ABSTRACT ALGEBRA

:
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
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
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
—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
×
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
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

−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) =

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
(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
[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
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
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
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

1
(b
1
), θ
2
(a
2

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

1
(b
1
), θ
2
(a
2

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
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

−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) =

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
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
_
_
_
_

_
_
_
_
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
[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
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
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
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

. . . . . SOLUTIONS 1 Integers 2 Functions 3 Groups 4 Polynomials 5 Commutative Rings 6 Fields BIBLIOGRAPHY INDEX . . . . . . . . . . .iv CONTENTS 33 33 33 35 49 57 87 93 101 104 105 6 FIELDS Review problems . . . . . . . . . . . . . . . . .

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

vi PREFACE .

—When working on questions involving divisibility you may ﬁnd it useful to go back to Deﬁnition 1. The examples in Section 1.1. you need to make sure that you are familiar with all of the deﬁnitions and theorems in the section. it is a signiﬁcant step forward into the realm of abstract algebra. simplify. one that is obtained by considering all powers of a particular element. actually tell you everything you will need to know about cyclic groups. the proofs of the theorems contain important techniques that you need to copy in solving the exercises in the text. —Another approach to proving that b|a is to use the division algorithm (see Theorem 1. If you expand the expression b|a by writing “a = bq for some q ∈ Z”.3) to write a = bq + r.1 Divisors Before working through the solved problems for this section. then you have an equation to work with.Chapter 1 INTEGERS Chapter 1 of the text introduces the basic ideas from number theory that are a prerequisite to studying abstract algebra. try to write down an expression for a and expand. In fact. where 0 ≤ r < b. constructed using congruence classes of integers. Then to prove that b|a you only 1 . Here are several useful approaches you should be able to use. One of the ﬁrst broad classes of groups that you will meet depends on the deﬁnition of a cyclic group.1.1. and so you can use all of the things you already know (from high school algebra) about working with equations. For example. in Chapter 3 of the text you will be introduced to the concept of a group. although Chapter 1 is very concrete. This equation involves ordinary integers. 1. Many of the concepts introduced there can be abstracted to much more general situations. —To show that b|a.4. In many cases. or substitute for terms in the expression until you can show how to factor out b.

This is really useful in working on questions involving greatest common divisors. CHAPTER 1. SOLVED PROBLEMS: §1. .1 22.6 states that any two nonzero integers a and b have a greatest common divisor. 26. then b = ±a.2 states that integers a and b are relatively prime if and only if there exist integers m and n with ma + nb = 1.  0 29.6). An integer is a linear combination of a and b if and only if it is a multiple of their greatest common divisor. gcd(n − 1. If n is a positive integer. Give a proof by induction to show that each number in the sequence 12. n + 10).. is divisible by 6. 27. INTEGERS —Theorem 1. n2 + n + 1) = 3. . 102. Prove that if n is a positive integer. . if you have a linear combination ma + nb = d. Prove that if a and b are nonzero integers for which a|b and b|a. Which of the integers 0. 377). Find gcd(435. . 23. ﬁnd the possible values of gcd(n. 30. . Prove that if n is an integer with n > 1. then gcd(n − 1. This is one of the most useful tools in working with relatively prime integers. then m2 − n2 is divisible by 8. . Find gcd(3553.1. which can be expressed as the smallest positive linear combination of a and b. it only shows that gcd(a. n are integers? 25.2 need to ﬁnd some way to check that r = 0. 10002. 527). More generally. 24. b) is a divisor of d (refer back to Theorem 1.2. 10 can be expressed in the form 12m + 20n. where m. and express it as a linear combination of 435 and 377. 1002. . 1. Remember that this only works in showing that gcd(a. n2 + n + 1) = 1 or n  1 0 −1 1 0  = 0 0 0 0 0 1 0  0 0  1 1. and express it as a linear combination of 3553 and 527.2 Primes Proposition 1. Since the fundamental theorem of arithmetic (on prime factorization) is proved in this section.1. . Prove that if m and n are odd integers. you now have some more familiar techniques to use. 28. then  0 1 if and only if 4|n. b) = 1.

3. and you can multiply both sides of a congruence by the same integer. you can only divide through by an integer that is relatively prime to n. If two integers are congruent modulo n.) What things are diﬀerent? In an ordinary equation you can divide through by a nonzero number. Just be very careful! One of the important techniques to understand is how to switch between congruences and ordinary equations. for all integers n > 1. but only guarantees that dividing by n produces the same remainder in each case. b2 ) = 1. For positive integers a. . Find all integer solutions of the equation xy + 2y − 3x = 25. What things are the same? You can add or subtract the same integer on both sides of a congruence. First. it is not exactly the same.3 Congruences In this section. b) = 1 if and only if gcd(a2 . This is usually expressed by saying that if gcd(a. the integers may diﬀer by some multiple of n. 24. any equation involving integers can be converted into a congruence by just reducing modulo n. (a) Use the Euclidean algorithm to ﬁnd gcd(1776. 1492). This works because if two integers are equal. In a congruence modulo n.3. 26. b. 1492). 30. 1089). n) = 1. Let m and n be positive integers. Give the lattice diagram of all divisors of 250. for all integers n > 1. n) = 1 and ac ≡ ad (mod n). 1089). 25. Prove that gcd(2n2 + 4n − 3. then are certainly congruent modulo n. The do the opposite conversion you must be more careful.3. (b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776. that doesn’t make them equal. then a ≡ c (mod n). and you can use the fact that if a ≡ b (mod n) and b ≡ c (mod n). Prove that gcd(2m − 1. (a) Use the Euclidean algorithm to ﬁnd gcd(1274. and the comments in the text both before and after the proof of the proposition. 3 27. In other words. 2n − 1) = 1 if and only if gcd(m. 1. (Review Proposition 1.1. You can use substitution. Do the same for 484. prove that gcd(a. CONGRUENCES SOLVED PROBLEMS: §1. Is the same true for 2n − 1 and 3n − 1? 29. 28.2 23. (b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274. Prove that n − 1 and 2n − 1 are relatively prime. then c ≡ d (mod n). 2n2 + 6n − 4) = 1. it is important to remember that although working with congruences is almost like working with equations.

3 26. because we have to introduce another variable. 27. Actually. Give integers a. Notice that converting to an equation makes it more complicated. for some q ∈ Z . Solve the congruence 42x ≡ 12 (mod 90).4 CHAPTER 1. 3x ≡ 2 (mod 13) . In the example.6 very carefully. Solve the system of congruences 30. n to provide an example of a system x ≡ a (mod m) that has no solution. we can reduce each term modulo 5. Solve the system of congruences x ≡ 2 (mod 9) x ≡ 4 (mod 10) . (b) Show that if x = m and y = n is an integer solution to the equation in part (a). m. INTEGERS The conversion process is illustrated in Example 1. because you might miss important algorithms or computational techniques.5 of the text. x ≡ b (mod n) . and all simultaneous linear equations of the form x ≡ a (mod n) and x ≡ b (mod m). You should read the proofs of Theorem 1. then so is x = m + 15q and y = n − 22q. where the congruence x ≡ 7 (mod 8) is converted into the equation x = 7 + 8q . 55x ≡ 35 (mod 75). 55x ≡ 36 (mod 75). b.3. These proofs actually show you the necessary techniques to solve all linear congruences of the form ax ≡ b (mod n). and you can’t aﬀord to skip over their proofs. SOLVED PROBLEMS: §1. for any integer q. 29. Solve the system of congruences 31. we really want a congruence modulo 5. x ≡ 23 (mod 32) .3. (a) Find all solutions to the congruence (b) Find all solutions to the congruence 28. 5x ≡ 14 (mod 17) x ≡ 5 (mod 25) 32.3. (a) Find one particular integer solution to the equation 110x + 75y = 45. so that we ﬁnally get x ≡ 2 + 3q (mod 5) . so the next step is to rewrite the equation as x ≡ 7 + 8q (mod 5) . Many of the theorems in the text should be thought of as “shortcuts”. where the moduli n and m are relatively prime.5 and Theorem 1.

4 of the text contain several deﬁnitions for elements of Zn . so that you can easily shift back and forth between them. They will provide n an important source of examples in Chapter 3. . and it is worthwhile to learn all of the approaches. written [a]n [x]n = [b]n . The exercises for Section 1. 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).4. If (a. any linear congruence of the form ax ≡ b (mod n) can be viewed as an equation in Zn . To be more precise. n n It is well worth your time to learn about the sets Zn and Z× . (a) Compute the last digit in the decimal expansion of 4100 . for all positive integers n. trying to divide by a in the congruence ax ≡ b (mod n) can get you into trouble unless gcd(a. SOLVED PROBLEMS: §1. Find the multiplicative inverse of each nonzero element of Z13 . the element [a] ∈ Zn is said to be n idempotent if [a]2 = [a]. Since |Z× | = ϕ(n). then the smallest positive integer k such that ak ≡ 1 (mod n) is called the multiplicative order of [a] in Z× . 1. Sometimes it helps to have various ways to think about a problem. Finally. Prove that 10n+1 + 4 · 10n + 4 is divisible by 9. INTEGERS MODULO N 33. 5 36.4 30. provided we think in terms of equivalence classes. (b) Is 4100 divisible by 3? 34. when we begin studying groups. n) = 1. Prove that the fourth power of an integer can only have 0. For example. Instead of thinking in terms of division. n) = 1. and choose whichever approach is the most convenient.1. 1. This gives you one more way to view problems involving congruences. or 6 as its units digit. Find all integers n for which 13 | 4(n2 + 1). 35. provided [a]−1 exists. 31. it is probably better to think of multiplying both sides of the equation [a]n [x]n = [b]n by [a]−1 . 5. 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]. and nilpotent if [a]k = [0] for some k.4 Integers Modulo n The ideas in this section allow us to work with equations instead of congruences.

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

which give a very important class of examples. 7 . 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 . and inverse functions. as introduced in this section. 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. and g(y) = ln y. as well as on your knowledge of the groups Zn and Z× . the functions that are one-to-one correspondences will be particularly important. functions will provide a way to compare two diﬀerent structures. 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. When you begin to study groups in Chapter 3. for all y ∈ R+ . n 2.1. The second goal of the chapter is to begin studying groups of permutations. for all x ∈ R. In this setting. the deﬁnition of function. provide one of the most important examples of a pair of inverse functions. is stated rather formally in terms of ordered pairs. Deﬁnition 2.Chapter 2 FUNCTIONS The ﬁrst goal of this chapter is to provide a review of functions.) In terms of actually using this deﬁnition. you will be able draw on your knowledge of permutation groups.1. That makes it especially important to understand the concept of an inverse function.1. In our study of algebraic structures in later chapters. one-to-one and onto functions. The functions f : R → R+ and g : R+ → R deﬁned by f (x) = ex .1 Functions Besides reading Section 2. (Think of this as a deﬁnition given in terms of the “graph” of the function. it might help to get out your calculus textbook and review composite functions.

(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. You are already familiar with one of the important examples: in Chapter 1 we split the set of integers up into subsets. 26.1. Let A be an n × n matrix with entries in R. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax. 2. 21.1. 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. 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.8 CHAPTER 2. This led to the concept of congruence modulo n. Deﬁne the function θ : Z× → Z× by 17 17 17 θ(x) = ax. Deﬁne a linear transformation L : Rn → Rn by L(x) = Ax.7. Let A be an m × n matrix with entries in R.1. where AT is the transpose of A. then it is invertible.1. and assume that m > n. 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. FUNCTIONS SOLVED PROBLEMS: §2. Explain why this agrees with Deﬁnition 2. Let A be an n × n matrix with entries in R. Note: A vector x is called an eigenvector of A if it is nonzero and there exists a scalar λ such a that Ax = λx. 25. for all x ∈ Z× . Explain why this agrees with Deﬁnition 2. for all x ∈ Rn . 24. for all x ∈ Rn . First there is the deﬁnition .4. 23.1 20. In this section you will ﬁnd three diﬀerent points of view. ﬁnd the 17 inverse function θ−1 . Deﬁne a linear transformation L : Rn → Rm by L(x) = Ax. Is θ one to one? Is θ onto? If possible. Let a be a ﬁxed element of Z× . more than one 22. which is a model for our general notion of an equivalence relation. depending on the remainder when the integer is divided by the ﬁxed integer n. Show that L is a one-to-one function if det(AT A) = 0. for all x ∈ Rn .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. Prove that L is one-to-one if and only if no eigenvalue of A is zero. looking at the one idea of splitting up a set S from three distinct vantage points.

Let u be a ﬁxed vector in R3 . it turns out that every partition (and equivalence relation) really comes from a function f : S → T . deﬁne (x1 . 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.2. Then there is the notion of a partition of S. Show that ∼ is an equivalence relation. . SOLVED PROBLEMS: §2.2. y1 ) ∼ (x2 . and give a geometric description of the equivalence classes of ∼. Your goal should be to learn about each point of view. On the set C of complex numbers. 20. Show that ∼ is an equivalence relation. 19. deﬁne z1 ∼ z2 if ||z1 || = ||z2 ||. b)} of all ordered pairs of positive integers. 17. Check that ∼ deﬁnes an equivalence relation. On the set {(a. deﬁne v ∼ w if v ·u = w ·u. where · denotes the standard dot product. for all x ∈ R. For the linear transformation L : R3 → R3 deﬁned by L(x. 12 Show that the formula f deﬁnes a function. where we say that x1 and x2 are equivalent if f (x1 ) = f (x2 ). For the function f : R → R deﬁned by f (x) = x2 .2 14. for all [x]12 ∈ Z12 . x + y + z. For vectors v and w. EQUIVALENCE RELATIONS 9 of an equivalence relation on S. so that you can easily switch from one to the other. describe the equivalence relation on R that is determined by f . x + y + z) . for all (x. and assume that u has length 1. z) ∈ R3 . 16. which is a big help in deciding which point of view to take. Finally. 15. y2 ) if x1 y2 = x2 y1 . which tells you when two diﬀerent elements of S belong to the same subset. z) = (x + y + z. Find the image of f and the set Z12 /f of equivalence classes determined by f . y. On the set of all n × n matrices over R. Show that this deﬁnes an equivalence relation. 18. y. deﬁne A ∼ B if there exists an invertible matrix P such that P AP −1 = B. which places the emphasis on describing the subsets. 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.

just as in calculus. Our point of view is that permutations are functions. so we have to do the computations from right to left. FUNCTIONS 2. You need to do enough computations so that you will feel comfortable in dealing with permutations. and Sym(S) is the set of all permutations on S. For 17. τ σ. we will say that G is a group of permutations if the following conditions hold. 16. then στ σ −1 has order m. Write σ as a product of disjoint cycles. τ . (iii) if σ ∈ G. . τ −1 . 9)(1. (i) If σ. then τ σ ∈ Sym(S). then σ −1 ∈ Sym(S). σ −1 . 4. τ ∈ G. 6) ∈ S9 .10 CHAPTER 2. What is the order of σ? Compute σ −1 .6. τ στ −1 . (iii) if σ ∈ Sym(S). In two of the problems. In the text we noted that if S is any set. )(6.3 Permutations This section introduces and studies the last major example that we need before we begin studying groups in Chapter 3. 7). We will see later that this agrees with Deﬁnition 3. Compute the order of τ = 1 2 3 4 5 7 2 11 4 6 σ = (3. then τ σ ∈ G. 4.1 of the text. 9. If you are reading another book along with Abstract Algebra. If G is a nonempty subset of Sym(S). 2. we need the following deﬁnition.3 13. 5. and we write functions on the left. 15. Let σ = (2. for any permutation σ ∈ Sn . write σ as a 7 5 6 9 2 4 8 1 3 product of disjoint cycles. What is the order of σ? Is σ an even permutation? Compute σ −1 . 8. then σ −1 ∈ G. (i) If σ. instead of the way we have deﬁned multiplication. compute the order of στ σ −1 . 6 7 8 9 8 10 9 10 1 3 11 5 . στ σ −1 . you need to be aware that some authors multiply permutations by reading from left to right. SOLVED PROBLEMS: §2. Prove that if τ ∈ Sn is a permutation with order m. 7. (ii) 1S ∈ G. 1 2 2 5 3 1 4 8 5 3 6 6 7 4 8 7 9 9 and 14. then we have the following properties. (ii) 1S ∈ Sym(S). write each of these permutations as a 1 5 4 7 2 6 8 9 3 product of disjoint cycles: σ. For the permutation σ = 1 2 3 4 5 6 7 8 9 . 8. στ . 6)(3. For the permutations σ = τ= 1 2 3 4 5 6 7 8 9 . τ ∈ Sym(S).

Prove that τ Gτ −1 = {σ ∈ Sym(S) | σ = τ γτ for some γ ∈ G} is a group of permutations. and let X be a subset of S. Let G = {σ ∈ Sym(S) | σ(X) ⊂ X}. PERMUTATIONS 18.3. Prove that G is a group of permutations. but not 11 or 13. 11 19. 20. for the set S. Let τ be a ﬁxed permutation in Sym(S).2. with G ⊆ Sym(S). 12. Let G be a group of permutations. . and 14. Let S be a set. Show that S10 has elements of order 10.

and use this to ﬁnd its order and its inverse. let α = (1. 6). 2. is a one-to-one and onto function. ﬁnd the inverse function φ−1 . 2). Show that f is a one-to-one function. Deﬁne f : R → R by f (x) = x3 + 3xz − 5. 3). 4. Hint: Use the derivative of f to show that f is a strictly increasing function. . 9). 6. deﬁne x ∼ y if x − y is an integer. Deﬁne the function φ : Z× → Z× by φ(x) = x−1 . In S10 . Compute φα . let α = (1. Show that φα : Sn → Sn deﬁned by φα (σ) = ασα−1 . write σ as a product of disjoint cycles. for all x ∈ R. for all σ ∈ Sn . β = (1. FUNCTIONS Review Problems 1. Is φ one to 17 17 17 one? Is φ onto? If possible. On the set Q of rational numbers. 5.12 CHAPTER 2. (b) In S3 . for all x ∈ R. 3. For σ = αβγ. 3. describe the equivalence relation on R that is determined by f . and γ = (1. for all x ∈ Z× . 5. For the function f : R → R deﬁned by f (x) = x2 . 7. (a) Let α be a ﬁxed element of Sn . 2. Show that ∼ is an equivalence relation. 2. Is σ even or odd? 5.

then a · x = a · y. so long as everyone knows the convention that we are using. for 13 . and ﬁnite group. especially the quantiﬁers (“for all”. This still works for the operation in a group. which must be stated in exactly the right order. subtraction. very soon we will just write ab instead of a ∗ b. which just represent various numbers. you must pay careful attention to each part. and the equation will still hold. and x = y. and has inverses for each of its elements. multiplication.Chapter 3 GROUPS The study of groups. But the operations are still the usual ones for numbers. The step from arithmetic to algebra involves starting to use variables. “for each”. is usually thought of as the real beginning of abstract algebra. “there exists”). addition. which we begin in this chapter. to show that ∗ might represent ordinary addition or multiplication. or possibly operations on matrices or functions. These deﬁnitions provide the language you will be working with. and you simply must know this language. From one point of view. even if it represents something new. At ﬁrst we will use ∗ or · to represent an operation. or maybe even something quite far from your experience. The step from algebra to abstract algebra involves letting the operation act like a variable. has an identity element. 3.3. abelian group. 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. and division. One of the things we try to do with notation is to make it look familiar. For example. since if x and y are elements of a group G.1. The precise statement is given in Deﬁnition 3. one of the rules you are used to says that you can multiply both sides of an equation by the same value. the axioms for a group give us just what we need to work with equations involving the operation in the group.1 Deﬁnition of a Group This section contains these deﬁnitions: binary operation. group.

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

Let F be the set of all fractional linear transformations of the complex plane. Prove that these weaker conditions (given only on the left) still imply that G is a group. If the subset is a group in its own right. (c) Show that 2 is the identity element for the operation ∗. Any group of n × n matrices (with entries in R) is a subgroup of GLn (R). az + b . If you only look at the operation of addition in a vector space. and suppose that a and b are any elements of G. SUBGROUPS 26. then ba = ab. 15 15 27. c. 32. for any positive integer n. Show that if (ab)2 = a2 b2 . 28. then it is called a subgroup. Show that F forms a group under composition of functions. Show that (aba−1 )n = abn a−1 . y ∈ G. Let G be a group. you are right. Write out the multiplication table for Z× . (a) Show that the operation ∗ is closed on G. b. . In Deﬁnition 3. 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.2. F is the set of all functions f (z) : C → C of the form f (z) = cz + d where the coeﬃcients a. deﬁne x ∗ y = xy − x − y + 2.3. For x. 3. replace condition (iii) with the condition that there exists e ∈ G such that e · a = a for all a ∈ G. (d) Show that for element a ∈ G there exists an inverse a−1 ∈ G. (b) Show that the associative law holds for ∗. 31. Now might be a good time to pick up your linear algebra text and review vector spaces and subspaces. and any subspace is automatically a subgroup. 30. and replace condition (iv) with the condition that for each a ∈ G there exists a ∈ G with a · a = e. it forms an abelian group. Let G be a group. For instance. 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.3 of the text. Let G = {x ∈ R | x > 1} be the set of all real numbers greater than 1.2 Subgroups Many times a group is deﬁned by looking at a subset of a known group. any group of permutations is a subgroup of Sym(S). using the same operation as the larger set. That is. for some set S.1. 29. If the idea of a subgroup reminds you of studying subspaces in your linear algebra course. and suppose that a and b are any elements of G. d are integers with ad − bc = 1.

ﬁnd the order of the √ √ √ 2 2 2 2 elements − + i and − − i. Show that N = {g ∈ G | g = an for some a ∈ G} is a subgroup of G. In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. 3)(5. and why they must form a subgroup. 27. show that 21 H = {[x]21 | x ≡ 1 (mod 3)} are subgroups of G. If the group is ﬁnite. GROUPS Lagrange’s theorem is very important. and K = {[x]21 | x ≡ 1 (mod 7)} . In G = Z× . (a) Find the cyclic subgroup of S7 generated by the element (1. p (b) Use part (a) to prove that n must be a power of 2. Let G be an abelian group. 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. then you only need to look at positive powers. 26. 28.16 CHAPTER 3. SOLVED PROBLEMS: §3. (b) Find a subgroup of S7 that contains 12 elements. show that cos θ − sin θ H= θ∈R sin θ cos θ is a subgroup of G. 24 24. This is a useful fact to know when you are looking for subgroups in a given group. Suppose that p is a prime number of the form p = 2n + 1. and let n be a ﬁxed positive integer. since in that case the inverse a−1 of any element can be expressed in the form an . In Z× . It is also important to remember that every element a in a group deﬁnes a subgroup a . 2 2 2 2 30. one that is cyclic and one that is not 20 cyclic. 7). This subgroup has o(a) elements. (a) Show that in Z× the order of [2]p is 2n. where o(a) is the order of a.2 23. for some n > 0. 2. ﬁnd two subgroups of order 4. consisting of all powers (positive and negative) of the element. 29. Find all cyclic subgroups of Z× . You do not have to list all of the elements if you can explain why there must be 12. In the multiplicative group C× of √ complex numbers. 25.

using ordered pairs.3. Note: Exercise 3. 17. [15]18 ) in the group Z12 × Z18 . This includes matrices with entries in the ﬁeld Zp . Show that H and K are subgroups of G. for any prime number p.3 Constructing Examples The most important result in this section is Proposition 3. CONSTRUCTING EXAMPLES 31.3.7. Find two groups G1 and G2 whose direct product G1 × G2 has a subgroup that is not of the form H1 × H2 . and list all of the generators for the group. and this allows us to construct very interesting ﬁnite groups as subgroups of GLn (Zp ). K= a b c d d = a. in which we can allow the entries in the matrix to come from any ﬁeld. ad − bc = 0 17 Show that K is a subgroup of GL2 (R). Compute the centralizer in GL2 (R) of the matrix 2 1 1 1 .3. 32. Find the order of the element ([9]12 . The second construction in this section is the direct product. which takes two known groups and constructs a new one. Show that Z5 × Z3 is a cyclic group.14 in the text deﬁnes the centralizer of an element a of the group G to be C(a) = {x ∈ G | xa = ax}.2. This generalizes the construction of n-dimensional vector spaces (that case is much simpler since every entry comes from the same set). c = −2b. 19. Show that if p is a prime number. 3. where the entry in the ith component comes from a group Gi . 20. which shows that the set of all invertible n × n matrices forms a group. 18.3 16. SOLVED PROBLEMS: §3. In the group G = Z× . and ﬁnd the subgroup HK. then the order of the general linear group GLn (Zp ) is (pn − 1)(pn − p) · · · (pn − pn−1 ). for subgroups H1 ⊆ G1 and H2 ⊆ G2 . . and ntuples are multiplied component-by-component. Let K be the following subset of GL2 (R). This can be extended to n-tuples. let H = {[x] | x ≡ 1 (mod 4)} and K = {[y] | y ≡ 36 1 (mod 9)}.

18  CHAPTER 3. Let H be the following subset of the group G = GL2 (Z5 ). 25. 2 0 2 1 1 2 1 1 . b ∈ G1 . b ∈ Z5 . then BA = A−1 B. since the inverse function φ−1 also respects the multiplication of the two groups. Let A = 1 1 −1 0 and B = . Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . (c) Show that every element of H can be written uniquely in the form Ai B j . where 0 ≤ i < 5 and 0 ≤ j < 2. The function φ can be thought of as simply renaming the elements of G1 . and show that C(A) ∩ C(B) = Z(G). where Z(G) is the center of G. 3. Compute the centralizer in GL2 (Z3 ) of the matrix 24. GROUPS  i 0 0 0  in the group GL3 (C). Find the centralizers C(A) and 0 1 0 1 C(B). H= m b 0 1 ∈ GL2 (Z5 ) m. b ∈ G1 makes certain that multiplication can be done in either group and the transferred to the other. since it is one-to-one and onto. Compute the centralizer in GL2 (Z3 ) of the matrix . Find the order of the element A =  0 −1 0 0 −i 22. 21.4 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. m = ±1 (a) Show that H is a subgroup of G with 10 elements. the existence of an isomorphism guarantees that there is a way to set up a correspondence between the elements of the groups in such a way that the group multiplication tables will look exactly the same. In terms of the respective group multiplication tables for G1 and G2 . (b) Show that if we let A = 1 1 0 1 and B = −1 0 0 1 . . The condition that φ(ab) = φ(a)φ(b) for all a. 23.

23. and then prove that φ is a group isomorphism. because there are inﬁnitely many possibilities. [m]2 ) = ([n]10 . ISOMORPHISMS 19 From an algebraic perspective. To decide on the function to use. then so is G2 .4. G2 . If G1 is abelian.5) that the answer to this particular question is the list Z8 . you should actually produce an isomorphism φ : G1 → G2 . we can show (in Section 7. 24. you probably need to see some similarity between the group operations. 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. . Prove that the group Z× × Z× is isomorphic to the group Z6 × Z10 . Each time you meet a new property of groups. Prove that φ is a group isomorphism. the two groups must have exactly the same number of elements of order n. Z4 × Z2 . and suppose that θ1 : G1 → H1 and θ2 : G2 → H2 are group isomorphisms. In some ways it is harder to show that two groups are not isomorphic. Deﬁne φ : Z30 × Z2 → Z10 × Z6 by φ([n]30 . if G1 is cyclic. up to isomorphism. then the question becomes manageable. Show that φ is a group isomorphism. x2 ) = (θ1 (x1 ). Furthermore. H1 . for all ([n]30 . θ2 (x2 )). x2 ) ∈ G1 × G2 . you should ask whether it is preserved by any isomorphism. for each positive integer n. 26. [m]2 ) ∈ Z30 × Z2 . for all x ∈ R. then so is G2 . In fact. and let H be a subgroup of G. then you can decide that two groups are not isomorphic (provided that the property would have been transferred by any isomorphism). Z2 × Z2 × Z2 . Let φ : R× → R× be deﬁned by φ(x) = x3 .4 21. we should think of isomorphic groups as being essentially the same. Deﬁne φ : G1 × G2 → H1 × H2 by φ(x1 . Show that Z× is isomorphic to Z16 . Suppose that G1 and G2 are isomorphic groups. To show that two groups G1 and G2 are isomorphic. 17 22. then the subset aHa−1 = {g ∈ G | g = aha−1 for some h ∈ H} is a subgroup of G that is isomorphic to H. H2 be groups. SOLVED PROBLEMS: §3. If you can show that one group has a property that the other one does not have. First prove that φ is a well-deﬁned function. Let G1 . The problem of ﬁnding all abelian groups of order 8 is impossible to solve. for all (x1 . Let G be a group. In this situation we would usually say that we have found all abelian groups of order 8.3. [4n + 3m]6 ). Prove that if a is any element of G. 7 11 25.

by looking at the group Z of integers. x2 ∈ S. 3. b ∈ Z3 . In this section.20 CHAPTER 3. Theorem 3. a22 = 1 is a subgroup of G.2 shows that every cyclic group is isomorphic to one of these concrete examples. Let H be the following subgroup of group G = GL2 (Z3 ). (b) Show that H is isomorphic to the group R of all real numbers. showing that they are in one-to-one correspondence with the . Show that for any prime number p. and hk = kh for all h ∈ H and k ∈ K. Prove that S is a group under this operation. for all x1 . which describes the subgroups of Zn . and the related groups Zn . 30. so all of the information about cyclic groups is already contained in these basic examples. then there are subgroups H and K in G such that H ∩ K = {e}. G1 . HK = G. show that H= a11 a21 a12 a22 ∈ GL2 (R) a11 = 1.5. Show that G is not isomorphic to the direct product R× × R. (a) In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. Prove that if G is isomorphic to G1 × G2 . and let S be any set for which there exists a one-toone and onto function φ : G → S. m = 0 Show that H is isomorphic to the symmetric group S3 . a21 = 0. We discovered that each of these groups is generated by a single element. Deﬁne an operation on S by setting x1 · x2 = φ(φ−1 (x1 )φ−1 (x2 )). You should pay particular attention to Proposition 3. G2 be groups. Let G be a group.5 Cyclic Groups We began our study of abstract algebra very concretely. Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . p p 29.5. 31. Let G.3. under addition. the subgroup of diagonal matrices in GL2 (Zp ) is isomorphic to Z× × Z× . 28. and that φ is actually a group isomorphism. 32. GROUPS 27. H= m b 0 1 ∈ GL2 (Z3 ) m. and this motivated the deﬁnition of an abstract cyclic group.

then the subgroups are “linearly ordered” in the sense that given any two subgroups.) SOLVED PROBLEMS: §3. pq 3. Find all cyclic subgroups of the quaternion group. then Z× is not a cyclic group. the subgroups of order n just get lost inside the larger symmetric group. Z× .6 Permutation Groups As with the previous section. 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 . 21 29. since every ﬁnite group of order n is isomorphic to a subgroup of the symmetric group Sn . In n is a prime power. ﬁnd the order of [24]30 .6. and Z× are isomorphic to each other. 25. 3. because as n gets larger and larger. 2. Find all generators of the cyclic group Z28 . 26. Use this information to show that the quaternion group cannot be isomorphic to the subgroup of S4 generated by (1. 3). then the direct product G1 × G2 is a cyclic group. 24.4 we will prove that every ﬁnite abelian group is isomorphic to a direct product of cyclic groups of prime power order. These cyclic groups have a particularly simple structure. ﬁnd the order of the subgroup [18]30 . one is a subset of the other. In Z30 . Show that any cyclic group of even order has exactly one element of order 2. 4) and (1. 28. and form the basic building blocks for all ﬁnite abelian groups. respectively. Give the lattice diagram of subgroups of Z100 . 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.3. Prove that if p and q are diﬀerent odd primes. 30. This does imply. however. (In Theorem 7. Is Z4 × Z15 isomorphic to Z6 × Z10 ? 23.5 20. 27. Show that the three groups Z6 . That isn’t as impressive as it sounds at ﬁrst. which has order n!. Use the the result in Problem 27 to show that the multiplicative groups Z× 15 and Z× are not cyclic groups.5. PERMUTATION GROUPS 21 positive divisors of n. 9 18 21. Is Z4 × Z10 isomorphic to Z2 × Z20 ? 22. Prove that if G1 and G2 are groups of order 7 and 11.

with o(a) = n. 26. In the dihedral group Dn = {ai bj | 0 ≤ i < n.22 CHAPTER 3.8. GROUPS group. The abstract deﬁnition of a group is useful simply because it can be more easily applied to a wide variety of situation. ﬁnd the subgroup H generated by (1. 3. 28. The most important result in this section is Theorem 3. which is a preliminary form of the Fundamental Homomorphism Theorem. 2. in terms of the rigid motions of an n-gon. It is easy to prove that the image φ(G1 ) is . We can write Dn = {ai bj | 0 ≤ i < n. ﬁnd the centralizer of a. Find the centralizer of (1. o(b) = 2.4 we introduced the concept of an isomorphism. for σ = (1. SOLVED PROBLEMS: §3. 0 ≤ j < 2} with o(a) = n. and ba = a−1 b. 3) and (1. 0 ≤ j < 2} with o(a) = n. and studied in detail what it means for two groups to be isomorphic. In doing computations in Dn it is useful to have at hand the formula bai = an−i b. we start with a group homomorphism φ : G1 → G2 . 0 ≤ j < 2} with o(a) = n. 2. o(b) = 2. 24. In S4 .6 22. in S4 . ﬁnd the corresponding subgroup σHσ −1 . 27. 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.8. 2). in fact. You should make every eﬀort to get to know the dihedral groups Dn . 0 ≤ j < 2. (The full statement is given in Theorem 3. and ba = a−1 b. For the subgroup H of S4 deﬁned in the previous problem. and in A4 . 23. homomorphisms provide the way to relate one group to another. In the dihedral group Dn = {ai bj | 0 ≤ i < n. 25. 3) in S3 .8.7. In the dihedral group Dn = {ai bj | 0 ≤ i < n. and ba = a−1 b. after we develop the concepts of cosets and factor groups. There are many important examples of group homomorphisms that are not isomorphisms. 4). shown in the ﬁrst of the solved problems given below. In this section we look at functions that respect the group operations but may not be one-to-one and onto. They have a concrete representation. and.7 Homomorphisms In Section 3. show that each element of the form ai b has order 2. and ba = a−1 b} . show that bai = an−i b.) In this formulation of the Fundamental Homomorphism Theorem. Show that each element in A4 can be written as a product of 3-cycles. o(b) = 2. for all 0 ≤ i < n. o(b) = 2.

with generator [3]7 .7.7.7. so that although the homomorphism may not be an isomorphism between G1 and G2 . and apply the fundamental 36 homomorphism theorem. 7 (b) Show that Z× is cyclic. 17 (c) Completely determine all group homomorphisms from Z× into Z× . φ2 (x)). Find all group homomorphisms from Z4 into Z10 . and show how elements of the image correspond to cosets of the kernel. (a) Show that φ is a well-deﬁned group homomorphism. 18. since for any group homomorphism φ : G1 → G2 it describes the connections between subgroups of G1 and subgroups of G2 . Let p and q be diﬀerent odd primes. and for this formula ﬁnd the kernel and image.5 are important. such that m is a divisor of n. because they give a complete description of all group homomorphisms between two cyclic groups. (b) Find the kernel and image of φ. 23. 19. Let G. p q . SOLVED PROBLEMS: §3.7 17. it does deﬁne an isomorphism between a subgroup of G2 and what we call a factor group of G1 . Deﬁne φ : Z4 × Z6 → Z4 × Z3 by φ(x. y). Examples 3.7. we can deﬁne a group operation on the equivalence classes of ∼. 21.4 and 3. for all x ∈ G. for all [x]n ∈ Z× . using the operation in G1 . G1 . The function φ has an equivalence relation associated with it. HOMOMORPHISMS 23 a subgroup of G2 . (a) Find the formulas for all group homomorphisms from Z18 into Z30 . Just as in Z.6 is also useful. Let n and m be positive integers. 24. Let φ1 : G → G1 and φ2 : G → G2 be group homomorphisms. where we let a ∼ b if φ(a) = φ(b). and apply the fundamental homomorphism theorem. 17 7 20. Prove that φ : G → G1 × G2 deﬁned by φ(x) = (φ1 (x). for a. and G2 be groups. (a) Show that Z× is cyclic.7. (b) Choose one of the nonzero formulas in part (a).3. 22. is a well-deﬁned group homomorphism. Show that φ : Z× → Z× deﬁned by φ([x]n ) = [x]m . where we use the equivalence relation deﬁned by congruence modulo n. Proposition 3. Then Theorem 3. for 36 12 all [x]36 ∈ Z× . ﬁnd the kernel and image of φ. Prove that Z× is isomorphic to the direct pq product Z× × Z× . with generator [3]17 . For the group homomorphism φ : Z× → Z× deﬁned by φ([x]36 ) = [x]12 . y) = (x + 2y.8 shows that this group is isomorphic to φ(G1 ). b ∈ G1 . is a well-deﬁned n m n group homomorphism.

and let K = {(0. The moral of this story is that if you deﬁne a function on G rather than G/N . with generators a of order n and b of order 2. if your function is deﬁned on cosets. and then in Section 1.24 CHAPTER 3. (b) You may assume that any abelian group of order 12 is isomorphic to either Z12 or Z6 × Z2 . and Factor Groups The notion of a factor group is one of the most important concepts in abstract algebra. Normal Subgroups. To construct a factor group. the most convenient way is use a formula deﬁned on representatives of the cosets of N . where at ﬁrst we thought of congruence classes as inﬁnite sets of integers. where we have a ≡ b (mod n) if and only if a − b ∈ nZ. you must prove that your formula actually deﬁnes a function. Once again.8 Cosets. GROUPS 3. it is important to let the theorem do its job. The easiest way to do this is to just deﬁne a homomorphism φ from G to G1 . Which answer is correct for G/H? For G/K? 29. 0)}. On the other hand. That is. we start with a normal subgroup and the equivalence classes it determines. Quite often we need to show that a factor group G/N that we have constructed is isomorphic to another group G1 . you ordinarily don’t need to worry that it is well-deﬁned.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. 0). since all subgroups of an abelian group are normal. The only complication is that the equivalence relation respects the operation in G only when the subgroup is a normal subgroup. Let the dihedral group Dn be given via generators and relations. let H = {(0. 0). List the cosets of 7 in Z× . Of course. showing that there exists a well-deﬁned isomorphism between G/N and G1 . (0. But then you must be careful to prove that the formula you are using does not depend on the particular choice of a representative. SOLVED PROBLEMS: §3. so that it does as much of the hard work as possible. list all cosets of K. making sure that N is the kernel of φ. satisfying ba = a−1 b. it can be much trickier to prove that it is one-to-one than to simply compute the kernel of a homomorphism deﬁned on G. The key idea is to begin thinking of equivalence classes as elements in their own right. in addition to proving that it is onto and respects the operations in the two groups. Let G = Z6 × Z4 . That is what we did in Chapter 1. Is the factor group Z× / 7 cyclic? 16 16 28. 2)}. Then you must prove that your function is one-to-one. (3. (a) List all cosets of H. This construction parallels the construction of Zn from Z. then the Fundamental Theorem does the rest of the work. if you deﬁne a function on the cosets of G/N . In actually using the Fundamental Homomorphism Theorem. If you prove that φ maps G onto G1 . in an abelian group we can use any subgroup.8 27. .

Which is correct? 32. (a) Show that N is a normal subgroup of G. Let G = D6 and let N be the subgroup a3 = {e. 34. (c) List all left cosets and all right cosets of b 30. if there exists g ∈ G such that b = gag −1 . (a) Prove that HN is a subgroup of G. For a. 33. and let N = {e. 25 (b) You may assume that G/N is isomorphic to either Z6 or S3 . Let G be the dihedral group D12 . then HN/N is isomorphic to H. COSETS. Find the conjugacy classes of D4 . (b) Prove that N is a normal subgroup of HN . (c) Prove that if H ∩ N = {e}. written b ∼ a.3. Let G be a group. (b) Is G/N abelian? 31. b ∈ G we say that b is conjugate to a. a9 }. a6 . NORMAL SUBGROUPS. (a) Let G be a group. (a) Prove that N is a normal subgroup of G. a3 . and list all cosets of N . Show that ∼ is an equivalence relation on G. (b) Show that any element of the form ai b has order 2.8. (b) Show that a subgroup N of G is normal in G if and only if N is a union of conjugacy classes. and let N and H be subgroups of G such that N is normal in G. . a3 } of G. AND FACTOR GROUPS (a) Show that bai = a−i b for all i with 1 ≤ i < n. The equivalence classes of ∼ are called the conjugacy classes of G.

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

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

11. 12.28 CHAPTER 4. and use your answers to ﬁnd [x2 + x]−1 . 8. Factor x4 + x + 1 over Z2 [x]/ x4 + x + 1 . 7. Are the following polynomials irreducible over Q? (a) 3x5 + 18x2 + 24x + 6 (b) 7x3 + 12x2 + 3x + 45 (c) 2x10 + 25x3 + 10x2 − 30 6. 10. (a) Show that x2 + 1 is irreducible over Z3 . ﬁnd [x]−1 and [x + 1]−1 . ﬁnd the multiplicative inverse of [x + 1]. POLYNOMIALS 5. 14. In Z5 [x]/ x3 + x + 1 . Express 2x3 + x2 + 2x + 2 as a product of polynomials irreducible over Z5 . 9. (b) Show that x3 + 2x2 + 3 is irreducible over Z5 . (a) Express x4 + x as a product of polynomials irreducible over Z5 . Factor x5 − 10x4 + 24x3 + 9x2 − 33x − 12 over Q. Construct an example of a ﬁeld with 343 = 73 elements. In Z2 [x]/ x3 + x + 1 . 13. show that [x + 1] is a generator. but [x] is not. . (b) List the elements of the ﬁeld F = Z3 [x]/ x2 + 1 . Find the multiplicative inverse of [x2 + x + 1] (a) in Q[x]/ x3 − 2 . Factor x5 − 2x4 − 2x3 + 12x2 − 15x − 2 over Q. (c) In the multiplicative group of nonzero elements of F . (b) in Z3 [x]/ x3 + 2x2 + x + 1 . 15.

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

For the element a = (0. (a) Find all ideals of R. (b) Find the units of R. COMMUTATIVE RINGS 5. = (c) Is I a prime ideal of R? Hint: If you use the fundamental homomorphism theorem. 8. (c) Find the idempotent elements of R. Show that the rings R and S in the two previous problems are isomorphic as abelian groups. Give an example to show that the set of all zero divisors of a ring need not be an ideal of the ring. n ∈ Z} . Let S be the ring Z2 [x]/ x3 + x . (c) Find the idempotent elements of R. Find all maximal ideals. ﬁnd Ann(a) = {r ∈ R | ra = 0}. . then 1−e is also idempotent. (b) Find the units of R. what is the largest possible order 180 of an element? 6. (b) Show that R/I ∼ Z2 [x]/ x2 + 1 . (b) Show that if e is idempotent. 7. prove that I is not maximal. 9. (a) Show that I is an ideal of R. Let Z[i] be the subring of the ﬁeld of complex numbers given by Z[i] = {m + ni ∈ C | m. 11. then R ∼ Re ⊕ R(1 − e). Let R be the ring Z2 [x]/ x4 + 1 . and let I be the set of all congruence classes in R of the form [f (x)(x2 + 1)]. and all prime ideals. of Z36 = Z/36Z. 15. 14. 13. but not as rings. = 12. (a) Show that if e is an idempotent element of R. Prove that I is a prime ideal.30 CHAPTER 5. Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients. 2) of the ring R = Z12 ⊕ Z8 . Let R be any commutative ring with identity 1. Show that Ann(a) is an ideal of R. (a) Find all ideals of S. 10. In the group Z× of units of the ring Z180 . Let R be the ring Z2 [x]/ x3 + 1 . you can do the ﬁrst two parts together.

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

COMMUTATIVE RINGS .32 CHAPTER 5.

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

FIELDS .34 CHAPTER 6.

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.

Adding and subtracting 20 makes it possible to get this term. is divisible by 6. We conclude that 4|n. you should read Appendix 4 in the text. for n = 1. The ﬁrst half of the proof is easier than the second. to divide n by 4. say n = 4q. and it is easier to state as the converse of the ﬁrst statement: if An = I.. A3 = A · A2 . then we can show that the statement must also be true for n = k + 1. and then it turns out that we can factor out 6. 30. Conversely. so r = 0 since A. In the second half of the proof. we need a statement that depends on an integer n. but we need to involve the expression 10k + 2 in some way. etc. We can easily factor a 10 out of 10k+1 . if An = I. Solution: To give a proof by induction. then An = A4q = (A4 )q = I q = I. (This “anchors” the induction argument. we can use the division algorithm to write n = 4q + r. 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 . 102. and A3 are not equal to I. . so we can prove the following statement: for each positive integer n. Then Ar = An−4q = An (A−4 )q = I · I q = I. Comment: If you are unsure about doing a proof by induction. . . Give a proof by induction to show that each number in the sequence 12. 10k+1 + 2 = (10)(10k ) + 20 − 20 + 2 = (10)(10k + 2) − 18 = (10)(6q) − (6)(3) = (6)(10q − 3) . Let’s start by assuming that 10k + 2 is divisible by 6. We can write the numbers in the given sequence in the form 10n + 2. A2 . then 4|n. since it just takes a computation. The next step is to prove that if we assume that the statement is true for n = k. and then show that the remainder has to be 0. A4 = A · A3 . with 0 ≤ r < 4. 10002. The ﬁrst step is to check that the statement is true for n = 1. the integer 10n + 2 is divisible by 6.) Clearly 12 is divisible by 6. for some q ∈ Z. and then look at the expression when n = k + 1.. . to get 10k+1 + 2 = (10)(10k ) + 2.     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. . say 10k + 2 = 6q.38 CHAPTER 1 SOLUTIONS We also need to show that An = I only when 4|n. if An = I then we will use the division algorithm. 2. 1002. .

(b) Use the prime factorizations of 1492 and 1776 to ﬁnd gcd(1776. (a) Use the Euclidean algorithm to ﬁnd gcd(1776. 1089). 1089) = 1. 72 = 68 · 1 + 4. we need to use one axis for each prime. 185 = 164 · 1 + 21. Proposition 1.2 SOLUTIONS 23. 24. 1089). These numbers go along one axis of the rectangular diagram. Thus 21 = 17 · 1 + 4. and 2. 1089 = 185 · 5 + 164. 164 = 21 · 7 + 17. Solution: The prime factorizations are 250 = 2 · 53 and 484 = 22 · 112 . gcd(1274. dividing 250 by 5 produces 50. This completes the induction. 1492). In each diagram. 10. 1492) = 22 .CHAPTER 1 SOLUTIONS 39 We have now shown that if 10k + 2 is divisible by 6. 68 = 4 · 17. 250 125 50 242 484 44 25 10 121 22 4 5 2 11 2 1 1 . (b) Use the prime factorizations of 1274 and 1089 to ﬁnd gcd(1274. Solution: Since 1274 = 2 · 72 · 13 and 1089 = 32 · 112 .2. to give the factors along the corresponding axis. For example. 1492). Give the lattice diagram of all divisors of 250. then 10k+1 + 2 is divisible by 6. 1492 = 284 · 5 + 72. Solution: We have 1274 = 1089 · 1 + 185. Then we can just divide (successively) by the prime. we see that 1274 and 1089 are relatively prime. (a) Use the Euclidean algorithm to ﬁnd gcd(1274. Solution: Since 1776 = 24 · 3 · 37 and 1492 = 22 · 373. 1. Do the same for 484. in succession. Solution: We have 1776 = 1492 · 1 + 284. Thus gcd(1776. 1492) = 4.9 shows that gcd(1776. 17 = 4 · 4 + 1. 25. 284 = 72 · 3 + 68.

Listing the solutions as ordered pairs (x. −16). Solution: If we had a product. That motivates one possible method of solution. which states that if gcd(2m − 1. Then a similar argument yields gcd(a2 . and so gcd(3n − 1. b2 ) = 1 if and only if gcd(a. For each of these values there is a corresponding value for y. you don’t have to explain how you found them. n) = 1. then gcd(m. since the complementary factor must be equal to y − 3. Let m and n be positive integers. 2n − 1) = 1 if and only if gcd(m. assuming that gcd(m. n − 1) = 1. n) = 1 and showing that this forces gcd(2m − 1. (−3. b ∈ Z with am + bn = 1. then gcd(2m − 1. (2)(3n − 1) + (−3)(2n − 1) = 1. Then we will prove the converse. prove that gcd(a. To prove the converse. we could use the prime factorization theorem. 2n − 1) = 1. the only way it can be factored is to have 1 · 19 = 19 or (−1) · (−19) = 19. 22).40 CHAPTER 1 SOLUTIONS 26. 2n − 1) = 1. we will prove that if gcd(m. for all integers n > 1. Prove that n − 1 and 2n − 1 are relatively prime. Is the same true for 2n − 1 and 3n − 1? Solution: We can write (1)(2n − 1) + (−2)(n − 1) = 1. Before beginning the proof. x + 2 = 19. For positive integers a. which proves that gcd(2n − 1. Find all integer solutions of the equation xy + 2y − 3x = 25. b2 ) = 1 if and only if gcd(a. 2n − 1) = 1. Solution: Proposition 1. Therefore we have 4 possibilities: x + 2 = 1. 29.2. 4). b2 ) = 1. 28. 2). x + 2 = −1. or x + 2 = −19. we recall that the following identity holds for all values of x: xk − 1 = (x − 1)(xk−1 + xk−2 + · · · + x + 1). we have the four solutions (−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. 2n − 1) = 1. Prove that gcd(2m − 1. 27. Solution: If gcd(m. y). b. First. n) = 1. (17. c) = 1. b2 ) = 1. n) = 1. Similarly. Comment: We need to do the proof in two parts. bc) = 1 if and only if gcd(a. Comment: Is this really a proof? Yes–producing the necessary linear combinations is enough. n) = 1. b) = 1 if and only if gcd(a2 . we will use a proof by contradiction.3 (d) states that gcd(a. b) = 1. then there exist a. and (−21. Substituting x = 2m and k = a in the identity given above shows that 2m − 1 . Using c = b gives gcd(a. b) = 1 and gcd(a.

The solution is x ≡ 11. Therefore gcd(2m − 1. Prove that gcd(2n2 + 4n − 3. we see that 7 is a factor of −14. 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. 2n − 1) = 1. 2n − 1) = gcd(2n − 1. 71. getting gcd(2n − 1. The numbers congruent to 1 modulo 15 are 16. which proves that gcd(2m − 1. Solution: We can use the Euclidean algorithm. for all integers n > 1. or 7x ≡ 2 (mod 15). so we multiply both sides of the congruence by −2 since (−2)(7) = −14 ≡ 1 (mod 15). so there is a solution since 6 is a factor of 12. 2n2 + 4n − 3) = gcd(2n2 + 4n − 3. 56. 1) = 1. −34. Thus we have −14x ≡ −4 (mod 15). n − 1) = gcd(n − 2. (Of course. We next use trial and error to look for the multiplicative inverse of 7 modulo 15. 41.CHAPTER 1 SOLUTIONS 41 is a factor of 2am − 1. 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. Solve the congruence 42x ≡ 12 (mod 90). 61. This reduces to 7x = 2 + 15q. The proof now involves what may look like a trick (but it is a useful one). say gcd(m. We have shown that gcd(2n2 + 6n − 4. q ∈ Z with m = dq and n = dp. 30.3 SOLUTIONS 26.. 31. for some s ∈ Z. or x ≡ 11 (mod 15). Equivalently. 26. 46. and this completes the proof. The next step is to divide 2n2 + 4n − 3 by 2n − 1. and this gives a quotient of n + 2 and a remainder of n − 1. Among these. 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. n) = d. etc. you could also continue with the Euclidean algorithm. for some t ∈ Z. Solution: We have gcd(42. n) = 1. The same argument shows that we can write 2bn − 1 = (2n − 1)(t). 2n2 + 6n − 4) = 1. If gcd(m. Solving the congruence 42x ≡ 12 (mod 90) is equivalent solving the equation 42x = 12 + 90q for integers x and q. −29. . and −14. 86 (mod 90). we obtain 7x ≡ 2 (mod 15) by dividing 42x ≡ 12 (mod 90) through by 6. etc. 90) = 6. 2n − 1) = 1.) 1. say 2am − 1 = (2m − 1)(s). 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. then there exist p.

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

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

1 −5 46 0 1 91 . [6]−1 = [11]13 . ﬁnd all nilpotent elements. and [6]13 and [−2]13 . Since [12]−1 = [−1]−1 = [−1]13 = [12]13 . [11]−1 = [11]20 . while 12. If ab ≡ −1 (mod n). and we see that [3]13 and [−4]13 are inverses. 31 | 651. and look at the various ways to factor them. n) = 1 if and only if gcd(n − a. 9. as are the pairs [4]13 and [−3]13 .4 SOLUTIONS 30. Using 40. and 3 · 5 = 15 ≡ 1 (mod 7). Using 14. 3379) = 31. 1 −5 46 −1 6 45 . as are [−a]n and [−b]n . and 40 are congruent to 1.44 CHAPTER 1 SOLUTIONS 1. 11. and 19. if possible (in Z× ). ﬁnd all idempotent elements. Find [3379]−1 . 2 · 4 = 8 ≡ 1 (mod 7). 501 501 Solution: We need to use the Euclidean algorithm. 13. 3. It is useful to list the integers with m with m ≡ ±1 (mod n). if possible (in Z× ). 34. the class [6]7 is its own inverse. 7. 31. Using 12. This helps to check your list. Comment: If ab ≡ 1 (mod n). They occur in pairs. 20 20 20 [13]−1 = [17]20 . and so (4061. 13 13 13 [5]−1 = [8]13 . [9]−1 = [9]20 . 501 33. 2 −11 −1 6 1 45 Thus [91]−1 = [−11]501 = [490]501 . we see that [5]13 and [8]13 are inverses. Find the multiplicative inverse of each nonzero element of Z13 . so [2]7 and [4]7 are inverses of each other. 1 −1 682 −4 5 651 1 0 . Furthermore. since gcd(a. 20 20 . Solution: Note that 14. 25. then [a]n and [−b]n are inverses. In Z20 : ﬁnd all units (list the multiplicative inverse of each). 32. n) = 1. Solution: Since 6 ≡ −1 (mod 7). 31 651 At the next step. 17. We have [3]−1 = [7]20 . 0 1 4061 3379 5 −6 −4 5 . and [3]7 and [5]7 are inverses of each other. [4]−1 = [10]13 . 27. 1 −1 682 0 1 3379 . and 39 are congruent to −1. here is the list of inverses: [2]−1 = [7]13 . as are [−a]n and [b]n . and [19]−1 = [19]20 . 1 0 501 0 1 91 . [3]−1 = [9]13 . this 13 13 13 13 takes care of all of the nonzero elements of Z13 . then [a]n and [b]n are inverses. Finally. Solution: The units of Z20 are the equivalence classes represented by 1. 4061 4061 Solution: The inverse does not exist. Find [91]−1 . we see that [2]13 and [7]13 are inverses. Comment: We know that Zn has ϕ(n) units. Find the multiplicative inverse of each nonzero element of Z7 .

then any solution to the congruences x ≡ 1 (mod b) and x ≡ 0 (mod c) will be idempotent modulo n. [−4]3 = [−4][16] = [6].4. c) = 1. Show that Z× is cyclic. Since [2]8 = [−1]2 = [1]. and the powers we have listed represent the only possible values of [8]i . 36. it follows that [2] has multiplicative order 8. They are [0]20 and [10]20 . Then [8]i−m = [−4]n−j . The nilpotent elements of Z20 can be found by using trial and error. we need to ﬁnd an element whose 17 multiplicative order is 16. Solution: The units of Z24 are the equivalence classes represented by 1. 11. [5]20 . so the next possibility to check is 8. and [−4]6 = [−4][−9] = [1]. They are [0]20 .13 of the text: if n = bc. ﬁnd all nilpotent elements. . [18]24 . If you want a more systematic approach. [1]20 . [12]24 . with gcd(b. this forces i = m and n = j. 7. 37. We have [3]2 = [9]. [2]3 = [8]. [−4]4 = [−4][6] = [11]. and 23. and we are done. so the multiplicative order of [8] is 4. and [3]8 = [16] = [−1]. ﬁnd all idempotent elements. and [16]20 . [6]24 . [3]4 = [81] = [−4]. We next try [3]. 35. [16]24 . Solution: We ﬁrst compute the powers of [8]: [8]2 = [−6]. 5. The only divisor of 16 that is left is 16 itself. The idempotent elements are [0]24 . [9]24 . for some 0 ≤ i < 4 and 0 ≤ j < 6 and 0 ≤ m < 4 and 0 ≤ n < 6. We have [2]2 = [4]. but there is no known algorithm for p actually ﬁnding the one element whose powers cover all of Z× . 17 Comment: To show that Z× is cyclic. In Z24 : ﬁnd all units (list the multiplicative inverse of each). There are 24 possible products of the form [8]i [−4]j . and [2]4 = [16] = [−1]. [1]24 . The solution just uses trial and error. so [3] does in fact have multiplicative order 16. or by using Problem 1.40. for 0 ≤ i < 4 and 0 ≤ j < 6. Problem 39 shows that the multiplicative order of an element has to be a divisor of 16. [−4]5 = [−4][11] = [−9]. and since the only power of [8] that is equal to a power of [−4] is [1] (as shown by our computations). Show that Z× is not cyclic but that each element has the form [8]i [−4]j . j. Are these all diﬀerent? Suppose that [8]i [−4]j = [8]m [−4]n . It is known than if p is prime. and the nilpotent elements are [0]24 . and [8]4 = [−6]2 = [1]. 19. so the multiplicative order of [−4] is 6. We next compute the powers of [−4]: [−4]2 = [16]. and so each element is its own inverse. then Z× is cyclic. 17. p Solution: We begin by trying [2]. you can use a the hint in Exercise 1. For each of these numbers we have x2 ≡ 1 (mod 24).4. 35 35 35 for some positive integers i. [8]3 = [8][−6] = [−13].CHAPTER 1 SOLUTIONS 45 The idempotent elements of Z20 can be found by using trial and error. 13.

prove that ω n = 1 if and only if 3|n. Thus gcd(7605. Since ak ≡ 1 (mod n). For ω = − + 2 2 . Solution: Assume that k is the smallest positive integer for which ak ≡ 1 (mod n). n) = 1. 5733) = 117. with [a]k = [0]. If k is the smallest positive t 1 2 integer such that kβi ≥ αi for all i. for any integer n. and this completes the proof. √ 3 1 i. and 117 = (−3) · 7605 + 4 · 5733. 2. Finally. 49 −65 −3 4 1 0 0 117 0 1 7605 5733 . .11. showing that Z35 is not cyclic. Let n be a positive integer. and so the solution is [x] = [−3] or [x] = [2]. and let a ∈ Z with gcd(a.6 implies that either [x] + [3] = [0] or [x] − [2] = [0]. n) = 1.46 CHAPTER 1 SOLUTIONS We conclude that since there are 24 elements of the form [8]i [−4]j . If n = pα1 pα2 · · · pαt is the prime factorization of n. Corollary 1. Solution: Use the matrix form of the Euclidean algorithm: 1 −1 1872 0 1 5733 . then it must be a divisor of a. if [a]n is nilpotent. where 0 ≤ βj ≤ αj for all j. then n | ak . 39. n Conversely. Find gcd(7605. and express it as a linear combination of 7605 and 5733. then n | ak . and q ∈ Z. so no element of Z35 has multiplicative order 24. which shows that aϕ(n) ≡ 1 (mod n). 38. so each prime n divisor of n is a divisor of ak . where 0 ≤ r < k. we know that gcd(a. Thus ar = aϕ(n)−kq = aϕ(n) (ak )−q ≡ 1 (mod n). 11 Solution: We can factor [x]2 + [x] − [6] = ([x] + [3])([x] − [2]). But if a prime p is a divisor of ak . Prove that if k is the smallest positive integer for which ak ≡ 1 (mod n). 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 so [a]k = [0]k . 1 −1 1872 −3 4 117 . every element in Z35 must be of this form.4. and so we can apply Theorem 1.4. 40. We can use the division algorithm to write ϕ(n) = qk + r. Solve the equation [x]2 + [x]11 − [6]11 = [0]11 . ([8]i [−4]j )12 = ([8]4 )3i ([−4]6 )2j = [1]. then we must have a = t 1 2 pβ1 pβ2 · · · pβt d. then k | ϕ(n). Solution: First assume that each prime divisor of n is a divisor of a. so we must have r = 0 since r < k and k is the smallest positive integer with ak ≡ 1 (mod n). 5733). SOLUTIONS TO THE REVIEW PROBLEMS 1.

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

27 of the text that ϕ(2n) = ϕ(2)ϕ(n) = ϕ(n). and this does not change the computation. the prime 2 does not occur in its prime factorization.4. then ϕ(2n) = ϕ(n). the integers n and 2n are relatively prime.48 CHAPTER 1 SOLUTIONS 6.8 shows that to compute ϕ(2n) in terms of ϕ(n) 1 we need to add 2 · (1 − 2 ). Solution: Since n is odd. Second solution: Since n is odd. Show that if n > 1 is an odd integer. and so it follows from Exercise 1. . The formula in Proposition 1.4.

y1 ) and (x2 . 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. y) belongs to the subset of the plane.1. Explain why this agrees with Deﬁnition 2. the “horizontal line test” is exactly the same as the condition given in Deﬁnition 2.4. 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.1 SOLUTIONS 20.1. then there will be points (x1 .1. y2 ) line on its graph.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. Solution: If a horizontal line intersects the graph of the function more than once. the uniqueness part of the deﬁnition translates directly into the “vertical line test”.1.4. a function is one-to-one if f (x1 ) = f (x2 ) implies x1 = x2 . 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. If a vertical line intersects the curve in two distinct points.4.1.Chapter 2 Functions 2. then we cannot have y1 = y2 while x1 = x2 .1 to the given curve. Thus if we apply Deﬁnition 2. 21. if (x1 . According to Deﬁnition 2. y2 ) on the curve with x1 = x2 and y1 = y2 . y1 ) and (x2 . Explain why this agrees with Deﬁnition 2. According to Deﬁnition 2. Equivalently. In this context. then the points of intersection represent points (x1 . more than one 49 . y2 ) for which x1 = x2 but y1 = y2 .1.1. y1 ) and (x2 .

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

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

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

for all [x]12 ∈ Z12 . then x1 ≡ x2 (mod 12). 2)(4. 7. What is the order of σ? Is σ an even permutation? Compute σ −1 . [±2]2 = 12 12 12 [4]12 . [9]12 }. 2. 9. 7. then P AP −1 = B and QBQ−1 = C for some P. To compute the images of f we have [0]2 = [0]12 . Check that ∼ deﬁnes an equivalence relation. 2. and so A ∼ C. σ −1 . Thus 12 12 12 12 f (Z12 ) = {[0]12 . στ = (1. [±5]12 }. then P AP −1 = B for some invertible matrix P . 3). and so its order is 12 since lcm[3. 8. 3)(4. {[±2]12 . [±4]12 }. 8)(2. 5. τ −1 = (2. write each of these permutations as a 1 5 4 7 2 6 8 9 3 product of disjoint cycles: σ. τ στ −1 = (1. 8. 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 . 8). {[±3]12 }. 4. τ = (2. 1). [±3]2 = [9]12 .CHAPTER 2 SOLUTIONS 53 parallel to this one. where I is the n × n identity matrix. 2. 5). τ . 8. 7.3 SOLUTIONS 13. Solution: We have A ∼ A since IAI −1 = A. [±1]2 = [1]12 . 7. [±4]2 = [4]12 . 9). 3. τ στ −1 . 8. which shows that f ([x1 ]12 ) = 1 2 f ([x2 ]12 ). If A ∼ B and B ∼ C. 4)(7. 9. 7). 9. 5. 4] = 12. 19. 9. [6]12 }. 8. The corresponding equivalence classes determined by f are {[0]12 . since it can be expressed as the product of 6 transpositions. Thus the partition consists of the planes perpendicular to the vector (1. 14. 9)(3. τ −1 . τ σ = (1. On the set of all n × n matrices over R. 8. Deﬁne the formula f : Z12 → Z12 by f ([x]12 ) = [x]2 . 6. 2. {[±1]12 . and [6]2 = [0]12 . 20. . [1]12 . and so we get A = P −1 B(P −1 )−1 . It is an even permutation. 5. [4]12 . 4. 5. τ σ. write σ as a 7 5 6 9 2 4 8 1 3 product of disjoint cycles. 2. 9). 4. 4. [±5]2 = [1]12 . If A ∼ B. Solution: σ = (1. deﬁne A ∼ B if there exists an invertible matrix P such that P AP −1 = B. and so x2 ≡ x2 (mod 12). 3. σ −1 = (1. 12 Show that the formula f deﬁnes a function. Substituting gives Q(P AP −1 )Q−1 = (QP )A(QP )−1 = C. 9). στ . 8). στ σ −1 = (1. 5)(3. 4). 5)(3. 1. 4. We have σ −1 = (1. 5)(3. Q. Find the image of f and the set Z12 /f of equivalence classes determined by f . 7)(2. 6). στ σ −1 . Solution: The formula for f is well-deﬁned since if [x1 ]12 = [x2 ]12 . 7. For the permutation σ = 1 2 3 4 5 6 7 8 9 . 5)(3. Solution: We have σ = (1.

but not 11 or 13. 5). Show that S10 has elements of order 10. 9)(8. 7. 19. 7. 8)(2. 7) has order 12. 20. 17. 2. 6. 4. 2)(3. 5. What is the order of σ? Compute σ −1 . since 11 and 13 are prime. Let σ = (2. 12. 8. 7. 8. for any permutation σ ∈ Sn . Prove that τ Gτ −1 = {σ ∈ Sym(S) | σ = τ γτ for some γ ∈ G} is a group of permutations. 8. 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). 7. 8. so it has order 15 = lcm[5. 7. Prove that G is a group of permutations. 10). while the element (1. 9)(3. 2)(3. 7). 5. 11. and there are no cycles of that length in S10 . 6)(3. 3. 7. and σ −1 = (1. 3]. 9)(3. 9)(2. 8) = (1. 6. with G ⊆ Sym(S). 5. 5. 7. 6. 11. 8. 9)(1. 5. 6. then στ σ −1 has order m. 9) has order 14. 8. Prove that if τ ∈ Sn is a permutation with order m.54 CHAPTER 2 SOLUTIONS 15. 3)(4. Let G be a group of permutations. 5. 8. 6. compute the order of στ σ −1 . On the other hand. 9. For Solution: Since τ = (1. 18. since (στ σ −1 )k = (1) implies στ k σ −1 = (1). 7)(1. 7) has order 10. 2. 6. 6 7 8 9 8 9 10 1 10 3 11 5 . Let G = {σ ∈ Sym(S) | σ(X) ⊂ X}. and thus στ σ −1 has order 6. 10)(3. 3. 11. 16. for the set S. and (1. 4. 3. any element of order 11 or 13 would have to be a cycle. 6) ∈ S9 . . 10). 6. it has order 6. Let S be a set. Solution: We have σ = (1. Compute the order of τ = 1 2 3 4 5 7 2 11 4 6 σ = (3. so the cycle structure of στ σ −1 is the same as that of τ . and 14. 9. the order of στ σ −1 cannot be less than n. We have στ σ −1 = (3. Write σ as a product of disjoint cycles. On the other hand. Let τ be a ﬁxed permutation in Sym(S). )(6. Solution: Assume that τ ∈ Sn has order m. 4. Solution: The permutation (1. 5. and then τ k = σ −1 σ = (1). 4. 6. 5. 7). and let X be a subset of S.

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

56 CHAPTER 2 SOLUTIONS .

24. 2 Solution: If a and b are nonzero rational numbers. again making it impossible to deﬁne a group structure. so the cross product cannot be used to make the set of all vectors in R3 into a group. This means that the dot product does not even deﬁne a binary operation on the set of vectors in R3 . Show that G is a group under this multiplication. z1 )×(x2 . Is R3 a group under this multiplication? Solution: The cross product of the zero vector and any other vector is the zero vector. the cross product is deﬁned by (x1 . For vectors (x1 . z2 ) in R3 . for all a. y1 . This means that the operation could not have an identity element. Use the dot product to deﬁne a multiplication on R3 . b ∈ G. y2 . z1 ) and (x2 . deﬁne a new multiplication ab by a∗b = . 23.1 SOLUTIONS 22. x1 y2 − y1 x2 ). not a vector. 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. y1 . showing that the operation is closed on the set 2 G. Does this make R3 into a group? Solution: The dot product of two vectors is a scalar. z2 ) = (y1 z2 − z1 y2 . On the set G = Q× of nonzero rational numbers.Chapter 3 Groups 3. and so is . y2 . z1 x2 − x1 z2 . The operation is associative since a ∗ (b ∗ c) = a ∗ bc 2 57 = a bc 2 2 = a(bc) 4 . then ab is a nonzero ab rational number.

4. · 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 . [8]9 }. −4. with the elements in a slightly diﬀerent order. 26. 9 Solution: Z× = {[1]9 . and if a is nonzero. 25. 2. 4 4 The number 2 acts as the multiplicative identity. gives a diﬀerent picture of the group.58 and (a ∗ b) ∗ c = ab 2 ∗c= CHAPTER 3 SOLUTIONS ab 2 c 2 = (ab)c . [4]9 . 7. Write out the multiplication table for Z× . [11]15 . [7]9 . and the second table shows what happens when we arrange the elements in order. We will 15 write the elements as {1. [4]15 . then a is a nonzero rational number that serves as the multiplicative inverse of a. [5]9 . −2. as successive powers of 2. [2]9 . Write out the multiplication table for Z× . [8]15 . [7]15 . We will write m for [m]9 . [14]15 }. 15 Solution: Z× = {[1]15 . [2]15 . [13]15 . −1}. · 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. −7. 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} when listing the congruence classes in the group. Now assume that the result holds for n = k. 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 . so we can multiply both sides of the equation on the left by a−1 without destroying the equality. Multiplication in G is well-deﬁned. If we are to be precise about using the associative law. ±2. denoted by a−1 and b−1 . Let G be a group. 28.CHAPTER 3 SOLUTIONS 59 Comment: Notice how much easier it makes it to use the representatives {±1. Let G be a group. Since G is a group. Show that (aba−1 )n = abn a−1 . The associative law for multiplication essentially says that parentheses don’t matter. ±4. Show that if (ab)2 = a2 b2 . Solution: To give a careful proof we need to use induction. Expanding this equation gives us (ab)(ab) = a2 b2 . Using this induction hypothesis. since we have shown that if (ab)2 = a2 b2 . both a and b have inverses. for any positive integer n. Solution: Assume that a and b are elements of G for which (ab)2 = a2 b2 . b(ab)b−1 = (ab2 )b−1 (ba)(bb−1 ) = (ab)(bb−1 ) ba = ab This completes the proof. respectively. The statement for n = 1 is simply that aba−1 = aba−1 . and suppose that a and b are any elements of G. (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 . so we don’t really need to include all of the steps we showed before. which is certainly true. we have the following calculation. and suppose that a and b are any elements of G. 27. we have to include the following steps. then ba = ab. then ba = ab.

and that the associative law holds. since the two equations 1 ∗ x = 1 and (−1) ∗ x = −1 have no simultaneous solution in G. In summary.60 CHAPTER 3 SOLUTIONS Thus the statement holds for n = k + 1. Prove that these weaker conditions (given only on the left) still imply that G is a group. so by induction it holds for all values of n.3 of the text. This shows that e is a multiplicative identity for G. deﬁne the operation a ∗ b = |a|b. az + b . even though it has a left identity element and right inverses. and then (as above) we can cancel a to get a · a = e. and so G is a group under the given operation. Finally. replace condition (iii) with the condition that there exists e ∈ G such that e · a = a for all a ∈ G. 1/|a| is a right inverse for any a ∈ G. Then a ∗ b = 0 if a = 0 and b = 0. Thus the original condition (iv) holds. We also have the equation a · (a · a ) = (a · a) · a = e · a = a = a · e . 31. we can cancel a from the left of the above equation. We will ﬁrst show that a · e = a. There is no right identity element. 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. for all a ∈ G. for all a. Solution: Assume that the two replacement conditions hold. Note the e·e = e. but the equation x ∗ a = 1 has no solution for a = −1. Then a · (a · e) = (a · a) · e = e · e = e = a · a . F is the set of all functions f (z) : C → C of the form f (z) = cz + d . That is. 29.1. and so the original condition (iii) is satisﬁed. and since there exists an element a ∈ G with a · a = e. The number 1 is a left identity element. to get a · e = a. The previous exercise shows that in the deﬁnition of a group it is suﬃcient to require the existence of a left identity element and the existence of left inverses. so we have deﬁned a binary operation on G. 30. Solution: On the set G of nonzero real numbers. 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. which shows that a is indeed the multiplicative inverse of a. In Deﬁnition 3. Let a be an element in G with a · a = e. we have shown that G is not a group. and replace condition (iv) with the condition that for each a ∈ G there exists a ∈ G with a · a = e. so −1 has no left inverse. since 1 ∗ a = |1|a = a for all a ∈ G. Let F be the set of all fractional linear transformations of the complex plane. b ∈ G.

c.1. for matrices). and f2 (z) = . 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 . we can use the formula for the inverse of a 2 × 2 matrix with determinant 1 to ﬁnd an inverse function for .CHAPTER 3 SOLUTIONS 61 where the coeﬃcients a. 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. 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. c = 0. Finally. We only need to check that it can be written in the correct form. a1 z + b1 a2 z + b2 Let f1 (z) = . Composition of functions is always associative (compare Exercise 3. b. We need a better way to look at the entire situation. 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 . (c2 a1 + d2 c1 )z + (c2 b1 + d2 d1 ) You can see that verifying all of the axioms is going to be painful. Furthermore. Solution: We ﬁrst need to check that composition of functions deﬁnes a binary operation on F . 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 ) . respectively. d are integers with ad − bc = 1. and d = 1. so let’s look at the following matrix product. 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 1.2 in the text. All of this means that it is fair to use what we already know about matrix multiplication.3.1. as a fractional linear transformation. b = 0. so we need to check the closure axiom in Deﬁnition 3. Show that F forms a group under composition of functions. and this can be shown by choosing coeﬃcients a = 1.

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

3)(5. The element [9] = [3]2 has order 2. 7). ±[9]} is closed. 7))2 = (1. 2. 1 2 and so g1 g2 = an an = (a1 a2 )n . 3)(5. These elements. 2). 2. 7))3 = (5. suppose that g1 and g2 belong to N .4 it is enough to show that H is closed under multiplication. In G = Z× . 7). show that 21 H = {[x]21 | x ≡ 1 (mod 3)} are subgroups of G. 3)(5. 3. 25. the subset N is nonempty since the identity element e can always be written in the form e = en . 3)(5. 2. You do not have to list all of the elements if you can explain why there must be 12. and so [3] is a cyclic subgroup of order 4. one that is cyclic and one that is not 20 cyclic. (b) Find a subgroup of S7 that contains 12 elements. 6. Next. together with (1. 21 27. Since H is a ﬁnite. 2. 4)(5. 2. Corollary 3. Solution: We only need to ﬁnd an element of order 12. ±9}. then g −1 = (an )−1 = (a−1 )n . 7))6 = (1). 1 2 Finally. Solution: First. form the cyclic subgroup generated by (1. 3. and therefore [x]21 · [y]21 = [xy]21 belongs to H. Then there must exist elements a1 and a2 in G with g1 = an and g2 = an . A similar argument shows that K is a subgroup of Z× . 2. H is not cyclic since no element of H has order 4. Solution: The subset H is ﬁnite and nonempty (it certainly contains [1]21 ). then x ≡ 1 (mod 3) and t ≡ 1 (mod 3). ((1. the element (1. ±7. 2. 26. Since the order of a product of disjoint cycles is the least common multiple of their lengths. ±3. 7). ((1. we need to check the orders of elements in Z× = {±1. 3)(5. 3.2. Solution: To ﬁnd a cyclic subgroup of order 4. which 20 turns out to have order 4. 2. (a) Find the cyclic subgroup of S7 generated by the element (1. If [x]21 and [y]21 belong to H. 2. and let n be a ﬁxed positive integer. ((1. 3)(5. if g ∈ N . with g = an . 2)(5. so it follows that xy ≡ 1 (mod 3). 7))4 = (1. It is easy to check that the subset H = {±[1]. 7). 3)(5. 2. Show that N = {g ∈ G | g = an for some a ∈ G} is a subgroup of G.4 implies that it is a subgroup. Solution: We have ((1. and so g −1 has the right form to belong to N . In Z× . since it will generate a cyclic subgroup with 12 elements. so by Corollary 3. 7). 3)(5. and K = {[x]21 | x ≡ 1 (mod 7)} . Let G be an abelian group. 7) has order 12.CHAPTER 3 SOLUTIONS 63 24. Finally. nonempty subset of a known group. It is natural to begin with [3]. 3). 7))5 = (1.2. and why they must form a subgroup. ﬁnd two subgroups of order 4. ((1. The last equality holds since G is abelian.

In the group G = GL2 (R) of invertible 2 × 2 matrices with real entries. We can use Demoivre’s Theorem (Theorem A. It follows from this theorem that (cos(3π/4) + i sin(3π/4))8 = √ √ cos(6π) + i sin(6π) = 1. so by part (a) p this implies that n is a divisor of 2n−1 . 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. (See Appendix A. Solution: The order of [2] is a divisor of |Z× | = p − 1 = 2n . A similar √ √ argument shows that − 22 − 22 i also has order 8.5 for a discussion of the properties of complex numbers. 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. In the multiplicative group C× of √ complex numbers. 29. and squaring this yields 22n ≡ 1 (mod p). 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. let θ = 0. and therefore n is a power of 2. we have 2n ≡ −1 (mod p). (a) Show that in Z× the order of [2]p is 2n. and for any proper divisor k of 2n we have k ≤ n. 30.) Each of the numbers has magnitude 1. show that cos θ − sin θ H= θ∈R sin θ cos θ is a subgroup of G. elements − 2 2 2 2 Solution: It is probably easiest to change these complex numbers from rectangular coordinates into polar coordinates.5. Thus the order of [2] is a divisor of 2n.64 CHAPTER 3 SOLUTIONS 28. so 2k ≡ 1 (mod p) since 2k − 1 < 2n + 1 = p.2) to compute powers of complex numbers. ﬁnd the order of the √ √ √ 2 2 2 2 + i and − − i. . p Solution: Since 2n +1 = p. Suppose that p is a prime number of the form p = 2n + 1. and so − 22 + 22 i has order 8 in C× . This shows that [2] has order 2n. 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). Existence of inverses: cos θ sin θ − sin θ cos θ −1 = cos(−θ) − sin(−θ) sin(−θ) cos(−θ) ∈ H.

since for any element in H the determinant is a2 + 2b2 . Compute the centralizer in GL2 (R) of the matrix Note: Exercise 3. There are 8 such elements. Show that Z5 × Z3 is a cyclic group. c = −2b.3 SOLUTIONS 16. then order of an element ([a]5 . the element ([a]5 . 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. [2]5 . Let K be the following subset of GL2 (R). [3]5 . while the second and third equations imply that a = b + d = c + d. and list all of the generators for the group. 32. Equating corresponding entries 1 1 c d a+c b+d shows that we must have 2a+b = 2a+c. which is always positive. The ﬁrst and last equations imply that b = c. any matrix of this form commutes with A. 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 = . [4]5 have order 5 in Z5 and [1]3 . = 2 1 1 1 . On the other hand.14 in the text deﬁnes the centralizer of an element a of the group G to be C(a) = {x ∈ G | xa = ax}. . and a b −2b a −1 .3. [b]3 ) is a generator if and only if [a]5 = [0]5 and [b]5 = [0]5 .4 (b). ad − bc = 0 65 Show that K is a subgroup of GL2 (R).CHAPTER 3 SOLUTIONS 31. or d = a − b. The 1 a −b . Since [1]5 . [2]3 have order 3 in Z3 . 2 1 a b . K= a b c d d = a. a+b = 2b+d. and suppose that X = belongs to 1 1 c d the centralizer of A in GL2 (R). Then we must have XA = AX.2. which can easily be listed. Solution: By Proposition 3. b ∈ R and ab = a2 + b2 . [b]3 ) in Z5 × Z3 is the least common multiple of the orders of the components. 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. 1 1 b a−b Solution: Let A = 3. 2c+d = a+c. and c+d = b+d.

1) has order 2. 12) = 3. the element (1. 20. we need to omit the p multiples of the ﬁrst row. so there are pn − 1 choices. Find two groups G1 and G2 whose direct product G1 × G2 has a subgroup that is not of the form H1 × H2 . Solution: We need to count the number of ways in which an invertible matrix can be constructed.66 CHAPTER 3 SOLUTIONS Comment: The other 7 elements in the group will have at least one component equal to zero. [17]. we have o([9]12 ) = o([3]12 ) = 4. This argument can be continued. then the order of the general linear group GLn (Zp ) is (pn − 1)(pn − p) · · · (pn − pn−1 ). For the third row. so it generates a cyclic subgroup that does not have the required form. In the group G = Z× . Solution: It can be shown (as in Problem 3. The ﬁrst row can be any nonzero vector. Show that H and K are subgroups of G. 21.26) that the given subsets are subgroups. let H = {[x] | x ≡ 1 (mod 4)} and K = {[y] | y ≡ 36 1 (mod 9)}. [15]18 ) in the group Z12 × Z18 . which is the number of vectors in the subspace spanned by the ﬁrst two rows that we have chosen. Adding the identity element to the list accounts for all 15 elements of Z5 × Z3 . cn . the set HK must contain 12 elements. we need to subtract p2 . but to be linearly independent of the ﬁrst row. Thus the order of ([9]12 . giving the stated result. 18. 6] = 12. Thus there are pn − p2 possibilities for the third row. [29]} and K = {[1]. 19. This is done by noting that we need n linearly independent rows. 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). (A more formal proof could be given by induction. [13]. it cannot be a scalar multiple of that row.2. 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  . o([15]18 ) = o([3]18 ) = 6. Therefore the total number of ways to construct a second row independent of the ﬁrst is pn − p.)   i 0 0 0  in the group GL3 (C). Find the order of the element ([9]12 . Show that if p is a prime number. [5]. Solution: In Z2 × Z2 . [25]. Similarly. Since x · [1] = x · [19] for x ∈ G. A short computation shows that H = {[1]. [19]}. [15]18 ) is lcm[4. for subgroups H1 ⊆ G1 and H2 ⊆ G2 . 17. Solution: Since gcd(9. and ﬁnd the subgroup HK. Since we have p possible scalars. There are pn possibilities for the second row. and so HK = G.

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

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

8 (c) that [3]k = [3]m since [3]17 has order 16 in Z× Therefore φ([k]16 ) = φ([m]16). we have to show that the formula for φ does not depend on the particular representative that is chosen. 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 Next. since Z16 is also a cyclic group. we ﬁrst apply the function φ to the two elements. This provides the clue at to how to deﬁne the isomorphism we need. giving all of the elements in H.1. φ([2]16 ) = [3]2 . Deﬁne φ : Z16 → Z× by setting φ([1]16 ) = [3]17 . and Proposition 3. If k ≡ m (mod 16). Since φ is deﬁned by 17 using a representative n of the equivalence class [n]16 . For any elements [n]16 and [m]16 in Z16 . with generator [1]16 .4. where 0 ≤ i < 5 and 0 ≤ j < 2. Show that Z× is isomorphic to Z16 .CHAPTER 3 SOLUTIONS 69 (c) Show that every element of H can be written uniquely in the form Ai B j . for all [n]16 ∈ Z16 . 33 = 27 ≡ 10.5 that φ is also an onto function.2. The 17 17 general formula is φ([n]16 ) = [3]n . and 17 17 17 so φ is a well-deﬁned function. [n]16 and [m]16 . we ﬁrst compute what happens if we combine [n]16 and [m]16 using the operation in Z16 . and so φ is a one-to-one function. and 38 ≡ 3 · 11 ≡ 33 ≡ −1 ≡ 1.8 (c) shows that φ([k]16 ) = φ([m]16 ) only if k ≡ m (mod 16).2. Solution: Since 1 0 b 1 1 c 0 1 = 1 b+c 0 1 . etc. .1. 17 17 17 Thus φ([n]16 + [m]16 ) = φ([n]16 ) · φ([m]16 ). the cyclic subgroup gen- 1 b . The proof that φ respects the two group operations follows the proof in Example 3. 17 Solution: The element [3] is a generator for Z× . since 32 = 9. 36 ≡ 3·5 ≡ 15. 17 34 ≡ 3·10 ≡ 30 ≡ 13.3 (a) implies that any isomorphism between cyclic groups must map a generator to a generator. it follows from Proposition 2. Proposition 3. Then because both Z16 and Z× have 16 17 elements. and then substitute the result into the function φ: φ([n]16 + [m]16 ) = φ([n + m]16 ) = [3]n+m . erated by A consists of all matrices of the form 3. then it follows from Proposition 3.4 SOLUTIONS 21. Multiplying on the 0 1 right by B will create 5 additional elements. Therefore Z× is a cyclic group with 16 17 elements. 35 ≡ 3·13 ≡ 39 ≡ 5.4. and this completes the proof that φ is a group isomorphism. 37 ≡ 3·15 ≡ 45 ≡ 11.

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

respectively. and 3 | q. 27. [c]2 ) + ([b]30 . so we may conclude that φ is a group isomorphism. We know (by Exercise 3. 26. We have now shown that if φ([n]30 .4. G2 be groups. Let H = φ(H ∗ ) and K = φ(K ∗ ) be the images in G of H ∗ and K ∗ . x2 ) | x2 = e} and K ∗ = {(x1 . is a group isomorphism. Solution: By Exercise 3. Let G be a group. This means that we can use Proposition 3. [4(a + b) + 3(c + d)]2 ) = ([a + b]10 . [4a + 4b + 3c + 3d]2 ) and so φ respects the operations in the two groups. it follows that 2 | m. If φ([n]30 . [c]2 ) and ([b]30 . Prove that if G is isomorphic to G1 × G2 . and let H be a subgroup of G. for some q ∈ Z.4. . Prove that if a is any element of G. By Exercise 3. and hk = kh for all h ∈ H and k ∈ K. G1 . and so the condition in Proposition 3. then the subset aHa−1 = {g ∈ G | g = aha−1 for some h ∈ H} is a subgroup of G that is isomorphic to H. it follows that φ must also be an onto function. HK = G.4. x2 ) | x1 = e} have the properties we are looking for. [4a + 3c + 4b + 3d]2 ) = ([a + b]10 . then ([n]10 .13 in the text. [d]2 )) = ([a]10 . Exercise 3. [4n + 3m]6 ) = ([0]10 . and hk = kh for all h ∈ H and k ∈ K.9 in the text shows that in G1 × G2 the subgroups H ∗ = {(x1 . so [n]30 = [10q]30 = [0]30 . Then since 2 and 3 are prime numbers. [0]6 ). and therefore 2 | 3m since 2 | 40q. [m]2 ) = ([0]10 . [c + d]2 )) = ([a + b]10 . [d]2 )) = φ(([a + b]30 . and 3 | 40q since 3 | 3m. [0]6 ). It is then clear that the function θ : H → aHa−1 deﬁned by θ(x) = axa−1 is an isomorphism. [4a + 3c]2 ) + ([b]10 . HK = G. so 10 | n. [4a + 4b + 3c + 3d]2 ) 71 φ(([a]30 . so [m]2 = [0]2 . [4b + 3d]2 ) = ([a + b]10 .4. then ([n]30 . so we only need to show that H ∩ K = {e}. We conclude that φ is a one-toone function. or 6 | (40q + 3m). so aHa−1 = φ(H) is a subgroup of G. Let G.15 the image under φ of any subgroup of G is again a subgroup of G. [d]2 ) we have φ(([a]30 . the function φ : G → G deﬁned by φ(x) = axa−1 . We have therefore checked all of the necessary conditions. and 6 | (4n + 3m). [c]2 )) + φ(([b]30 . for all x ∈ G. say n = 10q. then there are subgroups H and K in G such that H ∩ K = {e}. Since the two groups both have 60 elements. [m]2 ) = ([0]30 .CHAPTER 3 SOLUTIONS For any elements ([a]30 . [m]2 ) = ([0]10 .4. Solution: Let φ : G1 × G2 → G be an isomorphism. [0]2 ).15) that H and K are subgroups of G.3.4 is satisﬁed.4 to show that φ is one-to-one. It follows that 2 | (40q + 3m) and 3 | (40q + 3m). [0]6 ).

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

Solution: (Outline only) The operation is well-deﬁned on S. Prove that S is a group under this operation. This information should make it plausible that the function φ : S3 → H deﬁned by φ(ai bj ) = Ai B j . since φ and φ−1 are functions and the operation on G is well-deﬁned. we can show that BA = A−1 B. where 0 ≤ i < 3 and 0 ≤ j < 2.3 (on page 104 of the text). as given in Table 3. gives a one-to-one correspondence between the elements of the groups which also produces multiplication tables that look exactly the same. b ∈ Z3 . Let G be a group. and that φ is actually a group isomorphism. H= m b 0 1 ∈ GL2 (Z3 ) m. G is not abelian.25. such that all entries in the respective group tables also have the same one-to-one correspondence. By Proposition 3. m = 0 Show that H is isomorphic to the symmetric group S3 . Remember that constructing an isomorphism is the same as constructing a one-to-one correspondence between the elements of the group. since each factor is abelian. The direct product R× ×R is an abelian group.3 (b). In this case we can explain how this can be done. and let S be any set for which there exists a one-toone and onto function φ : G → S.3.3. Let H be the following subgroup of group G = GL2 (Z3 ). and that each element of H has the form can be written uniquely in the form Ai B j . Let A = and B = . since = 0 1 0 1 0 1 1 1 −1 0 −1 1 but = . Deﬁne an operation on S by setting x1 · x2 = φ(φ−1 (x1 )φ−1 (x2 )). The associative law holds . then the groups cannot be isomorphic. Solution: This group is small enough that we can just compare its multiplication table to that of S3 .CHAPTER 3 SOLUTIONS 30. 73 Show that G is not isomorphic to the direct product R× × R. without actually writing out 1 1 −1 0 the multiplication table. Let G be the subgroup of GL2 (R) deﬁned by G= m b 0 1 m=0 . if one of the groups is abelian but the other is not. for all x1 . 31.4. for all 0 ≤ i < 3 and 0 ≤ j < 2. On −1 0 1 1 −1 −1 the other hand. x2 ∈ S. Then just 0 1 0 1 as in Problem 3. Thus the two groups cannot be iso0 1 0 1 0 1 morphic. 32. 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.

. and so [5] must 9 = 18 have order 6. The list of generators 2 7 is {±1. and so we actually have [24]30 = [18]30 .74 CHAPTER 3 SOLUTIONS in S because it holds in G. then x−1 = φ((φ−1 (x))−1 ). the generators correspond to u the numbers less than 28 and relatively prime to 28. ±13}. and Z6 × Z10 ∼ = = Z2 × Z3 × Z2 × Z5 .5. ±11. ﬁnd the order of the subgroup [18]30 . and that Z20 ∼ = = Z4 × Z5 . The two groups are not isomorphic since the ﬁrst has an element of order 4. Thus all three groups are isomorphic.3 (a). Our theorems tell us that 18 Z× ∼ Z6 . and Z× are isomorphic to each other. In Z30 .1. 23. and |Z× | = 6. It then follows from Problem 3.0. Similarly. Given a known group G such as R× .23 that Z4 × Z10 ∼ Z4 × Z2 × Z5 . ±9. showing that Z× is cyclic of order 6. ±3. Note that nZ100 is used to mean all multiples of n in Z100 . Solution: The subgroups correspond to the divisors of 100. Then [18]30 = [6]30 . Find all generators of the cyclic group Z28 . Z4 × Z15 ∼ Z4 × Z3 × Z5 . Comment: This reveals the secret behind problems like Exercises 3. 9 18 9 and so [2] must have order 6. 24. we can use one-toone functions deﬁned on G to produce new groups with operations that look rather diﬀerent from the usual examples. the identity element in S is φ(e). 9 18 Solution: First. 53 ≡ 17 ≡ 1.5 SOLUTIONS 20. Is Z4 × Z10 isomorphic to Z2 × Z20 ? Solution: It follows from Theorem 3.5.4 that Z10 ∼ Z2 × Z5 . = and Z2 × Z20 ∼ Z2 ∼ Z4 × Z5 . = 22. In Z× . we have |Z× | = 6.11 and 3. while the second has none. Solution: By Proposition 3. 25. [24]30 = [6]30 . ﬁnd the order of [24]30 . and it is easy to check that if x ∈ S. where e is the identity of G.4. Solution: Using Proposition 3.1. ±5. Give the lattice diagram of subgroups of Z100 .4. Finally. Therefore = Z4 × Z10 ∼ Z2 × Z20 . 52 ≡ 7. Z× . showing that Z× is cyclic of order 6. and are given in Figure 3. The Euler ϕ-function allows us to compute how many there are: ϕ(28) = 1 · 6 · 28 = 12. and so the subgroup has 30/6 = 5 elements. Our 9 theorems tell us that Z× ∼ Z6 . 22 = 4. 18 = 21. 30) = 6.3 (b). In Z× . Is Z4 × Z15 isomorphic to Z6 × Z10 ? Solution: As in Problem 21. we ﬁrst ﬁnd gcd(18. it is possible to show that the obvious = = mapping from Z4 × Z2 × Z5 onto Z2 ∼ Z4 × Z5 is an isomorphism.12 in the text. Show that the three groups Z6 .5. 3. 23 = 8 ≡ 1.

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

but for each given n we only need to worry about a ﬁnite number of equations. In any isomorphism. 4) = a3 b. The cyclic subgroups −1 = {±1}. In the dihedral group Dn = {ai bj | 0 ≤ i < n. o(b) = 2. for all 0 ≤ i < n. 4) = (2. so a2 ≡ 1 (mod pq) since p and q are relatively prime. 4)(1. . 3. it has order 4. In S4 . ik = −j. and so it is impossible for this group to be isomorphic to the quaternion group. ±j}. [−1]pq is not a solution. jk = i. ±i}. 3). To ﬁnd the subgroup generated by a and b. 2)(3. ba2 = (1. we have ab = (1. 3)(2. 2) = (1. so [a]pq = [−1]pq . 3. a2 b. pq Solution: We know that [−1]pq has order 2. which has 3 cyclic subgroups of order 4. with a2 = (1. 3. 3) = (1. a2 b = (1. let (1. Since a is a cycle of length 4. and thus [a]pq has order 2. 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 ). 4)(2.6 SOLUTIONS 22. 2.7 of the text (see page 108). For example. 2. ki = j. 3) = b. This shows that the subgroup generated by a and b consists of the 8 elements {e. 4) = (1. ab.3. 2. Because q is an odd prime. we have ba = (1. a3 . The Chinese remainder theorem (Theorem 1. and ba3 = (1. 4) = a2 b. b. ±k} is deﬁned in Example 3. a3 b}. ±i = {±1.3. Comment: This is similar to a proof by induction. ±i. 4. 4). the equation bai = an−i b is just the relation that deﬁnes the group. cyclic subgroups would correspond to cyclic subgroups. 4) = a and (1. show that bai = an−i b. i2 = −1. 2)(3. 4)(2. and ba = a−1 b. 4)(1. ±j = {±1. But a2 ≡ 1 (mod p) and a2 ≡ 1 (mod q). 3) = (1. 2). 3. 3)(1. and a3 b = (1. 3. 3)(1. and i4 = i2 i2 = (−1)2 = 1. 4. so by Problem 27 it is enough to ﬁnd one other element of order 2. Furthermore. 4). a. a2 . 2)(1. 30. On the other side. kj = −i.76 CHAPTER 3 SOLUTIONS Solution: The quaternion group Q = {±1. 4) and a3 = a−1 = (1. 0 ≤ j < 2} with o(a) = n. If we assume that the result holds for i = k.6) states that the system of congruences x ≡ 1 (mod p) and x ≡ −1 (mod q) has a solution [a]pq . and ±k = {±1. 3. 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 . since p and q are relatively prime. Prove that if p and q are diﬀerent odd primes. 3) = ab. ±j. 3. 4. 3)(2. 3) = (2. ji = −k. then Z× is not a cyclic group. 3)(1. The elements satisfy the following identities: i2 = j2 = k2 = −1 and ij = k. ±k} can be found by using the given identities. 3)(2. i3 = i2 i = −i. Solution: For i = 1. This implies that the result must hold for all i with 0 ≤ i < n.

4)−1 (1. 4) = (2. 4). 4). For the subgroup H of S4 deﬁned in the previous problem. 3)(1. and in A4 . 3)(2. 4) = (1. 3) and b = (1. 2.10. 2). (2. 4)−1 = (4. 4)(2. o(b) = 2. 3. 3)(1. ﬁnd the corresponding subgroup σHσ −1 . 2. Solution: The centralizer C(a) contains all powers of a. 3. 3). 3. Then we can quickly do the other computations: (1. 3). 3). 3)σ −1 = (σ(1). 4. 3). In the dihedral group Dn = {ai bj | 0 ≤ i < n. o(b) = 2. This shows that C(a) has at least n elements. 3) and (1. so we have a ⊆ C(a).CHAPTER 3 SOLUTIONS 77 23. In S4 . (1. σ(2). 4. 3). 2. we can use Exercise 2. 4). 2. for each τ ∈ H. and this set of elements is closed under multiplication. 4)H(1. show that each element of the form ai b has order 2. As a shortcut. 4)}. ab = (1. 4) = (1. 2). 2. 3)(1. 3). 2. 4). (1. (2. 4). 3. Thus (1. 4. 3)(1. . and ba = a−1 b. a. Then H must contain a2 = (1. (1. (2. 26. 3). 3)(2. Lagrange’s theorem show that there is no subgroup that lies strictly between a and Dn . Show that each element in A4 can be written as a product of 3-cycles. (1. C(a) = Dn . (2. 2). 4). (2. 3. in S4 . ﬁnd the centralizer of a. 2). 4)−1 = (1. Find the centralizer of (1. (1. since by deﬁnition b does not belong to C(a). 2). 3. In the dihedral group Dn = {ai bj | 0 ≤ i < n. Solution: We ﬁrst list the 3-cycles: (1. a2 b} = {(1). so a ⊆ C(a) ⊆ Dn and C(a) = Dn together imply that C(a) = a . 27. Rather than starting with each of the other elements and then trying to write them as a product of 3-cycles. 4)(2. 2. it is easier to just look at the possible products of 3-cycles. 4)−1 (1. 4) = (1). (2. 4)(1. 3). 2). we have (ai b)2 = (ai b)(ai b) = ai (bai )b = ai (an−i b)b = (ai an−i )(b2 ) = an e = e. 3) in S3 . 28. and this accounts for all 12 of the elements in A4 . (1. 3)}. (We have just listed the elements of S3 . 4). 2. a2 . 2)(1. and (1. (1. 4)(1. 3). 3. On the other hand. 2. Solution: Let a = (1. 4). which shows that σ(1. Since a contains exactly half of the elements in Dn . 2. (1. b. Solution: Using the result from the previous problem. 2. 4). 0 ≤ j < 2} with o(a) = n. 2)(1. and ba = a−1 b. 3. (1. 3) = (2. 25. we have (1.) Thus H = {(1). for σ = (1. (1. σ(3)) = (4. 2. 2) = (4. 4. 4). ab. (3. 0 ≤ j < 2} with o(a) = n. 2. 3.3. 2). 3) = (1. We have (1. ﬁnd the subgroup H generated by (1. 4)(1)(1. 2. and (2. 4)(1. 2) = (4. 3. Since (1. Solution: We need to compute στ σ −1 . (1. 24. 4)−1 (1. 4)−1 = {(1). 4)(1. 2)(3. 3). 4)(1. 3) and a2 b = (2. 3).

3. 4). 4)a−1 = (2. for all [x]18 ∈ Z18 . 4)a−1 = (3. we need to check that xa = ax. 4) = (x. Under any group homomorphism φ : Z4 → Z10 . 4)(2. y. . and so no product of transpositions belongs to C(a). (a) Find the formulas for all group homomorphisms from Z18 into Z30 . 4)a−1 = (2. 4)a−1 = (1. 2)(3. That exercise shows that if (1. the centralizer of (1. 4). a(1. a(2. σ(2). 3. 4)(3.7. . which deﬁnes the zero function. so (u. Exercise 2. It helps to have some shortcuts when doing the necessary computations. and z.7. we can conclude that no 4-cycle belongs to C(a). Using Exercise 2. 3) do not commute with a. v. . 2) does not belong to the centralizer. k)σ −1 = (σ(1). y. 2)}. This accounts for an additional 6 elements. w. Finally. v. 2. then σ(1. . Since there are 6 elements of this form. 1). A similar argument shows that no 3-cycle that includes the number 4 as one of its entries can belong to C(a). (1. To see that x belongs to C(a). (1. 3) is equal to (1. for all [x]n ∈ Zn . or else φ([1]4 ) = [5]10 . 3)(2. we have a(1. 4). 4). Let a = (1. Thus φ([1]4 ) = [0]10 .3.5 shows that any group homomorphism from Zn into Zk must have the form φ([x]n ) = [mx]k . Solution: Example 3. For the products of the transposition. 2. Solution: Example 3. we now have a total of 21 elements that are not in C(a). the centralizer C(a) always contains the cyclic subgroup a generated by a. . 3).5 shows that any group homomorphism from Z18 into Z30 must have the form φ([x]18 ) = [mx]30 . Find all group homomorphisms from Z4 into Z10 . 3) in S4 we have to work a bit harder. so no transposition in S4 commutes with a. 3)(1. (2. 2.7 SOLUTIONS 17. To ﬁnd the centralizer of (1. 4)a−1 = (2. z. and (2. w. . Without doing all of the calculations. . 3). Since . From the computations in S3 . 2. This means that w = z. 4). 3)a−1 = (2. the order of φ([1]4 ) must be a divisor of 4 and of 10. which leads to the formula φ([x]4 ) = [5x]10 . 2. 2. or that axa−1 = x.10. 1)(3.78 CHAPTER 3 SOLUTIONS Solution: Since any power of an element a commutes with a. 4). 3. . 2. z. and (3.10 provides a quick way to do this in a group of permutations. 18. we know that (1. and by Lagrange’s theorem a proper subgroup of a group with 6 elements can have at most 3 elements. 3). 4). y. . In S3 . 2). 4). and therefore C(a) = a . σ(k)). 2. we have a(1. 4). Thus the centralizer of (1. The remaining transpositions in S4 are (1. . we will have a(x. (1. in A4 we must get the same answer: C(a) = a . If we do a similar computation with a 4-cycle. k) is a cycle of length k and σ is any permutation. . so the only possibilities are 1 and 2. 3) . and a(3. 4)a−1 = (u. and a(1. . since a just permutes the numbers x. for all [x]4 ∈ Z4 . 3) always contains the subgroup {(1). 4). since it is easy to check that (1.

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

for all [x]pq ∈ Z× . namely [1]36 . 24. 12) = 1. pq p q so φ is also an onto function. φ2 (x)). Solution: Given a. [25]36 . It follows that [x]pq = [1]pq . Let n and m be positive integers. for all [x]n ∈ Z× . ﬁnd the kernel and image of φ. φ is a homomorphism since for [a]n . since m | n. For the group homomorphism φ : Z× → Z× deﬁned by φ([x]36 ) = [x]12 . 36 12 The kernel of φ consists of the elements in Z× that are congruent to 1 mod 36 12. then gcd(x. and this implies that m | (x1 − x2 ). φ([a]n [b]n ) = φ([ab]n ) = n [ab]m = [a]m [b]m = φ([a]n )φ([b]n ). n then n | (x1 − x2 ). It follows that Z× ∼ Z× / [13]36 . for all [x]pq ∈ Z× . and this implies that pq | (x − 1). 22. for all [x]pq ∈ Z× . Solution: The previous problem shows that φ is a group homomorphism. φ is a well-deﬁned function. pq Using Problem 23. φ2 (b)) = (φ1 (a)φ1 (b). since if [x1 ]n = [x2 ]n in Z× . Prove that φ : G → G1 × G2 deﬁned by φ(x) = (φ1 (x). . is a well-deﬁned group homomorphism. and G2 be groups. φ2 (a)) · (φ1 (b). 36) = 1. Next. [13]36 . Show that φ : Z× → Z× deﬁned by φ([x]n ) = [x]m . p q Solution: Using Problem 21. and so φ([x1 ]n ) = φ([x2 ]n ). we can deﬁne group homomorphisms φ1 : Z× → pq Z× and φ2 : Z× → Z× by setting φ1 ([x]pq ) = [x]p . we have φ(ab) = (φ1 (ab). φ2 (a)φ2 (b)) and so φ : G → G1 × G2 is a group homomorphism. φ2 (ab)) = (φ1 (a)φ1 (b). for all x ∈ G. Let p and q be diﬀerent odd primes. for 36 12 all [x]36 ∈ Z× . Thus [x1 ]m = [x2 ]m . Solution: First.80 CHAPTER 3 SOLUTIONS 21. It follows that Z× and Z× × Z× have the same order. and apply the fundamental 36 homomorphism theorem. Prove that Z× is isomorphic to the direct pq product Z× × Z× . and this shows that φ is a one-to-one function. Exercise 1. so p | (x − 1) and q | (x − 1). φ2 (a)φ2 (b)) φ(a)φ(b) = (φ1 (a).4.27 in the text states that if m > 0 and n > 0 are relatively prime integers. since if gcd(x. since p adn q are relatively prime. then ϕ(mn) = ϕ(m)ϕ(n). If [x]pq ∈ pq ker(φ). φ2 ([x]pq )). [b]n ∈ Z× . we can deﬁne a group homomorphism φ : Z× → Z× × Z× q pq p by setting φ([x]pq ) = (φ1 ([x]pq ). It is evident that φ maps Z× onto Z× . b in G. then [x]p = [1]p and [x]q = [1]q . Let φ1 : G → G1 and φ2 : G → G2 be group homomorphisms. Let G. G1 . 12 = 36 23. is a well-deﬁned n m n group homomorphism. This completes the proof that φ is a group isomorphism. and pq q p pq φ2 ([x]pq ) = [x]q . such that m is a divisor of n.

1)} (1. 0) + K = {(1. with generators a of order n and b of order 2. (3. 0) + H = {(1. so it must have order 4. (3. 2)} (4. Solution: The cosets of H = {(0. 1). 0)}. . 3)} (0. and can be proved inductively: assuming bak = a−k b. (4. 0)} (2. 3. 1). 0) + H = {(3. 3)} (1. 3)} (0. (3. Thus the order of an element in G/H is at most 6. 0).CHAPTER 3 SOLUTIONS 81 3. 0). (5. 28. (4. 2)} (1. 3)} (5. 0) + K = {(2. 0)} (0. 3) + K = {(1. 1). 1). 5. 3)} The cosets of K = {(0. 1) + H = {(3. Let the dihedral group Dn be given via generators and relations. (0. (5. 1). 1). 0). the coset 3 7 does not have order 2. (1. 3). 2) + K = {(0. let H = {(0. (a) Show that bai = a−i b for all i with 1 ≤ i < n. List the cosets of 7 in Z× . (0. 1) + K = {(1. (4. 1) + K = {(2. (5. 2)} (1. 3) + K = {(0. 0) + H = {(4. 2). 1) + H = {(1. 2)} (3. 11. 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. (3. 5} 9 7 = {9. 9. 3)} (3. 1) + H = {(5. 15}. (1. 0). 0) + H = {(5. 2)} (2. producing an element in H. 1) + H = {(2. 2). Solution: The identity holds for all positive integers i. 0). 1). 0). (5. 3) + K = {(2. 1)} (0. (4. 3)} (2. 1). 7. 1) + K = {(0. (4. 0). 2)} (0. 0) + K = {(0. (0. (3. Let G = Z6 × Z4 . (3. 3). 0)} (1. 1). 2)} (2. 3). 0) + H = {(0. 0)} are (b) You may assume that any abelian group of order 12 is isomorphic to either Z12 or Z6 × Z2 . satisfying ba = a−1 b. 0). (2. 7} 2 3 7 = {3. 2)} (1. and let K = {(0. 0) + H = {(2. 0). 3)} (4. (5. 13. (3. 0). we have bak+1 = bak a = a−k ba = a−k a−1 b = a−(k+1) b. showing that the factor group is cyclic. 2) + K = {(1. 0). (a) List all cosets of H. 2).8 SOLUTIONS 27. 16 7 = {1. (5. list all cosets of K. (2. 1) + K has order 12 in G/K. 2) + K = {(2. On the other hand. 2)} (5. 1) + H = {(0. 0). Is the factor group Z× / 7 cyclic? 16 16 Solution: Z× = {1. 1) + H = {(4. (4. 13} Since 3 ∈ 7 . = ∼ Z12 . 2)}. (1. and so G/K = 29. 15} 11 7 = {11. 3)} (2. 1)} (2. 2)} are (0. (3. 0). (0. and so G/H ∼ Z6 × Z2 .

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

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

We could 13 also write this as Z× = [2]13 . and give the lattice diagram which shows the 11 inclusions between them. [4]. 24 = 5.     b 1 c d 1 a+c b+d 0  0 1 0  =  0 1 0 . and we have 22 = 4. 24 = 16 ≡ 3. 29 = 6. [10]} The lattice diagram forms a diamond.  −1  1 a b 1 −a  0 1 0  1 =  0 0 0 1 0 0 verses.   1 a b Solution: Deﬁne φ : G → R × R by φ  0 1 0  = (a. and 26 ≡ 2 · 25 ≡ 12. 13 Solution: The ﬁrst element to try is [2]. 3. (a) What are the possibilities for the order of an element of Z× ? Explain 13 your answer. so the subgroups are as follows. [9]. Solution: The group Z× has order 12. 13 2. 4.84 CHAPTER 3 SOLUTIONS SOLUTIONS TO THE REVIEW PROBLEMS 1. Solution: First check for cyclic subgroups. b ∈ R . Show that the group G in the previous problem is isomorphic to the direct product R × R. b). so 1 0 0 1 0 0 1 The identity matrix belongs to the set. 28 = 3. [5]. 3. and 12. 2. [2]2 . and the order of any element must be 13 a divisor of 12. 23 = 8. 210 = 1. so the possible orders are 1. [2]6 . 0 1 GL3 (R). (b) Show that Z× is a cyclic group. This shows that Z× is cyclic. in shorthand notation: 22 = 4. 23 = 8. 6. and  −b 0 . [2]4 . and thus [2] is a generator for Z× . in addition to Z× 11 11 and {[1]}: [2]2 = {[1]. so the order of [2] is greater than 6. 25 ≡ 2 · 24 ≡ 6. 26 = 9. [2]5 } = {[1]. [2]8 } = {[1]. By part (a) it must be 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. [3]} and [2]5 = {[1]. 4. This is 0 0 1 one-to-one and onto because it has an inverse function θ : R × R → G deﬁned . 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. 25 = 10. 27 = 7. Find all subgroups of Z× .

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

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

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

the fact that it has not roots in Q does not mean that it is irreducible over Q. x + = 1 x2 g(x) + (− 1 )(x − 1)f (x). ±2. the only possible factorization has the form x4 −2x2 +8x+1 = (x2 +ax+b)(x2 +cx+d). −6 as possibilities. 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. or b = d = −1. f (−2) = −294. 2 2 5. Then f (2) = 32. and these do not work. Beginning with the last equation and back-solving. 6. We have either b = d = 1. Factor x5 − 10x4 + 24x3 + 9x2 − 33x − 12 over Q. ±2. The second factor is irreducible over Q since it satisﬁes Eisenstein’s criterion for p = 3. so this eliminates all but ±2. ±3. (It is important to note that since the degree of the polynomial is greater than 3. so 2x + 1 is the greatest common divisor (we must then divide by 2 to make it monic). This implies that the polynomial is irreducible over Q. Solution: The possible rational roots of f (x) = x5 −10x4 +24x3 +9x2 −33x−12 are ±1. The only possible rational roots of the second factor are 1 and −1. and ﬁnally we obtain the factorization f (x) = (x − 4)(x4 − 6x3 + 9x + 3). Are the following polynomials irreducible over Q? (a) 3x5 + 18x2 + 24x + 6 Solution: Dividing by 3 we obtain x5 + 6x2 + 8x + 2. We have f (1) = 21. (c) 2x10 + 25x3 + 10x2 − 30 Solution: Eisenstein’s criterion is satisﬁed for p = 5. 7. and bd = 1. so for any root we must have (r − 1)|21. which is irreducible in Z2 [x]. ±4. and this satisﬁes Eisenstein’s criterion for p = 2.) Since the polynomial has no linear factors. and since 2 is a root we have the factorization x5 − 2x4 − 2x3 + 12x2 − 15x − 2 = (x − 2)(x4 − 2x2 + 8x + 1). and then g(x) = (x−1)(2x2 −x−1)+(2x+1). Solution: The possible rational roots are ±1. At the next step we obtain 2x2 − x − 1 = (x − 1)(2x + 1). ±6. Factor x5 − 2x4 − 2x3 + 12x2 − 15x − 2 over Q. (b) 7x3 + 12x2 + 3x + 45 Solution: Reducing the coeﬃcients modulo 2 gives the polynomial x3 + x + 1. ac + b + d = −2. ad + bc = 8. 4. in which . in which case a + c = 8.88 CHAPTER 4 SOLUTIONS f (x) = (x+1)g(x)+(2x2 −x−1). ±12. This leads to the equations a + c = 0.

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

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

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

21.94 CHAPTER 5 SOLUTIONS Solution: By part (b). Addition in the two rings is diﬀerent. while in Z3 ⊕ Z3 there is not. It follows that the largest possible order of an element is lcm[2. and this implies a = 0 and b = 0 since Z2 is a ﬁeld. y)(0. the only idempotent element is the identity) and the only nilpotent element that is also idempotent is the zero matrix. an element in R is either a unit or nilpotent. Solution: We need to solve (x. If φ(1) = 7. On the other hand. in R we have [x + 1] = [0] but [x + 1]2 = [x2 + 1] = [0]. Solution: In R we have a + a = 0. b)2 = (0. while that of Z3 ⊕ Z3 is not. 4. Thus R cannot be isomorphic to Z2 ⊕ Z2 . then the image is 14Z42 and the kernel is 3Z120 .5. 7. 28. since in that ring (a. 14. We only need 2y ≡ 0 (mod 8). and it can be checked that in 7Z42 . For the element a = (0. The only unit that is idempotent is the identity matrix (in a group. what is the largest possible order 180 of an element? Solution: Since 180 = 22 32 5. 28 are idempotent. Furthermore. If φ(1) = 28. then the image is 7Z42 and the kernel is 6Z120 . since in Z9 there is a nonzero solution to the equation x2 = 0. 180 = 4 9 5 = In the latter additive group. . Find all ring homomorphisms from Z120 into Z42 . If φ(1) = 21.10 shows that Z× ∼ Z× × Z× × Z× ∼ Z2 × Z6 × Z4 . 6. and then a = 0 and b = 0.) 5.2. then the image is 21Z42 and the kernel is 2Z120 . 3. 42) = 6. (In Z9 let x = 3. for all a ∈ R. The argument can be given using either addition or multiplication. so R is not isomorphic to Z4 . 21. 4] = 12. since the additive group of Z9 is cyclic. so the ﬁrst component x can be any element of Z12 . Then Example 5. y) ∈ Z12 ⊕ Z8 . 6. Let R be the ring Z2 [x]/ x2 + 1 . it follows from Theorem 3. Multiplication is also diﬀerent. The additive order of φ(1) must be a divisor of gcd(120. Show that although R has 4 elements. only 0. 2) of the ring R = Z12 ⊕ Z8 . 0) for (x. so it must belong to the subgroup 7Z42 = {0. it is not isomorphic to either of the rings Z4 or Z2 ⊕ Z2 . b)2 = (0. Show that Ann(a) is an ideal of R. 7. the order of an element is the least common multiple of the orders of its components. 0) implies a2 = 0 and b2 = 0. while in Z3 ⊕ Z3 the equation (a.4 that the ring Z180 is isomorphic to the ring Z4 ⊕ Z9 ⊕ Z5 . In the group Z× of units of the ring Z180 . 2. Solution: Let φ : Z120 → Z42 be a ring homomorphism. φ(1) must be idempotent. 2) = (0. 0) implies a2 = 0 and b2 = 0. 35}. Are Z9 and Z3 ⊕ Z3 isomorphic as rings? Solution: The answer is no. ﬁnd Ann(a) = {r ∈ R | ra = 0}.

you can do the ﬁrst two parts together. then Z36 /P is a ﬁnite integral domain. Prove that I is a prime ideal. Let I be the subset of Z[x] consisting of all polynomials with even coeﬃcients. with kernel equal to I. prove that I is not maximal. 1). 7. Then R/I is isomorphic to Z2 [x]. and these correspond to 2Z36 and 3Z36 . Find all maximal ideals. and hence P is maximal. Give an example to show that the set of all zero divisors of a ring need not be an ideal of the ring. The lattice of ideals of Z36 is exactly the same as the lattice of subgroups. (a) Show that I is an ideal of R. so I is not maximal. 10. of Z36 = Z/36Z. Hint: If you use the fundamental homomorphism theorem. and it is also closed under multiplication by any element of R since 4Z8 is an ideal of Z8 . Let R be the ring Z2 [x]/ x4 + 1 . the maximal ideals that contain 36Z are 2Z and 3Z. an obvious contradiction. Solution: If P is a prime ideal of Z36 . An alternate approach we can use Proposition 5. 0) and (0. This is an onto ring homomorphism with kernel I. 9.7. so the maximal ideals of Z36 correspond to the prime divisors of 36. and all prime ideals. so the relevant ideals correspond to the divisors of 36. Again.CHAPTER 5 SOLUTIONS 95 while y = 0. In Z every ideal is principal. 1) of Z × Z are zero divisors. Solution: The elements (1. = Solution: Deﬁne φ : Z2 [x]/ x4 + 1 → Z2 [x]/ x2 + 1 by φ(f (x) + x4 + 1 ) = (f (x) + x2 + 1 ). . 4. which is not a ﬁeld.3. Thus Ann((0. 8. Thus we only need to ﬁnd the maximal ideals of Z36 . This set is certainly closed under addition. Solution: Deﬁne φ : Z[x] → Z2 [x] by reducing coeﬃcients modulo 2. so it is a ﬁeld. and let I be the set of all congruence classes in R of the form [f (x)(x2 + 1)]. (b) Show that R/I ∼ Z2 [x]/ x2 + 1 . which shows that there is a one-to-one correspondence between the ideals of Z/36Z and the ideals of Z that contain 36Z. 2)) = Z12 ⊕ 4Z8 . It is not diﬃcult to show that φ is an onto ring homomorphism. but if the set of zero divisors were closed under addition it would include (1. The maximal ideals of Z36 are thus 2Z36 and 3Z36 . This mapping is well-deﬁned since x2 + 1 is a factor of x4 + 1 over Z2 . (c) Is I a prime ideal of R? Solution: No: (x + 1)(x + 1) ≡ 0 (mod x2 + 1).

Solution: We have [x]3 = [1]. for all r ∈ R. (c) Find the idempotent elements of R. (b) Show that if e is idempotent. s ∈ R we have φ(rs) = (rse. 12.7. rs(1 − e)2 ) = (rse. Furthermore. r(1−e)). Let R be any commutative ring with identity 1. It is not necessary Table 5. Solution: By Proposition 5. and adding the two equations gives r = s. [x + 1][x2 + x + 1] = [x3 + 1] = [0]. rs(1 − e)) and φ(r)φ(s) = (re. nonzero ideals are the principal ideals generated by [x + 1] and [x2 + x + 1]. (b) Find the units of R. CHAPTER 5 SOLUTIONS (a) Show that if e is an idempotent element of R. so [x + 1] and [x2 + x + 1] cannot be units.3. (a) Find all ideals of R.1 gives the multiplication table. s(1 − e)) = (rse2 . Then φ is one-to-one since if φ(r) = φ(s). Thus the only units are 1. = Solution: Note that e(1−e) = e−e2 = e−e = 0.96 11. Let R be the ring Z2 [x]/ x3 + 1 . φ is onto. so [x] and [x2 ] are units. and [x2 ]. since for any element (ae.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. so the only proper. . it is easy to check that φ preserves addition. [x]. b(1 − e)) we have (ae. then 1−e is also idempotent. and for any r. Finally. Solution: We have (1−e)2 = (1−e)(1−e) = 1−e−e+e2 = 1−e−e+e = 1−e. then re = se and r(1 − e) = s(1 − e). r(1 − e))(se. On the other hand. b(1 − e)) = φ(r) for r = ae + b(1 − e). We have the factorization x3 + 1 = x3 − 1 = (x − 1)(x2 + x + 1). This also excludes [x2 + x] = [x][x + 1] and [x2 + 1] = [x2 ][1 + x]. then R ∼ Re ⊕ R(1 − e). Deﬁne φ : R → Re⊕R(1−e) by φ(r) = (re. rs(1 − e)). the ideals of R correspond to the ideals of Z2 [x] that contain x3 + 1 . Solution: Note: Table 5.

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

It is clear that 1 + i is in the kernel. Then m and n are either both even or both odd.98 CHAPTER 5 SOLUTIONS (a) Deﬁne φ : Z[i] → Z2 by φ(m + ni) = [m + n]2 . 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) . Since the elements of Zp [x]/ x2 + 1 all have the form [a + bx]. Prove that φ is a ring homomorphism. Solution: We have the following computations. which show that θ is a ring homomorphism. r + J). with ker(φ) = I ∩ J. deﬁne θ : Z[i] → Zp [x]/ x2 + 1 by θ(m + ni) = [m + nx]. . We need to use the fact that [x2 ] = [−1] in Zp [x]/ x2 + 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 + bdx2 ] . We claim that ker(φ) is generated by 1 + i. and deﬁne the function φ : R → R/I ⊕ R/J by φ(r) = (r + I. it is clear the θ is an onto function. for all r ∈ R. 2 and so m + ni belongs to the principal ideal generated by 1 + i. (b) For any prime number p. 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 . and so it follows that m − n is always even. 16. Find ker(φ) and show that it is a principal ideal of Z[i]. Prove that θ is an onto ring homomorphism. Let m + ni ∈ ker(φ) = {m + ni | m + n ≡ 0 (mod 2)}. Let I and J be ideals in the commutative ring R. for some congruence classes a and b in Zp . and we note that (1 − i)(1 + i) = 2. (a) Show that φ is a ring homomorphism. Solution: We have the following computations.

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

100 CHAPTER 5 SOLUTIONS .

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

3 25. √ √ Find [Q( 7 16 + 3 7 8) : Q]. 7. 2 3√ is 25} √ √ √ a basis for Q( 3 5. 2. Therefore {1. and this extension contains u = 2 + 3 5. or α3 − 3iα2 − 3α + i = 2. 3 5. 3 5) over Q. Since x5 − 2 is also irre√ ducible by Eisenstein’s criterion. 3 25. 3 5. √ √ Show that [Q( 2 + 3 5) : Q] = 6. 2. 3 5. 3 5. 2. and remains irreducible √ over Q( 5 2). 3. Solution: Eisenstein’s criterion works with p = 2. . i). 8. 2 3 5. 3 25. 3 5) in over Q( 3 5). This root would have degree 3 over Q. If x3 + 6x2 − 12x + 2 could √ be factored over Q( 5 2). and so it would √ have a root in Q( 5 2). √ √ √ √ Solution: Let u = 7 16 + 3 7 8. √ √ √ Find the degree of 3 2 + i over Q.102 CHAPTER 6 SOLUTIONS 4. 5. √ √ √ Solution: The set {1.4 shows that the required basis √ √ √ so √ √ of is {1. Thus {1. 2 3 5. and so Q( 2 + i) = Q( 3 2. or 3 over Q. √ √ √ √ √ √ √ Since {1. and then u must have degree 7 over Q since [Q(u) : Q] = is a divisor of [Q( 7 2) : Q]. so 2 + 3 5 cannot satisfy √nonzero polynomial of degree 1. 5 3 25}. then √ √ √ √ √ √ a( 2 + 3 5)3 + b( 2 + 3 5)2 + c( 2 + 3 5) + d = √ √ √ √ √ √ √ √ √ a(2 2 + 6 3 5 + 3 2 3 25 + 5) + b(2 + 2 2 3 5 + 3 25) + c( 2 + 3 5) + d = √ √ √ √ √ √ √ (5a + 2b + d) · 1 + (6a + c) 3 5 + b 3 25 + (2a + c) 2 + 2b 2 3 5 + 3a 2 3 25 = 0. 2 3 25} are linearly independent over Q. 6. 5. and then c = d = 0 as well. Since u = ( 7 2 + 3)( 7 2)3 . it follows that √ u ∈ Q( 7√ Since x7 − 2 is irreducible over Q by Eisenstein’s criterion. [Q( 5 2) : Q] = 5. and since this √ extension has degree 3. We conclude a √ that [Q( 2 + 3 5) : Q] = 6. so that α − i = 3 2. so we have α3 − 3iα2 + 3i2 α − i3 = 2. 3 25} is a basis for Q( 3 5) over Q. or 6 over Q.√it √ follows immediately that a = b = 0. and that is impossible since 3 is not a divisor of 5. Then (α − i)3 = 2. 5 is a basis for Q( 5. the minimal polynomial √2 − √ of√ 2 remains irre2 √ √ x √ √ √ ducible over the extension Q( 3 5). Solving for i we get √ i = (α3 − 3α − 2)/(3α2 − 1). and √ the proof √ Theorem 6. the √ minimal polynomial x2 − 5 of 5 remains irre√ √ √ ducible √ the extension Q( 3 5). It follows that u has degree 2. we 2). √ √ √ Solution: The set {1. and this shows that i ∈ Q( 3 √+ i). √ √ We will show that u cannot have degree ≤ 3. If 2 + 3 5 is a root of a polynomial ax3 + bx2 + cx + d in Q[x]. 3 5. 3 25} is a basis for Q( 3 5) over Q. then it would have a linear factor. Show that x3 + 6x2 − 12x + 2 is irreducible over Q. It follows 2 √ √ √ 3 3 3 immediately that 2 ∈ Q( 2 + i). 5 3 5.2. and since this √ extension has degree 3. Does 4 2 belong to Q( 3 2 + i)? √ √ Solution: Let α = 3 2 + i. √ √ Find a basis for Q( 5. have [Q( 7 2) : Q] √ 7. 2) over Q.

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

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

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

6. 8. 94 nilpotent element. 26. ﬁnite. 2 lattice diagram. 24–26. 94. 31. 29. multiplicative. 44–46. 7. 94. of groups. 49 ideal. 15. 5. 62. of a ring homomorphism. 14 inverse element. 3. 30. 9. 10. 31. even. 23. parallel. 82. 95 ideal. 6. 95 polynomial. 62. 86 group. 22. 33. 94. 23. 76. 95. 39 linear combination. 29. 22. 11. 29. 23. 84 . 80 polynomial. 16–18. 5 nilpotent element. 10 permutation. 101 prime ideal. of subgroups. 26. 54 parallel plane. 13. 25. 53 matrix. modulo n. 2. 96. 97–99. 64 inverse. 95. 97. 102 isomorphic rings. 86 nullity. 44. 82. 44. 94. 5. 22. 50. 93. 31. 10. 93. prime. 101. 95 minimal polynomial. 53 plane. 21 group homomorphism. 5. 31. modulo n. fractional. 101–103 kernel. 95 Lagrange’s theorem. 33. of a permutation. 50 one-to-one function. 45. 10. 45. irreducible. 5 identity element. 30. 53 irreducible polynomial. 30. 60. 95–98 idempotent element. 47. 74 order. 7. 51 order.106 group. 66. 30. 98 horizontal line test. 28. 11. 8. 59 integers mod n. of permutations. 23. 101 ideal. 54 perpendicular plane. 28. 29. 95. 101 multiplicative inverse. 14. 47 nilpotent element. dihedral. 41. 90 multiplicative order. of groups. 93 normal subgroup. 6. 78–80 group isomorphism. 5. 5 multiplicative order. 97. 45. 15. 6. 21. 50. 8. 2. 53 plane. 38. 53 permutation group. 53 partition. 51 maximal ideal. 8. 30. 54 group. 6. 99 INDEX lattice diagram. 30. 6. maximal. 9 permutation. 21. 35 linear congruence. 41. of a ring. 29. 53 kernel. 22. 95 ideal. 95–98 74. 30. relatively. 33. principal ideal. 69–73 isomorphism. 57. 80 induction. 8. of a ring homomorphism. of rings. 20. 81. 19. 5. 16. symmetric. 51 onto function. 58. 13 group. 47 order. 30. 97 idempotent element. invertible. 51 matrix. of rings. modulo n. principal. 60. 23 homomorphism. 9. 77. minimal. 83. 89. 5 linear transformation. 94 image. of a group homomorphism. 7. 90 invertible matrix. 28. perpendicular. 15. 77. 65 image. 56. 50. 18. 29. 11. multiplicative. 94. 99 prime. 44. 33. 60 linear transformation. 88. 51 linearly independent vectors. 78–80 homomorphism. 8. 60. 99 isomorphism.

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