SYLLABUS

DKM22 – MEASURE THEORY AND COMPLEX ANALYSIS

Unit I: Lebesgue measure – Outer measure – Measurable sets and Lebesgue

measure – Measurable functions – Little wood’s three principles.

Unit II: Lebesgue integral – The Riemann integral – The lebesgue integral of a
bounded function over a set of finite measure – the integral of a non negative
function – the general lebesgue integral

Unit III : Complex numbers – Analytic functions – Elementary theory of power

series.

Unit IV : Cauchy’s Theorem – Cauchy’s integral formula – Singularities.

Unit V : Taylor’s Theorem – Maximum Principle – The calculus of Residues.

Texts: 1. Royden – Real Analysis Third Edition (PHI) Chapter 3 ( Excluding

section 3.4), Chapter 4(excluding section 4.5).

2. Ahlfors - Complex Analysis (Tata- McGraw Hill Third Edition) Chapter 1,

Chapter 2 (sections 1 and 2) and chapter 4 ( sections 1,2,3 and 5).

1
 Unit I

Measure Theory

We need the following definitions and results

Definition :
A Collection of subsets of X is called an algebra of sets or a
boolien algebra if,
i. A B whenever A,B
ii. is in whenever A

By Demorgan’s Law A B is in whenever A,B

Result 1:
Let be an algebra of subsets and {Ai} a sequence of sets in .Then
there is a sequence {Bi} of sets in , such that Bi Bj = for i j and ⋃
=⋃ .

Definition:
An algebra of sets is called a sigma algebra or a boral field, if
every union of a countable collection of sets in is again in .
By Demorgan’s Law, the intersection of countable of sets in is
again in

Result 2:
Given any collection δ of subsets of X, there is a smallest sigma
algebra that contains δ, (i.e) there is a sigma algebra containing δ such that if
ẞ is any sigma algebra containing δ such that ẞ.

Definition:
The collection ẞ of borel sets is the smallest sigma algebra which
contains all of the open sets.

2
 1.1 Lebesgue Measure

Definition:
The length (I) of the interval I is defined to be the difference of
end points of the interval, if I is bounded and if I is unbounded.

Definition:
A set function m that assigned to each sets E in some collection
of sets of real numbers a non-negative extended real numbers mE, called
measure of E.

Properties:
i. mE is defined for each set E of real numbers. (i.e)
ii. For an interval I ,
iii. If {En} is a sequence of disjoint sets (for which m is defined)
⋃ ∑
iv. m is translation invariant.
(i.e) If E is a set for which m is defined and if E+ y = {x+ y ; x E}
obtained by replacing each point x E by the point x+ y.
Then m (E+ y)= m(E).

Definition:
We say that m is a count ably additive measure. If it is a non-
negative extended real valued function whose domain of definition is a -
algebra of sets we have m ( En) = ∑ for each {En} of disjoint sets in

Properties:
Let m be a count ably additive measure defined for all sets in
Then we have the following properties.
i) m(E) ≥0, for all E

3
ii) If A,B and A C B ⇒ m(A) ≤ m(B)
Proof:
i) It follows from the definition .
ii) m(A B) = m(A) + m(B-A)
⇒ m(B) = m(A) + m(B-A)
⇒m(A)≤ m(B) [since m(B-A) ≥ 0]
This property is called monotonicity.
iii) Let the collection {En} be any sequence of sets in .
Then ⋃ ∑ . This property is called countably
subadditivity.
For,
Let {En} be a sequence of sets in By Result 1, there exists a { } of disjoint
sets in such that ⋃ =⋃ , where = En- (E1 E2 ……. En-1),
CEn

∑
∑
Observation:
If there is a set A such that mA < ∞. Then m = 0
For, any set A = A
m(A) = m(A ) = m(A) + m( )
⇒ m( ) = 0 [m(A) < ∞]
Example:

Let {
| |

4
 n is countable additive set function and translation invariant. It is defined for all
sets of real numbers. This measure is called counting measure.
Solution:
Let {En} be a sequence of disjoint sets in R. One of the sets, say, En is
infinite.
Then =| |
=∑ | |
=∑
If all the sets in {En}are finite and En if n m.
Then n =| |=∑ | |
=∑
⇒ n is countably additive.
Also, n(E+ y) = | |
=| |
= n(E)
⇒ n is translation invariant.

1.2 Outer Measure

Definition:
For each set A of real numbers. Consider a countable collection {In} of
open intervals that cover A. (i.e) Collections for which A In, and for each
such collections, consider the sum of the length of the interval in the collection.
Then the outer measure m* A to be the infimum of all such sums.

(i.e) m*A = ∑

Then the immediately the following is observed.

i) m*A ≥ 0
ii) m*A = mA , A
(i.e) m* = m/

5
 iii) Since m = 0 we have m* = 0
iv) Let , then m*A ≤ m*B
For, let
α = {{In}/A In }
β = {{In}/B In }

⇒β α
⇒ Inf ∑ α ≤ inf ∑ β
⇒ inf ∑ ≤ inf ∑

⇒ m*A ≤ m*B
v) m*({x}) = 0
For, {x} (x- , x+ ) = I

m*
⇒ m*{x} = 0 [since is arbitrary]
THEOREM :1
The outer measure of an interval is its length.
Proof:
First we consider the case of closed finite interval, say [a,b].
Now [a,b] ⊂ (a- , b+ )
m*[ ] = b-a+
Since is arbitrary, m*[ ] b-a
To prove: m [ ] b-a
Let [a,b] ⊂ ⋃
By Heine Borel theorem, there exists a finite sub collection I1,I2,……,Im
intervals such that I ⊂ ⋃ and since the sum of the length of the finite
collection is no greater than the sum of the length of the original collection and
hence it is enough toprove that ∑ Ik b-a for finite collections {In} that
cover [a, b].
Since a is contained in In, there must be one of the In that contains a.
Let this be the interval (a1,b1).
6
 Then we have a1< a < b1.If b1 < b, then b1 (a,b)
Since b1 ∉ (a1,b1), there exists an interval (a2,b2) in the collection {In} such
that b1 (a2,b2) (i.e) a2 < b1 < b2
Continue in this fashion , we obtain a sequence (a1,b1),(a2,b2),……,(ak,bk) from
the collection {In} such that ai < bi-1 < bi.
Since {In} is a finite collection, our process must terminate with some finite
interval (ak, bk).
(i.e) b (ak , bk) (or) ak < b < bk.
∑ ( Ik ) ∑ (ai,bi)
= (a1,b1)+……+ (ak,bk)
= b1- a1+………+bk- ak
= bk - (ak-bk-1) - (ak-1-bk-2) -……..- (a2-b1) - a 1
≥ bk-a1 [since ai < bi-1]
As a1 < a and b < bk
bk-a1 > b-a
∑ ( In ) > b-a
By taking inf we have inf ∑ ( In ) b-a
m*[a,b] b-a
If I is any finite interval then given > 0 there is a closed interval J ⊂ I such
(J) > (I) -
Now, (I) - < (J) = m*(J) ≤ m*(I) ≤ m*( )̅ = ( )̅ = (I).
(I) - < m*(I) ≤ (I)
If I is an infinite interval, then given any real number , there is a closed
interval J⊂ I with (J) = .
Hence, m*(I) ≥ m*(J) = (J) =
Since m*(I) ≥ , for each , we have m*(I) = = (I). Hence proved.

7
THEOREM:2
Let {An} be a countable collection of sets of real numbers. Then
m*( An) ≤ ∑ m*( An ) [This proposition is called count ably sub additivity
of m*].
Proof:
If one of the sets An has infinite outer measure then inequality holds trivially.
If m*An is finite then given there is a countable collection {In, i}i of
open intervals such that An⊂ ⋃ and ∑ (In, i) < m*An + 2-n (by definition
of m*).
Now the collection {In, i}n, i = ⋃ In, i}i is countable, being that union of the
countable number of countable collection and covers union of An.
m*( An) ≤ ∑ (In, i)
= ∑ ∑ (In, i )
≤∑ m*An + 2-n )
= ∑ m* An +
Since is arbitrary, m*( An) ≤ ∑ m*An.
COROLLORY:3
If A is countable, then m*A=0
Proof:
Given A is countable. Then A= ⋃ xn}
m*A = m*(⋃ xn})
≤∑ m*{xn}, as m*{xn}= 0
=0
m*A= 0 (as m*A ≥ 0)

COROLLORY:4
The set [0,1] is not countable.

8
Proof: m*[0,1] = [0,1]
Hence , [0,1] is not countable [by corollary :3]
Definition:
A set which is a countable union of closed sets is called .
We say that a set as if it is the intersection of countable collection of
open sets. Note that Complement of is and vise versa.

Theorem:5
Given any set A and any there is an open set O such that A⊂O and
m*O m*A + . There is a G such that A ⊂ G and m*A = m*G.
Proof:
Given , by the definition of m*, there is a countable collection {In} of
open intervals A⊂ In such that ∑ (In) < m*A + -------(1)
Let O = In ⇒ O is open
m*O = m* In
∑ m* In
= ∑ (In)
< m*A + ------- (2) (by (1)),

Let =

Then by (2), for all n there exists an open set Gn such that A ⊂ Gn and

m*Gn ≤ m*A+ --------(3)

Let G = Gn , then G is a set and A⊂G. ⇒ m*A ≤ m*G

Now A⊂ Gn n and Gn is open. Also m*G ≤ m*Gn ≤ m*A+ n (by(3))

⇒ m*G ≤ m*A. Hence, m*A = m*G.

LEMMA:6
If m*A=0 then m*(A B) = m*B

9
Proof:
B ⊆ A B ⇒ m*B ≤ m*(A B) (1)
By count ably sub additivity property, m*(A B) ≤ m*A +m*B
Given, m*A=0. Therefore, m*(A B) ≤ m*B (2)
From (1) and (2), m*(A B) = m*B

1.3 Measurable Sets and Lebesgue Measure

Definition:
A set E is said to be measurable if for each set A, we have

Remark:
(i) Since A= (A E) (A Ec)
⇒ m*(A) ≤ m*(A E) + m*(A Ec)
We have the following definition
E is measurable if for each A we have, m*A ≥ m*(A E) + m*(A Ec).
(ii) Since the definition of measurability the symmetric in E and Ec, we
have Ec is measurable whenever E is measurable. Clearly, and R are
measurable.

LEMMA:7
If m*E=0 then E is measurable.
Proof:
Let A be any set.
A E ⊂ E ⇒ m*(A E) ≤ m*E = 0
⇒ m*(A E) = 0
Also, A ∩ ⇒ (A∩ ) (A)
Therefore, (A) (A∩ )+0

10
 ⇒ (A) (A∩ )+ (A∩E) ⇒ E is measurable.

LEMMA:8
If and are measurable sets, then is measurable.
Proof:
Let A be any set. Since, is measurable we have,
( ) ( ) (1)
Since A ∩ ( ) = (A ∩ ) U (A ∩ ∩ ), we have
(A ∩ ( )) (A ∩ ) + (A ∩ ∩ )
⇒ (A∩( )) + (A∩( ∩ )) (A ∩ ) + (A ∩ ∩ )
+ (A ∩ ( ∩ ))
= (A ∩ ) + (A ∩ ) ( by (1) )
(A) [ since is measurable]
Therefore, (A) ((A ∩ ( )) + (A ∩ )
⇒ is measurable.

LEMMA :9
A family ɱ of measurable sets is an algebra of sets.
Proof:
Let , ɱ
⇒ is measurable ( by lemma 8)
⇒ ɱ
Also, E ɱ ⇒ ɱ (by definition)
Therefore, ɱ is an algebra.

LEMMA : 10
Let A be any set and , , ……., be a finite sequence of disjoint
measurable sets . Then (A ∩ (⋃ )) = ∑
Proof:
We prove the lemma by induction on n.
The result is clearly true when n = 1
Assume that the result is true if we have n-1 sets
Since = φ , i j we have , (A ∩ (⋃ )) =
(A ∩ (⋃ )) = (A ∩ (⋃ ))
11
Since is measurable we have,
(A ∩ (⋃ )) = ( A ∩ (⋃ ) + ( A ∩ (⋃ ) )
= + (( A ∩ (⋃ ) )
= + (A ∩ (⋃ )
= +∑ ( by induction hyp)
=∑
The theorem is true for all values of n .

THEOREM:11
The collection ɱ of measurable sets is a – algebra, that is,
complement of a measurable set is measurable , union of a countable collection
of measurable sets is measurable. Moreover, every set with outer measure zero is
measurable.
Proof:
By lemma 9, ɱ is an algebra of sets.

Claim: ɱ is a – algebra.
It is enough to prove, E = ⋃ , ɱ ⇒E ɱ
Let E = ⋃ , ɱ
By a result, we have E = ⋃ and ∩ = φ, and
⋃ =⋃
Let A be any set. Let = ⋃
⇒ is measurable (i.e) Є ɱ
Now , = ⋃ ⋃ E
= E , for all n = ⊃ , for all n
Since is measurable ,
= +
+ )
= ⋃
=∑ (by lemma 10)
Therefore , ∑ + )
This is true for every n and L.H.S is independent of n.
We have , ∑ + )
[ ]+ ) ( by countably
subadditive property)

12
 + )
⇒ E is measurable
Therefore , ɱ is –algebra
For, Intersection of countable collection of sets is measurable.
Let Єɱ ⇒ Єɱ ⇒⋃ Єɱ
⇒ ⋃ Єɱ ⇒∩ Єɱ
Also by lemma 7, every set with outer measure zero is measurable.

LEMMA: 12
Open interval (a, ) is measurable.
Proof:
Let A be any set
Let = (a, ) , = A ∩ (- ,a]
To prove (a, ) is measurable.
Claim: +
If = then there is nothing to prove
If ,then given Є 0 , there exists a countable collection of open
intervals { } which covers A and for which ∑ ) + Є → (1)
by the definition of outer measure .
Let = (a, ) and = ∩ (- ]
Then and are intervals (or) empty.
Now , )= + = + → (2)
Since ϹU ,
( U ∑
Similarly, U ,
( U ∑
Therefore, ( + ( ∑ +∑
∑ +∑ )
=∑ ) (by (2))
( + (by (1))
Since, is arbitrary , ( + ( ( ).

THEOREM:13
Every Borel set is measurable. In particular , each open set and each closed
set is measurable.

13
Proof:
Let Ɓ be the family of Borel sets.
By definition ,Ɓ is the smallest – algebra containing all the open sets →(1)
Also ɱ, the collection of all measurable sets is – algebra.
Since (a, ) Є ɱ, ⇒ (a, )c Є ɱ , for all a ⇒ (- ] Є ɱ , for all a
Now , (- ,b) = ⋃ ( +

⇒⋃ ( + Є ɱ [as ɱ is a – algebra]
⇒ (- , b) Є ɱ
Now , (a, b) = (- ,b) ∩ (a, )
Therefore , every open interval is measurable
Since each open set is the countable union of open intervals , every open set
belongs to ɱ. Then every closed set is measurable. Therefore, ɱ is a -algebra
containing all the open sets.Therefore, [by 1]. Hence the result.

Remark:
If E is measurable set, we define the lebesgue measure mE be the outer
measure of E. (ie.) m= m*/ ɱ. It is means that the domain of m is ɱ and the
domain of m* is (R). (ie.) If E is measurable set, mE= m*E.

THEOREM :14
Let {Ei} be a sequence of measurable sets, then m( Ei) ≤ ∑mEi. If the sets
{En} are pairwise disjoint then m( Ei) = ∑ m .

Proof:
If {Ei} is a finite sequence of disjoint measurable sets, then by lemma 10
(by taking A = R) we have, m* ⋃ =∑ m* .
m ⋃ =∑ m and so m is finitely additive.
Let {En} be an infinite sequence of pair wise disjoint measurable sets.
As ⋃ ⊃ ⋃
m ⋃ ≥ m(⋃ )
=∑ is true for every n.
Since Left hand side of inequality is independent of n, We have,
m ⋃ ≥ ∑ .

14
The reverse inequality follows from the countable sub additive property,
m ⋃ ≤∑ i. Therefore, m ⋃ =∑ .

Theorem:15
Let {En} be an infinite decreasing sequence of measurable sets, that is with
En+1 En , for each n. Let be finite and m (⋂ i) = i .
Proof:
Let E = ⋂ i . Let .
Then E1 - E = ⋃ i and the set Fi are pair wise disjoint.
Hence m(E1 E) = m(⋃ i) = ∑ i =∑ (Ei Ei+1)  (1)
Since E E1, E1 = E (E1 E) and mE1 = mE + m(E1 E)  (2)
Similarly, Since Ei+1 Ei Ei = Ei+1 (Ei Ei+1) and
mEi = mEi+1 m(Ei Ei+1). Also mE ≤ mE1< ∞
 m(E1-E) = mE1- mE [by (1)].
And m(Ei Ei+1) = mEi - mEi+1 [since, Ei+1 Ei E1 and mEi+1 ≤ mE1 < ∞ ]
Therefore, mE1 - mE = m(E1 E)
=∑ (Ei Ei+1) [by (1)]
=∑ Ei mEi+1)
= i ∑ Ei mEi+1)
= i (mE1 - mEn)
mE1- mE = mE1- i mEn
mE = i mEn [since, mE1< ∞]
m(⋂ i) = i .

THEOREM:16
Let E be the given set, then the following are equivalent
i) E is measurable.
ii) given there is an open set O contains E with m*(O-E)
iii) given there is a closed set F contained in E with m*(E-F)
iv) there is a G in with E and m*(G-E)=0
v) there is a F in with F E such that m*(E-F) = 0
m*E is finite ,then the above statements are equivalent to (vi).
vi) given ,there is a finite union of open intervals such that m*(U E)
Proof:
We prove the theorem as follows:
(i) => (ii)=> (iv)=>(i)
15
 (ii) => (iii)=> (v)=>(i) and (i) (vi)
step I:
(i) => (ii)
Given E is measurable.
Case (i):
Suppose m*E = mE < with m*(E-Fn) <
 Given , there exists a collection {In} of open intervals such that
E In and ∑ n) ≤ m*E + .
Let O = In.Then mO = m( In) ≤ ∑ n = ∑ n)
mO ≤ m*E +  (1)
Now E In = O
=> m*O = m*E + m*(O-E)
=> m*(O-E) = m*O - m*E [since mE is finite]
=> m*(O-E) ≤ m*E + - m*E
= [since mO = m*O and O is measurable]
=> m*(O-E) <
Case(ii):
Let m*E is infinite
Let E ⋃ n, where {In} is a collection of intervals of finite length.
Define En=E In => E= ⋃ n
As E, In are measurable , E In is measurable. ⇒ En is measurable for all n.
Now En In => m*En < m*In < ⇒ m*En is finite for all n.
By case(i), for given , there exists an open set Gn such that
En Gn, n = 1,2,3,…. and m*(Gn – En) < /2n
Let O = ⋃ n => O is open
Consider O - E = ⋃ n- ⋃ n
⋃ n- En)
m*(O - E) ≤ m*(⋃ n- En))
≤ ∑ *(Gn- En)
< ∑ = => m*(O - E) <
(ii) => (iv):
Given , there exists an open set O with E O such that m*(O-E) <
For each n, taking = 1/n we get an open set On such that E On with
m*(On-E) < 1/n  (1)
Let G= n, then G
Since E On, => G-E On-E,
=> m*(G-E) ≤ m*( On-E) < 1/n, n
16
 Therefore, m*(G-E) = 0
(iv)=>(i):
There is a G in with E G and m*(G-E) = 0
Since each open set is measurable, each set is measurable.
Therefore, G is measurable.
Also m*(G-E)=0,then by lemma 7, G-E is measurable.
But E = G-(G-E) and hence E is measurable.

Step II:
(ii) =>(iii):
Now (i) => (ii) follows from step I
⇒ E is measurable ⇒ Ec is measurable.
Given , there exist an open set O ⊃ Ec such that m*(O - Ec)<
 Oc E
Since O is open, F = Oc is closed.
Now F E and m*(E-F) = m*(E-Oc) = m*(O-Ec) <

(iii)=> (v):
Given , there exists a closed set C such that C E and m*(E- C) <
For each n, there exists a closed set Fn such that Fn E with m*(E- Fn) <

Let F = Fn. F is a Fζ set

Now, Fn E for all n. Then F E
Since Fn F => E-F E - Fn

m*(E-F) m*(E- Fn) < , for all n ⇒ m*(E-F) = 0

(v) => (i) :

Given, there exists a Fζ - set F E such that m*(E-F)=0
Since Fζ set is measurable, F is measurable.
Also, m*(E-F)=0, by lemma7, E-F is measurable.
But, E = (E-F) F. Since F is measurable, E is measurable
Step III:
(i)(vi)

17
 To Prove (i) => (vi)
Suppose E is measurable.

Given there exists an open set O ⊃ E such that m*(O-E) ---(1) (by (ii))

As mE is finite ((i.e) mE < ) and m(O) = mE + m(O-E) <

As O is an open set, O is the disjoint union of open intervals Ii, (i.e) O=⋃ i,
Ii Ij= (i j). Therefore, m(O) = m(⋃ i) = ∑ Ii = ∑ Ii )
Now, m(O) < ⇒ There exists n such that ∑ Ii) < /2
Let U = ⋃ i

E U = (E-U) (U-E)
E U (O-U) (O-E) [... E O & O U ]
m*(E U) m*(O-U)+m*(O-E) [... E O & O U]

< + = [O-U=∑ i- ∑ i =∑ i

m*(O-U) ∑ , by definition]
Therefore, m*(E U) <
(vi) => (i):
Conversely, suppose given there exists intervals {Ii} such that
U=⋃ *
i and m (E U) <

Let be given and also given m*E <

Therefore, there exists an open set O ⊃ E such that m*O < m*E + /3-----(1)
[by Theorem 5]
Also given there exists intervals {Ii} such that m*(E U) < /3 ----------(2),

where U= ⋃ i

Let J = U O . Then J is open

Also, J E = (J E) – (J E)
Now, J E = (U O) E = U (O E) = U E [... E O]
J E U E ⇒J E U E.

18
 m*(J E) m*(U E) < /3 ---------- (3) (by (2))
O E (O J) (J E) [as O-E (O-J) (J-E)
and E-O (E-J) (J-O)]
m*(O E) m*(O J) + m*(J E)
But O J=O-J [ O ⊃ J]
Therefore, m*(O J) = m* (O - J)
= m*(O) – m*(J) [ O, J open => O, J are measurable]
(4)
But E J (E-J) ⇒ E J (E J)

m*E m*J + m*(E J) ⇒ m*E < m*J + (by (3) ) (5)

Now, m*(O E) m*(O J) + m*(J E)

= m*(O) - m*(J) + m*(J E) (by(4))

m*E + – m*J + ( by(1) & (3))

m*J + – m*J + (by(5))

=
m*(O E) <
Since E O => O E=O-E
m*(O - E) < . This proves (ii)
By step II, (ii) => (i) (i.e.) E is measurable.

1.4 Measurable Functions

LEMMA:17
Let f be an extended real valued function whose domain is measurable. Then
the following statements are equivalent
(i) For each real number , the set {x/ f(x)> } is measurable.
(ii) For each real number , the set {x/ f(x) } is measurable.

19
 (iii) For each real number , the set {x/ f(x)< } is measurable.
(iv) For each real number , the set {x/ f(x) } is measurable.
These statements imply (v)
(v) For each extended real number , the set {x/f(x)= } is
measurable.
Proof:
Let D = the domain of f . D is measurable (given)

Given : {x/f(x)> } is measurable for all

Now, {x/f(x) } = D - {x / f(x) > }. Since, the difference between two
measurable sets is measurable we have {x / f(x) } is measurable.

Suppose {x / f(x) < } is measurable.

Now . Since D is measurable, we have

is measurable.

Suppose is measurable.

Now

Since the difference of two measurable sets is measurable

Therefore is measurable.

Hence, we have proved ⇒ ,

Similarly, we can prove ⇒ .

Since ⋂ and the

intersection of a sequence of measurable sets is measurable

Therefore , is measurable.

20
 ⇒

⋃ and the union of a

sequence of measurable sets is measurable. Therefore, is
measurable.

Hence all the above four statements are equivalent.

If is real, ⋂

Therefore, together implies

Since ⋂

By ,

⇒ ⋂

By ,

⇒ ⋂

Hence ⇒ .

Definition:

An extended real valued function f is said to be Lebesgue measurable, if its

domain is measurable and it satisfies one of the first four statements of the
above proposition.

Note:

A continuous function on measurable set is measurable.

For any real , is open.

⇒ is measurable for all .

21
 Therefore, ⇒

Recall a real valued function defined on an interval [ ]

is called a step function if there is a partition

such that for all , the function assumes only one value in the interval
( ).

If is a step function, then is an interval (or) union of

intervals.

⇒ is open for all real . ⇒ is measurable.

⇒ is measurable.

If is a measurable function and is the measurable subset of the

domain . Then is also measurable.

LEMMA:18

Let be a constant and let be two measurable real valued

functions defined on the same domain. Then the function

are also measurable.

Proof:

Let be any real number.

Now,

Since is measurable, is measurable.

Therefore, Left hand side is measurable.

⇒ is measurable.

Claim :

If then clearly ⇒ is measurable.

22
Suppose .

Then ⁄

Since is measurable, R.H.S is measurable for all real .

⇒ Left h nd side is measurable for all real ⇒ is measurable.

i Claim :

Suppose . Then

There exist a rational number r such that

Therefore ⋃ ⋂

Since are measurable functions we have, and

are measurable. Therefore, Right hand side is measurable.

Hence is measurable. ⇒ is measurable.

Since – is measurable by (ii)

Therefore, is measurable.

v Consider for all

Since is measurable, we have is measurable, for all .

Therefore, . ⇒ is measurable.

Now are measurable ⇒ is measurable.

⇒ re e sur e ⇒ is measurable.

⇒ [ ] is e sur e ⇒ is measurable.

23
LEMMA :19

Let be a sequence of measurable function (with the same domain

Of definition) then the functions sup inf ,

sup inf
f f f f re e sur e
n n
Proof:

Let sup

Now ⋃

Since each is measurable, R.H.S is measurable.

Therefore, , - Therefore, h is measurable.

Define .

Then ⋃

Since is measurable, and . We have,

⋃ ⇒ is measurable.

⇒ is measurable.

Similarly, we can prove inf and are measurable.

Since , Therefore, is measurable.

Similarly, is also measurable.

Definition:

A property is said to hold almost everywhere. If the set of points where it

fails to hold is a set of measure zero.

LEMMA 20

If is a measurable function, and is almost everywhere, then is

measurable.

24
Proof: Let

Since is almost everywhere, we have

Now ,

Since f is measurable, is measurable.

Since the second and third sets are subsets of E and

⇒ Both the sets h ve e sure zero and hence they are measurable.

Since , we have is measurable.

⇒ is measurable.

LEMMA :21

Let be a measurable function defined on [a,b] and assume that

takes the value only on a set of measure zero. Then given we can
find a step function and a continuous function such that
and

Proof:

To prove this we required following 4 lemmas.

Lemma:1
Given a measurable function f on [a , b] that takes the values only
on a set of measure zero and given there is an integer M such that
except on a set of measure less than ⁄ .

Proof:
Suppose for all M such that ⁄ .

Let ⇒ ⁄

Also, ⊃ ⊃ ⊃ and let ⋂

⇒ ⇒
25
By Theorem 15, i ⁄ . This is a contradiction.

Therefore given , there is M such that ⁄

Lemma:2
Let f be a measurable function on [a, b], given and M, there is a
simple function except where If
, then we may take so that .
Proof:

Given , there exists M with such that for some n.

Let

Define

Then on

Therefore, [ ] except where

If
Then by the above construction, there exists a simple function
.
Hence the lemma.

Lemma :3
Given a simple function [ ] , there is a step function g on [a , b]
such that except on a set of measure less than ⁄ .

Proof:
Let be a simple function and it assumes finite number of values
. Let

26
Then by, theorem 16 (vi)
Since is measurable, (by definition of simple function) for every
there is a finite union of intervals and

Vi = U1i ⋃ U2i ⋃…….⋃ Uni such that m*(Ei ∆ Vi) < ,1 i n.

Define g(x) = ci , x Vi
If x Ei ⋂Vi , then g(x) = φ(x). If g(x) φ(x), then x Ei ∆ Vi , for some i
Therefore, { x / g(x) φ(x)} ⋃ i i}
m*{ x / g(x) φ(x) } ∑ i i

∑ =

Therefore, g(x) = φ(x), except on a set of measure less than .

If m φ M then we take g so that m g M.

Definition:
if x
The function is defined by (x) = , is called the characteristic
if x
function of E.
A linear combination φ(x) = ∑ (x) is called a simple function if the
sets Ei are measurable.

Lemma :4
Given a step function g on [a,b] there is a continuous function h such that
g(x) = h(x) except on a set of measurable < . If m g M, then we may take h
so that m h M.

Proof:
Given a step function g on [a,b] such that g(x) = ci , x [xi-1 , xi] for some
subdivision of [a , b] and a = x0 ≤ x1≤ x2≤ ……..≤ xn= b.

27
Define

* +
h(x) = [ ]
( ) [ ]
{

Then h(x) = g(x), if x ⋃ [ ]

Now, m (⋃ [ ] ∑ [ ]

Therefore, h(x) = g(x) except on a set of measure less than .

Main proof:
Since f takes the value only on a set of measure zero, we may assume
that m f M.
Given 0, Let f be a measurable function. Then by lemma 2, there exists a
simple function with m M such that | | .

By lemma 3, there exists a step function g with m g M on [a,b] such that

| | .

⇒| | | | | |

+ = .

By lemma 4, there exists a continuous function h with m h M such that

| | .

⇒| | | | | | + = .

Observation:
is measurable iff A is measurable .

28
Proof:
Let be any real number .

{x / (x) > } = { ………….(1)

⇒ {x / (x) > } is measurable for all

Conversely, is measurable. ⇒ {x / (x) > } is measurable.
⇒ A is measurable. (by (1))

1.5 : Little Wood’s 3 principle

i) Every (measurable) set is nearly a finite union of intervals.
ii) Every (measurable) function is nearly continuous.
iii) Every (measurable) convergence sequence of function is nearly
uniformly convergence.

The following proposition gives one version of the third principle.

PROPOSITION :22
Let E be a measurable set of finite measure and {fn} be a sequence of
measurable functions defined on E. Let f be a real valued function such that for
each x in E we have fn(x) converges to f(x). Then given and δ there
is a measurable set A subset of E with m(A) and integer N such that x
and n , | |
Proof:
Let be given. Let Gn = {x /| | }
Let EN = ⋃∞ = {x /| | for some n N}
Since fn is measurable and fn converges to f point wise.
We have, f is measurable.
Since fn , f are measurable we have, Gn and hence EN are measurable.
Also we have EN+1⊆ EN

29
Since E1 E and m(E) we have m(E1) .
Then by Theorem 15, we have m(⋂ )= i .
By the definition of EN, for all x E , there exists N such that x EN as
⋂ = φ.
⇒ 0 = mφ = i .
⇒ there exist N such that
⇒
Take then
Now ,
.

PROPOSITION :23
Let E be a measurable set of finite measure and be a sequence of
measurable functions that converge to a real valued function f almost
everywhere on E . Then given there is a set A subset of E with
and an N such that for all , and all
Proof:
Given pointwise almost everywhere on E.
⇒ There exists

By proposition 22, Given there exists and a set

and such that every , .
Let
⇒ ⇒
⇒ and

30
 UNIT-II

Lebesgue Integral
2.1 The Riemann integral
Let f be a bounded real valued function defined on the interval [a, b]
and let be a subdivision of [a , b]. Then for each
subdivision we can define the sums ∑ and
∑ , where and
[ ]

. Then we define upper Riemann integral of f by

[ ]
̅
∫ inf , where the inf is taken over all possible sub divisions
of [a,b].

Similarly, we can define lower Riemann integral and ∫ sup

The upper integral is always at least as large as the lower integral and
if the two are equal, we say that f is Riemann integrable and call this
common value the Riemann integrable of f and we shall denote it by

By a step function we mean a function which has the form

x , for some subdivision of [a ,b] and for some set of
constant . Under this definition we have ∫ x dx ∑
̅
with this definition we have, ∫ inf ∫ x dx for all step
functions x f x .
sup
Similarly, ∫ x f x ∫ x dx

Example:
0 If x is irrational
Show that if 1 if x is rational

31
Then f is not Riemann integrable.
Soln:
Let
inf and sup [ ]
This is true for every subdivision of [a , b].
∑ ∑ [ ]

and ∑
̅
∫ inf and ∫ sup

̅
⇒ ∫ ∫ ⇒ is not Riemann integrable.

2.2: The Lebesgue Integral of a bounded function over a set of

finite measure.
The function defined by,

1 if

0 if

is called the characteristic function of E.

A linear combination ∑ is called a simple function
if the sets are measurable .
This representation for is not unique. However note that the function
is simple iff it measurable and assumes only finite numbers of values.

32
 If is a simple function and { } the set of non-zero values
of , then ∑ where , this representation of
is called the Canonical representation.
If vanishes outside a set of finite measure, we define the integral of
by, ∫ ∑ where has the canonical representation

∫∑
Note:
Some times we denote it by ∫ . If E is any measurable set then we define
∫ ∫

LEMMA :1
Let ∑ with for .

Suppose each set is a measurable set of finite measure , then

∫ ∑ .
Proof:
Consider
⋃
∑

⇒ ∑

⇒ ∑

∫ ∑

PROPOSITION:2
Let and be two simple functions which vanish out side a set of
finite measure then ∫ ∫ ∫ and if almost
everywhere, then ∫ ∫ .
Proof:
Let {Ai} and {Bi} be the sets which occur in the canonical representation
of
33
 Let and be the sets where are zero. Then the set
obtained by taking all the intersections from a finite disjoint Collection
of measurable sets.
And we may write ∑ and ∑ .

Then by lemma 1, ∫ ∑ and ∫ ∑

Now ∑

Then by lemma 1 , ∫ ∑

∫ ∑ ∑

∑ ∑ .

∫ ∫

Note that ∫ ∫ ∫
almost everywhere, then almost everywhere.
⇒ on F and
Now from a disjoint collection of measurable sets.
Now by definition,

∫ ∑

∑ where ,

Now
And are disjoint.
Since m is additive,
Now ∫ ∫ ∫
= ∑ ∑ [ ]
∑ ∑

= ∑ (as )

∫ ∫ ( as )
34
 ⇒∫ ∫

PROPOSITION:3
Let f be defined and bounded on a measurable sets E with . In
order that, ∫ ∫ for all simple function
it is necessary and sufficient that f be measurable.
Proof:
Let f be bounded by M. Suppose f is measurable.

Let , -

Then are measurable and disjoint.

⋃ ⇒ ∑ -------------------(1)
Define the simple function

⇒ ∫ ∫

= ∑ -------------(2)

Similarly, ∫ ∫ = ∑ --------------------(3)

From (1) and (2) ,

⇒ ∫ ∫

∑ ∑

35
 [ ]

Since n is arbitrary and mE < , we have

⇒ ∫ ∫

Conversely,

Suppose ∫ ∫

Given n, there exist a simple functions such that

(i)
(ii) ∫ x ∫
Define
sup
Then are measurable.
Also,
⇒ sup inf
⇒

Let

Now Claim that

Let

⇒ on

⇒ ∫ [∫ ]

36
 ⇒ ∫ ∫

⇒ ∫ ∫

Since n is arbitrary, ⇒ ( )

As ⋃ ⇒

⇒ except on a set of measure zero.

⇒ almost everywhere.
⇒ almost everywhere on E,
⇒ f is e sur e

If is a bounded measurable function defined on a measurable set E with

. We define the (lebesgue) integral of f over E by,

∫ ∫

Note:
(i) We write the integral as ∫ .
(ii) If [ ] we write ∫ instead of ∫[ ]
.
(iii) If f is a bounded measurable function which vanishes outside a set
E of finite measure, we write ∫ for ∫ .
(iv) ∫ is the same as ∫ .

PROPOSITION :4
Let f be a bounded function defined on [a , b]. If f is Riemann integrable
on [a ,b], then it is measurable ∫ ∫ .

Proof:
Since the step function is also a simple function, we have
37
 ∫ ∫ ∫

Since f is Riemann integerable we have, ∫ ∫

⇒ ∫ ∫

⇒ f is measurable (By proposition 3)

Also from the above relation we have, ∫ ∫ .

Therefore, ∫ ∫ .

PROPOSITION:5
If are bounded measurable functions defined on a set E of finite
measure, then
i. ∫ = a∫ + b∫
ii. If f = g almost everywhere then ∫ =∫
iii. If f ∫ and hence |∫ | ∫| |
g almost everywhere then ∫
iv. If A f B then AmE ∫ BmE
v. If A & B are disjoint measurable sets of finite measure then
∫ =∫ +∫

Proof: Suppose a 0

∫ = inf ∫ [ f a af ]

= a inf ∫

∫ = a∫ ⟶(1)

Suppose a 0

∫ = inf ∫ [ f aϕ af , (a 0) ]

38
 = a sup ∫

= a inf ∫ ( by proposition 3)

∫ = a∫ ⟶(2)

From (1) & (2), ∫ = a∫ ⟶(3)

If and are simple functions such that 1 f and 2 g.

Then 1+ 2 is a simple function and + f+g

∫ ∫

=∫ +∫

Now by taking infimum on R.H.S over 1 f and 2 g.

Then we have ∫ inf ∫ + inf ∫

=∫ +∫ ⟶(4)

On the other hand, if and are simple functions such that

ϕ1 f and ϕ2 g.
Then ϕ1+ϕ2 is a simple function and + f+g

∫ ∫ =∫ +∫

Now by taking sup on R.H.S over ϕ1 f and ϕ2 g

Then ∫ sup ∫ + sup ∫

=∫ +∫ ⟶(5)

From (4) & (5), we have ∫ =∫ +∫ ⟶(6)

i) ∫ =∫ +∫ (by (6))

= ∫ + ∫ (by (3))

ii) Given f = g almost everywhere

⇒ f - g = 0 almost everywhere
f-g⇒ 0 almost everywhere

39
 ⇒∫ 0 (by proposition 2)

Taking infimum we have, inf ∫ 0

⇒∫ 0

Similarly, we can prove ∫ 0

∫ =0

⇒∫ -∫ =0 (by (i))

⇒∫ =∫

iii) Suppose f g almost everywhere

⇒f-g 0 almost everywhere
ϕ f-g ⇒ϕ 0 almost everywhere ⇒ ∫ 0

⇒ sup ∫ 0

⇒∫ 0 ⇒∫ -∫ 0 ⇒∫ ∫

Since f | | and -f | |
⇒∫ ∫ | | and ∫ ∫| |

⇒∫ ∫ | | and -∫ ∫| |

⇒ |∫ | ∫| |

iv) Suppose A f B
⇒∫ ∫ ∫

⇒ A mE ∫ B mE [ ∫ = A∫ = A mE ]

v) Suppose A & B are disjoint measurable sets of finite measure

Now, ={

Since A & B are disjoint measurable sets, then we have

= +

40
 ∫ =∫ =∫ =∫ =∫ +∫

Therefore, ∫ =∫ +∫ .

PROPOSITON:6 [ Bounded Convergence Theorem ]

Let be a seqence of measurable functions defined on a set E of finite
measure and Suppose that there is a real number M such that | | M for
all n, for all x and f = i for all x E then ∫ = i .

Proof :

Given 0, there exists N 0 and a measurable set A B with mA

such that for all n N and x E - A, | | ⟶(1) (by
proposition 22 of unit I ) .

Now, |∫ ∫ | = |∫ |

∫| |

=∫| |+ ∫ | |

∫ +∫

= 2M mA + m 2M + = 𝜺

|∫ ∫ | 𝜺 for all n N. Therefore, ∫ = i ∫

Definition:
Let f be a non-negative measurable function defined on a
measurable set E, we define ∫ = sup ∫ , where h is a bounded
measurable function such that m ⁄ is finite.
[ (ie) h vanishes outside a set of finite measure ]
PROPOSITION:7
A bounded function f on [a,b] is Riemann integrable iff the set of
points at which f is discontinuous has measure zero.

41
PROPOSITION:8
If f and g are non-negative measurable functions then
i. ∫ = c∫ ,c 0
ii. ∫ =∫ +∫
iii. If f g almost everywhere , then ∫ ∫

Proof: i) Let f be a non-negative measurable function and c 0. For every

bounded measurable function h, h f ⇒ ch cf and cf ⇒ f.

Now ∫ = sup ∫ = sup ∫

= sup ∫ = c sup ∫ = c∫

ii) Let h and k be bounded measurable functions vanishing outside the set of
finite measure.
h f, k g ⇒ h+k f+g ⇒ ∫ ∫

⇒∫ +∫ ∫

Taking supremum on L.H.S over h f and k g then

⇒ sup ∫ + sup ∫ ∫

⇒∫ +∫ ∫ ⟶(1)

Let be a bounded measurable functions which vanishes outside a set of finite

measure and f + g. Define h = min and k = -h
Therefore, k = - h is defined at all points of its domain. ( since l is bounded, h
is bounded )
By definition, h f and also, f +g
⇒h +k = f +g
Therefore, 0 k g [as h = min (f, )]
Moreover, h, k ⇒ h, k are bounded measurable functions and they vanish
outside the set of finite measure.
h f and k g

42
 ⇒∫ =∫

⇒∫ =∫ +∫

⇒∫ ∫ +∫

Taking supremum on f + g, ⇒ ∫ ∫ +∫ ⟶(2)

From (1) &(2)

∫ ∫ +∫

iii) Suppose f g almost everywhere

Let h be a bounded measurable function which vanishes outside the set of finite
measure and h f - g
⇒h 0 almost everywhere [ f g a.e ⇒ h = f - g 0 a.e]
⇒∫ 0

By taking supremum we have, sup ∫ 0

⇒∫ 0

Assume ∫ . [ suppose ∫ = , then ∫ ∫ ]

Adding ∫ on both sides,

⇒∫ +∫ ∫

⇒∫ ∫

⇒∫ ∫

THEOREM:9. [ Fatou’s Lemma]

If {fn} is a sequence of non-negative measurable functions and fn(x) f(x)
almost everywhere on a set E, then ∫ lim ∫ .

Proof:
Without loss of generality we may assume that, the convergence is
everywhere, since the integrals over the sets of measure zero are zero.

43
 Let h be a bounded measurable function which is not greater than f and
which vanishes outside a set Eʹ of finite measure.
Define a function hn = min{h(x), fn(x)}
Then hn is bounded by the bound for h and vanishes outside Eʹ.
And also hn (x) h(x). Then by proposition 6 (Bounded convergence theorem),
we have ∫ ∫ ∫

Now taking sup, we have ∫ ∫ ∫ ∫ .

THEOREM : 10.[ Monotone Convergence Theorem]

Let {fn} be an increasing sequence of non-negative measurable functions and
let almost everywhere then ∫ ∫ .

Proof:
By Fatou’s Lemma we have, ∫ ∫ . (1)

But for each m we have, and also ∫ ∫

⇒ ∫ ∫ . (2)

From (1) & (2) ∫ ∫ ∫ ∫

⇒ ∫ ∫ ∫

⇒ ∫ ∫ .

COROLLARY : 11
Let {un} be a sequence of non-negative measurable functions and let f = ∑ un.
Then ∫ ∑ ∫ .
Proof:
∑ . Let ∑

44
Since, is an increasing sequence of non-negative
measurable functions and
By Monotone Convergence theorem, ∫
= i ∞ ∑ uk
= i ∑ uk = ∑ uk
∞

PROPOSITION :12
Let f be a non-negative function and {Ei} a disjoint sequence of
measurable sets. Let ⋃ . Then ∫ f ∑ ∫ f

Proof:
Let

Since, {Ei} are disjoint sequence of measurable sets and E=⋃ i ,we have
=∑ ⇒ ∑ = ∑ ui

By corollary 11, ∫ ∑ ∫
∫ ∑ ∫

∫ ∑ ∫

Definition :
A non-negative measurable function f is called integrable over the
measurable set E if ∫ f

PROPOSITION:13.
Let f and g be two non-negative measurable functions. If f is integrable
over E and g(x) < f(x) on E. Then g is also integrable on E and
∫ ∫ ∫ .

Proof:
∫ ∫

∫ ∫ ⟶ (1) [ ]

Since f integrable, ∫ ⇒∫ ∫ ⇒∫

45
⇒ g is integrable.
(1) ⇒ ∫ ∫ ∫ [ ∫ ]

PROSITION:14
Let f be a non-negative function which is integrable over a set E. Then
given > 0, there exists a > 0 such that for every set A subset of E with
mA < δ we have ∫

Proof:
Case (i):
Suppose f is bounded. Let be given.
If A E such that mA< δ then ∣ ∫ ∣ ∫ ∣ ∣

≤ ∫ = M . m(A) < M . δ = ε

⇒∣∫ ∣

Case(ii):
Given f, Define = min {f , n}
⇒ ≤ n ⇒ each is bounded and i
Also { } is an increasing sequence of measurable functions.
By Monotone Convergence Theorem, we have ∫ = i ∫

Let be given, there exists N such that ∫ >∫ -

Now, ∫ =∫ ∫ < . Let δ = .

If mA < δ then ∫ =∫

=∫ ∫

= + N mA [ min(f, N)]

∫ < +Nδ= +N =ε

46
 ⇒∫

Theorem: 15
Let { be a sequence of non-negative measurable functions which
converges to f and suppose for all n then ∫ = i ∫
Proof:
Let i ∫
By Fatou’s lemma, ∫ ∫ -----------(1)

Since we have ∫ ∫ , for all n.

⇒ ∫ ∫ ------------(2)

From (1) and (2), ∫ ∫ ∫

But ∫ ∫

⇒ ∫ ∫ ∫
⇒ ∫ exists and ∫ ∫

Example :
The Monotone Convergence theorem need not hold for decreasing sequence of
.
Soln: Consider the function [ ]

Then ∫ =∫ [ ] =m[ = , for all n.

Also is decreasing to zero function and so f = 0 = i
⇒∫ =0. But ∫ ≠ lim ∫ .

2.3 General Lebesgue Integral

Definition:
By the positive part of a function f, we mean the function =fvo
47
 (ie) = max {f(x), 0}
Similarly, we define the negative part by =-fvo
(ie) = max {-f(x), 0} or = - min {f(x), 0}
If f is measurable and so and are measurable.
We have and ∣ ∣ .

Definition:
A measurable function is said to be integrable over E if and
are both integrable over E and we define ∫ ∫ ∫ .

PROPOSITION:16
Let f and g be integrable over E, then
(i) the function cf is integrable over E and ∫ ∫
(ii) the function f+g is integrable over E and ∫ ∫ ∫
(iii) If f ≤ g is a.e , then ∫ ∫
(iv) If A and B are disjoint measurable sets contained in E, then
∫⋃ ∫ ∫

Proof:
(i) Suppose c ≥ 0
Then, cf = c
=
and are non-negative integrable functions.
⇒ cf is integrable, when c ≥ 0

∫ ∫

=∫ ∫ [By Definition]

= ∫ ∫

48
 = [∫ ∫ ]

∫ = ∫

Suppose c < 0, Let c = -d, d>0

Now, cf = (-d)f
cf = (-d) ( )=-d = -d
and d are non-negative integrable functions over E.
⇒ cf is integrable, when c < 0

∫ ∫

∫ ∫ (by definition)

∫ ∫

[∫ ∫ ]

[∫ ∫ ]

[∫ ∫ ]

∫ ∫

(ii) Let where are non-negative integrable functions.

⇒
⇒
⇒∫ ∫
∫ ∫ ∫ ∫ (by proposition 8)
⇒∫ ∫ ∫ ∫
⇒ ∫ ∫ ∫ --------------(1)
Now Suppose f and g are integrable, and

Also are non-negative integrable functons over E.

49
 By (1), ∫ ∫ ∫

∫ ∫ ∫ ∫

∫ ∫ ∫ ∫

∫ ∫ ∫

(iii) Let A = { x/ g(x) = , B = { x/ f(x) =

⇒ B⊆ (as f ≤ g)
On E-A, g - f is well-defined, finite and g - f ≥ 0 almost everywhere
[as f and g are integrable]
Also mA=0, ∫ ∫ ∫

= ∫ ∫

=∫ ∫

By proposition 8, ∫ [ g-f ≥ 0 a.e on E-A]

⇒∫ ∫ ∫ ∫ [ mA=0 ]

=∫

⇒ ∫ ≥ ∫

iv) ∫ =∫ =∫ ) =∫ +f

=∫ +∫ =∫ +∫ .

Dominated convergence theorem (or) Lebesgue convergence theorem

THEOREM:17
Let g be integrable over E and Let be a sequence of measurable
functions such that | | g on E and for almost all x on E, we have
f(x) = i (x). Then ∫ = i ∫ .

Proof:

50
 Since g is integrable , is measurable and | | We have
each is integrable.
Therefore, m |
So ignoring set of measure zero, we can assume | | for all x E.
Consider g , Now | | ⇒ g
Also f= i ⇒ i
By Fatou’s lemma , ∫ ∫

⇒ ∫ ∫ ∫ ̅̅̅̅̅ ∫

⇒ ∫ ̅̅̅̅̅ ∫ ⟶ [ ∫ ]
Also | | ⇒ g
⇒ g+ and i +f

By Fatou’s lemma , ∫ ∫
⇒ ∫ ∫ ∫ ∫

⇒ ∫ ∫ ⟶ [ ∫ ]
From and ,

∫ ∫ ∫ ∫

⇒ ∫ ∫ ∫ .
⇒ i ∫ exists and i ∫ ∫ .

PROPOSITION:18 [Generalization of Lebesgue convergenceTheorem]

Let be a sequence of integrable functions which converges
almost everwhere to an integrable function g. Let be a sequence of
measurable functions such that | | and →f almost everwhere if
∫ i ∫ then ∫ i ∫ .
Proof:

51
 Let almost everwhere.
Let D = { x / does not converges to g(x) }. Then m D = 0
⇒ ∫ , and ∫

So we can assume that , for all x E.

Similarly, we can assume that for all x .
Since is measurable, is measurable and | |
⇒ each is integrable as each is integrable.
Consider
and i = g – f , ignoring a set of measure zero.
By Fatou’s lemma , ∫ ∫

⇒ ∫ ∫ ∫ ∫

⇒ ∫ ∫ ∫ ∫ [ i ∫ ∫ ]

⇒ ∫ ∫ [ ∫ ]
Similarly, we can prove ∫ ∫ . ∫ i ∫ .

PROPOSITION:19
A measurable is integrable over E iff | | is integrable.
Proof:
Suppose f is integrable over E and f =
⇒ are integrable (by definition) over E.

∫ and ∫
Now , | |
⇒ ∫| | ∫ ∫
⇒ ∫| | ⇒ | | is integrable.
Conversely ,

52
 Suppose | | is integrable over E.
To Prove: f is integrable.
Given f is measurable. ⇒ are are measurable
Also | | =
Now | | and | |
⇒ ∫ ∫| | and ∫ ∫| |
⇒ and are integrable ⇒ f is integrable.
PROPOSITION:20
If f is integrable over E, then |∫ | ∫| | .
Proof:
Since f | | and f | |
⇒ ∫ ∫| | and ∫ ∫| |
⇒ ∫ ∫| |
⇒ ∫ ∫ ∫
⇒ ∫ ∫ .
Example:

Prove that the function is not lebesgue integrable over [ .

Soln :
We know that the measurable function f is integrable iff is
integrable.
Now, consider the integral,

∫ ∑ ∫

∑ ∫

53
 ]

∑ ∫

[ ]

∑ ∫ sin

∑ ∫ sin

∑ ∫ sin

∑ [– os ]

∑ [– os os ] ∑ ∑

i ∫

i ∑ ∑

∫ .

54
 UNIT III

COMPLEX ANALYSIS

3.1 Complex Numbers

Algebra of complex numbers:

Definition:
A complex number is an ordered pair of numbers C = {(a,b) / a,b ϵ R}
Notation:
The complex number (a, b) is written as a + ib where I =
REMARK:
The set C= {(a,b) /a,b ϵ R }is a field under the operation of
addition and multiplication defined by
i) (a ,b) + (c,d) = (a+c, b+d)
ii) (a,b) * (c,d) = (ac-bd ,ad+bc)
Conjugation and absolute value
The transformation states z = x+iy to ̅ = x-iy is called complex
conjugation.
NOTE:
1) A number is real iff it is equal to its conjugate .
2) Conjugation is involuntary (i.e) ̿ =z
̅
3).Re(z) = x =
̅
4).Im(z) = y =
5) i) ̅̅̅̅̅̅̅̅̅̅= ̅̅̅̅+ ̅
ii) ̅̅̅̅̅̅=̅̅̅̅ ̅
6) z ̅= (x+iy)(x-iy) = x2 + y2
7) ̅ is called the modules or the absolute value of z and it is
denoted by │z│,│z│2= z ̅= x2+y2.
8) i) │z1z2│= │z1││z2│
ii) │z│= │ ̅│
9) │z1+z2│ │z1│+│z2│
│z1+z2│2 = ( z1+z2)( ̅ + ̅ ) = z1 ̅ + z1 ̅ + z2 ̅ + z2 ̅
= │z1│2+│z2│2 +2Re (z1 ̅ )
│z1│2+│ z2│2 +2│z1 ̅ │
=│z1│2+│ z2│2 +2│z1││ z2│

55
 │z1+ z2│2 (│z1+ z2│)2
│z1+ z2│ │z1│+│ z2│
10) │z1+ z2│2 +│z1 z2│2 = 2│z1│2+ │z2│2
│z1+ z2│2 +│z1 - z2│2 = (z1+ z2)( ̅ + ̅ )+ (z1 z2))( ̅ ̅)
= │z1│ +│ z2│ +2Re (z1 ̅ ) + │z1│ +│ z2│ - 2Re (z1 ̅ )
2 2 2 2

│z1+ z2│ +│z1 - z2│2 = 2│z1│2+2│ z2│2

2

NOTE:
By induction hypothesis
i) │z1+ z2+……..+zn│ │z1│+│z2│+…..+│zn│
ii) -│z│ │z│
iii) -│z│ │z│

SQUARE ROOT

Let √
(x+iy)2 = x2 y2 + 2ixy
Equating real and imaginary parts
x2 y2 ……………….(1)

We have to solve for x and y

(x2+y2)2 =(x2-y2)2 +4 x2 y2
= 2+ 2
x2+y2 =√ -----------(2)

(1) + (2) gives 2x2=α+√ and 2y2=-α+ √

√ √
⇒ √ and y = √

The signs of x and y are so chosen that 2xy = β is satisfied

√ = x + iy

√ √
=± √ √ , provided β 0
| |

56
If β = 0 then square root is if α 0

Modulus – Amplitude form of a complex number:

P(r, )

r y

Fig. 3.1

Any complex number can be written of the form z=r ( cos Ɵ + i sin Ɵ),
where Ɵ = amp z = arg z and r = |z|. Then we have x=r cos Ɵ and
y = r sin Ɵ and r = |z| and Ɵ=tan-1(y/x) .

PROBLEM : 1. Show that the area of the triangle with vertices z1, z2, z3 is
| |
given by ∑ .

The area of the ABC = ½∑x1 (y2-y3)

= | |

= | |

= | |

̅̅̅

= || |
̅̅̅
|
̅̅̅

57
 ̅
= | ̅ |
̅

̅
= | | + | ̅ |
̅

̅ ̅
= | ̅ | = |̅ |
̅ ̅

= ∑̅

Problem 2: Show that the equation of the circle with centre α (complex) and
radius r is zz - αz -α z +|α|2 –r2 =0.

Solution: Let C be the centre and P(z) be any point on a circle then

CP=r ⇒|z-α|=r ⇒|z-α|2=r2 ⇒ ( )(̅̅̅̅̅̅̅)=r2

⇒ zz - zα - αz + αα = r2 ⇒ zz - αz - α z + |α|2 - r2 = 0

Problem 3: Prove (i) | ̅

| = 1 if either |a|=1 or |b|=1. when will be the
equation true if |a|=|b|=1 ?. (ii) If |a|<1 and |b|<1 then prove that | | <1

Solution: (i) Consider |a| = 1 ⇒ aa =1

Let w =

w = ̅ ̅
= ̅
= ̅
⇒w =1a

ww = (1 a) (1 a) = = =1
| |

|w|2 = 1 ⇒ |w| = 1

58
 | | = 1

Similarly, when |b|=1 we can prove that |w|=1.

| | = 1 if either |a| = 1 or |b| = 1

Let |a|= |b|=1

| |
Now | | = | | = | ̅|| |

Therefore, | | = 1 is true only if a b

But we have |a| = 1 = |b| and hence the equation is true only when arg a arg b

(ii) Given |a| < 1 and |b| < 1.

To Prove |a-b| < |1-ab| (ie) T.P |a-b|2 < | 1-a b|2 (ie) T.P |a-b|2 – |1-a b|2 <0

Consider |a-b|2 – |1-a b|2 = (a-b)(a -b) – (1-a b)(1-a b)

= aa – ab – ba + bb – 1 + ab + ab - aabb = |a|2 – 1 + |b|2 - |a|2|b|2

= ( |a|2-1) - |b|2( |a|2-1) = ( |a|2-1) (1- |b|2) < 0

Hence | | < 1.

Cauchy’s Inequality:
Let ai ,bi (i = 1, 2,.............,n) be complex numbers

|∑ |2 (∑ | | 2) (∑ | | 2)

Proof: Let λ be any complex number and we assume that not all b i’s are zero
[If all bi’s are zero, then the given in equation is clearly true.

Consider ∑ (| ̅| ) 0

59
 (ie) ∑ | | +|λ|2 ∑ | | - 2Re ̅ ∑ 0

∑
This is true for any complex number and for any λ = ∑ | |

Hence we get

|∑ | |∑ |
∑ | | + ∑ | | 2 - 2Re
(∑ | | ) ∑ | |

|∑ | |∑ |
∑ | | + -2 0
∑ | | ∑ | |

|∑ |
∑ | | - ∑ | |

|∑ | ∑ | | ∑ | |

Lagrange’s Identity

|∑ | ∑ | | ∑ | | -∑ | |

where a1, a2... an and b1, b2.... bn are arbitrary complex numbers. Deduce the
Cauchy’s Inequality.

Proof:

∑ | | ∑ | |

=(a1a1+ a2a2 +............+ anan )( b1b1+ b2b2 +............+ bnbn)

= a1a1b1b1 + a2a2 b2b2 +..............+ ananbnbn + a1a1 b2b2+ a1a1 b3b3+..........+

a1a1 bnbn + a2a2 b3b3+ a2a2 b4b4 +.............+ a2a2 bnbn+..............+ an-1an-1 bnbn+
b1b1 a2 a2 + ..........+ b1b1 anan + b2b2 a3a3+........+ b2b2 anan+............+ bn-1bn-1 a1a1

= |∑ | +∑ | | +∑ | | ……..(1)

60
We know that |a-b|2 = |a|2 + |b|2 -2 Re (ab)

∑ | ̅ ̅| = ∑ | ̅| + ∑ | ̅| e ∑ ̅ ̅ ----(2)

(1) - (2)

∑ | | ∑ | | ∑ | ̅ ̅|

=∑ | | +∑ | ̅| + ∑ | ̅ | -∑ | ̅| - ∑ | ̅|

+2Re ∑ ̅ ̅

= |∑ | + Re ∑ ̅ ……………..(3)

Now, | ∑ | = (a1b1+ a2b2+.........+ anbn) + (a1b1+ a2b2+ .............+ anbn)

= a1b1 a1b1 + a2b2 a2b2+............+ anbn anbn+ a1b1 a2b2+ a1b1

a3b3+.......+ a1b1 anbn +.................+ an-1bn-1 anbn + a1b1

a2b2+........+ a1b1 anbn+ a2b2 a3b3+.........+ a2b2 anbn+ ….

+ an-1bn-1 anbn

= |∑ | +∑ + ∑ ̅

= |∑ | +2Re ∑ ……….(4)

From (3) and (4)

|∑ | =∑ | | ∑ | | -∑ | |

Deduction:

Since ∑ | |

61
 -∑ | |

(ie) |∑ | ∑ | | ∑ | | , which is the required Cauchy

inequality.

NOTE:

The equation of a straight line can be written as z = a+bt, where t is real.

⇒ = t = real ⇒Im = 0. If Im 0 then it is right half

plane and if Im > then it is left half plane.

Spherical Representation

The system C of complex numbers can be extended by introducing the

symbol . Now its connection with finite numbers is given by
a+ = +a = finite a and b. = .b = b including b = .

Further, and 0. are not defined

= a 0 , =0 b . We call as the point at .

Extended complex plane

The points in the plane together with the point at form the extended
complex plane.

1. Every straight line shall pass through the at (Ideal point)

2. No half plane shall contain the ideal point.

Stereographic projection

It is a geometric model in which all points of the extended plane we have

a concrete representation.

62
 Consider the unit sphere S whose equation is x12+ x22+ x32=1 with every
point on S except (0, 0, 1) we can associate a complex number z = and
this correspondence is 1-1.

O
z
Z

FIG. 3.2 Stereographic projection.

Now, |z|2 = = [as x12+x22+ x32 =1]

| |
= ⇒ x3 =
| |

̅ ̅
Similarly, x1 = and x2 =
| | | |

The correspondence can be completed by letting the point at

corresponds to (0, 0, 1).

We can regard the sphere as a representation of the extended plane or of the

extended number system.

63
 Note that the hemisphere x3 < 0 corresponds to the disc |z| < 1 and the
hemisphere x3 > 0 corresponds to its outside |z| < 1.

In function theory the sphere ‘S’ is referred as the Riemann sphere.

If the complex plane is identified with (x1 , x2) plane with x1 axis and x2 axis
corresponding to real and Imaginary axis respectively then the transformation
z= takes on a simple geometrical meaning. Now writing z = x + iy

We have x + iy =

Equating real and Imaginary part

x= , y=

(or) x : y : -1= x1 : x2 : x3 - 1

The points (x, y, 0), (x1, x2, ) and (0, 0, 1) are in a straight line.

Hence the correspondence is a central projection from the center (0, 0, 1). It
is called a stereographic projection. It is geometrically evident that
stereographic projection transforms every straight line in the z-plane into the
circle on S which passes through the pole (0, 0,1) and the converse is also.

More generally, any circle on the sphere corresponds to circle or the straight
line z-plane. To prove this we observe that a circle on the sphere lies in a plane
α1x1+ α2x2+ α3x3 = α0 and α12+ α22+ α32 = 1 and 0 ≤ α0 < 1

̅ ̅ | |
α1 + α2 + α3 = α0
| | | | | |

α1( z + z) - iα2( z - z ) + α3(|z|2 - 1) = α0(1 + |z|2)

α12x -iα2(2iy) + (α3 – α0) |z|2 = α0 + α3

(x2 + y2)(α3 - α0) + 2 α1x + 2 α2y - (α0 + α3) = 0

64
For α0 α3 , this is the equation of the circle.

For α0 α3, it represents as a straight line.

Conversely, the equation of any circle or straight line can be written in this form.

This correspondence is consequently one to one.

To calculate the distance d(z,z′) between the stereographic projection of z and z′.

The points on the sphere are denoted by (x1, x2, x3) and (x1′, x2′, x3′)

d(z,z′) =√ x x x x x x

=√ x x x x x x

Consider,

̅ ̅ ̅ ̅ | | | |
x1 x1′+x2 x ′+x3 x3′ = | |
| |

̅ ̅ ̅̅ ̅ ̅ ̅̅ | | | |
= | | | |

̅ ̅ | | | | | | | |
= | | | |

| | ( | | ) | | | | | | | | ̅ ̅ | | | | | | | |
| | | |

| | | | ̅ ̅
= 1- | | | |

| |
=1 - | | | |

65
 | | | | | |
d(z,z′) =√ ( | | | |
) =√ | | | |
=
√ | | | |

For z′ = the corresponding formula is

d(z, ) = √ x x x

=√ x x x x =√ x =√ x

| | | |
= √ | |
= √ | |

= √ ( | |
) = √ | |
=
√ | |

Problem 1 Show that z and z′ corresponds to diametrically opposite points to

Riemann’s sphere iff z̅̅̅ = -1.

Solution: Let the diametrically opposite points be ( nd

z= ; z′ =

zz̅ = = = = -1 [ s ]

Conversely, zz̅ = -1

Let z = . Since zz̅′= -1 , ( z̅′= -1

z̅′ = ,z′ = = =

= =

z′ =

66
Therefore z = ( , then z′ = (-

Therefore , z and z′ are diametrically opposite points.

3.2 Analytic Functions

Introduction to the concept of analytic function

There are four different types of functions

1. Real function of a complex variable

2. Complex function of a real variable
3. Real functions of a real variable
4. Complex functions of a complex variable

Notation:

W = f(z) is to denote complex function of a complex variable for the

remaining three functions, we use y = f(x), where x and y be real or complex. If
a variable is definitely restricted by real values, then we denote it by t. All
functions must be defined and consequently single valued.

Limit and Continuity:

Definition :

The function f(x) is said to have the limit A as x tends to a.

i (1) if and only if the following is true

For every ,there exists a number with the property that

|f(x)-A| ˂ for all values of x such that |x-a| ˂ and x ≠ a.

Form eqn (1), i ̅̅̅̅̅̅ = ̅

From (1) and (2), i e(f(x)) = Re(A)

67
Similarly, i m(f(x)) =Im(A) )

Conversely, (1) is a consequence of equation (3a) and (3b).

Definition:

The function f(x) said to be continuous at x=a iff i f(x)=f(a).

f is continuous iff f is continuous at all points where it is defined. The sum and
product of two continuous functions are continuous. The quotient is defined
and continuous at a, provided g(a) 0.

If f is continuous, so are Re(f(x)), Im(f(x)) and |f(x)| is continuous.

The derivative of a function:

f′′(a) = i

The usual result for forming the derivative of a sum , a product or a quotient are
all valid. The derivative of a composite function is determined by the chain
rule.

There is a fundamental difference between the cases of a Real and Complex

Independent variable.

Result: The real function of a complex variable either has a derivative zero or
else the derivative does not exist.

Proof: Let f(x) be real function of a complex variable whose derivative exists
at z = a . Then f′(a) is on one side is real, for it is the limit of the quotient
as h through all real values, On the other side, it is also the limit
of the quotient and as such purely Imaginary.

Since f′(a) exists and is unique and it is both real and Imaginary ⇒ f′(a) = 0

Example: W = f(z) = |z|2

68
 w = |z+ z|2 - |z|2 (z+ z) (z̅+̅̅̅z) -zz̅ = zz̅+z̅̅̅z+z̅ z+ z̅̅̅z -zz̅

̅̅̅̅
= + z̅ +̅̅̅z (1)

When z = 0, i = i ̅̅̅z = 0 ⇒ =0

When z Let 0 through all real values

Take z=h , ̅̅̅z = h. Then , = z + z̅ h

= i = z + z̅

z through purely imaginary values, z ih and ̅̅̅z = -ih

When z ⇒h

From (1) , = -z +z̅ …………………………(3)

From (2) and (3), does not exist when z , since the limit is unique.

Therefore exists only at the origin.

The case of a complex function of a real variable can be reduced to the real
case.

z(t) = x(t) +iy(t) ⇒ z′(t) = x′(t) + y′(t)

The existence of z′(t) is equivalent to the simultaneously existence of x′(t) and

y′(t).

Analytic Function:

The class of analytic function is formed by the complex functions of a complex

variable which posses a derivative whenever the function is defined.

69
 The sum and product of two analytic functions is again analytic. The same
is true for the quotient of two analytic functions, provided that g(z) also not
vanish. In general case, it is necessary to exclude the points at which g(z) = 0.

The definition of the derivative can be written in the form

f′(z) = i .

As a first consequence, f(z) is necessarily continuous .

[ ]
For, f(z+h) - f(z) =

i f z h f z i

Therefore i f z h = f(z) . Therefore in general the converse is not

true.

Example: f(z) = |z|2. It is continuous at all the points. But it is not

differentiable when z ≠ 0.

If f(z) = u(z) + iv(z) is continuous then it implies u(z) and v(z) are both
continuous.

Theorem 1: Let w = f(z) = u(x,y)+iu(x,y) be differentiable at any point in a

region D . Then the partial derivatives ux,uy and vx,vy exist and satisfy the
Cauchy Riemann equations ux = vy , uy = -vx (i.e) = ; =- .

Proof:
Let f(z) = u(x,y) + iv(x,y) be analytic at any point z of the region D.

Therefore, f′(z) = i exists and is unique . (i.e) It is independent

of the path along which h 0.

If h = x then f′(z) = i +i i

70
= +i (

Therefore f (z) = ux+ ivx (1)

Since f′(z) exists, the above limit exists which means that ux and vx exist.

If h = i y , f (z) = i +i i

= -i + = -i

f′(z) = -i uy+vy (2)

Since f′(z) exists, the above limit exists which means that uy and vy exist

Since the limit should be unique, from (1) and (2) ux + ivx= -i uy + vy

Equating real and imaginary parts, we have ux = vy and uy = -vx

These are called C-R equations.

The following theorem is the sufficient condition for function is to be analytic.

Theorem 2: If u(x,y) and v(x,y) have continuous first order partial derivative
which satisfy the C-R equations , then f(z) = u(z) + i v(z) is analytic with
continuous derivative f′(z).

Proof: Let f(z) = u(x,y) + iv(x,y) where ux = vy , uy = -vx

Now, f(z+h+ik) - f(z) = f(x+iy+h+ik) - f(x+iy) = f(x+h+i(y+k)) - f(x+iy)

= f(x+h+i(y+k)) - f(x+iy)

= u(x+h,y+k) + i v(x+h,y+k) - u(x,y) - i v(x,y)

= u(x+h,y+k) - u(x+h,y) + u(x+h,y) - u(x,y)

71
 +i [v(x+h,y+k) - v(x+h,y) + v(x+h,y) - v(x,y)]

Using mean value theorem we get,

u(x+h , y+k) - u(x+h , y) = k (x + h , y + k) (0 < < 1) [ exist]

= k( (x,y) + ) where 0 as h k 0[ is continuous]

u(x+h ,y) – u(x , y) = h (x+h , y) (0< <1) [ exist]

= h[ ((x,y)+ ] where as h 0, k 0 [ is continuous]

u(x+h , y+k) – u(x , y) = h +k +h +k

=h +k + where = k+h &

where as h 0, k 0

Similarly,

v(x+h , y+k) – v(x , y) = h +k + where =k +h

where 0, 0 as h 0 , k 0

Taking limit h+ik 0

, , and 0 [as h k and 0]

as h+ik

f(z+h+ik) – f(z) = u(x+h , y+k) – u(x , y) + i(v(x+h , y+k) – v(x , y))

=h +k + + i (h +k + )

=h( +i )+k( +i )+ +i

72
 =h( +i )+k( +i )+ +i [ = , =- ]

=h( +i ) + ik ( + i )+ +i

= (h + ik) ( +i )+ +i

Hence, i = +i + i

= +i

Since and exist and are unique, (z) exists.

Hence f(z) is analytic at an arbitrary point z. It is analytic in a region.

Hence the theorem is proved.

It is observed that as the C.R equations are necessary condition for

differentiability, if they are not satisfied at a point then the function is not
differentiable at that point.
Note that as f ‘(z) = i then | z | =| i |
= +
= + (- ) [ =- , = ]
= -
We shall prove later, the derivative of an analytic function is itself
analytic.
⇒ u and v will have continuous partial derivatives of all order and in particular
the mixed derivatives will be equal.
From the C-R equation, = and =-
⇒ = and =-
⇒ =- [ = ]
⇒ + =0

73
(i.e) u= + = 0. Similarly, v= + =0

A function u satisfies the Laplace’s equation u = 0 is said to be

harmonic.
The real and imaginary parts of an analytic functions are harmonic.
If two harmonic function u and v satisfies the C-R equations then v is said to
be harmonic conjugate to u.
If v is a harmonic conjugate of u then –u is the harmonic conjugate of
v and conversely.
It is also true that the harmonic conjugate is unique except for an additive
constant.
Observation: f(z) is analytic function on D if and only if v is harmonic
conjugate of u.
If f(z) = u(x,y) + iv(x,y) is analytic.
⇒ u and v satisfy the C-R equations.
⇒ v is a harmonic conjugate of u.
Conversely
If v is a harmonic conjugate of u , by theorem 2 , the function f(z) =
u(x,y) + iv(x,y) is analytic.[for this purpose, we make the explicitly that u and v
have continuous first order partial derivatives]
Example 1:

Find the harmonic conjugate of a harmonic function u(x,y)= - .

= 2x , = -2y

Using C-R equations, = 2x and - = -2y ⇒ = 2y

Consider = 2y. Integrating w.r.to x keeping y as constant

v = 2yx + (y) ⇒ = 2x + ’(y)

⇒ 2x = 2x + ’(y) ⇒ ’(y) = 0

⇒ (y) = c (a constant)
74
 v = 2xy + c

f(z) = u+iv = - + i(2xy + c)

= - + i2xy + ic

= + ic = + ic

Example 2:

Consider a complex function f(x,y) of two real variables. Let z = x + iy ,

̅ ̅
̅ = x - iy and x = , y = -i( ). With the change of variables we can consider
f(x,y) as a function of z and ̅ which will be treated as independent variables.

Soln: = +

= + = ( –i )

̅
= ̅
+ ̅

= + = ( +i )

If f is analytic then

⇒ =0 ⇒ ̅
=0

⇒ any analytic function is independent of ̅ and a function z alone.

Corollary 3:

This formal reasoning supports that analytic functions are true functions of
a complex variable as opposed to functions which are more adequately described
as complex functions of two real variables.

75
 By similar formal arguments, we derive a simple method which allows to
compute without the use of integration.

The analytic function f(z) whose real part is given harmonic function u(x,y)
[given a harmonic function u without the use of integration we are now going to
determine the analytic function f(z)].

̅
Note that = 0 ⇒ ̅̅̅̅̅̅ may be considered as a function of ̅ , denote it
by (̅ )̅

⇒ u(x,y) = [ f(z) + i ̅̅̅̅̅̅]

= [ f(z) + i (̅ )̅ ]

= ̅
[ f(x+iy) + i (x-iy)]

This is a formal identity. It is reasonable to expect that it holds even when x

and y are complex.

Let x = ⁄ ,y= ⁄

u( ⁄ ̅
, ⁄ ) = [f(z) + (0)] ……………… (1)

Since f(z) is only determined upto a purely imaginary constant we may as well
assume that f(0) is real.

⇒ ̅ = u(0,0)
(0)

f(z) = 2u ( ⁄ , ⁄ ) - u(0,0) [By (1)]

A pure imaginary constant can be added at will.

NOTE: In this form,the method is definitely related to the function u(x,y) which
are rational in x and y for the function must have the meaning for the complex
values of the arguments.

76
Example 3: Show that the harmonic function satisfies the formal differential
equation ̅
.

Soln: Given, u is a harmonic function. ⇒ =0 ⇒ + =0

Now, ̅
= [ ̅] = [ . ̅
+ . ̅
]

= [ . + ( ] = [ +i ]

= [ . + . +i( . + . ]

= [ . + . +i . +i . ]

= [ -i +i .+ ]

= [ + ]= .0 =0

Aliter: + =4 ̅

Theorem 4:

If f(z) = u +iv be an analytic function in a region D,then prove that f(z) is

constant in D. If any one of the following conditions hold ,

(i) f ’(z) vanishes identically in D. (ii) R[f(z)] = u = constant.

(iii) Im(f(z)) = v = constant. (iv) | | = constant.

(v)arg f(z) = constant.

Proof: Now f(z) =u +iv

77
 f ‘(z) = +i = -i ( By C-R eqn)

(i) Now, f ‘(z) 0 ⇒ 0 ⇒ 0

⇒ =0, =0, =0, =0

⇒ u and v are constant on any line segments parallel to co-ordinate axis.

But any two points in D can be joined by such parallel lines.

⇒ f(z) is constant.

(ii) u = a is constant ⇒ =0, =0

f ‘(z) = +i = -i =0

By (i), f(z) is constant.

(iii) v = constant ⇒ =0, =0

f ‘(z) = +i = +i =0

By (i), f(z) is constant.

(iv) | | = constant ⇒ + = constant

⇒ = 0 …………..(1)

= 0 …………..(2)

u X (1) + v X (2)

⇒ =0

78
⇒ =0 [ ]

⇒ = 0 (or) =0

Similarly,

⇒ = 0 (or) =0

If 0 then u and v are constant

Therefore is constant.

If = 0 at a point and it is constantly zero and = 0.

(v) arg = c = constant

⇒t n ( ) ⇒ =t n ⇒ t n ⇒ ot

= kv where k = ot ⇒ ⇒ =0

⇒ =0 ⇒ = constant [By (ii)]

⇒ f(z) = constant

Polynomials:

Every constant is a analytic function with derivative zero. The simple non
constant analytic function is z whose derivative is one. Since the sum and the
product of two analytic functions are again analytic ⇒ every polynomial
p(z) = ………+ is an analytic function and its derivative
f‘(z) = ………..+ is analytic. If 0 then deg p(z) = n.

For formal reason the constant zero is regarded as a polynomial and its
degree is Therefore, the zero polynomial is excluded from our
consideration.

79
 By fundamental theorem of Algebra, P(z) = 0 has at least one root for n 0.

If P( ) = 0 ⇒ P(z) =(z- ) (z) where (z) is the polynomial of degree

n-1. The repetition of this process leads to a complete factorization
P(z) = (z- )……… (z- ), where ……., are not necessarily distinct.
Moreover, the factorization is uniquely determined except the order of the
factors.

If exactly h of coincide, their common values called a zero of P(z) of

order h. Sum of the orders of the zeros of the polynomial is equal to its degree.

Determination of the order of zero :

Suppose is a zero of P(z) of order h.

⇒ P(z) = (z) and ( 0 and successive derivative yields,

P( ) = P’( ) =…………= = 0.

(i.e) the order of a zero equal to order of the first non-vanishing derivative.

NOTE: Zero of order one is called a single zero and characterized by the
condition, P( ) = 0 and P’( ) 0.

THEOREM 5: (LUCAS)

If all zeros of a polynomial P(z) lie in a half plane, then all zeros of a
derivative P’(z) lie in the same half plane.

Proof: If 1, 2, 3,………, n are zeros of P(z). Then P(z) can be written as,
P(z) an(z- 1)(z- 2)……..(z- n),where an 0

Taking log on both sides

og og + og )+ og )+……+ og )

Differentiate with respect to z, we get

80
 + +……+ ⟶(1)

Let the half plane H be defined as the part of the plane where ( )

If k is in H and z is not in H

Then we have ( )= ( )

= ( )+ ( )

= ( )- ( )

But the imaginary parts of a reciprocal number have opposite signs.

Therefore, under the same assumption ( ) 0

If this is true for all k, therefore from (1)

b =∑

( ) (∑ ) =∑ ( ) < 0

P’(z)

All the zeros of a derivative P’(z) lie in the same half plane H.

RATIONAL FUNCTION

Let R(z) be the quotient of two polynomials. We can assume that P(z)
and Q(z) has no common factors and hence no common zero.

81
 R(z) will be given the value of at the zeros of Q(z). It must therefore must
be considered as the function with the values in the extended plane, and as such
it is continuous .

The zeros of Q(z) are called poles of R(Z). R’(z) only when
Q(z) .R’(z) has the same poles as R(z),the order of each poles being increased
by 1.

Poles and zeros of a rational function at .[ i ]

Consider ( ) . (i.e) R( (0)

If (0) 0 or , the order of the zero (or) the pole at is

defined as the order of the zero (or) pole of (z) at the origin

R(z)

We obtain,

By the power belongs either to the numerator or denominator.

Case(i) m > n
(z) has a zero of order m-n at the origin.
has a zero of order m-n at .

Case(ii) m < n
has a pole of order n-m at the origin
has a pole of order n-m at

Case(iii)
R
Since R( is neither zero nor
has neither zero nor a pole at

82
 Number of zeros Number of pole
In the finite In the finite
plane At Total plane At Total

m>n n m-n m m - m

m<n n - n m n-m n

m=n n - n m - m

NOTE:

We can now count the total no of zeros and poles in the extended plane. The
count shows that the no of zeros including those at is equal to bigger of m and
n.

This common number of zeros and poles is called order of the rational
function.

If a is any constant, the function R(z) - a has the same poles as R(z) and
consequently the same order .

The zeros of R(z) - a are roots of the equation R(z)=a .

Theorem 6

A rational function R(z) of order p has p zeros and p poles and every equation
R(z) = a has exactly p roots.

Proof:

Let R(z) be a rational function.

Consider R(z) – a = –a = -----------(1)

83
The numerator and denominator can not have a common factor.

For, if so, it would be a factor of P(z) and Q(z) both and therefore R(z) would
not be in the lowest form. R(z) is not a rational function. This is a contradiction.
It follows that the order of R(z) – a = p = order of R(z). Therefore, R(z) – a has
exactly p roots.

Theorem 7: Every rational function has a representation by partial fraction:

Proof: First to derive this representation R(z) has a pole at we carryout the
division of P(z) by Q(z) until the degree of the remainder is atmost equal to that
of the denominator.

⟶(1) Where G(z) is a polynomial without constant term

and H(z) is finite at .The degree of G(z) is the order of the pole at the
polynomial G(z) is called the singular part of R(z) at .

Let the distinct finite poles R(z) be denoted by . The function

R( ) be the rational function of with a pole at is equal to .

From decomposition (1), R( )

Let z = . Then ( )+ ( ) ⟶(2)

Here ( ) is a polynomial in without constant term called the singular

part of R(z) at z = .The function ( ) is finite for z =

Consider now the expression, R(z) - G(z) - ∑ ( ) ⟶(3)

This is a rational function which cannot have other poles than and
At z = we can find that the two terms with finite limits and the same is
true at .

84
Therefore, (3) has neither any finite pole not a pole at A rational function
without poles must reduce to a constant .

If this constant is observed in G(z), We obtain R(z) = G(z)+ ∑ ( )

3.3 POWER SERIES

A power series is of the form + z+ +……+ +…… where the

coefficients and the variable z are complex.

NOTE:

∑ is a power series with respect to the center .consider the

geomentric series 1+z+ +……..+ +…… Whose partial sum is
= 1+z+….+ . Since ⟶0 for | | and | | for | | ,

The geomentric series convergent to for | | and diverges for | | .

THEOREM 8 (ABEL)

For every power series ∑ there exists a number R, 0

called the radius of convergence with the following properties.

1. The series converges absolutely for every z with | | . If 0 ρ the

convergence is uniform for | | .

2. If | | the terms of the series are unbounded and the series is consequently
divergent.

3. In | | the sum of the series is an analytic function. The derivative can be

obtained by the term wise differentiation and the derivative series has the same
radius of convergence.

85
Proof: The circle | | is called the circle of convergence .We shall show
that the theorem holds if we choose R according to the formula

i | | ⟶(1)

This is known as Hadmard’s formula.

Let | | Then there exists such that | |

By the definition of limit superior and equation (1), there exists a positive
integer such that | | (i.e) | | for all n ⟶(2)

| |
| | ( ) for large n

Since the power series ∑ has a convergent geometric series as a majorant

and is consequently convergent.

To prove the uniform convergence for | | We choose with

From (2), We get | | for all n

| |
| | ( ) ( ) for all n [ | | ]

Since the major ant is convergent and has constant terms, we conclude by
Weierstrass M-test that the power series is uniformly convergent.

If | | we choose so that | |

Since i | | There are arbitrary large n such that | |

86
 | |
(i.e) | | and consequently | | ( ) for infinetly many n. Hence
the terms of the series is unbounded accordingly the series is divergent.

STEP:1

The derivative series ∑ has the same radius of convergent

Proof:

Let R and be the radii of convergence of the series ∑ and

∑ respectively.

Then ̅̅̅̅ | | and ̅̅̅̅ | |

= ̅̅̅̅ | |

Therefore the theorem is over if we show that i

To prove this, Let so that n

Hence (or) so that as n

i and so R’ = R.

STEP 2:

For | | We write ∑

where

87
 ∑ and also ∑ = i

To prove:

Consider the identity

. / ( )

⟶(3) where we assume that z ≠ z0 and |z0| < ρ < R

∑ ∑

| | ∑ │ak│ k-1
[since │z│< and │zo│< ]

Now ∑ │ak│ k-1

is the remainder term in a convergent series .

Hence, we can find no such that | |<€ 3 n no --------(4)

i = f1(z) for │z│< R and since │zo│< R , i = f1(zo)

=>There exists an n1 such that │Sn’(zo) - f1(zo)│< € 3 ----------(5) n n1.

Choose a fixed n no , n1 . We know that Sn’(zo) = i

By the definition of derivative , we can find > 0 such that 0<│z - zo│<

 | | < € 3 --------(6)

88
Using (4),(5) and (6) and it follows by (3) that,

| – | < € when 0 < │z-zo│<

 f’(zo) exists and f1(zo) = f ’(zo).

Remark: Every analytic function has a Taylor development .The power series
development of f(z) is uniquely determined if it exists.

A power series with positive radius convergences has derivatives of all orders.

They are given explicitly by

f(z) = ao+a1z+………..+an zn +…………………

f ‘(z) = a1+2a2z+……………..+nanzn-1+………..

f ”(z) = 2a2+6a3z+……………+n(n-1)anzn+……

…………………………………………………

f k(z) = k!ak+ ak+1z + a k+2 z2 +……….

In particular, ak =

The power series becomes f (z) = f (0) + z +…..+ +…….

This is the familiar Maclarian – Taylor development. But we have proved only
under the assumption that f(z) has a power series development.

The following theorem refers to the case where a power series converges at a
point on the circle of converges at a point on the circle of convergence and
note that R = 1.

89
Theorem : 9 (Abel’s Limit theorem)

∑an convergences .Then f(z) = ∑ tends to f(1) as z approaches 1. In

such a way that remains bounded.

Proof.

We may assume that ∑ =0, since this can be obtained by adding a

constant to ao. Now, f (1) = ∑ = 0. Let Sn= ao+a1+……………+an

Consider the identity (summation by parts)

Sn(z) = ao+a1z+………..+a n z n = so+(s1-so)z+…………….+(sn-sn-1)zn

= so(1-z) + s1(z-z2) +………..+ sn-1(zn-1-zn) + s n z n

= (1-z)(so+ s1z +……………+sn-1zn-1) +s n z n

But s n z n 0 as n (since ∑a n= 0, sn )

f (z) = (1-z) ∑

Since remains bounded, there exists a positive constant k such that

Since Sn , given choose m so large that │Sn│< for n

Now ∑

∑ ∑ --------(1)

and ∑ ∑

< ∑

90
 = | |
<

Therefore (1) becomes ∑

The first term on the right can be made arbitary small by choosing z sufficiently
close to1.

Therefore subject to the stated restriction.

PROBLEM 1:

1).Find the radius convergences of the following series.

i) ∑npzn ii)∑ iii)∑ n!zn iv)∑(1+1 n)n2zn

soln :

i) an= np an+1= (n+1)p

R= i = i = i ( )

= i =1

ii) ∑

an = an+1=

R= i = i = i = i =

iii) For z = 0 the series is convergent.

an=n!, an+1= (n+1)!.

91
Then R= i = i = i = i =0

iv) an = (1+

= i an )1/n = i (1+ )n = e . Therefore, R =

Problem 2. If ∑ an zn and ∑ bnzn have the radii of convergences R1 and R2 .Show

that the radius of convergences for ∑ zn is atleast R1R2 .

Soln: ̅̅̅̅̅ │an│1/n and = ̅̅̅̅̅ │bn│1/n

Let R be the radius of convergence of ∑anbnzn

̅̅̅̅̅ │1/n = ̅̅̅̅̅│an│1/n│bn│1/n =̅̅̅̅̅ │an│1/n̅̅̅̅̅̅ │bn│1/n = .

=> R=R1R2

Problem3. Find the radius of convergences of a power series f (z) =∑ and

prove that (2-z) f (z) - 2

Solution: a n = , a n+1 = . and R = i

R = i = i = = 2

f (z) = ∑ < ∑ =1+z/2+z2/22+………

= = [ since │z│< 2 ]

i (2-z) f (z) = i (2-z) = 2

Therefore

92
 UNIT IV

COMPLEX INTEGRATION

4.1 The Line Integrals :

Definite integral of complex function over a real interval . If f (t) = u(t)+i v(t) is
a continuous function defined in an interval (a,b) .Then define,

∫ ∫ ∫

Properties of the integral

Property 1 : ∫ ∫

Proof: Let c = and f(t)= u(t)+i v(t).

∫ ∫ ( )

=∫ [ ]

=∫ ∫ ……….(1) (by defn)

∫ dt = ( ∫

= ∫ ∫ ∫ ∫

=∫ ∫ ∫ ∫

=∫ ∫ ………..(2)

From (1) and (2), ∫ ∫

93
Property 2: When a |∫ | ∫ | | holds for arbitrary
complex function f(t).

Proof:

If ∫ then ∫ ∫

Clearly, ∫ ∫

Therefore, the given statement is true .

Now assume that ∫

From (1), Re*∫ + = Re* ∫ +

Since ‘c’ is arbitrary, we may set c = where is real but arbitrary.

Re* ∫ + = Re*∫ +

=∫ [ ) ] dt ∫

∫ …………….(1)

Since rg ∫

Then ∫ ∫ ………..(2)

Re* ∫ + Re* ∫ +

= Re ∫

= ∫ ……….(3)

94
From (1) and (3), We have ∫ ∫

Complex line integral of f(z) extended over the arc

Suppose is a smooth arc is given by z = z (t), a ≤ t ≤ b and f (z) is

continuous on . Then f (z (t)) is continuous in t. We define

∫ ∫ ( ) .

If is piecewise differentiable or if (t) is piecewise continuous the

interval can be subdivided in the obvious manner.

The integral is invariant under change of parameter. A change of

parameter is defined by increasing function t = t( ) which maps an interval
α ≤ ≤ β into a ≤ t ≤ b. we assume that t ( ) is piecewise differentiable.

By change of variables,

∫ ( ) ∫ ( ( )) ( )

But ( ) (z (t( ))).

∫ ( ) ∫ ( ( )) (z (t(z))) dz.

Hence the integral has the same value whether is represented by z = z (t) or by

z = z (t ( )).

Note: We define the opposite are – by the equation z = z (-t), -b ≤ t ≤ -a.

∫ ∫ ( )

= ∫ ( ) ( ) by the change of variable, z = - t

95
 = ∫

Further if + +……...+ then

∫ ∫ =

=∫ ∫ +……….+∫

Integration with respect to arc length

∫ ∣ ∣=∫ ( )∣ ∣ ∫ ( )∣ ∣dt

|∫ | =∣∫ ( ) (t) dt ∣ ∫ ∣ ∣∣ ∣dt

∫ ∣ ∣∣ ∣

Note : If f = 1 then ∣∫ | ∫ ∣ ∣

∫ ∫ ∣ ∣ =∫ ( )∣ ∣ (as ds = dx2+dy2)

To find the length of the circle with radius with centre at a:

The parametric equation of circle is z = z(t) = a+ ,0≤t<2 . .

∫ ∫ ∣ ∣ ∫ ∣ ∣ ∣ =∫ =

RECTIFIABLE ARCS

The length of an arc can also be defined as the least upper bound of all sums.

∣z ( ) z ( ) ∣ ∣z ( ) )∣ +∣z ( ) )∣

where a = < <……. < =b

96
If the l.u.b is finite we say that the arc is rectifiable.

Note : Piecewise differentiable arcs are rectifiable arcs.

Observation: An are z = z (t) is rectifiable iff real and imaginary part of z (t) are
of bounded variation.

For, Since ∣ ) ) ∣ ≤ ∣z ( ) )∣ and

∣ ( ) )∣ ≤ ∣ ) )∣

∣ ) )∣=∣ ∣

=∣[ ] [ ]∣

≤ ∣ ∣ + ∣ ) ∣

The sums ∣ ∣+∣ ∣+…………. +∣ ∣ and

the sums ∣ ∣ ∣ ∣+…………+∣ ∣,

∣ ∣+∣ ∣+…………. +∣ ∣ are

bounded at the same time.

When the later sums are bounded, one says that the functions x(t) and y(t) are of
bounded variation.

Therefore a arc z = z (t) is rectifiable iff real and imaginary part of z (t) are of
bounded variation.

LINE INTEGRALS AS FUNCTION OF ARC:

General line integrals of the form ∫ are often studied as

function or functional of the arc

Assume that in a region is

free to vary on Ω.

97
 Class of integrals having the property that the integral over an arc
depends on its end points.

If and have the same initial point and the same end points then
∫ ∫ .

Theorem 1:
Integrals depend only on the end points iff the integral over any
closed curve is zero.

Proof:

If is a closed curve then and – have the same end points and if the

integrals depend only on the end points.

 ∫ ∫ ∫

 ∫ .∫

Conversely, Suppose and have the same end points.

(i.e.) T.P the integral depends only on the end points.

(i.e.) To Prove: ∫ ∫ have the same end points.

By given hypothesis, is a closed curve

Since the integral over any closed curve is zero.

∫ ∫ ∫ ∫ ∫

∫ ∫

The following theorem gives the necessary and sufficient condition for a line
integral depends only on the end points

98
THEOREM 2:

The line integral ∫ defined in Ω, depends only on the end

points of U(x,y) in Ω with the partial
derivatives .

Proof:

Suppose there exists a function U(x,y) in Ω such that .

To Prove the integral depends only on the end points of

Let a, b are the end points of

FIG. 4.1

Then ∫ ∫

=∫

=∫ ( )

99
 =∫ ( )

=[ ]

= U(x (b),y (b)) – U (x (a), y (a))

The line integral ∫ depends only on the end points of .

Conversely, Suppose the line integral ∫ depends only on the end

points of .

To Prove: There exists a function U(x,y) in Ω such that .

Choose a fixed point ( joint it to (x, y) by a polygon contained in Ω

whose sides are parallel to the co-ordinate axes.

Define a function U(x, y) = ∫

Since the line integral depends only on the end points of , the function is well
defined. Choose the last segment of horizontal, we can keep y constant and let
x vary without changing the other segment.

On the last segment we can choose x for parameter and obtained

U(x,y) = ∫

The lower limit of the integral being irrelevant ,

In the same way, by choosing the last segment vertical by keeping x constant,

We have U(x,y) = ∫

100
Note:

If then is an exact
differential.

Therefore, the above theorem can be stated as, An integral depends only on the
end points if and only if the integrand is an exact differential.

Consider f (z) dz = f (z) dx + i f (z) dy

By the definition of an exact differential, there must exist a function F(z) in Ω

such that

which is the complex form of C – R equations.

Further, f (z) by assumption is continuous. (Otherwise∫ is not defined)

Hence F(z) is analytic.

The above theorem can be restated as follows

The integral ∫ with continuous f depends only on the end points of iff
f is the derivative of an analytical function in Ω.

Lemma 3:

We find that ∫ dz = 0 for all closed curve provided that the

integer n ≥ 0.

Proof:

n
For, since f (z) = (z-a) is continuous and f is the derivative of an analytic
function

F (z) = in the whole plane.

101
 By the result above, ∫ depends only on the end points of .

 the integral over any closed curve is zero.

∫ dz = 0 for any closed curve . …….(1)

If n is negative but not equal to -1.

The same result hold for all closed curves which do not pass through a. In the
complementary region of the point a the indefinite integral is still analytic.

For n = -1, the equation (1) does not always hold.

Consider an example.

Consider a circle C with centre a represented by the equation z = a + eit,

0 ≤ t ≤ 2 Then dz =

We obtain ∫ ∫ dt = 2

Example 1. Compute ∫ where the directed line segment from 0 to 1+ i.

Soln: Let z = x+iy , z = 0 , ⇒ x = 0, y = 0 and z = 1+i ⇒ x=1, y=1

Therefore y = x ⇒ dx = dy

Therefore ∫ ∫

Example 2: Compute ∫∣ ∣
∣ ∣∣ ∣

Soln: Given ∣z∣=1 then z = ∣z∣ = os t i sin t

102
 ∣ ∣ ∣ os ∣

Now, dz = i dt implies |dz| = dt. Then,

∫∣ ∣
∣ ∣∣ ∣ =∫ dt = 0 1 = -4(cos - cos0) = 8.

4.2 Cauchy’s Theorem for a rectangle

Consider a rectangle R defined by the inequality a ≤ x ≤ b and c ≤ y ≤ d.

This perimeter can be considered as a simple closed curve consisting of four line
segments whose direction we choose so that R lies to the left of the directed
segment. The order of the vertices is (a,c), (b,c), (b,d), (a,d). We refer to this
closed curve as the boundary curve (or) contour of arc and we denote it by

R is chosen as a closed point set and hence it is a region. Further, a function

is analytic on the rectangle R means that it is analytic on an open set which is
contained in R.

Theorem 4:

The function f(z) is analytic on R. Then ∫ = 0.

Proof: Proof is based on the method of bisection. Define =∫ ,

which we will use for any rectangle contained in the given one.

If R is divided into 4 congruent rectangles by joining the

mid points of opposite sides.

We denote the boundaries of the rectangles , k = 1, 2, 3, 4.

Therefore We find that since the

common sides cancel each other.

Now,∫ ∫ ∫ ∫ ∫

103
Denote ∫ ( )

Clearly for at least one (k = 1, 2, 3, 4), we have ∣ ɳ ∣ ≥ ɳ (R).

FIG. 4-2. Bisection of rectangle.

We denote this rectangle by . If several Rk have this property that choice

shall be made according to the definite rule. This process can be repeated
indefinitely and be obtain the sequence of nested rectangle or R ⊃ R1⊃ R2
⊃…….⊃ Rn⊃……
With the properties, ∣ɳ ( )∣≥ ∣ɳ ∣

≥ ( ∣ɳ ∣ ∣ ∣ .

≥ ∣ ∣

Therefore, ∣ɳ (Rn) ∣ ≥ ∣ ∣

The rectangle Rn converges to a point z* R in the sense that Rn will be

contained in the prescribed neighborhood ∣z-z*∣ < as soon as n sufficiently
large.

104
 First of all, we choose so small that f (z) is defined and analytic in
∣z-z*∣ < .

Secondly, if is given, we choose so that

∣z-z*∣ < ⇒∣ ∣

⇒∣ ∣ ∣ ∣ ∣ ∣ )

We assume that satisfies both conditions Rn is contained in ∣z-z*∣ < .

Also we have ∫ ∫ [ 1 and z are the derivative

of analytic function z and respectively]

 ∫

∫ ∫ ∫ ∫

∫ [( ) ]

Therefore | | ∫ | || |

≤ ∫ | || |

If dn denotes the length of the diagonal of Rn then z ϵ Rn and so | z - | d

If Ln denotes the length of the perimeter of Rn then

| |≤ ∫ | | = .

If d and L denotes the length of the diagonal and the perimeter of R

respectively, then dn = 2–n d and Ln = 2-n L. Therefore, | | ≤ 4-n
)

105
From (1) and (3) , | |≥ ∣ ∣

∣ ∣ ≤ 4n | | ≤ 4n 4-n ≤

Since is arbitrary, we have . Hence ∫ = 0.

Theorem 5

Let f(z) be analytic on the set R’ obtained from a rectangle R by omitting a

finite number of interior points . If it is true that i for
all I, then ∫

Proof: It is sufficient consider the case of a single exceptional point . (For

evidently R can be divided into smaller rectangle which contains at most one

We divided R into 9 rectangles as shown in the figure.

FIG. 4-3

Therefore, ∫ = ∑ ∫ .

Hence, by Cauchy theorem for rectangle applied to all rectangles except R 0

∫ = 0, i = 1, ….,8.

106
Therefore, ∫ = ∫ . Given, i

If ϵ > 0 we can choose the rectangle R0 so small that |z – | | | < ϵ.

Consider | | on R0.
| |

| |
|∫ |≤ ∫ | || | <ϵ∫ ….(1)
| |

Let us assume R0 is a square with centre and a be the side of the square R0.

Therefore, | z – | ≥ ⇒ .
| |

| |
Now, ∫ ∫ | |
| |

Therefore (1) becomes, |∫ | Since is arbitrary, the theorem

follows.

Note: The hypothesis of the theorem is fulfilled if f(z) is analytic and bounded
on R’.

Cauchy’s theorem in a disc

It is not proved that integral of an analytic function over a closed

curve is always is zero. ∫

In order to make sure that the integral vanishes, it is necessary to make a special
assumption concerning the region Ω in which f(z) is known to be analytic and to
which the curve γ is restricted.

We must restrict to a special case we assume that Ω is a open disc

|z – z0| < ρ to be denoted by Δ.

107
Theorem 6:

If f(z) is analytic in a open disc Δ, then ∫ for every closed curve γ

in Δ.

Proof: Let O be centre z0 = x0 +i y0 and P be any point z = x + iy inside Δ. We

define a function F(z) = ∫ ……..(1) where ζ consists of the horizontal
line segment OA from the centre (x0 ,y0) to (x ,y0) and the vertical segment AP
from (x ,y0) to (x, y).

P(x,y)


 

FIG. 4-4

F(z) = ∫ =∫ ∫

=∫ ∫ …….(3)

From the figure, OAPBO is a rectangle. By Cauchy theorem of rectangle,

∫

Let ζ1 be a curve consists of the vertical segment OB from (x0 ,y0) to (x0, y) and
the horizontal segment BP from (x0 ,y) to (x, y).

108
OAPBO = OAP + PBO = ζ + (-ζ1) = ζ - ζ1. Therefore, ∫

∫ ∫ ∫ ∫

∫ ∫

Therefore, F(z) = ∫ =∫ ∫

=∫ x ∫

= ∫ x ∫ …..(5)

From (3), = i f(x+iy) = i f(z)

From (5), = f(x+iy) = f(z) +i = f(z) – f(z) = 0

If F(z) = u + iv then +i =0

+i +i - =0 ( - )+i( =0

- = 0 and =0 ux = vy and vx = - uy

ux = vy and uy = -vx

u and v satisfy C.R equations.

Now, , = i f(z) and f(z) is continuous

ux , uy , vx , vy are all continuous. Therefore, F(z) = u + iv is analytic on Δ

f(z) dz is exact differential

109
 The integral depends only on the endpoints.

The integral over any closed curve γ in Δ is zero.

Therefore, ∫ .

Theorem 7:

Let f(z) be defined in the region Δ’ obtained by omitting a finite

number of points from an open disc Δ. If f(z) satisfies the condition
i for all i then ∫ holds for any closed curve γ
in Δ’.

Proof: Let O be the centre z0 = x0 + i y0 and and P be any point inside Δ. We

define F(z) by F(z) = ∫ where we let the curve ζ not passing
exceptional points. Assuming first that no lies on the line x = x0 and y = y0 by
letting ζ consists of three line segments as in the figure with the last segment is
vertical and consider ζ’ with the last segment is horizontal ( F(z) is independent
of the choice of the middle segment)

D P(x,y)

σ
AA B
C
C

FIG. 4-5

By Theorem 4 and Theorem 5 , ∫

110
 ∫ ∫

∫ ∫

∫ ∫

F
It is easy to verify that  if (z ) , Hence
y

F is analytic is exact differential

∫ = 0 for any closed curve γ.

4.3 Cauchy’s Integral Formula

It enables us to study the local property of an analytic function.
.
Lemma 8: (The index of a point with respect to a closed curve)
If the piecewise differentiable closed curve does not pass through the
point a, then the value of the integral ∫ is a multiple of

Proof
The equation of is

Let us consider the function ∫

It is defined and continuous on the closed interval [ ] and it has the derivative

when ever is continuous

Consider, [ ]

= 0 except perhaps at a finite

number of points , (using (1)).

111
 = a constant = k (say)…….(2)
When t = α, h(α) = 0 .Therefore, =k

Therefore, (2) becomes .

When t = β , . Since γ is a closed curve, z(α) = z(β).

Therefore, . h(β) is a multiple of 2πi.

Hence, h(β) = ∫ = multiple of 2πi.

Therefore, ∫ = h(β) = multiple of 2πi.

Definition: ( The index or winding number)

The index of the point a with respect to the curve by the equation

( ,a) = ∫ . It is also called the winding number of with respect to a.

Properties of winding number:

Property 1. Prove that (- ,a) = - ( ,a)
Proof:

( ,a) = ∫ = - ∫ = - ( ,a).

Property 2. ( ,a) = 0 for all closed curves γ in a disc ( or circle ) and for all
points of a outside the disc.

Proof: is analytic inside the disc. ( as a lies out side the disc )

Therefore, ∫ for all closed curve γ in the disc. ( by Theorem 6)

( ,a) = ∫ = 0.

Remark: As a point set γ is closed and bounded. Its complement is open. The
complement of the point set γ can be represented as a union of disjoint region.

112
If the complement regions are considered in the extended plane, there is exactly
one which contains the point at . Consequently, γ determines one and only
one unbounded region.
Property 3:
As a function of ‘a’ the index ( ,a) is constant in each of the region
determined by , and zero in the unbounded region.
Proof:
Join a and b by a line segment not intersecting outside the line segment
log ( is analytic whose derivative is .

Therefore, ∫ =0 ⇒∫ =∫

⇒ ∫ = ∫ ⇒ ( ,a) = ( ,b).

If | | is sufficiently large, is contained in a disc | | < <| |

Therefore by Property (2), ( ,a) =0
⇒ ( ,a) = 0 in the unbounded region.
NOTE 1:

We know that, ∫ = where C is a circle about a.

⇒ when =C, ∫ =1 ⇒ ( ,a) = 1.

Theorem 9: (The Integral Formula )

Suppose that f(z) is analytic in an open set ∆, and let be a closed

curve in ∆ for any point a not on . ∫ ----------->(A)
where is the index a with respect to .

Proof: Given f(z) is analytic in open set ∆. Also given that a closed curve in
∆ and a point a ϵ ∆ which does not lie on .

Define F(z) =

113
This function is analytic for z ≠ a and for z it is not defined.

But it satisfies the condition i i [ ]

Therefore By theorem (7) ∫ ---------->(1)

∫ ( )

∫ ∫

∫ ∫

Therefore, ∫

If a then

Since is analytic in then we h ve ∫ [by Cauchy’s

theorem for circular disc]

Therefore . Hence equation (A) is true for all a

Note: In the special case , we have ∫ and this

gives a representation formula to compute f(a) as soon as the value of f(z) on
is given, together with the fact that f(z) is analytic in . This is called Cauchy
representation formula. By the change of notation, we write
∫ . This is referred to as Cauchy’s integral formula.

Example 1 Compute ∫| |

Solution: ∫| |

114
 = [since ]

Example 2 Compute ∫| |
by decomposition of integral into partial
fraction.

Solution:

Put z=i, and z = -i,A=

∫| | ∫| |
* +

[ ∫| | ∫| |
]

[ ] where C is the circle, |z| = 2.

[ ] .

| |
Example 3. Compute ∫| |
under the condition | | .
| |

Solution: Given | | ⇒ ⇒
⇒ ̅̅̅ Then | | ̅̅̅

| |

| |
∫| | | |
∫| | ̅ ̅

∫| | ̅ ̅ ̅ ̅

∫| | | | ̅ ̅ ̅

115
 ∫| |
[ | | ]
[ ̅ ̅ ̅ ]

∫| | ̅ ̅

∫| | ̅

∫| | ̅

̅
∫| |
( ̅
)

( ̅) ( ̅)

. /
̅

̅
Put , ( ) | | and
̅ ̅ | |

| | ̅
Put ̅
( ̅
) ( ̅
) and | |

| |
∫ ∫
| | ̅
(
| | | | ̅)

[ ∫ ∫ ]
̅
| | | | ̅

0 . / . /1
̅ ̅ ̅

̅
* ( ̅
)+ ⟶ where C is | |

116
Since , | | ⇒| | | |

Case (i) If | | ⇒ a lies with in the circle C.

| ̅| | |
Therefore ̅
lies outside the circle C.

Therefore ( ̅
)

| |
Therefore becomes ∫ = =
| | ̅ ̅
| |

̅
[ | | ]
̅ | |

= | |

Case (ii) If | | ⇒ a lies outside the circle C.

| |
̅ | |

⇒ ̅
lies inside the circle C. ⇒ ( ̅
)

| |
Therefore ∫| | | |
= ̅
( ̅
)

̅
̅ | | | |

HIGHER DERIVATIVES:

The Cauchy integral formula gives us an ideal tool for the study of the local
properties of an analytic function.

117
Lemma 10: Suppose that is continuous on the arc Then the function ,
∫ is analytic in each of the region determined by and its
derivative is .

Proof: We first prove that (z) is continuous.

Let be a point not on and choose the neighborhood | |< so that it

does not meet . Choose the restricted neighborhood | | < . Then for all
| |=| | =| |

| |–| | ⁄ = ⁄

Therefore, | | ⁄ ………………………(1)

(z) - ( =∫ -∫

[( ]
=∫

=∫

| z | = |∫ |

| || || |
∫ | || |

| || |
=| |∫
| || |

| |
. M∫ | | where | | M

| z | | | L, where L=∫ | |

118
As z | z – |

z is continuous at . Since is arbitrary , z is continuous for all

z not on .

To Prove: z is analytic.

z =( )∫

=∫

i =∫ i

i =∫ ’( ) = ( )

The derivative of ( ) exists at and since is arbitrary, z is analytic

for all z.

The general case is proved by induction.

Suppose z is analytic and z = (n-1) z

Consider z =∫ -∫

=∫ -∫

=∫ -∫

=∫ +∫ - ∫

=∫ -∫ +∫ …(2)

119
 Take G( ) = , (2) becomes,

z =∫ -∫ + ∫ …….(3)

Let (z) = ∫

Therefore (3) becomes

z = (z) - ( +∫

By induction hypothesis applied to G( ),

(z) ( as z and the factor of z - is bounded in a

neighborhood of [ as (z) is continuous]

Therefore z 0 as z .

z z is continuous at a point

To Prove: z is analytic

= +∫

= +∫

When z , = )+∫

= )+

= (n-1) + ….(4) [ ) = (n-1) ]

∫ =∫ =∫ =

120
Therefore becomes,

Since is arbitrary, is analytic and .

Lemma: 11 Prove in detail an analytic function has derivatives of all orders

(or) An analytic function defined in a region Ω has derivatives of all orders
and these are analytic in Ω.

Proof: Let a Ω and f(z) be analytic in Ω. Consider a -neighbourhood ∆

about a and in ∆, for all z inside C, (C, z) = (C, a) = 1

Hence by Cauchy’s Integral Formula, f(z) = ∫

By the above Lemma 10, the integral on the R.H.S is analytic function where

derivative is ∫ . Therefore, ∫

By the same lemma, the integral on the R.H.S is analytic function. Therefore,
whenever f(z) is analytic in Ω then is also analytic in Ω.

Therefore, ∫

∫ are all analytic functions.

Theorem 12 (Morera’s Theorem)

If f(z) is defined and continuous in a region Ω, and ∫ for

all closed curve Ω, then f(z) is analytic in Ω.

121
Proof: Given ∫ for all closed curve Ω.

=> ∫ , with continuous f, depends only on the end pts of .

=> f is the derivative of the analytic function in Ω.

=> f is analytic in Ω. [by lemma 11]

Theorem 13 (Liouvilles’s theorem)

A function which is analytic and bounded in the whole plane must reduce
to a constant.

Proof: Let a Ω and C is any circle of radius r with centre a [ for

all z inside of C]

Now, f(a) = ∫

∫ [by lemma 3]

| || |
Therefore, | | ∫ | |

∫ | | where |z-a| = r and |f(z)| for all z Ω

.

This is true for all circle with centre a. Letting r => | |

=> for all a. => f(z) is a constant.

122
 Theorem 14 ( Fundamental theorem of Algebra )

Every polynomial of degree 1 has atleast one root.

Proof: Let us assume that the polynomial P(z) is of degree n 1 has no root.

Therefore, P(z) never vanishes in the complex plane.

=> is analytic everywhere =>

=> For every > 0 there exists a > 0 such that | | for |z| >

Since is continuous in the bounded closed domains |z| .

Therefore, there exists a number K such that | | | |

Let M = max( => | | for all z Ω

Therefore, by Leouvilles theorem, P(z) is constant.

This is a contradiction.[since P(z) is not a constant]

P(z) must be zero for at least one value of z.

=>The equation P(z) must have at least one root.

Theorem 15 (Cauchy’s Estimate)

Let f(z) be analytic in a region Ω and consider a circle C , |z-a|=

contained in Ω. If |f(z)| M on C then | |

123
Proof: We know that ∫

| || |
=>| | ∫ | |

Given |f(z)| on C and |z-a| =

=>| | ∫ | |

This is known as Cauchy’s Estimate.

124
 UNIT V

LOCAL PROPERTIES OF ANALYTIC FUNCTION

5.1 Removable singularities, Taylor’s Theorem:

Cauchy’s integral formula remains valid in the presece of a finite

number of exceptional points, all satisfying the fundamental condition of
theorem 5, provided that none of them coincides with a.

Cauchy’s formula provides us with a representation of f(z) through an

integral which in its dependence on z as the same character at the exceptional
points as everywhere else. Points with this character are called removable
singularities.

Theorem 1:
Suppose that f(z) is analytic in the region Ωʹ obtained by omitting
point a from a region Ω. A necessary and sufficient condition that there exists
an analytic function in Ω which coincides with f(z) in Ωʹ is that
i The extended function is uniquely determined.

Proof: Suppose f(z) is analytic in Ωʹ and suppose there exists an analytic

function F(z) in Ω such that F(z) = f(z) in Ω’.

To prove: i

Now, F(z) is analytic in Ω => F(z) is continuous on Ω.

=> F(z) is continuous at a in Ω => given there exists such that

|F(z) - F(a) | < whenever |z - a| < -------->(1)

Now, F(z) = f(z) in Ωʹ[i.e. z ≠ a ]

Therefore, (1) becomes |f(z) - A | < whenever |z - a| < where A = F(a)

i [ since F(a) = i i ]

And i i i

125
 =0.A=0

Let be the extended function in Ω where z ≠ a

, (i.e. in Ωʹ) (i.e. when z ≠ a, )

Moreover, i i

Therefore i = . Then for all z in Ω.

Therefore, the extended function is unique.

Conversely, Let a be an exceptional points and i

We draw a circle C about a so that C and its inside are contained in Ω.

Therefore by Cauchy’s Integral Formula, f(z) ∫ for all z≠a.

But the integral on the R.H.S represents an analytic function of z throughout the
inside of C.

Consequently, the function which is equal to f(z) for z ≠ a and which has the
value ∫ for z = a, is analytic in Ω and denote it by f(a).

Therefore the extended function is F(z) {

for z

Theorem 2: Taylor’s Theorem(Finite development)

If f(z) is analytic in a region Ω, containing a, it is possible to write

, where is analytic in Ω.

Proof: Define for z ≠ a

126
 i i

[ f is analytic in Ω ⇒ continuous in Ω]

i i

Hence there exists an analytic function f1(z) = {

Repeating this process we can define an analytic function

2 and so on.

The recursive scheme by which is defined and can be written in the form

From these equations which are trivially valid for z = a and we obtain

Differentiating n times and setting z = a we get,

for all n.

127
Note:

This finite development which is the most useful for the study of the local
properties of f(z).

Since is analytic in Ω , therefore by Cauchy Integral Formula, we

have

∫ ------------>(1)

where C is the circle about a so that C and its inside are contained in Ω.

Using Taylor’s theorem,

Therefore becomes,

∫ [ -

( )
]

∫ ∑ ∫ ------------>(2)

∫ ∫

[ ]
∫ ∫

128
 = 0 identically for all z inside of C.

By lemma 10,

When n = 1, . Therefore

Similarly, we get for all r ≥ 1

Therefore from (2) ∫

This representation is valid inside of C.

Zeroes and poles

Theorem 3:

Let f(z) be defined and analytic in the region Ω. Suppose for some point
a Ω, f(a) and the derivative all vanish. Then f(z) 0 on Ω.

Proof:

Let C be a circle with centre a and radius R in the region Ω.

By Taylor’s theorem,

for all n

where is analytic in Ω.

By hypothesis, we have f ⟶ and

(i.e.) the circumference C has to be contained in Ω in which f(z) is defined and

analytic.

129
f( is continuous on C and C is compact. ⇒ f is bounded on C

(i.e.) |f( | on C.

| || |
Therefore | | ∫ | || |

| |
∫ [ | | ]--------------->(2)
| |

| | | | | |

| | | | | |

| | | |

Therefore becomes,

| | ∫ | |
| |

| |

| |

Therefore | | | |

| | | || |

| |
| |

| |
( ) | |

| | | |
Therefore ⇒

130
 ⇒ inside of C ---------->(3)

T.P

Let E1 be the set on which f(z) and all its derivatives vanish and E 2 be the set
on which the function (or) one of the derivatives different from zero.

is open, is open because the functions and all its derivatives are
continuous.

Now, Ω . But Ω is connected either or

Suppose , then the function and all its derivatives can never vanish at any
point. This is a contradiction. Ω f(z) on Ω.

DEFINITION:

Let f(z) 0. If f(a) 0 then there exists a least positive integer h such that
(a) 0. Then a is the zero of order h.

NOTE: By the previous result, there are no zeros of infinite order.

Lemma 4 If a is a zero of order h then f(z) (z) where (z) is

analytic and (a) 0.

Proof: Given: f(z) is analytic in Ω. By Taylor’s theorem,

f(z) = f(a) + (z-a) +……+ + (z)

[ +…………..+ [ f(a)=f’(a)=……=

131
f(z) = (z) where (z) = +……..+ (z) is
analytic.

T.P: (a) 0

(a) = .

ISOLATED POINTS:

THEOREM 5:

The zeros of an analytic function which does not vanish identically are isolated
(or) The zeros are isolated.

Proof: Let f(z) be analytic function and let f(z) 0. Let z = a be a zero of order
h. f(z) = (z) where is analytic and .

Since is continuous, in a neighborhood of a and z = a is the only

zero of f(z) in this neighborhood.

COROLLARY 6:

If f(z) vanish on a set with an accumulation point in Ω then f(z) 0

Proof: Let S = {z Ω f(z) = 0} and S has an accumulation point a in Ω.

There exists a sequence ( ) in S such that a as n

f( ) as n . Since ( ) f( ) = 0 n f (a) = 0.

Claim: f(z) 0

Suppose f(z) ≢ 0. But then zeros are isolated

132
 (ie) not isolated not zeros and accumulation point not isolated

a is an accumulation point a is not a zero of f(z)

f(a) 0.This is a contradiction to f(a) = 0.Therefore f(z) 0 on .

THEOREM 7 (UNIQUENESS THEOREM )

If f(z) and g(z) are analytic in Ω and if f(z) = g(z) on a set which has an
accumulation point in Ω, then f(z) g(z) on Ω.

Proof: Given that f(z) = g(z) on a set which has an accumulation point in Ω

f(z) - g(z) = 0 (ie) (f-g)(z) = 0 on a set which has an accumulation point in Ω

By previous corollary, (f-g)(z) 0 on Ω f(z) g(z) on Ω.

NOTE:

1. If f(z) 0 in a sub-region of Ω, then f(z) 0 on

2. If f(z)=0 on arc f(z) 0 on Ω
3. An analytic function is uniquely determined by its values on any set with
its accumulation point in the region of analyticity [refer uniqueness
theorem]

DEFINITION ( Isolated Singularity)

Consider a function f(z) which is analytic in neighborhood of a except perhaps at

a itself then the point a is called an isolated singularity.

In other words, f(z) shall be analytic in a region o < |z-a| < then the point a
is called an isolated singularity of f(z).

DEFINITION: (Removable singularity)

If the function is not analytic at ‘a’ but can be made analytic by merely assigning
a suitable value to the function at a point a in region Ω.
133
DEFINITION: (Regular)

If a is the removable singularity and if f(z) is analytic in some(deleted)

neighborhood a then f(z) is said to be regular at a.

NOTE: 1. Regular is sometimes used as a synonym for analytic.

1. If a is taken as a Removable singular point then we can define f(a), so that

f(z) becomes analytic in the disc |z-a| < .

DEFINITION: (Pole)

If f(z) has an isolated singularity at z=a and f(z) at z a. Then f(z) is said to
have a pole at z=a

NOTE:

1. If a is a pole of f(z), we said f(a)= there exists such that f(z) 0

for 0 < |z-a| < [Since f(z) is analytic in region 0 < |z-a| < ]. In this
region the function g(z) = is defined and analytic.

But the singularity of g(z) at a is removable and g(z) has an analytic

extension with g(a) = 0.

Since g(z) does not vanish identically zero and so a is a zero of g(z) of
finite order. We write g(z) = (z) with (a) 0.

The number ‘h’ is called the order of the Pole and

f(z) = = = where = is analytic

and different from zero in a neighborhood of a.

DEFINITION: If f(z) has a pole at z = a and f(z) = (z) where

(z) is analytic and different from zero in a neighborhood of a. Then h is the
order of the pole of f(z) at z = a.

134
 DEFINITION:

A function f(z) which is analytic in a region Ω, except for poles, is

said to be meromorphic in Ω.

NOTE:

1. More precisely, to every a Ω there exists a neighborhood |z-a|<

contained in Ω, such that f(z) is analytic for 0 < |z-a |< and the isolated
singularity is a pole.
2. By definition, the poles of a meromorphic function are isolated.
3. The quotient of two analytic function in Ω is a meromorphic function
in Ω, provided that g(z) is not identically zero. The only possible poles
are the zero of g(z). But a common zero of f(z) and g(z) can also be a
removable singularity. In this case the values of the quotient is
determined by continuity. The sum, the product, the quotient of the two
meormorphic functions are meromorphic.

Detailed discussions of Isolated Singularity:

Consider the condition, (i) |f(z)| = 0

(ii) |f(z)| = for real values of ‘ .

If (i) holds for certain , it holds for all larger and hence for some integer m.
Then f(z) has a removable singularity and vanish for z = a.

Either f(z) ≡ 0, in which case (i) holds for all or f(z) has a zero of
finite order k. In the later case it follows at once that

(i) holds for all > h = m - k

(ii) holds for all < h.

The discussion shows that there are 3 possibilities:

1) Condition (i) holds for all and f(z) vanishes identically.

135
 2) There exists an integer h such that (i) holds for all > h and (ii) for h
3) Neither (i) nor (ii) holds for any .

CASE 1: Trivial

CASE 2: h may be called the algebraic order of f(z) at a. It is positive in the

case of a pole and negative in the case of a zero and zero if f(z) is analytic but
f(z) not equal to zero at a. The order is always an integer. In the case of a pole
of order h, apply Taylor’s theorem to an analytic for f(z). We have,

f(z) = + (z-a) + +………….+ +

φ(z) , where φ(z) is analytic at z = a.

For z ≠ a, we have

f(z) = + +………….+ + φ(z)

The part of this development which proceeds φ(z) is called the singular part of
f(z) at z = a.

Therefore A pole has not only an order but also a well defined singular part.

In the case (3) the point a is an essential singularity. In the neighborhood of an

essential singularity f(z) is at the same time unbounded and comes arbitrary
close to zero.

Note: The difference of two functions with the same singular part is analytic
at a.

CHARACTERIZATION OF THE BEHAVIOUR OF A FUNCTION IN

THE NEIGHBORHOOD OF AN ESSENTIAL SINGULARITY:

Theorem 8 ( WEIERSTRASS THEOREM)

An analytic function comes arbitrarily closed to any complex value in every

neighborhood of an essential singularity.

136
Proof: If the assertion were not true, we could find a complex number A and a
> 0 such that |f(z)-A| > in a neighborhood of a (except for z = a)

For any < 0, we have then i | | | | . Hence a would

not be an essential singularity of f(z) – A.

Accordingly, there exists a with i | | | | = 0 and we are

free to choose > 0.

Since i | | | | 0 =0 [ |f(z)-|A|| ≥ |f(z)-A| ]

a would not be an essential singularity of f(z).

This is a contradiction [ as a is an essential singularity of f(z)].

Theorem 9 (LOCAL MAPPING)

Let be the zeros of a function f(z) which is analytic in a disc ∆ and does not
vanish identically, each zero being counted as many as its order indicates. For
every closed curve in ∆ which does not pass through a zero
dz where the sum has only a finite number terms not equal to zero.

Proof: Given that f(z) is analytic and not identically zero in an open disc ∆ and
also given that is a closed curve in ∆ such that f(z) ≠ 0 on .

Case 1:

If f(z) has only a finite number of zeros in ∆,

Let them be z0, z1, z2,…... where each zero is repeated as many times as its
order indicates.

By the repeated application of Taylor’s theorem we have,

137
f(z) = (z- ) (z- )……………… (z- )g(z) where g(z) is analytic and g(z) ≠ 0
in ∆

Taking log on both sides,

log f(z) = log(z- ) + log(z- ) +…………+ log(z- ) + log g(z).

Differentiate with respect to z,

= + + ……………… + +

dz = + +…...+ +

dz

= ( , )+ ( , )+……….+ ( , )+ dz

= (1)

Since g(z) ≠ 0 and g(z) is analytic (z) is analytic.

Now, is analytic in ∆ and is a closed curve in ∆.

By Cauchy’s theorem on a disc, dz=0

(1) becomes =

CASE 2: Suppose f(z) has infinitely many zeros in ∆.

It is clear that is contained in a concentric disc ∆’ smaller than ∆.

138
If f(z) ≢0 , it has only a finite number of zeros in ∆’.

For, If there were infinitely many zeros they would have an accumulation point
in the closure of [By Balzano – Weierstrass theorem]. This is impossible.
The zeros outside of satisfies = 0 and hence do not contribute to the
sum in (1). Hence the theorem.

Observation 1: = ∫ yields a formula for which the total

number of zero enclosed by .

For, Applying the transformation w = f(z). Let 𝛤 be image of γ under this

transformation.

∫ ∫

= = . That is, (𝛤,0) = .

If each ( , ) must be either be 0 or 1 then the formula in Theorem 9

= ∫ yields a formula for which the total number of zero
enclosed by . This is evidently the case is the circle.

Observation 2: Let a be an arbitrary complex value. Apply Theorem 9 to

f(z) - a. The zeros of f(z) - a are the roots of the equation f(z) = a and we denote
them by (a).

Therefore, by Theorem 9, =

But ∫ = ∫ = ∫ = (𝛤,a)

(𝛤,a) = . It is necessary to assume that f(z) ≠ a on .

139
Observation 3: If a and b are in the same region determined by 𝛤 then
(𝛤,a) = (𝛤,b) = . If is a circle, It follows
that f(z) takes the values a and b equally many times inside of .

THEOREM 10

Suppose that f(z) is analytic at , f( ) = and that f(z) - has a zero of order
n at If > 0 is sufficiently small, there exists a corresponding δ > 0 such that
for all a with |a - < δ the equation f(z) = a has exactly n roots in the disc
|z - z0| < ϵ.

Proof: We can choose so that f(z) is defined and analytic for |z- | ≤ and so
that is the only zero of f(z) - in this disc. Let be a circle |z- | = and 𝛤
its image under the mapping w = f(z). Since Complement of the closed set
𝛤, there exists a neighborhood |w - < δ which does not intersect 𝛤. It
follows immediately that all values a in this neighborhood are taken in the same
number of times inside . The equation f(z) = has exactly n coinciding roots
inside of , and hence every value a is taken n times. It is understood that
multiple roots are counted according to their multiplicity. But if is sufficient
small, we can assert that all roots of the equation f(z) = a are simple for a ≠ .
Indeed, it is sufficient to choose so that (z) does not vanish for 0 < |z- | < .

z- plane w- plane

FIG. 5.1 Local correspondence.

140
Corollary 11

A non constant analytic function maps open sets onto open sets.

Proof: Since the image of every sufficiently small disc |z- | < contains a
neighborhood |w - | < δ (By Theorem 10 above)

Therefore, f maps open sets onto open sets. Therefore, f is open map.

Corollary 12

If f(z) is analytic at zo with (z0) ≠ 0. It maps a neighborhood of z0

conformally and topologically onto a region.

Proof: Given f ‘(z0) ≠ 0. Hence n=1, in this case there is a 1-1 correspondence
between the disc │w - w0│< δ and an open subset Δ of │z - z0│< ε. Since the
open sets in the z-plane corresponds to open sets in the w - plane. The inverse
function of f(z) is continuous . Then the mapping is topological. But the
mapping can be restricted to neighborhood of z0 contained in Δ.

... f ‘(z0) ≠ 0 => f is conformal.

5.2 Maximum Principle

Theorem 13 (The maximum principle)

If f(z) is analytic and non constant in a region Ω, then its absolutely value
│f(z)│has no maximum in Ω.

Proof: If w0 = f(z0) is any value taken in Ω then by Corollary 11, there exists a
neighborhood │w-w0│< δ contained in the image of Ω. In this neighborhood
there are points of modulus >│w0│.

Hence │f(z0)│ is not the maximum of │f(z) │ .

Alternative Proof for Maximum Principle:

141
Let z0 be any pt in Ω. Consider a circle γ with centre z0 at radius r.

=> = z0 + reiζ , 0 ≤ ζ ≤ 2π => = ireiζ dζ on γ

( )
By Cauchy Integral formula, f(z0) = ∫ = ∫

= ∫ ( ) ……….(1)

This formula shows that the value of an analytic function at the centre of a
circle is equal to arithmetic mean of its values on the circle subject to the
condition that the closed disc │z-z0│≤ r is contained in this region of analyticity.

(1) => │f(z0)│≤ ∫ ( ) ----------(2)

Suppose that │f(z0)│ were a maximum => │f(z0+reiζ)│≤│f(z0)│

If the strict inequality hold for a single value of ζ it would hold, by continuity
on a whole arc. Then the mean value of │f(z0) + reiζ)│ would be strictly less
than │f(z0)│. Therefore, (2) => │f(z0)│<│f(z0)│

This is a contradiction. Therefore, f(z) must be constantly equal to │f(z0)│ on

all sufficiently small circles │z-z0│= r and hence in a neighborhood of z0. =>
f(z) must reduce to a constant.

Theorem 13’

If f(z) is defined and continuous on a closed and bounded set E and analytic on
the interior of E, then the maximum │f(z)│ on E is assumed on the boundary of
E.

Proof: Since E is compact, │f(z)│ has a maximum on E .

Assume │f(z)│ has a maximum at z0.

Case (i): If z0 is on the boundary, there is nothing to prove.

142
Case (ii): Assume z0 is an interior point of E.

Then │f(z0)│ is also a maximum of │f(z)│ in a disc │z-z0│< δ contained in E.

This is not possible, unless f(z) is constant in the component of the interior of E
which contains z0.

=> By the continuity that │f(z)│ = the maximum on the whole boundary of
that component.

This boundary is non-empty and it is contained in the boundary of E.

Therefore the maximum is always assumed at boundary points.

Schwarz’s Lemma:

Theorem 14 : If f(z) is analytic for │z│< 1 and satisfies the conditions

│f(z)│≤ 1, f(0) = 0, then │f(z)│≤│z│ and │f ‘(0)│≤ 1. If │f(z)│=│z│ for
some z ≠ 0, or if │f ‘(0)│= 1, then f(z) = cz and with a constant c of absolute
value 1.

Proof: Since f(z) is analytic in the disc │z│< 1, Taylor’s expansion about the
origin gives f(z) = c0 + c1z + c2z2 +………+ cnzn +…......

By hypothesis, f(0) = 0 so that c0 = 0

. . . f(z) = c1z + c2z2 +…...............+ cnzn +………. -----------(1)

Consider the function, f1(z) = = c1+ c2z +……… in the unit disc │z│<1,
f1(z) is analytic. If we set f1(0) = c1 = f ‘(0)

(i.e) f1(z) = {

Let z = a be an arbitrary point of the unit disc.

143
We choose ‘r’ such that │a│< r < 1

Since │f(z)│≤ 1, on the circle │z│= r, we have │f1(z)│= ≤ ---------(2)

By the maximum principle , the inequality (2) also holds in this disc │z│≤ r

f ≤ If we let r → 1 we see that │f1(a)│≤ 1

(i.e) │ │≤1 (or) │f(a)│≤ │a│

Since a is arbitrary ,we have │f(z)│≤ │z │ ……….(3) for which │z│< 1

[ In particular, | | | | (given)]

If the equality in (3) holds at a single point it means that attains its
maximum and hence that must reduce to a constant [by maximum
modulus principle]

Therefore | | can hold only when = = (i.e)

where is a real constant or where | | | | . Hence the
theorem.

Cycles and chains

Definition : Chains
Consider the formal sums which need not be an arc
and we can define the corresponding integral ∫ ∫
∫ ∫ such formal sums of arcs are
called chains.
Definition : Cycles
A chain is a cycle if it can be represented as a sum of closed arcs (or) a
chain is cycle if and only if in any representation the initial and end pts of the
individual arcs are identical in pairs.

144
Definition :
A region is simply connected if its complement with the extended plane is
connected

Definition: (Homology)
A cycle in an open set Ω is said to be homologues to zero with respect to Ω if
ε(γ,a) = 0 for all points a in the complement of Ω.

The General statement of Cauchy’s Theorem

If f(z) is analytic in Ω then ∫ for every cycle γ which is

homologues to zero.

5.3 The calculus of Residues

Now the determination of line integrals of analytic functions over a closed

curve can be reduced to the determination of a period. We are thus possessing
of a method which in many cases permits to evaluate integrals without resorting
to explicit calculus.

In order to make this method more systematic a simple formulation known as

the calculus of residues was introduced by Cauchy.

Residue Theorem:

All the results which were derived as consequences of Cauchy’s

theorem for a disc remains valid in arbitrary region for all cycles which are
homologues to zero.

Cauchy’s Integral formula can be expressed in the form, if f(z) is analytic in a

region Ω.

ε (γ,a) f(a) = ∫ dz for every cycle γ which is homologus to zero in Ω .

We now turn to the discussion of a function f(z) which is analytic in a region Ω

except for isolated singularities.
145
Let us assume that there are only finite number singular points denoted by

a1 ,a2 ,…….an. The region obtained by excluding the points aj will be denoted
by Ω’ . To each aj, there exists a δj > 0 such that the doubly connected region is
contained in Ω’. Draw a circle Cj about aj of radius less than δj .

Let Pj =∫ be the corresponding period of f(z).

The particular function has the period 2πi.

Therefore Set Rj = ---------(1)

Now, f(z) - has a vanishing period. The constant Rj which produces this
result is called the residue at the point aj.

Definition: The residue of f(z) at an isolated singularity a is the unique

complex number R which makes f(z) - the derivative of a single valued
analytic function in an annulus 0 <│z-a│< δ.

Note: Notation R = . Since Rj = ∫ where Cj is the circle

about the isolated singular point aj , = ∫

Theorem 15: (Residue Theorem):

Let f(z) be analytic except for isolated singularities aj in a region Ω then

∫ =∑ for any cycle γ which is homologus to

zero in Ω and does not pass through any points aj.

Proof:

Case(i): Suppose there exists only finite number of isolated singularities (say)
, , ……., .

146
Define Ω’ = Ω - ⋃ . Therefore f(z) is analytic in the region Ω’.

Let γ be a cycle homologues to zero in Ω and does not pass through any one of
the aj’s.

Let Cj be a circle with centre aj and radius > 0. Then ∫ is defined

as the residue of f(z) at a singularity aj . Consider a cycle 𝛤 = γ - ∑ Cj,

Now, ε(𝚪,ak) = ε(γ-∑ Cj, ak) = ε(γ, ak) - ε( ∑ Cj ,ak)

= ε(γ, ak)- ∑ ) ε(Cj,ak) = ε(γ,ak) - ε(γ,ak) ε(Ck, ak)

= ε(γ,ak) - ε(γ,ak) = 0

Thus, ε(𝚪,ak) = 0

Let a does not belongs to Ω. ε (𝚪,a) = ε(γ-∑ Cj, a)

= ε(γ,a) - ε(∑ Cj ,a)= ε(γ,a) - ∑ ε(Cj ,a) = 0 – 0 = 0

Therefore is a cycle homologues to zero which does not pass through the aj’s.
(i. e) 𝚪 is a cycle homologues with respect to Ω’.

Therefore by Cauchy general theorem,

∫ = 0 => ∫ ∑
=0

=>∫ = ∑ ∫

=> ∫ = ∑ ∫

=∑

147
Case (ii): Suppose there exists infinitely many isolated singularities to Ω. The
set of points a such that ε(γ, a) = 0 is open . Since γ is compact there is a large
circular disc D such that γ and for a ~ , ε(γ,a) = 0.

Then the set of points b such that ε(γ,b) ≠ 0 equal to -A = where

A = {a ε(γ,a) = 0}.

Therefore - A is a closed and bounded set and hence - A is compact.

Hence, there exists only a finite number of aj such that ε(γ,aj) ≠ 0 and for this
aj case(i) applies.

Therefore ∫ =∑

Note1: In the applications, it is frequently the case that each ε(γ,aj) is either 0
or 1 . we have ∑ ,where the sum is extended over all singularities
enclosed by γ.

2) The residue of f(z) at a simple pole z = a is f(z) = 0.

Problem 1:

Find the poles and the residues at their poles of the following
function .

Solution: Let f(z) =

Poles of f(z) is given by (z-a)(z-b)=0 (i.e) z = a or z = b

(i.e.) z = a and z = b are the poles of f(z).

Residue of f(z) at the simple pole z = a

Residue of f(z) at the simple pole z = b

148
 i

When b = a, is a pole of order 2.

We know that if z = a is a pole of order h and if

f(z) = then

By Taylor’s theorem for

where is analytic

Divide by

∫ ∫ ∫ ∫ where C is the

circle with centre a.

149
The Argument principle:

Theorem 16:

If f(z) is meromorphic in Ω with zeros and poles bk. Then

∫ ∑ ( ) ∑ for every cycle which is

homologues to zeros in Ω which does not pass through any of the zeros or poles.

Proof: Let be the zero of f(z) of order h.

Therefore where ( ) and is analytic.

Taking log and differentiate with respect to z.

Therefore is a simple pole of with residue h.

Let be the pole of f(z) of order .

where and is analytic.

Taking log and differentiate with respect to z

Therefore is a simple pole of the function with residue – .

By Residue theorem, we have ∫ ∑ ( ) ∑

150
Corollary 17 (Rouche’s theorem )

Let be a cycle homologous to zero in Ω and such that is either zero

or one for any point z not on . Suppose that f(z) and g(z) are analytic in Ω and
satisfy the inequality | | | | Then f(z) and g(z) have the
same number of zeros enclosed by .

Proof: Given that f and g are analytic in Ω.

Further, f and g do not have a zero on .

For, f has a zero a on

By hypothesis, | | | | | | 0.This is a contradiction.

Therefore f has no zero on . Similarly, g has no zero on .

f and g can not have a zero on . Let

is meromorphic in Ω.

Let N = number of zeros of F(z) enclosed by .

= number of zeros of g(z) inside of .

Let P = number of poles of F(z) enclosed by .

= number of zeros of f(z) inside of .

Therefore, By the Argument principle, ∫

where = F(

Given | | | |

151
 | |

=>│1 - F(z)│< 1 , z

=> which is contained in the open unit circular disc of centre one
and radius 1.

=> => N - P = 0 => N = P

=> Number of zeros of g (z) inside of = Number of zeros of f (z) inside of

Therefore f and g have the same number of zeros enclosed by .

REMARK: f(z) and g(z) are interchangeable. Therefore, we have the condition
│ f(z) – g(z) │ < │g(z) │ => f and g have the same number of zeros.

Take =>

Therefore │ z

Therefore, g(z) and g(z)+

Therefore │ z

then g(z) and g(z)+

Problem 1:

How many roots does the equation z7-2z5+6z3-z+1=0 have in the disc │z│<1

Solution:

Of the coefficients 1,-2,6,-1,1,the coefficients 6 has the maximum absolute

value.

To use Rouche’s theorem ,

152
Let f(z) = 6z3 and g(z) = z7 - 2z5 – z + 1

On │z│=1, │f(z)│= 6 z3=6

│g(z)│= │ z7 - 2z5 – z + 1 │

< │z│7+2│z│5+│z│+1 = 5

=>│g(z) │< │f(z) │. But f(z)= 6 z3 has 3 roots z = 0

By Rouche’s theorem,

=> g + f and f have the same number of zeros inside the circle │z│=1

=> g+f = z7-2z5+6z3-z+1 = 0 has 3 roots inside of the circle │z│< 1

Problem 2. How many roots of the equation z4 - 6z + 3 = 0 have their modulus

between 1 and 2.

Solution: Consider the circle │z│= 1. Of the coefficients 1,-6,3

6 has the maximum absolute value.

To use the Rouche’s theorem, take f(z)= -6z and g(z)= z4+3

On │z│=1 , │f(z)│= 6│z│= 6 and │g(z)│= │ z4+3│≤│ z4│+3 = 4

Therefore | | | |

But f(z) = 6z has one root z = 0 inside of | |=1

By Rouche’s theorem

=> g+f and f have the same number of zeros inside the circle | |

=> g+f = z4-6z+3=0 has one root inside the circle | |

153
Consider the circle | |

Let f(z) = z4 and g(z) = -6z+3 . On | |

| | | |4 = 16 and │g(z) │≤ 6│z│+3 ≤ 15

Therefore, │g(z) │< │f(z) │. But f(z) = z4

Therefore, f(z) has 4 roots z = 0 inside the circle | |

Hence by Rouche’s theorem ,

f + g and f have the same number of zeros inside of | |

Therefore, f(z) + g(z) = z4 - 6z + 3 = 0 has 4 roots inside of | |

Therefore, the number roots lying between | | and | | is 4 – 1 = 3

5.4 Evaluation of definite integrals:

Type 1:

∫ os sin os sin is a rational

function of os sin

Put z = . dz = =

os , sin

os , sin

Also │z│=1 (i.e) C is the unit circle │z│= 1

∫ os sin ∫ ( ) = ∫
=2

154
Problem 1

Compute ∫ , a>1

Solution: ∫ =½∫

sin e os

Put z = dz = i dɵ => dɵ = =

∫ = ∫ where C is the unit circle |z| < 1

= ∫

= -i2∫

= -2i∫ where f(z) =

The poles of f(z) are given by

Z=

α=-a+ =-a-
αβ = . But |α||β|=1 => |α| =
| |

Now, |β | > 1 => < 1 => |α| < 1

| |

α lies inside the circle C, |z| = 1.

155
 es =

= = =
√

I= ∫

= .2πi [sum of the residue of the poles with in C]

= 4π ( )=

Therefore, ∫ =

∫ =1/2 ∫ = 1/2 =

Deduction: ∫ =

Diff. w.r.to a , -∫ =-1/2 .2a

∫ =

Problem 2

Evaluate ∫ , |a| > 1

Solution:

Take I = ∫ =∫

Put ɵ = 2x. when x = 0 , ɵ = 0 and when x = , ɵ = π then dɵ = 2 dx

156
Let I =∫ =½∫

Put z = , dz = i dɵ and dɵ =

Let C be the unit circle |z | = 1

I=½∫
( )

=1/2∫
. /

=-1/i∫

=i∫

=i∫ where f(z) =

= i 2πi (sum of the residues of f(z) at the poles with in C).

The poles of f(z) is given by

√ √
Z= =

√
= = 2a+1 ±2 √

Let α = 2a+1+2√ β = 2a+1- 2√

|α| = |2a+1+2 | |α||β| = 1

=> |β| = < 1, |β| < 1. Therefore lies with in the circle | |
| |

157
 i

( )

(sum of the residue at the poles with in C)

( )

Note:

If is a pole of order m then

For,

Since a is a pole order m,

Therefore

Therefore

Problem 3. Apply the Cauchy’s residue theorem and prove that

∫ os sin

158
Solution: ∫ os sin

Real part of ∫

Real part of ∫

Real part of ∫

Put z , and C is a unit circle | |

Real part of ∫

Real part of ∫

Real part of ∫ where

Real part of

Poles of

is a pole of order lies within C.

i f(z)) = i . )= .

I = R.P. ∫ = R.P. .2 (Sum of residues at the poles within C)

= R.P 2 . =

Lemma 18:

If i (z-a)f(z) = A and if C is the arc of the circle | | = r then

i ⟶ ∫ = iA( - )
159
Proof:

i (z-a)f(z) = A ⇒ given 0 ,there exists > 0 depending on such that

⟶

| |< for | |< . But | |=r

Therefore, if r < then | | < on the arc C.

Therefore, (z-a)f(z) = A + where | | < ⇒ f(z) =

Therefore, ∫ ∫ = (A+ ) ∫

= (A+ ) ∫ where |z| = r.

= i(A+ )( - )

= i A( - )+i ( - )

|∫ | | | = | || |< -

Since ⇒ 0. Therefore, i ⟶ ∫ = iA

NOTE:

1) In particular, if A = 0 then i ⟶ ∫ =0

2) i z-a)f(z) = f(z)
⟶

Lemma 19:

If C is the arc of the circle | | = R then i ⟶ =A then

160
 i ∫ = iA
⟶

Proof:

i zf(z) = A.Therefore we choose R so large that |zf(z) – A| <

⟶

On the arc C, (i.e) zf(z) = A + where | | <

Therefore ∫ =∫ .Ri

=i(A+ ) (

= Ai + i(

|∫ | | |

= | |( – < ( – )⟶ ⟶

Since ⟶ and consequently R ⟶

Therefore, we get i ⟶ ∫ = iA ( – )

NOTE:

1) In particular, if A = 0 then i ∫ = 0.
⟶

2) Res f(z) = i -z f(z)

⟶

Jordan’s lemma 20:

If is analytic except at finite number of singularities and if as

Then i ∫ where denotes the semi
circle | |

161
Proof:

Choose R so large that all the singularities of f(z) lie with in Γ and none on its
boundary. Since uniformly as | | . Therefore, there exists
such that | | for every on

Now, |∫ | ∫ | || || |

Put

∫ | | | |

= ∫ | |

= ∫ | || |

∫ [ | | ]

( )
∫ [ ]

[ ]

as and

i ∫

Result:

1. If i then i ∫

2. If i then i ∫

162
 3. If C is the arc of the circle | | such that and
i then i ∫

4. If C is the arc of the circle | | such that and

i then i ∫

TYPE II:

Evaluation of integrals of the form ∫ where is analytic in

the upper half plane except at a finite number of points and have no poles on the
real axis.

Above type of integrals are evaluated by integrating around C

consisting of a semi circle of radius R large enough to include all the poles of
and the part of the real axis from –R to R.

Therefore, by Cauchy’s Residue theorem

∫ Sum of the residue of the poles of f z within C

⇒∫ ∫ ∑

where ∑ denotes the sum of the residue of the poles in the upper half.

It can be shown that, if i ⇒ i ∫

Also, i ∫ ∫ and ∫ ∑

Problem 1: Evaluate ∫

Solution:

163
Consider ∫ , where C is the contour consisting of a large semi circle
of radius R along with the part of the real axis from –

∫ ∑

where ∑ Sum of the residue of the poles in the upper half plane

∫ ∫ ∑

i i i ∫

Also, i ∫ ∫ and ∫ ∑

∫ ∑

To find the pole of

Therefore has only one pole z = i of order 2 lies in the upper half of the
plane.

164
 [ ]

∫ ∑

∫ ∫

Problem 2. Evaluate ∫ (a is real and a > 0)

Solution: Here and i i …(1)

Consider ∫ where C is the contour consisting of a large semi circle

of radius R along with the part of the real axis from –

∫ ∑

∫ ∫ ∑

∫ ∫ ∑

165
From (1), We have i ∫

⇒ i ∫

Also, i ∫ ∫

i ∫ ∫

∫ ∑

To find the pole of

⇒ ⇒

Poles of

Therefore is the only pole of order 3 lies in the upper half plane.

i . /

i . /

i . /

i . /

166
 i . /

. /

Therefore ∫

Problem 3. Evaluate ∫

Solution: Here f(z) =

Consider ∫ where C is the contour consisting of a large semi circle of

radius R along with the part of the real axis from –

∫ ∑

where ∑ sum of the residue of the poles in the upper half plane

∫ ∫ ∑

i i . /

Therefore i ∫

167
Also, i ∫ ∫

∫ ∑

∫ ∑

To find the poles of

is a pole of order 1 lies with in the upper half plane.

is a pole of order 1 lies with in the upper half plane

168
 i
[ ]

∫ [ ]

* +

Problem 4. Evaluate ∫

Solution:

Consider ∫ where C is the contour consisting of a large semi circle of

radius R. along with the part of the real axis from –

Therefore ∫ ∑

where ∑ sum of the residue of the poles in the upper half plane

∫ ∫ ∑

169
 ∫ ∫ ∑

i i i ∫

Also, i ∫ ∫ . Therefore ∫ ∑

To find the pole of

( ) where

⁄ ⁄ ⁄ ⁄
Therefore, the poles are

⁄ ⁄
Therefore, and are the only poles with in C.

Let =a = =

̇ ̇
es = (z ) =

̇
= = = = =

Similarly, es ⁄ =

∫ = 2 i∑ = ( )

= ( ) = (2isin )

170
 = (sin ) =

∫ = ∫

= =

Type III :

Evaluation of the integrals of the form ∫ sin dx ,

∫ os dx (m > 0) where (i) P(x) and Q(x) are polynomials (ii).

Degree of Q(x) exceeds that of P(x) (iii). Q(x) has no real roots.

Above type of the integrals are evaluated by integrating ∫ f(z) dz

where f(z) = around a contour C consisting of a semicircle of radius R
large enough to include all the poles of the integrand in the upper half plane
and also part of the real axis from –R to +R

By Cauchy’s residue theorem, ∫ f(z) dz = 2 i∑

∫ f(x) dx + ∫ f(z) dz= 2 i∑

̇
By Jordan’s lemma, ∫ f(z) dz = 0 (m > 0)

[ implies f(z) = ]

̇
∫ f(x) dx = ∫ f(x) dx

∫ f(x) dx = 2 ∑

∫ os sin f(x) dx = 2 ∑

171
∫ os f(x) dx + i ∫ sin f(x) dx = 2 ∑

Equating the real and imaginary roots, we get the values of the given integral.

∫ dx, a real and ∫ dx , a real

Solution:

Consider ∫ dz = ∫ f(z) dz where C is the contour consisting of a

semicircle of radius R large enough to include all the poles of f(z) in the
upper half plane and also the part of the real axis from to .

Therefore, By Cauchy’s residue theorem,

∫ f(z) dz = 2 ∑

∫ f(x) dx + ∫ f(z) dz = 2 ∑

∫ dx + ∫ dz = 2 ∑

̇ ̇ ̇
Since f(z) = = 0. By Jordan lemma, ∫ dz = 0

̇
Also ∫ dx = ∫ dx

Therefore, ∫ dx = 2 ∑

To find the poles of f(z)

Z= are the poles of f(z) and Z = ai is the only pole lies within C.

̇ ̇
f(z) = (z - ai) f(z)

172
 ̇
= (z - ai) =

∫ dx = 2 ∑ =2 =

∫ dx + i∫ dx =

Equating real and imaginary parts, ∫ dx = (1)

∫ dx = 0

When m=1, ∫ dx =

∫ dx = ∫ dx = =

Differentiate (1) w.r.to m, ∫ xdx

= (-a)

Put m = 1, ∫ dx=

∫ = ∫ dx

Problem 2. ∫ dx =

Solution:

Consider∫ , where f(z) = . It has a singularity at z = 0 on the real

axis . Let the contour C consists of large semicircle | | indented at z = 0

173
and be the radius of this small semicircle of identation. Now there is no
singularities within C.

-R R
Fig.5.2

By Cauchy’s residue theorem, ∫ =2 ∑ =0

∫ ∫ ∫ ∫ =0

̇ ̇ ̇
∫ = 0 , since =0

̇
∫ = 0 [By Jordan lemma]

Consider ∫ where is the circle | | =

̇ ̇
=1

̇
∫ = = =

The negative sign is taken because the orientation is clockwise

Taking R and

174
∫ ∫ =0 ∫

∫ ∫ ( ) =i

Equating real and imaginary parts, ∫ = 0 and ∫ =

Therefore ∫ ∫

Type IV :

Evaluation of the integrals of the type ∫

Problem 1: ∫ and hence deduce that ∫ and

∫ og
og

Solution: Consider ∫ , where f(z) = , 0 < a < 2. z = 0 is a

singularity of f(z) for 0 < a < 1.

The poles of f(z) are given by z = [ ]

The contour C consists of the large semicircle | | = R in the upper half plane
and the real axis from to intented at z = 0 and by a small circle ɣ of
radius , the only pole lying within C is z = i.

es i

i i

= = [as i = ]

175
 γ

-R R
Fig.5.3

By Cauchy’s Residue Theorem, ∫

∫ ∫ ∫ ∫

= ( ) ....(1)

i i

i ∫

Also, i i i

= i [ ]

i ∫

Taking Limit and R in (1), We have

176
 ∫ ∫

∫ ∫ …..(2)

Consider ∫

Put x = -y ; x = - ;y= ;x=0;y=0

Therefore ∫ ∫

= ∫ ∫ [-1= ]

Substitute in (2), ∫ ∫

( )∫

( )∫

i sin ∫

∫ ose …...(3)

Diff. w.r.t. a, ∫ ose ot =

177
Diff. again w.r.t. a ∫

Put a = 1, ∫

Integrating (3) w.r.t. a,

∫ ( ) = og t n …..(4)
( )

Also ∫ og t n …..(5)

Now, (4) - (5)

∫ = log tan - log tan

= log

Problem 2. ∫ , 0<α<1

Solution:

Consider ∫ where f(z) = where C is the semi-circle

given by |z| =R where R is every large intended at z = 0 by a small semi-circle
of radius and the real axis from –R to R.

Therefore, f(z) is analytic with C.

Therefore, By Cauchy residue theorem, ∫ =0

∫ + ∫ + ∫ + ∫ = 0 ……..(1)

178
i = i

-R R

Fig.5.4

= i [Since, |z|<1]

= 0

i ∫ =0

i = i

= i log

= i

= i [ ]

= 2 i + i ...........)

= 0 [Since, i = 0]

179
i ∫ = 0 [by lemma (2)]

Taking limit R and 0 in (1)

∫ +0+ ∫ +0=0

∫ + ∫ =0 ……….(2)

Consider I = ∫

Put x = -y When x= - , y= and x= 0, y= 0

Therefore, I = ∫

= ∫ [since -1 = ]

= ∫

= ∫

= - ∫ …………… (3)

Sub (3) in (2)

0= ∫ + ∫

Therefore, (1- )∫ = 0

Therefore ∫ = 0

180
Problem 3. Evaluate ∫ and ∫

Solution:

Consider ∫ , where f(z) = , z = 0 the branch point of f(z)

and poles of f (z) is given by z = +i and z = -i.

Let C be the large semi-circle, |z| = R indented at z = 0 and the radius of

small semi-circle of indentation. The pole z = i lies within C.

es = i

= = = - [Since log i = i ]

Therefore, By Cauchy Residue Theorem,

Therefore, ∫ =2 i∑ +

= 2 i =

∫ + ∫ + ∫ + ∫ =0 …….(1)

i = i = i

= i i

= i = i

= 2 ( 0) [ Since, i =0]

181
 = 0

Therefore, i ∫ =0

i z f (z) = i

Put z =

Therefore, i z f(z) = i = i

= i = 0 (as above )

Therefore, i ∫ =0

Applying 0 and R we get

(1) ∫ + ∫ = - 3
/4

∫ + ∫ = …………….(2)

Consider I= ∫

Put x = -y when x = - , y= and x = 0 , y =0

I = ∫

= ∫

I=∫ [Since, log (-1) = ]

182
 = ∫

2
= ∫ +∫ +2 ∫

= t n + ∫ +2 ∫

= +∫ +2 ∫ ….(3)

From (2) and (3)

+∫ +2 ∫ +∫ =

Equating Real and Imaginary Part

+2 ∫ =

2∫ = + =

∫ = and ∫ =0

TYPE: V Integrals involving many valued function:

Problem 1:

Evaluate ∫ og sin x dx

Solution:

Consider ∫ where f (z) = log (1- ) and choose the contour C

as follows , within the contour C , f(z) is regular.

183
 1 2

Fig .5.5

By Cauchy Residue Theorem, ∫ = 0, where f (z) = log (1- )

∫ +∫ + ∫ +∫

+∫ +∫ =0

I1+ I2 + I3 + I4 + I5 + I6 = 0

Consider I3 = ∫

= ∫

=∫

I5 = ∫

= ∫

=∫

184
As n , 1, 2 0

I3 ∫

I5 ∫ =-∫

I3 + I5 0

Therefore i z f (z) = i z log ( )

= i

( )
= i

= i (By L’Hopital rule)

= i = 0

i ⟶
∫ .

Similarly, i ⟶
∫ =0

Consider I4 = ∫

= ∫ 0 as n

As n , 1, 2 0 we get ∫ .

∫ =0

= ( = (- 2i sinx)
185
∫ i sinx =0

∫ + ∫ og +∫ +∫ =0

log2 + log(-i) +∫ +∫ =0

log2 - + ( ) +∫ =0

log2 - + +∫ =0

186