This action might not be possible to undo. Are you sure you want to continue?

:

A STUDY GUIDE

FOR BEGINNERS

John A. Beachy

Northern Illinois University

2000

ii

This is a supplement to

Abstract Algebra, Second Edition

by John A. Beachy and William D. Blair

ISBN 0–88133–866–4, Copyright 1996

Waveland Press, Inc.

P.O. Box 400

Prospect Heights, Illinois 60070

847 / 634-0081

www.waveland.com

c _John A. Beachy 2000

Permission is granted to copy this document in electronic form, or to print it for

personal use, under these conditions:

it must be reproduced in whole;

it must not be modiﬁed in any way;

it must not be used as part of another publication.

Formatted February 8, 2002, at which time the original was available at:

http://www.math.niu.edu/

∼

beachy/abstract algebra/

Contents

PREFACE v

1 INTEGERS 1

1.1 Divisors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Primes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.3 Congruences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.4 Integers Modulo n . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2 FUNCTIONS 7

2.1 Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 Equivalence Relations . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.3 Permutations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3 GROUPS 13

3.1 Deﬁnition of a Group . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.2 Subgroups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.3 Constructing Examples . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.4 Isomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3.5 Cyclic Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3.6 Permutation Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.7 Homomorphisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.8 Cosets, Normal Subgroups, and Factor Groups . . . . . . . . . . . . 24

Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

4 POLYNOMIALS 27

Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

5 COMMUTATIVE RINGS 29

Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

iii

iv CONTENTS

6 FIELDS 33

Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

SOLUTIONS 33

1 Integers 35

2 Functions 49

3 Groups 57

4 Polynomials 87

5 Commutative Rings 93

6 Fields 101

BIBLIOGRAPHY 104

INDEX 105

PREFACE v

PREFACE

I ﬁrst taught an abstract algebra course in 1968, using Herstein’s Topics in

Algebra. It’s hard to improve on his book; the subject may have become broader,

with applications to computing and other areas, but Topics contains the core of any

course. Unfortunately, the subject hasn’t become any easier, so students meeting

abstract algebra still struggle to learn the new concepts, especially since they are

probably still learning how to write their own proofs.

This “study guide” is intended to help students who are beginning to learn

about abstract algebra. Instead of just expanding the material that is already

written down in our textbook, I decided to try to teach by example, by writing out

solutions to problems. I’ve tried to choose problems that would be instructive, and

in quite a few cases I’ve included comments to help the reader see what is really

going on. Of course, this study guide isn’t a substitute for a good teacher, or for

the chance to work together with other students on some hard problems.

Finally, I would like to gratefully acknowledge the support of Northern Illinois

University while writing this study guide. As part of the recognition as a “Presi-

dential Teaching Professor,” I was given leave in Spring 2000 to work on projects

related to teaching.

DeKalb, Illinois John A. Beachy

October 2000

vi PREFACE

Chapter 1

INTEGERS

Chapter 1 of the text introduces the basic ideas from number theory that are a

prerequisite to studying abstract algebra. Many of the concepts introduced there

can be abstracted to much more general situations. For example, in Chapter 3 of

the text you will be introduced to the concept of a group. One of the ﬁrst broad

classes of groups that you will meet depends on the deﬁnition of a cyclic group, one

that is obtained by considering all powers of a particular element. The examples

in Section 1.4, constructed using congruence classes of integers, actually tell you

everything you will need to know about cyclic groups. In fact, although Chapter 1

is very concrete, it is a signiﬁcant step forward into the realm of abstract algebra.

1.1 Divisors

Before working through the solved problems for this section, you need to make sure

that you are familiar with all of the deﬁnitions and theorems in the section. In

many cases, the proofs of the theorems contain important techniques that you need

to copy in solving the exercises in the text. Here are several useful approaches you

should be able to use.

—When working on questions involving divisibility you may ﬁnd it useful to go back

to Deﬁnition 1.1.1. If you expand the expression b[a by writing “a = bq for some

q ∈ Z”, then you have an equation to work with. This equation involves ordinary

integers, and so you can use all of the things you already know (from high school

algebra) about working with equations.

—To show that b[a, try to write down an expression for a and expand, simplify, or

substitute for terms in the expression until you can show how to factor out b.

—Another approach to proving that b[a is to use the division algorithm (see The-

orem 1.1.3) to write a = bq + r, where 0 ≤ r < b. Then to prove that b[a you only

1

2 CHAPTER 1. INTEGERS

need to ﬁnd some way to check that r = 0.

—Theorem 1.1.6 states that any two nonzero integers a and b have a greatest

common divisor, which can be expressed as the smallest positive linear combination

of a and b. An integer is a linear combination of a and b if and only if it is

a multiple of their greatest common divisor. This is really useful in working on

questions involving greatest common divisors.

SOLVED PROBLEMS: §1.1

22. Find gcd(435, 377), and express it as a linear combination of 435 and 377.

23. Find gcd(3553, 527), and express it as a linear combination of 3553 and 527.

24. Which of the integers 0, 1, . . . , 10 can be expressed in the form 12m + 20n,

where m, n are integers?

25. If n is a positive integer, ﬁnd the possible values of gcd(n, n + 10).

26. Prove that if a and b are nonzero integers for which a[b and b[a, then b = ±a.

27. Prove that if m and n are odd integers, then m

2

−n

2

is divisible by 8.

28. Prove that if n is an integer with n > 1, then gcd(n − 1, n

2

+ n + 1) = 1 or

gcd(n −1, n

2

+n + 1) = 3.

29. Prove that if n is a positive integer, then

_

_

0 0 −1

0 1 0

1 0 0

_

_

n

=

_

_

1 0 0

0 1 0

0 0 1

_

_

if and only if 4[n.

30. Give a proof by induction to show that each number in the sequence 12, 102,

1002, 10002, . . ., is divisible by 6.

1.2 Primes

Proposition 1.2.2 states that integers a and b are relatively prime if and only if there

exist integers m and n with ma + nb = 1. This is one of the most useful tools in

working with relatively prime integers. Remember that this only works in showing

that gcd(a, b) = 1. More generally, if you have a linear combination ma + nb = d,

it only shows that gcd(a, b) is a divisor of d (refer back to Theorem 1.1.6).

Since the fundamental theorem of arithmetic (on prime factorization) is proved

in this section, you now have some more familiar techniques to use.

1.3. CONGRUENCES 3

SOLVED PROBLEMS: §1.2

23. (a) Use the Euclidean algorithm to ﬁnd gcd(1776, 1492).

(b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776, 1492).

24. (a) Use the Euclidean algorithm to ﬁnd gcd(1274, 1089).

(b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274, 1089).

25. Give the lattice diagram of all divisors of 250. Do the same for 484.

26. Find all integer solutions of the equation xy + 2y −3x = 25.

27. For positive integers a, b, prove that gcd(a, b) = 1 if and only if gcd(a

2

, b

2

) = 1.

28. Prove that n−1 and 2n−1 are relatively prime, for all integers n > 1. Is the

same true for 2n −1 and 3n −1?

29. Let m and n be positive integers. Prove that gcd(2

m

− 1, 2

n

− 1) = 1 if and

only if gcd(m, n) = 1.

30. Prove that gcd(2n

2

+ 4n −3, 2n

2

+ 6n −4) = 1, for all integers n > 1.

1.3 Congruences

In this section, it is important to remember that although working with congruences

is almost like working with equations, it is not exactly the same.

What things are the same? You can add or subtract the same integer on both

sides of a congruence, and you can multiply both sides of a congruence by the same

integer. You can use substitution, and you can use the fact that if a ≡ b (mod n)

and b ≡ c (mod n), then a ≡ c (mod n). (Review Proposition 1.3.3, and the

comments in the text both before and after the proof of the proposition.)

What things are diﬀerent? In an ordinary equation you can divide through by

a nonzero number. In a congruence modulo n, you can only divide through by an

integer that is relatively prime to n. This is usually expressed by saying that if

gcd(a, n) = 1 and ac ≡ ad (mod n), then c ≡ d (mod n). Just be very careful!

One of the important techniques to understand is how to switch between con-

gruences and ordinary equations. First, any equation involving integers can be

converted into a congruence by just reducing modulo n. This works because if two

integers are equal, then are certainly congruent modulo n.

The do the opposite conversion you must be more careful. If two integers are

congruent modulo n, that doesn’t make them equal, but only guarantees that di-

viding by n produces the same remainder in each case. In other words, the integers

may diﬀer by some multiple of n.

4 CHAPTER 1. INTEGERS

The conversion process is illustrated in Example 1.3.5 of the text, where the

congruence

x ≡ 7 (mod 8)

is converted into the equation

x = 7 + 8q , for some q ∈ Z .

Notice that converting to an equation makes it more complicated, because we have

to introduce another variable. In the example, we really want a congruence modulo

5, so the next step is to rewrite the equation as

x ≡ 7 + 8q (mod 5) .

Actually, we can reduce each term modulo 5, so that we ﬁnally get

x ≡ 2 + 3q (mod 5) .

You should read the proofs of Theorem 1.3.5 and Theorem 1.3.6 very carefully.

These proofs actually show you the necessary techniques to solve all linear congru-

ences of the form ax ≡ b (mod n), and all simultaneous linear equations of the form

x ≡ a (mod n) and x ≡ b (mod m), where the moduli n and m are relatively prime.

Many of the theorems in the text should be thought of as “shortcuts”, and you can’t

aﬀord to skip over their proofs, because you might miss important algorithms or

computational techniques.

SOLVED PROBLEMS: §1.3

26. Solve the congruence 42x ≡ 12 (mod 90).

27. (a) Find all solutions to the congruence 55x ≡ 35 (mod 75).

(b) Find all solutions to the congruence 55x ≡ 36 (mod 75).

28. (a) Find one particular integer solution to the equation 110x + 75y = 45.

(b) Show that if x = m and y = n is an integer solution to the equation in

part (a), then so is x = m+ 15q and y = n −22q, for any integer q.

29. Solve the system of congruences x ≡ 2 (mod 9) x ≡ 4 (mod 10) .

30. Solve the system of congruences 5x ≡ 14 (mod 17) 3x ≡ 2 (mod 13) .

31. Solve the system of congruences x ≡ 5 (mod 25) x ≡ 23 (mod 32) .

32. Give integers a, b, m, n to provide an example of a system

x ≡ a (mod m) x ≡ b (mod n)

that has no solution.

1.4. INTEGERS MODULO N 5

33. (a) Compute the last digit in the decimal expansion of 4

100

.

(b) Is 4

100

divisible by 3?

34. Find all integers n for which 13 [ 4(n

2

+ 1).

35. Prove that 10

n+1

+ 4 10

n

+ 4 is divisible by 9, for all positive integers n.

36. Prove that the fourth power of an integer can only have 0, 1, 5, or 6 as its

units digit.

1.4 Integers Modulo n

The ideas in this section allow us to work with equations instead of congruences,

provided we think in terms of equivalence classes. To be more precise, any linear

congruence of the form

ax ≡ b (mod n)

can be viewed as an equation in Z

n

, written

[a]

n

[x]

n

= [b]

n

.

This gives you one more way to view problems involving congruences. Sometimes

it helps to have various ways to think about a problem, and it is worthwhile to learn

all of the approaches, so that you can easily shift back and forth between them, and

choose whichever approach is the most convenient. For example, trying to divide by

a in the congruence ax ≡ b (mod n) can get you into trouble unless gcd(a, n) = 1.

Instead of thinking in terms of division, it is probably better to think of multiplying

both sides of the equation [a]

n

[x]

n

= [b]

n

by [a]

−1

n

, provided [a]

−1

n

exists.

It is well worth your time to learn about the sets Z

n

and Z

×

n

. They will provide

an important source of examples in Chapter 3, when we begin studying groups.

The exercises for Section 1.4 of the text contain several deﬁnitions for elements

of Z

n

. If (a, n) = 1, then the smallest positive integer k such that a

k

≡ 1 (mod n)

is called the multiplicative order of [a] in Z

×

n

. The set Z

×

n

is said to be cyclic if

it contains an element of multiplicative order ϕ(n). Since [Z

×

n

[ = ϕ(n), this is

equivalent to saying that Z

×

n

is cyclic if has an element [a] such that each element

of Z

×

n

is equal to some power of [a]. Finally, the element [a] ∈ Z

n

is said to be

idempotent if [a]

2

= [a], and nilpotent if [a]

k

= [0] for some k.

SOLVED PROBLEMS: §1.4

30. Find the multiplicative inverse of each nonzero element of Z

7

.

31. Find the multiplicative inverse of each nonzero element of Z

13

.

6 CHAPTER 1. INTEGERS

32. Find [91]

−1

501

, if possible (in Z

×

501

).

33. Find [3379]

−1

4061

, if possible (in Z

×

4061

).

34. In Z

20

: ﬁnd all units (list the multiplicative inverse of each); ﬁnd all idempo-

tent elements; ﬁnd all nilpotent elements.

35. In Z

24

: ﬁnd all units (list the multiplicative inverse of each); ﬁnd all idem-

potent elements; ﬁnd all nilpotent elements.

36. Show that Z

×

17

is cyclic.

37. Show that Z

×

35

is not cyclic but that each element has the form [8]

i

35

[−4]

j

35

,

for some positive integers i, j.

38. Solve the equation [x]

2

11

+ [x]

11

−[6]

11

= [0]

11

.

39. Let n be a positive integer, and let a ∈ Z with gcd(a, n) = 1. Prove that if k

is the smallest positive integer for which a

k

≡ 1 (mod n), then k [ ϕ(n).

40. Prove that [a]

n

is a nilpotent element of Z

n

if and only if each prime divisor

of n is a divisor of a.

Review Problems

1. Find gcd(7605, 5733), and express it as a linear combination of 7605 and 5733.

2. For ω = −

1

2

+

√

3

2

i, prove that ω

n

= 1 if and only if 3[n, for any integer n.

3. Solve the congruence 24x ≡ 168 (mod 200).

4. Solve the system of congruences 2x ≡ 9 (mod 15) x ≡ 8 (mod 11) .

5. List the elements of Z

×

15

. For each element, ﬁnd its multiplicative inverse, and

ﬁnd its multiplicative order.

6. Show that if n > 1 is an odd integer, then ϕ(2n) = ϕ(n).

Chapter 2

FUNCTIONS

The ﬁrst goal of this chapter is to provide a review of functions. In our study of

algebraic structures in later chapters, functions will provide a way to compare two

diﬀerent structures. In this setting, the functions that are one-to-one correspon-

dences will be particularly important.

The second goal of the chapter is to begin studying groups of permutations,

which give a very important class of examples. When you begin to study groups in

Chapter 3, you will be able draw on your knowledge of permutation groups, as well

as on your knowledge of the groups Z

n

and Z

×

n

.

2.1 Functions

Besides reading Section 2.1, it might help to get out your calculus textbook and

review composite functions, one-to-one and onto functions, and inverse functions.

The functions f : R → R

+

and g : R

+

→ R deﬁned by f(x) = e

x

, for all x ∈ R,

and g(y) = ln y, for all y ∈ R

+

, provide one of the most important examples of a

pair of inverse functions.

Deﬁnition 2.1.1, the deﬁnition of function, is stated rather formally in terms of

ordered pairs. (Think of this as a deﬁnition given in terms of the “graph” of the

function.) In terms of actually using this deﬁnition, the text almost immediately

goes back to what might be a more familiar deﬁnition: a function f : S → T is a

“rule” that assigns to each element of S a unique element of T.

One of the most fundamental ideas of abstract algebra is that algebraic struc-

tures should be thought of as essentially the same if the only diﬀerence between

them is the way elements have been named. To make this precise we will say that

structures are the same if we can set up an invertible function from one to the other

that preserves the essential algebraic structure. That makes it especially important

to understand the concept of an inverse function, as introduced in this section.

7

8 CHAPTER 2. FUNCTIONS

SOLVED PROBLEMS: §2.1

20. The “Vertical Line Test” from calculus says that a curve in the xy-plane is

the graph of a function of x if and only if no vertical line intersects the curve

more than once. Explain why this agrees with Deﬁnition 2.1.1.

21. The “Horizontal Line Test” from calculus says that a function is one-to-one

if and only if no horizontal line intersects its graph more than once. Explain

why this agrees with Deﬁnition 2.1.4.

more than one

22. In calculus the graph of an inverse function f

−1

is obtained by reﬂecting the

graph of f about the line y = x. Explain why this agrees with Deﬁnition 2.1.7.

23. Let A be an n n matrix with entries in R. Deﬁne a linear transformation

L : R

n

→R

n

by L(x) = Ax, for all x ∈ R

n

.

(a) Show that L is an invertible function if and only if det(A) ,= 0.

(b) Show that if L is either one-to-one or onto, then it is invertible.

24. Let A be an mn matrix with entries in R, and assume that m > n. Deﬁne

a linear transformation L : R

n

→ R

m

by L(x) = Ax, for all x ∈ R

n

. Show

that L is a one-to-one function if det(A

T

A) ,= 0, where A

T

is the transpose

of A.

25. Let A be an n n matrix with entries in R. Deﬁne a linear transformation

L : R

n

→ R

n

by L(x) = Ax, for all x ∈ R

n

. Prove that L is one-to-one if

and only if no eigenvalue of A is zero.

Note: A vector x is called an eigenvector of A if it is nonzero and there exists

a scalar λ such a that Ax = λx.

26. Let a be a ﬁxed element of Z

×

17

. Deﬁne the function θ : Z

×

17

→ Z

×

17

by

θ(x) = ax, for all x ∈ Z

×

17

. Is θ one to one? Is θ onto? If possible, ﬁnd the

inverse function θ

−1

.

2.2 Equivalence Relations

In a variety of situations it is useful to split a set up into subsets in which the

elements have some property in common. You are already familiar with one of

the important examples: in Chapter 1 we split the set of integers up into subsets,

depending on the remainder when the integer is divided by the ﬁxed integer n. This

led to the concept of congruence modulo n, which is a model for our general notion

of an equivalence relation.

In this section you will ﬁnd three diﬀerent points of view, looking at the one idea

of splitting up a set S from three distinct vantage points. First there is the deﬁnition

2.2. EQUIVALENCE RELATIONS 9

of an equivalence relation on S, which tells you when two diﬀerent elements of S

belong to the same subset. Then there is the notion of a partition of S, which places

the emphasis on describing the subsets. Finally, it turns out that every partition

(and equivalence relation) really comes from a function f : S → T, where we say

that x

1

and x

2

are equivalent if f(x

1

) = f(x

2

).

The reason for considering several diﬀerent point of view is that in a given

situation one point of view may be more useful than another. Your goal should be

to learn about each point of view, so that you can easily switch from one to the

other, which is a big help in deciding which point of view to take.

SOLVED PROBLEMS: §2.2

14. On the set ¦(a, b)¦ of all ordered pairs of positive integers, deﬁne (x

1

, y

1

) ∼

(x

2

, y

2

) if x

1

y

2

= x

2

y

1

. Show that this deﬁnes an equivalence relation.

15. On the set C of complex numbers, deﬁne z

1

∼ z

2

if [[z

1

[[ = [[z

2

[[. Show that

∼ is an equivalence relation.

16. Let u be a ﬁxed vector in R

3

, and assume that u has length 1. For vectors v

and w, deﬁne v ∼ w if v u = w u, where denotes the standard dot product.

Show that ∼ is an equivalence relation, and give a geometric description of

the equivalence classes of ∼.

17. For the function f : R →R deﬁned by f(x) = x

2

, for all x ∈ R, describe the

equivalence relation on R that is determined by f.

18. For the linear transformation L : R

3

→R

3

deﬁned by

L(x, y, z) = (x +y +z, x +y +z, x +y +z) ,

for all (x, y, z) ∈ R

3

, give a geometric description of the partition of R

3

that

is determined by L.

19. Deﬁne the formula f : Z

12

→ Z

12

by f([x]

12

) = [x]

2

12

, for all [x]

12

∈ Z

12

.

Show that the formula f deﬁnes a function. Find the image of f and the set

Z

12

/f of equivalence classes determined by f.

20. On the set of all nn matrices over R, deﬁne A ∼ B if there exists an invert-

ible matrix P such that PAP

−1

= B. Check that ∼ deﬁnes an equivalence

relation.

10 CHAPTER 2. FUNCTIONS

2.3 Permutations

This section introduces and studies the last major example that we need before we

begin studying groups in Chapter 3. You need to do enough computations so that

you will feel comfortable in dealing with permutations.

If you are reading another book along with Abstract Algebra, you need to be

aware that some authors multiply permutations by reading from left to right, instead

of the way we have deﬁned multiplication. Our point of view is that permutations

are functions, and we write functions on the left, just as in calculus, so we have to

do the computations from right to left.

In the text we noted that if S is any set, and Sym(S) is the set of all permutations

on S, then we have the following properties. (i) If σ, τ ∈ Sym(S), then τσ ∈ Sym(S);

(ii) 1

S

∈ Sym(S); (iii) if σ ∈ Sym(S), then σ

−1

∈ Sym(S). In two of the problems,

we need the following deﬁnition.

If G is a nonempty subset of Sym(S), we will say that G is a group of permuta-

tions if the following conditions hold.

(i) If σ, τ ∈ G, then τσ ∈ G;

(ii) 1

S

∈ G;

(iii) if σ ∈ G, then σ

−1

∈ G.

We will see later that this agrees with Deﬁnition 3.6.1 of the text.

SOLVED PROBLEMS: §2.3

13. For the permutation σ =

_

1 2 3 4 5 6 7 8 9

7 5 6 9 2 4 8 1 3

_

, write σ as a

product of disjoint cycles. What is the order of σ? Is σ an even permutation?

Compute σ

−1

.

14. For the permutations σ =

_

1 2 3 4 5 6 7 8 9

2 5 1 8 3 6 4 7 9

_

and

τ =

_

1 2 3 4 5 6 7 8 9

1 5 4 7 2 6 8 9 3

_

, write each of these permutations as a

product of disjoint cycles: σ, τ, στ, στσ

−1

, σ

−1

, τ

−1

, τσ, τστ

−1

.

15. Let σ = (2, 4, 9, 7, )(6, 4, 2, 5, 9)(1, 6)(3, 8, 6) ∈ S

9

. Write σ as a product of

disjoint cycles. What is the order of σ? Compute σ

−1

.

16. Compute the order of τ =

_

1 2 3 4 5 6 7 8 9 10 11

7 2 11 4 6 8 9 10 1 3 5

_

. For

σ = (3, 8, 7), compute the order of στσ

−1

.

17. Prove that if τ ∈ on

is a permutation with order m, then στσ

−1

has order m,

for any permutation σ ∈ on

.

2.3. PERMUTATIONS 11

18. Show that S

10

has elements of order 10, 12, and 14, but not 11 or 13.

19. Let S be a set, and let X be a subset of S. Let G = ¦σ ∈ Sym(S) [ σ(X) ⊂ X¦.

Prove that G is a group of permutations.

20. Let G be a group of permutations, with G ⊆ Sym(S), for the set S. Let τ be

a ﬁxed permutation in Sym(S). Prove that

τGτ

−1

= ¦σ ∈ Sym(S) [ σ = τγτ for some γ ∈ G¦

is a group of permutations.

12 CHAPTER 2. FUNCTIONS

Review Problems

1. For the function f : R →R deﬁned by f(x) = x

2

, for all x ∈ R, describe the

equivalence relation on R that is determined by f.

2. Deﬁne f : R → R by f(x) = x

3

+ 3xz − 5, for all x ∈ R. Show that f is a

one-to-one function.

Hint: Use the derivative of f to show that f is a strictly increasing function.

3. On the set Q of rational numbers, deﬁne x ∼ y if x − y is an integer. Show

that ∼ is an equivalence relation.

4. In S

10

, let α = (1, 3, 5, 7, 9), β = (1, 2, 6), and γ = (1, 2, 5, 3). For σ = αβγ,

write σ as a product of disjoint cycles, and use this to ﬁnd its order and its

inverse. Is σ even or odd?

5. Deﬁne the function φ : Z

×

17

→Z

×

17

by φ(x) = x

−1

, for all x ∈ Z

×

17

. Is φ one to

one? Is φ onto? If possible, ﬁnd the inverse function φ

−1

.

6. (a) Let α be a ﬁxed element of S

n

. Show that φ

α

: S

n

→ S

n

deﬁned by

φ

α

(σ) = ασα

−1

, for all σ ∈ S

n

, is a one-to-one and onto function.

(b) In S

3

, let α = (1, 2). Compute φ

α

.

Chapter 3

GROUPS

The study of groups, which we begin in this chapter, is usually thought of as the real

beginning of abstract algebra. The step from arithmetic to algebra involves starting

to use variables, which just represent various numbers. But the operations are still

the usual ones for numbers, addition, subtraction, multiplication, and division.

The step from algebra to abstract algebra involves letting the operation act like

a variable. At ﬁrst we will use ∗ or to represent an operation, to show that ∗ might

represent ordinary addition or multiplication, or possibly operations on matrices or

functions, or maybe even something quite far from your experience. One of the

things we try to do with notation is to make it look familiar, even if it represents

something new; very soon we will just write ab instead of a ∗ b, so long as everyone

knows the convention that we are using.

3.1 Deﬁnition of a Group

This section contains these deﬁnitions: binary operation, group, abelian group, and

ﬁnite group. These deﬁnitions provide the language you will be working with, and

you simply must know this language. Try to learn it so well that you don’t have

even a trace of an accent!

Loosely, a group is a set on which it is possible to deﬁne a binary operation that

is associative, has an identity element, and has inverses for each of its elements.

The precise statement is given in Deﬁnition 3.1.3; you must pay careful attention

to each part, especially the quantiﬁers (“for all”, “for each”, “there exists”), which

must be stated in exactly the right order.

From one point of view, the axioms for a group give us just what we need to

work with equations involving the operation in the group. For example, one of the

rules you are used to says that you can multiply both sides of an equation by the

same value, and the equation will still hold. This still works for the operation in a

group, since if x and y are elements of a group G, and x = y, then a x = a y, for

13

14 CHAPTER 3. GROUPS

any element a in G. This is a part of the guarantee that comes with the deﬁnition

of a binary operation. It is important to note that on both sides of the equation,

a is multiplied on the left. We could also guarantee that x a = y a, but we can’t

guarantee that a x = y a, since the operation in the group may not satisfy the

commutative law.

The existence of inverses allows cancellation (see Proposition 3.1.6 for the precise

statement). Remember that in a group there is no mention of division, so whenever

you are tempted to write a ÷b or a/b, you must write a b

−1

or b

−1

a. If you are

careful about the side on which you multiply, and don’t fall victim to the temptation

to divide, you can be pretty safe in doing the familiar things to an equation that

involves elements of a group.

Understanding and remembering the deﬁnitions will give you one level of un-

derstanding. The next level comes from knowing some good examples. The third

level of understanding comes from using the deﬁnitions to prove various facts about

groups.

Here are a few of the important examples. First, the sets of numbers Z, Q, R,

and C form groups under addition. Next, the sets Q

×

, R

×

, and C

×

of nonzero

numbers form groups under multiplication. The sets Z and Z

n

are groups under

addition, while Z

×

n

is a group under multiplication. It is common to just list these

sets as groups, without mentioning their operations, since in each case only one of

the two familiar operations can be used to make the set into a group. Similarly, the

set Mn

(R) of all nn matrices with entries in R is a group under addition, but not

multiplication, while the set GL

n

(R) of all invertible n n matrices with entries

in R is a group under multiplication, but not under addition. There shouldn’t be

any confusion in just listing these as groups, without speciﬁcally mentioning which

operation is used.

In the study of ﬁnite groups, the most important examples come from groups

of matrices. I should still mention that the original motivation for studying groups

came from studying sets of permutations, and so the symmetric group on

still has

an important role to play.

SOLVED PROBLEMS: §3.1

22. Use the dot product to deﬁne a multiplication on R

3

. Does this make R

3

into

a group?

23. For vectors (x

1

, y

1

, z

1

) and (x

2

, y

2

, z

2

) in R

3

, the cross product is deﬁned by

(x

1

, y

1

, z

1

)(x

2

, y

2

, z

2

) = (y

1

z

2

− z

1

y

2

, z

1

x

2

− x

1

z

2

, x

1

y

2

− y

1

x

2

). Is R

3

a

group under this multiplication?

24. On the set G = Q

×

of nonzero rational numbers, deﬁne a new multiplication

by a∗b =

ab

2

, for all a, b ∈ G. Show that G is a group under this multiplication.

25. Write out the multiplication table for Z

×

9

.

3.2. SUBGROUPS 15

26. Write out the multiplication table for Z

×

15

.

27. Let G be a group, and suppose that a and b are any elements of G. Show that

if (ab)

2

= a

2

b

2

, then ba = ab.

28. Let G be a group, and suppose that a and b are any elements of G. Show that

(aba

−1

)

n

= ab

n

a

−1

, for any positive integer n.

29. In Deﬁnition 3.1.3 of the text, replace condition (iii) with the condition that

there exists e ∈ G such that e a = a for all a ∈ G, and replace condition (iv)

with the condition that for each a ∈ G there exists a

∈ G with a

a = e.

Prove that these weaker conditions (given only on the left) still imply that G

is a group.

30. The previous exercise shows that in the deﬁnition of a group it is suﬃcient to

require the existence of a left identity element and the existence of left inverses.

Give an example to show that it is not suﬃcient to require the existence of a

left identity element together with the existence of right inverses.

31. Let F be the set of all fractional linear transformations of the complex plane.

That is, F is the set of all functions f(z) : C →C of the form f(z) =

az +b

cz +d

,

where the coeﬃcients a, b, c, d are integers with ad − bc = 1. Show that F

forms a group under composition of functions.

32. Let G = ¦x ∈ R [ x > 1¦ be the set of all real numbers greater than 1. For

x, y ∈ G, deﬁne x ∗ y = xy −x −y + 2.

(a) Show that the operation ∗ is closed on G.

(b) Show that the associative law holds for ∗.

(c) Show that 2 is the identity element for the operation ∗.

(d) Show that for element a ∈ G there exists an inverse a

−1

∈ G.

3.2 Subgroups

Many times a group is deﬁned by looking at a subset of a known group. If the

subset is a group in its own right, using the same operation as the larger set, then

it is called a subgroup. For instance, any group of permutations is a subgroup of

Sym(S), for some set S. Any group of n n matrices (with entries in R) is a

subgroup of GL

n

(R).

If the idea of a subgroup reminds you of studying subspaces in your linear algebra

course, you are right. If you only look at the operation of addition in a vector space,

it forms an abelian group, and any subspace is automatically a subgroup. Now might

be a good time to pick up your linear algebra text and review vector spaces and

subspaces.

16 CHAPTER 3. GROUPS

Lagrange’s theorem is very important. It states that in a ﬁnite group the number

of elements in any subgroup must be a divisor of the total number of elements in

the group. This is a useful fact to know when you are looking for subgroups in a

given group.

It is also important to remember that every element a in a group deﬁnes a

subgroup ¸a), consisting of all powers (positive and negative) of the element. This

subgroup has o(a) elements, where o(a) is the order of a. If the group is ﬁnite, then

you only need to look at positive powers, since in that case the inverse a

−1

of any

element can be expressed in the form a

n

, for some n > 0.

SOLVED PROBLEMS: §3.2

23. Find all cyclic subgroups of Z

×

24

.

24. In Z

×

20

, ﬁnd two subgroups of order 4, one that is cyclic and one that is not

cyclic.

25. (a) Find the cyclic subgroup of S

7

generated by the element (1, 2, 3)(5, 7).

(b) Find a subgroup of S

7

that contains 12 elements. You do not have to list

all of the elements if you can explain why there must be 12, and why they

must form a subgroup.

26. In G = Z

×

21

, show that

H = ¦[x]

21

[ x ≡ 1 (mod 3)¦ and K = ¦[x]

21

[ x ≡ 1 (mod 7)¦

are subgroups of G.

27. Let G be an abelian group, and let n be a ﬁxed positive integer. Show that

N = ¦g ∈ G [ g = a

n

for some a ∈ G¦ is a subgroup of G.

28. Suppose that p is a prime number of the form p = 2

n

+ 1.

(a) Show that in Z

×

p

the order of [2]

p

is 2n.

(b) Use part (a) to prove that n must be a power of 2.

29. In the multiplicative group C

×

of complex numbers, ﬁnd the order of the

elements −

√

2

2

+

√

2

2

i and −

√

2

2

−

√

2

2

i.

30. In the group G = GL

2

(R) of invertible 2 2 matrices with real entries, show

that

H =

__

cos θ −sin θ

sin θ cos θ

_¸

¸

¸

¸

θ ∈ R

_

is a subgroup of G.

3.3. CONSTRUCTING EXAMPLES 17

31. Let K be the following subset of GL

2

(R).

K =

__

a b

c d

_¸

¸

¸

¸

d = a, c = −2b, ad −bc ,= 0

_

Show that K is a subgroup of GL

2

(R).

32. Compute the centralizer in GL

2

(R) of the matrix

_

2 1

1 1

_

.

Note: Exercise 3.2.14 in the text deﬁnes the centralizer of an element a of the

group G to be C(a) = ¦x ∈ G [ xa = ax¦.

3.3 Constructing Examples

The most important result in this section is Proposition 3.3.7, which shows that the

set of all invertible n n matrices forms a group, in which we can allow the entries

in the matrix to come from any ﬁeld. This includes matrices with entries in the

ﬁeld Z

p

, for any prime number p, and this allows us to construct very interesting

ﬁnite groups as subgroups of GL

n

(Z

p

).

The second construction in this section is the direct product, which takes two

known groups and constructs a new one, using ordered pairs. This can be extended

to n-tuples, where the entry in the ith component comes from a group G

i

, and n-

tuples are multiplied component-by-component. This generalizes the construction

of n-dimensional vector spaces (that case is much simpler since every entry comes

from the same set).

SOLVED PROBLEMS: §3.3

16. Show that Z

5

Z

3

is a cyclic group, and list all of the generators for the

group.

17. Find the order of the element ([9]

12

, [15]

18

) in the group Z

12

Z

18

.

18. Find two groups G

1

and G

2

whose direct product G

1

G

2

has a subgroup

that is not of the form H

1

H

2

, for subgroups H

1

⊆ G

1

and H

2

⊆ G

2

.

19. In the group G = Z

×

36

, let H = ¦[x] [ x ≡ 1 (mod 4)¦ and K = ¦[y] [ y ≡

1 (mod 9)¦. Show that H and K are subgroups of G, and ﬁnd the subgroup

HK.

20. Show that if p is a prime number, then the order of the general linear group

GL

n

(Z

p

) is (p

n

−1)(p

n

−p) (p

n

−p

n−1

).

18 CHAPTER 3. GROUPS

21. Find the order of the element A =

_

_

i 0 0

0 −1 0

0 0 −i

_

_

in the group GL

3

(C).

22. Let G be the subgroup of GL

2

(R) deﬁned by

G =

__

m b

0 1

_¸

¸

¸

¸

m ,= 0

_

.

Let A =

_

1 1

0 1

_

and B =

_

−1 0

0 1

_

. Find the centralizers C(A) and

C(B), and show that C(A) ∩ C(B) = Z(G), where Z(G) is the center of G.

23. Compute the centralizer in GL

2

(Z

3

) of the matrix

_

2 1

0 2

_

.

24. Compute the centralizer in GL

2

(Z

3

) of the matrix

_

2 1

1 1

_

.

25. Let H be the following subset of the group G = GL

2

(Z

5

).

H =

__

m b

0 1

_

∈ GL

2

(Z

5

)

¸

¸

¸

¸

m, b ∈ Z

5

, m = ±1

_

(a) Show that H is a subgroup of G with 10 elements.

(b) Show that if we let A =

_

1 1

0 1

_

and B =

_

−1 0

0 1

_

, then BA = A

−1

B.

(c) Show that every element of H can be written uniquely in the form A

i

B

j

,

where 0 ≤ i < 5 and 0 ≤ j < 2.

3.4 Isomorphisms

A one-to-one correspondence φ : G

1

→ G

2

between groups G

1

and G

2

is called

a group isomorphism if φ(ab) = φ(a)φ(b) for all a, b ∈ G

1

. The function φ can

be thought of as simply renaming the elements of G

1

, since it is one-to-one and

onto. The condition that φ(ab) = φ(a)φ(b) for all a, b ∈ G

1

makes certain that

multiplication can be done in either group and the transferred to the other, since

the inverse function φ

−1

also respects the multiplication of the two groups.

In terms of the respective group multiplication tables for G

1

and G

2

, the exis-

tence of an isomorphism guarantees that there is a way to set up a correspondence

between the elements of the groups in such a way that the group multiplication

tables will look exactly the same.

3.4. ISOMORPHISMS 19

From an algebraic perspective, we should think of isomorphic groups as being

essentially the same. The problem of ﬁnding all abelian groups of order 8 is im-

possible to solve, because there are inﬁnitely many possibilities. But if we ask for

a list of abelian groups of order 8 that comes with a guarantee that any possible

abelian group of order 8 must be isomorphic to one of the groups on the list, then

the question becomes manageable. In fact, we can show (in Section 7.5) that the

answer to this particular question is the list Z

8

, Z

4

Z

2

, Z

2

Z

2

Z

2

. In this

situation we would usually say that we have found all abelian groups of order 8, up

to isomorphism.

To show that two groups G

1

and G

2

are isomorphic, you should actually produce

an isomorphism φ : G

1

→G

2

. To decide on the function to use, you probably need

to see some similarity between the group operations.

In some ways it is harder to show that two groups are not isomorphic. If you can

show that one group has a property that the other one does not have, then you can

decide that two groups are not isomorphic (provided that the property would have

been transferred by any isomorphism). Suppose that G

1

and G

2

are isomorphic

groups. If G

1

is abelian, then so is G

2

; if G

1

is cyclic, then so is G

2

. Furthermore,

for each positive integer n, the two groups must have exactly the same number of

elements of order n. Each time you meet a new property of groups, you should ask

whether it is preserved by any isomorphism.

SOLVED PROBLEMS: §3.4

21. Show that Z

×

17

is isomorphic to Z

16

.

22. Let φ : R

×

→ R

×

be deﬁned by φ(x) = x

3

, for all x ∈ R. Show that φ is a

group isomorphism.

23. Let G

1

, G

2

, H

1

, H

2

be groups, and suppose that θ

1

: G

1

→ H

1

and θ

2

:

G

2

→ H

2

are group isomorphisms. Deﬁne φ : G

1

G

2

→ H

1

H

2

by

φ(x

1

, x

2

) = (θ

1

(x

1

), θ

2

(x

2

)), for all (x

1

, x

2

) ∈ G

1

G

2

. Prove that φ is a

group isomorphism.

24. Prove that the group Z

×

7

Z

×

11

is isomorphic to the group Z

6

Z

10

.

25. Deﬁne φ : Z

30

Z

2

→ Z

10

Z

6

by φ([n]

30

, [m]

2

) = ([n]

10

, [4n + 3m]

6

), for

all ([n]

30

, [m]

2

) ∈ Z

30

Z

2

. First prove that φ is a well-deﬁned function, and

then prove that φ is a group isomorphism.

26. Let G be a group, and let H be a subgroup of G. Prove that if a is any

element of G, then the subset

aHa

−1

= ¦g ∈ G [ g = aha

−1

for some h ∈ H¦

is a subgroup of G that is isomorphic to H.

20 CHAPTER 3. GROUPS

27. Let G, G

1

, G

2

be groups. Prove that if G is isomorphic to G

1

G

2

, then there

are subgroups H and K in G such that H ∩K = ¦e¦, HK = G, and hk = kh

for all h ∈ H and k ∈ K.

28. Show that for any prime number p, the subgroup of diagonal matrices in

GL

2

(Z

p

) is isomorphic to Z

×

p

Z

×

p

.

29. (a) In the group G = GL

2

(R) of invertible 2 2 matrices with real entries,

show that

H =

__

a

11

a

12

a

21

a

22

_

∈ GL

2

(R)

¸

¸

¸

¸

a

11

= 1, a

21

= 0, a

22

= 1

_

is a subgroup of G.

(b) Show that H is isomorphic to the group R of all real numbers, under

addition.

30. Let G be the subgroup of GL

2

(R) deﬁned by

G =

__

m b

0 1

_¸

¸

¸

¸

m ,= 0

_

.

Show that G is not isomorphic to the direct product R

×

R.

31. Let H be the following subgroup of group G = GL

2

(Z

3

).

H =

__

m b

0 1

_

∈ GL

2

(Z

3

)

¸

¸

¸

¸

m, b ∈ Z

3

, m ,= 0

_

Show that H is isomorphic to the symmetric group o3

.

32. Let G be a group, and let S be any set for which there exists a one-to-

one and onto function φ : G → S. Deﬁne an operation on S by setting

x

1

x

2

= φ(φ

−1

(x

1

)φ

−1

(x

2

)), for all x

1

, x

2

∈ S. Prove that S is a group under

this operation, and that φ is actually a group isomorphism.

3.5 Cyclic Groups

We began our study of abstract algebra very concretely, by looking at the group Z

of integers, and the related groups Z

n

. We discovered that each of these groups is

generated by a single element, and this motivated the deﬁnition of an abstract cyclic

group. In this section, Theorem 3.5.2 shows that every cyclic group is isomorphic

to one of these concrete examples, so all of the information about cyclic groups is

already contained in these basic examples.

You should pay particular attention to Proposition 3.5.3, which describes the

subgroups of Z

n

, showing that they are in one-to-one correspondence with the

3.6. PERMUTATION GROUPS 21

positive divisors of n. In n is a prime power, then the subgroups are “linearly

ordered” in the sense that given any two subgroups, one is a subset of the other.

These cyclic groups have a particularly simple structure, and form the basic building

blocks for all ﬁnite abelian groups. (In Theorem 7.5.4 we will prove that every ﬁnite

abelian group is isomorphic to a direct product of cyclic groups of prime power

order.)

SOLVED PROBLEMS: §3.5

20. Show that the three groups Z

6

, Z

×

9

, and Z

×

18

are isomorphic to each other.

21. Is Z

4

Z

10

isomorphic to Z

2

Z

20

?

22. Is Z

4

Z

15

isomorphic to Z

6

Z

10

?

23. Give the lattice diagram of subgroups of Z

100

.

24. Find all generators of the cyclic group Z

28

.

25. In Z

30

, ﬁnd the order of the subgroup ¸[18]

30

); ﬁnd the order of ¸[24]

30

).

26. Prove that if G

1

and G

2

are groups of order 7 and 11, respectively, then the

direct product G

1

G

2

is a cyclic group.

27. Show that any cyclic group of even order has exactly one element of order 2.

28. Use the the result in Problem 27 to show that the multiplicative groups Z

×

15

and Z

×

21

are not cyclic groups.

29. Find all cyclic subgroups of the quaternion group. Use this information to

show that the quaternion group cannot be isomorphic to the subgroup of o4

generated by (1, 2, 3, 4) and (1, 3).

30. Prove that if p and q are diﬀerent odd primes, then Z

×

pq

is not a cyclic group.

3.6 Permutation Groups

As with the previous section, this section revisits the roots of group theory that

we began to study in an earlier chapter. Cayley’s theorem shows that permutation

groups contain all of the information about ﬁnite groups, since every ﬁnite group

of order n is isomorphic to a subgroup of the symmetric group on

. That isn’t as

impressive as it sounds at ﬁrst, because as n gets larger and larger, the subgroups

of order n just get lost inside the larger symmetric group, which has order n!. This

does imply, however, that from the algebraists point of view the abstract deﬁnition

of a group is really no more general than the concrete deﬁnition of a permutation

22 CHAPTER 3. GROUPS

group. The abstract deﬁnition of a group is useful simply because it can be more

easily applied to a wide variety of situation.

You should make every eﬀort to get to know the dihedral groups Tn

. They have

a concrete representation, in terms of the rigid motions of an n-gon, but can also be

described more abstractly in terms of two generators a (of order n) and b (of order

2) which satisfy the relation ba = a

−1

b. We can write

Tn

= ¦a

i

b

j

[ 0 ≤ i < n, 0 ≤ j < 2, with o(a) = n, o(b) = 2, and ba = a

−1

b¦ .

In doing computations in Tn

it is useful to have at hand the formula ba

i

= a

n−i

b,

shown in the ﬁrst of the solved problems given below.

SOLVED PROBLEMS: §3.6

22. In the dihedral group Tn

= ¦a

i

b

j

[ 0 ≤ i < n, 0 ≤ j < 2¦ with o(a) = n,

o(b) = 2, and ba = a

−1

b, show that ba

i

= a

n−i

b, for all 0 ≤ i < n.

23. In the dihedral group Tn

= ¦a

i

b

j

[ 0 ≤ i < n, 0 ≤ j < 2¦ with o(a) = n,

o(b) = 2, and ba = a

−1

b, show that each element of the form a

i

b has order 2.

24. In o4

, ﬁnd the subgroup H generated by (1, 2, 3) and (1, 2).

25. For the subgroup H of o4

deﬁned in the previous problem, ﬁnd the corre-

sponding subgroup σHσ

−1

, for σ = (1, 4).

26. Show that each element in /4

can be written as a product of 3-cycles.

27. In the dihedral group Tn

= ¦a

i

b

j

[ 0 ≤ i < n, 0 ≤ j < 2¦ with o(a) = n,

o(b) = 2, and ba = a

−1

b, ﬁnd the centralizer of a.

28. Find the centralizer of (1, 2, 3) in o3

, in o4

, and in /4

.

3.7 Homomorphisms

In Section 3.4 we introduced the concept of an isomorphism, and studied in detail

what it means for two groups to be isomorphic. In this section we look at functions

that respect the group operations but may not be one-to-one and onto. There are

many important examples of group homomorphisms that are not isomorphisms,

and, in fact, homomorphisms provide the way to relate one group to another.

The most important result in this section is Theorem 3.7.8, which is a prelim-

inary form of the Fundamental Homomorphism Theorem. (The full statement is

given in Theorem 3.8.8, after we develop the concepts of cosets and factor groups.)

In this formulation of the Fundamental Homomorphism Theorem, we start with a

group homomorphism φ : G

1

→ G

2

. It is easy to prove that the image φ(G

1

) is

3.7. HOMOMORPHISMS 23

a subgroup of G

2

. The function φ has an equivalence relation associated with it,

where we let a ∼ b if φ(a) = φ(b), for a, b ∈ G

1

. Just as in Z, where we use the

equivalence relation deﬁned by congruence modulo n, we can deﬁne a group opera-

tion on the equivalence classes of ∼, using the operation in G

1

. Then Theorem 3.7.8

shows that this group is isomorphic to φ(G

1

), so that although the homomorphism

may not be an isomorphism between G

1

and G

2

, it does deﬁne an isomorphism

between a subgroup of G

2

and what we call a factor group of G

1

.

Proposition 3.7.6 is also useful, since for any group homomorphism φ : G

1

→

G

2

it describes the connections between subgroups of G

1

and subgroups of G

2

.

Examples 3.7.4 and 3.7.5 are important, because they give a complete description

of all group homomorphisms between two cyclic groups.

SOLVED PROBLEMS: §3.7

17. Find all group homomorphisms from Z

4

into Z

10

.

18. (a) Find the formulas for all group homomorphisms from Z

18

into Z

30

.

(b) Choose one of the nonzero formulas in part (a), and for this formula ﬁnd

the kernel and image, and show how elements of the image correspond to

cosets of the kernel.

19. (a) Show that Z

×

7

is cyclic, with generator [3]

7

.

(b) Show that Z

×

17

is cyclic, with generator [3]

17

.

(c) Completely determine all group homomorphisms from Z

×

17

into Z

×

7

.

20. Deﬁne φ : Z

4

Z

6

→Z

4

Z

3

by φ(x, y) = (x + 2y, y).

(a) Show that φ is a well-deﬁned group homomorphism.

(b) Find the kernel and image of φ, and apply the fundamental homomorphism

theorem.

21. Let n and m be positive integers, such that m is a divisor of n. Show that

φ : Z

×

n

→ Z

×

m

deﬁned by φ([x]

n

) = [x]

m

, for all [x]

n

∈ Z

×

n

, is a well-deﬁned

group homomorphism.

22. For the group homomorphism φ : Z

×

36

→ Z

×

12

deﬁned by φ([x]

36

) = [x]

12

, for

all [x]

36

∈ Z

×

36

, ﬁnd the kernel and image of φ, and apply the fundamental

homomorphism theorem.

23. Let G, G

1

, and G

2

be groups. Let φ

1

: G → G

1

and φ

2

: G → G

2

be

group homomorphisms. Prove that φ : G → G

1

G

2

deﬁned by φ(x) =

(φ

1

(x), φ

2

(x)), for all x ∈ G, is a well-deﬁned group homomorphism.

24. Let p and q be diﬀerent odd primes. Prove that Z

×

pq

is isomorphic to the direct

product Z

×

p

Z

×

q

.

24 CHAPTER 3. GROUPS

3.8 Cosets, Normal Subgroups, and Factor Groups

The notion of a factor group is one of the most important concepts in abstract

algebra. To construct a factor group, we start with a normal subgroup and the

equivalence classes it determines. This construction parallels the construction of

Z

n

from Z, where we have a ≡ b (mod n) if and only if a − b ∈ nZ. The only

complication is that the equivalence relation respects the operation in G only when

the subgroup is a normal subgroup. Of course, in an abelian group we can use any

subgroup, since all subgroups of an abelian group are normal.

The key idea is to begin thinking of equivalence classes as elements in their own

right. That is what we did in Chapter 1, where at ﬁrst we thought of congruence

classes as inﬁnite sets of integers, and then in Section 1.4 when we started working

with Z

n

we started to use the notation [a]

n

to suggest that we were now thinking

of a single element of a set.

In actually using the Fundamental Homomorphism Theorem, it is important to

let the theorem do its job, so that it does as much of the hard work as possible.

Quite often we need to show that a factor group G/N that we have constructed

is isomorphic to another group G

1

. The easiest way to do this is to just deﬁne a

homomorphism φ from G to G

1

, making sure that N is the kernel of φ. If you prove

that φ maps G onto G

1

, then the Fundamental Theorem does the rest of the work,

showing that there exists a well-deﬁned isomorphism between G/N and G

1

.

The moral of this story is that if you deﬁne a function on G rather than G/N,

you ordinarily don’t need to worry that it is well-deﬁned. On the other hand, if you

deﬁne a function on the cosets of G/N, the most convenient way is use a formula

deﬁned on representatives of the cosets of N. But then you must be careful to

prove that the formula you are using does not depend on the particular choice of

a representative. That is, you must prove that your formula actually deﬁnes a

function. Then you must prove that your function is one-to-one, in addition to

proving that it is onto and respects the operations in the two groups. Once again,

if your function is deﬁned on cosets, it can be much trickier to prove that it is

one-to-one than to simply compute the kernel of a homomorphism deﬁned on G.

SOLVED PROBLEMS: §3.8

27. List the cosets of ¸7) in Z

×

16

. Is the factor group Z

×

16

/ ¸7) cyclic?

28. Let G = Z

6

Z

4

, let H = ¦(0, 0), (0, 2)¦, and let K = ¦(0, 0), (3, 0)¦.

(a) List all cosets of H; list all cosets of K.

(b) You may assume that any abelian group of order 12 is isomorphic to either

Z

12

or Z

6

Z

2

. Which answer is correct for G/H? For G/K?

29. Let the dihedral group D

n

be given via generators and relations, with gener-

ators a of order n and b of order 2, satisfying ba = a

−1

b.

3.8. COSETS, NORMAL SUBGROUPS, AND FACTOR GROUPS 25

(a) Show that ba

i

= a

−i

b for all i with 1 ≤ i < n.

(b) Show that any element of the form a

i

b has order 2.

(c) List all left cosets and all right cosets of ¸b)

30. Let G = D

6

and let N be the subgroup

¸

a

3

_

= ¦e, a

3

¦ of G.

(a) Show that N is a normal subgroup of G.

(b) Is G/N abelian?

31. Let G be the dihedral group D

12

, and let N = ¦e, a

3

, a

6

, a

9

¦.

(a) Prove that N is a normal subgroup of G, and list all cosets of N.

(b) You may assume that G/N is isomorphic to either Z

6

or S

3

. Which is

correct?

32. (a) Let G be a group. For a, b ∈ G we say that b is conjugate to a, written

b ∼ a, if there exists g ∈ G such that b = gag

−1

. Show that ∼ is an equivalence

relation on G. The equivalence classes of ∼ are called the conjugacy classes

of G.

(b) Show that a subgroup N of G is normal in G if and only if N is a union

of conjugacy classes.

33. Find the conjugacy classes of D

4

.

34. Let G be a group, and let N and H be subgroups of G such that N is normal

in G.

(a) Prove that HN is a subgroup of G.

(b) Prove that N is a normal subgroup of HN.

(c) Prove that if H ∩ N = ¦e¦, then HN/N is isomorphic to H.

26 CHAPTER 3. GROUPS

Review Problems

1. (a) What are the possibilities for the order of an element of Z

×

13

? Explain

your answer.

(b) Show that Z

×

13

is a cyclic group.

2. Find all subgroups of Z

×

11

, and give the lattice diagram which shows the

inclusions between them.

3. Let G be the subgroup of GL

3

(R) consisting of all matrices of the form

_

_

1 a b

0 1 0

0 0 1

_

_

such that a, b ∈ R .

Show that G is a subgroup of GL

3

(R).

4. Show that the group G in the previous problem is isomorphic to the direct

product RR.

5. List the cosets of the cyclic subgroup ¸9) in Z

×

20

. Is Z

×

20

/ ¸9) cyclic?

6. Let G be the subgroup of GL

2

(R) consisting of all matrices of the form

_

m b

0 1

_

, and let N be the subset of all matrices of the form

_

1 b

0 1

_

.

(a) Show that N is a subgroup of G, and that N is normal in G.

(b) Show that G/N is isomorphic to the multiplicative group R

×

.

7. Assume that the dihedral group D

4

is given as ¦e, a, a

2

, a

3

, b, ab, a

2

b, a

3

b¦,

where a

4

= e, b

2

= e, and ba = a

3

b. Let N be the subgroup

¸

a

2

_

= ¦e, a

2

¦.

(a) Show by a direct computation that N is a normal subgroup of D

4

.

(b) Is the factor group D

4

/N a cyclic group?

8. Let G = D

8

, and let N = ¦e, a

2

, a

4

, a

6

¦.

(a) List all left cosets and all right cosets of N, and verify that N is a normal

subgroup of G.

(b) Show that G/N has order 4, but is not cyclic.

Chapter 4

POLYNOMIALS

In this chapter we return to several of the themes in Chapter 1. We need to talk

about the greatest common divisor of two polynomials, and when two polynomials

are relatively prime. The notion of a prime number is replaced by that of an

irreducible polynomial. We can work with congruence classes of polynomials, just

as we did with congruence classes of integers. The point of saying this is that it will

be worth your time to review the deﬁnitions and theorems in Chapter 1.

In addition to generalizing ideas from the integers to polynomials, we want to

go beyond high school algebra, to be able to work with coeﬃcients that may not be

real numbers. This motivates the deﬁnition of a ﬁeld, which is quite closely related

to the deﬁnition of a group (now there are two operations instead of just one). The

point here is that you can beneﬁt from reviewing Chapter 3.

Because you have a lot more experience now than when you started Chapter 1,

I didn’t break the problems up by section. Of course, you don’t have to wait until

you have ﬁnished the chapter to practice solving some of these problems.

Review Problems

1. Use the Euclidean algorithm to ﬁnd gcd(x

8

−1, x

6

−1) in Q[x] and write it

as a linear combination of x

8

−1 and x

6

−1.

2. Over the ﬁeld of rational numbers, use the Euclidean algorithm to show that

2x

3

−2x

2

−3x + 1 and 2x

2

−x −2 are relatively prime.

3. Over the ﬁeld of rational numbers, ﬁnd the greatest common divisor of

x

4

+ x

3

+ 2x

2

+ x + 1 and x

3

− 1, and express it as a linear combination of

the given polynomials.

4. Over the ﬁeld of rational numbers, ﬁnd the greatest common divisor of

2x

4

− x

3

+ x

2

+ 3x + 1 and 2x

3

− 3x

2

+ 2x + 2 and express it as a linear

combination of the given polynomials.

27

28 CHAPTER 4. POLYNOMIALS

5. Are the following polynomials irreducible over Q?

(a) 3x

5

+ 18x

2

+ 24x + 6

(b) 7x

3

+ 12x

2

+ 3x + 45

(c) 2x

10

+ 25x

3

+ 10x

2

−30

6. Factor x

5

−10x

4

+ 24x

3

+ 9x

2

−33x −12 over Q.

7. Factor x

5

−2x

4

−2x

3

+ 12x

2

−15x −2 over Q.

8. (a) Show that x

2

+ 1 is irreducible over Z

3

.

(b) List the elements of the ﬁeld F = Z

3

[x]/

¸

x

2

+ 1

_

.

(c) In the multiplicative group of nonzero elements of F, show that [x + 1] is

a generator, but [x] is not.

9. (a) Express x

4

+x as a product of polynomials irreducible over Z

5

.

(b) Show that x

3

+ 2x

2

+ 3 is irreducible over Z

5

.

10. Express 2x

3

+x

2

+ 2x + 2 as a product of polynomials irreducible over Z

5

.

11. Construct an example of a ﬁeld with 343 = 7

3

elements.

12. In Z

2

[x]/

¸

x

3

+x + 1

_

, ﬁnd the multiplicative inverse of [x + 1].

13. Find the multiplicative inverse of [x

2

+x + 1]

(a) in Q[x]/

¸

x

3

−2

_

;

(b) in Z

3

[x]/

¸

x

3

+ 2x

2

+x + 1

_

.

14. In Z

5

[x]/

¸

x

3

+x + 1

_

, ﬁnd [x]

−1

and [x +1]

−1

, and use your answers to ﬁnd

[x

2

+x]

−1

.

15. Factor x

4

+x + 1 over Z

2

[x]/

¸

x

4

+x + 1

_

.

Chapter 5

COMMUTATIVE RINGS

This chapter takes its motivation from Chapter 1 and Chapter 4, extending results

on factorization to more general settings than just the integers or polynomials over

a ﬁeld. The concept of a factor ring depends heavily on the corresponding deﬁnition

for groups, so you may need to review the last two sections of Chapter 3. Remember

that the distributive law is all that connects the two operations in a ring, so it is

crucial in many of the proofs you will see.

Review Problems

1. Let R be the ring with 8 elements consisting of all 3 3 matrices with entries

in Z

2

which have the following form:

_

_

a 0 0

0 a 0

b c a

_

_

You may assume that the standard laws for addition and multiplication of

matrices are valid.

(a) Show that R is a commutative ring (you only need to check closure and

commutativity of multiplication).

(b) Find all units of R, and all nilpotent elements of R.

(c) Find all idempotent elements of R.

2. Let R be the ring Z

2

[x]/

¸

x

2

+ 1

_

. Show that although R has 4 elements, it

is not isomorphic to either of the rings Z

4

or Z

2

⊕Z

2

.

3. Find all ring homomorphisms from Z

120

into Z

42

.

4. Are Z

9

and Z

3

⊕Z

3

isomorphic as rings?

29

30 CHAPTER 5. COMMUTATIVE RINGS

5. In the group Z

×

180

of units of the ring Z

180

, what is the largest possible order

of an element?

6. For the element a = (0, 2) of the ring R = Z

12

⊕ Z

8

, ﬁnd Ann(a) = ¦r ∈ R [

ra = 0¦. Show that Ann(a) is an ideal of R.

7. Let R be the ring Z

2

[x]/

¸

x

4

+ 1

_

, and let I be the set of all congruence classes

in R of the form [f(x)(x

2

+ 1)].

(a) Show that I is an ideal of R.

(b) Show that R/I

∼

= Z

2

[x]/

¸

x

2

+ 1

_

.

(c) Is I a prime ideal of R?

Hint: If you use the fundamental homomorphism theorem, you can do the

ﬁrst two parts together.

8. Find all maximal ideals, and all prime ideals, of Z

36

= Z/36Z.

9. Give an example to show that the set of all zero divisors of a ring need not

be an ideal of the ring.

10. Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients.

Prove that I is a prime ideal; prove that I is not maximal.

11. Let R be any commutative ring with identity 1.

(a) Show that if e is an idempotent element of R, then 1−e is also idempotent.

(b) Show that if e is idempotent, then R

∼

= Re ⊕R(1 −e).

12. Let R be the ring Z

2

[x]/

¸

x

3

+ 1

_

.

(a) Find all ideals of R.

(b) Find the units of R.

(c) Find the idempotent elements of R.

13. Let S be the ring Z

2

[x]/

¸

x

3

+x

_

.

(a) Find all ideals of S.

(b) Find the units of R.

(c) Find the idempotent elements of R.

14. Show that the rings R and S in the two previous problems are isomorphic as

abelian groups, but not as rings.

15. Let Z[i] be the subring of the ﬁeld of complex numbers given by

Z[i] = ¦m+ni ∈ C [ m, n ∈ Z¦ .

31

(a) Deﬁne φ : Z[i] → Z

2

by φ(m + ni) = [m + n]

2

. Prove that φ is a ring

homomorphism. Find ker(φ) and show that it is a principal ideal of Z[i].

(b) For any prime number p, deﬁne θ : Z[i] →Z

p

[x]/

¸

x

2

+ 1

_

by θ(m+ni) =

[m+nx]. Prove that θ is an onto ring homomorphism.

16. Let I and J be ideals in the commutative ring R, and deﬁne the function

φ : R →R/I ⊕R/J by φ(r) = (r +I, r +J), for all r ∈ R.

(a) Show that φ is a ring homomorphism, with ker(φ) = I ∩ J.

(b) Show that if I +J = R, then φ is onto, and thus R/(I ∩J)

∼

= R/I ⊕R/J.

17. Considering Z[x] to be a subring of Q[x], show that these two integral domains

have the same quotient ﬁeld.

18. Let p be an odd prime number that is not congruent to 1 modulo 4. Prove

that the ring Z

p

[x]/

¸

x

2

+ 1

_

is a ﬁeld.

Hint: Show that a root of x

2

= −1 leads to an element of order 4 in the

multiplicative group Z

×

p

.

32 CHAPTER 5. COMMUTATIVE RINGS

Chapter 6

FIELDS

These review problems cover only the ﬁrst three sections of the chapter. If you

are studying abstract algebra because you plan to be a high school teacher, it is

precisely these sections (along with the earlier material on polynomials) that are

the most relevant to what you will be teaching.

Review Problems

1. Let u be a root of the polynomial x

3

+3x+3. In Q(u), express (7−2u+u

2

)

−1

in the form a +bu +cu

2

.

2. (a) Show that Q(

√

2 +i) = Q(

√

2, i).

(b) Find the minimal polynomial of

√

2 +i over Q.

3. Find the minimal polynomial of 1 +

3

√

2 over Q.

4. Show that x

3

+ 6x

2

− 12x + 2 is irreducible over Q, and remains irreducible

over Q(

5

√

2).

5. Find a basis for Q(

√

5,

3

√

5) over Q.

6. Show that [Q(

√

2 +

3

√

5) : Q] = 6.

7. Find [Q(

7

√

16 + 3

7

√

8) : Q].

8. Find the degree of

3

√

2 +i over Q. Does

4

√

2 belong to Q(

3

√

2 +i)?

33

34 CHAPTER 6. FIELDS

Chapter 1

Integers

1.1 SOLUTIONS

22. Find gcd(435, 377), and express it as a linear combination of 435 and 377.

Comment: You deﬁnitely need to know how to do these computations.

Solution: We will use the Euclidean algorithm. Divide the larger number

by the smaller, which should give you a quotient of 1 and a remainder of 58.

Then divide the remainder 58 into 377, and continue the Euclidean algorithm

as in Example 1.1.4 in the text. That should give you the following equations.

435 = 1 377 + 58 gcd(435, 377) = gcd(377, 58)

377 = 6 58 + 29 = gcd(58, 29)

58 = 2 29 = 29

The repeated divisions show that gcd(435, 377) = 29, since the remainder in

the last equation is 0. To write 29 as a linear combination of 435 and 377 we

need to use the same equations, but we need to solve them for the remainders.

58 = 435 −1 377

29 = 377 −6 58

Now take the equation involving the remainder 29, and substitute for 58, the

remainder in the previous equation.

29 = 377 −6 58

= 377 −6 (435 −1 377)

= 7 377 −6 435

This gives the linear combination we need, 29 = (7)(377) −(6)(435).

35

36 CHAPTER 1 SOLUTIONS

23. Find gcd(3553, 527), and express it as a linear combination of 3553 and 527.

Comment: This time we will use the matrix form of the Euclidean algorithm.

You should be able to use both the back-solving form (as in Problem 22)

and the matrix form. In Chapter 4, the Euclidean algorithm is used for

polynomials, and the matrix method just gets too complicated, so we have to

adapt the back-solving method.

Solution: Just as in Problem 22, the ﬁrst step is to divide the smaller number

into the larger. We get 3553 = 6 527 + 391, so this tells us to multiply the

bottom row of the matrix

_

1 0 3553

0 1 527

_

by 6 and subtract from the ﬁrst

row. The rest of the steps in reducing the matrix to the form we want should

be clear. We have

_

1 0 3553

0 1 527

_

;

_

1 −6 391

0 1 527

_

;

_

1 −6 391

−1 7 136

_

;

_

3 −20 119

−1 7 136

_

;

_

3 −20 119

−4 27 17

_

;

_

31 −209 0

−4 27 17

_

.

Therefore gcd(3553, 527) = 17, and 17 = (−4)(3553) + (27)(527).

24. Which of the integers 0, 1, . . . , 10 can be expressed in the form 12m + 20n,

where m, n are integers?

Solution: Theorem 1.1.6 provides the answer. An integer k is a linear com-

bination of 12 and 20 if and only if it is a multiple of their greatest common

divisor, which is 4. Therefore we can express 0, 4, and 8 in the required form,

but we can’t do it for the rest.

Comment: Check out the answer in concrete terms. We can write

0 = 12 0 + 20 0; 4 = 12 2 + 20 (−1); 8 = 12 (−1) + 20 1.

25. If n is a positive integer, ﬁnd the possible values of gcd(n, n + 10).

Solution: Let d = gcd(n, n + 10). Then d[n and d[(n + 10), so we must have

d[10, and therefore d is limited to one of 1, 2, 5, or 10. Can each of these

occur for some n?

Yes: gcd(3, 13) = 1; gcd(2, 12) = 2; gcd(5, 15) = 5; gcd(10, 20) = 10.

26. Prove that if a and b are nonzero integers for which a[b and b[a, then b = ±a.

Comment: The ﬁrst step is to use Deﬁnition 1.1.1 to rewrite a[b and b[a as

equations, to give something concrete to work with.

Solution: Since a [ b, there is an integer m with b = ma. Since b [ a, there is

an integer k with a = kb. Substituting a = kb in the equation b = ma we get

b = m(kb), so since b is nonzero we can cancel it to get 1 = mk. Since both

m and k are integers, and [1[ = [m[ [k[, we must have [m[ = 1 and [k[ = 1, so

either b = a or b = −a.

CHAPTER 1 SOLUTIONS 37

27. Prove that if m and n are odd integers, then m

2

−n

2

is divisible by 8.

Solution: First, we need to use the given information about m and n. Since

they are odd, we can write them in the form m = 2k + 1 and n = 2q + 1,

for some integers k and q. We can factor m

2

−n

2

to get (m+ n)(m−n), so

substituting for m and n we get

m

2

−n

2

= (2k + 1 + 2q + 1)(2k + 1 −2q −1) = (2)(k +q + 1)(2)(k −q) .

Now we need to take two cases. If k −q is even, then k −q has 2 as a factor,

say k −q = 2p, for some integer p. Substituting for k −q gives us

m

2

−n

2

= (2)(k +q + 1)(2)(2)(p) = (8)(k +q + 1)(p) .

If k −q is odd, then k + q = (k −q) + (2q) is the sum of an odd integer and

an even integer, so it must also be odd. That means that k +q +1 is even, so

it has 2 as a factor. Now we can suppose that k +q +1 = 2t, for some integer

t. In this case, substituting for k +q + 1 gives us

m

2

−n

2

= (2)(2)(t)(2)(k −q) = (8)(t)(k −q) .

Showing that we can factor 8 out of m

2

− n

2

gives exactly what we were to

prove: if m and n are odd, then m

2

−n

2

is divisible by 8.

28. Prove that if n is an integer with n > 1, then gcd(n − 1, n

2

+ n + 1) = 1 or

gcd(n −1, n

2

+n + 1) = 3.

Comment: It’s not a bad idea to check this out for some values of n, just to

get a feeling for the problem. For n = 3, we have gcd(2, 13) = 1. For n = 4,

we have gcd(3, 21) = 3. For n = 5, we have gcd(4, 31) = 1. For n = 6, we

have gcd(5, 43) = 1. For n = 7, we have gcd(6, 57) = 1. These calculations

don’t prove anything, but maybe they do make the problem look plausible.

Solution: Problem 25 gives a hint. In that problem, since the gcd was a divisor

of n and n + 10, it had to be a divisor of 10. To use the same approach, we

would have to write n

2

+ n + 1 as n −1 plus something. That doesn’t work,

but we are very close. Dividing n

2

+ n + 1 by n − 1 (using long division of

polynomials) we get a quotient of n+2 and a remainder of 3, so n

2

+n+1 =

(n + 2)(n − 1) + 3. Now we can see that any common divisor of n − 1 and

n

2

+n + 1 must be a divisor of 3, so the answer has to be 1 or 3.

29. Prove that if n is a positive integer, then

_

_

0 0 −1

0 1 0

1 0 0

_

_

n

=

_

_

1 0 0

0 1 0

0 0 1

_

_

if and only if 4[n.

Comment: Let’s use A for the matrix, and I for the identity matrix. The

proof must be given in two pieces. We need to show that if 4[n, then A

n

= I.

38 CHAPTER 1 SOLUTIONS

We also need to show that A

n

= I only when 4[n, and it is easier to state

as the converse of the ﬁrst statement: if A

n

= I, then 4[n. The ﬁrst half of

the proof is easier than the second, since it just takes a computation. In the

second half of the proof, if A

n

= I then we will use the division algorithm, to

divide n by 4, and then show that the remainder has to be 0.

Solution: We begin by computing A

2

, A

3

= A A

2

, A

4

= A A

3

, etc.

_

_

0 0 −1

0 1 0

1 0 0

_

_

2

=

_

_

0 0 −1

0 1 0

1 0 0

_

_

_

_

0 0 −1

0 1 0

1 0 0

_

_

=

_

_

−1 0 0

0 1 0

0 0 −1

_

_

_

_

0 0 −1

0 1 0

1 0 0

_

_

3

=

_

_

0 0 −1

0 1 0

1 0 0

_

_

_

_

−1 0 0

0 1 0

0 0 −1

_

_

=

_

_

0 0 1

0 1 0

−1 0 0

_

_

_

_

0 0 −1

0 1 0

1 0 0

_

_

4

=

_

_

0 0 −1

0 1 0

1 0 0

_

_

_

_

0 0 1

0 1 0

−1 0 0

_

_

=

_

_

1 0 0

0 1 0

0 0 1

_

_

Now we can see that if 4[n, say n = 4q, then A

n

= A

4q

= (A

4

)

q

= I

q

= I.

Conversely, if A

n

= I, we can use the division algorithm to write n = 4q +r,

with 0 ≤ r < 4. Then A

r

= A

n−4q

= A

n

(A

−4

)

q

= I I

q

= I, so r = 0 since

A, A

2

, and A

3

are not equal to I. We conclude that 4[n.

30. Give a proof by induction to show that each number in the sequence 12, 102,

1002, 10002, . . ., is divisible by 6.

Comment: If you are unsure about doing a proof by induction, you should

read Appendix 4 in the text.

Solution: To give a proof by induction, we need a statement that depends

on an integer n. We can write the numbers in the given sequence in the form

10

n

+ 2, for n = 1, 2, . . ., so we can prove the following statement: for each

positive integer n, the integer 10

n

+ 2 is divisible by 6.

The ﬁrst step is to check that the statement is true for n = 1. (This “anchors”

the induction argument.) Clearly 12 is divisible by 6.

The next step is to prove that if we assume that the statement is true for

n = k, then we can show that the statement must also be true for n = k + 1.

Let’s start by assuming that 10

k

+ 2 is divisible by 6, say 10

k

+ 2 = 6q, for

some q ∈ Z, and then look at the expression when n = k + 1. We can easily

factor a 10 out of 10

k+1

, to get 10

k+1

+ 2 = (10)(10

k

) + 2, but we need to

involve the expression 10

k

+2 in some way. Adding and subtracting 20 makes

it possible to get this term, and then it turns out that we can factor out 6.

10

k+1

+ 2 = (10)(10

k

) + 20 −20 + 2 = (10)(10

k

+ 2) −18

= (10)(6q) −(6)(3) = (6)(10q −3)

CHAPTER 1 SOLUTIONS 39

We have now shown that if 10

k

+2 is divisible by 6, then 10

k+1

+2 is divisible

by 6. This completes the induction.

1.2 SOLUTIONS

23. (a) Use the Euclidean algorithm to ﬁnd gcd(1776, 1492).

Solution: We have 1776 = 1492 1 + 284; 1492 = 284 5 + 72;

284 = 72 3 + 68; 72 = 68 1 + 4; 68 = 4 17. Thus gcd(1776, 1492) = 4.

(b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776, 1492).

Solution: Since 1776 = 2

4

3 37 and 1492 = 2

2

373, Proposition 1.2.9 shows

that gcd(1776, 1492) = 2

2

.

24. (a) Use the Euclidean algorithm to ﬁnd gcd(1274, 1089).

Solution: We have 1274 = 1089 1 + 185; 1089 = 185 5 + 164;

185 = 164 1 +21; 164 = 21 7 +17; 21 = 17 1 +4; 17 = 4 4 +1. Thus

gcd(1274, 1089) = 1.

(b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274, 1089).

Solution: Since 1274 = 2 7

2

13 and 1089 = 3

2

11

2

, we see that 1274 and

1089 are relatively prime.

25. Give the lattice diagram of all divisors of 250. Do the same for 484.

Solution: The prime factorizations are 250 = 2 5

3

and 484 = 2

2

11

2

. In each

diagram, we need to use one axis for each prime. Then we can just divide

(successively) by the prime, to give the factors along the corresponding axis.

For example, dividing 250 by 5 produces 50, 10, and 2, in succession. These

numbers go along one axis of the rectangular diagram.

250 484

125 50 242 44

25 10 121 22 4

5 2 11 2

1 1

40 CHAPTER 1 SOLUTIONS

26. Find all integer solutions of the equation xy + 2y −3x = 25.

Solution: If we had a product, we could use the prime factorization theorem.

That motivates one possible method of solution.

xy + 2y −3x = 25

(x + 2)y −3x = 25

(x + 2)y −3x −6 = 25 −6

(x + 2)y −3(x + 2) = 19

(x + 2)(y −3) = 19

Now since 19 is prime, the only way it can be factored is to have 1 19 =

19 or (−1) (−19) = 19. Therefore we have 4 possibilities: x + 2 = 1,

x + 2 = −1, x + 2 = 19, or x + 2 = −19. For each of these values there is a

corresponding value for y, since the complementary factor must be equal to

y −3. Listing the solutions as ordered pairs (x, y), we have the four solutions

(−1, 22), (−3, −16), (17, 4), and (−21, 2).

27. For positive integers a, b, prove that gcd(a, b) = 1 if and only if gcd(a

2

, b

2

) = 1.

Solution: Proposition 1.2.3 (d) states that gcd(a, bc) = 1 if and only if

gcd(a, b) = 1 and gcd(a, c) = 1. Using c = b gives gcd(a, b

2

) = 1 if and

only if gcd(a, b) = 1. Then a similar argument yields gcd(a

2

, b

2

) = 1 if and

only if gcd(a, b

2

) = 1.

28. Prove that n−1 and 2n−1 are relatively prime, for all integers n > 1. Is the

same true for 2n −1 and 3n −1?

Solution: We can write (1)(2n − 1) + (−2)(n − 1) = 1, which proves that

gcd(2n − 1, n − 1) = 1. Similarly, (2)(3n − 1) + (−3)(2n − 1) = 1, and so

gcd(3n −1, 2n −1) = 1.

Comment: Is this really a proof? Yes–producing the necessary linear combi-

nations is enough; you don’t have to explain how you found them.

29. Let m and n be positive integers. Prove that gcd(2

m

− 1, 2

n

− 1) = 1 if and

only if gcd(m, n) = 1.

Comment: We need to do the proof in two parts. First, we will prove that if

gcd(m, n) = 1, then gcd(2

m

−1, 2

n

−1) = 1. Then we will prove the converse,

which states that if gcd(2

m

−1, 2

n

−1) = 1, then gcd(m, n) = 1, To prove the

converse, we will use a proof by contradiction, assuming that gcd(m, n) ,= 1

and showing that this forces gcd(2

m

−1, 2

n

−1) ,= 1.

Before beginning the proof, we recall that the following identity holds for all

values of x: x

k

−1 = (x −1)(x

k−1

+x

k−2

+ +x + 1).

Solution: If gcd(m, n) = 1, then there exist a, b ∈ Z with am + bn = 1.

Substituting x = 2

m

and k = a in the identity given above shows that 2

m

−1

CHAPTER 1 SOLUTIONS 41

is a factor of 2

am

−1, say 2

am

−1 = (2

m

−1)(s), for some s ∈ Z. The same

argument shows that we can write 2

bn

−1 = (2

n

−1)(t), for some t ∈ Z. The

proof now involves what may look like a trick (but it is a useful one). We have

1 = 2

1

−1

= 2

am+bn

−2

bn

+ 2

bn

−1

= 2

bn

(2

am

−1) + 2

bn

−1

= 2

bn

(s)(2

m

−1) + (t)(2

n

−1)

and so we have found a linear combination of 2

m

−1 and 2

n

−1 that equals

1, which proves that gcd(2

m

−1, 2

n

−1) = 1.

If gcd(m, n) ,= 1, say gcd(m, n) = d, then there exist p, q ∈ Z with m = dq

and n = dp. But then an argument similar to the one given for the ﬁrst part

shows that 2

d

− 1 is a common divisor of 2

dq

− 1 and 2

dp

− 1. Therefore

gcd(2

m

−1, 2

n

−1) ,= 1, and this completes the proof.

30. Prove that gcd(2n

2

+ 4n −3, 2n

2

+ 6n −4) = 1, for all integers n > 1.

Solution: We can use the Euclidean algorithm. Long division of polynomials

shows that dividing 2n

2

+ 6n −4 by 2n

2

+ 4n −3 gives a quotient of 1 and a

remainder of 2n −1. The next step is to divide 2n

2

+ 4n −3 by 2n −1, and

this gives a quotient of n + 2 and a remainder of n −1. We have shown that

gcd(2n

2

+6n−4, 2n

2

+4n−3) = gcd(2n

2

+4n−3, 2n−1) = gcd(2n−1, n−1)

and so we can use Problem 28 to conclude that 2n

2

+4n−3 and 2n

2

+6n−4

are relatively prime since 2n −1 and n −1 are relatively prime.

(Of course, you could also continue with the Euclidean algorithm, getting

gcd(2n −1, n −1) = gcd(n −2, 1) = 1.)

1.3 SOLUTIONS

26. Solve the congruence 42x ≡ 12 (mod 90).

Solution: We have gcd(42, 90) = 6, so there is a solution since 6 is a factor

of 12. Solving the congruence 42x ≡ 12 (mod 90) is equivalent solving the

equation 42x = 12 + 90q for integers x and q. This reduces to 7x = 2 + 15q,

or 7x ≡ 2 (mod 15). Equivalently, we obtain 7x ≡ 2 (mod 15) by dividing

42x ≡ 12 (mod 90) through by 6. We next use trial and error to look for the

multiplicative inverse of 7 modulo 15. The numbers congruent to 1 modulo

15 are 16, 31, 46, 61, etc., and −14, −29, −34, etc. Among these, we see

that 7 is a factor of −14, so we multiply both sides of the congruence by −2

since (−2)(7) = −14 ≡ 1 (mod 15). Thus we have −14x ≡ −4 (mod 15), or

x ≡ 11 (mod 15). The solution is x ≡ 11, 26, 41, 56, 71, 86 (mod 90).

42 CHAPTER 1 SOLUTIONS

27. (a) Find all solutions to the congruence 55x ≡ 35 (mod 75).

Solution: We have gcd(55, 75) = 5, which is a divisor of 35. Thus we have

55x ≡ 35 (mod 75); 11x ≡ 7 (mod 15); 44x ≡ 28 (mod 15);

−x ≡ 13 (mod 15); x ≡ 2 (mod 15). The solution is

x ≡ 2, 17, 32, 47, 62 (mod 75).

(b) Find all solutions to the congruence 55x ≡ 36 (mod 75).

Solution: There is no solution, since gcd(55, 75) = 5 is not a divisor of 36.

28. (a) Find one particular integer solution to the equation 110x + 75y = 45.

Solution: Any linear combination of 110 and 75 is a multiple of the gcd.

_

1 0 110

0 1 75

_

;

_

1 −1 35

0 1 75

_

;

_

1 −1 35

−2 3 5

_

;

_

15 −22 0

−2 3 5

_

Thus −2(110) + 3(75) = 5, and multiplying by 9 yields a solution x = −18,

y = 27.

Comment: The matrix computation shows that 110(15) + 75(−22) = 0, so

adding any multiple of the vector (15, −22) to the particular solution (−18, 27)

will also determine a solution.

Second solution: The equation reduces to the congruence 35x ≡ 45 (mod 75).

This reduces to 7x ≡ 9 (mod 15), and multiplying both sides by −2 gives

x ≡ −3 (mod 15). Thus 75y = 45 + 3(110) = 375 and so x = −3, y = 5 is a

solution.

(b) Show that if x = m and y = n is an integer solution to the equation in

part (a), then so is x = m+ 15q and y = n −22q, for any integer q.

Solution: If 110m + 75n = 45, then 110(m + 15q) + 75(n − 22q) = 45 +

110(15)q + 75(−22)q = 45, since 110(15) −75(22) = 0.

29. Solve the system of congruences x ≡ 2 (mod 9) x ≡ 4 (mod 10) .

Solution: Convert the second congruence to the equation x = 4 + 10q for

some q ∈ Z. Then 4 + 10q ≡ 2 (mod 9), which reduces to q ≡ 7 (mod 9).

Thus the solution is x ≡ 74 (mod 90).

30. Solve the system of congruences 5x ≡ 14 (mod 17) 3x ≡ 2 (mod 13) .

Solution: By trial and error, 7 5 ≡ 1 (mod 17) and 9 3 ≡ 1 (mod 13),

so 5x ≡ 14 (mod 17); 35x ≡ 98 (mod 17); x ≡ 13 (mod 17)

and 3x ≡ 2 (mod 13); 27x ≡ 18 (mod 13); x ≡ 5 (mod 13).

Having reduced the system to the standard form, we can solve it in the usual

way. We have x = 13 +17q for some q ∈ Z, and then 13 +17q ≡ 5 (mod 13).

This reduces to 4q ≡ 5 (mod 13), so 40q ≡ 50 (mod 13), or q ≡ 11 (mod 13).

This leads to the answer, x ≡ 13 + 17 11 ≡ 200 (mod 221).

CHAPTER 1 SOLUTIONS 43

31. Solve the system of congruences x ≡ 5 (mod 25) x ≡ 23 (mod 32) .

Solution: Write x = 23+32q for some q ∈ Z, and substitute to get 23+32q ≡

5 (mod 25), which reduces to 7q ≡ 7 (mod 25), so q ≡ 1 (mod 15). This

gives x ≡ 55 (mod 25 32).

32. Give integers a, b, m, n to provide an example of a system

x ≡ a (mod m) x ≡ b (mod n)

that has no solution.

Solution: In the example the integers m and n cannot be relatively prime.

This is the clue to take m = n = 2, with a = 1 and b = 0.

33. (a) Compute the last digit in the decimal expansion of 4

100

.

Solution: The last digit is the remainder when divided by 10. Thus we must

compute the congruence class of 4

100

(mod 10). We have 4

2

≡ 6 (mod 10),

and then 6

2

≡ 6 (mod 10). Thus 4

100

= (4

2

)

50

≡ 6

50

≡ 6 (mod 10).

(b) Is 4

100

divisible by 3?

Solution: No, since 4

100

≡ 1

100

≡ 1 (mod 3). Or you can write 2

200

as the

prime factorization, and then (3, 2

200

) = 1.

34. Find all integers n for which 13 [ 4(n

2

+ 1).

Solution: This is equivalent solving the congruence 4(n

2

+ 1) ≡ 0 (mod 13).

Since gcd(4, 13) = 1, we can cancel 4, to get n

2

≡ −1 (mod 13). Just

computing the squares modulo 13 gives us (±1)

2

= 1, (±2)

2

= 4, (±3)

2

= 9,

(±4)

2

≡ 3 (mod 13), (±5)

2

≡ −1 (mod 13), and (±6)

2

≡ −3 (mod 13). We

have done the computation for representatives of each congruence class, so

the answer to the original question is x ≡ ±5 (mod 13).

35. Prove that 10

n+1

+ 4 10

n

+ 4 is divisible by 9, for all positive integers n.

Solution: This could be proved by induction, but a more elegant proof can

be given by simply observing that 10

n+1

+ 4 10

n

+ 4 ≡ 0 (mod 9) since

10 ≡ 1 (mod 9).

36. Prove that the fourth power of an integer can only have 0, 1, 5, or 6 as its

units digit.

Solution: Since the question deals with the units digit of n

4

, it is really asking

to ﬁnd n

4

(mod 10). All we need to do is to compute the fourth power of each

congruence class modulo 10: 0

4

= 0, (±1)

4

= 1, (±2)

4

= 16 ≡ 6 (mod 10),

(±3)

4

= 81 ≡ 1 (mod 10), (±4)

4

≡ 6

2

≡ 6 (mod 10), and 5

4

≡ 5

2

≡

5 (mod 10). This shows that the only possible units digits for n

4

are 0, 1, 5,

and 6.

44 CHAPTER 1 SOLUTIONS

1.4 SOLUTIONS

30. Find the multiplicative inverse of each nonzero element of Z

7

.

Solution: Since 6 ≡ −1 (mod 7), the class [6]

7

is its own inverse. Further-

more, 2 4 = 8 ≡ 1 (mod 7), and 3 5 = 15 ≡ 1 (mod 7), so [2]

7

and [4]

7

are

inverses of each other, and [3]

7

and [5]

7

are inverses of each other.

31. Find the multiplicative inverse of each nonzero element of Z

13

.

Comment: If ab ≡ 1 (mod n), then [a]

n

and [b]

n

are inverses, as are [−a]

n

and [−b]

n

. If ab ≡ −1 (mod n), then [a]

n

and [−b]

n

are inverses, as are [−a]

n

and [b]

n

. It is useful to list the integers with m with m ≡ ±1 (mod n), and

look at the various ways to factor them.

Solution: Note that 14, 27, and 40 are congruent to 1, while 12, 25, and 39

are congruent to −1. Using 14, we see that [2]

13

and [7]

13

are inverses. Using

12, and we see that [3]

13

and [−4]

13

are inverses, as are the pairs [4]

13

and

[−3]

13

, and [6]

13

and [−2]

13

. Using 40, we see that [5]

13

and [8]

13

are inverses.

Finally, here is the list of inverses: [2]

−1

13

= [7]

13

; [3]

−1

13

= [9]

13

; [4]

−1

13

= [10]

13

;

[5]

−1

13

= [8]

13

; [6]

−1

13

= [11]

13

; Since [12]

−1

13

= [−1]

−1

13

= [−1]

13

= [12]

13

, this

takes care of all of the nonzero elements of Z

13

.

32. Find [91]

−1

501

, if possible (in Z

×

501

).

Solution: We need to use the Euclidean algorithm.

_

1 0 501

0 1 91

_

;

_

1 −5 46

0 1 91

_

;

_

1 −5 46

−1 6 45

_

;

_

2 −11 1

−1 6 45

_

Thus [91]

−1

501

= [−11]

501

= [490]

501

.

33. Find [3379]

−1

4061

, if possible (in Z

×

4061

).

Solution: The inverse does not exist.

_

1 0 4061

0 1 3379

_

;

_

1 −1 682

0 1 3379

_

;

_

1 −1 682

−4 5 651

_

;

_

5 −6 31

−4 5 651

_

At the next step, 31 [ 651, and so (4061, 3379) = 31.

34. In Z

20

: ﬁnd all units (list the multiplicative inverse of each); ﬁnd all idempo-

tent elements; ﬁnd all nilpotent elements.

Comment: We know that Z

n

has ϕ(n) units. They occur in pairs, since

gcd(a, n) = 1 if and only if gcd(n −a, n) = 1. This helps to check your list.

Solution: The units of Z

20

are the equivalence classes represented by 1, 3, 7,

9, 11, 13, 17, and 19. We have [3]

−1

20

= [7]

20

, [9]

−1

20

= [9]

20

, [11]

−1

20

= [11]

20

,

[13]

−1

20

= [17]

20

, and [19]

−1

20

= [19]

20

.

CHAPTER 1 SOLUTIONS 45

The idempotent elements of Z

20

can be found by using trial and error. They

are [0]

20

, [1]

20

, [5]

20

, and [16]

20

. If you want a more systematic approach, you

can use a the hint in Exercise 1.4.13 of the text: if n = bc, with gcd(b, c) = 1,

then any solution to the congruences x ≡ 1 (mod b) and x ≡ 0 (mod c) will

be idempotent modulo n.

The nilpotent elements of Z

20

can be found by using trial and error, or by

using Problem 1.4.40. They are [0]

20

and [10]

20

.

35. In Z

24

: ﬁnd all units (list the multiplicative inverse of each); ﬁnd all idem-

potent elements; ﬁnd all nilpotent elements.

Solution: The units of Z

24

are the equivalence classes represented by 1, 5, 7,

11, 13, 17, 19, and 23. For each of these numbers we have x

2

≡ 1 (mod 24),

and so each element is its own inverse.

The idempotent elements are [0]

24

, [1]

24

, [9]

24

, [16]

24

, and the nilpotent ele-

ments are [0]

24

, [6]

24

, [12]

24

, [18]

24

.

36. Show that Z

×

17

is cyclic.

Comment: To show that Z

×

17

is cyclic, we need to ﬁnd an element whose

multiplicative order is 16. The solution just uses trial and error. It is known

than if p is prime, then Z

×

p

is cyclic, but there is no known algorithm for

actually ﬁnding the one element whose powers cover all of Z

×

p

.

Solution: We begin by trying [2]. We have [2]

2

= [4], [2]

3

= [8], and [2]

4

=

[16] = [−1]. Problem 39 shows that the multiplicative order of an element

has to be a divisor of 16, so the next possibility to check is 8. Since [2]

8

=

[−1]

2

= [1], it follows that [2] has multiplicative order 8.

We next try [3]. We have [3]

2

= [9], [3]

4

= [81] = [−4], and [3]

8

= [16] =

[−1]. The only divisor of 16 that is left is 16 itself, so [3] does in fact have

multiplicative order 16, and we are done.

37. Show that Z

×

35

is not cyclic but that each element has the form [8]

i

35

[−4]

j

35

,

for some positive integers i, j.

Solution: We ﬁrst compute the powers of [8]: [8]

2

= [−6], [8]

3

= [8][−6] =

[−13], and [8]

4

= [−6]

2

= [1], so the multiplicative order of [8] is 4, and the

powers we have listed represent the only possible values of [8]

i

.

We next compute the powers of [−4]: [−4]

2

= [16], [−4]

3

= [−4][16] = [6],

[−4]

4

= [−4][6] = [11], [−4]

5

= [−4][11] = [−9], and [−4]

6

= [−4][−9] = [1],

so the multiplicative order of [−4] is 6.

There are 24 possible products of the form [8]

i

[−4]

j

, for 0 ≤ i < 4 and

0 ≤ j < 6. Are these all diﬀerent? Suppose that [8]

i

[−4]

j

= [8]

m

[−4]

n

, for

some 0 ≤ i < 4 and 0 ≤ j < 6 and 0 ≤ m < 4 and 0 ≤ n < 6. Then

[8]

i−m

= [−4]

n−j

, and since the only power of [8] that is equal to a power of

[−4] is [1] (as shown by our computations), this forces i = m and n = j.

46 CHAPTER 1 SOLUTIONS

We conclude that since there are 24 elements of the form [8]

i

[−4]

j

, every

element in Z

35

must be of this form.

Finally, ([8]

i

[−4]

j

)

12

= ([8]

4

)

3i

([−4]

6

)

2j

= [1], so no element of Z

35

has mul-

tiplicative order 24, showing that Z

35

is not cyclic.

38. Solve the equation [x]

2

11

+ [x]

11

−[6]

11

= [0]

11

.

Solution: We can factor [x]

2

+[x] −[6] = ([x] +[3])([x] −[2]). Corollary 1.4.6

implies that either [x] + [3] = [0] or [x] − [2] = [0], and so the solution is

[x] = [−3] or [x] = [2].

39. Let n be a positive integer, and let a ∈ Z with gcd(a, n) = 1. Prove that if k

is the smallest positive integer for which a

k

≡ 1 (mod n), then k [ ϕ(n).

Solution: Assume that k is the smallest positive integer for which a

k

≡

1 (mod n). We can use the division algorithm to write ϕ(n) = qk +r, where

0 ≤ r < k, and q ∈ Z. Since a

k

≡ 1 (mod n), we know that gcd(a, n) = 1,

and so we can apply Theorem 1.4.11, which shows that a

ϕ(n)

≡ 1 (mod n).

Thus a

r

= a

ϕ(n)−kq

= a

ϕ(n)

(a

k

)

−q

≡ 1 (mod n), so we must have r = 0 since

r < k and k is the smallest positive integer with a

k

≡ 1 (mod n).

40. Prove that [a]

n

is a nilpotent element of Z

n

if and only if each prime divisor

of n is a divisor of a.

Solution: First assume that each prime divisor of n is a divisor of a. If

n = p

α

1

1

p

α

2

2

p

α

t

t

is the prime factorization of n, then we must have a =

p

β

1

1

p

β

2

2

p

β

t

t

d, where 0 ≤ β

j

≤ α

j

for all j. If k is the smallest positive

integer such that kβ

i

≥ α

i

for all i, then n [ a

k

, and so [a]

k

n

= [0]

k

.

Conversely, if [a]

n

is nilpotent, with [a]

k

n

= [0], then n [ a

k

, so each prime

divisor of n is a divisor of a

k

. But if a prime p is a divisor of a

k

, then it must

be a divisor of a, and this completes the proof.

SOLUTIONS TO THE REVIEW PROBLEMS

1. Find gcd(7605, 5733), and express it as a linear combination of 7605 and 5733.

Solution: Use the matrix form of the Euclidean algorithm:

_

1 0 7605

0 1 5733

_

;

_

1 −1 1872

0 1 5733

_

;

_

1 −1 1872

−3 4 117

_

;

_

49 −65 0

−3 4 117

_

. Thus

gcd(7605, 5733) = 117, and 117 = (−3) 7605 + 4 5733.

2. For ω = −

1

2

+

√

3

2

i, prove that ω

n

= 1 if and only if 3[n, for any integer n.

CHAPTER 1 SOLUTIONS 47

Solution: Calculations in the introduction to Chapter 1 show that ω

2

=

−

1

2

−

√

3

2

i, and ω

3

= 1. If n ∈ Z, and 3[n, then n = 3q for some q ∈ Z. Then

ω

n

= ω

3q

= (ω

3

)

q

= 1

q

= 1. Conversely, if n ∈ Z and ω

n

= 1, use the division

algorithm to write n = q 3+r, where the remainder satisﬁes 0 ≤ r < 3. Then

1 = ω

n

= ω

3q+r

= (ω

3

)

q

ω

r

= ω

r

. Since r = 0, 1, 2 and we have shown that

ω ,= 1 and ω

2

,= 1, the only possibility is r = 0, and therefore 3[n.

3. Solve the congruence 24x ≡ 168 (mod 200).

Solution: First we ﬁnd that gcd(24, 200) = 8, and 8 [ 168, so the congruence

has a solution. The next step is to reduce the congruence by dividing each

term by 8, which gives 24x ≡ 168 (mod 200). To solve the congruence

3x ≡ 21 (mod 25) we could ﬁnd the multiplicative inverse of 3 modulo 25.

Trial and error shows it to be −8, we can multiply both sides of the congruence

by −8, and proceed with the solution.

24x ≡ 168 (mod 200)

3x ≡ 21 (mod 25)

−24x ≡ −168 (mod 25)

x ≡ 7 (mod 25)

The solution is x ≡ 7, 32, 57, 82, 107, 132, 157, 182 (mod 200).

4. Solve the system of congruences 2x ≡ 9 (mod 15) x ≡ 8 (mod 11) .

Solution: Write x = 8 +11q for some q ∈ Z, and substitute to get 16 +22q ≡

9 (mod 15), which reduces to 7q ≡ −7 (mod 15), so q ≡ −1 (mod 15). This

gives x ≡ −3 (mod 11 15).

5. List the elements of Z

×

15

. For each element, ﬁnd its multiplicative inverse, and

ﬁnd its multiplicative order.

Solution: There should be 8 elements since ϕ(15) = 8. By Problem 39, the

multiplicative order of any nontrivial element is 2, 4, or 8. The elements are

[1], [2], [4], [7], [8], [11], [13], and [14].

Computing powers, we have [2]

2

= [4], [2]

3

= [8], and [2]

4

= [1]. This shows

not only that the multiplicative order of [2] is 4, but that the multiplicative

order of [4] is 2. The same computation shows that [2]

−1

= [8] and [4]

−1

= [4].

We can also deduce that [13] = [−2] has multiplicative order 4, that [13]

−1

=

[−2]

−1

= [−8] = [7], and that [11]

−1

= [−4]

−1

= [−4] = [11].

Next, we have [7]

2

= [4], so [7] has multiplicative order 4 because [7]

4

= [4]

2

=

[1].

To compute the multiplicative order of [8], we can rewrite it as [2]

3

, and then

it is clear that the ﬁrst positive integer k with ([2]

3

)

k

= [1] is k = 4, since

3k must be a multiple of 4. (This can also be shown by rewriting [8] as

[−7].) Similarly, [11] = [−4] has multiplicative order 2, and [13] = [−2] has

multiplicative order 4.

48 CHAPTER 1 SOLUTIONS

6. Show that if n > 1 is an odd integer, then ϕ(2n) = ϕ(n).

Solution: Since n is odd, the prime 2 does not occur in its prime factorization.

The formula in Proposition 1.4.8 shows that to compute ϕ(2n) in terms of ϕ(n)

we need to add 2 (1 −

1

2

), and this does not change the computation.

Second solution: Since n is odd, the integers n and 2n are relatively prime, and

so it follows from Exercise 1.4.27 of the text that ϕ(2n) = ϕ(2)ϕ(n) = ϕ(n).

Chapter 2

Functions

2.1 SOLUTIONS

20. The “Vertical Line Test” from calculus says that a curve in the xy-plane is

the graph of a function of x if and only if no vertical line intersects the curve

more than once. Explain why this agrees with Deﬁnition 2.1.1.

Solution: We assume that the x-axis is the domain and the y-axis is the

codomain of the function that is to be deﬁned by the given curve. According

to Deﬁnition 2.1.1, a subset of the plane deﬁnes a function if for each element

x in the domain there is a unique element y in the codomain such that (x, y)

belongs to the subset of the plane. If a vertical line intersects the curve in two

distinct points, then there will be points (x

1

, y

1

) and (x

2

, y

2

) on the curve with

x

1

= x

2

and y

1

,= y

2

. Thus if we apply Deﬁnition 2.1.1 to the given curve,

the uniqueness part of the deﬁnition translates directly into the “vertical line

test”.

21. The “Horizontal Line Test” from calculus says that a function is one-to-one

if and only if no horizontal line intersects its graph more than once. Explain

why this agrees with Deﬁnition 2.1.4.

Solution: If a horizontal line intersects the graph of the function more than

once, then the points of intersection represent points (x

1

, y

1

) and (x

2

, y

2

) for

which x

1

,= x

2

but y

1

= y

2

. According to Deﬁnition 2.1.4, a function is

one-to-one if f(x

1

) = f(x

2

) implies x

1

= x

2

. Equivalently, if (x

1

, y

1

) and

(x

2

, y

2

) line on its graph, then we cannot have y

1

= y

2

while x

1

,= x

2

. In this

context, the “horizontal line test” is exactly the same as the condition given

in Deﬁnition 2.1.4.

more than one

49

50 CHAPTER 2 SOLUTIONS

22. In calculus the graph of an inverse function f

−1

is obtained by reﬂecting the

graph of f about the line y = x. Explain why this agrees with Deﬁnition 2.1.7.

Solution: We ﬁrst note that the reﬂection of a point (a, b) in the line y =

x is the point (b, a). This can be seen by observing that the line segment

joining (a, b) and (b, a) has slope −1, which makes it perpendicular to the line

y = x, and that this line segment intersects the line y = x at the midpoint

((a +b)/2, (a +b)/2) of the segment.

If f : R →R has an inverse, and the point (x, y) lies on the graph of f, then

y = f(x), and so f

−1

(y) = f

−1

(f(x)) = x. This shows that the point (x, y)

lies on the graph of f

−1

. Conversely, if (x, y) lies on the graph of f

−1

, then

x = f

−1

(y), and therefore y = f(f

−1

(y)) = f(x), which shows that (y, x) lies

on the graph of f.

On the other hand, suppose that the graph of the function g is deﬁned by

reﬂecting the graph of f in the line y = x. For any real number x, if y =

f(x) then we have g(f(x)) = g(y) = x and for any real number y we have

f(g(y)) = f(x) = y, where x = g(y). This shows that g = f

−1

, and so f has

an inverse.

23. Let A be an n n matrix with entries in R. Deﬁne a linear transformation

L : R

n

→R

n

by L(x) = Ax, for all x ∈ R

n

.

(a) Show that L is an invertible function if and only if det(A) ,= 0.

Solution: I need to assume that you know that a square matrix A is invertible

if and only if det(A) ,= 0.

First, if L has an inverse, then it can also be described by multiplication by a

matrix B, which must satisfy the conditions BA = I, and AB = I, where I is

the nn identity matrix. Thus A is an invertible matrix, and so det(A) ,= 0.

On the other hand, if det(A) ,= 0, then A is invertible, and so L has an inverse,

deﬁned by L

−1

(x) = A

−1

x, for all x ∈ R

n

.

(b) Show that if L is either one-to-one or onto, then it is invertible.

Solution: The rank of the matrix A is the dimension of the column space of

A, and this is the image of the transformation L, so L is onto if and only if A

has rank n.

On the other hand, the nullity of A is the dimension of the solution space of

the equation Ax = 0, and L is one-to-one if and only if the nullity of A is

zero, since Ax

1

= Ax

2

if and only if A(x

1

−x

2

) = 0.

To prove part (b) we need to use the Rank–Nullity Theorem, which states

that if A is an n n matrix, then the rank of A plus the nullity of A is n.

Since the matrix A is invertible if and only if it has rank n, it follows that L is

invertible if and only if L is onto, and then the Rank–Nullity Theorem shows

that this happens if and only if L is one-to-one.

CHAPTER 2 SOLUTIONS 51

24. Let A be an mn matrix with entries in R, and assume that m > n. Deﬁne

a linear transformation L : R

n

→ R

m

by L(x) = Ax, for all x ∈ R

n

. Show

that L is a one-to-one function if det(A

T

A) ,= 0, where A

T

is the transpose

of A.

Solution: If det(A

T

A) ,= 0, then A

T

A is an invertible matrix. If we deﬁne

K : R

m

→ R

n

by K(x) = (A

T

A)

−1

A

T

x, for all x ∈ R

m

, then KL is the

identity function on R

m

. It then follows from Exercise 17 that L is one-to-

one.

Comment: There is a stronger result that depends on knowing a little more

linear algebra. In some linear algebra courses it is proved that det(A

T

A)

gives the n-dimensional “content” of the parallepiped deﬁned by the column

vectors of A. This content is nonzero if and only if the vectors are linearly

independent, and so det(A

T

A) ,= 0 if and only if the column vectors of A are

linearly independent. According to the Rank–Nullity Theorem, this happens

if and only if the nullity of A is zero. In other words, L is a one-to-one linear

transformation if and only if det(A

T

A) ,= 0.

25. Let A be an n n matrix with entries in R. Deﬁne a linear transformation

L : R

n

→ R

n

by L(x) = Ax, for all x ∈ R

n

. Prove that L is one-to-one if

and only if no eigenvalue of A is zero.

Note: A vector x is called an eigenvector of A if it is nonzero and there exists

a scalar λ such a that Ax = λx.

Solution: As noted in the solution to problem 23, Ax

1

= Ax

2

if and only if

A(x

1

−x

2

) = 0, and so L is one-to-one if and only if Ax ,= 0 for all nonzero

vectors x. This is equivalent to the statement that there is no nonzero vector

x for which Ax = 0 x, which translates into the given statement about

eigenvalues of A.

26. Let a be a ﬁxed element of Z

×

17

. Deﬁne the function θ : Z

×

17

→ Z

×

17

by

θ(x) = ax, for all x ∈ Z

×

17

. Is θ one to one? Is θ onto? If possible, ﬁnd the

inverse function θ

−1

.

Solution: Since a has an inverse in Z

×

17

, we can deﬁne ψ : Z

×

17

→ Z

×

17

by

ψ(x) = a

−1

x, for all x ∈ Z

×

17

. Then ψ(θ(x)) = ψ(ax) = a

−1

(ax) = (a

−1

a)x =

x and θ(ψ(x)) = θ(a

−1

x) = a(a

−1

x) = (aa

−1

)x = x, which shows that

ψ = θ

−1

. This implies that θ is one-to-one and onto.

2.2 SOLUTIONS

14. On the set ¦(a, b)¦ of all ordered pairs of positive integers, deﬁne (x

1

, y

1

) ∼

(x

2

, y

2

) if x

1

y

2

= x

2

y

1

. Show that this deﬁnes an equivalence relation.

52 CHAPTER 2 SOLUTIONS

Solution: We ﬁrst show that the reﬂexive law holds. Given an ordered pair

(a, b), we have ab = ba, and so (a, b) ∼ (a, b).

We next check the symmetric law. Given (a

1

, b

1

) and (a

2

, b

2

) with (a

1

, b

1

) ∼

(a

2

, b

2

), we have a

1

b

2

= a

2

b

1

, and so a

2

b

1

= a

1

b

2

, which shows that (a

2

, b

2

) ∼

(a

1

, b

1

).

Finally, we verify the transitive law. Given (a

1

, b

1

), (a

2

, b

2

), and (a

3

, b

3

) with

(a

1

, b

1

) ∼ (a

2

, b

2

) and (a

2

, b

2

) ∼ (a

3

, b

3

), we have the equations a

1

b

2

= a

2

b

1

and a

2

b

3

= a

3

b

2

. If we multiply the ﬁrst equation by b

3

and the second

equation by b

1

, we get a

1

b

2

b

3

= a

2

b

1

b

3

= a

3

b

1

b

2

. Since b

2

,= 0 we can cancel

to obtain a

1

b

3

= a

3

b

1

, showing that (a

1

, b

1

) ∼ (a

3

, b

3

).

15. On the set C of complex numbers, deﬁne z

1

∼ z

2

if [[z

1

[[ = [[z

2

[[. Show that

∼ is an equivalence relation.

Solution: The reﬂexive, symmetric, and transitive laws can be easily veriﬁed

since ∼ is deﬁned in terms of an equality, and equality is itself an equivalence

relation.

16. Let u be a ﬁxed vector in R

3

, and assume that u has length 1. For vectors v

and w, deﬁne v ∼ w if v u = w u, where denotes the standard dot product.

Show that ∼ is an equivalence relation, and give a geometric description of

the equivalence classes of ∼.

Solution: The reﬂexive, symmetric, and transitive laws for the relation ∼

really depend on an equality, and can easily be veriﬁed. Since u has length 1,

v u represents the length of the projection of v onto the line determined by

u. Thus two vectors are equivalent if and only if they lie in the same plane

perpendicular to u. It follows that the equivalence classes of ∼ are the planes

in R

3

that are perpendicular to u.

17. For the function f : R →R deﬁned by f(x) = x

2

, for all x ∈ R, describe the

equivalence relation on R that is determined by f.

Solution: The equivalence relation determined by f is deﬁned by setting a ∼ b

if f(a) = f(b), so a ∼ b if and only if a

2

= b

2

, or, a ∼ b if and only if [a[ = [b[.

18. For the linear transformation L : R

3

→R

3

deﬁned by

L(x, y, z) = (x +y +z, x +y +z, x +y +z) ,

for all (x, y, z) ∈ R

3

, give a geometric description of the partition of R

3

that

is determined by L.

Solution: Since (a

1

, a

2

, a

3

) ∼ (b

1

, b

2

, b

3

) if L(a

1

, a

2

, a

3

) = L(b

1

, b

2

, b

3

), it

follows from the deﬁnition of L that (a

1

, a

2

, a

3

) ∼ (b

1

, b

2

, b

3

) if and only if

a

1

+a

2

+a

3

= b

1

+b

2

+b

3

. For example, ¦(x, y, z) [ L(x, y, z) = (0, 0, 0)¦ is the

plane through the origin whose equation is x +y +z = 0, with normal vector

(1, 1, 1). The other subsets in the partition of R

3

deﬁned by L are planes

CHAPTER 2 SOLUTIONS 53

parallel to this one. Thus the partition consists of the planes perpendicular

to the vector (1, 1, 1).

19. Deﬁne the formula f : Z

12

→ Z

12

by f([x]

12

) = [x]

2

12

, for all [x]

12

∈ Z

12

.

Show that the formula f deﬁnes a function. Find the image of f and the set

Z

12

/f of equivalence classes determined by f.

Solution: The formula for f is well-deﬁned since if [x

1

]

12

= [x

2

]

12

, then

x

1

≡ x

2

(mod 12), and so x

2

1

≡ x

2

2

(mod 12), which shows that f([x

1

]

12

) =

f([x

2

]

12

).

To compute the images of f we have [0]

2

12

= [0]

12

, [±1]

2

12

= [1]

12

, [±2]

2

12

=

[4]

12

, [±3]

2

12

= [9]

12

, [±4]

2

12

= [4]

12

, [±5]

2

12

= [1]

12

, and [6]

2

12

= [0]

12

. Thus

f(Z

12

) = ¦[0]

12

, [1]

12

, [4]

12

, [9]

12

¦. The corresponding equivalence classes de-

termined by f are ¦[0]

12

, [6]

12

¦, ¦[±1]

12

, [±5]

12

¦, ¦[±2]

12

, [±4]

12

¦, ¦[±3]

12

¦.

20. On the set of all nn matrices over R, deﬁne A ∼ B if there exists an invert-

ible matrix P such that PAP

−1

= B. Check that ∼ deﬁnes an equivalence

relation.

Solution: We have A ∼ A since IAI

−1

= A, where I is the n n identity

matrix. If A ∼ B, then PAP

−1

= B for some invertible matrix P, and

so we get A = P

−1

B(P

−1

)

−1

. If A ∼ B and B ∼ C, then PAP

−1

=

B and QBQ

−1

= C for some P, Q. Substituting gives Q(PAP

−1

)Q

−1

=

(QP)A(QP)

−1

= C, and so A ∼ C.

2.3 SOLUTIONS

13. For the permutation σ =

_

1 2 3 4 5 6 7 8 9

7 5 6 9 2 4 8 1 3

_

, write σ as a

product of disjoint cycles. What is the order of σ? Is σ an even permutation?

Compute σ

−1

.

Solution: We have σ = (1, 7, 8)(2, 5)(3, 6, 4, 9), and so its order is 12 since

lcm[3, 2, 4] = 12. It is an even permutation, since it can be expressed as the

product of 6 transpositions. We have σ

−1

= (1, 8, 7)(2, 5)(3, 9, 4, 6).

14. For the permutations σ =

_

1 2 3 4 5 6 7 8 9

2 5 1 8 3 6 4 7 9

_

and

τ =

_

1 2 3 4 5 6 7 8 9

1 5 4 7 2 6 8 9 3

_

, write each of these permutations as a

product of disjoint cycles: σ, τ, στ, στσ

−1

, σ

−1

, τ

−1

, τσ, τστ

−1

.

Solution: σ = (1, 2, 5, 3)(4, 8, 7); τ = (2, 5)(3, 4, 7, 8, 9); στ = (1, 2, 3, 8, 9);

στσ

−1

= (1, 8, 4, 7, 9)(3, 5); σ

−1

= (1, 3, 5, 2)(4, 7, 8); τ

−1

= (2, 5)(3, 9, 8, 7, 4);

τσ = (1, 5, 4, 9, 3); τστ

−1

= (1, 5, 2, 4)(7, 9, 8).

54 CHAPTER 2 SOLUTIONS

15. Let σ = (2, 4, 9, 7, )(6, 4, 2, 5, 9)(1, 6)(3, 8, 6) ∈ S

9

. Write σ as a product of

disjoint cycles. What is the order of σ? Compute σ

−1

.

Solution: We have σ = (1, 9, 6, 3, 8)(2, 5, 7), so it has order 15 = lcm[5, 3],

and σ

−1

= (1, 8, 3, 6, 9)(2, 7, 5).

16. Compute the order of τ =

_

1 2 3 4 5 6 7 8 9 10 11

7 2 11 4 6 8 9 10 1 3 5

_

. For

σ = (3, 8, 7), compute the order of στσ

−1

.

Solution: Since τ = (1, 7, 9)(3, 11, 5, 6, 8, 10), it has order 6. We have στσ

−1

=

(3, 8, 7)(1, 7, 9)(3, 11, 5, 6, 8, 10)(3, 7, 8) = (1, 3, 9)(8, 11, 5, 6, 7, 10), so the cycle

structure of στσ

−1

is the same as that of τ, and thus στσ

−1

has order 6.

17. Prove that if τ ∈ on

is a permutation with order m, then στσ

−1

has order m,

for any permutation σ ∈ on

.

Solution: Assume that τ ∈ on

has order m. It follows from the identity

(στσ

−1

)

k

= στ

k

σ

−1

that (στσ

−1

)

m

= στ

m

σ

−1

= σ(1)σ

−1

= (1). On the

other hand, the order of στσ

−1

cannot be less than n, since (στσ

−1

)

k

= (1)

implies στ

k

σ

−1

= (1), and then τ

k

= σ

−1

σ = (1).

18. Show that S

10

has elements of order 10, 12, and 14, but not 11 or 13.

Solution: The permutation (1, 2)(3, 4, 5, 6, 7) has order 10, while the element

(1, 2, 3)(4, 5, 6, 7) has order 12, and (1, 2)(3, 4, 5, 6, 7, 8, 9) has order 14. On the

other hand, since 11 and 13 are prime, any element of order 11 or 13 would

have to be a cycle, and there are no cycles of that length in S

10

.

19. Let S be a set, and let X be a subset of S. Let G = ¦σ ∈ Sym(S) [ σ(X) ⊂ X¦.

Prove that G is a group of permutations.

20. Let G be a group of permutations, with G ⊆ Sym(S), for the set S. Let τ be

a ﬁxed permutation in Sym(S). Prove that

τGτ

−1

= ¦σ ∈ Sym(S) [ σ = τγτ for some γ ∈ G¦

is a group of permutations.

CHAPTER 2 SOLUTIONS 55

SOLUTIONS TO THE REVIEW PROBLEMS

1. For the function f : R →R deﬁned by f(x) = x

2

, for all x ∈ R, describe the

equivalence relation on R that is determined by f.

2. Deﬁne f : R → R by f(x) = x

3

+ 3xz − 5, for all x ∈ R. Show that f is a

one-to-one function.

Hint: Use the derivative of f to show that f is a strictly increasing function.

3. On the set Q of rational numbers, deﬁne x ∼ y if x − y is an integer. Show

that ∼ is an equivalence relation.

4. In S

10

, let α = (1, 3, 5, 7, 9), β = (1, 2, 6), and γ = (1, 2, 5, 3). For σ = αβγ,

write σ as a product of disjoint cycles, and use this to ﬁnd its order and its

inverse. Is σ even or odd?

Solution: We have σ = (1, 6, 3, 2, 7, 9), so σ has order 6, and

σ

−1

= (1, 9, 7, 2, 3, 6). Since σ has length 6, it can be written as a product of

5 transpositions, so it is an odd permutation.

5. Deﬁne the function φ : Z

×

17

→Z

×

17

by φ(x) = x

−1

, for all x ∈ Z

×

17

. Is φ one to

one? Is φ onto? If possible, ﬁnd the inverse function φ

−1

.

Solution: For all x ∈ Z

×

17

we have φ(φ(x)) = φ(x

−1

) = (x

−1

)

−1

= x, so

φ = φ

−1

, which also shows that φ is one-to-one and onto.

6. (a) Let α be a ﬁxed element of S

n

. Show that φ

α

: S

n

→ S

n

deﬁned by

φ

α

(σ) = ασα

−1

, for all σ ∈ S

n

, is a one-to-one and onto function.

Solution: If φ

α

(σ) = φ

α

(τ), for σ, τ ∈ S

n

, then ασα

−1

= ατα

−1

. We can

multiply on the left by α

−1

and on the right by α, to get σ = τ, so φ

α

is

one-to-one. Finally, given τ ∈ S

n

, we have φ

α

(σ) = τ for σ = α

−1

τα, and so

φ

α

is onto.

Another way to show that φ

α

is one-to-one and onto is to show that it has an

inverse function. A short computation shows that (φ

α

)

−1

= φ

α

−1.

(b) In S

3

, let α = (1, 2). Compute φ

α

.

Solution: Since (1, 2) is its own inverse, direct computations show that

φ

α

((1)) = (1), φ

α

((1, 2)) = (1, 2), φ

α

((1, 3)) = (2, 3), φ

α

((2, 3)) = (1, 3),

φ

α

((1, 2, 3)) = (1, 3, 2), and φ

α

((1, 3, 2)) = (1, 2, 3).

56 CHAPTER 2 SOLUTIONS

Chapter 3

Groups

3.1 SOLUTIONS

22. Use the dot product to deﬁne a multiplication on R

3

. Does this make R

3

into

a group?

Solution: The dot product of two vectors is a scalar, not a vector. This means

that the dot product does not even deﬁne a binary operation on the set of

vectors in R

3

.

23. For vectors (x

1

, y

1

, z

1

) and (x

2

, y

2

, z

2

) in R

3

, the cross product is deﬁned by

(x

1

, y

1

, z

1

)(x

2

, y

2

, z

2

) = (y

1

z

2

− z

1

y

2

, z

1

x

2

− x

1

z

2

, x

1

y

2

− y

1

x

2

). Is R

3

a

group under this multiplication?

Solution: The cross product of the zero vector and any other vector is the

zero vector, so the cross product cannot be used to make the set of all vectors

in R

3

into a group.

Even if we were to exclude the zero vector we would still have problems. The

cross product of two nonzero vectors deﬁnes a vector that is perpendicular to

each of the given vectors. This means that the operation could not have an

identity element, again making it impossible to deﬁne a group structure.

24. On the set G = Q

×

of nonzero rational numbers, deﬁne a new multiplication

by a∗b =

ab

2

, for all a, b ∈ G. Show that G is a group under this multiplication.

Solution: If a and b are nonzero rational numbers, then ab is a nonzero

rational number, and so is

ab

2

, showing that the operation is closed on the set

G. The operation is associative since

a ∗ (b ∗ c) = a ∗

_

bc

2

_

=

a

_

bc

2

_

2

=

a(bc)

4

57

58 CHAPTER 3 SOLUTIONS

and

(a ∗ b) ∗ c =

_

ab

2

_

∗ c =

_

ab

2

_

c

2

=

(ab)c

4

.

The number 2 acts as the multiplicative identity, and if a is nonzero, then

4

a

is a nonzero rational number that serves as the multiplicative inverse of a.

25. Write out the multiplication table for Z

×

9

.

Solution: Z

×

9

= ¦[1]

9

, [2]

9

, [4]

9

, [5]

9

, [7]

9

, [8]

9

¦. We will write m for [m]

9

.

1 2 4 5 7 8

1 1 2 4 5 7 8

2 2 4 8 1 5 7

4 4 8 7 2 1 5

5 5 1 2 7 8 4

7 7 5 1 8 4 2

8 8 7 5 4 2 1

Comment: Rewriting the table, with the elements in a slightly diﬀerent order,

gives a diﬀerent picture of the group.

1 2 4 8 7 5

1 1 2 4 8 7 5

2 2 4 8 7 5 1

4 4 8 7 5 1 2

8 8 7 5 1 2 4

7 7 5 1 2 4 8

5 5 1 2 4 8 7

Each element in the group is a power of 2, and the second table shows what

happens when we arrange the elements in order, as successive powers of 2.

26. Write out the multiplication table for Z

×

15

.

Solution: Z

×

15

= ¦[1]

15

, [2]

15

, [4]

15

, [7]

15

, [8]

15

, [11]

15

, [13]

15

, [14]

15

¦. We will

write the elements as ¦1, 2, 4, 7, −7, −4, −2, −1¦.

1 -1 2 -2 4 -4 7 -7

1 1 -1 2 -2 4 -4 7 -7

-1 -1 1 -2 2 -4 4 -7 7

2 2 -2 4 -4 -7 7 -1 1

-2 -2 2 -4 4 7 -7 1 -1

4 4 -4 -7 7 1 -1 -2 2

-4 -4 4 7 -7 -1 1 2 -2

7 7 -7 -1 1 -2 2 4 -4

-7 -7 7 1 -1 2 -2 -4 4

CHAPTER 3 SOLUTIONS 59

Comment: Notice how much easier it makes it to use the representatives

¦±1, ±2, ±4, ±7¦ when listing the congruence classes in the group.

27. Let G be a group, and suppose that a and b are any elements of G. Show that

if (ab)

2

= a

2

b

2

, then ba = ab.

Solution: Assume that a and b are elements of G for which (ab)

2

= a

2

b

2

.

Expanding this equation gives us

(ab)(ab) = a

2

b

2

.

Since G is a group, both a and b have inverses, denoted by a

−1

and b

−1

,

respectively. Multiplication in G is well-deﬁned, so we can multiply both

sides of the equation on the left by a

−1

without destroying the equality.

If we are to be precise about using the associative law, we have to include the

following steps.

a

−1

((ab)(ab)) = a

−1

(a

2

b

2

)

(a

−1

(ab))(ab) = (a

−1

a

2

)b

2

((a

−1

a)b))(ab) = ((a

−1

a)a)b

2

(eb)(ab) = (ea)b

2

b(ab) = ab

2

The next step is to multiply on the right by b

−1

. The associative law for mul-

tiplication essentially says that parentheses don’t matter, so we don’t really

need to include all of the steps we showed before.

b(ab)b

−1

= (ab

2

)b

−1

(ba)(bb

−1

) = (ab)(bb

−1

)

ba = ab

This completes the proof, since we have shown that if (ab)

2

= a

2

b

2

, then

ba = ab.

28. Let G be a group, and suppose that a and b are any elements of G. Show that

(aba

−1

)

n

= ab

n

a

−1

, for any positive integer n.

Solution: To give a careful proof we need to use induction. The statement

for n = 1 is simply that aba

−1

= aba

−1

, which is certainly true. Now assume

that the result holds for n = k. Using this induction hypothesis, we have the

following calculation.

(aba

−1

)

k+1

= (aba

−1

)

k

(aba

−1

)

= (ab

k

a

−1

)(aba

−1

)

= (ab

k

)(a

−1

a)(ba

−1

)

= (ab

k

)(ba

−1

)

= ab

k+1

a

−1

60 CHAPTER 3 SOLUTIONS

Thus the statement holds for n = k +1, so by induction it holds for all values

of n.

29. In Deﬁnition 3.1.3 of the text, replace condition (iii) with the condition that

there exists e ∈ G such that e a = a for all a ∈ G, and replace condition (iv)

with the condition that for each a ∈ G there exists a

∈ G with a

a = e.

Prove that these weaker conditions (given only on the left) still imply that G

is a group.

Solution: Assume that the two replacement conditions hold. Note the ee = e,

and that the associative law holds.

We will ﬁrst show that a e = a, for all a ∈ G. Let a

be an element in G with

a

a = e. Then

a

(a e) = (a

a) e = e e = e = a

a ,

and since there exists an element a

∈ G with a

a

= e, we can cancel a

**from the left of the above equation, to get a e = a. This shows that e is a
**

multiplicative identity for G, and so the original condition (iii) is satisﬁed.

We also have the equation

a

(a a

) = (a

a) a

= e a

= a

= a

e ,

and then (as above) we can cancel a

to get a a

= e, which shows that a

is

indeed the multiplicative inverse of a. Thus the original condition (iv) holds,

and so G is a group under the given operation.

30. The previous exercise shows that in the deﬁnition of a group it is suﬃcient to

require the existence of a left identity element and the existence of left inverses.

Give an example to show that it is not suﬃcient to require the existence of a

left identity element together with the existence of right inverses.

Solution: On the set G of nonzero real numbers, deﬁne the operation a ∗ b =

[a[b, for all a, b ∈ G. Then a ∗ b ,= 0 if a ,= 0 and b ,= 0, so we have deﬁned a

binary operation on G. The operation is associative since a∗(b∗c) = a∗([b[c) =

[a[[b[c = [ab[c and (a ∗ b) ∗ c = ([a[b) ∗ c = [[a[b[c = [ab[c. The number 1 is a

left identity element, since 1 ∗ a = [1[a = a for all a ∈ G. There is no right

identity element, since the two equations 1 ∗ x = 1 and (−1) ∗ x = −1 have no

simultaneous solution in G. Finally, 1/[a[ is a right inverse for any a ∈ G, but

the equation x ∗ a = 1 has no solution for a = −1, so −1 has no left inverse.

In summary, we have shown that G is not a group, even though it has a left

identity element and right inverses.

31. Let F be the set of all fractional linear transformations of the complex plane.

That is, F is the set of all functions f(z) : C →C of the form f(z) =

az +b

cz +d

,

CHAPTER 3 SOLUTIONS 61

where the coeﬃcients a, b, c, d are integers with ad − bc = 1. Show that F

forms a group under composition of functions.

Solution: We ﬁrst need to check that composition of functions deﬁnes a binary

operation on F, so we need to check the closure axiom in Deﬁnition 3.1.3.

Let f

1

(z) =

a

1

z +b

1

c

1

z +d

1

, and f

2

(z) =

a

2

z +b

2

c

2

z +d

2

, with a

1

d

1

− b

1

c

1

= 1 and

a

2

d

2

−b

2

c

2

= 1. Then for any complex number z we have

f

2

◦ f

1

(z) = f

2

(f

1

(z)) =

a

2

f

z

(z) +b

2

c

2

f

z

(z) +d

2

=

a

2

_

a

1

z+b

1

c

1

z+d

1

_

+b

2

c

2

_

a

1

z+b

1

c

1

z+d

1

_

+d

2

=

a

2

(a

1

z +b

1

) +b

2

(c

1

z +d

1

)

c

2

(a

1

z +b

1

) +d

2

(c

1

z +d

1

)

=

(a

2

a

1

+b

2

c

1

)z + (a

2

b

1

+b

2

d

1

)

(c

2

a

1

+d

2

c

1

)z + (c

2

b

1

+d

2

d

1

)

.

You can see that verifying all of the axioms is going to be painful. We need a

better way to look at the entire situation, so let’s look at the following matrix

product.

_

a

2

b

2

c

2

d

2

_ _

a

1

b

1

c

1

d

1

_

=

_

a

2

a

1

+b

2

c

1

a

2

b

1

+b

2

d

1

c

2

a

1

+d

2

c

1

c

2

b

1

+d

2

d

2

_

If we associate with the fractional linear transformations f

2

(z) =

a

2

z +b

2

c

2

z +d

2

and f

1

(z) =

a

1

z +b

1

c

1

z +d

1

the matrices

_

a

2

b

2

c

2

d

2

_

and

_

a

1

b

1

c

1

d

1

_

, respectively,

then we can see that composition of two fractional linear transformations

corresponds to the product of the two associated matrices. Furthermore, the

condition that ad −bc = 1 for a fractional linear transformation corresponds

to the condition that the determinant of the associated matrix is equal to 1.

All of this means that it is fair to use what we already know about matrix

multiplication. The proof that the determinant of a product is the product of

the determinants can be used to show that in the composition f

2

◦ f

1

we will

still have the required condition on the coeﬃcients that we calculated.

Composition of functions is always associative (compare Exercise 3.1.2 in the

text, for matrices), and the identity function will serve as an identity element

for F. We only need to check that it can be written in the correct form, as a

fractional linear transformation, and this can be shown by choosing coeﬃcients

a = 1, b = 0, c = 0, and d = 1. Finally, we can use the formula for the

inverse of a 2 2 matrix with determinant 1 to ﬁnd an inverse function for

62 CHAPTER 3 SOLUTIONS

f(z) =

az +b

cz +d

. This gives f

−1

(z) =

dz −b

−cz +a

, and completes the proof that

F forms a group under composition of functions.

32. Let G = ¦x ∈ R [ x > 1¦ be the set of all real numbers greater than 1. For

x, y ∈ G, deﬁne x ∗ y = xy −x −y + 2.

(a) Show that the operation ∗ is closed on G.

Solution: If a, b ∈ G, then a > 1 and b > 1, so b − 1 > 0, and therefore

a(b −1) > (b −1). It follows immediately that ab −a −b + 2 > 1.

(b) Show that the associative law holds for ∗.

Solution: For a, b, c ∈ G, we have

a ∗ (b ∗ c) = a ∗ (bc −b −c + 2)

= a(bc −b −c + 2) −a −(bc −b −c + 2) + 2

= abc −ab −ac −bc +a +b +c .

On the other hand, we have

(a ∗ b) ∗ c = (ab −a −b + 2) ∗ c

= (ab −a −b + 2)c −(ab −a −b + 2) −c + 2

= abc −ab −ac −bc +a +b +c .

Thus a ∗ (b ∗ c) = (a ∗ b) ∗ c.

(c) Show that 2 is the identity element for the operation ∗.

Solution: Since the operation is commutative, the one computation 2 ∗ y =

2y −2 −y + 2 = y suﬃces to show that 2 is the identity element.

(d) Show that for element a ∈ G there exists an inverse a

−1

∈ G.

Solution: Given any a ∈ G, we need to solve a ∗ y = 2. This gives us

the equation ay − a − y + 2 = 2, which has the solution y = a/(a − 1).

This solution belongs to G since a > a − 1 implies a/(a − 1) > 1. Finally,

a∗(a/a−1) = a

2

/(a−1)−a−a/(a−1)+2 = (a

2

−a

2

+a−a)/(a−1)+2 = 2.

3.2 SOLUTIONS

23. Find all cyclic subgroups of Z

×

24

.

Solution: You can check that x

2

= 1 for all elements of the group. Thus each

nonzero element generates a subgroup of order 2, including just the element

itself and the identity [1]

24

.

CHAPTER 3 SOLUTIONS 63

24. In Z

×

20

, ﬁnd two subgroups of order 4, one that is cyclic and one that is not

cyclic.

Solution: To ﬁnd a cyclic subgroup of order 4, we need to check the orders

of elements in Z

×

20

= ¦±1, ±3, ±7, ±9¦. It is natural to begin with [3], which

turns out to have order 4, and so ¸[3]) is a cyclic subgroup of order 4.

The element [9] = [3]

2

has order 2. It is easy to check that the subset H =

¦±[1], ±[9]¦ is closed. Since H is a ﬁnite, nonempty subset of a known group,

Corollary 3.2.4 implies that it is a subgroup. Finally, H is not cyclic since no

element of H has order 4.

25. (a) Find the cyclic subgroup of S

7

generated by the element (1, 2, 3)(5, 7).

Solution: We have ((1, 2, 3)(5, 7))

2

= (1, 3, 2), ((1, 2, 3)(5, 7))

3

= (5, 7),

((1, 2, 3)(5, 7))

4

= (1, 2, 3), ((1, 2, 3)(5, 7))

5

= (1, 3, 2)(5, 7), ((1, 2, 3)(5, 7))

6

=

(1). These elements, together with (1, 2, 3)(5, 7), form the cyclic subgroup

generated by (1, 2, 3)(5, 7).

(b) Find a subgroup of S

7

that contains 12 elements. You do not have to list

all of the elements if you can explain why there must be 12, and why they

must form a subgroup.

Solution: We only need to ﬁnd an element of order 12, since it will generate a

cyclic subgroup with 12 elements. Since the order of a product of disjoint cy-

cles is the least common multiple of their lengths, the element (1, 2, 3, 4)(5, 6, 7)

has order 12.

26. In G = Z

×

21

, show that

H = ¦[x]

21

[ x ≡ 1 (mod 3)¦ and K = ¦[x]

21

[ x ≡ 1 (mod 7)¦

are subgroups of G.

Solution: The subset H is ﬁnite and nonempty (it certainly contains [1]

21

), so

by Corollary 3.2.4 it is enough to show that H is closed under multiplication.

If [x]

21

and [y]

21

belong to H, then x ≡ 1 (mod 3) and t ≡ 1 (mod 3), so it

follows that xy ≡ 1 (mod 3), and therefore [x]

21

[y]

21

= [xy]

21

belongs to H.

A similar argument shows that K is a subgroup of Z

×

21

.

27. Let G be an abelian group, and let n be a ﬁxed positive integer. Show that

N = ¦g ∈ G [ g = a

n

for some a ∈ G¦ is a subgroup of G.

Solution: First, the subset N is nonempty since the identity element e can

always be written in the form e = e

n

. Next, suppose that g

1

and g

2

belong to

N. Then there must exist elements a

1

and a

2

in G with g

1

= a

n

1

and g

2

= a

n

2

,

and so g

1

g

2

= a

n

1

a

n

2

= (a

1

a

2

)

n

. The last equality holds since G is abelian.

Finally, if g ∈ N, with g = a

n

, then g

−1

= (a

n

)

−1

= (a

−1

)

n

, and so g

−1

has

the right form to belong to N.

64 CHAPTER 3 SOLUTIONS

28. Suppose that p is a prime number of the form p = 2

n

+ 1.

(a) Show that in Z

×

p

the order of [2]

p

is 2n.

Solution: Since 2

n

+1 = p, we have 2

n

≡ −1 (mod p), and squaring this yields

2

2n

≡ 1 (mod p). Thus the order of [2] is a divisor of 2n, and for any proper

divisor k of 2n we have k ≤ n, so 2

k

,≡ 1 (mod p) since 2

k

−1 < 2

n

+ 1 = p.

This shows that [2] has order 2n.

(b) Use part (a) to prove that n must be a power of 2.

Solution: The order of [2] is a divisor of [Z

×

p

[ = p − 1 = 2

n

, so by part (a)

this implies that n is a divisor of 2

n−1

, and therefore n is a power of 2.

29. In the multiplicative group C

×

of complex numbers, ﬁnd the order of the

elements −

√

2

2

+

√

2

2

i and −

√

2

2

−

√

2

2

i.

Solution: It is probably easiest to change these complex numbers from rectan-

gular coordinates into polar coordinates. (See Appendix A.5 for a discussion

of the properties of complex numbers.) Each of the numbers has magnitude

1, and you can check that

−

√

2

2

+

√

2

2

i = cos(3π/4)+i sin(3π/4) and −

√

2

2

−

√

2

2

i = cos(5π/4)+i sin(5π/4).

We can use Demoivre’s Theorem (Theorem A.5.2) to compute powers of com-

plex numbers. It follows from this theorem that (cos(3π/4) +i sin(3π/4))

8

=

cos(6π) + i sin(6π) = 1, and so −

√

2

2

+

√

2

2

i has order 8 in C

×

. A similar

argument shows that −

√

2

2

−

√

2

2

i also has order 8.

30. In the group G = GL

2

(R) of invertible 2 2 matrices with real entries, show

that

H =

__

cos θ −sin θ

sin θ cos θ

_¸

¸

¸

¸

θ ∈ R

_

is a subgroup of G.

Solution: Closure: To show that H is closed under multiplication we need to

use the familiar trig identities for the sine and cosine of the sum of two angles.

_

cos θ −sin θ

sin θ cos θ

_ _

cos φ −sin φ

sin φ cos φ

_

=

_

cos θ cos φ −sin θ sin φ −cos θ sin φ −sin θ cos φ

sin θ cos φ + cos θ sin φ −sin θ sin φ + cos θ cos φ

_

=

_

cos θ cos φ −sin θ sin φ −(sin θ cos φ + cos θ sin φ)

sin θ cos φ + cos θ sin φ cos θ cos φ −sin θ sin φ

_

=

_

cos(θ +φ) −sin(θ +φ)

sin(θ +φ) cos(θ +φ)

_

∈ H.

Identity: To see that the identity matrix is in the set, let θ = 0.

Existence of inverses:

_

cos θ −sin θ

sin θ cos θ

_

−1

=

_

cos(−θ) −sin(−θ)

sin(−θ) cos(−θ)

_

∈ H.

CHAPTER 3 SOLUTIONS 65

31. Let K be the following subset of GL

2

(R).

K =

__

a b

c d

_¸

¸

¸

¸

d = a, c = −2b, ad −bc ,= 0

_

Show that K is a subgroup of GL

2

(R).

Solution: The closure axiom holds since

_

a

1

b

1

−2b

1

a

1

_ _

a

2

b

2

−2b

2

a

2

_

=

_

a

1

a

2

−2b

1

b

2

a

1

b

2

+b

1

a

2

−2(a

1

b

2

−b

1

a

2

) a

1

a

2

−2b

1

b

2

_

. The

identity matrix belongs K, and

_

a b

−2b a

_

−1

=

1

a

2

+ 2b

2

_

a −b

−2(−b) a

_

.

Comment: We don’t need to worry about the condition ad−bc ,= 0, since for

any element in H the determinant is a

2

+ 2b

2

, which is always positive.

32. Compute the centralizer in GL

2

(R) of the matrix

_

2 1

1 1

_

.

Note: Exercise 3.2.14 in the text deﬁnes the centralizer of an element a of the

group G to be C(a) = ¦x ∈ G [ xa = ax¦.

Solution: Let A =

_

2 1

1 1

_

, and suppose that X =

_

a b

c d

_

belongs to

the centralizer of A in GL

2

(R). Then we must have XA = AX, so do-

ing this calculation shows that

_

2a +b a +b

2c +d c +d

_

=

_

a b

c d

_ _

2 1

1 1

_

=

_

2 1

1 1

_ _

a b

c d

_

=

_

2a +c 2b +d

a +c b +d

_

. Equating corresponding entries

shows that we must have 2a+b = 2a+c, a+b = 2b+d, 2c+d = a+c, and c+d =

b+d. The ﬁrst and last equations imply that b = c, while the second and third

equations imply that a = b +d = c +d, or d = a −b. On the other hand, any

matrix of this form commutes with A, so the centralizer in GL

2

(R) of the ma-

trix

_

2 1

1 1

_

is the subgroup

__

a b

b a −b

_¸

¸

¸

¸

a, b ∈ R and ab ,= a

2

+b

2

_

.

3.3 SOLUTIONS

16. Show that Z

5

Z

3

is a cyclic group, and list all of the generators for the

group.

Solution: By Proposition 3.3.4 (b), then order of an element ([a]

5

, [b]

3

) in

Z

5

Z

3

is the least common multiple of the orders of the components. Since

[1]

5

, [2]

5

, [3]

5

, [4]

5

have order 5 in Z

5

and [1]

3

, [2]

3

have order 3 in Z

3

, the

element ([a]

5

, [b]

3

) is a generator if and only if [a]

5

,= [0]

5

and [b]

5

,= [0]

5

.

There are 8 such elements, which can easily be listed.

66 CHAPTER 3 SOLUTIONS

Comment: The other 7 elements in the group will have at least one component

equal to zero. There are 4 elements of order 5 (with [0]

3

as the second com-

ponent) and 2 elements of order 3 (with [0]

5

as the ﬁrst component). Adding

the identity element to the list accounts for all 15 elements of Z

5

Z

3

.

17. Find the order of the element ([9]

12

, [15]

18

) in the group Z

12

Z

18

.

Solution: Since gcd(9, 12) = 3, we have o([9]

12

) = o([3]

12

) = 4. Similarly,

o([15]

18

) = o([3]

18

) = 6. Thus the order of ([9]

12

, [15]

18

) is lcm[4, 6] = 12.

18. Find two groups G

1

and G

2

whose direct product G

1

G

2

has a subgroup

that is not of the form H

1

H

2

, for subgroups H

1

⊆ G

1

and H

2

⊆ G

2

.

Solution: In Z

2

Z

2

, the element (1, 1) has order 2, so it generates a cyclic

subgroup that does not have the required form.

19. In the group G = Z

×

36

, let H = ¦[x] [ x ≡ 1 (mod 4)¦ and K = ¦[y] [ y ≡

1 (mod 9)¦. Show that H and K are subgroups of G, and ﬁnd the subgroup

HK.

Solution: It can be shown (as in Problem 3.2.26) that the given subsets are

subgroups. A short computation shows that H = ¦[1], [5], [13], [17], [25], [29]¦

and K = ¦[1], [19]¦. Since x [1] ,= x [19] for x ∈ G, the set HK must contain

12 elements, and so HK = G.

20. Show that if p is a prime number, then the order of the general linear group

GL

n

(Z

p

) is (p

n

−1)(p

n

−p) (p

n

−p

n−1

).

Solution: We need to count the number of ways in which an invertible matrix

can be constructed. This is done by noting that we need n linearly independent

rows. The ﬁrst row can be any nonzero vector, so there are p

n

−1 choices.

There are p

n

possibilities for the second row, but to be linearly independent

of the ﬁrst row, it cannot be a scalar multiple of that row. Since we have p

possible scalars, we need to omit the p multiples of the ﬁrst row. Therefore

the total number of ways to construct a second row independent of the ﬁrst

is p

n

−p.

For the third row, we need to subtract p

2

, which is the number of vectors in

the subspace spanned by the ﬁrst two rows that we have chosen. Thus there

are p

n

− p

2

possibilities for the third row. This argument can be continued,

giving the stated result. (A more formal proof could be given by induction.)

21. Find the order of the element A =

_

_

i 0 0

0 −1 0

0 0 −i

_

_

in the group GL

3

(C).

Solution: For any diagonal 3 3 matrix we have

_

_

a 0 0

0 b 0

0 0 c

_

_

n

=

_

_

a

n

0 0

0 b

n

0

0 0 c

n

_

_

,

CHAPTER 3 SOLUTIONS 67

It follows immediately that the order of A is the least common multiple of the

orders of the diagonal entries i, −1, and −i. Thus o(A) = 4.

22. Let G be the subgroup of GL

2

(R) deﬁned by

G =

__

m b

0 1

_¸

¸

¸

¸

m ,= 0

_

.

Let A =

_

1 1

0 1

_

and B =

_

−1 0

0 1

_

. Find the centralizers C(A) and

C(B), and show that C(A) ∩ C(B) = Z(G), where Z(G) is the center of G.

Solution: Suppose that X =

_

m b

0 1

_

belongs to C(A) in G. Then we must

have XA = AX, and doing this calculation shows that

_

m m+b

0 1

_

=

_

m b

0 1

__

1 1

0 1

_

=

_

1 1

0 1

__

m b

0 1

_

=

_

m b + 1

0 1

_

.

Equating corresponding entries shows that we must have m + b = b + 1, and

so m = 1. On the other hand, any matrix of this form commutes with A, and

so C(A) =

__

1 b

0 1

_¸

¸

¸

¸

b ∈ R

_

.

Now suppose that X =

_

m b

0 1

_

belongs to C(B). Then XB = BX, and so

_

−m b

0 1

_

=

_

m b

0 1

__

−1 0

0 1

_

=

_

−1 0

0 1

__

m b

0 1

_

=

_

−m −b

0 1

_

.

Equating corresponding entries shows that we must have b = 0, and so C(B) =

__

m 0

0 1

_¸

¸

¸

¸

0 ,= m ∈ R

_

.

This shows that C(A) ∩C(B) is the identity matrix, and since any element in

the center of G must belong to C(A) ∩ C(B), our calculations show that the

center of G is the trivial subgroup, containing only the identity element.

23. Compute the centralizer in GL

2

(Z

3

) of the matrix

_

2 1

0 2

_

.

Solution: Let A =

_

2 1

0 2

_

, and suppose that X =

_

a b

c d

_

belongs to

the centralizer of A in GL

2

(Z

3

). Then XA = AX, and so

_

2a a + 2b

2c c + 2d

_

=

_

a b

c d

_ _

2 1

0 2

_

=

_

2 1

0 2

_ _

a b

c d

_

=

_

2a +c 2b +d

2c 2d

_

. Equating

corresponding entries shows that we must have 2a = 2a +c, a + 2b = 2b +d,

2c = 2c, and c + 2d = 2d. The ﬁrst equation implies that c = 0, while the

68 CHAPTER 3 SOLUTIONS

second equation implies that a = d. It follows that the centralizer in GL

2

(Z

3

)

of the matrix

_

2 1

0 2

_

is the subgroup

__

a b

0 a

_¸

¸

¸

¸

a, b ∈ Z

3

and a ,= 0

_

.

Comment: The centralizer contains 6 elements, while it follows from Prob-

lem 20 in this section that GL

2

(Z

3

) has (3

2

−1)(3

2

−3) = 48 elements.

24. Compute the centralizer in GL

2

(Z

3

) of the matrix

_

2 1

1 1

_

.

Solution: Let A =

_

2 1

1 1

_

, and suppose that X =

_

a b

c d

_

belongs to the

centralizer of A in GL

2

(Z

3

). Then XA = AX, and so

_

2a +b a +b

2c +d c +d

_

=

_

a b

c d

__

2 1

1 1

_

=

_

2 1

1 1

__

a b

c d

_

=

_

2a +c 2b +d

a +c b +d

_

. Equating cor-

responding entries shows that we must have 2a + b = 2a + c, a + b =

2b + d, 2c + d = a + c, and c + d = b + d. The ﬁrst equation implies

that c = b, while the second equation implies that d = a − b. It fol-

lows that the centralizer in GL

2

(Z

3

) of the matrix

_

2 1

1 1

_

is the subgroup

__

a b

b a −b

_¸

¸

¸

¸

a, b ∈ Z

3

and a ,= 0 or b ,= 0

_

.

Comment: In this case the centralizer contains 8 of the 48 elements in

GL

2

(Z

3

).

25. Let H be the following subset of the group G = GL

2

(Z

5

).

H =

__

m b

0 1

_

∈ GL

2

(Z

5

)

¸

¸

¸

¸

m, b ∈ Z

5

, m = ±1

_

(a) Show that H is a subgroup of G with 10 elements.

Solution: Since in the matrix

_

m b

0 1

_

there are two choices for m and 5

choices for b, we will have a total of 10 elements. The set is closed under mul-

tiplication since

_

±1 b

0 1

_ _

±1 c

0 1

_

=

_

±1 b ±c

0 1

_

, and it is certainly

nonempty, and so it is a subgroup since the group is ﬁnite.

(b) Show that if we let A =

_

1 1

0 1

_

and B =

_

−1 0

0 1

_

, then BA = A

−1

B.

Solution: We have BA =

_

−1 0

0 1

_ _

1 1

0 1

_

=

_

−1 −1

0 1

_

and A

−1

B =

_

1 −1

0 1

_ _

−1 0

0 1

_

=

_

−1 −1

0 1

_

.

CHAPTER 3 SOLUTIONS 69

(c) Show that every element of H can be written uniquely in the form A

i

B

j

,

where 0 ≤ i < 5 and 0 ≤ j < 2.

Solution: Since

_

1 b

0 1

_ _

1 c

0 1

_

=

_

1 b +c

0 1

_

, the cyclic subgroup gen-

erated by A consists of all matrices of the form

_

1 b

0 1

_

. Multiplying on the

right by B will create 5 additional elements, giving all of the elements in H.

3.4 SOLUTIONS

21. Show that Z

×

17

is isomorphic to Z

16

.

Solution: The element [3] is a generator for Z

×

17

, since 3

2

= 9, 3

3

= 27 ≡ 10,

3

4

≡ 310 ≡ 30 ≡ 13, 3

5

≡ 313 ≡ 39 ≡ 5, 3

6

≡ 35 ≡ 15, 3

7

≡ 315 ≡ 45 ≡ 11,

and 3

8

≡ 3 11 ≡ 33 ≡ −1 ,≡ 1. Therefore Z

×

17

is a cyclic group with 16

elements. This provides the clue at to how to deﬁne the isomorphism we need,

since Z

16

is also a cyclic group, with generator [1]

16

, and Proposition 3.4.3 (a)

implies that any isomorphism between cyclic groups must map a generator to

a generator.

Deﬁne φ : Z

16

→ Z

×

17

by setting φ([1]

16

) = [3]

17

, φ([2]

16

) = [3]

2

17

, etc. The

general formula is φ([n]

16

) = [3]

n

17

, for all [n]

16

∈ Z

16

. Since φ is deﬁned by

using a representative n of the equivalence class [n]

16

, we have to show that

the formula for φ does not depend on the particular representative that is

chosen. If k ≡ m (mod 16), then it follows from Proposition 3.2.8 (c) that

[3]

k

17

= [3]

m

17

since [3]

17

has order 16 in Z

×

17

Therefore φ([k]

16

) = φ([m]16), and

so φ is a well-deﬁned function.

Proposition 3.2.8 (c) shows that φ([k]

16

) = φ([m]

16

) only if k ≡ m (mod 16),

and so φ is a one-to-one function. Then because both Z

16

and Z

×

17

have 16

elements, it follows from Proposition 2.1.5 that φ is also an onto function.

The proof that φ respects the two group operations follows the proof in Ex-

ample 3.4.1. For any elements [n]

16

and [m]

16

in Z

16

, we ﬁrst compute what

happens if we combine [n]

16

and [m]

16

using the operation in Z

16

, and then

substitute the result into the function φ:

φ([n]

16

+ [m]

16

) = φ([n +m]

16

) = [3]

n+m

17

.

Next, we ﬁrst apply the function φ to the two elements, [n]

16

and [m]

16

, and

then combine the results using the operation in Z

×

17

:

φ([n]

16

) φ([m]

16

) = [3]

n

17

[3]

m

17

= [3]

n+m

17

.

Thus φ([n]

16

+[m]

16

) = φ([n]

16

) φ([m]

16

), and this completes the proof that

φ is a group isomorphism.

70 CHAPTER 3 SOLUTIONS

22. Let φ : R

×

→ R

×

be deﬁned by φ(x) = x

3

, for all x ∈ R. Show that φ is a

group isomorphism.

Solution: The function φ preserves multiplication in R

×

since for all a, b ∈ R

×

we have φ(ab) = (ab)

3

= a

3

b

3

= φ(a)φ(b). The function is one-to-one and

onto since for each y ∈ R

×

the equation φ(x) = y has the unique solution

x =

3

√

y.

23. Let G

1

, G

2

, H

1

, H

2

be groups, and suppose that θ

1

: G

1

→ H

1

and θ

2

:

G

2

→ H

2

are group isomorphisms. Deﬁne φ : G

1

G

2

→ H

1

H

2

by

φ(x

1

, x

2

) = (θ

1

(x

1

), θ

2

(x

2

)), for all (x

1

, x

2

) ∈ G

1

G

2

. Prove that φ is a

group isomorphism.

Solution: If (y

1

, y

2

) ∈ H

1

H

2

, then since θ

1

is an isomorphism there is a

unique element x

1

∈ G

1

with y

1

= θ

1

(x

1

). Similarly, since θ

2

is an isomor-

phism there is a unique element x

2

∈ G

2

with y

2

= θ

2

(x

2

). Thus there is a

unique element (x

1

, x

2

) ∈ G

1

G

2

such that (y

1

, y

2

) = φ(x

1

, x

2

), and so φ is

one-to-one and onto.

Given (a

1

, a

2

) and (b

1

, b

2

) in G

1

G

2

, we have

φ((a

1

, a

2

) (b

1

, b

2

)) = φ((a

1

b

1

, a

2

b

2

)) = (θ

1

(a

1

b

1

), θ

2

(a

2

b

2

))

= (θ

1

(a

1

)θ

1

(b

1

), θ

2

(a

2

)θ

2

(b

2

))

φ((a

1

, a

2

)) φ((b

1

, b

2

)) = (θ

1

(a

1

), θ

2

(a

2

)) (θ

1

(b

1

), θ

2

(b

2

))

= (θ

1

(a

1

)θ

1

(b

1

), θ

2

(a

2

)θ

2

(b

2

))

and so φ : G

1

G

2

→H

1

H

2

is a group isomorphism.

24. Prove that the group Z

×

7

Z

×

11

is isomorphic to the group Z

6

Z

10

.

Solution: You can check that Z

×

7

is cyclic of order 6, generated by [3]

7

, and

that Z

×

11

is cyclic of order 10, generated by [2]

11

. Just as in Problem 21, you

can show that θ

1

: Z

6

→ Z

×

7

deﬁned by θ

1

([n]

6

) = [3]

n

7

and θ

2

: Z

10

→ Z

×

11

deﬁned by θ

2

([m]

10

) = [2]

m

11

are group isomorphisms. It then follows from

Problem 23 that φ : Z

6

Z

10

→ Z

×

7

Z

×

11

deﬁned by φ(([n]

6

, [m]

10

)) =

([3]

n

7

, [2]

m

11

), for all [n]

6

∈ Z

6

and all [m]

10

∈ Z

10

, is a group isomorphism.

25. Deﬁne φ : Z

30

Z

2

→ Z

10

Z

6

by φ([n]

30

, [m]

2

) = ([n]

10

, [4n + 3m]

6

), for

all ([n]

30

, [m]

2

) ∈ Z

30

Z

2

. First prove that φ is a well-deﬁned function, and

then prove that φ is a group isomorphism.

Solution: If ([n]

30

, [m]

2

) and ([k]

30

, [j]

2

) are equal elements of Z

30

Z

2

,

then 30 [ n −k and 2 [ m−j. It follows that 10 [ n −k, and so [n]

10

= [k]

10

.

Furthermore, 30 [ 4(n−k), so 6 [ 4(n−k), and then 6 [ 3(m−j), which together

imply that 6 [ (4n + 3m) − (4k + 3j), showing that [4n + 3m]

6

= [4k + 3j]

6

.

Thus ([n]

10

, [4n + 3m]

6

) = ([k]

10

, [4k + 3j]

6

), which shows that the formula

for φ does yield a well-deﬁned function.

CHAPTER 3 SOLUTIONS 71

For any elements ([a]

30

, [c]

2

) and ([b]

30

, [d]

2

) we have

φ(([a]

30

, [c]

2

) + ([b]

30

, [d]

2

)) = φ(([a +b]

30

, [c +d]

2

))

= ([a +b]

10

, [4(a +b) + 3(c +d)]

2

)

= ([a +b]

10

, [4a + 4b + 3c + 3d]

2

)

φ(([a]

30

, [c]

2

)) +φ(([b]

30

, [d]

2

)) = ([a]

10

, [4a + 3c]

2

) + ([b]

10

, [4b + 3d]

2

)

= ([a +b]

10

, [4a + 3c + 4b + 3d]

2

)

= ([a +b]

10

, [4a + 4b + 3c + 3d]

2

)

and so φ respects the operations in the two groups. This means that we can use

Proposition 3.4.4 to show that φ is one-to-one. If φ([n]

30

, [m]

2

) = ([0]

10

, [0]

6

),

then ([n]

10

, [4n + 3m]

6

) = ([0]

10

, [0]

6

), so 10 [ n, say n = 10q, for some

q ∈ Z, and 6 [ (4n + 3m), or 6 [ (40q + 3m). It follows that 2 [ (40q + 3m)

and 3 [ (40q + 3m), and therefore 2 [ 3m since 2 [ 40q, and 3 [ 40q since

3 [ 3m. Then since 2 and 3 are prime numbers, it follows that 2 [ m, so

[m]

2

= [0]

2

, and 3 [ q, so [n]

30

= [10q]

30

= [0]

30

. We have now shown that

if φ([n]

30

, [m]

2

) = ([0]

10

, [0]

6

), then ([n]

30

, [m]

2

) = ([0]

30

, [0]

2

), and so the

condition in Proposition 3.4.4 is satisﬁed. We conclude that φ is a one-to-

one function. Since the two groups both have 60 elements, it follows that φ

must also be an onto function. We have therefore checked all of the necessary

conditions, so we may conclude that φ is a group isomorphism.

26. Let G be a group, and let H be a subgroup of G. Prove that if a is any

element of G, then the subset

aHa

−1

= ¦g ∈ G [ g = aha

−1

for some h ∈ H¦

is a subgroup of G that is isomorphic to H.

Solution: By Exercise 3.4.13 in the text, the function φ : G → G deﬁned by

φ(x) = axa

−1

, for all x ∈ G, is a group isomorphism. By Exercise 3.4.15 the

image under φ of any subgroup of G is again a subgroup of G, so aHa

−1

=

φ(H) is a subgroup of G. It is then clear that the function θ : H → aHa

−1

deﬁned by θ(x) = axa

−1

is an isomorphism.

27. Let G, G

1

, G

2

be groups. Prove that if G is isomorphic to G

1

G

2

, then there

are subgroups H and K in G such that H ∩K = ¦e¦, HK = G, and hk = kh

for all h ∈ H and k ∈ K.

Solution: Let φ : G

1

G

2

→ G be an isomorphism. Exercise 3.3.9 in the

text shows that in G

1

G

2

the subgroups H

∗

= ¦(x

1

, x

2

) [ x

2

= e¦ and K

∗

=

¦(x

1

, x

2

) [ x

1

= e¦ have the properties we are looking for. Let H = φ(H

∗

)

and K = φ(K

∗

) be the images in G of H

∗

and K

∗

, respectively. We know (by

Exercise 3.4.15) that H and K are subgroups of G, so we only need to show

that H ∩ K = ¦e¦, HK = G, and hk = kh for all h ∈ H and k ∈ K.

72 CHAPTER 3 SOLUTIONS

Let y ∈ G, with y = φ(x), for x ∈ G

1

G

2

. If y ∈ H ∩K, then y ∈ H, and so

x ∈ H

∗

. Since y ∈ K as well, we must also have x ∈ K

∗

, so x ∈ H

∗

∩K

∗

, and

therefore x = (e

1

, e

2

), where e

1

and e

2

are the respective identity elements in

G

1

and G

2

. Thus y = φ((e

1

, e

2

)) = e, showing that H ∩ K = ¦e¦. Since y

is any element of G, and we can write x = h

∗

k

∗

for some h

∗

∈ H

∗

and some

k

∗

∈ K

∗

, it follows that y = φ(h

∗

k

∗

) = φ(h

∗

)φ(k

∗

), and thus G = HK. It

is clear that φ preserves the fact that elements of h

∗

and K

∗

commute. We

conclude that H and K satisfy the desired conditions.

28. Show that for any prime number p, the subgroup of diagonal matrices in

GL

2

(Z

p

) is isomorphic to Z

×

p

Z

×

p

.

Solution: Since each matrix in GL

2

(Z

p

) has nonzero determinant, it is clear

that the mapping φ : Z

×

p

Z

×

p

→GL

2

(Z

p

) deﬁned by φ(x

1

, x

2

) =

_

x

1

0

0 x

2

_

,

for each (x

1

, x

2

) ∈ Z

×

p

Z

×

p

, is one-to-one and maps Z

×

p

Z

×

p

onto the subgroup

of diagonal matrices. This mapping respects the operations in the two groups,

since for (a

1

, a

2

), (b

1

, b

2

) ∈ Z

×

p

Z

×

p

we have

φ((a

1

, a

2

)(b

1

, b

2

)) = φ((a

1

b

1

, a

2

b

2

))

=

_

a

1

b

1

0

0 a

2

b

2

_

=

_

a

1

0

0 b

1

_ _

a

2

0

0 b

2

_

= φ((a

1

, a

2

))φ((b

1

, b

2

)) .

Thus φ is the desired isomorphism.

29. (a) In the group G = GL

2

(R) of invertible 2 2 matrices with real entries,

show that

H =

__

a

11

a

12

a

21

a

22

_

∈ GL

2

(R)

¸

¸

¸

¸

a

11

= 1, a

21

= 0, a

22

= 1

_

is a subgroup of G.

Solution: Closure:

_

1 a

0 1

_ _

1 b

0 1

_

=

_

1 a +b

0 1

_

.

Identity: The identity matrix has the correct form.

Existence of inverses:

_

1 a

0 1

_

−1

=

_

1 −a

0 1

_

∈ H.

(b) Show that H is isomorphic to the group R of all real numbers, under

addition.

Solution: Deﬁne φ : R → H by φ(x) =

_

1 x

0 1

_

, for all x ∈ R. You can

easily check that φ is an isomorphism. (The computation necessary to show

that φ preserves the respective operations is the same computation we used

to show that H is closed.)

CHAPTER 3 SOLUTIONS 73

30. Let G be the subgroup of GL

2

(R) deﬁned by

G =

__

m b

0 1

_¸

¸

¸

¸

m ,= 0

_

.

Show that G is not isomorphic to the direct product R

×

R.

Solution: Our approach is to try to ﬁnd an algebraic property that would be

preserved by any isomorphism but which is satisﬁed by only one of the two

groups in question. By Proposition 3.4.3 (b), if one of the groups is abelian

but the other is not, then the groups cannot be isomorphic.

The direct product R

×

Ris an abelian group, since each factor is abelian. On

the other hand, G is not abelian, since

_

−1 0

0 1

_ _

1 1

0 1

_

=

_

−1 −1

0 1

_

but

_

1 1

0 1

_ _

−1 0

0 1

_

=

_

−1 1

0 1

_

. Thus the two groups cannot be iso-

morphic.

31. Let H be the following subgroup of group G = GL

2

(Z

3

).

H =

__

m b

0 1

_

∈ GL

2

(Z

3

)

¸

¸

¸

¸

m, b ∈ Z

3

, m ,= 0

_

Show that H is isomorphic to the symmetric group o3

.

Solution: This group is small enough that we can just compare its multipli-

cation table to that of o3

, as given in Table 3.3.3 (on page 104 of the text).

Remember that constructing an isomorphism is the same as constructing a

one-to-one correspondence between the elements of the group, such that all

entries in the respective group tables also have the same one-to-one correspon-

dence.

In this case we can explain how this can be done, without actually writing out

the multiplication table. Let A =

_

1 1

0 1

_

and B =

_

−1 0

0 1

_

. Then just

as in Problem 3.3.25, we can show that BA = A

−1

B, and that each element

of H has the form can be written uniquely in the form A

i

B

j

, where 0 ≤ i < 3

and 0 ≤ j < 2. This information should make it plausible that the function

φ : o3

→H deﬁned by φ(a

i

b

j

) = A

i

B

j

, for all 0 ≤ i < 3 and 0 ≤ j < 2, gives

a one-to-one correspondence between the elements of the groups which also

produces multiplication tables that look exactly the same.

32. Let G be a group, and let S be any set for which there exists a one-to-

one and onto function φ : G → S. Deﬁne an operation on S by setting

x

1

x

2

= φ(φ

−1

(x

1

)φ

−1

(x

2

)), for all x

1

, x

2

∈ S. Prove that S is a group under

this operation, and that φ is actually a group isomorphism.

Solution: (Outline only) The operation is well-deﬁned on S, since φ and φ

−1

are functions and the operation on G is well-deﬁned. The associative law holds

74 CHAPTER 3 SOLUTIONS

in S because it holds in G; the identity element in S is φ(e), where e is the

identity of G, and it is easy to check that if x ∈ S, then x

−1

= φ((φ

−1

(x))

−1

).

Comment: This reveals the secret behind problems like Exercises 3.1.11 and

3.4.12 in the text. Given a known group G such as R

×

, we can use one-to-

one functions deﬁned on G to produce new groups with operations that look

rather diﬀerent from the usual examples.

3.5 SOLUTIONS

20. Show that the three groups Z

6

, Z

×

9

, and Z

×

18

are isomorphic to each other.

Solution: First, we have [Z

×

9

[ = 6, and [Z

×

18

[ = 6. In Z

×

9

, 2

2

= 4, 2

3

= 8 ,≡ 1,

and so [2] must have order 6, showing that Z

×

9

is cyclic of order 6. Our

theorems tell us that Z

×

9

∼

= Z

6

. In Z

×

18

, 5

2

≡ 7, 5

3

≡ 17 ,≡ 1, and so [5] must

have order 6, showing that Z

×

18

is cyclic of order 6. Our theorems tell us that

Z

×

18

∼

= Z

6

. Thus all three groups are isomorphic.

21. Is Z

4

Z

10

isomorphic to Z

2

Z

20

?

Solution: It follows from Theorem 3.5.4 that Z

10

∼

= Z

2

Z

5

, and that Z

20

∼

=

Z

4

Z

5

. It then follows from Problem 3.4.23 that Z

4

Z

10

∼

= Z

4

Z

2

Z

5

,

and Z

2

Z

20

∼

= Z

2

∼

= Z

4

Z

5

. Finally, it is possible to show that the obvious

mapping from Z

4

Z

2

Z

5

onto Z

2

∼

= Z

4

Z

5

is an isomorphism. Therefore

Z

4

Z

10

∼

= Z

2

Z

20

.

22. Is Z

4

Z

15

isomorphic to Z

6

Z

10

?

Solution: As in Problem 21, Z

4

Z

15

∼

= Z

4

Z

3

Z

5

, and Z

6

Z

10

∼

=

Z

2

Z

3

Z

2

Z

5

. The two groups are not isomorphic since the ﬁrst has an

element of order 4, while the second has none.

23. Give the lattice diagram of subgroups of Z

100

.

Solution: The subgroups correspond to the divisors of 100, and are given in

Figure 3.0.1. Note that nZ

100

is used to mean all multiples of n in Z

100

.

24. Find all generators of the cyclic group Z

28

.

Solution: By Proposition 3.5.3 (b), the generators correspond to u the num-

bers less than 28 and relatively prime to 28. The Euler ϕ-function allows us

to compute how many there are: ϕ(28) =

1

2

6

7

28 = 12. The list of generators

is ¦±1, ±3, ±5, ±9, ±11, ±13¦.

25. In Z

30

, ﬁnd the order of the subgroup ¸[18]

30

); ﬁnd the order of ¸[24]

30

).

Solution: Using Proposition 3.5.3 (a), we ﬁrst ﬁnd gcd(18, 30) = 6. Then

¸[18]

30

) = ¸[6]

30

), and so the subgroup has 30/6 = 5 elements.

Similarly, ¸[24]

30

) = ¸[6]

30

), and so we actually have ¸[24]

30

) = ¸[18]

30

).

CHAPTER 3 SOLUTIONS 75

Figure 3.1: for Problem 23

Z

100

¸ ¸

2Z

100

5Z

100

¸ ¸ ¸ ¸

4Z

100

10Z

100

25Z

100

¸ ¸ ¸ ¸

20Z

100

50Z

100

¸ ¸

¸0)

26. Prove that if G

1

and G

2

are groups of order 7 and 11, respectively, then the

direct product G

1

G

2

is a cyclic group.

Solution: Since 7 and 11 are primes, the groups are cyclic. If a has order 7 in

G

1

and b has order 11 in G

2

, then (a, b) has order lcm[7, 11] = 77 in G

1

G

2

.

Thus G

1

G

2

is cyclic since it has an element whose order is equal to the

order of the group.

27. Show that any cyclic group of even order has exactly one element of order 2.

Solution: If G is cyclic of order 2n, for some positive integer n, then it follows

from Theorem 3.5.2 that G is isomorphic to Z

2n

. Since isomorphisms preserve

orders of elements, we only need to answer the question in Z

2n

. In that group,

the elements of order 2 are the nonzero solutions to the congruence 2x ≡

0 (mod 2n), and since the congruence can be rewritten as x ≡ 0 (mod n), we

see that [n]

2n

is the only element of order 2 in Z

2n

.

28. Use the the result in Problem 27 to show that the multiplicative groups Z

×

15

and Z

×

21

are not cyclic groups.

Solution: In Z

×

15

, both [−1]

15

and [4]

15

are easily checked to have order 2.

In Z

×

21

, we have [8]

2

21

= [64]

21

= [1]

21

, and so [8]

21

and [−1]

21

have order 2.

29. Find all cyclic subgroups of the quaternion group. Use this information to

show that the quaternion group cannot be isomorphic to the subgroup of o4

generated by (1, 2, 3, 4) and (1, 3).

76 CHAPTER 3 SOLUTIONS

Solution: The quaternion group Q = ¦±1, ±i, ±j, ±k¦ is deﬁned in Exam-

ple 3.3.7 of the text (see page 108). The elements satisfy the following identi-

ties: i

2

= j

2

= k

2

= −1 and ij = k, jk = i, ki = j, ji = −k, kj = −i, ik = −j.

The cyclic subgroups ¸−1) = ¦±1¦, ¸±i) = ¦±1, ±i¦, ¸±j) = ¦±1, ±j¦, and

¸±k) = ¦±1, ±k¦ can be found by using the given identities. For example,

i

2

= −1, i

3

= i

2

i = −i, and i

4

= i

2

i

2

= (−1)

2

= 1.

In o4

, let (1, 2, 3, 4) = a and (1, 3) = b. Since a is a cycle of length 4, it

has order 4, with a

2

= (1, 3)(2, 4) and a

3

= a

−1

= (1, 4, 3, 2). To ﬁnd the

subgroup generated by a and b, we have ab = (1, 2, 3, 4)(1, 3) = (1, 4)(2, 3),

a

2

b = (1, 3)(2, 4)(1, 3) = (2, 4), and a

3

b = (1, 4, 3, 2)(1, 3) = (1, 2)(3, 4). On

the other side, we have ba = (1, 3)(1, 2, 3, 4) = (1, 2)(3, 4) = a

3

b, ba

2

=

(1, 3)(1, 3)(2, 4) = (2, 4) = a

2

b, and ba

3

= (1, 3)(1, 4, 3, 2) = (1, 4)(2, 3) =

ab. This shows that the subgroup generated by a and b consists of the 8

elements ¦e, a, a

2

, a

3

, b, ab, a

2

b, a

3

b¦. Furthermore, from the cycle structures

of the elements we can see that the only cyclic subgroup of order 4 is the

one generated by a (and a

3

). In any isomorphism, cyclic subgroups would

correspond to cyclic subgroups, and so it is impossible for this group to be

isomorphic to the quaternion group, which has 3 cyclic subgroups of order 4.

30. Prove that if p and q are diﬀerent odd primes, then Z

×

pq

is not a cyclic group.

Solution: We know that [−1]

pq

has order 2, so by Problem 27 it is enough

to ﬁnd one other element of order 2. The Chinese remainder theorem (The-

orem 1.3.6) states that the system of congruences x ≡ 1 (mod p) and x ≡

−1 (mod q) has a solution [a]

pq

, since p and q are relatively prime. Be-

cause q is an odd prime, [−1]

pq

is not a solution, so [a]

pq

,= [−1]

pq

. But

a

2

≡ 1 (mod p) and a

2

≡ 1 (mod q), so a

2

≡ 1 (mod pq) since p and q are

relatively prime, and thus [a]

pq

has order 2.

3.6 SOLUTIONS

22. In the dihedral group Tn

= ¦a

i

b

j

[ 0 ≤ i < n, 0 ≤ j < 2¦ with o(a) = n,

o(b) = 2, and ba = a

−1

b, show that ba

i

= a

n−i

b, for all 0 ≤ i < n.

Solution: For i = 1, the equation ba

i

= a

n−i

b is just the relation that deﬁnes

the group. If we assume that the result holds for i = k, then for i = k +1 we

have

ba

k+1

= (ba

k

)a = (a

n−k

b)a = a

n−k

(ba) = a

n−k

a

−1

b = a

n−(k+1)

b .

This implies that the result must hold for all i with 0 ≤ i < n.

Comment: This is similar to a proof by induction, but for each given n we

only need to worry about a ﬁnite number of equations.

CHAPTER 3 SOLUTIONS 77

23. In the dihedral group Tn

= ¦a

i

b

j

[ 0 ≤ i < n, 0 ≤ j < 2¦ with o(a) = n,

o(b) = 2, and ba = a

−1

b, show that each element of the form a

i

b has order 2.

Solution: Using the result from the previous problem, we have (a

i

b)

2

=

(a

i

b)(a

i

b) = a

i

(ba

i

)b = a

i

(a

n−i

b)b = (a

i

a

n−i

)(b

2

) = a

n

e = e.

24. In o4

, ﬁnd the subgroup H generated by (1, 2, 3) and (1, 2).

Solution: Let a = (1, 2, 3) and b = (1, 2). Then H must contain a

2

=

(1, 3, 2), ab = (1, 3) and a

2

b = (2, 3), and this set of elements is closed un-

der multiplication. (We have just listed the elements of o3

.) Thus H =

¦(1), a, a

2

, b, ab, a

2

b¦ = ¦(1), (1, 2, 3), (1, 3, 2), (1, 2), (1, 3), (2, 3)¦.

25. For the subgroup H of o4

deﬁned in the previous problem, ﬁnd the corre-

sponding subgroup σHσ

−1

, for σ = (1, 4).

Solution: We need to compute στσ

−1

, for each τ ∈ H. Since (1, 4)

−1

=

(1, 4), we have (1, 4)(1)(1, 4) = (1), and (1, 4)(1, 2, 3)(1, 4) = (2, 3, 4). As

a shortcut, we can use Exercise 2.3.10, which shows that σ(1, 2, 3)σ

−1

=

(σ(1), σ(2), σ(3)) = (4, 2, 3). Then we can quickly do the other computations:

(1, 4)(1, 3, 2)(1, 4)

−1

= (4, 3, 2)

(1, 4)(1, 2)(1, 4)

−1

= (4, 2)

(1, 4)(1, 3)(1, 4)

−1

= (4, 3)

(1, 4)(2, 3)(1, 4)

−1

= (2, 3).

Thus (1, 4)H(1, 4)

−1

= ¦(1), (2, 3, 4), (2, 4, 3), (2, 3), (2, 4), (3, 4)¦.

26. Show that each element in /4

can be written as a product of 3-cycles.

Solution: We ﬁrst list the 3-cycles: (1, 2, 3), (1, 2, 4), (1, 3, 2), (1, 3, 4), (1, 4, 2),

(1, 4, 3), (2, 3, 4), and (2, 4, 3). Rather than starting with each of the other

elements and then trying to write them as a product of 3-cycles, it is easier

to just look at the possible products of 3-cycles. We have (1, 2, 3)(1, 2, 4) =

(1, 3)(2, 4), (1, 2, 4)(1, 2, 3) = (1, 4)(2, 3), (1, 2, 3)(2, 3, 4) = (1, 2)(3, 4), and

this accounts for all 12 of the elements in /4

.

27. In the dihedral group Tn

= ¦a

i

b

j

[ 0 ≤ i < n, 0 ≤ j < 2¦ with o(a) = n,

o(b) = 2, and ba = a

−1

b, ﬁnd the centralizer of a.

Solution: The centralizer C(a) contains all powers of a, so we have ¸a) ⊆ C(a).

This shows that C(a) has at least n elements. On the other hand, C(a) ,= Tn

,

since by deﬁnition b does not belong to C(a). Since ¸a) contains exactly half

of the elements in Tn

, Lagrange’s theorem show that there is no subgroup

that lies strictly between ¸a) and Tn

, so ¸a) ⊆ C(a) ⊆ Tn

and C(a) ,= Tn

together imply that C(a) = ¸a).

28. Find the centralizer of (1, 2, 3) in o3

, in o4

, and in /4

.

78 CHAPTER 3 SOLUTIONS

Solution: Since any power of an element a commutes with a, the central-

izer C(a) always contains the cyclic subgroup ¸a) generated by a. Thus the

centralizer of (1, 2, 3) always contains the subgroup ¦(1), (1, 2, 3), (1, 3, 2)¦.

In o3

, the centralizer of (1, 2, 3) is equal to ¸ (1, 2, 3) ), since it is easy to check

that (1, 2) does not belong to the centralizer, and by Lagrange’s theorem a

proper subgroup of a group with 6 elements can have at most 3 elements. To

ﬁnd the centralizer of (1, 2, 3) in o4

we have to work a bit harder.

It helps to have some shortcuts when doing the necessary computations. To

see that x belongs to C(a), we need to check that xa = ax, or that axa

−1

= x.

Exercise 2.3.10 provides a quick way to do this in a group of permutations.

That exercise shows that if (1, 2, . . . , k) is a cycle of length k and σ is any

permutation, then σ(1, 2, . . . , k)σ

−1

= (σ(1), σ(2), . . . , σ(k)).

Let a = (1, 2, 3). From the computations in o3

, we know that (1, 2), (1, 3),

and (2, 3) do not commute with a. The remaining transpositions in o4

are

(1, 4), (2, 4), and (3, 4). Using Exercise 2.3.10, we have a(1, 4)a

−1

= (2, 4),

a(2, 4)a

−1

= (3, 4), and a(3, 4)a

−1

= (1, 4), so no transposition in o4

com-

mutes with a. For the products of the transposition, we have a(1, 2)(3, 4)a

−1

=

(2, 3)(1, 4), a(1, 3)(2, 4)a

−1

= (2, 1)(3, 4), and a(1, 4)(2, 3)a

−1

= (2, 4)(3, 1),

and so no product of transpositions belongs to C(a).

If we do a similar computation with a 4-cycle, we will have a(x, y, z, 4)a

−1

=

(u, v, w, 4), since a just permutes the numbers x, y, and z. This means that

w ,= z, so (u, v, w, 4) ,= (x, y, z, 4). Without doing all of the calculations, we

can conclude that no 4-cycle belongs to C(a). This accounts for an additional 6

elements. A similar argument shows that no 3-cycle that includes the number

4 as one of its entries can belong to C(a). Since there are 6 elements of this

form, we now have a total of 21 elements that are not in C(a), and therefore

C(a) = ¸a). Finally, in /4

we must get the same answer: C(a) = ¸a).

3.7 SOLUTIONS

17. Find all group homomorphisms from Z

4

into Z

10

.

Solution: Example 3.7.5 shows that any group homomorphism from Z

n

into

Z

k

must have the form φ([x]

n

) = [mx]

k

, for all [x]

n

∈ Z

n

. Under any group

homomorphism φ : Z

4

→ Z

10

, the order of φ([1]

4

) must be a divisor of 4

and of 10, so the only possibilities are 1 and 2. Thus φ([1]

4

) = [0]

10

, which

deﬁnes the zero function, or else φ([1]

4

) = [5]

10

, which leads to the formula

φ([x]

4

) = [5x]

10

, for all [x]

4

∈ Z

4

.

18. (a) Find the formulas for all group homomorphisms from Z

18

into Z

30

.

Solution: Example 3.7.5 shows that any group homomorphism from Z

18

into Z

30

must have the form φ([x]

18

) = [mx]

30

, for all [x]

18

∈ Z

18

. Since

CHAPTER 3 SOLUTIONS 79

gcd(18, 30) = 6, the possible orders of [m]

30

= φ([1]

18

) are 1, 2, 3, 6. The cor-

responding choices for [m]

30

are [0]

30

, of order 1, [15]

30

, of order 2, [10]

30

and

[20]

30

, of order 3, and [5]

30

and [25]

30

, of order 6.

(b) Choose one of the nonzero formulas in part (a), and for this formula ﬁnd

the kernel and image, and show how elements of the image correspond to

cosets of the kernel.

Solution: For example, consider φ([x]

18

) = [5x]

30

. The image of φ consists

of the multiples of 5 in Z

30

, which are 0, 5, 10, 15, 20, 25. We have ker(φ) =

¦0, 6, 12¦, and then cosets of the kernel are deﬁned by adding 1, 2, 3, 4, and

5, respectively. We have the following correspondence

¦0, 6, 12¦ ←→φ(0) = 0, ¦3, 9, 15¦ ←→φ(3) = 15,

¦1, 7, 13¦ ←→φ(1) = 5, ¦4, 10, 16¦ ←→φ(4) = 20,

¦2, 8, 14¦ ←→φ(2) = 10, ¦5, 11, 17¦ ←→φ(5) = 25.

19. (a) Show that Z

×

7

is cyclic, with generator [3]

7

.

Solution: Since 3

2

≡ 2 and 3

3

≡ 6, it follows that [3] must have order 6.

(b) Show that Z

×

17

is cyclic, with generator [3]

17

.

Solution: The element [3] is a generator for Z

×

17

, since 3

2

= 9, 3

3

= 27 ≡ 10,

3

4

≡ 3 10 ≡ 13, 3

5

≡ 3 13 ≡ 5, 3

6

≡ 3 5 ≡ 15, 3

7

≡ 3 15 ≡ 11,

3

8

≡ 3 11 ≡ 16 ,≡ 1.

(c) Completely determine all group homomorphisms from Z

×

17

into Z

×

7

.

Solution: Any group homomorphism φ : Z

×

17

→Z

×

7

is determined by its value

on the generator [3]

17

, and the order of φ([3]

17

) must be a common divisor

of 16 and 6, The only possible orders are 1 and 2, so either φ([3]

17

) = [1]

7

or

φ([3]

17

) = [−1]

7

. In the ﬁrst case, φ([x]

17

) = [1]

7

for all [x]

17

∈ Z

×

17

, and in

the second case φ(([3]

17

)

n

) = [−1]

n

7

, for all [x]

17

= ([3]

17

)n ∈ Z

×

17

.

20. Deﬁne φ : Z

4

Z

6

→Z

4

Z

3

by φ(x, y) = (x + 2y, y).

(a) Show that φ is a well-deﬁned group homomorphism.

Solution: If y

1

≡ y

2

(mod 6), then 2y

1

− 2y

2

is divisible by 12, so 2y

1

≡

2y

2

(mod 4), and then it follows quickly that φ is a well-deﬁned function. It

is also easy to check that φ preserves addition.

(b) Find the kernel and image of φ, and apply the fundamental homomorphism

theorem.

Solution: If (x, y) belongs to ker(φ), then y ≡ 0 (mod 3), so y = 0 or y = 3.

If y = 0, then x = 0, and if y = 3, then x = 2. Thus the elements of the

kernel K are (0, 0) and (2, 3).

It follows that there are 24/2 = 12 cosets of the kernel. These cosets are in

one-to-one correspondence with the elements of the image, so φ must map

Z

4

Z

6

onto Z

4

Z

3

. Thus (Z

4

Z

6

)/¦(0, 0), (2, 3)¦

∼

= Z

4

Z

3

.

80 CHAPTER 3 SOLUTIONS

21. Let n and m be positive integers, such that m is a divisor of n. Show that

φ : Z

×

n

→ Z

×

m

deﬁned by φ([x]

n

) = [x]

m

, for all [x]

n

∈ Z

×

n

, is a well-deﬁned

group homomorphism.

Solution: First, φ is a well-deﬁned function, since if [x

1

]

n

= [x

2

]

n

in Z

×

n

,

then n [ (x

1

− x

2

), and this implies that m [ (x

1

− x

2

), since m [ n. Thus

[x

1

]

m

= [x

2

]

m

, and so φ([x

1

]

n

) = φ([x

2

]

n

).

Next, φ is a homomorphism since for [a]

n

, [b]

n

∈ Z

×

n

, φ([a]

n

[b]

n

) = φ([ab]

n

) =

[ab]

m

= [a]

m

[b]

m

= φ([a]

n

)φ([b]

n

).

22. For the group homomorphism φ : Z

×

36

→ Z

×

12

deﬁned by φ([x]

36

) = [x]

12

, for

all [x]

36

∈ Z

×

36

, ﬁnd the kernel and image of φ, and apply the fundamental

homomorphism theorem.

Solution: The previous problem shows that φ is a group homomorphism. It is

evident that φ maps Z

×

36

onto Z

×

12

, since if gcd(x, 12) = 1, then gcd(x, 36) = 1.

The kernel of φ consists of the elements in Z

×

36

that are congruent to 1 mod

12, namely [1]

36

, [13]

36

, [25]

36

. It follows that Z

×

12

∼

= Z

×

36

/ ¸[13]

36

).

23. Let G, G

1

, and G

2

be groups. Let φ

1

: G → G

1

and φ

2

: G → G

2

be

group homomorphisms. Prove that φ : G → G

1

G

2

deﬁned by φ(x) =

(φ

1

(x), φ

2

(x)), for all x ∈ G, is a well-deﬁned group homomorphism.

Solution: Given a, b in G, we have

φ(ab) = (φ

1

(ab), φ

2

(ab))

= (φ

1

(a)φ

1

(b), φ

2

(a)φ

2

(b))

φ(a)φ(b) = (φ

1

(a), φ

2

(a)) (φ

1

(b), φ

2

(b))

= (φ

1

(a)φ

1

(b), φ

2

(a)φ

2

(b))

and so φ : G →G

1

G

2

is a group homomorphism.

24. Let p and q be diﬀerent odd primes. Prove that Z

×

pq

is isomorphic to the direct

product Z

×

p

Z

×

q

.

Solution: Using Problem 21, we can deﬁne group homomorphisms φ

1

: Z

×

pq

→

Z

×

p

and φ

2

: Z

×

pq

→ Z

×

q

by setting φ

1

([x]

pq

) = [x]

p

, for all [x]

pq

∈ Z

×

pq

, and

φ

2

([x]

pq

) = [x]

q

, for all [x]

pq

∈ Z

×

pq

.

Using Problem 23, we can deﬁne a group homomorphism φ : Z

×

pq

→Z

×

p

Z

×

q

by setting φ([x]

pq

) = (φ

1

([x]

pq

), φ

2

([x]

pq

)), for all [x]

pq

∈ Z

×

pq

. If [x]

pq

∈

ker(φ), then [x]

p

= [1]

p

and [x]

q

= [1]

q

, so p [ (x −1) and q [ (x −1), and this

implies that pq [ (x − 1), since p adn q are relatively prime. It follows that

[x]

pq

= [1]

pq

, and this shows that φ is a one-to-one function. Exercise 1.4.27

in the text states that if m > 0 and n > 0 are relatively prime integers, then

ϕ(mn) = ϕ(m)ϕ(n). It follows that Z

×

pq

and Z

×

p

Z

×

q

have the same order,

so φ is also an onto function. This completes the proof that φ is a group

isomorphism.

CHAPTER 3 SOLUTIONS 81

3.8 SOLUTIONS

27. List the cosets of ¸7) in Z

×

16

. Is the factor group Z

×

16

/ ¸7) cyclic?

Solution: Z

×

16

= ¦1, 3, 5, 7, 9, 11, 13, 15¦.

¸7) = ¦1, 7¦ 3 ¸7) = ¦3, 5¦ 9 ¸7) = ¦9, 15¦ 11 ¸7) = ¦11, 13¦

Since 3

2

,∈ ¸7), the coset 3 ¸7) does not have order 2, so it must have order 4,

showing that the factor group is cyclic.

28. Let G = Z

6

Z

4

, let H = ¦(0, 0), (0, 2)¦, and let K = ¦(0, 0), (3, 0)¦.

(a) List all cosets of H; list all cosets of K.

Solution: The cosets of H = ¦(0, 0), (0, 2)¦ are

(0, 0) +H = ¦(0, 0), (0, 2)¦ (1, 0) +H = ¦(1, 0), (1, 2)¦

(2, 0) +H = ¦(2, 0), (2, 2)¦ (3, 0) +H = ¦(3, 0), (3, 2)¦

(4, 0) +H = ¦(4, 0), (4, 2)¦ (5, 0) +H = ¦(5, 0), (5, 2)¦

(0, 1) +H = ¦(0, 1), (0, 3)¦ (1, 1) +H = ¦(1, 1), (1, 3)¦

(2, 1) +H = ¦(2, 1), (2, 3)¦ (3, 1) +H = ¦(3, 1), (3, 3)¦

(4, 1) +H = ¦(4, 1), (4, 3)¦ (5, 1) +H = ¦(5, 1), (5, 3)¦

The cosets of K = ¦(0, 0), (3, 0)¦ are

(0, 0) +K = ¦(0, 0), (3, 0)¦ (0, 1) +K = ¦(0, 1), (3, 1)¦

(0, 2) +K = ¦(0, 2), (3, 2)¦ (0, 3) +K = ¦(0, 3), (3, 3)¦

(1, 0) +K = ¦(1, 0), (4, 0)¦ (1, 1) +K = ¦(1, 1), (4, 1)¦

(1, 2) +K = ¦(1, 2), (4, 2)¦ (1, 3) +K = ¦(1, 3), (4, 3)¦

(2, 0) +K = ¦(2, 0), (5, 0)¦ (2, 1) +K = ¦(2, 1), (5, 1)¦

(2, 2) +K = ¦(2, 2), (5, 2)¦ (2, 3) +K = ¦(2, 3), (5, 3)¦

(b) You may assume that any abelian group of order 12 is isomorphic to either

Z

12

or Z

6

Z

2

. Which answer is correct for G/H? For G/K?

Solution: Adding an element of G to itself 6 times yields a 0 in the ﬁrst

component and either 0 or 2 in the second component, producing an element in

H. Thus the order of an element in G/H is at most 6, and so G/H

∼

= Z

6

Z

2

.

On the other hand, (1, 1) +K has order 12 in G/K, and so G/K

∼

= Z

12

.

29. Let the dihedral group D

n

be given via generators and relations, with gener-

ators a of order n and b of order 2, satisfying ba = a

−1

b.

(a) Show that ba

i

= a

−i

b for all i with 1 ≤ i < n.

Solution: The identity holds for all positive integers i, and can be proved in-

ductively: assuming ba

k

= a

−k

b, we have ba

k+1

= ba

k

a = a

−k

ba = a

−k

a

−1

b =

a

−(k+1)

b.

82 CHAPTER 3 SOLUTIONS

(b) Show that any element of the form a

i

b has order 2.

Solution: We have (a

i

b)

2

= a

i

ba

i

b = a

i

a

−i

b

2

= a

0

= e.

(c) List all left cosets and all right cosets of ¸b)

Solution: The left cosets of ¸b) have the form a

i

¸b) = ¦a

i

, a

i

b¦, for 0 ≤ i < n.

The right cosets of ¸b) have the form ¸b) a

i

= ¦a

i

, a

−i

b¦, for 0 ≤ i < n.

30. Let G = D

6

and let N be the subgroup

¸

a

3

_

= ¦e, a

3

¦ of G.

(a) Show that N is a normal subgroup of G.

Solution: The argument is the same as in the previous problem.

(b) Is G/N abelian?

Solution: For aN = ¦a, a

4

¦ and bN = ¦b, a

3

b¦, we have (aN)(bN) = abN =

¦ab, a

4

b¦, while (bN)(aN) = baN = a

5

bN = ¦a

5

b, a

2

b¦. Thus (aN)(bN) ,=

(bN)(aN), and G/N is not abelian.

31. Let G be the dihedral group D

12

, and let N = ¦e, a

3

, a

6

, a

9

¦.

(a) Prove that N is a normal subgroup of G, and list all cosets of N.

Solution: Since N =

¸

a

3

_

, it is a subgroup. It is normal since a

i

(a

3n

)a

−i

=

a

3n

and a

i

b(a

3n

)a

i

b = a

i

a

−3n

a

−i

= (a

3n

)

−1

. (We are using the fact that

ba

i

= a

−i

b.)

The cosets of N are

N = ¦e, a

3

, a

6

, a

9

¦, Nb = ¦ab, a

3

b, a

6

b, a

9

b¦,

Na = ¦a, a

4

, a

7

, a

10

¦, Nab = ¦ab, a

4

b, a

7

b, a

10

b¦,

Na

2

= ¦a

2

, a

5

, a

8

, a

11

¦, Na

2

b = ¦a

2

b, a

5

b, a

8

b, a

11

b¦.

(b) You may assume that G/N is isomorphic to either Z

6

or S

3

. Which is

correct?

Solution: The factor group G/N is not abelian, since NaNb = Nab but

NbNa = Na

2

b, because ba = a

11

b ∈ Na

2

b. Thus G/N

∼

= S

3

.

32. (a) Let G be a group. For a, b ∈ G we say that b is conjugate to a, written

b ∼ a, if there exists g ∈ G such that b = gag

−1

. Show that ∼ is an equivalence

relation on G. The equivalence classes of ∼ are called the conjugacy classes

of G.

Solution: We have a ∼ a since we can use g = e. If b ∼ a, the b = gag

−1

for some g ∈ G, and so a = g

−1

bg = g

−1

b(g

−1

)

−1

, which shows that a ∼ b.

If c ∼ b and b ∼ a, then c = gbg

−1

and b = hah

−1

for some g, h ∈ G, so

c = g(hah

−1

)g

−1

= (gh)a(gh)

−1

, which shows that c ∼ a. Thus ∼ is an

equivalence relation.

(b) Show that a subgroup N of G is normal in G if and only if N is a union

of conjugacy classes.

CHAPTER 3 SOLUTIONS 83

Solution: The subgroup N is normal in G if and only if a ∈ N implies

gag

−1

∈ G, for all g ∈ G. Thus N is normal if and only if whenever it

contains an element a it also contains the conjugacy class of a. Another way

to say this is that N is a union of conjugacy classes.

33. Find the conjugacy classes of D

4

.

Solution: Remember: the notion of a conjugacy class was just deﬁned in the

previous exercise. Let D

4

= ¦e, a, a

2

, a

3

, b, ab, a

2

b, a

3

b¦, with a

4

= e, b

2

= e,

and ba = a

−1

b. Since xex

−1

= e, the only element conjugate to e is e itself.

If x is any power of a, then x commutes with a, and so xax

−1

= a. If x = a

i

b,

then xax

−1

= a

i

baa

−i

b = a

i

a

i−1

b

2

= a

2i−1

, so this shows that a

3

is the only

conjugate of a (other than a itself).

The solution of an earlier problem shows that xa

2

x

−1

= a

2

in D

4

, so a

2

is not

conjugate to any other element.

If x = a

i

, then xbx

−1

= a

i

ba

−i

= a

i

a

i

b = a

2i

b. If x = a

i

b, then xbx

−1

=

(a

i

b)b(a

i

b)

−1

= a

i

a

i

b = a

2i

b. Thus a

2

b is the only conjugate of b.

If x = a

i

, then x(ab)x

−1

= a

i

aba

−i

= a

i+1

a

i

b = a

2i+1

b. If x = a

i

b, then

xabx

−1

= (a

i

b)ab(a

i

b)

−1

= a

i

a

−1

a

i

b = a

2i−1

b. Thus a

3

b is the only conjugate

of ab.

34. Let G be a group, and let N and H be subgroups of G such that N is normal

in G.

(a) Prove that HN is a subgroup of G.

Solution: See Proposition 3.3.2. It is clear that e = e e belongs to the set

HN, so HN is nonempty. Suppose that x, y belong to HN. Then x = h

1

n

1

and y = h

2

n

2

, for some h

1

, h

2

∈ H and some n

1

, n

2

∈ N. We have

xy

−1

= h

1

n

1

(h

2

n

2

)

−1

= h

1

n

1

n

−1

2

h

−1

2

= (h

1

h

−1

2

)(h

2

(n

1

n

−1

2

)h

−1

2

),

and this element belongs to HN since the assumption that N is normal guar-

antees that h

2

(n

1

n

−1

2

)h

−1

2

∈ N.

(b) Prove that N is a normal subgroup of HN.

Solution: Since N is normal in G, it is normal in the subgroup HN, which

contains it.

(c) Prove that if H ∩ N = ¦e¦, then HN/N is isomorphic to H.

Solution: Deﬁne φ : H → HN/N by φ(x) = xN for all x ∈ H. (Deﬁning

a function from HN/N into H is more complicated.) Then φ(xy) = xyN =

xNyN = φ(x)φ(y) for all x, y ∈ H. Any coset of N in HN has the form

hnN for some h ∈ H and some n ∈ N. But then hnN = hN = φ(h), and

so this shows that φ is onto. Finally, φ is one-to-one since if h ∈ H belongs

to the kernel of φ, then hN = φ(h) = N, and so h ∈ N. By assumption,

H ∩ N = ¦e¦, and so h = e.

84 CHAPTER 3 SOLUTIONS

SOLUTIONS TO THE REVIEW PROBLEMS

1. (a) What are the possibilities for the order of an element of Z

×

13

? Explain

your answer.

Solution: The group Z

×

13

has order 12, and the order of any element must be

a divisor of 12, so the possible orders are 1, 2, 3, 4, 6, and 12.

(b) Show that Z

×

13

is a cyclic group.

Solution: The ﬁrst element to try is [2], and we have 2

2

= 4, 2

3

= 8, 2

4

=

16 ≡ 3, 2

5

≡ 2 2

4

≡ 6, and 2

6

≡ 2 2

5

≡ 12, so the order of [2] is greater than

6. By part (a) it must be 12, and thus [2] is a generator for Z

×

13

. We could

also write this as Z

×

13

= ¸[2]

13

).

2. Find all subgroups of Z

×

11

, and give the lattice diagram which shows the

inclusions between them.

Solution: First check for cyclic subgroups, in shorthand notation: 2

2

= 4,

2

3

= 8, 2

4

= 5, 2

5

= 10, 2

6

= 9, 2

7

= 7, 2

8

= 3, 2

9

= 6, 2

10

= 1. This

shows that Z

×

11

is cyclic, so the subgroups are as follows, in addition to Z

×

11

and ¦[1]¦:

¸

[2]

2

_

= ¦[1], [2]

2

, [2]

4

, [2]

6

, [2]

8

¦ = ¦[1], [4], [5], [9], [3]¦ and

¸

[2]

5

_

=

¦[1], [2]

5

¦ = ¦[1], [10]¦ The lattice diagram forms a diamond.

3. Let G be the subgroup of GL

3

(R) consisting of all matrices of the form

_

_

1 a b

0 1 0

0 0 1

_

_

such that a, b ∈ R .

Show that G is a subgroup of GL

3

(R).

Solution: We have

_

_

1 a b

0 1 0

0 0 1

_

_

_

_

1 c d

0 1 0

0 0 1

_

_

=

_

_

1 a +c b +d

0 1 0

0 0 1

_

_

, so

the closure property holds. The identity matrix belongs to the set, and

_

_

1 a b

0 1 0

0 0 1

_

_

−1

=

_

_

1 −a −b

0 1 0

0 0 1

_

_

, so the set is closed under taking in-

verses.

4. Show that the group G in the previous problem is isomorphic to the direct

product RR.

Solution: Deﬁne φ : G → R R by φ

_

_

_

_

1 a b

0 1 0

0 0 1

_

_

_

_

= (a, b). This is

one-to-one and onto because it has an inverse function θ : RR →G deﬁned

CHAPTER 3 SOLUTIONS 85

by θ((a, b)) =

_

_

1 a b

0 1 0

0 0 1

_

_

. Finally, φ preserves the respective operations

since φ

_

_

_

_

1 a b

0 1 0

0 0 1

_

_

_

_

1 c d

0 1 0

0 0 1

_

_

_

_

= φ

_

_

_

_

1 a +c b +d

0 1 0

0 0 1

_

_

_

_

=

(a +c, b +d) = (a, b) + (c, d) = φ

_

_

_

_

1 a b

0 1 0

0 0 1

_

_

_

_

+φ

_

_

_

_

1 c d

0 1 0

0 0 1

_

_

_

_

.

5. List the cosets of the cyclic subgroup ¸9) in Z

×

20

. Is Z

×

20

/ ¸9) cyclic?

Solution: Z

×

20

= ¦±1, ±3, ±7, ±9¦.

¸9) = ¦1, 9¦ (−1) ¸9) = ¦−1, −9¦ 3 ¸9) = ¦3, 7¦ (−3) ¸9) = ¦−3, −7¦

Since x

2

∈ ¸9), for each element x of Z

×

20

, the factor group is not cyclic.

6. Let G be the subgroup of GL

2

(R) consisting of all matrices of the form

_

m b

0 1

_

, and let N be the subset of all matrices of the form

_

1 b

0 1

_

.

(a) Show that N is a subgroup of G, and that N is normal in G.

Solution: The set N is nonempty since it contains the identity matrix, and

it is a subgroup since

_

1 b

0 1

_ _

1 c

0 1

_

−1

=

_

1 b

0 1

_ _

1 −c

0 1

_

=

_

1 b −c

0 1

_

. N is normal in G since

_

m b

0 1

_ _

1 c

0 1

_ _

m b

0 1

_

−1

=

_

m mc +b

0 1

_ _

1/m −b/m

0 1

_

=

_

1 mc

0 1

_

∈ N.

(b) Show that G/N is isomorphic to the multiplicative group R

×

.

Solution: Deﬁne φ : G → R

×

by φ

__

m b

0 1

__

= m. Then we have

φ

__

m b

0 1

_ _

n c

0 1

__

= φ

__

mn mc +b

0 1

__

= mn =

φ

__

m b

0 1

__

φ

__

n c

0 1

__

. Since m can be any nonzero real number, φ

maps G onto R

×

, and φ

__

m b

0 1

__

= 1 if and only if m = 1, so N = ker(φ).

The fundamental homomorphism theorem implies that G/N

∼

= R

×

.

Note that this part of the proof covers part (a), since once you have determined

the kernel, it is always a normal subgroup. Thus parts (a) and (b) can be

proved at the same time, using the argument given for part (b).

86 CHAPTER 3 SOLUTIONS

7. Assume that the dihedral group D

4

is given as ¦e, a, a

2

, a

3

, b, ab, a

2

b, a

3

b¦,

where a

4

= e, b

2

= e, and ba = a

3

b. Let N be the subgroup

¸

a

2

_

= ¦e, a

2

¦.

(a) Show by a direct computation that N is a normal subgroup of D

4

.

Solution: We have a

i

a

2

a

−i

= a

2

and (a

i

b)a

2

(a

i

b)

−1

= a

i

a

−2

ba

i

b =

a

i

a

−2

a

−i

b

2

= a

−2

= a

2

, for all i, which implies that N is normal.

(b) Is the factor group D

4

/N a cyclic group?

Solution: The cosets of N are

N = ¦e, a

2

¦, Na = ¦a, a

3

¦, Nb = ¦b, a

2

b¦, and Nab = ¦ab, a

3

b¦.

Since b and ab have order 2, and a

2

∈ N, we see that each element in the

factor group has order 2, so G/N is not cyclic.

8. Let G = D

8

, and let N = ¦e, a

2

, a

4

, a

6

¦.

(a) List all left cosets and all right cosets of N, and verify that N is a normal

subgroup of G.

Solution: The right cosets of N are

N = ¦e, a

2

, a

4

, a

6

¦, Na = ¦a, a

3

, a

5

, a

7

¦,

Nb = ¦b, a

2

b, a

4

b, a

6

b¦, Nab = ¦ab, a

3

b, a

5

b, a

7

b¦.

The left cosets of N are more trouble to compute, but we get

N = ¦e, a

2

, a

4

, a

6

¦, aN = ¦a, a

3

, a

5

, a

7

¦,

bN = ¦b, a

6

b, a

4

b, a

2

b¦, abN = ¦ab, a

7

b, a

5

b, a

3

b¦.

The fact that the left and right cosets of N coincide shows that N is normal.

(b) Show that G/N has order 4, but is not cyclic.

Solution: It is clear that there are 4 cosets. We have NaNa = Na

2

= N,

NbNb = Ne = N, and NabNab = Ne = N, so each coset has order 2.

Chapter 4

Polynomials

SOLUTIONS TO THE REVIEW PROBLEMS

1. Use the Euclidean algorithm to ﬁnd gcd(x

8

−1, x

6

−1) in Q[x] and write it

as a linear combination of x

8

−1 and x

6

−1.

Solution: Let x

8

−1 = f(x) and x

6

−1 = g(x). We have f(x) = x

2

g(x)+(x

2

−

1), and g(x) = (x

4

+x

2

+1)(x

2

−1), so this shows that gcd(x

8

−1, x

6

−1) =

x

2

−1, and x

2

−1 = f(x) −x

2

g(x).

2. Over the ﬁeld of rational numbers, use the Euclidean algorithm to show that

2x

3

−2x

2

−3x + 1 and 2x

2

−x −2 are relatively prime.

Solution: Let 2x

3

− 2x

2

− 3x + 1 = f(x) and 2x

2

− x − 2 = g(x). We ﬁrst

obtain f(x) = (x −

1

2

)g(x) −

3

2

x. At the next step we can use x rather than

3

2

x, and then g(x) = (2x −1)g(x) −2. The constant remainder at the second

step implies that gcd(f(x), g(x)) = 1.

3. Over the ﬁeld of rational numbers, ﬁnd the greatest common divisor of

x

4

+ x

3

+ 2x

2

+ x + 1 and x

3

− 1, and express it as a linear combination of

the given polynomials.

Solution: Let x

4

+ x

3

+ 2x

2

+ x + 1 = f(x) and x

3

− 1 = g(x). We ﬁrst

obtain f(x) = (x + 1)g(x) + 2(x

2

+ x + 1), and then the next step yields

g(x) = (x−1)(x

2

+x+1), so gcd(f(x), g(x)) = x

2

+x+1, and (x

2

+x+1) =

1

2

f(x) −

1

2

(x + 1)g(x).

4. Over the ﬁeld of rational numbers, ﬁnd the greatest common divisor of

2x

4

− x

3

+ x

2

+ 3x + 1 and 2x

3

− 3x

2

+ 2x + 2 and express it as a linear

combination of the given polynomials.

Solution: To simplify the computations, let 2x

4

− x

3

+ x

2

+ 3x + 1 = f(x)

and 2x

3

−3x

2

+2x+2 = g(x). Using the Euclidean algorithm, we ﬁrst obtain

87

88 CHAPTER 4 SOLUTIONS

f(x) = (x+1)g(x)+(2x

2

−x−1), and then g(x) = (x−1)(2x

2

−x−1)+(2x+1).

At the next step we obtain 2x

2

− x − 1 = (x − 1)(2x + 1), so 2x + 1 is the

greatest common divisor (we must then divide by 2 to make it monic).

Beginning with the last equation and back-solving, we get

2x + 1 = g(x) −(x −1)(2x

2

−x −1)

= g(x) −(x −1)(f(x) −(x + 1)g(x))

= g(x) + (x

2

−1)g(x) −(x −1)f(x)

= x

2

g(x) −(x −1)f(x)

This gives the ﬁnal answer, x +

1

2

=

1

2

x

2

g(x) + (−

1

2

)(x −1)f(x).

5. Are the following polynomials irreducible over Q?

(a) 3x

5

+ 18x

2

+ 24x + 6

Solution: Dividing by 3 we obtain x

5

+6x

2

+8x +2, and this satisﬁes Eisen-

stein’s criterion for p = 2.

(b) 7x

3

+ 12x

2

+ 3x + 45

Solution: Reducing the coeﬃcients modulo 2 gives the polynomial x

3

+x+1,

which is irreducible in Z

2

[x]. This implies that the polynomial is irreducible

over Q.

(c) 2x

10

+ 25x

3

+ 10x

2

−30

Solution: Eisenstein’s criterion is satisﬁed for p = 5.

6. Factor x

5

−10x

4

+ 24x

3

+ 9x

2

−33x −12 over Q.

Solution: The possible rational roots of f(x) = x

5

−10x

4

+24x

3

+9x

2

−33x−12

are ±1, ±2, ±3, ±4, ±6, ±12. We have f(1) = 21, so for any root we must

have (r − 1)[21, so this eliminates all but ±2, 4, −6 as possibilities. Then

f(2) = 32, f(−2) = −294, and ﬁnally we obtain the factorization f(x) =

(x − 4)(x

4

− 6x

3

+ 9x + 3). The second factor is irreducible over Q since it

satisﬁes Eisenstein’s criterion for p = 3.

7. Factor x

5

−2x

4

−2x

3

+ 12x

2

−15x −2 over Q.

Solution: The possible rational roots are ±1, ±2, and since 2 is a root we have

the factorization x

5

−2x

4

−2x

3

+12x

2

−15x−2 = (x−2)(x

4

−2x

2

+8x+1).

The only possible rational roots of the second factor are 1 and −1, and these

do not work. (It is important to note that since the degree of the polynomial

is greater than 3, the fact that it has not roots in Q does not mean that it

is irreducible over Q.) Since the polynomial has no linear factors, the only

possible factorization has the form x

4

−2x

2

+8x+1 = (x

2

+ax+b)(x

2

+cx+d).

This leads to the equations a+c = 0, ac+b+d = −2, ad+bc = 8, and bd = 1.

We have either b = d = 1, in which case a + c = 8, or b = d = −1, in which

CHAPTER 4 SOLUTIONS 89

case a + c = −8. Either case contradicts a + c = 0, so x

4

− 2x

2

+ 8x + 1 is

irreducible over Q.

As an alternate solution, we could reduce x

4

− 2x

2

+ 8x + 1 modulo 3 to

get p(x) = x

4

+ x

2

+ 2x + 1. This polynomial has no roots in Z

3

, so the

only possible factors are of degree 2. The monic irreducible polynomials of

degree 2 over Z

3

are x

2

+ 1, x

2

+x + 2, and x

2

+ 2x + 2. Since the constant

term of p(x) is 1, the only possible factorizations are p(x) = (x

2

+ x + 2)

2

,

p(x) = (x

2

+ 2x + 2)

2

, or p(x) = (x

2

+ x + 2)(x

2

+ 2x + 2). In the ﬁrst the

coeﬃcient of x is 1; the second has a nonzero cubic term; in the third the

coeﬃcient of x is 0. Thus p(x) is irreducible over Z

3

, and hence over Q.

8. (a) Show that x

2

+ 1 is irreducible over Z

3

.

Solution: To show that p(x) = x

2

+ 1 is irreducible over Z

3

, we only need

to check that it has no roots in Z

3

, and this follows from the computations

p(0) = 1, p(1) = 2, and p(−1) = 2.

(b) List the elements of the ﬁeld F = Z

3

[x]/

¸

x

2

+ 1

_

.

Solution: The congruence classes are in one-to-one correspondence with the

linear polynomials, so we have the nine elements [0], [1], [2], [x], [x+1], [x+2],

[2x], [2x + 1], [2x + 2].

(c) In the multiplicative group of nonzero elements of F, show that [x + 1] is

a generator, but [x] is not.

Solution: The multiplicative group of F has 8 elements, and since [x]

2

= [−1],

it follows that [x] has order 4 and is not a generator. On the other hand,

[x + 1]

2

= [x

2

+ 2x + 1] = [−1 + 2x + 1] = [2x] = [−x], and so [x + 1]

4

=

[−x]

2

= [−1], which shows that [x + 1] does not have order 2 or 4. The only

remaining possibility (by Lagrange’s theorem) is that [x +1] has order 8, and

so it is a generator for the multiplicative group of F.

9. (a) Express x

4

+x as a product of polynomials irreducible over Z

5

.

Solution: In general, we have x

4

+ x = x(x

3

+ 1) = x(x + 1)(x

2

− x + 1).

The factor p(x) = x

2

− x + 1 is irreducible over Z

5

since it can be checked

that it has no roots in Z

5

. (We get p(0) = 1, p(1) = 1, p(−1) = 3, p(2) = 3,

p(−2) = 2.)

(b) Show that x

3

+ 2x

2

+ 3 is irreducible over Z

5

.

Solution: If p(x) = x

3

+ 2x

2

+ 3, then p(0) = 3, p(1) = 1, p(−1) = −1,

p(2) = 4, and p(−2) = 3, so p(x) is irreducible over Z

5

.

10. Express 2x

3

+x

2

+ 2x + 2 as a product of polynomials irreducible over Z

5

.

Solution: We ﬁrst factor out 2, using (2)(−2) = −4 ≡ 1 (mod 5). This

reduces the question to factoring p(x) = x

3

− 2x

2

+ x + 1. (We could also

multiply each term by 3.) Checking for roots shows that p(0) = 1, p(1) = 1,

p(−1) = −3, p(2) = 3, and p(−2) ≡ −2, so p(x) itself is irreducible over Z

5

.

90 CHAPTER 4 SOLUTIONS

11. Construct an example of a ﬁeld with 343 = 7

3

elements.

Solution: We only need to ﬁnd a cubic polynomial over Z

7

that has no roots.

The simplest case would be to look for a polynomial of the form x

3

+a. The

cube of any element of Z

7

gives either 1 or −1, so x

3

= 2 has no root over Z

7

,

and thus p(x) = x

3

−2 is an irreducible cubic over Z

7

. Using the modulus p(x),

the elements of Z

7

[x]/ ¸p(x)) correspond to polynomials of degree 2 or less,

giving the required 7

3

elements. With this modulus, the identities necessary

to determine multiplication are [x

3

] = [5] and [x

4

] = [5x].

12. In Z

2

[x]/

¸

x

3

+x + 1

_

, ﬁnd the multiplicative inverse of [x + 1].

Solution: We ﬁrst give a solution using the Euclidean algorithm. For p(x) =

x

3

+ x + 1 and f(x) = x + 1, the ﬁrst step of the Euclidean algorithm gives

p(x) = (x

2

+x)f(x)+1. Thus p(x)−(x

2

+x)f(x) = 1, and so reducing modulo

p(x) gives [−x

2

−x][f(x)] = [1], and thus [x + 1]

−1

= [−x

2

−x] = [x

2

+x].

We next give an alternate solution, which uses the identity [x

3

] = [x + 1] to

solve a system of equations. We need to solve [1] = [x + 1][ax

2

+bx +c] or

[1] = [ax

3

+bx

2

+cx +ax

2

+bx +c]

= [ax

3

+ (a +b)x

2

+ (b +c)x +c]

= [a(x + 1) + (a +b)x

2

+ (b +c)x +c]

= [(a +b)x

2

+ (a +b +c)x + (a +c)] ,

so we need a +b ≡ 0 (mod 2), a +b +c ≡ 0 (mod 2), and a +c ≡ 1 (mod 2).

This gives c ≡ 0 (mod 2), and therefore a ≡ 1 (mod 2), and then b ≡

1 (mod 2). Again, we see that [x + 1]

−1

= [x

2

+x].

13. Find the multiplicative inverse of [x

2

+x + 1]

(a) in Q[x]/

¸

x

3

−2

_

;

Solution: Using the Euclidean algorithm, we have

x

3

−2 = (x

2

+x + 1)(x −1) + (−1), and so [x

2

+x + 1]

−1

= [x −1].

This can also be done by solving a system of 3 equations in 3 unknowns.

(b) in Z

3

[x]/

¸

x

3

+ 2x

2

+x + 1

_

.

Solution: Using the Euclidean algorithm, we have

x

3

+ 2x

2

+x + 1 = (x + 1)(x

2

+x + 1) + (−x) and

x

2

+x + 1 = (−x −1)(−x) + 1. Then a substitution gives us

1 = (x

2

+x + 1) + (x + 1)(−x)

= (x

2

+x + 1) + (x + 1)((x

3

+ 2x

2

+x + 1) −(x + 1)(x

2

+x + 1))

= (−x

2

−2x)(x

2

+x + 1) + (x + 1)(x

3

+x

2

+ 2x + 1) .

Thus [x

2

+x +1]

−1

= [−x

2

−2x] = [2x

2

+x]. This can be checked by ﬁnding

[x

2

+x+1][2x

2

+x], using the identities [x

3

] = [x

2

−x−1] and [x

4

] = [x−1].

CHAPTER 4 SOLUTIONS 91

This can also be done by solving a system of equations, or, since the set is

ﬁnite, by taking successive powers of [x

2

+ x + 1]. The latter method isn’t

really practical, since the multiplicative group has order 26, and this element

turns out to have order 13.

14. In Z

5

[x]/

¸

x

3

+x + 1

_

, ﬁnd [x]

−1

and [x +1]

−1

, and use your answers to ﬁnd

[x

2

+x]

−1

.

Solution: Using the division algorithm, we obtain x

3

+x+1 = x(x

2

+1) +1,

and so [x][x

2

+ 1] = [−1]. Thus [x]

−1

= [−x

2

−1].

Next, we have x

3

+x+1 = (x+1)(x

2

−x+2)−1, and so [x+1]

−1

= [x

2

−x+2].

Finally, we have

[x

2

+x]

−1

= [x]

−1

[x + 1]

−1

= [−x

2

−1][x

2

−x + 2]

= [−x

4

+x

3

−2x

2

−x

2

+x −2] .

Using the identities [x

3

] = [−x −1] and [x

4

] = [−x

2

−x], this reduces to

[x

2

+x]

−1

= [x

2

+x −x −1 −3x

2

+x −2]

= [−2x

2

+x −3] = [3x

2

+x + 2] .

15. Factor x

4

+x + 1 over Z

2

[x]/

¸

x

4

+x + 1

_

.

Solution: There are 4 roots of x

4

+ x + 1 in the given ﬁeld, given by the

cosets corresponding to x, x

2

, x + 1, x

2

+ 1. This can be shown by using the

multiplication table, with the elements in the form 10, 100, 11, and 101, or

by computing with polynomials, using the fact that (a + b)

2

= a

2

+ b

2

since

2ab = 0. We have x

4

+x + 1 ≡ 0,

(x

2

)

4

+ (x

2

) + 1 = (x

4

)

2

+x

2

+ 1 ≡ (x + 1)

2

+x

2

+ 1 ≡ x

2

+ 1 +x

2

+ 1 ≡ 0,

(x + 1)

4

+ (x + 1) + 1 ≡ x

4

+ 1 +x ≡ x + 1 + 1 +x ≡ 0, and

(x

2

+1)

4

+(x

2

+1)+1 ≡ (x

4

)

2

+1+x

2

≡ (x+1)

2

+1+x

2

≡ x

2

+1+1+x

2

≡ 0.

Thus x

4

+x + 1 factors as a product of 4 linear terms.

92 CHAPTER 4 SOLUTIONS

Chapter 5

Commutative Rings

SOLUTIONS TO THE REVIEW PROBLEMS

1. Let R be the ring with 8 elements consisting of all 3 3 matrices with entries

in Z

2

which have the following form:

_

_

a 0 0

0 a 0

b c a

_

_

You may assume that the standard laws for addition and multiplication of

matrices are valid.

(a) Show that R is a commutative ring (you only need to check closure and

commutativity of multiplication).

Solution: It is clear that the set is closed under addition, and the following

computation checks closure under multiplication.

_

_

a 0 0

0 a 0

b c a

_

_

_

_

x 0 0

0 x 0

y z x

_

_

=

_

_

ax 0 0

0 ax 0

bx +ay cx +az ax

_

_

Because of the symmetry a ↔ x, b ↔ y, c ↔ z, the above computation also

checks commutativity.

(b) Find all units of R, and all nilpotent elements of R.

Solution: Four of the matrices in R have 1’s on the diagonal, and these are

invertible since their determinant is nonzero. Squaring each of the other four

matrices gives the zero matrix, and so they are nilpotent.

(c) Find all idempotent elements of R.

93

94 CHAPTER 5 SOLUTIONS

Solution: By part (b), an element in R is either a unit or nilpotent. The only

unit that is idempotent is the identity matrix (in a group, the only idempotent

element is the identity) and the only nilpotent element that is also idempotent

is the zero matrix.

2. Let R be the ring Z

2

[x]/

¸

x

2

+ 1

_

. Show that although R has 4 elements, it

is not isomorphic to either of the rings Z

4

or Z

2

⊕Z

2

.

Solution: In R we have a+a = 0, for all a ∈ R, so R is not isomorphic to Z

4

.

On the other hand, in R we have [x + 1] ,= [0] but [x + 1]

2

= [x

2

+ 1] = [0].

Thus R cannot be isomorphic to Z

2

⊕ Z

2

, since in that ring (a, b)

2

= (0, 0)

implies a

2

= 0 and b

2

= 0, and this implies a = 0 and b = 0 since Z

2

is a

ﬁeld.

3. Find all ring homomorphisms from Z

120

into Z

42

.

Solution: Let φ : Z

120

→Z

42

be a ring homomorphism. The additive order of

φ(1) must be a divisor of gcd(120, 42) = 6, so it must belong to the subgroup

7Z

42

= ¦0, 7, 14, 21, 28, 35¦. Furthermore, φ(1) must be idempotent, and it

can be checked that in 7Z

42

, only 0, 7, 21, 28 are idempotent.

If φ(1) = 7, then the image is 7Z

42

and the kernel is 6Z

120

. If φ(1) = 21, then

the image is 21Z

42

and the kernel is 2Z

120

. If φ(1) = 28, then the image is

14Z

42

and the kernel is 3Z

120

.

4. Are Z

9

and Z

3

⊕Z

3

isomorphic as rings?

Solution: The answer is no. The argument can be given using either addition

or multiplication. Addition in the two rings is diﬀerent, since the additive

group of Z

9

is cyclic, while that of Z

3

⊕ Z

3

is not. Multiplication is also

diﬀerent, since in Z

9

there is a nonzero solution to the equation x

2

= 0, while

in Z

3

⊕ Z

3

there is not. (In Z

9

let x = 3, while in Z

3

⊕ Z

3

the equation

(a, b)

2

= (0, 0) implies a

2

= 0 and b

2

= 0, and then a = 0 and b = 0.)

5. In the group Z

×

180

of units of the ring Z

180

, what is the largest possible order

of an element?

Solution: Since 180 = 2

2

3

2

5, it follows from Theorem 3.5.4 that the ring Z

180

is isomorphic to the ring Z

4

⊕Z

9

⊕Z

5

. Then Example 5.2.10 shows that

Z

×

180

∼

= Z

×

4

Z

×

9

Z

×

5

∼

= Z

2

Z

6

Z

4

.

In the latter additive group, the order of an element is the least common

multiple of the orders of its components. It follows that the largest possible

order of an element is lcm[2, 6, 4] = 12.

6. For the element a = (0, 2) of the ring R = Z

12

⊕ Z

8

, ﬁnd Ann(a) = ¦r ∈ R [

ra = 0¦. Show that Ann(a) is an ideal of R.

Solution: We need to solve (x, y)(0, 2) = (0, 0) for (x, y) ∈ Z

12

⊕Z

8

. We only

need 2y ≡ 0 (mod 8), so the ﬁrst component x can be any element of Z

12

,

CHAPTER 5 SOLUTIONS 95

while y = 0, 4. Thus Ann((0, 2)) = Z

12

⊕ 4Z

8

. This set is certainly closed

under addition, and it is also closed under multiplication by any element of R

since 4Z

8

is an ideal of Z

8

.

7. Let R be the ring Z

2

[x]/

¸

x

4

+ 1

_

, and let I be the set of all congruence classes

in R of the form [f(x)(x

2

+ 1)].

(a) Show that I is an ideal of R.

(b) Show that R/I

∼

= Z

2

[x]/

¸

x

2

+ 1

_

.

Solution: Deﬁne φ : Z

2

[x]/

¸

x

4

+ 1

_

→Z

2

[x]/

¸

x

2

+ 1

_

by

φ(f(x) +

¸

x

4

+ 1

_

) = (f(x) +

¸

x

2

+ 1

_

). This mapping is well-deﬁned since

x

2

+ 1 is a factor of x

4

+ 1 over Z

2

. It is not diﬃcult to show that φ is an

onto ring homomorphism, with kernel equal to I.

(c) Is I a prime ideal of R?

Solution: No: (x + 1)(x + 1) ≡ 0 (mod x

2

+ 1).

Hint: If you use the fundamental homomorphism theorem, you can do the

ﬁrst two parts together.

8. Find all maximal ideals, and all prime ideals, of Z

36

= Z/36Z.

Solution: If P is a prime ideal of Z

36

, then Z

36

/P is a ﬁnite integral domain,

so it is a ﬁeld, and hence P is maximal. Thus we only need to ﬁnd the maximal

ideals of Z

36

. The lattice of ideals of Z

36

is exactly the same as the lattice of

subgroups, so the maximal ideals of Z36 correspond to the prime divisors of

36. The maximal ideals of Z

36

are thus 2Z

36

and 3Z

36

.

An alternate approach we can use Proposition 5.3.7, which shows that there

is a one-to-one correspondence between the ideals of Z/36Z and the ideals

of Z that contain 36Z. In Z every ideal is principal, so the relevant ideals

correspond to the divisors of 36. Again, the maximal ideals that contain 36Z

are 2Z and 3Z, and these correspond to 2Z

36

and 3Z

36

.

9. Give an example to show that the set of all zero divisors of a ring need not

be an ideal of the ring.

Solution: The elements (1, 0) and (0, 1) of Z Z are zero divisors, but if

the set of zero divisors were closed under addition it would include (1, 1), an

obvious contradiction.

10. Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients.

Prove that I is a prime ideal; prove that I is not maximal.

Solution: Deﬁne φ : Z[x] →Z

2

[x] by reducing coeﬃcients modulo 2. This is

an onto ring homomorphism with kernel I. Then R/I is isomorphic to Z

2

[x],

which is not a ﬁeld, so I is not maximal.

96 CHAPTER 5 SOLUTIONS

11. Let R be any commutative ring with identity 1.

(a) Show that if e is an idempotent element of R, then 1−e is also idempotent.

Solution: We have (1−e)

2

= (1−e)(1−e) = 1−e−e+e

2

= 1−e−e+e = 1−e.

(b) Show that if e is idempotent, then R

∼

= Re ⊕R(1 −e).

Solution: Note that e(1−e) = e−e

2

= e−e = 0. Deﬁne φ : R →Re⊕R(1−e)

by φ(r) = (re, r(1−e)), for all r ∈ R. Then φ is one-to-one since if φ(r) = φ(s),

then re = se and r(1 − e) = s(1 − e), and adding the two equations gives

r = s. Furthermore, φ is onto, since for any element (ae, b(1 − e)) we have

(ae, b(1 −e)) = φ(r) for r = ae + b(1 −e). Finally, it is easy to check that φ

preserves addition, and for any r, s ∈ R we have φ(rs) = (rse, rs(1 −e)) and

φ(r)φ(s) = (re, r(1 −e))(se, s(1 −e)) = (rse

2

, rs(1 −e)

2

) = (rse, rs(1 −e)).

12. Let R be the ring Z

2

[x]/

¸

x

3

+ 1

_

.

Solution: Note: Table 5.1 gives the multiplication table. It is not necessary

Table 5.1: Multiplication in Z

2

[x]/

¸

x

3

+ 1

_

1 x x

2

x

2

+x + 1 x

2

+x x + 1 x

2

+ 1

1 1 x x

2

x

2

+x + 1 x

2

+x x + 1 x

2

+ 1

x x x

2

1 x

2

+x + 1 x

2

+ 1 x

2

+x x + 1

x

2

x

2

1 x x

2

+x + 1 x + 1 x

2

+ 1 x

2

+x

x

2

+x + 1 x

2

+x + 1 x

2

+x + 1 x

2

+x + 1 x

2

+x + 1 0 0 0

x

2

+x x

2

+x x

2

+ 1 x + 1 0 x

2

+x x + 1 x

2

+ 1

x + 1 x + 1 x

2

+x x

2

+ 1 0 x + 1 x

2

+ 1 x

2

+x

x

2

+ 1 x

2

+ 1 x + 1 x

2

+x 0 x

2

+ 1 x

2

+x x + 1

to compute the multiplication table in order to solve the problem.

(a) Find all ideals of R.

Solution: By Proposition 5.3.7, the ideals of R correspond to the ideals of

Z

2

[x] that contain

¸

x

3

+ 1

_

. We have the factorization x

3

+ 1 = x

3

− 1 =

(x −1)(x

2

+x +1), so the only proper, nonzero ideals are the principal ideals

generated by [x + 1] and [x

2

+x + 1].

(b) Find the units of R.

Solution: We have [x]

3

= [1], so [x] and [x

2

] are units. On the other hand,

[x +1][x

2

+x +1] = [x

3

+1] = [0], so [x +1] and [x

2

+x +1] cannot be units.

This also excludes [x

2

+ x] = [x][x + 1] and [x

2

+ 1] = [x

2

][1 + x]. Thus the

only units are 1, [x], and [x

2

].

(c) Find the idempotent elements of R.

CHAPTER 5 SOLUTIONS 97

Solution: Using the general fact that (a+b)

2

= a

2

+2ab +b

2

= a

2

+b

2

(since

Z

2

[x] has characteristic 2) and the identities [x

3

] = [1] and [x

4

] = [x], it is

easy to see that the idempotent elements of R are [0], [1], [x

2

+ x + 1], and

[x

2

+x].

13. Let S be the ring Z

2

[x]/

¸

x

3

+x

_

.

Solution: Note: Table 5.2 gives the multiplication table. It is not necessary

Table 5.2: Multiplication in Z

2

[x]/

¸

x

3

+x

_

1 x

2

+x + 1 x

2

x x

2

+x x + 1 x

2

+ 1

1 1 x

2

+x + 1 x

2

x x

2

+x x + 1 x

2

+ 1

x

2

+x + 1 x

2

+x + 1 1 x

2

x x

2

+x x + 1 x

2

+ 1

x

2

x

2

x

2

x

2

x x

2

+x x

2

+x 0

x x x x x

2

x

2

+x x

2

+x 0

x

2

+x x

2

+x x

2

+x x

2

+x x

2

+x 0 0 0

x + 1 x + 1 x + 1 x

2

+x x

2

+x 0 x

2

+ 1 x

2

+ 1

x

2

+ 1 x

2

+ 1 x

2

+ 1 0 0 0 x

2

+ 1 x

2

+ 1

to compute the multiplication table in order to solve the problem.

(a) Find all ideals of S.

Solution: Over Z

2

we have the factorization x

3

+x = x(x

2

+1) = x(x +1)

2

,

so by Proposition 5.3.7 the proper nonzero ideals of S are the principal ideals

generated by [x], [x + 1], [x

2

+ 1] = [x + 1]

2

, and [x

2

+x] = [x][x + 1].

¸

[x

2

+x]

_

= ¦[0], [x

2

+x]¦

¸

[x

2

+ 1]

_

= ¦[0], [x

2

+ 1]¦

¸[x]) = ¦[0], [x], [x

2

], [x

2

+x]¦ ¸[x + 1]) = ¦[0], [x + 1], [x

2

+ 1], [x

2

+x]¦

(b) Find the units of R.

Solution: Since no unit can belong to a proper ideal, it follows from part (a)

that we only need to check [x

2

+x+1]. This is a unit since [x

2

+x+1]

2

= [1].

(c) Find the idempotent elements of R.

Solution: Since [x

3

] = [1], we have [x

2

]

2

= [x

2

], and then [x

2

+1]

2

= [x

2

+1].

These, together with [0] and [1], are the only idempotents.

14. Show that the rings R and S in the two previous problems are isomorphic as

abelian groups, but not as rings.

Solution: Both R and S are isomorphic to Z

2

Z

2

Z

2

, as abelian groups.

They cannot be isomorphic as rings since R has 3 units, while S has only 2.

15. Let Z[i] be the subring of the ﬁeld of complex numbers given by

Z[i] = ¦m+ni ∈ C [ m, n ∈ Z¦ .

98 CHAPTER 5 SOLUTIONS

(a) Deﬁne φ : Z[i] → Z

2

by φ(m + ni) = [m + n]

2

. Prove that φ is a ring

homomorphism. Find ker(φ) and show that it is a principal ideal of Z[i].

Solution: We have the following computations, which show that φ is a ring

homomorphism.

φ((a +bi) + (c +di)) = φ((a +c) + (b +d)i) = [a +c +b +d]

2

φ((a +bi)) +φ((c +di)) = [a +b]

2

+ [c +d]

2

= [a +b +c +d]

2

φ((a +bi)(c +di)) = φ((ac −bd) + (ad +bc)i) = [ac −bd +ad +bc]

2

φ((a +bi))φ((c +di)) = [a +b]

2

[c +d]

2

= [ac +ad +bc +bd]

2

.

We claim that ker(φ) is generated by 1 + i. It is clear that 1 + i is in the

kernel, and we note that (1 −i)(1 +i) = 2. Let m+ni ∈ ker(φ) = ¦m+ni [

m + n ≡ 0 (mod 2)¦. Then m and n are either both even or both odd, and

so it follows that m−n is always even. Therefore

m+ni = (m−n) +n +ni = (m−n) +n(1 +i)

=

_

m−n

2

_

(1 −i)(1 +i) +n(1 +i)

=

_

1

2

(m−n)(1 −i) +n

_

(1 +i) ,

and so m+ni belongs to the principal ideal generated by 1 +i.

(b) For any prime number p, deﬁne θ : Z[i] →Z

p

[x]/

¸

x

2

+ 1

_

by θ(m+ni) =

[m+nx]. Prove that θ is an onto ring homomorphism.

Solution: We have the following computations, which show that θ is a ring

homomorphism. We need to use the fact that [x

2

] = [−1] in Z

p

[x]/

¸

x

2

+ 1

_

.

θ((a +bi) + (c +di)) = θ((a +c) + (b +d)i) = [(a +c) + (b +d)x]

θ((a +bi)) +θ((c +di)) = [a +bx] + [c +dx] = [(a +c) + (b +d)x]

θ((a +bi)(c +di)) = θ((ac −bd) + (ad +bc)i) = [(ac −bd) + (ad +bc)x]

θ((a +bi))φ((c +di)) = [a +bx][c +dx] = [ac + (ad +bc)x +bdx

2

] .

Since the elements of Z

p

[x]/

¸

x

2

+ 1

_

all have the form [a + bx], for some

congruence classes a and b in Z

p

, it is clear the θ is an onto function.

16. Let I and J be ideals in the commutative ring R, and deﬁne the function

φ : R →R/I ⊕R/J by φ(r) = (r +I, r +J), for all r ∈ R.

(a) Show that φ is a ring homomorphism, with ker(φ) = I ∩ J.

CHAPTER 5 SOLUTIONS 99

Solution: The fact that φ is a ring homomorphism follows immediately from

the deﬁnitions of the operations in a direct sum and in a factor ring. Since

the zero element of R/I ⊕R/J is (0+I, 0+J), we have r ∈ ker(φ) if and only

if r ∈ I and r ∈ J, so ker(φ) = I ∩ J.

(b) Show that if I +J = R, then φ is onto, and thus R/(I ∩J)

∼

= R/I ⊕R/J.

Solution: If I +J = R, then we can write 1 = x+y, for some x ∈ I and y ∈ J.

Given any element (a + I, b + J) ∈ R/I ⊕R/J, consider r = bx + ay, noting

that a = ax+ay and b = bx+by. We have a−r = a−bx−ay = ax−bx ∈ I,

and b−r = b−bx−ay = by−ay ∈ J. Thus φ(r) = (a+I, b+J), and φ is onto.

The isomorphism follows from the fundamental homomorphism theorem.

17. Considering Z[x] to be a subring of Q[x], show that these two integral domains

have the same quotient ﬁeld.

Solution: An element of the quotient ﬁeld of Q[x] has the form

f(x)

g(x)

, for

polynomials f(x) and g(x) with rational coeﬃcients. If m is the lcm of the

denominators of the coeﬃcients of f(x) and n is the lcm of the denominators

of the coeﬃcients of g(x), then we have

f(x)

g(x)

=

n

m

h(x)

k(x)

for h(x), k(x) ∈ Z[x],

and this shows that

f(x)

g(x)

belongs to the quotient ﬁeld of Z[x].

18. Let p be an odd prime number that is not congruent to 1 modulo 4. Prove

that the ring Z

p

[x]/

¸

x

2

+ 1

_

is a ﬁeld.

Hint: Show that a root of x

2

= −1 leads to an element of order 4 in the

multiplicative group Z

×

p

.

Solution: We must show that x

2

+ 1 is irreducible over Z

p

, or, equivalently,

that x

2

+ 1 has no root in Z

p

.

Suppose that a is a root of x

2

+ 1 in Z

p

. Then a

2

≡ −1 (mod p), and so

a

4

≡ 1 (mod p). The element a cannot be a root of x

2

−1, so it does not have

order 2, and thus it must have order 4. By Lagrange’s theorem, this means

that 4 is a divisor of the order of Z

×

p

, which is p −1. Therefore p = 4q +1 for

some q ∈ Z, contradicting the assumption.

100 CHAPTER 5 SOLUTIONS

Chapter 6

Fields

SOLUTIONS TO THE REVIEW PROBLEMS

1. Let u be a root of the polynomial x

3

+3x+3. In Q(u), express (7−2u+u

2

)

−1

in the form a +bu +cu

2

.

Solution: Dividing x

3

+ 3x + 3 by x

2

− 2x + 7 gives the quotient x + 2

and remainder −11. Thus u

3

+ 3u + 3 = (u + 2)(u

2

− 2u + 7) − 11, and so

(7 −2u +u

2

)

−1

= (2 +u)/11 = (2/11) + (1/11)u.

2. (a) Show that Q(

√

2 +i) = Q(

√

2, i).

Solution: Let u =

√

2 + i. Since (

√

2 + i)(

√

2 − i) = 2 − i

2

= 3, we have

√

2 − i = 3(

√

2 + i)

−1

∈ Q(u), and it follows easily that

√

2 ∈ Q(u) and

i ∈ Q(u), so Q(

√

2, i) ⊆ Q(u). The reverse inclusion is obvious.

(b) Find the minimal polynomial of

√

2 +i over Q.

Solution: We have Q ⊆ Q(

√

2) ⊆ Q(

√

2, i). Thus [Q(

√

2) : Q] = 2 since

√

2

is a root of a polynomial of degree 2 but is not in Q. We have [Q(

√

2, i) :

Q(

√

2)] = 2 since i is a root of a polynomial of degree 2 over Q(

√

2) but is

not in Q(

√

2). Thus [Q(

√

2 +i) : Q] = 4, and so the minimal polynomial for

√

2 +i must have degree 4.

Since u =

√

2 + i, we have u − i =

√

2, u

2

− 2iu + i

2

= 2, and u

2

− 3 = 2iu.

Squaring again and combining terms gives u

4

−2u

2

+9 = 0. Thus the minimal

polynomial for

√

2 +i is x

4

−2x

2

+ 9.

3. Find the minimal polynomial of 1 +

3

√

2 over Q.

Solution: Let x = 1 +

3

√

2. Then x − 1 =

3

√

2, and so (x − 1)

3

= 2, which

yields x

3

−3x

2

+3x−1 = 2, and therefore x

3

−3x

2

+3x−3 = 0. Eisenstein’s

criterion (with p = 3) shows that x

3

− 3x

2

+ 3x − 3 is irreducible over Q, so

this is the required minimal polynomial.

101

102 CHAPTER 6 SOLUTIONS

4. Show that x

3

+ 6x

2

− 12x + 2 is irreducible over Q, and remains irreducible

over Q(

5

√

2).

Solution: Eisenstein’s criterion works with p = 2. Since x

5

− 2 is also irre-

ducible by Eisenstein’s criterion, [Q(

5

√

2) : Q] = 5. If x

3

+6x

2

−12x+2 could

be factored over Q(

5

√

2), then it would have a linear factor, and so it would

have a root in Q(

5

√

2). This root would have degree 3 over Q, and that is

impossible since 3 is not a divisor of 5.

5. Find a basis for Q(

√

5,

3

√

5) over Q.

Solution: The set ¦1,

3

√

5,

3

√

25¦ is a basis for Q(

3

√

5) over Q, and since this

extension has degree 3, the minimal polynomial x

2

− 5 of

√

5 remains irre-

ducible in the extension Q(

3

√

5). Therefore ¦1,

√

5 is a basis for Q(

√

5,

3

√

5)

over Q(

3

√

5), and so the proof of Theorem 6.2.4 shows that the required basis

is ¦1,

√

5,

3

√

5,

√

5

3

√

5,

3

√

25,

√

5

3

√

25¦.

6. Show that [Q(

√

2 +

3

√

5) : Q] = 6.

Solution: The set ¦1,

3

√

5,

3

√

25¦ is a basis for Q(

3

√

5) over Q, and since this

extension has degree 3, the minimal polynomial x

2

− 2 of

√

2 remains irre-

ducible over the extension Q(

3

√

5). Thus ¦1,

3

√

5,

3

√

25,

√

2,

√

2

3

√

5,

√

2

3

√

25¦ is

a basis for Q(

3

√

5,

√

2) over Q, and this extension contains u =

√

2 +

3

√

5. It

follows that u has degree 2, 3, or 6 over Q.

We will show that u cannot have degree ≤ 3. If

√

2 +

3

√

5 is a root of a

polynomial ax

3

+bx

2

+cx +d in Q[x], then

a(

√

2 +

3

√

5)

3

+b(

√

2 +

3

√

5)

2

+c(

√

2 +

3

√

5) +d =

a(2

√

2 + 6

3

√

5 + 3

√

2

3

√

25 + 5) +b(2 + 2

√

2

3

√

5 +

3

√

25) +c(

√

2 +

3

√

5) +d =

(5a+2b +d) 1 +(6a+c)

3

√

5 +b

3

√

25 +(2a+c)

√

2 +2b

√

2

3

√

5 +3a

√

2

3

√

25 = 0.

Since ¦1,

3

√

5,

3

√

25,

√

2,

√

2

3

√

5,

√

2

3

√

25¦ are linearly independent over Q, it

follows immediately that a = b = 0, and then c = d = 0 as well, so

√

2 +

3

√

5

cannot satisfy a nonzero polynomial of degree 1, 2, or 3 over Q. We conclude

that [Q(

√

2 +

3

√

5) : Q] = 6.

7. Find [Q(

7

√

16 + 3

7

√

8) : Q].

Solution: Let u =

7

√

16 + 3

7

√

8. Since u = (

7

√

2 + 3)(

7

√

2)

3

, it follows that

u ∈ Q(

7

√

2). Since x

7

− 2 is irreducible over Q by Eisenstein’s criterion, we

have [Q(

7

√

2) : Q] = 7, and then u must have degree 7 over Q since [Q(u) : Q]

is a divisor of [Q(

7

√

2) : Q].

8. Find the degree of

3

√

2 +i over Q. Does

4

√

2 belong to Q(

3

√

2 +i)?

Solution: Let α =

3

√

2 +i, so that α−i =

3

√

2. Then (α−i)

3

= 2, so we have

α

3

− 3iα

2

+ 3i

2

α − i

3

= 2, or α

3

− 3iα

2

− 3α + i = 2. Solving for i we get

i = (α

3

− 3α − 2)/(3α

2

− 1), and this shows that i ∈ Q(

3

√

2 + i). It follows

immediately that

3

√

2 ∈ Q(

3

√

2 +i), and so Q(

3

√

2 +i) = Q(

3

√

2, i).

CHAPTER 6 SOLUTIONS 103

Since x

3

−2 is irreducible over Q, the number

3

√

2 has degree 3 over Q. Since

x

2

+ 1 is irreducible over Q, we see that i has degree 2 over Q. Therefore

[Q(

3

√

2 + i) : Q] ≤ 6. On the other hand, [Q(

3

√

2 + i) : Q] = [Q(

3

√

2 + i) :

Q(

3

√

2)][Q(

3

√

2) : Q] and [Q(

3

√

2 + i) : Q] = [Q(

3

√

2 + i) : Q(i)][Q(i) : Q] so

[Q(

3

√

2 +i) : Q] must be divisible by 2 and 3. Therefore [Q(

3

√

2 +i) : Q] = 6.

Finally,

4

√

2 has degree 4 over Q since x

4

−2 is irreducible over Q, so it cannot

belong to an extension of degree 6 since 4 is not a divisor of 6.

104 BIBLIOGRAPHY

BIBLIOGRAPHY

Allenby, R. B. J. T., Rings, Fields, and Groups: An Introduction to Abstract Algebra

London: Edward Arnold, 1983.

Artin, M., Algebra, Englewood Cliﬀs, N.J.: Prentice-Hall, Inc., 1991

Birkhoﬀ, G., and S. Mac Lane, A Survey of Modern Algebra (4

th

ed.). New York:

Macmillan Publishing Co., Inc., 1977.

Fraleigh, J., A First Course in Abstract Algebra (6

th

ed.). Reading, Mass.: Addison-

Wesley Publishing Co., 1999.

Gallian, J., Contemporary Abstract Algebra (4

th

ed.). Boston: Houghton Miﬄin

Co., 1998

Herstein, I. N., Abstract Algebra. (3

rd

ed.). New York: John Wiley & Sons, Inc.,

1996.

———, Topics in Algebra (2

nd

ed.). New York: John Wiley & Sons, Inc., 1975.

Hillman, A. P., and G. L. Alexanderson, Abstract Algebra: A First Undergraduate

Course. Prospect Heights: Waveland Press, 1999.

Maxﬁeld, J. E., and M. W. Maxﬁeld, Abstract Algebra and Solution by Radicals.

New York: Dover Publications, Inc., 1992.

Saracino, D., Abstract Algebra: A First Course. Prospect Heights: Waveland Press,

1992.

Van der Waerden, B. L., A History of Algebra: from al-Khwarizmi to Emmy

Noether. New York: Springer-Verlag, 1985.

INDEX 105

Index

abelian group, 13

algorithm, division, 1

algorithm, Euclidean, 35

alternating group, 22, 77

annihilator, 30, 94

associative law, 13, 15, 59, 62

basis, for an extension ﬁeld, 33, 102

binary operation, 13

cancellation law, 14

Cayley’s theorem, 21

centralizer, 17, 18, 22, 65, 67, 68, 77

Chinese remainder theorem, 76

closure, 15, 57, 62, 64, 65

combination, linear, 2

complex numbers, 30, 97

composite function, 7

congruence, linear, 5

congruence, 3–5, 41

conjugacy class, 25, 82, 83

coset, 24–26, 81, 82, 86

criterion, of Eisenstein, 101, 102

cross product, 14, 57

cycle, 54

cyclic, 1, 5, 6, 20, 21, 26, 45, 69, 76,

84, 86

cyclic group, 17, 23, 24, 65, 81

cyclic subgroup, 16, 63

determinant, 8, 51, 65

digit, units, 5, 43

dihedral group, 22, 24–26, 77, 78,

82, 83, 86

direct product, 17

disjoint cycles, 10, 53, 54

division algorithm, 1

division, 14

dot product, 9, 14, 52, 57

eigenvalue, 8, 51

Eisenstein’s criterion, 88, 101, 102

element, idempotent, 5, 6, 29, 30, 44, 45,

93, 94, 96, 97

element, nilpotent, 6, 29, 44–46, 56, 93,

94

equivalence relation, 8, 9, 51–53

Euclidean algorithm, for polynomials, 27,

87, 90

Euclidean algorithm, matrix form, 36, 46

Euclidean algorithm, 3, 35, 39, 41, 44, 46

Euler phi-function, 74

even permutation, 53

factor group, 24, 26, 81, 82, 86

ﬁeld, 27, 30, 31, 94, 95, 97, 99

ﬁeld, ﬁnite, 28, 89–91

ﬁeld, of quotients, 31, 99

ﬁeld, of rational numbers, 27, 28, 87–89

ﬁnite ﬁeld, 28, 89–91

ﬁnite group, 13

fractional linear transformation, 15, 60

function, composite, 7

function, inverse, 7, 8, 50

function, one-to-one, 7

function, onto, 7

fundamental homomorphism theorem, for

groups, 22–24, 80

fundamental homomorphism theorem, for

rings, 30, 95, 99

Gaussian integers, 30, 98

gcd, of integers 2, 3, 6, 35, 36, 40, 46

gcd, of polynomials, 27, 87, 88

general linear group, 14–18, 20, 26,

64–68, 72, 73, 84, 85

generator, 21, 65, 69, 74

group, 1, 13

group, abelian, 13

group, alternating, 22, 77

group, cyclic, 23

106 INDEX

group, dihedral, 22, 24–26, 76, 77, 81,

82, 83, 86

group, ﬁnite, 13

group, of permutations, 10, 11, 54

group, symmetric, 21

group homomorphism, 18, 22, 23, 78–80

group isomorphism, 23

homomorphism, of groups, 22, 23, 78–80

homomorphism, of rings, 29, 31, 94, 95,

98

horizontal line test, 8, 49

ideal, 30, 31, 97–99, 101

ideal, maximal, 30, 95

ideal, prime, 30, 95

ideal, principal, 31, 95–98

idempotent element, 5, 6, 29, 30, 44, 45,

93, 94, 96, 97

idempotent element, modulo n, 5

identity element, 13, 15, 57, 58, 60,

62, 65

image, of a ring homomorphism, 94

image, 23, 80

induction, 2, 38, 59

integers mod n, 5, 14

inverse element, 7, 8, 15, 50, 60, 62, 64

inverse, multiplicative, 5, 6, 28, 41, 44,

45, 47, 90

invertible matrix, 9, 53

irreducible polynomial, 28, 33, 88, 89,

101, 102

isomorphic rings, 29, 30, 94, 95, 97, 99

isomorphism, of groups, 19, 20, 22,

69–73

isomorphism, of rings, 29, 30, 94, 95,

97, 99

kernel, of a group homomorphism, 23, 80

kernel, of a ring homomorphism, 94, 95

Lagrange’s theorem, 16, 77, 99

lattice diagram, of subgroups, 21, 26,

74, 84

lattice diagram, 3, 39

linear combination, 2, 35

linear congruence, 5

linear transformation, fractional, 15, 60

linear transformation, 8, 50, 51

linearly independent vectors, 51

matrix, invertible, 9, 53

matrix, 8, 14, 50, 51

maximal ideal, 30, 95

minimal polynomial, 33, 101

multiplicative inverse, 5, 6, 28, 41, 44,

45, 60, 90

multiplicative order, modulo n, 5

multiplicative order, 6, 47

nilpotent element, 6, 29, 44–46, 56, 93,

94

nilpotent element, modulo n, 5

nilpotent element, of a ring, 29, 93

normal subgroup, 25, 26, 82, 86

nullity, 50

one-to-one function, 7, 8, 51

onto function, 7, 8, 51

order, 16–18, 21, 66, 74

order, multiplicative, 5, 6, 47

order, of a permutation, 10, 11, 54

parallel plane, 53

partition, 9

permutation, 10

permutation, even, 53

permutation group, 10, 11, 54

perpendicular plane, 53

plane, parallel, 53

plane, perpendicular, 53

polynomial, irreducible, 33, 101–103

polynomial, minimal, 33, 101

prime ideal, 30, 95

prime, relatively, 2

principal ideal, 31, 95–98

INDEX 107

quaternion group, 21, 75

quotient ﬁeld, 31, 99

rank, of a matrix, 50

rank nullity theorem, 50, 51

rational roots, 88

reﬂexive law, 52

relatively prime polynomials, 27, 87

relatively prime, 2

ring homomorphism, 29, 31, 94, 95, 98

root, of a polynomial, 33, 101, 102

root, rational, 88

subgroup, normal, 25, 26, 82, 86

subgroup, 15, 17, 65, 66

subring, 30, 31, 97, 99

subspace, 15

symmetric group, 14, 16, 21, 22, 63, 77

symmetric law, 52

system of congruences, 4, 6, 42, 47

theorem, of Lagrange, 99

transformation, linear, 8, 50

transitive law, 52

unit, of a ring, 29, 30, 93, 94, 96, 97

units, mod 7; 23, 79

units, mod 9; 14, 21, 58, 74

units, mod 13; 26, 84

units, mod 15; 15, 21, 58, 75

units, mod 17; 19, 23, 69, 79

units, mod 18; 21, 74

units, mod 20; 16, 26, 63, 85

units, mod 21; 16, 21, 63, 75

units, mod 24; 16, 62

units, mod 36; 17, 66

units, mod n, 5, 6, 14, 45

units, mod p, 16, 23, 64, 80

units digit, 5, 43

vector space, 15, 17

vertical line test, 8, 49

well-deﬁned function, 24

ii This is a supplement to

Abstract Algebra, Second Edition by John A. Beachy and William D. Blair ISBN 0–88133–866–4, Copyright 1996 Waveland Press, Inc. P.O. Box 400 Prospect Heights, Illinois 60070 847 / 634-0081 www.waveland.com

c John A. Beachy 2000 Permission is granted to copy this document in electronic form, or to print it for personal use, under these conditions: it must be reproduced in whole; it must not be modiﬁed in any way; it must not be used as part of another publication.

Formatted February 8, 2002, at which time the original was available at: http://www.math.niu.edu/∼ beachy/abstract algebra/

Contents

PREFACE 1 INTEGERS 1.1 Divisors . . . . . . 1.2 Primes . . . . . . . 1.3 Congruences . . . . 1.4 Integers Modulo n Review problems . . . . v 1 1 2 3 5 6 7 7 8 10 12 13 13 15 17 18 20 21 22 24 26 27 27 29 29

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

2 FUNCTIONS 2.1 Functions . . . . . . . 2.2 Equivalence Relations 2.3 Permutations . . . . . Review problems . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

3 GROUPS 3.1 Deﬁnition of a Group . . . . . . 3.2 Subgroups . . . . . . . . . . . . 3.3 Constructing Examples . . . . 3.4 Isomorphisms . . . . . . . . . . 3.5 Cyclic Groups . . . . . . . . . . 3.6 Permutation Groups . . . . . . 3.7 Homomorphisms . . . . . . . . 3.8 Cosets, Normal Subgroups, and Review problems . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . Factor . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Groups . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

4 POLYNOMIALS Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 COMMUTATIVE RINGS Review problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

. . . .iv CONTENTS 33 33 33 35 49 57 87 93 101 104 105 6 FIELDS Review problems . . . . . . . . . . . . . . . . . . . . . SOLUTIONS 1 Integers 2 Functions 3 Groups 4 Polynomials 5 Commutative Rings 6 Fields BIBLIOGRAPHY INDEX . . . . . . . .

I’ve tried to choose problems that would be instructive. especially since they are probably still learning how to write their own proofs. DeKalb. Instead of just expanding the material that is already written down in our textbook. using Herstein’s Topics in Algebra. Of course. with applications to computing and other areas. by writing out solutions to problems. but Topics contains the core of any course. this study guide isn’t a substitute for a good teacher. or for the chance to work together with other students on some hard problems. so students meeting abstract algebra still struggle to learn the new concepts. the subject may have become broader. This “study guide” is intended to help students who are beginning to learn about abstract algebra.” I was given leave in Spring 2000 to work on projects related to teaching. and in quite a few cases I’ve included comments to help the reader see what is really going on. Beachy . I would like to gratefully acknowledge the support of Northern Illinois University while writing this study guide. Unfortunately. Finally. I decided to try to teach by example. It’s hard to improve on his book. As part of the recognition as a “Presidential Teaching Professor. the subject hasn’t become any easier.PREFACE v PREFACE I ﬁrst taught an abstract algebra course in 1968. Illinois October 2000 John A.

vi PREFACE .

In fact. If you expand the expression b|a by writing “a = bq for some q ∈ Z”. —When working on questions involving divisibility you may ﬁnd it useful to go back to Deﬁnition 1. This equation involves ordinary integers. where 0 ≤ r < b. In many cases. For example.1. Many of the concepts introduced there can be abstracted to much more general situations.3) to write a = bq + r.Chapter 1 INTEGERS Chapter 1 of the text introduces the basic ideas from number theory that are a prerequisite to studying abstract algebra. in Chapter 3 of the text you will be introduced to the concept of a group. The examples in Section 1. Here are several useful approaches you should be able to use. the proofs of the theorems contain important techniques that you need to copy in solving the exercises in the text. One of the ﬁrst broad classes of groups that you will meet depends on the deﬁnition of a cyclic group. you need to make sure that you are familiar with all of the deﬁnitions and theorems in the section.4. —To show that b|a. then you have an equation to work with. simplify. actually tell you everything you will need to know about cyclic groups. try to write down an expression for a and expand. one that is obtained by considering all powers of a particular element.1.1. constructed using congruence classes of integers. although Chapter 1 is very concrete. 1. it is a signiﬁcant step forward into the realm of abstract algebra. Then to prove that b|a you only 1 . or substitute for terms in the expression until you can show how to factor out b. and so you can use all of the things you already know (from high school algebra) about working with equations. —Another approach to proving that b|a is to use the division algorithm (see Theorem 1.1 Divisors Before working through the solved problems for this section.

1. CHAPTER 1. n are integers? 25. An integer is a linear combination of a and b if and only if it is a multiple of their greatest common divisor.2 states that integers a and b are relatively prime if and only if there exist integers m and n with ma + nb = 1. Find gcd(3553. Which of the integers 0. . 28. which can be expressed as the smallest positive linear combination of a and b. n2 + n + 1) = 1 or n 1 0 −1 1 0 = 0 0 0 0 0 1 0 0 0 1 1. SOLVED PROBLEMS: §1. 30. 1002. gcd(n − 1. then 0 1 if and only if 4|n. . 527).6). 26.6 states that any two nonzero integers a and b have a greatest common divisor. Give a proof by induction to show that each number in the sequence 12. if you have a linear combination ma + nb = d. This is really useful in working on questions involving greatest common divisors. n2 + n + 1) = 3. b) is a divisor of d (refer back to Theorem 1.1. 27. n + 10). 24. then gcd(n − 1.2 need to ﬁnd some way to check that r = 0. . 10 can be expressed in the form 12m + 20n. b) = 1. Remember that this only works in showing that gcd(a. If n is a positive integer. This is one of the most useful tools in working with relatively prime integers. 377). is divisible by 6. Prove that if a and b are nonzero integers for which a|b and b|a. . Since the fundamental theorem of arithmetic (on prime factorization) is proved in this section. 0 29. . 10002. you now have some more familiar techniques to use. .. where m. Find gcd(435. More generally. Prove that if n is an integer with n > 1. 23. 102.2. INTEGERS —Theorem 1. and express it as a linear combination of 435 and 377. then m2 − n2 is divisible by 8. and express it as a linear combination of 3553 and 527.2 Primes Proposition 1. then b = ±a. . ﬁnd the possible values of gcd(n.1. . it only shows that gcd(a. Prove that if n is a positive integer.1 22. Prove that if m and n are odd integers.

n) = 1 and ac ≡ ad (mod n). 3 27.3. . Prove that gcd(2m − 1. 2n2 + 6n − 4) = 1. any equation involving integers can be converted into a congruence by just reducing modulo n.3 Congruences In this section. 1492). (a) Use the Euclidean algorithm to ﬁnd gcd(1274. In a congruence modulo n.1. CONGRUENCES SOLVED PROBLEMS: §1. Let m and n be positive integers. (b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776. 26. and you can use the fact that if a ≡ b (mod n) and b ≡ c (mod n). for all integers n > 1. If two integers are congruent modulo n. 1492). b) = 1 if and only if gcd(a2 . This is usually expressed by saying that if gcd(a. In other words. then are certainly congruent modulo n. Do the same for 484. but only guarantees that dividing by n produces the same remainder in each case. 24. 2n − 1) = 1 if and only if gcd(m. you can only divide through by an integer that is relatively prime to n. (a) Use the Euclidean algorithm to ﬁnd gcd(1776. 1089). the integers may diﬀer by some multiple of n. 30. First. Give the lattice diagram of all divisors of 250. Find all integer solutions of the equation xy + 2y − 3x = 25. it is not exactly the same. What things are the same? You can add or subtract the same integer on both sides of a congruence. (Review Proposition 1. Prove that gcd(2n2 + 4n − 3.3. Just be very careful! One of the important techniques to understand is how to switch between congruences and ordinary equations. 25.3. n) = 1. that doesn’t make them equal. For positive integers a. You can use substitution. The do the opposite conversion you must be more careful. 28. and the comments in the text both before and after the proof of the proposition. This works because if two integers are equal. (b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274. then c ≡ d (mod n).) What things are diﬀerent? In an ordinary equation you can divide through by a nonzero number. prove that gcd(a. for all integers n > 1. Prove that n − 1 and 2n − 1 are relatively prime. then a ≡ c (mod n). b2 ) = 1. and you can multiply both sides of a congruence by the same integer. it is important to remember that although working with congruences is almost like working with equations. b. 1089). Is the same true for 2n − 1 and 3n − 1? 29.2 23. 1.

then so is x = m + 15q and y = n − 22q. 3x ≡ 2 (mod 13) . because we have to introduce another variable.5 and Theorem 1. where the moduli n and m are relatively prime. SOLVED PROBLEMS: §1. where the congruence x ≡ 7 (mod 8) is converted into the equation x = 7 + 8q . because you might miss important algorithms or computational techniques. b. and all simultaneous linear equations of the form x ≡ a (mod n) and x ≡ b (mod m). x ≡ 23 (mod 32) . (b) Show that if x = m and y = n is an integer solution to the equation in part (a). (a) Find all solutions to the congruence (b) Find all solutions to the congruence 28. Solve the congruence 42x ≡ 12 (mod 90). 29.5 of the text.3.4 CHAPTER 1. 55x ≡ 35 (mod 75). for any integer q. Solve the system of congruences 30. x ≡ b (mod n) . we can reduce each term modulo 5. Give integers a. and you can’t aﬀord to skip over their proofs. we really want a congruence modulo 5. Notice that converting to an equation makes it more complicated. 27. Solve the system of congruences 31. These proofs actually show you the necessary techniques to solve all linear congruences of the form ax ≡ b (mod n).3. Many of the theorems in the text should be thought of as “shortcuts”. You should read the proofs of Theorem 1. Solve the system of congruences x ≡ 2 (mod 9) x ≡ 4 (mod 10) . so that we ﬁnally get x ≡ 2 + 3q (mod 5) . (a) Find one particular integer solution to the equation 110x + 75y = 45. n to provide an example of a system x ≡ a (mod m) that has no solution. INTEGERS The conversion process is illustrated in Example 1.6 very carefully. so the next step is to rewrite the equation as x ≡ 7 + 8q (mod 5) . 5x ≡ 14 (mod 17) x ≡ 5 (mod 25) 32. m. In the example.3 26. 55x ≡ 36 (mod 75). for some q ∈ Z . Actually.3.

For example. and it is worthwhile to learn all of the approaches. any linear congruence of the form ax ≡ b (mod n) can be viewed as an equation in Zn . this is n equivalent to saying that Z× is cyclic if has an element [a] such that each element n of Z× is equal to some power of [a]. Instead of thinking in terms of division. or 6 as its units digit. 35. provided we think in terms of equivalence classes.4 of the text contain several deﬁnitions for elements of Zn .1. 5 36. for all positive integers n. the element [a] ∈ Zn is said to be n idempotent if [a]2 = [a]. (a) Compute the last digit in the decimal expansion of 4100 . INTEGERS MODULO N 33. 1. provided [a]−1 exists. This gives you one more way to view problems involving congruences. If (a. it is probably better to think of multiplying both sides of the equation [a]n [x]n = [b]n by [a]−1 . and choose whichever approach is the most convenient. n) = 1. and nilpotent if [a]k = [0] for some k. Finally. To be more precise. Since |Z× | = ϕ(n). 31. so that you can easily shift back and forth between them. 1. Sometimes it helps to have various ways to think about a problem. They will provide n an important source of examples in Chapter 3. Find all integers n for which 13 | 4(n2 + 1).4 30.4 Integers Modulo n The ideas in this section allow us to work with equations instead of congruences. (b) Is 4100 divisible by 3? 34. written [a]n [x]n = [b]n . Find the multiplicative inverse of each nonzero element of Z13 . n) = 1. Find the multiplicative inverse of each nonzero element of Z7 . Prove that 10n+1 + 4 · 10n + 4 is divisible by 9.4. The exercises for Section 1. when we begin studying groups. then the smallest positive integer k such that ak ≡ 1 (mod n) is called the multiplicative order of [a] in Z× . . trying to divide by a in the congruence ax ≡ b (mod n) can get you into trouble unless gcd(a. Prove that the fourth power of an integer can only have 0. 5. n n It is well worth your time to learn about the sets Zn and Z× . The set Z× is said to be cyclic if n n it contains an element of multiplicative order ϕ(n). SOLVED PROBLEMS: §1.

ﬁnd all idempotent elements. Prove that if k is the smallest positive integer for which ak ≡ 1 (mod n). For ω = − + i. 35. 17 37. Let n be a positive integer. Solve the congruence 24x ≡ 168 (mod 200). 2 2 3. INTEGERS 34. ﬁnd all nilpotent elements. Find gcd(7605. Prove that [a]n is a nilpotent element of Zn if and only if each prime divisor of n is a divisor of a. then k | ϕ(n). 11 39. and ﬁnd its multiplicative order. and express it as a linear combination of 7605 and 5733. n) = 1. Show that Z× is cyclic. 5733). Solve the equation [x]2 + [x]11 − [6]11 = [0]11 . 35 35 35 for some positive integers i. for any integer n. ﬁnd its multiplicative inverse. if possible (in Z× ). 36. prove that ω n = 1 if and only if 3|n. if possible (in Z× ). Review Problems 1. In Z24 : ﬁnd all units (list the multiplicative inverse of each). 38. 501 501 33. Show that if n > 1 is an odd integer. 2x ≡ 9 (mod 15) x ≡ 8 (mod 11) . Find [3379]−1 . and let a ∈ Z with gcd(a. 4. 4061 4061 CHAPTER 1. . 6. 40. Solve the system of congruences Z× . ﬁnd all idempotent elements.6 32. ﬁnd all nilpotent elements. List the elements of For each element. √ 1 3 2. 15 5. then ϕ(2n) = ϕ(n). In Z20 : ﬁnd all units (list the multiplicative inverse of each). Find [91]−1 . j. Show that Z× is not cyclic but that each element has the form [8]i [−4]j .

Deﬁnition 2. One of the most fundamental ideas of abstract algebra is that algebraic structures should be thought of as essentially the same if the only diﬀerence between them is the way elements have been named.Chapter 2 FUNCTIONS The ﬁrst goal of this chapter is to provide a review of functions. which give a very important class of examples. for all y ∈ R+ . as well as on your knowledge of the groups Zn and Z× . is stated rather formally in terms of ordered pairs. it might help to get out your calculus textbook and review composite functions. the text almost immediately goes back to what might be a more familiar deﬁnition: a function f : S → T is a “rule” that assigns to each element of S a unique element of T . the functions that are one-to-one correspondences will be particularly important.1. 7 .1.1 Functions Besides reading Section 2. and inverse functions. one-to-one and onto functions. The functions f : R → R+ and g : R+ → R deﬁned by f (x) = ex . n 2. When you begin to study groups in Chapter 3. In this setting. The second goal of the chapter is to begin studying groups of permutations. provide one of the most important examples of a pair of inverse functions. for all x ∈ R. you will be able draw on your knowledge of permutation groups. (Think of this as a deﬁnition given in terms of the “graph” of the function.1. functions will provide a way to compare two diﬀerent structures.) In terms of actually using this deﬁnition. and g(y) = ln y. To make this precise we will say that structures are the same if we can set up an invertible function from one to the other that preserves the essential algebraic structure. as introduced in this section. the deﬁnition of function. In our study of algebraic structures in later chapters. That makes it especially important to understand the concept of an inverse function.

1. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax. First there is the deﬁnition . FUNCTIONS SOLVED PROBLEMS: §2. 24. for all x ∈ Rn . The “Horizontal Line Test” from calculus says that a function is one-to-one if and only if no horizontal line intersects its graph more than once.8 CHAPTER 2.1. Let A be an n × n matrix with entries in R. This led to the concept of congruence modulo n. The “Vertical Line Test” from calculus says that a curve in the xy-plane is the graph of a function of x if and only if no vertical line intersects the curve more than once. Let a be a ﬁxed element of Z× . which is a model for our general notion of an equivalence relation. more than one 22. for all x ∈ Rn . where AT is the transpose of A. In calculus the graph of an inverse function f −1 is obtained by reﬂecting the graph of f about the line y = x. for all x ∈ Rn . then it is invertible. looking at the one idea of splitting up a set S from three distinct vantage points.7. for all x ∈ Z× . and assume that m > n. Deﬁne the function θ : Z× → Z× by 17 17 17 θ(x) = ax. Show that L is a one-to-one function if det(AT A) = 0.2 Equivalence Relations In a variety of situations it is useful to split a set up into subsets in which the elements have some property in common. You are already familiar with one of the important examples: in Chapter 1 we split the set of integers up into subsets. Deﬁne a linear transformation L : Rn → Rm by L(x) = Ax. (b) Show that if L is either one-to-one or onto. 2. Let A be an m × n matrix with entries in R. (a) Show that L is an invertible function if and only if det(A) = 0. Note: A vector x is called an eigenvector of A if it is nonzero and there exists a scalar λ such a that Ax = λx. 25. Explain why this agrees with Deﬁnition 2. 23. Prove that L is one-to-one if and only if no eigenvalue of A is zero. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax.1 20. 21. depending on the remainder when the integer is divided by the ﬁxed integer n. Explain why this agrees with Deﬁnition 2.1.1. Let A be an n × n matrix with entries in R.4. In this section you will ﬁnd three diﬀerent points of view. Explain why this agrees with Deﬁnition 2. 26. Is θ one to one? Is θ onto? If possible. ﬁnd the 17 inverse function θ−1 .

12 Show that the formula f deﬁnes a function. x + y + z) . which places the emphasis on describing the subsets. give a geometric description of the partition of R3 that is determined by L. 16. Let u be a ﬁxed vector in R3 . The reason for considering several diﬀerent point of view is that in a given situation one point of view may be more useful than another. Find the image of f and the set Z12 /f of equivalence classes determined by f . y1 ) ∼ (x2 . so that you can easily switch from one to the other. for all (x. y. x + y + z. where we say that x1 and x2 are equivalent if f (x1 ) = f (x2 ). On the set C of complex numbers. y. SOLVED PROBLEMS: §2. deﬁne z1 ∼ z2 if ||z1 || = ||z2 ||. . describe the equivalence relation on R that is determined by f . Check that ∼ deﬁnes an equivalence relation. for all x ∈ R.2. Show that this deﬁnes an equivalence relation.2 14. and give a geometric description of the equivalence classes of ∼. z) = (x + y + z.2. On the set of all n × n matrices over R. for all [x]12 ∈ Z12 . deﬁne A ∼ B if there exists an invertible matrix P such that P AP −1 = B. where · denotes the standard dot product. b)} of all ordered pairs of positive integers. 15. which tells you when two diﬀerent elements of S belong to the same subset. Deﬁne the formula f : Z12 → Z12 by f ([x]12 ) = [x]2 . Your goal should be to learn about each point of view. which is a big help in deciding which point of view to take. EQUIVALENCE RELATIONS 9 of an equivalence relation on S. it turns out that every partition (and equivalence relation) really comes from a function f : S → T . z) ∈ R3 . Finally. Show that ∼ is an equivalence relation. deﬁne (x1 . 19. deﬁne v ∼ w if v ·u = w ·u. Then there is the notion of a partition of S. 17. For the linear transformation L : R3 → R3 deﬁned by L(x. For vectors v and w. For the function f : R → R deﬁned by f (x) = x2 . 20. y2 ) if x1 y2 = x2 y1 . On the set {(a. Show that ∼ is an equivalence relation. 18. and assume that u has length 1.

compute the order of στ σ −1 . For the permutation σ = 1 2 3 4 5 6 7 8 9 . What is the order of σ? Compute σ −1 . . then σ −1 ∈ G. then τ σ ∈ G. 8. Write σ as a product of disjoint cycles.1 of the text. τ −1 . Let σ = (2. we will say that G is a group of permutations if the following conditions hold. then σ −1 ∈ Sym(S). 7). τ ∈ G. If G is a nonempty subset of Sym(S). 1 2 2 5 3 1 4 8 5 3 6 6 7 4 8 7 9 9 and 14. τ στ −1 . 2. τ ∈ Sym(S). we need the following deﬁnition. instead of the way we have deﬁned multiplication. What is the order of σ? Is σ an even permutation? Compute σ −1 .10 CHAPTER 2. (i) If σ. write σ as a 7 5 6 9 2 4 8 1 3 product of disjoint cycles. σ −1 . 7. then τ σ ∈ Sym(S). For the permutations σ = τ= 1 2 3 4 5 6 7 8 9 . for any permutation σ ∈ Sn . Our point of view is that permutations are functions. (ii) 1S ∈ Sym(S). τ σ. FUNCTIONS 2. We will see later that this agrees with Deﬁnition 3. (iii) if σ ∈ Sym(S). For 17. and we write functions on the left. στ . (ii) 1S ∈ G. 9)(1. )(6. (iii) if σ ∈ G. In two of the problems. just as in calculus. then we have the following properties. you need to be aware that some authors multiply permutations by reading from left to right. SOLVED PROBLEMS: §2. τ . If you are reading another book along with Abstract Algebra. 6)(3. 16. 4. You need to do enough computations so that you will feel comfortable in dealing with permutations. In the text we noted that if S is any set. 9. Compute the order of τ = 1 2 3 4 5 7 2 11 4 6 σ = (3. 6 7 8 9 8 10 9 10 1 3 11 5 . στ σ −1 .6. 8. write each of these permutations as a 1 5 4 7 2 6 8 9 3 product of disjoint cycles: σ. and Sym(S) is the set of all permutations on S. 6) ∈ S9 . 4. so we have to do the computations from right to left. 5.3 13.3 Permutations This section introduces and studies the last major example that we need before we begin studying groups in Chapter 3. (i) If σ. Prove that if τ ∈ Sn is a permutation with order m. then στ σ −1 has order m. 15.

with G ⊆ Sym(S). 20. Let S be a set. Let τ be a ﬁxed permutation in Sym(S). Let G = {σ ∈ Sym(S) | σ(X) ⊂ X}. Let G be a group of permutations. . 11 19.3. PERMUTATIONS 18. but not 11 or 13.2. 12. Prove that τ Gτ −1 = {σ ∈ Sym(S) | σ = τ γτ for some γ ∈ G} is a group of permutations. Prove that G is a group of permutations. and let X be a subset of S. for the set S. and 14. Show that S10 has elements of order 10.

4. 9). β = (1. 2. Show that f is a one-to-one function. Deﬁne the function φ : Z× → Z× by φ(x) = x−1 . On the set Q of rational numbers. 6. for all x ∈ R. 6). FUNCTIONS Review Problems 1. let α = (1. Hint: Use the derivative of f to show that f is a strictly increasing function. 3. deﬁne x ∼ y if x − y is an integer. 3. For the function f : R → R deﬁned by f (x) = x2 . is a one-to-one and onto function. ﬁnd the inverse function φ−1 . Compute φα . (b) In S3 . Show that ∼ is an equivalence relation. 7. and γ = (1. let α = (1. describe the equivalence relation on R that is determined by f . and use this to ﬁnd its order and its inverse. Is φ one to 17 17 17 one? Is φ onto? If possible. 3). Deﬁne f : R → R by f (x) = x3 + 3xz − 5. write σ as a product of disjoint cycles. for all x ∈ R. for all σ ∈ Sn . 5. . 2. 5. 2. for all x ∈ Z× . Show that φα : Sn → Sn deﬁned by φα (σ) = ασα−1 . 2). In S10 . For σ = αβγ. (a) Let α be a ﬁxed element of Sn . Is σ even or odd? 5.12 CHAPTER 2.

and the equation will still hold. multiplication. to show that ∗ might represent ordinary addition or multiplication. has an identity element. “for each”. group. These deﬁnitions provide the language you will be working with. For example. and x = y. 3.1. The precise statement is given in Deﬁnition 3. one of the rules you are used to says that you can multiply both sides of an equation by the same value.3. One of the things we try to do with notation is to make it look familiar. which we begin in this chapter. Try to learn it so well that you don’t have even a trace of an accent! Loosely. At ﬁrst we will use ∗ or · to represent an operation. for 13 . or maybe even something quite far from your experience. But the operations are still the usual ones for numbers. so long as everyone knows the convention that we are using. which just represent various numbers. and you simply must know this language. abelian group. which must be stated in exactly the right order. addition. the axioms for a group give us just what we need to work with equations involving the operation in the group. “there exists”). or possibly operations on matrices or functions. subtraction. is usually thought of as the real beginning of abstract algebra. then a · x = a · y. and division. and has inverses for each of its elements. especially the quantiﬁers (“for all”. The step from arithmetic to algebra involves starting to use variables. you must pay careful attention to each part. even if it represents something new. The step from algebra to abstract algebra involves letting the operation act like a variable.Chapter 3 GROUPS The study of groups.1 Deﬁnition of a Group This section contains these deﬁnitions: binary operation. This still works for the operation in a group. and ﬁnite group. very soon we will just write ab instead of a ∗ b. a group is a set on which it is possible to deﬁne a binary operation that is associative. From one point of view. since if x and y are elements of a group G.

you can be pretty safe in doing the familiar things to an equation that involves elements of a group. but not multiplication. The next level comes from knowing some good examples. x1 y2 − y1 x2 ). I should still mention that the original motivation for studying groups came from studying sets of permutations. the cross product is deﬁned by (x1 . Does this make R3 into a group? 23. R× . y2 . Next. while Z× is a group under multiplication. you must write a · b−1 or b−1 · a. z1 )×(x2 . for all a.1. and C× of nonzero numbers form groups under multiplication. since the operation in the group may not satisfy the commutative law. the most important examples come from groups of matrices. On the set G = Q× of nonzero rational numbers. b ∈ G. but we can’t guarantee that a · x = y · a. Similarly. There shouldn’t be any confusion in just listing these as groups.6 for the precise statement). z1 ) and (x2 . the sets Q× . We could also guarantee that x · a = y · a. SOLVED PROBLEMS: §3. This is a part of the guarantee that comes with the deﬁnition of a binary operation. z2 ) in R3 . It is important to note that on both sides of the equation. Is R3 a group under this multiplication? 24. and so the symmetric group Sn still has an important role to play. The third level of understanding comes from using the deﬁnitions to prove various facts about groups. Here are a few of the important examples. the sets of numbers Z. so whenever you are tempted to write a ÷ b or a/b. a is multiplied on the left. Q. If you are careful about the side on which you multiply. while the set GLn (R) of all invertible n × n matrices with entries in R is a group under multiplication. deﬁne a new multiplication ab by a∗b = . without mentioning their operations. It is common to just list these n sets as groups. Use the dot product to deﬁne a multiplication on R3 . GROUPS any element a in G. R. and don’t fall victim to the temptation to divide. Show that G is a group under this multiplication.14 CHAPTER 3. without speciﬁcally mentioning which operation is used. y1 . In the study of ﬁnite groups. the set Mn (R) of all n × n matrices with entries in R is a group under addition. First. since in each case only one of the two familiar operations can be used to make the set into a group. Understanding and remembering the deﬁnitions will give you one level of understanding. 9 . y2 . Write out the multiplication table for Z× . 2 25. and C form groups under addition. y1 . The sets Z and Zn are groups under addition. Remember that in a group there is no mention of division. z2 ) = (y1 z2 − z1 y2 . z1 x2 − x1 z2 .1 22. For vectors (x1 . The existence of inverses allows cancellation (see Proposition 3. but not under addition.

Show that F forms a group under composition of functions. 29. If the subset is a group in its own right. using the same operation as the larger set. and any subspace is automatically a subgroup. 28. it forms an abelian group. That is. Let G be a group.3 of the text. any group of permutations is a subgroup of Sym(S). az + b . Let G be a group. c. d are integers with ad − bc = 1. 31. Let F be the set of all fractional linear transformations of the complex plane. 3. For x.3. SUBGROUPS 26. For instance. b.2. for any positive integer n. 15 15 27. and suppose that a and b are any elements of G. then it is called a subgroup. and suppose that a and b are any elements of G. replace condition (iii) with the condition that there exists e ∈ G such that e · a = a for all a ∈ G. (c) Show that 2 is the identity element for the operation ∗. If you only look at the operation of addition in a vector space. Show that (aba−1 )n = abn a−1 . and replace condition (iv) with the condition that for each a ∈ G there exists a ∈ G with a · a = e. Prove that these weaker conditions (given only on the left) still imply that G is a group. then ba = ab. Give an example to show that it is not suﬃcient to require the existence of a left identity element together with the existence of right inverses. In Deﬁnition 3. (a) Show that the operation ∗ is closed on G. 32.2 Subgroups Many times a group is deﬁned by looking at a subset of a known group. The previous exercise shows that in the deﬁnition of a group it is suﬃcient to require the existence of a left identity element and the existence of left inverses. F is the set of all functions f (z) : C → C of the form f (z) = cz + d where the coeﬃcients a. (b) Show that the associative law holds for ∗. you are right.1. 30. (d) Show that for element a ∈ G there exists an inverse a−1 ∈ G. Let G = {x ∈ R | x > 1} be the set of all real numbers greater than 1. If the idea of a subgroup reminds you of studying subspaces in your linear algebra course. Any group of n × n matrices (with entries in R) is a subgroup of GLn (R). deﬁne x ∗ y = xy − x − y + 2. for some set S. . Show that if (ab)2 = a2 b2 . Now might be a good time to pick up your linear algebra text and review vector spaces and subspaces. y ∈ G. Write out the multiplication table for Z× .

(b) Find a subgroup of S7 that contains 12 elements. You do not have to list all of the elements if you can explain why there must be 12. and let n be a ﬁxed positive integer. where o(a) is the order of a. 2. It is also important to remember that every element a in a group deﬁnes a subgroup a . Suppose that p is a prime number of the form p = 2n + 1. GROUPS Lagrange’s theorem is very important.2 23. show that 21 H = {[x]21 | x ≡ 1 (mod 3)} are subgroups of G. since in that case the inverse a−1 of any element can be expressed in the form an . show that cos θ − sin θ H= θ∈R sin θ cos θ is a subgroup of G. 25. and K = {[x]21 | x ≡ 1 (mod 7)} . (a) Show that in Z× the order of [2]p is 2n. 26. consisting of all powers (positive and negative) of the element. In the multiplicative group C× of √ complex numbers. In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. 24 24. 2 2 2 2 30. This subgroup has o(a) elements. (a) Find the cyclic subgroup of S7 generated by the element (1. ﬁnd two subgroups of order 4. 3)(5. If the group is ﬁnite. It states that in a ﬁnite group the number of elements in any subgroup must be a divisor of the total number of elements in the group. SOLVED PROBLEMS: §3. p (b) Use part (a) to prove that n must be a power of 2. In Z× . 28. one that is cyclic and one that is not 20 cyclic. for some n > 0. This is a useful fact to know when you are looking for subgroups in a given group. 7). then you only need to look at positive powers. Show that N = {g ∈ G | g = an for some a ∈ G} is a subgroup of G. 27. In G = Z× . Let G be an abelian group. Find all cyclic subgroups of Z× . 29.16 CHAPTER 3. and why they must form a subgroup. ﬁnd the order of the √ √ √ 2 2 2 2 elements − + i and − − i.

This generalizes the construction of n-dimensional vector spaces (that case is much simpler since every entry comes from the same set). Let K be the following subset of GL2 (R).3.14 in the text deﬁnes the centralizer of an element a of the group G to be C(a) = {x ∈ G | xa = ax}. using ordered pairs. 19.2.7. This can be extended to n-tuples. and this allows us to construct very interesting ﬁnite groups as subgroups of GLn (Zp ). 3. let H = {[x] | x ≡ 1 (mod 4)} and K = {[y] | y ≡ 36 1 (mod 9)}. ad − bc = 0 17 Show that K is a subgroup of GL2 (R). which takes two known groups and constructs a new one. The second construction in this section is the direct product. 18. 32. then the order of the general linear group GLn (Zp ) is (pn − 1)(pn − p) · · · (pn − pn−1 ). Find two groups G1 and G2 whose direct product G1 × G2 has a subgroup that is not of the form H1 × H2 .3. Compute the centralizer in GL2 (R) of the matrix 2 1 1 1 . . 20. and ﬁnd the subgroup HK. K= a b c d d = a. CONSTRUCTING EXAMPLES 31.3 16. [15]18 ) in the group Z12 × Z18 . which shows that the set of all invertible n × n matrices forms a group.3 Constructing Examples The most important result in this section is Proposition 3. c = −2b. in which we can allow the entries in the matrix to come from any ﬁeld. where the entry in the ith component comes from a group Gi . In the group G = Z× . for any prime number p. Note: Exercise 3. Find the order of the element ([9]12 . Show that if p is a prime number. for subgroups H1 ⊆ G1 and H2 ⊆ G2 . Show that Z5 × Z3 is a cyclic group. Show that H and K are subgroups of G. and list all of the generators for the group. and ntuples are multiplied component-by-component. This includes matrices with entries in the ﬁeld Zp . SOLVED PROBLEMS: §3.3. 17.

Compute the centralizer in GL2 (Z3 ) of the matrix 24. b ∈ Z5 .4 Isomorphisms A one-to-one correspondence φ : G1 → G2 between groups G1 and G2 is called a group isomorphism if φ(ab) = φ(a)φ(b) for all a. the existence of an isomorphism guarantees that there is a way to set up a correspondence between the elements of the groups in such a way that the group multiplication tables will look exactly the same. In terms of the respective group multiplication tables for G1 and G2 . where 0 ≤ i < 5 and 0 ≤ j < 2. . m = ±1 (a) Show that H is a subgroup of G with 10 elements. Find the centralizers C(A) and 0 1 0 1 C(B). Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . 3. 23. H= m b 0 1 ∈ GL2 (Z5 ) m. 2 0 2 1 1 2 1 1 . Let A = 1 1 −1 0 and B = .18 CHAPTER 3. b ∈ G1 . The condition that φ(ab) = φ(a)φ(b) for all a. GROUPS i 0 0 0 in the group GL3 (C). since it is one-to-one and onto. where Z(G) is the center of G. Let H be the following subset of the group G = GL2 (Z5 ). since the inverse function φ−1 also respects the multiplication of the two groups. Find the order of the element A = 0 −1 0 0 −i 22. 21. and show that C(A) ∩ C(B) = Z(G). (b) Show that if we let A = 1 1 0 1 and B = −1 0 0 1 . b ∈ G1 makes certain that multiplication can be done in either group and the transferred to the other. Compute the centralizer in GL2 (Z3 ) of the matrix . then BA = A−1 B. The function φ can be thought of as simply renaming the elements of G1 . 25. (c) Show that every element of H can be written uniquely in the form Ai B j .

. because there are inﬁnitely many possibilities. In fact. you probably need to see some similarity between the group operations. for all x ∈ R. 23. x2 ) ∈ G1 × G2 . and let H be a subgroup of G. Z2 × Z2 × Z2 . [4n + 3m]6 ). Furthermore. If you can show that one group has a property that the other one does not have. for each positive integer n. for all (x1 . First prove that φ is a well-deﬁned function. Let φ : R× → R× be deﬁned by φ(x) = x3 . Z4 × Z2 . To decide on the function to use.3.4 21. G2 . 7 11 25. ISOMORPHISMS 19 From an algebraic perspective. But if we ask for a list of abelian groups of order 8 that comes with a guarantee that any possible abelian group of order 8 must be isomorphic to one of the groups on the list. we can show (in Section 7. Deﬁne φ : G1 × G2 → H1 × H2 by φ(x1 . 26. [m]2 ) ∈ Z30 × Z2 . Prove that the group Z× × Z× is isomorphic to the group Z6 × Z10 . you should actually produce an isomorphism φ : G1 → G2 . H1 . To show that two groups G1 and G2 are isomorphic. then you can decide that two groups are not isomorphic (provided that the property would have been transferred by any isomorphism). then so is G2 . Show that φ is a group isomorphism. 24. Let G1 . [m]2 ) = ([n]10 .5) that the answer to this particular question is the list Z8 . In this situation we would usually say that we have found all abelian groups of order 8. 17 22. x2 ) = (θ1 (x1 ). and then prove that φ is a group isomorphism. If G1 is abelian. Deﬁne φ : Z30 × Z2 → Z10 × Z6 by φ([n]30 . The problem of ﬁnding all abelian groups of order 8 is impossible to solve. Prove that if a is any element of G. θ2 (x2 )). we should think of isomorphic groups as being essentially the same. if G1 is cyclic. then the question becomes manageable. Each time you meet a new property of groups. the two groups must have exactly the same number of elements of order n. you should ask whether it is preserved by any isomorphism. Show that Z× is isomorphic to Z16 . H2 be groups. Let G be a group. then so is G2 . then the subset aHa−1 = {g ∈ G | g = aha−1 for some h ∈ H} is a subgroup of G that is isomorphic to H. and suppose that θ1 : G1 → H1 and θ2 : G2 → H2 are group isomorphisms. up to isomorphism. In some ways it is harder to show that two groups are not isomorphic. Suppose that G1 and G2 are isomorphic groups. for all ([n]30 . Prove that φ is a group isomorphism. SOLVED PROBLEMS: §3.4.

5.3. for all x1 .5. Prove that if G is isomorphic to G1 × G2 . then there are subgroups H and K in G such that H ∩ K = {e}. (a) In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. a21 = 0. 31. b ∈ Z3 . a22 = 1 is a subgroup of G. HK = G. show that H= a11 a21 a12 a22 ∈ GL2 (R) a11 = 1. G2 be groups. and that φ is actually a group isomorphism. and hk = kh for all h ∈ H and k ∈ K. showing that they are in one-to-one correspondence with the . which describes the subgroups of Zn . Deﬁne an operation on S by setting x1 · x2 = φ(φ−1 (x1 )φ−1 (x2 )). the subgroup of diagonal matrices in GL2 (Zp ) is isomorphic to Z× × Z× . Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . and let S be any set for which there exists a one-toone and onto function φ : G → S. G1 . Let H be the following subgroup of group G = GL2 (Z3 ). Let G. m = 0 Show that H is isomorphic to the symmetric group S3 . 3. You should pay particular attention to Proposition 3. In this section. p p 29. under addition. by looking at the group Z of integers. x2 ∈ S. 28. so all of the information about cyclic groups is already contained in these basic examples.2 shows that every cyclic group is isomorphic to one of these concrete examples.5 Cyclic Groups We began our study of abstract algebra very concretely. and the related groups Zn . 30. Show that for any prime number p. Prove that S is a group under this operation. H= m b 0 1 ∈ GL2 (Z3 ) m. GROUPS 27. and this motivated the deﬁnition of an abstract cyclic group. (b) Show that H is isomorphic to the group R of all real numbers. 32. Show that G is not isomorphic to the direct product R× × R. Theorem 3. Let G be a group.20 CHAPTER 3. We discovered that each of these groups is generated by a single element.

then the subgroups are “linearly ordered” in the sense that given any two subgroups. ﬁnd the order of the subgroup [18]30 . Prove that if p and q are diﬀerent odd primes. and Z× are isomorphic to each other. ﬁnd the order of [24]30 . 4) and (1. That isn’t as impressive as it sounds at ﬁrst. 9 18 21.6. Cayley’s theorem shows that permutation groups contain all of the information about ﬁnite groups. 21 29.) SOLVED PROBLEMS: §3. because as n gets larger and larger. This does imply. 2. since every ﬁnite group of order n is isomorphic to a subgroup of the symmetric group Sn . that from the algebraists point of view the abstract deﬁnition of a group is really no more general than the concrete deﬁnition of a permutation . 25.5. Give the lattice diagram of subgroups of Z100 . 3. Is Z4 × Z10 isomorphic to Z2 × Z20 ? 22. 28. 27. 3). Show that any cyclic group of even order has exactly one element of order 2. Find all generators of the cyclic group Z28 . Z× . 30. 26. and form the basic building blocks for all ﬁnite abelian groups. one is a subset of the other. then Z× is not a cyclic group. Is Z4 × Z15 isomorphic to Z6 × Z10 ? 23. Use this information to show that the quaternion group cannot be isomorphic to the subgroup of S4 generated by (1. Prove that if G1 and G2 are groups of order 7 and 11. In n is a prime power. PERMUTATION GROUPS 21 positive divisors of n. Find all cyclic subgroups of the quaternion group. (In Theorem 7. which has order n!.5 20.3. this section revisits the roots of group theory that we began to study in an earlier chapter. respectively. Use the the result in Problem 27 to show that the multiplicative groups Z× 15 and Z× are not cyclic groups. the subgroups of order n just get lost inside the larger symmetric group. pq 3. In Z30 . 24. Show that the three groups Z6 . however. then the direct product G1 × G2 is a cyclic group.6 Permutation Groups As with the previous section. These cyclic groups have a particularly simple structure.4 we will prove that every ﬁnite abelian group is isomorphic to a direct product of cyclic groups of prime power order.

7. show that bai = an−i b. for σ = (1. in S4 . after we develop the concepts of cosets and factor groups. and ba = a−1 b. and ba = a−1 b. 4). and in A4 . for all 0 ≤ i < n. 26. which is a preliminary form of the Fundamental Homomorphism Theorem. in terms of the rigid motions of an n-gon. o(b) = 2. 23. but can also be described more abstractly in terms of two generators a (of order n) and b (of order 2) which satisfy the relation ba = a−1 b. There are many important examples of group homomorphisms that are not isomorphisms.7 Homomorphisms In Section 3. 28. 0 ≤ j < 2} with o(a) = n. You should make every eﬀort to get to know the dihedral groups Dn . with o(a) = n. 3. show that each element of the form ai b has order 2. 0 ≤ j < 2. Show that each element in A4 can be written as a product of 3-cycles. We can write Dn = {ai bj | 0 ≤ i < n. The most important result in this section is Theorem 3. 27. 3) and (1. and studied in detail what it means for two groups to be isomorphic. In the dihedral group Dn = {ai bj | 0 ≤ i < n.6 22. 3) in S3 . ﬁnd the centralizer of a. homomorphisms provide the way to relate one group to another. In the dihedral group Dn = {ai bj | 0 ≤ i < n. 25.) In this formulation of the Fundamental Homomorphism Theorem. shown in the ﬁrst of the solved problems given below. Find the centralizer of (1. o(b) = 2. In this section we look at functions that respect the group operations but may not be one-to-one and onto. SOLVED PROBLEMS: §3.4 we introduced the concept of an isomorphism. It is easy to prove that the image φ(G1 ) is . 2.8. The abstract deﬁnition of a group is useful simply because it can be more easily applied to a wide variety of situation. They have a concrete representation. In doing computations in Dn it is useful to have at hand the formula bai = an−i b. (The full statement is given in Theorem 3. 0 ≤ j < 2} with o(a) = n. and. In the dihedral group Dn = {ai bj | 0 ≤ i < n. GROUPS group.8. 24. we start with a group homomorphism φ : G1 → G2 . o(b) = 2. and ba = a−1 b.22 CHAPTER 3.8. in fact. In S4 . and ba = a−1 b} . ﬁnd the corresponding subgroup σHσ −1 . 2). o(b) = 2. 0 ≤ j < 2} with o(a) = n. 2. ﬁnd the subgroup H generated by (1. For the subgroup H of S4 deﬁned in the previous problem.

Let φ1 : G → G1 and φ2 : G → G2 be group homomorphisms. and show how elements of the image correspond to cosets of the kernel.5 are important. for all x ∈ G. 24. (b) Choose one of the nonzero formulas in part (a). 17 (c) Completely determine all group homomorphisms from Z× into Z× . since for any group homomorphism φ : G1 → G2 it describes the connections between subgroups of G1 and subgroups of G2 . Let n and m be positive integers. (a) Show that Z× is cyclic.7. using the operation in G1 . 18.7. 7 (b) Show that Z× is cyclic. such that m is a divisor of n. Prove that Z× is isomorphic to the direct pq product Z× × Z× . and G2 be groups. For the group homomorphism φ : Z× → Z× deﬁned by φ([x]36 ) = [x]12 . Examples 3. (a) Find the formulas for all group homomorphisms from Z18 into Z30 . it does deﬁne an isomorphism between a subgroup of G2 and what we call a factor group of G1 . SOLVED PROBLEMS: §3.8 shows that this group is isomorphic to φ(G1 ). because they give a complete description of all group homomorphisms between two cyclic groups. with generator [3]17 . 21. Proposition 3. Deﬁne φ : Z4 × Z6 → Z4 × Z3 by φ(x.7. where we use the equivalence relation deﬁned by congruence modulo n. The function φ has an equivalence relation associated with it. Find all group homomorphisms from Z4 into Z10 . (a) Show that φ is a well-deﬁned group homomorphism. ﬁnd the kernel and image of φ. 22. 23. for a.3. Let G. φ2 (x)). we can deﬁne a group operation on the equivalence classes of ∼. p q . Just as in Z. Prove that φ : G → G1 × G2 deﬁned by φ(x) = (φ1 (x). for all [x]n ∈ Z× . HOMOMORPHISMS 23 a subgroup of G2 . (b) Find the kernel and image of φ.7. 19. is a well-deﬁned n m n group homomorphism. where we let a ∼ b if φ(a) = φ(b). and for this formula ﬁnd the kernel and image.7 17. and apply the fundamental 36 homomorphism theorem.7. Then Theorem 3. Show that φ : Z× → Z× deﬁned by φ([x]n ) = [x]m . G1 . b ∈ G1 . is a well-deﬁned group homomorphism. and apply the fundamental homomorphism theorem. y) = (x + 2y.4 and 3.6 is also useful. with generator [3]7 . so that although the homomorphism may not be an isomorphism between G1 and G2 . for 36 12 all [x]36 ∈ Z× . Let p and q be diﬀerent odd primes. 17 7 20. y).

(a) List all cosets of H. . Once again. Then you must prove that your function is one-to-one. if you deﬁne a function on the cosets of G/N . The key idea is to begin thinking of equivalence classes as elements in their own right. SOLVED PROBLEMS: §3. in addition to proving that it is onto and respects the operations in the two groups. then the Fundamental Theorem does the rest of the work. Of course. 0). List the cosets of 7 in Z× . it can be much trickier to prove that it is one-to-one than to simply compute the kernel of a homomorphism deﬁned on G. the most convenient way is use a formula deﬁned on representatives of the cosets of N . This construction parallels the construction of Zn from Z. Is the factor group Z× / 7 cyclic? 16 16 28. making sure that N is the kernel of φ. Quite often we need to show that a factor group G/N that we have constructed is isomorphic to another group G1 . Let the dihedral group Dn be given via generators and relations. If you prove that φ maps G onto G1 . and then in Section 1. 0). The only complication is that the equivalence relation respects the operation in G only when the subgroup is a normal subgroup. (3. list all cosets of K. But then you must be careful to prove that the formula you are using does not depend on the particular choice of a representative. we start with a normal subgroup and the equivalence classes it determines. where at ﬁrst we thought of congruence classes as inﬁnite sets of integers. you ordinarily don’t need to worry that it is well-deﬁned. 0)}. GROUPS 3. it is important to let the theorem do its job. satisfying ba = a−1 b. if your function is deﬁned on cosets. and let K = {(0. Let G = Z6 × Z4 .4 when we started working with Zn we started to use the notation [a]n to suggest that we were now thinking of a single element of a set. Which answer is correct for G/H? For G/K? 29. That is what we did in Chapter 1. where we have a ≡ b (mod n) if and only if a − b ∈ nZ. in an abelian group we can use any subgroup. On the other hand. To construct a factor group. In actually using the Fundamental Homomorphism Theorem. with generators a of order n and b of order 2.24 CHAPTER 3. Normal Subgroups. The moral of this story is that if you deﬁne a function on G rather than G/N .8 Cosets. let H = {(0. (0. since all subgroups of an abelian group are normal. showing that there exists a well-deﬁned isomorphism between G/N and G1 . That is. and Factor Groups The notion of a factor group is one of the most important concepts in abstract algebra. (b) You may assume that any abelian group of order 12 is isomorphic to either Z12 or Z6 × Z2 . so that it does as much of the hard work as possible. The easiest way to do this is to just deﬁne a homomorphism φ from G to G1 .8 27. 2)}. you must prove that your formula actually deﬁnes a function.

if there exists g ∈ G such that b = gag −1 . then HN/N is isomorphic to H. b ∈ G we say that b is conjugate to a. The equivalence classes of ∼ are called the conjugacy classes of G. 34. a3 .3. a3 } of G. (b) Show that any element of the form ai b has order 2. Let G be the dihedral group D12 . (b) Prove that N is a normal subgroup of HN . and list all cosets of N . COSETS. 33.8. For a. (a) Show that N is a normal subgroup of G. and let N = {e. and let N and H be subgroups of G such that N is normal in G. Find the conjugacy classes of D4 . a6 . Let G = D6 and let N be the subgroup a3 = {e. (b) Show that a subgroup N of G is normal in G if and only if N is a union of conjugacy classes. Show that ∼ is an equivalence relation on G. . AND FACTOR GROUPS (a) Show that bai = a−i b for all i with 1 ≤ i < n. Which is correct? 32. (a) Prove that N is a normal subgroup of G. (c) List all left cosets and all right cosets of b 30. Let G be a group. a9 }. (c) Prove that if H ∩ N = {e}. (a) Prove that HN is a subgroup of G. (a) Let G be a group. NORMAL SUBGROUPS. 25 (b) You may assume that G/N is isomorphic to either Z6 or S3 . (b) Is G/N abelian? 31. written b ∼ a.

GROUPS Review Problems 1. . a3 . Assume that the dihedral group D4 is given as {e. and verify that N is a normal subgroup of G. (b) Show that Z× is a cyclic group. 7. Show that the group G in the previous problem is isomorphic to the direct product R × R. b ∈ R . ab. (b) Show that G/N is isomorphic to the multiplicative group R× . and let N be the subset of all matrices of the form . but is not cyclic. Let G = D8 . and give the lattice diagram which shows the 11 inclusions between them. List the cosets of the cyclic subgroup 9 in Z× . and let N = {e. a2 . a. b. 0 1 Show that G is a subgroup of GL3 (R). Is Z× / 9 cyclic? 20 20 6. a2 b. 5. and ba = a3 b. 4. a3 b}. (b) Is the factor group D4 /N a cyclic group? 8. Let N be the subgroup a2 = {e. b2 = e. a2 . (b) Show that G/N has order 4. a4 . Let G be the subgroup of 1 0 0 GL3 (R) consisting of all matrices of the form a b 1 0 such that a. 13 2. Find all subgroups of Z× . 0 1 0 1 (a) Show that N is a subgroup of G. (a) What are the possibilities for the order of an element of Z× ? Explain 13 your answer. Let G be the subgroup of GL2 (R) consisting of all matrices of the form m b 1 b . a6 }. (a) Show by a direct computation that N is a normal subgroup of D4 . where a4 = e. and that N is normal in G. a2 }.26 CHAPTER 3. 3. (a) List all left cosets and all right cosets of N .

27 . 4. which is quite closely related to the deﬁnition of a group (now there are two operations instead of just one). Review Problems 1. we want to go beyond high school algebra. Of course. ﬁnd the greatest common divisor of 2x4 − x3 + x2 + 3x + 1 and 2x3 − 3x2 + 2x + 2 and express it as a linear combination of the given polynomials. I didn’t break the problems up by section. x6 − 1) in Q[x] and write it as a linear combination of x8 − 1 and x6 − 1. to be able to work with coeﬃcients that may not be real numbers. Over the ﬁeld of rational numbers. In addition to generalizing ideas from the integers to polynomials. The point of saying this is that it will be worth your time to review the deﬁnitions and theorems in Chapter 1. use the Euclidean algorithm to show that 2x3 − 2x2 − 3x + 1 and 2x2 − x − 2 are relatively prime. 2. just as we did with congruence classes of integers. Over the ﬁeld of rational numbers. This motivates the deﬁnition of a ﬁeld. The notion of a prime number is replaced by that of an irreducible polynomial. Use the Euclidean algorithm to ﬁnd gcd(x8 − 1. 3. The point here is that you can beneﬁt from reviewing Chapter 3. ﬁnd the greatest common divisor of x4 + x3 + 2x2 + x + 1 and x3 − 1. We need to talk about the greatest common divisor of two polynomials. and when two polynomials are relatively prime. Because you have a lot more experience now than when you started Chapter 1. and express it as a linear combination of the given polynomials. you don’t have to wait until you have ﬁnished the chapter to practice solving some of these problems.Chapter 4 POLYNOMIALS In this chapter we return to several of the themes in Chapter 1. We can work with congruence classes of polynomials. Over the ﬁeld of rational numbers.

(b) in Z3 [x]/ x3 + 2x2 + x + 1 . Factor x5 − 2x4 − 2x3 + 12x2 − 15x − 2 over Q. Are the following polynomials irreducible over Q? (a) 3x5 + 18x2 + 24x + 6 (b) 7x3 + 12x2 + 3x + 45 (c) 2x10 + 25x3 + 10x2 − 30 6. In Z2 [x]/ x3 + x + 1 . POLYNOMIALS 5. Construct an example of a ﬁeld with 343 = 73 elements. ﬁnd [x]−1 and [x + 1]−1 . (b) List the elements of the ﬁeld F = Z3 [x]/ x2 + 1 . 15. (a) Show that x2 + 1 is irreducible over Z3 . Factor x5 − 10x4 + 24x3 + 9x2 − 33x − 12 over Q. 13. . ﬁnd the multiplicative inverse of [x + 1]. but [x] is not. 14. and use your answers to ﬁnd [x2 + x]−1 . 8.28 CHAPTER 4. 12. (c) In the multiplicative group of nonzero elements of F . (a) Express x4 + x as a product of polynomials irreducible over Z5 . 7. Express 2x3 + x2 + 2x + 2 as a product of polynomials irreducible over Z5 . 11. Factor x4 + x + 1 over Z2 [x]/ x4 + x + 1 . 10. Find the multiplicative inverse of [x2 + x + 1] (a) in Q[x]/ x3 − 2 . 9. (b) Show that x3 + 2x2 + 3 is irreducible over Z5 . show that [x + 1] is a generator. In Z5 [x]/ x3 + x + 1 .

(c) Find all idempotent elements of R. The concept of a factor ring depends heavily on the corresponding deﬁnition for groups. (b) Find all units of R. (a) Show that R is a commutative ring (you only need to check closure and commutativity of multiplication).Chapter 5 COMMUTATIVE RINGS This chapter takes its motivation from Chapter 1 and Chapter 4. 3. Review Problems 1. Let R be the ring with 8 elements consisting of all 3 × 3 matrices with entries in Z2 which have the following form: a 0 0 0 a 0 b c a You may assume that the standard laws for addition and multiplication of matrices are valid. Let R be the ring Z2 [x]/ x2 + 1 . Remember that the distributive law is all that connects the two operations in a ring. 4. so it is crucial in many of the proofs you will see. it is not isomorphic to either of the rings Z4 or Z2 ⊕ Z2 . Show that although R has 4 elements. and all nilpotent elements of R. 2. Find all ring homomorphisms from Z120 into Z42 . Are Z9 and Z3 ⊕ Z3 isomorphic as rings? 29 . extending results on factorization to more general settings than just the integers or polynomials over a ﬁeld. so you may need to review the last two sections of Chapter 3.

Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients. . prove that I is not maximal. Let R be the ring Z2 [x]/ x3 + 1 . Show that the rings R and S in the two previous problems are isomorphic as abelian groups. 7. 2) of the ring R = Z12 ⊕ Z8 . (c) Find the idempotent elements of R. (a) Show that if e is an idempotent element of R. Let R be any commutative ring with identity 1. 9. 13. then R ∼ Re ⊕ R(1 − e). you can do the ﬁrst two parts together. Show that Ann(a) is an ideal of R. In the group Z× of units of the ring Z180 . (a) Show that I is an ideal of R. (b) Show that if e is idempotent. 14. then 1−e is also idempotent. (a) Find all ideals of R. (a) Find all ideals of S. but not as rings. (b) Find the units of R. 10. 8. ﬁnd Ann(a) = {r ∈ R | ra = 0}. (b) Show that R/I ∼ Z2 [x]/ x2 + 1 . n ∈ Z} . Find all maximal ideals. Let R be the ring Z2 [x]/ x4 + 1 . (b) Find the units of R. and let I be the set of all congruence classes in R of the form [f (x)(x2 + 1)]. 15. COMMUTATIVE RINGS 5. what is the largest possible order 180 of an element? 6. Let Z[i] be the subring of the ﬁeld of complex numbers given by Z[i] = {m + ni ∈ C | m. For the element a = (0. = 12. Prove that I is a prime ideal. and all prime ideals.30 CHAPTER 5. of Z36 = Z/36Z. 11. (c) Find the idempotent elements of R. Give an example to show that the set of all zero divisors of a ring need not be an ideal of the ring. = (c) Is I a prime ideal of R? Hint: If you use the fundamental homomorphism theorem. Let S be the ring Z2 [x]/ x3 + x .

(a) Show that φ is a ring homomorphism. Let I and J be ideals in the commutative ring R. and deﬁne the function φ : R → R/I ⊕ R/J by φ(r) = (r + I. Considering Z[x] to be a subring of Q[x]. deﬁne θ : Z[i] → Zp [x]/ x2 + 1 by θ(m + ni) = [m + nx]. Prove that θ is an onto ring homomorphism.31 (a) Deﬁne φ : Z[i] → Z2 by φ(m + ni) = [m + n]2 . with ker(φ) = I ∩ J. for all r ∈ R. Find ker(φ) and show that it is a principal ideal of Z[i]. (b) For any prime number p. = 17. Let p be an odd prime number that is not congruent to 1 modulo 4. Hint: Show that a root of x2 = −1 leads to an element of order 4 in the multiplicative group Z× . and thus R/(I ∩ J) ∼ R/I ⊕ R/J . r + J). (b) Show that if I + J = R. Prove that φ is a ring homomorphism. then φ is onto. p . show that these two integral domains have the same quotient ﬁeld. Prove that the ring Zp [x]/ x2 + 1 is a ﬁeld. 18. 16.

COMMUTATIVE RINGS .32 CHAPTER 5.

Chapter 6 FIELDS These review problems cover only the ﬁrst three sections of the chapter. Find the minimal polynomial of 1 + 3 2 over Q. Show that x3 + 6x2 − 12x + 2 is irreducible over Q. If you are studying abstract algebra because you plan to be a high school teacher. and remains irreducible √ over Q( 5 2). it is precisely these sections (along with the earlier material on polynomials) that are the most relevant to what you will be teaching. Let u be a root of the polynomial x3 + 3x + 3. √ √ 5. Does 4 2 belong to Q( 3 2 + i)? 33 . i). (a) Show that Q( 2 + i) = Q( 2. √ √ 6. express (7 − 2u + u2 )−1 in the form a + bu + cu2 . √ √ 2. √ √ 7. Find the degree of 3 2 + i over Q. √ 3. 4. Review Problems 1. Show that [Q( 2 + 3 5) : Q] = 6. √ (b) Find the minimal polynomial of 2 + i over Q. √ √ √ 8. In Q(u). Find [Q( 7 16 + 3 7 8) : Q]. Find a basis for Q( 5. 3 5) over Q.

34 CHAPTER 6. FIELDS .

Chapter 1

Integers

1.1 SOLUTIONS

22. Find gcd(435, 377), and express it as a linear combination of 435 and 377. Comment: You deﬁnitely need to know how to do these computations. Solution: We will use the Euclidean algorithm. Divide the larger number by the smaller, which should give you a quotient of 1 and a remainder of 58. Then divide the remainder 58 into 377, and continue the Euclidean algorithm as in Example 1.1.4 in the text. That should give you the following equations. 435 = 1 · 377 + 58 377 = 6 · 58 + 29 58 = 2 · 29 gcd(435, 377) = gcd(377, 58) = gcd(58, 29) = 29

The repeated divisions show that gcd(435, 377) = 29, since the remainder in the last equation is 0. To write 29 as a linear combination of 435 and 377 we need to use the same equations, but we need to solve them for the remainders. 58 29 = = 435 − 1 · 377 377 − 6 · 58

Now take the equation involving the remainder 29, and substitute for 58, the remainder in the previous equation. 29 = 377 − 6 · 58 = 377 − 6 · (435 − 1 · 377) = 7 · 377 − 6 · 435

This gives the linear combination we need, 29 = (7)(377) − (6)(435). 35

36

CHAPTER 1 SOLUTIONS

23. Find gcd(3553, 527), and express it as a linear combination of 3553 and 527. Comment: This time we will use the matrix form of the Euclidean algorithm. You should be able to use both the back-solving form (as in Problem 22) and the matrix form. In Chapter 4, the Euclidean algorithm is used for polynomials, and the matrix method just gets too complicated, so we have to adapt the back-solving method. Solution: Just as in Problem 22, the ﬁrst step is to divide the smaller number into the larger. We get 3553 = 6 · 527 + 391, so this tells us to multiply the 1 0 3553 by 6 and subtract from the ﬁrst bottom row of the matrix 0 1 527 row. The rest of the steps in reducing the matrix to the form we want should be clear. We have 1 0 0 1 3553 527 ; 1 −6 391 0 1 527 ; ; 1 −6 391 −1 7 136 ; ; 0 17 .

3 −20 119 −1 7 136

3 −20 119 −4 27 17

31 −209 −4 27

Therefore gcd(3553, 527) = 17, and 17 = (−4)(3553) + (27)(527). 24. Which of the integers 0, 1, . . . , 10 can be expressed in the form 12m + 20n, where m, n are integers? Solution: Theorem 1.1.6 provides the answer. An integer k is a linear combination of 12 and 20 if and only if it is a multiple of their greatest common divisor, which is 4. Therefore we can express 0, 4, and 8 in the required form, but we can’t do it for the rest. Comment: Check out the answer in concrete terms. We can write 0 = 12 · 0 + 20 · 0; 4 = 12 · 2 + 20 · (−1); 8 = 12 · (−1) + 20 · 1. 25. If n is a positive integer, ﬁnd the possible values of gcd(n, n + 10). Solution: Let d = gcd(n, n + 10). Then d|n and d|(n + 10), so we must have d|10, and therefore d is limited to one of 1, 2, 5, or 10. Can each of these occur for some n? Yes: gcd(3, 13) = 1; gcd(2, 12) = 2; gcd(5, 15) = 5; gcd(10, 20) = 10. 26. Prove that if a and b are nonzero integers for which a|b and b|a, then b = ±a. Comment: The ﬁrst step is to use Deﬁnition 1.1.1 to rewrite a|b and b|a as equations, to give something concrete to work with. Solution: Since a | b, there is an integer m with b = ma. Since b | a, there is an integer k with a = kb. Substituting a = kb in the equation b = ma we get b = m(kb), so since b is nonzero we can cancel it to get 1 = mk. Since both m and k are integers, and |1| = |m| |k|, we must have |m| = 1 and |k| = 1, so either b = a or b = −a.

CHAPTER 1 SOLUTIONS 27. Prove that if m and n are odd integers, then m2 − n2 is divisible by 8.

37

Solution: First, we need to use the given information about m and n. Since they are odd, we can write them in the form m = 2k + 1 and n = 2q + 1, for some integers k and q. We can factor m2 − n2 to get (m + n)(m − n), so substituting for m and n we get m2 − n2 = (2k + 1 + 2q + 1)(2k + 1 − 2q − 1) = (2)(k + q + 1)(2)(k − q) . Now we need to take two cases. If k − q is even, then k − q has 2 as a factor, say k − q = 2p, for some integer p. Substituting for k − q gives us m2 − n2 = (2)(k + q + 1)(2)(2)(p) = (8)(k + q + 1)(p) . If k − q is odd, then k + q = (k − q) + (2q) is the sum of an odd integer and an even integer, so it must also be odd. That means that k + q + 1 is even, so it has 2 as a factor. Now we can suppose that k + q + 1 = 2t, for some integer t. In this case, substituting for k + q + 1 gives us m2 − n2 = (2)(2)(t)(2)(k − q) = (8)(t)(k − q) . Showing that we can factor 8 out of m2 − n2 gives exactly what we were to prove: if m and n are odd, then m2 − n2 is divisible by 8. 28. Prove that if n is an integer with n > 1, then gcd(n − 1, n2 + n + 1) = 1 or gcd(n − 1, n2 + n + 1) = 3. Comment: It’s not a bad idea to check this out for some values of n, just to get a feeling for the problem. For n = 3, we have gcd(2, 13) = 1. For n = 4, we have gcd(3, 21) = 3. For n = 5, we have gcd(4, 31) = 1. For n = 6, we have gcd(5, 43) = 1. For n = 7, we have gcd(6, 57) = 1. These calculations don’t prove anything, but maybe they do make the problem look plausible. Solution: Problem 25 gives a hint. In that problem, since the gcd was a divisor of n and n + 10, it had to be a divisor of 10. To use the same approach, we would have to write n2 + n + 1 as n − 1 plus something. That doesn’t work, but we are very close. Dividing n2 + n + 1 by n − 1 (using long division of polynomials) we get a quotient of n + 2 and a remainder of 3, so n2 + n + 1 = (n + 2)(n − 1) + 3. Now we can see that any common divisor of n − 1 and n2 + n + 1 must be a divisor of 3, so the answer has to be 1 or 3. n 0 0 −1 1 0 0 0 = 0 1 0 29. Prove that if n is a positive integer, then 0 1 0 0 1 1 0 0 if and only if 4|n. Comment: Let’s use A for the matrix, and I for the identity matrix. The proof must be given in two pieces. We need to show that if 4|n, then An = I.

but we need to involve the expression 10k + 2 in some way. The next step is to prove that if we assume that the statement is true for n = k. 30.. A4 = A · A3 . We conclude that 4|n. . A3 = A · A2 . 102. . and A3 are not equal to I. the integer 10n + 2 is divisible by 6. and then it turns out that we can factor out 6. for n = 1. then we can show that the statement must also be true for n = k + 1. Solution: We begin by 2 0 0 −1 0 0 1 0 = 0 1 0 0 1 3 0 0 −1 0 0 1 0 = 0 1 0 0 1 4 0 0 −1 0 0 1 0 = 0 1 0 0 1 computing A2 . and it is easier to state as the converse of the ﬁrst statement: if An = I. then 4|n. say 10k + 2 = 6q. 0 −1 0 0 −1 −1 0 0 1 0 0 1 0 = 0 1 0 0 0 1 0 0 0 0 −1 0 −1 −1 0 0 0 0 1 1 0 0 1 0 = 0 1 0 0 0 0 0 −1 −1 0 0 0 −1 0 0 1 1 0 0 1 0 0 1 0 = 0 1 0 0 0 −1 0 0 0 0 1 Now we can see that if 4|n. so we can prove the following statement: for each positive integer n. 10002. We can easily factor a 10 out of 10k+1 . etc. to divide n by 4. Comment: If you are unsure about doing a proof by induction. Conversely. . In the second half of the proof. . The ﬁrst step is to check that the statement is true for n = 1. we need a statement that depends on an integer n. since it just takes a computation. we can use the division algorithm to write n = 4q + r. if An = I. say n = 4q. 10k+1 + 2 = (10)(10k ) + 20 − 20 + 2 = (10)(10k + 2) − 18 = (10)(6q) − (6)(3) = (6)(10q − 3) . then An = A4q = (A4 )q = I q = I. is divisible by 6. The ﬁrst half of the proof is easier than the second.38 CHAPTER 1 SOLUTIONS We also need to show that An = I only when 4|n. 1002. Solution: To give a proof by induction. if An = I then we will use the division algorithm. We can write the numbers in the given sequence in the form 10n + 2. so r = 0 since A.) Clearly 12 is divisible by 6. Give a proof by induction to show that each number in the sequence 12. with 0 ≤ r < 4. 2. to get 10k+1 + 2 = (10)(10k ) + 2. you should read Appendix 4 in the text. A2 . (This “anchors” the induction argument. . Adding and subtracting 20 makes it possible to get this term. Then Ar = An−4q = An (A−4 )q = I · I q = I. and then look at the expression when n = k + 1. Let’s start by assuming that 10k + 2 is divisible by 6. for some q ∈ Z.. . and then show that the remainder has to be 0.

In each diagram. Solution: We have 1274 = 1089 · 1 + 185. 1089). Thus gcd(1776. we need to use one axis for each prime. Proposition 1. (a) Use the Euclidean algorithm to ﬁnd gcd(1776. Solution: Since 1776 = 24 · 3 · 37 and 1492 = 22 · 373. 1. 1492) = 22 . in succession. Solution: Since 1274 = 2 · 72 · 13 and 1089 = 32 · 112 . to give the factors along the corresponding axis. 24. (a) Use the Euclidean algorithm to ﬁnd gcd(1274. 1089). 284 = 72 · 3 + 68. then 10k+1 + 2 is divisible by 6. This completes the induction. Solution: The prime factorizations are 250 = 2 · 53 and 484 = 22 · 112 . 25.CHAPTER 1 SOLUTIONS 39 We have now shown that if 10k + 2 is divisible by 6. Then we can just divide (successively) by the prime.9 shows that gcd(1776. For example. 1089 = 185 · 5 + 164. dividing 250 by 5 produces 50. 10. we see that 1274 and 1089 are relatively prime. Do the same for 484. 1492). gcd(1274. Thus 21 = 17 · 1 + 4. Give the lattice diagram of all divisors of 250. 250 125 50 242 484 44 25 10 121 22 4 5 2 11 2 1 1 . These numbers go along one axis of the rectangular diagram. and 2. 72 = 68 · 1 + 4. 1492) = 4. 68 = 4 · 17. 1492 = 284 · 5 + 72. Solution: We have 1776 = 1492 · 1 + 284. 1492). 164 = 21 · 7 + 17. 17 = 4 · 4 + 1.2 SOLUTIONS 23. 1089) = 1. (b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274. (b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776.2. 185 = 164 · 1 + 21.

and (−21. bc) = 1 if and only if gcd(a. (−3.40 CHAPTER 1 SOLUTIONS 26. First. which states that if gcd(2m − 1. we will prove that if gcd(m. 29. we have the four solutions (−1. n − 1) = 1.3 (d) states that gcd(a. Is the same true for 2n − 1 and 3n − 1? Solution: We can write (1)(2n − 1) + (−2)(n − 1) = 1. b2 ) = 1. b2 ) = 1 if and only if gcd(a. assuming that gcd(m. the only way it can be factored is to have 1 · 19 = 19 or (−1) · (−19) = 19. which proves that gcd(2n − 1. 2).2. −16). Find all integer solutions of the equation xy + 2y − 3x = 25. Similarly. 28. That motivates one possible method of solution. Prove that gcd(2m − 1. and so gcd(3n − 1. b. y). Listing the solutions as ordered pairs (x. Before beginning the proof. 2n − 1) = 1 if and only if gcd(m. x + 2 = −1. Therefore we have 4 possibilities: x + 2 = 1. n) = 1. n) = 1. 2n − 1) = 1. prove that gcd(a. we recall that the following identity holds for all values of x: xk − 1 = (x − 1)(xk−1 + xk−2 + · · · + x + 1). xy + 2y − 3x (x + 2)y − 3x (x + 2)y − 3x − 6 (x + 2)y − 3(x + 2) (x + 2)(y − 3) = = = = = 25 25 25 − 6 19 19 Now since 19 is prime. Solution: If we had a product. Substituting x = 2m and k = a in the identity given above shows that 2m − 1 . b) = 1 and gcd(a. Solution: Proposition 1. For each of these values there is a corresponding value for y. Comment: We need to do the proof in two parts. 2n − 1) = 1. n) = 1 and showing that this forces gcd(2m − 1. you don’t have to explain how you found them. Then a similar argument yields gcd(a2 . b ∈ Z with am + bn = 1. 2n − 1) = 1. b2 ) = 1 if and only if gcd(a. since the complementary factor must be equal to y − 3. then there exist a. then gcd(m. b) = 1. c) = 1. Solution: If gcd(m. for all integers n > 1. x + 2 = 19. 2n − 1) = 1. Let m and n be positive integers. n) = 1. For positive integers a. b) = 1 if and only if gcd(a2 . b2 ) = 1. (17. 22). or x + 2 = −19. Comment: Is this really a proof? Yes–producing the necessary linear combinations is enough. then gcd(2m − 1. n) = 1. To prove the converse. we could use the prime factorization theorem. we will use a proof by contradiction. Prove that n − 1 and 2n − 1 are relatively prime. 4). (2)(3n − 1) + (−3)(2n − 1) = 1. Then we will prove the converse. Using c = b gives gcd(a. 27.

−34. n) = d. or x ≡ 11 (mod 15). etc. (Of course. Equivalently. 2n − 1) = 1. so we multiply both sides of the congruence by −2 since (−2)(7) = −14 ≡ 1 (mod 15). 2n2 + 4n − 3) = gcd(2n2 + 4n − 3. say 2am − 1 = (2m − 1)(s). 31.3 SOLUTIONS 26. 56. getting gcd(2n − 1. you could also continue with the Euclidean algorithm. 1) = 1. so there is a solution since 6 is a factor of 12. Solving the congruence 42x ≡ 12 (mod 90) is equivalent solving the equation 42x = 12 + 90q for integers x and q. n − 1) = gcd(n − 2. The proof now involves what may look like a trick (but it is a useful one). and this gives a quotient of n + 2 and a remainder of n − 1. 2n − 1) = gcd(2n − 1. We next use trial and error to look for the multiplicative inverse of 7 modulo 15. Prove that gcd(2n2 + 4n − 3. or 7x ≡ 2 (mod 15). 90) = 6. We have 1 = = = = 21 − 1 2am+bn − 2bn + 2bn − 1 2bn (2am − 1) + 2bn − 1 2bn (s)(2m − 1) + (t)(2n − 1) and so we have found a linear combination of 2m − 1 and 2n − 1 that equals 1. 46. for all integers n > 1. 86 (mod 90). we see that 7 is a factor of −14. The same argument shows that we can write 2bn − 1 = (2n − 1)(t). q ∈ Z with m = dq and n = dp. We have shown that gcd(2n2 + 6n − 4. 2n2 + 6n − 4) = 1.CHAPTER 1 SOLUTIONS 41 is a factor of 2am − 1. The next step is to divide 2n2 + 4n − 3 by 2n − 1. Solve the congruence 42x ≡ 12 (mod 90). −29. 26. and this completes the proof. which proves that gcd(2m − 1. Therefore gcd(2m − 1. we obtain 7x ≡ 2 (mod 15) by dividing 42x ≡ 12 (mod 90) through by 6. But then an argument similar to the one given for the ﬁrst part shows that 2d − 1 is a common divisor of 2dq − 1 and 2dp − 1. Long division of polynomials shows that dividing 2n2 + 6n − 4 by 2n2 + 4n − 3 gives a quotient of 1 and a remainder of 2n − 1. Solution: We can use the Euclidean algorithm. Thus we have −14x ≡ −4 (mod 15). Solution: We have gcd(42. n − 1) and so we can use Problem 28 to conclude that 2n2 + 4n − 3 and 2n2 + 6n − 4 are relatively prime since 2n − 1 and n − 1 are relatively prime. for some t ∈ Z. then there exist p. The solution is x ≡ 11. Among these. say gcd(m. 2n − 1) = 1. . 61. and −14. 30. The numbers congruent to 1 modulo 15 are 16. This reduces to 7x = 2 + 15q. If gcd(m. etc. n) = 1. 41.) 1.. 71. for some s ∈ Z.

(a) Find one particular integer solution to the equation 110x + 75y = 45. Solution: If 110m + 75n = 45. Thus the solution is x ≡ 74 (mod 90). and multiplying both sides by −2 gives x ≡ −3 (mod 15). Comment: The matrix computation shows that 110(15) + 75(−22) = 0. This reduces to 7x ≡ 9 (mod 15). CHAPTER 1 SOLUTIONS 55x ≡ 35 (mod 75). Then 4 + 10q ≡ 2 (mod 9). 27) will also determine a solution. (b) Show that if x = m and y = n is an integer solution to the equation in part (a). Solution: By trial and error. y = 27. 7 · 5 ≡ 1 (mod 17) and 9 · 3 ≡ 1 (mod 13). 62 (mod 75). Solve the system of congruences x ≡ 2 (mod 9) x ≡ 4 (mod 10) . 0 1 75 0 1 75 −2 3 5 −2 3 5 Thus −2(110) + 3(75) = 5. . Solution: Any linear combination of 110 and 75 is a multiple of the gcd. we can solve it in the usual way. Having reduced the system to the standard form. and multiplying by 9 yields a solution x = −18. The solution is 55x ≡ 36 (mod 75). Thus 75y = 45 + 3(110) = 375 and so x = −3. −x ≡ 13 (mod 15). x ≡ 2 (mod 15). since gcd(55. . Solve the system of congruences so 5x ≡ 14 (mod 17). Solution: We have gcd(55. for any integer q. Second solution: The equation reduces to the congruence 35x ≡ 45 (mod 75). 75) = 5. −22) to the particular solution (−18. . then so is x = m + 15q and y = n − 22q. (b) Find all solutions to the congruence Solution: There is no solution. We have x = 13 + 17q for some q ∈ Z. 11x ≡ 7 (mod 15). 44x ≡ 28 (mod 15). and 3x ≡ 2 (mod 13). so 40q ≡ 50 (mod 13).42 27. 5x ≡ 14 (mod 17) 3x ≡ 2 (mod 13) . 32. so adding any multiple of the vector (15. (a) Find all solutions to the congruence 55x ≡ 35 (mod 75). 1 0 110 1 −1 35 1 −1 35 15 −22 0 . x ≡ 13 + 17 · 11 ≡ 200 (mod 221). Thus we have x ≡ 2. 27x ≡ 18 (mod 13). or q ≡ 11 (mod 13). This reduces to 4q ≡ 5 (mod 13). y = 5 is a solution. 30. 75) = 5 is not a divisor of 36. 29. Solution: Convert the second congruence to the equation x = 4 + 10q for some q ∈ Z. and then 13 + 17q ≡ 5 (mod 13). 17. which is a divisor of 35. x ≡ 13 (mod 17) x ≡ 5 (mod 13). 35x ≡ 98 (mod 17). 28. which reduces to q ≡ 7 (mod 9). then 110(m + 15q) + 75(n − 22q) = 45 + 110(15)q + 75(−22)q = 45. This leads to the answer. 47. since 110(15) − 75(22) = 0.

CHAPTER 1 SOLUTIONS 31. and (±6)2 ≡ −3 (mod 13). 5. and 6. Since gcd(4. and 54 ≡ 52 ≡ 5 (mod 10). Just computing the squares modulo 13 gives us (±1)2 = 1. or 6 as its units digit. which reduces to 7q ≡ 7 (mod 25). We have done the computation for representatives of each congruence class. so q ≡ 1 (mod 15). This is the clue to take m = n = 2. 5. Solution: In the example the integers m and n cannot be relatively prime. Find all integers n for which 13 | 4(n2 + 1). for all positive integers n. Solution: This is equivalent solving the congruence 4(n2 + 1) ≡ 0 (mod 13). b. 2200 ) = 1. (±2)4 = 16 ≡ 6 (mod 10). Solution: The last digit is the remainder when divided by 10. 36. 1. 32. and substitute to get 23 + 32q ≡ 5 (mod 25). Or you can write 2200 as the prime factorization. and then 62 ≡ 6 (mod 10). 34. but a more elegant proof can be given by simply observing that 10n+1 + 4 · 10n + 4 ≡ 0 (mod 9) since 10 ≡ 1 (mod 9). 13) = 1. with a = 1 and b = 0. Thus we must compute the congruence class of 4100 (mod 10). 33. (±3)4 = 81 ≡ 1 (mod 10). Solve the system of congruences x ≡ 5 (mod 25) x ≡ 23 (mod 32) . Give integers a. 1. Solution: This could be proved by induction. This gives x ≡ 55 (mod 25 · 32). it is really asking to ﬁnd n4 (mod 10). We have 42 ≡ 6 (mod 10). m. (±5)2 ≡ −1 (mod 13). Prove that the fourth power of an integer can only have 0. (±2)2 = 4. (±4)4 ≡ 62 ≡ 6 (mod 10). All we need to do is to compute the fourth power of each congruence class modulo 10: 04 = 0. Prove that 10n+1 + 4 · 10n + 4 is divisible by 9. x ≡ b (mod n) . (±4)2 ≡ 3 (mod 13). so the answer to the original question is x ≡ ±5 (mod 13). Thus 4100 = (42 )50 ≡ 650 ≡ 6 (mod 10). 43 Solution: Write x = 23 + 32q for some q ∈ Z. to get n2 ≡ −1 (mod 13). n to provide an example of a system x ≡ a (mod m) that has no solution. and then (3. we can cancel 4. (b) Is 4100 divisible by 3? Solution: No. This shows that the only possible units digits for n4 are 0. Solution: Since the question deals with the units digit of n4 . (±1)4 = 1. since 4100 ≡ 1100 ≡ 1 (mod 3). (a) Compute the last digit in the decimal expansion of 4100 . 35. (±3)2 = 9.

Comment: If ab ≡ 1 (mod n). since gcd(a. 20 20 20 [13]−1 = [17]20 . Since [12]−1 = [−1]−1 = [−1]13 = [12]13 . 13 13 13 [5]−1 = [8]13 . Comment: We know that Zn has ϕ(n) units. and [6]13 and [−2]13 . n) = 1. In Z20 : ﬁnd all units (list the multiplicative inverse of each). 31 651 At the next step. Using 12. while 12. Solution: Since 6 ≡ −1 (mod 7). then [a]n and [b]n are inverses. as are [−a]n and [b]n . Find the multiplicative inverse of each nonzero element of Z7 . 1 −1 682 −4 5 651 1 0 . 1 −1 0 1 682 3379 . 2 −11 −1 6 1 45 Thus [91]−1 = [−11]501 = [490]501 . then [a]n and [−b]n are inverses. 31. 7. Solution: Note that 14. This helps to check your list. and 40 are congruent to 1. as are the pairs [4]13 and [−3]13 . Furthermore. 1 0 0 1 501 91 . We have [3]−1 = [7]20 . 31 | 651. if possible (in Z× ). this 13 13 13 13 takes care of all of the nonzero elements of Z13 . n) = 1 if and only if gcd(n − a. 20 20 . and 19. 9. 501 33. Using 40. 1 −5 46 −1 6 45 . and 3 · 5 = 15 ≡ 1 (mod 7). 3379) = 31. Solution: The units of Z20 are the equivalence classes represented by 1. and [19]−1 = [19]20 . 4061 4061 Solution: The inverse does not exist. Find [91]−1 . Finally. 1 −5 46 0 1 91 . [4]−1 = [10]13 . If ab ≡ −1 (mod n). here is the list of inverses: [2]−1 = [7]13 . [6]−1 = [11]13 . [11]−1 = [11]20 . Find [3379]−1 . ﬁnd all nilpotent elements. They occur in pairs. [3]−1 = [9]13 . 3. Using 14. 17. and we see that [3]13 and [−4]13 are inverses. and 39 are congruent to −1.44 CHAPTER 1 SOLUTIONS 1. and [3]7 and [5]7 are inverses of each other. Find the multiplicative inverse of each nonzero element of Z13 . and look at the various ways to factor them. so [2]7 and [4]7 are inverses of each other. 32. It is useful to list the integers with m with m ≡ ±1 (mod n). 11. ﬁnd all idempotent elements. 34. 501 501 Solution: We need to use the Euclidean algorithm. we see that [5]13 and [8]13 are inverses. 2 · 4 = 8 ≡ 1 (mod 7). 0 1 4061 3379 5 −6 −4 5 . we see that [2]13 and [7]13 are inverses.4 SOLUTIONS 30. 27. and so (4061. if possible (in Z× ). [9]−1 = [9]20 . as are [−a]n and [−b]n . the class [6]7 is its own inverse. 25. 13.

and so each element is its own inverse. 35. They are [0]20 . and [3]8 = [16] = [−1]. for 0 ≤ i < 4 and 0 ≤ j < 6. and [−4]6 = [−4][−9] = [1]. so the next possibility to check is 8. [1]20 . We have [3]2 = [9]. and 23. Are these all diﬀerent? Suppose that [8]i [−4]j = [8]m [−4]n . We next try [3]. Solution: The units of Z24 are the equivalence classes represented by 1. so the multiplicative order of [8] is 4. 17. Since [2]8 = [−1]2 = [1]. 5. we need to ﬁnd an element whose 17 multiplicative order is 16. If you want a more systematic approach. Solution: We ﬁrst compute the powers of [8]: [8]2 = [−6]. [2]3 = [8]. with gcd(b. 19. Show that Z× is not cyclic but that each element has the form [8]i [−4]j . and we are done. ﬁnd all idempotent elements. ﬁnd all nilpotent elements. for some 0 ≤ i < 4 and 0 ≤ j < 6 and 0 ≤ m < 4 and 0 ≤ n < 6. [16]24 . so the multiplicative order of [−4] is 6. 13. [1]24 . [8]3 = [8][−6] = [−13]. or by using Problem 1. j. [5]20 . but there is no known algorithm for p actually ﬁnding the one element whose powers cover all of Z× . p Solution: We begin by trying [2]. 35 35 35 for some positive integers i. The idempotent elements are [0]24 . 7. Then [8]i−m = [−4]n−j . this forces i = m and n = j. and [2]4 = [16] = [−1]. It is known than if p is prime. it follows that [2] has multiplicative order 8.40. [6]24 . [−4]3 = [−4][16] = [6]. and [16]20 . .4. The nilpotent elements of Z20 can be found by using trial and error. and the powers we have listed represent the only possible values of [8]i .4. c) = 1. [3]4 = [81] = [−4]. then any solution to the congruences x ≡ 1 (mod b) and x ≡ 0 (mod c) will be idempotent modulo n. and the nilpotent elements are [0]24 . The solution just uses trial and error. For each of these numbers we have x2 ≡ 1 (mod 24). [9]24 . so [3] does in fact have multiplicative order 16. There are 24 possible products of the form [8]i [−4]j . [−4]4 = [−4][6] = [11]. 17 Comment: To show that Z× is cyclic. 36. Show that Z× is cyclic. then Z× is cyclic. In Z24 : ﬁnd all units (list the multiplicative inverse of each). They are [0]20 and [10]20 . [−4]5 = [−4][11] = [−9]. [18]24 . Problem 39 shows that the multiplicative order of an element has to be a divisor of 16. 11. and since the only power of [8] that is equal to a power of [−4] is [1] (as shown by our computations). [12]24 . The only divisor of 16 that is left is 16 itself. We have [2]2 = [4].13 of the text: if n = bc. and [8]4 = [−6]2 = [1]. We next compute the powers of [−4]: [−4]2 = [16]. 37.CHAPTER 1 SOLUTIONS 45 The idempotent elements of Z20 can be found by using trial and error. you can use a the hint in Exercise 1.

and this completes the proof. and so we can apply Theorem 1. Thus ar = aϕ(n)−kq = aϕ(n) (ak )−q ≡ 1 (mod n). and so the solution is [x] = [−3] or [x] = [2]. If n = pα1 pα2 · · · pαt is the prime factorization of n. n) = 1. for any integer n. Solution: First assume that each prime divisor of n is a divisor of a. Prove that if k is the smallest positive integer for which ak ≡ 1 (mod n). which shows that aϕ(n) ≡ 1 (mod n). 39. We can use the division algorithm to write ϕ(n) = qk + r. and let a ∈ Z with gcd(a. if [a]n is nilpotent. Solve the equation [x]2 + [x]11 − [6]11 = [0]11 . Let n be a positive integer. 38.11. showing that Z35 is not cyclic. Solution: Assume that k is the smallest positive integer for which ak ≡ 1 (mod n).6 implies that either [x] + [3] = [0] or [x] − [2] = [0]. 49 −65 −3 4 1 0 0 117 0 1 7605 5733 . and q ∈ Z. then it must be a divisor of a. For ω = − + 2 2 . Finally. and so [a]k = [0]k .46 CHAPTER 1 SOLUTIONS We conclude that since there are 24 elements of the form [8]i [−4]j . Since ak ≡ 1 (mod n). we know that gcd(a. ([8]i [−4]j )12 = ([8]4 )3i ([−4]6 )2j = [1]. then n | ak . so no element of Z35 has multiplicative order 24. √ 3 1 i. with [a]k = [0]. Find gcd(7605. where 0 ≤ r < k. 11 Solution: We can factor [x]2 + [x] − [6] = ([x] + [3])([x] − [2]). prove that ω n = 1 if and only if 3|n. But if a prime p is a divisor of ak .4. SOLUTIONS TO THE REVIEW PROBLEMS 1. where 0 ≤ βj ≤ αj for all j. . then k | ϕ(n). and express it as a linear combination of 7605 and 5733. so each prime n divisor of n is a divisor of ak . every element in Z35 must be of this form. and 117 = (−3) · 7605 + 4 · 5733. Thus gcd(7605. 5733) = 117. Prove that [a]n is a nilpotent element of Zn if and only if each prime divisor of n is a divisor of a. Solution: Use the matrix form of the Euclidean algorithm: 1 −1 1872 0 1 5733 . 2. 1 −1 1872 −3 4 117 .4. 5733). n) = 1. If k is the smallest positive t 1 2 integer such that kβi ≥ αi for all i. then we must have a = t 1 2 pβ1 pβ2 · · · pβt d. n Conversely. Corollary 1. then n | ak . 40. so we must have r = 0 since r < k and k is the smallest positive integer with ak ≡ 1 (mod n).

To compute the multiplicative order of [8]. and ω 3 = 1. and 3|n. so [7] has multiplicative order 4 because [7]4 = [4]2 = [1]. To solve the congruence 3x ≡ 21 (mod 25) we could ﬁnd the multiplicative inverse of 3 modulo 25. 200) = 8. since 3k must be a multiple of 4. [4]. and [2]4 = [1]. By Problem 39. Trial and error shows it to be −8. so the congruence has a solution. [8]. 24x 3x −24x x ≡ 168 ≡ 21 ≡ −168 ≡ 7 (mod (mod (mod (mod 200) 25) 25) 25) The solution is x ≡ 7. and 15 ﬁnd its multiplicative order. Computing powers. ﬁnd its multiplicative inverse. and substitute to get 16 + 22q ≡ 9 (mod 15). Then 1 = ω n = ω 3q+r = (ω 3 )q ω r = ω r . and 8 | 168. List the elements of Z× . Since r = 0. and proceed with the solution. the only possibility is r = 0. The next step is to reduce the congruence by dividing each term by 8. but that the multiplicative order of [4] is 2. 157. and [13] = [−2] has multiplicative order 4. If n ∈ Z. 107. This shows not only that the multiplicative order of [2] is 4. 82. the multiplicative order of any nontrivial element is 2. [7]. This gives x ≡ −3 (mod 11 · 15). we have [2]2 = [4]. we have [7]2 = [4]. Solve the system of congruences 2x ≡ 9 (mod 15) x ≡ 8 (mod 11) . and that [11]−1 = [−4]−1 = [−4] = [11]. if n ∈ Z and ω n = 1. 4. [2]3 = [8]. we can rewrite it as [2]3 .) Similarly. that [13]−1 = [−2]−1 = [−8] = [7]. 132.CHAPTER 1 SOLUTIONS 47 Solution: Calculations in the introduction to Chapter 1 show that ω 2 = √ 1 3 − − i. Solve the congruence 24x ≡ 168 (mod 200). (This can also be shown by rewriting [8] as [−7]. and [14]. 57. then n = 3q for some q ∈ Z. 2 and we have shown that ω = 1 and ω 2 = 1. Solution: There should be 8 elements since ϕ(15) = 8. [11] = [−4] has multiplicative order 2. which reduces to 7q ≡ −7 (mod 15). Solution: Write x = 8 + 11q for some q ∈ Z. The same computation shows that [2]−1 = [8] and [4]−1 = [4]. [13]. The elements are [1]. where the remainder satisﬁes 0 ≤ r < 3. Conversely. . and therefore 3|n. use the division algorithm to write n = q · 3 + r. we can multiply both sides of the congruence by −8. We can also deduce that [13] = [−2] has multiplicative order 4. and then it is clear that the ﬁrst positive integer k with ([2]3 )k = [1] is k = 4. Next. so q ≡ −1 (mod 15). 3. [2]. 1. 32. which gives 24x ≡ 168 (mod 200). [11]. 4. Then 2 2 ω n = ω 3q = (ω 3 )q = 1q = 1. 182 (mod 200). 5. For each element. Solution: First we ﬁnd that gcd(24. or 8.

Show that if n > 1 is an odd integer. the integers n and 2n are relatively prime.8 shows that to compute ϕ(2n) in terms of ϕ(n) 1 we need to add 2 · (1 − 2 ). and so it follows from Exercise 1. Second solution: Since n is odd. .48 CHAPTER 1 SOLUTIONS 6. and this does not change the computation.4. The formula in Proposition 1. then ϕ(2n) = ϕ(n).27 of the text that ϕ(2n) = ϕ(2)ϕ(n) = ϕ(n). Solution: Since n is odd.4. the prime 2 does not occur in its prime factorization.

Solution: If a horizontal line intersects the graph of the function more than once. y2 ) line on its graph. a function is one-to-one if f (x1 ) = f (x2 ) implies x1 = x2 .4. According to Deﬁnition 2.1.Chapter 2 Functions 2. Explain why this agrees with Deﬁnition 2.1. Solution: We assume that the x-axis is the domain and the y-axis is the codomain of the function that is to be deﬁned by the given curve. Thus if we apply Deﬁnition 2. The “Vertical Line Test” from calculus says that a curve in the xy-plane is the graph of a function of x if and only if no vertical line intersects the curve more than once. then there will be points (x1 . The “Horizontal Line Test” from calculus says that a function is one-to-one if and only if no horizontal line intersects its graph more than once.4. Equivalently.4.1.1. the uniqueness part of the deﬁnition translates directly into the “vertical line test”.1. y1 ) and (x2 . y1 ) and (x2 .1 to the given curve. y2 ) on the curve with x1 = x2 and y1 = y2 . y1 ) and (x2 . 21. the “horizontal line test” is exactly the same as the condition given in Deﬁnition 2. According to Deﬁnition 2. Explain why this agrees with Deﬁnition 2. then the points of intersection represent points (x1 .1. If a vertical line intersects the curve in two distinct points.1 SOLUTIONS 20. more than one 49 . if (x1 .1. a subset of the plane deﬁnes a function if for each element x in the domain there is a unique element y in the codomain such that (x. then we cannot have y1 = y2 while x1 = x2 .1. y) belongs to the subset of the plane. In this context. y2 ) for which x1 = x2 but y1 = y2 .

For any real number x. which states that if A is an n × n matrix. This shows that g = f −1 . then x = f −1 (y). a) has slope −1. and so det(A) = 0. Explain why this agrees with Deﬁnition 2. First. Thus A is an invertible matrix. and this is the image of the transformation L. where I is the n × n identity matrix. then A is invertible. and so f −1 (y) = f −1 (f (x)) = x. a). and therefore y = f (f −1 (y)) = f (x). b) in the line y = x is the point (b. then the rank of A plus the nullity of A is n. which must satisfy the conditions BA = I. In calculus the graph of an inverse function f −1 is obtained by reﬂecting the graph of f about the line y = x. This shows that the point (x. Solution: We ﬁrst note that the reﬂection of a point (a. b) and (b. if y = f (x) then we have g(f (x)) = g(y) = x and for any real number y we have f (g(y)) = f (x) = y. then it is invertible. if det(A) = 0. Conversely. which shows that (y. On the other hand. then y = f (x).50 CHAPTER 2 SOLUTIONS 22. y) lies on the graph of f . for all x ∈ Rn . and so L has an inverse. On the other hand. and AB = I. and so f has an inverse. the nullity of A is the dimension of the solution space of the equation Ax = 0.7.1. y) lies on the graph of f −1 . and the point (x. If f : R → R has an inverse. Let A be an n × n matrix with entries in R. Solution: I need to assume that you know that a square matrix A is invertible if and only if det(A) = 0. and L is one-to-one if and only if the nullity of A is zero. so L is onto if and only if A has rank n. if L has an inverse. Since the matrix A is invertible if and only if it has rank n. . since Ax1 = Ax2 if and only if A(x1 − x2 ) = 0. suppose that the graph of the function g is deﬁned by reﬂecting the graph of f in the line y = x. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax. then it can also be described by multiplication by a matrix B. To prove part (b) we need to use the Rank–Nullity Theorem. 23. it follows that L is invertible if and only if L is onto. y) lies on the graph of f −1 . where x = g(y). (b) Show that if L is either one-to-one or onto. and then the Rank–Nullity Theorem shows that this happens if and only if L is one-to-one. and that this line segment intersects the line y = x at the midpoint ((a + b)/2. for all x ∈ Rn . Solution: The rank of the matrix A is the dimension of the column space of A. On the other hand. x) lies on the graph of f . deﬁned by L−1 (x) = A−1 x. if (x. which makes it perpendicular to the line y = x. (a + b)/2) of the segment. (a) Show that L is an invertible function if and only if det(A) = 0. This can be seen by observing that the line segment joining (a.

we can deﬁne ψ : Z× → Z× by 17 17 17 ψ(x) = a−1 x. Solution: If det(AT A) = 0. and assume that m > n. ﬁnd the 17 inverse function θ−1 . On the set {(a. Note: A vector x is called an eigenvector of A if it is nonzero and there exists a scalar λ such a that Ax = λx. Deﬁne the function θ : Z× → Z× by 17 17 17 θ(x) = ax. Is θ one to one? Is θ onto? If possible. and so det(AT A) = 0 if and only if the column vectors of A are linearly independent. Ax1 = Ax2 if and only if A(x1 − x2 ) = 0.CHAPTER 2 SOLUTIONS 51 24. for all x ∈ Z× . y2 ) if x1 y2 = x2 y1 . Show that L is a one-to-one function if det(AT A) = 0. 25. It then follows from Exercise 17 that L is one-toone. y1 ) ∼ (x2 . L is a one-to-one linear transformation if and only if det(AT A) = 0. 26. for all x ∈ Rn . where AT is the transpose of A. then AT A is an invertible matrix. Let a be a ﬁxed element of Z× . for all x ∈ Rm . Solution: Since a has an inverse in Z× . In other words. According to the Rank–Nullity Theorem. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax. Let A be an m × n matrix with entries in R. Then ψ(θ(x)) = ψ(ax) = a−1 (ax) = (a−1 a)x = 17 x and θ(ψ(x)) = θ(a−1 x) = a(a−1 x) = (aa−1 )x = x. Comment: There is a stronger result that depends on knowing a little more linear algebra. If we deﬁne K : Rm → Rn by K(x) = (AT A)−1 AT x. . which shows that ψ = θ−1 . deﬁne (x1 . 2. Let A be an n × n matrix with entries in R. Solution: As noted in the solution to problem 23. which translates into the given statement about eigenvalues of A. this happens if and only if the nullity of A is zero. This implies that θ is one-to-one and onto. b)} of all ordered pairs of positive integers. Show that this deﬁnes an equivalence relation. In some linear algebra courses it is proved that det(AT A) gives the n-dimensional “content” of the parallepiped deﬁned by the column vectors of A. then KL is the identity function on Rm . and so L is one-to-one if and only if Ax = 0 for all nonzero vectors x. This is equivalent to the statement that there is no nonzero vector x for which Ax = 0 · x. for all x ∈ Rn . for all x ∈ Z× . This content is nonzero if and only if the vectors are linearly independent. Prove that L is one-to-one if and only if no eigenvalue of A is zero.2 SOLUTIONS 14. Deﬁne a linear transformation L : Rn → Rm by L(x) = Ax.

we verify the transitive law. b1 ). 15. b) ∼ (a. It follows that the equivalence classes of ∼ are the planes in R3 that are perpendicular to u. and transitive laws for the relation ∼ really depend on an equality. x + y + z) . For the linear transformation L : R3 → R3 deﬁned by L(x. a3 ) = L(b1 . b2 . deﬁne v ∼ w if v ·u = w ·u. for all (x. a ∼ b if and only if |a| = |b|. a2 . b2 ) ∼ (a3 . z) ∈ R3 . For the function f : R → R deﬁned by f (x) = x2 . z) = (x + y + z. we have the equations a1 b2 = a2 b1 and a2 b3 = a3 b2 . a2 . Given an ordered pair (a. for all x ∈ R. b1 ) ∼ (a2 . we have ab = ba. describe the equivalence relation on R that is determined by f . Solution: The reﬂexive. Given (a1 . deﬁne z1 ∼ z2 if ||z1 || = ||z2 ||. and transitive laws can be easily veriﬁed since ∼ is deﬁned in terms of an equality. b2 . b1 ) ∼ (a3 . b1 ) and (a2 . b). 1). give a geometric description of the partition of R3 that is determined by L. (a2 . b2 ). b2 ). symmetric. Thus two vectors are equivalent if and only if they lie in the same plane perpendicular to u. and can easily be veriﬁed. b1 ) ∼ (a2 . Solution: The equivalence relation determined by f is deﬁned by setting a ∼ b if f (a) = f (b). v · u represents the length of the projection of v onto the line determined by u. {(x. Let u be a ﬁxed vector in R3 . where · denotes the standard dot product. If we multiply the ﬁrst equation by b3 and the second equation by b1 . y. b3 ) if L(a1 . b3 ). Solution: The reﬂexive. and assume that u has length 1. we get a1 b2 b3 = a2 b1 b3 = a3 b1 b2 . y. so a ∼ b if and only if a2 = b2 .52 CHAPTER 2 SOLUTIONS Solution: We ﬁrst show that the reﬂexive law holds. 0)} is the plane through the origin whose equation is x + y + z = 0. We next check the symmetric law. b3 ) with (a1 . b2 ) with (a1 . b3 ). Finally. 16. 0. 17. Since b2 = 0 we can cancel to obtain a1 b3 = a3 b1 . 18. Solution: Since (a1 . or. with normal vector (1. and so a2 b1 = a1 b2 . Show that ∼ is an equivalence relation. showing that (a1 . and equality is itself an equivalence relation. x + y + z. For vectors v and w. it follows from the deﬁnition of L that (a1 . b1 ). For example. we have a1 b2 = a2 b1 . symmetric. b). 1. and so (a. z) = (0. b2 ) and (a2 . Given (a1 . The other subsets in the partition of R3 deﬁned by L are planes . a2 . b2 . and (a3 . b3 ) if and only if a1 + a2 + a3 = b1 + b2 + b3 . b3 ). a3 ) ∼ (b1 . b2 ) ∼ (a1 . a3 ) ∼ (b1 . which shows that (a2 . On the set C of complex numbers. and give a geometric description of the equivalence classes of ∼. y. y. Since u has length 1. z) | L(x. Show that ∼ is an equivalence relation.

[±3]2 = [9]12 . The corresponding equivalence classes determined by f are {[0]12 . [±4]12 }. 5. 12 Show that the formula f deﬁnes a function. and so we get A = P −1 B(P −1 )−1 . and [6]2 = [0]12 . Deﬁne the formula f : Z12 → Z12 by f ([x]12 ) = [x]2 . 9)(3. Solution: We have A ∼ A since IAI −1 = A. 4. τ σ = (1. 4. 5). 8)(2. [4]12 . then P AP −1 = B and QBQ−1 = C for some P. 4. 2. 9. 5. 5)(3. σ −1 = (1. Q. 4] = 12. 3)(4. τ −1 . 7. [6]12 }. στ = (1. 5. write each of these permutations as a 1 5 4 7 2 6 8 9 3 product of disjoint cycles: σ. We have σ −1 = (1. For the permutations σ = τ= 1 2 2 3 5 1 4 5 8 3 6 7 6 4 8 9 7 9 and 1 2 3 4 5 6 7 8 9 . 2. If A ∼ B and B ∼ C. {[±2]12 . 2. 9. 4. What is the order of σ? Is σ an even permutation? Compute σ −1 . Find the image of f and the set Z12 /f of equivalence classes determined by f . Substituting gives Q(P AP −1 )Q−1 = (QP )A(QP )−1 = C. where I is the n × n identity matrix. 9).3 SOLUTIONS 13. 4)(7. 5)(3. 7.CHAPTER 2 SOLUTIONS 53 parallel to this one. τ = (2. 7)(2. 8. 3. If A ∼ B. 1). τ στ −1 . {[±3]12 }. τ −1 = (2. [9]12 }. It is an even permutation. 8). 7. 4. τ . [±4]2 = [4]12 . deﬁne A ∼ B if there exists an invertible matrix P such that P AP −1 = B. 7. which shows that f ([x1 ]12 ) = 1 2 f ([x2 ]12 ). 5)(3. Thus the partition consists of the planes perpendicular to the vector (1. Solution: The formula for f is well-deﬁned since if [x1 ]12 = [x2 ]12 . [1]12 . {[±1]12 . 2. and so x2 ≡ x2 (mod 12). 19. 8). 1. 8. 8. write σ as a 7 5 6 9 2 4 8 1 3 product of disjoint cycles. τ στ −1 = (1. Solution: We have σ = (1. On the set of all n × n matrices over R. then P AP −1 = B for some invertible matrix P . for all [x]12 ∈ Z12 . 7. 9). 8. 14. and so A ∼ C. τ σ. To compute the images of f we have [0]2 = [0]12 . 9). and so its order is 12 since lcm[3. For the permutation σ = 1 2 3 4 5 6 7 8 9 . στ . 3. 2. 2)(4. Thus 12 12 12 12 f (Z12 ) = {[0]12 . [±5]2 = [1]12 . then x1 ≡ x2 (mod 12). . 9. 5)(3. 8. 9. [±5]12 }. since it can be expressed as the product of 6 transpositions. Solution: σ = (1. 8. 5. [±1]2 = [1]12 . 7). Check that ∼ deﬁnes an equivalence relation. στ σ −1 = (1. 6). στ σ −1 . σ −1 . 20. 3). [±2]2 = 12 12 12 [4]12 . 4). 6.

it has order 6. 2. 9) has order 14. 8. since 11 and 13 are prime. 5. 9)(2. Prove that τ Gτ −1 = {σ ∈ Sym(S) | σ = τ γτ for some γ ∈ G} is a group of permutations. 6. 20. Show that S10 has elements of order 10. 7. Let σ = (2. so the cycle structure of στ σ −1 is the same as that of τ . 3. 12. 6)(3. 8. and there are no cycles of that length in S10 . 11. 6. Write σ as a product of disjoint cycles. 8. Let τ be a ﬁxed permutation in Sym(S). 17. 3. 18. 5. 5. 6 7 8 9 8 9 10 1 10 3 11 5 . 6. and (1. 4. with G ⊆ Sym(S). 7). compute the order of στ σ −1 . 7. )(6. and thus στ σ −1 has order 6. 4. Solution: Assume that τ ∈ Sn has order m. and then τ k = σ −1 σ = (1). 7. for the set S. 7)(1. 10). 7. so it has order 15 = lcm[5. 7. 9)(8. Prove that G is a group of permutations. For Solution: Since τ = (1. 9. Solution: The permutation (1. and 14. since (στ σ −1 )k = (1) implies στ k σ −1 = (1). then στ σ −1 has order m. 3)(4. 6. 9)(3. 19. the order of στ σ −1 cannot be less than n. 10)(3. 9)(3. 5). 16. 3. 2. 8. 7. 11. 10). for any permutation σ ∈ Sn . 8)(2. any element of order 11 or 13 would have to be a cycle. Let G be a group of permutations. We have στ σ −1 = (3. but not 11 or 13. 9)(1. Let G = {σ ∈ Sym(S) | σ(X) ⊂ X}. 8. Solution: We have σ = (1. On the other hand. 8. 8. Let S be a set. On the other hand. 2)(3. 5. 8) = (1. 5.54 CHAPTER 2 SOLUTIONS 15. 6) ∈ S9 . 5. 7). 11. and σ −1 = (1. 6. 3]. 7. What is the order of σ? Compute σ −1 . 9. . Compute the order of τ = 1 2 3 4 5 7 2 11 4 6 σ = (3. 7) has order 12. and let X be a subset of S. 4. 5. It follows from the identity (στ σ −1 )k = στ k σ −1 that (στ σ −1 )m = στ m σ −1 = σ(1)σ −1 = (1). 6. 7) has order 10. 6. while the element (1. 6. 5. Prove that if τ ∈ Sn is a permutation with order m. 4. 2)(3.

for all σ ∈ Sn . Deﬁne the function φ : Z× → Z× by φ(x) = x−1 . Is φ one to 17 17 17 one? Is φ onto? If possible. 3. 5. On the set Q of rational numbers. 4. for all x ∈ Z× . Show that f is a one-to-one function. 9. Solution: If φα (σ) = φα (τ ). 2. Finally. φα ((1. Hint: Use the derivative of f to show that f is a strictly increasing function. 2. for all x ∈ R. and use this to ﬁnd its order and its inverse. A short computation shows that (φα )−1 = φα−1 . (a) Let α be a ﬁxed element of Sn . 2. 2. 6). for σ. we have φα (σ) = τ for σ = α−1 τ α. 3)) = (1. 9). to get σ = τ . 7. and φα ((1. direct computations show that φα ((1)) = (1). Show that φα : Sn → Sn deﬁned by φα (σ) = ασα−1 . (b) In S3 . so σ has order 6. β = (1. Another way to show that φα is one-to-one and onto is to show that it has an inverse function. 2)) = (1. φα ((2. and γ = (1. 2. 2) is its own inverse. 9). Since σ has length 6. We can multiply on the left by α−1 and on the right by α. 3). 2. For σ = αβγ. 2). write σ as a product of disjoint cycles. deﬁne x ∼ y if x − y is an integer. Solution: For all x ∈ Z× we have φ(φ(x)) = φ(x−1 ) = (x−1 )−1 = x. 3). 3. 5. 2). is a one-to-one and onto function. τ ∈ Sn . 5. Is σ even or odd? Solution: We have σ = (1. 6. 7. For the function f : R → R deﬁned by f (x) = x2 . 2). and σ −1 = (1. which also shows that φ is one-to-one and onto. 3)) = (1. Solution: Since (1. so 17 φ = φ−1 . 2)) = (1. 6. so it is an odd permutation. Show that ∼ is an equivalence relation. . 7. In S10 . it can be written as a product of 5 transpositions. φα ((1. 3). for all x ∈ R. Deﬁne f : R → R by f (x) = x3 + 3xz − 5. so φα is one-to-one. 3. 3. let α = (1. Compute φα . let α = (1. ﬁnd the inverse function φ−1 . describe the equivalence relation on R that is determined by f . 6). 3). and so φα is onto.CHAPTER 2 SOLUTIONS 55 SOLUTIONS TO THE REVIEW PROBLEMS 1. 3)) = (2. 3. 3. given τ ∈ Sn . then ασα−1 = ατ α−1 . φα ((1. 2.

56 CHAPTER 2 SOLUTIONS .

This means that the operation could not have an identity element. The operation is associative since a ∗ (b ∗ c) = a ∗ bc 2 57 = a bc 2 2 = a(bc) 4 . Even if we were to exclude the zero vector we would still have problems. z1 ) and (x2 . 23. On the set G = Q× of nonzero rational numbers. z2 ) = (y1 z2 − z1 y2 . so the cross product cannot be used to make the set of all vectors in R3 into a group. y2 . then ab is a nonzero ab rational number. x1 y2 − y1 x2 ). Is R3 a group under this multiplication? Solution: The cross product of the zero vector and any other vector is the zero vector. Show that G is a group under this multiplication. y1 . for all a. 2 Solution: If a and b are nonzero rational numbers. not a vector. For vectors (x1 . z1 )×(x2 . again making it impossible to deﬁne a group structure. z1 x2 − x1 z2 . Use the dot product to deﬁne a multiplication on R3 . deﬁne a new multiplication ab by a∗b = .Chapter 3 Groups 3. z2 ) in R3 . showing that the operation is closed on the set 2 G. 24. the cross product is deﬁned by (x1 . The cross product of two nonzero vectors deﬁnes a vector that is perpendicular to each of the given vectors. Does this make R3 into a group? Solution: The dot product of two vectors is a scalar.1 SOLUTIONS 22. b ∈ G. This means that the dot product does not even deﬁne a binary operation on the set of vectors in R3 . y2 . y1 . and so is .

We will 15 write the elements as {1. 25. and if a is nonzero. · 1 2 4 8 7 5 1 1 2 4 8 7 5 2 2 4 8 7 5 1 4 4 8 7 5 1 2 8 8 7 5 1 2 4 7 7 5 1 2 4 8 5 5 1 2 4 8 7 Each element in the group is a power of 2. 15 Solution: Z× = {[1]15 . 2. −7.58 and (a ∗ b) ∗ c = ab 2 ∗c= CHAPTER 3 SOLUTIONS ab 2 c 2 = (ab)c . as successive powers of 2. and the second table shows what happens when we arrange the elements in order. [13]15 . −1}. · 1 -1 2 -2 4 -4 7 -7 1 1 -1 2 -2 4 -4 7 -7 -1 -1 1 -2 2 -4 4 -7 7 2 2 -2 4 -4 -7 7 -1 1 -2 -2 2 -4 4 7 -7 1 -1 4 4 -4 -7 7 1 -1 -2 2 -4 -4 4 7 -7 -1 1 2 -2 7 7 -7 -1 1 -2 2 4 -4 -7 -7 7 1 -1 2 -2 -4 4 . [2]15 . Write out the multiplication table for Z× . 9 · 1 2 4 5 7 8 1 1 2 4 5 7 8 2 2 4 8 1 5 7 4 4 8 7 2 1 5 5 5 1 2 7 8 4 7 7 5 1 8 4 2 8 8 7 5 4 2 1 Comment: Rewriting the table. [14]15 }. We will write m for [m]9 . with the elements in a slightly diﬀerent order. then a is a nonzero rational number that serves as the multiplicative inverse of a. 4. 7. gives a diﬀerent picture of the group. Write out the multiplication table for Z× . [7]9 . [4]15 . −2. 9 Solution: Z× = {[1]9 . [2]9 . [8]9 }. [7]15 . [11]15 . [4]9 . 26. 4 4 The number 2 acts as the multiplicative identity. [5]9 . [8]15 . −4.

Solution: To give a careful proof we need to use induction.CHAPTER 3 SOLUTIONS 59 Comment: Notice how much easier it makes it to use the representatives {±1. which is certainly true. Show that (aba−1 )n = abn a−1 . (aba−1 )k+1 = = = = = (aba−1 )k (aba−1 ) (abk a−1 )(aba−1 ) (abk )(a−1 a)(ba−1 ) (abk )(ba−1 ) abk+1 a−1 . b(ab)b−1 = (ab2 )b−1 (ba)(bb−1 ) = (ab)(bb−1 ) ba = ab This completes the proof. a−1 ((ab)(ab)) (a−1 (ab))(ab) ((a−1 a)b))(ab) (eb)(ab) b(ab) = = = = = a−1 (a2 b2 ) (a−1 a2 )b2 ((a−1 a)a)b2 (ea)b2 ab2 The next step is to multiply on the right by b−1 . Since G is a group. Show that if (ab)2 = a2 b2 . 28. Let G be a group. denoted by a−1 and b−1 . ±7} when listing the congruence classes in the group. then ba = ab. so we can multiply both sides of the equation on the left by a−1 without destroying the equality. The statement for n = 1 is simply that aba−1 = aba−1 . Expanding this equation gives us (ab)(ab) = a2 b2 . 27. since we have shown that if (ab)2 = a2 b2 . and suppose that a and b are any elements of G. Multiplication in G is well-deﬁned. respectively. then ba = ab. we have to include the following steps. we have the following calculation. and suppose that a and b are any elements of G. ±4. The associative law for multiplication essentially says that parentheses don’t matter. Now assume that the result holds for n = k. both a and b have inverses. for any positive integer n. so we don’t really need to include all of the steps we showed before. Using this induction hypothesis. Let G be a group. ±2. If we are to be precise about using the associative law. Solution: Assume that a and b are elements of G for which (ab)2 = a2 b2 .

The previous exercise shows that in the deﬁnition of a group it is suﬃcient to require the existence of a left identity element and the existence of left inverses. so we have deﬁned a binary operation on G. for all a. We will ﬁrst show that a · e = a.3 of the text. and since there exists an element a ∈ G with a · a = e.60 CHAPTER 3 SOLUTIONS Thus the statement holds for n = k + 1. even though it has a left identity element and right inverses. 30. 1/|a| is a right inverse for any a ∈ G. 31. replace condition (iii) with the condition that there exists e ∈ G such that e · a = a for all a ∈ G. Note the e·e = e. since the two equations 1 ∗ x = 1 and (−1) ∗ x = −1 have no simultaneous solution in G. This shows that e is a multiplicative identity for G. Prove that these weaker conditions (given only on the left) still imply that G is a group. There is no right identity element. and so the original condition (iii) is satisﬁed. The operation is associative since a∗(b∗c) = a∗(|b|c) = |a||b|c = |ab|c and (a ∗ b) ∗ c = (|a|b) ∗ c = ||a|b|c = |ab|c. That is. Finally. we have shown that G is not a group. Then a ∗ b = 0 if a = 0 and b = 0. and replace condition (iv) with the condition that for each a ∈ G there exists a ∈ G with a · a = e. az + b . Thus the original condition (iv) holds. F is the set of all functions f (z) : C → C of the form f (z) = cz + d . Let a be an element in G with a · a = e. we can cancel a from the left of the above equation. Let F be the set of all fractional linear transformations of the complex plane. Give an example to show that it is not suﬃcient to require the existence of a left identity element together with the existence of right inverses. and so G is a group under the given operation. since 1 ∗ a = |1|a = a for all a ∈ G. but the equation x ∗ a = 1 has no solution for a = −1. Solution: On the set G of nonzero real numbers. which shows that a is indeed the multiplicative inverse of a. In Deﬁnition 3. for all a ∈ G. and then (as above) we can cancel a to get a · a = e. Then a · (a · e) = (a · a) · e = e · e = e = a · a . Solution: Assume that the two replacement conditions hold. so −1 has no left inverse. deﬁne the operation a ∗ b = |a|b.1. We also have the equation a · (a · a ) = (a · a) · a = e · a = a = a · e . and that the associative law holds. to get a · e = a. so by induction it holds for all values of n. The number 1 is a left identity element. 29. b ∈ G. In summary.

a2 c2 b2 d2 a1 c1 b1 d1 = a2 a1 + b2 c1 c2 a1 + d2 c1 a2 b1 + b2 d1 c2 b1 + d2 d2 a2 z + b2 c2 z + d2 If we associate with the fractional linear transformations f2 (z) = and f1 (z) = a1 z + b1 a2 b2 a1 b1 the matrices and . the condition that ad − bc = 1 for a fractional linear transformation corresponds to the condition that the determinant of the associated matrix is equal to 1. Furthermore. The proof that the determinant of a product is the product of the determinants can be used to show that in the composition f2 ◦ f1 we will still have the required condition on the coeﬃcients that we calculated. c2 d2 c1 d1 c1 z + d1 then we can see that composition of two fractional linear transformations corresponds to the product of the two associated matrices. Show that F forms a group under composition of functions. so we need to check the closure axiom in Deﬁnition 3. b = 0. for matrices). Solution: We ﬁrst need to check that composition of functions deﬁnes a binary operation on F . and this can be shown by choosing coeﬃcients a = 1.3. All of this means that it is fair to use what we already know about matrix multiplication. We only need to check that it can be written in the correct form. b.1. Finally.2 in the text. and f2 (z) = . as a fractional linear transformation. we can use the formula for the inverse of a 2 × 2 matrix with determinant 1 to ﬁnd an inverse function for . Then for any complex number z we have f2 ◦ f1 (z) = f2 (f1 (z)) = a2 = c2 = = a1 z+b1 c1 z+d1 a1 z+b1 c1 z+d1 a2 fz (z) + b2 c2 fz (z) + d2 + b2 + d2 a2 (a1 z + b1 ) + b2 (c1 z + d1 ) c2 (a1 z + b1 ) + d2 (c1 z + d1 ) (a2 a1 + b2 c1 )z + (a2 b1 + b2 d1 ) . (c2 a1 + d2 c1 )z + (c2 b1 + d2 d1 ) You can see that verifying all of the axioms is going to be painful. Composition of functions is always associative (compare Exercise 3. We need a better way to look at the entire situation. and the identity function will serve as an identity element for F . respectively. d are integers with ad − bc = 1. c. a1 z + b1 a2 z + b2 Let f1 (z) = .1. with a1 d1 − b1 c1 = 1 and c1 z + d1 c2 z + d2 a2 d2 − b2 c2 = 1. c = 0. and d = 1. so let’s look at the following matrix product.CHAPTER 3 SOLUTIONS 61 where the coeﬃcients a.

Let G = {x ∈ R | x > 1} be the set of all real numbers greater than 1. y ∈ G. Finally. This gives f −1 (z) = . we have (a ∗ b) ∗ c = (ab − a − b + 2) ∗ c = (ab − a − b + 2)c − (ab − a − b + 2) − c + 2 = abc − ab − ac − bc + a + b + c . On the other hand. Solution: If a. the one computation 2 ∗ y = 2y − 2 − y + 2 = y suﬃces to show that 2 is the identity element. (c) Show that 2 is the identity element for the operation ∗. b. (a) Show that the operation ∗ is closed on G.62 f (z) = CHAPTER 3 SOLUTIONS az + b dz − b . It follows immediately that ab − a − b + 2 > 1. 24 Solution: You can check that x2 = 1 for all elements of the group. Solution: Given any a ∈ G. Thus a ∗ (b ∗ c) = (a ∗ b) ∗ c. which has the solution y = a/(a − 1). c ∈ G.2 SOLUTIONS 23. This solution belongs to G since a > a − 1 implies a/(a − 1) > 1. we have a ∗ (b ∗ c) = a ∗ (bc − b − c + 2) = a(bc − b − c + 2) − a − (bc − b − c + 2) + 2 = abc − ab − ac − bc + a + b + c . b ∈ G. a∗(a/a−1) = a2 /(a−1)−a−a/(a−1)+2 = (a2 −a2 +a−a)/(a−1)+2 = 2. so b − 1 > 0. For x. Solution: Since the operation is commutative. . Thus each nonzero element generates a subgroup of order 2. This gives us the equation ay − a − y + 2 = 2. we need to solve a ∗ y = 2. and therefore a(b − 1) > (b − 1). 3. (b) Show that the associative law holds for ∗. Find all cyclic subgroups of Z× . 32. (d) Show that for element a ∈ G there exists an inverse a−1 ∈ G. including just the element itself and the identity [1]24 . then a > 1 and b > 1. Solution: For a. deﬁne x ∗ y = xy − x − y + 2. and completes the proof that cz + d −cz + a F forms a group under composition of functions.

3.CHAPTER 3 SOLUTIONS 63 24. 1 2 and so g1 g2 = an an = (a1 a2 )n . 3. Solution: The subset H is ﬁnite and nonempty (it certainly contains [1]21 ). we need to check the orders of elements in Z× = {±1. 2. ±7. A similar argument shows that K is a subgroup of Z× . 7). 3)(5. (b) Find a subgroup of S7 that contains 12 elements. Solution: First. It is easy to check that the subset H = {±[1]. 3)(5. 3)(5. and let n be a ﬁxed positive integer. and so g −1 has the right form to belong to N . (a) Find the cyclic subgroup of S7 generated by the element (1. 2. Since the order of a product of disjoint cycles is the least common multiple of their lengths. which 20 turns out to have order 4. nonempty subset of a known group. suppose that g1 and g2 belong to N . 7))5 = (1. H is not cyclic since no element of H has order 4. 3)(5. with g = an . Since H is a ﬁnite. 7).2. Solution: We have ((1. ±[9]} is closed. 3. 3)(5. 2. 7). 7))3 = (5. 2. together with (1. ±9}. and why they must form a subgroup. ±3. The element [9] = [3]2 has order 2. 7))6 = (1). You do not have to list all of the elements if you can explain why there must be 12. and therefore [x]21 · [y]21 = [xy]21 belongs to H. In Z× . 26. show that 21 H = {[x]21 | x ≡ 1 (mod 3)} are subgroups of G.4 it is enough to show that H is closed under multiplication. 1 2 Finally. Corollary 3. ((1. If [x]21 and [y]21 belong to H. and so [3] is a cyclic subgroup of order 4. 25. In G = Z× . ﬁnd two subgroups of order 4. 21 27. 3)(5. so by Corollary 3. then x ≡ 1 (mod 3) and t ≡ 1 (mod 3). if g ∈ N . 6. Finally. Solution: We only need to ﬁnd an element of order 12. ((1. one that is cyclic and one that is not 20 cyclic. 7) has order 12. 2. Solution: To ﬁnd a cyclic subgroup of order 4.2. then g −1 = (an )−1 = (a−1 )n . 3). It is natural to begin with [3]. ((1. Show that N = {g ∈ G | g = an for some a ∈ G} is a subgroup of G. form the cyclic subgroup generated by (1. so it follows that xy ≡ 1 (mod 3). 3)(5. 4)(5. 7))2 = (1. 7). since it will generate a cyclic subgroup with 12 elements. Let G be an abelian group. 3)(5. 7))4 = (1. Then there must exist elements a1 and a2 in G with g1 = an and g2 = an .4 implies that it is a subgroup. the subset N is nonempty since the identity element e can always be written in the form e = en . 2. The last equality holds since G is abelian. 2). and K = {[x]21 | x ≡ 1 (mod 7)} . the element (1. 7). 2. Next. 2)(5. ((1. 2. 2. These elements. 2.

Solution: The order of [2] is a divisor of |Z× | = p − 1 = 2n . We can use Demoivre’s Theorem (Theorem A. It follows from this theorem that (cos(3π/4) + i sin(3π/4))8 = √ √ cos(6π) + i sin(6π) = 1. This shows that [2] has order 2n. (a) Show that in Z× the order of [2]p is 2n.5 for a discussion of the properties of complex numbers. 30. let θ = 0. ﬁnd the order of the √ √ √ 2 2 2 2 + i and − − i. Solution: Closure: To show that H is closed under multiplication we need to use the familiar trig identities for the sine and cosine of the sum of two angles. sin(θ + φ) cos(θ + φ) Identity: To see that the identity matrix is in the set. 29. Suppose that p is a prime number of the form p = 2n + 1. and you can check that − √ √ 2 2 2 + 2 i = cos(3π/4)+i sin(3π/4) and − √ √ 2 2 2 − 2 i = cos(5π/4)+i sin(5π/4). and so − 22 + 22 i has order 8 in C× . (See Appendix A. and squaring this yields 22n ≡ 1 (mod p). Existence of inverses: cos θ sin θ − sin θ cos θ −1 = cos(−θ) − sin(−θ) sin(−θ) cos(−θ) ∈ H. cos θ − sin θ cos φ − sin φ sin θ cos θ sin φ cos φ = = = cos θ cos φ − sin θ sin φ − cos θ sin φ − sin θ cos φ sin θ cos φ + cos θ sin φ − sin θ sin φ + cos θ cos φ cos θ cos φ − sin θ sin φ −(sin θ cos φ + cos θ sin φ) sin θ cos φ + cos θ sin φ cos θ cos φ − sin θ sin φ cos(θ + φ) − sin(θ + φ) ∈ H. A similar √ √ argument shows that − 22 − 22 i also has order 8.) Each of the numbers has magnitude 1. we have 2n ≡ −1 (mod p). . so 2k ≡ 1 (mod p) since 2k − 1 < 2n + 1 = p. In the multiplicative group C× of √ complex numbers. elements − 2 2 2 2 Solution: It is probably easiest to change these complex numbers from rectangular coordinates into polar coordinates. and for any proper divisor k of 2n we have k ≤ n. so by part (a) p this implies that n is a divisor of 2n−1 .64 CHAPTER 3 SOLUTIONS 28. Thus the order of [2] is a divisor of 2n. p Solution: Since 2n +1 = p. and therefore n is a power of 2. (b) Use part (a) to prove that n must be a power of 2.2) to compute powers of complex numbers. show that cos θ − sin θ H= θ∈R sin θ cos θ is a subgroup of G.5. In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries.

3.14 in the text deﬁnes the centralizer of an element a of the group G to be C(a) = {x ∈ G | xa = ax}. There are 8 such elements. [4]5 have order 5 in Z5 and [1]3 . 2 1 a b . Equating corresponding entries 1 1 c d a+c b+d shows that we must have 2a+b = 2a+c. Since [1]5 .3 SOLUTIONS 16. Show that Z5 × Z3 is a cyclic group. and list all of the generators for the group. then order of an element ([a]5 . On the other hand. Solution: The closure axiom holds since a1 b1 a2 b2 a1 a2 − 2b1 b2 a1 b2 + b1 a2 = −2b1 a1 −2b2 a2 −2(a1 b2 − b1 a2 ) a1 a2 − 2b1 b2 identity matrix belongs K. a2 + 2b2 −2(−b) a Comment: We don’t need to worry about the condition ad − bc = 0. which can easily be listed. a+b = 2b+d. The 1 a −b . b ∈ R and ab = a2 + b2 . and a b −2b a −1 . 1 1 b a−b Solution: Let A = 3.4 (b). or d = a − b. while the second and third equations imply that a = b + d = c + d.CHAPTER 3 SOLUTIONS 31. = 2 1 1 1 . 2c+d = a+c. c = −2b. [b]3 ) in Z5 × Z3 is the least common multiple of the orders of the components. . and suppose that X = belongs to 1 1 c d the centralizer of A in GL2 (R). ad − bc = 0 65 Show that K is a subgroup of GL2 (R). and c+d = b+d. [b]3 ) is a generator if and only if [a]5 = [0]5 and [b]5 = [0]5 . so the centralizer in GL2 (R) of the ma2 1 a b trix is the subgroup a. any matrix of this form commutes with A. Solution: By Proposition 3. [3]5 .2. the element ([a]5 . which is always positive. so do2a + b a + b a b 2 1 ing this calculation shows that = = 2c + d c + d c d 1 1 2 1 a b 2a + c 2b + d = . [2]3 have order 3 in Z3 . K= a b c d d = a. [2]5 . 32. since for any element in H the determinant is a2 + 2b2 . Then we must have XA = AX. Let K be the following subset of GL2 (R). The ﬁrst and last equations imply that b = c. Compute the centralizer in GL2 (R) of the matrix Note: Exercise 3.

Solution: It can be shown (as in Problem 3. [5]. For the third row. Since x · [1] = x · [19] for x ∈ G. 12) = 3. 17. Similarly. 6] = 12. [19]}. the set HK must contain 12 elements. [15]18 ) in the group Z12 × Z18 . for subgroups H1 ⊆ G1 and H2 ⊆ G2 . o([15]18 ) = o([3]18 ) = 6. 18. 1) has order 2. Therefore the total number of ways to construct a second row independent of the ﬁrst is pn − p. we need to subtract p2 . [13]. cn . so it generates a cyclic subgroup that does not have the required form. and so HK = G. 20. but to be linearly independent of the ﬁrst row. we need to omit the p multiples of the ﬁrst row.) i 0 0 0 in the group GL3 (C). [25]. Solution: Since gcd(9. giving the stated result. and ﬁnd the subgroup HK. [15]18 ) is lcm[4. A short computation shows that H = {[1]. it cannot be a scalar multiple of that row. which is the number of vectors in the subspace spanned by the ﬁrst two rows that we have chosen. 19. Show that if p is a prime number. Thus there are pn − p2 possibilities for the third row. Adding the identity element to the list accounts for all 15 elements of Z5 × Z3 . This argument can be continued.26) that the given subsets are subgroups. There are 4 elements of order 5 (with [0]3 as the second component) and 2 elements of order 3 (with [0]5 as the ﬁrst component). let H = {[x] | x ≡ 1 (mod 4)} and K = {[y] | y ≡ 36 1 (mod 9)}.2. so there are pn − 1 choices. [17].66 CHAPTER 3 SOLUTIONS Comment: The other 7 elements in the group will have at least one component equal to zero. Find the order of the element A = 0 −1 0 0 −i Solution: For any diagonal 3 × 3 matrix we n n a 0 0 a 0 b 0 = 0 0 0 c 0 have 0 bn 0 0 0 . we have o([9]12 ) = o([3]12 ) = 4. Show that H and K are subgroups of G. the element (1. This is done by noting that we need n linearly independent rows. 21. Find two groups G1 and G2 whose direct product G1 × G2 has a subgroup that is not of the form H1 × H2 . (A more formal proof could be given by induction. then the order of the general linear group GLn (Zp ) is (pn − 1)(pn − p) · · · (pn − pn−1 ). Solution: In Z2 × Z2 . Find the order of the element ([9]12 . Thus the order of ([9]12 . Solution: We need to count the number of ways in which an invertible matrix can be constructed. Since we have p possible scalars. The ﬁrst row can be any nonzero vector. There are pn possibilities for the second row. In the group G = Z× . [29]} and K = {[1].

Equating corresponding entries shows that we must have b = 0. Then XB = BX. 22. while the . 23. and show that C(A) ∩ C(B) = Z(G). any matrix of this form commutes with A. Equating corresponding entries shows that we must have m + b = b + 1. and −i. and 1 b b∈R . where Z(G) is the center of G. Then we must 0 1 have XA = AX. and doing this calculation shows that Solution: Suppose that X = m 0 m+b 1 = m b 0 1 1 1 0 1 = 1 0 1 1 m b 0 1 = m b+1 0 1 . On the other hand. Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . Let A = 1 1 −1 0 and B = . The ﬁrst equation implies that c = 0. and so m = 1. a + 2b = 2b + d. Equating c d 0 2 0 2 c d 2c 2d corresponding entries shows that we must have 2a = 2a + c. and since any element in the center of G must belong to C(A) ∩ C(B). Find the centralizers C(A) and 0 1 0 1 C(B). our calculations show that the center of G is the trivial subgroup. −1. Thus o(A) = 4. 0 1 This shows that C(A) ∩ C(B) is the identity matrix. 2c = 2c. Compute the centralizer in GL2 (Z3 ) of the matrix Solution: Let A = 2 1 0 2 2 0 1 2 . and suppose that X = . and so C(B) = m 0 0=m∈R . Then XA = AX. a b belongs to c d 2a a + 2b the centralizer of A in GL2 (Z3 ). containing only the identity element.CHAPTER 3 SOLUTIONS 67 It follows immediately that the order of A is the least common multiple of the orders of the diagonal entries i. and c + 2d = 2d. and so −1 0 0 1 m b 0 1 −m −b 0 1 = = = . so C(A) = 0 1 Now suppose that X = −m 0 b 1 m b 0 1 m b 0 1 −1 0 0 1 belongs to C(B). and so = 2c c + 2d a b 2 1 2a + c 2b + d 2 1 a b = = . m b belongs to C(A) in G.

a + b = 2b + d.68 CHAPTER 3 SOLUTIONS second equation implies that a = d. Solution: Since in the matrix (b) Show that if we let A = Solution: We have BA = 1 −1 0 1 −1 0 0 1 = 1 1 0 1 −1 0 0 1 −1 −1 0 1 and B = 1 1 0 1 . while it follows from Problem 20 in this section that GL2 (Z3 ) has (32 − 1)(32 − 3) = 48 elements. Equating corc d 1 1 1 1 c d a+c b+d responding entries shows that we must have 2a + b = 2a + c. b ∈ Z3 and a = 0 . we will have a total of 10 elements. and A−1 B = −1 −1 0 1 . Compute the centralizer in GL2 (Z3 ) of the matrix Solution: Let A = 2 1 1 1 . 24. It fol2 1 lows that the centralizer in GL2 (Z3 ) of the matrix is the subgroup 1 1 a b a. H= m b 0 1 ∈ GL2 (Z5 ) m. m = ±1 (a) Show that H is a subgroup of G with 10 elements. a b belongs to the c d 2a + b a + b centralizer of A in GL2 (Z3 ). Let H be the following subset of the group G = GL2 (Z5 ). and it is certainly 0 1 0 1 0 1 nonempty. b ∈ Z3 and a = 0 or b = 0 . Then XA = AX. then BA = A−1 B. and c + d = b + d. b ∈ Z5 . and suppose that X = 2 1 1 1 . while the second equation implies that d = a − b. 2c + d = a + c. The ﬁrst equation implies that c = b. 0 2 0 a Comment: The centralizer contains 6 elements. and so it is a subgroup since the group is ﬁnite. It follows that the centralizer in GL2 (Z3 ) 2 1 a b of the matrix is the subgroup a. −1 0 = 0 1 . b a−b Comment: GL2 (Z3 ). In this case the centralizer contains 8 of the 48 elements in 25. and so = 2c + d c + d a b 2 1 2 1 a b 2a + c 2b + d = = . m b there are two choices for m and 5 0 1 choices for b. The set is closed under mul±1 b ±1 c ±1 b ± c tiplication since = .

17 Solution: The element [3] is a generator for Z× .3 (a) implies that any isomorphism between cyclic groups must map a generator to a generator.5 that φ is also an onto function. 17 Next.1. and 38 ≡ 3 · 11 ≡ 33 ≡ −1 ≡ 1. and then substitute the result into the function φ: φ([n]16 + [m]16 ) = φ([n + m]16 ) = [3]n+m . Since φ is deﬁned by 17 using a representative n of the equivalence class [n]16 . we have to show that the formula for φ does not depend on the particular representative that is chosen.8 (c) that [3]k = [3]m since [3]17 has order 16 in Z× Therefore φ([k]16 ) = φ([m]16). 33 = 27 ≡ 10. Solution: Since 1 0 b 1 1 c 0 1 = 1 b+c 0 1 . Multiplying on the 0 1 right by B will create 5 additional elements. etc.1. and 17 17 17 so φ is a well-deﬁned function. For any elements [n]16 and [m]16 in Z16 . we ﬁrst apply the function φ to the two elements. If k ≡ m (mod 16). since Z16 is also a cyclic group. 35 ≡ 3·13 ≡ 39 ≡ 5. Deﬁne φ : Z16 → Z× by setting φ([1]16 ) = [3]17 . φ([2]16 ) = [3]2 . Therefore Z× is a cyclic group with 16 17 elements.4 SOLUTIONS 21. since 32 = 9. and Proposition 3.4. erated by A consists of all matrices of the form 3. with generator [1]16 . 17 17 17 Thus φ([n]16 + [m]16 ) = φ([n]16 ) · φ([m]16 ). [n]16 and [m]16 . and so φ is a one-to-one function.CHAPTER 3 SOLUTIONS 69 (c) Show that every element of H can be written uniquely in the form Ai B j . This provides the clue at to how to deﬁne the isomorphism we need. the cyclic subgroup gen- 1 b . Proposition 3.8 (c) shows that φ([k]16 ) = φ([m]16 ) only if k ≡ m (mod 16). and this completes the proof that φ is a group isomorphism. we ﬁrst compute what happens if we combine [n]16 and [m]16 using the operation in Z16 . and then combine the results using the operation in Z× : 17 φ([n]16 ) · φ([m]16 ) = [3]n [3]m = [3]n+m . 36 ≡ 3·5 ≡ 15.2. The 17 17 general formula is φ([n]16 ) = [3]n . 37 ≡ 3·15 ≡ 45 ≡ 11.2. giving all of the elements in H. Show that Z× is isomorphic to Z16 . 17 34 ≡ 3·10 ≡ 30 ≡ 13.4. then it follows from Proposition 3. . Then because both Z16 and Z× have 16 17 elements. it follows from Proposition 2. The proof that φ respects the two group operations follows the proof in Example 3. for all [n]16 ∈ Z16 . where 0 ≤ i < 5 and 0 ≤ j < 2.

θ2 (a2 b2 )) = (θ1 (a1 )θ1 (b1 ). then since θ1 is an isomorphism there is a unique element x1 ∈ G1 with y1 = θ1 (x1 ). a2 ) · (b1 . b2 )) = φ((a1 b1 . and suppose that θ1 : G1 → H1 and θ2 : G2 → H2 are group isomorphisms. showing that [4n + 3m]6 = [4k + 3j]6 . which together imply that 6 | (4n + 3m) − (4k + 3j). Thus there is a unique element (x1 . Similarly. Let G1 . is a group isomorphism. x2 ) ∈ G1 × G2 such that (y1 . First prove that φ is a well-deﬁned function. [m]10 )) = 7 11 ([3]n . a2 b2 )) = (θ1 (a1 b1 ). x2 ) = (θ1 (x1 ). since θ2 is an isomorphism there is a unique element x2 ∈ G2 with y2 = θ2 (x2 ).70 CHAPTER 3 SOLUTIONS 22. Let φ : R× → R× be deﬁned by φ(x) = x3 . 23. Furthermore. It then follows from 11 Problem 23 that φ : Z6 × Z10 → Z× × Z× deﬁned by φ(([n]6 . 30 | 4(n−k). The function is one-to-one and onto since for each y ∈ R× the equation φ(x) = y has the unique solution √ x = 3 y. Prove that the group Z× × Z× is isomorphic to the group Z6 × Z10 . It follows that 10 | n − k. for all (x1 . [4k + 3j]6 ). [2]m ). Prove that φ is a group isomorphism. y2 ) ∈ H1 × H2 . generated by [3]7 . Solution: The function φ preserves multiplication in R× since for all a. x2 ) ∈ G1 × G2 . so 6 | 4(n−k). G2 . for all x ∈ R. a2 )) · φ((b1 . b2 )) and so φ : G1 × G2 → H1 × H2 is a group isomorphism. and so φ is one-to-one and onto. which shows that the formula for φ does yield a well-deﬁned function. Solution: If ([n]30 . H1 . and 7 that Z× is cyclic of order 10. then 30 | n − k and 2 | m − j. 7 11 25. Just as in Problem 21. 24. Solution: If (y1 . Deﬁne φ : Z30 × Z2 → Z10 × Z6 by φ([n]30 . and so [n]10 = [k]10 . [4n + 3m]6 ). for all ([n]30 . Thus ([n]10 . generated by [2]11 . [j]2 ) are equal elements of Z30 × Z2 . you 11 can show that θ1 : Z6 → Z× deﬁned by θ1 ([n]6 ) = [3]n and θ2 : Z10 → Z× 7 7 11 deﬁned by θ2 ([m]10 ) = [2]m are group isomorphisms. 7 11 Solution: You can check that Z× is cyclic of order 6. y2 ) = φ(x1 . Show that φ is a group isomorphism. H2 be groups. x2 ). [m]2 ) ∈ Z30 × Z2 . b ∈ R× we have φ(ab) = (ab)3 = a3 b3 = φ(a)φ(b). and then 6 | 3(m−j). [m]2 ) = ([n]10 . θ2 (a2 )θ2 (b2 )) φ((a1 . . a2 ) and (b1 . [m]2 ) and ([k]30 . θ2 (a2 )) · (θ1 (b1 ). and then prove that φ is a group isomorphism. b2 ) in G1 × G2 . θ2 (x2 )). θ2 (a2 )θ2 (b2 )) = (θ1 (a1 ). for all [n]6 ∈ Z6 and all [m]10 ∈ Z10 . [4n + 3m]6 ) = ([k]10 . we have φ((a1 . θ2 (b2 )) = (θ1 (a1 )θ1 (b1 ). Deﬁne φ : G1 × G2 → H1 × H2 by φ(x1 . Given (a1 .

4 is satisﬁed. HK = G. for some q ∈ Z. We know (by Exercise 3. [c]2 ) and ([b]30 . [4a + 4b + 3c + 3d]2 ) 71 φ(([a]30 . [0]6 ).3.4. Let G be a group. [d]2 ) we have φ(([a]30 . [m]2 ) = ([0]30 . Solution: Let φ : G1 × G2 → G be an isomorphism. x2 ) | x1 = e} have the properties we are looking for. [4a + 3c]2 ) + ([b]10 . and hk = kh for all h ∈ H and k ∈ K. and 3 | 40q since 3 | 3m. This means that we can use Proposition 3. We have therefore checked all of the necessary conditions. Let H = φ(H ∗ ) and K = φ(K ∗ ) be the images in G of H ∗ and K ∗ . By Exercise 3.13 in the text. say n = 10q. so aHa−1 = φ(H) is a subgroup of G.9 in the text shows that in G1 × G2 the subgroups H ∗ = {(x1 .4 to show that φ is one-to-one. the function φ : G → G deﬁned by φ(x) = axa−1 . It is then clear that the function θ : H → aHa−1 deﬁned by θ(x) = axa−1 is an isomorphism. . [0]2 ). and so the condition in Proposition 3.15) that H and K are subgroups of G. it follows that φ must also be an onto function. so [m]2 = [0]2 .4. and hk = kh for all h ∈ H and k ∈ K. We conclude that φ is a one-toone function. [d]2 )) = φ(([a + b]30 . [4n + 3m]6 ) = ([0]10 . [m]2 ) = ([0]10 . [4(a + b) + 3(c + d)]2 ) = ([a + b]10 . G2 be groups. We have now shown that if φ([n]30 . then there are subgroups H and K in G such that H ∩ K = {e}. respectively. so [n]30 = [10q]30 = [0]30 . Exercise 3. then the subset aHa−1 = {g ∈ G | g = aha−1 for some h ∈ H} is a subgroup of G that is isomorphic to H. or 6 | (40q + 3m). [c + d]2 )) = ([a + b]10 . Prove that if G is isomorphic to G1 × G2 . for all x ∈ G. [4b + 3d]2 ) = ([a + b]10 . HK = G. If φ([n]30 .4. G1 .15 the image under φ of any subgroup of G is again a subgroup of G. It follows that 2 | (40q + 3m) and 3 | (40q + 3m). it follows that 2 | m. [4a + 3c + 4b + 3d]2 ) = ([a + b]10 . and let H be a subgroup of G. 27.4. so we only need to show that H ∩ K = {e}. so we may conclude that φ is a group isomorphism. Solution: By Exercise 3. [0]6 ). Prove that if a is any element of G. [m]2 ) = ([0]10 . so 10 | n.CHAPTER 3 SOLUTIONS For any elements ([a]30 . [c]2 )) + φ(([b]30 . [4a + 4b + 3c + 3d]2 ) and so φ respects the operations in the two groups. is a group isomorphism. then ([n]30 . Let G. x2 ) | x2 = e} and K ∗ = {(x1 . Then since 2 and 3 are prime numbers. and therefore 2 | 3m since 2 | 40q. [0]6 ). and 3 | q. and 6 | (4n + 3m).4. 26. Since the two groups both have 60 elements. then ([n]10 . [d]2 )) = ([a]10 . [c]2 ) + ([b]30 .

(The computation necessary to show that φ preserves the respective operations is the same computation we used to show that H is closed. Identity: The identity matrix has the correct form. If y ∈ H ∩ K. then y ∈ H. show that H= a11 a21 a12 a22 ∈ GL2 (R) a11 = 1. showing that H ∩ K = {e}. a2 ))φ((b1 . under addition. it is clear x1 0 that the mapping φ : Z× ×Z× → GL2 (Zp ) deﬁned by φ(x1 . p p 0 x2 for each (x1 . for x ∈ G1 × G2 . This mapping respects the operations in the two groups. Since y ∈ K as well. x2 ) = . so x ∈ H ∗ ∩ K ∗ . a2 b2 )) a1 b1 0 = 0 a2 b2 = a1 0 0 b1 a2 0 0 b2 = φ((a1 . since for (a1 . Show that for any prime number p. and so x ∈ H ∗ . and thus G = HK. (b) Show that H is isomorphic to the group R of all real numbers. and we can write x = h∗ k ∗ for some h∗ ∈ H ∗ and some k ∗ ∈ K ∗ . It is clear that φ preserves the fact that elements of h∗ and K ∗ commute. the subgroup of diagonal matrices in GL2 (Zp ) is isomorphic to Z× × Z× . where e1 and e2 are the respective identity elements in G1 and G2 . Thus y = φ((e1 . a2 )(b1 . e2 ). with y = φ(x). b2 )) = φ((a1 b1 . b2 ) ∈ Z× × Z× we have p p φ((a1 . for all x ∈ R. and therefore x = (e1 . is one-to-one and maps Z× ×Z× onto the subgroup p p p p of diagonal matrices. e2 )) = e.72 CHAPTER 3 SOLUTIONS Let y ∈ G. (a) In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. p p Solution: Since each matrix in GL2 (Zp ) has nonzero determinant. Solution: Closure: 1 a 0 1 1 a 0 1 1 0 −1 b 1 = 1 a+b 0 1 . b2 )) . You can 0 1 easily check that φ is an isomorphism.) . (b1 . We conclude that H and K satisfy the desired conditions. Thus φ is the desired isomorphism. Existence of inverses: = 1 −a 0 1 ∈ H. a21 = 0. a22 = 1 is a subgroup of G. 28. it follows that y = φ(h∗ k ∗ ) = φ(h∗ )φ(k ∗ ). Since y is any element of G. 29. 1 x Solution: Deﬁne φ : R → H by φ(x) = . a2 ). we must also have x ∈ K ∗ . x2 ) ∈ Z× ×Z× .

as given in Table 3. then the groups cannot be isomorphic. for all 0 ≤ i < 3 and 0 ≤ j < 2. if one of the groups is abelian but the other is not. By Proposition 3. 73 Show that G is not isomorphic to the direct product R× × R.3 (b). This information should make it plausible that the function φ : S3 → H deﬁned by φ(ai bj ) = Ai B j . we can show that BA = A−1 B. where 0 ≤ i < 3 and 0 ≤ j < 2. Let A = and B = . and let S be any set for which there exists a one-toone and onto function φ : G → S. G is not abelian. 31. Let G be a group. and that φ is actually a group isomorphism.3. Solution: This group is small enough that we can just compare its multiplication table to that of S3 .3 (on page 104 of the text). since φ and φ−1 are functions and the operation on G is well-deﬁned.CHAPTER 3 SOLUTIONS 30.3. without actually writing out 1 1 −1 0 the multiplication table. The direct product R× ×R is an abelian group. Solution: (Outline only) The operation is well-deﬁned on S. 32. since = 0 1 0 1 0 1 1 1 −1 0 −1 1 but = . since each factor is abelian. Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . The associative law holds . and that each element of H has the form can be written uniquely in the form Ai B j . Deﬁne an operation on S by setting x1 · x2 = φ(φ−1 (x1 )φ−1 (x2 )). for all x1 . Remember that constructing an isomorphism is the same as constructing a one-to-one correspondence between the elements of the group. m = 0 Show that H is isomorphic to the symmetric group S3 . In this case we can explain how this can be done.4. Prove that S is a group under this operation. such that all entries in the respective group tables also have the same one-to-one correspondence. Let H be the following subgroup of group G = GL2 (Z3 ).25. Then just 0 1 0 1 as in Problem 3. H= m b 0 1 ∈ GL2 (Z3 ) m. On −1 0 1 1 −1 −1 the other hand. Solution: Our approach is to try to ﬁnd an algebraic property that would be preserved by any isomorphism but which is satisﬁed by only one of the two groups in question. gives a one-to-one correspondence between the elements of the groups which also produces multiplication tables that look exactly the same. x2 ∈ S. b ∈ Z3 . Thus the two groups cannot be iso0 1 0 1 0 1 morphic.

Show that the three groups Z6 . we have |Z× | = 6. 3.1. 30) = 6. 23 = 8 ≡ 1.0. Therefore = Z4 × Z10 ∼ Z2 × Z20 .3 (b). ±3. and that Z20 ∼ = = Z4 × Z5 . Find all generators of the cyclic group Z28 . it is possible to show that the obvious = = mapping from Z4 × Z2 × Z5 onto Z2 ∼ Z4 × Z5 is an isomorphism.1.74 CHAPTER 3 SOLUTIONS in S because it holds in G. and so [5] must 9 = 18 have order 6.5. In Z30 . and so we actually have [24]30 = [18]30 . Similarly. ±5. showing that Z× is cyclic of order 6. Given a known group G such as R× . where e is the identity of G. = 22.5 SOLUTIONS 20. ±13}. 9 18 Solution: First. Solution: The subgroups correspond to the divisors of 100. 18 = 21. Thus all three groups are isomorphic. The two groups are not isomorphic since the ﬁrst has an element of order 4.5. the identity element in S is φ(e). 24.23 that Z4 × Z10 ∼ Z4 × Z2 × Z5 . Our theorems tell us that 18 Z× ∼ Z6 . Solution: By Proposition 3. showing that Z× is cyclic of order 6. Is Z4 × Z15 isomorphic to Z6 × Z10 ? Solution: As in Problem 21. we ﬁrst ﬁnd gcd(18. ﬁnd the order of [24]30 . ±9. and Z× are isomorphic to each other. The list of generators 2 7 is {±1. Solution: Using Proposition 3. and are given in Figure 3. Comment: This reveals the secret behind problems like Exercises 3.4. 53 ≡ 17 ≡ 1. and Z6 × Z10 ∼ = = Z2 × Z3 × Z2 × Z5 . the generators correspond to u the numbers less than 28 and relatively prime to 28. = and Z2 × Z20 ∼ Z2 ∼ Z4 × Z5 . Z4 × Z15 ∼ Z4 × Z3 × Z5 . and it is easy to check that if x ∈ S. 22 = 4. and |Z× | = 6. Give the lattice diagram of subgroups of Z100 . The Euler ϕ-function allows us to compute how many there are: ϕ(28) = 1 · 6 · 28 = 12. In Z× . and so the subgroup has 30/6 = 5 elements. Is Z4 × Z10 isomorphic to Z2 × Z20 ? Solution: It follows from Theorem 3.4. ±11. Note that nZ100 is used to mean all multiples of n in Z100 .3 (a).12 in the text.5. Finally. we can use one-toone functions deﬁned on G to produce new groups with operations that look rather diﬀerent from the usual examples. ﬁnd the order of the subgroup [18]30 .11 and 3. Then [18]30 = [6]30 . 9 18 9 and so [2] must have order 6.4 that Z10 ∼ Z2 × Z5 . . Our 9 theorems tell us that Z× ∼ Z6 . while the second has none. Z× . then x−1 = φ((φ−1 (x))−1 ). 25. It then follows from Problem 3. 23. [24]30 = [6]30 . 52 ≡ 7. In Z× .

4) and (1. Use the the result in Problem 27 to show that the multiplicative groups Z× 15 and Z× are not cyclic groups. then the direct product G1 × G2 is a cyclic group. then it follows from Theorem 3. In that group.1: for Problem 23 Z100 2Z100 5Z100 4Z100 10Z100 25Z100 20Z100 50Z100 0 26. 21 Solution: In Z× . for some positive integer n. . If a has order 7 in G1 and b has order 11 in G2 . 3). and so [8]21 and [−1]21 have order 2. b) has order lcm[7. 11] = 77 in G1 × G2 . 27. 3. 21 21 29. both [−1]15 and [4]15 are easily checked to have order 2. Thus G1 × G2 is cyclic since it has an element whose order is equal to the order of the group. we only need to answer the question in Z2n . respectively. then (a. Solution: Since 7 and 11 are primes. Since isomorphisms preserve orders of elements. the elements of order 2 are the nonzero solutions to the congruence 2x ≡ 0 (mod 2n). we see that [n]2n is the only element of order 2 in Z2n .2 that G is isomorphic to Z2n . the groups are cyclic. Solution: If G is cyclic of order 2n. 28. 15 In Z× . 2. Find all cyclic subgroups of the quaternion group. we have [8]2 = [64]21 = [1]21 .CHAPTER 3 SOLUTIONS 75 Figure 3. Prove that if G1 and G2 are groups of order 7 and 11. Show that any cyclic group of even order has exactly one element of order 2. Use this information to show that the quaternion group cannot be isomorphic to the subgroup of S4 generated by (1.5. and since the congruence can be rewritten as x ≡ 0 (mod n).

so a2 ≡ 1 (mod pq) since p and q are relatively prime. 3. 2) = (1. 4)(2. with a2 = (1. cyclic subgroups would correspond to cyclic subgroups. Solution: For i = 1. 0 ≤ j < 2} with o(a) = n. ±k} is deﬁned in Example 3. a3 b}. we have ab = (1. 4) = a3 b. 3) = ab. the equation bai = an−i b is just the relation that deﬁnes the group. 4. ±j = {±1. ±i}. [−1]pq is not a solution. show that bai = an−i b. ab. a. The elements satisfy the following identities: i2 = j2 = k2 = −1 and ij = k. ±i. 2)(3. On the other side. and i4 = i2 i2 = (−1)2 = 1. for all 0 ≤ i < n. so by Problem 27 it is enough to ﬁnd one other element of order 2. 3. and a3 b = (1. 4) = a2 b. ba2 = (1. 4)(2. a2 b = (1. we have ba = (1. and so it is impossible for this group to be isomorphic to the quaternion group. and ba3 = (1. For example. 4) = (1. since p and q are relatively prime. from the cycle structures of the elements we can see that the only cyclic subgroup of order 4 is the one generated by a (and a3 ). then for i = k + 1 we have bak+1 = (bak )a = (an−k b)a = an−k (ba) = an−k a−1 b = an−(k+1) b . 2. This implies that the result must hold for all i with 0 ≤ i < n. 4). 4) = a and (1. 3) = (2.7 of the text (see page 108). ji = −k. The cyclic subgroups −1 = {±1}. and ±k = {±1. 3)(1.6 SOLUTIONS 22. o(b) = 2. jk = i. pq Solution: We know that [−1]pq has order 2. ±i = {±1. ±j. . If we assume that the result holds for i = k. 3)(2.3. so [a]pq = [−1]pq . To ﬁnd the subgroup generated by a and b. and ba = a−1 b. 4) and a3 = a−1 = (1. 4)(1. 3)(2.6) states that the system of congruences x ≡ 1 (mod p) and x ≡ −1 (mod q) has a solution [a]pq . But a2 ≡ 1 (mod p) and a2 ≡ 1 (mod q). 4) = (2.76 CHAPTER 3 SOLUTIONS Solution: The quaternion group Q = {±1. 2). 3. 3)(1. 2. Prove that if p and q are diﬀerent odd primes. Since a is a cycle of length 4. In S4 . ±j}. 3)(1. Comment: This is similar to a proof by induction. In the dihedral group Dn = {ai bj | 0 ≤ i < n. i3 = i2 i = −i. 3)(2. 4)(1. kj = −i. 3. 3. b. 3) = (1. 3) = b. 3. ik = −j. 2)(3. a2 b. 3. ±k} can be found by using the given identities. let (1. and thus [a]pq has order 2. In any isomorphism. but for each given n we only need to worry about a ﬁnite number of equations. This shows that the subgroup generated by a and b consists of the 8 elements {e. it has order 4. which has 3 cyclic subgroups of order 4. 3). 2)(1. 30. 4. Furthermore. 2. ki = j. a3 . 4.3. i2 = −1. 4). Because q is an odd prime. then Z× is not a cyclic group. a2 . The Chinese remainder theorem (Theorem 1. 3) = (1.

4. 2. 2)(3. 3) and a2 b = (2. 4). 3) and (1. 28. so a ⊆ C(a) ⊆ Dn and C(a) = Dn together imply that C(a) = a . 3)(2. o(b) = 2. ﬁnd the centralizer of a. and this set of elements is closed under multiplication. 4). 2. 3) in S3 . 2. (1. Thus (1. (1. 2. 3. 3. Lagrange’s theorem show that there is no subgroup that lies strictly between a and Dn . Then we can quickly do the other computations: (1. 3. so we have a ⊆ C(a). 2). 4)−1 (1. 4) = (2. 4). 3. show that each element of the form ai b has order 2. 3). Then H must contain a2 = (1. 3)(1. ab = (1. 3. 4) = (1. 3)(1. 4). In the dihedral group Dn = {ai bj | 0 ≤ i < n. 4)−1 (1. 3) = (2. 4)−1 = {(1). a2 b} = {(1). and ba = a−1 b. 4)(2. (2. 2). 4)−1 (1. Solution: We need to compute στ σ −1 . 3. . 4)}. (2. 4)(2. 2. 3). 4) = (1). 3)(1. for σ = (1. 2. 3) = (1. 3. As a shortcut. a2 . 2). (1. 25. 2. o(b) = 2. 3). 2. 4. for each τ ∈ H. Solution: We ﬁrst list the 3-cycles: (1. 0 ≤ j < 2} with o(a) = n. 3). and (1. it is easier to just look at the possible products of 3-cycles. (2. Solution: Using the result from the previous problem. (1. σ(3)) = (4. 4). 3). which shows that σ(1. On the other hand. ﬁnd the subgroup H generated by (1. 3. 3) and b = (1. 3). 4)(1. 4). ab. 3). In S4 . 3)σ −1 = (σ(1). 3). 4) = (1. (2. 4). 26. (1. (1. ﬁnd the corresponding subgroup σHσ −1 . 2). since by deﬁnition b does not belong to C(a). 3). 4)(1. (1. Rather than starting with each of the other elements and then trying to write them as a product of 3-cycles. 4). 2. 3)(2. In the dihedral group Dn = {ai bj | 0 ≤ i < n. 4)(1. (1. σ(2). we have (ai b)2 = (ai b)(ai b) = ai (bai )b = ai (an−i b)b = (ai an−i )(b2 ) = an e = e. 4)−1 = (4. 2) = (4. 3).CHAPTER 3 SOLUTIONS 77 23. in S4 . Solution: The centralizer C(a) contains all powers of a.) Thus H = {(1). and (2.3. 4. 3. 2)(1. and this accounts for all 12 of the elements in A4 . and in A4 . 2) = (4. and ba = a−1 b. 4)(1. For the subgroup H of S4 deﬁned in the previous problem.10. 2. Show that each element in A4 can be written as a product of 3-cycles. Since (1. 4)(1. (2. 2. 0 ≤ j < 2} with o(a) = n. 2. b. 4)H(1. (1. Since a contains exactly half of the elements in Dn . C(a) = Dn . 27. Solution: Let a = (1. 24. 4). 3. We have (1. (1. This shows that C(a) has at least n elements. 4). 2). a. 4)(1)(1. we have (1. 4)−1 = (1. we can use Exercise 2. 2)(1. (1. 2). (2. (3. 2). 2. 3). 3)(1. Find the centralizer of (1. 3)}. 2. (We have just listed the elements of S3 . 4.

(1. we have a(1. 2. Solution: Example 3. (1. 4)(2. Without doing all of the calculations. Let a = (1. If we do a similar computation with a 4-cycle. . .7. 3)(1. for all [x]4 ∈ Z4 . y. Exercise 2. 2.10. 3) . 4). 3)(2. we have a(1. z. (a) Find the formulas for all group homomorphisms from Z18 into Z30 . 3). we need to check that xa = ax.3. 4). and a(1. which leads to the formula φ([x]4 ) = [5x]10 . (1. Thus φ([1]4 ) = [0]10 . 1). 2. . since a just permutes the numbers x. Under any group homomorphism φ : Z4 → Z10 .7 SOLUTIONS 17. we now have a total of 21 elements that are not in C(a). or else φ([1]4 ) = [5]10 . or that axa−1 = x. and z. 4). then σ(1. 4)a−1 = (u. This means that w = z. 2)}. 3)a−1 = (2. 2.78 CHAPTER 3 SOLUTIONS Solution: Since any power of an element a commutes with a. 4)a−1 = (2. For the products of the transposition.5 shows that any group homomorphism from Zn into Zk must have the form φ([x]n ) = [mx]k . . . 4)(3.7. 4)a−1 = (2. A similar argument shows that no 3-cycle that includes the number 4 as one of its entries can belong to C(a). . 18. (2. From the computations in S3 . a(1. which deﬁnes the zero function. Using Exercise 2.5 shows that any group homomorphism from Z18 into Z30 must have the form φ([x]18 ) = [mx]30 . 2. 4). To see that x belongs to C(a). we can conclude that no 4-cycle belongs to C(a). the centralizer C(a) always contains the cyclic subgroup a generated by a. v. 3) do not commute with a. v. 4). Since . 4). 4). 3) is equal to (1. and therefore C(a) = a .10 provides a quick way to do this in a group of permutations. . 4). 3. 3.3. . a(2. 4)a−1 = (2. for all [x]n ∈ Zn . w. z. . the centralizer of (1. 3). Find all group homomorphisms from Z4 into Z10 . . Finally. In S3 . σ(k)). so the only possibilities are 1 and 2. the order of φ([1]4 ) must be a divisor of 4 and of 10. since it is easy to check that (1. 2. k) is a cycle of length k and σ is any permutation. 4). Since there are 6 elements of this form. Solution: Example 3. we will have a(x. 4) = (x. 2)(3. y. 1)(3. 3). w. 4)a−1 = (1. 3) in S4 we have to work a bit harder. y. and by Lagrange’s theorem a proper subgroup of a group with 6 elements can have at most 3 elements. 2). and (3. 2. 4)a−1 = (3. Thus the centralizer of (1. σ(2). The remaining transpositions in S4 are (1. and (2. for all [x]18 ∈ Z18 . and so no product of transpositions belongs to C(a). 3) always contains the subgroup {(1). 2) does not belong to the centralizer. . That exercise shows that if (1. in A4 we must get the same answer: C(a) = a . so (u. It helps to have some shortcuts when doing the necessary computations. k)σ −1 = (σ(1). 4). and a(3. 2. so no transposition in S4 commutes with a. To ﬁnd the centralizer of (1. This accounts for an additional 6 elements. we know that (1. .

11. If y = 0. consider φ([x]18 ) = [5x]30 . of order 3. The only possible orders are 1 and 2. (b) Choose one of the nonzero formulas in part (a). 19. Solution: For example. (a) Show that Z× is cyclic. 17 34 ≡ 3 · 10 ≡ 13. {1. y) belongs to ker(φ). 2. 8. then y ≡ 0 (mod 3). 6. 4. the possible orders of [m]30 = φ([1]18 ) are 1. 33 = 27 ≡ 10. 7 Solution: Since 32 ≡ 2 and 33 ≡ 6. of order 2. 15. {3. then x = 0. 17} ←→ φ(5) = 25. 30) = 6. 17 Solution: The element [3] is a generator for Z× . 3. (b) Show that Z× is cyclic. 36 ≡ 3 · 5 ≡ 15. so 2y1 ≡ 2y2 (mod 4). We have ker(φ) = {0. It is also easy to check that φ preserves addition. [15]30 . 13} ←→ φ(1) = 5. We have the following correspondence {0. 20. and then it follows quickly that φ is a well-deﬁned function. 6. 2. Solution: If (x. Thus (Z4 × Z6 )/{(0. and [5]30 and [25]30 . y) = (x + 2y. then x = 2. 16} ←→ φ(4) = 20. 10. and if y = 3. 37 ≡ 3 · 15 ≡ 11. φ([x]17 ) = [1]7 for all [x]17 ∈ Z× . then 2y1 − 2y2 is divisible by 12. it follows that [3] must have order 6. 5. for all [x]17 = ([3]17 )n ∈ Z× . and apply the fundamental homomorphism theorem. 0). and for this formula ﬁnd the kernel and image. since 32 = 9. (c) Completely determine all group homomorphisms from Z× into Z× . 15} ←→ φ(3) = 15. 12} ←→ φ(0) = 0. of order 1. and the order of φ([3]17 ) must be a common divisor of 16 and 6. and in 17 the second case φ(([3]17 )n ) = [−1]n . respectively. with generator [3]7 . 17 7 Solution: Any group homomorphism φ : Z× → Z× is determined by its value 17 7 on the generator [3]17 . 6. so φ must map Z4 × Z6 onto Z4 × Z3 . 38 ≡ 3 · 11 ≡ 16 ≡ 1. 0) and (2. 35 ≡ 3 · 13 ≡ 5. (2. It follows that there are 24/2 = 12 cosets of the kernel. 3. {2. and show how elements of the image correspond to cosets of the kernel. = . so either φ([3]17 ) = [1]7 or φ([3]17 ) = [−1]7 . (a) Show that φ is a well-deﬁned group homomorphism. Solution: If y1 ≡ y2 (mod 6). 12}. These cosets are in one-to-one correspondence with the elements of the image. The image of φ consists of the multiples of 5 in Z30 . so y = 0 or y = 3. 25.CHAPTER 3 SOLUTIONS 79 gcd(18. {5. 7. Thus the elements of the kernel K are (0. 7 17 20. (b) Find the kernel and image of φ. 14} ←→ φ(2) = 10. 3). {4. In the ﬁrst case. and 5. of order 6. 9. Deﬁne φ : Z4 × Z6 → Z4 × Z3 by φ(x. y). [10]30 and [20]30 . and then cosets of the kernel are deﬁned by adding 1. 10. The corresponding choices for [m]30 are [0]30 . which are 0. 3)} ∼ Z4 × Z3 . with generator [3]17 .

and this shows that φ is a one-to-one function. φ2 (x)).4. we can deﬁne group homomorphisms φ1 : Z× → pq Z× and φ2 : Z× → Z× by setting φ1 ([x]pq ) = [x]p . This completes the proof that φ is a group isomorphism. pq p q so φ is also an onto function. we can deﬁne a group homomorphism φ : Z× → Z× × Z× q pq p by setting φ([x]pq ) = (φ1 ([x]pq ). It follows that Z× and Z× × Z× have the same order. for all [x]pq ∈ Z× . and so φ([x1 ]n ) = φ([x2 ]n ). so p | (x − 1) and q | (x − 1). and apply the fundamental 36 homomorphism theorem. Let n and m be positive integers. for 36 12 all [x]36 ∈ Z× . then [x]p = [1]p and [x]q = [1]q . [13]36 . φ2 (a)φ2 (b)) = (φ1 (a). p q Solution: Using Problem 21. and this implies that pq | (x − 1). and pq q p pq φ2 ([x]pq ) = [x]q . Solution: First. φ2 (b)) = (φ1 (a)φ1 (b). for all x ∈ G. 36) = 1. 12) = 1. namely [1]36 . we have φ(ab) = (φ1 (ab).80 CHAPTER 3 SOLUTIONS 21. 36 12 The kernel of φ consists of the elements in Z× that are congruent to 1 mod 36 12. Let φ1 : G → G1 and φ2 : G → G2 be group homomorphisms. It follows that Z× ∼ Z× / [13]36 . is a well-deﬁned n m n group homomorphism. Solution: The previous problem shows that φ is a group homomorphism. φ([a]n [b]n ) = φ([ab]n ) = n [ab]m = [a]m [b]m = φ([a]n )φ([b]n ). φ is a homomorphism since for [a]n . and this implies that m | (x1 − x2 ). n then n | (x1 − x2 ). Solution: Given a. Next. φ2 (ab)) = (φ1 (a)φ1 (b). since if [x1 ]n = [x2 ]n in Z× . [25]36 . φ2 ([x]pq )). for all [x]pq ∈ Z× . It is evident that φ maps Z× onto Z× . It follows that [x]pq = [1]pq . Exercise 1. Let p and q be diﬀerent odd primes. For the group homomorphism φ : Z× → Z× deﬁned by φ([x]36 ) = [x]12 . is a well-deﬁned group homomorphism. [b]n ∈ Z× . Show that φ : Z× → Z× deﬁned by φ([x]n ) = [x]m . and G2 be groups. 12 = 36 23. G1 . since if gcd(x. Let G. since m | n.27 in the text states that if m > 0 and n > 0 are relatively prime integers. φ2 (a)φ2 (b)) φ(a)φ(b) and so φ : G → G1 × G2 is a group homomorphism. Thus [x1 ]m = [x2 ]m . then ϕ(mn) = ϕ(m)ϕ(n). φ is a well-deﬁned function. φ2 (a)) · (φ1 (b). Prove that Z× is isomorphic to the direct pq product Z× × Z× . since p adn q are relatively prime. for all [x]pq ∈ Z× . Prove that φ : G → G1 × G2 deﬁned by φ(x) = (φ1 (x). ﬁnd the kernel and image of φ. . for all [x]n ∈ Z× . pq Using Problem 23. such that m is a divisor of n. 22. 24. b in G. If [x]pq ∈ pq ker(φ). then gcd(x.

3. (2. (2. (3. 0)}. 2) + K = {(2. 15} 11 7 = {11. 9. 3)} (4. 2)} (2. 1) + H = {(3. 7. (3. (3. 2)} (5. 3). List the cosets of 7 in Z× . Solution: The cosets of H = {(0. 1) + K = {(0. 1). 3)} (5. 0). 7} 2 3 7 = {3. 15}. 0). (4. 0). with generators a of order n and b of order 2. (4. 0) + K = {(1. 2). 1)} (2. 2) + K = {(1. 1) + H = {(4. Thus the order of an element in G/H is at most 6. 1). (a) List all cosets of H. 3). 0). 3)} (0. 0). 1) + H = {(0. 1) + H = {(2. 1). and so G/H ∼ Z6 × Z2 . the coset 3 7 does not have order 2. 3)} (1. (4. Which answer is correct for G/H? For G/K? Solution: Adding an element of G to itself 6 times yields a 0 in the ﬁrst component and either 0 or 2 in the second component. 0)} (1. 0)} (0. (0. 0) + K = {(2. 16 7 = {1. list all cosets of K. 0). . 0) + H = {(4. 1) + K = {(2. 2). 1) + K = {(1. showing that the factor group is cyclic. 3) + K = {(1. producing an element in H. 0) + H = {(2. 1). 1) + H = {(5. 1) + K has order 12 in G/K. satisfying ba = a−1 b. 2)} are (0. 0)} are (b) You may assume that any abelian group of order 12 is isomorphic to either Z12 or Z6 × Z2 . (5. Let the dihedral group Dn be given via generators and relations. (1. (3. (4. and let K = {(0. (5. On the other hand. (0. 28. 5. 0) + H = {(5. we have bak+1 = bak a = a−k ba = a−k a−1 b = a−(k+1) b. (3. (3. 2)}. (1. (5. 2)} (2. 0). 3)} (0. 0) + H = {(3. 0). (0. let H = {(0. 2) + K = {(0. 0)} (2. (5. 3)} (2. Let G = Z6 × Z4 . (a) Show that bai = a−i b for all i with 1 ≤ i < n. 1)} (0. 1). 1)} (1. 13. 2)} (1. 2)} (0. 3)} The cosets of K = {(0. 0). and so G/K = 29. 0). (1. 1) + H = {(1. 2)} (1. 1). 0) + H = {(1. (4. 2). 3)} (3. 2)} (4. 1).8 SOLUTIONS 27. 13} Since 3 ∈ 7 . = ∼ Z12 . Is the factor group Z× / 7 cyclic? 16 16 Solution: Z× = {1. (4. 0). (0. 1). 2)} (1. 5} 9 7 = {9. 11. so it must have order 4. (5. 0). 3) + K = {(0. 3)} (2.CHAPTER 3 SOLUTIONS 81 3. (3. (5. Solution: The identity holds for all positive integers i. 0) + K = {(0. 0). 3). (3. 2)} (3. 3) + K = {(2. 1). and can be proved inductively: assuming bak = a−k b. 0) + H = {(0.

a9 }. (b) Is G/N abelian? Solution: For aN = {a. (a) Prove that N is a normal subgroup of G. (c) List all left cosets and all right cosets of b Solution: The left cosets of b have the form ai b = {ai . Thus ∼ is an equivalence relation. for 0 ≤ i < n. a4 b. so c = g(hah−1 )g −1 = (gh)a(gh)−1 . while (bN )(aN ) = baN = a5 bN = {a5 b. (a) Show that N is a normal subgroup of G.) The cosets of N are N = {e. a11 }. and let N = {e. and list all cosets of N . since N aN b = N ab but N bN a = N a2 b. written b ∼ a. and G/N is not abelian. a3 b}. N ab = {ab. (b) You may assume that G/N is isomorphic to either Z6 or S3 . a8 . a6 . The equivalence classes of ∼ are called the conjugacy classes of G. If b ∼ a. N a = {a. Thus G/N ∼ S3 . Which is correct? Solution: The factor group G/N is not abelian. h ∈ G. a8 b. 30. and so a = g −1 bg = g −1 b(g −1 )−1 . a5 b. For a. a4 . Let G be the dihedral group D12 .82 CHAPTER 3 SOLUTIONS (b) Show that any element of the form ai b has order 2. It is normal since ai (a3n )a−i = a3n and ai b(a3n )ai b = ai a−3n a−i = (a3n )−1 . a3 . Show that ∼ is an equivalence relation on G. Thus (aN )(bN ) = (bN )(aN ). a6 b. . (a) Let G be a group. a9 b}. which shows that a ∼ b. it is a subgroup. a2 b}. a5 . a10 b}. a3 } of G. for 0 ≤ i < n. N a2 b = {a2 b. a3 b. (We are using the fact that bai = a−i b. which shows that c ∼ a. Solution: We have (ai b)2 = ai bai b = ai a−i b2 = a0 = e. Solution: Since N = a3 . a7 . Solution: We have a ∼ a since we can use g = e. then c = gbg −1 and b = hah−1 for some g. The right cosets of b have the form b ai = {ai . a4 b}. a10 }. a4 } and bN = {b. a11 b}. N a2 = {a2 . If c ∼ b and b ∼ a. we have (aN )(bN ) = abN = {ab. a3 . if there exists g ∈ G such that b = gag −1 . because ba = a11 b ∈ N a2 b. a7 b. Let G = D6 and let N be the subgroup a3 = {e. ai b}. a9 }. a6 . 31. a−i b}. N b = {ab. Solution: The argument is the same as in the previous problem. the b = gag −1 for some g ∈ G. (b) Show that a subgroup N of G is normal in G if and only if N is a union of conjugacy classes. = 32. b ∈ G we say that b is conjugate to a.

b2 = e. y ∈ H. b. 34. h2 ∈ H and some n1 . then xbx−1 = ai ba−i = ai ai b = a2i b. which contains it. and let N and H be subgroups of G such that N is normal in G. But then hnN = hN = φ(h). the only element conjugate to e is e itself. then xabx−1 = (ai b)ab(ai b)−1 = ai a−1 ai b = a2i−1 b. If x is any power of a. then HN/N is isomorphic to H. for all g ∈ G. a3 . Let D4 = {e. We have xy −1 = h1 n1 (h2 n2 )−1 = h1 n1 n−1 h−1 = (h1 h−1 )(h2 (n1 n−1 )h−1 ). 2 2 2 2 2 and this element belongs to HN since the assumption that N is normal guarantees that h2 (n1 n−1 )h−1 ∈ N . a3 b}. and ba = a−1 b. so this shows that a3 is the only conjugate of a (other than a itself). Thus a2 b is the only conjugate of b. 33. (c) Prove that if H ∩ N = {e}. Then x = h1 n1 and y = h2 n2 . and so h = e. so HN is nonempty. then x(ab)x−1 = ai aba−i = ai+1 ai b = a2i+1 b. Any coset of N in HN has the form hnN for some h ∈ H and some n ∈ N . Solution: Remember: the notion of a conjugacy class was just deﬁned in the previous exercise. ab. If x = ai b.2. If x = ai b. If x = ai . (Deﬁning a function from HN/N into H is more complicated. H ∩ N = {e}. Thus a3 b is the only conjugate of ab. Find the conjugacy classes of D4 . If x = ai b. It is clear that e = e · e belongs to the set HN . then x commutes with a. Since xex−1 = e. φ is one-to-one since if h ∈ H belongs to the kernel of φ. so a2 is not conjugate to any other element. 2 2 (b) Prove that N is a normal subgroup of HN . a2 . (a) Prove that HN is a subgroup of G. The solution of an earlier problem shows that xa2 x−1 = a2 in D4 . By assumption. for some h1 . Solution: See Proposition 3. and so this shows that φ is onto. Suppose that x. it is normal in the subgroup HN . If x = ai . n2 ∈ N . Finally.CHAPTER 3 SOLUTIONS 83 Solution: The subgroup N is normal in G if and only if a ∈ N implies gag −1 ∈ G. then xax−1 = ai baa−i b = ai ai−1 b2 = a2i−1 . Solution: Deﬁne φ : H → HN/N by φ(x) = xN for all x ∈ H. Let G be a group. a. and so xax−1 = a. and so h ∈ N . . with a4 = e. y belong to HN . Thus N is normal if and only if whenever it contains an element a it also contains the conjugacy class of a. then xbx−1 = (ai b)b(ai b)−1 = ai ai b = a2i b. Another way to say this is that N is a union of conjugacy classes. Solution: Since N is normal in G. a2 b.) Then φ(xy) = xyN = xN yN = φ(x)φ(y) for all x. then hN = φ(h) = N .3.

[5]. 25 ≡ 2 · 24 ≡ 6. 23 = 8. [4]. [3]} and [2]5 = {[1]. 29 = 6. Show that the group G in the previous problem is isomorphic to the direct product R × R. in shorthand notation: 22 = 4. so the set is closed under taking in1 Show that G is a subgroup of 1 a Solution: We have 0 1 0 0 the closure property holds. and we have 22 = 4. 26 = 9. so the possible orders are 1. Find all subgroups of Z× . This shows that Z× is cyclic. and give the lattice diagram which shows the 11 inclusions between them. Let G be the subgroup of 1 0 0 GL3 (R) consisting of all matrices of the form a b 1 0 such that a. b 1 c d 1 a+c b+d 0 0 1 0 = 0 1 0 . and −b 0 . so the subgroups are as follows. We could 13 also write this as Z× = [2]13 . [2]2 . [2]8 } = {[1]. and the order of any element must be 13 a divisor of 12. 23 = 8. [2]4 . so the order of [2] is greater than 6. 4. 4. 6. so 1 0 0 1 0 0 1 The identity matrix belongs to the set. 1 a b Solution: Deﬁne φ : G → R × R by φ 0 1 0 = (a. and thus [2] is a generator for Z× . 28 = 3. 210 = 1. 3. and 12.84 CHAPTER 3 SOLUTIONS SOLUTIONS TO THE REVIEW PROBLEMS 1. (b) Show that Z× is a cyclic group. Solution: The group Z× has order 12. 13 Solution: The ﬁrst element to try is [2]. −1 1 a b 1 −a 0 1 0 1 = 0 0 0 1 0 0 verses. 2. By part (a) it must be 12. 27 = 7. [10]} The lattice diagram forms a diamond. Solution: First check for cyclic subgroups. in addition to Z× 11 11 and {[1]}: [2]2 = {[1]. (a) What are the possibilities for the order of an element of Z× ? Explain 13 your answer. 0 1 GL3 (R). 24 = 5. [2]6 . This is 0 0 1 one-to-one and onto because it has an inverse function θ : R × R → G deﬁned . b). [2]5 } = {[1]. [9]. 25 = 10. 3. b ∈ R . 13 2. 24 = 16 ≡ 3. and 26 ≡ 2 · 25 ≡ 12.

using the argument given for part (b). and −1 1 b 1 c 1 b 1 −c it is a subgroup since = = 0 1 0 1 0 1 0 1 1 b−c 0 1 m 0 . so N = ker(φ). 0 0 1 0 0 1 5. Solution: The set N is nonempty since it contains the identity matrix. 0 1 The fundamental homomorphism theorem implies that G/N ∼ R× . 6. ±9}. b) + (c. Finally.CHAPTER 3 SOLUTIONS 85 1 a b by θ((a. ±7. b)) = 0 1 0 . Thus parts (a) and (b) can be proved at the same time. 20 9 = {1. ±3. and let N be the subset of all matrices of the form . 20 (−3) 9 = {−3. d) = φ 0 1 0 + φ 0 1 0 . and φ Note that this part of the proof covers part (a). b + d) = (a. Is Z× / 9 cyclic? 20 20 Solution: Z× = {±1. it is always a normal subgroup. N is normal in G since 1/m −b/m 0 1 = m b 0 1 1 mc 0 1 1 0 ∈ N. −9} 3 9 = {3. Then we have 0 1 mc + b = mn = 1 . and that N is normal in G. 9} (−1) 9 = {−1. List the cosets of the cyclic subgroup 9 in Z× . Let G be the subgroup of GL2 (R) consisting of all matrices of the form m b 1 b . φ preserves the respective operations 0 0 1 1 a b 1 c d 1 a+c b+d 1 0 = since φ 0 1 0 0 1 0 = φ 0 0 0 1 0 0 1 0 0 1 1 a b 1 c d (a + c. since once you have determined the kernel. . = maps G onto R× . 7} Z× . c 1 m b 0 1 −1 = mc + b 1 (b) Show that G/N is isomorphic to the multiplicative group R× . Solution: φ φ m b 0 1 m b 0 1 Deﬁne φ : G → R× by φ n 0 φ c 1 n 0 =φ c 1 mn 0 m b = m. φ m b = 1 if and only if m = 1. for each element x of the factor group is not cyclic. 0 1 0 1 (a) Show that N is a subgroup of G. −7} Since x2 ∈ 9 . Since m can be any nonzero real number.

. and N ab = {ab. Let G = D8 . a2 . a. a6 }. a4 . ab. a3 b. so G/N is not cyclic. a4 b. a3 b}. 8. a5 b. bN = {b. Let N be the subgroup a2 = {e. a4 b. a6 }. a7 }. N a = {a. (b) Is the factor group D4 /N a cyclic group? Solution: The cosets of N are N = {e. a3 b}. Assume that the dihedral group D4 is given as {e. and ba = a3 b. a7 b. but is not cyclic. a3 . b. a2 . a3 . a3 . N ab = {ab. abN = {ab. a3 b}. a6 b}. a5 . N bN b = N e = N . a2 }. N a = {a.86 CHAPTER 3 SOLUTIONS 7. (a) List all left cosets and all right cosets of N . and verify that N is a normal subgroup of G. a5 b. which implies that N is normal. b2 = e. a3 }. but we get The fact that the left and right cosets of N coincide shows that N is normal. The left cosets of N are more trouble to compute. Solution: The right cosets of N are N = {e. and let N = {e. a7 }. Solution: We have ai a2 a−i = a2 and (ai b)a2 (ai b)−1 = ai a−2 bai b = ai a−2 a−i b2 = a−2 = a2 . a2 b. where a4 = e. so each coset has order 2. N b = {b. a2 b}. We have N aN a = N a2 = N . a2 b. for all i. a6 }. aN = {a. and a2 ∈ N . we see that each element in the factor group has order 2. a7 b}. (a) Show by a direct computation that N is a normal subgroup of D4 . a4 . a2 }. a6 b. Since b and ab have order 2. a4 . and N abN ab = N e = N . a2 . N = {e. a5 . a2 . a2 b}. Solution: It is clear that there are 4 cosets. N b = {b. (b) Show that G/N has order 4.

4. Over the ﬁeld of rational numbers. Solution: Let x8 −1 = f (x) and x6 −1 = g(x). Using the Euclidean algorithm. Over the ﬁeld of rational numbers. and g(x) = (x4 + x2 + 1)(x2 − 1). and then the next step yields g(x) = (x − 1)(x2 + x + 1).Chapter 4 Polynomials SOLUTIONS TO THE REVIEW PROBLEMS 1. 2. let 2x4 − x3 + x2 + 3x + 1 = f (x) and 2x3 − 3x2 + 2x + 2 = g(x). ﬁnd the greatest common divisor of x4 + x3 + 2x2 + x + 1 and x3 − 1. g(x)) = 1. Solution: To simplify the computations. use the Euclidean algorithm to show that 2x3 − 2x2 − 3x + 1 and 2x2 − x − 2 are relatively prime. so this shows that gcd(x8 − 1. Over the ﬁeld of rational numbers. x6 − 1) in Q[x] and write it as a linear combination of x8 − 1 and x6 − 1. Solution: Let 2x3 − 2x2 − 3x + 1 = f (x) and 2x2 − x − 2 = g(x). The constant remainder at the second step implies that gcd(f (x). Solution: Let x4 + x3 + 2x2 + x + 1 = f (x) and x3 − 1 = g(x). so gcd(f (x). g(x)) = x2 + x + 1. and (x2 + x + 1) = 1 1 2 f (x) − 2 (x + 1)g(x). and then g(x) = (2x − 1)g(x) − 2. and x2 − 1 = f (x) − x2 g(x). 3. we ﬁrst obtain 87 . x6 − 1) = x2 − 1. and express it as a linear combination of the given polynomials. We ﬁrst obtain f (x) = (x + 1)g(x) + 2(x2 + x + 1). We have f (x) = x2 g(x)+(x2 − 1). Use the Euclidean algorithm to ﬁnd gcd(x8 − 1. At the next step we can use x rather than 2 2 3 2 x. We ﬁrst obtain f (x) = (x − 1 )g(x) − 3 x. ﬁnd the greatest common divisor of 2x4 − x3 + x2 + 3x + 1 and 2x3 − 3x2 + 2x + 2 and express it as a linear combination of the given polynomials.

The second factor is irreducible over Q since it satisﬁes Eisenstein’s criterion for p = 3. Are the following polynomials irreducible over Q? (a) 3x5 + 18x2 + 24x + 6 Solution: Dividing by 3 we obtain x5 + 6x2 + 8x + 2. The only possible rational roots of the second factor are 1 and −1. in which . Beginning with the last equation and back-solving.88 CHAPTER 4 SOLUTIONS f (x) = (x+1)g(x)+(2x2 −x−1). ±12. or b = d = −1. At the next step we obtain 2x2 − x − 1 = (x − 1)(2x + 1). and then g(x) = (x−1)(2x2 −x−1)+(2x+1).) Since the polynomial has no linear factors. We have either b = d = 1. in which case a + c = 8. (It is important to note that since the degree of the polynomial is greater than 3. ±6. ±2. 2 2 5. so 2x + 1 is the greatest common divisor (we must then divide by 2 to make it monic). ad + bc = 8. and since 2 is a root we have the factorization x5 − 2x4 − 2x3 + 12x2 − 15x − 2 = (x − 2)(x4 − 2x2 + 8x + 1). ±2. Factor x5 − 10x4 + 24x3 + 9x2 − 33x − 12 over Q. so this eliminates all but ±2. 7. which is irreducible in Z2 [x]. f (−2) = −294. This leads to the equations a + c = 0. Solution: The possible rational roots of f (x) = x5 −10x4 +24x3 +9x2 −33x−12 are ±1. Then f (2) = 32. and these do not work. we get 2x + 1 = = = = g(x) − (x − 1)(2x2 − x − 1) g(x) − (x − 1)(f (x) − (x + 1)g(x)) g(x) + (x2 − 1)g(x) − (x − 1)f (x) x2 g(x) − (x − 1)f (x) 1 2 This gives the ﬁnal answer. We have f (1) = 21. the fact that it has not roots in Q does not mean that it is irreducible over Q. x + = 1 x2 g(x) + (− 1 )(x − 1)f (x). 6. the only possible factorization has the form x4 −2x2 +8x+1 = (x2 +ax+b)(x2 +cx+d). ac + b + d = −2. Factor x5 − 2x4 − 2x3 + 12x2 − 15x − 2 over Q. 4. ±4. (c) 2x10 + 25x3 + 10x2 − 30 Solution: Eisenstein’s criterion is satisﬁed for p = 5. −6 as possibilities. (b) 7x3 + 12x2 + 3x + 45 Solution: Reducing the coeﬃcients modulo 2 gives the polynomial x3 + x + 1. ±3. Solution: The possible rational roots are ±1. This implies that the polynomial is irreducible over Q. so for any root we must have (r − 1)|21. and this satisﬁes Eisenstein’s criterion for p = 2. and bd = 1. and ﬁnally we obtain the factorization f (x) = (x − 4)(x4 − 6x3 + 9x + 3).

[x].CHAPTER 4 SOLUTIONS 89 case a + c = −8. Solution: We ﬁrst factor out 2. p(−1) = 3. Solution: The congruence classes are in one-to-one correspondence with the linear polynomials. and so [x + 1]4 = [−x]2 = [−1]. Solution: In general. [2x + 1]. On the other hand. 9. so x4 − 2x2 + 8x + 1 is irreducible over Q. or p(x) = (x2 + x + 2)(x2 + 2x + 2). we have x4 + x = x(x3 + 1) = x(x + 1)(x2 − x + 1). The only remaining possibility (by Lagrange’s theorem) is that [x + 1] has order 8. and p(−2) ≡ −2. p(x) = (x2 + 2x + 2)2 . p(1) = 1. p(1) = 1. Express 2x3 + x2 + 2x + 2 as a product of polynomials irreducible over Z5 . (c) In the multiplicative group of nonzero elements of F . (We could also multiply each term by 3. p(2) = 4. [x + 1]2 = [x2 + 2x + 1] = [−1 + 2x + 1] = [2x] = [−x]. In the ﬁrst the coeﬃcient of x is 1. the only possible factorizations are p(x) = (x2 + x + 2)2 . so we have the nine elements [0]. This reduces the question to factoring p(x) = x3 − 2x2 + x + 1. p(−1) = −3. Solution: If p(x) = x3 + 2x2 + 3. which shows that [x + 1] does not have order 2 or 4. (a) Express x4 + x as a product of polynomials irreducible over Z5 . in the third the coeﬃcient of x is 0. p(2) = 3. p(−2) = 2. we only need to check that it has no roots in Z3 . As an alternate solution.) Checking for roots shows that p(0) = 1. [1]. Solution: The multiplicative group of F has 8 elements. 10. (b) List the elements of the ﬁeld F = Z3 [x]/ x2 + 1 . The factor p(x) = x2 − x + 1 is irreducible over Z5 since it can be checked that it has no roots in Z5 . but [x] is not. and hence over Q. the second has a nonzero cubic term. Since the constant term of p(x) is 1. using (2)(−2) = −4 ≡ 1 (mod 5). Either case contradicts a + c = 0. so the only possible factors are of degree 2. The monic irreducible polynomials of degree 2 over Z3 are x2 + 1. . p(1) = 2. [2x]. it follows that [x] has order 4 and is not a generator.) (b) Show that x3 + 2x2 + 3 is irreducible over Z5 . (We get p(0) = 1. and x2 + 2x + 2. and p(−1) = 2. [x+2]. show that [x + 1] is a generator. Thus p(x) is irreducible over Z3 . 8. p(−1) = −1. [2x + 2]. and so it is a generator for the multiplicative group of F . and this follows from the computations p(0) = 1. and p(−2) = 3. we could reduce x4 − 2x2 + 8x + 1 modulo 3 to get p(x) = x4 + x2 + 2x + 1. This polynomial has no roots in Z3 . p(1) = 1. x2 + x + 2. [x+1]. Solution: To show that p(x) = x2 + 1 is irreducible over Z3 . [2]. then p(0) = 3. so p(x) itself is irreducible over Z5 . so p(x) is irreducible over Z5 . (a) Show that x2 + 1 is irreducible over Z3 . p(2) = 3. and since [x]2 = [−1].

90

CHAPTER 4 SOLUTIONS

11. Construct an example of a ﬁeld with 343 = 73 elements. Solution: We only need to ﬁnd a cubic polynomial over Z7 that has no roots. The simplest case would be to look for a polynomial of the form x3 + a. The cube of any element of Z7 gives either 1 or −1, so x3 = 2 has no root over Z7 , and thus p(x) = x3 −2 is an irreducible cubic over Z7 . Using the modulus p(x), the elements of Z7 [x]/ p(x) correspond to polynomials of degree 2 or less, giving the required 73 elements. With this modulus, the identities necessary to determine multiplication are [x3 ] = [5] and [x4 ] = [5x]. 12. In Z2 [x]/ x3 + x + 1 , ﬁnd the multiplicative inverse of [x + 1]. Solution: We ﬁrst give a solution using the Euclidean algorithm. For p(x) = x3 + x + 1 and f (x) = x + 1, the ﬁrst step of the Euclidean algorithm gives p(x) = (x2 +x)f (x)+1. Thus p(x)−(x2 +x)f (x) = 1, and so reducing modulo p(x) gives [−x2 − x][f (x)] = [1], and thus [x + 1]−1 = [−x2 − x] = [x2 + x]. We next give an alternate solution, which uses the identity [x3 ] = [x + 1] to solve a system of equations. We need to solve [1] = [x + 1][ax2 + bx + c] or [1] = = = = [ax3 + bx2 + cx + ax2 + bx + c] [ax3 + (a + b)x2 + (b + c)x + c] [a(x + 1) + (a + b)x2 + (b + c)x + c] [(a + b)x2 + (a + b + c)x + (a + c)] ,

so we need a + b ≡ 0 (mod 2), a + b + c ≡ 0 (mod 2), and a + c ≡ 1 (mod 2). This gives c ≡ 0 (mod 2), and therefore a ≡ 1 (mod 2), and then b ≡ 1 (mod 2). Again, we see that [x + 1]−1 = [x2 + x]. 13. Find the multiplicative inverse of [x2 + x + 1] (a) in Q[x]/ x3 − 2 ; Solution: Using the Euclidean algorithm, we have x3 − 2 = (x2 + x + 1)(x − 1) + (−1), and so [x2 + x + 1]−1 = [x − 1]. This can also be done by solving a system of 3 equations in 3 unknowns. (b) in Z3 [x]/ x3 + 2x2 + x + 1 . Solution: Using the Euclidean algorithm, we have x3 + 2x2 + x + 1 = (x + 1)(x2 + x + 1) + (−x) and x2 + x + 1 = (−x − 1)(−x) + 1. Then a substitution gives us 1 = (x2 + x + 1) + (x + 1)(−x) = (x2 + x + 1) + (x + 1)((x3 + 2x2 + x + 1) − (x + 1)(x2 + x + 1)) = (−x2 − 2x)(x2 + x + 1) + (x + 1)(x3 + x2 + 2x + 1) .

Thus [x2 + x + 1]−1 = [−x2 − 2x] = [2x2 + x]. This can be checked by ﬁnding [x2 + x + 1][2x2 + x], using the identities [x3 ] = [x2 − x − 1] and [x4 ] = [x − 1].

CHAPTER 4 SOLUTIONS

91

This can also be done by solving a system of equations, or, since the set is ﬁnite, by taking successive powers of [x2 + x + 1]. The latter method isn’t really practical, since the multiplicative group has order 26, and this element turns out to have order 13. 14. In Z5 [x]/ x3 + x + 1 , ﬁnd [x]−1 and [x + 1]−1 , and use your answers to ﬁnd [x2 + x]−1 . Solution: Using the division algorithm, we obtain x3 + x + 1 = x(x2 + 1) + 1, and so [x][x2 + 1] = [−1]. Thus [x]−1 = [−x2 − 1]. Next, we have x3 +x+1 = (x+1)(x2 −x+2)−1, and so [x+1]−1 = [x2 −x+2]. Finally, we have [x2 + x]−1 = [x]−1 [x + 1]−1 = [−x2 − 1][x2 − x + 2] = [−x4 + x3 − 2x2 − x2 + x − 2] .

Using the identities [x3 ] = [−x − 1] and [x4 ] = [−x2 − x], this reduces to [x2 + x]−1 = [x2 + x − x − 1 − 3x2 + x − 2] = [−2x2 + x − 3] = [3x2 + x + 2] .

15. Factor x4 + x + 1 over Z2 [x]/ x4 + x + 1 . Solution: There are 4 roots of x4 + x + 1 in the given ﬁeld, given by the cosets corresponding to x, x2 , x + 1, x2 + 1. This can be shown by using the multiplication table, with the elements in the form 10, 100, 11, and 101, or by computing with polynomials, using the fact that (a + b)2 = a2 + b2 since 2ab = 0. We have x4 + x + 1 ≡ 0, (x2 )4 + (x2 ) + 1 = (x4 )2 + x2 + 1 ≡ (x + 1)2 + x2 + 1 ≡ x2 + 1 + x2 + 1 ≡ 0, (x + 1)4 + (x + 1) + 1 ≡ x4 + 1 + x ≡ x + 1 + 1 + x ≡ 0, and (x2 +1)4 +(x2 +1)+1 ≡ (x4 )2 +1+x2 ≡ (x+1)2 +1+x2 ≡ x2 +1+1+x2 ≡ 0. Thus x4 + x + 1 factors as a product of 4 linear terms.

92

CHAPTER 4 SOLUTIONS

(c) Find all idempotent elements of R. c ↔ z. and these are invertible since their determinant is nonzero. and the following computation checks closure under multiplication. 93 . the above computation also checks commutativity. (a) Show that R is a commutative ring (you only need to check closure and commutativity of multiplication). ax 0 0 x 0 0 a 0 0 0 a 0 0 x 0 = 0 ax 0 b c a y z x bx + ay cx + az ax Because of the symmetry a ↔ x. (b) Find all units of R. Solution: It is clear that the set is closed under addition. and so they are nilpotent. Solution: Four of the matrices in R have 1’s on the diagonal. Squaring each of the other four matrices gives the zero matrix.Chapter 5 Commutative Rings SOLUTIONS TO THE REVIEW PROBLEMS 1. and all nilpotent elements of R. b ↔ y. Let R be the ring with 8 elements consisting of all 3 × 3 matrices with entries in Z2 which have the following form: a 0 0 0 a 0 b c a You may assume that the standard laws for addition and multiplication of matrices are valid.

ﬁnd Ann(a) = {r ∈ R | ra = 0}. so the ﬁrst component x can be any element of Z12 .4 that the ring Z180 is isomorphic to the ring Z4 ⊕ Z9 ⊕ Z5 . 7. and this implies a = 0 and b = 0 since Z2 is a ﬁeld. 7. then the image is 7Z42 and the kernel is 6Z120 . For the element a = (0. the only idempotent element is the identity) and the only nilpotent element that is also idempotent is the zero matrix. If φ(1) = 21. 2) = (0. Then Example 5. If φ(1) = 7. Solution: In R we have a + a = 0. while in Z3 ⊕ Z3 the equation (a.5. then the image is 14Z42 and the kernel is 3Z120 . Furthermore. Solution: We need to solve (x. 42) = 6. It follows that the largest possible order of an element is lcm[2. only 0. an element in R is either a unit or nilpotent. for all a ∈ R. Are Z9 and Z3 ⊕ Z3 isomorphic as rings? Solution: The answer is no. 28 are idempotent. y) ∈ Z12 ⊕ Z8 . b)2 = (0. 2. Show that although R has 4 elements. 4] = 12. so R is not isomorphic to Z4 . Find all ring homomorphisms from Z120 into Z42 .10 shows that Z× ∼ Z× × Z× × Z× ∼ Z2 × Z6 × Z4 . then the image is 21Z42 and the kernel is 2Z120 . Show that Ann(a) is an ideal of R. 6. 4. what is the largest possible order 180 of an element? Solution: Since 180 = 22 32 5. since in Z9 there is a nonzero solution to the equation x2 = 0. and then a = 0 and b = 0.) 5. The only unit that is idempotent is the identity matrix (in a group. 0) implies a2 = 0 and b2 = 0. The argument can be given using either addition or multiplication. 3. (In Z9 let x = 3.94 CHAPTER 5 SOLUTIONS Solution: By part (b). 21. the order of an element is the least common multiple of the orders of its components. in R we have [x + 1] = [0] but [x + 1]2 = [x2 + 1] = [0]. We only need 2y ≡ 0 (mod 8). it is not isomorphic to either of the rings Z4 or Z2 ⊕ Z2 .2. 2) of the ring R = Z12 ⊕ Z8 . 180 = 4 9 5 = In the latter additive group. 0) for (x. and it can be checked that in 7Z42 . 21. it follows from Theorem 3. . φ(1) must be idempotent. Multiplication is also diﬀerent. since the additive group of Z9 is cyclic. On the other hand. The additive order of φ(1) must be a divisor of gcd(120. 6. b)2 = (0. Let R be the ring Z2 [x]/ x2 + 1 . In the group Z× of units of the ring Z180 . 28. so it must belong to the subgroup 7Z42 = {0. while that of Z3 ⊕ Z3 is not. 14. since in that ring (a. y)(0. 35}. If φ(1) = 28. while in Z3 ⊕ Z3 there is not. Solution: Let φ : Z120 → Z42 be a ring homomorphism. Addition in the two rings is diﬀerent. Thus R cannot be isomorphic to Z2 ⊕ Z2 . 0) implies a2 = 0 and b2 = 0.

the maximal ideals that contain 36Z are 2Z and 3Z. Thus we only need to ﬁnd the maximal ideals of Z36 . 1) of Z × Z are zero divisors. It is not diﬃcult to show that φ is an onto ring homomorphism. = Solution: Deﬁne φ : Z2 [x]/ x4 + 1 → Z2 [x]/ x2 + 1 by φ(f (x) + x4 + 1 ) = (f (x) + x2 + 1 ). 10. Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients.3. 7. Solution: The elements (1. Solution: If P is a prime ideal of Z36 . prove that I is not maximal. 4. Again. This set is certainly closed under addition. 9. and hence P is maximal. Let R be the ring Z2 [x]/ x4 + 1 . The lattice of ideals of Z36 is exactly the same as the lattice of subgroups. of Z36 = Z/36Z. which is not a ﬁeld. Hint: If you use the fundamental homomorphism theorem. and all prime ideals. so the relevant ideals correspond to the divisors of 36. The maximal ideals of Z36 are thus 2Z36 and 3Z36 . with kernel equal to I. This mapping is well-deﬁned since x2 + 1 is a factor of x4 + 1 over Z2 . 0) and (0. and let I be the set of all congruence classes in R of the form [f (x)(x2 + 1)]. In Z every ideal is principal. Give an example to show that the set of all zero divisors of a ring need not be an ideal of the ring. This is an onto ring homomorphism with kernel I. so the maximal ideals of Z36 correspond to the prime divisors of 36. which shows that there is a one-to-one correspondence between the ideals of Z/36Z and the ideals of Z that contain 36Z. 8. an obvious contradiction. so I is not maximal. you can do the ﬁrst two parts together. 2)) = Z12 ⊕ 4Z8 . and these correspond to 2Z36 and 3Z36 . (a) Show that I is an ideal of R. Then R/I is isomorphic to Z2 [x]. 1). Find all maximal ideals. . so it is a ﬁeld. and it is also closed under multiplication by any element of R since 4Z8 is an ideal of Z8 . but if the set of zero divisors were closed under addition it would include (1. (b) Show that R/I ∼ Z2 [x]/ x2 + 1 . Solution: Deﬁne φ : Z[x] → Z2 [x] by reducing coeﬃcients modulo 2.CHAPTER 5 SOLUTIONS 95 while y = 0. (c) Is I a prime ideal of R? Solution: No: (x + 1)(x + 1) ≡ 0 (mod x2 + 1). then Z36 /P is a ﬁnite integral domain. An alternate approach we can use Proposition 5. Thus Ann((0. Prove that I is a prime ideal.7.

[x + 1][x2 + x + 1] = [x3 + 1] = [0]. Let R be the ring Z2 [x]/ x3 + 1 . so the only proper. This also excludes [x2 + x] = [x][x + 1] and [x2 + 1] = [x2 ][1 + x]. it is easy to check that φ preserves addition. then R ∼ Re ⊕ R(1 − e). (a) Find all ideals of R. (b) Show that if e is idempotent. the ideals of R correspond to the ideals of Z2 [x] that contain x3 + 1 . s ∈ R we have φ(rs) = (rse. nonzero ideals are the principal ideals generated by [x + 1] and [x2 + x + 1]. Solution: Note: Table 5.1: Multiplication in Z2 [x]/ x3 + 1 × 1 x x2 1 1 x x2 2 x x x 1 x2 x2 1 x x2 + x + 1 x2 + x + 1 x2 + x + 1 x2 + x + 1 x2 + x x2 + x x2 + 1 x+1 x+1 x+1 x2 + x x2 + 1 x2 + 1 x2 + 1 x+1 x2 + x x2 + x + 1 x2 + x + 1 x2 + x + 1 x2 + x + 1 x2 + x + 1 0 0 0 x2 + x x2 + x x2 + 1 x+1 0 x2 + x x+1 x2 + 1 x+1 x+1 x2 + x x2 + 1 0 x+1 x2 + 1 x2 + x x2 + 1 x2 + 1 x+1 x2 + x 0 x2 + 1 x2 + x x+1 to compute the multiplication table in order to solve the problem. Thus the only units are 1. . We have the factorization x3 + 1 = x3 − 1 = (x − 1)(x2 + x + 1). Deﬁne φ : R → Re⊕R(1−e) by φ(r) = (re. rs(1 − e)). so [x] and [x2 ] are units.3. Furthermore. (b) Find the units of R. rs(1 − e)2 ) = (rse. = Solution: Note that e(1−e) = e−e2 = e−e = 0. CHAPTER 5 SOLUTIONS (a) Show that if e is an idempotent element of R. r(1−e)). φ is onto. Solution: We have (1−e)2 = (1−e)(1−e) = 1−e−e+e2 = 1−e−e+e = 1−e. (c) Find the idempotent elements of R. 12. since for any element (ae.1 gives the multiplication table. b(1 − e)) we have (ae. Solution: By Proposition 5. and adding the two equations gives r = s. It is not necessary Table 5. s(1 − e)) = (rse2 . Then φ is one-to-one since if φ(r) = φ(s). and [x2 ]. On the other hand. and for any r.96 11. so [x + 1] and [x2 + x + 1] cannot be units. then 1−e is also idempotent. Let R be any commutative ring with identity 1. r(1 − e))(se. Solution: We have [x]3 = [1]. then re = se and r(1 − e) = s(1 − e). [x]. for all r ∈ R. Finally.7. b(1 − e)) = φ(r) for r = ae + b(1 − e). rs(1 − e)) and φ(r)φ(s) = (re.

[x2 + 1]} [x + 1] = {[0]. (a) Find all ideals of S. it is easy to see that the idempotent elements of R are [0]. [x2 + x]} . (c) Find the idempotent elements of R. [x]. They cannot be isomorphic as rings since R has 3 units. 14. 13. Solution: Since no unit can belong to a proper ideal. Let S be the ring Z2 [x]/ x3 + x . as abelian groups. [x2 + x] = {[0]. [x2 ]. together with [0] and [1]. 15.2: Multiplication in Z2 [x]/ x3 + x × 1 x2 + x + 1 x2 2 1 1 x +x+1 x2 2 2 x +x+1 x +x+1 1 x2 2 2 2 x x x x2 x x x x x2 + x x2 + x x2 + x x2 + x x+1 x+1 x+1 x2 + x x2 + 1 x2 + 1 x2 + 1 0 x x x x x2 x2 + x x2 + x 0 x2 + x x2 + x x2 + x x2 + x x2 + x 0 0 0 x+1 x+1 x+1 x2 + x x2 + x 0 x2 + 1 x2 + 1 x2 + 1 x2 + 1 x2 + 1 0 0 0 x2 + 1 x2 + 1 to compute the multiplication table in order to solve the problem. so by Proposition 5.7 the proper nonzero ideals of S are the principal ideals generated by [x]. [x2 + 1] = {[0]. [x2 + 1] = [x + 1]2 . Solution: Over Z2 we have the factorization x3 + x = x(x2 + 1) = x(x + 1)2 . we have [x2 ]2 = [x2 ]. [x2 + 1]. Solution: Note: Table 5. [x2 + x]} (b) Find the units of R. This is a unit since [x2 +x+1]2 = [1]. [x + 1]. [1]. while S has only 2. Solution: Both R and S are isomorphic to Z2 × Z2 × Z2 . [x2 + x + 1]. it follows from part (a) that we only need to check [x2 +x+1]. It is not necessary Table 5. Let Z[i] be the subring of the ﬁeld of complex numbers given by Z[i] = {m + ni ∈ C | m. These. Show that the rings R and S in the two previous problems are isomorphic as abelian groups. n ∈ Z} . and [x2 + x] = [x][x + 1].2 gives the multiplication table. [x + 1].3. [x2 + x]} [x] = {[0]. are the only idempotents. Solution: Since [x3 ] = [1].CHAPTER 5 SOLUTIONS 97 Solution: Using the general fact that (a + b)2 = a2 + 2ab + b2 = a2 + b2 (since Z2 [x] has characteristic 2) and the identities [x3 ] = [1] and [x4 ] = [x]. and [x2 + x]. and then [x2 + 1]2 = [x2 + 1]. but not as rings.

2 and so m + ni belongs to the principal ideal generated by 1 + i. It is clear that 1 + i is in the kernel. for some congruence classes a and b in Zp . Prove that φ is a ring homomorphism. θ((a + bi) + (c + di)) = θ((a + c) + (b + d)i) = [(a + c) + (b + d)x] θ((a + bi)) + θ((c + di)) = [a + bx] + [c + dx] = [(a + c) + (b + d)x] θ((a + bi)(c + di)) = θ((ac − bd) + (ad + bc)i) = [(ac − bd) + (ad + bc)x] θ((a + bi))φ((c + di)) = [a + bx][c + dx] = [ac + (ad + bc)x + bdx2 ] . (a) Show that φ is a ring homomorphism. for all r ∈ R. We need to use the fact that [x2 ] = [−1] in Zp [x]/ x2 + 1 . which show that φ is a ring homomorphism. Solution: We have the following computations. r + J). and deﬁne the function φ : R → R/I ⊕ R/J by φ(r) = (r + I. Then m and n are either both even or both odd. Since the elements of Zp [x]/ x2 + 1 all have the form [a + bx]. it is clear the θ is an onto function. which show that θ is a ring homomorphism.98 CHAPTER 5 SOLUTIONS (a) Deﬁne φ : Z[i] → Z2 by φ(m + ni) = [m + n]2 . and so it follows that m − n is always even. We claim that ker(φ) is generated by 1 + i. Therefore m + ni = (m − n) + n + ni = (m − n) + n(1 + i) m−n = (1 − i)(1 + i) + n(1 + i) 2 1 = (m − n)(1 − i) + n (1 + i) . (b) For any prime number p. φ((a + bi) + (c + di)) = φ((a + c) + (b + d)i) = [a + c + b + d]2 φ((a + bi)) + φ((c + di)) = [a + b]2 + [c + d]2 = [a + b + c + d]2 φ((a + bi)(c + di)) = φ((ac − bd) + (ad + bc)i) = [ac − bd + ad + bc]2 φ((a + bi))φ((c + di)) = [a + b]2 · [c + d]2 = [ac + ad + bc + bd]2 . deﬁne θ : Z[i] → Zp [x]/ x2 + 1 by θ(m + ni) = [m + nx]. with ker(φ) = I ∩ J. Let I and J be ideals in the commutative ring R. Find ker(φ) and show that it is a principal ideal of Z[i]. and we note that (1 − i)(1 + i) = 2. Let m + ni ∈ ker(φ) = {m + ni | m + n ≡ 0 (mod 2)}. . 16. Prove that θ is an onto ring homomorphism. Solution: We have the following computations.

We have a − r = a − bx − ay = ax − bx ∈ I. then we have f (x) = m h(x) for h(x). The isomorphism follows from the fundamental homomorphism theorem. and thus R/(I ∩ J) ∼ R/I ⊕ R/J . p Solution: We must show that x2 + 1 is irreducible over Zp . By Lagrange’s theorem. for g(x) polynomials f (x) and g(x) with rational coeﬃcients. Prove that the ring Zp [x]/ x2 + 1 is a ﬁeld. Suppose that a is a root of x2 + 1 in Zp . and b−r = b−bx−ay = by −ay ∈ J. or. 0 + J). b + J) ∈ R/I ⊕ R/J . for some x ∈ I and y ∈ J. this means that 4 is a divisor of the order of Z× . Then a2 ≡ −1 (mod p). Therefore p = 4q + 1 for p some q ∈ Z. so ker(φ) = I ∩ J. and so a4 ≡ 1 (mod p). The element a cannot be a root of x2 − 1. and thus it must have order 4. we have r ∈ ker(φ) if and only if r ∈ I and r ∈ J. Let p be an odd prime number that is not congruent to 1 modulo 4. that x2 + 1 has no root in Zp . Considering Z[x] to be a subring of Q[x]. and φ is onto. Thus φ(r) = (a+I. show that these two integral domains have the same quotient ﬁeld. Given any element (a + I. then φ is onto. (b) Show that if I + J = R. Hint: Show that a root of x2 = −1 leads to an element of order 4 in the multiplicative group Z× . which is p − 1. k(x) ∈ Z[x]. then we can write 1 = x+y. Solution: An element of the quotient ﬁeld of Q[x] has the form f (x) . g(x) k(x) and this shows that f (x) g(x) belongs to the quotient ﬁeld of Z[x]. . 17.CHAPTER 5 SOLUTIONS 99 Solution: The fact that φ is a ring homomorphism follows immediately from the deﬁnitions of the operations in a direct sum and in a factor ring. b+J). 18. so it does not have order 2. noting that a = ax + ay and b = bx + by. contradicting the assumption. = Solution: If I +J = R. If m is the lcm of the denominators of the coeﬃcients of f (x) and n is the lcm of the denominators n of the coeﬃcients of g(x). consider r = bx + ay. Since the zero element of R/I ⊕ R/J is (0 + I. equivalently.

100 CHAPTER 5 SOLUTIONS .

express (7 − 2u + u2 )−1 in the form a + bu + cu2 . Eisenstein’s criterion (with p = 3) shows that x3 − 3x2 + 3x − 3 is irreducible over Q. Thus u3 + 3u + 3 = (u + 2)(u2 − 2u + 7) − 11. In Q(u). which yields x3 − 3x2 + 3x − 1 = 2. √ 3. u2 − 2iu + i2 = 2. √ √ √ √ Solution: We have Q ⊆ Q( 2) ⊆ Q( 2. √ √ 2. √ √ √ Solution: Let u = 2 + i. and u2 − 3 = 2iu. i) : √ √ Q( 2)] =√ since i is a √ of a polynomial of degree 2 over Q( 2) but is 2 root not in Q( 2). √ (b) Find the minimal polynomial of 2 + i over Q. we have i2 √ √ −1 2 − i = 3( 2√ i) + ∈ Q(u). Since ( 2 + i)( 2 − i) = 2 − √ = 3. (a) Show that Q( 2 + i) = Q( 2. and so the minimal polynomial for √ 2 + i must have degree 4. and so (x − 1)3 = 2. Thus [Q( 2) : Q] = 2 since 2 √ is a root of a polynomial of degree 2 but is not in Q. so Q( 2. √ √ Solution: Let x = 1 + 3 2. and it follows easily that 2 ∈ Q(u) and i ∈ Q(u). Thus the minimal and polynomial for 2 + i is x4 − 2x2 + 9. i). Find the minimal polynomial of 1 + 3 2 over Q. Squaring again √ combining terms gives u4 −2u2 +9 = 0. i). Let u be a root of the polynomial x3 + 3x + 3. i) ⊆ Q(u). 101 . Solution: Dividing x3 + 3x + 3 by x2 − 2x + 7 gives the quotient x + 2 and remainder −11. and therefore x3 − 3x2 + 3x − 3 = 0. We have [Q( 2. √ √ Since u = 2 + i. The reverse inclusion is obvious. so this is the required minimal polynomial. and so (7 − 2u + u2 )−1 = (2 + u)/11 = (2/11) + (1/11)u. Thus [Q( 2 + i) : Q] = 4. we have u − i = 2. Then x − 1 = 3 2.Chapter 6 Fields SOLUTIONS TO THE REVIEW PROBLEMS 1.

Does 4 2 belong to Q( 3 2 + i)? √ √ Solution: Let α = 3 2 + i. 3 5. √ √ Find [Q( 7 16 + 3 7 8) : Q]. and this shows that i ∈ Q( 3 √+ i). 2) over Q. Since x5 − 2 is also irre√ ducible by Eisenstein’s criterion. Show that x3 + 6x2 − 12x + 2 is irreducible over Q. then it would have a linear factor. 2. We conclude a √ that [Q( 2 + 3 5) : Q] = 6. and since this √ extension has degree 3. Solution: Eisenstein’s criterion works with p = 2. or α3 − 3iα2 − 3α + i = 2. 3 5) over Q. and remains irreducible √ over Q( 5 2). 3 5. 3 25. √ √ √ Find the degree of 3 2 + i over Q. It follows that u has degree 2. and so it would √ have a root in Q( 5 2). It follows 2 √ √ √ 3 3 3 immediately that 2 ∈ Q( 2 + i). 2. and this extension contains u = 2 + 3 5. 3 25. 3. 3 5) in over Q( 3 5). the √ minimal polynomial x2 − 5 of 5 remains irre√ √ √ ducible √ the extension Q( 3 5). so 2 + 3 5 cannot satisfy √nonzero polynomial of degree 1. 3 25} is a basis for Q( 3 5) over Q. 6. If x3 + 6x2 − 12x + 2 could √ be factored over Q( 5 2).102 CHAPTER 6 SOLUTIONS 4. This root would have degree 3 over Q. have [Q( 7 2) : Q] √ 7. and √ the proof √ Theorem 6. and then c = d = 0 as well. and so Q( 2 + i) = Q( 3 2. so we have α3 − 3iα2 + 3i2 α − i3 = 2. 3 5. 5 3 5. 3 5. Then (α − i)3 = 2. the minimal polynomial √2 − √ of√ 2 remains irre2 √ √ x √ √ √ ducible over the extension Q( 3 5). and that is impossible since 3 is not a divisor of 5. 2 3√ is 25} √ √ √ a basis for Q( 3 5. 2 3 5. so that α − i = 3 2. √ √ Find a basis for Q( 5. i). 5 3 25}. then √ √ √ √ √ √ a( 2 + 3 5)3 + b( 2 + 3 5)2 + c( 2 + 3 5) + d = √ √ √ √ √ √ √ √ √ a(2 2 + 6 3 5 + 3 2 3 25 + 5) + b(2 + 2 2 3 5 + 3 25) + c( 2 + 3 5) + d = √ √ √ √ √ √ √ (5a + 2b + d) · 1 + (6a + c) 3 5 + b 3 25 + (2a + c) 2 + 2b 2 3 5 + 3a 2 3 25 = 0.4 shows that the required basis √ √ √ so √ √ of is {1. √ √ √ √ Solution: Let u = 7 16 + 3 7 8. 5. 5. 2 3 5. √ √ Show that [Q( 2 + 3 5) : Q] = 6. Therefore {1. [Q( 5 2) : Q] = 5. 7. 3 25. Since u = ( 7 2 + 3)( 7 2)3 . Thus {1.√it √ follows immediately that a = b = 0. . it follows that √ u ∈ Q( 7√ Since x7 − 2 is irreducible over Q by Eisenstein’s criterion. √ √ We will show that u cannot have degree ≤ 3. If 2 + 3 5 is a root of a polynomial ax3 + bx2 + cx + d in Q[x]. √ √ √ Solution: The set {1. 2. 3 25} is a basis for Q( 3 5) over Q. and since this √ extension has degree 3. Solving for i we get √ i = (α3 − 3α − 2)/(3α2 − 1). 5 is a basis for Q( 5. or 3 over Q. or 6 over Q. 2 3 25} are linearly independent over Q. √ √ √ √ √ √ √ Since {1. and then u must have degree 7 over Q since [Q(u) : Q] = is a divisor of [Q( 7 2) : Q]. 8. we 2). √ √ √ Solution: The set {1.2. 3 5.

[Q( 3√ + i) : Q] = [Q( 3 2 + i) : : 2 √ √ Q( 3 2)][Q( 3 2) : Q] and [Q( 3 2 + i) : Q] = [Q( 3 2 + i) : Q(i)][Q(i) : Q] so √ √ [Q( 3 2 + i) : Q] must be divisible by 2 and 3. we see that i has√ degree 2 over Q. the number 3 2 has degree 3 over Q. so it cannot belong to an extension of degree 6 since 4 is not a divisor of 6. On the other hand. 4 2 has degree 4 over Q since x4 − 2 is irreducible over Q. . Therefore √ √ [Q( 3 2 + i)√ Q] ≤ 6.CHAPTER 6 SOLUTIONS 103 √ Since x3 − 2 is irreducible over Q. Therefore [Q( 3 2 + i) : Q] = 6. √ Finally. Since x2 + 1 is irreducible over Q.

M..J. Abstract Algebra: A First Undergraduate Course. I. G. and Groups: An Introduction to Abstract Algebra London: Edward Arnold. and M..). Maxﬁeld. Artin. 1975. Mac Lane. A.). Englewood Cliﬀs. 1999. A Survey of Modern Algebra (4th ed. Gallian. Abstract Algebra. 1996... Prospect Heights: Waveland Press. 1991 Birkhoﬀ.). Van der Waerden. 1998 Herstein. 1983. J. ———.). L.. J.. Saracino. T. Abstract Algebra and Solution by Radicals.. A First Course in Abstract Algebra (6th ed. B.: AddisonWesley Publishing Co. P. Fields. Fraleigh.. and G. Maxﬁeld.. Inc. 1992. (3rd ed. L. Mass.. Topics in Algebra (2nd ed. Inc.). Rings. Reading. 1992. D. and S. Abstract Algebra: A First Course. B.. Inc. R. New York: John Wiley & Sons. Boston: Houghton Miﬄin Co. N.. Inc. 1977. Algebra. New York: Springer-Verlag.. New York: Macmillan Publishing Co. N. 1985.: Prentice-Hall. A History of Algebra: from al-Khwarizmi to Emmy Noether... 1999. Hillman. New York: Dover Publications. W. Alexanderson.. J. Prospect Heights: Waveland Press. Contemporary Abstract Algebra (4th ed. New York: John Wiley & Sons.104 BIBLIOGRAPHY BIBLIOGRAPHY Allenby.. J. Inc. E. ..

86 cyclic group. 22. 28. 22. 99 ﬁeld. 26. abelian. 77 Chinese remainder theorem. 3. of polynomials. 27. idempotent. 17. 31. 30. 18. 8. 1. 56. for rings. 7 function. 65 combination. 1 division. 27. 72. 17. 65 digit. 13 group. 88. 23. 97. 89–91 ﬁnite group. Euclidean. 7 function. linear. 2 complex numbers. 14–18. 35. 46 Euler phi-function. 53 factor group. 24. 28. 5. 33. units. 44. 94. 27. 7. 89–91 ﬁeld. 14. 65. 22. 102 element. 52. of Eisenstein. 14 dot product. alternating. 14 Cayley’s theorem. 81 cyclic subgroup. 102 binary operation. 30. one-to-one. 20. 35 alternating group. 43 dihedral group. 46 gcd. 23 . 26. 64. 6. 14. 51 Eisenstein’s criterion. 53. 35. 62. of rational numbers. 101. 36. 95. 83 coset. 96. 87. 65. of integers 2. 68. 83. 57. 99 Gaussian integers. 87–89 ﬁnite ﬁeld. 77. 74 even permutation. inverse. division. 5. 40. 82. 94 associative law. for polynomials. 8. 81. 51. matrix form. 7 congruence. 30. 1 algorithm. 24–26. 27. 10. 5 congruence. 57 eigenvalue. 94 equivalence relation. 78. 54 division algorithm. 65. 21. 22–24. 3–5. 15. 80 fundamental homomorphism theorem. 98 gcd. 50 function. 93. 41 conjugacy class. 13 algorithm. 13 cancellation law. 22. 7 fundamental homomorphism theorem. 95. 69. 51–53 Euclidean algorithm. 94. 36. 16. 1. 28. for groups. 44–46. composite. nilpotent. of quotients. cyclic. 17 disjoint cycles. 73. 45. 90 Euclidean algorithm. 30. 82. 46 Euclidean algorithm. for an extension ﬁeld. 31. 86 ﬁeld. 84. 25. onto. 85 generator. 13. 26. 97 element. 82. 63 determinant. 84. 8. 9. 15. 77 annihilator. 6. 64–68. 24. 101. 13 fractional linear transformation. 86 direct product. 45. 29. 30. 41. 9. 24–26. 6. 86 criterion. 62 basis. 5. 60 function. 87. 21 centralizer. 76. 39. 3. 30. 82. 88 general linear group. 8. 15. 20. linear.INDEX 105 Index abelian group. 44. 59. 76 closure. 21. 102 cross product. 67. ﬁnite. 69. 77 group. 13 group. 81. 74 group. 6. 29. 54 cyclic. 99 ﬁeld. 57 cycle. 93. 97 composite function.

94. 89. 6. modulo n. 83. 94 image. 5 linear transformation. 41. 44. multiplicative. 50. modulo n. 29. 6. of a ring homomorphism. parallel. 54 perpendicular plane. of subgroups. principal. 6. of a ring homomorphism. 11. 21. 95–98 74. 51 linearly independent vectors. 22. prime. 51 maximal ideal. 51 order. 66. 44. 29. 15. 8. 47. of a ring. 65 image. invertible. 45. 20. 95. 81. of groups. 28. 13. principal ideal. 7. 62. of permutations. 8. 5. 30. dihedral. 95–98 idempotent element. 101 ideal. 31. 41. 47 nilpotent element. 23 homomorphism. 15. 5. 50 one-to-one function. 80 induction. 30. 18. 82. of rings. 29. 95 ideal. 58. 101 multiplicative inverse. 93. 80 polynomial. ﬁnite. 77. 94. 31. 101–103 kernel. 22. 47 order. 5 nilpotent element. 77. 38. 26. 57. 93. modulo n. 11. 5. 7. 2 lattice diagram. 90 multiplicative order. 30. of a permutation. 60. minimal. 97–99. 14 inverse element. 45. 86 nullity. 11. 2. 29. 26. 95 Lagrange’s theorem. 94. 59 integers mod n. 98 horizontal line test. 60.106 group. 30. 53 irreducible polynomial. 95 ideal. 21. 78–80 group isomorphism. 82. 84 . 99 INDEX lattice diagram. 30. 51 onto function. 8. 29. 19. 6. 31. 28. 33. 30. 53 matrix. 6. 64 inverse. 16–18. 28. 33. 94. 86 group. 50. 53 kernel. 90 invertible matrix. 30. 5 multiplicative order. 76. 93 normal subgroup. 2. 23. 96. of groups. 51 matrix. fractional. 22. 97 idempotent element. 60. 97. 49 ideal. 60 linear transformation. 5. 101 prime ideal. of a group homomorphism. 8. 33. relatively. 78–80 homomorphism. 30. 23. 94. 22. 13 group. 24–26. 56. symmetric. 25. 9 permutation. 3. 95. 94 nilpotent element. 16. even. 97. 53 plane. 45. 10 permutation. 6. 99 isomorphism. 95 minimal polynomial. 50. maximal. 31. 53 partition. multiplicative. 23. 95. 8. 10. 9. 15. 9. 29. 5 identity element. 62. 99 prime. 8. 33. 23. 54 group. 53 permutation group. 21 group homomorphism. 10. 7. 54 parallel plane. 101. 39 linear combination. 69–73 isomorphism. 53 plane. 35 linear congruence. 74 order. 88. 10. 95 polynomial. perpendicular. 5. 44–46. 102 isomorphic rings. of rings. 44. irreducible. 14.

80 units digit. 75 quotient ﬁeld. 26. 5. 2 ring homomorphism. 21. 21. 21. mod n. 8. 22.INDEX quaternion group. 31. 27. of a matrix. 63. 97 units. 52 relatively prime polynomials. 88 subgroup. 93. 23. 75 units. mod 17. 82. 74 units. 5. 26. 4. 66 subring. 94. 84 units. 24 107 . 66 units. mod 15. 87 relatively prime. 102 root. 98 root. 96. 16. 74 units. 15. 21. 77 symmetric law. 58. 30. 6. 45 units. 63. 99 rank. 95. 31. 94. 79 units. of a ring. of a polynomial. 62 units. 17 vertical line test. normal. 14. 21. 101. 17. mod 21. 16. 43 vector space. 51 rational roots. 16. 14. mod 9. mod 18. 6. 29. mod p. 85 units. mod 24. of Lagrange. 15 symmetric group. 30. 52 system of congruences. 99 transformation. 15. 25. 31. 86 subgroup. 79 units. 21. mod 36. 50 transitive law. 99 subspace. 15. 42. 49 well-deﬁned function. 65. mod 7. mod 20. 16. 75 units. 23. 8. 58. 14. 64. 19. 97. 52 unit. linear. 29. 50 rank nullity theorem. 50. rational. 47 theorem. 69. 88 reﬂexive law. mod 13. 23. 63. 17. 16. 33. 26.

- Abstract Algebra (Beachy, Blair) 7.1 Problems and Solutions
- Problems and Solutions to Abstract Algebra (Beachy, Blair)
- 3. ABSTRACT ALGEBRA.pdf
- Abstract Algebra I. N. Herstein p265
- Abstract Algebra Solutions
- Modern Algebra Durbin
- Guide
- Abstract Algebra
- Abstract Algebra
- "Fuckin' Concrete Contemporary Abstract Algebra Introduction..." by Nicolas Bourbaki Junior
- Abstract Algebra
- [I. N. Herstein] Topics in Algebra, 2nd Edition (1(Bookos.org)
- Aspects of Infinite Groups
- [O.Saracino,Abstract algebra,A first course.pdf
- Beachyguide2008
- aljabar abstrak
- Abstract Algebra
- Abstract Algebra (Herstein 3rd Ed)
- A first course in abstract algebra
- Abstract Algebra
- Introduction to Analysis Problems and Solutions
- Herstein Abstract Algebra Student's Solution Manual
- Solution's Manual Abstract Algebra Rotman
- Hunger Ford Solution Manual
- Elementary Analysis Solutions
- algebra
- 2. Joseph A. Gallian, Contemporary Abstract Algebra (4th Edition), Narosa Publishing House,
- J.rotman. .a.first.course.in.Abstract.algebra
- Spivak Calculus
- (eBook - PDF - Mathematics) - Abstract Algebra

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd