230
DOC.
18
DISCUSSION OF DOC.
16
experiment,
and
it
would be
interesting
to
hear how this
matter
is viewed
by
the
astronomers
present
here.
Einstein: It
is
outside
my competence
to
lay
down the details of how the
astronomers
should do that. It involves the
photographing
of the fixed
stars
near
the
Sun
during a
total solar
eclipse,
so as
to decide whether
or
not
the
nearness
of the
Sun influences the
apparent
locations of the
stars.
Jäger:
Are
astrophysicists
not
of
the
opinion
that
changes
in the
pictures
of
the
fixed
stars
occur
depending
on
whether the Sun
is
nearby
or
not,
and that the
changes
looked for
by
Einstein
completely
vanish in
comparison?
Einstein:
Experts
will have
to decide about
that;
for the time
being,
we
have
to
wait and
see
how the
photographs
will
turn out.
Mie: I would like
to
draw attention
to
another
experimental consequence
of the
various theories
of
gravitation. According
to
Einstein's
theory,
the oscillation
periods
of
atoms at
a
place
of
high gravitational
potential
must
be
different from those where
the
gravitational potential
is
zero.
Hence the lines of
a
line
spectrum
from
a
fixed
star
of
large
mass
must
be shifted with
respect
to
the lines observed
on
the Earth.
[23]
According
to
my theory,
this
is not
so.
My theory,
as
I would
like
to
note
here,
is
based
on a specific principle.
To be
sure,
I
have abandoned the
principle
of
the
identity
of
gravitational
and inertial
mass,
and
I
also believe that
one
cannot
base
any
theory
on
it.
Instead,
I
have the
principle
that the absolute
value
of
the
gravitational
potential
has
no
effect whatsoever
on
physical phenomena.
I call this the
principle
of
the
relativity
of the
gravitational potential.
Thus,
according
to
my theory
the shifts
of the
spectral
lines
predicted by
Einstein's
theory are
not to
be
expected.
Einstein:
Yes,
that's
right; according
to
my theory,
and also
according
to that of
Nordström,
this
must
take
place;
an
oscillator
transported
from here
to
the Sun
must
oscillate slower there. It
is
unfortunately
true that other
causes
also
produce
line
shifts,
and it is therefore
very
difficult
to
check whether
such
a
shift is
produced
[24]
exactly by
this
cause.