ON ELECTRICAL THEORIES. 147 
From this equation we see that the H.M.F. due to the rotation vanishes 
for each plate, so that, according to this theory, there should be no current 
through the galvanometer. 
This experiment of Schiller shows that both Grassmann’s and Clau- 
sius’ theories must be wrong, as well as Ampére’s and Korteweg’s, for 
we can easily see that they would make the disc rotate in the way in 
which Schiller first tried the experiment, and if this were so, it follows 
from dynamical principles that a current must be produced in the second 
form of the experiment. 
This would seem to be the case even if we take into account the cur- 
rents in the dielectric, unless we suppose that all the circuits are closed, 
for if all the circuits are closed then the disc will not rotate, as all the 
theories agree. If the circuits are not closed we may divide the currents 
in the disc into two parts, one part being of such magnitude as to form 
with the dielectric currents closed circuits; then the forces on this part 
and the dielectric will form a system in equilibrium; and there remains 
the other part of the currents, the action of the magnet on which ought to 
set the disc in rotation. Taking Schiller’s experiments together, we may 
say that they show that the dielectric must be taken into account, and 
that some form of the potential theory is the only one of the theories we 
are considering which can give the expression for the forces due toa 
distribution of currents. 
Although these two experiments of Schiller’s show that of the 
theories we have discussed only the dielectric ones can be retained, we 
shall describe one or two more experiments which have been or could be 
made to distinguish between the various theories. Clausius’ and Grass- 
-mann’s theories lead to the same expression for the force between two 
elements of current, so that these theories stand or fall together. Grass- 
-%Imann in his paper! describes an experiment which would distinguish 
between his theory and Ampére’s, or, in fact, any other except Clausius’ 
which has ever been published. 
Suppose that NS and SN are two mag- 
nets whose north and south poles are de- 
noted by N and § respectively, and that 
these magnets are fastened together by a 
rod NS, the system being suspended by a 
cocoon thread attached to the middle point 
of NS. Let AB be an unclosed circuit, say yy 
@ Wire joining the plates of a charged con- i 
denser ; then, according to Grassmann’s and 
Clausius’ theories, the system will rotate in Swern 
such a way that the sense of rotation is re- ; 
lated to a vertical line drawn downwards s : “s 
like rotation and translation in a right- 
handed screw. According to every other theory it will rotate in the 
opposite direction. 
Another experiment has been made by v. Helmholtz,? which shows 
that the potential theory leads to wrong results unless the action of the 
dielectric is taken into account. bb is a rotating conductor, to the ends 
of which large condenser plates are attached, which, when in rotation, 
come very near to the similar plates c, c. The plates b and care segments 
’ Pogg., Ixiv. 1, 1845. 2 Wissenschaftliche Abhandlungen, vol. . 783. 
L2 
