DOC.

45

239

rigidly electrified

body,

and in

the

second

case

it consists

of

a

number

of

uniformly

moving mass

points which

do

not

exert

any

forces

on

each

other.

Before

I start

with the investigation, I

must

insert here

a

remark

on

the

presumed

range

of

validity

of

Maxwell's

equations

for

empty

space

so as

to

meet

a

naturally

arising

objection. In

earlier

papers

I

showed

that

our

present electromechanical world

view

is

not suited

for

explaining

the entropic

properties

of

radiation

and

the

laws

governing

the emission

and

absorption

of

radiation

and

those

governing

the

specific

heat; rather, it

seems

to

me

necessary

to

assume

that the

nature of

any

periodic

process

is

such

that

the

conversion of

energy

can

only proceed

in certain

quanta of

finite

magnitude

(light quanta),

i.e., that the

manifold

of

processes

possible in reality is

smaller than the manifold of

processes

possible

according

to

our

present

theoretical

views.1

In

particular,

we

would

have

to

imagine

that

in

a

radia-

tion

process

the instantaneous

electromagnetic state

in

a

region

of

space

is

completely

determined

by a

finite

number

of

quantities--in

contrast to

the

vector theory

of radiation.

But

as

long

as we

do not

possess

a

picture

that

corresponds to

the

requirements

mentioned, it will

be

natural

to

use

the

current

theory

for all

problems

not

concerned with

entropy

relations

or

conversions

of

elementary

small quantities of

energy

without

having

to fear

that

we

will

thereby

arrive

at

incorrect results.

I

can

illustrate

most

graphically

how

I

see

the

present

situation

regarding

these questions with the

following imaginary

case.

Let

us imagine

that the molecular-kinetic

theory

of

heat

has not

yet

been

propounded,

but that it

has been

demonstrated with

complete

certainty

that the

Brownian motion

(motion

of

particles

suspended

in liquids) is

not due

to

any

external

supply

of

energy,

while it is

clearly

recognized

that these

motions

cannot

be

explained

with

the

help

of

mechanics and

thermodynamics.

In

such

a

situation

one

would

rightly conclude that

a

radical

change

of

theore-

tical principles

must

take

place.

In

spite of

that,

nobody

would shrink

from

applying

the fundamental

equations

of

mechanics and

thermodynamics

to

handle

problems

not

related

to instantaneous

states

in small

regions

of

space.

In

this

sense,

in

my

opinion,

we

can

confidently

base

our

considerations

on

Maxwell's equations.

1A.

Einstein,

Ann.

d.

Phys.

17 (1905):

132; 20

(1906):

199;

and

22 (1907):

180.

[3]