8

NATURE OF

MOLECULAR FORCES

deprecatingly

about his two

papers

on

molecular

forces. He sent Stark

a

collection of

his

offprints, omitting

"my two worthless

beginners'

works"

("meine zwei

wertlosen

Erstlingsarbeiten").[36]

Einstein

published one

final

paper on capillarity

and molecular

forces,

Einstein

1911a.

In this

paper,

he

demonstrated,

without

mentioning

his earlier

papers,

that

the

assumption

of

a

universal

range

for the molecular cohesive force

is

incompatible

with certain

empiri-

cally

established laws

obeyed by

the surface tension. He stressed the need

to

introduce

a range

for the force between two molecules that

is

dependent on

the nature

of

the

molecules.

Einstein

1902a

(Doc.

2) opens

with

a

discussion

of

the conditions for the

validity

of

the

second law

of

thermodynamics, a

discussion that

proved

to be

of

some significance

for

Einstein's

work. He

points

out that the theories

of

dissociation and

of

dilute

solutions

are

based

on

the

application

of

the second law

of

thermodynamics

to

processes

involving

idealized

semipermeable

membranes that

can separate any

two

(or more)

substances.

Al-

though even

the

approximate realizability

of

such

processes

is

often

doubtful, the

predic-

tions

of

the theories

are

confirmed

by experiment.[37]

It

seems,

therefore,

that

one can

draw

valid conclusions from the consideration

of

highly

idealized

processes.

In

particular,

it

seems

that

one may

apply

the second

law

to

mixtures acted

on by

external

conservative

forces,

which

produce

the

same

effects

as

semipermeable

membranes.

Generalizing

this

conclusion, Einstein formulated the

hypothesis

that

the

second law

may

be

applied

to

mixtures, the

components

of

which

are subjected

to

arbitrary

conservative

forces. The

arguments

in the

body

of

the

paper

make extensive

use

of

this

hypothesis.

Einstein's

next

paper,

which initiated his

study

of

the

statistical foundations

of

thermodynamics,

may

have

been

stimulated

by

the

need he

felt

to

justify

this

hypothesis.

In

any

case,

the

final

section

of

that

paper

is devoted

to

providing

such

a

justification.[38]

[36]

Einstein

to Stark, 7

December 1907. He

did not mention his

theory

of

molecular forces

in his Autobiographical Notes

(see

Einstein

1979).

[37]

See

Planck

1891 for

a

discussion of

the

role

of

ideal

processes

in

thermodynamics.

Nernst

1898,

pp.

102-103, discusses

the

ther-

modynamical

use

of

idealized

semipermeable

membranes.

See also

Einstein

1902a

(Doc. 2),

note 1.

[38]

See Einstein 1902b

(Doc. 3),

p.

433, and

the editorial

note,

"Einstein

on

the Foundations

of

Statistical

Physics,"

p.

46.