78 Art. 9.- T. Terada : 



quadrants 0'^-6!^ 6^-12;^ 12"-18f 18"-24'; the phase angle differs 

 successively by -, if /o be of the same sign throughout; or û 

 clianges the sign alternatively, if ç remains constant. It we 

 are to identify /o with the equatorial portion of the circular 

 current assumed in § 19, then it will suffice to suppose that a 

 portion of the circular belt or ring carrying the current is subjected 

 to a radial vibration, as if it were a portion of the ring carrying 

 out a dilatator]/ vibration in such a way that the amplitude is 

 given by cos-V, say, where ^' is the longitude counted from noon. 

 The fluctuation of the intensity must be considered as keeping 

 pace with the vibration and the period must be different for day 

 and night. 



Though the above solution is by no means unique, yet it 

 may at least serve as a provisional explanation as far as the 

 phenomena of rotatory pulsation are concerned. It is, moreover, 

 not contradictory to the principal facts of observation already 

 enumerated: viz., not irreconcilable with the mode of the daily 

 variation of the azimuth of tbe horizontal components, since 

 the circular current assumed may l)e considered to have its 

 maximum intensity alwaj^s near the equator; and again not 

 inconsistent with the observed amplitude ratio and phase relation 

 of JZ and JX, since the see-saw motion chiefly affects JY, 

 and tlie discussions on these relations given earlier hold good 

 liere also without any essential modification. 



According to van Bemmeln, the ratio /iZjilX at Batavia 

 seems to be invariabl}^ of insensible magnitudes. This fact is 

 in harmony with the above assumption that the maximum current 

 is near the equator. On a theoretical ground, this seems also 

 admissible if we may attribute the origin of the disturbing 

 current to the vertical motion of tJie atmosphere, since general 

 planetary circulation of the whole atmosphere must be in any 

 case vertical in these regions, where the horizontal intensity 

 of the magnetic field is also greatest. 



29. Tlie above taken as granted, it seems plausible to 

 assume next that th(? impulse causing the vertical vibration, if 

 not solely mechanical, is to be attributed to the reciprocal action 



