Discriminant of an algebraic number field

Last updated August 22, 2023

A fundamental domain of the ring of integers of the field K obtained from Q by adjoining a root of x - x - 2x + 1. This fundamental domain sits inside K [?]Q R. The discriminant of K is 49 = 7 . Accordingly, the volume of the fundamental domain is 7 and K is only ramified at 7. Discriminant49CubicFieldFundamentalDomain.png — A fundamental domain of the ring of integers of the field K obtained from Q by adjoining a root of x − x − 2x + 1. This fundamental domain sits inside K ⊗_Q R. The discriminant of K is 49 = 7 . Accordingly, the volume of the fundamental domain is 7 and K is only ramified at 7.

In mathematics, the discriminant of an algebraic number field is a numerical invariant that, loosely speaking, measures the size of the (ring of integers of the) algebraic number field. More specifically, it is proportional to the squared volume of the fundamental domain of the ring of integers, and it regulates which primes are ramified.

The discriminant is one of the most basic invariants of a number field, and occurs in several important analytic formulas such as the functional equation of the Dedekind zeta function of K, and the analytic class number formula for K. A theorem of Hermite states that there are only finitely many number fields of bounded discriminant, however determining this quantity is still an open problem, and the subject of current research.^[1]

The discriminant of K can be referred to as the absolute discriminant of K to distinguish it from the relative discriminant of an extension K/L of number fields. The latter is an ideal in the ring of integers of L, and like the absolute discriminant it indicates which primes are ramified in K/L. It is a generalization of the absolute discriminant allowing for L to be bigger than Q; in fact, when L = Q, the relative discriminant of K/Q is the principal ideal of Z generated by the absolute discriminant of K.

Definition

Let K be an algebraic number field, and let O_K be its ring of integers. Let b₁, ..., b_n be an integral basis of O_K (i.e. a basis as a Z-module), and let {σ₁, ..., σ_n} be the set of embeddings of K into the complex numbers (i.e. injective ring homomorphisms K → C). The discriminant of K is the square of the determinant of the n by n matrix B whose (i,j)-entry is σ_i(b_j). Symbolically,

\Delta _{K}=\det \left({\begin{array}{cccc}\sigma _{1}(b_{1})&\sigma _{1}(b_{2})&\cdots &\sigma _{1}(b_{n})\\\sigma _{2}(b_{1})&\ddots &&\vdots \\\vdots &&\ddots &\vdots \\\sigma _{n}(b_{1})&\cdots &\cdots &\sigma _{n}(b_{n})\end{array}}\right)^{2}.

Equivalently, the trace from K to Q can be used. Specifically, define the trace form to be the matrix whose (i,j)-entry is Tr_K/Q(b_ib_j). This matrix equals B^TB, so the square of the discriminant of K is the determinant of this matrix.

The discriminant of an order in K with integral basis b₁, ..., b_n is defined in the same way.

Examples

Quadratic number fields: let d be a square-free integer, then the discriminant of $K=\mathbf {Q} ({\sqrt {d}})$ is^[2]

\Delta _{K}=\left\{{\begin{array}{ll}d&{\text{if }}d\equiv 1{\pmod {4}}\\4d&{\text{if }}d\equiv 2,3{\pmod {4}}.\\\end{array}}\right.

An integer that occurs as the discriminant of a quadratic number field is called a fundamental discriminant.^[3]

Cyclotomic fields: let n > 2 be an integer, let ζ_n be a primitive nth root of unity, and let K_n = Q(ζ_n) be the nth cyclotomic field. The discriminant of K_n is given by^[2]^[4]

\Delta _{K_{n}}=(-1)^{\varphi (n)/2}{\frac {n^{\varphi (n)}}{\displaystyle \prod _{p|n}p^{\varphi (n)/(p-1)}}}

where

\varphi (n)

is Euler's totient function, and the product in the denominator is over primes p dividing n.

Power bases: In the case where the ring of integers has a power integral basis, that is, can be written as O_K = Z[α], the discriminant of K is equal to the discriminant of the minimal polynomial of α. To see this, one can choose the integral basis of O_K to be b₁ = 1, b₂ = α, b₃ = α², ..., b_n = αⁿ⁻¹. Then, the matrix in the definition is the Vandermonde matrix associated to α_i = σ_i(α), whose determinant squared is

\prod _{1\leq i<j\leq n}(\alpha _{i}-\alpha _{j})^{2}

which is exactly the definition of the discriminant of the minimal polynomial.

Let K = Q(α) be the number field obtained by adjoining a root α of the polynomial x³ − x² − 2x − 8. This is Richard Dedekind's original example of a number field whose ring of integers does not possess a power basis. An integral basis is given by {1, α, α(α + 1)/2} and the discriminant of K is −503.^[5]^[6]
Repeated discriminants: the discriminant of a quadratic field uniquely identifies it, but this is not true, in general, for higher-degree number fields. For example, there are two non-isomorphic cubic fields of discriminant 3969. They are obtained by adjoining a root of the polynomial x³− 21x + 28 or x³− 21x− 35, respectively.^[7]

Basic results

Brill's theorem:^[8] The sign of the discriminant is (−1)^r₂ where r₂ is the number of complex places of K.^[9]
A prime p ramifies in K if and only if p divides Δ_K .^[10]
Stickelberger's theorem:^[11]

\Delta _{K}\equiv 0{\text{ or }}1{\pmod {4}}.

Minkowski's bound :^[12] Let n denote the degree of the extension K/Q and r₂ the number of complex places of K, then

|\Delta _{K}|^{1/2}\geq {\frac {n^{n}}{n!}}\left({\frac {\pi }{4}}\right)^{r_{2}}\geq {\frac {n^{n}}{n!}}\left({\frac {\pi }{4}}\right)^{n/2}.

Minkowski's theorem:^[13] If K is not Q, then |Δ_K| > 1 (this follows directly from the Minkowski bound).
Hermite–Minkowski theorem :^[14] Let N be a positive integer. There are only finitely many (up to isomorphisms) algebraic number fields K with |Δ_K| <N. Again, this follows from the Minkowski bound together with Hermite's theorem (that there are only finitely many algebraic number fields with prescribed discriminant).

History

Richard Dedekind showed that every number field possesses an integral basis, allowing him to define the discriminant of an arbitrary number field. Dedekind.jpeg — Richard Dedekind showed that every number field possesses an integral basis, allowing him to define the discriminant of an arbitrary number field.

The definition of the discriminant of a general algebraic number field, K, was given by Dedekind in 1871.^[15] At this point, he already knew the relationship between the discriminant and ramification.^[16]

Hermite's theorem predates the general definition of the discriminant with Charles Hermite publishing a proof of it in 1857.^[17] In 1877, Alexander von Brill determined the sign of the discriminant.^[18] Leopold Kronecker first stated Minkowski's theorem in 1882,^[19] though the first proof was given by Hermann Minkowski in 1891.^[20] In the same year, Minkowski published his bound on the discriminant.^[21] Near the end of the nineteenth century, Ludwig Stickelberger obtained his theorem on the residue of the discriminant modulo four.^[22]^[23]

Relative discriminant

The discriminant defined above is sometimes referred to as the absolute discriminant of K to distinguish it from the relative discriminant Δ_K/L of an extension of number fields K/L, which is an ideal in O_L. The relative discriminant is defined in a fashion similar to the absolute discriminant, but must take into account that ideals in O_L may not be principal and that there may not be an O_L basis of O_K. Let {σ₁, ..., σ_n} be the set of embeddings of K into C which are the identity on L. If b₁, ..., b_n is any basis of K over L, let d(b₁, ..., b_n) be the square of the determinant of the n by n matrix whose (i,j)-entry is σ_i(b_j). Then, the relative discriminant of K/L is the ideal generated by the d(b₁, ..., b_n) as {b₁, ..., b_n} varies over all integral bases of K/L. (i.e. bases with the property that b_i ∈ O_K for all i.) Alternatively, the relative discriminant of K/L is the norm of the different of K/L.^[24] When L = Q, the relative discriminant Δ_K/Q is the principal ideal of Z generated by the absolute discriminant Δ_K . In a tower of fields K/L/F the relative discriminants are related by

\Delta _{K/F}={\mathcal {N}}_{L/F}\left({\Delta _{K/L}}\right)\Delta _{L/F}^{[K:L]}

where ${\mathcal {N}}$ denotes relative norm.^[25]

Ramification

The relative discriminant regulates the ramification data of the field extension K/L. A prime ideal p of L ramifies in K if, and only if, it divides the relative discriminant Δ_K/L. An extension is unramified if, and only if, the discriminant is the unit ideal.^[24] The Minkowski bound above shows that there are no non-trivial unramified extensions of Q. Fields larger than Q may have unramified extensions: for example, for any field with class number greater than one, its Hilbert class field is a non-trivial unramified extension.

Root discriminant

The root discriminant of a degree n number field K is defined by the formula

\operatorname {rd} _{K}=|\Delta _{K}|^{1/n}.

^[26]

The relation between relative discriminants in a tower of fields shows that the root discriminant does not change in an unramified extension.

Asymptotic lower bounds

Given nonnegative rational numbers ρ and σ, not both 0, and a positive integer n such that the pair (r,2s) = (ρn,σn) is in Z × 2Z, let α_n(ρ, σ) be the infimum of rd_K as K ranges over degree n number fields with r real embeddings and 2s complex embeddings, and let α(ρ, σ) = liminf_n→∞α_n(ρ, σ). Then

\alpha (\rho ,\sigma )\geq 60.8^{\rho }22.3^{\sigma }

,

and the generalized Riemann hypothesis implies the stronger bound

\alpha (\rho ,\sigma )\geq 215.3^{\rho }44.7^{\sigma }.

^[27]

There is also a lower bound that holds in all degrees, not just asymptotically: For totally real fields, the root discriminant is > 14, with 1229 exceptions.^[28]

Asymptotic upper bounds

On the other hand, the existence of an infinite class field tower can give upper bounds on the values of α(ρ, σ). For example, the infinite class field tower over Q(√-m) with m = 3·5·7·11·19 produces fields of arbitrarily large degree with root discriminant 2√m ≈ 296.276,^[27] so α(0,1) < 296.276. Using tamely ramified towers, Hajir and Maire have shown that α(1,0) < 954.3 and α(0,1) < 82.2,^[26] improving upon earlier bounds of Martinet.^[27]^[29]

Relation to other quantities

When embedded into $K\otimes _{\mathbf {Q} }\mathbf {R}$ , the volume of the fundamental domain of O_K is ${\sqrt {|\Delta _{K}|}}$ (sometimes a different measure is used and the volume obtained is $2^{-r_{2}}{\sqrt {|\Delta _{K}|}}$ , where r₂ is the number of complex places of K).
Due to its appearance in this volume, the discriminant also appears in the functional equation of the Dedekind zeta function of K, and hence in the analytic class number formula, and the Brauer–Siegel theorem.
The relative discriminant of K/L is the Artin conductor of the regular representation of the Galois group of K/L. This provides a relation to the Artin conductors of the characters of the Galois group of K/L, called the conductor-discriminant formula.^[30]

Notes

↑ Cohen, Diaz y Diaz & Olivier 2002
1 2 Manin, Yu. I.; Panchishkin, A. A. (2007), Introduction to Modern Number Theory, Encyclopaedia of Mathematical Sciences, vol. 49 (Second ed.), p. 130, ISBN 978-3-540-20364-3, ISSN 0938-0396, Zbl 1079.11002
↑ Definition 5.1.2 of Cohen 1993
↑ Proposition 2.7 of Washington 1997
↑ Dedekind 1878, pp. 30–31
↑ Narkiewicz 2004, p. 64
↑ Cohen 1993 , Theorem 6.4.6
↑ Koch 1997 , p. 11
↑ Lemma 2.2 of Washington 1997
↑ Corollary III.2.12 of Neukirch 1999
↑ Exercise I.2.7 of Neukirch 1999
↑ Proposition III.2.14 of Neukirch 1999
↑ Theorem III.2.17 of Neukirch 1999
↑ Theorem III.2.16 of Neukirch 1999
1 2 Dedekind's supplement X of the second edition of Peter Gustav Lejeune Dirichlet's Vorlesungen über Zahlentheorie ( Dedekind 1871 )
↑ Bourbaki 1994
↑ Hermite 1857.
↑ Brill 1877.
↑ Kronecker 1882.
↑ Minkowski 1891a.
↑ Minkowski 1891b.
↑ Stickelberger 1897.
↑ All facts in this paragraph can be found in Narkiewicz 2004 , pp. 59, 81
1 2 Neukirch 1999 , §III.2
↑ Corollary III.2.10 of Neukirch 1999 or Proposition III.2.15 of Fröhlich & Taylor 1993
1 2 Hajir, Farshid; Maire, Christian (2002). "Tamely ramified towers and discriminant bounds for number fields. II". J. Symbolic Comput. 33: 415–423. doi: 10.1023/A:1017537415688 .
1 2 3 Koch 1997 , pp. 181–182
↑ Voight 2008
↑ Martinet, Jacques (1978). "Tours de corps de classes et estimations de discriminants". Inventiones Mathematicae (in French). 44: 65–73. Bibcode:1978InMat..44...65M. doi:10.1007/bf01389902. S2CID 122278145. Zbl 0369.12007.
↑ Section 4.4 of Serre 1967

Related Research Articles

In number theory, the ideal class group of an algebraic number field $K$ is the quotient group $J K / P K$ where $J K$ is the group of fractional ideals of the ring of integers of $K$ , and $P K$ is its subgroup of principal ideals. The class group is a measure of the extent to which unique factorization fails in the ring of integers of $K$ . The order of the group, which is finite, is called the class number of $K$ .

Algebraic number theory is a branch of number theory that uses the techniques of abstract algebra to study the integers, rational numbers, and their generalizations. Number-theoretic questions are expressed in terms of properties of algebraic objects such as algebraic number fields and their rings of integers, finite fields, and function fields. These properties, such as whether a ring admits unique factorization, the behavior of ideals, and the Galois groups of fields, can resolve questions of primary importance in number theory, like the existence of solutions to Diophantine equations.

In transcendental number theory, the Lindemann–Weierstrass theorem is a result that is very useful in establishing the transcendence of numbers. It states the following:

In algebraic number theory, a quadratic field is an algebraic number field of degree two over $, the rational numbers.$

In mathematics, the field trace is a particular function defined with respect to a finite field extension L/K, which is a K-linear map from L onto K.

In mathematics, the ring of integers of an algebraic number field $is the ring of all algebraic integers contained in . An algebraic integer is a root of a monic polynomial with integer coefficients: . This ring is often denoted by or . Since any integer belongs to and is an integral element of, the ring is always a subring of .$

In commutative algebra, the norm of an ideal is a generalization of a norm of an element in the field extension. It is particularly important in number theory since it measures the size of an ideal of a complicated number ring in terms of an ideal in a less complicated ring. When the less complicated number ring is taken to be the ring of integers, Z, then the norm of a nonzero ideal I of a number ring R is simply the size of the finite quotient ring R/I.

In number theory, more specifically in local class field theory, the ramification groups are a filtration of the Galois group of a local field extension, which gives detailed information on the ramification phenomena of the extension.

Chebotarev's density theorem in algebraic number theory describes statistically the splitting of primes in a given Galois extension K of the field $of rational numbers. Generally speaking, a prime integer will factor into several ideal primes in the ring of algebraic integers of K . There are only finitely many patterns of splitting that may occur. Although the full description of the splitting of every prime p in a general Galois extension is a major unsolved problem, the Chebotarev density theorem says that the frequency of the occurrence of a given pattern, for all primes p less than a large integer N, tends to a certain limit as N goes to infinity. It was proved by Nikolai Chebotaryov in his thesis in 1922, published in.$

In abstract algebra and number theory, Kummer theory provides a description of certain types of field extensions involving the adjunction of nth roots of elements of the base field. The theory was originally developed by Ernst Eduard Kummer around the 1840s in his pioneering work on Fermat's Last Theorem. The main statements do not depend on the nature of the field – apart from its characteristic, which should not divide the integer n – and therefore belong to abstract algebra. The theory of cyclic extensions of the field K when the characteristic of K does divide n is called Artin–Schreier theory.

In abstract algebra, Hilbert's Theorem 90 (or Satz 90) is an important result on cyclic extensions of fields (or to one of its generalizations) that leads to Kummer theory. In its most basic form, it states that if L/K is an extension of fields with cyclic Galois group G = Gal(L/K) generated by an element $and if is an element of L of relative norm 1, that is$

In mathematics, the interplay between the Galois group G of a Galois extension L of a number field K, and the way the prime ideals P of the ring of integers O_K factorise as products of prime ideals of O_L, provides one of the richest parts of algebraic number theory. The splitting of prime ideals in Galois extensions is sometimes attributed to David Hilbert by calling it Hilbert theory. There is a geometric analogue, for ramified coverings of Riemann surfaces, which is simpler in that only one kind of subgroup of G need be considered, rather than two. This was certainly familiar before Hilbert.

In mathematics, an affine Lie algebra is an infinite-dimensional Lie algebra that is constructed in a canonical fashion out of a finite-dimensional simple Lie algebra. Given an affine Lie algebra, one can also form the associated affine Kac-Moody algebra, as described below. From a purely mathematical point of view, affine Lie algebras are interesting because their representation theory, like representation theory of finite-dimensional semisimple Lie algebras, is much better understood than that of general Kac–Moody algebras. As observed by Victor Kac, the character formula for representations of affine Lie algebras implies certain combinatorial identities, the Macdonald identities.

In algebraic number theory, the different ideal is defined to measure the (possible) lack of duality in the ring of integers of an algebraic number field K, with respect to the field trace. It then encodes the ramification data for prime ideals of the ring of integers. It was introduced by Richard Dedekind in 1882.

The Artin reciprocity law, which was established by Emil Artin in a series of papers, is a general theorem in number theory that forms a central part of global class field theory. The term "reciprocity law" refers to a long line of more concrete number theoretic statements which it generalized, from the quadratic reciprocity law and the reciprocity laws of Eisenstein and Kummer to Hilbert's product formula for the norm symbol. Artin's result provided a partial solution to Hilbert's ninth problem.

In theoretical physics, the superconformal algebra is a graded Lie algebra or superalgebra that combines the conformal algebra and supersymmetry. In two dimensions, the superconformal algebra is infinite-dimensional. In higher dimensions, superconformal algebras are finite-dimensional and generate the superconformal group.

In number theory, the class number formula relates many important invariants of a number field to a special value of its Dedekind zeta function.

In mathematics, an algebraic number field is an extension field $of the field of rational numbers such that the field extension has finite degree . Thus is a field that contains and has finite dimension when considered as a vector space over .$

In mathematics, especially in algebraic number theory, the Hermite–Minkowski theorem states that for any integer N there are only finitely many number fields, i.e., finite field extensions K of the rational numbers Q, such that the discriminant of K/Q is at most N. The theorem is named after Charles Hermite and Hermann Minkowski.

In the mathematical field of algebraic number theory, the concept of principalization refers to a situation when, given an extension of algebraic number fields, some ideal of the ring of integers of the smaller field isn't principal but its extension to the ring of integers of the larger field is. Its study has origins in the work of Ernst Kummer on ideal numbers from the 1840s, who in particular proved that for every algebraic number field there exists an extension number field such that all ideals of the ring of integers of the base field become principal when extended to the larger field. In 1897 David Hilbert conjectured that the maximal abelian unramified extension of the base field, which was later called the Hilbert class field of the given base field, is such an extension. This conjecture, now known as principal ideal theorem, was proved by Philipp Furtwängler in 1930 after it had been translated from number theory to group theory by Emil Artin in 1929, who made use of his general reciprocity law to establish the reformulation. Since this long desired proof was achieved by means of Artin transfers of non-abelian groups with derived length two, several investigators tried to exploit the theory of such groups further to obtain additional information on the principalization in intermediate fields between the base field and its Hilbert class field. The first contributions in this direction are due to Arnold Scholz and Olga Taussky in 1934, who coined the synonym capitulation for principalization. Another independent access to the principalization problem via Galois cohomology of unit groups is also due to Hilbert and goes back to the chapter on cyclic extensions of number fields of prime degree in his number report, which culminates in the famous Theorem 94.

References

Primary sources

Brill, Alexander von (1877), "Ueber die Discriminante", Mathematische Annalen, 12 (1): 87–89, doi:10.1007/BF01442468, JFM 09.0059.02, MR 1509928, S2CID 120947279 , retrieved 2009-08-22
Dedekind, Richard (1871), Vorlesungen über Zahlentheorie von P.G. Lejeune Dirichlet (2 ed.), Vieweg, retrieved 2009-08-05
Dedekind, Richard (1878), "Über den Zusammenhang zwischen der Theorie der Ideale und der Theorie der höheren Congruenzen", Abhandlungen der Königlichen Gesellschaft der Wissenschaften zu Göttingen, 23 (1), retrieved 2009-08-20
Hermite, Charles (1857), "Extrait d'une lettre de M. C. Hermite à M. Borchardt sur le nombre limité d'irrationalités auxquelles se réduisent les racines des équations à coefficients entiers complexes d'un degré et d'un discriminant donnés", Crelle's Journal , 1857 (53): 182–192, doi:10.1515/crll.1857.53.182, S2CID 120694650 , retrieved 2009-08-20
Kronecker, Leopold (1882), "Grundzüge einer arithmetischen Theorie der algebraischen Grössen", Crelle's Journal , 92: 1–122, JFM 14.0038.02 , retrieved 2009-08-20
Minkowski, Hermann (1891a), "Ueber die positiven quadratischen Formen und über kettenbruchähnliche Algorithmen", Crelle's Journal, 1891 (107): 278–297, doi:10.1515/crll.1891.107.278, JFM 23.0212.01 , retrieved 2009-08-20
Minkowski, Hermann (1891b), "Théorèmes d'arithmétiques", Comptes rendus de l'Académie des sciences , 112: 209–212, JFM 23.0214.01
Stickelberger, Ludwig (1897), "Über eine neue Eigenschaft der Diskriminanten algebraischer Zahlkörper", Proceedings of the First International Congress of Mathematicians, Zürich, pp. 182–193, JFM 29.0172.03

Secondary sources

Bourbaki, Nicolas (1994). Elements of the history of mathematics. Translated by Meldrum, John. Berlin: Springer-Verlag. ISBN 978-3-540-64767-6. MR 1290116.
Cohen, Henri (1993), A Course in Computational Algebraic Number Theory, Graduate Texts in Mathematics, vol. 138, Berlin, New York: Springer-Verlag, ISBN 978-3-540-55640-4, MR 1228206
Cohen, Henri; Diaz y Diaz, Francisco; Olivier, Michel (2002), "A Survey of Discriminant Counting", in Fieker, Claus; Kohel, David R. (eds.), Algorithmic Number Theory, Proceedings, 5th International Syposium, ANTS-V, University of Sydney, July 2002, Lecture Notes in Computer Science, vol. 2369, Berlin: Springer-Verlag, pp. 80–94, doi:10.1007/3-540-45455-1_7, ISBN 978-3-540-43863-2, ISSN 0302-9743, MR 2041075
Fröhlich, Albrecht; Taylor, Martin (1993), Algebraic number theory, Cambridge Studies in Advanced Mathematics, vol. 27, Cambridge University Press, ISBN 978-0-521-43834-6, MR 1215934
Koch, Helmut (1997), Algebraic Number Theory, Encycl. Math. Sci., vol. 62 (2nd printing of 1st ed.), Springer-Verlag, ISBN 3-540-63003-1, Zbl 0819.11044
Narkiewicz, Władysław (2004), Elementary and analytic theory of algebraic numbers, Springer Monographs in Mathematics (3 ed.), Berlin: Springer-Verlag, ISBN 978-3-540-21902-6, MR 2078267
Neukirch, Jürgen (1999). Algebraische Zahlentheorie. Grundlehren der mathematischen Wissenschaften. Vol. 322. Berlin: Springer-Verlag. ISBN 978-3-540-65399-8. MR 1697859. Zbl 0956.11021.
Serre, Jean-Pierre (1967), "Local class field theory", in Cassels, J. W. S.; Fröhlich, Albrecht (eds.), Algebraic Number Theory, Proceedings of an instructional conference at the University of Sussex, Brighton, 1965, London: Academic Press, ISBN 0-12-163251-2, MR 0220701
Voight, John (2008), "Enumeration of totally real number fields of bounded root discriminant", in van der Poorten, Alfred J.; Stein, Andreas (eds.), Algorithmic number theory. Proceedings, 8th International Symposium, ANTS-VIII, Banff, Canada, May 2008, Lecture Notes in Computer Science, vol. 5011, Berlin: Springer-Verlag, pp. 268–281, arXiv: 0802.0194 , doi:10.1007/978-3-540-79456-1_18, ISBN 978-3-540-79455-4, MR 2467853, S2CID 30036220, Zbl 1205.11125
Washington, Lawrence (1997), Introduction to Cyclotomic Fields, Graduate Texts in Mathematics, vol. 83 (2 nd ed.), Berlin, New York: Springer-Verlag, ISBN 978-0-387-94762-4, MR 1421575, Zbl 0966.11047