Examples 1.1.

1. Use induction to prove, for all n  1, 2, 3, , that

n(n  1)(2n  1)
12  22  32   n2 
6
Solution
We observe that
1(1  1)(2  1)
12  1
6
This means that for n  1 , the statement is true.

Now by the second induction hypothesis we have

n(n  1)(2n  1)
12  22  32   n2  ( n  1) 2   ( n  1) 2
6
n(2n  1)  6( n  1)
 ( n  1)[ ]
6
2n 2  7 n  6)
 ( n  1)[ ]
6
(n  1)(n  2)(2n  3)
 .
6
These show that the statement is true for all positive integers.

2. Establish the Bernoulli inequality:

If (1  a)  0, then (1  a)n  1  na for all n  1.

Solution
For n  1 , we have (1  a)1  1  1 a , and so the statement is true.
Now by the second induction hypothesis we have:
(1  a)n1  (1  a)n (1  a)
 (1  na )(1  a )  1  (n  1)a  na 2  1  (n  1)a.
Indeed since na 2  0 .

Exercises 1.1.4

1. For all n  1, 2, 3, prove that

1 n
1  5  52   5n1  (5  1) .
4

2. For all n  1, 2, 3, prove that

1 3  5   (2n 1)  n2 .

3. Verify that for all n  1,

(2n)!
2  6 10 14   (4n  2)  .
n!
1.2 The Binomial Theorem

Closely connected with factorial notation are the binomial

coefficient
n
 .
k 
For any positive integer n and any integers k satisfying
0  k  n, these are defined by
n n!
  .
 k  k !(n  k )!
n
By canceling out either k ! or (n  k )!,   can be written as
k 
 n  n(n  1) (k  1) n(n 1) (n  k  1)
  
k  (n  k )! k!

For example
 9  9! 9  8  7  6  5 9  8  7  6
     126
 4  4!5! 5! 4!
Also observe that
n n
      1 , since we have 0!  1
0 n

There are many useful identities connecting binomial

coefficients.

Theorem 1.2.1 ( Pascal’s Rule )

 n   n   n  1
    1 k  n
 k   k  1  k  1
Proof

n  n  n! n!
   
 k   k  1 k !(n  k )! (k  1)!(n  k  1)!
n! 1 1
= [  ]
(k  1)!(n  k )! k n  k  1
n! n 1
= [ ]
(k  1)!(n  k )! k (n  k  1)
(n  1)!  n  1
= = .
k !(n  k  1)!  k  1

1.3 The Binomial Theorem

The binomial theorem is a formula for the complex expansion of

(a  b)n , n  1, into a sum of powers of a and b .

By a direct multiplication, it is easy to verify that:

(a  b)1  a  b
(a  b) 2  a 2  2ab  b 2
(a  b)3  a 3  3a 2b  3ab 2  b3
(a  b) 4  a 3  4a 3b  6a 2b 2  4ab3  b 4 , ect.

These leads us to the general form:

n n n  n  n 1  n  n

(a  b)n    a n    a n 1b    a n 2b 2    ab    b .
0 1  2  n  1 n
Theorem 1.3.1 ( The Binomial Theorem )

For n  1, we have
n
n
( a  b) n     a n  k b k .
k 0  k 

Proof. (by mathematical induction)

When n  1 , the formula reduces to

1
1 1  1
(a  b)1     a1k b k    a1b0    a0b1  a  b, , and so
k 0  k   0  1
the statement is true.

Now by the second induction hypothesis we have:

(a  b) n 1  a (a  b) n  b (a  b) n
n
 n  n  k 1 k n
 n  n  j j 1
=   
k 0  k 
a b    a b
j 0  j 

n
 n  n  k 1 k n
 n  n k 1 k
= a    a
n 1
b   a b  b n 1
k 1  k  k 1  k  1 

n
n  n 
 a n 1   [      ] a
( n 1)  k k
b  b n 1
k 1   
k k  1 
By Pascal’s rule, we have:

n
 n  1 ( n 1)  k k
(a  b) n 1  a n 1    a b  b n 1
k 1  k 
n
 n  1 ( n 1) k k
=   a b .
k 0  k 

So the theorem is true for all n  1.

Solved problem 1.3.2

Derive Newton’s identity:

 n  k   n  n  r 
       ; nk r 0.
k   r   r  k r
Solution
 n  k  n! k!
    ( )( )
 k  r  k !(n  k )! r !(k  r )!
n!

r !(n  k )!(k  r )!
n! (n  r )!  n  n  r 
( )( )    .
r !(n  r )! (n  k )!(k  r )!  r  k  r 
 9. 0 | n, then n  0 . (zero divides only zero).

10. If d , n are positive and d | n then d  n .

Proof.
We shall prove for example (2)

Since d | n and n | m , then there are x, y such that

n  dx, m  ny .
So
m  ny  (dx) y  d ( xy) .
Hence
d |m

To prove (3)

Since d | n and d | m , then there are x, y such that

n  dx, m  dy .
Now
an  bm  a(dx)  b(dy)  d (ax  by) .
So
d | an  bm .

To prove (10)

Since d , n are positive and d | n , then there is a positive integer

x such that n  dx .
Whence
n  dx  d as asserted.

Solved problems 2.1.2 .

1. Use the definition of divisibility to prove that

If a | b and b | c then a 2 | (b2  3bc) .

Solution:
Suppose
a | b and b|c ,
so that
b  ak and c  bq  a(kq) for some integers k , q .
Hence
b2  3bc  (ak )2  3(ak )(akq)  a 2 (k 2  3k 2q) , where (k 2  3k 2 q) is integer
and so
a 2 | (b2  3bc) .

2. Let a, b, c be integers. Show that: If a / b then a / bc

Solution:
Suppose a | b , so that b  ak for some k .
By multiply it by c , we get
bc  (ak )c  a(kc) .
Whence
a / bc .
3. Prove that a | b if and only if ac | bc , where c  0 .
Solution:

Assume a | b , so that b  as for some integer s .

Multiply this equation by a positive integer c ,we have
bc  asc ,
whence
bc  (ac)s .
Thus
ac | bc .

Conversely, assume ac | bc with c  0 , we deduce that

bc  (ac)s for some s .
Thus
bc  asc , since c  0 we have b  as .
Hence
a |b.

4. Disprove the following: If a | b  c , then a | b or a | c .

This is not true. For instance put
a  3, b  2, c  4 .
We have a | b  c while neither a | b nor a | c .

5. Use induction to prove, for all positive integer n , that

22 n  5 is divisible by 3 .

Solution
For n  1 , we have 22  5  9 , and so the statement is true
Assume the statement is true for n  k i.e. 3 | 22k  5 hence for
some integer q , we have 22k  5  3q .
Now
22( k 1)  5  22 k 22  5
 22 k (1  3)  5
 (22 k  5)  3(22 k )
 3q  3(22 k )
 3(q  22 k ).
Hence 3 | 22( k 1)  5 as needed.
Thus the statement is true for any n .

6. Show that any integer in the form 6 x  5 can be written in

the form 3k  1 and the convers is not true.
Indeed
6x  5  6 x  6 1  3k 1 , where k  (2 x  1) .
But the converse is not true, for instance 2  3 1 1 while the
equation 2  6 x  5 has no integer solution.

Exercises 2.1.3.

1. Prove each of the properties 1 through 10 in Theorem 2.2.

2.2. The Division Algorithm

Theorem 2.2.1

If a, b are integers and b  0 then there exist unique integers

q, r satisfying the two conditions a  bq  r and 0  r  b . q is
called the quotient and r is the remainder. In particular if r  0
we have a  bq whence b | a otherwise if r  0, then b | a .

Examples 2.2.2

2 | 6 , since 6  2  3  0 with q  3, r  0 .

3 | 18 , since 18  3  (6)  0 with q  6, r  0 .

3 | 16 , since 16  3  5  1 with q  5, r  1 .

For 117, 7 , we have 117  7 q  r where q  16, r  5 .

For 117,7 , we have 117  7 q  r where q  17, r  2 .

For 3, 7 , we have 3  7 q  r where q  0, r  3 .

For 3, 7 , we have 3  7 q  r where q  1, r  4 .

Examples 2.2.3

1. For any integer n is either even or odd.

Indeed
n  2q  r; r  0,1.
Whence
if r  0 then n  2q i.e. n is even,
and
if r  1 then n  2q  1 i.e. n is odd.

2. For any integer n , the number n3  n always has 3 as a

factor.
( consider the three cases: n  3q; n  3q  1; n  3q  2 ).

3. The product of three consecutive integers has 6 as a factor.

( consider the six cases:
n  6q; n  6q  1; n  6q  2; n  6q  3; n  6q  4; n  6q  5 ).
Definition 2.3.3

Two integers a  0, b  0 are said to be coprime (or relatively

prime) whenever gcd(a, b)  1.

Examples 2.3.4

gcd(3,7)  1; gcd(3,10)  1; gcd(9,14) 1 .

Theorem 2.3.5. ( Greatest Common Divisor properties).

1. Given nonzero integers a and b, gcd(a, b) always exist and

unique.

2. Euclid’s Algorithm. gcd(a, b)  d can be written as a linear

combination of a and b . i.e. there exist integers x and y such
that d  ax  by . See (3) in Theorem 2.1.1.

3. Given nonzero integers a and b ,

we have
gcd(a, b)  1 if and only if ar  bs  1, for some r , s  .
Proof:

We shall prove (3) If a, b are relatively prime so that gcd(a, b)  1,

then Euclid’s Algorithm guarantees the existence of integers
x, y satisfying
1  ax  by .
As for the converse suppose that 1  ax  by for some x, y  , and
that d  gcd(a, b) . Since d | a and d | b so that d | (ax  by), or d |1 ,
whence d  1 or gcd(a, b)  1 .

Theorem 2.3.6.

For any positive integer m prove that gcd(ma, mb)  m gcd(a, b) .

Proof.
Let gcd(ma, mb)  k , gcd(a, b)  d , and we need to prove that
k  md .

Since gcd(a, b)  d , then d | a  d | b implies md | ma  md | mb

but gcd(ma, mb)  k , then md  k (1)

Also gcd(a, b)  d implies d  sa  tb for some integers s, t .

Multiply both sides by m , we obtain md  s(ma)  t (mb) as a
linear combination of ma, mb . But the greatest common divisors
is the least linear combination, so k  md (2) .
From (1), (2) we get the required equality.

For example, we have

gcd(12,30)  3gcd(4,10)  3  2gcd(2,5)  6 1  6 .

Corollary 2.3.7

If d is a common divisor of a, b , then d  gcd(a, b) if and only if

gcd(a / d , b / d )  1.

The greatest common divisor can be found by a more efficient

process, involving repetition application of the Division
Algorithm. This method is referred to as Euclidean Algorithm.
Let us see how the Euclidean Algorithm works in an example.

Example 2.3.8

Find gcd(12378, 3054) and use Euclidean Algorithm to obtain

x, y satisfying
gcd(12378, 3054)  12378 x  3054 y .
The appropriate application of the Division Algorithm produce
the equations
12378  4  3054  162
3054  18  162 + 138
162 = 1  138 + 24
138 = 5  24 + 18
24 = 1  18 + 6
18 = 3  6 + 0

The process tells us that the last remainder appearing in these

equations, namely, the integer 6, is the greatest common divisor
of 12378 and 3054 .

To represent 6 as a linear combination of the integers

12378 and 3054 , we start with the next-to-last of the displayed
equations and successively eliminate the remainders
18, 24, 138 and 162;

6  24  18
 24  (138  5  24)
 6  24  138
 6(162  138)  138
 6 162  7  138
 6 162  7  (3054  18 162)
 132 162  7  3054
 132(12378  4  3054)  7  3054
 132 12378  ( 535) 3054

Thus, we have
6  gcd(12378, 3054)  12378 x  3054 y

Where x  132 and y  535 .

Lemma 2.3.9 (Euclid’s Lemma).

If a | bc , with gcd(a, b)  1, then a | c .

Proof
The relation gcd(a, b)  1 allows us to write 1  ar  bs for some
choice integers r , s .
Multiplying the last equation by c , produces
c  1 c  (ar  bs)c  acr  bcs .
Since a | ac and a | bc it follows that a | (acr  bcs)  c .

That is
a | c as asserted.
Solved Problem 2.3.10.

1. Show that gcd(a, a  1)  1. Indeed , a  (1)  (a  1) 1  1 and

Theorem 2.3.5. (3) leads to the required.

2. Show that gcd(a, a  2)  1 or 2 .

Proof.

In the start, we have

a  (1)  (a  2) 1  2 .

Thus if gcd(a, a  2)  d , therefore d | 4 and d | (a  2) .

Whence
d | (a  (1)  (a  2) 1)  2 .
Now
d | 2 leads to d  1 or d  2 .

3. If gcd(4, a)  gcd(4, b)  2 , then show that gcd(4, a  b)  4 .

Proof:

Since gcd(4, a)  gcd(4, b)  2 then 2 | a and 2 | b . Thus

a  2k and b  2q for some odd integers k , q ; otherwise if k or q
is even then gcd(a, 4)  4 or gcd(b, 4)  4 which is not correct.
Therefore k +q is even, so k +q  2 x for some x . Now
a  b  2(k +q)  2(2 x)  4 x and so gcd(4, a  b)  4 as required.
4. Recall the following statement: For any positive integer m
we have gcd(ma, mb)  m gcd(a, b) , then find gcd(70500,64500) .
Solution

gcd(70500,64500)  1500gcd(47, 43) 1500 1 1500 .

5. Show that if gcd(a, b)  gcd(b, c)  1 , then gcd(a, bc)  1 .

Solution:
Since gcd(a, b)  gcd(b, c)  1 , then we have
ax  by  1 and as  ct  1 for some integers x, y, r, s .
By multiplying the two equations, we get
a 2 xs  axct  byas  bcyt  1 .
Thus
a(axs  xct  bys)  bc( yt )  1 .
So that
gcd(a, bc)  1 .

6. If a | c and b | c, with gcd(a, b)  1, then ab | c .

Proof

If a | c and b | c then c  ar  bs for some integers r , s .

Now the assumption gcd(a, b)  1 allow us to write 1  ax  by for
some x, y  . Multiply the last equation by c , it appears that
c  1 c  (ax  by)  c  acx  bcy .
So that
c  a(bs) x  b(ar ) y  ab(sx  ry)
or as a divisibility statement, ab | c .
The following concept is a parallel to that of the greatest
common divisor of two nonzero integers a and b, whenever
a | c and b | c .

2.4. The Least Common Multiply

Definition 2.4.1

The least common multiply of two nonzero integers a and b,

denoted by lcm[a, b] , is the positive integer m satisfying the
following:

(1) a | m and b | m and (2) if a | c and b | c then m  c .

As an example, the positive common multiply of the integers

12 and 30 are 60, 120, 180, ; hence lcm[12,30]  60 .

The following remark is clear from our discussion.

Given nonzero integers a and b, lcm[a, b]  | ab | .

The relationship between the ideas of greatest common divisor

and the least common multiply is given by:

Theorem 2.4.2

For positive integers a and b we have gcd(a, b) lcm[a, b]  ab .

For example, of this statement. gcd(3054, 12378)  6,
whence , lcm[3054, 12378]  3054 12378 / 6  6300402 .
Corollary 2.4.3

For positive integers a and b , lcm[a, b]  ab if and only if

gcd(a, b)  1.

Solved problems 2.4.4

1. gcd(99,10!)  9  gcd(11,(10)(8)!)  9 1  9 .
While
(99)(10!) (99)(10!)
lcm(99,10!)    (11)10!  11! .
gcd(99,10!) 9

2. There are infinitely many pairs of integers a, b satisfy

lcm[a, b]  gcd(a, b) . Indeed, for any positive integer n we
have
lcm[n, n]  n  gcd(n, n) .

3. If k  0 , then lcm(ka, kb)  k lcm(a, b) .

Indeed: If k  0 , then
ka  kb k 2 ab
lcm(ka, kb)    k lcm(a, b) .
gcd(ka, kb) k gcd(a, b)

4. Evaluate gcd(a, a  1); lcm[a, a  1] where a is a positive integer.

Solution:
Since a(1)  (a  1)(1)  1 , so that gcd(a, a  1)  1.

a(a  1) a(a  1)
Now lcm[a, a  1]    a(a  1) .
gcd(a, a  1) 1
 Indeed,
ax  by  a[ x0  (b / d )t ]  b[ y0  (a / d )t ]
 (ax0  by0 )  (ab / d  ab / d )t  c  0t  c

Thus there are infinite numbers of solutions. One for every value
of t .

Corollary 2.5.2.

In particular, if gcd(a, b)  1 and if x0 , yo is a solution of the

linear Diophantine equation ax  by  c , then all solutions are
given by x  x0  bt , y  y0  at for integral value of t .

Solved Problems 2.5.3.

1. The linear Diophantine equation 6 x  14 y  15 has no

solution, since gcd(6,14)  2 and 2 | 15 .

2. Since gcd(7,13)  1 and 1| 666 , then the linear Diophantine

equation 7 x  13 y  666 has a solution, and it follows that
7(r )  13(s)  1 with r  2, s  1. Multiply the relation
7(2)  13(1)  1 by 666 , we arrive at 666  7(1332)  13(6661) so
that x  1332, y  666 provide one solution to the
Diophantine equation 7 x  13 y  666 . All other solutions are
expressed by x  1332  13t, y  666  7t .
3. The general solutions of the linear Diophantine equation
172 x  20 y  1000 can be given by x  500  5t , y  4250  43t .
Find the solutions in the positive integers. For this, t must be
chosen to satisfy the two inequalities
36
5t  500  0,  34t  4250  0 or, equivalently 98  t  100 ,
43
since t  99 we have only one positive solution, that is
x  5, y  7 corresponding to the value t  99 .

4. Find the general solution of the Diophantine equation:

341x  217 y  31.
Solution:
Since gcd(7,13)  1 , so 7(s)  13(t )  1 has solution at s  2, t  1 .
Multiplying the relation
7(2)  13(1)  1 by 666 ,
we arrive at 666  7(1332)  13(666) so that x  1332, y  666
provide one solution to the Diophantine equation
7 x  13 y  666.
All other solutions are expressed by
x  1332  13t , y  666  7t .
5. For any prime number p and nonzero integer a we have
gcd(a, p)  p or gcd(a, p)  1 according as p divides a or not.

6. If n  1 is composite then n has a prime divisor p  n .

Examples 3.1.3

1. 6, 8, 9, 10, 12, 14, 15, 16 are 8 consecutive composite numbers.

2. There is no three consecutive prime numbers.

3. The list of primes in order goes 2,3,5,7,11,13,17,19,

4. To obtain all the primes between 1 and 100 . Let’s look at a

number table from 1 to 100 and see how they were
discovered. In the start take out the special number 1 .
Then take out numbers that have a composite factors
2, 3, 5, 7 .(see 3.1.2 no 6.). Here is how you do it.

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
Solved problems 3.1.6

1. 1  2  3  5  7  211 is prime; 1  2  3  5  7 11  2311 is prime,

while 1  2  3  5  7 1113  59  509 , where 59 and 509 are primes.

2. 221 is composite while 211 is prime.

The smallest prime divides 221 is 13 . Thus 221  13 17 i.e. it is
composite.

Consider the integer 211 . Since 13  211  17 , we need only try

out the primes which are not larger than 13 as possible divisors,
namely the primes 2,3,5,7,11,13 . But it is easy to recognize that
211 is not divisible by 2,3,5 .

Also,
7 | 211, for, 211  7(30) 1
11 | 211, for, 211  11(19)  2 .
13 | 211, for, 211  13(16)  3

Thus 211 is prime.

3. 1387 is composite while 1013 is prime.

The smallest prime divides 1387 is 19 . Thus 1387  19  73 i.e. it

is composite. Consider the integer a  1013 . Since
31  1013  32 , we need only try out the primes which are not
larger than 31 as possible divisors, namely the primes
 2,3,5,7,11,13,17,19, 23, 29,31. But it is easy to recognize that
a  1013 is not divisible by 2,3,5 .
Also,
7 | 1013, for, 1013  144(7)  5;
11 | 1013, for, 1013  92(11) 1;
13 | 1013, for, 1013  77(13) 12;
17 | 1013, for, 1013  59(17) 10;
19 | 1013, for, 1013  53(19)  6
23 | 1013, for, 1013  44(23) 1;
29 | 1013, for, 1013  34(29)  27;
31 | 1013, for, 1013  32(31)  21.

Thus 1013 is prime.

4. In the following table we obtain all the primes between 100

and 200.

101 103 107 109

113
127
131 137 139
149
151 157
163 167
173 179
181
191 193 197 199
3.2. Unique Factorization.

Definition 3.2.1

Let n  2 , if n  p1 p2 ps where p1 , p2 , , ps are primes, we call

this the prime factorization of n .

Theorem 3.2.2 ( The Fundamental Theorem of Athematic )

Every integer n  2 can be written uniquely in the form

n  p1 p2 ps , where p1 , p2 , , ps are primes satisfying
p1  p2   ps .

In other words sometimes we say. Every integer n  2 can be

expressed as a product of primes; this representation is unique
except of the order of the prime occur. Note for example that
600  2  2  2  3  5  5  2  3  2  5  2  5  3  5  2  2  2  5 etc. Perhaps, the
nicest way to write the prime factorization of
600 is 600  23  3  52 .

Corollary 3.2.3

Any integer n  2 can be written uniquely in a canonical form

n  p1k p2k
1
prk , where for i  1, 2, , r each ki is a positive integer
2 r

and each pi is prime, with p1  p2   pr .

As further examples we site

4725  33  52  7 and 17640  23  32  5  72 .
Theorem 3.2.4

Let m and n be positive integers and p1 , p2 , , pr be the distinct

primes which divide at least one of m or n . Then m and n may be
written in the form

m  p1k1 p2k2 prkr with ki  0 for i  1, 2, ,r

.
n  p1j1 p2j2 prjr with ji  0 for i  1, 2, ,r

Whence, gcd(m, n)  p1u p2u 1 2

prur where ui  min{ki , ji } , the smaller
of ki and ji ,

lcm[m, n]  p1v1 p2v2 prvr where vi  max{ki , ji } , the largest

of ki and ji .

For example, if a  108 and b  225 , then

a  22 33 50 and b  20 32 52 .
Thus

gcd(a, b)  20 32 5o  9; lcm[a, b]= 22 33 52  2700 .

Solved problems. 3.2.5

1. The prime factorization of the number n  13832000 is given

by
n  26.53.7.13.19 .
2. If a  4725 and b  17640 , then a  20 33 52 7 and b  23 32 51 72
Thus by theorem 3.2.4
gcd(4725, 17640)  20  32  5  7  315;
.
lcm[4725, 17640]= 23  33  52  7 2  264600.

3. 4725  33 52 7 and 17460  22 32 5 (97) ,

thus by theorem 2.3.6
gcd(17460, 4725)  gcd(22 32 5 (97), 33 52 7)
 32  5 gcd(22 (97), 3  5  7 ) .
 (45)(1)  45.
Also by problem (3) in 2.4.4
lcm[17460, 4725]  lcm[22 32 5 (97), 33 52 7]
= 32  5 lcm(22 (97), 3  5  7 )
 45  22  3  5  7  (97)
 1833300