1 Complex Introduction

Complex Analysis
Note: This module is prepared from the text book (Complex Variables and Applications by R.V.
Churchill and J.W. Brown, 8th Edition, 2009, TMH) just to help the students. The study material
is expected to be useful but not exhaustive. For detailed study, the students are advised to attend
the lecture/tutorial classes regularly, and consult the text book.
Appeal: Please do not print this e-module unless it is really necessary.
Dr. Suresh Kumar, Department of Mathematics, BITS Pilani, Pilani Campus
1
Complex Analysis Dr. Suresh Kumar, BITS Pilani 2
Chapter 1
Complex Numbers
Complex numbers are defined as ordered pairs (x, y) of real numbers which are interpreted as points
(x, y) in XY-plane of cartesian system. The complex numbers of the form (x, 0) are points on the
X-axis (usually taken as the real number line) and are called pure real numbers. The complex
numbers of the form (0, y) are points on the Y-axis and are called pure imaginary numbers. The
real numbers x and y are respectively known as the real and imaginary parts of the complex
number (x, y). We denote the complex number (x, y) by z, its real part by Rez, and imaginary
part by Imz. Thus, we have z = (x, y), Rez = x and Imz = y. Further, two complex numbers z1
and z2 are said to be equal, that is, z1 = z2 if and only if Rez1 =Rez2 and Imz1 =Imz2 .
x z= H x ,yL
X
O
Addition and Multiplication of complex numbers

The addition and multiplication of two complex numbers z1 = (x1 , y1 ) and z2 = (x2 , y2 ) are defined
as follows:
z1 + z2 = (x1 , y1 ) + (x2 , y2 ) = (x1 + x2 , y1 + y2 ),
z1 z2 = (x1 , y1 )(x2 , y2 ) = (x1 x2 − y1 y2 , x1 y2 + y1 x2 ).

Note that the above operations reduce to the usual operations of addition and multiplication
when restricted to the real numbers.
(x1 , 0) + (x2 , 0) = (x1 + x2 , 0)
(x1 , 0)(x2 , 0) = (x1 x2 , 0).

The complex number system is, therefore, a natural extension of the real number system.
Simplified representation of complex number

Making use of the definitions of addition and multiplication operations of complex numbers, we
have
z = (x, y) = (x, 0) + (0, y) = (x, 0) + (0, 1)(y, 0)

We shall denote the imaginary number (0, 1) by i. For simplified representation, we write x for
(x, 0) and y for (y, 0). Thus, the simplified representation of z is
z = x + iy.
With this representation, the addition and multiplication operations become
z1 + z2 = (x1 + iy1 ) + (x2 + iy2 ) = (x1 + x2 ) + i(y1 + y2 ),
z1 z2 = (x1 + iy1 )(x2 + iy2 ) = (x1 x2 − y1 y2 ) + i(x1 y2 + y1 x2 ).

Further, we have i2 = −1 since i2 = i.i = (0, 1)(0, 1) = (−1, 0).
Some algebraic properties of complex numbers

It is easy to verify the following properties of complex numbers:
(1) Commutative property: z1 + z2 = z2 + z1 , z1 z2 = z2 z1 .
(2) Associative property: z1 + (z2 + z3 ) = (z1 + z2 ) + z3 , z1 (z2 z3 ) = (z1 z2 )z3 .
(3) Identity: z + 0 = (x, y) + (0, 0) = (x, y) = z, z.1 = (x, y)(1, 0) = (x, y) = z. Therefore, 0 and
1 are respectively the additive and multiplicative identities.
(4) Inverse: For any complex number z = (x, y), there exists a complex number (−x, −y) denoted
by −z and called negative of z such that z + (−z) = 0. We call −z as additive inverse of z. For any
non-zero complex number z = (x, y), there exists a complex number (x/(x2 + y 2 ), −y/(x2 + y 2 ))
denoted by 1/z or z −1 and called reciprocal of z such that zz −1 = 1. We call z −1 as multiplicative
inverse of z. Please also see the footnote1
(5) Distributive property: z1 (z2 + z3 ) = z1 z2 + z1 z3 .
Modulus of a complex number

Recall that the absolute value a real number is its distance from origin on the real number line.
p of a complex numberpz = x + iy, denoted by |z|, is
In analogy, the modulus or the absolute value
defined as the non-negative real number x2 + y 2 , that is, |z| = x2 + y 2 . Geometrically, it is
the distance of the point (x, y) (representing the complex number z) from the origin (0, 0). Some
straightforward but important properties of modulus are given below.
(1) For any two complex numbers z1 and z2 , we have |z1 | < |z2 | or |z1 | = |z2 | or |z1 | > |z2 |. Note
that the expressions z1 < z2 or z1 > z2 are not meaningful unless z1 and z2 are pure real numbers.
(2) The equation |z − z0 | = r represents a circle with centre at z0 and radius r in the complex
plane. The inequality |z − z0 | < r represents the interior region of the circle |z − z0 | = r while
the inequality |z − z0 | > r represents the entire region in the complex plane exterior to the circle
|z − z0 | = r.
(3) Let a point P (x, y) in the XY-plane corresponds to a complex number z = x + iy. If we assign
−→ −→
the position vector OP of P to represent the complex number z, then |z| = |OP |. Considering the
vector representation of complex numbers, one can easily demonstrate the addition and difference
The expression z1 + (−z2 ) denoted by z1 − z2 is called the difference of z1 from z2 . The expression z1 z2−1
1
denoted by zz21 is called the division of z1 by z2 .

x z= H x ,yL
ÈzÈ
y
X
O
of two complex numbers in the complex plane as per the rules of vector algebra. However, note
that the product of two complex numbers can not be associated with the dot product or scalar
product of vectors.
(4) For any two complex numbers z1 and z2 , we have |z1 z2 | = |z1 ||z2 | and |z1 /z2 | = |z1 |/|z2 |
provided z2 6= 0. Further, we have |z1 + z2 | ≤ |z1 | + |z2 |, known as the triangle inequality. It simply
says that the length of any side of a triangle is less than or equal to the sum of the lengths of the
other two sides. Also, it can be proved that |z1 + z2 | ≥ ||z1 | − |z2 ||.
(5) |z|2 = (Rez)2 + (Imz)2 , Rez ≤ |Rez| ≤ |z| and Imz ≤ |Imz| ≤ |z|.
Conjugate of a complex number

The conjugate of a complex number z = x + iy, denoted by z, is defined as the complex number
x − iy, that is, z = x + iy = x − iy. Geometrically, z is the reflection of z in the X-axis. Some
straightforward properties of complex conjugates are given below.
(1) z = z, z1 ± z2 = z1 ± z2 , z1 /z2 = z1 /z2 .
z+z z−z
(2) zz = |z|2 , Rez = 2
and Imz = 2i
.
Polar or exponential form of a complex number

Let (r, θ) be polar coordinates of a point that corresponds to a non-zero complex number z = x+iy
in the complex plane. Then x = r cos θ and y = r sin θ. So we have z = r(cos θ + i sin θ). It is the
polar representation of the complex number z. Using the Euler’s formula2 eiθ = cos θ + i sin θ, we
get a more compact form z = reiθ , known as the exponential form of z. Obviously, the modulus
of z is |z| = r. The angle θ is called the argument of z and is denoted by argz. It is measured in
radians and is undefined for z = 0. Further, for a non-zero complex number there exists infinitely
many values of θ but all differing by an integer multiple of 2π. In the range −π < θ ≤ π, we can
find a unique value of θ for any given non-zero complex number z. It is called principal argument
of z and is denoted by Argz. So we have argz = Argz + 2kπ, k is an integer. Some useful points
are given below.
2
It can be proved that (cos θ + i sin θ)n = cos(nθ) + i sin(nθ), that is, (eiθ )n = einθ . It is known as de Moivre’s
formula
x z= H x ,yL
ÈzÈ = r
y
Θ
X
O
(1) Let z1 = r1 eiθ1 and z2 = r2 eiθ2 . Then z1 = z2 if and if r1 = r2 and θ1 = θ2 + 2kπ, k being an
integer.
(2) The nth roots of a non-zero complex number z0 are the roots of the equation z n = z0 . If
√
z0 = r0 eiθ0 and z = reiθ , then z n = z0 yields rn = r0 and nθ = θ0 + 2kπ. So r = n r0 and
θ = θn0 + 2kπ
n
, and hence the nth roots of z0 are given by

√ θ0 2kπ
z = r0 exp i
n
+ , k = 0, 1, 2, ......, n − 1.
n n
Note that the integer values of k less than 0 and greater than n − 1 do not yield any roots dif-
√
ferent from the listed ones. Obviously, the nth roots of z0 lie symmetrically on the circle |z| = n r0 .

(3) One can easily prove that arg(z1 z2 ) = arg(z1 ) + arg(z2 ) and arg zz12 = arg(z1 ) − arg(z2 ).
However, one should be careful that these equalities hold in the sense that given any two argu-
ments in the equality there exists some value of the third argument satisfying the equality.
(4) Note that z = reiθ (0 ≤ θ ≤ 2π) is parametric representation of the circle |z| = r. As
the parameter θ increases from 0 to 2π, the point z traverses the circle |z| = r once in the
counterclockwise direction starting from the positive real axis.
Regions in complex plane

The following definitions are simple but crucial in the forthcoming analysis. So read and understand
these carefully.
Neighbourhoods
Let z0 be a complex number. Then for any > 0, the interior of the circle |z − z0 | = , that is,
the region |z − z0 | < defines a neighbourhood of z0 . The neighbourhood with z0 deleted, that is,
0 < |z − z0 | < is called deleted neighbourhood of z0 . For example, |z| < 1 is neighbourhood of 0.
In fact, it is neighbourhood of its every point. 0 < |z| < 1 is deleted neighbourhood of 0.
0.5
X
-1 -0.5 0 0.5 1
-0.5
-1
Figure 1: Shaded region stands for |z| < 1, the neighbourhood of z = 0.
Interior, exterior and boundary points

Let S be a set of complex numbers. A complex number z0 is said to be an interior point of S if
there exists at least one neighbourhood of z0 that completely lies within S. On the other hand,
z0 is an exterior point of S if there exists at least one neighbourhood of z0 that completely lies
outside S. Next, z0 is a boundary point of S if it neither an interior nor an exterior point of S.
In other words, if each neighbourhood of z0 contains points from exterior as well as interior of S.
For example, z0 = 0.5, z1 = 1 + i and z2 = i are respectively the interior, exterior and boundary
points of S = {z : |z| ≤ 1}.
Y
1
1+i
0.5
X
-1 -0.5 0 0.5 1
-0.5
-1
Figure 2: z0 = 0.5, z1 = 1 + i and z2 = i are respectively the interior, exterior and boundary points
of S = {z : |z| ≤ 1}.
Open and closed sets

If all points of a set are its interior points, it is said be an open set. A set containing all its
boundary points is a closed set. For example, the |z| < 1 is an open set whereas the set |z| ≤ 1 is
a closed set. Note that the punctured unit disc 0 < |z| < 1 is an open set while the set 0 < |z| ≤ 1
is neither open nor closed. Likewise the annular region 1 < |z| ≤ 2 is neither open nor closed.
X
-3 -2 -1 0 1 2
-1
-2
-3
Figure 3: 1 < |z| ≤ 2 is neither open nor closed.
Connected sets
A connected set is the set in which any two points can be joined by a polygonal line consisting of
finite number of line segments without leaving the set. For example, 1 < |z| ≤ 2 is a connected
set, but the set S = {z : Rez < −1 or Rez > 1} is not connected since points of the region
Rez < −1 can not be joined with the points of the region Rez > 1 by any polygonal line.
Y
X
-3 -2 -1 0 1 2
-1
-2
-3
Figure 4: S = {z : Rez < −1 or Rez > 1} is not connected.
Domains and regions

A non-empty open connected set is called a domain. A region is a domain with some, none or all
the boundary points. For example, 1 < |z| < 2 is domain as well as a region. The set 0 < |z| ≤ 1
is a region but not a domain.
Bounded sets
A set is bounded if it can be enclosed inside the circle |z| = r for some finite value of r. For
example, the set |z| < 1 is bounded while the set Rez > 0 being the right half of the complex
plane is unbounded.
Limit points of a set

A point z0 is said to be limit point or accumulation point of a set S if every neighbourhood of
z0 carries infinitely many points of S. Obviously, S has no limit points if it is a finite set. Next,
all the interior and boundary points of a set are its limit points. A set is closed if and only if it
contains all its boundary points. A set may have finite number of limit points. For example, 0 is
the only limit point of the set {i/n : n = 1, 2, 3....}. Note that 0 is not a member of the set. The
set |z| ≤ 1 is the set of all limit points of the set |z| < 1.
Chapter 2
Function, Domain and Range
Let B be a subset of the set C of complex numbers. Then a function f from B to C, also written
as f : B → C, is a rule that assigns to each member z of S a complex number w in C. The
complex number w is called the value of f at z (or image of z under f ) and is denoted by f (z). So
w = f (z). The set B is called the domain of definition of f , and the set {f (z) : z ∈ S} is called
the range of f .
The domain, here, should not be confused with the domain (open connected set) defined in the
previous chapter. Here, domain appears in analogy with the real functions, and simply stands for
discrete or continuous collection of points in the complex plane permissible for the given function.
For a given function, the set of all permissible points in the complex plane is called its natural
domain. Any proper subset of the natural domain is the restricted domain. We may define
functions on restricted domains. Of course, the domain of a function can be an open connected
set. For example, f : {z : |z| < 1} → C given by f (z) = z 2 + 5 is a function defined on the unit
circular disc |z| < 1. Here |z| < 1 is the restricted domain of f as its natural domain is C. Also,
|z| < 1 is an open connected set.
Since z = x + iy, we can express the function f (z) = z 2 + 5 in terms of x and y as follows:
f (x + iy) = (x + iy)2 + 5 = x2 − y 2 + 5 + i(2xy) = u(x, y) + iv(x, y),
where u(x, y) = x2 − y 2 + 5 and v(x, y) = 2xy are respectively the real and imaginary parts of
f (z). If we choose the polar form z = reiθ , then u and v can be determined as functions of r and
θ.
Also, note that a function of complex variable need not be complex valued. For example,
f (z) = |z|2 = x2 + y 2 with u(x, y) = x2 + y 2 and v(x, y) = 0, is a pure real valued function.
Multiple Valued Functions

Multiple valued functions, which assign more than one value to the domain points, do arise in
complex system as these appear in the real system3 . Such functions are studied by constructing
single valued functions√in a systematic manner. For instance, the function w = z 1/2 assigns two
values namely w = ± reiΘ/2 corresponding to each non-zero complex number
√ iΘ/2 z = reiΘ (−π <
√ iΘ/2
Θ ≤ π). The functions given by f1 (z) = re , f1 (0) = 0 and f2 (z) = − re , f2 (0) = 0 are
both single valued functions defined in the entire complex plane.
Limit at an interior point

Let a function f (z) be defined in some deleted neighbourhood of z0 . Then the limit of f (z) as
z → z0 is said to be w0 denoted as
lim f (z) = w0
z→z0
3
√
The real function√given by x2 + y 2 = √1 assigns two values namely y = ± 1 − x2 to each value of x ∈ [−1, 1].
The functions y1 = 1 − x2 and y2 = − 1 − x2 are single valued functions. The graphs of these functions are
respectively the upper half and lower half arcs of the circle x2 + y 2 = 1
if given any real number > 0 (howsoever small), there exists a real number δ > 0 (depending on
) such that
|f (z) − w0 | < for 0 < |z − z0 | < δ.
Geometrically speaking, the limit of f (z) as z → z0 is w0 if given any -neighbourhood of w0 , there

exists a deleted δ-neighbourhood of z0 such that the images f (z) of all points z of the deleted
δ-neighbourhood of z0 are contained in the -neighbourhood of w0 . Also, existence of the limit of
f (z) as z → z0 means that its value w0 is unique and is independent of the path along which z → z0 .
Ex. Show that lim (2z + 3) = 5.

z→1
Sol. Let > 0 be given. Then we have
|2z + 3 − 5| = |2z − 2| = 2|z − 1| < for |z − 1| < /2.
Choosing δ = /2, we have
|2z + 3 − 5| < for |z − 1| < δ.
∴ lim (2z + 3) = 5.
z→1
z
Ex. Show that lim does not exist.
z→0 z
Sol. We have
z
x + iy x + imx 1 + im
lim = lim = lim = ,
z→0 z (x,y)→(0,0) x − iy x→0 x − imx 1 − im
where we have chosen the path y = mx along which z → 0. Therefore, the given limit depends on
m, that is, on the chosen path and is, therefore, not unique. Hence, it does not exist.
Limit at a boundary point

The definition of the limit at interior point can be extended to the case when z0 is boundary point
of the region where f (z) is defined. In this case, the limit of f (z) as z → z0 is said to be w0 if
given > 0, there exists δ > 0 such that |f (z) − w0 | < for all z common to the δ-neighbourhood
of z0 and the domain region of f (z).
Ex. Suppose f (z) be defined in |z| < 1. Show that lim(iz/2) = i/2.
z→1
Sol. Here z = 1 is boundary point of |z| < 1. Let > 0 be given. Then we have
|iz/2 − i/2| = (1/2)|z − 1| < for |z − 1| < 2 and |z| < 1.
Choosing δ = 2, we have
|iz/2 − i/2| < for |z − 1| < δ and |z| < 1.
∴ lim (iz/2) = i/2.

z→1
Algebra of limits
Let f (z) and g(z) be defined in some deleted neighbourhood of z0 . Then following results can be
proved.
(1) If z0 = x0 + iy0 and f (z) = u(x, y) + iv(x, y), then
lim f (z) = lim u(x, y) + i lim v(x, y).

z→z0 (x,y)→(x0 ,y0 ) (x,y)→(x0 ,y0 )
(2) lim [f (z)g(z)] = lim f (z) lim g(z).

z→z0 z→z0 z→z0
(3) lim [f (z) ± g(z)] = lim f (z) ± lim g(z).

z→z0 z→z0 z→z0
(4) lim [f (z)/g(z)] = lim f (z)/ lim g(z), provided lim g(z) 6= 0.
z→z0 z→z0 z→z0 z→z0
(5) lim [kf (z)] = k lim f (z), where k is a constant.

z→z0 z→z0
(6) If f (z) = a0 + a1 z + ..... + an z n , then lim f (z) = f (z0 ).

z→z0
Limits involving infinity

It is useful to include the point at infinity (∞) in the complex plane. Such a complex plane is
called extended complex plane. To visualize the point at infinity in the complex plane, consider
the complex plane passing through the equator of a unit sphere centred at the origin. Now, each
point z in the complex plane corresponds to a unique point P on the surface of unit sphere, where
P is the point of intersection of the line passing through the north pole N of the sphere and the
point z, with the surface of the sphere. Obviously, N does not correspond to any finite point in
the complex plane. If we agree N to correspond to the point at infinity, there is a one to one
correspondence between the points on the unit sphere and the points in the extended complex
plane. Such a correspondence is called stereographic projection, and the unit sphere is called
Riemann sphere.
Let > 0 be a real number. Then the points z in the complex plane such that |z| > 1/
correspond to a circular region centred at N on the surface of sphere. This in turn suggests that
|z| > 1/ is a neighbourhood of ∞.
Some results on limits involving ∞ are given below.
(1) lim f (z) = ∞ if and only if lim [1/f (z)] = 0.

z→z0 z→z0
(2) lim f (z) = lim f (1/z)

z→∞ z→0
(3) lim f (z) = ∞ if and only if lim f (1/z) = 0.

z→∞ z→0
z + 2i z−1
Ex. lim = ∞ since lim = 0.
z→1 z − 1 z→1 z + 2i
iz + 2 i + 2/z i+0
Ex. lim = lim = = i.
z→∞ z + i z→∞ 1 + i/z 1+0
Continuity
Let a function f (z) be defined in some neighbourhood of z0 . Then f (z) is said to be continuous at
z0 if given any real number > 0 (howsoever small), there exists a real number δ > 0 (depending
on ) such that
|f (z) − f (z0 )| < for |z − z0 | < δ.
So f (z) is continuous at z0 if and only if lim f (z) = f (z0 ). Further, f (z) is said to be continuous
z→z0
in a region if and only if it is continuous at every point of the region. Some results pertaining to
continuous functions are given below.
(1) If f (z) and g(z) are continuous functions at z0 , then so are the functions f (z)±g(z), f (z)g(z),
(f og)(z) and f (z)/g(z) (g(z0 ) 6= 0).
(2) sin z, cos z, sinh z, cosh z, ez and polynomials all are continuous at every point in the complex
plane.
(3) If f (z) is continuous and non-zero at z0 , then there exists some neighbourhood of z0 such
that f (z) is non-zero at each point of the neighbourhood.
For, f (z) is continuous at z0 . Choosing = |f (z0 )|/2, there exists some δ > 0 such that
|f (z) − f (z0 )| < |f (z0 )|/2 for |z − z0 | < δ.
Now, if there is any point z in the δ-neighbourhood |z − z0 | < δ of z0 such that f (z) = 0,
then we get |f (z0 )| < |f (z0 )|/2, which is not true. So f (z) is non-zero at each point of the
δ-neighbourhood |z − z0 | < δ of z0 .
(4) A function f (z) = u(x, y)+iv(x, y) is continuous at z0 = x0 +iy0 if and only if the component
functions u(x, y) and v(x, y) are continuous at (x0 , y0 ).
(5) If f (z) is continuous in a closed and bounded region R, then f (z) is bounded in R, that is,
there exists some constant M > 0 such that |f (z)| < M for all z ∈ R.
For, let f (z) = u(x, y) + iv(x, y). Then continuity of f (z) in R implies the continuity of
the component functions u(x, y) and v(x, y) in R. Now, from the theory of real functions,
u(x, y) and v(x, y) being continuous in the closed and bounded region R are bounded in R.
So there exists constants M1 > 0p and M2 > 0 such that |u(x, y)| < M1 and |v(x, y)| < M2
for all (x, y) ∈ R. Also, |f (z)| = [u(x, y)]2 + [v(x, y)]2 . It follows that f (z) is bounded in
R.
Differentiability
Let a function w = f (z) be defined in some neighbourhood of z0 . Then w = f (z) is said to be
f (z) − f (z0 )
differentiable at z0 if and only if the limit lim exists. The derivative of w = f (z) at
z→z0 z − z0
z0 is denoted by f 0 (z0 ) or dw

dz z=z0
. So we have
f (z) − f (z0 )
f 0 (z0 ) = lim .
z→z0 z − z0
If we choose z = z0 + δz, then we have
f (z0 + δz) − f (z0 )

f 0 (z0 ) = lim .
δz→0 δz
Some results pertaining to differentiable functions are given below.
d
(1) [f (z) ± g(z)] = f 0 (z) ± g 0 (z).
dz
d
(2) Product rule: [f (z)g(z)] = f 0 (z)g(z) + f (z)g 0 (z).
dz
f 0 (z)g(z) − f (z)g 0 (z)

d f (z)
(3) Quotient rule: = provided g(z) 6= 0.
dz g(z) [g(z)]2
d
(4) Chain rule: [f (g(z))] = f 0 (g(z))g 0 (z).
dz
(5) If f (z) and g(z) are differentiable functions at z0 , then so are the functions f (z) ± g(z),
f (z)g(z), (f og)(z) and f (z)/g(z) (g(z0 ) 6= 0).
(6) sin z, cos z, sinh z, cosh z, ez and polynomials all are differentiable at every point in the
complex plane.
Ex. If f (z) = z 2 + 1, then
f (2 + δz) − f (2) (2 + δz)2 + 1 − (22 + 1)

f 0 (2) = lim = lim = lim (4 + δz) = 4.
δz→0 δz δz→0 δz δz→0
Ex. Show that f (z) = |z|2 is not differentiable at any z except z = 0.

Sol. We have
|z + δz|2 − |z|2
f 0 (z) = lim
δz→0 δz
(z + δz)(z + δz) − zz
= lim
δz→0 δz
(z + δz)(z + δz) − zz
= lim
δz→0 δz
δz
= z + lim δz + z lim
δz→0 δz→0 δz
δx − iδy
= z+ lim (δx − iδy) + z lim
(δx,δy)→(0,0) (δx,δy)→(0,0) δx + iδy
1 − im
= z+z , (δy = mδx)
1 + im
Clearly, f 0 (0) = 0 and f 0 (z) depends on m if z 6= 0. Thus, f (z) = |z|2 is not differentiable at any
z except z = 0.
Ex. If f (z) is differentiable at z0 , then show that f (z) is continuous at z0 . However, the converse
need not be true.
Sol. If f 0 (z0 ) exists, then we have

f (z) − f (z0 ) f (z) − f (z0 )
lim [f (z)−f (z0 )] = lim (z − z0 ) = lim . lim (z−z0 ) = f 0 (z0 ).0 = 0.
z→z0 z→z0 z − z0 z→z0 z − z0 z→z0
This shows that f (z) is continuous at z0 . However, the converse need not be true. In the previous
example, we showed that the function f (z) = |z|2 = x2 + y 2 is not differentiable at any z except
z = 0. However, it is continuous for all z as its real part u(x, y) = x2 + y 2 and imaginary part
v(x, y) = 0, both are continuous everywhere in the complex plane.
Necessary condition for differentiability

Let f (z) = u(x, y) + iv(x, y) be differentiable at z0 = x0 + iy0 . Then first order partial derivatives
of u(x, y) and v(x, y) exist at (x0 , y0 ), and the partial differential equations, known as the Cauchy-
Riemann (CR) equations,
ux = vy , uy = −vx
are satisfied at (x0 , y0 ).

Proof. We have
f (z0 + δz) − f (z0 )
f 0 (z0 ) = lim
δz→0 δz
u(x0 + δx, y0 + δy) + iv(x0 + δx, y0 + δy) − u(x0 , y0 ) − iv(x0 , y0 )
= lim
(δx,δy)→(0,0) δx + iδy
u(x0 + δx, y0 + δy) − u(x0 , y0 ) v(x0 + δx, y0 + δy) − v(x0 , y0 )
= lim +i lim
(δx,δy)→(0,0) δx + iδy (δx,δy)→(0,0) δx + iδy
First solving f 0 (z0 ) along the path δy = 0, we get

u(x0 + δx, y0 ) − u(x0 , y0 ) v(x0 + δx, y0 ) − v(x0 , y0 )
f 0 (z0 ) = lim +i lim = ux (x0 , y0 )+ivx (x0 , y0 ).
δx→0 δx δx→0 δx
Next solving f 0 (z0 ) along the path δx = 0, we get
u(x0 , y0 + δx) − u(x0 , y0 ) v(x0 , y0 + δy) − v(x0 , y0 )
f 0 (z0 ) = −i lim +i lim = −iuy (x0 , y0 )+vy (x0 , y0 ).
δy→0 δx δy→0 δy
Since f 0 (z0 ) exists uniquely, so its values along the paths δx = 0 and δy = 0 must be equal. It
follows that the CR equations ux = vy and uy = −vx are satisfied at (x0 , y0 ).
Ex. The function f (z) = z 2 is differentiable throughout the complex plane as its derivative
f 0 (z) = 2z exists uniquely for all z. Now, let us verify the CR equations. We have f (z) = z 2 =
x2 − y 2 + i(2xy). Therefore, u(x, y) = x2 − y 2 and v(x, y) = 2xy. We find that ux = 2x = vy
and uy = −2y = −vx . So CR equations are satisfied for all (x, y) in the complex plane, as expected.
Ex. We know that the function f (z) = |z|2 is differentiable only at the origin in the complex
plane. Now, f (z) = |z|2 = x2 + y 2 . Therefore, u(x, y) = x2 + y 2 and v(x, y) = 0. So ux = 2x
uy = 2y, vx = 0 and vy = 0. We see that the CR equations are satisfied only at (0, 0), as expected.
CR equations in polar form: If z = reiθ , then f (reiθ ) = u(r, θ) + iv(r, θ) and CR equations in
polar form read as rur = vθ and uθ = −rvr .
Proof: We have
f (reiθ ) = u(r, θ) + iv(r, θ). (1)
Differentiating (1) partially with respect to r and θ, we get
f 0 (reiθ )eiθ = ur + ivr , (2)
and
f 0 (reiθ )reiθ i = uθ + ivθ , (3)
respectively. From (2) and (3), we obtain
ir(ur + ivr ) = uθ + ivθ
or irur − rvr = uθ + ivθ

Equating real and imaginary parts, we get rur = vθ and uθ = −rvr , the CR equations in polar
form.
Note. If CR equations are satisfied at a point, then the function need not be differentiable at that
point as illustrated in the following example.
x3 (1 + i) − y 3 (1 − i)
Ex. If f (z) = for z = (x, y) 6= 0 and f (0) = 0, then show that CR equations
x2 + y 2
are satisfied at z = 0 but f (z) is not differentiable at z = 0.
Sol. Let f (z) = u(x, y) + iv(x, y). Then we have
x3 − y 3 x3 + y 3
u(x, y) = and v(x, y) =
x2 + y 2 x2 + y 2
Also, f (0) = 0 implies u(0, 0) + iv(0, 0) = 0. So u(0, 0) = 0 and v(0, 0) = 0.
u(δx, 0) − u(0, 0) δx
∴ ux (0, 0) = lim = lim = 1.
u(0, δy) − u(0, 0) −δy

uy (0, 0) = lim = lim = −1.
δy→0 δy δy→0 δy
v(δx, 0) − v(0, 0) δx
vx (0, 0) = lim = lim = 1.
v(0, δy) − u(0, 0) δy

vy (0, 0) = lim = lim = 1.
δy→0 δy δy→0 δy
Thus, ux (0, 0) = 1 = vy (0, 0) and uy (0, 0) = −1 = −vx (0, 0). So the CR equations are satisfied at
z = 0.
Next, we have
f (z) − f (0)
f 0 (0) = lim
z→0 z−0
x3 (1 + i) − y 3 (1 − i)
= lim
(x,y)→(0,0) (x2 + y 2 )(x + iy)
x3 (1 + i) − m3 x3 (1 − i)
= lim (along the path y = mx)
x→0 (x2 + m2 x2 )(x + imx)
(1 + i) − m3 (1 − i)
= .
(1 + m2 )(1 + im)
This shows that the value of f 0 (0) depends on m, and hence f 0 (0) does not exist.
Sufficient condition for differentiability

Let f (z) = u(x, y) + iv(x, y) be defined in some neighbourhood z0 = x0 + iy0 . If the first order
partial derivatives of u(x, y) and v(x, y) exist in the neighbourhood of (x0 , y0 ), and are continuous
at (x0 , y0 ) such that the CR equations ux = vy , uy = −vx are satisfied at (x0 , y0 ), then f (z) is
differentiable at z0 .
Proof. The continuity of the partial derivatives ux and uy at (x0 , y0 ) implies the differentiability
of u(x, y) at (x0 , y0 ), that is,
u(x0 + δx, y0 + δy) − u(x0 , y0 ) = ux (x0 , y0 )δx + uy (x0 , y0 )δy + 1 δx + 2 δy,
where 1 and 2 tend to 0 as (δx, δy) → (0, 0). Similarly, v(x, y) is differentiable at (x0 , y0 ), that is,
v(x0 + δx, y0 + δy) − v(x0 , y0 ) = vx (x0 , y0 )δx + vy (x0 , y0 )δy + 3 δx + 4 δy,
where 3 and 4 tend to 0 as (δx, δy) → (0, 0).

It follows that
f (z0 + δz) − f (z0 ) = u(x0 + δx, y0 + δy) + iv(x0 + δx, y0 + δy) + i[u(x0 , y0 ) + iv(x0 , y0 )]
= ux (x0 , y0 )δx + uy (x0 , y0 )δy + 1 δx + 2 δy
+i[vx (x0 , y0 )δx + vy (x0 , y0 )δy + 3 δx + 4 δy]
Since CR equations are satisfied at (x0 , y0 ), so we replace uy (x0 , y0 ) by −vx (x0 , y0 ) and vy (x0 , y0 )
by −ux (x0 , y0 ) in the above expression, and then divide both sides by δz = δx + iδy to obtain
f (z0 + δz) − f (z0 ) δx δy

= ux (x0 , y0 ) + ivx (x0 , y0 ) + (1 + i3 ) + (2 + i4 )
δz δz δz
and δy ≤ 1 and δy ≤ 1. So in the limit δz → 0, that is,

Both δxδz δz
are bounded since δx δz δz
(δx, δy) → (0, 0), the above expression leads to
f 0 (z0 ) = ux (x0 , y0 ) + ivx (x0 , y0 ).
This shows that f is differentiable at z0 .

Analyticity
A function f (z) is said to be analytic at a point z0 if it is differentiable at each point in some
neighbourhood of z0 .
Ex. Show that f (z) = z 2 is analytic at z = 0.

Sol. Consider a neighbourhood of 0, say |z| < 1. Then f 0 (z) = 2z, the derivative of f (z) = z 2 ,
exists at all points of |z| < 1. So f (z) = z 2 is analytic at z = 0.
Ex. Show that f (z) = |z|2 is not analytic at any point in the complex plane.
Sol. Let z0 be any point in the complex plane. Since f (z) = |z|2 is nowhere differentiable except
z = 0, so it can not be differentiable at all points of any neighbourhood of z0 . So f (z) = |z|2 is
not analytic at z0 , and hence it is not analytic anywhere in the complex plane.
Remarks:
1. By definition, analyticity at a point implies the differentiability at that point. However, the
converse is not true. For example, f (z) = |z|2 is differentiable at z = 0 but is not analytic
at this point.
2. If a function f (z) is differentiable at each point of an open set S then it is analytic at every
point of S since S being an open set is neighbourhood of its every point, and we say that
f (z) is analytic in the open set S. For example, f (z) = z 2 is analytic in the open set |z| < 1.
In case, S is not open and we say that f (z) is analytic in S, then it should be understood
that f (z) is analytic in some open set containing S.
3. The definition of analyticity suggests that the necessary and sufficient conditions for analyt-
icity are same as of differentiability. So, if a function f (z) = u(x, y) + iv(x, y) is analytic in
a domain D, then the CR equations ux = vy and uy = −vx are satisfied at all points of the
domain D. Conversely, if the CR equations ux = vy and uy = −vx are satisfied at all points
of the domain D, and the partial derivatives ux , uy , vx and vy are continuous in D, then the
function f (z) is analytic in D.
4. If f (z) and g(z) are analytic functions in a domain D, then so are the functions f (z) ± g(z),
f (z)g(z), (f og)(z) and f (z)/g(z) (g(z0 ) 6= 0).
5. The terms ‘regular’ and ‘holomorphic’ are synonyms of the term ‘analytic’ in the literature.
Entire function
If a function is analytic at all points in the complex plane, then it is said to be an entire function.
Ex. sin z, cos z, sinh z, cosh z, ez and polynomials all are entire functions as these are differentiable
at every point in the complex plane.
Singularity
If a function f (z) fails to be analytic at z0 but is analytic at some point in every neighbourhood
of z0 , then z0 is said to be singularity or singular point of f (z).
1
Ex. The function f (z) = is analytic everywhere except z = 0. So z = 0 is the only singular
z
1
point of f (z) = .
z
Ex. The function f (z) = |z|2 has no singularity as it is not analytic anywhere.
1
Ex. The function f (z) = has two singularities z = 1 and z = 2.
(z − 1)(z − 2)
1
Ex. The singularities of the function f (z) = are z = nπ, where n is any integer.
sin z
Theorem. If f 0 (z) = 0 everywhere in a domain D, then f (z) must be constant throughout D.
Proof. Let f (z) = u(x, y) + iv(x, y). Given that f 0 (z) = 0 everywhere in a domain D. It implies
that f (z) is differentiable everywhere in D, and hence f (z) is analytic everywhere in D. Thus, the
CR equations ux = vy , uy = −vx are satisfied at all points in D. Also, f 0 (z) = ux + ivx = vy − iuy .
So f 0 (z) = 0 everywhere in D implies that ux = 0, uy = 0, vx = 0 and vy = 0 everywhere in D.
Now, consider a line segment P Q inside D. Let s denote the arc length from P along P Q, and
n̂ be a unit vector along P Q in the direction of increasing s. Then directional derivative of u(x, y)
along P Q reads as
du
= ∇u.n̂ = (ux î + uy ĵ).n̂ = 0,
ds
for all (x, y) on P Q since ux = 0 and uy = 0 everywhere in D. This shows that u is constant for
all points on the line segment P Q. Let u(x, y) = a for all (x, y) on P Q. If QR is any line segment
inside D, then using the argument as used for the line segment P Q, it follows that u is constant
for all points on the line segment QR. But value of u(x, y) is a at Q. It implies that u(x, y) = a for
all (x, y) on QR. Now, D being a domain is open and connected. So any two points of D can be
joined by a finite number of line segments inside D. It follows that u(x, y) = a for all points in D.
Similarly, v(x, y) is constant, say b, for all points in D. Hence, f (z) = u(x, y) + iv(x, y) = a + ib,
a constant, everywhere in D.
Theorem. If a function f (z) and its conjugate f (z), both are analytic everywhere in a domain
D, then f (z) is constant throughout D.
Proof. Let f (z) = u(x, y) + iv(x, y). Then f (z) = u(x, y) − iv(x, y). Given that f (z) and f (z),
both are analytic everywhere in the domain D. So CR equations for f (z) = u(x, y) + iv(x, y) given
by
ux = vy , uy = −vx ,
and the CR equations for f (z) = u(x, y) − iv(x, y) given by
ux = −vy , uy = vx ,
are satisfied everywhere in D. Adding and subtracting the CR equations in pairs for f (z) and
f (z), we get ux = 0, uy = 0 and vx = 0, vy = 0, respectively, everywhere in D. So by previous
theorem, f (z) is constant throughout D.
Theorem. If a function f (z) is analytic and |f (z)| is constant everywhere in a domain D, then
f (z) is constant throughout D.
Proof. Let |f (z)| = c, a constant for all z ∈ D. If c = 0, then f (z) = 0 for all z ∈ D and we are
through. Assume that c 6= 0. Then we have c2 = |f (z)|2 = f (z)f (z) or f (z) = c2 /f (z). Given that
f (z) is analytic and |f (z)| = c 6= 0 throughout D. It implies that f (z) = c2 /f (z) is also analytic
throughout D. So by previous theorem, f (z) is constant throughout D.
Harmonic function
Let h(x, y) be a real valued function defined in a domain D in the xy-plane. Then h(x, y) is said
to be harmonic in D if there exist continuous partial derivatives of the first and second order of
h(x, y) throughout D, and satisfy the Laplace equation hxx + hyy = 0.
Ex. Consider the function h(x, y) = e−y sin x. Then we have

hx = e−y cos x, hy = −e−y sin x, hxx = −e−y sin x, hyy = e−y sin x.
We see that h(x, y) has continuous partial derivatives of the first and second order throughout the
xy-plane. Also, hxx + hyy = −e−y sin x + e−y sin x = 0. Therefore, h(x, y) = e−y sin x is harmonic
in the entire xy-plane.
Ex. If f (z) = u(x, y) + iv(x, y) is analytic in a domain D, then show that u(x, y) and v(x, y) are
harmonic in D.
Sol. Given that f (z) = u(x, y) + iv(x, y) is analytic in D. So CR equations are satisfied in D,
that is,
ux = vy and uy = −vx .
It will be proved later that if f (z) = u(x, y) + iv(x, y) is analytic, then u(x, y) and v(x, y) possess
continuous partial derivatives of all orders. In particular, uxy , uyx , vxy and vyx are continuous.
Then from the theory of partial differentiation, it follows that uxy = uyx and vxy = vyx .
Now, differentiating the first CR equation partially with respect to x while second with respect
to y, and adding the resulting equations, we get
uxx + uyy = vxy − vyx = vxy − vxy = 0.
This shows that u(x, y) is harmonic in D.
Similarly, differentiating the first CR equation partially with respect to y while second with
respect to x, and adding the resulting equations, we get
vxx + vyy = uxy − uyx = uxy − uxy = 0.
This shows that v(x, y) is harmonic in D.
Harmonic conjugate
If two real valued functions u(x, y) and v(x, y) are harmonic in a domain D and satisfy the CR
equations throughout D, then v(x, y) is known as harmonic conjugate of u(x, y) in D.
It is easy to deduce that a function f (z) = u(x, y) + iv(x, y) is analytic in a domain D if and
only if v(x, y) is harmonic conjugate of u(x, y) in D.
Note that if v(x, y) is harmonic conjugate of u(x, y), then u(x, y) need not be harmonic con-
jugate of v(x, y). For instance, consider the function f (z) = z 2 = x2 − y 2 + i(2xy). Here,
u(x, y) = x2 − y 2 and v(x, y) = 2xy. So ux = 2x = vy and uy = −2y = −vx . Thus, CR equations
are satisfied. Also, uxx + uyy = 2 − 2 = 0 and vxx + vyy = 0 + 0 = 0. Hence, v(x, y) is harmonic
conjugate of u(x, y). Now, let u(x, y) = 2xy and v(x, y) = x2 − y 2 . Then, ux = 2y 6= −2y = vy .
This shows that the CR equations are not satisfied in the opposite case. Thus, u(x, y) is not
harmonic conjugate of v(x, y).
Ex. Find harmonic conjugate of u(x, y) = y 3 − 3x2 y.

Sol. Let v(x, y) be harmonic conjugate of u(x, y) = y 3 − 3x2 y. So by CR equations, we have
vx = −uy = −3y 2 + 3x2 , and vy = ux = −6xy.
Now integrating vx = −3y 2 + 3x2 with respect to x, we get
v(x, y) = −3xy 2 + x3 + φ(y)
Differentiating partially with respect to y, we have
vy = −6xy + φ0 (y).
But from the second CR equation, vy = −6xy. It follows that φ0 (y) = 0. So φ(y) = C, a constant.
Thus, v(x, y) = −3xy 2 + x3 + C is the required harmonic conjugate.
Chapter 3
Exponential function
Complex exponential function, denoted by ez or exp z, is defined as
ez = ex eiy ,
where z = x + iy. Obviously, |ez | = ex and arg(ez ) = y + 2kπ, where k is any integer. Also,
the complex exponential function ez reduces to the real exponential function ex for y = 0. Some
properties of ez different from ex are given below.
• ex is positive for all real values of x. However, ez can attain negative values. For instance,
eiπ = cos π + i sin π = −1.
• Since e2πi = 1, so ez+2πi = ez e2πi = ez . This shows that ez is a periodic function with
imaginary period 2πi. On the other hand, ex is not a periodic function.
Logarithmic function
The logarithm of a non-zero complex number z, denoted by log z is defined as the complex number
w such that ew = z. To determine w explicitly, let w = u + iv and z = reiΘ . Then we have
eu eiv = reiΘ .
It follows that
eu = r and v = Θ + 2kπ, where k is any integer.
Thus, the logarithm of z is given by
w = log z = ln r + i(Θ + 2kπ), k = 0, ±1, ±2, .......
Obviously, log z is a multiple valued function. The value of log z corresponding to k = 0 is called
its principal value, and is denoted by Logz, that is,
Log z = ln r + iΘ.
It is, of course, a well defined single valued function of z provided z 6= 0. It reduces to the usual
logarithm of real numbers when z is a positive real number. Now, the multiple valued function
log z can be rewritten as
log z = Log z + 2kπi, k = 0, ±1, ±2, .......
Next, we have
elog z = eLog z e2kπi = eln r+iΘ .1 = eln r eiΘ = reiΘ = z.
On the other hand
log ez = log(ex eiy ) = ln ex + i(y + 2kπ) = x + iy + 2kπi = z + 2kπi.

Complex exponents
Let c be any complex number. Then the function z c , where z 6= 0, is defined as
z c = ec log z .
Since log z is a multiple valued function, so is z c . For instance,
ii = ei log i = ei[ln 1+i(2kπ+π/2)] = e−(4k+1)π/2 , k = 0, ±1, ±2, .....
Following the definition of z c , we have
cz = ez log c ,
the exponential function with base c(6= 0). It is indeed a multiple valued function. The usual
interpretation of ez occurs when the principal value of logarithm is taken into account. For the
principal value, log e = 1.
Trigonometric functions
By Euler’s formula, we know that eix = cos x + i sin x and e−ix = cos x − i sin x. So the sine and
cosine functions of the real variable x can be written as
eix − e−ix eix + e−ix
sin x = , cos x = .
2i 2
In analogy, the sine and cosine functions of a complex variable z are defined as
eiz − e−iz eiz + e−iz

sin z = , cos z = .
2i 2
Clearly, these functions reduce to sin x and cos x in case z is restricted to the real number x.
This fact serves as a motivation for the above definitions of sin z and cos z. Next, we define
the other elementary trigonometric functions given by tan z = sin z/ cos z, cot z = cos z/ sin z,
sec z = 1/ cos z and csc z = 1/ sin z, allowing the values of z for which the denominator of the
function under consideration is non-zero.
Let z, z1 and z2 be any complex numbers. Then the following identities can easily be established:
(1) sin(−z) = − sin z, cos(−z) = cos z
(2) sin(z1 + z2 ) + sin(z1 − z2 ) = 2 sin z1 cos z2
(3) sin(z1 + z2 ) = sin z1 cos z2 + cos z1 sin z2
(4) cos(z1 + z2 ) = cos z1 cos z2 − sin z1 sin z2
(5) sin2 z + cos2 z = 1, 1 + tan2 z = sec2 z, 1 + cot2 z = csc2 z
(6) sin 2z = 2 sin z cos z, cos 2z = cos2 z − sin2 z
(7) sin(z + π/2) = cos z, cos(z + π/2) = − sin z
(8) sin(iy) = i sinh y, cos(iy) = cosh y, where sinh y = (ey − e−y )/2 and cosh y = (ey + e−y )/2
(9) sin z = sin(x + iy) = sin x cos(iy) + cos x sin(iy) = sin x cosh y + i cos x sinh y
(10) cos z = cos(x + iy) = cos x cos(iy) + sin x sin(iy) = cos x cosh y + i sin x sinh y
(11) sin(z + 2π) = sin z, cos(z + 2π) = cos z
(12) | sin z|2 = sin2 x + sinh2 y

For, | sin z|2 = sin2 x cosh2 y+cos2 x sinh2 y = sin2 x cosh2 y+(1−sin2 x) sinh2 y = sin2 x(cosh2 y−
sinh2 y) + sinh2 y = sin2 x + sinh2 y
(13) | cos z|2 = cos2 x + sinh2 y
Note that the identities in (11) indicate that sin z and cos z are periodic functions with period
2π each. The identities (12) and (13) tell us that sin z and cos z are unbounded functions in the
complex plane since sinh2 y → ∞ as y → ∞ and therefore | sin z| → ∞ as well as | cos z| → ∞
in the limit y → ∞. In contrast, sin x and cos x are bounded functions in their domain of real
numbers since −1 ≤ sin x ≤ 1 and −1 ≤ cos x ≤ 1 for all x ∈ R.
Inverse Trigonometric functions

If z = sin w, then w defines inverse sine of z and is denoted by sin−1 z. So w = sin−z is the inverse
sine function. Now, the relation
eiw − e−iw
z = sin w = .
2i
yields the following quadratic equation in eiw .
(eiw )2 − 2iz(eiw ) − 1 = 0.
Solving for eiw , we get
eiw = iz + (1 − z 2 )1/2 or w = −i log[iz + (1 − z 2 )1/2 ].
Thus, we have
sin−1 z = −i log[iz + (1 − z 2 )1/2 ].
Note that (1 + z 2 )1/2 is a double valued function. Also, the logarithmic function is multiple valued
function. Consequently, sin−1 z is a multiple valued function.
In analogy to sin−1 z, we define inverse cosine and tangent functions denoted by cos−1 z and
tan−1 z from the relations w = cos z and w = tan z, respectively. Further, we can prove that the
multiple valued functions cos−1 z and tan−1 z are given by
cos−1 z = −i log[z + i(1 − z 2 )1/2 ].

i i+z
tan−1 z = log .
2 i−z
Hyperbolic functions
The hyperbolic sine and cosine functions, denoted by sinh z and cosh z respectively, are defined as
ez − e−z ez + e−z
sinh z = , cosh z = .
2 2
These definitions are motivated by the fact that these functions reduce to their counterparts
sinh x = (ex − e−x )/2 and cosh x = (ex + e−x )/2 in real system when z is restricted to the real
variable x. The other elementary hyperbolic functions are defined as tanh z = sinh z/ cosh z,
coth z = cosh z/ sinh z, sech z = 1/ cosh z and csch z = 1/ sinh z, allowing the values of z for which
the denominator of the function under consideration is non-zero. Some useful identities related to
the hyperbolic functions are given below.
(1) sinh(−z) = − sinh z, cosh(−z) = cosh z
(2) cosh2 z − sinh2 z = 1
(3) −i sinh(iz) = sin z, cosh(iz) = cos z, −i sin(iz) = sinh z, cos(iz) = cosh z
(4) sinh(z1 + z2 ) = sinh z1 cosh z2 + cosh z1 sinh z2
(5) cosh(z1 + z2 ) = cosh z1 cosh z2 + sinh z1 sinh z2
(6) sinh z = sinh(x + iy) = sinh x cosh(iy) + cosh x sinh(iy) = sinh x cos y + i cosh x sin y
(7) cosh z = cosh(x + iy) = cosh x cosh(iy) + sinh x sinh(iy) = cosh x cos y + i sinh x sin y
(8) | sinh z|2 = sinh2 x + sin2 y
(9) | cosh z|2 = sinh2 x + cos2 y
Inverse hyperbolic functions

If z = sinh w, then w defines inverse hyperbolic sine of z and is denoted by sinh−1 z. So w = sinh−1 z
is the inverse hyperbolic sine function. Now, the relation
ew − e−w
z = sinh w = .
2
yields
w = log[z + (z 2 + 1)1/2 ].
Thus, we have
sinh−1 z = log[z + (z 2 + 1)1/2 ].
Note that (1 + z 2 )1/2 is a double valued function. Also, the logarithmic function is multiple valued
function. Consequently, sinh−1 z is a multiple valued function.
In analogy to sinh−1 z, we define inverse hyperbolic cosine and tangent functions denoted by
cosh−1 z and tanh−1 z from the relations w = cosh z and w = tanh z, respectively. Further, we can
prove that the multiple valued functions cosh−1 z and tanh−1 z are given by
cosh−1 z = log[z + (z 2 − 1)1/2 ].
1 1+z
tan−1 z = log .
2 1−z
Chapter 4
Definition . (Differentiation of a complex-valued function of real variable) If w(t) =
u(t) + iv(t) is a complex-valued function of a real variable t, then its derivative with respect to t
is defined as
w0 (t) = u0 (t) + iv 0 (t),
where prime denotes the derivative with respect to t.

Obviously, the existence of the derivatives u0 (t) and v 0 (t) is essential for the existence of w0 (t).
Remark . Note that complex-valued functions of real variable do not respect the mean value
theorem of differential calculus4 . For, consider the function w(t) = eit , t ∈ [0, 2π]. Then w(t) is
continuous in [0, 2π], and is differentiable in (0, 2π) since w0 (t) = ieit exists for all t ∈ (0, 2π). Also
w(0) = 1 = w(2π). But |w0 (t)| = 1 for all t ∈ (0, 2π). Thus, w0 (t) does not vanish at any point in
(0, 2π).
Definition . (Integration of a complex-valued function of real variable) Let w(t) =

u(t) + iv(t) be a piecewise continuous function in [a, b]. Then the integral of w(t) = u(t) + iv(t)
over [a, b] is defined as
Z b Z b Z b
w(t)dt = u(t)dt + i v(t)dt.
a a a
Remark . Mean value theorem of integral calculus5 is also not respected by the complex-valued
functions of real variable. For, consider the function w(t) = eit , t ∈ [0, 2π]. Then
2π b 2π
eit
Z Z
it
w(t)dt = e dt = = 0.
0 a i 0
But |w(c)(2π − 0)| = 2π|eic | = 2π for all c ∈ (0, 2π). So

Z 2π
w(t)dt = 0 6= w(c)(2π − 0)
0
for any c ∈ (0, 2π).
Contours
Definition . (Simple Arc) The function z(t) = x(t) + iy(t), a ≤ t ≤ b, where x(t) and y(t) are
continuous on [a, b], defines an arc in the complex plane from z(a) to z(b). The curve is traversed
with increasing values of t in [a, b], that is, from z(a) to z(b). If the curve does not cross itself,
that is, z(t1 ) 6= z(t2 ) when t1 6= t2 , it is said to be a simple arc or Jordan arc.
4
If a function f is continuous in [a, b], differentiable in (a, b) and f (a) = f (b), then f 0 (x) vanishes at least once
in (a, b).
Z b
5
If a function f is continuous in [a, b], then f (x)dx = c(b − a) for some c ∈ (a, b).
a
Definition . (Simple Closed Curve) If the arc z(t) = x(t) + iy(t), a ≤ t ≤ b does not cross
itself except for z(a) = z(b), it is said to be simple closed curve or Jordan curve. If the simple
closed curve traverses from z(a) to z(b) in counterclockwise direction, then it is said to be positively
oriented.
Definition . (Smooth Arc) If z 0 (t) = x0 (t) + iy 0 (t) is continuous on [a, b] and non-zero on (a, b),
then the arc or curve z(t) = x(t) + iy(t), a ≤ t ≤ b is said to be a smooth arc.
Y
Y Y
4 1.0 1.0
0.5 0.5
3
-1.5 -1.0 - 0.5 0.5 1.0

X -1.5 -1.0 - 0.5 0.5 1.0
X
2
- 0.5 - 0.5
1
-1.0 -1.0
0
0 1 2 3 4
X -1.5 -1.5
Figure 5: Left: z(t) = t + it, 0 ≤ t ≤ 4; Middle: z(t) = eit = cos t + i sin t, 0 ≤ t ≤ 2π; Right:
z(t) = e−it = cos t − i sin t, 0 ≤ t ≤ 2π.
Example . z(t) = t + it, 0 ≤ t ≤ 4 is a smooth arc. It is the line segment y = x, 0 ≤ x ≤ 4 (See

left plot in Figure 5).
Example . z(t) = eit = cos t + i sin t, 0 ≤ t ≤ 2π is a positively oriented simple closed smooth
curve. It is the unit circle x2 + y 2 = 1 traversed in counterclockwise direction (See middle plot in
Figure 5).
Example . z(t) = e−it = cos t − i sin t, 0 ≤ t ≤ 2π is a negatively oriented simple closed smooth
curve. It is the unit circle x2 + y 2 = 1 traversed in clockwise direction (See right plot in Figure 5).
Definition . (Contour) A contour is a piecewise smooth arc, that is, an arc consisting of finite
number of smooth arcs joined end to end. A simple closed contour is a contour that does not
intersect itself except for the end points.
Example . z(t) = t + it, 0 ≤ t ≤ 4 is a simple contour but not closed.
Example . z(t) = eit = cos t + i sin t, 0 ≤ t ≤ 2π is a simple closed contour.
Example . Any triangle is a simple closed contour consisting of three smooth arcs (the three
sides) joined end to end. Similarly, any rectangle is a simple closed contour.
Contour Integrals
Definition . (Contour Integral) Let f be a piecewise continuous function on a contour Z C :
z(t) = x(t) + iy(t), a ≤ t ≤ b. Then the integral of f along the contour C, denoted by f (z)dz
C
is defined as
Z Z b
f (z)dz = f [z(t)]z 0 (t)dt.
C a
Note that the piecewise continuity of f on C and the piecewise continuity of z 0 (t) (as C : z(t) =
x(t) + iy(t) is a contour) in [a, b] imply the piecewise continuity of the integrand f [z(t)]z 0 (t) in
[a, b], and hence ensure the existence of the contour integral.
Next, we have
Z Z b Z a Z
0 0
f (z)dz = f [z(t)]z (t)dt = − f [z(t)]z (t)dt = − f (z)dz.
C a b −C
where −C is the notation for the contour C : z(t) = x(t) + iy(t), a ≤ t ≤ b, when it is traversed
from z(b) to z(a).
If the contour C : z(t) = x(t) + iy(t), a ≤ t ≤ b consists of two smooth arcs C1 : z1 (t) =
x1 (t) + iy1 (t) (a ≤ t ≤ c) and C2 : z2 (t) = x2 (t) + iy2 (t) (c ≤ t ≤ b), then we have
Z Z b Z c Z b Z Z
0 0 0
f (z)dz = f [z(t)]z (t)dt = f [z1 (t)]z1 (t)dt+ f [z2 (t)]z2 (t)dt = f (z)dz + f (z)dz.
C a a c C1 C2
In general, if the contour C consists of n piecewise smooth arcs C1 , C2 , ...., Cn joined end to end
in succession, then
Z Z Z Z
f (z)dz = f (z)dz + f (z)dz + ..... + f (z)dz.
C C1 C2 Cn
Z
Example . Evaluate zdz, where C : |z| = 1.
C
Solution . Given contour C can be written as z = eit , 0 ≤ t ≤ 2π.

Z Z 2π Z 2π
−it it
∴ zdz = e e .idt = idt = 2πi.
C 0 0
Z
1

Example . Evaluate f (z)dz, where f (z) = z 1/2 = exp 2
log z , (|z| > 0, 0 < argz < 2π) and
C
C is the semicircular path z = 2eiθ , 0 ≤ θ ≤ π (Figure 6).
H Θ = 0L X
-2 -1 1 2
Figure 6: z = 2eiθ , 0 ≤ θ ≤ π.
Solution . We notice that the point z = 2 of the semicircular path C : z(θ) = 2eiθ , 0 ≤ θ ≤ π,
lies on the ray θ = 0, where f (z) is not defined. So first we test the piecewise continuity of the
integrand f [z(θ)]z 0 (θ) in the interval [0, π]. We have
√

1
f [z(θ)] = exp (ln 2 + iθ) = 2eiθ/2 ,
2
√
√ √ √
f [z(θ)]z 0 (θ) =
2eiθ/2 2ieiθ = 2 2iei3θ/2 = −2 2 sin(3θ/2) + i2 2 cos(3θ/2), (0 < θ ≤ π).
√
So the right hand limit of f [z(θ)]z 0 (θ) as θ → 0 is i2 2, which is finite. This shows that f [z(θ)]z 0 (θ)
is piecewise continuous on [0, π]. Hence the given integral exists, and is given by
i3θ/2 π
√
Z π √ Z π √
Z
2e 4 2
f (z)dz = f [z(θ)]z 0 (θ)dθ = 2 2i ei3θ/2 dθ = 2 2i =− (1 + i).
C 0 0 3i 0 3
Boundedness of Contour Integrals

Z b

Theorem . If w(t) is a complex-valued piecewise continuous function in [a, b], then w(t)dt ≤
Z b a
|w(t)|dt.
a
Z b

Proof. Let r0 = w(t)dt. If r0 = 0, then the required inequality holds clearly and we have
a
nothing Z
to prove. So let us assume r0 6= 0. Then we have
b Z b
r0 eiθ0 = w(t)dt or r0 = e−iθ0 w(t)dt.
a a
Since r0 isZ real, so right hand

Z b side must also be real.
b
∴ r0 = Re e−iθ0 w(t)dt = Re[e−iθ0 w(t)]dt.
a a
Now, Re[e−iθ0 w(t)] ≤ |e−iθ0 w(t)| = |e−iθ0 ||w(t)| = |w(t)|.

Z b Z b Z b

∴ r0 ≤ |w(t)|dt or w(t)dt ≤ |w(t)|dt.
a a a
Theorem
Z . If a function f is piecewise continuous on a contour C : z(t) = x(t) + iy(t), a ≤ t ≤ b,
then f (z)dz is bounded.
C
Proof. Given that f is piecewise continuous on C : z(t) = x(t) + iy(t), a ≤ t ≤ b. It implies that
f [z(t)] is piecewise continuous on [a, b]. Therefore, f [z(t)] must be bounded on [a, b], that is, there
exists a constant M such that |f [z(t)]| ≤ M for all t ∈ [a, b]. The length of the contour C reads as
Z b Z bp
0
|z (t)|dt = x0 (t)2 + y 0 (t)2 dt = L (say).
a a
Then, in view of the previous theorem, we have

Z Z b Z b Z b Z b
0 0 0
|z 0 (t)|dt = M L.

f (z)dz =
f [z(t)]z (t)dt ≤ |f [z(t)]z (t)|dt = |f [z(t)]||z (t)|dt ≤ M
C a a a a
Z
Thus, f (z)dz is bounded.
C
Use of antiderivatives in contour integrals

Definition . (Antiderivative) Let f be continuous in a domain D (open connected set). Then
antiderivative of f is a function F such that F 0 (z) = f (z) for all z ∈ D.
Theorem . If f is a continuous function in a domain D, then the following statements are equiv-
alent:
(i) Antiderivative of f exists throughout D.
Z
(ii) If C is any contour inside D from a point z1 to z2 , then f (z)dz = F (z2 ) − F (z1 ), where
C
F 0 (z) = f (z).
(iii) Integral of f along any closed contour lying completely inside D is 0.
Remark . It should be noted that the three statements in the above theorem imply each other,
and the independent existence/validity of the statements may not be true. Further, the concept of
antiderivative and path independence is very useful in evaluating contour integrals as illustrated
in the examples given below.
Remark . If antiderivative of a function f exists Z zthroughout D, and C is any contour inside D
2
from a point z1 to z2 , then we use the notation f (z)dz for the contour integral of f along C.
Z z1
In other words, when the contour integral f (z)dz is independent of the path joining the points
C
z1 and z2 , we write
Z Z z2
f (z)dz = f (z)dz.
C z1
R
Example . Evaluate C
zdz, where C is any contour extending from 0 and 1 − i.
Solution . The integrand function z is continuous everywhere in the complex plane. ZAlso, its
antiderivative is z 2 /2, which exists everywhere in the complex plane. It implies that zdz is
C
independent of the contour C joining the points 0 and 1 − i. So we have
Z 1−i 1−i
z2
Z
1
zdz = zdz = = (1 − i)2 = i.
C 0 2 0 2
Z
1
Example . Evaluate 3
dz, where C is any contour extending from 2 and i but not passing
C z
through origin.
Solution . The integrand function 1/z 3 is continuous everywhere in the complex plane except at
z = 0. Also, its antiderivative is −3/z 2 , which exists everywhere in the complex plane except at
zZ = 0. The given contour C extending from 2 to i does not pass through z = 0. It implies that
zdz is independent of the contour C. So we have
C
Z Z i i
1 3 3 15
zdz = dz = − =3+ = .
C 2 z3 z2 2 4 4
Z
1
Example . Use the approach of antiderivatives to evaluate dz, where C : |z| = 2.
C z
Solution . The integrand function f (z) = 1/z is continuous everywhere in the complex plane
except at z = 0. Also, its antiderivative is F (z) = log z = ln r + iθ (α < θ < α + 2π), which exists
everywhere in the complex plane except at z = 0 and the branch cut θ = α. Obviously, for any
value of α, the antiderivative F (z) does not exist at the point of intersection of the branch cut
θ = α and the given contour C : z = 2eiθ , 0 ≤ θ ≤ 2π. To resolve this issue, we evaluate the
integral along C in two parts: (i) along upper half C1 : z = 2eiθ , 0 ≤ θ ≤ π and, (ii) along lower
half C2 : z = 2eiθ , π ≤ θ ≤ 2π. Notice that the branch cut θ = −π/2 of F (z) = log z = ln r + iθ
(−π/2 < θ < 3π/2) does not intersect C1 , which extends from z = 2 (θ = 0 ) to z = −2 (θ = π).
So we have
Z
1
dz = log z]−2
2 = log(−2) − log(2) = (ln 2 + iπ) − ln 2 = πi.
C1 z
Similarly, the branch cut θ = π/2 of F (z) = log z = ln r + iθ (π/2 < θ < 5π/2) does not
intersect C2 , which traverses from z = −2 (θ = π ) to z = 2 (θ = 2π). So we have
Z
1
dz = log z]2−2 = log(2) − log(−2) = (ln 2 + i.2π) − (ln 2 + iπ) = πi.
C2 z
Finally, we have
Z Z Z
1 1 1
dz = dz + dz = πi + πi = 2πi.
C z C1 z C2 z
Cauchy’s Theorem
0
Theorem Z . (Cauchy) If f is analytic and f is continuous within and on a simple closed contour
C, then f (z)dz = 0.
C
Figure 7: A simple closed contour C.

Proof. Let f (z) = u(x, y) + iv(x, y) and C : z(t) = x(t) + iy(t), a ≤ t ≤ b. Then we have
Z Z b
f (z)dz = f [z(t)]z 0 (t)dt (4)
C a
Z b
= [u(x(t), y(t)) + iv(x(t), y(t))][x0 (t) + iy 0 (t)]dt
a
Z b Z b
0 0
= [u(x(t), y(t))x (t) − v(x(t), y(t))y (t)]dt + i [v(x(t), y(t))x0 (t) − u(x(t), y(t))y 0 (t)]dt
Za Z a
= [u(x, y)dx − v(x, y)dy] + i [v(x, y)dx + u(x, y)dy]

C C
It is given that f is analytic. So f (z) = ux + ivx = vy − iux , ux = vy and uy = −vx . Further, f 0 is
0
continuous within and on C. It follows that u, v, ux , uy , vx and vy all are continuous within and
on C. So by Green’s theorem6 , (4) becomes
Z ZZ ZZ ZZ ZZ
f (z)dz = (−vx −uy )dxdy+i (ux −vy )dxdy = (−vx +vx )dxdy+i (vy −vy )dxdy = 0,
C R R R R
where R is the closed region carrying all points within and on C.
Cauchy-Goursat Theorem
Goursat proved that the condition of continuity of f 0 , as mentioned in Cauchy’s theorem, can be
dropped.
Theorem . (Cauchy-Goursat) If f is analytic within and on a simple closed contour C, then

Z
f (z)dz = 0.
C
It is a very usefulZresult regarding the contour integrals. For example, suppose we wish to solve
2
the contour integral ez dz, where C is some simple closed contour. Solving this integral directly
C
without the aid of Cauchy-Goursat theorem would be a tricky business (you can try!). We know
z2
that the function
Z e is analytic everywhere in the complex plane. So by Cauchy-Goursat theorem
2
it follows that ez dz = 0
C
Simply and multiply connected domains

Definition . (Simply and multiply connected domains) A domain D is said to be simply
connected if any closed contour inside D contains points of D only. A domain which is not simply
connected is called multiply connected.
Example . |z| < 1 is a simply connected domain since any simple closed contour inside |z| < 1
would carry points of |z| < 1 only.
6
Z If φ, ψ, φy and
Z Zψx are continuous within and on a simple closed curve C in the XY-plane, then
[φdx + ψdy] = (ψx − φy )dxdy, where R is the closed region carrying all points within and on C.
C R
Example . The ring shaped domain 1 < |z| < 2 is multiply connected. For, the simple closed
contour |z| = 3/2 lies inside 1 < |z| < 2. The region |z| ≤ 1 is a part of interior of |z| = 3/2.
However, |z| ≤ 1 is not lying in 1 < |z| < 2.
Y Y
1 2
0.5 1
X X
-1 -0.5 0 0.5 1 -3 -2 -1 0 1 2
-0.5 -1
-1 -2
-3
Figure 8: Left: |z| < 1 is simply connected. Right: The ring shaped domain 1 < |z| < 2 is
multiply connected
Z
Theorem . If a function f is analytic throughout a simply connected domain D, then f (z)dz =
C
0 for every closed contour lying in D. Furthermore, f must have antiderivative everywhere in D.
Proof. If C is any simple Zclosed contour inside D, then f is analytic within and on C. So by
Cauchy-Goursat theorem, f (z)dz = 0. If C is closed but not simple, then C consists of a finite
C
number of simple closed contours, say C1 , C2 , ..., Cn such that the integral of f along C is equal
to the sum of integrals along the simple closedZ contours C1 , C2 , ..., Cn . But integral of f along
any simple closed contour is 0. It follows that f (z)dz = 0.
Z C
We have proved that f (z)dz = 0 for every closed contour C inside the domain D. So from
C
theorem , it follows that antiderivative of f exists throughout D.
Extension of Cauchy-Goursat theorem

In the following, we discuss some theorems which pertain to the extension of Cauchy-Goursat
theorem to multiply connected regions.
Theorem . (Principle of deformation of path) Suppose a positively oriented simple closed

contour C1 completely lies in the interior of a positively oriented simple closed contour C. If a
function f is analytic withinZand on the closed
Z region constituted by the interior and boundary of
C, and exterior of C1 , then f (z)dz = f (z)dz.
C C1
S
P C
B C1
A
Q
R
Figure 9: The curve ABP QBARSA constitutes a simple closed contour as it is traversed along
the path shown by arrowheads starting from A and ending at A.
Proof. Let AB be a line segment such that A lies on C and B lies on C1 . Also, choose distinct
points P , Q on C1 while R, S on C as shown in Figure 9. Then we see that the path ABP QBARSA
constitutes a simple closed contour within and on which the function f is analytic. So by Cauchy-
Goursat theorem, we have
Z
f (z)dz = 0.
ABP QBARSA
Z Z Z Z
=⇒ f (z)dz + f (z)dz + f (z)dz + f (z)dz = 0.
AB BP QB BA ARSA
Z Z Z Z
=⇒ f (z)dz − f (z)dz − f (z)dz + f (z)dz = 0.
AB BQP B AB ARSA
Z Z
=⇒ − f (z)dz + f (z)dz = 0.
C1 C
Z Z
=⇒ f (z)dz = f (z)dz.
C C1
This result is called principle of deformation of path in the sense that the contour C is
deformed to the contour C1 for evaluating the integral. It proves to be very useful for solving the
integrals along contours of arbitrary shape.
Theorem . Suppose two positively oriented and non-intersecting simple closed contour C1 and
C2 completely lie in the interior of a positively oriented simple closed contour C. If a function
f is analytic within and on
Z the closedZregion constituted
Z by the interior and boundary of C and
exteriors of C1 , C2 , then f (z)dz = f (z)dz + f (z)dz.
C C1 C2
Proof. Let AB, DE and F G be line segments or polygonal paths connecting C and C1 , C1 and
C2 , and C2 and C, respective as shown in the Figure 10. Choose points P , Q on Also, choose
distinct points P , Q on C1 while R, S on C2 and T , U on C as shown in Figure 10. Then we see
that the path ABP DERF GT AU GSEDQBA constitutes a simple closed contour within and on
which the function f is analytic. So by Cauchy-Goursat theorem, we have
Z
f (z)dz = 0.
ABP DERF GT AU GSEDQBA
T
P C
C1 C2 R G
B D F
E
A Q
S
U
Figure 10: The curve ABP DERF GT AU GSEDQBA constitutes a simple closed contour as it is
traversed along the path shown by arrowheads starting from A and ending at A.
Z Z Z Z Z Z
=⇒ f (z)dz + f (z)dz + f (z)dz + f (z)dz + f (z)dz + f (z)dz +
Z AB Z BP D Z DE Z ERF Z FG Z GT A
f (z)dz + f (z)dz + f (z)dz + f (z)dz + f (z)dz + f (z)dz = 0.

AU G GF F SE ED DQB BA
Z Z Z Z Z Z
=⇒ f (z)dz + f (z)dz + f (z)dz + f (z)dz + f (z)dz − f (z)dz −
Z AB Z BP D Z DE Z ERF Z FGZ AT G
f (z)dz − f (z)dz + f (z)dz − f (z)dz + f (z)dz − f (z)dz = 0.

GU A FG F SE DE DQB AB
Z Z Z Z Z Z
=⇒ f (z)dz+ f (z)dz+ f (z)dz+ f (z)dz− f (z)dz− f (z)dz = 0.
BP D DQB ERF F SE AT G GU A
Z Z Z
=⇒ f (z)dz + f (z)dz − f (z)dz = 0.
BP DQB ERF SE AT GU A
Z Z Z
=⇒ f (z)dz + f (z)dz − f (z)dz = 0.
C1 C2 C
Z Z Z
=⇒ f (z)dz = f (z)dz + f (z)dz.
C C1 C2
Remark . Let C be a positively oriented simple closed contour, and C1 , C2 , ...., Cn be positively
oriented non-intersecting simple closed contours lying inside C. If a function f is analytic within
and on the closed region constituted by the interior and boundary of C, and exteriors of C1 ,
C2 ,......., Cn , then following the analogy of theorem , we get
Z Z Z Z
f (z)dz = f (z)dz + f (z)dz + ...... + f (z)dz.
C C1 C2 Cn
This result is also known as Cauchy-Goursat theorem for multiply connected domains.
Z
dz
Example . Solve , given that C is any contour such that a circle of radius less than 2 and
C z
centred at origin lies inside C.
Solution . Consider the unit circle C1 : |z| = 1. Then the function 1/z is analytic in the interior
of C but exterior of C1 . So by principle of deformation of path, we have
Z Z Z 2π it
dz dz ie dt
= = = 2πi.
C z C1 z 0 eit
Cauchy Integral Formula and its extension

Theorem . (Cauchy Integral Formula) If f is analytic within and on a positively oriented
simple closed contour C and z0 is a point in the interior of C, then
Z
1 f (z)dz
f (z0 ) = . (5)
2πi C z − z0
C
r
z0 C1
Figure 11:
f (z)
Proof. Given that f is analytic within and on the simple closed contour C. So the function z−z 0
has the only singularity z = z0 that lies inside C. Let C1 : |z − z0 | = r be a circle with radius r so
f (z)
small that it completely lies in the interior of C. Then the function z−z 0
is analytic in the interior
of C but exterior of C1 . So by principle of deformation of path, we have
Z 2π Z 2π
f (z0 + reit )ireit dt
Z Z
f (z)dz f (z)dz
= = it
=i f (z0 + reit )dt.
C z − z0 C1 z − z0 0 re 0
Here, r is at our discretion. In the limit r → 0, the continuity of f implies that

Z Z 2π
f (z)dz
=i f (z0 )dt = 2πif (z0 ),
C z − z0 0
which leads to the Cauchy Integral Formula.

Note that the Cauchy Integral Formula can be generalized to obtain
Z
f (z)dz
n+1
= 2πif (n) (z0 ) n!.
C (z − z0 )
For, let z be a point inside C. Then by Cauchy Integral Formula, we have
Z
1 f (s)ds
f (z) = . (6)
2πi C s − z
Without any argument (in analogy of differentiation of real functions under integral sign7 ), let us
differentiate both sides of (6) with respect to z to obtain
Z
0 1 f (s)ds
f (z) = . (7)
2πi C (s − z)2
Z b(α) Z b(α)
7 d ∂ db
Leibniz Rule for differentiation under integral sign: f (x, α)dx = f (x, α)dx + f (b, α) −
dα a(α) a(α) ∂α dα
da
f (a, α)
dα
Now, we prove the formula obtained in (7). Let d be the smallest distance of z from points s on
C and 0 < |δz| < d. Then, in view of (6), we have
Z
f (z + δz) − f (z)
Z
1 1 1 f (s) 1 f (s)ds
= − ds = . (8)
δz 2πi C s − z − δz s − z δz 2πi C (s − z − δz)(s − z)
Subtracting right hand side of (7) from both sides of (8), we get
f (z + δz) − f (z)
Z Z
1 f (s)ds 1 δzf (s)ds
− 2
= . (9)
δz 2πi C (s − z) 2πi C (s − z − δz)(s − z)2
Now, let |f (s)| ≤ M on C and L be the length of C. We notice that |s − z| ≥ d and |δ| < d, and
therefore
|s − z − δz| = |(s − z) − δz| ≥ ||s − z| − |δz|| ≥ d − |δz| > 0.
Then, from (9), it follows that

Z
f (z + δz) − f (z)
≤ |δz|M L .
Z
1 f (s)ds 1 δzf (s)ds
− 2
= (10)
δz 2πi C (s − z) 2π C (s − z − δz)(s − z) (d − |δz|)d2
2
In the limit δz → 0, right side of the inequality (10) tends to 0 and hence we get
f (z + δz) − f (z)
Z
1 f (s)ds
lim − = 0.
δz→0 δz 2πi C (s − z)2
Z
0 1 f (s)ds
∴ f (z) = .
2πi C (s − z)2
Differentiating again with respect to z and following the procedure adopted for proving (7), it
can be shown that
Z
00 2! f (s)ds
∴ f (z) = .
2πi C (s − z)3
Repeating the same procedure n − 2 more times, we get the general formula
Z
(n) n! f (s)ds
f (z) = , n = 0, 1, 2, ......
2πi C (s − z)n+1
If we denote z by z0 and replace s by z, the above formula reads as

Z
(n) n! f (z)dz
f (z0 ) = , n = 0, 1, 2, ......
2πi C (z − z0 )n+1
We shall refer this formula to as the generalized Cauchy Integral Formula since it yields the Cauchy
Integral Formula (5) for n = 0. It is very useful for evaluating contour integrals as illustrated in
the following examples.
Z
dz
Example . Evaluate , C : |z| = 1.
C z
Z
f (z)dz
Solution . Comparing the given integral with n+1
, we have f (z) = 1, z0 = 0 and
C (z − z0 )
n = 0. We see that f (z) = 1 is analytic within and on C. Also, z0 = 0 lies inside C : |z| = 1. So
by Cauchy integral formula, we have
Z
dz
= 2πif (0) = 2πi.
C z
Z 3
z + 2z 2 − 1
Example . Evaluate dz, C : |z| = 2.
C (z − 1)3
Z
f (z)dz
Solution . Comparing the given integral with n+1
, we have f (z) = z 3 + 2z 2 − 1, z0 = 1
C (z − z0 )
and n = 2. So f 00 (z) = 6z + 4. We see that f (z) = z 3 + 2z 2 − 1 is analytic within and on C. Also,
z0 = 1 lies inside C : |z| = 2. So by Cauchy integral formula, we have
Z 3
z + 2z 2 − 1
3
dz = 2πif 00 (1) = 2πi(10) = 20πi.
C (z − 1)
Z 3
z + 2z 2 − 1
Example . Evaluate dz, C : |z| = 1/2.
C (z − 1)3
z 3 +2z 2 −1
Solution . The integrand function (z−1)3
has only one singularity z = 1 that lies outside
z 3 +2z 2 −1
C : |z| = 1/2. So is analytic within and on C : |z| = 1/2. So by Cauchy-Goursat theorem,
(z−1)3
the given integral vanishes.
z2 + 2
Z
Example . Evaluate dz, C : |z| = 3.
C (z − 1)(z − 4)
z2 + 2
Solution . The given integrand function has two singularities namely z = 1 and
(z − 1)(z − 4)
(z 2 + 2)/(z − 4)
Z
z = 4. But z = 4 lies outside C : |z| = 3. So rewriting the given integral as dz
Z C z−1
f (z)dz
and comparing with n+1
, we have f (z) = (z 2 + 2)/(z − 4), z0 = 1 and n = 0. We see
C (z − z0 )
that f (z) = (z 2 + 2)/(z − 4) is analytic within and on C. Also, z0 = 1 lies inside C : |z| = 3. So
by Cauchy integral formula, we have
z2 + 2
Z
dz = 2πif (1) = 2πi(−1) = −2πi.
C (z − 1)(z − 4)
Implications of Cauchy Integral Formula

In this section, we discuss some remarkable implications of the Cauchy Integral Formula.
Theorem . (Derivatives of an analytic function) If a function f is analytic at z0 , then

derivatives of all orders of f are also analytic at z0 .
Proof. Given that f is analytic at z0 . Therefore, f is analytic in some neighbourhood of z0 , say,

|z − z0 | < . Then f is analytic within and on the circle C : |z − z0 | = /2. So by Cauchy Integral
Formula, f 00 exists at all points inside C. It implies that f 0 is analytic at z0 . Applying the same
argument on f 0 , it is easy to show that f 00 is analytic at z0 . In general, we reach to the conclusion
that derivatives of all orders of f are analytic at z0 . This completes the proof.
Remark . Note that the above result is not true in case of real functions. For, consider f (x) = x3/2 .
Then the first derivative of f , that is, f 0 (x) = (3/2)x1/2 exists at x = 0, which is f 0 (0) = 0. But
the second derivative f 00 (x) = (3/4)x−1/2 does not exists at x = 0.
Theorem . If a function f (z) = u(x, y) + iv(x, y) is analytic in a domain D, then u(x, y) and
v(x, y) possess continuous partial derivatives of all orders in D.
Proof. Given that f (z) = u(x, y)+iv(x, y) is analytic D. So by theorem , derivatives of all orders of
f exist in D. Also, by theorem (necessary condition for differentiability), f 0 (z) = ux +ivx = vy −iuy ,
f 00 (z) = uxx + ivxx = vxy − iuxy and so forth. Now by theorem(?), differentiability of f in D implies
continuity of u and v in D, differentiability of f 0 in D implies continuity of ux , uy , vx , vy in D and
so forth. This completes the proof. It
Z
Theorem . (Morera’s Theorem) If a function f is continuous in a domain D, and f (z)dz = 0
C
for every closed contour C inside D, then f is analytic throughout the domain D.
Z
Proof. Given that f is continuous in the domain D, and f (z)dz = 0 for every closed contour C
C
inside D. So by the theorem of antiderivatives, there exists a function F such that F 0 (z) = f (z)
for all z ∈ D. It implies that F (z) is analytic in D. We know that derivative of analytic function
is also analytic. Therefore, f being derivative of F is analytic in D. This completes the proof.
Theorem . (Liouville’s Theorem) Every entire and bounded function is constant.
Proof. Let f be an entire and bounded function. Let z0 be any point in the complex plane. Then
to prove the Liouville’s theorem, it is sufficient to show that f 0 (z0 ) = 0. Now consider a circle
C : |z − z0 | = R in the complex plane. Then f being entire function is analytic within and on C.
So by Cauchy integral formula, we have
Z
0 1 f (z)
f (z0 ) = dz
2πi C (z − z0 )2
Now f being a bounded function there exists a constant M such that |f (z)| ≤ M for all z on
C : |z − z0 | = R. So we have
Z Z 2π it it
Z 2π
0
1 f (z) 1 f (z0 + Re )Re idt 1 1 M
|f (z0 )| =

2
dz =

2 i2t
= |f (z0 +Reit )|dt ≤ M.2π =
2πi C (z − z0 ) 2πi 0 R e 2πR 0 2πR R
Since R is at our discretion, in the limit R → ∞, the inequality |f 0 (z0 )| ≤ M

R
implies that f 0 (z0 ) = 0.
This completes the proof.
Ex. The function sin z is unbounded in the complex plane. For, if sin z is bounded, then sin z
being an entire function should be constant by Liouville’s theorem.
Remark . f If a function f is analytic within and on a positively oriented circle C : |z−z0 | = R and
|f (z)| ≤ M for all z on C, then using the generalized Cauchy Integral Formula and the procedure
adopted in last step of the proof of Liouville’s theorem, it is easy to show that |f (n) (z0 )| ≤ n!M/Rn .
Theorem . (Fundamental theorem of algebra) Any polynomial of degree n ≥ 1 has at least

one zero in the complex plane.
Proof. Let Pn (z) be a polynomial of degree n ≥ 1. Then there exists constants a0 , a1 , ......., an−1 ,
an 6= 0 such that Pn (z) = a0 + a1 z + ......... + an−1 z n−1 + an z n . Assume, if possible, that Pn (z) has
no zero in the complex plane. Then the function 1/Pn (z) = f (z) (say) being a rational function
with denominator non-zero for all z, is analytic throughout the complex plane. Now, we shall
prove that f (z) is bounded. We have
a a1 an−1
0
|Pn (z)| = n + n−1 + .......... + + an z n (11)

z za a z a
0 1 n−1
≥ |an | − n − n−1 − .......... − |z|n (12)
z z z
a |an |
We can choose a positive real number R sufficiently large such that zn−i i
< 2n , (i = 0, 1, .., n − 1),
for all z with |z| > R. So we have

|an | |an | n
|Pn (z)| > |an | − Rn = R .
2 2
It leads to |f (z)| = |1/Pn (z)| < |an2|Rn for all z with |z| > R, which is turn implies that f (z) is
bounded in the exterior of the disk |z| = R. Also f (z) being an entire function is bounded within
and on the disk |z| = R. Thus f (z) is bounded function. By Liouville’s theorem, therefore, f (z)
must be constant, which is not true. So our assumption that Pn (z) has no zero in the complex
plane is wrong. This completes the proof.
Remark . By the above theorem, Pn (z) has at least one zero, say zn . Then z − zn is a factor of
Pn (z) and consequently Pn (z) = (z − zn )Qn−1 (z), where Qn−1 (z) is a polynomial of degree n − 1.
Again, by the above theorem, the polynomial Qn−1 (z) has at least one zero, say zn−1 . So z − zn−2
is a factor of Qn−1 (z) and this in turn implies that Pn (z) = (z − zn )(z − zn−1 )Rn−2 (z), where
Rn−2 (z) is a polynomial of degree n − 2. Continuing the same process for n − 2 more times, we
find
Pn (z) = (z − zn )(z − zn−1 ).......(z − z1 )an
This shows that the polynomial Pn (z) has exactly n zeros given by z1 ,...., zn .
Chapter 5
Theorem . (Laurent series) If a function f is analytic throughout an open annular or ring
shaped domain R1 < |z − z0 | < R2 , and C is any positively oriented simple closed contour C inside
the ring shaped domain encircling z0 , then for any z in R1 < |z − z0 | < R2 , f (z) has the series
representation given by
∞
X ∞
X
f (z) = n
an (z − z0 ) + bn (z − z0 )−n ,
n=0 n=1
where
Z
1 f (z)dz
an = , n = 0, 1, 2, ....
2πi C (z − z0 )n+1
Z
1 f (z)dz
bn = , n = 1, 2, ....
2πi C (z − z0 )−n+1
Remark . If f is analytic throughout the domain |z − z0 | < R2 , then f (z)(z − z0 )n−1 (n = 1, 2, ...)
is analytic within and on closed the contour C. So by Cauchy Goursat theorem, it follows that
Z
1 f (z)dz
bn = = 0, n = 1, 2, ....
2πi C (z − z0 )−n+1
and consequently the above Laurent series of f reduces to the Taylor series of f in the domain
|z − z0 | < R2 .
1 1 1
Example . The function = − has two singularities z = 1 and z = 2.
(z − 1)(z − 2) z−2 z−1
So this function is analytic in three distinct domains |z| < 1, 1 < |z| < 2 and |z| > 2. Let us find
find series of f valid (convergent) in the three domains.
Solution . (i) For |z| < 1, we have |z/2| < 1. It follows that
∞ ∞ ∞
1 1 1 1 1 1X X X
− =− + =− n
(z/2) + n
z = (1 − 2−n−1 )z n
z−2 z−1 2 1 − z/2 1 − z 2 n=0 n=0 n=0
It is Taylor series of the given function in the domain |z| < 1. Equivalently, it is Laurent series
with all bn = 0.
(ii) For 1 < |z| < 2, we have |z/2| < 1 and |1/z| < 1. It follows that
∞ ∞ ∞ ∞
1 1 1 1 1 1 1X 1 X −n 1X X
− =− − =− (z/2)n − z =− (z/2)n − z −n
z−2 z−1 2 1 − z/2 z 1 − 1/z 2 n=0 z n=0 2 n=0 n=1
It is Laurent series of the given function in the domain 1 < |z| < 2.
(iii) For |z| > 2, we have |2/z| < 1 and |1/z| < 1. It follows that
∞ ∞ ∞
1 1 1 1 1 1 1X 1 X −n X n
− = − = (2/z)n − z = (2 − 1)z −n
z−2 z−1 z 1 − 2/z z 1 − 1/z z n=0 z n=0 n=1
It is Laurent series of the given function in the domain 1 < |z| < 2.
Chapter 6
Isolated Singularity
We know that if a function fails to be analytic at z0 but is analytic at some point in each neighbour-
hood of z0 , then z0 is a singularity of f . Further, if f is analytic in some deleted neighbourhood
of z0 , then z0 is said to be isolated singularity of f . In other words, z0 is isolated singularity of f
provided there exists some neighbourhood of z0 that contains no singularity of f other than z0 .
1
Ex. The function (z−1)(z−2) has two singularities namely z = 1 and z = 2. Both are isolated since
singularity free deleted neighbourhoods can easily be constructed for both the singularities. For
instance, 0 < |z − 1| < 1/2 and 0 < |z − 2| < 1/2 are deleted neighbourhoods of z = 1 and z = 2
respectively, not containing any other singularity of f .
1
Ex. Consider the function sin(π/z) . We see that z = 0 is a singularity of this function as it is
not defined at z = 0. The other singularities are given by sin(π/z) = 0 or π/z = nπ or z = 1/n
where n is any non-zero integer. Thus, the given function has infinitely many singularities. The
singularities z = 1/n are all isolated. But z = 0 is not isolated as each neighbourhood of 0 carries
infinitely many singularities of the type 1/n.
Types of Isolated Singularities

Let z0 be an isolated singularity of a function f . Then there exists some deleted neighbourhood of
z0 , say 0 < |z − z0 | < R, in which f is analytic. So f is analytic in the ring shaped annular region
0 < |z −z0 | < R (We may treat z = 0 as the point circle). So Laurent series of f in 0 < |z −z0 | < R
reads as
∞
X ∞
X
f (z) = n
an (z − z0 ) + bn (z − z0 )−n ,
n=0 n=1
where
Z
1 f (z)dz
an = , n = 0, 1, 2, ....
2πi C (z − z0 )n+1
Z
1 f (z)dz
bn = , n = 1, 2, ....
2πi C (z − z0 )−n+1
and C is any positively oriented simple closed contour inside 0 < |z − z0 | < R around z0 .
∞
X
The first part of Laurent series namely an (z − z0 )n is called regular part while the second
n=0
∞
X
part bn (z − z0 )−n is called principal part. Depending on the number terms in the principal part
n=1
of the Laurent series, we classify the isolated singularity in the following three ways.
Removable singularity
If the principal part of Laurent series contains no term, that is, bn = 0 for all n, then z0 is called
removable singularity of f .
Ex. Consider the function sinz z . Obviously, 0 is an isolated singularity of sin z

z
. Next, we have
z3 z2

sin z 1
= z− + ....... = 1 − + .......
z z 3! 3!
the Laurent series of sinz z valid in the domain 0 < |z| < ∞. Note that we cancelled z in the series
expansion since z 6= 0 in the domain 0 < |z| < ∞. We see that principal part of the Laurent
series contains no term, and therefore 0 is a removable singularity of sinz z . If we define the value of
sin z
z
equal to 1 at z = 0, then 0 is no more a singularity of sinz z . In other words the singularity is
removed by redefining the function. That is why the name removable singularity is there.
Pole
If the principal part of Laurent series contains finite number of terms, that is, there exists some
positive integer m such that bm = 0 and bn = 0 for all n > m, then z0 is called pole order m of f .
A pole of order 1 is called simple pole.
1
Ex. Consider the function f (z) = (z−1) 2 . We see that z = 1 is the only singularity of f . Also,
−2
f (z) = (z − 1) is already in the Laurent series form about z = 1 with b1 = 0, b2 = 1 and bn = 0
for all n > 2. So z = 1 is a pole of order 2.
Note: Let f (z) = g(z)/(z − z0 )m , where g(z) is non-zero and analytic at z0 . Then z0 is a pole of
order m of f (z). For, g(z) being analytic at z0 , its Taylor series about z0 reads as
∞
X
g(z) = an (z − z0 )n = a0 + a1 (z − z0 ) + ...... + am (z − z0 )m + am+1 (z − z0 )m+1 + ......
n=0
∴ f (z) = a0 (z − z0 )−m + a1 (z − z0 )−m+1 + ...... + am + am+1 (z − z0 ) + ......

Now g(z0 ) = a0 6= 0 since g(z) is non-zero at z0 . So from Laurent series of f (z), we observe that
z = z0 is a pole of order m.
1
Ex. The function f (z) = (z−1)(z−3) 2 has two isolated singularities namely 1 and 3. To see the
behavior of the singularity 1, we rewrite f (z) as

1/(z − 3)2
f (z) = ,
z−1
and notice that 1/(z − 3)2 is analytic and non-zero at z = 1. So z = 1 is a simple pole.
Again, rewriting f (z) as

1/(z − 1)
f (z) = ,
(z − 3)2
we see that 1/(z − 1) is analytic and non-zero at z = 3. So z = 3 is a pole of order 2.
Essential singularity
If the principal part of Laurent series contains infinitely number of terms, then z0 is called an
essential singularity of f .
2 2
Ex. The function e1/z has an isolated singularity at z = 0. The Laurent series of e1/z about
z = 0 is given by
2 1 1 1
e1/z = 1 + 2
+ + ...........
z 2! z 4
We see that there are infinite number of terms with negative powers of z. Hence, 0 is an essential
2
singularity of e1/z .
Residue at Isolated Singularity

In Laurent series of f (z) about the isolated singularity z = z0 , the coefficient of (z − z0 )−1 is given
by
Z
1
b1 = f (z)dz.
2πi C
It is called residue of f at z0 and we shall denote it by Res[f (z), z0 ]. Therefore, we have

Z
1
Res[f (z), z0 ] = b1 = f (z)dz.
2πi C
It can be rewritten as
Z
f (z)dz = 2πi Res[f (z), z0 ] = 2πib1 .
C
This shows
Z that the value of b1 or Res[f (z), z0 ] can readily provide us the value of the contour
integral f (z)dz.
C
Z
2
Ex. Evaluate the contour integral e1/z dz, C : |z| = 1.
C
1/z 2
Sol. The function e has only one singularity namely z = 0 inside C. The Laurent series of
2
e1/z is given by
2 1 1 1
e1/z = 1 + + + ...........
z 2 2! z 4
We see that the coefficient of z −1 is 0, that is, b1 = 0. Hence, the value of the given integral is 0.
Remark. The above example shows that if integral of a function along a simple closed contour
is 0, the function need not be analytic. On the other hand, if a function is analytic within and
on a simple closed contour C, then the integral of the function along C is 0, by Cauchy-Goursat
theorem.
Residue at pole
If z0 is a pole of order m, then Laurent series of f (z) reads as
∞
X
f (z) = an (z − z0 )n + b1 (z − z0 )−1 + b2 (z − z0 )−2 + ..... + bm (z − z0 )−m .
n=0
Multiplying both sides by (z − z0 )m , we have

∞
X
(z − z0 )m f (z) = an (z − z0 )n+m + b1 (z − z0 )m−1 + b2 (z − z0 )m−2 + ..... + bm .
n=0
Differentiating both sides m − 1 times with respect z, and evaluating the resulting two sides at
z = z0 , we get
φm−1 (z) = b1 (m − 1)!,
where φ(z) = (z − z0 )m f (z).

φm−1 (z)
∴ Res[f (z), z0 ] = b1 =
(m − 1)!
1
Ex. Find the residues at the poles of f (z) = (z−1)(z−3)2.
Sol. The given function has two poles namely 1 and 3 of orders 1 and 2, respectively.
Residue at the simple pole z = 1
In this case, we have φ(z) = (z − 1)f (z) = 1/(z − 3)2 . So φ(1) = 1/4 and
φ(1)
Res[f (z), 1] = = 1/4.
0!
Residue at z = 3
In this case, we have φ(z) = (z − 3)2 f (z) = 1/(z − 1). So φ0 (z) = −1/(z − 1)2 , φ0 (3) = −1/4 and
φ0 (3)
Res[f (z), 3] = = −1/4.
1!
Cauchy Residue Theorem

If a function is analytic within and on a positively oriented simple closed contour C except some
finite number of singular points, say z1 , z2 , ....., zn inside C, then
Z Xn
f (z)dz = 2πi Res[f (z), zk ].
C k=1
Proof. The singularities z1 , z2 , ....., zn being finite in number are isolated. Let us enclose these
singularities inside non-overlapping n circles Ck : |z − zk |, (k = 1, 2, ...., n) completely lying inside
C. Then f is analytic inside the contour C and outside the circles Ck . So by Cauchy-Goursat
theorem, we have
Z Xn Z Xn
f (z)dz = f (z)dz = 2πi Res[f (z), zk ]
C k=1 Ck k=1
since
Z
f (z)dz = 2πi Res[f (z), zk ], (k = 1, 2, ...., n).
Ck
Remark. The Cauchy Residue Theorem is very useful for calculating contour integral of a func-
tion having finite number of singularities inside the contour.
Z
1
Ex. Use Cauchy residue theorem to evaluate 2
dz, C : |z| = 4.
C (z − 1)(z − 3)
1
Sol. We see that the integrand f (z) = (z−1)(z−3) 2 has two poles namely 1 and 3, both lying in the
interior of the circle C : |z| = 4. So by Cauchy residue theorem

Z
1
2
dz = 2πi[Res[f (z), 1] + Res[f (z), 3]] = 2πi(1/4 − 1/4) = 0
C (z − 1)(z − 3)
Z
1
Ex. Use Cauchy residue theorem to evaluate 2
dz, C : |z| = 2.
C (z − 1)(z − 3)
Sol. In this case, the pole 1 lies inside C : |z| = 2 while the pole 3 lies outside it. So by Cauchy
residue theorem
Z
1
2
dz = 2πiRes[f (z), 1] = 2πi(1/4) = πi/2.
C (z − 1)(z − 3)
Caution. Observe carefully the positions of singularities with respect to the given contour C. In
the formula of Cauchy residue theorem, use residues at only those singularities which lie inside C.
5z − 2
Z
Ex. Use Cauchy residue theorem to evaluate dz, C : |z| = 2.
C z(z − 1)
5z−2
Sol. We see that the integrand f (z) = z(z−1) has two simple poles, namely 0 and 1 both lying
inside the circle C : |z − 1| = 1/2. So by Cauchy residue theorem, we have
Z
f (z)dz = 2πi [Res[f (z), 0] + Res[f (z), 1]]
C
Residue at z = 0
5z−2
In this case, we have φ(z) = zf (z) = z−1
. So φ(0) = 2 and
φ(0)
Res[f (z), 1] = = 2.
0!
Residue at z = 3
5z−2
In this case, we have φ(z) = (z − 1)f (z) = z
. So φ(1) = 3 and
φ(1)
Res[f (z), 1] = = 3.
0!
Z
∴ f (z)dz = 2πi(2 + 3) = 10πi.
C
Residue at Infinity
Sometimes it is useful to include the point at ∞ in the complex plane, which we call the extended
complex plane. If a function f is analytic in the domain R1 < |z| < ∞, then ∞ is said to be an
isolated singularity of f . Further, residue of f at ∞ is defined as
Z
1
Res[f (z), ∞] = f (z)dz, (13)
2πi C0
where C0 : |z| = R0 (R0 > R1 ) is a negatively oriented circle.

Next assume that f is analytic in the entire complex plane except some finite number of singular
points inside a positively oriented simple closed contour C that lies in the interior of C1 : |z| = R1 .
Then f is analytic in the domain between C and C0 . Therefore by principle of the deformation of
path
Z Z
f (z)dz = − f (z)dz = −2πiRes[f (z), ∞]. (14)
C C0
Here minus sign appears since C0 is negatively oriented.

The function f is analytic in R1 ≤ |z| < ∞ with isolated singularity at ∞. So the function
Z |w| ≤ 1/R1 with isolated singularity at w = 0. Now let us use the
f (1/w) is analytic in 0 <
1
transformation z = w
in f (z)dz so that
C0
Z Z
1 1
f (z)dz = − f dw. (15)
C0 C∗ w2 w
where C∗ : |w| = 1/R1 is a circle with positive orientation since the transformation z = 1/w
1 1
changes the orientation. Now w2 f w is analytic within and on the positively oriented circle C∗
except 0. So by Cauchy residue theorem
Z
1 1 1 1
2
f dw = 2πiRes 2 f ,0 . (16)
C∗ w w z z
From (14), (15) and (16), we get

1 1
Res[f (z), ∞] = −Res 2 f ,0 . (17)
z z
and,
Z
1 1
f (z)dz = 2πiRes 2 f ,0 .
C z z
This formula is more economical than the Cauchy residue theorem for evaluating the integral of f
along the contour C as it involves the calculation of single residue. Remember the hypothesis for
this formula that the function f is analytic in the entire complex plane except a finite number of
points inside the positively oriented simple closed contour C.
5z − 2
Z
Ex. Evaluate dz, C : |z| = 2.
C z(z − 1)
5z−2
Sol. We see that the integrand f (z) = z(z−1) is analytic in the entire complex plane except the
two singularities, namely 0 and 1, which lie inside the circle C : |z| = 2. So we have
Z
1 1
f (z)dz = 2πiRes 2 f ,0 .
C z z
Now,

1 1 5 − 2z
f = .
z2 z z(1 − z)
We see that 0 is a simple pole of z12 f 1

. Here φ(z) = z. z12 f 1 5−2z

z z
= 1−z
and so φ(0) = 5. It follows
that
Z
f (z)dz = 2πi(5) = 10πi.
C
Zeros and Poles

Let a function f be analytic at z0 . Then z0 is said to be a zero of order m of f if f (z0 ) =
0,f (n) (z0 ) = 0 (n = 1, 2, ..., m − 1) and f (m) (z0 ) = 0. For example, z = 1 is a zero of order 2 of the
function f (z) = (z − 1)2 since f (1) = 0, f 0 (1) = 0 and f 00 (1) = 2 6= 0.
Further, it can be shown that f has a zero z0 of order m if and only if f (z) = (z − z0 )m g(z),
where g(z) is analytic and non-zero at z0 .
We know that a function f having a pole z0 of order m can be written in the form f (z) =
g(z)/(z − z0 )m , where g(z) is analytic and non-zero at z0 . Then the function h(z) = 1/g(z) is also
analytic and non-zero at z0 . It then implies that the function 1/f (z) = (z − z0 )m h(z) has a zero
of order m at z0 .
More generally, we can conclude that if two functions p and q are analytic at z0 such that
p(z0 ) 6= 0 and q has a zero of order m at z0 , then the function p/q has a pole of order m at z0 . In
case, m = 1, that is, z0 is a simple pole of p/q, then it can easily be shown that
Res [p(z)/q(z), z0 ] = p(z0 )/q 0 (z0 )
Ex. Find residues at the poles of cot z.
Sol. We have cot z = cos z/ sin z. The singularities of cot z are given by sin z = 0 or z = nπ, where
n is any integer. For each integer value of n, the singularity z = nπ is a simple pole of cot z. Also,
cos z is analytic and non-zero at z = nπ as cos nπ = (−1)n . So residue at z = nπ is given by
Res [cot z, nπ] = Res [cos z/ sin z, nπ] = cos nπ/ cos nπ = 1.
Ex. We know z = π is a zero of sin z. So sin z = (z − π)g(z), where g(z) is analytic and non-zero
at z = π. Determine g(z).
Sol. It is given by
g(z) = sin z/(z − π) for z 6= π and g(π) = −1.
Note that the value g(π) = −1 is defined so that g(z) is analytic at z = π.

1 Complex Introduction

Uploaded by

Copyright:

Available Formats

1 Complex Introduction

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 Complex Introduction

Uploaded by

Copyright:

Available Formats

Complex Analysis

Appeal: Please do not print this e-module unless it is really necessary.

Dr. Suresh Kumar, Department of Mathematics, BITS Pilani, Pilani Campus

Addition and Multiplication of complex numbers

z1 + z2 = (x1 , y1 ) + (x2 , y2 ) = (x1 + x2 , y1 + y2 ),

z1 z2 = (x1 , y1 )(x2 , y2 ) = (x1 x2 − y1 y2 , x1 y2 + y1 x2 ).

(x1 , 0) + (x2 , 0) = (x1 + x2 , 0)

(x1 , 0)(x2 , 0) = (x1 x2 , 0).

Simplified representation of complex number

z = (x, y) = (x, 0) + (0, y) = (x, 0) + (0, 1)(y, 0)

With this representation, the addition and multiplication operations become

z1 + z2 = (x1 + iy1 ) + (x2 + iy2 ) = (x1 + x2 ) + i(y1 + y2 ),

z1 z2 = (x1 + iy1 )(x2 + iy2 ) = (x1 x2 − y1 y2 ) + i(x1 y2 + y1 x2 ).

Some algebraic properties of complex numbers

Modulus of a complex number

denoted by zz21 is called the division of z1 by z2 .

Conjugate of a complex number

(1) z = z, z1 ± z2 = z1 ± z2 , z1 /z2 = z1 /z2 .

Polar or exponential form of a complex number

Regions in complex plane

Figure 1: Shaded region stands for |z| < 1, the neighbourhood of z = 0.

Interior, exterior and boundary points

Open and closed sets

Figure 3: 1 < |z| ≤ 2 is neither open nor closed.

Figure 4: S = {z : Rez < −1 or Rez > 1} is not connected.

Domains and regions

Limit points of a set

f (x + iy) = (x + iy)2 + 5 = x2 − y 2 + 5 + i(2xy) = u(x, y) + iv(x, y),

Multiple Valued Functions

Limit at an interior point

|f (z) − w0 | <  for 0 < |z − z0 | < δ.

Geometrically speaking, the limit of f (z) as z → z0 is w0 if given any -neighbourhood of w0 , there

Ex. Show that lim (2z + 3) = 5.

|2z + 3 − 5| = |2z − 2| = 2|z − 1| <  for |z − 1| < /2.

Choosing δ = /2, we have

|2z + 3 − 5| <  for |z − 1| < δ.

Limit at a boundary point

|iz/2 − i/2| = (1/2)|z − 1| <  for |z − 1| < 2 and |z| < 1.

Choosing δ = 2, we have

|iz/2 − i/2| <  for |z − 1| < δ and |z| < 1.

∴ lim (iz/2) = i/2.

(1) If z0 = x0 + iy0 and f (z) = u(x, y) + iv(x, y), then

lim f (z) = lim u(x, y) + i lim v(x, y).

(2) lim [f (z)g(z)] = lim f (z) lim g(z).

(3) lim [f (z) ± g(z)] = lim f (z) ± lim g(z).

(5) lim [kf (z)] = k lim f (z), where k is a constant.

(6) If f (z) = a0 + a1 z + ..... + an z n , then lim f (z) = f (z0 ).

Limits involving infinity

(1) lim f (z) = ∞ if and only if lim [1/f (z)] = 0.

(2) lim f (z) = lim f (1/z)

(3) lim f (z) = ∞ if and only if lim f (1/z) = 0.

|f (z) − f (z0 )| <  for |z − z0 | < δ.

|f (z) − f (z0 )| < |f (z0 )|/2 for |z − z0 | < δ.

If we choose z = z0 + δz, then we have

f (z0 + δz) − f (z0 )

Ex. If f (z) = z 2 + 1, then

f (2 + δz) − f (2) (2 + δz)2 + 1 − (22 + 1)

Ex. Show that f (z) = |z|2 is not differentiable at any z except z = 0.

Necessary condition for differentiability

|f (z) − w0 | < for 0 < |z − z0 | < δ.

Geometrically speaking, the limit of f (z) as z → z0 is w0 if given any -neighbourhood of w0 , there

|2z + 3 − 5| = |2z − 2| = 2|z − 1| < for |z − 1| < /2.

Choosing δ = /2, we have

|2z + 3 − 5| < for |z − 1| < δ.

|iz/2 − i/2| = (1/2)|z − 1| < for |z − 1| < 2 and |z| < 1.

Choosing δ = 2, we have

|iz/2 − i/2| < for |z − 1| < δ and |z| < 1.

|f (z) − f (z0 )| < for |z − z0 | < δ.

u(x0 + δx, y0 + δy) − u(x0 , y0 ) = ux (x0 , y0 )δx + uy (x0 , y0 )δy + 1 δx + 2 δy,

v(x0 + δx, y0 + δy) − v(x0 , y0 ) = vx (x0 , y0 )δx + vy (x0 , y0 )δy + 3 δx + 4 δy,

where 3 and 4 tend to 0 as (δx, δy) → (0, 0).