FOUNDATIONS OF

STATISTICAL

PHYSICS 45

Aside from his

reading

of

material

directly

concerned with kinetic

theory,

Einstein

was

studying

at least three other

topics

in 1901

and 1902 that

may

have

suggested

the need to

extend the foundations

of

thermodynamics

and kinetic

theory.[35]

First,

since at least the

spring

of

1901,

Einstein had been

reading

Planck's

papers on

irreversible radiation

processes,

in which Planck

sought

to extend the

concept

of

entropy

to radiation.[36]

Second,

Einstein

was studying

the work

of

Drude and others

on

the electron

theory

of

metals,

in which the

apparatus

of

kinetic

theory

is

employed

to

explain

such

phenomena

as

electrical and thermal

conductivity.[37] Especially important

for

Einstein's

conception

of

the foundations

of

statistical

physics was

his

reading

in

late

May

1901

of

a

paper

by

Reinganum

that convinced him

completely

of

the

validity

of

the electron

theory.[38]

Rein-

ganum

showed that the

principal

results

of Drude's

electron

theory can

be

derived

without

any assumptions

about the nature

of

the

elementary charge

carriers

or

the mechanism

of

their interactions with the atoms

of

a

metal. The

only

crucial

assumption

is

that

the

equi-

partition

theorem

applies

to

a gas

of

moving charge

carriers.

Reinganum

concluded:

It

appears

therefore

.

. .

that

in metals,

too,

electricity moves

in discrete

quantities

with

the

magnitude

of

the

charges on electrolytic

ions,

and that the

principles

of

the

gas

theory

are

applicable

to

the

masses moving

with the

charges.

Es erscheint daher

. . .

dass auch

in

den Metallen die Elektricität in

discreten

Mengen von

der Grösse

elektrolytischer Ionenladungen

sich

bewegt,

und dass

für die mit den

Ladungen bewegten

Massen die

Principien

der Gastheorie

anzuwenden sind.[39]

Reinganum's

results

helped

to

confirm

Einstein's

commitment to

an

atomistic view

of

electricity, as

well

as

of

matter;

but

they may

also have

suggested

that the

equipartition

theorem,

as a

vital

principle

of

statistical

physics,

needed

a more general

mechanical foun-

dation than it had hitherto been

given.

In

fact,

one

of

the

major

results of

Einstein

1902b

(Doc.

3)

is

precisely

a

demonstration that the

equipartition

theorem

is

valid not

only

for

a

[35]

Anticipations

of Einstein's

later distinc-

tion between

"theories

of

principle"

and

"con-

structive

theories"

(see

Einstein

1919) may

also

have inclined him

to

seek

more

general

founda-

tions for

thermodynamics

and the kinetic

theory.

For

a

discussion

of

the

distinction, see

the Intro-

duction,

pp.

xxi-xxii.

[36]

Einstein

apparently

studied

one

or

both

of

Planck

1900a, 1900b;

see

Einstein to Mileva

Maric,

4

April

1901

and

10

April

1901

(Vol. 1,

Docs. 96 and

97).

For

a

discussion

of Einstein's

interest in

Planck's

work,

see

Vol.

1,

the edito-

rial

note,

"Einstein

on

Thermal, Electrical,

and

Radiation Phenomena,"

pp.

235-237.

See also

Einstein

to Maric,

second

half

of

May 1901,

4

June

1901,

and

7

July

1901 (Vol.

1,

Docs.

110,

112,

and

114),

as

well

as

Einstein to Jost Win-

teler, 8 July 1901 (Vol.

1,

Doc.

115).

[37]

His

reading

included at least

Drude

1900a,

1900b

(see

Einstein

to Mileva

Maric,

4

April

1901,

and 28

May 1901,

Vol.

1,

Docs. 96

and

111).

For

a

discussion

of Einstein's

interest

in the electron

theory

of

metals, see

Vol.

1,

the

editorial

note,

"Einstein

on Thermal, Electrical,

and Radiation Phenomena,"

pp.

235-237.

[38]

See

Reinganum

1900. For Einstein's

com-

ments

on

it,

see

Einstein

to

Mileva

Maric,

28

May

1901 (Vol.

1,

Doc.

111).

[39]

Reinganum

1900,

p.

403.

Reinganum's

comments

concern primarily

Drude

1900a.