A, 



jiroral Curtains. On pages 605 — 610 of my »Norwegian Aurora 

 Polaris Expedition, 1902 — 1903«, I formed a special theory to explain the 

 manner in which it may be conceived that northern-light draperies are 

 built up. 



In developing this theory, I drew attention principally to the fol- 

 lowing characteristic circumstance, namely that this auroral light often 

 appears as a vertically hanging curtain, consisting of denseh- co-ordinated 

 parallel rays. The longitudinal direction of the curtain is most frequently 

 magnetic east and west. 



As a further characteristic feature I would mention that the auroral 

 curtain is also formed from east to west, or vice versa, in such a way 

 that the rays, one after another, seem to be precipitated from the sky, and 

 this so rapidly that the curtain can be completel}' formed and extend right 

 across the heavens in a few seconds. 



I resorted, as I have always done when studying magnetic and auroral 

 phenomena on the earth, to my experiments with the magnetic terrella. 

 These experiments led me to conclude that the continuous luminous ring 

 in the auroral zone on the terrella was produced by a countless succession 

 of secondary precipitations overlapping one another in such a way that 

 the luminous ring appeared to be continuous. We remember, for instance, 

 having once counted, on the night side of the terrella, about twenty distinct 

 secondary precipitations, of which those of a higher order la}- to the east 

 of those of a lower. The number of these precipitations was greatl}' 

 multiplied in proportion to the increase in the magnetisation of the terrella, 

 and they appeared one after another as the magnetism was increased. 

 This opinion with regard to the construction of the luminous ring is firmly 

 maintained in my endeavour to develope a theory concerning the forma- 

 tion of auroral draperies. 



The cosmic rays approach the earth in the same manner as our 

 cathode rays approach the terrella. We must therefore suppose that the 



