Mag. S. 6. Vol. 42. No. 217. July 1 921 . F 



