50 
PHYSICS: S. J. BARNETT 
Without, however, the introduction of hypotheses unproved by experiment 
it does not seem possible to determine from the experiments on rotation what 
would happen in the case of translatory motion, or to obtain from them any 
answer to the question of the existence of the aether. The present investiga- 
tion was undertaken with the hope of shedding some light upon these problems. 
2. If an air condenser with horizontal parallel plates, short-circuited and 
placed in a uniform magnetic field whose lines of induction are parallel to the 
plates, is moved, together with the short-circuiting wire, in a direction parallel 
to the plates and perpendicular to the lines of induction, it will become charged 
to a potential difference equal to the motional electromotive force, which will 
be denoted by E, in the wire, the charges (provided, as will be assumed, that 
edge effects are negligible) being restricted to the inner or opposing faces of 
the -plates. 
3. If the region below the plane of the upper surface of the lower con- 
denser plate were filled with a medium of zero permeability, the tubes of 
magnetic induction would be confined to the region above this plane, and the 
condenser would become charged exactly as before for the same motion. 
4. If, with the arrangement of §3, the condenser and wire were to remain 
fixed and the agent producing the magnetic field were to move, the relative 
motion being the same as before, the results to be expected would be different 
according to the hypothesis adopted with respect to the aether: 
A. If there is no aether, and the principle of relativity is valid, the con- 
denser would become charged exactly as in §2 and 3. 
B. If the aether exists, the electric intensity produced by the motion of 
the tubes of induction in the aether between the condenser plates would be 
equal and opposite to the field intensity there due to the potential differences 
produced by the electromotive force in the short-circuiting wire, and there 
would thus be no charge on the lower condenser plate. 
Thus we should have a method of discriminating between the two 
hypotheses. 
5. There being of course no medium of zero permeability in existence, I 
have tried to secure an arrangement equivalent to that of §4, so far as the 
effect under investigation is concerned, by using an artifice analogous to the 
electric guard-ring. 
Two similar electromagnets, referred to as the upper and lower magnets, 
with their coils in series are used to produce the field. The four poles are all 
alike, each being a rectangular parallelepiped with the largest sides horizontal 
and parallel to the condenser plates. The lower magnet and condenser are 
fixed to the floor. The upper magnet forms the bob of a huge pendulum 
swung from the ceiling, and has translatory motion parallel to the condenser 
plates when in its lowest position. When the upper magnet is in this position, 
the upper and lower poles are about a centimeter apart, and the upper surface 
of the lower condenser plate is symmetrically located with reference to all 
