DOC. 197 JANUARY 1910
147
effective
deceleration
is
really mainly
rectilinear,[4]
so
that
the
elementary process seems
to be
completely
determined
by
the
electron,
its
direction of
motion,
and the orientation
of
the
wall.
Hence
I do
not
see
any
factor that
could
prevent
this
entire
elementary
process,
including
the
Roentgen
emission
(1st.
fraction),[5]
from
being completely
symmetric
about
this
axis
of
motion. But the
assumption
of
a
lateral direction of
emission
does
not
fit in at all with this
symmetry;
on
the
contrary,
it
violates
it.
One
might actually
view this
as a
proof
against
the
assumption
of directed
emission.
And
now
the counter-advocate!
Assume
the
above
process
and
assume
in
addition
that
the
system
is
enclosed
by a
spherical
shell
that
is impermeable
to
Roentgens
and has
a
small hole in
it.
Behind
the hole there
is
a
metal
plate
P.
We
know
that
in this
case
the
metal
plate
emits
secondary
rays
whose
kinetic
energy
is
of
the
same
order of
magnitude
as
that of
the
incident
electrons
E, and
is independent
of
the
lumen of
the
screen opening.
Does the
plate
P
possess
the
property
of
frugally storing
up fragments
of
the
spherical Roentgen
waves
until it
is
able
to
outfit
one
of
its
electronic
chil-
dren
with
a
respectable
amount
of
energy so
that
the
latter
can
make
its
trip through
space
with
a
vehemence
befitting
its
Roentgenian birth?[6]
Which
of the advocates
is
the
mischief-maker.
I
think
that,
as was
the
case
prior to
the
relativity theory,
here,
too,
a
prejudice is
the
source
of the
difficulties.
I
know
of
such
a
prejudice, though
I
do
not know
whether it
is
essential.
Namely,
the localization
that
we
ascribe to
the
electromagnetic
energy
in Maxwell's
theory is totally
arbitrary;[7]
but
so
far
this
insight
has
not
helped
me
much
in
clarifying
the
question.
Or
maybe
the
electron
is
not
to
be conceived
as
such
a simple
structure
as we
think? There
is nothing
one
would
not
consider
when
one
is
in
a
predicament!
And
now
to
the other
problem child,
the
rigid body.[8]
I
occupy
myself very
little with
it.
For
it
seems
to
me
that
the
empirical
data
do not suffice for
the construction of
a
theory
of
arbitrarily
accelerated
bodies.
Had
it not
been
for
Fizeau's
experiment
and
the
measurements
concerning
the
velocity
of
light
in
vacuum,
we
would
not
have
had
the
material needed for the construction of the
relativity theory;
we
are,
in
my
opinion,
in
a
similar
situation
with
respect
to
acceleration. It
is
only
about
infinitely slowly
accelerated
systems
that
anything
at all
can
be
asserted
at
the
moment, in
my opinion.
Nevertheless,
one
should
try
to
devise
hypotheses
about
the
behavior of
rigid
bodies that
would allow
a
uniform rotation.
I
now
have
very
little time
because
my
new
job
demands
more
than
I
thought
it
would;[9]
this
is
due
to
my
poor memory
and to
the
circumstance
that
until
now
I have
dabbled
in
my
area
of
specialization
only as a
kind
of
amateur.
With
cordial
greetings, yours very truly,
A.
Einstein
Moussonstr.
12