342
DOC.
17
THE
THEORY
OF RELATIVITY
The front and
rear
ends
of
the
tube
in
the
accompanying
sketch
are
closed
by
a
glass
plate.
Pipe
connections attached
at
the
two ends
make
it
possible
for
a
liquid
to flow
through
the tube
in
the direction of
its axis.
How does
the
velocity
with which
the
liquid
flows
through
the tube
affect
the
propagation
velocity
of
a
light ray passing axially
through
the tube? If
it
is
true
that the luminiferous ether
moves
with
the
matter
that
flows
through
the
tube,
then the
following
picture
obtains.
If
we assume
that
in
water
at rest
light
propagates
with
velocity
V,
V thus
being
the
velocity
of
light
relative to
water,
while
v
is
the
velocity
of the
water relative to
the
tube,
then
we
must
say:
If
the
luminiferous ether adheres
to
the
water,
then the
velocity
of
light
relative
to
the
water
is
always
the
same, regardless
of whether the
water
is
in
motion
or
not.
Accordingly, one
should
expect
that the
velocity
of the
propagation
of
light
relative to
the tube
is greater
by
v
if
the
liquid is
in
motion than
if it is at rest.
In Fizeau's
experiment,
one
of
two
beams
of
light
capable
of interference traversed
the
tube
in
the
manner
described. From
the influence of the
known
velocity
of motion of the
liquid
on
the
position
of the
interference
fringes,
it
was
possible
to
calculate the influence that the
water
moving
with
velocity v
exerted
on
the
velocity
of the
propagation
of
light
relative to
the tube
at rest.
Fizeau found that the motion of the
liquid
did not
increase the
velocity
of
light
relative
to
the tube
by
v,
but
only by a
fraction of
this value
(v
1-1
n2
if
n
denotes the
refractive
capacity
of the
liquid).
If
this
refractive
capacity
is
very
close to
1,
i.e.,
if
light
propagates
almost
as
fast in
the
liquid
as
in
empty
space,
then the motion of the
liquid
has
practically
no
influence.
From
this it
had
to be
concluded that the
conception
according
to which
light always
propagates
with
the
same velocity
V
relative
to water
is
not
compatible
with
experience.
The
next
simplest hypothesis was
that the luminiferous ether
does not
participate
in
the motion of the
matter.
From
this
hypothesis,
as
a
basis, we
cannot
deduce
in such
a
simple manner
how
the
optical
phenomena
are
influenced
by
the motion of
matter. But
in
the
mid-'90s H. A.
Lorentz succeeded
in
formulating
a
theory
based
on
the
assumption
of
a
completely stationary
luminiferous ether.
His
theory
provides a
completely
correct
account
of almost
all known
phenomena
in
the
optics
and
electrodynamics
of
moving
bodies,
including
the
experiment
of Fizeau
we
have
just
discussed.
Let
me
add
at
once
that
a theory fundamentally
different from that of
Lorentz,
which would
be based
on
simple
and intuitive
assumptions
and would
accomplish
the
same ends,
could not
be
formulated.
For
that
reason,
the
theory
of the
stationary
luminiferous
ether
had to
be
accepted
for the
time
being
as
the
only
theory
compatible
with
the
totality
of
experience.
[4]
Let
us now
consider
this
theory
of the
stationary
ether from the
standpoint
of the
principle
of
relativity.
If
we
designate
as
acceleration-free
all
those
systems
with
respect
to which
material
points
not
subjected
to
external forces
move uniformly,
the
principle
of
relativity
states
the
following:
The
laws
of
nature
are
identical
in all
acceleration-free