82 



BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



Fig. 39 is taken from his paper, "Sur les trajectoires des corpuscules electrises dans 1'espace sous 

 1'action du magnetisme terrestre'^ 1 ). It will here be seen that rays answering to y 0.5 and 0.7 

 fall in towards the earth very much as do the greater number of the rays in the experiment with the 

 terrella in position b. There will also certainly be rays coming in towards the terrella, that answer to 

 the other mathematically possible paths; but it is not so easy to demonstrate them with this experimental 

 arrangement with a screen. 



The phenomenon re- 

 presented in fig. 37, of the 

 ring of light round the equa- 

 tor, should answer to paths 

 where y = about i. The 

 stronger the magnetism, the 

 larger will the ring be. In 

 the experiments shown in 

 fig. 38, the magnetism is so 

 strong that the equator-ring 

 is not formed, owing to the 

 glass walls of the discharge- 

 tube. By the experiment with 

 the terrella, it is also easy 

 to show a phenomenon that 

 is most easily explained by 

 the presence of rays answer- 

 ing to the calculated paths 

 for y = between 0.5 and 

 - i. I have mentioned in 

 a former work( 2 ) that, just 

 within the equator-ring, the 

 terrella sometimes has a 

 clearly phosphorescent line 

 along the equator. 1 had 

 formerly to have recourse to 



Fig. 39- 



the assumption of secondary rays in order to explain this phenomenon; but it is now explained most 

 naturally by rays answering to Stermer's calculated trajectories. 



What we have to notice, however, is that the bulk of the rays in the experiment turn round in 

 front of the terrella on the afternoon side. The mathematical treatment has hitherto given only the 

 mathematically possible trajectories, but has not stated where the bulk of the rays pass the earth, partly 

 because the nature of the rays emitted by the sun is not sufficiently known. 



As the current-arrows during our perturbations are directed towards the east, the perturbation 

 cannot be explained by a ring such as this round the earth. If, on the other hand, we assumed the per- 

 manent existence of such a ring, we might imagine the perturbation to be explained by a diminution in 

 the strength of this current. This explanation is very improbable and unnecessary. It seems necessary, 

 owing to the connection of these perturbations with the polar storms, to suppose that the equatorial 



(') Archives des Sciences Physiques et Naturelles. Geneva. Vol. XXIV, 1907, chap. IV. 

 (*) Expedition Norvegienne de 18991900, 1. c., p. 46. 



