On Rapid Periodic Variations of Terrestrial Magnetism. 



75 



the pole in the morning lionr, decidedly deviating from the 

 earth's pole. But in this case the phase relation of JZ an JX 

 must be opposite on Ijoth sides of the current which is never 

 actuall}^ the case. If the above be the case, the magnetic 

 waves will naturally be of progressive character. 



Agairi, to explain the phenomena in question on the 

 assumption of a progressive wave of the sort considered in § 2o, 

 it may probably suffice theoretically to consider two intersecting 

 systems of waves propagated with proper velocities, though as a 

 matter of fact, we have neither an evidence nor a physical 

 ground for assuming the existence of such a complicated system 

 of currents. The wave length of sucli waves must, however, 

 be of the order of the earth circumference, in order to explain 

 Fig. 23; hence the case is essentially similar to that considered 

 at the outset of this paragraph. 



Fje. -24. 



Finally, we may su})})(ise that tlie atmosi)heric electric 

 current is not purehj horizontal as was generally assumed in tlio 

 altove discussions, and also that tlie vertical component current 

 is subjected to a periodic fluctuation Avitli a proper phase relation 



