MAT 221 LECTURE NOTE

REAL ANALYSIS I

April 3, 2021
Contents

1 Concept of Sets and Functions 3

1.1 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Construction of Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 More on Construction of Sets . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.6 The Real Number System . . . . . . . . . . . . . . . . . . . . . . . . 19
1.7 Absolute Values of Numbers . . . . . . . . . . . . . . . . . . . . . . . 23
1.8 Infimum and Supremum . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.9 Completeness of Axiom: . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.10 Some Properties of the Supremum: . . . . . . . . . . . . . . . . . . . 26
1.11 Problem: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

2 Dedekind Cut 31
2.1 Constructing the Real Numbers . . . . . . . . . . . . . . . . . . . . . 36

3 Sequences 39
3.1 Modulus of Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.2 Uniqueness of limits . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

1
 3.3 Arithmetic Operations on Convergent Sequences . . . . . . . . . . . . 44
3.4 Cauchy Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

4 Series 52

2
Chapter 1

Concept of Sets and Functions

1.0.1 Definition. 1. A set is a collection of distinct objects e.g

(i.) Nigerians

(ii.) The Planets

(iii). The odd numbers etc.

2. The objects making up a set are called members or elements of the set. e.g

(i.) You are a Nigerian.

(ii.) Saturn is a member of the Planets.

(iii). 3 is a member of the odd numbers.

1.1 Notation
Let S represent a set. Then the notation a ∈ S is read: a belongs to S and it means
a is a member (element) of the set S

3
1.1.1 Definition. Two sets A and B are equal(denoted A = B) if and only if they
have the same members. A 6= B denote ”A is not equal to B”.

1.2 Notation
(1). It is usual to denote a set by listing its members within braces. Thus {1, 2, 3}
is the set whose only members are 1, 2, 3. {a, e} and {0} are other examples.
The above way of designating sets is called roster notation and the list of
members appearing within the braces is referred to as the roster of the sets.

(2). Along with (1) above, a set can also be denoted thus: {x ∈ U : f (x)} where
f (x) is a sentence defining the members of the set. It is read ” the set of all x
in U such that f (x)”

1.2.1 Example. (a). {x ∈ Q : 2x = 1} is the set of all rational numbers such that
2x = 1.

(b). {x ∈ N : 1 6 x 6 3}

1.2.2 Definition. (i). A set that has no element is called the empty set denoted
∅ or { }.

(ii). The set of all things under discussion in a particular context is called the
universal set denoted by E or U. We can represent N as N := {1, 2, 3, ..., n, ...}
and 5 < x < 50 := {6, 7, 8, ..., 49}
In the same vein, we define the family {A1 , A2 , ..., An } of the set A0i s with its
clear meaning. Another variant is the following: For each n ∈ N, let xn be an object.
The collection of all such objects is displayed as {xn : n ∈ N}, N here is called an
indexing set. More generally, let A be a set and for each α ∈ A, let xα be an object
which may be a set. The collection of all such objects is denoted by {xα : α ∈ A}.

4
1.3 Construction of Sets
This involves ways of obtaining new sets from given ones, thus generating more sets.

1.3.1 Definition. Let A be a set. Then a set B is called a subset of A if and only
if every member of B is a member of A. This is denoted B ⊂ A read B is included
in A or A includes (contains) B.
Symbolically,
(B ⊂ A) := (x ∈ B ⇒ x ∈ A). Thus going back to the equality of sets , we can see
that A and B are equal ⇐⇒ A ⊂ B and B ⊂ A. (Define strict inclusion).
N.B : The empty set and the entire set are subsets of the given sets.

1.3.2 Definition. (Intersection) The intersection of two sets A and B is the set
of all members that are simultaneously in A and B. This is denoted by A ∩ B read
”A intersection B”. A ∩ B := {x : (x ∈ A) ∧ (x ∈ B)}.

1.3.3 Definition. (Disjoint)Two sets A and B are said to be disjoint if they do

not have any members in common.

1.3.4 Definition. (Union) The union of two sets A and B is the set of all the
elements that belong to either or both of them. This is denoted by A ∪ B i.e
A ∪ B := {x : (x ∈ A) ∨ (x ∈ B)}.

1.3.5 Definition. Set difference The set difference between A and B is the set
whose members are in A but not in B. This is denoted by A − B

1.3.6 Definition. (Complement) Let B ⊂ A as above , then A − B is called the

complement of B relative to A. While U − B is called the complement of B. This
0
complement is a special type of set difference. This is denoted by ∼ A or A or Ac .

1.3.7 Example. For A, B ∈ S , A ⊂ B ⇔ S − A ⊃ S − B. (Show)???

5
1.4 More on Construction of Sets

CARTESIAN PRODUCT OF SETS:

1.4.1 Definition. An ordered pair is a two element set one member of which is
designated the first entry, the other is designated the second entry.

1.4.2 Definition. If A and B are sets, the Cartesian product of A and B is the set
of all ordered pairs with first entry of a member of A and second entry a member
of B. This is denoted A × B and reads ”the Cartesian product of A and B”. i.e
A × B := {(x, y) : x ∈ A, y ∈ B}. The two ordered pairs (x, y) and (a, b) are equal
if and only if a = x and b = y.

1.4.3 Example. Suppose B = {w, x, y, z} and A = {i, o, u}. Then A × B =

{(i, w), (i, x), (i, y), (i, z), (o, w), (o, w), (o, x), (o, y), (o, z), (u, w), (u, x), (u, y), (u, z)} or
(Draw a table).

1.4.4 Remark. The ordered pair (a, b) differ from the pairs {a, b} in the two fol-
lowing ways.

(1). A pair {a, b} is the set with two distinct elements, hence {a, b} is only used
when a 6= b. However (a, a) ∈ A × B when a is in both A and B.

(2). If a 6= b, (a, b) 6= (b, a) but {a, b} = {b, a}. Notice also that S × T 6= T × S
unless S = T , S = ∅ or T = ∅

Problem:

(1) Let A be a set. Why is A × ∅ = ∅?.

(2) Consider the set E, E1 , E2 arbitrary. show that

6
 E × (E1 ∩ E2 ) = (E × E1 ) ∩ (E × E2 )
E × (E1 ∪ E2 ) = (E × E1 ) ∪ (E × E2 )

(3) Let S = {1, 2} and T = {1, 3, 4}.

(i) List the ordered pairs of S × T .

(ii) List the ordered pairs of T × S

(iii) Is S × T = T × S??? (Justify your answer).

(iv) Determine the conditions on S and T that makes the equality in (iii) to
hold.

1.4.5 Example. (1) Read up power sets, Family of sets and Cardinality of sets.

(2) Show that the Cardinality of a set with n elements is 2n .

(3) State and prove De Morgan’s duality theorems. If A, B, C are sets, then

(a) A − (B ∪ C) = (A − B) ∩ (A − C)

(b) A − (B ∩ C) = (A − B) ∪ (A − C)

(4) Prove De Morgan’s laws i.e A = U.

1.5 Relations
1.5.1 Definition. Let X and Y be any sets. A relation R from X to Y is a subset
of X × Y . A relation from X into X is simple and called a relation in X. xRy is
read x is related to y and it means (x, y) ∈ R.

1.5.2 Example. Let X := {2, 3, 4, 7} and Y := {2, 3, 5}. Then X × Y has 12

members. The power sets P(X × Y ) for X × Y has 212 members. Thus there are
212 relations from X into Y including the empty relation. From the above, define

7
 • R1 := {(x, y) ∈ X × Y : x = y} = {(2, 2), (3, 3)}

• R2 := {(x, y) ∈ X × Y : x < y} = {(2, 3), (2, 5), (3, 5), (4, 5)}

• R3 := {(x, y) ∈ X × Y : x2 = y} = ∅

• R4 := {(x, y) ∈ X × Y : x is divisible by y} = {(2, 2), (3, 3)}

• R5 := {(x, y) ∈ X × Y : x > y} =??

• R6 := {(x, y) ∈ X × Y : x|y} =??

Problem: Use each of the following to define a relation: =, >, 6, ⊆, =⇒, ∼, ∼

=, ⊥, |
or k

1.5.3 Definition. The domain of the relation R is the set {x : (x, y) ∈ R} and it
is denoted dom(R). The range of the relation is the set {y : (x, y) ∈ R} and it is
denoted by Rge(R).

1.5.4 Definition. Let R be a relation

(1.) Then y is called the image of x under the R or a value of R at x iff (x, y) ∈ R.
Moreover, R(x) := {(x, y) ∈ R} is called the image of x under R. Where
R(x) is the singleton, {y} say, write R(x) = y.

(2.) More generally, if S ⊂ dom(R), then R(S) := ∪s∈S R(s) is called the image of
S under R.

1.5.5 Definition. Given a relation R , the new relation R−1 defined by

R−1 = {(y, x) : (x, y) ∈ R}

is called the converse of R. Thus (a, b) ∈ R iff (b, a) ∈ R.

8
1.5.6 Definition. Let R be a relation from X into Y . Let R0 be another relation.

0
(1). The complementary relation to R is denoted R and it is defined by

0
R := {(x, y) ∈ X × Y : (x, y) ∈
/ R}

(2). The relation R0 is said to be a subrelation of the relation R iff R0 ⊂ R

(3). R0 and R are said to be disjoint iff R0 ∩ R = ∅.

(4). X × Y is called the universal relation from X into Y .

(5). R0 is called the empty relation from X into Y iff R0 ∩ (X × Y ) = ∅.

(6). R0 and R are said to be equal iff they are equal as sets.

1.5.7 Definition. Let N and M be two relations. Then N · M , the composition of

N and M is defined by

N · M := {(x, z) : ∃y 3 (x, y) ∈ M and (y, z) ∈ N }

Let R be a relation on a set S. Then

(i). R is reflexive if aRa ∀ a ∈ S i.e (x, x) ∈ R.

(ii). R is symmetric if (x, y) ∈ R ⇒ (y, x) ∈ R.

(iii). R is transitive if (x, y) ∈ R and (y, z) ∈ R ⇒ (x, y) ∈ R.

(iv.) R is antisymmetric if (x, y) ∈ R and (y, x) ∈ R ⇒ x = y.

A relation that satisfies (i) − (iii) is called an equivalence relation.

9
Exercise:

(1.) Let | be the integer divisor relation on the set Z of integers. Is | reflexive,
symmetric, or transitive? . Is it an equivalence or linear.

Functions
0
1.5.8 Definition. A relation f is called a function iff (x, y) ∈ f and (x, y ) ∈ f
0
⇒ y = y . Since the function is a relation, all the foregoing on relations hold for the
functions and therefore will be taken for granted in the sequel.

Functions as Relation

(1) For a f , if (x, y) ∈ f , then the image of x under f is the singleton {y}. AS
such f (x) = y and f (x) is said to be the value of f at x.

(2) Two functions f1 and f2 are equal iff dom(f1 ) = dom(f2 ) and f1 (x) = f2 (x) ∀
x ∈ dom(f1 ).

(3) Recall the converse relation. Note that the relation R4 though is a function,
is such that its converse R−1
4 is not a function. Now in the special case where

in addition to f being a function, its converse f −1 is also a function , call f −1

the inverse of f .

(5) By f : X −→ Y is a function, it is meant that f is a function with dom(f ) = X

and rge(f ) ⊂ Y . It is read: f is a function of X into Y . Y is called the range
space of f .

1.5.9 Remark. Note that other forms of functions exist.

(i). X −→f Y

10
 Figure 1.1: Diagram of Composition of f and g

(ii). f := x 7−→ y : X −→ Y

(iii). x 7−→ f (x) : X −→ Y

1.5.10 Definition. Let f : A −→ B and g : B −→ C be two functions. The

composition g·f : A −→ C is defined by g·f (a) = g(f (a)). For example, f : x 7−→ x2
and g : x 7−→ x + 3, then g · f : x 7−→ x2 + 3 and f · g : x 7−→ (x + 3)2

1.5.11 Theorem. The composition f · g of two functions f : X1 −→ Y1 and

g : X2 −→ Y2 is a function. Moreover, dom()f · g) = {x : g(x) ∈ dom(f )} and
f · g(x) = f (g(x)) ∀x ∈ dom(f · g).

Proof. By definition,

f · g = {(x, y) : ∃z 3 (x, z) ∈ g and (z, y) ∈ f }

. We shall prove that f · g is a function. Take (x, y) ∈ f · g. Then, by the definition.

0 0 0 0
∃z, z 3 (x, z) ∈ g, (z, y) ∈ f, (x, z ) ∈ g, (z , y ) ∈ f g being a function, it follows
0 0 0
that z = z, and so (z, y) ∈ f and (z , y) ∈ f . Thus y = y also. Hence we have

11
shown that f ·g is a function. Now from the definition, (x, y) ∈ f ·g iff ∃x 3 (x, z) ∈ g
and (z, y) ∈ f iff x ∈ dom(g), z = g(x) and (z, y) ∈ f iff x ∈ dom(g), g(x) ∈ dom(f )
and y = f (g(x)). Thus

dom(f ·g) = {x : x ∈ dom(g) and g(x) ∈ dom(f ) and (f ·g) = f (g(x)) ∀x ∈ dom(f ·g)}

Classification of Functions

1.5.12 Definition. Let f : X −→ Y be a function.

0
(i.) f is said to be an injection (or one-to-one) iff (x, y) ∈ f and (x , y) ∈ f
0
⇒x=x.

(ii). f is said to be a surjection (or an onto function) iff y ∈ Y implies that ∃x ∈

X 3 (x, y) ∈ f .

(iii.) If f is both an injection and a surjection, then it is said to be a bijection.

1.5.13 Remark. :Note that an injection has an inverse.

1.5.14 Example. (1). The identity function: Let X be a set. The function i :=
x 7−→ x : X −→ X is called the identity function. Note that i (x) = x ∀
x ∈ X. This function is a bijection.

(2). The inclusion function: Let S be a subset of a set Z. The function j := x 7−→
x : S −→ Z is called the inclusion function. Note that R, from X = {2, 3, 4, 7}
into Y = {2, 3, 5} given above is an inclusion function.
Reason: R1 := {(x, y) ∈ X × Y : x = y} = {(2, 2), (3, 3)}. Thus R1 := x 7−→
k : {2, 3} −→ Y . R1 is an injection. Note how this function is to the identity
function.

12
 (3). The characteristic function: Again let S be a subset of a set Z. The func-
tion 
 1 if x ∈ S
χS (x) := : Z −→ {0, 1}
 0 if x ∈ Z/S

is called the characteristic function of S with respect to Z. What is the range

of χS if S = Z.

(4). The relation R4 = {(2, 2), (4, 2), (3, 3)} from X into Y given above is a func-
tion. In fact
R4 := x 7−→ y 3 y|x : {2, 3, 4} −→ {2, 3}

Sequences

1.5.15 Definition. Let Y be a set. Any function x : N −→ Y is called a sequence

(inY ). Let x : N −→ Y be a sequence since for each n ∈ N, x(n) is unique , the
sequence x is identified completely when the set {x(n) : n ∈ N} is known. Write xn
for {x(n) : n ∈ N}. Call (xn ) the sequence x : N −→ Y and xn its nth term.

Problem

(1) Is the empty relation a function?? Give reasons.

(2) Prove that the composition of two injections ia an injection

(3) Let f : X −→ Y and g : Y −→ Z be two bijections. Prove that g◦f : X −→ Z

is a bijection.

(4) Prove that a map f : A −→ B is a bijection iff there is a map g : B −→ A

such that f ◦ g := identity and g ◦ f := identity. Show also that g = f −1 and
is uniquely determined.

13
 (5) Let f : A −→ B and g : B −→ C be bijections. Then (g ◦ f ) is a bijection
and (g ◦ f )−1 = f −1 ◦ g −1 (Hint: use (4) above).

(6) Let f : A −→ B be a function, C1 , C2 ⊂ B and D1 , D2 ⊂ A. Prove

(a). f −1 (C1 ∪ C2 ) = f −1 (C1 ) ∪ f −1 (C2 )

(b). f (D1 ∪ D2 ) = f (D1 ) ∪ f (D2 )

(c). f −1 (C1 ∩ C2 ) = f −1 (C1 ) ∩ f −1 (C2 )

(d). f (D1 ∩ D2 ) = f (D1 ) ∩ (D2 )

Countability of Sets

Before introducing the concept of countability of sets, we need the following:

1.5.16 Definition. (Equivalence Sets) Two sets A and B are said to be equivalent
iff there is a bijection g : A −→ B.

1.5.17 Example. (1) The sets {1, 2, 3, 4, 5} and {a, e, i, o, u} are equivalent.
Reason:

g
1 /a

2 / e

3 / i

4 / o

5 / u

14
 Figure 1.2:

The matching g displayed above is a bijection between {1, 2, 3, 4, 5} and {a, e, i, o, u}.

(2) Consider the set Y of all world capitals and the set X of all countries. The
function {(x, y): y is a capital of the country x} ⊂ X ×Y is a bijection between
X and Y . Hence X and Y are equivalent.

Finite Sets

1.5.18 Definition. A set F is finite iff it is empty or equivalent to {1, 2, 3, ..., n}

for some fixed non-negative integer n. Moreover, we denote by |F | the cardinality
of the set F . In this case, |F | = n. A set which is not finite is infinite.

1.5.19 Theorem. A finite set cannot equivalent to a proper subset of itself.

Proof. We shall prove by contradiction by assuming that a finite set is equivalent

to a subset of itself. To do this, let F be a finite set with |F | = n and F0 ⊂ F .
Let |F0 | = n0 < n. Thus ∃ a bijection g : {1, 2, 3, ..., n0 } −→ F0 ; i.e ∃ a bijection
h : F0 −→ F . Therefore, the composition h ◦ g(displayed below ) is a bijection
between {1, 2, 3, ..., n0 } and F .

15
 g
{1, 2, 3, · · · , n} /
? F0

h◦g h
&
F

Figure 1.3: Diagram of Composition of g and h

Thus by definition of cardinality of sets, F has both n and n0 elements with

n0 < n. This is a contradiction. Hence F0 is not equivalent with F .

1.5.20 Definition. A set X is said to be countable iff it is finite or equivalent to

N, otherwise the set is said to be uncountable. An infinite countable set is called a
countably infinite set.

1.5.21 Theorem. Any subset of a countable set is countable.

16
Proof. Let S be the given countable set and assume A ⊆ S. If A is finite, then
there is nothing to prove, so we can assume that A is infinite ( which means S
is also infinite). Let s = {sn } be an infinite sequence of distinct terms such that
S = {s1 , s2 , ...}. Define a function on the positive integers as follows: Let k(1) be the
smallest positive integer m such that sm ∈ A. Assuming that k(1), k(2), ..., k(n − 1)
have been defined , let k(n) be the smallest positive integer m > k(n − 1) such
that sm ∈ A. Then k is order preserving (i.e m > n ⇒ k(m) > k(n)). Form the
composition function S · K. The domain of s ◦ K is the set of positive integers and
the range of s ◦ k is A. Furthermore, s ◦ k is one-to-one since

s[k(n)] = s[k(m)]

implies
sk(n) = sk(m) ,

which implies k(n) = k(m) and this implies n = m. This proves the theorem. 2

1.5.22 Theorem. (1) The union of two countable sets is countable.

(2) The union of a finite family of countable set is countable.

(3) The union of a countable family of countable sets is countable.

1.5.23 Theorem. Let F be a collection of sets. Then for any set B, we have
[ \
B− A= (B − A)
A∈F A∈F

and
\ [
B− A= (B − A)
A∈F A∈F

Proof. If x ∈ B − S, then x ∈ B but x ∈

/ S. Hence, it is not true that x belong to
at least one A in F . Hence, for every A ∈ F , x ∈ B − A. But this implies x ∈ T ,

17
so that B − S ⊆ T . Reversing the steps, we obtain T ⊆ B − S and this proves that
B − S = T . (Prove the second statement)

1.5.24 Definition. If F is a collection of sets such that every two distinct sets in
F are disjoint, then F is said to be a collection of disjoint set.

1.5.25 Theorem. If F is a countable collection of disjoint sets, say F = {A1 , A2 , ...},

such that each set An is countable, then the union ∞
S
k=1 Ak is also countable.
∞
S
Proof. Let An = {a1,n , a2,n , ...}, n = 1, 2, 3, ... and let S = Ak . Then every
k=1
element x of S is in at least one of the sets in F and hence x = am,n for some pair of
integers (m, n). The pair (m, n) is uniquely determined by x, since F is a collection
of disjoint sets. Hence the function f defined by f (x) = (m, n) if x = am,n , x ∈ S,
has domain S. The rang f (S) is a subset of Z+ × Z+ (where Z+ is the set of positive
integers) and hence is countable. But f is one-to-one and therefore S is equivalent
to f (S) =⇒ S is countable.
From the foregoing , we have the following.

1.5.26 Theorem. If F is a countable collection of countable sets, then the union

of all sets in F is also countable.

Exercise

(1.) Prove that the set {..., −2, −1, 0, 1, 2, ...} is countable.
Solution: Define ϕ : {..., −2, −1, 0, 1, 2, ...} ⇒ {1, 2, 3, ...}



 2n, if n > 0

ϕ= 1, if n = 0 i.e.N ∪ {0} ∪ {−n, n ∈ Z}


 −2n + 1, if n < 0


Verify that this is a bijection.

18
(2.) Prove that the set Q of the rational number is countable.
Solution: Let A0 = {0}, A1 = {± 11 , ± 21 , ±...± n1 } = Z {0}, A2 = {± 12 , ± 22 , ±...± n2 }
... Ak = {± k1 , ± k2 , ±..., ± nk } We note that
∞
[ [ m
Q= Aj = { : m ∈ Z}
j∈N n=1
n

= a union of a countable collection of countable sets

Therefore, Q is countable.

1.6 The Real Number System

1.6.1 Definition. (Field) Let F be a set on which two binary operations + and ·
are defined such that ∀x, y ∈ F ,

(1) x + y ∈ F

(2) x · y ∈ F. We say that F is closed under the binary operations + and ·

and (F, +, ·) is called a field if in addition, the following properties also hold
∀ x, y, z ∈ F ,

(3) x + (y + z) = (x + y) + z

(4) x · (y · z) = (x · y) · z (Associative property)

(5) x + y = y + x

(6) x · y = y · x (commutative property)

(7) x · (y + z) = (x · y) + (x · z) (Distributive property)

(8) ∃0 ∈ F 3 0 + x = x + 0 = x. 0 is called the additive identity

19
 (9) ∃e ∈ F 3 e.x = x = x.e. e is called the multiplicative identity.

(10) For each element x ∈ F, ∃ an element −x ∈ F 3 x + (−x) = (−x) + x = 0.

The element (−x) is the additive inverse of x

(11) For each element x ∈ F , x 6= 0∃ an element x−1 ∈ F such that x · (x−1 ) =

(x−1 ) · x = e. x−1 is called the multiplicative inverse of x

1.6.2 Remark. (i) N is not a field.

(ii) Z is not a field.

(iii) Q is a field.

(iv) R is a field.
To define the real number system, we notice that not all known numbers are in
Q as seen in the following theorem.
√
1.6.3 Theorem. The number 2∈
/ Q.

Proof. To prove this , we need the following lemma.

1.6.4 Lemma. : If x is odd, then x2 is odd and if y is even, then y 2 is even.

Proof. If x is odd, it can be written in the form, x = 2n + 1 and

x2 = (2n + 1)2 = 4n2 + 4n + 1 = 2(2n2 + 1) + 1 = 2m + 1
∴ x2 is an odd number.
If y is even , then y = 2n and y 2 = 4n2 = (2n)2 = 2(2n2 ) = 2m
∴ y 2 is an even number.

√
We now prove the theorem. Suppose that 2 ∈ Q, it follows that
√ p
2= . (i)
q

20
 √
Let (p, q) = 1, and from 2 = pq , we have

p2 = 2q 2 . (ii)

From (ii), we conclude that p2 is an even number =⇒ p is even. Since p and q are
relatively prime =⇒ q is an odd number since p is even , let p = 2m. From (ii),
2q 2 = p2 = (2m)2 = 2q 2 = 4m2 or q 2 = 2m2 =⇒ q 2 is even =⇒ q is even. This is a
√ √
contradiction that (p, q) = 1. ∴ 2 6= pq and 2 ∈
/ Q.

1.6.5 Remark. (1.) Note that the following inclusions exists N ⊂ Z ⊂ Q ⊂ R.

√
(2.) The set where 2 belongs (= real number that are not rational) irrational
numbers denoted I. e.g e, π, eπ .

(3.) I ∩ Q = ∅.

(4.) I ∪ Q = R.

(5.) R is a field under the binary operations + and · and Q is a subfield while I is
not a field.

1.6.6 Definition. The set of positive elements in R is given by R+ with the following
properties:

(1.) 0 ∈
/ R+

(2.) For any element x ∈ R, only one of the following relation holds: −x ∈ R+ ,
x = 0, x ∈ R+ .

(3.) ∀ x, y ∈ R+ , x + y ∈ R+ and x · y ∈ R+ .

1.6.7 Definition. (R as an ordered field ) ORDER AXIOM

(1.) For any x, y ∈ R, we shall denote x + (−y) = x − y.

21
 (2.) Define > on R by saying that if x > y, then x − y ∈ R+ . Thus x > 0 =⇒ x ∈
R+ .

(3.) We shall write x > y ⇐⇒ y < x.

(4.) The relation < or > is called an order on R and we call R an ordered field.

1.6.8 Theorem. Let x, y ∈ R, then

(i.) x · 0 = 0 · x = 0.

(ii.) x · (−y) = (−x) · y = −xy.

(iii.) (−x)(−y) = xy.

(iv.) The identities 0 and 1 are unique.

(v.) −x is uniquely determined for x ∈ R.

(vi.) If x 6= 0, x ∈ R, x−1 is unique.

1.6.9 Definition. (Interval in R) Let a, b ∈ R with a < b, then the set

(i) (a, b) = {x : a < x < b} is called an open interval in R. Note that a ∈

/ (a, b)
and b ∈
/ (a, b).

(ii) [a, b] = {x : a 6 x 6 b} is called a closed interval in R. a ∈ [a, b] and b ∈ [a, b].

(iii) [a, b) = {x : a 6 x < b} and (a, b] = {x : a < x 6 b}. These are called
half-open or half-closed intervals.

22
1.7 Absolute Values of Numbers
1.7.1 Definition. If x ∈ R, we define the absolute value of x by



 +x, if x > 0

|x| = 0, if x = 0


 −x,

if x < 0

1.7.2 Theorem. For all a, bc ∈ R,

(1) | − a| = |a|.

(2) |ab| = |a||b|

(3) If c > 0, then |a| < c iff −c < a < c.

(4) −|a| 6 a 6 |a|.

Proof. (1) If a = 0, then |a| = 0. |0| = 0 = | − 0|.

If a > 0, then −a < 0. |a| = a = −(−a) = −|a|.
If a < 0, then −a > 0. | − a| = −a = |a|.

(2) If a = 0 and b = 0, then |ab| = |0 · 0| = |0||0| = |a||b| = 0

If a > 0, b > 0, then ab > 0. |ab| = ab = |a||b|.
If a > 0 and b < 0, then ab < 0. |ab| = −ab = a(−b) = |a||b|.

(3) If |a| 6 c, then +a 6 c and −a 6 c. i.e −c 6 a... − c 6 a 6 c

(4) −|a| 6 a 6 |a|. Let c = |a|, From (iii) above, =⇒ −c 6 a 6 c.

1.7.3 Theorem. (Triangle Inequality) For any a, b ∈ R, then |a + b| 6 |a| + |b|.

Proof. Since −|a| 6 a 6 |a| and −|b| 6 b 6 |a|. Adding, we give −(|a| + |b|) 6
a + b 6 |a| + |b| , this implies a + b 6 |a| + |b|

23
Also
−(|a| + |b|) 6 a + b =⇒ −(a + b) 6 |a| + |b| , ∴, |a + b| 6 |a| + |b|

1.7.4 Corollary. For any x, y ∈ R

(i) ||x| − |y|| 6 |x − y|

(ii) |x − y| 6 |x| + |y|

Proof. (i) x = x − y + y, this implies |x| = |x − y + y| 6 |x − y| + |y|, then we

have |x| − |y| 6 |x − y|
y = y − x + x, then we have |y| = |y − x + x| 6 |y − x| + |x|
|y| − |x| 6 |y − x| = |x − y|, ∴, we have −(|x| − |y|) 6 |x − y| and ∴, we have
also ||x| − |y|| 6 |x − y|.

(ii) |x − y| 6 |x| + |y| and |x − y| = |x + (−y)| 6 |x| + | − y| = |x| + |y|

1.7.5 Corollary. For any x1 , x2 , ..., xn ∈ R,

then |x1 + x2 + ... + xn | 6 |x1 | + |x2 | + ... + |xn |

1.8 Infimum and Supremum

1.8.1 Definition. Let A be a non-empty subset of R, then

(i). A is said to be bounded above if there exists b ∈ R such that x 6 b ∀ x ∈ A.

(ii). A is said to be bounded below if there exists a ∈ R such that a 6 x ∀ x ∈ A.

b is called an upper bound and a is called a lower bound.

(iii). A is said to be bounded if it is bounded from above and from below; otherwise,
it is unbounded.

24
1.8.2 Example. (i). If A = (−20, −19] . A is bounded below by −20 and above
by −19.

(ii). A = {x ∈ Z : x > 5}. Lower bounds of A = {y : y 6 5}. A is unbounded

above. Hence A is not bounded since it is not bounded above but only bounded
below.

1.8.3 Definition. Let A be a non-empty subsets of R.

(i). An element c ∈ A is called a minimum element of A if c 6 x ∀ x ∈ A.

(ii). An element d ∈ A is called a maximum element of A if x 6 d ∀ x ∈ A.

1.8.4 Remark. (1). A set A may have no maximum or minimum element, though
may be bounded e.g A = ( 21 , 1). A is bounded below by 1
2
and above by 1 but
1
2
∈
/ A and 1 ∈
/ A. ∴ A has no maximum or minimum element.

(2). If a set is bounded from above, then it has infinitely many upper bounds. If a
set is bounded below, then it has infinitely many lower bounds.

1.8.5 Definition. (1). If A is bounded from above, the minimum of all the upper
bounds of A is called the least upper bound (l.u.b) or supremum of A i.e
sup A := {y ∈ R: if b ∈ R is an upper bound of A, y 6 b}.

(2) If A is bounded from below, the maximum of all the lower bounds of A is called
the greatest lower bound of A or the infimum of A defined by: inf A := {z ∈ R :
if a ∈ R is a lower bound of A, then z > a}

1.8.6 Remark. The Supremum or Infimum of a set A need not belong to the A.

1.8.7 Proposition. Let A be a non-empty subset of R, if A is bounded, then

(1) Supremum of A is unique.

25
 (2) Infimum of A is unique.

Proof. (1) Suppose A is bounded above and b = sup A, c = sup A. B y definition,

b 6 c and c 6 b implies b = c.

(2) Exercise.

1.9 Completeness of Axiom:

(1) Every non-empty subset of R that is bounded above has a supremum in R.

(2) Every non-empty subset of R that is bounded below has an infimum in R.

1.10 Some Properties of the Supremum:

The first property called the approximation property states that a set with a supre-
mum contains numbers arbitrary close to its supremum.

1.10.1 Theorem. (Approximation Property) Let S be a non-empty set of real

numbers with a supremum, say b = sup S. Then for every a < b , there exists some
x ∈ S such that a < x 6 b.

1.10.2 Theorem. (Additive Property) Given non-empty subsets of A and B of R,

let C denote the set
C = {x + y : x ∈ A, y ∈ B}

If each of A and B has a supremum, then C has a supremum and sup C = sup A +
sup B.

1.10.3 Theorem. (Comparison Property) Given non-empty subsets S and T of R

such that s < t ∀ s ∈ S and t ∈ T . If T has a supremum, then S has a supremum

26
and
sup S 6 sup T

1.10.4 Theorem. Archimedean Property (Eudoxus Theorem) If x ∈ R, then there

exists n ∈ N such that x < n.

Proof. Suppose(by contradiction) that n 6 x, i.e x is an upper bound of N. By

completeness axiom, N has a supremum in R. So let U = sup N. We have u−1 < U .
It follows that there exists m ∈ N such that u − 1 < m < u. Then u < m + 1. But
m + 1 ∈ N, thus this last inequality violates the definition of n as a supremum of N:
hence the assumption n 6 x is wrong and therefore x < n.

1.10.5 Corollary. Let x, y be positive real numbers

(i) ∃n ∈ N such that x < ny.

1
(ii) ∃n ∈ N such that 0 < n
< y.

(iii) ∃n ∈ N such that n − 1 < x < n

x
Proof. (i) x, y > 0 thus z = y
> 0. By Archimedean property, ∃n ∈ R such that
x
y
= z < n. ∴ x < ny.

1
(ii) In (i) above, set x = 1. Then 1 < ny or n
< y.

(iii) By the Archimedean property, the subset A = {m ∈ N : x < m} =

6 ∅. Let
n = inf A, then n − 1 < x < n

1.10.6 Remark. Q has been shown to be countable while I = R−Q is uncountable.

We shall show that between any two real numbers, ∃ a rational number, i.e Q is
dense in R.

27
1.10.7 Theorem. (Density theorem)
If x, y ∈ R, with x < y, then ∃ a rational number r ∈ Q such that x < r < y.

1
Proof. Suppose that x > 0, then y − x > 0. Let z = y−x
> 0. By Archmedean
1
theorem, ∃ n ∈ N such that y−x
< n. ∴ ny − nx > 1. By corollary (iii) above, ∃
m ∈ N succh that nx < m, and m < ny. Since m < nx + 1 < ny or nx < m < ny
m
or x < n
< y.

1.10.8 Theorem. The set of all real numbers is uncountable

Proof. It suffices to show that the set of x satisfying 0 < x < 1 is uncountable. If
the real numbers in this interval were countable, there would be a sequence s = {sn }
whose terms would constitute the whole interval. We shall show this is impossible
by constructing, in the interval, a real number which is not a term of this sequence.
Now write each sn as an infinite decimal:

sn = 0.un,1 un,2 un,3 ...

where each un,i is 0, 1, ..., or 9. Consider the real number y which has the decimal
expansion
y = 0.v1 v2 ...

where 
n,n 6
 1 if u = A
Vn =
 2 if u = 1
n,n

Then no term of the sequence {sn } can be equal to y, since y differs from s1
in the first decimal place, differs from s2 in the second decimal place ,..., sn in the
nth decimal place. (A situation like sn = 0.19999... and y = 0.20000... cannot occur
here because of the way vn are chosen) since 0 < y < 1, the theorem is proved.

1.10.9 Corollary. No interval on R is countable.

28
Proof. Let (a, b) be an interval. Suppose (a, b) is countable. Then enumerate (a, b)
as x1 , x2 , ..., xn , .... As above, this leads to be length of (a, b) is 0, contrary to the
fact that its length is b − a > 0. Hence (a, b) is not countable.

1.10.10 Corollary. (1) The irrational numbers in any interval are uncountable.

(2) Between any two real numbers, there is always an irrational number and so
an infinite number of them.
Note: Let r1 , r2 ∈ Q with r2 6= 0, and let t be an irrational number. Then r1 +r2 t
is irrational number.

Proof. Suppose r1 + r2 t is a rational number, q say. Then r1 + r2 t = q. From where

q−r1
it follows that t = r2
which is a rational number. Hence r1 + r2 is an irrational
number.

1.10.11 Theorem. Let Z+ denote the set of all positive integers. Then the Carte-
sian product Z+ × Z+ is countable.

Proof. Define a function f on Z+ × Z+ as follows. f (m, n) = 2m 3n if (m, n) ∈

Z+ × Z+ . Then f is one-to-one on Z+ × Z+ and the range of f is a subset of Z+ .
Whence the result.

1.11 Problem:
Use the Eudoxus theorem to established each of the following statements.

t
(1) Suppose that z, t ∈ R+ . Then ∃ m ∈ N such that real number m
satisfies
t t
0< m
< z. If in addition, t is an irrational number. then m
is irrational.

(2) Suppose that t is an irrational number and that y, zR are such that y < z.
Then

29
(i) ∃ r1 , r2 ∈ Q such that the irrational number r1 +r2 t satisfies y < r1 +r2 t <
z

(ii) ∃ r ∈ Q such that the irrational number rt satisfies y < rt < z

30
Chapter 2

Dedekind Cut

There are two standardly used approaches of constructing the real numbers from
the rational namely:

(1) Using equivalence classes of Cauchy sequences of rational numbers.

(2) Dedekind cuts.

The initiative behind Dedekind cuts is simple, even though their formal definitions
appear technical. Here, we want to use the rational numbers to construct the real
numbers, and the key observation is that every real numbers can be characterized
by the set of rational numbers that are greater than it.
For example, the real number 1 is characterized by the set {x ∈ Q : x > 1}. The
number 1 ofcourse is a rational number and so the set {x ∈ Q : x > 1} is defined in
the realm of rational numbers, which is all we require. On the other hand, the set
√ √
{x ∈ Q : x > 2} is not describable using only rational numbers, because 2 is not
a rational number. Instead, we define Dedekind cuts to be subsets of Q that behave
as sets of the form {x ∈ Q : x > r} for real numbers r while using a definition that
is strictly in terms of the rational numbers.

31
2.0.1 Definition. Let A ⊆ Q be a set. Then, A is a Dedekind cut if the following
three properties hold:

(1) A 6= ∅, A 6= Q.

(2) Let x ∈ A. If y ∈ Q and y > x, then y ∈ A.

(3) Let x ∈ A. Then there is some y ∈ A such that y < x.

The following lemma show that Dedekind cuts exists and are plentiful.

2.0.2 Lemma. Let r ∈ Q. Then the set {x ∈ Q : x > r} is a Dedekind cut.

Proof. Let D = {x ∈ Q : x > r}. We show that D satisfies the three parts of the
definition of Dedekind cuts.

(a) We know that r − 1, r + 1 ∈ Q and that r − 1 < r < r + 1. Hence r − 1 ∈

/D
and r + 1 ∈ D, and therefore, D =
6 ∅ and D =
6 Q.

(b) Let m ∈ D and let y ∈ Q. Suppose y > m. Because m > r, it follows that
y > r. Hence y ∈ D.

n+r n+r
(c) Let n ∈ D. Then n > r. It follows that 2
∈ Q and r < 2
< n. Hence
n+r
2
∈ D.

2.0.3 Remark. To find a Dedekind cut that is not of the form in the above lemma,
we consider the set of rational numbers greater than a real number that is not
rational. For example: Let T = {x ∈ Q : x > 0 and x2 > 2}. Then, T is a Dedekind
cut, since it has the form {x ∈ Q : x > r} for some r ∈ Q, then r2 = 2. By there is
no rational number x such that x2 = 2 and it follow that T is a Dedekind cut that
is not of the form given in the lemma.

32
2.0.4 Definition. Let r ∈ Q. The rational cut at r, denoted Dr is the Dedekind
cut Dr = {x ∈ Q : x > r}. An irrational cut is a Dedekind cut that is not a rational
cut at any rational number. Properties of Dedekind’s cut will be given below:

Properties of Dedekind cuts

2.0.5 Lemma (1). Let A ⊂ Q be a Dedekind cut.

(1) Q − A = {x ∈ Q : x < a ∀ a ∈ A}.

(2) x ∈ Q − A. If y ∈ Q and y 6 x, then y ∈ Q − A.

Proof. (1) Let x ∈ Q − A . Let a ∈ A. We know that x < a or x = a or x > a.

If x = a, then x ∈ A. by definition of a, which is a contradiction. If x > a,
then x ∈ A by (ii) of Dedekind cut definition which is a contradiction. Hence,
x < a. We conclude that Q − A ⊆ {x ∈ Q : x < a ∀ a ∈ A}. Now let
y ∈ {x ∈ Q : x < a ∀ a ∈ A}. If y ∈ A, we would then have y < y, which is a
contradiction . Hence, y ∈ Q − A. We deduce that Q − A ⊇ {x ∈ Q : x < a ∀
a ∈ A}. Therefore, Q − A = {x ∈ Q : x < a ∀ a ∈ A}.

(2) The part easily follows from part(1).

2.0.6 Lemma (2). Let A, B ⊆ Q be Dedekind cuts. Then one has A $ B or A = B

or B $ A holds.

Proof. If A = B, there is nothing to prove. So, assume that A ∈

/ B. There are now
two cases. First, suppose that there is some a ∈ A, such that a ∈ Q − B. Then by
the first part of lemma (1), we know that a < b ∀ b ∈ B. By (ii) of Dedekind cuts
definition, it follows that b ∈ A ∀ b ∈ B. Hence, B ⊂ A. Since we are assuming
B ∈
/ A, then B $ A. Secondly, there is some d ∈ B such that d ∈ Q − A, and a
similar argument shows that A $ B (complete this!).

33
2.0.7 Lemma (3). Let A be a non-empty family of subsets of Q. Suppose that X
S S
is a Dedekind cut for all X ∈ A. If X∈
/ Q, then X∈ / Q is a Dedekind cut.
X∈A X∈A
S
Proof. Let B = X. Assume that B ∈
/ Q. We will show that B satisfies the
X∈A
three parts of Dedekind cut definition.

(a) We know that X 6= ∅ ∀ X ∈ A, so, B 6= ∅. By hypothesis, B 6= Q.

(b) Let b ∈ B and y ∈ Q. Suppose that y > b. We know that b ∈ X for some
X ∈ A. By (ii) of Dedekind cut definition, applied to X, we see that y ∈ X.
Hence y ∈ B.

(c) Let c ∈ B. Then c ∈ D for some D ∈ A. By (iii) of Dedekind cut definition,

there is some z ∈ D such that z < c. Hence y ∈ B.

The following lemma provides the law for some of basic operations for the real
numbers.

2.0.8 Lemma. (4) Let A, B ⊆ Q be Dedekind cuts.

(1) The set {r ∈ Q : r = a + b for some a ∈ A and b ∈ B} is a Dedekind cut.

(2) The set {r ∈ Q : −r < c for some c ∈ Q − A} is a Dedekind cut.

(3) Suppose that 0 ∈ Q − A and 0 ∈ Q − B. The set {r ∈ Q : r = ab for some

a ∈ A and b ∈ B} is a Dedekind cut.

Proof. We shall prove (1), (2) and (4) only.

(1) Let M = {r ∈ Q : r = a + b for some a ∈ A and b ∈ B}. We will show that

M satisfies the (3) properties of Dedekind cuts definition.

34
 (i) We know that A 6= ∅ and A 6= Q, and B 6= ∅, B 6= Q. Let x ∈ A,
p ∈ Q − A, y ∈ B and q ∈ Q − B. Then, x + y ∈ M , so, M 6= ∅. By
lemma (1) above, p < a ∀ a ∈ A and q < b ∀ b ∈ B. It follows that
p + q < a + b ∀ a ∈ A, b ∈ B. Hence, p + q ∈ Q − M and so, M 6= Q.

(ii) Let c ∈ M , and let y ∈ Q. Suppose that y > c. We know that c = a + b

for some a ∈ A, b ∈ B. Then y = [a + (y − c)] + b. Because y > c, then
a + (y − c) > a and hence by (ii) of Dedekind cuts definition, we see that
a + (y − c) ∈ A. It follows that y ∈ M .

(iii) Let d ∈ M . We know that d = s + t for some s ∈ A, and t ∈ B. Applying

(iii) of Dedekind cuts definition to A, we see that there is some g ∈ A
such that g < s. Then g + t ∈ M and g + t < s + t = d

(2) Let N = {r ∈ Q : −r < c for some c ∈ Q − A}. We show that N satisfies the
three parts of the Dedekind cuts definition.

(i) Clearly, A 6= ∅ and A 6= Q. Let b ∈ Q − A. By lemma1(2), we deduce

that b − 1 ∈ Q − A. Then −(b − 1) ∈ N . Hence N 6= ∅. Next, let
a ∈ A. Then −(−a) ∈
/ Q − A and therefore by lemma1(2), we know that
−(−a) ≮ g for all g ∈ Q − A. Hence −a ∈ Q − N . So, N 6= Q.

(ii) Let d ∈ N and let y ∈ Q. Suppose that y > d. It follow that −d > −y.
By definition of N , we know that −d < c for some c ∈ Q − A. Then
−y < c and therefore, y ∈ N .
e+(−c)
(iii) Let c ∈ N . Then, −e < c for some c ∈ Q − A. Hence e > −c. Let 2
.
If follow that s ∈ Q and −c < s < e. Hence, −s < c. It follows that
s ∈ N.

(4) Because q ∈ Q − A and q > 0, it follow from lemma1(1) that 0 < q < a ∀

35
 1
a ∈ A. Let R = {r ∈ Q : r > 0 and r
< c for some c ∈ Q − A}. We show that
R satisfies the three parts of a Dedekind cuts definition.

q
(i) Clearly 0 ∈
/ R, so R 6= Q. Thus, we have that 0 < 2
< q. Then
2
q
= ( 2q )−1 ∈ R, and hence R ∈
/ ∅.

(ii) Let w ∈ R and let y ∈ Q. Suppose y > w. We know that w > 0 and
1 1
hence y > 0. It follows that y
6 w
. By definition of R, we know that
1 1
w
< c for some c ∈ Q − A. Then y
< c and hence y ∈ R.
1 1
(iii) Let m ∈ R. Therefore, m > 0, and m
∈ Q − A and m
< c for some
1 m+ 1c
c ∈ Q − A. Hence 0 < c
< m. Let t = 2
. It follows that t ∈ Q and
1 1
0< c
< t < m. Therefore, 0 < t
< c, and hence t ∈ R.

The next lemma will be used to show that the real numbers satisfy some basic
algebraic properties. Though, it applies the well-ordering principle.

2.0.9 Lemma. (5) Let A ⊂ Q be a Dedekind cut. Let y ∈ Q.

(1) Suppose that y > 0. Then there are u ∈ A and v ∈ Q − A such that y = u − v
and v < e for some e ∈ Q − A.

2 Suppose that y > 1 and that there is some q ∈ Q − A such that q > 0. Then,
there are r ∈ A and s ∈ Q − A such that s > 0, and y > rs , and s < g for some
g ∈ Q − A.

Proof (Rudin, Bloch, ...). .

2.1 Constructing the Real Numbers

We want the set of real numbers to contain a copy of the set of rational numbers,
√
and it should also have numbers such as 2, where the set of rational numbers has

36
 ”gaps”.

2.1.1 Definition. The set of real numbers, denoted R is defined by R := {A ⊆ Q : A

is Dedekind cut}. Given that Dedekind cuts are sets of rational numbers, we define
the relation ”<” on the real numbers in terms of the relation ”⊂” on the sets of
rational numbers.

2.1.2 Definition. The relation < on R is defined by A < B ⇐⇒ A % B, for all

A, B ∈ R. The relation A 6 B on R is defined by A 6 B iff A ⊇ B ∀ A, B ∈ R.
The following definition of addition and negation for real numbers makes sense.

2.1.3 Definition. The binary operation ” + ” on R is defined by A + B = {r ∈ Q :

r = a + b for some a ∈ A and b ∈ B}, ∀ A, B ∈ R.
The unary operation ” − ” on R is defined by −A = {r ∈ Q : −r < c for some
c ∈ Q − A} , ∀ A ∈ R. In order to define multiplication and multiplicative inverse
for real numbers which require various cases, we shall need the following lemma.

2.1.4 Lemma. (Proposition 1) Let A ∈ R and let r ∈ Q

(1) A > Dr if f there is some q ∈ Q − A such that q > r.

(2) A > Dr iff r ∈ Q − A iff a > r ∀ a ∈ A.

(3) If A < D0 , then −A > D0 .

Proof.

(1) and (2) follow from the statements that if A ⊆ Q is a Dedekind cut and r ∈ Q,
then A $ D iff there is some q ∈ Q − A such that r < q. And also A ⊆ Dr iff
r ∈ Q − A iff r < a ∀ a ∈ A.

(3) Suppose A < D0 . Then A % D0 . Because D0 = {x ∈ Q : x > 0}, it follows that

there is some q ∈ A such that q 6 0. By (ii) of the definition of Dedekind cuts, we

37
deduce that 0 ∈ A. Hence, 0 ∈
/ Q − A and therefore, −0 ∈
/ Q − A, which implies that
0∈
/ −A and hence 0 ∈ Q − (−A). By (2) of this lemma, we have that −A > D0 .
The following definition of multiplication and multiplicative inverse makes sense due
to lemma4 (3),(4) and Proposition1.

2.1.5 Definition. The binary operation ” · ” on R is defined by




 {r ∈ Q : r = ab f or some a ∈ A and b ∈ B}; if A > D0 and B > D0


−[(−A).B]; if A < D0 and B > D0


A·B =


 −[A.(−B)]; if A > D0 and B < D0



 (−A).(−B); if A < D and B < D
0 0

The unary operation −1 on R − D0 is defined by


 {r ∈ Q : r > 0 and 1 < 0 f or some c ∈ Q − A}; if A > D
0
A−1 = r
 −A−1 ; if A < D0

38
Chapter 3

Sequences

Recall that a sequence in X is a function x := x 7−→ xn : N −→ X.

3.0.1 Definition. A sequence in R is a function x : N −→ R : n 7−→ xn . Here, xn

is called the n-th term of the sequence.
Notation: Sequences whose terms are xn is denoted by {xn } or (xn )

Example of Sequences

(1) 1, 2, 3, ..., n, n + 1, ... = (n)

1 1 1
(2) , , , ..., 21n , ...
2 22 23
= ( 21 )n .

(3) ( n1p ), p > 0, p ∈ Q.

(4) (sin nπ).

(5) ((1 − n1 )n )

3.0.2 Definition. Let {xn } be a sequence. Suppose that ∃ x0 ∈ R, such that given
 > 0, ∃ N ∈ N such that n > N =⇒ |xn − x0 | < . Then, the sequence {xn } is

39
said to converge to x0 . Symbolically, this is written as lim xn = x0 or xn −→ x0 as
x→∞
n −→ ∞.

3.0.3 Example. In this example, we prove that the sequence { n1 } converges to 0.

Proof. Given  > 0, we have

1 1 1
| − 0| = <  ⇐⇒ < n
n n 
By the Eudoxus theorem, ∃ N ∈ N such that N > 1 . Hence, n > N ⇒ n > N > 1

1 1 1
⇒ n
< N
<  i.e n > N ⇒ n
< , i.e n > N ⇒ | n1 − 0| = 1
n
< , ∴ lim1
=0
n→∞ n
2n2 +5
3.0.4 Example. Prove that: lim n2 +3n+1 =2
n→∞

Proof. Let  > 0 be given, then

2n2 + 5 −6n + 3 6n + 3 6n + 3 9
| 2
− 2| = | 2 |6 2 6 2
<
n + 3n + 1 n + 3n + 1 n + 3n + 1 n n
Now, by Eudoxus theorem, ∃N ∈ N such that N > 9 . Hence, n > N ⇒ n > N >
9 9 2 2n2 +5

⇒ n
<  ∴ n > N ⇒ | n22n+3n+1
+5
− 2| < 9
n
< . Therefore, lim n2 +3n+1 =2
n→∞

Exercises
1
(1) Prove that lim = 0 if a ∈ R+ .
n→∞ n+a

3n2 +4n+3
(2) Prove that lim 2 = 43 .
n→∞ 4n +8

3.0.5 Definition. A sequence which is not convergent is said to be divergent. Di-

vergent sequences can take four forms namely:

(1) Divergence to ∞.

(2) Divergence to −∞.

(3) finite oscillation.

(4) infinite oscillation.

40
Divergence Sequences

(i) A sequence (xn ) is said to diverge to ∞ iff given any positive real number
k(k ∈ R+ ), ∃ a natural number N depending on k (Nk ∈ N) 3 xn > k ∀
n > N.

3.0.6 Example. If xn := rn , where r > 1, then the sequence diverges to ∞.

To see this set r = 1 + n. Then, x > 0. By Binomial theorem, rn = (1 + x)n =
1 + nx + ... + xn > 1 + nx. Given k ∈ R+ , ∃ n ∈ N 3 nx > k. Hence
by Eudoxus theorem xn > 1 + nx > k i.e (xn ) diverges to ∞ by definition.
Notation: Symbolically, if a sequence (xn ) diverges to ∞, we write lim xn = ∞
n→∞
or xn −→ ∞ as n −→ ∞

(ii) A sequences (un ) is said to diverge to −∞ iff given k ∈ R+ , ∃ Nk ∈ N

3 un < −k ∀ n > N . And we write lim un = −∞.
n→∞

3.0.7 Example. Let un = b − nd where d > 0. Then un diverges to −∞.

Proof. Let k > max{0, b} be given. Then, by Eudoxus theorem, ∃ n ∈ N 3

nd > k + b. Therefore −nd < −k − b i.e b − nd < −k. Here, lim un = −∞.
n→∞

(iii) To define a finitely oscillating sequence, we need the following definitions.

3.0.8 Definition. A sequence (un ) is said to be bounded iff given N ∈ N, ∃

M ∈ R+ 3 |un | 6 M ∀ n > N .

3.0.9 Definition. A sequence (un ) is said to oscillate finitely iff ∃ s, t, M ∈ R

with the following properties: s < t and given any N ∈ N, ∃ n1 , n2 > N 3
un1 < s and t < un2 and addition (un ) is bounded i.e ∃ M ∈ R+ 3 |un | 6 M
∀ n > N.

41
 3.0.10 Example. Suppose that u2n−1 = −1 ∀ n ∈ N, u2n = 1 ∀ n ∈ N.
Hence, (un ) = −1, 1, −1, 1, ..., This sequence oscillate finitely.

−1 1
Proof. Let s = 2
,t = 2
and M = 1. Then, |un | 6 1 ∀ n ∈ N. Also, given
−1
any N ∈ N, then 2N, 2N + 1 > N . Moreover, u2N −1 := −1 < 2
=s<t=
1
2
< 1 = u2N . Hence, un oscillates finitely.

(iv) An unbounded oscillating sequence (un ) is said to oscillate infinitely:

3.0.11 Example. If un := n sin nπ

2
, then the sequence {un } oscillate infinitely.

Proof. Let k ∈ R and N ∈ N be given with k > 0. Then, by Eudoxus

theorem, ∃ N0 ∈ N 3 N0 > max{k, N }. Now, (4N0 + 1) sin 4N20 +1 π = 4N0 + 1
,(4N0 + 3) sin 4N20 +3 π = 4N0 + 3. If we set n1 = 4N0 + 3 and n2 = 4N0 + 1,
then n1 > N0 > N , n2 > N0 > N and un1 = n1 sin n21 π = −n1 < −N0 6 −k,
un2 = n2 sin n22 π = n2 > N0 > k. Hence, (un ) oscillates and is unbounded,
Therefore, (un ) oscillate infinitely.

Exercise

(1) Suppose that un is a sequence converging to −u0 and that a is a non-zero

real number. Prove that aun −→ au0 as n −→ ∞.

(2) Prove that un := n| sin nπ

4
| oscillates infinitely and that un := sin nπ
4
.

3.1 Modulus of Sequences

3.1.1 Definition. A sequence (un ) is said to be

(1) bounded if ∃ M ∈ R+ 3 |un | 6 M ∀ n ∈ N.

42
 (2) eventually bounded off-zero if ∃ N ∈ N and m ∈ R+ 3 m 6 |un | ∀ n > N .

(3) bounded off-zero if N = 1 in (2) above.

(4) absolutely convergent if the sequence (|un |) converges.

Note: A finitely oscillating sequence is a bounded sequence. Thus a bounded se-

quence is not necessarily convergent. However, the next theorem shows that a
convergent sequence is bounded.

3.1.2 Theorem. Let (un ) be a convergent sequence converging to u0 . Then,

(1) (|un |) converges to |u0 |.

(2) (|un |) is bounded.

(3) (un ) is eventually bounded off-zero provided u0 6= 0.

Proof. (1) (un ) converges to u0 . Therefore, given  > 0, ∃ N ∈ N 3 n > N ⇒

||un | − |u0 || < . Therefore, n > N ⇒ |u0 | −  < |un | < |u0 | + . Hence (un )
converges to |u0 |.

(2) Set  = 1. Then |u0 | −  < |un | < |u0 | +  gives n > N ⇒ |un | < |u0 | + 1. Now
each |un |, n = 1, 2, ..., N1 − 1 is finite. Hence, K := max16n6N1 −1 |un | is finite.
M = max{K, 1 + |u0 |}. Then |un | 6 M ∀ n ∈ N. Thus, (un ) is bounded.

(3) Set  := 21 |u0 | in |u0 | −  < |un | < |u0 | + . Then, the first inequality gives
n > N 1 |u0 | ⇒ |u0 | − 12 |u0 | < |un | i.e n > N 1 |u0 | ⇒ 1
|u |
2 0
< |un |. Since
2 2
1
|u |
2 0
> 0, we conclude that (un ) is eventually bounded off-zero.

3.1.3 Remark. Converse of the last theorem is not true. Thus, a sequence may
converge absolutely without being convergent. For example, (un ) = 1, −1, 1, −1, ...
(un ) diverges. However, |un | = 1 ∀ n ∈ N. So, (|un |) converges to 1.

43
3.2 Uniqueness of limits
3.2.1 Theorem. A sequence converges to at most one limit.

0
Proof. Suppose that the sequence (un ) converges to u0 and u0 . Then, given  > 0,
0  0 0
∃N, N ∈ N, n > N ⇒ |un − u0 | < 2
and n > N ⇒ |un − u0 | < 2 . Now take
0 0 0 0
n0 > max{N, N }. Then, |u0 − u0 | = |u0 − un + un − u0 | 6 |un − u0 | + |un − u0 | <
 0 0
2
+ 2 = . Thus, since  is arbitrarily small, then u0 − u0 = 0 ⇒ u0 = u0 ∀ n0 > N .
0 0 0
Reason: u0 6= u0 , then |u0 − u0 | = d > 0. Set  = d2 . Then |u0 − u0 | < d2 . This
0
gives a contradiction for the large. Thus u0 = u0 .

3.3 Arithmetic Operations on Convergent Sequences

3.3.1 Theorem. Let (un ) and (vn ) be convergent sequences, converging to u0 and
v0 respectively. Then we have the following:

(1) lim (un ± vn ) = u0 ± v0 .

n→∞

(2) lim (aun ) = au0 .

n→∞

(3) lim (un vn ) = u0 v0 .

n→∞

u0
(4) lim ( uvnn ) = v0
.
n→∞

(5) u0 6 v0 if ∃N ∈ N 3 un 6 vn ∀ n > N .

(6) For any fixed N ∈ N, lim un+N = lim un

n→∞ n→∞

 
Proof. (1) Given  > 0, ∃N1 , N2 ∈ N 3 |un − u0 | < 2
if n > N1 and |vn − v0 | < 2

∀ n > N2 . Now choose N = max{N1 , N2 }. Then n > N ⇒ |(un + vn ) − (u0 +

 
v0 )| 6 |un − u0 | + |vn − v0 | < 2
+ 2
= . ∴ lim (un + vn ) = u0 + v0 .
n→∞

44
 
(2) Let  > 0 be given. Then, 1+|a|
> 0. Since (un ) is convergent to u0 , ∃N ∈ N 3
 |a|
n > N ⇒ |un − u0 | < 1+|a|
. ∴ n > N ⇒ |aun − au0 | = |a||un − u0 | < 1+|a|
.
Hence, lim aun = au0 .
n→∞

(3) Let  > 0be given. We note that

|un vn − u0 v0 | = |un vn − un v0 + un v0 − u0 v0 |

6 |un vn − un v0 | + |un v0 − u0 v0 |

6 |un ||vn − v0 | + |v0 ||un − u0 |

6 M |vn − v0 | + |v0 ||un − u0 |

since un is convergent,∃ M ∈ R+ 3 |un | 6 M . Then, ∃N1 , N2 ∈ N 3 |un −u0 | <

 
2(M +1)
∀n > N1 and |vn − v0 | < 2(|v0 |+1)
∀n > N2 . Let n > max{N1 , N2 } ⇒
M M  
|un vn − u0 v0 | 6 2(M +1)
 + 2(|v0 |+1)
 = 2
+ 2
= . Hence, lim un vn = u0 v0 .
n−→∞

1 1
(4) We shall show that vn
−→ v0
as n −→ ∞ and then invoke (3) above to
|vn −v0 |
conclude.Therefore, given  > 0, ∃N ∈ N 3 ∀ n > N ,| v1n − 1
v0
| 6 ||vn |v0 |
6
|vn −v0 |
1
|v |2
∀ n > N since for some N, n > N ⇒ 12 |v0 | < |v0 |, given  > 0, ∃N ∗ ∈ N,
2 0
|vn −v0 |
|vn − v0 | < 21 |v0 |2  ∀ n > N ∗ , thus | v1n − 1
v0
| 6 1
|v |2
=  ∀ n > max{N, N ∗ }.
2 0
1 1
∴, lim = .
n−→∞ vn v0

To complete the proof, notice that from (3), we have that if un −→ u0 as

un
n −→ ∞ and vn −→ v0 as n −→ ∞, then un vn −→ u0 v0 . Therefore, vn
=
u0
un . v1n −→ u0 v10 = v0
as n −→ ∞. And this conclude the proof.

(5) As a consequence of (1) and (2), un −→ u0 , vn −→ v0 ⇒ un − vn −→ u0 − v0 .

u0 −v0
Suppose u0 − v0 > 0. Then, given  > 0, with  = 2
,∃ N ∈ N 3 0 <
u0 −v0
2
< un − vn ∀ n > N . This is a contradiction since ∃N 3 un − vn 6 0 ∀
n > N . Hence, u0 − v0 6 0. i.e u0 6 v0 .

45
 6 This follows from the next theorem.

3.3.2 Theorem. If (un ) and (vn ) are two sequences whose terms differ only for
a finite number of values of n, then the limiting behaviour of the two sequences
is the same.

3.3.3 Example. If a sequence (un ) converges to u0 and m ∈ N, then um

n −→

um
0 as n −→ ∞.

Proof. We note that since un −→ u0 as n −→ ∞, then u2n −→ u20 as n −→ ∞

by the product rule for limit of sequences. Assume that for m = k, ukn −→ uk0
as n −→ ∞. Then, uk+1
n = ukn .un −→ uk0 .u0 = uk+1
0 as n −→ ∞. Hence by
induction, ukn −→ uk0 as n −→ ∞.

The general principle of convergence

Here, we consider the general condition for the convergence of sequence in R. It

gives in general the completeness axiom for the real line. We shall need the following
concepts:

(1) Monotone Sequence:

3.3.4 Definition. A sequence (un ) is said to be increasing (respectively de-

creasing) iff un+1 > un ∀ n ∈ N (respectively un+1 6 un n ∈ N). Strictly
increasing and Strictly decreasing sequences are defined analogously. A se-
quence which is either increasing or decreasing is called a monotone sequence.
Strict monotone has its obvious meaning.

3.3.5 Theorem. A bounded monotone sequence is convergent.

(a) Suppose that (un ) is bounded increasing. Then ∃M ∈ R+ 3 |un | 6 M

∀ n ∈ N. Then by the l.u.b theorem, there exists a real number u0 3

46
 sup un = u0 . Now, given  > 0, there exists N ∈ N 3 u0 −  < uN since
n∈N
(un ) is increasing. Therefore, u0 −  < un ∀ n > N . u0 −  < un < u0 <
u0 +  ∀ n > N ⇒ |un − u0 | <  ∀ n > N . Thus lim un = u0 .
n−→∞

(b) If (un ) is decreasing, then (−un ) is increasing. M 6 un . Note that

(−un ) = −k(un ).
1+ 12 +...+ n
1
 
3.3.6 Example. (1) Prove that the sequence given by an = n

is convergent.
1+ 12 +...+ n
1 1+ 12 +...+ n
1 1
+ n+1
Solution: Set σn = n
and σn+1 = n
.

1 + 21 + ... + n1 + n+1 1
1 + 12 + ... + 1
n
σn+1 − σn = −
 n1 1
 n
n 1+ +...+
Then n+1
− 2
n
n

=
n(n + 1)
1 1
− 12 + n+11 1
− 31 + ... + n+1 − n1





− 1 + n+1
n+1
=
n(n + 1)
Each term in the numerator is negative. Therefore, σn+1 − σn < 0.
Hence, the sequence (an ) is decreasing. It is also bounded below by
0. Hence it is convergent.
√
(2) Consider the sequence given by a1 = 1, an+1 = 2 + an , n > 2.
Determine whether (an ) converges.
Solution Clearly, an > 0 ∀ n > 1 since a1 < 2, by induction,
√ √
we obtain that an+1 = 2 + an < 2 + 2 = 2 ∀ n > 1. Since
√
an+1 − an = 2 + an − an > 0. This is true if (2 − an )(1 + an ) > 0.
Since an 6 2, it follows that the sequence is monotone increasing and
bounded. Hence it is convergent.
(3) Discuss the convergence of the sequence given by: an = 2, an+1 =
 
1 2
2
a n + an
n > 2.
Solution: an > 0 ∀ n > 1 since a1 = 2, thus by induction, we have

47
    
1 2 1 2
an+1 = 2
an + an
< 2 ∀ n > 1. Thus an+1 − an = 2
an + an
−
an < 0 ∀ n > 1 since an 6 2. ⇒ the sequence is monotone decreasing
and bounded above by 2. Thus, it is convergent.

Subsequences

3.3.7 Definition. A subsequence of a sequence {un } is a sequence of the form

un1 , un2 , ..., unk , ...

0
where the nks are the natural numbers with

n1 < n2 < ... < nk , ...

3.3.8 Example. Let the sequence (un ) be give and then consider the sequence (un ).
Here, n1 = 4, n2 = 8, n3 = 12, ...., nk = 4k. Hence, (un ) is a subsequence of (un ).

3.3.9 Theorem. If a sequence (un ) converges to u0 , then every subsequence of (un )

converges to u0 .

Proof. The sequence (un ) converges to u0 . Then given  > 0, there exists N ∈ N 3
n > N ⇒ |un − u0 | < . Now, choose k0 3 nk0 > N . Then, nk > nk0 > N ∀ k > k0 .
Therefore ⇒ |unk − u0 | < . Thus, the subsequence (unk ) converges to u0 .

3.3.10 Example. Prove that the sequence (xn ), −1 < x < 1 converges to zero.

Proof. Case 1: 0 6 x < 1. Clearly, the sequence converges since (xn ) is monotone
decreasing and bounded below by zero. Let its limit be a. Then (x2n ) is a subse-
quence of (xn ) also converges to a by the last theorem. But x2n = (xn )2 . Hence
lim (xn )2 = a2 . Hence a2 = a,. Therefore, a = 0 or 1 since (xn ) is a decreasing
n−→∞
sequence with x1 = x < 1, it follows that lim xn = inf(xn ) = 0.
n−→∞

48
 Case 2: −1 < x < 0. Note that (|x|n ) converges to zero by case1. Therefore,
given  > 0, ∃N ∈ N 3 n > N ⇒ |xn − 0| = |xn | < . Thus , lim xn = 0.
n−→∞

Problem

Give an example to show that a subsequence of a sequence may converge without

the sequence converging.

3.4 Cauchy Sequence

3.4.1 Definition. A sequence (un ) is called a Cauchy sequence if given  > 0∃N ∈
N 3 n, m > N ⇒ |un − u0 | < .

(−1)n
3.4.2 Example. Let an = n
. We show that (an ) is Cauchy. To do this, let
m > n. Thus,

(−1)n−1 (−1)m−1 1 1 2
|an − am | = | − |6 + < <
n m n m n

Thus, (an ) is Cauchy.

3.4.3 Theorem. Every convergent sequence is Cauchy.

Proof. Suppose (un ) converges to u0 . Then, given  > 0, ∃N ∈ N 3 n > N ⇒ |un −

u0 | < 2 , therefore |un − um | = |un − u0 + u0 − um | 6 |un − u0 | + |u0 − um | < 2 + 2 = .

3.4.4 Theorem. Every Cauchy sequence is bounded.

Proof. Let (un ) be Cauchy sequence. Then, given  > 0, ∃N ∈ N 3 n, m > N ⇒

|un − um | <  ⇒ − < un − um <  ⇒ − + um < un < um + . Since (un ) is Cauchy,
then take M = max{um + , 1}. Thus, ∃M ∈ R+ such that n > N ⇒ |un | 6 M .
Therefore (un ) is bounded being Cauchy.

49
3.4.5 Definition. n ∈ N is called a peak point of a sequence (un ) iff um < un ∀
m > n.

3.4.6 Theorem. Let (un ) be a sequence. Then (un ) has a subsequence which is
either decreasing or increasing.

Proof. Suppose that the sequence (un ) contains infinitely many peak points, namely;
n1 , n2 , ..., nk , ... Then un1 > un2 > un3 > ... > unk > ... in which case (unk ) is the
decreasing subsequence desired. Now suppose the sequence has only finitely many
peak points namely: m1 , m2 , m3 , ..., mk . Then let n1 be such that n1 > mj , j =
1, 2, 3, ..., k. n1 is not a peak point of (un ). Therefore, ∃n2 3 n2 > n1 and un2 > un1 .
Note that n2 is not a peak point of (un ) and n2 is greater than the peak points.
Therefore, ∃n3 3 n3 > n2 and un3 > un2 . By iteration, we obtain a sequence (unk )
which is increasing.

3.4.7 Corollary. (Bolzano Weierstrass Theorem) Every bounded sequence has a

convergent subsequence.

Proof. Let (un ) be a bounded sequence. Then, it has a monotone subsequence (unk )
is bounded as a subsequence of a bounded sequence. Therefore, (unk ) is convergent
since it is monotone and bounded and this proves the theorem.

3.4.8 Theorem. (General Principle of Convergence) A sequence is convergent iff

it is a Cauchy sequence.

Proof. The first part of this theorem is to assume that if (un ) is a Cauchy sequence.
This has already been achieved in the previous theorem that every convergent se-
quence is Cauchy. Conversely, assume that (un ) is a Cauchy sequence. Then, it is
bounded. Thus, it contains a convergent subsequence (unk )(Bolzano- Weierstrass
theorem). Let unk −→ l. Then, given  > 0, ∃j 3 k > j ⇒ |unk − l| < 2 . Moreover,

50
(un ) is a Cauchy. Therefore, ∃N 3 m, n > N ⇒ |un −um | < 2 . Set k = max{nj , N },
 
choose j0 3 nl > k. Then n > k ⇒ |un − l| 6 |un − unj0 | + |unj0 − l| < 2
+ 2
= .
Therefore, |un − l| as n −→ ∞

Exercise
1 2
(1) If un := n
+ n
+ ... + nn , prove that the sequence diverges to ∞.

1 2 n
(2) If un := n2
+ n2
+ ... + n2
, prove that lim un = 12 .
n−→∞

(3) Deduce that the sum, product and scalar multiple of Cauchy sequences are
Cauchy sequences.

(4) Prove that the sequence (un ) with un > 0 ∀ n ∈ N may be a Cauchy sequence.
Give a sufficient condition for ( u1n ) to be a Cauchy sequence.

51
Chapter 4

Series

4.0.1 Definition. An expression of the form

u1 + u2 + u3 + ... + un + ...
0
where uns are real numbers called a series. un is called the n-th term of the se-
P∞
ries. The series defined above is also written as : un for short. If we set
n=1

S1 = u1

S2 = u1 + u2

S3 = u1 + u2 + u3

Sn = u1 + u2 + ... + un

Then, Sn is called the n-th partial sum of the series above. And (Sn ) is called the
sequence of partial sums of the series.

52
 ∞
P
4.0.2 Definition. We say that the series un converges or has a sum S iff the
n=1
sequence of partial sums converges to S. Otherwise, the series is said to diverge.
In particular, a series can diverge to ±∞, oscillate finitely or infinitely according as
(Sn ), its sequence of partial sums converges to ±∞, oscillate finitely or infinitely.
The limit S to which a series converges is called the sum of the series.

4.0.3 Example. (1) Consider the series

∞
X
a + ar + ar2 + ar3 + ... + arn + ... = arn−1
n=1

We consider the conditions for convergence or divergence in the above.

Case I a 6= 0, r = 1. Then Sn = na.By the Eudoxus theorem, we see that for

a > 0, lim Sn = ∞ and for a < 0, lim Sn = −∞.
n−→∞ n−→∞

Case II a 6= 0, r > 1. Note that the series above is the usual G.P of elementary
a(1−rn )
mathematics. Hence Sn = 1−r
. Therefore, for a > 0, lim Sn =
n→∞
a(1−rn ) a arn

−arn
lim − 1−r
= lim = −∞ since lim = ∞. And for
n→∞ 1−r n→∞ 1−r n→∞ 1−r
n) n
a < 0, lim Sn = lim a(1−r
1−r
= +∞ since lim −ar
1−r
= +∞
n→∞ n→∞ n→∞

∞
1
is convergent and its sum is 21 .
P
(2) The series (n+1)(n+2)
n=1
k ∞
1 1 1 1 1
P P
Reason: Sk = (n+1)(n+2)
= (n+1)
− (n+2)
= 2
− k+2
. Therefore, the
n=1 n=1
∞
1 1
= 12 . Hence, 1
converges to 12 .
P
lim Sk = lim − lim
n−→∞ n−→∞ 2 n−→∞ k+1 n=1
(n+1)(n+2)

∞ ∞
1 1
P P
4.0.4 Remark. In example (2) above, notice that neither n+1
nor n+2
is
n=1 n=1
∞
1
P
convergent. This shall be provided later when we shall show that n
is divergent.
n=1
A series is really a sequence of partial sums in the manner explained above and the
correspondence is total. For thsi reason, all the properties of convergence can be
transferred without loss of generality to series.

53
Arithmetic Operations on Convergent Series
∞
P ∞
P
4.0.5 Theorem. Suppose the series un and the series vn converges respec-
n=1 n=1
tively to u0 and v0 , then
∞
P
(1) the series (un + vn ) converges to u0 + v0 .
n=1

∞
P
(2) the series cun converges to cu0 provided c ∈ R
n=1

k k
P 0 P
Proof. (1) Sk = un , Sk = vn . Then, by hypothesis, lim Sk = u0 and
n=1 n=1 k−→∞
k k k
0 00 P P P 0
lim Sk = v0 . Now, Sk = (un + vn ) = un + vn = Sk + Sk .Hence
k−→∞ n=1 n=1 n=1
00 0
lim Sk = lim Sk + lim Sk = u0 + v0
k−→∞ k−→∞ k−→∞

k k
0 P 0 P 0 0
(2) Now Sk = cun . Then Sk = c un = cSk . Therefore, lim Sk =
n=1 n=1 n−→∞
lim cSk = c lim Sk = cu0
n−→∞ n−→∞

Cauchy Criterion for convergent of Series

∞
P
4.0.6 Theorem. The series un converges iff given  > 0, ∃N ∈ N such that
n=1
m
P
n, m > N ⇒ | uk | < .
k=n+1

Proof. By hypothesis, the series is convergent iff the sequence of partial of sums is
convergent iff given  > 0, ∃N ∈ N 3 m > k > N ⇒ |Sm − Sk | <  iff |Sm − Sk | =
m
P Pk m
P
| un − un | = |uk+1 + uk+2 + ... + um | <  which implies that | un | < 
n=1 n=1 n=k+1

Behaviour of the terms of a Convergent Series

∞
P
4.0.7 Theorem. The series un converges implies that lim un = 0.
n=1 n−→∞

54
 k
P ∞
P
Proof. Sk = un . Suppose that un converges to u0 ∈ R. Then, given  >
n=1 n=1
0, ∃N ∈ N 3 k > N ⇒ |Sk −u0 | < 2 . Hence, uk = |Sk −Sk−1 | = |Sk −u0 +u0 −Sk−1 | 6
 
|Sk − u0 | + |u0 − Sk−1 | < 2
+ 2
=  provided k − 1 > N . Therefore, lim uk = 0.
k−→∞

4.0.8 Remark. The converse of the above theorem is in general false i.e it is not
∞
P
true that lim un ⇒ un is convergent.
n−→∞ n=1
∞
1 1
P
To see this, recall that lim = 0. But we show what follows however that
n−→∞ n n=1
n

called the harmonic series is divergent. This shows by showing that the sequence
of partial sum is not Cauchy and therefore not convergent. Now, S2N − SN =
1
N +1
+ N1+2 + ... + 2N
1
> N
2N
= 1
∀ N ∈ N. Hence, (Sk ) cannot be a Cauchy sequence.
2
∞
1
P
Thus (Sk ) is not convergent, showing that n
is divergent.
n=1

Exercise

Prove the convergence or otherwise of each of the following series and obtain their
limits when they exists.
∞
2n2 +5
P
(1) n2 +3n+1
n=1

∞
1
P
(2) (2n−1)(2n+1)
n=1

∞
1
P
(3) (n+1)(n+2)(n+3)
n=1

Absolute Series of a Series

∞
P
4.0.9 Definition. A series un is said to be absolute iff un > 0 ∀ n ∈ N. The
n=1
P∞ ∞
P ∞
P
absolute series of the series un is the series |un |. And a series un is said
n=1 n=1 n=1
∞
P
to be absolutely convergent iff the series |un | is convergent.
n=1

55
4.0.10 Theorem. An absolutely convergent series is convergent.
∞
P ∞
P
Proof. Suppose that un is an absolutely convergent series. Then |un | is
n=1 n=1
convergent. Let  > 0, be given. Then there exists N ∈ N 3 m > n > N ⇒
|un+1 | + |un+2 | + |un+3 | + ... + |um | <  (by Cauchy Criterion) ⇒ |un+1 + un+2 +
un+3 + ... + um | <  6 |un+1 | + |un+2 | + |un+3 | + ... + |un | <  (by triangle inequality
). Here, by Cauchy criterion, the series is convergent.
∞
P
4.0.11 Theorem. (The Boundedness Criterion) The absolute series |un | of a
n=1
∞
P ∞
P
series un is convergent iff the sequence of partial sums for |un | is bounded.
n=1 n=1

Proof. Mimick that of the absolutely converging sequence and transfer |S|n =
∞
P ∞
P
|un |. Then, (|S|n ), the sequence of partial sums for |un | is increasing. There-
n=1 n=1
fore, (|S|n ) converges iff it is bounded. Therefore, the required results follows from
the definition.
∞
an−1 , for −1 < a < 1. |S|n := 1 + |a| +
P
4.0.12 Example. (1) Again, consider
n=1
n |a|n
|a|2 + |a|n−1 = 1−|a|
1−|a|
1
= 1−|a| − 1−|a|
1
6 1−|a| . 1
Thus, |S|n < 1−|a| ∀ n ∈ N.
∞
an−1 is absolutely convergent and is convergent.
P
Therefore,
n=1

∞
1
P
(2) We show directly that the series n
is divergent. Now,
n=1

z }| { z }| { z }| { z }| {
1 1 1 1 1 1 1 1 1 1
1+ + + + + + + + + + ... +
2 3 4 5 6 7 8 9 10 16

Each of the sums under the trace is greater than 12 . In fact, the following
holds:

56
 1 1
S21 = 1 + >1+
2 2
1 1 1 1 2
S22 = S21 + + > 1 + + > 1 +
3 4 2 2 2
1 1 1 1 2 1 3
S23 = S22 + + + + > 1 + + > 1 +
5 6 7 8 2 2 2
1 1 1 3 1 4
S24 = S23 + + + ... + >1+ + >1+
9 10 16 2 2 2

1
We claim that S2n > 1 + 2n
∀ n ∈ N. To see this, we observe that for n = 1, it
is true. Suppose that it is true for n = k, S2k > 1 + k2 . Then,

1 1 1 1 1 1 2k k 1
S2k+1 = S2k + k+1
+ k+2
+...+ 2k
> S2 k + k+1
+ k+1
+...+ k+1
> S2 k + k+1
> 1+ +
2 2 2 2 2 2 2 2 2
Hence, by induction, the claim is provided. This given M ∈ R+ ∃ by Eudoxus
theorem, N ∈ N 3 N > 2M − 1. Therefore, S2n > (1 + 12 N ) > 1 + 12 (2M − 1) =
1
M+ > N . (Sn ) is a strictly increasing sequence. Therefore, Sn > S2N > M ∀
2
∞
1
n > 2N . Thus, lim Sn = ∞. Hence, the series
P
n
diverges.
n−→∞ n=1
∞
P
4.0.13 Theorem. Let a series un be given. Set
n=1

 U if Un > 0
n
Un+ =
 0 if Un < 0

 0 if Un > 0
Un− =
 U if Un < 0
n
∞ ∞ ∞
u+ u−
P P P
Then the series un is absolutely convergent iff n and n are conver-
n=1 n=1 n=1
gent.
∞
P ∞
P ∞
P
Proof. Let un be absolutely convergent, thus |un | is convergent. Thus, un
n=1 n=1 n=1
is convergent by the fact that an absolutely convergent series is convergent. Now

57
 ∞ ∞
 ∞ ∞

1 1 1
u+ u+
P P P P
n = 2
(un + |un |) . Therefore, n = 2
(un + |un |) = 2
un + |un |
n=1 n=1 n=1 n=1
is convergent being the sum of the two convergence series multiplied by a constant.
∞ ∞ ∞
 ∞
u− 1
u+
P P P P
Similarly, n = 2
un − |un | is convergent. Now suppose that n
n=1 n=1 n=1 n=1 
∞ ∞ ∞ ∞
 ∞
u− (u+ −
u+ u−
P P P P P
and n are convergent. Then, |un | = n − un ) = n + − n
n=1 n=1 n=1 n=1 n=1
converges being the sum of two convergent series.

Conditional Convergence of Series

4.0.14 Definition. A series is said to converge conditional if its convergent but

not absolutely convergent. The existence of conditionally convergent series shall be
proved in what follows and we establish the converse of the theorem which states
that every absolutely convergent series is convergent is in general, false. First, we
establish the theorem called Leibniz theorem.

4.0.15 Theorem. Suppose that (un ) is a decreasing sequence in [0, ∞). Suppose
∞
(−1)n−1 un converges.
P
further that lim un = 0. Then the series
n−→∞ n=1
∞
(−1)n−1 un = u1 − u2 + u3 − ... + (−1)n+1 un . Then, S2n =
P
Proof. Note that
n=1
(u1 − u2 ) + (u3 − u4 ) + ... + (u2n−1 − u2n ) The content of each of the brackets
is positive. Hence, the first equality shows that (S2n ) is an increasing sequence
and the second equality shows that |S2n | 6 u1 ∀ n ∈ N. Therefore, the sequence
(S2n ) is convergent to l say. Now, S2n+1 = S2n + u2n+1 . Therefore, lim S2n+1 =
n−→∞
lim S2n + u2n+1 = l + 0 = l. Set
n−→∞

 S if n is odd
n
wn =
 l if n is even

 S if n is even
n
vn =
 l if n is odd

58
 Then, wn −→ l as n −→ ∞ and vn −→ l as n −→ ∞. But Sn = wn + vn − l.
∴ lim Sn = l + l − l = l i.e (Sn ) is convergent to l.
n−→∞

∞
P (−1)n+1
4.0.16 Example. Consider the series n
. We observe that the sequence
n=1
∞
1
P (−1)n+1
(un ) is decreasing, and lim = 0. Therefore, converges conditionally
n−→∞ n n=1
n

by Leibniz rule. However, the series is not absolutely convergent.

Comparison Test
∞ ∞
vn be two given series. Suppose that k ∈ R+
P P
4.0.17 Theorem. Let un and
n=1 n=1
and N ∈ N 3 |un | 6 k|vn | ∀ n > N
∞
P
(1) Then the absolute convergence of vn implies the absolute convergence of
n=1
∞
P
un .
n=1

∞
P ∞
P
(2) The absolute divergence of un i.e the divergence of |un | implies the
n=1 n=1
∞
P
absolute divergence of vn .
n=1

∞
P ∞
P
Proof. (1) Consider now |vn | converges implies kvn converges to l say.
n=1 n=1
∞
P ∞
P
Therefore, Sm = |un | 6 |u2 | + ... + |uN −1 | + k|vn | = |u1 | + |u2 | + ... +
n=1 n=1
|uN −1 | + l. For m ∈ N. Therefore, (Sm ) is a bounded sequence and so by
∞
P
boundedness criterion, |un | is convergent and hence the result.
n=1

(2) This is the contrapositive of (1). Therefore, it follows from the truth of (1)

4.0.18 Example. Test the following series for convergence.

∞
1
P
(1) nn
.
n=1

59
 ∞
1
P
(2) np
, p is a rational number with 0 < p 6 1.
n=1

∞
P 4+(−1)n
(3) 3n +n3
.
n=1

Solution
∞
1 1 1
(1) n2 6 nn ∀ n ∈ N. Therefore, 0 <
P
nn
< n2
∀ n ∈ N. By n2
converges.
n=1
∞
1
P
Therefore, by comparison test, nn
is also convergent.
n=1

∞
1
with 0 < p 6 1. Now, note that 0 < np 6 n.
P
(2) Consider the p series np
n=1
∞
1 1 1
P
Therefore, 0 < n
< np
. But n
.
n=1

∞ ∞
4+(−1)n 5
P 5
P 1
(3) 0 6 3n +n3
6 3n
. Now, 3n
= 5 3n
is convergent. Hence, by the
n=1 n=1
∞
P 4+(−1)n
comparison test, 3n +n3
is convergent.
n=1

Variants of Comaprison Test-Root Test

∞
P
4.0.19 Theorem. (Root Test) Suppose that un is given series with lim1 = r,
n=1 |un | n
∞
P ∞
P
then, r < 1 implies that un converges absolutely and r > 1 implies that un
n=1 n=1
diverges absolutely.

Proof. Suppose that 0 6 r < 1. Then we can choose s 3 r < s < 1. Now
1 1
lim (un ) n = r. Set  : s − r, then, ∃N0 ∈ N, ||un | n − r| < s − r ∀ n > N . Therefore,
n−→∞
0 n
(S ) 6 |un | ∀ n > N0 . It is now clear from comparison with the geometric series
∞ ∞
0 0
(S )n , S > 1, that
P P
un is divergent absolutely.
n=1 n=1

∞
P
4.0.20 Theorem. (Ration Test) Let un be a series. Suppose that there exists
n=1

60
 ∞
r ∈ (0, 1) and N ∈ N 3 |un+1 | 6 r|un | ∀ n ∈ N or lim | uun+1
P
n
|. Then, un
n−→∞ n=1
converges absolutely.

Proof. Suppose that |un+1 | 6 r|un | holds. Then, there exists N0 ∈ N 3 || uun+1
n
−r|| <
. Thus, |un+1 | 6 (r + )|un | ∀ n > N0 . Thus, by comparison test lim | uun+1
n
|
n−→∞
holds. Now suppose r < 1. Then we can choose a real number s with r < s < 1.
Then, s − r > 0. Set  = s − r. Then, lim | uun+1
n
| becomes: |un+1 | 6 s|un | ∀
n−→∞
n > N0 . Therefore, |uN0 +1 | 6 s|uN0 | and |uN0 +2 | 6 s2 |uN0 |. By iteration, |uN0 +k | 6
∞
sk |uN0 | ∀ k > N . The geometry series S k being convergent, its constant multiple
P
n=1
∞ ∞ ∞
k
P P P
S |uN0 |is also convergent. Hence, by comparison test, |un | = |uN0 +1 | is
n=1 n=N0 k=0
∞
P
convergent and therefore, |un | is convergent.
n=1

∞
nm
P
4.0.21 Example. Discuss the convergence of n!
,m ∈ N.
n=1
 
nm (n+1)m un+1 (n+1)m n!

(n+1)m 1
Proof. Let un = n!
and un+1 = (n+1)!
. Then, un
= (n+1)! nm
= n
. n+1 .
m
Hence, | (n+1)
n
1
. n+1 | = r. Therefore, lim un+1
< 1. Thus convergence by ratio test.
n→∞ un

∞
P
4.0.22 Theorem. Let the series un be given, Suppose that x ∈ R, with r > 1 3
n=1
for some N ∈ N, we have |un+1 | > r|un | ∀ n ∈ N or lim | uun+1
n
| = r > 1, then the
n−→∞
P∞
|un | diverges.
n=1

Implication of the Ratio test

∞
un is a series such that lim | uun+1
P
If n
| = l exists. Then.
n=1 n−→∞

∞
P
(i) If l < 1, then un converges absolutely.
n=1

∞
P
(ii) If l > 1, then un diverges absolutely.
n=1

61
 ∞
P
(iii) If l = 1, then the convergence un is inconclusive.
n=1

Convergent of Re-arrangement of an absolutely Convergent

Series
∞
P
4.0.23 Definition. Let un be a series and f : N −→ N be a bijective mapping.
n=1
∞
P ∞
P
Then uf (x) is called the re-arrangement series of un . The following theorem
n=1 n=1
is pertinent.
∞
P
4.0.24 Theorem. A series un is absolutely convergent iff every re-arrangement
n=1
∞
P ∞
P ∞
P ∞
P
series uf (x) of un converges absolutely with uf (x) = un .
n=1 n=1 n=1 n=1

∞
P ∞
P ∞
P
Proof. Suppose that un is absolutely convergent, then both |un | and un
n=1 n=1 n=1
∞ n n
P P 0 P
are convergent with un = Sn . Set Sn = uk and Sk = uf (x) . Then,
n=1 k=1 k=1

given  > 0, ∃N ∈ N 3 n, m > N ⇒ ||S|n − |S|m | < 2
⇒ |S0 − Sn | < 2 . Now,
choose k0 3 {1, 2, 3, ..., N } ⊂ {f (1), f (2), ..., f (k0 )}. In fact, we may set k0 =
0 0 0
max{f (1)−1 , f (2)−1 , ..., f (N )−1 }. Then, ||S |k − |S |N | < 2 k > k0 and |S0 − Sk | 6
∞
0
|S0 − SN | + |Sk0 − SN | = 2 + 2 ∀ k > k0 . Thus,
P
uf (x) converges to S0 . The
n=1
converse also follows immediately .

4.0.25 Example. (1) Discuss the convergence or divergence of the following se-
ries.
∞
1
P
(i) 2n
.
n=1
∞
2
P
(ii) n2 +3
n=1

(2) Determine whether or not the following series converges or diverges.

62
 ∞
P (n!)2
(i) (2n)!
.
n=1
P∞
1
√ √
(ii) n
{ n+1− n}.
n=1
∞  n2
n
P
(iii) (n+1)
.
n=1

(3) Find the limit of the following series to determine whether they converge or
diverge.
∞
2n +3n
P
(i) 6n
.
n=1
∞
n29
P
(ii) (n+1)!
n=1

Solution:
∞ ∞ ∞ ∞ ∞
1 1 1 1 1 1
P P P P P
(1) (i). 2n
can be written as 2n
= 2 n
. Since n
diverges , then 2n
n=1 n=1 n=1 n=1 n=1
diverges.
∞ ∞ ∞ ∞
2 2 2 2
P P P P
(ii). 2
n +3
. By comparison test, n2 +3
6 n2
converges, then n2 +3
n=1 n=1 n=1 n=1
converges.
∞
P (n!)2 (n!)2 ((n+1)!)2 an+1 ((n+1)!)2 (2n)!
(2) (i). (2n)!
. Let an = (2n)!
and an+1 = (2(n+1))!
. Then an
= .
(2(n+1))! (n!)2
=
n=1
∞
(n+1)2 n +2n+1 2 an+1 1
P (n!)2
(2n+1)(2n+2)
, then it implies | 4n 2 +4n+1 | and lim | a
n
|= 4
< 1. Thus, (2n)!
n−→∞ n=1
is convergent by the ratio test.
∞
P 1
√ √ ∞
P √
n+1 P∞ √
n
∞
P n+1
∞
P n
(ii). n
{ n + 1− n} its equal to n
− n
= √
n n+1
− √
n n
=
n=1 n=1 n=1 n=1 n=1
∞ ∞
n+1 n n+1 n
P P
3 2
1 − 3
1 . Since lim 1 < 1 and lim 1 < 1. So
n=1 (n +n ) 2 n=1 (n ) 2 n−→∞ (n3 +n2 ) 2 n−→∞ (n3 ) 2
P∞
n+1
∞
P n
P∞
1
√ √
therefore since √
n n+1
− √
n n
is convergent, then n
{ n + 1 − n}
n=1 n=1 n=1
convergent.
∞  n2 ∞ ∞ ∞ ∞
n n n 1 1
P P P P P
(iii). (n+1)
6 ( (n+1)
6 (n)2
6 n
. Since n
diverges, then
n=1 n=1 n=1 n=1 n=1
∞  n2
n
P
(n+1)
diverges by comparison.
n=1

63
 k k k
 k k

2n +3n 1 n 1 n 1 n 1 n
= 72 .
P P
P
P
P

(3) (i). 6n
= 3
+ 2
. Then, lim 3
+ 2
n=1 n=1 n=1 k−→∞ n=1 n=1
∞
2n +3n 7
P
⇒ 6n
converges to 2
.
n=1
∞
P n29 n29 (n+1)29 un+1 (n+1)29 (n+1)!
(ii). (n+1)!
. Let un = (n+1)!
and un+1 = (n+2)!
. Then, un
= (n+2)!
. n29 =
n=1
 29
(n+1)
n
1
. n+2 . Hence, lim | uun+1
n
| = 0 < 1. Therefore, it implies that
n−→∞
∞
P n29
(n+1)!
is convergent.
n=1

64