Chapter 8

Topics in Number Theory

8.1 The Greatest Common Divisor

Preview Activity 1 (The Greatest Common Divisor)

1. Explain what it means to say that a nonzero integer m divides an integer n.

Recall that we use the notation m j n to indicate that the nonzero integer m
divides the integer n.

2. Let m and n be integers with m ¤ 0. Explain what it means to say that m

does not divide n.

Definition. Let a and b be integers, not both 0. A common divisor of a and

b is any nonzero integer that divides both a and b. The largest natural number
that divides both a and b is called the greatest common divisor of a and b.
The greatest common divisor of a and b is denoted by gcd .a; b/.

3. Use the roster method to list the elements of the set that contains all the
natural numbers that are divisors of 48.

4. Use the roster method to list the elements of the set that contains all the
natural numbers that are divisors of 84.

5. Determine the intersection of the two sets in Parts (3) and (4). This set
contains all the natural numbers that are common divisors of 48 and 84.

6. What is the greatest common divisor of 48 and 84?

414
8.1. The Greatest Common Divisor 415

7. Use the method suggested in Parts (3) through (6) to determine each of the
following: gcd.8; 12/, gcd.0; 5/, gcd.8; 27/, and gcd.14; 28/.

8. If a and b are integers, make a conjecture about how the common divisors
of a and b are related to the greatest common divisor of a and b.

Preview Activity 2 (The GCD and the Division Algorithm)

When we speak of the quotient and the remainder when we “divide an integer a
by the positive integer b,” we will always mean the quotient q and the remainder r
guaranteed by the Division Algorithm. (See Section 3.5, page 143.)

1. Each row in the following table contains values for the integers a and b.
In this table, the value of r is the remainder (from the Division Algorithm)
when a is divided by b. Complete each row in this table by determining
gcd.a; b/, r, and gcd.b; r/.

a b gcd.a; b/ Remainder r gcd.b; r/

44 12
75 21
50 33

2. Formulate a conjecture based on the results of the table in Part (1).

The System of Integers

Number theory is a study of the system of integers, which consists of the set of
integers, Z D f : : : ; 3; 2; 1; 0; 1; 2; 3; : : : g and the various properties of this
set under the usual operations of addition and multiplication and under the usual
ordering relation of “less than.” The properties of the integers in Table 8.1 will be
considered axioms in this text.
We will also assume the properties of the integers shown in Table 8.2. These
properties can be proven from the properties in Table 8.1. (However, we will not
do so here.)
416 Chapter 8. Topics in Number Theory

For all integers a, b, and c:

Closure Properties for Addition and a C b 2 Z and
Multiplication ab 2 Z
Commutative Properties for Addition a C b D b C a, and
and Multiplication ab D ba
Associative Properties for Addition .a C b/ C c D a C .b C c/ and
and Multiplication .ab/ c D a .bc/
Distributive Properties of Multiplica- a .b C c/ D ab C ac, and
tion over Addition .b C c/ a D ba C ca
Additive and Multiplicative Identity a C 0 D 0 C a D a, and
Properties a
1 D1
a Da
Additive Inverse Property a C . a/ D . a/ C a D 0

Table 8.1: Axioms for the Integers

Zero Property of Multipli- If a 2 Z, then a
0 D 0
a D 0.

cation
Cancellation Properties If a; b; c 2 Z and a C b D a C c, then b D c.
of Addition and Multipli- If a; b; c 2 Z, a ¤ 0 and ac D bc, then b D c.
cation

Table 8.2: Properties of the Integers

We have already studied a good deal of number theory in this text in our dis-
cussion of proof methods. In particular, we have studied even and odd integers,
divisibility of integers, congruence, and the Division Algorithm. See the summary
for Chapter 3 on page 166 for a summary of results concerning even and odd in-
tegers as well as results concerning properties of divisors. We reviewed some of
these properties and the Division Algorithm in the Preview Activities.

The Greatest Common Divisor

One of the most important concepts in elementary number theory is that of the
greatest common divisor of two integers. The definition for the greatest common
divisor of two integers (not both zero) was given in Preview Activity 1.

1. If a; b 2 Z and a and b are not both 0, and if d 2 N, then d D gcd.a; b/

provided that it satisfies all of the following properties:
8.1. The Greatest Common Divisor 417

 d j a and d j b. That is, d is a common divisor of a and b.

 If k is a natural number such that k j a and k j b, then k  d. That is,
any other common divisor of a and b is less than or equal to d.

2. Consequently, a natural number d is not the greatest common divisor of a

and b provided that it does not satisfy at least one of these properties. That
is, d is not equal to gcd.a; b/ provided that

 d does not divide a or d does not divide b; or

 There exists a natural number k such that k j a and k j b and k > d.

This means that d is not the greatest common divisor of a and b provided
that it is not a common divisor of a and b or that there exists a common
divisor of a and b that is greater than d.

In the preview activities, we determined the greatest common divisors for several
pairs of integers. The process we used was to list all the divisors of both integers,
then list all the common divisors of both integers and, finally, from the list of all
common divisors, find the greatest (largest) common divisor. This method works
reasonably well for small integers but can get quite cumbersome if the integers are
large. Before we develop an efficient method for determining the greatest common
divisor of two integers, we need to establish some properties of greatest common
divisors.
One property was suggested in Preview Activity 1. If we look at the results
in Part (7) of that preview activity, we should observe that any common divisor of
a and b will divide gcd.a; b/. In fact, the primary goals of the remainder of this
section are

1. To find an efficient method for determining gcd.a; b/, where a and b are
integers.

2. To prove that the natural number gcd.a; b/ is the only natural number d that
satisfies the following properties:

 d divides a and d divides b; and

 if k is a natural number such that k j a and k j b, then k j d.

The second goal is only slightly different from the definition of the greatest com-
mon divisor. The only difference is in the second condition where k  d is re-
placed by k j d.
418 Chapter 8. Topics in Number Theory

We will first consider the case where a and b are integers with a ¤ 0 and
b > 0. The proof of the result stated in the second goal contains a method
(called the Euclidean Algorithm) for determining the greatest common divisors
of the two integers a and b. The main idea of the method is to keep replacing the
pair of integers .a; b/ with another pair of integers .b; r/, where 0  r < b and
gcd.b; r/ D gcd.a; b/. This idea was explored in Preview Activity 2. Lemma 8.1
is a conjecture that could have been formulated in Preview Activity 2.

Lemma 8.1. Let c and d be integers, not both equal to zero. If q and r are integers
such that c D d
q C r, then gcd.c; d / D gcd.d; r/.

Proof. Let c and d be integers, not both equal to zero. Assume that q and r are
integers such that c D d
q C r. For ease of notation, we will let

m D gcd.c; d / and n D gcd.d; r/:

Now, m divides c and m divides d . Consequently, there exist integers x and y such
that c D mx and d D my. Hence,

r Dc d
q
r D mx .my/q
r D m.x yq/:

But this means that m divides r. Since m divides d and m divides r, m is less than
or equal to gcd.d; r/. Thus, m  n.
Using a similar argument, we see that n divides d and n divides r. Since
c D d
q C r, we can prove that n divides c. Hence, n divides c and n divides d .
Thus, n  gcd.c; d / or n  m. We now have m  n and n  m. Hence, m D n
and gcd.c; d / D gcd.d; r/. 

Progress Check 8.2 (Illustrations of Lemma 8.1)

We completed several examples illustrating Lemma 8.1 in Preview Activity 2. For
another example, let c D 56 and d D 12. The greatest common divisor of 56 and
12 is 4.

1. According to the Division Algorithm, what is the remainder r when 56 is

divided by 12?

2. What is the greatest common divisor of 12 and the remainder r?

8.1. The Greatest Common Divisor 419

The key to finding the greatest common divisor (in more complicated cases) is
to use the Division Algorithm again, this time with 12 and r. We now find integers
q2 and r2 such that
12 D r
q2 C r2 :

3. What is the greatest common divisor of r and r2?

The Euclidean Algorithm

The example in Progress Check 8.2 illustrates the main idea of the Euclidean Al-
gorithm for finding gcd.a; b/, which is explained in the proof of the following
theorem.
Theorem 8.3. Let a and b be integers with a ¤ 0 and b > 0. Then gcd.a; b/ is
the only natural number d such that

(a) d divides a and d divides b, and

(b) if k is an integer that divides both a and b, then k divides d .

Proof. Let a and b be integers with a ¤ 0 and b > 0, and let d D gcd.a; b/. By
the Division Algorithm, there exist integers q1 and r1 such that

a D b
q1 C r1 , and 0  r1 < b: (1)

If r1 D 0, then equation (1) implies that b divides a. Hence, b D d D gcd.a; b/

and this number satisfies Conditions (a) and (b).
If r1 > 0, then by Lemma 8.1, gcd.a; b/ D gcd.b; r1 /. We use the Division
Algorithm again to obtain integers q2 and r2 such that

b D r1
q2 C r2 , and 0  r2 < r1 : (2)

If r2 D 0, then equation (2) implies that r1 divides b. This means that r1 D

gcd.b; r1 /. But we have already seen that gcd.a; b/ D gcd.b; r1 /. Hence, r1 D
gcd.a; b/. In addition, if k is an integer that divides both a and b, then, using
equation (1), we see that r1 D a b
q1 and, hence k divides r1 . This shows that
r1 D gcd.a; b/ satisfies Conditions (a) and (b).
If r2 > 0, then by Lemma 8.1, gcd.b; r1 / D gcd.r1 ; r2 /. But we have already
seen that gcd.a; b/ D gcd.b; r1 /. Hence, gcd.a; b/ D gcd.r1 ; r2 /. We now con-
tinue to apply the Division Algorithm to produce a sequence of pairs of integers
420 Chapter 8. Topics in Number Theory

(all of which have the same greatest common divisor). This is summarized in the
following table:

Original Equation from Inequality from New

Pair Division Algorithm Division Algorithm Pair
.a; b/ a D b
q1 C r1 0  r1 <b .b; r1 /
.b; r1 / b D r1
q2 C r2 0  r2 < r1 .r1 ; r2 /
.r1 ; r2 / r1 D r2
q3 C r3 0  r3 < r2 .r2 ; r3 /
.r2 ; r3 / r2 D r3
q4 C r4 0  r4 < r3 .r3 ; r4 /
.r3 ; r4 / r3 D r4
q5 C r5 0  r5 < r4 .r4 ; r5 /
:: :: :: ::
: : : :

From the inequalities in the third column of this table, we have a strictly decreasing
sequence of nonnegative integers .b > r1 > r2 > r3 > r4


/. Consequently, a
term in this sequence must eventually be equal to zero. Let p be the smallest
natural number such that rpC1 D 0. This means that the last two rows in the
preceding table will be

Original Equation from Inequality from New

Pair Division Algorithm Division Algorithm Pair



rp 2 ; rp 1 rp 2 D rp 1
qp C rp 0  rp < rp 1 rp 1 ; rp


rp 1 ; rp rp 1 D rp
qpC1 C 0

Remember that this table was constructed by repeated use of Lemma 8.1 and that
the greatest common divisor of each pair of integers produced equals gcd.a; b/.
Also, the last row in the table indicates that rp divides rp 1 . This means that
gcd.rp 1 ; rp / D rp and hence rp D gcd.a; b/.
This proves that rp D gcd.a; b/ satisfies Condition (a) of this theorem. Now
assume that k is an integer such that k divides a and k divides b. We proceed
through the table row by row. First, since r1 D a b
q, we see that

k must divide r1 .

The second row tells us that r2 D b r1
q2 . Since k divides b and k divides r1 ,

we conclude that

k divides r2 .
8.1. The Greatest Common Divisor 421

Continuing with each row, we see that k divides each of the remainders r1 , r2 , r3 ,
: : : ; rp . This means that rp D gcd.a; b/ satisfies Condition (b) of the theorem. 

Progress Check 8.4 (Using the Euclidean Algorithm)

1. Use the Euclidean Algorithm to determine gcd.180; 126/. Notice that we

have deleted the third column (Inequality from Division Algorithm) from
the following table. It is not needed in the computations.

Original Equation from New

Pair Division Algorithm Pair
.180; 126/ 180 D 126
1 C 54 .126; 54/
.126; 54/ 126 D

Consequently, gcd.180; 126/ D .

2. Use the Euclidean Algorithm to determine gcd.4208; 288/.

Original Equation from New

Pair Division Algorithm Pair
.4208; 288/ 4208 D 288
14 C 176 .288; /

Consequently, gcd.4208; 288/ D .

Some Remarks about Theorem 8.3

Theorem 8.3 was proven with the assumptions that a; b 2 Z with a ¤ 0 and b > 0.
A more general version of this theorem can be proven with a; b 2 Z and b ¤ 0.
This can be proven using Theorem 8.3 and the results in the following lemma.
Lemma 8.5. Let a; b 2 Z with b ¤ 0. Then

1. gcd.0; b/ D jbj.

2. If gcd.a; b/ D d , then gcd.a; b/ D d .

The proofs of these results are in Exercise (4). An application of this result is given
in the next example.
422 Chapter 8. Topics in Number Theory

Example 8.6 (Using the Euclidean Algorithm)

Let a D 234 and b D 42. We will use the Euclidean Algorithm to determine
gcd.234; 42/.

Step Original Equation from New

Pair Division Algorithm Pair
1 .234; 42/ 234 D 42
5 C 24 .42; 24/
2 .42; 24/ 42 D 24
1 C 18 .24; 18/
3 .24; 18/ 24 D 18
1 C 6 .18; 6/
4 .18; 6/ 18 D 6
3

So gcd.234; 42/ D 6 and hence gcd.234; 42/ D 6.

Writing gcd.a; b/ in Terms of a and b

We will use Example 8.6 to illustrate another use of the Euclidean Algorithm. It
is possible to use the steps of the Euclidean Algorithm in reverse order to write
gcd.a; b/ in terms of a and b. We will use these steps in reverse order to find
integers m and n such that gcd.234; 42/ D 234m C 42n. The idea is to start
with the row with the last nonzero remainder and work backward as shown in the
following table:

Explanation Result
First, use the equation in Step 3 to 6 D 24 18
1
write 6 in terms of 24 and 18.
Use the equation in Step 2 to write 6 D 24 18
1
18 D 42 24
1. Substitute this into D 24 .42 24
1/
the preceding result and simplify. D 42
. 1/ C 24
2
We now have written 6 in terms of 6 D 42
. 1/ C 24
2
42 and 24. Use the equation in D 42
. 1/ C .234 42
5/
2
Step 1 to write 24 D 234 42
5. D 234
2 C 42
. 11/
Substitute this into the preceding
result and simplify.

Hence, we can write

gcd.234; 42/ D 234
2 C 42
. 11/:
(Check this with a calculator.) In this case, we say that we have written
gcd.234; 42/ as a linear combination of 234 and 42. More generally, we have
the following definition.
8.1. The Greatest Common Divisor 423

Definition. Let a and b be integers. A linear combination of a and b is an

integer of the form ax C by, where x and y are integers.
Progress Check 8.7 (Writing the gcd as a Linear Combination)
Use the results from Progress Check 8.4 to

1. Write gcd.180; 126/ as a linear combination of 180 and 126.

2. Write gcd.4208; 288/ as a linear combination of 4208 and 288.

The previous example and progress check illustrate the following important
result in number theory, which will be used in the next section to help prove some
other significant results.
Theorem 8.8. Let a and b be integers, not both 0. Then gcd.a; b/ can be written
as a linear combination of a and b. That is, there exist integers u and v such that
gcd.a; b/ D au C bv.

We will not give a formal proof of this theorem. Hopefully, the examples
and activities provide evidence for its validity. The idea is to use the steps of the
Euclidean Algorithm in reverse order to write gcd.a; b/ as a linear combination of
a and b. For example, assume the completed table for the Euclidean Algorithm is

Step Original Equation from New

Pair Division Algorithm Pair
1 .a; b/ a D b
q1 C r1 .b; r1 /
2 .b; r1 / b D r1
q2 C r2 .r1 ; r2 /
3 .r1 ; r2 / r1 D r2
q3 C r3 .r2 ; r3 /
4 .r2 ; r3 / r2 D r3
q4 C 0 .r3 ; r4 /

We can use Step 3 to write r3 D gcd.a; b/ as a linear combination of r1 and r2 . We

can then solve the equation in Step 2 for r2 and use this to write r3 D gcd.a; b/ as
a linear combination of r1 and b. We can then use the equation in Step 1 to solve
for r1 and use this to write r3 D gcd.a; b/ as a linear combination of a and b.
In general, if we can write rp D gcd.a; b/ as a linear combination of a pair
in a given row, then we can use the equation in the preceding step to write rp D
gcd.a; b/ as a linear combination of the pair in this preceding row.
The notational details of this induction argument get quite involved. Many
mathematicians prefer to prove Theorem 8.8 using a property of the natural num-
bers called the Well-Ordering Principle. The Well-Ordering Principle for the
424 Chapter 8. Topics in Number Theory

natural numbers states that any nonempty set of natural numbers must contain a
least element. It can be proven that the Well-Ordering Principle is equivalent to the
Principle of Mathematical Induction.

Exercises 8.1

1. Find each of the following greatest common divisors by listing all of the
common divisors of each pair of integers.
? (a) gcd.21; 28/ ? (c) gcd.58; 63/ (e) gcd.110; 215/
? (b) gcd. 21; 28/ ? (d) gcd.0; 12/ (f) gcd.110; 215/

2. ? (a) Let a 2 Z and let k 2 Z with k ¤ 0. Prove that if k j a and k j .a C 1/,

then k j 1, and hence k D ˙1.
(b) Let a 2 Z. Find the greatest common divisor of the consecutive inte-
gers a and a C 1. That is, determine gcd.a; a C 1/.
3. (a) Let a 2 Z and let k 2 Z with k ¤ 0. Prove that if k j a and k j .a C 2/,
then k j 2.
(b) Let a 2 Z. What conclusions can be made about the greatest common
divisor of a and a C 2?
? 4. Let a; b 2 Z with b ¤ 0. Prove each of the following:

(a) gcd.0; b/ D jbj

(b) If gcd.a; b/ D d , then gcd.a; b/ D d . That is,
gcd .a; b/ D gcd.a; b/.

5. For each of the following pairs of integers, use the Euclidean Algorithm to
find gcd.a; b/ and to write gcd.a; b/ as a linear combination of a and b. That
is, find integers m and n such that d D am C bn.
?
(a) a D 36; b D 60 (d) a D 12628; b D 21361
?
(b) a D 901; b D 935 (e) a D 36, b D 60
?
(c) a D 72; b D 714 (f) a D 901, b D 935

6. (a) Find integers u and v such that 9u C 14v D 1 or explain why it is not
possible to do so. Then find integers x and y such that 9x C 14y D 10
or explain why it is not possible to do so.
8.1. The Greatest Common Divisor 425

(b) Find integers x and y such that 9x C 15y D 10 or explain why it is not
possible to do so.
(c) Find integers x and y such that 9x C 15y D 3162 or explain why it is
not possible to do so.

7. ? (a) Notice that gcd.11; 17/ D 1. Find integers x and y such that
11x C 17y D 1.
? (b) Let m; n 2 Z. Write the sum m C n as a single fraction.
11 17
(c) Find two rational numbers with denominators of 11 and 17, respec-
10
tively , whose sum is equal to . Hint: Write the rational numbers
187
m n
in the form and , where m; n 2 Z. Then write
11 17
m n 10
C D :
11 17 187
Use Exercises (7a) and (7b) to determine m and n.
(d) Find two rational numbers with denominators 17 and 21, respectively,
326
whose sum is equal to or explain why it is not possible to do so.
357
(e) Find two rational numbers with denominators 9 and 15, respectively,
10
whose sum is equal to or explain why it is not possible to do so.
225

Explorations and Activities

8. Linear Combinations and the Greatest Common Divisor

(a) Determine the greatest common divisor of 20 and 12?

(b) Let d D gcd.20; 12/. Write d as a linear combination of 20 and 12.
(c) Generate at least six different linear combinations of 20 and 12. Are
these linear combinations of 20 and 12 multiples of gcd.20; 12/?
(d) Determine the greatest common divisor of 21 and 6 and then generate
at least six different linear combinations of 21 and 6. Are these linear
combinations of 21 and 6 multiples of gcd.21; 6/?
(e) The following proposition was first introduced in Exercise (18) on
page 243 in Section 5.2. Complete the proof of this proposition if you
have not already done so.
426 Chapter 8. Topics in Number Theory

Proposition 5.16 Let a, b, and t be integers with t ¤ 0. If t divides a

and t divides b, then for all integers x and y, t divides (ax + by).
Proof . Let a, b, and t be integers with t ¤ 0, and assume that t divides
a and t divides b. We will prove that for all integers x and y, t divides
.ax C by/.
So let x 2 Z and let y 2 Z. Since t divides a, there exists an integer m
such that : : : :
(f) Now let a and b be integers, not both zero and let d D gcd.a; b/. The-
orem 8.8 states that d is a linear combination of a and b. In addition,
let S and T be the following sets:

S D fax C by j x; y 2 Zg and T D fkd j k 2 Zg:

That is, S is the set of all linear combinations of a and b, and T is the
set of all multiples of the greatest common divisor of a and b. Does the
set S equal the set T ? If not, is one of these sets a subset of the other
set? Justify your conclusions.
Note: In Parts (c) and (d), we were exploring special cases for these
two sets.

8.2 Prime Numbers and Prime Factorizations

Preview Activity 1 (Exploring Examples where a Divides b
c)

1. Find at least three different examples of nonzero integers a, b, and c such

that a j .bc/ but a does not divide b and a does not divide c. In each case,
compute gcd.a; b/ and gcd.a; c/.
2. Find at least three different examples of nonzero integers a, b, and c such
that gcd.a; b/ D 1 and a j .bc/. In each example, is there any relation
between the integers a and c?
3. Formulate a conjecture based on your work in Parts (1) and (2).

Preview Activity 2 (Prime Factorizations)

Recall that a natural number p is a prime number provided that it is greater than
1 and the only natural numbers that divide p are 1 and p. A natural number other
than 1 that is not a prime number is a composite number. The number 1 is neither
prime nor composite. (See Exercise 13 from Section 2.4 on page 78.)
8.2. Prime Numbers and Prime Factorizations 427

1. Give examples of four natural numbers that are prime and four natural num-
bers that are composite.

Theorem 4.9 in Section 4.2 states that every natural number greater than 1 is
either a prime number or a product of prime numbers.
When a composite number is written as a product of prime numbers, we say
that we have obtained a prime factorization of that composite number. For exam-
ple, since 60 D 22
3
5, we say that 22
3
5 is a prime factorization of 60.

2. Write the number 40 as a product of prime numbers by first writing 40 D

2
20 and then factoring 20 into a product of primes. Next, write the number
40 as a product of prime numbers by first writing 40 D 5
8 and then factoring
8 into a product of primes.

3. In Part (2), we used two different methods to obtain a prime factorization

of 40. Did these methods produce the same prime factorization or different
prime factorizations? Explain.

4. Repeat Parts (2) and (3) with 150. First, start with 150 D 3
50, and then
start with 150 D 5
30.

Greatest Common Divisors and Linear Combinations

In Section 8.1, we introduced the concept of the greatest common divisor of two
integers. We showed how the Euclidean Algorithm can be used to find the greatest
common divisor of two integers, a and b, and also showed how to use the results
of the Euclidean Algorithm to write the greatest common divisor of a and b as a
linear combination of a and b.
In this section, we will use these results to help prove the so-called Fundamen-
tal Theorem of Arithmetic, which states that any natural number greater than 1 that
is not prime can be written as product of primes in “essentially” only one way. This
means that given two prime factorizations, the prime factors are exactly the same,
and the only difference may be in the order in which the prime factors are written.
We start with more results concerning greatest common divisors. We first prove
Proposition 5.16, which was part of Exercise (18) on page 243 in Section 5.2 and
Exercise (8) on page 425in Section 8.1.
Proposition 5.16 Let a, b, and t be integers with t ¤ 0. If t divides a and t divides
b, then for all integers x and y, t divides (ax + by).
428 Chapter 8. Topics in Number Theory

Proof. Let a, b, and t be integers with t ¤ 0, and assume that t divides a and t
divides b. We will prove that for all integers x and y, t divides .ax C by/.
So let x 2 Z and let y 2 Z. Since t divides a, there exists an integer m such
that a D mt and since t divides b, there exists an integer n such that b D nt .
Using substitution and algebra, we then see that
ax C by D .mt /x C .nt /y
D t .mx C ny/
Since .mx C ny/ is an integer, the last equation proves that t divides ax C by and
this proves that for all integers x and y, t divides .ax C by/. 

We now let a; b 2 Z, not both 0, and let d D gcd.a; b/. Theorem 8.8 states
that d can be written as a linear combination of a and b. Now, since d j a and
d j b, we can use the result of Proposition 5.16 to conclude that for all x; y 2 Z,
d j .ax C by/. This means that d divides every linear combination of a and b.
In addition, this means that d must be the smallest positive number that is a linear
combination of a and b. We summarize these results in Theorem 8.9.
Theorem 8.9. Let a; b 2 Z, not both 0.

1. The greatest common divisor, d , is a linear combination of a and b. That is,

there exist integers m and n such that d D am C bn.
2. The greatest common divisor, d , divides every linear combination of a and
b. That is, for all x; y 2 Z , d j .ax C by/.
3. The greatest common divisor, d , is the smallest positive number that is a
linear combination of a and b.

Relatively Prime Integers

In Preview Activity 1, we constructed several examples of integers a, b, and c such

that a j .bc/ but a does not divide b and a does not divide c. For each example,
we observed that gcd.a; b/ ¤ 1 and gcd.a; c/ ¤ 1.
We also constructed several examples where a j .bc/ and gcd.a; b/ D 1. In all
of these cases, we noted that a divides c. Integers whose greatest common divisor
is equal to 1 are given a special name.

Definition. Two nonzero integers a and b are relatively prime provided that
gcd.a; b/ D 1.
8.2. Prime Numbers and Prime Factorizations 429

Progress Check 8.10 (Relatively Prime Integers)

1. Construct at least three different examples where p is a prime number, a 2

Z , and p j a. In each example, what is gcd.a; p/? Based on these examples,
formulate a conjecture about gcd.a; p/ when p j a.

2. Construct at least three different examples where p is a prime number, a 2

Z , and p does not divide a. In each example, what is gcd.a; p/? Based
on these examples, formulate a conjecture about gcd.a; p/ when p does not
divide a.

3. Give at least three different examples of integers a and b where a is not

prime, b is not prime, and gcd.a; b/ D 1, or explain why it is not possible to
construct such examples.

Theorem 8.11. Let a and b be nonzero integers, and let p be a prime number.

1. If a and b are relatively prime, then there exist integers m and n such that
am C bn D 1. That is, 1 can be written as linear combination of a and b.

2. If p j a, then gcd.a; p/ D p.

3. If p does not divide a, then gcd.a; p/ D 1.

Part (1) of Theorem 8.11 is actually a corollary of Theorem 8.9. Parts (2)
and (3) could have been the conjectures you formulated in Progress Check 8.10.
The proofs are included in Exercise (1).

Given nonzero integers a and b, we have seen that it is possible to use the
Euclidean Algorithm to write their greatest common divisor as a linear combination
of a and b. We have also seen that this can sometimes be a tedious, time-consuming
process, which is why people have programmed computers to do this. Fortunately,
in many proofs of number theory results, we do not actually have to construct this
linear combination since simply knowing that it exists can be useful in proving
results. This will be illustrated in the proof of Theorem 8.12, which is based on
work in Preview Activity 1.

Theorem 8.12. Let a, b, be nonzero integers and let c be an integer. If a and b are
relatively prime and a j .bc/, then a j c.
430 Chapter 8. Topics in Number Theory

The explorations in Preview Activity 1 were related to this theorem. We will

first explore the forward-backward process for the proof. The goal is to prove that
a j c. A standard way to do this is to prove that there exists an integer q such that

c D aq: (1)

Since we are given that a j .bc/, there exists an integer k such that

bc D ak: (2)

It may seem tempting to divide both sides of equation (2) by b, but if we do so,
we run into problems with the fact that the integers are not closed under division.
Instead, we look at the other part of the hypothesis, which is that a and b are
relatively prime. This means that gcd.a; b/ D 1. How can we use this? This
means that a and b have no common factors except for 1. In light of equation (2),
it seems reasonable that any factor of a must also be a factor of c. But how do we
formalize this?
One conclusion that we can use is that since gcd.a; b/ D 1, by Theorem 8.11,
there exist integers m and n such that

am C bn D 1: (3)

We may consider solving equation (3) for b and substituting this into equa-
tion (2). The problem, again, is that in order to solve equation (3) for b, we need
to divide by n.
Before doing anything else, we should look at the goal in equation (1). We
need to introduce c into equation (3). One way to do this is to multiply both sides
of equation (3) by c. (This keeps us in the system of integers since the integers are
closed under multiplication.) This gives

.am C bn/ c D 1
c
acm C bcn D c: (4)

Notice that the left side of equation (4) contains a term, bcn, that contains bc. This
means that we can use equation (2) and substitute bc D ak in equation (4). After
doing this, we can factor the left side of the equation to prove that a j c.

Progress Check 8.13 (Completing the Proof of Theorem 8.12)

Write a complete proof of Theorem 8.12.
8.2. Prime Numbers and Prime Factorizations 431

Corollary 8.14.

1. Let a; b 2 Z, and let p be a prime number. If p j .ab/, then p j a or p j b.

2. Let p be a prime number, let n 2 N, and let a1 ; a2 ; : : : ; an 2 Z. If

p j .a1 a2


an /, then there exists a natural number k with 1  k  n
such that p j ak .

Part (1) of Corollary 8.14 is a corollary of Theorem 8.12. Part (2) is proved us-
ing mathematical induction. The basis step is the case where n D 1, and Part (1) is
the case where n D 2. The proofs of these two results are included in Exercises (2)
and (3).

Historical Note

Part (1) of Corollary 8.14 is known as Euclid’s Lemma. Most people associate
geometry with Euclid’s Elements, but these books also contain many basic results
in number theory. Many of the results that are contained in this section appeared
in Euclid’s Elements.

Prime Numbers and Prime Factorizations

We are now ready to prove the Fundamental Theorem of Arithmetic. The first part
of this theorem was proved in Theorem 4.9 in Section 4.2. This theorem states
that each natural number greater than 1 is either a prime number or is a product of
prime numbers. Before we state the Fundamental Theorem of Arithmetic, we will
discuss some notational conventions that will help us with the proof. We start with
an example.
We will use n D 120. Since 5 j 120, we can write 120 D 5
24. In addition,
we can factor 24 as 24 D 2
2
2
3. So we can write

120 D 5
24
D 5 .2
2
2
3/ :

This is a prime factorization of 120, but it is not the way we usually write this
factorization. Most often, we will write the prime number factors in ascending
order. So we write

120 D 2
2
2
3
5 or 120 D 23
3
5:
432 Chapter 8. Topics in Number Theory

Now, let n 2 N. To write the prime factorization of n with the prime factors in
ascending order requires that if we write n D p1 p2


pr , where p1 ; p2 ;


pr are
prime numbers, we will have p1  p2 


 pr .

Theorem 8.15 (The Fundamental Theorem of Arithmetic).

1. Each natural number greater than 1 is either a prime number or is a product

of prime numbers.

2. Let n 2 N with n > 1. Assume that

n D p1 p2


pr and that n D q1 q2


qs ;

where p1 ; p2 ;


pr and q1 ; q2 ;


qs are primes with p1  p2 


 pr

and q1  q2 


 qs . Then r D s, and for each j from 1 to r, pj D qj .

Proof. The first part of this theorem was proved in Theorem 4.9. We will prove
the second part of the theorem by induction on n using the Second Principle of
Mathematical Induction. (See Section 4.2.) For each natural number n with n > 1,
let P .n/ be

If n D p1 p2


pr and n D q1 q2


qs , where p1 ; p2 ;


pr and q1 ; q2;


qs

are primes with p1  p2 


 pr and q1  q2 


 qs , then r D s,
and for each j from 1 to r, pj D qj .
For the basis step, we notice that since 2 is a prime number, its only factoriza-
tion is 2 D 1
2. This means that the only equation of the form 2 D p1 p2


pr ,
where p1 ; p2 ;


pr are prime numbers, is the case where r D 1 and p1 D 2. This
proves that P .2/ is true.
For the inductive step, let k 2 N with k  2. We will assume that
P .2/; P .3/; : : : ; P .k/ are true. The goal now is to prove that P .k C 1/ is true.
To prove this, we assume that .k C 1/ has two prime factorizations and then prove
that these prime factorizations are the same. So we assume that

k C 1 D p1 p2


pr and that k C 1 D q1 q2


qs , where p1 ; p2 ;


pr and

q1 ; q2 ;


qs are primes with p1  p2 


 pr and q1  q2 


 qs .

We must now prove that r D s, and for each j from 1 to r, pj D qj . We can

break our proof into two cases: (1) p1  q1 ; and (2) q1  p1 . Since one of these
must be true, and since the proofs will be similar, we can assume, without loss of
generality, that p1  q1 .
8.2. Prime Numbers and Prime Factorizations 433

Since k C 1 D p1 p2


pr , we know that p1 j .k C 1/, and hence we may

conclude that p1 j .q1 q2


qs /. We now use Corollary 8.14 to conclude that there
exists a j with 1  j  s such that p1 j qj . Since p1 and qj are primes, we
conclude that
p1 D qj :

We have also assumed that q1  qj for all j and, hence, we know that q1  p1 .
However, we have also assumed that p1  q1 . Hence,

p1 D q1 :

We now use this and the fact that k C 1 D p1 p2


pr D q1 q2


qs to conclude

that
p2


pr D q2


qs :

The product in the previous equation is less that k C 1. Hence, we can apply our
induction hypothesis to these factorizations and conclude that r D s, and for each
j from 2 to r, pj D qj .
This completes the proof that if P .2/; P .3/; : : : ; P .k/ are true, then P .k C 1/
is true. Hence, by the Second Principle of Mathematical Induction, we conclude
that P .n/ is true for all n 2 N with n  2. This completes the proof of the
theorem. 

Note: : We often shorten the result of the Fundamental Theorem of Arithmetic by

simply saying that each natural number greater than one that is not a prime has a
unique factorization as a product of primes. This simply means that if n 2 N,
n > 1, and n is not prime, then no matter how we choose to factor n into a product
of primes, we will always have the same prime factors. The only difference may
be in the order in which we write the prime factors.

Further Results and Conjectures about Prime Numbers

1. The Number of Prime Numbers

Prime numbers have fascinated mathematicians for centuries. For example,
we can easily start writing a list of prime numbers in ascending order. Fol-
lowing is a list of the prime numbers less than 100.
2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71,
73, 79, 83, 89, 97
434 Chapter 8. Topics in Number Theory

This list contains the first 25 prime numbers. Does this list ever stop? The
question was answered in Euclid’s Elements, and the result is stated in The-
orem 8.16. The proof of this theorem is considered to be one of the classical
proofs by contradiction.

Theorem 8.16. There are infinitely many prime numbers.

Proof. We will use a proof by contradiction. We assume that there are only
finitely many primes, and let

p1 ; p2 ; : : : ; pm

be the list of all the primes. Let

M D p1 p2


pm C 1: (1)

Notice that M ¤ 1. So M is either a prime number or, by the Fundamental

Theorem of Arithmetic, M is a product of prime numbers. In either case,
M has a factor that is a prime number. Since we have listed all the prime
numbers, this means that there exists a natural number j with 1  j  m
such that pj j M . Now, we can rewrite equation (1) as follows:

1DM p1 p2


pm : (2)

We have proved pj j M , and since pj is one of the prime factors of p1 p2


pm ,

we can also conclude that pj j .p1 p2


pm /. Since pj divides both of the
terms on the right side of equation (2), we can use this equation to conclude
that pj divides 1. This is a contradiction since a prime number is greater
than 1 and cannot divide 1. Hence, our assumption that there are only finitely
many primes is false, and so there must be infinitely many primes. 

2. The Distribution of Prime Numbers

There are infinitely many primes, but when we write a list of the prime num-
bers, we can see some long sequences of consecutive natural numbers that
contain no prime numbers. For example, there are no prime numbers be-
tween 113 and 127. The following theorem shows that there exist arbitrarily
long sequences of consecutive natural numbers containing no prime num-
bers. A guided proof of this theorem is included in Exercise (15).

Theorem 8.17. For any natural number n, there exist at least n consecutive
natural numbers that are composite numbers.
8.2. Prime Numbers and Prime Factorizations 435

There are many unanswered questions about prime numbers, two of which will
now be discussed.

3. The Twin Prime Conjecture

By looking at the list of the first 25 prime numbers, we see several cases
where consecutive prime numbers differ by 2. Examples are: 3 and 5; 11
and 13; 17 and 19; 29 and 31. Such pairs of prime numbers are said to be
twin primes. How many twin primes exist? The answer is not known. The
Twin Prime Conjecture states that there are infinitely many twin primes.
As of June 25, 2010, this is still a conjecture as it has not been proved or
disproved.
For some interesting information on prime numbers, visit the Web site The
Prime Pages (http://primes.utm.edu/), where there is a link to The Largest
Known Primes Web site. According to information at this site as of June 25,
2010, the largest known twin primes are



65516468355  2333333 1 and 65516468355  2333333 C 1 :

Each of these prime numbers contains 100355 digits.

4. Goldbach’s Conjecture
Given an even natural number, is it possible to write it as a sum of two prime
numbers? For example,

4D 2C2 6D 3C3 8 D 5C3

78 D 37 C 41 90 D 43 C 47 138 D 67 C 71

One of the most famous unsolved problems in mathematics is a conjecture

made by Christian Goldbach in a letter to Leonhard Euler in 1742. The
conjecture, now known as Goldbach’s Conjecture, is as follows:
Every even integer greater than 2 can be expressed as the sum of two
(not necessarily distinct) prime numbers.
As of June 25, 2010, it is not known if this conjecture is true or false, al-
though most mathematicians believe it to be true.

Exercises 8.2
?
1. Prove the second and third parts of Theorem 8.11.
436 Chapter 8. Topics in Number Theory

(a) Let a be a nonzero integer, and let p be a prime number. If p j a, then

gcd.a; p/ D p.
(b) Let a be a nonzero integer, and let p be a prime number. If p does not
divide a, then gcd.a; p/ D 1.

? 2. Prove the first part of Corollary 8.14.

Let a; b 2 Z, and let p be a prime number. If p j .ab/, then p j a or p j b.
Hint: Consider two cases: (1) p j a; and (2) p does not divide a.

?
3. Use mathematical induction to prove the second part of Corollary 8.14.
Let p be a prime number, let n 2 N, and let a1 ; a2 ; : : : ; an 2 Z. If
p j .a1 a2


an /, then there exists a k 2 N with 1  k  n such that
p j ak .

? 4. (a) Let a and b be nonzero integers. If there exist integers x and y such that
ax C by D 1, what conclusion can be made about gcd.a; b/? Explain.
(b) Let a and b be nonzero integers. If there exist integers x and y such that
ax C by D 2, what conclusion can be made about gcd.a; b/? Explain.

5. (a) Let a 2 Z. What is gcd.a; a C 1/? That is, what is the greatest com-
mon divisor of two consecutive integers? Justify your conclusion.
Hint: Exercise (4) might be helpful.
(b) Let a 2 Z. What conclusion can be made about gcd.a; a C 2/? That
is, what conclusion can be made about the greatest common divisor of
two integers that differ by 2? Justify your conclusion.

6. (a) Let a 2 Z. What conclusion can be made about gcd.a; a C 3/? That
is, what conclusion can be made about the greatest common divisor of
two integers that differ by 3? Justify your conclusion.
(b) Let a 2 Z. What conclusion can be made about gcd.a; a C 4/? That
is, what conclusion can be made about the greatest common divisor of
two integers that differ by 4? Justify your conclusion.

7. ? (a) Let a D 16 and b D 28. Determine the

 value of d D gcd.a; b/, and
a b
then determine the value of gcd ; .
d d
(b) Repeat Exercise (7a) with a D 10 and b D 45.
8.2. Prime Numbers and Prime Factorizations 437

(c) Let a; b 2 Z, not both equal to 0, and let d 

D gcd.a;
 b/. Explain why
a b a b
and are integers. Then prove that gcd ; D 1. Hint: Start
d d d d
by writing d as a linear combination of a and b.
This says that if you divide both a and b by their greatest common divisor,
the result will be two relatively prime integers.

8. Are the following propositions true or false? Justify your conclusions.

(a) For all integers a; b; and c, if a j c and b j c, then .ab/ j c.

(b) For all integers a; b; and c, if a j c, b j c and gcd.a; b/ D 1, then
.ab/ j c.
? 9. In Exercise
(16) in Section 3.5, it was proved that if n is an odd integer, then
8 j n2 1 . (This result was also proved in Exercise (19) in Section 7.4.)
Now, prove the following proposition:


If n is an odd integer and 3 does not divide n, then 24 j n2 1 .

10. (a) Prove the following proposition:

For all a; b; c 2 Z, gcd.a; bc/ D 1 if and only if gcd.a; b/ D 1 and
gcd.a; c/ D 1.
(b) Use mathematical induction to prove the following proposition:
Let n 2 N and let a; b1 ; b2 ; : : : ; bn 2 Z. If gcd.a; bi / D 1 for all i 2 N
with 1  i  n, then gcd.a; b1 b2


bn / D 1.
? 11. Is the following proposition true or false? Justify your conclusion.
For all integers a; b; and c, if gcd.a; b/ D 1 and c j .a C b/, then
gcd.a; c/ D 1 and gcd.b; c/ D 1.

12. Is the following proposition true or false? Justify your conclusion.

If n 2 N, then gcd.5n C 2; 12n C 5/ D 1.

13. Let y 2 N. Use the Fundamental Theorem of Arithmetic to prove that

there exists an odd natural number x and a nonnegative integer k such that
y D 2k x.

14. (a) Determine five different primes that are congruent to 3 modulo 4.
(b) Prove that there are infinitely many primes that are congruent to 3 mod-
ulo 4.

15. (a) Let n 2 N. Prove that 2 divides Œ.n C 1/Š C 2.

438 Chapter 8. Topics in Number Theory

(b) Let n 2 N with n  2. Prove that 3 divides Œ.n C 1/Š C 3.

(c) Let n 2 N. Prove that for each k 2 N with 2  k  .n C 1/, k divides
Œ.n C 1/Š C k.
(d) Use the result of Exercise (15c) to prove that for each n 2 N, there
exist at least n consecutive composite natural numbers.

16. The Twin Prime Conjecture states that there are infinitely many twin primes,
but it is not known if this conjecture is true or false. The answers to the
following questions, however, can be determined.

(a) How many pairs of primes p and q exist where q p D 3? That is,
how many pairs of primes are there that differ by 3? Prove that your
answer is correct. (One such pair is 2 and 5.)
(b) How many triplets of primes of the form p, p C 2, and p C 4 are there?
That is, how many triplets of primes exist where each prime is 2 more
than the preceding prime? Prove that your answer is correct. Notice
that one such triplet is 3, 5, and 7.
Hint: Try setting up cases using congruence modulo 3.

17. Prove the following proposition:

Let n 2 N. For each a 2 Z, if gcd.a; n/ D 1, then for every b 2 Z,
there exists an x 2 Z such that ax  b .mod n/.
Hint: One way is to start by writing 1 as a linear combination of a and n.

18. Prove the following proposition:

For all natural numbers m and n, if m and n are twin primes other than
the pair 3 and 5, then 36 divides mnC1 and mnC1 is a perfect square.
Hint: Look at several examples of twin primes. What do you notice about
the number that is between the two twin primes? Set up cases based on this
observation.

Explorations and Activities

19. Square Roots and Irrational

p pNumbers. In Chapter 3, we proved that some
square roots (such as 2 and 3) are irrational numbers. In this activity, we
will use the Fundamental Theorem of Arithmetic to prove that if a natural
number is not a perfect square, then its square root is an irrational number.
8.3. Linear Diophantine Equations 439

(a) Let n be a natural number. Use the Fundamental Theorem of Arith-

metic to explain why if n is composite, then there exist prime numbers
p1 ; p2 ; : : : ; pr and natural numbers ˛1 ; ˛2 ; : : : ; ˛r such that

n D p1˛1 p2˛2


pr˛r : (1)

Then, if we use r D 1 and ˛1 D 1 for a prime number, explain why we

can write any natural number in the form given in equation (1).
(b) A natural number b is a perfect square if and only if there exists a
natural number a such that b D a2 . Explain why 36, 400, and 15876
are perfect squares. Then determine the prime factorization of these
perfect squares. What do you notice about these prime factorizations?
(c) Let n be a natural number written in the form given in equation (1) in
part (a). Prove that n is a perfect square if and only if for each natural
number k with 1  k  r, ˛k is even.
(d) Prove that for all natural numbers n, if n is not a perfect square, then
p
n is an irrational number. Hint: Use a proof by contradiction.

8.3 Linear Diophantine Equations

Preview Activity 1 (Integer Solutions for Linear Equations in One Variable)

1. Does the linear equation 6x D 42 have a solutions that is an integer? Ex-

plain.

2. Does the linear equation 7x D 21 have a solution that is an integer? Ex-

plain.

3. Does the linear equation 4x D 9 have a solution that is an integer? Explain.

4. Does the linear equation 3x D 20 have a solution that is an integer? Ex-

plain.

5. Prove the following theorem:

Theorem 8.18. Let a; b 2 Z with a ¤ 0.

 If a divides b, then the equation ax D b has exactly one solution that

is an integer.
440 Chapter 8. Topics in Number Theory

 If a does not divide b, then the equation ax D b has no solution that is

an integer.

Preview Activity 2 (Linear Equations in Two Variables)

1. Find integers x and y so that 2x C 6y D 25 or explain why it is not possible

to find such a pair of integers.

2. Find integers x and y so that 6x 9y D 100 or explain why it is not possible

to find such a pair of integers.

3. Notice that x D 2 and y D 1 is a solution of the equation 3x C 5y D 11,

and that x D 7 and y D 2 is also a solution of the equation 3x C 5y D 11.

(a) Find two pairs of integers x and y so that x > 7 and 3x C 5y D 11.
(Try to keep the integer values of x as small as possible.)
(b) Find two pairs of integers x and y so that x < 2 and 3x C 5y D 11.
(Try to keep the integer values of x as close to 2 as possible.)
(c) Determine formulas (one for x and one for y) that will generate pairs
of integers x and y so that 3x C 5y D 11.
Hint: The two formulas can be written in the form x D 2 C km and
y D 1 C k n, where k is an arbitrary integer and m and n are specific
integers.

4. Notice that x D 4 and y D 0 is a solution of the equation 4x C 6y D 16,

and that x D 7 and y D 2 is a solution of the equation 4x C 6y D 16.

(a) Find two pairs of integers x and y so that x > 7 and 4x C 6y D 16.
(Try to keep the integer values of x as small as possible.)
(b) Find twopairs of integers x and y so that x < 4 and 4x C 6y D 16.
(Try to keep the integer values of x as close to 4 as possible.)
(c) Determine formulas (one for x and one for y) that will generate pairs
of integers x and y so that 4x C 6y D 16.
Hint: The two formulas can be written in the form x D 4 C km and
y D 0 C k n, where k is an arbitrary integer and mand n are specific
integers.

In the two preview activities, we were interested only in integer solutions for
certain equations. In such instances, we give the equation a special name.
8.3. Linear Diophantine Equations 441

Definition. An equation whose solutions are required to be integers is called

a Diophantine equation.

Diophantine equations are named in honor of the Greek mathematician Dio-

phantus of Alexandria (circa 300 c.e.). Very little is known about Diophantus’ life
except that he probably lived in Alexandria in the early part of the fourth century
c.e. and was probably the first to use letters for unknown quantities in arith-
metic problems. His most famous work, Arithmetica, consists of approximately
130 problems and their solutions. Most of these problems involved solutions of
equations in various numbers of variables. It is interesting to note that Diophantus
did not restrict his solutions to the integers but recognized rational number solu-
tions as well. Today, however, the solutions for a so-called Diophantine equation
must be integers.

Definition. If a and b are integers with a ¤ 0, then the equation ax D b is a

linear Diophantine equation in one variable.

Theorem 8.18 in Preview Activity 1 provides us with results that allows us to

determine which linear diophantine equations in one variable have solutions and
which ones do not have a solution.
A linear Diophantine equation in two variables can be defined in a manner
similar to the definition for a linear Diophantine equation in one variable.

Definition. Let a, b, and c be integers with a ¤ 0 and b ¤ 0. The Diophan-

tine equation ax C by D c is called a linear Diophantine equation in two
variables.

The equations that were investigated in Preview Activity 2 were linear Dio-
phantine equations in two variables. The problem of determining all the solutions
of a linear Diophantine equation has been completely solved. Before stating the
general result, we will provide a few more examples.

Example 8.19 (A Linear Diophantine Equation in Two Variables)

The following example is similar to the examples studied in Preview Activity 2.
We can use substitution to verify that x D 2 and y D 1 is a solution of the
linear Diophantine equation
4x C 3y D 5:
The following table shows other solutions of this Diophantine equation.
442 Chapter 8. Topics in Number Theory

x y x y
2 1 1 3
5 5 4 7
8 9 7 11
11 13 10 15

It would be nice to determine the pattern that these solutions exhibit. If we consider
the solution x D 2 and y D 1 to be the “starting point,” then we can see that
the other solutions are obtained by adding 3 to x and subtracting 4 from y in the
previous solution. So we can write these solutions to the equation as

x D 2 C 3k and yD 1 4k;

where k is an integer. We can use substitution and algebra to verify that these
expressions for x and y give solutions of this equation as follows:

4x C 3y D 4 .2 C 3k/ C 3 . 1 4k/
D .8 C 12k/ C . 3 12k/
D 5:

We should note that we have not yet proved that these solutions are all of the
solutions of the Diophantine equation 4x C 3y D 5. This will be done later.
If the general form for a linear Diophantine equation is ax C by D c, then for
this example, a D 4 and b D 3. Notice that for this equation, we started with one
solution and obtained other solutions by adding b D 3 to x and subtracting a D 4
from y in the previous solution. Also, notice that gcd.3; 4/ D 1.

Progress Check 8.20 (An Example of a Linear Diophantine Equation)

1. Verify that the following table shows some solutions of the linear Diophan-
tine equation 6x C 9y D 12.

x y x y
2 0 1 2
5 2 4 4
8 4 7 6
11 6 10 8

2. Follow the pattern in this table to determine formulas for x and y that will
generate integer solutions of the equation 6x C 9y D 12. Verify that the
formulas actually produce solutions for the equation 6x C 9y D 12.
8.3. Linear Diophantine Equations 443

Progress Check 8.21 (Revisiting Preview Activity 2)

Do the solutions for the linear Diophantine equations in Preview Activity 2 show
the same type of pattern as the solutions for the linear Diophantine equations in
Example 8.19 and Progress Check 8.20? Explain.

The solutions for the linear Diophantine equations in Preview Activity 2, Ex-
ample 8.19, and Progress Check 8.20 provide examples for the second part of The-
orem 8.22.
Theorem 8.22. Let a, b, and c be integers with a ¤ 0 and b ¤ 0, and let
d D gcd.a; b/.

1. If d does not divide c, then the linear Diophantine equation ax C by D c

has no solution.

2. If d divides c, then the linear Diophantine equation ax C by D c has in-

finitely many solutions. In addition, if .x0 ; y0 / is a particular solution of this
equation, then all the solutions of this equation can be written in the form
b a
x D x0 C k and y D y0 k;
d d
for some integer k.

Proof. The proof of Part (1) is Exercise (1). For Part (2), we let a, b, and c be
integers with a ¤ 0 and b ¤ 0, and let d D gcd.a; b/. We also assume that d j c.
Since d D gcd.a; b/, Theorem 8.8 tells us that d is a linear combination of a and
b. So there exist integers s and t such that

d D as C bt: (1)

Since d j c, there exists an integer m such that c D d m. We can now multiply

both sides of equation (1) by m and obtain

d m D .as C bt /m
c D a.sm/ C b.t m/:

This means that x D sm, y D t m is a solution of ax C by D c, and we have

proved that the Diophantine equation ax C by D c has at least one solution.
Now let x D x0 , y D y0 be any particular solution of ax C by D c, let k 2 Z,
and let
b a
x D x0 C k y D y0 k: (2)
d d
444 Chapter 8. Topics in Number Theory

We now verify that for each k 2 Z, the equations in (2) produce a solution of
ax C by D c.  
b  a 
ax C by D a x0 C k C b y0 k
d d
ab ab
D ax0 C k C by0 k
d d
D ax0 C by0
D c:
This proves that the Diophantine equation ax C by D c has infinitely many solu-
tions.
We now show that every solution of this equation can be written in the form
described in (2). So suppose that x and y are integers such that ax C by D c. Then

.ax C by/ .ax0 C by0 / D c c D 0;

and this equation can be rewritten in the following form:

a.x x0 / D b.y0 y/: (3)

Dividing both sides of this equation by d , we obtain

a  
b
.x x0 / D .y0 y/ :
d d
This implies that  
a b
divides .y0 y/:
d d
 
a b
However, by Exercise (7) in Section 8.2, gcd ; D 1, and so by Theo-
d d
a
rem 8.12, we can conclude that divides .y0 y/. This means that there exists
d
a
an integer k such that y0 y D k, and solving for y gives
d
a
y D y0 k:
d
Substituting this value for y in equation (3) and solving for x yields
b
x D x0 C k:
d
This proves that every solution of the Diophantine equation ax C by D c can be
written in the form prescribed in (2). 
8.3. Linear Diophantine Equations 445

The proof of the following corollary to Theorem 8.22 is Exercise (2).

Corollary 8.23. Let a, b, and c be integers with a ¤ 0 and b ¤ 0. If a and b are
relatively prime, then the linear Diophantine equation ax C by D c has infinitely
many solutions. In addition, if x0 , y0 is a particular solution of this equation, then
all the solutions of the equation are given by
x D x0 C bk y D y0 ak
where k 2 Z.

Progress Check 8.24 (Linear Diophantine Equations)

1. Use the Euclidean Algorithm to verify that gcd.63; 336/ D 21. What con-
clusion can be made about linear Diophantine equation 63x C 336y D 40
using Theorem 8.22? If this Diophantine equation has solutions, write for-
mulas that will generate the solutions.
2. Use the Euclidean Algorithm to verify that gcd.144; 225/ D 9. What con-
clusion can be made about linear Diophantine equation 144x C 225y D 27
using Theorem 8.22? If this Diophantine equation has solutions, write for-
mulas that will generate the solutions.

Exercises 8.3

1. Prove Part (1) of Theorem 8.22:

Let a, b, and c be integers with a ¤ 0 and b ¤ 0, and let d D gcd.a; b/. If
d does not divide c, then the linear Diophantine equation ax C by D c has
no solution.
2. Prove Corollary 8.23.
Let a, b, and c be integers with a ¤ 0 and b ¤ 0. If a and b are relatively
prime, then the linear Diophantine equation ax Cby D c has infinitely many
solutions. In addition, if .x0 ; y0 / is a particular solution of this equation, then
all the solutions of the equation are given by
x D x0 C bk y D y0 ak;
where k 2 Z.
3. Determine all solutions of the following linear Diophantine equations.
446 Chapter 8. Topics in Number Theory

(a) 9x C 14y D 1 (e) 200x C 49y D 10

(b) 18x C 22y D 4 (f) 200x C 54y D 21
(c) 48x 18y D 15 (g) 10x 7y D 31
(d) 12x C 9y D 6 (h) 12x C 18y D 6

4. A certain rare artifact is supposed to weigh exactly 25 grams. Suppose that

you have an accurate balance scale and 500 each of 27 gram weights and 50
gram weights. Explain how to use Theorem 8.22 to devise a plan to check
the weight of this artifact.
Hint: Notice that gcd.50; 27/ D 1. Start by writing 1 as a linear combination
of 50 and 27.

5. On the night of a certain banquet, a caterer offered the choice of two dinners,
a steak dinner for $25 and a vegetarian dinner for $16. At the end of the
evening, the caterer presented the host with a bill (before tax and tips) for
$1461. What is the minimum number of people who could have attended the
banquet? What is the maximum number of people who could have attended
the banquet?

6. The goal of this exercise is to determine all (integer) solutions of the linear
Diophantine equation in three variables 12x1 C 9x2 C 16x3 D 20.

(a) First, notice that gcd.12; 9/ D 3. Determine formulas that will gener-
ate all solutions for the linear Diophantine equation
3y C 16x3 D 20.
(b) Explain why the solutions (for x1 and x2 ) of the Diophantine equation
12x1 C 9x2 D 3y can be used to generate solutions for
12x1 C 9x2 C 16x3 D 20.
(c) Use the general value for y from Exercise (6a) to determine the solu-
tions of 12x1 C 9x2 D 3y.
(d) Use the results from Exercises (6a) and (6c) to determine formulas that
will generate all solutions for the Diophantine equation
12x1 C 9x2 C 16x3 D 20.
Note: These formulas will involve two arbitrary integer parameters.
Substitute specific values for these integers and then check the resulting
solution in the original equation. Repeat this at least three times.
(e) Check the general solution for 12x1 C 9x2 C 16x3 D 20 from Exer-
cise (6d).
8.3. Linear Diophantine Equations 447

7. Use the method suggested in Exercise (6) to determine formulas that will
generate all solutions of the Diophantine equation 8x1 C 4x2 6x3 D 6.
Check the general solution.

8. Explain why the Diophantine equation 24x1 18x2 C 60x3 D 21 has no

solution.

9. The purpose of this exercise will be to prove that the nonlinear Diophantine
equation 3x 2 y 2 D 2 has no solution.

(a) Explain why if there is a solution of the Diophantine equation 3x 2

y 2 D 2, then that solution must also be a solution of the congruence
3x 2 y 2  2 .mod 3/.
(b) If there is a solution to the congruence 3x 2 y 2  2 .mod 3/,
explain why there then must be an integer y such that y 2  2 .mod 3/.
(c) Use a proof by contradiction to prove that the Diophantine equation
3x 2 y 2 D 2 has no solution.

10. Use the method suggested in Exercise (9) to prove that the Diophantine equa-
tion 7x 2 C 2 D y 3 has no solution.

Explorations and Activities

11. Linear Congruences in One Variable. Let n be a natural number and let
a; b 2 Z with a ¤ 0. A congruence of the form ax  b .mod n/ is called a
linear congruence in one variable. This is called a linear congruence since
the variable x occurs to the first power.
A solution of a linear congruence in one variable is defined similarly to
the solution of an equation. A solution is an integer that makes the result-
ing congruence true when the integer is substituted for the variable x. For
example,

 The integer x D 3 is a solution for the congruence 2x  1 .mod 5/

since 2
3  1 .mod 5/ is a true congruence.
 The integer x D 7 is not a solution for the congruence 3x  1 .mod 6/
since 3
7  1 .mod 6/ is a not a true congruence.

(a) Verify that x D 2 and x D 5 are the only solutions the linear congru-
ence 4x  2 .mod 6/ with 0  x < 6.
448 Chapter 8. Topics in Number Theory

(b) Show that the linear congruence 4x  3 .mod 6/ has no solutions with
0  x < 6.
(c) Determine all solutions of the linear congruence 3x  7 .mod 8/ with
0  x < 8.

The following parts of this activity show that we can use the results of Theo-
rem 8.22 to help find all solutions of the linear congruence 6x  4 .mod 8/.

(d) Verify that x D 2 and x D 6 are the only solutions for the linear
congruence 6x  4 .mod 8/ with 0  x < 8.
(e) Use the definition of “congruence” to rewrite the congruence
6x  4 .mod 8/ in terms of “divides.”
(f) Use the definition of “divides” to rewrite the result in part (11e) in the
form of an equation. (An existential quantifier must be used.)
(g) Use the results of parts (11d) and (11f) to write an equation that will
generate all the solutions of the linear congruence 6x  4 .mod 8/.
Hint: Use Theorem 8.22. This can be used to generate solutions for x
and the variable introduced in part (11f). In this case, we are interested
only in the solutions for x.

Now let n be a natural number and let a; c 2 Z with a ¤ 0. A general linear

congruence of the form ax  c .mod n/ can be handled in the same way
that we handled in 6x  4 .mod 8/.

(h) Use the definition of “congruence” to rewrite ax  c .mod n/ in terms

of “divides.”
(i) Use the definition of “divides” to rewrite the result in part (11h) in the
form of an equation. (An existential quantifier must be used.)
(j) Let d D gcd.a; n/. State and prove a theorem about the solutions of
the linear congruence ax  c .mod n/ in the case where d does not
divide c.
Hint: Use Theorem 8.22.
(k) Let d D gcd.a; n/. State and prove a theorem about the solutions of
the linear congruence ax  c .mod n/ in the case where d divides c.
8.4. Chapter 8 Summary 449

8.4 Chapter 8 Summary

Important Definitions

 Greatest common divisor of two integers, page 414

 Linear combination of two integers, page 423

 Prime number, page 426

 Composite number, page 426

 Prime factorization, page 427

 Relatively prime integers, page 428

 Diophantine equation, page 441

 Linear Diophantine equation in two variables, page 441

Important Theorems and Results about Relations, Equivalence Rela-

tions, and Equivalence Classes

 Theorem 8.3. Let a and b be integers with a ¤ 0 and b > 0. Then gcd.a; b/
is the only natural number d such that

(a) d divides a,
(b) d divides b, and
(c) if k is an integer that divides both a and b, then k divides d.

 Theorem 8.8. Let a and b be integers, not both 0. Then gcd.a; b/ can be
written as a linear combination of a and b. That is, there exist integers u
and v such that gcd.a; b/ D au C bv.

 Theorem 8.9.

1. The greatest common divisor, d , is a linear combination of a and b.

That is, there exist integers m and n such that d D am C bn.
2. The greatest common divisor, d , divides every linear combination of a
and b. That is, for all x; y 2 Z , d j .ax C by/.
3. The greatest common divisor, d , is the smallest positive number that is
a linear combination of a and b.
450 Chapter 8. Topics in Number Theory

 Theorem 8.11. Let a and b be nonzero integers, and let p be a prime num-
ber.

1. If a and b are relatively prime, then there exist integers m and n such
that am C bn D 1. That is, 1 can be written as linear combination of
a and b.
2. If p j a, then gcd.a; p/ D p.
3. If p does not divide a, then gcd.a; p/ D 1.

 Theorem 8.12 Let a, b, and c be integers. If a and b are relatively prime

and a j .bc/, then a j c.

 Corollary 8.14

1. Let a; b 2 Z, and let p be a prime number. If p j .ab/, then p j a or

p j b.
2. Let p be a prime number, let n 2 N, and let a1 ; a2 ; : : : ; an 2 Z. If
p j .a1 a2


an /, then there exists a natural number k with 1  k  n
such that p j ak .

 Theorem 8.15, The Fundamental Theorem of Arithmetic

1. Each natural number greater than 1 is either a prime number or is a

product of prime numbers.
2. Let n 2 N with n > 1. Assume that

n D p1 p2


pr and that n D q1 q2


qs ;

where p1 ; p2 ;


pr and q1 ; q2 ;


qs are primes with p1  p2 

 pr and q1  q2 


 qs . Then r D s, and for each j
from 1 to r, pj D qj .

 Theorem 8.16. There are infinitely many prime numbers.

 Theorem 8.22. Let a, b, and c be integers with a ¤ 0 and b ¤ 0, and let

d D gcd.a; b/.

1. If d does not divide c, then the linear Diophantine equation ax Cby D

c has no solution.
8.4. Chapter 8 Summary 451

2. If d divides c, then the linear Diophantine equation ax C by D c

has infinitely many solutions. In addition, if .x0 ; y0 / is a particular
solution of this equation, then all the solutions of the equation are given
by
b a
x D x0 C k and y D y0 k;
d d
where k 2 Z.

 Corollary 8.23. Let a, b, and c be integers with a ¤ 0 and b ¤ 0. If a and

b are relatively prime, then the linear Diophantine equation ax C by D c
has infinitely many solutions. In addition, if x0 , y0 is a particular solution
of this equation, then all the solutions of the equation are given by

x D x0 C bk and y D y0 ak;

where k 2 Z.