86 Nature of the Cosmic Electric Rays. 



auroral streamer, an explanation o£ this sort can no longer 

 be upheld ; but the change of colour must be due to some 

 change as to the way in which the light is excited. In other 

 words, it must be due to some change o£ properties of the 

 cosmic electric rays. 



But the question is now : Ts it necessary to assume that 

 change of colour means a change of carrier ? The fact 

 that an ordinary aurora changes into a red colour without 

 change of place can hardly be brought into harmony with 

 the assumption that a change of carrier is necessary to 

 produce changes of colour. Although our observational 

 material in this respect is very limited, and although we are 

 unable to solve the problem, it may be of value to suggest 

 an explanation which does not require the assumption of 

 change of carrier, but only change of velocity. 



Now we may at once remark that if all electric rays r 

 which were engaged in the production of an aurora,, were 

 completely absorbed near the bottom edge, light should 

 be emitted in the atmosphere from the bombardment of 

 rays, which had velocities that varied from zero near the 

 bottom edge to the initial velocity they possess when 

 they enter the atmosphere. Under these conditions the 

 colour might change from the bottom edge and upward s r 

 but if there were no change in the composition of the 

 atmosphere all the aurorse ought to have shown the same 

 colouring. 



From the study of the luminosity distribution and its 

 variations we were led to the view that the electric rays 

 did not pass straight down parallel to the lines of force 

 to be absorbed at the bottom edge, but they were turning- 

 round the magnetic lines of force and could be driven out 

 into space from the effect of the magnetic field or from the 

 combined action of magnetic and electric fields ; and if so, 

 it is possible to explain variations of colour even when 

 we would suppose that all rays that enter the atmosphere 

 are strictly identical both as regards carrier, charge, and 

 velocity. It is thus a well-known fact * that the spectrum 

 of a gas, and especially that of nitrogen, varies enormously 

 with the velocity of the cathode rays which produce it ; and, 

 as a matter of fact, the luminosity produced in nitrogen 

 by cathode rays of small velocity in a vacuum tube has 

 a colour very similar to the red colour which the aurora 



* See E. Gehrke & R. Seeliger. "Ueber das Leuchten der Gase unter 

 Einfiuss von Katkodenstrahlen,' 1 Verh. d. Phys. Ges. xiv. pp. 335, 1023 

 (1912). Fulcker, Astroph. Journal, xxxiv. p. 388 (1911). L. Vegard, 

 Ann. d. Physik, xli. p. 625 (1913) ; lii. p. 72 (1917). 



