l
INTRODUCTION
TO
VOLUME
8
determined
by
the
masses
of
bodies”) nor
in the formulation
of
the
equivalence
principle
(“inertia and
gravity are
of
precisely
the
same nature”).
Both
new
formu-
lations
can
be
traced
to Einstein’s
debate
with De Sitter.
Einstein
formulated what
would
become
Mach’s
principle
to
express
what he found
objectionable
about De
Sitter’s
vacuum
solution
of
the
field
equations
with
cosmological
term.
The
formu-
lation
of
the
equivalence principle
was probably
at
least in
part
in reaction
to
attempts by
De Sitter and Mie
to
distinguish
between
an
inertial
and
a gravitational
part
of
the metric field in Einstein’s
cosmological
model.[32]
VIII
The
preceding
section
highlighted some
of
the
dramatic
changes
in Einstein’s
con-
ception
of
the foundations
of
the “Entwurf”
theory
and
general
relativity. Looking
at the debate
over energy-momentum
conservation in
general
relativity, however,
which
comprises a significant
fraction
of
the scientific
correspondence
in this vol-
ume,
one
is struck
by
the
continuity
of
Einstein’s
position.
His basic
ideas about
the role
of
energy-momentum
in the
gravitational
field
equations
and about
energy-
momentum conservation did
not
change
much
in
going
from
the final version
of
his
scalar
theory
for
static
fields[33]
to
the
“Entwurf”
theory, or
in
going
from the “Ent-
wurf”
theory
to
the
generally
covariant
versions
of
the tensor
theory
in 1915 and
1917.
At
each
stage,
Einstein
made
sure
that
gravitational energy(-momentum)
enters into the field
equations
in
exactly
the
same way
as
any
other
energy(-momen-
tum).
He
derived
an identity
satisfied
by
the
gravitational potential,
whether
repre-
sented
by a
scalar
or
by a
metric
tensor,
with the
help
of
which he could derive
energy-momentum
conservation from the field
equations.
Once Einstein found
a
Lagrangian
formulation of
the
theory,[34]
he derived this
identity
from the invari-
ance
of
the action
integral
under coordinate transformations-under
“justified”
transformations in the “Entwurf”
theory,
under
arbitrary
coordinate transforma-
tions in the final
theory.
The
developments
for the
theory
of
November
1915
can
be traced in the
corre-
spondence
in this volume. In
a
letter
to Ehrenfest
of
early
1916
(Doc. 185),
which
can
be
seen as
a blueprint
for the sections
on
the field
equations
and
energy-
momentum conservation in
Einstein 1916e
(Vol. 6,
Doc.
30),
Einstein
showed how
energy-momentum
conservation
can
be derived from the field
equations. Shortly
after the
publication
of
Einstein
1916o
(Vol.
6,
Doc.
41),
he
explained
to
Ehrenfest
(Doc. 275)
that
the
identity
needed
for
such
a
derivation follows from the invari-
ance
of
the action
integral
for the
theory
under
arbitrary
coordinate transforma-
tions.
Shortly
before
submitting
the
paper,
he had told
Ehrenfest
that the
purpose
of
the
Lagrangian
formulation
of
the
theory
had been to show the intimate
con-