DOC. 22 SOLVAY DISCUSSION REMARKS
267
opinion-is of
great importance
for
a
clear
understanding
of the
general
content
of
Nernst's theorem.
Nernst seeks first
to
prove
the
impossibility
of the existence of
an
adiabatic
process occurring
in the finite
through
which absolute
zero
can
be reached. For
if
such adiabatic
processes were
to exist,
then
there would exist
cyclic
Carnot
processes
in which the lower
temperature
would be
zero
(T2
=
0), thus,
processes
described
by
the
diagram
in
Fig.
2.
FIG.
2
It
follows from the familiar
general equation
q1/T1
=
q2/T2
that
in
our
case,
in which
T1
T2
T2
=
0,
q1
vanishes.
Thus,
the isotherm
CD
would have
to
be viewed
simultaneously
as an
adiabatic. The
system
would be able
to
go through
it
without
necessitating
the
use
of
a
heat reservoir of
temperature
T2
=
0.
Thus,
the
cyclic process
would
require
only
one
heat reservoir
of
temperature
T
=
T1
and would be able
to convert
its
heat
to
work-in
contradiction
to
Carnot's
principle.
Nernst concludes
from this that
an
adiabatic
of
the
type
BC
cannot exist.
I
doubt the
cogency
of this
argument
for the
following
reason.
In
my opinion,
it
is
impossible
in
principle
to
carry
out
the
partial process
CD
adiabatically.
Because
every irreversibility,
no
matter
how
small,
must
have the effect that the
system
arrives
back
at
B,
instead of
at
D,
from the
state
it
occupies during
the
compression along
the
adiabatic. For since
absolutely
exact
reversible
processes
do
not
exist in
nature,
a
compression
of
our
system starting
from C
cannot
occur
in
an
absolutely
reversible
manner;
minute
quantities
of
energy
will
always
be transformed into disordered
energy (heat).
However small these
quantities
of
energy may be, they certainly
do
exist and force the
system
from the T
=
0 axis into the adiabatic
CB.
The
cyclic
process
under consideration is therefore unrealizable in
principle.
Even
if
it
is
now recognized
that that
proof
is
untenable,
it must
nevertheless be
admitted that the
assumption
of the existence of such adiabatics is
very
hurtful
to
one's
physical
sensibilities. It
seems
much
more promising
to start out conversely,
from the
assumption
that it is
impossible
to
attain absolute
zero
by
a
finite
process,