ON ELECTRICAL THEORIES. 145 
suspended magnet, and find that it is too small to be observed ; as, how- 
ever, the time of charging up the wire will be very small compared with 
the time of vibration of the magnet, the effect will be of the nature of an 
impulse, so that in this case there will be considerable advantage in 
having the moment of inertia of the suspended magnet small; while, 
as Schiller arranged the experiment, there was no such advantage, as the 
thing expected was a steady deflection. Thus if the ring magnet were 
retained it would be desirable to make the opening of it as small as pos- 
sible, retaining the same cross action. I think the arrangement could 
be made sensitive enough to be deflected if the value of = were any 
considerable fraction of the rate of increase of the electricity at the end 
of the wire. 
There is another way in which the continuity or discontinuity of the 
current might be tested, and which might perhaps be more delicate than 
the last. We saw on p. 141 that at any point of a current at which 3 
had a finite value the mechanical force on the element is not at right 
angles to the element. In addition to the ordinary force at right angles 
to the element, there is a force in the direction of the vector potential 
equal in magnitude to the product of the values of the vector potential 
and &. 
The existence of this force could be tested 
_ by an arrangement of the following kind :— 
AB and CD are light movable segments 
of the same circle, having balls covered with 
_ paraffin A, B, C, D fastened to their ends. 
_ These segments are connected with a very light 
framework which can rotate about an axis per- 
_ pendicular to the plane of the segments; the 
segments touch at their middle points contact- 
pieces which are connected with a Holtz ma- 
chine. EF is the section of an electromagnet 
concentric with AB and CD; the whole is surrounded with a metal 
cylinder to screen it from external electric influences. When a current 
is passing through the electromagnet it produces a vector potential, 
whose direction is at right angles to the radius from O, the centre of the 
_ electromagnet perpendicular to its axis. Thus if » exists there will be a 
couple tending to twist the system AB, CD about its axis, but if } exists 
at all it will be when the electrical condition of the balls A, B, C, D is 
changing, so that unless the currents are continuous we should expect the 
System to rotate when the balls are being charged up. I have calculated 
that the system might easily be made sensitive enough to be sensibly 
deflected on charging or discharging, provided 3 is an appreciable fraction 
of the rate of change of the surface-density of the electricity on the balls. 
Schiller’s Second Huperiment. 
Schiller has made another experiment, which shows that Ampére’s 
theory fails for unclosed circuits. The first form of the experiment con- 
sisted in having a solenoid placed over a condenser one of whose plates 
could rotate about a vertical axis coinciding with the axis of the solenoid. 
One end of the solenoid was connected to one plate of the condenser and 
the other end to the other plate. When the solenoid ‘is connected toa 
' Pogg. Ann., clix. p. 456; clx. p. 333. 
1885. L 
