ComplexAnalysis Notes

Lecture Notes for
MATH29141 Complex Analysis
and
Part I of
MATH34011 Complex Analysis
and Applications
Lecturer: James Montaldi

Department of Mathematics
The University of Manchester
2023–24
ii
Contents
Introduction 1
Chapter 1 Complex functions, derivatives and integrals 3

1.1 Open sets and domains 3
1.2 Complex functions and continuity 6
1.3 Differentiable functions 8
1.3A What is the derivative? 10
1.3B The Cauchy-Riemann equations 10
1.4 A first look at integration 13
1.4A Paths and contours 14
1.4B Contour integration 17
1.4C Antiderivatives 19
1.5 Exercises 22
Chapter 2 Power series 25

2.1 Recap on (absolute) convergence of real series 25
2.2 Limits of complex sequences 28
2.3 On (absolute) convergence of complex series 29
2.4 Power series and their radius of convergence 31
2.5 Differentiation of power series 34
2.6 Elementary functions 39
2.6A The exponential function 39
2.6B Trigonometric functions 40
2.6C Hyperbolic functions 41
2.6D Periods of the exponential and trigonometric functions 42
2.6E The logarithm 42
2.7 Exercises 44
Chapter 3 Cauchy’s theorem & Cauchy’s formula 49

3.1 Winding numbers 49
3.2 Cauchy’s Theorem 54
3.3 The Estimation Lemma 58
3.4 Cauchy’s Integral Formula 61
3.5 Exercises 63
iii
iv
Chapter 4 Taylor series and Laurent series 67

4.1 Taylor series 67
4.2 Applications of the Cauchy-Taylor theorem 71
4.2A Cauchy’s Estimate 71
4.2B Radius of convergence 71
4.2C Liouville’s Theorem 73
4.2D Uniqueness of holomorphic functions 74
4.2E The Fundamental Theorem of Algebra 75
4.3 Laurent series 77
4.3A Calculating Laurent series 80
4.4 Exercises 83
Chapter 5 Cauchy’s Residue Theorem 85

5.1 Singularities 85
5.1A Removable singularities 86
5.1B Poles 87
5.1C Isolated essential singularities 87
5.2 Zeros and poles of meromorphic functions 88
5.3 Residues and Cauchy’s Residue Theorem 90
5.4 Calculating residues 92
5.5 Straightforward examples 96
5.6 Proof of Cauchy’s Residue Theorem 98
5.7 Exercises 100
Chapter 6 Applications of the residue theorem 103

6.1 Infinite real integrals 103
6.2 Trigonometric integrals 107
6.3 Summation of series 109
6.4 Exercises 113
2023–24
© University of Manchester
Introduction
Where we are going?
Complex analysis is the version of analysis for functions of a complex variable. It is natural
for a student to ask, ‘where do such functions come from?’ The first step is easy: we all know
what a polynomial is, for example f (x) = x 3 − x 2 + x − 1, for x ∈ R. And this cubic can be
factorized, f (x) = (x − 1)(x 2 + 1). The first factor corresponds to the (only) real root x = 1.
To make this a complex function is easy: we just allow x ∈ C. Conventionally, we call the
complex variable z instead of x, so here we would have f (z) = z 3 − z 2 + z − 1. Now with z ∈ C
this factorizes even further: f (z) = (z − 1)(z + i )(z − i ), showing f has 3 complex roots. This is
a particular case of a beautiful theorem: every polynomial of degree n has n complex roots
(counting multiplicity—it might for example have double roots: what this really means is that
such a polynomial can be factored into n linear factors).
What about functions which are not polynomial? Clearly rational functions are straight-
forward (a rational function is the quotient of two polynomials): for example
x2 − 1 z2 − 1
f (x) = (x ∈ R), becomes f (z) = (z ∈ C).
x 2 + 3x + 2 z 2 + 3z + 2
But what about more general functions, like ex or cos(x) or ln(x)? The approach here is to
express the function as a power series (Taylor series). For example
1 1 3 1 4
ex = 1 + x + x 2 + x + x +··· .
2 3! 4!
It would be nice to just replace x ∈ R again by z ∈ C. However, first we need to ensure that
the series converges to something when z ∈ C, at least for some values of z. When you know
that it does, then you can just replace x by z in the power series. Chapter 2 is dedicated to
addressing this question.
Before learning about power series, in Chapter 1 we discuss what complex functions are,
and how to differentiate and integrate them.
You are already familiar with how to differentiate and integrate real-valued functions de-
fined on the real line. For example, if f : R → R is defined by f (x) = 3x 2 + 2x then you know
1
2
that f ′ (x) = 6x +2 and that f (x) d x = x 3 + x 2 +c. You saw how to formally define differentia-
R
tion and integration for functions that map the reals to the reals in your Real Analysis course.
In this course, we will look at what it means for functions defined on the complex plane to
be differentiable or integrable and look at ways in which one can integrate complex-valued
functions. Surprisingly, the theory turns out to be considerably easier than the real-valued
case! Thus the word ‘complex’ in the title refers to the presence of complex numbers, and not
that the analysis is harder!
If f (z) is a complex function, then f ′ (z) is defined as a limit in the same way as for real
functions. The surprising difference between real and complex differentiability is that if a
complex function is differentiable once, then it is differentiable twice, three times, and so on,
but that is not an obvious fact and requires us to develop integration first. (This is in stark
contrast to the real case: for example f (x) = x|x| is only differentiable once at the origin.)
Perhaps the main reason complex analysis is so important in mathematics is the theory
of integration. Here we find a very rich theory with some remarkable results. These are dis-
cussed in Chapters 3, 4 and 5.
Many of the important applications of complex analysis are based on the Estimation
Lemma (Lemma 3.12), so watch out for that when it comes (in Sec. 3.3).
One of the highlights towards the end of Part 1 of the course is Cauchy’s Residue Theorem.
This theorem leads to a new method for calculating real integrals that would be difficult or
impossible just using techniques that you know from real analysis. For example, let 0 < a < b
and consider
x sin x
Z∞
2 2 2 2
d x. (∗)
−∞ (x + a )(x + b )
If you try calculating this using techniques that you know (integration by substitution, inte-
gration by parts, etc) then you will quickly hit an impasse. However, using complex analysis
and the theory of residues one can evaluate (∗) in about five lines of work!1 . You will see
examples like this in Part 2.
1 The answer is π (e −a − e −b ) if b 6= a as you’ll see in Part 2 of the course

b 2 −a 2
2023–24
Chapter 1
Complex functions, derivatives and

integrals
If we begin with a real polynomial function such as f (x) = x 3 + 2x, then to make it a complex
function is straightforward: we just replace x ∈ R by z ∈ C to get f (z) = z 3 + 2z. This is defined
for all z ∈ C. For example f (1 + i ) = (1 + i )3 + 2(1 + i ) = 4i .
One can also ‘complexify’ rational functions1 in the same way. For example g (x) = 1−x x
3
z
becomes g (z) = 1−z 3 . In this case the real rational function is defined for all x ∈ R except
x = 1, while the complex rational function is defined for all z ∈ C except the three cube roots
of unity (i.e., solutions of z 3 = 1).
In Chapter 2 we will ask about ‘complexifying’ more general functions such as e x or tan x,
but for this chapter we will discuss what complex functions are, and what derivatives and
integrals mean, using just polynomials and rational functions as examples.
1.1 Open sets and domains
In real analysis, one normally studies functions f : (a, b) → R, f : [a, b] → R or f : R → R

where (a, b) ⊂ R is an open interval or [a, b] ⊂ R is a closed interval. Usually, when talking
about continuity or differentiability in real analysis, one studies functions defined on open
intervals or on the entire real line. This is often because, when studying whether a function
f : (a, b) → R is continuous or differentiable at a point x 0 ∈ (a, b), we want to be able to look
at points x near to x 0 (either to the left or to the right of x 0 ) and then take the limit at x → x 0 .
We will need the appropriate generalisation to complex analysis of the notion of ‘defined on
an open interval’ that captures similar behaviour.
A major difference between real and complex analysis is that the geometry of the complex
plane is far richer than that of the real line. For example, the only connected subsets of R are
1 a rational function is the ratio of two polynomials
3
4 § 1.1 Open sets and domains
r
b
a
F IGURE 1.1: The disc D(a; r ) — the circle itself is broken (dashed) to empha-
size the fact that it’s not part of the disc.
intervals, whereas there are far more complicated connected subsets in C. We need to make
precise what we mean by convergence and differentiability, and this requires functions to be
defined on open sets (generalising open intervals) in C.
Notation Throughout, we use ⊂ (rather than ⊆) to denote ‘is a subset of’. Thus A ⊂ B means
that A is a subset of (and possibly equal to) B .
We will make a great deal of use of open discs in the complex plane so we introduce some
notation: the disc of radius r and centre a ∈ C is defined to be the set of points whose distance
from a is less than r (see Fig. 1.1):
D(a; r ) := {z ∈ C | |z − a| < r } . (1.1)
Definition 1.1. Let D ⊂ C. We say that D is an open set if for every a ∈ D there exists ε > 0
such that D(a; ε) ⊂ D. A set Z is closed if its complement C \ Z is open ✯
Warning: Note that a set is closed precisely when the complement is open. A very common
mistake is to think that ‘closed’ means ‘not open’: this is not the case, and it is easy to find
examples of sets that are neither open nor closed—even in R (can you think of any?). "
One can often decide whether a subset of C is open or not by looking at it and thinking
carefully. For example, any open disc D(z 0 ; r ) is an open set: see Figure 1.2.
We will also need the notion of a polygonal arc in C. Let z 0 , z 1 ∈ C. We denote the straight
line from z 0 to z 1 by [z 0 , z 1 ]. Now let z 0 , z 1 , . . . , z r ∈ C. We call the union of the straight lines
[z 0 , z 1 ], [z 1, z 2 ], . . . , [z r −1 , z r ] a polygonal arc joining z 0 to z r (see Fig. 1.3).
Open subsets of C may be very complicated. We will only be interested in ‘nice’ open sets
called domains.
Definition 1.2. A non-empty set D ⊂ C is said to be a domain if it is open and connected. A

set D is connected if given any two point z 1 , z 2 ∈ D, there exists a polygonal arc contained in
D that joins z 1 to z 2 . ✯
2023–24
C OMPLEX FUNCTIONS , DERIVATIVES AND INTEGRALS 5
r
b
a
b
z
F IGURE 1.2: This illustrates why the open disc D(a; r ) is an open set: given
any point z ∈ D, we can find another open disc centred at z that is contained
in D(a; r ).
w D
b
F IGURE 1.3: A polygonal arc joining z and w in a domain D.
Remark There are other definitions of connected; eg joining points by a smooth curve
rather than a polygonal one. But for open sets they are all equivalent. ❞
Examples 1.3.
(i). C itself is a domain;
(ii). Any open disc D(a; r ) is a domain.
(iii). If a ∈ C then C \{a} is a domain, and more generally if Z ⊂ C is a finite subset then C \ Z
is a domain.
(iv). An annulus {z ∈ C | r 1 < |z − z 0 | < r 2 } is a domain.
(v). A half-plane such as {z ∈ C | Re(z) > a} is a domain.
(vi). A closed disc {z ∈ C | |z − z 0 | ≤ r } or a closed half-plane {z ∈ C | Re(z) ≥ a} are not

domains as they are not open sets.
2023–24
6 § 1.2 Complex functions and continuity
(vii). The set D = {z ∈ C | Im(z) 6= 0}, corresponding to the complex plane with the real axis
deleted, is not a domain. Although it is an open set, there are points (such as i , −i )
that cannot be connected by a polygonal arc lying entirely in D.
See Figure 1.4 for other examples of domains.
D
D1 D2
(i) (ii)
F IGURE 1.4: In (i), D is a domain (with a hole in it). In (ii) D = D 1 ∪ D 2 is not a

domain as it is not connected; however each of D 1 and D 2 are domains.
Note: It will be sufficient in this course to draw a picture and describe informally why a
set is a domain or not.
1.2 Complex functions and continuity
We will only consider functions defined on domains.

Let D ⊂ C be a non-empty domain. A function f : D → C is a rule that assigns to each
point z ∈ D an image f (z) ∈ C.
Write z = x +i y. Then saying that f is a function is equivalent to saying that there are two
real-valued functions u(x, y) and v(x, y) of the two real variables x, y such that
f (z) = u(x, y) + i v(x, y).
Examples 1.4. (i). Let f (z) = z 2 . Then f (x+i y) = (x+i y)2 = x 2 −y 2 +2i x y. Here u(x, y) =
2 2
x − y , v(x, y) = 2x y.
(ii). Let f (z) = z̄. Then f (x + i y) = x − i y. Here u(x, y) = x, v(x, y) = −y.

2023–24
1
Both of these examples have domain D = C. On the other hand f (z) = 1−z
is defined for
all z 6= 1 so D = C \ {1}.
Let D be a domain and let f : D → C. Let z 0 ∈ D. We say that limz→z0 f (z) = ℓ (or, equiva-
lently, f (z) tends to ℓ as z tends to z 0 ) if, for all ε > 0, there exists δ > 0 such that if z ∈ D and
0 < |z − z 0 | < δ then | f (z) − ℓ| < ε.
That is, f (z) → ℓ as z → z 0 means that if z is very close (but not equal to) z 0 then f (z)
is very close to ℓ. Note that in this definition we do not need to know the value of f (z 0 ).
Example 1.5. Let f : C → C be defined by f (z) = 1 if z 6= 0 and f (0) = 0. Then limz→0 f (z) =
1. Here limz→0 f (z) 6= f (0) so f is not continuous.
We will be interested in functions which do behave nicely when taking limits.
Definition 1.6. Let D be a domain and let f : D → C be a function. We say that f is continu-
ous at z 0 ∈ D if
lim f (z) = f (z 0 ).
z→z 0
We say that f is continuous on D if it is continuous at z 0 for all z 0 ∈ D. ✯
It is not hard to show (and follows from discussions in Chapter 2 on convergence) that if
f (z) → ℓ and ℓ = a + i b then u(x, y) → a and v(x, y) → b.
Continuity obeys the same rules as in Real Analysis. In particular, suppose that f , g : D →
C are complex functions which are continuous at z 0 . Then
f (z) + g (z), f (z)g (z), c f (z) (c ∈ C)
are all continuous at z 0 , as is f (z)/g (z) provided that g (z 0 ) 6= 0.

Notation For future reference, we list some special domains which occur in the
course, with their notation:
• C, the entire complex plane
• C∗ = C \ {0}, often called the punctured plane
• C− = C \ {x ∈ R | x ≤ 0}, often called the cut plane (see Fig. 2.1 on p.43)
• D(z 0 ; R) the disc of radius R and centre z 0
• D ∗ (z 0 ; R) the punctured disc of radius R and centre z 0 , equal to D(z 0 , ; R) \ {z 0 }

(a disc without its centre),
• A(z 0 ; R 1 , R 2 ) = {z ∈ C | R 1 < |z − z 0 | < R 2 }, the annulus with centre z 0 and radii

R 1 and R 2 (R 1 is the inner radius, R 2 the outer radius) — see Fig 4.2 on p.78.
Notice that A(z 0 ; 0, R) = D ∗ (z 0 , R).
2023–24
8 § 1.3 Differentiable functions
1.3 Differentiable functions
You will recall how derivatives are defined for real functions (from Real Analysis). We make
the same definition for complex functions:
Definition 1.7. Let D ⊂ C be a domain and let f : D → C be a function. Let z 0 ∈ D. We say

that f is differentiable at z 0 if
f (z) − f (z 0 )
f ′ (z 0 ) = lim (1.2)
z→z 0 z − z0
exists. We call f ′ (z 0 ) the derivative of f at z 0 . If f is differentiable at every point z 0 ∈ D then
we say that f is differentiable on D. ✯
One can also use the notation

df
(z 0 )
dz
to denote the derivative of f at z 0 .
Remark In the definition of differentiability for functions of a real variable x, the limit is as
x → x 0 , and there we have just two possibilities, x → x 0+ (limit from the right) and x → x 0−
(limit from the left). For functions of a complex variable one must allow z to converge to z 0
from any direction, and the limiting complex number must always be the same. That is one
reason that differentiability for a complex function is so special (as we’ll see). ❞
Remark Although the definition of differentiability of a function in complex analysis is,

essentially, the same as the definition in real analysis, we lose many of the geometrical inter-
pretations of the derivative. For example, one cannot easily interpret f ′ (z 0 ) as the gradient
or slope of f at z 0 . As another example, in real analysis one can normally interpret points x 0
for which f ′ (x 0 ) = 0 as turning points or local maxima/minima of f . The notion of a local
maximum or local minimum does not exist in complex analysis; this is because there is no
natural ordering on the set of complex numbers. ❞
Differentiability is a very strong property for a complex function to possess; it is much

stronger than the real case. For example, there are many example of functions where u(x, y)
and v(x, y) are both differentiable but f (z) is not. For this reason, we shall often use the
following alternative terminology.
Definition 1.8. Suppose that f : D → C is differentiable on a domain D. Then we say that f

is holomorphic on D. We write O (D) for the set of all holomorphic functions on D. ✯
The higher derivatives are defined similarly, each the derivative of the one before, and we
denote them by
f ′′ (z 0 ), f ′′′ (z 0 ), . . . , f (n) (z 0 ).
2023–24
Example 1.9. Let f (z) = z 2 , defined on C. Let z 0 ∈ C be any point. Then
f (z) − f (z 0 ) z 2 − z 02 (z + z 0 )(z − z 0 )
lim = lim = lim = lim z + z 0 = 2z 0 .
z→z 0 z − z0 z→z 0 z − z0 z→z 0 z − z0 z→z 0
Hence f ′ (z 0 ) = 2z 0 for all z 0 ∈ C. Thus f is differentiable at every point in C and so is a

holomorphic function on C (or f ∈ O (C)).
All of the standard rules of differentiable functions continue to hold in the complex case:
Proposition 1.10. Let f , g be holomorphic on a domain D. Let c ∈ C. Then the following

functions are also holomorphic on D, with derivatives given by the familar rules,
(i) sum rule: ( f + g )′ = f ′ + g ′ ,
(ii) scalar rule: (c f )′ = c f ′ ,
(iii) product rule: ( f g )′ = f ′ g + f g ′ ,

³ ´′
f f ′g − f g ′
(iv) quotient rule: g = g 2 (at points where g 6= 0),
(v) chain rule: ( f ◦ g )′ = f ′ ◦ g · g ′ .
Proof: The proofs are all very simple modifications of the corresponding arguments in
the real variable case. (Usually the only modification needed is to replace the absolute
value | · | defined on R with the modulus | · | defined on C.) ❒
Parts (i) and (ii) of the proposition imply that for any domain D, the set O (D) of homo-
morphic functions on D is a (complex) vector space.
We will need the following useful fact.
Proposition 1.11. Suppose that f is differentiable at z 0 . Then f is continuous at z 0 .
Proof: ¡To show that ¢

f is continuous, we need to show that limz→z0 f (z) = f (z 0 ), i.e.
limz→z0 f (z) − f (z 0 ) = 0. Note that
f (z) − f (z 0 )
lim f (z) − f (z 0 ) = lim (z − z 0 ) = f ′ (z 0 ) × 0 = 0,
z→z 0 z→z 0 z − z0
as required. ❒
2023–24
In the exercises, we show that all polynomials are differentiable (by the same argument
as in the real setting). Consequently (by Proposition 1.10(iv)) so are all rational functions on
1
their domain of definition (eg 1−z is not defined at z = 1, so its natural domain is C \ {1}). It
then follows from the proposition above that all polynomials are continuous on C, and all
rational functions are continuous on their domain of definition.
1.3A What is the derivative?
For functions of a real variable, the derivative represents the gradient of the function (or of
the graph) at a point. This is made precise by writing
f (x + h) − f (x)
f ′ (x) = lim
h→0 h
as
f (x + h) − f (x) = f ′ (x)h + o(h)
where o(h) represents a function, say ε(h), satisfying limh→0 ε(h)/h = 0. (A.k.a. first order
Taylor series.) The interpretation here is that if we change x by a small amount h then f (x)
changes by (approximately) f ′ (x)h — in other words you multiply the change by f ′ (x). That’s
what we call the gradient.
It is similar in the complex case:
f (z + h) − f (z) = f ′ (z)h + o(h).
(where now h ∈ C). So if z is changed by a small amount h then f (z) changes by (approxi-
mately) f ′ (z)h; again h is multiplied by the derivative f ′ (z). So the derivative represents an
amplitwist (see complex numbers section on blackboard) rather than a gradient.
1.3B The Cauchy-Riemann equations
Throughout, let D be a domain. Let z = x + i y ∈ D. Let f : D → C be a complex valued

function. We write f as the sum of its real part and imaginary part by setting
f (z) = u(x, y) + i v(x, y)
where u, v : D → R are real-valued functions.
Example 1.12. Let f (z) = z 3 . Then
f (z) = z 3 = (x + i y)3 = x 3 − 3x y 2 + i (3x 2 y − y 3 ) = u(x, y) + i v(x, y)
where u(x, y) = x 3 − 3x y 2 and v(x, y) = 3x 2 y − y 3 .

2023–24
Recall definition of partial derivatives:

Suppose that g (x, y) is a real-valued function depending on two real variables x, y.
Define
∂g g (x + h, y) − g (x, y) ∂g g (x, y + k) − g (x, y)
(x, y) = lim , (x, y) = lim
∂x h→0 h ∂y k→0 k
(if these limits exist). For brevity (and provided there is no confusion), we often leave
out the (x, y) and write
∂g ∂g
, , or g x , g y .
∂x ∂y
Thus, to calculate g x = ∂g /∂x we treat y as a constant and differentiate with respect
to x, and to calculate g y we treat x as a constant and differentiate with respect to y.
If f is differentiable, then the Cauchy-Riemann equations give two relationships between

the partial derivatives of u and v.
Definition 1.13. Let u(x, y) and v(x, y) be two differentiable functions defined on a domain
in R2 . The Cauchy-Riemann equations are the two equations,
∂u ∂v

 ∂x = ,


∂y

(1.3)
 ∂u ∂v
= − .



∂y ∂x
They are often known as the C-R equations. ✯
Notice that one of these has a minus sign—this comes from the fact that i −1 = −i . We will
often write u x for ∂u/∂x etc., thus the C-R equations are
ux = v y , and v x = −u y .
Theorem 1.14 (The Cauchy-Riemann Theorem). Suppose u(x, y) and v(x, y) are real-
valued functions defined on a domain D with continuous first-order partial derivatives,
and let f (z) = u(x, y) + i v(x, y) where z = x + i y.
Then f is holomorphic on D if and only if u and v satisfy the Cauchy-Riemann equa-
tions on D.
Although mostly known to d’Alembert and Euler before Cauchy, this central theorem goes
by the name of the Cauchy-Riemann theorem, because they made important use of it, Cauchy
for the analysis of complex functions and Riemann for their geometry.
Below we prove that if f is holomorphic then u, v satisfy the Cauchy-Riemann equations,
but first we see some examples. A proof of the converse can be found for example in Sec. 4.2
of Stewart and Tall.
2023–24
Examples 1.15. (i). Let us check these C-R equations for Example 1.12 above. We have
f (z) = z 3 (which we know is holomorphic on C with derivative 3z 2 ), and u(x, y) =
x 3 − 3x y 2 and v(x, y) = 3x 2 y − y 3 . Then
u x = 3x 2 − 3y 2 = v y ,
and
u y = −6x y = −v x ,
as required.
(ii). We can use the Cauchy-Riemann equations to examine whether the function f (z) = z̄
might be differentiable on C. Writing z = x + i y allows us to write f (z) = z̄ = x − i y.
Hence f (z) = u(x, y) + i v(x, y) with u(x, y) = x and v(x, y) = −y. Now
u x = 1, u y = 0, v x = 0, v y = −1.
Hence there are no points at which u x = v y so that f (z) = z̄ is not differentiable on any
domain in C.
Remark Notice however that f (z) = z̄ is continuous at every point in C. Hence f (z) = z̄ is an
example of an everywhere continuous but nowhere differentiable function. Such functions
also exist in real analysis, but they are much harder to write down and considerably harder
to study (one of the simplest is known as Weierstrass’ function w (x) = ∞ n=0 2
−nα
cos 2πb n x
P
where α ∈ (0, 1), b ≥ 2; such functions are still of interest in current research). ❞
Proof: Suppose that f is holomorphic. Recall from (1.2) that to calculate f ′ (z 0 ) we look
at points that are close to z 0 and then let these points tend to z. The trick is to calculate
f ′ (z 0 ) in two different ways: by looking at points that converge to z 0 horizontally, and by
looking at points that converge to z 0 vertically: the derivative must be the same indepen-
dently of how the limit z → z 0 is taken, as pointed out in an earlier remark.
Let h be real and consider z 0 + h = (x 0 + h) + i y 0. Then as h → 0 we have z 0 + h → z 0 .

Hence
f (z 0 + h) − f (z 0 )
f ′ (z 0 ) = lim
h→0 h
u(x 0 + h, y 0 ) + i v(x 0 + h, y 0 ) − u(x 0 , y 0 ) − i v(x 0 , y 0 )
= lim
h→0 h
u(x 0 + h, y 0 ) − u(x 0 , y 0 ) v(x 0 + h, y 0 ) − v(x 0 , y 0 )
= lim +i
h→0 h h
∂u ∂v
= (x 0 , y 0 ) + i (x 0 , y 0 ). (1.4)
∂x ∂x
2023–24
Now take k to be real and consider z 0 +i k = x 0 +i (y 0 +k). Then as k → 0 we have z 0 +i k →

z 0 . Hence
f (z 0 + i k) − f (z 0 )
f ′ (z 0 ) = lim
k→0 ik
u(x 0 , y 0 + k) + i v(x 0, y 0 + k) − u(x 0 , y 0 ) − i v(x 0 , y 0 )
= lim
k→0 ik
u(x 0 , y 0 + k) − u(x 0 , y 0 ) v(x 0 , y 0 + k) − v(x 0 , y 0 )
= lim +i
k→0 ik ik
∂u ∂v
= −i (x 0 , y 0 ) + (x 0 , y 0 ), (1.5)
∂y ∂y
recalling that 1/i = −i . Comparing the real and imaginary parts of (1.4) and (1.5) gives the
result.
For the converse, the interested student can consult any text on Complex Analysis (for
example Theorem 4.12 of Stewart and Tall). ❒
Remark The statement of the theorem requires f to be holomorphic in a domain D. How-

ever, the proof shows that if f ′ (z) exists just at z = z 0 then the C-R equations are also satisfied
at that point. It is the converse (which we haven’t proved) which requires that u and v be
differentiable with continuous derivatives on an open set. ❞
1.4 A first look at integration
The theory of integration is the heart of complex analysis, and is behind all (or certainly most)
of its many applications. Here we see a first glimpse of this theory, and we will return to it in
greater depth in later chapters.
Consider the real integral
Zb
f (x) dx.
a
We often read this as ‘the integral of f from a to b’. That is, we think of starting at the point a
and moving along the real axis to b, integrating f as we go.
Now let z 0 , z 1 ∈ C. How might we define
Zz 1
f (z) dz ?
z0
We want to start at z 0 , move through the complex plane to z 1 , integrating f as we go. But in
the complex plane there are lots of ways of moving from z 0 to z 1 . Suppose γ is a path from
2023–24
14 § 1.4 A first look at integration
z 0 to z 1 (defined below to be a continuous curve starting at z 0 and ending at z 1 ). Then, using

similar ideas to those from Vector Calculus, we can define
Z
f (z) dz.
γ
A priori this looks like it will depend on the path γ. However, as we shall see, in complex anal-
ysis in many cases this quantity is independent of the path chosen (analogous to integrating
conservative vector fields in vector calculus).
1.4A Paths and contours
First we need to make precise what we mean by a path.
Definition 1.16. A path is a continuous function γ : [a, b] → C where [a, b] is a real interval.
✯
So, for each t ∈ [a, b], γ(t ) is a point on the path. We say that the path γ starts at γ(a) and
ends at γ(b).
Remark Note that a path is a function. Sometimes, it is convenient to regard a path as a

set of points (a curve) in C, i.e. we identify the function γ with its image. However, we should
regard this set of points as having an orientation: a path starts at one end-point and ends at
the other. If we think of the path γ in this way then we sometimes call the function γ(t ) a
parametrization of the path γ. Note that the same path can have different parametrizations.
For example
γ1 (t ) = t + i t , γ2 (t ) = t 2 + i t 2 , 0 ≤ t ≤ 1
are both parametrizations of the straight line that starts at 0 and ends at 1 + i . We shall see
later (Proposition 1.27) that when we calculate an integral along a path then it is independent
of the choice of parametrization. ❞
As an example of a path, let z 0 , z 1 ∈ C. Define
γ(t ) = (1 − t )z 0 + t z 1 , 0 ≤ t ≤ 1. (1.6)
Then γ(0) = z 0 , γ(1) = z 1 and the image of γ is the straight line joining z 0 to z 1 . We sometimes
denote this path by [z 0 , z 1 ]. See Figure 1.5.
Definition 1.17. Let γ : [a, b] → C be a path. If γ(a) = γ(b) (i.e. if γ starts and ends at the
same point) then we say that γ is a closed path or a closed loop. ✯
2023–24
z1
γ
z0
F IGURE 1.5: The path γ(t ) = (1−t )z 0 +t z 1 , 0 ≤ t ≤ 1, describes the straight line
joining z 0 to z 1 , sometimes denoted [z 0 , z 1 ].
F IGURE 1.6: The circular path γ(t ) = cos(t ) + i sin(t ), 0 ≤ t ≤ 2π. It starts at 1
and travels anticlockwise around the unit circle, ending up back at 1.
Example 1.18. An important example of a closed path is given by
γ(t ) = cos t + i sin t , 0 ≤ t ≤ 2π. (1.7)
[After Chapter 2, we will write this as γ(t ) = e i t .] This is the path that describes the circle in
C with centre 0 and radius 1, starting and ending at the point 1, travelling around the circle
in an anticlockwise direction (we say the path is oriented anticlockwise). See Figure 1.6.
If instead we want to parametrize the circle with centre z 0 and radius r we use
γ(t ) = z 0 + r (cos(t ) + i sin(t ), 0 ≤ t ≤ 2π. (1.8)
Definition 1.19. A path γ is said to be smooth if γ : [a, b] → C is differentiable and γ′ is con-

tinuous. (By differentiable at the end-points a and b we mean that the one-sided derivative
exists at each.) ✯
2023–24
All of the examples of paths above are smooth.

We can use integrals to define the length of a smooth path (similar to the length of a curve
in the plane R2 ):
Definition 1.20. Let γ : [a, b] → C be a smooth path. Then the length of γ is defined to be
Zb
length(γ) = |γ′ (t )| dt .
a
Example 1.21. It is straightforward to check from (1.6) that length([z 0 , z 1 ]) = |z 1 − z 0 |, as it

should be. If γ(t ) is the path given in (1.7) then
length(γ) = 2π,
and the length of the path in (1.8) is 2πr as you should check.
Suppose that γ : [a, b] → C is a path that starts at γ(a) and ends at γ(b). Then we can
consider the reverse of this path, where we start at γ(b) and, travelling backwards along γ,
end at γ(a). More formally, we make the following definition.
Definition 1.22. Let γ : [a, b] → C be a path. Define the reverse path γ− : [a, b] → C to be the
path
γ− (t ) = γ(a + b − t ).
✯
Note that the formula implies γ− (a) = γ(b) and γ− (b) = γ(a), as you should check.
Often we will want to integrate over a number of paths joined together. One could make
the latter a path by constructing a suitable reparametrization, but in practice this makes
things complicated; in particular the joins may not be smooth. It is simpler to give a name to
several smooth paths joined together.
Definition 1.23. A contour γ is a collection of smooth paths γ1 , . . . , γn where the end-point

of γr coincides with the start point of γr +1 , 1 ≤ r ≤ n − 1. We write
γ = γ1 + · · · + γn .
If the end-point of γn coincides with the start point of γ1 then we call γ a closed contour.
Further, the length of the contour γ (above) is defined to be
length(γ) = length(γ1 ) + · · · + length(γn ).
That is, it’s the sum of the lengths of the individual components. ✯
2023–24
γ2
γ4
−R γ3 −ε ε γ1 R
F IGURE 1.7: The contour γ1 + γ2 + γ3 + γ4 .
Thus a contour is a path that is smooth except at finitely many places: so it looks like a
smooth path with finitely many corners. The polygonal arcs introduced above are all con-
tours.
Example 1.24. Let 0 < ε < R. Define
γ1 : [ε, R] → C γ1 (t ) = t ,
γ2 : [0, π] → C γ2 (t ) = Re i t ,
γ3 : [−R, −ε] → C γ3 (t ) = t ,
γ4 : [−π, 0] → C γ4 (t ) = εe −i t .
(where we use e i t as an abbreviation for cos(t ) + i sin(t )!). Then γ = γ1 + γ2 + γ3 + γ4 is a

closed contour, shown in Figure 1.7. (Later in the course there will be many integrals over
contours similar to this one.)
1.4B Contour integration
Let f : D → C be a complex functions defined on a domain D. Let γ : [a, b] → D be a smooth

path in D.
Definition 1.25. The integral of f along γ is defined to be

Z Zb
f (z) dz := f (γ(t )) γ′ (t ) dt . (1.9)
γ a
R R
We will often write γ f for γ f (z) dz. ✯
2023–24
Remark RIn practice we write f (γ(t )) γ′ (t ) = u(t ) + i v(t ) where u, v : [a, b] → R and define
R b Rb
γ f to be a u(t ) dt + i a v(t ) dt and so find the real and imaginary parts of the integral. ❞
Example 1.26. Take f (z) = z 2 and γ(t ) = t 2 +i t , 0 ≤ t ≤ 1 (part of a parabola in the complex
plane going from 0 to 1+i ). Then f (γ(t )) = (t 2 +i t )2 = t 4 − t 2 +2i t 3 and γ′ (t ) = 2t +i . Hence
Z Z1 Z1
f (z) dz = ′
f (γ(t ))γ (t ) dt = (t 4 − t 2 + 2i t 3 )(2t + i ) dt
γ 0 0
Z1 Z1
= 2t 5 − 4t 3 dt + i 5t 4 − t 2 dt
0 0
1 6 4 1 1 3 1
· ¸ · ¸
5
= t −t +i t − t
3 0 3 0
2 2
= − +i .
3 3
The following proposition shows that the definition (1.9) is independent of the choice of
parametrization of the path.
Proposition 1.27. Let γ : [a, b] → C be a smooth path. Let φ : [c, d ] → [a, b] be an in-
creasing smooth bijection. Then γ ◦ φ : [c, d ] → C is a path that has the same image as γ.
Moreover, Z Z
f = f
γ◦φ γ
for any continuous function f .
Proof: It is clear that both γ and γ◦φ have the same image. Thus γ and γ◦φ are different
parametrizations of the same path. Note that
Z Zd
f = f (γ(φ(t )))(γ ◦ φ)′ (t ) dt
γ◦φ c
Zd
= f (γ(φ(t )))γ′ (φ(t ))φ′ (t ) dt (by the chain rule)
c
Zb
= f (γ(t ))γ′ (t ) dt (by the change of variables formula).
a
Remark If φ in Proposition 1.27 is a decreasing smooth bijection then γ ◦ φ has the same
image as γ but the path traverses this in the opposite direction, i.e. γ ◦ φ is a parametrization
2023–24
of γ− . Following the above calculation we see that

R R
γ◦φ f =− γ f , corresponding to the fact
R R
stated in the proposition below that γ− f = − γ f . ❞
Now suppose that γ = γ1 + · · · + γn is a contour in D. We define

Z Z Z
f = f +··· + f.
γ γ1 γn
The following basic properties of contour integration follow easily from this definition.
Proposition 1.28. Let f , g : D → C be continuous and let c ∈ C. Suppose that γ, γ1 , γ2 are

contours in D. Suppose that the end point of γ1 is the start point of γ2 . Then
(i) Z Z Z
f = f+ f;
γ1 +γ2 γ1 γ2
(ii) Z Z Z
(f + g) = f+ g;
γ γ γ
(iii) Z Z
cf = c f;
γ γ
(iv) Z Z
f = − f.
γ− γ
1.4C Antiderivatives
Recall from real calculus that one way to calculate the integral of f is to find an antideriva-
tive2 , i.e. find a function F such that F ′ = f . The Fundamental Theorem of Calculus then says
Rb
that a f (x) dx = F (b)−F (a). We have an analogue of this in for the complex integral. We first
need the following definition.
Definition 1.29. Let D be a domain and f : D → C be a continuous function. We say that a

function F : D → C is an antiderivative of f on D if, throughout D, F ′ = f . ✯
Theorem 1.30 (The Fundamental Theorem of Contour Integration). Suppose that f :

D → C is continuous and F : D → C is an antiderivative of f on D. Let γ be a contour from
2 an antiderivative is called a primitive in some texts

2023–24
z 0 to z 1 . Then Z
f = F (z 1 ) − F (z 0 ). (1.10)
γ
Proof: This follows from the usual (real) fundamental theorem of calculus, as follows.
Suppose first that γ : [a, b] → D, γ(a) = z 0 , γ(b) = z 1 , is a single smooth path.
Let w (t ) = f (γ(t ))γ′ (t ) and let W (t ) = F (γ(t )). Then by the chain rule
W ′ (t ) = F ′ (γ(t ))γ′ (t ) = f (γ(t ))γ′ (t ) = w (t ).
Write w (t ) = u(t ) + i v(t ) and W (t ) = U (t ) + i V (t ) so that U ′ = u, V ′ = v. Hence

Z Zb
f = f (γ(t ))γ′ (t ) dt
γ a
Zb
= w (t ) dt
a
Zb Zb
= u(t ) dt + i v(t ) dt
a a
= U (t )|ba + i V (t )|ba (by the Fund. Thm of Calculus)
= W (t )|ba
= F (z 1 ) − F (z 0 ).
Now, if γ is not itself smooth, then it is a collection of smooth paths (Definition 1.23). One
can then apply the above argument to each smooth part, ending with the same result. ❒
Notice that (1.10) does not depend on the choice of path γ from z 0 to z 1 provided the path
remains within the domain D.
Example 1.31. Let f (z) = z 2 and let γ be any contour from z 0 = 0 to z 1 = 1 + i . Then F (z) =
z 3 /3 is an antiderivative for f (valid everywhere in C) and
1 3 1 3 (1 + i )3 2 2
Z
z 2 dz = z1 − z0 = = − + i.
γ 3 3 3 3 3
(Compare with Example 1.26.)
In this example, how did we find the antiderivative? Essentially it’s by guesswork and
experience. Given f (z) = z 2 , what function do we know who’s derivative is z 2 ? Well, 13 z 3 is
one; another is 13 z 3 +1, etc. Since we know how to differentiate any polynomial (Exercise 1.3),
we can find an antiderivative of any polynomial, in the same way as for real polynomials.
The following useful fact is well-known to you in the real case: a function with zero deriva-
tive must be constant.
2023–24
Corollary 1.32. Suppose that F is holomorphic on a domain D and F ′ (z) = 0 for all
z ∈ D. Then F is constant on D.
Proof: Fix z 0 ∈ D and let z 1 ∈ D be an arbitrary point. Let γ be any contour from z 0 to z 1
contained in D. The theorem tells us that
Z Z
F (z 1 ) − F (z 0 ) = F ′ (z) dz = 0 dz = 0.
γ γ
Thus F (z 1 ) = F (z 0 ), but this holds for every z 1 ∈ D and hence F is constant. ❒
Remark If γ is a closed path (i.e. γ starts and

H
ends at the same point) and f has an an-
tiderivative on a domain that contains γ then γ f = 0. However, possessing an antiderivative
is a very strong hypothesis on f and D (see the following remarks). ❞
Example 1.33. In real analysis, any sufficiently nice function f has an antiderivative on
any interval: we define Zx
F (x) = f (t ) dt .
0
then F ′ = f . In complex analysis, however, the existence of an antiderivative on a domain

D is a strong hypothesis, because domains can be much more complicated than intervals.
Consider for example f (z) = 1/z defined on D = C \ {0}. Does this function have an an-
tiderivative on D? Let us answer this by doing an integral. Let f (z) = 1/z and γ(t ) = e i t ,
0 ≤ t ≤ 2π denote the unit circle in C oriented anticlockwise, and note that γ is a closed
H H
path. If f has an antiderivative on D then γ f = 0. To evaluate γ f we need to use the
definition of the contour integral given in (1.9). We have
Z2π Z2π Z2π
1
Z
f = f (γ(t ))γ′ (t ) dt = it
i e i t dt = i dt = 2πi .
γ 0 0 e 0
Z2π Z2π Z2π

1
Z
′
f = f (γ(t ))γ (t ) dt = i E (t )dt = i dt = 2πi .
γ 0 0 E (t ) 0
This is not zero, and thus f does not have an antiderivative on D. [It’s tempting to think log z
will be an antiderivative of z1 , but we haven’t defined log for complex numbers, and when
we do (in the next chapter) you will see it is a subtle function.]
2023–24
22 § 1.5 Exercises
1.5 Exercises
Ex 1.1 Which of the following sets are domains? (Recall that it is sufficient to draw a picture
and describe informally why a set is a domain or not.)
(i) {z ∈ C | Im(z) > 0}, (ii) {z ∈ C | Re(z) > 0, |z| < 2},
(iii) {z ∈ C | |z| ≤ 6}, (iv) {z ∈ C | |z| < 2} ∪ {z ∈ C | |z| > 4},
(v) {z ∈ C | |z| < 2} \ {z ∈ C | |z| ≤ 1}.
Ex 1.2 Using the definition in (1.2), differentiate the following complex functions from first
principles:
(i) f (z) = z 3 − z 2 ; (ii) f (z) = 1/z (z 6= 0).
Ex 1.3 Let p(z) = z n (for n ∈ N). Show from the definition of derivative that p ′ (z) = nz n−1 .
Deduce that all polynomials are differentiable.
Ex 1.4 Consider the two complex functions,
1
(a) f (z) = z 2 ; (b) f (z) = (z 6= 0).
z
(i) For each of the functions, write f (z) in the form u(x, y) + i v(x, y) where z = x + i y
and u, v are real-valued functions.
(ii) Show that u and v satisfy the Cauchy-Riemann equations everywhere for (a), and for
all z 6= 0 in (b).
(iii) Write the function f (z) = |z| in the form u(x, y)+i v(x, y). Using the Cauchy-Riemann
equations, decide whether there are any points in C at which f is differentiable.
Ex 1.5 (i) Show that the Cauchy-Riemann equations hold for the functions u, v given by
u(x, y) = x 3 − 3x y 2 , v(x, y) = 3x 2 y − y 3 . Show that u, v are the real and imaginary
parts of a holomorphic function f : C → C.
(ii) Show that the Cauchy-Riemann equations hold for the functions u, v given by
x 4 − 6x 2 y 2 + y 4 4x y 3 − 4x 3 y
u(x, y) = , v(x, y) =
(x 2 + y 2 )4 (x 2 + y 2 )4
where (x, y) 6= (0, 0). Deduce that u, v are the real and imaginary parts of a holomor-
phic function f : C \ {0} → C.
Ex 1.6 Let f (z) = |z|2 .
(i) Show from the definition (1.2) that f ′ (0) = 0.

(ii) Use the Cauchy-Riemann theorem to show that f is not differentiable in any domain.
2023–24
Ex 1.7 Suppose that f (z) = u(x, y)+i v(x, y) is holomorphic. Use the Cauchy-Riemann equa-
tions to show that both u and v satisfy Laplace’s equation:
∂2 u ∂2 u ∂2 v ∂2 v
+ = 0, + =0
∂x 2 ∂y 2 ∂x 2 ∂y 2
(you may assume that the second partial derivatives exist and are continuous). Note:
functions which satisfy Laplace’s equation are called harmonic functions.
Ex 1.8 Let f (z) = z 3 , f : C → C. Determine real-valued functions u, v so that f (z) = u(x, y) +

i v(x, y) (where z = x + i y). Verify that both u and v are harmonic functions.
We see below that the Cauchy Riemann equations have the following remarkable implica-
tion: Suppose f (z) = u(x, y) + i v(x, y) is holomorphic and that we know a formula for u, then
we can recover v (up to a constant); similarly, if we know v then we can recover u (up to a
constant). Hence for complex differentiable functions, the real part of a function determines
the imaginary part (up to constants), and vice versa.
Ex 1.9 Suppose f (z) = u(x, y)+i v(x, y) is holomorphic on C. Suppose we know that u(x, y) =
x 5 − 10x 3 y 2 + 5x y 4 . By using the Cauchy-Riemann equations, find all the possible forms
of v(x, y).
Ex 1.10 Suppose that

u(x, y) = x 3 − k x y 2 + 12x y − 12x
for some constant k ∈ C. Find all values of k for which u is the real part of a holomorphic
function f : C → C.
Ex 1.11 Show that if f : C → C is holomorphic and f has a constant real part then f is con-
stant.
Ex 1.12 Show that the only holomorphic function f : C → C of the form f (x + i y) = u(x) +
i v(y) is given by f (z) = λz + c for some λ ∈ R and c ∈ C.
Ex 1.13 Suppose that f (z) = u(x, y) + i v(x, y), f : C → C, is a holomorphic function and that
2u(x, y) + v(x, y) = 5 for all z = x + i y ∈ C.
Show that f is constant.
Ex 1.14 Let f : C → C be a holomorphic function. Define g : C → C by,
g (z) = f (z).
(i). For f (z) = z 2 − i z find g (z).

(ii). For arbitrary holomorphic f show that g is also holomorphic. [Hint: Write f = u + i v
and g = U + i V and express U ,V in terms of u, v .]
2023–24
24 Exercises
And now for some integration . . .
Ex 1.15 Draw the following paths:

(i) γ(t ) = e −i t , 0 ≤ t ≤ π
(ii) γ(t ) = 1 + i + 2e i t , 0 ≤ t ≤ 2π,
(iii) γ(t ) = t + i cosh t , −1 ≤ t ≤ 1,
(iv) γ(t ) = cosh t + i sinh t , −1 ≤ t ≤ 1.
Ex 1.16 Find the values of Z

x − y + i x 2 dz
γ
where z = x + i y and γ is:
(i) the straight line joining 0 to 1 + i ;

(ii) the imaginary axis from 0 to i ;
(iii) the line parallel to the real axis from i to 1 + i .
Does the integrand satisfy the Cauchy-Riemann equations?
Ex 1.17 Let
γ1 (t ) = 2 + 2e i t , 0 ≤ t ≤ 2π,
γ2 (t ) = i + e −i t , 0 ≤ t ≤ π/2.
Draw the paths γ1 , γ2 .

Rb
From the definition γ f = a f (γ(t ))γ′ (t ) dt , calculate
R
dz dz
Z Z
(i) , (ii) .
γ1 z −2 γ2 (z − i )3
Ex 1.18 Let γ denote the circular path with centre 1 and radius 1, described once, oriented
anticlockwise and starting at the point 2. Let f (z) = |z|2 . Write down a parametrization
of γ. Hence calculate γ |z|2 dz. Why does this tell you the function f (z) = |z|2 has no
R
antiderivative?
Ex 1.19 Let γ denote the semi-circular path in the upper half-plane along a circle with centre
0 and radius 3 which starts at 3 and ends at −3. Let f (z) = 1/z 2 defined on the domain
D = C \ {0}.
R
(i) Write down a parametrization of γ and calculate γ f .
(ii) Write down a parametrization of γ− . Calculate γ− f from the definition of the contour
R
R R
integral and check that, in this case, γ− f = − γ f .
Ex 1.20 Show that the function Re : C → R defined by taking the real part of a complex num-
ber, is continuous. Do the same for Im : C → R.
2023–24
Chapter 2
Power series and elementary analytic

functions
In Chapter 1 we studied some properties of complex functions, armed only with polynomials
and rational functions as examples, obtained by complexifying corresponding real functions.
The question we want to address now is, can we ‘complexify’ other functions such as e x
or sin(x)? The answer is ‘yes’, using the expression of these functions as power series. For
example
e x = 1 + x + 12 x 2 + 3!
1 3
x +···
becomes
e z = 1 + z + 21 z 2 + 3!1 z 3 + · · ·
In order to do this we need to discuss when such power series converge (ie, for which z ∈ C
is this expression meaningful), and for this we need some properties of convergent series—
firstly we recall those for real series and then we see how this extends to complex series.
2.1 Recap on (absolute) convergence of real series
Recall (we won’t write it here) the ε-N definition of convergence of (real) sequences. We will
make much use of the results on convergence of real series (from earlier courses on analy-
sis), so here is a quick recap of the pertinent properties. For proofs you should refer to earlier
course notes. Almost all the material in this chapter goes back to a text book written in 1821
by A.-L. Cauchy1 . He gave a lecture course in Analysis at the famous École Polythechnique in
Paris, and this text book (called Cours d’Analyse) was foundational. Prior to that, mathemati-
cians had a good idea of what the sum of an infinite series meant (Newton, Leibniz, Euler and
1 Augustin-Louis Cauchy (1789–1857)
25
26 § 2.1 Recap on (absolute) convergence of real series
others all used them) but Cauchy was the first to suggest a precise definition. And his defi-
nition of convergent sequences (and what we now call Cauchy sequences) was introduced in
order to study and make sense of infinite series.
Consider the series of real numbers
∞
X
an . (2.1)
n=1
PN
To understand convergence, one forms the sequence of partial sums, namely S N := n=1 a n
for N = 1, 2, 3, . . .
The series (2.1) is said to converge if the sequence (S N ) of partial sums converges to a limit
S, and this value S is called the sum of the given series. A series which does not converge is
said to diverge.
The series (2.1) is said to converge absolutely if ∞ n=1 |a n | converges.
P
Recall that if a series converges absolutely then it converges (see Lemma 2.6 for proof for
complex series).
P∞ n
Examples 2.1. (i) The series n=0 (−1) = 1 − 1 + 1 − 1 + 1 − 1 + · · · diverges, because the
sequence of partial sums is 1, 0, 1, 0, 1, 0, 1, . . . which obviously doesn’t converge (it oscillates).
1 1 1 1
(ii) Famously, the harmonic series ∞ n=1 n = 1 + 2 + 3 + 4 + · · · diverges. (It diverges very,
P
very slowly. For example, to get the partial sum up to 20 takes over 400 million terms!)
1
(iii) A very useful fact is the following: the series ∞ n=1 n a converges if and only if a > 1
P
(this can be proved by the integral test, which we don’t discuss in this course).
(−1)n+1
(iv) The alternating harmonic series ∞ = 1 − 21 + 31 − 41 + · · · converges (it sums
P
n=1 n
to log 2—see Exercise 2.22 if interested), but (ii) tells us it doesn’t converge absolutely.
Convergence tests Recall the following tests for absolute convergence of real series—some
of which we will revisit in the context of complex series in §2.3 below:
• Comparison test: if b n is absolutely convergent and |a n | ≤ |b n | then is abso-

P P
n an
lutely convergent.
¯ ¯
• Ratio test: let limn→∞ ¯ aan+1 ¯ = ℓ (if the limit exists).
¯ ¯
n
– If ℓ < 1 then the series converges absolutely,

– If ℓ > 1 the series diverges,
– If ℓ = 1 no conclusion can be drawn.
• Root test: let limn→∞ |a n |1/n = ℓ (if the limit exists), then the same conclusions hold
depending on ℓ: ℓ < 1 (abs. conv.), ℓ > 1 (div.) and ℓ = 1 (inconclusive)
2023–24
P OWER SERIES 27
• Cauchy-Hadamard test: if the limit in the root test doesn’t exist, then one can modify the
expression so that the limit does exist:
ℓ = lim sup |a n |1/n .

n→∞
This limit always exists (although may be infinite) as we discuss below, and then the
conclusions are the same according to whether ℓ is larger or smaller than 1.
The limsup of a sequence: Let x n be a sequence of positive real numbers. For each n,
consider supk≥n x k (that is, ignore the first n − 1 terms and look at the sequence from the n th
term on, and find the supremum of that).
• if the sequence is not bounded above then this is infinite for all n;
• if the original sequence is bounded, then the new sequence y n := supk≥n x k is decreas-
ing (because there are fewer terms to take the supremum of).
Since the x n ≥ 0, it follows that y n ≥ 0. Recall that any decreasing sequence of non-
negative reals converges. Hence
µ ¶
lim y n = lim sup x k
n→∞ n→∞ k≥n
exists (although in the unbounded case it will be equal to ∞). We denote the limit by lim supn→∞ x n .
Thus lim sup x n exists (allowing ∞ as ‘existing’) for any postive sequence x n . (And one can
show that if limn→∞ x n = ℓ then lim supn→∞ x n = ℓ.)
The limsup shares serval properties with ordinary limits, but not all. For example,
lim sup(x n + a) = (lim sup x n ) + a, lim sup(ax n ) = a lim sup x n (for a ≥ 0)
However, lim sup(x n + v n ) ≤ lim sup x n + lim sup v n (simple examples show you don’t neces-
sarily get equality: see below).
Example 2.2 (of a lim sup). Consider the sequence | sin(n)| (n ∈ Z). Since n is never a
multiple of π/2 the terms are never equal to 1, but one can make it as close as one likes.
Thus for any n, supk≥n sin(n) = 1. And so lim supn→∞ | sin(n)| = 1.
On the other hand, n| sin(n)| is unbounded, so lim supn→∞ n| sin(n)| = ∞.
Finally, consider
(
1 + 1/n for n odd,
x n :=
0 for n even.
Then the limit of the even terms is 0, while the limit of the odd terms is 1, so the sequence
doesn’t have a limit. However, the supremum construction ignores the even terms (the sup
will always be the next odd term), and so lim supn→∞ x n = 1.
Notice that if we define v n similarly to x n but with odd and even swapped round (so
2023–24
28 § 2.2 Limits of complex sequences
v n = 0 for odd n), then

x n + v n = 1 + 1/n, (∀n)
hence lim supn→∞ (x n + v n ) = 1 as well, showing that the limsup’s don’t add.
Example 2.3. The different tests described above are all proved by using the comparison
test, based on the geometric series, so let us recall the argument for that. Let
S n = 1 + x + x 2 + · · · + x n−1 .
Then S n − xS n = 1 − x n , so that (for x 6= 1)
1 − xn
Sn = ,
1−x
(so far it’s just algebra of polynomials, and not analysis, and valid for all x ∈ R except x = 1;
note that if x = 1 then S n = n). To find the sum of the infinite series S ∞ , we need to take the
limit as n → ∞ (by definition of the sum of an infinite series).
1
Clearly, if |x| < 1 then x n → 0 so that limn→∞ S n = 1−x . On the other hand, if |x| > 1 then
S n → ±∞ (the sum diverges). It also diverges for x = −1, but not to ±∞.
Alternating sequences There are some sequences that converge, but don’t converge ab-
solutely (an example is given in Example 2.1(iv) above). The principal method for proving
convergence in such cases is Leibnitz’s test:
• if |a n | → 0 monotonically, and the a n alternate in sign, then n a n converges.
P
See the optional Exercise 2.21 for a proof of this test.
2.2 Limits of complex sequences
Let z n ∈ C be a sequence of complex numbers. The definition of convergence to a limit is the

same as for real sequences, with absolute value replaced by modulus: we say that z n → z as
n → ∞ if, for all ε > 0 there exists N ∈ N such that if n ≥ N then |z n − z| < ε.
Lemma 2.4. Let z n ∈ C and write z n = x n + i y n , x n , y n ∈ R. Then the sequence z n converges

if and only if both x n and y n converge.
Proof: Suppose that z n → z and write z = x + i y. Then

q
|x n − x| ≤ |x n − x|2 + |y n − y|2 = |z n − z| → 0
2023–24
P OWER SERIES 29
as n → ∞. Hence x n → x. A similar argument shows that y n → y.

Conversely, suppose that x n → x and y n → y. Then
q
|z n − z| = |x n − x|2 + |y n − y|2 → 0
so that z n → z. ❒
2.3 On (absolute) convergence of complex series
Let z k ∈ C. We say that the series ∞ z converges if the sequence of partial sums s n =
P
Pn k=0 k
z converges. The limit of this sequence of partial sums is called the sum of the series.
k=0 k
In Exercise 2.1 we show that ∞ z n is convergent if, and only if, both ∞n=0 Re(z n ) and
P P
P∞ n=0
n=0 Im(z n ) are convergent.
We will need a stronger property than just convergence.
P∞
Definition 2.5. Let z n ∈ C. We say that n=0 z n is absolutely convergent if the real series
P∞
n=0 |z n | is convergent. ✯
P∞ P∞
Lemma 2.6. Suppose that n=0 z n is absolutely convergent. Then n=0 z n is convergent.
P∞
Proof: Suppose that n=0 z n is absolutely convergent. Let z n = x n +i y n . Then |x n |, |y n | ≤
|z n |. Hence by the comparison test, the real series ∞
P∞
n=0 x n and n=0 y n are absolutely
P
convergent. As x n , y n are real, we know that

¯∞ ¯ ∞ ¯∞ ¯ ∞
¯X ¯ X ¯X ¯ X
¯
¯ xn ¯ ≤
¯ |x n |, ¯
¯ y n ¯¯ ≤ |y n |,
n=0 n=0 n=0 n=0
P∞ P∞ P∞
so that n=0 x n and n=0 y n are convergent. By the above remark, n=0 z n is convergent.
❒
We saw in Example 2.1(iv) an example of a (real) series which converges but doesn’t con-
verge absolutely. One can similarly construct examples of complex series with the same
property—one example is n i n /n (which sums to − 12 log 2 + 41 i π; see Exercise 2.7).
P
One reason for working with absolutely convergent series is that they behave well when
multiplied together. Indeed, two series which converge absolutely may be multiplied in a
similar way to two finite sums. First note that if we have two finite sums then we can multiply
them together systematically as follows:
(a 0 + a 1 + a 2 + a 3 + · · · + a n )(b 0 + b 1 + b 2 + b 3 + · · · + b n )
= (a 0 b 0 ) + (a 0 b 1 + a 1 b 0 ) + (a 0 b 2 + a 1 b 1 + a 2 b 0 ) + (a 0 b 3 + a 1 b 2 + a 2 b 1 + a 3 b 0 ) + · · ·
2023–24
30 § 2.3 On (absolute) convergence of complex series
For absolutely convergent series the following proposition holds. (We remark that Propo-
sition 2.7 is not true in general if one of the infinite series converges but is not absolutely
convergent.)
P∞ P∞
Proposition 2.7. Let a n , b n ∈ C. Suppose that n=0 a n and n=0 b n are absolutely con-
vergent. Then µ ∞
X
¶µ ∞
X
¶ ∞
X
an bn = cn
n=0 n=0 n=0
P∞
where c n = a 0 b n + a 1 b n−1 + a 2 b n−2 + · · · + a n b 0 and moreover n=0 c n is absolutely con-
vergent.
Proof: Omitted. (It can be found in Stewart and Tall, p.71.) ❒
The convergence tests for real series we discussed above continue to hold for complex
series and we state them below as propositions.
Proposition 2.8 (The ratio test). Let w n ∈ C. Suppose that

|w n+1 |
lim = ℓ. (2.2)
n→∞ |w n |
P∞ P∞
If ℓ < 1 then n=0 w n is absolutely convergent. If ℓ > 1 then n=0 w n diverges.
Proposition 2.9 (The root test). Let w n ∈ C. Suppose that
lim |w n |1/n = ℓ. (2.3)

n→∞
P∞ P∞
Proposition 2.10 (The Cauchy-Hadamard test). Let w n ∈ C. Let
ℓ = lim sup |w n |1/n . (2.4)

n→∞
P∞ P∞
The advantage of the Cauchy-Hadamard test is that the limit ℓ always exists (although it
may be infinite), see p. 27. The disadvantage is that it’s harder to calculate.
Remark In all three tests, if ℓ = 1 then we can conclude nothing: the series may converge
absolutely, it may converge but not absolutely converge, or it may diverge. We will see such
examples below. ❞
2023–24
P OWER SERIES 31
Example 2.11. Consider the series

X∞ in
.
n=0 2
n
Here w n = i n /2n . We can use the ratio test to show that this series converges absolutely.
Indeed, note that
¯ w n+1 ¯ ¯ i n+1 2n ¯ ¯ i ¯ 1
¯ ¯ ¯ ¯ ¯ ¯
¯ w ¯ ¯ 2n+1 i n ¯ = ¯ 2 ¯ = 2 .
¯ ¯=¯ ¯ ¯ ¯
n
Hence limn→∞ |w n+1 /w n | = 1/2 < 1 and so by the ratio test the series converges absolutely.
We could also have used the root test to show that this series converges absolutely. To
see this, note that ¯ n ¯1/n µ ¶1/n
1/n
¯i ¯ 1 1
|w n | = ¯¯ n ¯¯ = n = .
2 2 2
Hence limn→∞ |w n |1/n = 12 < 1 and so by the root test the series converges absolutely.
2.4 Power series and their radius of convergence
The reason we considered the convergence tests above is that we need to apply them to de-
termine the convergence of power series.
The simplest form of power series is an ’infinite polynomial’:
∞
an z n ,
X
(2.5)
n=0
with coefficients a n ∈ C. This is called a power series at 0 (or about 0). If we replace 0 by
another point z 0 we get the following more general definition:
P∞ n
Definition 2.12. A series of the form n=0 a n (z − z 0 ) where a n ∈ C, z ∈ C is called a power
series at z 0 . ✯
By changing variables and replacing z − z 0 by z we return to the expression (2.5), so the

two are very similar.
The principal question we wish to address here is, for which values of z does a given power
series converge?
Let R be the largest modulus of any z for which (2.5) converges. More precisely,
½ ∞ ¾
n
X
R = sup r ≥ 0 | ∃ z ∈ C such that |z| = r and a n z converges . (2.6)
n=0
Note that (2.5) certainly converges when z = 0 (the sum is a 0 ), so that R ≥ 0 (and it’s possible
that R = 0). We allow R = ∞ if no finite supremum exists.
2023–24
32 § 2.4 Power series and their radius of convergence
P∞ n
Theorem 2.13. Let n=0 a n z be a power series and let R be defined as above. Then
P∞ n
(i) n=0 a n z converges absolutely for |z| < R;
P∞ n
(ii) n=0 a n z diverges for |z| > R.
Recall from Chapter 1 that the disc of radius r and centre a ∈ C is denoted
D(a; r ) := {z ∈ C | |z − a| < r } .
Thus (i) says that absolute convergence occurs for all z ∈ D(0; R).
The proof of (i) is based on the Cauchy-Hadamard test, and we separate out that part of
the argument in the following lemma.
Lemma. Suppose n a n w n converges and |z| < |w | then a n z n converges absolutely.
P P
Proof: We use the Cauchy-Hadamard test. Let ℓw = lim supn→∞ |a n w n |1/n and ℓz sim-
n
ilarly. Now since converges it follows from the Cauchy-Hadamard test that
P
n an w
ℓw ≤ 1.
Let ρ = |w |/|z|. Then ρ < 1. We see,
ℓz = lim sup |a n z n |1/n

n→∞
= lim sup |a n |1/n |z|
n→∞
= ρ lim sup |a n |1/n |w |
n→∞
= ρℓw .
Since ℓw ≤ 1 it follows that ℓz ≤ ρ < 1 and hence a n z n converges absolutely (by the
P
Cauchy-Hadamard test). ❒
Proof of Theorem 2.13: (i) Fix z ∈ D(0; R). If there were a point w with |w | = R for which
n
converges then the lemma would guarantee the result, since |z| < |w |.
P
n an w
But since R is the supremum of a set, it might not be achieved in that set, so a more
delicate argument is needed. Since |z| < R there is a w for which n a n w n converges,
P
with |z| < |w | < R (for otherwise R would not be the supremum). The result then follows
from the lemma using this value of w .
(ii) This is just the definition of R; if |z| > R then the series diverges. ❒
The remarkable fact in this theorem is that the domain of convergence of any power series
is always a circular disc (possibly the whole plane if R = ∞). Note however that this theorem
leaves out the case |z| = R (a circle—the boundary of the disc). And for good reason: the
power series may converge everywhere on the circle, it may converge for some z but not all,
or it may diverge for every z in the circle. We’ll see some examples below and in the exercises.
2023–24
P OWER SERIES 33
Definition 2.14. The number R defined in Eq. (2.6) is called the radius of convergence of
n
the power series ∞n=0 a n z . The disc D(0; R) is called the disc of convergence.
P
✯
We would like some ways of computing the radius of convergence of a power series.
Proposition 2.15 (Radius of convergence formulæ).

P∞ n
Let n=0 a n z be a power series with radius of convergence R.
|a n+1 |
(i) If ℓ = lim exists, then R = 1/ℓ.
n→∞ |a n |
(ii) If ℓ = lim |a n |1/n exists, then R = 1/ℓ;
n→∞
(iii) Let ℓ = lim supn→∞ |a n |1/n , then R = 1/ℓ.
(We interpret 1/0 as ∞ and 1/∞ as 0.)
Since the lim sup always exists (see p.27), part (iii) gives a formula for the radius of con-
vergence, known as the Cauchy-Hadamard formula2 , and having a formula is very useful
for proving general statements, as we will see. However, for many power series that arise in
practice, one can use one of the two simpler expressions (i) and (ii).
Futhermore, if the limit in (i) exists then the limit in (ii) exists and they give the same
answer for the radius of convergence. It is straightforward to find examples of sequences a n
for which the limit in (ii) exists but the limit in (i) does not. If neither limit exists, one needs
the Cauchy-Hadamard formula.
If R = ∞ then the power series converges for every z ∈ C, and we say the resulting function
is an entire function.
Example 2.16. Consider

X∞ zn
.
n=0 n
Here a n = 1/n. In this case ¯ ¯

¯ a n+1 ¯ n
¯ a ¯ = n +1 → 1
¯ ¯
n
as n → ∞. Hence the radius of convergence is equal to 1 and the series converges for all
z ∈ D(0; 1).
Example 2.17. Consider

X∞ zn
.
n=0 2
n
2 named after Cauchy and Jaques Hadamard (1865–1963): Cauchy proved the root test while Hadamard ex-
tended it to this version with the lim sup (although it is said that Riemann used this earlier but didn’t publish
it)
2023–24
34 § 2.5 Differentiation of power series
Here a n = 1/2n . Using Proposition 2.15(i) we can calculate the radius of convergence as
|a n+1 | 2n 1 1
ℓ = lim = lim n+1 = lim =
n→∞ |a n | n→∞ 2 n→∞ 2 2
so that R = ℓ−1 = 2. Alternatively, we could use Proposition 2.15(ii) and see that
¶1/n
1 1 1
µ
1/n
ℓ = lim |a n | = lim n = lim = ,
n→∞ n→∞ 2 n→∞ 2 2
so that again R = ℓ−1 = 2.
Proof of Proposition 2.15: First we prove (i). Suppose that |a n+1 /a n | converges to a limit,
say ℓ, as n → ∞, i.e. ¯ ¯
¯ a n+1 ¯
lim ¯¯ ¯ = ℓ.
n→∞ a ¯
n
Then
|a n+1 z n+1 | |a n+1 |
lim n
= lim |z| → ℓ|z|.
n→∞ |a n z | n→∞ |a n |
n
By the ratio test, the power series ∞ n=0 a n z converges for ℓ|z| < 1 and diverges for ℓ|z| >
P
1. Hence the radius of convergence R = 1/ℓ.

Now we prove (ii). Suppose that |a n |1/n → ℓ as n → ∞. By the root test, the power
n 1/n
series ∞ n
n=0 a n z converges if limn→∞ |a n z | = limn→∞ |a n |1/n |z| = ℓ|z| < 1 and di-
P
n 1/n 1/n
verges if limn→∞ |a n z | = limn→∞ |a n | |z| = ℓ|z| > 1. Hence the radius of conver-
gence R = 1/ℓ.
The proof of (iii) is the same as the proof of (ii), replacing lim by lim sup. ❒
2.5 Differentiation of power series
In this section, we recall the derivative for functions of a complex variable and use this to find
the derivative of a power series.
Recall from (1.2) that the derivative of a function f at z 0 ∈ C is the limit
f (z) − f (z 0 )
f ′ (z 0 ) := lim
z→z 0 z − z0
provided the limit exists.

It is easy to show that for a polynomial
p(z) = a 0 + a 1 z + · · · + a n z n
2023–24
P OWER SERIES 35
the derivative is given by

p ′ (z) = a 1 + 2a 2 z + · · · + na n z n−1 .
(See Exercise 1.3). This suggests that a power series
∞
an z n
X
f (z) = (2.7)
n=0
can be differentiated term by term to give
∞
f ′ (z) = na n z n−1 .
X
(2.8)
n=1
However, because we are dealing with infinite sums, this needs to be proved. There are two
steps to this: (i) we have to show that if (2.7) converges for |z| < R then (2.8) converges for
|z| < R, and (ii) that f (z) is differentiable for |z| < R and the derivative is given by (2.8).
Lemma 2.18. Consider the two power series,

∞
an z n ,
X
f (z) =
n=0
∞
na n z n−1 .
X
g (z) =
n=1
The power series f and g have the same radius of convergence.
Proof: Using the Cauchy-Hadamard formula (Proposition 2.15(iii)), we need to show

that their respective ‘lim sups’ are equal. Call them ℓ f and ℓg . Then
ℓg = lim sup |na n |1/n , and ℓ f = lim sup |a n |1/n .

n→∞ n→∞
We use the well-known limit,

lim 1 log x = 0. (2.9)
x→∞ x
(This is for real x, and can be proved by L’Hopital’s rule, for example). It follows from this
that as n → ∞, one has n 1/n → 1 (take logs of this and replace n by x).
Finally, ℓg = lim supn→∞ |na n |1/n = lim supn→∞ n 1/n |a n |1/n and since n 1/n → 1, this
¡ ¢
reduces just to ℓ f as required. ❒
P∞ n
Theorem 2.19. Let f (z) = have radius of convergence R > 0. Then f (z) is
n=0 a n z
n−1
differentiable on the disc of convergence D(0; R) and f ′ (z) = ∞
P
n=1 na n z .
2023–24
P∞ n−1
Proof: Let g (z) = n=1 na n z . By Lemma 2.18 we know that this converges for |z| < R.
Fix z 0 ∈ C with |z 0 | < R. We have to show that f (z) is differentiable at z 0 and, moreover,
the derivative is equal to g (z 0 ), i.e. we have to show that if |z 0 | < R then
f (z) − f (z 0 )
f ′ (z 0 ) := lim = g (z 0 )
z→z 0 z − z0
or equivalently
f (z) − f (z 0 )
µ ¶
lim − g (z 0 ) = 0.
z→z 0 z − z0
It will be convenient to define the polynomials,
q n (z) = z n−1 + z 0 z n−2 + · · · + z 0n−2 z + z 0n−1 .
Then q n (z)(z − z 0 ) = z n − z 0n . Moreover, q(z 0 ) = nz 0n−1 .

For any N ≥ 1 we have the following
f (z) − f (z 0 ) ∞ z n − z 0n
µ ¶
− na n z 0n−1
X
− g (z 0 ) = an
z − z0 n=1 z − z0
∞ ¡
a n q n (z) − na n z 0n−1
X ¢
=
n=1
∞
a n q n (z) − nz 0n−1 .
X ¡ ¢
=
n=1
[Note: we need to take the limit as z → z 0 , and if we could swap the infinite sum and the
limit we’d be done, since q n (z 0 ) = nz 0n−1 ; but this needs to be proved!]
Now fix N > 0 and split the infinite sum into 2 parts:
N ∞
a n (q n (z) − nz 0n−1 ) + a n (q n (z) − nz 0n−1 )
X X
=
n=1 n=N +1
= Σ1,N (z) + Σ2,N (z), say.
We are interested in the limit of this sum as z → z 0 :
f (z) − f (z 0 )
µ ¶
lim − g (z 0 ) = lim Σ1,N (z) + lim Σ2,N (z). (2.10)
z→z 0 z − z0 z→z 0 z→z 0
Now (for any N > 0), Σ1,N is a polynomial, so the (finite) sum commutes with the limit:
N
a n lim (q n (z) − nz 0n−1 ) = 0,
X
lim Σ1,N (z) = (2.11)
z→z 0 z→z 0
n=1
since q n (z 0 ) = nz 0n−1 .
2023–24
P OWER SERIES 37
It remains to consider the Σ2,N term. For this, fix ε > 0. Choose r such that |z 0 | < r < R.
n−1
Then, as in the proof of Lemma 2.18, ∞ is absolutely convergent. Hence we
P
n=1 na n r
can choose N = N (ε) such that
∞ ε
|na n r n−1 | < .
X
n=N +1 2
Since |z 0 | < r , provided z is close enough to z 0 so that |z| < r then we have that
∞ 2ε
2n|a n |r n−1 <
¯ ¯ X
¯Σ2,N (z)¯ ≤ = ε.
n=N +1 2
It follows that ¯ ¯
lim ¯Σ2,N (z)¯ < ε. (2.12)
z→z 0
Consequently, for each ε > 0 it follows from (2.10), (2.11) and (2.12) that
¯ f (z) − f (z 0 )
¯ ¯
¯
lim ¯¯ − g (z 0 )¯¯ < ε.
z→z 0 z − z0
Since this is true for any ε > 0, it follows that,
¯ f (z) − f (z 0 )
¯ ¯
¯
lim ¯¯ − g (z 0 )¯¯ = 0
z→z 0 z − z0
and hence f ′ (z 0 ) exists and equals g (z 0 ). ❒
n
The above two results have a very important consequence. If f (z) = ∞ n=0 a n z has disc
P
of convergence D(0; R) then we can differentiate it as many times as we like within this disc.
P∞ n
Corollary 2.20. Let f (z) = n=0 a n z have radius of convergence R. Then all of the
′ ′′ ′′′ (k)
higher derivatives f , f , f , . . . , f , . . . of f exist for z within the disc of convergence.
Moreover,
∞ ∞ n!
f (k) (z) = n(n − 1) · · · (n − k + 1)a n z n−k = a n z n−k .
X X
n=k n=k (n − k)!
Proof: This is a simple induction on k. ❒
n
Antiderivatives We start with the series f (z) = . Now integrate this term by term to
P
n an z
obtain a new power series
X a n n+1
F (z) = z .
n n +1
Question: is this the antiderivative of f ? Answer:
2023–24
Corollary 2.21. Let f (z) = a z n be a power series with radius of convergence R > 0.
P
Pn nan n+1
Then the power series F (z) = n n+1 z also has radius of convergence R and moreover
′
satisfies F (z) = f (z) for all z ∈ D(0; R).
Proof: Suppose the power series F has radius of convergence R 1 . By the theorem, the
derivative F ′ of F has the same radius of convergence as F and is given by term-by-term
differentiation of the power series: that is, F ′ = f and R 1 = R. ❒
Power series about a different centre. Instead of using a power series at the origin, by
replacing z by z − z 0 we can consider a power series at the point z 0 . (This will be useful
n
later on when we look at Taylor series in Chapter 4.) Suppose that f (z) = ∞ n=0 a n z has disc
P
of convergence D(0; R). Then, replacing z by z − z 0 , we have that the power series g (z) =
P∞ n
n=0 a n (z −z 0 ) has disc of convergence D(z 0 ; R). That is, the power series g (z) converges for
all z inside the disc with centre z 0 and radius R. Moreover, inside this disc of convergence all
the higher derivatives of g exist and
∞ n!
g (k) (z) = a n (z − z 0 )n−k .
X
n=k (n − k)!
Uniqueness of power series It is useful to realize that power series are unique. For exam-
ple, we know
1
= 1 + z + z2 + z3 + · · ·
1−z
for z ∈ D(0; 1), but conceivably there could be another power series which converges to f in a
neighbourhood of 0. But in fact there can’t be as the following result shows.
Proposition 2.22. Suppose that

∞ ∞
a n (z − z 0 )n = b n (z − z 0 )n
X X
(2.13)
n=0 n=0
for all z ∈ D(z 0 ; ε) (for some ε > 0). Then a n = b n for all n ∈ N.
Proof: Subtracting the right-hand side from the left-hand side in (2.13), it is sufficient to
prove that if
∞
c n (z − z 0 )n = 0
X
(2.14)
n=0
for all z ∈ C with |z − z 0 | < ε then c n = 0 (putting c n = a n − b n into (2.13)). First substitute
z = z 0 to obtain c 0 = 0. Next differentiate both sides of (2.14) and put z = z 0 to obtain
c 1 = 0. Continue in this way, differentiating k times, to see that, for each k ∈ N, ∞
P
n=k
n(n −
n−k
1) · · · (n−(k −1))c k (z−z 0 ) = 0. Putting z = z 0 into this expression then shows that c k = 0;
that is, a k = b k . ❒
2023–24
P OWER SERIES 39
Remark 2.23. Sometimes one encounters (and we will see more later) series with negative
powers of z. For example,
1 1 1 1 1
+ 2 + 3 + 4 +··· + n +··· .
z z z z z
This is a geometric progression, with common ration z −1 , so converges for |z −1 | < 1, or equiv-
alently |z| > 1. The sum is
1 z
= for |z| > 1.
1 − z −1 z − 1
❞
2.6 Elementary functions
‘Elementary functions’ are functions formed from polynomials, exponentials, logarithms,

trigonometric functions and radicals (roots), using the four arithmetic operations +, −, ×, ÷.
All but one of the functions we define here are defined by power series (the logarithm being
the one exception).
2.6A The exponential function
You will already have met the (real) exponential function e x whose power series (or Taylor
n
series) is ∞n=0 x /n!. We use this as a basis for the (complex) exponential function.
P
Definition 2.24. The exponential function is defined to be the power series

X∞ zn
exp z = .
n=0 n!
By Proposition 2.15(i) we see that the radius of convergence R for exp z satisfies
1 n! 1
= lim = lim =0
R n→∞ (n + 1)! n→∞ n + 1
so that R = ∞. Hence this series has radius convergence ∞, and so converges absolutely for
all z ∈ C: functions whose power series converges for all z ∈ C are said to be entire functions.
d
By Theorem 2.19 we may differentiate term-by-term to obtain dz exp z = exp z (as in the
real case). This is because
d X∞ z n−1 X ∞ z n−1 X∞ zn
exp z = n = = = exp z.
dz n=1 n! n=1 (n − 1)! n=0 n!
In the real case we know that if x, y ∈ R then e x+y = e x e y . This is also true in the complex
case, and follows from a calculation based on Proposition 2.7 (see Exercise 2.4). An entirely
different argument is given in Exercise 2.15.
2023–24
40 § 2.6 Elementary functions
Proposition 2.25. Let z 1 , z 2 ∈ C. Then exp(z 1 + z 2 ) = exp(z 1 ) exp(z 2 ).
In particular, if we take z 1 = z and z 2 = −z in Proposition 2.25 then we have that
exp(z) exp(−z) = exp(z − z) = exp0 = 1.
Hence exp z 6= 0 for any z ∈ C. (We already knew that e x = 0 has no real solutions; now we
know that it has no complex solutions either.)
Finally, we want to connect the real number e to the complex exponential function. We
define e to be the real number e = exp1. Then, iterating Proposition 2.25 inductively, we
obtain
e n = exp(1)n = exp(1 + · · · + 1) = expn.
For a rational number m/n (n > 0) we have that
(exp(m/n))n = exp(nm/n) = exp(m) = e m
so that exp(m/n) = e m/n . Thus the notation e z = exp z does not conflict with the usual defini-
tion of e x when z is real. Hence we shall normally write e z for exp z. In particular, if we write
z = x + i y then Proposition 2.25 tells us that
e x+i y = e x e i y .
We already understand real exponentials e x . Hence to understand complex exponentials we

need to understand expressions of the form e i y .
2.6B Trigonometric functions
As real functions, we first define sine and cosine as ratios in triangles. We later derive their
(real) power series. To define the complex versions, we just use the same power series, but
now of course z is a complex variable. Thus we define,
∞ z 2n 1 1
(−1)n = 1 − z2 + z4 − · · ·
X
cos z =
n=0 (2n)! 2 4!
∞ z 2n+1 1 1
(−1)n = z − z3 + z5 + · · ·
X
sin z =
n=0 (2n + 1)! 3! 5!
It is straightforward to check (using Proposition 2.15(i)) that these converge absolutely for all
z ∈ C, so are entire functions.
Replacing z by −z in the series we see that cos is an even function and that sin is an odd
function, i.e.
cos(−z) = cos z, sin(−z) = − sin z.
Moreover, cos(0) = 1, sin(0) = 0.
2023–24
P OWER SERIES 41
By Theorem 2.19 we can differentiate term-by-term to see that
d d
cos z = − sin z, sin z = cos z.
dz dz
Term-by-term addition of the power series for cos z and sin z shows that
exp i z = cos z + i sin z. (2.15)
In particular, if θ ∈ R we have Euler’s famous formula

ei θ = cos θ + i sin θ
and hence we can write any complex number in exponential form as
z = r ei θ
where r = |z| and θ = arg z.

Replacing z by −z we see that e −i z = cos z − i sin z. Hence
1 1 iz
cos z = (e i z + e −i z ), sin z = (e − e −i z ).
2 2i
Squaring the above expressions and adding them gives cos2 z + sin2 z = 1. These are all ex-
pressions that we already knew in the case when z is a real number; now we know that they
continue to hold when z is any complex number. Carrying on in the same way, one can prove
the addition formulæ cos(z 1 + z 2 ) = cos z 1 cos z 2 − sin z 1 sin z 2 , etc, for complex z 1 , z 2 , and all
the other usual trigonometric identities (see Exercise 2.8).
sin z
Moreover, we can define the other trig functions as usual, such as tan z = cos z , and cot,
sec, csc similarly.
Some of these are covered in the exercises at the end of the chapter.
2.6C Hyperbolic functions
Define
e z + e −z e z − e −z
cosh z = , sinh z = .
2 2
If we write out the resulting series for cosh and sinh we see that cosh z has all the even terms
from exp z, while sinh z has all the odd terms.
Differentiating these definitions we see that
d d
cosh z = sinh z, sinh z = cosh z.
dz dz
One can also prove addition formulæ for the hyperbolic trigonometric functions, and other
identities including (for example)
cosh2 z − sinh2 z = 1, for all z ∈ C

2023–24
42 § 2.6 Elementary functions
(again, we knew this already when z ∈ R).

We also have the relations
cos i z = cosh z, sin i z = i sinh z;
(see Exercise 2.8).
2.6D Periods of the exponential and trigonometric functions
Definition 2.26. Let f : C → C. We say that a number p ∈ C is a period for f if

f (z + p) = f (z) for all z ∈ C. ✯
Clearly if p ∈ C is a period and n ∈ Z is any integer then np is also a period.

For the exponential function, we have that
e 2πi = cos 2π + i sin 2π = 1
so that
e z+2πi = e z e 2πi = e z .
Hence 2πi is a period for exp, as is 2nπi for any integer n. In Exercise 2.12 we shall see that
these are the only periods for exp.
We shall also see in the exercises that the only complex periods for sin and cos are 2nπ.
2.6E The logarithm
In real analysis, the (natural) logarithm is the inverse function to the exponential function.
That is, if e x = y then x = log y. (Throughout, when the argument is real and positive then log
will mean what it traditionally means.) Here we consider the complex analogue of this. It is
the one function in this chapter that we do not define by a power series.
Let z ∈ C, z 6= 0, and consider the equation
exp w = z. (2.16)
By §2.6D, if w 1 is a solution to (2.16) then so is w 1 + 2nπi (for n ∈ Z). Each of these values is
called a logarithm of z, and we denote any of these values by log z. Thus, unlike in the real
case, the complex logarithm is a multi-valued function, but any two values differ by 2nπi for
some n ∈ Z.
We want to find a formula for log z. In (2.16) write w = x + i y. Then
z = exp w = exp(x + i y) = e x (cos y + i sin y). (2.17)
By taking the modulus of both sides of (2.17) we see that e x = |z|. Note that both x and |z| are
real numbers. Hence x = log |z|. By taking the argument of both sides of (2.17) we see that
y = arg z. Hence we can make the following definition.
2023–24
P OWER SERIES 43
Definition 2.27. Let z ∈ C, z 6= 0. Then a complex logarithm of z is
log z = log |z| + i arg z
where arg z is any argument of z. The principal value of log z is the value of log z when arg z
has its principal value Arg z, i.e. the unique value of the argument in (−π, π]. We denote the
principal logarithm by Log z:
Log z = log |z| + i Arg z. (2.18)
If we write z = r ei θ then log z = log r + i θ is a logarithm of z. ✯
Note that we say a complex logarithm (rather than the complex logarithm) to emphasise
the fact that the complex logarithm is multi-valued.
Dealing with multivalued functions is tricky (you’ll see more about this in part 2 of this
course). One way is to only consider the logarithm function on a subset of C.
Notice also that, like the real logarithm, log(z) is not defined for z = 0.
Definition 2.28. The complex plane with the negative real-axis (including 0) removed is
called the cut plane, and denoted C− . Formally,
n o
C− = z = r ei θ ∈ C | r > 0, θ ∈ (−π, π) .
It is a domain. See Figure 2.1. ✯
F IGURE 2.1: The cut plane C− : this is the complex plane with the negative real
axis and 0 removed.
The reason for considering the cut plane is that it is the natural domain for the definition
of Log, and Log is holomorphic there. It is easy to see first that Log is continuous there: look-
ing at (2.18) we know the real log is continuous and clearly Arg is continuous on C− (it has a
discontinuity at points of the negative real axis where it jumps between π and −π, but that is
precisely the set we are excluding in C− ).
2023–24
44 § 2.7 Exercises
Proposition 2.29. The principal logarithm Log : C− → C is holomorphic and
d 1
Log z = for z ∈ C− .
dz z
Proof: Let w = Log z. Then z = exp w . Let Log(z + h) = w + k. Now as already pointed
out, Log is continuous on the cut plane so we have that k → 0 as h → 0. Then
d Log(z + h) − Log(z)
Log z = lim
dz h→0 h
(w + k) − w
= lim
k→0 exp(w + k) − exp(w )
exp(w + k) − exp(w ) −1
µ ¶
= lim
k→0 k
¶−1
d
µ
= exp(w )
dw
1
= .
z
Having defined the complex logarithm we can go on to define complex powers. For b, z ∈
C with b 6= 0 we define the principal value of b z to be
b z = exp(z Log b)
and the subsidiary values to be exp(z log b).

So this means we know how to raise any complex number to the power of any other com-
plex number. More on this and other uses of log in Part 2.
2.7 Exercises
Ex 2.1 Let z n ∈ C. Show that ∞

P∞
n=0 z n is convergent if, and only if, both n=0 Re(z n ) and
P
P∞
n=0 Im(z n ) are convergent. [ Hint: Use Lemma 2.4]
Ex 2.2 Find the radius of convergence of each of the following power series:
∞ 2n z n ∞ zn ∞ ∞
n! z n , n p z n (p ∈ N).
X X X X
(i) , (ii) , (iii) (iv)
n=1 n n=1 n! n=1 n=1
[Note that (iii) shows that not every power series leads to a function.]
2023–24
P OWER SERIES 45
Ex 2.3 Consider the power series

∞
an z n
X
n=0
where a n = 1/2 if n is even and a n = 1/3n if n is odd. Show that neither of the formulæ
n
for the radius of convergence given in Proposition 2.15(i) and (ii) converge for this power
series. Show by using either the limsup formula or the comparison test that this power
series converges for |z| < 2. Can you find the sum when z = 1?
Ex 2.4 (i) By multiplying two series together, show using Proposition 2.7 that for |z| < 1,
we have
∞ 1
nz n−1 =
X
.
n=1 (1 − z)2
(ii) By multiplying two series together, show using Proposition 2.7 that for z, w ∈ C we
have
X∞ zn X ∞ wn X∞ (z + w )n
= .
n=0 n! n=0 n! n=0 n!
What does this tell us about the exponential function?
Ex 2.5 Recall that if |z| < 1 then we can sum the geometric progression with common ratio z
and initial term 1 as follows:
1
1 + z + z2 + z3 + · · · + zn + · · · = .
1−z
Use Theorem 2.19 to show that for each k ≥ 1
1 ∞ µ ¶
n
z n−(k−1)
X
=
(1 − z)k n=k−1 k − 1
for |z| < 1. (When k = 2 this gives an alternative proof of the result in Exercise 2.4 (i).)
Ex 2.6 Show that the series

1 1 1 1
1− p + p − p + p −···
2 3 4 5
converges. Does it converge absolutely?
zn
Ex 2.7 Consider the series ∞ n=1 n , which has radius of convergence 1. It diverges for z =
P
1 (the harmonic series) and converges for z = −1 (the alternating harmonic series). By
considering the real and imaginary parts, determine whether or not it converges for z = i .
Ex 2.8 Show that for z ∈ C we have
e i z + e −i z e i z − e −i z
(i) cos z = , (ii) sin z = .
2 2i
and deduce cos(i z) = cosh(z) and sin(i z) = i sinh(z).
Show also that, for z, w ∈ C
2023–24
46 Exercises
(iii) cos i z = cosh z, sin i z = i sinh z;

(iv) sin(z + w ) = sin z cos w + cos z sin w ,
(v) cos(z + w ) = cos z cos w − sin z sin w .
Ex 2.9 For each of the complex functions exp, cos, sin, cosh, sinh find the set of points on
which it assumes (i) real values, and (ii) purely imaginary values.
Ex 2.10 We know that the only real numbers x ∈ R for which sin x = 0 are x = nπ, n ∈ Z. Show
that there are no further complex zeros for sin, i.e., if sin z = 0, z ∈ C, then z = nπ for some
n ∈ Z. Also show that if cos z = 0, z ∈ C then z = (n + 1/2)π, n ∈ Z.
Ex 2.11 Find the zeros of the following functions
(i) 1 + e z ; (ii) 1 + i − e z .
Ex 2.12 (i) Recall that a complex number p ∈ C is called a period of f : C → C if f (z + p) =

f (z) for all z ∈ C. Calculate the set of periods of sin z.
(ii) We know that p = 2nπi , n ∈ Z, are periods of exp z. Show that there are no other
periods.
Ex 2.13 (So far, there has been little difference between the real and the complex versions of
elementary functions. Here is one instance of where they can differ.)
Let z 1 , z 2 ∈ C \ {0}. Show that
Log z 1 z 2 = Log z 1 + Log z 2 + 2nπi .
where n is an integer depending on z 1 and z 2 which need not be zero. Give an explicit
example of two complex numbers z 1 , z 2 for which Log z 1 z 2 6= Log z 1 + Log z 2 .
Ex 2.14 Calculate the principal value of i i and the subsidiary values. (Do you find it surpris-
ing that these turn out to be real?)
Ex 2.15 Use Corollary 1.32 to prove exp(z 1 + z 2 ) = exp(z 1 ) exp(z 2 ) (see Proposition 2.25) as
follows. Fix c ∈ C and consider the function defined on C by,
F (z) = exp(z) exp(c − z).
• Show that F ′ (z) = 0,

• Deduce exp(z) exp(c − z) = exp(c),
• Put z = z 1 and c = z 1 + z 2 to obtain the result.
Ex 2.16 Derive formulæ for the real and imaginary parts of the following complex functions
and check that they satisfy the Cauchy-Riemann equations:
(i) sin z, (ii) cos z, (iii) sinh z, (iv) cosh z.

2023–24
P OWER SERIES 47
Ex 2.17 Use Proposition 2.29 to determine the power series for Log(1 + z) near z = 0. What is
its radius of convergence?
Ex 2.18 Let z = r ei θ belong to the cut plane (so r > 0 and θ ∈ (−π, π)). Consider the contour
γ from 1 to z consisting of the straight line from 1 to r , and then the arc of the circle of
radius r from r to z. Sketch a diagram showing this contour. Find
dz
Z
γ z
and compare your answer to Proposition 2.29.
Ex 2.19 (Here we prove the binomial theorem for non-integer and complex powers, first
proved by Abel in 1826.)
Let α ∈ C and suppose that α 6∈ N (where N = {0, 1, 2, 3, . . . }). Define the power series
α(α − 1) 2 α(α − 1)(α − 2) 3

f (z) = 1 + αz + z + z +···
2! 3!
∞ α(α − 1) · · · (α − n + 1)
zn .
X
= 1+
n=1 n!
(Note that, as α 6∈ N, this is an infinite series.)
(i) Show that the this power series has radius of convergence 1.
α f (z)
(ii) Show that, for |z| < 1, we have f ′ (z) = .
1+z
f (z)
(iii) By considering the derivative of the function g (z) = , show that f (z) = (1+z)α
(1 + z)α
for |z| < 1.
[Why do we assume α 6∈ N?]
The final three exercises are for interest and not part of the syllabus:
Ex 2.20 As well as infinite sums one can also study infinite products, and Euler was a master
of finding such expressions. Here is one famous example of Euler’s (which we won’t prove):
x2 x2 ∞ µ x2
µ ¶µ ¶ ¶
sin(πx) = πx (1 − x 2 ) 1 −
Y
1− · · · = πx 1− 2
4 9 j =1 n
To see why this might be true, consider where each side vanishes (but that’s not a proof!).
Rather than prove Euler’s example, we consider this simpler one:
1 ∞ ³ k
´
1 + x2 .
Y
= (2.19)
1−x k=0
2023–24
48 Exercises
First we must ask what we mean by this infinite product . . . Cauchy pointed out one can
analyze it in an analogous way to infinite sums. For n ≥ 1 let p n be the polynomial,
n ³ k
´
1 + x2 .
Y
p n (x) =
k=0
And now consider the limit as n → ∞.

(i) Simplify (1 − x)p n (x) (you will probably need to try n = 2, 3, 4 to see the pattern: lots of
uses of the difference between two squares)3 .
(ii) Divide by (1 − x) to obtain an exact expression for p n (x).
(iii) Now take the limit as n → ∞, under certain conditions on x. What condition?
(iv) Note that the argument works for x ∈ C as well as x ∈ R. Now, the right hand side of
(2.19) has many complex zeros which are not zeros of the left hand side. How are they
related to the condition you found in (iii)?
Ex 2.21 (Proof of Leibniz test for alternating series—again for interest.)

Consider the series
a0 − a1 + a2 − a3 + · · ·
where the sequence (a n ) is monotonically decreasing to 0 (in particular this means the a n
are positive). Let s n = nj=0 (−1)n a j .
P
(i). Show s 0 > s 2 > s 4 > · · · > 0.

(ii). Show s 1 < s 3 < s 5 < · · · < a 0 .
(iii). Deduce that the sequences (s 2k ) and (s 2k+1) both converge. Call their limits s + and
s − respectively.
(iv). Show s 2k − s 2k−1 → 0.
(v). Deduce s + = s − and hence the sequence of partial sums converges.
Ex 2.22 We show the alternating harmonic series sums to log 2. This series is
1 1 1 ∞ 1
(−1)n+1 .
X
1− + − +··· =
2 3 4 n=1 n
Here’s one proof that this series converges to ln 2:

R1 x n 1
(i). Show that 0 1+x dx ≤ n+1 .
Ã !
n−1
(−1)k x k − 1 = (−1)n+1 x n
X
(ii). Show that (1 + x)
k=0
(iii). Divide through by (1 + x).
(iv). Integrate the LHS over x ∈ [0, 1] and use (i) to provide a useful upper bound.
(v). Finish the argument.
3 Hint: for n = 2 you should get (1 − x)p (x) = 1 − x 8

2
2023–24
Chapter 3
Cauchy’s theorem & Cauchy’s formula
In general, looking for an antiderivative is not the best way of calculating complex integrals.
There are more powerful techniques that allow us to calculate many complex integrals with-
out having to search for antiderivatives, particularly since for many interesting integrals the
antiderivative won’t exist. One such technique that applies in the case when γ is a closed
contour is Cauchy’s Theorem.
Before stating the theorem, we need to discuss the notion of the winding number of a
path about a point.
3.1 Winding numbers
Suppose that f : D → C. The Fundamental Theorem of Contour Integration (Theorem 1.30)

tells us that if f has an antiderivative F in D and γ is any path in D from z 0 to z 1 then
Z
f = F (z 1 ) − F (z 0 ).
γ
We say that a path (or contour) γ : [a, b] → D is closed if it begins and ends at the same
point, i.e. if z 0 = γ(a) = γ(b) = z 1 .
From now on we will be considering integrals around closed contours, and to emphasize
this we use the notation I
f.
γ
The circle in the integral sign says γ is a closed contour.

In particular, it follows from Theorem 1.30 that if f has an antiderivative F on D then
I
f =0 (3.1)
γ
49
50 § 3.1 Winding numbers
for all closed paths γ in D. What happens if we do not know if f has an antiderivative? In this
case, Cauchy’s Theorem gives conditions under which (3.1) continues to hold. (Actually, there
are many different theorems of this kind, most of which are either due to, or were known to,
Cauchy and are often referred to as ‘Cauchy’s Theorem’. We will give one version expressed in
terms of winding numbers.)
Winding numbers There are several ways to view or calculate winding numbers. Here we
give the main two geometric interpretations and below we give an analytic one (in terms of
an integral).
Let γ be a closed path in C and let z 0 be a point that is not on γ.
(i) Imagine you have a piece of string. Tie one end to (say) a pencil and place the tip of the
pencil on the point z 0 . Now trace around the closed path γ with the other end of the piece
of string. When you get back to where you started, the string will be wrapped around the
pencil some number of times. This number (counted positively for anti-clockwise turns and
negatively for clockwise turns) is the winding number of γ at z 0 .
(ii) Instead of a piece of string, an easier way to calculate the winding number about z 0 is
to draw a ray (a straight line) from z 0 to outside of any disc enclosing γ (in any direction, but
making sure the ray is never tangent to γ). Now count the number of points of intersection
of γ with the ray you have chosen, but count them as +1 if γ crosses from right to left as you
look away from z 0 , and as −1 if γ crosses from left to right. (A moment’s thought shows that
right to left corresponds to γ going anticlockwise around z 0 .)
It is not hard to see these two geometric points of view are equivalent: if γ wraps r times
around the the point z 0 , then it must pass any ray at least r times, possibly more if it goes
back and forth. But each ‘back’ will cancel with a ‘forth’, with the nett total being r .
See Figure 3.1 for examples of winding numbers.
z2
z1 z1
γ1
z0 z1
z0 z0
γ2
γ3
(i) (ii) (iii)
F IGURE 3.1: In (i), w (γ1 ; z 0 ) = 1 and w (γ1 ; z 1 ) = 0. In (ii), w (γ2; z 0 ) = −1 and

w (γ2 ; z 1 ) = 0 as γ2 winds clockwise around z 0 . In (iii), w (γ3; z 0 ) = 2, w (γ3 ; z 1 ) =
1, w (γ3 ; z 2 ) = 0 as γ3 winds anticlockwise twice around z 0 , anticlockwise once
around z 1 and does not wind at all around z 2 .
2023–24
C AUCHY ’ S THEOREM & C AUCHY ’ S FORMULA 51
In examples, it is easy to calculate winding numbers by eye and this is how we shall al-
ways do it. The following results are obvious in terms of the geometric meaning of winding
number.
Proposition 3.1. (i) Let γ1 , γ2 be closed paths that do not pass through z 0 . Then
w (γ1 + γ2 ; z 0 ) = w (γ1; z 0 ) + w (γ2 ; z 0 ).
(ii) Let γ be a closed path that does not pass through z 0 . Then
w (γ− ; z 0 ) = −w (γ; z 0 ).
Proof: For (ii), each time γ− crosses our chosen ray it does so in the direction opposite to
γ. So for a particular crossing, if γ contributes +1 then γ− contributes −1, and vice versa.
❒
However, in order to use winding numbers to develop the theory of integration, we shall
need an analytic expression for the winding number w (γ; z) of a closed path γ around a point
z. Let us first consider the case when the closed path γ does not pass through the origin 0.
We need the following lemma, which we state without proof.
Lemma 3.2. Let γ be a path in C \ {0}. Then there exists a parametrization γ : [a, b] → C \ {0}
of γ for which t 7→ arg γ(t ) is a continuous function. Any other choice of parametrization with
a continuous choice of argument differs from this argument by a constant integer multiple of
2π.
Example 3.3. For example, consider
eit , 0 ≤ t ≤ π
½
γ(t ) =
e i (t +2π), π < t ≤ 2π.
Then γ describes the unit circle with centre 0 and radius 1. Here
t, 0 ≤ t ≤ π
½
arg γ(t ) =
t + 2π, π < t ≤ 2π.
and this is not continuous. However, we can find a parametrization of γ for which the argu-
ment is continuous, for example
γ(t ) = e i t , 0 ≤ t ≤ 2π
and note that arg γ(t ) = t , 0 ≤ t ≤ 2π, is continuous.

2023–24
52 § 3.1 Winding numbers
Now consider the closed path γ. We can reinterpret the winding number w (γ; 0) of γ
around 0 as the multiple of 2π giving the total change in argument along γ.
Proposition 3.4. Let γ be a closed path that does not pass through the origin. Then
1 1
I
dz = w (γ; 0).
2πi γz
This gives an analytic (rather than geometric) way of defining the winding number — by
doing an integral.
Examples 3.5. (i) Let γ(t ) = e 4πi t , 0 ≤ t ≤ 1. This winds around the origin twice anticlock-
wise, and so should have winding number w (γ; 0) = 2. We can check this using Proposi-
tion 3.4 as follows:
I1
1 1 1 1
I
dz = 4πi
4πi e4πi t dt
2πi γ z 2πi 0 e t
Z1
= 2 dt = 2.
0
(ii) Let γ(t ) = e −i t , 0 ≤ t ≤ 2π. In this case, γ winds around the origin once, but clockwise.
Thus w (γ; 0) = −1. Again, we can check this using Proposition 3.4 as follows:
Z2π
1 1 1 1
I
dz = (−i )e −i t dt
2πi γz 2πi 0 e −i t
Z2π Z2π
1 1
= −i dt = −1 dt = −1.
2πi 0 2π 0
Proof of Proposition 3.4: Intuitively this is clear: let γ : [a, b] → C \ {0} be a closed path
that does not pass through the origin. Note that γ(a) = γ(b). Then (and we put quotes
around the following to indicate that this is not a valid proof)
Zb
1 1 ′
Z
“ dz = γ (t )dt
γz a γ(t )
£ ¤b
= log(γ(t )) a (really???)
¡ ¢ ¡ ¢
= ln |γ(b)| + i arg γ(b) − ln |γ(a)| + i arg γ(a)
¡ ¢
= i arg γ(b) − arg γ(a)
= 2πi w (γ; 0)”.
The reason that the above computation does not work is that 1/z does not have log(z) (or,
indeed, the principal logarithm Log(z)) as an antiderivative on C \ {0} ). This is because
2023–24
Log(z) is not continuous on C \ {0} and so cannot be differentiable. However, Log(z) is

continuous and is an antiderivative for 1/z on the cut plane, where we remove the neg-
ative real axis from C. More generally, one can define a logarithm continuously on a cut
plane where one removes any ray from C. (A ray is an infinite straight line starting at 0;
for example, the negative real axis is a ray.)
For each α ∈ [−π, π) define the cut plane at angle α to be
Cα = C \ {r e i α | r > 0},
i.e. the complex plane with the ray inclined at angle α from the positive x-axis removed
(note that 0 ∈ Cα ). On Cα we can define arg z to be argα z = θ where
z = r e i θ , r > 0, α − 2(m + 1)π < θ ≤ α − 2mπ
where we have the freedom to choose any m ∈ Z. (The case α = π, m = 0 corresponds to

the usual principal value of the argument.)
Let γ be a closed path that does not pass through the origin. In general, γ will not lie
in one cut plane. Split γ up into pieces γ1 , . . . , γn defined on [t 0, t 1 ], . . . , [t n−1 , t n ] so that
each γr lies in a single cut plane, Cαr , say. Along each γr we will choose a value of the
argument argαr which is continuous on Cαr and such that argαr γr (t r ) = argαr +1 γr +1 (t r ),
0 ≤ r ≤ n − 1. Hence
1
I
dz = log γ(t r ) − log(γ(t r −1 ))
γr z
¡ ¢
= log |γ(t r )| − log |γ(t r −1 )| + i argαr (γ(t r )) − argαr (γ(t r −1 )) .
Now
I
1 X n Z 1
dz = dz
γz r =1 γr z
X n ¡ ¢ n ¡
X ¢
= log |γ(t r )| − log |γ(t r −1 )| + i argαr (γ(t r )) − argαr (γ(t r −1 )) .
r =1 r =1
The real parts cancel. The imaginary parts sum to
argαn (γ(t n )) − argα0 (γ(t 0 )),
the total change in argument around γ, i.e. 2πw (γ; 0). ❒
More generally, we have the following formula for the winding number around z 0 for a
closed path that does not pass through z 0 .
Corollary 3.6. Let γ be a closed path that does not pass through z 0 . Then
1 1
I
dz = w (γ; z 0 ).
2πi γ z − z0
2023–24
54 § 3.2 Cauchy’s Theorem
Proof: This is just a change-of-origin argument. Let γ : [a, b] → C be a closed path that
does not pass through z 0 . Consider the path γ1 (t ) = γ(t ) − z 0 ; this is γ translated by z 0 .
Then w (γ; z 0 ) = w (γ1 ; 0). Now
Zb
1 1 1 1
I
dz = γ′ (t ) dt
2πi γ z − z0 2πi a γ(t ) − z 0
Zb
1 1
= γ′ (t ) dt (as γ′ (t ) = γ′1 (t ))
2πi a γ1 (t ) 1
1 1
I
= dz
2πi γ1 z
= w (γ1 ; 0).
3.2 Cauchy’s Theorem
We can now state Cauchy’s Theorem.
Theorem 3.7 (Cauchy’s Theorem). Let f be holomorphic on a domain D and let γ be a

Zin D that does not wind around any point outside D (i.e. w (γ; z) = 0 for all
closed contour
z 6∈ D). Then f = 0.
γ
See Figure 3.2 for examples of paths involved in the hypotheses of Cauchy’s Theorem.
Remark The strength of Cauchy’s Theorem is that we do not need to know if f has an an-
R
tiderivative on D. (If f did have an antiderivative on D then γ f = 0 follows immediately from
the Fundamental Theorem of Contour Integration and we wouldn’t need the assumption on
winding numbers; however, possessing an antiderivative is an extremely strong assumption
on f (or on D). See Theorem 1.30 and the remark following it.) ❞
Proof: There are many proofs of Cauchy’s Theorem; here we give one based on Green’s
Theorem1 . We assume (in addition to the hypotheses stated) that f has continuous par-
tial derivatives.
1 see MATH10121 Calculus and Vectors and/or MATH20401/20411 PDEs and Vector Calculus
2023–24
γ
D
γ
γ
D
D
✔ ✔ ✗
(i) (ii) (iii)
F IGURE 3.2: In (i) and (ii), γ has winding number zero around every point
outside D, so the hypotheses of Cauchy’s Theorem (Theorem 3.7) hold. In (iii)
γ has winding number 1 around points inside the ‘hole’ in D (whose points
are not in D), hence the hypothesis of Cauchy’s Theorem do not hold.
Green’s theorem states the following: suppose that γ is a piecewise smooth closed
contour bounding a region Γ, g , h are C 1 functions defined on an open set containing Γ,
then Ïµ ¶
∂h ∂g
Z
g (x, y) dx + h(x, y) dy = − dx dy. (3.2)
γ Γ ∂x ∂y
Let f be as in the hypotheses and write f (z) = f (x + i y) = u(x, y) + i v(x, y). Note that
dz = dx + i dy. Then
Z Z
f dz = (u + i v)(dx + i dy)
γ γ
Z Z
= u dx − v dy + i v dx + u dy
γ γ
Ïµ ¶ Ïµ ¶
∂v ∂u ∂u ∂v
= − − dx dy + − dx dy
Γ ∂x ∂y Γ ∂x ∂y
= 0
as, by the Cauchy-Riemann equations, ∂u/∂x = ∂v/∂y and ∂u/∂y = −∂v/∂x hold every-
where on Γ. ❒
Remark In many ways, this proof is cheating: Green’s Theorem is a deep theorem and
not easy to prove. There are direct proofs of Cauchy’s theorem, but they are lengthy and
difficult. (The idea is to build D up from smaller pieces, often starting with the case when D
is a triangle or rectangle; see Stewart and Tall, Chapter 8. In fact it is similar to the proof of
Green’s theorem.)
Another reason for why the above proof is cheating is that Green’s theorem requires the
partial derivatives in (3.2) to be continuous. In general, the statement of Cauchy’s Theorem
2023–24
56 § 3.2 Cauchy’s Theorem
only requires the partial derivatives to exist in D (i.e. we do not need to assume that they are
continuous). In fact, as we shall see, the existence of the derivative on a domain forces the
derivative (and so the partial derivatives) to be continuous (indeed, if the derivative exists on
a domain then the function is differentiable infinitely many times). However the proof of this
fact uses Cauchy’s Theorem. ❞
There are many variants of Cauchy’s Theorem. Here we give just two simple modifica-
tions.
Our first variant deals with simply connected domains. Heuristically, a domain is simply
connected if it does not have any holes in it. (For example, in Figure 3.2(i) the domain D is
simply connected; however the domains D in Figures 3.2(ii) and (iii) are not simply connected
as they have holes in them.) More precisely:
Definition 3.8. A domain D is simply connected if for all closed contours γ in D and for all
z 6∈ D, we have w (γ; z) = 0. ✯
Theorem 3.9 (Cauchy’s Theorem for simply connected domains). Suppose that D is
a simply connected domain and f is a holomorphic function on D. Then for any closed
R
contour γ in D we have that γ f = 0.
More generally, it is sometimes useful to integrate around several closed contours.
Theorem 3.10 (The Generalized Cauchy Theorem). Let D be a domain and let f be
holomorphic on D. Let γ1 , . . . , γn be closed contours in D. Suppose that
w (γ1; z) + · · · + w (γn ; z) = 0 for all z 6∈ D.
Then Z Z
f + ··· + f = 0.
γ1 γn
Remark The hypotheses of the Generalized Cauchy Theorem (Theorem 3.10) give one way
of extending Cauchy’s Theorem to non-simply connected domains. Consider the example in
Figure 3.3. Here, if z is ‘outside’ D then clearly w (γ1 ; z) = w (γ2 ; z) = 0. If z is in the ‘hole’ in
D then w (γ1; z) = 1, w (γ2 ; z) = −1 so that w (γ1; z) + w (γ2 ; z) = 0. Hence the hypotheses of the
Generalized Cauchy Theorem hold. ❞
Proof of Theorem 3.10: Suppose that γr starts and ends at z j ∈ D, 1 ≤ j ≤ n. Choose

any z 0 ∈ D and contours σ1 , . . . , σn in D which join z 0 to z 1 , . . . , z n , respectively. (See Fig-
ure 3.4.)
2023–24
γ1
γ2
D
F IGURE 3.3: An example of closed contours that satisfy the hypotheses in the
Generalized Cauchy Theorem (Theorem 3.10).
Note that, for each j , σ j + γ j − σ j is a closed contour that starts and ends at z 0 and,
moreover, that for z 6∈ D we have w (σ j + γ j − σ j ; z) = w (γ j ; z). We see that
γ = σ1 + γ1 − σ1 + · · · + σn + γn − σn
is a closed contour that starts and ends at z 0 . Let z 6∈ D. Then, using Proposition 3.1,
w (γ; z) = w (σ1 + γ1 − σ1 + · · · + σn + γn − σn ; z)
Xn
= w (σ j + γ j − σ j ; z)
j =1
n
X
= w (γ j ; z)
j =1
= 0.
R
Hence by Cauchy’s Theorem γ f = 0. Hence
n
X
µZ Z Z ¶ n Z
X
f+ f+ f = f
j =1 σj γj −σ j j =1 γ j
R R
as −σ j f =− σj f. ❒
Before proceeding to Cauchy’s Integral Formula, we need an important technical result

about integration known as the Estimation Lemma, one that we will be using many times.
2023–24
58 § 3.3 The Estimation Lemma
γ2
σ2 γ1
zb 0
γ3 σ1
σ3
F IGURE 3.4: (For proof of Theorem 3.10.) The path γ is formed by starting at
z 0 , traversing σ1 , then around γ1 , then back along σ1 , then along σ2 , around
γ2 , back along σ2 , along σ3 , around γ3 and back along σ3 , ending at z 0 . In
symbols,
γ = σ1 + γ1 + σ− − −
1 + σ2 + γ2 + σ2 + σ3 + γ3 + σ3 , or
γ = σ1 + γ1 − σ1 + σ2 + γ2 − σ2 + σ3 + γ3 − σ3
3.3 The Estimation Lemma
There are two results about real integration that are obvious from considering the integral of
f (x) over [a, b] as the area underneath the graph of f . Firstly
¯Zb ¯ Zb
¯ ¯
¯
¯ f (x) dx ¯¯ ≤ | f (x)| dx (3.3)
a a
(see Figure 3.5), and secondly, if | f (x)| ≤ M then

¯Zb ¯
¯ ¯
¯
¯ f (x) dx ¯¯ ≤ M (b − a). (3.4)
a
See Figure 3.6.

Both of these results have analogies in the context of complex analysis. However, the
proofs are surprisingly intricate.
Here is the complex analogue of (3.3).
2023–24
+ + + +
+
−
F IGURE 3.5: If f [a, b] → R is negative on some subset of [a, b] then the area
underneath that part of the graph is negative. When f is replaced by | f |, this
area becomes positive.
a b
F IGURE 3.6: The graph of f is contained inside the rectangle of width b − a

and height M . Hence the area underneath the graph cannot be more than
M (b − a).
Lemma 3.11. Let a, b ∈ R and let φ : [a, b] → C be a continuous function. Then

¯Zb ¯ Zb
¯ ¯
¯
¯ φ(t ) dt ¯ ≤
¯ |φ(t )| dt . (3.5)
a a
Rb
Proof: First note that if a φ(t ) dt = 0 then the result is obvious. So assume it is non-zero
and write
Zb
φ(t ) dt = R ei θ
a
with R > 0. Then, since θ is a constant,

Zb
e−i θ φ(t ) dt = R. (3.6)
a
Now write e −i θ φ(t ) = u(t ) + i v(t ). Then by (3.6)

Zb Zb
u(t ) dt = R, and v(t ) dt = 0.
a a
2023–24
60 § 3.3 The Estimation Lemma
Moreover, u(t ) ≤ |e −i θ φ(t )| = |φ(t )|. Hence,

¯Zb ¯ Zb Zb
¯ ¯ ¯ ¯
¯
¯ φ(t ) dt ¯=R =
¯ u(t ) dt ≤ ¯φ(t )¯ dt
a a a
as required. ❒
We can now prove the following important result—the complex analogue of (3.4)—which
we will use many times in the remainder of the course. (Many results in Analysis are based
on a useful estimation like this.)
Lemma 3.12 (The Estimation Lemma). Let f : D → C be continuous and let γ be a contour
in D of length L. Suppose that | f (z)| ≤ M for all z on γ. Then
¯Z ¯
¯ ¯
¯ f ¯ ≤ M L.
¯ ¯
γ
(Some authors call this the ML Lemma for obvious reasons.)
Proof: This follows simply from Lemma 3.11 as follows:

¯Z ¯ ¯Zb ¯
′
¯ ¯ ¯ ¯
¯ f¯ = ¯
¯ ¯ ¯ f (γ(t ))γ (t ) dt ¯
¯
γ a
Zb
≤ | f (γ(t ))| |γ′ (t )| dt (by Lemma 3.11)
a
Zb
≤ M |γ′ (t )| dt
a
= M L.
Remark We use the Estimation Lemma in two different ways: either M is small or L is small.
That is, (i) suppose f is a function which takes small (in modulus) values along a contour γ,
R
then γ f is small; (ii) if f is any continuous function and γ is a contour with small length,
R
then γ f is small. ❞
Example 3.13. Let f (z) = 1/(z 2 + z + 1) and let γ(t ) = 5eRi t , 0 ≤ t ≤ 2π, be the circle of radius
5 centred at 0. We use the Estimation Lemma to bound γ f (z) dz.
2023–24
First note that if z is a point on γ then |z| = 5. Hence
|z 2 + z + 1| ≥ |z|2 − |z + 1| by the reverse triangle inequality

≥ |z|2 − |z| − 1 by the triangle inequality
= 25 − 5 − 1 = 19.
Thus for z on γ we have that

1 ¯ 1
¯ ¯
¯
| f (z)| = ¯¯ 2
¯≤ .
z + z + 1 ¯ 19
Next we note that L = length(γ) = 2π × 5 = 10π.

Thus, by the Estimation Lemma,
¯ f (z) dz ¯ ≤ 10π .
¯Z ¯
¯ ¯
¯
γ
¯ 19
3.4 Cauchy’s Integral Formula
One of the most remarkable facts in complex analysis is Cauchy’s Integral Formula.
In real analysis, we say that a function is C r if it can be differentiated r times and the r th
derivative is continuous. For example f (x) = x|x| is C 1 (the derivative is 2|x|), but not C 2 .
Then C 1 ⊃ C 2 ⊃ C 3 ⊃ · · · and we think of a function that is C r for a large r as being ‘nice’. If
we differentiate a C r function then we obtain a C r −1 function, i.e. differentiation takes nice
functions and makes them slightly ‘less nice’. Integration, however, works the other way: the
indefinite integral of a C r function is C r +1 . Hence integration makes nice functions ‘even
nicer’. In terms of complex analysis, this distinction into C r functions does not have any
meaning: as we shall see in the next chapter, if a function is differentiable once then it is
differentiable infinitely many times! That is, the complex world, C 1 = C 2 = C 3 = C 4 = · · · . And
ultimately, this follows from Cauchy’s integral formula.
Theorem 3.14 (Cauchy’s Integral Formula for a circle). Suppose that f is holomorphic
on the disc D(z 0 ; R). For 0 < r < R let C r be the path C r (t ) = z 0 + r e i t , 0 ≤ t ≤ 2π (so that
C r is the circle with centre z 0 and radius r ). Then for each w ∈ D(z 0 ; r ) we have that
1 f (z)
Z
f (w ) = dz. (3.7)
2πi Cr z −w
Equation (3.7) has the following remarkable consequence: if we know the value of the
holomorphic function f along the closed path C r then we know the values of the function at
all points inside the disc D r . This does not have any analogue for functions of a real variable.
2023–24
62 § 3.4 Cauchy’s Integral Formula
Remark Theorem 3.14 is formulated in terms of the function being holomorphic on a disc
and integrating around circles. This is not necessary, and a more general version of Cauchy’s
Integral Formula holds provided that f is holomorphic on a simply connected domain D and
we replace C r by any simple closed loop. (A closed loop γ is simple if, for every point z not on
γ, the winding number is either 0 or 1.) ❞
Proof: Fix w ∈ D(z 0 , r ). Consider the function
f (z) − f (w )
g (z) = .
z −w
Then g is differentiable in the domain D = {z ∈ C | |z − z 0 | < R, z 6= w }. Define the circle S ε

to be the circle centred at w and of radius ε > 0.
S ε (t ) = w + εe i t , 0 ≤ t ≤ 2π.
Then, provided ε > 0 is sufficiently small, both C r and S ε lie inside D.

We apply the Generalized Cauchy Theorem (Theorem 3.10) to the contours S ε and
−C r . Suppose that z is not in the domain D. Then either |z − z 0 | > R or z = w . In the first
case, if |z −z 0 | > R then w (S ε, z) = w (C r , z) = 0. In the second case, if z = w then w (S ε, z) =
1 and w (−C r , z) = −1. Hence we have that w (S ε, z) + w (−C r , z) = 0 for all z 6∈ D. Noting
R R
that −C r g = − C r g we have that, by the Generalized Cauchy Theorem (Theorem 3.10),
Z Z
g (z) dz = g (z) dz. (3.8)
Cr Sε
Now, from the definition of g , we have that limz→w g (z) = f ′ (w ). As | f ′ (w )| is finite, it

follows that g (z) is bounded for z sufficiently close to w , i.e. there exist δ > 0 and M > 0
such that if 0 < |w − z| < δ then |g (z)| < M .
Hence, if ε < δ, the Estimation Lemma (Lemma 3.12) implies that
¯Z ¯
¯ ¯
¯ g (z) dz ¯ ≤ M 2πε.
¯ ¯
Sε
By (3.8) it follows that

¯Z ¯
¯ ¯
¯
¯ g (z) dz ¯¯ ≤ M 2πε,
Cr
and since we can take ε > 0 to be arbitrarily small, it follows that

Z
g (z) dz = 0. (3.9)
Cr
2023–24
Recalling that g (z) = ( f (z) − f (w ))/(z − w ) and that f (w ) is a constant, we can substi-
tute this expression for g into (3.9) to obtain
f (z) f (w )
Z Z
dz = dz
Cr z − w Cr − w
z
1
Z
= f (w ) dz
Cr z − w
= f (w )2πi w (C r , w )
= f (w )2πi
as C r winds once anticlockwise around w . Hence

1 f (z)
Z
f (w ) = dz.
2πi C r z − w
❒
3.5 Exercises
Ex 3.1 Let f , g : D → C be holomorphic. Let γ be a smooth path in D starting at z 0 and ending

at z 1 . Prove the complex analogue of the integration by parts formula:
Z Z
′
f g = f (z 1 )g (z 1 ) − f (z 0 )g (z 0 ) − f ′ g .
γ γ
Ex 3.2 Consider the function f : C → C, f (z) = z 2 sinh z. Find an antiderivative and calculate
the integral along any smooth path from 0 to i π.
Ex 3.3 Prove Proposition 1.28(iv): Let D be a domain, γ a contour in D, and let f : D → C be

continuous. Let γ− denote the reverse path of γ. Show that
Z Z
f =− f.
γ− γ
Ex 3.4 Calculate (by eye) the winding number around every point which is not on the path.
Ex 3.5 Let
1
γ1 (t ) = −1 + e i t , 0 ≤ t ≤ 2π,
2
1
γ2 (t ) = 1 + e i t , 0 ≤ t ≤ 2π,
2
γ(t ) = 2e i t , 0 ≤ t ≤ 2π.
Let f (z) = 1/(z 2 − 1). Use the Generalized Cauchy Theorem to deduce that
Z Z Z
f dz = f dz + f dz.
γ γ1 γ2
2023–24
64 Exercises
F IGURE 3.7: See Exercise 3.4.
Ex 3.6 Let γ1 denote the unit circle centred at 0, radius 1, oriented anti-clockwise. Let f (z) =
R
1/z. Show that γ1 f = 2πi . Let γ2 be the closed contour as illustrated in Figure 3.8. Use
R
the Generalized Cauchy Theorem on an appropriate domain to calculate γ2 f .
Ex 3.7 Let z 1 , z 2 be two distinct points in C and let D be the domain C\{z 1 , z 2 }. Suppose that f
is holomorphic in D and that γ, γ1 , γ2 are closed contours in D as illustrated in Figure 3.9.
Suppose that Z Z
f = 3 + 4i , f = 5 + 6i .
γ1 γ2
R
Use the Generalized Cauchy Theorem to calculate γ f.
2023–24
γ2
γ1
F IGURE 3.8: Here γ1 denotes the unit circle oriented anticlockwise and γ2 is
an arbitrary closed contour that winds once around 0 (see Exercise 3.6).
Ex 3.8 (Sometimes one can use Cauchy’s Integral formula even in the case when f is not
holomorphic: the trick is to replace f by a holomorphic function g which agrees with f
on the contour of integration—hence having the same integral.)
Let f (z) = |z + 1|2 . Let γ(t ) = e i t , 0 ≤ t ≤ 2π be the path that describes the unit circle with
centre 0 and oriented anticlockwise.
(i) Show that f is not holomorphic on any domain in C. [Hint: use the Cauchy-Riemann
Theorem.]
(ii) Find a function g that is holomorphic on some domain that contains γ and such that
f (z) = g (z) at all points on the unit circle γ. [Hint: recall that if w ∈ C then |w|2 = w w̄ .]
(iii) Use Cauchy’s Integral formula to show that
Z
|z + 1|2 dz = 2πi .
γ
2023–24
66
γ1 γ
z1 γ2
z2
F IGURE 3.9: See Exercise 3.7.

Chapter 4
Taylor series and Laurent series
In this chapter we consider two types of power series. The first is the power series of a holo-
morphic function—that is, given a holomorphic function, how do we find its power series
about any given point in its domain. The answer is (to some extent) like the real case: we
can find its Taylor series; however, it differs from the real case in two surprising ways: firstly
the coefficients can be found not only by differentiating the function at the point in question,
but also by integrating it along a contour around the point—this is based on Cauchy’s integral
formula from the previous chapter. The other difference with the real case is that for a holo-
morphic function, the Taylor series always converges to the function—for real differentiable
functions this is often the case but not always (see Remark ?? below).
The second type of power series allows for functions that have singularities (in particular
points where it tends to infinity), such as f (z) = 1/z at z = 0. This function of course doesn’t
have a Taylor series—it is not holomorphic in any neighbourhood of 0. Instead we allow
negative powers of z as well as positive powers. An example would be
1 1 1 7 3
= + z+ z +··· ,
sin z z 6 360
in a punctured neighbourhood of 0. The coefficients of these series cannot be calculated
by differentiation at the point since the function is not even defined there, let alone differ-
entiable. But it can be calculated using the integral approach based on Cauchy’s integral
formula, as we see below. These series with negative powers are called Laurent series.
4.1 Taylor series
As we’ll see, Cauchy’s integral formula allows us to express any differentiable function as a
power series (its Taylor series expansion). One of the most remarkable properties of complex
differentiable functions then follows from Theorem 2.19: if f is differentiable once then it is
differentiable arbitrarily many times.
67
68 § 4.1 Taylor series
Theorem 4.1 (Cauchy-Taylor Theorem for holomorphic functions). Suppose that f ∈

O(D) for some domain D. Then all of the higher derivatives of f exist in D and, on any
disc D(z 0 ; R) ⊂ D, f has a Taylor series expansion given by
∞
a n (z − z 0 )n ,
X
f (z) =
n=0
1
where a n = n!
f (n) (z 0 ). Furthermore, if 0 < r < R and C r (t ) = z 0 + r e i t , 0 ≤ t ≤ 2π, then
1 f (z)
Z
an = dz.
2πi Cr (z − z 0 )n+1
Note here the crucial fact: f (z) is equal to its Taylor series. This is not true in general
in real analysis: there one often needs a remainder term. A well-known example (also first
recognized by Cauchy!) of a real function that is differentiable an arbitrary number of times
but not equal to its Taylor series is given by
2
e −1/x , x 6= 0
½
f (x) =
0, x = 0.
One can check (by induction) that f (n) (0) = 0 for all n, so the Taylor series of f at 0 is identi-
cally 0. As f 6= 0 near 0, it follows that the Taylor series does not converge to f on any interval
(−ε, ε).
A second difference with the real Taylor series is that, even if its Taylor series does con-
verge to f , the coefficients are not determined by (real) integrals.
Definition 4.2. If, for each z 0 ∈ D, a function f : D → C is equal to its Taylor series at z 0 on
some open set containing z 0 then one says that f is analytic. ✯
It follows from Theorem 4.1 that all holomorphic functions are analytic; however the ex-
ample above shows that not all infinitely real-differentiable functions are analytic. It also
follows that if f is differentiable once, it is differentiable arbitrarily many times.
Proof of Theorem 4.1: First recall that for any w ∈ C we have
1 − w m+1
1+ w +··· + wm = .
1−w
Put w = h/(z − z 0 ) where h ∈ C (the reader might like to put z 0 = 0 for a first reading of this
2023–24
TAYLOR SERIES AND L AURENT SERIES 69
proof). Then
µ ¶m+1
h
1−
h hm z − z0
1+ +··· + =
z − z0 (z − z 0 )m h
1−
z − z0
µ µ ¶m+1 ¶
h
1−
z − z0
= × (z − z 0 ).
z − z0 − h
If we divide through by (z − z 0 ) and rearrange, we obtain, for z 6= z 0 ,
1 1 1 h hm h m+1
= = + + · · · + + .
z − (z 0 + h) (z − z 0 ) − h z − z 0 (z − z 0 )2 (z − z 0 )m+1 (z − z 0 )m+1 (z − z 0 − h)
Fix h such that 0 < |h| < R and suppose, for the moment, that |h| < r < R. Then
Cauchy’s Integral formula, together with the above identity, gives
1 f (z)
Z
f (z 0 + h) = dz
2πi C r z − (z 0 + h)
1 1 hm
µ
h
Z
= f (z) + + · · · +
2πi C r z − z 0 (z − z 0 )2 (z − z 0 )m+1
h m+1
¶
+ dz
(z − z 0 )m+1 (z − z 0 − h)
m
an h n + Am .
X
=
n=0
where
1 f (z)
Z
an = dz
2πi Cr (z − z 0 )n+1
and
h m+1 f (z)
Z
Am = dz.
2πi Cr (z − z 0 )m+1 (z − z 0 − h)
We show that A m → 0 as m → ∞.
As f is differentiable on C r , it is bounded. So there exists M > 0 such that | f (z)| ≤ M
for all z on C r .
By the reverse triangle inequality, using the facts that |h| < r = |z − z 0 | for z on C r , we
have that
|z − z 0 − h| ≥ ||z − z 0 | − |h|| = r − |h|.
Hence, by the Estimation Lemma (Lemma 3.12)
1 M |h|m+1 M |h| |h| m

µ ¶
|A m | ≤ 2πr = .
2π r m+1 (r − |h|) r − |h| r
2023–24
70 § 4.1 Taylor series
Since |h| < r , this tends to zero as m → ∞. Hence

∞
an h n
X
f (z 0 + h) =
n=0
for |h| < R with

1 f (z)
Z
an = dz
2πi Cr (z − z 0 )n+1
provided that r satisfies |h| < r < R. However, the integral is unchanged if we vary r in the
whole range 0 < r < R. Hence this formula is valid for the whole of this range of r .
Finally, we put h = z − z 0 . Then we have that
∞
a n (z − z 0 )n
X
f (z) =
n=0
for |z − z 0 | < R, with a n given as above. From Theorem 2.19 we know that a power series
can be differentiated term-by-term as many times as we please and that
f (n) (z 0 )
an = .
n!
One property of the Taylor series of a function is that it is the only power series that con-
verges to the given function in a neighbourhood of the given point. This is a consequence of
the uniqueness of power series, as stated in Proposition 2.22.
This fact sometimes allows us to calculate Taylor series without using the formula for the
coefficients given in Theorem 4.1.
Example 4.3. We can find the Taylor series for f (z) = sin2 z as follows.
Recall that sin2 z = (1 − cos 2z)/2, and that cos z = n 2n
P∞
n=0 (−1) z /(2n)! (which converges
for all z ∈ C). Hence
1
sin2 z = (1 − cos 2z)
2
∞ 2n ¶
1 n (2z)
µ
X
= 1− (−1)
2 n=0 (2n)!
2z 2 23 z 4 2n−1 2n
= − + · · · + (−1)n+1 z +··· .
2! 4! (2n)!
As this is a power series that is equal to f (z) and is valid for all z ∈ C, by Proposition 2.22 this
must be the Taylor series of f on C.
2023–24
4.2 Applications of the Cauchy-Taylor theorem
We’ve seen one application of the Integral Formula part of the Cauchy-Taylor theorem above
(namely, that every holomorphic function is analytic) but it has many more applications; here
we give just five (‘just’?).
4.2A Cauchy’s Estimate
The theorem gives us a relation between the derivatives of a function and Cauchy-style inte-
grals. As a consequence we can apply the Estimation Lemma to obtain a useful estimate.
Lemma 4.4 (Cauchy’s Estimate). Suppose that f is holomorphic on the disc D(z 0 ; R). If
0 < r < R and | f (z)| ≤ M for all z such that |z − z 0 | = r then, for all n ≥ 0,
Mn!
| f (n) (z 0 )| ≤ .
rn
Proof: By Theorem 4.1 we know that
n! f (z)
Z
f (n) (z 0 ) = dz,
2πi Cr (z − z 0 )n+1
where C r is the circle of radius r and centre z 0 . By the Estimation Lemma (Lemma 3.12),
f (z)
¯Z ¯
(n) n! ¯¯ ¯
| f (z 0 )| = dz ¯¯
2π C r (z − z 0 )
¯ n+1
n! M
≤ 2πr
2π r n+1
Mn!
= .
rn
❒
4.2B Radius of convergence
1
Recall that the Taylor series at the origin of f (z) = 1−z is 1 + z + z 2 + z 3 + · · · . We know that the
radius of convergence in this case is R = 1: it obviously couldn’t be any larger since for z = 1
the function itself is infinite, and so the power series will diverge there.
Similarly, the power series for Log(1 + z) = z − 21 z 2 + 13 z 3 − · · · has radius of convergence
R = 1, and couldn’t be any greater than 1 because the function is infinite at z = −1.
2023–24
72 § 4.2 Applications of the Cauchy-Taylor theorem
Theorem 4.5. Let f ∈ O(D), for some domain D, and let z 0 ∈ D. Let ρ be the distance
from z 0 to C \D (and if D = C then ρ = ∞). Then the Taylor series for f about z 0 converges
to f in D(z 0 , ρ).
In other words, the radius of convergence of the Taylor series is always as large as it can
possibly be, see Figure 4.1.
b z0
F IGURE 4.1: For f ∈ O (D), the Taylor series about z 0 converges inside the disc
shown.
Examples 4.6. (i). If f is entire then R = ∞ (as for example, with f (z) = exp(z)).
(ii). Let f (z) = 1/z. Then D = C \ {0}. The Taylor series about z = 2i will have radius of
convergence R = 2 (as 2 is the distance between 0 and 2i ).
Example 4.7. A common mistake is to think that f (z) (or its Taylor series) must have an
infinite value somewhere on the boundary of the disc of convergence. That this is not the
case is shown by the power series
X∞ zn
f (z) = 2
.
n=1 n
This has R = 1, but for any z = ei θ , the sum ei nθ /n 2 converges absolutely (because (1/n 2 )
P P
converges). However, as soon as |z| > 1 the series diverges.
Proof: Let the Taylor series of f at z 0 be

∞
a n (z − z 0 )n ,
X
f (z 0 ) =
n=1
2023–24
and denote its radius of convergence by R. We wish to show R = ρ. We know R ≤ ρ as

explained before the theorem.
Choose any r > 0 such that r < ρ. Then f is finite on the circle C r and holomorphic in
D(z 0 , r ). By Cauchy’s estimate above, it follows that
1 (n) M
|a n | = f ≤ n.
n! r
Thus
M 1/n
|a n |1/n ≤ .
r
If M = 0 then f is identically 0 and there is nothing to prove. If M > 0 then limn→∞ M 1/n =
1, and hence
M 1/n 1
lim sup |a n |1/n ≤ lim sup = .
n→∞ n→∞ r r
It follows from the Cauchy-Hadamard formula that the radius of convergence of the power
series satisfies R ≥ r . But this holds for any r < ρ and hence R ≥ ρ as required. ❒
4.2C Liouville’s Theorem
Theorem 4.8 (Liouville’s Theorem). Suppose that the entire holomorphic function f is
bounded on the whole of C. Then f is constant.
By bounded we mean that there exists M > 0 such that | f (z)| ≤ M for all z ∈ C.
Remark This theorem has no analogue in real analysis. It is easy to think of functions
f : R → R that are differentiable and bounded but not constant. (For example f (x) = sin x.) ❞
Proof: Choose M such that | f (z)| ≤ M for all z ∈ C. Let z 0 ∈ C. Since f is holomorphic
on the whole of C, it is holomorphic in the disc D(z 0 ; R) of radius R centred at z 0 for R as
large as we please. By Cauchy’s Estimate (Lemma 4.4), we have for 0 < r < R
M
| f ′ (z 0 )| ≤ .
r
Since we can choose R as large as we please, so we can choose r as large as we please.

Hence we can let r → ∞. Hence f ′ (z 0 ) = 0 for every z 0 ∈ C, whence f is constant. ❒
2023–24
4.2D Uniqueness of holomorphic functions
We begin with a property of holomorphic functions (one which is also true for real analytic
functions). The fact that domains are connected is essential here as we see in this first lemma.
Lemma 4.9. Let D be a domain, and suppose U ⊂ D is both open and closed. Then either
U = ; or U = D.
Proof: Recall that a map is continuous if and only if the inverse image of every open set
is open. Equivalently, the inverse image of every closed set is closed.
Suppose for a contradiction that U is non-empty and not equal to D. Let V = D \U .
Then both U and V are open and non-empty. Let z 1 ∈ U and z 2 ∈ V . By the hypothesis
that V is a domain, there is a continuous path (in fact a polygonal arc) γ : [a, b] → D such
that γ(a) = z 1 , γ(b) = z 2 .
Now, since γ is continuous and U is open then γ−1 (U ) ⊂ [a, b] is open. And since V is
open γ−1 (V ) ⊂ [a, b] is open.
It follows that [a, b] is the union of two disjoint open intervals which is not possible!
(To see this consider inf(γ−1 (V )) = inf{t ∈ [a, b] | γ(t ) ∈ V } and show it is neither in γ−1 (U )
nor γ−1 (V ).) ❒
Proposition 4.10. Let f be a holomorphic function defined on a domain D. If any one

of the following are true then f is identically zero on D.
(i). There is a point z 0 ∈ D at which f and all its derivatives vanish (i.e., f (n) (z 0 ) = 0 for
all n ≥ 0).
(ii). There is an open subset U ⊂ D for which f U

= 0 ( f vanishes identically on U ). Then
f = 0.
(iii). There is a convergent sequence of distinct points z n → z 0 in D such that f (z n ) = 0

for all n.
Part (iii) often arises by having a line or curve in the plane where the function vanishes: it
is easy to extract a convergent sequence of points on a line or curve.
Proof: Let Z ⊂ D be the set of points at which f and all its derivatives vanish.
Since all these functions f (n) are continuous it follows that Z is closed in D.
Now consider a point w ∈ Z . The Taylor series of f at w is therefore zero. By Taylor’s
theorem, there is an ε > 0 such that f is identically 0 on the (open) disc D(w, ε). If follows
that D(w, ε) ⊂ Z , and hence Z is open.
2023–24
Since Z is both open and closed, we either have Z = ; or Z = D (the case Z = D saying
that f is identically 0).
We argue in each case that Z 6= ;.
(i) This states z 0 ∈ Z .
(ii) Since f = 0 on U then at any z 0 ∈ U all derivatives of f vanish.
(iii) Suppose f has a zero of order m ≥ 0 at z 0 and consider the Taylor series of f at z 0 :
f (z) = a m (z − z 0 )m + a m+1 (z − z 0 )m+1 + a m+2 (z − z 0 )m+2 + · · ·
in a disc D(z 0 ; ε). Dividing through by (z − z 0 )m we have a new function
g (z) = a m + a n+1 (z − z 0 ) + a n+1 (z − z 0 )2 + · · · .
also defined on D(z 0 ; ε).

Suppose N is large enough that n > N implies z n ∈ D(z 0 ; ε) (which exists by virtue
of being a convergent sequence). We have f (z n ) = 0. Since (z 0 − z n ) 6= 0 it follows that
g (z n ) = 0 for all n > N . By continuity of g it follows that g (z 0 ) = 0. That is, a m = 0, which
is a contradiction. Therefore z 0 ∈ Z and so Z 6= ;. ❒
In previous chapters we defined complexifications of well-known real functions such as

sin and e x using their power series expansions. The following corollary shows that these are
the only possible complexifications.
Corollary 4.11. Let f and g be holomorphic functions defined on C such that f R

= g R.
Then f = g .
Proof: Apply Proposition 4.10(iii) to f − g . ❒
An example of this would be another proof that exp(w + z) = exp(w ) exp(z) since we al-
ready know this is true for real arguments.
4.2E The Fundamental Theorem of Algebra
Consider the equation x − n = 0 where n ∈ N. This equation always has solutions x ∈ N (in-
deed, x = n). If, however, we consider x + n = 0, n ∈ N, then we need to introduce negative
integers to be able to solve this equation. More generally, consider the equation p x − q = 0
where p, q ∈ Z; then we need to introduce rational numbers Q to be able to solve this equa-
tion. Continuing this theme, one can see that one needs to introduce surds (to solve x 2 −2 = 0)
and complex numbers (to solve x 2 +1 = 0). Let us ask the ultimate question along these lines:
if we have a polynomial equation where the coefficients are complex numbers, do we need
to invent a larger class of numbers to be able to solve this equation or will complex numbers
suffice? The answer is that complex numbers are indeed sufficient.
2023–24
Theorem 4.12 (The Fundamental Theorem of Algebra). Let p(z) = z n +a n−1 z n−1 +· · ·+
a 1 z + a 0 be a polynomial of degree n ≥ 1 with coefficients a j ∈ C. Then there exists α ∈ C
such that p(α) = 0.
Corollary 4.13. Let p(z) = z n + a n−1 z n−1 + · · · + a 1 z + a 0 be a polynomial of degree n ≥ 1

with coefficients a j ∈ C. Then we can factorise p(z) into linear factors: there exist α j ∈ C,
1 ≤ j ≤ n such that
n
Y
p(z) = (z − α j ).
j =1
This theorem has many proofs all naturally involving complex analysis. Here we present
one based on Liouville’s theorem.
Proof of Theorem 4.12: Suppose for a contradiction that there are no solutions to p(z) =
0, i.e. suppose that p(z) 6= 0 for all z ∈ C.
If p(z) 6= 0 for all z ∈ C then 1/p(z) is holomorphic for all z ∈ C. We shall show that
1/p(z) is bounded and then use Liouville’s theorem to show that p is constant.
For z 6= 0
p(z) a n−1 a1 a0
n
= 1+ + · · · + n−1 + n → 1
z z z z
as |z| → ∞. Hence there exists K > 0 such that if |z| > K then
¯ p(z) ¯ 1
¯ ¯
¯ zn ¯ ≥ 2 .
¯ ¯
Re-arranging this implies that for |z| > K we have that

¯ 1 ¯ 2 2
¯ ¯
¯ p(z) ¯ ≤ |z n | ≤ K n .
¯ ¯
Hence 1/p(z) is bounded if |z| > K .

We shall show that this bound continues to hold for |z| ≤ K . Let z 0 ∈ C, |z 0 | ≤ K .
Let C r (t ) = z 0 + r e i t , 0 ≤ t ≤ 2π, denote the circular path with centre z 0 and radius r .
By choosing r sufficiently large, we can assume that C r is contained in {z ∈ C | |z| > K }.
Hence, for such an r , if z is any point on C r then |z| > K . Hence if z is any point on C r
then |1/p(z)| ≤ 2/K n . By Cauchy’s Estimate (Lemma 4.4) it follows that
¯ 1 ¯ 2
¯ ¯
¯ p(z ) ¯ ≤ K n .
¯ ¯
0
Hence |1/p(z)| ≤ 2/K n for all z ∈ C, so that p is a bounded holomorphic function on

C. By Liouville’s Theorem (Theorem 4.8), this implies that p is constant, a contradiction.
❒
2023–24
Proof of Corollary 4.13: Let p(z) be a degree n polynomial with coefficients in C. By

Theorem 4.12 we can find α1 ∈ C such that p(α1 ) = 0. Write p(z) = (z −α1 )q(z) where q(z)
is a degree n − 1 polynomial with coefficients in C. The proof then follows by induction
on n. ❒
We have already seen that a holomorphic function f can be expressed as a Taylor series:
i.e. if f is differentiable on a domain D and z 0 ∈ D then we can write
∞
a n (z − z 0 )n
X
f (z) = (4.1)
n=0
for suitable coefficients a n , and this expression is valid for z such that |z − z 0 | < R, for some
R > 0. The idea of Laurent series is to generalize (4.1) to allow negative powers of (z −z 0 ). This
turns out to be a remarkably useful tool.
4.3 Laurent series
Definition 4.14. A Laurent series is a series of the form

∞
a n (z − z 0 )n ,
X
(4.2)
n=−∞
which has negative as well as possibly positive powers. ✯
As (4.2) is a doubly infinite sum, we need to take care as to what it means. To define (4.2)
we split it into two sums:
∞ ∞
a −n (z − z 0 )−n + a n (z − z 0 )n = Σ− + Σ+ .
X X
n=1 n=0
This Σ− has all the negative powers of (z − z 0 ), and Σ+ the constant term and all the positive
powers.
The first question to address is when does (4.2) converge? For this, we need both Σ− and
+
Σ to converge.
Now Σ+ converges for |z − z 0 | < R 2 for some R 2 ≥ 0, where R 2 is the radius of convergence
of Σ+ .
We can recognise Σ− as a power series in (z −z 0 )−1 . This has a radius of convergence equal
to, say, R 1−1 ≥ 0. That is, Σ− converges when |(z − z 0 )−1 | < R 1−1. In other words, Σ− converges
when |z − z 0 | > R 1 .
Combining these, we see that if 0 ≤ R 1 < R 2 ≤ ∞ then (4.2) converges on the set
{z ∈ C | R 1 < |z − z 0 | < R 2 }.
2023–24
78 § 4.3 Laurent series
R1
z0
b
R2
F IGURE 4.2: The annulus A(z 0 ; R 1 , R 2 ) in C with centre z 0 and radii R 1 < R 2 .
See Figure 4.2. Such a set is called an annulus1 , and we will denote it
A(z 0 ; R 1 , R 2 ).
Notice that if R 1 = 0 then A(z 0 ; 0, R 2 ) = D ∗ (z 0 , R 2 )—the punctured disc.

The following theorem says that given a function f that is holomorphic on an annulus
then it can be expressed as a Laurent series (similar to Taylor’s theorem for a holomorphic
function on a disc). Moreover, we can obtain an expression for the coefficients a n in terms of
the function f .
Theorem 4.15 (Laurent’s theorem). Suppose that f is holomorphic on the annulus
A(z 0 ; R 1 , R 2 ) = {z ∈ C | R 1 < |z − z 0 | < R 2 },
where 0 ≤ R 1 < R 2 ≤ ∞. Then we can write f as a Laurent series: on A(z 0 ; R 1 , R 2 ) we have

∞ ∞
a n (z − z 0 )n + a −n (z − z 0 )−n
X X
f (z) = (4.3)
n=0 n=1
Moreover, let R 1 < r < R 2 and let C r (t ) = z 0 + r e i t , 0 ≤ t ≤ 2π be the circular path with
centre z 0 and radius r . Then, for n ∈ Z,
1 f (z)
Z
an = dz (4.4)
2πi Cr (z − z 0 )n+1
Note that in this case we cannot conclude that a n = f (n) (z 0 )/n! as we do not know that f is
differentiable at z 0 (indeed, it may not even be defined at z 0 ): so integration is more powerful
than differentiation.
1 annulus is Latin for ring: the plural is annuli (not annuluses)
2023–24
We omit the proof — it is similar to the proof of Taylor’s Theorem and details can be found
in Stewart and Tall’s book (Theorem 11.1).
We call the series (4.3) the Laurent series of f (z) about z 0 or the Laurent expansion of
f (z).
We call
−1
a n (z − z 0 )n
X
n=−∞
the principal part of the Laurent series. Thus the principal part of a Laurent series is the part
that contains all the negative powers of (z − z 0 ) (which we denoted Σ− earlier).
The following result tells us that the coefficients in the Laurent series expansion are uniquely
determined (compare with the analogous result for ordinary power series in Proposition 2.22).
Proposition 4.16 (Uniqueness of Laurent series). Suppose that

∞ ∞
a n (z − z 0 )n = b n (z − z 0 )n
X X
(4.5)
n=−∞ n=−∞
for all z ∈ C such that R 1 < |z − z 0 | < R 2 . Then a n = b n for all n ∈ Z.
Proof: Let
∞ ∞
a n (z − z 0 )n = b n (z − z 0 )n
X X
f (z) =
n=−∞ n=−∞
on A(z 0 ; R 1 , R 2 ). Then Laurent’s theorem states that
1 f (z)
Z
an = dz
2πi Cr (z − z 0 )n+1
and
1 f (z)
Z
bn = dz,
2πi Cr (z − z 0 )n+1
where C r is a circular path with centre z 0 and radius r (with R 1 < r < R 2 ), described once
and oriented anticlockwise.
But the right hand sides are the same, so therefore a n = b n for all n. ❒
Warning: A common mistake is to find the coefficients of the Laurent series of a function
f (z) about say z = 0, and assume that the same coefficients can be used about some other
point z 0 6= 0. This is emphatically not the case (it is already not the case for Taylor series in real
analysis: the coefficients of f (x) = sin(x) about x = 0 and about x = π/2 are not the same). We
shall see some specific examples of this in the context of Laurent series below. "
2023–24
4.3A Calculating Laurent series
Given a specific function f that is holomorphic on an annulus, we want to calculate the Lau-
rent series of f ; that is, we want to be able to calculate the coefficients a n . If we were to appeal
directly to Theorem 4.15 we would have to evaluate the integral in (4.4). In general, this is dif-
ficult or time-consuming (Exercise 5.9 in the next chapter leads you through one example of
this). Instead, we can appeal to Proposition 4.5: given a function f that is holomorphic on an
annulus R 1 < |z − z 0 | < R 2 , if we can find an expression of the form (4.2) that is equal to f on
this annulus then it must be the Laurent series.
Example 4.17. Let f (z) = e z + e 1/z . Recall that e z = n

/n! for all z ∈ C. Hence e 1/z =
P∞
P∞ n=0 z
−n
n=0 z /n! for all z 6= 0. Hence
∞ 1 1 1 z2 zn
an z n = · · · +
X
f (z) = + · · · + + + 2 + z + + · · · + +··· .
n=−∞ n!z n 2!z 2 z 2! n!
Thus,
1
a 0 = 2, and a ±n = for n ≥ 1,
n!
This expansion if valid for all z 6= 0, i.e. on D ∗ (0; ∞) = A(0; 0, ∞).
Example 4.18. Let

1
f (z) =
z(1 − z)
and let us calculate the Laurent series at z 0 = 0.
There are two possible approaches.
First write f using partial fractions:
1 1
f (z) = +
z 1−z
Now 1/z is already a Laurent series at 0 (the only non-zero coefficient is a −1 = 1). Note that
this converges if z 6= 0 (in this case, as there is only one term, checking convergence just
means checking when this formula makes sense!).
n
Now, by summing a geometric progression, we have that 1/(1 − z) = ∞ n=0 z and this
P
power series converges for |z| < 1.

Hence f (z) has Laurent series
1 ∞
+ 1 + z + z2 + z3 + · · · = zn
X
f (z) = (4.6)
z n=−1
2023–24
1
and this expression is valid on the annulus {z ∈ C | 0 < |z| < 1}. The term z
is the principal
part of this Laurent series.
The second approach is to write
1 1
µ ¶
f (z) = .
z 1−z
Then, using the geometric progression for the term in brackets,
1¡
1 + z + z2 + z3 + · · ·
¢
f (z) =
z
which expands to give the same Laurent series as (4.6).
Example 4.19. Let

1 1 1
f (z) = = −
(z − 1)(z − 2) z −2 z −1
We will expand f as three different Laurent series about z 0 = 0, valid in three different annuli.
First note that we can write
1 −1 ∞
zn
X
= =− (4.7)
z −1 1−z n=0
(summing a geometric progression) and that this is valid for |z| < 1. We can also write
1 1 1 1 X∞ ∞ 1
−n
X
= ¡ = z =
z − 1 z 1 − 1z n
¢
z n=0 n=1 z
by again noting that ∞ −n

= 1/(1 − z −1 ) is the sum to infinity of a geometric progression
P
n=0 z
with common ratio z −1 . This converges for |z −1 | < 1, i.e. |z| > 1. Hence
1 ∞ 1
X
= (valid for |z| > 1). (4.8)
z − 1 n=1 z n
Similarly, we can write
1 −1 1
µ
1
¶
1 X∞ ³ z ń
= =− z =− (4.9)
z −2 2−z 2 1− 2 2 n=0 2
n
by noting that ∞ n=0 (z/2) = 1/(1 − z/2) is the sum of a geometric progression with common
P
ratio z/2. This expansion is valid when |z/2| < 1, i.e. when |z| < 2.
2023–24
We can also write
1 1 1 1 X∞ ³ z ´−n X∞ 2n−1
= ¡ 2
¢= = (4.10)
z −2 z 1− z z n=0 2 n=1 z
n
by recognising the middle term as the sum of a geometric progression with common ratio
(z/2)−1 . This converges when |(2/z)−1 | < 1, i.e. when |z| > 2.
Using (4.7) and (4.9) we see that we can expand
∞ 1 X∞ ³ z ń ∞ µ 1
¶
n
1 − n+1 z n
X X
f (z) = z − =
n=0 2 n=0 2 n=0 2
and this is valid on the annulus (disc) D(0; 1) = {z ∈ C | 0 ≤ |z| < 1}.
Using (4.8) and (4.9) we can expand
1 X∞ ³ z ń X∞ 1
f (z) = − −
2 n=0 2 n=1 z
n
1 1 1 z zn
= ···+ + · · · + + + + · · · + +···
zn z 2 22 2n+1
and this is valid on the annulus A(0; 1, 2) = {z ∈ C | 1 < |z| < 2}.
Using (4.8) and (4.10) we can expand
X∞ 2n−1 X∞ 1
f (z) = − n
+ n
n=1 z n=1 z
X∞ 1 − 2n−1
=
n=1 zn
and this is valid on the annulus A(0; 2, ∞) = {z ∈ C | 2 < |z| < ∞}.
In the above examples we have expanded functions as Laurent series on annuli centred at
the origin. If we want to expand a function f (z) as a Laurent series on an annulus centred at
z 0 then it is often convenient to first change coordinates to w = z − z 0 , calculate the Laurent
series centred at w = 0 and in terms of w , and then change coordinates back to z.
Example 4.20. Let

ez
f (z) = .
(z − 1)2
We will expand f as a Laurent series on the annulus A(1; 0, ∞) = {z ∈ C | 0 < |z − 1| < ∞}.
We first change coordinates and let w = z − 1. Then z = 1 + w and we are interested in
expanding
e 1+w
.
w2
2023–24
Now e 1+w = ee w = e n
P∞
n=0 w /n!. Hence
e 1+w e X∞ wn e e e e e e
2
= 2
= 2 + + + w + w 2 + · · · + w n−2 + · · · ,
w w n=0 n! w w 2! 3! 4! n!
and this is valid provided that w 6= 0. Changing coordinates back to z we obtain

e e e e e e
f (z) = + + + (z − 1) + (z − 1)2 + · · · + (z − 1)n−2 + · · · ,
(z − 1)2 z − 1 2! 3! 4! n!
valid for z 6= 1, i.e. on {z ∈ C | 0 < |z − 1|}.
4.4 Exercises
Ex 4.1 Find the Taylor expansion of the following functions around 0 and find the radius of
convergence:
2
(i) cos2 z, (ii) (2z + 1)−1 , (iii) f (z) = e z .
Ex 4.2 Calculate the Taylor series expansion of Log(1 + z) around 0 and determine its radius
of convergence.
Ex 4.3 Let f , g be two holomorphic functions defined on a domain D. Suppose there is a

sequence of distinct points z n converging to z 0 in D for which f (z n ) = g (z n ). Deduce that
f = g on D.
Ex 4.4 Show that every polynomial p of degree at least 1 is surjective (that is, for all a ∈ C
there exists z ∈ C such that p(z) = a).
Ex 4.5 Suppose that f is holomorphic on the whole of C and suppose that | f (z)| ≤ K |z|k for
some real constant K > 0 and some positive integer k ≥ 0. Prove that f is a polynomial
function of degree at most k.
[Hint: Calculate the coefficients of z n , n ≥ k , in the Taylor expansion of f around 0.]
Ex 4.6 Let f ∈ O (D), and suppose z 0 ∈ D is a zero of f which is not isolated. Show that f is
identically 0. [Hint: Use Proposition 4.10]
1
Ex 4.7 Let f (z) = .
1 + z2
(i) Write down the power series for f obtained as a geometric progression.
(ii) Express f in (complex) partial fractions, and use Taylor’s theorem to rederive the Taylor
series about z = 0. (You should of course obtain the same answer.)
Ex 4.8 The Fibonacci sequence F n satisfies the recurrence relation
F n+2 = F n+1 + F n , (∗)

2023–24
84 Exercises
with F 0 = 0, F 1 = 1.
n
(i) Show that the series ∞ n=0 F n z converges to a rational function f (z).
P
[Hint: Multiply (∗) by z n+2

and sum, being careful of the range of the sum.]
(ii) Show that the rational function f is holomorphic on D = C \ {−ϕ, ϕ−1 }, where ϕ is the
p
Golden Mean: ϕ = 12 ( 5 + 1).
(iii) Apply Theorem 4.5 to obtain the radius of convergence of this power series.
(iv) Apply Taylor’s theorem to show that
ÃÃ p !n Ã p !n !
1 1+ 5 1− 5
Fn = p − .
5 2 2
Ex 4.9 Find the Laurent expansions of the following around z = 0:
(i) (z − 3)−1 , valid for 3 < |z| < ∞;

(ii) 1/(z(1 − z)), valid for 0 < |z| < 1;
(iii) z 3 e 1/z , valid for 0 < |z| < ∞;
(iv) cos(1/z), valid for 0 < |z| < ∞.
Ex 4.10 Find Laurent expansions for the function

1 1
f (z) = + .
z +1 z −3
valid on the annuli
(i) 0 ≤ |z| < 1, (ii) 1 < |z| < 3, (iii) 3 < |z| < ∞.
Ex 4.11 Let
1
f (z) = .
z 2 (z − 1)
(i) Find a Laurent series expansion for f valid for 0 < |z| < 1.
(ii) Find a Laurent series expansion for f valid for 0 < |z − 1| < 1.
[Hint: introduce w = z − 1 and recall that 1/(1 − w)2 = ∞ n−1
, provided that |w| < 1.]
P
n=1 nw
Ex 4.12 Let f (z) = (z − 1)−2 . Find Laurent series for f valid on the following annuli:
(i) {z ∈ C | 0 < |z − 1| < ∞},

(ii) {z ∈ C | 0 ≤ |z| < 1},
(iii) {z ∈ C | 1 < |z| < ∞}.
Ex 4.13 Let D ⊂ C be a domain, and f : D → C be a continuous function. Suppose that for

every closed path γ in D we know Z
f = 0.
γ
Show that f ∈ O (D). (This is called Morera’s theorem.)
[Hint: First show f has an antiderivative on D , and then use the fact that the antiderivative is
differentiable.]
2023–24
Chapter 5
Cauchy’s Residue Theorem
One of the more remarkable applications of integration in the complex plane in general, and
Cauchy’s Theorem in particular, is that it gives a method for calculating real integrals that,
up until now, would have been difficult or even impossible. As another application: you may
remember from Real Analysis or Sequences and Series that you studied whether an infinite
series ∞ n=0 a n converged or not. However, in only very few examples were you able to say
P
what the limit actually is! Using complex analysis, it becomes very easy to evaluate (some)
infinite series; for example one can show that
X∞ 1 π2
2
= .
n=1 n 6
See the end of Chapter 6. (This sum was in fact first discovered by Euler using clever summa-
tion arguments.)
Before proceeding with the definition of residues, we need to discuss the three types of
isolated singularity of functions; these are based on the Laurent series of a function.
5.1 Singularities
Definition 5.1. If a function f is holomophic on a domain D with the exception of a point

w ∈ D then we say w is an isolated singularity of f . ✯
A common way for a singularity to occur is if f is not defined at w , for then it cannot be
differentiable at w and hence not holomorphic.
Examples 5.2. (i). A simple example is f (z) = sinz z . Substituting z = 0 gives 0/0 which is
not defined. Hence f has a singularity at z = 0, and for z 6= 0 the function f is certainly
defined and homorphic.
85
86 § 5.1 Singularities
(ii). If f (z) = 1/z then f is not defined at the origin (we are not allowed to divide by 0).
Hence f has a singularity at z = 0, and is holomorphic for z 6= 0.
(iii). A third example is given by f (z) = exp(1/z). This is an example of the ‘worst’ type of
singularity as we will see.
We will see below that there are 3 types of isolated singularity: ‘removable’, ’pole’ and
‘essential’. These examples illustrate one of each type.
Suppose that f has an isolated singularity at z 0 . Then f is holomorphic on a punctured

disc of the form D ∗ (z 0 ; R) = {z ∈ C | 0 < |z − z 0 | < R}. We expand f as a Laurent series around
z 0 on this punctured disc to obtain
∞ ∞
b n (z − z 0 )−n + a n (z − z 0 )n ,
X X
f (z) =
n=1 n=0
and this is valid for 0 < |z − z 0 | < R. Consider the principal part of the Laurent series
∞
b n (z − z 0 )−n .
X
(5.1)
n=1
There are three possibilities: the principal part of f may have

(i) no terms,
(ii) a finite number of terms, or
(iii) an infinite number of terms.
These three possibilities correspond to the three types, as we’ll now discuss.
5.1A Removable singularities
Suppose that f has an isolated singularity at z 0 and that the principal part of the Laurent
series (5.1) has no terms in it. In this case, for 0 < |z − z 0 | < R we have that
f (z) = a 0 + a 1 (z − z 0 ) + · · · + a n (z − z 0 )n + · · · .
The radius of convergence of this power series is at least R, and so f (z) extends to a function
that is differentiable at z 0 provided we define f (z 0 ) = a 0 .
In other words, a removable singularity is not really a singularity at all, it is just a problem
with the way the function was defined.
Example 5.3. Let

sin z
f (z) = , z 6= 0.
z
Then f has an isolated singularity at 0 as f (z) is not defined at z = 0. However, we know that
sin z z2 z4
= 1− + −···
z 3! 5!
2023–24
C AUCHY ’ S R ESIDUE T HEOREM 87
for z 6= 0. Define f (0) = 1. Then f (z) is differentiable for all z ∈ C. Hence f has a removable
singularity at z = 0.
5.1B Poles
Suppose that f has an isolated singularity at z 0 and that the principal part of the Laurent
series (5.1) has finitely many terms in it. In this case, for 0 < |z − z 0 | < R we can write
bm b1 ∞
a n (z − z 0 )n
X
f (z) = + · · · + +
(z − z 0 )m z − z 0 n=0
where b m 6= 0. In this case, we say that f has a pole of order m at z 0 . A pole of order 1 is called
a simple pole.
Example 5.4. Let

sin z
, z 6= 0.
f (z) =
z4
Then f has an isolated singularity at z = 0. We can write
sin z 1 11 1 1
4
= 3− + z − z3 + · · · .
z z 3! z 5! 7!
Hence f has a pole of order 3 at z = 0.
We will often consider functions f : D → C defined on a domain D that are differentiable

except at finitely many points in D and f has either removable singularities or poles at these
points.
Definition 5.5. Let D be a domain. A function f : D → C is said to be meromorphic if f is

holomorphic on D except for a set of isolated singularities, and these are either removable
singularities or poles. ✯
We will make considerable use of meromorphic functions in the next chapter.
5.1C Isolated essential singularities
Suppose that f has an isolated singularity at z 0 and that the principal part of the Laurent se-
ries (5.1) has infinitely many terms in it. In this case we say that z 0 is an essential singularity
of f .
Essential singularities are difficult to deal with and we will not consider them in this
course.
2023–24
88 § 5.2 Zeros and poles of meromorphic functions
Example 5.6. Let f (z) = sin(1/z), z 6= 0. Then f has a singularity at z = 0 and
1 1 1 1
µ ¶
sin = − 3
+ −··· .
z z 3!z 5!z 5
Hence f has an isolated essential singularity at z = 0.
Remark One of the differences between a pole and an essential singularity is that if z 0 is the
isolated singularity in question then
• if z 0 is a pole then | f (z)| → ∞ as z → z 0 , while
• if z 0 is an essential singularity and w ∈ C is any complex number, then there exists a se-
quence of points z j converging to z 0 such that f (z j ) → w as z j → z 0 . [This is known as the
Casorati-Weierstrass theorem, but is not part of our syllabus. See eg, Stewart and Tall p.234,
or Ahlfors p.129 for more details.] ❞
5.2 Zeros and poles of meromorphic functions
Recall that a function f has a singularity at z 0 if f is not differentiable at z 0 . We will only

consider the case when f has poles as singularities. In the examples we have seen so far f (z)
has a pole at z 0 because we have been able to write f (z) = p(z)/q(z) and q(z 0 ) = 0 (so that f
is not even defined at z 0 ). Thus it makes sense to first study zeros of functions.
Definition 5.7. A function f defined on a domain D has a zero at z 0 ∈ D if f (z 0 ) = 0. ✯
We will only be interested in isolated zeros. Intuitively, a function f has an isolated zero
at z 0 if there are no other zeros nearby. More formally, we have the following definition.
Definition 5.8. A function f defined on a domain D has an isolated zero at z 0 if f (z 0 ) = 0

and there exists ε > 0 such that f (z) 6= 0 for all z ∈ D(z 0 ; ε) \ {z 0 }. ✯
Let f : D → C be holomorphic and suppose that f has an isolated zero at z 0 . By Taylor’s

Theorem (Theorem 4.1), we can expand f as a Taylor series in some neighbourhood around
z 0 . That is we can wrote
∞
a n (z − z 0 )n
X
f (z) = (5.2)
n=0
for all z in some disc that contains z 0 .
Definition 5.9. We say that f has a zero of order m at z 0 if a 0 = a 1 = · · · = a m−1 = 0 but

a m 6= 0. We say that z 0 is a simple zero if it is a zero of order 1. ✯
2023–24
Example 5.10. (i) Let f (z) = z 2 . Then f has a zero of order 2 at 0.
(ii) Let f (z) = z(z + 2i )3 . Then f has a zero of order 1 at 0 and a zero of order 3 at −2i .
(iii) Let f (z) = z 2 + 4. Then, noting that z 2 + 4 = (z − 2i )(z + 2i ), we see that f has simple
zeros at ±2i .
Remark The coefficients a n in the Taylor expansion are given by a n = f (n) (z 0 )/n!. Thus f
has a zero of order m at z 0 if and only if f (k) (z 0 ) = 0 for 0 ≤ k ≤ m − 1 but f (m) (z 0 ) 6= 0. In
particular, if f (z 0 ) = 0 but f ′ (z 0 ) 6= 0 then z 0 is a simple zero. ❞
Example 5.11. (i) Let f (z) = sin z. Then f (z) has zeros at kπ, k ∈ Z. Note that f ′ (kπ) =
cos kπ = (−1)k 6= 0. Hence all the zeros are simple zeros.
(ii) Let f (z) = 1 − cos z. Then f (z) has zeros at 2kπ, k ∈ Z. Now f ′ (z) = sin z and f ′ (2kπ) =
0, but f ′′ (2kπ) = cos 2kπ = 1 6= 0. Hence all the zeros have order 2.
Lemma 5.12. A holomorphic function f has a zero of order m at z 0 if and only if there is a
holomorphic function g (z) defined on some disc centre z 0 with g (z 0 ) 6= 0 such that
f (z) = (z − z 0 )m g (z).
Proof: ’Only if’: Suppose f has a zero of order m at z 0 . Then by (5.2) we can write
∞
f (z) = a m (z − z 0 )m + a m+1 (z − z 0 )m+1 + · · · = (z − z 0 )m a n+m (z − z 0 )n
X
n=0
n
where a m 6= 0. Take g (z) = ∞ n=0 a n+m (z − z 0 ) . Then g is holomorphic on an open disc
P
centred on z 0 and g (z 0 ) = a m 6= 0.
’If’: Left to the reader (begin by writing g (z) = b n (z − z 0 )n with b 0 6= 0).
P
We can now link poles of a function f (z) = p(z)/q(z) with zeros of the function q.
Lemma 5.13. Suppose that f (z) = p(z)/q(z) where
(i) p is holomorphic and p(z 0 ) 6= 0,
(ii) q is holomorphic and q has a zero of order m at z 0 .

2023–24
90 § 5.3 Residues and Cauchy’s Residue Theorem
Then f has a pole of order m at z 0 .
Proof: By Lemma 5.12, we can write q(z) = (z − z 0 )m r (z) where r is holomorphic and
r (z 0 ) 6= 0. Define g (z) = p(z)/r (z). Then g (z) is holomorphic at z 0 , and so we can expand
it as a Taylor series at z 0 as
∞
a n (z − z 0 )n
X
g (z) =
n=0
and this expression is valid in some disc |z − z 0 | < R, for some R > 0. Then
p(z)
f (z) =
q(z)
p(z)
=
(z − z 0 )m r (z)
g (z)
=
(z − z 0 )m
1 ∞
a n (z − z 0 )n
X
=
(z − z 0 )m n=0
a0 a1 a2
= + + +··· .
(z − z 0 )m (z − z 0 )m−1 (z − z 0 )m−2
However, a 0 = g (z 0 ) = p(z 0 )/r (z 0 ) 6= 0, as p(z 0 ) 6= 0. Hence f has a pole of order m at z 0 . ❒
Example 5.14. (i) Let

sin z
f (z) = .
(z − 3)2
Then f has a pole of order 2 at z = 3. This is because sin z 6= 0 when z = 3 and (z − 3)2
has a zero of order 2 at z = 3.
(ii) Let
z +3
. f (z) =
sin z
Then f has a simple pole at kπ for each k ∈ Z. This is because sin z has a simple zero
at z = kπ for each k ∈ Z but z + 3 6= 0 when z = kπ.
5.3 Residues and Cauchy’s Residue Theorem
We begin with the following important definition.
Definition 5.15. Suppose that f is holomorphic on a domain D except for an isolated sin-
gularity at z 0 ∈ D. Suppose that on {z ∈ C | 0 < |z − z 0 | < R} ⊂ D, f has Laurent expansion
∞ ∞
a n (z − z 0 )n + b n (z − z 0 )−n .
X X
f (z) =
n=0 n=1
2023–24
The residue of f at z 0 is defined to be
Res( f ; z 0 ) = b 1 .
That is, the residue of f at the isolated singularity z 0 is the coefficient of (z − z 0 )−1 in the
Laurent expansion. ✯
Let 0 < r < R. By Laurent’s Theorem (Theorem 4.15) we have the alternative expression
1
I
Res( f ; z 0 ) = f (z) dz
2πi C r
where C r (t ) = z 0 + r e i t , 0 ≤ t ≤ 2π is a circular anticlockwise path around z 0 in the annulus of

convergence. This shows that residues are related to integration.
Cauchy’s Residue Theorem relies on using Cauchy’s Theorem in just the right way. In
particular, we have to be careful about the paths that we integrate over. We make the following
definition.
Definition 5.16. A closed contour γ is said to be a simple closed loop if, for every point z not
on γ, the winding number is either w (γ; z) = 0 or w (γ; z) = 1. If w (γ; z) = 1 then we say that z
is inside γ. ✯
γ2
γ3
γ1
F IGURE 5.1: Here γ1 is a simple closed loop. The closed loops γ2 and γ3 are
not simple because there are points where the winding number is −1.
Thus a simple closed loop is a loop that goes round anticlockwise in a loop once, and
without intersecting itself; see Figure 5.1. In practice, we will look at simple closed loops that
are made up of line segments and arcs of circles.
2023–24
92 § 5.4 Calculating residues
We can now state the main result of this section.
Theorem 5.17 (Cauchy’s Residue Theorem). Let D be a domain containing a simple

closed loop γ and the points inside γ. Suppose that f is meromorphic on D with finitely
many poles at z 1 , z 2 , . . . , z n inside γ. Then
I n
X
f (z) dz = 2πi Res( f ; z j ).
γ j =1
Warning: A word of warning: you will have noticed that many expressions in complex anal-
ysis have a factor of 2πi in them. A very common mistake is to either miss a 2πi out, or put
one in by mistake. "
We defer the proof of Cauchy’s Residue Theorem until later.
5.4 Calculating residues
In order to use Cauchy’s Residue Theorem we need to be able to easily calculate residues. In
some cases, ad hoc manipulations have to be used to calculate the Laurent series, but there
are many cases where one can calculate them more systematically.
First recall that if f (z) has Laurent series
bm b1 ∞
a n (z − z 0 )n
X
f (z) = + · · · + + (5.3)
(z − z 0 )m (z − z 0 ) n=0
with b m 6= 0 then we say that f has a pole of order m at z 0 , and a pole of order 1 is called a
simple pole. The residue of this expression is Res( f ; z 0 ) = b 1 . The question we address in this
section is how do we find b 1 for a given function? One method is clearly to calculate the entire
Laurent series (5.3), but since we only require one coefficient this is often overkill and there
are simpler methods.
Remark If we can write f (z) = p(z)/q(z) where p and q are differentiable and p(z) 6= 0
when q(z) = 0 then the poles of f occur at the zeros of q. Moreover f has a pole of order m at
z 0 if q has a zero of order m at z 0 . ❞
It is easy to calculate the residue at a simple pole, for then the only non-zero b n is b 1 .
Lemma 5.18. (i) If f has a simple pole at z 0 then
Res( f ; z 0 ) = lim (z − z 0 ) f (z).

z→z 0
2023–24
(ii) If f (z) = p(z)/q(z) where p, q are differentiable, p(z 0 ) 6= 0, q(z 0 ) = 0 but q ′ (z 0 ) 6= 0, then
p(z 0 )
Res( f ; z 0 ) = .
q ′ (z 0 )
Proof:
(i) If f has a simple pole at z 0 then it has a Laurent series
b1 ∞
a n (z − z 0 )n
X
f (z) = +
z − z 0 n=0
valid on some punctured disc 0 < |z − z 0 | < R. Hence
∞
a n (z − z 0 )n+1
X
(z − z 0 ) f (z) = b 1 +
n=0
so that limz→z0 (z − z 0 ) f (z) = b 1 .

(ii) The hypotheses imply that f has a simple pole at z 0 . By part (i) and the fact that
q(z 0 ) = 0, the residue is
(z − z 0 )p(z) p(z) p(z 0 )

lim = lim ³ ´= ′ .
z→z 0 q(z) z→z 0 q(z)−q(z 0 ) q (z 0 )
z−z 0
Example 5.19. For example, let

cos πz
f (z) = .
(1 − z 3 )
This has a simple pole at z = 1 and satisfies the hypothesis of Lemma 5.18 with p(z) =
cos(πz) and q(z) = (1 − z 3 ). Then p(1) = (−1) and q ′ (1) = −3(1)2 = −3. Hence
−1 1
Res( f ; 1) = = .
−3 3
We can generalize Lemma 5.18 to poles of order m.
Lemma 5.20. Suppose that f has a pole of order m at z 0 . Then
1 dm−1 ¡
µ ¶
m
¢
Res( f ; z 0 ) = lim (z − z 0 ) f (z) .
z→z 0 (m − 1)! dz m−1
2023–24
94 § 5.4 Calculating residues
Proof: If f has a pole of order m at z 0 then it has a Laurent series
bm b1 ∞
a n (z − z 0 )n
X
f (z) = + · · · + +
(z − z 0 )m z − z 0 n=0
valid for 0 < |z − z 0 | < R, for some R > 0. Hence

∞
(z − z 0 )m f (z) = b m + (z − z 0 )b m−1 + · · · + (z − z 0 )m−1 b 1 + a n (z − z 0 )m+n .
X
n=0
Differentiating this m − 1 times gives
dm−1 ∞ (m + n)!
(z − z 0 )m f (z) = (m − 1)!b 1 + a n (z − z 0 )n+1 .
X
dz m−1
n=0 (n + 1)!
Dividing by (m − 1)! and letting z → z 0 gives the result. ❒
Example 5.21. Let ¶3

z +1
µ
f (z) = .
z −1
This has a pole of order 3 at z = 1. To calculate the residue we note that (z −1)3 f (z) = (z +1)3 .
Hence
1 d2 ¡ ¢ 6 6
2
(z − 1)3 f (z) = (z + 1)−→ × 2 = 6
2! dz 2! 2!
as z → 1. Hence Res( f ; 1) = 6.
Let us check this by calculating the Laurent series. First let us change variables by writing
w = z − 1. Then z = w + 1 and we can write
z +1 3 (w + 2)3
µ ¶
=
z −1 w3
w 3 + 6w 2 + 12w + 8
=
w3
8 12 6
= 3
+ 2 + +1
w w w
8 12 6
= + + + 1.
(z − 1)3 (z − 1)2 (z − 1)
Hence f has a pole of order 3 at z = 1 and we can read off Res( f ; 1) = 6 as the coefficient of
1/(z − 1).
In other cases, one has to manipulate the formula for f to calculate the residue.
2023–24
Example 5.22. Let

1
f (z) = .
z 2 sin z
This has singularities whenever the denominator is zero. Hence the singularities are at z =
0, kπ. We will use Laurent series to calculate the residue at z = 0.
Recalling the power series for sin z we can write
1
f (z) =
z 2 sin z
1
=
z3
µ ¶
2
z z− +···
6
¶−1
1 z2
µ
= 1− +···
z3 6
1 z2
µ ¶
= 1+ +···
z3 6
1 1
= 3
+ +···
z 6z
where we have omitted higher order terms. (Note that when doing computations such as
these, one can usually ignore terms that will not contribute to the coefficient of 1/z.) Hence
we have a pole of order 3, with Res( f ; 0) = 1/6.
For the poles at kπ, k 6= 0, we could change variables to w = z − kπ and calculate the
Laurent series. Alternatively, we can use Lemma 5.18(ii). First note that we can write
p(z)
f (z) =
q(z)
where p(z) = 1 and q(z) = z 2 sin z. Now, for k 6= 0, kπ is a simple zero of sin z (as sin′ kπ =
cos kπ 6= 0) and so is a simple zero of q(z). Hence
p(kπ) (−1)k
Res( f ; kπ) = =
q ′ (kπ) (kπ)2
as q ′ (z) = 2z sin z + z 2 cos z so that q ′ (kπ) = (kπ)2 cos kπ = (−1)k (kπ)2 .

Alterntively, let p(z) = 1/z 2 and q(z) = sin z. Now kπ is a simple zero of q(z) and not a
zero of p(z), so we can apply Lemma 5.18(ii) to find
p(kπ) (1/kπ)2 (−1)k

Res( f ; kπ) = = = .
q ′ (kπ) (−1)k (kπ)2
2023–24
96 § 5.5 Straightforward examples
5.5 Straightforward examples
There are many different applications of the residue theorem, and you will see more in the
next chapter and many in Part 2 of the course. Here we limit ourselves to some simple exam-
ples.
We shall evaluate some simple integrals around the circular contours C 2 (t ) = 2e i t , 0 ≤
t ≤ 2π and C 4 (t ) = 4e i t , 0 ≤ t ≤ 2π. Thus C 2 is the circle of radius 2 centred at 0 oriented
anticlockwise, and C 4 is the circle of radius 4 centred at 0 oriented anticlockwise. Hence both
C 2 and C 4 are simple closed loops.
Consider first the function
3
f (z) =
z −1
Then f has a pole at z = 1 and no other poles. We can read off from the definition of f that
Res( f ; 1) = 3. As the pole at z = 1 lies inside C 2 , by Cauchy’s Residue Theorem we have that
I
f dz = 2πi Res( f ; 1) = 6πi .
C2
Similarly, the pole at z = 1 lies inside C 4 , hence

I
f dz = 2πi Res( f ; 1) = 6πi .
C4
See Figure 5.2.
C4
C2
F IGURE 5.2: The function f (z) = 3/(z − 1) has a pole at z = 1 which lies inside
both C 2 and C 4 .
Now consider the function

1
f (z) = .
z 2 + (i − 3)z − 3i
Then f has a pole when the denominator has a zero. To find the poles we first factorise the
denominator
z 2 + (i − 3)z − 3i = (z − 3)(z + i )
2023–24
(to do this we could either use the quadratic formula or inspired guesswork). Thus f has
simple poles z = 3 and z = −i . Using Lemma 5.18 we can calculate that
−1 1
Res( f ; −i ) = , Res( f ; 3) = .
3+i 3+i
See Figure 5.3.
C4
C2
3
−i b
F IGURE 5.3: The function f (z) = 1/(z 2 +(i −3)z −3i ) has simple poles at z = −i
and z = 3.
H
Now consider C2 f dz. The pole z = −i is inside C 2 but the pole z = 3 is outside. Hence
−2πi (3 − i )
I µ ¶
−1
f dz = 2πi Res( f ; −i ) = 2πi =
C2 3+i 10
−2π − 6πi −π
= = (1 + 3i ).
10 5
H
Now consider C4 f dz. In this case, both the poles at z = −i and z = 3 lie inside C 4 . Hence
1
I µ ¶
¡ ¢ −1
f dz = 2πi Res( f ; −i ) + Res( f ; 3) = 2πi + = 0.
C4 3+i 3+i
2023–24
98 § 5.6 Proof of Cauchy’s Residue Theorem
5.6 Proof of Cauchy’s Residue Theorem
Let us first recall the statement of Cauchy’s Residue Theorem:
Theorem 5.23 (Cauchy’s Residue Theorem). Let D be a domain containing a simple

closed loop γ and the points inside γ. Suppose that f is holomorphic on D except for
finitely many poles at z 1 , z 2 , . . . , z n inside γ. Then
I n
X
f (z) dz = 2πi Res( f ; z j ).
γ j =1
Proof: The proof is a simple application of the Generalized Cauchy Theorem (Theo-
rem 3.10).
Since D is open, for each j = 1, . . . , n, we can find circles
S j (t ) = z j + ε j e i t , 0 ≤ t ≤ 2π
centred at z j and of radii ε j , each oriented anticlockwise, such that S j and the points
inside S j lie in D and such that S j contains no singularity other than z j (see Figure 5.4).
D S1
z1 γ
S2
S3 z2
z3
F IGURE 5.4: Here we have 3 poles at z 1 , z 2 , z 3 inside γ. The circles S 1 , S 2 , S 3

(centred on z 1 , z 2 , z 3 , respectively) have been chosen so that they lie inside γ
and do not intersect each other.
2023–24
Let D ′ = D \ {z 1 , . . . , z n }. We claim that the collection of paths
γ− , S 1 , . . . , S n
satisfy the hypotheses of the Generalized Cauchy Theorem (Theorem 3.10) with respect
to D ′ : i.e. their winding numbers sum to zero for every point not in D ′ .
To see this, first note that
w (γ− ; z) = w (S j ; z) = 0 for z 6∈ D.
Hence the hypotheses of the Generalized Cauchy Theorem hold for points z not in D. It
remains to consider points in D that are not in D ′ , i.e. the poles z j .
Since each pole z j lies inside γ, we have that
w (γ−; z j ) = −w (γ; z j ) = −1.
Moreover,
0 if k 6= j
½
w (S k ; z j ) =
1 if k = j.
Hence
w (γ−; z j ) + w (S 1; z j ) + · · · + w (S n ; z j ) = 0.
Hence, by the Generalized Cauchy Theorem,

I I I
f+ f +··· + f = 0.
γ− S1 Sn
By Laurent’s Theorem (Theorem 4.15) we have that
1
I
Res( f ; z j ) = f (z) dz.
2πi Sj
Hence
I I I
f = f + ··· + f
γ S1 Sn
¡ ¢
= 2πi Res( f ; z 1 ) + · · · + Res( f ; z n ) ,
concluding the proof. ❒

2023–24
100 § 5.7 Exercises
5.7 Exercises
Ex 5.1 Find the poles and their orders of the functions
1 1 1 1
(i) , (ii) , (iii) , (iv) .
z2 + 1 z 4 + 16 z 4 + 2z 2 + 1 z2 + z − 1
Ex 5.2 Describe the type of singularity at 0 of each of the following functions:
cos z − 1
(i) sin(1/z), (ii) z −3 sin2 z, (iii) .
z2
Ex 5.3 Let D be a domain and let z 0 ∈ D. Suppose that f is holomorphic on D \ {z 0 } and is

bounded on D \ {z 0 } (that is, there exists M > 0 such that | f (z)| ≤ M for all z ∈ D \ {z 0 }).
Show that f has a removable singularity at z 0 .
Ex 5.4 Use Lemma 5.13 to determine the poles of the following functions. Use Lemmas 5.18,
and 5.20 (as appropriate) to calculate the residue at each pole.
¶2
1 z +1
µ
z
(i) , (ii) tan z, (iii) , (iv) 2 .
z(1 − z 2 ) 1 + z4 z +1
Ex 5.5 Determine the singularities of the following functions. By considering Taylor series,
calculate the residue at each singularity.
sin z sin2 z
(i) , (ii) .
z2 z4
1
Ex 5.6 (i) Let f (z) = . Then f has singularities at 0 and 1. Expand f as a Laurent
z(1 − z)2
series at 0 and as a Laurent series at 1. In each case, read off from the Laurent series
the order of the pole and the residue of the pole.
(Hint: recall that 1/(1 − z)2 = 1 + 2z + 3z 2 + · · · + nz n−1 + · · · if |z| < 1.)
(ii) Check your answer by using Lemmas 5.13, 5.18 and 5.20.
Ex 5.7 Suppose that f , g : D → C are holomorphic and that z 0 ∈ D. Suppose that f has a zero
of order n at z 0 and g has a zero of order m at z 0 . Show that f (z)g (z) has a zero of order
n + m at z 0 .
Ex 5.8 Let C r be the circle C r (t ) = r e i t , 0 ≤ t ≤ 2π, with centre 0 and radius r . Use Cauchy’s
Residue Theorem to evaluate the integrals
1 1 e az
I I I
(i) 2
dz, (ii) 2
dz, (iii) dz (a ∈ R).
C4 z − 5z + 6 C 5/2 z − 5z + 6 C2 1 + z2
2023–24
Ex 5.9 (This optional exercise shows how to use Cauchy’s Residue Theorem to explicitly cal-
culate the coefficients in Laurent’s Theorem using the integral formula.)
Recall that Laurent’s Theorem (Theorem 4.15) says the following:
Suppose that f is holomorphic on the annulus A(z 0 ; R 1 , R 2 ) = {z ∈ C | R 1 < |z − z 0 | < R 2 }.
Then f can be written as a Laurent series on this annulus in the form
∞
a n (z − z 0 )n .
X
f (z) =
n=−∞
The coefficients are given by
1 f (z)
I
an = dz
2πi Cr (z − z 0 )n+1
where C r (t ) = z + r e i t , 0 ≤ t ≤ 2π, denotes the circular path centre z 0 of radius r and r is

chosen such that R 1 < r < R 2 .
By using Cauchy’s Residue Theorem to evaluate a n , determine the Laurent series for the
function
1
f (z) =
z(z − 1)
valid on the annuli
(i) A(0; 0, 1) = D ∗ (0; 1) (0 < |z| < 1),

(ii) A(0; 1, ∞) (ie 1 < |z| < ∞),
(iii) D ∗ (1; 1) (ie 0 < |z − 1| < 1),
(iv) A(1; 1, ∞) (ie 1 < |z − 1| < ∞).
(Thus we are using Cauchy’s Residue Theorem to evaluate C r f (z)/(z − z 0 )n+1 dz around
H
a suitable closed contour C r of radius r and centred at an appropriate z 0 (z 0 = 0 in (i),(ii)

and z 0 = 1 in (iii),(iv)) by locating the poles of f (z)/(z−z 0 ) that lie inside C r and calculating
their residues.)
In each case, check your answer by directly calculating the Laurent series using the meth-
ods described in §4.3.
2023–24
2023–24 102 Exercises
Chapter 6
Applications of the residue theorem
6.1 Infinite real integrals
In this section we shall show how to use Cauchy’s Residue Theorem to calculate some infinite
real integrals, i.e. integrals of the form
Z∞
f (x) d x (6.1)
−∞
where f is a real-valued function defined on the real line.

R∞
First we need to make precise what (6.1) means. Formally, we say that −∞ f (x) d x exists
if ZB
lim f (x) d x (6.2)
A,B →∞ −A
R∞
converges, where the limits can be taken in either order. We then define −∞ f (x) d x to be
equal to this limit.
R∞
If −∞ f (x) d x exists then it is equal to its principal value, defined by
Z∞ ZR
P f (x) d x = lim f (x) d x. (6.3)
−∞ R→∞ −R
However, there are many functions f for which the principal value of the integral (6.3) exists
but (6.2) does not. For example, take f (x) = x. Then
1 2 ¯¯R R2 R2
ZR ZR ¯
f (x) d x = x dx = x ¯ = − =0
−R −R 2 x=−R 2 2
R∞
and so converges to 0 as R → ∞. Hence P −∞ x d x = 0. However
1 2 ¯¯B B 2 A2
ZB ZB ¯
f (x) d x = x dx = x ¯ = −
−A −A 2 x=−A 2 2
103
104 § 6.1 Infinite real integrals
R∞
does not converge if we first let B tend to ∞ and then let A tend to ∞. Hence −∞ x d x does
not exist.
The following gives a criterion for (6.2) to converge.
Lemma 6.1. Suppose that f : R → C is a continuous function and there exist constants K > 0,
C > 0 and r > 1 such that for |x| ≥ K we have
C
| f (x)| ≤ . (6.4)
|x|r
R∞ R∞
Then −∞ f (x) d x exists and is equal to its principal value P −∞ f (x) d x.
Instead of giving a general theorem, let us consider an example that will illustrate the
method. We will show how to use Cauchy’s Residue Theorem to evaluate
1
Z∞
2 2
dx (6.5)
−∞ (x + 1)(x + 4)
(the fact that 1 and 4 are squares will make the calculations notationally easier, but this is not
essential to the method).
R RR
First note that the complex contour integral [−R,R] f is equal to the real integral −R f (x) d x.
To see this, first recall from (1.6) that [−R, R] denotes the straight line path from −R to R and
that this has parametrisation γ(t ) = t , −R ≤ t ≤ R. Hence
Z ZR ZR
f = f (γ(t )) γ′ (t ) dt = f (t ) dt .
[−R,R] −R −R
Note that there exists a constant C > 0 such that

1
¯ ¯
¯ ¯ C
¯ (x 2 + 1)(x 2 + 4) ¯ ≤ x 4 .
¯ ¯
Hence, by Lemma 6.1, the infinite integral −∞ 1/(x 2 + 1)(x 2 + 4) d x exists and is equal to its
R∞
principal value
ZR
1 1
Z∞
P 2 2
d x = lim d x.
−∞ (x + 1)(x + 4) R→∞ −R (x + 1)(x 2 + 4)
2
We will calculate the principal value of integral using Cauchy’s Residue Theorem.
Let
1
f (z) =
(z 2 + 1)(z 2 + 4)
(note that we have introduced a complex variable). Let [−R, R] denote the path along the real
axis that starts at −R and ends at R. This has parametrisation t , −R ≤ t ≤ R. Note that we can
equate the real integral (6.5) with the complex integral as follows:
ZR
1
Z
2 2
d x = f (z) dz.
−R (x + 1)(x + 4) [−R,R]
2023–24
A PPLICATIONS OF THE RESIDUE THEOREM 105
-R R
F IGURE 6.1: The ‘D-shaped’ contour ΓR . It starts at −R, travels along the real
axis to R, and then anticlockwise along the semicircle S R with centre 0 and
radius R.
To use Cauchy’s Residue Theorem, we need a closed contour. Introduce a semi-circular

path S R (t ) = Re i t , 0 ≤ t ≤ π and the ‘D-shaped’ contour ΓR = [−R, R] + S R (see Figure 6.1).
Now ΓR is a simple closed loop. To use Cauchy’s Residue Theorem, we need to know the
poles and residues of f (z). Now
1 1
f (z) = = .
(z 2 + 1)(z 2 + 4) (z − i )(z + i )(z − 2i )(z + 2i )
Hence f (z) has simple poles at z = +i , −i , +2i , −2i . If we take R > 2 then the poles at z = i , 2i
lie inside ΓR (note that the poles at z = −i , −2i lie outside ΓR ). Now by Lemma 5.18,
Res( f ; i ) = lim(z − i ) f (z)

z→i
1
= lim
z→i (z + i )(z − 2i )(z + 2i )
1
=
6i
and
Res( f ; 2i ) = lim (z − 2i ) f (z)

z→2i
1
= lim
z→2i (z − i )(z + i )(z + 2i )
−1
= .
12i
Hence by Cauchy’s Residue Theorem
Z Z I
f (z), dz + f (z) dz = f (z) dz
[−R,R] SR ΓR
¡ ¢
= 2πi Res( f ; i ) + Res( f ; 2i )
1 1
µ ¶
π
= 2πi − = .
6i 12i 6
2023–24
106 § 6.1 Infinite real integrals
If we can show that Z

lim f (z) dz = 0 (6.6)
R→∞ S R
then we will have that

1
Z∞
π
Z
2 + 1)(x 2 + 4)
d x = lim f (z) dz = .
−∞ (x R→∞ [−R,R] 6
To complete the calculation, we show that (6.6) holds. We shall use the Estimation Lemma.
Let z be a point on S R . Note that |z| = R. Hence
|(z 2 + 1)(z 2 + 4)| ≥ (R 2 − 1)(R 2 − 4)
so that
1 1
¯ ¯
¯ ¯
¯ (z 2 + 1)(z 2 + 4) ¯ ≤ (R 2 − 1)(R 2 − 4) .
¯ ¯
Hence, by the Estimation Lemma,
1
¯Z ¯
¯ ¯
¯ f (z) dz ¯¯ ≤ length(S R )
¯
SR (R 2 − 1)(R 2 − 4)
πR
=
(R 2 − 1)(R 2 − 4)
→ 0
as R → ∞, which is what we wanted to check.
Remark 6.2. As a general method, to evaluate

ZR
f (x) d x
−R
one uses the following steps:
(i) Check that f (x) satisfies the hypotheses of Lemma 6.1.
(ii) Construct a ‘D-shaped’ contour ΓR as in Figure 6.1.
(iii) Find the poles and residues of f (z) that lie inside ΓR when R is large.
R
(iv) Use Cauchy’s Residue Theorem to write down ΓR f (z) dz.
(v) Split this integral into an integral over [−R, R] and an integral over S R . Use the Estima-
tion Lemma to conclude that the integral over S R converges to 0 as R → ∞.
For a particular example, one may need to make small modifications to the above process,
but the general method is normally as above. ❞
2023–24
Remark 6.3. It is very easy to lose minus signs or factors of 2πi when doing these computa-
tions. You should always check that your answer makes sense. For example, if I had missed
out a factor of i in the above then I would have obtained an expression of the form
1
Z∞
i
2 + 1)(x 2 + 4)
dx = .
−∞ (x 6
This is obviously wrong: the left-hand side is a real number, whereas the (incorrect) right-
hand side is imaginary. Similarly, in this example the integrand on the left-hand side is a
positive function, and so the integral must be positive; hence if the right-hand side is negative
then there must be a mistake somewhere in the calculation. ❞
6.2 Trigonometric integrals
We can use Cauchy’s Residue Theorem to calculate integrals of the form

Z2π
Q(cos t , sin t ) dt (6.7)
0
where Q is some function. (Integrands such as cos4 t sin3 t − 7 sin t , or cos t + sin2 t , etc, fall
into this category.)
The first step is to turn (6.7) into a complex integral. Set z = e i t . Then
z + z −1 z − z −1
cos t = , sin t = .
2 2i
Also [0, 2π] transforms into the unit circle C 1 (t ) = e i t , 0 ≤ t ≤ 2π. Finally, note that dz = i e i t dt
so that
dz
dt = .
iz
Hence Z2π
z + z −1 z − z −1 dz
Z µ ¶
Q(cos t , sin t ) dt = Q , .
0 C1 2 2i iz
Then in principle we can evaluate this integral by finding the poles of
z + z −1 z − z −1 1
µ ¶
Q ,
2 2i iz
inside C 1 , together with their associated residues, and then use Cauchy’s Residue Theorem.
Instead of stating a general theorem, we shall compute some examples to illustrate the
method.
2023–24
108 § 6.2 Trigonometric integrals
Example 6.4. We shall show how to compute

Z2π
(cos3 t + sin2 t ) dt .
0
Let z = e i t so that dt = dz/i z. Then

Z2π
(cos3 t + sin2 t ) dt
0
Z Ãµ ¶3 µ ¶2 !
z + z −1 z − z −1 dz
= +
C1 2 2i iz
Z µ 3
3z 3z −1 z −3 z 2 1 z −2 dz
¶
z
= + + + − + −
C1 8 8 8 8 4 2 4 iz
2 −2 −4 −1
1 z 3 3z z −3
µ ¶
z z z
Z
= + + + − + − dz
C1 i 8 8 8 8 4 2 4
The integrand has a pole of order 4 at z = 0 with residue 1/2i , and no other poles. Hence
Z2π
1
(cos3 t + sin2 t ) dt = 2πi = π.
0 2i
Example 6.5. We shall compute

Z2π
cos t sin t dt .
0
Again, substituting z = e i t we have that

Z2π
cos t sin t dt
0
1 dz
Z
= (z + z −1 )(z − z −1 )
C1 4i iz
1 2 dz
Z
= (z − z −2 )
C 4i iz
Z1 µ ¶µ
1
¶
−1
= z − 3 dz.
C1 4 z
The integrand has a pole of order 3 at z = 0 with residue 0. There are no other poles. Hence
Z2π
cos t sin t dt = 0.
0
2023–24
6.3 Summation of series
These are all based on the function cotπz which conveniently has a pole at each integer point.
Recall that cot πz = cos πz/ sin πz. Then cot πz has a pole whenever sin πz = 0, i.e. when-
ever z = n, n ∈ Z. First note that sin πz has a simple zero at z = n (as sin′ πz = π cos πz 6= 0
when z = n). Hence cot πz has a simple pole at z = n. By Lemma 5.18(ii) we have
cos πn 1
Res(cot πz; n) = = .
π cos πn π
This suggests a method for summing infinite series of the form ∞ n=1 a n . Let f (z) be a mero-
P
morphic function defined on C such that f (n) = a n . Consider the function f (z) cot πz. Then,
if f (n) 6= 0, we have
an
Res( f (z) cot πz ; n) =
π
and we can use Cauchy’s Residue Theorem to calculate ∞ n=1 a n . For example, we will show
P
P∞ 2
how to use this method to calculate n=1 1/n .
There are two technicalities to overcome. First of all, we need to choose a good contour
to integrate around. We will want to use the Estimation Lemma along this contour, so we
will need some bounds on | f (z) cot(πz)|. Secondly, f (z) may have poles of its own and these
2
will need to be taken into account. (In the above example, to calculate ∞n=1 1/n we will take
P
f (z) = 1/z 2 , which has a pole at z = 0.)

Instead of choosing a D-shaped contour, here we use a square contour. Let C N denote
the square in C with vertices at
1 1 1 1
µ ¶ µ ¶ µ ¶ µ ¶
N + −i N + , N + +i N + ,
2 2 2 2
1 1 1 1
µ ¶ µ ¶ µ ¶ µ ¶
− N + +i N + , − N + −i N +
2 2 2 2
(see Figure 6.2). This is a square with each side having length 2N + 1. (The 12 ’s are there so
that the sides of this square do not pass through the integer points on the real axis.)
Lemma 6.6. There is a bound, independent of N , on cot πz for z ∈ C N , i.e. there exists M > 0
such that for all N and all z ∈ C N , we have | cot πz| ≤ M .
Proof: Consider the square C N . This has two horizontal sides and two vertical sides,
parallel to the real and imaginary axes, respectively.
Consider first the horizontal sides. Let z = x + i y be a point on one of the horizontal
2023–24
110 § 6.3 Summation of series
CN
b b b b b b b
-(N+1)-N -1 1 N N+1
F IGURE 6.2: The square contour C N .
sides of C N . Then |y| ≥ 1/2. Hence
¯ i πz
+ e −i πz ¯¯
¯
¯e
| cot πz| = ¯
¯ e i πz − e −i πz ¯
¯¯ ¯¯
¯ ¯e i πz ¯ + ¯e −i πz ¯ ¯
¯ ¯
≤
¯ ¯
¯ ¯ i πz ¯ ¯ −i πz ¯ ¯
¯ e
¯ ¯ − e
¯ ¯ ¯
¯ −πy
¯e πy ¯
+ e ¯¯
= ¯
¯ e −πy − e πy ¯
= coth |πy|
³π´
≤ coth
2
as |y| ≥ 1/2.
Consider now the vertical sides. If z is on a vertical side of C N then
1
µ ¶
z = ± N + + i y.
2
so x = ±(N + 12 ). Then
e 2i πz = e ±i π(2N +1) e −2πy = −e −2πy = .

2023–24
Hence,
¯ i πz
+ e −i πz ¯¯
¯
¯e
| cot πz| = ¯
¯ e i πz − e −i πz ¯
¯ 2πi z
+ 1 ¯¯
¯
¯e
= ¯
¯ e 2πi z − 1 ¯
¯ −2πi y
+ 1 ¯¯
¯
¯ −e
= ¯
¯ −e −2πi y − 1 ¯
1 − e −2πy
=
1 + e 2πy
≤ 1.
Now coth(π/2) ≈ 1.0903 > 1, hence | cot πz| < 2 for all z ∈ C N . ❒
Instead of stating a general theorem on how to use Cauchy’s Residue Theorem to eval-
uate infinite sums, we will work through an example to illustrate the method. Very similar
techniques and slight modifications to the argument work for many other examples.
2 2
We will evaluate ∞ n=0 1/n . Let f (z) = 1/z and consider the function
P
cot πz cos πz
f (z) cot πz = 2
= 2 .
z z sin πz
This has a pole whenever the denominator has a zero. These occur when z 2 sin πz = 0, i.e.
when z = n, n ∈ Z. Note that when n 6= 0 we have a simple pole and when n = 0 we have a
pole of order 3.
Let us calculate the residue when n 6= 0. We use Lemma 5.18(ii). Then
cot πz cos πn
µ ¶
Res ,n =
z2 πn 2 cos πn + 2n sinπn
1
= .
πn 2
Now consider the pole at z = 0. There are (at least) three ways to work out the residue
here, and for completeness we’ll discuss them all. Firstly, if we know the series expansion for
cot(z) we can use that:
1 z z 3 2z 5
cot z = − − − −··· .
z 3 45 945
Hence
cot πz 1 π π3 z 2π5 z 3
= − − − −···
z2 πz 3 3z 45 945
from which it is clear that z = 0 is a pole of order 3 with residue −π/3.
2023–24
112 § 6.3 Summation of series
Alternatively, if we don’t know the series for cot(z) then we can write
(πz)2 (πz)4
³ ´
1 cos πz 1 1 − 2! + 4! − · · ·
2
= 2
z sin πz
³ ´
3
z (πz) − (πz) + (πz)5 − · · ·
3! 5!
(πz)2 (πz)4
³ ´
1 1 − 2! + 4! − · · ·
=
πz 3 1 − (πz)2 + (πz)4 − · · ·
³ ´
3! 5!
2
1 (πz) (πz)2
µ ¶µ ¶
4 4
= 1− + O(z ) 1 + + O(z )
πz 3 2! 3!
1 (πz)2
µ ¶
= 1 − +···
πz 3 3
so that Res(cot πz/z 2 , 0) = −π/3. (We used the expansion (1−x)−1 = 1+x+x 2 +· · · , and looking
ahead in the calculation we saw we don’t need the z 4 terms.) Note that to calculate the residue
we need only calculate the coefficient of the term involving 1/z; hence we need to be very
careful when manipulating these infinite sums to ensure that we account for all the possible
terms which may contribute towards 1/z.
Finally, as a third method of calculating the residue at 0, one could use Lemma 5.20.
Now let C N be the square contour illustrated above. Note that each side of the square has
length 2N + 1. Hence the length of C N is 4(2N + 1).
Note that the poles that lie inside C N occur at z = 0, ±1, · · · , ±N . By Cauchy’s Residue
Theorem we have that
N
X
µ
cot πz
¶ Z
cot πz
2πi Res 2
,n = dz.
n=−N z CN z2
Recall from Lemma 6.6 that | cot πz| ≤ M on C N , where M is independent of N (we can take
M = 2). Also note that |1/z 2 | ≤ 1/N 2 for z on C N . By the Estimation Lemma we have
cot πz ¯¯ M
¯Z ¯
¯ M
¯
2
dz ¯ ≤ N 2 lengthC N = N 2 4(2N + 1)
¯
CN z
which tends to 0 as N → ∞. Hence
N
X cot πz
µ ¶
lim Res , n = 0. (6.8)
N →∞ n=−N z2
Now
N
X cot πz
µ ¶
Res , n
n=−N z2
−1 ¶ N
cot πz cot πz cot πz
µ ¶ µ µ ¶
X X
= Res , n + Res , 0 + Res , n
n=−N z2 z2 n=1 z2
N
X 1 π
= 2 2
−
n=1 πn 3
2023–24
and combining this with (6.8) we see that

∞
X 1 π
2 2
− = 0.
n=1 πn 3
This rearranges to give

X∞ 1 π2
2
= .
n=1 n 6
6.4 Exercises
Ex 6.1 (a) Consider the following real integral:

1
Z∞
dx.
−∞ x2 + 1
(i) Explain why this integral is equal to its principal value.
(ii) Use Cauchy’s Residue Theorem to evaluate this integral. (How would you have
done this without using complex analysis?)
(b) (i) Now evaluate, using Cauchy’s Residue Theorem, the integral
e 2i x
Z∞
dx.
−∞ x2 + 1
(ii) By taking real and imaginary parts, calculate
cos 2x sin 2x
Z∞ Z∞
2
dx, 2
dx.
−∞ x + 1 −∞ x + 1
(Why is it obvious, without having to use complex integration, that one of these
integrals is zero?)
(iii) Why does the ‘D-shaped’ contour used in the lectures for calculating such inte-
grals fail when we try to integrate
e −2i x
Z∞
2
dx?
−∞ x + 1
By choosing a different contour, explain how one could evaluate this integral
using Cauchy’s Residue Theorem.
Ex 6.2 Use Cauchy’s Residue Theorem to evaluate the following real integrals:
1 1
Z∞ Z∞
(i) 2 2
dx, (ii) 2 4
dx.
−∞ (x + 1)(x + 3) −∞ 28 + 11x + x
Ex 6.3 Consider the integral

Z2π
1
dt
0 13 + 5 cos t
2023–24
114 Exercises
(i) Use the substitution z = e i t to show that

Z2π
1 2 1
I
dt = 2
dz
0 13 + 5 cos t i C 1 5z + 26z + 5
where C 1 is the circle with centre 0, radius 1, oriented anticlockwise.
(ii) Show that f (z) = 1/(5z 2 +26z +5) has simple poles at z = −5 and z = −1/5. Show that
Res( f ; −1/5) = 1/24.
(iii) Use Cauchy’s Residue Theorem to show that
Z2π
1 π
dt = .
0 13 + 5 cos t 6
Ex 6.4 Convert the following real integrals into complex integrals around the unit circle in
the complex plane. Hence use Cauchy’s Residue Theorem to evaluate them.
Z2π Z2π
1
(i) 2 cos3 t + 3 cos2 t dt , (ii) dt .
0 0 1 + cos2 t
Ex 6.5 By considering the function
eiz
f (z) =
z 2 + 4z + 5
integrated around a suitable contour, show that
sin x −π sin 2
Z∞
2
dx = .
−∞ x + 4x + 5 e
4
= π4 /90.
P∞
Ex 6.6 Use the method of summation of series to show that n=1 1/n
3
Why doesn’t the method work for evaluating ∞ n=1 1/n ?
P
Ex 6.7 (This exercise uses Cauchy’s Residue Theorem to calculate an integral that is (I be-
lieve) impossible to calculate using common techniques of real analysis/calculus.)
Let 0 < a < b. Evaluate the integral:
x sin x
Z∞
dx
−∞ (x 2 + a 2 )(x 2 + b 2 )
by integrating a suitable function around a suitable contour.
Ex 6.8 Suppose a 6= 0. Consider the function

cot πz
z2 + a2
Show that this function has poles at z = n, n ∈ Z and z = ±i a. Calculate the residues at
these poles.
Hence show that
∞
X 1 π 1
= coth πa − 2 .
n=1 n2 + a2 2a 2a
2023–24
Ex 6.9 (The method used in §6.2 can be used evaluate other, more complicated, integrals.)
Let C 1 (t ) = e i t , 0 ≤ t ≤ 2π, denote the unit circle in C centred at 0 and with radius 1.
(i) Prove, using Cauchy Residue Theorem, that
ez
Z
dz = 2πi .
C1 z
(ii) By using the substitution z = e i t , prove that

Z2π Z2π
cos t
e cos(sin t ) dt = 2π, e cos t sin(sin t ) dt = 0.
0 0
Ex 6.10 (Sometimes, to calculate an indefinite integral, one has to be rather creative in pick-
ing the right contour.)
Let 0 < a < 1. Show that
e ax
Z∞
π
dx =
−∞ 1 + e x sin aπ
using the following steps.
(i) Show that this integral exists and is equal to its principal value.
(ii) Let f (z) = e az /(1 + e z ). Show that f is meromorphic with simple poles at z = (2k +
1)πi , k ∈ Z. Draw a diagram to illustrate where the poles are. Calculate the residue
Res( f ; πi ).
(iii) On the diagram from (ii), draw the contour ΓR = γ1,R + γ2,R + γ3,R + γ4,R where:
γ1,R is the horizontal straight line from −R to R,
γ2,R is the vertical straight line from R to R + 2πi ,
γ3,R is the horizontal straight line from R + 2πi to −R + 2πi ,
γ4,R is the vertical straight line from −R + 2πi to −R.
Which poles does ΓR wind around? Use Cauchy’s Residue Theorem to calculate
R
ΓR f .
(iv) Show, by choosing suitable parametrisations of the paths γ1,R and γ3,R and direct
computation, that γ3 f = −e 2πi a γ1 f .
R R
(v) Show, using the Estimation Lemma, that

Z Z
lim f = lim f = 0.
R→∞ γ2,R R→∞ γ4,R
(vi) Deduce the claimed result.

2023–24

ComplexAnalysis Notes

Uploaded by

ComplexAnalysis Notes

Uploaded by

Lecture Notes for

MATH29141 Complex Analysis

Lecturer: James Montaldi

Chapter 1 Complex functions, derivatives and integrals 3

Chapter 2 Power series 25

Chapter 3 Cauchy’s theorem & Cauchy’s formula 49

Chapter 4 Taylor series and Laurent series 67

Chapter 5 Cauchy’s Residue Theorem 85

Chapter 6 Applications of the residue theorem 103

Where we are going?

1 The answer is π (e −a − e −b ) if b 6= a as you’ll see in Part 2 of the course

Complex functions, derivatives and

1.1 Open sets and domains

In real analysis, one normally studies functions f : (a, b) → R, f : [a, b] → R or f : R → R

D(a; r ) := {z ∈ C | |z − a| < r } . (1.1)

Definition 1.2. A non-empty set D ⊂ C is said to be a domain if it is open and connected. A

F IGURE 1.3: A polygonal arc joining z and w in a domain D.

(i). C itself is a domain;

(ii). Any open disc D(a; r ) is a domain.

(iv). An annulus {z ∈ C | r 1 < |z − z 0 | < r 2 } is a domain.

(v). A half-plane such as {z ∈ C | Re(z) > a} is a domain.

(vi). A closed disc {z ∈ C | |z − z 0 | ≤ r } or a closed half-plane {z ∈ C | Re(z) ≥ a} are not

See Figure 1.4 for other examples of domains.

F IGURE 1.4: In (i), D is a domain (with a hole in it). In (ii) D = D 1 ∪ D 2 is not a

1.2 Complex functions and continuity

We will only consider functions defined on domains.

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

(ii). Let f (z) = z̄. Then f (x + i y) = x − i y. Here u(x, y) = x, v(x, y) = −y.

We will be interested in functions which do behave nicely when taking limits.

We say that f is continuous on D if it is continuous at z 0 for all z 0 ∈ D. ✯

f (z) + g (z), f (z)g (z), c f (z) (c ∈ C)

are all continuous at z 0 , as is f (z)/g (z) provided that g (z 0 ) 6= 0.

• C∗ = C \ {0}, often called the punctured plane

• D(z 0 ; R) the disc of radius R and centre z 0

• D ∗ (z 0 ; R) the punctured disc of radius R and centre z 0 , equal to D(z 0 , ; R) \ {z 0 }

• A(z 0 ; R 1 , R 2 ) = {z ∈ C | R 1 < |z − z 0 | < R 2 }, the annulus with centre z 0 and radii

1.3 Differentiable functions

Definition 1.7. Let D ⊂ C be a domain and let f : D → C be a function. Let z 0 ∈ D. We say

One can also use the notation

Remark Although the definition of differentiability of a function in complex analysis is,

Differentiability is a very strong property for a complex function to possess; it is much

Definition 1.8. Suppose that f : D → C is differentiable on a domain D. Then we say that f

Example 1.9. Let f (z) = z 2 , defined on C. Let z 0 ∈ C be any point. Then

Hence f ′ (z 0 ) = 2z 0 for all z 0 ∈ C. Thus f is differentiable at every point in C and so is a

Proposition 1.10. Let f , g be holomorphic on a domain D. Let c ∈ C. Then the following

(i) sum rule: ( f + g )′ = f ′ + g ′ ,

(ii) scalar rule: (c f )′ = c f ′ ,

(iii) product rule: ( f g )′ = f ′ g + f g ′ ,

(v) chain rule: ( f ◦ g )′ = f ′ ◦ g · g ′ .

Proposition 1.11. Suppose that f is differentiable at z 0 . Then f is continuous at z 0 .

Proof: ¡To show that ¢

1.3A What is the derivative?

f (z + h) − f (z) = f ′ (z)h + o(h).

1.3B The Cauchy-Riemann equations

Throughout, let D be a domain. Let z = x + i y ∈ D. Let f : D → C be a complex valued

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

where u, v : D → R are real-valued functions.

Example 1.12. Let f (z) = z 3 . Then

f (z) = z 3 = (x + i y)3 = x 3 − 3x y 2 + i (3x 2 y − y 3 ) = u(x, y) + i v(x, y)

where u(x, y) = x 3 − 3x y 2 and v(x, y) = 3x 2 y − y 3 .

Recall definition of partial derivatives:

If f is differentiable, then the Cauchy-Riemann equations give two relationships between

They are often known as the C-R equations. ✯

Let h be real and consider z 0 + h = (x 0 + h) + i y 0. Then as h → 0 we have z 0 + h → z 0 .

Now take k to be real and consider z 0 +i k = x 0 +i (y 0 +k). Then as k → 0 we have z 0 +i k →

Remark The statement of the theorem requires f to be holomorphic in a domain D. How-

1.4 A first look at integration

z 0 to z 1 (defined below to be a continuous curve starting at z 0 and ending at z 1 ). Then, using

1.4A Paths and contours

First we need to make precise what we mean by a path.

Remark Note that a path is a function. Sometimes, it is convenient to regard a path as a

As an example of a path, let z 0 , z 1 ∈ C. Define

Example 1.18. An important example of a closed path is given by

γ(t ) = cos t + i sin t , 0 ≤ t ≤ 2π. (1.7)

γ(t ) = z 0 + r (cos(t ) + i sin(t ), 0 ≤ t ≤ 2π. (1.8)

Definition 1.19. A path γ is said to be smooth if γ : [a, b] → C is differentiable and γ′ is con-