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 was given leave in Spring 2000 to work on projects related to teaching. this study guide isn’t a substitute for a good teacher. It’s hard to improve on his book. the subject hasn’t become any easier. by writing out solutions to problems. using Herstein’s Topics in Algebra. and in quite a few cases I’ve included comments to help the reader see what is really going on. I’ve tried to choose problems that would be instructive. but Topics contains the core of any course. As part of the recognition as a “Presidential Teaching Professor. so students meeting abstract algebra still struggle to learn the new concepts. or for the chance to work together with other students on some hard problems. Instead of just expanding the material that is already written down in our textbook. Unfortunately. I decided to try to teach by example. Finally. I would like to gratefully acknowledge the support of Northern Illinois University while writing this study guide. Illinois October 2000 John A. especially since they are probably still learning how to write their own proofs. DeKalb. Of course. with applications to computing and other areas. Beachy .PREFACE v PREFACE I ﬁrst taught an abstract algebra course in 1968. This “study guide” is intended to help students who are beginning to learn about abstract algebra. the subject may have become broader.

vi PREFACE .

simplify.3) to write a = bq + r. where 0 ≤ r < b.1 Divisors Before working through the solved problems for this section. —When working on questions involving divisibility you may ﬁnd it useful to go back to Deﬁnition 1. For example. actually tell you everything you will need to know about cyclic groups. Here are several useful approaches you should be able to use. The examples in Section 1. in Chapter 3 of the text you will be introduced to the concept of a group. —Another approach to proving that b|a is to use the division algorithm (see Theorem 1. one that is obtained by considering all powers of a particular element.Chapter 1 INTEGERS Chapter 1 of the text introduces the basic ideas from number theory that are a prerequisite to studying abstract algebra. Then to prove that b|a you only 1 . then you have an equation to work with. This equation involves ordinary integers. If you expand the expression b|a by writing “a = bq for some q ∈ Z”. 1. the proofs of the theorems contain important techniques that you need to copy in solving the exercises in the text. try to write down an expression for a and expand. In fact.1. it is a signiﬁcant step forward into the realm of abstract algebra.1.4. One of the ﬁrst broad classes of groups that you will meet depends on the deﬁnition of a cyclic group. although Chapter 1 is very concrete. and so you can use all of the things you already know (from high school algebra) about working with equations. or substitute for terms in the expression until you can show how to factor out b. you need to make sure that you are familiar with all of the deﬁnitions and theorems in the section.1. In many cases. constructed using congruence classes of integers. —To show that b|a. Many of the concepts introduced there can be abstracted to much more general situations.

n + 10).. if you have a linear combination ma + nb = d. . This is one of the most useful tools in working with relatively prime integers. you now have some more familiar techniques to use. b) = 1. If n is a positive integer. n2 + n + 1) = 3. . This is really useful in working on questions involving greatest common divisors. INTEGERS —Theorem 1. Remember that this only works in showing that gcd(a. 527). then gcd(n − 1. then b = ±a. Find gcd(3553. n2 + n + 1) = 1 or n 1 0 −1 1 0 = 0 0 0 0 0 1 0 0 0 1 1. 28. where m. 1002. then m2 − n2 is divisible by 8. 0 29. . . 102. . Which of the integers 0. More generally. .2 need to ﬁnd some way to check that r = 0. which can be expressed as the smallest positive linear combination of a and b. 27. ﬁnd the possible values of gcd(n. .2. 30.1 22. SOLVED PROBLEMS: §1. An integer is a linear combination of a and b if and only if it is a multiple of their greatest common divisor. 24. Prove that if m and n are odd integers. 377). Prove that if n is a positive integer.6 states that any two nonzero integers a and b have a greatest common divisor. Prove that if n is an integer with n > 1. CHAPTER 1. 10 can be expressed in the form 12m + 20n.1. then 0 1 if and only if 4|n. Give a proof by induction to show that each number in the sequence 12. and express it as a linear combination of 3553 and 527. it only shows that gcd(a. n are integers? 25. 26. b) is a divisor of d (refer back to Theorem 1. 23. is divisible by 6. gcd(n − 1. 10002.2 states that integers a and b are relatively prime if and only if there exist integers m and n with ma + nb = 1. 1. Since the fundamental theorem of arithmetic (on prime factorization) is proved in this section. Prove that if a and b are nonzero integers for which a|b and b|a. and express it as a linear combination of 435 and 377.6). Find gcd(435.2 Primes Proposition 1. .1.

any equation involving integers can be converted into a congruence by just reducing modulo n. Prove that gcd(2m − 1. b. Prove that gcd(2n2 + 4n − 3. The do the opposite conversion you must be more careful. that doesn’t make them equal. Find all integer solutions of the equation xy + 2y − 3x = 25. 1492). For positive integers a. . 26. 28. What things are the same? You can add or subtract the same integer on both sides of a congruence. 1. First. the integers may diﬀer by some multiple of n. n) = 1 and ac ≡ ad (mod n). you can only divide through by an integer that is relatively prime to n.3. Is the same true for 2n − 1 and 3n − 1? 29. and you can use the fact that if a ≡ b (mod n) and b ≡ c (mod n). (Review Proposition 1.3.2 23. Let m and n be positive integers. then are certainly congruent modulo n.3 Congruences In this section. 1089). but only guarantees that dividing by n produces the same remainder in each case. b2 ) = 1. b) = 1 if and only if gcd(a2 . 2n − 1) = 1 if and only if gcd(m.) What things are diﬀerent? In an ordinary equation you can divide through by a nonzero number. then c ≡ d (mod n). (a) Use the Euclidean algorithm to ﬁnd gcd(1274. You can use substitution. 30. 2n2 + 6n − 4) = 1. If two integers are congruent modulo n. 25. for all integers n > 1. In a congruence modulo n. (b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776. it is not exactly the same. n) = 1. In other words.1. for all integers n > 1. prove that gcd(a. 1089).3. 3 27. and you can multiply both sides of a congruence by the same integer. Give the lattice diagram of all divisors of 250. Do the same for 484. Prove that n − 1 and 2n − 1 are relatively prime. it is important to remember that although working with congruences is almost like working with equations. This is usually expressed by saying that if gcd(a. (a) Use the Euclidean algorithm to ﬁnd gcd(1776. then a ≡ c (mod n). and the comments in the text both before and after the proof of the proposition. This works because if two integers are equal. Just be very careful! One of the important techniques to understand is how to switch between congruences and ordinary equations. 1492). 24. (b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274. CONGRUENCES SOLVED PROBLEMS: §1.

Many of the theorems in the text should be thought of as “shortcuts”. b. These proofs actually show you the necessary techniques to solve all linear congruences of the form ax ≡ b (mod n). In the example. for any integer q. 55x ≡ 35 (mod 75). Solve the system of congruences x ≡ 2 (mod 9) x ≡ 4 (mod 10) . Notice that converting to an equation makes it more complicated. Give integers a. (b) Show that if x = m and y = n is an integer solution to the equation in part (a). because we have to introduce another variable.4 CHAPTER 1. and you can’t aﬀord to skip over their proofs. 27. INTEGERS The conversion process is illustrated in Example 1. x ≡ 23 (mod 32) . Solve the congruence 42x ≡ 12 (mod 90). You should read the proofs of Theorem 1. m. because you might miss important algorithms or computational techniques. 3x ≡ 2 (mod 13) . (a) Find one particular integer solution to the equation 110x + 75y = 45. we really want a congruence modulo 5. Solve the system of congruences 30. so the next step is to rewrite the equation as x ≡ 7 + 8q (mod 5) . 5x ≡ 14 (mod 17) x ≡ 5 (mod 25) 32.6 very carefully. for some q ∈ Z . where the congruence x ≡ 7 (mod 8) is converted into the equation x = 7 + 8q . SOLVED PROBLEMS: §1. we can reduce each term modulo 5.3. where the moduli n and m are relatively prime. 55x ≡ 36 (mod 75). n to provide an example of a system x ≡ a (mod m) that has no solution. so that we ﬁnally get x ≡ 2 + 3q (mod 5) . then so is x = m + 15q and y = n − 22q. Solve the system of congruences 31. (a) Find all solutions to the congruence (b) Find all solutions to the congruence 28.3.5 and Theorem 1. and all simultaneous linear equations of the form x ≡ a (mod n) and x ≡ b (mod m).5 of the text. 29. x ≡ b (mod n) .3. Actually.3 26.

5 36. Instead of thinking in terms of division. INTEGERS MODULO N 33. provided we think in terms of equivalence classes. n) = 1. 35. provided [a]−1 exists. then the smallest positive integer k such that ak ≡ 1 (mod n) is called the multiplicative order of [a] in Z× . The exercises for Section 1. it is probably better to think of multiplying both sides of the equation [a]n [x]n = [b]n by [a]−1 .4 of the text contain several deﬁnitions for elements of Zn . the element [a] ∈ Zn is said to be n idempotent if [a]2 = [a]. . and choose whichever approach is the most convenient.1. Find all integers n for which 13 | 4(n2 + 1). 1. or 6 as its units digit. n n It is well worth your time to learn about the sets Zn and Z× . This gives you one more way to view problems involving congruences. Find the multiplicative inverse of each nonzero element of Z13 . (a) Compute the last digit in the decimal expansion of 4100 . 5. so that you can easily shift back and forth between them. SOLVED PROBLEMS: §1.4 30. any linear congruence of the form ax ≡ b (mod n) can be viewed as an equation in Zn . To be more precise. written [a]n [x]n = [b]n . 31. trying to divide by a in the congruence ax ≡ b (mod n) can get you into trouble unless gcd(a. (b) Is 4100 divisible by 3? 34. 1. Prove that 10n+1 + 4 · 10n + 4 is divisible by 9. If (a. and it is worthwhile to learn all of the approaches. Since |Z× | = ϕ(n). For example. for all positive integers n. Prove that the fourth power of an integer can only have 0. and nilpotent if [a]k = [0] for some k. They will provide n an important source of examples in Chapter 3. when we begin studying groups. Finally.4 Integers Modulo n The ideas in this section allow us to work with equations instead of congruences. 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].4. Find the multiplicative inverse of each nonzero element of Z7 . The set Z× is said to be cyclic if n n it contains an element of multiplicative order ϕ(n). n) = 1. Sometimes it helps to have various ways to think about a problem.

For ω = − + i. ﬁnd its multiplicative inverse. INTEGERS 34. . Review Problems 1. Find [3379]−1 . 4061 4061 CHAPTER 1. prove that ω n = 1 if and only if 3|n. then k | ϕ(n).6 32. n) = 1. 36. Show that Z× is cyclic. Find gcd(7605. Solve the system of congruences Z× . Find [91]−1 . ﬁnd all idempotent elements. List the elements of For each element. 2x ≡ 9 (mod 15) x ≡ 8 (mod 11) . Let n be a positive integer. 4. and express it as a linear combination of 7605 and 5733. if possible (in Z× ). Show that if n > 1 is an odd integer. for any integer n. Solve the congruence 24x ≡ 168 (mod 200). Solve the equation [x]2 + [x]11 − [6]11 = [0]11 . Prove that if k is the smallest positive integer for which ak ≡ 1 (mod n). 35 35 35 for some positive integers i. ﬁnd all nilpotent elements. 38. In Z24 : ﬁnd all units (list the multiplicative inverse of each). 501 501 33. 40. 5733). 2 2 3. ﬁnd all idempotent elements. 35. √ 1 3 2. Show that Z× is not cyclic but that each element has the form [8]i [−4]j . 17 37. then ϕ(2n) = ϕ(n). if possible (in Z× ). Prove that [a]n is a nilpotent element of Zn if and only if each prime divisor of n is a divisor of a. ﬁnd all nilpotent elements. In Z20 : ﬁnd all units (list the multiplicative inverse of each). 15 5. 11 39. 6. j. and ﬁnd its multiplicative order. and let a ∈ Z with gcd(a.

which give a very important class of examples. When you begin to study groups in Chapter 3. 7 . The second goal of the chapter is to begin studying groups of permutations. is stated rather formally in terms of ordered pairs.1. 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 f : R → R+ and g : R+ → R deﬁned by f (x) = ex .1 Functions Besides reading Section 2. and g(y) = ln y. In our study of algebraic structures in later chapters. the deﬁnition of function. the functions that are one-to-one correspondences will be particularly important. 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. 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.1. one-to-one and onto functions. That makes it especially important to understand the concept of an inverse function. as well as on your knowledge of the groups Zn and Z× . it might help to get out your calculus textbook and review composite functions. functions will provide a way to compare two diﬀerent structures. provide one of the most important examples of a pair of inverse functions.Chapter 2 FUNCTIONS The ﬁrst goal of this chapter is to provide a review of functions. you will be able draw on your knowledge of permutation groups.1. In this setting. for all x ∈ R.) In terms of actually using this deﬁnition. (Think of this as a deﬁnition given in terms of the “graph” of the function. n 2. for all y ∈ R+ . and inverse functions. Deﬁnition 2.

Let A be an m × n matrix with entries in R. for all x ∈ Rn . This led to the concept of congruence modulo n.4. Explain why this agrees with Deﬁnition 2.7. (a) Show that L is an invertible function if and only if det(A) = 0. ﬁnd the 17 inverse function θ−1 . more than one 22. First there is the deﬁnition . Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax.8 CHAPTER 2. 25. where AT is the transpose of A. Show that L is a one-to-one function if det(AT A) = 0. FUNCTIONS SOLVED PROBLEMS: §2. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax. Prove that L is one-to-one if and only if no eigenvalue of A is zero. In this section you will ﬁnd three diﬀerent points of view. for all x ∈ Rn . 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. In calculus the graph of an inverse function f −1 is obtained by reﬂecting the graph of f about the line y = x. Let a be a ﬁxed element of Z× . then it is invertible. 23. 2. 24. 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.1. Deﬁne the function θ : Z× → Z× by 17 17 17 θ(x) = ax. Deﬁne a linear transformation L : Rn → Rm by L(x) = Ax. looking at the one idea of splitting up a set S from three distinct vantage points. Explain why this agrees with Deﬁnition 2.1 20. Let A be an n × n matrix with entries in R. and assume that m > 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.1. (b) Show that if L is either one-to-one or onto. Let A be an n × n matrix with entries in R. 26. depending on the remainder when the integer is divided by the ﬁxed integer n. for all x ∈ Rn . You are already familiar with one of the important examples: in Chapter 1 we split the set of integers up into subsets.1. for all x ∈ Z× .1. Explain why this agrees with Deﬁnition 2. which is a model for our general notion of an equivalence relation. Is θ one to one? Is θ onto? If possible.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. 21.

2 14. and give a geometric description of the equivalence classes of ∼. x + y + z) . b)} of all ordered pairs of positive integers. deﬁne (x1 . y2 ) if x1 y2 = x2 y1 . Show that ∼ is an equivalence relation. Then there is the notion of a partition of S. . On the set {(a. 19. which tells you when two diﬀerent elements of S belong to the same subset. Find the image of f and the set Z12 /f of equivalence classes determined by f . For the linear transformation L : R3 → R3 deﬁned by L(x. deﬁne v ∼ w if v ·u = w ·u. For the function f : R → R deﬁned by f (x) = x2 . deﬁne z1 ∼ z2 if ||z1 || = ||z2 ||. 17. which is a big help in deciding which point of view to take. z) ∈ R3 . For vectors v and w. 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. z) = (x + y + z. y. deﬁne A ∼ B if there exists an invertible matrix P such that P AP −1 = B. so that you can easily switch from one to the other. and assume that u has length 1. Your goal should be to learn about each point of view. Check that ∼ deﬁnes an equivalence relation. describe the equivalence relation on R that is determined by f . y. 18. for all (x.2. Show that this deﬁnes an equivalence relation. y1 ) ∼ (x2 . it turns out that every partition (and equivalence relation) really comes from a function f : S → T . for all [x]12 ∈ Z12 . where we say that x1 and x2 are equivalent if f (x1 ) = f (x2 ). Deﬁne the formula f : Z12 → Z12 by f ([x]12 ) = [x]2 . give a geometric description of the partition of R3 that is determined by L. 12 Show that the formula f deﬁnes a function. which places the emphasis on describing the subsets. x + y + z. Show that ∼ is an equivalence relation. On the set of all n × n matrices over R. where · denotes the standard dot product. On the set C of complex numbers. SOLVED PROBLEMS: §2. EQUIVALENCE RELATIONS 9 of an equivalence relation on S. 20. for all x ∈ R.2. Let u be a ﬁxed vector in R3 . Finally. 15. 16.

What is the order of σ? Compute σ −1 . You need to do enough computations so that you will feel comfortable in dealing with permutations. then σ −1 ∈ Sym(S). στ σ −1 . τ −1 . (i) If σ. for any permutation σ ∈ Sn . τ ∈ Sym(S). 1 2 2 5 3 1 4 8 5 3 6 6 7 4 8 7 9 9 and 14. compute the order of στ σ −1 . instead of the way we have deﬁned multiplication.10 CHAPTER 2. 7). 6)(3. Let σ = (2. τ σ. then τ σ ∈ G. 5. . 8. τ στ −1 . In two of the problems. 6) ∈ S9 . τ ∈ G.6.3 Permutations This section introduces and studies the last major example that we need before we begin studying groups in Chapter 3. you need to be aware that some authors multiply permutations by reading from left to right. (i) If σ.3 13. 7. 15. so we have to do the computations from right to left. Compute the order of τ = 1 2 3 4 5 7 2 11 4 6 σ = (3. For the permutations σ = τ= 1 2 3 4 5 6 7 8 9 . 9)(1. In the text we noted that if S is any set. just as in calculus. 6 7 8 9 8 10 9 10 1 3 11 5 . then we have the following properties. (iii) if σ ∈ Sym(S). What is the order of σ? Is σ an even permutation? Compute σ −1 . then σ −1 ∈ G. write σ as a 7 5 6 9 2 4 8 1 3 product of disjoint cycles. SOLVED PROBLEMS: §2. and we write functions on the left. For the permutation σ = 1 2 3 4 5 6 7 8 9 . If you are reading another book along with Abstract Algebra. τ . FUNCTIONS 2. For 17. then τ σ ∈ Sym(S). σ −1 . and Sym(S) is the set of all permutations on S. write each of these permutations as a 1 5 4 7 2 6 8 9 3 product of disjoint cycles: σ. 4. We will see later that this agrees with Deﬁnition 3. (ii) 1S ∈ Sym(S). 2. 16. (ii) 1S ∈ G. στ . If G is a nonempty subset of Sym(S). (iii) if σ ∈ G. 9. )(6. Prove that if τ ∈ Sn is a permutation with order m. Write σ as a product of disjoint cycles. 8. we need the following deﬁnition.1 of the text. then στ σ −1 has order m. 4. Our point of view is that permutations are functions. we will say that G is a group of permutations if the following conditions hold.

3. and let X be a subset of S. Let S be a set. Let G be a group of permutations. for the set S. with G ⊆ Sym(S). 12. 11 19. but not 11 or 13. 20. Prove that G is a group of permutations. and 14. Let τ be a ﬁxed permutation in Sym(S). . Let G = {σ ∈ Sym(S) | σ(X) ⊂ X}. Prove that τ Gτ −1 = {σ ∈ Sym(S) | σ = τ γτ for some γ ∈ G} is a group of permutations. PERMUTATIONS 18.2. Show that S10 has elements of order 10.

(b) In S3 . ﬁnd the inverse function φ−1 . Is σ even or odd? 5. 3. β = (1. 6. . 6). 2). for all x ∈ R. Deﬁne f : R → R by f (x) = x3 + 3xz − 5. In S10 . write σ as a product of disjoint cycles. 3). let α = (1. Hint: Use the derivative of f to show that f is a strictly increasing function. 5. For the function f : R → R deﬁned by f (x) = x2 . describe the equivalence relation on R that is determined by f . 9). On the set Q of rational numbers. Deﬁne the function φ : Z× → Z× by φ(x) = x−1 . 7. Is φ one to 17 17 17 one? Is φ onto? If possible. let α = (1. Show that φα : Sn → Sn deﬁned by φα (σ) = ασα−1 . deﬁne x ∼ y if x − y is an integer. 5. is a one-to-one and onto function. For σ = αβγ. 2. for all x ∈ Z× . 3. and γ = (1. 4.12 CHAPTER 2. and use this to ﬁnd its order and its inverse. FUNCTIONS Review Problems 1. for all σ ∈ Sn . Show that f is a one-to-one function. for all x ∈ R. 2. 2. Show that ∼ is an equivalence relation. Compute φα . (a) Let α be a ﬁxed element of Sn .

“there exists”). which just represent various numbers. Try to learn it so well that you don’t have even a trace of an accent! Loosely. is usually thought of as the real beginning of abstract algebra. abelian group. This still works for the operation in a group. The precise statement is given in Deﬁnition 3. and x = y. you must pay careful attention to each part. subtraction.1. and ﬁnite group. “for each”. the axioms for a group give us just what we need to work with equations involving the operation in the group. or possibly operations on matrices or functions. multiplication. group. For example. At ﬁrst we will use ∗ or · to represent an operation.1 Deﬁnition of a Group This section contains these deﬁnitions: binary operation.3. which we begin in this chapter. and division.Chapter 3 GROUPS The study of groups. 3. very soon we will just write ab instead of a ∗ b. One of the things we try to do with notation is to make it look familiar. especially the quantiﬁers (“for all”. even if it represents something new. to show that ∗ might represent ordinary addition or multiplication. has an identity element. But the operations are still the usual ones for numbers. and the equation will still hold. The step from algebra to abstract algebra involves letting the operation act like a variable. which must be stated in exactly the right order. From one point of view. one of the rules you are used to says that you can multiply both sides of an equation by the same value. These deﬁnitions provide the language you will be working with. The step from arithmetic to algebra involves starting to use variables. a group is a set on which it is possible to deﬁne a binary operation that is associative. for 13 . since if x and y are elements of a group G. and you simply must know this language. and has inverses for each of its elements. addition. then a · x = a · y. so long as everyone knows the convention that we are using. or maybe even something quite far from your experience.

without mentioning their operations. z1 ) and (x2 . z2 ) in R3 .1.1 22. I should still mention that the original motivation for studying groups came from studying sets of permutations. since the operation in the group may not satisfy the commutative law. Similarly. and C× of nonzero numbers form groups under multiplication. If you are careful about the side on which you multiply. This is a part of the guarantee that comes with the deﬁnition of a binary operation. without speciﬁcally mentioning which operation is used. Write out the multiplication table for Z× . The existence of inverses allows cancellation (see Proposition 3. 9 . R× . a is multiplied on the left. It is important to note that on both sides of the equation. z1 )×(x2 . Use the dot product to deﬁne a multiplication on R3 . but we can’t guarantee that a · x = y · a. and C form groups under addition. you can be pretty safe in doing the familiar things to an equation that involves elements of a group.6 for the precise statement). R. z1 x2 − x1 z2 . and so the symmetric group Sn still has an important role to play. but not under addition. The next level comes from knowing some good examples. SOLVED PROBLEMS: §3. the sets Q× . 2 25. Show that G is a group under this multiplication. the most important examples come from groups of matrices. Here are a few of the important examples.14 CHAPTER 3. y1 . y1 . Is R3 a group under this multiplication? 24. For vectors (x1 . so whenever you are tempted to write a ÷ b or a/b. The sets Z and Zn are groups under addition. while the set GLn (R) of all invertible n × n matrices with entries in R is a group under multiplication. Understanding and remembering the deﬁnitions will give you one level of understanding. There shouldn’t be any confusion in just listing these as groups. y2 . Next. while Z× is a group under multiplication. GROUPS any element a in G. It is common to just list these n sets as groups. since in each case only one of the two familiar operations can be used to make the set into a group. but not multiplication. First. for all a. We could also guarantee that x · a = y · a. the set Mn (R) of all n × n matrices with entries in R is a group under addition. Remember that in a group there is no mention of division. and don’t fall victim to the temptation to divide. you must write a · b−1 or b−1 · a. the sets of numbers Z. b ∈ G. The third level of understanding comes from using the deﬁnitions to prove various facts about groups. x1 y2 − y1 x2 ). Does this make R3 into a group? 23. y2 . In the study of ﬁnite groups. On the set G = Q× of nonzero rational numbers. the cross product is deﬁned by (x1 . z2 ) = (y1 z2 − z1 y2 . Q. deﬁne a new multiplication ab by a∗b = .

For x. F is the set of all functions f (z) : C → C of the form f (z) = cz + d where the coeﬃcients a. Write out the multiplication table for Z× . Prove that these weaker conditions (given only on the left) still imply that G is a group. SUBGROUPS 26. it forms an abelian group. and replace condition (iv) with the condition that for each a ∈ G there exists a ∈ G with a · a = e. Let F be the set of all fractional linear transformations of the complex plane. 29.1. Show that F forms a group under composition of functions. d are integers with ad − bc = 1. That is. and suppose that a and b are any elements of G. and any subspace is automatically a subgroup. Any group of n × n matrices (with entries in R) is a subgroup of GLn (R). 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. then it is called a subgroup. 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. (c) Show that 2 is the identity element for the operation ∗. If you only look at the operation of addition in a vector space. (a) Show that the operation ∗ is closed on G. In Deﬁnition 3. 3.2. Let G = {x ∈ R | x > 1} be the set of all real numbers greater than 1. for some set S. 28. Let G be a group.3. then ba = ab. any group of permutations is a subgroup of Sym(S). y ∈ G. . 31. If the subset is a group in its own right. Show that if (ab)2 = a2 b2 . c. and suppose that a and b are any elements of G.3 of the text. (b) Show that the associative law holds for ∗. you are right. 15 15 27. Show that (aba−1 )n = abn a−1 . 30. (d) Show that for element a ∈ G there exists an inverse a−1 ∈ G. az + b . Now might be a good time to pick up your linear algebra text and review vector spaces and subspaces. replace condition (iii) with the condition that there exists e ∈ G such that e · a = a for all a ∈ G. 32. for any positive integer n. If the idea of a subgroup reminds you of studying subspaces in your linear algebra course. Let G be a group. For instance. b. deﬁne x ∗ y = xy − x − y + 2. using the same operation as the larger set.2 Subgroups Many times a group is deﬁned by looking at a subset of a known group.

and K = {[x]21 | x ≡ 1 (mod 7)} . In G = Z× . ﬁnd the order of the √ √ √ 2 2 2 2 elements − + i and − − i. 2. Let G be an abelian group. for some n > 0. 2 2 2 2 30. You do not have to list all of the elements if you can explain why there must be 12.16 CHAPTER 3. since in that case the inverse a−1 of any element can be expressed in the form an . In the multiplicative group C× of √ complex numbers. In Z× . consisting of all powers (positive and negative) of the element. p (b) Use part (a) to prove that n must be a power of 2. one that is cyclic and one that is not 20 cyclic. 24 24. (a) Show that in Z× the order of [2]p is 2n. This is a useful fact to know when you are looking for subgroups in a given group. Show that N = {g ∈ G | g = an for some a ∈ G} is a subgroup of G. 25. In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. then you only need to look at positive powers. GROUPS Lagrange’s theorem is very important. It is also important to remember that every element a in a group deﬁnes a subgroup a . and let n be a ﬁxed positive integer. SOLVED PROBLEMS: §3. 3)(5. 7). Find all cyclic subgroups of Z× . where o(a) is the order of a. 26. (a) Find the cyclic subgroup of S7 generated by the element (1. 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. 28.2 23. show that cos θ − sin θ H= θ∈R sin θ cos θ is a subgroup of G. 29. (b) Find a subgroup of S7 that contains 12 elements. If the group is ﬁnite. Suppose that p is a prime number of the form p = 2n + 1. and why they must form a subgroup. This subgroup has o(a) elements. show that 21 H = {[x]21 | x ≡ 1 (mod 3)} are subgroups of G. 27. ﬁnd two subgroups of order 4.

19. let H = {[x] | x ≡ 1 (mod 4)} and K = {[y] | y ≡ 36 1 (mod 9)}. Show that H and K are subgroups of G. and ntuples are multiplied component-by-component. and list all of the generators for the group. This includes matrices with entries in the ﬁeld Zp . This generalizes the construction of n-dimensional vector spaces (that case is much simpler since every entry comes from the same set). Compute the centralizer in GL2 (R) of the matrix 2 1 1 1 .3 Constructing Examples The most important result in this section is Proposition 3.3 16.14 in the text deﬁnes the centralizer of an element a of the group G to be C(a) = {x ∈ G | xa = ax}. in which we can allow the entries in the matrix to come from any ﬁeld. then the order of the general linear group GLn (Zp ) is (pn − 1)(pn − p) · · · (pn − pn−1 ). The second construction in this section is the direct product. using ordered pairs. c = −2b. ad − bc = 0 17 Show that K is a subgroup of GL2 (R). 3.3. 17. SOLVED PROBLEMS: §3. Show that Z5 × Z3 is a cyclic group. Note: Exercise 3.2. Find the order of the element ([9]12 . In the group G = Z× . CONSTRUCTING EXAMPLES 31. K= a b c d d = a. 20.3. Show that if p is a prime number. and ﬁnd the subgroup HK.3. Find two groups G1 and G2 whose direct product G1 × G2 has a subgroup that is not of the form H1 × H2 . This can be extended to n-tuples.7. [15]18 ) in the group Z12 × Z18 . for subgroups H1 ⊆ G1 and H2 ⊆ G2 . and this allows us to construct very interesting ﬁnite groups as subgroups of GLn (Zp ). 32. Let K be the following subset of GL2 (R). for any prime number p. 18. . which takes two known groups and constructs a new one. which shows that the set of all invertible n × n matrices forms a group. where the entry in the ith component comes from a group Gi .

and show that C(A) ∩ C(B) = Z(G). b ∈ G1 . H= m b 0 1 ∈ GL2 (Z5 ) m. 25. Compute the centralizer in GL2 (Z3 ) of the matrix 24. . where Z(G) is the center of G.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. then BA = A−1 B. 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. m = ±1 (a) Show that H is a subgroup of G with 10 elements. Find the order of the element A = 0 −1 0 0 −i 22. since it is one-to-one and onto. b ∈ G1 makes certain that multiplication can be done in either group and the transferred to the other. Let H be the following subset of the group G = GL2 (Z5 ). GROUPS i 0 0 0 in the group GL3 (C).18 CHAPTER 3. 3. (b) Show that if we let A = 1 1 0 1 and B = −1 0 0 1 . (c) Show that every element of H can be written uniquely in the form Ai B j . Let A = 1 1 −1 0 and B = . 23. Compute the centralizer in GL2 (Z3 ) of the matrix . The condition that φ(ab) = φ(a)φ(b) for all a. Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . In terms of the respective group multiplication tables for G1 and G2 . The function φ can be thought of as simply renaming the elements of G1 . 2 0 2 1 1 2 1 1 . since the inverse function φ−1 also respects the multiplication of the two groups. where 0 ≤ i < 5 and 0 ≤ j < 2. b ∈ Z5 . Find the centralizers C(A) and 0 1 0 1 C(B). 21.

If you can show that one group has a property that the other one does not have. 26. if G1 is cyclic. for all ([n]30 . H1 . for all (x1 . 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. [4n + 3m]6 ). . Let G be a group. In some ways it is harder to show that two groups are not isomorphic. Let G1 . Prove that if a is any element of G. Deﬁne φ : Z30 × Z2 → Z10 × Z6 by φ([n]30 . and suppose that θ1 : G1 → H1 and θ2 : G2 → H2 are group isomorphisms. and let H be a subgroup of G. In this situation we would usually say that we have found all abelian groups of order 8. Deﬁne φ : G1 × G2 → H1 × H2 by φ(x1 . Z4 × Z2 . Z2 × Z2 × Z2 . you should actually produce an isomorphism φ : G1 → G2 . for each positive integer n. θ2 (x2 )). x2 ) ∈ G1 × G2 . 23. Prove that φ is a group isomorphism. because there are inﬁnitely many possibilities. the two groups must have exactly the same number of elements of order n. we can show (in Section 7. x2 ) = (θ1 (x1 ). SOLVED PROBLEMS: §3. Show that φ is a group isomorphism. Furthermore. ISOMORPHISMS 19 From an algebraic perspective. Show that Z× is isomorphic to Z16 . If G1 is abelian. H2 be groups. we should think of isomorphic groups as being essentially the same. Prove that the group Z× × Z× is isomorphic to the group Z6 × Z10 . then the subset aHa−1 = {g ∈ G | g = aha−1 for some h ∈ H} is a subgroup of G that is isomorphic to H. 24. then you can decide that two groups are not isomorphic (provided that the property would have been transferred by any isomorphism).4. G2 . then the question becomes manageable.5) that the answer to this particular question is the list Z8 . The problem of ﬁnding all abelian groups of order 8 is impossible to solve. 17 22.3. Let φ : R× → R× be deﬁned by φ(x) = x3 . In fact. Suppose that G1 and G2 are isomorphic groups. To show that two groups G1 and G2 are isomorphic. Each time you meet a new property of groups. First prove that φ is a well-deﬁned function. and then prove that φ is a group isomorphism.4 21. for all x ∈ R. [m]2 ) ∈ Z30 × Z2 . [m]2 ) = ([n]10 . To decide on the function to use. you probably need to see some similarity between the group operations. up to isomorphism. then so is G2 . 7 11 25. then so is G2 . you should ask whether it is preserved by any isomorphism.

and that φ is actually a group isomorphism. Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 .2 shows that every cyclic group is isomorphic to one of these concrete examples. under addition.20 CHAPTER 3. Show that G is not isomorphic to the direct product R× × R. and let S be any set for which there exists a one-toone and onto function φ : G → S. (b) Show that H is isomorphic to the group R of all real numbers.5.5 Cyclic Groups We began our study of abstract algebra very concretely. 3. b ∈ Z3 . Show that for any prime number p. Let G be a group. the subgroup of diagonal matrices in GL2 (Zp ) is isomorphic to Z× × Z× . Prove that S is a group under this operation. a21 = 0. H= m b 0 1 ∈ GL2 (Z3 ) m. m = 0 Show that H is isomorphic to the symmetric group S3 .3. 31. a22 = 1 is a subgroup of G. Theorem 3. and hk = kh for all h ∈ H and k ∈ K. Let H be the following subgroup of group G = GL2 (Z3 ). then there are subgroups H and K in G such that H ∩ K = {e}. 28. show that H= a11 a21 a12 a22 ∈ GL2 (R) a11 = 1. Let G. In this section. for all x1 . G1 . (a) In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. and this motivated the deﬁnition of an abstract cyclic group. which describes the subgroups of Zn . and the related groups Zn . 30. GROUPS 27. so all of the information about cyclic groups is already contained in these basic examples. G2 be groups. by looking at the group Z of integers. We discovered that each of these groups is generated by a single element.5. showing that they are in one-to-one correspondence with the . Prove that if G is isomorphic to G1 × G2 . x2 ∈ S. HK = G. 32. p p 29. You should pay particular attention to Proposition 3. Deﬁne an operation on S by setting x1 · x2 = φ(φ−1 (x1 )φ−1 (x2 )).

the subgroups of order n just get lost inside the larger symmetric group. 28.4 we will prove that every ﬁnite abelian group is isomorphic to a direct product of cyclic groups of prime power order. respectively. pq 3. since every ﬁnite group of order n is isomorphic to a subgroup of the symmetric group Sn . however. 3). Find all cyclic subgroups of the quaternion group. 24. ﬁnd the order of the subgroup [18]30 . then Z× is not a cyclic group. this section revisits the roots of group theory that we began to study in an earlier chapter. ﬁnd the order of [24]30 . These cyclic groups have a particularly simple structure. 25.3. Give the lattice diagram of subgroups of Z100 . Show that the three groups Z6 .) SOLVED PROBLEMS: §3.6. That isn’t as impressive as it sounds at ﬁrst. Is Z4 × Z15 isomorphic to Z6 × Z10 ? 23. In Z30 . Prove that if p and q are diﬀerent odd primes. 4) and (1. then the direct product G1 × G2 is a cyclic group. 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 . (In Theorem 7. Cayley’s theorem shows that permutation groups contain all of the information about ﬁnite groups.5. PERMUTATION GROUPS 21 positive divisors of n. 2. 9 18 21. In n is a prime power. Z× . and form the basic building blocks for all ﬁnite abelian groups. 27. 26. and Z× are isomorphic to each other. then the subgroups are “linearly ordered” in the sense that given any two subgroups. because as n gets larger and larger. which has order n!. This does imply.6 Permutation Groups As with the previous section. 3. Prove that if G1 and G2 are groups of order 7 and 11. 21 29.5 20. 30. one is a subset of the other. Use this information to show that the quaternion group cannot be isomorphic to the subgroup of S4 generated by (1. Is Z4 × Z10 isomorphic to Z2 × Z20 ? 22. Find all generators of the cyclic group Z28 . Show that any cyclic group of even order has exactly one element of order 2. Use the the result in Problem 27 to show that the multiplicative groups Z× 15 and Z× are not cyclic groups.

8. It is easy to prove that the image φ(G1 ) is . for all 0 ≤ i < n. In this section we look at functions that respect the group operations but may not be one-to-one and onto. and ba = a−1 b. after we develop the concepts of cosets and factor groups. in fact.4 we introduced the concept of an isomorphism. There are many important examples of group homomorphisms that are not isomorphisms. They have a concrete representation. shown in the ﬁrst of the solved problems given below.8. 25. 0 ≤ j < 2} with o(a) = n. GROUPS group. 0 ≤ j < 2} with o(a) = n. show that each element of the form ai b has order 2. and ba = a−1 b} . ﬁnd the centralizer of a. ﬁnd the corresponding subgroup σHσ −1 . in S4 . and ba = a−1 b. SOLVED PROBLEMS: §3.7 Homomorphisms In Section 3. which is a preliminary form of the Fundamental Homomorphism Theorem. and ba = a−1 b. Find the centralizer of (1. We can write Dn = {ai bj | 0 ≤ i < n.6 22. The most important result in this section is Theorem 3. o(b) = 2. and in A4 . 3) and (1. 2). 2. In the dihedral group Dn = {ai bj | 0 ≤ i < n. In S4 . Show that each element in A4 can be written as a product of 3-cycles. 24. homomorphisms provide the way to relate one group to another. 3) in S3 . 28. 0 ≤ j < 2. For the subgroup H of S4 deﬁned in the previous problem. 3. 23. and studied in detail what it means for two groups to be isomorphic. ﬁnd the subgroup H generated by (1. o(b) = 2.) In this formulation of the Fundamental Homomorphism Theorem. 4). 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. (The full statement is given in Theorem 3.7. o(b) = 2. 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 Dn . for σ = (1. 27.8. o(b) = 2. In doing computations in Dn it is useful to have at hand the formula bai = an−i b.22 CHAPTER 3. show that bai = an−i b. in terms of the rigid motions of an n-gon. with o(a) = n. 26. 2. In the dihedral group Dn = {ai bj | 0 ≤ i < n. In the dihedral group Dn = {ai bj | 0 ≤ i < n. we start with a group homomorphism φ : G1 → G2 . 0 ≤ j < 2} with o(a) = n. and.

18. The function φ has an equivalence relation associated with it. so that although the homomorphism may not be an isomorphism between G1 and G2 . 17 (c) Completely determine all group homomorphisms from Z× into Z× . Show that φ : Z× → Z× deﬁned by φ([x]n ) = [x]m .7. y). since for any group homomorphism φ : G1 → G2 it describes the connections between subgroups of G1 and subgroups of G2 . (a) Show that Z× is cyclic. Then Theorem 3. HOMOMORPHISMS 23 a subgroup of G2 . with generator [3]7 . Let φ1 : G → G1 and φ2 : G → G2 be group homomorphisms. 19. and apply the fundamental 36 homomorphism theorem.4 and 3. y) = (x + 2y. and show how elements of the image correspond to cosets of the kernel. for a. For the group homomorphism φ : Z× → Z× deﬁned by φ([x]36 ) = [x]12 . Prove that Z× is isomorphic to the direct pq product Z× × Z× . is a well-deﬁned n m n group homomorphism. for 36 12 all [x]36 ∈ Z× . where we let a ∼ b if φ(a) = φ(b). Deﬁne φ : Z4 × Z6 → Z4 × Z3 by φ(x. 7 (b) Show that Z× is cyclic. where we use the equivalence relation deﬁned by congruence modulo n. is a well-deﬁned group homomorphism. G1 . SOLVED PROBLEMS: §3. Examples 3.3. Let n and m be positive integers.7. ﬁnd the kernel and image of φ. Prove that φ : G → G1 × G2 deﬁned by φ(x) = (φ1 (x). (b) Choose one of the nonzero formulas in part (a). and for this formula ﬁnd the kernel and image. and apply the fundamental homomorphism theorem.8 shows that this group is isomorphic to φ(G1 ). with generator [3]17 . using the operation in G1 . 24. Let p and q be diﬀerent odd primes. (a) Find the formulas for all group homomorphisms from Z18 into Z30 . Proposition 3.5 are important. we can deﬁne a group operation on the equivalence classes of ∼.7. 21. 17 7 20.6 is also useful. 23. because they give a complete description of all group homomorphisms between two cyclic groups. 22. for all x ∈ G. and G2 be groups. Let G. φ2 (x)). such that m is a divisor of n.7 17. p q . (a) Show that φ is a well-deﬁned group homomorphism. b ∈ G1 . (b) Find the kernel and image of φ.7. Find all group homomorphisms from Z4 into Z10 . for all [x]n ∈ Z× . Just as in Z.7. it does deﬁne an isomorphism between a subgroup of G2 and what we call a factor group of G1 .

Let G = Z6 × Z4 . The easiest way to do this is to just deﬁne a homomorphism φ from G to G1 . Of course. Normal Subgroups. list all cosets of K. Let the dihedral group Dn be given via generators and relations. But then you must be careful to prove that the formula you are using does not depend on the particular choice of a representative. List the cosets of 7 in Z× . (0. in addition to proving that it is onto and respects the operations in the two groups.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. it is important to let the theorem do its job.8 27. 0)}. (b) You may assume that any abelian group of order 12 is isomorphic to either Z12 or Z6 × Z2 . If you prove that φ maps G onto G1 . then the Fundamental Theorem does the rest of the work. The moral of this story is that if you deﬁne a function on G rather than G/N . and let K = {(0. you must prove that your formula actually deﬁnes a function. if you deﬁne a function on the cosets of G/N . 0). where at ﬁrst we thought of congruence classes as inﬁnite sets of integers. making sure that N is the kernel of φ. (a) List all cosets of H. and then in Section 1. 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. and Factor Groups The notion of a factor group is one of the most important concepts in abstract algebra. This construction parallels the construction of Zn from Z.24 CHAPTER 3. Then you must prove that your function is one-to-one. so that it does as much of the hard work as possible. That is. (3. if your function is deﬁned on cosets. In actually using the Fundamental Homomorphism Theorem. To construct a factor group. let H = {(0. Is the factor group Z× / 7 cyclic? 16 16 28. 0). That is what we did in Chapter 1. The key idea is to begin thinking of equivalence classes as elements in their own right. in an abelian group we can use any subgroup. GROUPS 3. Which answer is correct for G/H? For G/K? 29. . On the other hand. showing that there exists a well-deﬁned isomorphism between G/N and G1 . satisfying ba = a−1 b. SOLVED PROBLEMS: §3. 2)}. 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. you ordinarily don’t need to worry that it is well-deﬁned. Quite often we need to show that a factor group G/N that we have constructed is isomorphic to another group G1 . with generators a of order n and b of order 2. the most convenient way is use a formula deﬁned on representatives of the cosets of N .8 Cosets. we start with a normal subgroup and the equivalence classes it determines. since all subgroups of an abelian group are normal. Once again.

(b) Show that any element of the form ai b has order 2. a3 } of G. (b) Prove that N is a normal subgroup of HN . and list all cosets of N . (b) Is G/N abelian? 31. and let N = {e. a6 . . For a. 33. (a) Let G be a group. COSETS. a3 .3. then HN/N is isomorphic to H. written b ∼ a. 34. The equivalence classes of ∼ are called the conjugacy classes of G. Find the conjugacy classes of D4 .8. and let N and H be subgroups of G such that N is normal in G. NORMAL SUBGROUPS. Let G be the dihedral group D12 . AND FACTOR GROUPS (a) Show that bai = a−i b for all i with 1 ≤ i < n. a9 }. Show that ∼ is an equivalence relation on G. (c) List all left cosets and all right cosets of b 30. b ∈ G we say that b is conjugate to a. (c) Prove that if H ∩ N = {e}. if there exists g ∈ G such that b = gag −1 . (a) Prove that N is a normal subgroup 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. Which is correct? 32. Let G be a group. Let G = D6 and let N be the subgroup a3 = {e. (a) Show that N is a normal subgroup of G. (a) Prove that HN is a subgroup of G. 25 (b) You may assume that G/N is isomorphic to either Z6 or S3 .

4. GROUPS Review Problems 1. 0 1 0 1 (a) Show that N is a subgroup of G. where a4 = e. Let N be the subgroup a2 = {e. a. Assume that the dihedral group D4 is given as {e. 0 1 Show that G is a subgroup of GL3 (R). 5. (b) Show that Z× is a cyclic group. a3 b}. a6 }. 7. and ba = a3 b. a4 .26 CHAPTER 3. (a) Show by a direct computation that N is a normal subgroup of D4 . (b) Is the factor group D4 /N a cyclic group? 8. (b) Show that G/N has order 4. a2 . Is Z× / 9 cyclic? 20 20 6. and give the lattice diagram which shows the 11 inclusions between them. Show that the group G in the previous problem is isomorphic to the direct product R × R. and that N is normal in G. Let G = D8 . a3 . and let N = {e. 13 2. b ∈ R . b2 = e. a2 }. ab. and verify that N is a normal subgroup of G. a2 . 3. Let G be the subgroup of GL2 (R) consisting of all matrices of the form m b 1 b . . Find all subgroups of Z× . a2 b. (b) Show that G/N is isomorphic to the multiplicative group R× . and let N be the subset of all matrices of the form . b. but is not cyclic. (a) What are the possibilities for the order of an element of Z× ? Explain 13 your answer. (a) List all left cosets and all right cosets of N . 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. List the cosets of the cyclic subgroup 9 in Z× .

Over the ﬁeld of rational numbers. 2. I didn’t break the problems up by section. We can work with congruence classes of polynomials. Over the ﬁeld of rational numbers. use the Euclidean algorithm to show that 2x3 − 2x2 − 3x + 1 and 2x2 − x − 2 are relatively prime. you don’t have to wait until you have ﬁnished the chapter to practice solving some of these problems. The point here is that you can beneﬁt from reviewing Chapter 3. The notion of a prime number is replaced by that of an irreducible polynomial. The point of saying this is that it will be worth your time to review the deﬁnitions and theorems in Chapter 1. Over the ﬁeld of rational numbers.Chapter 4 POLYNOMIALS In this chapter we return to several of the themes in Chapter 1. and express it as a linear combination of the given polynomials. and when two polynomials are relatively prime. Use the Euclidean algorithm to ﬁnd gcd(x8 − 1. ﬁnd the greatest common divisor of x4 + x3 + 2x2 + x + 1 and x3 − 1. just as we did with congruence classes of integers. Review Problems 1. 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. Because you have a lot more experience now than when you started Chapter 1. we want to go beyond high school algebra. x6 − 1) in Q[x] and write it as a linear combination of x8 − 1 and x6 − 1. Of course. In addition to generalizing ideas from the integers to polynomials. 3. We need to talk about the greatest common divisor of two polynomials. 4. This motivates the deﬁnition of a ﬁeld. to be able to work with coeﬃcients that may not be real numbers. 27 . which is quite closely related to the deﬁnition of a group (now there are two operations instead of just one).

15. Factor x4 + x + 1 over Z2 [x]/ x4 + x + 1 . POLYNOMIALS 5. show that [x + 1] is a generator. Are the following polynomials irreducible over Q? (a) 3x5 + 18x2 + 24x + 6 (b) 7x3 + 12x2 + 3x + 45 (c) 2x10 + 25x3 + 10x2 − 30 6. Find the multiplicative inverse of [x2 + x + 1] (a) in Q[x]/ x3 − 2 . In Z5 [x]/ x3 + x + 1 . 7. (a) Show that x2 + 1 is irreducible over Z3 . 12. ﬁnd the multiplicative inverse of [x + 1]. Factor x5 − 2x4 − 2x3 + 12x2 − 15x − 2 over Q. 11. In Z2 [x]/ x3 + x + 1 . (b) Show that x3 + 2x2 + 3 is irreducible over Z5 . but [x] is not. (b) in Z3 [x]/ x3 + 2x2 + x + 1 . Express 2x3 + x2 + 2x + 2 as a product of polynomials irreducible over Z5 . 14. 13. (b) List the elements of the ﬁeld F = Z3 [x]/ x2 + 1 . ﬁnd [x]−1 and [x + 1]−1 . (a) Express x4 + x as a product of polynomials irreducible over Z5 . Construct an example of a ﬁeld with 343 = 73 elements. 9.28 CHAPTER 4. 10. (c) In the multiplicative group of nonzero elements of F . and use your answers to ﬁnd [x2 + x]−1 . Factor x5 − 10x4 + 24x3 + 9x2 − 33x − 12 over Q. 8. .

(c) Find all idempotent elements of R. Show that although R has 4 elements. so it is crucial in many of the proofs you will see. Remember that the distributive law is all that connects the two operations in a ring. and all nilpotent elements of R. so you may need to review the last two sections of Chapter 3. 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. 3. Let R be the ring Z2 [x]/ x2 + 1 . The concept of a factor ring depends heavily on the corresponding deﬁnition for groups. (a) Show that R is a commutative ring (you only need to check closure and commutativity of multiplication). (b) Find all units of R. Find all ring homomorphisms from Z120 into Z42 . Review Problems 1. 4.Chapter 5 COMMUTATIVE RINGS This chapter takes its motivation from Chapter 1 and Chapter 4. 2. Are Z9 and Z3 ⊕ Z3 isomorphic as rings? 29 . it is not isomorphic to either of the rings Z4 or Z2 ⊕ Z2 . extending results on factorization to more general settings than just the integers or polynomials over a ﬁeld.

what is the largest possible order 180 of an element? 6. Let R be the ring Z2 [x]/ x4 + 1 . Let Z[i] be the subring of the ﬁeld of complex numbers given by Z[i] = {m + ni ∈ C | m. Let S be the ring Z2 [x]/ x3 + x . = (c) Is I a prime ideal of R? Hint: If you use the fundamental homomorphism theorem. For the element a = (0. and all prime ideals. Let R be any commutative ring with identity 1. (a) Show that if e is an idempotent element of R. 15.30 CHAPTER 5. n ∈ Z} . Prove that I is a prime ideal. then R ∼ Re ⊕ R(1 − e). Give an example to show that the set of all zero divisors of a ring need not be an ideal of the ring. 7. 2) of the ring R = Z12 ⊕ Z8 . prove that I is not maximal. COMMUTATIVE RINGS 5. then 1−e is also idempotent. (c) Find the idempotent elements of R. of Z36 = Z/36Z. (b) Show that if e is idempotent. In the group Z× of units of the ring Z180 . 8. (a) Show that I is an ideal of R. Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients. . Find all maximal ideals. Let R be the ring Z2 [x]/ x3 + 1 . 10. (a) Find all ideals of S. (b) Find the units of R. = 12. 14. Show that Ann(a) is an ideal of R. 9. 13. but not as rings. (a) Find all ideals of R. ﬁnd Ann(a) = {r ∈ R | ra = 0}. (c) Find the idempotent elements of R. 11. Show that the rings R and S in the two previous problems are isomorphic as abelian groups. you can do the ﬁrst two parts together. (b) Show that R/I ∼ Z2 [x]/ x2 + 1 . and let I be the set of all congruence classes in R of the form [f (x)(x2 + 1)]. (b) Find the units of R.

Hint: Show that a root of x2 = −1 leads to an element of order 4 in the multiplicative group Z× . then φ is onto. (b) For any prime number p.31 (a) Deﬁne φ : Z[i] → Z2 by φ(m + ni) = [m + n]2 . (a) Show that φ is a ring homomorphism. show that these two integral domains have the same quotient ﬁeld. 16. 18. and deﬁne the function φ : R → R/I ⊕ R/J by φ(r) = (r + I. deﬁne θ : Z[i] → Zp [x]/ x2 + 1 by θ(m + ni) = [m + nx]. Prove that the ring Zp [x]/ x2 + 1 is a ﬁeld. (b) Show that if I + J = R. = 17. p . and thus R/(I ∩ J) ∼ R/I ⊕ R/J . Considering Z[x] to be a subring of Q[x]. Find ker(φ) and show that it is a principal ideal of Z[i]. with ker(φ) = I ∩ J. Let p be an odd prime number that is not congruent to 1 modulo 4. Prove that φ is a ring homomorphism. Prove that θ is an onto ring homomorphism. for all r ∈ R. Let I and J be ideals in the commutative ring R. r + J).

COMMUTATIVE RINGS .32 CHAPTER 5.

4. √ √ 2. √ (b) Find the minimal polynomial of 2 + i over Q. Find the degree of 3 2 + i over Q. √ √ 7. √ √ √ 8. Review Problems 1. In Q(u). Find [Q( 7 16 + 3 7 8) : Q]. 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. i). √ 3. express (7 − 2u + u2 )−1 in the form a + bu + cu2 .Chapter 6 FIELDS These review problems cover only the ﬁrst three sections of the chapter. Show that x3 + 6x2 − 12x + 2 is irreducible over Q. Does 4 2 belong to Q( 3 2 + i)? 33 . 3 5) over Q. √ √ 6. If you are studying abstract algebra because you plan to be a high school teacher. Find a basis for Q( 5. Show that [Q( 2 + 3 5) : Q] = 6. (a) Show that Q( 2 + i) = Q( 2. Find the minimal polynomial of 1 + 3 2 over Q. √ √ 5.

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 3553 0 1 527 ; 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.

102. Let’s start by assuming that 10k + 2 is divisible by 6.. Conversely. etc. . Adding and subtracting 20 makes it possible to get this term. for some q ∈ Z. A4 = A · A3 . since it just takes a computation. Solution: To give a proof by induction. then we can show that the statement must also be true for n = k + 1. if An = I. (This “anchors” the induction argument. . we can use the division algorithm to write n = 4q + r. . 2. but we need to involve the expression 10k + 2 in some way. . the integer 10n + 2 is divisible by 6. We conclude that 4|n.) Clearly 12 is divisible by 6. say 10k + 2 = 6q. and then look at the expression when 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 . to divide n by 4. 30. say n = 4q. so r = 0 since A. . . The ﬁrst step is to check that the statement is true for n = 1. In the second half of the proof. and then it turns out that we can factor out 6. The ﬁrst half of the proof is easier than the second. and it is easier to state as the converse of the ﬁrst statement: if An = I. 1002. then An = A4q = (A4 )q = I q = I. you should read Appendix 4 in the text. The next step is to prove that if we assume that the statement is true for n = k. 10k+1 + 2 = (10)(10k ) + 20 − 20 + 2 = (10)(10k + 2) − 18 = (10)(6q) − (6)(3) = (6)(10q − 3) . and then show that the remainder has to be 0.38 CHAPTER 1 SOLUTIONS We also need to show that An = I only when 4|n. with 0 ≤ r < 4. 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. for n = 1. to get 10k+1 + 2 = (10)(10k ) + 2. A3 = A · A2 . is divisible by 6. we need a statement that depends on an integer n. 10002. We can easily factor a 10 out of 10k+1 .. Give a proof by induction to show that each number in the sequence 12. if An = I then we will use the division algorithm. We can write the numbers in the given sequence in the form 10n + 2. Comment: If you are unsure about doing a proof by induction. then 4|n. Then Ar = An−4q = An (A−4 )q = I · I q = I. A2 . so we can prove the following statement: for each positive integer n. and A3 are not equal to I.

284 = 72 · 3 + 68. Give the lattice diagram of all divisors of 250. 72 = 68 · 1 + 4. 1. 1089). 1492 = 284 · 5 + 72. This completes the induction. Solution: Since 1776 = 24 · 3 · 37 and 1492 = 22 · 373. 24. Solution: We have 1776 = 1492 · 1 + 284. Thus 21 = 17 · 1 + 4. Do the same for 484. to give the factors along the corresponding axis. (b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776. 185 = 164 · 1 + 21. 25. we see that 1274 and 1089 are relatively prime.9 shows that gcd(1776. then 10k+1 + 2 is divisible by 6. Solution: Since 1274 = 2 · 72 · 13 and 1089 = 32 · 112 .2. For example. 1089) = 1. 250 125 50 242 484 44 25 10 121 22 4 5 2 11 2 1 1 . 10. (a) Use the Euclidean algorithm to ﬁnd gcd(1776.2 SOLUTIONS 23. gcd(1274. 17 = 4 · 4 + 1. (b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274.CHAPTER 1 SOLUTIONS 39 We have now shown that if 10k + 2 is divisible by 6. Proposition 1. we need to use one axis for each prime. 1089 = 185 · 5 + 164. 1492) = 22 . Solution: The prime factorizations are 250 = 2 · 53 and 484 = 22 · 112 . 68 = 4 · 17. 164 = 21 · 7 + 17. Thus gcd(1776. 1492) = 4. These numbers go along one axis of the rectangular diagram. in succession. 1492). and 2. Solution: We have 1274 = 1089 · 1 + 185. In each diagram. 1492). dividing 250 by 5 produces 50. Then we can just divide (successively) by the prime. 1089). (a) Use the Euclidean algorithm to ﬁnd gcd(1274.

First. since the complementary factor must be equal to y − 3. then gcd(2m − 1. b2 ) = 1 if and only if gcd(a. 2n − 1) = 1. you don’t have to explain how you found them. −16). Solution: If gcd(m. Prove that n − 1 and 2n − 1 are relatively prime. Then we will prove the converse. x + 2 = 19. b) = 1. Using c = b gives gcd(a. 2). then gcd(m. n) = 1. Is the same true for 2n − 1 and 3n − 1? Solution: We can write (1)(2n − 1) + (−2)(n − 1) = 1. Substituting x = 2m and k = a in the identity given above shows that 2m − 1 . (−3. which proves that gcd(2n − 1. Comment: We need to do the proof in two parts. For each of these values there is a corresponding value for y. bc) = 1 if and only if gcd(a. (17. 2n − 1) = 1. b) = 1 if and only if gcd(a2 . and so gcd(3n − 1. we recall that the following identity holds for all values of x: xk − 1 = (x − 1)(xk−1 + xk−2 + · · · + x + 1). or x + 2 = −19. n − 1) = 1. then there exist a. c) = 1.3 (d) states that gcd(a. and (−21. b2 ) = 1. for all integers n > 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. Prove that gcd(2m − 1. 22). 2n − 1) = 1. we will prove that if gcd(m. we have the four solutions (−1. x + 2 = −1. (2)(3n − 1) + (−3)(2n − 1) = 1. Comment: Is this really a proof? Yes–producing the necessary linear combinations is enough. 2n − 1) = 1 if and only if gcd(m. Solution: If we had a product. 28. Find all integer solutions of the equation xy + 2y − 3x = 25. 4). Therefore we have 4 possibilities: x + 2 = 1. b) = 1 and gcd(a. b ∈ Z with am + bn = 1. Before beginning the proof. For positive integers a. y). b2 ) = 1 if and only if gcd(a. Let m and n be positive integers. Then a similar argument yields gcd(a2 . 2n − 1) = 1. we will use a proof by contradiction. n) = 1. Similarly. To prove the converse.2. Listing the solutions as ordered pairs (x. b.40 CHAPTER 1 SOLUTIONS 26. prove that gcd(a. which states that if gcd(2m − 1. Solution: Proposition 1. b2 ) = 1. n) = 1. 27. n) = 1. the only way it can be factored is to have 1 · 19 = 19 or (−1) · (−19) = 19. we could use the prime factorization theorem. That motivates one possible method of solution. assuming that gcd(m. n) = 1 and showing that this forces gcd(2m − 1. 29.

Among these. etc. 86 (mod 90).. 2n2 + 4n − 3) = gcd(2n2 + 4n − 3. Equivalently. This reduces to 7x = 2 + 15q. 31. Prove that gcd(2n2 + 4n − 3. etc. 1) = 1. Solution: We have gcd(42. we see that 7 is a factor of −14. 56. 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. getting gcd(2n − 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. . n) = 1. We next use trial and error to look for the multiplicative inverse of 7 modulo 15. 26. 90) = 6. which proves that gcd(2m − 1. 30. Therefore gcd(2m − 1. 46. Thus we have −14x ≡ −4 (mod 15). and this gives a quotient of n + 2 and a remainder of n − 1.) 1. If gcd(m. The next step is to divide 2n2 + 4n − 3 by 2n − 1. q ∈ Z with m = dq and n = dp. −34. (Of course. 61. 41. say 2am − 1 = (2m − 1)(s). The numbers congruent to 1 modulo 15 are 16. Solution: We can use the Euclidean algorithm. you could also continue with the Euclidean algorithm. n) = d. The same argument shows that we can write 2bn − 1 = (2n − 1)(t). 2n − 1) = 1. We have shown that gcd(2n2 + 6n − 4. and this completes the proof. The solution is x ≡ 11. then there exist p. −29. 71. 2n − 1) = 1. so there is a solution since 6 is a factor of 12. or 7x ≡ 2 (mod 15). so we multiply both sides of the congruence by −2 since (−2)(7) = −14 ≡ 1 (mod 15).3 SOLUTIONS 26. for some s ∈ Z. 2n − 1) = gcd(2n − 1. and −14.CHAPTER 1 SOLUTIONS 41 is a factor of 2am − 1. for all integers n > 1. Solve the congruence 42x ≡ 12 (mod 90). n − 1) = gcd(n − 2. for some t ∈ Z. 2n2 + 6n − 4) = 1. 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. we obtain 7x ≡ 2 (mod 15) by dividing 42x ≡ 12 (mod 90) through by 6. The proof now involves what may look like a trick (but it is a useful one). say gcd(m. or x ≡ 11 (mod 15). 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. Solving the congruence 42x ≡ 12 (mod 90) is equivalent solving the equation 42x = 12 + 90q for integers x and q.

and 3x ≡ 2 (mod 13).42 27. We have x = 13 + 17q for some q ∈ Z. Solution: Convert the second congruence to the equation x = 4 + 10q for some q ∈ Z. 75) = 5 is not a divisor of 36. (a) Find one particular integer solution to the equation 110x + 75y = 45. 17. 27) will also determine a solution. . since 110(15) − 75(22) = 0. Comment: The matrix computation shows that 110(15) + 75(−22) = 0. for any integer q. 27x ≡ 18 (mod 13). Solution: We have gcd(55. Then 4 + 10q ≡ 2 (mod 9). y = 5 is a solution. Solve the system of congruences x ≡ 2 (mod 9) x ≡ 4 (mod 10) . or q ≡ 11 (mod 13). (a) Find all solutions to the congruence 55x ≡ 35 (mod 75). then so is x = m + 15q and y = n − 22q. 7 · 5 ≡ 1 (mod 17) and 9 · 3 ≡ 1 (mod 13). . 5x ≡ 14 (mod 17) 3x ≡ 2 (mod 13) . The solution is 55x ≡ 36 (mod 75). so 40q ≡ 50 (mod 13). Solve the system of congruences so 5x ≡ 14 (mod 17). Solution: If 110m + 75n = 45. . 35x ≡ 98 (mod 17). which is a divisor of 35. 11x ≡ 7 (mod 15). −x ≡ 13 (mod 15). 1 0 110 1 −1 35 1 −1 35 15 −22 0 . 29. This leads to the answer. and multiplying by 9 yields a solution x = −18. Second solution: The equation reduces to the congruence 35x ≡ 45 (mod 75). (b) Find all solutions to the congruence Solution: There is no solution. which reduces to q ≡ 7 (mod 9). we can solve it in the usual way. so adding any multiple of the vector (15. and multiplying both sides by −2 gives x ≡ −3 (mod 15). Thus the solution is x ≡ 74 (mod 90). Solution: Any linear combination of 110 and 75 is a multiple of the gcd. (b) Show that if x = m and y = n is an integer solution to the equation in part (a). CHAPTER 1 SOLUTIONS 55x ≡ 35 (mod 75). 0 1 75 0 1 75 −2 3 5 −2 3 5 Thus −2(110) + 3(75) = 5. 32. 44x ≡ 28 (mod 15). x ≡ 13 (mod 17) x ≡ 5 (mod 13). 62 (mod 75). Solution: By trial and error. 47. since gcd(55. then 110(m + 15q) + 75(n − 22q) = 45 + 110(15)q + 75(−22)q = 45. −22) to the particular solution (−18. This reduces to 4q ≡ 5 (mod 13). Thus 75y = 45 + 3(110) = 375 and so x = −3. x ≡ 13 + 17 · 11 ≡ 200 (mod 221). 75) = 5. and then 13 + 17q ≡ 5 (mod 13). 28. 30. This reduces to 7x ≡ 9 (mod 15). Thus we have x ≡ 2. x ≡ 2 (mod 15). y = 27. Having reduced the system to the standard form.

(±4)2 ≡ 3 (mod 13). Or you can write 2200 as the prime factorization. This shows that the only possible units digits for n4 are 0. We have done the computation for representatives of each congruence class. n to provide an example of a system x ≡ a (mod m) that has no solution. and substitute to get 23 + 32q ≡ 5 (mod 25). 13) = 1. 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). and 6.CHAPTER 1 SOLUTIONS 31. 32. Give integers a. 1. We have 42 ≡ 6 (mod 10). 1. This gives x ≡ 55 (mod 25 · 32). which reduces to 7q ≡ 7 (mod 25). (±3)4 = 81 ≡ 1 (mod 10). Prove that the fourth power of an integer can only have 0. m. (±3)2 = 9. Solution: This is equivalent solving the congruence 4(n2 + 1) ≡ 0 (mod 13). since 4100 ≡ 1100 ≡ 1 (mod 3). and then 62 ≡ 6 (mod 10). This is the clue to take m = n = 2. it is really asking to ﬁnd n4 (mod 10). All we need to do is to compute the fourth power of each congruence class modulo 10: 04 = 0. 5. Since gcd(4. for all positive integers n. to get n2 ≡ −1 (mod 13). we can cancel 4. Just computing the squares modulo 13 gives us (±1)2 = 1. or 6 as its units digit. (±5)2 ≡ −1 (mod 13). and then (3. 35. 2200 ) = 1. x ≡ b (mod n) . Solve the system of congruences x ≡ 5 (mod 25) x ≡ 23 (mod 32) . and 54 ≡ 52 ≡ 5 (mod 10). (±1)4 = 1. Solution: In the example the integers m and n cannot be relatively prime. Solution: Since the question deals with the units digit of n4 . (±4)4 ≡ 62 ≡ 6 (mod 10). (±2)2 = 4. 36. Thus we must compute the congruence class of 4100 (mod 10). 43 Solution: Write x = 23 + 32q for some q ∈ Z. (±2)4 = 16 ≡ 6 (mod 10). 34. so q ≡ 1 (mod 15). (b) Is 4100 divisible by 3? Solution: No. (a) Compute the last digit in the decimal expansion of 4100 . with a = 1 and b = 0. Solution: The last digit is the remainder when divided by 10. so the answer to the original question is x ≡ ±5 (mod 13). 5. Find all integers n for which 13 | 4(n2 + 1). Thus 4100 = (42 )50 ≡ 650 ≡ 6 (mod 10). and (±6)2 ≡ −3 (mod 13). Solution: This could be proved by induction. 33. Prove that 10n+1 + 4 · 10n + 4 is divisible by 9. b.

if possible (in Z× ). 11. 34. 20 20 . 31 | 651. 1 −1 682 −4 5 651 1 0 . 1 −5 46 −1 6 45 . We have [3]−1 = [7]20 . [4]−1 = [10]13 . and [19]−1 = [19]20 . and we see that [3]13 and [−4]13 are inverses. 1 −5 46 0 1 91 . 7. 25. Solution: Note that 14. while 12.4 SOLUTIONS 30. if possible (in Z× ). Solution: The units of Z20 are the equivalence classes represented by 1. 9. and 3 · 5 = 15 ≡ 1 (mod 7). and 40 are congruent to 1. 4061 4061 Solution: The inverse does not exist. this 13 13 13 13 takes care of all of the nonzero elements of Z13 . 27. 501 501 Solution: We need to use the Euclidean algorithm. 1 −1 682 0 1 3379 . It is useful to list the integers with m with m ≡ ±1 (mod n). 20 20 20 [13]−1 = [17]20 . Comment: We know that Zn has ϕ(n) units. 2 · 4 = 8 ≡ 1 (mod 7). n) = 1. as are [−a]n and [−b]n . then [a]n and [−b]n are inverses. Using 14. and [6]13 and [−2]13 . ﬁnd all nilpotent elements. 2 −11 −1 6 1 45 Thus [91]−1 = [−11]501 = [490]501 . and so (4061. Finally. 13. then [a]n and [b]n are inverses. They occur in pairs. [9]−1 = [9]20 . Comment: If ab ≡ 1 (mod n). as are the pairs [4]13 and [−3]13 . 0 1 4061 3379 5 −6 −4 5 .44 CHAPTER 1 SOLUTIONS 1. Using 40. [3]−1 = [9]13 . Find the multiplicative inverse of each nonzero element of Z13 . 17. Solution: Since 6 ≡ −1 (mod 7). 501 33. and 19. If ab ≡ −1 (mod n). and [3]7 and [5]7 are inverses of each other. 1 0 501 0 1 91 . ﬁnd all idempotent elements. since gcd(a. so [2]7 and [4]7 are inverses of each other. Find the multiplicative inverse of each nonzero element of Z7 . n) = 1 if and only if gcd(n − a. Since [12]−1 = [−1]−1 = [−1]13 = [12]13 . and look at the various ways to factor them. and 39 are congruent to −1. as are [−a]n and [b]n . we see that [5]13 and [8]13 are inverses. 13 13 13 [5]−1 = [8]13 . Using 12. Find [91]−1 . 3379) = 31. 32. the class [6]7 is its own inverse. In Z20 : ﬁnd all units (list the multiplicative inverse of each). [11]−1 = [11]20 . here is the list of inverses: [2]−1 = [7]13 . 31. This helps to check your list. Find [3379]−1 . [6]−1 = [11]13 . Furthermore. 3. 31 651 At the next step. we see that [2]13 and [7]13 are inverses.

It is known than if p is prime. but there is no known algorithm for p actually ﬁnding the one element whose powers cover all of Z× . so the multiplicative order of [8] is 4. 13. c) = 1. There are 24 possible products of the form [8]i [−4]j . [9]24 . so the next possibility to check is 8. you can use a the hint in Exercise 1. We have [3]2 = [9]. this forces i = m and n = j. [−4]3 = [−4][16] = [6]. and since the only power of [8] that is equal to a power of [−4] is [1] (as shown by our computations). and [−4]6 = [−4][−9] = [1]. 5. [18]24 . 35. and 23. Since [2]8 = [−1]2 = [1]. and the nilpotent elements are [0]24 . and so each element is its own inverse. [−4]5 = [−4][11] = [−9]. For each of these numbers we have x2 ≡ 1 (mod 24). 7. j. it follows that [2] has multiplicative order 8. and [8]4 = [−6]2 = [1]. ﬁnd all idempotent elements.4. [8]3 = [8][−6] = [−13]. [−4]4 = [−4][6] = [11]. We next try [3]. so [3] does in fact have multiplicative order 16.40. and we are done. [16]24 . and [3]8 = [16] = [−1]. They are [0]20 and [10]20 . so the multiplicative order of [−4] is 6. [12]24 . 19. Problem 39 shows that the multiplicative order of an element has to be a divisor of 16. 36. 17 Comment: To show that Z× is cyclic. and [16]20 . p Solution: We begin by trying [2]. [5]20 . [6]24 . 11. then any solution to the congruences x ≡ 1 (mod b) and x ≡ 0 (mod c) will be idempotent modulo n. 37. Show that Z× is cyclic. 35 35 35 for some positive integers i. [3]4 = [81] = [−4]. The idempotent elements are [0]24 . or by using Problem 1. They are [0]20 . If you want a more systematic approach. The only divisor of 16 that is left is 16 itself.13 of the text: if n = bc. We have [2]2 = [4]. 17. . Solution: We ﬁrst compute the powers of [8]: [8]2 = [−6]. [1]20 . we need to ﬁnd an element whose 17 multiplicative order is 16. and [2]4 = [16] = [−1]. Then [8]i−m = [−4]n−j . ﬁnd all nilpotent elements. The solution just uses trial and error. [1]24 .CHAPTER 1 SOLUTIONS 45 The idempotent elements of Z20 can be found by using trial and error. for some 0 ≤ i < 4 and 0 ≤ j < 6 and 0 ≤ m < 4 and 0 ≤ n < 6. The nilpotent elements of Z20 can be found by using trial and error. In Z24 : ﬁnd all units (list the multiplicative inverse of each). Show that Z× is not cyclic but that each element has the form [8]i [−4]j .4. Solution: The units of Z24 are the equivalence classes represented by 1. with gcd(b. for 0 ≤ i < 4 and 0 ≤ j < 6. We next compute the powers of [−4]: [−4]2 = [16]. Are these all diﬀerent? Suppose that [8]i [−4]j = [8]m [−4]n . then Z× is cyclic. and the powers we have listed represent the only possible values of [8]i . [2]3 = [8].

([8]i [−4]j )12 = ([8]4 )3i ([−4]6 )2j = [1]. so we must have r = 0 since r < k and k is the smallest positive integer with ak ≡ 1 (mod n). Thus ar = aϕ(n)−kq = aϕ(n) (ak )−q ≡ 1 (mod n). where 0 ≤ βj ≤ αj for all j. which shows that aϕ(n) ≡ 1 (mod n). Solution: Assume that k is the smallest positive integer for which ak ≡ 1 (mod n). Find gcd(7605. 49 −65 −3 4 1 0 0 117 0 1 7605 5733 . then we must have a = t 1 2 pβ1 pβ2 · · · pβt d. 5733). then k | ϕ(n). and let a ∈ Z with gcd(a. for any integer n.4.11. then n | ak . where 0 ≤ r < k. Since ak ≡ 1 (mod n). then it must be a divisor of a. 5733) = 117. If n = pα1 pα2 · · · pαt is the prime factorization of n. and express it as a linear combination of 7605 and 5733. so each prime n divisor of n is a divisor of ak . and 117 = (−3) · 7605 + 4 · 5733. n Conversely. SOLUTIONS TO THE REVIEW PROBLEMS 1. showing that Z35 is not cyclic. and so we can apply Theorem 1. and this completes the proof.4. Solution: First assume that each prime divisor of n is a divisor of a. n) = 1. Prove that if k is the smallest positive integer for which ak ≡ 1 (mod n). prove that ω n = 1 if and only if 3|n. 11 Solution: We can factor [x]2 + [x] − [6] = ([x] + [3])([x] − [2]). 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 . 2. √ 3 1 i.6 implies that either [x] + [3] = [0] or [x] − [2] = [0]. Solve the equation [x]2 + [x]11 − [6]11 = [0]11 . so no element of Z35 has multiplicative order 24. every element in Z35 must be of this form. We can use the division algorithm to write ϕ(n) = qk + r. But if a prime p is a divisor of ak . 38. we know that gcd(a. 40. 1 −1 1872 −3 4 117 . Prove that [a]n is a nilpotent element of Zn if and only if each prime divisor of n is a divisor of a. and q ∈ Z. Corollary 1. Finally. 39. If k is the smallest positive t 1 2 integer such that kβi ≥ αi for all i. with [a]k = [0]. n) = 1. . and so the solution is [x] = [−3] or [x] = [2]. if [a]n is nilpotent. then n | ak . Let n be a positive integer. Solution: Use the matrix form of the Euclidean algorithm: 1 −1 1872 0 1 5733 . Thus gcd(7605. For ω = − + 2 2 .

and [2]4 = [1]. [4]. Trial and error shows it to be −8. The same computation shows that [2]−1 = [8] and [4]−1 = [4]. . if n ∈ Z and ω n = 1. To compute the multiplicative order of [8]. [13]. [7]. since 3k must be a multiple of 4. Solution: Write x = 8 + 11q for some q ∈ Z. so [7] has multiplicative order 4 because [7]4 = [4]2 = [1]. [11] = [−4] has multiplicative order 2. 182 (mod 200). and proceed with the solution. we have [2]2 = [4]. This shows not only that the multiplicative order of [2] is 4. 3. 57. Then 2 2 ω n = ω 3q = (ω 3 )q = 1q = 1. ﬁnd its multiplicative inverse. 24x 3x −24x x ≡ 168 ≡ 21 ≡ −168 ≡ 7 (mod (mod (mod (mod 200) 25) 25) 25) The solution is x ≡ 7. 200) = 8. [8]. the only possibility is r = 0. 157. Then 1 = ω n = ω 3q+r = (ω 3 )q ω r = ω r . Solve the system of congruences 2x ≡ 9 (mod 15) x ≡ 8 (mod 11) . we can multiply both sides of the congruence by −8. [2]. so the congruence has a solution. and 15 ﬁnd its multiplicative order. (This can also be shown by rewriting [8] as [−7]. and substitute to get 16 + 22q ≡ 9 (mod 15). then n = 3q for some q ∈ Z. Solve the congruence 24x ≡ 168 (mod 200). 1. 2 and we have shown that ω = 1 and ω 2 = 1. 107. 5. 132. By Problem 39. use the division algorithm to write n = q · 3 + r. but that the multiplicative order of [4] is 2. If n ∈ Z. [11]. We can also deduce that [13] = [−2] has multiplicative order 4. and 8 | 168. [2]3 = [8]. 4. To solve the congruence 3x ≡ 21 (mod 25) we could ﬁnd the multiplicative inverse of 3 modulo 25. and [14]. which reduces to 7q ≡ −7 (mod 15). and that [11]−1 = [−4]−1 = [−4] = [11]. 4.) Similarly. Since r = 0. The next step is to reduce the congruence by dividing each term by 8. This gives x ≡ −3 (mod 11 · 15). the multiplicative order of any nontrivial element is 2. List the elements of Z× . we can rewrite it as [2]3 . 32. or 8. Solution: There should be 8 elements since ϕ(15) = 8. and ω 3 = 1. which gives 24x ≡ 168 (mod 200). and then it is clear that the ﬁrst positive integer k with ([2]3 )k = [1] is k = 4. where the remainder satisﬁes 0 ≤ r < 3. For each element. Computing powers. that [13]−1 = [−2]−1 = [−8] = [7]. Conversely. and [13] = [−2] has multiplicative order 4. Solution: First we ﬁnd that gcd(24. and therefore 3|n. and 3|n. we have [7]2 = [4].CHAPTER 1 SOLUTIONS 47 Solution: Calculations in the introduction to Chapter 1 show that ω 2 = √ 1 3 − − i. 82. Next. so q ≡ −1 (mod 15). The elements are [1].

.8 shows that to compute ϕ(2n) in terms of ϕ(n) 1 we need to add 2 · (1 − 2 ). and this does not change the computation. Solution: Since n is odd. the integers n and 2n are relatively prime.27 of the text that ϕ(2n) = ϕ(2)ϕ(n) = ϕ(n). and so it follows from Exercise 1. the prime 2 does not occur in its prime factorization.48 CHAPTER 1 SOLUTIONS 6.4. Second solution: Since n is odd. then ϕ(2n) = ϕ(n).4. The formula in Proposition 1. Show that if n > 1 is an odd integer.

the uniqueness part of the deﬁnition translates directly into the “vertical line test”.4. then we cannot have y1 = y2 while x1 = x2 .4. According to Deﬁnition 2. y1 ) and (x2 . if (x1 . In this context. the “horizontal line test” is exactly the same as the condition given in Deﬁnition 2.1.1.1. According to Deﬁnition 2.1. more than one 49 . y2 ) line on its graph. y2 ) for which x1 = x2 but y1 = y2 . 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. Thus if we apply Deﬁnition 2. Explain why this agrees with Deﬁnition 2.1.1 to the given curve. 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.1. y2 ) on the curve with x1 = x2 and y1 = y2 .Chapter 2 Functions 2. then the points of intersection represent points (x1 . If a vertical line intersects the curve in two distinct points. Equivalently.1.1 SOLUTIONS 20.1. y1 ) and (x2 .4. 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. 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. 21. y1 ) and (x2 . y) belongs to the subset of the plane. a function is one-to-one if f (x1 ) = f (x2 ) implies x1 = x2 . Solution: If a horizontal line intersects the graph of the function more than once. Explain why this agrees with Deﬁnition 2.

x) lies on the graph of f . if (x. where x = g(y). and that this line segment intersects the line y = x at the midpoint ((a + b)/2. a) has slope −1. On the other hand. 23. This shows that the point (x.50 CHAPTER 2 SOLUTIONS 22. 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 A is invertible. where I is the n × n identity matrix. Explain why this agrees with Deﬁnition 2. if det(A) = 0. (b) Show that if L is either one-to-one or onto. then y = f (x). Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax. and AB = I. To prove part (b) we need to use the Rank–Nullity Theorem. Since the matrix A is invertible if and only if it has rank n.7. b) and (b. and so f has an inverse. and L is one-to-one if and only if the nullity of A is zero. This can be seen by observing that the line segment joining (a. On the other hand. then x = f −1 (y). and this is the image of the transformation L. for all x ∈ Rn . y) lies on the graph of f −1 . and the point (x. y) lies on the graph of f . This shows that g = f −1 . and so L has an inverse. (a + b)/2) of the segment. and therefore y = f (f −1 (y)) = f (x). First. y) lies on the graph of f −1 . Conversely. and so f −1 (y) = f −1 (f (x)) = x. Thus A is an invertible matrix. For any real number x. (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. deﬁned by L−1 (x) = A−1 x. the nullity of A is the dimension of the solution space of the equation Ax = 0. . and so det(A) = 0. then it is invertible. If f : R → R has an inverse. since Ax1 = Ax2 if and only if A(x1 − x2 ) = 0. In calculus the graph of an inverse function f −1 is obtained by reﬂecting the graph of f about the line y = x. b) in the line y = x is the point (b. Solution: We ﬁrst note that the reﬂection of a point (a. a). then it can also be described by multiplication by a matrix B. Let A be an n × n matrix with entries in R. suppose that the graph of the function g is deﬁned by reﬂecting the graph of f in the line y = x. then the rank of A plus the nullity of A is n. which must satisfy the conditions BA = I. and then the Rank–Nullity Theorem shows that this happens if and only if L is one-to-one. On the other hand. which states that if A is an n × n matrix. if L has an inverse. so L is onto if and only if A has rank n.1. it follows that L is invertible if and only if L is onto. Solution: The rank of the matrix A is the dimension of the column space of A. which makes it perpendicular to the line y = x. which shows that (y. for all x ∈ Rn .

Solution: As noted in the solution to problem 23. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax. for all x ∈ Rn . Let A be an m × n matrix with entries in R. Prove that L is one-to-one if and only if no eigenvalue of A is zero. then KL is the identity function on Rm . for all x ∈ Rm . Ax1 = Ax2 if and only if A(x1 − x2 ) = 0. Solution: If det(AT A) = 0. and assume that m > n. and so det(AT A) = 0 if and only if the column vectors of A are linearly independent. and so L is one-to-one if and only if Ax = 0 for all nonzero vectors x. 26. According to the Rank–Nullity Theorem. 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. It then follows from Exercise 17 that L is one-toone.2 SOLUTIONS 14. for all x ∈ Z× . y2 ) if x1 y2 = x2 y1 . Solution: Since a has an inverse in Z× . This implies that θ is one-to-one and onto. Deﬁne the function θ : Z× → Z× by 17 17 17 θ(x) = ax. Deﬁne a linear transformation L : Rn → Rm by L(x) = Ax. which translates into the given statement about eigenvalues of A. This content is nonzero if and only if the vectors are linearly independent. this happens if and only if the nullity of A is zero. 2. If we deﬁne K : Rm → Rn by K(x) = (AT A)−1 AT x. 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. b)} of all ordered pairs of positive integers. then AT A is an invertible matrix. 25. Show that this deﬁnes an equivalence relation. L is a one-to-one linear transformation if and only if det(AT A) = 0. for all x ∈ Z× . Show that L is a one-to-one function if det(AT A) = 0. Is θ one to one? Is θ onto? If possible. In other words. On the set {(a. . deﬁne (x1 . for all x ∈ Rn . Let A be an n × n matrix with entries in R. which shows that ψ = θ−1 . we can deﬁne ψ : Z× → Z× by 17 17 17 ψ(x) = a−1 x. Let a be a ﬁxed element of Z× . Comment: There is a stronger result that depends on knowing a little more linear algebra. y1 ) ∼ (x2 . where AT is the transpose of A.CHAPTER 2 SOLUTIONS 51 24. This is equivalent to the statement that there is no nonzero vector x for which Ax = 0 · x. 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. ﬁnd the 17 inverse function θ−1 .

Solution: The reﬂexive. z) | L(x. a3 ) ∼ (b1 . a2 . deﬁne z1 ∼ z2 if ||z1 || = ||z2 ||. b2 ) ∼ (a3 . b1 ) ∼ (a2 . For vectors v and w. so a ∼ b if and only if a2 = b2 . If we multiply the ﬁrst equation by b3 and the second equation by b1 . y. and assume that u has length 1. symmetric. x + y + z. y. we have ab = ba. We next check the symmetric law. Solution: Since (a1 . and (a3 . b2 ) ∼ (a1 . a2 .52 CHAPTER 2 SOLUTIONS Solution: We ﬁrst show that the reﬂexive law holds. b1 ) ∼ (a3 . and transitive laws for the relation ∼ really depend on an equality. give a geometric description of the partition of R3 that is determined by L. b2 ). we verify the transitive law. and so a2 b1 = a1 b2 . 0)} is the plane through the origin whose equation is x + y + z = 0. and transitive laws can be easily veriﬁed since ∼ is deﬁned in terms of an equality. and equality is itself an equivalence relation. For the linear transformation L : R3 → R3 deﬁned by L(x. we get a1 b2 b3 = a2 b1 b3 = a3 b1 b2 . we have the equations a1 b2 = a2 b1 and a2 b3 = a3 b2 . For example. for all x ∈ R. or. Since b2 = 0 we can cancel to obtain a1 b3 = a3 b1 . 15. Given an ordered pair (a. b2 ) and (a2 . and give a geometric description of the equivalence classes of ∼. symmetric. y. 1. On the set C of complex numbers. Since u has length 1. b3 ). describe the equivalence relation on R that is determined by f . It follows that the equivalence classes of ∼ are the planes in R3 that are perpendicular to u. z) ∈ R3 . Given (a1 . and can easily be veriﬁed. b) ∼ (a. The other subsets in the partition of R3 deﬁned by L are planes . {(x. it follows from the deﬁnition of L that (a1 . a2 . For the function f : R → R deﬁned by f (x) = x2 . (a2 . b3 ). 16. b3 ) with (a1 . for all (x. 18. Thus two vectors are equivalent if and only if they lie in the same plane perpendicular to u. b2 . where · denotes the standard dot product. z) = (x + y + z. b2 . with normal vector (1. 0. b1 ) and (a2 . a ∼ b if and only if |a| = |b|. b2 . b1 ). Solution: The equivalence relation determined by f is deﬁned by setting a ∼ b if f (a) = f (b). showing that (a1 . z) = (0. Given (a1 . a3 ) ∼ (b1 . deﬁne v ∼ w if v ·u = w ·u. b1 ) ∼ (a2 . b). Show that ∼ is an equivalence relation. b1 ). 1). x + y + z) . b3 ) if and only if a1 + a2 + a3 = b1 + b2 + b3 . and so (a. Let u be a ﬁxed vector in R3 . v · u represents the length of the projection of v onto the line determined by u. Finally. b3 ) if L(a1 . Solution: The reﬂexive. y. b3 ). b). b2 ). a3 ) = L(b1 . Show that ∼ is an equivalence relation. b2 ) with (a1 . we have a1 b2 = a2 b1 . 17. which shows that (a2 .

3). 4. σ −1 = (1. What is the order of σ? Is σ an even permutation? Compute σ −1 . {[±2]12 . The corresponding equivalence classes determined by f are {[0]12 . 9. 8). and so its order is 12 since lcm[3. στ σ −1 . 4. 8. 4)(7. 9). 8). Thus 12 12 12 12 f (Z12 ) = {[0]12 . deﬁne A ∼ B if there exists an invertible matrix P such that P AP −1 = B. If A ∼ B. 5. 2)(4. and so A ∼ C. 4). 4. 3)(4. 1). 9. 7)(2. 9). τ . [±5]12 }. [1]12 . [±2]2 = 12 12 12 [4]12 . which shows that f ([x1 ]12 ) = 1 2 f ([x2 ]12 ). [9]12 }. στ . τ σ. 6). 2. στ σ −1 = (1. 9). On the set of all n × n matrices over R. Solution: We have σ = (1. then P AP −1 = B for some invertible matrix P . τ σ = (1. 6. Solution: The formula for f is well-deﬁned since if [x1 ]12 = [x2 ]12 . For the permutation σ = 1 2 3 4 5 6 7 8 9 . 5)(3. 9)(3. Thus the partition consists of the planes perpendicular to the vector (1. . Solution: σ = (1. 8. for all [x]12 ∈ Z12 . 4. To compute the images of f we have [0]2 = [0]12 . [±5]2 = [1]12 . 3. 7). 9. 8. and so x2 ≡ x2 (mod 12).CHAPTER 2 SOLUTIONS 53 parallel to this one. 7. We have σ −1 = (1. σ −1 . and [6]2 = [0]12 . If A ∼ B and B ∼ C. [±4]2 = [4]12 . 5)(3. 12 Show that the formula f deﬁnes a function. 8. since it can be expressed as the product of 6 transpositions. τ στ −1 . {[±3]12 }. 8)(2. στ = (1.3 SOLUTIONS 13. Solution: We have A ∼ A since IAI −1 = A. 5)(3. 19. [6]12 }. 2. 9. τ στ −1 = (1. [±4]12 }. 2. 5. 8. [±1]2 = [1]12 . 20. 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. and so we get A = P −1 B(P −1 )−1 . Q. then P AP −1 = B and QBQ−1 = C for some P. write σ as a 7 5 6 9 2 4 8 1 3 product of disjoint cycles. 1. 14. then x1 ≡ x2 (mod 12). 7. 2. 7. 5. Check that ∼ deﬁnes an equivalence relation. 2. 7. It is an even permutation. 5. Deﬁne the formula f : Z12 → Z12 by f ([x]12 ) = [x]2 . 8. 3. 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 . τ = (2. [±3]2 = [9]12 . τ −1 . 5)(3. 4. 5). 7. write each of these permutations as a 1 5 4 7 2 6 8 9 3 product of disjoint cycles: σ. τ −1 = (2. 4] = 12. where I is the n × n identity matrix. [4]12 . {[±1]12 .

7). then στ σ −1 has order m. 6. 11. On the other hand. but not 11 or 13. 9. 9) has order 14. 7. 20. so it has order 15 = lcm[5. 6. Let G be a group of permutations. any element of order 11 or 13 would have to be a cycle. 3. What is the order of σ? Compute σ −1 . 12. 8) = (1. 6. 3. 8. 7. 5. 7. . 6. and let X be a subset of S. 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). Let σ = (2. 6. 2. 10). 6 7 8 9 8 9 10 1 10 3 11 5 . Prove that if τ ∈ Sn is a permutation with order m. Prove that τ Gτ −1 = {σ ∈ Sym(S) | σ = τ γτ for some γ ∈ G} is a group of permutations. 2)(3. 8. 6. 8. 5. 18. 10)(3. 7. We have στ σ −1 = (3. 6) ∈ S9 . 17. 2. since (στ σ −1 )k = (1) implies στ k σ −1 = (1). 8. For Solution: Since τ = (1. 5. Prove that G is a group of permutations. On the other hand. )(6. 19. Solution: The permutation (1. since 11 and 13 are prime. the order of στ σ −1 cannot be less than n. 7)(1. Solution: Assume that τ ∈ Sn has order m. 5. so the cycle structure of στ σ −1 is the same as that of τ . Solution: We have σ = (1. Let G = {σ ∈ Sym(S) | σ(X) ⊂ X}. 7). 9)(3. Compute the order of τ = 1 2 3 4 5 7 2 11 4 6 σ = (3. 5). 4. Let S be a set. and then τ k = σ −1 σ = (1). 9. 7. and 14. it has order 6. 9)(3. 8. 9)(1. for any permutation σ ∈ Sn . 7. 11. 4. while the element (1. with G ⊆ Sym(S). 11. 6. 9)(8. 5. 6. 5. 16. 8. and there are no cycles of that length in S10 . and thus στ σ −1 has order 6. compute the order of στ σ −1 . 5. 3. Let τ be a ﬁxed permutation in Sym(S). 4. 2)(3. and σ −1 = (1. 4.54 CHAPTER 2 SOLUTIONS 15. 3]. 8. 7. Show that S10 has elements of order 10. 9)(2. 5. 8)(2. Write σ as a product of disjoint cycles. 6)(3. and (1. 7) has order 10. 7) has order 12. 10). 3)(4. for the set S.

2). write σ as a product of disjoint cycles. so φα is one-to-one. In S10 . Deﬁne the function φ : Z× → Z× by φ(x) = x−1 . given τ ∈ Sn . 7. for all σ ∈ Sn . Show that f is a one-to-one function. Solution: If φα (σ) = φα (τ ). and so φα is onto. 2. ﬁnd the inverse function φ−1 . Deﬁne f : R → R by f (x) = x3 + 3xz − 5. 5. Solution: Since (1. and φα ((1. 3. Finally. A short computation shows that (φα )−1 = φα−1 . let α = (1. . 7. let α = (1. (a) Let α be a ﬁxed element of Sn . For σ = αβγ. 6. for all x ∈ R. 6. and σ −1 = (1. it can be written as a product of 5 transpositions. Is φ one to 17 17 17 one? Is φ onto? If possible. Show that φα : Sn → Sn deﬁned by φα (σ) = ασα−1 . Hint: Use the derivative of f to show that f is a strictly increasing function. 3)) = (1. and use this to ﬁnd its order and its inverse. On the set Q of rational numbers. 6). 2)) = (1. then ασα−1 = ατ α−1 . 2. deﬁne x ∼ y if x − y is an integer. 3. so σ has order 6. is a one-to-one and onto function. 2)) = (1. 9. 7. β = (1. 2. direct computations show that φα ((1)) = (1). Solution: For all x ∈ Z× we have φ(φ(x)) = φ(x−1 ) = (x−1 )−1 = x. 2. 2. 2. 5. so 17 φ = φ−1 . Another way to show that φα is one-to-one and onto is to show that it has an inverse function. describe the equivalence relation on R that is determined by f . 2). 9). for all x ∈ Z× . we have φα (σ) = τ for σ = α−1 τ α. 3). 3)) = (1. 3). 9). 2). Since σ has length 6. which also shows that φ is one-to-one and onto. 3. Is σ even or odd? Solution: We have σ = (1. for all x ∈ R. Show that ∼ is an equivalence relation. τ ∈ Sn . 5. 3. 3)) = (2. and γ = (1. to get σ = τ . 3). φα ((2. 3.CHAPTER 2 SOLUTIONS 55 SOLUTIONS TO THE REVIEW PROBLEMS 1. so it is an odd permutation. φα ((1. Compute φα . 4. 2. 6). 2) is its own inverse. 3). For the function f : R → R deﬁned by f (x) = x2 . for σ. φα ((1. We can multiply on the left by α−1 and on the right by α. (b) In S3 . φα ((1. 3.

56 CHAPTER 2 SOLUTIONS .

so the cross product cannot be used to make the set of all vectors in R3 into a group. not a vector. y2 . 24. the cross product is deﬁned by (x1 . z1 ) and (x2 . 23. showing that the operation is closed on the set 2 G. deﬁne a new multiplication ab by a∗b = . y1 . b ∈ G. z1 x2 − x1 z2 . Is R3 a group under this multiplication? Solution: The cross product of the zero vector and any other vector is the zero vector.Chapter 3 Groups 3. z2 ) = (y1 z2 − z1 y2 . Show that G is a group under this multiplication. Does this make R3 into a group? Solution: The dot product of two vectors is a scalar. Use the dot product to deﬁne a multiplication on R3 . and so is . On the set G = Q× of nonzero rational numbers. y1 . 2 Solution: If a and b are nonzero rational numbers. This means that the dot product does not even deﬁne a binary operation on the set of vectors in R3 . y2 . again making it impossible to deﬁne a group structure.1 SOLUTIONS 22. then ab is a nonzero ab rational number. for all a. For vectors (x1 . Even if we were to exclude the zero vector we would still have problems. 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 . The cross product of two nonzero vectors deﬁnes a vector that is perpendicular to each of the given vectors. x1 y2 − y1 x2 ). z2 ) in R3 . z1 )×(x2 .

4 4 The number 2 acts as the multiplicative identity. We will write m for [m]9 . and the second table shows what happens when we arrange the elements in order. [2]9 . [4]9 . 25.58 and (a ∗ b) ∗ c = ab 2 ∗c= CHAPTER 3 SOLUTIONS ab 2 c 2 = (ab)c . · 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. Write out the multiplication table for Z× . We will 15 write the elements as {1. with the elements in a slightly diﬀerent order. [7]9 . Write out the multiplication table for Z× . [8]9 }. [13]15 . 7. −2. · 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 . [4]15 . 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. −7. 4. as successive powers of 2. [14]15 }. [7]15 . [5]9 . [2]15 . then a is a nonzero rational number that serves as the multiplicative inverse of a. 26. and if a is nonzero. gives a diﬀerent picture of the group. 9 Solution: Z× = {[1]9 . [8]15 . 2. −4. 15 Solution: Z× = {[1]15 . [11]15 . −1}.

The associative law for multiplication essentially says that parentheses don’t matter. we have to include the following steps. denoted by a−1 and b−1 . both a and b have inverses. If we are to be precise about using the associative law. Since G is a group. Show that (aba−1 )n = abn a−1 . Using this induction hypothesis. b(ab)b−1 = (ab2 )b−1 (ba)(bb−1 ) = (ab)(bb−1 ) ba = ab This completes the proof. Solution: Assume that a and b are elements of G for which (ab)2 = a2 b2 . we have the following calculation. so we don’t really need to include all of the steps we showed before. and suppose that a and b are any elements of G. ±2. Show that if (ab)2 = a2 b2 . Solution: To give a careful proof we need to use induction. then ba = ab. 27. then ba = ab. Let G be a group. for any positive integer n. respectively. Multiplication in G is well-deﬁned. which is certainly true. ±4.CHAPTER 3 SOLUTIONS 59 Comment: Notice how much easier it makes it to use the representatives {±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 . Let G be a group. Now assume that the result holds for n = k. 28. Expanding this equation gives us (ab)(ab) = a2 b2 . and suppose that a and b are any elements of G. The statement for n = 1 is simply that aba−1 = aba−1 . ±7} when listing the congruence classes in the group. 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 we have shown that if (ab)2 = a2 b2 . so we can multiply both sides of the equation on the left by a−1 without destroying the equality.

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. so by induction it holds for all values of n. and so G is a group under the given operation. Then a · (a · e) = (a · a) · e = e · e = e = a · a . Solution: Assume that the two replacement conditions hold. In summary. Solution: On the set G of nonzero real numbers. That is. F is the set of all functions f (z) : C → C of the form f (z) = cz + d .60 CHAPTER 3 SOLUTIONS Thus the statement holds for n = k + 1. This shows that e is a multiplicative identity for G. There is no right identity element. Thus the original condition (iv) holds. 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. 31. we can cancel a from the left of the above equation. 30. even though it has a left identity element and right inverses. replace condition (iii) with the condition that there exists e ∈ G such that e · a = a for all a ∈ G.3 of the text. 29. and replace condition (iv) with the condition that for each a ∈ G there exists a ∈ G with a · a = e. 1/|a| is a right inverse for any a ∈ G. b ∈ G. and then (as above) we can cancel a to get a · a = e. az + b . We also have the equation a · (a · a ) = (a · a) · a = e · a = a = a · e . In Deﬁnition 3. deﬁne the operation a ∗ b = |a|b. Let F be the set of all fractional linear transformations of the complex plane. and so the original condition (iii) is satisﬁed. so we have deﬁned a binary operation on G.1. Let a be an element in G with a · a = e. We will ﬁrst show that a · e = a. Then a ∗ b = 0 if a = 0 and b = 0. and since there exists an element a ∈ G with a · a = e. since 1 ∗ a = |1|a = a for all a ∈ G. but the equation x ∗ a = 1 has no solution for a = −1. for all a. Note the e·e = e. which shows that a is indeed the multiplicative inverse of a. for all a ∈ G. and that the associative law holds. we have shown that G is not a group. since the two equations 1 ∗ x = 1 and (−1) ∗ x = −1 have no simultaneous solution in G. Prove that these weaker conditions (given only on the left) still imply that G is a group. Finally. The number 1 is a left identity element. so −1 has no left inverse. to get a · e = a. 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.

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 = 1.1.2 in the text. 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 ) . 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. Show that F forms a group under composition of functions. b = 0.CHAPTER 3 SOLUTIONS 61 where the coeﬃcients a. c = 0. and d = 1. 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 .1. We need a better way to look at the entire situation. b. c. Finally. 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. Composition of functions is always associative (compare Exercise 3. d are integers with ad − bc = 1. for matrices). respectively.3. and f2 (z) = . 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. we can use the formula for the inverse of a 2 × 2 matrix with determinant 1 to ﬁnd an inverse function for . so let’s look at the following matrix product. Furthermore. with a1 d1 − b1 c1 = 1 and c1 z + d1 c2 z + d2 a2 d2 − b2 c2 = 1. and the identity function will serve as an identity element for F . as a fractional linear transformation. We only need to check that it can be written in the correct form. (c2 a1 + d2 c1 )z + (c2 b1 + d2 d1 ) You can see that verifying all of the axioms is going to be painful. so we need to check the closure axiom in Deﬁnition 3. a1 z + b1 a2 z + b2 Let f1 (z) = . All of this means that it is fair to use what we already know about matrix multiplication.

3. (b) Show that the associative law holds for ∗. including just the element itself and the identity [1]24 . (a) Show that the operation ∗ is closed on G. 32. Thus each nonzero element generates a subgroup of order 2. we need to solve a ∗ y = 2. and therefore a(b − 1) > (b − 1). . It follows immediately that ab − a − b + 2 > 1. (c) Show that 2 is the identity element for the operation ∗. This solution belongs to G since a > a − 1 implies a/(a − 1) > 1. This gives f −1 (z) = . y ∈ G.62 f (z) = CHAPTER 3 SOLUTIONS az + b dz − b . a∗(a/a−1) = a2 /(a−1)−a−a/(a−1)+2 = (a2 −a2 +a−a)/(a−1)+2 = 2. 24 Solution: You can check that x2 = 1 for all elements of the group. Finally. the one computation 2 ∗ y = 2y − 2 − y + 2 = y suﬃces to show that 2 is the identity element. This gives us the equation ay − a − y + 2 = 2. and completes the proof that cz + d −cz + a F forms a group under composition of functions. Let G = {x ∈ R | x > 1} be the set of all real numbers greater than 1. c ∈ G. (d) Show that for element a ∈ G there exists an inverse a−1 ∈ G. On the other hand. Solution: If a. so b − 1 > 0. Solution: For a. 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 . b ∈ G.2 SOLUTIONS 23. 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. For x. deﬁne x ∗ y = xy − x − y + 2. Thus a ∗ (b ∗ c) = (a ∗ b) ∗ c. Solution: Since the operation is commutative. Solution: Given any a ∈ G. then a > 1 and b > 1. Find all cyclic subgroups of Z× . which has the solution y = a/(a − 1).

2. Finally. the element (1. Let G be an abelian group. 2)(5. 6. 3. 3)(5. ﬁnd two subgroups of order 4. H is not cyclic since no element of H has order 4. 2. 2. since it will generate a cyclic subgroup with 12 elements. ((1. ±7. and therefore [x]21 · [y]21 = [xy]21 belongs to H. In G = Z× . so by Corollary 3. 7) has order 12. ((1.2. and let n be a ﬁxed positive integer. 4)(5. with g = an .2. 3)(5. It is easy to check that the subset H = {±[1]. It is natural to begin with [3]. These elements. and so g −1 has the right form to belong to N . 3)(5. 2. and so [3] is a cyclic subgroup of order 4. 26. 2. ±3. 7))3 = (5. 2. Since the order of a product of disjoint cycles is the least common multiple of their lengths. and K = {[x]21 | x ≡ 1 (mod 7)} . Solution: To ﬁnd a cyclic subgroup of order 4. then x ≡ 1 (mod 3) and t ≡ 1 (mod 3). 2. (a) Find the cyclic subgroup of S7 generated by the element (1. Solution: The subset H is ﬁnite and nonempty (it certainly contains [1]21 ). ((1.4 implies that it is a subgroup. Then there must exist elements a1 and a2 in G with g1 = an and g2 = an . 1 2 and so g1 g2 = an an = (a1 a2 )n .CHAPTER 3 SOLUTIONS 63 24. (b) Find a subgroup of S7 that contains 12 elements. 7))2 = (1. which 20 turns out to have order 4. 3. 2. nonempty subset of a known group. If [x]21 and [y]21 belong to H. 1 2 Finally. ±9}. Next. 25. 3)(5. 2). 7). Since H is a ﬁnite. show that 21 H = {[x]21 | x ≡ 1 (mod 3)} are subgroups of G. In Z× . so it follows that xy ≡ 1 (mod 3). 3. 3)(5. 7))5 = (1. 7). ((1. 3)(5. 7). The element [9] = [3]2 has order 2. form the cyclic subgroup generated by (1. 7))4 = (1. and why they must form a subgroup. 3). Show that N = {g ∈ G | g = an for some a ∈ G} is a subgroup of G.4 it is enough to show that H is closed under multiplication. 7). 2. Solution: First. we need to check the orders of elements in Z× = {±1. 3)(5. ±[9]} is closed. 3)(5. suppose that g1 and g2 belong to N . Solution: We only need to ﬁnd an element of order 12. 7). A similar argument shows that K is a subgroup of Z× . if g ∈ N . the subset N is nonempty since the identity element e can always be written in the form e = en . then g −1 = (an )−1 = (a−1 )n . together with (1. 2. Solution: We have ((1. one that is cyclic and one that is not 20 cyclic. The last equality holds since G is abelian. Corollary 3. 7))6 = (1). You do not have to list all of the elements if you can explain why there must be 12. 21 27.

we have 2n ≡ −1 (mod p). p Solution: Since 2n +1 = p. (a) Show that in Z× the order of [2]p is 2n. 30.5 for a discussion of the properties of complex numbers. It follows from this theorem that (cos(3π/4) + i sin(3π/4))8 = √ √ cos(6π) + i sin(6π) = 1. We can use Demoivre’s Theorem (Theorem A. elements − 2 2 2 2 Solution: It is probably easiest to change these complex numbers from rectangular coordinates into polar coordinates.2) to compute powers of complex numbers. Suppose that p is a prime number of the form p = 2n + 1. 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 . ﬁnd the order of the √ √ √ 2 2 2 2 + i and − − i. let θ = 0. Solution: The order of [2] is a divisor of |Z× | = p − 1 = 2n . 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. Thus the order of [2] is a divisor of 2n. (See Appendix A. A similar √ √ argument shows that − 22 − 22 i also has order 8. and therefore n is a power of 2. 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).64 CHAPTER 3 SOLUTIONS 28. show that cos θ − sin θ H= θ∈R sin θ cos θ is a subgroup of G. In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. so 2k ≡ 1 (mod p) since 2k − 1 < 2n + 1 = p. In the multiplicative group C× of √ complex numbers. sin(θ + φ) cos(θ + φ) Identity: To see that the identity matrix is in the set. (b) Use part (a) to prove that n must be a power of 2. 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.5. Existence of inverses: cos θ sin θ − sin θ cos θ −1 = cos(−θ) − sin(−θ) sin(−θ) cos(−θ) ∈ H. and squaring this yields 22n ≡ 1 (mod p). .) Each of the numbers has magnitude 1. This shows that [2] has order 2n. and so − 22 + 22 i has order 8 in C× . 29.

. Since [1]5 . and c+d = b+d. any matrix of this form commutes with A. Equating corresponding entries 1 1 c d a+c b+d shows that we must have 2a+b = 2a+c. Let K be the following subset of GL2 (R). 32. Compute the centralizer in GL2 (R) of the matrix Note: Exercise 3. 2c+d = a+c. 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.4 (b). 1 1 b a−b Solution: Let A = 3. There are 8 such elements. Show that Z5 × Z3 is a cyclic group. then order of an element ([a]5 . ad − bc = 0 65 Show that K is a subgroup of GL2 (R). or d = a − b. On the other hand. [3]5 . c = −2b. K= a b c d d = a. and suppose that X = belongs to 1 1 c d the centralizer of A in GL2 (R). [2]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 = . the element ([a]5 . Then we must have XA = AX. since for any element in H the determinant is a2 + 2b2 .3 SOLUTIONS 16. The ﬁrst and last equations imply that b = c. b ∈ R and ab = a2 + b2 . which can easily be listed. and list all of the generators for the group. a+b = 2b+d. Solution: By Proposition 3.CHAPTER 3 SOLUTIONS 31. and a b −2b a −1 . The 1 a −b .14 in the text deﬁnes the centralizer of an element a of the group G to be C(a) = {x ∈ G | xa = ax}.3. [b]3 ) in Z5 × Z3 is the least common multiple of the orders of the components. [2]3 have order 3 in Z3 . 2 1 a b . [4]5 have order 5 in Z5 and [1]3 . [b]3 ) is a generator if and only if [a]5 = [0]5 and [b]5 = [0]5 . = 2 1 1 1 . while the second and third equations imply that a = b + d = c + d.2. a2 + 2b2 −2(−b) a Comment: We don’t need to worry about the condition ad − bc = 0. so the centralizer in GL2 (R) of the ma2 1 a b trix is the subgroup a.

it cannot be a scalar multiple of that row. 20.26) 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). Therefore the total number of ways to construct a second row independent of the ﬁrst is pn − p.2. 12) = 3. Similarly. and ﬁnd the subgroup HK. (A more formal proof could be given by induction. o([15]18 ) = o([3]18 ) = 6. A short computation shows that H = {[1]. 17. Find two groups G1 and G2 whose direct product G1 × G2 has a subgroup that is not of the form H1 × H2 . the set HK must contain 12 elements. let H = {[x] | x ≡ 1 (mod 4)} and K = {[y] | y ≡ 36 1 (mod 9)}. Show that H and K are subgroups of G. Thus the order of ([9]12 . Solution: In Z2 × Z2 . Since x · [1] = x · [19] for x ∈ G. This is done by noting that we need n linearly independent rows. 19. so it generates a cyclic subgroup that does not have the required form. which is the number of vectors in the subspace spanned by the ﬁrst two rows that we have chosen. 21. but to be linearly independent of the ﬁrst row. cn . and so HK = G. [13].) i 0 0 0 in the group GL3 (C).66 CHAPTER 3 SOLUTIONS Comment: The other 7 elements in the group will have at least one component equal to zero. This argument can be continued. [19]}. 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. Thus there are pn − p2 possibilities for the third row. Find the order of the element ([9]12 . There are pn possibilities for the second row. For the third row. the element (1. In the group G = Z× . [25]. The ﬁrst row can be any nonzero vector. 18. [5]. Solution: We need to count the number of ways in which an invertible matrix can be constructed. then the order of the general linear group GLn (Zp ) is (pn − 1)(pn − p) · · · (pn − pn−1 ). we need to omit the p multiples of the ﬁrst row. so there are pn − 1 choices. [17]. giving the stated result. Show that if p is a prime number. 6] = 12. [15]18 ) in the group Z12 × Z18 . [29]} and K = {[1]. for subgroups H1 ⊆ G1 and H2 ⊆ G2 . 1) has order 2. [15]18 ) is lcm[4. Solution: Since gcd(9. we need to subtract p2 . Adding the identity element to the list accounts for all 15 elements of Z5 × Z3 . Since we have p possible scalars. Solution: It can be shown (as in Problem 3.

and doing this calculation shows that Solution: Suppose that X = m m+b 0 1 = m b 0 1 1 1 0 1 = 1 0 1 1 m b 0 1 = m b+1 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). where Z(G) is the center of G. 2c = 2c. The ﬁrst equation implies that c = 0. a b belongs to c d 2a a + 2b the centralizer of A in GL2 (Z3 ). Equating corresponding entries shows that we must have b = 0. Equating corresponding entries shows that we must have m + b = b + 1. and 1 b b∈R . our calculations show that the center of G is the trivial subgroup. Let A = 1 1 −1 0 and B = . Equating c d 0 2 0 2 c d 2c 2d corresponding entries shows that we must have 2a = 2a + c. Then XA = AX. Compute the centralizer in GL2 (Z3 ) of the matrix Solution: Let A = 2 1 0 2 2 0 1 2 . and since any element in the center of G must belong to C(A) ∩ C(B). 22. Then we must 0 1 have XA = AX. while the .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. a + 2b = 2b + d. 0 1 This shows that C(A) ∩ C(B) is the identity matrix. and c + 2d = 2d. −1. any matrix of this form commutes with A. and suppose that X = . and so −1 0 0 1 m b 0 1 −m −b 0 1 = = = . On the other hand. 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 . m b belongs to C(A) in G. and so m = 1. Then XB = BX. and so C(B) = m 0 0=m∈R . and show that C(A) ∩ C(B) = Z(G). 23. containing only the identity element. and so = 2c c + 2d a b 2 1 2a + c 2b + d 2 1 a b = = . Thus o(A) = 4. and −i.

then BA = A−1 B. a b belongs to the c d 2a + b a + b centralizer of A in GL2 (Z3 ). while it follows from Problem 20 in this section that GL2 (Z3 ) has (32 − 1)(32 − 3) = 48 elements. and suppose that X = 2 1 1 1 . we will have a total of 10 elements. and so it is a subgroup since the group is ﬁnite. Let H be the following subset of the group G = GL2 (Z5 ). Equating corc d 1 1 1 1 c d a+c b+d responding entries shows that we must have 2a + b = 2a + c. 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 . and A−1 B = −1 −1 0 1 . a + b = 2b + d. and it is certainly 0 1 0 1 0 1 nonempty. b ∈ Z3 and a = 0 or b = 0 . −1 0 = 0 1 . 0 2 0 a Comment: The centralizer contains 6 elements.68 CHAPTER 3 SOLUTIONS second equation implies that a = d. b ∈ Z5 . It fol2 1 lows that the centralizer in GL2 (Z3 ) of the matrix is the subgroup 1 1 a b a. The ﬁrst equation implies that c = b. In this case the centralizer contains 8 of the 48 elements in 25. 2c + d = a + c. It follows that the centralizer in GL2 (Z3 ) 2 1 a b of the matrix is the subgroup a. 24. m b there are two choices for m and 5 0 1 choices for b. b ∈ Z3 and a = 0 . m = ±1 (a) Show that H is a subgroup of G with 10 elements. b a−b Comment: GL2 (Z3 ). The set is closed under mul±1 b ±1 c ±1 b ± c tiplication since = . Then XA = AX. Compute the centralizer in GL2 (Z3 ) of the matrix Solution: Let A = 2 1 1 1 . H= m b 0 1 ∈ GL2 (Z5 ) m. and so = 2c + d c + d a b 2 1 2 1 a b 2a + c 2b + d = = . and c + d = b + d. while the second equation implies that d = a − b.

erated by A consists of all matrices of the form 3. for all [n]16 ∈ Z16 . For any elements [n]16 and [m]16 in Z16 . If k ≡ m (mod 16). Deﬁne φ : Z16 → Z× by setting φ([1]16 ) = [3]17 . we have to show that the formula for φ does not depend on the particular representative that is chosen. 17 34 ≡ 3·10 ≡ 30 ≡ 13. we ﬁrst compute what happens if we combine [n]16 and [m]16 using the operation in Z16 . where 0 ≤ i < 5 and 0 ≤ j < 2. 17 Next. and then substitute the result into the function φ: φ([n]16 + [m]16 ) = φ([n + m]16 ) = [3]n+m .4 SOLUTIONS 21. Since φ is deﬁned by 17 using a representative n of the equivalence class [n]16 . The proof that φ respects the two group operations follows the proof in Example 3.4. then it follows from Proposition 3. The 17 17 general formula is φ([n]16 ) = [3]n . Show that Z× is isomorphic to Z16 .3 (a) implies that any isomorphism between cyclic groups must map a generator to a generator.8 (c) that [3]k = [3]m since [3]17 has order 16 in Z× Therefore φ([k]16 ) = φ([m]16). and Proposition 3. we ﬁrst apply the function φ to the two elements. it follows from Proposition 2. 17 Solution: The element [3] is a generator for Z× . and then combine the results using the operation in Z× : 17 φ([n]16 ) · φ([m]16 ) = [3]n [3]m = [3]n+m . 17 17 17 Thus φ([n]16 + [m]16 ) = φ([n]16 ) · φ([m]16 ). Solution: Since 1 0 b 1 1 c 0 1 = 1 b+c 0 1 . etc. Multiplying on the 0 1 right by B will create 5 additional elements. with generator [1]16 . Proposition 3. and so φ is a one-to-one function.5 that φ is also an onto function. [n]16 and [m]16 . and this completes the proof that φ is a group isomorphism. φ([2]16 ) = [3]2 . since 32 = 9. giving all of the elements in H.8 (c) shows that φ([k]16 ) = φ([m]16 ) only if k ≡ m (mod 16).CHAPTER 3 SOLUTIONS 69 (c) Show that every element of H can be written uniquely in the form Ai B j .1.2. 33 = 27 ≡ 10. 36 ≡ 3·5 ≡ 15. 35 ≡ 3·13 ≡ 39 ≡ 5. Then because both Z16 and Z× have 16 17 elements. . since Z16 is also a cyclic group.4. and 38 ≡ 3 · 11 ≡ 33 ≡ −1 ≡ 1. 37 ≡ 3·15 ≡ 45 ≡ 11. Therefore Z× is a cyclic group with 16 17 elements. and 17 17 17 so φ is a well-deﬁned function.1. the cyclic subgroup gen- 1 b . This provides the clue at to how to deﬁne the isomorphism we need.2.

Similarly. b2 )) and so φ : G1 × G2 → H1 × H2 is a group isomorphism. Given (a1 . for all x ∈ R. 24. [m]2 ) ∈ Z30 × Z2 . then 30 | n − k and 2 | m − j. and then prove that φ is a group isomorphism. a2 ) · (b1 . [j]2 ) are equal elements of Z30 × Z2 . Let G1 . generated by [3]7 . Deﬁne φ : Z30 × Z2 → Z10 × Z6 by φ([n]30 . θ2 (a2 )θ2 (b2 )) φ((a1 . generated by [2]11 . y2 ) = φ(x1 . b2 )) = φ((a1 b1 . G2 . 7 11 25. Solution: If (y1 . and then 6 | 3(m−j). θ2 (a2 )θ2 (b2 )) = (θ1 (a1 ). It then follows from 11 Problem 23 that φ : Z6 × Z10 → Z× × Z× deﬁned by φ(([n]6 . θ2 (b2 )) = (θ1 (a1 )θ1 (b1 ). Solution: If ([n]30 . a2 b2 )) = (θ1 (a1 b1 ). and 7 that Z× is cyclic of order 10. First prove that φ is a well-deﬁned function. It follows that 10 | n − k. 7 11 Solution: You can check that Z× is cyclic of order 6. Thus there is a unique element (x1 . θ2 (x2 )). 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. and suppose that θ1 : G1 → H1 and θ2 : G2 → H2 are group isomorphisms. . θ2 (a2 b2 )) = (θ1 (a1 )θ1 (b1 ). Prove that the group Z× × Z× is isomorphic to the group Z6 × Z10 . for all [n]6 ∈ Z6 and all [m]10 ∈ Z10 . [m]2 ) = ([n]10 . x2 ) ∈ G1 × G2 . [4n + 3m]6 ). 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. is a group isomorphism. Deﬁne φ : G1 × G2 → H1 × H2 by φ(x1 .70 CHAPTER 3 SOLUTIONS 22. H2 be groups. a2 )) · φ((b1 . Furthermore. H1 . Let φ : R× → R× be deﬁned by φ(x) = x3 . x2 ) ∈ G1 × G2 such that (y1 . [4k + 3j]6 ). we have φ((a1 . θ2 (a2 )) · (θ1 (b1 ). and so [n]10 = [k]10 . 30 | 4(n−k). and so φ is one-to-one and onto. showing that [4n + 3m]6 = [4k + 3j]6 . which shows that the formula for φ does yield a well-deﬁned function. Thus ([n]10 . b2 ) in G1 × G2 . [4n + 3m]6 ) = ([k]10 . for all ([n]30 . which together imply that 6 | (4n + 3m) − (4k + 3j). [2]m ). a2 ) and (b1 . so 6 | 4(n−k). b ∈ R× we have φ(ab) = (ab)3 = a3 b3 = φ(a)φ(b). x2 ) = (θ1 (x1 ). y2 ) ∈ H1 × H2 . since θ2 is an isomorphism there is a unique element x2 ∈ G2 with y2 = θ2 (x2 ). then since θ1 is an isomorphism there is a unique element x1 ∈ G1 with y1 = θ1 (x1 ). [m]10 )) = 7 11 ([3]n . Prove that φ is a group isomorphism. [m]2 ) and ([k]30 . for all (x1 . x2 ). Solution: The function φ preserves multiplication in R× since for all a. Just as in Problem 21. 23. Show that φ is a group isomorphism.

then the subset aHa−1 = {g ∈ G | g = aha−1 for some h ∈ H} is a subgroup of G that is isomorphic to H. is a group isomorphism. so we may conclude that φ is a group isomorphism. and hk = kh for all h ∈ H and k ∈ K. 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. it follows that 2 | m. [m]2 ) = ([0]10 . HK = G. it follows that φ must also be an onto function. It follows that 2 | (40q + 3m) and 3 | (40q + 3m). say n = 10q.4 to show that φ is one-to-one. Exercise 3.13 in the text. Then since 2 and 3 are prime numbers. so 10 | n.3. G2 be groups. for all x ∈ G. and 6 | (4n + 3m). the function φ : G → G deﬁned by φ(x) = axa−1 . respectively. [0]6 ). for some q ∈ Z. We have therefore checked all of the necessary conditions. [d]2 ) we have φ(([a]30 .4. By Exercise 3.4. so [m]2 = [0]2 . and hk = kh for all h ∈ H and k ∈ K. then ([n]10 . and 3 | 40q since 3 | 3m.4. Solution: Let φ : G1 × G2 → G be an isomorphism. [4a + 3c + 4b + 3d]2 ) = ([a + b]10 .4. [c]2 ) + ([b]30 . [c]2 )) + φ(([b]30 .15) that H and K are subgroups of G. Prove that if G is isomorphic to G1 × G2 . We conclude that φ is a one-toone function. [d]2 )) = φ(([a + b]30 . so we only need to show that H ∩ K = {e}. HK = G. [m]2 ) = ([0]10 . Let G be a group. or 6 | (40q + 3m). x2 ) | x1 = e} have the properties we are looking for. and so the condition in Proposition 3. We know (by Exercise 3. [0]2 ). [4b + 3d]2 ) = ([a + b]10 . Let H = φ(H ∗ ) and K = φ(K ∗ ) be the images in G of H ∗ and K ∗ .4. [0]6 ).4 is satisﬁed. and 3 | q.CHAPTER 3 SOLUTIONS For any elements ([a]30 . 26. [m]2 ) = ([0]30 . then ([n]30 .15 the image under φ of any subgroup of G is again a subgroup of G. 27. This means that we can use Proposition 3. x2 ) | x2 = e} and K ∗ = {(x1 . [0]6 ). Let G. [4a + 4b + 3c + 3d]2 ) and so φ respects the operations in the two groups. then there are subgroups H and K in G such that H ∩ K = {e}. and let H be a subgroup of G.9 in the text shows that in G1 × G2 the subgroups H ∗ = {(x1 . [4a + 3c]2 ) + ([b]10 . [c + d]2 )) = ([a + b]10 . [4n + 3m]6 ) = ([0]10 . G1 . If φ([n]30 . Since the two groups both have 60 elements. [4(a + b) + 3(c + d)]2 ) = ([a + b]10 . . Prove that if a is any element of G. so [n]30 = [10q]30 = [0]30 . [c]2 ) and ([b]30 . [d]2 )) = ([a]10 . We have now shown that if φ([n]30 . [4a + 4b + 3c + 3d]2 ) 71 φ(([a]30 . and therefore 2 | 3m since 2 | 40q. Solution: By Exercise 3.

showing that H ∩ K = {e}. (b) Show that H is isomorphic to the group R of all real numbers. x2 ) ∈ Z× ×Z× . a21 = 0. (a) In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. it is clear x1 0 that the mapping φ : Z× ×Z× → GL2 (Zp ) deﬁned by φ(x1 . e2 ). Thus φ is the desired isomorphism. b2 )) = φ((a1 b1 . Since y ∈ K as well. and so x ∈ H ∗ . It is clear that φ preserves the fact that elements of h∗ and K ∗ commute. 29. and thus G = HK. for x ∈ G1 × G2 . a2 b2 )) a1 b1 0 = 0 a2 b2 = a1 0 0 b1 a2 0 0 b2 = φ((a1 . so x ∈ H ∗ ∩ K ∗ . p p 0 x2 for each (x1 . b2 ) ∈ Z× × Z× we have p p φ((a1 . This mapping respects the operations in the two groups. a2 ). with y = φ(x). since for (a1 . Thus y = φ((e1 . 1 x Solution: Deﬁne φ : R → H by φ(x) = . 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 . is one-to-one and maps Z× ×Z× onto the subgroup p p p p of diagonal matrices. You can 0 1 easily check that φ is an isomorphism. under addition.) . If y ∈ H ∩ K. and therefore x = (e1 .72 CHAPTER 3 SOLUTIONS Let y ∈ G. show that H= a11 a21 a12 a22 ∈ GL2 (R) a11 = 1. a2 ))φ((b1 . Since y is any element of G. where e1 and e2 are the respective identity elements in G1 and G2 . and we can write x = h∗ k ∗ for some h∗ ∈ H ∗ and some k ∗ ∈ K ∗ . Identity: The identity matrix has the correct form. then y ∈ H. We conclude that H and K satisfy the desired conditions. 28. the subgroup of diagonal matrices in GL2 (Zp ) is isomorphic to Z× × Z× . e2 )) = e. a22 = 1 is a subgroup of G. (The computation necessary to show that φ preserves the respective operations is the same computation we used to show that H is closed. Existence of inverses: = 1 −a 0 1 ∈ H. a2 )(b1 . Show that for any prime number p. (b1 . we must also have x ∈ K ∗ . b2 )) . it follows that y = φ(h∗ k ∗ ) = φ(h∗ )φ(k ∗ ). for all x ∈ R. x2 ) = .

if one of the groups is abelian but the other is not. Solution: (Outline only) The operation is well-deﬁned on S. and that φ is actually a group isomorphism. for all x1 . On −1 0 1 1 −1 −1 the other hand. G is not abelian. for all 0 ≤ i < 3 and 0 ≤ j < 2. m = 0 Show that H is isomorphic to the symmetric group S3 . Let A = and B = . and let S be any set for which there exists a one-toone and onto function φ : G → S.4. gives a one-to-one correspondence between the elements of the groups which also produces multiplication tables that look exactly the same. x2 ∈ S.25. as given in Table 3. since φ and φ−1 are functions and the operation on G is well-deﬁned. Remember that constructing an isomorphism is the same as constructing a one-to-one correspondence between the elements of the group. Let G be a group. In this case we can explain how this can be done.3. By Proposition 3. Thus the two groups cannot be iso0 1 0 1 0 1 morphic. Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . 32. 73 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.3. without actually writing out 1 1 −1 0 the multiplication table.3 (on page 104 of the text). since = 0 1 0 1 0 1 1 1 −1 0 −1 1 but = . we can show that BA = A−1 B. Then just 0 1 0 1 as in Problem 3. Prove that S is a group under this operation. Deﬁne an operation on S by setting x1 · x2 = φ(φ−1 (x1 )φ−1 (x2 )). Solution: This group is small enough that we can just compare its multiplication table to that of S3 .CHAPTER 3 SOLUTIONS 30. 31. The associative law holds . since each factor is abelian. then the groups cannot be isomorphic.3 (b). The direct product R× ×R is an abelian group. and that each element of H has the form can be written uniquely in the form Ai B j . Let H be the following subgroup of group G = GL2 (Z3 ). H= m b 0 1 ∈ GL2 (Z3 ) m. such that all entries in the respective group tables also have the same one-to-one correspondence. This information should make it plausible that the function φ : S3 → H deﬁned by φ(ai bj ) = Ai B j . b ∈ Z3 . where 0 ≤ i < 3 and 0 ≤ j < 2.

Our theorems tell us that 18 Z× ∼ Z6 .4.12 in the text. Z4 × Z15 ∼ Z4 × Z3 × Z5 . Is Z4 × Z15 isomorphic to Z6 × Z10 ? Solution: As in Problem 21. . Find all generators of the cyclic group Z28 . and so the subgroup has 30/6 = 5 elements. The list of generators 2 7 is {±1. and it is easy to check that if x ∈ S. Give the lattice diagram of subgroups of Z100 .5. where e is the identity of G.23 that Z4 × Z10 ∼ Z4 × Z2 × Z5 . = 22.11 and 3. Show that the three groups Z6 . 24. 23 = 8 ≡ 1. Therefore = Z4 × Z10 ∼ Z2 × Z20 . 53 ≡ 17 ≡ 1. and so [5] must 9 = 18 have order 6. and |Z× | = 6. and Z× are isomorphic to each other.5 SOLUTIONS 20. we can use one-toone functions deﬁned on G to produce new groups with operations that look rather diﬀerent from the usual examples. Note that nZ100 is used to mean all multiples of n in Z100 . Similarly. The Euler ϕ-function allows us to compute how many there are: ϕ(28) = 1 · 6 · 28 = 12. ±13}. while the second has none.0. the identity element in S is φ(e). Our 9 theorems tell us that Z× ∼ Z6 . Given a known group G such as R× . Solution: By Proposition 3.74 CHAPTER 3 SOLUTIONS in S because it holds in G. we ﬁrst ﬁnd gcd(18. In Z30 . Z× .4. In Z× . Is Z4 × Z10 isomorphic to Z2 × Z20 ? Solution: It follows from Theorem 3. [24]30 = [6]30 . and that Z20 ∼ = = Z4 × Z5 . and Z6 × Z10 ∼ = = Z2 × Z3 × Z2 × Z5 . and so we actually have [24]30 = [18]30 . The two groups are not isomorphic since the ﬁrst has an element of order 4. ±9. = and Z2 × Z20 ∼ Z2 ∼ Z4 × Z5 . 18 = 21. ﬁnd the order of the subgroup [18]30 .4 that Z10 ∼ Z2 × Z5 . showing that Z× is cyclic of order 6.3 (a). ﬁnd the order of [24]30 .1. 9 18 Solution: First. and are given in Figure 3. 30) = 6. Solution: Using Proposition 3.5. Then [18]30 = [6]30 . 22 = 4. 52 ≡ 7. ±11. Finally. 9 18 9 and so [2] must have order 6.5. 25. Solution: The subgroups correspond to the divisors of 100. it is possible to show that the obvious = = mapping from Z4 × Z2 × Z5 onto Z2 ∼ Z4 × Z5 is an isomorphism. Comment: This reveals the secret behind problems like Exercises 3. 3. It then follows from Problem 3. ±5. ±3. In Z× .1.3 (b). showing that Z× is cyclic of order 6. Thus all three groups are isomorphic. we have |Z× | = 6. then x−1 = φ((φ−1 (x))−1 ). 23. the generators correspond to u the numbers less than 28 and relatively prime to 28.

then the direct product G1 × G2 is a cyclic group. 27.CHAPTER 3 SOLUTIONS 75 Figure 3. Since isomorphisms preserve orders of elements. 15 In Z× . 21 21 29. Use the the result in Problem 27 to show that the multiplicative groups Z× 15 and Z× are not cyclic groups.5. then it follows from Theorem 3. and since the congruence can be rewritten as x ≡ 0 (mod n). we only need to answer the question in Z2n . Use this information to show that the quaternion group cannot be isomorphic to the subgroup of S4 generated by (1. If a has order 7 in G1 and b has order 11 in G2 .2 that G is isomorphic to Z2n . we see that [n]2n is the only element of order 2 in Z2n . Prove that if G1 and G2 are groups of order 7 and 11. and so [8]21 and [−1]21 have order 2. In that group. Thus G1 × G2 is cyclic since it has an element whose order is equal to the order of the group. Show that any cyclic group of even order has exactly one element of order 2. .1: for Problem 23 Z100 2Z100 5Z100 4Z100 10Z100 25Z100 20Z100 50Z100 0 26. 4) and (1. Solution: Since 7 and 11 are primes. 3. then (a. both [−1]15 and [4]15 are easily checked to have order 2. for some positive integer n. respectively. we have [8]2 = [64]21 = [1]21 . Find all cyclic subgroups of the quaternion group. 21 Solution: In Z× . 3). 28. the groups are cyclic. 11] = 77 in G1 × G2 . the elements of order 2 are the nonzero solutions to the congruence 2x ≡ 0 (mod 2n). b) has order lcm[7. Solution: If G is cyclic of order 2n. 2.

4)(2. 3). 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 ). 3. a2 b. 2)(1. and i4 = i2 i2 = (−1)2 = 1. . 3) = (1. 2. 4) = a3 b. which has 3 cyclic subgroups of order 4. ±k} is deﬁned in Example 3. This implies that the result must hold for all i with 0 ≤ i < n. with a2 = (1. 4) = (1. In any isomorphism. 3. The Chinese remainder theorem (Theorem 1. o(b) = 2. 4) = a2 b. 4) = a and (1. 3)(1. ki = j. ±i = {±1. 2.6 SOLUTIONS 22. On the other side. and ±k = {±1. let (1. 4)(2. 2) = (1. 4. we have ab = (1. 4) and a3 = a−1 = (1. 2)(3. 2). The elements satisfy the following identities: i2 = j2 = k2 = −1 and ij = k. ±j = {±1. 2)(3. 4. 4)(1. show that bai = an−i b. then Z× is not a cyclic group. To ﬁnd the subgroup generated by a and b. and a3 b = (1. 3)(1. 4) = (2. 3) = (1. a. 3) = b. 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 . For example. jk = i. 3. for all 0 ≤ i < n. This shows that the subgroup generated by a and b consists of the 8 elements {e. so [a]pq = [−1]pq . 4). a2 b = (1. Comment: This is similar to a proof by induction. ±k} can be found by using the given identities. [−1]pq is not a solution. If we assume that the result holds for i = k. kj = −i. ±j}. 3. In the dihedral group Dn = {ai bj | 0 ≤ i < n. Solution: For i = 1. Since a is a cycle of length 4. a3 . 3. and thus [a]pq has order 2. and ba = a−1 b. ik = −j. we have ba = (1. and ba3 = (1.76 CHAPTER 3 SOLUTIONS Solution: The quaternion group Q = {±1. 3)(1. 4). 3)(2. but for each given n we only need to worry about a ﬁnite number of equations. a2 . ba2 = (1. 3)(2. ji = −k. ab. 3. 3)(2. since p and q are relatively prime. a3 b}. pq Solution: We know that [−1]pq has order 2. In S4 . i3 = i2 i = −i.3. so a2 ≡ 1 (mod pq) since p and q are relatively prime. i2 = −1. 2. Because q is an odd prime. ±j. the equation bai = an−i b is just the relation that deﬁnes the group. so by Problem 27 it is enough to ﬁnd one other element of order 2. ±i}. it has order 4.3. b. Prove that if p and q are diﬀerent odd primes. 3) = ab. and so it is impossible for this group to be isomorphic to the quaternion group.6) states that the system of congruences x ≡ 1 (mod p) and x ≡ −1 (mod q) has a solution [a]pq . cyclic subgroups would correspond to cyclic subgroups. 3.7 of the text (see page 108). The cyclic subgroups −1 = {±1}. ±i. 0 ≤ j < 2} with o(a) = n. 4. 3) = (2. But a2 ≡ 1 (mod p) and a2 ≡ 1 (mod q). 4)(1. Furthermore. 30.

In the dihedral group Dn = {ai bj | 0 ≤ i < n. 3) in S3 . 2). 4. (2. 2). 3). 4). ab. show that each element of the form ai b has order 2. 4)−1 (1.10. 4)−1 = (1. Rather than starting with each of the other elements and then trying to write them as a product of 3-cycles. 26. 3. 3. (1. 4)(1)(1. (1. 2. 3)(2. (1. a. we have (1. 2). 4). Then we can quickly do the other computations: (1. o(b) = 2. Solution: Let a = (1. ab = (1. since by deﬁnition b does not belong to C(a). 2. 3. 2. 2.) Thus H = {(1). ﬁnd the corresponding subgroup σHσ −1 . 4) = (1). 3. and (1. On the other hand. 4). Show that each element in A4 can be written as a product of 3-cycles. 2. 4. which shows that σ(1. 2)(1. 4)}. 3. Since (1. 3)(1. 3). 2. 2. 2. Solution: We need to compute στ σ −1 . 4)(1. we have (ai b)2 = (ai b)(ai b) = ai (bai )b = ai (an−i b)b = (ai an−i )(b2 ) = an e = e. (1. 4)(1. 3. 3) = (2. C(a) = Dn . σ(2). 4) = (1. 3)(2. (1. and in A4 . and (2. (2. 2. 2)(1. 3. 2) = (4. 28. 4)H(1. o(b) = 2. 4)(1. 2) = (4. 2). 3). Solution: Using the result from the previous problem. a2 . 3). As a shortcut. 3). 3)(1. and ba = a−1 b. (1. a2 b} = {(1). (3. Since a contains exactly half of the elements in Dn . Find the centralizer of (1. 3)(1. 3)}. 4)−1 (1. (We have just listed the elements of S3 . 2. so we have a ⊆ C(a). (2. 3) and a2 b = (2. (1. 3). 2). for σ = (1. 4)(1. (1. 3) and (1. 4. and this accounts for all 12 of the elements in A4 . (1. 4).3. 3)(1. 2)(3. (2.CHAPTER 3 SOLUTIONS 77 23. 3). 3). 24. 2. 2). (1. Thus (1. 4). 0 ≤ j < 2} with o(a) = n. 2. 4). ﬁnd the centralizer of a. 4) = (2. 4). . 3. 4). 27. 3). (1. Then H must contain a2 = (1. (2. 4)(2. and this set of elements is closed under multiplication. so a ⊆ C(a) ⊆ Dn and C(a) = Dn together imply that C(a) = a . 4)−1 (1. and ba = a−1 b. 3). 0 ≤ j < 2} with o(a) = n. (2. 2. σ(3)) = (4. For the subgroup H of S4 deﬁned in the previous problem. 3. 2). 4)−1 = {(1). ﬁnd the subgroup H generated by (1. This shows that C(a) has at least n elements. Solution: We ﬁrst list the 3-cycles: (1. 4)−1 = (4. in S4 . 3) and b = (1. 4. 3). We have (1. 4)(1. Lagrange’s theorem show that there is no subgroup that lies strictly between a and Dn . for each τ ∈ H. it is easier to just look at the possible products of 3-cycles. 4)(2. b. 4). 4) = (1. we can use Exercise 2. 3) = (1. 3)σ −1 = (σ(1). 3. In S4 . 25. 2. Solution: The centralizer C(a) contains all powers of a. 4). In the dihedral group Dn = {ai bj | 0 ≤ i < n.

3. k) is a cycle of length k and σ is any permutation. 4)a−1 = (2. and therefore C(a) = a . 4).10 provides a quick way to do this in a group of permutations. . 2. 4)a−1 = (u.7. 2)}. Let a = (1. Finally.3. and by Lagrange’s theorem a proper subgroup of a group with 6 elements can have at most 3 elements. (a) Find the formulas for all group homomorphisms from Z18 into Z30 . Solution: Example 3. 4). 1)(3. y. and a(1. which leads to the formula φ([x]4 ) = [5x]10 . or else φ([1]4 ) = [5]10 .78 CHAPTER 3 SOLUTIONS Solution: Since any power of an element a commutes with a. 3). for all [x]4 ∈ Z4 .7. w. 3)(2. σ(2). so (u. and so no product of transpositions belongs to C(a). 1). 2. Find all group homomorphisms from Z4 into Z10 . z. the centralizer of (1. the order of φ([1]4 ) must be a divisor of 4 and of 10. Since . w. . This means that w = z. 4)(2. 3) . To ﬁnd the centralizer of (1. Thus the centralizer of (1. (1. . 4). y. 3) do not commute with a. 4)a−1 = (3. Thus φ([1]4 ) = [0]10 . A similar argument shows that no 3-cycle that includes the number 4 as one of its entries can belong to C(a). σ(k)). and z. 2. 2). This accounts for an additional 6 elements. 3) in S4 we have to work a bit harder. 2) does not belong to the centralizer. The remaining transpositions in S4 are (1. . 2. in A4 we must get the same answer: C(a) = a . and a(3. 4)a−1 = (1. since it is easy to check that (1. For the products of the transposition. Since there are 6 elements of this form. Without doing all of the calculations. 3) always contains the subgroup {(1). 3.5 shows that any group homomorphism from Z18 into Z30 must have the form φ([x]18 ) = [mx]30 . 4)a−1 = (2.10. . .7 SOLUTIONS 17. 2. a(2. . and (2. and (3. so the only possibilities are 1 and 2. . 4) = (x. 4). 4). 3)(1. 2. the centralizer C(a) always contains the cyclic subgroup a generated by a. To see that x belongs to C(a). 2. 3) is equal to (1. or that axa−1 = x. If we do a similar computation with a 4-cycle. so no transposition in S4 commutes with a. we can conclude that no 4-cycle belongs to C(a). (2. 4). . It helps to have some shortcuts when doing the necessary computations. 18. we will have a(x. 3). we now have a total of 21 elements that are not in C(a). we know that (1. for all [x]18 ∈ Z18 . we need to check that xa = ax. 3)a−1 = (2. 3. 4)a−1 = (2.5 shows that any group homomorphism from Zn into Zk must have the form φ([x]n ) = [mx]k . . 4). then σ(1. y. 2. . Solution: Example 3. From the computations in S3 . since a just permutes the numbers x. 3). z. v. In S3 . 4)(3. (1. (1. Exercise 2. That exercise shows that if (1. a(1. Using Exercise 2. we have a(1. 2)(3. 4). k)σ −1 = (σ(1). . 4). Under any group homomorphism φ : Z4 → Z10 . which deﬁnes the zero function. v. for all [x]n ∈ Zn . we have a(1. 4).

16} ←→ φ(4) = 20. These cosets are in one-to-one correspondence with the elements of the image. 13} ←→ φ(1) = 5. 4. 10. then y ≡ 0 (mod 3). 11. [15]30 . so y = 0 or y = 3. 6. y). and then it follows quickly that φ is a well-deﬁned function. and apply the fundamental homomorphism theorem. 2. of order 1. {2. The image of φ consists of the multiples of 5 in Z30 . and the order of φ([3]17 ) must be a common divisor of 16 and 6. Solution: If y1 ≡ y2 (mod 6). {5. (a) Show that φ is a well-deﬁned group homomorphism. 17 7 Solution: Any group homomorphism φ : Z× → Z× is determined by its value 17 7 on the generator [3]17 . We have the following correspondence {0. for all [x]17 = ([3]17 )n ∈ Z× . 3). 7 17 20. The corresponding choices for [m]30 are [0]30 . 0). y) = (x + 2y. {1. [10]30 and [20]30 . Solution: If (x. 6. 38 ≡ 3 · 11 ≡ 16 ≡ 1. 15. consider φ([x]18 ) = [5x]30 . and if y = 3. 7. 0) and (2. respectively. 15} ←→ φ(3) = 15. of order 6. 12}. {4. 17} ←→ φ(5) = 25. so 2y1 ≡ 2y2 (mod 4). 17 Solution: The element [3] is a generator for Z× . 9. (b) Show that Z× is cyclic. 19. It is also easy to check that φ preserves addition. 37 ≡ 3 · 15 ≡ 11. 20. 3. 6. We have ker(φ) = {0. then 2y1 − 2y2 is divisible by 12. which are 0. 7 Solution: Since 32 ≡ 2 and 33 ≡ 6. 30) = 6. Thus the elements of the kernel K are (0. (b) Find the kernel and image of φ. 36 ≡ 3 · 5 ≡ 15. 2. (a) Show that Z× is cyclic. 25. it follows that [3] must have order 6. 10. and 5. 35 ≡ 3 · 13 ≡ 5. 5. since 32 = 9. y) belongs to ker(φ). and [5]30 and [25]30 . It follows that there are 24/2 = 12 cosets of the kernel. and show how elements of the image correspond to cosets of the kernel. Solution: For example. and in 17 the second case φ(([3]17 )n ) = [−1]n . so φ must map Z4 × Z6 onto Z4 × Z3 . with generator [3]7 . Deﬁne φ : Z4 × Z6 → Z4 × Z3 by φ(x. 8. φ([x]17 ) = [1]7 for all [x]17 ∈ Z× . 3)} ∼ Z4 × Z3 . (b) Choose one of the nonzero formulas in part (a). = . In the ﬁrst case. The only possible orders are 1 and 2. so either φ([3]17 ) = [1]7 or φ([3]17 ) = [−1]7 . 17 34 ≡ 3 · 10 ≡ 13. and for this formula ﬁnd the kernel and image. (2. 12} ←→ φ(0) = 0. 33 = 27 ≡ 10. then x = 2. of order 3. with generator [3]17 . {3. If y = 0. (c) Completely determine all group homomorphisms from Z× into Z× . of order 2. 14} ←→ φ(2) = 10. then x = 0. and then cosets of the kernel are deﬁned by adding 1. 3.CHAPTER 3 SOLUTIONS 79 gcd(18. the possible orders of [m]30 = φ([1]18 ) are 1. Thus (Z4 × Z6 )/{(0.

Next. φ2 (x)). Thus [x1 ]m = [x2 ]m .80 CHAPTER 3 SOLUTIONS 21. Solution: First. 36) = 1. namely [1]36 . . pq p q so φ is also an onto function. φ is a homomorphism since for [a]n . φ is a well-deﬁned function. [25]36 . we can deﬁne group homomorphisms φ1 : Z× → pq Z× and φ2 : Z× → Z× by setting φ1 ([x]pq ) = [x]p . pq Using Problem 23. so p | (x − 1) and q | (x − 1). It follows that [x]pq = [1]pq . φ2 (a)φ2 (b)) and so φ : G → G1 × G2 is a group homomorphism. Let p and q be diﬀerent odd primes. b in G. Exercise 1. then ϕ(mn) = ϕ(m)ϕ(n). φ2 ([x]pq )). and apply the fundamental 36 homomorphism theorem. p q Solution: Using Problem 21. 36 12 The kernel of φ consists of the elements in Z× that are congruent to 1 mod 36 12. This completes the proof that φ is a group isomorphism. 12 = 36 23. and this shows that φ is a one-to-one function. and G2 be groups. for all x ∈ G. n then n | (x1 − x2 ). Solution: Given a. Prove that Z× is isomorphic to the direct pq product Z× × Z× . and this implies that pq | (x − 1). we have φ(ab) = (φ1 (ab). for all [x]pq ∈ Z× . φ2 (a)φ2 (b)) φ(a)φ(b) = (φ1 (a). since m | n. for all [x]pq ∈ Z× . 24. then [x]p = [1]p and [x]q = [1]q . ﬁnd the kernel and image of φ. Let φ1 : G → G1 and φ2 : G → G2 be group homomorphisms. 22. Let n and m be positive integers. Prove that φ : G → G1 × G2 deﬁned by φ(x) = (φ1 (x). It is evident that φ maps Z× onto Z× . Let G. Show that φ : Z× → Z× deﬁned by φ([x]n ) = [x]m . for 36 12 all [x]36 ∈ Z× . is a well-deﬁned group homomorphism. It follows that Z× and Z× × Z× have the same order. φ2 (a)) · (φ1 (b).4. φ2 (ab)) = (φ1 (a)φ1 (b). since if gcd(x. 12) = 1. and so φ([x1 ]n ) = φ([x2 ]n ). such that m is a divisor of n. and pq q p pq φ2 ([x]pq ) = [x]q . Solution: The previous problem shows that φ is a group homomorphism. and this implies that m | (x1 − x2 ). for all [x]pq ∈ Z× . [b]n ∈ Z× . then gcd(x. since p adn q are relatively prime. For the group homomorphism φ : Z× → Z× deﬁned by φ([x]36 ) = [x]12 . φ2 (b)) = (φ1 (a)φ1 (b). for all [x]n ∈ Z× . is a well-deﬁned n m n group homomorphism. If [x]pq ∈ pq ker(φ).27 in the text states that if m > 0 and n > 0 are relatively prime integers. since if [x1 ]n = [x2 ]n in Z× . G1 . we can deﬁne a group homomorphism φ : Z× → Z× × Z× q pq p by setting φ([x]pq ) = (φ1 ([x]pq ). φ([a]n [b]n ) = φ([ab]n ) = n [ab]m = [a]m [b]m = φ([a]n )φ([b]n ). It follows that Z× ∼ Z× / [13]36 . [13]36 .

2)} (1. 1). with generators a of order n and b of order 2. 3)} (3. 0) + K = {(2. 5} 9 7 = {9. and so G/H ∼ Z6 × Z2 . 13. Let the dihedral group Dn be given via generators and relations. 3)} (5. (0. 13} Since 3 ∈ 7 . (3. 1) + H = {(1. 1). 1). list all cosets of K. 1) + K = {(2. 11. 1) + H = {(2. = ∼ Z12 . (5. (4. showing that the factor group is cyclic. (5. 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. 0). 0) + H = {(2. 1) + K = {(0. producing an element in H. 1)} (1. 0). and so G/K = 29. 15} 11 7 = {11. 0) + H = {(3. (5.CHAPTER 3 SOLUTIONS 81 3. (1. 0) + K = {(0. 1) + H = {(0. 5. 3). 0)}. (3. 1). 0)} (0. 2)} (1. On the other hand. 0) + H = {(4. 3) + K = {(0. the coset 3 7 does not have order 2. 2) + K = {(1. (1. 1)} (0. 9. 1). (4. 3)} (0. 7} 2 3 7 = {3. 3). (4. (a) List all cosets of H. 1). 3). 1) + H = {(3. 15}. List the cosets of 7 in Z× . 2). (0. 1) + K has order 12 in G/K. 2)}. 1). let H = {(0. . 2)} (5. 0). 0). 0) + H = {(5. 2). we have bak+1 = bak a = a−k ba = a−k a−1 b = a−(k+1) b. 1). (3. (4. 1) + H = {(5. 2)} (2. (4. 3) + K = {(2. 2) + K = {(2. and let K = {(0. 2) + K = {(0. (3. (5. 0) + K = {(1. 2)} (2.8 SOLUTIONS 27. 2)} are (0. 1). 3)} (2. (a) Show that bai = a−i b for all i with 1 ≤ i < n. 0)} (1. 3)} (4. Solution: The cosets of H = {(0. 0)} are (b) You may assume that any abelian group of order 12 is isomorphic to either Z12 or Z6 × Z2 . (2. and can be proved inductively: assuming bak = a−k b. (2. 1) + K = {(1. 28. 0). (4. 0). (5. (3. 1)} (2. 2)} (0. 0). 2). so it must have order 4. 16 7 = {1. (3. 2)} (3. 1) + H = {(4. 3)} The cosets of K = {(0. (1. 0)} (2. 0). 0). Solution: The identity holds for all positive integers i. Thus the order of an element in G/H is at most 6. 3. 0). (5. 0). 2)} (1. (0. (3. 0) + H = {(1. 0). 0) + H = {(0. (3. 7. 3)} (0. Is the factor group Z× / 7 cyclic? 16 16 Solution: Z× = {1. satisfying ba = a−1 b. 0). Let G = Z6 × Z4 . 3)} (1. 2)} (4. 3) + K = {(1. 3)} (2.

a6 b. Thus (aN )(bN ) = (bN )(aN ). . (c) List all left cosets and all right cosets of b Solution: The left cosets of b have the form ai b = {ai . (We are using the fact that bai = a−i b. It is normal since ai (a3n )a−i = a3n and ai b(a3n )ai b = ai a−3n a−i = (a3n )−1 . a4 . a6 . for 0 ≤ i < n. Thus G/N ∼ S3 . a3 } of G. and so a = g −1 bg = g −1 b(g −1 )−1 . since N aN b = N ab but N bN a = N a2 b.82 CHAPTER 3 SOLUTIONS (b) Show that any element of the form ai b has order 2. and G/N is not abelian. N ab = {ab. a8 . a9 b}. which shows that c ∼ a. (a) Let G be a group. Solution: The argument is the same as in the previous problem. Thus ∼ is an equivalence relation. = 32. and let N = {e. because ba = a11 b ∈ N a2 b. (a) Show that N is a normal subgroup of G. a3 b. b ∈ G we say that b is conjugate to a. a10 b}. the b = gag −1 for some g ∈ G. Show that ∼ is an equivalence relation on G. a5 b. a4 b}. If b ∼ a. a7 b. Which is correct? Solution: The factor group G/N is not abelian. N b = {ab. (a) Prove that N is a normal subgroup of G. then c = gbg −1 and b = hah−1 for some g. a3 . N a2 = {a2 . (b) Is G/N abelian? Solution: For aN = {a. If c ∼ b and b ∼ a. it is a subgroup. a9 }. a6 . For a. a7 . a9 }. for 0 ≤ i < n. 30. Solution: We have a ∼ a since we can use g = e. a11 b}. while (bN )(aN ) = baN = a5 bN = {a5 b. a4 } and bN = {b. The right cosets of b have the form b ai = {ai . a2 b}. N a2 b = {a2 b. The equivalence classes of ∼ are called the conjugacy classes of G. if there exists g ∈ G such that b = gag −1 . we have (aN )(bN ) = abN = {ab. Let G be the dihedral group D12 . h ∈ G. ai b}. which shows that a ∼ b. a11 }. Let G = D6 and let N be the subgroup a3 = {e. N a = {a. a8 b. written b ∼ a. and list all cosets of N . Solution: We have (ai b)2 = ai bai b = ai a−i b2 = a0 = e. Solution: Since N = a3 . a3 b}. a5 . a3 . 31.) The cosets of N are N = {e. a−i b}. so c = g(hah−1 )g −1 = (gh)a(gh)−1 . a10 }. (b) Show that a subgroup N of G is normal in G if and only if N is a union of conjugacy classes. a4 b. (b) You may assume that G/N is isomorphic to either Z6 or S3 .

Finally. so HN is nonempty. and let N and H be subgroups of G such that N is normal in G. 33.) Then φ(xy) = xyN = xN yN = φ(x)φ(y) for all x. Then x = h1 n1 and y = h2 n2 . If x = ai . But then hnN = hN = φ(h). Let D4 = {e. The solution of an earlier problem shows that xa2 x−1 = a2 in D4 . the only element conjugate to e is e itself. and so this shows that φ is onto. Thus a2 b is the only conjugate of b. Another way to say this is that N is a union of conjugacy classes. Thus N is normal if and only if whenever it contains an element a it also contains the conjugacy class of a. a2 . 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 . ab. a3 b}. Solution: Deﬁne φ : H → HN/N by φ(x) = xN for all x ∈ H. for all g ∈ G. which contains it. then xabx−1 = (ai b)ab(ai b)−1 = ai a−1 ai b = a2i−1 b. and so h = e. for some h1 . and so h ∈ N . H ∩ N = {e}. If x = ai b. It is clear that e = e · e belongs to the set HN . y belong to HN .3. If x is any power of a. then xax−1 = ai baa−i b = ai ai−1 b2 = a2i−1 . so this shows that a3 is the only conjugate of a (other than a itself). b. h2 ∈ H and some n1 . then x(ab)x−1 = ai aba−i = ai+1 ai b = a2i+1 b. φ is one-to-one since if h ∈ H belongs to the kernel of φ. We have xy −1 = h1 n1 (h2 n2 )−1 = h1 n1 n−1 h−1 = (h1 h−1 )(h2 (n1 n−1 )h−1 ). If x = ai . Suppose that x. Solution: Since N is normal in G. y ∈ H. Solution: See Proposition 3. Any coset of N in HN has the form hnN for some h ∈ H and some n ∈ N . If x = ai b. Let G be a group. Solution: Remember: the notion of a conjugacy class was just deﬁned in the previous exercise. a. so a2 is not conjugate to any other element. 2 2 (b) Prove that N is a normal subgroup of HN . By assumption. Thus a3 b is the only conjugate of ab. then xbx−1 = ai ba−i = ai ai b = a2i b. a3 . (Deﬁning a function from HN/N into H is more complicated. b2 = e. Since xex−1 = e. then HN/N is isomorphic to H. If x = ai b.2. n2 ∈ N . 34.CHAPTER 3 SOLUTIONS 83 Solution: The subgroup N is normal in G if and only if a ∈ N implies gag −1 ∈ G. and so xax−1 = a. then hN = φ(h) = N . with a4 = e. Find the conjugacy classes of D4 . (a) Prove that HN is a subgroup of G. then x commutes with a. it is normal in the subgroup HN . a2 b. . and ba = a−1 b. (c) Prove that if H ∩ N = {e}. then xbx−1 = (ai b)b(ai b)−1 = ai ai b = a2i b.

and give the lattice diagram which shows the 11 inclusions between them. [10]} The lattice diagram forms a diamond. Solution: First check for cyclic subgroups. −1 1 a b 1 −a 0 1 0 1 = 0 0 0 1 0 0 verses. 2. We could 13 also write this as Z× = [2]13 . b). This shows that Z× is cyclic. so the order of [2] is greater than 6. 29 = 6. 28 = 3. b ∈ R . and 12. [2]5 } = {[1]. and 26 ≡ 2 · 25 ≡ 12. and we have 22 = 4. [2]6 . and the order of any element must be 13 a divisor of 12. 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. so 1 0 0 1 0 0 1 The identity matrix belongs to the set. 23 = 8. 6. [3]} and [2]5 = {[1]. 24 = 16 ≡ 3. 23 = 8. and thus [2] is a generator for Z× . 25 ≡ 2 · 24 ≡ 6. in addition to Z× 11 11 and {[1]}: [2]2 = {[1]. 13 2. 4. 3. [2]4 . 25 = 10. Find all subgroups of Z× .84 CHAPTER 3 SOLUTIONS SOLUTIONS TO THE REVIEW PROBLEMS 1. 210 = 1. [2]8 } = {[1]. 1 a b Solution: Deﬁne φ : G → R × R by φ 0 1 0 = (a. [2]2 . 3. Solution: The group Z× has order 12. 27 = 7. 4. This is 0 0 1 one-to-one and onto because it has an inverse function θ : R × R → G deﬁned . [4]. 26 = 9. [9]. b 1 c d 1 a+c b+d 0 0 1 0 = 0 1 0 . (b) Show that Z× is a cyclic group. [5]. Show that the group G in the previous problem is isomorphic to the direct product R × R. so the possible orders are 1. 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 Solution: The ﬁrst element to try is [2]. so the subgroups are as follows. and −b 0 . (a) What are the possibilities for the order of an element of Z× ? Explain 13 your answer. By part (a) it must be 12. 24 = 5. 0 1 GL3 (R). in shorthand notation: 22 = 4.

it is always a normal subgroup. ±3. 0 0 1 0 0 1 5. Is Z× / 9 cyclic? 20 20 Solution: Z× = {±1. d) = φ 0 1 0 + φ 0 1 0 . 20 9 = {1. . 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. 20 (−3) 9 = {−3. Then we have 0 1 mc + b = mn = 1 . Since m can be any nonzero real number. 0 1 The fundamental homomorphism theorem implies that G/N ∼ R× . since once you have determined the kernel. b) + (c. Finally. b + d) = (a. 6.CHAPTER 3 SOLUTIONS 85 1 a b by θ((a. and φ Note that this part of the proof covers part (a). 0 1 0 1 (a) Show that N is a subgroup of G. and let N be the subset of all matrices of the form . and that N is normal in G. List the cosets of the cyclic subgroup 9 in Z× . c 1 m b 0 1 −1 = m mc + b 0 1 (b) Show that G/N is isomorphic to the multiplicative group R× . b)) = 0 1 0 . ±7. φ m b = 1 if and only if m = 1. so N = ker(φ). N is normal in G since 1/m −b/m 0 1 = m b 0 1 1 mc 0 1 1 0 ∈ N. Thus parts (a) and (b) can be proved at the same time. using the argument given for part (b). Let G be the subgroup of GL2 (R) consisting of all matrices of the form m b 1 b . 7} Z× . = maps G onto R× . 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 . Solution: The set N is nonempty since it contains the identity matrix. 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. −7} Since x2 ∈ 9 . for each element x of the factor group is not cyclic. −9} 3 9 = {3. ±9}. 9} (−1) 9 = {−1.

where a4 = e. so each coset has order 2. a6 b. a2 b}. 8. b2 = e. a2 }. a3 b}. (a) Show by a direct computation that N is a normal subgroup of D4 . and N abN ab = N e = N . a4 . a7 b. Assume that the dihedral group D4 is given as {e. and a2 ∈ N . (a) List all left cosets and all right cosets of N . a3 b}. a3 }. (b) Show that G/N has order 4. Let G = D8 . Since b and ab have order 2. . which implies that N is normal. we see that each element in the factor group has order 2. a. a4 b. Solution: The right cosets of N are N = {e. N b = {b. a6 }. N ab = {ab. bN = {b. a7 }. a6 }. a2 . a2 }. a3 . for all i. b. a4 . and verify that N is a normal subgroup of G. and let N = {e. a2 b. a2 . so G/N is not cyclic. a3 . 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 . but we get The fact that the left and right cosets of N coincide shows that N is normal. a6 b}. a2 . Let N be the subgroup a2 = {e. a5 b. a5 b. a4 . a7 }. N a = {a. a3 . ab. (b) Is the factor group D4 /N a cyclic group? Solution: The cosets of N are N = {e. N b = {b. We have N aN a = N a2 = N . N a = {a. a4 b. aN = {a. a2 b. but is not cyclic. and N ab = {ab. a7 b}. N bN b = N e = N . a3 b. a2 b}. N = {e. a5 . a6 }. and ba = a3 b. The left cosets of N are more trouble to compute. Solution: It is clear that there are 4 cosets. abN = {ab. a2 . a5 . a3 b}.86 CHAPTER 3 SOLUTIONS 7.

and then g(x) = (2x − 1)g(x) − 2. 4. Solution: To simplify the computations. x6 − 1) in Q[x] and write it as a linear combination of x8 − 1 and x6 − 1. use the Euclidean algorithm to show that 2x3 − 2x2 − 3x + 1 and 2x2 − x − 2 are relatively prime. and g(x) = (x4 + x2 + 1)(x2 − 1). The constant remainder at the second step implies that gcd(f (x). Over the ﬁeld of rational numbers. We ﬁrst obtain f (x) = (x + 1)g(x) + 2(x2 + x + 1). ﬁnd the greatest common divisor of x4 + x3 + 2x2 + x + 1 and x3 − 1. Solution: Let x8 −1 = f (x) and x6 −1 = g(x). and then the next step yields g(x) = (x − 1)(x2 + x + 1). At the next step we can use x rather than 2 2 3 2 x. x6 − 1) = x2 − 1. Over the ﬁeld of rational numbers. so gcd(f (x). Use the Euclidean algorithm to ﬁnd gcd(x8 − 1. 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. 3. we ﬁrst obtain 87 . Using the Euclidean algorithm.Chapter 4 Polynomials SOLUTIONS TO THE REVIEW PROBLEMS 1. 2. g(x)) = 1. and (x2 + x + 1) = 1 1 2 f (x) − 2 (x + 1)g(x). let 2x4 − x3 + x2 + 3x + 1 = f (x) and 2x3 − 3x2 + 2x + 2 = g(x). and x2 − 1 = f (x) − x2 g(x). Over the ﬁeld of rational numbers. We ﬁrst obtain f (x) = (x − 1 )g(x) − 3 x. We have f (x) = x2 g(x)+(x2 − 1). and express it as a linear combination of the given polynomials. Solution: Let x4 + x3 + 2x2 + x + 1 = f (x) and x3 − 1 = g(x). so this shows that gcd(x8 − 1. g(x)) = x2 + x + 1. Solution: Let 2x3 − 2x2 − 3x + 1 = f (x) and 2x2 − x − 2 = g(x).

7. 6. ±6. so for any root we must have (r − 1)|21. We have f (1) = 21. −6 as possibilities. ad + bc = 8. ±4. and since 2 is a root we have the factorization x5 − 2x4 − 2x3 + 12x2 − 15x − 2 = (x − 2)(x4 − 2x2 + 8x + 1). Factor x5 − 10x4 + 24x3 + 9x2 − 33x − 12 over Q. Factor x5 − 2x4 − 2x3 + 12x2 − 15x − 2 over Q. and ﬁnally we obtain the factorization f (x) = (x − 4)(x4 − 6x3 + 9x + 3). x + = 1 x2 g(x) + (− 1 )(x − 1)f (x). ±12. and this satisﬁes Eisenstein’s criterion for p = 2. so this eliminates all but ±2.88 CHAPTER 4 SOLUTIONS f (x) = (x+1)g(x)+(2x2 −x−1). The only possible rational roots of the second factor are 1 and −1. f (−2) = −294. so 2x + 1 is the greatest common divisor (we must then divide by 2 to make it monic). 4. which is irreducible in Z2 [x]. At the next step we obtain 2x2 − x − 1 = (x − 1)(2x + 1). Then f (2) = 32. (c) 2x10 + 25x3 + 10x2 − 30 Solution: Eisenstein’s criterion is satisﬁed for p = 5. ±2. and then g(x) = (x−1)(2x2 −x−1)+(2x+1). We have either b = d = 1. the fact that it has not roots in Q does not mean that it is irreducible over Q. Solution: The possible rational roots are ±1. This leads to the equations a + c = 0. ±3. The second factor is irreducible over Q since it satisﬁes Eisenstein’s criterion for p = 3. This implies that the polynomial is irreducible over Q. the only possible factorization has the form x4 −2x2 +8x+1 = (x2 +ax+b)(x2 +cx+d). 2 2 5. ac + b + d = −2. ±2. in which case a + c = 8.) Since the polynomial has no linear factors. Solution: The possible rational roots of f (x) = x5 −10x4 +24x3 +9x2 −33x−12 are ±1. and these do not work. in which . and bd = 1. Are the following polynomials irreducible over Q? (a) 3x5 + 18x2 + 24x + 6 Solution: Dividing by 3 we obtain x5 + 6x2 + 8x + 2. or b = d = −1. Beginning with the last equation and back-solving. (b) 7x3 + 12x2 + 3x + 45 Solution: Reducing the coeﬃcients modulo 2 gives the polynomial x3 + x + 1. 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. (It is important to note that since the degree of the polynomial is greater than 3.

and p(−2) = 3. p(−2) = 2. [2]. we could reduce x4 − 2x2 + 8x + 1 modulo 3 to get p(x) = x4 + x2 + 2x + 1. [2x + 2]. .CHAPTER 4 SOLUTIONS 89 case a + c = −8. using (2)(−2) = −4 ≡ 1 (mod 5). (We could also multiply each term by 3. p(−1) = −3. [x+1]. p(−1) = 3. show that [x + 1] is a generator. so x4 − 2x2 + 8x + 1 is irreducible over Q. Solution: The multiplicative group of F has 8 elements. (a) Express x4 + x as a product of polynomials irreducible over Z5 . and x2 + 2x + 2. Express 2x3 + x2 + 2x + 2 as a product of polynomials irreducible over Z5 . This polynomial has no roots in Z3 . and so [x + 1]4 = [−x]2 = [−1].) Checking for roots shows that p(0) = 1. Solution: If p(x) = x3 + 2x2 + 3. then p(0) = 3.) (b) Show that x3 + 2x2 + 3 is irreducible over Z5 . and hence over Q. but [x] is not. in the third the coeﬃcient of x is 0. [2x + 1]. [x]. [x+2]. Since the constant term of p(x) is 1. The factor p(x) = x2 − x + 1 is irreducible over Z5 since it can be checked that it has no roots in Z5 . the second has a nonzero cubic term. p(1) = 2. (c) In the multiplicative group of nonzero elements of F . and p(−2) ≡ −2. so we have the nine elements [0]. This reduces the question to factoring p(x) = x3 − 2x2 + x + 1. and so it is a generator for the multiplicative group of F . which shows that [x + 1] does not have order 2 or 4. (We get p(0) = 1. 10. Solution: The congruence classes are in one-to-one correspondence with the linear polynomials. Solution: To show that p(x) = x2 + 1 is irreducible over Z3 . p(2) = 3. (b) List the elements of the ﬁeld F = Z3 [x]/ x2 + 1 . In the ﬁrst the coeﬃcient of x is 1. Thus p(x) is irreducible over Z3 . 8. (a) Show that x2 + 1 is irreducible over Z3 . so p(x) is irreducible over Z5 . As an alternate solution. we only need to check that it has no roots in Z3 . and since [x]2 = [−1]. p(1) = 1. p(1) = 1. [1]. so the only possible factors are of degree 2. 9. The monic irreducible polynomials of degree 2 over Z3 are x2 + 1. x2 + x + 2. The only remaining possibility (by Lagrange’s theorem) is that [x + 1] has order 8. [x + 1]2 = [x2 + 2x + 1] = [−1 + 2x + 1] = [2x] = [−x]. On the other hand. Either case contradicts a + c = 0. it follows that [x] has order 4 and is not a generator. or p(x) = (x2 + x + 2)(x2 + 2x + 2). so p(x) itself is irreducible over Z5 . we have x4 + x = x(x3 + 1) = x(x + 1)(x2 − x + 1). the only possible factorizations are p(x) = (x2 + x + 2)2 . Solution: We ﬁrst factor out 2. [2x]. and this follows from the computations p(0) = 1. p(x) = (x2 + 2x + 2)2 . and p(−1) = 2. Solution: In general. p(1) = 1. p(−1) = −1. p(2) = 4. p(2) = 3.

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

the above computation also checks commutativity. 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. Solution: It is clear that the set is closed under addition. c ↔ z. b ↔ y.Chapter 5 Commutative Rings SOLUTIONS TO THE REVIEW PROBLEMS 1. and these are invertible since their determinant is nonzero. (a) Show that R is a commutative ring (you only need to check closure and commutativity of multiplication). Solution: Four of the matrices in R have 1’s on the diagonal. Squaring each of the other four matrices gives the zero matrix. (b) Find all units of R. and all nilpotent elements of R. and so they are nilpotent. and the following computation checks closure under multiplication. (c) Find all idempotent elements of R. 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. 93 .

then the image is 7Z42 and the kernel is 6Z120 . Thus R cannot be isomorphic to Z2 ⊕ Z2 . 2) = (0. while in Z3 ⊕ Z3 there is not. For the element a = (0. while in Z3 ⊕ Z3 the equation (a. while that of Z3 ⊕ Z3 is not. If φ(1) = 21. Are Z9 and Z3 ⊕ Z3 isomorphic as rings? Solution: The answer is no. then the image is 21Z42 and the kernel is 2Z120 . If φ(1) = 28. b)2 = (0. 7. 21. 7. y)(0. It follows that the largest possible order of an element is lcm[2.5. so R is not isomorphic to Z4 . φ(1) must be idempotent. Find all ring homomorphisms from Z120 into Z42 . Let R be the ring Z2 [x]/ x2 + 1 .94 CHAPTER 5 SOLUTIONS Solution: By part (b). Furthermore. 21. Show that although R has 4 elements. 4. what is the largest possible order 180 of an element? Solution: Since 180 = 22 32 5. (In Z9 let x = 3. 28. 0) implies a2 = 0 and b2 = 0. 180 = 4 9 5 = In the latter additive group. so it must belong to the subgroup 7Z42 = {0. and then a = 0 and b = 0. 3. Show that Ann(a) is an ideal of R. an element in R is either a unit or nilpotent. 35}. 2. Solution: Let φ : Z120 → Z42 be a ring homomorphism. If φ(1) = 7. In the group Z× of units of the ring Z180 . Solution: In R we have a + a = 0. On the other hand.) 5. the only idempotent element is the identity) and the only nilpotent element that is also idempotent is the zero matrix. 4] = 12.2. . Addition in the two rings is diﬀerent. We only need 2y ≡ 0 (mod 8). 6. it is not isomorphic to either of the rings Z4 or Z2 ⊕ Z2 . since in that ring (a. and this implies a = 0 and b = 0 since Z2 is a ﬁeld. Then Example 5. 28 are idempotent. 2) of the ring R = Z12 ⊕ Z8 . in R we have [x + 1] = [0] but [x + 1]2 = [x2 + 1] = [0]. Multiplication is also diﬀerent. 0) for (x. the order of an element is the least common multiple of the orders of its components. 6. since the additive group of Z9 is cyclic. since in Z9 there is a nonzero solution to the equation x2 = 0. Solution: We need to solve (x. for all a ∈ R. b)2 = (0. then the image is 14Z42 and the kernel is 3Z120 . and it can be checked that in 7Z42 . so the ﬁrst component x can be any element of Z12 . ﬁnd Ann(a) = {r ∈ R | ra = 0}. The argument can be given using either addition or multiplication. y) ∈ Z12 ⊕ Z8 .4 that the ring Z180 is isomorphic to the ring Z4 ⊕ Z9 ⊕ Z5 . 14.10 shows that Z× ∼ Z× × Z× × Z× ∼ Z2 × Z6 × Z4 . 42) = 6. it follows from Theorem 3. The only unit that is idempotent is the identity matrix (in a group. only 0. 0) implies a2 = 0 and b2 = 0. The additive order of φ(1) must be a divisor of gcd(120.

but if the set of zero divisors were closed under addition it would include (1. This set is certainly closed under addition. 2)) = Z12 ⊕ 4Z8 .3. (a) Show that I is an ideal of R. = Solution: Deﬁne φ : Z2 [x]/ x4 + 1 → Z2 [x]/ x2 + 1 by φ(f (x) + x4 + 1 ) = (f (x) + x2 + 1 ). and these correspond to 2Z36 and 3Z36 . Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients. 10. an obvious contradiction. so the maximal ideals of Z36 correspond to the prime divisors of 36. The maximal ideals of Z36 are thus 2Z36 and 3Z36 . prove that I is not maximal. The lattice of ideals of Z36 is exactly the same as the lattice of subgroups. then Z36 /P is a ﬁnite integral domain. . Give an example to show that the set of all zero divisors of a ring need not be an ideal of the ring. 8.7. and hence P is maximal. Hint: If you use the fundamental homomorphism theorem. with kernel equal to I. Solution: The elements (1. This mapping is well-deﬁned since x2 + 1 is a factor of x4 + 1 over Z2 . 1). Solution: Deﬁne φ : Z[x] → Z2 [x] by reducing coeﬃcients modulo 2. In Z every ideal is principal. Again. you can do the ﬁrst two parts together. 1) of Z × Z are zero divisors. the maximal ideals that contain 36Z are 2Z and 3Z. 9. Let R be the ring Z2 [x]/ x4 + 1 .CHAPTER 5 SOLUTIONS 95 while y = 0. Find all maximal ideals. which shows that there is a one-to-one correspondence between the ideals of Z/36Z and the ideals of Z that contain 36Z. and all prime ideals. (b) Show that R/I ∼ Z2 [x]/ x2 + 1 . Thus we only need to ﬁnd the maximal ideals of Z36 . 0) and (0. Thus Ann((0. so I is not maximal. (c) Is I a prime ideal of R? Solution: No: (x + 1)(x + 1) ≡ 0 (mod x2 + 1). which is not a ﬁeld. Prove that I is a prime ideal. so the relevant ideals correspond to the divisors of 36. An alternate approach we can use Proposition 5. and it is also closed under multiplication by any element of R since 4Z8 is an ideal of Z8 . It is not diﬃcult to show that φ is an onto ring homomorphism. and let I be the set of all congruence classes in R of the form [f (x)(x2 + 1)]. 7. Solution: If P is a prime ideal of Z36 . so it is a ﬁeld. 4. Then R/I is isomorphic to Z2 [x]. of Z36 = Z/36Z. This is an onto ring homomorphism with kernel I.

Let R be the ring Z2 [x]/ x3 + 1 . (a) Find all ideals of R. and [x2 ]. This also excludes [x2 + x] = [x][x + 1] and [x2 + 1] = [x2 ][1 + x]. so the only proper. 12. Then φ is one-to-one since if φ(r) = φ(s). the ideals of R correspond to the ideals of Z2 [x] that contain x3 + 1 . Thus the only units are 1. Solution: By Proposition 5.7.96 11. then re = se and r(1 − e) = s(1 − e). s(1 − e)) = (rse2 . (b) Show that if e is idempotent. φ is onto. nonzero ideals are the principal ideals generated by [x + 1] and [x2 + x + 1]. Solution: Note: Table 5. (c) Find the idempotent elements of R. it is easy to check that φ preserves addition. rs(1 − e)2 ) = (rse. then R ∼ Re ⊕ R(1 − e). Finally. rs(1 − e)) and φ(r)φ(s) = (re. since for any element (ae. [x + 1][x2 + x + 1] = [x3 + 1] = [0]. for all r ∈ R. (b) Find the units of R. Let R be any commutative ring with identity 1. Solution: We have [x]3 = [1]. . Furthermore. then 1−e is also idempotent. Solution: We have (1−e)2 = (1−e)(1−e) = 1−e−e+e2 = 1−e−e+e = 1−e. and for any r. On the other hand.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. [x]. Deﬁne φ : R → Re⊕R(1−e) by φ(r) = (re. b(1 − e)) = φ(r) for r = ae + b(1 − e). r(1−e)). We have the factorization x3 + 1 = x3 − 1 = (x − 1)(x2 + x + 1). b(1 − e)) we have (ae.1 gives the multiplication table. so [x] and [x2 ] are units. It is not necessary Table 5. CHAPTER 5 SOLUTIONS (a) Show that if e is an idempotent element of R. so [x + 1] and [x2 + x + 1] cannot be units. and adding the two equations gives r = s. s ∈ R we have φ(rs) = (rse. rs(1 − e)).3. = Solution: Note that e(1−e) = e−e2 = e−e = 0. r(1 − e))(se.

Let Z[i] be the subring of the ﬁeld of complex numbers given by Z[i] = {m + ni ∈ C | m. [x2 ]. Solution: Over Z2 we have the factorization x3 + x = x(x2 + 1) = x(x + 1)2 . and then [x2 + 1]2 = [x2 + 1]. [x2 + x + 1]. [x2 + x]} . Show that the rings R and S in the two previous problems are isomorphic as abelian groups. it is easy to see that the idempotent elements of R are [0]. This is a unit since [x2 +x+1]2 = [1]. Solution: Note: Table 5. [x2 + x]} [x] = {[0]. together with [0] and [1]. n ∈ Z} . we have [x2 ]2 = [x2 ]. and [x2 + x]. but not as rings. [x + 1]. it follows from part (a) that we only need to check [x2 +x+1]. Let S be the ring Z2 [x]/ x3 + x .3. as abelian groups. [x2 + x]} (b) Find the units of R. (c) Find the idempotent elements of R. Solution: Both R and S are isomorphic to Z2 × Z2 × Z2 . and [x2 + x] = [x][x + 1]. They cannot be isomorphic as rings since R has 3 units. [x2 + 1]. It is not necessary Table 5. Solution: Since no unit can belong to a proper ideal. [x2 + x] = {[0]. are the only idempotents. (a) Find all ideals of S.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. 13. [x2 + 1] = {[0]. 15. [x]. These. [x + 1].7 the proper nonzero ideals of S are the principal ideals generated by [x]. Solution: Since [x3 ] = [1]. [x2 + 1] = [x + 1]2 .2 gives the multiplication table. [x2 + 1]} [x + 1] = {[0]. 14.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]. while S has only 2. [1].

which show that θ is a ring homomorphism. 16. It is clear that 1 + i is in the kernel.98 CHAPTER 5 SOLUTIONS (a) Deﬁne φ : Z[i] → Z2 by φ(m + ni) = [m + n]2 . Solution: We have the following computations. We need to use the fact that [x2 ] = [−1] in Zp [x]/ x2 + 1 . Prove that θ is an onto ring homomorphism. We claim that ker(φ) is generated by 1 + i. Since the elements of Zp [x]/ x2 + 1 all have the form [a + bx]. and we note that (1 − i)(1 + i) = 2. which show that φ is a ring homomorphism. Prove that φ is a ring homomorphism. for all r ∈ R. for some congruence classes a and b in Zp . 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 . . 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 ] . and deﬁne the function φ : R → R/I ⊕ R/J by φ(r) = (r + I. Let I and J be ideals in the commutative ring R. Then m and n are either both even or both odd. and so it follows that m − n is always even. with ker(φ) = I ∩ J. deﬁne θ : Z[i] → Zp [x]/ x2 + 1 by θ(m + ni) = [m + nx]. 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) . Let m + ni ∈ ker(φ) = {m + ni | m + n ≡ 0 (mod 2)}. Find ker(φ) and show that it is a principal ideal of Z[i]. 2 and so m + ni belongs to the principal ideal generated by 1 + i. Solution: We have the following computations. it is clear the θ is an onto function. (b) For any prime number p. (a) Show that φ is a ring homomorphism. r + J).

for g(x) polynomials f (x) and g(x) with rational coeﬃcients. Thus φ(r) = (a+I. or. (b) Show that if I + J = R. so ker(φ) = I ∩ J. k(x) ∈ Z[x]. 17. We have a − r = a − bx − ay = ax − bx ∈ I. noting that a = ax + ay and b = bx + by. equivalently.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. then we can write 1 = x+y. Considering Z[x] to be a subring of Q[x]. Given any element (a + I. and so a4 ≡ 1 (mod p). b + J) ∈ R/I ⊕ R/J . 18. which is p − 1. Since the zero element of R/I ⊕ R/J is (0 + I. and φ is onto. we have r ∈ ker(φ) if and only if r ∈ I and r ∈ J. Prove that the ring Zp [x]/ x2 + 1 is a ﬁeld. then we have f (x) = m h(x) for h(x). p Solution: We must show that x2 + 1 is irreducible over Zp . Therefore p = 4q + 1 for p some q ∈ Z. and thus it must have order 4. . show that these two integral domains have the same quotient ﬁeld. Hint: Show that a root of x2 = −1 leads to an element of order 4 in the multiplicative group Z× . for some x ∈ I and y ∈ J. 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). that x2 + 1 has no root in Zp . b+J). The element a cannot be a root of x2 − 1. and b−r = b−bx−ay = by −ay ∈ J. Solution: An element of the quotient ﬁeld of Q[x] has the form f (x) . and thus R/(I ∩ J) ∼ R/I ⊕ R/J . By Lagrange’s theorem. 0 + J). then φ is onto. Suppose that a is a root of x2 + 1 in Zp . contradicting the assumption. Let p be an odd prime number that is not congruent to 1 modulo 4. consider r = bx + ay. so it does not have order 2. = Solution: If I +J = R. Then a2 ≡ −1 (mod p). g(x) k(x) and this shows that f (x) g(x) belongs to the quotient ﬁeld of Z[x]. The isomorphism follows from the fundamental homomorphism theorem. this means that 4 is a divisor of the order of Z× .

100 CHAPTER 5 SOLUTIONS .

√ 3. Thus u3 + 3u + 3 = (u + 2)(u2 − 2u + 7) − 11. Solution: Dividing x3 + 3x + 3 by x2 − 2x + 7 gives the quotient x + 2 and remainder −11. In Q(u). u2 − 2iu + i2 = 2. Find the minimal polynomial of 1 + 3 2 over Q. √ √ √ √ Solution: We have Q ⊆ Q( 2) ⊆ Q( 2. and it follows easily that 2 ∈ Q(u) and i ∈ Q(u). Then x − 1 = 3 2. Eisenstein’s criterion (with p = 3) shows that x3 − 3x2 + 3x − 3 is irreducible over Q. 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. 101 . (a) Show that Q( 2 + i) = Q( 2. We have [Q( 2. and so (7 − 2u + u2 )−1 = (2 + u)/11 = (2/11) + (1/11)u. Thus the minimal and polynomial for 2 + i is x4 − 2x2 + 9. we have u − i = 2. Squaring again √ combining terms gives u4 −2u2 +9 = 0. i). express (7 − 2u + u2 )−1 in the form a + bu + cu2 . which yields x3 − 3x2 + 3x − 1 = 2. √ √ Solution: Let x = 1 + 3 2. and u2 − 3 = 2iu. √ √ √ Solution: Let u = 2 + i. √ (b) Find the minimal polynomial of 2 + i over Q. i). Since ( 2 + i)( 2 − i) = 2 − √ = 3. √ √ Since u = 2 + i. and so the minimal polynomial for √ 2 + i must have degree 4. so this is the required minimal polynomial. √ √ 2. The reverse inclusion is obvious. and therefore x3 − 3x2 + 3x − 3 = 0. i) ⊆ Q(u). Let u be a root of the polynomial x3 + 3x + 3. Thus [Q( 2 + i) : Q] = 4. i) : √ √ Q( 2)] =√ since i is a √ of a polynomial of degree 2 over Q( 2) but is 2 root not in Q( 2).Chapter 6 Fields SOLUTIONS TO THE REVIEW PROBLEMS 1. we have i2 √ √ −1 2 − i = 3( 2√ i) + ∈ Q(u). so Q( 2.

Then (α − i)3 = 2. 3 25. 2.102 CHAPTER 6 SOLUTIONS 4. 3 5. 3. and remains irreducible √ over Q( 5 2). 5 3 25}. 3 25. This root would have degree 3 over Q. 2 3 5. √ √ √ Solution: The set {1. 3 5. 2. √ √ Find a basis for Q( 5. Does 4 2 belong to Q( 3 2 + i)? √ √ Solution: Let α = 3 2 + i. then it would have a linear factor. 2) over Q. Therefore {1. Since x5 − 2 is also irre√ ducible by Eisenstein’s criterion. i). 3 5. and that is impossible since 3 is not a divisor of 5. 7. have [Q( 7 2) : Q] √ 7. √ √ √ Find the degree of 3 2 + i over Q. Solution: Eisenstein’s criterion works with p = 2.2. √ √ √ √ √ √ √ Since {1. √ √ √ Solution: The set {1. 2 3 25} are linearly independent over Q. It follows that u has degree 2. and so it would √ have a root in Q( 5 2). 5. Solving for i we get √ i = (α3 − 3α − 2)/(3α2 − 1). If x3 + 6x2 − 12x + 2 could √ be factored over Q( 5 2). Since u = ( 7 2 + 3)( 7 2)3 . √ √ We will show that u cannot have degree ≤ 3.√it √ follows immediately that a = b = 0. 2 3 5. 5 is a basis for Q( 5. If 2 + 3 5 is a root of a polynomial ax3 + bx2 + cx + d in Q[x]. 3 5) in over Q( 3 5). 3 5. the √ minimal polynomial x2 − 5 of 5 remains irre√ √ √ ducible √ the extension Q( 3 5). or 3 over Q. or 6 over Q. 3 25. or α3 − 3iα2 − 3α + i = 2. 3 5. and then c = d = 0 as well. 2. so that α − i = 3 2. 2 3√ is 25} √ √ √ a basis for Q( 3 5. √ √ Show that [Q( 2 + 3 5) : Q] = 6. It follows 2 √ √ √ 3 3 3 immediately that 2 ∈ Q( 2 + i). and so Q( 2 + i) = Q( 3 2. √ √ √ √ Solution: Let u = 7 16 + 3 7 8. 3 25} is a basis for Q( 3 5) over Q. 5. [Q( 5 2) : Q] = 5. Thus {1. and then u must have degree 7 over Q since [Q(u) : Q] = is a divisor of [Q( 7 2) : Q]. 6. We conclude a √ that [Q( 2 + 3 5) : Q] = 6. 3 25} is a basis for Q( 3 5) over Q. 5 3 5.4 shows that the required basis √ √ √ so √ √ of is {1. √ √ Find [Q( 7 16 + 3 7 8) : Q]. 8. so 2 + 3 5 cannot satisfy √nonzero polynomial of degree 1. 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. and this shows that i ∈ Q( 3 √+ i). so we have α3 − 3iα2 + 3i2 α − i3 = 2. the minimal polynomial √2 − √ of√ 2 remains irre2 √ √ x √ √ √ ducible over the extension Q( 3 5). we 2). 3 5) over Q. and since this √ extension has degree 3. and this extension contains u = 2 + 3 5. it follows that √ u ∈ Q( 7√ Since x7 − 2 is irreducible over Q by Eisenstein’s criterion. Show that x3 + 6x2 − 12x + 2 is irreducible over Q. and since this √ extension has degree 3. and √ the proof √ Theorem 6. .

Therefore [Q( 3 2 + i) : Q] = 6. [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. On the other hand.CHAPTER 6 SOLUTIONS 103 √ Since x3 − 2 is irreducible over Q. Since x2 + 1 is irreducible over Q. Therefore √ √ [Q( 3 2 + i)√ Q] ≤ 6. so it cannot belong to an extension of degree 6 since 4 is not a divisor of 6. . we see that i has√ degree 2 over Q. the number 3 2 has degree 3 over Q. 4 2 has degree 4 over Q since x4 − 2 is irreducible over Q. √ Finally.

Inc. Saracino. and G. 1992... Fraleigh. 1996.)... 1991 Birkhoﬀ. New York: John Wiley & Sons.). Inc. Hillman. 1977. Boston: Houghton Miﬄin Co. J. I. Alexanderson.. B.: Prentice-Hall. 1999. A First Course in Abstract Algebra (6th ed. D. Topics in Algebra (2nd ed. 1975. A. Inc... 1992.. New York: Dover Publications. Mac Lane. ———.). N. Abstract Algebra: A First Undergraduate Course. Contemporary Abstract Algebra (4th ed. (3rd ed. New York: Springer-Verlag. 1985. Fields. .. R. J. G... A History of Algebra: from al-Khwarizmi to Emmy Noether.J. 1983. New York: Macmillan Publishing Co. Englewood Cliﬀs.: AddisonWesley Publishing Co. 1998 Herstein.. Abstract Algebra: A First Course. W. L. M.). Artin. and Groups: An Introduction to Abstract Algebra London: Edward Arnold. P.. Mass. Algebra. J. Gallian. and S. Maxﬁeld. A Survey of Modern Algebra (4th ed. Abstract Algebra. T. N. Prospect Heights: Waveland Press. Prospect Heights: Waveland Press.. and M. Inc. B.104 BIBLIOGRAPHY BIBLIOGRAPHY Allenby.. Abstract Algebra and Solution by Radicals. 1999. New York: John Wiley & Sons. Reading. Maxﬁeld... Inc. J. L. E.. Rings.). Van der Waerden.

84. 16. 102 cross product. 95. 23. 65 combination. 17 disjoint cycles. 33. 54 division algorithm. 44–46. 24. 86 criterion. 88. 77 annihilator. 40. 45. composite. 77 Chinese remainder theorem. 76 closure. onto. 14. 18. 15. 6. 7 function. for an extension ﬁeld. 14–18. 101. 29. 82. 22. 27. matrix form. 89–91 ﬁeld. 22. 67. 53. 64–68. 13 algorithm. 73. 26. 43 dihedral group. 97. 27. 6. 87. linear. 6. 13 group. 3. alternating. 94. 35. 20. 5. 13 fractional linear transformation. 8. 78. 15. 76. 8. 83 coset. 89–91 ﬁnite group. 99 ﬁeld. 84. 9. 30. abelian. 8. Euclidean. 22. 88 general linear group. 80 fundamental homomorphism theorem. 21 centralizer. 69. 53 factor group. 54 cyclic. 82. 101. 62 basis. 26. 7 function. 59. 86 cyclic group. 60 function. 9. 24. 56. 20. 41. 36. 63 determinant. 44. 97 composite function. 31. 99 ﬁeld. 95. 30. 65. 86 ﬁeld. 44. 10. 81. ﬁnite. 21. for polynomials. 5. 64. of integers 2. 3–5. 30. 74 group. 87–89 ﬁnite ﬁeld. 7 congruence. 28. 3. 28. 17. 35. 86 direct product. 90 Euclidean algorithm. 15. 72. 35 alternating group. 27. 28. for rings. 102 element. 30. 31. 24–26. 7 fundamental homomorphism theorem. 68. 1 division. 50 function. linear. 36. 93. 8. inverse. 102 binary operation. 98 gcd. 51 Eisenstein’s criterion. 65. 22. 46 Euclidean algorithm. 83. 25. 94. 87. 85 generator. 81 cyclic subgroup. 13 group. for groups. 13 cancellation law. 13. 23 . 41 conjugacy class. 24–26. 65. of Eisenstein. 99 Gaussian integers. 30. 21. 22–24. 57. 97 element. 29. nilpotent. units. 57 eigenvalue. of polynomials. 81. 27. 69. cyclic. 39. 57 cycle. 96. 26. idempotent. 52. 77. 2 complex numbers. 14. 14 Cayley’s theorem. 1. 94 associative law. 46 Euler phi-function. 1. one-to-one. 46 gcd. 51. 93. 82. 82. 94 equivalence relation. 14 dot product.INDEX 105 Index abelian group. of quotients. 6. 62. 51–53 Euclidean algorithm. 17. 77 group. 5. division. 45. 74 even permutation. 7. 30. 65 digit. of rational numbers. 1 algorithm. 5 congruence.

99 prime. principal. 94. of rings. 81. 3. 62. 6. 30. 28. 10. modulo n. 54 parallel plane. 29. 95 polynomial. 94 nilpotent element. 54 group. relatively. 22. 50. 15. 14 inverse element. 51 linearly independent vectors. of a ring homomorphism. of a ring homomorphism. 6. 22. 29. 29. 93 normal subgroup. 30. 13. 28. of a permutation. 23. 7. 53 plane. 93. 44. 45. prime. 21. 39 linear combination. 54 perpendicular plane. 18. 8. 2. 89. 95. 86 group. 66. 78–80 homomorphism. 21 group homomorphism. irreducible. 30. modulo n. of permutations. 97. 8. 28. 50 one-to-one function. 60. 84 . 99 isomorphism. 33. 51 order. 60. 95–98 74. 5. 8. 31. 8. 53 partition. 25. 10 permutation. 2. 96. 14. 30. 101–103 kernel. 82. 50. minimal. 22. 60 linear transformation. 15. 6. 45. 90 multiplicative order. modulo n. maximal. of rings. multiplicative. 94. 6. symmetric. 62. 77. 33. 102 isomorphic rings. 9 permutation. 80 polynomial. 64 inverse. 10. 101. of groups. 99 INDEX lattice diagram. 95 ideal. 5 identity element. 29. 59 integers mod n. 5 linear transformation. 83. 23. 80 induction. 47. 24–26. principal ideal. 69–73 isomorphism. 29. 57. 5. 26. 33. 8. 95 ideal. 30. 82. 47 order. multiplicative. 41. 44. 53 matrix. 77. 94 image. 86 nullity. 101 prime ideal. of subgroups. 47 nilpotent element. 95 Lagrange’s theorem. of a ring. 53 kernel. 35 linear congruence. of groups. 16–18. 7. 38. 60. 30. 94. 56. 26. even. 7. 6. 23. 53 permutation group. 93.106 group. 29. 5. 53 plane. 49 ideal. 22. 94. 11. dihedral. 44–46. 8. 11. 90 invertible matrix. 98 horizontal line test. 95 minimal polynomial. 9. 94. 97 idempotent element. 13 group. 11. parallel. 23 homomorphism. 16. 31. 65 image. invertible. perpendicular. 5. 53 irreducible polynomial. 101 multiplicative inverse. 51 maximal ideal. 97–99. 10. 44. 20. 33. 95–98 idempotent element. fractional. 5 nilpotent element. 30. 6. ﬁnite. 95. 31. 5 multiplicative order. 78–80 group isomorphism. 41. 2 lattice diagram. 97. 9. 101 ideal. 51 matrix. 45. 30. 23. 74 order. of a group homomorphism. 95. 21. 58. 51 onto function. 50. 15. 19. 31. 88. 5. 76.

mod 20. 19. 69. 27. 15. 22. 85 units. 63. 26. 17. of a ring. mod n. of a matrix. 21. 93. 21. 33. of Lagrange. 23. 50 transitive law. 102 root. 21. 14. 23. 98 root. 63. 62 units. 99 transformation. 86 subgroup. mod 9. 15. 31. 97. 96. 74 units. mod 18. 88 subgroup. 14. 6. 15 symmetric group. 101. 82. 16. 75 units. 2 ring homomorphism. 50. 16. 52 relatively prime polynomials. 97 units. 64. 21. mod 13. 51 rational roots.INDEX quaternion group. linear. 63. 8. 5. mod 17. 66 subring. 99 subspace. 16. 79 units. 74 units. mod 21. 29. 17. 87 relatively prime. 75 quotient ﬁeld. 99 rank. 52 unit. 17 vertical line test. 50 rank nullity theorem. normal. mod 36. 49 well-deﬁned function. 95. 8. 65. mod 24. 31. 15. 84 units. 75 units. 21. 42. 24 107 . 88 reﬂexive law. 58. 79 units. 23. of a polynomial. 94. 94. 21. rational. 29. 30. 26. 16. 16. mod 15. 77 symmetric law. 5. 52 system of congruences. mod 7. 43 vector space. 25. 30. 58. mod p. 14. 66 units. 45 units. 4. 6. 31. 47 theorem. 80 units digit. 26.

Sign up to vote on this title

UsefulNot useful- Problems and Solutions to Abstract Algebra (Beachy, Blair)
- Abstract Algebra (Beachy, Blair) 7.1 Problems and Solutions
- 3. ABSTRACT ALGEBRA.pdf
- Abstract Algebra Solutions
- Guide
- Solution's Manual Abstract Algebra Rotman
- Abstract Algebra I. N. Herstein p265
- Abstract Algebra
- [O.Saracino,Abstract algebra,A first course.pdf
- Modern Algebra Durbin
- Abstract Algebra
- Abstract Algebra
- A first course in abstract algebra
- Hunger Ford Solution Manual
- Beachyguide2008
- aljabar abstrak
- Abstract Algebra
- Introduction to Analysis Problems and Solutions
- Elementary Analysis Solutions
- Abstract Algebra (Herstein 3rd Ed)
- J.rotman. .a.first.course.in.Abstract.algebra
- [I. N. Herstein] Topics in Algebra, 2nd Edition (1(Bookos.org)
- "Fuckin' Concrete Contemporary Abstract Algebra Introduction..." by Nicolas Bourbaki Junior
- Abstract Algebra
- Abstract Algebra I. N. Herstein (Solution).
- Herstein Abstract Algebra Student's Solution Manual
- Introduction to Topology
- Dummit Solutions
- 2. Joseph A. Gallian, Contemporary Abstract Algebra (4th Edition), Narosa Publishing House,
- (eBook - PDF - Mathematics) - Abstract Algebra