458 Messrs. Watson and Burbury on the Laio 



normal force only on every elementary current (canons III. 

 and VI.), this being the only experimental fact inconsistent 

 with the simple law. 



The experiments quoted in support of this are the follow- 

 ing:— 



(1) A current in a circular 

 conductor, movable about an axis 

 through its centre perpendicular 

 to its plane, remains unmoved in 

 the field of another closed cir- 

 cuit. 



(2) A current in a rectangular 

 conductor A B C D, movable 

 about one side A D, which coin- 

 cides with a line through the 

 centre, 0, of a circular conduc- 

 tor, the line A D being perpen- 

 dicular to the plane of the latter, 

 remains immovable when a cur- 

 rent passes through the circular 

 conductor. 



(3) (Maxwell, vol. ii. p. 149, cited above.) A circular 

 conductor, C D, movable about its centre in its own 

 plane, passes through two mercury-cups fixed at A 

 and B, at which a current enters and leaves the 

 conductor respectively ; and no motion ensues when 

 a closed current is brought into the neighbourhood. 



Now, in point of fact, all these three experiments 

 would be equally satisfied by any law which gave a 

 potential between two currents. For in (1) and (2) 

 no possible displacement of the movable circuit 

 could alter that potential, if it existed, as is obvious 

 from the symmetry of the arrangements in the ori- 

 ginal and displaced positions respectively. Since, therefore, 

 no displacement could alter the potentials if they existed, it 

 follows that if there be a potential there would be no force 

 producing motion. 



The same thing is true of No. 3, although at first sight not 

 so evident; for since the current enters and leaves the con- 

 ductor at the points A and B fixed in space, and the arc A B 

 is fixed in space, it follows that, notwithstanding any motion 

 of the conductor, the current remains fixed in space, and the 

 potential must remain the same whether the conductor be 

 moved or not. Hence there can be no tendency to move the 

 conductor. 



These three experiments, therefore, do not prove that the 



