﻿the liiUj.i producing Aurora Borealis. 229 



, The simplest way of explaining the indistinctness of the 

 maxima is to suppose a great part of the aurorse to correspond 

 to the interval (I 2 ). The great variety of forms of the orbits 

 will also best account for the great variety of forms of 

 auroral bands, and their rapid and often irregnlar motions. 



7. Finally, we shall consider the question as to whether 

 the distribution of aurorse is consistent with the assumption 

 of a positive radiation. 



Let, in fig. 5, the north pole be turned upwards, then the 

 orbits drawn correspond to negative rays. The positive rays 

 will have quite a similar course, only that they turn round 

 the magnetic axis in the opposite direction, or orbits of posi- 

 tive rays are the mirror image of the negative ones with 

 respect to a plane through the sun and the magnetic axis. 



Now if the rays are negative the aurora occurring before 

 magnetic midnight should correspond to <J> less than 180°. 

 In accordance with the suggestion made by Birkeland I 

 suggested, in a note to ' Nature '*, that the greater frequency 

 of aurorse on the evening side compared with that on the 

 morning side should indicate a negative radiation. This 

 supposition is based on the assumption, which indeed might 

 seem quite a reasonable one, that the simplest orbits, those 

 corresponding to the interval U«£< 180°, should occur more 

 frequently than orbits for which <J> is greater than 180°. 



The mathematical theory, however, is not yet carried so 

 far that the probability of the occurrence of various orbits has 

 been found ; but still the calculations made enable us to 

 draw certain conclusions. 



Looking at the curve A w r e notice that for values of <l> 

 less than about ,35°, ty is greater than any of the values 

 Sm can have, or radiation from the sun cannot at all follow 

 those very simplest orbits passing from the earth towards 

 infinity, because the sun cannot take up the position required. 



Moreover, the results for Kingua Fjord are an exception to 

 the rule that more aurorse occur round the evening than 

 round the morning maximum. But even in the cases where 

 the rule of the greater evening frequency holds, a negative 

 charge of the rays is by no means a necessary consequence. 



Provided the interval I 2 is excluded, <l> is found to be less 

 than 360° and the value of <J> corresponding to the evening 

 maximum is 180° + a, where a. at the various stations has 

 the fairly small values given in Table III., and the sign -f- 

 corresponds to a positive, the sign — to a negative radiation. 



The theory gave two night maxima about symmetrical! a 

 situated with respect to magnetic midnight. Thus both 



* Loc. cit. 



