Chapter 4

INTEGRAL
1 Derivatives of Functions w (t)
In order to introduce integrals of f (z) in a fairly simple way, we need to first consider
derivatives of complex-valued functions w of a real variable t. We write

w (t) = u (t) + iv (t) , (1)

where the functions u and v are real-valued functions of t. The derivative of the
function (1) at a point t is defined as

w0 (t) = u0 (t) + iv 0 (t) , (2)

provided each of u0 and v 0 exists at t.

From (2), it follows that for every complex constant z0 = x0 + iy0 ,

d
[z0 w (t)] = [(x0 + iy0 ) (u + iv)]0 = [(x0 u − y0 v) + i (y0 u + x0 v)]0
dt
= (x0 u − y0 v)0 + i (y0 u + x0 v)0 = x0 u0 − y0 v 0 + i (y0 u0 + x0 v 0 )
= (x0 + iy0 ) (u0 + iv 0 ) = z0 w0 (t) .

Another expected rule that we shall often use is

d z0 t
e = z0 ez0 t , (3)
dt
where z0 = x0 + iy0 . To verify this, we write

ez0 t = ex0 t eiy0 t = ex0 t cos y0 t + iex0 t sin y0 t

and refer to definition (2) to see that

d z0 t
0
0
e = ex0 t cos y0 t + i ex0 t sin y0 t .
dt
Familiar rules from calculus and some simple algebra then lead us to the expression
d z0 t
= (x0 + iy0 ) ex0 t cos y0 t + iex0 t sin y0 t


e
dt
= (x0 + iy0 ) ex0 t eiy0 t = z0 ez0 t .

1
2 Definite Integrals of Functions w (t)
When w (t) is a complex-valued function of a real variable t and is written
w (t) = u (t) + iv (t) ,
where u and v are real-valued, the definite integral of w (t) over an interval a ≤ t ≤ b
is defined as Z b Z b Z b
w (t) dt = u (t) dt + i v (t) dt, (4)
a a a
provided the individual integrals on the right exist. Thus
Z b Z b
Re w (t) dt = Re [w (t)] dt
a a

and Z b Z b
Im w (t) dt = Im [w (t)] dt.
a a

Example 2.1 For an illustration of definition (4),

Z 1 Z 1 Z 1
2 2

2
(1 + it) dt = 1 − t dt + i 2tdt = + i.
0 0 0 3
Improper integral of w (t) over unbounded intervals are defined in a similar way.
The existence of the integrals of u and v in (4) is ensured if those functions are
piecewise continuous on the interval a ≤ t ≤ b. Such a function is continuous every
where in the stated interval except possibly for a finite number of points where,
although discontinuous, it has one-sided limits. Of course, only the right-hand limit
is required at a; and only the left-hand limit is required at b. When both u and v are
piecewise continuous, the function w is said to have that property.
Anticipated rules for integrating a complex constant times a function w (t), for
integrating sums of such functions, and for interchanging limits of integration are all
valid. Those rules, as well as the property
Z b Z c Z b
w (t) dt = w (t) dt + w (t) dt,
a a c

are easy to verify by recalling corresponding results in calculus.

The fundamental theorem of calculus, involving antiderivatives, can, moreover, be
extended so as to apply to integrals of the type (4). To be specific, suppose that the
functions
w (t) = u (t) + iv (t) and W (t) = U (t) + iV (t)
are continuous on the interval a ≤ t ≤ b. If W 0 (t) = w (t) when a ≤ t ≤ b, then
U 0 (t) = u (t) and V 0 (t) = v (t). Hence, in view of (4),
Z b
w (t) dt = [U (t)]ba + i [V (t)]ba = [U (b) + iV (b)] − [U (a) + iV (a)]
a
= W (b) − W (a)
= [W (t)]ba

2
Example 2.2 Since by (3)

eit
 
d 1 d it 1 it
= e = ie = eit ,
dt i i dt i

one can see that

Z π/4 π/4
eit eiπ/4 1

it 1 π π 
e dt = = − = cos + i sin − 1
0 i
0 i i i 4 4
   
1 1 i 1 1 1
= √ + √ −1 = √ + √ −1
i 2 2 2 i 2
 
1 1
= √ +i 1− √ .
2 2

3 Contours
Integrals of complex-valued functions of a complex variable are defined on curves in
the complex plane, rather than on just intervals of the real line. Classes of curves
that are adequate for the study of such integrals are introduced in this section.
A set of points z = (x, y) in the complex plane is said to be an arc if

x = x (t) , y = y (t) , a ≤ t ≤ b,

where x (t) and y (t) are continuous functions of the real parameter t. This definition
establishes a continuous mapping of the interval a ≤ t ≤ b into the z plane; and
the image points are ordered according to increasing values of t. It is convenient to
describe the points of C by means of the equation

z = z (t) , a ≤ t ≤ b, (5)

where
z (t) = x (t) + iy (t) . (6)
The arc C is a simple arc or Jordan arc, if it does not cross itself; that is, C is
simple if z (t1 ) 6= z (t2 ) when t1 6= t2 , except possibly when t1 = a and t2 = b. When
the arc C is simple except for the fact that z (b) = z (a), we say that C is a simple
closed curve or a Jordan curve. Such a curve is positively oriented when it is in the
counterclockwise direction.

Example 3.1 The polygonal line defined by means of the equations


x + ix when 0 ≤ x ≤ 1
z= (7)
x + i when 1 ≤ x ≤ 2

and consisting of a line segment from 0 to 1 + i followed by one from 1 + i to 2 + i

(Figure ) is a simple arc.

3
Example 3.2 The unit circle

z = eiθ , 0 ≤ θ ≤ 2π (8)

about the origin is a simple closed curve, oriented in the counterclockwise direction.
So is the circle
z = z0 + Reiθ , 0 ≤ θ ≤ 2π, (9)
centered at the point z0 and with radius R.

The same set of points can make up different arcs.

Example 3.3 The arc

z = e−iθ , 0 ≤ θ ≤ 2π (10)
is not the same as the arc described by (8). The set of points is the same, but now
the circle is traversed in the clockwise direction.

Example 3.4 The points on the arc

z = ei2θ , 0 ≤ θ ≤ 2π

are the same as those making up the arcs (8) and (10). The arc here differs, how-
ever, from each of those arc since the circle is traversed twice in the counterclockwise
direction.

The parametric representation used for any given C is not unique. It is possible
to change the interval over which the parameter ranges to any other interval. To be
specific, suppose that
t = φ (τ ) , α ≤ τ ≤ β, (11)
where φ is a real-valued function mapping an interval α ≤ τ ≤ β onto the interval
a ≤ t ≤ b in (5) (Figure ). We assume that φ is continuous with a continuous
derivative. We also assume that φ0 (τ ) > 0 for each τ ; this ensures that t increase
with τ . Representation (5) is then transformed by equation (11) into

z = Z (τ ) , α ≤ τ ≤ β, (12)

where  
b−a αβ − bα
Z (τ ) = z [φ (τ )] = z τ+ . (13)
β−α β−α
Suppose that the components x0 (t) and y 0 (t) of the derivative

z 0 (t) = x0 (t) + iy 0 (t)

of function (6), used to represent C, are continuous on the entire interval a ≤ t ≤ b.

The arc is then called a differentiable arc, and the real-valued function
q
|z (t)| = [x0 (t)]2 + [y 0 (t)]2
0

4
is integrable over the interval a ≤ t ≤ b. In fact, according to the definition of arc
length in calculus, the length of C is the number
Z b
L= |z 0 (t)| dt. (14)
a

The value of L is invariant under certain changes in the representation for C that
is used, as one would expect. More precisely, with the change of variable indicated
in (11), expression (14) takes the form
Z β
L= |z 0 [φ (τ )]| φ0 (τ ) dτ .
α

So, if representation (12) is used for C, the derivative

Z 0 (τ ) = z 0 [φ (τ )] φ0 (τ )

enable us to write expression (14) as

Z β
L= |Z 0 (τ )| dτ .
α

Thus the same length of C would be obtained if representation (12) were to be used.
If equation (5) represents a differentiable arc and if z 0 (t) 6= 0 anywhere in the
interval a ≤ t ≤ b, then the unit tangent vector

z 0 (t)
T=
|z 0 (t)|

is well defined for all t in that interval, with angle of inclination arg z 0 (t). Also, when
T turns, it does so continuously as the parameter t varies over the entire interval
a ≤ t ≤ b. This expression for T is the one learned in calculus when z (t) is interpreted
as a radius vector. Such an arc is said to be smooth. In referring to a smooth arc
z = z (t) , a ≤ t ≤ b, then we agree that the derivative z 0 (t) is continuous and nonzero
on the closed interval a ≤ t ≤ b.
A contour or piecewise smooth arc, is an arc consisting of a finite number of
smooth arcs joined end to end. Hence if equation (5) represent a contour, z (t) is
continuous, whereas its derivative z 0 (t) is piecewise continuous. The polygonal line
(7) is for example, a contour. When only the initial and final values of z (t) are the
same, a contour C is called a simple closed contour. Examples are the circles (8) and
(9), as well as the boundary of a triangle or a rectangle taken in a specific direction.
The length of a contour or a simple closed contour is the sum of the lengths of the
smooth arcs that make up the contour.
The points on any simple closed curve or simple closed contour C are boundary
points of two distinct domains, one of which is the interior of C and is bounded. The
other, which is the exterior of C, is unbounded.

5
4 Contour Integrals
We turn now to integrals of complex-valued functions f of the complex variable z.
Such an integral is defined in terms of the values f (z) along a given contour C,
extending from a point z = z1 to a point z = z2 in the complex plane. It is, therefore,
a line integral; and its value depends, in general, on the contour C as well as on the
function f . It is written
Z Z z2
f (z) dz or f (z) dz,
C z1

the latter notation often being used when the value of the integral is independent of
the choice of the contour taken between two fixed end points. While the integral may
be defined directly as the limit of a sum, we choose to define it in terms of a definite
integral.
Suppose that the equation
z = z (t) , a ≤ t ≤ b (15)
represents a contour C, extending from z1 = z (a) to z2 = z (b). We assume that
f [z (t)] is piecewise continuous on C. We then define the line integral or contour
integral, of f along C in terms of the parameter t :
Z Z b
f (z) dz = f [z (t)] z 0 (t) dt. (16)
C a
0
Since C is a contour, z (t) is also piecewise continuous on a ≤ t ≤ b; and so the
existence of integral (16) is ensured.
The value of a contour integral is invariant under a change in the representation
of its contour when the change is of the type (13). This can be seen by following
the same general procedure that was used in Section 3 to show the invariance of arc
length.
Proof. Assume that the parametric representation of contour C given by
z = z (t) , a ≤ t ≤ b.
Let this representation transformed to
z = Z (τ ) , α ≤ τ ≤ β
where
Z (τ ) = z (φ (τ )) and t = φ (τ ) .
Hence
Z Z β
f (z) dz = f (Z (τ )) Z 0 (τ ) dτ
Z(τ ) α
Z β
= f [z (φ (τ ))] z 0 (φ (τ )) φ0 (τ ) dτ
α
Z b
= f [z (t)] z 0 (t) dt
Za
= f (z) dz.
z(t)

6
 It follows immediately from (16) and properties of integrals of complex-valued
functions w (t) mentioned in Section 2 that
Z Z
z0 f (z) dz = z0 f (z) dz,
C C

for any complex constant z0 , and

Z Z Z
[f (z) + g (z)] dz = f (z) dz + g (z) dz.
C C C

Associated with the contour C used in integral (16) is the contour −C, consisting
of the same set of points but with the order reversed so that the new contour extends
from z2 to z1 (Figure ). The contour −C has parameter representation
z = z (−t) , − b ≤ t ≤ −a.
Hence, by taking τ = −t, we have dτ = −dt and a ≤ τ ≤ b. Therefore
Z Z −a
f (z) dz = f [z (−t)] z 0 (−t) dt
−C −b
Z b
= f [z (τ )] z 0 (τ ) (−dτ )
a
Z b
= − f [z (τ )] z 0 (τ ) dτ
Za
= − f (z) dz
C

where z 0 (−t) denotes the derivative of z (t) with respect to t, evaluated at −t.
Consider now a path C, with representation (15), that consists of a contour C1
from z1 to z2 followed by a contour C2 from z2 to z3 , the initial point of C2 being
the final point of C1 (Figure ). There is a value c of t, where a < c < b, such that
z (c) = z2 . Consequently, C1 is represented by
z = z (t) , a ≤ t ≤ c
and C2 is represented by
z = z (t) , c ≤ t ≤ b.
Also, by a rule for integrals of function w (t) that was noted in Section 2,
Z b Z c Z b
0 0
f [z (t)] z (t) dt = f [z (t)] z (t) dt + f [z (t)] z 0 (t) dt.
a a c

Evidently, then Z Z Z
f (z) dz = f (z) dz + f (z) dz.
C C1 C2
Sometimes the contour C is called the sum of its legs C1 and C2 and is denoted by
C1 + C2 . The sum of two contours C1 and −C2 is well defined when C1 and C2 have
the same final points, and it is written C1 − C2 (Figure ).
Definite integrals in calculus can be interpreted as areas, and they have other
interpretations as well.

7
4.1 Some Examples
Example 4.1 Let us find the value of the integral
Z
I= zdz
C

when C is the right-hand half

π π
z = 2eiθ , − ≤θ≤
2 2
of the circle |z| = 2 from z = −2i to z = 2i (Figure ). According to Definition (16),
Z π/2 Z π/2
iθ 0

0
eiθ eiθ dθ;


I= 2eiθ 2e dθ = 4
−π/2 −π/2

and since
0
eiθ = e−iθ and eiθ = ieiθ ,
this means that Z π/2 Z π/2
−iθ iθ
I=4 e ie dθ = 4i dθ = 4πi.
−π/2 −π/2

Note that zz = |z|2 = 4 when z is a point on the semicircle C. Hence the result
Z
zdz = 4πi
C

can also be written Z

dz
= πi.
C z

If f (z) is given in the form f (z) = u (x, y) + iv (x, y) , where z = x + iy, one can
sometime apply definition (16), using one of the variable x and y as the parameter.

Example 4.2 Here we first let C1 denote the polygonal line OAB shown in Figure
and evaluate the integral
Z Z Z
I1 = f (z) dz = f (z) dz + f (z) dz, (17)
C1 OA AB

where
f (z) = y − x − i3x2 , z = x + iy.
The leg OA may be represented parametrically as z = 0 + iy, 0 ≤ y ≤ 1; and, since
x = 0 at points on that line segment, the value of f there vary with the parameter y
according to the equation f (z) = y, 0 ≤ y ≤ 1. Consequently,
Z Z 1 Z 1
i
f (z) dz = yidy = i ydy = .
OA 0 0 2

8
On the leg AB, the points are z = x + i, 0 ≤ x ≤ 1; and, since y = 1 on this segment,
Z Z 1 Z 1 Z 1
2 1
x2 dx = − i.


f (z) dz = 1 − x − i3x · 1dx = (1 − x) dx − 3i
AB 0 0 0 2
In view of (17), we now see that
i 1 1−i
I1 = + − i = .
2 2 2
If C2 denotes the segment OB of the line y = x, with parametric representation
z = x + ix, 0 ≤ x ≤ 1, the fact that y = x on OB enables us to write
Z Z 1 Z 1
2
I2 = f (z) dz = −i3x (1 + i) dx = 3 (1 − i) x2 dx = 1 − i.
C2 0 0

Then the integrals of f (z) along the two path C1 and C2 have different values even
though those paths have the same initial and the same final points.
Observe how it follows that the integral of f (z) over the simple closed contour
OABO or C1 − C2 , has the nonzero value
1−i −1 + i
I1 − I2 = − (1 − i) = .
2 2
Example 4.3 We begin here by letting C denote an arbitrary smooth arc
z = z (t) , a ≤ t ≤ b
from a fixed point z1 to a fixed point z2 (Figure ). In order to evaluate the integral
Z Z b
zdz = z (t) z 0 (t) dt,
C a
we note that
d [z (t)]2
= z (t) z 0 (t) .
dt 2
Because z (a) = z1 and z (b) = z2 , we have
#b
[z (t)]2 [z (b)]2 − [z (a)]2 z 2 − z12
Z
zdz = = = 2 .
C 2 2 2
a
Inasmuch as the value of this integral depends only on the end points of C and is
otherwise independent of the arc that is taken, we may write
Z z2
z 2 − z12
zdz = 2 . (18)
z1 2
Expression (18) also valid when C is a contour that is not necessarily smooth since
a contour consists of a finite number of smooth arcs Ck , k = 1, 2, . . . , n, joined end to
end. More precisely, suppose that each Ck extends from zk to zk+1 . Then
n Z n Z zk+1 n 2
zk+1 − zk2 z 2 − z12
Z X X X
zdz = zdz = zdz = = n+1 , (19)
C k=1 Ck k=1 zk k=1
2 2
where this last summation has telescoped and z1 is the initial point of C and zn+1 is
its final point.
It follows from (19) that the integral of the function f (z) = z around each closed
contour in the plane has value zero.

9
4.2 Examples With Branch Cuts
The path in a contour integral can contain a point on a branch cut of the integrand
involved. The next examples illustrate this.

Example 4.4 Let C denote the semicircular path

z = 3eiθ , 0 ≤ θ ≤ π
from the point z = 3 to the point z = −3 (Figure ). Although the branch (from
Chapter 3)  
1/2 1
f (z) = z = exp log z , |z| > 0, 0 < arg z < 2π
2
of the multiple-valued function z 1/2 is not defined at the initial point z = 3 of the
contour C, the integral Z
I= z 1/2 dz
C
nevertheless exists. For the integrand is piecewise continuous on C. To see that this
is so, we first observe that when z (θ) = 3eiθ ,
√
 
1
f [z (θ)] = exp (ln 3 + iθ) = 3eiθ/2 .
2
Hence the right-hand limits of the real and imaginary components of the function
√ iθ/2 iθ √
f [z (θ)] z 0 (θ) = 3e 3ie = 3 3iei3θ/2
√ 3θ √ 3θ
= −3 3 sin + i3 3 cos , 0 < θ ≤ π
2 2
√
at θ = 0 exist, those limits being 0 and i3 3, respectively. This means that f [z (θ)] z 0 (θ)
continuous on the closed interval 0 ≤ θ ≤ π when its value at θ = 0 is defined as
is √
i3 3. Consequently,
Z
1/2
Z Z π
0
√ Z π i3θ/2
I = z dz = f (z) dz = f [z (θ)] z (θ) dθ = 3 3i e dθ
C C 0 0
π
√ √

2 i3θ/2
= 2 3 ei3π/2 − 1


= 3 3i e
3i 0
√
= −2 3 (1 + i) .

Example 4.5 Suppose that C is the positively oriented circle (Figure )

z = R eiθ , −π ≤θ ≤π
about the origin, and let a denote any nonzero real number. Using the principal branch
f (z) = z a−1 = exp [(a − 1) Log z] , |z| > 0, −π < Arg z < π
of the power function z a−1 , let us evaluate the integral
Z
I= z a−1 dz. (20)
C

10
 When z (θ) = R eiθ , it is easy to see that
f [z (θ)] z 0 (θ) = iRa eiaθ = −Ra sin aθ + iRa cos aθ,
where the positive value of Ra is to be taken. Inasmuch as this function is piecewise
continuous on −π < θ < π, integral (20) exists. In fact,
Z Z π Z π
a−1 0 a
I = z dz = f [z (θ)] z (θ) dθ = iR eiaθ dθ
C −π −π
 iaθ π  iaπ −iaπ

e e −e
= iRa = iRa
ia ia
a
 iaπ−π −iaπ 
2R e − e 2Ra
= i =i sin aπ.
a 2i a
Note that if a is a nonzero integer n, this result tells us that
Z
z n−1 dz = 0, n ∈ Z.
C

If a is allowed to be zero, we have

Z Z π Z π
dz 1 iθ
= iθ
iR e dθ = i dθ = 2πi.
C z −π R e −π

5 Upper Bound for Moduli of Contour Integrals

Lemma 5.1 If w (t) is a piecewise continuous complex-valued function defined on an
interval a ≤ t ≤ b, then Z b  Z b
 

 w (t) dt ≤ |w (t)| dt.
a a

Proof. Exercise (refer book page 137).

Theorem 5.2 Let C denote a contour of length L, and suppose that a function f (z)
is piecewise continuous on C. If M is a nonnegative constant such that
|f (z)| ≤ M (21)
for all point z on C at which f (z) is defined, then
Z 
 
 f (z) dz  ≤ M L. (22)
 
C

Proof. Let z = z (t) , a ≤ t ≤ b be a parametric representation of C. According to

the above lemma,
Z  Z b  Z b
0
|f [z (t)] z 0 (t)| dt
   
 f (z) dz  =
 

 f [z (t)] z (t) dt ≤
C a a
Z b Z b
= |f [z (t)]| |z 0 (t)| dt ≤ M |z 0 (t)| dt
a a
Z b
= M |z 0 (t)| dt
a

11
when a ≤ t ≤ b. Since the integral on the right here represents the length L of C,
inequality (22) is established. It is a strict inequality if inequality (21) is strict.
Example 5.1 Let C be the arc of the circle |z| = 2 from z = 2 to z = 2i that lies in
the first quadrant (Figure ). Inequality (22) can be used to show that
Z 
 z + 4  6π
 z 3 − 1 dz  ≤ 7 . (23)

C

This is done by noting first that if z is a point on C, so that |z| = 2, then

|z + 4| ≤ |z| + 4 = 6
and
z − 1 ≥ |z|3 − 1 = 7.
 3   

Thus, when z lies on C,  

 z+4  |z + 4| 6
 z 3 − 1  = |z 3 − 1| ≤ 7 .
 

Writing M = 6/7 and observing that L = π is the length of C, we may now use
inequality (22) to obtain inequality (23).
Example 5.2 Here CR is the semicircular path
z = R eiθ , 0 ≤ θ ≤ π,
and z 1/2 denotes the branch
√
 
1/2 1 π 3π
z = exp log z = reiθ/2 , r > 0, − < θ <
2 2 2
of the square root function (Figure ). Without actually finding the value of the
integral, one can easily show that
z 1/2
Z
lim dz = 0. (24)
R→∞ C z 2 + 1
R

For, when |z| = R > 1,

 1/2  √ iθ/2  √
z  =  Re  = R

and  2    
z + 1 ≥ z 2  − 1 = R2 − 1.
Consequently, at points on CR ,
 1/2  √
 z  R
 z 2 + 1  ≤ MR where MR = R2 − 1 .
 

Since the length of CR is the number L = πR, it follows from inequality (22) that
1/2
Z 
 z 

2
dz  ≤ MR L.
CR z + 1

But √ √
πR R 1/R2 π/ R
MR L = 2 · = → 0 as R → ∞.
R − 1 1/R2 1 − (1/R2 )
Limit (24) is therefore established.

12
6 Antiderivatives
Although the value of a contour integral of a function f (z) from a fixed point z1 to
a fixed point z2 depends on the path that is taken, there are certain functions whose
integrals from z1 to z2 have values that are independent of path. (Recall Examples
4.2 and 4.3). The examples just cited also illustrate the fact that the values of
integrals around closed paths are sometimes, but not always, zero. Our next theorem
is useful in determining when integration is independent of path and, moreover, when
an integral around a closed path has value zero.
The theorem contains an extension of the fundamental theorem of calculus that
simplifies the evaluation of many contour integrals. The extension involves the con-
cept on an antiderivative of a continuous function f (z) on a domain D, or a function
F (z) such that F 0 (z) = f (z) for all z ∈ D.
Theorem 6.1 Suppose that a function f (z) is continuous on a domain D. If any
one of the following statements is true, then so are the others:
(a) f (z) has an antiderivative F (z) throughout D;
(b) the integral of f (z) along contours lying entirely in D and extending from any
fixed point z1 to any fixed point z2 all have the same value, namely
Z z2
f (z) dz = F (z)]zz21 = F (z2 ) − F (z1 )
z1

where F (z) is the antiderivative in statement (a) ;

(c) the integrals of f (z) around closed contours lying entirely in D all have value
zero.
It should be emphasized that the theorem does not claim that any of these state-
ments is true for a given function f (z). It says only that all of them are true or that
none of them is true.
Example 6.1 The continuous function f (z) = z 2 has an antiderivative F (z) = z 3 /3
throughout the plane. Hence
Z 1+i 1+i
z3 1 2
2
z dz = = (1 + i)3 = (−1 + i)
0 3 0 3 3
for every contour from z = 0 to z = 1 + i.
Example 6.2 The function f (z) = 1/z 2 , which is continuous everywhere except at
the origin, has an antiderivative F (z) = −1/z in the domain |z| > 0, consisting of
the entire plane with the origin deleted. Consequently,
Z
dz
2
=0
C z

when C is the positively oriented circle (Figure )

z = 2eiθ , −π ≤θ ≤π
about the origin.

13
 Note that the integral of the function f (z) = 1/z around the same circle cannot
be evaluated in a similar way. For, although the derivative of any branch F (z) of log z
is 1/z, F (z) is not differentiable, or even defined, along its branch cut. In particular
if a ray θ = α from the origin is used to form the branch cut, F 0 (z) fails to exist
at the point where that ray intersects the circle C (Figure ). So C does not lie in
any domain throughout which F 0 (z) = 1/z, and one cannot make direct use of an
antiderivative. The following example illustrates how a combination of two different
antiderivatives can be used to evaluate f (z) = 1/z around C.

Example 6.3 Let C1 denote the right half

π π
z = 2eiθ , − ≤θ≤ (25)
2 2
of the circle C in Example 6.2 (Figure ). The principal branch

Log z = ln r + iΘ, r > 0, −π < Θ < π

of the logarithmic function serves as an antiderivative of the function 1/z in the

evaluation of the integral of 1/z along C1 :
Z Z 2i
dz dz
= = Log z]2i
−2i = Log (2i) − Log (−2i)
C1 z −2i z
 π  π
= ln 2 + i − ln 2 − i = πi.
2 2
This integral was evaluated in another way in Example 4.1, where representation (25)
for the semicircle was used.
Next, let C2 denote the left half
π 3π
z = 2eiθ , ≤θ≤
2 2
of the same circle C and consider the branch

log z = ln r + iθ, r > 0, 0 < θ < 2π

of the logarithmic function (Figure ). One can write

Z Z −2i
dz dz
= = log z]−2i
2i = log (−2i) − log (2i)
C2 z 2i z
  
3π π
= ln 2 + i − ln 2 + i = πi.
2 2

The value of the integral of 1/z around the entire circle C = C1 + C2 is thus obtained:
Z Z Z
dz dz dz
= + = πi + πi = 2πi.
C z C1 z C2 z

14
7 Cauchy-Goursat Theorem
Theorem 7.1 (Cauchy) If a function f is analytic at all points interior to and on
simple closed contour C and f 0 is continuous there, then
Z
f (z) dz = 0.
C

Example 7.1 If C is any simple closed contour, in either direction, then

Z
exp z 3 dz = 0.


C

This is because the composite function f (z) = exp (z 3 ) is analytic everywhere and its
derivative f 0 (z) = 3z 2 exp (z 3 ) is continuous everywhere.

Theorem 7.2 (Cauchy-Goursat Theorem) If a function f is analytic at all points

interior to and on simple closed contour C, then
Z
f (z) dz = 0.
C

8 Simply Connected Domains

A simply connected domain D is a domain such that every simple closed contour
within it encloses only points of D. The set of point interior to a simple closed contour
is an example. The annular domain between two concentric circle is, however, not
simply connected.
The closed contour in the Cauchy-Goursat Theorem need not be simple when the
theorem is adapted to simply connected domains. More precisely, the contour can
actually cross itself. The following theorem allows for this possibility.

Theorem 8.1 If a function f is analytic throughout a simply connected domain D,

then Z
f (z) dz = 0 (26)
C
for every closed contour C lying in D.

Example 8.1 If C denotes any closed contour lying in the open disk |z| < 2 (Figure
), then
zez
Z
2 5 dz = 0.
C (z + 9)

This is because the disk is a simply connected domain and the two singularities z = ±3i
of the integrand are exterior to the disk.

Corollary 8.1 A function f that is analytic throughout a simply connected domain

D must have an antiderivative everywhere in D.

15
Proof. Observe that a function f is continuous on a domain D when it is ana-
lytic there. Consequently, since (26) holds for the function in the hypothesis of this
corollary and for each closed contour C in D, f has an antiderivative throughout D,
according to the Theorem 6.1.

Note that, since the complex plane is simply connected, the corollary tells us that
entire functions always possess antiderivatives.

9 Cauchy Integral Formula

Theorem 9.1 Let f be analytic everywhere inside and on a simple closed contour
C, taken in the positive sense. If z0 is any point interior to C, then
Z
1 f (z)
f (z0 ) = dz. (27)
2πi C z − z0
Formula ( 27 ) is called Cauchy integral formula. It tells us that if a function f is
to be analytic within and on a simple closed contour C, then the values of f interior
to C are completely determined by the values of f on C.
When the Cauchy integral formula is written as
Z
f (z) dz
= 2πif (z0 ) , (28)
C z − z0

it can be used to evaluate certain integrals along simple closed contours.

Example 9.1 Let C be the positively oriented circle |z| = 2. Since the function
z
f (z) =
9 − z2
is analytic within and on C and since the point z0 = −i is interior to C, formula (28)
tells us that
z/ (9 − z 2 )
 
−i
Z Z
zdz π
2
= dz = 2πi = .
C (9 − z ) (z + i) C z − (−i) 10 5

9.1 An Extension of The Cauchy-Integral Formula

The Cauchy integral formula in the Theorem 9.1 can be extended so as to provide an
integral representation for derivative of f at z0 . To obtain the extension, we consider
a function f that is analytic everywhere inside and on a simple closed contour, taken
in the positive sense. We then write the Cauchy integral formula as
Z
1 f (s) ds
f (z) = , (29)
2πi C s − z
where z is interior to C and s denotes points on C. Differentiating formally with
respect to z under the integral sign here, without rigorous justification, we find that
Z
0 1 f (s) ds
f (z) = .
2πi C (s − z)2

16
 Continuing the same technique, we obtain
Z
00 1 f (s) ds
f (z) = . (30)
πi C (s − z)3

Mathematical induction can be used to obtain the formula

Z
(n) n! f (s) ds
f (z) = , n = 1, 2, . . . . (31)
2πi C (s − z)n+1

Note that with the agreement that

f (0) (z) = f (z) and 0! = 1,

expression (31) is also valid when n = 0, in which case it becomes the Cauchy integral
formula (29)
When written in the from
Z
f (z) dz 2πi (n)
n+1 = f (z0 ) , n = 0, 1, 2, . . . , (32)
C (z − z0 ) n!

expressions (29) and (31) can be useful in evaluating certain integrals when f is
analytic inside and on a simple closed contour C, taken in the positive sense, and z0
is any point interior to C.

Example 9.2 If C is the positively oriented unit circle |z| = 1 and

f (z) = exp (2z) ,

then Z Z
exp (2z) dz f (z) dz 2πi 000 8πi
= 3+1 = f (0) = .
C z4 C (z − 0) 3! 3

Example 9.3 Let z0 be any point interior to a positively oriented simple closed con-
tour C. When f (z) = 1, expression (32) shows that
Z
dz
= 2πi
C z − z0

and Z
dz
= 0, n = 1, 2, . . . .
C (z − z0 )n+1

9.2 Some Consequences of The Extension

Theorem 9.2 If a function f is analytic at a given point, then its derivative of all
orders are analytic there too.

17
Proof. Assume that a function f is analytic at a point z0 . There must, then, be a
neighborhood |z − z0 | < ε of z0 throughout which f is analytic. Consequently, there
is a positively oriented circle C0 , centered at z0 with radius ε/2, such that f is analytic
inside and on C0 (Fig ). From (30), we know that
Z
00 1 f (s) ds
f (z) =
πi C0 (s − z)3

at each point z interior to C0 , and the existence of f 00 (z) throughout the neighborhood
|z − z0 | < ε/2 means that f 0 is analytic at z0 . One can apply to the analytic function
f 0 to conclude that its derivative f 00 is analytic, etc. Theorem 9.2 is now established.

Corollary 9.1 If a function f (z) = u (x, y)+iv (x, y) is analytic at a point z = (x, y),
then the component functions u and v have continuous partial derivatives of all orders
at that point.

Proof. When a function

f (z) = u (x, y) + iv (x, y)
is analytic at a point z = (x, y), then f 0 is analytic at z by Theorem 9.2. The
differentiability of f 0 , i.e., f 00 exists, ensures the continuity of f 0 there. Since

f 0 (z) = ux + ivx = vy − iuy ,

we may conclude that the first-order partial derivatives of u and v are continuous at
that point. Furthermore, since f 00 is analytic at z, f 00 continuous there and since

f 00 (z) = uxx + ivxx = vyx − iuyx ,

the second-order partial derivatives of u and v are continuous at that point, etc.
Corollary 9.1 is now established.

The proof of the next theorem, due to E. Morera (1856 - 1909), depends on the
fact that the derivative of an analytic function is itself analytic, as stated in Theorem
9.2.

Theorem 9.3 Let f be continuous on a domain D. If

Z
f (z) dz = 0
C

for every closed contour C in D, then f is analytic throughout D.

Proof. When its hypothesis is satisfied, the Theorem 6.1 ensures that f has an
antiderivative in D; that is, there exists an analytic function F such that F 0 (z) = f (z)
at each point in D. Since f is the derivative of F , it then follows from Theorem 9.2
that f is analytic in D.

18
Theorem 9.4 Suppose that a function f is analytic inside and on a positively ori-
ented circle CR , centered at z0 with radius R (Figure ). If MR denotes the maximum
value of |f (z)| on CR , then

f (z0 ) ≤ n!MR , n = 1, 2, . . . .
 (n) 
(33)
Rn
Inequality (33) is called Cauchy’s inequality.
Proof. Inequality (33) is an immediate consequence of the expression
Z
(n) n! f (z) dz
f (z0 ) = , n = 1, 2, . . . ,
2πi CR (z − z0 )n+1

which is a slightly different form of (31), when n ∈ Z+ . We need only apply the
Theorem 5.2, which gives upper bound for the moduli of the values of contour integrals
 
 f (z)  |f (z)| |f (z)| MR
 (z − z )n+1  = |z − z |n+1 = Rn+1 ≤ Rn+1 ,
 
0 0

to see that
Z 
 (n)  n!  f (z) dz  n! MR n!MR
n+1  ≤ · n+1 2πR =
f (z0 ) =   , n = 1, 2, . . . ,
2π CR (z − z0 )
 2π R Rn

where MR is as in the statement of Theorem 9.4.

10 Liouville’s Theorem and The Fundamental The-

orem of Algebra
Theorem 10.1 (Liouville’s Theorem) If a function f is entire and bounded in
the complex plane, then f (z) is constant throughout the plane.

Proof. Assume that f is as stated and since f is entire, Cauchy’s inequality in

Theorem 9.4 tells us that when n = 1,
MR
|f 0 (z0 )| ≤ . (34)
R
Moreover, the boundedness condition on f tells us that a nonnegative constant M
exists such that |f (z)| ≤ M for all z; and, because the constant MR in (34) is always
less than or equal to M , it follows that
M
|f 0 (z0 )| ≤ , (35)
R
where R can be arbitrarily large. Now the number M in (35) is independent of the
value of R that is taken. Hence that inequality holds for arbitrarily large values of
R only if f 0 (z0 ) = 0. Since the choice of z0 was arbitrary, this means that f 0 (z) = 0
everywhere in the complex plane. Consequently, f is a constant function, according
to the Theorem in Chapter 2.

19
Theorem 10.2 (Fundamental Theorem of Algebra) Any polynomial

P (z) = a0 + a1 z + a2 z 2 + · · · + an z n , an 6= 0

of degree n, n ≥ 1 has at least one zero. That is, there exists at least one point z0
such that P (z0 ) = 0.

Proof. Suppose that P (z) is not zero for any value of z. Then the reciprocal
1
f (z) =
P (z)
is clearly entire, and it is also bounded in the complex plane.
To show that it is bounded, we first write
a0 a1 a2 an−1
w= n
+ n−1 + n−2 + · · · + , (36)
z z z z
so that
P (z) = (an + w) z n . (37)
Next we observe that a sufficiently large positive number R can be found such that
the modulus of each of the quotients in (36) is less than the number |an | / (2n) where
|z| > R. The generalized triangle inequality from Chapter 1

|z1 + z2 + · · · + zn | ≤ |z1 | + |z2 | + · · · + |zn |

thus shows that

|an |
|w| < whenever |z| > R.
2
Consequently,
|an |
|an + w| ≥ ||an | − |w|| > whenever |z| > R.
2
This inequality and (37) enable us to write
|an | n |an | n
|P (z)| = |an + w| |z|n > |z| > R whenever |z| > R.
2 2
Evidently, then
1 2
|f (z)| = < whenever |z| > R.
|P (z)| |an | Rn
So f is bounded in the region exterior to the disk |z| ≤ R. But f is continuous in
that closed disk, and this means that f is bounded there too. Hence f is bounded
in the entire plane. It follows from Liouville’s theorem that f (z) and consequently
P (z) is constant. But P (z) is not constant, and we have reached a contradiction.

The fundamental theorem tells us that any polynomial P (z) of degree n, n ≥ 1

can be expressed as a product of linear factors:

P (z) = c (z − z1 ) (z − z2 ) · · · (z − zn ) , (38)

20
where c and zk , k = 1, 2, . . . , n are complex constants. More precisely, the theorem
ensures that P (z) has a zero z1 . Then,

P (z) = (z − z1 ) Q1 (z) ,

where Q1 (z) is a polynomial of degree n − 1. The same argument, applied to Q1 (z)

reveals that there is a number z2 such that

P (z) = (z − z1 ) (z − z2 ) Q2 (z) ,

where Q2 (z) is a polynomial of degree n − 2. Continuing in this way, we arrive at

expression (38). Some of the constants zk in (38) may appear more than once, and it
is clear that P (z) can have no more than n distinct zeros.

21