Professional Documents
Culture Documents
Steven G Krantz - Dictionary of Algebra, Arithmetic, and Trigonometry (2001)
Steven G Krantz - Dictionary of Algebra, Arithmetic, and Trigonometry (2001)
Steven G Krantz - Dictionary of Algebra, Arithmetic, and Trigonometry (2001)
Comprehensive Dictionary
of Mathematics
DICTIONARY OF
ALGEBRA,
ARITHMETIC,
AND
TRIGONOMETRY
Edited by
Steven G. Krantz
CRC Press
Boca Raton London New York Washington, D.C.
Library of Congress Cataloging-in-Publication Data
This book contains information obtained from authentic and highly regarded sources. Reprinted material is
quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts
have been made to publish reliable data and information, but the author and the publisher cannot assume
responsibility for the validity of all materials or for the consequences of their use.
Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, microfilming, and recording, or by any information storage or retrieval
system, without prior permission in writing from the publisher.
All rights reserved. Authorization to photocopy items for internal or personal use, or the personal or internal
use of specific clients, may be granted by CRC Press LLC, provided that $.50 per page photocopied is paid
directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923 USA. The fee code for
users of the Transactional Reporting Service is ISBN 1-58488-052-X/01/$0.00+$.50. The fee is subject to
change without notice. For organizations that have been granted a photocopy license by the CCC, a separate
system of payment has been arranged.
The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for creating
new works, or for resale. Specific permission must be obtained in writing from CRC Press LLC for such
copying.
Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used
only for identification and explanation, without intent to infringe.
The second volume of the CRC Press Comprehensive Dictionary of Mathematics covers algebra,
arithmetic and trigonometry broadly, with an overlap into differential geometry, algebraic geometry,
topology and other related fields. The authorship is by well over 30 mathematicians, active in
teaching and research, including the editor.
Because it is a dictionary and not an encyclopedia, definitions are only occasionally accompanied
by a discussion or example. In a dictionary of mathematics, the primary goal is to define each term
rigorously. The derivation of a term is almost never attempted.
The dictionary is written to be a useful reference for a readership that includes students, scientists,
and engineers with a wide range of backgrounds, as well as specialists in areas of analysis and
differential equations and mathematicians in related fields. Therefore, the definitions are intended
to be accessible, as well as rigorous. To be sure, the degree of accessibility may depend upon the
individual term, in a dictionary with terms ranging from Abelian cohomology to z intercept.
Occasionally a term must be omitted because it is archaic. Care was taken when such circum-
stances arose to ensure that the term was obsolete. An example of an archaic term deemed to be
obsolete, and hence not included, is “right line”. This term was used throughout a turn-of-the-century
analytic geometry textbook we needed to consult, but it was not defined there. Finally, reference to
a contemporary English language dictionary yielded “straight line” as a synonym for “right line”.
The authors are grateful to the series editor, Stanley Gibilisco, for dealing with our seemingly
endless procedural questions and to Nora Konopka, for always acting efficiently and cheerfully with
CRC Press liaison matters.
Douglas N. Clark
Editor-in-Chief
absolute value of real number For a real acyclic chain complex An augmented, pos-
number r, the nonnegative real number |r|, given itive chain complex
by
r if r ≥ 0 ∂n+1 ∂n ∂n−1
· · · −→ Xn −→ Xn−1 −→ . . .
|r| =
−r if r < 0 . ∂2 ∂1
· · · −→ X1 −→ X0 → A → 0
abstract algebraic variety A set that is anal-
forming an exact sequence. This in turn means
ogous to an ordinary algebraic variety, but de-
that the kernel of ∂n equals the image of ∂n+1
fined only locally and without an imbedding.
for n ≥ 1, the kernel of equals the image of
abstract function (1) In the theory of gen- ∂1 , and is surjective. Here the Xi and A are
eralized almost-periodic functions, a function modules over a commutative unitary ring.
mapping R to a Banach space other than the
complex numbers. addend In arithmetic, a number that is to be
(2) A function from one Banach space to an- added to another number. In general, one of the
other Banach space that is everywhere differen- operands of an operation of addition. See also
tiable in the sense of Fréchet. addition.
abstract variety A generalization of the no- addition (1) A basic arithmetic operation
tion of an algebraic variety introduced by Weil, that expresses the relationship between the
in analogy with the definition of a differentiable number of elements in each of two disjoint sets
manifold. An abstract variety (also called an and the number of elements in the union of those
abstract algebraic variety) consists of (i.) a two sets.
family {Vα }α∈A of affine algebraic sets over a (2) The name of the binary operation in an
given field k, (ii.) for each α ∈ A a family of Abelian group, when the notation “+” is used
open subsets {Wαβ }β∈A of Vα , and (iii.) for each for that operation. See also Abelian group.
pair α and β in A a birational transformation be- (3) The name of the binary operation in a
tween Wαβ and Wαβ such that the composition ring, under which the elements form an Abelian
of the birational transformations between sub- group. See also Abelian group.
sets of Vα and Vβ and between subsets of Vβ (4) Sometimes, the name of one of the opera-
and Vγ are consistent with those between sub- tions in a multi-operator group, even though the
sets of Vα and Vγ . operation is not commutative.
additive group (1) Any group, usually adele Following Weil, let k be either a finite
Abelian, where the operation is denoted +. See algebraic extension of Q or a finitely generated
group, Abelian group. extension of a finite prime field of transcendency
(2) In discussing a ring R, the commutative degree 1 over that field. By a place of k is meant
group formed by the elements of R under the the completion of the image of an isomorphic
addition operation. embedding of k into a local field (actually the
equivalence class of such completions under the
additive identity In an Abelian group G, the equivalence relation induced by isomorphisms
unique element (usually denoted 0) such that of the local fields). A place is infinite if the local
g + 0 = g for all g ∈ G. field is R or C, otherwise the place is finite. For
a place v, kv will denote the completion, and if
additive identity a binary operation that is v is a finite place, rv will denote the maximal
called addition and is denoted by “+.” In this compact subring of kv . An adele is an element
situation, an additive identity is an element i ∈ S of
that satisfies the equation
kv × rv ,
i+s =s+i =s v∈P v ∈P
/
algebraic cycle By an algebraic cycle of di- algebraic equivalence of divisors Two di-
mension m on an algebraic variety V is meant a visors f and g on an irreducible variety X are
finite formal sum algebraically equivalent if there exists an alge-
c i Vi braic family of divisors, ft , t ∈ T , and points
t1 and t2 ∈ T , such that f = ft1 , and g = ft2 .
where the ci are integers and the Vi are irre- Thus, algebraic equivalence is an algebraic ana-
ducible m-dimensional subvarieties of V . The log of homotopy, though the analogy is not par-
cycle is said to be effective or positive if all the ticularly fruitful.
coefficients ci are non-negative. The support of Algebraic equivalence has the important
the cycle is the union of the subvarieties hav- property of preserving the degree of divisors;
ing non-zero coefficients. The set of cycles of that is, two algebraically equivalent divisors have
dimension m forms an Abelian group under ad- the same degree. It also preserves principal
dition, which is denoted Zm (V ). divisors; that is, if one divisor of an algebrai-
cally equivalent pair is principal, then so is the
algebraic dependence The property shared other one. (A divisor is principal if it is the di-
by a set of elements in a field, when they sat- visor of a rational function.) Thus, the group
isfy a non-trivial polynomial equation. Such an D0 /P is a subgroup of the divisor class group
equation demonstrates that the set of elements Cl 0 (X) = D/P . Here, D0 is the group of divi-
is not algebraically independent. sors algebraically equivalent to 0, P is the group
of principal divisors, and D is the group of di-
algebraic differential equation (1) An equa- visors of degree 0. The group D0 /p is exactly
tion of the form the subgroup of the divisor class group realized
F x, y, y
, y
over the field F is called an algebraic function algebraic multiplication In elementary al-
field over F . gebra, the multiplication of algebraic expres-
sions, which extends the operation of multipli-
algebraic fundamental group A generaliza- cation of numbers in arithmetic.
tion of the concept of fundamental group defined
for an algebraic variety over a field of character- algebraic multiplicity The multiplicity of an
istic p > 0, formed in the context of finite étale eigenvalue λ of a matrix A as a root of the char-
coverings. acteristic polynomial of A. See also geometric
multiplicity, index.
algebraic geometry Classically, algebraic
geometry has meant the study of geometric prop- algebraic number A complex number z is
erties of solutions of algebraic equations. In an algebraic number if it satisfies a non-trivial
modern times, algebraic geometry has become polynomial equation P (z) = 0, for which the
synonymous with the study of geometric objects coefficients of the polynomial are rational num-
associated with commutative rings. bers.
ϕj ◦ ϕi−1 : ϕi Ui ∩ Uj → ϕj Ui ∩ Uj
cohomology.
are analytic.
Amitsur complex See Amitsur cohomology.
analytic variety A set that is the simulta-
ample See ample vector bundle, ample divi- neous zero set of a finite collection of analytic
sor. functions.
ample divisor A divisor D such that nD is analytic vector A vector v in a Hilbert space
very ample for some positive integer n. A divi- H is called an analytic vector for a finite set
sor is very ample if it possesses a certain type of {Tj }m
j =1 of (unbounded) operators on H if there
canonical projective immersion. exist positive constants C and N such that
Tj · · · Tj v
≤ CN k k!
ample vector bundle A vector bundle E 1 k H
where the line bundle OE ∨ (1) on P (E ∨ ) is am- for all ji ∈ {1, . . . , m} and every positive integer
ple. That is, there is a morphism f from P (E ∨ ) k.
to a projective space Pn with OE ∨ (1)⊗ = f ∗
m
Archimedian ordered field A set which, in Arens–Royden Theorem Let C(MA ) denote
addition to satisfying the axioms for a field, also the continuous functions on the maximal ideal
possesses an Archimedian ordering. That is, the space MA of the Banach algebra A. Suppose
field F is ordered in that it contains a subset P that f ∈ C(MA ) and f does not vanish. Then
and the following properties hold: there exists a g ∈ A, for which g −1 ∈ A, and for
(i.) F is the disjoint union of P , {0}, and −P . which f/ĝ has a continuous logarithm on MA .
In other words, each x ∈ F belongs either to P , (Here ĝ denotes the Gelfand transform of g.)
or equals 0, or −x belongs to P , and these three
possibilities are mutually exclusive. arithmetic The operations of addition, sub-
(ii.) If x, y ∈ P , then x +y ∈ P and xy ∈ P . traction, multiplication, and division and their
The ordered field is also Archimedian in that properties for the integers.
the absolute value function
arithmetical equivalence An equivalence
x, if x ∈ P relation on the integers which is consistent with
|x| = 0, if x = 0 the four operations of arithmetic. (a ∼ b and
−x, if x ∈ −P c ∼ d imply a ± c ∼ b ± d, etc.) An example
is finite. That is, there exists an N such that For each prime ℘ of K there is a σ = K/k ℘
In = In+1 for all n ≥ N. ∈ G =Gal(K/k) such that
Artin L-function The function L(s, ϕ), de- Aσ ≡ AN(℘) (mod ℘), A ∈ D ;
K/k
fined as follows. Let K be a finite Galois exten- is called the Artin symbol of ℘ for the
℘
sion of a number field k with G = Gal(K/k). Abelian extension K/k. For an ideal a = ℘ e
Let ϕ : G → GL(V ) be a finite dimensional rep- of k relatively prime to the relative discriminant
resentation (characteristic 0). For each prime ℘
e
of K/k, define K/k = K/k .
of k, set L℘ (s, ϕ) = det(I − ϕ℘ N (℘)−s )−1 , a ℘
where ϕ℘ = 1e τ ∈T ϕ(σ τ ), T is the inertia ascending central series A sequence of sub-
group of ℘, |T | = e and σ is the Frobenius groups
automorphism of ℘. Then
{1} = H0 < H1 < H2 < · · · < G
L(s, ϕ) = L℘ (s, ϕ), for s > 1 .
of a group G with identity 1, where Hn+1 is the
℘
unique normal subgroup of Hn for which the
quotient group Hn+1 /Hn is the center of G/Hn .
Artin-Rees Lemma Let R be a Noetherian
ring, I an ideal of R, F a finitely generated sub- ascending chain of subgroups A sequence
module over R, and E a submodule of F . Then, of subgroups
there exists an integer m ≥ 1 such that, for all H1 < · · · < Hn < Hn+1 < · · · < G
integers n ≥ m, it follows that I n F ∩ E =
I n−m (I m F ∩ E). of a group G.
One then solves for xn and then back substitutes Or, −524.37 in the base 8 system stands for the
this value for xn into the equation number
− 5 82 + 2 81 + 4 + 3 8−1 + 7 8−2 .
tn−1 n−1 xn−1 + tnn xn = cn−1
The base 10 number system is called the deci-
and solves for xn−1 . Continuing in this way, all
mal system. For base n, the term n-ary is used;
of the variables x1 , x2 , . . . , xn can be solved for.
for example, ternary, in base 3.
backward error analysis A technique for
base point The point in a set to which a bun-
estimating the error in evaluating f (x1 , . . . , xn ),
dle of (algebraic) objects is attached. For exam-
assuming one knows f (a1 , . . . , an ) = b and has
ple, a vector bundle V defined over a manifold
control of |xi − ai | for 1 ≤ i ≤ n.
M will have to each point b ∈ M an associated
vector space Vb . The point b is the base point
Baer’s sum For given R-modules A and C,
for the vectors in Vb .
the sum of two elements of the Abelian group
ExtR (C, A).
base term For a spectral sequence E =
{E r , d r } where d r : Ep,q
r r
→ Ep−r,q+r−1 ,r =
Bairstow method of solving algebraic equa- r
2, 3, . . . and Ep,q = 0 whenever p < 0 or
tions An iterative method for finding qua- r .
q < 0, a term of the form Ep,o
dratic factors of a polynomial. The goal being
to obtain complex roots that are conjugate pairs.
basic feasible solution A type of solution of
the linear equation Ax = b. If A is an m × n
Banach algebra An algebra over the com-
matrix with m ≤ n and b ∈ Rm , suppose A =
plex numbers with a norm · , under which it
A1 + A2 when A1 is an m × m nonsingular
is a Banach space and such that
matrix. It is a solution of the form A(x1 , 0) =
xy ≤ xy A1 x1 + A2 0 = b where x1 ∈ Rm and x1 ≥ 0.
for all x, y ∈ B. If B is an algebra over the real basic form of linear programming problem
numbers, then B is called a real Banach algebra. The following form of a linear programming
basic invariants For a commutative ring K for m ≥ 1. Third, form the sequence of quo-
with identity and a ring R containing K and G tients rm = xm+1 /xm for m ≥ 1. If the polyno-
a subgroup of AutK (R), a minimal set of gener- mial has a single root, r, of largest magnitude,
ators of the ring R G . then the sequence {rm } will converge to r.
basic optimal solution A solution of a linear Bernoulli number Consider numbers Bn∗ de-
programming problem that minimizes the ob- fined by the functional equation
jective function (cost function) and is basic in ∞
the sense that, in the linear constraints x x (−1)n−1 B ∗ x 2n
n
+ − 1 =
ex − 1 2 (2n)!
a11 x1 + a12 x2 + · · · + a1n xn = b1 n=1
a21 x1 + a22 x2 + · · · + a2n xn = b2 B1∗ x 2 B2∗ x 4 B3∗ x 6
......... = − + − +···
2! 4! 6!
am1 x1 + am2 x2 + · · · + amn xn = bm
or, alternatively, by the equation
for the problem, n − m values of the n variables
x1 , x2 , . . . , xn are zero. x x ∞
Bn∗ x 2n
1− cot =
2 2 (2n)!
basic variable A variable that has value zero n=1
in a linear programming problem. The basic B1∗ x 2 B2∗ x 4 B3∗ x 6
variables lie on the boundaries of the convex re- = + + + ··· .
2! 4! 6!
gions determined by the constraints in the prob-
lem. The numbers Bn∗ are called the Bernoulli num-
bers. The definition given here is the classical
basis A subset B of a vector space V which one. There are several alternative, and more
has the property that every vector v ∈ V can modern, definitions. Bernoulli numbers arise
be expressed uniquely as a finite linear com- in the theory of special functions, in the study
bination of elements of B. That is, if V is a of hypergeometric functions, and as the coeffi-
vector space over the field F , then for a given cients of the Taylor expansions of many classical
v ∈ V , there exists a unique, finite, collec- transcendental functions.
tion of vectors x1 , x2 , . . . , xn ∈ B and scalars
α1 , α2 , . . . , αn ∈ F such that x = α1 x1 +α2 x2 + Betti numbers The nth Betti number Bn ,
· · · + αn xn . of a manifold M, is the dimension of the nth
By definition, V is finite dimensional if it has cohomology group, H n (M, R). [The group H n
a finite basis. In an infinite dimensional vec- (M, R) is the quotient group consisting of equiv-
tor space, if there is a topology on V , the sum alence classes of the closed n forms modulo the
representing a vector x may be allowed to be differentials of (n − 1) forms.]
and n
= n!
k
b(x, αy + βz) = αb(x, y) + βb(x, z) k!(n − k)!
Ñ = (N \{p}) ∪ Ũ , Bp (N ) = Ñ / ∼ ,
birational mapping For V and W irreducible
algebraic varieties defined over k, a closed irre- where z ∼ w if z ∈ N \{p} and w = (z, l) ∈ Ũ .
ducible subset T of V × W where the closure of The blowing up of N at p is π : Bp (N ) → N .
the projection T → V is V , the closure of the
projection T → W is W , and k(V ) = k(T ) = BN-pair A pair of subgroups (B, N ) of a
k(W ). Also called birational transformation. group G such that:
(i.) B and N generate G;
birational transformation See birational (ii.) B ∩ N = H *N ; and
mapping. (iii.) the group W = N/H has a set of genera-
tors R such that for any r ∈ R and any w ∈ W
Birch-Swinnerton-Dyer conjecture The (a) rBw ⊂ BwB ∩ BrwB,
rank of the group of rational points of an el- (b) rBr = B.
liptic curve E is equal to the order of the 0 of
L(s, E) at s = 1. Consider the elliptic curve Bochner’s Theorem A function, defined on
E : y 2 = x 3 − ax − b where a and b are inte- R, is a Fourier-Stieltjes transform if and only if it
gers. If E(Q) = E ∩ (Q × Q), by Mordell’s is continuous and positive definite. [A function
Theorem E(Q) is a finitely generated Abelian f , defined on R, is defined to be positive definite
group. Let N be the conductor of E, and if if
p | N , let ap + p be the number of solutions of f (y)f (x − y)dy > 0
y 2 = (x 3 − ax − b) (mod p). The L-function R
of E, for all x-values.]
L(s, E) = 1 − εp p −s Borel subalgebra A maximal solvable sub-
p|N algebra of a reductive Lie algebra defined over
an algebraically closed field of characteristic 0.
1 − ap p −s + p 1−2s
p|N Borel subgroup A maximal solvable sub-
group of a complex, connected, reductive Lie
where εp = 0 or ± 1. group.
block A term used in reference to vector bun- Borel-Weil Theorem If Gc is the complex-
dles, permutation groups, and representations. ification of a compact connected group G, any
bounded torsion group A torsion group T Bravais group The group of all orthogonal
where there is an integer n ≥ 0 such that t n = 1 transformations that leave invariant a given lat-
for all t ∈ T . tice L.
Cartan’s criterion of semisimplicity A Lie Carter subgroup Any finite solvable group
algebra L is semisimple if and only if the Killing contains a self-normalizing, nilpotent subgroup,
form K of L is nondegenerate. called a Carter subgroup.
Cartan’s criterion of solvability Let gl(n, Cartesian product If X and Y are sets, then
K) be the general linear Lie algebra of degree the Cartesian product of X and Y , denoted X ×
n over a field K and let L be a subalgebra of Y , is the set of all ordered pairs (x, y) with x ∈ X
gl(n, K). Then L is solvable if and only if and y ∈ Y .
tr(AB) = 0 (tr(AB) = trace of AB), for every
A ∈ L and B ∈ [L, L]. Cartier divisor A divisor which is linearly
equivalent to the divisor 0 on a neighborhood of
each point of an irreducible variety V .
Cartan’s Theorem (1) E. Cartan’s Theo-
rem. Let W1 and W2 be the highest weights of
Casimir element Let β1 , . . . , βn be a basis of
irreducible representations w1 , w2 of the Lie al-
the semisimple Lie algebra L. Using the Killing
gebra L, respectively. Then w1 is equivalent to
form K of L, let mij = K(βi , βj ). Also, let
w2 if and only if W1 = W2 .
mij represent the inverse of the matrix (mij ) and
(2) H. Cartan’s Theorem. The sheaf of let c be an element of the quotient
associative
ideals defined by an analytic subset of a com- algebra Q(L), defined by c = mij βi βj . This
plex manifold is coherent. element c is called the Casimir element of the
semisimple Lie algebra L.
Cartan subalgebra A subalgebra A of a Lie
algebra L over a field K, such that A is nilpotent Casorati’s determinant The n × n determi-
and the normalizer of A in L is A itself. nant
D(c1 (x), . . . , cn (x)) =
Cartan subgroup A subgroup H of a group
c1 (x) c2 (x) ··· cn (x)
G such that H is a maximal nilpotent subgroup
c1 (x + 1) c2 (x + 1) ··· cn (x + 1)
of G and, for every subgroup K of H of finite
,
index in H , the normalizer of K in G is also of
···
finite index in K.
c1 (x + n − 1) c2 (x + n − 1) ··· cn (x + n − 1)
character group The set of all characters λk + an−1 λk−1 + · · · + an−k+1 λ + an−k = 0 ,
of a group G, with addition defined by (X1 +
which is again called the characteristic equation
X2 )(x) = X1 (x)·X2 (x). The character group is
associated with this given difference equation.
Abelian and is sometimes called the dual group
(3) The above two definitions can be extended
of G. See character.
for a system of linear differential (difference)
characteristic Let F be a field with identity equations.
1. If there is a natural number c such that c1 = (4) Moreover, if M = (mij ) is a square ma-
1 + · · · + 1 (c 1s) = 0, then the smallest such c trix of degree n over a field F , then the algebraic
is a prime number p, called the characteristic of equation |λI − M| = 0 is also called the char-
the field F . If there is no natural number c such acteristic equation of M.
that c1 = 0, then we say that the characteristic
characteristic linear system Let S be a non-
of the field F is 0.
singular surface and let A be an irreducible al-
characteristic class (1) Of an R-module gebraic family of positive divisors of dimension
extension 0 → N → X → M → 0, d on S such that a generic member M of A is an
the element 00 (idN ) in the extension module irreducible non-singular curve. Then, the char-
Ext1R (M, N ), where idN is the identity map on acteristic set forms a (d − 1)-dimensional linear
system and contains TrM |M| (the trace of |M| on
N in HomR (N, N ) ∼ = Ext0R (N, N ) and 00 is
M) as a subfamily. This linear system is called
the connecting homomorphism Ext0R (N, N ) →
the characteristic linear system of A.
Ext1R (M, N ) obtained from the extension se-
quence. See connecting homomorphism. characteristic multiplier Let Y (t) be a fun-
(2) Of a vector bundle over base space X, damental matrix for the differential equation
any of a number of constructions of a particu-
lar cohomology class of X, chosen so that the y = A(t)y . (∗)
bundle induced by a map f : Y → X is the
image of the characteristic class of the bundle Let ω be a period for the matrix A(t). Suppose
over X under the associated cohomological map that H is a constant matrix that satisfies
f ∗ : H ∗ (X) → H ∗ (Y ). See Chern class, Euler
Y (t + ω) = Y (t)H , t ∈ (−∞, ∞) .
class, Pontrjagin class, Stiefel-Whitney class,
Thom class. Then an eigenvalue µ for H of index k and mul-
tiplicity m is called a characteristic multiplier
characteristic equation (1) If we substitute for (∗), or for the periodic matrix A(t), of index
y = eλx in the general nth order linear differen- k and multiplicity m.
tial equation
y (n) (x) + an−1 y (n−1) (x) + . . . characteristic multiplier Let Y (t) be a fun-
damental matrix for the differential equation
+a1 y (x) + a0 y(x) = 0
y = A(t)y . (∗)
with constant coefficients ai (i = n − 1, . . . , 0)
and then divide by eλx , we obtain Let ω be a period for the matrix A(t). Suppose
n
λ + an−1 λ n−1
+ · · · + a1 λ + a 0 = 0 , that H is a constant matrix that satisfies
circulant See cyclic determinant. (iii.) the family {GK } is closed under intersec-
tion and conjugation;
circular units The collection of units of the
s (iv.) 8 GK is the trivial subgroup of G;
form 1−ζ
1−ζ t , where ζ is a pn th root of unity, p is
a prime, and s ≡ t (modp) (and p |s, t). (2) a G-module A, the formation module,
such that A is the union of its submodules A(GK )
class (1) (Algebra.) A synonym of set that that are fixed by GK ;
is used when the members are closely related, (3) cohomology groups H r (L/K), defined
like an equivalence class or the class of residues as H r (GK /GL , A(GK ) ), for which H 1 (L/K)
modulo m. = 0 whenever GL is normal in GK ;
(2) (Logic.) A generalization of set, includ- (4) for each field K, there is an isomorphism
ing objects that are “too big” to be sets. Con- A → invK A of the Brauer group H 2 (∗ /K)
sideration of classes allows one to avoid such into Q/Z such that
class of curve The degree of the tangential Clifford numbers The elements of the Clif-
equation of a curve. ford algebra. See Clifford algebra.
closed under operation A set S, with a bi- where the composition of any two successive
nary operation ∗ such that a ∗ b ∈ S for all δ j is zero. In this context, the elements of C i
a, b ∈ S. are called i-cochains, the elements of the ker-
nel of δ i , i-cocycles, the elements of the im-
closure property The property of a set of be- age of δ i−1 , i-coboundaries, the mapping δ i ,
ing closed under a binary operation. See closed the ith coboundary operator, the factor module
under operation. H i (C) = ker δ i /im δ i−1 , the ith cohomology
module and the set H (C) = {H i (C) : i ∈ Z},
coalgebra Let ρ and ρ be linear mappings the cohomology module of C. If c is an i-cocycle
defined as ρ : V → V ⊗F V , and ρ : V → (i.e., an element of ker δ i ), the corresponding el-
F , where F is a field, V is vector space over ement c + im δ i−1 of H i (C) is called the coho-
F , and ⊗ denotes tensor product. Then, the mology class of c. Two i-cocycles belong to the
triple (V , ρ, ρ ) is said to be a coalgebra over same cohomology class if and only if they dif-
F , provided (1V ⊗ ρ) ◦ ρ = (ρ ⊗ 1V ) ◦ ρ and fer by an i-coboundary; such cocycles are called
(1V ⊗ ρ ) ◦ ρ = (ρ ⊗ 1V ) ◦ ρ = 1V . cohomologous.
p q
OV |U → OV |U → F|U → 0 and B(I ) is zero.
column nullity For an m × n matrix A, the common divisor A number that divides all
number m − r(A), where r(A) is the rank of A. the numbers in a list. For example, 3 is a com-
mon divisor of 6, 9, and 12. See also greatest
column vector A matrix with only one col- common divisor.
umn, i.e., a matrix of the form
a1 common fraction A quotient of the form a/b
a2 where a, b are integers and b = 0.
.. .
.
common logarithm Logarithm to the base
am
10 (the logarithm that was most often used in
Also called column matrix. arithmetic calculations before electronic calcu-
lators were invented). See also logarithm.
combination of things When r objects are
selected from a collection of n objects, the r common multiple A number that is a multi-
selected objects are called a combination of r ple of all the numbers in a list. For example, 12
objects from the collection of n objects, pro- is a common multiple of 3, 4, and 6. See also
vided that the selected objects are all regarded least common multiple.
as having equal status and not as being in any
particular order. (If the r objects are put into a commutant If S is a subset of a ring R, the
particular order, they are then called a permu- commutant of S is the set S = {a ∈ R : ax =
tation, not a combination, of r objects from n.) xa for all x ∈ S}. Also called: commutor.
The number of different combinations
ofr ob-
n n
jects from n is denoted by Cr or and commutative algebra (1) The name given to
r
the subject area that considers rings and modules
n!
equals . in which multiplication obeys the commutative
r!(n − r)! law, i.e., xy = yx for all elements x, y.
commensurable Two non-zero numbers a (2) An algebra in which multiplication obeys
and b such that a = nb,√for some rational
√ num- the commutative law. See algebra.
ber n. For example, 2√ 2 and 3√2 are com-
mensurable because 2 2 = 23 (3 2). All ra- commutative field A field in which multi-
tional numbers are commensurable with each plication is commutative, i.e., xy = yx, for all
other. No irrational number is commensurable elements x, y. (The axioms for a field require
with any rational number. that addition is always commutative. Some ver-
sions of the axioms insist that multiplication has
common denominator When two or more to be commutative too. When these versions of
fractions are about to be added, it is helpful to re- the axioms are in use, a commutative field be-
express them first, so that they have the same de- comes simply a field, and the term division ring
nominator, called a common denominator. This is used for structures that obey all the field ax-
process is also described as putting the fractions ioms except commutativity of multiplication.)
over a common denominator. For example, to See also field.
3
simplify x−1 − x+1
2
we might write
commutative group See Abelian group.
3 2
−
x−1 x+1
commutative law The requirement that a bi-
3(x + 1) 2(x − 1) nary operation ∗, on a set X, satisfy x ∗y = y ∗x,
= −
(x − 1)(x + 1) (x − 1)(x + 1) for all x, y ∈ X. Addition and multiplication of
x+5 real numbers both obey the commutative law;
= . matrix multiplication does not.
(x − 1)(x + 1)
complementary slackness This refers to complete linear system The set of all effec-
Tucker’s Theorem on Complementary Slackness, tive divisors linearly equivalent to a divisor.
which asserts that, for any real matrix A, the in-
equalities Ax = 0, x ≥ 0 and t uA ≥ 0 have complete local ring See local ring.
solutions x, u satisfying At u + x > 0.
completely positive mapping For von Neu-
mann algebras A and B, a linear mapping T :
complementary submodule Let N be a sub- A → B such that for all n ≥ 1 the induced
module of a module M. Then a complementary mappings
submodule to N in M is a submodule N of M
such that M = N ⊕ N . In ⊗ T : Mn (A) = Mn (C) ⊗ A → Mn (B)
E = E1 ⊕ E2 ⊕ · · · ⊕ Em ,
complete group A group G whose center
is trivial and all its automorphisms are inner.
such that each Ei is irreducible. See irreducible
Thus, G and the automorphism group of G are
representation.
canonically isomorphic.
Also called semi-simple.
complete integral closure Let O be an in- completely solvable group A group that is
tegral domain in a field K and M an O-module the direct product of simple groups.
contained in K. Let S be the set of all val-
uations on K that are nonnegative on O. If complete pivoting A process of solving an
υ ∈ S,let Rυ be the valuation ring of υ. Then n × n linear system of equations Ax = b. By
M = v∈S Rn M is the completion of M. If O a succession of row and column operations, one
is the integral closure of O, M is the complete may solve this equation once A has been trans-
integral closure of the O-module M = OM. formed to an upper triangular matrix. Suppose
complete valuation ring Let k be a field. A complex analytic geometry (1) Analytic
subring R of k is called a valuation ring if, for geometry, i.e., the study of geometric shapes
any x ∈ k, we have either x ∈ R or x −1 ∈ R. A through the use of coordinate systems, but
valuation ring gives rise to a valuation, or norm, within a complex vector space rather than the
on K. The valuation ring is complete if every more usual real vector space so that the coor-
Cauchy sequence in this valuation converges. dinates are complex numbers rather than real
numbers.
(2) The study of analytic varieties (the sets of
complete Zariski ring See Zariski ring. common zeros of systems of analytic functions),
complex semisimple Lie algebra A complex complex variable A variable whose values
Lie algebra that is semisimple, i.e., does not have are complex numbers.
an Abelian invariant subalgebra.
component (1) When a vector or force is
complex semisimple Lie group A complex expressed as an ordered pair (a1 , a2 ) in two di-
Lie group whose Lie algebra is semisimple, i.e., mensions, as a triple (a1 , a2 , a3 ) in three dimen-
does not have an Abelian invariant subalgebra. sions, or as an n-tuple (a1 , a2 , . . . , an ) in n di-
mensions, the numbers a1 , a2 , . . . , an are called
complex simple Lie algebra A complex Lie its components.
algebra that is simple, i.e., not Abelian and hav- (2) When a vector v in a vector space V is
ing no proper invariant subalgebra. expressed in the form a1 e1 + a2 e2 + · · · + an en
where a1 , a2 , . . . , an are scalars and e1 , e2 , . . . ,
en is a basis of V , the scalars a1 , a2 , . . . , an are
complex simple Lie group A complex Lie
called the components of v with respect to the
group whose Lie algebra is simple, i.e., not
basis e1 , e2 , . . . , en .
Abelian and having no proper invariant subal-
gebra. (3) When a vector or force v is expressed as
w+x, where w is parallel to a given direction and
x is perpendicular to that direction, w is called
complex special orthogonal group See com-
the component of v in the given direction.
plex orthogonal matrix.
(4) The word component is also used loosely
to mean simply a part of a mathematical expres-
complex sphere A complex n-sphere with
sion.
center z0 and radius r is the set of all points at
distance r from z0 in the n-dimensional com-
composite field The smallest subfield of a
plex metric space Cn . (Cn is the set of all n-
given field K containing a given collection {kα :
tuples (a1 , a2 , . . . , an ), where each ai is a com-
α ∈ A} of subfields of K.
plex number.)
composite number An integer that is not
complex spinor group The universal cover zero, not 1, not −1, and not prime.
of SO(n, C) = {A ∈ GL(n, C) : At A = I and
det A = 1} it is denoted Spin(n, C). composition algebra An alternative algebra
A over a field F (characteristic = 2), with iden-
complex structure A complex analytic struc- tity 1 and a quadratic norm n : A → F such
ture on a differentiable manifold. Complex that n(x, y) = n(x)n(y).
structures may also be put on real vector spaces
and on pseudo groups. See analytic structure. composition factor See composition series.
complex symplectic group The set of all n× composition factor series For a group G
n matrices A with complex numbers as entries with composition series G = G0 ⊃ G1 ⊃
and having the property that AT J A = J (where · · · ⊃ Gr = {e}, the sequence G0 /G1 , . . . ,
AT is the transpose of A, J is the matrix Gr−1 /Gr .
$ %
0 I composition series A series of subgroups
−I 0 G0 , G1 , . . . , Gn of a group G such that G0 = 1,
continuous cocycle A matrix that arises as coradical Of a coalgebra, the sum of its sim-
the derivative of a group action on a manifold. ple subcoalgebras. See coalgebra, simple co-
algebra.
contragredient representation Of a repre-
sentation ρ of a group G on some vector space coregular representation The contragredi-
V , the representation ρ∗ of G on the dual space ent representation of the right regular represen-
V ∗ defined by ρ ∗ (g) = ρ(g −1 )∗. See repre- tation reciprocal to a given left regular represen-
sentation, dual space. tation. See contragredient representation.
This theorem was first proved by Fischer for Coxeter group A group with generators ri ,
matrices (1905) and later (1920) it was extended i ∈ I , and relations of the form (ri rj )aij = 1,
by Courant to differential operators. where all aii = 1 and whenever i = j , aij > 1
(aij may be infinite, implying no relation be-
covariant A term describing a type of func- tween ri and rj in such a case).
tor, in constrast with a contravariant functor. A
covariant functor F : C → D assigns to every Cramer’s rule Assume we are given n linear
object c of the domain category C an object F (c) equations in n unknowns. That is, we are given
of the codomain category D and to every arrow the linear equations Li (X) = n
j =1 Cij Xj ,
α : c → c of C an arrow F (α) : F (c) → F (c ) where the Cij are given numbers, the Xj are
of D in such a way that F (idC ) = idD , where idC unknowns and i = 1, . . . , n. We are asked
and idD are the identity arrows of the respective to solve the n equations Li (X) = bj , where
categories, and that T (β ◦ α) = T (β) ◦ T (α) (◦ j = 1, . . . , n. One can write these equations in
represents composition of arrows in both cate- matricial form AX = B, where A is a square
gories) whenever β ◦ α is defined in C. See also n × n matrix and X = (X1 , . . . , Xn ) and B =
contravariant. (b1 , . . . , bn ) are vectors in Rn . Cramer gave
a formula for solving these equations provid-
covering Of a (nonsingular) algebraic curve ing the matrix A has an inverse. Let | · | de-
X over the field k, a curve Y for which there note the determinant of a matrix, which maps a
exists a k-rational map Y → X which induces square matrix to a number. (See determinant.)
an inclusion of function fields k(x) X→ k(Y ) The matrix A will have an inverse provided the
making k(Y ) separably algebraic over k(X). number |A| is not zero. So, assuming A has an
inverse (this is also referred to by saying that A
covering family A family of morphisms in a
is non-singular), the solution values of Xj are
category C which define the Grothendieck topol-
given as follows. In the matrix A, replace the
ogy on C.
ith column of A (that is, the column made up of
Ci1 , . . . , Cin ) by B. We again obtain a square
coversine function The function
matrix Ai and the solution numbers Xi = |A i|
|A| .
covers α = 1 − sin α . If X is the n vector made up of these numbers,
it will solve the system and this is the only so-
lution.
Coxeter complex A thin chamber complex in
which, to every pair of adjacent chambers, there Cremona transformation A birational map
exists a root containing exactly one of the cham- from a projective space over some field to itself.
bers. When C ⊂ C we say that C is a face of C . See birational mapping.
Two chambers are adjacent if their intersection
has codimension 1 in each: the codimension of cremona transformation A birational map
C in C is the number of minimal nonzero faces of the projective plane to itself.
cubic equation An equation of the form ax 3 + cycle (1) In graph theory, a graph Cn on ver-
bx 2 + cx + d = 0, where a, b, c, d are constants tices v1 , v2 , . . . , vn whose only edges are be-
(and a = 0). tween v1 and v2 , v2 and v3 , . . . , vn and v1 .
(2) In a permutation group Sn , a permutation
cup product (1) In a lattice or Boolean alge- with at most one orbit containing more than one
bra, the fundamental operation a ∨ b, also called object.
the join or sum of the elements a and b. (3) In homology theory, any element in the
(2) In cohomology theory, where H r (X, Y ; kernel of a homomorphism in a homology com-
G) is the cohomology of the pair (X, Y ) with plex.
coefficients in the group G, the operation that
sends the pair (f, g), f ∈ H r (X, Y ; G1 ), g ∈ cyclic algebra A crossed product algebra of
H s (X, Z; G2 ), to the image a cyclic extension field F (over a base field k)
with its (cyclic) Galois group G = /g0 with
f ∪ g ∈ H r+s (X, Y ∪ Z; G3 )
respect to a factor set that is determined by a
of the element f ⊗ g under the map single nonzero element a of k. The elements
of this algebra are uniquely of the form α0 +
H r (X, Y ; G1 ) ⊗ H s (X, Z; G2 ) α1 b
+ · · · + αn−1 bn−1 ,where the αi come from
→ H r+s (X, Y ∪ Z; G3 ) , F , 1, b, b2 , . . . , bn−1 are the formal basis el-
ements of the algebra, and n is the order of G.
induced by the diagonal map The multiplication is determined by the relations
0 : (X, Y ∪ Z) → (X, Y ) × (X, Z) . bα = α g b, bi bj = bi+j and bn = a.
cyclotomic field Any extension of a field ob- where the product is over the proper divisors d
tained by adjoining roots of unity. of n.
D 238.74999· · · .
dn−1 dn−2 · · · d1 d0 · d−1 d−2 · · · , decomposition field Let the ring A be closed
in its quotient field K. Suppose that B is its
in which the di are the digits of x; d0 and d−1 integral closure in a finite Galois extension L,
are separated by the decimal point (.). The digits with group G. Then B is preserved by elements
of x are determined
recursively by the formulas of G. Let ℘ be a maximal ideal of A and B
dn−1 = 10xn−1 and for k > 1, a maximal ideal of B that lies above ℘. Now
GB is the subgroup of G consisting of those
k−1
x− j =1 (dn−j 10
n−j ) elements that preserve B. Observe that GB acts
dn−k = , in a natural way on the residue class field B/B
10n−k
and it leaves A/℘ fixed. To any σ ∈ GB , we
can associate an element σ ∈ B/B over A/℘;
where n is the unique integer that satisfies 10n−1 the map
≤ x < 10n and t is the floor function (the σ −→ σ
3
greatest integer ≤ t). For example, if x = 238 , thereby induces a homomorphism of GB into
4
then n = 3, d2 = 2, d1 = 3, d0 = 8, d−1 = the group of automorphisms of B/B over A/℘.
7, d−2 = 5. The only possible digits are 0, 1, 2, The fixed field in GB is called the decompo-
3, 4, 5, 6, 7, 8, 9. The digit d0 is called the ones sition field of B, and is denoted Ldec .
digit, d1 the tens digit, d2 the hundreds digit, d3
the thousands digit, etc.; also, d−1 is the tenths decomposition field Let the ring A be closed
digit, d−2 the hundredths digit, etc. It can be in its quotient field K. Suppose that B is its inte-
shown that, precisely when x is rational, the se- gral closure in a finite Galois extension L, with
quence of digits of x eventually repeats. That is, group G. Then B is preserved by elements of
there is a smallest integer p for which di−p = di G. Let ℘ be a maximal ideal of A and B a maxi-
for every i less than some fixed index. Here we mal ideal of B that lies above ℘. Now GB is the
call the string of digits di−1 di−2 · · · di−p a re- subgroup of G consisting of those elements that
peating block and p the period of the represen- preserve B. Observe that GB acts in a natural
tation of x. In the special case that the repeating way on the residue class field B/B and it leaves
block is the single digit 0, the convention is to A/℘ fixed. To any σ ∈ GB we can associate an
drop all the trailing zeros from the representation element σ ∈ B/B over A/℘; the map
and say that x has a finite or terminating decimal
σ −→ σ
expansion. Further, it is possible in this case to
give a second, distinct expansion of x: if x has thereby induces a homomorphism of GB into the
finite decimal expansion with final nonzero digit group of automorphisms of B/B over A/℘.
definite quadratic form Let E be a finite di- degenerate A term found in numerous sub-
mensional vector space over the complex num- jects in mathematics. For example, in algebraic
bers C . Let L be a symmetric bilinear form on geometry, when considering the homogeneous
E. (See bilinear form, symmetric form.) The bar resolutions of Abelian groups, one uses sub-
quadratic form associated with L is the func- groups generated by (n + 1)-tuples (y0 , y1 , . . . ,
tion Q(x) = L(x, x). If Q(x) > 0, [resp., yn ) with yi = yi+1 for at least one value of i;
Q(x) ≥ 0, < 0, ≤ 0] the form is called posi- such an (n + 1)-tuple is called degenerate.
tive definite [resp., nonnegative definite, nega-
tive definite, nonpositive definite]. As an im- degree of divisor If a polynomial p(x) is
portant application, assume F (x, y) is a smooth factored as follows:
function of two real variables x and y. Let f be
the quadratic part of the Taylor expansion of F p(x) = a0 (x − x1 )n1 · · · (x − xk )nk ,
deflation A process of finding other eigenval- degree of polynomial term See degree of
ues of a matrix when one eigenvalue and eigen- polynomial.
vector are known. More specifically, if A is an
n × n matrix with eigenvalues λ1 , . . . , λn , and De Moivre’s formula See De Moivre’s The-
Av = λ1 v, with v a nonzero (column) vector, orem.
then, for any other vector u, the eigenvalues of
the matrix B = A − vuT are De Moivre’s Theorem For any integer n and
any angle θ the complex equation (De Moivre’s
λ1 − uT v, λ2 , . . . , λn . formula)
By choosing uT to be a multiple of the first row (cos θ + i sin θ)n = cos(nθ ) + i sin(nθ )
of A and scaled so that uT v = λ1 , the first row
of B becomes identically zero. holds.
derived equation See derivation of equation. the determinant is a11 a22 − a12 a21 . For a larger
n×n matrix A, the determinant is defined recur-
derived functor If T is a functor then its sively, as follows: Let Ai,j be the (n−1)×(n−1)
left-derived functors are defined inductively as matrix created from A by removing the ith row
follows: Let T0 = T . If Sn is any connected and the j th column. Then
sequence of (additive) functors, then each natu- n
.
ral transformation S0 → T0 extends to a unique det A = (−1)i+j aij det Ai,j
morphism {Sn : n ≥ 0} → {Tn : n ≥ 0} of con- i=1
nected sequences of functors. The right derived
functors are defined similarly. This sum can be computed, and is the same, for
any choice of j between 1 and n.
derived series Given a Lie algebra G, we
define its derived series G0 , G1 , . . . , inductively determinant factor If A is a matrix with ele-
by G0 = G, Gn+1 = [Gn , Gn ], n ≥ 0, where, ments in a principal ideal ring (for example, the
for any subsets S and T of G, [S, T ] denotes the integers or a polynomial ring over a field), then
differential subring If R is a differential ring directed set A set X equipped with a partial
with derivations D1 , . . . , Dk , then a subring of ordering and such that if x, y ∈ X then there
S of R is a differential subring if Di S ⊂ S for exists z ∈ X such that x ≤ z and y ≤ z.
all i. See differential ring.
direct factor Either H or K, in the direct
differential variable See differential poly- product H × K or H ⊗ K. See direct product.
nomial.
direct integral See integral direct sum.
differentiation (1) In a chain complex, a
map, usually denoted dn or δn , from one module direct limit Suppose {Gµ }µ∈I is an indexed
to the next. An example of a differentiation is family of Abelian groups, where I is a directed
the boundary operator encountered in the study set. Suppose that there is also a family of ho-
of simplicial complexes. See also chain com- momorphisms ϕµν : Gµ → Gν , defined for all
plex, boundary operator. µ < ν, such that: ϕµµ : Gµ → Gµ is the iden-
(2) Of a function f (x) of a real variable, at a tity, and if µ < ν < κ then ϕνκ ◦ ϕµν = ϕµκ .
real number x = a, the limit Consider the disjoint union of the groups Gµ
and form an equivalence relation by xµ ∼ xν ,
f (a + h) − f (a) xµ ∈ Gµ and xν ∈ Gν , if for some upper bound
f (a) = lim .
h→0 h κ of µ and ν we have ϕµκ (xµ ) = ϕνκ (xν ). Then
the direct limit is defined to be the set of equiv-
Many generalizations to other topological spaces alence classes and is denoted by
exist.
lim Gµ .
→ µ∈I
digit A symbol in a number system. For
example, in the binary number system, the only See also directed set.
digits that are used are 0 and 1. See also duodec-
imal number system. direct product (1) A group G is called the
internal direct product of subgroups H and K
dihedral group An algebraic group Dn , gen- if the following three conditions hold: H and
erated by two elements: a, which is a rotation K are normal subgroups of G, H ∩ K contains
of the Euclidean plane about the origin through only the identity element, and G = H K. This
n , and b, which is reflection through
angle of 2π internal direct product is denoted H × K.
the y-axis. Dn is the group of symmetries of the (2) If H and K are any two groups, then the
regular n-gon, and has order 2n. external direct product of H and K, denoted
H ⊗ K, is the Cartesian product: {(h, k) : h ∈
dimension The number of vectors in the ba- H, k ∈ K}. H ⊗K is a group, defining multipli-
sis of a vector space V . If the basis is finite, cation componentwise, i.e., (h1 , k1 ) · (h2 , k2 ) =
then V is called finite dimensional. In this case, (h1 · h2 , k1 · k2 ). See also internal product, ex-
if V is a vector space of dimension n over the ternal product.
real numbers R, then it is isomorphic to the Eu-
clidean space Rn . If the basis is infinite, then V direct proportion Quantity x is directly pro-
is called infinite dimensional. See also basis. portional to quantity y, or varies directly as y,
if there is a constant k = 0 such that x = ky.
Diophantine equation An equation in which (k is called the constant of proportionality.) See
solutions are restricted to the integers. also inverse proportion, joint proportion.
direct decomposition A group G has a di- direct sum (1) In the case where V1 , V2 , . . . ,
rect decomposition G = H1 × H2 × · · · × Hn Vn are all vector spaces over the same field F ,
dual curve Suppose that is an irreducible dual linear programming problem (1) An-
plane curve of degree m ≥ 1 in a projective other linear programming problem which is in-
plane. The dual curve ˆ in the dual projective timately related to a given one. Consider the
plane is the closure of the set of tangent lines to linear programming problem: minimize c1 x1 +
at its nonsingular points. The dual of ˆ is . c2 x2 + · · · + cn xn , where all xi ≥ 0 and subject
to the system of constraints:
dual graded module Suppose A = n≤0 An
a11 x1 + a12 x2 + · · · + a1n xn ≥ b1
If A∗n is the
is a graded module over a field k.
∗ a21 x1 + a22 x2 + · · · + a2n xn ≥ b2
dual of the module An , then A = A∗n is the ..
dual graded module of A. .
am1 x1 + am2 x2 + · · · + amn xn ≥ bm .
dual homomorphism A mapping φ : L1 →
L2 , of one lattice to another, such that φ(x∩y) = The following linear programming problem is
φ(x) ∪ φ(y) and φ(x ∪ y) = φ(x) ∩ φ(y). known as the dual problem: maximize b1 y1 +
E
eigenspace in the weaker sense Suppose
that L is a linear space, T is a linear transfor-
mation on L, and λ is an eigenvalue of T . Then
the collection of all elements v in L such that
(T − λI )k v = 0, for some integer k > 0, is the
e One of the most important constants in eigenspace in the weaker sense, corresponding
mathematics. The following are two common to the eigenvalue λ. Such an element v is some-
ways of approximating this number: times called a root vector of T .
1 n
e = lim 1 + eigenvalue For an n × n matrix (or a linear
n→∞ n operator) A, a number λ, such that there exists a
and non-zero n × 1 vector v satisfying the equation
∞
1 Av = λv. The product of the eigenvalues of a
e= , matrix equals its determinant. See also eigen-
n!
n=0 vector.
where n! = 1 · 2 · 3 · · · n is the factorial.
The number e is transcendental and its value eigenvector A non-zero n × 1 vector v that
is approximately 2.7182818 . . . . The letter e satisfies the equation Av = λ v for some num-
stands for Euler. ber λ, for a given n × n matrix A. See also
eigenvalue.
effective divisor See integral divisor.
Eisenstein series One of the simplest exam-
effective genus Suppose that is an irre- ples of a modular form, defined as a sum over
ducible algebraic curve and ˜ is a nonsingular a lattice. In detail: Let be a discontinuous
curve that is birationally equivalent to . Then group of finite type operating on the upper half
the genus of ˜ is the effective genus of . An al- plane H , and let κ1 , . . . , κh be a maximal set
gebraic curve is rational if its effective genus of cusps of which are not equivalent with re-
is 0 and elliptic if its effective genus is 1. spect to . Let i be the stabilizer in of κi ,
and fix an element σi ∈ G = SL(2, R) such that
eigenfunction A function f which satisfies σi ∞ = κi and such that σi−1 i σi is equal to the
the equation T (f ) = λf , for some number λ, group 0 of all matrices of the form
where T is a linear transformation from a space
of functions into itself. For example, if T is 1 b
the linear transformation on the space of twice 1
differentiable functions of one variable:
with b ∈ Z. Denote by y(z) the imaginary part
d 2f of z ∈ H . The Eisenstein series Ei (z, s) for the
T (f ) =
dx 2 cusp κi is then defined by
then cos(x) is an eigenfunction of T , with λ =
−1. Eigenfunctions arise in the analysis of par- Ei (z, s) = y(σi−1 σ z)s , σ ∈ 1 \ ,
tial differential equations such as the heat equa-
tion. See also eigenvalue. where s is a complex variable.
χ (K) is a generalization of the Euler charac- evolution Another term for the process of
teristic, V − E + F , where V , E, F are, re- extracting a root. See extraction of root.
spectively, the numbers of vertices, edges, and
faces of a simple closed polyhedron (namely, a exact sequence Consider a sequence of mod-
polyhedron that is topologically equivalent to ules {Mi } and a sequence of homomorphisms
a sphere). Euler’s Theorem in combinatorial {hi } with
topology states that V − E + F = 2. See also hi : Mi−1 −→ Mi
Lefschetz number. for all i = 1, 2, . . . The sequence of homomor-
phisms is called exact at Mi if Imhi = Kerhi+1 .
Euler product Consider the Riemann zeta The sequence is called exact if it is exact at every
function Mi . For a sequence
1 1 h2
ζ (s) = 1 + + ··· + s + ··· , 0 −→ M1 −→ M2 ,
2s k
Here H n are the cohomology groups. exact sequence of homology (1) Suppose A
(2) If 0 → A → B → C → 0 is a short ex- is a subspace of X. Then there exists a long
act sequence of complexes, then there are natural exact sequence
maps δ i : H i (C) → H i+1 (A), giving rise to an
· · · → Hn (A) → Hn (X)s
exact sequence
→ Hn (X, A) → Hn−1 (A) → · · · ,
· · · → H i (A) → H i (B) → H i (C)
called the exact sequence of homology, where
→H i+1
(A) → · · · Hn (X) denotes the nth homology group of X
and Hn (X, A) is the nth relative homology group.
of cohomology. (2) If 0 → A → B → C → 0 is a short ex-
act sequence of complexes, then there are natural
exact sequence of ext (contravariant) If maps δi : Hi (C) → Hi−1 (A), giving rise to an
exact sequence
0 −→ A −→ B −→ C −→ 0
· · · → Hi (A) → Hi (B) → Hi (C)
is a short exact sequence of modules over a ring
R and M is an R-module (on the same side as → Hi−1 (A) → · · ·
A), then the associated contravariant exact se- of homology.
quence of ext is a certain long exact sequence of
the form exact sequence of Tor Let A be a right -
module and let B → B → B be an exact
0 → Hom(C, M) → Hom(B, M) sequence of left -modules. Then there exists
→ Hom(A, M) an exact sequence
→ Ext1R (C, M) → Ext1R (B, M) Torn (A, B ) → Torn (A, B) → Torn (A, B )
→ Ext1R (A, M) → · · · → Tor1 (A, B ) → Tor1 (A, B)
→ Tor1 (A, B ) → A ⊗ B → A ⊗ B
→ · · · → ExtnR (C, M) → ExtnR (B, M)
→ A ⊗ B → 0 .
→ ExtnR (A, M) → · · · .
exceptional compact real simple Lie algebra
exact sequence of ext (covariant) If Since compact real semisimple Lie algebras are
in one-to-one correspondence with complex
0 −→ A −→ B −→ C −→ 0 semisimple Lie algebras (via complexification),
the classification of compact real simple Lie al-
is a short exact sequence of modules over a ring gebras reduces to the classification of complex
R and M is an R-module (on the same side as A), simple Lie algebras. Thus, the compact real
The sum of digits of 683, 256, and 939 are 17 ≡ exponential function of a matrix Let A be
8 mod 9, 13 ≡ 4 mod 9 and 21 ≡ 3 mod 9, an n × n matrix over the complex numbers C.
respectively. Indeed, 8 + 4 = 12 ≡ 3 mod 9. The exponential function f (A) = eA is defined
F ∞
1
k!
k=0
1 1
= 1 + + + ···
1! 2!
finer If T and U are two topologies on a space is a finite continued fraction usually denoted by
X, and if T ⊆ U, then U is said to be finer than 1 + 2+ 1 1
3.
T.
finite field A field F which is a finite set. In
finite A term associated with the number of such a case, the prime field of F can be identified
elements in a set. A set S is finite if there ex- with Zp , the field of residue classes, modulo p,
ists a natural number n and a one-to-one corre- for some prime integer p. (See prime field.) It
spondence between the elements of S and the follows that |F | = pn for some integer n. We
0 → N → N → N → 0 ,
first syzygy Let R = k[x1 , . . . , xn ] be a poly-
nomial ring of n variables x1 , . . . , xn over a field the induced sequence
k and, relying on the natural gradation of R [i.e.,
deg (xi ) = 1, deg (c) = 0 for c ∈ k], let M be 0 → M ⊗R N → M ⊗R N → M ⊗R N → 0
a finitely generated graded R-module. Desig- is exact, then R is a flat R-module.
nate by (f1 , . . . , fm ) a minimal basis of M over Here M ⊗R N is a tensor product of a right
R consisting of homogeneous elements. Intro- R-module M and a left R-module N .
duce m indeterminates gi (i = 1, . . . , m) and Remark: In view of the isomorphism between
m
the free R-module F , F = Rgi generated by M ⊗R N and N ⊗R M, we could drop the qual-
i=1 ifiers left and right.
flat resolution Let B be a left R module, form ring Let (R, P ) be a local ring and Q
where R is a ring with unit. A flat resolution of its P -primary ideal. Set
B is an exact sequence,
Fi = Qi /Qi+1 , i = 0, 1, . . . ; Q0 = R ,
φ2 φ1 φ0
· · · −→ E1 −→ E0 −→ B −→ 0 ,
and for A = A (mod Qi+1 ) ∈ Fi B = B (mod
where every Ei is a flat left R module. (We Qj +1 ) ∈ Fj , require
shall define exact sequence shortly.) There is
a companion notion for right R modules. Flat AB = A B modQi+j +1 ∈ Fi+j .
resolutions are important in homological alge-
bra and enter into the dimension theory of rings
Then the form ring of R with respect to Q is
and modules. See also flat dimension, flat mod-
defined as a graded ring F generated by F1 over
ule, injective resolution, projective resolution.
F0 and equal to the direct sum of modules
An exact sequence is a sequence of left R
modules, such as the one above, where every φi ∞
is a left R module homomorphism (the φi are F = Fi ,
called connecting homomorphisms), such that i=0
Im(φi+1 ) = Ker(φi ). Here Im(φi+1 ) is the im-
age of φi+1 , and Ker(φi ) is the kernel of φi . In where Fi is a module of homogeneous elements
the particular case above, because the sequence of degree i.
ends with 0, it is understood that the image of
φ0 is B, that is φ0 is onto. There is a companion formula (1) A formal expression of a propo-
notion for right R modules. sition in terms of local symbols.
(2) A formal expression of some rule or other
formal group Formal groups are analogs of results (e.g., Frenet formula, Stirling formulas,
the local Lie groups, for the case of algebraic etc.).
k-groups with a nonzero prime characteristic. (3) Any sequence of symbols of a formal cal-
culus.
formal scheme A topological local ringed
space that is locally isomorphic to a formal spec- forward elimination A step in the Gauss
trum Spf(A) of a Noetharian ring A. elimination method of solving a system of linear
equations
formal spectrum Let R be a Noetherian ring
which is complete with respect to its ideal I n
(in the I -adic topology), so that its completion aij xj = bi , (i = 1, . . . , n)
along I can be identified with R. The formal j =1
spectrum Spf(R), of R, is a pair (X , OX ) con-
sisting of a formal scheme X = V (I ) ⊂ Spec consisting of the following steps
(R) and a sheaf of topological rings OX defined
as follows: (m+1) (m) (m) (m)
(m)
aij = aij − aim amj /amm ,
>(D(f ) ∩ X , OX ) = limn>0 Rf /I n Rf , f ∈ R .
← (m+1) (m) (m)
(m) (m)
bi = bi − aim bm /amm ,
Here V (I ) is a set of primitive ideals of R con-
taining I , D(f ) = Spec(R) − V (f ), f ∈ R
are elementary open sets forming a base of the i, j = m + 1, . . . , n, with
Zariski topology, and >(Q, O) designates the
(1) (1)
set of sections over Q of a sheaf space O over aij ≡ aij , bi = bi ,
(1), (12)(34), (13)(24), (14)(23) . where cν1 , cν2 , . . . , cνq is obtained from c1 , c2 ,
. . . , cp by deleting the vanishing terms ci = 0.
It is Abelian and, as a transitive permutation Defining, finally
group, it is imprimitive with the imprimitivity
system {12}, {34}. W (x) = V f (x), f (x), . . . , f (n) (x)
Also called Klein’s four group or Vierergruppe
V. and
fractional ideal (1) (Of an algebraic num- free group A free product of infinite cyclic
ber field k) Let k be an algebraic number field groups. The number of free factors is called the
of finite degree (i.e., an extension field of Q of rank of the group F . Alternatively, (i.) a free
finite degree) and I an integral domain consist- group Fn on n generators is a group generated
ing of all algebraic integers. Further, let p be by a set of free generators (i.e., by the generators
the principle order of k (i.e., an integral domain that satisfy no relations other than those implied
k ∩ I whose field of quotients is k) and a an by the group axioms); (ii.) a free group is a group
integral ideal of k (i.e., an ideal of the principle with an empty set of defining relations.
order p). Then a fractional ideal of k is a p- To form a free group we can start with a free
module that is contained in k (i.e., pa ⊂ a) such semigroup, defined on a set of symbols S =
that αa ⊂ p for some α (α ∈ k, α = 0). See {a1 , a2 , . . . }, that consists of all words (i.e., fi-
algebraic number, algebraic integer. nite strings of symbols from S, repetitions being
(2) (of a ring R) An R-submodule a of the allowed), including an empty word representing
ring Q of total quotients of R such that there the unity. Next, we extend S to S that contains
exists a non-zero divisor q of R such that qa ⊂ the inverses and the identity e
R. See ring of total quotients.
S = e, a1 , a1−1 , a2 , a2−1 , . . . .
fractional programming Designating: x an
n-dimensional vector of decisive variables, m
an n-dimensional constant vector, Q a positive Then a set of equivalence classes of words
definite, symmetric, constant n × n matrix and formed from S with the law of composition de-
( , ) the scalar product, the problem of fractional fined by juxtaposition (to obtain a product αβ of
programming is to maximize the expression two words α and β, we attach β to the end of
α; to obtain the inverse of α reverse the order of
(x, m) symbols ai while replacing ai by ai−1 and vice
√ , versa) is the free group F on the set S.
(x, Qx)
The equivalence relation (designated by “∼”)
subject to nonnegativity of x (x ≥ 0) and linear used to define the equivalence classes is defined
constraints Ax ≤ b. via the elementary equivalences ee ∼ e, ai ai−1
∼ e, ai−1 ai ∼ e, ai e ∼ ai , ai−1 e ∼ ai−1 , eai =
fractional root A root of a polynomial p(x)
ai , eai−1 = ai−1 , so that α ∼ β if α is obtain-
with integral coefficients
able from β through a sequence of elementary
p(x) = a0 x n + a1 x n−1 + · · · + an equivalences.
that has the form r/s. One can show that r and s free module For a ring R, an R-module that
are such that an is divisible by r and a0 is divisi- has a basis.
ble by s. Any rational root of monic polynomial Remarks: (i.) If R is a field, then every R-
having integral coefficients is thus an integer. module is free (i.e., a linear space over R).
Also called rational root. (ii.) A finitely generated module V is free if
there is an isomorphism φ : R n → V , where R
free Abelian group The direct product (fi- is a commutator ring with unity.
nite or infinite) of infinite cyclic groups. Equiv- (iii.) A free Z-module is also called a free
alently, a free Abelian group is a free Z-module. Abelian ring.
Newton polynomials
F : Fq → Fq , F : x (→ x p
n
is injective and thus an automorphism of Fq . sr (x) = xir .
(See also Frobenius automorphism.) In fact, F i=1
generates the cyclic Galois group G(Fq /Fp ). See also Frobenius generating function.
Generally, for a scheme X over a finite field
Fq of q(= pn ) elements, the Frobenius endo- Frobenius generating function For an arbi-
morphism is an endomorphism ψ : X → X trary partition (α) of n, written in the form
such that ψ = I d (the identity mapping) on the
n
set of Fq –points of X (i.e., on the set of points of α1 α2
αn
X defined over Fq ) and the mapping of the struc- (α) ≡ 1 2 ...n , iαi = n ,
i=1
ture sheaf ψ ∗ : OX → OX raises the elements
of OX to the qth power. the Frobenius generating function F(α) (x) ≡
Thus, for an affine variety X ⊂ An defined F(α) (x1 , . . . , xn ) is given by the product
over Fq , we have
q F(α) (x) = J(x)s(α) (x) ,
q
ψ (x1 , . . . , xn ) = x1 , . . . , xn .
where J(x) is a polynomial, expressible as a
Vandermonde determinant
[λ]
Frobenius formula The characters χ(α) of
the irreducible representation [λ] ≡ (λ1 , λ2 , J(x) ≡ J (x1 , x2 , . . . , xn ) = xi − x j
i<j
. . . , λn ), associated with the class (α) ≡ (1α1
2α2 3α3 . . . nαn ) of the symmetric (≡ permuta- n−1
x x2n−1 ... xnn−1
tion) group Sn , where the partitions [λ] and (α) 1
. ..
of n satisfy the relations .. .
= x2 x22 ... xn ,
2
n n 1
x1 x2 ... xn
λi = n = iαi ,
1 1 ... 1
i=1 i=1
λ1 ≥ λ2 ≥ · · · ≥ λn ≥ 0 , and s(α) (x) ≡ ni=1 (sr (x))αi is a product of
αi th powers of power sums (or Newton polyno-
are given by the following Frobenius formula: mials) sr ,
[λ] n
χ(α) J[λ] (x) = F(α) (x) , sr (x) = xir .
λ i=1
for all x ∈ G. Then the elements of G that do where N [N] labels the classes DN of H that are
not lie in H , together with the identity element contained in (CN ∩ H ) and have sN elements.
(composed of inflation, restriction, and trans- fundamental system (Of solutions of a sys-
gression mappings) is exact and is referred to as tem of linear homogeneous equations)
the fundamental exact sequence. n
aij xj = 0, i = 1, . . . , m) . (1)
fundamental operations of arithmetic The j =1
operations of addition, subtraction, multiplica-
tion, and division. See operation. Often, extrac- n
tion of square roots is added to this list. Let fi = aij Xj , (i = 1, . . . , m) be linear
Starting with the set N of natural numbers, j =1
which is closed with respect to the first three forms over a field F, fi : F n → F . Clearly,
fundamental operations, one carries out an ex- the solutions of (1) form a (right) linear space
tension to the field of rational numbers Q rep- V over F , since if xki ≡ (x1(k) , . . . , xn(k) ) ∈
resenting the smallest domain in which the four F n , (k = 1, . . . , r) are solutionsof (1), so
fundamental operations can be carried out in- is their (right) linear combination rj =1 xj cj ,
discriminately, excepting division by zero. Ad- cj ∈ F . In fact, V is the kernel of the (left) lin-
joining the operation of square root extractions, ear mapping G : F n → F m given by G : x ( →
we arrive at the field of complex numbers C. (f1 (x), . . . , fm (x)). Designating the dimension
These operations can be defined for various of V by d, d = dim V , we can distinguish the
algebraic systems, in which case the commuta- following cases:
tive and associative laws may not hold. (i.) d = 0: In this case the system (1) has
only the trivial solution x = 0.
fundamental point See fundamental curve. (ii.) d > 0: Choosing a basis {x1 , . . . , xd }
for V , we see that any solution of (1) is a (right)
fundamental root system (of a (complex) linear combination of the xk , (k = 1, . . . , d).
semi-simple Lie algebra g) A similar concept One then says that x1 , . . . , xd form a fundamen-
arises in Kac-Moody algebras, in algebraic tal system of solutions of (1). Clearly, a nontriv-
groups, algebraic geometry, and other fields. Let ial solution is found if and only if r < n, r being
sec α = cos α ,
1
csc α = sin α ,
1 orem, the unit group Ek of an algebraic number
1 + tan2 α = sec2 α, 1 + cot 2 α = csc2 α. field k is the direct product of a cyclic group of
Furthermore, we have that a finite order and the free Abelian multiplicative
group of rank r (note that r = r1 +r2 −1, where
sin2 α + cos2 α = 1 , r1 and r2 designate, respectively, the number of
real and complex conjugates x (i) , i = 1, . . . , n
implying that the points P = (x, y) lying on of any x ∈ k, so that r1 + 2r2 = n, n being the
the unit circle x 2 + y 2 = 1 with its center at the degree of k over Q). A basis (e1 , e2 , . . . , er ) of
origin (0, 0) can be expressed in terms of the this free Abelian group is referred to as a system
angle α as P = (cos α, sin α). of fundamental units of k. See Dirichlet Unit
The important addition formulas read Theorem, unit group.
sin(α ± β) = sin α cos β ± cos α sin β,
cos(α ± β) = cos α cos β ∓ sin α sin β, fundamental vectors The vectors e1 , e2 , . . . ,
tan(α ± β) = (tan α ± tan β)/(1 ∓ tan α tan β). en of an n-dimensional vector space V over F ,
When α = β we have forming a basis of V , are referred to as funda-
sin 2α = 2 sin α cos α, of V . Clearly, for any x ∈ V , we
mental vectors
cos 2α = cos2 α − sin2 α = 2 cos2 α − 1 = have that x = ni=1 ξi ei , ξi ∈ F , and the ξi are
1 − 2 sin2 α, called the components of x, with respect to the
and for a general integral multiple of α fundamental vectors e1 , . . . , en .
Galois cohomology Let K/k be a finite Ga- Galois extension A finite field extension K/k
lois extension with the Galois group G(K/k). such that the order of the Galois group G(K/k)
Suppose, further, that G(K/k) acts on some is equal to the degree [K : k] = dimk K of the
Abelian group A. The Galois cohomology field extension K, i.e.,
groups H n (G(K/k), A) ≡ H n (K/k, A), n ≥
0 are then the cohomology groups defined by the |G(K/k)| = [K : k] = dimk K .
(cochain) complex (F n , d), with F n consisting Remarks:
of all mappings G(K/k)n → A and d designat- (i.) The degree [K : k] of the field exten-
ing the coboundary operator (see cohomology sion K/k, k ⊂ K, equals the dimension of K
groups). When the extension K/k is of an in- as a k-vector space, [K : k] = dimk K. One
finite degree, one also requires that the Galois distinguishes quadratic ([K : k) = 2), cubic
topological group acts continuously on the dis- ([K : k] = 3), biquadratic (K : k] = 4), finite
crete group A and the mappings for the cochains ([K : k] < ∞), etc., extensions.
in F n are also continuous. (ii.) All quadratic and biquadratic extensions
One also defines Galois cohomology for a are Galois extensions.
non-Abelian group A, in which case one usually (iii.) For any finite field extension K/k, the
restricts oneself to zero- and one-dimensional order of the Galois group g ≡ |G(K/k)| divides
cohomology groups, H 0 and H 1 , respectively. the degree of the extension [K : k], i.e., g|[K :
In the first case, H 0 (K/k, A) = AG(K/k) repre- k].
sents a set of fixed points in A under the action (iv.) For a Galois extension K/k with the Ga-
of the Galois group G(K/k), while in the sec- lois group G(K/k), the fixed field K G is given
ond case H 1 (K/k, A) is the quotient set of the by k, i.e., K G = k. See also Galois theory.
set of 1-dimensional cocycles.
The concept of Galois cohomology enables Galois field Finite fields Fq , q = p n , are
one to define the cohomological dimension of also called Galois fields. See finite field.
the Galois group Gk of a field k. See coho-
mological dimension. Non-Abelian Galois co- Galois group For an extension K of a field k,
homology enables the classification of principal the group of all k-automorphisms of K is called
homogeneous spaces of group schemes and, in the Galois group of the field extension K/k and
particular, to classify types of algebraic varieties is denoted by G(K/k).
(using Galois cohomology groups of algebraic Remarks:
groups). (i.) A k-automorphism of an extension field
Also called cohomology of a Galois group. K is an automorphism that acts as the identity
See also Tamagawa number, Tate-Shafarevich on the subfield k. It is also referred to as an
group. automorphism of a field extension K.
(ii.) Since every Galois extension is a split-
Galois equation Let K/k be a finite Galois ting field of some polynomial f (x) ∈ k[x],
extension with the Galois group G(K/k). Then and any two splitting fields K of a polynomial
K is a minimal splitting field of a separable poly- f (x) ∈ k[x] are isomorphic, the Galois group
nomial f (X) ∈ k[X], and we call G(K/k) the G(K/k) depends only on f (up to isomorphism).
Galois group of f (X) or the Galois group of See Galois extension. Thus, if K is the splitting
the algebraic equation f (X) = 0. (See min- field of f (x) ∈ k[x], the Galois group G(K/k)
imal splitting field, separable polynomial.) If is also referred to as the Galois group of the
G(K/k) is Abelian or cyclic, we call the equa- polynomial over k.
geometric series A series of the form global Hecke algebra Let F be either an al-
gebraic number field of finite degree or an alge-
∞
braic function field of one variable over a finite
ar j = ar + ar 2 + · · · + ar n + · · · .
field. Consider the general linear group GL2 (F )
j =0
of degree 2 over F , and denote by GA the group
Sometimes also referred to as a geometric pro- of rational points of GL2 (F ) over the adele ring
gression. of F . For each place v of F , let Hv be the Hecke
If |r| < 1, the above series converges to the algebra of the standard maximal compact sub-
sum ar/(1 − r). group of the general linear group of degree 2
over the completion of F at v. Denote by εv
Geršgorin’s Theorem The eigenvalues of an the normalized Haar measure of Hv . The re-
n × n matrix A = (aij ), with entries from the stricted tensor product of the local Hecke alge-
complex field, lie in the union of n closed discs bras Hv with respect to the family {εv } is called
(known as the Geršgorin discs), the global Hecke algebra of GA .
n
G-mapping Suppose that G is a group, and
z ∈ C : |z − aii | ≤ |aik | . X and Y are G-sets. A G-mapping is a mapping
f : X → Y such that f (gx) = gf (x) for each
i=1 k=i
g in G and each x in X.
As a consequence of Geršgorin’s Theorem,
every strictly diagonally dominant matrix is in- GNS construction A means of construct-
vertible. The latter fact is also known as the ing a cyclic representation of a C ∗ -algebra on
Lévy–Desplanques Theorem. See also ovals of a Hilbert space from a state of the C ∗ -algebra.
Cassini. The letters G, N, and S refer to Gel’fand,
Naı̆mark (Neumark), and Siegel (Segal), respec-
Givens method of matrix transformation tively.
A method for transforming a symmetric ma-
trix M into a tridiagonal matrix N = P MP −1 . good reduction For a discrete valuation ring
Here P is a product of two-dimensional rotation R, with quotient field K, and an Abelian variety
matrices. A over K, we denote by A the Neron minimal
ground form Consider the general linear group inverse See generalized inverse.
group of some degree over a field F of charac-
teristic zero. A finite number of homogeneous group minimization problem Suppose that
forms with coefficients in F that are algebrai- A is an m × n matrix with real entries, b a vec-
cally independent is called a set of ground forms. tor in Euclidean m-space Rm , and c a vector in
See general linear group. Rn . Consider the vector xB of basic variables
associated with a dual feasible basis of the linear
group One of the basic structures in algebra, programming problem of minimizing cT x sub-
consisting of a set G and a (composition) map ject to the conditions that Ax = b and xj ≥ 0
m : G × G → G, usually written as m(g, g ) = for j = 1, . . . , n. Denote by S the set of vectors
g ◦ g , g + g or g · g , satisfying the following x such that Ax = b, xj ≥ 0 for j = 1, . . . , n,
axioms: and some of the coordinates of x are restricted
(i.) (associativity) g ◦(g ◦g ) = (g ◦g )◦g to be integers. The problem of minimizing cT x
for all g, g , g ∈ G; subject to the condition that x is an element of
(ii.) (identity) there is an element e ∈ G such the group generated from S by ignoring the non-
that e ◦ g = g = g ◦ e, for all g ∈ G. negativity constraints on the coordinates of the
(iii.) (inverses) for each g ∈ G, there is g ∈ vector xB is called a group minimization prob-
G (called the inverse of g) such that g ◦ g = lem. One may solve this problem by finding
e = g ◦ g. the shortest path on a directed graph that has a
One can think of a group as describing sym- special structure. If each coordinate of the vec-
metries of certain objects. tor of basic variables for the optimal solution of
the group minimization problem is nonnegative,
group algebra Let R be a commutative ring then that solution is optimal for the original lin-
with identity and let G be a group with ele- ear programming problem; otherwise one may
ments {gα }α . The R-group algebra is the R- use a branch-and-bound algorithm to investigate
3
haversine The complex valued function h(z)
2 max (xi /mi ) < xi /mi < 1 , = (1 − cos z)/2.
i=1
then A(m1 , m2 , m3 ) is even. In 1943, Hasse Hecke algebra (1) Let H be a subgroup of
made the conjecture, which arose in his inves- a group G, and suppose that for each g in G,
tigations of class numbers of Abelian number the index of H ∩ gH g −1 in H is finite. Denote
fields. by H \G/H the set of double cosets of G by
H . If R is a commutative ring, then the module
Hasse symbol See Hasse-Minkowski char- R H \G/H has an R-algebra structure and is called
acter. the Hecke algebra of (G, H ) over R.
(2) Let H be a subgroup of a finite group G.
Hasse-Witt map See Hasse-Witt matrix. Suppose that F is a field, and e is an idempotent
in F H such that the left ideal F H e affords an
Hasse-Witt matrix Given a complete smooth F -character. Then the subalgebra e · F G · e is
algebraic variety X of dimension n defined over called a Hecke algebra.
a perfect field k of positive characteristic p, there
is a natural Frobenius morphism F : X → X, Hecke character Consider an algebraic num-
whose action on the structure sheaf OX of X is ber field of finite degree, denote its idele group
just as the pth power map. This action induces by J and its idele class group by C. A char-
another action on the nth cohomology group acter of J that is a character of C is called a
H n (X, OX ), called the Hasse-Witt map of X Hecke character (or Grössencharakter). See
and denoted also by F . The group H n (X, OX ) idele class, idele group.
is a finite-dimensional k-vector space and the
matrix corresponding to F is called the Hasse- Hecke L-function A function L(s), of a com-
Witt matrix of X. plex variable s, defined as follows. Suppose that
F is an algebraic number field of finite degree,
Hasse zeta function The function ζ (s), of a and m is an integral divisor of F . Denote by χ
complex variable s, defined as follows. Suppose the character of the ideal class group of F mod-
that V is a nonsingular complete algebraic vari- ulo m. For each integral ideal a of F , denote its
ety defined over a finite algebraic number field absolute norm by N (a). Denote by I the set of
F . For each prime ideal P of F , denote by VP all integral ideals of F . Then
the reduction of V modulo P , denote by FP the L(s) = χ (a)/N (a)s .
residue field of P , and denote by Nm the number a∈I
Held group A finite simple group of order Hessenberg method of matrix transformation
4,030,387,200. A method for transforming a matrix into a matrix
(bij ) by means of a triangular similarity trans-
Hensel’s Lemma Let A be a complete lo- formation so that bij = 0 whenever i − j ≥ 2.
cal ring with m its maximal ideal. Assume that
A is m-adically complete. Let f (x) ∈ A[x] Hessian The Jacobian of the first order deriva-
be a monic polynomial of degree n. Let π : tives of a differentiable function f = f (x1 , . . . ,
A → A/m be the canonical map and extend xn ) of n real variables; i.e., H (f ) = J (∂f/∂x1 ,
π to π : A[x] → (A/m)[x]. If g1 (x) and . . . , ∂f/∂xn ) = |∂ 2 f/∂xi ∂xj |.
g2 (x) are relatively prime monic polynomials
over A/m of degree r and n − r (respectively), Hey zeta function The function ζ (s), of a
such that π(f (x)) = g1 (x)g2 (x), then there ex- complex variable s, defined as follows. Sup-
ist h1 (x), h2 (x) ∈ A[x] having degree r and n− pose that A is a simple algebra over an algebraic
r such that f (x) = h1 (x)h2 (x) and π(hi (x)) = number field of finite degree, o is a maximal or-
gi (x), for i = 1, 2. der of A, and a is an integral left o-ideal. Denote
by N (a) the number of elements in o/a, and by
Herbrand quotient Let G = {g1 , . . . , gn } I the set of all integral left o-ideals. Then
be a finite group and let M be a G-module. Let
N : M → M be the norm defined by N (m) = ζ (s) = N (a)−s .
g1 m + · · · + gn m. Then N (M) is contained in a∈I
M G . Let h0 = M G /N (M) and let h1 be the first
cohomology group H 1 (G, M). If h0 and h1 are
finite, then the quotient of their orders is h(M), higher algebra (1) Algebra at the undergrad-
the Herbrand quotient. The Herbrand quotient uate level.
is multiplicative for short exact sequences of G- (2) Modern, or abstract, algebra.
modules: if we have 1 → M → M → M →
0, then h(M) = H (M )h(M ). higher-degree Diophantine equation An al-
gebraic equation of degree three or higher whose
Herbrand’s Lemma Let G be a finite group. coefficients are integers, such that the solutions
If M is a sub-G-module of a G-module M and sought are to be integers. See also Diophantine
the Herbrand quotient h(M ) exists, then h(M) equation.
of maps such that, for every x and y in R and Hilbert modular group If M ∈ PSL(2, Od )
every i and j in N, is a matrix and M stands for the matrix ob-
(i.) δi (x + y) =
δi x + δi y; tained by replacing each entry of M by its Galois
(ii.) δi (xy) = p,q∈N:p+q=i δp x · δq y; conjugate, then the map M → (M, M) sends
(iii.) δi (δj x) = i+j PSL(2, Od ) to an irreducible lattice in G (where
i δi+j x; and
(iv.) δ0 x = x. G = PSL(2, R)×PSL(2, R)). This last is called
a Hilbert modular group. It is known that any
highest common factor See greatest com- irreducible lattice in G is commensurable to one
mon divisor. of the Hilbert modular groups.
highest weight (1) Suppose that g is a com- Hilbert modular surface Denote by G a
plex semisimple Lie algebra, h is a Cartan subal- Hilbert modular group, and denote by H 2 the
gebra of g, and O is a lexicographic ordering on product of the upper half-space with itself. After
the real linear subspace of the complex-valued adding a finite number of points to H 2 /G and
forms on h, spanned by the root system of g rel- obtaining the minimal resolution of the space,
ative to h. If ρ is a representation of g, then the one obtains a nonsingular surface over the com-
maximal element of the set of weights of ρ with plex numbers. This surface is called the Hilbert
respect to O is called the highest weight of ρ. modular surface.
(2) If V is a vector space over the field C of
complex numbers, g is a semisimple Lie algebra Hilbert norm-residue symbol Suppose that
over C, π is an irreducible finite-dimensional F is an algebraic number field that contains a
representation of g in V , and h is a Cartan sub- primitive nth root of unity, that a and b are
algebra of g, then g has a Cartan decomposition nonzero elements of F , and that P is a prime
g = h ⊕ (⊕α gα ), with each α being an eigen- divisor of F . Put
√
value of the action of h on π . Such an eigen- a, F n b /F
value is called a weight of π . A nonzero vector σ = ,
v ∈ V is called a highest weight vector of π if P
v is an eigenvector for the action of h on π and
if gα (v) = 0 for each positive root α of g. The where the symbol on the
right
is the norm-residue
highest weight of π is the weight of the highest a,b
symbol. The symbol P defined by
weight vector of π . n
√ σ
a, b n
Higman-Sims group A finite simple group = b
P n
of order 44,352,000.
is called the Hilbert norm-residue symbol. See
Hilbert-Hasse norm residue symbol For a also norm-residue symbol.
fixed prime p, let ζn be a primitive p n th root of
1 in some algebraic closure of Qp . Let Ln = Hilbert polynomial Let R be an Artinian
Qp (ζn ), and let (·, ·)n be the p n th Hilbert norm ring and let R[x1 , . . . , xn ] be the (graded) poly-
residue symbol of Ln∗ . nomial ring (graded by degree). Let M = M0 ⊕
M1 ⊕ . . . be a finitely generated graded R[x1 ,
Hilbert modular group If M ∈ PSL(2, Od ) . . . , xn ]-module. Let LM (n) = L(Mn ) denote
is a matrix and M stands for the matrix ob- the length of the R-module. Then there is a
tained by replacing each entry of M by its Galois polynomial f (x) with rational entries such that
conjugate, then the map M → (M, M) sends LM (n) = f (n), for sufficiently large n. This
Hodge theory A body of results concern- Hölder’s Theorem If n = 2, 6, then the sym-
ing cohomology groups of manifolds. Three of metric group Sn is complete. See symmetric
these results are listed here. group, complete group.
(1) Suppose that X is a closed oriented Rie-
mannian manifold. Then each element of a co-
homology group of X with complex coefficients holomorphic function See analytic function.
has a unique harmonic representative.
(2) Suppose that V is a compact complex holomorphic functional calculus If X is a
manifold of dimension n that is the image of complex Banach space (a complete normed vec-
a holomorphic mapping from a compact Kähler tor space), T : X → X is a bounded linear
manifold of dimension n. If q = max(n − p + operator and f (z) is a holomorphic function on
1, 0), then H n (V , C) carries the Hodge structure (a neighborhood of) the spectrum of T (the set
induced by the type (p, q)-decomposition of the of eigenvalues of T , if X is finite dimensional),
space of differential forms on the manifold. then one can define f (T ) via a Cauchy type inte-
(3) If X is a smooth noncomplete irreducible gral. The map f → f (T ) is a homomorphism
variety, then H n (X, C) carries a mixed Hodge from the algebra of holomorphic functions in a
structure that is independent of the choice of neighborhood of the spectrum to T into the Ba-
complete algebraic variety X such that X − X nach algebra of bounded linear operators on T .
is a subvariety.
See Hodge structure. holosymmetric class Suppose that T is an n-
dimensional lattice in n-dimensional Euclidean
Hölder inequality If p = 0, 1, 1/p + 1/q = space, and that K is a finite subgroup of the or-
1 and ai , bi > 0, then we have thogonal group. Denote by A, B, and G the sets
1/p 1/q of arithmetic crystal classes, of Bravais types,
p q
ai bi ≤ ai bi , and of geometric crystal classes of (T , K), re-
i i i spectively. There are surjective mappings s1 :
A → B and s2 : A → G, and there is an in-
if p > 1 and jective mapping i : B → A such that s1 ◦ i is
1/p 1/q the identity mapping on B. A holosymmetric
p
q
ai bi ≥ ai bi , class is a class that belongs to the image of the
i i i
mapping s2 ◦ i.
See arithmetic crystal class, Bravais type, ge-
if 0 < p < 1. See also Hölder integral inequal-
ometric crystal class.
ity.
Hölder integral inequality Suppose that f homogeneous Of the same kind. For ex-
and g are measurable positive functions defined ample, a homogeneous polynomial is a polyno-
on a measurable set E, and suppose that p and q mial each of whose monomials has the same de-
are positive real numbers such that 1/p +1/q = gree. See homogeneous polynomial. See also
1. If p > 1, then homogeneous bounded domain, homogeneous
1/p 1/q coordinate ring, homogeneous equation, homo-
p q geneous ideal, homogeneous n-chain.
fg ≤ f g .
E E E
homogeneous element See graded ring. homological algebra Originally called mul-
tilinear algebra, this branch of algebra deals
homogeneous equation An equation, in an with the category of left (or right) R-modules
unknown function f , having the form L[f ] = over some ring R (generally assumed to be as-
0, where L acts linearly (L[c1 f1 + c2 f2 ] = sociative and to possess an identity element).
cL l[f1 ] + c2 L[f2 ]). Thus, L could be a linear A left R-module is an Abelian group A, to-
differential operator (ordinary or partial). See gether with a ring homomorphism of the ring
also homogeneous difference equation. R into the ring of endomorphisms (homomor-
phisms) of A into itself. A right R-module is de-
homogeneous function Let F be a field. A fined similarly except the ring homomorphism
function f = f (x1 , . . . , xn ) : F n → F , such is replaced by a ring anti-homomorphism (i.e.,
that, for all c ∈ F and all (x1 , . . . , xn ) ∈ F n , the order of products is reversed). When R is
we have f (cx1 , . . . , cxn ) = ck f (x1 , . . . , xn ) is commutative, a structure as a left R-module in-
called a homogeneous function of degree k. See duces a structure as a right R-module and vice
also homogeneous polynomial. versa. The morphisms of this category are the
H ∗ (X) ⊗ H ∗ (X) via a Künneth isomorphism. where 0 < |an | < 1 (n = 1, 2, · · · ) and (1 −
In this situation µ∗ is called a Hopf comultipli- |an |) < ∞. The set of points on C at which
cation and (X, µ) is called an H -space. If α is the H -series converges is called its set of con-
an element of H ∗ (X), then µ∗ (α) is called the vergence. The subject of representation theory
Hopf coproduct of α. also considers H -series.
hull-kernel topology A topology on the set hyperbolic cotangent function See hyper-
of primitive ideals (the structure space I) of a bolic function.
Banach algebra R. Under this topology, the clo-
sure of a set U is the set of primitive ideals con- hyperbolic function The six complex func-
taining the intersection of the ideals in U. tions:
hyperbolic sine: sinh(z) = exp z−exp(−z)
2
Hurwitz’s relation A relation between the hyperbolic cosine: cosh(z) = exp z+exp(−z)
2
Riemann matrices of two Abelian varieties T1 sinh(z)
hyperbolic tangent: tanh(z) = cosh(z)
and T2 that implies the existence of a homo- cosh(z)
hyperbolic cotangent: coth(z) =
morphism from T1 to T2 . Let T1 and T2 be sinh(z)
Abelian varieties with Riemann matrices F1 = hyperbolic secant: sech(z) = cosh(z)
1
(1) (1) (2) (2)
(ω1 , . . . , ω2n ) and F2 = (ω1 , . . . , ω2m ). hyperbolic cosecant: csch(z) = sinh(z)
1
.
There is a homomorphism λ : T1 → T2 if and Due to the fact that sinh(iz) = i · sin(z) and
only if there is a complex matrix W , and a ma- cosh(iz) = cos(z), where sin(z) and cos(z) are
trix M with integer entries, such that W F1 = the ordinary sine and cosine of the complex an-
F2 M. In particular, for every homomorphism gle z, the hyperbolic functions are rarely used
λ : T1 → T2 there is a representation matrix outside of certain specialized applications.
W (λ) with complex coefficients, and a represen-
tation matrix M(λ) with respect to the real coor- hyperbolic secant function See hyperbolic
dinate systems (ω1(1) , . . . , ω2n
(1) (2)
) and (ω1 , . . . , function.
(2)
ω2m ), that has coefficients in Z, such that W (λ)
F1 = F2 M(λ). hyperbolic sine function See hyperbolic
function.
Hurwitz’s Theorem If every member of a
normal family (of analytic functions on a con- hyperbolic tangent function See hyperbolic
nected, open, planar set F) is never zero on function.
F, then the limit functions are either identically
hyperbolic transformation A linear frac-
zero or never zero on F.
tional function f (z) = az+b
cz+d where a, b, c, and
A family of analytic functions on a set F is
d are complex constants with ad − bc = 0 has
normal on F if every sequence from the fam-
a pair of (not necessarily distinct) fixed points.
ily contains a subsequence that converges uni-
(See linear fractional function.) This is so be-
formly on compact subsets of F.
cause the equation z = az+bcz+d is linear or qua-
dratic in z. When the two fixed points coin-
Hurwitz zeta function The function
cide, the transformation is said to be parabolic.
∞
When there are distinct fixed points, say α and
1
ζ (s, a) = , 0<a≤1. β, then we can write the transformation in the
(n + a)s w−α
form w−β z−α
= k z−β . When k is real (necessarily
n=0
non-zero) the transformation is said to be hyper-
A generalization of the Riemann zeta function bolic. When |k| = 1, the transformation is said
considered by Hurwitz (1862). to be elliptic.
idele class Members of the idele class group identity map See identity function.
of a global field. See idele.
identity matrix The identity element E in the
idele class group The group of all idele ring of n×n matrices over some coefficient ring.
classes of a global field. See idele. The entries of E are given by the Kronecker
delta-function
idele group The group of all ideles of a global
1 if i = j
field. See idele. δij =
0 if i = j .
integer programming The general linear where X is a set and µ is a measure on X, or,
programming problem asks for the maximum more precisely, where X is a set, B is a σ -field
value of a linear function L of n variables, sub- of subsets of X, and µ is a measure on B. See
ject to linear constraints. In other words, the integral direct sum. Here, l(x) is a unitary rep-
problem is to maximize L(x1 , x2 , . . . , xn ) sub- resentation of G for each x ∈ X. If L! is another
ject to the conditions AX ≤ B, where X is the unitary representation of G, and if L! can also
column vector formed from x1 , . . . , xn , B is a be represented as a direct integral,
column vector, and A is a matrix. The general
integer programming problem is the same, ex- L! = l ! (x) dµ(x) ,
cept the solution (x1 , . . . , xn ) is to consist of X
isomorphism (1) In group theory, a map- Isomorphism Theorem of Class Field Theory
ping φ : G → H between two groups, G and Let k be an algebraic number field. Let I (m) be
H , which is one-to-one (injective), onto (sur- the multiplicative group of all fractional ideals
jective), and which preserves the group opera- of k which are relatively prime to a given integral
tion, that is φ(g1 · g2 ) = φ(g1 ) · φ(g2 ). The divisor m of k. The Galois group of a class field
notion extends to rings, where φ is required to K/k for an ideal group H (m) is isomorphic to
preserve the ring addition and multiplication, to I (m)/H (m). Therefore, every class field K/k
vector spaces, where φ is required to preserve is an Abelian extension of k.
vector addition and scalar multiplication (i.e.,
φ(λ1 v1 + λ2 v2 ) = λ1 φ(v1 ) + λ2 φ(v2 )), and to
isomorphism theorems of groups The three
completely general algebraic contexts (see (2)
standard theorems describing the relationship
below).
between homomorphisms, quotient groups, and
(2) In universal algebra, a mapping φ : G →
normal subgroups. Let G and H be groups, and
H between two (universal) algebras G and H,
let φ : G → H be a homomorphism with ker-
which is one-to-one (injective), onto (surjec-
nel K. (The kernel of φ is the set K = {g ∈
tive), and which preserves the operations of G
G : φ(g) = e}, where e is the group identity el-
and H. In more detail, G = (G, FG ) and H =
ement in H .) The First Isomorphism Theorem
(H, FH ), where G is a set and FG is a set of
states that K is a normal subgroup of G (i.e.,
functions from finite Cartesian products of G
gK = Kg for every g ∈ G), and the quotient
with itself to G (FG is called the set of operations
group (factor group) G/K is isomorphic to the
on G), and similarly for H. Thus the functions
image of φ. Let S and T be subgroups of G, with
in FG are G-valued functions f (g1 , . . . , gn ) of
T normal. The Second Isomorphism Theorem
n G-valued variables, and the value of n may
states that S ∩ T is normal in S, and S/(S ∩ T )
vary with the function f . To be an isomor-
is isomorphic to T S/T . Let K ⊂ H ⊂ G,
phism, φ is required to be a one-to-one and onto
with both K and H normal in G. The Third
function from the set G to the set H , and for
Isomorphism Theorem states that H /K is a nor-
every function f ∈ FG , there must be a func-
mal subgroup of G/K, and (G/K)/(H /K) is
tion h ∈ FH , such that φ(f (g1 , . . . , gn )) =
isomorphic to G/H .
h(φ(g1 ), . . . , φ((gn )), and vice-versa. (This is
what is meant by “preserving the operations of There is an additional theorem which is some-
G and H.”) times called the Fourth Isomorphism Theorem,
In the special case where G and H are groups, but is more commonly called Zassenhaus’s
FG equals the singleton set containing the group Lemma. Let A0 , A1 , B0 , and B1 be subgroups
operation of G (the group multiplication), and of G. Suppose A0 is normal in A1 , and B0
similarly for FH . We thus recapture the moti- is normal in B1 . Zassenhaus’s Lemma states
vating case of group isomorphisms. that A0 (A1 ∩ B0 ) is normal in A0 (A1 ∩ B1 ),
(3) In category theory, a morphism φ : A → B0 (A0 ∩ B1 ) is normal in B0 (A1 ∩ B1 ), and
B between two objects of a category with an A0 (A1 ∩ B1 )/A0 (A1 ∩ B0 ) is isomorphic to
inverse morphism. In other words, for φ to be B0 (A1 ∩ B1 )/B0 (A0 ∩ B1 ). See also factor
an isomorphism, there must also be a morphism group, normal subgroup.
ψ : B → A (the inverse of φ) such that ψ ◦ φ =
ιA and φ ◦ ψ = ιB . Here, ιA and ιB are the isotropic (1) In physics and other sciences,
identity morphisms on A and B, respectively. a material or substance which responds the
The category theoretic definition captures all same way to physical forces in all directions is
of cases (1) and (2) above, and also includes isotropic. Antonym: anisotropic.
J
to the Jacobi rotation method, where the ma-
trices Uk are chosen to be particularly simple
unitary matrices called planar rotation matri-
ces. See Jacobi rotation method.
Jacobian variety The Picard variety of a Jacobi method of finding key matrix A step
smooth, irreducible, projective curve. See Pi- in solving a linear system Ax = b by the linear
card variety. stationary iterative process. If the linear sta-
tionary iterative process is written as x (k+1) =
Jacobi identity The identity (x · y) · z + (y · x (k) + R(b − Ax (k) ), then the Jacobi method
z) · x + (z · x) · y = 0 satisfied by any Lie alge- chooses R to be the inverse of the diagonal sub-
bra. For example, if A is any associative alge- matrix of A. See also linear stationary iterative
bra, and [x, y] denotes the commutator, [x, y] = process.
xy−yx, then the commutator satisfies the Jacobi
identity [[x, y], z]+[[y, z], x]+[[z, x], y] = 0. Jacobi rotation method An iterative method
See Lie algebra. for approximating all of the eigenvalues (char-
acteristic values) of a Hermitian matrix A. The
Jacobi method for solving linear equations method begins by choosing A0 = A, and then
An iterative numerical method, also called the produces a sequence of matrices A1 , A2 , . . . ,
total-step method, for approximating the solu- AN , culminating in a nearly diagonal (all entries
tions to a system of linear equations. In more off the diagonal are small) matrix AN with good
detail, suppose we wish to approximate the solu- approximations to the eigenvalues of A down
tion to the equation Ax = b, where A is an n×n the diagonal. At each step, Ak = Uk∗ Ak−1 Uk ,
square matrix. Write A = L + D + U , where L where Uk is the (unitary) matrix of a planar rota-
is lower triangular, D is diagonal, and U is up- tion annihilating the off diagonal entry of Ak−1
per triangular. The matrix D is easy to invert, so with largest modulus, hence the name rotation
replace the exact equation Dx = −(L+U )x +b method, and Uk∗ is the conjugate transpose of
by the relation Dxk = −(L + U )xk−1 + b, and
(k)
solve for xk in terms of xk−1 : Uk . The matrix Uk = ui,j differs from the
identity matrix only in four entries, u(k) (k)
p,p , uq,q ,
xk = −D −1 (L + U )xk−1 + D −1 b . (k) (k)
up,q , and uq,p . The formulas for these entries
This gives us the core of the Jacobi iteration are particularly simple in the case where the
method. We choose a convenient starting vec- original matrix
A is a real symmetric matrix: If
(k−1) (k) (k)
tor x0 and use the above formula to compute Ak−1 = ai,j , then up,p = uq,q = cos(φ),
successive approximations x1 , x2 , x3 , . . . to the
and u(k) (k)
p,q = −uq,p = sin(φ), where
actual solution x. Under suitable conditions, the
sequence of successive approximations does in- (k−1)
deed converge to x. See iteration matrix. 2ap,q
tan(2φ) = (k−1) (k−1)
,
ap,p − aq,q
Jacobi method of computing eigenvalues
Any of the several iterative methods for approxi- and − π4 ≤ φ ≤ π4 . In the commonly oc-
mating all of the eigenvalues (characteristic val- curring case where the original matrix A has
ues) of a Hermitian matrix A by constructing no repeated eigenvalues, the method converges
a finite sequence of matrices A0 , A1 , . . . , AN , quadratically. The method is named after its
where A0 = A and, for 0 < k ≤ N , Ak = originator, Gustav Jacob Jacobi (1804–1851).
Uk∗ Ak−1 Uk , Uk is a unitary matrix, and Uk∗ de-
notes the conjugate transpose of Uk . The method Jacobi’s inverse problem The problem of
ends with a nearly diagonal matrix AN (all en- inverting Abelian integrals of the first kind on
tries off the diagonal are small) with good ap- a compact Riemann surface R of genus g ≥ 1.
Jacobson radical The set of all elements Janko-Ree group Any member of the family
r in a ring R such that rs is quasi-regular for of all finite simple groups for which the central-
all s ∈ R. In more detail, let R be an arbi- izer of every involution (element of order 2) has
trary ring, possibly non-commutative, possibly the form Z2 × PSL2 (q), q odd. These groups
without a unit element. An element r ∈ R is consist of the Ree groups 2G2 (3n ), for n odd,
quasi-regular if there is an element r ∈ R such and the Janko group J1 . See Janko groups, Ree
that r + r + rr = 0. (In the special case group.
where R has a unit element 1, this is equiva-
lent to (1 + r)(1 + r ) = 1.) For example, every Jensen measure (1) A positive measure µ
nilpotent element (r n = 0 for some n) is quasi- on the closure of an open subset ' of Cn (C the
regular, but there are often other quasi-regular complex numbers) such that for x ∈ ',
elements. The Jacobson radical of R is the set
log |f (x)| ≤ log |f (t)| dµ(t) ,
J of all elements r ∈ R such that rs is quasi-
regular for all s ∈ R. for all f belonging to some appropriate class of
The Jacobson radical is a two-sided ideal, and holomorphic functions (such as the holomorphic
it generalizes the notion of the ordinary radical functions on ' with continuous extensions to the
(the set of all nilpotent elements) of a commuta- closure of '). Jensen measures are named after
tive ring, though even in the special case where Jensen’s inequality, which states that normal-
R is commutative, the Jacobson radical often ized Lebesgue measure on the unit circle (1/2π
differs from the ordinary radical. Both derive times arclength measure) is a Jensen measure.
much of their importance from the following In this case, ' is taken to be the open unit disk
theorem: Let R be commutative. Then (i.) R is {z ∈ C : |z| < 1} in the complex plane.
isomorphic to a subring of a direct sum of fields (2) More generally, a positive measure µ on
if and only if its radical vanishes, and (ii.) R the maximal ideal space M of a commutative
is isomorphic to a subdirect sum of fields if and Banach algebra A is called a Jensen measure
only if its Jacobson radical vanishes. See also for an element , ∈ M if
radical, subdirect sum of rings, Wedderburn’s
Theorem. log |,(f )| ≤ log |t (f )| dµ(t)
j-algebra A concept which reduces the study for all f ∈ A. See also Banach algebra, Jensen’s
of homogeneous bounded domains to algebraic inequality, maximal ideal space.
problems. Let G be a Lie algebra over R, H a
subalgebra of G, (j ) a collection of linear endo- Jensen’s inequality (1) In complex variable
morphisms of G, and ω a linear form on G. The theory. Let f be holomorphic on a neighbor-
K
The theorem remains true if restated using
the strong operator topology instead of the weak
operator topology. Sometimes, the statement of
the theorem includes the following additional
information. The set of self-adjoint elements of
K3 surface A class of algebraic surface in A1 is strongly dense in the set of self-adjoint
abstract algebraic geometry, defined in a pro- elements of M1 , the set of positive elements of
jective space over an algebraically closed field. A1 is strongly dense in M1 , and if A contains 1,
In projective 3-space they can be regarded as de- the unitary group of A is strongly dense in the
formations of quartic surfaces. A K3 surface is unitary group of M.
characterized as a nonsingular, nonrational sur-
face, in several ways including: k-compact group A connected algebraic
(i.) irregularity, Kodaira dimension, and the group defined over a perfect field k whose k-
canonical divisor are zero; Borel subgroups are reduced to the identity
(ii.) irregularity is zero and the arithmetic, group. The name k-anisotropic group is used
geometric, and first plurigenus are all one; also. See also k-isotropic group.
(iii.) as a compact complex analytic surface,
the first Chern class is zero and it has Betti num- k-complete scheme Let f : X −→ Y be a
bers b0 = 1, b1 = 0, b2 = 22, b3 = 0, b4 = 1. morphism of schemes X, Y . When f has a prop-
The space of one-dimensional differential erty, it is customary to say that X has the prop-
forms on a K3 surface is zero. An example of a erty over Y , or that X is a Y -(property) scheme.
K3 surface is any smooth surface of order four The property of being complete is connected
in projective three-dimensional space. with the property of being proper. A morphism
K3 surfaces were early examples of surfaces f : X −→ Y is proper if it is separated, of fi-
satisfying Weil’s conjecture concerning the ana- nite type, and is universally closed. Then X is
log of the Riemann Hypothesis for algebraic va- called proper over Y . See separated morphism,
rieties. morphism of finite type.
Now, let k be an algebraically closed field.
Kakeya-Eneström Theorem Let f be a Let X be a scheme of finite type over k which is
polynomial with real coefficients, say reduced (i.e., for any element x the local ring at
x has no nilpotent elements) and is irreducible
f (x) = an x n + an−1 x n−1 + · · · + a0 , (i.e., the underlying topological space is not the
union of proper closed subsets). If X is proper
for each real number x. Suppose
over k (actually over the spectrum of k), then X
an ≥ an−1 ≥ · · · ≥ a0 > 0 . is called a k-complete scheme.
Let r be any root of the polynomial. Then |r| ≤ kernel (1) In algebra, where a homomor-
1. phism f is defined between two algebraic sys-
tems A and B, if the group identity of B is de-
Kaplansky’s Density Theorem A funda- noted by e, then the kernel of f is
mental theorem from the theory of von Neu- ker(f ) = {x ∈ A : f (x) = e} .
mann algebras proved by Kaplansky in 1951.
The closure M with respect to the weak oper- Alternately, ker(f ) may be denoted f −1 ({e}).
ator topology of a C ∗ -subalgebra A of the set of The kernel is a subset of A that usually has
bounded linear operators on a separable Hilbert special properties. If A and B are groups, then
space is a von Neumann algebra. Furthermore, the kernel of a homomorphism is a normal sub-
if A1 is the set of elements of A with norm ≤ 1 group of A. If A and B are R-modules over a
in A (unit ball of A) and M1 is the set of ele- ring R, the kernel is a submodule of A. If A and
ments of M with norm ≤ 1 (unit ball of M), B are topological linear spaces, then the kernel
β
l−1
→ Tor1 (Hm (R) ⊗ Hq (L)) → 0
fm (t) = i m−1 t i , m+q=n−1
i=1
is exact.
for m > 1, and Bernoulli number Bn . (2) The next Künneth Theorem exhibits the
isomorphism. Let A be a ring with identity. Let
Kummer surface A class of K3 surface L be a complex of left A-modules and R be
in abstract algebraic geometry first studied by a complex of right A-modules. Assume that
E. Kummer in 1864. It is a quartic surface which for each n the homology modules Hn (R) and
has the maximum number (16) of double points cycle modules Cn (R) are projective (in this case
possible for a quartic surface in projective three- Tor1 (Hm (R) ⊗ Hq (L)) = 0). Then for each n
dimensional space. It can be described as the there is an isomorphism α such that the sequence
quotient variety of a two-dimensional Abelian
variety by the automorphism subgroup gener- Hm (R) ⊗ Hq (L) ∼
= Hn (R ⊗ L) .
ated by the sign-change automorphism s(x) = m+q=n
−x. See also K3 surface.
In projective three-dimensional space, the (3) Let G be an Abelian group. For simplicial
surface given by complexes L and R and Cartesian product L×R,
the homology group Hn (L × R; G) splits into
x 4 + y 4 + z4 + w 4 = 0 the direct sum
is a Kummer surface. Hn (L × R; G) ∼ = Hm (L) ⊗ Hq (R)
m+q=n
Künneth’s formula See Künneth Theorem.
⊕ Tor Hm (L) ⊗ Hq (R) .
Künneth Theorem The theorem is formu- m+q=n−1
lated for several different areas of mathemat- (4) Similar Künneth formulas and Künneth
ics. Occurring in the statement of the theorem Theorems generalized (e.g., to spectral se-
are one or more formulas concerning exact se- quences) or stated with other criteria on
quences known as Künneth formulas. The theo- the groups (e.g., torsion free groups) have been
rem itself is sometimes called Künneth formula. studied.
All Künneth-type formulas are related to study-
ing the theory of products (e.g., tensor products) k-Weyl group See k-rank.
in homology and cohomology for various math-
ematical objects and structures. Examples fol-
low.
left Noetherian ring A ring having the prop- is exact, where X is a positive chain complex
erty that (considered as a left module over itself) of A-modules Xk , then the homomorphisms ∂k
every nonempty set of submodules (meaning left have the property that the composition ∂k ∂k+1 =
ideals in this context) has a maximal element. 0 for all k, and < : X → M is a sequence
For such a ring, any ascending chain of left ide- of homomorphisms <k : Xk → M such that
als is finite. Compare with left Artinian ring. <k−1 ∂k = 0 for all k, then X is called a left reso-
For a left Noetherian ring, the prime radical lution of M. This concept is used in homology
is the largest nilpotent ideal. See largest nilpo- theory in the computing of extension groups.
tent ideal. For a left Noetherian ring, any left
left satellite Let R1 and R be rings. Let
module is Artinian if and only if it is Noethe-
CR1 and CR be categories of R1 -modules and R-
rian.
modules, respectively. Let T : CR1 → CR be an
Noetherian rings are named after A.E. additive functor. A left satellite of T is an ad-
Noether (1882–1935). ditive functor defined over the same categories
as T and having the same variance (contra- or
left order Let g be an integral domain in co-). Once one left satellite of T (to be denoted
which every ideal is uniquely decomposed into by S1 T ) is determined, a sequence {Sn T } of left
a product of principal ideals (i.e., a Dedekind satellites may be defined by iteration.
domain). Let F be the field of quotients of g. In order to define the functor S1 T by describ-
Let A be a separable algebra of finite degree ing its action on modules and homomorphisms,
over F . Let L be a g-lattice of A. Then the set some background notation and remarks on ho-
{x ∈ A : xL ⊂ L} is an order of A called a left mology theory need to be presented. Left satel-
order of A. See order. See also ZG-lattice. lites may be defined for covariant functors and
contravariant functors. Here, the procedure for
left projective resolution A left resolution creating a left satellite from a covariant functor
X of an A-module M (where A is a ring with will be described in some detail, with the cor-
unity) such that each Xn in the exact sequence responding procedure for contravariant functors
only sketched. There is also a parallel develop-
· · · → Xn → Xn−1 → · · · → X0 → M → 0 ment for right satellites.
Let A be an R1 -module. From homology
theory, it is always possible to construct an exact
is a projective A-module, and is called a left pro- sequence
jective resolution of M. See also left resolution.
α β
0→M→P →A→0
left regular representation Let R be a ring,
M a left R-module, and HomZ (M, M) the ring with P projective. For what follows, since the
of module endomorphisms (where M is viewed module S1 T (A) (to be defined) is only unique
as a module over Z). For each r ∈ R, let ρr : within isomorphism (with respect to the choice
M −→ M be a left translation (i.e., ρr (x) = rx of P ), a particular choice of exact sequence
for all x ∈ M). If the ring homomorphism needs to be prescribed (which can always be
ρ : R → HomZ (M, M), given by ρ(r) = ρr for done).
all r ∈ M is an injection, then ρ is called a left (i.) Suppose T is covariant. Let A ∈ CR1 .
regular representation of R in H omZ (M, M). Let T (α) : T (M) → T (P ). Define S1 T (A) to
Left (and right) regular representations are very be Ker(T (α)). Then the sequence
important tools in ring theory with many appli-
cations throughout the theory. 0 → S1 T → T (M) → T (P )
less than or equal to See greater than. Levi subgroup Let G be a Lie group. Let
RadG denote the radical of G. A Lie subgroup
level structure A notion occurring in the L of G is called a Levi subgroup of G if (i.) the
study of modulii spaces of Abelian varieties. identity imbedding of L into G is a Lie group
Usually a letter intrinsic to the variety discussed homomorphism, (ii.) G = (RadG) L, (iii.) the
is appended to the term. dimension of (RadG) ∩ L is zero. If the Lie
In such a circumstance, the initial value problem Lie group A group G, together with the struc-
dx ture of a differentiable manifold over the real or
= f (t, x), x(a) = c complex number field, such that the functions
dt
f : G × G → G and g : G → G defined
has a unique solution for all c ∈ K n . by f (x, y) = xy and g(x) = x −1 are differen-
tiable.
Lie algebra A theory named after the Norwe- The additive groups of the fields of real or
gian mathematician M.S. Lie (1842–1899) who complex numbers are Lie groups. The group of
pioneered the study of what are now called Lie invertible n×n matrices over the real or complex
Groups. number fields is a Lie group.
Over a commutative ring with unit K, a Lie In addition to providing a thriving branch of
algebra is a K-module together with a K-module mathematics, Lie groups are used in many mod-
homomorphism x ⊗ y → [x, y] of L ⊗ L into ern physical theories including that of gravita-
L such that, for all x, y, z ∈ L, tion and quantum mechanics.
Lie algebra of Lie group There are three Lie’s Theorem There are several theorems
Lie algebras that are called a Lie algebra of a which bear the name and honor the Norwe-
Lie group G. (See Lie algebra.) gian mathematician M.S. Lie (1842–1899) who,
(1) The Lie algebra of all left invariant deriva- among other things, founded the theory of Lie
tions on G with the bracket product [D1 , D2 ] groups. For solvable Lie algebras, the follow-
defined for two derivations D! , D2 as D1 D2 − ing version of Lie’s Theorem is considered an
D2 D1 . important tool.
(2) The Lie algebra of all left invariant vector The irreducible representations of a finite di-
fields on G with the bracket product of two left mensional, solvable Lie algebra over an algebra-
invariant vector fields X and Y defined using the ically closed field of characteristic zero all are
natural isomorphism between the vector space one-dimensional.
of left invariant vector fields and the vector space
of left derivations of G. Lie subalgebra Let K be a commutative ring
(3) The Lie algebra of all tangent vectors to with unit. A Lie subalgebra is a subalgebra of
G at e with the bracket product between tangent a Lie algebra (meaning a K-submodule that is
vectors defined using the natural isomorphism closed under the bracket product of the Lie al-
between the space of left invariant vector fields gebra). A Lie subalgebra is a Lie algebra. See
of G and the space of tangent vectors of G at e. Lie algebra.
minimal basis A basis for a vector space V minimal Weierstrass equation A Weier-
with the property that, if any element is removed strass equation for an elliptic curve E/K is one
from the basis, it no longer forms a basis (i.e., of the form
some element of V cannot be expressed as a
y 2 + a1 xy + a3 y = x 3 + a2 x 2 + a4 x + a6 .
finite linear combination of the smaller set). See
basis. We call such an equation minimal if, among all
Often such minimality is part of the definition possible Weierstrass equations, it has least dis-
of basis. criminant |D|.
module of cocycles The dual of the module module of finite presentation A module M
of cycles. More precisely, if X0 , X1 , X2 , . . . over a ring R such that there is a positive in-
is a graded chain of cycles over a ring R (often teger n and an exact sequence of R-modules
the integers or the reals) with boundary maps 0 → K → R n → M → 0 where K is finitely
∂n : Xn → Xn−1 , then the module of cocycles generated.
is the graded chain X̂0 , X̂1 , X̂2 , . . . where each
X̂n is the set of all ring homomorphisms from Xn module
of finite type (1) A graded module
to R. The induced coboundary map ∂ˆn : X̂n → i≥0 A i over a field k for which each Ai is finite
X̂n+1 is defined by ∂ˆn ĉn (xn+1 ) = ĉn (∂n xn+1 ) dimensional.
for ĉn ∈ X̂n and xn+1 ∈ Xn+1 . (2) A sheaf of O-modules (called an O-
Module) in a ringed space (X, O), which is lo-
module of cycles A concept that arises in the cally generated by a finite number of sections
subject of graded modules. Regard the graded over O.
module X as a sequence of modules X0 , X1 , X2 , (3) A finitely generated module.
. . . with maps ∂n : Xn → Xn−1 (called bound-
ary maps) with ∂n−1 ◦ ∂n = 0. A cycle cn ∈ Xn module of homomorphisms Let M and N
is one that has zero boundary (∂n cn = 0). The be modules over a commutative ring R. The set
set of all cycles is a module over the underlying of all homomorphisms of module M to module
ring (often the integers). This concept arises N is called the module of homomorphisms of
most commonly in topology where the graded M to N , with addition defined pointwise and
modules are chains of simplicies in a topological multiplication given by (r · f )(x) = r · (f (x)),
space X. In this case, Xk is the free module gen- for r ∈ R, f : M → N a homomorphism
erated (say, over the integers) by the continuous and x ∈ M. The module of homomorphisms is
images of the standard k-dimensional simplex denoted HomR (M, N ).
in Euclidean space
k
module with operator domain A module
M, together with a set A and a map from A × M
k
(x1 , . . . , xk ) ∈ R : xi ≥ 0, xi ≤ 1
into M satisfying the following conditions. (i.)
i=1
For every a ∈ A and x ∈ M there is a unique
(into X). The boundary map of a chain is a element ax ∈ M. (ii.) If a ∈ A and x, y ∈ M,
weighted sum of (the continuous images of) its then a(x + y) = ax + ay. In this situation, M
k − 1 dimensional faces where the weights are is called a module with operator domain A. It
either plus one or minus one depending on the is also called a module over A or an A-module.
Mordell’s conjecture Any algebraic curve morphism of local ringed spaces Let X be
of genus ≥ 2 defined over an algebraic number a topological space, and suppose that to each
field k of finite degree has finitely many rational x ∈ X is associated a local ring BX,x in a natural
points. (Conjectured in 1922 and settled in the way. Let Y be another such space. A mapping
affirmative by G. Faltings in 1983. The original f : X → Y is called a morphism of local ringed
1922 conjecture was stated for the special case spaces if, whenever, f (x) = y, there is induced
k = Q.) a homomorphism BY,y → BX,x .
multiplicity of weight Let g denote a com- mutually associated diagrams An O(n) di-
plex semisimple Lie algebra, h a Cartan subal- agram is a Young diagram T for which the sum
gebra of g, and R the corresponding root system. of the lengths of the first column and the sec-
Let V be a g-module (not necessarily finite di- ond column is ≤ n. Two O(n) diagrams T1
mensional), and let w ∈ h∗ a linear form on h. and T2 are called mutually associated diagrams
We will let V n denote the set of all v ∈ V such if the lengths of their first columns sum up to
that H v = w(H )v, for all H ∈ h. This is a n and their corresponding columns (except the
vector subspace of V . An element of V w is said first ones) have equal lengths. See Young dia-
to have weight w. The dimension of V w is called gram.
N
The addition satisfies the following laws:
A1 Closure Law: n + m ∈ N
A2 Commutative Law: n + m = m + n
A3 Associative Law: m+(n+p) = (m+n)+p
A4 Cancellation Law: If m + p = n + p, then
Nakai-Moishezon criterion For a ringed m=n
space (X, O), we let I denote a coherent sheaf Multiplication on N is defined by
of ideals of O. When X is a k-complete scheme (i.) n · 1 = n
(where k is a field), then for any r-dimensional (ii.) n·m∗ = n·m+n whenever n·m is defined.
closed subvariety W of X and any invertible M1 Closure Law: n · m ∈ N
sheaf S we let (S r · W ) denote the intersection M2 Commutative Law: n · m = m · n
number of S r with O/I. The Nakai-Moishezon M3 Associative Law: m · (n · p) = (m · n) · p
criterion states that if (S r · W ) > 0 for any r- M4 Cancellation Law: If m · p = n · p, then
dimensional closed subvariety W of X, then S m=n
is ample.
Addition and multiplication are subject to the
Naperian logarithm See natural logarithm. Distributive Law
D1 For all n, m, p ∈ N, m·(n+p) = m·n+m·p
natural logarithm The logarithm in the base The elements of N are called natural num-
e. See e. The natural logarithm of x is denoted bers.
log x or, in elementary textbooks, ln x.
The adjective natural is used, due to the sim- natural positive cone For a weight w on
plicity of the defining formula the positive elements of a von Neumann algebra
x A, we set Nw = {A ∈ A : w(A∗ A) < ∞}.
log x =
1
dt , Let 0 ≤ s ≤ 21 . Then the closure W s of the
1 t set of vectors sw η(A), where A ranges over
all positive elements in Nw ∩ Nw∗ , has some
for x real and positive.
of the properties of the von Neumann algebra
Also called Naperian logarithm. See also
A. (Here, w is a modular operator and η is
logarithm, logarithmic function.
a complex linear mapping on Nw defined by
setting (η(A◦ ), η(A)) = w(A∗ A).) The special
natural number A positive integer. The sys- 1
tem N of natural numbers was developed by the case W 4 is called the natural positive cone. It
Italian mathematician Peano, using a few sim- is a self-dual convex cone and is independent of
ple properties known as Peano Postulates. Let the weight w.
there exist a non-empty set N such that
Postulate I: 1 is an element in N. negation Given a proposition P , the propo-
Postulate II: For every n ∈ N, there exists a sition (not P ) is called the negation of P and is
unique n∗ ∈ N , called the successor of n. denoted by ∼ P or ¬P .
Postulate III: 1 is not the successor of any ele-
ment in N. negative See negative element.
Postulate IV: If n, m ∈ N and n∗ = m∗ , then
n = m. negative angle An angle XOP , with ver-
Postulate V: Any subset K ⊂ N having the prop- tex O, initial side OP and terminal side OX,
erties such that the direction of rotation is counter-
(i.) 1 is an element of K clockwise.
(ii.) k ∗ ∈ K whenever k ∈ K
satisfies K = N. negative cochain complex A cochain com-
Addition on N is defined by: plex C in an Abelian category such that C n = 0
(i.) n + 1 = n∗ , for every n ∈ N for n > 0. See cochain complex.
nilpotent group A group G such that Cn (G) Noetherian local ring A Noetherian ring
= G, for some n, where 1 ⊂ C1 (G) ⊂ C2 (G) having a unique maximal ideal.
⊂ · · · is the ascending central series of G. See
ascending central series. Noetherian module A left module M, over
a ring R such that, for every ascending chain
nilpotent ideal A (left, right, two-sided) ideal M1 ⊂ M2 ⊂ · · · of submodules of M, there
I of a ring R is nil if every element of I is nilpo- exists an integer p such that Mk = Mp , for all
tent; I is a nilpotent ideal if I n = 0 for some k ≥ p.
integer n.
Noetherian ring A ring R is left (resp. right)
nilpotent Lie algebra A Lie algebra such
Noetherian if R is Noetherian as a left (resp.
that there exists a number M such that any com-
right) module over itself. See Noetherian mod-
mutator of order M is zero.
ule. R is said to be Noetherian if R is both left
and right Noetherian.
nilpotent Lie group A connected Lie group
whose Lie algebra is a nilpotent Lie algebra.
Noetherian scheme A locally Noetherian
nilpotent matrix A square matrix A such scheme whose underlying topological space is
that Ap = 0 where p is a positive integer. If p compact. A scheme X is a locally Noetherian
is the least positive integer for which Ap = 0, scheme if it has an affine covering {Ui }, where
then A is said to be nilpotent of index p. each Ui is the spectrum of some Noetherian ring
Ri .
nilpotent radical Let G be a Lie algebra.
The radical of G is the union S of all solvable Noetherian semilocal ring A Noetherian
ideals of G; it is itself a solvable ideal of G. The ring with only a finite number of maximal
largest nilpotent ideal of G is the union N of all ideals. See Noetherian ring. See also Noethe-
nilpotent ideals of G. The ideal I = [S, G] is rian local ring.
K of degree n which is one of the valuations Cholesky method to this normal equation is one
vi , 1 ≤ i ≤ n, defined as follows: There are way of solving the linear least squares problem.
exactly n mutually distinct injections ϕ1 , . . . , ϕn See Cholesky method of factorization.
of K into the complex number field C. We may
assume that ϕi (K) ⊂ R if and only if 1 ≤ i ≤ r1 normal extension An extension field K of
and that ϕi (a) and ϕn−i+1+r1 (a) are complex a field F , such that K is a finite extension of F
conjugates for r1 ≤ i ≤ r2 = (n − r1 )/2. Let and F is the fixed field of G(K, F ), the group
vi (a) = |ϕi (a)| for 1 ≤ i ≤ r1 and vi (a) = of automorphism of K relative to F .
|ϕi (a)|2 for r1 ≤ i ≤ r2 = (n − r1 )/2. See
Archimedian valuation. normal form A canonical choice of repre-
sentatives for equivalence classes with respect
normal basis Let L be a Galois extension of to some group action. See also Jordan normal
a field K and G(L/K) the Galois group of L/K. form.
(See Galois extension, Galois group.) If L/K is
a finite Galois extension, then there is an element normal function A continuous, strictly
l ∈ L such that the set {l g : g ∈ G(L/K)} forms monotone function of the ordinal numbers.
a basis for L over K called a normal basis. See
also Normal Basis Theorem. Normalization Theorem for Finitely Gener-
The existence of a normal basis implies that ated Rings Let R be a finitely generated ring
the regular representation of G(L/K) is equiv- over an integral domain I. Then there exist an
alent to the K-linear representation of G(L/K) element α ∈ I (with α = 0) and algebraically
by means of L. independent elements Y1 , . . . , Yn of R over I
such that the ring of quotients RS is integral
Normal Basis Theorem Any (finite dimen- over I[α −1 , Y1 , . . . , Yn ]. Here, S = {α i : i =
sional) Galois extension field L/K that has a 1, 2, . . . }.
normal basis.
Normalization Theorem for Polynomial
normal chain of subgroups A normal chain Rings Let I be an ideal in a polynomial ring
of subgroups of a group G is a finite sequence K[X1 , X2 , . . . , Xn ] in n variables over a field
G = G0 ⊃ G1 ⊃ · · · ⊃ Gn = {e} of subgroups K. Then there exist Y1 , Y2 , . . . , Yn in K[X1 ,
of G with the property that each Gi is a normal X2 , . . . , Xn ] such that (i.) Y1 , Y2 , . . . , Yn gen-
subgroup of Gi−1 for 1 ≤ i ≤ n. erate I ∩K[Y1 , Y2 , . . . , Yn ] and (ii.) K[X1 , X2 ,
. . . , Xn ] is integral over K[Y1 , Y2 , . . . , Yn ].
normal crossings Let D ≥ 0 be a divisor
(i.e., a cycle of codimension 1) on a non-singular normalized cochain A cochain f ,
variety V and let x ∈ V . Then D is a divisor in Hochschild cohomology, satisfying
with only normal crossings at x if there is a sys- f (ł1 , . . . , łn ) = 0 whenever one of the łi
tem of local coordinates (f1 , . . . , fn ) around x is 1. Here, : is an algebra over a commuta-
d
S1 and S2 are bounded linear operators and :
=
(j ) (j )
c 1 x1 + · · · + c n xn is multiplication by a sequence {λn ≥ 0} in l1 .
Then the nuclear norm of T is given by T 1 =
j =1
inf S1 :l1 S2 . When X and Y are Hilbert
is a form of degree d with rational coefficients. spaces, this norm coincides with the trace norm.
(j )
Here, ci denotes a conjugate of ci , 1 ≤ i ≤ n. (2) A locally convex topological vector space
Z is called a nuclear space if for each abso-
normic form A normic form of order p ≥ 0 lutely convex neighborhood U of 0 there is an-
in a field k is a homogeneous polynomial f of other absolutely convex neighborhood V ⊂ U
degree d ≥ 1 in n = pd variables with coeffi- of 0 such that the natural linear mapping τV ,U :
cients in k such that the equation f = 0 has no Z(V ) −→ Z(U ) is a nuclear operator. Here,
solution in k except (0, . . . , 0). Z(U ) and Z(V ) are the normed spaces obtained
from seminorms corresponding to U and V , re-
norm-residue symbol Let k be a finite ex- spectively.
tension of Qp and
null sequence Any sequence tending to 0 (in
X = k{{t1 }} . . . {{tn−1 }} a topological space with a 0 element).
Especially, one has a null sequence in the I-
be an n-dimensional local field. See local field. adic topology. Let I be an ideal of a ring R
Assume that k contains the roots of unity µ of and let M be an R-module. A null sequence
order q = p ν , p = 2. Let ζ be a generator
in M is a sequence which converges to zero
of µ. According to Vostokov, there is a skew-
symmetric map in the I-adic topology of M. A base for the
neighborhood system of zero in this topology is
H : X∗ × · · · × X∗ → µ given by {I n M : n = 1, 2, . . . }.
H(α1 , α2 , . . . , αn+1 ) = ζ tr res(φ/s) , αi ∈ X∗ ,
null space A synonym for the elements of a
with the property that αi + αj = 1 ⇐⇒ H(α1 , matrix, when the matrix is considered as a linear
α2 , . . . , αn+1 ) = 1 for i = j . Here tr is the transformation.
trace operator of the inertia subfield of k, s is
determined in X by an expansion of ζ in X, and number (1) Any member of the real or com-
φ(α1 , α2 , . . . , αn+1 ) is given by an expansion plex numbers.
of the αs in X. Here res is the residue of φ/s, (2) We say that two sets have the same num-
i.e., the coefficient of 1/t1 t2 · · · tn−1 . ber if there is a bijection between them. This is
number field Any subfield of the field of numerical range The numerical range (or
complex numbers. field of values) of an n × n matrix A with entries
from the complex field is the set
number of irregularity The dimension of
the Picard variety of an irreducible algebraic va- F (A) = x ∗ Ax : x ∈ Cn , x ∗ x = 1 .
riety.
The numerical range is a compact, convex set
number system Any one of the fields Q, R, that contains the eigenvalues of A, is invari-
or C (rational numbers, real numbers, or com- ant under unitary equivalence and is subadditive
plex numbers). (i.e., F (A + B) ⊂ F (A) + F (B)). The numer-
ical radius of A is defined in terms of F (A) by
A
numerator The quantity A, in the fraction B
(B is called the denominator). r(A) = max{|z| : z ∈ F (A)} .
order of function The order of an analytic orthogonal matrix An element O of the or-
function f (z) at point a is the exponent of the thogonal group O(n). See orthogonal group.
lowest power in the Taylor Series of f (z) at a.
orthogonal measure A measure concentrat-
order of polynomial Let ed on a set at measure 0.
Pn (x) = an x n + an−1 x n−1 + · · · + a1 x + a0 orthogonal subset Let S be a subset of a
vector (Hilbert) space V. The orthogonal subset
be a polynomial with complex coefficients and S ⊥ of S is the subset
an = 0. The number n is the order of the poly-
nomial Pn (x). S ⊥ = {
v ∈ V : v · w
= 0 for all w
∈ S} .
P
= ∂p,q−1
◦ ∂p,q
+ ∂p−1,q
◦ ∂p,q =0.
We define the associated chain complex (Xn , ∂)
p-adic numbers The completion of the ratio- by setting
nal field Q with respect to the p-adic valuation
Xn = Xp,q , ∂n = ∂p,q + ∂p,q .
|·|p . See p-adic valuation. See also completion.
p+q=n p+q=n
for all x ∈ V and for some ω ∈ V . The number permutation representation A permutation
ω is called a period of f (x), and f (x) with a representation of a group G is a homomorphism
period ω = 0 is call a periodic function. from G to the group SX of all permutations of a
set X. The most common example is when X =
period relation Conditions on an n×n matrix G and the permutation of G obtained from g ∈
which help determine when a complex torus is G is given by x → gx (or x → xg, depending
an Abelian manifold. In Cn , let * be generated on whether a product of permutations is read
by (1, 0, . . . , 0), (0, 1, 0, . . . , 0), . . . , (0, 0, . . . , right-to-left or left-to-right).
0, 1), (a11 , a12 , . . . , a1n ), (a21 , a22 , . . . , a2n ),
. . . (an1 , an2 , . . . , ann ). Then Cn / * is an Peron-Frobenius Theorem See Frobenius
Abelian manifold if there are integers d1 , d2 , . . . , Theorem on Non-Negative Matrices.
dn = 0 such that, if A = (aij ) and D = (δij di ),
then (i.) AD is symmetric; and (ii.) (AD) is Perron’s Theorem of Positive Matrices If
positive symmetric. Conditions (i.) and (ii.) are A is a positive n × n matrix, A has a positive
the period relations. real eigenvalue λ with the following properties:
(i.) λ is a simple root of the characteristic equa-
permanent Given an m×n matrix A = (aij ) tion.
with m ≤ n, the permanent of A is defined by (ii.) λ has a positive eigenvector u .
(iii.) If µ is any other eigenvalue of A, then
permA = a1i1 a2i2 . . . amim , |µ| < λ.
where the summation is taken over all m- Peter-Weyl theory Let G be a compact Lie
permutations (i1 , i2 , . . . , im ) of the set {1, 2, group and let C(G) be the commutative asso-
. . . , n}. When A is a square matrix, the per- ciative algebra of all complex valued continuous
manent therefore has an expansion similar to functions defined on G. The multiplicative law
that of the determinant, except that the factor defined on C(G) is just the usual composition
place value The value given to a digit, de- Poincaré Let R be a commutative ring with
pending on that digit’s position in relation to the unit. Let U be an orientation over R of a com-
units place. For example, in 239.71, 9 repre- pact n-manifold X with boundary. Then for all
sents 9 units, 3 represents 30 units, 2 represents indices q and R-modules G there is an isomor-
7
200 units, 7 represents 10 units and 1 represents phism
1
100 units.
γU : Hq (X; G) ≈ H n−q (X; G) .
Plancherel formula Let G be a unimodular
locally compact group and Ĝ be its quasidual. This is called Poincaré-Lefschetz duality. The
Let U be a unitary representation of G and U ∗ analogous result for a manifold X without bound-
be its adjoint. For any f , g ∈ L1 (G) ∩ L2 (G), ary is called Poincaré duality.
the Plancherel formula
Poincaré-Birkhoff-Witt Theorem Let G be
f (x)g(x) dx = t (Ug∗ (ξ )Uf (ξ )) dµ(ξ ) a Lie algebra over a number field K. Let X1 , . . . ,
G Ĝ Xn be a basis of G, and let R = K[Y1 , . . . , Yn ]
be a polynomial ring on K in n indeterminates
holds, where Uf (ξ ) = Ĝ f (x)Ux (ξ )dx. The Y1 , . . . , Yn . Then there exists a unique alge-
measure µ is called the Plancherel measure. bra homomorphism ψ : R → G such that
ψ(1) = 1 and ω(Yj ) = Xj , j = 1, . . . , n.
plane trigonometry Plane trigonometry is
Moreover, ψ is bijective, and the j th homoge-
related to the study of triangles, which were
neous component Rj is mapped by ψ onto G j .
studied long ago by the Babylonians and an-
Thus, the set of monomials {X1k1 X2k2 . . . Xnkn },
cient Greeks. The word trigonometry is derived
k1 , . . . , kn ≥ 0, forms a basis of U (G) over
from the Greek word for “the measurement of
K. This is the so-called Poincaré-Birkhoff-Witt
triangles.” Today trigonometry and trigonomet-
Theorem. Here U (G) = T (G)/J is the quotient
ric functions are indispensable tools not only in
associative algebra of G where J is the two-sided
mathematics, but also in many practical appli-
ideal of T (G) generated by all elements of the
cations, especially those involving oscillations
form X ⊗ Y − Y ⊗ X − [X, Y ] and T (G) is the
and rotations.
tensor algebra over G.
Plücker formulas Let m be the class, n the
degree, and δ, χ , i, and τ be the number of Poincaré differential invariant Let w =
nodes, cups, inflections, tangents, and bitan- α(z − z◦ )/(1 − z◦ z) with |α| = 1 and |z◦ | <
gents. Then 1, be a conformal mapping of |z| < 1 onto
|w| < 1. Then the quantity |dw|/(1 − |w|2 ) =
n(n − 1) = m + 2δ + 3χ |dz|/(1 − |z|2 ) is called Poincaré’s differen-
tial invariant. The disk {|z| < 1} becomes
m(m − 1) = n + 2τ + 3i
a non-Euclidean space using any metric with
3n(n − 2) = i + 6δ + 8χ ds = |dz|/(1 − |z|2 ).
3m(m − 2) = χ + 6τ + 8i
3(m − n) = i − χ . Poincaré duality Any theorem general-
izing the following: Let M be a com-
pact n-dimensional manifold without bound-
plurigenera For an algebraic surface S with a ary. Then, for each p, there is an isomor-
canonical divisor K of S, the collection of num- phism H p (M; Z2 ) ∼
= Hn−p (M; Z2 ). If, in
is called the Poincaré metric for the unit disc in z = x + iy = r(cos θ + i sin θ)
the complex plane. where r = x 2 + y 2 and θ = tan−1 yx .
pointed co-algebra Let V be a co-algebra. pole divisor Suppose X is a smooth affine va-
A nonzero subco-algebra W of V is said to riety of dimension r and suppose Y ⊂ X is a sub-
positive number A real number greater than power of a complex number Let z = x +
zero. iy = r(cos θ + i sin θ) be a complex number
with r = x 2 + y 2 and θ = tan−1 xy . Let n be
positive root Let S be a basis of a root system a positive number. The nth power of z will be
φ in a vector space the complex number r n (cos nθ + i sin nθ ).
V such that each root β can be
written as β = a∈S ma a, where the integers
ma have the same sign. Then β is a positive root power-residue symbol Let n be a positive
if all ma ≥ 0. integer and let K be an algebraic number field
containing the nth roots of unity. Let α ∈ K ×
positive semidefinite matrix An n×n matrix and let ℘ be a prime ideal of the ring such that
A such that, for all u ∈ Rn , we have ℘ is relatively prime to n and α. The nth power
is a positive integer and let K be an algebraic
(A(
u), u) ≥ 0 . number field containing the nth roots of unity.
such that ρU U = identity and such that ρU W = prime divisor For an integer n, a prime di-
ρU V ρV W whenever U ⊆ V ⊆ W . The col- visor is a prime that occurs in the prime factor-
lection of Abelian groups along with the homo- ization of n. For an algebraic number field or
morphisms ρU V is called a presheaf of Abelian for an algebraic function field of one variable (a
prime ideal Let R be a commutative ring primitive hypercubic set A finite subgroup
with identity 1 and let I = R be an ideal of R. K of the orthogonal group O(V ) is called fully
Then I is prime if whenever a, b ∈ R are such transitive if there is a set S = {e1 , . . . , es } that
that ab ∈ I , then at least one of a and b is in I . spans V on which K acts transitively and K has
no invariant subspace in V . In this case, one can
prime number A positive integer p is said choose S as either
to be prime if (i.) the primitive hypercubic type:
(i.) p > 1,
(ii.) p has no positive divisors except 1 and p. S = {e1 , . . . , en } , ei , ej = δij ;
The first few prime numbers are 2, 3, 5, 7, 11,
13, 17. or
(ii.) the primitive hyperbolic type:
prime
rational divisor A divisor p =
ni Pi on X over k satisfying the following S = {f1 , . . . , fn+1 } ,
three conditions: (i.) p is invariant under any au- 1, i=1,...n+1,i=j
tomorphism σ of k̄/k; (ii.) for any j , there exists (fi , fj ) = .
σ − n1 , i,j =1,...,n+1,i=j
an automorphism σj of k̄/k such that Pj = P1 j ;
(iii.) n1 = · · · = nt = [k(P1 ) : k]i , where X is primitive ideal Let R be a Banach algebra.
a nonsingular irreducible complete curve, k is a A two-sided ideal I of R is primitive if there is a
subfield of the universal domain K such that X regular maximal left ideal J such that I is the set
is defined over k. Prime rational divisors gen- of elements r ∈ R with rR ⊆ I . The regularity
erate a subgroup of the group of divisors G(X), of J means that there is an element u ∈ R such
which is called a group of k-rational divisors. that r − ru ∈ J for all r ∈ R.
principal idele Let K be an algebraic number principal value (1) The principal values of
field. The multiplicative group K × injects di- arcsin, arccos, and arctan are the inverse func-
agonally into the group IK of ideles. The image tions of the functions sin x, cos x, and tan x, re-
is called the set of principal ideles. stricted to the domains − π2 ≤ x ≤ π2 , 0 ≤ x ≤
π , and − π2 < x < π2 , respectively. See arc sine,
principal matrix Suppose A = [Aij ] is an arc cosine, arc tangent.
n × n matrix. The principal matrices associated (2) Let f (x) have a singularity at x = c,
with A are A(k) = [Aij ], 1 ≤ i, j ≤ k ≤ n. with a ≤ c ≤ b. The Cauchy principal value of
b
a f (x) dx is
principal minor See principal submatrix.
c−F b
principal order Let K be a finite extension lim f (x) dx + f (x) dx .
F→0 a c+F
of the rational field Q. The ring of all algebraic
integers in K is called the principal order of K. The Cauchy
∞ principal value of an improper
c
integral −∞ f (x)dx is limc→∞ −c f (x)dx.
principal root A root with largest real part (if
this root is unique) of the characteristic equation principle of counting constants Let X and
of a differential-difference equation. Y be algebraic varieties and let C be an ir-
reducible subvariety of X × Y . Let pX and
principal series For a semisimple Lie group, pY denote the projection maps onto the fac-
those unitary representations induced from finite tors of X × Y . Let a1 = dim(pX (C)) and
dimensional unitary representations of a mini- a2 = dim(pY (C)). There exist a nonempty
mal parabolic subgroup. open subset U1 of pX (C), contained in pX (C),
and a nonempty open subset U2 of pY (C)), con-
principal solution A solution F (x) of the tained in pY (C), such that all irreducible com-
equation SF (x)/Sx = g(x), where SF (x) = ponents of C(x) = {y ∈ Y : (x, y ∈ C} have
F (x + Sx) − F (x). Such a solution F (x) can the same dimension b2 for all x ∈ U1 and such
be obtained by a formula in terms of integral, that all irreducible components of C −1 (y) =
series, and limits. {x ∈ X : (x, y) ∈ C} have the same dimension
b2 for all y ∈ U2 . These dimensions satisfy
principal submatrix A submatrix of an m×n a1 + b2 = a2 + b1 .
matrix A is an (m − k) × (n − >) matrix obtained
from A by deleting certain k rows (k < m) and principle of reflection Two complex num-
> columns (> < n) of A. If m = n and if the set bers z1 and z2 are said to be symmetric with
of deleted rows coincides with the set of deleted respect to a circle of radius r and center z0 if
columns, we call the submatrix obtained a prin- (z1 − z0 )(z2 − z0 ) = r 2 . The principle of
(3) Let A1 and A2 be objects in a category C. projection matrix A square matrix M such
A triple (P , π1 , π2 ) is called the product of A1 that M 2 = M.
and A2 if P is an object of C, πi : P → Ai is
a morphism for i = 1, 2, and if whenever X is projective algebraic variety Let K be a
another object with morphisms fi : X → Ai , field, let K̄ be its algebraic closure, and let Pn (K̄)
for i = 1, 2, then there is a unique morphism be n-dimensional projective space over K̄. Let S
f : X → P such that πi f = fi for i = 1, 2. be a set of homogeneous polynomials in X0 , . . . ,
Xn . The set of common zeros Z of S in Pn (K̄)
(4) See also bracket product, cap product, is called a projective algebraic variety. Some-
crossed product, cup product, direct product, times, the definition also requires the set Z to be
Euler product, free product, Kronecker prod- irreducible, in the sense that it is not the union
uct, matrix multiplication, partial product, ten- of two proper subvarieties.
sor product, torsion product, wedge product.
projective class Let A be a category. A pro-
product complex Let C1 be a complex of jective class is a class P of objects in A such
right modules over a ring R and let C2 be a that for each A ∈ A there is a P ∈ P and a
complex of left R-modules. The tensor product P-epimorphism f : P → A.
C1 ⊗R C2 gives a complex of Abelian groups,
projective class group Consider left mod-
called the product complex.
ules over a ring R with 1. Two finitely gener-
ated projective modules P1 and P2 are said to
product double chain complex The dou- be equivalent if there are finitely generated free
ble chain complex (Zp,q , ∂ , ∂ ) obtained from modules F1 and F2 such that P1 ⊕F1 , P2 ⊕F2 .
a chain complex X of right A-modules with The set of equivalence classes, with the opera-
boundary operator ∂p and a chain complex Y tion induced from direct sums, forms a group
of left A-modules with boundary operator ∂q in called the projective class group of R.
projective space Let K be a field and con- proper fraction A positive rational number
sider the set of (n + 1)-tuples (x0 , . . . , xn ) in such that the numerator is less than the denom-
K n+1 with at least one coordinate nonzero. Two inator. See also improper fraction.
tuples (. . . , xi , . . . ) and (. . . , xi , . . . ) are equiv-
alent if there exists λ ∈ K × such that xi = proper intersection Let Y and Z be subva-
λxi for all i. The set Pn (K) of all equivalence rieties of an algebraic variety X. If every irre-
classes is called n-dimensional projective space ducible component of Y ∩ Z has codimension
over K. It can be identified with the set of equal to codimY +codimZ, then Y and Z are said
lines through the origin in K n+1 . The equiv- to intersect properly.
alence class of (x0 , x1 , . . . , xn ) is often denoted
(x0 : x1 : · · · : xn ). More generally, let R be a proper Lorentz group The group formed
commutative ring with 1. The scheme Pn (R) is by the Lorentz transformations whose matrices
given as the set of homogeneous prime ideals of have determinants greater than zero.
R[X0 , . . . , Xn ] other than (X0 , . . . , Xn ), with a
structure sheaf defined in terms of homogeneous proper morphism of schemes A morphism
rational functions of degree 0. It is also possible of schemes f : X → Y such that f is sepa-
to define projective space Pn (S) for a scheme rated and of finite type and such that for every
S by patching together the projective spaces for morphism T → Y of schemes, the induced mor-
appropriate rings. phism X ×T Y → T takes closed sets to closed
sets.
projective special linear group The quotient
group defined as the group of matrices (of a fixed proper orthogonal matrix An orthogonal
size) of determinant 1 modulo the subgroup of matrix with determinant +1. See orthogonal
scalar matrices of determinant 1. matrix.
Prüfer ring An integral domain R such that purely inseparable extension An extension
all finitely generated ideals in R are inversible L/K of fields such that every element of L is
in the field of quotients of R. See also integral purely inseparable over K. See purely insepa-
domain. rable element.
R
group.
(5) The radical of a Lie algebra is the largest
solvable ideal.
rational map (1) Informally, a rational func- rational representation Let R be a commu-
tion. See rational function. tative ring with 1 and let G be a subgroup of
(2) In algebraic geometry, a function from an GLn (R) for some n. Let ρ : G → GLm (R)
open subset U of a variety X to another variety for some m be a homomorphism. Suppose
Y , which preserves regular functions. (Regu- there exist m2 rational functions in n2 vari-
lar functions are functions which can be rep- ables pst ({Xij }) ∈ R[{Xij }] with ρ((gij )) =
resented locally as quotients of polynomials.) (pst (guv )) for all (gij ) ∈ G. Then ρ is called a
In more detail, let X and Y be varieties, and rational representation.
rational variety A variety over a field k real Lie algebra A Lie algebra over the field
which, for some n, is birationally equivalent to of real numbers. See Lie algebra.
an n-dimensional projective space over k. See
also birational mapping, variety. real number The real numbers, denoted R,
form a field with binary operations of + (addi-
R-basis Let R be a ring contained in a pos- tion) and · (multiplication). This field is also an
sibly larger ring S. A left S module V has an ordered field: There exists a subset P ⊂ R such
R-basis if V is freely generated as an R mod- that (i.) if a, b ∈ P then a + b ∈ P and a · b ∈ P
ule by a family of generators, vα , α ∈ V . This and (ii.) the sets P , −P = {x ∈ R : −x ∈ P } ,
means (i.) each vα ∈ V ; (ii.) if a finite sum {0} (0 denotes the additive identity) are pairwise
r1 vα1 + · · · + rn vαn = 0, where r1 , . . . rn ∈ R, disjoint and their union is R. Furthermore, the
then ri = 0 for i = 1, . . . n; (iii.) the set of all fi- field R satisfies the completeness axiom: If S
nite sums r1 vα1 + · · · + rn vαn , where r1 , . . . rn ∈ is a non-empty subset of R and if S is bounded
R, is equal to V . The family vα , α ∈ A, is called above, then S has a least upper bound in R. With
an R-basis for V . these axioms, R is known as a complete ordered
The most important case of this occurs when field. The existence of such a field is taken as
R is chosen to be the ring of integers, Z, in which an assumption by many. Others prefer to postu-
case an R basis is called a Z-basis. See also Z- late the existence of the natural numbers N ⊂ R
basis. by means of the Peano postulates. See natu-
ral number. With these postulated, the integers,
real axis in complex plane A complex num- Z are constructed from N and then the rational
ber p + iq, p, q real, can be identified with the numbers Q are constructed as the field of quo-
point (p, q) in Cartesian coordinates. Under tients of Z. To construct the real numbers from
this identification, numbers of the form p + i0, the rationals requires more work. One way is to
that is, real numbers, are identified with points use the method of Dedekind cuts and another in-
of the form (p, 0). The collection of all such volves viewing the reals as equivalence classes
points forms one of the axes in this coordinate of Cauchy sequences. See also field.
system; this axis is called the real axis.
real part of complex number If z = a + bi,
real closed field A real field F such that any where a and b are real, is a complex number, the
algebraic extension of F which is real must be real part of z is a. This is denoted as (z) = a.
equal to F . That is, F is maximal with respect
to the property of reality in an algebraic closure. real prime divisor Let K be an algebraic
See real field. number field of degree n; then there are n injec-
tions σ1 , . . . , σn of K into C. We may assume
real field A field F such that −1 is not a sum that these are ordered so that n = r1 + 2r2 , and
of squares in F . so that for 1 ≤ j ≤ r1 , σi maps to the real num-
regular function (1) In complex analysis, an regular module Let R be a commutative ring
analytic or holomorphic function. See analytic with 1 and M an R-module. An element x ∈ M
function, holomorphic function. is said to be regular if, for every r ∈ R, there
(2) In algebraic geometry, a function which exists t ∈ R such that rx = rtrx. A submodule
can be represented locally as a quotient of poly- N of M is regular if each element of N is regular.
nomials. In more detail, let X be a variety. (A In particular, M is regular if each of its elements
variety is the solution set of a system of poly- is regular.
nomial equations.) Let V be an open subset of
X. Let f : V → k be a function from V to a regular module Let R be a commutative ring
field k. The function f is regular at the point with 1 and M an R-module. An element x ∈ M
p ∈ V if there is an open neighborhood U of is said to be regular if for every r ∈ R, there
p, contained in V , on which f = g/ h, where exists t ∈ R such that rx = rtrx. A submodule
g and h are polynomials and h = 0 on U . f is N of M is regular if each element of N is regular.
regular if it is regular at each point of its domain In particular, M is regular if each of its elements
V. is regular.
regular ideal A left ideal J in a ring R, for regular polyhedral group The group of mo-
which there exists an e ∈ R such that r −re ∈ J tions which preserve a regular polygon in the
for every r ∈ R. Similarly, a right ideal J , for plane, or the group of motions in three dimen-
which there exists an e ∈ R such that r −er ∈ R sional space that preserve a regular polyhedron.
for every r ∈ R. Note that if R has an identity, The group of motions preserving a regular n-gon
then all ideals are regular. in the plane is called the dihedral group Dn ; it
has order 2n. The alternating group A4 can be
regular local ring Let R be a local ring and realized as the group of motions preserving a
set k = R/m. Then k is a field. A local Noethe- tetrahedron, and A5 can be realized as the group
rian ring is called regular if dimk m/m2 = of motions preserving an iscosahedron; they are
dim R. (Here dim is the Krull dimension. See called the tetrahedral and iscosahedral groups,
Krull dimension.) See also local ring. respectively. Similarly, the symmetric group S4
is called the octahedral group because it can be
regular mapping (1) A differentiable map realized as the group of motions preserving an
with nonvanishing Jacobian determinant be- octahedron.
tween two manifolds of the same dimension.
(2) A mapping given by an n-tuple of regular regular prime A prime number which is not
functions from a variety to a variety contained irregular. See irregular prime.
in affine n-space. See regular function.
regular representation (1) Let K be a com-
regular matrix (1) A Markov chain is a prob- mutative ring with identity and let A be a K alge-
ability model for describing a system that ran- bra. An element a ∈ A can then be viewed as an
domly moves among different states over suc- element of the algebra of K endomorphisms on
cessive periods of time. For such a process, A via x " → a x for all x ∈ A. Such a representa-
relative invariant (1) Let K be a field and let remainder (1) If a is an integer and b is a
F be an extension field of K. Let G be a group positive integer, then there is a unique pair of
of automorphisms of F over K, and suppose that integers q, r with 0 ≤ r < b such that a =
K is the fixed field of the group G. A relative qb + r. The integer q is called the quotient and
invariant of G in F is a nonzero element Q ∈ r is called the remainder.
F such that, for each σ ∈ G, there exists an
(2) More generally, the terminology can be
element χ (σ ) in K for which σ (Q) = χ (σ )Q.
applied in any Euclidean domain; here, by def-
(2) In algebraic geometry, a quantity invari- inition, there is such a division algorithm. This
ant under birational transformation. can also be applied in polynomial rings: If
R is an integral domain and if a(x), b(x) ∈
relatively ample Let (X, OX ) be a scheme. R[X], then there exists q(x), r(x) ∈ R[X] with
Let F be a sheaf of O-modules and construct a deg r(x) < deg b(x) or r(x) = 0 such that
projective bundle P (F ) on X. A locally closed a(x) = b(x)q(x) + r(x). Again, q(x) is called
S-subscheme f : X → S of p : P (F ) → S the quotient of the division of a(x) by b(x) and
is called quasiprojective over S. For such a r(x) is called the remainder.
quasiprojective scheme, an invertible sheaf L on
X is called ample over S if there exist a locally
Remainder Theorem Suppose R is a ring
free OS -module of finite type on S and an im-
with identity, for example, the integers. Suppose
mersion ι : X → P (F ) such that OX (1) = L.
that f (x) = an x n +an−1 x n−1 +· · ·+a1 x +a0 ∈
If L⊗n is ample for some n, L is said to be rel-
R[x], that is, f (x) is a polynomial with coeffi-
atively ample over S.
cients in R. The Remainder Theorem states that
relatively minimal See relatively minimal for any c ∈ R there exists a unique q(x) ∈ R[x]
model. such that f (x) = q(x)(x − c) + f (c).
relatively minimal variety See relatively replica Let V be a vector space over a field
minimal model. K and V ∗ its dual. For A in EndK V (the K-
endomorphisms of V ) define A∗ ∈ EndK (V ∗ )
relatively prime numbers Two integers via A∗ u(x) = u(Ax) for every x ∈ V and
which have no common factor other than 1 or u ∈ V ∗ . Let Vsr denote the tensor product of
-1. s copies of V and r copies of V ∗ . Define Ars on
Vsr by
relative norm Let K and k be algebraic num- Ars (x1 ⊗ · · · ⊗ xr ⊗ u1 ⊗ · · · ⊗ ur ) =
s
ber fields with k ⊂ K and set n = [K : k]. Then i=1 x1 ⊗ · · · ⊗ Axi ⊗ · · · ⊗ xs ⊗ u1 ⊗ · · · ⊗ ur
there are n k-linear embeddings φi : K → C. − rj =1 x1 ⊗· · ·⊗xs ⊗u1 ⊗· · ·⊗A∗ uj ⊗· · ·⊗ur .
For an ideal U of K, U (i) = {φi (u) : u ∈ U } is An element B ∈ EndK (V ) is called a replica of
an ideal of φi (K). Let L be the field generated A if the nullspace of Ars is a subspace of the
by φi (K), i = 1, . . . , n. The ideal in L gener- nullspace of Bsr for every r and s.
G is called the direct product of the groups Gi . Here the first m rows are formed with the coeffi-
The set of all elements {gi }i∈G of G such that cients ai and the last n rows are formed with the
gi = e for all but a finite number of i forms a coefficients bi . The two polynomial equations
right annihilator Let S be a subset of a ring right ideal A nonempty subset I of a ring R
R. The set {r ∈ R : Sr = 0} is called the right such that, whenever x and y are in I and r is in
annihilator of S. Note that the right annihilator R, x − y and xr are in I . See also left ideal,
of S is a right ideal of R and is an ideal of R if ideal.
S is a right ideal.
right injective resolution See right resolu-
right Artinian ring A ring that satisfies the tion.
descending chain condition on right ideals; that
is, for every chain I1 ⊃ I2 ⊃ I3 ⊃ . . . of right right invariant (1) Let G be a locally com-
ideals of the ring, there is an integer m such that pact topological group. A Borel measure µ on
Ii = Im for all i ≥ m. See also right Noethe- G is said to be right invariant if µ(Ea) = µ(E)
rian ring, left Artinian ring, left Noetherian ring, for every a ∈ G and Borel set E ⊂ G.
Artinian ring, Noetherian ring. (2) Let G be a Lie group. An element g ∈ G
defines an operator of right translation by g on
right-balanced functor Let T be a functor of G: Rg (x) = xg for x ∈ G. If T is a tensor field
several variables, some covariant and some con- on G and if Rg T = T for every g ∈ G (where
travariant, from the category of modules over a Rg T is defined in the natural way), then T is
ring R to the category of modules over a ring S. said to be right invariant.
T is said to be right-balanced if (i.) when any
of the covariant variables of T is replaced by an right inverse element Suppose S is a non-
injective module, T becomes an exact functor in empty set with an associative binary operation
the remaining variables, and (ii.) when any one ∗. Assume that there is a right identity element
of the contravariant variables of T is replaced by e ∈ S, with respect to ∗. If a ∈ S, a right inverse
a projective module, T becomes an exact functor element for a is an element b ∈ S such that
of the remaining variables.
a∗b =e.
right coset Any set, denoted by Sa, of all
right multiples sa of the elements s of a sub- If every element of such an S has an inverse
group S of a group G and a fixed element a of element, then S is called a group. See also left
G. See left coset. inverse element.
right derived functor See left derived func- right Noetherian ring A ring which satisfies
tor. the ascending chain condition on right ideals.
That is, for every chain I1 ⊂ I2 ⊂ I3 ⊂ . . .
right global dimension For a ring A, the of right ideals there exists an integer m such
smallest n ≥ 0 for which A has the property that Ii = Im for every i ≥ m. See also right
that every right A-module has projective dimen- Artinian ring, left Noetherian ring, left Artinian
sion ≤ n. This is equal to the smallest n ≥ 0 for ring, Noetherian ring, Artinian ring.
which each right A-module has injective dimen-
sion ≤ n. See projective dimension, injective right order Let g be an integral domain in
dimension. See also left global dimension. which every ideal is uniquely decomposed into
right satellite Let A be an Abelian category ringed space A topological space X, together
and let R be a selective Abelian category. If with a sheaf of rings OX on X. The main import
A has enough proper injectives, then a P con- of this statement is that to each open set U of X,
nected pair (T , E∗ , S) of covariant functors is there is associated a ring OX (U ). Example: Let
right universal if and only if each proper short X be an open subset of Cn , or, more generally, a
exact sequence in A, 0 → C → J → K → 0, complex analytic manifold. For each open sub-
with J proper injective, induces a right exact set U , let OX (U ) = the ring of analytic (holo-
sequence T (J ) → T (K) → S(C) → 0 in R. morphic) functions defined on U . Then the pair
Given T , the S with this property is uniquely (X, OX ) is an important example of a ringed
determined. It is called the right satellite of T . space. The sheaf OX is called the Oka sheaf.
See also left satellite. See sheaf.
row nullity For A an m×n matrix, the dimen- ruled surface A surface S that contains, for
sion of the row null space (the space of solutions each point p in S, a line that contains p. Some
described by Ax = 0). If A is of rank r, the row simple examples of ruled surfaces are planes and
nullity is equal to m − r. hyperboloids of one sheet.
is the L-function of E over Q, p = 0 or ±1, 1− scheme A ringed space X for which there
ap u+pu2 = p1 (u, E mod p) = 1−ap u+pu2 = is an affine scheme Spec(A) such that Spec(A)
√
(1 − πp u)(1 − π̄p u) with πp = peiθp (0 < and X are locally isomorphic.
θp < π ). When E has complex multiplication,
the distribution of θp for half of p is uniform in Schmidt’s Theorem If a field F is perfect,
[0, π ], and θp is π2 for the remaining half of p. then every extension E/F is separable.
isfying µi (Ci ) = C−i , µi (Ei ) = I−j , µi (Ii ) = The inequality is commonly generalized in
E−i . Further, µ−i is defined to be µ−1 i ,i = abstract inner product spaces to mean |(x, y)| ≤
1, . . . , k. Then the group generated by µ1 , . . . , x y for vectors x and y, where (·, ·) is the in-
µk is called a Schottky group. See also Möbius ner product and · is its norm. Also known as
transformation. the Cauchy-Schwarz or Bunyakovskiǐ inequal-
ity, it is a special case of the Hölder integral
Schreier conjecture A conjecture which inequality. See Hölder’s Theorem.
states that for an arbitrary finite simple group,
G, the outer automorphisms of G form a solv- scientific notation The convention in applied
able group. The conjecture is true for all known mathematics wherein real numbers are repre-
finite simple groups. sented as decimals (between one and ten) mul-
tiplied by powers of ten. For example, 36000 is
Schur index Let G be a finite group, F ⊂ S 3.6×104 and 0.00031 is 3.1×10−4 in scientific
where S is a splitting field for G, and ϕ ∈ notation. The notation is particularly useful in
IrrS (G). If , is an irreducible S-representation dealing with numbers that are very large or very
which affords ϕ and - is an irreducible F - small in magnitude.
representation such that , is a constituent of
-S , then the multiplicity of , as a constituent secant function One of the fundamental
of -S is called the Schur index of ϕ over F . See trigonometric functions, denoted sec x. By
also splitting field, irreducible representation, definition, sec x = cos1 x , and hence the se-
constituent. cant function is (i.) periodic, satisfying sec(x +
2π ) = sec x; (ii.) bounded below, satisfying
Schur product See Hadamard product. 1 ≤ | sec x|, for all real x, and (iii.) undefined
where the cosine is 0 (x = ± π2 , ± 3π 2 , . . . ).
Schur’s Lemma (1) Let M and N be sim- Many other properties of the secant function can
ple A-modules, and let f : M → N be an A- be derived from those of the cosine and sine. See
homomorphism. Then f is either an isomor- cosine function. See also secant of an angle.
phism or f is the zero homomorphism.
(2) Let A be a ring and let M be a simple A- secant method The iterative sequence of ap-
module. Let D = EA (M) be the endomorphism proximate solutions of the equation f (x) = 0,
ring of M. Then D is a skew field. given by
xn − xn+1
Schur subgroup The subgroup of the Brauer xn+1 = xn − f (xn ) ·
f (xn ) − f (xn+1 )
group Br(k) consisting of those algebra classes
that contain a Schur algebra A over k, where k for f (xn ) = f (xn+1 ), n ≥ 1.
is a field of characteristic 0.
secant of an angle Written secα, the recip-
Schur-Zassenhaus Theorem Suppose N is rocal of the x-coordinate of the point where the
a normal subgroup of a finite group G such that terminal ray of the angle α whose initial ray lies
|N | and |G : N| are relatively prime. Then G along the positive x-axis intersects the unit cir-
contains subgroups of order |G : N| and any two cle. If 0 < α < π2 (α in radians) so that the
of these subgroups are conjugate in G. angle is one of the angles in a right triangle with
See also normal subgroup, conjugate sub- adjacent side a, opposite side b, and hypotenuse
group. c, then sec α = ac .
Selberg zeta function The zeta function semidirect product A group G which can be
∞
written as the product of a normal subgroup N of
Z4 (s, M) = det I − M (γi ) N (γi )−s−n , G and a subgroup H of G where N ∩ H = {1}.
i n=0
where 4 ⊂ SL(2, R) is a Fuchsian group, op- semifinite function A measurable function
erating on the complex upper half-plane H, P1 , f with the property that {x : |f (x)| = ∞} is σ -
exist for all multi-indices (α1 , . . . , αk ) ∈ (Z+ )k semiprime ring A ring R is semiprime if 0
such that |α| = α1 + · · · + αk ≤ n. Under these is the only nilpotent ideal.
conditions,
φv (u) = semiprimitive ring A ring in which the rad-
i α1 +···+αk ical is the set containing only the zero element.
m(α1 ,...,αk ) uα1 · · · uαk +o |u|n , See radical.
v k
α1 ! · · · αk ! 1
|α|≤n
semireductive action A rational action de-
where |u| = |u1 |+· · ·+|uk | and for sufficiently fined by ρ such that N = f1 K + · · · fr K is
small |u|, the principal value of log φv (u) can be a G-admissible module (f1 , . . . , fr ∈ R) and
expressed by Taylor’s expansion as that if f0 mod N (f0 ∈ R) is G-invariant, then
there is a homogeneous form h in f0 , . . . , fr
log φv (u) = of positive degree with coefficients in K such
i α1 +···+αk that h is monic in f0 and is G-invariant, where
a (α1 ,...,αk ) uα1 1 · · · uαk k + o |u|n K is a commutative ring with identity and G is
α1 ! · · · αk ! v a matrix group over K (i.e., a subgroup of the
|α|≤n
general linear group GL(n, K)). A homomor-
(α ,...,α )
where the coefficients av 1 k
are called phism ρ of G into GL(m, K) is called a ratio-
(mixed) semi-invariants, or cumulants, of order nal representation of G if there exist rational
α = (α1 , . . . , αk ) of the vector v = (v1 , . . . , vk ). functions φkl , 1 ≤ k, l ≤ m, in n2 variables xij
in addition, A is closed under countable unions, is an example of an L-matrix. Such matrices first
then it is called a σ -field. arose in the study of sign-solvable systems in
economics models. A sign-nonsingular matrix
signature The ordered pair (p, q) corre- A is a square L-matrix; that is, every matrix in
sponding to a real quadratic form Q of rank Q(A) is nonsingular.
p + q, where Q is equivalent to
p q+p similar (1) (Similar figures) Geometric fig-
xi2 − xj2 . ures such that the ratio of the distance between
i=1 j =p+1
pairs of points in one figure and the distance be-
tween the corresponding pairs of points in an-
See also signature of Hermitian form. other figure is constant for every pair of points in
the first figure and corresponding pair of points
signature of Hermitian form The ordered in the second figure.
pair (p, q) corresponding to a Hermitian form (2) (Similar terms) Terms in an expression
H of rank p + q, where H is equivalent to which have the same unknowns and each un-
p q+p
known is raised to the same power in each term.
For example, in the expression 5x 2 y 3 + 8x 3 +
x i xi − x j xj .
3y 3 +7x 2 y 3 , 5x 2 y 3 and 7x 2 y 3 are similar terms.
i=1 j =p+1
(3) (Similar triangles) Triangles which have
proportional corresponding sides.
signed numbers Numbers that are either pos-
itive or negative are said to be signed or directed similar decimals Numbers that have the same
numbers. number of decimal places are said to be similar.
For example, 6.003, 2.232, and 100.000 are all
significant digits In a real number in decimal similar decimals.
form, the digits to the left of the decimal point
beginning with first non-zero digit, together with similar fractions Simple fractions in which
the digits to the right of the decimal point ending the denominators are equal. See simple fraction,
with the last non-zero digit. denominator.
simple algebra An algebra A such that the simple root A root of an equation that is
ring A is both a simple ring and a semisimple not repeated, i.e., a root of multiplicity 1. See
ring. multiplicity of root.
simple co-algebra A co-algebra which has simplest alternating polynomial The poly-
no nontrival proper subco-algebras. See coalge- nomial p(X1 , . . . , Xn ) = (X2 − X1 )(X3 − X2 )
bra. . . . (Xn − Xn−1 ). See alternating polynomial.
simple component (1) Suppose R is a ring simplex (1) In geometry, the convex hull of
written as R = R1 ⊗· · ·⊗Rn where R1 , . . . , Rn an affine independent set of points, the vertices
are ideals satisfying the condition that if Ri = of the simplex.
Ri ⊗Ri , where Ri and Ri are ideals, then either (2) In topology, a homeomorphic image of a
Ri = 0 or Ri = 0. Then the ideals R1 , . . . , Rn geometric simplex (as in (1)).
are called the simple components of R.
(2) Suppose P is a polynomial
in n variables simplex method A method for optimizing a
over a field F . Write P = ni=1 Piαi where the basic feasible solution of a primary linear pro-
Pi are the distinct, irreducible (over F ) factors gramming problem in a finite number of itera-
of P . Pj is called a simple component of P if tions. Let I be the set of basic variables and J be
αj = 1. the set of non-basic variables. Let x be a basic
a2 x2 + · · · + an xn ≤ b may be converted to a
linear equality by adding a nonnegative variable solution Anything that satisfies a given set
z, called a slack variable. Namely, a1 x1 +a2 x2 + of constraints is called a solution to that set of
· · · + an xn ≤ b becomes a1 x1 + a2 x2 + · · · + constraints. A number may be a solution to an
an xn + z = b. See also linear programming. equation or a problem. A function may be a
solvable ideal An ideal in which the derived spatial ∗-isomorphism Suppose that Hi is
series becomes zero. a Hilbert space and that Mi is a von Neumann
algebra contained in the set of bounded linear
solvable Lie group A Lie group G, which, operators on Hi , for i = 1, 2, and consider a
ignoring its Lie group structure and considering ∗-isomorphism π : M1 → M2 . If there exists
it only as an abstract group, is solvable. See a bijective isometric linear mapping U : H1 →
solvable group. H2 such that U AU ∗ = π A, for each A in M1
(with U ∗ denoting the adjoint of U ), then π is
solving a triangle The act of using the size called spatial.
of certain sides and angles of a triangle to deter-
mine the remaining sides and angles. See also spatial tensor product The C ∗ -algebra re-
solution of oblique spherical triangle, solution sulting when the algebraic tensor product of two
of plane triangle, solution of right spherical tri- C ∗ algebras A and B is equipped with the spa-
angle. tial, or minimal, C ∗ cross-norm,
space (1) A set of objects, usually called xmin = sup (ρ ⊗ η)(x)
ρ,η
points. Often equipped with some additional
structure, e.g., a topological space, if equipped where x ∈ A ⊗ B, and the supremum runs over
with a topology, or a metric space, if equipped all ∗-representations of A and B. There may
with a metric. be more than one norm on the algebraic tensor
(2) The unbounded three-dimensional region product A ⊗ B whose completion gives a C ∗ -
R3 in which all physical objects exist. algebra.
space coordinates A set of three numbers special Clifford group If R is a ring with
sufficient to specify uniquely the location of a identity, M a free R-module, and Q a quadratic
stable vector bundle A vector bundle E with standard position of angle An angle with its
the property that for any proper subbundle E of vertex at the origin and its initial ray lying on the
E, the inequality positive x-axis. The magnitude of the angle is
then measured in a counter-clockwise direction
rank(E) · degree(E ) < rank(E ) · degree(E)
with respect to the side lying on the x-axis.
holds.
star of element (1) (Of an element A of a
standard complex Suppose that R is a com- complex 2) The set of all elements of 2 that
mutative ring, and that g is a Lie algebra over contain A.
R that is R-free. Denote byU the enveloping (2) (Of an element x of a Banach algebra
algebra of g, and denote by R (g) the exterior B with involution ∗) The image of x under ∗
R-module g. The U -free reso-
algebra of the (denoted x ∗ ).
lution U ⊗ R (g) of R is called the standard
complex of g. state (1) A positive linear functional of norm
See enveloping algebra, exterior algebra. 1 on a C ∗ -algebra.
(2) The state of a system of N particles in
standard form of difference equation If y classical mechanics is described by the instan-
is a function from the integers to the real or com- taneous position and velocity of each of the par-
plex numbers, and 2 is the finite difference op- ticles. If these particles are moving in three-
erator defined by dimensional Euclidean space, the state is then
described by a point in R6N . Given Newton’s
(2y)(x) = y(x + 1) − y(x) ,
laws, the knowledge of the state of a system at
a difference equation is an equation relating y, any given time allows one to predict its state
and its finite differences, i.e., an equation of the at any time in the future or past. These states
form F (x, y, 2y, . . . , 2k y) = 0. This stan- of the system are sometimes referred to as the
dard form of this equation results by rewriting dynamical or microscopic state of the system.
it in such a way that the finite difference opera- In thermodynamics, the state of a system is
tor does not explicitly occur, i.e., by choosing T often specified by a much smaller number of
so that the equation becomes G(x, y(x), y(x + quantities, such as its density, volume, and tem-
1), . . . , y(x + T)) = 0. perature. In this case the number of quantities
chosen to specify the state of the system are
standard parabolic k-subgroup Suppose the minimum needed to predict its macroscopic
that k is a field, and G is a connected reduc- properties, and these states correspond to a sta-
tive group defined over k. Suppose that S is a tistical average over a very great number of mi-
maximal k-split torus, denote by Z the central- croscopic states of the system.
izer of S in G, and by P a minimal parabolic
k-subgroup of G. Denote by r the set of all k- Steenrod algebra Suppose that p is a prime,
roots of G with respect to S, choose an ordering and denote by Zp the field of order p. Suppose
of r, and suppose that θ is a subset of the funda- that q is a nonnegative integer, and denote by
mental system of r with respect to that ordering. Aq the Abelian group that consists of all stable
See Weyl group, highest weight. Stoker multiplier A concept from the theory
of water waves.
Steinberg type group A connected, semisim-
ple, algebraic group G, defined over a field F , Stokes multiplier With respect to a given
of
formal fundamental solution Y
such that G has a maximal, connected, solvable,
closed subgroup defined over F . (x) = H
Y (x) ×T eQ(x) ,
stereographic projection A particular meth- the Stokes multiplier (or Stokes matrix) Sα ∈
od of associating to each point on a sphere a GL(n, C) corresponding to the singular line α
point in the plane. Let N be the north pole of of 2 at 0 is defined by Yα − = Yα + + Sα on α.
the sphere, and W the equatorial plane. Through
each point, p, of the sphere, draw the line pass- Stone space The space of all uniform ultra-
ing through N . The stereographic projection of filters on ω with the natural topology. Equiva-
p is the point q at which this line passes through lently, the remainder βω \ ω in the Cech-Stone
W. Note that this is a one-to-one (and hence in- compactification of the integers. This construc-
vertible) mapping of the sphere minus the north tion is denoted St(P(ω)/fin), where P(ω)/fin
called the structural constants of g relative to Sturm method of solving algebraic equations
X1 , . . . , Xn . Suppose that f is a real-valued polynomial of
degree d in one variable, and that f has no mul-
structure A description of a mathematical tiple zero. For real numbers a and b with a < b,
object in terms of sets and relations, subject to a denote by N (a, b) the number of real roots of
set of axioms is called a mathematical structure. the equation f = 0 in the interval (a, b). By
For example, a group, defined in terms of a set applying the division algorithm, one may con-
and an operation, subject to the usual axioms for struct polynomials f0 , f1 , . . . , fd and q1 , . . . ,
a group, is a mathematical structure. qd−1 such that f0 = f , fi−1 = fi qi + fi+1
for 1 ≤ i ≤ d − 1, and fd is constant. Denote
structure sheaf (1) Suppose that V is an by V (a) the number of changes in sign of the
affine variety. Denote by AU the ring of regular terms of the finite sequence f0 (a), . . . , fd (a),
functions on the open set U . The sheaf obtained and denote by V (b) the number of changes in
from the assignment of AU to U for each open sign of the terms of the finite sequence f0 (b),
set U is called the structure sheaf of V (or the . . . , fd (b); in each case we ignore zero terms.
sheaf of germs of regular functions on V ). Then N (a, b) = V (a) − V (b). Sturm’s method
(2) Suppose that V is a topological space, F uses this result to approximate the real roots of
is a field, and O is a sheaf of germs of map- the equation f = 0.
pings from V to F . Suppose that there is a finite
open covering U1 , . . . , Un of V such that each Sturm’s Theorem A theorem which relates
pair (Ui , OUi ) is isomorphic to an affine variety. the position of zeros of solutions of two lin-
Then O is called the structure sheaf of the pair ear, second order, ordinary differential equa-
(V , O). tions. Suppose that y1 (x) and y2 (x) are non-
(3) Suppose that X is a topological space, and zero solutions of
O is a sheaf on X of commutative rings, each
with a unit, such that the stalk of O at each point y1 + q(x)y1 = 0
x of X does not equal {0}. Then O is called the y2 + p(x)y2 = 0
structure sheaf of the pair (X, O).
on some interval I . If q(x) > p(x) for all x in
structure space The set of kernels of alge- some interval I , then between any two zeros of
braically irreducible representations of a Banach y2 , there must be a zero of y1 .
algebra.
subalgebra A subset S of an algebra A, with
Structure Theorem for Type-III von Neu- the property that S is itself an algebra with re-
mann Algebras Suppose that M is a von spect to the operations of addition and multipli-
Neumann algebra of type III. Then there exist a cation of A.
von Neumann algebra N of type II∞ , a faithful
normal trace τ , and a one-parameter group G of subbialgebra A subspace of a bialgebra B,
∗-automorphisms θt with τ ◦ θt = e−t τ , such which is both a subalgebra and a subco-algebra
that M is a crossed product of N with G. of B.
subgroup Let G be a group and H a subset substitution The act of replacing one math-
of G. We call H a subgroup of G if H contains ematical expression by another equivalent one.
the identity element, H is closed under multi- For instance, if one wishes to solve the system
sum of vectors See addition of vectors. Suzuki group A finite simple group of order
q 2 (q − 1)(q 2 + 1), where n is a positive integer,
superabundance Suppose that S is a nonsin- and q = 22n+1 .
gular surface, and D is a curve on S. Denote by
O(D) the sheaf of germs of holomorphic cross- sweepable bounded domain A bounded do-
sections of D. The dimension of H 1 (S, O(D)) main D in Cn such that there is a compact subset
is called the superabundance of D. K of D and a subgroup 4 of the group of holo-
morphic automorphisms of D with 4K = D.
supersolvable group A finite group G such
that there are subgroups G0 , . . . , Gn of G such Sylow’s Theorems Suppose that G is a finite
that G0 is trivial, Gi is a normal subgroup of group.
Gi+1 for i = 0, 1, . . . , n − 1, each Gi is normal (1) If m and n are relatively prime and the
in G, and each quotient group Gi+1 /Gi is cyclic order of G is pn m, then G contains a subgroup of
and of prime order. order p i for each i = 1, 2, . . . , n. If 1 ≤ i ≤ n
and S is a subgroup of G of order pi , then S is
supplemented algebra Let R be a commu- a normal subgroup of some subgroup of G of
tative ring with a unit. If A is an R-algebra and order p i+1 .
ε : A → R is an R-algebra homomorphism, (2) If H is a p-subgroup of G and S is a Sylow
then (A, ε) is called a supplemented algebra (or p-subgroup of G, then there is some element g
augmented algebra). of G such that H is a subgroup of gSg −1 . In
particular, every two Sylow p-subgroups of G
are conjugate in G.
support (1) The support of a complex-valued
function is the closure of the set of points at (3) Denote by s the number of distinct Sylow
which the function is nonzero. Equivalently, it is p-subgroups of G. Then k is a factor of the order
the complement of the largest open set on which of G and s ≡ 1(mod p). If H is a p-subgroup
the function vanishes. of G, denote by t the number of distinct Sy-
low p-subgroups of G that contain H . Then
(2) The support of a measure is the comple-
t ≡ 1pmodp.
ment of the largest open set of zero measure.
See Sylow subgroup.
supremum The least upper bound or join of a Sylow subgroup A maximal p-subgroup of
set of elements of a lattice. The term is most fre- a group G, for a prime number p. Also called
quently used with regard to sets of real numbers. Sylow p-subgroup of G.
If S is a set of real numbers, the supremum of S See Sylow’s Theorems.
is the unique real number B = sup S defined by
the following two conditions: (i.) x ≤ B, for all Sylvester’s elimination method A method
x ∈ S; (ii.) if x ≤ C, for all x ∈ S, then B ≤ C. used in elimination theory for obtaining the re-
See also infimum. sultant of two polynomials. Suppose that I is
a unique factorization domain and a0 , . . . , am
surd√ An irrational root of a positive integer, and b0 , . . . , bn are elements
nof I . Put f (x) =
e.g., 3 2, or 52/5 , or a sum of such expressions. m m−i and g(x) = n−i . Define
i=0 a i x i=0 b i x
an m×n matrix C = (cij ) as follows. For i = 1,
surjection A function e from a set X to a set . . . , n and j = i, . . . , i + m, put cij = aj −i .
Y such that e(X) = Y , that is, such that every For i = n + 1, . . . , n + m and j = i − n, . . . ,
element y ∈ Y is the image e(x) of an element i, put cij = bj +n−i . Put cij = 0 for all other
x ∈ X. Surjections are also called surjective appropriate values of i and j . If a0 and b0 have
functions, surjective mappings, and onto func- no common factor, then the determinant of C
tions. The notion of a surjection generalizes to equals the resultant of f and g.
symplectic group The set of all symplectic system of equations A set of equations in the
transformations from a symplectic vector space variables x1 , x2 , . . . , xn , which are all required
(V , ω) to itself form the symplectic group of to be satisfied simultaneously.
(V , ω). If V is R2n , with the standard symplec-
tic structure, the notations Sp(2n) and Sp(2n, R)
are used. See symplectic transformation. system of fundamental solutions (1) Sup-
pose that F is a field, and consider a system S of
symplectic manifold A manifold equipped linear homogeneous equations with coefficients
with a distinguished closed 2-form of maximal in F . Denote by V the vector space over F of
rank. solutions of S. If the dimension of V is pos-
itive, then a basis for V is called a system of
symplectic transformation A linear trans- fundamental solutions of S.
formation T from a symplectic vector space (2) Suppose that D is a division ring, and con-
(V1 , ω1 ) to a symplectic vector space (V2 , ω2 ), sider a system S of linear homogeneous equa-
which preserves the symplectic form; i.e., tions with left coefficients in D. Denote by M
ω2 (T (v1 ) , T (v2 )) = ω1 (v1 , v2 ) the unitary right D-module of solutions of S. If
the rank of M is positive, then a basis for M is
for any vectors v1 and v2 in V1 . called a system of fundamental solutions of S.
T of g −1 under
adI∂, and g λ is the λ-eigenspace of
adE = 2 k zk ∂zk + α iuα ∂u∂ α in gh .
Tate cohomology A variation on the coho- terminating decimal A real number that can
mology of finite groups. It forms a free resolu- be expressed as m + 10pn for integers p, m and
tion of the finite group Z as a ZG module. See n. Rational numbers can always be expressed as
also free resolution. either a terminating or repeating decimal. See
also repeating decimal.
Tate’s conjecture For a projective nonsingu-
lar variety V over a finite algebraic number field term of polynomial An individual summand
K, let U r (V̄ ) denote the group ofalgebraic cy- in the polynomial expression.
cles of codimension r on V̄ = V K C modulo
homological equivalence and let U r (V ) be the tetrahedral group The alternating group of
subgroup of U r (V̄ ) generated by the algebraic degree 4, denoted A4 . It is called the tetrahedral
cycles rational over K. Tate conjectured that the group because it can be realized as the group of
rank of U r (V ) is equal to the order of the pole rigid motions that preserve a tetrahedron.
of the Hasse ζ -function ζ2r (s, V ) at s = r + 1.
theory of equations The study of polynomial
Tate-Shafarevich group The set of elements equations, the most general of which is
of the Weil-Châtelet group that are everywhere
locally trivial. x n + a1 x n−1 + · · · + an−1 x + an = 0 .
Tate’s Theorem Let K/k be a Galois ex- Some of the questions studied in the theory of
tension with Galois group G. Tate’s Theorem equations are the existence of roots, methods
states that the n − 2 cohomology group of G of solutions, and the possibility of solutions by
with coefficients in Z (denoted Ĥ (G, Z)) is iso- radicals.
morphic to the nth cohomology group of G with
coefficients in the idele class group Ck of K (de- theory of moduli Given a class of objects,
noted Ĥ (G, Ck )). The isomorphism is given by such as “projective surfaces of general type” or
6n (α) = ζK/k 7 α, where 7 denotes the cross “rank two vector bundles on a given Riemann
product and ζK/k is the canonical cohomology surface,” one may classify them (up to isomor-
class for K/k. phism, say, or some other similar equivalence
relation) in two steps:
tautological line bundle The subspace of First, identify numerical invariants (invari-
Pn XR n+1 (where Pn is projective space) that ants under the chosen equivalence). Then, for
consists of the pairs (L, x) where L ∈ Pn and the objects having a fixed set of numerical in-
x ∈ L. variants, describe an algebraic variety (or an an-
alytic variety) which parametrizes these. The
tensor product Given two linear spaces, M parametrization should satisfy certain natural-
and N , the tensor product, M ⊗ N , is the unique ity conditions; most often, the parameter space
linear space with the property that, given a bi- (called moduli space, since the parameters are
torsion-free group An Abelian group with totally real field A number field with no
no elements of finite order except for the identity. imaginary infinite prime divisors.
torsion group An Abelian group with the total matrix algebra The set of all n × n
property that all of its elements are of finite order. matrices over a field K. It is usually denoted
Mn (K).
torsion product A left-derived functor of the
tensor product of modules. See also Tor. total step method for solving linear equations
See Jacobi method for solving linear equations.
torsion subgroup The subgroup of an Abel-
ian group consisting of the set of all elements of total transform Let T be a rational mapping
finite order. from the algebraic variety V to the algebraic
variety W and let T
be an irreducible subvariety
torus embedding A normal scheme X, lo- of T whose projections have the closed images
cally of finite type, over a closed field k on which V
and W
which are irreducible subvarieties of
an algebraic torus T acts with a dense open orbit V and W , respectively. The total transform of
isomorphic to T . V
is the set of points of W that correspond to
V
by T .
total boundary operator For the associated
chain complex, (Xn , ∂), ∂ is called the total trace The trace of a square matrix, A = (aij )
boundary operator. is the sum of the elements along the main diag-
onal,
total degree See filtration degree, comple- n
mentary degree. Tr(A) = aii .
i=1
total differential Let f be a real-valued func-
tion of the real variables, x1 , x2 , . . . , xn . The trace formula Let T be the regular represen-
total differential is a generalization of the differ- tation of a connected semisimple Lie group G on
ential for a function of one variable. Its existence / G, where / is a discrete subgroup of G, and
for a function of two variables is equivalent to let χk be the characters of the irreducible unitary
the existence of a unique tangent plane to the representations in the irreducible decomposition
surface at a point. When it does exist, the total of T . The trace formula is an identity between
differential of f at a point (a1 , a2 , . . . , an ) is the two ways of calculating
the trace of the integral
function operator with kernel y∈/ f (x −1 λy):
df (dx1 , dx2 , . . . , dxn ) ∞
= fx1 dx1 + fx2 dx2 + · · · + fxn dxn . χk (g) dg = f x −1 γ x dx ,
k=1 G γ
where the partials are evaluated at (a1 , a2 , . . . ,
an ). Note, though, that the existence of the par- where γ is the conjugate class of γ in / and Dγ
tials fx1 , fx2 , . . . , fxn is not sufficient for the is the quotient space of the centralizer Gγ of γ
existence of the total differential when n > 1. in G by the centralizer /γ of γ in /.
Tucker’s Theorem on Complementary Slack- type-I von Neumann algebra A von Neu-
ness For any matrix A, the inequalities Ax = mann algebra that contains an Abelian projec-
0, and u
A ≥ 0 have solutions x, u that satisfy tion E for which I is the only central projection
A
u + x > 0. covering E. Also called a discrete von Neumann
algebra.
two-phase simplex method A method used
to introduce new artificial variables so that there type-II von Neumann algebra A von Neu-
is a solution to start with. This usually happens mann algebra that is semifinite and contains no
if (0, 0, . . . , 0) is not a feasible solution to the Abelian projection.
linear programming problem. The first phase
is to use the new variables to have a solution to type-III von Neumann algebra A von Neu-
start with, then after performing some row trans- mann algebra for which a semi-finite normal
formations, we can discard all the artificial vari- trace does not exist. Also called a purely in-
ables. This reduces the original question to an finite von Neumann algebra.
ordinary linear programming problem. Solving
this problem is the second phase of this algo- type of a group A generic phrase that could
rithm. refer to (i.) the homotopy type of a group, (ii.) the
property of the group being of Lie type, or
type-I factor A factor that is isomorphic to (iii.) the property of the group being of Schot-
the algebra of bounded operators on an n-dimen- tky type, or any of several other commonly used
sional Hilbert space. phrases.
undefined term The initial objects about uniqueness theorem In general, a theorem
which the basic hypotheses of an axiom system that states that an object that satisfies certain
are written. An example from set theory of an specified conditions must be unique. Examples
undefined term is element. are theorems that state that certain differential
equations have unique solutions that satisfy ini-
undetermined coefficients Referring to a tial values and theorems that state that functions
method of solving a system of n linear differ- (especially analytic functions) that satisfy cer-
ential equations, when the matrix of the system tain conditions must be unique.
has unique distinct real eigenvalues and the ini-
tial conditions are given. The method uses the unirational variety An irreducible algebraic
fact that each solution xi of such a system is of variety over a field k that has a finite algebraic
the form extension which is purely transcendental over k.
xi (t) = ci1 etλ1 + · · · + cin etλn . uniserial algebra An algebra that can be de-
composed into a direct sum of ideals that are
From this fact, the initial conditions, and the primary rings.
original system of differential equations, a sys-
tem of linear equations in the unknown coeffi- unit (1) An invertible element, especially in
cients cij can be derived and solved, yielding the a ring.
final solution. (2) The multiplicative identity element.
unipotent component The unipotent linear unitary algebra An algebra with a unitary
transformation that is one of the products of the element.
unitary ring A ring with a unitary element. universal domain An algebraically closed
field K that has infinite transcendence degree
unitary transformation A lineartransfor- over a given field k.
mation that leaves the inner product ni=1 x i xi
invariant. See also unitary matrix. universal enveloping algebra A universal
enveloping algebra of a Lie algebra G over a
unit circle The set of all points in the plane commutative ring R with a unit element is an
that are unit distance from the origin. associative R-algebra U (G) with a unit element,
together with a mapping σ : G → U (G) for
T = z = eit : 0 ≤ t < 2π . which the following properties hold:
(i.) σ is a homomorphism of Lie algebras, i.e.,
σ is R-linear and
unit element In a group G or ring R, the σ ([X, Y ]) = σ (X)σ (Y ) − σ (Y )σ (X) ,
element e that is the multiplicative identity. That
is, either e · g = g for all g ∈ G or e · r = r for for all X, Y ∈ G;
all r ∈ R. (ii.) For every associative R-algebra A with a
unmixed ideal An ideal which is not mixed. upper semi-continuous function Let X be
See mixed ideal. a topological space and f : X → R a function.
Then f is said to be upper semi-continuous if
Unmixedness Theorem If R is a locally f −1 ((−∞, α)) is open for every real α.
Macaulay ring, and an ideal a of the polynomial
ring R[x1 , . . . , xn ] over R is generated by r el- upper triangular matrix A square matrix
ements, with height a = r, then a is unmixed. having only zero entries below the main diago-
See local Macaulay ring, unmixed ideal. nal.
V det
1
·
xn
·
···
2
·
=
xnn−1
j <k
xj − x k .
A = A1 A2 · · · Aj −1
valuation ideal Let K be a number field and
let v be an additive valuation of the field K into and Aj ∩ Ak = {e}, if j = k. If is odd, Van-
an ordered additive group G. It follows that the diver’s conjecture states that the group A2 A4 A6
set Rv = {a ∈ K : v(a) ≥ 0} is a subring · · · A−1 is trivial.
of K and this ring has only one maximal ideal
{a : v(a) = 0} which is the valuation ideal of vanish To equal zero, e.g., a function f from
v (or of Rv ). a set to the real or complex numbers is said
to vanish on a subset B of if it maps every
element of B to zero.
valuation ring If A ⊆ B ⊆ K are two local
rings with maximal ideals a and b, respectively, vanishing cocycle A cocycle which is equal
then one says that B dominates A if a ⊂ b. to some coboundary in a cochain complex.
Dominance is a partial order relation on the set
of subrings of K. The maximal elements of this variable A quantity that is allowed to repre-
set are exactly the valuation rings of K. sent any element of a given set. Any number in
the set is called a value of the variable and the
set itself is called the domain of the variable.
value When a variable stands in place of con-
stants, then any one of these constants represents variable component A term that stands in
a value of the variable. contrast to the fixed component of a linear sys-
tem. See fixed component. The precise mean-
value group The submodule {v(a) : a ∈ ing of variable component, however, may vary
K\{0}} of G, where G is a totally ordered group depending upon the context.
(i.e., G is a lattice-ordered commutative group
variation (1) A term, introduced by
that is totally ordered, with the group operation
Lagrange, used to denote a small displacement
expressed by addition), K is a field, and v is an
of an independent variable or of a functional. If
additive valuation v : K → G∪{∞} of the field
f is a functional defined on a space , a varia-
K (i.e., a mapping satisfying (i.) v(a) = ∞
tion of the argument ω0 ∈ is a curve ω(t, v),
if and only if a = 0; (ii.) v(ab) = v(a) +
where a ≤ t ≤ b, a ≤ 0, b ≥ 0, passing through
v(b) for all a, b = 0; and (iii.) v(a + b) ≥
ω0 . As v ranges through V (space of parame-
min{v(a), v(b)}).
ters), the variations range through a family of
curves where, on each curve, t = 0 corresponds
Vandermonde determinant The determi- to position ω0 . In the case where V = , v is
nant of an n × n square matrix of which each referred to as variation and is called the direc-
row consists of the 0th to (n − 1)st powers of tional variation.
where the supremum is taken over all finite par- and direction. Commonly represented by a di-
titions π of E into disjoint subsets of E, each rected line segment whose length and orienta-
element of π belonging to A. If |µ|() < ∞, tion in space give the magnitude and direction,
then µ is said to be of bounded variation. respectively. Two directed line segments repre-
sent the same vector if they are equivalent; that
variety If An denotes the affine n-space over is, if they have the same length and orientation.
a field k and k[x1 , x2 , . . . , xn ] the polynomial (2) An element of a vector space. See vector
ring in n variables over k, an affine algebraic space.
variety (or simply affine variety) is a subset of
An of the form
vector bundle A mapping π of a vector
space E onto a vector space M such that for
p ∈ An : f (p) = 0 for all f ∈ T ,
each p ∈ M the inverse image π −1 (p) =
where T is some subset of K[x1 , x2 , . . . , xn ]. If {e ∈ E : π(e) = p} is a real vector space.
Pn denotes the projective n-space over a field More specifically, a vector bundle is a five-tuple
k and k[x0 , x1 , . . . , xn ] the polynomial ring in (E, M, V k , W (k), π ) where (i.) E is the bundle
n + 1 variables over k, a projective algebraic space, (ii.) M is the base space, (iii.) V k is a k-
variety (or simply projective variety) is a subset dimensional vector space, called the fiber space,
of Pn of the form (iv.) π is a mapping of E onto M, called the pro-
jection, such that π −1 (p) is homeomorphic to
p ∈ Pn : f (p) = 0 for all f ∈ T , V k for all p ∈ M, and (v.) W (k) is the general
linear group of all non-singular k × k matrices
where T is a subset of homogeneous polynomi- acting on V k in such a way that if Av = v for
als in K[x0 , x1 , . . . , xn ]. all v ∈ V k then A = I , the identity matrix.
weak (inductive) dimension The empty set Weak Lefschetz Theorem Let X be an al-
∅ is declared to have dimension −1. Proceed gebraic subvariety of complex dimension n in
inductively as follows. If R is a set and if for the complex projective space CPN , let P ⊂
every neighborhood U (p) of each point p ∈ CPN be a hyperplane passing through all sin-
R there exists an open neighborhood V such gular points of X (if any) and let Y = X ∩ P
that p ∈ V ⊂ U (p) and the dimension of the be a hyperplane section of X. Then the relative
boundary of V is n − 1, then we say that R has homology groups Hi (X, Y ; Z) vanish for i < n.
weak inductive dimension ≤ n. This implies that the natural homomorphism
If it is the case that the dimension of R is less
than or equal to n and it is not the case that the Hi (Y ; Z) −→ Hi (X; Z)
dimension of R is less than or equal to n − 1,
then the dimension of R is n. is an isomorphism for i < n − 2 and surjective
for i = n − 1.
weak convergence (1) Convergence in the
weak topology. A sequence {xn } in a linear topo- Weak Mordell-Weil Theorem If A is an
logical space V is said to converge weakly to an Abelian variety of dimension n, defined over
element x in V if, for every continuous linear an algebraic number field k of finite degree, and
functional f on V , limn→∞ f (xn ) = f (x). Ak the group of all k-rational points on A, then
wedge product (1) If X and Y are topological weight (1) The smallest cardinal number
spaces and x0 ∈ X, y0 ∈ Y fixed points, the which is the cardinality of an open base of a
subspace X × {y0 } ∪ {x0 } × Y of the product topological space X is called the weight of the
X × Y is called a wedge product of X and Y . space X.
(2) A multiplication operation defined on the (2) Let G be a complex semisimple Lie al-
alternating multilinear forms on a K-module V , gebra and fix a Cartan subalgebra H of G. Fur-
where K is a commutative ring with identity. thermore, fix a lexicographic linear ordering on
The wedge product is also called exterior prod- HR∗ , the linear space of all C-valued forms on H .
uct and is defined to be associative. Assume L Let (ρ, V ) be a representation of G. For each
and M are alternating multilinear forms on V of λ ∈ H ∗ let Vλ = {v ∈ V : ρ(H )v = λ(H )v}.
degree r and s, respectively. Let Sr+s be the set Then Vλ is a subspace of V and if Vλ = {0},
of permutations of {1, 2, . . . , r + s} and G ⊂ S then λ is called the weight of the representation
the set of all permutations σ which permute the ρ (with respect to H ). The set of weights is fi-
sets {1, 2, . . . , r} and {r + 1, r + 2, . . . , r + s} nite and its maximum element with respect to
within themselves. Let S be any set of permuta- the ordering on HR∗ is called the highest weight
tions of {1, 2, . . . , r + s} which contains exactly of ρ.
one element from each left coset of G. The (3) Let R be a root system in a Euclidean
wedge product of L and M is defined by the space E. Let / be the set of all elements λ in
equation E for which 2(λ, r)/(r, r) is an integer, for all
r ∈ R. The elements of / are called weights.
L∧M = (sgn σ )(L ⊗ M)σ ,
σ ∈S
weight function (1) If p and f are nonneg-
where ative functions defined and integrable on 0, let
Mr (f ) be defined as
(L ⊗ M)σ (α1 , . . . , αr+s )
r 1/r
0 pf dx
= (L ⊗ M) ασ 1 , . . . , ασ (r+s) . Mr (f ) = , r = 0
0 p dx
Weierstrass canonical form (1) An equa- and if Mr (f ) > 0 for some r > 0, define
tion representing the reciprocal of the gamma M0 (f ) = limr→0+ Mr (f ). The quantity Mr (f )
function as an infinite product, is called the mean of degree r of f with respect
∞ to the weight function p. If p = 1, then M1 (f ),
1 x −x/n M0 (f ), and M−1 (f ) are the arithmetic, geomet-
= xeCx 1+ e .
((x) n ric, and harmonic means of f , respectively.
n=1
Weyl chamber (1) See Weyl group, for the where G is a compact, connected, semisimple
Weyl chambers of a root system in a finite di- Lie group, (ρ, V ) is an irreducible representa-
mensional Euclidian space. tion of G, ξρ is the character of ρ (χρ (g) =
(2) Let G be a connected affine algebraic trρ(g)), /
is the highest weight of ρ, λ ∈ P ,
group, T a maximal torus in G, and B a Borel ξρ (h) = w∈W det(w)e(w(λ))(X) , h = exp X,
subgroup of G containing T . The set of all semi- and δ = 21 α∈;+ α.
regular one-parameter subgroups in T whose as-
sociated Borel groups equal B is called a Weyl Weyl’s Theorem A theorem of H. Weyl stat-
chamber of B with respect to T in G. ing that any finite dimensional representation
Y
are inserted in the blocks of a Young diagram
in some order, the diagram is called a Young
tableau.
Z an ideal of A.
zero exponent The integer 0, as a power zeta function (1) The series
to raise elements of a multiplicative group. A
consequence of one of the laws of exponents is 1 1 1
ζ (s) = 1 + + s + s + ···
a m /a n = a m−n . Thus, when the base a is differ- 2 s 3 4
ent from zero and the two exponents are equal,
which converges for all real numbers s > 1 is
n = m the equation becomes 1 = a m /a n =
known as the Riemann ζ -function. B. Riemann
a m−n = a 0 .
was the first to treat ζ (s) as a function of a com-
plex variable and showed that ζ (s) is holomor-
zero homomorphism A homomorphism φ : phic and has no zeros in s > 1. Furthermore,
G −→ G
that maps every element of G to the Riemann proved that ζ (s) can be extended to
identity in G. A zero homomorphism is also a meromorphic function in the whole complex
called the trivial homomorphism. plane whose only pole is a simple pole a s = 1.
See also Riemann hypothesis.
zero matrix A matrix in which every entry (2) Any of several special functions called ζ -
is zero. functions that (i.) are meromorphic on the whole