Introduction to Complex Numbers

Richard Earl

Michaelmas Term 2024

SYLLABUS:
Complex numbers and their arithmetic. The Argand diagram (complex plane). Modulus and
argument of a complex number. Simple transformations of the complex plane. De Moivre’s Theorem;
roots of unity. Euler’s theorem; polar form reiθ of a complex number. Polynomials and a statement
of the Fundamental Theorem of Algebra.
READING LIST:

• Richard Earl – Complex Numbers on [Link]

study/bridging-gap

• Richard Earl – Towards Higher Mathematics, Cambridge, 2017, Chapter 1.

FURTHER EADING:

• Roger Fenn – Geometry, Springer, Chapter 4

• Liang-shin Hahn – Complex Numbers and Geometry, Mathematical Association of America,

1994, Chapter 1.

1 The Need for Complex Numbers

It is well known that the two roots of the quadratic equation ax2 + bx + c = 0 (a, b, c real and a ̸= 0)
are √
−b ± b2 − 4ac
x= (1)
2a
and mathematicians have been solving quadratic equations since the time of the Babylonians. When
the discriminant b2 − 4ac is positive then these two roots are real and distinct; graphically they are
where the curve y = ax2 + bx + c cuts the x-axis. When b2 − 4ac = 0 then we have one real root and
the curve just touches the x-axis here. But what happens when b2 − 4ac < 0? Then there are no real
solutions to the equation as no real number squares to give the negative b2 − 4ac. From the graphical

1
point of view the curve y = ax2 + bx + c lies entirely above or below the x-axis.

distinct real roots repeated real root complex roots

It is only comparatively recently that mathematicians have been comfortable with the roots in (1)
when b2 − 4ac < 0. During the Renaissance the quadratic would
√ have been considered unsolvable or its
roots would have been called imaginary. But if we imagine −1 to exist, and that it behaves (adds
and multiplies) much the same as other numbers, then the two roots in (1) can be written in the form
√
x = A ± B −1 (2)
√
where A = −b/(2a) and B = 4ac − b2 /(2a) are real numbers. But what meaning can such roots
have? It was this philosophical point which pre-occupied mathematicians until the start of the 19th
century when these ‘imaginary’ numbers started proving so useful (especially in the work of Cauchy)
that the philosophical concerns ultimately became side-issues.

Notation 1 We shall, from now on, write

√
i= −1

though many books, particularly those written for engineers and physicists, use j instead. The nota-
tion i was first introduced by the Swiss mathematician Leonhard Euler in 1777. (See p.8 for a brief
biography.)

Definition 2 A complex number is a number of the form a + bi where a and b are real numbers.
We will usually denote a complex number with a single letter like z or w. If z = a + bi, where a and
b are real, then a is known as the real part of z and b as the imaginary part. We write

a = Re z and b = Im z.

When we write ‘let z = a + bi’ we will implicitly assume that a and b are real so that a = Re z and
b = Im z.

• Note that real numbers are complex numbers; a real is just a complex number with zero imaginary
part.

Notation 3 We write C for the set of all complex numbers.

One of the first major results concerning complex numbers, and which conclusively demonstrated
their usefulness, was proved rigorously by Argand in 1806. From the quadratic formula (1) we know
that all quadratic equations can be solved using complex numbers – what Gauss was the first to prove
was the much more general result:

2
Theorem 4 (Fundamental Theorem of Algebra) Let p(z) = a0 + a1 z + · · · + an z n be a polynomial
of degree n ⩾ 1 with real (or complex) coefficients ak . Then the roots of the equation p(z) = 0 are
complex. That is, there are n (not necessarily distinct) complex numbers γ1 , . . . , γn such that

a0 + a1 z + · · · + an z n = an (z − γ1 )(z − γ2 ) · · · (z − γn ).

In particular the theorem shows that a degree n polynomial has, counting repetitions, n roots in C.

The proof of this theorem is far beyond the scope of this text. Note that the theorem only guaran-
tees the existence of the roots of a polynomial somewhere in C unlike the quadratic formula which
determines exactly the roots. The theorem gives no hints as to where in C these roots are to be found.

2 Their Algebra
We add, subtract, multiply and divide complex numbers in obvious sensible ways. To add or subtract
complex numbers, we just add or subtract their real and imaginary parts:

(a + bi) + (c + di) = (a + c) + (b + d)i; (a + bi) − (c + di) = (a − c) + (b − d)i. (3)

Note that these equations are entirely comparable with adding or subtracting two vectors in the xy-
plane. Unlike with vectors, we can also multiply complex numbers by expanding the brackets in the
usual fashion and remembering that i2 = −1:

(a + bi)(c + di) = ac + bci + adi + bdi2 = (ac − bd) + (ad + bc)i. (4)

To divide complex numbers, we firstly note that (c + di)(c − di) = c2 + d2 is real. So

   
a + bi a + bi c − di ac + bd bc − ad
= × = + i. (5)
c + di c + di c − di c2 + d2 c2 + d2
Example 5 Calculate, in the form a + bi, the following complex numbers.

(1 + 3i) + (2 − 6i), (1 + 3i) − (2 − 6i), (1 + 3i)(2 − 6i), (1 + 3i)/(2 − 6i).

Solution

(1 + 3i) + (2 − 6i) = (1 + 2) + (3 + (−6))i = 3 − 3i;

(1 + 3i) − (2 − 6i) = (1 − 2) + (3 − (−6))i = −1 + 9i;
(1 + 3i)(2 − 6i) = 2 + 6i − 6i − 18i2 = 2 + 18 = 20.

Division takes a little more care, and we need to remember to multiply both numerator and denomi-
nator by 2 + 6i.
1 + 3i (1 + 3i)(2 + 6i) 2 + 6i + 6i + 18i2 −16 + 12i 2 3
= = 2 2
= = − + i.
2 − 6i (2 − 6i)(2 + 6i) 2 +6 40 5 10

Remark 6 Division of complex numbers √ is very√ similar to the method√of rationalizing a surd. Recall
that to write a quotient such as (2+ 3 2)/(1+ 2 2) in the form q1 +q√2 2 where q1 and q2 are rational
numbers, then we multiply the numerator and denominator by 1 − 2 2 to get
√ √ √ √ √
2+3 2 2+3 2 1−2 2 2 + 3 2 − 4 2 − 12 10 1 √
√ = √ × √ = = + 2. ■
1+2 2 1+2 2 1−2 2 1−8 7 7

3
The number c − di,used in (5), as relating to c + di, has a special name and some useful properties
(Proposition 18).

Definition 7 Let z = a + bi. The conjugate of z is the number a − bi and is denoted as z̄ (or
sometimes as z ∗ ).

• Note z is real if and only if z = z̄. Also

z + z̄ z − z̄
Re z = ; Im z = . (6)
2 2i

• Note that two complex numbers are equal if and only if their real and imaginary parts correspond.
That is
z = w if and only if Re z = Re w and Im z = Im w.
This is called comparing real and imaginary parts.

• Note from equation (2) that when the real quadratic equation ax2 +bx+c = 0 has complex roots
then these roots are conjugates of one another. More generally, if z0 is a root of the polynomial
an z n + an−1 z n−1 + · · · + a0 = 0, where the coefficients ak are real, then its conjugate z0 is also
a root (Corollary 19).

We present the following problem because it highlights a potential early misconception involving
complex numbers – if we need a new number i as the square root of −1, then shouldn’t we need
another one for the square root of i? However z 2 = i is just another polynomial equation, with
complex coefficients, and two roots in C are guaranteed by the fundamental theorem of algebra. They
are also quite straightforward to calculate.

Example 8 Find all those complex numbers z that satisfy z 2 = i.

Solution Suppose that z 2 = i and z = a + bi. Then i = (a + bi)2 = (a2 − b2 ) + 2abi. Comparing real
and imaginary parts in the above we obtain two simultaneous equations

a2 − b2 = 0 and 2ab = 1.
√
So b = ±a from√the first equation. Substituting b = a into the second equation gives a = b = 1/ 2
or a = b = −1/ 2. Substituting b = −a into the second equation gives −2a2 = 1 which has√no real
solution a. 2
√ So the two z which satisfy z = i, i.e. the two square roots of i, are (1 + i)/ 2 and
(−1 − i)/ 2.

Remark 9 Notice, as with the square roots of real numbers, that the two square roots of i are negative
one another, and this is generally the case. However, it’s typically the case that neither of the two
√
square roots is more preferential than the other. So we reserve the notation z, denoting a preferred
√
choice of square root of z, for a positive number z and define z to be the positive root of z in such a
case. Further the following rules

(xy)r = xr y r , xr xs = xr+s

hold true for positive x, y and real r, s. It’s unclear what these rules might mean for complex numbers,
let alone whether they are true. Below is a famous fallacy of mathematics which shows care needs to
be taken with powers of complex numbers.

4
Example 10 Clearly r r
−1 1 −1 1
= so that =
1 −1 1 −1
p √ √
and as a/b = a/ b then √
√
i −1 1 1
= √ =√ = .
1 1 −1 i
2 2
This rearranges to give i = 1, which is plainly false as i = −1.
Example 11 Use the quadratic formula to find the two solutions of z 2 − (3 + i)z + (2 + i) = 0.
Solution We have a = 1, b = −3 − i and c = 2 + i (in the notation of the quadratic formula). So
b2 − 4ac = (−3 − i)2 − 4 × 1 × (2 + i) = 9 − 1 + 6i − 8 − 4i = 2i.
√ √
Knowing that i = ±(1 + i)/ 2 from Example 8, we see that the two square roots of 2i are ±(1 + i).
So
√ √
−b ± b2 − 4ac (3 + i) ± 2i (3 + i) ± (1 + i) 4 + 2i 2
z= = = = or = 2 + i or 1.
2a 2 2 2 2
Note that the two roots are not conjugates of one another – this need not be the case when the
coefficients a, b, c are not all real.

3 The Argand Diagram

The real numbers are often represented on the real line, each point of which corresponds to a unique
real number. This number increases as we move from left to right along the line. The complex
numbers, having two components, their real and imaginary parts, can be represented on a plane;
indeed C is sometimes referred to as the complex plane, but more commonly, when we represent C
in this manner, we call it the Argand diagram 1 .
In the Argand diagram the point (a, b) represents the complex number a + bi so that the x-axis
contains all the real numbers, and so is termed the real axis, and the y-axis contains all those complex
numbers which are purely imaginary (i.e. have no real part) and so is referred to as the imaginary
axis.

the Argand diagram cartesian and polar co-ordinates

1
After the Swiss mathematician Jean-Robert Argand (1768-1822), although the Norwegian mathematician Caspar
Wessel (1745-1818) had previously had the same idea for representing complex numbers, but his work went unnoticed.

5
Remark 12 We can think of z0 = a + bi as a point in an Argand diagram, but it is often useful to
think of it as a vector as well. Adding z0 to another complex number translates that number by the
vector (a, b). That is the map z 7→ z + z0 represents a translation a units to the right and b units up
in the complex plane. Note also that the conjugate z̄ of a point z is its mirror image in the real axis.
So the map z 7→ z̄ represents reflection in the real axis. We shall discuss in more detail the geometry
of the Argand diagram in §5. ■

A complex number z in the complex plane can be represented by Cartesian co-ordinates, its real
and imaginary parts, x and y, but equally useful is the representation of z by polar co-ordinates,
r and θ. If we let r be the distance of z from the origin and, if z ̸= 0, we let θ be the angle that the
line connecting z to the origin makes with the positive real axis (see the right diagram above), then
we can write
z = x + yi = r cos θ + (r sin θ)i. (7)
The relations between z’s Cartesian and polar co-ordinates are simple:
p
x = r cos θ and y = r sin θ; r = x2 + y 2 and tan θ = y/x.

Definition
p 13 The number r is called the modulus of z and is written |z| . If z = x + yi then
|z| = x + y 2 .
2

• Note that |z| ⩾ 0 for all z and that if |z| = 0 then z = 0. Note also

|Re z| ⩽ |z| and |Im z| ⩽ |z| . (8)

Definition 14 The number θ is called the argument of z and is written arg z. If z = x + yi then
y x y
sin arg z = p ; cos arg z = p ; tan arg z = .
x + y2
2 x + y2
2 x

Note that the argument of 0 is undefined. Note also that arg z is defined only up to multiples of 2π.
For example, the argument of 1 + i could be π/4 or 9π/4 or −7π/4 etc.. Here π/4 would be the
preferred choice as for definiteness we shall take the principal values for argument to be in the
range 0 ⩽ θ < 2π.

Notation 15 For ease of notation we shall write cis θ for cos θ + i sin θ, so that complex numbers in
polar form as in (7) will now be written z = r cis θ.

Remark 16 In due course we will prove Euler’s famous result, eiθ = cos θ + i sin θ, and so write the
polar form as reiθ after that. ■

Proposition 17 Let α and β be real numbers. Then

cis(α + β) = cis(α) cis(β); cis α = cis(−α) = (cis α)−1 .

Proof Recalling the formulae for cos(α + β) and sin(α + β) we have

cis(α) cis(β) = (cos α + i sin α)(cos β + i sin β)

= (cos α cos β − sin α sin β) + i(sin α cos β + cos α sin β)
= cos(α + β) + i sin(α + β) = cis(α + β).

6
From this we then have cis(−α)cis α = cis (−α + α) = cis(0) = 1, and finally
cis α = cos α + i sin α = cos α − i sin α = cos(−α) + i sin(−α) = cis(−α).

We now prove some important formulae about properties of the modulus, argument and conjuga-
tion.
Proposition 18 The following identities and inequalities hold for complex numbers z, w.

|zw| = |z||w| (9) z ± w = z̄ ± w̄ (15)

|z/w| = |z|/|w| if w ̸= 0 (10) zw = z̄ w̄ (16)
arg(zw) = arg z + arg w if z, w ̸= 0 (11) z/w = z̄/w̄ if w ̸= 0 (17)
arg(z/w) = arg z − arg w if z, w ̸= 0 (12) |z̄| = |z| (18)
arg z̄ = − arg z if z ̸= 0 (13) |z + w| ⩽ |z| + |w| (19)
2
z z̄ = |z| (14) ||z| − |w|| ⩽ |z − w| (20)

(19) is known as the triangle inequality.

Proof We prove here a selection of these identities. The remainder are left as exercises.
• Identity (9): |zw| = |z| |w| . Let z = a + bi and w = c + di. Then zw = (ac − bd) + (bc + ad)i so
that
p p
|zw| = (ac − bd)2 + (bc + ad)2 = a2 c2 + b2 d2 + b2 c2 + a2 d2
p p p
= (a2 + b2 )(c2 + d2 ) = a2 + b2 c2 + d2 = |z| |w| .

• Identity (11): arg(zw) = arg z + arg w. Let z = r cis θ and w = R cis Θ. Then
zw = (r cis θ)(R cis Θ) = rR cis θ cis Θ = rR cis (θ + Θ),
by Proposition 17. We can read off that |zw| = rR = |z| |w| , which is a second proof of (9), and
also that
arg(zw) = θ + Θ = arg z + arg w up to multiples of 2π.
• Identity (16): zw = z w. With z = a + bi, w = c + di then
zw = (ac − bd) + (bc + ad)i = (ac − bd) − (bc + ad)i = (a − bi)(c − di) = zw.

• Identity (19): the triangle inequality |z + w| ⩽ |z| + |w|.

Im A geometric proof of this is simple, explains the in-

equality’s name, and is represented diagrammatically
w
z to the left.
Note that the shortest distance between 0 and z + w
z w is |z + w|. This is the length of one side of the triangle
with vertices 0, z, z + w and so is shorter in length
Re
than the path which goes straight from 0 to z then
straight on to z + w. The total length of this second
w
path is |z| + |w|.
For an algebraic proof, we note for complex numbers
z, w using (6), (8), (9) and (18) that

7
 zw + zw
= Re(zw) ⩽ |zw| = |z| |w| = |z| |w| . (21)
2
Hence, using (14), (15), (21) we have

|z + w|2 = (z + w)(z + w)
= (z + w)(z + w)
= zz + zw + zw + ww
⩽ |z|2 + 2 |z| |w| + |w|2 = (|z| + |w|)2 ,

to give the required result. There is equality only if zw ⩾ 0 so that Re(zw) = |zw|; if z and w
are non-zero this is equivalent to requiring z = µw for some µ > 0, meaning that 0, z, w are on
the same half-line from 0.

Leonhard Euler (1707-1783) (pronounced ”oil-er”) was a

prolific Swiss mathematician (over 800 papers bear his name)
and indisputably the greatest mathematician of the 18th cen-
tury. He made major contributions in many areas of mathe-
matics, but especially in the study of infinite series and in par-
ticular he determined the sum of the series 1 + 41 + 19 + 16
1
+· · ·
π2
to be 6 (the so-called Basel problem). Euler’s Identity
eiθ = cos θ + i sin θ, and in particular eiπ = −1, can be
proved using infinite series for the exponential and trigono-
metric functions.

But Euler’s name appears widely throughout pure and applied mathematics: Euler’s equation is
fundamental in the Calculus of Variations (which involves problems such as showing the shortest
curve between two points is a straight line); Euler’s Method is the simplest numerical method for
computing approximate solutions of differential equations; Euler’s Theorem in number theory is a
generalization of Fermat’s Little Theorem; three important centres of a triangle lie on the Euler
line. Euler also produced some of the first topological results: he famously showed that it was
impossible to traverse the seven bridges in Königsberg without repetition, in fact determining more
generally which networks (or graphs) can be traversed. He also showed for a (convex) polyhedron
that V − E + F = 2 where V, E, F respectively denote the number of vertices (corners), edges, faces
that the polyhedron has. Both these results depend on shape (e.g. how the points are connected)
rather than geometry (e.g. the lengths of the edges).

Corollary 19 (Conjugate Pairs) Suppose that z0 is a root of the degree n polynomial p(z) =
an z n + · · · + a1 z + a0 where the coefficients ak are all real. Then the conjugate z0 is also a root of
p(z). Consequently
p(z) = an (z − α1 ) · · · (z − αr )q1 (z) · · · qs (z)
where α1 , . . . αr are the real roots of p(z) and q1 (z), . . . , qs (z) are real quadratic polynomials with
conjugate complex roots. In particular, an odd degree real polynomial has at least one real root.

8
Proof Note

p(z0 ) = an (z0 )n + an−1 (z0 )n−1 + · · · + a1 z0 + a0

= an (z0 )n + an−1 (z0 )n−1 + · · · + a1 z0 + a0 [as the coefficients ak are real]
= an z0n + an−1 z0n−1 + · · · + a1 z0 + a0 [using (15) and (16)]
= p(z0 ) = 0̄ = 0.

The fundamental theorem of algebra tells us a polynomial’s roots can be found amongst the complex
numbers. So the roots of p(z) are either real, call these α1 , . . . , αr or come in conjugate complex pairs
β1 , β1 , . . . , βs , βs . Now

(z − βk )(z − βk ) = z 2 − (βk + βk )z + βk βk = z 2 − (2 Re βk )z + |βk |2 .

If we denote this real quadratic as qk (z) then p(z) = an (z−α1 ) · · · (z−αr )q1 (z) · · · qs (z) where n = r+2s
from equating the degrees of the polynomials. Note that if n is odd then r ⩾ 1 and so p(z) has at
least one real root.

4 Roots of Unity

Im
z0 3
z0 2 Consider the complex number z0 = cisθ where θ is
some real number in the range 0 ⩽ θ < 2π. The
z0 modulus of z0 is then 1 and the argument of z0 is θ.
1 1 Re
In Proposition 18 we proved that |zw| = |z| |w| and
for z, w ̸= 0 that arg(zw) = arg z + arg w. Hence

|z n | = |z|n and arg(z n ) = n arg z

for any integer n and z ̸= 0. So the modulus of (z0 )n is

1n = 1, and the argument of (z0 )n is nθ up to multiples
of 2π. Putting this another way, we have the following
powers of z0 famous theorem due to De Moivre.

Theorem 20 (De Moivre’s theorem 2 ) For a real number θ and integer n we have that

cos nθ + i sin nθ = (cos θ + i sin θ)n ,

or more succinctly, cis(nθ) = (cis θ)n .

Example 21 Find expressions for cos 4θ and sin 4θ.

2
Abraham De Moivre (1667-1754), a French protestant who fled religious persecution in France to move to England,
is best remembered for this formula but he also made important contributions in probability which appeared in his The
Doctrine Of Chances (1718).

9
Solution Writing c for cos θ and s for sin θ we have

cos 4θ + i sin 4θ = (c + is)4 = c4 + 4ic3 s − 6c2 s2 − 4ics3 + s4 .

Comparing real and imaginary parts we have

cos 4θ = c4 − 6c2 (1 − c2 ) + (1 − c2 )2 = 8c4 − 8c2 + 1; sin 4θ = 4c3 s − 4c(1 − c2 )s = 8c3 s − 4cs.

We now apply the ideas from the proof of De Moivre’s theorem to the following problem, that of

finding the roots of z n = 1 where n ⩾ 1 is an integer.

We know from the fundamental theorem of algebra that there are (counting multiplicities) n solutions
– these are known as the nth roots of unity. Let’s first solve directly z n = 1 for n = 2, 3, 4.

• n = 2: we have 0 = z 2 − 1 = (z − 1)(z + 1) and so the square roots of 1 are ±1.

√
• n = 3: 0 = z 3 − 1 = (z − 1)(z 2 + z + 1). The cube roots of unity are then 1 and (−1 ± i 3)/2.

• n = 4: 0 = z 4 − 1 = (z 2 − 1)(z 2 + 1), so that the fourth roots of 1 are ±1 and ±i.

Plotting these roots on Argand diagrams we can see a pattern developing

1.5
1.5 1.5

1.0
1.0 1.0

0.5
0.5 0.5

2 1 1 2
2 1 1 2 2 1 1 2

0.5
0.5 0.5

1.0
1.0 1.0

1.5
1.5 1.5

square roots cube roots fourth roots

Returning to the general case, suppose that

z = r cis θ and satisfies z n = 1.

Then z n has modulus rn and has argument nθ, whilst 1 has modulus 1 and argument 0. Comparing
their moduli we see rn = 1 and hence r = 1 (as r > 0). Comparing arguments we see nθ = 0 up to
multiples of 2π. That is, nθ = 2kπ for some integer k giving θ = 2kπ/n. So if z n = 1 then z has the
form z = cis(2kπ/n) where k is an integer. At first glance there seems to be an infinite number of
roots but, as cosine and sine have period 2π, then cis(2kπ/n) repeats with period n. Hence we have
shown:

Proposition 22 (Roots of Unity) The nth roots of unity, that is the solutions of the equation
z n = 1, are
z = cis(2kπ/n) where k = 0, 1, 2, . . . , n − 1.

10
 • When plotted these nth roots of unity form a regular n-gon inscribed within the unit circle with
a vertex at 1. More generally, for c ̸= 0, the n solutions of z n = c make a regular n-gon inscribed
in the circle |z| = |c|1/n .

Example 23 Find all the solutions of the cubic z 3 = −2 + 2i.

√
Solution If we write −2 + 2i in its polar form we
√ have −2 + 2i = 8 cis(3π/4). So if z 3 = −2 + 2i
and z has modulus r and argument θ then r3 = 8 and 3θ = 3π/4 up to multiples of 2π,which gives
√
r = 2 and θ = π/4 + 2kπ/3 for some integer k.

As before we need only consider k = 0, 1, 2 (as other values of k lead to repeats) and so the three cube
roots are
√ √ √
π     
11π 19π
2 cis = 1 + i, 2 cis , 2 cis .
4 12 12

• Note that the above method generalizes naturally to solving any equation of the form z n = c
where c is a complex number and n ⩾ 1 is an integer.

5 Their Geometry
Using the modulus and argument functions, we can measure distances and angles in the Argand
diagram.

Definition 24 Given two complex numbers z = z1 + z2 i and w = w1 + w2 i then the distance between
them is |z − w| . This follows from Pythagoras’ Theorem as
p
(z1 − w1 )2 + (z2 − w2 )2 = |(z1 − w1 ) + (z2 − w2 )i| = |z − w| .

Definition 25 Given three complex numbers a, b, c then the angle ∡abc equals
 
c−b
arg(c − b) − arg(a − b) = arg .
a−b

5
Im On the left we see the distance
from 1 + i and 5 + 4i is
4 5 4i p
(5 − 1)2 + (4 − 1)2 = 5.
3

3 In the right diagram a = 2+i,

2
b = 1 + 2i and c = 3 + 4i. The
angle at b is
 
1 1 i
4 (3 + 4i) − (1 + 2i)
arg
(2 + i) − (1 + 2i)
Re
0
0 1 2 3 4 5 6 = arg(2i) = π/2.

11
 • Notice that Definition 25 gives a signed angle in the sense that it is measured in an anti-clockwise
fashion from the segment ba round to the segment bc, and the above formula would give negative
this result if the roles of a and c were swapped. Note also that this signed angle is defined only
up to multiples of 2π.

Example 26 Find the smaller angle ∡cab where a = 1 + i, b = 3 + 2i and c = 4 − 3i.

Solution The angle ∡cab is given by

     
b−a 2+i 2 + 11i
arg = arg = arg .
c−a 3 − 4i 25

So tan ∡cab = 11/2 with the number (2 + 11i)/25 clearly in the first quadrant. (There is a solution
to tan θ = 11/2 in the opposite quadrant.) Hence ∡cab = tan−1 (11/2) ≈ 1.39 radians.
We can use these definitions of distance and angle to describe various regions of the plane.

Example 27 Describe the following regions of the Argand diagram.

• |z − a| = r, where a is a complex number and r > 0.

This is clearly the locus of points at distance r from a, i.e. the circle with centre a and radius r. The
regions |z − a| < r and |z − a| > r are then the interior and exterior of the circle respectively.

• |z − a| = |z − b|, where a, b are distinct complex numbers.

This is the set of points equidistant from a and b and so is the perpendicular bisector of the line segment
connecting a and b. The regions |z − a| < |z − b| and |z − a| > |z − b| are then the half-planes either
side of this line and which respectively contain a and b.

• arg(z − a) = θ, where a is a complex number and 0 ⩽ θ < 2π.

This is a half-line emanating from the point a and making an angle θ with the positive real axis. Note
that it doesn’t include the point a itself.

Example 28 Describe the set of points z which satisfy the equation

 
z+1 π
arg = . (22)
z−1 2

Solution Method One: parametrization. We can rewrite (22) as

2 z+1
1+ = = it where t > 0,
z−1 z−1
as these points it are precisely those with argument π/2. Solving for z we have
 2   
1 + ti t −1 −2t
z= = 2 +i . (23)
−1 + ti t +1 1 + t2

It is quite easy to spot, from the first expression for z in (23), that the denominator and numerator
have the same modulus and so |z| = 1. Those with knowledge of the half-angle tangent formulae may
also have spotted them in (23).

12
 So z must lie on the unit circle |z| = 1. We see that the real part x(t) in (23) varies across the
range −1 < x(t) < 1 as t varies across the positive numbers, but the imaginary part y(t) is always
negative. Thus (22) is the equation of the lower semicircle from |z| = 1, not including the points −1
and 1.
Method Two: a geometric approach. In the diagram on the left
arg(z − 1) = −α and arg(z + 1) = −β where α, β represent the
Im 1.5
actual magnitudes of the angles irrespective of their sense. By
(12) we have
1.0 z 1
arg((z + 1)/(z − 1)) = π/2 ⇐⇒ arg(z + 1) − arg(z − 1) = π/2
0.5

⇐⇒ −β − (−α) = π/2
Re
1.5 1.0 0.5 0.5 1.0 1.5 ⇐⇒ β + (π − α) = π/2
0.5
i.e. when the angle at z is a right angle, which holds when −1,
z 1.0
1 is the diameter of the circle through −1, 1, z. This is Thales’
Theorem. It can be similarly shown that arg((z + 1)/(z − 1)) =
1.5
−π/2 on the upper semicircle.
An important class of maps of the complex plane is that of the isometries, i.e. the distance
preserving maps.

Definition 29 We say a map f from C to C is an isometry if it preserves distance, that is

|f (z) − f (w)| = |z − w| for any complex z, w.

Example 30 Of particular note are the following three isometries.

• z 7→ (cis θ) z which is rotation anticlockwise by θ about 0;

in particular z 7→ iz is rotation by a right angle anticlockwise about 0.

• z 7→ z + k which is translation by Re k to the right, and Im k up.

• z 7→ z̄ which is reflection in the real axis.

Example 31 The triangle abc in C (with the vertices taken in anticlockwise order) is equilateral if
and only if a + ωb + ω 2 c = 0 where ω = cis(2π/3) is a cube root of unity.

Solution Note first that 1 + ω + ω 2 = 0. The triangle abc is equilateral if and only if c − b is the side
b − a rotated through 2π/3 anticlockwise – i.e. if and only if

c − b = ω(b − a)
c b c b ⇐⇒ ωa − (1 + ω)b + c = 0
⇐⇒ ωa + ω 2 b + c = 0
⇐⇒ a + ωb + ω 2 c = 0.
b a
a b

13
Proposition 32 (Isometries of C) Let f be an isometry of C.
If f is orientation-preserving then f has the form f (z) = az + b for some complex a, b with |a| = 1.
These maps are all rotations. If f reverses orientation then f has the form f (z) = az̄ + b for
some complex a, b with |a| = 1. These include, but are not limited to, reflections.

Proof This result (and more generally all the isometries of Euclidean space) are classified the Geometry
course this term.

Example 33 Express in the form f (z) = az̄ + b reflection in the line x + y = 1.

Solution Knowing from Proposition 32 that the reflection has the form f (z) = az̄ + b we can find a
and b by considering where two points are mapped. As 1 and i both lie on the line of reflection then
they are both fixed. So

1 = a1̄ + b = a1 + b; i = ai + b = −ai + b.

Substituting b = 1 − a into the second equation we find a = −i and b = 1 + i. Hence f (z) = −iz + 1 + i.

There are other important maps of the complex plane which aren’t isometries. Here are some
examples which involve determining the images of certain regions. The methods used involve either
relying on the fact that the map in question has an inverse or parametrizing the region with Cartesian
or polar co-ordinates.

Example 34 In the two cases below, find the image in C of the given subset under the given map.

• A is the region Im z > Re z > 0 and f (z) = z 2 .

A general point in A can be 2

Im
2
Im
written in the form z = rcisθ
where π/4 < θ < π/2 and A
r > 0. 1 f A 1

De Moivre’s Theorem gives us

that z 2 = r2 cis2θ which has an Re Re
argument of 2θ in the range 2 1 1 2 2 1 1 2

π/2 < 2θ < π and indepen-

dently a modulus of r2 in the 1 1

range r2 > 0.
So f (A) is the second quad-
rant. 2 2

• B is the unit disc |z| < 1 and g(z) = 1+z

1−z .
The map g maps from the set {z ∈ C : z ̸= 1} to {z ∈ C : z ̸= −1}, with an inverse g −1 (z) =
(z − 1)/(z + 1). To see this we note

w = (1 + z)/(1 − z) ⇔ w − wz = 1 + z ⇔ w − 1 = z(1 + w) ⇔ z = (w − 1)/(w + 1).

14
 Note z is in g(B) 2 2
Im Im
⇔ g −1 (z) is in B
gB
⇔ |(z − 1)/(z + 1)| < 1
⇔ |z − 1| < |z + 1| 1 1
g Im
which is the half-plane with B
points closer to 1 than −1, Re Re
2 1 1 2 2 1 1 g Re 2
(Example 27) or equiva-
lently the half-plane Rez >
0. 1 1

2 2

We now prove a selection of basic geometric facts. Here is a reminder of some identities which will
prove useful.
z + z̄ Re z Im z
Re z = ; z z̄ = |z|2 ; cos arg z = ; sin arg z = . (24)
2 |z| |z|

Theorem 35 (Cosine Rule). Let ABC be a triangle with angles Â, B̂, Ĉ. Then

|BC|2 = |AB|2 + |AC|2 − 2 |AB| |AC| cos Â. (25)

Proof We can introduce co-ordinates in the plane so that A is at the origin and B is at 1. Let C be
at the point z. So in terms of our co-ordinates: |AB| = 1, |BC| = |z − 1| , |AC| = |z| , Â = arg z.
Hence, by (24),

RHS of (25) = 1 + |z|2 − 2 |z| cos arg z = 1 + z z̄ − 2 |z| × Re z

|z|
= 1 + z z̄ − 2 × (z + z̄)/2 = 1 + z z̄ − z − z̄
= (z − 1)(z̄ − 1) = LHS of (25).

Whilst a simple enough theorem to prove in a geometric manner, Thales’ Theorem is arrived at
nicely with the use of complex numbers, and the theorem’s converse also comes naturally using algebra.

Theorem 36 (Thales 3 ) Let A, B be distinct points in the plane and let C be the circle with AB as
a diameter. Then the point P lies on C if and only if the angle ∡AP B is a right angle.

Proof Without loss of generality we may assume that A, B, P have complex co-ordinates −1, 1, z
respectively so that the circle C is the circle |z| = 1. By the cosine rule, ∡AP B is a right angle if and
only if
22 = |z − 1|2 + |z + 1|2 ⇐⇒ 4 = 2z z̄ + 2 ⇐⇒ |z| = 1.

3
After the Greek mathematician and philosopher Thales (c.624BC–c.547BC). Thales might reasonably be considered
the first Western philosopher – in that he sought to explain phenomena without reference to mythology – though it is
unclear to what extent he was able to prove the various theorems of geometry that are attributed to him. At that time
it is more likely that these results were appreciated as useful rules of thumb employed by engineers.

15
6 Euler’s Identity
The exponential function exp z and trigonometric functions, sin z and cos z, can be defined for complex
numbers by the power series
∞ ∞ ∞
X zn X (−1)n z 2n X (−1)n z 2n+1
exp z = , cos z = , sin z = .
n! (2n)! (2n + 1)!
n=0 n=0 n=0

A rigorous treatment of these functions appears in the Analysis I course this term, but we will
somewhat informally discuss properties of these functions.

• These power series converge for all complex numbers z.

• Each function is differentiable and

exp′ z = exp z, cos′ z = − sin z, sin′ z = cos z.

• Cosine is an even function and sine is an odd function. Both have period 2π.
• The identity
exp(z + w) = exp z × exp w
holds for all complex numbers z, w.
• The identity
cos2 z + sin2 z = 1
holds for all complex numbers z. This does not imply that sin z and cos z are bounded functions.
In fact they are unbounded functions.
• Euler’s Identity
exp z = cos z + i sin z
holds for all complex numbers z. Consequently exp z is periodic with period 2πi.

Proof Note that the sequence ik , where k ⩾ 0, equals 1, i, −1, −i.1.i, −1, −i, . . . repeating with period
4. Hence
∞
X (iz)n
exp(iz) =
n!
n=0
∞ ∞
X (iz)2k X (iz)2k+1
= +
(2k)! (2k + 1)!
k=0 k=0
∞ ∞
X (−1)k z 2k X (−1)k z 2k+1
= +i
(2k)! (2k + 1)!
k=0 k=0
= cos z + i sin z.

• The identities
exp (iz) + exp (−iz) exp (iz) − exp (−iz)
cos z = , sin z =
2 2i
hold for general complex z.

16
 • When θ is real then cos θ = Re(exp iθ) and sinθ = Im(exp iθ). This is not true for general
complex z.

• It can be shown that exp : R → (0, ∞) defines a bijection with differentiable inverse log : (0, ∞) →
R, called the natural logarithm (and also denoted ln).

• Given positive a > 0 and real x we can define

ax = exp (x log a) .

Note, with this definition, ex = exp x for real x.

• For non-zero complex z the equation exp w = z has a solution w. In fact, as exp is periodic,
w = log z can take infinitely many values.

• So for complex numbers a, b, the definition

ab = exp(b log a)

might take infinitely many values, though only finitely many values if b is rational and just one
value if b is an integer.

7 Further topics – off syllabus

Below are discussed some further topics relating to complex numbers. Note that the topics are all
beyond the exam syllabus of this short course, but may be of interest, in particular because
they include some details on how to solve cubic equations – the problem which led to the initial use
of complex numbers.

7.1 Solving Cubic Equations

A cubic equation is one of the form Az 3 + Bz 2 + Cz + D = 0 where A ̸= 0. By dividing through by A
if necessary we can assume A = 1. Further we can make a substitution to simplify a cubic equation.

Proposition 37 The substitution z = Z − a/3 turns the equation z 3 + az 2 + bz + c = 0 into one of

the form Z 3 + mZ + n = 0.

Proof The proof is a matter of elementary algebra.

This leaves us in a position to introduce Cardano’s method for solving such cubics.

Algorithm 38 (Cardano’s Method for Solving Cubics) Consider the cubic equation

z 3 + mz + n = 0, (26)

where m and n are real numbers. Let D be such that D2 = m3 /27 + n2 /4. (As D2 is real then D is a
real or purely imaginary number.) We then define t and u by

t = −n/2 + D and u = n/2 + D,

and let T and U respectively be cube roots of t and u. It follows that tu is real, and if T and U are
chosen appropriately, so that T U = m/3, then z = T − U is a solution of the original cubic equation.

17
Remark 39 The discriminant of a cubic is ∆ = (α − β)2 (β − γ)2 (γ − α)2 where α, β, γ are the
roots of the cubic. It can be shown that ∆ = −4m3 − 27n2 for (26). So D2 = −∆/108. For m ⩽ 0, it
is the case that (26) has three, two or one real root when ∆ < 0, ∆ = 0 or ∆ > 0. ■

Proof Putting z = T − U into the LHS of (26) we get

LHS = (T − U )3 + m(T − U ) + n
= T 3 − 3U T 2 + 3U 2 T − U 3 + mT − mU + n
= t − u + (m − 3U T )(T − U ) + n
= (−n/2 + D) − (n/2 + D) + (m − 3U T )(T − U ) + n
= (m − 3T U )(T − U ).

Now
(T U )3 = tu = (−n/2 + D)(n/2 + D) = −n2 /4 + D2 = m3 /27 = (m/3)3 .
The cube roots of tu = (m/3)3 are m/3, ωm/3 and ω 2 m/3, where ω ̸= 1 is a cube root of unity. If T
is a cube root of t and U0 is a cube root of u, then T U0 is a cube root of tu. The other cubes roots
of u are ωU0 and ω 2 U0 ; if we choose U appropriately from U0 , ωU0 , ω 2 U0 we have T U = m/3. Thus,
with these T and U , we’ve shown z = T − U satisfies z 3 + mz + n = (m − 3T U )(T − U ) = 0. Note
that if T and U are chosen carefully (so that T U = m/3) then the other two roots are

ωT − ω 2 U , ω 2 T − ωU

as ωT, ω 2 T and ωU, ω 2 U are respectively cube roots of t and u and (ωT )(ω 2 U ) = (ω 2 T )(ωU ) = T U =
m/3.

18
 Solving The Cubic The story of how the cubic equation came to be
solved is a colourful one, but is also a window on a time when the habits
of the mathematical community were very different from today. The
three roots of the cubic equation z 3 + mz + n = 0 appear in (27) below,
where k equals 0, 1 or 2 and ω is a cube root of unity other than 1. The
formula is considerably more complicated than the quadratic formula. A
further complication, for the time, came with the unease 16th century
mathematicians had with negative numbers, let alone complex ones. So a
cubic equation of the form z 3 + mz = n would only have been considered
meaningful if m and n were positive and only positive roots would have
been of interest; an equation of the form z 3 = mz +n (with m and n again
positive) would have been considered an entirely different type of cubic
equation. The first person to solve equations of the form z 3 + mz = n
was Scipione Del Ferro (1456-1526) who was a mathematician at Bologna
University. Before he died, Del Ferro shared, in secret, his solution with
a student, Fiore.
Armed with this knowledge, which he thought himself to be the sole possessor of, Fiore rather fancied
that he could make a reputation for himself as a teacher of mathematics and challenged the Venetian
mathematician, Nicolo Tartaglia (1499-1557), to a competition to solve cubic equations which they would
set one another. Unfortunately for Fiore, Tartaglia was more than up to the task as he proved both able
to solve equations of the form z 3 +mz = n but could also pose and solve equations of the form z 3 = mz +n
which were beyond Fiore’s ken. Girolamo Cardano (1501-1576, pictured left), heard of Tartaglia’s victory
and – after some considerable effort on his part – eventually convinced Tartaglia to share his methods,
but only after Cardano had made an oath to also keep them secret which Cardano largely did. But the
situation became yet more confused when Cardano shared the solution of the cubic with his student and
secretary Lodovico Ferrari (1522-1565) who subsequently used it to solve quartic (degree four) equations:
both the solutions to the cubic and quartic were now held in secrecy by Cardano’s oath to Tartaglia.
The situation eventually resolved itself when Cardano heard that Fiore had gained his solution from Del
Ferro. Cardano then travelled to Bologna to see Del Ferro’s papers and was able to see that his original
method had been the same as Tartaglia’s. Feeling unburdened of the responsibilities of his oath, Cardano
published the solutions to the cubic and quartic in his Ars Magna (1545) citing priority with Del Ferro
and citing Tartaglia as an independent later solver of the cubic. Despite Cardano acknowledging the two
as the original solvers, the method of solution is usually referred to as Cardano’s method.
q q
3 2k 3
k
p p
zk = ω 2 3
−n/2 + n /4 + m /27 + ω −n/2 − n2 /4 + m3 /27 (27)

Example 40 Find the three roots of z 3 − 12z + 8 = 0.

√
Solution We have m = −12 and n = 8. As D2 = m3 /27 + n2 /4 = −48 then we can take D = 4 3i.
We set √ √
t = −n/2 + D = −4 + 4 3i; u = n/2 + D = 4 + 4 3i.
Now T is a cube root of t. As |t| = 8 and arg t = 2π/3, then t’s three cube roots are

T1 = 2 cis (2π/9); T2 = 2 cis (8π/9); T3 = 2 cis (14π/9).

Similarly |u| = 8 and arg u = π/3 giving

U1 = 2 cis (7π/9); U2 = 2 cis (π/9); U3 = 2 cis (13π/9)

19
respectively, where the Ui are chosen so that Ti Ui = m/3 = −4. Hence the three roots of z 3 −12z+8 = 0
are
T1 − U1 = 2(cis (2π/9) − cis (7π/9)) = 4 cos(2π/9);
T2 − U2 = 2(cis (8π/9) − cis (π/9)) = 4 cos(8π/9);
T3 − U3 = 2(cis (14π/9) − cis(13π/9)) = 4 cos(14π/9).
Numerically these roots are 3.06418, −3.75877, and 0.69459 to 5 decimal places. Note that the three
roots are all real even though none of t, u, Ti , Ui is real.

The Quintic Despite the successes in the 16th cen-

tury in solving cubics and quartics, more than two cen-
turies passed before progress was made with the quintic.
During that time, perhaps the most insightful view of
the mathematics to unfold came in 1770 when Joseph-
Louis Lagrange (1736-1813) rederived the solutions for
quadratic, cubic and quartic equations through an anal-
ysis of permutations of their roots; at that time though
it seemed that the mathematical community largely ex-
pected quintic equations to be likewise solvable and
Paolo Ruffini (1765-1822), Niels Henrik Abel (1802-
1829) and Evariste Galois (1811-1832) encountered var-
ious difficulties convincing contemporaries of the insolv-
ability of the general quintic. Insolvability, here, means
that there is no formula for the solutions of a general
quintic polynomial that involves only addition, subtrac-
tion, multiplication, division and taking roots, as is the
case for polynomials of degree 4 or less; that is to say
the general quintic is not solvable by radicals.
Ruffini gave an incomplete proof of this fact in 1799 but, despite writing to Lagrange several times,
his work failed to interest other mathematicians even to the extent of the gap in the proof being
pointed out. Ruffini’s work contains many important ideas about permutation groups and it may
have been the novelty of his work that led to it going unheeded. It was to be Abel, instead, who
provided the first complete proof in 1824. But it was Galois (pictured left), around 1830, who saw
deepest into the problem. Abel had shown that certain quintics are insolvable by radicals, though
some (e.g. x5 = 0) clearly are. Galois’ genius was to give a criterion showing precisely which
polynomials (of any degree) are solvable by radicals. He associated with a polynomial an algebraic
structure now known as its Galois group. If the roots of the polynomial could be extracted through
a succession of everyday algebra and taking roots, then its Galois group could correspondingly be
”built up” in a certain technical sense – the group would be what is now known as solvable. And
the Galois group of the general quintic is not solvable. Galois’ legacy in mathematics is enormous
and many important ideas relating to groups and fields – two fundamental structures in modern
algebra – can be traced back to his work; in fact finite fields are commonly called Galois fields after
him. However Galois, too, found it difficult to share his ideas during his lifetime despite three times
submitting his work to the Paris Academy. It was only thanks to Joseph Liouville (1809-1882) that
Galois’ work posthumously reached a wider audience, though this wasn’t until 1846.

Example 41 By making the substitution z = k cos θ, and recalling that cos 3θ = 4 cos3 θ − 3 cos θ,
solve the equation z 3 − 2z − 4 = 0.

20
 p
Solution
p We need that k 3 : 2k = 4 : 3 so that k = 8/3. The subsitution then gives cos 3θ =
3 3/2. Clearly this does not have real solutions θ, but using the identity cos(x + iy) = cos x cosh y −
i sin x sinh y, we see that p
cos x cosh y = 3 3/2, sin x sinh y = 0.
As y ̸= 0 then we have that
 x = nπ where n = 0, 2, 4 (with other n giving repetitions or contradictions)
−1
p
and y = cosh 3 3/2 . We then have the three roots
r   r
8 iy
8 y 
z1 = cos cosh = = 2;
3 33 3
r   r " √ #
8 2π iy 8 −1 y  3 y 
z2 = cos + = cosh −i sinh = −1 − i;
3 3 3 3 2 3 2 3
r   r " √ #
8 4π iy 8 −1 y  3 y 
z3 = cos + = cosh +i sinh = −1 + i.
3 3 3 3 2 3 2 3

7.2 The Extended Complex Plane

There is a useful general equation which encompasses both circles and lines. This fact is less surprising
with an appreciation of the extended complex plane, circlines and the Riemann sphere.

Proposition 42 (Circles and Lines in C) Let A and C be real and B complex, with A, B not both
zero. Then
Az z̄ + B z̄ + B̄z + C = 0, (28)
represents: (a) a line in direction iB when A = 0; q
(b) a circle, if A ̸= 0 and |B|2 ⩾ AC, with centre −B/A and radius |A|−1 |B|2 − AC;
and otherwise has no solutions. Moreover every circle and line can be represented in the form of (28).

Proof If A ̸= 0 then we can rearrange (28) as

z z̄ + (B/A)z̄ + (B̄/A)z + C/A = 0,

(z + B/A)(z + B/A) = B B̄/A2 − C/A, [as A is real and using (15)]
|z + B/A|2 = (|B|2 − AC)/A2 ,
[using (14)].
q
If |B|2 ⩾ AC then this is a circle with centre −B/A and radius |A|−1 |B|2 − AC and otherwise there
are no solutions to (28). Conversely, note that the equation of a general circle is |z − a| = r where a
is a complex number and r ⩾ 0. This can be rearranged using (14) as

z z̄ − az̄ − āz + (|a|2 − r2 ) = 0

which is in the form of (28) with A = 1, B = −a and C = |a|2 − r2 .

If A = 0 then we have the equation B z̄ + B̄z + C = 0. If we write B = u + iv and z = x + yi then

(u + iv)(x − yi) + (u − vi)(x + yi) + C = 0 which rearranges to 2ux + 2vy + C = 0, (29)

which is the equation of a line. Moreover we see that every line appears in this form by choosing u, v, C
appropriately. The line (29) is parallel to the vector (v, −u) or equivalently v − ui = i(u + iv) = iB.

21
 The fact that lines and circles can both be expressed in this way is not coincidental. Using the
form (28) above one can see that the map z 7→ 1/z maps the set of circles and lines to the set of circles
and lines. Specifically

• a line through the origin maps to a line through the origin;

• a circle through the origin maps to a line not through the origin (and vice versa);

• a circle not through the origin maps to a circle not through the origin.

Note that the the map z 7→ 1/z isn’t particularly well-defined on C. It isn’t defined at z = 0. But
note that points near 0 map to distant points, with great modulus and any argument. So it makes
sense to think of the image of 0 as a ‘infinity’ which is ‘out there’ in all directions (as opposed to ±∞
which extend the real line). The extended complex plane is then C∞ = C ∪ {∞} and the map
z 7→ 1/z is a bijection of the extended complex plane. Viewed this way, lines can be thought of as
‘circles’ that pass through infinity and the term ‘circline’ is used to describe both lines and circles.
The extended complex plane can be naturally identified with a sphere as follows. The sphere S,
with centre (0, 0, 0) and radius one, has equation x2 + y 2 + z 2 = 1.
Thinking of C as the xy-plane, every complex
number P = X + Y i can be identified with a
point Q on S by drawing a line from (X, Y, 0) z
in the xy-plane to the sphere’s north pole N =
(0, 0, 1); this line intersects the sphere at two N
points Q and N . We define a map f from C
to the sphere S by setting f (P ) = Q. (i) The
Q1
S
image point f (X + Y i) equals P2 P1 Re
X2 + Y 2 − 1
 
2X 2Y
, , .
1 + X2 + Y 2 1 + X2 + Y 2 1 + X2 + Y 2
Q2
(ii) The image of f equals all of S except N .

(iii) The inverse map π = f −1 is called stereographic projection. For (x, y, z) ̸= N , then

x + yi
π(x, y, z) = .
1−z
• Note that π maps points near N to complex numbers with large moduli. It is consequently
natural to identify a single ‘infinite’ point, written ∞, with the north pole N .

• If we identify C∞ via stereographic projection with the sphere S, then S is known as the Rie-
mann sphere.

• Under stereographic projection, lines in C correspond to circles on S which pass through N , and
circles in C correspond to circles on S which don’t pass through N .

• Viewed on the Riemann sphere, the map z 7→ 1/z simply turns the sphere upside down, rotating
it about the x-axis.

22
7.3 Complex Analysis

Complex analysis is a central topic in mathematics and widely

regarded as both one of the most aesthetic and harmonius but also
highly applicable. Most undergraduates meet analysis for the first
time at university in real analysis courses (pure calculus essentially)
containing results they might have otherwise considered obvious, or
”known” from school, treated in a seemingly pedantic manner and
with a focus on the pathological. By and large, a first course in
real analysis contains more you-can’t-do-that moments than most
undergraduates welcome. By comparison, complex analysis is the
can-do sibling of real analysis. The main subject of study is the set
of holomorphic functions, essentially consisting of those complex-
valued functions that have a derivative. It turns out that being
differentiable in this complex sense is a more demanding require-
ment; consequently it is possible to build up a richer theory about
these functions – for example, holomorphic functions have in fact
derivatives of all orders and are defined locally by a Taylor series.
But the set of holomorphic functions is still wide enough to include
all the important functions you are likely to be interested in. Com-
plex analysis, though, is more than simply another version of real
analysis in which the results are cleaner and more positive. The
very nature of the subject is different and has a much more topolog-
ical flavour. The first major result of complex analysis is Cauchy’s
Theorem, after Augustin-Louis Cauchy (1789-1857), pictured left.
In fact you will see Cauchy’s name populating much of a complex
analysis course as he almost single-handedly developed the subject.
This theorem states that the integral around any closed curve Γ
(essentially a loop) in the complex plane of a holomorphic function
f (z) is zero. If the integrand isn’t holomorphic everywhere – like
1/z, where only z = 0 is a problem – then the integral now only
depends on whether the loop encloses 0. If not, the answer is still
zero; if it wraps around the origin once anticlockwise the answer is
2πi; twice and the answer is 4πi. You might reasonably raise the
issue that a real integral often represents something, well, real – e.g. area or arc-length: what meaning
can a complex answer have? But by taking the real or imaginary parts of such an answer, real integrals
can be determined, often ones that are difficult to calculate by real methods alone. By such means a
first course in complex analysis would likely have you determining integrals and infinite sums such as

∞ 2π ∞ ∞
π4
Z Z Z
sin x π dt 2π dx π π  X 1
dx = ; =√ ; = csc ; = .
0 x 2 0 2 + sin t 3 0 1 + xn n n n4 90
1

It is usually via complex analysis that one first sees a proof of the Fundamental Theorem of Algebra
(Theorem 4). Complex analysis is also widely applicable. In pure mathematics it leads naturally into
the study of Riemann surfaces (for example, x2 +y 2 = 1, rather than defining a curve in the real plane
R2 , defines a two-dimensional surface in four-dimensional C2 ) and aspects of algebraic geometry; also
it is important in analytic number theory via Riemann’s Zeta Function. Complex methods are widely
used in the solution of partial differential equations with the links between Laplace’s equation and
holomorphic functions proving very useful in the study of ideal (inviscid) fluid flow in two dimensions.
23