Homepage › Solution manuals › David A. Cox › Galois Theory › Exercise 13.2.18
Exercise 13.2.18
This exercise is concerned with the polynomials (13.28). As in the Historical Notes, we will assume that they lie in and are irreducible.
- (a)
- Show that is a root of .
- (b)
- Prove that the Galois group of over is isomorphic to .
- (c)
- Prove that over , the first two polynomials of (13.28) have cyclic Galois group while the third has Galois group isomorphic to .
Answers
Proof.
- (a)
-
As in Cardan’s method, we substitute
to
in
We obtain
Then we verify that is a solution of the system
Indeed
and, since ,
Therefore
is a root of . If we replace by another fifth root (where ), in the equation
then is replaced by , and is replace by , therefore is a solution of the preceding system, so is also a root of . So
are roots of , where
These roots are the five roots of , as proved in the following expansion:
Since
we obtain
so
Therefore the roots of are .
This was perhaps the Euler’s starting point.
- (b)
-
We obtain the discriminant of
with
R.<P,Q,u,v,e,x> = QQ[] f = x^5 - 5*P*x^2 - 5*Q*x -e; Delta = f.discriminant(x).subs(e = Q^2/P + P^3/Q).factor() Deltaso
Thus is not a square in .
Using the resolvent() function of Exercise 14, we obtain the sextic resolvent:
K.<P,Q,e> = QQ[] S.<x> =PolynomialRing(K, order = ’degrevlex’) f = x^5 - 5*P*x^2 - 5*Q*x -e theta = resolvent(f)[0]; theta.subs(e = Q^2/P + P^3/Q)which has root ( ). By Section 13.2, the Galois group of is up to conjugacy.
- (c)
-
By Exercise 13, we know that the Galois group of
over
is
which is not cyclic, so there is a misprint in the sentence.
-
The polynomial
, is irreducible over
by hypothesis. Let
be a root of
. Since
and
, we obtain that
divides
, where
. Therefore
so . If is the minimal polynomial of over , then , and divides , therefore , and we have proved that is irreducible over .
Moreover
is a square in , and is a root of the resolvent
If , and , then
But , so . doesn’t split completely over , where is a root of . By Theorem 13.2.6 and the table 13.2.C,
where is the splitting field of .
-
Now
, and
is assumed irreducible over
.
With the same proof as in the first bullet, remains irreducible over .
By part (b),
is a square in , and
has root . By part (a),
We prove that is not real. If , then , so .
Then, using and ,
Since , we would have , and then the root , in contradiction with the irreducibility of over .
So is not in the field , and doesn’t split completely over .
By the table 13.2.C, .
-
The third Euler’s polynomial is
.
If , we obtain , which is Example 13.2.10. We know from this example and from Exercise 14 that the Galois group of over is .
With the same proof as in the first bullet, remains irreducible over .
By Exercise 14,
is a square in the field , and
has the root .
It remains to know if splits completely over .
Note that
Therefore if , , so we find (see Exercise 14) that
So is a root of if
Therefore are the roots of and satisfy . If is a root, so is (where ).
Conversely
So the roots of are
where is a solution of the system
so
(where we choose the real roots, so ).
We obtained the factorization of :
If the root is real, then , so , and this imply , in contradiction with the assumed separability of .
Therefore , where is a root of , so doesn’t split completely over .
is isomorphic to .
We obtained the same Galois group for the 3 Euler’s polynomials of (13.28).