20 DOC.

10

MAY 1914

satisfy

the relation

udQ

=

dlgW,

where

lg

W

=

lJdrflgy.

(6)

However,

contrary

examples

of such

G(q, p,

a)'s

also exist in which

udQ

=

dlgW

is

violated.

(7)

In all extensions of

the

quanta

approach,

we

remain within

the

weight

class

G(q,

P, a) =

r(i).

(8)

Then

I

hope

also to be able to

present

to

you orally

my

reflections

on

why

the

Fokker

systems satisfy

the

2nd law

even

though

it is

impossible

to

describe

them

as

the

“most

probable”

distributions

(!!!!).

(9)

As

a

closing

remark

[Schlussbemerkung]

a) Planck,

Einstein,

Debye

work with

G(q, p,

a),

therefore it

is

worthwhile

to

examine

why

these

people

do

come up

with

udQ

=

dlgW

with

such anti-

Boltzmann

spirited

G's.

b)

Only

once

did

anyone

work

with

G(q,

p,

a,

T):

Herzfeld.[14]

This

displeases

me.

c)

Ideal

gas can

“of

its

own

accord”

happen to

shrink to

half

its

volume

on one

occasion and

to

a

1/3

on

another,

as

if labeled item

(a)

had

compelled

it.-Classical

Hertzian resonators at

frequency

v0

could

“by

coincidence” all be

“stunned” at

once.

On arrival

at

the Planck

ellipses they

will

belong

in

frequency

v(a1,

B1),

in

another instance

by

chance

to

Planck

ellipses

that

belong

in

frequency

u(a2,

B2)-as

if

they had

been

pressed

on

to

these

ellipses through

corresponding

a,

ß

values

with the

aid of the

quantum

hypothesis

lever.-Calculate

the

quotients

of

the

probabilities of

both

these coincidences.

Yes

sir-this would be

the

entropy

calculation in Boltzmann’s

spirit.

You

see

I

understand

your

comment.

But

did

Planck,

you,

and

Debye

calculate it

like

this?-

No!-

Rather with

G(q,

p,

a)

see

e.g.,

Einstein,

Ann.

d. Phys.

22

(1907)

p.

182 bottom.[15]

Einstein:[16] "You

are

inclined to

the

view

that

lg

W deviates from

thermodynamic

entropy.

You

prove

that

this

is

really

the

case

if

the

probability

function

G(q,

p,

a)

depends

on

the

a

parameters.

(1)

I

believe, however,

that the

assumption

of such

a

dependency

is

not

permissible,

that

it

is

even

completely

contrary

to Boltzmann’s

conception.

(2)

For if

you

only

have

a

look

at

the

states which

the

system

assumes

of

its

own

accord

in the

course

of

time,

it

is

evident

that then

variable

parameters

a

are

altogether

inconceivable

as

independent

entities.

(3)

The

scale

by

which

the