DOC.
51
337
difference to
the condenser
plates,
and
varies
the strip
velocity from lower
to
higher values,
then
at
first both the
charge
of the condenser plates,
which
is
proportional to
the
vector ®, and
the
magnetic
induction
® in the
strip will increase.
When
v
reaches the value
c/eji,
both the condenser's
charge
and
the
magnetic
induction
become
infinitely
large. Hence,
in this
case even an
arbitrarily
small
applied potential
difference
would
destroy
the
strip.
For
all
v
c/(ql there result
negative
values
for
D and
fB.
Thus,
in the last
case a
potential difference
applied to
the condenser
plates
would charge
the condenser in the
sense
opposite to
the potential difference.
Finally,
we
consider the
case
of the
presence
of
a
magnetic
field
Sj
excited
from
the outside.
We
then
have
the
equation
1
^
1
-
^,
,
V2'
1
"
^ % +
f(e^
"
my
which
yields
a
relation
between
Ez
and
Dz
at
a
given
ny
.
If
one
restricts oneself
to quantities
of the
first
order in
v/c,
one
has
(2)
Tz
=
az
+
V-(eß
-
my
,
while Lorentz's
theory
leads
to
the
expression
(3)
Vz
=
eBz
+
f(e
-
l)pS}y .
[11]
As
we
know,
the latter
equation
has been
experimentally
tested
by
H. A.
Wilson
(Wilson
effect).
One
sees
that
(2) and (3)
differ in
terms
of first
[12]
order. If
we
would
have
a
dielectric
body
of
considerable
permeability, it
would be
possible to decide
experimentally
between
equations (2)
and
(3).
[13]
If
one
connects
the
plates
A1
and
A2
by a
conductor,
a
charge
of
magnitude
Dz
per
unit
area
is
generated
on
the condenser plates;
one
z
obtains it
from
equation (2)
by
taking
into
account
that for
connected
condenser plates,
Ez =
0.
One
gets
Dz
=
v/c(em