DOC.
18
DISCUSSION OF DOC.
17 355
hand,
when
he
was explaining
Michelson's
experiment,
stated
outright
his
conjecture
that
it
may
well
be
good
to
assume
that
such
a change
in
length is brought
about
by
the
influence
of
the
ether
or
of molecular
forces.
These
are
two
things
that
I cannot
[6]
reconcile.
Prof.
Einstein: Allow
me
to
answer
with
a
comparison.
It
has to do with
the second
law
of
thermodynamics,
the
law
of the limited
convertibility
of thermal
energy.
If
one
takes
the
assumption
of the
impossibility
of
a perpetuum
mobile
of the second
kind
as
the
starting point
of the
argument,
then
our
law
appears
as
almost
an
immediate
consequence
of the
basic
premise
of the
theory.
But
if
one
bases
the
theory
of heat
on
the
equations
of motion of
molecules,
then
our
law
appears
as
the result of
a long
series
of
most subtle
arguments.
Just
as
here both of these
routes have
their undeniable
justification, so
the above-mentioned
points
of
view
of
Minkowski
on
the
one
hand,
and
of
H. A.
Lorentz,
on
the
other,
also
seem
to
me completely
justified.
Prof.
Meissner: As
far
as
I
know,
Minkowski
used the
theory
of
relativity
in
order
to
[7]
derive the
general equations
for
moving
bodies
from
the fundamental
equations
of the
electrodynamics
of
bodies at rest.
He
set
up
a
system
of
formulas
that
does not coincide
either
with
the formulas of
Cohn
nor
with
those of
H. A.
Lorentz.
Objections
have
been
raised
against
this
new system.
Since I know
the
whole
theory
of
relativity more
from
the
[8]
mathematical
point
of
view,
I would
very
much
like to know
the
reasons
that
prompt
the
physicists
to
decide
against
Minkowski's
equations
and
in favor
of those of Cohn and
Lorentz. It
seems
to
me,
from the mathematical
point
of
view,
that there
must exist
only
one
system
of
equations,
namely
that of
Minkowksi.
Prof.
Einstein:
If
one
starts out from
the
theory
of bodies at
rest,
then
one can
derive
only
the
laws
of
electrodynamics
for
uniformly moving
bodies
by means
of the
relativity
transformation. Do the
equations
of
electrodynamics
for
uniformly moving
bodies
also
hold
for
bodies
in
spatially
and
temporally
non-uniform
motion?
This
is
possible
but
not
certain. To that
extent,
Minkowski's
equations
are a
hypothetical
extension of those
existing
before.
Regarding
the theories of Cohn and
Lorentz,
the
following
should
be noted. The
theory
of Lorentz
deviates from
that of
Minkowski insofar
as a
small
inaccuracy
crept
into
it,
due
to
the
much
more
difficult
method of
derivation.
In
fact,
there
are no
fundamental differences between
Minkowski's
and Lorentz's
theory.
On the other
hand,
Cohn's
electrodynamics
must
be
viewed
as
fundamentally
different.
[9]
Fritz Müller:
According
to
the
explanations
given
in
the
lecture,
when
a
clock
is
set
up
at
the North
Pole,
and
a
synchronously
running
clock
is
located at the
equator,
then,
if
we
consider the rotation of the
earth,
the
clock at
the North Pole
is
at
rest,
while
the