MT (N) 201

REAL ANALYSIS MT(N) - 201
BACHELOR OF SCIENCE/ BACHELOR OF ARTS

(New Education Policy-2020)
REAL ANALYSIS
MT(N)-201
RR
DEPARTMENT OF MATHEMATICS
SCHOOL OF SCIENCES
UTTARAKHAND OPEN UNIVERSITY
HALDWANI, UTTARAKHAND
263139
COURSE NAME: REAL ANALYSIS
COURSE CODE: MT(N) 201
Department of Mathematics
School of Science
Uttarakhand Open University
Haldwani, Uttarakhand, India,
263139
Real Analysis MT(N) 201
BOARD OF STUDIES - 2023

Chairman
Prof. O.P.S. Negi
Honorable Vice Chancellor
Prof. P. D. Pant Prof. Harish Chandra Prof. Manoj Kumar

Director Senior Professor Professor and Head
School of Sciences Department of Mathematics Department of Mathematics,
Uttarakhand Open University Institute of Science Statistics and Computer Science
Haldwani, Uttarakhand Banaras Hindu University G.B. Pant University of
Varanasi Agriculture & Technology,
Pantnagar
Prof. Sanjay Kumar Dr. Arvind Bhatt Dr. Jyoti Rani

Professor Programme Cordinator Assistant Professor
Department of Mathematics Associate Professor Department of Mathematics
DeenDayalUpadhyaya College Department of Mathematics Uttarakhand Open University
University of Delhi Uttarakhand Open University Haldwani, Uttarakhand
New Delhi Haldwani, Uttarakhand
Dr. Kamlesh Bisht Dr. Shivangi Upadhyay

Assistant Professor(AC) Assistant Professor (AC)
Department of Mathematics Department of Mathematics
Uttarakhand Open University Uttarakhand Open University
Haldwani, Uttarakhand Haldwani, Uttarakhand
Editor
Dr. Arvind Bhatt
Associate Professor
Haldwani, Uttarakhand
Unit Writer Blocks Units
Dr. Deepak Kumar Sharma I, II and III 1 to 14

Assistant Professor
CONTENTS
MT(N) - 201
BLOCK I: REAL NUMBERS, SEQUENCE Page Number 01 - 85

AND SERIES
Unit – 1 Sets and functions 02-18
Unit – 2 Real Numbers 19 - 32
Unit - 3 Limit points, open and 33 - 48

closed sets
Unit - 4 Real sequences 49 - 67
Unit - 5 Infinite Series 68 - 85
BLOCK II: FUNCTIONS SINGLE VARIABLE Page Number 86 - 140
Unit – 6 Limit 87 - 109
Unit - 7 Continuous Function 110 - 122
Unit - 8 Differentiation 123 - 140
BLOCK III: RIEMANN INTEGRAL AND IMPROPER Page Number 141 - 229
INTEGRAL
Unit – 9 Riemann integral I 142 - 155
Unit - 10 Riemann integral II 156 - 177
Unit - 11 Sequence and series of 178 - 191

function
Unit - 12 Improper integral I 192 - 204
Unit - 13 Improper integral II 205 - 218

Unit - 14 Dirichlet and Abel’s tests 219 - 229
for improper integrals
COURSE INFORMATION
The present self-learning material “Real Analysis” has been
designed for B.Sc. (Third Semester) learners of Uttarakhand Open
University, Haldwani. This course is divided into 14 units of study. This
Self Learning Material is a mixture of Four Block.
First block is Real Numbers, Sequence and Series, in this block Sets
and functions, Real Numbers, Limit points, open and closed sets, Real
sequences, Infinite Series defined Clearly.
Second block is Functions Single Variable, in this block Limits of

function, continuous function, Properties of Continuous function, Uniform
Continuity, Monotone and Inverse function. Derivative, Mean Value
theorem, L Hospital rule defined clearly.
Third block is Riemann Integration, Uniform convergence and
Improper integral, in this block Riemann integral, Integrability of
continuous and monotonic functions, Fundamental theorem of integral
calculus, First mean value theorem, Pointwise and uniform convergence of
sequence and series of functions, Weierstrass’s M-test, Dirichlet test and
Abel’s test for uniform convergence, Uniform convergence and continuity,
Uniform convergence and differentiability, Improper integrals, Dirichlet and
Abel’s tests for improper integrals are defined.
Adequate number of illustrative examples and exercises have also been
included to enable the leaners to grasp the subject easily.
Real Analysis MT(N) - 201
Course Name: REAL ANALYSIS

Course Code: MT(N) 201
BLOCK-I
Uttarakhand Open University 1
UNIT 1: SETS AND FUNCTIONS
CONTENTS:
1.1 Introduction
1.2 Objectives
1.3 Sets
1.4 Methods of describing a set
1.5 Types of sets
1.6 Subset, Superset and Power set
1.7 Operations on a set
1.8 De Morgan’s Laws
1.9 Cartesian Product of two sets
1.10 Functions or Mappings
1.11 Kinds of Functions
1.12 Inverse Function
1.13 Composite of Functions
1.14 Summary
1.15 Glossary
1.16 References
1.17 Suggested Reading
1.18 Terminal questions
1.19 Answers
1.1 INTRODUCTION
Set theory, branch of mathematics that deals with the properties

of well-defined collections of objects, which may or may not be of a
mathematical nature, such as numbers or functions. The theory is less
valuable in direct application to ordinary experience than as a basis for
precise and adaptable terminology for the definition of complex and
sophisticated mathematical concepts.
Between the years 1874 and 1897, the German mathematician and
logician Georg Cantor created a theory of abstract sets of entities and
made it into a mathematical discipline. This theory grew out of his
investigations of some concrete problems regarding certain types
of infinite sets of real numbers. A set, wrote Cantor, is a collection of
definite, distinguishable objects of perception or thought conceived as a
whole. The objects are called elements or members of the set.
1.2 OBJECTIVES
After studying this unit, learner will be able to
i. To analyze and predict the behavior of these systems over time.

ii. To provide solutions to problems that cannot be solved using other
mathematical techniques.
iii. To understand the definition of differential equation.
1.3 SETS
A set is a well - defined collection of distinct objects.

By a ‘well – defined’ collection of objects we mean that there is a rule by
means of which it is possible to say, without ambiguity, whether a
particular object belongs to the collection or not. The objects in a set are
‘distinct’ means we do not repeat an object over and over again in a set.
Each object belonging to a set is called an element of the set. Sets are
usually denoted by capital letters A, B, N, Q, S etc. and the elements by
lower case letters a, b, c, x etc.
The symbol ∈ is used to indicate ‘belongs to’. Thus x ∈ A ⟹ x is an

element of the set A.
The symbol ∉ is used to indicate ‘does not belong to’. Thus x ∉ A ⟹ x is

not an element of the set A.
Example: Let A = {1, 2, 3, 4, 5} be a set then we say 1∈ 𝐴 , 2∈ A, 3∈ 𝐴,

4∈ 𝐴, 5∈ 𝐴 but 6∉ A, 7∉ A, 8∉ A.
1.4 METHODS OF DESCRIBING A SET
There are two methods of describing a set.

(1) Roster Method.
In this method, a set is described by listing all its element, separating by
commas and enclosing within curly brackets.
For Example. (i) If A is the set of odd natural numbers less than 10,
then in roster form.
A = {1, 3, 5, 7, 9}
(ii) if B is the set of letters of the world FOLLOW, then in roster form.
B = {F, O, L, W}
(2) Set Builder Method.
Listing the element of a set is sometimes difficult and sometimes
impossible. We do not have a roster form of the set or rational number or
the set of real numbers. In set builder method, a set is described by
means of some property which is shared by all the element of the set.
For Example. (i) If P is the set of all prime numbers, then
P = {x : x is a prime number}
(ii) if A is the set of all natural numbers between 5 and 50, then
A = {x : x ∈ 𝑁 and 5 < x < 50}
1.5 TYPES OF SETS
(i) Finite set. A set is said to be finite if the number of its elements is
Finite i.e. its elements can be counted, by one by one, with counting
coming to end.
For Example. (a) the set of letters in the English alphabet is finite set
since it has 26 elements.
(b) Set of all multiples of 10 less than 10000 is a finite set.
(ii) Infinite set. A set is said to be infinite if the number of its elements
is infinite i.e. we count its elements, one by one, the counting never
comes to an end.
For Example. (a) the set of all points in a straight line is an infinite set.
(b) the sets ℕ, ℤ, ℚ, ℝ all are infinite sets.
(iii) Null Set. A set having no element is known as a null set or void set
or an empty set and is denoted by ∅ or {}.
For Example. (a) {x : x is an integer and 𝑥 2 = 3} = ∅, because there is
no integer whose square is 3.
(iv) Singleton Set. A set having only one element is called a singleton
set.
For Example. (a) {a} is a singleton set.
(b) {x: 𝑥 3 + 1 = 0 and x∈ ℝ} = { -1 } is a singleton set.
1.6 SUBSET, SUPERSET AND POWER SET
Set A is said to be a subset of Set B if all the elements of Set A are also
present in Set B. In other words, set A is contained inside Set B.
Example: If set A has {X, Y} and set B has {X, Y, Z}, then A is the
subset of B because elements of A are also present in set B.
Subset Symbol
In set theory, a subset is denoted by the symbol ⊆ and read as ‘is a

subset of’.
Using this symbol we can express subsets as follows:
A ⊆ B; which means Set A is a subset of Set B.
Note: A subset can be equal to the set. That is, a subset can contain all
the elements that are present in the set.
All Subsets of a Set
The subsets of any set consists of all possible sets including its elements
and the null set. Let us understand with the help of an example.
Example: Find all the subsets of set A = {1,2,3,4}
Solution: Given, A = {1,2,3,4}
Subsets are {},{1}, {2}, {3}, {4}, {1,2}, {1,3}, {1,4}, {2,3}, {2,4},
{3,4},{1,2,3}, {2,3,4}, {1,3,4}, {1,2,4},{1,2,3,4}.
Superset Definition
In set theory, set A is considered as the superset of B, if all the elements

of set B are the elements of set A. For example, if set A = {1, 2, 3, 4}
and set B = {1, 3, 4}, we can say that set A is the superset of B. As the
elements of B [(i.e.,)1, 3, 4] are in set A. We can also say that B is not a
superset of A.
Superset Symbol
The superset relationship is represented using the symbol “⊃”. For

instance, the set A is the superset of set B, and it is symbolically
represented by A ⊃ B.
Consider another example,
X = {set of polygons}, Y = {set of irregular polygons}
Then X is the superset of Y (X⊃Y). In other words, we can say that Y is

a subset of X (Y⊂X).
Proper Superset
The proper superset is also known as a strict superset. The set B is the
proper superset of set A, then all the elements of set A are in B, but set B
must contain at least one element which is not present in set A.
For example, let us take four sets.
A = {a, b, c}, B = {a, b, c, d}, C = (a, b, c}, D = {a, b, e}
From the sets given above,
B is the proper superset of A, as B is not equal to A
C is a superset of set A, but the set C is not a proper superset of set A, as

C=A
D is not a superset of A, as the set D does not contain the element “c”
which is present in set A.
Power Set
The set of all subsets of a set A is called the power set of A and denoted
by P(A).
i.e. P(A) = {S: S ⊂ 𝐴 }.
For Example. (i) if A = {a}, then P(A) = {∅, A}
(ii) If B = {1, 2} then P{B} = {∅,{1}, {2}, B}
Theorem 1. Every set a subset of itself.
Proof. Let A is any set. Since x ∈ A ⟹ x ∈ A, therefore A ⊂ 𝐴.
Theorem 2. Empty set is a subset of every set.
Proof. Given two sets A and B, let A=∅.

By definition, A is a subset of B if and only if every element in A is also
in B.
This means that A would not be a subset of B if there exists an element
in A that is not in B.
However, there are no elements in A. This means there cannot exist an

element in A that is not in B. Thus, A is a subset of B.
Since A = ∅ and B is an arbitrary set, the ∅ must be a subset of all sets.
Theorem 3. The empty set is unique.

Proof. Let ∅1 𝑎𝑛𝑑 ∅2 be two empty sets.
Since empty set is a subset of every set .
Therefore ∅1 ⊂ ∅2 and ∅2 ⊂ ∅1
⟹ ∅1 = ∅2 that proves the uniqueness of ∅.
Note: if a set has n elements, then the number of subsets is 2𝑛 .
1.7 OPERATIONS ON A SETS
1. Union of Sets. The union of two sets X and Y is equal to the set of
elements that are present in set X, in set Y, or in both the sets X and Y.
This operation can be represented as;
X ∪ Y = {a: a ∈ X or a ∈ Y}
Let us consider an example, say; set A = {1, 3, 5} and set B = {1, 2, 4}
Then A ∪ B = {1, 2, 3, 4, 5}
Properties of Union of Sets
(i) For any two Sets A and B, A ⊂ 𝐴 ∪ 𝐵 or B ⊂ 𝐴 ∪ 𝐵
Proof. Let x be any element of A. then
x∈A ⟹x∈𝐴∪𝐵
therefore A⊂ 𝐴∪𝐵
similarly, we can prove B ⊂ 𝐴 ∪ 𝐵
(ii) For any set A, 𝐴 ∪ ∅ = A.
Proof. 𝐴 ∪ ∅ = {x: x ∈ 𝐴 or x ∈ ∅ }
= {x: x ∈ 𝐴 } [∵ ∅ has no element]
=A
(iii) Union of sets is idempotent i.e. foe any set A, A∪ 𝐴 = A.
Proof. A ∪ 𝐴 = {x: x ∈ 𝐴 or x ∈ A}
= {x: x ∈ 𝐴 }
=A
(iv) Union of sets is commutative.
Proof. A ∪ 𝐵 = {x: x ∈ 𝐴 or x ∈ B}
= {x: x ∈ B or x ∈ 𝐴 }
=B∪𝐴
Note: Union of sets is Associative.
2. Intersection of Sets. The intersection of two sets X and Y is the set of

all elements which belong to both X and Y. This operation can be
represented as;
X ∩ Y = {a: a ∈ X and a ∈ Y}
Let us consider an example, say; set A = {1, 3, 5} and set B = {1, 2, 4}
Then A ∩ B = {1}
Properties of Intersection of Sets
(i) For any two sets A and B, A ∩ B ⊂ 𝐴 and A ∩ B ⊂ 𝐵.
Proof. Let x be any element of A ∩ B. then
x ∈ A ∩ B ⟹ x ∈ A and x ∈ B
⟹ x ∈ A (in particular)
Therefore A∩B⊂𝐴
Similarly, we can prove A ∩ B ⊂ 𝐵.
(ii) Intersection of sets is idempotent i.e. foe any set A, A ∩ 𝐴 = A.

Proof. A ∩ 𝐴 = {x: x ∈ 𝐴 and x ∈ A}
= {x: x ∈ 𝐴 }
=A
(iii) Intersection of sets is commutative.
Proof. A ∩ 𝐵 = {x: x ∈ 𝐴 and x ∈ B}
= {x: x ∈ B and x ∈ 𝐴 }
=B∩𝐴
Note: Intersection of sets is Associative.
3. Difference of Sets. The difference of two sets A and B is the set of all
elements which are in A but not in B.
The difference of sets A and B is denoted by A – B.
i.e. A – B = {x: x ∈ 𝐴 and x ∉ B}
For example. (i) if A = {1, 2, 3, 4, 5} and B {2, 4, 6, 8}, then A – B =
{1, 3, 5}, B – A = {6, 8}.
Clearly, A–B≠B-A
Note. The difference of sets is not commutative.
4. Complement of a Set. Let U be the universal set and A ⊂ U. then

complement of A is the set of those elements of U which are not in A.
the complement of A is denoted by 𝐴𝑐 .
Symbolically, 𝐴𝑐 = U – A = {x: x ∈ U and x ∉ A} = {x: x ∉ A}
For example. If U is the set of all natural numbers and A is the set of
even natural numbers, then
𝐴𝑐 = U – A
= the set of those natural numbers which are not even
= the set of odd natural numbers.
5. Symmetric Difference of Sets. If A and B are any two sets, then the
sets (A – B) ∪ (B − A) is called the symmetric difference of A and B.
The symmetric difference of A and B is denoted by A ∆ B and read as ‘A
symmetric difference B’.
For Example. If A = {a, b, c, d, e} and B = {c, d, e, f, g}, then
A – B = {a, b}, B – A = {f, g}
Therefore A ∆ B = (A – B) ∪ (B − A)
= {a, b} ∪ {f, g} = {a, b, f, g}.
1.8 DE MORGAN’S LAWS
For any two sets A and B, prove that
(a) (𝐀 ∪ 𝐁)𝐜 = 𝐀𝐜 ∩ 𝐁 𝐜 (b) (𝐀 ∩ 𝐁)𝐜 = 𝐀𝐜 ∪ 𝐁 𝐜
Proof. (a) We need to prove, (A ∪ B)c = Ac ∩ B c
Let X = (A ∪ B)c and Y = Ac ∩ B c
Let p be any element of X, then p ∈ X ⇒ p ∈ (A ∪ B)c

⇒ p ∉ (A ∪ B)
⇒ p ∉ A or p ∉ B
⇒ p ∈ A’ and p ∈ B’
⇒ p ∈ A’ ∩ B’
⇒p∈Y
∴X⊂Y . . . (i)
Again, let q be any element of Y, then q ∈ Y ⇒ q ∈ A’ ∩ B’
⇒ q ∈ Ac and q ∈ B c
⇒ q ∉ A or q ∉ B
⇒ q ∉ (A ∪ B)
⇒ q ∈ (A ∪ B )c
⇒q∈X
∴Y⊂X . . . (ii)
From (i) and (ii) X = Y
(A ∪ B)c = Ac ∩ B c
(b) We need to prove, (A ∩ B)c = Ac ∪ B c
Let X = (A ∩ B)c and Y = Ac ∪ B c
Let p be any element of X, then p ∈ X ⇒ p ∈ (A ∩ B)c
⇒ p ∉ (A ∩ B)
⇒ p ∉ A and p ∉ B
⇒ p ∈ Ac or p ∈ B c
⇒ p ∈ Ac ∪ B c⇒ p ∈ Y
∴ X ⊂ Y ————–(i)
Again, let q be any element of Y, then q ∈ Y ⇒ q ∈ Ac ∪ B c
⇒ q ∈ Ac or q ∈ B c
⇒ q ∉ A and q ∉ B
⇒ q ∉ (A ∩ B)
⇒ q ∈ (A ∩ B )c
⇒q∈X
∴ Y ⊂ X ————–(ii)
From (i) and (ii) X = Y
(A ∩ B)c = Ac ∪ B c
1.9 CARTESIAN PRODUCT OF TWO SETS
Given two non-empty sets A and B. The Cartesian product A × B is the

set of all ordered pairs of elements from A and B,
i.e., A × B = {(p, q): p ∈ A, q ∈ B}
If either P or Q is the null set, then A × B will also be an empty set,
i.e., A × B = φ
For Example: if A = {1, 2} and B = {3, 4, 5}, then the Cartesian

Product of A and B is A × B = {(1, 3), (1, 4), (1, 5), (2, 3), (2, 4), (2, 5)}.
Cardinality of Cartesian Product?

The cardinality of Cartesian products of sets A and B will be the total
number of ordered pairs in the A × B.
Let p be the number of elements of A and q be the number of elements in
B.
So, the number of elements in the Cartesian product of A and B is pq.
i.e. if n(A) = p, n(B) = q, then n(A × B) = pq.
1.10 FUNCTIONS OR MAPPINGS
A function can be visualized as an input/output device.
Let A & B be any two non-empty sets. If there exists a rule ‘f ’ which
associates to every element x ∈ A, a unique element y ∈ B, then such
rule ‘f ’ is called a function or mapping from the A to the set A to the
set B.
We write f: A ⟶ 𝐵 read ‘f ’ is a function from X to Y.
The set A is called the domain of f and the set B is called the Co-domain
of f.
Range of f = f(A) = {f(x): x ∈ A} , clearly f(A) ⊂ B.
1.11 KINDS OF FUNCTIONS
(1) Equal Functions. Let A and B be sets and f: A ⟶ 𝐵 and f: B ⟶

𝐴 be functions. We say that f and g are equal and write f = g if f(a) =
g(b) for all a ∈ A. If f and g are not equal, we write f ≠ 𝑔.
(2) One – One Function (Injective Function). A function f is one-

to-one if every element of the range of g corresponds to exactly one
element of the domain of f. One-to-one is also written as 1-1.
Formally, it is stated as, if f(x) = f(y) implies x=y, then f is one-to-one
mapped, or f is 1-1.
Example. Show that f: R→ R defined as f(a) = 3a3 – 4 is one to one

function?
Solution: Let f ( a1 ) = f ( a2 ) for all a1 , a2 ∈ R
so 3a13 – 4 = 3a23 – 4
a13 = a23
a13 – a23 = 0
(a1 – a2) (a1 + a1a2 + a22) = 0
a1 = a2 and (a12 + a1a2 + a22) = 0
(a12 + a1a2 + a22) = 0 is not considered because there are no real values of
a1 and a2.
Therefore, the given function f is one-one.
(3) Onto Function (Surjective Function). Onto function could be

explained by considering two sets, Set A and Set B, which consist of
elements. If for every element of B, there is at least one or more than one
element matching with A, then the function is said to be onto
function or surjective function.
Note: To show that a function f is an onto function, put y = f(x), and

show that we can express x in terms of y for any y ∈ B.
Example 1. Let A = {1, 5, 8, 9) and B {2, 4} And f ={(1, 2), (5, 4), (8,
2), (9, 4)}. Then prove f is a onto function.
Solution: From the question itself we get,
A={1, 5, 8, 9), B = {2, 4} & f={(1, 2), (5, 4), (8, 2), (9, 4)}
So, all the element on B has a domain element on A or we can say

element 1 and 8 & 5 and 9 has same range 2 & 4 respectively.
Therefore, f: A → B is a surjective function.
Example 2. How to tell if this function is an onto function? g: R → R

defined by g(x) = 1 + x2
Solution: Given the function g(x) = 1 + x2.
For real numbers, we know that x2 > 0. So 1 + x2 > 1. g(x) > 1 and hence
the range of the function is (1, ∞). Whereas, the second set is R (Real
Numbers). So the range is not equal to codomain and hence the function
is not onto.
Example 3. If f: R → R defined as f(x) = 2x.
𝑦
Solution. Let y = 2x then x = 2
𝑦
Thus, for every y R, we have x = 2 ∈ R such that f(x) = y.
Thus, f is onto.
Example 4. Consider the function f: R → R defined as f(x)= x2.
Solution. Let y = x2 therefore x = ± √𝑦
The square of any real number is non-negative.
It means that y ≥ 0.
Thus, for y ≤ 0, we cannot find an element x such that f(x) = y.
Thus, the range of f(x) is the set of non-negative real numbers and the
negative real numbers are not in the image of f(x).
As a result, f(x) is not onto.
Note: If you restrict the co-domain to ℝ+ ∪ {0}, which is the set of

non-negative real numbers, the function becomes onto.
1.12 INVERSE FUNCTION
Let f: A → B be a one – one and onto function. Then the function

g: B → A which associates to each element b ∈ 𝐵 the unique element
a ∈ 𝐴 such that f(a) = b is called the inverse function of f. the inverse
function of f is denoted by 𝑓 −1 .
Note: every function does not have an inverse. A function f: A → B has
inverse iff f is one – one and onto. If f has inverse, then f is said to be
invertible and 𝑓 −1 : B → A. also if a ∈ 𝐴, then f(a) = b where b ∈ 𝐵
⟹ a = 𝑓 −1 (𝑏).
1.13 COMPOSITE OF FUNCTION
Let f : A → B and g : B → C be two functions. Then the composition of

f and g, denoted by g ∘ f, is defined as the function g ∘ f : A → C given
by g ∘ f (x) = g(f (x)), ∀ x ∈ A.
Domain: f(g(x)) is read as f of g of x. In the composition of (f o g) (x)
the domain of function f becomes g(x). The domain is a set of all values
which go into the function.
Example: If f(x) = 3x+1 and g(x) = x2, then f of g of x,
f(g(x)) = f(x2) = 3x2+1.
If we reverse the function operation, such as f of f of x,
g(f(x)) = g(3x+1) = (3x+1)2
CHECK YOUR PROGRESS
True or false Questions
Problem 1. function f: R→ R, then f(x) = 2x is injective.
Problem 2. function f: R→ R, then f(x) = 2x+1 is not injective.

Problem 3. The onto function is also called the surjective function.
Problem 4. function f: R→ R, f(x) = |x| is an onto function.
Problem 5. In the surjective function, the range of the function “f” is
equal to the codomain.
1.14 SUMMARY
1. A set is a well - defined collection of distinct objects.
2. A ⊆ B; which means Set A is a subset of Set B.
3. For any two sets A and B, A ∩ B ⊂ 𝐴 and A ∩ B ⊂ 𝐵.
4. (a) (A ∪ B)c = Ac ∩ B c (b) (A ∩ B)c = Ac ∪ B c
5. Let f: A → B and g: B → C be two functions. Then the composition

of f and g, denoted by g ∘ f, is defined as the function g ∘ f: A → C given
by g ∘ f (x) = g(f (x)), ∀ x ∈ A.
1.15 GLOSSARY
Numbers
letters
Collections of objects
1.16 REFERENCES
1. T. M. Apostol, Mathematical Analysis (2nd Edition), Narosa

Publishing House, 2002.
2. R.G. Bartle and D.R. Sherbert, Introduction of real analysis (3 rd
Edition), John Wiley and Sons (Asia) P. Ltd., Inc. 2000.
3. W. Rudin, Principles of Mathematical Analysis (3rd Edition),
McGraw-Hill Publishing, 1976.
1.17 SUGGESTED READING
4. S.C. Malik and Savita Arora, Mathematical Analysis (6th Edition),

New Age International Publishers, 2021.
5. Shanti Narayan, A course of Mathematical Analysis (29th Edition),
S. Chand and Co., 2005.
6. K. A. Ross, Elementary Analysis, The Theory of Calculus (2nd

edition), Springer, 2013.
1.18 TERMINAL AND MODEL QUESTIONS
Q 1. Prove that the function f:ℕ ⟶ ℕ is given by f(x) = 𝑥 2 is one – one

function.
Q 2. Prove that the function f:ℕ ⟶ ℕ is given by f(x) = 𝑥 2 is not onto

function.
Q 3. Let A = [– 1, 1]. Then, discuss whether the following functions

defined on A are one-one, onto or bijective.
𝑥
(a) f(x) = 2. (b) f(x) = 𝑥 2
Q 4. If f(x) = 3x2, then find ( f ∘ f )(x).
Q 5. If f (x) = 2x and g(x) = x+1, then find ( f ∘ g )(x) if x = 1.
1.19 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. False
CYQ 3. True
CYQ 4. False
CYQ 5. True
TERMINAL QUESTIONS
TQ 3. (a) One - One but not Onto.
(b) Not One - One and not Onto.
TQ 4. 27 𝑋 2
TQ 5. 4
Real Analysis MT(N)-201
UNIT 2: REAL NUMBERS
CONTENTS:
2.1 Introduction
2.2 Objectives
2.3 Real numbers
2.4 Algebraic properties of ℝ
2.5 Order properties of ℝ
2.6 Absolute value
2.7 Triangle inequality
2.8 Completeness property of ℝ
2.9 Supremum and Infimum property of ℝ
2.10 Archimedean property of ℝ
2.11 Extended set of real number
2.12 Summary
2.13 Glossary
2.14 References
2.17 Answers
2.1 INTRODUCTION
The modern study of set theory was initiated by the German

mathematicians Richard Dedekind and Georg Cantor in the 1870s. In particular,
Georg Cantor is commonly considered the founder of set theory. The non-
formalized systems investigated during this early stage go
2.2 OBJECTIVES
(i) Real numbers

(ii) Algebraic properties of ℝ
(iii) Order properties of ℝ
(iv) Completeness property of ℝ
(v) Supremum and Infimum property of ℝ
(vi) Archimedean property of ℝ
2.3 REAL NUMBERS
A set containing all rational as well as irrational numbers is called the set
of all real numbers. The set of real number is denoted by ℝ.
We now describe some fundamental properties of the set ℝ.
1. Algebraic properties of ℝ.
2. Order properties of ℝ.
3. Completeness property of ℝ.
4. Archimedean property of ℝ.
2.4 ALGEBRAIC PROPERTIES OF ℝ
Addition and multiplication are defined on the set ℝ satisfying the

following properties:
A1. a + b ∈ ℝ for all a, b in ℝ.
A2. (a + b) + c = a + (b + c) for all a, b, c in ℝ.
A3. There exists an element 0 in ℝ (called zero element) such that
a + 0 = a for all a in ℝ.
A4. For each a in ℝ there exists an element – a in ℝ such that
a + (-a) = 0.
A5. a + b = b + a for all a, b in ℝ.
M1. a . b ∈ ℝ for all a, b in ℝ.
M2. (a . b) . c = a . (b . c) for all a, b, c in ℝ.
M3. There exists an element 1 in ℝ (called unity element) such that
a .1 = a for all a in ℝ.
1
M4. For each element a ≠ 0 in ℝ there exists an element in ℝ such
𝑎
1
that a . (𝑎) = 1.
M5. a . b = b . a for all a, b in ℝ.
D. a . (b + c) = a . b + a . c for all a, b in ℝ.
Theorem 2.4.1 Let a, b, c ∈ ℝ. Then
(i) a + b = a + c implies b = c (cancellation law for addition).
(ii) a ≠ 0 and a . b = a . c implies b = c (cancellation law for

multiplication).
Proof. (i) since a + b = a + c
-a ∈ ℝ, since 𝑎 ∈ ℝ. Therefore -a + (a + b) = -a + (a + c)
Or (-a + a) + b = (-a + a) + c, by A2
Or 0 + b = 0 + c, by A4
Or b = c.
(ii) since a.b = a.c

1 1 1
∈ ℝ, since 𝑎 ≠ 0. Therefore (𝑎) (𝑎. 𝑏) = (𝑎) (𝑎. 𝑐)
𝑎
1 1
Or, (𝑎 . 𝑎) . 𝑏 = (𝑎 . 𝑎) . 𝑐, by M2
Or, 1. 𝑏 = 1. 𝑐, by M4
Or, 𝑏 = 𝑐.
Theorem 2.4.2 Let 𝑎, 𝑏, 𝑐 ∈ ℝ. Then 𝑎. 𝑏 = 0 implies 𝑎 = 0 or 𝑏 = 0.
1 1
Proof. Let 𝑎 ≠ 0 then 𝑎 ∈ ℝ and 𝑎 . 𝑎 = 1.
1 1 1
𝑎. 𝑏 = 0 ⟹ 𝑎 . (𝑎. 𝑏) = 𝑎 . 0 ⟹ (𝑎 . 𝑎) . 𝑏 = 0 ⟹ 𝑏 = 0.
Therefore 𝑎 ≠ 0 ⟹ 𝑏 = 0. Contrapositively, 𝑏 ≠ 0 ⟹ 𝑎 = 0.
Therefore either 𝑎 = 0 or 𝑏 = 0.
2.5 ORDER PROPERTIES OF ℝ
On the set of ℝ, a linear order relation < is defined by “𝑎 < 𝑏 if 𝑎 ∈ ℝ,

𝑏 ∈ ℝ and 𝑎 is less than 𝑏” and it satisfies the following conditions:
O1. If 𝑎, 𝑏 ∈ ℝ, then exactly one of the following statements holds –
𝑎 < 𝑏, or 𝑎 = 𝑏, or 𝑏 < 𝑎 (law of trichotomy);
O2. 𝑎 < 𝑏 and 𝑏 < 𝑐 ⟹ 𝑎 < 𝑐 for 𝑎, 𝑏, 𝑐 ∈ ℝ (transitivity);
O3. 𝑎 < 𝑏 and 𝑎 + 𝑐 < 𝑏 + 𝑐 for 𝑎, 𝑏, 𝑐 ∈ ℝ;
O4. 𝑎 < 𝑏 and 0 < 𝑐 ⟹ 𝑎𝑐 < 𝑏𝑐 for 𝑎, 𝑏, 𝑐 ∈ ℝ.
Note: 1. The field ℝ together with the order relation defined on ℝ
satisfying O1 – O4 becomes an ordered field.
Note: 2. n > 0 for all 𝑛 ∈ ℕ
1
Note: 3. For all 𝑛 ∈ ℕ, 𝑛 > 0.
𝑎+𝑏
Theorem 2.5.1 Let 𝑎, 𝑏 ∈ ℝ. Then 𝑎 < 𝑏 ⟹ 𝑎 < < 𝑏.
2
Proof. 𝑎 < 𝑏 ⟹ 𝑎 + 𝑎 < 𝑎 + 𝑏

⟹ 2𝑎 < 𝑎 + 𝑏
1 1 1 1
⟹ 2 . 2𝑎 < 2 (𝑎 + 𝑏), since 2 ∈ ℝ and 2 > 0
𝑎+𝑏
⟹𝑎< .
2
Also 𝑎 < 𝑏 ⟹ 𝑎 + 𝑏 < 𝑏 + 𝑏

⟹ 𝑎 + 𝑏 < 2𝑏
1 1 1 1
⟹ (𝑎 + 𝑏) < . 2𝑏, since ∈ ℝ and > 0
2 2 2 2
𝑎+𝑏
⟹ < 𝑏.
2
𝑎+𝑏
Therefore 𝑎 < < 𝑏.
2
Corollary. There is no least positive real number.

If possible, let 𝑎 be the least positive real number. Then 𝑎 > 0.
1
0 < 𝑎 ⟹ 0 < 2 𝑎 < 𝑎 by theorem.
1 1
This shows that 2 𝑎 is a positive real number and 2 𝑎 < 𝑎 indicates that a
is not the least positive real number.
It follows that there is no least positive real number.
2.6 ABSOLUTE VALUE
Let 𝑎 ∈ ℝ. The absolute value of 𝑎, denoted by |𝑎|, is defined by
𝑎, 𝑖𝑓 𝑎 > 0
|𝑎| = { 0, 𝑖𝑓 𝑎 = 0
−𝑎, 𝑖𝑓 𝑎 < 0
For example, |4| = 4, |−10| = 10, |0| = 0.
It follows from definition that |𝑎| is a non-negative real number. |𝑎| = 0

if and only if 𝑎 = 0.
Theorem 2.6.1 Prove that
(i) |−𝑎| = |𝑎| for all 𝑎 ∈ ℝ.
(ii) |𝑎𝑏| = |𝑎||𝑏| for all 𝑎, 𝑏 ∈ ℝ.
(iii) if 𝑎, 𝑐 ∈ ℝ 𝑐 > 0, then |𝑎| < 𝑐 ⟺ −𝑐 < 𝑎 < 𝑐.
Proof. (i) Let 𝑎 > 0 then −𝑎 < 0 and |−𝑎| = −(−𝑎) = 𝑎 = |𝑎|.
Now let 𝑎 < 0 then −𝑎 > 0 and |−𝑎| = −𝑎 = |𝑎|.
Let 𝑎 = 0 then −𝑎 = 0 and |−𝑎| = 0 = |𝑎|.
Combining the cases, we have |−𝑎| = |𝑎| for all 𝑎 ∈ ℝ.
(ii) Let one or both of 𝑎, 𝑏 be 0. Then 𝑎𝑏 = 0.
In this case |𝑎𝑏| = 0 and |𝑎||𝑏| = 0. Therefore |𝑎𝑏| = |𝑎||𝑏|.
Now let 𝑎 > 0, 𝑏 > 0 then 𝑎𝑏 > 0 and |𝑎𝑏| = 𝑎𝑏, |𝑎| = 𝑎, |𝑏| = 𝑏
Therefore |𝑎𝑏| = |𝑎||𝑏|.
Now let 𝑎 < 0, 𝑏 > 0 then 𝑎𝑏 < 0 and |𝑎𝑏| = −𝑎𝑏, |𝑎| = −𝑎, |𝑏| = 𝑏
Now let 𝑎 < 0, 𝑏 > 0 then proof is similar.
Now let 𝑎 < 0, 𝑏 < 0 then 𝑎𝑏 > 0 and |𝑎𝑏| = −𝑎𝑏, |𝑎| = −𝑎, |𝑏| = −𝑏
Combining the cases, we have |𝑎𝑏| = |𝑎||𝑏| for all 𝑎, 𝑏 ∈ ℝ.
(iii) let |𝑎| < 𝑐 then if 𝑎 ≥ 0, 𝑎 < 𝑐 and if 𝑎 < 0, −𝑎 < 𝑐
this implies that
−𝑐 < 𝑎. Therefore |𝑎| < 𝑐 ⟹ −𝑐 < 𝑎 < 𝑐.
Conversely, let 𝑐 > 0 and −𝑐 < 𝑎 < 𝑐
Then we have 𝑎 < 𝑐, 0 < 𝑐 and −𝑎 < 𝑐.
Combining, we have |𝑎| < 𝑐.
2.7 TRIANGLE INEQUALITY
For all 𝑎, 𝑏 ∈ ℝ, |𝑎 + 𝑏| ≤ |𝑎| + |𝑏|
Proof. We have −|𝑎| ≤ 𝑎 ≤ |𝑎| and −|𝑏| ≤ 𝑏 ≤ |𝑏|.
Then −(|𝑎| + |𝑏|) ≤ 𝑎 + 𝑏 ≤ |𝑎| + |𝑏|.

This implies |𝑎 + 𝑏| ≤ |𝑎| + |𝑏| for all 𝑎, 𝑏 ∈ ℝ.
Corollary 1. |𝑎 − 𝑏| ≤ |𝑎| + |𝑏| for all 𝑎, 𝑏 ∈ ℝ.
Proof. Replacing 𝑏 by −𝑏 in triangle inequality we get the inequality.
Corollary 2. ||𝑎| − |𝑏|| ≤ |𝑎 − 𝑏|.
Proof. |𝑎| = |𝑎 − 𝑏 + 𝑏| ≤ |𝑎 − 𝑏|+≤ |𝑏|
Or |𝑎| − |𝑏| ≤ |𝑎 − 𝑏|.
Again |𝑏| = |𝑏 − 𝑎 + 𝑎| ≤ |𝑏 − 𝑎| + |𝑎|
Or |𝑏| − |𝑎| ≤ |𝑏 − 𝑎| = |𝑎 − 𝑏|
So we have −|𝑎 − 𝑏| ≤ |𝑎| − |𝑏| ≤ |𝑎 − 𝑏|.
This implies ||𝑎| − |𝑏|| ≤ |𝑎 − 𝑏|, since −𝑐 ≤ 𝑎 ≤ 𝑐 ⟹ |𝑎| ≤ 𝑐.
Corollary 3. Let 𝑎1 , 𝑎2 , . . . . , 𝑎𝑛 ∈ ℝ then
|𝑎1 + 𝑎2 + . . . . + 𝑎𝑛 | ≤ |𝑎1 | + |𝑎2 | + ⋯ + |𝑎𝑛 |.
x+2
Example. Solve the equation |2x−1| = 3.
x+2 x+2
Solution. Given |2x − 1| = 3 ⟹ 2x−1 = ±3.
x+2
If 2x − 1 = 3 ⟹ 𝑥 + 2 = 6𝑥 − 3 ⟹ 𝑥 = 1
x+2 1
If 2x − 1 = −3 ⟹ 𝑥 + 2 = −6𝑥 − 3 ⟹ 𝑥 = 7
1
Therefore 𝑥 = 7 ,1 .
2.8 COMPLETENESS PROPERTY OF ℝ
Let P be a Subset of ℝ. A real number 𝑢 is said to be an upper bound of

P if 𝑥 ∈ 𝑃 ⟹ 𝑥 ≤ 𝑢. A real number 𝑙 is said to be a lower bound of P of
𝑥 ∈ 𝑃 ⟹ 𝑥 ≥ 𝑙.
Let P be a subset of ℝ. P is said to be bounded above if P has an upper

bound. P is said to be bounded below if P has lower bound.
P is said to be bounded set if P be bounded above as well as bounded

below.
1 1
Example 1. Let P = {1, 2 , 3 , … }. P is bounded above, 1 being an upper
bound. P is bounded below, 0 being a lower bound.
Example 2. Let P = {𝑥 ∈ 𝑃: 1 < x < 2} . P is bounded above, 2 being

an upper bound. P is bounded below, 1 being a lower bound.
Example 3. Let P = ∅. Every real number 𝑥 is an upper bound of the set

P. Every real number is a lower bound of the set P. therefore P is
bounded set.
2.9 SUPREMUM AND INFIMUM PROPERTY
OF ℝ
Let P be a subset of ℝ. If P is bounded above, then an upper bound

of P is said to be the supremum of P (or least upper bound of P) if it is
less than every other upper bound of P.
If P is bounded below then a lower bound of P is said to be the infimum

of P (or the greatest lower bound of P) if it is greater than every other
lower bound of S.
Note: 1. Every non-empty subset of ℝ that is bounded above has a least

upper bound or supremum.
2. Every non-empty subset of ℝ that is bounded below has a greatest

lower bound or infimum.
Properties of the supremum and the infimum.
Let P be a non-empty subset of ℝ, which is bounded above. Then

supremum of P exists, Let M = sup P. then 𝑀 ∈ ℝ and M satisfies the
following conditions:
(i) 𝑥 ∈ P ⟹ 𝑥 ≤ M, and
(ii) for each 𝜖 > 0, there exist an element 𝑦 (depends on 𝜖) in P such that
M − 𝜖 < 𝑦 ≤ M.
Let P be a non-empty subset of ℝ, which is bounded below. Then infimum

of P exists, Let 𝑚 = inf P. then 𝑚 ∈ ℝ and 𝑚 satisfies the following
conditions:
(i) 𝑥 ∈ P ⟹ 𝑥 ≥ 𝑚, and
(ii) for each 𝜖 > 0, there exist an element 𝑦 (depends on 𝜖) in P such that
𝑚 < 𝑦 ≤ 𝑚 + 𝜖.
Question 1. Prove that the set of natural number (ℕ) is not bounded
above.
Solution: The set ℕ is a non – empty subset of ℝ, since 1 ∈ ℕ. Let ℕ be

bounded above. Then ℕ being a non – empty subset of ℝ, sup ℕ exist by
supremum property of ℝ. Let u = sup ℕ. Then
(i) 𝑥 ∈ ℕ ⟹ 𝑥 ≤ u, and
(ii) for each 𝜖 > 0, there exist an element 𝑦 (depends on 𝜖) in ℕ such that
𝑢 − 𝜖 < 𝑦 ≤ u.
Let us choose 𝜖 = 1. Then there exists an element 𝑘 in ℕ such that
𝑢 − 1 < 𝑘 ≤ u. then 𝑢 − 1 < 𝑘 ⟹ 𝑢 < 𝑘 + 1.
Since 𝑘 is a natural number, 𝑘 + 1 is also a natural number. 𝑘 + 1 > 𝑢

implies that 𝑢 is not an upper bound of the set ℕ. Thus we arrive a
contradiction. So aur assumption was wrong. Hence the set of natural
number (ℕ) is not bounded above.
Question 2. Let P be a non – empty subset of ℝ which is bounded above

and T = {−𝑥 ∶ 𝑥 ∈ 𝑃}. Prove that the set T is bounded below and
𝑖𝑛𝑓𝑇 = −𝑠𝑢𝑝𝑃.
Solution: since P is bounded above therefore 𝑠𝑢𝑝𝑃 exists. Let 𝑢 = 𝑠𝑢𝑝 𝑃.
Then 𝑥 ∈ 𝑃 ⟹ 𝑥 ≤ 𝑢. let 𝑦 ∈ 𝑇 then −𝑦 ∈ 𝑃 and therefore −𝑦 ≤ 𝑢, i.e.
𝑦 ≥ −𝑢. This implies that −𝑢 is a lower bound of T. therefore the set T is
bounded below.
Let us choose 𝜖 > 0. Since 𝑢 = 𝑠𝑢𝑝𝑃, there exists an element p in P such
that 𝑢 − 𝜖 < 𝑝 ≤ 𝑢. Therefore −𝑢 ≤ −𝑝 < −𝑢 + 𝜖. ……(i)
Let 𝑞 = −𝑝. Then 𝑞 ∈ 𝑇.
(i) shows that for a pre-assigned positive 𝜖 there exists an element 𝑞 in T
such that −𝑢 ≤ 𝑞 < −𝑢 + 𝜖.
This proves that −𝑢 = 𝑖𝑛𝑓𝑇. Therefore 𝑖𝑛𝑓𝑇 = − 𝑠𝑢𝑝𝑃.
Question 3. Let P be a non – empty bounded subset of ℝ with 𝑠𝑢𝑝𝑃 = 𝑀

and 𝑖𝑛𝑓𝑃 = 𝑚. Prove that the set T = {|𝑥 − 𝑦|: 𝑥 ∈ 𝑃, 𝑦 ∈ 𝑃 } is bounded
above and 𝑠𝑢𝑝𝑇 = 𝑀 − 𝑚.
Solution: 𝑥 ∈ 𝑃 ⟹ 𝑚 ≤ 𝑥 ≤ 𝑀, 𝑦 ∈ 𝑃 ⟹ 𝑚 ≤ 𝑦 ≤ 𝑀.
Therefore 𝑚 − 𝑀 ≤ 𝑥 − 𝑦 ≤ 𝑀 − 𝑚 i.e. |𝑥 − 𝑦| ≤ 𝑀 − 𝑚.
This shows that T is bounded above, 𝑀 − 𝑚 being an upper bound.
Let 𝑎 ∈ 𝑃. Then |𝑎 − 𝑎| ∈ 𝑇 showing that T is non-empty. By the
supremum property of ℝ, 𝑠𝑢𝑝𝑇 exists.
We prove that no real number less than 𝑀 − 𝑚 is an upper bound of T.
if possible, let 𝑝 < 𝑀 − 𝑚 be an upper bound of T.
Let (𝑀 − 𝑚) − 𝑝 = 2𝜖. Then 𝜖 > 0 and 𝑝 + 𝜖 = 𝑀 − 𝑚 − 𝜖.
Since 𝑠𝑢𝑝𝑃 = 𝑀, there exist an element 𝑥 ∈ 𝑃 such that
𝜖
𝑀−2<𝑥≤𝑀
Since 𝑖𝑛𝑚𝑃 = 𝑀, there exist an element 𝑦 ∈ 𝑃 such that

𝜖
𝑚 < 𝑥 ≤ 𝑚 + 2.
Now 𝑥 − 𝑦 > 𝑀 − 𝑚 − 𝜖 i.e. 𝑥 − 𝑦 < 𝑝 + 𝜖.

This shows that 𝑝 is not an upper bound of T.
Therefore, no real number less than 𝑀 − 𝑚 is an upper bound of T.
Hence 𝑠𝑢𝑝𝑇 = 𝑀 − 𝑚.
Note: Let 𝐴, 𝐵 be bounded subset of ℝ such that 𝑥 ∈ 𝐴, 𝑦 ∈ 𝐵 ⟹ 𝑥 ≤ 𝑦.

Then 𝑠𝑢𝑝𝐴 ≤ 𝑖𝑛𝑓𝐵.
2.10 ARCHIMEDEAN PROPERTY OF ℝ
If 𝑥, 𝑦 ∈ ℝ and 𝑥 > 0, 𝑦 > 0, then there exists a natural number 𝑛 such

that 𝑛𝑦 > 𝑥.
Proof: If possible, let there exist no natural number n for which 𝑛𝑦 > 𝑥.
Then for every natural number 𝑘, 𝑘𝑦 ≤ 𝑥.
Thus, the set S = {𝑘𝑦: 𝑘 ∈ ℕ} is bounded above, 𝑥 being an upper bound.
S is non-empty because 𝑦 ∈ 𝑆.
By the supremum property of ℝ, 𝑠𝑢𝑝𝑆 exists. Let 𝑠𝑢𝑝𝑆 = b. then 𝑘𝑦 ≤ 𝑏

for all 𝑘 ∈ ℕ.
𝑏 − 𝑦 < 𝑏 since 𝑦 > 0. This shows that 𝑏 − 𝑦 is not an upper bound of 𝑆
and therefore there exists a natural number 𝑝 such that
𝑏 − 𝑦 < 𝑝𝑦 ≤ 𝑏. This implies (𝑝 + 1)𝑦 > 𝑏 …….. (i)
But 𝑝 ∈ ℕ ⟹ 𝑝 + 1 ∈ ℕ and therefore (𝑝 + 1)𝑦 ∈ 𝑆.
(i) shows that 𝑏 is not the supremum of S, a contradiction.
Therefore, our assumption is wrong and the existence of a natural number
𝑛 satisfying 𝑛𝑦 > 𝑥 is proved.
Note: (1) If 𝑥 ∈ ℝ, then there exists a natural number n such that 𝑛 > 𝑥.
Case (i). If 𝒙 > 𝟎.
Taking 𝑦 = 1, by Archimedean property of ℝ there exists a
natural number 𝑛 such that 𝑛. 1 > 𝑥 and hence existence is proved.
Case (ii). If 𝒙 ≤ 𝟎. Then 𝒏 = 𝟏.
(2). If 𝑥 ∈ ℝ and 𝑥 > 0, then there exists a natural number 𝑛 such that
1
0 < 𝑛 < 𝑥.
Taking 𝑦 = 1, by Archimedean property of ℝ there exists a natural

number 𝑛 such that 𝑛𝑥 > 1.
1 1
Since 𝑛 is a natural number, 𝑛 > 0 and therefore > 0 and also < 𝑥.
𝑛 𝑛
1
Therefore, we have 0 < 𝑛 < 𝑥.
(3). If 𝑥 ∈ ℝ and 𝑥 > 0, then there exists a natural number 𝑚 such that
𝑚 − 1 ≤ 𝑥 < 𝑚.
Taking 𝑦 = 1 and 𝑥 > 0, by Archimedean property of ℝ there exist a
natural number 𝑛 such that 𝑛. 1 > 𝑥, i.e. 𝑛 > 𝑥.
Let 𝑆 = {𝑘 ∈ ℕ: 𝑘 > 𝑥}. Then 𝑆 is non-empty subset of ℕ, since 𝑛 ∈ 𝑆. By
well ordering property of the set ℕ, 𝑆 has a least element, say 𝑚. Since
𝑚 ∈ 𝑆, 𝑚 > 𝑥.
As 𝑚 is least element of 𝑆. 𝑚 − 1 ≯ 𝑥. i.e. 𝑚 − 1 ≤ 𝑥.
Hence 𝑚 − 1 ≤ 𝑥 < 𝑚.
2.11 EXTENDED SET OF REAL NUMBER
It is often convenient to extend the set ℝ by the addition of two elements

∞ and −∞. This enlarged set is called the extended set of real numbers
and is often denoted by ℝ∗ .
In the extended set ℝ∗ we define –
For all 𝑥 ∈ ℝ, 𝑥 + ∞ = ∞ + 𝑥 = ∞
𝑥 + (−∞) = (−∞) + 𝑥 = −∞.
For all 𝑥 > 0, 𝑥. ∞ = ∞. 𝑥 = ∞ and
𝑥. (−∞) = (−∞). 𝑥 = −∞.
For all 𝑥 < 0, 𝑥. ∞ = ∞. 𝑥 = −∞ and
𝑥. (−∞) = −∞. 𝑥 = ∞.
Now ∞ + ∞ = ∞, (−∞) + (−∞) = −∞
∞. ∞ = ∞, (-∞). ∞ = −∞, (-∞). (−∞) = ∞.
And ∞ + (−∞), (-∞) + ∞, 0. ∞, ∞. 0, 0 − ∞, −∞. 0 are not defined.
Now, if 𝑆 be a non-empty subset of ℝ having no upper bound, we define
𝑠𝑢𝑝𝑆 = ∞. If 𝑆 be a non-empty subset of ℝ having no longer bound, we
define 𝑖𝑛𝑓𝑆 = −∞.
CHECK YOUR PROGRESS
Problem 1. The set of real number is not bounded above.

Problem 2. The set of natural number is bounded below.
𝑎+𝑏
Problem 3. Let 𝑎, 𝑏 ∈ ℝ. Then 𝑎 < 𝑏 ⟹ 𝑎 < < 𝑏.
2
Problem 4. |𝑎𝑏| = |𝑎||𝑏| for all 𝑎, 𝑏 ∈ ℝ.

Problem 5. The supremum of the interval (1, 3) is 4.
2.12 SUMMARY
1. A set containing all rational as well as irrational numbers is called the

set of all real numbers. The set of real number is denoted by ℝ.
2. There is no least positive real number.

3. Every non-empty subset of ℝ that is bounded above has a least upper
bound or supremum.
4. If 𝑥, 𝑦 ∈ ℝ and 𝑥 > 0, 𝑦 > 0, then there exists a natural number 𝑛

such that 𝑛𝑦 > 𝑥 is called Archimedean property of ℝ.
5. The set of natural number (ℕ) is not bounded above.
2.13 GLOSSARY
Numbers
Sets
Intervals
Modulus
2.14 REFERENCES

2. R.G. Bartle and D.R. Sherbert, Introduction of real analysis (3rd


Q 1. Prove that the set of natural number ℕ is not bounded above.
Q 2. Prove that the set {1, 2, 3, …, 10} is not bounded above.
1
Q 3. Find the Supremum and infimum of {𝑛 : 𝑛 ∈ ℕ}.
Q 4. Prove that the set T = {|𝑥 − 𝑦|: 𝑥 ∈ 𝑃, 𝑦 ∈ 𝑃 } is bounded above.

x+3
Q 5. Solve the equation |2x−6| ≤ 1.
2.17 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. True
CYQ 3. True
CYQ 4. True
CYQ 5. False
TERMINAL QUESTIONS
TQ 3. 1, 0
TQ 5. Solution set is {𝑥 ∈ ℝ: 𝑥 ≥ 9} ∪ {𝑥 ∈ ℝ: 𝑥 ≤ 1}.
UNIT 3: LIMIT POINT, OPEN SET AND

CLOSED SETS
CONTENTS:
3.1 Introduction
3.2 Objectives
3.3 Neighbourhood
3.4 Interior point
3.5 Open set
3.6 Limit point
3.7 Closed set
3.8 Summary
3.9 Glossary
3.10 References
3.13 Answers
3.1 INTRODUCTION
A point is called a limit point of a set in the Euclidean plane if

there is no minimum distance from that point to members of the set; for
example, the set of all numbers less than 1 has 1 as a limit point. A set is
not connected if it can be divided into two parts such that a point of one
part is never a limit point of the other part. The set is connected if it cannot
be so divided. For example, if a point is removed from an arc, any
remaining points on either side of the break will not be limit points of the
other side, so the resulting set is disconnected. If a single point is removed
from a simple closed curve such as a circle or polygon, on the other hand,
it remains connected; if any two points are removed, it becomes
disconnected. A figure-eight curve does not have this property because
one point can be removed from each loop and the figure will remain
connected.
3.2 OBJECTIVES
(i) Neighbourhood
(ii) Interior point
(iii) Open set
(iv) Limit point
3.3 NEIGHBOURHOOD
Let 𝑐 ∈ ℝ. A subset 𝑆 ⊂ ℝ is said to be neighbourhood of 𝑐 if there

exist an open interval (𝑎, 𝑏) such that 𝑐 ∈ (𝑎, 𝑏) ⊂ 𝑆.
Clearly an open bounded interval containing the point 𝑐 is a

neighbourhood of 𝑐. Such a neighbourhood of 𝑐 is denoted by
𝑁(𝑐).
A closed bounded interval containing the point 𝑐 may not be a
neighbourhood of 𝑐.
For example: 1 ∈ [1, 5] is not a neighbourhood of 1.
Theorem 3.3.1. Let 𝑐 ∈ ℝ. The union of two neighbourhoods of 𝑐

is a neighbourhood of 𝑐.
Proof. Let 𝑆1 ⊂ ℝ, 𝑆2 ⊂ ℝ be two neighbourhoods of 𝑐. Then there
exists open interval (𝑎1 , 𝑏1 ), (𝑎2 , 𝑏2 ) such that 𝑐 ∈ (𝑎1 , 𝑏1 ) ⊂ 𝑆1 and
𝑐 ∈ (𝑎2 , 𝑏2 ) ⊂ 𝑆2 .
Then 𝑎1 < 𝑏1 , 𝑎2 < 𝑏1 and 𝑎1 < 𝑏2 , 𝑎2 < 𝑏2 . Let 𝑎3 = min{𝑎1 , 𝑎2 },
𝑏3 = max{𝑏1 , 𝑏2 }. Then (𝑎1 , 𝑏1 ) ∪ (𝑎2 , 𝑏2 ) = (𝑎3 , 𝑏3 ) and 𝑐 ∈ (𝑎2 , 𝑏2 ).
(𝑎1 , 𝑏1 ) ⊂ 𝑆1 ∪ 𝑆2 and (𝑎2 , 𝑏2 ) ⊂ 𝑆1 ∪ 𝑆2
⟹ (𝑎3 , 𝑏3 ) = (𝑎1 , 𝑏1 ) ∪ (𝑎2 , 𝑏2 ) ⊂ 𝑆1 ∪ 𝑆2.
Thus 𝑐 ∈ (𝑎3 , 𝑏3 ) ⊂ 𝑆1 ∪ 𝑆2 .
This prove that 𝑆1 ∪ 𝑆2 is a neighbourhood of 𝑐.
Note: The union of finite number of neighbourhoods of 𝑐 is a
Theorem 3.3.2. Let 𝑐 ∈ ℝ. The intersection of two

neighbourhoods of 𝑐 is a neighbourhood of 𝑐.
Proof. Let 𝑆1 ⊂ ℝ, 𝑆2 ⊂ ℝ be two neighbourhoods of 𝑐. Then there
exists open interval (𝑎1 , 𝑏1 ), (𝑎2 , 𝑏2 ) such that 𝑐 ∈ (𝑎1 , 𝑏1 ) ⊂ 𝑆1 and
𝑐 ∈ (𝑎2 , 𝑏2 ) ⊂ 𝑆2 .
Then 𝑎1 < 𝑏1 , 𝑎2 < 𝑏1 and 𝑎1 < 𝑏2 , 𝑎2 < 𝑏2 . Let 𝑎3 = max{𝑎1 , 𝑎2 },
𝑏3 = min{𝑏1 , 𝑏2 }. Then (𝑎1 , 𝑏1 ) ∩ (𝑎2 , 𝑏2 ) = (𝑎3 , 𝑏3 ) and 𝑐 ∈ (𝑎2 , 𝑏2 ).
(𝑎3 , 𝑏3 ) = (𝑎1 , 𝑏1 ) ∩ (𝑎2 , 𝑏2 ) ⊂ (𝑎1 , 𝑏1 ) ⊂ 𝑆1 and (𝑎3 , 𝑏3 ) =
(𝑎1 , 𝑏1 ) ∩ (𝑎2 , 𝑏2 ) ⊂ (𝑎2 , 𝑏2 ) ⊂ 𝑆2
⟹ (𝑎3 , 𝑏3 ) ⊂ 𝑆1 ∩ 𝑆2 .
Thus 𝑐 ∈ (𝑎3 , 𝑏3 ) ⊂ 𝑆1 ∩ 𝑆2 .
This prove that 𝑆1 ∩ 𝑆2 is a neighbourhood of 𝑐.
Note: The intersection of finite number of neighbourhoods of 𝑐 is a

Note: The intersection of infinite number of neighbourhoods of 𝑐 may or
may not be neighbourhood of 𝑐.
1 1
Example: for every 𝑛 ∈ ℕ, (− 𝑛 , 𝑛) is neighbourhood of 0.
1 1
⋂∞
𝑛=1 (− 𝑛 , 𝑛 ) = {0}. This is not a neighbourhood of 0.
3.4 INTERIOR POINT
Let 𝑆 be a subset of ℝ. A point 𝑥 in 𝑆 is said to be an interior point

of 𝑆 if there exists a neighbourhood 𝑁(𝑥) of 𝑥 such that
𝑁(𝑥) ⊂ 𝑆.
The set of all interior points of 𝑆 is said to be interior of 𝑆 and denoted
by int S or by 𝑆 0 .
From definition it follows that 𝑆 0 ⊂ 𝑆 for any set 𝑆 ⊂ ℝ.
1 1
Examples 1. Let 𝑆 = {1, , , … }.
2 3
Let 𝑥 ∈ 𝑆, every neighbourhood of 𝑥 contain some points not in 𝑆.

So 𝑥 can not be an interior point of 𝑆. Therefore, int S = ∅.
Examples 2. Let 𝑆 = ℕ.
So 𝑥 can not be an interior point of 𝑆. Therefore, int S = ∅.
Examples 3. Let 𝑆 = ℚ.
So 𝑥 can not be an interior point of 𝑆. Therefore, 𝑆 0 = ∅.
Examples 4. Let 𝑆 = { 𝑥 ∈ ℝ: 1 < 𝑥 < 3}. Each point of 𝑆 is an
interior point of 𝑆. so interior of 𝑆 = 𝑆.
Examples 5. Let 𝑆 = ℝ. Each point of 𝑆 is an interior point of 𝑆. so

interior of 𝑆 0 = 𝑆.
3.5 OPEN SET
Let 𝑆 be a subset of ℝ. 𝑆 is said to be open set if each point of 𝑆 is

an interior point of 𝑆.
1 1
Examples 1. Let 𝑆 = {1, , , … }. no point of 𝑆 is an interior point
2 3
of 𝑆. therefore, 𝑆 is not an open set.

Examples 2. Let 𝑆 = ℤ.
no point of 𝑆 is an interior point of 𝑆. therefore, 𝑆 is not an open
set.
Examples 3. Let 𝑆 = ℚ.
no point of 𝑆 is an interior point of 𝑆. therefore, 𝑆 is not an open
set.
Examples 4. Let 𝑆 = { 𝑥 ∈ ℝ: 1 < 𝑥 < 3}. Each point of 𝑆 is an
interior point of 𝑆. therefore, 𝑆 is an open set.
Examples 5. Let 𝑆 = ℝ. Each point of 𝑆 is an interior point of 𝑆.

therefore, 𝑆 is an open set.
Theorem 3.5.1. Let 𝑆 ⊂ ℝ. then 𝑆 is an open set if and only if

𝑆 = 𝑖𝑛𝑡 𝑆.
Proof. Observe in general that 𝑖𝑛𝑡 𝑆. ⊆ 𝑆. Now suppose that S is open.
Then for every 𝑥 ∈ 𝑆, there is 𝜖 > 0 so that 𝑁(𝑥, 𝜖 ) ⊆ 𝑆 and this just says
𝑥 ∈ 𝑖𝑛𝑡 𝑆 and since 𝑥 was arbitrary we have shown that 𝑆 ⊆ 𝑖𝑛𝑡 𝑆 or
equivalently 𝑆 = 𝑖𝑛𝑡 𝑆.
Conversely suppose the 𝑖𝑛𝑡 𝑆= 𝑆. The if 𝑥 ∈ S then 𝑥 is an interior point
and there is an 𝜖 > 0 so that 𝑁(𝑥, 𝜖 ) ⊆ 𝑆. But this says S is open.
Theorem 3.5.2. The union of two open sets in ℝ is an open set.

Proof. Let 𝐺1 and 𝐺1 be two open Sets in ℝ.
Let 𝑥 ∈ 𝐺1 ∪ 𝐺2. Then 𝑥 ∈ 𝐺1 or 𝑥 ∈ 𝐺2 .
Let 𝑥 ∈ 𝐺1. Since 𝐺1 is an open set and 𝑥 ∈ 𝐺1 , 𝑥 is an interior point of
𝐺1. Therefore, there exists a neighbourhood 𝑁(𝑥) of 𝑥 such that
𝑁(𝑥) ⊂ 𝐺1 ⟹ 𝑁(𝑥) ⊂ 𝐺1 ∪ 𝐺2.
This shows that 𝑥 is an interior point of 𝐺1 ∪ 𝐺2.
Since 𝑥 is arbitrary, every point of 𝐺1 ∪ 𝐺2 is an interior point of
𝐺1 ∪ 𝐺2. Therefore 𝐺1 ∪ 𝐺2 is an open set.
If however, 𝑥 ∈ 𝐺2 , we can prove in similar manner that 𝐺1 ∪ 𝐺2 is
an open set. This completes the proof.
Theorem 3.5.3. The intersection of two open sets in ℝ is an open

set.
Proof. Let 𝐺1 and 𝐺1 be two open Sets in ℝ.
Case 1. 𝐺1 ∩ 𝐺2 = ∅. Since ∅ is an open set, 𝐺1 ∩ 𝐺2 is an open set.
Case 2. 𝐺1 ∩ 𝐺2 ≠ ∅. Let 𝑥 ∈ 𝐺1 ∩ 𝐺2. Then 𝑥 ∈ 𝐺1 and 𝑥 ∈ 𝐺2 .
Since 𝐺1 is an open set and 𝑥 ∈ 𝐺1, 𝑥 is an interior point of 𝐺1.
Hence there exists a positive 𝛿1 such that the neighbourhood
𝑁(𝑥, 𝛿1 ) ⊂ 𝐺1 .
Since 𝐺2 is an open set and 𝑥 ∈ 𝐺2 , 𝑥 is an interior point of 𝐺2 .
Hence there exists a positive 𝛿2 such that the neighbourhood
𝑁(𝑥, 𝛿2 ) ⊂ 𝐺1 .
Let 𝛿 = min{𝛿1 , 𝛿2 }. Then 𝛿.
𝑁(𝑥, 𝛿) ⊂ 𝑁(𝑥, 𝛿1 ) ⊂ 𝐺1 and 𝑁(𝑥, 𝛿) ⊂ 𝑁(𝑥, 𝛿2 ) ⊂ 𝐺2 .
Consequently, 𝑁(𝑥, 𝛿) ⊂ 𝐺1 ∩ 𝐺2.
This shows that 𝑥 is an interior point of 𝐺1 ∩ 𝐺2. Since x is arbitrary,
𝐺1 ∩ 𝐺2 is an open set and this completes the proof.
Theorem 3.5.4. The union of an arbitrary collection of open sets in

ℝ is an open set.
Proof. Let {𝐺𝛼 : 𝛼 ∈ ∧}, ∧ being the index set, be an arbitrary

collection of open sets in ℝ. Let G = ⋃𝛼∈ ∧ 𝐺𝛼 .
𝑥 ∈ G. then 𝑥 belongs to at least one open set of the collection, say
𝐺𝜆 , (𝜆 ∈ ∧).
Since 𝐺𝜆 is an open set and 𝑥 ∈ 𝐺𝜆 , 𝑥 is an interior point of 𝐺𝜆 .
Therefore, there exists a neighbourhood 𝑁(𝑥) of 𝑥 such that 𝑁(𝑥) ⊂
𝐺𝜆 . 𝑁(𝑥) ⊂ 𝐺𝜆 ⟹ 𝑁(𝑥) ⊂ G.
This shows that 𝑥 is an interior point of G. Since x is arbitrary, G is an
open set and this completes the proof.
Note: The intersection of an infinite number of open sets in ℝ is not

necessarily an open set.
Let us consider the sets 𝐺𝑖 Where
𝐺1 = {𝑥 ∈ ℝ: − 1 < 𝑥 < 1}
1 1
𝐺2 = {𝑥 ∈ ℝ: − 2 < 𝑥 < 2}
… … … … …
1 1
𝐺𝑛 = {𝑥 ∈ ℝ: − 𝑛 < 𝑥 < 𝑛}
… … … … …
Each 𝐺𝑖 is an open set. ⋂∞
𝑖=1 𝐺𝑖 = {0}. This is not an open set.
Let us consider the sets 𝐺𝑖 Where

𝐺1 = {𝑥 ∈ ℝ: − 1 < 𝑥 < 1}
𝐺2 = {𝑥 ∈ ℝ: − 2 < 𝑥 < 2}
… … … … …
𝐺𝑛 = {𝑥 ∈ ℝ: − 𝑛 < 𝑥 < 𝑛}
… … … … …
Each 𝐺𝑖 is an open set. ⋂∞
𝑖=1 𝐺𝑖 = 𝐺1 . This is an open set.
From these two examples we conclude that the intersection of an

infinite number of open sets in ℝ is not necessarily an open set.
Note: Every open interval is open set.
Theorem 3.5.5. Let 𝑆 is a subset of ℝ then 𝑖𝑛𝑡 𝑆 is an open set.

Proof.
Case 1. 𝑖𝑛𝑡 𝑆 = ∅. Since ∅ is an open set, 𝑖𝑛𝑡 𝑆 is an open set.
Case 2. 𝑖𝑛𝑡 𝑆 ≠ ∅. let 𝑥 ∈ 𝑖𝑛𝑡 𝑆. then 𝑥 is an interior point of 𝑆.
therefore, there exist a neighbourhood 𝑁(𝑥) of 𝑥 such that 𝑁(𝑥) ⊂ 𝑆.
let 𝑥 ∈ 𝑁(𝑥). Then 𝑁(𝑥) is neighbourhood of 𝑦 also and since
𝑁(𝑥) ⊂ 𝑆, 𝑦 is an interior point of 𝑆.
Thus 𝑦 ∈ 𝑁(𝑥). Then 𝑁(𝑥) ⟹ 𝑦 ∈ 𝑖𝑛𝑡 𝑆. therefore 𝑁(𝑥) ) ⊂ 𝑖𝑛𝑡 𝑆.
This shows that 𝑥 is an interior point of 𝑖𝑛𝑡 𝑆.
Thus 𝑥 ∈ 𝑖𝑛𝑡 𝑆 ⟹ 𝑥 is an interior point of 𝑖𝑛𝑡 𝑆.
Therefore 𝑖𝑛𝑡 𝑆 is an open set. This completes the proof.
3.6 LIMIT POINT
A point 𝑥 ∈ ℝ is said to be limit point of a subset of 𝑆 of ℝ if every

neighbourhood of 𝑥 has a point of 𝑆 other that 𝑥.
In symbols, a point 𝑥 ∈ ℝ is said to be limit point of a subset 𝑆 of ℝ if
for each neighbourhood 𝑁 of 𝑥,
(𝑁 ∩ 𝑆) − {𝑝} ≠ ∅.
Note: 1. Limit point of set is also called a limiting point or a cluster
point or a condensation point or an accumulation point of the set.
Note: 2. A finite set has no limit point.
Note: 3. A limit point of 𝑆 may or may not belongs to 𝑆.
Note: 4. ISOLATED POINT.
A point 𝑥 ∈ 𝑆 is called an isolated point of 𝑆 if 𝑥 is not a limit point of 𝑆.
Note:5. Set of all limit points of 𝑆 is called the derived set of 𝑆 and is
denoted by 𝑆 ′ . thus
𝑆 ′ = {𝑥: 𝑥 𝑖𝑠 𝑎 𝑙𝑖𝑚𝑖𝑡 𝑝𝑜𝑖𝑛𝑡 𝑜𝑓 𝑆}.
Example: 1. Prove that 0 is the limit point of the set

1
𝑆 = {𝑛 : 𝑛 ∈ ℕ}.
Solution: for each 𝜖 > 0, (−𝜖, 𝜖) is neighbourhood of 0.

By Archimedean property of reals, for each 𝜖 > 0, ∃ 𝑛 ∈ ℕ
1
such that 𝑛 > 𝜖
1 1 1
⟹ < 𝜖 ⟹ −𝜖 < 0 < 𝑛 < 𝜖 ⟹ ∈ (−𝜖, 𝜖)
𝑛 𝑛
1
Thus, every neighbourhood of 0 contains a point of 𝑆, namely 𝑛.
⟹ 0 is limit point of 𝑆.
Example: 2. Find the derived set of each of the following:

(i) (1, ∞) (ii) (−∞, −1)
Solution:1. Let 𝑥 be any real number.
If 𝑥 < 1, then for 0 < 𝜖 < 1 − 𝑥, (𝑥 − 𝜖, 𝑥 + 𝜖 ) ∩ (1, ∞) = ∅.
⟹ Any real number < 1 is not a limit point of (1, ∞).
If 𝑥 ∈ [1, ∞), then for every 𝜖 > 0, (𝑥 − 𝜖, 𝑥 + 𝜖 ) contains infinitely
many points of (1, ∞) to the right of 1.
⟹ Every element of [1, ∞) is a limit point of (1, ∞).
∴ (1, ∞)′ = [1, ∞).
(ii) Please try yourself. Answer: (−∞, −1)
Example: 3. Find the derived set of each of the following:

1+(−1)𝑛
𝑆={ : 𝑛 ∈ ℕ}.
𝑛
1+(−1)𝑛
Solution: Let 𝑆 = { : 𝑛 ∈ ℕ}
𝑛
1+(−1)𝑛 1−1
When 𝑛 is odd, = =0
𝑛 𝑛
1+(−1)𝑛 1+1 2
When 𝑛 is even, = =𝑛
𝑛 𝑛
1
Therefore, 𝑆 = {0} ∪ {𝑛 : 𝑛 ∈ ℕ}
⟹ 𝑆 ′ = {0} .
Example: 4. Give one example of each of the following:

(i) an infinite set having no limit point.
(ii) an infinite set having one limit point.
(iii) a set having two limit point.
(iv) a set having an infinite number of limit points.

(v) a set every point of which is a limit point.
(vi) a set with only √2 is a limit point.
Solution:
(i) The set of all natural numbers is an infinite set having no limit point
set.
1
(ii) The set {𝑛 : 𝑛 ∈ ℕ} is an infinite set having only one limit point,
namely 0.
1 1
(iii) The set {𝑛 : 𝑛 ∈ ℕ} ∪ {1 − 𝑛 : 𝑛 ∈ ℕ} has two limit points, namely
0 and 1.
(iv) The sets ℚ, ℝ and (1, 2) have an infinite number of limit points.
(v) Every point of a closed interval [1, 2] is a limit point.
1
(vi) The set {√2 + 𝑛 : 𝑛 ∈ ℕ} has only √2 as a limit point.
Example: 5. Give one example of each of the following:

(i) an unbounded set having limit points.
(ii) a bounded set having limit points.
(iii) an unbounded set having no limit point.
Solution:
(i) The set of all rational numbers is an unbounded set and derived set of
ℚ is ℝ.
(ii) The set [1, 2] is bounded and [1, 2]′ = [1, 2].
(iii) The set ℤ is unbounded and ℤ′ = ∅.
Example: 6. Prove that a finite set has no limit point.

Solution: Let 𝑆 = {𝑝1 , 𝑝2 , … … , 𝑝𝑛 } be a finite subset of ℝ. Let p be any
real number.
If we choose 𝜖 = min. {|𝑝 − 𝑝1 |, |𝑝 − 𝑝2 |, … … , |𝑝 − 𝑝𝑛 | }, then (𝑝 −
𝜖, 𝑝 + 𝜖) is a neighbourhood of p which contains no element of 𝑆, i.e.
(𝑝 − 𝜖, 𝑝 + 𝜖 ) ∩ 𝑆 = ∅
Therefore, p is not a limit point of 𝑆.
Since p is arbitrary, therefore 𝑆 has no limit point.
Some important theorem:

Theorem 1. (Bolzano – Weierstrass Theorem)
Every infinite and bounded subset of ℝ has a limit point.
Proof: Let 𝑆 be an infinite bounded subset of ℝ.
(i) if 𝑆 is bounded ⟹ ∃ real number 𝑘 and K such that 𝑘 ≤ 𝑠 ≤ 𝐾 ∀
𝑠 ∈ 𝑆.
(ii) Let a set T be defined as follows
T = {𝑡: 𝑡 > finitely many elements of 𝑆}
(iii) To prove that T ≠ ∅.
𝑘 ≤ 𝑠 ∀ 𝑠 ∈ 𝑆 ⟹ k is greater than no elements of 𝑆
⟹𝑘∈T ⟹ T ≠ ∅.
(iv) To prove that T is bounded above.
For any 𝜖 > 0, 𝐾 + 𝜖 > 𝐾 ≥ 𝑠 ∀ 𝑠 ∈ 𝑆 ⟹ 𝐾 + 𝜖 ∉ T, 𝐾 ∉ T
⟹ ∀ 𝑡 ∈ 𝑇, 𝑡 < 𝐾 ⟹ T is bounded above.
∵ T is a non-empty bounded subset of ℝ
∴ T has a least upper bound say 𝑢.
(v) To prove that 𝑢 is a limit point of 𝑆.
Let (𝑢 − 𝜖, 𝑢 + 𝜖 ) be any neighbourhood of 𝑢.
𝑢 is least upper bound of T ⟹ ∃ some 𝑡 ∈ T such that 𝑡 > 𝑢 − 𝜖, 𝜖 >
0.
Now 𝑡 ∈ T ⟹ 𝑡 > finitely many elements of 𝑆
⟹ 𝑢 − 𝜖 > finitely many elements of 𝑆
⟹ finitely many elements of 𝑆 lie to the left of 𝑢 − 𝜖
⟹ infinitely many elements of S lie to right of 𝑢 − 𝜖 ……. (1)
Also 𝑢 = 𝑙. 𝑢. 𝑏 𝑜𝑓 𝑇 ⟹𝑢+𝜖∉T
⟹ 𝑢 + 𝜖 > infinitely many elements of S
⟹ infinitely many elements of S lie to left of 𝑢 + 𝜖 ……. (2)
Combining (1) and (2), (𝑢 − 𝜖, 𝑢 + 𝜖) has infinitely many

elements of S. but (𝑢 − 𝜖, 𝑢 + 𝜖) is any neighbourhood of 𝑢.
Therefore, every neighbourhood of 𝑢 has infinitely many elements of S.
Hence 𝑢 is a limit point of S.
Theorem 2. If S is an infinite bounded above subset of ℝ and

𝑢 = 𝑙. 𝑢. 𝑏 𝑜𝑓 S such that 𝑢 ∉ S, then 𝑢 ∈ 𝑆 ′ , i.e. 𝑢 is limit point of S.
Proof: For each 𝜖 > 0, (𝑢 − 𝜖, 𝑢 + 𝜖) is neighbourhood of 𝑢.
Since 𝑢 = 𝑙. 𝑢. 𝑏. of S, 𝑢 − 𝜖 is not an upper bound od S.
Therefore ∃ some 𝑥 ∈ S such that 𝑥 > 𝑢−𝜖
Also 𝑥 <𝑢 < 𝑢+𝜖
Therefore 𝑢−𝜖 <𝑥 < 𝑢+𝜖
⟹ (𝑢 − 𝜖, 𝑢 + 𝜖) ∩ S −{𝑢} contains at least one point 𝑥 of S.
⟹ 𝑢 is a limit point of S.
Note: The derived set of an infinite bounded subset of ℝ is bounded.
3.7 CLOSED SET
Let 𝐴 be a subset of ℝ then 𝐴 is said to be closed set if its complement

𝐴𝑐 = ℝ − 𝐴 is an open set.
i.e. a set is closed if its complements is open.
For example. (i) ℝ𝑐 = ℝ − ℝ = ∅ which is open ⟹ ℝ is closed.
(ii) ∅𝑐 = ℝ − ∅ = ℝ which is open ⟹ ∅ is closed.
Note: ℝ and ∅ are only two sets which are both open and closed.
(iii) [𝑎, 𝑏]𝑐 = (−∞, 𝑎) ∪ (𝑏, ∞) being the union of two open sets is
itself open ⟹ every closed interval [a, b] is a closed set.
Question. Prove that the set ℤ of all integer is a closed set.

Solution. ℤ will be the closed set if ℤ𝑐 is open an open set.
Let 𝑥 ∈ ℤ𝑐 then 𝑥 ∉ ℤ i.e. 𝑥 is not an integer.
|𝑥−𝑛|
If 𝑛 is an integer nearest to 𝑥, then ∃ 𝜖 = > 0 such that
2
(𝑥 − 𝜖, 𝑥 + 𝜖) does not contain any integer

Therefore, (𝑥 − 𝜖, 𝑥 + 𝜖) ⊂ ℤ𝑐 ⟹ ℤ𝑐 is neighbourhood of 𝑥.
But 𝑥 is any point of ℤ𝑐 .
Therefore, ℤ𝑐 is neighbourhood of each of its points.
⟹ ℤ𝑐 is open ⟹ ℤ is closed.
Similarly, the set ℕ of all natural numbers is a closed set.
Question. The set of all rational number is not a closed set.

Solution. Let 𝑥 be any element of ℚ𝑐 = ℝ − ℚ, the set of irrational
numbers.
For every 𝜖 > 0, (𝑥 − 𝜖, 𝑥 + 𝜖) contains infinitely many rational
numbers.
Therefore, ∃ no open interval such that 𝑥 ∈ 𝐼 ⊂ ℚ𝑐
⟹ ℚ𝑐 is not a neighbourhood of 𝑥
⟹ ℚ𝑐 is not an open set.
⟹ ℚ is not a closed set.
Theorem 3.7.1. The union of two closed sets is a closed set.

Proof: Let A and B be two set closed seats.
⟹ 𝐴𝑐 and 𝐵𝑐 are open sets.
⟹ 𝐴𝑐 ∩ 𝐵𝑐 is an open set.
⟹ (𝐴 ∪ 𝐵)𝑐 is an open set.
[since 𝐴𝑐 ∩ 𝐵𝑐 = (𝐴 ∪ 𝐵)𝑐 De Morgan’s Law]
⟹ 𝐴 ∪ 𝐵 is an open set.
Theorem 3.7.2. The union of a finite number of closed sets is a closed
set.
Proof: Let 𝐴1 , 𝐴2 , …, 𝐴𝑛 be 𝑛 closed sets and S = ⋃𝑛𝑖=1 𝐴𝑖
⟹ 𝐴𝑐 1 , 𝐴𝑐 2 ,…, 𝐴𝑐 𝑛 are 𝑛 open sets.
⟹ ⋃𝑛𝑖=1 𝐴𝑐 𝑖 is an open set.
[ since the intersection of a finite collection of open sets is an open set]
⟹ (⋃𝑛𝑖=1 𝐴𝑖 )𝑐 is an open set.
⟹ ⋃𝑛𝑖=1 𝐴𝑖 is a closed set.
Theorem 3.7.3. The union of an infinite family of closed sets need not
be a closed set.
1
Proof: Let 𝐴𝑛 = [𝑛 , 1] ∀ 𝑛 ∈ ℕ, then {𝐴𝑛 }𝑛 ∈ ℕ is an infinite family of
closed sets.
1 1
𝐴1 = {1}, 𝐴2 = [2 , 1], 𝐴3 = [3 , 1], …
1 1
Therefore, ⋃∞
𝑛=1 𝐴𝑛 = {1} ∪ [2 , 1] ∪ [3 , 1] ∪ … = (0, 1] which is not
closed.
Theorem 3.7.4. The intersection of two closed sets is a closed set.
Proof: Let A and B be two set closed seats.
⟹ 𝐴𝑐 and 𝐵𝑐 are open sets.
⟹ 𝐴𝑐 ∪ 𝐵𝑐 is an open set.
⟹ (𝐴 ∩ 𝐵)𝑐 is an open set.
[since 𝐴𝑐 ∪ 𝐵𝑐 = (𝐴 ∩ 𝐵)𝑐 De Morgan’s Law]
⟹ 𝐴 ∪ 𝐵 is closed set.
Note: The intersection of an arbitrary family of closed set is closed.

Note: A set S is said to be perfect if 𝑆 = 𝑆 ′ .
Example: The set of all real number is perfect but the set of all rational
number is not a perfect set.
CHECK YOUR PROGRESS
Problem 1. The set {𝑥: 0 ≤ 𝑥 ≤ 1} 𝑖𝑠 𝑎 𝑐𝑙𝑜𝑠𝑒𝑑 𝑠𝑒𝑡.
Problem 2. The set of natural number is open set.

Problem 3. The set of rational number is closed set.
Problem 4. The set {𝑥: 0 ≤ 𝑥 ≤ 1} is neighbourhood of each of its
point.
Problem 5. The set {𝑥: 0 ≤ 𝑥 ≤ 1} has limit point 0.
3.8 SUMMARY
1. Let 𝑐 ∈ ℝ. A subset 𝑆 ⊂ ℝ is said to be neighbourhood of 𝑐 if
there exist an open interval (𝑎, 𝑏) such that 𝑐 ∈ (𝑎, 𝑏) ⊂ 𝑆.

Clearly an open bounded interval containing the point 𝑐 is a
2. Let 𝑆 be a subset of ℝ. 𝑆 is said to be open set if each point of 𝑆
is an interior point of 𝑆.
3. A set is closed if its complements is open.
4. Bolzano – Weierstrass Theorem:
Every infinite and bounded subset of ℝ has a limit point.
3.9 GLOSSARY
Numbers
Intervals
Sets
Functions
3.10 REFERENCES

2. R.G. Bartle and D.R. Sherbert, Introduction of real analysis (3rd



Q 1. Prove that the set of natural number ℕ is closed set.
Q 2. Prove that the set of rational number ℚ is not closed set.
Q 3. The set {𝑥: 4 ≤ 𝑥 < 8} is neither open nor closed.
Q 4. Prove that intersection of an arbitrary family of closed sets is a closed
set.
Q 5. With the help of examples, prove that intersection of an infinite
family of open sets may or may not be an open set.
3.13 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. False
CYQ 3. False
CYQ 4. True
CYQ 5. True
UNIT 4: REAL SEQUENCES
CONTENTS:
4.1 Introduction
4.2 Objectives
4.3 Real sequence
4.4 Limit of a sequence
4.5 Convergent sequence
4.6 Divergent sequence
4.7 Oscillatory sequence
4.8 Null sequence
4.9 Monotonic sequence
4.10 Limit point of a sequence
4.11 Bolzano – Weierstrass theorem for the sequence
4.12 Limit superior and limit inferior of a sequence
4.13 Subsequence
4.14 Cauchy Sequence
4.15 Summary
4.16 Glossary
4.17 References
4.20 Answers
4.1 INTRODUCTION
So far, we have introduced sets as well as the number systems that

we will use in this text. Next, we will study sequences of numbers.
Sequences are, basically, countably many numbers arranged in an order
that may or may not exhibit certain patterns. Here is the formal definition
of a sequence:
4.2 OBJECTIVES
(i) Neighborhoods
(ii) Interior point
(iii) Open set
(iv) Limit point
4.3 SEQUENCE
A real sequence is a function whose domain is the set ℕ of all natural

numbers and range is a subset of the set ℝ of real numbers.
Symbolically 𝑓: ℕ ⟶ ℝ , {𝑥𝑛 }, {𝑎𝑛 }, 〈𝑏𝑛 〉 etc.
1
Example: 〈𝑛〉 , 〈2𝑛 〉, (−1)𝑛 etc.
Range of sequence: The set of all distinct terms of a sequence is called

its range.
Note: In a sequence {𝑥𝑛 }, since 𝑛 ∈ ℕ and ℕ is an infinite set, the

number of terms of a sequence is always infinite. The range of the
sequence may be a finite set.
Example: if 𝑥𝑛 = (−1)𝑛 , then {𝑥𝑛 } = {−1, 1, −1, 1, … }, the range
of this sequence {𝑥𝑛 } = {−1, 1} which is a finite set.
Type of Sequence:
1. Constant sequence: A sequence {𝑥𝑛 } defined by 𝑥𝑛 = 𝑐 ∈ ℝ ∀

𝑛 ∈ ℕ is called a constant sequence.
Thus {𝑥𝑛 } = {𝑐, 𝑐, 𝑐 … … . . } is a constant sequence with range = {𝑐 }, a
singletion.
2. Bounded above sequence: A sequence {𝑥𝑛 } is said to be
bounded above if ∃ a real number 𝐾 such that
𝑥𝑛 ≤ 𝐾 ∀ 𝑛 ∈ ℕ
i.e. if range of the sequence is bounded above.
3. Bounded Below sequence: A sequence {𝑥𝑛 } is said to be

bounded below if ∃ a real number 𝑘 such that
𝑥𝑛 ≥ 𝑘 ∀ 𝑛 ∈ ℕ
i.e. if range of the sequence is bounded below.
4. Bounded sequence: A sequence {𝑥𝑛 } is said to be bounded if it

is bounded above as well as bounded below.
Thus, a sequence {𝑥𝑛 } is said to be bounded if ∃ a real numbers 𝑘 and 𝐾,
such that
𝑘 ≤ 𝑥𝑛 ≤ 𝐾 ∀ 𝑛 ∈ ℕ
i.e. if range of the sequence is bounded.
5. Unbounded sequence: A sequence {𝑥𝑛 } is said to be unbonded

if it is not bounded.
6. Unbounded above sequence: A sequence {𝑥𝑛 } is said to be

unbounded above if it not bounded above.
i.e. for every real number 𝐾, ∃ 𝑚 ∈ ℕ such that 𝑥𝑚 > 𝐾.
7. Unbounded below sequence: A sequence {𝑥𝑛 } is said to be

unbounded below if it not bounded below.
i.e. for every real number 𝑘, ∃ 𝑚 ∈ ℕ such that 𝑥𝑚 < 𝑘.
1
Example (i) The sequence {𝑥𝑛 } defined by 𝑥𝑛 = 𝑛 is bounded,
since 0 < 𝑥𝑛 ≤ 1.
(ii) The sequence {𝑥𝑛 } defined by 𝑥𝑛 = 𝑛 is bounded below,
because 𝑥𝑛 ≥ 1 ∀ 𝑛 ∈ ℕ.
(iii) The sequence {𝑥𝑛 } defined by 𝑥𝑛 = (−1)𝑛 is bounded,
since 1 ≤ 𝑥𝑛 ≤ 1.
(iv) every constant sequence is bounded.
(v) The sequence {𝑥𝑛 } defined by 𝑥𝑛 = (−1)𝑛 . 𝑛 is neither bounded
below nor bounded above.
Theorem 4.3.1. A sequence {𝑥𝑛 } is bounded iff ∃ a positive real number

𝑀 such that |𝑥𝑛 | ≤ 𝑀 ∀ 𝑛 ∈ ℕ.
Proof: Necessary part
Let {𝑥𝑛 } be bounded. Then ∃ two real number ℎ and 𝑘 such that
ℎ ≤ 𝑥𝑛 ≤ 𝑘 ∀ 𝑛 ∈ ℕ ……. (1)
Let 𝑀 =Max.{|ℎ|, |𝑘|}, then |ℎ| ≤ 𝑀 and |𝑘| ≤ 𝑀
⟹ −𝑀 ≤ ℎ ≤ 𝑀 and −𝑀 ≤ 𝑘 ≤ 𝑀 …… (2)
From (1) and (2), we have −𝑀 ≤ ℎ ≤ 𝑥𝑛 ≤ 𝑘 ≤ 𝑀 ∀ 𝑛 ∈ ℕ
⟹ −𝑀 ≤ 𝑥𝑛 ≤ 𝑀 ∀ 𝑛 ∈ ℕ
⟹ |𝑥 𝑛 | ≤ 𝑀 ∀ 𝑛 ∈ ℕ
Sufficient part
Let 𝑀 be a positive real number such that
|𝑥 𝑛 | ≤ 𝑀 ∀ 𝑛 ∈ ℕ
Then −𝑀 ≤ 𝑥𝑛 ≤ 𝑀 ∀ 𝑛 ∈ ℕ
⟹ {𝑥𝑛 } is bounded.
Note: The above theorem is used as a definition of a bounded sequence

and should be committed to memory.
4.4 LIMIT OF A SEQUENCE
Let {𝑥𝑛 } be a sequence and 𝑙 ∈ ℝ. The real number 𝑙 is said to be limit

of the sequence {𝑥𝑛 } it to each 𝜖 > 0, ∃ 𝑚 ∈ ℕ (𝑚 depending on 𝜖)
such that |𝑥𝑛 − 𝑙 | < 𝜖 ∀ 𝑛 ≥ 𝑚.
If 𝑙 is the limit of the sequence {𝑥𝑛 }, then we write 𝑥𝑛 ⟶ 𝑙 as 𝑛 ⟶ ∞ or
lim 𝑥𝑛 = 𝑙.
𝑛⟶∞
Note: if |𝑥𝑛 − 𝑙 | < 𝜖 ∀ 𝑛≥𝑚

Then 𝑙 − 𝜖 < 𝑥𝑛 < 𝑙 + 𝜖 ∀ 𝑛≥𝑚
Then 𝑥𝑛 ∈ (𝑙 − 𝜖, 𝑙 + 𝜖) ∀ 𝑛 ≥ 𝑚.
4.5 CONVERGENT SEQUENCE
If lim 𝑥𝑛 = 𝑙, then we say that the sequence {𝑥𝑛 } converges to 𝑙.

𝑛⟶∞
Theorem 4.5.1. Every convergent sequence has a unique limit.

Or
A sequence cannot converge to more than one limit.
Proof: if possible, let a sequence {𝑥𝑛 } converse to two distinct real

numbers 𝑙 and 𝑙′.
Let 𝜖 = |𝑙 − 𝑙′|. Since 𝑙 ≠ 𝑙′, |𝑙 − 𝑙′| > 0 so that 𝜖 > 0.
Now the sequence {𝑥𝑛 } converges to 𝑙
⟹ given 𝜖 > 0, ∃ a positive integer 𝑚1 such that
𝜖
|𝑥 𝑛 − 𝑙 | < ∀ 𝑛 ≥ 𝑚1
2
Also, the sequence {𝑥𝑛 } converges to 𝑙′

⟹ given 𝜖 > 0, ∃ a positive integer 𝑚2 such that
𝜖
|𝑥𝑛 − 𝑙′| < ∀ 𝑛 ≥ 𝑚2
2
Let 𝑚 = 𝑚𝑎𝑥. {𝑚1 , 𝑚2 }

𝜖 𝜖
Then |𝑥 𝑛 − 𝑙 | < and |𝑥𝑛 − 𝑙′| < ∀𝑛 ≥𝑚 ……. (1)
2 2
Now |𝑙 − 𝑙′| = |(𝑙 − 𝑥𝑛 ) + (𝑥𝑛 − 𝑙 ′ )| ≤ |𝑙 − 𝑥𝑛 | + |𝑥𝑛 − 𝑙′|

= |𝑥𝑛 − 𝑙 | + |𝑥𝑛 − 𝑙′| (∵ |−𝑥 | = |𝑥 | )
𝜖 𝜖
< + =𝜖 ∀𝑛 ≥ 𝑚
2 2
∴ |𝑙 − 𝑙′| < 𝜖 ∀ 𝑛 ≥ 𝑚
1
Which contradicts the assumption that 𝜖 = 2 |𝑙 − 𝑙′|
⟹ Our assumption was wrong. Hence 𝑙 = 𝑙′.
4.6 DIVERGENT SEQUENCE
(i) A sequence {𝑥𝑛 } is said to be diverge to +∞ if given any

positive real number 𝐾, however large there exists a positive integer
𝑚 (depending on 𝐾) such that 𝑥𝑛 > 𝐾 ∀ 𝑛 ≥ 𝑚
And we write lim 𝑥𝑛 = ∞ or 𝑥𝑛 ⟶ ∞ as 𝑛 ⟶ ∞
𝑛⟶∞
(ii) A sequence {𝑥𝑛 } is said to be diverge to −∞ if given any positive

real number 𝐾, however large there exists a positive integer
𝑚 (depending on 𝐾) such that 𝑥𝑛 < −𝐾 ∀ 𝑛 ≥ 𝑚
And we write lim 𝑥𝑛 = −∞ or 𝑥𝑛 ⟶ ∞ as 𝑛 ⟶ ∞
𝑛⟶∞
(iii) A sequence {𝑥𝑛 } is said to be divergent sequence if it diverse to

+∞ or −∞.
i.e. if 𝑥𝑛 ⟶ ∞ or 𝑥𝑛 ⟶ −∞ as 𝑛 ⟶ ∞.
Example. The sequence {𝑛} and {𝑛2 } diverge to +∞. Whereas the
sequences {−𝑛} and {−𝑛2 } diverge to −∞.
4.7 OSCILLATORY SEQUENCE
If a sequence {𝑥𝑛 } neither converges to a finite number nor diverges to

+∞ or −∞, it is called an oscillatory sequence. Oscillatory sequences
are of two types:
(i) A bounded sequence which do not converse is said to be oscillate
finitely.
For example: Consider a bounded sequence {(−1)𝑛 }.
Here 𝑥𝑛 = (−1)𝑛
Then lim 𝑥2𝑛 = lim (−1)2𝑛 = 1 and
𝑛⟶∞ 𝑛⟶∞
lim 𝑥2𝑛+1 = lim (−1)2𝑛+1 = −1.

𝑛⟶∞ 𝑛⟶∞
Thus lim 𝑥𝑛 does not exist

𝑛⟶∞
⟹ the sequence does not converse.

Hence this sequence oscillates finitely.
(ii) An unbounded sequence which do not diverge is said to be oscillate
infinitely.
For example: Consider a bounded sequence {(−1)𝑛 . 𝑛}.
Here 𝑥𝑛 = (−1)𝑛 . 𝑛
Then lim 𝑥2𝑛 = lim (−1)2𝑛 . 2𝑛 = +∞ and
𝑛⟶∞ 𝑛⟶∞
lim 𝑥2𝑛+1 = lim (−1)2𝑛+1 . (2𝑛 + 1) = −∞.

𝑛⟶∞ 𝑛⟶∞
Thus, the sequence does not diverge.

⟹ this sequence oscillates infinitely.
Note: When we say lim 𝑥𝑛 = 𝑙, it means lim 𝑥2𝑛 = lim 𝑥2𝑛+1 = 𝑙

𝑛⟶∞ 𝑛⟶∞ 𝑛⟶∞
Similarly, lim 𝑥𝑛 = +∞, it means lim 𝑥2𝑛 = lim 𝑥2𝑛+1 = +∞

𝑛⟶∞ 𝑛⟶∞ 𝑛⟶∞
4.8 NULL SEQUENCE
A sequence {𝑥𝑛 } is said to be null sequence if it converges to zero

i.e. if lim 𝑥𝑛 = 0.
𝑛⟶∞
1 1 1 (−1)𝑛−1
For example: The sequence {𝑛}, {𝑛2}, {2𝑛 } and { } are null
𝑛
sequences.
4.9 MONOTONIC SEQUENCE
(i) A sequence {𝑥𝑛 } is said to be monotonically increasing.

If 𝑥𝑛+1 ≥ 𝑥𝑛 ∀ 𝑛 ∈ ℕ
i.e. if 𝑥1 ≤ 𝑥2 ≤ 𝑥3 ≤ ⋯ ≤ 𝑥𝑛 ≤ 𝑥𝑛+1 ≤ ⋯
(ii) A sequence {𝑥𝑛 } is said to be monotonically decreasing.
If 𝑥𝑛+1 ≤ 𝑥𝑛 ∀ 𝑛 ∈ ℕ
i.e. if 𝑥1 ≥ 𝑥2 ≥ 𝑥3 ≥ ⋯ ≥ 𝑥𝑛 ≥ 𝑥𝑛+1 ≥ ⋯
(iii) A sequence {𝑥𝑛 } is said to be monotonic if it is either monotonically
increasing or monotonically decreasing.
(iv) A sequence {𝑥𝑛 } is said to be strictly monotonically increasing.
If 𝑥𝑛+1 > 𝑥𝑛 ∀ 𝑛 ∈ ℕ
(v) A sequence {𝑥𝑛 } is said to be strictly monotonically decreasing.
If 𝑥𝑛+1 < 𝑥𝑛 ∀ 𝑛 ∈ ℕ
(vi) A sequence {𝑥𝑛 } is said to be strictly monotonic if it is either
strictly monotonically increasing or strictly monotonically decreasing.
Theorem 4.9.1. Every convergent sequence is bounded.

Proof: Let {𝑥𝑛 } be a convergent sequence, converging to 𝑙.
For 𝜖 = 1, there exists a positive integer 𝑚 such that
|𝑥 𝑛 − 𝑙 | < 1 ∀𝑛≥𝑚
⟹ 𝑙 − 1 < 𝑥𝑛 < 𝑙 + 1 ∀𝑛 ≥ 𝑚
Let 𝑘 = min.{𝑥1, 𝑥2 , … , 𝑥𝑚−1, 𝑙 − 1} and

𝐾 = mix.{𝑥1, 𝑥2 , … , 𝑥𝑚−1, 𝑙 + 1}
Then 𝑘 ≤ 𝑥𝑛 ≤ 𝐾 ∀𝑛≥𝑚
⟹ The sequence {𝑥𝑛 } is a bounded sequence.
Monotone convergence theorem.

Theorem 4.9.2. Every monotonically increasing sequence which is
bounded above converge to its least upper bound.
Proof: Let {𝑥𝑛 } be a monotonically increasing sequence which is
bounded above. Let 𝑢 be the 𝑙. 𝑢. 𝑏. of the sequence {𝑥𝑛 }.
We shall show that {𝑥𝑛 } converges to 𝑢.
Let 𝜖 > 0 be given.
Since 𝑢 − 𝜖 < 𝑢, therefore, 𝑢 − 𝜖 is not an upper bound of {𝑥𝑛 }
⟹ ∃ a positive integer 𝑚 such that 𝑥𝑛 > 𝑢 − 𝜖
Since {𝑥𝑛 } is monotonically increasing.
Therefore 𝑥𝑛 ≥ 𝑥𝑚 > 𝑢 − 𝜖 ∀𝑛≥𝑚 ……. (i)
Then 𝑥𝑛 > 𝑢 − 𝜖 ∀𝑛≥𝑚
Also, 𝑢 is the 𝑙. 𝑢. 𝑏. of the sequence {𝑥𝑛 }
⟹ 𝑥𝑛 ≤ 𝑢 + 𝜖 ∀𝑛∈ℕ
⟹ 𝑥𝑛 < 𝑢 + 𝜖 ∀𝑛∈ℕ ……. (ii)
From (1) and (2), 𝑢 − 𝜖 < 𝑥𝑛 < 𝑢 + 𝜖 ∀𝑛≥𝑚
⟹ |𝑥 𝑛 − 𝑢 | < 𝜖 ∀𝑛≥𝑚
⟹ lim 𝑥𝑛 = 𝑢
𝑛⟶∞
⟹ {𝑥𝑛 } converges to 𝑢.
Theorem 4.9.3. Every monotonically decreasing sequence which is
bounded below converges to its greatest lower bound.
Proof: Similar as Theorem 4.9.2.
Note: Every monotonically increasing sequence which is not bounded

above diverges to ∞.
Note: Every monotonically decreasing sequence which is not bounded
below diverges to −∞.
Note: Every monotonic sequence either converges or diverges.
Examples 4.9.1. By definition show that

1
(i) The sequence {𝑛} converges to 0.
1
(ii) The sequence {𝑛2} converges to 0.
1
(iii) The sequence {3𝑛 } converges to 0.
(iv) The sequence {√𝑛 + 1 − √𝑛} is a null sequence.

1
Sol. (i) let 𝑎𝑛 = 𝑛 . let 𝜖 > 0 be given.
1 1 1 1
Now |𝑎𝑛 − 0| = | − 0| = | | = < 𝜖 if 𝑛 >
𝑛 𝑛 n 𝜖
1
If 𝑚 is a positive integer > 𝜖 , then |𝑎𝑛 − 0| < 𝜖 ∀ 𝑛 ≥ 𝑚
⟹ lim 𝑎𝑛 = 0
𝑛⟶∞
1
⟹ i.e. sequence {𝑛} converges to 0.
1
(ii) let 𝑎𝑛 = 𝑛2 . let 𝜖 > 0 be given.
1 1 1 1
Now | 𝑎𝑛 − 0 | = | − 0| = |𝑛2| = 𝑛2 < 𝜖 if 𝑛2 > 𝜖
𝑛2
1
If 𝑚 is a positive integer > , then |𝑎𝑛 − 0| < 𝜖 ∀ 𝑛 ≥ 𝑚
√𝜖
𝑛⟶∞
1
⟹ i.e. sequence {𝑛2} converges to 0.
1
(iii) let 𝑎𝑛 = 3𝑛 . let 𝜖 > 0 be given.
1 1 1 1
Now | 𝑎𝑛 − 0 | = | − 0| = |3𝑛 | = 3𝑛 < 𝜖 if 3𝑛 > 𝜖
3𝑛
1
i.e. if 𝑛 𝑙𝑜𝑔𝑛 > 𝑙𝑜𝑔 (𝜖 )
1
𝑙𝑜𝑔( )
𝜖
i.e. if 𝑛 > 𝑙𝑜𝑔(3) [since 𝑙𝑜𝑔(3) > 0]
1
𝑙𝑜𝑔( )
𝜖
If 𝑚 is a positive integer > 𝑙𝑜𝑔(3) , then |𝑎𝑛 − 0| < 𝜖 ∀ 𝑛 ≥ 𝑚
𝑛⟶∞
1
⟹ i.e. sequence {3𝑛 } converges to 0.
(iv) let 𝑎𝑛 = √𝑛 + 1 − √𝑛
(√𝑛−1−√𝑛)(√𝑛+1+√𝑛) (n+1)−n 1
= = =
√𝑛+1+√𝑛 √𝑛+1+√𝑛 √𝑛+1+√𝑛
let 𝜖 > 0 be given.
1 1 1 1 1
Now |𝑎𝑛 − 0| = | − 0| = | |= <2 < <𝜖
√𝑛+1+√ 𝑛 √𝑛+1+√𝑛 √𝑛+√𝑛 √𝑛 √𝑛
1 1
if √𝑛 > 𝜖 i.e. if 𝑛 > 𝜖2
1
If 𝑚 is a positive integer > 𝜖2 , then |𝑎𝑛 − 0| < 𝜖 ∀ 𝑛 ≥ 𝑚
𝑛⟶∞
⟹ The given sequence {√𝑛 + 1 − √𝑛} is a null sequence.
4.10 LIMIT POINT OF A SEQUENCE
A real number 𝑙 is said to be a limit point of a sequence {𝑎𝑛 } if

every neighbourhood of 𝑙 contains infinitely many terms of the sequence.
Note: If 𝑎𝑛 = 𝑙 for infinitely many values of 𝑛 then 𝑙 is a limit point of
{𝑎𝑛 }.
Note: Limit of a sequence is a limit point but limit point need not be a
limit of sequence.
Note: Limit point of a sequence need not be a term of the sequence.
1 1
For example: Sequence {𝑛} has 0 as a limit point but no term of {𝑛} is 0.
1
Example 4.10.1. Prove that 0 is a limit point of the Sequence {𝑛}.
1
Sol. For 𝜖 > 0, ∃ 𝑚 ∈ ℕ such that < 𝜖
𝑚
1 1
Therefore, for 𝑛 ≥ 𝑚, 0<𝑛≤𝑚<𝜖
1
⟹ −𝜖 < 0 < 𝑛 < 𝜖 ∀ 𝑛 ≥ 𝑚
1
⟹ ∈ (−𝜖, 𝜖) ∀𝑛≥𝑚
𝑛
⟹ Every neighbourhood of 0 contains infinitely many terms of the

1
sequence {𝑛}.
1
⟹ 0 is a limit point of the sequence {𝑛}.
Example 4.10.2. The sequence {(−1)𝑛 } has two limit points.

Sol. Let 𝑎𝑛 = (−1)𝑛 , then 𝑎𝑛 = −1 when 𝑛 is odd and 𝑎𝑛 = 1 when 𝑛
is even.
Thus, every neighbourhood of −1 contains all the odd terms of
the sequence.
Therefore, −1 is a limit point.

Also, every neighbourhood of 1 contains all the even terms of the
sequence.
Therefore, 1 is a limit point.
Therefore, the given sequence has only two limit points.
Example 4.10.3. The sequence {𝑛} has no limit points.

1 1
Sol. Let 𝑙 be any real number, then the neighbourhood (𝑙 − 4 , 𝑙 + 4) of 𝑙
contains at most one term of the sequence {𝑛}.

⟹ 𝑙 is not a limit point of the sequence {𝑛}.
Theorem 4.10.1. If 𝑙 is a limit point of the range of a sequence {𝑎𝑛 } , then

𝑙 is a limit point of the sequence {𝑎𝑛 }.
Proof: Let S = range of the sequence {𝑎𝑛 }.
Since 𝑙 is a limit point of S, every neighbourhood of 𝑙 contains infinitely
many elements of S.
But each element of S is a term of the sequence {𝑎𝑛 }.
∴ Every neighbourhood of 𝑙 contains infinitely many terms of the
sequence {𝑎𝑛 }.
⟹ 𝑙 is a limit point of the sequence {𝑎𝑛 }.
Note: The converse of above theorem may not be true.
0, 𝑤ℎ𝑒𝑛 𝑛 𝑖𝑠 𝑜𝑑𝑑
Consider 𝑎𝑛 = 1 + (−1)𝑛 = {
2, 𝑤ℎ𝑒𝑛 𝑛 𝑖𝑠 𝑒𝑣𝑒𝑛
Therefore, 0, 2 are the limit points of the sequence {𝑎𝑛 }.
But the range of the sequence = {0, 2} is a finite set.
Since the finite set has no limit point, the range of {𝑎𝑛 } has no limit point.
Note: If the terms of the sequence are different the limit points of the
sequence are the limit points of the range set.
4.11 BOLZANO – WEIERSTRASS THEOREM

FOR SEQUENCES
Every Bounded sequence has at least one limit point.

Proof: Let {𝑎𝑛 } be a bounded sequence and S be its range,
i.e. S = {𝑎𝑛 : 𝑛 ∈ ℕ}
since {𝑎𝑛 } is bounded, then S is bounded.
Case 1. Let S be a finite set.
Then ∃ a real number 𝑙 such that 𝑎𝑛 = 𝑙 for an infinite number of values
of n ∈ ℕ.
⟹ Given 𝜖 > 0, 𝑎𝑛 ∈ (𝑙 − 𝜖, 𝑙 + 𝜖) for an infinite number of values of
n ∈ ℕ.
⟹ Every neighbourhood of 𝑙 contains infinitely many terms of the
Therefore 𝑙 is limit point of the sequence {𝑎𝑛 }.
Case 2. Let S be an infinite set.
Since S is an infinite bounded set, by Bolzano-Weierstrass theorem, S has
at least one limit point, say 𝑙.
Now, 𝑙 is limit point of S.
⟹ Every neighbourhood of 𝑙 contains infinitely many numbers of the
elements of S.
But each element of S is a term of the sequence {𝑎𝑛 }.
∴ every neighbourhood of 𝑙 contains an infinite number of terms of the
Therefore 𝑙 is a limit point of the sequence {𝑎𝑛 }.
Corollary: If S is a closed and bounded set, then every sequence in S has

a limit point in S.
Corollary: The set of the limit point of the bounded sequence is bounded.
Corollary: The set of the limit points of a sequence is a closed set.
Corollary: The set of limit points of an unbounded sequence may or may
not be unbounded.
1 1
For example. (i) The sequence {1, , 2, , 3, … } is unbounded but the
2 3
set of its limit points is {0}, which is bounded.
1 1 1 1 1
(i) The sequence {2, 1 + 2 , 2 + 2 , 1 + 3 , 2 + 3 , 3 + 3 , … … } is
unbounded but the set of its limit points is ℕ, which is unbounded.
4.12 LIMIT SUPERIOR AND LIMIT INFERIOR

OF A SEQUENCES
Let {𝑎𝑛 } be a bounded sequence, then the sequence has the least and
greatest limit points.
The least limit point of {𝑎𝑛 } is called the limit inferior of {𝑎𝑛 } and is
denoted by lim 𝑖𝑛𝑓. 𝑎𝑛 .
𝑛⟶∞
The greatest limit point of {𝑎𝑛 } is called the limit superior of {𝑎𝑛 } and is
denoted by lim 𝑠𝑢𝑝. 𝑎𝑛 .
𝑛⟶∞
Note 1. If {𝑎𝑛 } is unbounded above then lim 𝑠𝑢𝑝. 𝑎𝑛 = ∞.

𝑛⟶∞
And If {𝑎𝑛 } is unbounded below then lim 𝑖𝑛𝑓. 𝑎𝑛 = −∞.

𝑛⟶∞
Note 2. Since the greatest limit point of the sequence {𝑎𝑛 } ≥ least limit
point.
Therefore, lim 𝑠𝑢𝑝. 𝑎𝑛 ≥ lim 𝑖𝑛𝑓. 𝑎𝑛 .
𝑛⟶∞ 𝑛⟶∞
Example 4.12.1. Find lim 𝑠𝑢𝑝. 𝑎𝑛 and lim 𝑖𝑛𝑓. 𝑎𝑛 for the sequence
𝑛⟶∞ 𝑛⟶∞
{𝑎𝑛 } = {(−1)𝑛 }.
Sol. Since the sequence {(−1)𝑛 } has only two limit points −1 and 1.
Therefore, the set of limit points = {−1, 1} which is bounded.
Therefore, lim 𝑖𝑛𝑓. 𝑎𝑛 = −1 and lim 𝑠𝑢𝑝. 𝑎𝑛 = 1.
𝑛⟶∞ 𝑛⟶∞
Example 4.12.2. Find lim 𝑠𝑢𝑝. 𝑎𝑛 and lim 𝑖𝑛𝑓. 𝑎𝑛 for the sequence
𝑛⟶∞ 𝑛⟶∞
𝑛
{𝑎𝑛 } = {(−1) . 𝑛}.
Sol. Since the sequence {(−1)𝑛 . 𝑛} is unbounded above and unbounded
below both.
Therefore, lim 𝑖𝑛𝑓. 𝑎𝑛 = −∞ and lim 𝑠𝑢𝑝. 𝑎𝑛 = ∞.
𝑛⟶∞ 𝑛⟶∞
4.13 SUBSEQUENCE
Let {𝑎𝑛 } be a sequence of real numbers, and

let 𝑛1 < 𝑛2 < · · · < 𝑛𝑘 < · · ·
be a strictly increasing sequence of natural numbers.
Then {𝑎𝑛1 , 𝑎𝑛2 , · · · , 𝑎𝑛𝑘 , · · ·} is called a subsequence of {𝑎𝑛 } and is
denoted by {𝑎𝑛𝑘 }.
Note: A subsequence is formed from a sequence by selecting certain terms
from the sequence in order.
For Example. (i) {𝑎2𝑛 }, {𝑎2𝑛−1 }, {𝑎𝑛2 } are all subsequence of {𝑎𝑛 }.
(ii) {2, 4, 6, ….}, {1, 3, 5, …}, {1, 4, 9, 16, …} are all subsequences of
the sequence {n}.
Note: Every sequence is a subsequence of itself.
Theorem 4.13.1. If a sequence {𝑎𝑛 } converges to 𝑙, then every

subsequence of {𝑎𝑛 } also converges to 𝑙.
Proof: Let {𝑎𝑛𝑘 } be a subsequence of {𝑎𝑛 }.
Since {𝑎𝑛 }converges to 𝑙.
Therefore, given 𝜖 > 0, ∃ a positive integer 𝑚
such that |𝑎𝑛 − 𝑙 | < 𝜖 ∀ 𝑛 ≥ 𝑚 ……. (i)
We can find a natural number 𝑛𝑘0 ≥ 𝑚
If 𝑛𝑘 ≥ 𝑛𝑘0 , then 𝑛𝑘 ≥ 𝑚.
Therefore, from (i), we have |𝑎𝑛𝑘 − 𝑙| < 𝜖 ∀ 𝑛𝑘 ≥ 𝑚
Hence {𝑎𝑛𝑘 }converges to 𝑙.
4.14 CAUCHY SEQUENCES
A sequence {𝑎𝑛 } is said to be a Cauchy sequence if given 𝜖 > 0, however

small, there exists a positive integer 𝑚 (depending on 𝜖) such that
|𝑎𝑛 − 𝑎𝑚 | < 𝜖 ∀ 𝑛 ≥ 𝑚.
Theorem 4.14.1. Every Cauchy sequence is bounded.

Proof: Let {𝑎𝑛 } be a Cauchy sequence.
Taking 𝜖 = 1, by definition, there exists a positive integer 𝑚 such that

| 𝑎𝑛 − 𝑎𝑚 | < 1 ∀𝑛≥𝑚
⟹ 𝑎𝑚 − 1 < 𝑎𝑛 < 𝑎𝑚 + 1 ∀𝑛≥𝑚
Let 𝑘 = 𝑚𝑖𝑛. { 𝑎1 , 𝑎2 , … , 𝑎𝑚−1 , 𝑎𝑚 − 1}
And 𝐾 = 𝑚𝑎𝑥. { 𝑎1 , 𝑎2 , … , 𝑎𝑚−1 , 𝑎𝑚 + 1}
Then 𝑘 ≤ 𝑎𝑛 ≤ 𝐾 ∀𝑛
Hence the sequence {𝑎𝑛 } is bounded.
Note: The converse of above theorem is not true, i.e. every bounded
sequence need not be a Cauchy sequence.
For Example. The sequence {(−1)𝑛 } is bounded but it is not a Cauchy
sequence.
Theorem 4.14.2. (Cauchy’s convergence criterion)

A sequence is convergent if and only if it is a Cauchy sequence.
Proof: First, let {𝑎𝑛 } be a convergent sequence, converging to 𝑙.
We shall show that it is a Cauchy sequence.
Let 𝜖 > 0 be given. Then there exists a positive integer m such that
𝜖
| 𝑎𝑛 − 𝑙 | < ∀𝑛≥𝑚 ….. (i)
2
𝜖
In particular, for 𝑛 = 𝑚, we have |𝑎𝑛 − 𝑙 | < 2 …… (ii)
Now, |𝑎𝑛 − 𝑎𝑚 | = |(𝑎𝑛 − 𝑙) − (𝑎𝑚 − 𝑙)| ≤ |(𝑎𝑛 − 𝑙)| + |(𝑎𝑚 − 𝑙)|

𝜖 𝜖
< 2 + 2 = 𝜖. ∀ 𝑛 ≥ 𝑚 [using (i) and (ii)]
Thus | 𝑎𝑛 − 𝑎𝑚 | < 𝜖 ∀ 𝑛 ≥ 𝑚
⟹ {𝑎𝑛 } is a Cauchy sequence.
Conversely: Let {𝑎𝑛 } is a Cauchy sequence.
Since Every Cauchy sequence is bounded, therefore {𝑎𝑛 } is bounded.
Since every bounded sequence has a limit point, {𝑎𝑛 } has limit point 𝑙
(say).
We shall show that {𝑎𝑛 } converges to 𝑙.
Let 𝜖 > 0 be given. Since {𝑎𝑛 } is Cauchy sequence, ∃ a positive integer
𝜖
𝑚 such that | 𝑎𝑛 − 𝑎𝑚 | < ∀𝑛≥𝑚 …… (iii)
3
Since 𝑙 is limit point of {𝑎𝑛 }, every neighbourhood of 𝑙 contains

infinitely many terms of {𝑎𝑛 }.
𝜖 𝜖
⟹ 𝑎𝑛 ∈ (𝑙 − 3 , 𝑙 + 3) for infinitely many values of 𝑛.
In particular, we can find a positive integer 𝑘 > 𝑚 such that

𝜖 𝜖
𝑎𝑘 ∈ (𝑙 − 3 , 𝑙 + 3) ….... (iv)
𝜖
i.e. | 𝑎𝑘 − 𝑙 | < ……. (v)
3
𝜖
Also, since 𝑘 > 𝑚, therefore, we have |𝑎𝑛 − 𝑎𝑚 | < 3 …… (vi)
Now, |𝑎𝑛 − 𝑙 | = |(𝑎𝑛 − 𝑎𝑚 ) + (𝑎𝑚 − 𝑎𝑘 ) + (𝑎𝑘 − 𝑙)|

≤ | 𝑎 𝑛 − 𝑎𝑚 | + | 𝑎𝑚 − 𝑎𝑘 | + | 𝑎𝑘 − 𝑙 |
= | 𝑎𝑛 − 𝑎𝑚 | + | 𝑎𝑘 − 𝑎𝑚 | + | 𝑎𝑘 − 𝑙 |
𝜖 𝜖 𝜖
<3+3+3 =𝜖
Thus, | 𝑎𝑛 − 𝑙 | < 𝜖 ∀𝑛≥𝑚

⟹ {𝑎𝑛 } converges to 𝑙.
Example. 4.14.1. Prove that the sequence whose nth terms are given
below are Cauchy sequence.
1 𝑛 (−1)𝑛
(i) 𝑛 (ii) 𝑛+1 (iii) 𝑛
1
Sol. (i) Here 𝑎𝑛 = 𝑛
𝜖 > 0 be given and let 𝑛 > 𝑚.

1 1 1 1
Now | 𝑎𝑛 − 𝑎𝑚 | = | − | = −
𝑛 𝑚 𝑚 𝑛
1 1
< 𝑚 < 𝜖 whenever 𝑚 > 𝜖
∴ For each 𝜖 > 0, ∃ a +𝑣𝑒 integer 𝑚 such that |𝑎𝑛 − 𝑎𝑚 | < 𝜖 ∀ 𝑛 > 𝑚.
𝑛
(ii) Here 𝑎𝑛 = 𝑛+1

𝑛 𝑚 1 1
Now | 𝑎𝑛 − 𝑎𝑚 | = | − 𝑚+1| = |(1 − 𝑛+1) − (1 − 𝑚+1)|
𝑛+1
1 1 1 1
= |𝑛+1 − 𝑚+1| = 𝑚+1 − 𝑛+1
1 1 1
< 𝑚+1 < 𝑚 < 𝜖 whenever 𝑚 > 𝜖
(−1)𝑛
(iii) Here 𝑎𝑛 = 𝑛

(−1)𝑛 (−1)𝑚 (−1)𝑛 (−1)𝑚
Now | 𝑎𝑛 − 𝑎𝑚 | = | − |≤| |+| |
𝑛 𝑚 𝑛 𝑚
1 1
=𝑛+𝑚
1 1 2 2
< 𝑚 + 𝑚 < 𝑚 < 𝜖 whenever 𝑚 > 𝜖
CHECK YOUR PROGRESS
1
Problem 1. Limit superior and inferior are equal for sequence {𝑛}.
1
Problem 2. The limit point of the sequence {10 + 𝑛} is 10.
Problem 3. The sequence {1 + sin 𝑛} is unbounded.

Problem 4. The set of all limit point of a sequence is open set.
1 1 1 1
Problem 5. The Sequence 1 + 1! + 2! + 3! + ⋯ + 𝑛! is convergent.
4.15 SUMMARY
1. Range of sequence: The set of all distinct terms of a sequence is

called its range.
2. The sequence {𝑛} has no limit points.
3. A sequence {𝑥𝑛 } is said to be null sequence if it converges to zero
i.e. if lim 𝑥𝑛 = 0.
𝑛⟶∞
4. A sequence is convergent if and only if it is a Cauchy sequence.

5. Bolzano – Weierstrass Theorem for a sequence:
Every Bounded sequence has at least one limit point.
4.16 GLOSSARY
Numbers
Intervals
Limit points
Functions
Bounded, Unbounded sets
4.17 REFERENCES



1 𝑛+1
Q1. Prove that the sequence (1 + 𝑛) is convergent.
1 1 1 1
Q2. Prove that the sequence { + 𝑛+2 + 𝑛+3 + ⋯ + 2𝑛} is
𝑛+1
monotonically increasing.
Q3. If {𝑎𝑛 } and {𝑏𝑛 } are null sequences, show that {𝑎𝑛 + 𝑏𝑛 } is also null
sequence.
Q4. If {𝑎𝑛 } and {𝑏𝑛 } are null sequences, show that {𝑎𝑛 . 𝑏𝑛 } is also null
sequence.
Q5. States and prove Bolzano – Weierstrass Theorem for a sequence.
4.20 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. True
CYQ 3. False
CYQ 4. False
CYQ 5. True
UNIT 5: SERIES
5.1 Introduction
5.2 Objectives
5.3 Series or real number
5.4 Positive terms series
5.5 Comparison test
5.6 D’Alembert’s ratio test
5.7 Cauchy’s roots test
5.8 Alternating Series
5.9 Absolute and Conditional Convergence
5.10 Summary
5.11 Glossary
5.12 Suggested Readings
5.13 References
5.14 Terminal Questions
5.15 Answers
5.1 INTRODUCTION
In mathematics, a series is, roughly speaking, the operation
of adding infinitely many quantities, one after the other, to a given starting
quantity. The study of series is a major part of calculus and its
generalization, mathematical analysis. Series are used in most areas of
mathematics, even for studying finite structures (such as in combinatorics)
through generating functions. In addition to their ubiquity in mathematics,
infinite series are also widely used in other quantitative disciplines such
as physics, computer science, statistics and finance.
For a long time, the idea that such a potentially infinite summation could
produce a finite result was considered paradoxical. This paradox was resolved
using the concept of a limit during the 17th century. Zeno's
paradox of Achilles and the tortoise illustrates this counterintuitive property
of infinite sums: Achilles runs after a tortoise, but when he reaches the
position of the tortoise at the beginning of the race, the tortoise has reached a
second position; when he reaches this second position, the tortoise is at a third
position, and so on. Zeno concluded that Achilles could never reach the
tortoise, and thus that movement does not exist. Zeno divided the race into
infinitely many sub-races, each requiring a finite amount of time, so that the
total time for Achilles to catch the tortoise is given by a series. The resolution
of the paradox is that, although the series has an infinite number of terms, it
has a finite sum, which gives the time necessary for Achilles to catch up with
the tortoise.
A brief introduction to Infinite series and some results in infinite series will
be discussed.
5.2 OBJECTIVES
After reading this unit learners will be able to
 infinite series
 positive term series
5.3 SERIES
The sum of the terms of a sequence is said to be a series. Thus if
𝑦1 , 𝑦2 , 𝑦3 , ….is a sequence then the sum 𝑦1 + 𝑦2 + 𝑦3 + ⋯ of all the terms is
called an infinite series and is expressed by ∑∞ 𝑛=1 𝑦𝑛 or ∑ 𝑦𝑛 .
Evidently, we cannot just add up all the infinite number of terms of the series
in ordinary way and in fact it is not obvious that this kind of sum has any
meaning. Therefore, we start by associating with the given series, a sequence
{𝑆𝑛 }, where 𝑆𝑛 denotes the sum of the first 𝑛 terms of the series.
Hence 𝑆𝑛 = 𝑦1 + 𝑦2 + ⋯ . +𝑦𝑛 ∀𝑛
And this sequence {𝑆𝑛 } is said to be the sequence of partial sums of the series.
The partial sums
𝑺𝟏 = 𝒚𝟏 ; 𝑺𝟐 = 𝒚𝟏 + 𝒚𝟐 ; 𝑺𝟑 = 𝒚𝟏 + 𝒚𝟐 + 𝒚𝟑 + ⋯ .. and so on.
The series is convergent if the sequence {𝑆𝑛 } of partial sums converges and
lim 𝑆𝑛 is called the sum of the series.
If {𝑆𝑛 } does not tend to a limit then the sum of the infinite series does not exist
or we can say that the series does not converges.
An infinite series is converge, diverge or oscillate according as its sequence

of partial sums {𝑆𝑛 } converges, diverges and oscillates.
Necessary condition of convergences of an infinite series

Theorem: A Necessary condition of convergences of an infinite series
∑ 𝒚𝒏 is lim 𝑦𝑛 = 0.
𝑛→∞
Proof. Let 𝑆𝑛 = 𝑦1 + 𝑦2 + ⋯ … … . +𝑦𝑛 , so that {𝑆𝑛 } is the sequence of

partial sums.
It is given that series is converges
Thus, the sequence {𝑆𝑛 } is also converges.
Let lim 𝑆𝑛 = 𝑡. Now 𝑦𝑛 = 𝑆𝑛 − 𝑆𝑛−1 , 𝑛 > 1.

𝑛→∞
Therefore, lim 𝑦𝑛 = lim (𝑆𝑛 − 𝑆𝑛−1 ) = lim 𝑆𝑛 − lim 𝑆𝑛−1 = 𝑡 − 𝑡 = 0

𝑛→∞ 𝑛→∞ 𝑛→∞ 𝑛→∞
Thus lim 𝑦𝑛 = 0
𝑛→∞
NOTE:
A series cannot converge if 𝑛𝑡ℎ term does not tend to zero.
Cauchy’s General Principle of Convergence for Series
Theorem: A necessary and sufficient condition for the convergence of an

infinite series ∑ 𝒚𝒏 is that the sequence of its partial sums {𝑺𝒏 } is
convergent
Or
An infinite series ∑ 𝒚𝒏 converges iff for every 𝜺 > 𝟎 there exists a positive
integer M such that |𝒚𝟏 + 𝒚𝟐 + 𝒀𝟑 + ⋯ … + 𝒚𝒏 | < 𝜺 whenever
𝒎≥𝒏≥𝑴
Proof. Let 𝑆𝑛 = ∑ 𝑦𝑛 = 𝑦1 + 𝑦2 + 𝑦3 + ⋯ … + 𝑦𝑛 and 𝑆𝑚 = ∑ 𝑦𝑚 = 𝑦1 +

𝑦2 + 𝑦3 + ⋯ … + 𝑦𝑚 be the 𝑛𝑡ℎ and 𝑚𝑡ℎ partial sum of series respectively,
where 𝑚 ≥ 𝑛.
⇒ |𝑆𝑚 − 𝑆𝑛 | = |(𝑦1 + 𝑦2 + 𝑦3 + ⋯ + 𝑦𝑚 ) − (𝑦1 + 𝑦2 + 𝑦3 + ⋯ + 𝑦𝑛 )|
= |𝑦𝑚+1 + 𝑦𝑚+2 + ⋯ + 𝑦𝑛 | .
Let 𝜀 > 0 and for every 𝜀 the series ∑ 𝑦𝑛 converges iff the sequence of partial
sums {Sn} converges
⇔|𝑆𝑚 − 𝑆𝑛 | < 𝜀 ∀ 𝑚 ≥ 𝑛 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑀 ∈ ℕ
⇔|𝑦𝑚+1 + 𝑦𝑚+2 + ⋯ + 𝑦𝑛 | < 𝜀 ∀ 𝑚 ≥ 𝑛 𝑓𝑜𝑟 𝑠𝑜𝑚𝑒 𝑀 ∈ ℕ

1
Example: Prove that ∑ 𝑛 does not converge.
Proof. Let the given series be converges.
Therefore, for any given 𝜀 > 0, there exists a positive integer 𝑚 such that
1 1 1
| + 𝑛+2 + ⋯ + 𝑛+𝑝| < 𝜀 ∀ 𝑛 ≥ 𝑚 𝑎𝑛𝑑 𝑝 ≥ 1.
𝑛+1
If 𝑛 = 𝑚 and 𝑝 = 𝑚, we get
1 1 1 1 1 1
+ + ⋯+ = + +⋯+
𝑛+1 𝑛+2 𝑛+𝑝 𝑚+1 𝑚+2 𝑚+𝑚
1 1 1
= + + ⋯+
𝑚+1 𝑚+2 2𝑚
1 1
> 𝑚. 2𝑚 > 2 > 𝜀
1 1 1
i.e. 𝑛+1 + 𝑛+2 + ⋯ + 𝑛+𝑝 > 𝜀, a contradiction.
1
Therefore ∑ 𝑛 does not converge
NOTE:
𝒊 𝟏
We can see that 𝒍𝒊𝒎𝒏→∞ (𝒏) = 𝟎 but ∑ 𝒏 does not converge
If ∑ 𝑦𝑛 = 𝑦 then ∑ 𝑐𝑦𝑛 = 𝑐𝑦 independent of n.
Example: If 𝑦𝑛 > 0 and ∑ 𝑦𝑛 is convergent with the sum 𝑆, then prove that
𝑦𝑛 2𝑦𝑛
< , when 𝑛 is sufficiently large. Also prove that
𝑦1 +𝑦2 +⋯+𝑦𝑛 𝑆
𝑦𝑛
∑ is convergent.
𝑦1 +𝑦2 +⋯+𝑦𝑛
Proof. It is given that ∑ 𝑦𝑛 is convergent with the sum S.
Hence for 𝜀 > 0 ∃ 𝑚 ∈ ℤ+
|𝑆𝑛 − 𝑆| < 𝜀 ∀ 𝑛 ≥ 𝑚 where 𝑆𝑛 = 𝑦1 + 𝑦2 + ⋯ + 𝑦𝑛 ,
or 𝜀 < 𝑆𝑛 − 𝑆 < 𝜀 ⇒ 𝑆 − 𝜀 < 𝑆𝑛 < 𝑆 + 𝜀, ∀ 𝑛 ≥ 𝑚

1
For 𝜀 = 2 𝑆 > 0
1 1 𝑆 3𝑆 2 1 2
𝑆 − 2 𝑆 < 𝑆𝑛 < 𝑆 + 2 𝑆 ⇒ 2 < 𝑆𝑛 < ⇒ > 𝑆 > 3𝑆 , ∀ 𝑛 ≥ 𝑚
2 𝑆 𝑛
2 1 2𝑦𝑛 𝑦
or > 𝑆 ,∀𝑛 ≥ 𝑚 ⇒ > 𝑆𝑛 , ∀ 𝑛 ≥ 𝑚 .
𝑆 𝑛 𝑆 𝑛
𝒚 𝒚 𝒚 𝒚
Now 𝑺𝒏+𝟏 + 𝑺𝒏+𝟐 + 𝑺𝒏+𝟑 + ⋯ + 𝑺𝒏+𝒑
𝒏+𝟏 𝒏+𝟐 𝒏+𝟑 𝒏+𝒑
𝟐
< (𝒚𝒏+𝟏 + 𝒚𝒏+𝟐 + 𝒚𝒏+𝟑 + ⋯ + 𝒚𝒏+𝒑 ), ∀ 𝑛 ≥ 𝑚, 𝑝 ≥ 1
𝑺
𝒚 𝒚 𝒚 𝒚 𝟐
⇒ 𝑺𝒏+𝟏 + 𝑺𝒏+𝟐 + 𝑺𝒏+𝟑 + ⋯ + 𝑺𝒏+𝒑 < 𝑺 (𝑺𝒏+𝒑 − 𝑺𝒏 ), ∀ 𝑛 ≥ 𝑚, 𝑝 ≥ 1.
𝒏+𝟏 𝒏+𝟐 𝒏+𝟑 𝒏+𝒑
As ∑ 𝑦𝑛 is convergent, then given 𝜀 > 0, there exists a positive integer 𝑚1 ,

such that
𝜀𝑆
𝑺𝒏+𝒑 − 𝑺𝒏 < , ∀ 𝑛 ≥ 𝑚1
2
Therefore,
𝒚𝒏+𝟏 𝒚 𝒚 𝒚 𝟐 𝜀𝑆
+ 𝑺𝒏+𝟐 + 𝑺𝒏+𝟑 + ⋯ + 𝑺𝒏+𝒑 < 𝑺 < 𝜀, ∀ 𝑛 ≥ max(𝑚1 , 𝑚) , 𝑝 ≥ 1
𝑺𝒏+𝟏 𝒏+𝟐 𝒏+𝟑 𝒏+𝒑 2
𝑦𝑛
Therefore, by Cauchy’s General Principle of convergence, ∑ 𝑦 is
1 +𝑦2 +⋯+𝑦𝑛
convergent.
5.4 POSITIVE TERM SERIES

Let ∑ 𝑦𝑛 be an infinite series of positive term series of positive terms (𝑦𝑛 ≥
0) and {𝑆𝑛 } be the sequence of its partial sums such that
𝑆𝑛 = 𝑦1 + 𝑦2 + ⋯ + 𝑦𝑛 ≥ 0, ∀𝑛
⇒ 𝑆𝑛 − 𝑆𝑛−1 = 𝑦𝑛 ≥ 0 ⇒ 𝑆𝑛 ≥ 𝑆𝑛−1 , ∀ 𝑛 > 1
Therefore, the sequence {𝑆𝑛 } of partial sums of a series of positive terms is a

monotonic increasing sequence.
Hence {𝑆𝑛 } can either converge or diverge to +∞.
Theorem: A positive term series converges if and only if the sequence of its
partial sums is bounded above.
Proof. Let ∑ 𝑦𝑛 and {𝑆𝑛 } be positive term series and a sequence of its partial
sums respectively.
⇒ {𝑆𝑛 } be a monotonic increasing sequence.
As we know that monotonic increasing sequence converges iff it is bounded

above.
Hence {𝑆𝑛 } converges if and only if the sequence of its partial sums is
bounded above.
Necessary Conditions for convergence of positive term series
Theorem:(Pringsheim’s theorem) If a series ∑ 𝑦𝑛 of positive monotonic

decreasing terms converges then 𝑦𝑛 → 0 and also lim 𝑛𝑦𝑛 = 0.
𝑛→∞
Proof. Let ∑ 𝑦𝑛 be the convergent series of positive monotonic decreasing

terms.
By the definition of convergent series, for any 𝜀 > 0, there exists a positive
integer 𝑀 such that
𝜀
|𝑦𝑚+1 + y𝑚+2 + ⋯ + y𝑚+𝑝 | < , ∀𝑚 ≥ 𝑀, 𝑝 ≥ 1
2
Let 𝑚 + 𝑝 = 𝑛 > 2𝑀 and

𝑛 𝑛
𝑚 = [ 2 ] 𝒊. 𝒆. 𝒎 = 𝑔𝑟𝑒𝑎𝑡𝑒𝑠𝑡 𝑖𝑛𝑡𝑒𝑔𝑒𝑟 𝑛𝑜𝑡 𝑔𝑟𝑒𝑎𝑡𝑒𝑟 𝑡ℎ𝑎𝑛 2 .
Hence
𝜀
𝑦𝑚+1 + y𝑚+2 + ⋯ + y𝑛 < 2
But ∑ 𝑦𝑛 is positive monotonic decreasing.
i.e. 𝑦𝑚+1 > y𝑚+2 > ⋯ > y𝑛 ⇒ 𝑦𝑚+1 + y𝑚+2 + ⋯ + y𝑛 >

𝑦
⏟𝑛 + 𝑦𝑛 + ⋯ . . +𝑦𝑛
(𝑛−𝑚)𝑡𝑖𝑚𝑒𝑠
⇒ 𝑦𝑚+1 + y𝑚+2 + ⋯ + y𝑛 > (𝑛 − 𝑚)𝑦𝑛

𝜀
Therefore (𝑛 − 𝑚)𝑦𝑛 < 𝑦𝑚+1 + y𝑚+2 + ⋯ + y𝑛 < 2
𝑛 𝜀 𝑛
(𝑛 − 2 ) 𝑦𝑛 < 2 𝑏𝑒𝑐𝑎𝑢𝑠𝑒 𝑚 = [ 2 ]
𝑛 𝜀
⇒ < 2 ⇒ 𝑛𝑦𝑛 < 𝜀
2
Hence lim 𝑛𝑦𝑛 = 0

𝑛→∞
NOTE:
lim 𝑛𝑦𝑛 = 0 is only necessary not sufficient condition. If lim 𝑛𝑦𝑛 ≠ 0 then
𝑛→∞ 𝑛→∞
1
the series ∑ 𝑦𝑛 is obviously divergent..Example ∑ 𝑛 diverges because
lim 𝑛𝑦𝑛 = 1 ≠ 0 and positive monotonic decreasing terms.\
𝑛→∞
𝟏
Theorem Let ∑ 𝒏𝒑 be positive term series then it is convergent iff 𝒑 > 𝟏.
1 1 1
Proof. Let 𝑆𝑛 = 1𝑝 + 2𝑝 + ⋯ + 𝑛𝑝
Case 1. When 𝑝 > 1
Now
1
=1 …………(1)
1𝑝
1 1 1 1 2 1
+ 3𝑝 < 2𝑝 + 2𝑝 < 2𝑝 = 2𝑝−1 ………(2)
2𝑝
1 1 1 1 1 1 1 1 4 1 1 2
+ + + < + + + < = =( ) ..…..(3)
4𝑝 5𝑝 6𝑝 7𝑝 4𝑝 4𝑝 4𝑝 4𝑝 4𝑝 4𝑝−1 2𝑝−1
1 1 1 1 1 1 8 1 1 3
+ 9𝑝 + ⋯ + 15𝑝 < 8⏟𝑝 + 8𝑝 + ⋯ + 8𝑝 < 8𝑝 = 8𝑝−1 = (2𝑝−1 ) ..…..(4)
8𝑝
8 𝑡𝑖𝑚𝑒𝑠
……… ……………………… ……… ………………………
……… ……………………… ……… ………………………
……… ……………………… ……… ………………………

1 1 1 1 1 1
+ (2𝑛 +1)𝑝 + ⋯ + (2𝑛+1−1)𝑝 < (2𝑛)𝑝 + (2𝑛)𝑝 + ⋯ + (2𝑛 )𝑝
(2𝑛 )𝑝 ⏟
2𝑛 𝑡𝑖𝑚𝑒𝑠
2𝑛 1 𝑛
= (2𝑛)𝑝 = (2𝑝−1 ) ..(n)
Adding (1), (2),…..,(n), we get
1 1 1 1 𝑝−1 1 𝑛
+ 2𝑝 + ⋯ + (2𝑛+1−1)𝑝 = 𝑆2𝑛+1 −1 < 1 + (2) + ⋯ + (2𝑝−1 )
1𝑝
𝑛+1
1 𝑝−1 1 𝑛+1
1(1−(( ) ) ) 2𝑝−1 (1−( 𝑝−1 ) )
2 2
= 1 𝑝−1
=
1−( ) 2𝑝−1 −1
2
Therefore
2𝑝−1
𝑆2𝑛+1−1 < 2𝑝−1 −1 , 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑛
As we know that when 𝑛 is any positive integer.
2𝑛+1 − 1 > 2𝑛 > 𝑛
Therefore
2𝑝−1
𝑆𝑛 < 𝑆2𝑛 < 𝑆2𝑛+1 −1 < 2𝑝−1 −1
2𝑝−1
Since for a given 𝑝, 2𝑝−1−1 is a fixed number.
Hence, the sequence {𝑆𝑛 } of partial sums of given positive term series is
bounded above.
Therefore, the series converges for 𝑝 > 1.
Case II: When 𝑝 ≤ 1
As we know, if n is any positive integer and 𝑝 ≤ 1 then

1 1
𝑛𝑝 ≤ 𝑛 implies 𝑛𝑝 ≥ 𝑛
Therefore
1 1 1
1 + 2𝑝 ≥ 1 + 2 > 2 ……….(1’)
1 1 1 1 2 1
+ 4𝑝 ≥ 3 + 4 > 4 = 2 …………. (2’)
3𝑝
1 1 1 1 1 1 1 1 4 1
+ + + ≥ + + + > = ………… (3’)
5𝑝 6𝑝 7𝑝 8𝑝 5 6 7 8 8 2
1 1 1 1 1 1
𝑝
+ 𝑝 + ⋯+ 𝑝 ≥ + + ⋯+
9 10 16 9 10 16
1 1 1 8 1
≥⏟ + 16 + ⋯ + 16 = 16 = 2 ….. (4’)
16
8 𝑡𝑖𝑚𝑒𝑠
……… ……………………… ……… ………………………
……… ……………………… ……… ………………………
……… ……………………… ……… ………………………
1 1 1
+ 𝑚−1 + ⋯+ 𝑚 𝑝
(2𝑚−1+ 1) 𝑝 (2 + 2) 𝑝 (2 )
1 1 1 1 1 1
≥ 2𝑚−1 +1 + 2𝑚−1+2 +. . + 2𝑚 > 2⏟𝑚 + 2𝑚 + ⋯ + 2𝑚
2𝑚−1 𝑡𝑖𝑚𝑒𝑠
2𝑚−1 1
= =2 …….(m’)
2𝑚
Adding (1’), (2’),…and (m’), we get

1 1 1 1 1 1 𝑚
+ 2𝑝 + ⋯ + (2𝑚 )𝑝 = 𝑆2𝑚 > 2⏟+ 2 + ⋯ … . + 2 =
1𝑝 2
𝑚 𝑡𝑖𝑚𝑒𝑠
𝑚
i.e. 𝑆2𝑚 > 2 .
Now we try to prove that {𝑆𝑛 } is not bounded above.
𝑚′
Let 𝐾 be any number and there exists 𝑚′ ∈ ℕ such that >𝐾
2
′
Let 𝑛 > 2𝑚
Hence 𝑆𝑛 > 𝑆2𝑚′ > 𝐾
Therefore, we conclude that the sequence of partial sums {𝑆𝑛 } of given series
is not bounded above.
Thus, the series diverges for 𝑝 ≤ 1.

1
Therefore, ∑ 𝑛𝑝 converges if 𝑝 > 1.
5.5 COPARISION TEST

Test 1. If ∑ 𝑢𝑛 and ∑ 𝑣𝑛 are series of positive terms and ∑ 𝑣𝑛 is convergent
and there is a positive constant k such that 𝑢𝑛 ≤ 𝑘𝑣𝑛 , ∀ 𝑛, then ∑ 𝑢𝑛 is also
convergent.
Test 2. If ∑ 𝑢𝑛 and ∑ 𝑣𝑛 are series of positive terms and ∑ 𝑣𝑛 is divergent and

there is a positive constant k such that 𝑢𝑛 ≥ 𝑘𝑣𝑛 , ∀ 𝑛, then ∑ 𝑢𝑛 is also
divergent.
Test 3. If ∑ 𝑢𝑛 and ∑ 𝑣𝑛 are series of positive terms and

𝑢𝑛
(i) if lim = 1(finite and non-zero), then ∑ 𝑢𝑛 and ∑ 𝑣𝑛 both converge or
𝑛⟶∞ 𝑣𝑛
diverge together.
𝑢𝑛
(ii) if lim = 0 and ∑ 𝑣𝑛 converges, then ∑ 𝑢𝑛 is also converges.
𝑛⟶∞ 𝑣𝑛
𝑢𝑛
(iii) if lim = ∞ and ∑ 𝑣𝑛 diverges, then ∑ 𝑢𝑛 is also diverges.
𝑛⟶∞ 𝑣𝑛
5.6 D’ALEMBERT’S RATION TEST

𝑢𝑛
If ∑ 𝑢𝑛 is a series of positive terms such that lim = 𝑙, then
𝑛⟶∞ 𝑢𝑛 +1
(i) ∑ 𝑢𝑛 is convergent if 𝑙 > 1.
(ii) ∑ 𝑢𝑛 is divergent if 𝑙 < 1.
(iii) ∑ 𝑢𝑛 may converge or diverge if 𝑙 = 1.(i.e. the test fails if 𝑙 = 1).

𝑢𝑛
And if lim = ∞, then ∑ 𝑢𝑛 is convergent.
𝑛⟶∞ 𝑢𝑛 +1
5.7 CAUCHY’S ROOT TEST
If ∑ 𝑢𝑛 is a series of positive terms such that lim (𝑢𝑛 )1/𝑛 = 𝑙, then

𝑛⟶∞
(i) ∑ 𝑢𝑛 is convergent if 𝑙 < 1.
(ii) ∑ 𝑢𝑛 is divergent if 𝑙 > 1.
(iii) ∑ 𝑢𝑛 may converge or diverge if 𝑙 = 1.(i.e. the test fails if 𝑙 = 1).
And if lim (𝑢𝑛 )1/𝑛 = ∞, then ∑ 𝑢𝑛 is divergent.

𝑛⟶∞
5.8 ALTERNATING SERIES
A series with terms alternately positive and negative is called an alternating

series.
Thus, the series 𝑢1 − 𝑢2 + 𝑢3 − 𝑢4 … where 𝑢𝑛 > 0.
For each n, is an alternating series and is briefly written as ∑(−1)𝑛−1 𝑢𝑛 .
Leibnitz’s test on alternating series
The alternating series ∑(−1)𝑛−1 𝑢𝑛 = 𝑢1 − 𝑢2 + 𝑢3 − 𝑢4 … where 𝑢𝑛 >

0 ∀ 𝑛 converges if
(i) 𝑢𝑛 ≥ 𝑢𝑛+1 ∀ 𝑛 and (ii) lim 𝑢𝑛 = 0.

𝑛⟶∞
5.9 ABSOLUTE AND CONDITIONAL

CONVERGENCE
Definition 1. A series ∑∞
𝑛=1 𝑢𝑛 is said to be absolute convergent if the
series ∑∞
𝑛=1|𝑢𝑛 | is convergent.
Definition 2. If ∑∞ ∞
𝑛=1 𝑢𝑛 converges but not absolutely then the series ∑𝑛=1 𝑢𝑛
is called conditionally convergent.
Theorem: Every absolutely convergent series is convergent.
Proof. Let ∑∞ ∞
𝑛=1 𝑢𝑛 be absolutely convergent series, then ∑𝑛=1|𝑢𝑛 | is
convergent.
Therefore, By Cauchy’s general principle of convergence, given 𝜖 > 0, ∃ a

positive integer m such that ||𝑢𝑚+1 | + |𝑢𝑚+2 | + ⋯ + |𝑢𝑛 || < 𝜖 ∀ 𝑛 > 𝑚
i.e. |𝑢𝑚+1 | + |𝑢𝑚+2 | + ⋯ + |𝑢𝑛 | < 𝜖 ∀𝑛>𝑚 ……. (1)
now, by triangle inequality, we have
|𝑢𝑚+1 + 𝑢𝑚+2 + ⋯ + 𝑢𝑛 | ≤ |𝑢𝑚+1 | + |𝑢𝑚+2 | + ⋯ + |𝑢𝑛 |

<𝜖 ∀𝑛 >𝑚
Therefore, By Cauchy’s general principle of convergence, the series ∑∞

𝑛=1 𝑢𝑛
is convergent.
Hence ∑∞
𝑛=1|𝑢𝑛 | is convergent ⟹ 𝑢𝑛 is convergent.
Note 1. Absolute convergence ⟹ convergent, but convergence need not

imply absolute convergence i.e. the convergence of above theorem need not
be true.
For Example: Consider the series

(−1) 𝑛−1 1 1 1
∑∞
𝑛=1 =1−2+3−4+⋯
𝑛
1
It is an alternating series. Here 𝑢𝑛 = 𝑛. Clearly 𝑢𝑛 > 0 ∀ 𝑛
1 1
Since 𝑛 > 𝑛+1 , 𝑢𝑛 > 𝑢𝑛+1 ∀ 𝑛
1
Also lim 𝑢𝑛 = lim =0
𝑛⟶∞ 𝑛⟶∞ 𝑛
(−1)𝑛−1
Therefore, by Leibnitz’s test ∑∞
𝑛=1 is convergent.
𝑛
Note: If ∑∞ 𝑛=1 𝑢𝑛 is an absolutely convergent series, then the series of its

positive terms and the series of its negative terms are both convergent.
Note: If ∑∞𝑛=1 𝑢𝑛 is conditionally convergent, then the series of its positive

terms and the series of its negative terms are both divergent.
Note: A series with mixed signs cannot converge if the series of its positive
terms is convergent (divergent) and the series of its negative terms is
divergent (convergent).
Example 1. Test the convergence of the series

1 1 1
(i) ∑∞
𝑛=1 sin (ii) ∑∞
𝑛=1 sin
𝑛 𝑛 𝑛
1
Sol. (i) Here 𝑢𝑛 = sin 𝑛
1 1 1 1 1
= − . + . −⋯
𝑛 3! 𝑛3 5! 𝑛5
1 1 1 1 1
= 𝑛 [1 − 3! . 𝑛2 + 5! . 𝑛4 − ⋯ ]
1
Take 𝑣𝑛 =
𝑛
𝑢𝑛 1 1 1 1
lim = lim [1 − 3! . 𝑛2 + 5! . 𝑛4 − ⋯ ] = 1 which is finite and ≠ 0.
𝑛⟶∞ 𝑣𝑛 𝑛⟶∞
Therefore, ∑∞ ∞
𝑛=1 𝑢𝑛 and ∑𝑛=1 𝑣𝑛 converges and diverges together.
1 1
Since ∑∞ ∞ ∞
𝑛=1 𝑣𝑛 = ∑𝑛=1 𝑛 is of the form ∑𝑛=1 𝑛 𝑝 with p = 1.
𝑛=1 𝑣𝑛 is divergent ⟹ ∑𝑛=1 𝑢𝑛 is divergent.
1 1
(ii) Here 𝑢𝑛 = 𝑛 sin 𝑛
1 1 1 1 1 1
= 𝑛 [𝑛 − 3! . 𝑛3 + 5! . 𝑛5 − ⋯ ]
1 1 1 1 1
= 𝑛2 [1 − 3! . 𝑛2 + 5! . 𝑛4 − ⋯ ]
1
Take 𝑣𝑛 =
𝑛2
𝑢𝑛 1 1 1 1
lim = lim [1 − 3! . 𝑛2 + 5! . 𝑛4 − ⋯ ] = 1 which is finite and ≠ 0.
𝑛⟶∞ 𝑣𝑛 𝑛⟶∞
𝑛=1 𝑢𝑛 and ∑𝑛=1 𝑣𝑛 converges and diverges together.
1 1
Since ∑∞ ∞ ∞
𝑛=1 𝑣𝑛 = ∑𝑛=1 𝑛 2 is of the form ∑𝑛=1 𝑛 𝑝 with p = 2.
𝑛=1 𝑣𝑛 is convergent ⟹ ∑𝑛=1 𝑢𝑛 is convergent.
Example 2. Discuss the convergence or divergence of the following series:
𝑛2
(i) ∑∞
𝑛=1 𝑛!
𝑛2
Sol. Here, 𝑢𝑛 = 𝑛!
(𝑛+1)2 (𝑛+1)2 𝑛+1

Therefore, 𝑢𝑛+1 = = (𝑛+1)𝑛! =
(𝑛+1)! 𝑛!
𝑢𝑛 𝑛2 1
Therefore, 𝑢 = 𝑛+1 = 1 1
𝑛+1 +
𝑛 𝑛2
𝑢𝑛 1 1
Therefore, lim = lim 1 1 =0=∞
𝑛⟶∞ 𝑢𝑛+1 𝑛⟶∞ +
𝑛 𝑛2
Therefore, by D’Alembert’s Ration test, ∑∞

𝑛=1 𝑢𝑛 is convergent.
Example 3. Test the convergence of the following series:

2
𝑛 𝑛
∑∞
𝑛=1 𝑛+1)
( .
𝑛 𝑛2
Sol. Here, 𝑢𝑛 = (𝑛+1)
1/𝑛
𝑛 𝑛2 𝑛 𝑛 𝑛+1 −𝑛
Therefore, (𝑢𝑛 )1/𝑛 = [(𝑛+1) ] = (𝑛+1) = ( )
𝑛
−1
1 𝑛
= [(1 + 𝑛) ]
−1
1 𝑛 1
Therefore, lim (𝑢𝑛 )1/𝑛 = lim [(1 + 𝑛) ] = 𝑒 −1 = 𝑒 < 1
𝑛⟶∞ 𝑛⟶∞
Therefore, by Cauchy’s Root Test, the given series ∑∞

𝑛=1 𝑢𝑛 is convergent.
CHECK YOUR PROGRESS
Problem 1. Every absolutely convergent series is convergent.

Problem 2. Every convergent series is absolutely convergent.
2
1 −𝑛
Problem 3. The series ∑∞𝑛=1 (1 + 𝑛 ) is convergent.
Problem 4. A positive term series converges if and only if the
sequence of its partial
sums is bounded above.
Problem 5. The series ∑∞𝑛=1 𝑛 is convergent.
5.10 SUMMARY
1. If ∑∞
𝑛=1 𝑢𝑛 is an absolutely convergent series, then the series of its positive
terms and the series of its negative terms are both convergent.
2. If ∑∞
𝑛=1 𝑢𝑛 is conditionally convergent, then the series of its positive terms
and the series of its negative terms are both divergent.
3. A series with mixed signs cannot converge if the series of its positive terms
is convergent (divergent) and the series of its negative terms is divergent
(convergent).
4. Test 1. If ∑ 𝑢𝑛 and ∑ 𝑣𝑛 are series of positive terms and ∑ 𝑣𝑛 is convergent

and there is a positive constant k such that 𝑢𝑛 ≤ 𝑘𝑣𝑛 , ∀ 𝑛, then ∑ 𝑢𝑛 is also
convergent.
Test 2. If ∑ 𝑢𝑛 and ∑ 𝑣𝑛 are series of positive terms and ∑ 𝑣𝑛 is divergent and

there is a positive constant k such that 𝑢𝑛 ≥ 𝑘𝑣𝑛 , ∀ 𝑛, then ∑ 𝑢𝑛 is also
divergent.
Test 3. If ∑ 𝑢𝑛 and ∑ 𝑣𝑛 are series of positive terms and

𝑢𝑛
(i) if lim = 1(finite and non-zero), then ∑ 𝑢𝑛 and ∑ 𝑣𝑛 both converge or
𝑛⟶∞ 𝑣𝑛
diverge together.
𝑢𝑛
(ii) if lim = 0 and ∑ 𝑣𝑛 converges, then ∑ 𝑢𝑛 is also converges.
𝑛⟶∞ 𝑣𝑛
𝑢𝑛
(iii) if lim = ∞ and ∑ 𝑣𝑛 diverges, then ∑ 𝑢𝑛 is also diverges.
𝑛⟶∞ 𝑣𝑛
5.11 GLOSSARY
sequence
limit
5.12 REFERENCES
1. T. M. Apostol, Mathematical Analysis (2nd Edition), Narosa Publishing

House, 2002.
2. R.G. Bartle and D.R. Sherbert, Introduction of real analysis (3rd Edition), John
Wiley and Sons (Asia) P. Ltd., Inc. 2000.
3. W. Rudin, Principles of Mathematical Analysis (3rd Edition), McGraw-Hill
Publishing, 1976.
4. S.C. Malik and Savita Arora, Mathematical Analysis (6th Edition), New Age
International Publishers, 2021.
5. Shanti Narayan, A course of Mathematical Analysis (29th Edition), S. Chand
and Co., 2005.
6. K. A. Ross, Elementary Analysis, The Theory of Calculus (2nd edition),
Springer, 2013.

2𝑛 . 𝑛!
Q 1. Examine the convergence of the series ∑∞
𝑛=1 .
𝑛𝑛
1
𝑛=1 𝑛! .
1
𝑛=1 (𝑙𝑜𝑔𝑛)𝑛 .
Q 4. Define Cauchy’s Roots Test.
Q 5. Test the convergence and absolutely convergence of the series
(−1)𝑛−1 𝑛
∑∞
𝑛=1 .
𝑛+2
5.15 ANSWERS
TQ1. Convergent.
TQ2. Convergent.
TQ3. Convergent.
TQ5. Not convergent
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. False
CYQ 3. True
CYQ 4. True
CYQ 5. False

BLOCK-II
FUNCTIONS SINGLE VARIABLE
UNIT 6: LIMITS
CONTENTS:
6.1 Introduction
6.2 Objectives
6.3 Limit on the basis of (𝜀, 𝛿)
6.4 Variable
6.5 Limit on the basis of L.H.L. and R.H.L.
6.6 Infinites limits
6.7 L Hospital rule
6.8 Summary
6.9 Glossary
6.10 References
6.13 Answers
6.1 INTRODUCTION
In Mathematics, the limit of a function is a fundamental concept

in calculus and analysis concerning the behaviour of that function near a
particular input which may or may not be in the domain of the function.
In previous unit we discussed about sequence and series. In this unit we
will be discussed about limit of one variable function.
The limit of a function is defined as the unique real number that the
functions take when the variable of the function approaches a particular
point. For any given function f(x), and a real number ‘c’, the limit of the
function is defined as,
𝐥𝐢𝐦 𝒇(𝒙) = L
𝒙→𝒂
This is read as, “limit of f(x), as x approaches a equals L”
6.2 OBJECTIVES
(i) Neighborhood
(ii) Interior point
(iii) Open set
(iv) Limit point
6.3 LIMIT
∎(𝜺, 𝜹) definition of Limit.

Let 𝐴 ⊆ 𝑅, and let 𝑐 be a cluster point of A. For a function 𝑓 ∶ 𝐴 → ℝ, a
real number 𝑙 is said to be a limit of 𝑓 at 𝑐 if, given any 𝜀 > 0, there exists
a 𝛿 > 0 such that
if 𝑥 ∈ 𝐴 and 0 < |𝑥 − 𝑐| < 𝑑, then |𝑓(𝑥 ) − 𝑙 | < 𝜀.
Note: If 𝑓 ∶ 𝐴 → ℝ and if 𝑐 is a cluster point of 𝐴, then 𝑓 can have only

one limit at 𝑐.
Proof: Let 𝑙 and 𝑙’ be the limits of function 𝑓.
𝜀
For any 𝜀 > 0, there exists 𝛿1 (2) > 0 such that if 𝑥 ∈ 𝐴 and
𝜀 𝜀
0 < |𝑥 − 𝑐 | < 𝛿1 (2), then |𝑓 (𝑥 ) − 𝑙 | < 2.
𝜀 𝜀
and there exists 𝛿2 (2) > 0 such that if 𝑥 ∈ 𝐴 and 0 < |𝑥 − 𝑐 | < 𝛿2 (2),
𝜀
then |𝑓 (𝑥 ) − 𝑙 ′ | < 2.
𝜀 𝜀
Let 𝛿 = inf {𝛿1 (2) , 𝛿2 (2)}. Then if 𝑥 ∈ 𝐴 and 0 < |𝑥 − 𝑐 | < 𝛿.
The Triangle Inequality implies that
|𝑙 − 𝑙 ′ | = |𝑙 + (−𝑓(𝑥) + 𝑓 (𝑥 )) − 𝑙 ′ | ≤ |𝑙 − 𝑓(𝑥 )| + |𝑓(𝑥 ) − 𝑙 ′ |
𝜀 𝜀
< 2 + 2 < 𝜀.
Because 𝜀 > 0 is arbitrary. Therefore, 𝑙 − 𝑙 ′ = 0 ⇒ 𝑙 = 𝑙 ′ .
Example 6.3.1. Prove that 𝐥𝐢𝐦 𝒄 = 𝒄

𝒙→𝒂
Sol. Let 𝑓 (𝑥 ) = 𝑐 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥 ∈ ℝ .

Now we will try to prove that lim 𝑓 (𝑥 ) = 𝑐.
𝑥→𝑎
Let 𝜀 > 0 and 𝛿 = 1.
Then if 0 < |𝑥 − 𝑎| < 1, we have
|𝑓 (𝑥 ) − 𝑐 | = |𝑐 − 𝑐 | = 0 < 𝜀.
As 𝜀 > 0 is arbitrary, by definition of limit we get
lim 𝑓(𝑥) = 𝑐 .
𝑥→𝑎
Example 6.3.2. Prove that 𝐥𝐢𝐦 𝒙𝟐 = 𝒃𝟐

𝒙→𝒃
Sol. Let 𝑓 (𝑥 ) = 𝒙𝟐 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥 ∈ ℝ .

Now we will try to prove that lim 𝑓 (𝑥 ) = 𝑏2 .
𝑥→𝑎
Now we try to prove that |𝑓(𝑥 ) − 𝑏2 | = |𝑥 2 − 𝑏2 | less than a
preassigned 𝜀 > 0 by taking x sufficiently close to 𝑏.
Now
𝑥 2 − 𝑏2 = (𝑥 − 𝑏)(𝑥 + 𝑏).
If | 𝑥 − 𝑏| < 1, then
|𝑥 | < |𝑏 | + 1
Hence |𝑥 + 𝑏| ≤ |𝑥 | + |𝑏| < |𝑏| + 1 + |𝑏| < 2|𝑏| + 1
Thus, if |𝑥 − 𝑏| < 1 then
|𝑥 2 − 𝑏2 | ≤ |𝑥 − 𝑏||𝑥 + 𝑏| < (2|𝑏| + 1)|𝑥 − 𝑏| ……(1)
𝜀 𝜀
Let |𝑥 − 𝑏| < 2|𝑐|+1 and we choose 𝛿(𝜀 ) = inf {1, 2|𝑐|+1},
Then if 0 < |𝑥 − 𝑏| < 𝛿 (𝜀 ),

Now if |𝑥 − 𝑏| < 1, then equation (1) is valid.
𝜀
If |𝑥 − 𝑏| < 2|𝑏|+1 then
𝜀
|𝑥 2 − 𝑏2 | < (2|𝑏| + 1)|𝑥 − 𝑏| < (2|𝑏| + 1). < 𝜀
2|𝑐|+1
As we have choice to choose 𝛿 (𝜀 ) > 0 for an arbitrary choice of 𝜀 > 0.
We deduce that lim 𝑥 2 = 𝑏2
𝑥→𝑏
𝟏 𝟏
Example 6.3.3. Prove that 𝐥𝐢𝐦 = 𝒃 𝒊𝒇 𝒃 > 𝟎
𝒙→𝒃 𝒙
𝟏
Proof. Let 𝑓 (𝑥 ) = 𝒙 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥 > 0 and assume 𝒃 > 𝟎
1
Now we will try to prove that lim 𝑓 (𝑥 ) = 𝑏.
𝑥→𝑎
1 1 1
Therefore, we will try to prove that the difference |𝑓(𝑥) − 𝑏| = |𝑥 − 𝑏 |
less than a preassigned 𝜀 > 0 by taking x sufficiently close to 𝑏 > 0.
Now
1 1 1 1
| − | = | (𝑏 − 𝑥 )| = |𝑥 − 𝑏| 𝑓𝑜𝑟 𝑥 > 0.
𝑥 𝑏 𝑏𝑥 𝑏𝑥
1
Now if | 𝑥 − 𝑏| < 2 𝑏 then
1 1 1 3 1 2 1
− 2 𝑏 < 𝑥 − 𝑏 < 2 𝑏 ⇒ 2 𝑏 < 𝑥 < 2 𝑏 ⇒ 2 𝑏2 < 𝑏𝑥 ⇒ 𝑏2 > 𝑏𝑥 .
Therefore
1 2 1
0 < 𝑏𝑥 < 𝑏2 for | 𝑥 − 𝑏| < 2 𝑏
Hence, for these values of 𝑥 we have
1 2
|𝑓(𝑥) − | < |𝑥 − 𝑏 | ………(1)
𝑏 𝑏2
In order to make this last term less than 𝜀 it suffices to take
1 1 1
|𝑥 − 𝑏| < 𝑏2 𝜀. Consequently, if we choose 𝛿 (𝜀 ) = inf { 𝑏, 𝑏2 𝜀},
2 2 2
Then if 0 < |𝑥 − 𝑏| < 𝛿 (𝜀 ),
1
Now if |𝑥 − 𝑏| < 2 𝑏, then equation (1) is valid.
1
Therefore, since |𝑥 − 𝑏| < 2 𝑏2 𝜀, that
1 1 1
|𝑓(𝑥) − | = | − | < 𝜀
𝑏 𝑥 𝑏
Since we have a way of choosing 𝛿 (𝜀 ) > 0 for an arbitrary choice of 𝜀 >
0, we conclude that
1 1
lim = 𝑏 𝑖𝑓 𝑏 > 0.
𝑥→𝑏 𝑥
Theorem 6.3.1. Let 𝑋 ⊆ ℝ and 𝑓, 𝑔: 𝑋 → ℝ and let 𝑏 ∈ ℝ be a cluster

point of 𝑋 and ∈ ℝ .
(a) If lim 𝑓 = 𝑙1 and lim 𝑔 = 𝑙2 , then
𝑥→𝑏 𝑥→𝑏
(i) lim 𝑓 + 𝑔 = 𝑙1 + 𝑙2 (ii) lim 𝑓 − 𝑔 = 𝑙1 − 𝑙2
(iii) lim 𝑓𝑔 = 𝑙1 𝑙2
𝑥→𝑏
(b) If ℎ: 𝑋 → ℝ and ℎ(𝑥) ≠ 0 for all 𝑥 ∈ 𝑋, if lim ℎ = 𝑙3 ≠ 0, then
𝑥→𝑏
𝑓 𝑙1
lim = 𝑙
𝑥→𝑏 ℎ 3
Proof. (a) It is given that lim 𝑓 = 𝑙1 and lim 𝑔 = 𝑙2 , Hence for any
𝜀 > 0 there exists a positive number 𝛿1 and 𝛿2 such that
𝜀 𝜀
|𝑓 (𝑥 ) − 𝑙1 | < when 0 < |𝑥 − 𝑏| < 𝛿1 and |𝑔(𝑥 ) − 𝑙2 | < when
2 2
0 < |𝑥 − 𝑏| < 𝛿2
Let 𝛿 = min(𝛿1 , 𝛿2 ), then
𝜀
|𝑓 (𝑥 ) − 𝑙1 | < when 0 < |𝑥 − 𝑏| < 𝛿 ………(1)
2
and
𝜀
|𝑔(𝑥) − 𝑙2 | < when 0 < |𝑥 − 𝑏| < 𝛿 ………(2)
2
Now, when 0 < |𝑥 − 𝑏| < 𝛿
|(𝑓 + 𝑔)(𝑥) − (𝑙1 + 𝑙2 )| = |𝑓 (𝑥 ) − 𝑙1 + 𝑔(𝑥 ) − 𝑙2 |
≤ |𝑓 (𝑥 ) − 𝑙1 | + |𝑔(𝑥 ) − 𝑙2 |
𝜀 𝜀
<2+2 =𝜀
Therefore
|(𝑓 + 𝑔)(𝑥) − (𝑙1 + 𝑙2 )| < 𝜀 when 0 < |𝑥 − 𝑏| < 𝛿
Thus, lim 𝑓 + 𝑔 = 𝑙1 + 𝑙2
𝑥→𝑏
(ii) When 0 < |𝑥 − 𝑏| < 𝛿
|(𝑓 − 𝑔)(𝑥) − (𝑙1 − 𝑙2 )| = |𝑓 (𝑥 ) − 𝑙1 + 𝑔(𝑥 ) − 𝑙2 |
≤ |𝑓 (𝑥 ) − 𝑙1 | + |𝑙2 − 𝑔(𝑥 )|
𝜀 𝜀
= |𝑓 (𝑥 ) − 𝑙1 | + |𝑔(𝑥 ) − 𝑙2 | < 2 + 2 = 𝜀
(From (1) and (2))
|(𝑓 − 𝑔)(𝑥) − (𝑙1 − 𝑙2 )| < 𝜀 when 0 < |𝑥 − 𝑏| < 𝛿
Thus, lim 𝑓 − 𝑔 = 𝑙1 − 𝑙2
𝑥→𝑏
(iii) |(𝑓𝑔)(𝑥) − (𝑙1 𝑙2 )| = |𝑓 (𝑥 )𝑔(𝑥 ) − 𝑙1 𝑙2 |
= |𝑓 (𝑥 )𝑔(𝑥 ) − 𝑓(𝑥 )𝑙2 + 𝑓(𝑥)𝑙2 − 𝑙1 𝑙2 |
= |𝑓 (𝑥 )(𝑔(𝑥 ) − 𝑙2 ) + 𝑙2 (𝑓(𝑥) − 𝑙1 )|
≤ |𝑓 (𝑥 )||𝑔(𝑥) − 𝑙2 | + |𝑙2 ||𝑓(𝑥 ) − 𝑙1 | .….(3)
As we know that lim 𝑓 = 𝑙1 . Hence for any 𝜀 = 1 there exists a positive
𝑥→𝑏
number 𝛿′1 such that
|𝑓 (𝑥 ) − 𝑙1 | < 1 when 0 < |𝑥 − 𝑏| < 𝛿′1 .
Now
|𝑓 (𝑥 )| = |𝑓(𝑥 ) − 𝑙1 + 𝑙1 | ≤ |𝑓(𝑥) − 𝑙1 | + |𝑙1 |
< 1 + |𝑙1 | , when 0 < |𝑥 − 𝑏| < 𝛿′1 …….(4)
lim 𝑔 = 𝑙2 , there exists a positive number 𝛿′2 such that
𝑥→𝑏
𝜀
|𝑔 (𝑥 ) − 𝑙2 | < 2
when 0 < |𝑥 − 𝑏| < 𝛿′2 …….(5)
1+|𝑙1 |
Again lim 𝑓 = 𝑙1 , there exists a positive number 𝛿′3 such that

𝑥→𝑏
𝜀
|𝑓 (𝑥 ) − 𝑙1 | < 2
when 0 < |𝑥 − 𝑏| < 𝛿′3 …...(6)
|𝑙2 |
Let 𝛿 ′ = min{𝛿1′ , 𝛿2′ , 𝛿3′ }. Then from (3), (4), (5) and (6),
when 0 < |𝑥 − 𝑏| < 𝛿′
𝜀 𝜀
|(𝑓𝑔)(𝑥) − (𝑙1 𝑙2 )| < (1 + |𝑙1 |) 2
+ |𝑙2 | |𝑙 | < 𝜀.
2
1+|𝑙1 | 2
Hence lim 𝑓𝑔 = 𝑙1 𝑙2 .
𝑥→𝑏
|𝑚|
(b) lim ℎ = 𝑙3 ≠ 0 therefore for 𝜀 = > 0 there exists 𝛿3 > 0 such that
𝑥→𝑏 2
|𝑚|
|ℎ (𝑥 ) − 𝑙3 | < when 0 < |𝑥 − 𝑏| < 𝛿3
2
Now
|𝑙3 | = |𝑙3 − ℎ(𝑥 ) + ℎ(𝑥)| ≤ |𝑙3 − ℎ(𝑥)| + |ℎ(𝑥 )| = |ℎ(𝑥 ) − 𝑙3 | +
|ℎ(𝑥 )| or
|𝑙3 | |𝑙3 | |𝑙3 | 1 2
or |𝑙3 | < + |ℎ(𝑥 )| ⇒ |ℎ(𝑥 )| > |𝑙3 | − = . 𝑜𝑟 |ℎ(𝑥)|
< |𝑙
2 2 2 3|
It implies that there exists a deleted neighbourhood of 𝑏 on which ℎ(𝑥)
does not vanish.
Now, when 0 < |𝑥 − 𝑏| < 𝛿3
𝑓 𝑙1 𝑓(𝑥 ) 𝑙1 𝑓(𝑥 )𝑙3 − ℎ(𝑥 )𝑙1
|( ) (𝑥 ) − | = | − |=| |
ℎ 𝑙3 ℎ (𝑥 ) 𝑙3 ℎ(𝑥 )𝑙3
𝑓(𝑥 )𝑙3 − 𝑙1 𝑙3 + 𝑙1 𝑙3 − ℎ(𝑥 )𝑙1
=| |
ℎ(𝑥 )𝑙3
𝑙 (𝑓(𝑥)−𝑙1 )+𝑙1 (𝑙3 −ℎ(𝑥))
=|3 |
ℎ(𝑥)𝑙3
1 |𝑙1 |
≤ |ℎ(𝑥)| |𝑓(𝑥 ) − 𝑙1 | + |𝑙 |ℎ(𝑥 ) − 𝑙3 |
3 ||ℎ(𝑥)|
2 2 |𝑙1 |
< . |𝑓 (𝑥) − 𝑙1 | + | ℎ (𝑥 ) − 𝑙3 |
|𝑙 3 | |𝑙3 | |𝑙3 |
2 2|𝑙 |
= |𝑙 . |𝑓 (𝑥 ) − 𝑙1 | + 12 |ℎ(𝑥 ) − 𝑙3 | ............(7)
3| |𝑙3 |
Let 𝜀 > 0 be given.
It is given that lim 𝑓 = 𝑙1 and lim ℎ = 𝑙3 , hence there exists positive
numbers 𝛿1" and 𝛿2" such that
1
|𝑓 (𝑥 ) − 𝑙1 | < 𝜀 |𝑙3 |, when 0 < |𝑥 − 𝑏| < 𝛿1" …….(8)
4
1 |𝑙3 |2
|ℎ (𝑥 ) − 𝑙3 | < 𝜀
|𝑙1 |
, when 0 < |𝑥 − 𝑏| < 𝛿2" .….(9)
4
Let 𝛿" = min {𝛿3 , 𝛿1" , 𝛿2" } . Then from (7), (8) and (9), we get
𝑓 𝑙 2 1 2|𝑙 | 1 |𝑙 |2 𝜀 𝜀
|( ) (𝑥) − 1 | < |𝑙 | . 𝜀|𝑙3 | + |𝑙 1|2 . 𝜀 |𝑙3 | = + = 𝜀 when
ℎ 𝑙3 3 4 3 4 1 2 2
0 < |𝑥 − 𝑏| < 𝛿"
𝑓 𝑙
Therefore |( ) (𝑥 ) − 1 | < 𝜀 when 0 < |𝑥 − 𝑏| < 𝛿"
ℎ 𝑙3
𝑓 𝑙1
Hence lim ℎ = 𝑙
𝑥→𝑏 3
√𝟒+𝒙−𝟐
Ex.6.3.4. Find 𝐥𝐢𝐦𝐱→𝟎 𝒙
√4+𝑥−2
Proof. It is given that 𝑙𝑖𝑚𝑥→0 𝑥
√4+𝑥−2 √4+𝑥−2 √4+𝑥+2 4+𝑥−4
𝑙𝑖𝑚𝑥→0 = 𝑙𝑖𝑚𝑥→0 . = 𝑙𝑖𝑚𝑥→0 . =
𝑥 𝑥 √4+𝑥−2 √4+𝑥+2
𝑥 1
𝑙𝑖𝑚𝑥→0 =4.
√4+𝑥+2
6.4 VARIABLE
A symbol such as x or y, used to represent an arbitrary element of a set is

called a variable. For example y = f (x).
The symbol x which represents an element in the domain is called the

independent variable, and the symbol y which represent the element
corresponding to x is called the dependent variable. This is based on the
fact that value of x can be arbitrary chosen, then y has a value which
depends upon the chosen value of x.
6.5 LIMIT
∎ Definition of Limit.
f(x) is said to tend to a limit as 𝑥 tends to ‘𝑎’ if both the left and right
hand limits exist and equal, and their common value is called the limit of
the function.
lim− 𝑓 (𝑥 ) = lim 𝑓 (𝑎 − ℎ) where, ℎ > 0 is called left hand limit (L.H.L.)
𝑥→𝑎 ℎ→0
And
lim+ 𝑓 (𝑥 ) = lim 𝑓 (𝑎 + ℎ) where, ℎ > 0 is called right hand limit
𝑥→𝑎 ℎ→0
(R.H.L.)
If L.H.L. = R.H.L. then lim 𝑓(𝑥 ) exist.

𝑥→𝑎
And if L.H.L. ≠ R.H.L. then lim 𝑓 (𝑥 ) does not exist.
𝑥→𝑎
Example 6.5.1. Do the following limits exists? if yes, find them.

1
(i) lim sin 𝑥−1
𝑥→1
1
(ii) lim 𝑥 sin 𝑥
𝑥→0
1
(iii) lim 2 𝑥−1
𝑥→1
𝑒 1/𝑥
(iv) lim 𝑒 1/𝑥 +1
𝑥→0
Sol. (i) L.H.L. = lim− 𝑓(𝑥) = lim 𝑓(1 − ℎ)

𝑥→1 ℎ→0
1 1
= lim sin 1−ℎ−1 = −lim sin ℎ .
ℎ→0 ℎ→0
1
Now as ℎ → 0, sin ℎ is finite and oscillates between −1 and 1, so it does
not tend to any unique and definite value as ℎ → 0. Hence L.H.L. does not
exist.
Similarly, the right-hand limit also does not exist as 𝑥 → 1.
1
Thus lim sin 𝑥−1 does not exist.
𝑥→1
(ii) L.H.L. = lim− 𝑓 (𝑥 ) = lim 𝑓(0 − ℎ)
𝑥→0 ℎ→0
1 1
= lim (0 − ℎ) sin = lim h sin
ℎ→0 0−ℎ ℎ→0 ℎ
= 0 × a finite quantity between −1 and 1
= 0.
Similarly, R.H.L. = lim+ 𝑓 (𝑥 ) = lim 𝑓(0 + ℎ)
𝑥→0 ℎ→0
1 1
= lim (0 + ℎ) sin 0+ℎ = lim h sin ℎ
ℎ→0 ℎ→0
= 0 × a finite quantity between −1 and 1
= 0.
1
Thus L.H.L. and R.H.L. both exist and are equal, and hence lim 𝑥 sin 𝑥
𝑥→0
exists and is equal to zero.
1
∴ lim 𝑥 sin = 0.
𝑥→0 𝑥
(iii) L.H.L. = lim− 𝑓 (𝑥 ) = lim 𝑓(1 − ℎ)

𝑥→1 ℎ→0
1 1
1 1
= lim 2 1−ℎ−1 = lim 2 −ℎ = 2−∞ = 2∞ = ∞ = 0.
ℎ→0 ℎ→0
R.H.L. = lim+ 𝑓 (𝑥 ) = lim 𝑓 (1 + ℎ)
𝑥→1 ℎ→0
1 1
= lim 2 1+ℎ−1 = lim 2 ℎ = 2∞ = ∞.
ℎ→0 ℎ→0
Since L.H.L.≠ R.H.L.
1
∴ lim 2 𝑥−1 does not exist.
𝑥→1
(iv) L.H.L. = lim− 𝑓 (𝑥 ) = lim 𝑓(0 − ℎ)
𝑥→0 ℎ→0
1 1
−
𝑒 0−ℎ 𝑒 ℎ 0
= lim 1 = lim 1 = 0+1 = 0
ℎ→0 ℎ→0 𝑒 −
𝑒 0−ℎ + 1 ℎ +1
R.H.L. = lim+ 𝑓(𝑥) = lim 𝑓(0 + ℎ)
𝑥→0 ℎ→0
1 1
𝑒0 + ℎ 𝑒ℎ 1
= lim 1 = lim 1 = lim 1
ℎ→0 𝑒 0 + ℎ + 1 ℎ→0 𝑒 ℎ + 1 ℎ→0 1 − 𝑒 − ℎ
1 1
= 1 + 𝑒 −∞ = 1 +0 = 1.
Since L.H.L.≠ R.H.L.

𝑒 1/𝑥
∴ lim 𝑒 1/𝑥 +1 does not exist.
𝑥→0
𝑥2
− 𝑎 𝑓𝑜𝑟 0 < 𝑥 < 𝑎
Example 6.5.2. Find lim 𝑓(𝑥) where 𝑓 (𝑥 ) = { 𝑎 𝑎3 .
𝑥→0
𝑎 − 𝑥2 𝑓𝑜𝑟 𝑥 > 𝑎
Sol. L.H.L. = lim− 𝑓 (𝑥 ) = lim 𝑓 (𝑎 − ℎ)

𝑥→𝑎 ℎ→0
(𝑎−ℎ)2 𝑎2
= lim [ − 𝑎] = −𝑎 = 𝑎−𝑎 =0
ℎ→0 𝑎 𝑎
R.H.L. = lim+ 𝑓 (𝑥 ) = lim 𝑓 (𝑎 + ℎ)
𝑥→𝑎 ℎ→0
𝑎3 𝑎3
= lim [𝑎 − (𝑎+ℎ)2 ] = 𝑎 − 𝑎2 = 𝑎 − 𝑎 = 0.
ℎ→0
Therefore, L.H.L. and R.H.L. both exist and each equal to 0.
∴ lim 𝑓 (𝑥 ) = 0.
𝑥→0
6.6 INFINITE LIMITS
A function f (x) is said to approach + ∞ or - ∞ as x → a, if for given   0

there exists   0 such that
f ( x)   or f ( x)    whenever 0  | x  a |   .
Then in other words, Limit f ( x)   or Limit f ( x)    .

x a x a
sin x
Example 6.6.1. Find Limit .
x0 x
sin x
Solution. Let f ( x)  Here
x
sin h
f (0  0)  Limit f (0  h)  Limit f (h)  Limit
h 0 h 0 h 0 h
3 5 7
hh h h  ....
3! 5! 7!
 Limit
h 0 h
2 4 6
 Limit 1  h h h  ....  1
h 0 3! 5! 7!
And
sin (h)
f (0  0)  Limit f (0  h)  Limit f (h)  Limit
h 0 h 0 h 0 h
sin (h)
 Limit  1.
h 0 h
sin x
Since f (0 + 0) = f (0 - 0) = 1 and hence Limit  1.
h 0 x
sin x
Example 6.6.2.. Find Limit .
x x
sin x
Solution. Let f ( x)  . Put x = 1/y so as x →∞, y→ 0. Then
x
sin x sin (1 / y ) 1

Limit  Limit  Limit y sin  
x x y 0 1/ y y 0
 y
1
Let g ( y )  y sin   . Then, right hand limit is
 y
g (0  0)  Limit g (0  h)  Limit g (h)

h 0 h 0
1
 Limit h sin  
h 0
h
 0  finite quantity which lies between -1 and +1
0
and the left hand limit is
g (0  0)  Limit g (0  h)  Limit g ( h)
h 0 h 0
1
 Limit h sin    0
h 0
h
1
Since g (0 + 0) = g (0 - 0) = 0 therefore Limit y sin    0 and hence
y 0
 y
sin x
Limit  0.
x  x
1
Example 6.6.3. Find lim sin (𝑥).
𝑥⟶∞
1
Solution. Let f ( x)  sin   . Here
 x
1
f (0  0)  Limit f (0  h)  Limit f (h)  Limit sin  
h 0 h 0 h 0
h
1
As h → 0, the value of sin   oscillates between -1 and +1 passing
h
1
through zero. Hence there is no definite number l to which sin   tends
h
to as h → 0. Therefore right hand limit does not exist. Similarly left hand
limit f (0 – 0) also does not exists.
1
Thus lim sin (𝑥) does not exist.
𝑥⟶∞
Example 6.6.4. Find Limit 1  x  x .

1
x0
Solution. Let f ( x)  Limit 1  x  x . Now right hand limit is

1
x 0
f (0  0)  Limit f (0  h)  Limit f (h)  Limit 1  h 

1
h
h 0 h 0 h 0
 11  1  1  1  
 1   1   1  2  
 Limit 1  .h  
h h  2 h  h  h  . h 3  ....
.h 
h 0  h 2! 3! 
 
 
 1 1. 1  h  1. 1  h1  2h  
 Limit 1     ....
h 0
 1! 2! 3! 
1 1 1
1     ....  e
1! 2! 3!
Similarly, the left hand limit is
f (0  0)  Limit f (0  h)  Limit f (h)  Limit 1  h 

1
h e
h 0 h 0 h 0
Thus both f (0 + 0) and f (0 – 0) exists and equal to e. Hence
Limit 1  x 
1
x  e.
x 0
x2
Example 6.6.5. Show that Limit does not exist.
x2 ( x  2)
x2
Solution. Let f ( x)  Limit . Now right hand limit is
x2 ( x  2)
2h2
f (2  0)  Limit f (2  h)  Limit
h 0 h 0 (2  h  2)
h h
 Limit  Limit 1
h 0 ( h) h 0 h
and the left hand limit is
2h2
f (2  0)  Limit f (2  h)  Limit
h 0 h 0 (2  h  2)
h h
 Limit  Limit  1
h 0 (  h) h 0 h
x2
Since f (2 + 0) ≠ f (2 – 0). Hence Limit does not exist.
x2 ( x  2)
1 1x
Example 6.6.6. Find Limit e .
x0 x
1 1x
Solution. Let f ( x)  Limit e . Then
x0 x
1 1h
f (0  0)  Limit f (0  h)  Limit f (h)  Limit e
h 0 h 0 h 0 h
1 1
  (since   and e h   as h → 0)
h
and
1 1
f (0  0)  Limit f (0  h)  Limit f (h)  Limit  e h
h 0 h 0 h 0 h
1
 Limit  1
h 0
he h
1
 Limit 0
h 0  1 1 1 1 1 
h 1   2
 3
 ....
 h 2! h 3! h 
1 1x
Since f (0 + 0) ≠ f (0 – 0). Hence Limit e does not exist.
x0 x
6.7 L HOSPITAL RULE
L’Hospital’s rule is totally different from the quotient law of

differentiation. There is a solid logical base that why we only differentiate
numerator and denominator directly, instead of using quotient law of
differentiation.
(2) It must be clearly remembered that L’Hospital’s method be used only

0 
in the situations of and not in other cases.
0 
(3) In L’Hospital’s rule, numerator f(x) and denominator g(x) are to be

differentiated separately.
(4) It may be helpful for students that

log e 1  0, log e 0  , log e   , e 0  1, e   0 , e   .
sin 𝑥
Example 6.7.1. Evaluate lim .
𝑥⟶0 𝑥
sin x 0
Sol. Clearly, lim is a form.
x 0 x 0
L’Hospital’sRule:
sin 𝑥 (sin 𝑥)′ cos 𝑥
lim = lim = lim = 1.
𝑥→0 𝑥 𝑥→0 (𝑥)′ 𝑥→0 1
Note: In second method, dash ( ′ ) above sin 𝑥 and 𝑥 represents the first
derivative with respect to 𝑥 (variable with respect to the limit has been
taken).
log 𝑥
Example 6.7.2. Evaluate lim .
𝑥⟶0 𝑥−1
log x 0
Sol. Clearly, lim is a   form.
x 0 x  1
0
L’Hospital’s Rule:
1
log 𝑥 (log 𝑥)′ ( )
𝑥
lim = lim (𝑥 – 1)′ = lim = 1.
𝑥→1 𝑥 – 1 𝑥→1 𝑥→1 1
x  sin x
Example 6.7.3. Find lim
x 0 x2
Sol. L’Hospital’s Rule:

(𝑥−sin 𝑥) (𝑥−sin 𝑥)′
lim 𝑥3
= lim (𝑥 3)′
𝑥→0 𝑥→0
(1−cos 𝑥)
= lim 3𝑥 2
𝑥→0
(1−cos 𝑥)′
= lim (3𝑥 2)′
𝑥→0
{−(− sin 𝑥)}

= lim 6𝑥
𝑥→0
(sin 𝑥)′
= lim
𝑥→0 (6𝑥)′
(cos 𝑥)
= lim 6
𝑥→0
1
= .
6
ex 1
lim
Example 6.7.4. Find x 0 x
Sol. L’Hospital’s Method:

𝑒 𝑥 −1 (𝑒 𝑥 −1)′ 𝑒𝑥
lim = lim = lim = 1.
𝑥→0 𝑥 𝑥→0 (𝑥)′ 𝑥→0 1
x cos x  log( 1  x)
lim
Example 6.7.5. Evaluate x 0 x2
Sol. L’Hospital’s Rule:

𝑥𝑐𝑜𝑠 𝑥−log(1+𝑥) 𝟎
lim (𝟎form)
𝑥→0 𝑥2
{𝑥𝑐𝑜𝑠 𝑥−log(1+𝑥)}′
= lim (𝑥 2)′
𝑥→0
1 𝟎
1.cos 𝑥−𝑥(sin 𝑥)−(1+𝑥) (𝟎form)
= lim
𝑥→0 2𝑥
1
(cos 𝑥−𝑥 sin 𝑥−(1+𝑥))′
= lim
𝑥→0 (2𝑥)′
1
− sin 𝑥−(1.sin 𝑥+𝑥.cos 𝑥)+
(1+𝑥) 2
= lim
𝑥→0 2
1
− 2sin 𝑥−𝑥.cos 𝑥+
(1+𝑥)2
= lim
𝑥→0 2
1
= 2.
Note: Observe that L’Hospital’s rule is sometimes easier than the

algebraic method. We will explain next examples only by L’Hospital’s
rule.
(1  x) n  1
lim
Example 6.7.6. Find x0 x
(1+𝑥)𝑛−1 {(1+𝑥)𝑛 −1}′

Sol. lim = lim {𝑥}′
𝑥→0 𝑥 𝑥→0
𝑛(1+𝑥)𝑛−1
= lim
𝑥→0 1
=𝑛
a x  xa
lim
Example 6.7.7. Evaluate x 0 x  a
x a
𝑎 𝑥 −𝑥 𝑎 (𝑎 𝑥 −𝑥 𝑎 )′ 𝑎 𝑥 log 𝑎−𝑎.𝑥 𝑎−1

Sol. lim = lim = lim
𝑥→𝑎 𝑥 𝑥 −𝑎 𝑎 𝑥→𝑎 (𝑥 𝑥 −𝑎 𝑎 )′ 𝑥→𝑎 𝑥 𝑥 (log 𝑥+1)
𝑎 𝑎 log 𝑎−𝑎.𝑎 𝑛−1 𝑎 𝑎 (log 𝑎−1) log 𝑎−1

= = 𝑎𝑎(log 𝑎+1) =
𝑎 𝑎 log 𝑎+𝑎 𝑎 log 𝑎+1
Note: The first derivate of 𝑥 𝑥 in above example calculated as follows:
𝑦 = 𝑥𝑥
Taking logarithms
log 𝑦 = 𝑥 log 𝑥
Now differentiating both sides with respect to 𝑥
1 𝑑𝑦
( ) = log 𝑥 + 1
𝑦 𝑑𝑥
𝑑𝑦
( ) = y(log 𝑥 + 1) = 𝑥 𝑥 (log 𝑥 + 1).
𝑑𝑥
5 sin x  7 sin 2 x  3 sin 3 x

lim
Example 6.7.8. Evaluate x0 tan x  x
5 sin 𝑥−7 sin 2𝑥+3 sin 3𝑥

Sol. lim
𝑥→0 tan 𝑥
(5 sin 𝑥−7 sin 2𝑥+3 sin 3𝑥)′

= lim
𝑥→0 (tan 𝑥−𝑥)′
5 cos 𝑥−7×2 cos 2𝑥+3×3 cos 3𝑥

= lim
𝑥→0 𝑠𝑒𝑐 2 𝑥−1
(5 cos 𝑥−14 cos 2𝑥+9 cos 3𝑥)′

= lim
𝑥→0 (𝑠𝑒𝑐 2 𝑥−1)′
−5 sin 𝑥+14.2 sin 2𝑥−9.3 sin 3𝑥

= lim
𝑥→0 2 sec 𝑥 tan 𝑥
−5 cos 𝑥+28.2 cos 2𝑥−27.3 cos 3𝑥

= lim
𝑥→0 2(sec 𝑥 𝑠𝑒𝑐 2𝑥+sec 𝑥 tan 𝑥 tan 𝑥)
−5+56−81 −30
= = = −15 .
2 2
e x  e sin x
lim
Example 6.7.9. Evaluate x0 x  sin x
𝑒 𝑥 −𝑒 sin 𝑥 (𝑒 𝑥 −𝑒 sin 𝑥 )′
Sol. lim = lim
𝑥→0 𝑥−sin 𝑥 𝑥→0 (𝑥−sin 𝑥)′
𝑒 𝑥 −cos 𝑥.𝑒 sin 𝑥

= lim
𝑥→0 1−cos 𝑥
(𝑒 𝑥 −cos 𝑥.𝑒 sin 𝑥 )′

= lim
𝑥→0 (1−cos 𝑥)′
𝑒 𝑥 −{cos 𝑥.cos 𝑥.𝑒 sin 𝑥 +(− sin 𝑥).𝑒 sin 𝑥 }

= lim
𝑥→0 sin 𝑥
{𝑒 𝑥 −cos2 𝑥.𝑒 sin 𝑥 +sin 𝑥.𝑒 sin 𝑥 }′

= lim
𝑥→0 (sin 𝑥)′
𝑒 𝑥 − {2 cos 𝑥 . (−sin 𝑥). 𝑒 sin 𝑥 + cos 3 𝑥 . 𝑒 sin 𝑥 } + {cos 𝑥. 𝑒 sin 𝑥 + sin 𝑥 cos 𝑥. 𝑒 sin 𝑥 }
= lim
𝑥→0 cos 𝑥
𝑒 𝑥 +3 sin 𝑥 cos 𝑥. 𝑒 sin 𝑥 −cos3 𝑥.𝑒 sin 𝑥 +cos 𝑥. 𝑒 sin 𝑥
= lim cos 𝑥
𝑥→0
1−1+1
= 1
= 1.
∞
Case II: Form
∞
𝑛 2+5
Example 6.7.10. Evaluate 𝑙𝑖𝑚𝑥→0 .
𝑛 2+4𝑛+3
n2  5 
Sol. Clearly, lim is a form.
n  n  4 n  3
2

Algebraic Method:
5 5
𝑛 2+5 𝑛 2(1+ 2 ) (1+ 2 )
𝑛 𝑛
𝑙𝑖𝑚 = 𝑙𝑖𝑚 4 3 = 𝑙𝑖𝑚 4 3 =1 .
𝑛→∞ 𝑛 2+4𝑛+3 𝑛→∞ 𝑛 2(1+𝑛+ 2 ) 𝑛→∞ (1+𝑛+ 2 )
𝑛 𝑛
L’Hospital’s Method:
𝑛 2+5
𝑙𝑖𝑚
𝑛→∞ 𝑛 2+4𝑛+3
∞
(Again ∞ form)
(2𝑛)′
= 𝑙𝑖𝑚 (2𝑛+4)′
𝑛→∞
2
= 𝑙𝑖𝑚
𝑛→∞ 2
=1.
log x
lim
Example 6.7.11. Evaluate x 0 cot x
∞
Sol. This is of the form ∞. We have therefore,
1
𝑙𝑜𝑔 𝑥 (𝑙𝑜𝑔 𝑥 )′ (𝑥 ) ∞
lim = lim ′
= lim 2
( 𝑓𝑜𝑟𝑚)
𝑥→0 𝑐𝑜𝑡 𝑥 𝑥→0 (𝑐𝑜𝑡 𝑥 ) 𝑥→0 −𝑐𝑜𝑠𝑒𝑐 𝑥 ∞
− 𝑠𝑖𝑛2 𝑥 0
= lim ( 𝑓𝑜𝑟𝑚)
𝑥→0 𝑥 0
−2𝑠𝑖𝑛 𝑥 𝑐𝑜𝑠 𝑥
= lim = 0.
𝑥→0 1

log( x  )
lim 2
 tan x
x
Example 6.7.12. Find 2

log( x  )
2 is a ∞
Sol. lim form.
 tan x ∞
x
2
We have,
𝜋 𝜋
log(𝑥− ) [log(𝑥− )]′
2 2
lim𝜋 = lim𝜋 =
𝑥→ tan 𝑥 𝑥→ (tan 𝑥)′
2 2
1
( 𝜋 )
(𝑥− ) ∞
2
lim (∞ 𝑓𝑜𝑟𝑚)
𝑥→
𝜋 𝑠𝑒𝑐 2 𝑥
2
𝑐𝑜𝑠 2𝑥 0
= lim𝜋 𝜋 ( 0 𝑓𝑜𝑟𝑚)
𝑥→ (𝑥− 2 )
2
(𝑐𝑜𝑠 2𝑥)′
= lim𝜋 𝜋
𝑥→ (𝑥− )′
2 2
−2 cos 𝑥 sin 𝑥
= lim𝜋
𝑥→ 1
2
= 0.
log( x  a )
lim
x  a log( e x  e a )
Example 6.7.13. Evaluate
log(𝑥−𝑎) ∞ (log(𝑥−𝑎))′
Sol. lim ( 𝑓𝑜𝑟𝑚) = lim (log(𝑒 𝑥 −𝑒 𝑎))′
𝑥→𝑎 log(𝑒 𝑥 −𝑒 𝑎 ) ∞ 𝑥→𝑎
1
( )
𝑥−𝑎
= lim 1
𝑥→𝑎 ( 𝑥 𝑎 )𝑒 𝑥
𝑒 −𝑒
𝑒 𝑥 −𝑒 𝑎 0
= lim (𝑥−𝑎)𝑒 𝑥 ( 𝑓𝑜𝑟𝑚)
𝑥→𝑎 0
(𝑒 𝑥 −𝑒 𝑎 )′
= lim [(𝑥−𝑎)𝑒 𝑥 ]′
𝑥→𝑎
𝑒𝑥
= lim (𝑥−𝑎)𝑒 𝑥 +𝑒 𝑥
𝑥→𝑎
𝑒𝑥
= lim [(𝑥−𝑎)+1]𝑒 𝑥
𝑥→𝑎
1
= lim [(𝑥−𝑎)+1]
𝑥→𝑎
=1
e x  3x 3
lim
Example 6.7.14. Find x 4e  4 x
x
Sol.
𝑒 𝑥 +3𝑥 3 ∞
lim (∞ 𝑓𝑜𝑟𝑚)
𝑛→∞ 4𝑒 𝑥 +4𝑥
′
(𝑒 𝑥 +3𝑥 3 )
= lim
𝑛→∞ (4𝑒 𝑥 +4𝑥)′
𝑒 𝑥 +9𝑥 2 ∞
= lim (∞ 𝑓𝑜𝑟𝑚)
𝑛→∞ 4𝑒 𝑥 +4
(𝑒 𝑥 +9𝑥 2 )′
= lim
𝑛→∞ (4𝑒 𝑥 +4)′
𝑒 𝑥 +18𝑥 1 ∞
𝑛→∞ 4𝑒 𝑥
(𝑒 𝑥 +18𝑥 1)′
= lim (4𝑒 𝑥 )′
𝑛→∞
(𝑒 𝑥 +18) ∞
𝑛→∞ 4𝑒 𝑥
𝑒𝑥
= lim
𝑛→∞ 4𝑒 𝑥
1
= 4.
log(tan 2 2 x)
lim
x0 log(tan 2 x )
Sol. We have,
log(𝑡𝑎𝑛2 2𝑥) ∞
lim ( 𝑓𝑜𝑟𝑚)
𝑥→0 log(𝑡𝑎𝑛 2 𝑥 ) ∞
2 log(tan 2𝑥) ∞
𝑥→0 2 log(tan 𝑥)
1
(log(tan 2𝑥))′ ( ).2 𝑠𝑒𝑐 2 2𝑥
tan 2𝑥
= lim = lim 1
𝑥→0 (log(tan 𝑥))′ 𝑥→0 ( ). 𝑠𝑒𝑐 2𝑥
tan 𝑥
2 tan 𝑥 𝑐𝑜𝑠 2𝑥 2 sin 𝑥 cos 𝑥

= lim tan 2𝑥 𝑐𝑜𝑠 22𝑥 = lim sin 2𝑥 cos 2𝑥
𝑥→0 𝑥→0
sin 2𝑥 1 1
= lim sin 2𝑥 cos 2𝑥 = lim cos 2𝑥 = 1 = 1.
𝑥→0 𝑥→0
log(sin x)
lim
x 0 cot x
Sol. We have,
log(sin 𝑥) ∞ (log(sin 𝑥))′ ′

lim (∞ 𝑓𝑜𝑟𝑚) = lim
𝑥→0 cot 𝑥 𝑥→0 (cot 𝑥)′
1
.cos 𝑥 cos 𝑥
= lim sin 𝑥
= lim (− sin 𝑥 . 𝑠𝑖𝑛2 𝑥)
𝑥→0−𝑐𝑜𝑠𝑒𝑐 2 𝑥 𝑥→0
= lim (− cos 𝑥. sin 𝑥) = 0.

𝑥→0
xn
Example 6.7.17. Find lim , where 𝑛 is a positive integer.
n  e x
Sol. We have,
𝑥𝑛 ∞ (𝑥 𝑛 )′ 𝑛𝑥 𝑛−1 ∞
lim ( 𝑓𝑜𝑟𝑚) = lim = lim (∞ 𝑓𝑜𝑟𝑚)
𝑥→∞ 𝑒𝑥 ∞ 𝑥→∞ (𝑒 𝑥 )′ 𝑥→∞ 𝑒𝑥
′
(𝑛𝑥 𝑛−1) 𝑛(𝑛−1)𝑥 𝑛−2 ∞
= lim = lim (∞ 𝑓𝑜𝑟𝑚)
𝑥→∞ (𝑒 𝑥 )′ 𝑛→∞ 𝑒𝑥
′
(𝑛(𝑛−1)𝑥 𝑛−2) 𝑛(𝑛−1)(𝑛−2)𝑥 𝑛−3 ∞
= lim = lim (∞ 𝑓𝑜𝑟𝑚)
𝑛→∞ (𝑒 𝑥 )′ 𝑛→∞ 𝑒𝑥
Repeating this process, we get

(𝑛(𝑛−1)(𝑛−2)…𝑛 𝑓𝑎𝑐𝑡𝑜𝑟𝑠)
= lim
𝑛→∞ 𝑒𝑥
𝑛! 𝑛! 𝑛!
= lim = = ∞ = 0.
𝑛→∞ 𝑒 𝑥 𝑒∞
log sin 2 x
Example 6.7.18. Find lim
x 0 log sin x
Sol. We have,
log sin 2𝑥 ∞
lim ( 𝑓𝑜𝑟𝑚)
𝑥→0 log sin 𝑥 ∞
(log sin 2𝑥)′

= lim
𝑥→0 (log sin 𝑥)′
2
( .cos 2𝑥)
sin 2𝑥
= lim 1
𝑥→0 ( .cos 𝑥)
sin 𝑥
2 cot 2𝑥 ∞
𝑥→0 cot 𝑥
(2 cot 2𝑥)′
= lim
𝑥→0 (cot 𝑥)′
−4 𝑐𝑜𝑠𝑒𝑐 2 2𝑥 ∞
𝑥→0 −𝑐𝑜𝑠𝑒𝑐 2 𝑥
4 𝑠𝑖𝑛 2 𝑥
= lim 𝑠𝑖𝑛2 2𝑥
𝑥→0
4 𝑠𝑖𝑛 2 𝑥
= lim (2 sin 𝑥 cos 𝑥)2
𝑥→0
1
= lim 𝑐𝑜𝑠 2𝑥 = 1.
𝑥→0
log 𝑥
Example 6.7.19. Find lim , 𝑎 > 1.
𝑥⟶∞ 𝑎 𝑥
Sol. We have,
log 𝑥 ∞ (log 𝑥)′
lim (∞ 𝑓𝑜𝑟𝑚) = lim
𝑥→∞ 𝑎 𝑥 𝑥→∞ (𝑎 𝑥 )′
1
( )
𝑥
= lim
𝑥→∞ 𝑎 𝑥 log 𝑎
1 1
= lim
log 𝑎 𝑥→∞ 𝑥 𝑎 𝑥
1
= log 𝑎 × 0 = 0.
CHECK YOUR PROGRESS
sin 𝑥
Problem 1. lim is 0.
𝑥⟶0 𝑥
log 𝑥
Problem 2. lim is 1.
𝑥⟶0 𝑥−1
1
Problem 3. lim sin (𝑥) does not exist.
𝑥⟶∞
√4+𝑥−2
Problem 4. 𝑙𝑖𝑚𝑥→0 is 1.
𝑥
𝑛 2 +5
Problem 5. 𝑙𝑖𝑚𝑥→0 is 1.
𝑛 2 +4𝑛+3
6.8 SUMMARY
1. 𝐥𝐢𝐦 𝒇(𝒙) = L
𝒙→𝒂
This is read as, “limit of f(x), as x approaches a equals L”.
2. A symbol such as x or y, used to represent an arbitrary element of a set
is called a variable. For example y = f (x).
0 
3. L’Hospital’s method be used only in the situations of and not in
0 
other cases.
6.9 GLOSSARY
Numbers
Intervals
Limit points
Functions
Bounded, Unbounded sets
6.10 REFERENCES


5𝑛 2−5
Q1. Prove that 𝑙𝑖𝑚𝑥→0 is 5.
𝑛 2 +4𝑛+3
1
Q2. Prove that 𝑙𝑖𝑚𝑥→0 sin (𝑥 2) does not exist.
1
Q3. Prove that 𝑙𝑖𝑚𝑥→0 𝑥 2 sin (𝑥 2) exist and equal to 0.
1
Q4. Prove that lim 9 𝑥−1 does not exist.
𝑥→1
15 log 𝑥
Q5. Prove that lim , 𝑎 > 1 is 0.
𝑥⟶∞ 𝑎𝑥
6.13 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. False
CYQ 2. True
CYQ 3. True
CYQ 4. False
CYQ 5. True
UNIT 7: CONTINUITY
CONTENTS:
7.1 Introduction
7.2 Objectives
7.3 Continuity (𝜀, 𝛿) definition
7.4 Continuity (by L.H.L. and R.H.L.)
7.5 Discontinuity
7.6 Type of Discontinuity
7.7 Uniformly continuity
7.8 Summary
7.9 Glossary
7.10 References
7.13 Answers
7.1 INTRODUCTION
In mathematics, a continuous function is a function such that a

small variation of the argument induces a small variation of the value of
the function. This implies there are no abrupt changes in value, known
as discontinuities. More precisely, a function is continuous if arbitrarily
small changes in its value can be assured by restricting to sufficiently
small changes of its argument. A discontinuous function is a function
that is not continuous. Until the 19th century, mathematicians largely
relied on intuitive notions of continuity and considered only continuous
functions.
Continuity is one of the core concepts of calculus and mathematical

analysis, where arguments and values of functions
are real and complex numbers.
7.2 OBJECTIVES
(i) Continuity
(ii) Discontinuity
(iii) Type of Discontinuity
(iv) Uniformly continuous
7.3 CONTINUITY (𝝐, 𝜹) DEFINITION
Definition 1. A real valued function f (x) defined on an interval I is said to

be continuous at x  a  I if and only if for any arbitrarily chosen positive
number  , however small, we can find a corresponding number   0 such
that
f ( x)  f (a)   whenever x  a   .
We say that f (x) is continuous if it is continuous at every x  I .
Or
f (x) is continuous at x = a is given   0 , we can find a   0 such that
x  a    f ( x)  f ( a )   .
7.4 CONTINUITY FROM LEFT AND

CONTINUITY FROM RIGHT
A function f (x) is said to be continuous from left at x = a if Limit f ( x)

x  a 0
exists and equal to f (a) i.e.,
Limit f (a  h)  f (a)
h 0
Similarly, f (x) is said to be continuous from right at x = a if Limit f ( x)

x a  0
exists and equal to f (a) i.e.,
Limit f (a  h)  f (a)
h 0
and f (x) is continuous at x = a iff
Limit f ( x)  Limit f ( x)  f (a)

x  a 0 x a 0
Limit f (a  h)  Limit f (a  h)  f (a)

h 0 h 0
7.5 DISCONTINUITY
If a function is not continuous at a point, then it is said to be

discontinuous at that point and the point is called a point of discontinuity
of the function.
1
e. g. The function f ( x )  does not exists at x = a so f (x) is not
xa
continuous at x = a.
7.6 TYPES OF DISCONTINUITY
(1) Removable discontinuity:
A function f (x) is said to have a removable discontinuity at a point x = a

if Limit f (x) exist but is not equal to f (a) i.e., if
x a
f (a  0)  f (a  0)  f (a)
The function can be made continuous by defining it in such a way that
Limit f ( x)  f (a) .
x a
(2) Discontinuity of the first kind:
A function f (x) is said to have a discontinuity of the first kind or ordinary

discontinuity at x = a if f (a + 0) and f (a – 0) both exist but not equal. The
point x = a is said to be a point of discontinuity from the left or right
according as
f (a  0)  f (a)  f (a  0) or f (a  0)  f (a)  f (a  0) .
(3) Discontinuity of the second kind:
A function f (x) is said to have a discontinuity of the second kind at x = a

if none of the limits f (a + 0) and f (a – 0) exist. The point x = a is said to
be a point of discontinuity of second kind from the left or right according
as f ( a  0) or f ( a  0) does not exist.
(4) Mixed discontinuity:
If a function f has a mixed discontinuity at ‘𝑎’if either
(i) lim− 𝑓(𝑥) does not exist and lim+ 𝑓(𝑥) exists, however lim+ 𝑓(𝑥)
𝑥⟶𝑎 𝑥⟶𝑎 𝑥⟶𝑎
may or may not equal to 𝑓(𝑎).
(ii) lim+ 𝑓(𝑥) does not exist and lim− 𝑓(𝑥) exists, however lim− 𝑓(𝑥)
𝑥⟶𝑎 𝑥⟶𝑎 𝑥⟶𝑎
may or may not equal to 𝑓(𝑎).
(5) infinite discontinuity:
A function f (x) is said to have an infinite discontinuity at x = a if f (a + 0)

or f (a – 0) is + ∞ or - ∞ i.e., if f (x) is discontinuous at x = a and f (x) is
unbounded in every neighbourhood of x = a.
(6) Piecewise continuous function:
A function 𝑓: 𝐴 ⟶ ℝ is said to be piecewise continuous on 𝐴 if 𝐴 can be

divided into a finite number of parts so that 𝑓 is continuous on each part.
Clearly, in such a case, 𝑓 has a finite number of discontinuities and the set
𝐴 is divided at the points of discontinuities.
Note: Jump of a function at a point.
If both f (a + 0) and f (a – 0) exists, then the jump in the function at x = a

is defined as the non-negative difference f (a  0) ~ f (a  0) .
A function having a finite number of jumps in a given interval is called

piecewise continuous.
Illustrative Examples
Example 1. Test the continuity of f (x) at x = 1 when
x 2  2 if x  1

f ( x )  2 x  1 if x  1
3 if x  1

Solution. Here f (1)  2.1  1  3
f (1  0)  Limit f (1  h)  Limit (1  h) 2  2
h 0 h0
 Limit 1  h 2  2 h  2  3 as1  h 1.

h0
f (1  0)  Limit f (1  h)  Limit (1  h) 2  2
h0 h0
 Limit 1  h  2 h  2  3 as1  h  1.
2
h 0
So f (1)  f (1  0)  f (1  0) .Hence f (x) is continuous at x = 1.
1
Example 2. Discuss the continuity of the function f ( x)  1
when
1 e x
x ≠ 0 and f (0)  0 for all values of x.
Solution. Test the continuity at x = 0
f (0  0)  Limit f (0  h)  Limit f (h)

h0 h0
1
 Limit 1
1
h0
1 e h
f (0  0)  Limit f (0  h)  Limit f (h)

h0 h0
1
 Limit 1
0
h0
1 e h
Thus we have f (0 + 0) ≠ f (0 – 0) = f (0). So f (x) is not continuous at x =

0 and it is a discontinuity of first kind i.e., f (x) is continuous on the left
and has a discontinuity of first kind on right at x = 0.
Now test the continuity at x = a ≠ 0
1
f (a)  1
1 e a
1
f (a  0)  Limit f (a  h)  Limit 1
h0 h0 ah
1 e
1
 1
 f (a)
1 e a
1
f (a  0)  Limit f (a  h)  Limit 1
h0 h0 ah
1 e
1
 1
 f (a)
1 e a
Thus we have f (a + 0) = f (a – 0) = f (a). Hence f (x) is continuous at

every point except x = 0.
Example 3. Test the continuity of the function
 1
 x sin if x  0
f ( x)   x

0 if x  0
Solution. Here
f (0  0)  Limit f (0  h)  Limit f (h), h  0

h0 h0
1
 Limit h sin 0
h0 h
f (0  0)  Limit f (0  h)  Limit f (h), h  0

h0 h0
 1 1
 Limit (h ) sin     Limit h sin  0
h0
 h  h0 h
Thus we have f (0 + 0) = f (0 – 0) = f (0). Hence f (x) is continuous at x =

0.
Note. 1. If we check the continuity at x = c ≠ 0 of the above function,

then we see that
1
Limit f ( x)  Limit x sin
xc xc x
1
 c sin  f (c )
c
So f (x) is continuous at x = c. Thus f (x) is continuous for all x R i.e., f

(x) is continuous on the whole real line.
Note 2. If we take f (0) = 2, in the above function, then f (0 + 0) = f (0 –

0) ≠ f (0). The function becomes discontinuities at x = 0 and has a
removable discontinuity at x = 0.
Example 4. If a function f (x) is defined by f ( x)  x  x, where x is a

positive variable and [x] denotes the integral part of x. Show that it is
discontinuous for integral values of x and continuous for all others. Draw
the graph.
Solution. From the definition of the function f (x) we have
 x  (n  1) for n  1  x  n

f ( x)  0 for x  n
x  n for n  x  n  1
 where n is an integer
First we test the continuity of f (x) at x = n. We have f (n) = 0.
f (n  0)  Limit f (n  h)  Limit (n  h)  n
h 0 h 0
 Limit h  0 [as n < n + h < n+1]

h 0
f (n  0)  Limit f (n  h)  Limit (n  h)  n  1
h0 h0
 Limit 1  h  1 [ as n-1 < n-h < n]

h 0
Since f (n – 0) ≠ f (n + 0), so the function f (x) is discontinuous at x = n.

Thus f (x) is discontinuous for all integral values of x. it is obviously
continuous for all other values of x.
Since x is a positive variable putting = 1, 2, 3, 4, 5, …., we see that graph

of the function consists of the following straight lines.
x when 0  x  1
0 when x  1

 x 1 when 1  x  2

0 when x  2
y  f ( x)  
x  2 when 2  x  3
0 when x  3

x  3 when 3  x  4
0 when x  5

and so on.
𝑌
𝑦=1
𝑥=3 𝑋
𝑥=1 𝑥=2 𝑥=4
It is clear from the graph that
(1) The function is discontinuous for all integral values of x but

continuous for other values of x.
(2) The function is bounded between 0 and 1 in every domain which
includes an integer.
(3) The lower bound 0 is attained but upper bound 1 is not attained since
f (x) ≠ 1 for any value of x.
Example 5. Show that the function f (x) = [x] + [-x] has a removable
discontinuity for integral values of x.
Solution. We see that f (x) = 0 when x is an integer and f (x) = -1 when x

is not an integer. Hence if n is an integer then
f (n – 0) = f (n + 0) = -1 and f (n) = 0.
So the function f (x) has a removable discontinuity at x = n, where n is an

integer.
x
Example 6. Prove that the function f ( x)  for x ≠ 0 and f (0) = 0, is
x
continuous at all the points except x = 0.
x x
Solution. If x > 0 then, f ( x)   1 and if x < 0 then, f ( x)   1.
x x
Therefore the given function can define as:
 1, if x  0

f ( x)  0, if x  0
 1, if x  0

If x < 0, then f (x) = -1 i.e., f (x) is a constant function and a constant

function is always continuous at each point of its domain. This implies
that f (x) is continuous for all x < 0.
Similarly, we can show that f (x) is continuous for all x > 0. Now we see
the continuity at x = 0.
f (0  0)  Limit f (0  h)  Limit f (h), h  0

h0 h0
 Limit 1  1
h0
f (0  0)  Limit f (0  h)  Limit f (h), h  0

h0 h0
 Limit  1  1
h0
Here f (0  0) ≠ f (0  0) ≠ f (0) . Hence f (x) is not continuous at x = 0.
Example 7. Discuss the continuity of the following functions at 𝑥 = 0 of

1
𝑥 2 sin , 𝑥 ≠ 0
the function 𝑓(𝑥 ) = { 𝑥 .
0 , 𝑥=0
Sol.
7.7 UNIFORM CONTINUITY
A function f: X → Y is said to be uniformly continuous on A ⊆ X if for

every ε > 0, there exists δ > 0 such that x, y ∈ A, |𝑥 − 𝑦| < δ implies
|𝑓(𝑥) − 𝑓(𝑦)| < ε.
Note: Uniform continuity is domain base property i.e. Uniform continuity

is defined on a set.
Example 7.6.1 Prove that the function f (x) = 3x + 1 is uniformly

continuous on ℝ.
Solution: Since |f (x) − f (y)| = |(3x + 1) − (3y + 1)| = 3|x − y|
so, given ε > 0, we choose δ = ε / 3. Then, |x − c| < δ implies
|f (x) − f (c)| = 3|x − c| < 3(ε / 3) = ε.
Hence the function f (x) = 3x + 1 is uniformly continuous on ℝ.
Example 7.6.2 Prove that the function f (x) = 𝑥 2 is not uniformly

continuous on ℝ.
Solution: for any ϵ > 0and for any x, y ∈ ℝ we have a δ

such that |x−y| < δ ⟹ |f(x)−f(y)| < ϵ. Our claim here is now
|𝑥 2 −𝑦 2 | < ϵ. That is the distance between 𝑥 2 and 𝑦 2 is at
most ϵ everywhere on the vertical axis as long as we keep 𝑥 and 𝑦 at
most δ away from each other. The problem arises if we let 𝑥 and 𝑦 getting
larger and larger.
So for the given ϵ we have fixed δ and we may play with 𝑥 and 𝑦 values.
δ ϵ ϵ
Let's make them large enough to satisfy < | 𝑥 − y | < δ and 𝑥 > , y > .
2 δ δ
Then | 𝑥 − y |< δ
δ 2ϵ
But |𝑥 2 −𝑦 2 | = | 𝑥 − y || 𝑥 + y | > ⟹ |𝑥 2 −𝑦 2 | > δ
2 δ
which is a contraction with the definition which is assumed true at the

beginning. Therefore f (x) = 𝑥 2 can not be uniformly continuous.
Note: Every Uniformly continuous function is continuous function.
CHECK YOUR PROGRESS
1
Problem 1. The function 𝑓 (𝑥 ) = sin at 𝑥 = 0 has a discontinuity of
𝑥
second kind.
Problem 2. The function 𝑓 (𝑥 ) = 𝑒 𝑥 is not continuous at 𝑥 = 0.
Problem 3. Every uniformly continuous function is continuous
function.
Problem 4. Every polynomial function is uniformly continuous on ℝ.
Problem 5. Every continuous function uniformly continuous.
7.8 SUMMARY
1. If Limit f (a  h)  Limit f (a  h)  f (a) then function f(x) is

h 0 h 0
continuous at 𝑥 = 𝑎.
2. A function f: X → Y is said to be uniformly continuous on A ⊆ X if

for every ε > 0, there exists δ > 0 such that x, y ∈ A, |𝑥 − 𝑦| < δ implies
|𝑓(𝑥) − 𝑓(𝑦)| < ε.
7.9 GLOSSARY
Numbers
Intervals
Sets
Functions
Limits
7.10 REFERENCES
1. T. M. Apostol, Mathematical Analysis (2 nd Edition), Narosa

3. W. Rudin, Principles of Mathematical Analysis (3 rd Edition),
4. S.C. Malik and Savita Arora, Mathematical Analysis (6 th Edition),

5. Shanti Narayan, A course of Mathematical Analysis (29 th Edition),
6. K. A. Ross, Elementary Analysis, The Theory of Calculus (2 nd

Q 1. Prove that every uniform continuous function is continuous.
Q 2. Prove that every polynomial function is continuous.
1
Q 3. Prove that 𝑓 (𝑥 ) = is not uniformly continuous in (0, 1).
𝑥
Q 4. Prove that 𝑓 (𝑥 ) = 𝑥 2 is uniformly continuous on [-2, 2].

Q 5. Prove that 𝑓 (𝑥 ) = 21/𝑥 is not continuous at 0.
7.13 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. False
CYQ 3. True
CYQ 4. False
CYQ 5. False
UNIT 8: DIFFERENTIATION
Contents
8.1 Introduction
8.2 Objectives
8.3 Derivative
8.4 Mean Value Theorem
8.5 Taylor’s theorem
8.6 Summary
8.7 Glossary
8.9 References
8.11 Answers
8.1 INTRODUCTION
In previous unit we discussed about limit and continuity. In this

unit we will discussed about Differentiability and Mean Value theorem.
Prior to the seventeenth century, a curve was generally described as a
locus of points satisfying some geometric condition, and tangent lines
were obtained through geometric construction. This viewpoint changed
dramatically with the creation of analytic geometry in the 1630s by Rene
Descartes (1596–1650) and Pierre de Fermat (1601–1665).
In this new setting geometric problems were recast in terms of algebraic
expressions, and new classes of curves were defined by algebraic rather
than geometric conditions. The concept of derivative evolved in this new
context. The problem of finding tangent lines and the seemingly unrelated
problem of finding maximum or minimum values were first seen to have
a connection by Fermat in the 1630s. And the relation between tangent
lines to curves and the velocity of a moving particle was discovered in the
late 1660s by Isaac Newton. Newton’s theory of ‘‘fluxions,’’ which was
based on an intuitive idea of limit, would be familiar to any modern
student of differential calculus once some changes in terminology and
notation were made.
But the vital observation, made by Newton and, independently, by
Gottfried Leibniz in the 1680s, was that areas under curves could be
calculated by reversing the differentiation
process. This exciting technique, one that solved previously difficult area
problems with ease, sparked enormous interest among the mathematicians
of the era and led to a coherent theory
that became known as the differential and integral calculus.
In this Unit we will develop the theory of differentiation. Integration
theory, including the fundamental theorem that relates differentiation and
integration, will be the
subject of the next chapter. Consequently, we will concentrate on the
mathematical aspects of the derivative and not go into its applications in
geometry, physics, economics, and so on.
8.2 OBJECTIVES
In this Unit, we will Discussed about
 Improper integral
 Test of convergence
 Absolute integral
8.3 DERIVATIVE
We begin with the definition of the derivative of a function.
Derivative: Let I ⊆ ℝ be an interval, f: (x, y) → ℝ and b ∈ I. Then l ∈ ℝ

is said to be derivative of f at b if for any given ε > 0 there exists δ(ε)
such that if x ∈ I then
f(x)−f(b)
| − l| < ε whenever 0 < |x − b| < δ.
x−b
We can also say that f is differentiable at b, and we write f ′ (c)
Or
f(x)−f(b)
The derivative of f at b is given by f ′ (c) = lim provided this
h→0 x−h
limit exists.
Note: We now show that continuity of f at a point b is a necessary
(but not sufficient) condition for the existence of the derivative at b.
Theorem: If f ∶ 𝐼 → ℝ has a derivative at 𝑏 ∈ 𝐼, then 𝑓 is continuous

at 𝑏.
Proof. We have
f(x)−f(b)
f(x) − f(b) = ( ) (x − b) For all x ∈ I; x ≠ b
x−b
Because f ′ (b) exists, Therefore
f(x) − f(b)
lim (f(x) − f(b)) = lim (( ) (x − b))
x→b x→b x−b
f(x)−f(b)
= lim ( ) lim (x − b)
x→b x−b x→b
= f ′ (b). 0 = 0
Therefore, lim (f(x) − f(b)) = 0 ⇒ lim f(x) − lim f(b) ⇒ lim f(x) = f(b)
x→b x→b x→b x→b
Hence f is continuous at b.
NOTE: The continuity of f ∶ I → ℝ at a point does not promise the

existence of the derivative at that point.
Theorem: Let I ⊆ ℝ be an Interval and f, g: X → ℝ be functions that are

differentiable at b ∈ ℝ Then
(i) If α ∈ ℝ, then the function αf is differentiable at b and
(αf)′ (b) = αf ′ (b)
(ii) The function f+g is differentiable at b and
(f + g)′ (b) = f ′ (b) + g′ (b)
(iii) The function f and g is differentiable at b and
(fg)′ (b) = f ′ (b)g(b) + f(b)g′ (b)
(iv) If g(b) ≠ 0, then the function f and g is differentiable at b and
f ′ f′ (b)g(b)−f(b)g′ (b)
( ) ( b) = 2
g (g(b))
Proof. (i) Let h1 = αf, then for x ∈ I and x ≠ b, we have

h1(x)−h1(b) (αf)(x)−(αf)(b) f(x)−f(b)
= =α
x−b x−b x−b
Since f is differentiable at b implies f ′ (b) exists. Therefore

h1(x)−h1(b) f(x)−f(b) f(x)−f(b)
lim = lim α = αlim = αf ′ (c)
x→b x−b x→b x−b x→b x−b
Hence (αf)′ (b) = αf ′ (b)
(ii) Let h2 = f + g, then for x ∈ I and x ≠ b, we have

h2(x)−h2(b) (f+g)(x)−(f+g)(b) f(x)+g(x)−f(b)−g(b) f(x)−f(b)+g(x)−g(b)
= = =
x−b x−b x−b x−b
f(x)−f(b) g(x)−g(b)
= +
x−b x−b
Since f and g are differentiable at b implies f ′ (b) and g′ (b) exists.

Therefore
h2(x)−h2(b) f(x)−f(b) g(x)−g(b)
lim = lim + lim = f ′ (b) + g′ (b)
x→b x−b x→b x−b x→b x−b
Hence (f + g) ′ (b) = f ′ (b) + g′ (b)

(iii) Let h3 = fg, then for x ∈ I and x ≠ b, we have
h3 (x) − h3 (b) (fg)(x) − (fg)(b) f(x)g(x) − f(b)g(b)

= =
x−b x−b x−b
f(x)g(x) − f(b)g(x) + f(b)g(x) − f(b)g(b)
=
x−b
g(x)(f(x)−f(b))+f(b)(g(x)−g(b)) f(x)−f(b) g(x)−g(b)
= = g ( x) + f(b) .
x−b x−b x−b
It is given that f and g is differentiable at b and
g is differentiable at b ⇒ g is continuous i.e. lim g(x) = g(b) (by

x→b
previous theorem)
Therefore
h3(x)−h3(b) f(x)−f(b) g(x)−g(b)
lim = lim {g(x) + f(b) }
x→b x−b x→b x−b x−b
f(x)−f(b) g(x)−g(b) f(x)−f(b)

= lim g(x) + lim f(b) = lim g(x) lim +
x→b x−b x→b x−b x→b x→b x−b
g(x)−g(b)
f(b) lim = f ′ (b)g(b) + f(b)g′ (b)
x→b x−b
Hence (fg)′ (b) = f ′ (b)g(b) + f(b)g′ (b)

f
(iv) Let h4 = , since g is differentiable at b ⇒ since g is continuous at b.
g
It is given that g(b) ≠ 0, therefore there exists an interval I1 ⊆ I with

b ∈ I1 such that
g(x) ≠ 0 for all x ∈ I1 .
Now for x ∈ I1 , x ≠ b, we get
f f f ( x) f( b )
h4 (x) − h4 (b) g ( x ) − ( b) −
g g(x) g(b) f(x)g(b) − f(b)g(x)
= = =
x−b x−b x−b g(b)g(b)(x − b)
f(x)g(b) − f(b)g(x) f(x)g(b) − f(b)g(b) + f(b)g(b) − f(b)g(x)

= =
g (b)g(b)(x − b) g(b)g(b)(x − b)
(f(x)−f(b))g(b)−f(b)(g(x)−g(b)) 1 f(x)−f(b) g(x)−g(b)

= [ . g(b) − f(b). ]
g(x)g(b)(x−b) g(x)g(b) x−b x−b
Therefore
h4(x)−h4(b) 1 f(x)−f(b) g(x)−g(b)
lim = lim [ . g(b) − f(b). ]
x→b x−b x→b g(x)g(b) x−b x−b
1 f(x)−f(b)
= lim [lim ( ) . g(b) −
x→b g(x)g(b) x→b x−b
g(x)−g(b)
f(b). lim ( )]
x→b x−b
1
= 2 . [f ′ (b)g(b) − f(b)g′ (b)]
g′ (b)
Hence
f ′ f′ (b)g(b)−f(b)g′ (b)
( ) (b) = 2
g (g(b))
8.4 MEAN VALUE THEOREM
The Mean Value Theorem, which relates the values of a function to

values of its derivative,
is one of the most useful results in real analysis
We begin by looking at the relationship between the relative extrema of

a function and
the values of its derivative.
Relative Maximum: The function f ∶ I → ℝ is said to have a

relative maximum at b ∈ I if there exists a neighborhood V = Vδ (b) of b
such that f(x) ≤ f(b), for all x in V ∩ I.
Relative Minimum: The function f ∶ I → ℝ is said to have a

relative minimum at b ∈ I if there exists a neighborhood V ′ = V ′ δ′ (b)
of b such that f(x) ≥ f(b), for all x in V ′ ∩ I.
Relative Extremum: f has a relative extremum at b ∈ I if it has

either a relative maximum or a relative minimum at b.
Interior Extremum Theorem
Theorem: Let b be an interior point of the interval I at which

f ∶ I → R has a relative extremum. If the derivative of f at b exists, then
f ′ (b) = 0.
Proof. If f ’ (b) > 0, Then there exists a neighborhood V ⊆ I of b such

that
f(x)−f(b)
> 0 for x ∈ V, x ≠ b
x−b
If x ∈ V, x > b, then we get

f(x)−f(b)
f(x) − f(b) = (x − b). >0
x−b
But this contradicts the hypothesis that f has a relative maximum at b.
Hence, we cannot have f ’ (b) > 0.
Similarly, we cannot have f ’ (b) < 0.
Therefore, f ’ (b) = 0.
Rolle’s Theorem
Theorem: Consider that f is continuous on a closed interval I = [a, b]
and
the derivative f ′ (0) exists at every point of the open interval (a, b), and
f(a) = f(b) = 0.
Then there exists at least one point c in (a, b) such that f ′ (c) = 0
Proof. If f(x) = 0 for all x in I or vanishes identically on I, then any c

in (a, b) will satisfy the result of the theorem.
Hence Let f does not vanish identically or f ≠ 0.
Now replacing f by (−f) and consider f assumes some positive values.
So by the Maximum Minimum Theorem,
The function f attains the value sup{f(x): x ∈ I} > 0 at some point c in I.
Since f(a) = f(b) = 0. the point c must lie in (a, b).
Hence f ′ (c) exists.
Since f has a relative maximum at c.
By the Interior Extremum Theorem, we get
f ′ (c ) = 0
Geometrical Representation of Rolle’s theorem
Fig.
In the given graph, the curve y = f(x) is continuous between x = a

and x = b and
at every point, within the interval, it is possible to draw a tangent and

ordinates corresponding to the abscissa and are equal then there exists at
least one tangent to the curve which is parallel to the x-axis.
Algebraically, this theorem tells us that if f (x) is representing a

polynomial function in x and the two roots of the equation f(x) = 0 are
x = a and x = b, then there exists at least one root of the equation
f′(x) = 0 lying between these values.
The converse of Rolle’s theorem is not true and it is also possible that
there exists more than one value of x, for which the theorem holds good
but there is a definite chance of the existence of one such value.
NOTE:
 Rolle’s theorem does not hold good if

(i) f (x) is discontinuous in the closed interval [a, b].
(ii) f (x) does not exist at some point in (a, b).
(iii) f (a) ≠ f (b).
Example: Rolle’s Theorem can be used for the location of roots of a

function.
For, if a function g can be identified as the derivative of a function f,

then between any two roots of f there is at least one root of g.
For example: let g(x) = cos x then g is known to be
the derivative of f(x) = sin x. Hence, between any two roots of sin x
there is at least one
root of cos x .
On the other hand, g′ (x) = −sinx = −f(x) .
Another application of Rolle’s Theorem informed us that between any

two roots of cos there is at least one root of sin.Therefore, we conclude
that the roots of sin and cos interlace each other
Mean Value Theorem:

Suppose that f is continuous on a closed interval I = [a, b]and f has a derivative
in the open interval (a, b). Then there exists atleast one point c in (a, b) such
that f(b) − f(a) = f ′ (c)(b − a)
Proof. Assume the function Φ defined on I such that

f(b)−f(a)
Φ(x) = f(x) − f(a) − (x − a)
b−a
We can easily see that The Conditions of Rolle’s Theorem are satisfied by Φ
since Φ is continuous on [a, b], differentiable on (a, b), and Φ(a) = Φ(b).
Therefore, there exists a point b in (a, b) such that

f(b)−f(a)
0 = Φ′ (c) = f ′ (c) − .
b−a
f(b)−f(a)
Therefore f ′ (c) = ⇒ f(b) − f(a) = f ′ (c)(b − a)
b−a
Geometrical Interpretation
The geometric view of the Mean Value Theorem is that there is some
point on the curve y = f(x) at which the tangent line is parallel to the
line segment through the points (a, f (a)) and (b, f (b)). Thus it is easy to
remember the statement of the Mean Value
Theorem by drawing appropriate diagrams. While this should not be

discouraged, it tends to suggest that its importance is geometrical in
nature, which is quite misleading. In fact the
Mean Value Theorem is a wolf in sheep’s clothing and is the

Fundamental Theorem of Differential Calculus.
Cauchy Mean Value Theorem:

Theorem: Let f and g be continuous on [a, b] and differentiable on
(a, b), and assume that g(x) ≠ 0 for all x in (a, b). Then there exists c in
(a, b) such that
f(b)−f(a) f′ (c)
=
g(b)−g(a) g′ (c)
Proof. Since g′ (x) ≠ 0 for all x in (a, b), therefore
Using Rolle’s Theorem, we get
g(a) ≠ g(b).
For x in [a, b], now new define

f(b)−f(a)
φ ( x) = (g(x) − g(a)) − (f(x) − f(a))
g(b)−g(a)
Then h is continuous on [a, b], differentiable on (a, b), and
φ(a) = φ(b) = 0 .
Therefore, According, to Rolle’s Theorem
there exists a point c in (a, b) such that

f(b)−f(a)
0 = φ′(c) = g′ (c) − f ′ (c)
g(b)−g(a)
As we know g′ (c) ≠ 0 , we obtain required result that is

f′ (c) f(b)−f(a)
=
g′ (c) g(b)−g(a)
Strictly Increasing: A function f is said to be strictly increasing on

an interval I if for any points x1 and x 2 in I such that x1 < x 2 , we have
(x1 ) < f( x 2 ) .
Strictly decreasing: A function f is said to be strictly increasing on

an interval I if for any points x1 and x 2 in I such that x1 < x 2 , we have
f(x1 ) < f( x 2 ).
Derivatives can be used to determine whether a function is increasing,

decreasing or constant on an interval:
f(x) is increasing if derivative f ′ (x) > 0,
f(x) is decreasing if derivative f ′ (x) < 0,
f(x) is constant if derivative f ′ (x) = 0.
A critical number, c, is one where f ′ (c) = 0 or f ′ (c) does not exist; a

critical point is (c, f(c)).
After locating the critical number(s), choose test values in each interval
between these critical numbers, then calculate the derivatives at the test
values to decide whether the function is increasing or decreasing in each
given interval.
(In general, identify values of the function which are discontinuous, so,
in addition to critical numbers, also watch for values of the function
which are not defined, at vertical asymptotes or singularities (“holes”).)
8.5 TAYLOR’S THEOREM

Example 1. Find the first 4 terms of the Taylor series for the following
functions:
(a) 𝑓(𝑥 ) = 𝑙𝑜𝑔𝑥 centered at 𝑎 = 1.

1
(b) 𝑓 (𝑥 ) = centered at 𝑎 = 1.
𝑥
𝜋
(c) 𝑓(𝑥 ) = 𝑠𝑖𝑛𝑥 centered at 𝑎 = .
4
Sol. (i)
(ii)
(iii)
Example 2.
Sol.
Example 3.
Sol.
CHECK YOUR PROGRESS
Problem 1. The derivatives of 𝑓(𝑥 ) = sin 𝑥 is 1 at 𝑥 = 0.

Problem 2. The derivatives 𝑓(𝑥 ) = 𝑒 𝑥 is 1 at 𝑥 = 0.
Problem 3. Every Differentiable function is continuous function.
Problem 4. Rolle’s Theorem can be used for the location of roots of a

function.
Problem 5. Every continuous function is Differentiable.
8.6 SUMMARY
1. Theorem: Let b be an interior point of the interval I at which
f ∶ I → R has a relative extremum. If the derivative of f at b exists, then
f ′ (b) = 0.
2. Rolle’s Theorem
Theorem: Consider that f is continuous on a closed interval I = [a, b]

and the derivative f ′ (0) exists at every point of the open interval (a, b),
and f(a) = f(b) = 0. Then there exists at least one point c in (a, b) such
that f ′ (c) = 0
8.7 GLOSSARY
Numbers
Intervals
Continuity function
Functions
Limits
8.8 REFERENCES
1. T. M. Apostol, Mathematical Analysis (2 nd Edition), Narosa

3. W. Rudin, Principles of Mathematical Analysis (3 rd Edition),
4. S.C. Malik and Savita Arora, Mathematical Analysis (6 th Edition),


6. K. A. Ross, Elementary Analysis, The Theory of Calculus (2 nd

𝑥2
Q 1. Using Taylor’s theorem, show that 𝑐𝑜𝑠𝑥 ≥ 1 − ∀ 𝑥 ∈ ℝ.
2
𝑥3
Q 2. Using Taylor’s theorem, show that 𝑥 − < 𝑠𝑖𝑛𝑥 < 𝑥, 𝑥 > 0.
3!
1
Q 3. Prove that 𝑓 (𝑥 ) = is not differentiable at 0.
𝑥
Q 4. Prove that 𝑓 (𝑥 ) = 𝑥 2 is differentiable.
8.11 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. True
CYQ 3. True
CYQ 4. True
CYQ 5. False

BLOCK-III
RIEMANN INTEGRAL AND IMPROPER
INTEGRAL
Department Of Mathematics
UNIT 9: RIEMANN INTEGRAL I
Contents
9.1 Introduction
9.2 Objectives
9.3 Riemann Integral
9.4 Inequalities for integrals
9.5 Refinement of partitions and tagged partitions
9.6 Summary
9.7 Glossary
9.9 References
9.11 Answers
9.1 INTRODUCTION
In the branch of mathematics known as real analysis,

the Riemann integral, created by Bernhard Riemann, was the first rigorous
definition of the integral of a function on an interval. It was presented to
the faculty at the University of Göttingen in 1854, but not published in a
journal until 1868.[1] For many functions and practical applications, the
Riemann integral can be evaluated by the fundamental theorem of
calculus or approximated by numerical integration, or simulated
using Monte Carlo integration.
9.2 OBJECTIVES
 Upper Riemann Sums

 Lower Riemann Sums
 Riemann Integral
 Construct mean value theorem of calculus
9.3 RIEMANN INTEGRAL
Now we will discuss the definition of Riemann integral of a function 𝑓 on

an interval [𝑎, 𝑏].
We first define some basic terms that will be frequently used.
Partition of 𝑰: If 𝐼 = [𝑎, 𝑏] is a closed bounded interval in ℝ, then a

partition of 𝐼 is a finite, ordered set 𝑃 = (𝑥0 , 𝑥1 , … , 𝑥𝑛−1 , 𝑥𝑛 ) of points in
𝐼 such that
𝑎 = 𝑥0 < 𝑥1 < ⋯ < 𝑥𝑛−1 < 𝑥𝑛 = 𝑏.
The points of 𝑃 are used to divide 𝐼 = [𝑎, 𝑏] into non-overlapping

subintervals
𝐼1 = [𝑥0 , 𝑥1 ], 𝐼2 = [𝑥1 , 𝑥2 ], … , 𝐼𝑛 = [𝑥𝑛−1 , 𝑥𝑛 ]
𝑥0 = 𝑎 𝑥1 𝑥2 𝑥𝑛 𝑥𝑛 = 𝑏
Fig. Partition of 𝑰 = [𝒂, 𝒃]

Let 𝑓 be a bounded real function on [𝑎, 𝑏]. Obviously f is bounded on each
sub-interval corresponding to each partition P. Let 𝑀𝑖 and 𝑚𝑖 be the
supremum and infimum respectively of 𝑓 in ∆𝑥𝑖 . Then
Upper Darboux Sums:

𝑼(𝑷, 𝒇) = 𝑴𝟏 ∆𝒙𝟏 + 𝑴𝟐 ∆𝒙𝟐 + ⋯ + 𝑴𝒏 ∆𝒙𝒏 = ∑𝒏𝒊=𝟏 𝑴𝒊 ∆𝑥𝑖
is called Upper Darboux Sums of 𝑓 corresponding to the partition 𝑃.
Lower Darboux Sums:

𝑳(𝑷, 𝒇) = 𝒎𝟏 ∆𝒙𝟏 + 𝒎𝟐 ∆𝒙𝟐 + ⋯ + 𝒎𝒏 ∆𝒙𝒏 = ∑𝒏𝒊=𝟏 𝒎𝒊 ∆𝑥𝑖
is called Lower Darboux Sums of 𝑓 corresponding to the partition 𝑃.
Note: Let M and m are the bounds of f in [a, b]. Then

m ≤ mi ≤ Mi ≤ M ⇒ m∆xi ≤ mi ∆xi ≤ Mi ∆xi ≤ M∆xi
⇒∑ni=1 m∆xi ≤ ∑ni=1 mi ∆xi ≤ ∑ni=1 Mi ∆xi ≤ ∑ni=1 M∆xi
⇒ m ∑ni=1 ∆xi ≤ ∑ni=1 mi ∆xi ≤ ∑ni=1 Mi ∆xi ≤ M ∑ni=1 ∆xi
⇒ 𝐦(𝐚 − 𝐛) ≤ 𝐋(𝐏, 𝐟) ≤ 𝐔(𝐏, 𝐟) ≤ 𝐌(𝐚 − 𝐛)
Note:
Therefore U(P, f) is increasing L(P, f) is decreasing function.
Example 1. Compute L(P, f) and U(P, f) if 𝑓(𝑥 ) = 𝑥 For 𝑥 ∈ [0, 3] and

let 𝑃 = {0, 1, 2, 3} be the partition of [0, 3].
Solution: Partition P divides the interval [0, 3] into sub-intervals
𝐼1 =[0, 1], 𝐼2 =[1, 2], 𝐼3 =[2, 3]
The length of these intervals are given by
𝛿1 = 1 − 0 = 1, 𝛿2 = 2 − 1 = 1, 𝛿3 = 3 − 2 = 1.
Also, if 𝑀𝑟 and 𝑚𝑟 be respectively the l.u.b. and g.l.b. of the function 𝑓 in
[𝑥𝑟−1 , 𝑥𝑟 ], then here we get
𝑀1 = 1, 𝑚1 = 0, 𝑀2 = 2, 𝑚2 = 1 and 𝑀3 = 3, 𝑚3 = 2
Therefore, U(P, f) = ∑3𝑟=1 𝑀𝑟 𝛿𝑟 = 𝑀1 𝛿1 + 𝑀2 𝛿2 + 𝑀3 𝛿3
= 1.1 + 2.1 + 3.1 = 6.
And L(P, f) = ∑3𝑟=1 𝑚𝑟 𝛿𝑟 = 𝑚1 𝛿1 + 𝑚2 𝛿2 + 𝑚3 𝛿3
= 0.1 + 1.1 + 2.1 = 3.
Upper Integral: The infimum of the set of upper sums is called Upper
Integral.
−𝑏
i.e. ∫𝑎 𝑓 𝑑𝑥 = inf 𝑈 = inf {𝑈 (𝑃, 𝑓): 𝑃 is a partition of [𝑎, 𝑏]}
Lower Integral: The supremum of the set of lower sums is called Lower
Integral.
𝑏
i.e. ∫−𝑎 𝑓 𝑑𝑥 = sup 𝐿 = sup {𝐿(𝑃, 𝑓): 𝑃 is a partition of [𝑎, 𝑏]}
Darboux’s condition of integrability:
When Upper integral and lower integral are equal then 𝑓 is said to be
Riemann Integral over [ 𝑎, 𝑏].
𝑏 −𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓 𝑑𝑥 = ∫−𝑎 𝑓 𝑑𝑥
Another definition of Riemann Integrable: A function 𝑓 ∶ [ 𝑎, 𝑏] → ℝ

is said to be Riemann integrable on [𝑎, 𝑏] if there exists a number 𝐿 ∈ ℝ
such that for every 𝜀 > 0 there exists 𝛿 > 0 such that if 𝑃′ is any tagged
partition of [𝑎, 𝑏] with ||𝑃′ || > 0, then |𝑆 (𝑓, 𝑃′ ) − 𝐿| < 𝜀
The set of all Riemann integrable functions on [𝑎, 𝑏] will be denoted by

𝑅[𝑎, 𝑏].
Example 2. Show that a constant function 𝛼 is integrable and

𝑏
∫𝑎 𝑑𝑥 = 𝛼(𝑏 − 𝑎).
Proof. Let P be any partiion of the interval [a,b], then

𝐿(𝑃, 𝑓) = 𝛼∆𝑥1 + 𝛼∆𝑥2 + ⋯ + 𝛼∆𝑥𝑛
= 𝛼 (∆𝑥1 + ∆𝑥2 + ⋯ + ∆𝑥𝑛 ) = 𝛼(𝑏 − 𝑎)
Similarly, 𝑈 (𝑃, 𝑓) = 𝛼∆𝑥1 + 𝛼∆𝑥2 + ⋯ + 𝛼∆𝑥𝑛 = 𝛼(𝑏 − 𝑎)
Therefore
𝑏
∫−𝑎 𝛼𝑑𝑥 = sup 𝐿(𝑃, 𝑓) = 𝛼(𝑏 − 𝑎) and
𝑏
∫−𝑎 𝛼𝑑𝑥 = inf 𝑈(𝑃, 𝑓) = 𝛼(𝑏 − 𝑎)
𝑏 𝑏
⇒∫−𝑎 𝛼𝑑𝑥 = ∫−𝑎 𝛼𝑑𝑥 = 𝛼(𝑏 − 𝑎)
𝑏
Therefore, the constant function is R-integrable and ∫𝑎 𝛼𝑑𝑥 = 𝛼 (𝑏 − 𝑎).
Example 3. Prove that function 𝑓 defines as
0, when 𝑥 is rational
𝑓 (𝑥 ) = { is not integrable on any interval.
1, when 𝑥 𝑖s irrational
Proof. Let P be any partiion of the interval [a,b], then
𝐿(𝑃, 𝑓) = ∑𝑛𝑖=1 𝑚𝑖 ∆𝑥𝑖 = 0∆𝑥1 + 0∆𝑥2 + ⋯ + 0∆𝑥𝑛 = 0
Similarly, 𝑈 (𝑃, 𝑓) = ∑𝑛𝑖=1 𝑀𝑖 ∆𝑥𝑖 = 1∆𝑥1 + 1∆𝑥2 + ⋯ + 1∆𝑥𝑛 = 𝑏 − 𝑎
Therefore
𝑏
∫−𝑎 𝛼𝑑𝑥 = sup 𝐿(𝑃, 𝑓) = 0 and
𝑏
∫−𝑎 𝛼𝑑𝑥 = inf 𝑈(𝑃, 𝑓) = 𝑏 − 𝑎
𝑏 𝑏
⇒∫−𝑎 𝛼𝑑𝑥 ≠ ∫−𝑎 𝛼𝑑𝑥
Therefore, the given function is not R-integrable on any interval.
Example 4. Show that function 𝑓(𝑥 ) = 𝑥 3 is integrable on any interval

[0, 𝑏].
Proof. Let 𝑃 be any partiion of the interval [0, 𝑏] obtained by dividing

0 𝑏 2𝑏 3𝑏 𝑛𝑏
interval into 𝑛 −equal parts. i.e. 𝑃 = [𝑛 = 0, 𝑛 , , ,…, = 𝑏]
𝑛 𝑛 𝑛
(𝑖−1)𝑘 3
Let lower bounds of function in ∆𝑥𝑖 = ( ) and Upper bounds of
𝑛
𝑖𝑘 3
function in ∆𝑥𝑖 = ( 𝑛 )
Therefore
𝐿(𝑃, 𝑓) = ∑𝑛𝑖=1 𝑚𝑖 ∆𝑥𝑖 = 𝑚1 ∆𝑥1 + 𝑚2 ∆𝑥2 + ⋯ + 𝑚𝑛 ∆𝑥𝑛
𝑏 𝑏 3 𝑏 2𝑏 3 𝑏 𝑏(𝑛−1) 3 𝑏 𝑏4
= 0. 𝑛 + (𝑛) . 𝑛 + ( 𝑛 ) . 𝑛 + ⋯ + ( ) . 𝑛 = 𝑛 4 [ 13 + 23 +
𝑛
⋯ + (𝑛 − 1)3 ]
𝑏4 (𝑛−1)2 𝑛 2 𝑏4 1 2
= = (1 − 𝑛)
4𝑛 4 4
Similarly
𝑈(𝑃, 𝑓) = ∑𝑛𝑖=1 𝑀𝑖 ∆𝑥𝑖 = 𝑀1 ∆𝑥1 + 𝑀2 ∆𝑥2 + ⋯ + 𝑀𝑛 ∆𝑥𝑛
𝑏 3 𝑏 2𝑏 3 𝑏 2𝑏 3 𝑏 𝑏𝑛 3 𝑏 𝑏4
= ( 𝑛 ) . 𝑛 + ( 𝑛 ) . 𝑛 + ( 𝑛 ) . 𝑛 + ⋯ + ( 𝑛 ) . 𝑛 = 𝑛 4 [ 13 +
23 + ⋯ + 𝑛 3 ]
𝑏4 𝑛 2(𝑛+1)2 𝑏4 1 2
= = (1 + 𝑛)
4𝑛 4 4
Therefore
𝑏 𝑏4
∫−0 𝛼𝑑𝑥 = sup 𝐿(𝑃, 𝑓) = 4
and
𝑏 𝑏4
∫−0 𝛼𝑑𝑥 = inf 𝑈(𝑃, 𝑓) = 4
𝑏 𝑏 𝑏4
⇒∫−0 𝛼𝑑𝑥 = ∫−0 𝛼𝑑𝑥 = 4
𝑏 𝑏4
Therefore, the given function is R-integrable and ∫0 𝛼𝑑𝑥 = .
4
𝜋
Example 5. Show that the function 𝑓(𝑥 ) = 𝑠𝑖𝑛𝑥 is integrable in [0, 2 ] and
𝜋
∫0 𝑠𝑖𝑛𝑥 𝑑𝑥 = 1.
2
𝜋
Solution: Let any partition of [0, 2 ] be
0𝜋 𝜋 2𝜋 𝑟𝜋 𝑛𝜋 𝜋
P = {0 = 2𝑛 , 2𝑛 , 2𝑛 , … , 2𝑛 , … , 2𝑛 = 2 }
𝜋
Which dissects [0, 2 ] into 𝑛 equal parts.
𝜋
The length of each subinterval = 2𝑛 and the 𝑟 𝑡ℎ sub - interval is
(𝑟−1)𝜋 𝑟𝜋
𝐼𝑟 = [ , 2𝑛].
2𝑛
𝜋
As 𝑓(𝑥 ) = 𝑠𝑖𝑛𝑥 is increasing in [0, 2 ], so we have
𝑠𝑖𝑛(𝑟−1)𝜋 𝑠𝑖𝑛𝑟𝜋
𝑚𝑟 = and 𝑀𝑟 = , 𝑟 = 1,2,3, … , 𝑛
2𝑛 2𝑛
𝑟𝜋 𝜋
Therefore, 𝑈 (𝑃, 𝑓) = ∑𝑛𝑖=1 𝑀𝑖 ∆𝑥𝑖 = ∑𝑛𝑖=1 (𝑠𝑖𝑛 ).
2𝑛 2𝑛
𝜋 𝜋 2𝜋 𝑛𝜋
= 2𝑛 [ 𝑠𝑖𝑛 2𝑛 + 𝑠𝑖𝑛 2𝑛 + ⋯ + 𝑠𝑖𝑛 2𝑛 ]
𝜋 𝑛−1 𝜋 𝑛𝜋
𝜋 sin( + . ) 𝑠𝑖𝑛
2𝑛 𝑛 2𝑛 4𝑛
= 2𝑛 [ 𝜋 ]
𝑠𝑖𝑛
4𝑛
∵ sin 𝑎 + sin(𝑎 + 𝑑 ) + ⋯ + sin[𝑎 + (𝑛 − 1)𝑑]

𝑛−1 𝑛𝑑
sin(𝑎+ .𝑑 ) 𝑠𝑖𝑛
2 2
= sin(𝑑/2)
𝜋 (𝑛+1)𝜋 𝜋 𝜋
Or 𝑈(𝑃, 𝑓) = 2𝑛 [{sin sin ( 4 )} / sin (4𝑛)]
4𝑛
𝜋 𝜋 𝜋 1 𝜋
= 2𝑛 [{sin( 4 + 4𝑛) √2} / sin (4𝑛)]
𝜋 𝜋 𝜋 𝜋 𝜋 𝜋
= 2√2𝑛 [{sin 4 𝑐𝑜𝑠 4𝑛 + cos 4 𝑠𝑖𝑛 4𝑛} / sin (4𝑛)]
𝜋 𝜋
= 4𝑛 (cot 4𝑛 + 1)
𝜋 𝜋
Similarly, 𝐿(𝑃, 𝑓) = 4𝑛 (cot 4𝑛 − 1)
𝜋 𝜋
Now, Riemann lower integral = lim 𝐿(𝑃, 𝑓) = lim (cot 4𝑛 − 1)
𝑛⟶∞ 𝑛⟶∞ 4𝑛
𝜋/4𝑛 𝜋
= lim 𝜋 − lim
𝑛⟶∞ tan (4𝑛) 𝑛⟶∞ 4𝑛
=1–0=1 …… (1)
𝜋 𝜋
And Riemann upper integral = lim 𝑈(𝑃, 𝑓) = lim (cot 4𝑛 + 1)
𝑛⟶∞ 𝑛⟶∞ 4𝑛
=1 ……. (2)
From equation (1) and (2) we get

𝜋
𝜋
𝑓 (𝑥 ) = 𝑠𝑖𝑛𝑥 is integrable in [0, 2 ] and ∫02 𝑠𝑖𝑛𝑥 𝑑𝑥 = 1.
9.4 INEQUALITIES FOR INTEGRALS

We already prove that
𝑏
𝑚(𝑏 − 𝑎) ≤ ∫𝑎 𝑓 𝑑𝑥 ≤ 𝑀(𝑏 − 𝑎) 𝑤ℎ𝑒𝑛 𝑏 ≥ 𝑎 …………………..(I)
If 𝑏 < 𝑎, so that 𝑎 > 𝑏 and

𝑎 𝑎
𝑚(𝑎 − 𝑏) ≤ ∫𝑏 𝑓 𝑑𝑥 ≤ 𝑀(𝑎 − 𝑏) ⇒−𝑚(𝑎 − 𝑏) ≤ − ∫𝑏 𝑓 𝑑𝑥 ≤
−𝑀(𝑎 − 𝑏)
𝑏
⇒ 𝑚(𝑏 − 𝑎) ≤ ∫𝑎 𝑓 𝑑𝑥 ≤ 𝑀(𝑏 − 𝑎) when 𝑏 < 𝑎………………(II)
Deduction 1: If 𝑓 is bounded and integrable on [𝑎, 𝑏], then there exists a

𝑎
number 𝑘 lying between bounds of 𝑓 such that ∫𝑏 𝑓 𝑑𝑥 = 𝑘(𝑏 − 𝑎)
Deduction 2: If 𝑓 is continuous and integrable on [𝑎, 𝑏], then there exists

𝑎
a number 𝑐 lying between 𝑎 and 𝑏 such that ∫𝑏 𝑓 𝑑𝑥 = 𝑓(𝑐)(𝑏 − 𝑎)
Deduction 3: If 𝑓 is bounded and integrable on [𝑎, 𝑏], and 𝛼 > 0 is a

number such that |𝑓 (𝑥 )| ≤ 𝛼 for all 𝑥 ∈ [𝑎, 𝑏], then
𝑎
|∫𝑏 𝑓 𝑑𝑥 | ≤ 𝛼|𝑏 − 𝑎|.
Proof. Let 𝑀 and 𝑚 be the upper bounds and lower bounds of 𝑓(𝑥)
respectively.
Let 𝛼 > 0 is a number such that |𝑓 (𝑥)| ≤ 𝛼 for all 𝑥 ∈ [𝑎, 𝑏]
Hence for 𝑏 > 𝑎, −𝛼 ≤ 𝑓(𝑥 ) ≤ 𝛼
⇒ −𝛼 ≤ 𝑚 ≤ 𝑓(𝑥 ) ≤ 𝑀 ≤ 𝛼
𝑏
⇒ −𝛼(𝑏 − 𝑎) ≤ 𝑚(𝑏 − 𝑎) ≤ ∫𝑎 𝑓 (𝑥 ) ≤ 𝑀(𝑏 − 𝑎) ≤ 𝛼(𝑏 − 𝑎)
𝑏
⇒ |∫𝑎 𝑓(𝑥 )| ≤ 𝛼(𝑏 − 𝑎)
If 𝑎 > 𝑏, we have
𝑏
|∫𝑎 𝑓 (𝑥 )| ≤ 𝛼(𝑎 − 𝑏)
𝑏
Therefore |∫𝑎 𝑓 (𝑥 )| ≤ 𝛼|𝑏 − 𝑎|.
The result is trivial for 𝑎 = 𝑏.
Deduction 4: If 𝑓 is bounded and integrable on [𝑎, 𝑏] and 𝑓(𝑥) ≥ 0 for

all 𝑥 ∈ [𝑎, 𝑏], then
𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 ≥ 0 when 𝑏 ≥ 𝑎 and ∫𝑎 𝑓 𝑑𝑥 ≤ 0 when 𝑏 ≤ 𝑎
Proof. Because 𝑓(𝑥) ≥ 0 for all 𝑥 ∈ [𝑎, 𝑏], then the lower bound of
𝑓 (𝑥 ) 𝑖. 𝑒. 𝑚 ≥ 0
From Inequality (I) and (II) , we get

𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 ≥ 0 when 𝑏 ≥ 𝑎 and ∫𝑎 𝑓 𝑑𝑥 ≤ 0 when 𝑏 ≤ 𝑎
Deduction 5 : If 𝑓 and 𝑔 are bounded and integrable on [𝑎, 𝑏], such that
𝑓 (𝑥 ) ≥ 𝑔(𝑥)., for all 𝑥 ∈ [𝑎, 𝑏].then
𝑏 𝑏 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 ≥ ∫𝑎 𝑔 𝑑𝑥 when 𝑏 ≥ 𝑎 and ∫𝑎 𝑓 𝑑𝑥 ≤ ∫𝑎 𝑓 𝑑𝑥 when 𝑏 ≤ 𝑎
Proof. It is given that 𝑓 ≥ 𝑔 then 𝑓 − 𝑔 ≥ 0 for all 𝑥 ∈ [𝑎, 𝑏].
Using deduction 4, we have

𝑏
∫𝑎 (𝑓 − 𝑔)𝑑𝑥 ≥ 0 𝑖𝑓 𝑏 ≥ 𝑎
𝑏 𝑏
⇒∫𝑎 𝑓𝑑𝑥 ≥ ∫𝑎 𝑔𝑑𝑥 𝑖𝑓 𝑏 ≥ 𝑎
Similarly
𝑏 𝑏
∫𝑎 𝑓𝑑𝑥 ≤ ∫𝑎 𝑔𝑑𝑥 𝑖𝑓 𝑏 ≤ 𝑎
9.5 REFINEMENT OF PARTITIONS AND

TAGGED PARTITIONS
Norm: The norm (or mesh) of 𝑃 to be the number
𝜇(𝑃) = max {𝑥1 − 𝑥0 , 𝑥2 − 𝑥1 , … , 𝑥𝑛 − 𝑥𝑛−1 }
OR
the norm of a partition is merely the length of the largest subinterval into
which the partition divides [𝑎, 𝑏].
Refinement: A partition 𝑃∗ is said to be a refinement of 𝑃 if 𝑃∗ ⊇ 𝑃 i.e.

every point of P is a point of 𝑃∗ .
Or we can say that 𝑃∗ refines 𝑃 or 𝑃∗ is finer than 𝑃.
If 𝑃1 and 𝑃2 are two partitions, then 𝑃∗ = 𝑃1 ∪ 𝑃2 .
Theorem 9.5.1. Suppose that 𝑓 ∶ [𝑎, 𝑏] → 𝑅 is bounded and 𝑃 and 𝑃 ∗

be partitions of [a, b] and refinement of 𝑃 respectively. Then
(i) 𝐿(𝑃, 𝑓) ≤ 𝐿(𝑃∗ , 𝑓)
(ii) 𝑈(𝑃∗ , 𝑓) ≤ 𝑈(𝑃, 𝑓)
Proof. Let 𝑃 be partition of [𝑎, 𝑏] and 𝑷∗ contains just one more point ′𝛼′
than 𝑃.
Let 𝛼 ∈ ∆𝑥𝑖 𝑖. 𝑒 𝑥𝑖−1 < 𝛼 < 𝑥𝑖 .
It is given that the function 𝑓 is bounded over the interval [𝑎, 𝑏].
⇒ It is bounded in every subinterval ∆𝑥𝑖 .
Let 𝛽1 , 𝛽2 and 𝑚𝑖 be the infimum of f in the interval [𝑥𝑖−1 , 𝛼 ], [𝛼, 𝑥𝑖 ] and

[𝑥𝑖−1 , 𝑥𝑖 ] respectively.
Obviously 𝑚𝑖 ≤ 𝛽1 and 𝑚𝑖 ≤ 𝛽2 .
Hence
𝐿(𝑃∗ , 𝑓) − 𝐿(𝑃, 𝑓) = 𝑚1 ∆𝑥1 + 𝑚2 ∆𝑥2 + ⋯ + 𝛽1 (𝛼 − 𝑥𝑖−1 ) + 𝛽1 (𝑥𝑖 −

𝛼 ) + 𝑚𝑖+1 ∆𝑥𝑖+1 + ⋯ + 𝑚𝑛 ∆𝑥𝑛 − (𝑚1 ∆𝑥1 + 𝑚2 ∆𝑥2 + ⋯ + 𝑚𝑖 ∆𝑥𝑛 +
𝑚𝑛 ∆𝑥𝑛 )
= 𝛽1 (𝛼 − 𝑥𝑖−1 ) + 𝛽2 (𝑥𝑖 − 𝛼 ) − 𝑚𝑖 (𝑥𝑖 − 𝑥𝑖−1 )
= 𝛽1 𝛼 − 𝛽1 𝑥𝑖−1 + 𝛽2 𝑥𝑖 − 𝛽2 𝛼 − 𝑚𝑖 𝑥𝑖 + 𝑚𝑖 𝑥𝑖−1
= 𝛽1 𝛼 − 𝛽1 𝑥𝑖−1 − 𝑚𝑖 𝛼 + 𝑚𝑖 𝛼 + 𝛽2 𝑥𝑖 − 𝛽2 𝛼 −
𝑚𝑖 𝑥𝑖 + 𝑚𝑖 𝑥𝑖−1
= 𝛼(𝛽1 − 𝑚𝑖 ) − 𝑥𝑖−1 (𝛽1 − 𝑚𝑖 ) − 𝑚𝑖 (𝑥𝑖 − 𝛼 ) +

𝛽2 (𝑥𝑖 − 𝛼 )
= (𝛼 − 𝑥𝑖−1 )(𝛽1 − 𝑚𝑖 ) + (𝛽2 − 𝑚𝑖 ) (𝑥𝑖 − 𝛼 )
𝑥𝑖 > 𝛼 > 𝑥𝑖−1 and 𝛽1 , 𝛽2 ≥ 𝑚𝑖 ⇒(𝛼 − 𝑥𝑖−1 ), (𝑥𝑖 − 𝛼 ), (𝛽1 − 𝑚𝑖 ) and

(𝛽2 − 𝑚𝑖 ) are positive.
Therefore, 𝐿(𝑃∗ , 𝑓) − 𝐿(𝑃, 𝑓) ≥ 0
If 𝑃∗ contains 𝑝 points more than 𝑃, we repeat the above reasoning 𝑝 times

and conclude that
𝐿(𝑃∗ , 𝑓) ≥ 𝐿(𝑃, 𝑓)
Similarly, we can prove that 𝑈(𝑃∗ , 𝑓) ≤ 𝑈(𝑃, 𝑓)
Corollary If a refinement 𝑷∗ of 𝑷 contains k points more than 𝑷 and

|𝒇(𝒙)| ≤ 𝑲, for all 𝒙 ∈ [𝒂, 𝒃], then
(i) 𝐿(𝑃, 𝑓) ≤ 𝐿(𝑃∗ , 𝑓) ≤ 𝐿(𝑃, 𝑓) + 2𝑘𝐾𝜇
(ii) 𝑈(𝑃, 𝑓) ≥ 𝑈(𝑃∗ , 𝑓) ≥ 𝑈(𝑃, 𝑓) − 2𝑘𝐾𝜇
Proof. . Let 𝑃 be partition of [𝑎, 𝑏] and 𝑷∗ contains just one more point
′𝛼′ than 𝑃.
Let 𝛼 ∈ ∆𝑥𝑖 𝑖. 𝑒 𝑥𝑖−1 < 𝛼 < 𝑥𝑖 .
It is given that the function 𝑓 is bounded over the interval [𝑎, 𝑏].
⇒ It is bounded in every subinterval ∆𝑥𝑖 .
Let 𝛽1 , 𝛽2 and 𝑚𝑖 be the infimum of 𝑓 in the interval [𝑥𝑖−1 , 𝛼 ], [𝛼, 𝑥𝑖 ] and

[𝑥𝑖−1 , 𝑥𝑖 ] respectively.
Obviously 𝑚𝑖 ≤ 𝛽1 and 𝑚𝑖 ≤ 𝛽2 .
Hence
𝐿(𝑃∗ , 𝑓) − 𝐿(𝑃, 𝑓) = 𝑚1 ∆𝑥1 + 𝑚2 ∆𝑥2 + ⋯ + 𝛽1 (𝛼 − 𝑥𝑖−1 ) +
𝛽1 (𝑥𝑖 − 𝛼 ) + 𝑚𝑖+1 ∆𝑥𝑖+1 + ⋯ + 𝑚𝑛 ∆𝑥𝑛 − (𝑚1 ∆𝑥1 + 𝑚2 ∆𝑥2 + ⋯

+ 𝑚𝑖 ∆𝑥𝑛 + 𝑚𝑛 ∆𝑥𝑛 )
= 𝛽1 (𝛼 − 𝑥𝑖−1 ) + 𝛽2 (𝑥𝑖 − 𝛼 ) − 𝑚𝑖 (𝑥𝑖 − 𝑥𝑖−1 )
= 𝛽1 𝛼 − 𝛽1 𝑥𝑖−1 + 𝛽2 𝑥𝑖 − 𝛽2 𝛼 − 𝑚𝑖 𝑥𝑖 + 𝑚𝑖 𝑥𝑖−1
= 𝛽1 𝛼 − 𝛽1 𝑥𝑖−1 − 𝑚𝑖 𝛼 + 𝑚𝑖 𝛼 + 𝛽2 𝑥𝑖 − 𝛽2 𝛼 − 𝑚𝑖 𝑥𝑖 + 𝑚𝑖 𝑥𝑖−1
= 𝛼(𝛽1 − 𝑚𝑖 ) − 𝑥𝑖−1 (𝛽1 − 𝑚𝑖 ) − 𝑚𝑖 (𝑥𝑖 − 𝛼 ) + 𝛽2 (𝑥𝑖 − 𝛼 )
= (𝛼 − 𝑥𝑖−1 )(𝛽1 − 𝑚𝑖 ) + (𝛽2 − 𝑚𝑖 ) (𝑥𝑖 − 𝛼 )
It us given that |𝑓 (𝑥 )| ≤ 𝐾 for all 𝑥 ∈ [𝑎, 𝑏], therefore
−𝐾 ≤ 𝑚𝑖 ≤ 𝛽1 ≤ 𝐾 ⇒ 𝐾 ≥ −𝑚𝑖 and 𝐾 ≥ 𝛽1 ⇒ 2𝐾 ≥ 𝛽1 − 𝑚𝑖 or
2𝐾 ≥ 𝛽1 − 𝑚𝑖 ≥ 0
Similarly
2𝐾 ≥ 𝛽2 − 𝑚𝑖 ≥ 0
Therefore
𝐿(𝑃∗ , 𝑓) − 𝐿(𝑃, 𝑓) ≤ 2𝐾(𝛼 − 𝑥𝑖−1 ) + 2𝐾 (𝑥𝑖 − 𝛼 ) = 2𝐾(𝛼 − 𝑥𝑖−1 +

𝑥𝑖 − 𝛼 ) = 2𝐾(𝑥𝑖 − 𝑥𝑖−1 )
Therefore
𝐿(𝑃∗ , 𝑓) − 𝐿(𝑃, 𝑓) ≤ 2𝐾∆𝑥𝑖
Let 𝜇 be the norm of 𝑃, hence
𝐿(𝑃∗ , 𝑓) − 𝐿(𝑃, 𝑓) ≤ 2𝐾𝜇
Let each additional point is introduced one by one, by repeating the

above reasoning 𝒌 times, we get
𝐿(𝑃∗ , 𝑓) − 𝐿(𝑃, 𝑓) ≤ 2𝐾𝑘𝜇 ⇒𝐿(𝑃∗ , 𝑓) ≤ 𝐿(𝑃, 𝑓) + 2𝐾𝑘𝜇
Also, 𝐿(𝑃, 𝑓) ≤ 𝐿(𝑃∗ , 𝑓)
Hence 𝐿(𝑃, 𝑓) ≤ 𝐿(𝑃, 𝑓) + 2𝐾𝑘𝜇
Similarly, we can prove that 𝑈(𝑃, 𝑓) ≥ 𝑈(𝑃∗ , 𝑓) ≥ 𝑈(𝑃, 𝑓) − 2𝑘𝐾𝜇
CHECK YOUR PROGRESS
True or false/MCQ Questions
𝜋
Problem 1. The function 𝑓(𝑥 ) = 𝑠𝑖𝑛𝑥 is integrable in [0, 2 ] and
𝜋
∫0 𝑠𝑖𝑛𝑥 𝑑𝑥 = 1.
2
Problem 2. If the function 𝑓: ℝ ⟶ ℝ is defined as 𝑓(𝑥 ) = [𝑥] where,

[. ] represent the greatest integer function then
(a) 𝑓(𝑥) is continuous function on ℝ.
(b) 𝑓(𝑥) is Differential function on ℝ.
(c) 𝑓(𝑥) is Riemann integrable.
(d) 𝑓(𝑥) is not Riemann integrable.
Problem 3. Every Riemann integrable function is continuous function.
Problem 4. Every polynomial function is Riemann integrable on ℝ.
Problem 5. U(P, f) is increasing L(P, f) is decreasing function.
9.6 SUMMARY
1. If Limit f (a  h)  Limit f (a  h)  f (a) then function f(x) is

h 0 h 0
continuous at 𝑥 = 𝑎.
2. A function f: X → Y is said to be uniformly continuous on A ⊆ X if

for every ε > 0, there exists δ > 0 such that x, y ∈ A, |𝑥 − 𝑦| < δ implies
|𝑓(𝑥) − 𝑓(𝑦)| < ε.
9.7 GLOSSARY
integration
continuity
Functions
Limits
9.8 REFERENCES




Q 1. Prove that every constant function is Reimann integrable.
Q 2. Prove that every polynomial function is Riemann integrable.
𝜋
Q 3. Show that the function 𝑓 (𝑥 ) = 𝑠𝑖𝑛𝑥 is integrable in [0, 2 ]
1 1
Q 4. Using Riemann integration prove ∫0 𝑥 𝑑𝑥 = 2 .
Q 5. Define upper and lower Riemann sums.
9.11 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. (c)
CYQ 3. False
CYQ 4. True
CYQ 5. True
UNIT 10: RIEMANN INTEGRAL II

Contents
10.1 Objectives
10.2 Introduction
10.3 Darboux Theorem
10.4 Condition for integrability and some properties
10.5 Some important theorem
10.6 Riemann sum
10.7 Summary
10.8 Glossary
10.10 References
10.12 Answers
10.1 INTRODUCTION
During a century and a half of development and refinement of techniques,
calculus consisted of these paired operations and their applications,
primarily to physical problems.
In the 1850s, Bernhard Riemann adopted a new and different viewpoint.

He separated the concept of integration from its companion,
differentiation, and examined the motivating summation and limit process
of finding areas by itself. He broadened the scope by considering all
functions on an interval for which this process of ‘‘integration’’ could be
defined: the class of ‘‘integrable’’ functions. The Fundamental Theorem
of Calculus became a result that held only for a restricted set of integrable
functions. The viewpoint of Riemann led others to invent other integration
theories, the most significant being Lebesgue’s theory of integration. But
there have been some advances made in more recent times that extend
even the Lebesgue theory to a considerable extent.
10.2 OBJECTIVES
In this Unit, we will
 Discussed about Riemann Integral

 Construct mean value theorem of calculus
10.3 DARBOUX THEOREM

Darboux Theorem
Theorem 10.3.1. If 𝒇 is bounded function on [𝒂, 𝒃] then to every
𝜺 > 𝟎, there corresponds 𝜹 > 𝟎 such that

−𝒃
(i) 𝑼(𝑷, 𝒇) < ∫𝒂 𝒇 𝒅𝒙 + 𝜺
𝒃
(ii) 𝑳(𝑷, 𝒇) > ∫−𝒂 𝒇 𝒅𝒙 − 𝜺
For every partition 𝑷 of [𝒂, 𝒃] with norm 𝝁(𝑷) < 𝜹
Proof. It is given that 𝑓 is bounded on [𝑎, 𝑏]. Hence there exists 𝛼 > 0
such that
𝑓 (𝑥 ) ≤ 𝛼 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥 ∈ [𝑎, 𝑏]
Now
−𝑏
∫𝑎 𝑓 𝑑𝑥 = inf 𝑈 = inf {𝑈(𝑃, 𝑓): 𝑃 is a partition of [𝑎, 𝑏]}
Hence for every 𝜀 > 0 there exists a partition 𝑃′ = {𝑥0 , 𝑥1 , 𝑥3 , … , 𝑥𝑘 } of

[𝑎, 𝑏] such that
−𝑏 1
𝑈 (𝑃1 , 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 ………………………..(1)
Also partition 𝑃′ contains 𝑘 − 1 points other than 𝑎 and 𝑏.
Let 𝛿 be a positive number such that

1
2(𝑘 − 1)𝛼𝛿 = 2 𝜀 ………………………………….(2).
Let 𝑃 be any partition such that 𝑃 = {𝑥0 , 𝑥1 , 𝑥3 , … , 𝑥𝑛 } with norm
𝜇 (𝑃) < 𝛿.
Assume 𝑃∗ be a refinement of 𝑃 and 𝑃’ such that 𝑃∗ = 𝑃 ∪ 𝑃′
𝑃 ∗ be a refinement of 𝑃⇒ 𝑃∗ have 𝑝 − 1 more point than 𝑃 and also

𝑓 (𝑥 ) ≤ 𝛼
Therefore
𝑈 (𝑃, 𝑓) ≥ 𝑈 (𝑃∗ , 𝑓) ≥ 𝑈 (𝑃, 𝑓) − 2(𝑝 − 1)𝛼𝛿 (Using previous corollary)
⇒ 𝑈(𝑃, 𝑓) − 2(𝑝 − 1)𝛼𝛿 ≤ 𝑈(𝑃∗ , 𝑓)
≤ 𝑈 (𝑃 ′ , 𝑓 )
−𝑏 1
< ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 (Using eq (1))
Therefore
−𝑏 1
𝑈 (𝑃, 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 + 2(𝑝 − 1)𝛼𝛿
Using equation (2), we get

−𝑏 1 1 −𝑏
𝑈 (𝑃, 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 + 2 𝜀 < ∫𝑎 𝑓 𝑑𝑥 + 𝜀
𝑏
Similarly, we can prove that 𝐿(𝑃, 𝑓) > ∫−𝑎 𝑓 𝑑𝑥 − 𝜀
Note:
 Tags: If a point 𝑡𝑖 has been selected from each subinterval 𝐼𝑖 = [𝑥𝑖−1 , 𝑥𝑖 ]

for 𝑖 = 1,2, . . , 𝑛, then the points are called tags of the subintervals 𝐼𝑖 .
 Tagged Partition of 𝑰: A set of ordered 𝑃 = {([𝑥𝑖−1 , 𝑥𝑖 ], 𝑡𝑖 ); 𝑖 =
1,2, … , 𝑛} of subintervals and corresponding tags is called a tagged
partition of 𝐼.
10.4 CONDITION OF INTEGRABILITY AND

SOME PROPERTIES OF INTEGRABLE
FUNCTIONS
We already discussed that the bounded function is integrable if upper and
lower integral are equal. Now we try to study the necessary and sufficient
condition for integrability of a function.
FIRST FORM
Theorem 10.4.1. The necessary and sufficient condition for

integrability of a bounded function f is for every 𝜺 > 𝟎 there exists
𝜹 > 𝟎 such that for every partition 𝑷 of [𝒂, 𝒃]with norm 𝝁(𝑷) < 𝜹 and
𝑼(𝑷, 𝒇) − 𝑳(𝑷, 𝒇) < 𝜺
Proof. Necessary condition
Let 𝑓 be a bounded function and integrable over interval [𝑎, 𝑏],

𝑏 −𝑏 𝑏
Hence ∫−𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓 𝑑𝑥
Let 𝜀 > 0 be any positive number.
By Darbaux’s Theorem there exists a positive number 𝛿 such that foe

every partition 𝑃 with norm 𝜇 (𝑃) < 𝛿
−𝑏 1
𝑈 (𝑃, 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 …………………..(1)
𝑏 1
𝐿(𝑃, 𝑓) > ∫−𝑎 𝑓 𝑑𝑥 − 2 𝜀 ……………………(2)
𝑏 1
⇒ −𝐿(𝑃, 𝑓) < − ∫−𝑎 𝑓 𝑑𝑥 + 2 𝜀………………………………….(3)
By adding inequality (1) and (3), we get

−𝑏 1 𝑏 1
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 − ∫−𝑎 𝑓 𝑑𝑥 + 2 𝜀 = 𝜀
Hence for every partition 𝑃 of [𝑎, 𝑏]with norm 𝜇 (𝑃) < 𝛿
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀
Sufficient Condition
Assume for every partition 𝑃 of [𝑎, 𝑏]with norm 𝜇 (𝑃) < 𝛿 and
𝑈(𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀…………………………………………..(4)
for any partition 𝑃 of [𝑎, 𝑏], we have

−𝑏 −𝑏
𝑈 (𝑃, 𝑓) ≥ ∫𝑎 𝑓 𝑑𝑥 ⇒∫𝑎 𝑓 𝑑𝑥 ≤ 𝑈(𝑃, 𝑓)…………………………..(5)
𝑏 𝑏
𝐿(𝑃, 𝑓) ≤ ∫−𝑎 𝑓 𝑑𝑥 ⇒− ∫−𝑎 𝑓 𝑑𝑥 ≤ −𝐿(𝑃, 𝑓)………………………..(6)
Adding inequality (5) and (6), we get

−𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 − ∫−𝑎 𝑓 𝑑𝑥 ≤ 𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓)
Using inequality (4), we get

−𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 − ∫−𝑎 𝑓 𝑑𝑥 < 𝜀
Because 𝜀 is any arbitrary positive number and also we know that a non
negative number is less than every positive number.
Therefore it should be equal to 0.

−𝑏 𝑏
i.e. ∫𝑎 𝑓 𝑑𝑥 − ∫−𝑎 𝑓 𝑑𝑥 < 𝜀 = 0
−𝑏 𝑏
Therefore ∫𝑎 𝑓 𝑑𝑥 = ∫−𝑎 𝑓 𝑑𝑥 which implies that 𝑓 is integrable over
interval [𝑎, 𝑏].
SECOND FORM
Theorem 10.4.2. A bounded function 𝒇 is integrable on [𝒂, 𝒃] iff for

every 𝜺 > 𝟎 there exists a partition 𝑷 of [𝒂, 𝒃] such that 𝑼(𝑷, 𝒇) −
𝑳(𝑷, 𝒇) < 𝜺
Proof. Necessary condition
Let 𝑓 be a bounded function and integrable over interval [𝑎, 𝑏],

𝑏 −𝑏 𝑏
Hence ∫−𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓 𝑑𝑥
As we know that the

𝑏 −𝑏
∫−𝑎 𝑓 𝑑𝑥 =supremum of lower sums and ∫𝑎 𝑓 𝑑𝑥 = infimum of upper
sums
Hence there exists a partition 𝑃′ and 𝑃′′ such that

−𝑏 1
𝑈 (𝑃′ , 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀
𝑏 1
⇒ 𝑈 (𝑃′ , 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 …………..(1)
𝑏 1
𝐿(𝑃′′ , 𝑓) > ∫−𝑎 𝑓 𝑑𝑥 − 2 𝜀
𝑏 1
⇒ 𝐿(𝑃′′ , 𝑓) > ∫𝑎 𝑓 𝑑𝑥 − 2 𝜀
𝑏 1
⇒ ∫𝑎 𝑓 𝑑𝑥 < 𝐿(𝑃′′ , 𝑓) + 𝜀 …………(2)
2
Assume 𝑃 be the commom refinement of partitions 𝑃′ and 𝑃′′ i.e.
𝑃 = 𝑃′ ∪ 𝑃′′
Therefore
𝑏 1
𝑈 (𝑃, 𝑓) ≤ 𝑈 (𝑃′ , 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 2 𝜀 (using inequality (1))
1 1
⇒ 𝑈 (𝑃, 𝑓) < 𝐿(𝑃′′ , 𝑓) + 2 𝜀 + 2 𝜀 = 𝐿(𝑃′′ , 𝑓) + 𝜀
Therefore, 𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀 for a partition 𝑃.
Sufficient Condition
Assume 𝜀 < 0 be any positive number. Consider 𝑃 be a partitions such

that
𝑈(𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀…………………………………………..(3)
Now for any partition 𝑃 of [𝑎, 𝑏], we have

−𝑏 −𝑏
𝑈 (𝑃, 𝑓) ≥ ∫𝑎 𝑓 𝑑𝑥 ⇒∫𝑎 𝑓 𝑑𝑥 ≤ 𝑈(𝑃, 𝑓)…………………………..(4)
𝑏 𝑏
𝐿(𝑃, 𝑓) ≤ ∫−𝑎 𝑓 𝑑𝑥 ⇒− ∫−𝑎 𝑓 𝑑𝑥 ≤ −𝐿(𝑃, 𝑓)………………………..(5)
Adding inequality (4) and (5), we get

−𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 − ∫−𝑎 𝑓 𝑑𝑥 ≤ 𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓)
−𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 − ∫−𝑎 𝑓 𝑑𝑥 < 𝜀
Because 𝜀 is any arbitrary positive number and also we know that a non
negative number is less than every positive number.
Therefore, it should be equal to 0.

−𝑏 𝑏
i.e. ∫𝑎 𝑓 𝑑𝑥 − ∫−𝑎 𝑓 𝑑𝑥 < 𝜀 = 0
−𝑏 𝑏
Therefore ∫𝑎 𝑓 𝑑𝑥 = ∫−𝑎 𝑓 𝑑𝑥 which implies that 𝑓 is integrable over
Integrability of the sum and difference of Integrable functions
Theorem 10.4.3. Let 𝒇𝟏 and 𝒇𝟐 are two bounded and integrable

function on [𝒂, 𝒃] then 𝒇 = 𝒇𝟏 + 𝒇𝟐 is also integrable on [𝒂, 𝒃] and
𝑏 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓1 𝑑𝑥 + ∫𝑎 𝑓2 𝑑𝑥
Proof. Let 𝑓1 and 𝑓2 are two bounded ⇒ 𝑓 = 𝑓1 + 𝑓2 is bounded on [𝑎, 𝑏].
Let 𝑃 be any partition 𝑃 of [𝑎, 𝑏] such that
𝑃 = {𝑎 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 = 𝑏}.
Let 𝑀𝑖′ and 𝑚𝑖′ are the upper and lower bound of 𝑓1 respectively and 𝑀𝑖′′
and 𝑚𝑖′′ are the upper and lower bound of 𝑓2 respectively in ∆𝑥𝑖 .
Assume 𝑀𝑖 and 𝑚𝑖 are the upper and lower bound of 𝑓 respectively in

∆𝑥𝑖 .
Therefore
𝑚𝑖′ + 𝑚𝑖′′ ≤ 𝑚𝑖 ≤ 𝑀𝑖 ≤ 𝑀𝑖′ + 𝑀𝑖′′ ……………(1)
Multiplying inequality (1) by ∆𝑥𝑖 , we get
(𝑚𝑖′ + 𝑚𝑖′′ )∆𝑥𝑖 ≤ 𝑚𝑖 ∆𝑥𝑖 ≤ 𝑀𝑖 ∆𝑥𝑖 ≤ (𝑀𝑖′ + 𝑀𝑖′′ )∆𝑥𝑖
Adding all these inequalities for 𝑖 = 1,2,3, … , 𝑛, we get

𝑛 𝑛 𝑛
∑(𝑚𝑖′ + 𝑚𝑖′′ )∆𝑥𝑖 ≤ ∑ 𝑚𝑖 ∆𝑥𝑖 ≤ ∑ 𝑀𝑖 ∆𝑥𝑖

𝑖=1 1=1 1=1
≤ ∑𝑛1=1(𝑀𝑖′ + 𝑀𝑖′′ )∆𝑥𝑖
⇒ 𝐿(𝑃, 𝑓1 ) + 𝐿(𝑃, 𝑓2 ) ≤ 𝐿(𝑃, 𝑓) ≤ 𝑈(𝑃, 𝑓) ≤ 𝑈(𝑃, 𝑓1 ) + 𝑈(𝑃, 𝑓2 )
𝑈 (𝑃, 𝑓) ≤ 𝑈 (𝑃, 𝑓1 ) + 𝑈(𝑃, 𝑓2 ) ……….(2)
𝐿(𝑃, 𝑓1 ) + 𝐿(𝑃, 𝑓2 ) ≤ 𝐿(𝑃, 𝑓)
−𝐿(𝑃, 𝑓) ≤ −(𝐿(𝑃, 𝑓1 ) + 𝐿(𝑃, 𝑓2 )) ………….(3)
It is given that 𝑓1 and 𝑓2 are integrable. Hence for any partition 𝑃 there
exists 𝛿 > 0 such that the norm 𝜇 (𝑃) < 𝛿, we have
1
𝑈(𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓1 ) < 𝜀………………………….(4)
2
1
𝑈 (𝑃, 𝑓2 ) − 𝐿(𝑃, 𝑓2 ) < 2 𝜀…………………………...(5)
From (2),(3),(4) and (5), we get
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) ≤ 𝑈(𝑃, 𝑓1 ) + 𝑈 (𝑃, 𝑓2 ) − (𝐿(𝑃, 𝑓1 ) + 𝐿(𝑃, 𝑓2 ))

1
= 𝑈(𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓1 ) + 𝑈(𝑃, 𝑓2 ) − 𝐿(𝑃, 𝑓2 ) <2 𝜀 +
1
𝜀
2
Therefore
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀 .
Hence the function 𝑓 is integrable.
𝑓1 and 𝑓2 are integrable and 𝜀 > 0 is any positive number.
Using Darboux’s theorem, there exists 𝛿 > 0 such that for all partitions 𝑃
whose norm 𝜇(𝑃) < 𝛿, we have
𝑏 1
𝑈 (𝑃, 𝑓1 ) < ∫𝑎 𝑓1 𝑑𝑥 + 2 𝜀 ……………(6)
And
𝑏 1
𝑈(𝑃, 𝑓2 ) < ∫𝑎 𝑓2 𝑑𝑥 + 2 𝜀 ……………(7)

𝑏
∫𝑎 𝑓 𝑑𝑥 ≤ 𝑈 (𝑃, 𝑓) ≤ 𝑈 (𝑃, 𝑓1 ) + 𝑈(𝑃, 𝑓2 )
Using inequalities (6) and (7), we get

𝑏 𝑏 1 𝑏 1 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 < ∫𝑎 𝑓1 𝑑𝑥 + 2 𝜀 + ∫𝑎 𝑓2 𝑑𝑥 + 2 𝜀 = ∫𝑎 𝑓1 𝑑𝑥 + ∫𝑎 𝑓2 𝑑𝑥 + 𝜀
As we know 𝜀 is arbitrary, therefore
𝑏 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 ≤ ∫𝑎 𝑓1 𝑑𝑥 + ∫𝑎 𝑓2 𝑑𝑥 ……………..(8)
Now replacing 𝑓1 and 𝑓2 with (−𝑓1 ) and (−𝑓2 ) respectively, we get

𝑏 𝑏 𝑏
∫𝑎 (−𝑓) 𝑑𝑥 ≤ ∫𝑎 (−𝑓1 ) 𝑑𝑥 + ∫𝑎 (−𝑓2 ) 𝑑𝑥
𝑏 𝑏 𝑏
i.e. ∫𝑎 𝑓 𝑑𝑥 ≥ ∫𝑎 𝑓1 𝑑𝑥 + ∫𝑎 𝑓2 𝑑𝑥 …………..(9)
From inequality (8) and (9), we get

𝑏 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓1 𝑑𝑥 + ∫𝑎 𝑓2 𝑑𝑥
Theorem 10.4.4. Let 𝒇𝟏 and 𝒇𝟐 are two bounded and integrable

function on [𝒂, 𝒃] then 𝒇 = 𝒇𝟏 − 𝒇𝟐 is also integrable on [𝒂, 𝒃] and
𝑏 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓1 𝑑𝑥 − ∫𝑎 𝑓2 𝑑𝑥
Proof. Let 𝑓1 and 𝑓2 are two bounded ⇒ 𝑓 = 𝑓1 + (−𝑓2 ) is bounded on

[𝑎, 𝑏].
Let 𝑃 be any partition 𝑃 of [𝑎, 𝑏] such that 𝑃 = {𝑎 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 =

𝑏 }.
Let 𝑀𝑖′ and 𝑚𝑖′ are the upper and lower bound of 𝑓1 respectively and 𝑀𝑖′′
and 𝑚𝑖′′ are the upper and lower bound of 𝑓2 respectively in ∆𝑥𝑖 .
⇒ −𝑀𝑖′′ and −𝑚𝑖′′ are the upper and lower bound of (−𝑓2 ) respectively
in ∆𝑥𝑖 .
Assume 𝑀𝑖 and 𝑚𝑖 are the upper and lower bound of 𝑓 respectively in

∆𝑥𝑖 .
Therefore
m′i + (−m′′i ) ≤ mi ≤ Mi ≤ Mi′ + (− Mi′′ )
⇒ m′i − Mi′ ≤ mi ≤ Mi ≤ Mi′′ − m′′i …………(1)
Multiplying inequality (1) by ∆xi , we get
(m′i − Mi′′ )∆xi ≤ mi ∆xi ≤ Mi ∆xi ≤ (Mi′ − m′′i )∆xi
Adding all these inequalities for i = 1,2,3, … , n, we get

𝑛
∑(𝑚𝑖′ − 𝑀𝑖′ )∆𝑥𝑖 ≤

𝑖=1
∑𝑛1=1 𝑚𝑖 ∆𝑥𝑖 ≤ ∑𝑛1=1 𝑀𝑖 ∆𝑥𝑖 ≤ ∑𝑛1=1(𝑀𝑖′ − 𝑚𝑖′′ )∆𝑥𝑖
⇒ 𝐿(𝑃, 𝑓1 ) − 𝑈 (𝑃, 𝑓2 ) ≤ 𝐿(𝑃, 𝑓) ≤ 𝑈 (𝑃, 𝑓) ≤ 𝑈 (𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓2 )
𝑈 (𝑃, 𝑓) ≤ 𝑈 (𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓2 ) ………….(2)
𝐿(𝑃, 𝑓1 ) − 𝑈(𝑃, 𝑓2 ) ≤ 𝐿(𝑃, 𝑓)
−𝐿(𝑃, 𝑓) ≤ 𝑈(𝑃, 𝑓2 ) − 𝐿(𝑃, 𝑓1 ) …………….(3)
It is given that 𝑓1 and 𝑓2 are integrable. Hence for any partition 𝑃 there
exists 𝛿 > 0 such that the norm 𝜇 (𝑃) < 𝛿, we have
1
𝑈(𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓1 ) < 2 𝜀 ……………….(4)
1
𝑈 (𝑃, 𝑓2 ) − 𝐿(𝑃, 𝑓2 ) < 𝜀 ………………...(5)
2
From (2), (3), (4) and (5), we get
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) ≤ 𝑈(𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓2 ) + 𝑈(𝑃, 𝑓2 ) − 𝐿(𝑃, 𝑓1 )
= 𝑈 (𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓1 ) + 𝑈 (𝑃, 𝑓2 ) − 𝐿(𝑃, 𝑓2 )

1 1
<2𝜀 + 2𝜀
Therefore
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀 .
Hence the function 𝑓 is integrable.
𝑓1 and 𝑓2 are integrable and 𝜀 > 0 is any positive number.
Using Darboux’s theorem, there exists 𝛿 > 0 such that for all partitions 𝑃
whose norm 𝜇(𝑃) < 𝛿, we have
𝑏 1
𝑈 (𝑃, 𝑓1 ) < ∫𝑎 𝑓1 𝑑𝑥 + 2 𝜀 ……………(6)
And
𝑏 1
𝐿(𝑃, 𝑓2 ) > ∫𝑎 𝑓2 𝑑𝑥 + 2 𝜀
𝑏 1
⇒ −𝐿(𝑃, 𝑓2 ) < − ∫𝑎 𝑓2 𝑑𝑥 + 2 𝜀 ……………(7)

𝑏
∫𝑎 𝑓 𝑑𝑥 ≤ 𝑈 (𝑃, 𝑓) ≤ 𝑈 (𝑃, 𝑓1 ) − 𝐿(𝑃, 𝑓2 )
Using inequalities (6) and (7), we get

𝑏 𝑏 1 𝑏 1 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 < ∫𝑎 𝑓1 𝑑𝑥 + 2 𝜀 − ∫𝑎 𝑓2 𝑑𝑥 + 2 𝜀 = ∫𝑎 𝑓1 𝑑𝑥 − ∫𝑎 𝑓2 𝑑𝑥 + 𝜀
As we know 𝜀 is arbitrary, therefore

𝑏 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 ≤ ∫𝑎 𝑓1 𝑑𝑥 − ∫𝑎 𝑓2 𝑑𝑥 …………………………..(8)
Now replacing 𝑓1 and 𝑓2 with (−𝑓1 ) and (−𝑓2 ) respectively, we get

𝑏 𝑏 𝑏
∫𝑎 (−𝑓) 𝑑𝑥 ≤ ∫𝑎 (−𝑓1 ) 𝑑𝑥 − ∫𝑎 (−𝑓2 ) 𝑑𝑥
𝑏 𝑏 𝑏
i.e. ∫𝑎 𝑓 𝑑𝑥 ≥ ∫𝑎 𝑓1 𝑑𝑥 − ∫𝑎 𝑓2 𝑑𝑥…………………………..(9)
From inequality (8) and (9), we get

𝑏 𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 𝑓1 𝑑𝑥 − ∫𝑎 𝑓2 𝑑𝑥 .
Oscillation: The oscillation of a bounded function f on an interval

[𝒂, 𝒃] is the supremum of the set {|𝒇(𝒙𝟏 ) − 𝒇(𝒙𝟐 )|: 𝒙𝟏 , 𝒙𝟐 ∈ [𝒂, 𝒃]} of
numbers.
Let 𝑀 and 𝑚 be the upper and lower bounds of 𝑓 on [𝑎, 𝑏] respectively.
⇒𝑚 ≤ 𝑓 (𝑥1 ) ≤ 𝑀 and 𝑚 ≤ 𝑓 (𝑥2 ) ≤ 𝑀 for all 𝑥1 , 𝑥2 ∈ [𝑎, 𝑏]
⇒ |𝑓(𝑥1 ) − 𝑓 (𝑥2 )| ≤ 𝑀 − 𝑚 for all 𝑥1 , 𝑥2 ∈ [𝑎, 𝑏] ………..(1)
⇒ 𝑀 − 𝑚 is an upper bound of {𝑓 (𝑥1 ) − 𝑓 (𝑥2 ) , for all 𝑥1 , 𝑥2 ∈ [𝑎, 𝑏]}
Let 𝜀 > 0 be any positive number, because 𝑀 is supremum of 𝑓.
Therefore, there exists 𝑦 ∈ [𝑎, 𝑏] such that

1
𝑓 (𝑦 ) > 𝑀 − 𝜀 …………(2)
2
Similarly, there exists 𝑧 ∈ [𝑎, 𝑏] such that

1
𝑓 (𝑧 ) > 𝑚 + 2 𝜀 …………(3)
From inequalities (2) and (3), we conclude that there exist 𝑥, 𝑦 ∈ [𝑎, 𝑏]
such that
1 1
𝑓 (𝑦 ) − 𝑓 (𝑧 ) > 𝑀 − 2 𝜀 − 𝑚 − 2 𝜀 = 𝑀 − 𝑚 − 𝜀
Or |𝑓(𝑦) − 𝑓 (𝑧)| > 𝑀 − 𝑚 − 𝜀 …………(4)
From inequalities (1) and (4), we conclude that
𝑀 − 𝑚 is an upper bound and also number less than 𝑀 − 𝑚 cannot be

upper bound of given set.
Hence 𝑀 − 𝑚 = sup{ |𝑓(𝑦) − 𝑓(𝑧)|: 𝑦, 𝑧 ∈ [𝑎, 𝑏]} .......(A)
10.5 SOME IMPORTANT THEOREM
Theorem 10.5.1. If 𝒇 and 𝒈 are two bounded and integrable functions

on [𝒂, 𝒃] then the product 𝒇𝒈 is also bounded and integrable on [𝒂, 𝒃].
Proof. It is given that 𝑓 and 𝑔 are two bounded therefore there exists 𝛼
such tha
|𝑓 (𝑥 )| ≤ 𝛼 and |𝑔(𝑥 )| ≤ 𝛼 for all 𝑥 ∈ [𝑎, 𝑏]
⇒ |𝑓𝑔(𝑥 )| = |𝑓(𝑥)||𝑔(𝑥 )| ≤ 𝛼. 𝛼 ≤ 𝛼 2
It implies that 𝑓𝑔 is bounded on [𝑎, 𝑏].
Let 𝑃 = {𝑎 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 = 𝑏} be any partition of [𝑎, 𝑏].
Let 𝑀𝑖′ and 𝑚𝑖′ are the upper and lower bound of 𝑓 respectively and 𝑀𝑖′′
and 𝑚𝑖′′ are the upper and lower bound of 𝑔 respectively in ∆𝑥𝑖 .
Assume 𝑀𝑖 and 𝑚𝑖 are the upper and lower bound of 𝑓𝑔 respectively in

∆𝑥𝑖 .
Now for all 𝑥, 𝑥 ′ ∈ ∆𝑥𝑖 ,
(𝑓𝑔)(𝑥 ′ ) − (𝑓𝑔)(𝑥) = 𝑓 (𝑥 ′ )𝑔(𝑥 ′ ) − 𝑓 (𝑥 )𝑔(𝑥 )
= 𝑓 (𝑥 ′ )𝑔(𝑥 ′ ) − 𝑓 (𝑥 )𝑔(𝑥 ′ ) + 𝑓 (𝑥 )𝑔(𝑥 ′ ) − 𝑓(𝑥 )𝑔(𝑥 )
= 𝑔(𝑥 ′ )(𝑓 (𝑥 ′ ) − 𝑓(𝑥 )) + 𝑓(𝑥)(𝑔(𝑥 ′ ) − 𝑔(𝑥 ))
It implies that
|(𝑓𝑔)(𝑥 ′ ) − (𝑓𝑔)(𝑥)| = |𝑔(𝑥 ′ )(𝑓 (𝑥 ′ ) − 𝑓 (𝑥 )) + 𝑓(𝑥 )(𝑔(𝑥 ′ ) − 𝑔(𝑥 ))|
≤ |𝑔(𝑥 ′ )||𝑓(𝑥 ′ ) − 𝑓(𝑥)| + |𝑓(𝑥 )|| 𝑔(𝑥 ′ ) − 𝑔(𝑥 )|
Hence, From inequality (A), we get
𝑀 − 𝑚 ≤ 𝛼(𝑀′ − 𝑚′ ) + 𝛼(𝑀′′ − 𝑚′′ ) ……………………(1)
Let 𝜀 > 0 be given number and it is given that 𝑓 and 𝑔 integrable on

Therefore there exists a positive number 𝛿 > 0 such that for any partition
𝑃 with norm 𝜇 (𝑃) < 𝛿
𝜀
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) ≤ 2𝛼 …………………………………..(2) and
𝜀
𝑈 (𝑃, 𝑔) − 𝐿(𝑃, 𝑔) ≤ 2𝛼 …………………………………..(3)
Now multiply inequality (1) with ∆𝑥𝑖 , we get
(𝑀 − 𝑚)∆𝑥𝑖 ≤ 𝛼(𝑀′ − 𝑚′ )∆𝑥𝑖 + 𝛼 (𝑀′′ − 𝑚′′ )∆𝑥𝑖
∑𝑛1=1(𝑀 − 𝑚)∆𝑥𝑖 ≤ ∑𝑛1=1 𝛼 (𝑀′ − 𝑚′ )∆𝑥𝑖 + ∑𝑛1=1 𝛼(𝑀′′ − 𝑚′′ )∆𝑥𝑖
⇒ ∑𝑛1=1 𝑀∆𝑥𝑖 − ∑𝑛1=1 𝑚∆𝑥𝑖 ≤ 𝛼 (∑𝑛1=1 𝑀′ ∆𝑥𝑖 − ∑𝑛1=1 𝑚′ ∆𝑥𝑖 ) +

𝛼 (∑𝑛1=1 𝑀 ′′ ∆𝑥𝑖 − ∑𝑛1=1 𝑚′′ ∆𝑥𝑖 )
⇒ 𝑈(𝑃, 𝑓𝑔) − 𝐿(𝑃, 𝑓𝑔) ≤ 𝛼(𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓)) +
𝛼(𝑈 (𝑃, 𝑔) − 𝐿(𝑃, 𝑔))

𝜀 𝜀
≤ 𝛼 2𝛼 + 𝛼 2𝛼
Therefore 𝑈(𝑃, 𝑓𝑔) − 𝐿(𝑃, 𝑓𝑔) ≤ 𝜀
Hence, we conclude that 𝑓𝑔 is integrable on [𝑎, 𝑏].
Theorem 10.5.2. If 𝒇 and 𝒈 are two bounded and integrable functions

on [𝒂, 𝒃] and there exists a positive number 𝒌 such that |𝒈| ≥
𝒌 𝒇𝒐𝒓 𝒂𝒍𝒍 𝒙 ∈ [𝒂, 𝒃] then the 𝒇/𝒈 is also bounded and integrable on
[𝒂, 𝒃].
Proof. It is given that 𝑓 and 𝑔 are two bounded therefore there exists 𝛼
such that
1 1 1
|𝑓 (𝑥 )| ≤ 𝛼 and 𝑘 ≤ |𝑔(𝑥 )| ≤ 𝛼 ⇒ ≥ ≥ 𝛼 for all 𝑥 ∈ [𝑎, 𝑏]
𝑘 |𝑔(𝑥)|
1 𝛼
⇒ |(𝑓/𝑔)(𝑥 )| = |𝑓 (𝑥 )|/|𝑔(𝑥 )| ≤ 𝛼. 𝑘 ≤ 𝑘
It implies that 𝑓𝑔 is bounded on [𝑎, 𝑏].
Let 𝑃 = {𝑎 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 = 𝑏} be any partition of [𝑎, 𝑏].
Let 𝑀𝑖′ and 𝑚𝑖′ are the upper and lower bound of 𝑓 respectively and 𝑀𝑖′′
and 𝑚𝑖′′ are the upper and lower bound of 𝑔 respectively in ∆𝑥𝑖 .
Assume 𝑀𝑖 and 𝑚𝑖 are the upper and lower bound of 𝑓/𝑔 respectively in
∆𝑥𝑖 .
𝑓 𝑓 𝑓(𝑥 ′ ) 𝑓(𝑥) 𝑓(𝑥 ′)𝑔(𝑥)−𝑓(𝑥)𝑔(𝑥 ′)

|( ) (𝑥 ′ ) − ( ) (𝑥 )| = | ′) − | =| |
𝑔 𝑔 𝑔(𝑥 𝑔(𝑥) 𝑔(𝑥)𝑔(𝑥 ′)
𝑓(𝑥 ′ )𝑔(𝑥)−𝑓(𝑥)𝑔(𝑥)+𝑓(𝑥)𝑔(𝑥)−𝑓(𝑥)𝑔(𝑥 ′)
=| |
𝑔(𝑥)𝑔(𝑥 ′)
𝑔(𝑥)(𝑓(𝑥 ′ )−𝑓(𝑥))−𝑓(𝑥)(𝑔(𝑥 ′)−𝑔(𝑥))

=| |
𝑔(𝑥)𝑔(𝑥 ′)
|𝑓(𝑥 ′ )−𝑓(𝑥)| |𝑔(𝑥 ′)−𝑔(𝑥)|

≤𝛼 +𝛼
|𝑔(𝑥)𝑔(𝑥 ′)| |𝑔(𝑥)𝑔(𝑥 ′)|
Hence, From inequality (A), we get

1 1
𝑀 − 𝑚 ≤ 𝛼. (𝑀′ − 𝑚′ ). 𝑘 2 + 𝛼. (𝑀 ′′ − 𝑚′′ ). 𝑘 2
Hence
𝛼 𝛼
𝑀−𝑚 ≤ (𝑀 ′ − 𝑚 ′ ) + (𝑀′′ − 𝑚′′ ) ………(1)
𝑘2 𝑘2
Let 𝜀 > 0 be given number and it is given that 𝑓 and 𝑔 integrable on

Therefore, there exists a positive number 𝛿 > 0 such that for any partition
𝑃 with norm
𝜇 (𝑃 ) < 𝛿
𝜀𝑘 2
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) ≤ ……..(2) and
2𝛼
𝜀𝑘 2
𝑈 (𝑃, 𝑔) − 𝐿(𝑃, 𝑔) ≤ ………..(3)
2𝛼

𝛼 𝛼
(𝑀 − 𝑚)∆𝑥𝑖 ≤ 𝑘 2 (𝑀 ′ − 𝑚′ )∆𝑥𝑖 + 𝑘 2 (𝑀′′ − 𝑚′′ )∆𝑥𝑖

𝛼
∑𝑛1=1(𝑀 − 𝑚)∆𝑥𝑖 ≤ ∑𝑛1=1 2 (𝑀′ − 𝑚′ )∆𝑥𝑖 + ∑𝑛1=1 𝛼 (𝑀′′ − 𝑚′′ )∆𝑥𝑖
𝑘
𝛼
⇒ ∑𝑛1=1 𝑀∆𝑥𝑖 − ∑𝑛1=1 𝑚∆𝑥𝑖 ≤ 𝑘 2 (∑𝑛1=1 𝑀′ ∆𝑥𝑖 − ∑𝑛1=1 𝑚′ ∆𝑥𝑖 ) +
𝛼
(∑𝑛1=1 𝑀′′ ∆𝑥𝑖 − ∑𝑛1=1 𝑚′′ ∆𝑥𝑖 )
𝑘2
𝛼 𝛼
⇒ 𝑈 (𝑃, 𝑓𝑔) − 𝐿(𝑃, 𝑓𝑔 ) ≤ 𝑘 2 (𝑈(𝑃, 𝑓) − 𝐿(𝑃, 𝑓)) + 𝑘 2 (𝑈(𝑃, 𝑔) −
𝐿(𝑃, 𝑔))
𝛼 𝜀𝑘 2 𝛼 𝜀𝑘 2
≤ 𝑘2 + 𝑘2
2𝛼 2𝛼
Therefore 𝑈(𝑃, 𝑓𝑔) − 𝐿(𝑃, 𝑓𝑔) ≤ 𝜀
Hence we conclude that 𝑓/𝑔 is integrable on [𝑎, 𝑏].
Theorem 10.5.3. If 𝒇 ís bounded and integrable functions on [𝒂, 𝒃]

𝒃
then |𝒇| is also bounded and integrable on [𝒂, 𝒃] and also |∫𝒂 𝒇 𝒅𝒙| ≤
𝒃
∫𝒂 |𝒇| 𝒅𝒙.
Proof. It is given that 𝑓 is bounded therefore there exists 𝛼 such that
|𝑓 (𝑥 )| ≤ 𝛼 for all 𝑥 ∈ [𝑎, 𝑏]
It implies that the function |𝑓| is bounded.
Since 𝑓 is integrable, for a given positive number 𝜀 > 0 there exists a

partition 𝑃 = {𝑎 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 = 𝑏} of [𝑎, 𝑏] and such that
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀 …………………..(1)
Let 𝑀𝑖 and 𝑚𝑖 are the upper and lower bound of 𝑓 respectively and 𝑀𝑖′
and 𝑚𝑖′ are the upper and lower bound of 𝑔 respectively in ∆𝑥𝑖 .
||𝑓|(𝑥) − |𝑓|(𝑦)| = ||𝑓(𝑥)| − |𝑓 (𝑦)|| ≤ |𝑓 (𝑥 ) − 𝑓 (𝑦)|
⇒ 𝑀𝑖′ − 𝑚𝑖′ ≤ 𝑀 − 𝑚………………………..(2)
(𝑀𝑖′ − 𝑚𝑖′ )∆𝑥𝑖 ≤ (𝑀𝑖 − 𝑚𝑖 )∆𝑥𝑖
∑𝑛1=1(𝑀𝑖′ − 𝑚𝑖′ )∆𝑥𝑖 ≤ ∑𝑛1=1(𝑀𝑖 − 𝑚𝑖 )∆𝑥𝑖
⇒ ∑𝑛1=1 𝑀𝑖′ ∆𝑥𝑖 − ∑𝑛1=1 𝑚𝑖′ ∆𝑥𝑖 ≤ ∑𝑛1=1 𝑀𝑖 ∆𝑥𝑖 − ∑𝑛1=1 𝑚𝑖 ∆𝑥𝑖
⇒ 𝑈 (𝑃, |𝑓|) − 𝐿(𝑃, |𝑓|) ≤ 𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓)
𝑈 (𝑃, |𝑓|) − 𝐿(𝑃, |𝑓|) < 𝜀 . Hence |𝑓| is integrable on [𝑎, 𝑏].
We Know that if 𝑓 and 𝑔 are bounded and integrable on [𝑎, 𝑏] such that
𝑓 ≥ 𝑔 then
𝑏 𝑏
∫𝑎 𝑓 𝑑𝑥 ≤ ∫𝑎 𝑔 𝑑𝑥 when 𝑏 ≤ 𝑎
𝑏 𝑏
Hence ∫𝑎 𝑓 𝑑𝑥 ≤ ∫𝑎 |𝑓| 𝑑𝑥
𝑏 𝑏 𝑏
and − ∫𝑎 𝑓 𝑑𝑥 = ∫𝑎 (−𝑓) 𝑑𝑥 ≤ ∫𝑎 |𝑓| 𝑑𝑥
𝑏 𝑏
⇒ |∫𝑎 𝑓 𝑑𝑥 | ≤ ∫𝑎 |𝑓| 𝑑𝑥
Note: The Converse of the above theorem is not true. For example, the
function
1, 𝑤ℎ𝑒𝑛 𝑥 𝑖𝑠 𝑟𝑎𝑡𝑖𝑜𝑛𝑎𝑙
𝑓 (𝑥 ) = {
−1, 𝑤ℎ𝑒𝑛 𝑥 𝑖𝑠 𝑖𝑟𝑟𝑎𝑡𝑖𝑜𝑛𝑎𝑙
−𝑏 −𝑏
Here ∫𝑎 𝑓 𝑑𝑥 = 𝑏 − 𝑎 but ∫𝑎 𝑓 𝑑𝑥 = 𝑎 − 𝑏
It implies that 𝑓 is not integrable.

𝑏
But |𝑓 (𝑥 )| = 1 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥, therefore ∫𝑎 |𝑓| 𝑑𝑥 exists and equal to 𝑏 − 𝑎.
Here we observe that |𝑓| is integrable.
Theorem 10.5.4. Every Monotonic function 𝑓 is Riemann integrable.
Proof: Let us suppose that the function 𝑓 is monotonically increasing

function on [𝑎, 𝑏].
Now for a given positive number 𝜀, there exists a partition P =

{𝑎 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 = 𝑏} of
𝜀
[𝑎, 𝑏] such that the length of each sub – interval <
[𝑓(𝑎)−𝑓(𝑏)+1]
𝜀
i.e. (𝑥𝑟 − 𝑥𝑟−1 ) < for 𝑟 = 1, 2, … , 𝑛 ……. (1)
[𝑓(𝑎)−𝑓(𝑏)+1]
Again, the function 𝑓being monotonically increasing on [𝑎, 𝑏], it is

bounded and monotonically increasing on each sub – interval [𝑥𝑟−1 , 𝑥𝑟 ].
Let the bounds of function f on the sun – interval [𝑥𝑟−1 , 𝑥𝑟 ] be 𝑀𝑟 and 𝑚𝑟 ,

then
𝑀𝑟 = 𝑓(𝑥𝑟 ) and 𝑚𝑟 = 𝑓(𝑥𝑟−1 ) ……. (2)
Therefore, for this partition P, we find that
𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) = ∑𝑛𝑖=1(𝑀𝑟 − 𝑚𝑟 )(𝑥𝑟 − 𝑥𝑟−1 )

𝜀
< [𝑓(𝑎)−𝑓(𝑏)+1] ∑𝑛𝑖=1[𝑓(𝑥𝑟 ) − 𝑓(𝑥𝑟−1 )]
𝜀
Therefore, 𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < ∑𝑛𝑖=1[𝑓(𝑥𝑛 ) − 𝑓(𝑥0 )]
[𝑓(𝑎)−𝑓(𝑏)+1]
𝜀
Therefore, 𝑈 (𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < [𝑓(𝑎)−𝑓(𝑏)+1] ∑𝑛𝑖=1[𝑓(𝑏) − 𝑓(𝑎)]
Therefore, 𝑈(𝑃, 𝑓) − 𝐿(𝑃, 𝑓) < 𝜀
Therefore, Every Monotonically increasing function 𝑓 is Riemann

integrable.
Similarly, we can prove that Every Monotonically decreasing function

𝑓 is Riemann integrable.
Therefore, every monotonic function is Riemann integrable.
10.6 RIEMANN SUM

Riemann Sum: Let 𝑃′ is the tagged partition then the Riemann sum of a
function 𝑓 ∶ [𝑎, 𝑏] → ℝ corresponding to 𝑃′ can be defined as
𝑆 (𝑓, 𝑃′ ) = ∑𝑛𝑖=1 𝑓 (𝑡𝑖 ) (𝑥𝑖 − 𝑥𝑖−1 )
If the function 𝑓 is positive on [𝑎, 𝑏], then the Riemann Sum is the sum of
the areas of 𝑛 rectangles whose bases are the subintervlas 𝐼1 = [𝑥𝑖−1 , 𝑥𝑖 ]
and whose heights are 𝑓(𝑡𝑖 ). See Fig 5.1.
𝑥0 = 𝑎 𝑡1 𝑥1 𝑡2 𝑥2 𝑡3 𝑥3 𝑡𝑛 𝑥𝑛 = 𝑏
𝑥𝑛−1
Fig 5.1. A Riemann Sum
Theorem. If 𝑓: [𝑎, 𝑏] → ℝ is continuous, then 𝑓 is Riemann integrable.
Proof. Let 𝜀 > 0 be given.
Now 𝑓 is continuous on [𝑎, 𝑏] ⇒ It is also uniformly continuous.

𝜀
Therefore, there exists a 𝛿 > 0 such that |𝑓(𝑥) − 𝑓(𝑦)| < 𝑏−𝑎 whenever
|𝑥 − 𝑦| < 𝛿.
For any large integer 𝑁 we assume an equally spaced partition

𝑏−𝑎
𝑥𝑘 = 𝑎 + 𝑘ℎ, with ℎ = and 𝑘 = 0,1, … , 𝑁. We choose 𝑁 so large
𝑁
𝑏−𝑎
that < 𝛿.
𝑁
Now function 𝑓 is continuous on any of the intervals [𝑥𝑘−1 , 𝑥𝑘 ],
Hence there must exist points 𝑐𝑘 , 𝑑𝑘 ∈ [𝑥𝑘−1 , 𝑥𝑘 ] where 𝑓 attains its

minimum and maximum, respectively, i.e.
𝑓(𝑐𝑘 ) ≤ 𝑓(𝑥) ≤ 𝑓(𝑑𝑘 ) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥 ∈ [𝑥𝑘−1 , 𝑥𝑘 ].

Let 𝑠, 𝑡: [𝑎, 𝑏] → ℝ are two step functions such that on each interval
[𝑥𝑘−1 , 𝑥𝑘 )
𝑠(𝑥) = 𝑓(𝑐𝑘 ) and 𝑡(𝑥 ) = 𝑓 (𝑑𝑘 ).
Therefore, we conclude that 𝑠(𝑥 ) ≤ 𝑓 (𝑥) ≤ 𝑡(𝑥 ) for some
𝑥 ∈ [𝑥𝑘−1 , 𝑥𝑘 )
𝑏−𝑎
Since |𝑐𝑘 − 𝑑𝑘 | ≤ < 𝛿 then for any 𝑥 ∈ [𝑥𝑘−1 , 𝑥𝑘 )
𝑁
𝜀
𝑡(𝑥 ) − 𝑠 (𝑥) = 𝑓 (𝑑𝑘 ) − 𝑓 (𝑐𝑘 ) < 𝑏−𝑎.
This also holds for each interval [𝑥𝑘−1 , 𝑥𝑘 ) (𝑘 = 1,2, … , 𝑁)
𝜀
Hence we shown that 0 ≤ 𝑡(𝑥 ) − 𝑠 (𝑥) < 𝑏−𝑎 for all 𝑥 ∈ [𝑎, 𝑏]
𝜀
Now compare the integrals of 𝑡 and 𝑠 and since 𝑡 ≤ 𝑠 + 𝑏−𝑎
𝑏 𝑏 𝜀 𝑏
Then ∫𝑎 𝑡(𝑥)𝑑𝑥 ≤ ∫𝑎 (𝑠(𝑥 ) + 𝑏−𝑎) 𝑑𝑥 = ∫𝑎 𝑠𝑑𝑥 + 𝜀.
Fundamental Theorem of Calculus

Theorem. A function 𝒇 is bounded and integrable on [𝒂, 𝒃] and there
𝒃
exists a function F such that 𝑭′ = 𝒇 on [𝒂, 𝒃], then ∫𝒂 𝒇(𝒙)𝒅𝒙 =
𝑭(𝒃) − 𝑭(𝒂)
Proof. It is given that 𝐹 ′ = 𝑓 is bounded and integrable on [𝑎, 𝑏].
Therefore, for every given 𝜀 > 0 there exists a positive number 𝛿 such
that for every partition 𝑃 = {𝑎 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 = 𝑏}, with norm
𝜇 (𝑃) < 𝛿.
𝑏
|∑𝑛𝑖=1 𝑓(𝑡𝑖 ) ∆𝑥𝑖 − ∫𝑎 𝑓 (𝑥 )𝑑𝑥| < 𝜀 …………………….(1)
For every choice of points 𝑡𝑖 in ∆𝑥𝑖 .

Because we have freedom in the selection of points 𝑡𝑖 in ∆𝑥𝑖 ., we choose
them in a particular way as follows:
By Lagrange Mean value theorem, we have
𝐹 (𝑥𝑖 ) − 𝐹(𝑥𝑖−1 ) = 𝐹 ′ (𝑡𝑖 )∆𝑥𝑖 (𝑖 = 1,2, … , 𝑛)
Hence 𝐹 (𝑥𝑖 ) − 𝐹(𝑥𝑖−1 ) = 𝑓(𝑡𝑖 )∆𝑥𝑖
It implies that ∑𝑛𝑖=1 𝑓 (𝑡𝑖 ) ∆𝑥𝑖 = ∑𝑛𝑖=1(𝐹(𝑥𝑖 ) − 𝐹 (𝑥𝑖−1 )) = 𝐹(𝑏) − 𝐹 (𝑎).
From inequality (1), we get
𝑏
∫𝑎 𝑓 (𝑥 )𝑑𝑥 = 𝐹(𝑏) − 𝐹(𝑎)
This theorem is also known as the Second Fundamental theorem of
Integral Calculus.
First Mean Value theorem
Theorem. A function 𝑓 is continuous on [𝑎, 𝑏], then there exists a number
𝑏
𝑘 in [𝑎, 𝑏] uch that ∫𝑎 𝑓 𝑑𝑥 = 𝑓(𝑘)(𝑏 − 𝑎).
Proof. It is given that 𝑓 is continuous on [𝑎, 𝑏], therefore f is Riemann
Integrable on [𝑎, 𝑏].
Let 𝑀 and 𝑚 are the upper and lower bound of 𝑓 on [𝑎, 𝑏] respectively.
As we know that
𝑏
𝑚(𝑏 − 𝑎) ≤ ∫𝑎 𝑓 𝑑𝑥 ≤ 𝑀(𝑏 − 𝑎)
Hence there exists a real number 𝛾 ∈ [𝑚, 𝑀] such that
𝑏
∫𝑎 𝑓 𝑑𝑥 = 𝛾(𝑏 − 𝑎)
Because 𝑓 is continuous on [𝑎, 𝑏], it attains every value between m and
M.
Hence, there exists a number 𝑘 ∈ [𝑎, 𝑏] such that 𝑓 (𝑘) = 𝛾.
𝑏
Therefore, ∫𝑎 𝑓 𝑑𝑥 = 𝑓(𝑘)(𝑏 − 𝑎)
CHECK YOUR PROGRESS
True or false/MCQ Questions
Problem 1.
then
(a)
(b)
(c) .
(d) None of the above.
Problem 2. Every continuous function on closed interval is not

Riemann integrable.
Problem 3. Every continuous function on closed interval is Riemann

integrable.
Problem 4. For every polynomial function Riemann upper integral is

equal to Riemann lower integral.
Problem 5. U(P, f) is decreasing, L(P, f) is decreasing function.
10.7 SUMMARY
1. Darboux Theorem
If 𝑓 is bounded function on [𝑎, 𝑏] then to every
𝜀 > 0, there corresponds 𝛿 > 0 such that

−𝑏
(i) 𝑈(𝑃, 𝑓) < ∫𝑎 𝑓 𝑑𝑥 + 𝜀
𝑏
(ii) 𝐿(𝑃, 𝑓) > ∫−𝑎 𝑓 𝑑𝑥 − 𝜀
For every partition 𝑃 of [𝑎, 𝑏] with norm 𝜇 (𝑃) < 𝛿
2. Every Monotonic function 𝑓 is Riemann integrable.
3. If 𝑓: [𝑎, 𝑏] → ℝ is continuous, then 𝑓 is Riemann integrable.
10.8 GLOSSARY
integration
continuity
Functions
Limits
10.9 REFERENCES




Q 1. Prove that every constant function is Reimann integrable.
Q 2. Prove that every polynomial function is Riemann integrable.
𝜋
Q 3. Show that the function 𝑓 (𝑥 ) = 𝑠𝑖𝑛𝑥 is integrable in [0, 2 ]
1 1
Q 4. Using Riemann integration prove ∫0 𝑥 𝑑𝑥 = 2 .
Q 5. Define upper and lower Riemann sums.
10.12 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. (c)
CYQ 2. False
CYQ 3. True
CYQ 4. True
CYQ 5. False
UNIT 11: SEQUENCE AND SERIES OF FUNCTION

Contents
11.1 Introduction
11.2 Objectives
11.3 Sequence of functions
11.4 Pointwise Convergence
11.5 Refinement of partitions and tagged partitions
11.7 Continuous Limit Theorem
11.8 Uniform Convergence and Differentiation
11.9 Series of Functions
11.10 Criterion for Uniform Convergence of Series
11.11 Abel’s test
11.12 Dirichlet’s test
11.13 Summary
11.14 Glossary
11.16 References
11.18 Answers
11.1 INTRODUCTION
Mathematics allows us to create sequences and series not only
for real numbers but also for functions. This article will give you a deeper
understanding of how to construct sequences and series for real-valued
functions. We will also delve into the concept of convergence in
sequences and series of functions. To solidify these concepts, we will
include some solved problems on sequences and series of functions.
11.2 OBJECTIVES
 Sequence of functions
 Series of functions
 Abel’s test
 Dirichlet’s test
11.3 SEQUENCE OF FUNCTION

Let 𝑓𝑛 be a real – valued function defined on an interval I and for each 𝑛 ∈
ℕ. Then
< 𝑓1 , 𝑓2 , 𝑓3 , …, 𝑓𝑛 , … > is called a sequence of real valued function.
Denoted by {𝑓𝑛 } 𝑜𝑟 〈𝑓𝑛 〉.
𝑠𝑖𝑛𝑛𝑥
Example: {𝑓𝑛 } = {𝑥 𝑛 , 0 ≤ 𝑥 ≤ 1} and { , 0 ≤ 𝑥 ≤ 1 } are sequence
𝑛
of functions.
11.4 POINTWISE CONVERGENCE
For each 𝑛 ∈ N, let 𝑓𝑛 : 𝐴 → R be a real-valued function on 𝐴. The

sequence ( 𝑓𝑛 ) of functions converges pointwise on 𝐴 to a function 𝑓 if,
for all 𝑥 ∈ 𝐴, the sequence of real numbers (𝑓𝑛 (𝑥)) converges to the real
number 𝑓(𝑥).
We often write lim 𝑛→∞ 𝑓𝑛 (𝑥) = 𝑓 (𝑥) or lim 𝑛→∞ 𝑓𝑛 = 𝑓
𝑥 2 + 𝑛𝑥 𝑥2
Example 1. 𝑓𝑛 (𝑥) = = +𝑥
𝑛 𝑛
𝑥2
lim ( + 𝑥) = 0 + 𝑥 = 𝑥.
n→∞ 𝑛
If 𝑓 (𝑥) = 𝑥, then 𝑓𝑛 → 𝑓 as 𝑛 → ∞. In this case, the functions 𝑓𝑛 are

everywhere continuous and differentiable, and the limit function is also
everywhere continuous and differentiable.
Example 2. Let 𝑓𝑛 (𝑥) = 𝑥 𝑛 on the set [0, 1].
0, 0 ≤ x < 1
lim 𝑓𝑛 (𝑥) = lim 𝑥 𝑛 = f(𝑥) = { .
n→∞ n→∞ 1, 𝑥=1
In this case, the functions 𝑓𝑛 (𝑥) are continuous on [0, 1], but the limit
function f(𝑥) is not continuous at every point of [0, 1].
Note: Suppose 𝑓 (𝑥) = lim 𝑓𝑛 (𝑥) for all 𝑥 ∈ 𝐴. then

n→∞
(i) If each 𝑓𝑛 is continuous on 𝐴, then 𝑓 is continuous on 𝐴.
(ii) if each 𝑓𝑛 is differentiable on 𝐴, then 𝑓 is differentiable on 𝐴.
11.5 UNIFORM CONVERGENCE

Let ( 𝑓𝑛 ) be a sequence of functions defined on 𝐴 ⊆ R. We say that (𝑓𝑛 )
converges uniformly on 𝐴 to the limit function 𝑓 defined on 𝐴 if for
every 𝜖 > 0 there exists an N ∈ ℕ such that |𝑓𝑛 (x) − f (x)| < ϵ
for all 𝑥 ∈ 𝐴, whenever 𝑛 ≥ 𝑁.
Note: In the definition, the value of 𝑁 is independent of 𝑥.
1
Example1: 𝑓𝑛 (𝑥) = √𝑥 2 + 𝑛2
1
Therefore, lim 𝑓𝑛 (𝑥) = lim √𝑥 2 + = |𝑥|.
n→∞ n→∞ 𝑛2
So, 𝑓𝑛 (𝑥) → 𝑓 (𝑥) = |𝑥| pointwise.
Let 𝜖 > 0 be given. Choose 𝑁 ∈ N large enough such that 1 𝑁 < 𝜖. Then
for any 𝑥 ∈ R and 𝑛 ≥ 𝑁 we have
1
|√𝑥 2+ 2 + |x||
1 1 𝑛
| 𝑓𝑛 (𝑥) − 𝑓 (𝑥)| = |√𝑥 2 + 𝑛2 − |x|| = |√𝑥 2 + 𝑛2 − |x|| ( 1
)
|√𝑥 2+ 2 + |x||
𝑛
1 1
𝑛2 𝑛2 1
= 1
≤ 1
=𝑛<𝜖
|√𝑥 2+ 2 + |x|| |√0 + 2 + 0|
𝑛 𝑛
This shows that (𝑓𝑛 ) → 𝑓 uniformly on R. Note that each 𝑓𝑛 (𝑥) is both
continuous and differentiable on R, but 𝑓 (𝑥) = |𝑥| is continuous on R and
not differentiable at 𝑥 = 0.
1
𝑛𝑥 𝑖𝑓 0 ≤ 𝑥 ≤ 1 𝑛
Example2: 𝑓𝑛 (𝑥) = {2 − 𝑛𝑥 𝑖𝑓 1
≤ 𝑥 ≤
2
𝑛 𝑛
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Therefore, lim 𝑓𝑛 (𝑥 ) = 𝑓(𝑥) = 0
n→∞
If f(𝑥) = 0, then (𝑓𝑛 ) → 𝑔 pointwise.

1
Let 𝜖 =1/2 and 𝑥𝑛 = 𝑛 then
| 𝑓𝑛 (𝑥𝑛 ) − 𝑓 (𝑥𝑛 )| = | 1− 0| = 1 > 𝜖 =1/2.
So, it is not true that for all 𝜖 > 0, there exist an 𝑁 ∈ ℕ large enough such
that 𝑛 ≥ 𝑁 implies | 𝑓𝑛 (𝑥𝑛 ) − 𝑓 (𝑥𝑛 )| < for all 𝑥.
So 𝑓𝑛 (𝑥 ) does not converge to 𝑓(𝑥) uniformly.
11.6 CAUCHY CRITERION FOR UNIFORM

CONVERGENCE
A sequence of functions ( 𝑓𝑛 ) defined on a set 𝐴 ⊆ R converges
uniformly on 𝐴 if and only if for every 𝜖 > 0 there exists an 𝑁 ∈ ℕ such
that | 𝑓𝑛 (𝑥) − 𝑓𝑚 (𝑥)| < 𝜖 whenever 𝑚, 𝑛 ≥ 𝑁 and 𝑥 ∈ 𝐴.
Proof: (⇒) Assume the sequence ( 𝑓𝑛 ) converges uniformly on 𝐴 to a

limit function 𝑓. Let 𝜖 > 0 be given. Then there exists an 𝑁 ∈ ℕ such that
ϵ
| 𝑓𝑛 (𝑥) − 𝑓 (𝑥)| < ,
2
whenever 𝑛 ≥ 𝑁 and 𝑥 ∈ 𝐴. Then if 𝑛, 𝑚 ≥ 𝑁 and 𝑥 ∈ 𝐴, we have
| 𝑓𝑛 (𝑥) − 𝑓𝑚 (𝑥)| = | 𝑓𝑛 (𝑥) – 𝑓 (𝑥) + 𝑓 (𝑥) - 𝑓𝑚 (𝑥)|
≤ | 𝑓𝑛 (𝑥) − 𝑓 (𝑥)| + | 𝑓𝑚 (𝑥) − 𝑓 (𝑥)|

ϵ ϵ
<2+2=𝜖
(⇐) Conversely, assume that for every 𝜖 > 0 there exists an 𝑁 ∈ N such
that | 𝑓𝑛 (𝑥) − 𝑓𝑚 (𝑥)| < 𝜖 whenever 𝑚, 𝑛 ≥ 𝑁 and 𝑥 ∈ 𝐴. This
hypothesis implies that, for each 𝑥 ∈ 𝐴, ( 𝑓𝑛 (𝑥)) is a Cauchy sequence.
By Cauchy’s Criterion, this sequence converges to a point, which we
will call 𝑓 (𝑥). So, the uniformly Cauchy sequence converges pointwise
to the function 𝑓 (𝑥). We must show that the convergence is also
uniform. For the value of 𝜖 given above, we use the corresponding 𝑁.
Then for 𝑛, 𝑚 ≥ 𝑁 and all 𝑥 ∈ 𝐴,
| 𝑓𝑛 (𝑥) − 𝑓𝑚 (𝑥)| < 𝜖
Taking the limit as 𝑚 → ∞ gives
| 𝑓𝑛 (𝑥) − 𝑓𝑚 (𝑥)| ≤ 𝜖 for all 𝑥 ∈ 𝐴, which shows that ( 𝑓𝑛 ) converges

uniformly to 𝑓 on 𝐴.
This completes the proof.
11.7 CONTINUOUS LIMIT THEOREM
Let ( 𝑓𝑛 ) be a sequence of functions defined on 𝐴 ⊆ R that converges

uniformly on 𝐴 to a function 𝑓. If each 𝑓𝑛 is continuous at 𝑐 ∈ 𝐴, then 𝑓
is continuous at 𝑐.
Proof. Let 𝜖 > 0 be given. Fix 𝑐 ∈ 𝐴. Since 𝑓𝑛 → 𝑓 uniformly, there

exists an 𝑁 ∈ ℕ such that
ϵ
| 𝑓𝑁 (𝑥) − 𝑓 (𝑥) | < 3 for all 𝑥 ∈ 𝐴.
Since 𝑓𝑁 is continuous at 𝑐, there exists 𝛿 > 0 such that

ϵ
| 𝑓𝑁 (𝑥) − 𝑓𝑁 (c) | < 3 whenever |𝑥 − 𝑐| < 𝛿.
If |𝑥 − 𝑐| < 𝛿, then
| 𝑓 (𝑥) − 𝑓 (c) | = | 𝑓 (𝑥) + 𝑓𝑁 (𝑥) −𝑓𝑁 (𝑥) − 𝑓𝑁 (c) + 𝑓𝑁 (c) − 𝑓 (c) |
≤ | 𝑓 (𝑥) − 𝑓𝑁 (c) | + | 𝑓𝑁 (𝑥) − 𝑓𝑁 (c) | + | 𝑓𝑁 (c) - 𝑓 (c) |

ϵ ϵ ϵ
<3+3+3=ϵ
The first and third 𝜖/3 are due to uniform convergence and the choice of
𝑁. The second 𝜖/3 is due to the choice of 𝛿. This shows that 𝑓 is
continuous at 𝑐, as desired.
11.8 UNIFORM CONVERGENCE AND

DIFFERENTIATION
 Differentiable Limit Theorem
Let (𝑓𝑛 ) → 𝑓 pointwise on the closed interval [𝑎, 𝑏] and
assume each 𝑓𝑛 is differentiable. If (𝑓𝑛′ ) → 𝑔 uniformly on
[𝑎, 𝑏], then 𝑓 is differentiable and 𝑓 ′ = 𝑔.
Proof: Fix 𝑐 ∈ [𝑎, 𝑏] and let 𝜖 > 0. We’ll show there exists 𝛿 > 0 such
that
whenever 0 < |𝑥 − 𝑐| < 𝛿 and 𝑥 ∈ [𝑎, 𝑏].
For 𝑥 ≠ 𝑐, consider the following:
From Cauchy’s Criterion for uniform convergence, since the sequence (

𝑓′𝑛 ) converges uniformly to 𝑔, there exists an 𝑁2 ∈ ℕ such that
whenever 𝑚, 𝑛 ≥ 𝑁2 and 𝑥 ∈ [𝑎, 𝑏]. Set 𝑁 = max{𝑁1 , 𝑁2 }
11.9 SERIES OF FUNCTION

Let {𝑓𝑛 } is a sequence of real valued functions on an interval I,
then 𝑓1 + 𝑓2 + ⋯ + 𝑓𝑛 + ⋯ is called a series of real valued function

defined on I.
this series is denoted by ∑∞

𝑛=1 𝑓𝑛 .
𝑥𝑛
Examples: 𝑓 (𝑥) = ∑∞
𝑛=1 (This equals 𝑒 𝑥 for all 𝑥 ∈ R.)
𝑛!
Note: Let 𝑓 and 𝑓𝑛 for 𝑛 ∈ N be functions defined on a set 𝐴 ⊆ R.
(a) The infinite series ∑︁𝑓𝑛 (𝑥) = 𝑓1 (𝑥) + 𝑓2 (𝑥) + 𝑓3 (𝑥) + · · · converges
pointwise on 𝑨 to 𝒇 (𝒙) if the sequence of partial sums
𝑠𝑘 (𝑥) = 𝑓1 (𝑥) + 𝑓2 (𝑥) + 𝑓3 (𝑥) + · · · +𝑓𝑘 (𝑥) converges pointwise to 𝑓

(𝑥) on 𝐴.
(b) The infinite series converges uniformly on 𝑨 to 𝒇 (𝒙) if the sequence

of partial sums converges uniformly on 𝐴 to 𝑓 (𝑥).
Note: Since an infinite series of functions is defined in terms of the limit

of a sequence of partial sums, everything we already know about
sequences applies to series. For the sum ∑∞ 𝑛=1 𝑓𝑛 (𝑥), we merely restate
all of the previous theorems for the sequence of 𝑘th partial sums 𝑠𝑘 (𝑥) =
𝑓1 (𝑥) + 𝑓2 (𝑥) + 𝑓3 (𝑥) + · · · +𝑓𝑘 (𝑥).
Note: Let 𝑓𝑛 be continuous functions defined on a set 𝐴 ⊆ R, and assume

∑∞𝑛=1 𝑓𝑛 (𝑥 ) converges uniformly on 𝐴 to a function 𝑓. Then 𝑓 is
continuous on 𝐴.
Note: Term-by-term Differentiability Theorem
Suppose the following three statements:
(i) Let 𝑓𝑛 be differentiable functions defined on an interval 𝐴 = [𝑎, 𝑏].
(ii) Assume ∑∞ ′
𝑛=1 𝑓𝑛 (𝑥) converges uniformly to a limit 𝑔(𝑥) on 𝐴.
(iii) There exists a point 𝑥0 ∈ [𝑎, 𝑏] where ∑∞𝑛=1 𝑓𝑛 (𝑥0 ) converges. Then
∞
the series ∑𝑛=1 𝑓𝑛 (𝑥 ) converges uniformly to a differentiable function 𝑓
(𝑥) satisfying 𝑓 ′ (𝑥) = 𝑔(𝑥) on 𝐴. In other words,
𝑓 (𝑥) = ∑∞
𝑛=1 𝑓𝑛 (𝑥 ) and 𝑓 ′ (𝑥) = ∑∞ ′
𝑛=1 𝑓𝑛 (𝑥).
11.10 CRITERION FOR UNIFORM

CONVERGENCE OF SERIES
1. Cauchy Criterion for Uniform Convergence of Series:
A series ∑∞
𝑛=1 𝑓𝑛 (𝑥 ) converges uniformly on 𝐴 ⊆ R if and only if for
every 𝜖 > 0 there exists an 𝑁 ∈ ℕ such that
|𝑓𝑚+1 (𝑥 ) + 𝑓𝑚+2 (𝑥 ) + ⋯ + 𝑓𝑛 (𝑥 )| < 𝜖 whenever 𝑛 > 𝑚 ≥ 𝑁 and 𝑥 ∈ A.
2. Weierstrass 𝑀-Test:
For each 𝑛 ∈ ℕ, let 𝑓𝑛 be a function defined on a set 𝐴 ⊆ R, and let 𝑀𝑛 >

0 be a real number satisfying |𝑓𝑛 (x)| < 𝑀𝑛 for all 𝑥 ∈ 𝐴. If ∑∞
𝑛=1 𝑀𝑛
∞
converges, then ∑𝑛=1 𝑓𝑛 (𝑥) converges uniformly on 𝐴.
cos 2𝑛 𝑥
Examples 1. The continuous functions for 𝑛 ∈ {0, 1, 2, 3, . . .}
2𝑛
satisfy
cos 2𝑛 𝑥 1
| | ≤ 𝑀𝑛 = 𝑛
2𝑛 2
for all 𝑥 ∈ R and 𝑛 ∈ {0, 1, 2, 3, . . .}. Since

1 1
∑∞ ∞
𝑛=0 𝑀𝑛 = ∑𝑛=0 2𝑛 = 1 = 2 < ∞,
1−( )
2
cos 2𝑛 𝑥
by the Weierstrass 𝑀-test, the series ∑∞
𝑛=0 converges uniformly to
2𝑛
a continuous function
cos 2𝑛 𝑥
g(x) =∑∞
𝑛=0 .
2𝑛
𝑥 2𝑛
Examples 2. Define 𝑓 (𝑥) = ∑∞
𝑛=0 1+𝑥 2𝑛 . Find the values of 𝑥 where the
series converges and show that we get a continuous function on this set.
Sol. If |𝑥| < 1, then by the Comparison Test, the series converges as
follows:
If |𝑥| ≥ 1, then the series diverges by the Divergence Test since
Now let 0 < 𝐾 < 1. Then on the interval [−𝐾, 𝐾] we have
Since
the series converges uniformly on [−𝐾, 𝐾] to a continuous function.

Since 𝐾 was arbitrary, the series
is a continuous function on (−1, 1).
What about the derivative? Consider
For 0 < 𝐾 < 1, we apply the Weierstrass 𝑀-test on the interval [−𝐾, 𝐾].
For 𝑥 ∈ [−𝐾, 𝐾],
The sum ∑∞ ∞
𝑛=0 𝑁𝑛 = ∑𝑛=0 𝑛𝐾
2𝑛−1
converges and by the Weierstrass 𝑀-
test the series of derivatives converges uniformly on the interval [−𝐾, 𝐾].
By the Differentiable Limit Theorem, 𝑓 ′ (𝑥) = 𝑔(𝑥) for 𝑥 ∈ [−𝐾, 𝐾].
Since 𝐾 was arbitrary, 𝑓 ′ (𝑥) = 𝑔(𝑥) for 𝑥 ∈ (−1, 1). That is,
11.11 ABEL’S TEST

Let (i) the series of functions ∑∞
𝑛=1 𝑓𝑛 (𝑥 ) be uniformly convergent
on [𝑎, 𝑏]
And (ii) The sequence of functions (𝑔𝑛 (𝑥)) be monotonic for every 𝑥 ∈
[𝑎, 𝑏] and uniformly bounded on [𝑎, 𝑏].
Then the series ∑∞

𝑛=1 𝑓𝑛 (𝑥 )𝑔𝑛 (𝑥 ) is uniformly convergent on [𝑎, 𝑏].
(−1)𝑛−1𝑥 𝑛
Example: Prove that the series ∑∞
𝑛=1 is uniformly convergent
𝑛(1+𝑥 𝑛 )
on [0, 1].
𝑥𝑛
Sol. Let 𝑔𝑛 (𝑥 ) = (1+𝑥 𝑛) , 𝑥 ∈ [0, 1]
𝑥 𝑛(𝑥−1)
Then 𝑔𝑛+1 − 𝑔𝑛 = ≤ 0 for all 𝑥 ∈ [0, 1].
(1+𝑥 𝑛)(1+𝑥 𝑛+1 )
For each 𝑥 ∈ [0, 1], the sequence (𝑔𝑛 ) is monotonic and for all 𝑥 ∈
[0, 1], |𝑔𝑛 | < 1 for all 𝑛 ∈ ℕ.
(−1)𝑛−1
The series ∑∞
𝑛=1 is convergent series of real numbers and
𝑛
therefore it is uniformly convergent on [0, 1].
(−1)𝑛−1 𝑥 𝑛
By Abel’s test, the series ∑∞
𝑛=1 is uniformly convergent on
𝑛(1+𝑥 𝑛 )
[0, 1].
11.12 DIRICHLET’S TEST
Let (i) the sequence of partial sums (𝑠𝑛 ) of the series of functions
∑∞
𝑛=1 𝑓𝑛 (𝑥 ) be uniformly bounded on [𝑎, 𝑏].
(ii) The sequence of functions (𝑔𝑛 (𝑥 )) be monotonic for every 𝑥 ∈
[𝑎, 𝑏] And
(iii) The sequence of functions (𝑔𝑛 (𝑥 )) is uniformly convergent
to 0 on [𝑎, 𝑏].
Then Then the series ∑∞

𝑛=1 𝑓𝑛 (𝑥 )𝑔𝑛 (𝑥 ) is uniformly convergent on
[𝑎, 𝑏].
(−1)𝑛 (𝑥 2+𝑛)
Example: Prove that the series ∑∞
𝑛=1 is uniformly
𝑛2
convergent in any closed and bounded interval [0, 1].
(𝑥 2+𝑛)
Sol. Let 𝑓𝑛 (𝑥) = (−1)𝑛 , 𝑔𝑛 (𝑥) = , 𝑥 ∈ [𝑎, 𝑏] .
𝑛2
Let 𝑠𝑛 = 𝑓1 + 𝑓2 + ⋯ + 𝑓𝑛 . Then the sequence (𝑠𝑛 ) is bounded.
𝑥 2+𝑛+1 𝑥 2 +𝑛
Then 𝑔𝑛+1 − 𝑔𝑛 = (𝑛+1)2
− < 0 for all 𝑥 ∈ [𝑎, 𝑏].
𝑛2
This shows that (𝑔𝑛 ) is a monotone decreasing sequence for each 𝑥 in

[𝑎, 𝑏].
lim 𝑔𝑛 (𝑥) = 0 each 𝑥 in [𝑎, 𝑏].

𝑛⟶∞
Thus, the sequence of functions (𝑔𝑛 ) is such that each 𝑔𝑛 is continuous

on [𝑎, 𝑏], the sequence converges to a continuous function on [𝑎, 𝑏] and
(𝑔𝑛 ) is monotone decreasing sequence on [𝑎, 𝑏].
(−1)𝑛 (𝑥 2+𝑛)
Therefore, by Dirichlet’s test ∑∞
𝑛=1 is uniformly convergent.
𝑛2
CHECK YOUR PROGRESS
Problem 1. Every pointwise convergent is uniform convergent.

Problem 2. Every uniform convergent is pointwise convergent.
𝑥𝑛
Problem 3. The series ∑∞
𝑛=1 , 0 ≤ 𝑥 ≤ 1 is pointwise convergent
𝑛
to 0.
𝑥𝑛
Problem 4. The series ∑∞
𝑛=1 , 0 ≤ 𝑥 ≤ 1 is uniformly convergent
𝑛
to 0.
11.13 SUMMARY
1. A sequence of functions ( 𝑓𝑛 ) defined on a set 𝐴 ⊆ R converges

uniformly on 𝐴 if and only if for every 𝜖 > 0 there exists an 𝑁 ∈ ℕ such
that | 𝑓𝑛 (𝑥) − 𝑓𝑚 (𝑥)| < 𝜖 whenever 𝑚, 𝑛 ≥ 𝑁 and 𝑥 ∈ 𝐴.
2. Let (i) the series of functions ∑∞

𝑛=1 𝑓𝑛 (𝑥 ) be uniformly convergent
on [𝑎, 𝑏]
And (ii) The sequence of functions (𝑔𝑛 (𝑥)) be monotonic for every 𝑥 ∈
[𝑎, 𝑏] and uniformly bounded on [𝑎, 𝑏].
Then the series ∑∞

𝑛=1 𝑓𝑛 (𝑥 )𝑔𝑛 (𝑥 ) is uniformly convergent on [𝑎, 𝑏].
11.14 GLOSSARY
sequence
series
11.15 REFERENCES


Q 1. Prove that the 𝑥 = 0 is a point of non – uniform convergence of the

𝑛𝑥
sequence of functions 〈𝑓𝑛 〉 Where, 𝑓𝑛 (𝑥 ) = 1+𝑛2𝑥 2 .
𝑛2 𝑥
Q 2. Prove that the sequence of functions 〈𝑓𝑛 〉 Where, 𝑓𝑛 (𝑥 ) = is
1+𝑛 2 𝑥 2
non-uniformly convergent on [0, 1].
𝑛𝑥
Q 3. Prove that the sequence of functions 〈𝑓𝑛 〉 Where, 𝑓𝑛 (𝑥 ) = 1+𝑛2 𝑥 2 is
uniformly convergent on [a, b], a > 0 but is only pointwise convergent on

[0, b].
11.18 ANSWERS
CHECK YOUR PROGRESS
CYQ 1. False
CYQ 2. True
CYQ 3. True
CYQ 4. True
UNIT 12: IMPROPER INTEGRAL I

Contents
12.1 Introduction
12.2 Objectives
12.3 Proper integral
12.4 Improper integral
12.5 Improper integral as limit of proper integral
12.6 Summary
12.7 Glossary
12.9 References
12.11 Answers
12.1 INTRODUCTION
In mathematical analysis, an improper integral is an extension
of the notion of a definite integral to cases that violate the usual
assumptions for that kind of integral. In the context of Riemann
integrals (or, equivalently, Darboux integrals), this typically involves
unboundedness, either of the set over which the integral is taken or of the
integrand (the function being integrated), or both. It may also involve
bounded but not closed sets or bounded but not continuous functions.
While an improper integral is typically written symbolically just like a
standard definite integral, it actually represents a limit of a definite
integral or a sum of such limits; thus improper integrals are said to
converge or diverge. If a regular definite integral (which
may metonymically be called a proper integral) is worked out as if it is
improper, the same answer will result. The concept of Riemann integrals
as developed in previous chapter requires that the range of integration is
finite and the integrand remains bounded on that domain. if either (or both)
of these assumptions is not satisfied it is necessary to attach a new
interpretation to the integral.
12.2 OBJECTIVES
 Proper integral
 Type of improper integral
12.3 PROPER INTEGRAL

𝑏
The definite integral ∫𝑎 𝑓(𝑥)𝑑𝑥 is called a proper integral if
(i) the interval of integration [𝑎, 𝑏] is finite or bounded.
(ii) the integrand 𝑓 is bounded on [𝑎, 𝑏].

𝑏
If F(x) is an indefinite integral of 𝑓 (𝑥), then ∫𝑎 𝑓 (𝑥)𝑑𝑥 = 𝐹 (𝑏) − 𝐹(𝑎).
12.4 IMPROPER INTEGRAL

𝑏
The definite integral ∫𝑎 𝑓 (𝑥 )𝑑𝑥 is called a improper integral if either or
𝑏
both the above conditions are not satisfied. Thus ∫𝑎 𝑓 (𝑥 )𝑑𝑥 is an
improper integral if either the interval of integration [𝑎, 𝑏] is not finite or
𝑓 is not bounded on [𝑎, 𝑏] or neither the interval [𝑎, 𝑏] is finite nor 𝑓 is
bounded over it.
𝑏
(i) In the definite integral ∫𝑎 𝑓(𝑥 )𝑑𝑥, if either 𝑎 or 𝑏 or both 𝑎 and 𝑏 are
infinite so that the interval of integration is unbounded but 𝑓 is bounded,
𝑏
then ∫𝑎 𝑓 (𝑥 )𝑑𝑥 is called an improper integral of the first kind.
∞ 𝑑𝑥 0
For example: ∫1 , ∫−∞ 𝑒 2𝑥 𝑑𝑥 are improper integral of the first kind.
√𝑥
𝑏
(ii) In the definite integral ∫𝑎 𝑓(𝑥 )𝑑𝑥, if both 𝑎 and 𝑏 are finite so that
the interval of integration is finite but 𝑓 has one or more point of infinite
𝑏
discontinuity i.e. 𝑓 is not bounded on [𝑎, 𝑏], then ∫𝑎 𝑓(𝑥 )𝑑𝑥 is called an
improper integral of the second kind.
2 1 2 1
For example: ∫1 𝑑𝑥 , ∫1 𝑑𝑥 are improper integral of second kind.
𝑥2 2−𝑥
𝑏
(iii) In the definite integral ∫𝑎 𝑓(𝑥 )𝑑𝑥, if the interval of integration is
𝑏
unbounded and 𝑓 is also unbounded, ∫𝑎 𝑓(𝑥 )𝑑𝑥 is called an improper
integral of the third kind.
∞ 𝑒 −𝑥
For example: ∫0 𝑑𝑥 is an improper integral of third kind.
√𝑥
12.5 IMPROPER ITEGRAL AS THE LIMIT OF

A PROPER INTEGRAL
(a) When the improper integral is of the first kind, either 𝑎 or 𝑏 or both 𝑎
and 𝑏 are infinite but 𝑓 is bounded. We define
∞ 𝑡
(i) ∫𝑎 𝑓 (𝑥 )𝑑𝑥 = lim ∫𝑎 𝑓 (𝑥 )𝑑𝑥, (𝑡 > 𝑎)
𝑡⟶∞
∞
The improper integral ∫𝑎 𝑓(𝑥 )𝑑𝑥 is said to be convergent if the limit
of right-hand side exists finitely and the integral is said to be divergent
if the limit is +∞ or −∞.
If the integral is neither convergent or divergent, then it is said to be

oscillating.
𝑏 𝑎
(ii) ∫−∞ 𝑓 (𝑥 )𝑑𝑥 = lim ∫𝑡 𝑓(𝑥 )𝑑𝑥, (𝑡 < 𝑏)
𝑡⟶−∞
𝑏
The improper integral ∫−∞ 𝑓(𝑥 )𝑑𝑥 is said to be convergent if the limit
∞ 𝑐 ∞
(iii) ∫−∞ 𝑓 (𝑥 )𝑑𝑥 =∫−∞ 𝑓 (𝑥 )𝑑𝑥 + ∫𝑐 𝑓 (𝑥 )𝑑𝑥 where c is any real number
𝑐 𝑡
= lim ∫𝑡 𝑓(𝑥 )𝑑𝑥 + lim ∫𝑐 2 𝑓(𝑥 )𝑑𝑥
𝑡1 ⟶−∞1 𝑡2 ⟶∞
∞
The improper integral ∫−∞ 𝑓(𝑥 )𝑑𝑥 is said to be convergent if both the
limits on the right-hand side exist finitely and independent of each other,
otherwise it is said to be divergent.
∞ 𝑐 𝑡
Note: ∫−∞ 𝑓(𝑥 )𝑑𝑥 ≠ lim [ ∫−𝑡 𝑓(𝑥 )𝑑𝑥 + ∫𝑐 𝑓 (𝑥)𝑑𝑥].
𝑡⟶∞
(b) When the improper integral is second kind, both 𝑎 and 𝑏 are finite
but 𝑓 has one points of infinite discontinuity on [𝑎, 𝑏].
(i) If 𝑓(𝑥) becomes infinite at 𝑥 = 𝑏 only, we define

𝑏 𝑏−𝜖
∫𝑎 𝑓(𝑥)𝑑𝑥 = lim ∫𝑎 𝑓 (𝑥 )𝑑𝑥.
𝜖⟶0+
𝑏
The improper integral ∫𝑎 𝑓(𝑥 )𝑑𝑥 is said to be convergent if the limit on
the right- hand side exists finitely and the interval is said to be divergent
(ii) If 𝑓(𝑥) becomes infinite at 𝑥 = 𝑎 only, we define

𝑏 𝑐
∫𝑎 𝑓 (𝑥)𝑑𝑥 = lim ∫𝑎+𝜖 𝑓 (𝑥 )𝑑𝑥.
𝜖⟶0+
𝑏
The improper integral ∫𝑎 𝑓(𝑥 )𝑑𝑥 is said to be convergent if the limit on
the right- hand side exists finitely, otherwise it is said to be divergent.
(iii) If 𝑓(𝑥) becomes infinite at 𝑥 = 𝑐 only where 𝑎 < 𝑐 < 𝑏, we define

𝑏 𝑐 𝑏
∫𝑎 𝑓 (𝑥)𝑑𝑥 = ∫𝑎 𝑓 (𝑥 )𝑑𝑥 + ∫𝑐 𝑓 (𝑥 )𝑑𝑥
𝑐−𝜖1 𝑏
= lim ∫𝑎 𝑓(𝑥)𝑑𝑥 + lim ∫𝑐+𝜖 𝑓(𝑥 )𝑑𝑥.
𝜖1 ⟶0+ 𝜖2 ⟶0+ 2
𝑏
The improper integral ∫𝑎 𝑓(𝑥 )𝑑𝑥 is said to be convergent if both the
limits on the right- hand side exist finitely and independent of each
other, otherwise it is said to be divergent.
Note: (1) if 𝑓 (𝑥 ) has finite discontinuity at the end point of the interval
of integration, then the point of discontinuity is approached from within
the interval.
Thus, if the interval of integration is [𝑎, 𝑏] and
(i) 𝑓 has infinite discontinuity at ′𝑎′, we consider [𝑎 + 𝜖, 𝑏] as 𝜖 ⟶ 0 +.
(ii) 𝑓 has infinite discontinuity at ′𝑏′, we consider [𝑎, 𝑏 − 𝜖] as 𝜖 ⟶ 0 +.
Note: (2) A proper integral is always convergent.

𝑏
Note: (3) If ∫𝑎 𝑓 (𝑥 )𝑑𝑥 is convergent, then
𝑏
(i) ∫𝑎 𝑘𝑓(𝑥 )𝑑𝑥 is convergent, 𝑘 ∈ ℝ,
𝑏 𝑐 𝑏
(ii) ∫𝑎 𝑓 (𝑥 )𝑑𝑥 = ∫𝑎 𝑓 (𝑥 )𝑑𝑥 + ∫𝑐 𝑓 (𝑥 )𝑑𝑥 where 𝑎 < 𝑐 < 𝑏 and each
interval or right-hand side is convergent.
Note: (4) For any 𝑐 between 𝑎 and 𝑏, 𝑖. 𝑒. 𝑎 < 𝑐 < 𝑏, we have

𝑏 𝑐 𝑏
∫𝑎 𝑓(𝑥)𝑑𝑥 = ∫𝑎 𝑓 (𝑥 )𝑑𝑥 + ∫𝑐 𝑓(𝑥 )𝑑𝑥
𝑏 𝑏
If ∫𝑐 𝑓 (𝑥 )𝑑𝑥 is a proper integral, then the two integrals ∫𝑎 𝑓(𝑥 )𝑑𝑥 and
𝑐
∫𝑎 𝑓 (𝑥 )𝑑𝑥 converges or diverge together.
𝑏
Thus, while testing the interval ∫𝑎 𝑓 (𝑥 )𝑑𝑥 convergence at 𝑎 it may be
𝑐
replaced by ∫𝑎 𝑓(𝑥 )𝑑𝑥 for any convenient 𝑐 such that 𝑎 < 𝑐 < 𝑏.
ILLUSTRATIVE EXAMPLES
Example 1. Examine the convergence of the following improper

integrals:
∞1 ∞ 1 ∞ 1 ∞ 1
(i) ∫0 𝑑𝑥 (ii) ∫1 𝑑𝑥 (iii) ∫1 𝑑𝑥 (iv) ∫0 𝑑𝑥
𝑥 √𝑥 𝑥 3/2 1+𝑥 2
∞1 𝑡1
Sol. (i) By definition, ∫0 𝑑𝑥 = lim ∫0 𝑥 𝑑𝑥
𝑥 𝑡⟶∞
= lim [𝑙𝑜𝑔𝑥]1𝑡 = lim 𝑙𝑜𝑔𝑡 = ∞

𝑡⟶∞ 𝑡⟶∞
∞1
Therefore, ∫0 𝑑𝑥 is divergent.
𝑥
∞ 1 𝑡 1 𝑡
(ii) By definition, ∫1 𝑑𝑥 = lim ∫1 𝑑𝑥 = lim ∫1 𝑥 −1/2 𝑑𝑥
√𝑥 𝑡⟶∞ √𝑥 𝑡⟶∞
= lim [2√𝑥]1𝑡 = lim (2√𝑡 − 2) = ∞

𝑡⟶∞ 𝑡⟶∞
∞ 1
Therefore, ∫0 𝑑𝑥 is divergent.
√𝑥
∞ 1 𝑡
(iii) By definition, ∫1 𝑑𝑥 = lim ∫1 𝑥 −3/2 𝑑𝑥
𝑥 3/2 𝑡⟶∞
𝑡
𝑥 −1/2 −2 𝑡
= lim [ −1/2 ] = lim [ 𝑥 ]
𝑡⟶∞ 1 𝑡⟶∞ √ 1
−2
= lim ( + 2) = 0 + 2 = 2, which is finite.
𝑡⟶∞ √𝑡
∞ 1
Therefore, ∫1 𝑑𝑥 is convergent and its value is 2.
𝑥 3/2
∞ 1 𝑡 1
(iv) By definition, ∫0 𝑑𝑥 = lim ∫0 𝑑𝑥
1+𝑥 2 𝑡⟶∞ 1+𝑥 2
= lim (tan−1 𝑡)𝑡0

𝑡⟶∞
= lim (tan−1 𝑡 − tan−1 0)

𝑡⟶∞
𝜋
= 2 which is finite.
∞ 1 𝜋
Therefore, ∫0 𝑑𝑥 is convergent and its value is 2 .
1+𝑥 2
Example 2. Examine the convergence of the following improper

integrals:
∞ ∞ 𝑥 ∞ 1
(i) ∫0 𝑒 −𝑚𝑥 𝑑𝑥 (𝑚 > 0) (ii) ∫𝑎 𝑑𝑥 (iii) ∫0 𝑑𝑥
1+𝑥 2 (1+𝑥)3
∞
(iv) ∫0 𝑠𝑖𝑛𝑥 𝑑𝑥
∞ 𝑡
Sol. (i) By definition, ∫0 𝑒 −𝑚𝑥 𝑑𝑥 = lim ∫0 𝑒 −𝑚𝑥 𝑑𝑥
𝑡⟶∞
1
= lim − 𝑚 (𝑒 −𝑚𝑡 − 1)
𝑡⟶∞
1 1
= − 𝑚 (0 − 1) = 𝑚 which is finite.
∞ 1
⟹ ∫0 𝑒 −𝑚𝑥 𝑑𝑥 is convergent and its value is 𝑚.
∞ 𝑥 𝑡 𝑥
(ii) By definition, ∫𝑎 𝑑𝑥 = lim ∫0 1+𝑥 2 𝑑𝑥
1+𝑥 2 𝑡⟶∞
𝑡1 2𝑥
= lim ∫0 2 (1+𝑥 2 )𝑑𝑥
𝑡⟶∞
1 𝑡
= lim (2 log(1 + 𝑥 2 ))
𝑡⟶∞ 𝑎
1
= lim 2 [log(1 + 𝑡 2 ) − log(1 + 𝑎2 )] = ∞
𝑡⟶∞
∞ 𝑥
Therefore, ∫𝑎 𝑑𝑥 is divergent.
1+𝑥 2
∞ 1 𝑡
(iii) By definition, ∫0 𝑑𝑥 = lim ∫0 (1 + 𝑥)−3 𝑑𝑥
(1+𝑥)3 𝑡⟶∞
𝑡
(1+𝑥)−2
= lim [ ]
𝑡⟶∞ −2 0
1 1 1 1
= lim − 2 [(1+𝑡)2 − 1] = − 2 (0 − 1) = 2 which is finite.
𝑡⟶∞
∞ 1 1
(1+𝑥)3
∞ 𝑡
(iv) By definition, ∫0 𝑠𝑖𝑛𝑥 𝑑𝑥 = lim ∫0 𝑠𝑖𝑛𝑥 𝑑𝑥 = lim [−𝑐𝑜𝑠 𝑥 ]𝑡0
𝑡⟶∞ 𝑡⟶∞
= lim (1 − 𝑐𝑜𝑠𝑡)
𝑡⟶∞
Which does not exist uniquely since cost oscillates between -1 and +1
when 𝑡 ⟶ ∞.
∞
Therefore, ∫0 𝑠𝑖𝑛𝑥 𝑑𝑥 oscillates.
Example 3. Examine for convergence the integrals:

∞ ∞
(i) ∫1 𝑥𝑒 −𝑥 𝑑𝑥 (ii) ∫0 𝑥 2 𝑒 −𝑥 𝑑𝑥
∞ 𝑡
Sol. (i) ∫1 𝑥𝑒 −𝑥 𝑑𝑥 = lim ∫1 𝑥𝑒 −𝑥 𝑑𝑥
𝑡⟶∞
= lim [−𝑥𝑒 −𝑥 − 𝑒 −𝑥 ]1𝑡

𝑡⟶∞
= lim [−𝑡𝑒 −𝑡 − 𝑒 −𝑡 + 𝑒 −1 + 𝑒 −1 ]
𝑡⟶∞
−𝑡 2
= lim ( 𝑒 𝑡 ) − lim (𝑒 −𝑡 ) + 𝑒 (applying L’ Hospital rule to first limit)
𝑡⟶∞ 𝑡⟶∞
−1 2 2 2
= lim ( 𝑒 𝑡 ) − 0 + 𝑒 = 0 + 𝑒 = 𝑒 which is finite.
𝑡⟶∞
∞ 2
Therefore, ∫1 𝑥𝑒 −𝑥 𝑑𝑥 is convergent and its value is 𝑒.
∞ 𝑡
(ii) ∫0 𝑥 2 𝑒 −𝑥 𝑑𝑥 = lim ∫0 𝑥 2 𝑒 −𝑥 𝑑𝑥
𝑡⟶∞
= lim [−𝑥 2 𝑒 −𝑥 − 2𝑥𝑒 −𝑥 − 2𝑒 −𝑥 ]𝑡0

𝑡⟶∞
= lim [−𝑡 2 𝑒 −𝑡 − 2𝑡𝑒 −𝑡 − 2𝑒 −1 + 2]

𝑡⟶∞
−𝑡 2 𝑡
= lim ( ) − 2 lim ( 𝑡 ) − 0 + 2 (Applying L’ Hospital rule)
𝑡⟶∞ 𝑒𝑡 𝑡⟶∞ 𝑒
−2𝑡 𝑡
= lim ( 𝑒 𝑡 ) − 2 lim (𝑒 𝑡 ) + 2
𝑡⟶∞ 𝑡⟶∞
(Again, applying L’ Hospital rule)

−2
= lim ( 𝑒 𝑡 ) −2× 0 + 2 = 0 + 2 = 2 which is finite.
𝑡⟶∞
∞
Therefore, ∫0 𝑥 2 𝑒 −𝑥 𝑑𝑥 is convergent and its value is 2.
Example 4. Examine for convergence of the integrals:

0 0 1
(i) ∫−∞ 𝑒 2𝑥 𝑑𝑥 (ii) ∫−∞ 𝑝2 +𝑞2 𝑥 2 𝑑𝑥
0 0
Sol. (i) ∫−∞ 𝑒 2𝑥 𝑑𝑥 = lim ∫𝑡 𝑒 2𝑥 𝑑𝑥
𝑡⟶−∞
0
𝑒 2𝑥 1 1 1
= lim [ ] = lim (1 − 𝑒 2𝑡 ) = 2 (1 − 0) = 2 which is finite.
𝑡⟶−∞ 2 𝑡 𝑡⟶−∞ 2
0 1
Therefore, ∫−∞ 𝑒 2𝑥 𝑑𝑥 is convergent and its value is 2.
0 1 0 1
(ii) ∫−∞ 𝑝2 +𝑞2 𝑥 2 𝑑𝑥 = lim ∫𝑡 𝑑𝑥
𝑡⟶−∞ 𝑝2 +𝑞2 𝑥 2
0 1
= lim ∫𝑡 𝑝2
𝑑𝑥
𝑡⟶−∞ 𝑞2 ( 2 +𝑥 2 )
𝑞
1 1 𝑥 0
= lim [𝑞2 . 𝑝/𝑞 tan−1 𝑝/𝑞]
𝑡⟶−∞ 𝑡
1 𝑞𝑡 1 𝜋 𝜋
= lim [0 − tan−1 ] = − (− 2 ) = 2𝑝𝑞 which is finite.
𝑡⟶−∞ 𝑝𝑞 𝑝 𝑝𝑞
0 1 𝜋
Therefore, ∫−∞ 𝑝2 +𝑞2 𝑥 2 𝑑𝑥 is convergent and its value is 2𝑝𝑞 .

∞ ∞ 1
(i) ∫−∞ 𝑒 −𝑥 𝑑𝑥 (ii) ∫−∞ 1+𝑥 2 𝑑𝑥
∞ 0 ∞
Sol. (i) ∫−∞ 𝑒 −𝑥 𝑑𝑥 = ∫−∞ 𝑒 −𝑥 𝑑𝑥 + ∫0 𝑒 −𝑥 𝑑𝑥
0 𝑡
= lim ∫𝑡 𝑒 −𝑥 𝑑𝑥 + lim ∫0 2 𝑒 −𝑥 𝑑𝑥
𝑡1 ⟶−∞
1 𝑡2 ⟶∞
𝑡
= lim [−𝑒 −𝑥 ]0𝑡1 + lim [−𝑒 −𝑥 ]02
𝑡1 ⟶−∞ 𝑡2 ⟶−∞
= lim (−1 + 𝑒 −𝑡1 ) + lim (−𝑒 −𝑡2 + 1)

𝑡1 ⟶−∞ 𝑡2 ⟶−∞
= (-1 + ∞) + (0 + 1) = ∞
∞
Therefore, ∫−∞ 𝑒 −𝑥 𝑑𝑥 is divergent to ∞.
∞ 1 0 1 ∞ 1
(ii) ∫−∞ 1+𝑥 2 𝑑𝑥 = ∫−∞ 1+𝑥 2 𝑑𝑥 + ∫0 𝑑𝑥
1+𝑥 2
0 1 𝑡 1
= lim ∫𝑡 𝑑𝑥 + lim ∫0 2 𝑑𝑥
1 1+𝑥 2
𝑡1 ⟶−∞ 1+𝑥 2 𝑡2 ⟶∞
𝑡
= lim [tan−1 𝑥 ]0𝑡1 + lim [tan−1 𝑥 ]02
𝑡1 ⟶−∞ 𝑡2 ⟶−∞
= lim [−tan−1 𝑡1 ] + lim [−tan−1 𝑡2 ]

𝑡1 ⟶−∞ 𝑡2 ⟶∞
𝜋 𝜋
= − (− 2 ) + 2 = 𝜋 which is finite.
∞ 1
Therefore, ∫−∞ 1+𝑥 2 𝑑𝑥 is convergent and its value is 𝜋.

1 1/𝑒 1
(i) ∫0 𝑙𝑜𝑔𝑥 𝑑𝑥 (ii) ∫0 𝑑𝑥
𝑥(𝑙𝑜𝑔𝑥)2
Sol. (i) 0 is only point of infinite discontinuity of the integrand on [𝑎, 𝑏].
1 1
Therefore, ∫0 𝑙𝑜𝑔𝑥 𝑑𝑥 = lim ∫0+𝜖(𝑙𝑜𝑔𝑥 ). 1𝑑𝑥
𝜖⟶0+
Integration by parts
= lim [𝑥 𝑙𝑜𝑔𝑥 − 𝑥 ]1𝜖 = lim (−1 − 𝜖 log 𝜖 + 𝜖)

𝜖⟶0+ 𝜖⟶0+
= −1 which is finite. [𝑠𝑖𝑛𝑐𝑒 lim 𝑥 𝑛 𝑙𝑜𝑔𝑥 = 0, 𝑛 > 0 ]

𝜖⟶0
1
Therefore, ∫0 𝑙𝑜𝑔𝑥 𝑑𝑥 is convergent and its value is −1.
(ii) since lim 𝑥(𝑙𝑜𝑔𝑥)𝑛 = 0, 𝑛 > 0, therefore, 0 is the only point of

𝑥⟶0
1
infinite discontinuity of the integrand on [0, 𝑒 ].
1/𝑒 1 1/𝑒 1
Therefore, ∫0 𝑑𝑥 = lim ∫0+𝜖 (𝑙𝑜𝑔𝑥 )−2 𝑥 𝑑𝑥
𝑥(𝑙𝑜𝑔𝑥)2 𝜖⟶0+
1/𝑒
(𝑙𝑜𝑔𝑥)−1
= lim [ ]
𝜖⟶0+ −1 𝜖
1 1
= lim [ 1 − log 𝜖 ] = −[−1 − 0] = 1
𝜖⟶0+ log𝑒
which is finite.
1/𝑒 1
𝑥(𝑙𝑜𝑔𝑥)2
CHECK YOUR PROGRESS
Problem 1. Every proper integral is always convergent.

1 1
Problem 2. ∫0 𝑑𝑥 is convergent.
𝑥2
1 1
Problem 3. ∫0 𝑥 2 𝑑𝑥 is divergent to +∞.
𝑒 1 1
Problem 4. ∫0 𝑥(𝑙𝑜𝑔𝑥)3 𝑑𝑥 is convergent to − 2.
2 1
Problem 5. ∫1 𝑥(𝑙𝑜𝑔𝑥)3 𝑑𝑥 is divergent to +∞.
12.6 SUMMARY
𝑏
1. In the definite integral ∫𝑎 𝑓 (𝑥 )𝑑𝑥, if either 𝑎 or 𝑏 or both 𝑎 and 𝑏 are
infinite so that the interval of integration is unbounded but 𝑓 is bounded,
𝑏
then ∫𝑎 𝑓 (𝑥 )𝑑𝑥 is called an improper integral of the first kind.
𝑏
2. In the definite integral ∫𝑎 𝑓 (𝑥 )𝑑𝑥, if both 𝑎 and 𝑏 are finite so that the
interval of integration is finite but 𝑓 has one or more point of infinite
𝑏
discontinuity i.e. 𝑓 is not bounded on [𝑎, 𝑏], then ∫𝑎 𝑓(𝑥 )𝑑𝑥 is called an
improper integral of the second kind.
𝑏
3. In the definite integral ∫𝑎 𝑓 (𝑥 )𝑑𝑥, if the interval of integration is
𝑏
unbounded and 𝑓 is also unbounded, ∫𝑎 𝑓(𝑥 )𝑑𝑥 is called an improper
integral of the third kind.
∞ 𝑡
4. (i) ∫𝑎 𝑓(𝑥 )𝑑𝑥 = lim ∫𝑎 𝑓 (𝑥 )𝑑𝑥, (𝑡 > 𝑎)
𝑡⟶∞
∞
The improper integral ∫𝑎 𝑓(𝑥 )𝑑𝑥 is said to be convergent if the limit
If the integral is neither convergent or divergent, then it is said to be

oscillating.
𝑏 𝑎
(ii) ∫−∞ 𝑓 (𝑥 )𝑑𝑥 = lim ∫𝑡 𝑓(𝑥 )𝑑𝑥, (𝑡 < 𝑏)
𝑡⟶−∞
𝑏
The improper integral ∫−∞ 𝑓(𝑥 )𝑑𝑥 is said to be convergent if the limit
∞ 𝑐 ∞
(iii) ∫−∞ 𝑓 (𝑥 )𝑑𝑥 =∫−∞ 𝑓 (𝑥 )𝑑𝑥 + ∫𝑐 𝑓 (𝑥 )𝑑𝑥 where c is any real number
𝑐 𝑡
= lim ∫𝑡 𝑓(𝑥 )𝑑𝑥 + lim ∫𝑐 2 𝑓(𝑥 )𝑑𝑥
𝑡1 ⟶−∞ 1 𝑡2 ⟶∞
∞
The improper integral ∫−∞ 𝑓(𝑥 )𝑑𝑥 is said to be convergent if both the
limits on the right-hand side exist finitely and independent of each other,
otherwise it is said to be divergent.
5. For any 𝑐 between 𝑎 and 𝑏, 𝑖. 𝑒. 𝑎 < 𝑐 < 𝑏, we have

𝑏 𝑐 𝑏
∫𝑎 𝑓(𝑥)𝑑𝑥 = ∫𝑎 𝑓 (𝑥 )𝑑𝑥 + ∫𝑐 𝑓(𝑥 )𝑑𝑥
𝑏 𝑏
If ∫𝑐 𝑓 (𝑥 )𝑑𝑥 is a proper integral, then the two integrals ∫𝑎 𝑓(𝑥 )𝑑𝑥 and
𝑐
∫𝑎 𝑓 (𝑥 )𝑑𝑥 converges or diverge together.
𝑏
Thus, while testing the interval ∫𝑎 𝑓 (𝑥 )𝑑𝑥 convergence at 𝑎 it may be
𝑐
replaced by ∫𝑎 𝑓(𝑥 )𝑑𝑥 for any convenient 𝑐 such that 𝑎 < 𝑐 < 𝑏.
12.7 GLOSSARY
sequence
series
12.8 REFERENCES



∞ 1
Q 1. Examine for convergence of the integral ∫−∞ 𝑒 𝑥 + 𝑒 −𝑥 𝑑𝑥 .
∞ 1
Q 2. Examine for convergence of the integral ∫−∞ 𝑥 2 + 2𝑥+2 𝑑𝑥 .
1 1
Q 3. Examine for convergence of the integral ∫0 𝑑𝑥 .
𝑥 2−3𝑥+2
𝜋 1
𝑠𝑖𝑛𝑥
𝜋 1
Q 5. Examine for convergence of the integral ∫0 𝑑𝑥.
1+𝑐𝑜𝑠𝑥
12.11 ANSWERS
𝜋
TQ1. Convergent to 2 .
TQ2. Convergent to 𝜋.
TQ3. Divergent to ∞.
CHECK YOUR PROGRESS
CYQ 1. True
CYQ 2. False
CYQ 3. True
CYQ 4. True
CYQ 5. True
UNIT 13: IMPROPER INTEGRAL II

Contents
13.1 Introduction
13.2 Objectives
b
13.3 Tests for convergence of ∫a f(𝑥 )d𝑥 at ‘a’
13.4 Comparison test I
13.5 Comparison test II (Limit form)
13.6 General test for convergence (Integrand may change sign)
13.7 Absolute convergence
13.8 convergence at ∞
13.9 Summary
13.10 Glossary
13.12 References
13.14 Answers
13.1 INTRODUCTION
In mathematical analysis, an improper integral is an extension

of the notion of a definite integral to cases that violate the usual
assumptions for that kind of integral. In this unit we discussed
b
tests for convergence of ∫a f(𝑥 )d𝑥 at ‘𝑎’, general test for convergence and
absolute convergence of some functions, also tests for convergence of
b
∫a f(𝑥 )d𝑥 at ‘∞’.
13.2 OBJECTIVES
 Test of convergence
 Absolute integral
𝒃
13.3 TEST FOR CONVERGENCE OF ∫𝒂 𝒇(𝒙)𝒅𝒙
AT 𝒂
Let a be the only point of infinite discontinuity of 𝑓(𝑥) on [𝑎, 𝑏]. The
case when 𝑏 is the only point of infinite discontinuity can be dealt with
in the same way.
Without any loss of generality, we assume that 𝑓(𝑥) is positive (or non-
negative) on [𝑎, 𝑏].
In case 𝑓(𝑥) is negative, we can replace it by (−𝑓) for testing the

b
convergence of ∫a f(𝑥 )d𝑥.
Theorem: A necessary and sufficient condition for the convergence of

b
the improper integral ∫𝑎 f(𝑥)d𝑥 at ‘𝑎′ where f is positive on (𝑎, 𝑏], is
that there exists a positive number M, independent of 𝜖 > 0 such that
b
∫𝑎+𝜖 f(𝑥 )d𝑥 < 𝑀 ∀ 𝜖 in (0, 𝑏 − 𝑎)
Proof: Since 𝑎 is the only point of infinite discontinuity of 𝑓 on [𝑎, 𝑏],

therefore, 𝑓 is continuous on (𝑎, 𝑏].
Also 𝑓 is positive on (𝑎, 𝑏].
⟹ For 𝑎 < 𝑎 + 𝜖 < 𝑏 i.e. for 0 < 𝜖 < 𝑏 − 𝑎, 𝑓 is positive and

continuous on [𝑎 + 𝜖, 𝑏].
b
⟹ ∫𝑎+𝜖 f(𝑥 )d𝑥 = 𝐴( 𝜖) represents the area bounded by 𝑓 on [𝑎 + 𝜖, 𝑏]
and 𝑥 −axis.
⟹ As 𝜖 ⟶ 0 +, i.e. as 𝜖 decrease, 𝐴(𝜖) increases since the length of the

interval increases.
b
⟹ lim 𝐴(𝜖) = lim ∫𝑎+𝜖 f(𝑥 )d𝑥 will exist finitely iff 𝐴(𝜖) is bounded
𝜖⟶0+ 𝜖⟶0+
above.
b
⟹ ∫𝑎 f(𝑥 )d𝑥 will converge iff ∃ a real number M > 0 and independent
of 𝜖 such that 𝐴(𝜖 ) < 𝑀
b b
⟹ ∫𝑎 f(𝑥 )d𝑥 converges iff ∫𝑎+𝜖 f(𝑥 )d𝑥 < 𝑀 ∀ 𝜖 in (0, b - 𝑎).
Note: If for every M > 0 and some 𝜖 in (0, b - 𝑎). 𝐴(𝜖 ) > 𝑀, then
b
∫𝑎+𝜖 f(𝑥 )d𝑥 is not bounded above.
b
Therefore, ∫𝑎+𝜖 f(𝑥 )d𝑥 tend to +∞ as 𝜖 tend to 0+ and hence, the
b
improper integral ∫𝑎 f(𝑥 )d𝑥 diverges to +∞.
13.4 COMPARISION TEST I

If f and 𝑔 are two positive functions with f(𝑥) ≤ 𝑔(𝑥) for all 𝑥 in (𝑎, 𝑏]
and 𝑎 is only point of infinite discontinuity on [𝑎, 𝑏], then
b b
(i) ∫𝑎 g(𝑥 )d𝑥 is convergent ⟹ ∫𝑎 f(𝑥 )d𝑥 is convergent
b b
(ii) ∫𝑎 f(𝑥)d𝑥 is divergent ⟹ ∫𝑎 g(𝑥)d𝑥 is divergent.
13.5 COMPARISION TEST II (LIMIT FORM)

If f and 𝑔 are two positive functions on (𝑎, 𝑏], a being the only point of
𝑓(𝑥)
infinity discontinuity, and lim = 1 where 𝑙 is non-zero finite
𝑥⟶𝑎+ 𝑔(𝑥)
b b
number, then two ∫𝑎 f(𝑥 )d𝑥 and ∫𝑎 g(𝑥 )d𝑥 converge or diverge
together.
Note: let f and 𝑔 be two positive functions on (𝑎, 𝑏], 𝑎 being the only
point of infinite discontinuity. Then
𝑓(𝑥) b b
(i) lim = 0 and ∫𝑎 g(𝑥 )d𝑥 converges ⟹ ∫𝑎 f(𝑥 )d𝑥 is converges
𝑓(𝑥) b b
(ii) lim = ∞ and ∫𝑎 g(𝑥 )d𝑥 diverges ⟹ ∫𝑎 f(𝑥 )d𝑥 is diverges.
Note:
b 1
(i) The improper integral ∫𝑎 (𝑥−𝑎)𝑛
d𝑥 is convergent
if and only if 𝑛 < 1.

b 1
(ii) The improper integral ∫𝑎 (𝑏−𝑥)𝑛
d𝑥 is convergent
if and only if 𝑛 < 1.
Note:
(i) if 𝑎 is the only point of infinite discontinuity of 𝑓 on [𝑎, 𝑏] and

b
lim (𝑥 − 𝑎)𝜇 𝑓(𝑥) exists and non-zero finite, then ∫𝑎 𝑓(𝑥)d𝑥
𝑥⟶𝑎+
converges if and only if 𝜇 < 1.
(ii) if 𝑏 is the only point of infinite discontinuity of 𝑓 on [𝑎, 𝑏] and

b
lim (𝑏 − 𝑥 )𝜇 𝑓(𝑥) exists and non-zero finite, then ∫𝑎 𝑓(𝑥)d𝑥
𝑥⟶𝑏+
 ILLUSTRATIVE EXAMPLES:
Example 1: Examine the convergence of the integrals.
1 1 2 1
(i) ∫0 d𝑥 (ii) ∫1 d𝑥
√𝑥 2 +𝑥 (1+𝑥)√2− 𝑥
1 1
Sol. (i) Here 𝑓 (𝑥 ) = =
√𝑥 2 + 𝑥 √ 𝑥 √𝑥 + 𝑥
0 is the point of infinite discontinuity of 𝑓 on [0, 1].

1 𝑓(𝑥) 1
Take 𝑔(𝑥 ) = , then lim = lim = 1 which is
√𝑥 𝑥⟶0+ 𝑔(𝑥) 𝑥⟶0+ √ 𝑥+1
non-zero and finite.

1 1
Therefore, By comparison test, ∫0 𝑓(𝑥)d𝑥 and ∫0 𝑔(𝑥)d𝑥 converge
or diverge together.
1 1 1
But ∫0 𝑔(𝑥)d𝑥 = ∫0 d𝑥
√𝑥
b 1
(From, ∫𝑎 (𝑥−𝑎)𝑛
d𝑥 with a = 0 converges.
since n = ½ < 1)
1 1 1
Therefore, ∫0 𝑓(𝑥)d𝑥 = ∫0 d𝑥 is convergent.
√ 𝑥+1
1
(ii) Here 𝑓 (𝑥 ) = (1+𝑥)
√2− 𝑥
2 is the point of infinite discontinuity of f on [1, 2].

1 𝑓(𝑥) 1 1
Take 𝑔(𝑥 ) = , then lim = lim = 3 which is
√2−𝑥 𝑥⟶2− 𝑔(𝑥) 𝑥⟶2− 𝑥+1

2 2
Therefore, by comparison test, ∫1 𝑓(𝑥)d𝑥 and ∫1 𝑔(𝑥)d𝑥 converge
2 2 1
√2−𝑥
b 1
(From, ∫𝑎 (𝑏−𝑥)𝑛
d𝑥 with b = 2 converges.)
since n = ½ < 1)
2 1 1
Therefore, ∫1 𝑓(𝑥)d𝑥 = ∫0 d𝑥 is convergent.
(1+𝑥)√2− 𝑥
Example 2: Examine the convergence of the integral.

1 1
∫0 𝑥 3( 2+ 𝑥 2)5
d𝑥
1
Sol. Here 𝑓 (𝑥) = 𝑥 3 ( 2+ 𝑥 2 )5

1 𝑓(𝑥) 1 1
Take 𝑔(𝑥 ) = 𝑥 3 , then lim = lim = 32 which is
𝑥⟶0+ 𝑔(𝑥) 𝑥⟶0+ ( 2+ 𝑥 2)5

1 1
Therefore, by comparison test, ∫0 𝑓(𝑥)d𝑥 and ∫0 𝑔(𝑥)d𝑥 converge
1 1 1
𝑥3
b 1
(From, ∫𝑎 (𝑥−𝑎)𝑛
d𝑥 with 𝑎 = 0 diverges.
since n = 3 > 1)
1 1 1
Therefore, ∫0 𝑓(𝑥)d𝑥 = ∫0 d𝑥 is divergent.
𝑥 3 ( 2+ 𝑥 2 )5

1 𝑙𝑜𝑔𝑥
∫0 d𝑥
1+𝑥
𝑙𝑜𝑔𝑥 𝑙𝑜𝑔𝑥
Sol. Since 1 + 𝑥 is negative on (0, 1], we take 𝑓(𝑥) = − 1 + 𝑥

1 𝑓(𝑥) 𝑥 𝑛 𝑙𝑜𝑔𝑥
Take 𝑔(𝑥 ) = 𝑥 𝑛 , then lim = lim − = 0 if 𝑛 > 0.
𝑥⟶0+ 𝑔(𝑥) 𝑥⟶0+ 1+𝑥
1
Taking n between o and 1, the integral ∫0 𝑔(𝑥)d𝑥 is convergent.
1
Therefore, by comparison test, ∫0 𝑓(𝑥)d𝑥 is convergent.

π/2 sin 𝑥
∫0 d𝑥
𝑥𝑝
Sol. If p is negative or zero, the given integral is a proper integral and

hence convergent when 𝑝 ≤ 0.
When p > 0, the only point of discontinuity is 0.

sin 𝑥
Let 𝑓 (𝑥 ) = 𝑥𝑝
1
Take 𝑓 (𝑥 ) = 𝑥 𝜇 then
𝑓(𝑥) 𝑠𝑖𝑛𝑥
lim = lim 𝑥 𝜇−𝑝 𝑠𝑖𝑛𝑥 = lim 𝑥 𝜇−𝑝+1 ( )
𝑥⟶0+ 𝑔(𝑥) 𝑥⟶0+ 𝑥⟶0+ 𝑥
= 1 if 𝜇 − 𝑝 + 1 = 0
= 0 if 𝜇 − 𝑝 + 1 > 0
= ∞ if 𝜇 − 𝑝 + 1 < 0
By taking 0 < 𝜇 < 1 and also 𝜇 = 𝑝 − 1 so that
0 < 𝑝 − 1 < 1 i.e. 1 < 𝑝 < 2.

π/2 π/2
Therefore, ∫0 𝑔(𝑥)d𝑥 is convergent and hence ∫0 𝑓(𝑥)d𝑥 is
convergent.
By taking 0 < 𝜇 < 1 and also 𝜇 > 𝑝 − 1 so that
−1 < 𝑝 − 1 < 𝜇 < 1 i.e. 0 < 𝑝 < 2.

π/2 π/2
Therefore, ∫0 𝑔(𝑥)d𝑥 is convergent and hence ∫0 𝑓(𝑥)d𝑥 is
convergent.
π/2 sin 𝑥
Hence ∫0 d𝑥 is convergent if 𝑝 < 2 and divergent if 𝑝 ≥ 2.
𝑥𝑝
π/2 𝑠𝑖𝑛 𝑚 𝑥
Example 5: Show that ∫0 𝑑𝑥 exists if and only if 𝑛 < 𝑚 + 1.
𝑥𝑛
𝑠𝑖𝑛 𝑚 𝑥 𝑠𝑖𝑛𝑥 𝑛 1
Sol. Here 𝑓 (𝑥 ) = =( ) . 𝑥 𝑛−𝑚
𝑥𝑛 𝑥
0 𝑖𝑓 𝑛 − 𝑚 < 0
lim 𝑓(𝑥) = { 1 𝑖𝑓 𝑛 − 𝑚 = 0
𝑥⟶0+
∞ 𝑖𝑓 𝑛 − 𝑚 > 0
Therefore, the given integral is proper integral if 𝑛 − 𝑚 ≤ 0 𝑖. 𝑒.
𝑛 ≤ 𝑚 and an improper integral if 𝑛 − 𝑚 > 0; 0 is only the point of

𝜋
infinite discontinuity of 𝑓 on [0, 2 ].
1
When 𝑛 − 𝑚 > 0, take 𝑔(𝑥 ) = 𝑥 𝑛−𝑚
𝑓(𝑥) 𝑠𝑖𝑛𝑥 𝑛
Therefore, lim = lim ( ) = 1 which is non zero and
𝑥⟶0+ 𝑔(𝑥) 𝑥⟶0+ 𝑥
finite.
π/2 π/2 1
Also ∫0 𝑔(𝑥) 𝑑𝑥 = ∫0 𝑑𝑥 is convergent iff 𝑛 − 𝑚 < 1 i.e.
𝑥 𝑛−𝑚
𝑛 < 𝑚 + 1.
Therefore, by comparison test, the given interval is convergent iff

𝑛 − 𝑚 < 1.
13.6 GENERAL TEST FOR CONVERGENCE

This test for convergence of an improper integral (finite limits of
integration but discontinuous integrand) holds whether or not the
integrand keeps the same sign.
b
 Cauchy’s test. The improper integral ∫𝑎 𝑓(𝑥) 𝑑𝑥, 𝑎 being the
only point of infinite discontinuity, converges at 𝑎 if and only if
to each 𝜖 > 0, there corresponds a 𝛿 > 0 such that
𝑎+𝜆
|∫𝑎+𝜆 2 𝑓 (𝑥 )𝑑𝑥| < 𝜖 for all 0 < 𝜆1 , 𝜆2 < 𝛿.
1
13.7 ABSOLUTE CONVERGENCE

b
Definition: The improper integral ∫𝑎 𝑓(𝑥) 𝑑𝑥 is said to be absolutely
b
convergent if ∫𝑎 |𝑓(𝑥)| 𝑑𝑥 is convergent.
Theorem: Every absolutely convergent integral is convergent.
b b
or ∫𝑎 |𝑓(𝑥)| 𝑑𝑥 exits ⟹ ∫𝑎 𝑓(𝑥) 𝑑𝑥 exists.
b
Proof. Since ∫𝑎 |𝑓(𝑥)| 𝑑𝑥 exists, therefore by Cauchy’s test , for every
𝜀 > 0, there corresponds a 𝛿 > 0 such that.
𝑎+𝜆
|∫𝑎+𝜆 2 |𝑓(𝑥)|𝑑𝑥 | < 𝜀, ∀ 0 < 𝜆1 , 𝜆2 < 𝛿 ….. (1)
1
𝑎+𝜆 𝑎+𝜆
Also, we know that |∫𝑎+𝜆 2 𝑓(𝑥 )𝑑𝑥| ≤ |∫𝑎+𝜆 2 |𝑓(𝑥)|𝑑𝑥 | …… (2)
1 1
𝑎+𝜆
From (1) and (2), we have |∫𝑎+𝜆 2 𝑓 (𝑥 )𝑑𝑥| < 𝜀, , ∀0 < 𝜆1 , 𝜆2 < 𝛿
1
b
∴ By Cauchy’s test ∫𝑎 𝑓(𝑥) 𝑑𝑥 is exists.
Note: Since |𝑓(𝑥)| is always positive, the comparison tests can be

b
applied for examining the convergence of ∫𝑎 |𝑓(𝑥)| 𝑑𝑥, i.e.,
b
absolute convergence ∫𝑎 𝑓(𝑥) 𝑑𝑥.
Note 2: The converse of the above theorem is not true. Every

convergent integral is not absolutely convergent. A convergent integral
which is not absolutely convergent is called a conditionally Convergent
Integral.
1
1 sin
Example 1. Test the convergence of ∫0 𝑥𝑥 𝑑𝑥.
√
1
sin
𝑥
Sol. Let f(x) =
√𝑥
Clearly, 𝑓 does not keep the same sign in a neighborhood of 0.

1 1
sin | sin
Now |𝑓(𝑥)| = | √𝑥 | = 𝑥 𝑥
|√𝑥 |
|
≤
1
√𝑥
, ∀ 𝑥 ∈ (0, 1]
1 1 1
But ∫0 𝑑𝑥 is convergent at 0. (Since 𝑛 = 2 < 1)
√𝑥
1
Therefore, by comparison test, ∫0 |𝑓(𝑥)|𝑑𝑥 is convergent at 0.
Since absolute convergence ⟹ convergence.

1
1 sin 𝑥
Therefore, ∫0 𝑑𝑥 is convergent.
√𝑥
1
1 sin
Example 2. Show that ∫0 𝑥 𝑝𝑥 𝑑𝑥 , p > 0, converges absolutely for p < 1.
1
sin
Sol. Let 𝑓 (𝑥) = 𝑥
p>0
𝑥𝑝
Clearly, 𝑓 does not keep the same sign in a neighborhood of 0.

1 1
sin | sin | 1
𝑥
Now, |𝑓(𝑥)| = | 𝑥
|= ≤ 𝑥 𝑝 , ∀ 𝑥 ∈ (0, 1]
𝑥𝑝 |𝑥 𝑝 |
1 1
Also ∫0 𝑑𝑥 is convergent iff p < 1.
𝑥𝑝
1
Therefore, by convergent test, ∫0 |𝑓(𝑥)|𝑑𝑥 converges if p < 1.
1
Hence ∫0 𝑓 (𝑥 )𝑑𝑥 converges absolutely for p < 1.
13.8 CONVERGENT AT ∞
Theorem: A necessary and sufficient condition for convergence of

∞
∫𝑎 𝑓 (𝑥 )𝑑𝑥, where 𝑓 (𝑥) > 0 ∀ 𝑥 ∈ [𝑥, 𝑡], is that there exists a positive
𝑡
number M, independent of t, such ∫𝑎 𝑓(𝑥 )𝑑𝑥 < M ∀ 𝑡 ≥ 𝑎.
𝑡
Proof. Let F(t) = ∫𝑎 𝑓(𝑥)𝑑𝑥
Since 𝑓 is positive in [𝑎, 𝑡], the function F(t) monotonically increases

with 𝑡 and will therefore, tend to a finite limit if and only if it is bounded
above, i.e. there exists a positive number M, independent of 𝑡, such that
𝐹(𝑡) < M ∀ 𝑡 ≥ 𝑎
𝑡
⟹ ∫𝑎 𝑓(𝑥)𝑑𝑥 < 𝑀 ∀ 𝑡 ≥ 𝑎
Note: if no such number M exists, then the monotonic increasing

function F(t) is unbounded above and therefore tends to ∞ as 𝑡 ⟶ ∞.
𝑡
Therefore, ∫𝑎 𝑓(𝑥)𝑑𝑥 diverges to ∞.
 Comparison test I.
If 𝑓 and 𝑔 are two functions such that
0 < 𝑓(𝑥) ≤ 𝑔(𝑥) ∀ 𝑥 ∈ [𝑎, ∞), then
∞ ∞
(i) ∫𝑎 𝑔(𝑥 )𝑑𝑥 is convergent ⟹ ∫𝑎 𝑓(𝑥 )𝑑𝑥 is
convergent
∞ ∞
(ii) ∫𝑎 𝑓(𝑥 )𝑑𝑥 is divergent ⟹ ∫𝑎 𝑔(𝑥 )𝑑𝑥 is
divergent.
 Comparison test II.

𝑓(𝑥)
If 𝑓 and 𝑔 are two positive functions on [𝑎, ∞) and lim =𝑙
𝑥⟶∞ 𝑔(𝑥)
then (i) if 𝑙 is non-zero finite, the two integrals
∞ ∞
∫𝑎 𝑓 (𝑥 )𝑑𝑥 and ∫𝑎 𝑔(𝑥 )𝑑𝑥 converges or diverges
together.
∞ ∞
(ii) if 𝑙 = 0 and ∫𝑎 𝑔(𝑥 )𝑑𝑥 converges, then ∫𝑎 𝑓(𝑥 )𝑑𝑥
converges.
∞ ∞
(iii) if 𝑙 = ∞ and ∫𝑎 𝑔(𝑥 )𝑑𝑥 diverges, then ∫𝑎 𝑓(𝑥)𝑑𝑥
diverges.
Note: A useful comparison integral.

∞ 1
The improper integral ∫𝑎 𝑑𝑥 (𝑎 > 0) convergent if and only if
𝑥𝑛
𝑛 > 1.
Example 1. Examine the convergence of the following integral
∞ 𝑥3 ∞ 1
(i) ∫1 (1+𝑥)5
𝑑𝑥 (ii) ∫1 𝑑𝑥
(2+𝑥)√𝑥
𝑥3 𝑥3 1
Sol. (i) Let f(x) = (1+𝑥)5 = 1 5
= 1 5
𝑥 5 (1+ ) 𝑥 2(1+ )
𝑥 𝑥
1
Take g(x) = 𝑥 2
𝑓(𝑥) 1
Therefore, lim = lim 5 =1 which is non-zero and finite.
𝑥⟶∞ 𝑔(𝑥) 𝑥⟶∞ (1+1 )
𝑥
∞ ∞
By comparison test, the two integrals ∫1 𝑓 (𝑥 )𝑑𝑥 and ∫1 𝑔(𝑥 )𝑑𝑥
converge or diverge together.
∞ ∞ 1
But ∫1 𝑔(𝑥 )𝑑𝑥 = ∫1 𝑥2
𝑑𝑥 is convergent (since n = 2 >1)
∞ ∞ 𝑥3
Therefore, ∫1 𝑓 (𝑥 )𝑑𝑥 = ∫1 (1+𝑥)5
𝑑𝑥 is convergent.
1 1
(ii) Let f(x) = (2+𝑥) = 3
√𝑥 2
𝑥 2 (1+ )
𝑥
1
Take g(x) = 𝑥 3/2
𝑓(𝑥) 1
Therefore, lim = lim 2 =1 which is non-zero and finite.
𝑥⟶∞ 𝑔(𝑥) 𝑥⟶∞ 1+𝑥
∞ ∞
By comparison test, the two integrals ∫1 𝑓 (𝑥 )𝑑𝑥 and ∫1 𝑔(𝑥 )𝑑𝑥
converge or diverge together.
∞ ∞ 1 3
But ∫1 𝑔(𝑥 )𝑑𝑥 = ∫1 𝑥 3/2
𝑑𝑥 is convergent (since n = 2 > 1)
∞ ∞ 1
Therefore, ∫1 𝑓 (𝑥 )𝑑𝑥 = ∫1 𝑑𝑥 is convergent.
(2+𝑥)√𝑥
Example 2. Examine the convergence of the following integral

∞ 𝑥 2𝑚
∫0 𝑑𝑥 𝑚, 𝑛 > 0
1 + 𝑥 2𝑛
∞ 𝑥 2𝑚 𝑎 𝑥 2𝑚 ∞ 𝑥 2𝑚
Sol. ∫0 𝑑𝑥 = ∫0 𝑑𝑥 + ∫𝑎 𝑑𝑥 where, 𝑎 > 0
1 + 𝑥 2𝑛 1 + 𝑥 2𝑛 1 + 𝑥 2𝑛
The first integral on the right is a proper integral and therefore,

convergent. The given integral will be convergent or divergent according
∞ 𝑥 2𝑚
as ∫𝑎 𝑑𝑥 is convergent or divergent.
1 + 𝑥 2𝑛
𝑥 2𝑚 𝑥 2𝑚 1
Let 𝑓 (𝑥 ) = 1 + 𝑥 2𝑛 = 1 = 1
𝑥 2𝑛 (1+ 2𝑛 ) 𝑥 2𝑛−2𝑚 (1+ 2𝑛 )
𝑥 𝑥
1
Take 𝑔(𝑥 ) = 𝑥 2𝑛−2𝑚
𝑓(𝑥) 1
lim = lim 1 =1 (Since n > 0)
𝑥⟶∞ 𝑔(𝑥) 𝑥⟶∞ 1 + 2𝑛
𝑥
Which is non-zero and finite.

∞ ∞
Therefore, by comparison test, ∫𝑎 𝑓(𝑥)𝑑𝑥 and ∫𝑎 𝑔(𝑥)𝑑𝑥 converge or
diverge together.
∞ ∞ 1
But ∫0 𝑔(𝑥 )𝑑𝑥 = ∫0 𝑑𝑥 converges if and only if 2𝑛 − 2𝑚 >
𝑥 2𝑛 − 2𝑚
1
1 i.e. 𝑛 − 𝑚 > 2 .
∞ 1
Therefore, ∫𝑎 𝑓(𝑥)𝑑𝑥 converges if and only if 𝑛 − 𝑚 > 2. Hence the
1
given integral converges if and only if 𝑛 − 𝑚 > 2.
CHECK YOUR PROGRESS
Problem 1. Every improper integral is always convergent.

∞ 1
Problem 2. ∫𝑎 𝑑𝑥 (a > 0) is convergent if n > 1.
𝑥𝑛
∞ 1
Problem 3. ∫2 𝑑𝑥 is divergent.
√𝑥 2 −1
∞ 𝑙𝑜𝑔𝑥
Problem 4. ∫1 𝑥 2 𝑑𝑥 is convergent.
Problem 5. Every absolute convergent is convergent.
13.9 SUMMARY
1. A necessary and sufficient condition for the convergence of the

b
improper integral ∫𝑎 f(𝑥 )d𝑥 at ‘𝑎′ where f is positive on (𝑎, 𝑏], is that
there exists a positive number M, independent of 𝜖 > 0 such that
b
∫𝑎+𝜖 f(𝑥 )d𝑥 < 𝑀 ∀ 𝜖 in (0, 𝑏 − 𝑎).
2. Comparison test: If f and 𝑔 are two positive functions with
f(𝑥) ≤ 𝑔(𝑥) for all 𝑥 in (𝑎, 𝑏] and 𝑎 is only point of infinite

discontinuity on [𝑎, 𝑏], then
b b
(i) ∫𝑎 g(𝑥 )d𝑥 is convergent ⟹ ∫𝑎 f(𝑥 )d𝑥 is convergent
b b
(ii) ∫𝑎 f(𝑥)d𝑥 is divergent ⟹ ∫𝑎 g(𝑥)d𝑥 is divergent.
3. if 𝑎 is the only point of infinite discontinuity of 𝑓 on [𝑎, 𝑏] and

b
lim (𝑥 − 𝑎)𝜇 𝑓(𝑥) exists and non-zero finite, then ∫𝑎 𝑓(𝑥)d𝑥
𝑥⟶𝑎+
13.10 GLOSSARY
sequence
series
13.11 REFERENCES



∞ 𝑥 𝑝−1
1+𝑥
Q 2. Prove that every absolute convergent integral is convergent.
∞ 𝑥 3+1
𝑥4
∞ 𝑐𝑜𝑠𝑥
1+𝑥 2
∞ 1
Q 5. Examine for convergence of the integral ∫𝑒 𝑑𝑥.
𝑥(𝑙𝑜𝑔𝑥)𝑛+1
13.14 ANSWERS
TQ1. Convergent if 0 < 𝑝 < 1 and divergent if 𝑝 ≥ 1.
TQ3. Convergent.
TQ4. Convergent.
TQ5. Divergent if 𝑛 ≤ 0, convergent if n < 0.
CHECK YOUR PROGRESS
CYQ 1. False
CYQ 2. True
CYQ 3. True
CYQ 4. True
CYQ 5. True
UNIT 14: DIRICHLET AND ABEL’S TEST FOR

IMPROPER INTEGRALS
Contents
14.1 Introduction
14.2 Objectives
14.3 Abel’s Test
14.4 Dirichlet’s Test
14.5 Summary
14.6 Glossary
14.8 References
14.10 Answers
14.1 INTRODUCTION
In mathematics, Abel's test (also known as Abel's criterion) is

a method of testing for the convergence of an infinite series. The test is
named after mathematician Niels Henrik Abel, who proved it in
1826.[1] There are two slightly different versions of Abel's test – one is
used with series of real numbers, and the other is used with power
series in complex analysis. Abel's uniform convergence test is a criterion
for the uniform convergence of a series of functions dependent
on parameters.
In mathematics, there are several integrals known as the Dirichlet
integral, after the German mathematician Peter Gustav Lejeune Dirichlet,
one of which is the improper integral of the sine function over the positive
real line.
14.2 OBJECTIVES
 Abel’s Test
 Dirichlet’s Test
14.3 ABEL’S TEST

∞
If ∫a f(𝑥 )d𝑥 is convergent at ∞ and 𝑔(𝑥) is bounded and monotonic for
∞
𝑥 ≥ 𝑎, then ∫a f(𝑥 )𝑔(𝑥)d𝑥 is convergent at ∞.
Or
An infinite integral which converges (not necessarily absolutely) will

remain convergent after the insertion of a factor which is bounded and
monotonic.
Proof: Since 𝑔 is monotonic on [𝑎, ∞), it is integrable on [𝑎, 𝑡], for all
𝑡 ≥ 𝑎.
Also, since 𝑓 is integrable on [𝑎, 𝑡], we have by second mean value

theorem.
𝑡2 p 2 𝑡
∫𝑡1 f(𝑥 )g(𝑥)d𝑥 = 𝑔( 𝑡1 ) ∫𝑡1 f(𝑥)d𝑥 = 𝑔(𝑡2 ) ∫𝑝 f(𝑥 )d𝑥 ………. (1)
Where 𝑎 < 𝑡1 ≤ 𝑝 ≤ 𝑡2
Since 𝑔 is bounded on [𝑎, ∞), there exists a positive number 𝑘 such that
|𝑔(𝑥)| ≤ 𝑘 ∀ 𝑥 ≥ 𝑎
In particular |𝑔(𝑡1 )| ≤ 𝑘, |𝑔(𝑡2 )| ≤ 𝑘 ………. (2)
Let ∈ > 0 be given,

∞
Since ∫a f(𝑥 )d𝑥 is convergent, there exists a number 𝑡0 such that
𝑡 𝜖
|∫𝑡 2 f(𝑥)d𝑥| ≤ ∀ 𝑡1 , 𝑡2 ≥ 𝑡0 ……… (3)
1 2𝑘
Let the number 𝑡1 , 𝑡2 in (1) be ≥ 𝑡0 so that the number 𝑝 which lies

between 𝑡1 and 𝑡2 , is also ≥ 𝑡0 . Hence from (3),
p 𝜖 𝑡 𝜖
|∫𝑡 f(𝑥 )d𝑥 | ≤ , |∫p 2 f(𝑥 )d𝑥 | ≤ ……… (4)
1 2𝑘 2𝑘
From (1), (2) and (4), it follows that a positive number 𝑡0 exists such
that for all 𝑡1 , 𝑡2 ≥ 𝑡0 .
𝑡 p 𝑡
|∫𝑡 2 f(𝑥)g(𝑥)d𝑥| = |𝑔(𝑡1 ) ∫𝑡 f(𝑥 )d𝑥 + 𝑔(𝑡2 ) ∫p 2 f(𝑥 )d𝑥|
1 1
p 𝑡 𝜖 𝜖
≤ |𝑔(𝑡1 )| |∫𝑡 f(𝑥 )d𝑥 | + |𝑔(𝑡2 )| |∫𝑝 2 f(𝑥 )d𝑥| < 𝑘. 2𝑘 + 𝑘. 2𝑘 = 𝜖
1
∞
Hence, by Cauchy’s test, ∫a f(𝑥 )g(𝑥)d𝑥 is convergent at ∞.
14.4 DRICHLET’S TEST

t
If ∫a f(𝑥 )d𝑥 is bounded for all 𝑡 ≥ 𝑎 and 𝑔(𝑥) is a bounded and
monotonic function for 𝑥 ≥ 𝑎, tending to 0 as 𝑥 ⟶ ∞, then
∞
∫a f(𝑥 )g(𝑥)d𝑥 is convergent at ∞.
Or
An infinite integral which oscillates finitely becomes convergent after

the insertion of a monotonic factor which tends to zero as limit.
Proof. Since g is monotonic on [𝑎, ∞), it is integrable on [𝑎, 𝑡], for all
𝑡 ≥ 𝑎.
Also, since 𝑓 is integrable on [𝑎, 𝑡], we have by second mean values

theorem,
𝑡2 p 2 𝑡
∫𝑡1 f(𝑥 )g(𝑥)d𝑥 = 𝑔( 𝑡1 ) ∫𝑡1 f(𝑥 )d𝑥 = 𝑔(𝑡2 ) ∫𝑝 f(𝑥 )d𝑥 ………. (1)
Where 𝑎 < 𝑡1 ≤ 𝑝 ≤ 𝑡2
t
Since ∫a f(𝑥 )d𝑥 is bounded for all 𝑡 ≥ 𝑎, there exists a positive number
𝑘 such that
p
|∫𝑡 f(𝑥 )d𝑥| ≤ 𝑘 ∀ 𝑡 ≥ 𝑎 ……… (2)
1
p a p
Now, |∫𝑡 f(𝑥 )d𝑥 | = |∫𝑡 f(𝑥 )d𝑥 + ∫a f(𝑥 )d𝑥|
1 2
p 𝑡
= |∫a f(𝑥 )d𝑥 − ∫a 1 f(𝑥 )d𝑥|
p 𝑡
≤ |∫a f(𝑥 )d𝑥| + |∫a 1 f(𝑥 )d𝑥|
≤ 𝑘 + 𝑘 = 2𝑘 ∀ 𝑡1 , 𝑝 ≥ 𝑎 …….. (3)
𝑡
Similarly, |∫𝑝 2 f(𝑥 )d𝑥| ≤ 2𝑘 ∀ 𝑡2 , 𝑝 ≥ 𝑎 …….. (4)
Let 𝜖 > 0 be given
Since lim 𝑔(𝑥 ) = 0, there exists a number 𝑡0 such that

𝑥⟶∞
𝜖
|𝑔(𝑥)| < ∀ 𝑥 ≥ 𝑡0
4𝑘
Let the number 𝑡1 , 𝑡2 in (1) be ≥ 𝑡0 , then

𝜖 𝜖
|𝑔(𝑡1 )| < and |𝑔(𝑡2 )| < 4𝑘 ……… (5)
4𝑘
From (1), (3), (4) and (5) it follows that a positive number 𝑡0 exists such
that for all 𝑡1 , 𝑡2 ≥ 𝑡0
𝑡 p 𝑡
|∫𝑡 2 f(𝑥)g(𝑥)d𝑥| = |𝑔(𝑡1 ) ∫𝑡 f(𝑥 )d𝑥 + 𝑔(𝑡2 ) ∫p 2 f(𝑥 )d𝑥|
1 1
p 𝑡 𝜖 𝜖
≤ |𝑔(𝑡1 )| |∫𝑡 f(𝑥 )d𝑥 | + |𝑔(𝑡2 )| |∫𝑝 2 f(𝑥 )d𝑥| < 4𝑘 . 2𝑘 + 4𝑘 . 2𝑘 = 𝜖
1
∞
Hence, by Cauchy’s test, ∫a f(𝑥 )g(𝑥)d𝑥 is convergent at ∞.
Examples 1. Examine the convergence of the integrals:
∞ 𝑠𝑖𝑛𝑥 ∞ 𝑠𝑖𝑛𝑥 ∞ 𝑠𝑖𝑛𝑥

(i) ∫0 d𝑥 (ii) ∫0 d𝑥 (iii) ∫0 d𝑥
𝑥 √𝑥 𝑥 3/2
∞ 𝑠𝑖𝑛𝑥
(iv) ∫a d𝑥 where a and m both are positive.
𝑥𝑚
𝑠𝑖𝑛𝑥
Sol. (i) Since lim = 1, therefore 0 is not a point of infinite
𝑥⟶∞ 𝑥
discontinuity.
∞ 𝑠𝑖𝑛𝑥 1 𝑠𝑖𝑛𝑥 ∞ 𝑠𝑖𝑛𝑥
Now, ∫0 d𝑥 = ∫0 d𝑥 + ∫1 d𝑥
𝑥 𝑥 𝑥
1 𝑠𝑖𝑛𝑥
Also ∫0 d𝑥 is a proper integral. Let us examine the convergence of
𝑥
∫1 d𝑥 at ∞.
𝑥
1
Let 𝑓 (𝑥 ) = 𝑠𝑖𝑛𝑥 and 𝑔(𝑥 ) = 𝑥
t t
Since |∫1 f(𝑥 )d𝑥| = |∫1 sin 𝑥 d𝑥|
= |𝑐𝑜𝑠1 − 𝑐𝑜𝑠𝑡| ≤ |𝑐𝑜𝑠1| + |𝑐𝑜𝑠𝑡| ≤ 2

t
Therefore, ∫1 f(𝑥 )d𝑥 is bounded for all 𝑡 ≥ 1.
Also 𝑔(𝑥) is bounded and monotonically decreasing function tending to

0 as 𝑥 ⟶ ∞ .
∞ ∞ 𝑠𝑖𝑛𝑥
By Dirichlet’s test, ∫1 f(𝑥 )g(𝑥)d𝑥 = ∫1 d𝑥 is convergent.
𝑥
Hence, from (1), ∫0 d𝑥 is convergent.
𝑥
𝑠𝑖𝑛𝑥 𝑠𝑖𝑛𝑥
(ii) Since lim = lim . √𝑥 = 1 × 0 = 0.
𝑥⟶∞ √𝑥 𝑥⟶∞ 𝑥
Therefore, 0 is not a point of infinite discontinuity.

Now ∫0 d𝑥 = ∫0 d𝑥 + ∫1 d𝑥 …….. (1)
√𝑥 √𝑥 √𝑥
1 𝑠𝑖𝑛𝑥
Also ∫0 d𝑥 is a proper integral. So, let us examine the convergence
√𝑥
of ∫1 𝑥 d𝑥 at ∞.
√
1
Let 𝑓(𝑥 ) = 𝑠𝑖𝑛𝑥 and 𝑔(𝑥 ) =
√𝑥
t
Since |∫1 f(𝑥 )d𝑥 | ≤ 2 [see part (i)]
t
Therefore, ∫1 f(𝑥 )d𝑥 is bounded for all 𝑡 ≥ 1.
Also 𝑔(𝑥) is a bounded and monotonically decreasing function tending

to 0 as 𝑥 ⟶ ∞ .
√𝑥
√𝑥

(iii) ∫0 d𝑥 = ∫0 d𝑥 + ∫1 d𝑥 ……… (1)
𝑥 3/2 𝑥 3/2 𝑥 3/2
1 𝑠𝑖𝑛𝑥
For the integral ∫0 d𝑥, 0 is a point of infinite discontinuity.
𝑥 3/2
𝑠𝑖𝑛𝑥 𝑠𝑖𝑛𝑥 1
Let 𝑓 (𝑥 ) = 𝑥 3/2 = .
𝑥 √𝑥
1
Take 𝑔(𝑥 ) =
√𝑥
𝑓(𝑥) 𝑠𝑖𝑛𝑥
Therefore, lim = lim = 1 which is non-zero and finite.
𝑥⟶0+ 𝑔(𝑥) 𝑥⟶0+ 𝑥
1 1 1
Since ∫0 𝑔(𝑥 )d𝑥 = ∫0 d𝑥 is convergent.
√𝑥
1 1 𝑠𝑖𝑛𝑥
∴ By comparison test, ∫0 𝑓(𝑥 )d𝑥 = ∫0 d𝑥 is convergent.
𝑥 3/2
Convergence of ∫1 d𝑥 at ∞.
𝑥 3/2
1
Let 𝑓 (𝑥 ) = 𝑠𝑖𝑛𝑥 and 𝑔(𝑥 ) = 𝑥 3/2
t
Since |∫1 f(𝑥 )d𝑥| ≤ 2 [see part (i)]
t
∴ ∫1 f(𝑥 )d𝑥 is bounded for all 𝑡 ≥ 1.

to 0 as 𝑥 ⟶ ∞ .
𝑥 3/2
𝑥 3/2
1
(iv) Let 𝑓 (𝑥) = 𝑠𝑖𝑛𝑥 and 𝑔(𝑥 ) = 𝑥 𝑚 , m > 0
t t
Since |∫𝑎 f(𝑥 )d𝑥 | = |∫𝑎 sin 𝑥 d𝑥 |
= |cos 𝑎 − cos 𝑡| ≤ |cos 𝑎| + |cos 𝑡| ≤ 2

t
∴ ∫𝑎 f(𝑥 )d𝑥 is bounded for all 𝑡 ≥ 1.

to 0 as 𝑥 ⟶ ∞ for m > 0.
By Dirichlet’s test, ∫𝑎 f(𝑥 )g(𝑥)d𝑥 = ∫a d𝑥 where m and a are both
𝑥𝑚
positive, is convergent.

∞ ∞ 𝑥
(i) ∫0 sin 𝑥 2 d𝑥 (ii) ∫0 sin 𝑥 d𝑥
1+𝑥 2
∞ 1 ∞
Sol. (i) We have ∫0 sin 𝑥 2 d𝑥 = ∫0 sin 𝑥 2 d𝑥 + ∫1 sin 𝑥 2 d𝑥 …… (1)
1
But ∫0 sin 𝑥 2 d𝑥 is a proper integral and therefore convergent.
∞
Convergence of ∫𝟏 𝐬𝐢𝐧 𝒙𝟐 𝐝𝒙 at ∞.
∞ ∞ 1
∫1 sin 𝑥 2 d𝑥 = ∫1 (2𝑥 sin 𝑥 2 ). 2𝑥 d𝑥
1
Let 𝑓 (𝑥 ) = 2𝑥 sin 𝑥 2 and 𝑔(𝑥 ) = 2𝑥
t t
Since, |∫1 f(𝑥 )d𝑥| = |∫1 2𝑥 sin 𝑥 2 d𝑥 | = |{− cos 𝑥 2 }1𝑡 |
= |cos 𝑎 − cos 𝑡 2 | ≤ |cos 𝑎| + |cos 𝑡 2 | ≤ 2

t

to 0 as 𝑥 ⟶ ∞.
∞ ∞
By Dirichlet’s test, ∫1 f(𝑥 )g(𝑥)d𝑥 = ∫1 sin 𝑥 2 d𝑥 is convergent.
∞
Hence, from (1) ∫0 sin 𝑥 2 d𝑥 is convergent.
(ii) We have
∞ 𝑥
∫0 sin 𝑥 d𝑥
1+𝑥 2
1 𝑥 ∞ 𝑥
= ∫0 sin 𝑥 d𝑥 + ∫1 sin 𝑥 d𝑥 …… (1)
1+𝑥 2 1+𝑥 2
1 𝑥
But ∫0 sin 𝑥 d𝑥 is a proper integral and therefore convergent.
1+𝑥 2
∞ 𝒙
Convergence of ∫𝟏 𝐬𝐢𝐧 𝒙 𝐝𝒙 ∞.
𝟏+𝒙𝟐
𝑥
Let 𝑓 (𝑥 ) = 𝑠𝑖𝑛 𝑥 and 𝑔(𝑥 ) = 1+𝑥 2
t
Since |∫1 f(𝑥 )d𝑥| ≤ 2
t
𝑥
Also, lim 𝑔(𝑥) = lim =0
𝑥⟶∞ 𝑥⟶∞ 1+𝑥 2
𝑔(𝑥) is a bounded and monotonically decreasing function tending to 0

as 𝑥 ⟶ ∞.
∞ ∞ 𝑥
By Dirichlet’s test, ∫1 f(𝑥 )g(𝑥)d𝑥 = ∫1 sin 𝑥 d𝑥 is convergent.
1+𝑥 2
𝑥
Hence, from (1) 1+𝑥 2 sin 𝑥 is convergent.

∞ 𝑠𝑖𝑛𝑥 ∞ 𝑠𝑖𝑛𝑥
(i) ∫0 𝑒 −𝑎𝑥 d𝑥, 𝑎 ≥ 0 (ii) ∫𝑎 𝑒 −𝑥 d𝑥, 𝑎 > 0
𝑥 𝑥2
𝑠𝑖𝑛𝑥
Sol. (i) Let 𝑓 (𝑥 ) = and 𝑔(𝑥 ) = 𝑒 −𝑎𝑥 , 𝑎 ≥ 0.
𝑥
∞
Since ∫0 𝑓(𝑥)d𝑥 is convergent and 𝑔(𝑥 ) is bounded and monotonically
decreasing function of 𝑥 for 𝑥 > 0.
∴ By Abel’s test, ∫0 𝑓 (𝑥 )𝑔(𝑥)d𝑥 = ∫0 𝑒 −𝑎𝑥 d𝑥 is convergent.
𝑥
𝑠𝑖𝑛𝑥
(ii) Let 𝑓 (𝑥 ) = and 𝑔(𝑥 ) = 𝑒 −𝑥
𝑥2
𝒔𝒊𝒏 𝒙 1 ∞ 1
Since |𝑓(𝑥)| = | |≤ and ∫a d𝑥 is convergent.
𝑥2 𝑥2 𝑥2
∞
Therefore, ∫a 𝑓 (𝑥 )d𝑥 is also convergent.
Again 𝑔(𝑥 ) is monotonic decreasing and bounded function for 𝑥 > 𝑎.

Therefore, by Abel’s test, ∫𝑎 𝑓(𝑥 ) g(𝑥)d𝑥 = ∫𝑎 𝑒 −𝑥 d𝑥, 𝑎 > 0 is
𝑥2
convergent.
CHECK YOUR PROGRESS
Problem 1. Improper integral is not convergent.

∞
Problem 2. If ∫a f(𝑥 )d𝑥 is convergent at ∞ and 𝑔(𝑥) is bounded and
∞
monotonic for 𝑥 ≥ 𝑎, then ∫a f(𝑥 )𝑔(𝑥)d𝑥 is convergent at ∞.
∞
Problem 3. ∫0 cos 𝑥 3 𝑑𝑥 is divergent.
∞ 𝑙𝑜𝑔𝑥
Problem 4. ∫1 𝑑𝑥 is convergent.
𝑥2
Problem 5. Every absolute convergent need not be convergent.
14.5 SUMMARY
∞
1. If ∫a f(𝑥 )d𝑥 is convergent at ∞ and 𝑔(𝑥) is bounded and monotonic
∞
for 𝑥 ≥ 𝑎, then ∫a f(𝑥 )𝑔(𝑥)d𝑥 is convergent at ∞. (This is Abel’s Test).
t
2. If ∫a f(𝑥 )d𝑥 is bounded for all 𝑡 ≥ 𝑎 and 𝑔(𝑥) is a bounded and
monotonic function for 𝑥 ≥ 𝑎, tending to 0 as 𝑥 ⟶ ∞, then
∞
∫a f(𝑥 )g(𝑥)d𝑥 is convergent at ∞. (This is Dirichlet’s Test).
14.6 GLOSSARY
Proper integral
Improper integral
14.7 REFERENCES




∞ cos 𝑥
Q 1. Test the convergence of the integral ∫0 𝑑𝑥.
√𝑥+𝑥 2
Q 2. Define Abel’s Test with example.
∞ log 𝑥 sin 𝑥
Q 3. Examine for convergence of the integral ∫𝑒 𝑑𝑥 .
𝑥
∞ 𝑐𝑜𝑠𝑥
Q 4. Test the convergence of the integral ∫𝑎 (1 − 𝑒 −𝑥 ). 𝑑𝑥 , 𝑎 > 0.
𝑥2
Q 5. Define Dirichlet’s Test with example.
14.10 ANSWERS
TQ1. Convergent.
TQ3. Convergent.
TQ4. Convergent.
CHECK YOUR PROGRESS
CYQ 1. False
CYQ 2. True
CYQ 3. False
CYQ 4. True
CYQ 5. False
REAL ANALYSIS MT(N) - 201
E-mail:info@uou.ac.in
Website: https://www.uou.ac.in/
Teen Pani Bypass Road, Transport Nagar

Uttarakhand Open University, Haldwani, Nainital-263139
Uttarakhand, India
Phone No. 05946-261122, 261123
Toll free No. 18001804025
Fax No. 05946-264232,

MT (N) 201

Uploaded by

Copyright:

Available Formats

MT (N) 201

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MT (N) 201

Uploaded by

Copyright:

Available Formats

REAL ANALYSIS MT(N) - 201

BACHELOR OF SCIENCE/ BACHELOR OF ARTS

COURSE CODE: MT(N) 201

BOARD OF STUDIES - 2023

Prof. P. D. Pant Prof. Harish Chandra Prof. Manoj Kumar

Prof. Sanjay Kumar Dr. Arvind Bhatt Dr. Jyoti Rani

Dr. Kamlesh Bisht Dr. Shivangi Upadhyay

Uttarakhand Open University

Unit Writer Blocks Units

Dr. Deepak Kumar Sharma I, II and III 1 to 14

BLOCK I: REAL NUMBERS, SEQUENCE Page Number 01 - 85

Unit – 1 Sets and functions 02-18

Unit – 2 Real Numbers 19 - 32

Unit - 3 Limit points, open and 33 - 48

Unit - 4 Real sequences 49 - 67

Unit - 5 Infinite Series 68 - 85

BLOCK II: FUNCTIONS SINGLE VARIABLE Page Number 86 - 140

Unit – 6 Limit 87 - 109

Unit - 7 Continuous Function 110 - 122

Unit - 8 Differentiation 123 - 140

Unit – 9 Riemann integral I 142 - 155

Unit - 10 Riemann integral II 156 - 177

Unit - 11 Sequence and series of 178 - 191

Unit - 13 Improper integral II 205 - 218

Second block is Functions Single Variable, in this block Limits of

Course Name: REAL ANALYSIS

UNIT 1: SETS AND FUNCTIONS

Set theory, branch of mathematics that deals with the properties

i. To analyze and predict the behavior of these systems over time.

A set is a well - defined collection of distinct objects.

The symbol ∈ is used to indicate ‘belongs to’. Thus x ∈ A ⟹ x is an

The symbol ∉ is used to indicate ‘does not belong to’. Thus x ∉ A ⟹ x is

Example: Let A = {1, 2, 3, 4, 5} be a set then we say 1∈ 𝐴 , 2∈ A, 3∈ 𝐴,

1.4 METHODS OF DESCRIBING A SET

There are two methods of describing a set.

1.5 TYPES OF SETS

1.6 SUBSET, SUPERSET AND POWER SET

In set theory, a subset is denoted by the symbol ⊆ and read as ‘is a

Using this symbol we can express subsets as follows:

A ⊆ B; which means Set A is a subset of Set B.

All Subsets of a Set

Example: Find all the subsets of set A = {1,2,3,4}

Solution: Given, A = {1,2,3,4}

In set theory, set A is considered as the superset of B, if all the elements

The superset relationship is represented using the symbol “⊃”. For

Consider another example,

X = {set of polygons}, Y = {set of irregular polygons}

Then X is the superset of Y (X⊃Y). In other words, we can say that Y is

For example, let us take four sets.

A = {a, b, c}, B = {a, b, c, d}, C = (a, b, c}, D = {a, b, e}

From the sets given above,

B is the proper superset of A, as B is not equal to A

C is a superset of set A, but the set C is not a proper superset of set A, as

i.e. P(A) = {S: S ⊂ 𝐴 }.

For Example. (i) if A = {a}, then P(A) = {∅, A}

(ii) If B = {1, 2} then P{B} = {∅,{1}, {2}, B}

Theorem 1. Every set a subset of itself.

Proof. Let A is any set. Since x ∈ A ⟹ x ∈ A, therefore A ⊂ 𝐴.

Theorem 2. Empty set is a subset of every set.