DOC.

55

FEBRUARY

1915 67

55.

To Michael Polányi

[Berlin,]

10 February 1915

Dear

Colleague,

We

are now

just

dealing

with the

case

that

absolute

zero

is

attainable

through

finite

repetitions

of

your

process,[1]

We

would

like

to arrive at

A.

Well,

you

say:

Now

we

enlarge

V

more

adiabati-

cally

until

we

have arrived at B.

Af-

ter this,

isothermally

until

C,

then

adiabatically

to

B, etc.

The conclu-

sion

thus

is

that the

isotherm B

C

corresponds

no more

to

an

entropy

difference

than the adiabat

C B

does,

q.e.d.

CAB

This

proof applies

in

the

case

that the adiabats

A

B

or

C B

actually

exist.

E.g.,

in

the

case

of

a

change

in volume

of

a

solid

body,

it

is

difficult

to

doubt

the

existence

of

that

zero-adiabat. But whether such kinds of

adiabats

always

exist

is

very questionable.

Then

the

petitio

principii

lies

in the

assumption

of

the

existence of

the

zero-adiabat. There

are cases

in which

this

existence

(in

principle)

is

highly improbable.

Think

of

a

fixed

solution,

in which

the

[...][2]

An adiabatic

re-

versible

change

in

volume of solution

V

at

absolute

zero

obvi-

ously

does

not

exist

here in principle.[3]

Volume

=

V

semipermeable membrane

Solution

Pure

solvent

Solvent

This

is

suspicious,

anyway,

in all

cases

where

the adiabats

touch

the

(T

=

0)-

axis

at

angles

other than

0

[degrees].

With

greetings

from

your

colleague

A.

Einstein.

I wish

you

luck

in

performing your military

service.