DOCS.
62,
63
MARCH
1915 75
Now I
am
going
to reformulate
the
expression
for 6I. Since
Euvg
is
a “slowly”
variable function of
the
coordinates,
I
can
treat
the
Wuv/g
quantities
on
the
right–
hand
side of
(1)
in
the
integral
as
constants, disregarding infinitely
small amounts
of
higher
order and set
''SIM'™’''«}
or
(1a)
yF9
Since,
according
to
the
assumption
6I
is
an
invariant
and, according
to
the
above,
Auv
is
a
contravariant
tensor
with
independently
selectable
components,
thus
9
is
a
covariant tensor.
In
the
cheerful
hope
that
you
will not
find
any significant
holes in
this
proof,
and
with best
regards,
I
am yours,
A. Einstein.
63. To
Erwin Freundlich
[Berlin,
19
March
1915]
Dear
Freundlich,
I
have
thought
about
the
question
of
whether matter
can
produce
a
gravita
tional
field
other than
a
g44
field.
This
is not
the
case.[1]
The
question
is
whether
the
integral covering
the
mass (e.g.,
Sun)
ƒ
T\dxdydz
becomes infinitesimal.
Within
the
accuracy applicable
here[2]
dXi
3T?
d%\ d%\

H I


0.n
dx
dy
dz
dt