11 



with a force varying inversely as the square of the distance from 

 the respective centres. Commencing with the atmosphere. — If 

 the globe be a magnet, we ought to observe at the poles some in- 

 dications of the convergence of the air towards the poles, i. e. 

 something similar to the inverted conical appearances which we 

 observe in the ingredients at the poles of an artificial magnet, 

 Plate I. It is true that the air is an invisible substance, yet, as 

 it occasionally becomes saturated with visible fluids, it would, 

 under such condition, with the advantage of reflected light, which 

 may also be expected to vary according to the curves formed by 

 the currents, show the phenomena of convergences. This fact 

 we have in the aurora borealis and australis. These are the 

 luminous appearances seen in the atmosphere connected with 

 the poles of our earth, the general appearances of which corre- 

 spond to the curves of convergence towards the poles. When 

 these luminous phenomena display unusual brightness and ac- 

 tivity, the magnetic needle is also found very fluctuating, both 

 in dip and direction ; and also the mercury in the barometer is 

 subject to similar action ; this coincidence of the variable move- 

 ments indicates that they are produced by the same cause, viz. the 

 disturbance of the equilibrium of the magnetic curves, or tension 

 of the fluid. 



The aurora is not the cause of the needle being disturbed, nor 

 are its luminous rays required to produce a rise or fall of the mer- 

 cury in the barometer ; on the contrary, the whole phenomena 

 appear to be the effects of the oscillating movements of the mag- 

 netic currents. 



The mercury in the barometer, like the magnetic needle, fluc- 

 tuates without being accompanied with a visible aurora. Pro- 

 bably the light is produced by a change in the constituent ele- 

 ments ; but the light is not essential to prove the existence of 

 the currents ; nor is visibility necessary to produce undulations 

 in such currents : such variations can only be ascertained by 

 their effects. 



That the barometer is subject to perpetual oscillation is a phe- 

 nomenon so well known as not to require comment ; but that 

 the magnetic needle should be also subject to similar oscillations 

 appears to those who have not paid attention to the subject some- 

 what sti-ftnge tT ^M'ever, the fact is, that the former is the effect 



