DOC.
467
FEBRUARY
1918
481
467. To
Rudolf Förster
[Berlin,]
19 February 1918
Esteemed
Colleague,
I
was
very pleased
with
your
new letter,[1]
from which
I
see
how
actively you
are
involved with
the
problems
of
relativity.
I
am
sorry
that
you
must devote
a
major part
of
your energy
to such
pathetic
endeavors! If
you
follow
your
ideas
through
and
publish them, though,
we can
hope
that
you
will
succeed in
exchang
ing your profession
with
that
of
a
teacher,
so
that
you
no longer
have to do
what
is
becoming
the
curse
of
mankind.[2] I
shall
reply
in
sequence
again:
1)
It
seems
to
me
that
B12,34 = B13,24 = B14,23 =
0 is
valid
for
an
arbitrary
coordinate
system only
when
Bik,lm
=
0
for
any
choice
of
the
indices.
2)
and
3).
Here
you
are
entirely right.
I had overlooked
that
the relation
guv
=Suv
+
auv
does not
correspond
to
the
relation
cf"
=
aT
+
s^.
Your
generalization
of
the
Riemann tensor thus
seems
not
to be
trivial,
no
less
than
your
attempt at
a
novel formulation
of
electromagnetics. Otherwise,
we
both
seem
to
be
abandoning
the
nonsymmetrical
guv's
for
the
reasons
mentioned.
Obviously,
I
can
only
form
a
judgment
about the
article
you
are publishing
under the
pseudonym[3]
when
I
have it in front
of
me.
There
is
one
thing I
would
just like
to
point
out, though.
The
infinitely
remote
can
be
regarded
as a
point
like
any
other[4]
only
when the world
is not
infinitely large,
when measured
with
a
cmgauge,
that
means,
when the
guv's
degenerate
at
00
in such
a way
that[5]
/+00
\f^gdr
=
finite
OO
/+OO
^fgïïdxx
=
finite
for
any
x2,
x3,
x4,
etc.,
so
that, for
ex.,
OO
lim
(
[
y/jfadxi')

0.
Then
an
arbitrarily
small
body
can
fill
the
“infinite realm” and
the
world
is
only apparently
infinite
(just owing
to
the
choice of
coordinates). If, however,
the
guv's
do not reduce
to
zero
in
the
indicated
way
for
infinitely large
xu's