On the Repulsion of a Rectilinear Electrical Current on itself. 141 



experiments, two small mercury cups C were placed upon a suit- 

 able support. In these cups were placed the two pointed ends 

 of a conductor of copper wire about a millimetre in diameter ; 

 they were bent in such a manner as to neutralize the action of 

 the earth's magnetism, and to present a horizontal part D D at 

 right angles to the partition between the two compartments. 

 By altering the height of the support, the distance between this 

 horizontal part of the conductor and the surface of the mercury 

 in the compartments could be changed. Two fixed conductors, 

 E E, connect one end of each of these compartments with one of 

 the mercury cups. By connecting the two other ends of these 

 compartments by means of the binding-screws F F with the poles 

 of a battery, the current passed through the mercury from one 

 end of the compartment to the other, then by one of the fixed 

 conductors through the moveable conductor, and then returned 

 to the other pole of the battery through the second fixed con- 

 ductor, and the mercury in the second compartment. When, 

 first, the two cups were fixed at such a height that the part 

 D D of the moveable conductor was only about a centimetre 

 distant from the mercury in the trough (that is, about half as 

 distant as the same part of one of Ampere's floating conductors), 

 and when the current was closed, there was no perceptible motion 

 in the conductor. At this distance, therefore, the crossed cur- 

 rents were not strong enough to displace visibly a part of a con- 

 ductor which a breath could deviate from its direction ; and yet 

 an Ampere's floater placed in the mercury in the same trough 

 after the moveable conductor had been removed, was briskly 

 repelled by the current of the same battery, which consisted of 

 six Bunsen's elements united in a series of three double elements. 

 You see the result of this experiment is not favourable to Mr. 

 Croll's point of view. To make it conclusive, it was necessary 

 to determine the mechanical force necessary to move each of the 

 two conductors, and to show that it was less for our moveable 

 conductor than for Ampere's floater. We intended to make it, 

 but in repeating the experiment with this floater a very simple 

 expedient presented itself, which, had the idea occurred to us 

 sooner, would have rendered superfluous all the apparatus de- 

 scribed above. It is to place this conductor, so to speak, in the 

 contrary direction in the mercury, that is to say, so that the parts 

 ab and dc (see the illustration to Mr. Croll's paper, page 248, 

 vol. xxi. of the Philosophical Magazine), instead of being directed 

 from the curvature b c towards the points P and N where this cur- 

 rent enters, should be directed towards the other end of the trough. 

 Now, reasoning on Mr. Croll's hypothesis, the floater ought still 

 to move from the points P and N when the current is closed ; for 

 there has been no change in its direction either in the mercury or 



