588 DOC.
564
JUNE
1918
Energy
cur[rent]
[c,
f)]
+
energy
density[6] are
correct.
Tav
degenerates
for
electrostatics
to
the
expressions,
as
Cohn has found
through
variation
of
ƒ
edö.[7]
However, tensor,
as
well
as
Tuv is
no
longer symmetric.
Momentum
density
[ö,
b]
=
[c,
f)]
+
[p,
b]
+
[c,
m]
This
looks
as
though these[8] certainly
denoted momentum
but
not
energy
cur.
This stems from
that the
electrical
dipole
at
buildup
of the
electr.
field
e
in
magnetic
field
b
obtains
a
force
that
is just
equal
to
d/dt[p,
b].
The
corresponding energy
current
is
free
of
divergence already
for
the
self-contained
individual
dipole:
The
mean
value for
the
momentum
thus
differs from
zero,
while
the
mean
value for these
energy
currents[9]
vanishes.
Tuv is
thereby
asymmetric
and
does
not
contradict the
prin-
ciple
of
the inertia
of
energy,
after
all.[10]
Tav
is
the
tensor
that
must
be inserted in
order to
obtain the
correct force
density
and
actually
has
nothing
to do with
the
mean
value for
the
energy-momentum
tensor in
a vacuum.
Ka can
be converted into
the
following
simple
form
Ka-F^J
--J
dxj
dFaß.
This[11]
d/dxo
is
meant for
a spatio-temp[orally]
constant
field.
The
integral
has
the
same
meaning
as
above. The result for matter at
rest,
divided into
space
&
time
is:
•
'«i,23
=
[i, b]
+
pe
-
J
grads(c,
de)
«
-
ƒ
grad
(-jpj
(b, db)
r
de
r
9
(-)
=
t,
db)
e
=
Dielectr. const.
=
tan a
u
=
Permeability