DOC.
594
AUGUST
1918
617
The
coordinates
are
directly measurable,
obtained
upon
measurement
with
a
standard
measuring
rod
at rest
relative to
the
relevant
system.
Time
is
measured
directly
with standard
clocks at rest
relative to
the
system.
Standard
measur-
ing
rods
and standard
clocks of all
systems
should all
agree
with
one
another
when
they
are
brought
to
the
same
place
at
rest relative to
one
another and
are
compared
to
one
another.
If this is
required,
the
quantities
x,
y,
z,
t
are
thus
assigned physical meaning.
Then
arbitrary
transformations
are no
longer
admissible.
Admissibility
then
de-
pends
on
the
physical
laws.
Hence,
in
your
case
of
the
classical
assumptions
(a)
and
(b), no
transformations
other than the Galilean
transformations
are justified.
You
arrive at
other transformations
just
by denying
that
choice of
coordinates
suitable
for
a
simple
interpretation
of
the
formulas.
These remarks
were
of
a general
nature. I
now
turn to
the
second
chapter.[5]
I
have
no
alternative
to
making
detailed comments
on
the
text.
Re
p.
41.[6]
It
is
self-evident
that
I
treat all times
of
the
system
as
equivalent,
since
I
set out from
the
postulate of
relativity.
I
cannot be
reproached
for not
making any
use
of
your
“zonal time.” From
the
explanations
at
the bottom
of
p.
41, one
sees
that
your symmetry system plays
the
role
of
Lorentz’s
ether.[7]
You
thus
immediately
encounter
the
difficulty
springing
from it for
your system
that
our
terrestrially
observable
space
lacks
electromagnetic “asymmetries.”
Your
sentence “The
error
of
Einstein’s
...”
is
completely incomprehensible
to
me.
If
a system’s
clocks
are
not set
rel.
to
one
another
(in
some way or
other),
their
data
are
then
incoherent and
cannot
serve
for
a
time
definition,[8]
not
even
in
a
“symmetry system,” p. 52.[9]
It
is
not
my
view
that
“system
time”
corresponded
to
a
kind of
“higher
time
concept.” My
view is
just that,
in
seeking
to
base
time
on
identical clocks and
signals
between
them, only
a
system
time
can
be
attained,
but
not
a
universal time. One could arrive
at
the latter
only
through
instantaneous
signals
or
through clocks
whose rates
are
not influenced
by
motion.
The existence of
either
of these
things
must
be
questioned
in
principle,
though,
a
priori.- P.
46.[10]
It
is
not correct
that
I
based
the
time
concept exclusively
on
that
of
simultaneity.
The
concept of
the
standard
clock is
added
to
it
as
well;
the
standard
clock
provides
the
unit
of
time.
P.
47.[11]
What
you accuse
me
of here
is incorrect.[12] I
avoid
an
uncertainty
regarding
the time unit
in
the
different
systems by determining
that,
in all
of them,
standard
clocks of
the
same type
be
used, i.e.,
standard
clocks
which,
when
placed
next to
one
another
at
a
state of
relative
rest,
are
identical
(or
at least identical
running
at
the
same rate).
Re
p.
50.[13]
A pendulum
clock
is
no
“clock”
in
my sense, only
a
pendulum
clock
with the
gravitational
Earth; this also
only
if
it
is
admissible to
abstract
from
the
spatial