70 Art. 9.-T. Terada: 



conviction, tliat if the tiuctuations of the current be either of tlie 

 kinds here considered, it is not merely due to the mechanical 

 wave motion of the upper atmosphere. This leads the question 

 towards the weakest side of our knowledge. 



The natural period of the usual electrical oscillations cannot 

 in any case be as long as several minutes, even if we take into 

 account the effect of the upper atmospheric shell; nor can it be 

 of such indefinite periods as actually observed. 



Though it seems .50 far difficult to say anything definite 

 aVjout the physical cause Avhich may produce the fluctuations in 

 question, the existence of the fluctuations above considered 

 seems to be suggested, on the other liand, l)y the phenomena 

 of tlie aurora polaris. Firstly, the frequently observed auroral 

 arc vividly reminds us of our hypothesis of the difïused 

 zonal current considered in § 23. According to the numerous 

 descriptions of the phenomena, the luminosity seems to undergo 

 different modes of variation, thougli we could find no record 

 definitely stating a periodicity of a few minutes. Such a slow 

 variation, even if present, would prul)al)ly have eluded discovery 

 by ordinary simple eye-ol)servations, not equipped Avitli special 

 means for that purpose. A special investigation in this direction 

 seems in any case desirable; for example a minute photometrical 

 study of cinematographical I'ecords as obtained b}^ Stürmer.'^ 

 Again, among the numerous records of the auroral displays,"^ 

 w^e frequently meet with descriptions of peculiar bands of 

 luminosity, running nearly perpendicular to the meridian and 

 showing wave-like propagation mostly toward S. It is very 

 probable that similar phenomena are of quite frequent, or rather 

 of almost daily occurrence, although usually the intensity may 

 be so weak that it rarely shows a luminousity strong enough to 

 attract the attention of the naked eye. Indeed, it is well known 

 that the characteristic "auroral lines" of spectra are almost 

 invariably visible in different latitudes. If such bands of 



1) C. Stormer, Videnskapsselskabets tkrifter, Math.-Naturv. Klasse, I'Jll, No. 17; 

 Astrophys. Jonr., 38, 1913, p. .311 ; Bull. Soc. Astr. France, 1913, Xov. 



2) E. E. Barnard, Ob.servations of Aurora made at the Yerkes Obs., 1897-1902, Astrophys. 

 Jour. 14. 190J. p. 135. M. Brondel, Uober das Xordliclit von 30, Juni, Met. Zs., 1908, p. 552. 



