ANT10

Introduction to Algebraic Number Theory
F. Oggier
2
A few words
These are lecture notes for the class on introduction to algebraic number theory,
given at NTU from January to April 2009 and 2010.
These lectures notes follow the structure of the lectures given by C. W uthrich
at EPFL. I would like to thank Christian for letting me use his notes as basic
material.
I also would like to thank Martianus Frederic Ezerman, Nikolay Gravin and
LIN Fuchun for their comments on these lecture notes.
At the end of these notes can be found a short bibliography of a few classical
books relevant (but not exhaustive) for the topic: [3, 6] are especially friendly
for a rst reading, [1, 2, 5, 7] are good references, while [4] is a reference for
further reading.
3
4
Contents
1 Algebraic Numbers and Algebraic Integers 7
1.1 Rings of integers . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2 Norms and Traces . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 Ideals 19
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2 Factorization and fractional ideals . . . . . . . . . . . . . . . . . 22
2.3 The Chinese Theorem . . . . . . . . . . . . . . . . . . . . . . . . 28
3 Ramication Theory 33
3.1 Discriminant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2 Prime decomposition . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.3 Relative Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.4 Normal Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4 Ideal Class Group and Units 49
4.1 Ideal class group . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.2 Dirichlet Units Theorem . . . . . . . . . . . . . . . . . . . . . . . 53
5 p-adic numbers 57
5.1 p-adic integers and p-adic numbers . . . . . . . . . . . . . . . . . 59
5.2 The p-adic valuation . . . . . . . . . . . . . . . . . . . . . . . . . 62
6 Valuations 67
6.1 Denitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
6.2 Archimedean places . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.3 Non-archimedean places . . . . . . . . . . . . . . . . . . . . . . . 71
6.4 Weak approximation . . . . . . . . . . . . . . . . . . . . . . . . . 74
5
6 CONTENTS
7 p-adic elds 77
7.1 Hensels way of writing . . . . . . . . . . . . . . . . . . . . . . . . 79
7.2 Hensels Lemmas . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7.3 Ramication Theory . . . . . . . . . . . . . . . . . . . . . . . . . 85
7.4 Normal extensions . . . . . . . . . . . . . . . . . . . . . . . . . . 86
7.5 Finite extensions of Q
p
. . . . . . . . . . . . . . . . . . . . . . . . 88
Chapter 1
Algebraic Numbers and Algebraic
Integers
1.1 Rings of integers
We start by introducing two essential notions: number eld and algebraic inte-
ger.
Denition 1.1. A number eld is a nite eld extension K of Q, i.e., a eld
which is a Q-vector space of nite dimension. We note this dimension [K : Q]
and call it the degree of K.
Examples 1.1. 1. The eld
Q(
2) = x + y
2 [ x, y Q
is a number eld. It is of degree 2 over Q. Number elds of degree 2 over
Q are called quadratic elds. More generally, Q[X]/f(X) is a number eld
if f is irreducible. It is of degree the degree of the polynomial f.
2. Let
n
be a primitive nth root of unity. The eld Q(
n
) is a number eld
called cyclotomic eld.
3. The elds C and R are not number elds.
Let K be a number eld of degree n. If K, there must be a Q-linear
dependency among 1, , . . . ,
n
, since K is a Q-vector space of dimension n.
In other words, there exists a polynomial f(X) Q[X] such that f(X) = 0.
We call an algebraic number.
Denition 1.2. An algebraic integer in a number eld K is an element K
which is a root of a monic polynomial with coecients in Z.
7
8 CHAPTER 1. ALGEBRAIC NUMBERS AND ALGEBRAIC INTEGERS
Example 1.2. Since X
2
2 = 0,
2 Q(
2) is an algebraic integer. Similarly,

i Q(i) is an algebraic integer, since X
2
+1 = 0. However, an element a/b Q
is not an algebraic integer, unless b divides a.
Now that we have the concept of an algebraic integer in a number eld, it is
natural to wonder whether one can compute the set of all algebraic integers of
a given number eld. Let us start by determining the set of algebraic integers
in Q.
Denition 1.3. The minimal polynomial f of an algebraic number is the
monic polynomial in Q[X] of smallest degree such that f() = 0.
Proposition 1.1. The minimal polynomial of has integer coecients if and
only if is an algebraic integer.
Proof. If the minimal polynomial of has integer coecients, then by denition
(Denition 1.2) is algebraic.
Now let us assume that is an algebraic integer. This means by denition
that there exists a monic polynomial f Z[X] such that f() = 0. Let g Q[X]
be the minimal polyonial of . Then g(X) divides f(X), that is, there exists a
monic polynomial h Q[X] such that
g(X)h(X) = f(X).
(Note that h is monic because f and g are). We want to prove that g(X)
actually belongs to Z[X]. Assume by contradiction that this is not true, that
is, there exists at least one prime p which divides one of the denominators of
the coecients of g. Let u > 0 be the smallest integer such that p
u
g does
not have anymore denominators divisible by p. Since h may or may not have
denominators divisible by p, let v 0 be the smallest integer such that p
v
h has
no denominator divisible by p. We then have
p
u
g(X)p
v
h(X) = p
u+v
f(X).
The left hand side of this equation does not have denominators divisible by p
anymore, thus we can look at this equation modulo p. This gives
p
u
g(X)p
v
h(X) 0 F
p
[X],
where F
p
denotes the nite eld with p elements. This give a contradiction,
since the left hand side is a product of two non-zero polynomials (by minimality
of u and v), and F
p
[X] does not have zero divisor.
Corollary 1.2. The set of algebraic integers of Q is Z.
Proof. Let
a
b
Q. Its minimal polynomial is X
a
b
. By the above proposition,
a
b
is an algebraic integer if and only b = 1.
Denition 1.4. The set of algebraic integers of a number eld K is denoted
by O
K
. It is usually called the ring of integers of K.
1.1. RINGS OF INTEGERS 9
The fact that O
K
is a ring is not obvious. In general, if one takes a, b two
algebraic integers, it is not straightforward to nd a monic polynomial in Z[X]
which has a + b as a root. We now proceed to prove that O
K
is indeed a ring.
Theorem 1.3. Let K be a number eld, and take K. The two statements
are equivalent:
1. is an algebraic integer.
2. The Abelian group Z[] is nitely generated (a group G is nitely generated
if there exist nitely many elements x
1
, ..., x
s
G such that every x G
can be written in the form x = n
1
x
1
+ n
2
x
2
+ ... + n
s
x
s
with integers
n
1
, ..., n
s
).
Proof. Let be an algebraic integer, and let m be the degree of its minimal
polynomial, which is monic and with coecients in Z by Proposition 1.1. Since
all
u
with u m can be written as Z-linear combination of 1, , . . . ,
m1
, we
have that
Z[] = Z Z . . . Z
m1
and 1, , . . . ,
m1
generate Z[] as an Abelian group. Note that for this
proof to work, we really need the minimal polynomial to have coecients in Z,
and to be monic!
Conversely, let us assume that Z[] is nitely generated, with generators
a
1
, . . . , a
m
, where a
i
= f
i
() for some f
i
Z[X]. In order to prove that is
an algebraic integer, we need to nd a monic polynomial f Z[X] such that
f() = 0. Let N be an integer such that N > deg f
i
for i = 1, . . . , m. We have
that
N
=
m
j=1
b
j
a
j
, b
j
Z
that is
j=1
b
j
f
j
() = 0.
Let us thus choose
f(X) = X
N
j=1
b
j
f
j
(X).
Clearly f Z[X], it is monic by the choice of N > deg f
i
for i = 1, . . . , m, and
nally f() = 0. So is an algebraic integer.
Example 1.3. We have that
Z[1/2] =
_
a
b
[ b is a power of 2
_
is not nitely generated, since
1
2
is not an algebraic integer. Its minimal poly-
nomial is X
1
2
.
Corollary 1.4. Let K be a number eld. Then O
K
is a ring.
Proof. Let , O
K
. The above theorem tells us that Z[] and Z[] are nitely
generated, thus so is Z[, ]. Now, Z[, ] is a ring, thus in particular and
Z[, ]. Since Z[] and Z[] are subgroups of Z[, ], they are nitely
generated. By invoking again the above theorem, and O
K
.
Corollary 1.5. Let K be a number eld, with ring of integers O
K
. Then
QO
K
= K.
Proof. It is clear that if x = b QO
K
, b Q, O
K
, then x K.
Now if K, we show that there exists d Z such that d O
K
(that
is d = O
K
, or equivalently, = /d). Let f(X) Q[X] be the minimal
polynomial of . Choose d to be the least common multiple of the denominators
of the coecients of f(X), then (recall that f is monic!)
d
deg(f)
f
_
X
d
_
= g(X),
and g(X) Z[X] is monic, with d as a root. Thus d O
K
.
1.2 Norms and Traces
Denition 1.5. Let L/K be a nite extension of number elds. Let L.
We consider the multiplication map by , denoted by
, such that
: L L
x x.
This is a K-linear map of the K-vector space L into itself (or in other words, an
endomorphism of the K-vector space L). We call the norm of the determinant
of
, that is
N
L/K
() = det(
) K,
and the trace of the trace of
, that is
Tr
L/K
() = Tr(
) K.
Note that the norm is multiplicative, since
N
L/K
() = det(
) = det(
) = det(
) det(
) = N
L/K
()N
L/K
()
while the trace is additive:
Tr
L/K
(+) = Tr(
+
) = Tr(
) = Tr(
)+Tr(
) = Tr
L/K
()+Tr
L/K
().
In particular, if n denotes the degree of L/K, we have that
N
L/K
(a) = a
n
N
L/K
(), Tr
L/K
(a) = aTr
L/K
(), a K.
1.2. NORMS AND TRACES 11
Indeed, the matrix of
a
is given by a diagonal matrix whose coecients are all
a when a K.
Recall that the characteristic polynomial of L is the characteristic poly-
nomial of
, that is
L/K
(X) = det(XI
) K[X].
This is a monic polynomial of degree n = [L : K], the coecient of X
n1
is
Tr
L/K
() and its constant term is N
L/K
().
Example 1.4. Let L be the quadratic eld Q(
2), K = Q, and take

Q(
2). In order to compute
, we need to x a basis of Q(
2) as Q-vector
space, say
1,
2.
Thus, can be written = a + b
2, a, b Q. By linearity, it is enough to
compute
on the basis elements:
(1) = a + b
2,
2) = (a + b
2)
2 = a
2 + 2b.
We now have that
_
1,
2
_
_
a 2b
b a
_
. .
M
=
_
a + b
2, 2b + a
2
_
and M is the matrix of
in the chosen basis. Of course, M changes with a

change of basis, but the norm and trace of are independent of the basis. We
have here that
N
Q(
2)/Q
() = a
2
2b
2
, Tr
Q(
2)/Q
() = 2a.
Finally, the characteristic polynomial of
is given by
L/K
(X) = det
_
XI
_
a b
2b a
__
= det
_
X a b
2b X a
_
= (X a)(X a) 2b
2
= X
2
2aX + a
2
2b
2
.
We recognize that the coecient of X is indeed the trace of with a minus
sign, while the constant coecient is its norm.
We now would like to give another equivalent denition of the trace and
norm of an algebraic integer in a number eld K, based on the dierent
roots of the minimal polynomial of . Since these roots may not belong to K,
we rst need to introduce a bigger eld which will contain all the roots of the
polynomials we will consider.
Denition 1.6. The eld

F is called an algebraic closure of a eld F if all the
elements of

F are algebraic over F and if every polynomial f(X) F[X] splits
completely over

F.
We can think that

F contains all the elements that are algebraic over F, in
that sense, it is the largest algebraic extension of F. For example, the eld of
complex numbers C is the algebraic closure of the eld of reals R (this is the
fundamental theorem of algebra). The algebraic closure of Q is denoted by

Q,
and

Q C.
Lemma 1.6. Let K be number eld, and let

K be its algebraic closure. Then
an irreducibe polynomial in K[X] cannot have a multiple root in

K.
Proof. Let f(X) be an irreducible polynomial in K[X]. By contradiction, let
us assume that f(X) has a multiple root in

K, that is f(X) = (X)
m
g(X)
with m 2 and g() ,= 0. We have that the formal derivative of f
(X) is given
by
f
(X) = m(X )
m1
g(X) + (X )
m
g
(X)
= (X )
m1
(mg(X) + (X )g
(X)),
and therefore f(X) and f
(X) have (X )
m1
, m 2, as a common factor
in

K[X]. In other words, is root of both f(X) and f
(X), implying that the

minimal polynomial of over K is a common factor of f(X) and f
(X). Now
since f(X) is irreducible over K[X], this common factor has to be f(X) itself,
implying that f(X) divides f
(X). Since deg(f
(X)) < deg(f(X)), this forces

f
(X) to be zero, which is not possible with K of characteristic 0.

Thanks to the above lemma, we are now able to prove that an extension of
number eld of degree n can be embedded in exactly n dierent ways into its
algebraic closure. These n embeddings are what we need to redene the notions
of norm and trace. Let us rst recall the notion of eld monomorphism.
Denition 1.7. Let L
1
, L
2
be two eld extensions of a eld K. A eld
monomorphism from L
1
to L
2
is a eld homomorphism, that is a map from
L
1
to L
2
such that, for all a, b L
1
,
(ab) = (a)(b)
(a + b) = (a) + (b)
(1) = 1
(0) = 0.
A eld homomorphism is automatically an injective map, and thus a eld
monomorphism. It is a eld K-monomorphism if it xes K, that is, if (c) = c
for all c K.
Example 1.5. We consider the number eld K = Q(i). Let x = a +ib Q(i).
If is eld Q-homomorphism, then (x) = a + (i)b since it has to x Q.
Furthermore, we need that
(i)
2
= (i
2
) = 1,
so that (i) = i. This gives us exactly two Q-monomorphisms of K into
K C, given by:
1
: a + ib a + ib,
2
: a + ib a ib,
that is the identity and the complex conjugation.
Proposition 1.7. Let K be a number eld, L be a nite extension of K of degree
n, and

K be an algebraic closure of K. There are n distinct K-monomorphisms
of L into

K.
Proof. This proof is done in two steps. In the rst step, the claim is proved in
the case when L = K(), L. The second step is a proof by induction on
the degree of the extension L/K in the general case, which of course uses the
rst step. The main idea is that if L ,= K() for some L, then one can nd
such intermediate extension, that is, we can consider the tower of extensions
K K() L, where we can use the rst step for K()/K and the induction
hypothesis for L/K().
Step 1. Let us consider L = K(), L with minimal polynomial f(X)
K[X]. It is of degree n and thus admits n roots
1
, . . . ,
n
in

K, which are
all distinct by Lemma 1.6. For i = 1, . . . , n, we thus have a K-monomorphism
i
: L

K such that
i
() =
i
.
Step 2. We now proceed by induction on the degree n of L/K. Let L and
consider the tower of extensions K K() L, where we denote by q, q > 1,
the degree of K()/K. We know by the rst step that there are q distinct K-
monomomorphisms from K() to

K, given by
i
() =
i
, i = 1, . . . , q, where
i
are the q roots of the minimal polynomial of .
Now the elds K() and K(
i
()) are isomorphic (the isomorphism is given
by
i
) and one can build an extension L
i
of K(
i
()) and an isomorphism
i
: L L
i
which extends
i
(that is,
i
restricted to K() is nothing else than
i
):
L L
i
K() K(
i
())
K
n
q
-
i
n
q
-
i
q
q
Now, since
[L
i
: K(
i
())] = [L : K()] =
n
q
< n,
we have by induction hypothesis that there are
n
q
distinct K(
i
())-monomorphisms
ij
of L
i
into

K. Therefore,
ij

i
, i = 1, . . . , q, j = 1, . . . ,
n
q
provide n distinct
K-monomorphisms of L into

K.
Corollary 1.8. A number eld K of degree n over Q has n embeddings into C.
Proof. The proof is immediate from the proposition. It is very common to nd
in the literature expressions such as let K be a number eld of degree n, and
1
, . . . ,
n
be its n embeddings, without further explanation.
Denition 1.8. Let L/K be an extension of number elds, and let L. Let
1
, . . . ,
n
be the n eld K-monomorphisms of L into

K C given by the above
proposition. We call
1
(), . . . ,
n
() the conjugates of .
Proposition 1.9. Let L/K be an extension of number elds. Let
1
, . . . ,
n
be
the n distinct embeddings of L into C which x K. For all L, we have
N
L/K
() =
n
i=1
i
(), Tr
L/K
() =
n
i=1
i
().
Proof. Let L, with minimal polynomial f(X) K[X] of degree m, and let
K()/K
(X) be its characteristic polynomial.
Let us rst prove that f(X) =
K()/K
(X). Note that both polynomials are
monic by denition. Now the K-vector space K() has dimension m, thus m
is also the degree of
K()/K
(X). By Cayley-Hamilton theorem (which states
that every square matrix over the complex eld satises its own characteristic
equation), we have that
K()/K
(
) = 0.
Now since
K()/K
(
) =
K()/K
()
,
(see Example 1.7), we have that is a root of
K()/K
(X). By minimality
of the minimal polynomial f(X), f(X) [
K()/K
(X), but knowing that both
polynomials are monic of same degree, it follows that
f(X) =
K()/K
(X). (1.1)
We now compute the matrix of
in a K-basis of L. We have that

1, , . . . ,
m1
is a K-basis of K(). Let k be the degree [L : K()] and let

1
, . . . ,
k
be a
K()basis of L. The set
i
j
, 0 i < m, 1 j k is a K-basis of L. The
multiplication
by can now be written in this basis as
=
_
_
_
_
_
B 0 . . . 0
0 B 0
.
.
.
.
.
.
0 0 . . . B
_
_
_
_
_
. .
k times
, B =
_
_
_
_
_
_
_
0 1 . . . 0
0 0 0
.
.
.
.
.
.
0 0 1
a
0
a
1
. . . a
m1
_
_
_
_
_
_
_
where a
i
, i = 0, . . . , m1 are the coecients of the minimal polynomial f(X)
(in other words, B is the companion matrix of f). We conclude that
N
L/K
() = N
K()/K
()
k
,
Tr
L/K
() = kTr
K()/K
(),
L/K
(X) = (
K()/K
)
k
= f(X)
k
,
where last equality holds by (1.1). Now we have that
f(X) = (X
1
)(X
2
) (X
m
)

Q[X]
= X
m
i=1
i
X
m1
+ . . .
m
i=1
i
Q[X]
= X
m
Tr
K()/K
()X
m1
+ . . . N
K()/K
() Q[X]
where last equality holds by (1.1), so that
N
L/K
() =
_
m
i=1
i
_
k
,
Tr
L/K
() = k
m
i=1
i
.
To conclude, we know that the embeddings of K() into

Q which x K are
determined by the roots of , and we know that there are exactly m distinct
such roots (Lemma 1.6). We further know (see Proposition 1.7) that each of
these embeddings can be extended into an embedding of L into

Q in exactly k
ways. Thus
N
L/K
() =
n
i=1
i
(),
Tr
L/K
() =
n
i=1
i
(),
which concludes the proof.
Example 1.6. Consider the eld extension Q(
2)/Q. It has two embeddings
1
: a + b
2 a + b
2,
2
: a + b
2 a b
2.
Take the element = a + b
2 Q(
2). Its two conjugates are

1
() = =
a + b
2,
2
() = a b
2, thus its norm is given by

N
Q(
2)/Q
() =
1
()
2
() = a
2
2b
2
,
while its trace is
Tr
Q(
2)/Q
() =
1
() +
2
().
It of course gives the same answer as what we computed in Example 1.4.
Example 1.7. Consider again the eld extension Q(
2)/Q. Take the element

= a+b
2 Q(
2), whose characteristic polynomial is given by, say, (X) =

p
0
+ p
1
X + p
2
X
2
. Thus () = p
0
+ p
1
(a + b
2) + p
2
(a
2
+ 2ab
2 + 2b
2
) =
(p
0
+ p
1
a + p
2
a
2
+ p
2
2b
2
) + (p
1
b + p
2
2ab)
2, and
()
=
_
p
0
+ p
1
a + p
2
a
2
+ p
2
2b
2
2bp
1
+ 4p
2
ab
p
1
b + p
2
2ab p
0
+ p
1
a + p
2
a
2
+ p
2
2b
2
_
(see Example 1.4). On the other hand, we have that
(
) = p
0
I + p
1
_
a 2b
b a
_
+ p
2
_
a 2b
b a
_
2
.
Thus we have that
()
= (
).
Example 1.8. Consider the number eld extensions Q Q(i) Q(i,
2).
There are four embeddings of Q(i,
2), given by
1
: i i,
2
: i i,
3
: i i,
4
: i i,
2
We have that
N
Q(i)/Q
(a + ib) =
1
(a + ib)
2
(a + ib) = a
2
+ b
2
, a, b Q
but
N
Q(i,
2)/Q
(a + ib) =
1
(a + ib)
2
(a + ib)
3
(a + ib)
4
(a + ib)
=
1
(a + ib)
2
(a + ib)
1
(a + ib)
2
(a + ib)
=
1
(a + ib)
2
2
(a + ib)
2
= (a
2
+ b
2
)
2
since a, b Q.
Corollary 1.10. Let K be a number eld, and let K be an algebraic integer.
The norm and the trace of belong to Z.
Proof. The characteristic polynomial
K/Q
(X) is a power of the minimal poly-
nomial (see inside the proof of the above theorem), thus it belongs to Z[X].
Corollary 1.11. The norm N
K/Q
() of an element of O
K
is equal to 1 if
and only if is a unit of O
K
.
Proof. Let be a unit of O
K
. We want to prove that its norm is 1. Since
is a unit, we have that by denition 1/ O
K
. Thus
1 = N
K/Q
(1) = N
K/Q
()N
K/Q
(1/)
by multiplicativity of the norm. By the above corollary, both N
K/Q
() and its
inverse belong to Z, meaning that the only possible values are 1.
Conversely, let us assume that O
K
has norm 1, which means that the
constant term of its minimal polynomial f(X) is 1:
f(X) = X
n
+ a
n1
X
n1
+ 1.
Let us now consider 1/ K. We see that 1/ is a root of the monic polynomial
g(X) = 1 + a
n1
X + X
n
,
with g(X) Z[X]. Thus 1/ is an algebraic integer.
Let us prove a last result on the structure of the ring of integers. Recall that a
group G is nitely generated if there exist nitely many elements x
1
, . . . , x
s
G
such that every x G can be written in the form x = n
1
x
1
+ . . . + n
s
x
s
,
with n
1
, . . . , n
s
integers. Such a group is called free if it is isomorphic to Z
r
,
r 0, called the rank of G. We now prove that O
K
is not only a ring, but
it is furthermore a free Abelian group of rank the degree of the corresponding
number eld.
Proposition 1.12. Let K be a number eld. Then O
K
is a free Abelian group
of rank n = [K : Q].
Proof. We know by Corollary 1.5 that there exists a Q-basis
1
, . . . ,
n
of
K with
i
O
K
for i = 1, . . . , n (take a basis of K with elements in K,
and multiply the elements by the proper factors to obtain elements in O
K
as
explained in Corollary 1.5). Thus, an element x O
K
can be written as
x =
n
i=1
c
i
i
, c
i
Q.
Our goal is now to show that the denominators of c
i
are bounded for all c
i
and
all x O
K
. To prove this, let us assume by contradiction that this is not the
case, that is, that there exists a sequence
x
j
=
n
i=1
c
ij
i
, c
ij
Q
such that the greatest denominator of c
ij
, i = 1, . . . , n goes to innity when
j . Let us look at the norm of such an x
j
. We know that N
K/Q
(x
j
) is the
determinant of an n n matrix with coecients in Q[c
ij
] (that is coecients
are Q-linear combinations of c
ij
). Thus the norm is a homogeneous polynomial
in c
ij
, whose coecients are determined by the eld extension considered. Fur-
thermore, it belongs to Z (Corollary 1.10). Since the coecients are xed and
the norm is in Z, the denominators of c
ij
cannot grow indenitely. They have
to be bounded by a given constant B. Thus we have shown that
O
K

1
B
n
i=1
Z
i
.
Since the right hand side is a free Abelian group, O
K
is free. Furthermore, O
K
contains n elements which are linearly independent over Q, thus the rank of O
K
is n.
Example 1.9. Let
p
be a primitive pth root of unity. One can show that the
ring of integers of Q(
p
) is
Z[
p
] = Z Z
p
Z
p2
p
.
Proposition 1.13. Let K be a number eld. Let K. If is the zero of a
monic polynomial f with coecients in O
K
, then O
K
. We say that O
K
is
integrally closed.
Proof. Let us write f(X) = X
m
+ a
m1
X
m1
+ . . . + a
0
, with a
i
O
K
. We
know by the above proposition that O
K
is a free abelian group which is nitely
generated. Since
m
= a
m1
m1
a
0
,
we have that O
K
[] is nitely generated as Abelian group. Thus Z[] O
K
[]
is also nitely generated, and is an algebraic integer by Theorem 1.3.
The main denitions and results of this chapter are
Denition of a number eld K of degree n and its ring
of integers O
K
.
Properties of O
K
: it is a ring with a Z-basis of n
elements, and it is integrally closed.
The fact that K has n embeddings into C.
Denition of norm and trace, with their characteri-
zation as respectively product and sum of the conju-
gates.
Chapter 2
Ideals
For the whole chapter, K is a number eld of degree n and O = O
K
is its ring
of integers.
2.1 Introduction
Historically, experience with unique prime factorization of integers led mathe-
maticians in the early days of algebraic number theory to a general intuition
that factorization of algebraic integers into primes should also be unique. A
likely reason for this misconception is the actual denition of what is a prime
number. The familiar denition is that a prime number is a number which is
divisible only by 1 and itself. Since units in Z are 1, this denition can be
rephrased as: if p = ab, then one of a or b must be a unit. Equivalently over
Z, a prime number p satises that if p[ab, then p[a or p[b. However, these two
denitions are not equivalent anymore over general rings of integers. In fact,
the second property is actually stronger, and if one can get a factorization with
primes satisfying this property, then factorization will be unique, which is
not the case for primes satisfying the rst property. To distinguish these two
denitions, we say in modern terminology that a number satisfying the rst
property is irreducible, while one satisfying the second property is prime.
Consider for example Z[
6]. We have that

6 = 2 3 =
6.
We get two factorizations into irreducibles (we have a factorization but it is not
unique). However this is not a factorization into primes, since
6 divides 2 3
but
6 does not divide 2 and does not divide 3 either. So in the case where
primes and irreducibles are dierent, we now have to think what we are looking
for when we say factorization. When we attempt to factorize an element x in a
domain D, we naturally mean proper factors a, b such that x = ab, and if either
19
20 CHAPTER 2. IDEALS
of these factors can be further decomposed, we go on. That means, we look for
writing
x = a
1
a
2
. . . a
n
into factors that cannot be reduced any further. The denition of irreducible
captures what it means for the factorization to terminate: one of the term has
to be a unit.
Thus what we are interested in is to understand the factorization into irre-
ducibles inside rings of integers. Before starting, let us make a few more remarks.
Note rst that this factorization into irreducibles may not always be possible
in general rings, since the procedure may continue indenitely. However the
procedure does stop for rings of integers. This comes from the fact that rings of
integers are, again in modern terminology, what we call noetherian rings. The
big picture can nally be summarized as follows:
In general rings, factorization even into irreducibles may not be possible.
In rings of integers, factorization into irreducibles is always possible, but
may not be unique.
For rings of integers which furthermore have a generalized Euclidean di-
vision, then the notions of prime and irreducible are equivalent, thus fac-
torization is unique.
Let us now get back to the problem we are interested in, namely, factorization
into product of irreducibles in rings of integers.
Example 2.1. Let K = Q(
5) be a quadratic number eld, with ring of

integers O
K
= Z[
5], since d 3 (mod 4). Let us prove that we do not have

a unique factorization into product of irreducibles in Z[
5]. We have that

21 = 7 3 = (1 + 2
5)(1 2
5)
with 3, 7, 1 2
5 irreducible. Let us show for example that 3 is irreducible.

Let us write
3 = , , O
K
.
We need to see that either or is a unit (that is an invertible element of O
K
).
The norm of 3 is given by
9 = N
K/Q
(3) = N
K/Q
()N
K/Q
(),
by multiplicativity of the norm. By Corollary 1.10, we know that N
K/Q
(), N
K/Q
()
Z. Thus we get a factorization of 9 in Z. There are only two possible factoriza-
tions over the integers:
N
K/Q
() = 1, N
K/Q
() = 9 (or vice versa): by Corollary 1.11, we
know that the element of O
K
of norm 1 is a unit, and we are done.
2.1. INTRODUCTION 21
Figure 2.1: Ernst Kummer (1810-1893) and Richard Dedekind (1831-1916)
N
K/Q
() = 3, N
K/Q
() = 3 (or vice versa): however, we will now show
that there is no element of O
K
with norm 3. Let us indeed assume that
there exists a + b
5 O
K
, a, b Z such that
N(a + b
5) = a
2
+ 5b
2
= 3.
We can check that this equation has no solution modulo 5, yielding a
contradiction.
On the other hand, we will see that the ideal 21O
K
can be factorized into 4
prime ideals p
1
, p
2
, p
3
, p
4
such that
7O
K
= p
1
p
2
, 3O
K
= p
3
p
4
, (1 + 2
5)O
K
= p
1
p
3
, (1 2
5)O
K
= p
2
p
4
,
namely
21O
K
= p
1
p
2
p
3
p
4
.
After the realization that uniqueness of factorization into irreducibles is
unique in some rings of integers but not in others, the mathematician Kum-
mer had the idea that one way to remedy to the situation could be to work with
what he called ideal numbers, new structures which would enable us to regain
the uniqueness of factorization. Ideals numbers then got called ideals by an-
other mathematician, Dedekind, and this is the terminology that has remained.
Ideals of O will the focus of this chapter.
Our goal will be to study ideals of O, and in particular to show that we get
a unique factorization into a product of prime ideals. To prove uniqueness, we
need to study the arithmetic of non-zero ideals of O, especially their behaviour
under multiplication. We will recall how ideal multiplication is dened, which
will appear to be commutative and associative, with O itself as an identity.
However, inverses need not exist, so we do not have a group structure. It turns
out that we can get a group if we extend a bit the denition of ideals, which we
will do by introducing fractional ideals, and showing that they are invertible.
2.2 Factorization and fractional ideals
Let us start by introducing the notion of norm of an ideal. We will see that in
the case of principal ideals, we can relate the norm of the ideal with the norm
of its generator.
Denition 2.1. Let I be a non-zero ideal of O, we dene the norm of I by
N(I) = [O/I[.
Lemma 2.1. Let I be a non-zero ideal of O.
1. We have that
N(O) = [N
K/Q
()[, O.
2. The norm of I is nite.
Proof. 1. First let us notice that the formula we want to prove makes sense,
since N() Z when O
K
, thus [N()[ is a positive integer. By
Proposition 1.12, O is a free Abelian group of rank n = [K : Q], thus
there exists a Z-basis
1
, . . . ,
n
of O, that is O = Z
1
Z
n
. It is
now a general result on free Abelian groups that if H is as subgroup of
G, both of same rank, with Z-bases x
1
, . . . , x
r
and y
1
, . . . , y
r
respectively,
with y
i
=
j
a
ij
x
j
, then [G/H[ = [ det(a
ij
)[. We thus apply this theorem
in our case, where G = O and H = O. Since one basis is obtained from
the other by multiplication by , we have that
[O/O[ = [ det(
)[ = [N
K/Q
()[.
2. Let 0 ,= I. Since I is an ideal of O and O I, we have a surjective
map
O/O O/I,
so that the result follows from part 1.
17) be a quadratic number eld, with ring of

integers O
K
= Z[
17]. Then
N(18)) = 18
2
.
2.2. FACTORIZATION AND FRACTIONAL IDEALS 23
Starting from now, we need to build up several intermediate results which
we will use to prove the two most important results of this section. Let us begin
with the fact that prime is a notion stronger than maximal. You may want to
recall what is the general result for an arbitrary commutative ring and compare
with respect to the case of a ring of integers (in general, maximal is stronger
than prime).
Proposition 2.2. Every non-zero prime ideal of O is maximal.
Proof. Let p be a non-zero prime ideal of O. Since we have that
p is a maximal ideal of O O/p is a eld,
it is enough to show that O/p is a eld. In order to do so, let us consider
0 ,= x O/p, and show that x is invertible in O/p. Since p is prime, O/p is
an integral domain, thus the multiplication map
x
: O/p O/p, z xz, is
injective (that is, its kernel is 0). By the above lemma, the cardinality of O/p
is nite, thus
x
being injective, it has to be also bijective. In other words
x
is invertible, and there exists y =
1
x
(1) O/p. By denition, y is the inverse
of x.
To prove the next result, we need to rst recall how to dene the multipli-
cation of two ideals.
Denition 2.2. If I and J are ideals of O, we dene the multiplication of ideals
as follows:
IJ =
_
_
_
finite
xy, x I, y J
_
_
_
.
Example 2.3. Let I = (
1
,
2
) =
1
O +
2
O and J = (
1
,
2
) =
1
O +
2
O,
then
IJ = (
1
1
,
1
2
,
2
1
,
2
2
).
Lemma 2.3. Let I be a non-zero ideal of O. Then there exist prime ideals
p
1
, . . . , p
r
of O such that
p
1
p
2
p
r
I.
Proof. The idea of the proof goes as follows: we want to prove that every non-
zero ideal I of O contains a product of r prime ideals. To prove it, we dene
a set o of all the ideals which do not contain a product of prime ideals, and
we prove that this set is empty. To prove that this set is empty, we assume by
contradiction that o does contain at least one non-zero ideal I. From this ideal,
we will show that we can build another ideal I
1
such that I is strictly included
in I
1
, and by iteration we can build a sequence of ideals strictly included in each
other, which will give a contradiction.
Let us now proceed. Let o be the set of all ideals which do not contain a
product of prime ideals, and let I be in o. First, note that I cannot be prime
(otherwise I would contain a product of prime ideals with only one ideal, itself).
By denition of prime ideal, that means we can nd , O with I, but
, I, , I. Using these two elements and , we can build two new ideals
J
1
= O + I I, J
2
= O + I I,
with strict inclusion since , , I.
To prove that J
1
or J
2
belongs to o, we assume by contradiction that none
are. Thus by denition of o, there exist prime ideals p
1
, . . . , p
r
and q
1
, . . . , q
s
such that p
1
p
r
J
1
and q
1
. . . q
s
J
2
. Thus
p
1
, . . . , p
r
q
1
, . . . , q
s
J
1
J
2
I
where the second inclusion holds since I and
J
1
J
2
= (O + I)(O + I) = O + I + I + I
2
I
But p
1
p
r
q
1
q
s
I contradicts the fact that I o. Thus J
1
or J
2
is in o.
Starting from assuming that I is in o, we have just shown that we can nd
another ideal, say I
1
(which is either J
1
or J
2
), such that I I
1
. Since I
1
is in
o we can iterate the whole procedure, and nd another ideal I
2
which strictly
contains I
1
, and so on and so forth. We thus get a strictly increasing sequence
of ideals in o:
I I
1
I
2
. . .
Now by taking the norm of each ideal, we get a strictly decreasing sequence of
integers
N(I) > N(I
1
) > N(I
2
) > . . . ,
which yields a contradiction and concludes the proof.
Note that an ideal I of O is an O-submodule of O with scalar multiplication
given by O I I, (a, i) ai. Ideals of O are not invertible (with respect to
ideal multiplication as dened above), so in order to get a group structure, we
extend the denition and look at O-submodules of K.
Denition 2.3. A fractional ideal I is a nitely generated O-module contained
in K.
Let
1
, . . . ,
r
K be a set of generators for the fractional ideal I as O-
module. By Corollary 1.5, we can write
i
=
i
/
i
,
i
,
i
O for i = 1, . . . , r.
Set
=
r
i=1
i
.
Since O is a ring, O. By construction, J := I is an ideal of O. Thus for
any fractional ideal I O, there exists an ideal J O and O such that
I =
1
J.
This yields an equivalent denition of fractional ideal.
Denition 2.4. An O-submodule I of K is called a fractional ideal of O if
there exists some non-zero O such that I O, that is J = I is an ideal
of O and I =
1
J.
It is easier to understand the terminology fractional ideal with the second
denition.
Examples 2.4. 1. Ideals of O are particular cases of fractional ideals. They
may be called integral ideals if there is an ambiguity.
2. The set
3
2
Z =
_
3x
2
Q [ x Z
_
is a fractional ideal.
It is now time to introduce the inverse of an ideal. We rst do it in the
particular case where the ideal is prime.
Lemma 2.4. Let p be a non-zero prime ideal of O. Dene
p
1
= x K [ xp O.
1. p
1
is a fractional ideal of O.
2. O p
1
.
3. p
1
p = O.
Proof. 1. Let us start by showing that p
1
is a fractional ideal of O. Let
0 ,= a p. By denition of p
1
, we have that ap
1
O. Thus ap
1
is an
integral ideal of O, and p
1
is a fractional ideal of O.
2. We show that O p
1
. Clearly O p
1
. It is thus enough to nd
an element which is not an algebraic integer in p
1
. We start with any
0 ,= a p. By Lemma 2.3, we choose the smallest r such that
p
1
p
r
(a)O
for p
1
, . . . , p
r
prime ideals of O. Since (a)O p and p is prime, we have
p
i
p for some i by denition of prime. Without loss of generality, we can
assume that p
1
p. Hence p
1
= p since prime ideals in O are maximal
(by Proposition 2.2). Furthermore, we have that
p
2
p
r
, (a)O
by minimality of r. Hence we can nd b p
2
p
r
but not in (a)O.
We are now ready to show that we have an element in p
1
which is not in
O. This element is given by ba
1
.
Using that p = p
1
, we thus get bp (a)O, so ba
1
p O and
ba
1
p
1
.
We also have b , (a)O and so ba
1
, O.
This concludes the proof that p
1
,= O.
3. We now want to prove that p
1
p = O. It is clear from the denition
of p
1
that p = pO pp
1
= p
1
p O. Since p is maximal (again
by Proposition 2.2), pp
1
is equal to p or O. It is now enough to prove
that pp
1
= p is not possible. Let us thus suppose by contradiction that
pp
1
= p. Let
1
, . . . ,
r
be a set of generators of p as O-module, and
consider again d := ab
1
which is in p
1
but not in O (by the proof of 2.).
Then we have that
d
i
p
1
p = p and dp p
1
p = p.
Since dp p, we have that
d
i
=
r
j=1
c
ij
j
p, c
ij
O, i = 1, . . . , r,
or equivalently
0 =
r
j=1,j=i
c
ij
j
+
i
(c
ii
d), i = 1, . . . , r.
The above r equations can be rewritten in a matrix equation as follows:
_
_
_
_
_
c
11
d c
1r
c
21
c
2r
.
.
.
.
.
.
c
r1
c
rr
d
_
_
_
_
_
. .
C
_
_
_
_
_
2
.
.
.
r
_
_
_
_
_
= 0.
Thus the determinant of C is zero, while det(C) is an equation of degree
r in d with coecients in O of leading term 1. By Proposition 1.13, we
have that d must be in O, which is a contradiction.
Example 2.5. Consider the ideal p = 3Z of Z. We have that
p
1
= x Q [ x 3Z Z = x Q [ 3x Z =
1
3
Z.
We have
p Z p
1
Q.
We can now prove that the fractional ideals of K form a group.
Theorem 2.5. The non-zero fractional ideals of a number eld K form a mul-
tiplicative group, denoted by I
K
.
Proof. The neutral element is O. It is enough to show that every non-zero
fractional ideal of O is invertible in I
K
. By Lemma 2.4, we already know this
is true for every prime (integral) ideals of O.
Let us now show that this is true for every integral ideal of O. Suppose
by contradiction that there exists a non-invertible ideal I with its norm N(I)
minimal. Now I is included in a maximal integral p, which is also a prime ideal.
Thus
I p
1
I p
1
p = O,
where last equality holds by the above lemma. Let us show that I ,= p
1
I, so
that the rst inclusion is actually a strict inclusion. Suppose by contradiction
that I = p
1
I. Let d p
1
but not in O and let
1
, . . . ,
r
be the set of
generators of I as O-module. We can thus write:
d
i
p
1
I = I, dI p
1
I = I.
By the same argument as in the previous lemma, we get that d is in O, a
contradiction. We have thus found an ideal with I p
1
I, which implies that
N(I) > N(p
1
I). By minimality of N(I), the ideal p
1
I is invertible. Let
J I
K
be its inverse, that is Jp
1
I = O. This shows that I is invertible, with
inverse Jp
1
.
If I is a fractional ideal, it can be written as
1
d
J with J an integral ideal of
O and d O. Thus dJ
1
is the inverse of I.
We can now prove unique factorization of integral ideals in O = O
K
.
Theorem 2.6. Every non-zero integral ideal I of O can be written in a unique
way (up to permutation of the factors) as a product of prime ideals.
Proof. We thus have to prove the existence of the factorization, and then the
unicity up to permutation of the factors.
Existence. Let I be an integral ideal which does not admit such a factorization.
We can assume that it is maximal among those ideals. Then I is not prime, but
we will have I p for some maximal (hence prime) ideal. Thus Ip
1
O is an
integral ideal and I Ip
1
O, which strict inclusion, since if I = Ip
1
, then
it would imply that O = p
1
. By maximality of I, the ideal Ip
1
must admit
a factorization
Ip
1
= p
2
p
r
that is
I = pp
2
p
r
,
a contradiction.
Unicity. Let us assume that there exist two distinct factorizations of I, that is
I = p
1
p
2
p
r
= q
1
q
s
where p
i
, q
j
are prime ideals, i = 1, . . . , r, j = 1, . . . , s. Let us assume by
contradiction that p
1
is dierent from q
j
for all j. Thus we can choose
j
q
j
but not in p
1
, and we have that
q
j
= I p
1
,
which contradicts the fact that p
1
is prime. Thus p
1
must be one of the q
j
, say
q
1
. This gives that
p
2
p
r
= q
2
q
s
.
We conclude by induction.
Example 2.6. In Q(i), we have
(2)Z[i] = (1+)
2
Z[i].
Corollary 2.7. Let I be a non-zero fractional ideal. Then
I = p
1
p
r
q
1
1
q
1
s
,
where p
1
, . . . , p
r
, q
1
, . . . , q
s
are prime integral ideals. This factorization is unique
up to permutation of the factors.
Proof. A fractional ideal can be written as d
1
I, d O, I an integral ideal. We
write I = p
1
p
t
and dO = q
1
q
n
. Thus
(dO)
1
I = p
1
p
t
q
1
q
q
1
n
.
It may be that some of the terms will cancel out, so that we end up with a
factorization with p
1
, . . . , p
r
and q
1
, . . . , q
s
. Unicity is proved as in the above
theorem.
2.3 The Chinese Theorem
We have dened in the previous section the group I
K
of fractional ideals of a
number eld K, and we have proved that they have a unique factorization into
a product of prime ideals. We are now interested in studying further properties
of fractional ideals. The two main properties that we will prove are the fact
that two elements are enough to generate these ideals, and that norms of ideals
are multiplicative. Both properties can be proved as corollary of the Chinese
theorem, that we rst recall.
Theorem 2.8. Let I =
m
i=1
p
k
i
i
be the factorization of an integral ideal I into
a product of prime ideals p
i
with p
i
,= p
j
of i ,= j. Then there exists a canonical
isomorphism
O/I
m
i=1
O/p
k
i
i
.
Corollary 2.9. Let I
1
, . . . , I
m
be ideals which are pairwise coprime (that is
I
i
+ I
j
= O for i ,= j). Let
1
, . . . ,
m
be elements of O. Then there exists
O with

i
mod I
i
, i = 1, . . . , m.
2.3. THE CHINESE THEOREM 29
Proof. We can write I
i
=
p
n
ij
ij
. By hypothesis, no prime ideal occurs more
than once as a p
ij
, and each congruence is equivalent to the nite set of con-
gruences

i
mod p
n
ij
ij
, i, j.
Now write I =
I
i
. Consider the vector (
1
, . . . ,
m
) in
m
i=1
O/I
i
. The map
O/I = O/
I
i

m
i=1
O/I
i
is surjective, thus there exists a preimage O of (
1
, . . . ,
m
).
Corollary 2.10. Let I be a fractional ideal of O, I. Then there exists
I such that
(, ) =< , >= O + O = I.
Proof. Let us rst assume that I is an integral ideal. Let p
1
, . . . , p
m
be the
prime factors of O I, so that I can be written as
I =
m
i=1
p
k
i
i
, k
i
0.
Let us choose
i
in p
k
i
i
but not in p
k
i
+1
i
. By Corollary 2.9, there exists O
such that
i
mod p
k
i
+1
i
for all i = 1, . . . , m. Thus
I =
m
i=1
p
k
i
i
and p
k
i
i
is the exact power of p
i
which divides O. In other words, OI
1
is
prime to O, thus OI
1
+ O = O, that is
O + I = I.
To conclude, we have that
I = O + I O + O I.
If I is a fractional ideal, there exists by denition d O such that I =
1
d
J
with J an integral ideal. Thus d J. By the rst part, there exists J
such that J = (d, ). Thus
I = O +

d
O.
We are now left to prove properties of the norm of an ideal, for which we
need the following result.
Proposition 2.11. Let p be a prime ideal of O and n > 0. Then the O-modules
O/p and p
n
/p
n+1
are (non-canonically) isomorphic.
Proof. We consider the map
: O p
n
/p
n+1
,
for any in p
n
but not in p
n+1
. The proof consists of computing the kernel and
the image of , and then of using one of the ring isomorphism theorems. This
will conclude the proof since we will prove that ker() = p and is surjective.
Let us rst compute the kernel of . If () = 0, then = 0, which
means that p
n+1
that is p.
Given any p
n
, by Corollary 2.9, we can nd
1
O such that
1
mod p
n+1
,
1
0 mod Op
n
,
since Op
n
is an ideal coprime to p
n+1
. Since p
n
, we have that
1
belongs to Op
n
p
n
= O since I J = IJ when I and J are coprime
ideals. In other words,
1
/ O. Its image by is
(
1
/) = (
1
/) mod p
n+1
= .
Thus is surjective.
Corollary 2.12. Let I and J be two integral ideals. Then
N(IJ) = N(I)N(J).
Proof. Let us rst assume that I and J are coprime. The chinese theorem tells
us that
O/IJ O/I O/J,
thus [O/IJ[ = [O/I[[O/J[. We are left to prove that N(p
k
) = N(p)
k
for k 1,
p a prime ideal. Now one of the isomorphism theorems for rings allows us to
write that
N(p
k1
) = [O/p
k1
[ =
O/p
k
p
k1
/p
k
=
[O/p
k
[
[p
k1
/p
k
[
.
By the above proposition, this can be rewritten as
[O/p
k
[
[O/p[
=
N(p
k
)
N(p)
.
Thus N(p
k
) = N(p
k1
)N(p), and by induction on k, we conclude the proof.
2.3. THE CHINESE THEOREM 31
Example 2.7. In Q(i), we have that (2)Z[i] = (1 + i)
2
Z[i] and thus
4 = N(2) = N(1 + i)
2
and
N(1 + i) = 2.
Denition 2.5. If I = J
1
J
1
2
is a non-zero fractional ideal with J
1
, J
2
integral
ideals, we set
N(I) =
N(J
1
)
N(J
2
)
.
This extends the norm N into a group homomorphism N : I
K
Q
. For
example, we have that N
_
2
3
Z
_
=
2
3
.
Denition of fractional ideals, the fact that they form
a group I
K
.
Denition of norm of both integral and fractional ide-
als, and that the norm of ideals is multiplicative.
The fact that ideals can be uniquely factorized into
products of prime ideals.
The fact that ideals can be generated with two ele-
ments: I = (, ).
Chapter 3
Ramication Theory
This chapter introduces ramication theory, which roughly speaking asks the
following question: if one takes a prime (ideal) p in the ring of integers O
K
of
a number eld K, what happens when p is lifted to O
L
, that is pO
L
, where L
is an extension of K. We know by the work done in the previous chapter that
pO
L
has a factorization as a product of primes, so the question is: will pO
L
still
be a prime? or will it factor somehow?
In order to study the behavior of primes in L/K, we rst consider absolute
extensions, that is when K = Q, and dene the notions of discriminant, inertial
degree and ramication index. We show how the discriminant tells us about
ramication. When we are lucky enough to get a nice ring of integers O
L
,
that is O
L
= Z[] for L, we give a method to compute the factorization of
primes in O
L
. We then generalize the concepts introduced to relative extensions,
and study the particular case of Galois extensions.
3.1 Discriminant
Let K be a number eld of degree n. Recall from Corollary 1.8 that there are
n embeddings of K into C.
Denition 3.1. Let K be a number eld of degree n, and set
r
1
= number of real embeddings
r
2
= number of pairs of complex embeddings
The couple (r
1
, r
2
) is called the signature of K. We have that
n = r
1
+ 2r
2
.
Examples 3.1. 1. The signature of Q is (1, 0).
2. The signature of Q(
d), d > 0, is (2, 0).

33
34 CHAPTER 3. RAMIFICATION THEORY
d), d < 0, is (0, 1).

3
2) is (1, 1).
Let K be a number eld of degree n, and let O
K
be its ring of integers. Let
1
, . . . ,
n
be its n embeddings into C. We dene the map
: K C
n
x (
1
(x), . . . ,
n
(x)).
Since O
K
is a free abelian group of rank n, we have a Z-basis
1
, . . . ,
n
of
O
K
. Let us consider the n n matrix M given by
M = (
i
(
j
))
1i,jn
.
The determinant of M is a measure of the density of O
K
in K (actually of
K/O
K
). It tells us how sparse the integers of K are. However, det(M) is only
dened up to sign, and is not necessarily in either R or K. So instead we
consider
det(M
2
) = det(M
t
M)
= det
_
n
k=1
k
(
i
)
k
(
j
)
_
i,j
= det(Tr
K/Q
(
i
j
))
i,j
Z,
and this does not depend on the choice of a basis.
Denition 3.2. Let
1
, . . . ,
n
K. We dene
disc(
1
, . . . ,
n
) = det(Tr
K/Q
(
i
j
))
i,j
.
In particular, if
1
, . . . ,
n
is any Z-basis of O
K
, we write
K
, and we call
discriminant the integer
K
= det(Tr
K/Q
(
i
j
))
1i,jn
.
We have that
K
,= 0. This is a consequence of the following lemma.
Lemma 3.1. The symmetric bilinear form
K K Q
(x, y) Tr
K/Q
(xy)
is non-degenerate.
Proof. Let us assume by contradiction that there exists 0 ,= K such that
Tr
K/Q
() = 0 for all K. By taking =
1
, we get
Tr
K/Q
() = Tr
K/Q
(1) = n ,= 0.
3.2. PRIME DECOMPOSITION 35
Now if we had that
K
= 0, there would be a non-zero column vector
(x
1
, . . . , x
n
)
t
, x
i
Q, killed by the matrix (Tr
K/Q
(
i
j
))
1i,jn
. Set =
n
i=1
i
x
i
, then Tr
K/Q
(
j
) = 0 for each j, which is a contradiction by the
above lemma.
Example 3.2. Consider the quadratic eld K = Q(
5). Its two embeddings

into C are given by
1
: a + b
5 a + b
5,
2
: a + b
5 a b
5.
Its ring of integers is Z[(1 +
5)/2], so that the matrix M of embeddings is

M =
_

1
(1)
2
(1)
1
_
1+
5
2
_

2
_
1+
5
2
_
_
and its discriminant
K
can be computed by
K
= det(M
2
) = 5.
3.2 Prime decomposition
Let p be a prime ideal of O. Then pZ is a prime ideal of Z. Indeed, one easily
veries that this is an ideal of Z. Now if a, b are integers with ab p Z, then
we can use the fact that p is prime to deduce that either a or b belongs to p and
thus to p Z (note that p Z is a proper ideal since p Z does not contain 1,
and p Z ,= , as N(p) belongs to p and Z since N(p) = [O/p[ < ).
Since p Z is a prime ideal of Z, there must exist a prime number p such
that p Z = pZ. We say that p is above p.
p O
K
K
pZ Z Q
We call residue eld the quotient of a commutative ring by a maximal ideal.
Thus the residue eld of pZ is Z/pZ = F
p
. We are now interested in the residue
eld O
K
/p. We show that O
K
/p is a F
p
-vector space of nite dimension. Set
: Z O
K
O
K
/p,
where the rst arrow is the canonical inclusion of Z into O
K
, and the second
arrow is the projection , so that = . Now the kernel of is given by
ker() = a Z [ a p = p Z = pZ,
so that induces an injection of Z/pZ into O
K
/p, since Z/pZ Im() O
K
/p.
By Lemma 2.1, O
K
/p is a nite set, thus a nite eld which contains Z/pZ and
we have indeed a nite extension of F
p
.
Denition 3.3. We call inertial degree, and we denote by f
p
, the dimension of
the F
p
-vector space O/p, that is
f
p
= dim
F
p
(O/p).
Note that we have
N(p) = [O/p[ = [F
dim
F
p
(O/p)
p
[ = [F
p
[
f
p
= p
f
p
.
Example 3.3. Consider the quadratic eld K = Q(i), with ring of integers
Z[i], and let us look at the ideal 2Z[i]:
2Z[i] = (1 + i)(1 i)Z[i] = p
2
, p = (1 + i)Z[i]
since (i)(1 + i) = 1 i. Furthermore, p Z = 2Z, so that p = (1 + i) is said
to be above 2. We have that
N(p) = N
K/Q
(1 + i) = (1 + i)(1 i) = 2
and thus f
p
= 1. Indeed, the corresponding residue eld is
O
K
/p F
2
.
Let us consider again a prime ideal p of O. We have seen that p is above
the ideal pZ = p Z. We can now look the other way round: we start with the
prime p Z, and look at the ideal pO of O. We know that pO has a unique
factorization into a product of prime ideals (by all the work done in Chapter
2). Furthermore, we have that p p, thus p has to be one of the factors of pO.
Denition 3.4. Let p Z be a prime. Let p be a prime ideal of O above p.
We call ramication index of p, and we write e
p
, the exact power of p which
divides pO.
We start from p Z, whose factorization in O is given by
pO = p
e
p
1
1
p
e
p
g
g
.
We say that p is ramied if e
p
i
> 1 for some i. On the contrary, p is non-ramied
if
pO = p
1
p
g
, p
i
,= p
j
, i ,= j.
Both the inertial degree and the ramication index are connected via the degree
of the number eld as follows.
Proposition 3.2. Let K be a number eld and O
K
its ring of integers. Let
p Z and let
pO = p
e
p
1
1
p
e
p
g
g
be its factorization in O. We have that
n = [K : Q] =
g
i=1
e
p
i
f
p
i
.
Proof. By Lemma 2.1, we have
N(pO) = [N
K/Q
(p)[ = p
n
,
where n = [K : Q]. Since the norm N is multiplicative (see Corollary 2.12), we
deduce that
N(p
e
p
1
1
p
e
p
g
g
) =
g
i=1
N(p
i
)
e
p
i
=
g
i=1
p
f
p
i
e
p
i
.
There is, in general, no straightforward method to compute the factorization
of pO. However, in the case where the ring of integers O is of the form O = Z[],
we can use the following result.
Proposition 3.3. Let K be a number eld, with ring of integers O
K
, and let
p be a prime. Let us assume that there exists such that O = Z[], and let f
be the minimal polynomial of , whose reduction modulo p is denoted by

f. Let
f(X) =
g
i=1
i
(X)
e
i
be the factorization of f(X) in F
p
[X], with
i
(X) coprime and irreducible. We
set
p
i
= (p, f
i
()) = pO + f
i
()O
where f
i
is any lift of
i
to Z[X], that is

f
i
=
i
mod p. Then
pO = p
e
1
1
p
e
g
g
is the factorization of pO in O.
Proof. Let us rst notice that we have the following isomorphism
O/pO = Z[]/pZ[]
Z[X]/f(X)
p(Z[X]/f(X))
Z[X]/(p, f(X)) F
p
[X]/
f(X),
where

f denotes f mod p. Let us call A the ring
A = F
p
[X]/
f(X).
The inverse of the above isomorphism is given by the evaluation in , namely, if
(X) F
p
[X], with (X) mod

f(X) A, and g Z[X] such that g = , then
its preimage is given by g(). By the Chinese Theorem, recall that we have
A = F
p
[X]/
f(X)
g
i=1
F
p
[X]/
i
(X)
e
i
,
since by assumption, the ideal (

f(X)) has a prime factorization given by (

f(X)) =
g
i=1
(
i
(X))
e
i
.
We are now ready to understand the structure of prime ideals of both O/pO
and A, thanks to which we will prove that p
i
as dened in the assumption is
prime, that any prime divisor of pO is actually one of the p
i
, and that the power
e
i
appearing in the factorization of

f are bigger or equal to the ramication index
e
p
i
of p
i
. We will then invoke the proposition that we have just proved to show
that e
i
= e
p
i
, which will conclude the proof.
By the factorization of A given above by the Chinese theorem, the maximal
ideals of A are given by (
i
(X))A, and the degree of the extension A/(
i
(X))A
over F
p
is the degree of
i
. By the isomorphism A O/pO, we get similarly
that the maximal ideals of O/pO are the ideals generated by f
i
() mod pO.
We consider the projection : O O/pO. We have that
(p
i
) = (pO + f
i
()O) = f
i
()O mod pO.
Consequently, p
i
is a prime ideal of O, since f
i
()O is. Furthermore, since
p
i
pO, we have p
i
[ pO, and the inertial degree f
p
i
= [O/p
i
: F
p
] is the degree
of
i
, while e
p
i
denotes the ramication index of p
i
.
Now, every prime ideal p in the factorization of pO is one of the p
i
, since
the image of p by is a maximal ideal of O/pO, that is
pO = p
e
p
1
1
p
e
p
g
g
and we are thus left to look at the ramication index.
The ideal
e
i
i
A of A belongs to O/pO via the isomorphism between O/pO
A, and its preimage in O by
1
contains p
e
i
i
(since if p
e
i
i
, then is a sum of
products
1

e
i
, whose image by will be a sum of product (
1
) (
e
i
)
with (
i
)
i
A). In O/pO, we have 0 =
g
i=1
i
()
e
i
, that is
pO =
1
(0) =
g
i=1
1
(
e
i
i
A)
g
i=1
p
e
i
i
=
g
i=1
p
e
i
i
.
We then have that this last product is divided by pO =
p
e
p
i
i
, that is e
i
e
p
i
.
Let n = [K : Q]. To show that we have equality, that is e
i
= e
p
i
, we use the
previous proposition:
n = [K : Q] =
g
i=1
e
p
i
f
p
i

g
i=1
e
i
deg(
i
) = dim
F
p
(A) = dim
F
p
Z
n
/pZ
n
= n.
The above proposition gives a concrete method to compute the factorization
of a prime pO
K
:
1. Choose a prime p Z whose factorization in pO
K
is to be computed.
2. Let f be the minimal polynomial of such that O
K
= Z[].
3. Compute the factorization of

f = f mod p:
f =
g
i=1
i
(X)
e
i
.
4. Lift each
i
in a polynomial f
i
Z[X].
5. Compute p
i
= (p, f
i
()) by evaluating f
i
in .
6. The factorization of pO is given by
pO = p
e
1
1
p
e
g
g
.
Examples 3.4. 1. Let us consider K = Q(
3
2), with ring of integers O

K
=
Z[
3
2]. We want to factorize 5O

K
. By the above proposition, we compute
X
3
2 (X 3)(X
2
+ 3X + 4)
(X + 2)(X
2
2X 1) mod 5.
We thus get that
5O
K
= p
1
p
2
, p
1
= (5, 2 +
3
2), p
2
= (5,
3
4 2
3
2 1).
2. Let us consider Q(i), with O
K
= Z[i], and choose p = 2. We have = i
and f(X) = X
2
+ 1. We compute the factorization of

f(X) = f(X)
mod 2:
X
2
+ 1 X
2
1 (X 1)(X + 1) (X 1)
2
mod 2.
We can take any lift of the factors to Z[X], so we can write
2O
K
= (2, i 1)(2, i + 1) or 2 = (2, i 1)
2
which is the same, since (2, i 1) = (2, 1 + i). Furthermore, since 2 =
(1 i)(1 +i), we see that (2, i 1) = (1 +i), and we recover the result of
Example 3.3.
Denition 3.5. We say that p is inert if pO is prime, in which case we have
g = 1, e = 1 and f = n. We say that p is totally ramied if e = n, g = 1, and
f = 1.
The discriminant of K gives us information on the ramication in K.
Theorem 3.4. Let K be a number eld. If p is ramied, then p divides the
discriminant
K
.
Proof. Let p [ pO be an ideal such that p
2
[ pO (we are just rephrasing the fact
that p is ramied). We can write pO = pI with I divisible by all the primes
above p (p is voluntarily left as a factor of I). Let
1
, . . . ,
n
O be a Z-basis
of O and let I but , pO. We write
= b
1
1
+ . . . + b
n
n
, b
i
Z.
Since , pO, there exists a b
i
which is not divisible by p, say b
1
. Recall that
K
= det
_
_
_
1
(
1
) . . .
1
(
n
)
.
.
.
.
.
.
n
(
1
) . . .
n
(
n
)
_
_
_
2
where
i
, i = 1, . . . , n are the n embeddings of K into C. Let us replace
1
by
, and set
D = det
_
_
_
1
() . . .
1
(
n
)
.
.
.
.
.
.
n
() . . .
n
(
n
)
_
_
_
2
.
Now D and
K
are related by
D =
K
b
2
1
,
since D can be rewritten as
D = det
_
_
_
_
_
_
_
_
1
(
1
) . . .
1
(
n
)
.
.
.
.
.
.
n
(
1
) . . .
n
(
n
)
_
_
_
_
_
_
_
_
b
1
0 . . . 0
b
2
1 0
.
.
.
b
n
. . . 1
_
_
_
_
_
_
_
_
_
_
2
.
We are thus left to prove that p [ D, since by construction, we have that p does
not divide b
2
1
.
Intuitively, the trick of this proof is to replace proving that p[
K
where we
have no clue how the factor p appears, with proving that p[D, where D has been
built on purpose as a function of a suitable which we will prove below is such
that all its conjugates are above p.
Let L be the Galois closure of K, that is, L is a eld which contains K, and
which is a normal extension of Q. The conjugates of all belong to L. We
know that belongs to all the primes of O
K
above p. Similarly, K L
belongs to all primes P of O
L
above p. Indeed, P O
K
is a prime ideal of O
K
above p, which contains .
We now x a prime P above p in O
L
. Then
i
(P) is also a prime ideal of
O
L
above p (
i
(P) is in L since L/Q is Galois,
i
(P) is prime since P is, and
p =
i
(p)
i
(P)). We have that
i
() P for all
i
, thus the rst column of
the matrix involves in the computation of D is in P, so that D P and D Z,
to get
D P Z = pZ.
3.3. RELATIVE EXTENSIONS 41
We have just proved that if p is ramied, then p[
K
. The converse is also
true.
Examples 3.5. 1. We have seen in Example 3.2 that the discriminant of
K = Q(
5) is
K
= 5. This tells us that only 5 is ramied in Q(
5).
2. In Example 3.3, we have seen that 2 ramies in K = Q(i). So 2 should
appear in
K
. One can actually check that
K
= 4.
Corollary 3.5. There is only a nite number of ramied primes.
Proof. The discriminant only has a nite number of divisors.
3.3 Relative Extensions
Most of the theory seen so far assumed that the base eld is Q. In most cases,
this can be generalized to an arbitrary number eld K, in which case we consider
a number eld extension L/K. This is called a relative extension. By contrast,
we may call absolute an extension whose base eld is Q. Below, we will gen-
eralize several denitions previously given for absolute extensions to relative
extensions.
Let K be a number eld, and let L/K be a nite extension. We have
correspondingly a ring extension O
K
O
L
. If P is a prime ideal of O
L
, then
p = P O
K
is a prime ideal of O
K
. We say that P is above p. We have a
factorization
pO
L
=
g
i=1
P
e
P
i
|p
i
,
where e
P
i
/p
is the relative ramication index. The relative inertial degree is
given by
f
P
i
|p
= [O
L
/P
i
: O
K
/p].
We still have that
[L : K] =
e
P|p
f
P|p
where the summation is over all P above p.
Let M/L/K be a tower of nite extensions, and let T, P, p be prime ideals
of respectively M, L, and K. Then we have that
f
P|p
= f
P|P
f
P|p
e
P|p
= e
P|P
e
P|p
.
Let I
K
, I
L
be the groups of fractional ideals of K and L respectively. We
can also generalize the application norm as follows:
N : I
L
I
K
P p
f
P|p
,
which is a group homomorphism. This denes a relative norm for ideals, which
is itself an ideal!
In order to generalize the discriminant, we would like to have an O
K
-basis
of O
L
(similarly to having a Z-basis of O
K
), however such a basis does not exist
in general. Let
1
, . . . ,
n
be a K-basis of L where
i
O
L
, i = 1, . . . , n. We
set
disc
L/K
(
1
, . . . ,
n
) = det
_
_
_
1
(
1
) . . .
n
(
1
)
.
.
.
.
.
.
1
(
n
) . . .
n
(
n
)
_
_
_
2
where
i
: L C are the embeddings of L into C which x K. We dene
L/K
as the ideal generated by all disc
L/K
(
1
, . . . ,
n
). It is called relative
discriminant.
3.4 Normal Extensions
Let L/K be a Galois extension of number elds, with Galois group G = Gal(L/K).
Let p be a prime of O
K
. If P is a prime above p in O
L
, and G, then (P)
is a prime ideal above p. Indeed, (P) O
K
K, thus (P) O
K
= P O
K
since K is xed by .
Theorem 3.6. Let
pO
L
=
g
i=1
P
e
i
i
be the factorization of pO
L
in O
L
. Then G acts transitively on the set P
1
, . . . , P
g
.
Furthermore, we have that
e
1
= . . . = e
g
= e where e
i
= e
P
i
|p
f
1
= . . . = f
g
= f where f
i
= f
P
i
|p
and
[L : K] = efg.
Proof. G acts transitively. Let P be one of the P
i
. We need to prove that
there exists G such that (P
j
) = P for P
j
any other of the P
i
. In the proof
of Corollary 2.10, we have seen that there exists P such that O
L
P
1
is
an integral ideal coprime to pO
L
. The ideal
I =
G
(O
L
P
1
)
is an integral ideal of O
L
(since O
L
P
1
is), which is furthermore coprime to
pO
L
(since (O
L
P
1
) and (pO
L
) are coprime and (pO
L
) = (p)(O
L
) =
pO
L
).
3.4. NORMAL EXTENSIONS 43
Thus I can be rewritten as
I =
G
()O
L
G
(P)
=
N
L/K
()O
L
G
(P)
and we have that
I
G
(P) = N
L/K
()O
L
.
Since N
L/K
() =
G
(), P and one of the is the identity, we have
that N
L/K
() P. Furthermore, N
L/K
() O
K
since O
L
, and we get
that N
L/K
() P O
K
= p, from which we deduce that p divides the right
hand side of the above equation, and thus the left hand side. Since I is coprime
to p, we get that p divides
G
(P). In other words, using the factorization
of p, we have that
G
(P) is divisible by pO
L
=
g
i=1
P
e
i
i
and each of the P
i
has to be among (P)
G
.
All the ramication indices are equal. By the rst part, we know that
there exists G such that (P
i
) = P
k
, i ,= k. Now, we have that
(pO
L
) =
g
i=1
(P
i
)
e
i
= pO
L
=
g
i=1
P
e
i
i
where the second equality holds since p O
K
and L/K is Galois. By comparing
the two factorizations of p and its conjugates, we get that e
i
= e
k
.
All the inertial degrees are equal. This follows from the fact that
induces the following eld isomorphism
O
L
/P
i
O
L
/(P
i
).
Finally we have that
[G[ = [L : K] = efg.
For now on, let us x P above p.
Denition 3.6. The stabilizer of P in G is called the decomposition group,
given by
D = D
P/p
= G [ (P) = P < G.
The index [G : D] must be equal to the number of elements in the orbit GP
of P under the action of G, that is [G : D] = [GP[ (this is the orbit-stabilizer
theorem).
By the above theorem, we thus have that [G : D] = g, where g is the number
of distinct primes which divide pO
L
. Thus
n = efg
= ef
[G[
[D[
and
[D[ = ef.
If P
is another prime ideal above p, then the decomposition groups D

P/p
and D
P
/p
are conjugate in G via any Galois automorphism mapping P to P
(in formula, we have that if P
= (P), then D
P/p
1
= D
(P)/p
).
Proposition 3.7. Let D = D
P/p
be the decomposition group of P. The subeld
L
D
= L [ () = , D
is the smallest subeld M of L such that (PO
M
)O
L
does not split. It is called
the decomposition eld of P.
Proof. We rst prove that L/L
D
has the property that(P O
L
D)O
L
does not
split. We then prove its minimality.
We know by Galois theory that Gal(L/L
D
) is given by D. Furthermore, the
extension L/L
D
is Galois since L/K is. Let Q = PO
L
D be a prime below P.
By Theorem 3.6, we know that D acts transitively on the set of primes above
Q, among which is P. Now by denition of D = D
P/p
, we know that P is xed
by D. Thus there is only P above Q.
Let us now prove the minimality of L
D
. Assume that there exists a eld
M with L/M/K, such that Q = P O
M
has only one prime ideal of O
L
above it. Then this unique ideal must be P, since by denition P is above
Q. Then Gal(L/M) is a subgroup of D, since its elements are xing P. Thus
M L
D
.
L P
L
D
Q
K p
n
g D
g G/D
terminology e f g
inert 1 n 1
totally ramied n 1 1
(totally) split 1 1 n
Table 3.1: Dierent prime behaviors
The next proposition uses the same notation as the above proof.
Proposition 3.8. Let Q be the prime of L
D
below P. We have that
f
Q/p
= e
Q/p
= 1.
If D is a normal subgroup of G, then p is completely split in L
D
.
Proof. We know that [G : D] = g(P/p) which is equal to [L
D
: K] by Galois
theory. The previous proposition shows that g(P/Q) = 1 (recall that g counts
how many primes are above). Now we compute that
e(P/Q)f(P/Q) =
[L : L
D
]
g(P/Q)
= [L : L
D
]
=
[L : K]
[L
D
: K]
.
Since we have that
[L : K] = e(P/p)f(P/p)g(P/p)
and [L
D
: K] = g(P/p), we further get
e(P/Q)f(P/Q) =
e(P/p)f(P/p)g(P/p)
g(P/p)
= e(P/p)f(P/p)
= e(P/Q)f(P/Q)e(Q/p)f(Q/p)
where the last equality comes from transitivity. Thus
e(Q/p)f(Q/p) = 1
and e(Q/p) = f(Q/p) = 1 since they are positive integers.
If D is normal, we have that L
D
/K is Galois. Thus
[L
D
: K] = e(Q/p)f(Q/p)g(Q/p) = g(Q/p)
and p completely splits.
Let be in D. Then induces an automorphism of O
L
/P which xes
O
K
/p = F
p
. That is we get an element () Gal(F
P
/F
p
). We have thus
constructed a map
: D Gal(F
P
/F
p
).
This is a group homomorphism. We know that Gal(F
P
/F
p
) is cyclic, generated
by the Frobenius automorphism dened by
Frob
P
(x) = x
q
, q = [F
p
[.
Denition 3.7. The inertia group I = I
P/p
is dened as being the kernel of .
Example 3.6. Let K = Q(i) and O
K
= Z[i]. We have that K/Q is a Galois
extension, with Galois group G = 1, where : a + ib a ib.
We have that
(2) = (1 + i)
2
Z[i],
thus the ramication index is e = 2. Since efg = n = 2, we have that
f = g = 1. The residue eld is Z[i]/(1 + i)Z[i] = F
2
. The decomposition
group D is G since ((1 +i)Z[i]) = (1 +i)Z[i]. Since f = 1, Gal(F
2
/F
2
) =
1 and () = 1. Thus the kernel of is D = G and the inertia group is
I = G.
We have that
(13) = (2 + 3i)(2 3i),
thus the ramication index is e = 1. Here D = 1 for (2 3i) since
((2+3i)Z[i]) = (23i)Z[i] ,= (2+3i)Z[i]. We further have that g = 2, thus
efg = 2 implies that f = 1, which as for 2 implies that the inertia group is
I = G. We have that the residue eld for (23i) is Z[i]/(23i)Z[i] = F
13
.
We have that (7)Z[i] is inert. Thus D = G (the ideal belongs to the base
eld, which is xed by the whole Galois group). Since e = g = 1, the
inertial degree is f = 2, and the residue eld is Z[i]/(7)Z[i] = F
49
. The
Galois group Gal(F
49
/F
7
) = 1, with : x x
7
, x F
49
. Thus the
inertia group is I = 1.
We can prove that is surjective and thus get the following exact sequence:
1 I D Gal(F
P
/F
p
) 1.
The decomposition group is so named because it can be used to decompose
the eld extension L/K into a series of intermediate extensions each of which
has a simple factorization behavior at p. If we denote by L
I
the xed eld of I,
then the above exact sequence corresponds under Galois theory to the following
tower of elds:
L P
L
I
L
D
K p
e
f
g
Intuitively, this decomposition of the extension says that L
D
/K contains all
of the factorization of p into distinct primes, while the extension L
I
/L
D
is the
source of all the inertial degree in P over p. Finally, the extension L/L
I
is
responsible for all of the ramication that occurs over p.
Note that the map plays a special role for further theories, including
reciprocity laws and class eld theory.
Denition of discriminant, and that a prime ramies if and
only if it divides the discriminant.
Denition of signature.
The terminology relative to ramication: prime
above/below, inertial degree, ramication index, residue
eld, ramied, inert, totally ramied, split.
The method to compute the factorization if O
K
= Z[].
The formula [L : K] =
g
i=1
e
i
f
i
.
The notion of absolute and relative extensions.
If L/K is Galois, that the Galois group acts transitively on
the primes above a given p, that [L : K] = efg, and the
concepts of decomposition group and inertia group.
Chapter 4
Ideal Class Group and Units
We are now interested in understanding two aspects of ring of integers of number
elds: how principal they are (that is, what is the proportion of principal
ideals among all the ideals), and what is the structure of their group of units.
For the former task, we will introduce the notion of class number (as the measure
of how principal a ring of integers is), and prove that the class number is nite.
We will then prove Dirichlets Theorem for the structure of groups of units.
Both results will be derived in the spirit of geometry of numbers, that is as a
consequence of Minkowskis theorem, where algebraic results are proved thanks
to a suitable geometrical interpretation (mainly the fact that a ring of integers
can be seen as a lattice in R
n
via the n embeddings of its number eld).
4.1 Ideal class group
Let K be a number eld, and O
K
be its ring of integers. We have seen in
Chapter 2 that we can extend the notion of ideal to fractional ideal, and that
with this new notion, we have a group structure (Theorem 2.5). Let I
K
denote
the group of fractional ideals of K. Let P
K
denote the subgroup of I
K
formed
by the principal ideals, that is ideals of the form O
K
, K
.
Denition 4.1. The ideal class group, denoted by Cl(K), is
Cl(K) = I
K
/P
K
.
Denition 4.2. We denote by h
K
the cardinality [Cl
K
[, called the class num-
ber.
In particular, if O
K
is a principal ideal domain, then Cl(K) = 0, and h
K
= 1.
Our goal is now to prove that the class number is nite for ring of integers
of number elds. The lemma below is a version of Minkowskis theorem.
49
50 CHAPTER 4. IDEAL CLASS GROUP AND UNITS
Lemma 4.1. Let be a lattice of R
n
. Let X R
n
be a convex, compact set
(that is a closed and bounded set since we are in R
n
), which is symmetric with
respect to 0 (that is, x X x X). If
Vol(X) 2
n
Vol(R
n
/),
then there exists 0 ,= such that X.
Proof. Let us rst assume that the inequality is strict: Vol(X) > 2
n
Vol(R
n
/).
Let us consider the map
:
1
2
X =
x
2
R
n
[ x X R
n
/.
If were injective, then
Vol
_
1
2
X
_
=
1
2
n
Vol(X) Vol(R
n
/)
that is Vol(X) 2
n
Vol(R
n
/), which contradicts our assumption. Thus
cannot be injective, which means that there exist x
1
,= x
2

1
2
X such that
(x
1
) = (x
2
). By symmetry, we have that x
2

1
2
X, and by convexity of X
(that is (1 t)x + ty X for t [0, 1]), we have that
_
1
1
2
_
x
1
+
1
2
(x
2
) =
x
1
x
2
2

1
2
X.
Thus 0 ,= = x
1
x
2
X, and (since (x
1
x
2
) = 0).
Let us now assume that Vol(X) = 2
n
Vol(R
n
/). By what we have just
proved, there exists 0 ,=
such that
(1 + )X for all > 0, since

Vol((1 + )X) = (1 + )
n
Vol(X)
= (1 + )
n
2
n
Vol(R
n
/)
> 2
n
Vol(R
n
/), for all > 0.
In particular, if < 1, then
2X. The set 2X is compact and discrete

(since is discrete), it is thus nite. Let us now understand what is happening
here. On the one hand, we have a sequence
with innitely many terms since

there is one for every 0 < < 1, while on the other hand, those innitely many
terms are all lattice points in 2X, which only contains nitely many of them.
This means that this sequence must converge to a point 0 ,= which belongs
to (1 + )X for innitely many > 0. Thus (
0
(1 + )X 0). Since
X is closed, we have that X.
Let n = [K : Q] be the degree of K and let (r
1
, r
2
) be the signature of
K. Let
1
, . . . ,
r
1
be the r
1
real embeddings of K into R. We choose one of
the two embeddings in each pair of complex embeddings, which we denote by
4.1. IDEAL CLASS GROUP 51
r
1
+1
, . . . ,
r
1
+r
2
. We consider the following map, called canonical embedding
of K:
: K R
r
1
C
r
2
R
n
(
1
(), . . . ,
r
1
(),
r
1
+1
(), . . . ,
r
1
+r
2
()). (4.1)
We have that the image of O
K
by is a lattice (O
K
) in R
n
(we have that
(O
K
) is a free abelian group, which contains a basis of R
n
). Let
1
, . . . ,
n
be
a Z-basis of O
K
. Let M be the generator matrix of the lattice (O
K
), given by
_
_
_
1
(
1
) . . .
r
1
(
1
) Re(
r
1
+1
(
1
)) Im(
r
1
+1
(
1
)) . . . Re(
r
1
+r
2
(
1
)) Im(
r
1
+r
2
(
1
))
.
.
.
.
.
.
1
(
n
) . . .
r
1
(
n
) Re(
r
1
+1
(
n
)) Im(
r
1
+1
(
n
)) . . . Re(
r
1
+r
2
(
n
)) Im(
r
1
+r
2
(
n
))
_
_
_
whose determinant is given by
Vol(R
n
/(O
K
)) = [ det(M)[ =
_
[
K
[
2
r
2
.
Indeed, we have that Re(x) = (x + x)/2 and Im(x) = (x x)/2i, x C, and
[ det(M)[ = [ det(M
)[
where M
is given by
_
_
_
_
1
(
1
) . . .
r
1
(
1
)
r
1
+1
(
1
)

r
1
+1
(
1
)
r
1
+1
(
1
)
2i
. . .
r
1
+r
2
(
1
)

r
1
+r
2
(
1
)
r
1
+r
2
(
1
)
2i
.
.
.
.
.
.
1
(
n
) . . .
r
1
(
n
)
r
1
+1
(
n
)

r
1
+1
(
n
)
r
1
+1
(
n
)
2i
. . .
r
1
+r
2
(
n
)

r
1
+r
2
(
n
)
r
1
+r
2
(
n
)
2i
_
_
_
_
.
Again, we have that [ det(M
)[ = 2
r
2
[ det(M
)[, with M
given by this time

_
_
_
1
(
1
) . . .
r
1
(
1
)
r
1
+1
(
1
)
r
1
+1
(
1
) . . .
r
1
+r
2
(
1
)
r
1
+r
2
(
1
)
.
.
.
.
.
.
1
(
n
) . . .
r
1
(
n
)
r
1
+1
(
n
)
r
1
+1
(
n
) . . .
r
1
+r
2
(
n
)
r
1
+r
2
(
n
)
_
_
_,
which concludes the proof, since (recall that complex embeddings come by pairs
of conjugates)
[ det(M)[ = 2
r
2
[ det(M
)[ = 2
r
2
_
K
.
We are now ready to prove that Cl(K) = I
K
/P
K
is nite.
Theorem 4.2. Let K be a number eld with discriminant
K
.
1. There exists a constant C = C
r
1
,r
2
> 0 (which only depends on r
1
and
r
2
) such that every ideal class (that is every coset of Cl(K)) contains an
integral ideal whose norm is at most
C
_
[
K
[.
2. The group Cl(K) is nite.
Proof. Recall rst that by denition, a non-zero fractional ideal J is a nitely
generated O
K
-submodule of K, and there exists K
such that J O
K
(if
i
span J as O
K
-module, write
i
=
i
/
i
and set =
i
). The fact that
J O
K
exactly means that J
1
by denition of the inverse of a fractional
ideal (see Chapter 2). The idea of the proof consists of, given a fractional ideal
J, looking at the norm of a corresponding integral ideal J, which we will prove
is bounded as claimed.
Let us pick a non-zero fractional ideal I. Since I is a nitely generated O
K
-
module, we have that (I) is a lattice in R
n
, and so is (I
1
), with the property
that
Vol(R
n
/(I
1
)) = Vol(R
n
/(O
K
))N(I
1
) =
_
[
K
[
2
r
2
N(I)
,
where the rst equality comes from the fact that the volume is given by the
determinant of the generator matrix of the lattice. Now since we have two
lattices, we can write the generator matrix of (I
1
) as being the generator
matrix of (O
K
) multiplied by a matrix whose determinant in absolute value
is the index of the two lattices. Let X be a compact convex set, symmetrical
with respect to 0. In order to get a set of volume big enough to use Minkowski
theorem, we set a scaling factor
n
= 2
n
Vol(R
n
/(I
1
))
Vol(X)
,
so that the volume of X is
Vol(X) =
n
Vol(X) = 2
n
Vol(R
n
/(I
1
)).
By Lemma 4.1, there exists 0 ,= () (I
1
) and () X. Since I
1
,
we have that I is an integral ideal in the same ideal class as I, and
N(I) = [N
K/Q
()[N(I) = [
n
i=1
i
()[N(I) M
n
N(I),
where M = max
xX
[x
i
[, x = (x
1
, . . . , x
n
), so that the maximum over X
gives
n
M. Thus, by denition of
n
, we have that
N(I)
2
n
Vol(R
n
/(I
1
))
Vol(X)
MN(I)
=
2
n
M
2
r
2
Vol(X)
_
K
=
2
r
1
+r
2
M
Vol(X)
. .
C
_
K
.
This completes the rst part of the proof.
4.2. DIRICHLET UNITS THEOREM 53
Figure 4.1: Johann Peter Gustav Lejeune Dirichlet (1805-1859)
We are now left to prove that Cl(K) is a nite group. By what we have
just proved, we can nd a system of representatives J
i
of I
K
/P
K
consisting
of integral ideals J
i
, of norm smaller than C
_
[
K
[. In particular, the prime
factors of J
i
have a norm smaller than C
_
[
K
[. Above the prime numbers
p < C
_
[
K
[, there are only nitely many prime ideals (or in other words,
there are only nitely many integrals with a given norm).
4.2 Dirichlet Units Theorem
By abuse of language, we call units of K the units of O
K
, that is the invertible
elements of O
K
. We have seen early on (Corollary 1.11) that units are charac-
terized by their norm, namely units are exactly the elements of O
K
with norm
1.
Theorem 4.3. (Dirichlet Units Theorem.) Let K be a number eld of
degree n, with signature (r
1
, r
2
). The group O
K
of units of K is the product of
the group (O
K
) of roots of unity in O
K
, which is cyclic and nite, and a free
group on r
1
+ r
2
1 generators. In formula, we have that
O
K
Z
r
1
+r
2
1
(O
K
).
The most dicult part of this theorem is actually to prove that the free
group has exactly r
1
+ r
2
1 generators. This is nowadays usually proven
using Minkowskis theorem. Dirichlet though did not have Minkowskis theorem
available: he proved the unit theorem in 1846 while Minkowski developed the
geometry of numbers only around the end of the 19th century. He used instead
the pigeonhole principle. It is said that Dirichlet got the main idea for his proof
while attending a concert in the Sistine Chapel.
Proof. Let : K R
r
1
C
r
2
R
n
be the canonical embedding of K (see
(4.1)). The logarithmic embedding of K is the mapping
: K
R
r
1
+r
2
(log [
1
()[, . . . , log [
r
1
+r
2
())[.
Since () = () +(), is a homomorphism from the multiplicative group
K
to the additive group of R

r
1
+r
2
.
Step 1. We rst prove that the kernel of restricted to O
K
is a nite
group. In order to do so, we prove that if C is a bounded subset of R
r
1
+r
2
, then
C
= x O
K
, (x) C is a nite set. In words, we look at the preimage of
a bounded set by the logarithmic embedding (more precisely, at the restriction
of the preimage to the units of O
K
).
Proof. Since C is bounded, all [
i
(x)[, x O
K
, i = 1, . . . , n belong to some
interval say [a
1
, a], a > 1. Thus the elementary polynomials in the
i
(x) will
also belong to some interval of the same form. Now they are the coecients
of the characteristic polynomial of x, which has integer coecients since x
O
K
. Thus there are only nitely many possible characteristic polynomials of
elements x C
, hence only nitely many possible roots of minimal polynomials

of elements x C
, which shows that x can belong to C
for only nitely many

x. Now if we set C = 0, C
is the kernel ker()[

O
K
of restricted to O
K
and
is thus nite.
Step 2. We now show that ker()[
O
K
consists of exactly all the roots of
unity (O
K
).
Proof. That it does consist of roots of unity (and is cyclic) is a known property
of any subgroup of the multiplicative group of any eld. Thus if x ker()[
O
K
then x is a root of unity. Now conversely, suppose that x
m
= 1. Then x is an
algebraic integer, and
[
i
(x)[
m
= [
i
(x
m
)[ = [1[ = 1
so that [
i
(x)[ = 1, and thus log [
i
(x)[ = 0 for all i, showing that x ker()[
O
K
.
Step 3. We are now ready to prove that O
K
is a nite generated abelian
group, isomorphic to (O
K
) Z
s
, s r
1
+ r
2
.
Proof. By Step 1, we know that (O
K
) is a discrete subgroup of R
r
1
+r
2
, that is,
any bounded subset of R
r
1
+r
2
contains only nitely many points of (O
K
). Thus
(O
K
) is a lattice in R
s
, hence a free Z-module of rank s, for some s r
1
+r
2
.
Now by the rst isomorphism theorem, we have that
(O
K
) O
K
/(O
K
)
with (x) corresponding to the coset x(O
K
). If x
1
(O
K
), . . . , x
s
(O
K
) form a
basis for O
K
/(O
K
) and x O
K
, then x(O
K
) is a nite produce of powers of
the x
i
G, so x is an element of (O
K
) times a nite product of powers of the x
i
.
Since the (x
i
) are linearly independent, so are the x
i
(provided that the notion
of linear independence is translated to a multiplicative setting: x
1
, . . . , x
s
are
multiplicatively independent if x
m
1
1
x
m
s
s
= 1 implies that m
i
= 0 for all i,
4.2. DIRICHLET UNITS THEOREM 55
from which it follows that x
m
1
1
x
m
s
s
= x
n
1
1
x
n
s
s
implies m
i
= n
i
for all i).
The result follows.
Step 4. We now improve the estimate of s and show that s r
1
+ r
2
1.
Proof. If x is a unit, then we know that its norm must be 1. Then
1 = N(x) =
n
i=1
i
(x) =
r
1
i=1
i
(x)
r
1
+r
2
j=r
1
+1
j
(x)
j
(x).
By taking the absolute values and applying the logarithmic embedding, we get
0 =
r
1
i=1
log [
i
(x)[ +
r
1
+r
2
j=r
1
+1
log([
j
(x)[[
j
(x)[)
and (x) = (y
1
, . . . , y
r
1
+r
2
) lies in the hyperplane W whose equation is
r
1
i=1
y
i
+ 2
r
1
+r
2
j=r
1
+1
y
j
= 0.
The hyperplane has dimension r
1
+r
2
1, so as above, (O
K
) is a free Z-module
of rank s r
1
+ r
2
1.
Step 5. We are left with showing that s = r
1
+r
2
1, which is actually the
hardest part of the proof. This uses Minkowski theorem. The proof may come
later... one proof can be found in the online lecture of Robert Ash.
Example 4.1. Consider K an imaginary quadratic eld, that is of the form
K = Q(
d), with d a positive square free integer. Its signature is (r

1
, r
2
) =
(0, 1). We thus have that its group of units is given by
Z
r
1
+r
2
1
G = G,
that is only roots of unity. Actually, we have that the units are the 4rth roots
of unity if K = Q(
1) (that is 1, i), the 6th roots of unity if K = Q(
3)
(that is 1,
3
,
2
3
), and only 1 otherwise.
Example 4.2. For K = Q(
3), we have (r
1
, r
2
) = (2, 0), thus r
1
+r
2
1 = 1,
and (O
K
) = 1. The unit group is given by
O
K
(2 +
3)
Z
.
Denition of ideal class group and class number.
The fact that the class number of a number eld is
nite.
The structure of units in a number eld (the state-
ment of Dirichlets theorem)
Chapter 5
p-adic numbers
The p-adic numbers were rst introduced by the German mathematician K.
Hensel (though they are foreshadowed in the work of his predecessor E. Kum-
mer). It seems that Hensels main motivation was the analogy between the ring
of integers Z, together with its eld of fractions Q, and the ring C[X] of polyno-
mials with complex coecients, together with its eld of fractions C(X). Both
Z and C[X] are rings where there is unique factorization: any integer can be
expressed as a product of primes, and any polynomial can be expressed uniquely
as
P(X) = a(X
1
)(X
2
) . . . (X
n
),
where a and
1
, . . . ,
n
are complex numbers. This is the main analogy Hensel
explored: the primes p Z are analogous to the linear polynomials X
C[X]. Suppose we are given a polynomial P(X) and C, then it is possible
(for example using a Taylor expansion) to write the polynomial in the form
P(X) =
n
i=0
a
i
(X )
i
, a
i
C.
This also works naturally for the integers: given a positive integer m and a
prime p, we can write it in base p, that is
m =
n
i=0
a
i
p
i
, a
i
Z
and 0 a
i
p 1.
The reason such expansions are interesting is that they give local infor-
mation: the expansion in powers of (X ) shows if P(X) vanishes at , and
to what order. Similarly, the expansion in base p will show if m is divisible by
p, and to what order.
57
58 CHAPTER 5. P-ADIC NUMBERS
Figure 5.1: Kurt Hensel (1861-1941)
Now for polynomials, one can go a little further, and consider their Laurent
expansion
f(X) =
in
0
a
i
(X )
i
,
that is any rational function can be expanded into a series of this kind in terms
of each of the primes (X ). From an algebraic point of view, we have
two elds: C(X) of all rational functions, and another eld C((X )) which
consists of all Laurent series in (X ). Then the function
f(X) expansion around (X )
denes an inclusion of elds
C(X) C((X )).
Hensels idea was to extend the analogy between Z and C[X] to include the
construction of such expansions. Recall that the analogous of choosing is
choosing a prime number p. We already know the expansion for a positive
integer m, it is just the base p representation. This can be extended for rational
numbers
x =
a
b
=
nn
0
a
n
p
n
yielding for every rational number x a nite-tailed Laurent series in powers of
p, which is called a p-adic expansion of x.
5.1. P-ADIC INTEGERS AND P-ADIC NUMBERS 59
We will come back to this construction in this chapter, and also see that it
achieves Hensels goal, since the set of all nite-tailed Laurent series in powers
of p is a eld, denoted by Q
p
, and that we similarly get a function
f(X) expansion around (X )
which denes an inclusion of elds
Q Q
p
.
Of course, more formalism has been further introduced since Hensels idea,
which will be presented in this chapter.
5.1 p-adic integers and p-adic numbers
We start this chapter by introducing p-adic integers, both intuitively by referring
to writing an integer in a given base p, and formally by dening the concept of
inverse limit. This latter approach will allow to show that p-adic integers form
a ring, denoted by Z
p
. We will then consider fractions of p-adic integers, that
is p-adic numbers, which we will show form the eld Q
p
.
Let p be a prime number. Given an integer n > 0, we can write n in base p:
n = a
0
+ a
1
p + a
2
p
2
+ . . . + a
k
p
k
with 0 a
i
< p.
Denition 5.1. A p-adic integer is a (formal) serie
= a
0
+ a
1
p + a
2
p
2
+
with 0 a
i
< p.
The set of p-adic integers is denoted by Z
p
. If we cut an element Z
p
at
its kth term
k
= a
0
+ a
1
p + + a
k1
p
k1
we get a well dened element of Z/p
k
Z. This yields mappings
Z
p
Z/p
k
Z.
A sequence of
k
, k > 0, such that
k
mod p
k

k
for all k
< k denes a
unique p-adic integer Z
p
(start with k = 1,
1
= a
0
, then for k = 2, we
need to have
2
= a
0
+ a
1
p for it to be a partial sum coherent with
1
). We
thus have the following bijection:
Z
p
= lim
Z/p
k
Z.
The notation on the right hand side is called inverse limit. Here we have an
inverse limit of rings (since Z/p
k
Z is a ring). The formal denition of an inverse
limit involves more formalism than we need for our purpose. To dene an inverse
limit of rings, we need a sequence of rings, which is suitably indexed (here the
sequence Z/p
k
Z is indexed by the integer k). We further need a sequence of
ring homomorphisms
ij
with the same index (here
ij
with i and j integers,
i j) satisfying that
1.
ii
is the identity on the ring indexed by i for all i,
2. for all i, j, k, i j k, we have
ij

jk
=
ik
.
In our case,
ij
: Z/p
j
Z Z/p
i
Z is the natural projection for i j, and the
inverse limit of rings we consider is dened by
lim
Z/p
i
Z = (x
i
)
i

i
Z/p
i
Z [
ij
(x
j
) = x
i
, i j.
Example 5.1. We can write 1 as a p-adic integer:
1 = (p 1) + (p 1)p + (p 1)p
2
+ (p 1)p
3
+ . . .
The description of Z
p
as limit of Z/p
k
Z allows to endow Z
p
with a commuta-
tive ring structure: given , Z
p
, we consider their sequences
k
,
k
Z/p
k
Z.
We then form the sequence
k
+
k
Z/p
k
Z which yields a well dened element
+ Z
p
. We do the same for multiplication.
Example 5.2. Let us compute the sum of = 2+13+. . . and = 1+23+. . .
in Z
3
. We have
1
2 mod 3 and
1
1 mod 3, thus
( + )
1
=
1
+
1
0 mod 3.
Then
2
5 mod 3
2
and
2
7 mod 3
2
, so that
( + )
2
=
2
+
2
= 12 3 mod 3
2
.
This yields
+ = 0 + 1 3 + . . . Z
3
.
We are just computing the addition in base 3!
Note that Z is included in Z
p
.
Let us now look at fractions instead of integers. The fraction 3/2 is the
solution of the equation 2x + 3 = 0. Does this equation have a solution in Z
3
?
We have that
3
2
=
3
1 3
= 3(1 + 3 + 3
2
+ . . .)
since
1
1 x
= 1 + x + x
2
+ .
Thus
3
2
= 1 3 + 1 3
2
+ 1 3
3
+ . . .
5.1. P-ADIC INTEGERS AND P-ADIC NUMBERS 61
Actually, if x = a/b and p does not divide b, then x = a/b Z
p
. Indeed, there
is an inverse b
1
Z/p
k
Z and the sequence ab
1
converges towards an x Z
p
such that bx = a. On the contrary, 1/p , Z
p
, since for all x Z
p
, we have that
(px)
1
= 0 ,= 1.
Denition 5.2. The p-adic numbers are series of the form
a
n
1
p
n
+ a
n+1
1
p
n1
+ + a
1
1
p
+ a
0
+ a
1
p + . . .
The set of p-adic numbers is denoted by Q
p
. It is a eld. We have an inclusion
of Q into Q
p
. Indeed, if x Q, then there exists N 0 such that p
N
x Z
p
. In
other words, Q can be seen as a subeld of Q
p
.
Example 5.3. Let p = 7. Consider the equation
X
2
2 = 0
in Z
7
. Let = a
0
+ a
1
7 + a
2
7
2
+ . . . be the solution of the equation. Then
we have that a
2
0
2 0 mod 7. We thus two possible values for a
0
:
1
= a
0
= 3,
1
= a
0
= 4.
We will see that those two values will give two solutions to the equation. Let
us choose a
0
= 3, and set
2
= a
0
+ a
1
7 Z/49Z.
We have that
2
2
2 0 mod 7
2
a
2
0
+ a
2
1
7
2
+ 2 7a
0
a
1
2 0 mod 7
2
3
2
+ 2 3 7 a
1
2 0 mod 7
2
7 + 6 7 a
1
0 mod 7
2
1 + 6 a
1
0 mod 7
a
1
1 mod 7.
By iterating the above computations, we get that
= 3 + 1 7 + 2 7
2
+ 6 7
3
+ 1 7
4
+ 2 7
5
+ . . .
The other solution is given by
= 4 + 5 7 + 4 7
2
+ 0 7
3
+ 5 7
4
+ 4 7
5
+ . . .
Note that X
2
2 does not have solutions in Q
2
or in Q
3
.
In the above example, we solve an equation in the p-adic integers by solving
each coecient one at a time modulo p, p
2
, . . . If there is no solution for one
coecient with a given modulo, then there is no solution for the equation, as
this is the case for Q
2
or Q
3
.
In the similar spirit, we can consider looking for roots of a given equation in
Q. If there are roots in Q, then there are also roots in Q
p
for every p (that
is, in all the Q
p
and in R). Hence we can conclude that there are no rational
roots if there is some p for which there are no p-adic roots. The fact that
roots in Q automatically are roots in Q
p
for every p means that a global root
is also a local root everywhere (that is at each p).
Much more interesting would be a converse: that local roots could be
patched together to give a global root. That putting together local in-
formation at all p should give global information is the idea behind the
so-called local-global principle, rst clearly stated by Hasse. A good example
where this principle is successful is the Hasse-Minkowski theorem:
Theorem 5.1. (Hasse-Minkowski) Let F(X
1
, . . . , X
n
) Q[X
1
, . . . , X
n
] be a
quadratic form (that is a homogeneous polynomial of degree 2 in n variables).
The equation
F(X
1
, . . . , X
n
) = 0
has non-trivial solutions in Q if and only if it has non-trivial solutions in Q
p
for each p .
5.2 The p-adic valuation
We now introduce the notion of p-adic valuation and p-adic absolute value. We
rst dene them for elements in Q, and extend them to elements in Q
p
after
proving the so-called product formula. The notion of absolute value on Q
p
enables to dene Cauchy sequences, and we will see that Q
p
is actually the
completion of Q with respect to the metric induced by this absolute value.
Let be a non-zero element of Q. We can write it as
= p
k
g
h
, k Z,
and g, h, p coprime to each other, with p prime. We set
ord
p
() = k
[[
p
= p
k
ord
p
(0) =
[0[
p
= 0.
We call ord
p
() the p-adic valuation of and [[
p
the p-adic absolute value of
. We have the following properties for the p-adic valuation:
ord
p
: Q Z
ord
p
(ab) = ord
p
(a) + ord
p
(b)
ord
p
(a + b) min(ord
p
(a), ord
p
(b))
ord
p
(a) = a = 0.
5.2. THE P-ADIC VALUATION 63
Let us now look at some properties of the p-adic absolute value:
[ [
p
: Q R
0
[ab[
p
= [a[
p
[b[
p
[a + b[
p
max([a[
p
, [b[
p
) [a[
p
+[b[
p
[a[
p
= 0 a = 0.
Note that in a sense, we are just trying to capture for this new absolute value
the important properties of the usual absolute value. Now the p-adic absolute
value induces a metric on Q, by setting
d
p
(a, b) = [a b[
p
,
which is indeed a distance (it is positive: d
p
(a, b) 0 and is 0 if and only if
a = b, it is symmetric: d
p
(a, b) = d
p
(b, a), and it satises the triangle inequality:
d
p
(a, c) d
p
(a, b) + d
p
(b, c)). With that metric, two elements a and b are close
if [a b[
p
is small, which means that ord
p
(a b) is big, or in other words, a big
power of p divides a b.
The following result connects the usual absolute value of Q with the p-adic
absolute values.
Lemma 5.2. (Product Formula) Let 0 ,= Q. Then
[[
= 1
where , 2, 3, 5, 7, . . . and [[
is the real absolute value of .

Proof. We prove it for a positive integer, the general case will follow. Let
be a positive integer, which we can factor as
= p
a
1
1
p
a
2
2
p
a
k
k
.
Then we have
_
_
_
[[
q
= 1 if q ,= p
i
[[
p
i
= p
a
i
i
for i = 1, . . . , k
[[
= p
a
1
1
p
a
k
k
The result follows.
In particular, if we know all but one absolute value, the product formula
allows us to determine the missing one. This turns out to be surprisingly impor-
tant in many applications. Note that a similar result is true for nite extensions
of Q, except that in that case, we must use several innite primes (actually
one for each dierent inclusion into R and C). We will come back to this result
in the next chapter.
The set of primes together with the innite prime, over which the product
is taken in the product formula, is usually called the set of places of Q.
Denition 5.3. The set
/
Q
= , 2, 3, . . .
is the set of places of Q.
Let us now get back to the p-adic numbers. Let = a
k
p
k
+ a
k+1
p
k+1
+
a
k+2
p
k+2
+ . . . Q
p
, with a
k
,= 0, and k possibly negative. We then set
ord
p
() = k
[[
p
= p
k
.
This is an extension of the denition of absolute value dened for elements of
Q.
Before going on further, let us recall two denitions:
Recall that a sequence of elements x
n
in a given eld is called a Cauchy
sequence if for every > 0 one can nd a bound M such that we have
[x
n
x
m
[ < whenever m, n M.
A eld K is called complete with respect to an absolute value [ [ if every
Cauchy sequence of elements of K has a limit in K.
Let Q
p
. Recall that
l
is the integer 0
l
< p
l
obtained by cutting
after a
l1
p
l1
. If n > m, we have
[
n
m
[
p
= [a
k
p
k
+ . . . + a
m
p
m
+ . . . + a
n1
p
n1
a
k
p
k
. . . a
m1
p
m1
[
= [a
m
p
m
+ a
m+1
p
m+1
+ . . . + a
n1
p
n1
[
p
p
m
.
This expression tends to 0 when m tends to innity. In other words, the
sequence (
n
)
n0
is a Cauchy sequence with respect to the metric induced by
[ [
p
.
Now let (
n
)
n1
be a Cauchy sequence, that is [
n
m
[
p
0 when m
with n > m, that is,
n

m
is more and more divisible by p, this is just the
interpretation of what it means to be close with respect to the p-adic absolute
value. The writing of
n
and
m
in base p will thus be the same for more and
more terms starting from the beginning, so that (
n
) denes a p-adic number.
This may get clearer if one tries to write down two p-adic numbers. If a, b
are p-adic integers, a = a
0
+a
1
p+a
2
p
2
+. . ., b = b
0
+b
1
p+b
2
p
2
+. . ., if a
0
,= b
0
,
then [ab[
p
= p
0
= 1 if p does not divide a
0
b
0
, and [ab[
p
= p
1
if p[a
0
b
0
,
but [a b[
p
cannot be smaller than 1/p, for which we need a
0
= b
0
. This
works similarly for a, b p-adic numbers. Then we can write a = a
k
1/p
k
+ . . .,
b = b
l
1/p
l
+ . . .. If k ,= l, say k > l, then [a b[
p
= [b
l
1/p
l
+ . . . + a
k
/p
k
+
. . . [
p
= p
l
, which is positive. The two p-adic numbers a and b are thus very far
apart. We see that for the distance between a and b to be smaller than 1, we
rst need all the coecients a
i
, b
i
, to be the same, for i = k, . . . , 1. We are
then back to the computations we did for a and b p-adic integers.
We have just shown that
5.2. THE P-ADIC VALUATION 65
Theorem 5.3. The eld of p-adic numbers Q
p
is a completion of Q with respect
to the p-adic metric induced by [ [
p
.
Now that we have a formal denition of the eld of the p-adic numbers, let
us look at some of its properties.
Proposition 5.4. Let Q
p
be the eld of the p-adic numbers.
1. The unit ball Q
p
[ [[
p
1 is equal to Z
p
.
2. The p-adic units are
Z
p
= Z
p
[ 0 ,= a
0
(Z/pZ)
= Z
p
[ [[
p
= 1.
3. The only non-zero ideals of Z
p
are the principal ideals
p
k
Z
p
= Q
p
[ ord
p
() k.
4. Z is dense in Z
p
.
Proof. 1. We look at the unit ball, that is Q
p
such that [[
p
1. By
denition, we have
[[
p
1 p
ord
p
()
1 ord
p
() 0.
This is exactly saying that belongs to Z
p
.
2. Let us now look at the units of Z
p
. Let be a unit. Then
Z
p
Z
p
and
1
Z
p
[[
p
1 and [1/[
p
1 [[
p
= 1.
3. We are now interested in the ideals of Z
p
. Let I be a non-zero ideal of Z
p
,
and let be the element of I with minimal valuation ord
p
() = k 0.
We thus have that
= p
k
(a
k
+ a
k+1
p + . . .)
where the second factor is a unit, implying that
Z
p
= p
k
Z
p
I.
We now prove that I p
k
Z
p
, which concludes the proof by showing that
I = p
k
Z
p
. If I is not included in p
k
Z
p
, then there is an element in I out of
p
k
Z
p
, but then this element must have a valuation smaller than k, which
cannot be by minimality of k.
4. We now want to prove that Z is dense in Z
p
. Formally, that means that for
every element Z
p
, and every > 0, we have B(, ) Z is non-empty
(where B(, ) denotes an open ball around of radius ).
Let us thus take Z
p
and > 0. There exists a k big enough so that
p
k
< . We set Z the integer obtained by cutting the serie of after
a
k1
p
k1
. Then
= a
k
p
k
+ a
k+1
p
k+1
+ . . .
implies that
[ [
p
p
k
< .
Thus Z is dense in Z
p
. Similarly, Q is dense in Q
p
.
Denition of p-adic integers using p-adic expansions, inverse
limit, and that they form a ring Z
p
Denition of p-adic numbers using p-adic expansions, and
that they form a eld Q
p
Denition of p-adic valuation and absolute value
The product formula
The formal denition of Q
p
as completion of Q, and that Z
p
can then be dened as elements of Q
p
with positive p-adic
valuation.
Ideals and units of Z
p
.
Chapter 6
Valuations
In this chapter, we generalize the notion of absolute value. In particular, we will
show how the p-adic absolute value dened in the previous chapter for Q can be
extended to hold for number elds. We introduce the notion of archimedean and
non-archimedean places, which we will show yield respectively innite and nite
places. We will characterize innite and nite places for number elds, and show
that they are very well known: innite places correspond to the embeddings of
the number eld into C while nite places are given by prime ideals of the ring
of integers.
6.1 Denitions
Let K be a eld.
Denition 6.1. An absolute value on K is a map [ [ : K R
0
which satises
[[ = 0 if and only if = 0,
[[ = [[[[ for all , K
there exists a > 0 such that [ + [
a
[[
a
+[[
a
.
We suppose that the absolute value [ [ is not trivial, that is, there exists
K with [[ , = 0 and [[ , = 1.
Note that when a = 1 in the last condition, we say that [ [ satises the
triangle inequality.
Example 6.1. The p-adic absolute valuation [ [
p
of the previous chapter,
dened by [[
p
= p
ord
p
()
satises the triangle inequality.
Denition 6.2. Two absolute values are equivalent if there exists a c > 0 such
that [[
1
= ([[
2
)
c
. An equivalence class of absolute value is called a place of
K.
67
68 CHAPTER 6. VALUATIONS
Example 6.2. Ostrowskis theorem, due to the mathematician Alexander Os-
trowski, states that any non-trivial absolute value on the rational numbers Q is
equivalent to either the usual real absolute value ([ [) or a p-adic absolute value
([ [
p
). Since [ [ = [ [
, we have that the places of Q are [ [

p
, p . By
analogy we also call p places of Q.
Note that any valuation makes K into a metric space with metric given by
d(x
1
, x
2
) = [x
1
x
2
[
a
. This metric does depend on a, however the induced
topology only depends on the place. This is what the above denition really
means: two absolute values on a eld K are equivalent if they dene the same
topology on K, or again in other words, that every set that is open with respect
to one topology is also open with respect to the other (recall that by open set,
we just mean that if an element belongs to the set, then it also belongs to an
open ball that is contained in the open set).
Lemma 6.1. Let [ [
1
and [ [
2
be absolute values on a eld K. The following
statements are equivalent:
1. [ [
1
and [ [
2
dene the same topology;
2. for any K, we have [[
1
< 1 if and only if [[
2
< 1;
3. [ [
1
and [ [
2
are equivalent, that is, there exists a positive real c > 0 such
that [[
1
= ([[
2
)
c
.
Proof. We prove 1. 2. 3. 1.
(1. 2.) If [ [
1
and [ [
2
dene the same topology, then any sequence that
converges with respect to one absolute value must also converge in the other.
But given any K, we have that
lim
n
n
= 0 lim
n
[
n
[ = 0
with respect to the topology induced by an absolute value [ [ ( may it be [ [
1
or [ [
2
) if and only if [[ < 1. This gives 2.
(2. 3.) Since [ [
1
is not trivial, there exists an element x
0
K such that
[x
0
[
1
< 1. Let us set c > 0, c R, such that
[x
0
[
c
1
= [x
0
[
2
.
We can always do that for a given x
0
, the problem is now to see that this holds
for any x K. Let 0 ,= x K. We can assume that [x[
1
< 1 (otherwise just
replace x by 1/x). We now set R such that
[x[
1
= [x
0
[
1
.
Again this is possible for given x and x
0
. We can now combine that
[x[
1
= [x
0
[
1
[x[
c
1
= [x
0
[
c
1
6.2. ARCHIMEDEAN PLACES 69
with
[x
0
[
c
1
= [x
0
[
2
[x
0
[
c
1
= [x
0
[
2
to get that
[x[
c
1
= [x
0
[
c
1
= [x
0
[
2
.
We are left to connect [x
0
[
2
with [x[
2
.
If m/n > , with m, n Z, n > 0, then
x
m
0
x
n
1
= [x
mn
0
[
1
x
n
0
x
n
1
= [x
0
[
mn
1
< 1.
Thus, by assumption from 2.,
x
m
0
x
n
2
< 1
that is
[x[
2
> [x
0
[
m/n
2
for all
m
n
> ,
or in other words,
[x[
2
> [x
0
[
+
2
, > 0 [x[
2
[x
0
[
2
.
Similarly, if m/n < , we get that [x[
2
< [x
0
[
m/n
2
[x[
2
[x
0
[
2
. Thus
[x[
2
= [x
0
[
2
= [x
0
[
c
1
= [x[
c
1
for all x K.
(3. 1.) If we assume 3., we get that
[ a[
1
< r [ a[
c
2
< r [ a[
2
< r
1/c
,
so that any open ball with respect to [ [
1
is also an open ball (albeit of dierent
radius) with respect to [ [
2
. This is enough to show that the topologies dened
by the two absolute values are identical. Note that having balls of dierent
radius tells us that the metrics are dierent.
6.2 Archimedean places
Let K be a number eld.
Denition 6.3. An absolute value on a number eld K is archimedean if for
all n > 1, n N, we have [n[ > 1.
The story goes that since for an Archimedean valuation, we have [m[ tends
to innity with m, the terminology recalls the book that Archimedes wrote,
called On Large Numbers.
Proposition 6.2. The only archimedean place of Q is the place of the real
absolute value [ [
.
Proof. Let [ [ be an archimedean absolute value on Q. We can assume that the
triangle inequality holds (otherwise, we replace [ [ by [ [
a
). We have to prove
that there exists a constant c > 0 such that [x[ = [x[
c
for all x Q. Let us

rst start by proving that this is true for positive integers.
Let m, n > 1 be integers. We write m in base n:
m = a
0
+ a
1
n + a
2
n
2
+ . . . + a
r
n
r
, 0 a
i
< n.
In particular, m n
r
, and thus
r
log m
log n
.
Thus, we can upper bound [m[ as follows:
[m[ [a
0
[ +[a
1
[[n[ + . . . +[a
r
[[n[
r
([a
0
[ +[a
1
[ + . . . +[a
r
[)[n[
r
since [n[ > 1
(1 + r)[n[
r+1
_
1 +
log m
log n
_
[n[
log m
log n
+1
.
Note that the second inequality is not true for example for the p-adic absolute
value! We can do similarly for m
k
, noticing that the last term is of order at
most n
rk
. Thus
[m[
k
_
1 +
k log m
log n
_
[n[
k log m
log n
+1
,
and
[m[
_
1 +
k log m
log n
_
1/k
[n[
log m
log n
+1/k
.
If we take the limit when k (recall that
n
n 1 when n ), we nd
that
[m[ [n[
log m
log n
.
If we exchange the role of m and n, we nd that
[n[ [m[
log n
log m
.
Thus combining the two above inequalities, we conclude that
[n[
1/ log n
= [m[
1/ log m
which is a constant, say e
c
. We can then write that
[m[ = e
c log m
= m
c
= [m[
c
since m > 1. We have thus found a suitable constant c > 0, which concludes
the proof when m is a positive integer.
To complete the proof, we notice that the absolute value can be extended
to positive rational number, since [a/b[ = [a[/[b[, which shows that [x[ = [x[
c
for 0 < x Q. Finally, it can be extended to arbitrary elements in Q by noting

that [ 1[ = 1.
6.3. NON-ARCHIMEDEAN PLACES 71
Let K be a number eld and : K C be an embedding of K into C, then
[x[
= [(x)[ is an archimedean absolute value.

Theorem 6.3. Let K be a number eld. Then there is a bijection
archimedean places embeddings of K into C up to conjugation .
The archimedean places are also called places at innity. We say that [ [ is
a real place if it corresponds to a real embedding. A pair of complex conjugate
embeddings is a complex place.
6.3 Non-archimedean places
Let K be a number eld. By denition, an absolute value: [ [ : K R
0
is
non-archimedean if there exists n > 1, n N, such that [n[ < 1.
Lemma 6.4. For a non-archimedean absolute value on Q, we have that
[m[ 1, for all m Z.
Proof. We can assume that [ [ satises the triangle inequality. Let us assume by
contradiction that there exists m Z such that [m[ > 1. There exists M = m
k
such that
[M[ = [m[
k
>
n
1 [n[
,
where n is such that [n[ < 1, which exists by denition. Let us now write M in
base n:
M = a
0
+ a
1
n + . . . + a
r
n
r
which is such that
[M[ [a
0
[ +[a
1
[[n[ + . . . +[a
r
[[n[
r
< n(1 +[n[ + . . . +[n[
r
)
since [a
i
[ = [1 + . . . + 1[ a
i
[1[ < n. Thus
[M[ < n
j0
[n[
j
=
n
1 [n[
which is a contradiction.
Lemma 6.5. Let [ [ be a non-archimedean absolute value which satises the
triangle inequality. Then
[ + [ max[[, [[
for all , K. We call [ [ ultrametric.
Proof. Let k > 0. We have that
[ + [
k
= [( + )
k
[
=
j=0
_
k
j
_
kj
j=0
_
k
j
_
[[
j
[[
kj
.
By the previous lemma, we have that
_
k
j
_
1, so that
[ + [
k
(k + 1) max[[, [[
k
.
Thus
[ + [
k
k + 1 max[[, [[.
We get the result by observing k .
Proposition 6.6. let K be a number eld, and [[ be a non-archimedean absolute
value. Let ,= 0. Then there exists a prime ideal p of O
K
and a constant C > 1
such that
[[ = C
ord
p
()
,
where ord
p
() is the highest power of p which divides O
K
.
Denition 6.4. We call
ord
p
: K
Z
the p-adic valuation.
Proof. We can assume that [ [ satises the triangle inequality. It is enough to
show the formula for O
K
.
We already know that [m[ 1 for all m Z. We now extend this result for
elements of O
K
.
([[ 1 for O
K
). For O
K
, we have an equation of the form
m
+ a
m1
m1
+ . . . + a
1
+ a
0
= 0, a
i
Z.
Let us assume by contradiction that [[ > 1. By Lemma 6.4, we have that
[a
i
[ 1 for all i. In the above equation, the term
m
is thus the one with
maximal absolute value. By Lemma 6.5, we get
[[
m
= [a
m1
m1
+ . . . + a
0
[
max[a
m1
[[[
m1
, . . . , [a
1
[[[, [a
0
[
max[[
m1
, . . . , 1
thus a contradiction. We have thus shown that [[ 1 for all O
K
. We now
set
p = O
K
[ [[ < 1.
6.3. NON-ARCHIMEDEAN PLACES 73
(p is a prime ideal of O
K
.) Let us rst show that p is an ideal of O
K
.
Let p and O
K
. We have that
[[ = [[[[ [[ < 1
showing that p and + p since
[ + [ max[[, [[ < 1
where the rst inequality follows from Lemma 6.5. Let us now show that p is a
prime ideal of O
K
. If , O
K
are such that p, then [[[[ < 1, which
means that at least one of the two terms has to be < 1, and thus either or
are in p.
(There exists a suitable C > 1.) We now choose in p but not in p
2
and
let be an element of O
K
. We set m = ord
p
(). We consider /
m
, which is
of valuation 0 (by choice of and m). We can write
m
O
K
= IJ
1
with I and J are integral ideals, both prime to p. By the Chinese Remainder
Theorem, there exists O
K
, J and prime to p. We furthermore set
=

m
I O
K
.
Since both and are elements of O
K
not in p, we have that [[ = 1 and
[[ = 1 (if this is not clear, recall the denition of p above). Thus
= 1.
We have nally obtained that
[[ = [[
m
for all O
K
, so that we conclude by setting
C =
1
[[
.
Corollary 6.7. For a number eld K, we have the following bijection
places of K real embeddingspairs of complex embeddingsprime ideals .
For each place of a number eld, there exists a canonical choice of absolute
values (called normalized absolute values).
real places:
[[ = [()[
R
,
where is the associated embedding.
complex places:
[[ = [()[
2
C
= [() ()[
R
,
where (, ) is the pair of associated complex embeddings.
nite places (or non-archimedean places):
[[ = N(p)
ord
p
()
where p is the prime ideal associated to [ [.
Proposition 6.8. (Product Formula). For all 0 ,= K, we have
[[
= 1
where the product is over all places , and all the absolute values are normalized.
Proof. Let us rewrite the product as
[[
nite
[[
innite
[[
We now compute N(O

K
) in two ways, one which will make appear the nite
places, and the other the innite places. First,
N(O
K
) =
p
N(p)
ord
p
()
=
nite
[[
1
which can be alternatively computed by

N(O
K
) = [N
K/Q
()[
R
=
[()[
C
=
innite
[[
.
6.4 Weak approximation
We conclude this chapter by proving the weak approximation theorem. The
term weak can be thought by opposition to the strong approximation the-
orem, where in the latter, we will state the existence of an element in O
K
,
while we are only able to guarantee this element to exist in K for the former.
Those approximation theorems (especially the strong one) restate the Chinese
Remainder Theorem in the language of valuations.
Let K be a number eld.
Lemma 6.9. Let be a place of K and
1
, . . . ,
N
be places dierent from .
Then there exists K such that [[
> 1 and [[
i
< 1 for all i = 1, . . . , N.
6.4. WEAK APPROXIMATION 75
Proof. We do a proof by induction on N.
(N=1). Since [ [
1
is dierent from [ [
, they induce dierent topologies,

and thus there exists K with
[[
1
< 1 and [[
1
(recall that we proved above that if the two induced topologies are the same,
then [[
1
[ < 1 implies [[
[ < 1). Similarly, there exists K with

[[
< 1 and [[
1
1.
We thus take =
1
.
(Assume true for N 1). We assume N 2. By induction hypothesis,
there exists K with
[[
> 1 and [[
i
< 1, i = 1, . . . , N 1.
Again, as we proved in the case N = 1, we can nd with
[[
> 1 and [[
N
< 1.
We have now 3 cases:
if [[
N
< 1: then take = . We have that [[
> 1, [[
i
< 1,
i = 1, . . . , N 1 and [[
N
< 1.
if [[
N
= 1: we have that
r
0 in the
i
-adic topology, for all i < N.
There exists thus r >> 0 such that
=
r
which satises the required inequalities. Note that [[
> 1 and [[
N
> 1
are immediately satised, the problem is for
i
, i = 1, . . . , N 1 where we
have no control on [[
i
and need to pick r >> 0 to satisfy the inequality.
if [[
N
> 1: we then have that
r
1 +
r
=
1
1 +
1
r
_
1 for [ [
N
0 for [ [
i
, i < N
Take
=

r
1 +
r
, r >> 0.
Theorem 6.10. Let K be a number eld, > 0,
1
, . . . ,
m
be distinct places
of K, and
1
, . . . ,
m
K. Then there exist K such that
[
i
[
i
< .
Proof. By the above lemma, there exist
j
K with [
j
[
j
> 1 and [
j
[
i
< 1
for i ,= j. Set
r
=
m
j=1
r
j
1 +
r
j
j
.
When r , we have
r

j
for the
j
-adic topology, since as in the above
proof
r
1 +
r
=
1
1 +
1
r
_
1 for [
j
[
j
> 1
0 for [
j
[
i
< 1, i ,= j.
Thus take =
r
, r >> 0.
Let K
i
be the completion of K with respect to the
i
-adic topology. We
can restate the theorem by saying that the image of
K
m
i=1
K
i
, x (x, x, . . . , x)
is dense.
Denition of absolute value, of place, of archimedean and
non-archimedean places
What are the nite/innite places for number elds
The product formula
Chapter 7
p-adic elds
In this chapter, we study completions of number elds, and their ramication
(in particular in the Galois case). We then look at extensions of the p-adic num-
bers Q
p
and classify them through their ramication, though they are actually
completion of number elds. We will address again the question of ramication
in number elds, and see how ramication locally can help us to understand
ramication globally.
By p-adic elds, we mean, in modern terminology, local elds of character-
istic zero.
Denition 7.1. Let K be a number eld, and let p be a prime. Let be
the place associated with p and [ [
= N(p)
ord
p
()
(recall that a place is an
equivalence class of absolute values, inside which we take as representative the
normalized absolute value). We set K
or K
p
the completion of K with respect
to the [ [
-adic topology. The eld K
admits an absolute value, still denoted

by [ [
, which extends the one of K.

In other words, we can also dene K
as
K
=
(x
n
) [ (x
n
) is a Cauchy sequence with respect to [ [
(x
n
) [ x
n
0
.
This is a well dened quotient ring, since the set of Cauchy sequence has a ring
structure, and those which tend to zero form a maximal ideal inside this ring.
Intuitively, this quotient is here to get the property that all Cauchy sequences
whose terms get closer and closer to each other have the same limit (and thus
dene the same element in K
).
Example 7.1. The completion of Q with respect to the induced topology by
[ [
p
is Q
p
.
Below is an example with an innite prime.
77
78 CHAPTER 7. P-ADIC FIELDS
Example 7.2. If is a real place, then K
= R. If is a complex place, then

K
= C.
Let us now compute an example where K is not Q.
7). We want to compute its completion K
where
is a place above 3. Since
3O
K
= (2
7)(2 +
7),
there are two places
1
,
2
above 3, corresponding to the two nite primes
p
1
= (2
7)O
K
, p
2
= (2 +
7)O
K
.
Now the completion K
where is one of the

i
, i = 1, 2, is an extension of Q
3
,
since the
i
-adic topology on K extends the 3-adic topology on Q.
Since K = Q[X]/(X
2
7), we have that K contains a solution for the
equation X
2
7. We now look at this equation in Q
3
, and similarly to what we
have computed in Example 5.3, we have that a solution is given by
1 + 3 + 3
2
+ 2 3
4
+ . . .
Thus
K
Q
3
.
One can actually show that the two places correspond to two embeddings of K
into Q
3
.
In the following, we consider only nite places. Let be a nite place of a
number eld.
Denition 7.2. We dene the integers of K
by
O
= x K
[ [x[
1.
The denition of absolute value implies that O
is a ring, and that

m
= x K
[ [x[
< 1
is its unique maximal ideal (an element of O
not in m
is a unit of O
). Such
a ring is called a local ring.
Example 7.4. The ring of integers O
of K
= Q
p
is Z
p
, and m
= pZ
p
.
We have the following diagram
K K
O
K
O
p m
-
dense
-
dense
-
dense
7.1. HENSELS WAY OF WRITING 79
We already have the notion of residue eld for p, given by
F
p
= O
K
/p.
We can similarly dene a residue eld for m
by
F
= O
/m
.
We can prove that
O
K
/p O
/m
.
7.1 Hensels way of writing
Let
be in m
but not in m
2
, so that ord
m
) = 1. We call
a uniformizer
of m
(or of O
). For example, for Z

p
, we can take = p. We now choose a
system of representatives of O
/m
:
( = c
0
= 0, c
1
, . . . , c
q1
,
where q = [F
p
[ = N(p). For example, for Z
p
, we have ( = 0, 1, 2, . . . , p 1.
The set
c
0
,
k
c
1
, . . . ,
k
c
q1
=
k
(
is a system of representatives for m
k
/m
k+1
.
Lemma 7.1. 1. Every element O
can be written in a unique way as

= a
0
+ a
1
+ a
2
+ . . .
with a
i
(.
2. An element of K
can be written as
= a
k
+ a
k+1
k+1
+ . . . .
3. The uniformizer generates the ideal m
, that is
= m
k
.
4. [[
= [F
[
k
, where = a
k
+ . . ., a
k
,= 0.
Proof. 1. Let O
. Let a
0
( be the representative of the class + m
in O
/m
. We set
1
=
a
0
.
We have that
1
O
, since
[
1
[
=
[ a
0
[
1.
Indeed, a
0
+m
implies that a
0
m
and thus [ a
0
[
.
By replacing by
1
, we nd a
1
( such that
2
=

1
a
1
.
By iterating this process k times, we get
= a
0
+
1
= a
0
+ a
1
+
2
.
.
.
= a
0
+ a
1
+ a
2
+ . . . +
k+1
k+1
.
Thus
[ (a
0
+ a
1
+ a
2
+ . . . + a
k
)[
= [
k+1
[
[
k+1
0
when k , since
and thus by denition of m
, [
< 1.
2. We multiply K
by
ord
m
()
, so that
ord
m
()
and we conclude by 1.
3. It is clear that
m
k
.
Conversely, let us take m
k
. We then have that

a
0
= a
1
= . . . = a
k1
= 0
and thus
= a
k
+ . . .
k
.
4. Since = a
k
+ . . ., a
k
,= 0, we have that
k
= m
k
but not in
m
k+1
, and

k
.
Thus
[[
= [
[
k
.
Now note that if
and
are two uniformizers, then [
[ = [
[, and
thus, we could have taken a uniformizer in the number eld rather than
in its completion, that is,
p but not in p
2
, which yields
[
[ = N(p)
ord
p
(
)
= N(p)
1
= [F
p
[
1
= [F
[
1
.
7.2. HENSELS LEMMAS 81
7.2 Hensels Lemmas
Lemma 7.2. (First Hensels Lemma). Let f(X) O
[X] be a monic poly-

nomial, and let

f(X) F
[X] be the reduction of f modulo m
. Let us assume
that there exist two coprime monic polynomials
1
and
2
in F
[X] such that
f =
1
2
.
Then there exists two monic polynomials f
1
and f
2
in O
[X] such that

f = f
1
f
2
,

f
1
=
1
,

f
2
=
2
.
Proof. We rst prove by induction that we can construct polynomials f
(k)
1
, f
(k)
2
in O
[X], k 1, such that

(1) f f
(k)
1
f
(k)
2
mod m
k
(2) f
(k)
i
f
(k1)
i
mod m
k1
.
(k=1). Since we know by assumption that there exist
1
,
2
such that
f =
1
2
, we lift
i
in a monic polynomial f
(1)
i
O
[X], and we have deg f

(1)
i
=
deg
i
.
(True up to k). We have already built f
(k)
i
. Using the condition (1), there
exists a polynomial g O
[X] such that

f = f
(k)
1
f
2
(k) +
k
g.
Using Bezouts identity for the ring F
[X], there exists polynomials

1
and
2
in F
[X] such that

g =
1
1
+
2
2
since
1
and
2
are coprime. We now lift
i
in a polynomial h
i
O
[X] of
same degree, and set
f
(k+1)
i
= f
(k)
i
+
k
h
i
.
We now need to check that (1) and (2) are satised. (2) is clearly satised by
construction. Let us check (1). We have
f
(k+1)
1
f
(k+1)
2
= (f
(k)
1
+
k
h
1
)(f
(k)
2
+
k
h
2
)
= f
(k)
1
f
(k)
2
+
k
(f
(k)
1
h
2
+ f
(k)
2
h
1
) +
2k
h
1
h
2
(f
k
g) +
k
(f
(k)
1
h
2
+ f
(k)
2
h
1
) mod m
k+1
.
We are now left to show that
(g + f
(k)
1
h
2
+ f
(k)
2
h
1
) 0 mod m
k+1
,
that is
g + f
(k)
1
h
2
+ f
(k)
2
h
1
0 mod m

or again in other words, after reduction mod m
g +

f
(k)
1
h
2
+

f
(k)
2
h
1
0,
which is satised by construction of h
1
and h
2
. So this concludes the proof by
induction.
Let us know conclude the proof of the lemma. We set
f
i
= lim
k
f
(k)
i
which converges by (2). By (1) we have that
f
1
f
2
= lim
k
f
(k)
1
f
(k)
2
= f.
Example 7.5. The polynomial f(X) = X
2
2 Z
7
[X] is factorized as
1
= (X 3),
2
= (X 4)
in F
7
[X].
Corollary 7.3. Let K be a number eld, be a nite place of K, and K
be
its completion. Denote q = [F
[. Then the set

q1
of (q 1)th roots of unity
belongs to O
.
Proof. Let us look at the polynomial X
q1
1. On the nite eld F
with q
elements, this polynomial splits into linear factors, and all its roots are exactly all
the invertible elements of F
. By Hensels lemma, f O
[X] can be completely

factorized. That is, it has exactly q 1 roots in O
. More precisely, we can

write
X
q1
1 =
q1
(X ) O
[X].
Of course, one can rewrite that
q1
belongs to O
since roots of unity are

clearly invertible in O
.
Lemma 7.4. (Second Hensels Lemma). Let f be a monic polynomial in
O
[X] and let f
be its formal derivative. We assume that there exists O
such that
[f()[
< [f
()[
2
.
Then there exists O
such that
f() = 0
and
[ [

[f()[
[f
()[
< [f
()[
.
7.2. HENSELS LEMMAS 83
Proof. We set
0
=
n+1
=
n
n
where
n
=
f(
n
)
f
(
n
)
.
(First part of the proof.) We rst show by induction that
1. [f(
n
)[
< [f(
n1
)[
2. [f
(
n
)[
= [f
()[
.
Let us assume these are true for n 1, and show they still hold for n + 1.
Let us rst note that
[
n
[
=
[f(
n
)[
[f
(
n
)[
<
[f()[
[f
(
n
)[
by 1.
=
[f()[
[f
()[
by 2.
< [f
()[
by assumption.
Since f O
[X] and O
, this means that [f
()[
1, and in particular
implies that
n
O
.
Let us write f(X) = a
0
+ a
1
X + a
2
X
2
+ . . . + a
n
X
n
, so that
f(X +
n
) = a
0
+ a
1
(X +
n
) + a
2
(X
2
+ 2X
n
+
2
n
) + . . . + a
n
(X
n
+ . . . +
n
n
)
= (a
0
+ a
1
n
+ a
2
2
n
+ . . . + a
n
n
n
) + X(a
1
+ a
2
2
n
+ . . . + a
n
n
n1
n
) + X
2
g(X)
= f(
n
) + f
(
n
)X + g(X)X
2
with g(X) O
[X]. We are now ready to prove that the two properties are
satised.
1. Let us rst check that [f(
n+1
)[
< [f(
n
)[
. We have that
f(
n+1
) = f(
n
n
)
= f(
n
) + f
(
n
)(
n
) + g(
n
)
2
n
take X =
n
= g(
n
)
2
n
recall the denition of
Let us now consider its absolute value
[f(
n+1
)[
= [g(
n
)[
[
n
[
2
[
n
[
2

n
O
, g O
[X]
< [f(
n
)[
[f()[
[f
()[
2
by 1. and 2.
< [f(
n
)[
by assumption
2. We now need to prove that [f
(
n+1
)[
= [f
()[
. We have that
[f
(
n+1
)[
= [f
(
n
n
)[
= [f
(
n
)
n
h(
n
)[
take again X =
n
max[f
(
n
)[
, [
n
[
[h(
n
)
[
= max[f
()[
, [
n
[
[h(
n
)
[ by 2.
and equality holds if the two arguments of the maximum are distinct. Now
the rst argument is [f
()[
, while the second is

[
n
[
[h(
n
)
[ [
n
[
h(
n
) O
< [f
()[
,
which completes the rst part of the proof.
(Second part of the proof.) We are now ready to prove that there exists
an element O
which satises the claimed properties. We set

= lim
n
n
.
Note that this sequence converges, since this is a Cauchy sequence. Indeed, for
n > m, we have
[
n
m
[
max[
n
n1
[
, . . . , [
m+1
m
[
= max[
n1
[, . . . , [
m
[
=
1
[f
()[
max[f(
n1
)[
, . . . , [f(
m
)[
by rst part of the proof, part 2.

=
[f(
m
)[
[f
()[
by rst part of the proof, part 1.

which tends to zero by 1. Let us check that as dened above satises the
required properties. First, we have that
f() = f( lim
n
n
) = lim
n
f(
n
) = 0.
Since O
is closed, O
, and we have that

[ [
= lim
n
[
n
[
lim
n
max[
n
n1
[
, . . . , [
1
[
= max[
n1
[, . . . , [
0
[
[f()[
[f
()[
< [f
()[
.
7.3. RAMIFICATION THEORY 85
7.3 Ramication Theory
Let L/K be a number eld extension. Let P and p be primes of L and K
respectively, with P above p. Since nite places correspond to primes, P and p
each induce a place (respectively w and v) such that the restriction of w to K
coincides with v, that is
([ [
w
)
K
= [ [
v
.
This in turn corresponds to a eld extension L
w
/K
v
. We can consider the
corresponding residue class elds:
F
P
= O
L
/P O
w
/m
w
= F
w
F
p
= O
K
/p O
v
/m
v
= F
v
and we have a nite eld extension F
w
/F
v
of degree f = f
P/|p
= f
w|v
. Note
that this means that the inertial degree f is the same for a prime in L/K and
the completion L
w
/K
v
with respect to this prime.
Lemma 7.5. Let
v
be a uniformizer of K
v
. Then
[
v
[
w
= [
w
[
e
w
where e = e
P/|p
= e
w|v
is the ramication index.
Note that this can be rewritten as m
v
O
w
= m
e
w
, which looks more like the
original denition of ramication index.
Proof. We can take
v
K and
w
L. Then
v
p but not in p
2
, and
w
P but not in P
2
. Thus
v
O
K
= pI where I is an ideal coprime to p. If
we lift p and
v
in O
L
, we get
pO
L
=
P
e
P
i
|p
i
,
v
O
L
=
P
e
P
i
|p
i
IO
L
where IO
L
is coprime to the P
i
. Now
ord
P
(
v
) = ord
P
(
P
e
P
i
|p
i
IO
L
) = e
P|p
= e
and
[
v
[
w
= N(P)
ord
P
(
v
)
= (N(P)
1
)
e
= [
w
[
e
w
.
This lemma also means that the ramication index coincides in the eld
extension and in its completion (this completes the same observation we have
just made above for the inertial degree).
Example 7.6. Let
K
v
= Q
p
L
w
= Q
p
(
n
p) = Q
p
[X]/(X
n
p)
The uniformizers are given by
v
= p,
w
=
n
p.
Thus
[
w
[
w
= 1/p
[
v
[
w
= 1/p
n
which can be seen by noting that
[
v
[
w
= [p[
w
= [
n
p[
n
w
which is the result of the Lemma. Thus
e = n
and the extension is totally ramied.
Example 7.7. Consider
K
v
= Q
p
L
w
= Q
p
(
) = Q
p
[X]/(X
2
)
with Z
p
, / (Q
p
)
2
. We have that
w
is still a uniformizer for L
w
, but
that [F
w
: F
v
] = 2.
The next theorem is a local version of the fact that if K is a number eld,
then O
K
is a free Z-module of rank [K : Q].
Theorem 7.6. The O
v
-module O
w
is free of rank
n
w|v
= [L
w
: K
v
] = f
w|v
e
w|v
.
We give no proof, but just mention that the main point of the proof is the
following: if
1
, . . . ,
f
O
v
is a set such that the reductions

i
generates
F
w
as an F
v
-vector space, then the set
k
w
0ke,1jf
is an O
v
-basis of O
w
.
7.4 Normal extensions
Let L/K be a Galois extension of number elds. Recall that the decomposition
group D of a prime P L is given by
D = Gal(L/K) [ (P) = P
and that the inertia group I is the kernel of the map that sends an element
of the Galois group in D to the Galois group Gal(F
P
/F
p
). The corresponding
xed subelds help us to understand the ramication in L/K:
L
L
I
L
D
K
e
f
g
We further have that
[L : K] = efg
(note the contrast with the local case, where we have that
[L
w
: K
v
] = ef
by Theorem 7.6).
To analyze local extensions, that is, the extensions of completions, we can
distinguish three cases:
Case 1. if p completely splits in L, that is g = [L : K] and e = f = 1,
then
[L
w
: K
v
] = ef = 1
and L
w
= K
v
. This is the case described in Example 7.3, namely
K = Q, L = Q(
7), K
v
= Q
3
, L
w
= Q
3
.
Case 2. if p is inert, that is g = e = 1 and f = [L : K], then
[L
w
: K
v
] = [L : K].
In this case,
v
is still a uniformizer for L
w
, but F
w
,= F
v
. This is a
non-ramied extension. For example, consider
K = Q, L = Q(
7), K
v
= Q
5
, L
w
= Q
5
(
7).
Case 3. If p is totally ramied, that is e = [L : K], then
[L
w
: K
v
] = [L : K]
but this time
v
is not a uniformizer for L
w
, and F
w
= F
v
. For example,
consider
K = Q, L = Q(
7), K
v
= Q
7
, L
w
= Q
7
(
7).
Example 7.8. When does the Golden ratio (1+
5)/2 belongs to Q
p
? It is easy
to see that this question can be reformulated as: when is Q
p
(
5) an extension
of Q
p
? Let us consider
K = Q, L = Q(
5), K
v
= Q
p
, L
w
= Q
p
(
5).
Using the above three cases, we see that if p is inert or ramied in Q(
5), then
[L
w
: K
v
] = [L : K] = 2
and the Golden ratio cannot be in Q
p
. This is the case for example for p = 2, 3
(inert), or p = 5 (ramied). On the contrary, if p splits, then Q
p
= Q
p
(
5).
This is for example the case for p = 11 (11 = (4 +
5)(4
5)).
To conclude this section, let us note the following:
Proposition 7.7. If L/K is Galois, we have the following isomorphism:
D
w|v
Gal(L
w
/K
v
).
Compare this local result with the its global counterpart, where we have
that D is a subgroup of Gal(L/K) of index [Gal(L/K) : D] = g.
7.5 Finite extensions of Q
p
Let F/Q
p
be a nite extension of Q
p
. Then one can prove that F is the com-
pletion of a number eld. In this section, we forget about this fact, and start
by proving that
Theorem 7.8. Let F/Q
p
be a nite extension. Then there exists an absolute
value on F which extends [ [
p
.
Proof. Let O be the set of F whose minimal polynomial over Q
p
has
coecients in Z
p
. The set O is actually a ring (the proof is the same as in
Chapter 1 to prove that O
K
is a ring).
We claim that
O = F [ N
F/Q
p
() Z
p
.
To prove this claim, we show that both inclusions hold. First, let us take O,
and prove that its norm is in Z
p
. If O, then the constant coecient a
0
of
its minimal polynomial over Q
p
is in Z
p
by denition of O, and
N
F/Q
p
() = a
m
0
Z
p
for some positive m. For the reverse inclusion, we start with F with
N
F/Q
p
() Z
p
. Let
f(X) = X
m
+ a
m1
X
m1
+ . . . + a
1
X + a
0
7.5. FINITE EXTENSIONS OF Q
P
89
be its minimal polynomial over Q
p
, with a priori a
i
Q
p
, i = 1, . . . , m1. Since
N
F/Q
p
() Z
p
, we have that [a
m
0
[
p
1, which implies that [a
0
[
p
1, that is
a
0
Z
p
. We now would like to show that all a
i
Z
p
, which is the same thing as
proving that if p
k
is the smallest power of p such that g(X) = p
k
f(X) Z
p
[X],
then k = 0. Now let r be the smallest index such that p
k
a
r
Z
p
(r 0 and
r > 0 if k > 0 since then p
k
a
0
cannot be a unit). We have (by choice of r) that
g(X) p
k
X
m
+ . . . + p
k
a
r
X
r
mod p
X
r
(p
k
X
mr
+ . . . + p
k
a
r
) mod p.
Hensels lemma tells that g should have a factorization, which is in contradiction
which the fact that g(X) = p
k
f(X) with f(X) irreducible. Thus r = 0 and
p
k
a
0
Z
p
proving that k = 0.
Let us now go back to the proof of the theorem. We now set for all F:
[[
F
= [N
F/Q
p
()[
1/n
p
where n = [F : Q
p
]. We need to prove that this is an absolute value, which
extends [ [
p
.
To show that it extends [ [
p
, let us restrict to Q
p
. Then
[[
F
= [N
F/Q
p
()[
1/n
p
= [
n
[
1/n
p
= [[
p
.
The two rst axioms of the absolute value are easy to check:
[[
F
= 0 = 0, [[
F
= [[
F
[[
F
.
To show that [ + [
F
max[[
F
, [[
F
, it is enough to show, up to
division by or , that
[[
F
1 [ + 1[
F
1.
Indeed, if say [/[
F
1, then
[/ + 1[
F
1 max[/[
F
, 1
and vice versa. Now we have that
[[
F
1 [N
F/Q
p
()[
1/n
p
1
[N
F/Q
p
()[
p
1
N
F/Q
p
() Z
p
O
by the claim above. Now since O is a ring, we have that both 1 and are
in O, thus + 1 O which implies that [ + 1[
F
1 and we are done.
We set
m = F [ [[
F
< 1
the unique maximal ideal of O and F = O/m is its residue class eld, which is
a nite extension of F
p
. We set the inertial degree to be f = [F : F
p
], and e to
be such that pO = m
e
, which coincide with the denitions of e and f that we
have already introduced.
We now proceed with studying nite extensions of Q
p
based on their rami-
cation. We start with non-ramied extensions.
Denition 7.3. A nite extension F/Q
p
is non-ramied if f = [F : Q
p
], that
is e = 1.
Finite non-ramied extensions of Q
p
are easily classied.
Theorem 7.9. For each f, there is exactly one unramied extension of degree
f. It can be obtained by adjoining to Q
p
a primitive (p
f
1)th root of unity.
Proof. Existence. Let F
p
f = F
p
( ) be an extension of F
p
of degree f, and let
g(X) = X
f
+ a
f1
X
f1
+ . . . + a
1
X + a
0
be the minimal polynomial of over F
p
. Let us now lift g(X) to g(X) Z
p
[X],
which yields an irreducible polynomial over Q
p
. If is a root of g(X), then
clearly Q
p
() is an extension of degree f of Q
p
. To complete the proof, it is
now enough to prove that Q
p
()/Q
p
is a non-ramied extension of Q
p
, for which
we just need to prove that is residue class eld, say F
p
, is of degree f over F
p
.
Since the residue class eld contains a root of g mod p (this is just mod p),
we have that
[F
p
: F
p
] f.
On the other hand, we have that
[F
p
: F
p
] [Q
p
() : Q
p
]
which concludes the proof of existence.
Unicity. We prove here that any extension F/Q
p
which is unramied and
of degree f is equal to the extension obtained by adjoining a primitive (p
f
1)th
root of unity. We already know by Corollary 7.3 that F must contain all the
(p
f
1)th roots of unity. We then need to show that the smallest eld extension
of Q
p
which contains the (p
f
1)th roots of unity is of degree f. Let be a
(p
f
1)th root of unity. We have that
Q
p
Q
p
() F.
But now, the residue class eld of Q
p
() also contains all the (p
f
1)th roots
of unity, so it contains F
p
f , which implies that
[Q
p
() : Q
p
] f.
P
91
Let us now look at totally ramied extensions.
Denition 7.4. A nite extension F of Q
p
is totally ramied if F = F
p
(that
is f = 1 and e = n).
Totally ramied extensions will be characterized in terms of Eisenstein poly-
nomials.
Denition 7.5. The monic polynomial
f(X) = X
m
+ a
m1
X
m1
+ . . . + a
0
Z
p
[X]
is called an Eisenstein polynomial if the two following conditions hold:
1. a
i
pZ
p
,
2. a
0
, p
2
Z
p
.
An Eisenstein polynomial is irreducible.
The classication theorem for nite totally ramied extensions of Q
p
can
now be stated.
Theorem 7.10. 1. If f is an Eisenstein polynomial, then Q
p
[X]/f(X) is
totally ramied.
2. Let F/Q
p
be a totally ramied extension and let
F
be a uniformizer.
Then the minimal polynomial of
F
is an Eisenstein polynomial.
Example 7.9. X
m
p is an Eisenstein polynomial for all m 2, then Q
p
(
m
p)
is totally ramied.
Proof. 1. Let F = Q
p
[X]/f(X), where
f(X) = X
m
+ a
m1
X
m1
+ . . . + a
1
X + a
0
and let e be the ramication index of F. Set m = [F : Q
p
]. We have to
show that e = m.
Let be a root of f, then
m
+ a
m1
m1
+ . . . + a
1
+ a
0
= 0
and
ord
m
(
m
) = ord
m
(a
m1
m1
+ . . . + a
0
).
Since f is an Eisenstein polynomial by assumption, we have that a
i

pZ
p
pO = m
e
, so that
ord
m
(a
m1
m1
+ . . . + a
0
) e
and ord
m
(
m
) e. In particular, ord
m
() 1. Let s be the smallest
integer such that
s
e
ord
m
()
.
Then m e s. If ord
m
(
m
) = e, then ord
m
() =
e
m
and thus s =
e
e/m
= m, and m e m which shows that m = e. To conclude the
proof, we need to show that ord
m
(
m
) > e cannot possibly happen. Let
us thus assume that ord
m
(
m
) > e. This implies that
ord
m
(a
0
) = ord
m
(
m
+ (a
m1
m1
+ . . . + a
1
)) > e.
Since ord
m
(a
0
) = ord
p
(a
0
)e, the second condition for Eisenstein polyno-
mial shows that ord
m
(a
0
) = e, which gives a contradiction.
2. We know from Theorem 7.6 that O is a free Z
p
-module, whose basis is
given by
p
j
k
F
0ke,1jf
so that every element in F can be written as
j,k
b
jk
k
F
p
j
and F = Q
p
[
F
]. Let
f(X) = X
m
+ a
m1
X
m1
+ . . . + a
1
X + a
0
be the minimal polynomial of
F
. Then a
0
= N
F/Q
p
(
F
) is of valuation
1, since
F
is a uniformizer and F/Q
p
is totally ramied. Let us look at

f,
the reduction of f in F
p
[X]. Since F
p
[X] is a unique factorization domain,
we can write
f(X) =
k
i
i
where
i
are irreducible distinct polynomials in F
p
[X]. By Hensels lemma,
we can lift this factorization into a factorization f =
f
i
such that

f
i
=
k
i
i
. Since f is irreducible (it is a minimal polynomial), we have only one
factor, that is f = f
1
, and

f
1
=
k
1
1
. In words, we have that

f is a power
of an irreducible polynomial in F
p
[X]. Then

f = (X a)
m
since

f must
have a root in F
p
= F. Since a
0
0 mod p, we must have a 0 mod p
and

f X
m
mod p. In other words, a
i
pZ
p
for all i. This tells us that
f(X) is an Eisenstein polynomial.
P
93
Denition of the completion K
of a number eld K,
of uniformizer.
Hensels Lemmas.
Local ramication index e
w|v
and inertial degree f
w|v
,
and the local formula n
w|v
= e
w|v
f
w|v
.
Classication of extensions of Q
p
: either non-ramied
(there a unique such extension) or totally ramied.
Bibliography
[1] H. Cohn. A Classical Invitation to Algebraic Numbers and Class Fields.
Springer-Verlag, NY, 1978.
[2] A. Fr ohlich and M.J. Taylor. Algebraic number theory. Cambridge University
Press, Great Britain, 1991.
[3] F.Q. Gouvea. p-adic Numbers: An Introduction. Springer Verlag, 1993.
[4] G. Gras. Class Field Theory. Springer Verlag, 2003.
[5] P. Samuel. Theorie algebrique des nombres. Hermann, 1971.
[6] I.N. Stewart and D.O. Tall. Algebraic Number Theory. Chapman and Hall,
1979.
[7] H.P.F. Swinnerton-Dyer. A Brief Guide to Algebraic Number Theory. Uni-
versity Press of Cambridge, 2001.
95

ANT10

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

ANT10

Enviado por

Direitos autorais:

Formatos disponíveis

Introduction to Algebraic Number Theory

2) is an algebraic integer. Similarly,

2), K = Q, and take

2). In order to compute

on the basis elements:

in the chosen basis. Of course, M changes with a

(X), implying that the

(X). Since deg(f

(X)) < deg(f(X)), this forces

(X) to be zero, which is not possible with K of characteristic 0.

in a K-basis of L. We have that

is a K-basis of K(). Let k be the degree [L : K()] and let

by can now be written in this basis as

2)/Q. It has two embeddings

2). Its two conjugates are

2, thus its norm is given by

2)/Q. Take the element

2), whose characteristic polynomial is given by, say, (X) =

6]. We have that

5) be a quadratic number eld, with ring of

5], since d 3 (mod 4). Let us prove that we do not have

5]. We have that

5 irreducible. Let us show for example that 3 is irreducible.

17) be a quadratic number eld, with ring of

d), d > 0, is (2, 0).

d), d < 0, is (0, 1).

5). Its two embeddings

5)/2], so that the matrix M of embeddings is

2), with ring of integers O

2]. We want to factorize 5O

is another prime ideal above p, then the decomposition groups D

(in formula, we have that if P

(1 + )X for all > 0, since

2X. The set 2X is compact and discrete

with innitely many terms since

given by this time

to the additive group of R

, hence only nitely many possible roots of minimal polynomials

, which shows that x can belong to C

for only nitely many

is the kernel ker()[

d), with d a positive square free integer. Its signature is (r

1) (that is 1, i), the 6th roots of unity if K = Q(

is the real absolute value of .

, we have that the places of Q are [ [

for all x Q. Let us

for 0 < x Q. Finally, it can be extended to arbitrary elements in Q by noting

= [(x)[ is an archimedean absolute value.

We now compute N(O

which can be alternatively computed by

, they induce dierent topologies,

[ < 1). Similarly, there exists K with

which satises the required inequalities. Note that [[

-adic topology. The eld K

admits an absolute value, still denoted

, which extends the one of K.

= R. If is a complex place, then

7). We want to compute its completion K

where is one of the

is a ring, and that

). For example, for Z

can be written in a unique way as

and thus by denition of m