442 Mr. J. J, Waterston on Electric Conduction. 



in motion " is conveyed by the entire thickness of the conductor 

 L L," call upon us to believe that, while at the beginning and 

 end of any minute interval in the above motion the charge is 

 confined to the surface, yet during the minute interval in which 

 the electricity is in motion it burrows below the surface and 

 moves through the body of the conductor. 



That the actual potentiality of the electric force resides, not 

 upon the charged surface, but in the space adjacent and exterior 

 to it, has, T think, been experimentally established (see § 54, 

 "Electrical Experiments/' Phil. Mag. February 1865). That 

 such is also the case with the electricity of the galvanic current 

 we have some proof in the phenomena of induced currents. It 

 is true that a current of this kind (M), while the circuit is closed, 

 does not induce a current in an adjacent wire circuit (N) at rest, 

 but the inductive influence is manifested when N approaches 

 or recedes from M ; therefore that very potential influence ema- 

 nating from M must exist in the space occupied by N before N 

 begins to move, and before the inductive current consequent to 

 that motion makes its appearance. That change of distance 

 between N and M should be accompanied by change of electric 

 tension in N is a direct consequence of electricity being an ema- 

 native force. That change of tension accompanied with trans- 

 ference of dynamic or work force is the primary condition of a 

 current we may clearly recognize in the example, fig. 1. The 

 difference of tension in a current giving rise to a gradient of ten- 

 sion has been experimentally established in a happy manner by 

 Mr. Latimer Clark (Report on the Atlantic Cable, Appendix II. 

 §§ 23, 25). 



Should the crucial experiment above suggested prove that con- 

 duction in metal conductors is a surface phenomenon, I think it 

 will be found possible to avoid the retarding effect of induction 

 in submarine cables. The mere covering of gutta percha upon 

 a wire does not cause retardation (Atlantic Cable Report, Appen- 

 dix II. § 69). The presence of a conducting medium on the 

 outside of covered wire is necessary. Hitherto experiments seem 

 to have been confined to iron and water as the outside conductors. 

 So far as I know, the effect of copper as outside conductor has 

 not yet been tried. Iron has one-sixth or one-seventh the con- 

 ducting-power of copper; water only about 1000 millionth. 

 Electric conduction as a whole depends on the freedom of motion 

 allowed to the force as a whole. The induced current should thus 

 have the same freedom to move as the primary. 



Practical considerations would probably require that the 

 outside copper conductor should be insulated ; so that, to make 

 the trial effectively, it would be necessary to connect the 

 outside coating (of woven thin copper wire?) with one pole 



