54 Prof. L. Vegard : Results of Northlight Investigations 



to position, which are too large and regular to be accidental 

 or due to errors. On the northern hemisphere the auroral 

 zenith is always situated somewhat below the magnetic 

 zenith. 



In individual cases observers have found difference of 

 10°-20° between the positions of the auroral and magnetic 

 zeniths. 



If from these facts we were able to draw the conclusion 

 that the ray-streamers might have a direction different from 

 that of the magnetic lines of force, it would mean that the 

 cosmic electric rays had a very small magnetic deflectibility, 

 of the same order of magnitude as that shown by the 

 a-rays; and rays composed of electrons were in these cases 

 excluded. 



We are not, however, justified in drawing that conclusion, 

 for we must remember that we have found the direction of 

 the magnetic lines of force as it is observed on the surface 

 of the earth, while the auroral rays are playing at a height 

 of 100-350 km. above the ground, and at the place where 

 the aurora appear the direction of the magnetic lines of 

 force may be different from that which we observe on the 

 ground. 



The polar magnetic disturbances which usually accompany 

 the aurorse will according to Birkeland be produced by 

 current systems situated about 200-400 km. above the 

 ground ; and alinough these systems may only slightly alter 

 the direction of the magnetic lines of force at the ground, 

 they may within the region of the aurorse produce a mag- 

 netic force of the same order of magnitude as that of the 

 normal field of the earth. Thus, even if we suppose the 

 streamers always to take up the direction of the magnetic 

 lines of force, the deviation of the radiation point from 

 the magnetic zenith may be accounted for by perturbing 

 influences. 



In this way we may explain the deviations shown in 

 individual cases. For one certain station, it was not ex- 

 cluded that perturbing influences might also produce a 

 deviation of the mean position of the radiation point relative 

 to the magnetic zenith. But if the deviation was due to 

 polar magnetic storms only, we should expect that it would 

 have a different direction at various places. Thus if the 

 perturbing influence at a station to the south of the auroral 

 zone on' an average produced a lowering of the radiation 

 point relative to the magnetic zenith, places to the north of 

 the zone ought to show a deviation in the opposite direction. 



We see, however, that all stations independent of relative 



