346

DOC.

31

VARIATIONAL PRINCIPLE

[2]See

Einstein

1916o (Doc. 41)

for

a

published paper

on

the

same topic

that shows

some

similarities with this

manuscript (see note

6

for

a

major

difference).

See also Cattani and De

Maria

1993

and

Kichenassamy

1993

for historical discussion of variational

principles

in

gen-

eral

relativity.

[3]At this

point

in

the

original text

Einstein indicates

a

note

he has

appended at

the foot of

the

page:

"Die

von

Hilbert

im

Anschluss

an

Mie

eingeführte Voraussetzung,

dass sich die

Funktion

D

durch

die

Komponenten

eines Vierervektors

und

dessen

erste

Ableitungen

darstellen

lasse,

halte

ich für

wenig

aussichtsvoll."

Einsteinqp

refers

to

the

theory developed

by

Hilbert in Hilbert

1915.

It

is based

on

Gustav Mie's

theory

of

matter,

which

implied

in

partic-

ular that the

energy-momentum

tensor

of

matter

should be

a

function

of

electromagnetic quan-

tities

only.

Later

in 1916

Einstein criticized Hilbert's

theory

rather

vehemently, calling

his

assumptions

for the Hamiltonian

"infantile, in

the

sense

of

the

child,

who

is innocent

of the

guile

of

the

external world"

("kindlich,

im Sinne des

Kindes,

das

keine

Tücken der Aussenwelt

kennt." Einstein

to

Hermann

Weyl,

23

November

1916).

See

also Mehra

1973

and Earman and

Glymour

1978b for historical discussions of Hilbert's

and

Einstein's theories.

[4]At this

point

in

the

original text

Einstein indicates

a

note

he has

appended at

the foot

of

the

page:

"Hilbert und Lorentz haben

zuerst

diesen

Weg

eingeschlagen." In

the

corresponding

passage

in

Einstein 1916o

(Doc. 41),

there

is

a

reference

to

Hilbert

1915

and Lorentz

1915,

1916b, 1916c,

and

1916d.

[5]All references

to equations

with numbers lower than

47

are

to equations

in

Einstein 1916e

(Doc. 30).

[6]In contrast to

the

discussion

in this

manuscript,

Einstein

left

completely unspecified

the

form of the Hamiltonian

in

Einstein 1916o

(Doc. 41).

The choice made here

was

already

dis-

cussed

by

Einstein

in

January

1916 in

two

letters

to

Lorentz

(see

Einstein

to

H. A. Lorentz,

17

January

1916

and

19 January 1916).

[7]"m"

should

be

"0."

[8]In

Einstein 1916o

(Doc.

41),

eq.

(90) is

used

to

derive the

energy-momentum

conserva-

tion laws. The

same

could

be

done

here,

starting

from

eq.

(90)

and the

explicit

form of

0,

but

as

Einstein admitted

a

few

months later

he

did

not

carry

out

the calculation

(see

Einstein

to

Theophile

de Donder, 23

July

1916).