6lO BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



On account of the absorbing power of the atmosphere, it does not appear that this theory can be 

 maintained. 



Other reasons telling against this theory are advanced by STORMER ( ! ) in his well-known essay, 

 'Sur les trajectoires des corpuscules 6lectris6s dans 1'espace". The only circumstance that ST&RMER 

 finds in favour of VILLARD'S theory and against mine, is that the auroral zone has a diameter ot 

 about 45, while according to his own calculations it should be much smaller (4 to 12 degrees for 

 cathodic rays and /?-rays of radium, and for a-rays 24 to 36 degrees). In order to explain this 

 disagreement, STORMER takes up for discussion the idea that the terrestrial magnetic field outside 

 the earth is greatly modified by exterior currents, especially by the equatorial ring discovered 

 through my experiments. This supposition is less natural, it appears to me, than the one advanced by 

 me as to the rigidity of the rays, viz. that H .Q must be between i and 10 millions. 



LENARD also makes the same suggestion in a recent paper on this subject, as stated on p. 596. 



Further, ST6RMER, in the same essay, paragraph 19, has advanced a very interesting theory on 

 the creation of the auroral curtains based upon his mathematical studies on my theories. 



In admitting an average value of H . Q of 315 for cathodic rays, he finds (1. c., page 119), the 

 theoretical dimensions of an auroral drapery. He arrives, for instance, at a length of 275 kilometres 

 corresponding to a thickness of 72 metres. 



In going through the same calculations and choosing H.Q = 3.1 X 10, I find the length of the 

 drapery almost unaltered, while the thickness has to be multiplied by 10. It will consequently be quite 

 700 metres. Nothing has here been added for the thickness of the auroral rays, as is done by STORMER. 

 It cannot be conceived here, in fact, that the auroral rays can be formed as STORMER supposes, as 

 in that case they would have a thickness of about 100 kilometres, which is contrary to all experience. 

 It will be observed that the dimensions, calculated in the manner indicated above, do not fit so 

 badly to a real auroral drapery; but it must be remembered that STORMER has here only calculated 

 the space in which the rays going to the centre of the elementary magnet, approach the earth. He 

 presupposes that the rays which in reality occur in the auroral curtains keep close to such rays through 

 the centre. We have seen from the experiments, however, that the cosmic rays lying nearest to those which 

 penetrate the auroral curtain, can swing entirely underneath the magnetic equator and penetrate the 

 southern auroral zone. 



From certain positions of the magnetic axis of the terrella in relation to the cathode, we observe, 

 however, that the luminous spot which always occurs on the afternoon side to the north of the luminous 

 ring, stretches itself into a ribbon (see fig. 220). These spots are formed by rays which are drawn 

 directly towards the polar regions of the terrella without swinging above or below the equator, and it is 

 perhaps these rays which are most likely to agree with the bundle of rays in STORMER'S interesting 

 calculation. 



I 1 ) STORMER : Archives des sciences physiques et naturelles, juillet, aout, sept, et Oct. 1907. 



