Prof. Thomson on the Mechanical Theory of Electrolysis. 441 



from the sm'face about the poles, and downwards to the earth 

 in the equatorial region s, and we may conclude that it does not 

 produce galvanic currents. 



18. From the smallness of the electro-motive intensity in this 

 extreme case, we may infer that no part of the phsenomena of 

 atmospheric electricity can be attributed to the inductive action 

 of the terrestrial magnetism on masses of air or water in motion 

 near the surface of the earth. 



19. If the space surrounding the earth, beyond the limits of the 

 atmosphere, were capable of conducting electricity, and were 

 affected as a fixed conductor by the motion of a magnet in the 

 neighbourhood of it, there would be electrical currents in com- 

 plete external circuits, induced both by the earth's rotatory mo- 

 tion, on account of the distribution of magnetism not being sym- 

 metrical about the axis of rotation, and by its motion through 

 space ; and it is I think far from improbable that the phsenomena 

 of aurora borealis and australis are so produced. It is quite im- 

 possible, in the present state of science regarding the relative 

 motion of the earth or of the solar system, and the medium fill- 

 ing all space, which by its undulations transmits light and radiant 

 heat, to form any estimate on satisfactory principles of the induc- 

 tive electro-motive forces which may arise from the motion of 

 translation of the terrestrial magnet through this medium ; but 

 we may form some idea of those which its rotatory motion may 

 produce by calculating the total electro-motive force on a closed 

 conductor held externally in a fixed position with reference to the 

 earth's centre. Thus let us conceive a circular conductor, of ra- 

 dius R, to be held with a diameter coincident vdth the earth's 

 axis of rotation ; and let i be the intensity of the total electro- 

 motive force which it would experience if it were made to revolve 

 round the earth once in 23^ 56°^ 4^, and the earth held at rest. 

 Denoting by P the radial component of the terrestrial magnetic 

 force at any element of this conductor, and in other respects 

 using the same notation as before, we have 



i^r"" V .^^r co^e .^dd^\^ n V co^OdO. 



J r if^ ^ - 



If we assume the distribution of magnetic force at the earth^s 

 surface to be of the simplest type, the force at either magnetic 

 pole to be 15, and the magnetic axis to be inclined at an angle 

 of 20° to the axis of rotation, we have, at the time when the 

 moving conductor is passing over the earth's magnetic poles, 



P = 15^,sin ie^%(fm^' ^^''' '^'^^\Y^^' 



and in these circumstances we have consequently S^oid^ o-gm 



