0PD8
Proof. Let be a non-zero element of
.
Let . Given , we put
. We put
.
Let be two distinct elements of and
let .
If , then
.
Assume and .
We have
where
|
|
|
We have
|
|
|
Ir follows that .
Assume finally and .
We have where
|
|
|
|
|
|
|
|
|
We have
|
|
|
It follows that .
It follows from Remark 7.4.11 that .
Define and as above.
Let . We have and
for in
.
Given in with , we have
.
Consider now another
non-zero element of and assume .
We have and
. The discussion above shows that
for .
Note also that for .
As a consequence, if .
Let , , ,
and .
Since , it follows that
, ,
, and
. We deduce that for
.
We proceed now by induction on to show
that determines , for .
Assume there is such that .
Let and . We have
. Define an element of
as follows.
Given , we put
if ,
if . Given , we
put . Given , we put
. This defines an element of
. Furthermore, .
We define similarly , ,
and starting with and .
We have and , hence also .
We have
, hence by
induction. Since and
, it follows that .
Assume for all .
We have .
Note that for .
Let be the unique
increasing bijection.
We define
and an element of as follows.
We put for ,
for
and
for .
Let , and . We have
|
|
|
Define similarly, starting with instead
of .
We have . By induction, we deduce
that , hence .
This completes the proof that the restriction of to is injective.
We deduce that the restriction of to is injective using
Remark 8.2.6
∎