Professional Documents
Culture Documents
MANUALES UEX
ndice general
5
1. Anillos de enteros
1.1. Introduccin . . . . . . . . . . . . . . . . . . .
1.2. Dominios de factorizacin nica . . . . . . .
1.3. Curvas y anillos de enteros . . . . . . . . . .
1.4. Puntos singulares de un anillo de nmeros
1.5. Anillos de Dedekind . . . . . . . . . . . . . .
1.6. Finitud del morfismo de desingularizacin
1.7. Apndice: Mtrica de la traza . . . . . . . .
1.8. Cuestionario . . . . . . . . . . . . . . . . . . .
1.9. Biografa de Dedekind . . . . . . . . . . . . .
1.10.Problemas . . . . . . . . . . . . . . . . . . . .
2. Fibras de los morfismos finitos
2.1. Introduccin . . . . . . . . . . . . . . . . . . .
2.2. Longitud de un mdulo . . . . . . . . . . . .
2.3. Multiplicidades y grados en dimensin cero
2.4. Puntos de las fibras de un morfismo finito .
2.5. Automorfismo de Frbenius . . . . . . . . .
2.6. Aplicaciones . . . . . . . . . . . . . . . . . . .
2.7. Cuestionario . . . . . . . . . . . . . . . . . . .
2.8. Biografa de Frbenius . . . . . . . . . . . .
2.9. Problemas . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
7
10
14
15
17
20
22
23
24
28
.
.
.
.
.
.
.
.
.
29
29
29
31
32
35
36
38
39
43
.
.
.
.
.
.
.
.
.
45
45
48
50
53
54
55
58
59
64
MANUALES UEX
Introduccin
NDICE GENERAL
MANUALES UEX
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
65
65
69
70
72
73
75
77
79
81
83
84
87
Bibliografa
91
ndice de trminos
93
MANUALES UEX
Introduccin
Introduccin
MANUALES UEX
Captulo 1
Anillos de enteros y anillos de
curvas algebraicas
1.1.
Introduccin
a. 2000 = 7 266 + 138. b. 266 = 1 138 + 128. c. 138 = 1 128 + 10. d. 128 = 12 10 + 8 5.
10 = 1 8 + 2. Luego, m.c.d (2000, 266) = 2.
Ahora, 2 = 10 1 8 = 10 1 (128 12 10) = 128 + 13 10 = 128 + 13(138 128) =
13 138 14 128 = 13 138 14(266 138) = 14 266 + 27 138 = 14 266 + 27(2000 7
266) = 27 2000 203 266.
Por tanto, una solucin particular de nuestro sistema de ecuaciones diofnticas
es x0 = 2 27 = 54, y0 = 2 203 = 406. Las soluciones de la ecuacin homognea
2000 x 266 y = 0 son las soluciones de 1000 x 133 y = 0, que son x = n 133, y = n 1000.
Todas las soluciones de nuestro sistema de ecuaciones diofnticas son
x = 54 + n 133
y = 406 + n 1000
MANUALES UEX
1.1. Introduccin
Anillos de enteros
MANUALES UEX
p1
2
p1
2
p1
a2 + b2 = 2178
Tenemos que calcular los enteros de Gauss a + bi Z[ i ], tales que N (a + bi ) = (a + bi )(a
bi ) = a2 + b2 = 2178 = 2 32 112 . Observemos que 3, 11 = 3 mod 4, luego son primos en
Z[ i ] y han de dividir a a + bi , es decir, a + bi = 3 11 (a0 + b0 i ) y N (a0 + b0 i ) = 2. Por tanto,
{(a0 , b0 ) = (1, 1), (1, 1), (1, 1), (1, 1)} y
{(a, b) = (33, 33), (33, 33), (33, 33), (33, 33)}.
Anillos de enteros
1.1. Introduccin
a + bi (33 + 33 i ) {
c2 d 2
c2 + d 2
+ c22cd
i . Por tanto,
+d 2
c2 d 2
2 cd
33( c d )2
c2 d 2 + 2 cd
+
i
:
c,
d
Z}
=
{
+
33
i : c, d Z}
c2 + d 2 c2 + d 2
c2 + d 2
c2 + d 2
P
siendo una raz primitiva n-sima de la unidad y trabaj con los nmeros a i i ,
a i Z. Es decir, trabaj en el anillo (concepto general introducido ms tarde por Dedekind) de enteros Z[]. Argumentando sobre la factorizacin nica, prob que la descomposicin anterior no es posible, con x, y, z Z no nulos. Dirichlet le hizo observar
a Kummer el error (cometido tambin por Cauchy y Lam) de suponer que todos los
anillos de enteros
eran dominios de factorizacin nica. Consideremos por sencillez
p
el anillo
p Z[ 5],
p tenemos dos descomposiciones en factores irreducibles 6 = 2 3 =
(1 + 5) (1 5). Para restaurar la factorizacin nica Kummer
introdujo los np
meros ideales (no diopuna definicin general). Si bien 1 + 5 y 2 son irreducibles
2
observemos
de 2. Es como
si
un m.c.d. ideal de
p hubiese
p
p que (1 + 5) es mltiplo
p
p
2 y 1 + 5. En la extensin Z
2, 2]ptenemos la factorizap[ 2 5] ,pZ[(1 +p 5)/ p
cin nica por irreducibles 6 = 2 ((1 + 5)/ 2) (1 5)/ 2 (si bien ya estamos
en anillos de enteros que no son los de partida). Dedekind observ que lo que estaba
definiendo Kummer era el concepto de ideal (recordemos que en los dominios de ideales principales (a 1 , . . . , a n ) = ( m.c.d.(a 1 , . . . , a n )), el concepto de ideal primo y que haba
probado que en tales anillos (dominios de Dedekind) todo ideal es producto de ideales
primos. Hilbert (con las torres de Hilbert) prob que todo anillo de enteros se mete
en otro anillo mayor donde sus ideales se hacen principales.
Sea x n + c 1 x n1 + + c n Z[ x] un polinomio irreducible y sean 1 , . . . , n sus races.
Consideremos Z[1 ] C y la norma N : Z[1 ] N, donde dado z Q[1 ], N ( z) es el
determinante de la homotecia de factor b en Q[1 ]. Si z = a + b1 , entonces N ( b) =
Q
i (a + b i ).
Resolver la ecuacin diofntica a n + c 1 a n1 b + + c n b n = c, equivale a encontrar
los z = a b1 Z[1 ], tales que N ( z) = c. Advirtamos, que en general, Z[1 ] no es
un dominio de factorizacin nica, ni sus invertibles son simplemente las races de la
unidad incluidas en Z[1 ].
Por desgracia los anillos de la Teora de Nmeros y los anillos de funciones algebraicas de las curvas algebraicas no son dominios de factorizacin nica. Tampoco son
anillos localmente de ideales principales, si lo fuesen seran localmente dominios de
factorizacin nica, pero pueden incluirse en anillos un poco ms grandes que s lo
son. Este captulo trata pues de los problemas de la factorizacin nica en estos anillos
y cmo resolverlos.
MANUALES UEX
x n = z n yn = ( z 1 y) ( z n y),
1.2.
Anillos de enteros
MANUALES UEX
gr p( x).
Recurrentemente obtendr que
10
Anillos de enteros
I = ( p 1 , . . . , p r ) A[x] + I { A + Ax + + Ax m1 }
Ahora bien, I { A + Ax + + Ax m1 } es un A -mdulo finito generado ya que es submdulo de { A + Ax + + Ax m1 }, que es un A -mdulo noetheriano. En conclusin, si escribimos I { A + Ax + + Ax m1 } = q 1 , . . . , q s A , tenemos que I = ( p 1 , . . . , p r , q 1 , . . . , q s ).
7. Corolario : Si A es un anillo noetheriano entonces A [ x1 , . . . , xn ]/ I es un anillo noetheriano.
Demostracin. A [ x1 , . . . , xn ] = A [ x1 , . . . , xn1 ][ xn ] es noetheriano por el teorema de la
base de Hilbert y por induccin sobre n. Por tanto, el cociente A [ x1 , . . . , xn ]/ I es un
anillo noetheriano.
8. Definicin : Sea A un anillo ntegro. Un elemento propio (no nulo ni invertible) de
A se dice que es irreducible si no descompone en producto de dos elementos propios.
9. Proposicin : Un mdulo M es noetheriano si y slo si toda cadena creciente de
submdulos de M , M1 M2 M n estabiliza, es decir, para n >> 0, M n = M m ,
para todo m n.
Demostracin. Si M es noetheriano y M1 M2 M n una cadena creciente de
submdulos de M , consideremos el submdulo N := i M i = m 1 , . . . , m r . Para n >> 0,
m 1 , . . . , m r M n , luego M n N M n , es decir, N = M n y M n = M m , para todo m n.
Veamos el recproco. Sea N un submdulo, si N , 0 sea 0 , m 1 N y M1 := m 1 .
Si M1 , N , sea m 2 N \ M1 y M2 := m 1 , m 2 . As sucesivamente vamos construyendo
una cadena 0 M1 M2 M3 que por la propiedad exigida a M ha de ser finita.
,
MANUALES UEX
11
Anillos de enteros
11. Definicin : Se dice que A es un dominio de factorizacin nica si todo elemento propio de A es igual a producto de irreducibles de modo nico, salvo factores por
invertibles y orden.
Z, k[ x] y en general los anillos eucldeos son dominios de factorizacin nica.
13. Definicin : Un polinomio P ( x) A [ x] se dice primitivo cuando sus coeficientes no
admiten un divisor comn no invertible, es decir, si P ( x) = a Q ( x) con a A , entonces
a es invertible.
14. Lema : Sea A un dominio de factorizacin nica con cuerpo de fracciones . Sean
P ( x),Q ( x) A [ x] dos polinomios primitivos. Entonces,
1. P ( x) Q ( x) es primitivo.
2. Si existen a, b A tales que a P ( x) = b Q ( x), entonces b = a u, para cierto invertible u A . Por tanto, si P ( x) = ab Q ( x) en [ x], entonces ab = u A es un invertible
de A .
Demostracin. 1. Supongamos que P ( x) Q ( x) = a R ( x), con R ( x) A [ x] y a A no
invertible. Sea p A irreducible que divida a a. Haciendo cociente en A [ X ] por p A [ x],
tenemos que
MANUALES UEX
P ( x) Q ( x) = 0 ( A / pA )[ x]
lo cual es contradictorio, porque ( A / pA )[ x] es ntegro y P ( x) y Q ( x) son no nulos.
2. Sea p un elemento irreducible que divida a a. Haciendo cociente en A [ X ] por
p A [ x], tenemos que 0 = b Q ( x), luego b = 0 y p divide a b. Dividiendo a a y b a la
vez por p y repitiendo sucesivamente este proceso obtendremos que a divide a b, y por
simetra que b divide a a. Luego, b = a u, para cierto invertible u A .
15. Teorema : Sea A un dominio de factorizacin nica con cuerpo de fracciones .
Un polinomio no constante primitivo, P ( x) A [ x], es irreducible en A [ x] si y slo si es
irreducible en [ x].
Demostracin. Supongamos que P ( x) es irreducible en [ x]. Si P ( x) = P1 ( x) P2 ( x), con
P1 ( x), P2 ( x) A [ x], entonces como P ( x) es irreducible en [ x], uno de los dos polinomios
P1 ( x) o P2 ( x) ha de ser de grado cero, digamos P1 ( x) = a. Como P ( x) es primitivo P1 ( x) =
a A es invertible. En conclusin, P ( x), es irreducible en A [ x].
12
Anillos de enteros
P ( x) =
a
P 1 ( x) P 2 ( x)
b
a
b
= u A , luego P ( x) no es irre-
16. Teorema (Gauss): Si A es un dominio de factorizacin nica, entonces A [ x] tambin lo es.
Demostracin. Sea = A A \{0} el cuerpo de fracciones. Sea P ( x) A [ x] y escribamos
P ( x) = a Q ( x), con a A y Q ( x) A [ x] primitivo. Sea
e 1 ( x) Q
e r ( x)
Q ( x) = Q
Q ( x) =
con Q i ( x) =
ai e
bi Q i
b
Q 1 ( x) Q r ( x)
c
()
A [ x] primitivos.
en A [ x].
Unicidad: Si P ( x) = q 1 q l P1 ( x) P t ( x), entonces cada P i ( x) es irreducible en [ x]
por el teorema 1.2.15. Por tanto, los polinomios P i ( x) (una vez reordenados) difieren de
los Q i ( x) en invertibles de A . Tachando los trminos polinmicos comunes se obtiene,
salvo invertibles de A , la igualdad q 1 q l = p 1 p s , de donde salvo permutacin de
los factores es q i = p i (salvo invertibles de A ).
Por el teorema de Gauss, Z[ x1 , . . . , xn ] y k[ x1 , . . . , xn ] son dominios de factorizacin
nica.
17. Definicin : Se dice que un anillo es un dominio de ideales principales si es un
anillo ntegro y todos sus ideales son principales (es decir, generados por un elemento).
MANUALES UEX
13
Anillos de enteros
1.3.
es epiyectivo. Escribiremos B = A [1 , . . . , n ]. Ker A [ x1 , . . . , xn ] es un ideal, que estar generado por ciertos polinomios p 1 ( x1 , . . . , xn ), . . . , p r ( x1 , . . . , xn ). Por tanto,
B ' A [ x1 , . . . , xn ]/( p 1 ( x1 , . . . , xn ), . . . , p r ( x1 , . . . , xn ))
MANUALES UEX
(1 , . . . , n ) 7 ( x 1 1 , . . . , x n n )
3. Definicin : Diremos que Spec A es una curva ntegra afn si A es una k-lgebra
de tipo finito ntegra y de dimensin de Krull 1.
4. Ejemplos :
14
Anillos de enteros
1.4.
1. Teorema : Sea O un anillo local noetheriano de dimensin de Krull mayor que cero
y sea m el ideal maximal de O . Entonces, m = ( t) es un ideal principal si y slo si O es
un dominio de ideales principales.
Demostracin. Veamos la implicacin directa. Dado a O no nulo, si a no es invertible entonces a = t a 1 . Si a 1 no es invertible, entonces (a) (a 1 ), porque si son iguales
,
a 1 = b a y a = tba, luego (1 tb)a = 0 y como 1 tb es invertible a = 0 y llegamos a contradiccin. Si a 1 no es invertible de nuevo a 1 = t a 2 y a = t2 a 2 . Si a 2 no es invertible,
de nuevo, (a 1 ) (a 2 ), y seguimos el proceso. Por noetherianidad, este proceso termina
2. Corolario : Sea O un anillo local noetheriano de dimensin de Krull mayor que
cero. Entonces, O es un dominio de ideales principales si y slo si dimO /m m/m2 = 1.
MANUALES UEX
15
Anillos de enteros
f f ( x) m x /m2x .
6. Proposicin : Sea m x A un ideal maximal. Sea I = ( f 1 , . . . , f n ) A un ideal in x A / I el ideal de las clases de m x . Se cumple que
cluido en m x y sea m
x /m
2x = (m x /m2x )/ d x f 1 , . . . , d x f n
m
x = m x / I . Por tanto,
Demostracin. Observemos que m
x /m
2x = m x /( I + m2x ) = (m x /m2x )/ I = (m x /m2x )/ d x f 1 , . . . , d x f n
m
7. Ejemplo : Sea m := ( x1 1 , . . . , xn n ) k[ x1 , . . . , xn ] y p( x1 , . . . , xn ) k[ x1 , . . . , xn ].
P p
P
Entonces, p( x1 , . . . , xn ) = p() + i x ()( x i i ) + i j ( x i i )( x j ) h i j ( x). Por tanto,
i
d p( x1 , . . . , xn ) =
p
x1
() d x1 + +
p
xn
() d xn
MANUALES UEX
(,) /m
2(,) = (m(,) /m2(,) )/( d (,) p( x, y)),
m
(,) /m
2(,) = 1 si y slo si d (,) p( x, y) , 0.
dim m
Luego, O = ( k[ x, y]/( p( x, y)))(,) es un dominio de ideales principales si y slo si
p
(,) /m
2(,) = ( d (,) x), luego
d (,) p( x, y) , 0. Por ejemplo, si y (, ) , 0, entonces m
(,) O = ( x ).
m
8. Ejemplo : Spec C[ x, y]/( y2 x3 ) tiene un nico punto singular: el origen. En efecto,
0 = d (,) ( y2 x3 ) = 32 d (,) x + 2 d (,) y si y slo si (, ) = (0, 0).
9. Proposicin : Sea i : A , B un morfismo finito, x Spec A un punto cerrado y
i 1 ( x) = Spec(B/m x B) = { y1 , . . . , yr }. Si B/m x B es un anillo reducido (por ejemplo, cuando sea una A /m x -lgebra separable), entonces m x B yi = m yi B yi , para todo i ; si adems
m x A x es principal, entonces m yi B yi es principal.
y1 m
yr = 0. Por tanto, m y1 m yr m x B y
Demostracin. Como B/m x B es reducida, m
m yi B yi = m y1 m yr B yi mx B yi
Luego, m x B yi = m yi B yi .
16
Anillos de enteros
d
(5)
,
0,
luego
y
es
no
singular.
En
conclusin,
y
Spec
Z
[
y
p
m y = (2, 5 + 1), es el nico punto singular.
11. Ejemplo : Sea m = e2 i/m C una raz primitiva m-sima de la unidad. Veamos
que Z[m ] es no singular al localizar en todo x Spec Z[m ]. Supongamos m = p n , con p
n
primo. El polinomio mnimo anulador de p n , p n ( x), que divide a x p 1, es separable
mdulo todo primo q , p. Por tanto, si m y Z[m ], cumple que m y Z = ( q), tenemos
que m y Z[ p n ] y = ( q), para q , p. El nico punto singular posible de Spec Z[ p n ] =
Spec Z[ x]/( p n ( x)), es m y = ( p, x 1). Observemos que
n1
) = (x p
n1
) p1 + + x p
n1
+1
1.5.
MANUALES UEX
p n ( x) = p ( x p
Anillos de Dedekind
17
Anillos de enteros
a
a
( )n + a 1 ( )n1 + . . . + a n = 0
b
b
Multiplicando por b n tendremos que a n es mltiplo de b, lo que contradice que b es
primo con a. En conclusin, los nicos elementos de enteros sobre A son los de A .
Si consideramos el nodo C y2 x2 + x3 = 0, la curva C que se obtiene de despegar
las dos ramas y el morfismo natural C C pegar las dos ramas, resulta que este
morfismo finito fuera del nodo es isomorfismo, luego es birracional (es decir, quitando un nmero finito conveniente de puntos en C y en C es un isomorfismo). Parece
claro intuitivamente que entre curvas no singulares en todo punto, no existen ms
morfismos finitos birracionales que los isomorfismos.
2. Teorema : Sea O un anillo ntegro local noetheriano de dimensin de Krull 1. Las
siguientes condiciones son equivalentes:
1. O es dominio de ideales principales.
2. O es ntegramente cerrado en su cuerpo de fracciones .
Demostracin. 1. 2. O es un dominio de ideales principales, luego dominio de factorizacin nica y es ntegramente cerrado en su cuerpo de fracciones.
2. 1. Sea f un elemento no nulo del ideal maximal m de O . O / f O es un anillo local
de dimensin cero. Por tanto, el ideal maximal m en O / f O es nilpotente. Es decir, existe
un n N de modo que mn f O . Sea n N mnimo verificando mn f O . Sea g mn1 de
f
modo que g f O . Basta probar que m = g O , pues tendramos que m es un O -mdulo
principal y O un dominio de ideales principales. Se verifica que
MANUALES UEX
g
f
m , O , tendremos que
g
f
m m. Por tanto,
g
f
pues
1
f
mn O . Si
g
f
es entero sobre O , as
O . Contradiccin porque g f O .
3. Lema : Sea S A un sistema multiplicativo. Si
f
0 N M M0 0
es una sucesin exacta de A -mdulos, entonces
fS
gS
0 NS M S M S0 0
es una sucesin exacta de A S -mdulos.
18
es un endomorfismo de m, que ha de
g
f
Anillos de enteros
4. Proposicin :
M = 0.
A S = ( A )S
En particular, si A es ntegramente cerrado, entonces A S tambin.
Un anillo ntegro es ntegramente cerrado en su cuerpo de fracciones si y slo si es
localmente ntegramente cerrado.
Demostracin. A S ( A )S es un morfismo entero, luego ( A )S A S . Sea f A S . Existe
una relacin entera
con a i A y s i S
6. Definicin : Un anillo A se dice que es un dominio de Dedekind si es noetheriano
de dimensin 1 e ntegramente cerrado en su cuerpo de fracciones.
MANUALES UEX
f n + a 1 / s 1 f n1 + + a n / s n = 0
19
Anillos de enteros
1.6.
MANUALES UEX
gG
una
2. Teorema : Sea A un anillo de nmeros enteros de cuerpo de fracciones y
20
Anillos de enteros
k[ x1 , . . . , xn ] , A
coincide con el cierre entero de k[ x1 , . . . , xn ]
finito e inyectivo. El cierre entero de A en
MANUALES UEX
5. Teorema : Sea A el anillo de una curva afn ntegra (resp. un anillo de nmeros
una extensin finita de cuerpos y A
enteros). Sea el cuerpo de fracciones de A , ,
21
Anillos de enteros
p
6. Ejemplo
: Calculemos
la desingularizacin de Z[ n]. Si n = m2 n0 , entonces
p
p
p Z[ n] =
p
0
0
Z[ m n ] Z[ n ] y la desingularizacin de Z[ n] coincide con la de Z[ n0 ]. As
p
pues, podemos suponer que n carece de factores cuadrticos. Z[ n] = Z[ x]/( x2 n) y
x2 n es separable mdulo p, salvo para p = 2 y p divisor de n. Por tanto, los nicos
puntos singulares posibles son m y = ( p, x), con p divisor de n, y m y = (2, x + 1) cuando n es impar. Observemos, en el primer caso, que que (Z[ x]/( x2 n, x) y = Z/2Z, luego
m y (Z[ x]/( x2 n, x)) y = ( x) e y es no singular. Veamos que sucede cuando m y = (2, x + 1).
Observemos que
d y ( x2 n) = d y (( x + 1)2 2 ( x + 1) 2
n1
n1
) = (
)d y 2 = 0
2
2
1
si y slo si n
od 4. Por tanto, y es singular, si n = 1 mod 4.
2 es par, es decir, n = 1 m
Supongamos que esta es la situacin. Observemos que
x+1 2 x+1 n1
)
=0
2
2
4
p
p
n+1
Por tanto, 2 es entero sobre Z, luego sobre Z[ n]. Si A es el cierre entero de
p
p
p pn+1
n+1
Z[ n], entonces A contiene a Z[ n, 2 ] = Z[ 2 ]. Los nicos puntos singulares
(
de Z[
p
n+1
2 ],
1
estn sobre la fibra de (2). Ahora bien, el polinomio y2 y n
4 , que anula
n+1
n+1
2 ]
MANUALES UEX
1.7.
Dada una k-extensin finita de cuerpos K separable (o una k-lgebra finita separable), existe una k-extensin de Galois que la trivializa, es decir,
n
K k =
22
Anillos de enteros
1.8. Cuestionario
h a : K K, h a ( b) := a b
Por cambio, de cuerpo base k , , tenemos el endomorfismo -lineal
h a 1 : K k K k , ( h a 1)( b ) := a b = h a1 ( b )
Si la matriz de h a en una base { e i } es (a i j ) la matriz de h a 1 es (a i j ). Por lo tanto,
n
tr( h a ) = tr( h a 1). Ahora bien, recordemos que va el isomorfismo K k = ,
n
a1 = ( g 1 (a), . . . , g n (a)), luego la matriz de h a 1 = h a1 en la base estndar de
es la matriz diagonal de coeficientes g i (a). Por tanto,
X
tr ( h a ) = g i (a)
i
iT2 : K K := Homk (K, k), a 7 iT2 (a), donde iT2 (a)(a0 ) := T2 (a, a0 )
1.8.
Cuestionario
MANUALES UEX
es un isomorfismo k-lineal.
23
Anillos de enteros
5]
como
producto
de ideales
p
p
primos los ideales (2), (3), (1 + 5) y (1 5). Descomponer como producto de
ideales primos el ideal (6).
14. Calcular la desingularizacin de C[ x, y]/( y2 x2 + x3 ).
1.9.
Biografa de Dedekind
MANUALES UEX
DEDEKIND BIOGRAPHY
24
Richard Dedekinds father was a professor at the Collegium Carolinum in Brunswick. His mother was the daughter of a professor who also worked at the Collegium Carolinum. Richard
was the youngest of four children and never married. He was to
live with one of his sisters, who also remained unmarried, for
most of his adult life.
He attended school in Brunswick from the age of seven and at
this stage mathematics was not his main interest. The school,
Martino-Catharineum, was a good one and Dedekind studied
science, in particular physics and chemistry. However, physics
became less than satisfactory to Dedekind with what he considered an imprecise logical structure and his attention turned
towards mathematics.
The Collegium Carolinum was an educational institution between a high school
and a university and he entered it in 1848 at the age of 16. There he was to receive
a good understanding of basic mathematics studying differential and integral calculus, analytic geometry and the foundations of analysis. He entered the University of
Gttingen in the spring of 1850 with a solid grounding in mathematics.
Gttingen was a rather disappointing place to study mathematics at this time, and
it had not yet become the vigorous research centre that it turned into soon afterwards.
Mathematics was directed by M.A. Stern and G. Ulrich. Gauss also taught courses in
mathematics, but mostly at an elementary level. The physics department was directed
by Listing and Wilhelm Weber. The two departments combined to initiate a seminar
which Dedekind joined from its beginning. There he learnt number theory which was
the most advanced material he studied. His other courses covered material such as
the differential and integral calculus, of which he already had a good understanding.
The first course to really make Dedekind enthusiastic was, rather surprisingly, a course on experimental physics taught by Weber. More likely it was Weber who inspired
Dedekind rather than the topic of the course.
In the autumn term of 1850, Dedekind attended his first course given by Gauss. It
was a course on least squares:
... fifty years later Dedekind remembered the lectures as the most beautiful he had
ever heard, writing that he had followed Gauss with constantly increasing interest
and that he could not forget the experience.
Dedekind did his doctoral work in four semesters under Gausss supervision and
submitted a thesis on the theory of Eulerian integrals. He received his doctorate from
Gttingen in 1852 and he was to be the last pupil of Gauss. However he was not well
trained in advanced mathematics and fully realised the deficiencies in his mathematical education.
At this time Berlin was the place where courses were given on the latest mathematical developments but Dedekind had not been able to learn such material at
Gttingen. By this time Riemann was also at Gttingen and he too found that the
mathematical education was aimed at students who were intending to become secondary school teachers, not those with the very top abilities who would go on to research
careers. Dedekind therefore spent the two years following the award of his doctorate
learning the latest mathematical developments and working for his habilitation.
In 1854 both Riemann and Dedekind were awarded their habilitation degrees within a few weeks of each other. Dedekind was then qualified as a university teacher
and he began teaching at Gttingen giving courses on probability and geometry.
Gauss died in 1855 and Dirichlet was appointed to fill the vacant chair at Gttingen. This was an extremely important event for Dedekind who found working with
Dirichlet extremely profitable. He attended courses by Dirichlet on the theory of numbers, on potential theory, on definite integrals, and on partial differential equations.
Dedekind and Dirichlet soon became close friends and the relationship was in many
ways the making of Dedekind, whose mathematical interests took a new lease of life
with the discussions between the two. Bachmann, who was a student in Gttingen at
this time wrote:
... recalled in later years that he only knew Dedekind by sight because Dedekind
always arrived and left with Dirichlet and was completely eclipsed by him.
Dedekind wrote in a letter in July 1856:
What is most useful to me is the almost daily association with Dirichlet, with whom
I am for the first time beginning to learn properly; he is always completely amiable
towards me, and he tells me without beating about the bush what gaps I need to fill
and at the same time he gives me the instructions and the means to do it. I thank him
already for infinitely many things, and no doubt there will be many more.
Dedekind certainly still continued to learn mathematics at this time as a student
would by attending courses, such as those by Riemann on abelian functions and elliptic
functions. Around this time Dedekind studied the work of Galois and he was the first
MANUALES UEX
Anillos de enteros
25
MANUALES UEX
26
Anillos de enteros
to lecture on Galois theory when he taught a course on the topic at Gttingen during
this period.
While at Gttingen, Dedekind applied for J L Raabes chair at the Polytechnikum
in Zrich. Dirichlet supported his application writing that Dedekind was an exceptional pedagogue. In the spring of 1858 the Swiss councillor who made appointments
came to Gttingen and Dedekind was quickly chosen for the post. Dedekind was appointed to the Polytechnikum in Zrich and began teaching there in the autumn of
1858.
In fact it was while he was thinking how to teach differential and integral calculus,
the first time that he had taught the topic, that the idea of a Dedekind cut came
to him. He recounts that the idea came to him on 24 November 1858. His idea was
that every real number r divides the rational numbers into two subsets, namely those
greater than r and those less than r. Dedekinds brilliant idea was to represent the
real numbers by such divisions of the rationals.
Dedekind and Riemann travelled together to Berlin in September 1859 on the occasion of Riemanns election to the Berlin Academy of Sciences. In Berlin, Dedekind
met Weierstrass, Kummer, Borchardt and Kronecker.
The Collegium Carolinum in Brunswick had been upgraded to the Brunswick Polytechnikum by the 1860s, and Dedekind was appointed to the Polytechnikum in 1862.
With this appointment he returned to his home town and even to his old educational
establishment where his father had been one of the senior administrators for many
years. Dedekind remained there for the rest of his life, retiring on 1 April 1894. He
lived his life as a professor in Brunswick:
... in close association with his brother and sister, ignoring all possibilities of change or attainment of a higher sphere of activity. The small, familiar world in which he
lived completely satisfied his demands: in it his relatives completely replaced a wife
and children of his own and there he found sufficient leisure and freedom for scientific
work in basic mathematical research. He did not feel pressed to have a more marked
effect in the outside world: such confirmation of himself was unnecessary.
After he retired, Dedekind continued to teach the occasional course and remained
in good health in his long retirement. The only spell of bad health which Dedekind
had experienced was 10 years after he was appointed to the Brunswick Polytechnikum when he had a serious illness, shortly after the death of his father. However he
completely recovered and, as we mentioned, remained in good health.
Dedekind made a number of highly significant contributions to mathematics and
his work would change the style of mathematics into what is familiar to us today.
One remarkable piece of work was his redefinition of irrational numbers in terms of
Dedekind cuts which, as we mentioned above, first came to him as early as 1858. He
published this in Stetigkeit und Irrationale Zahlen in 1872. In it he wrote:
Now, in each case when there is a cut ( A 1 , A 2 ) which is not produced by any rational
number, then we create a new, irrational number a, which we regard as completely
defined by this cut; we will say that this number a corresponds to this cut, or that it
produces this cut.
As well as his analysis of the nature of number, his work on mathematical induction, including the definition of finite and infinite sets, and his work in number theory,
MANUALES UEX
Anillos de enteros
27
1.10. Problemas
Anillos de enteros
duced a new style of mathematics that been a major influence on mathematicians ever
since. As Edwards writes:
Dedekinds legacy ... consisted not only of important theorems, examples, and concepts, but a whole style of mathematics that has been an inspiration to each succeeding generation.
Many honours were given to Dedekind for his outstanding work, although he always remained extraordinarily modest regarding his own abilities and achievements.
He was elected to the Gttingen Academy (1862), the Berlin Academy (1880), the Academy of Rome, the Leopoldino-Carolina Naturae Curiosorum Academia, and the Acadmie des Sciences in Paris (1900). Honorary doctorates were awarded to him by the
universities of Kristiania (Oslo), Zurich and Brunswick.
Article by: J.J. OConnor and E.F. Robertson (http://www-history.mcs.st-and.ac.uk/
Biographies/).
1.10.
Problemas
1. Probar que todo anillo de Dedekind que tenga slo un nmero finito de ideales
primos es un anillo de ideales principales.
2. Sea A un dominio de Dedekind e 0 , I A un ideal. Probar que A / I es un anillo
de ideales principales.
3. Probar que en un anillo de Dedekind todos los ideales estn generados por dos
elementos.
4. Sea A = C[ x1 , . . . , xn ]/( p 1 , . . . , p n1 ) y Specmax A . probar que A es un dominio
p
de ideales principales si y slo si ( x i ()) i j es una matriz de rango n 1.
j
p
p
5. Probar que Z[ 5] es un anillo de Dedekind. Sea m = (2, 1 + 5). Probar que
m2 = (2) pero que m no es principal.
p
p
6. Descomponer 33 + 11 7 en producto de elementos irreducibles de Q[ 7].
MANUALES UEX
7. Es
p
3+2p 6
1 6
entero sobre Z?
8. Sea p( x, y) = 0 una curva ntegra tal que el origen es un punto singular de multiplicidad k (es decir, p( x, y) ( x, y)k \( x, y)k+1 ). Supongamos que p(0, y) = yk q( y),
con q(0) , 0 (es decir, x = 0 corta transversalmente a la curva en el origen). Probar que y/ x es entero sobre el anillo (C[ x, y]/( p( x, y))) q(y) , y que
(C[ x, y]/( p( x, y)))[ y/ x] = C[ x, y/ x]/( r ( x, y/ x)),
donde x k r ( x, y/ x) = p( x, y).
9. Desingularizar la curva y2 y3 + x4 = 0 en un entorno del origen.
28
Captulo 2
Fibras de los morfismos finitos
2.1.
Introduccin
2.2.
MANUALES UEX
Longitud de un mdulo
29
es muy prctico. En los A -mdulos libres se define el rango del A -mdulo libre como el
nmero de elementos de sus bases.
Si intuimos que R3 es de dimensin 3 es porque observamos la cadena de inclusiones irrefinable: punto, recta, plano, espacio. Puede definirse la dimensin de un espacio
vectorial, como la longitud de las cadenas irrefinables de subespacios vectoriales. En
los A -mdulos pueden no existir bases, pero si podemos hablar de la longitud de las
cadenas irrefinables de submdulos de un mdulo. En trminos de stas definiremos
la longitud del mdulo, concepto que no coincide con el de rango, en general.
1. Definicin : Diremos que un A -mdulo M , 0 es simple cuando sus nicos submdulos son los triviales: 0 y M .
Si M es un A -mdulo simple entonces M = m, luego M ' A / Anul m. Ahora bien,
los submdulos de A / Anul m se corresponden con los ideales de A que contienen a
Anul m. Por tanto, M es simple si y slo si Anul m es un ideal maximal, es decir, M
es simple si y slo si M ' A /m, donde m es un ideal maximal de A .
2. Definicin : Diremos que una cadena finita de submdulos 0 = M0 M1
M n = M es una serie de composicin en M , si los cocientes sucesivos M i / M i1 son
A -mdulos simples. Diremos que la longitud de esta serie de composicin es n.
Como los submdulos de M i / M i1 se corresponden biyectivamente con los submdulos de M i que contienen a M i1 , el que M i / M i1 sea simple equivale a que
no existe una cadena M i1 N M i . Por tanto, que una cadena de submdulos
,
MANUALES UEX
30
i =1
n
X
i =1
2.3.
A A x1 A xn , a 7 (a/1, . . . , a/1)
MANUALES UEX
es un isomorfismo.
Demostracin. Si x i , x j , ( A x i ) x j = 0, porque Spec( A x i ) x j = ;, pues es igual al conjunto
de los ideales primos de A contenido en m x i y m x j . Obviamente, ( A x i ) x i = A x i . Por
n
31
2. Definicin : Sea A una k-lgebra finita, sea Y = Spec A , que es un nmero finito
de puntos cerrados, y sea y Y .
1. Llamaremos nmero de puntos de Y contando grados y multiplicidades a dimk A .
2. Llamaremos multiplicidad con la que aparece y en Y a m y (Y ) := l A ( A y ).
3. Llamaremos grado de y a gr y := dimk A /m y .
3. Proposicin : Se cumple que
Nmero de puntos de Y contando grados y multiplicidades =
m y (Y ) gr y
yY
Demostracin. En efecto, A =
l A ( A y ). Con todo, se concluye.
yY
A y , luego dimk A =
yY
dimk A y = dimk ( A /m y )
2.4.
MANUALES UEX
el morfismo inducido. Dado x Spec A , 1 ( x) = Spec B/m x B y B/m x B es una A /m x lgebra finita. Llamaremos nmero de puntos de 1 ( x) a dim A/mx B/m x B, multiplicidad con la que aparece y 1 ( x) en 1 ( x) a m y := l B ((B/m x B) y ), y grado de y sobre x
a gr x y := dim A/mx B/m y . Por tanto,
Nmero de puntos de 1 ( x) contando grados y multiplicidades =
m y gr x y
y1 (x)
32
B = nAx
n
3. Definicin : Sea : A B un morfismo finito entre dominios de Dedekind. Sea m y
e
un ideal maximal de B y m x := m y B. Entonces m x B y = m yy B y , para cierto e y N, que
llamaremos ndice de ramificacin de y.
4. Proposicin : Sea : A B un morfismo finito entre dominios de Dedekind. Sea
x Spec A un ideal maximal e y un punto en la fibra de x. La multiplicidad con la que
aparece y en la fibra de x es igual al ndice de ramificacin de y.
Demostracin. Escribamos m y B y = ( t). Entonces,
e
l B (B/m x B) y = l B (B y /m yy B y ) = l B (B y /( t e y )) = e y l B (B y /( t)) = e y ,
Sea A , B un morfismo finito inyectivo y : Spec B Spec A el morfismo inducido.
MANUALES UEX
5. Definicin : Se dice que no ramifica en y si B y /m x B y es una A /m x -lgebra separable. Si ramifica en y se dice que y es un punto de ramificacin de y que ( y) es
un punto rama de .
33
7. Teorema : Sea A , B un morfismo finito inyectivo entre anillos noetherianos ntegros de dimensin de Krull 1. Sean A y B los cuerpos de fracciones de A y B respectivamente. Supongamos que A , B es una extensin separable de cuerpos. Entonces,
el morfismo Spec B Spec A ramifica en un nmero finito de puntos.
Demostracin. Escribamos B = A [1 , . . . , n ], donde los i son enteros y separables
sobre A . Los puntos rama del morfismo A A [ i ] son un nmero finito. Luego los
puntos rama del morfismo A A [1 ] A A A [n ] son un nmero finito. Como B es
un cociente de A [1 ] A A A [n ], los puntos rama de A B son un nmero finito.
Luego Spec B Spec A ramifica en un nmero finito de puntos.
p
8. Ejercicio : Consideremos el morfismo finito Z Z[ 3]. Calcular el nmero de puntos, contando grados y multiplicidades, de las fibras de este morfismo, los puntos de
ramificacin e ndices de ramificacin.
Sea G un grupo finito de automorfismos de un anillo B. Dado g G , el automorfismo g : B B, induce el automorfismo g : Spec B Spec B. G opera sobre Spec B de
modo natural: dado g G y y Spec B, g y := g1 ( y), es decir, p g y := g(p y ).
9. Teorema : Sea G un grupo finito de automorfismos de un anillo B. Denotemos por
BG := { b B : g( b) = b para todo g G } y por (Spec B)/G := { y , con y Spec B, donde
decimos que y = z si y slo si existe g G tal que z = g y}. Se cumple que
Spec(BG ) = (Spec B)/G
donde BG = { b B : g( b) = b, para todo g G }.
MANUALES UEX
Q
gG
( x g( f )) es
10. Ejercicio : Sea : C[ x, y]/( y2 x) C[ x, y]/( y2 x) el automorfismo de C-lgebras
definido por ( y) = y y ( x) = x y sea G = {Id, }. Explicitar la operacin de G sobre
X = Spec C[ x, y]/( y2 x). Calcular X /G .
p
p
p
el automorfismo de anillos definido por ( 2) =
11.
pEjercicio : Sea : Z[ 2] Z[ 2]p
2 y sea G = {Id, }. Sea X = Spec Z[ 2]. Calcular X /G .
34
13. Ejercicio : Calcular el grupo de descomposicin de x X ,cuando X es el definido
en el ejercicio 2.4.10 o 2.4.11.
2.5.
Automorfismo de Frbenius
1. Teorema : Sea A un anillo de enteros tal que su cuerpo de fracciones A sea una
Q-extensin de Galois de grupo G y tal que G A = A . Sea m y A un ideal maximal
y sea ( p) = m y Z. El automorfismo de Frbenius, F , de A /m y est inducido por algn
automorfismo F p G de A , y ste es nico cuando A / pA es reducida (es decir, el morfismo Spec A Spec Z no ramifica en y), en este caso se dice que F p es el automorfismo
de Frbenius de A en el primo p.
MANUALES UEX
35
2.6. Aplicaciones
2. Observaciones :
1. F p : A A , es el automorfismo de A que deja estable m y ,
determinado por la condicin F p (a) = a p mod m y , para todo a A .
2. En el teorema, en la fibra de ( p), si en vez de tomar y consideramos otro punto y0 ,
entonces como G opera transitivamente en las fibras, existe g G de modo que
y0 = g y. Por tanto, el grupo de descomposicin de y0 es gD g1 y el automorfismo
que asociaramos a F sera gF p g1 .
3. Si A / pA es reducida y yi Spec( A / pA ), entonces ( A / pA ) yi = A /m yi . Por tanto,
m yi A yi = p A yi . Es decir, todos los puntos de la fibra del ideal primo ( p) son no
singulares. Si A es el cierre entero de A en A , entonces A yi = A yi , A / pA = A / p A
y el automorfismo de Frbenius de A en p no depende del anillo A considerado.
MANUALES UEX
2.6.
Aplicaciones
36
2.6. Aplicaciones
fija alguna raz de q( x), entonces F ( i ) = i , para algn i . Por tanto, q( x) tiene
alguna raz en Z/ pZ.
p
Considerando K = Q[ i, 2], vemos que el polinomio ( x2 + 1)( x2 2)( x2 + 2) tiene
raz modular en todo primo p, aunque carece de races racionales.
3. El grupo de Galois, G , de la extensin ciclotmica n-sima, Q[ e
2 p i
n
2 i
n
2 i
n
] es (Z/ nZ) :
)=e
2 p i
n
, luego
n
1, si n es un resto cuadrtico mdulo q ( n = a2 , para cierto a F q )
( ) :=
1, en otro caso
q
As pues, ( nq ) = n
q1
2
Demos un algoritmo para un clculo rpido de ( nq ). Descomponiendo n en producto de primos podemos suponer que n = p es primo. El grupo de Galois G
de Q[ e2 i/q ] es isomorfo a Fq , que es cclico. Observemos que va el isomorfismo G ' Fq , F p se aplica en p . Tenemos que ver cundo p (Fq )2 . (Fq )2 es el
nico subgrupo de ndice 2. La nica subextensin de grado dos de Q[ e2 i/q ], es
p
q1
p
2
K := Q[ e2 i/q ](F q ) = Q[ ] = Q[ q ], donde q = (1) 2 q. Luego, p (Fq )2 si y
slo si F p es la identidad sobre K .
Por otra parte, si p , 2, x2 q mdulo p es separable. As pues, el automorfismo
de Frbenius de K en p es la identidad cuando q Fp2 . Por tanto,
MANUALES UEX
p1 q1
p
q
1 q1 q
q
( ) = ( ) = ( ) 2 ( ) = (1) 2 2 ( )
q
p
p
p
p
p
Supongamos p = 2. Desgraciadamente Z , Z[p q ] ramifica en (2), pero si consip
q +1
deramos el cierre entero de Z[ q ], que es Z[ 2 ] ya no ramifica. El polinomio
37
2.7. Cuestionario
q +1
q 1
1. Ejercicio : Probar que el grupo de Galois de x3 + 2 x2 + 4 x + 1 es igual a S 3 , argumentando con los morfismos de Frbenius en 2 y 3.
2. Ejercicio : Si un polinomio con coeficientes enteros mnico es irreducible mdulo
un nmero primo entonces, es irreducible? Es x4 + 2 x2 + x + 1 irreducible mdulo 2?
Es x4 + 2 x2 + x + 1 Q[ x] irreducible? Probar que el grupo de Galois de x4 + 2 x2 + x + 1
es igual a S 4 , argumentando con los morfismos de Frbenius en 2 y 5.
2.7.
Cuestionario
MANUALES UEX
38
2.8.
Biografa de Frbenius
FRBENIUS BIOGRAPHY
MANUALES UEX
39
MANUALES UEX
40
dards at the University of Berlin, in the words of Frbenius, to the rank of a technical
school ... Even so, Fuchs and Schwarz yielded to him, and later Schottky, who was indebted to him alone for his call to Berlin. Frbenius was the leading figure, on whom
the fortunes of mathematics at Berlin university rested for 25 years. Of course, it did
not escape him, that the number of doctorates, habilitations, and docents slowly but
surely fell off, although the number of students increased considerably. That he could
not prevent this, that he could not reach his goal of maintaining unchanged the times
of Weierstrass, Kummer and Kronecker also in their external appearances, but to witness helplessly these developments, was doubly intolerable for him, with his choleric
disposition.
We should not be too hard on Frbenius for, as Haubrich explained:
They all felt deeply obliged to carry on the Prussian neo-humanistic tradition of
university research and teaching as they themselves had experienced it as students.
This is especially true of Frbenius. He considered himself to be a scholar whose duty
it was to contribute to the knowledge of pure mathematics. Applied mathematics, in
his opinion, belonged to the technical colleges.
The view of mathematics at the University of Gttingen was, however, very different. This was a time when there was competition between mathematians in the
University of Berlin and in the University of Gttingen, but it was a competition that
Gttingen won, for there mathematics flourished under Klein, much to Frbeniuss
annoyance. Biermann wrote:
The aversion of Frbenius to Klein and S. Lie knew no limits ...
Frbenius hated the style of mathematics which Gttingen represented. It was a
new approach which represented a marked change from the traditional style of German universities. Frbenius, as we said above, had extremely traditional views. In a
letter to Hurwitz, who was a product of the Gttingen system, he wrote on 3 February
1896:
If you were emerging from a school, in which one amuses oneself more with rosy
images than hard ideas, and if, to my joy, you are also gradually becoming emancipated
from that, then old loves dont rust. Please take this joke facetiously.
One should put the other side of the picture, however, for Siegel, who knew Frbenius for two years from 1915 when he became a student until Frbeniuss death,
related his impression of Frbenius as having a warm personality and expresses his
appreciation of his fast-paced varied and deep lectures. Others would describe his lectures as solid but not stimulating.
To gain an impression of the quality of Frbeniuss work before the time of his
appointment to Berlin in 1892 we can do no better than to examine the recommendations of Weierstrass and Fuchs when Frbenius was elected to the Prussian Academy
of Sciences in 1892. We quote a short extract to show the power, variety and high quality of Frbeniuss work in his Zrich years. Weierstrass and Fuchs listed 15 topics on
which Frbenius had made major contributions:
-On the development of analytic functions in series.
-On the algebraic solution of equations, whose coefficients are rational functions of
one variable.
-The theory of linear differential equations.
MANUALES UEX
41
Theorem. It was a burst of activity which set up the foundations of the whole of the
machinery of representation theory.
In a letter to Dedekind on 26 April 1896 Frbenius gave the irreducible characters for the alternating groups A 4 , A 5 , the symmetric groups S 4 , S 5 and the group
PSL(2, 7) of order 168. He completely determined the characters of symmetric groups
in 1900 and of characters of alternating groups in 1901, publishing definitive papers
on each. He continued his applications of character theory in papers of 1900 and 1901
which studied the structure of Frbenius groups.
Only in 1897 did Frbenius learn of Moliens work which he described in a letter to
Dedekind as very beautiful but difficult. He reformulated Moliens work in terms of
matrices and then showed that his characters are the traces of the irreducible representations. This work was published in 1897. Frbeniuss character theory was used
with great effect by Burnside and was beautifully written up in Burnsides 1911 edition of his Theory of Groups of Finite Order.
MANUALES UEX
42
2.9.
2.9. Problemas
Problemas
MANUALES UEX
43
MANUALES UEX
2.9. Problemas
44
Captulo 3
Anillos de valoracin discreta
3.1.
1. Definicin : Sea un cuerpo y = \{0}. Una valoracin real de es una aplicacin v : R que verifica
1. v( f g) = v( f ) + v( g), para todo f , g .
2. v( f + g) mn{v( f ), v( g)}, para todo f , g .
Si Im v = {0} se dice que v es trivial. Si Im v = Z, se dice que v es una valoracin
discreta. Seguiremos la convencin v(0) = .
Observemos que v(1) = v(1 1) = v(1) + v(1), luego v(1) = 0. Por tanto, 0 = v(1) =
v( f f 1 ) = v( f ) + v( f 1 ), luego v( f 1 ) = v( f ).
Sea O un anillo local de ideal maximal m y cuerpo de fracciones . Supongamos
que O cumple que si f mn \mn+1 y g mm \mm+1 entonces f g mn+m \mn+m+1 . Para
cada f O no nula, denotemos vm ( f ) al mximo nmero natural n tal que f mn . Es
fcil ver que la aplicacin
vm : Z
est bien definida y es una valoracin discreta de . Esta valoracin se denomina
valoracin m-dica.
Si O es de ideales principales, es inmediato ver que O = { f | vm ( f ) 0}. Veamos
el recproco.
2. Proposicin : Sea v una valoracin real de un cuerpo y denotemos O v = { f
: v( f ) 0}. Entonces O v es un anillo local de ideal maximal pv = { f : v( f ) > 0}, de
cuerpo de fracciones . Si v es discreta entonces O v es un domimio de ideales principales
de dimensin de Krull 1, y v = vpv .
MANUALES UEX
f / g 7 vm ( f / g) = vm ( f ) vm ( g)
45
Conjunto de valoraciones
discretas de
v 7 O v
3. Ejercicio : Sea v una valoracin discreta y sea f O v . Probar que v( f ) = l (O v /( f )).
4. Definicin : Dada una valoracin v : R diremos que O v es un anillo de valoracin de . Si v es discreta diremos que O v es un anillo de valoracin discreta de
.
Por la proposicin anterior, un anillo es un anillo de valoracin discreta de si y
slo si es un subanillo propio local de ideales principales de , de cuerpo de fracciones
.
La valoracin v : R, v( ) = 0 se dice que es trivial. O v = si y slo si v es
trivial. Se dice que es el anillo de valoracin trivial.
5. Proposicin : Los anillos de valoracin son ntegramente cerrados en su cuerpo de
fracciones.
Demostracin. Sea O v un anillo de valoracin de y a entero sobre O v . Existen
c i O v tales que a n + c 1 a n1 + + c n = 0. Entonces a n = ( c 1 a n1 + + c n ), luego nv(a) =
v(a n ) = nf{v( c 1 a n1 ), . . . , v( c n )} nf{( n 1)v(a), . . . , 0}, luego v(a) 0 y a O v .
6. Lema : Sea A un anillo ntegro (luego A est incluido en su cuerpo de fracciones
y al localizar por un sistema multiplicativo tambin). Entonces,
MANUALES UEX
A=
Demostracin. Sea
a
b
xSpecmax A
xSpecmax A
Ax
a
b
A.
46
Z
v:
v( A )0
MANUALES UEX
A=
Ov
v : K Z
47
3.1.1.
Variedad de Riemann
MANUALES UEX
que x y x0 estn ambos en uno de los abiertos afines Uh ,Uh ,Uh + , luego O C 0 ,x , O C 0 ,x0 .
i
48
v x ( f ) grk x
MANUALES UEX
xC,v x ( f )0
xC
49
3.2.
MANUALES UEX
Luego, si dos valores absolutos definen la misma topologa sus respectivas bolas unidad son iguales.
Fijemos un punto x con | x| > 1, es decir, 1/ x B(0, 1). Dado y, tendremos que | y| =
Por tanto, si | |0 es equivalente a | |, tenemos que | y|0 = | x|0 . Si definimos r := log| x| | x|0 ,
| y|0 = | x|0 = (| x|r ) = | y|r , para todo y.
5. Definicin : Un valor absoluto | | : A R se dice arquimediano si la imagen de la
aplicacin natural N R, n 7 | n| no est acotada, es decir, para toda constante C > 0
existe un natural n tal que | n| > C .
Evidentemente, todo cuerpo dotado de un valor absoluto arquimediano debe ser de
caracterstica cero.
50
n = a0 + a1 d + . . . + a k d k ,
con 0 a i < d
De donde
| n| d + d | d | + . . . + d | d |k d (1 + k) d (1 + logd n)
Por tanto,
| n k | d (1 + k logd n)
Entonces,
1 lm
d (1 + k logd n)
| n| k
=0
7. Primer teorema de Ostrowski, 1917 : Todo valor absoluto arquimediano sobre
Q es equivalente al valor absoluto usual.
Demostracin. Por el lema |2| > 1. Sustituyendo | | por | |r , con r > 0 conveniente, podemos suponer que |2| = 2. Entonces, |3| |2| + |1| = 3 y 4 = |2| |2| = |4| |3| + 1, luego
|3| = 3. Entonces, |5| |4| + |1| = 5 y 6 = |2| |3| = |6| |5| + 1, luego |5| = 5. As sucesivamente, obtenemos que | | es el valor absoluto usual sobre N, luego lo es sobre Q.
MANUALES UEX
3. || e|| = || || e||.
9. Ejemplo : Si E es un K -espacio vectorial con una base finita { e 1 , . . . , e n }, se define
la norma infinita como sigue:
X
{|1 |, . . . , |n |}.
|| i e i || := max
i
51
10. Proposicin : Si F es un subespacio vectorial cerrado de un espacio vectorial normado (E, || ||), entonces
|| e|| := nf{|| e0 || : e0 e + F }
es una norma sobre E /F, y la proyeccin natural E E /F es continua.
11. Proposicin : Sean (K, | |) un cuerpo completo y E un K -espacio vectorial de dimensin finita. Todas las normas sobre E son topolgicamente equivalentes y completas.
Demostracin. Es rutinario comprobar que E es completo para la norma infinita || ||, y
por tanto tambin es completo para cualquier otra norma topolgicamente equivalente
a la norma infinita.
Ya sabemos que cualquier norma || ||0 sobre E es menos fina que la norma infinita. Para la afirmacin inversa procedamos por induccin sobre n = dimK E . Por hiptesis de induccin, todo subespacio de E de dimensin menor que n es completo
para la norma || ||0 luego tambin es cerrado. Por tanto, las proyecciones j : E K e j ,
P
j ( i i e i ) := j e j , son continuas tomando en E la norma || ||0 y en K e j la norma cociente (que equivale, como todas, a la norma infinita). Por tanto, la aplicacin identidad
jj
MANUALES UEX
52
morfismos K C
md. conjugacin
= Spec(K Q R)
3.3.
de donde
|a + b| (1 + n)1/n | b|,
a
valores absolutos sobre Q,
= Spec Z {| | }
mdulo equivalencia
7. Corolario: Sea K un un cuerpo de nmeros y A el anillo de enteros de K . Entonces,
= Spec A
MANUALES UEX
Spec(K Q R)
53
3.4.
gr x
| f |x
=1
3. Definicin : Sea B una k-lgebra finita separable. Dada b B consideremos el kendomorfismo lineal b : B B, b0 7 bb0 . Se define la norma de b, que denotamos N ( b),
como N ( b) := det( b).
Obviamente, N (1) = 1 y N ( bb0 ) = N ( b) N ( b0 ). Sea una k-extensin que trivialice
a B y {1 , . . . , n } = Homkal g (B, ). Entonces, argumentando como hacamos con la
traza,
Q
N ( b) = i i ( b)
Si K es un cuerpo de nmeros, entonces es una Q-lgebra finita separable y tenemos la norma N : K Q.
MANUALES UEX
54
Demostracin. Existen sendas bases de los Z-mdulos A y A en las que el endomorfismo a : A A diagonaliza. El determinante de la matriz de a en estas bases es igual
salvo signos a | A /aA |, y es igual, salvo signos al determinante del endomorfismo a,
con lo que concluimos.
Sea K un cuerpo de nmeros y A el anillo de enteros de K . Denotemos X = Spec A
el conjunto de valores absolutos no arquimedianos de K (mdulo equivalencia), X :=
Spec(K Q R) el conjunto de valores absolutos arquimedianos de K (mdulo equivalen`
cia), y X = X X el conjunto de valores absolutos de K (mdulo equivalencia).
Sea | | el valor absoluto usual de C. Dado y X , sea | | y el valor absoluto arquimediano de K asociado a y definido por | f | y = | f ( y)|, donde f ( y) es igual a la clase de f en
(K Q R)/p y . Dicho de otro modo, si y se corresponde con : K C, entonces f ( y) = ( f )
y | f | y = |( f )|. Dado y X , denotemos gr y := dimR (K Q R)/m y .
Dado x X = Spec A , sea | | x el valor absoluto no arquimediano asociado a x de gr x
finido por |a| x := evx (a) . Recordemos que gr x := ln | A /m x |. Observemos que |a| x
=
v x (a)
| A /m x |
.
x X
gr x
| f |x
=1
mvxx (a) | =
| A /m x |vx (a) =
gr x
| a| x
x X
Q
Q
gr y
Por otra parte, | N (a)| = HomQal g (K,C) |(a)| = y X |a| y .
Q
gr x
Luego, x X |a| x = 1.
x X
x X
6. Ejercicio : Comprobar la frmula del teorema 3.4.5, para K = Q[ i ] y f = i + 1.
Denotemos los invertibles de A , A .
7. Proposicin : A = {a A : N (a) = 1}.
Demostracin. Sea a A . | N (a)| = | A /(a)| = 1 si y slo si a A .
Pn
8. Observacin : Dado a A sea p c ( x) = i=0 a i x n i el polinomio caracterstico de la
homotecia a : A A . Sabemos que N (a) = (1)n a n y por otra parte 0 = p(a) = b a + a n ,
con b A . En conclusin, N (a) = a c, con c A .
9. Proposicin : Sea c N y d = dimQ K . Consideremos la accin natural por multiplicacin de A en { f A : | N ( f )| = c}, entonces
|{ f A : | N ( f )| = c}/ A | c d
{ f A : | N ( f )| = c}/ A ( A / cA )/ A , f
7 f
3.5.
MANUALES UEX
55
I n = I Rn.
4. Ejercicio : Probar que un ideal homogneo p R es primo si y slo si cumple que
si el producto de dos elementos homogneos pertenece a p entonces uno de los dos
pertenece a p.
5. Definicin : Llamaremos ideal irrelevante de R al ideal ( R n ) R .
n ,0
6. Definicin : Llamaremos espectro proyectivo de R , y lo denotaremos Proj R , al conjunto de ideales primos homogneos de R que no contienen al ideal irrelevante.
Evidentemente Proj R Spec R . Consideraremos Proj R como espacio topolgico
con la topologa inicial heredada de la topologa de Zariski de Spec R . Si denotamos ( f )0h = { x Proj R, f p x } y escribimos f = f n + f n+1 + f m , es obvio que ( f )0h =
( f n , . . . , f m )0h = ( f n )0h ( f m )0h . Por tanto, una base de cerrados de la topologa de
Proj R son los cerrados ( f )0h , con f R homogneo, y una base de abiertos de la topologa de Proj R son los abiertos
U fh = { x Proj R, f p x }, ( f homogneo)
7. Definicin : Llamaremos espacio proyectivo de dimensin n (sobre k) a
Pnk = Proj k[ x0 , . . . , xn ]
MANUALES UEX
56
2. Si I es un ideal homogneo de R entonces R / I es un lgebra graduada homognea, de modo que el morfismo R R / I es un morfismo graduado que induce un
isomorfismo
Proj(R / I ) = ( I )0h
Dada un lgebra graduada R denotaremos por R n al conjunto de los elementos
homogneos de grado n de R . Observemos que R 0 es un subanillo de R .
11. Proposicin : Sea R un lgebra graduada y f R un elemento homogneo de
grado 1. Entonces,
U fh = Proj R f = Spec[R f ]0
Demostracin. Veamos que la composicin de los morfismos naturales
Proj R f , Spec R f Spec[R f ]0 ,
que asigna a cada ideal primo homogneo p R f el ideal primo [p]0 := p [R f ]0 , es el
homeomorfismo buscado. Observemos que el ideal primo p R f est determinado por
sus elementos homogneos de grado cero: un elemento homogneo g R f de grado m
pertenece a p si y slo si g/ f m pertenece a [p]0 . Por tanto, Proj R f Spec[R f ]0 es inyectivo. Observemos que R f = nZ [R f ]0 f n . Si q [R f ]0 es un ideal primo, entonces el
ideal homogneo p := nZ q f n R f es un ideal primo homogneo: Si g, g0 R f son dos
elementos homogneos de grados m y m0 respectivamente, tales que g g0 p, entonces
0
0
0
( g/ f m ) ( g0 / f m ) = ( g g0 )/ f m+m q, luego g/ f m g0 / f m pertenece a q, y por tanto g g0
pertenece a p. Observemos p [R f ]0 = q. En conclusin, Proj R f Spec[R f ]0 es biyectivo. Finalmente, si g R es homognea de grado m, la biyeccin anterior transforma
( g)0h = ( g/ f m )0h en ( g/ f m )0 . Luego la biyeccin continua dada es un homeomorfismo.
Por sencillez, supondremos a partir de ahora que R = R 0 [0 , . . . , n ], donde cada i
es de grado 1. En este caso, [R 0 [0 , . . . , n ] i ]0 = R 0 [0 / i , . . . , n / i ] donde R 0 [0 / i , . . . , n / i ]
es la R 0 -sublgebra de R 0 [0 , . . . , n ] i generada por 0 / i , . . . , n / i .
1. Proj R = U i .
i =0
2. U i es homeomorfo a Spec R 0 [ 0 , . . . , n ].
i
Diremos que U i es un abierto afn de Proj R . Por tanto, el espectro proyectivo admite
un recubrimiento por abiertos afines.
n
i =0
i =0
MANUALES UEX
57
3.6. Cuestionario
en cada uno de los abiertos afines, complementario del cerrado ( x i )0h (deshomogeneizar x02 + x12 + x22 por cada variable x i ).
2. Demostrar que el epimorfismo C[ x0 , x1 , x2 ] C[ x0 , x1 , x2 ]/( x02 + x12 + x22 ) define una
inmersin cerrada Proj C[ x0 , x1 , x2 ]/( x02 + x12 + x22 ) , P2
3. Definir una curva proyectiva plana que en uno de los abiertos afines sea la curva
plana afn y + x2 = 0. Corta la recta x = 0, a la curva y + x2 = 0, en algn punto
del infinito?
3.6.
Cuestionario
MANUALES UEX
58
p
3
12. Calcular los valores absolutos arquimedianos de Q( e2 i/5 ) y Q( 2).
3.7.
Biografa de Riemann
MANUALES UEX
RIEMANN BIOGRAPHY
59
MANUALES UEX
60
would avoid long computations as much as possible. His manner suited Riemann, who
adopted it and worked according to Dirichlets methods.
Riemanns work always was based on intuitive reasoning which fell a little below
the rigour required to make the conclusions watertight. However, the brilliant ideas
which his works contain are so much clearer because his work is not overly filled with
lengthy computations. It was during his time at the University of Berlin that Riemann
worked out his general theory of complex variables that formed the basis of some of
his most important work.
In 1849 he returned to Gttingen and his Ph.D. thesis, supervised by Gauss, was
submitted in 1851. However it was not only Gauss who strongly influenced Riemann
at this time. Weber had returned to a chair of physics at Gttingen from Leipzig during
the time that Riemann was in Berlin, and Riemann was his assistant for 18 months.
Also Listing had been appointed as a professor of physics in Gttingen in 1849. Through Weber and Listing, Riemann gained a strong background in theoretical physics
and, from Listing, important ideas in topology which were to influence his ground
breaking research.
Riemanns thesis studied the theory of complex variables and, in particular, what
we now call Riemann surfaces. It therefore introduced topological methods into complex function theory. The work builds on Cauchys foundations of the theory of complex
variables built up over many years and also on Puiseuxs ideas of branch points. However, Riemanns thesis is a strikingly original piece of work which examined geometric
properties of analytic functions, conformal mappings and the connectivity of surfaces.
In proving some of the results in his thesis Riemann used a variational principle
which he was later to call the Dirichlet Principle since he had learnt it from Dirichlets
lectures in Berlin. The Dirichlet Principle did not originate with Dirichlet, however, as
Gauss, Green and Thomson had all made use of it. Riemanns thesis, one of the most
remarkable pieces of original work to appear in a doctoral thesis, was examined on 16
December 1851. In his report on the thesis Gauss described Riemann as having:
... a gloriously fertile originality.
On Gausss recommendation Riemann was appointed to a post in Gttingen and
he worked for his Habilitation, the degree which would allow him to become a lecturer. He spent thirty months working on his Habilitation dissertation which was on
the representability of functions by trigonometric series. He gave the conditions of a
function to have an integral, what we now call the condition of Riemann integrability.
In the second part of the dissertation he examined the problem which he described in
these words:While preceding papers have shown that if a function possesses such and such a
property, then it can be represented by a Fourier series, we pose the reverse question:
if a function can be represented by a trigonometric series, what can one say about its
behaviour.
To complete his Habilitation Riemann had to give a lecture. He prepared three
lectures, two on electricity and one on geometry. Gauss had to choose one of the three
for Riemann to deliver and, against Riemanns expectations, Gauss chose the lecture
on geometry. Riemanns lecture ber die Hypothesen welche der Geometrie zu Grunde
liegen (On the hypotheses that lie at the foundations of geometry), delivered on 10
MANUALES UEX
61
MANUALES UEX
Abel and Jacobi. However Riemann was not the only mathematician working on such
ideas. Klein writes:
... when Weierstrass submitted a first treatment of general abelian functions to
the Berlin Academy in 1857, Riemanns paper on the same theme appeared in Crelles
Journal, Volume 54. It contained so many unexpected, new concepts that Weierstrass
withdrew his paper and in fact published no more.
The Dirichlet Principle which Riemann had used in his doctoral thesis was used by
him again for the results of this 1857 paper. Weierstrass, however, showed that there
was a problem with the Dirichlet Principle. Klein writes:
The majority of mathematicians turned away from Riemann ... Riemann had quite a different opinion. He fully recognised the justice and correctness of Weierstrasss
critique, but he said, as Weierstrass once told me, that he appealed to Dirichlets Principle only as a convenient tool that was right at hand, and that his existence theorems
are still correct.
We return at the end of this article to indicate how the problem of the use of Dirichlets Principle in Riemanns work was sorted out.
In 1858 Betti, Casorati and Brioschi visited Gttingen and Riemann discussed with
them his ideas in topology. This gave Riemann particular pleasure and perhaps Betti
in particular profited from his contacts with Riemann. These contacts were renewed
when Riemann visited Betti in Italy in 1863. In [16] two letter from Betti, showing the
topological ideas that he learnt from Riemann, are reproduced.
In 1859 Dirichlet died and Riemann was appointed to the chair of mathematics at
Gttingen on 30 July. A few days later he was elected to the Berlin Academy of Sciences. He had been proposed by three of the Berlin mathematicians, Kummer, Borchardt
and Weierstrass. Their proposal read:
Prior to the appearance of his most recent work [Theory of abelian functions], Riemann was almost unknown to mathematicians. This circumstance excuses somewhat
the necessity of a more detailed examination of his works as a basis of our presentation. We considered it our duty to turn the attention of the Academy to our colleague
whom we recommend not as a young talent which gives great hope, but rather as a
fully mature and independent investigator in our area of science, whose progress he
in significant measure has promoted.
A newly elected member of the Berlin Academy of Sciences had to report on their
most recent research and Riemann sent a report on On the number of primes less
than a given magnitude another of his great masterpieces which were to change the
direction of mathematical research in a most significant way. In it Riemann examined
the zeta function
( s ) =
X
Y
(1/ n s ) = (1 ps )1
n
which had already been considered by Euler. Here the sum is over all natural
numbers n while the product is over all prime numbers. Riemann considered a very
different question to the one Euler had considered, for he looked at the zeta function
as a complex function rather than a real one. Except for a few trivial exceptions, the
roots of ( s) all lie between 0 and 1. In the paper he stated that the zeta function had
62
infinitely many nontrivial roots and that it seemed probable that they all have real
part 1/2. This is the famous Riemann hypothesis which remains today one of the most
important of the unsolved problems of mathematics.
Riemann studied the convergence of the series representation of the zeta function
and found a functional equation for the zeta function. The main purpose of the paper
was to give estimates for the number of primes less than a given number. Many of
the results which Riemann obtained were later proved by Hadamard and de la Valle
Poussin.
In June 1862 Riemann married Elise Koch who was a friend of his sister. They had
one daughter. In the autumn of the year of his marriage Riemann caught a heavy cold
which turned to tuberculosis. He had never had good health all his life and in fact his
serious heath problems probably go back much further than this cold he caught. In
fact his mother had died when Riemann was 20 while his brother and three sisters all
died young. Riemann tried to fight the illness by going to the warmer climate of Italy.
The winter of 1862-63 was spent in Sicily and he then travelled through Italy, spending time with Betti and other Italian mathematicians who had visited Gttingen. He
returned to Gttingen in June 1863 but his health soon deteriorated and once again he
returned to Italy. Having spent from August 1864 to October 1865 in northern Italy,
Riemann returned to Gttingen for the winter of 1865-66, then returned to Selasca on
the shores of Lake Maggiore on 16 June 1866. Dedekind writes:
His strength declined rapidly, and he himself felt that his end was near. But still,
the day before his death, resting under a fig tree, his soul filled with joy at the glorious
landscape, he worked on his final work which unfortunately, was left unfinished.
Finally let us return to Weierstrasss criticism of Riemanns use of the Dirichlets
Principle. Weierstrass had shown that a minimising function was not guaranteed by
the Dirichlet Principle. This had the effect of making people doubt Riemanns methods.
Freudenthal writes:
All used Riemanns material but his method was entirely neglected. ... During the
rest of the century Riemanns results exerted a tremendous influence: his way of thinking but little.
Weierstrass firmly believed Riemanns results, despite his own discovery of the
problem with the Dirichlet Principle. He asked his student Hermann Schwarz to try
to find other proofs of Riemanns existence theorems which did not use the Dirichlet
Principle. He managed to do this during 1869-70. Klein, however, was fascinated by
Riemanns geometric approach and he wrote a book in 1892 giving his version of Riemanns work yet written very much in the spirit of Riemann. Freudenthal writes:
It is a beautiful book, and it would be interesting to know how it was received.
Probably many took offence at its lack of rigour: Klein was too much in Riemanns
image to be convincing to people who would not believe the latter.
In 1901 Hilbert mended Riemanns approach by giving the correct form of Dirichlets Principle needed to make Riemanns proofs rigorous. The search for a rigorous
proof had not been a waste of time, however, since many important algebraic ideas
were discovered by Clebsch, Gordan, Brill and Max Noether while they tried to prove
Riemanns results. Monastyrsky writes:
It is difficult to recall another example in the history of nineteenth-century mathe-
MANUALES UEX
63
3.8. Problemas
matics when a struggle for a rigorous proof led to such productive results.
Article by: J.J. OConnor and E.F. Robertson (http://www-history.mcs.st-and.ac.uk/
Biographies/).
3.8.
Problemas
MANUALES UEX
64
Captulo 4
MANUALES UEX
Teoremas fundamentales de la
Teora de Nmeros
4.1.
Divisores afines
65
2. Definicin : Cada f K , no nula, define un divisor afn, llamado divisor afn principal, que denotamos D ( f ):
X
D( f ) =
vx ( f ) x
xSpecmax A
Se dice que dos divisores afines D, D 0 son afnmente equivalentes si existe f K tal que
D = D 0 + D ( f ). El conjunto de los divisores afines principales de Div A , es un subgrupo
y el cociente de Div A por el subgrupo de los divisores afines principales se denota
Pic A = Div A / y se llama grupo de clases de ideales de A o grupo de Picard de A .
3. Ejercicio : Probar que Pic Z = {0}.
4. Ejercicio : Probar que Pic A = {0} si y slo si A es un dominio de ideales principales.
Si dos ideales no nulos a, a0 A son isomorfos, localizando en el punto genrico
obtenemos un isomorfismo de K -mdulos de K , que es multiplicar por una f K , luego
a0 = f a.
5. Proposicin : Se cumplen las igualdades
Conj. de ideales no nulos de A = Conj. de divisores afines efectivos
P
a = mnx11 mnxrr
7
D (a) := i n i x i
MANUALES UEX
66
I = mnx11 m xmm
s
n0x
n x + n0x
Q
Q
n Q
Observemos que xSpecmax A m x x xSpecmax A m x = xSpecmax A m x
.
Si dos ideales fraccionarios no nulos I, I 0 K son isomorfos, localizando en el punto
genrico obtenemos un isomorfismo de K -mdulos de K , que es multiplicar por una
f K , luego I 0 = f I .
/
o
{Ideales fraccionarios de K }
Q n
I D := { f K : D ( f ) D } = i m x ii
I
nx
MANUALES UEX
10. Definicin : Dado un ideal fraccionario I = mnx11 mnxrr de K definimos la norma
de I , que denotamos N ( I ), como el nmero racional positivo
Y
N ( I ) = | A /m x i | n i
i
67
N ( I ) = egr(D(I))
Es decir, el diagrama
{Divisores afines}
{Ideales fraccionarios}
gr
e
R+
es conmutativo.
Demostracin. Las aplicaciones egr , N I : {Divisores afines de K } R+ son morfismos
de grupos. Para ver que son iguales basta comprobar que coinciden sobre los puntos
x Specmax A . Efectivamente, egr(x)) = | A /m x | = N (m x ) = N ( I ( x)).
12. Proposicin : Dado un ideal a A , entonces N (a) = | A /a|. Dados dos ideales fraccionarios I 0 I , se cumple que N ( I 0 )/ N ( I ) = | I / I 0 |.
Q
n
Demostracin. Escribamos a = mnx11 mnxrr , entonces A /a = i A /m x ii y
Y
Y
n
| A /a| = | A /m x ii | = | A /m x i |n i = N (a)
i
13. Ejercicio : Sea 0 , f K . Probar que | N ( f )| = N ( f A ).
Solucin: Escribamos f = a/ b, a, b A . Entonces, ( f ) ( b) = (a) y
MANUALES UEX
68
4.2.
Divisores completos
`
1. Notacin : Sea X = Specmax A , X = Spec K Q R y X = X X .
2. Definicin : Llamaremos grupo de los divisores completos de X , que denotaremos
Div( X ), al grupo
Div( X ) = ( x X Z x) ( y X R y)
P
P
y diremos que D = x X n x x + y X y y es un divisor completo. Diremos que D | X :=
P
P
D ( f ) =
vx ( f ) x
x X
n x gr x +
x X
y gr y
y X
6. Ejercicio : Sea X el conjunto de valores absolutos de Q, mdulo equivalencia. Probar que Pic X = R.
7. Proposicin: Sea c Z. Salvo multiplicacin por invertibles existe un nmero finito de a A tal que N (a) = c.
MANUALES UEX
Demostracin. | N (a)| = | A /aA | = | c| si y slo si gr D (a) = ln | c|. Ahora bien, divisores afines efectivos de grado dado slo existen un nmero finito. Por tanto, existen a 1 , . . . , a m
de modo que gr D (a i ) = ln | c| y si gr D (a) = ln | c|, entonces Da = Da i . Luego a es igual
salvo multiplicacin por invertibles a alguno de los a i .
69
4.3. Discriminante
4.3.
1. Definicin : Sea E un R-espacio vectorial de dimensin n. Diremos que un subgrupo aditivo de E es una red si est generado por alguna base { e 1 , . . . , e n } del espacio
vectorial, es decir, = Z e 1 Z e n y E = Z R = R e 1 R e n .
Se dice que E / es el paraleleppedo generado por e 1 , . . . , e n .
2. Ejemplo : Sea K un cuerpo de nmeros, de grado d sobre Q. Sea K un Z-mdulo
libre de rango d . Consideremos la inclusin cannica
:= det((T2 ( e i , e j )))
Sean e01 , . . . , e0n E , con e0i =
j i j e j ,
MANUALES UEX
Si e01 , . . . , e0n es otra base del Z-mdulo , entonces det( i j ) = 1. Por tanto, no depende de la base de escogida.
Se define el volumen del paraleleppedo generado por e 1 , . . . , e n , por
V ol (E /) :=
p
p
| | = | det(T2 ( e i , e j ))|
70
4.3. Discriminante
5. Como,
(T2 (a i , a j )) = (tr(a i a j )) = (
d
X
k (a i a j )) = ( i (a j )) t ( i (a j )),
k=1
entonces
= det(( i (a j )))2
(donde ( i (a j )) es una matriz cuadrada de nmeros complejos de orden d ) y
V ol (O /) =
p
| | = | det(( i (a j )))|
0 = |/0 |2
V ol (O /0 ) = |/0 | V ol (O /)
Recordemos que si I 0 I son ideales fraccionarios, | I / I 0 | = N ( I 0 )/ N ( I ).
7. Proposicin : Si I es un ideal fraccionario, entonces
p
V ol (O / I ) = N ( I ) | A |
Demostracin. Sean a, b A ideales tales que I = a b1 (luego, I b1 y A b1 ).
Entonces,
V ol (O / I ) =
p
N (I )
N (I )
1
1
V
ol
(
O
/
b
)
=
N
(
b
)
V
ol
(
O
/
A
)
=
N
(
I
)
| A |
N (b1 )
N (b1 )
i< j
MANUALES UEX
71
4.4.
ID =
f K : vxi ( f ) n i , i
f K : v yi ( f ) i , yi
v x ( f ) 0, x , x i
n m \
n1
= m
m
{( i ) Rr Cs = O : | i | e i , i }
xm
x1
2. Propiedades:
n m
n1
2. El conjunto ID es finito porque es la interseccin de la red m
x1 m xm con el
compacto {( j ) Rr Cs = O : | j | e j , j }, que es finito.
MANUALES UEX
cuando
gr D ln
|K | s ln(/2)
V ol (O / I ) = N ( I )
|K | = egr(D(I))
p
| K |
72
p
6. Corolario: Si D es un divisor completo y gr D ln |K |, entonces D es linealmente
equivalente a un divisor completo efectivo.
4.5.
MANUALES UEX
73
8. Teorema de Hermite: Slo hay un nmero finito de extensiones de Q de grado y
discriminantes dados.
Demostracin. Sea K una extensin de discriminante y grado d .
Podemos suponer que i K : si i K , entonces |K[i] | | A[i] | = 4d | A |2 = 4d |K |2 ,
y como probaremos, el nmero de cuerpos cuyo valor absoluto del discriminante es
menor que 4d |2 | y grado 2 d , que contienen a i , es finito y cada uno de stos contiene
un nmero finito de subextensiones. En conclusin, el nmero de cuerpos de discriminante y grado d es finito.
Suponemos, pues, que i K (luego r = 0). Consideremos en el infinito el divisor
p
( d + ln |K |) y1 y2 . . . ys
El teorema de Riemann-Roch dbil afirma la existencia de una f A tal que | i ( f )|
e1 < 1, para todo i > 1. Como N ( f ) es un nmero entero, se sigue |1 ( f )| = | f | > 1.
Sea H = { HomQal g (K, C) : ( f ) = f }, tendremos que |( f )| > 1, para todo H . Por
tanto, H = {1 } y K = Q[ f ] (o bien, H = {1 , 1 }, en este caso K = Q[ i f ] y tomaramos i f
p
en vez de f ). Observemos, adems, que |1 ( f )| e d |K |. Por tanto, los coeficientes
del polinomio anulador de f estn acotados, pues sus races i ( f ) lo estn, y como son
nmeros enteros slo hay un nmero finito de tales polinomios.
9. Proposicin : Sea K un cuerpo de nmeros y d = dimQ K . Dado un ideal fraccionario I K , existe f I no nula, de modo que
p
| N ( f )| c| N ( I )|, con c = d ! d d (4/)s |K |
MANUALES UEX
C = {(1 , . . . , r , . . . , r+s ) O = Rr Cs :
X
ir
| i | +
2| j | t }
j>r
= d !d
i
d
(4/)
|K | | N ( I )|
74
4.6.
1 K A Div0 Pic0 0
son exactas. Sabemos que Pic( X ) es un grupo finito.
2. Proposicin : Pic0 es compacto.
p
0
= grcy1 y1 Div . Sea
Demostracin. Fijemos un divisor de grado c := ln |K |, D
c
Div
el conjunto de los divisores con soporte en el infinito de grado c. Obviamente,
0
c
c
c
0
Div = Div
, D 7 D + D
, Pic0 = Div0 /D ( A ) = Div
/D ( A ) =: Pic
y basta demosc
trar que Pic es compacto.
c
Dado D Div
, por el teorema de Riemann-Roch dbil existe f K tal que
MANUALES UEX
D + D ( f ) 0
75
c c
C := {D 00 Div
: D 00 0}
c
Es decir, D pertenece al compacto C f := C D ( f ) Div
. Observemos que c c0 0,
luego c0 c. Ahora bien, el nmero de f A , salvo multiplicacin por invertibles, tales
que gr D ( f ) c es finito. Por tanto, existe un nmero finito de funciones f i A de modo
c
que para cada D Div
, existe i tal que D C f i mod D ( A ). Por tanto,
c
Pic
= i C f i ,
3. Teorema de Dirichlet: Pic0 es un toro de dimensin r + s 1 y los invertibles
A es un grupo finito generado de rango r + s 1 y de torsin las races de la unidad
contenidas en K .
Demostracin. A es un subconjunto discreto de O , luego A es un subgrupo discreto
de O
. Consideremos el epimorfismo de grupos (que tiene seccin)
X
D : O
= (Rr Cs ) Div , ( i ) 7 (ln | i |) yi
i
4. Ejercicio : Probar que existen 1 , . . . , r+s1 A , de modo que a A si y slo si
n
MANUALES UEX
a = 1n1 r+r+ss11
para ciertos nmeros enteros n 1 , . . . , n r+s1 Z (nicos) y una raz n-sima de la unidad
K (nica).
5. Proposicin : El subgrupo de enteros de K de norma 1, {a A : N (a) = 1}, es un
grupo abeliano libre de rango r + s 1 si dimk K es impar, y es un grupo abeliano finito
generado de rango r + s 1 y torsin K si dimk K es par.
Demostracin. Si dimk K es impar, entonces r > 0, luego K R y K = {1}. Adems,
N (1) = 1, luego {a A : N (a) = 1} es un subgrupo de ndice dos de A y K {a
A : N (a) = 1} = {1}. Por tanto, {a A : N (a) = 1} es un grupo de rango r + s 1 sin torsin,
luego libre.
Si dimk K es par, entonces N () = 1 para todo K : Obviamente N (1) = 1. Si
Q
K es imaginaria entonces r = 0. Entonces, N (a) = si=1 i (a) i (a) > 0, para todo
76
6. Ejercicio : Probar que existen 1 , . . . , r+s1 A de norma 1, de modo que a A es
de norma 1, si y slo
n1
n
1 r+r+ss11
si dimk K impar.
a=
n r+s1
n1
1 r+s1 para un (nico) K , si dimk K es par.
para ciertos nmeros enteros n 1 , . . . , n r+s1 Z (nicos). Probar que existen adems
1 , . . . , i A de norma c Z, de modo que N (a) = c Z si y slo
a=
i 1 1 r+r+ss11
n
n
i 1 1 r+r+ss11
si dimk K impar.
para un (nico) K , si dimk K es par.
p
a+b
A ={
, a, b Z : a2 b2 = 4}
2
4.7.
MANUALES UEX
Demostracin. En virtud de la finitud de Pic A , basta probar el teorema para el nmero T ( n) de ideales de norma n en una clase de isomorfismos dada. El conjunto de
ideales de A est en correspondencia biunvoca con el conjunto de divisores afines efectivos y recordemos que si I es un ideal de norma n, entonces D ( I ) es un divisor de grado ln n. Por tanto, T ( n) es el nmero de divisores afines efectivos, D 0 , de grado ln n,
afnmente equivalentes a un divisor afn efectivo dado (que es equivalente a un divisor D (a), para cierto ideal a A ). Sea m = gr D (a)). La condicin D 0 = D ( f ) D (a) 0
77
(Rr Cs ) = O
(1 , . . . , r+s )
7
Div
P
i (ln | i |) yi
gr
Observemos que gr(D ( f )) = gr(D ( f )), ya que gr D ( f ) = 0. Sea G el ncleo del mor P
fismo de grupos D y Div O
, i i yi 7 ( e1 , . . . , er+s ) una seccin de D , lue
go O = G Div . Sea R Div , t 7 dt ( y1 + + yr+s ) una seccin de gr. Luego
Div = Div0 R y
O
= G Div0 R
y la homotecia por R en O
se corresponde con la traslacin por ln d en el tercer
factor de G Div0 R. Sea P Div0 el paraleleppedo fundamental de la red D ( A )
en Div0 . Para cada conjunto f A , existe f u f A tal que D ( f u) P R Div y
todos los que cumplen esta condicin son f u K (observemos adems que gr(D ( f v)) =
gr D ( f ), para todo v A ). Luego, si w = |K |, entonces w T ( n) es el nmero de elementos de la red a en el conjunto
Un := G P (, ln n + m] = n1/d U1
Es decir, w T ( n) es el nmero de elementos de la red n1/d a en U1 . Por el lema1 4.7.2,
d 1
w T ( n) = v n + O ( n d ).
2. Lema: Sea U un recinto acotado y limitado por un nmero finito de hipersuperficies
diferenciables en un espacio vectorial real E de dimensin d y sea E una red. Si
P () denota el nmero de puntos de U 1 , existe una constante no nula v tal que
MANUALES UEX
P () = vd + O (d 1 )
Demostracin. Podemos suponer que E = Rd y = Zd . Sea C = { x Rd : 0 x i 1 , i }.
`
Considerando la unin pU 1 p + C , obtenemos una figura que casi coincide con U ,
pues le faltan algunos puntos de U y le sobranpotros, pero tales puntos estn en el
compacto C de los puntos a una distancia = d / del borde C de U . Luego,
V ol (U ) V ol (C ) P ()d V ol (U ) + V ol (C )
y se concluye al observar que V ol (C ) = O ().
`
Donde E = O , = a y U = U1 {0}. U1 = G P (, m] es acotado porque si denotamos por la
igualdad G Div0 R = O
, entonces (G P (, m]) = (0, e m/d ] (G P {0}). Observemos adems
que el cierre de G P (, m] en O es igual a este conjunto unin 0 O
1
78
4.8.
La funcin zeta
1. Teorema : La serie ( x) =
n=1 n
lm ( x 1) ( x) = 1
x1
( x ) =
(1
p primo
Demostracin. La serie
n=1 n
1 1
)
px
es la funcin zeta de K .
4. Ejercicio : Probar que ( x) = Q ( x).
5. Teorema : La funcin K ( x) es continua en la semirecta x > 1,
x1
K ( x) =
Y
(1
p
1 1
)
N (p) x
MANUALES UEX
lm ( x 1) K ( x) = v
x1
P
Q
La igualdad a N (a) x = p (1 N (p) x )1 expresa la unicidad de la descomposicin
de cada ideal no nulo de A en producto de ideales primos.
79
Demostracin. Por hiptesis existe una constante c > 0 tal que | b n | < cn . Ahora, para
cada pareja de nmeros naturales m < r ,
r
X
a n n x =
n= m
r
X
n= m
( b n b n1 ) n x = b r r x b m m x +
Como | b n ( n x ( n + 1) x )| cn x
|
r
X
an n
| 2 cm
x+
R n+1
n
t x1 dt c x
Z
+cx
n= m
rX
1
n= m
R n+1
n
b n ( n x ( n + 1) x )
t x1+ dt,
t x1+ dt = (2 c +
cx
) m x+ ,
x +
(1
MANUALES UEX
grZ y=1
N (p y ) x
)1
(1
)1
x
x
N (p y )
N (p y )
grZ y>1
grZ y=1
Y
Y
Y
1
1
=
)1
(1 m x )r
(1
x
N (p y )
(p )
grZ y=1
m>1,mr d pP m,r
Q
y pP m,r (1 (p 1m )x )r definen funciones continuas en la semirrecta x > 1/2 segn 4.8.2,
hemos concluido.
80
4.8.1.
Aplicaciones
0 = q r ( x1 , . . . , xn )
tenga alguna solucin compleja es que admita soluciones modulares en infinitos primos
Demostracin. Si el sistema no tiene soluciones complejas, el teorema de los ceros
de Hilbert afirma que 0 = C[ x1 , . . . , xn ]/( q 1 , . . . , q r ) = Q[ x1 , . . . , xn ]/( q 1 , . . . , q r ) Q C, y por
tanto
Q[ x1 , . . . , xn ]/( q 1 , . . . , q r ) = 0
P
Luego existen polinomios h 1 , . . . , h r Q[ x1 , . . . , xn ] tales que i h i q i = 1. Multiplicando
P
por N N conveniente tenemos que i h0i q i = N , con h 1 , . . . , h r Z[ x1 , . . . , xn ]. Ahora
es evidente que, salvo en los primos que dividan a N , la reduccin q 1 = 0, . . . , q r = 0
mdulo p del sistema dado carece de soluciones en Z/ pZ.
Recprocamente, si el sistema considerado tiene alguna raz compleja, entonces
y el teorema de los ceros de Hilbert afirma que el sistema admite alguna solucin en
una extensin finita K de Q. Sea A el anillo de enteros de K . Como K = A Z Q, tal
solucin ser
an
a1
, . . . , xn =
x1 =
m1
mn
Q
a
con a i A y m i Z. Sea m = i m i , entonces x i = mi A m , para todo i . Como el
i
teorema 4.8.10 afirma la existencia de infinitos primos p de grado 1 en A m , se concluye
la existencia de infinitos primos p, tales que el sistema considerado tiene solucin en
Z/ pZ = A /p.
MANUALES UEX
Q[ x1 , . . . , xn ]/( q 1 , . . . , q r ) , 0
12. Corolario: Todo polinomio no constante con coeficientes enteros q( x) tiene infinitas
races modulares. Ms an, hay infinitos nmeros primos p en los que q( x) Z/ pZ tiene
todas sus races en Z/ pZ.
81
x1
x1
lo cual es contradictorio.
MANUALES UEX
16. Corolario: Si la reduccin de q( x) Z[ x] mdulo p descompone totalmente en casi
todo p, entonces q( x) descompone totalmente en Q.
Demostracin. Podemos suponer que q( x) es irreducible. Sea K = Q[ x]/( q( x)) y A =
Z[ x]/( q( x)). Observemos que un primo p Z descompone totalmente en A si y slo si
q( x) descompone totalmente en Z/ pZ[ x]. Por hiptesis, casi todo primo p Z descompone totalmente en K , luego por el teorema anterior, Q = K y q( x) = ( x ) descompone
totalmente en Q.
17. Corolario : Si un nmero entero es resto cuadrtico mdulo casi todo primo, entonces es un cuadrado perfecto.
Demostracin. Considrese en el corolario anterior q( x) = x2 n.
82
4.9. Cuestionario
4.9.
Cuestionario
MANUALES UEX
5. Consideremos R3 con el producto escalar estndar. Calcular el volumen del paraleleppedo definido por los vectores (1, 3, 2), (1, 2, 1) y (0, 1, 1).
2 i
3
83
4.10.
Biografa de Dirichlet
MANUALES UEX
DIRICHLET BIOGRAPHY
84
Lejeune Dirichlets family came from the Belgium town of Richelet where Dirichlets grandfather lived. This explains the
origin of his name which comes from Le jeune de Richele meaning Young from Richelet. Dirichlets came from the neighbourhood of Lige in Belgium and not, as many had claimed,
from France. His father was the postmaster of Dren, the town
of his birth situated about halfway between Aachen and Cologne. Even before he entered the Gymnasium in Bonn in 1817, at
the age of 12, he had developed a passion for mathematics and
spent his pocket-money on buying mathematics books. At the
Gymnasium he was a model pupil being:
... an unusually attentive and well-behaved pupil who was particularly interested
in history as well as mathematics.
After two years at the Gymnasium in Bonn his parents decided that they would
rather have him attend the Jesuit College in Cologne and there he had the good fortune to be taught by Ohm. By the age of 16 Dirichlet had completed his school qualifications and was ready to enter university. However, the standards in German universities were not high at this time so Dirichlet decided to study in Paris. It is interesting
to note that some years later the standards in German universities would become the
best in the world and Dirichlet himself would play a hand in the transformation.
Dirichlet set off for France carrying with him Gausss Disquisitiones arithmeticae a
work he treasured and kept constantly with him as others might do with the Bible. In
Paris by May 1822, Dirichlet soon contracted smallpox. It did not keep him away from
his lectures in the Collge de France and the Facult des Sciences for long and soon
he could return to lectures. He had some of the leading mathematicians as teachers
and he was able to profit greatly from the experience of coming in contact with Biot,
Fourier, Francoeur, Hachette, Laplace, Lacroix, Legendre, and Poisson.
From the summer of 1823 Dirichlet was employed by General Maximilien Sbastien Foy, living in his house in Paris. General Foy had been a major figure in the army
during the Napoleonic Wars, retiring after Napoleons defeat at Waterloo. In 1819 he
was elected to the Chamber of Deputies where he was leader of the liberal opposition
until his death. Dirichlet was very well treated by General Foy, he was well paid yet
treated like a member of the family. In return Dirichlet taught German to General
Foys wife and children.
Dirichlets first paper was to bring him instant fame since it concerned the famous
Fermats Last Theorem. The theorem claimed that for n > 2 there are no non-zero
integers x, y, z such that x n + yn = z n . The cases n = 3 and n = 4 had been proved by
Euler and Fermat, and Dirichlet attacked the theorem for n = 5. If n = 5 then one of
x, y, z is even and one is divisible by 5. There are two cases: case 1 is when the number
divisible by 5 is even, while case 2 is when the even number and the one divisible by 5
are distinct. Dirichlet proved case 1 and presented his paper to the Paris Academy in
July 1825. Legendre was appointed one of the referees and he was able to prove case
2 thus completing the proof for n = 5. The complete proof was published in September
1825. In fact Dirichlet was able to complete his own proof of the n = 5 case with an
argument for case 2 which was an extension of his own argument for case 1. It is worth
noting that Dirichlet made a later contribution proving the n = 14 case (a near miss
for the n = 7 case!).
On 28 November 1825 General Foy died and Dirichlet decided to return to Germany. He was encouraged in this by Alexander von Humboldt who made recommendations on his behalf. There was a problem for Dirichlet since in order to teach in a
German university he needed an habilitation. Although Dirichlet could easily submit
an habilitation thesis, this was not allowed since he did not hold a doctorate, nor could
he speak Latin, a requirement in the early nineteenth century. The problem was nicely solved by the University of Cologne giving Dirichlet an honorary doctorate, thus
allowing him to submit his habilitation thesis on polynomials with a special class of
prime divisors to the University of Breslau. There was, however, much controversy
over Dirichlets appointment.
From 1827 Dirichlet taught at Breslau but Dirichlet encountered the same problem
which made him choose Paris for his own education, namely that the standards at the
university were low. Again with von Humboldts help, he moved to the Berlin in 1828
where he was appointed at the Military College. The Military College was not the
attraction, of course, rather it was that Dirichlet had an agreement that he would be
able to teach at the University of Berlin. Soon after this he was appointed a professor
at the University of Berlin where he remained from 1828 to 1855. He retained his
position in the Military College which made his teaching and other administrative
duties rather heavier than he would have liked.
Dirichlet was appointed to the Berlin Academy in 1831 and an improving salary
from the university put him in a position to marry, and he married Rebecca Mendelssohn, one of the composer Felix Mendelssohns two sisters. Dirichlet had a lifelong
friend in Jacobi, who taught at Knigsberg, and the two exerted considerable influence
on each other in their researches in number theory.
In the 1843 Jacobi became unwell and diabetes was diagnosed. He was advised by
his doctor to spend time in Italy where the climate would help him recover. However,
Jacobi was not a wealthy man and Dirichlet, after visiting Jacobi and discovering his
plight, wrote to Alexander von Humboldt asking him to help obtain some financial
assistance for Jacobi from Friedrich Wilhelm IV. Dirichlet then made a request for as-
MANUALES UEX
85
MANUALES UEX
86
sistance from Friedrich Wilhelm IV, supported strongly by Alexander von Humboldt,
which was successful. Dirichlet obtained leave of absence from Berlin for eighteen
months and in the autumn of 1843 set off for Italy with Jacobi and Borchardt. After
stopping in several towns and attending a mathematical meeting in Lucca, they arrived in Rome on 16 November 1843. Schlfli and Steiner were also with them, Schlflis
main task being to act as their interpreter but he studied mathematics with Dirichlet
as his tutor.
Dirichlet did not remain in Rome for the whole period, but visited Sicily and then
spent the winter of 1844/45 in Florence before returning to Berlin in the spring of 1845.
Dirichlet had a high teaching load at the University of Berlin, being also required
to teach in the Military College and in 1853 he complained in a letter to his pupil
Kronecker that he had thirteen lectures a week to give in addition to many other
duties. It was therefore something of a relief when, on Gausss death in 1855, he was
offered his chair at Gttingen.
Dirichlet did not accept the offer from Gttingen immediately but used it to try
to obtain better conditions in Berlin. He requested of the Prussian Ministry of Culture that he be allowed to end lecturing at the Military College. However he received
no quick reply to his modest request so he wrote to Gttingen accepting the offer of
Gausss chair. After he had accepted the Gttingen offer the Prussian Ministry of Culture did try to offer him improved conditions and salary but this came too late.
The quieter life in Gttingen seemed to suit Dirichlet. He had more time for research and some outstanding research students. However, sadly he was not to enjoy
the new life for long. In the summer of 1858 he lectured at a conference in Montreux
but while in the Swiss town he suffered a heart attack. He returned to Gttingen, with
the greatest difficulty, and while gravely ill had the added sadness that his wife died
of a stroke.
We should now look at Dirichlets remarkable contributions to mathematics. We
have already commented on his contributions to Fermats Last Theorem made in 1825.
Around this time he also published a paper inspired by Gausss work on the law of
biquadratic reciprocity.
He proved in 1837 that in any arithmetic progression with first term coprime to
the difference there are infinitely many primes. This had been conjectured by Gauss.
Davenport wrote in 1980:
Analytic number theory may be said to begin with the work of Dirichlet, and in
particular with Dirichlets memoir of 1837 on the existence of primes in a given arithmetic progression.
Shortly after publishing this paper Dirichlet published two further papers on analytic number theory, one in 1838 with the next in the following year. These papers introduce Dirichlet series and determine, among other things, the formula for the class
number for quadratic forms.
His work on units in algebraic number theory Vorlesungen ber Zahlentheorie
(published 1863) contains important work on ideals. He also proposed in 1837 the
modern definition of a function
If a variable y is so related to a variable x that whenever a numerical value is
assigned to x, there is a rule according to which a unique value of y is determined,
4.11. Problemas
4.11.
MANUALES UEX
Problemas
87
4.11. Problemas
p
K K .
a i j m j | < c i, i
y X
Probar que existe 0 , f I , tal que | f | y < c y para todo y X .
MANUALES UEX
11. La batalla de Hastings (14 de octubre de 1066). Los hombres de Harold permanecan bien juntos, como solan hacer, y formaban 13 escuadrones, con el mismo nmero de hombres en cada escuadrn, y hostigaban a los esforzados normandos que se aventuraban entrar en sus reductos; porque un nico golpe de un
hacha de guerra sajona poda romper sus lanzas y cortar sus mayas... Cuando
Harold se lanz l mismo al ataque, los sajones formaban una poderoso escuadrn de hombres, gritando exclamaciones de guerra... Cuntos sajones haba
en la batalla de Hastings?
12. Probar que si dimQ K >> 0 entonces |K | >> 0.
88
4.11. Problemas
MANUALES UEX
89
MANUALES UEX
4.11. Problemas
90
Bibliografa
[1] A NDREWS, G.E.: Number Theory, Dover, 1994.
[2] A NGLIN, W.S.: The queen of mathematics. An introduction to number theory, Kluwer A.P./Texts in the Math. Sc., vol. 8 1995.
[3] B AKER , A.: Breve introduccin a la teora de nmeros, Alianza Editorial/472 AU
Ciencias, 1986.
[4] B OREVICH , Z.I. AND S HAFAREVICH , I.R.: Number Theory, Academic Press, Inc.
1966.
[5] E VEREST, W.: An introduction to number theory, Springer-Verlag/Graduate Texts
in Math., vol. 232 Versin digital en http://lope.unex.es.
[6] F ROHLICH , A.: Algebraic number theory, Cambridge U.P./Cambr. Stud. Adv.
Math., vol. 27, 1991.
[7] H ASSE H.: Number theory, Springer-Verlag/Grundl. Math. Wissensch., vol. 229,
1969.
[8] I RELAND, K.; R OSEN, M.: A classical introduction to modern number theory,
Springer-Verlag/Graduate Texts in Math., vol. 84, 1982.
[9] L ANG, S.: Algebraic number theory, Springer-Verlag/Graduate Texts in Math.,
vol. 110, 1994.
[11] M ILLER , S.J.; T AKLOO -B IGHASH , R.: An invitation to modern number theory,
Princeton University Press, 2006.
[12] N ATHANSON, M.B.: Elementary methods in
Verlag/Graduate Texts in Math., vol. 195, 2000.
number
theory,
Springer-
MANUALES UEX
91
BIBLIOGRAFA
[16] R OSE , H.E.: A course in number theory, Oxford University Press Inc., 2007.
[17] E LEMENTARY NUMBER THEORY: , International Thomson Publ. PWS Publ.CO,
1994.
[18] T ATTERSAL , J.J.: Elementary number theory in nine chapters, Cambridge University Press, 1999
[19] W EIL , A.: Number theory for beginners, Springer-Verlag, 1979.
MANUALES UEX
[20] W EIL , A.: Basic number theory, . Springer-Verlag/Grundl. Math. Wissensch., vol.
144, 1974.
92
ndice alfabtico
lgebra graduada, 53
Nmero de puntos contando grados y multiplicidades, 32
Anillo de enteros de un cuerpo, 21
Anillo de nmeros enteros, 15
Polinomio primitivo, 12
Anillo de valoracin, 44
Automorfismo de Frbenius en un primo Punto de ramificacin, 32
Punto no singular, 15
p, 35
Punto singular, 15
Cuerpo de nmeros, 21
Curva ntegra afn, 14
Curva proyectiva, 56
Red, 68
Serie de composicin de mdulos, 30
Smbolo de Legendre, 36
Valor absoluto, 48
Valor absoluto arquimediano, 48
Valoracin discreta, 43
Valoracin m-dica, 43
Valoracin real, 43
Valores absolutos equivalentes, 48
Variedad de Riemann, 46
Variedad proyectiva, 56
Volumen de un paraleleppedo, 68
Grado de un divisor, 66
Grupo de Picard, 64
Grupo de Picard completo, 66
Ideal fraccionario, 64
Ideal homogneo, 54
Ideal irrelevante, 54
ndice de ramificacin, 33
Lema de Euclides, 13
Ley de reciprocidad cuadrtica de Gauss,
36
Longitud de un mdulo, 30
MANUALES UEX
Teorema de Dirichlet, 74
Teorema de Hermite, 72
Teorema de la base de Hilbert, 10
Teorema de Riemann-Roch dbil, 70
Teorema del punto de la red de Minkowski, 70
Modulo simple, 30
Norma, 52
Norma de un ideal fraccionario, 66
93
93