1916. No. I. ARE THE SOLAR CORPL'5CLE RAYS NKGATIVE OR POSITIVE? 25 



for instance that an auroral drapery with a length of 275 kilometres and 

 a thickness of 72 metres may be formed by ordinary cathode rays, on 

 the supposition that the cathode rays are following near to the rays 

 which pass through the centre of the elementary magnet. But this suppo- 

 sition is as false as anything can be: anyone who has seen a fine north- 

 light curtain in formation knows that. When he sees the auroral rays 

 fall down from the sky and form a curtain coordinating themselves one 

 after another with an interval of time of about one fifth of a second 

 between each ray, he thinks after some reflection : if these rays came 

 from the sun, they would have proceeded in very different paths altogether. 



Professor Stormer, however, still frequently quotes his theory. He 

 says himself in the paper in -Terrestrial Magnetism ^ referred to above 

 »the mathematical theory of auroras which I have worked out in several 

 memoirs can be applied to the case of negative corpuscles as well as to 

 that of positive corpuscles*. 



This statement may well be impressive for those who do not under- 

 stand it well. Any theory having corpuscular rays for its starting point 

 has eo ipso the same property of application. But the facts observed on 

 the earth speak clearly in favour of the negative rays. It appears from 

 mv experimental discovery of the physical nature of auroral curtains that 

 a mathematical solution of the problem concerning this phenomenon would 

 onlv be within the limit of possibility if some mathematician had the good 

 luck to find the general solution of the equations of motion for electrified 

 corpuscles from the sun, when the rays arrived under the influence of the 

 earth's magnetism .\nd even then I do not believe that anyone could 

 have invented this special theory only by mathematical studies. At any 

 rate it would hive required a mathematician of almost unknown intimate 

 contact with nature to reveal her deep secrets only by means of mathe- 

 matical anal3sis. 



The experimental method offers indeed a much more natural way to 

 explore nature at first hand, even if the way is long and troublesome. 



The method of experimental analogies especially will more and more 

 prove to be highly valuable for the natural philosopher. When, however, 

 discoveries are made, the mathematical analysis may do splendid work in 

 clearly formulating discovery and theory, in drawing far reaching conse- 

 quences and deepening details. 



To return to our problem, we may say that at the present day the 

 general solution of the equation in question does not exist at all. 



As is well known, Professor Stormer has put forward a very fine 

 method of calculating the separate possible paths that electric corpuscle 



