DOC.
21
THEORY OF RELATIVITY 247
respect
to matter.
In
particular,
the
question
arises: Does ether take
part
in
the
motions of
ponderable
matter? This
question
led the brilliant
physicist
Fizeau
to
an
experiment
of fundamental
importance,
which will
now
be discussed
schematically.
[5]
Suppose
that
a ray
of
light
L
strikes
a
semitransparent
mirror
S1
and
is
split
into
two
rays.
The first
ray
arrives
at
E
via
a
and
b,
after
having
been reflected
at
the mirror
s2
and the
Vy
semitransparent mirror
S2.
The second
ray
arrives
at
E via
c
and
d,
after
having
been reflected
at
S1
and
s1
and
having passed through
s2.
The
two
rays
interfere
at
E;
there arise interference
fringes,
the intervals of which
depend on
the
adjustment
of the
apparatus.
The
position
of
these
interference
fringes
depends on
the difference in the times needed
by
the
two
rays
to traverse
their
paths.
If this time difference
changes by
even
such
a
small
fraction
as
10-8, i.e.,
one
hundred-millionth of the total
path time,
even
this
can
be
perceived
from the
displacement
of the interference
fringes.
Fig. I.
In each of the
path segments
a
and
d,
Fizeau installed
a pipe
filled with
water,
through
which the
rays passed longitudinally.
Each of these
pipes
was
provided
with
attachments at its ends
so
that water could be made
to
flow
axially through
the
pipes.
The aim of the
experiment
was
to
ascertain the influence of the
velocity
of the
flowing
water
on
the
position
of the interference
fringes.
From this influence it
is
possible
to
calculate how fast the
light propagates through
the
moving
water
relative
to
the
stationary
pipe.
Assuming
that the luminiferous ether
participates
in
the motions of
matter,
and
thus
in
the motion of
water in
the
present case,
the
following
was
to
be
expected
for
the
case
where the
water
flows with
velocity
v
along
the
path segment
a
in
the
direction of the
propagation
of the
light.
The
velocity
of
propagation
of the
light
relative
to
the
water
will
always
equal
the
same
value
V0,
whether the
water
flows
or
not.
However,
the
velocity
of
light V
relative
to
the
pipe
will
always
have
to
exceed
V0
by
the
velocity
of the
water
flow
v.
One would therefore
expect
V
-
V0
=
v.
Since
V
-
V0
could be determined from the
displacement
of the
interference
fringes,
and the
velocity
of the
water
was
known
directly,
Fizeau's
experiment permitted
a
test
of
this
formula. But the latter
was
not confirmed
by
the
experiment.
It
turned
out
that the difference
V
- V0
is
smaller than
v. Experiments
with different
liquids
showed that this difference
depends
not
only
on v
but also
on
the refraction index
n