## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

**Functions of One Complex Variable I, Second Edition
**

1

©Copyright by Andreas Kleefeld, 2009

All Rights Reserved

2

1

by John B. Conway

2

Last updated on January, 7th 2010

PREFACE

Most of the exercises I solved were assigned homeworks in the graduate courses Math 713 and Math 714

”Complex Analysis I and II” at the University of Wisconsin – Milwaukee taught by Professor Dashan Fan

in Fall 2008 and Spring 2009.

The solutions manual is intented for all students taking a graduate level Complex Analysis course. Students

can check their answers to homework problems assigned from the excellent book “Functions of One Com-

plex Variable I”, Second Edition by John B. Conway. Furthermore students can prepare for quizzes, tests,

exams and ﬁnal exams by solving additional exercises and check their results. Maybe students even study

for preliminary exams for their doctoral studies.

However, I have to warn you not to copy straight of this book and turn in your homework, because this

would violate the purpose of homeworks. Of course, that is up to you.

I strongly encourage you to send me solutions that are still missing to kleefeld@tu-cottbus.de (L

A

T

E

Xpreferred

but not mandatory) in order to complete this solutions manual. Think about the contribution you will give

to other students.

If you ﬁnd typing errors or mathematical mistakes pop an email to kleefeld@tu-cottbus.de. The recent

version of this solutions manual can be found at

http://www.math.tu-cottbus.de/INSTITUT/kleefeld/Files/Solution.html.

The goal of this project is to give solutions to all of the 452 exercises.

CONTRIBUTION

I thank (without special order)

Christopher T. Alvin

Martin J. Michael

for contributions to this book.

2

Contents

1 The Complex Number System 1

1.1 The real numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 The ﬁeld of complex numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.3 The complex plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.4 Polar representation and roots of complex numbers . . . . . . . . . . . . . . . . . . . . . 5

1.5 Lines and half planes in the complex plane . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.6 The extended plane and its spherical representation . . . . . . . . . . . . . . . . . . . . . 7

2 Metric Spaces and the Topology of C 9

2.1 Deﬁnitions and examples of metric spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.2 Connectedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.3 Sequences and completeness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.4 Compactness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.5 Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.6 Uniform convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3 Elementary Properties and Examples of Analytic Functions 21

3.1 Power series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.2 Analytic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3.3 Analytic functions as mappings. Möbius transformations . . . . . . . . . . . . . . . . . . 30

4 Complex Integration 40

4.1 Riemann-Stieltjes integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.2 Power series representation of analytic functions . . . . . . . . . . . . . . . . . . . . . . . 46

4.3 Zeros of an analytic function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

4.4 The index of a closed curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

4.5 Cauchy’s Theorem and Integral Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.6 The homotopic version of Cauchy’s Theorem and simple connectivity . . . . . . . . . . . 61

4.7 Counting zeros; the Open Mapping Theorem . . . . . . . . . . . . . . . . . . . . . . . . 64

4.8 Goursat’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

5 Singularities 66

5.1 Classiﬁcation of singularities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

5.2 Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

5.3 The Argument Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

3

6 The Maximum Modulus Theorem 82

6.1 The Maximum Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

6.2 Schwarz’s Lemma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

6.3 Convex functions and Hadamard’s Three Circles Theorem . . . . . . . . . . . . . . . . . 86

6.4 The Phragmén-Lindelöf Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

7 Compactness and Convergence in the Space of Analytic Functions 91

7.1 The space of continuous functions C(G, Ω) . . . . . . . . . . . . . . . . . . . . . . . . . 91

7.2 Spaces of analytic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

7.3 Spaces of meromorphic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

7.4 The Riemann Mapping Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

7.5 The Weierstrass Factorization Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

7.6 Factorization of the sine function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

7.7 The gamma function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

7.8 The Riemann zeta function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

8 Runge’s Theorem 121

8.1 Runge’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

8.2 Simple connectedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

8.3 Mittag-Leﬄer’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

9 Analytic Continuation and Riemann Surfaces 128

9.1 Schwarz Reﬂection Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

9.2 Analytic Continuation Along a Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

9.3 Monodromy Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

9.4 Topological Spaces and Neighborhood Systems . . . . . . . . . . . . . . . . . . . . . . . 130

9.5 The Sheaf of Germs of Analytic Functions on an Open Set . . . . . . . . . . . . . . . . . 131

9.6 Analytic Manifolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

9.7 Covering spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

10 Harmonic Functions 135

10.1 Basic properties of harmonic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

10.2 Harmonic functions on a disk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

10.3 Subharmonic and superharmonic functions . . . . . . . . . . . . . . . . . . . . . . . . . . 142

10.4 The Dirichlet Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

10.5 Green’s Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

11 Entire Functions 149

11.1 Jensen’s Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

11.2 The genus and order of an entire function . . . . . . . . . . . . . . . . . . . . . . . . . . 151

11.3 Hadamard Factorization Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

12 The Range of an Analytic Function 159

12.1 Bloch’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

12.2 The Little Picard Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

12.3 Schottky’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

12.4 The Great Picard Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

4

Chapter 1

The Complex Number System

1.1 The real numbers

No exercises are assigned in this section.

1.2 The ﬁeld of complex numbers

Exercise 1. Find the real and imaginary parts of the following:

1

z

;

z − a

z + a

(a ∈ R); z

3

;

3 + 5i

7i + 1

;

_

¸

¸

¸

¸

_

−1 + i

√

3

2

_

¸

¸

¸

¸

_

3

;

_

¸

¸

¸

¸

_

−1 − i

√

3

2

_

¸

¸

¸

¸

_

6

; i

n

;

_

1 + i

√

2

_

n

for 2 ≤ n ≤ 8.

Solution. Let z = x + iy. Then

a)

Re

_

1

z

_

=

x

x

2

+ y

2

=

Re(z)

¦z¦

2

Im

_

1

z

_

= −

y

x

2

+ y

2

= −

Im(z)

¦z¦

2

b)

Re

_

z − a

z + a

_

=

x

2

+ y

2

− a

2

x

2

+ y

2

+ 2ax + a

2

=

¦z¦

2

− a

2

¦z¦

2

+ 2aRe(z) + a

2

Im

_

z − a

z + a

_

=

2ya

x

2

+ y

2

+ 2xa + a

2

=

2Im(z)a

¦z¦

2

+ 2aRe(z) + a

2

c)

Re

_

z

3

_

= x

3

− 3xy

2

= Re(z)

3

− 3Re(z)Im(z)

2

Im

_

z

3

_

= 3x

2

y − y

3

= 3Re(z)

2

Im(z) − Im(z)

3

1

d)

Re

_

3 + 5i

7i + 1

_

=

19

25

Im

_

3 + 5i

7i + 1

_

= −

8

25

e)

Re

_

¸

¸

¸

¸

¸

¸

_

_

¸

¸

¸

¸

_

−1 + i

√

3

2

_

¸

¸

¸

¸

_

3

_

¸

¸

¸

¸

¸

¸

_

= 1

Im

_

¸

¸

¸

¸

¸

¸

_

_

¸

¸

¸

¸

_

−1 + i

√

3

2

_

¸

¸

¸

¸

_

3

_

¸

¸

¸

¸

¸

¸

_

= 0

f)

Re

_

¸

¸

¸

¸

¸

¸

_

_

¸

¸

¸

¸

_

−1 − i

√

3

2

_

¸

¸

¸

¸

_

6

_

¸

¸

¸

¸

¸

¸

_

= 1

Im

_

¸

¸

¸

¸

¸

¸

_

_

¸

¸

¸

¸

_

−1 − i

√

3

2

_

¸

¸

¸

¸

_

6

_

¸

¸

¸

¸

¸

¸

_

= 0

g)

Re (i

n

) =

_

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

_

0, n is odd

1, n ∈ ¦4k : k ∈ Z¦

−1, n ∈ ¦2 + 4k : k ∈ Z¦

Im(i

n

) =

_

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

_

0, n is even

1, n ∈ ¦1 + 4k : k ∈ Z¦

−1, n ∈ ¦3 + 4k : k ∈ Z¦

2

h)

Re

__

1 + i

√

2

_

n

_

=

_

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

_

0, n = 2

−

√

2

2

, n = 3

−1, n = 4

−

√

2

2

, n = 5

0, n = 6

√

2

2

, n = 7

1, n = 8

Im

__

1 + i

√

2

_

n

_

=

_

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

_

1, n = 2

√

2

2

, n = 3

0, n = 4

−

√

2

2

, n = 5

−1, n = 6

−

√

2

2

, n = 7

0, n = 8

Exercise 2. Find the absolute value and conjugate of each of the following:

−2 + i; −3; (2 + i)(4 + 3i);

3 − i

√

2 + 3i

;

i

i + 3

; (1 + i)

6

; i

17

.

Solution. It is easy to calculate:

a)

z = −2 + i, ¦z¦ =

√

5, ¯ z = −2 − i

b)

z = −3, ¦z¦ = 3, ¯ z = −3

c)

z = (2 + i)(4 + 3i) = 5 + 10i, ¦z¦ = 5

√

5, ¯ z = 5 − 10i

d

z =

3 − i

√

2 + 3i

, ¦z¦ =

1

11

√

110, ¯ z =

3 + i

√

2 − 3i

e)

z =

i

i + 3

=

1

10

+

3

10

i, ¦z¦ =

1

10

√

10, ¯ z =

1

10

−

3

10

i

f)

z = (1 + i)

6

= −8i, ¦z¦ = 8, ¯ z = 8i

g)

i

17

= i, ¦z¦ = 1, ¯ z = −i

Exercise 3. Show that z is a real number if and only if z = ¯ z.

3

Solution. Let z = x + iy.

⇒: If z is a real number, then z = x (y = 0). This implies ¯ z = x and therefore z = ¯ z.

⇔: If z = ¯ z, then we must have x +iy = x −iy for all x, y ∈ R. This implies y = −y which is true if y = 0 and

therefore z = x. This means that z is a real number.

Exercise 4. If z and w are complex numbers, prove the following equations:

¦z + w¦

2

= ¦z¦

2

+ 2Re(z ¯ w) + ¦w¦

2

.

¦z − w¦

2

= ¦z¦

2

− 2Re(z ¯ w) + ¦w¦

2

.

¦z + w¦

2

+ ¦z − w¦

2

= 2

_

¦z¦

2

+ ¦w¦

2

_

.

Solution. We can easily verify that

¯

¯ z = z. Thus

¦z + w¦

2

= (z + w)(z + w) = (z + w)(¯ z + ¯ w) = z¯ z + z ¯ w + w¯ z + w¯ w

= ¦z¦

2

+ ¦w¦

2

+ z ¯ w + ¯ zw = ¦z¦

2

+ ¦w¦

2

+ z ¯ w + ¯ z

¯

¯ w

= ¦z¦

2

+ ¦w¦

2

+ z ¯ w + z ¯ w = ¦z¦

2

+ ¦w¦

2

+ 2

z ¯ w + z ¯ w

2

= ¦z¦

2

+ ¦w¦

2

+ 2Re(z ¯ w) = ¦z¦

2

+ 2Re(z ¯ w) + ¦w¦

2

.

¦z − w¦

2

= (z − w)(z − w) = (z − w)(¯ z − ¯ w) = z¯ z − z ¯ w − w¯ z + w¯ w

= ¦z¦

2

+ ¦w¦

2

− z ¯ w − ¯ zw = ¦z¦

2

+ ¦w¦

2

− z ¯ w − ¯ z

¯

¯ w

= ¦z¦

2

+ ¦w¦

2

− z ¯ w − z ¯ w = ¦z¦

2

+ ¦w¦

2

− 2

z ¯ w + z ¯ w

2

= ¦z¦

2

+ ¦w¦

2

− 2Re(z ¯ w) = ¦z¦

2

− 2Re(z ¯ w) + ¦w¦

2

.

¦z + w¦

2

+ ¦z − w¦

2

= ¦z¦

2

+ Re(z ¯ w) + ¦w¦

2

+ ¦z¦

2

− Re(z ¯ w) + ¦w¦

2

= ¦z¦

2

+ ¦w¦

2

+ ¦z¦

2

+ ¦w¦

2

= 2¦z¦

2

+ 2¦w¦

2

= 2

_

¦z¦

2

+ ¦w¦

2

_

.

Exercise 5. Use induction to prove that for z = z

1

+ . . . + z

n

; w = w

1

w

2

. . . w

n

:

¦w¦ = ¦w

1

¦ . . . ¦w

n

¦; ¯ z = ¯ z

1

+ . . . + ¯ z

n

; ¯ w = ¯ w

1

. . . ¯ w

n

.

Solution. Not available.

Exercise 6. Let R(z) be a rational function of z. Show that R(z) = R(¯ z) if all the coeﬃcients in R(z) are real.

Solution. Let R(z) be a rational function of z, that is

R(z) =

a

n

z

n

+ a

n−1

z

n−1

+ . . . a

0

b

m

z

m

+ b

m−1

z

m−1

+ . . . b

0

where n, m are nonnegative integers. Let all coeﬃcients of R(z) be real, that is

a

0

, a

1

, . . . , a

n

, b

0

, b

1

, . . . , b

m

∈ R.

Then

R(z) =

a

n

z

n

+ a

n−1

z

n−1

+ . . . a

0

b

m

z

m

+ b

m−1

z

m−1

+ . . . b

0

=

a

n

z

n

+ a

n−1

z

n−1

+ . . . a

0

b

m

z

m

+ b

m−1

z

m−1

+ . . . b

0

=

a

n

z

n

+ a

n−1

z

n−1

+ . . . a

0

b

m

z

m

+ b

m−1

z

m−1

+ . . . b

0

=

a

n

¯ z

n

+ a

n−1

¯ z

n−1

+ . . . a

0

b

m

¯ z

m

+ b

m−1

¯ z

m−1

+ . . . b

0

= R(¯ z).

4

1.3 The complex plane

Exercise 1. Prove (3.4) and give necessary and suﬃcient conditions for equality.

Solution. Let z and w be complex numbers. Then

¦¦z¦ − ¦w¦¦ = ¦¦z − w + w¦ − ¦w¦¦

≤ ¦¦z − w¦ + ¦w¦ − ¦w¦¦

= ¦¦z − w¦¦

= ¦z − w¦

Notice that ¦z¦ and ¦w¦ is the distance from z and w, respectively, to the origin while ¦z − w¦ is the distance

between z and w. Considering the construction of the implied triangle, in order to guarantee equality, it is

necessary and suﬃcient that

¦¦z¦ − ¦w¦¦ = ¦z − w¦

⇐⇒ (¦z¦ − ¦w¦)

2

= ¦z − w¦

2

⇐⇒ (¦z¦ − ¦w¦)

2

= ¦z¦

2

− 2Re(z ¯ w) + ¦w¦

2

⇐⇒ ¦z¦

2

− 2¦z¦¦w¦ + ¦w¦

2

= ¦z¦

2

− 2Re(z ¯ w) + ¦w¦

2

⇐⇒ ¦z¦¦w¦ = Re(z ¯ w)

⇐⇒ ¦z ¯ w¦ = Re(z ¯ w)

Equivalently, this is z ¯ w ≥ 0. Multiplying this by

w

w

, we get z ¯ w

w

w

= ¦w¦

2

z

w

≥ 0 if w 0. If

t =

z

w

=

_

1

¦w¦

2

_

¦w¦

2

z

w

. Then t ≥ 0 and z = tw.

Exercise 2. Show that equality occurs in (3.3) if and only if z

k

/z

l

≥ 0 for any integers k and l, 1 ≤ k, l ≤ n,

for which z

l

0.

Solution. Not available.

Exercise 3. Let a ∈ R and c > 0 be ﬁxed. Describe the set of points z satisfying

¦z − a¦ − ¦z + a¦ = 2c

for every possible choice of a and c. Now let a be any complex number and, using a rotation of the plane,

describe the locus of points satisfying the above equation.

Solution. Not available.

1.4 Polar representation and roots of complex numbers

Exercise 1. Find the sixth roots of unity.

Solution. Start with z

6

= 1 and z = rcis(θ), therefore r

6

cis(6θ) = 1. Hence r = 1 and θ =

2kπ

6

with k ∈

¦−3, −2, −1, 0, 1, 2¦. The following table gives a list of principle values of arguments and the corresponding

value of the root of the equation z

6

= 1.

θ

0

= 0 z

0

= 1

θ

1

=

π

3

z

1

= cis(

π

3

)

θ

2

=

2π

3

z

2

= cis(

2π

3

)

θ

3

= π z

3

= cis(π) = −1

θ

4

=

−2π

3

z

4

= cis(

−2π

3

)

θ

5

=

−π

3

z

5

= cis(

−π

3

)

5

Exercise 2. Calculate the following:

a) the square roots of i

b) the cube roots of i

c) the square roots of

√

3 + 3i

Solution. c) The square roots of

√

3 + 3i.

Let z =

√

3 + 3i. Then r = ¦z¦ =

_

_ √

3

_

2

+ 3

2

=

√

12 and α = tan

−1

_

3

√

3

_

=

π

3

. So, the 2 distinct roots of z

are given by

2

√

r

_

cos

α+2kπ

n

+ i sin

α+2kπ

n

_

where k = 0, 1. Speciﬁcally,

√

z =

4

√

12

_

cos

π

3

+ 2kπ

2

+ i sin

π

3

+ 2kπ

2

_

.

Therefore, the square roots of z, z

k

, are given by

z

0

=

4

√

12

_

cos

π

6

+ i sin

π

6

_

=

4

√

12

_

√

3

2

+

1

2

i

_

z

1

=

4

√

12

_

cos

7π

6

+ i sin

7π

6

_

=

4

√

12

_

−

√

3

2

−

1

2

i

_

.

So, in rectangular form, the second roots of z are given by

_

4

√

108

2

,

4

√

12

2

_

and

_

−

4

√

108

2

, −

4

√

12

2

_

.

Exercise 3. A primitive nth root of unity is a complex number a such that 1, a, a

2

, ..., a

n−1

are distinct nth

roots of unity. Show that if a and b are primitive nth and mth roots of unity, respectively, then ab is a kth root

of unity for some integer k. What is the smallest value of k? What can be said if a and b are nonprimitive

roots of unity?

Solution. Not available.

Exercise 4. Use the binomial equation

(a + b)

n

=

n

k=0

_

n

k

_

a

n−k

b

k

,

where

_

n

k

_

=

n!

k!(n − k)!

,

and compare the real and imaginary parts of each side of de Moivre’s formula to obtain the formulas:

cos nθ = cos

n

θ −

_

n

2

_

cos

n−2

θ sin

2

θ +

_

n

4

_

cos

n−4

θ sin

4

θ − . . .

sin nθ =

_

n

1

_

cos

n−1

θ sin θ −

_

n

3

_

cos

n−3

θ sin

3

θ + . . .

Solution. Not available.

Exercise 5. Let z = cis

2π

n

for an integer n ≥ 2. Show that 1 + z + . . . + z

n−1

= 0.

6

Solution. The summation of the ﬁnite geometric sequence 1, z, z

2

, . . . , z

n−1

can be calculated as

_

n

j=1

z

j−1

=

z

n

−1

z−1

. We want to show that z

n

is an n

th

root of unity. So, using de Moivre’s formula, z

n

=

_

cis

_

2π

n

__

n

=

cis

_

n

2π

n

_

= cis(2π) = 1. It follows that 1 + z + z

2

+ ... + z

n−1

=

z

n

−1

z−1

=

1−1

z−1

= 0 as required.

Exercise 6. Show that ϕ(t) = cis t is a group homomorphism of the additive group R onto the multiplicative

group T = ¦z : ¦z¦ = 1¦.

Solution. Not available.

Exercise 7. If z ∈ C and Re(z

n

) ≤ 0 for every positive integer n, show that z is a non-negative real number.

Solution. Let n be an arbitrary but ﬁxed positive integer and let z ∈ C and Re(z

n

) ≥ 0. Since z

n

=

r

n

(cos(nθ) + i sin(nθ)), we have

Re(z

n

) = r

n

cos(nθ) ≥ 0.

If z = 0, then we are done, since r = 0 and Re(z

n

) = 0. Therefore, assume z 0, then r > 0. Thus

Re(z

n

) = r

n

cos(nθ) ≥ 0

implies cos(nθ) ≥ 0 for all n. This implies θ = 0 as we will show next. Clearly, θ [π/2, 3π/2]. If

θ ∈ (0, π/2), then there exists a k ∈ ¦2, 3, . . .¦ such that

π

k+1

≤ θ <

π

k

. If we choose n = k + 1, we have

π ≤ nθ <

(k + 1)π

k

which is impossible since cos(nθ) ≥ 0. Similarly, we can derive a contradiction if we assume θ ∈ (3π/2, 2π).

Then 2π − π/k ≤ θ < 2π − π/(k + 1) for some k ∈ ¦2, 3, . . .¦.

1.5 Lines and half planes in the complex plane

Exercise 1. Let C be the circle ¦z : ¦z − c¦ = r¦, r > 0; let a = c + rcis α and put

L

β

=

_

z : Im

_

z − a

b

_

= 0

_

where b = cisβ. Find necessary and suﬃcient conditions in terms of β that L

β

be tangent to C at a.

Solution. Not available.

1.6 The extended plane and its spherical representation

Exercise 1. Give the details in the derivation of (6.7) and (6.8).

Solution. Not available.

Exercise 2. For each of the following points in C, give the corresponding point of S : 0, 1 + i, 3 + 2i.

Solution. We have

Φ(z) =

_

2x

¦z¦

2

+ 1

,

2y

¦z¦

2

+ 1

,

¦z¦

2

− 1

¦z¦

2

+ 1

_

.

If z

1

= 0, then ¦z

1

¦ = 0 and therefore

Φ(z

1

) = (0, 0, −1).

7

Thus, z

1

= 0 corresponds to the point (0, 0, −1) on the sphere S .

If z

2

= 1 + i, then ¦z

2

¦ =

√

2 and therefore

Φ(z

2

) =

_

2

3

,

2

3

,

1

3

_

.

Thus, z

2

= 1 + i corresponds to the point

_

2

3

,

2

3

,

1

3

_

on the sphere S .

If z

3

= 3 + 2i, then ¦z

3

¦ =

√

13 and therefore

Φ(z

3

) =

_

3

7

,

2

7

,

6

7

_

.

Thus, z

3

= 3 + 2i corresponds to the point

_

3

7

,

2

7

,

6

7

_

on the sphere S .

Exercise 3. Which subsets of S correspond to the real and imaginary axes in C.

Solution. If z is on the real axes, then z = x which implies ¦z¦

2

= x

2

. Thus

Φ(z) =

_

2x

x

2

+ 1

, 0,

x

2

− 1

x

2

+ 1

_

.

Therefore the set

__

2x

x

2

+ 1

, 0,

x

2

− 1

x

2

+ 1

_

¦x ∈ R

_

⊂ S

corresponds to the real axes in C. That means the unit circle x

2

+ z

2

= 1 lying in the xz−plane corresponds

to the real axes in C.

If z is on the imaginary axes, then z = iy which implies ¦z¦

2

= y

2

. Thus

Φ(z) =

_

0,

2y

y

2

+ 1

,

y

2

− 1

y

2

+ 1

_

.

Therefore the set

__

0,

2y

y

2

+ 1

,

y

2

− 1

y

2

+ 1

_

¦y ∈ R

_

⊂ S

corresponds to the imaginary axes in C. That means the unit circle y

2

+ z

2

= 1 lying in the yz−plane

corresponds to the imaginary axes in C.

Exercise 4. Let Λ be a circle lying in S . Then there is a unique plane P in R

3

such that P ∩ S = Λ. Recall

from analytic geometry that

P = ¦(x

1

, x

2

, x

3

) : x

1

β

1

+ x

2

β

2

+ x

3

β

3

= l¦

where (β

1

, β

2

, β

3

) is a vector orthogonal to P and l is some real number. It can be assumed that β

2

1

+β

2

2

+β

2

3

=

1. Use this information to show that if Λ contains the point N then its projection on C is a straight line.

Otherwise, Λ projects onto a circle in C.

Solution. Not available.

Exercise 5. Let Z and Z

·

be points on S corresponding to z and z

·

respectively. Let W be the point on S

corresponding to z + z

·

. Find the coordinates of W in terms of the coordinates of Z and Z

·

.

Solution. Not available.

8

Chapter 2

Metric Spaces and the Topology of C

2.1 Deﬁnitions and examples of metric spaces

Exercise 1. Show that each of the examples of metric spaces given in (1.2)-(1.6) is, indeed, a metric space.

Example (1.6) is the only one likely to give any diﬃculty. Also, describe B(x; r) for each of these examples.

Solution. Not available.

Exercise 2. Which of the following subsets of C are open and which are closed: (a) ¦z : ¦z¦ < 1¦; (b) the

real axis; (c) ¦z : z

n

= 1 for some integer n ≥ 1¦; (d) ¦z ∈ C is real and 0 ≤ z < 1¦; (e) ¦z ∈ C : z is real and

0 ≤ z ≤ 1¦?

Solution. We have

(a) A := ¦z ∈ C : ¦z¦ < 1¦

Let z ∈ A and set ε

z

=

1−¦z¦

2

, then B(z, ε

z

) ⊂ A is open and therefore A =

_

z∈A

B(z, ε

z

) is open also. A

cannot be closed, otherwise A and C − A were both closed and open sets yet C is connected.

(b) B := ¦z ∈ C : z = x + iy, y = 0¦ (the real axis)

Let z ∈ C − B, then Imz 0. Set ε

z

=

¦Imz¦

2

, then B(z, ε

z

) ⊂ C − B.Hence B is closed since its

complement is open.

For any real x and any ε > 0 the point x + i

ε

2

∈ B(x, ε) but x + i

ε

2

∈ C − R. Thus B is not open.

(c) C := ¦z ∈ C : z

n

= 1 for some integer n ≥ 1¦

Claim: C is neither closed nor open.

C is not open because if z

n

= 1 then z = rcis(θ) with r = 1 and any ε-ball around z would contain an

element z

·

:= (1 +

ε

4

)cis(θ).

To show that C cannot be closed, note that for

p

q

∈ Q with p ∈ Z and q ∈ N the number z :=

cis

_

p

q

2π

_

∈ C since

z

q

= cis(p2π) = cos(p2π) + i sin(p2π) = 1.

Now ﬁx x ∈ R − Q and let ¦x

n

¦

n

be a rational sequence that converges to x. Let m be any natural

number. Now z

m

= 1 implies that sin(mx2π) = 0 which in turn means that mx ∈ Z contradicting the

choice x ∈ R − Q. We have constructed a sequence of elements of C that converges to a point that is

not element of C. Hence C is not closed.

9

(d) D := ¦z ∈ C : z is real and 0 ≤ z < 1¦

The set cannot be open by the argument given in 2.1) and it is not closed because z

n

:= 1 −

1

n

is a

sequence in D that converges to a point outside of D.

(e) E := ¦z ∈ C : z is real and 0 ≤ z ≤ 1¦

This set E is not open by the observation in 2.1) and it is closed because its complement is open: If

z E and z real, then B(z,

min¦¦x¦,¦x−1¦¦

2

) is contained in the complement of E, if z is imaginary then

B(z,

¦Imy¦

2

) is completely contained in the complement of E.

Exercise 3. If (X, d) is any metric space show that every open ball is, in fact, an open set. Also, show that

every closed ball is a closed set.

Solution. Not available.

Exercise 4. Give the details of the proof of (1.9c).

Solution. Not available.

Exercise 5. Prove Proposition 1.11.

Solution. Not available.

Exercise 6. Prove that a set G ⊂ X is open if and only if X − G is closed.

Solution. Not available.

Exercise 7. Show that (C

∞

, d) where d is given (I. 6.7) and (I. 6.8) is a metric space.

Solution. To show (C

∞

, d) with d(z, z

·

) =

2¦z−z

·

¦

[(1+¦z¦

2

)(1+¦z

·

¦

2

)]

1/2

for z, z

·

∈ C and d(z, ∞) =

2

(1+¦z¦

2

)

1/2

, z ∈ C is a

metric space.

i) d(z, z

·

) ≥ 0.

Since ¦z−z

·

¦ ≥ 0, we have d(z, z

·

) =

2¦z−z

·

¦

[(1+¦z¦

2

)(1+¦z

·

¦

2

)]

1/2

≥ 0 for all z, z

·

∈ C (the denominator is always positive).

Obviously, d(z, ∞) =

2

(1+¦z¦

2

)

1/2

≥ 0 for all z ∈ C.

ii) d(z, z

·

) = 0 iﬀ z = z

·

.

We have 2¦z − z

·

¦ = 0 iﬀ z = z

·

. Therefore, d(z, z

·

) =

2¦z−z

·

¦

[(1+¦z¦

2

)(1+¦z

·

¦

2

)]

1/2

= 0 iﬀ z = z

·

. d(∞, ∞) =

lim

z→∞

2

(1+¦z¦

2

)

1/2

= 0. Thus, d(∞, 0) = 0 iﬀ z = ∞.

iii) d(z, z

·

) = d(z

·

, z).

We have d(z, z

·

) =

2¦z−z

·

¦

[(1+¦z¦

2

)(1+¦z

·

¦

2

)]

1/2

=

2¦z

·

−z¦

[(1+¦z

·

¦

2

)(1+¦z¦

2

)]

1/2

= d(z

·

, z) for all z, z

·

∈ C. Also d(z, ∞) =

2

(1+¦z¦

2

)

1/2

=

d(∞, z) by the symmetry of d(z, z

·

) in general.

iv) d(z, z

·

) ≤ d(z, x) + d(x, z

·

).

We have

d(z, z

·

) =

2¦z − z

·

¦

[(1 + ¦z¦

2

)(1 + ¦z

·

¦

2

)]

1/2

≤

2¦z − x¦

[(1 + ¦z¦

2

)(1 + ¦x¦

2

)]

1/2

+

2¦x − z

·

¦

[(1 + ¦x¦

2

)(1 + ¦z

·

¦

2

)]

1/2

= d(z, x) + d(x, z

·

)

for all x, z, z

·

∈ C. The ﬁrst inequality will be shown next.

First, we need the following equation

(z − z

·

)(1 + ¯ xx) = (z − x)(1 + ¯ xz

·

) + (x − z

·

)(1 + ¯ xz), (2.1)

10

which follows from a simple computation

(z − x)(1 + ¯ xz

·

) + (x − z

·

)(1 + ¯ xz) = z − x + ¯ xz

·

z + x ¯ xz

·

+ x − z

·

+ x ¯ xz − ¯ xz

·

z

= z + x ¯ xz − z

·

− x ¯ xz

·

= (z − z

·

)(1 + ¯ xx).

Using (2.1) and taking norms gives

¦(z − z

·

)(1 + ¯ xx)¦ = ¦(z − z

·

)(1 + ¦x¦

2

)¦ = ¦(z − z

·

)¦(1 + ¦x¦

2

)

= ¦(z − x)(1 + ¯ xz

·

) + (x − z

·

)(1 + ¯ xz)¦

≤ ¦z − x¦(1 + ¦x¦

2

)

1/2

(1 + ¦z

·

¦

2

)

1/2

+ ¦x − z

·

¦(1 + ¦x¦

2

)

1/2

(1 + ¦z¦

2

)

1/2

(2.2)

where the last inequality follows by

¦z + ¯ xz

·

¦ ≤ (1 + ¦x¦

2

)

1/2

(1 + ¦z

·

¦

2

)

1/2

⇔ (1 + ¯ xz

·

)(1 + x¯ z

·

) ≤ (1 + ¦x¦

2

)(1 + ¦z

·

¦

2

)

and

¦z + ¯ xz¦ ≤ (1 + ¦x¦

2

)

1/2

(1 + ¦z¦

2

)

1/2

⇔ (1 + ¯ xz)(1 + x¯ z) ≤ (1 + ¦x¦

2

)(1 + ¦z¦

2

)

since

(1 + ¯ xz)(1 + x¯ z) ≤ (1 + ¦x¦

2

)(1 + ¦z¦

2

)

⇔ ¯ xz + x¯ z ≤ ¦x¦

2

+ ¦y¦

2

⇔ 2Re(x¯ z) ≤ ¦x¦

2

+ ¦y¦

2

which is true by Exercise 4 part 2 on page 3. Thus, dividing (2.2) by (1 + ¦x¦

2

)

1/2

yields

¦z − z

·

¦(1 + ¦x¦

2

)

1/2

≤ ¦z − x¦(1 + ¦z

·

¦

2

)

1/2

+ ¦x − z

·

¦(1 + ¦z¦

2

)

1/2

.

Multiplying this by

2

(1 + ¦x¦

2

)

1/2

(1 + ¦z

·

¦

2

)

1/2

(1 + ¦z¦

2

)

1/2

gives the assertion above.

Exercise 8. Let (X, d) be a metric space and Y ⊂ X. Suppose G ⊂ X is open; show that G ∩ Y is open in

(Y, d). Conversely, show that if G

1

⊂ Y is open in (Y, d), there is an open set G ⊂ X such that G

1

= G ∩ Y.

Solution. Set G

1

= G ∩ Y, let G be open in (X, d) and Y ⊂ X. In order to show that G

1

is open in (Y, d),

pick an arbitrary point p ∈ G

1

. Then p ∈ G and since G is open, there exists an > 0 such that

B

X

(p; ) ⊂ G.

But then

B

Y

(p; ) = B

X

(p; ) ∩ Y ⊂ G ∩ Y = G

1

which proves that p is an interior point of G

1

in the metric d. Thus G

1

is open in Y (Proposition 1.13a).

Let G

1

be an open set in Y. Then, for every p ∈ G

1

, there exists an −ball

B

Y

(p : ) ⊂ G

1

.

11

Thus

G

1

=

_

p∈G

1

B

Y

(p; ).

Since we can write

B

Y

(p; ) = B

X

(p; ) ∩ Y,

we get

G

1

=

_

p∈G

1

B

Y

(p; ) =

_

p∈G

1

_

B

X

(p; ) ∩ Y

_

=

_

p∈G

1

B

X

(p; ) ∩ Y = G ∩ Y

where G =

_

p∈G

1

B

X

(p; ) is open in (X, d). (Proposition 1.9c since each B

X

(p; ) is open).

Exercise 9. Do Exercise 8 with “closed” in place of “open.”

Solution. Not available.

Exercise 10. Prove Proposition 1.13.

Solution. Not available.

Exercise 11. Show that ¦cis k : k a non-negative integer ¦ is dense in T = ¦z ∈ C : ¦z¦ = 1¦. For which

values of θ is ¦cis(kθ) : k a non-negative integer ¦ dense in T?

Solution. Not available.

2.2 Connectedness

Exercise 1. The purpose of this exercise is to show that a connected subset of R is an interval.

(a) Show that a set A ⊂ R is an interval iﬀ for any two points a and b in A with a < b, the interval

[a, b] ⊂ A.

(b) Use part (a) to show that if a set A ⊂ R is connected then it is an interval.

Solution. Not available.

Exercise 2. Show that the sets S and T in the proof of Theorem 2.3 are open.

Solution. Not available.

Exercise 3. Which of the following subsets X of C are connected; if X is not connected, what are its

components: (a) X = ¦z : ¦z¦ ≤ 1¦ ∪ ¦z : ¦z − 2¦ < 1¦. (b) X = [0, 1) ∩

_

1 +

1

n

: n ≥ 1

_

. (c) X = C − (A ∩ B)

where A = [0, ∞) and B = ¦z = r cis θ : r = θ, 0 ≤ θ ≤ ∞¦?

Solution. a) Deﬁne X = ¦z : ¦z¦ ≤ 1¦ ∪¦z : ¦z −2¦ < 1¦ := A∪B. It suﬃces to show that X is path-connected.

Obviously A is path-connected and B is path-connected. Next, we will show that X is path-connected.

Recall that a space is path-connected if for any two points x and y there exists a continuous function f from

the interval [0, 1] to X with f (0) = x and f (1) = y (this function f is called the path from x to y).

Let x ∈ A and y ∈ B and deﬁne the function f (t) : [0, 1] → X by

f (t) =

_

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

_

(1 − 3t)x + 3tRe(x), 0 ≤ t ≤

1

3

(2 − 3t)Re(x) + (3t − 1)Re(y),

1

3

< t ≤

2

3

(3 − 3t)Re(y) + (3t − 2)y,

2

3

< t ≤ 1.

12

This function is obviously continuous, since f (1/3) = lim

t→

1

3

− f (t) = lim

t→

1

3

+ f (t) = Re(x) ∈ X and

f (2/3) = lim

t→

2

3

− f (t) = lim

t→

2

3

+ f (t) = Re(y) ∈ X. In addition, we have f (0) = x and f (1) = y. Therefore

X is path-connected and hence X is connected.

b) There is no way to connect ¦2¦ and ¦

3

2

¦. Therefore X is not connected. The components are [0, 1), ¦2¦,

¦

3

2

¦, ¦

4

3

¦, . . ., ¦1 +

1

n

¦.

c) X is not connected, since there is no way to connect (2, 1) and (1, −2). The k − th component is given by

C − ¦A ∩ B¦ where

A = [2πk − 2π, 2πk)

and

B = ¦z = r cis θ : r = θ, 2πk − 2π ≤ θ < 2πk¦, k ∈ ¦1, 2, 3, . . .¦.

Exercise 4. Prove the following generalization of Lemma 2.6. If ¦D

j

: j ∈ J¦ is a collection of connected

subsets of X and if for each j and k in J we have D

j

∩ D

k

then D =

_

¦D

j

: j ∈ J¦ is connected.

Solution. Let D =

_

j∈J

D

j

and ( = ¦D

j

: j ∈ J¦. If D is connected, we could write D as the disjoint union

A ∪ B where A and B are nonempty subsets of X. Thus, for each C ∈ ( either C ⊂ A or C ⊂ B.

We have C ⊂ A ∀C ∈ ( or C ⊂ B ∀C ∈ (. If not, then there exist E, F ∈ ( such that E ⊂ A and F ⊂ B.

But, we assume A ∪ B is disjoint and thus E ∪ F is disjoint which contradicts the assumption E ∩ F ∅.

Therefore, all members of ( are contained in either A or all B. Thus, either D = A and B = ∅ or D = B and

A = ∅. Both contradicting the fact that A, B are assumed to be nonempty. Hence,

D =

_

j∈J

D

j

is connected.

Exercise 5. Show that if F ⊂ X is closed and connected then for every pair of points a, b in F and each

> 0 there are points z

0

, z

1

, . . . , z

n

in F with z

0

= a, z

n

= b and d(z

k−1

, z

k

) < for 1 ≤ k ≤ n. Is the

hypothesis that F be closed needed? If F is a set which satisﬁes this property then F is not necessarily

connected, even if F is closed. Give an example to illustrate this.

Solution. Not available.

2.3 Sequences and completeness

Exercise 1. Prove Proposition 3.4.

Solution. a) A set is closed iﬀ it contains all its limit points.

Let S ⊂ X be a set.

“⇐”: Assume S contains all its limit points. We have to show that S is closed or that S

c

is open. Let

x ∈ S

c

. By assumption x is not a limit point and hence there exists an open −ball around x, B(x; ) such

that B(x; ) ∩ S = ∅ (negation of Proposition 1.13f). So B(x; ) ⊂ S

c

and therefore S

c

is open.

“⇒”: Let S be closed and x be a limit point. We claim x ∈ S . If not, S

c

(open) would be an open

neighborhood of x, that does not intersect S . (Proposition 1.13f again) which contradicts the fact that x is

a limit point of S .

b) If A ⊂ X, then A

−

= A ∪ ¦x : x is a limit point of A¦ := A ∪ A

·

.

Let A ⊂ X be a set.

“⊆”: Let x ∈

¯

A. We want to show x ∈ A ∪ A

·

. If x ∈ A, then obviously x ∈ A ∪ A

·

. Suppose x A. Since

x ∈

¯

A, we have (Proposition 1.13f) that for every > 0, B(x; ) ∩ A ∅. Because x A, B(x; ) must

13

intersect A in a point diﬀerent from x. In particular, for every integer n there is a point x

n

in B(x;

1

n

) ∩ A.

Thus d(x, x

n

) <

1

n

which implies x

n

→ x (see Proof of Proposition 3.2). Then x ∈ A

·

, so x ∈ A ∪ A

·

.

“⊇”: To show A ∪ A

·

⊆ A

−

. Let x ∈ A ∪ A

·

. If x ∈ A, then x ∈ A

−

(A ⊂ A

−

). Now, assume x ∈ A

·

but not

in A. Then there exists ¦x

n

¦ ⊂ A with lim

n→∞

x

n

= x. It follows, ∀ > 0 B(x; ) ∩ A ∅ since ¦x

n

¦ ⊂ A. By

Proposition 1.13f we get x ∈ A

−

.

Exercise 2. Furnish the details of the proof of Proposition 3.8.

Solution. Not available.

Exercise 3. Show that diam A = diam A

−

.

Solution. Not available.

Exercise 4. Let z

n

, z be points in C and let d be the metric on C

∞

. Show that ¦z

n

− z¦ → 0 if and only if

d(z

n

, z) → 0. Also show that if ¦z

n

¦ → ∞then ¦z

n

¦ is Cauchy in C

∞

. (Must ¦z

n

¦ converge in C

∞

?)

Solution. First assume that ¦z

n

− z¦ → 0, then

d(z

n

, z) =

2

_

(1 + ¦z

n

¦

2

)(1 + ¦z¦

2

)

¦z

n

− z¦ → 0

since the denominator

_

(1 + ¦z

n

¦

2

)(1 + ¦z¦

2

) ≥ 1 is bounded below away from 0. To see the converse, let

d(z

n

, z) → 0 or equivalently d

2

(z

n

, z) → 0. We need to show that if z

n

→ z in the d-norm, then ¦z

n

¦ ∞

because otherwise the denominator grows without bounds. In fact we will show that if ¦z

n

¦ → ∞, then

d(z

n

, z) 0 for any z ∈ C. Then

d

2

(z

n

, z) =

4

_

¦z

n

¦

2

− z

n

¯ z − ¯ z

n

z + ¦z¦

2

_

_

1 + ¦z

n

¦

2

_ _

1 + ¦z¦

2

_

=

4

_

¦z

n

¦

2

− 2Re(z

n

¯ z) + ¦z¦

2

_

_

1 + ¦z¦

2

_

¦z

n

¦

2

+ 1 + ¦z¦

2

=

4

_

1 −

2Re(z

n

¯ z)

¦z

n

¦

2

+

¦z¦

2

¦z

n

¦

2

_

1 + ¦z¦

2

+

1+¦z¦

2

¦z

n

¦

2

if ¦z

n

¦ 0

in particular if ¦z

n

¦ → ∞, d(z

n

, z) →

4

1+¦z¦

2

0. This shows that the denominator remains bounded as

d(z

n

, z) → 0 and therefore the numerator 2¦z

n

− z¦ → 0. Hence convergence in d-norm implies convergence

in ¦¦-norm for numbers z

n

, z ∈ C. Next assume that ¦z

n

¦ → ∞. Then clearly also

_

1 + ¦z

n

¦

2

→ ∞ and

therefore d(z

n

, ∞) =

2

√

1+¦z

n

¦

2

→ 0. The last thing to show is that if z

n

→ ∞ in (C

∞

, d), also ¦z

n

¦ → ∞. But

d(z

n

, ∞) → 0 implies

_

1 + ¦z

n

¦

2

→ ∞which is equivalent to ¦z

n

¦ → ∞.

Exercise 5. Show that every convergent sequence in (X, d) is a Cauchy sequence.

Solution. Let ¦x

n

¦ be a convergent sequence with limit x. That is, given > 0 ∃N such that d(x

n

, x) <

2

if

n > N and d(x, x

m

) <

2

if m > N. Thus

d(x

n

, x

m

) ≤ d(x

n

, x) + d(x, x

m

) <

2

+

2

= , ∀n, m ≥ N

and therefore ¦x

n

¦ is a Cauchy sequence.

14

Exercise 6. Give three examples of non complete metric spaces.

Solution. Example 1: Let X = C[−1, 1] and the metric d( f , g) =

_

_

1

−1

[ f (x) − g(x)]

2

dx, f , g ∈ X. Consider

the Cauchy sequence

f

n

(x) =

_

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

_

0, −1 ≤ x ≤ 0

nx, 0 < x ≤

1

n

1,

1

n

< x ≤ 1

.

It is obvious that the limit function f is discontinuous. Hence, the metric space (X, d) is not complete.

Example 2: Let X = (0, 1] with metric d(x, y) = ¦x − y¦, x, y ∈ X. The sequence ¦

1

n

¦ is Cauchy, but converges

to 0, which is not in the space. Thus, the metric space is not complete.

Example 3: Let X = Q, the rationals, with metric d(x, y) = ¦x −y¦, x, y ∈ X. The sequence deﬁned by x

1

= 1,

x

n+1

=

x

n

2

+

1

x

n

is a Cauchy sequence of rational numbers. The limit

√

2 is not a rational number. Therefore,

the metric space is not complete.

Exercise 7. Put a metric d on R such that ¦x

n

− x¦ → 0 if and only if d(x

n

, x) → 0, but that ¦x

n

¦ is a Cauchy

sequence in (R, d) when ¦x

n

¦ → ∞. (Hint: Take inspiration from C

∞

.)

Solution. Not available.

Exercise 8. Suppose ¦x

n

¦ is a Cauchy sequence and ¦x

n

k

¦ is a subsequence that is convergent. Show that

¦x

n

¦ must be convergent.

Solution. Since ¦x

n

k

¦ is convergent, there is a x such that x

n

k

→ x as k → ∞. We have to show that x

n

→ x

as n → x. Let > 0. Then we have ∃N ∈ N such that d(x

n

, x

m

) <

2

∀n, m ≥ N since ¦x

n

¦ is Cauchy and

∃M ∈ N such that d(x

n

k

, x) <

2

∀n

k

≥ M since x

n

k

→ x as k → ∞. Now, ﬁx n

k

0

> M + N, then

d(x

n

, x) ≤ d(x

n

, x

n

k

0

) + d(x

n

k

0

, x) <

2

+

2

= , ∀n ≥ N.

Thus, x

n

→ x as n → ∞and therefore ¦x

n

¦ is convergent.

2.4 Compactness

Exercise 1. Finish the proof of Proposition 4.4.

Solution. Not available.

Exercise 2. Let p = (p

1

, . . . , p

n

) and q = (q

1

, . . . , q

n

) be points in R

n

with p

k

< q

k

for each k. Let

R = [p

1

, q

1

] × . . . × [p

n

, q

n

] and show that

diam R = d(p, q) =

_

¸

¸

¸

¸

¸

_

n

k=1

(q

k

− p

k

)

2

_

¸

¸

¸

¸

¸

_

1

2

.

Solution. By deﬁnition

diam(R) = sup

x∈R,y∈R

d(x, y).

Obviously, R is compact, so we have

diam(R) = max

x∈R,y∈R

d(x, y).

15

Let x = (x

1

, . . . , x

n

) ∈ R and y = (y

1

, . . . , y

n

) ∈ R. Then clearly, p

i

≤ x

i

≤ q

i

and p

i

≤ y

i

≤ q

i

for all

i = 1, . . . , n. We also have

(y

i

− x

i

)

2

≤ (q

i

− p

i

)

2

, ∀i = 1, . . . , n. (2.3)

Therefore, by (2.3) we obtain

d(x, y) =

¸

_

n

i=1

(y

i

− x

i

)

2

≤

¸

_

n

i=1

(q

i

− p

i

)

2

= d(p, q).

Note that we get equality if for example x

i

= p

i

and y

i

= q

i

for all i = 1, . . . , n. Thus the maximum distance

is obtained, so

diam(R) = sup

x∈R,y∈R

d(x, y) =

¸

_

n

i=1

(q

i

− p

i

)

2

= d(p, q).

Exercise 3. Let F = [a

1

, b

1

] × . . . × [a

n

, b

n

] ⊂ R

n

and let > 0; use Exercise 2 to show that there are

rectangles R

1

, . . . , R

m

such that F =

_

m

k=1

R

k

and diam R

k

< for each k. If x

k

∈ R

k

then it follows that

R

k

⊂ B(x

k

; ).

Solution. Not available.

Exercise 4. Show that the union of a ﬁnite number of compact sets is compact.

Solution. Let K =

_

n

i=1

K

i

be a ﬁnite union of compact sets. Let ¦G

λ

¦

λ∈Γ

be an open cover of K, that is

K ⊂

_

λ∈Γ

G

λ

.

Of course, ¦G

λ

¦

λ∈Γ

is an open cover of each K

i

, i = 1, . . . , n, that is

K

i

⊂

_

λ∈Γ

G

λ

, ∀i = 1, . . . , n.

Since each K

i

is compact,

K

i

⊂

k

i

_

j=1

G

i, j

for each i = 1, . . . , n. Thus

K =

n

_

j=1

K

i

⊂

n

_

i=1

k

i

_

j=1

G

i, j

which is a ﬁnite union and therefore K is compact.

Exercise 5. Let X be the set of all bounded sequences of complex numbers. That is, ¦x

n

¦ ∈ X iﬀ sup¦¦x

n

¦ :

n ≥ 1¦ < ∞. If x = ¦x

n

¦ and y = ¦y

n

¦, deﬁne d(x, y) = sup¦¦x

n

− y

n

¦ : n ≥ 1¦. Show that for each x in X and

> 0,

¯

B(x; ) is not totally bounded although it is complete. (Hint: you might have an easier time of it if

you ﬁrst show that you can assume x = (0, 0, . . .).)

Solution. Not available.

Exercise 6. Show that the closure of a totally bounded set is totally bounded.

16

Solution. Let (X, d) be a given metric space and let S ⊂ X be totally bounded. Let > 0. Since S is totally

bounded, we have by Theorem 4.9 d) p. 22 that there exist a ﬁnite number of points x

1

, . . . , x

n

∈ S such that

S ⊆

n

_

k=1

B(x

k

; /2).

Taking the closure, gives the desired result

¯

S ⊆

n

_

k=1

B(x

k

; ).

2.5 Continuity

Exercise 1. Prove Proposition 5.2.

Solution. Not available.

Exercise 2. Show that if f and g are uniformly continuous (Lipschitz) functions from X into C then so is

f + g.

Solution. The distance in C is ρ(x, y) = ¦x − y¦. The distance in X is d(x, y). Let f : X → C be Lipschitz,

that is, there is a constant M > 0 such that

ρ( f (x), f (y)) = ¦ f (x) − f (y)¦ ≤ Md(x, y), ∀x, y ∈ X

and g : X → C be Lipschitz, that is, there is a constant N > 0 such that

ρ(g(x), g(y)) = ¦g(x) − g(y)¦ ≤ Nd(x, y), ∀x, y ∈ X.

We have

ρ( f (x) + g(x), f (y) + g(y)) = ¦ f (x) + g(x) − f (y) − g(y)¦ = ¦ f (x) − f (y) + g(x) − g(y)¦

≤ ¦ f (x) − f (y)¦ + ¦g(x) − g(y)¦

≤ Md(x, y) + Nd(x, y) = (M + N)d(x, y), ∀x, y ∈ X.

Since there is a constant K = M + N > 0 such that

ρ( f (x) + g(x), f (y) + g(y)) ≤ Kd(x, y), ∀x, y ∈ X,

we have that f + g is Lipschitz.

Now, let f , g be both uniformly continuous, that is,

∀

1

> 0 ∃δ

1

> 0 such that ρ( f (x), f (y)) = ¦ f (x) − f (y)¦ <

1

whenever d(x, y) < δ

1

and

∀

2

> 0 ∃δ

2

> 0 such that ρ( f (x), f (y)) = ¦ f (x) − f (y)¦ <

2

whenever d(x, y) < δ

2

.

We have

ρ( f (x) + g(x), f (y) + g(y)) = ¦ f (x) + g(x) − f (y) − g(y)¦ = ¦ f (x) − f (y) + g(x) − g(y)¦

≤ ¦ f (x) − f (y)¦ + ¦g(x) − g(y)¦

≤

1

+

2

,

17

whenever d(x, y) < δ

1

and d(x, y) < δ

2

, that is, whenever d(x, y) < min(δ

1

, δ

2

). So, choosing =

1

+

2

and

δ = min(δ

1

, δ

2

), we have shown that

∀ > 0 ∃δ > 0 such that ρ( f (x), f (y)) = ¦ f (x) − f (y)¦ < whenever d(x, y) < δ.

Thus f + g is uniformly continuous.

Exercise 3. We say that f : X → C is bounded if there is a constant M > 0 with ¦ f (x)¦ ≤ M for all x in X.

Show that if f and g are bounded uniformly continuous (Lipschitz) functions from X into C then so is f g.

Solution. Let f be bounded, that is, there exists a constant M

1

> 0 with ¦ f (x)¦ < M

1

for all x ∈ X and let g

be bounded, that is, there exists a constant M

2

> 0 with ¦g(x)¦ < M

2

for all x ∈ X. Obviously f g is bounded,

because

¦ f (x)g(x)¦ ≤ ¦ f (x)¦ ¦g(x)¦ ≤ M

1

M

2

∀x ∈ X.

So, there exists a constant M = M

1

M

2

with

¦ f (x)g(x)¦ ≤ M ∀x ∈ X.

Let f be Lipschitz, that is, there exists a constant N

1

> 0 such that

ρ( f (x), f (y)) ≤ N

1

d(x, y) ∀x, y ∈ X

and let g be Lipschitz, that is, there exists a constant N

2

> 0 such that

ρ(g(x), g(y)) ≤ N

1

d(x, y) ∀x, y ∈ X.

Now,

ρ( f (x)g(x), f (y)g(y)) = ¦ f (x)g(x) − f (y)g(y)¦ =

= ¦ f (x)g(x) − f (x)g(y) + f (x)g(y) − f (y)g(y)¦

≤ ¦ f (x)g(x) − f (x)g(y)¦ + ¦ f (x)g(y) − f (y)g(y)¦

≤ ¦ f (x)¦ ¦g(x) − g(y)¦ + ¦ f (x)¦ ¦g(y) − g(y)¦

≤ M

1

N

2

d(x, y) + M

2

N

1

d(x, y) = (M

1

N

2

+ M

2

N

1

)d(x, y) ∀x, y ∈ X.

Thus, there exists a constant K = M

1

N

2

+ M

2

N

1

> 0 with

ρ( f (x)g(x), f (y)g(y)) ≤ Kd(x, y) ∀x, y ∈ X,

so f g is Lipschitz and bounded.

Now, let f and g be both bounded and uniformly continuous, that is

∃ M

1

such that ¦ f (x)¦ ≤ M

1

∀x ∈ X

∃ M

2

such that ¦g(x)¦ ≤ M

2

∀x ∈ X

∀

1

> 0 ∃δ

1

> 0 such that ρ( f (x), f (y)) = ¦ f (x) − f (y)¦ <

1

whenever d(x, y) < δ

1

∀

2

> 0 ∃δ

2

> 0 such that ρ( f (x), f (y)) = ¦ f (x) − f (y)¦ <

2

whenever d(x, y) < δ

2

.

18

It remains to verify that f g is uniformly continuous, since we have already shown that f g is bounded. We

have

ρ( f (x)g(x), f (y)g(y)) = ¦ f (x)g(x) − f (y)g(y)¦ =

= ¦ f (x)g(x) − f (x)g(y) + f (x)g(y) − f (y)g(y)¦

≤ ¦ f (x)g(x) − f (x)g(y)¦ + ¦ f (x)g(y) − f (y)g(y)¦

≤ ¦ f (x)¦ ¦g(x) − g(y)¦ + ¦ f (x)¦ ¦g(y) − g(y)¦

≤ M

1

2

+ M

2

1

,

whenever d(x, y) < min(δ

1

, δ

2

). So choosing = M

1

2

+ M

2

1

and δ = min(δ

1

, δ

2

), we have ∀ > 0 ∃δ > 0

such that

¦ f (x)g(x) − f (y)g(y)¦ <

whenever d(x, y) < δ. Thus, f g is uniformly continuous and bounded.

Exercise 4. Is the composition of two uniformly continuous (Lipschitz) functions again uniformly continu-

ous (Lipschitz)?

Solution. Not available.

Exercise 5. Suppose f : X →Ω is uniformly continuous; show that if ¦x

n

¦ is a Cauchy sequence in X then

¦ f (x

n

)¦ is a Cauchy sequence in Ω. Is this still true if we only assume that f is continuous? (Prove or give

a counterexample.)

Solution. Assume f : X → Ω is uniformly continuous, that is, for every > 0 there exists δ > 0 such that

ρ( f (x), f (y)) < if d(x, y) < δ. If ¦x

n

¦ is a Cauchy sequence in X, we have, for every

1

> 0 there exists

N ∈ N such that d(x

n

, x

m

) <

1

for all n, m ≥ N. But then, by the uniform continuity, we have that

ρ( f (x

n

), f (x

m

)) < ∀n, m ≥ N

whenever d(x

n

, x

m

) < δ which tells us that ¦ f (x

n

)¦ is a Cauchy sequence in Ω.

If f is continuous, the statement is not true. Here is a counterexample: Let f (x) =

1

x

which is continuous

on (0, 1). The sequence x

n

=

1

n

is apparently convergent and therefore a Cauchy sequence in X. But

¦ f (x

n

)¦ = ¦ f (

1

n

)¦ = ¦n¦ is obviously not Cauchy. Note that f (x) =

1

x

is not uniformly continuous on (0, 1).

To see that pick = 1. Then there is no δ > 0 such that ¦ f (x) − f (y)¦ < 1 whenever ¦x −y¦ < δ. Assume there

exists such a δ. WLOG assume δ < 1 since the interval (0, 1) is considered. Let y = x +δ/2 and set x = δ/2,

then

¦ f (x) − f (y)¦ =

¸

¸

¸

¸

¸

1

x

−

1

y

¸

¸

¸

¸

¸

=

y − x

xy

=

δ/2

δ/2 δ

=

1

δ

> 1,

that is no matter what δ < 1 we choose, we always obtain ¦ f (x) − f (y)¦ > 1. Therefore f (x) =

1

x

cannot be

uniformly continuous.

Exercise 6. Recall the deﬁnition of a dense set (1.14). Suppose that Ω is a complete metric space and that

f : (D, d) → (Ω; ρ) is uniformly continuous, where D is dense in (X, d). Use Exercise 5 to show that there

is a uniformly continuous function g : X →Ω with g(x) = f (x) for every x in D.

Solution. Not available.

Exercise 7. Let G be an open subset of C and let P be a polygon in G from a to b. Use Theorems 5.15

and 5.17 to show that there is a polygon Q ⊂ G from a to b which is composed of line segments which are

parallel to either the real or imaginary axes.

19

Solution. Not available.

Exercise 8. Use Lebesgue’s Covering Lemma (4.8) to give another proof of Theorem 5.15.

Solution. Suppose f : X → Ω is continuous and X is compact. To show f is uniformly continuous. Let

> 0. Since f is continuous, we have for all x ∈ X there is a δ

x

> 0 such that ρ( f (x), f (y)) < /2 whenever

d(x, y) < δ

x

. In addition,

X =

_

x∈X

B(x; δ

x

)

is an open cover of X. Since X is by assumption compact (it is also sequentially compact as stated in

Theorem 4.9 p. 22), we can use Lebesgue’s Covering Lemma 4.8 p. 21 to obtain a δ > 0 such that x ∈ X

implies that B(x, δ) ⊂ B(z; δ

z

) for some z ∈ X. More precisely, x, y ∈ B(z; δ

z

) and therefore

ρ( f (x), f (z)) ≤ ρ( f (x), f (z)) + ρ( f (z), f (y)) <

2

+

2

=

and hence f is uniformly continuous on X.

Exercise 9. Prove the following converse to Exercise 2.5. Suppose (X, d) is a compact metric space having

the property that for every > 0 and for any points a, b in X, there are points z

0

, z

1

, . . . , z

n

in X with z

0

= a,

z

n

= b, and d(z

k−l

, z

k

) < for 1 ≤ k ≤ n. Then (X, d) is connected. (Hint: Use Theorem 5.17.)

Solution. Not available.

Exercise 10. Let f and g be continuous functions from (X, d) to (Ω, p) and let D be a dense subset of X.

Prove that if f (x) = g(x) for x in D then f = g. Use this to show that the function g obtained in Exercise 6

is unique.

Solution. Not available.

2.6 Uniform convergence

Exercise 1. Let ¦ f

n

¦ be a sequence of uniformly continuous functions from (X, d) into (Ω, ρ) and suppose

that f = u − −lim f

n

exists. Prove that f is uniformly continuous. If each f

n

is a Lipschitz function with

constant M

n

and sup M

n

< ∞, show that f is a Lipschitz function. If sup M

n

= ∞, show that f may fail to

be Lipschitz.

Solution. Not available.

20

Chapter 3

Elementary Properties and Examples of

Analytic Functions

3.1 Power series

Exercise 1. Prove Proposition 1.5.

Solution. Not available.

Exercise 2. Give the details of the proof of Proposition 1.6.

Solution. Not available.

Exercise 3. Prove that limsup(a

n

+b

n

) ≤ limsup a

n

+limsup b

n

and liminf(a

n

+b

n

) ≥ liminf a

n

+liminf b

n

for bounded sequences of real numbers ¦a

n

¦ and ¦b

n

¦.

Solution. Let r > limsup

n→∞

a

n

(we know there are only ﬁnitely many by deﬁnition) and let s > limsup

n→∞

(same here, there are only ﬁnitely many by deﬁnition). Then r + s > a

n

+ b

n

for all but ﬁnitely many n’s.

This however, implies that

r + s ≥ limsup

n→∞

(a

n

+ b

n

).

Since this holds for any r > limsup

n→∞

a

n

and s > limsup

n→∞

b

n

, we have

limsup

n→∞

(a

n

+ b

n

) ≤ limsup

n→∞

a

n

+ limsup

n→∞

b

n

.

Let r < liminf

n→∞

a

n

(we know there are only ﬁnitely many by deﬁnition) and let s < liminf

n→∞

(same

here, there are only ﬁnitely many by deﬁnition). Then r + s < a

n

+ b

n

for all but ﬁnitely many n’s. This

however, implies that

r + s ≤ liminf

n→∞

(a

n

+ b

n

).

Since this holds for any r < liminf

n→∞

a

n

and s < liminf

n→∞

b

n

, we have

liminf

n→∞

(a

n

+ b

n

) ≥ liminf

n→∞

a

n

+ liminf

n→∞

b

n

.

Exercise 4. Show that liminf a

n

≤ limsup a

n

for any sequence in R.

21

Solution. Let m = liminf

n→∞

a

n

and b

n

= inf¦a

n

, a

n+1

, . . .¦. Let M = limsup

n→∞

a

n

. Take any s > M.

Then, by deﬁnition of the limsup

n→∞

a

n

= M, we obtain that a

n

< s for inﬁnitely many n’s which implies

that b

n

< s for all n and hence limsup

n→∞

b

n

= m < s. This holds for all s > M. But the inﬁmum of all

these s’s is M. Therefore m ≤ M which is

liminf

n→∞

a

n

≤ limsup

n→∞

a

n

.

Exercise 5. If ¦a

n

¦ is a convergent sequence in R and a = lima

n

, show that a = liminf a

n

= limsup a

n

.

Solution. Suppose that ¦a

n

¦ is a convergent sequence in R with limit a = lim

n→∞

a

n

. Then by deﬁnition, we

have: ∀ > 0 ∃N > 0 such that ∀n ≥ N, we have ¦a

n

−a¦ ≤ , that is a − ≤ a

n

≤ a +. This means that all

but ﬁnitely many a

n

’s are ≤ a + and ≥ a − . This shows that

a − ≤ liminf

n→∞

a

n

.,,.

=:m

≤ a +

and

a − ≤ limsup

n→∞

a

n

.,,.

=:M

≤ a + .

By the previous Exercise 4, we also have

a − ≤ m ≤ M ≤ a + .

Hence,

0 ≤ M − m ≤ 2.

Since > 0 is arbitrary, we obtain m = M and further

a − ≤ liminf

n→∞

a

n

≤ limsup

n→∞

a

n

≤ a + ,

we obtain

liminf

n→∞

a

n

= limsup

n→∞

a

n

= a.

Exercise 6. Find the radius of convergence for each of the following power series: (a)

_

∞

n=0

a

n

z

n

, a ∈ C;

(b)

_

∞

n=0

a

n

2

z

n

, a ∈ C; (c)

_

∞

n=0

k

n

z

n

, k an integer 0; (d)

_

∞

n=0

z

n!

.

Solution. a) We have

_

∞

n=0

a

n

z

n

=

_

∞

k=0

b

k

z

n

with b

k

= a

k

, a ∈ C. We also have,

limsup

k→∞

¦b

k

¦

1/k

= limsup

k→∞

¦a

k

¦

1/k

= limsup

k→∞

¦a¦ = ¦a¦.

Therefore, R = 1/¦a¦, so

R =

_

¸

¸

_

¸

¸

_

1

¦a¦

, a 0

∞, a = 0

.

b) In this case, b

n

= a

n

2

where a ∈ C.

R = lim

n→∞

¸

¸

¸

¸

¸

b

n

b

n+1

¸

¸

¸

¸

¸

= lim

n→∞

¸

¸

¸

¸

¸

¸

a

n

2

a

(n+1)

2

¸

¸

¸

¸

¸

¸

= lim

n→∞

¸

¸

¸

¸

¸

¸

a

n

2

a

n

2

+2n+1

¸

¸

¸

¸

¸

¸

= lim

n→∞

¸

¸

¸

¸

¸

1

a

2n+1

¸

¸

¸

¸

¸

= lim

n→∞

1

¦a¦

2n+1

=

_

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

_

0, ¦a¦ > 1

1, ¦a¦ = 1

∞, ¦a¦ < 1

.

22

c) Now, b

n

= k

n

, k is an integer 0. We have

R = limsup

n→∞

¦b

n

¦

1/n

= limsup

n→∞

¦k

n

¦

1/n

= limsup

n→∞

¦k¦ = ¦k¦.

So

R =

1

¦k¦

=

_

¸

¸

_

¸

¸

_

1

k

, k > 0, k integer

−

1

k

, k < 0, k integer

.

d) We can write

_

∞

n=0

z

n!

=

_

∞

k=0

a

k

z

k

where

a

k

=

_

¸

¸

¸

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

¸

¸

¸

_

0, k = 0

2, k = 1

1, k = n!, n ∈ N, n > 1

0, otherwise

Thus,

limsup

k→∞

¦a

k

¦

1/k

= limsup

k→∞

¦1¦

1/k!

= 1.

Therefore 1/R = 1 which implies R = 1.

Exercise 7. Show that the radius of convergence of the power series

∞

n=1

(−1)

n

n

z

n(n+1)

is 1, and discuss convergence for z = 1, −1, and i. (Hint: The nth coeﬃcient of this series is not (−1)

n

/n.)

Solution. Rewrite the power series in standard form, then

∞

n=1

(−1)

n

n

z

n(n+1)

=

∞

n=1

a

k

z

k

with a

k

=

_

¸

¸

_

¸

¸

_

(−1)

n

n

if ∃ n ∈ N s.t. k = n(n + 1)

0 else

.

To ﬁnd the radius of convergence we use the root criterion and therefore need the estimates

1 ≤

n(n+1)

√

n ≤

n

√

n for n ∈ N.

The ﬁrst inequality is immediate from the fact that n ≥ 1 and hence n

1

n(n+1)

≥ 1. For the second inequality

note that

n ≤ n

n+1

⇔ n

1

n(n+1)

≤ n

1

n

⇔

n(n+1)

√

n ≤

n

√

n .

Using this one obtains

n(n+1)

_

¦

(−1)

n

n

¦ =

1

n(n+1)

√

n

≤ 1

and

n(n+1)

_

¦

(−1)

n

n

¦ =

1

n(n+1)

√

n

≥

1

n

√

n

.

23

Vague memories of calculus classes tell me that

n

√

n → 1, thus

1

R

= limsup

n(n+1)

√

a

n

= 1, i.e. R = 1.

If z = 1 the series reduces to

_

∞

n=1

(−1)

n

n

which converges with the Leibniz Criterion.

If z = −1 we note that the exponents n(n + 1) are always even integers and therefore the series is the

same as in the previous case of z = 1.

Now let z = i. The expression i

n(n+1)

will always be real, so if the series converges at z = i, it converges

to a real number. We also note that formally

∞

n=1

(−1)

n

n

i

n(n+1)

=

∞

n=1

c

n

with

_

¸

¸

_

¸

¸

_

1

n

ifn mod 4 ∈ ¦0, 1¦

−

1

n

if n mod 4 ∈ ¦2, 3¦

.

Deﬁne the partial sums S

k

:=

_

k

n=0

c

k

. We claim that the following chain of inequalities holds

0 ≤ S

4k+3

a)

< S

4k

b)

< S

4k+4

c)

< S

4k+2

d)

< S

4k+1

≤ 1.

To verify this, note that

S

4k+3

− S

4k

= c

4k+1

+ c

4k+2

+ c

4k+3

= −

16k

2

+ 8k − 1

(4k + 1)(4k + 2)(4k + 3)

< 0, hence a)

S

4k+4

− S

4k

= c

4k+3

+ c

4k+4

= −

1

4k + 3

+

1

4k + 4

< 0, hence c)

S

4k+2

− S

4k+1

= c

4k+2

< 0, hence d).

Relation b) is obvious and so are the upper bound c

1

= 1 and the non-negativity constraint. We remark that

¦S

4k+l

¦

k≥1

, l ∈ ¦0, 1, 2, 3¦ describe bounded and monotone subsequences that converge to some point. Now

that ¦c

n

¦ ¸ 0 the diﬀerence between S

4k+l

and S

4k+m

, l, m ∈ 0, 1, 2, 3 tends to zero, i.e. all subsequences

converge to the same limit. Therefore the power series converges also in the case of z = i.

3.2 Analytic functions

Exercise 1. Show that f (z) = ¦z¦

2

= x

2

+ y

2

has a derivative only at the origin.

Solution. The derivative of f at z is given by

f

·

(z) = lim

h→0

f (z + h) − f (z)

h

, h ∈ C

provided the limit exist. We have

f (z + h) − f (z)

h

=

¦z + h¦

2

− ¦z¦

2

h

=

(z + h)(¯ z +

¯

h) − z¯ z

h

=

z¯ z + h¯ z + z

¯

h + h

¯

h − z¯ z

h

= ¯ z +

¯

h + z

¯

h

h

=: D.

If the limit of D exists, it may be found by letting the point h = (x, y) approach the origin (0,0) in the complex

plane C in any manner.

1.) Take the path along the real axes, that is y = 0. Then

¯

h = h and thus

D = ¯ z + h + z

h

h

= ¯ z + h + z

24

and therefore, if the limit of D exists, its value has to be ¯ z + z.

2.) Take the path along the imaginary axes, that is x = 0. Then

¯

h = −h and thus

D = ¯ z − h − z

h

h

= ¯ z − h − z

and therefore, if the limit of D exists, its value has to be ¯ z − z.

Because of the uniqueness of the limit of D, we must have

¯ z + z = ¯ z − z ⇐⇒ z = −z ⇐⇒ z = 0,

if the limit of D exists. It remains to show that the limit of D exists at z = 0. Since z = 0, we have that D =

¯

h

and thus the limit of D is 0.

In summary, the function f (z) = ¦z¦

2

= x

2

+ y

2

has a derivative only at the origin with value 0.

Exercise 2. Prove that if b

n

, a

n

are real and positive and 0 < b = limb

n

< ∞, a = limsup a

n

then

ab = limsup(a

n

b

n

). Docs this remain true if the requirement of positivity is dropped?

Solution. Let a = limsup

n→∞

a

n

< ∞. Then there exists a monotonic subsequence ¦a

n

k

¦ of ¦a

n

¦ that converges

to a. Since lim

k→∞

b

n

k

= b, lim

n→∞

a

n

k

b

n

k

= ab. Hence, ¦a

n

k

b

n

k

¦ is a subsequence of a

n

b

n

that converges to ab. So

ab ≤ limsup

n→∞

a

n

b

n

. Hence, limsup

n→∞

a

n

b

n

≥ b limsup

n→∞

a

n

.

Now, let a = limsup

n→∞

a

n

= ∞. Then there exists a subsequence ¦a

n

k

¦ of ¦a

n

¦ such that lim

k→∞

a

n

k

= a > 0.

And, since lim

n→∞

b

n

> 0, lim

k→∞

a

n

k

b

n

k

= ∞. Hence limsup

n→∞

a

n

b

n

= ∞. In this second case, limsup

n→∞

a

n

b

n

≥

b limsup

n→∞

a

n

.

In both cases, we have established that ab ≤ limsup

n→∞

a

n

b

n

. Now, since for all n ∈ N, a

n

> 0 and b

n

> 0,

consider lim

n→∞

1

b

n

=

1

b

. Applying the inequality we have established replacing b

n

with

1

b

n

and a

n

replaced

with a

n

b

n

:

limsup

n→∞

a

n

= limsup

n→∞

1

b

n

(a

n

b

n

) ≥

1

b

limsup

n→∞

a

n

b

n

.

Rearranging,

limsup

n→∞

a

n

b

n

≤ b limsup

n→∞

a

n

.

It follows that ab = limsup

n→∞

a

n

b

n

as required.

Now, consider the case where we drop the positivity requirement. Let b

n

= 0, −

1

2

, 0, −

1

3

, 0, −

1

4

, . . . and

note 0 = b = lim

n→∞

b

n

< ∞. Also let a

n

= 0, −2, 0, −3, 0, −4 . . . and note limsup

n→∞

a

n

= a = 0. In this case,

ab = 0 1 = limsup

n→∞

a

n

b

n

.

Exercise 3. Show that limn

1/n

= 1.

25

Solution. Let n ∈ N. Also let a = n

1/n

. Then

a = n

1/n

⇐⇒ log a = log n

1/n

⇐⇒ log a =

log n

n

⇐⇒ lim

n→∞

log a = lim

n→∞

log n

n

Now, let f (x) =

log(x)

x

. Then lim

n→∞

f (x) = 0 by L’Hopital’s Rule. Thus, lim

n→∞

log n

n

= lim

n→∞

f (n) = 0 also. So

lim

n→∞

a = lim

n→∞

n

1/n

= 1.

Exercise 4. Show that (cos z)

·

= −sin z and (sin z)

·

= cos z.

Solution. We have by deﬁnition

(cos z)

·

=

_

1

2

_

e

iz

+ e

−iz

_

_

·

=

i

2

_

e

iz

− e

−iz

_

= −

1

2i

_

e

iz

− e

−iz

_

= −sin z

and similarly

(sin z)

·

=

_

1

2i

_

e

iz

− e

−iz

_

_

·

=

i

2i

_

e

iz

+ e

−iz

_

=

1

2

_

e

iz

+ e

−iz

_

= cos z.

Exercise 5. Derive formulas (2.14).

Solution. Not available.

Exercise 6. Describe the following sets: ¦z : e

z

= i¦, ¦z : e

z

= −1¦, ¦z : e

z

= −i¦, ¦z : cos z = 0¦,

¦z : sin z = 0¦.

Solution. Using the deﬁnition we obtain

¦z : e

z

= i¦ =

__

1

2

+ 2k

_

πi

_

, ¦z : e

z

= −1¦ = ¦(1 + 2k) πi¦ ,

¦z : e

z

= −i¦ =

__

−

1

2

+ 2k

_

πi

_

, ¦z : cos z = 0¦ =

__

1

2

+ k

_

π

_

,

and

¦z : sin z = 0¦ = ¦kπ¦

where k ∈ Z.

Exercise 7. Prove formulas for cos(z + w) and sin(z + w).

Solution. We have

cos(z) =

e

iz

+ e

−iz

2

sin(z) =

e

iz

− e

−iz

2i

.

26

1. Claim: cos(z) cos(w) − sin(z) sin(w) = cos(z + w).

Proof:

cos(z) cos(w) − sin(z) sin(w) =

e

iz

+ e

−iz

2

e

iw

+ e

−iw

2

−

e

iz

− e

−iz

2i

e

iw

− e

−iw

2i

=

1

4

_

e

iz

e

iw

+ e

−iz

e

−iw

_

+

1

4

_

e

iz

e

iw

+ e

−iz

e

−iw

_

=

1

2

e

iz

e

iw

+

1

2

e

−iz

e

−iw

=

1

2

_

e

iz

e

iw

+ e

−iz

e

−iw

_

= cos(z + w).

2. Claim: sin(z) cos(w) + cos(z) sin(w) = sin(z + w).

Proof:

sin(z) cos(w) + cos(z) sin(w) =

e

iz

− e

−iz

2i

e

iw

+ e

−iw

2

+

e

iz

+ e

−iz

2

e

iw

− e

−iw

2i

=

1

4i

_

e

iz

e

iw

− e

−iz

e

−iw

_

+

1

4i

_

e

iz

e

iw

− e

−iz

e

−iw

_

=

1

2i

e

iz

e

iw

−

1

2

e

−iz

e

−iw

=

1

2i

_

e

iz

e

iw

− e

−iz

e

−iw

_

= sin(z + w).

Exercise 8. Deﬁne tan z =

sin z

cos z

; where is this function deﬁned and analytic?

Solution. Not available.

Exercise 9. Suppose that z

n

, z ∈ G = C − ¦z : z ≤ 0¦ and z

n

= r

n

e

iθ

n

, z = re

iθ

where −π < θ, θ

n

< π. Prove

that if z

n

→ z then θ

n

→ θ and r

n

→ r.

Solution. Not available.

Exercise 10. Prove the following generalization of Proposition 2.20. Let G and Ωbe open in C and suppose

f and h are functions deﬁned on G, g : Ω → C and suppose that f (G) ⊂ Ω. Suppose that g and h are

analytic, g

·

(ω) 0 for any ω, that f is continuous, h is one-one, and that they satisfy h(z) = g( f (z)) for z in

G. Show that f is analytic. Give a formula for f

·

(z).

Solution. Not available.

Exercise 11. Suppose that f : G → C is a branch of the logarithm and that n is an integer. Prove that

z

n

= exp(nf (z)) for all z in G.

Solution. Not available.

Exercise 12. Show that the real part of the function z

1

2

is always positive.

Solution. We know that we can write z = re

iθ

0, −π < θ < π and Log(z) = ln(r) + iθ. Thus

z

1

2

= e

Logz

1

2

= e

1

2

Logz

= e

1

2

(lnr+iθ)

= e

1

2

lnr

e

1

2

θi

= e

1

2

lnr

_

cos

_

θ

2

_

+ i sin

_

θ

2

__

.

27

Hence,

Re(z) = e

1

2

lnr

cos

_

θ

2

_

> 0,

since e

1

2

lnr

> 0 and cos

_

θ

2

_

> 0 since −π < θ < π. Thus, the real part of the function z

1

2

is always positive.

Exercise 13. Let G = C − ¦z ∈ R : z ≤ 0¦ and let n be a positive integer. Find all analytic functions

f : G → C such that z = ( f (z))

n

for all z ∈ G.

Solution. Let Log(z) be the principal branch, then log(z) = Log(z) + 2kπi for some k ∈ Z. Thus, we can

write

f (z) = z

1/n

= e

log(z)/n

= e

(Log(z)+2kπi)/n

= e

Log(z)/n

e

2kπi/n

.

We know that the latter factor are the n-th roots of unity and depend only on k and n. They correspond to

the n distinct powers of the expression ζ = e

2πi/n

. Therefore, the branches of z

1/n

on the set U are given by

f (z) = ζ

k

e

Log(z)/n

,

where k = 0, . . . , n − 1 and therefore they are all constant multiples of each other.

Exercise 14. Suppose f : G → C is analytic and that G is connected. Show that if f (z) is real for all z in

G then f is constant.

Solution. First of all, we can write f : G → C as

f (z) = u(z) + iv(z)

where u, v are real-valued functions. Since f : G → C is analytic, that is f is continuously diﬀerentiable

(Deﬁnition 2.3), we have that u and v have continuous partial derivatives. By Theorem 2.29, this implies

that u, v satisfy the Cauchy-Riemann equations. That is,

∂u

∂x

=

∂v

∂y

and

∂u

∂y

= −

∂v

∂x

. (3.1)

Since f (z) is real ∀z ∈ G, this implies

v(z) ≡ 0

and therefore f (z) = u(z). So, since v(z) = 0, we have

∂v

∂x

=

∂v

∂y

= 0

and by (3.1) we obtain

∂u

∂x

=

∂u

∂y

= 0

and thus u

·

(z) = 0 (see reasoning of equation 2.22 and 2.23 on page 41). Hence f

·

(z) = 0. Since G is

connected and f : G → C is diﬀerentiable with f

·

(z) = 0 ∀z ∈ G, we have that f is constant.

Exercise 15. For r > 0 1et A =

_

ω : ω = exp

_

1

z

_

where 0 < ¦z¦ < r

_

; determine the set A.

28

Solution. Deﬁne the set S = ¦z : 0 < ¦z¦ < r¦ where r > 0. The image of this set under 1/z is clearly

T =

_

z :

1

r

< ¦z¦

_

.

To ﬁnd the image of A is the same as ﬁnding the image of T under e

z

.

Claim: The image of A is C − ¦0¦ (thus not depending on r).

To proof the claim, we need to show that for w 0, the equation e

z

= w has a solution z ∈ T. Using polar

coordinates we can write w = ¦w¦e

iθ

. We have to ﬁnd a complex number z = x +iy such that x

2

+y

2

>

1

r

and

e

x

e

iy

= w = ¦w¦e

iθ

. We want e

x

= ¦w¦ and y = θ + 2kπ, for some k ∈ Z. Using x = log ¦w¦ and k > 0 such

that (log ¦w¦)

2

+ (θ + 2kπ)

2

>

1

r

, then we found z = x + iy.

Exercise 16. Find an open connected set G ⊂ C and two continuous functions f and g deﬁned on G such

that f (z)

2

= g(z)

2

= 1 − z

2

for all z in G. Can you make G maximal? Are f and g analytic?

Solution. Not available.

Exercise 17. Give the principal branch of

√

1 − z.

Solution. Not available.

Exercise 18. Let f : G → C and g : G → C be branches of z

a

and z

b

respectively. Show that f g is a branch

of z

a+b

and f /g is a branch of z

a−b

. Suppose that f (G) ⊂ G and g(G) ⊂ G and prove that both f ◦ g and

g ◦ f are branches of z

ab

.

Solution. Not available.

Exercise 19. Let G be a region and deﬁne G

∗

= ¦z : ¯ z ∈ G¦. If f : G → C is analytic prove that

f

∗

: G

∗

→ C, deﬁned by f

∗

(z) = f ( ¯ x), is also analytic.

Solution. Let z = x + iy and let f (z) = u(x, y) + iv(x, y). By assumption f is analytic and therefore

u and v have continuous partial derivatives. In addition the Cauchy-Riemann Equations u

x

= v

y

and

u

y

= −v

x

are satisﬁed. Since f

∗

(z) = f ( ¯ x), we have f

∗

(z) = u

∗

(x, y) + iv

∗

(x, y) where u

∗

(x, y) = u(x, −y)

and v

∗

(x, y) = −v(x, −y). Hence, we have u

∗

x

(x, y) = u

x

(x, −y) = v

y

(x, −y), u

∗

y

(x, y) = −u

y

(x, −y) = v

x

(x, −y),

v

∗

x

(x, y) = −v

x

(x, −y) and v

∗

y

(x, y) = v

y

(x, −y) and therefore u

∗

x

= v

∗

y

and u

∗

y

= −v

∗

x

so f

∗

is analytic.

Exercise 20. Let z

1

, z

2

, . . . , z

n

be complex numbers such that Re z

k

> 0 and Re(z

l

. . . z

k

) > 0 for 1 ≤ k ≤ n.

Show that log(z

1

. . . z

n

) = log z

1

+ . . . + log z

n

, where log z is the principal branch of the logarithm. If the

restrictions on the z

k

are removed, does the formula remain valid?

Solution. Let z

1

, . . . , z

n

∈ C such that Re(z

j

) > 0 and Re(z

1

z

j

) > 0 for 1 ≤ j ≤ n. The proof will be

by induction. Consider ﬁrst the case where n = 2. Let z

1

, z

2

∈ C as above. Then −

π

2

< arg z

1

<

π

2

, −

π

2

<

arg z

2

<

π

2

and −

π

2

< arg(z

1

z

2

) <

π

2

. But note arg(z

1

z

2

) = arg z

1

+arg z

2

implies that −π < arg z

1

+arg z

2

< π.

Now, recall

log(z

1

z

2

) = ln ¦z

1

z

2

¦ + i arg(z

1

z

2

)

= ln ¦z

1

¦ + ln ¦z

2

¦ + i arg(z

1

) + i arg(z

2

)

= log(z

1

) + log(z

2

)

Assume this formula is true for n = k − 1. Let z

1

, . . . , z

k

∈ C as above. Then

log(z

1

. . . z

k

) = log((z

1

z

k−1

)z

k

)

= log(z

1

z

k−1

) + log z

k

by the case when n = 2

= log z

1

+ log z

2

+ . . . + log z

k−1

+ log z

k

by the case when n = k − 1

29

Hence, this is true for all n such that z

i

, 1 ≤ i ≤ n, satisfy the restrictions.

If the restriction on the z

k

are removed, does the formula remain valid? Consider the complex numbers

z

1

= −1 + 2i, z

2

= −2 + i. Then z

1

z

2

= 0 − 5i. Clearly, z

1

, z

2

, and z

1

z

2

do not meet the restrictions as stated

above. Now,

log(z

1

) = ln ¦z

1

¦ + i arg z

1

log(z

2

) = ln ¦z

2

¦ + i arg z

2

Thus log z

1

+ log z

2

= ln

√

5 + ln

√

5 + i(

3π

2

) = ln 5 + i(

3π

2

), but note that

3π

2

(−π, π) and log(z

1

z

2

) =

ln 5 + i(−

π

2

). Thus log z

1

+ log z

2

log(z

1

z

2

) where log z is the principal branch of the logarithm. Hence,

the formula is invalid.

Exercise 21. Prove that there is no branch of the logarithm deﬁned on G = C− ¦0¦. (Hint: Suppose such a

branch exists and compare this with the principal branch.)

Solution. Deﬁne the subset

ˆ

G of G by

ˆ

G = C − ¦z ∈ R : z ≤ 0¦. We use the notation Log for the principal

part of the log on

ˆ

G, that is

Log(z) = log ¦z¦ + i arg(z)

where arg(z) ∈ (−π, π). We will prove the statement above by contradiction.

Assume f (z) is a branch of the logarithm deﬁned on G. Then restricting f to

ˆ

G gives us a branch of the log

on

ˆ

G. Therefore its only diﬀerence to the principal branch is 2πik for some k ∈ Z. This yields

f (z) = log ¦z¦ + i arg(z) + 2πik

where z ∈

ˆ

G. Since f is analytic in G, it is continuous at −1. But we can check that this is not the case, thus

we have derived a contradiction

3.3 Analytic functions as mappings. Möbius transformations

Exercise 1. 1. Find the image of ¦z : Re z < 0, ¦Im z¦ < π¦ under the exponential function.

Solution. We have

¦z : Re z < 0, ¦Im z¦ < π¦ = ¦z = x + iy : x < 0, −π < y < π¦.

The image of ¦z = x + iy : x < 0, −π < y < π¦ under the exponential function is given by

¦e

x+iy

: x < 0, −π < y < π¦ = ¦e

x

e

iy

: x < 0, −π < y < π¦

= ¦e

x

(cos(y) + i sin(y)) : x < 0, −π < y < π¦

= ¦r(cos(y) + i sin(y)) : 0 < r < 1, −π < y < π¦.

If 0 < r < 1 is ﬁxed, then r(cos(y) + i sin(y)) describes a circle with radius r without the point (−r, 0)

centered at (0, 0).

Since r varies between 0 and 1, we get that the image is a solid circle with radius 1, where the boundary

does not belong to it, the negative x-axis does not belong to it and the origin does not belong to is.

Exercise 2. Do exercise 1 for the set ¦z : ¦Im z¦ < π/2¦.

30

Solution. We have

¦z : ¦Im z¦ < π/2¦ = ¦z = x + iy : x ∈ R, −

π

2

< y <

π

2

¦.

The image of ¦z = x + iy : x ∈ R, −

π

2

< y <

π

2

¦ under the exponential function is given by

¦e

x+iy

: x ∈ R, −

π

2

< y <

π

2

¦ = ¦e

x

(cos(y) + i sin(y)) : x ∈ R, −

π

2

< y <

π

2

¦

= ¦r(cos(y) + i sin(y)) : r > 0, −

π

2

< y <

π

2

¦.

If r > 0 is ﬁxed, then r(cos(y) + i sin(y)) describes a half circle with radius r centered at (0, 0) lying in the

right half plane not touching the imaginary axis.

Since r varies between 0 and inﬁnity, we get that the image is the right half plane without the imaginary

axis.

Exercise 3. Discuss the mapping properties of cos z and sin z.

Solution. Not available.

Exercise 4. Discuss the mapping properties of z

n

and z

1/n

for n ≥ 2. (Hint: use polar coordinates.)

Solution. Not available.

Exercise 5. Find the ﬁxed points of a dilation, a translation and the inversion on C

∞

.

Solution. In general, we have

S (z) =

az + b

cz + d

.

To get a dilation S (z) = az, we have that b = 0, c = 0, d = 1 and a > 0. To ﬁnd a ﬁxed point, we have to ﬁnd

all z such that S (z) = z. In this case az = z. Obviously z = 0 is a ﬁxed point. Also z = ∞ is a ﬁxed point,

since a ∞ = ∞(a > 0) or S (∞) =

a

c

=

a

0

= ∞.

To get a translation S (z) = z + b, we have that a = 1, c = 0, d = 1 and b ∈ R. To ﬁnd a ﬁxed point, we have

to ﬁnd all z such that S (z) = z. In this case z +b = z, which is true if z = ∞. To see that S (∞) =

a

c

=

1

0

= ∞.

To get a inversion S (z) =

1

z

, we have that a = 0, b = 1, c = 1 and d = 0. To ﬁnd a ﬁxed point, we have to

ﬁnd all z such that S (z) = z. In this case

1

z

= z which is equivalent to z

2

= 1 and thus z = 1 and z = −1 are

ﬁxed points.

Exercise 6. Evaluate the following cross ratios: (a) (7 + i, 1, 0, ∞) (b) (2, 1 − i, 1, 1 + i) (c) (0, 1, i, −1) (d)

(i − 1, ∞, 1 + i, 0).

Solution. We have

S (z) =

z − z

3

z − z

4

/

z

2

− z

3

z

2

− z

4

, if z

2

, z

3

, z

4

∈ C (3.2)

S (z) =

z − z

3

z − z

4

, if z

2

= ∞ (3.3)

S (z) =

z − z

3

z

2

− z

3

, if z

4

= ∞. (3.4)

a) By (3.4) we get

(7 + i, 1, 0, ∞) =

7 + i − 0

1 − 0

= 7 + i.

31

b) By (3.2) we get

(2, 1 − i, 1, 1 + i) =

2 − 1

2 − 1 − i

/

1 − i − 1

1 − i − 1 − i

=

1

1 − i

/

−i

−2i

=

2

1 − i

= 2

1 + i

(1 − i)(i + 1)

= 2

1 + i

2

= 1 + i.

c) By (3.2) we get

(0, 1, i, −1) =

0 − i

0 + 1

/

1 − i

1 + 1

= −i/

1 − i

2

= −i

2

1 − i

1 + i

1 + i

= −i

2

2

(1 + i) = −i − i

2

= 1 − i.

d) By (3.3) we get

(i − 1, ∞, 1 + i, 0) =

i − 1 − 1 − i

i − 1 − 0

=

−2

i − 1

i + 1

i + 1

=

−2

−2

(i + 1) = 1 + i.

Exercise 7. If Tz =

az+b

cz+d

, ﬁnd z

2

, z

3

, z

4

(in terms of a, b, c, d) such that Tz = (z, z

2

, z

3

, z

4

).

Solution. The inverse of T is given by

T

−1

(z) =

dz − b

−cz + a

as shown on p. 47. We have T

−1

(1) =

d−b

a−c

, T

−1

(0) = −

b

a

and T

−1

(∞) = −

d

c

. Set z

2

=

d−b

a−c

, z

3

= −

b

a

and

z

4

= −

d

c

to obtain Tz = (z, z

2

, z

3

, z

4

).

Exercise 8. If Tz =

az+b

cz+d

show that T(R

∞

) = R

∞

iﬀ we can choose a, b, c, d to be real numbers.

Solution. Not available.

Exercise 9. If Tz =

az+b

cz+d

, ﬁnd necessary and suﬃcient conditions that T(Γ) = Γ where Γ is the unit circle

¦z : ¦z¦ = 1¦.

Solution. We want if z¯ z = 1, then T(z)T(z) = 1.

T(z)T(z) = 1 ⇐⇒

az + b

cz + d

az + b

cz + d

= 1 ⇐⇒

(az + b)(¯ a¯ z +

¯

b)

(cz + d)(¯ c¯ z +

¯

d)

= 1

⇐⇒ az¯ a¯ z + b¯ a¯ z + az

¯

b + b

¯

b = cz¯ c¯ z + d¯ c¯ z +

¯

dcz + d

¯

d

⇐⇒ z¯ z(a¯ a − c¯ c) + z(a

¯

b − c

¯

d) + ¯ z(b¯ a + d¯ c) − d

¯

d = 0.

This is the same as the equation z¯ z − 1 = 0 if

a

¯

b − c

¯

d = 0

b¯ a − d¯ c = 0

and

a¯ a − c¯ c = 1

b

¯

b − d

¯

d = −1

which is equivalent to ¦a¦

2

− ¦c¦

2

= 1 and 1 = ¦d¦

2

− ¦b¦

2

. Hence, we have the two conditions

a

¯

b − c

¯

d = 0 and ¦a¦

2

+ ¦b¦

2

= ¦c¦

2

+ ¦d¦

2

.

32

These are the suﬃcient conditions. Let c = λ

¯

b, then a

¯

b − c

¯

d = 0 yields a

¯

b − λ

¯

b

¯

d = 0 iﬀ a = λ

¯

d iﬀ d =

¯ a

¯

λ

.

Insert this into ¦c¦

2

+ ¦d¦

2

= ¦a¦

2

+ ¦b¦

2

yields

¦λ¦

2

¦b¦

2

+

¦a¦

2

¦λ¦

2

= ¦a¦

2

+ ¦b¦

2

,

and therefore we can take ¦λ¦ = 1. Then

1

¯

λ

= λ, and so the form of the Möbius transformation is

T(z) =

az + b

λ(

¯

bz + ¯ a)

, where ¦λ¦ = 1

or

T(z) =

¯

λ

az + b

¯

bz + ¯ a

, where ¦λ¦ = 1.

Take λ = e

−iθ

, then

T(z) = e

iθ

az + b

¯

bz + ¯ a

, for some θ.

This mapping transforms ¦z¦ = 1 into ¦T(z)¦ = 1.

Exercise 10. Let D = ¦z : ¦z¦ < 1¦ and ﬁnd all Möbius transformations T such that T(D) = D.

Solution. Take an α ∈ D = ¦z : ¦z¦ < 1¦ such that T(α) = 0. Its symmetric point with respect to the unit

circle is

α

∗

=

1

¯ α

since

z

∗

− a =

R

2

¯ z − ¯ a

(see page 51) with a = 0 and R = 1 (the unit circle). Therefore T(α

∗

) = ∞. Thus T looks like

T(z) = K

z − α

¯ αz − 1

where K is a constant. (It is easy to check that T(α) = 0 and T(α

∗

) = T(

1

¯ α

) = ∞). Finally, we are going to

choose the constant K in such a way that ¦T(z

0

)¦ = 1 where z

0

= e

iθ

. We have

T(z

0

) = K

e

iθ

− α

¯ αe

iθ

− 1

and therefore

1 = ¦T(z

0

)¦ = ¦K¦

¦e

iθ

− α¦

¦ ¯ αe

iθ

− 1¦

= ¦K¦

_

e

iθ

− α

_ _

e

−iθ

− ¯ α

_

¦e

iθ

¦ ¦ ¯ α − e

−iθ

¦

= ¦K¦

_

e

iθ

− α

_ _

e

−iθ

− ¯ α

_

_

¯ α − e

−iθ

_ _

α − e

iθ

_ = ¦K¦.

So ¦K¦ = 1 implies K = e

iθ

for some real θ. We arrive at

T(z) = e

iθ

z − α

¯ αz − 1

, for some real θ.

Exercise 11. Show that the deﬁnition of symmetry (3.17) does not depend on the choice of points z

2

, z

3

, z

4

.

That is, showthat if ω

2

, ω

3

, ω

4

are also in Γ then equation (3.18) is satisﬁed iﬀ (z

∗

, ω

2

, ω

3

, ω

4

) = (z, ω

2

, ω

3

, ω

4

).

(Hint: Use Exercise 8.)

33

Solution. Not available.

Exercise 12. Prove Theorem 3.4.

Solution. Not available.

Exercise 13. Give a discussion of the mapping f (z) =

1

2

(z + 1/z).

Solution. The function f (z) =

1

2

_

z +

1

z

_

=

z

2

+1

2z

can be deﬁned for all z ∈ C − ¦0¦ and therefore also in the

punctured disk 0 < ¦z¦ < 1. To see that in this domain the function is injective, let z

1

, z

2

be two numbers in

the domain of f with the same image, then

0 =f (z

1

) − f (z

2

) =

z

2

1

+ 1

z

1

−

z

2

2

+ 1

z

2

=

z

2

1

z

2

+ z

2

− z

1

z

2

2

− z

1

z

1

z

2

=

(z

1

z

2

− 1)(z

1

− z

2

)

z

1

z

2

.

With the assumption 0 < ¦z

i

¦ < 1, i = 1, 2 the factor z

1

z

2

−1, is always nonzero and we conclude that z

1

= z

2

;

hence f (z) is injective.

The range of the function is C − ¦z ∈ C ¦ Tz ∈ [−1, 1] and Imz = 0¦. To see this, write z = re

iθ

in polar

coordinates and let f (z) = w = a + ib, a, b ∈ R.

f (z) = f (re

iθ

) =

1

2

_

re

iθ

+

1

r

e

−iθ

_

=

1

2

__

r +

1

r

_

cos θ + i

_

r −

1

r

_

sin θ

_

.

For the real and imaginary part of w the following equations must hold

a =

1

2

_

r +

1

r

_

cos θ

b =

1

2

_

r −

1

r

_

sin θ.

(3.5)

If f (z) = w has imaginary part b = 0 then sin θ = 0 and ¦cos θ¦ = 1. Therefore points of the form

a + ib, a ∈ [−1, 1], b = 0 cannot be in the range of f . For all other points the equations (3.5) can be solved

for r and θ uniquely (after restricting the argument to [−π, π)).

Given any value of r ∈ (0, 1), the graph of f (re

iθ

) as a function of θ looks like an ellipse. In fact from

formulas (3.5) we see that

_

a

1

2

(r+

1

r

)

_

2

+

_

b

1

2

(1−

1

r

)

_

2

= 1.

If we ﬁx the argument θ and let r vary in (0, 1) it follows from from equation (3.5) that the graph of f (re

iθ

)

is a hyperbola and it degenerates to rays if z is purely real or imaginary. In the case θ ∈ ¦(2k + 1)π ¦ k ∈ Z¦

the graph of f in dependence on r is on the imaginary axis and for θ ∈ ¦2kπ ¦ k ∈ Z¦ the graph of f (re

iθ

) is

either (−∞, −1) or (1, ∞). If cos θ 0 and sin θ 0 then

_

a

cos θ

_

2

−

_

b

sin θ

_

2

= 1.

Exercise 14. Suppose that one circle is contained inside another and that they are tangent at the point a.

Let G be the region between the two circles and map G conformally onto the open unit disk. (Hint: ﬁrst try

(z − a)

−1

.)

Solution. Using the hint, deﬁne the Möbius transformation T(z) = (z − a)

−1

which sends the region G

between two lines. Afterward applying a rotation followed by a translation, it is possible to send this region

to any region bounded by two parallel lines we want. Hence, choose S (z) = cz + d where ¦c¦ = 1 such that

S (T(G)) =

_

x + iy : 0 < y <

π

2

_

.

Applying the exponential function to this region yields the right half plane

exp(S (T(G))) = ¦x + iy : x > 0¦.

34

Finally, the Möbius transformation

R(z) =

z − 1

z + 1

maps the right half plane onto the unit disk (see page 53). Hence the function f deﬁned by R(exp(S (T(z))))

maps G onto D and is the desired conformal mapping ( f is a composition of conformal mappings). Doing

some simpliﬁcations, we obtain

f (z) =

e

c

z−a

+d

− 1

e

c

z−a

+d

+ 1

where the constants c and d will depend on the circle location.

Exercise 15. Can you map the open unit disk conformally onto ¦z : 0 < ¦z¦ < 1¦?

Solution. Not available.

Exercise 16. Map G = C − ¦z : −1 ≤ z ≤ 1¦ onto the open unit disk by an analytic function f . Can f be

one-one?

Solution. Not available.

Exercise 17. Let G be a region and suppose that f : G → C is analytic such that f (G) is a subset of a

circle. Show that f is constant.

Solution. Not available.

Exercise 18. Let −∞ < a < b < ∞and put Mz =

z−ia

z−ib

. Deﬁne the lines L

1

= ¦z : Im z = b¦, L

2

= ¦z : Im z =

a¦ and L

3

= ¦z : Re z = 0¦. Determine which of the regions A, B, C, D, E, F in Figure 1, are mapped by M

onto the regions U, V, W, X, Y, Z in Figure 2.

Solution. We easily see that we have M(ia) = 0. Therefore the region B, C, E and F which touch the line

ia are mapped somehow to the region U, V, X and Y which touch 0. Similarly we have M(ib) = ∞ and

therefore the region B and E which touch the line ib are mapped somehow to the region U and X which

touch ∞. Thus we conclude that C or F goes to either V or Y. Let us ﬁnd out. Let x, y be small positive

real numbers such that the poin z = x + iy + ia ∈ E. Thus, the imaginary part of Mz is a positive number

multiplied by x(b − a) and therefore also positive. Therefore we conclude that M maps E to U and B to X.

Because B and C meet at the line ia, we conclude that X and M(C) do, too. Hence, M maps C to Y and F

to V. By a similar argument, we obtain that M maps A to Z and D to W.

Exercise 19. Let a, b, and M be as in Exercise 18 and let log be the principal branch of the logarithm.

(a) Show that log(Mz) is deﬁned for all z except z = ic, a ≤ c ≤ b; and if h(z) = Im [log Mz] then

0 < h(z) < π for Re z > 0.

(b) Show that log(z − ic) is deﬁned for Re z > 0 and any real number c; also prove that ¦Imlog(z − ic)¦ <

π

2

if Re z > 0.

(c) Let h be as in (a) and prove that h(z) = Im [log(z − ia) − log(z − ib)].

(d) Show that

_

b

a

dt

z − it

= i[log(z − ib) − log(z − ia)]

(Hint: Use the Fundamental Theorem of Calculus.)

(e) Combine (c) and (d) to get that

h(x + iy) =

_

b

a

x

x

2

+ (y − t)

2

dt = arctan

_

y − a

x

_

− arctan

_

y − b

x

_

(f) Interpret part (e) geometrically and show that for Re z > 0, h(z) is the angle depicted in the ﬁgure.

35

Solution. Not available.

Exercise 20. Let S z =

az+b

cz+d

and Tz =

αz+β

γz+δ

, show that S = T iﬀ there is a non zero complex number λ such

that α = λa, β = λb, γ = λc, δ = λd.

Solution. ⇐: Let λ 0 be a complex number such that

α = λa

β = λb

γ = λc

δ = λd.

Then

T(z) =

αz + β

γz + δ

=

λaz + λb

λcz + λd

=

λ(az + b)

λ(cz + d)

=

(az + b)

(cz + d)

= S (z).

Thus, S = T.

⇒: Let S = T, that is S (z) = T(z). Then S (0) = T(0), S (1) = T(1) and S (∞) = T(∞) which is equivalent

to

b

d

=

β

δ

= λ

1

⇐⇒ b = λ

1

d, β = λ

1

δ (3.6)

a + b

c + d

=

α + β

γ + δ

(3.7)

a

c

=

α

γ

= λ

3

⇐⇒ a = λ

3

c, α = λ

3

γ. (3.8)

Insert b = λ

1

d, β = λ

1

δ, a = λ

3

c and α = λ

3

γ into (3.7), then

λ

3

c + λ

1

d

c + d

=

λ

3

γ + λ

1

δ

γ + δ

⇐⇒ λ

3

cγ + λ

1

dγ + λ

3

cδ + λ

1

dδ = λ

3

γc + λ

1

δc + λ

3

γd + λ

1

δd

⇐⇒ λ

1

dγ + λ

3

cδ − λ

1

δc − λ

3

γd = 0

⇐⇒ λ

1

(dγ − δc) − λ

3

(dγ − δc) = 0

⇐⇒ (λ

1

− λ

3

)(dγ − δc) = 0.

Thus, either λ

1

− λ

3

= 0 or dγ − δc = 0.

If λ

1

−λ

3

= 0, then from (3.6) and (3.7), we get λ

3

=

a

c

=

b

d

= λ

1

which implies ad −bc = 0 a contradiction

(Not possible, otherwise we do not have a Möbius transformation).

Thus, dγ − δc = 0 ⇐⇒

c

d

=

γ

δ

⇐⇒

γ

c

=

δ

d

:= λ, λ 0 or

γ = cλ (3.9)

δ = dλ. (3.10)

Insert (3.9) and (3.10) into (3.6) to obtain

b

d

=

β

δ

⇒

b

d

=

β

dλ

⇒ β = λb.

Insert (3.9) and (3.10) into (3.8) to obtain

a

c

=

α

γ

⇒

a

c

=

α

cλ

⇒ α = λa.

Therefore, we have

α = λa, β = λb, γ = λc, δ = λd.

36

Exercise 21. Let T be a Möbius transformation with ﬁxed points z

1

and z

2

. If S is a Möbius transformation

show that S

−1

TS has ﬁxed points S

−1

z

1

, and S

−1

z

2

.

Solution. To show

S

−1

TS (S

−1

z

1

) = S

−1

z

1

.

We have

S

−1

TS (S

−1

z

1

) = S

−1

TS S

−1

z

1

= S

−1

Tz

1

= S

−1

z

1

,

where the ﬁrst step follows by the associativity of compositions, the second step since S S

−1

= id and the

last step by Tz

1

= z

1

since z

1

is a ﬁxed point of T.

To show

S

−1

TS (S

−1

z

2

) = S

−1

z

2

.

We have

S

−1

TS (S

−1

z

2

) = S

−1

TS S

−1

z

2

= S

−1

Tz

2

= S

−1

z

2

,

where the ﬁrst step follows by the associativity of compositions, the second step since S S

−1

= id and the

last step by Tz

2

= z

2

since z

2

is a ﬁxed point of T.

Note that S

−1

TS is a well deﬁned Möbius transformation, since S and T are Möbius transformation. The

composition of Möbius transformation is a Möbius transformation.

Exercise 22. (a) Show that a Möbius transformation has 0 and ∞as its only ﬁxed points iﬀ it is a dilation,

but not the identity.

(b) Show that a Möbius transformation has ∞as its only ﬁxed point iﬀ it is a translation, but not the identity.

Solution. Not available.

Exercise 23. Show that a Möbius transformation T satisﬁes T(0) = ∞ and T(∞) = 0 iﬀ Tz = az

−1

for

some a in C.

Solution. Not available.

Exercise 24. Let T be a Möbius transformation, T the identity. Show that a Möbius transformation S

commutes with T if S and T have the same ﬁxed points. (Hint: Use Exercises 21 and 22.)

Solution. Let T and S have the same ﬁxed points. To show TS = S T, T id.

: Suppose T and S have two ﬁxed points, say z

1

and z

2

. Let M be a Möbius transformation with M(z

1

) = 0

and M(z

2

) = ∞. Then

MS M

−1

(0) = MS M

−1

Mz

1

= MS z

1

= Mz

1

= 0

and

MS M

−1

(∞) = MS M

−1

Mz

2

= MS z

2

= Mz

2

= ∞.

Thus MS M

−1

is a dilation by exercise 22 a) since MS M

−1

has 0 and ∞ as its only ﬁxed points. Similar,

we obtain MT M

−1

(0) = 0 and MT M

−1

(∞) = ∞ and therefore is also a dilation. It is easy to check that

dilations commute (deﬁne C(z) = az, a > 0 and D(z) = bz, b > 0, then CD(z) = abz = baz = DC(z)), thus

(MT M

−1

)(MS M

−1

) = (MS M

−1

)(MT M

−1

)

⇒ MTS M

−1

= MS T M

−1

⇒ TS = S T.

37

: Suppose T and S have one ﬁxed points, say z. Let M be a Möbius transformation with M(z) = ∞. Then

MS M

−1

(∞) = MS M

−1

Mz = MS z = Mz = ∞.

Thus MS M

−1

is a translation by exercise 22 b) since MS M

−1

has ∞ as its only ﬁxed point. Similar, we

obtain MT M

−1

(∞) = ∞ and therefore is also a translation. It is easy to check that translations commute

(deﬁne C(z) = z + 1 and D(z) = z + b, b > 0, then CD(z) = z + a + b = z + b + a = DC(z)), thus

(MT M

−1

)(MS M

−1

) = (MS M

−1

)(MT M

−1

)

⇒ MTS M

−1

= MS T M

−1

⇒ TS = S T.

Exercise 25. Find all the abelian subgroups of the group of Möbius transformations.

Solution. Not available.

Exercise 26. 26. (a) Let GL

2

(C) = all invertible 2 × 2 matrices with entries in C and let A be the group

of Möbius transformations. Deﬁne ϕ : GL

2

(C) → A by ϕ

_

a b

c d

_

=

az+b

cz+d

. Show that ϕ is a group

homomorphism of GL

2

(C) onto A. Find the kernel of ϕ.

(b) Let S L

2

(C) be the subgroup of GL

2

(C) consisting of all matrices of determinant 1. Show that the image

of S L

2

(C) under ϕ is all of A. What part of the kernel of ϕ is in S L

2

(C)?

Solution. a) We have to check that if

A =

_

a b

c d

_

and B =

_

ˆ a

ˆ

b

ˆ c

ˆ

d

_

then ϕ(AB) = ϕ(A) ◦ ϕ(B). A simple calculation shows that this is true. To ﬁnd the kernel of the group

homomorphism we have to ﬁnd all z such that

az+b

cz+d

= z. This is equivalent to az + b = cz

2

+ dz and by

comparing coeﬃcients we obtain b = c = 0 and a = d. Therefore, the kernel is given by N = ker(ϕ) = ¦λI :

λ ∈ C

×

¦. Note that the kernel is a normal subgroup of GL

2

(C).

b) Restricting ϕ to S L

2

(C) still yields a surjective map since for any matrix A ∈ GL

2

(C) both A and

the modiﬁcation M =

1

√

det A

A have the same image and the modiﬁcation matrix M has by construction

determinant 1. The kernel of the restriction is simply N ∩ S L

2

(C) = ¦±I¦.

Exercise 27. If gis a group and A is a subgroup then A is said to be a normal subgroup of gif S

−1

TS ∈ A

whenever T ∈ A and S ∈ g. g is a simple group if the only normal subgroups of g are ¦I¦ (I = the identity

of g) and g itself. Prove that the group Aof Möbius transformations is a simple group.

Solution. Not available.

Exercise 28. Discuss the mapping properties of (1 − z)

i

.

Solution. Not available.

Exercise 29. For complex numbers α and β with ¦α¦

2

+ ¦β¦

2

= 1

u

α,β

=

αz −

¯

β

βz + ¯ α

and let U = ¦u

α,β

: ¦α¦

2

+ ¦β¦

2

= 1¦

(a) Show that U is a group under composition.

(b) If S U

2

is the set of all unitary matrices with determinant 1, show that S U

2

is a group under matrix

38

multiplication and that for each A in S U

2

there are unique complex numbers α and β with ¦α¦

2

+ ¦β¦

2

= 1

and

A =

_

α β

−

¯

β ¯ α

_

(c) Show that

_

α β

−

¯

β ¯ α

_

·→ u

α,β

is an isomorphism of the group S U

2

onto U. What is its kernel?

(d) If l ∈ ¦0,

1

2

, 1,

3

2

, . . .¦ let H

l

= all the polynomials of degree ≤ 2l. For u

α,β

= u in U deﬁne T

(l)

u

: H

l

→ H

l

,

by (T

(l)

u

f )(z) = (βz + ¯ α)

2l

f (u(z)). Show that T

(l)

u

is an invertible linear transformation on H

l

, and u ·→ T

(l)

u

is an injective homomorphism of U into the group of invertible linear transformations of H

l

, onto H

l

.

Solution. Not available.

Exercise 30. For ¦z¦ < l deﬁne f (z) by

f (z) = exp

_

¸

¸

_

¸

¸

_

−i log

_

i

_

1 + z

1 − z

__

1/2

_

¸

¸

_

¸

¸

_

(a) Show that f maps D = ¦z : ¦z¦ < 1¦ conformally onto an annulus G.

(b) Find all Möbius transformations S (z) that map D onto D and such that f (S (z)) = f (z) when ¦z¦ < 1.

Solution. Not available.

39

Chapter 4

Complex Integration

4.1 Riemann-Stieltjes integrals

Exercise 1. Let γ : [a, b] → R be non decreasing. Show that γ is of bounded variation and V(γ) =

γ(b) − γ(a).

Solution. Let γ : [a, b] → R be a monotone, non-decreasing function. For a given partition of [a, b],

¦a = t

0

< t

1

< . . . < t

m

= b¦, for all n ∈ N, γ(t

n

) ≥ γ(t

n−1

). Therefore, ¦γ(t

n

) − γ(t

n−1

)¦ = γ(t

n

) − γ(t

n−1

) ≥ 0.

Let P = ¦a = t

0

< t

1

< . . . < t

m

= b¦, any partition of [a, b]. Then,

m

n=1

¦γ(t

n

) − γ(t

n−1

)¦ = [γ(t

m

) − γ(t

m−1

)] + [γ(t

m−1

) − γ(t

m−2

)] + . . . + [γ(t

2

) − γ(t

1

)] + [γ(t

1

) − γ(t

0

)]

= [γ(b) − γ(t

m−1

)] + [γ(t

m−1

) − γ(t

m−2

)] + . . . + [γ(t

2

) − γ(t

1

)] + [γ(t

1

) − γ(a)]

= γ(b) − γ(a)

Since γ(a), γ(b) ∈ R, γ(b) − γ(a) < ∞. It follows that γ is of bounded variation for any partition P. And

also v(γ; P) = sup¦v(γ; P) : P a partition of [a, b]¦ = V(γ) = γ(b) − γ(a).

Exercise 2. Prove Proposition 1.2.

Solution. Let γ : [a, b] → C be of bounded variation.

(a) If P and Q are partitions of [a, b] and P ⊂ Q then v(γ; P) ≤ v(γ; Q).

Proof will be by induction. Let P = ¦a = t

0

< t

1

< . . . < t

m

= b¦. Suppose Q = P ∪ ¦x¦, where t

k

< x < t

k+1

.

Then,

v(γ; P) =

ik+1

¦γ(t

i

) − γ(t

i−1

)¦ + ¦γ(t

k+1

) − γ(t

k

)¦

≤

ik+1

¦γ(t

i

) − γ(t

i−1

)¦ + ¦γ(x) − γ(t

k

)¦ + ¦γ(t

k+1

) − γ(x)¦

= v(γ; Q)

Now consider the general case on the number of elements of Q \ P. The smallest partition set for Q \ P

contains the trivial partition composed of ¦a, b¦. Using the case deﬁned above, we have v(γ; Q) ≥ v(γ; P) ≥

v(γ; Q \ P) ≥ ¦γ(b) − γ(a)¦, whenever P ⊂ Q.

(b) If σ : [a, b] → C is also of bounded variation and α, β ∈ C, then αγ + βσ is of bounded variation and

40

V(αγ + βσ) ≤ ¦α¦V(γ) + ¦β¦V(σ).

Let P = ¦a = t

0

< t

1

< . . . < t

m

= b¦ be a partition. Since γ and σ are of bounded variation for [a, b] ⊂ R,

there exists constants M

γ

> 0 and M

σ

> 0, respectively, such that v(γ; P) =

m

k=1

¦γ(t

k

) − γ(t

k−1

)¦ ≤ M

γ

and

v(γ; P) =

m

k=1

¦σ(t

k

) − σ(t

k−1

)¦ ≤ M

σ

. Then,

v(αγ + βσ; P) =

m

k=1

¦(αγ + βσ)(t

k

) − (αγ + βσ)(t

k−1

)¦

=

m

k=1

¦αγ(t

k

) + βσ(t

k

) − αγ(t

k−1

) − βσ(t

k−1

)¦

=

m

k=1

¦(αγ(t

k

) − αγ(t

k−1

)) + (βσ(t

k

) − βσ(t

k−1

))¦

≤

m

k=1

¦αγ(t

k

) − αγ(t

k−1

)¦ + ¦βσ(t

k

) − βσ(t

k−1

)¦

=

m

k=1

¦αγ(t

k

) − αγ(t

k−1

)¦ +

m

k=1

¦βσ(t

k

) − βσ(t

k−1

)¦

=

m

k=1

¦α¦ ¦(γ(t

k

) − γ(t

k−1

))¦ +

m

k=1

¦β¦ ¦(σ(t

k

) − σ(t

k−1

))¦

= ¦α¦

m

k=1

¦γ(t

k

) − γ(t

k−1

)¦ + ¦β¦

m

k=1

¦σ(t

k

) − σ(t

k−1

)¦

= ¦α¦v(γ; P) + ¦β¦v(σ; P)

≤ ¦α¦sup¦v(γ; P) : P a partition of [a, b]¦ + ¦β¦sup¦v(σ; P) : P a partition of [a, b]¦

= ¦α¦V(γ) + ¦β¦V(σ)

< ∞

The result is two-fold, V(αγ + βσ) ≤ ¦α¦V(γ) + ¦β¦V(σ) and it follows αγ + βσ is of bounded variation by

¦α¦V(γ) + ¦β¦V(σ).

Exercise 3. Prove Proposition 1.7.

Solution. Not available.

Exercise 4. Prove Proposition 1.8 (Use induction).

Solution. Not available.

Exercise 5. Let γ(t) = exp((−1 + i)/t

−1

) for 0 < t ≤ 1 and γ(0) = 0. Show that γ is a rectiﬁable path and

ﬁnd V(γ). Give a rough sketch of the trace of γ.

Solution. Not available.

Exercise 6. Show that if γ : [a, b] → C is a Lipschitz function then γ is of bounded variation.

41

Solution. Let γ : [a, b] → C be Lipschitz, that is ∃ a constant C > 0 such that

¦γ(x) − γ(y)¦ ≤ C¦x − y¦, ∀x, y ∈ [a, b].

Then, for any partition P = ¦a = t

0

< t

1

< . . . < t

m

= b¦ of [a, b], we have

v(γ; P) =

m

k=1

¦γ(t

k

) − γ(t

k−1

)¦ ≤

m

k=1

C¦t

k

− t

k−1

¦ = C

m

k=1

¦t

k

− t

k−1

¦ = C(b − a) =: M > 0.

Hence, the function γ : [a, b] → C, for [a, b] ⊂ R, is of bounded variation, since there is a constant M > 0

such that for any partition P = ¦a = t

0

< t

1

< . . . < t

m

= b¦ of [a, b]

v(γ; P) =

m

k=1

¦γ(t

k

) − γ(t

k−1

)¦ ≤ M.

Exercise 7. Show that γ : [0, 1] → C, deﬁned by γ(t) = t + it sin

1

t

for t 0 and γ(0) = 0, is a path but is

not rectiﬁable. Sketch this path.

Solution. Not available.

Exercise 8. Let γ and σ be the two polygons [1, i] and [1, 1 + i, i]. Express γ and σ as paths and calculate

_

γ

f and

_

σ

f where f (z) = ¦z¦

2

.

Solution. For γ we have γ(t) = (1 − t) + it for 0 ≤ t ≤ 1. So γ

·

(t) = −1 + i. Therefore,

_

γ

¦z¦

2

dz =

_

1

0

((1 − t)

2

+ t

2

)(−1 + i)dt

= (−1 + i)

_

1

0

(2t

2

− 2t + 1)dt

= (−1 + i)

_

2

3

t

3

− t

2

+ t

_

1

0

= (−1 + i)

_

2

3

− 1 + 1

_

= −

2

3

+ i

2

3

.

Since σ is composed of multiple lines, the path is given by ¦γ

1

(t) = 1 + it : 0 ≤ t ≤ 1; γ

2

(t) = (1 − t) + i :

42

0 ≤ t ≤ 1¦. So γ

·

1

(t) = i and γ

·

2

(t) = −1. Therefore,

_

γ

f =

_

γ

1

f +

_

γ

2

f

=

_

γ

1

¦z¦

2

dz +

_

γ

2

¦z¦

2

dz

=

_

1

0

(t

2

+ 1)(i)dt +

_

1

0

((1 − t)

2

+ 1)(−1)dt

= i

_

1

0

(t

2

+ 1)dt −

_

1

0

(t

2

− 2t + 2)dt

= i

_

t

3

3

+ t

_

1

0

−

_

t

3

3

− t

2

+ 2t

_

1

0

=

4

3

i −

_

1

3

− 1 + 2

_

=

4

3

i −

4

3

=

4

3

(−1 + i)

Exercise 9. Deﬁne γ : [0, 2π] → C by γ(t) = exp(int) where n is some integer (positive, negative, or zero).

Show that

_

γ

1

z

dz = 2πin.

Solution. Clearly, γ(t) = e

int

is continuous and smooth on [0, 2π]. Thus

_

γ

z

−1

dz =

_

2π

0

e

−int

ine

int

dt =

_

2π

0

in dt = in(2π − 0) = 2πin.

Exercise 10. Deﬁne γ(t) = e

it

for 0 ≤ t ≤ 2π and ﬁnd

_

γ

z

n

dz for every integer n.

Solution. Clearly, γ(t) = e

int

is continuous and smooth on [0, 2π] (It is the unit circle).

Case 1: n = −1

_

γ

z

−1

dz =

_

2π

0

e

−it

ie

it

dt = i

_

2π

0

dt = 2πi.

Case 2: n −1

_

γ

z

n

dz =

_

2π

0

e

int

ie

it

dt = i

_

2π

0

e

i(n+1)t

dt = t

_

e

i(n+1)t

i(n + 1)

_

2π

0

=

1

n + 1

_

e

i(n+1)t

_

2π

0

=

1

n + 1

_

e

i(n+1)2π

− 1

_

=

1

n + 1

[cos((n + 1)2π) − i sin((n + 1)2π)1] =

1

n + 1

[1 − i 0 − 1]

= 0.

Exercise 11. Let γ be the closed polygon [1 − i, 1 + i, −1 + i, −1 − i, 1 − i]. Find

_

γ

1

z

dz.

43

Solution. Deﬁne

γ

1

(t) = 1 + it, t from − 1 to 1 → γ

·

1

(t) = i

γ

2

(t) = t + i, t from 1 to − 1 → γ

·

2

(t) = 1

γ

3

(t) = −1 + it, t from 1 to − 1 → γ

·

3

(t) = i

γ

4

(t) = t − i, t from − 1 to 1 → γ

·

4

(t) = 1.

Thus

_

γ

1

z

dz

=

_

1

−1

1

1 + it

i dt +

_

−1

1

1

t + i

dt +

_

−1

1

1

−1 + it

i dt +

_

1

−1

1

t − i

dt

=

_

1

−1

_

−1

t + i

+

1

t − i

_

dt + i

_

1

−1

_

1

1 + it

−

1

−1 + it

_

dt

=

_

1

−1

−t + i + t + i

(t + i)(t − i)

dt + i

_

1

−1

−1 + it − 1 − it

(1 + it)(−1 + it)

dt

= 2i

_

1

−1

1

t

2

+ 1

dt − 2i

_

1

−1

1

−1 − t

2

dt

= 2i

_

1

−1

1

t

2

+ 1

dt + 2i

_

1

−1

1

t

2

+ 1

dt

= 4i

_

1

−1

1

t

2

+ 1

dt = 4i [arctan(z)]

1

−1

= 4i

_

π

4

−

_

−

π

4

__

= 4i

π

2

= 2πi.

Exercise 12. Let I(r) =

_

γ

e

iz

z

dz where γ : [0, π] → C is deﬁned by γ(t) = re

it

. Show that lim

r→∞

I(r) = 0.

Solution. To show lim

r→∞

I(r) = 0 which is equivalent to ∀ > 0, ∃n ∈ N such that if r > N, then ¦I(r)¦ < .

Let > 0, then

¦I(r) − 0¦ = ¦I(r)¦ =

¸

¸

¸

¸

¸

¸

_

γ

e

iz

z

dz

¸

¸

¸

¸

¸

¸

=

¸

¸

¸

¸

¸

¸

_

π

0

e

ire

it

re

it

rie

it

dt

¸

¸

¸

¸

¸

¸

=

¸

¸

¸

¸

¸

i

_

π

0

e

ire

it

dt

¸

¸

¸

¸

¸

≤

_

π

0

¸

¸

¸

¸

e

ire

it

¸

¸

¸

¸

dt.

Now, recall that ¦e

z

¦ = e

Re(z)

(p. 38 eq. 2.13), so

¸

¸

¸

¸

e

ire

it

¸

¸

¸

¸

= e

Re(ire

it

)

= e

Re(ir cos(t)+iri sin(t))

= e

Re(ir cos(t)−r sin(t))

= e

−r sin(t)

.

Hence,

_

π

0

¸

¸

¸

¸

e

ire

it

¸

¸

¸

¸

dt =

_

π

0

e

−r sin(t)

dt =

_

δ

1

0

e

−r sin(t)

dt +

_

π−δ

2

δ

1

e

−r sin(t)

dt +

_

π

π−δ

2

e

−r sin(t)

dt

for δ

1

> 0 and δ

2

> 0. Since e

−r sin(t)

is continuous, so δ

1

> 0 can be chosen such that

_

δ

1

0

e

−r sin(t)

dt <

3

, ∀r > 0.

Similar, δ

2

> 0 can be chosen such that

_

π

π−δ

2

e

−r sin(t)

dt <

3

, ∀r > 0.

44

Finally, because of the Lebesgue’s dominated convergence Theorem

lim

r→∞

_

π−δ

2

δ

1

e

−r sin(t)

dt =

_

π−δ

2

δ

1

lim

r→∞

e

−r sin(t)

dt

we have

lim

r→∞

_

π−δ

2

δ

1

e

−r sin(t)

dt <

3

.

Hence,

lim

r→∞

I(r) = 0.

Exercise 13. Find

_

γ

z

−

1

2

dz where: (a) γ is the upper half of the unit circle from +1 to −1: (b) γ is the

lower half of the unit circle from +1 to −1.

Solution. Not available.

Exercise 14. Prove that if ϕ : [a, b] → [c, d] is continuous and ϕ(a) = c, ϕ(b) = d then ϕ is one-one iﬀ ϕ is

strictly increasing.

Solution. Not available.

Exercise 15. Show that the relation in Deﬁnition 1.16 is an equivalence relation.

Solution. Not available.

Exercise 16. Show that if γ and σ are equivalent rectiﬁable paths then V(γ) = V(σ).

Solution. Not available.

Exercise 17. Show that if γ : [a, b] → C is a path then there is an equivalent path σ : [0, 1] → C.

Solution. Not available.

Exercise 18. Prove Proposition 1.17.

Solution. Not available.

Exercise 19. Let γ = 1 + e

it

for 0 ≤ t ≤ 2π and ﬁnd

_

γ

(z

2

− 1)

−1

dz.

Solution. Not available.

Exercise 20. Let γ = 2e

it

for −π ≤ t ≤ π and ﬁnd

_

γ

(z

2

− 1)

−1

dz.

Solution. Not available.

Exercise 21. Show that if F

1

and F

2

are primitives for f : G → C and G is open end connected then there

is a constant c such that F

1

(z) = c + F

2

(z) for each z in G.

Solution. Let F

1

and F

2

be primitives for f : G → C. Recall then that F

·

1

= f = F

·

2

. Let h = F

1

− F

2

for

all z ∈ G. Then for all z ∈ G,

h

·

(z) = F

·

1

(z) − F

·

2

(z) = f (z) − f (z) = 0

Since G is open and connected with h

·

(z) = 0 for all z ∈ G, then h is a constant, say c: h = c. So

h = F

1

− F

2

= c and F

1

(z) = F

2

(z) + c for all z ∈ G.

45

Exercise 22. Let γ be a closed rectiﬁable curve in an open set G and a G. Show that for n ≥ 2,

_

γ

(z − a)

−n

dz = 0.

Solution. Let f (z) = (z − a)

−n

for all z ∈ G. Then f has the antiderivative F(z) =

(z − a)

1−n

1 − n

on G. Since

F

·

(z) = f (z) where n ≥ 2 (note F is undeﬁned for n = 1). Now, let γ be a closed rectiﬁable curve in G with

endpoints z

1

, z

2

∈ C. Since z

1

= z

2

, F(z

1

) = F(z

2

). So,

_

γ

(z − a)

−n

dz = F(z

2

) − F(z

1

) = 0.

Exercise 23. Prove the following integration by parts formula. Let f and g be analytic in G and let γ be a

rectiﬁable curve from a to b in G. Then

_

γ

f g

·

= f (b)g(b) f (a)g(a) −

_

γ

f

·

g.

Solution. Let f , g be analytic in G and let γ be a rectiﬁable curve from a to b in G. Then

_

γ

f g

·

+

_

γ

f

·

g =

(def)

_

b

a

f (γ(t))g

·

(γ(t)) dγ(t) +

_

b

a

f

·

(γ(t))g(γ(t)) dγ(t)

=

Prop. 1.7 (a)

_

b

a

f (γ(t))g

·

(γ(t)) + f

·

(γ(t))g(γ(t)) dγ(t)

=

(def)

_

γ

f g

·

+ f

·

g =

_

γ

( f g)

·

Since ( f g)

·

is continuous ( f and g are analytic) by Theorem 1.18 we obtain

_

γ

( f g)

·

= ( f g)(b) − ( f g)(a) = f (b)g(b) − f (a)g(a).

Hence,

_

γ

f g

·

= f (b)g(b) − f (a)g(a) −

_

γ

f

·

g.

4.2 Power series representation of analytic functions

Exercise 1. Show that the function deﬁned by (2.2) is continuous.

Solution. Not available.

Exercise 2. Prove the following analogue of Leibniz.s rule (this exercise will be frequently used in the later

sections.) Let G be an open set and let γ be a rectiﬁable curve in C. Suppose that ϕ : ¦y¦ × G → C is a

continuous function and deﬁne g : G → C by

g(z) =

_

γ

ϕ(w, z) dw

then g is continuous. If

∂ϕ

∂z

exists for each (w, z) in ¦γ¦ × G and is continuous then g is analytic and

g

·

(z) =

_

γ

∂ϕ

∂z

(w, z) dw.

Solution. Not available.

46

Exercise 3. Suppose that γ is a rectiﬁable curve in C and ϕ is deﬁned and continuous on ¦γ¦. Use Exercise

2 to show that

g(z) =

_

γ

ϕ(w)

w − z

dw

is analytic on C − ¦γ¦ and

g

(n)

(z) = n!

_

γ

ϕ(w)

(w − z)

n+1

dw.

Solution. Not available.

Exercise 4. (a) Prove Abel’s Theorem: Let

_

a

n

(z − a)

n

have radius of convergence 1 and suppose that

_

a

n

converges to A. Prove that

lim

r→1

−

a

n

r

n

= A.

(Hint: Find a summation formula which is the analogue of integration by parts.)

(b) Use Abel’s Theorem to prove that log 2 = 1 −

1

2

+

1

3

− . . ..

Solution. Not available.

Exercise 5. Give the power series expansion of log z about z = i and ﬁnd its radius of convergence.

Solution. Assume a ∈ C is not zero, then

1

z

=

1

a + z − a

=

1

a

1

a+z−a

a

=

1

a

1

1 +

z−a

a

=

1

a

∞

n=0

(−1)

n

a

n

(z − a)

n

=

∞

n=0

(−1)

n

a

n+1

(z − a)

n

,

where the radius of convergence is ¦a¦. Let a = i, then integrate term by term yields

log(z) =

π

2

i +

∞

n=0

_

i

2

_

n

(n + 1)i

n+1

(z − i)

n+1

+

π

2

i −

∞

n=0

i

n+1

n + z

(z − i)

n+1

=

π

2

i −

∞

n=1

(1 + iz)

n

n

,

since Log(i) =

π

2

i. The convergence radius is ¦i¦ = 1.

Exercise 6. Give the power series expansion of

√

z about z = 1 and ﬁnd its radius of convergence.

Solution. From the deﬁnition of numbers with rational exponents z

a

= exp(a log z) we see that the square

root cannot be analytic on all of C. In fact, taking the ﬁrst derivatives we see that f

(n)

(z) = 2

−n

n−1

k=0

(1 −

2k)z

−n+

1

2

and therefore

a

n

=

1

n!

f

(n)

(1) =

1

n!2

n

n−1

_

k=0

,

and with these a

n

,

f (z) =

√

z =

∞

n=0

a

n

(z − 1)

z

=

∞

k=0

1

n!2

n

n−1

_

k=0

(1 − 2k)(z − 1)

n

.

Since the logarithm is undeﬁned at zero, no matter which branch we consider, the radius of convergence is

1.

47

Exercise 7. Use the results of this section to evaluate the following integrals:

(a)

_

γ

e

iz

z

2

dz, γ(t) = e

it

, 0 ≤ t ≤ 2π;

(b)

_

γ

dz

z − a

, γ(t) = a + re

it

, 0 ≤ t ≤ 2π;

(c)

_

γ

sin(z)

z

3

dz, γ(t) = e

it

, 0 ≤ t ≤ 2π;

(d)

_

γ

log z

z

n

dz, γ(t) = 1 +

1

2

e

it

, 0 ≤ t ≤ 2π and n ≥ 0.

Solution. a) Let a = 0, r = 1, n = 1, f (z) = e

iz

and f

·

(z) = ie

iz

. Clearly f (z) is analytic on C and

¯

B(0; 1) ⊂ C. Now use Corollary 2.13.

f

·

(0) =

1!

2πi

_

γ

e

iz

(z − 0)

2

dz ⇐⇒ i =

1

2πi

_

γ

e

iz

z

2

dz ⇐⇒

_

γ

e

iz

z

2

dz = −2π.

b) Let a = a, r = r, n = 0, and f (z) = 1. Clearly f (z) is analytic on C and

¯

B(a; r) ⊂ C. Now use Corollary

2.13.

f

(0)

(a) =

0!

2πi

_

γ

1

(z − a)

1

dz ⇐⇒ 1 =

1

2πi

_

γ

1

z − a

dz ⇐⇒

_

γ

1

z − a

dz = 2πi.

c) Let a = 0, r = 1, n = 2 and f (z) = sin(z). Then f

·

(z) = cos(z) and f

··

(z) = −sin(z). Clearly f (z) is

analytic on C (p. 38) and

¯

B(0; 1) ⊂ C. Now use Corollary 2.13.

f

··

(0) =

2!

2πi

_

γ

sin(z)

(z − 0)

3

dz ⇐⇒ −sin(0) =

1

πi

_

γ

sin(z)

z

3

dz ⇐⇒ 0 =

1

πi

_

γ

sin(z)

z

3

dz

⇐⇒

_

γ

sin(z)

z

3

dz = 0.

d) We have that

f (z) =

log(z)

z

n

is analytic in the disk B(1;

1

2

) where n ≥ 0. Furthermore, γ is a closed rectiﬁable curve (γ(t) is a circle with

radius r = 1/2 and center (1, 0) in C). Obviously γ(t) = 1 +

1

2

e

it

⊆ B(1;

1

2

). Hence, by Proposition 2.15, f

has a primitive and therefore

_

γ

f = 0. Therefore,

_

γ

log(z)

z

n

dz = 0.

Exercise 8. Use a Möbuis transformation to show that Proposition 2.15 holds if the disk B(a; R) is replaced

by a half plane.

Solution. Not available.

48

Exercise 9. Evaluate the following integrals:

(a)

_

γ

e

z

− e

−z

z

n

dz where n is a positive integer and γ(t) = e

it

, 0 ≤ t ≤ 2π;

(b)

_

γ

dz

_

z −

1

2

_

n

where n is a positive integer and γ(t) =

1

2

+ e

it

, 0 ≤ t ≤ 2π;

(c)

_

γ

dz

z

2

+ 1

where γ(t) = 2e

it

, 0 ≤ t ≤ 2π. (Hint: expand (z

2

+ 1)

−1

by means of partial fractions);

(d)

_

γ

sin z

z

dz where γ(t) = e

it

, 0 ≤ t ≤ 2π;

(e)

_

γ

z

1/m

(z − 1)

m

dz where γ(t) = 1 +

1

2

e

it

, 0 ≤ t ≤ 2π;

Solution. a) Let a = 0, r = 1, n = m − 1 and

f (z) = e

z

− e

−z

f

·

(z) = e

z

+ e

−z

f

··

(z) = e

z

− e

−z

.

.

.

f

(m−1)

(z) = e

z

− (−1)

m+1

e

−z

.

Clearly f (z) is analytic on C and

¯

B(0; 1) ⊂ C. Now use Corollary 2.13.

f

(m−1)

(0) =

(m − 1)!

2πi

_

γ

e

z

− e

−z

(z − 0)

m

dz

⇐⇒

(m − 1)!

2πi

_

γ

e

z

− e

−z

z

m

dz =

_

¸

¸

_

¸

¸

_

0 , m odd

2 , m even

⇐⇒

_

γ

e

z

− e

−z

z

m

dz =

_

¸

¸

_

¸

¸

_

0 , m odd

4πi

(m−1)!

, m even

(m > 0).

b) Let a = 1/2, r = 1, n = m − 1 and

f (z) = 1

f

·

(z) = 0

f

··

(z) = 0

.

.

.

f

(m−1)

(z) = 0.

49

Clearly f (z) is analytic on C and

¯

B(

1

2

; 1) ⊂ C. Now use Corollary 2.13.

f

(m−1)

_

1

2

_

=

(m − 1)!

2πi

_

γ

1

_

z −

1

2

_

m

dz

⇐⇒

(m − 1)!

2πi

_

γ

1

_

z −

1

2

_

m

dz =

_

¸

¸

_

¸

¸

_

1 , m = 1

0 , m > 1

⇐⇒

_

γ

1

_

z −

1

2

_

m

dz =

_

¸

¸

_

¸

¸

_

2πi , m = 1

0 , m > 1

.

c) Using partial fraction decomposition, we obtain

1

z

2

+ 1

!

=

A

z − i

+

B

z + i

=

Az + Ai + Bz − Bi

z

2

+ 1

which implies

A + B = 0 (4.1)

Ai − Bi = 1. (4.2)

From (4.1), we get A = −B inserted into (4.2) yields −2Bi = 1 and therefore B = −

1

2i

and A =

1

2i

. Thus

_

γ

dz

z

2

+ 1

=

1

2i

_

γ

dz

z − i

−

1

2i

_

γ

dz

z + i

.

For the ﬁrst integral on the right side, use Proposition 2.6. Let a = 0, r = 2 and f (w) = 1, γ(t) = 2e

it

,

0 ≤ t ≤ 2π. Clearly f (z) is analytic on C. We also have

¯

B(0; 2) ⊂ C. Then

f (z) =

1

2πi

_

γ

f (w)

w − i

dw ⇐⇒ 1 =

1

2πi

_

γ

1

w − i

dw ⇐⇒

_

γ

1

z − i

dz = 2πi,

where ¦ − i − 0¦ < 2 ⇐⇒ 1 < 2.

For the second integral on the right side, use Proposition 2.6. Let a = 0, r = 2 and f (w) = 1, γ(t) = 2e

it

,

0 ≤ t ≤ 2π. Clearly f (z) is analytic on C. We also have

¯

B(0; 2) ⊂ C. Then

f (z) =

1

2πi

_

γ

f (w)

w + i

dw ⇐⇒ 1 =

1

2πi

_

γ

1

w + i

dw ⇐⇒

_

γ

1

z + i

dz = 2πi,

where ¦i¦ < 2 ⇐⇒ 1 < 2. Hence

_

γ

dz

z

2

+ 1

=

1

2i

_

γ

dz

z − i

−

1

2i

_

γ

dz

z + i

=

1

2i

2πi −

1

2i

2πi = π − π = 0.

So

_

γ

dz

z

2

+ 1

= 0.

d) Let a = 0, r = 1, n = 0 and f (z) = sin(z). Clearly f (z) is analytic on C and

¯

B(0; 1) ⊂ C. Now use

Corollary 2.13.

f

(0)

(0) =

0!

2πi

_

γ

sin(z)

(z − 0)

1

dz ⇐⇒ 0 =

1

2πi

_

γ

sin(z)

z

dz ⇐⇒

_

γ

sin(z)

z

dz = 0.

50

e) Let a = 1, r = 1/2, n = m − 1 and

f (z) = z

1/m

f

·

(z) =

1

m

z

1/m−1

f

··

(z) =

1

m

_

1

m

− 1

_

z

1/m−2

.

.

.

f

(m−1)

(z) =

1

m

_

1

m

− 1

_

. . .

_

1

m

− (m − 2)

_

z

1/m−(m−1)

.

Clearly f (z) is analytic on C and

¯

B(1;

1

2

) ⊂ C. Now use Corollary 2.13.

f

(m−1)

(1) =

(m − 1)!

2πi

_

γ

z

1/m

(z − 1)

m

dz

⇐⇒

1

m

_

1

m

− 1

_

. . .

_

1

m

− (m − 2)

_

1 =

(m − 1)!

2πi

_

γ

z

1/m

(z − 1)

m

dz

⇐⇒

_

γ

z

1/m

(z − 1)

m

dz =

2πi

(m − 1)!

1

m

_

1 − m

m

_ _

1 − 2m

m

_

. . .

_

1 − (m − 1)m

m

_

⇐⇒

_

γ

z

1/m

(z − 1)

m

dz =

2πi

(m − 1)!m

m−1

m−2

_

i=0

(1 − im).

Exercise 10. Evaluate

_

γ

z

2

+1

z(z

2

+4)

dz where γ(t) = re

it

, 0 ≤ t ≤ 2π, for all possible values of r, 0 < r < 2 and

2 < r < ∞.

Solution. Let f (z) =

z

2

+ 1

z(z

2

+ 4)

. Then f (z) can be decomposed using partial fractions as

f (z) =

z

2

+ 1

z(z

2

+ 4)

=

1/4

z

+

3/8

z − 2i

+

3/8

z + 2i

.

It follows,

_

γ

z

2

+ 1

z(z

2

+ 4)

dz =

1

4

_

γ

1

z

dz +

3

8

_

γ

1

z − 2i

dz +

3

8

_

γ

1

z + 2i

dz. By Proposition 2.6,

_

γ

1

z

dz =

2πi f (0) = 2πi where f ≡ 1 as f ≡ 1 is analytic on B(0, 2) and γ = 0 + re

it

with 0 < r < 2. Now, note

1

z−2i

is

analytic on B(0, 2) and γ for 0 < r < 2 is in B(0, 2). Thus, by Proposition 2.15,

_

γ

1

z − 2i

dz = 0. Similarly,

_

γ

1

z + 2i

dz = 0. It follows for 0 < r < 2,

_

γ

z

2

+ 1

z(z

2

+ 4)

dz =

1

4

2πi =

π

2

i. Now, for 2 < r < ∞, note γ(t) = re

it

so γ

·

(t) = ire

it

. Calculating the integrals in turn,

_

γ

1

z

dz =

_

2π

0

ire

it

re

it

dt = i

_

2π

0

dt = 2πi.

For the second integral, we use Proposition 2.6 with a = 0, 2 < r < ∞, f (w) = 1, and γ(t) = 0 + re

it

for

51

0 ≤ t ≤ 2π. Clearly f (z) is analytic on C and

¯

B(0; 2 < r < ∞) ⊂ C. Then,

f (−2i) =

1

2πi

_

γ

f (w)

w + 2i

dw ⇐⇒ 1 =

1

2πi

_

γ

f (w)

w + 2i

dw

⇐⇒

_

γ

1

z + 2i

dz = 2πi

if 2 < r < ∞. Similarly,

f (2i) =

1

2πi

_

γ

f (w)

w − 2i

dw ⇐⇒

_

γ

1

z − 2i

dz = 2πi.

It follows that,

_

γ

z

2

+ 1

z(z

2

+ 4)

dz =

1

4

_

γ

1

z

dz +

3

8

_

γ

1

z − 2i

dz +

3

8

_

γ

1

z + 2i

dz

=

1

4

(2πi) +

3

8

(2πi) +

3

8

(2πi)

= 2πi

Exercise 11. Find the domain of analyticity of

f (z) =

1

2i

log

_

1 + iz

1 − iz

_

;

also, show that tan f (z) = z (i.e., f is a branch of arctan z). Show that

f (z) =

∞

k=0

(−1)

k

z

2k+1

2k + 1

, for ¦z¦ < 1

(Hint: see Exercise III. 3.19.)

Solution. Not available.

Exercise 12. Show that

sec z = 1 +

∞

k=1

E

2k

(2k)!

z

2k

for some constants E

2

, E

4

, . These numbers are called Euler’s constants. What is the radius of conver-

gence of this series? Use the fact that 1 = cos z sec z to show that

E

2n

−

_

2n

2n − 2

_

E

2n−2

+

_

2n

2n − 4

_

E

2n−4

+ + (−1)

n−1

_

2n

2

_

E

2

+ (−1)

n

= 0.

Evaluate E

2

, E

4

, E

6

, E

8

. (E

10

= 50521 and E

12

= 2702765).

Solution. It is easily seen that

sec z = 1 +

∞

k=1

E

2k

(2k)!

z

2k

, (for some E

2

, E

4

, . . .),

52

since sec(0) =

1

cos(0)

=

1

1

= 1 and sec z =

1

cos z

is an even function. The radius of convergence is the closest

distance from 0 to the nearest non-analytic point which is precisely

π

2

.

Since 1 = cos z sec z, we can multiply the series for sec z and cos z together to obtain

1 =

∞

i=0

_

¸

¸

¸

¸

¸

_

n

k=0

(−1)

i−k

E

2k

(2k)!(2i − 2k)!

_

¸

¸

¸

¸

¸

_

z

2i

.

Finally, we can compare the coeﬃcients of z

2i

and multiply by (2i)! to obtain the formula

n

i=0

(−1)

n−k

E

2k

_

2n

2k

_

= 0

which is equivalent to

E

2n

−

_

2n

2n − 2

_

E

2n−2

+

_

2n

2n − 4

_

E

2n−4

+ + (−1)

n−1

_

2n

2

_

E

2

+ (−1)

n

= 0.

In addition, we have E

2

= 1, E

4

= 5, E

6

= 61 and E

8

= 1285.

Exercise 13. Find the series expansion of

e

z

−1

z

about zero and determine its radius of convergence. Consider

f (z) =

z

e

z

−1

and let

f (z) =

∞

k=0

a

k

k!

z

k

be its power series expansion about zero. What is the radius of convergence? Show that

0 = a

0

+

_

n + 1

1

_

a

1

+ +

_

n + 1

n

_

a

n

.

Using the fact that f (z) +

1

2

z is an even function show that a

k

= 0 for k odd and k > 1. The numbers

B

2n

= (−1)

n−1

a

2n

are called the Bernoulli numbers for n ≥ 1. Calculate B

2

, B

4

, , B

l0

.

Solution. The Bernoulli numbers B

n

are deﬁned by

B

n

= lim

z→∞

d

n z

e

z

−1

dz

n

, n = 0, 1, 2, . . .

We determine the largest R such that

z

e

z

− 1

=

∞

k=0

B

k

k!

z

k

for 0 < ¦z¦ < R.

The function f (z) =

z

e

z

−1

has a removable singularity at z = 0, because it can be extended to 0 by an

analytic function g on a small neighborhood B(0, r) of 0, setting

g(z) :=

_

¸

¸

_

¸

¸

_

z

e

z

−1

if 0 < ¦z¦ < r

1 if x = 0

.

Furthermore the function f has essential singularities exactly at z = 2πk, k ∈ Z − ¦0¦. Therefore f (z) has

a Laurent series expansion on the punctured disk ann(0; 0, 2π) and the negative indices disappear because

the singularity at the origin is a removable one. The uniqueness of the Laurent expression together with the

53

formula for the coeﬃcients in the expansion (Theorem 2.8, p. 72) give the desired result.

First we compute B

0

= lim

z→0

z

e

z

−1

= 1 and note that the result in question does not hold for B

0

. Also

compute B

1

. Taking the limits, B

1

= lim

z→0

(1−z)e

z

−1

(e

z

−1)

2

= −

1

2

. Consider the equation

z

e

z

−1

=

_

∞

k=0

B

k

k!

z

k

for

z ∈ ann(0; 0, 2π). This is equivalent to

z =

_

¸

¸

¸

¸

¸

_

∞

l=1

z

l

l!

_

¸

¸

¸

¸

¸

_

_

¸

¸

¸

¸

¸

_

∞

k=0

B

k

k!

z

k

_

¸

¸

¸

¸

¸

_

.

Divide by z 0, then

1 =

_

¸

¸

¸

¸

¸

_

∞

l=1

z

l−1

l!

_

¸

¸

¸

¸

¸

_

_

¸

¸

¸

¸

¸

_

∞

k=0

B

k

k!

_

¸

¸

¸

¸

¸

_

=

_

¸

¸

¸

¸

¸

_

∞

l=0

z

l

(l + 1)!

_

¸

¸

¸

¸

¸

_

_

¸

¸

¸

¸

¸

_

∞

k=0

B

k

k!

_

¸

¸

¸

¸

¸

_

=

∞

n=0

n

m=0

B

m

m!

z

m

z

n−m

(n + 1 − m)!

=

∞

n=0

z

n

n

m=0

B

m

m!(n + 1 − m)!

=

∞

n=0

z

n

(n + 1)!

n

m=0

_

n + 1

m

_

B

m

= B

0

+

∞

n=1

z

n

(n + 1)!

n

m=0

_

n + 1

m

_

B

m

.

Since B

0

= 0 conclude that

0 =

∞

n=1

z

n

(n + 1)!

n

m=0

_

n + 1

m

_

B

m

for all nonzero z ∈ ann(0; 0, 2π). This can only happen if the inner summation equals zero for all n ∈ N.

This prove the ﬁrst part of the statement. For the second claim, one only needs to show that f (z) +

1

2

z is an

even function. It is true that

f (z) +

1

2

z =

2z + (e

z

− 1)z

2(e

z

− 1)

=

z(e

z

+ 1)

2(e

z

− 1)

.

as well as t’

f (−z) −

1

2

z =

−z(e

z

+ 1)

e

−z

− 1

=

−z(1 + e

z

)

2(1 − e

z

)

=

z(1 + e

z

)

2(e

z

− 1)

= f (z) +

1

2

z.

With the power series expansion of f and recalling that B

1

= −

1

2

this is the same as saying

1 +

∞

k=2

B

k

k!

(−z)

k

= 1 +

∞

k=2

B

k

k!

(z)

k

or, yet in other terms,

B

k

= −B

k

= 0 for k odd , k > 1.

54

In part b) the values of B

0

= 1 and B

1

= −

1

2

have been computed already. Also B

3

= B

5

= B

7

= 0 by

the second statement of part b). It remains to compute B

2

, B

4

, B

6

and B

8

. The formula involving binomial

coeﬃcients in the part above can be solved for B

n

and

B

n

= −

1

n + 1

_

¸

¸

¸

¸

¸

¸

_

n−1

k=0

_

n + 1

k

_

B

k

_

¸

¸

¸

¸

¸

¸

_

.

This gives

B

2

= −

1

3

(B

0

+ 3B

1

) = −

1

3

_

1 −

3

2

_

=

1

6

,

B

4

= −

1

5

(B

0

+ 5B

1

+ 10B

2

) = −

1

30

,

B

6

= −

1

7

(B

0

+ 7B

1

+ 21B

2

+ 35B

4

) =

1

42

,

and

B

8

= −

1

9

(B

0

+ 9B

1

+ 36B

2

+ 126B

4

+ 84B

6

) = −

1

30

.

Exercise 14. Find the power series expansion of tan z about z = 0, expressing the coeﬃcients in terms of

Bernoulli numbers. (Hint: use Exercise 13 and the formula cot 2z =

1

2

cot z −

1

2

tan z.)

Solution. To ﬁnd the power series expansion of tan z about z = 0, we replace z by 2iz in the function

f (z) =

z(e

z

+ 1)

2(e

z

− 1)

which we obtained in the previous problem. We obtain

f (2iz) =

2iz(e

2iz

+ 1)

2(e

2iz

− 1)

=

iz(e

iz

+ e

−iz

)

e

iz

− e

−iz

= z

cos(z)

sin(z)

= z cot(z).

Next, we replace z by 2iz of the power series expansion for f (z) to obtain

z cot(z) = 1 −

∞

n=1

4

n

B

2n

(2n)!

z

2n

.

Multiplying the formula cot 2z =

1

2

cot z −

1

2

tan z with 2z gives

z tan z = z cot z − 2z cot 2z

and therefore

z tan z =

_

¸

¸

¸

¸

¸

_

1 −

∞

n=1

4

n

B

2n

(2n)!

z

2n

_

¸

¸

¸

¸

¸

_

−

_

¸

¸

¸

¸

¸

_

1 −

∞

n=1

4

2n

B

2n

(2n)!

z

2n

_

¸

¸

¸

¸

¸

_

=

∞

n=1

4

n

(4

n

− 1)B

2n

(2n)!

z

2n

.

Hence,

tan(z) =

∞

n=1

4

n

(4

n

− 1)B

2n

(2n)!

z

2n−1

.

55

4.3 Zeros of an analytic function

Exercise 1. Let f be an entire function and suppose there is a constant M, an R > 0, and an integer n ≥ 1

such that ¦ f (z)¦ ≤ M¦z¦

n

for ¦z¦ > R. Show that f is a polynomial of degree ≤ n.

Solution. Since f is an entire function, that is f ∈ A(C), we have by Proposition 3.3 p. 77, that f has a

power series expansion

f (z) =

∞

k=0

a

k

z

k

, wherea

k

=

f

(k)

(0)

k!

with inﬁnite radius of convergence. We have to show that f

(k)

(0) = 0 for k ≥ n + 1, then we are done. Let

r > R and let γ(t) = re

it

, 0 ≤ t ≤ 2π. Note that γ is a closed and rectiﬁable curve. Then

¦ f

(k)

(0)¦ =

Corollary 2.13 p.73

¸

¸

¸

¸

¸

¸

k!

2πi

_

γ

f (z)

z

k+1

dz

¸

¸

¸

¸

¸

¸

≤

k!

2π

_

γ

¦ f (z)¦

¦z¦

k+1

¦dz¦.

Since r > R, ¦γ(t)¦ = ¦re

it

¦ = r > R, 0 ≤ t ≤ 2π. So ¦ f (z)¦ ≤ M¦z¦

n

, z ∈ ¦γ¦, we get

k!

2π

_

γ

¦ f (z)¦

¦z¦

k+1

¦dz¦ ≤

k!

2π

_

γ

M¦z¦

n

¦z¦

k+1

¦dz¦ =

k!M

2π

_

γ

¦z¦

n−k−1

¦dz¦

=

¦z¦=r

k!M

2π

r

n−k−1

_

γ

¦dz¦ =

k!M

2π

r

n−k−1

2πr

= k!Mr

n−k

= k!M

1

r

k−n

→ 0

as r → ∞since k ≥ n + 1 ⇐⇒ k − n ≥ 1. Hence, ¦ f

(k)

(0)¦ = 0 ∀k ≥ n + 1 and thus

f (z) =

n

k=0

f

(k)

(0)

k!

z

k

;

a polynomial of degree less or equal n.

Exercise 2. Give an example to show that G must be assumed to be connected in Theorem 3.7.

Solution. Not available.

Exercise 3. Find all entire functions f such that f (x) = e

x

for x in R.

Solution. Let f (z) = e

z

. Clearly f is analytic on C since it is entire. Then for all x ∈ R, f (x) = e

x

. Now let

g be an entire function such that g(x) = e

x

for all x ∈ R. Now, set H = ¦z ∈ G : f (z) = g(z)¦. So, R ⊆ H.

Clearly 0 is a limit point of H and 0 ∈ G. Hence, by Corollary 3.8, f (z) ≡ g(z). So f (z) = e

z

is the set of all

entire functions f such that f (x) = e

x

for all x ∈ R.

Exercise 4. Prove that e

z+a

= e

z

e

a

by applying Corollary 3.8.

Solution. Not available.

Exercise 5. Prove that cos(a + b) = cos a cos b − sin a sin b by applying Corollary 3.8.

Solution. Not available.

Exercise 6. Let G be a region and suppose that f : G → C is analytic and a ∈ G such that ¦ f (a)¦ ≤ ¦ f (z)¦

for all z in G. Show that either f (a) = 0 or f is constant.

56

Solution. Let f : G → C be analytic, where G is a region. Let a ∈ G such that

¦ f (a)¦ ≤ ¦ f (z)¦ ∀z ∈ G. (4.3)

Obviously, f (a) satisﬁes (4.3) since 0 ≤ ¦ f (z)¦ ∀z ∈ G is true.

Now, suppose f (a) 0 ∀a ∈ G. Deﬁne

g(z) =

1

f (z)

.

Clearly, g ∈ A(G) (since f (a) 0 ∀a ∈ G) and

¦g(z)¦ =

(def)

1

¦ f (z)¦

≤

(4.3)

1

¦ f (a)¦

=

(def)

¦g(a)¦, ∀z ∈ G.

According to the Maximum Modulus Theorem 3.11 p. 79, g is constant and thus f is constant. Thus, we

conclude: Either f (a) = 0 or f is constant.

Exercise 7. Give an elementary proof of the Maximum Modulus Theorem for polynomials.

Solution. Not available.

Exercise 8. Let G be a region and let f and g be analytic functions on G such that f (z)g(z) = 0 for all z in

G. Show that either f ≡ 0 or g ≡ 0.

Solution. It is easy to verify the following claim:

If f (z) =constant in B(a; R) ⊂ G, then f (z) =same constant for all z ∈ G.

Proof: Suppose f (z) = c in B(a; R). Deﬁne g(z) = f (z) − c, then g(z) ∈ A(G). g

(n)

(a) = 0 for all

n = 0, 1, 2, . . .. So g(z) = f (z) − c ≡ 0 on G.

Let G be a region and let f , g ∈ A(G) such that

f (z)g(z) = 0, ∀z ∈ G.

Assume f (z) 0, ∀z ∈ G. Let a ∈ G such that f (a) 0, then (since f is continuous, because f ∈ A(G))

there is an R > 0 such that f (z) 0 whenever z ∈ B(a; R). Hence g(z) = 0 ∀z ∈ B(a; R) to fulﬁll f (z)g(z) = 0

∀z ∈ B(a; R). By the claim above we obtain that

g(z) = 0, ∀z ∈ G.

So g ≡ 0.

Similar, assume g(z) 0 ∀z ∈ G, then by the same reasoning, we obtain f ≡ 0. Thus, we conclude: Either

f ≡ 0 or g ≡ 0.

Exercise 9. Let U : C → R be a harmonic function such that U(z) ≥ 0 for all z in C; prove that U is

constant.

Solution. There exists an analytic function f : C → C such that

U(z) = Re( f (z)),

since U : C → R is assumed to be harmonic. Hence, Re( f (z)) ≥ 0 ∀z ∈ C by assumption. For all z ∈ C,

deﬁne g(z) = e

−f (z)

∈ A(C). Then

¦g(z)¦ =

(def)

¦e

−f (z)

¦ = e

Re(−f (z))

= e

−Re( f (z))

≤ 1, ∀z ∈ C,

where the last step follows since Re( f (z)) ≥ 0 by assumption. Thus, g(z) is bounded and an entire function,

and therefore g is constant according to Liouville’s Theorem 3.4 p. 77 and hence f (z) is constant, too. It

follows that U is constant.

57

Exercise 10. Show that if f and g are analytic functions on a region G such that

¯

f g is analytic then either

f is constant or g ≡ 0.

Solution. Not available.

4.4 The index of a closed curve

Exercise 1. Prove Proposition 4.3.

Solution. Not available.

Exercise 2. Give an example of a closed rectiﬁable curve γ in C such that for any integer k there is a point

a ¦γ¦ with n(γ; a) = k.

Solution. Not available.

Exercise 3. Let p(z) be a polynomial of degree n and let R > 0 be suﬃciently large so that p never vanishes

in ¦z : ¦z¦ ≥ R¦. If γ(t) = Re

it

, 0 ≤ t ≤ 2π, show that

_

γ

p

·

(z)

p(z)

dz = 2πin.

Solution. Since p(z) is a polynomial of degree n, we can write

p(z) = c

n

_

k=1

(z − a

k

)

for some constant c (see Corollary 3.6 p. 77) where a

1

, . . . , a

n

are the zeros of p(z). By the product rule we

obtain

p

·

(z) = c

n

k=1

(z − a

k

)

·

n

_

l=1

lk

(z − a

l

) = c

n

k=1

n

_

l=1

lk

(z − a

l

).

Thus

_

γ

p

·

(z)

p(z)

dz =

_

γ

c

_

n

k=1

n

l=1

lk

(z − a

l

)

c

n

k=1

(z − a

k

)

dz

=

_

γ

_

1

z − a

1

+

1

z − a

2

+ . . . +

1

z − a

n

_

dz

=

_

γ

(z − a

1

)

−1

dz +

_

γ

(z − a

2

)

−1

dz + . . . +

_

γ

(z − a

n

)

−1

dz

= n(γ, a

1

)2πi + n(γ, a

2

)2πi + . . . + n(γ, a

n

)2πi =

n

k=1

2πi = 2πin,

since by Deﬁnition 4.2 p. 81 γ(t) = Re

it

, 0 ≤ t ≤ 2π or γ(t) = Re

2πit

, 0 ≤ t ≤ 1 is a closed and rectiﬁable

curve and a

1

, . . . , a

n

¦γ¦ since R > 0 suﬃciently large so that p never vanishes in ¦z : ¦z¦ ≥ R¦. In addition,

we have n(γ; a

i

) = 1 ∀1 ≤ i ≤ n. Hence,

_

γ

p

·

(z)

p(z)

dz = 2πin.

Exercise 4. Fix w = re

iθ

0 and let γ be a rectiﬁable path in C − ¦0¦ from 1 to w. Show that there is an

integer k such that

_

γ

z

−1

dz = log r + iθ + 2πik.

Solution. Not available.

58

4.5 Cauchy’s Theorem and Integral Formula

Exercise 1. Suppose f : G → C is analytic and deﬁne ϕ : G × G → C by ϕ(z, w) = [ f (z) − f (w)](z − w)

−1

if z w and ϕ(z, z) = f

·

(z). Prove that ϕ is continuous and for each ﬁxed w, z → ϕ(z, w) is analytic.

Solution. Not available.

Exercise 2. Give the details of the proof of Theorem 5.6.

Solution. Not available.

Exercise 3. Let B

±

=

¯

B(±1;

1

2

), G = B(0; 3)−(B

+

∪B

−

). Let γ

1

, γ

2

, γ

3

be curves whose traces are ¦z−1¦ = 1,

¦z+1¦ = 1, and ¦z¦ = 2, respectively. Give γ

1

, γ

2

, and γ

3

orientations such that n(γ

1

; w)+n(γ

2

; w)+n(γ

3

; w) =

0 for all w in C − G.

Solution. Not available.

Exercise 4. Show that the Integral Formula follows from Cauchy’s Theorem.

Solution. Assume the conditions of Cauchy’s Theorem, that is: Let G be an open subset of the plane and

f : G → C an analytic function. If γ is a closed rectiﬁable curve in G such that n(γ; w) = 0 ∀w ∈ G − C,

then

_

γ

f = 0.

Now, let a ∈ G, where a ¦γ¦. Deﬁne

g(z) =

_

¸

¸

_

¸

¸

_

f (z)−f (a)

z−a

, z a

f

·

(a), z = a

.

Clearly, g ∈ A(G). (Check the power series of f (z) like in the proof of Theorem 3.7 c) ⇒ b) p. 78) By

Cauchy’s Theorem, we obtain

_

γ

g = 0 ⇒

aγ

_

γ

f (z) − f (a)

z − a

dz = 0

⇒

_

γ

f (z)

z − a

dz −

_

γ

f (a)

z − a

dz = 0

⇒

_

γ

f (z)

z − a

dz = f (a)

_

γ

1

z − a

dz

⇒

Def. 4.2 p.81

_

γ

f (z)

z − a

dz = f (a)n(γ; a)2πi

⇒

1

2πi

_

γ

f (z)

z − a

dz = n(γ; a) f (a).

This is exactly Cauchy’s Integral Formula.

Exercise 5. Let γ be a closed rectiﬁable curve in C and a ¦γ¦. Show that for n ≥ 2

_

γ

(z − a)

−n

dz = 0.

Solution. Since γ is a closed rectiﬁable curve in C, ¦γ¦ is bounded. So there exists R > 0 such that

¦γ¦ ⊂ B(a; R). Hence C\B(a; R) belongs to the unbounded component of B(a; R). Therefore n(γ; w) = 0

∀w ∈ C\B(a; R) by Theorem 4.4 p.82. We can apply Corollary 5.9 p.86 if we let G = B(a; R) the open set

and f : G → C given by f (z) = 1 is analytic. Further γ is a closed and rectiﬁable curve in G such that

n(γ; w) = 0, ∀w ∈ C\G,

59

then for a in B(a; R)\¦γ¦

f

(n−1)

(a)n(γ; a) =

(n − 1)!

2πi

_

γ

1

(z − a)

n

dz

⇒ 0 n(γ; a) =

(n − 1)!

2πi

_

γ

1

(z − a)

n

dz

since f

(n−1)

(a) = 0 if n ≥ 2. Hence

_

γ

(z − a)

−n

dz = 0, for n ≥ 2.

Exercise 6. Let f be analytic on D = B(0; 1) and suppose ¦ f (z)¦ ≤ 1 for ¦z¦ < 1. Show ¦ f

·

(0)¦ ≤ 1.

Solution. We can directly apply Cauchy’s estimate 2.14 p. 73 with a = 0, R = 1, M = 1 and N = 1, that is

by assumption f is analytic on B(0; 1) and ¦ f (z)¦ ≤ 1 for all z in B(0; 1) (since ¦z¦ < 1). Then

¦ f

·

(0)¦ ≤

1! 1

1

1

= 1.

So ¦ f

·

(0)¦ ≤ 1.

Exercise 7. Let γ(t) = 1 + e

it

for 0 ≤ t ≤ 2π. Find

_

γ

_

z

z−1

_

n

dz for all positive integers n.

Solution. Let G = B(1; R) (open) where 1 < R < ∞ and let f : G → C be given by f (z) = z

n

which

is analytic on G. By Theorem 4.4 p. 82, n(γ; w) = 0 ∀w ∈ C\G since C\G belongs to the unbounded

component of G and γ is a closed and rectiﬁable curve in C.

Now, we can apply Corollary 5.9 with G and f deﬁned above such that n(γ; w) = 0 ∀w ∈ C\G and γ is a

closed rectiﬁable curve in G, then for a = 1 in G\¦γ¦

f

(n−1)

(1)n(γ; a) =

(n − 1)!

2πi

_

γ

z

n

(z − 1)

n

dz =

(n − 1)!

2πi

_

γ

_

z

z − 1

_

n

dz

⇒ n! =

(n − 1)!

2πi

_

γ

_

z

z − 1

_

n

dz

⇒

_

γ

_

z

z − 1

_

n

dz = 2πin,

since f

(n−1)

(1) = n! and n(γ; a) = 1.

Exercise 8. Let G be a region and suppose f

n

: G → C is analytic for each n ≥ 1. Suppose that ¦ f

n

¦

converges uniformly to a function f : G → C Show that f is analytic.

Solution. Let G be a region and suppose f

n

: G → C is analytic for each n ≥ 1, that us f

n

is continuous for

each n ≥ 1.

Let T be an arbitrary triangular path in G. Then T is closed and rectiﬁable. We also have, that n(T; w) = 0

∀w ∈ C\G, since w is an element of the unbounded component of G (Theorem 4.4 p. 82).

By Cauchy’s Theorem (First Version), p. 85 with m = 1, we have

_

T

f

n

= 0, ∀n ∈ N, (4.4)

60

since we suppose f

n

: G → C is analytic for each n ≥ 1 and T is closed and rectiﬁable. We also have

_

T

f = lim

n→∞

_

T

f

n

=

(4.4)

lim

n→∞

0 = 0

The ﬁrst equality follows by Lemma 2.7 p. 71 since T is closed and rectiﬁable and since each ¦ f

n

¦ ∈ A(G),

we have f

n

is continuous for each f

n

and by the uniform convergence to a function f , we get f is continuous

(Theorem 6.1 p. 29).

Finally, we can apply Morera’s Theorem 5.10 p. 86 to obtain that f is analytic in G, since f is continuous

and

_

γ

f = 0 for every triangular path T in G.

Exercise 9. Show that if f : C → C is a continuous function such that f is analytic oﬀ [−1, 1] then f is an

entire function.

Solution. By Morera’s Theorem, it suﬃce to show that for every triangular path T in C, we have

_

T

f = 0,

since we already assume f is continuous. There are several cases to consider:

Case 1: A triangle does not intersect I. In this case, we obtain by Cauchy’s Theorem, that

_

T

f = 0,

since we can ﬁnd an open neighborhood G containing T (is closed and rectiﬁable) such that f ∈ A(G) (by

assumption) and n(T; w) = 0 ∀w ∈ C\G (by Theorem 4.4 p. 82).

Case 2: A triangle touches I exactly at one point P. This single point of intersection P is a removable

singularity, since f is continuous. Again apply Cauchy’s Theoremto obtain

_

T

f = 0. (Procedure: Translate

the triangle by ±i depending if it lies above or below the x-axis. By Cauchy’s Theorem

_

T

f = 0 over the

translated triangle and let → 0 since f is continuous).

Case 3: One edge of the triangle touches I. Like in case 2 we can translate the triangle by ± We can

also argue this way: Let G be an open neighborhood containing the triangle T. Let ¦T

n

¦ be a sequence of

triangles that are not intersecting I, but whose limit is the given T. Then by the continuity of f , we get

_

T

f = lim

n→∞

_

T

n

f = lim

n→∞

0 = 0,

where the ﬁrst equality follows by Lemma 2.7 and the second one by case 1.

Case 4: A triangle T is cut into 2 parts by I. Then, we can always decompose the triangle T in three parts.

Two triangles are of the kind explained in case 3 and one is of the kind explained in case 2 (a sketch might

help). Thus,

_

T

f = 0 again.

Case 5: A triangle T contains parts of I (also here a sketch might help). In this instance, we can decompose

the triangle into 5 parts. Two triangles are of the kind explained in case 3 and the other three triangles are

of the kind explained in case 2. Hence,

_

T

f = 0.

Summary: Since

_

T

f = 0 for every triangular path in C and f is continuous, we get by Morera’s Theorem:

f ∈ A(C), that is f is an entire function.

Exercise 10. Use Cauchy’s Integral Formula to prove the Cayley–Hamilton Theorem: If A is an n × n

matrix over C and f (z) = det(z − A) is the characteristic polynomial of A then f (A) = 0. (This exercise was

taken from a paper by C. A. McCarthy, Amer. Math. Monthly, 82 (1975), 390–391).

Solution. Not available.

4.6 The homotopic version of Cauchy’s Theorem and simple connec-

tivity

Exercise 1. Let G be a region and let σ

1

, σ

2

: [0, 1] → G be the constant curves σ

1

(t) ≡ a, σ

2

(t) ≡ b.

Show that if γ is closed rectiﬁable curve in G and γ ∼ σ

1

, then γ ∼ σ

2

. (Hint: connect a and b by a curve.)

61

Solution. Assume γ is a closed and rectiﬁable curve in G and γ ∼ σ

1

. Since σ

1

(t) ≡ a and σ

2

(t) ≡ b, we

surely have that σ

1

and σ

2

are closed and rectiﬁable curves in G. If we can show that σ

1

∼ σ

2

, then we get

γ ∼ σ

2

, since ∼ is an equivalence relation, that is:

γ ∼ σ

1

and σ

1

∼ σ

2

⇒ γ ∼ σ

2

.

So, now we show σ

1

∼ σ

2

, that is, we have to ﬁnd a continuous function Γ : [0, 1] × [0, 1] →G such that

Γ(s, 0) = σ

1

(s) = a and Γ(s, 1) = σ

2

(s) = b, for 0 ≤ s ≤ 1

and Γ(0, t) = Γ(1, t) for 0 ≤ t ≤ 1. Let Γ(s, t) = tb + (1 − t)a. Clearly Γ is a continuous function. (it is

constant with respect to s and a line with respect to t). In addition, it satisﬁes Γ(s, 0) = a ∀0 ≤ s ≤ 1 and

Γ(s, 1) = b ∀0 ≤ s ≤ 1 and Γ(0, t) = tb + (1 − t)a = Γ(1, t) ∀0 ≤ t ≤ 1. So σ

1

∼ σ

2

.

Exercise 2. Show that if we remove the requirement “Γ(0, t) = Γ(1, t) for all t” from Deﬁnition6.1 then the

curve γ

0

(t) = e

2πit

, 0 ≤ t ≤ 1, is homotopic to the constant curve γ

1

(t) ≡ 1 in the region G = C − ¦0¦.

Solution. Not available.

Exercise 3. 3. Let ( = all rectiﬁable curves in G joining a to b and show that Deﬁnition 6.11 gives an

equivalence relation on (.

Solution. Not available.

Exercise 4. Let G = C − ¦0¦ and show that every closed curve in G is homotopic to a closed curve whose

trace is contained in ¦z : ¦z¦ = 1¦.

Solution. Not available.

Exercise 5. Evaluate the integral

_

γ

dz

z

2

+1

where γ(θ) = 2¦ cos 2θ¦e

iθ

for 0 ≤ θ ≤ 2π.

Solution. A sketch shows that the two zeros of z

2

+ 1 (they are ±i) are inside of the closed and rectiﬁable

curve γ (The region looks like a clover with four leaves ). Using partial fraction decompositions gives

_

γ

dz

z

2

+ 1

=

1

2i

_

γ

1

z − i

dz −

1

2i

_

γ

1

z + i

dz

= π

1

2πi

_

γ

1

z − i

dz − π

1

2i

_

γ

1

z + i

dz

=

Def 4.2 p. 81

π (n(γ; i) − n(γ; −i))

= π 0 = 0,

since n(γ; i) = n(γ; −i) since i and −i are contained in the region generated by γ. Hence

_

γ

dz

z

2

+ 1

= 0.

Exercise 6. Let γ(θ) = θe

iθ

for 0 ≤ θ ≤ 2π and γ(θ) = 4π − θ for 2π ≤ θ ≤ 4π. Evaluate

_

γ

dz

z

2

+π

2

.

Solution. A sketch reveals that the zero −πi is inside the region and the zero πi is outside the region. Using

partial fraction decomposition yields

_

γ

dz

z

2

+ π

2

=

1

2πi

_

γ

1

z − iπ

dz −

1

2πi

_

γ

1

z + iπ

dz

=

Def 4.2 p. 81

n(γ; iπ) − n(γ; −iπ) = 0 − 1 = −1,

62

since iπ is not contained in the region generated by γ, so n(γ; iπ) = 0 and −iπ is contained in the region, so

n(γ; −iπ) = 1. Hence,

_

γ

dz

z

2

+ π

2

= −1.

Exercise 7. Let f (z) = [(z−

1

2

−i) (z−1−

3

2

i) (z−1−

i

2

) (z−

3

2

−i)]

−1

and let γ be the polygon [0, 2, 2+2i, 2i, 0].

Find

_

γ

f .

Solution. Not available.

Exercise 8. Let G = C − ¦a, b¦, a b, and let γ be the curve in the ﬁgure below.

(a) Show that n(γ; a) = n(γ; b) = 0.

(b) Convince yourself that γ is not homotopic to zero. (Notice that the word is “convince” and not “prove”.

Can you prove it?) Notice that this example shows that it is possible to have a closed curve γ in a region

such that n(γ; z) = 0 for all z not in G without γ being homotopic to zero. That is, the converse to Corollary

6.10 is false.

Solution. Let γ be the path depicted on p. 96. We can write it as a sum of 6 paths. Two of them will

be closed and have a and b in their unbounded component. Therefore, two integrals will be zero and we

will have another 2 pair of non-closed paths. The ﬁrst pair begins at the leftmost crossing pair and goes

around a in opposite direction. The second pair begins at the middle crossing pair and goes around b in

opposite direction. They also meet at the rightmost crossing point. If we integrate over the path around a

is equivalent to evaluate

_

γ

1

−γ

2

1

z−a

dz where γ

1

(t) = a + re

it

, 0 ≤ t ≤ π and γ

2

(t) = a + re

it

, π ≤ t ≤ 2π for

some r > 0. It is easily seen that we have

_

γ

1

1

z − a

dz =

_

γ

2

1

z − a

dz = πi

and therefore n(γ; a) = 0 and similarly we obtain n(γ; b) = 0.

Exercise 9. Let G be a region and let γ

0

and γ

1

be two closed smooth curves in G. Suppose γ

0

∼ γ

1

, and Γ

satisﬁes (6.2). Also suppose that γ

t

(s) = Γ(s, t) is smooth for each t. If w ∈ C−G deﬁne h(t) = n(γ

t

; w) and

show that h : [0, 1] → Z is continuous.

Solution. Not available.

Exercise 10. Find all possible values of

_

γ

dz

1+z

2

, where γ is any closed rectiﬁable curve in C not passing

through ±i.

Solution. Let γ be a closed rectiﬁable curve in C not passing through ±i. Using partial fraction decompo-

sition and the deﬁnition of the winding number we obtain

_

γ

dz

1 + z

2

=

1

2i

__

γ

1

z − 1

dz −

_

γ

1

z + i

dz

_

=

1

2i

(2πin(γ; i) − 2πin(γ; −i)) = π (n(γ; i) − n(γ; −i)) .

Exercise 11. Evaluate

_

γ

e

z

−e

−z

z

4

dz where γ is one of the curves depicted below. (Justify your answer.)

Solution. Using Corollary 5.9 p. 86 we have ∀a ∈ G − ¦γ¦

_

γ

f (z)

(z − a)

k+1

dz = 2πi f

(k)

(a)n(γ; a)

1

k!

,

63

where γ is a closed rectiﬁable curve in G (open) such that n(γ; w) = 0 for all w ∈ C − G and f : G → C is

analytic.

In our case a = 0 and γ satisﬁes the above assumptions for exercise a), b) and c). Let f (z) = e

z

−e

−z

which

is clearly analytic on any region G. We have f

(3)

(0) = e

0

+ e

−0

= 2 and using the formula above for k = 3

we obtain

_

γ

e

z

− e

−z

z

4

dz = 2πi 2n(γ; 0)

1

6

=

2πi

3

n(γ; 0).

Hence, n(γ; 0) = 1 for part a) and thus the result is

2πi

3

. We have n(γ; 0) = 1 for part b) and thus the result

is

4πi

3

. We obtain the same result for part c).

4.7 Counting zeros; the Open Mapping Theorem

Exercise 1. Show that if f : G → C is analytic and γ is a rectiﬁable curve in G then f ◦ γ is also a

rectiﬁable curve.

Solution. Not available.

Exercise 2. Let G be open and suppose that γ is a closed rectiﬁable curve in G such that γ ≈ 0. Set

r = d(¦γ¦, ∂G) and H = ¦z ∈ C : n(γ; z) = 0¦. (a) Show that ¦z : d(z, ∂G) <

1

2

r¦ ⊂ H. (b) Use part (a)

to show that if f : G → C is analytic then f (z) = α has at most a ﬁnite number of solutions z such that

n(γ; z) 0.

Solution. Not available.

Exercise 3. Let f be analytic in B(a; R) and suppose that f (a) = 0. Show that a is a zero of multiplicity m

iﬀ f

(m−1)

(a) = . . . = f (a) = 0 and f

(a)

(a) 0.

Solution. ⇐: Let f

(n)

(a) = 0 for all n = 1, 2, . . . , m − 1 (we also have f (a) = 0) and f

(m)

(a) 0. Since

f ∈ A(B(a; R)), we can write

f (z) =

∞

k=0

a

k

(z − a)

k

where a

k

=

f

(k)

(a)

k!

and ¦z − a¦ < R

=

∞

k=0

f

(k)

(a)

k!

(z − a)

k

=

∞

k=m

f

(k)

(a)

k!

(z − a)

k

= (z − a)

m

∞

k=0

f

(k)

(a)

k!

(z − a)

k−m

.,,.

g(z)

=

g(a)0, since f

(m)

0

(z − a)

m

g(z)

where g ∈ A(B(a; R)). Hence, a is a zero of multiplicity m.

⇒: Suppose a is a zero of multiplicity m ≥ 1 of f . By deﬁnition, there exists g ∈ A(B(0; R)) such that

f (z) = (z − a)

m

g(z), g(a) 0 (p. 97).

Then after a lengthy calculation, we obtain

f

(n)

(z) =

n

k=0

_

n

k

_

m!

(m − (n − k))!

(z − a)

m−(n−k)

g

(k)

(z), n = 1, 2, . . . , m.

64

If z = a, then clearly f

(n)

(a) = 0 for m−(n −k) > 0 and f

(n)

(a) 0 for m−(n −k) = 0. If k = 0 and m = n,

then f

(m)

(a) = f

(n)

(a) = m!g(a) 0 since g(a) 0 by assumption. Hence, f

(n)

(a) = 0 for n = 1, 2, . . . , m−1

and f

(n)

(a) 0 for n = m.

Exercise 4. Suppose that f : G → C is analytic and one-one; show that f

·

(z) 0 for any z in G.

Solution. Assume f

·

(a) = 0 for some a ∈ G. Then f (a) = α where α is a constant. Deﬁne

F(z) = f (z) − α.

Then F(a) = f (a) − α = α − α = 0 and F

·

(a) = f

·

(a) = 0 by assumption. Hence, F has a as a root of

multiplicity m ≥ 2.

Then, we can use Theorem 7.4 p. 98 to argue that there is an > 0 and δ > 0 such that for 0 < ¦ξ − a¦ < δ,

the equation f (z) = ξ has exactly m simple roots in B(a; ), since f is analytic in B(a; R) and α = f (a). Also

f (z) − α has a zero of order m ≥ 2 at z = a.

Since f (z) = ξ has exactly m (our case m ≥ 2) simple roots in B(a; ), we can ﬁnd at least two distinct

points z

1

, z

2

∈ B(a; ) ⊂ G such that

f (z

1

) = ξ = f (z

2

)

contradicting that f is 1 − 1. So f

·

(z) 0 for any z in G.

Exercise 5. Let X and Ω be metric spaces and suppose that f : X → Ω, is one-one and onto. Show that f

is an open map iﬀ f is a closed map. (A function f is a closed map if it takes closed sets onto closed sets.)

Solution. Not available.

Exercise 6. Let P : C → R be deﬁned by P(z) = Re z; show that P is an open map but is not a closed map.

(Hint: Consider the set F = ¦z : Im z = (Re z)

−1

and Re z 0¦.)

Solution. Not available.

Exercise 7. Use Theorem 7.2 to give another proof of the Fundamental Theorem of Algebra.

Solution. Not available.

4.8 Goursat’s Theorem

No exercises are assigned in this section.

65

Chapter 5

Singularities

5.1 Classiﬁcation of singularities

Exercise 1. Each of the following functions f has an isolated singularity at z = 0. Determine its nature; if

it is a removable singularity deﬁne f (0) so that f is analytic at z = 0; if it is a pole ﬁnd the singular part; if

it is an essential singularity determine f (¦z : 0 < ¦z¦ < δ¦) for arbitrarily small values of δ.

a)

f (z) =

sin z

z

;

b)

f (z) =

cos z

z

;

c)

f (z) =

cos z − 1

z

;

d)

f (z) = exp(z

−1

);

e)

f (z) =

log(z + 1)

z

2

;

f)

f (z) =

cos(z

−1

)

z

−1

;

g)

f (z) =

z

2

+ 1

z(z − 1)

;

h)

f (z) = (1 − e

z

)

−1

;

i)

f (z) = z sin

1

z

;

66

j)

f (z) = z

n

sin

1

z

.

Solution. a) According to Theorem 1.2 p.103, z = 0 is a removable singularity, since

lim

z→∞

(z − 0)

sin(z)

z

= lim

z→∞

z

sin(z)

z

= lim

z→∞

sin(z) = 0.

If we deﬁne,

f (z) =

_

¸

¸

_

¸

¸

_

sin(z)

z

, z 0

1, z = 0

,

then f is analytic at z = 0. A simple computation using L’Hospital’s rule yields lim

z→0

sin(z)

z

= 1.

b) We have (see p. 38)

cos(z) =

∞

k=0

(−1)

k

z

2k

(2k)!

and therefore

cos(z)

z

=

∞

k=0

(−1)

k

z

2k−1

(2k)!

=

1

z

+ analytic part.

Thus

cos z

z

has a simple pole at z = 0.

c) According to Theorem 1.2 p.103, z = 0 is a removable singularity, since

lim

z→∞

(z − 0)

cos(z) − 1

z

= lim

z→∞

z

cos(z) − 1

z

= lim

z→∞

cos(z) − 1 = 1 − 1 = 0.

If we deﬁne,

f (z) =

_

¸

¸

_

¸

¸

_

cos(z)−1

z

, z 0

1, z = 0

,

then f is analytic at z = 0. A simple computation using L’Hospital’s rule yields lim

z→0

cos(z)−1

z

= 0.

e) We know

log(1 + z) =

∞

k=1

(−1)

k+1

z

k

k

.

So we can write

log(1 + z)

z

2

=

1

z

2

∞

k=1

(−1)

k+1

z

k

k

=

∞

k=1

(−1)

k+1

z

k−2

k

=

1

z

−

1

2

+

∞

k=3

(−1)

k+1

z

k−2

k

and hence f (z) =

log(1+z)

z

2

has a pole of order 1 at z = 0 and the singular part is

1

z

(by Equation 1.7 and

Deﬁnition 1.8 p. 105).

g) First, we simplify

z

2

+1

z

2

−z

= 1 +

z+1

z(z−1)

. Now partial fraction decomposition of

z+1

z(z−1)

yields

z + 1

z(z − 1)

= −

1

z

+

2

z − 1

so

f (z) = 1 −

1

z

+

2

z − 1

67

and hence f (z) =

z

2

+1

z

2

−z

has poles of order 1 at z = 0 and z = 1 and the singular part is

2

z−1

−

1

z

(by Equation

1.7 and Deﬁnition 1.8 p. 105).

i) We know (see p. 38)

sin(z) =

∞

k=0

(−1)

k

z

2k+1

(2k + 1)!

so

f (z) = z sin

1

z

= z

∞

k=0

(−1)

k

_

1

z

_

2k+1

(2k + 1)!

=

∞

k=0

(−1)

k

_

1

z

_

2k

(2k + 1)!

= 1 −

1

3!z

2

+

1

5!z

4

−

1

7!z

6

+ . . .

Therefore f (z) = z sin

1

z

has an essential singularity at z = 0 (it is neither a pole nor a removable singularity

by Corollary 1.18c). We have

f (¦z : 0 < ¦z¦ < δ¦) = f [ann(0; 0, δ)] = C

by the Casorati-Weierstrass Theorem p. 109. But f (¦z : 0 < ¦z¦ < δ¦) = f [ann(0; 0, δ)] is either C or C\¦a¦

for one a ∈ C by the Great Picard Theorem (p. 300). In this case,

f (¦z : 0 < ¦z¦ < δ¦) = f [ann(0; 0, δ)] = C.

Exercise 2. Give the partial fraction expansion of r(z) =

z

2

+1

(z

2

+z+1)(z−1)

2

.

Solution. Not available.

Exercise 3. Give the details of the derivation of (1.17) from (1.16).

Solution. Not available.

Exercise 4. Let f (z) =

1

z(z−1)(z−2)

; give the Laurent Expansion of f (z) in each of the following annuli: (a)

ann(0; 0, 1); (b) ann(0; 1, 2); (c) ann(0; 2, ∞).

Solution. a) We have 0 < ¦z¦ < 1 and using partial fraction decomposition yields

f (z) =

1

z(z − 1)(z − 2)

=

1

z

1

(z − 1)(z − 2)

=

1

z

_

−1

z − 1

+

1

z − 2

_

=

1

z

_

1

1 − z

−

1

2

1

1 −

z

2

_

=

¦z¦<1and ¦z/2¦<1

1

z

_

¸

¸

¸

¸

¸

_

∞

n=0

z

n

−

1

2

∞

n=0

_

z

2

_

n

_

¸

¸

¸

¸

¸

_

=

∞

n=0

z

n−1

−

∞

n=0

_

1

2

_

n+1

z

n−1

=

∞

n=−1

_

¸

¸

¸

¸

¸

_

1 −

_

1

2

_

n+2

_

¸

¸

¸

¸

¸

_

z

n

=

∞

n=−1

_

1 − 2

−n−2

_

z

n

, (0 < ¦z¦ < 1).

b) We have 1 < ¦z¦ < 2 and thus

f (z) =

1

z(z − 1)(z − 2)

=

1

z

_

¸

¸

¸

¸

¸

_

−

1

z

1

1 −

1

z

−

1

2

1

1 −

z

2

_

¸

¸

¸

¸

¸

_

=

¦1/z¦<1and ¦z/2¦<1

1

z

_

¸

¸

¸

¸

¸

_

−

1

z

∞

n=0

_

1

z

_

n

−

1

2

∞

n=0

_

z

2

_

n

_

¸

¸

¸

¸

¸

_

=

∞

n=0

−

_

1

z

_

n+2

−

∞

n=0

_

1

2

_

n+1

z

n−1

=

∞

n=0

−z

−n−2

−

∞

n=0

_

1

2

_

n+1

z

n−1

=

−2

n=−∞

−z

n

−

∞

n=−1

_

1

2

_

n+2

z

n

, (1 < ¦z¦ < 2).

68

c) We have 2 < ¦z¦ < ∞and thus

f (z) =

1

z(z − 1)(z − 2)

=

1

z

_

¸

¸

¸

¸

¸

_

−

1

z

1

1 −

1

z

+

1

z

1

1 −

2

z

_

¸

¸

¸

¸

¸

_

=

¦1/z¦<1and ¦2/z¦<1

1

z

_

¸

¸

¸

¸

¸

_

−

1

z

∞

n=0

_

1

z

_

n

+

1

z

∞

n=0

_

2

z

_

n

_

¸

¸

¸

¸

¸

_

=

∞

n=0

−

_

1

z

_

n+2

+

∞

n=0

2

n

_

1

z

_

n+2

=

∞

n=0

(2

n

− 1)z

−n−2

=

−2

n=−∞

(2

−n−2

− 1)z

n

, (2 < ¦z¦ < ∞).

Exercise 5. Show that f (z) = tan z is analytic in C except for simple poles at z =

π

2

+ nπ, for each integer

n. Determine the singular part of f at each of these poles.

Solution. Deﬁne z

n

=

π

2

+ nπ. We know tan(z) =

sin(z)

cos(z)

and (cos(z))

·

= −sin(z). Hence, tan(z) has a simple

pole at each z

n

with residue −1 and with singular part −

1

z−z

n

at each z

n

.

Exercise 6. If f : G → C is analytic except for poles show that the poles of f cannot have a limit point in

G.

Solution. Not available.

Exercise 7. Let f have an isolated singularity at z = a and suppose f (z) 0. Show that if either (1.19) or

(1.20) holds for some s in R then there is an integer m such that (1.19) holds if s > m and (1.20) holds if

s < m.

Solution. Let lim

z→a

¦z − a¦

s

¦ f (z)¦ = 0 hold, that is lim

z→a

(z − a)

s

f (z) = 0 for some s in R. Then it holds

obviously for any greater s’s. Hence, there exist an integer n > 0 such that

lim

z→a

(z − a)

n

f (z) = 0 (5.1)

and

lim

z→a

(z − a)

n+1

f (z) = 0 (5.2)

(In fact we are free to choose any positive integer greater than s). Now, (z − a)

n

f (z) thus has a removable

singularity at z = a (by Theorem 1.2 p. 103) applied to (5.2)) and its extended value is also zero by (5.2).

Thus (z − a)

n

f (z) has a zero of ﬁnite order k at z − a. So

(z − a)

n

f (z) = (z − a)

k

h(z)

where h(z) is analytic at a and h(a) 0. Hence

lim

z→a

(z − a)

s

f (z) = lim

z→a

(z − a)

s−n

(z − a)

n

f (z) = lim

z→a

(z − a)

s−n+k

h(z) =

_

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

_

0, s > n − k

±∞, s < n − k

h(a) 0, s = n − k.

Thus,

lim

z→a

¦z − a¦

s

¦ f (z)¦ =

_

¸

¸

_

¸

¸

_

0, s > m

∞, s < m

69

where m = n − k.

Now let lim

z→a

¦z − a¦

s

¦ f (z)¦ = ∞for some s in R. Then there exists an integer n < 0 so that

lim

z→a

¦z − a¦

n

¦ f (z)¦ = ∞ (5.3)

(In fact we are free to choose any negative integer less than s). Now (z − a)

n

f (z) thus has a pole at z = a

(by Deﬁnition 1.3 p.105 and (5.1)) and (z − a)

n

f (z) has a pole of ﬁnite order l at z = a and so

(z − a)

n

f (z) = (z − a)

−l

k(z)

where k(z) is analytic at a and k(a) 0 (Proposition 1.4 p.105). Hence

lim

z→a

(z − a)

s

f (z) = lim

z→a

(z − a)

s−n+n

f (z) = lim

z→a

(z − a)

s−n−l

k(z) =

_

¸

¸

¸

¸

¸

_

¸

¸

¸

¸

¸

_

0, s > n + l

±∞, s < n + l

k(a) 0, s = n + l.

Thus,

lim

z→a

¦z − a¦

s

¦ f (z)¦ =

_

¸

¸

_

¸

¸

_

0, s > m

∞, s < m

where m = n + l.

Exercise 8. Let f , a, and m be as in Exercise 7. Show: (a) m = 0 iﬀ z = a is a removable singularity and

f (a) 0; (b) m < 0 iﬀ z = a is a removable singularity and f has a zero at z = a of order −m; (c) m > 0 iﬀ

r = a is a pole of f of order m.

Solution. a) ⇒: Let m = 0. Then by Exercise 7, we have

lim

z→a

(z − a)

2

f (z) =

_

¸

¸

_

¸

¸

_

0, s > 0 = m

0, s = 0 = m

. (5.4)

Thus using s = 1, f (z) has a removable singularity (by Theorem 1.2 p. 103) with the extended value equal

to f (a) = lim

z→a

f (z) 0.

⇐: Let z = a be a removable singularity and f (a) = lim

z→a

f (z) 0. Then lim

z→a

(z − a) f (z) = 0 by

Theorem 1.2 p.103. By Exercise 7, we obtain that there is an integer m such that

lim

z→a

¦z − a¦

s

¦ f (z)¦ =

_

¸

¸

_

¸

¸

_

0, s > m

∞, s < m

.

But m = 0, since

lim

z→a

(z − a)

s

f (z) =

_

¸

¸

_

¸

¸

_

0, s > 0

∞, s < 0

,

which follows by (5.4).

b) ⇒: Let the algebraic order be deﬁned at z = a and be equal to −m (m > 0). Then by Exercise 7, we have

lim

z→a

f (z) = 0 and lim

z→a

(z − a) f (z) = 0.

Recall lim

z→a

(z − a)

s

f (z) = 0 if s > −m so it is true of s = 0 and s = 1. Thus, f (z) has a removable

singularity at z = a and also the extended function has the value lim

z→a

f (z) = 0 at z = a (p. 103 Theorem

70

1.2). Now h(z) := (z − a)

−m

f (z) has a removable singularity at z = a (since lim

z→a

(z − a)

−m+1

f (z) = 0) but

h(z) has extended value

h(a) = lim

z→a

(z − a)

−m

f (z) 0

(by Exercise 7). Therefore f (z) = (z − a)

m

h(z) has a zero of order m at z = a.

⇐: Assume f (z) has a removable singularity at z = a and let z = a be a zero of order m > 0. Then

f (z) = (z − a)

m

h(z) where h(z) is analytic in ¦z − a¦ < δ, δ > 0 and h(a) 0. Then

lim

z→a

¦z − a¦

s

¦ f (z)¦ = lim

z→a

¦z − a¦

s+m

¦h(z)¦ =

_

¸

¸

_

¸

¸

_

0, s > −m

∞, s < −m

by Exercise 7. Hence, the algebraic order of f at z = a is equal to −m (m > 0).

c) ⇒: Let the algebraic order be m > 0. Then

lim

z→a

(z − a)

s

f (z) = 0

for s > m by Exercise 7, that is

lim

z→a

(z − a)

m+1

f (z) = 0 and lim

z→a

(z − a)

m

f (z) 0

by Exercise 7. Then for z a, we have h(z) = (z − a)

m

f (z) has a removable singularity at z = a with

h(a) 0. Now f (z) = (z − a)

−m

h(z) where h ∈ A(B(a; δ)) and h(a) 0. Thus f (z) has a pole of order m at

z = a.

⇐: Let f (z) have a pole of order m at z = a. Then f (z) = (z−a)

−m

g(z) where g(z) ∈ A(B(a; δ)) and g(a) 0.

Thus lim

z→a

(z − a)

m

f (z) = g(a) 0 and

lim

z→a

(z − a)

m+1

f (z) = lim

z→a

(z − a)g(z) = 0.

So

lim

z→a

¦z − a¦

s

¦ f (z)¦ = lim

z→a

¦z − a¦

s−m

¦z − a¦

m

¦ f (z)¦ =

_

¸

¸

_

¸

¸

_

0, s > m

∞, s < m

by Exercise 7. Therefore, m > 0 is the algebraic order of f at z = a.

Exercise 9. A function f has an essential singularity at z = a iﬀ neither (1.19) nor (1.20) holds for any real

number s.

Solution. ⇒: We prove this direction by proving the contrapositive (P → Q ⇐⇒ ¬Q ⇒ ¬P). Assume

(1.19) or (1.20) hold for some real number s. Then by Exercise 7, there exists an integer m such that (1.19)

holds if s > m and (1.20) holds if s < m. Then by Exercise 8, z = a is either a removable singularity or a

pole. Hence z = a is not an essential singularity (by Deﬁnition 1.3 p.105).

⇐: Also this direction by is being proved by showing the contrapositive. Assume that f has an isolated

singularity at z = a which is not an essential singularity. Then either z = a is a removable singularity or a

pole (by Deﬁnition 1.3 p.105). Again by Problem 7 and 8, in either case there exists an m such that (1.19)

or (1.20) holds. (If z = a is removable, then either m = 0 or m < 0; if z = a is a pole, then m > 0 by

Exercise 8).

Exercise 10. Suppose that f has an essential singularity at z = a. Prove the following strengthened version

of the Casorati–Weierstrass Theorem. If c ∈ C and > 0 are given then for each δ > 0 there is a number α,

¦c − α¦ < , such that f (z) = α has inﬁnitely many solutions in B(a; δ).

71

Solution. Not available.

Exercise 11. Give the Laurent series development of f (z) = exp

_

1

z

_

. Can you generalize this result?

Solution. Not available.

Exercise 12. (a) Let λ ∈ C and show that

exp

_

1

2

λ

_

z +

1

z

__

= a

0

+

∞

n=1

a

n

_

z

n

+

1

z

n

_

for 0 < ¦z¦ < ∞, where for n ≥ 0

a

n

=

1

π

_

π

0

e

λ cos t

cos nt dt.

(b) Similarly, show

exp

_

1

2

λ

_

z −

1

z

__

= b

0

+

∞

n=1

b

n

_

z

n

+

(−1)

n

z

n

_

for 0 < ¦z¦ < ∞, where

b

n

=

1

π

_

π

0

cos(nt − λ sin t) dt.

Solution. Not available.

Exercise 13. Let R > 0 and G = ¦z : ¦z¦ > R¦; a function f : G → C has a removable singularity, a pole, or

an essential singularity at inﬁnity if f (z

−1

) has, respectively, a removable singularity, a pole, or an essential

singularity at z = 0. If f has a pole at ∞then the order of the pole is the order of the pole of f (z

−1

) at z = 0.

(a) Prove that an entire function has a removable singularity at inﬁnity iﬀ it is a constant.

(b) Prove that an entire function has a pole at inﬁnity of order m iﬀ it is a polynomial of degree m.

(c) Characterize those rational functions which have a removable singularity at inﬁnity.

(d) Characterize those rational functions which have a pole of order m at inﬁnity.

Solution. Not available.

Exercise 14. Let G = ¦z : 0 < ¦z¦ < 1¦ and let f : G → C be analytic. Suppose that γ is a closed rectiﬁable

curve in G such that n(γ; a) = 0 for all a in C − G. What is

_

γ

f ? Why?

Solution. Not available.

Exercise 15. Let f be analytic in G = ¦z : 0 < ¦z − a¦ < r¦ except that there is a sequence of poles ¦a

n

¦ in G

with a

n

→ a. Show that for any ω in C there is a sequence ¦z

n

¦ in G with a = limz

n

and ω = lim f (z

n

).

Solution. Not available.

Exercise 16. Determine the regions in which the functions f (z) =

_

sin

1

z

_

−1

and g(z) =

_

1

0

(t − z)

−1

dt are

analytic. Do they have any isolated singularities? Do they have any singularities that are not isolated?

Solution. Not available.

Exercise 17. Let f be analytic in the region G =ann(a; 0, R). Show that if

G

¦ f (x +iy)¦

2

dx dy < ∞then f

has a removable singularity at z = a. Suppose that p > 0 and

G

¦ f (x + iy)¦

p

dx dy < ∞; what can be said

about the nature of the singularity at z = a?

Solution. Not available.

72

5.2 Residues

Exercise 1. Calculate the following integrals:

a)

_

∞

0

x

2

dx

x

4

+ x

2

+ 1

b)

_

∞

0

cos x − 1

x

2

dx

c)

_

π

0

cos 2θ dθ

1 − 2a cos θ + a

2

where a

2

< 1

d)

_

π

0

dθ

(a + cos θ)

2

where a > 1.

Solution. Note that

_

∞

0

x

2

dx

x

4

+ x

2

+ 1

=

1

2

_

∞

−∞

x

2

dx

x

4

+ x

2

+ 1

since

x

2

x

4

+x

2

+1

is even. Let γ

R

= [−R, R] ∪ C

R

where C

R

= Re

it

, 0 ≤ t ≤ π and R > 0 (draw a sketch).

f (z) =

z

2

z

4

+z

2

+1

has two simple poles at a

1

= e

i

π

3

, a

2

= e

i

2π

3

enclosed by γ

R

. By the residual Theorem, we have

2πi(Res( f , a

1

) + Res( f , a

2

)) =

_

γ

R

f =

_

R

−R

x

2

x

4

+ x

2

+ 1

dx +

_

C

R

z

2

z

4

+ z

2

+ 1

dz.

Thus,

_

∞

−∞

x

2

dx

x

4

+ x

2

+ 1

= − lim

R→∞

_

C

R

z

2

z

4

+ z

2

+ 1

dz + 2πi

2

k=1

Res( f , a

k

).

We have

_

C

R

f (z) dz =

_

C

R

z

2

z

4

+ z

2

+ 1

dz ≤

¸

¸

¸

¸

¸

¸

_

C

R

z

2

z

4

+ z

2

+ 1

dz

¸

¸

¸

¸

¸

¸

≤

_

C

R

¦z¦

2

¦z

4

+ z

2

+ 1¦

¦dz¦ ≤

_

C

R

¦z¦

2

¦z¦

4

− ¦z¦

2

− 1

¦dz¦

=

R

2

R

4

− R

2

− 1

_

C

R

¦dz¦ =

πR

3

R

4

− R

2

− 1

→ 0 as R → ∞.

Next, we calculate Res( f , a

1

).

Res( f , e

i

π

3

) = lim

z→e

i

π

3

(z − e

i

π

3

)

z

2

z

4

+ z

2

+ 1

= e

i

2π

3

lim

z→e

i

π

3

(z − e

i

π

3

)

1

z

4

+ z

2

+ 1

= e

i

2π

3

lim

z→e

i

π

3

1

4z

3

+ 2z

=

e

i

2π

3

4e

i

3π

3

+ 2e

i

π

3

=

e

i

π

3

4e

i

2π

3

+ 2

=

1

2

+

1

2

√

3i

4(−

1

2

+

1

2

√

3i) + 2

=

1

2

+

1

2

√

3i

2

√

3i

=

1

4

√

3i

+

1

4

=

1

4

−

1

4

√

3

i.

73

Finally, we calculate Res( f , a

2

).

Res( f , e

i

2π

3

) = lim

z→e

i

2π

3

(z − e

i

2π

3

)

z

2

z

4

+ z

2

+ 1

= e

i

4π

3

lim

z→e

i

2π

3

(z − e

i

2π

3

)

1

z

4

+ z

2

+ 1

= e

i

4π

3

lim

z→e

i

2π

3

1

4z

3

+ 2z

=

e

i

4π

3

4e

i

6π

3

+ 2e

i

2π

3

=

e

i

2π

3

4e

i

4π

3

+ 2

=

−

1

2

+

1

2

√

3i

4(−

1

2

−

1

2

√

3i) + 2

=

−

1

2

+

1

2

√

3i

−2

√

3i

=

1

4

√

3i

−

1

4

= −

1

4

−

1

4

√

3

i.

So

2πi

2

k=1

Res( f , a

k

) = 2πi

_

1

4

−

1

4

√

3

i +

1

4

√

3i

−

1

4

_

= 2πi

_

−

1

2

√

3

i

_

= −

π

√

3

i

2

=

π

√

3

.

Hence,

_

∞

0

x

2

dx

x

4

+ x

2

+ 1

=

1

2

π

√

3

.

b) Let f (z) =

e

iz

−1

z

2

and γ = [−R, −r] + γ

r

+ [r, R] + γ

R

where γ

R

= Re

it

, 0 ≤ t ≤ π and γ

r

= re

−it

, 0 ≤ t ≤ π

and 0 < r < R (A sketch might help to see that γ is a closed and rectiﬁable curve). Clearly

_

γ

f (z) dz = 0

since f is analytic on the region enclosed by γ. Also

_

γ

f (z) dz =

_

−r

−R

e

ix

− 1

x

2

dx +

_

R

r

e

ix

− 1

x

2

dx +

_

γ

r

f +

_

γ

R

f .

We have

_

−r

−R

e

ix

− 1

x

2

dx =

_

R

r

e

−ix

− 1

x

2

dx.

So

_

−r

−R

e

ix

− 1

x

2

dx +

_

R

r

e

ix

− 1

x

2

dx =

_

R

r

e

−ix

+ e

−ix

− 2

x

2

dx = 2

_

R

r

e

−ix

+e

−ix

2

− 1

x

2

dx

= 2

_

R

r

cos(x) − 1

x

2

dx

Therefore,

_

R

r

cos(x) − 1

x

2

dx = −

1

2

_

γ

R

e

iz

− 1

z

2

dz −

1

2

_

γ

r

e

iz

− 1

z

2

dz

which implies that

_

∞

0

cos(x) − 1

x

2

dx = lim

R→∞

_

−

1

2

_

γ

R

e

iz

− 1

z

2

dz

_

− lim

r→0

_

1

2

_

γ

r

e

iz

− 1

z

2

dz

_

The ﬁrst term on the right hand side is zero, since

¸

¸

¸

¸

¸

¸

_

γ

R

e

iz

− 1

z

2

dz

¸

¸

¸

¸

¸

¸

≤

_

γ

R

¦e

iz

− 1¦

¦z¦

2

¦dz¦ ≤

1

R

2

__

γ

R

¦dz¦ +

_

γ

R

¦e

iz

¦ ¦dz¦

_

≤

1

R

2

(πR + πR) =

2π

R

→ 0 as R → ∞.

74

The second term on the right hand side is π, since

lim

r→0

__

γ

r

e

iz

− 1

z

2

dz

_

= −

1

2

2πi Res( f (z), 0) = −πi

2

= π

( f (z) can be written as

i

z

+analytic part and therefore Res( f (z), 0) = i). Hence,

_

∞

0

cos(x) − 1

x

2

dx = −

1

2

0 −

1

2

π = −

π

2

.

c) If a = 0, then

_

π

0

cos 2θ dθ

1 − 2a cos θ + a

2

=

_

π

0

cos 2θ dθ =

1

2

[sin 2θ]

π

0

=

1

2

[0 − 0] = 0.

Assume a 0. We can write

_

π

0

cos 2θ dθ

1 − 2a cos θ + a

2

=

1

2

_

2π

0

cos 2θ dθ

1 − 2a cos θ + a

2

=

1

2

_

2π

0

e

2θi

+e

2θi

2

1 − 2a

e

θi

+e

θi

2

+ a

2

dθ.

Let z = e

iθ

. Then dz = iz dθ =⇒ dθ =

dz

iz

. Hence,

1

2

_

2π

0

e

2θi

+e

2θi

2

1 − 2a

e

θi

+e

θi

2

+ a

2

dθ =

1

2

_

¦z¦=1

z

2

+

1

z

2

2

1 − a

_

z +

1

z

_

+ a

2

dz

iz

=

1

4i

_

¦z¦=1

z

4

+ 1

z

2

_

z − az

2

− a + a

2

z

_ dz

= −

1

4ia

_

¦z¦=1

z

4

+ 1

z

2

_

z

2

−

1+a

2

a

z + 1

_ dz = −

1

4ia

_

¦z¦=1

z

4

+ 1

z

2

(z − a)

_

z −

1

a

_ dz

By the Residue Theorem and the fact that ¦a¦ < 1, since a

2

< 1

_

¦z¦=1

z

4

+ 1

z

2

(z − a)

_

z −

1

a

_ dz = 2πi

_

Res f ( f , 0) + Res( f , a)

_

where f (z) =

z

4

+1

z

2

(z−a)(z−

1

a

)

. It is easily obtained

Res( f , 0) = lim

z→0

d

dz

_

¸

¸

¸

¸

¸

¸

_

z

2

z

4

+ 1

z

2

(z − a)

_

z −

1

a

_

_

¸

¸

¸

¸

¸

¸

_

= lim

z→0

d

dz

_

¸

¸

¸

¸

¸

¸

_

z

4

+ 1

(z − a)

_

z −

1

a

_

_

¸

¸

¸

¸

¸

¸

_

= lim

z→0

−4z

3

_

z

2

− az −

1

a

z + 1

_

− (z

4

− 1)

_

2z − a −

1

a

_

_

z

2

− az −

1

a

z + 1

_

2

= a +

1

a

and

Res( f , a) = lim

z→a

(z − a)

z

4

+ 1

z

2

(z − a)

_

z −

1

a

_ = lim

z→a

z

4

+ 1

z

2

_

z −

1

a

_ =

a

4

+ 1

a

2

_

a −

1

a

_ =

a

4

+ 1

a(a

2

− 1)

.

So

_

¦z¦=1

z

4

+ 1

z

2

(z − a)

_

z −

1

a

_ dz = 2πi

_

a +

1

a

+

a

4

+ 1

a(a

2

− 1)

_

= 2πi

2a

4

a(a

2

= 1)

= −

4πia

3

1 − a

2

.

Finally,

_

π

0

cos 2θ dθ

1 − 2a cos θ + a

2

= −

1

4ia

_

−

4πia

3

1 − a

2

_

= π

a

2

1 − a

2

.

75

Exercise 2. Verify the following equations:

a)

_

∞

0

dx

(x

2

+ a

2

)

2

=

π

4a

3

, a > 0;

b)

_

∞

0

(log x)

3

1 + x

2

dx = 0;

c)

_

∞

0

cos ax

(1 + x

2

)

2

dx =

π(a + 1)e

−a

4

, if a > 0;

d)

_

π/2

0

dθ

a + sin

2

θ

=

π

2 [a(a + 1)]

1

2

, if a > 0;

e)

_

∞

0

log x

(1 + x

2

)

2

dx = −

π

4

;

f)

_

∞

0

dx

1 + x

2

=

π

2

;

g)

_

∞

−∞

e

ax

1 + e

x

dx =

π

sin aπ

, if 0 < a < 1;

h)

_

2π

0

log sin

2

2θ dθ = 4

_

4π

0

log sin θ dθ = −4π log 2.

Solution. a) Let f (z) =

1

(z

2

+a

2

)

2

and γ

R

= [−R, R] + C

R

where C

R

= Re

it

, 0 ≤ t ≤ π. So f has z

1

= ai as a

pole of order 2 enclosed by γ

R

. So we have by the Residue Theorem

_

γ

R

f (z) dz = 2πi Res( f , z

1

)

where

Res( f , z

1

) = lim

z→ai

d

dz

_

(z − ai)

2

1

(z − ai)

2

(z + ai)

2

_

= lim

z→ai

d

dz

_

1

(z + ai)

2

_

= lim

z→ai

−2

(z + ai)

3

= −

2

(2ai)

3

=

1

4a

3

i

.

Also

2πi

1

4a

3

i

=

π

2a

3

=

_

R

−R

dx

(x

2

+ a

2

)

2

+

_

C

R

f (z) dz

where the latter integral on the right hand side is zero since

¸

¸

¸

¸

¸

¸

_

C

R

1

(z

2

+ a

2

)

2

dz

¸

¸

¸

¸

¸

¸

≤

_

C

R

1

¦z

2

+ a

2

¦

2

¦dz¦ ≤

_

C

R

1

(¦z¦

2

− ¦a¦

2

)

2

¦dz¦ ≤

1

(R

2

− a

2

)

2

_

C

R

¦dz¦

=

πR

(R

2

− a

2

)

2

→ 0 as R → ∞.

76

Hence,

_

∞

0

dx

(x

2

+ a

2

)

2

=

1

2

_

∞

−∞

dx

(x

2

+ a

2

)

2

=

1

2

lim

R→∞

_

R

−R

dx

(x

2

+ a

2

)

2

=

1

2

π

2a

3

=

π

4a

3

.

c) Let f (z) =

e

iaz

(1+z

2

)

2

and γ

R

= [−R, R] + C

R

where C

R

= Re

it

, 0 ≤ t ≤ π. So f has z

1

= i as a pole of order 2

enclosed by γ

R

. By the Residue Theorem, we have

_

γ

R

f (z) dz = 2π Res( f , z

1

)

where

Res( f , z

1

) = lim

z→i

d

dz

_

(z − i)

2

e

iaz

(z − i)

2

(z + i)

2

_

= lim

z→i

d

dz

_

e

iaz

(z + i)

2

_

= lim

z→i

_

(z + i)

2

iae

iaz

− 2(z + i)e

iaz

(z + i)

4

_

= lim

z→i

_

(z + i)iae

iaz

− 2e

iaz

(z + i)

3

_

=

2i

2

ae

i

2

a

− 2e

i

2

a

(2i)

3

=

−2ae

−a

− 2e

−a

−8i

=

(a + 1)e

−a

4i

.

Also

2πi

(a + 1)e

−a

4i

=

π(a + 1)e

−a

2

=

_

R

−R

f (x) dx +

_

C

R

f (z) dz

where the latter integral on the right hand side is zero since

¸

¸

¸

¸

¸

¸

_

C

R

f (z) dz

¸

¸

¸

¸

¸

¸

=

¸

¸

¸

¸

¸

¸

_

C

R

e

iaz

(1 + z

2

)

2

dz

¸

¸

¸

¸

¸

¸

≤

_

C

R

¦e

iaz

¦

¦1 + z

2

¦

2

¦dz¦ ≤

_

C

R

1

(¦z¦

2

− 1)

2

¦dz¦

=

1

(R

2

− 1)

2

Rπ → 0 as R → ∞,

since ¦e

iaz

¦ = e

−Im(az)

≤ 1 since Im(az) ≥ 0 (a > 0 by assumption). Hence

_

∞

0

cos ax

(1 + x

2

)

2

dx =

1

2

_

∞

−∞

cos ax

(1 + x

2

)

2

dx =

1

2

lim

R→∞

_

R

−R

cos ax

(1 + x

2

)

2

dx =

1

2

π(a + 1)e

−a

2

=

π(a + 1)e

−a

4

provided a > 0.

g) First of all substitute u = e

x

, so

_

∞

−∞

e

ax

1 + e

x

dx =

_

∞

0

u

a

1 + u

du

u

=

_

∞

0

u

a−1

1 + u

du.

Now, let u = t

2

, then

_

∞

0

u

a−1

1 + u

du = 2

_

∞

0

t

2a−2

t

1 + t

2

dt = 2

_

∞

0

t

2a−1

1 + t

2

dt.

Hence,

_

∞

−∞

e

ax

1 + e

x

dx = 2

_

∞

0

x

2a−1

1 + x

2

dx.

Let f (z) =

z

2a−1

1+z

2

and γ = [−R, −r] + γ

r

+ [r, R] + γ

R

(0 < r < R). Deﬁne log z on G = ¦z ∈ C : z 0, −

π

2

<

arg(z) <

3π

2

¦ with log(z) = log ¦z¦ + iθ, θ = arg(z). f (z) has a simple pole i enclosed by γ. By the Residue

77

Theorem

_

γ

f (z) dz = 2πRes( f , i) = 2πi lim

z→i

(z − i)

z

2a−1

(z − i)(z + i)

= 2πi

i

2a−1

2i

= πi

2a−1

= πe

(2a−1) log i

= πe

(2a−1)(0+i

π

2

)

= πe

aπi

e

−i

π

2

= −πie

aπi

.

Also

_

γ

f (z) dz =

_

−r

−R

f (x) dx

.,,.

I

+

_

γ

r

f (z) dz

.,,.

II

+

_

R

r

f (x) dx

.,,.

III

+

_

γ

R

f (z) dz

.,,.

IV

.

I)

_

−r

−R

f (x) dx =

_

R

r

f (−x) dx =

_

R

r

(−x)

2a−1

1 + x

2

dx =

_

R

r

e

(2a−1)(log ¦−x¦+iπ)

1 + x

2

dx

= e

(2a−1)iπ

_

R

r

e

(2a−1) log x

1 + x

2

dx = −e

2aπi

_

R

r

e

(2a−1) log x

1 + x

2

dx

= −e

2aπi

_

R

r

x

2a−1

1 + x

2

dx.

IV)

¸

¸

¸

¸

¸

¸

_

γ

R

f (z) dz

¸

¸

¸

¸

¸

¸

=

¸

¸

¸

¸

¸

¸

_

γ

R

z

2a−1

1 + z

2

dz

¸

¸

¸

¸

¸

¸

≤

_

γ

R

¦z¦

2a−1

¦1 + z

2

¦

¦dz¦ ≤

R

2a−1

R

2

− 1

πR =

πR

2a

R

2

− 1

→ 0 as R → ∞

since a < 1.

II) Similar to IV)

¸

¸

¸

¸

¸

¸

_

γ

r

f (z) dz

¸

¸

¸

¸

¸

¸

≤

πr

2a

r

2

− 1

→ 0 as r → 0

since a > 0.

Hence, if r → 0 and R → ∞, we get

−πie

aπi

=

_

γ

f (z) dz = −e

2aπi

_

∞

0

f (x) dx + 0 +

_

∞

0

f (x) dx + 0 = (1 − e

2aπi

)

_

∞

0

f (x) dx.

Thus,

−πie

aπi

= (1 − e

2aπi

)

_

∞

0

f (x) dx

and so

_

∞

0

f (x) dx =

−πie

aπi

1 − e

2aπi

=

−iπ

e

−aπi

− e

aπi

=

iπ

e

aπi

− e

−aπi

=

iπ

2i sin(aπ)

=

π

2 sin(aπ)

.

Therefore,

_

∞

−∞

e

ax

1 + e

x

dx = 2

_

∞

0

x

2a−1

1 + x

2

dx = 2

π

2 sin(aπ)

=

π

sin(aπ)

provided 0 < a < 1.

Exercise 3. Find all possible values of

_

γ

exp z

−1

dz where γ is any closed curve not passing through z = 0.

Solution. Not available.

78

Exercise 4. Suppose that f has a simple pole at z = a and let g be analytic in an open set containing a.

Show that Res( f g; a) = g(a)Res( f ; a).

Solution. Not available.

Exercise 5. Use Exercise 4 to show that if G is a region and f is analytic in G except for simple poles at

a

1

, . . . , a

n

; and if g is analytic in G then

1

2πi

_

γ

f g =

n

k=1

n(γ; a

k

)g(a

k

)Res( f ; a

k

)

for any closed rectiﬁable curve γ not passing through a

1

, . . . , a

n

such that γ ≈ 0 in G.

Solution. Not available.

Exercise 6. Let γ be the rectangular path [n +

1

2

+ ni, −n −

1

2

+ ni, −n −

1

2

− ni, n +

1

2

− ni, n + i + ni] and

evaluate the integral

_

γ

π(z + a)

−2

cot πz dz for a an integer. Show that lim

n→∞

_

γ

π(z + a)

−2

cot πz dz = 0

and, by using the ﬁrst part, deduce that

π

2

sin

2

πa

=

∞

n=−∞

1

(a + n)

2

(Hint: Use the fact that for z = x + iy, ¦ cos z¦

2

= cos

2

x + sinh

2

y and ¦ sin z¦

2

= sin

2

x + sinh

2

y to show that

¦ cot πz¦ ≤ 2 for z on γ if n is suﬃciently large.)

Solution. Not available.

Exercise 7. Use Exercise 6 to deduce that

π

2

8

=

∞

n=0

1

(2n + 1)

2

Solution. Not available.

Exercise 8. Let γ be the polygonal path deﬁned in Exercise 6 and evaluate

_

γ

π(z

2

− a

2

)

−1

cot πz dz for a

an integer. Show that lim

n→∞

π(z

2

− a

2

)

−1

cot πz dz = 0, and consequently

π cot πa =

1

a

+

∞

n=1

2a

a

2

− n

2

for a an integer.

Solution. Not available.

Exercise 9. Use methods similar to those of Exercises 6 and 8 to show that

π

sin πa

=

1

a

+

∞

n=1

2(−1)

n

a

a

2

− n

2

for a an integer.

Solution. Not available.

79

Exercise 10. Let γ be the circle ¦z¦ = 1 and let m and n be non-negative integers. Show that

1

2πi

_

γ

(z

2

± 1)

m

dz

z

m+n+1

=

_

¸

¸

_

¸

¸

_

(±1)

p

(n+2p)!

p!(n+p)!

, if m = 2p + n, p ≥ 0

0, otherwise

Solution. Not available.

Exercise 11. In Exercise 1.12, consider a

n

and b

n

as functions of the parameter λ and use Exercise 10 to

compute power series expansions for a

n

(λ) and b

n

(λ). (b

n

(λ) is called a Bessel function.)

Solution. Not available.

Exercise 12. Let f be analytic in the plane except for isolated singularities at a

1

, a

2

, . . . , a

m

. Show that

Res( f ; ∞) = −

∞

k=1

Res( f , a

k

).

(Res( f ; ∞) is deﬁned as the residue of −z

−2

f (z

−1

) at z = 0. Equivalently, Res( f ; ∞) = −

1

2πi

_

f when

γ(t) = Re

it

, 0 ≤ t ≤ 2π, for suﬃciently large R.) What can you say if f has inﬁnitely many isolated

singularities?

Solution. Not available.

Exercise 13. Let f be an entire function and let a, b ∈ C such that ¦a¦ < R and ¦b¦ < R. If γ(t) = Re

it

,

0 ≤ t ≤ 2pi, evaluate

_

γ

[(z − a)(z − b)]

−1

f (z) dz. Use this result to give another proof of Liouville’s

Theorem.

Solution. Not available.

5.3 The Argument Principle

Exercise 1. Prove Theorem 3.6.

Solution. Not available.

Exercise 2. Suppose f is analytic on

¯

B(0; 1) and satisﬁes ¦ f (z)¦ < 1 for ¦z¦ = l. Find the number of solutions

(counting multiplicities) of the equation f (z) = z

n

where n is an integer larger than or equal to 1.

Solution. Not available.

Exercise 3. Let f be analytic in

¯

B(0; R) with f (0) = 0, f

·

(0) 0 and f (z) 0 for 0 < ¦z¦ ≤ R. Put

ρ = min¦¦ f (z)¦ : ¦z¦ = R¦ > 0. Deﬁne g : B(0; ρ) → C by

g(ω) =

1

2πi

_

γ

z f

·

(z)

f (z) − ω

dz

where γ is the circle ¦z¦ = R. Show that g is analytic and discuss the properties of g.

Solution. Not available.

Exercise 4. If f is meromorphic on G and

˜

f : G → C

∞

is deﬁned by

˜

f (z) = ∞ when z is a pole of f and

˜

f (z) = f (z) otherwise, show that

˜

f is continuous.

80

Solution. Not available.

Exercise 5. Let f be meromorphic on the region G and not constant; show that neither the poles nor the

zeros of f have a limit point in G.

Solution. Not available.

Exercise 6. Let G be a region and let H(G) denote the set of all analytic functions on G. (The letter

“H” stands for holomorphic. Some authors call a diﬀerentiable function holomorphic and call functions

analytic if they have a power series expansion about each point of their domain. Others reserve the term

“analytic” for what many call the complete analytic function, which we will not describe here.) Show that

H(G) is an integral domain; that is, H(G) is a commutative ring with no zero divisors. Show that M(G), the

meromorphic functions on G, is a ﬁeld.

We have said that analytic functions are like polynomials; similarly, meromorphic functions are analogues

of rational functions. The question arises, is every meromorphic function on G the quotient of two analytic

functions on G? Alternately, is M(G) the quotient ﬁeld of H(G)? The answer is yes but some additional

theory will be required before this answer can be proved.

Solution. Not available.

Exercise 7. State and prove a more general version of Rouche’s Theorem for curves other than circles in

G.

Solution. Not available.

Exercise 8. Is a non-constant meromorphic function on a region G an open mapping of G into C? Is it an

open mapping of G into C

∞

?

Solution. Not available.

Exercise 9. Let λ > 1 and show that the equation λ − z − e

−z

= 0 has exactly one solution in the half plane

¦z : Re z > 0¦. Show that this solution must be real. What happens to the solution as λ → 1?

Solution. Not available.

Exercise 10. Let f be analytic in a neighborhood of D =

¯

B(0; 1). If ¦ f (z)¦ < 1 for ¦z¦ = 1, show that there is

a unique z with ¦z¦ < 1 and f (z) = z. If ¦ f (z)¦ ≤ 1 for ¦z¦ = 1, what can you say?

Solution. Let g(z) = f (z) − z and h(z) = z. Clearly g, h are analytic in a neighborhood of

¯

B(0, 1) since f is

analytic. Note that g, h have no poles. Let γ = ¦z : ¦z¦ = 1¦. We have

¦g(z) + h(z)¦ = ¦ f (z) − z + z¦ = ¦ f (z)¦ < 1 = ¦z¦ = ¦h(z)¦ ≤ ¦g(z)¦ + ¦h(z)¦

on γ since ¦ f (z)¦ < 1 for ¦z¦ = 1. By Roché’s Theorem, we get

Z

g

− P

g

= Z

h

− P

h

.

But P

g

= P

h

= 0, so

Z

g

= Z

h

(inside the unit circle).

Since h(z) = z has only one zero (z = 0) inside the unit circle, we get that

g(z) = f (z) − z

has exactly one zero inside the unit circle. Hence, there is a unique z such that f (z) − z = 0 ⇐⇒ f (z) = z

with ¦z¦ < 1.

If we choose f (z) = z, then ¦ f (z)¦ ≤ 1 for ¦z¦ = 1 and therefore we have inﬁnitely many ﬁxed points.

If we choose f (z) = 1, then ¦ f (z)¦ ≤ 1 for ¦z¦ = 1 and therefore we do not have any ﬁxed points.

81

Chapter 6

The Maximum Modulus Theorem

6.1 The Maximum Principle

Exercise 1. Prove the following Minimum Principle. If f is a non-constant analytic function on a bounded

open set G and is continuous on G

−

, then either f has a zero in G or ¦ f ¦ assumes its minimum value on ∂G.

(See Exercise IV. 3.6.)

Solution. Since f ∈ C(

¯

G) we have ¦ f ¦ ∈ C(

¯

G). Hence ∃ a ∈

¯

G such that ¦ f (a)¦ ≤ ¦ f (z)¦ ∀z ∈

¯

G.

If a ∈ ∂G, then ¦ f ¦ assumes its minimum value on ∂G and we are done. Otherwise, if a ∂G, then a ∈ G

and we can write G =

_

A

i

where A

i

are the components of G, that is a ∈ A

i

for some i. But each A

i

is a

region, so we can use Exercise IV 3.6 which yields either f (a) = 0 or f is constant. But f was assumed

to be non-constant, so f has to have a zero in G. Therefore, either f has a zero in G or ¦ f ¦ assumes its

minimum value on ∂G.

Exercise 2. Let G be a bounded region and suppose f is continuous on G

−

and analytic on G. Show that

if there is a constant c ≥ 0 such that ¦ f (z)¦ = c for all z on the boundary of G then either f is a constant

function or f has a zero in G.

Solution. Assume there is a constant c ≥ 0 such that ¦ f (z)¦ = c for all z ∈ ∂G. According to the Maximum

Modulus Theorem (Version 2), we get

max¦¦ f (z)¦ : z ∈

¯

G¦ = max¦¦ f (z)¦ : z ∈ ∂G¦ = c.

So

¦ f (z)¦ ≤ c, ∀z ∈

¯

G. (6.1)

Since f ∈ C(G) which implies ¦ f ¦ ∈ C(

¯

G) and hence there exists an a ∈ G such that

¦ f (a)¦ ≤ ¦ f (z)¦ ≤

(6.1)

c,

then by Exercise IV 3.6 either f is constant of f has a zero in G.

Exercise 3. (a) Let f be entire and non-constant. For any positive real number c show that the closure of

¦z : ¦ f (z)¦ < c¦ is the set ¦z : ¦ f (z)¦ ≤ c¦.

(b) Let p be a polynomial and show that each component of ¦z : ¦ p(z)¦ < c¦ contains a zero of p. (Hint: Use

82

Exercise 2.)

(c) If p is a polynomial and c > 0 show that ¦z : ¦ p(z)¦ = c¦ is the union of a ﬁnite number of closed paths.

Discuss the behavior of these paths as c → ∞.

Solution. Not available.

Exercise 4. Let 0 < r < R and put A = ¦z : r ≤ ¦z¦ ≤ R¦. Show that there is a positive number > 0 such

that for each polynomial p,

sup¦¦ p(z) − z

−1

¦ : z ∈ A¦ ≥

This says that z

−1

is not the uniform limit of polynomials on A.

Solution. Not available.

Exercise 5. Let f be analytic on

¯

B(0; R) with ¦ f (z)¦ ≤ M for ¦z¦ ≤ R and ¦ f (0)¦ = a > 0. Show that the

number of zeros of f in B(0;

1

3

R) is less than or equal to

1

log 2

log

_

M

a

_

. Hint: If z

1

, . . . , z

n

are the zeros of f

in B(0;

1

3

R); consider the function

g(z) = f (z)

_

¸

¸

¸

¸

¸

_

n

_

k=1

_

1 −

z

z

k

_

_

¸

¸

¸

¸

¸

_

−1

,

and note that g(0) = f (0). (Notation:

n

k=1

a

k

= a

1

a

2

. . . a

n

.)

Solution. Not available.

Exercise 6. Suppose that both f and g are analytic on

¯

B(0; R) with ¦ f (z)¦ = ¦g(z)¦ for ¦z¦ = R. Show that if

neither f nor g vanishes in B(0; R) then there is a constant λ, ¦λ¦ = 1, such that f = λg.

Solution. Not available.

Exercise 7. Let f be analytic in the disk B(0; R) and for 0 ≤ r < R deﬁne A(r) = max¦Re f (z) : ¦z¦ = r¦.

Show that unless f is a constant, A(r) is a strictly increasing function of r.

Solution. Not available.

Exercise 8. Suppose G is a region, f : G → C is analytic, and M is a constant such that whenever z is on

∂

∞

G and ¦z

n

¦ is a sequence in G with z = limz

n

we have limsup ¦ f (z

n

)¦ ≤ M. Show that ¦ f (z)¦ ≤ M, for

each z in G.

Solution. We need to show

limsup

z→a

¦ f (z)¦ ≤ M ∀u ∈ ∂

∞

G.

Then we can use the Maximum Modulus Theorem (Version 3). Instead of showing that

limsup ¦ f (z

n

)¦ ≤ M ⇒ limsup

z→a

¦ f (z)¦ ≤ M,

we show the contrapositive, that is

limsup

z→a

¦ f (z)¦ > M ⇒ limsup ¦ f (z

n

)¦ > M.

So assume limsup

z→a

¦ f (z)¦ > M. But this implies limsup

z→a

¦ f (z

n

)¦ > M since z

n

→ z as z → ∞, that is z

n

gets arbitrarily close to z and since f is analytic, that is continuous, we have f (z

n

) gets arbitrarily close to

f (z).

83

6.2 Schwarz’s Lemma

Exercise 1. Suppose ¦ f (z)¦ ≤ 1 for ¦z¦ < 1 and f is a non-constant analytic function. By considering the

function g : D → D deﬁned by

g(z) =

f (z) − a

1 − ¯ af (z)

where a = f (0), prove that

¦ f (0)¦ − ¦z¦

1 + ¦ f (0)¦ ¦z¦

≤ ¦ f (z)¦ ≤

¦ f (0)¦ + ¦z¦

1 − ¦ f (0)¦ ¦z¦

for ¦z¦ < 1.

Solution. We claim that ¦g(z)¦ ≤ 1 and g(0) = 0. It is easy to check that g(0) = 0 since

g(0) =

f (0) − a

1 − ¯ af (0)

=

a − a

1 − ¯ aa

= 0.

We also have

¦g(z)¦ =

¦ f (z) − a¦

¦1 − ¯ af (z)¦

≤ 1

since ¦ f (z)¦ ≤ 1. (Can you see it?) Now, apply Schwarz Lemma.

¦g(z)¦ ≤ ¦z¦ for ¦z¦ < 1

⇐⇒ ¦ f (z) − a¦ ≤ ¦z¦ ¦1 − ¯ a f (z)¦

⇒ ¦¦ f (z)¦ − ¦a¦¦ ≤ ¦ f (z) − a¦ ≤ ¦z¦ ¦1 − ¯ af (z)¦ ≤ ¦z¦ + ¦z¦ ¦a¦ ¦ f (z)¦

⇒ ¦¦ f (z)¦ − ¦a¦¦ ≤ ¦z¦ + ¦z¦ ¦a¦ ¦ f (z)¦

⇒ −¦z¦ − ¦z¦ ¦a¦ ¦ f (z)¦ ≤ ¦ f (z)¦ − ¦a¦ ≤ ¦z¦ + ¦z¦ ¦a¦ ¦ f (z)¦

⇒

¦a¦ − ¦z¦

1 + ¦a¦ ¦z¦

≤ ¦ f (z)¦ ≤

¦a¦ + ¦z¦

1 − ¦a¦ ¦z¦

.

Set a = f (0) to obtain the desired result.

Exercise 2. Does there exist an analytic function f : D → D with f (

1

2

) =

3

4

and f

·

(

1

2

) =

2

3

?

Solution. The answer is no. Assume there exist an analytic function f : D → D with f (

1

2

) =

3

4

. According

to the deﬁnition, we have

E

a

α

= ¦ f ∈ A(D) : ¦ f (z)¦ ≤ 1, f (a) = α¦.

In our case, we have a =

1

2

and α =

3

4

. We know,

¸

¸

¸ f

·

(a)

¸

¸

¸ ≤

1 − ¦α¦

2

1 − ¦a¦

2

and therefore we must have

¸

¸

¸

¸

¸

¸

f

·

_

1

2

_

¸

¸

¸

¸

¸

¸

≤

1 −

_

3

4

_

2

1 −

_

1

2

_

2

=

7

16

3

4

=

7

16

3

4

=

7

12

= 0.58

¯

3.

But f

·

(

1

2

) =

2

3

= 0.

¯

6 which is not possible.

84

Exercise 3. Suppose f : D → C satisﬁes Re f (z) ≥ 0 for all z in D and suppose that f is analytic.

(a) Show that Re f (z) > 0 for all z in D.

(b) By using an appropriate Möbius transformation, apply Schwarz’s Lemma to prove that if f (0) = 1 then

¦ f (z)¦ ≤

1 + ¦z¦

1 − ¦z¦

for ¦z¦ < 1. What can be said if f (0) 1?

(c) Show that f (0) = 1, f also satisﬁes

¦ f (z)¦ ≥

1 − ¦z¦

1 + ¦z¦

(Hint: Use part (a)).

Solution. Not available.

Exercise 4. Prove Caratheodory’s Inequality whose statement is as follows: Let f be analytic on

¯

B(0; R)

and let M(r) = max¦¦ f (z)¦ : ¦z¦ = r¦, A(r) = max¦Re f (z) : ¦z¦ = r¦; then for 0 < r < R, if A(r) ≥ 0,

M(r) ≤

R + r

R − r

[A(R) + ¦ f (0)¦]

(Hint: First consider the case where f (0) = 0 and examine the function g(z) = f (Rz)[2A(R) + f (Rz)]

−1

for

¦z¦ < 1.)

Solution. Not available.

Exercise 5. Let f be analytic in D = ¦z : ¦z¦ < 1¦ and suppose that ¦ f (z)¦ ≤ M for all z in D. (a) If f (z

k

) = 0

for 1 ≤ k ≤ n show that

¦ f (z)¦ ≤ M

n

_

k=1

¦z − z

k

¦

¦1 − ¯ z

k

z¦

for ¦z¦ < 1. (b) If f (z

k

) = 0 for 1 ≤ k ≤ n, each z

k

0, and f (0) = Me

ia

(z

1

z

2

, . . . , z

n

), ﬁnd a formula for f .

Solution. Not available.

Exercise 6. Suppose f is analytic in some region containing

¯

B(0; 1) and ¦ f (z)¦ = 1 where ¦z¦ = 1. Find a

formula for f . (Hint: First consider the case where f has no zeros in B(0; 1).)

Solution. Not available.

Exercise 7. Suppose f is analytic in a region containing

¯

B(0; 1) and ¦ f (z)¦ = 1 when ¦z¦ = 1. Suppose that

f has a zero at z =

1

4

(1 + i) and a double zero at z =

1

2

. Can f (0) =

1

2

?

Solution. Not available.

Exercise 8. Is there an analytic function f on B(0; 1) such that ¦ f (z)¦ < 1 for ¦z¦ < 1, f (0) =

1

2

, and

f

·

(0) =

3

4

? If so, ﬁnd such an f . Is it unique?

Solution. Not available.

85

6.3 Convex functions and Hadamard’s Three Circles Theorem

Exercise 1. Let f : [a, b] → R and suppose that f (x) > 0 for all x and that f has a continuous second

derivative. Show that f is logarithmically convex iﬀ f

··

(x) f (x) − [ f

·

(x)]

2

≥ 0 for all x.

Solution. Not available.

Exercise 2. Show that if f : (a, b) → R is convex then f is continuous. Does this remain true if f is deﬁned

on the closed interval [a, b]?

Solution. Since f is convex on (a, b) and if a < s < t < u < b, we have

f (t) − f (s)

t − s

≤

f (u) − f (s)

u − s

≤

f (u) − f (t)

u − t

(6.2)

which we show now. Let a < s < t < u < b, t = λs + (1 − λ)u with λ =

u−t

u−s

∈ (0, 1). Since f is convex on

(a, b), we get

f (t) ≤

u − t

u − s

f (s) +

t − s

u − s

f (u)

which implies

(t − s) f (u) + (u − t) f (s) − (u − s) f (t) ≥ 0

which is equivalent to

t f (u) − s f (u) + uf (s) − t f (s) − uf (t) + s f (t) ≥ 0. (6.3)

From (6.3), we get (rearrange terms and add/subtract a term)

t f (u) − t f (s) − s f (u) + s f (s) − uf (t) − uf (s) + s f (t) − s f (s) ≥ 0

⇐⇒ (t − s)( f (u) − f (s)) − (u − s)( f (t) − f (s)) ≥ 0

⇐⇒

f (t) − f (s)

t − s

≤

f (u) − f (s)

u − s

.

Similar, from (6.3), we get

uf (u) − s f (u) − uf (t) + s f (t) − uf (u) + uf (s) + t f (u) − t f (s) ≥ 0

⇐⇒ (u − s)( f (u) − f (t)) − (u − t)( f (u) − f (s)) ≥ 0

⇐⇒

f (u) − f (s)

u − s

≤

f (u) − f (t)

u − t

.

Thus, we have proved (6.2).

Given x ∈ (a, b) choose δ > 0 such that [x − δ, x + δ] ⊂ (a, b).

Claim:

f (x) − f (x − δ)

δ

≤

f (z) − f (x)

z − x

≤

f (x + δ) − f (x)

δ

∀z ∈ (x − δ, x + δ). (6.4)

If the claim is true, (6.4) is equivalent to

f (x) − f (x − δ)

δ

(z − x) ≤ f (z) − f (x) ≤

f (x + δ) − f (x)

δ

(z − x).

Taking the limit z → x, we have that

f (x)−f (x−δ)

δ

(z − x) → 0 and

f (x+δ)−f (x)

δ

(z − x) → 0. Thus f (z) − f (x) → 0,

that is ¦z − x¦ < δ ⇒¦ f (z) − f (x)¦ < which shows that f is continuous.

86

Thus it remains to show the claim. The proof uses (6.2).

First, consider a point z ∈ (x − δ, x) and apply the second inequality in (6.2) gives

f (x) − f (x − δ)

δ

≤

f (z) − f (x)

z − x

which gives the ﬁrst inequality in (6.4).

Applying the outer inequality in (6.2) to the three points z < x < x + δ gives

f (x) − f (z)

x − z

≤

f (x + δ) − f (x)

x + δ − x

⇐⇒

f (z) − f (x)

z − x

≤

f (x + δ) − f (x)

δ

which gives the second inequality in (6.4).

Now, consider the case x < z < x + δ. Then, the ﬁrst inequality in (6.4) follows from the outer inequality in

(6.2) applied to the three points x − δ, x, z.

f (x) − f (x − δ)

x − x + δ

≤

f (z) − f (x)

z − x

⇐⇒

f (x) − f (x − δ)

δ

≤

f (z) − f (x)

z − x

.

The second inequality in (6.4) follows from the ﬁrst inequality in (6.2) applied to x, z, x + δ.

f (z) − f (x)

z − x

≤

f (x + δ) − f (x)

x + δ − x

⇐⇒

f (z) − f (x)

z − x

≤

f (x + δ) − f (x)

δ

.

Thus, we have proved the claim.

The statement is not true if f is deﬁned on the closed interval [a, b]! Here is a counterexample: Deﬁne

f : [a, b] → R by

f (x) =

_

¸

¸

_

¸

¸

_

0, x ∈ [a, b)

1, x = b

.

Clearly, f is convex on [a, b], but f is not continuous at x = b (but the function is continuous on (a, b)).

Exercise 3. Show that a function f : [a, b] → R is convex iﬀ any of the following equivalent conditions is

satisﬁed:

(a) a ≤ x < u < y ≤ b gives det

_

¸

¸

¸

¸

¸

¸

¸

¸

_

f (u) u 1

f (x) x 1

f (y) y 1

_

¸

¸

¸

¸

¸

¸

¸

¸

_

≥ 0;

(b) a ≤ x < u < y ≤ b gives

f (u)−f (x)

u−x

≤

f (y)−f (x)

y−x

;

(c) a ≤ x < u < y ≤ b gives

f (u)−f (x)

u−x

≤

f (y)−f (u)

y−u

;

Interpret these conditions geometrically.

Solution. Not available.

Exercise 4. Supply the details in the proof of Hadamard’s Three Circle Theorem.

Solution. Not available.

Exercise 5. Give necessary and suﬃcient conditions on the function f such that equality occurs in the

conclusion of Hadamard’s Three Circle Theorem.

Solution. Not available.

87

Exercise 6. Prove Hardy’s Theorem: If f is analytic on

¯

B(0; R) and not constant then

I(r) =

1

2π

_

2π

0

¦ f (re

iθ

)¦ dθ

is strictly increasing and log I(r) is a convex function of log r. Hint: If 0 < r

1

< r < r

2

ﬁnd a con-

tinuous function ϕ : [0, 2π] → C such that ϕ(θ) f (re

iθ

) = ¦ f (re

iθ

)¦ and consider the function F(z) =

_

2π

0

f (ze

iθ

)ϕ(θ) dθ. (Note that r is ﬁxed, so ϕ may depend on r.)

Solution. Not available.

Exercise 7. Let f be analytic in ann(0; R

1

, R

2

) and not identically zero; deﬁne

I

2

(r) =

1

2π

_

2π

0

¦ f (re

iθ

)¦

2

dθ.

Show that log I

2

(r) is a convex function of log r, R

1

< r < R

2

.

Solution. Not available.

6.4 The Phragmén-Lindelöf Theorem

Exercise 1. In the statement of the Phragmén-Lindelöf Theorem, the requirement that G be simply con-

nected is not necessary. Extend Theorem 4.1 to regions G with the property that for each z in ∂

∞

G there is

a sphere V in C

∞

centered at z such that V ∩ G is simply connected. Give some examples of regions that

are not simply connected but have this property and some which don’t.

Solution. Not available.

Exercise 2. In Theorem 4.1 suppose there are bounded analytic functions ϕ

1

, ϕ

2

, . . . , ϕ

n

on G that never

vanish and ∂

∞

G = A ∪ B

1

∪ . . . ∪ B

n

such that condition (a) is satisﬁed and condition (b) is also satisﬁed

for each ϕ

k

and B

k

. Prove that ¦ f (z)¦ ≤ M for all z in G.

Solution. We have ϕ

1

, . . . , ϕ

n

∈ A(G), bounded and ϕ

k

0 by assumption. Hence, ¦ϕ

k

(z)¦ ≤ κ

k

∀ z in G for

each k = 1, . . . , n. Further a) for every a in A, limsup

z→a

¦ f (z)¦ ≤ M and b) for every b in B

k

and η

k

> 0,

limsup

z→b

¦ f (z)¦¦ϕ

k

(z)¦

η

k

≤ M ∀ k = 1, . . . , n.

Also because G is simply connected, there is an analytic branch of log ϕ

k

(z) on G for each k = 1, . . . , n

(Corollary IV 6.17). Hence, g

k

(z) = exp¦η

k

log ϕ

k

(z)¦ is an analytic branch of ϕ

k

(z)

η

k

for η

k

> 0 and

¦g

k

(z)¦ = ¦ϕ

k

(z)¦

η

k

∀ k = 1, . . . , n. Deﬁne F : G → C by

F(z) = f (z)

n

_

k=1

g

k

(z)κ

−η

k

k

;

then F is analytic on G and

¦F(z)¦ ≤ ¦ f (z)¦

n

_

k=1

¦ϕ

k

(z)¦

η

k

κ

−η

k

k

≤

¦ϕ

k

(z)¦≤κ

k

∀k

¦ f (z)¦

n

_

k=1

κ

η

k

k

κ

−η

k

k

¦ϕ

k

(z)¦ = ¦ f (z)¦ ∀z ∈ G.

Hence,

limsup

z→a

¦ f (z)¦ =

(a),(b)

_

¸

¸

_

¸

¸

_

M, a ∈ A

Mκ

−η

k

k

, a ∈ B

k

∀k

.

88

By the Maximum Modulus Theorem III, we have

¦F(z)¦ ≤ max¦M, Mκ

−η

1

1

, . . . , Mκ

−η

n

n

¦ ∀z ∈ G.

Thus,

¦ f (z)¦ = ¦F(z)¦

n

_

k=1

¦g

k

(z)¦

−1

κ

η

k

k

≤

n

_

k=1

¸

¸

¸

¸

¸

κ

k

ϕ

k

(z)

¸

¸

¸

¸

¸

η

k

max¦M, Mκ

−η

1

1

, . . . , Mκ

−η

n

n

¦

for all z in G and for all η

k

> 0. Letting η

k

→ 0

+

∀k = 1, . . . , n gives that ¦ f (z)¦ ≤ M for all z in G.

Exercise 3. Let G = ¦z : ¦Im z¦ <

1

2

π¦ and suppose f : G → C is analytic and limsup

z→w

¦ f (z)¦ ≤ M for w

in ∂G. Also, suppose A < ∞and a < 1 can be found such that

¦ f (z)¦ < exp[Aexp(a¦Re z¦)]

for all z in G. Show that ¦ f (z)¦ ≤ M for all z in G. Examine exp(exp z) to see that this is the best possible

growth condition. Can we take a = 1 above?

Solution. Not available.

Exercise 4. Let f : G → C be analytic and suppose M is a constant such that limsup ¦ f (z

n

)¦ ≤ M for each

sequence ¦z

n

¦ in G which converges to a point in ∂

∞

G. Show that ¦ f (z)¦ ≤ M. (See Exercise 1.8).

Solution. Claim: Let M be a constant such that limsup

n→∞

¦ f (z

n

)¦ ≤ M for each sequence ¦z

n

¦ in G with

z = lim

n→∞

z

n

, ∀z ∈ ∂

∞

G (6.5)

implies

limsup

z→a

¦ f (z)¦ ≤ M for all a ∈ ∂

∞

G. (6.6)

If G would be a region together with the claim would give ¦ f (z)¦ ≤ M ∀z ∈ G by the Maximum Modulus

Theorem III.

However, G is assumed to be an open set. But every open set can be written as a union of components

(components are open and connected). So each component is a region.

Therefore, we can apply the Maximum Modulus Theorem III to each component once we have shown the

claim and additionally we have to show that the boundary of G is the same as taking the boundaries of the

components:

Clearly, since the components of G add up to G, we have that the boundary of the components must include

the boundary of G.

So it remains to show that the boundary of the component does not include more than the boundary of G.

Assume it would be the case. Then there needs to be a boundary of a component lying inside G (since every

component is lying inside G). But this would imply that it must be connected to another component of G

contradicting the fact a component is a maximally connected subset.

Hence, the boundary of G equals the boundary of all its components.

It remains to show the claim which we will do by showing the contrapositive.

Proof of the claim: Assume not (6.6), that is

∀M > 0 ∃a ∈ ∂

∞

G : limsup

z→a

¦ f (z)¦ > M

which is equivalent to

∀M > 0 ∃a ∈ ∂

∞

G : lim

r→0

+

sup¦¦ f (z)¦ : z ∈ G ∩ B(a; r)¦ > M

89

by deﬁnition. Then, clearly ∃δ > 0 such that lim

r→0

+ sup¦¦ f (z)¦ : z ∈ G ∩ B(a; r)¦ > M + δ. Let ¦z

n

¦ be a

sequence with lim

n→∞

= a and let r

n

= 2¦z

n

− a¦. Obviously, we have lim

n→∞

r

n

= 0. Next, consider the

sequence α

n

= sup¦¦ f (z)¦ : z ∈ F ∩ B(a; r

n

)¦. Then lim

n→∞

α

n

> M + δ. In addition, ¦α

n

¦ is a nonincreasing

sequence, therefore lim

n→∞

α

n

> M + δ implies α

n

> M + δ ∀n. Thus, ∃ y

n

∈ G ∩ B(a; r

n

) such that

¦ f (y

n

)¦ > M + δ. Taking limsup in this inequality yields

limsup

n→∞

¦ f (y

n

)¦ ≥ M + δ,

that is

limsup

n→∞

¦ f (y

n

)¦ ≥ M

where lim

n→∞

y

n

= a, which gives “not (6.5)”. Hence (6.5) implies (6.6).

Exercise 5. Let f : G → C be analytic and suppose that G is bounded. Fix z

0

in ∂G and suppose that

limsup

z→w

¦ f (z)¦ ≤ M for w in ∂G, w z

0

. Show that if lim

z→z

0

¦z − z

0

¦

**¦ f (z)¦ = 0 for every > 0 then
**

¦ f (z)¦ ≤ M for every z in ∂G. (Hint: If a G, consider ϕ(z) = (z − z

0

)(z − a)

−1

.)

Solution. Not available.

Exercise 6. Let G = ¦z : Re z > 0¦ and let f : G → C be an analytic function with limsup

z→w

¦ f (z)¦ ≤ M

for w in ∂G, and also suppose that for every > 0,

lim

r→∞

sup¦exp(−/r)¦ f (re

iθ

)¦ : ¦θ¦ <

1

2

π¦ = 0.

Show that ¦ f (z)¦ ≤ M for all z in G.

Solution. Not available.

Exercise 7. Let G = ¦z : Re z > 0¦ and let f : G → C be analytic such that f (1) = 0 and such that

limsup

z→w

¦ f (z)¦ ≤ M for w in ∂G. Also, suppose that for every δ, 0 < δ < 1, there is a constant P such that

¦ f (z)¦ ≤ Pexp(¦z¦

1−δ

).

Prove that

¦ f (z)¦ ≤ M

_

(1 − x)

2

+ y

2

(1 + x)

2

+ y

2

_

1

2

.

Hint: Consider f (z)

_

1+z

1−z

_

.

Solution. Not available.

90

Chapter 7

Compactness and Convergence in the

Space of Analytic Functions

7.1 The space of continuous functions C(G, Ω)

Exercise 1. Prove Lemma 1.5 (Hint: Study the function f (t) =

t

1+t

for t > −1.)

Solution. To show that µ : S × S → R, µ(s, t) =

d(s,t)

1+d(s,t)

is a metric we proof that µ satisﬁes the conditions

of a metric function.

For s, t ∈ S the value µ(s, t) =

d(s,t)

1+d(s,t)

is well deﬁned in the nonnegative real numbers because d is a metric.

Also µ(s, t) = 0 if and only if d(s, t) = 0 and since d is a metric it follows that s = t. The metric d is

symmetric and therefore µ is also. It remains to show the triangle inequality. Let therefore s, t, u ∈ S . The

metric d satisﬁes the triangle inequality

d(s, u) ≤ d(s, t) + d(t, u). (7.1)

The function f (t) =

t

1+t

is a strictly increasing, concave function on the interval [0, ∞). Thus if d(s, u) ≤

max¦d(s, t), d(t, u)¦, then also µ(s, u) ≤ max¦µ(s, t), µ(t, u)¦ and by nonnegativity of µ also

µ(s, u) ≤ µ(s, t) + µ(t, u).

If instead max¦d(s, t), d(t, u)¦ < d(s, u) then

1

1 + d(s, u)

< min

_

1

1 + d(s, t)

,

1

1 + d(t, u)

_

.

Together with equation (7.1) we have

µ(s, u) =

d(s, u)

1 + d(s, u)

≤

d(s, t) + d(t, u)

1 + d(s, u)

=

d(s, t)

1 + d(s, u)

+

d(t, u)

1 + d(s, u)

≤

d(s, t)

1 + d(s, t)

+

d(t, u)

1 + d(t, u)

= µ(s, t) + µ(t, u),

which gives the triangle inequality.

To show that the metrics d and µ are equivalent on S , let O be on open set in (S, d) and let x ∈ O. There is

91

ε ∈ (0, 1) such that B

d

(x, ε) ⊂ O. Choose δ positive such that δ <

ε

1+ε

, then B

µ

(x, δ) ⊂ B

d

(x, ε). Similarly, if

O is an open set in (S, µ) and δ is such that x ∈ B

µ

(x, δ) ⊂ O then choose ε > 0 with ε ≤

δ

1−δ

which implies

B

d

(x, ε) ⊂ B

µ

(x, δ). Since this can be done for any element x ∈ O, open sets in (S, d) are open in (S, µ) and

vice-versa.

The fact that exactly the open sets in (S, d) are open in (S, µ) leads to the statement about Cauchy sequences.

Assume that ¦x

n

¦

n

is a Cauchy sequence in (S, d). To see that it is Cauchy in (S, µ), let ε > 0 be arbitrary

but ﬁxed. By the above there is a δ small enough that B

d

(x, δ) ⊂ B

µ

(x, ε) and δ depends only on ε, but not

on x ∈ S . Since ¦x

n

¦

n

is Cauchy in (S, d) there is N ∈ N such that d(x

m

, x

n

) < δ whenever m, n ≥ N and

therefore also µ(x

m

, x

n

) < ε for m, n ≥ N. The opposite statement holds with a similar argument.

Exercise 2. Find the sets K

n

obtained in Proposition 1.2 for each of the following choices of G: (a) G is an

open disk; (b) G is an open annulus; (c) G is the plane with n pairwise disjoint closed disks removed; (d) G

is an inﬁnite strip; (e) G = C − Z.

Solution. a) G is an open disk,

Let a ∈ C and r > 0 such that G = B(a, r). Set

K

n

= B

_

a, r −

1

n

_

where a ball of negative radius is to be understood as the empty set. Depending on the size of r, a

ﬁnite number of K

n

may be empty, not violating the condition K

n

⊂ int K

n+1

, and obviously G =

_

∞

n=1

.

b) G is an open annulus,

Let a ∈ C be the center of the annulus with radii r, R, 0 ≤ r < R < ∞so that G = ann(a, r, R). Deﬁne

K

n

= ann

_

a, r +

1

n

, R −

1

n

_

,

again with the interpretation that an annulus with inner radius larger than the outer radius is the

empty set.

d) G = ¦z ∈ C : ¦Imz¦ < 1¦,

For this open set G deﬁne

K

n

= ¦z = x + iy ∈ C : x, y ∈ R, ¦x¦ ≤ n¦ ∩

_

z = x + iy ∈ C : x, y ∈ R¦y¦ ≤ 1 −

1

n

_

.

Although G is unbounded each K

n

is bounded and closed and hence compact and G =

_

∞

n=1

K

n

.

e) G = C − Z.

Here we can deﬁne

K

n

= ¦z ∈ C : ¦z¦ ≤ n¦ ∩

_

¸

¸

¸

¸

¸

¸

_

n

_

j=−n

B

_

( j,

1

n

_

_

¸

¸

¸

¸

¸

¸

_

C

,

an intersection of closed set, one of which is bounded, hence K

n

is compact. Also K

n

⊂ int K

n+1

by

deﬁnition and G =

_

∞

n=1

.

Exercise 3. Supply the omitted details in the proof of Proposition 1.18.

92

Solution. Not available.

Exercise 4. Let F be a subset of a metric space (X, d) such that F

−

is compact. Show that F is totally

bounded.

Solution. Not available.

Exercise 5. Suppose ¦ f

n

¦ is a sequence in C(G, Ω) which converges to f and ¦z

n

¦ is a sequence in G which

converges to a point z in G. Show lim f

n

(z

n

) = f (z).

Solution. Not available.

Exercise 6. (Dini’s Theorem) Consider C(G, R) and suppose that ¦ f

n

¦ is a sequence in C(G, R) which is

monotonically increasing (i.e., f

n

(z) ≤ f

n+1

(z) for all z in G) and lim f

n

(z) = f (z) for all z in G where

f ∈ C(G, R). Show that f

n

→ f .

Solution. By Proposition 1.2 we can ﬁnd a sequence ¦K

n

¦ of compact subsets of G such that

_

∞

n=1

K

n

. Thus,

WLOG we can assume that G is compact. Deﬁne g

n

= f − f

n

∀n. Clearly g

n

is continuous ∀n, since f

and f

n

are continuous by assumption. Since f

n

is monotonically increasing, g

n

is monotonically decreasing.

Since f

n

converges to f , g

n

converges pointwise to zero, that is

∀ > 0 ∃N ∈ N ∀x ∈ G : 0 ≤ g

N

(x) <

2

.

Since g

N

is continuous, there is an open neighborhood B(x; r) around x (with r < ∞). For any y ∈ B(x; r)

we have g

N

(y) < , that is

∀ > 0 ∀x ∈ G ∃N ∈ N ∃B(x; r) : ∀y ∈ B(x; r) : g

N

(y) < .

Since g

N

is monotonically decreasing, we surely get g

n

(y) < for every n ≥ N and for every y ∈ B(x; r).

These open neighborhoods will cover G. Since G is compact, we can ﬁnd ﬁnitely many x

1

, . . . , x

k

such that

B(x

i

; r) cover G for every value of . Note that the values of N can change the chosen center x of each

those neighborhoods, say N

i

. Therefore, let M = max

i

N

i

. Because of the monotonicity of g

n

, we have that

g

n

(y) < ∀ n ≥ M in every open neighborhood B(x

i

; r), that is, for every value in G. Hence,

∀ > 0 ∃M ∈ N, ∀n ≥ M ∀x ∈ G : ¦ f

n

(x) − f (x)¦ < ,

which shows f

n

→ f .

Exercise 7. Let ¦ f

n

¦ ⊂ C(G, Ω) and suppose that ¦ f

n

¦ is equicontinuous at each point of G. If f ∈ C(G, Ω)

and f (z) = lim f

n

(z) for each z then show that f

n

→ f .

Solution. By Proposition 1.2 we can ﬁnd a sequence ¦K

n

¦ of compact subsets of G such that

_

∞

n=1

K

n

. Thus,

WLOG we can assume that G is compact. Since f

n

is equicontiuous at each point of G, we have that

∀ > 0 ∃δ > 0 : d( f

n

(x), f

n

(y)) <

3

whenever ¦x − y¦ < δ. (7.2)

Taking the limit as n → ∞, we also get

d( f (x), f (y)) <

3

whenever ¦x − y¦ < δ. (7.3)

by the pointwise convergence ( f (z) = lim f

n

(z) ∀z ∈ G by assumption). As in the previous exercise, we

can ﬁnd ﬁnitely many points y

1

, . . . , y

k

∈ G such that the open neighborhoods B(y

i

; δ) cover G, since G is

93

compact. To be more precise: For every point x in G, we can ﬁnd an open neighborhood B(y

i

; δ) for some

i = 1, . . . , k. Since f

n

→ f pointwise, we have

d( f

n

(y

i

), f (y

i

)) <

3

∀n ≥ N

i

.

Set N = max

i

N

i

, then we have

d( f

n

(y

i

), f (y

i

)) <

3

∀n ≥ N ∀i = 1, . . . , k. (7.4)

Finally, for an arbitrary x ∈ G and n ≥ N we have

d( f

n

(x), f (x)) ≤ d( f

n

(x), f

n

(y

i

)) + d( f

n

(y

i

), f (y

i

)) + d( f (y

i

), f (x)) <

(7.2),(7.3),(7.4)

3

+

3

+

3

=

where we choose y

i

such that ¦x − y

i

¦ < δ. This shows f

n

→ f .

Exercise 8. (a) Let f be analytic on B(0; R) and let f (z) =

_

∞

n=0

a

n

z

n

for ¦z¦ < R. If f

n

(z) =

_

n

k=0

a

k

z

k

, show

that f

n

→ f in C(G; C).

(b) Let G = ann(0; 0, R) and let f be analytic on G with Laurent series development f (z) =

_

∞

n=−∞

a

n

z

n

. Put

f

n

(z) =

_

n

k=−∞

a

k

z

k

and show that f

n

→ f in C(G; C).

Solution. a) Let G = B(0; R). This result is proved with the Arzela-Ascoli Theorem. With the given function

f deﬁne an analytic function

g : G → C, g(z) :=

∞

k=0

¦a

k

¦z

k

. (7.5)

Let z ∈ G and deﬁne r :=

¦z¦+R

2

(< R) then z ∈ B(0; r). Let / = ¦ f

n

¦

n

. Then for this ﬁxed z and for any n ∈ N

we have

¦ f

n

(z)¦ = ¦

n

k=0

a

k

z

k

¦ ≤

n

k=0

¦a

k

¦¦z¦

k

= g(¦z¦) < ∞.

Thus for a given z ∈ G, ¦ f

n

(z) : f

n

∈ /¦ is bounded and therefore has compact closure.

Next, we want to show equicontinuity at each point of G. Fix z

0

∈ G and let ε be positive. Let r > 0 be such

that B(z

0

; r) ⊂ G. With the function g(z) deﬁned in equation (7.5) and the computations from above we have

for all w ∈ ∂B(z

0

; r) and for all n ∈ N

¦ f

n

(w)¦ ≤ g(¦w¦).

The boundary of the disk with radius r is compact and so is g(∂B(z

0

; r)). Thus there is a positive number M

such that ¦g(∂B(z

0

, r))¦ ≤ M.

Choose δ ∈

_

0, min¦

r

2

,

εr

2M

¦

_

and let ¦z − z

0

¦ < δ then as in the proof of Montel’s theorem the following

estimate holds

¦ f

n

(z) − f

n

(z

0

)¦ = ¦

1

2πi

_

¦w−z

0

¦=r

f

n

(w)(z − z

0

)

(w − z)(w − z

0

)

dw¦

≤

2M

r

δ ≤ ε.

By Arzela-Ascoli the sequence ¦ f

n

¦ is normal in C(G) so there is a subsequence ¦ f

n

k

¦ that converges in C(G).

Since f is analytic and ¦ f

n

¦ is the sequence of partial sums of f , ¦ f

n

¦ itself is an analytic Cauchy sequence

in C(G). Hence every subsequence converges to the same limit, which shows that f

n

→ f in C(G). As a last

step recall that H(G) is complete as a subspace of C(G) (Corollary 2.3) and ¦ f

n

¦ ⊂ H(G) thus also the limit

function f ∈ H(G).

94

7.2 Spaces of analytic functions

Exercise 1. Let f , f

1

, f

2

, . . . be elements of H(G) and show that f

n

→ f iﬀ for each closed rectiﬁable curve

γ in G, f

n

(z) → f (z) uniformly for z in ¦γ¦.

Solution. Not available.

Exercise 2. Let G be a region, let a ∈ R, and suppose that f : [a, ∞] × G → C is a continuous func-

tion. Deﬁne the integral F(z) =

_

∞

a

f (t, z) dt to be uniformly convergent on compact subsets of G if

lim

b→∞

_

b

a

f (t, z) dt exists uniformly for z in any compact subset of G. Suppose that this integral does con-

verge uniformly on compact subsets of G and that for each t in (a, ∞), f (t, ) is analytic on G. Prove that F

is analytic and

F

(k)

(z) =

_

∞

a

∂

k

f (t, z)

∂z

k

dt

Solution. Not available.

Exercise 3. The proof of Montel’s Theorem can be broken up into the following sequence of deﬁnitions and

propositions: (a) Deﬁnition. A set / ⊂ C(G, C) is locally Lipschitz if for each a in G there are constants

M and r > 0 such that ¦ f (z) − f (a)¦ ≤ M¦z − a¦ for all f in / and ¦z − a¦ < r. (b) If / ∈ C(G, C) is locally

Lipschitz then / is equicontinuous at each point of G. (c) If / ⊂ H(G) is locally bounded then / is locally

Lipschitz.

Solution. Not available.

Exercise 4. Prove Vitali’s Theorem: If G is a region and ¦ f

n

¦ ⊂ H(G) is locally bounded and f ∈ H(G) that

has the property that A = ¦z ∈ G : lim f

n

(z) = f (z)¦ has a limit point in G then f

n

→ f .

Solution. The sequence ¦ f

n

¦ is a locally bounded sequence in H(G) and by Montel’s Theorem then ¦ f

n

¦ is

normal, so there is a subsequence ¦ f

n

k

¦ ⊂ ¦ f

n

¦ that converges to f in H(G).

Suppose that f

n

f in H(G) then there must be a compact set K ⊂ G the convergence is not uniform on

K. In other words there must be an ε > 0 so that for all n ∈ N there is z

n

∈ K with ¦ f

n

(z

n

) − f (z

n

)¦ ≥ ε. By

compactness of K extract a subsequence of ¦z

n

¦, say ¦z

n

m

¦ that converges to a point z

0

∈ K.

The sequence ¦ f

n

¦ is locally bounded, so in particular the subsequence ¦ f

n

m

¦ with indices corresponding

to ¦z

n

m

¦ is, and again by Montel’s Theorem now applied to the subsequence and again the completeness

of H(G) there is an analytic function g such that f

n

m

→ g in H(G). On the set A of points of pointwise

convergence f

n

(z) → f and f

n

(z) → g.

Now G is a region and A has a limit point in G so by the Identity Theorem (Chapter 4, Corollary 3.8)

already f = g on G which gives a contradiction on the set K and the point z

0

because

¦ f

n

m

(z

n

m

) − f (z

n

m

)¦ → ¦g(z

0

) − f (z

0

)¦ ≥ ε.

Hence we can conclude that f

n

→ f in H(G).

Exercise 5. Show that for a set / ⊂ H(G) the following are equivalent conditions:

(a) / is normal;

(b) For every > 0 there is a number c > 0 such that ¦c f : f ∈ /¦ ⊂ B(0; ) (here B(0; ) is the ball in

H(G) with center at 0 and radius ).

Solution. Not available.

Exercise 6. Show that if / ⊂ H(G) is normal then /

·

= ¦ f

·

: f ∈ /¦ is also normal. Is the converse true?

Can you add something to the hypothesis that /

·

is normal to insure that / is normal?

95

Solution. Not available.

Exercise 7. Suppose / is normal in H(G) and Ω is open in C such that f (G) ⊂ Ω for every f in /. Show

that if g is analytic on Ω and is bounded on bounded sets then ¦g ◦ f : f ∈ /¦ is normal.

Solution. Not available.

Exercise 8. Let D = ¦z : ¦z¦ < 1¦ and show that / ⊂ H(D) is normal iﬀ there is a sequence ¦M

n

¦ of positive

constants such that limsup

n

√

M

n

≤ 1 and if f (z) =

_

∞

0

a

n

z

n

is in / then ¦a

n

¦ ≤ M

n

for all n.

Solution. Not available.

Exercise 9. Let D = B(0; 1) and for 0 < r < 1 let γ

r

(t) = re

2πit

, 0 ≤ t ≤ 1. Show that a sequence ¦ f

n

¦ in

H(D) converges to f iﬀ

_

γ

r

¦ f (r) − f

n

(z)¦ ¦dz¦ → 0 as n → ∞for each r, 0 < r < 1.

Solution. ⇒: Assume lim

n→∞

f

n

(z) = f (z) ∀z ∈ D. Then

_

γ

r

¦ f (z) − f

n

(z)¦ ¦dz¦ ≤ sup

z∈γ

r

¦ f (z) − f

n

(z)¦

_

γ

r

¦dz¦

=

γ

r

=re

2πit

,0≤t≤1

sup

z∈γ

r

¦ f (z) − f

n

(z)¦ 2πr, ∀0 < r < 1.

But sup

z∈γ

r

¦ f (z) − f

n

(z)¦ →as n → ∞since lim

n→∞

f

n

(z) = f (z) ∀ z ∈ D. Therefore,

_

γ

r

¦ f (z) − f

n

(z)¦ ¦dz¦ → 0

as n → ∞for each r, 0 < r < 1.

⇐: Assume

_

γ

r

¦ f (z) − f

n

(z)¦ ¦dz¦ → 0 as n → ∞for each r, 0 < r < 1. Let K be an arbitrary compact subset

of D = B(0; 1) and choose 0 < r < 1 such that K ⊂ B(0; r) and denote the shortest distance between an

arbitrary point in K and an arbitrary point on ∂B by δ > 0. By the assumption, we have

_

∂B

¦ f (z) − f

n

(z)¦ ¦dz¦ < 2πδ. (7.6)

Let a ∈ K, then by Cauchy’s Integral formula, we have

f (a) − f

n

(a) =

1

2πi

_

∂B

f (z) − f

n

(z)

z − a

dz.

Thus,

¦ f (a) − f

n

(a)¦ ≤

1

2π

_

∂B

¦ f (z) − f

n

(z)¦

¦z − a¦

¦dz¦ ≤

1

2πδ

_

∂B

¦ f (z) − f

n

(z)¦ ¦dz¦ <

(7.6)

2πδ

2πδ

= .

Hence, f

n

(a) → f (z) uniformly ∀a ∈ K. Since it converges uniformly on all compact subsets of D, Proposi-

tion 1.10 b) yields that ¦ f

n

¦ converges to f ∀ z ∈ D.

Exercise 10. Let ¦ f

n

¦ ⊂ H(G) be a sequence of one-one functions which converge to f . If G is a region,

show that either f is one-one or f is a constant function.

Solution. Assume ¦ f

n

¦ ⊂ H(G) is a sequence of one-one functions which converge to f where G is a region

(open and connected). Suppose f is not the constant function. Then, we have to show that f is one-one.

Choose an arbitrary point z

0

∈ G and deﬁne the sequence g

n

(z) = f

n

(z) − f

n

(z

0

). Clearly, the sequence ¦g

n

¦

converges to g(z) = f (z) − f (z

0

) on the open connected set G\¦z

0

¦, which is again a region. In addition, we

96

have that g

n

never vanishes on G\¦z

0

¦, since f

n

are assumed to be one-one functions.

Therefore, the sequence ¦g

n

¦ satisﬁes the conditions of the Corollary 2.6 resulting from Hurwitz’s Theorem,

where the region is G\¦z

0

¦. Thus, either g ≡ 0 or g never vanishes (g has no zeros). Since g(z) = f (z) − f (z

0

)

is not identically zero on G\¦z

0

¦, we must have g(z) = f (z) − f (z

0

) has no zeros in G\¦z

0

¦. This implies

f (z) f (z

0

) ∀z ∈ G\¦z

0

¦.

Since z

0

was an arbitrary point in G, we have shown that f is one-one on G.

Exercise 11. Suppose that ¦ f

n

¦ is a sequence in H(G), f is a non-constant function, and f

n

→ f in H(G).

Let a ∈ G and α = f (a); show that there is a sequence ¦a

n

¦ in G such that: (i) a = lima

n

; (ii) f

n

(a

n

) = α

for suﬃciently large n.

Solution. Not available.

Exercise 12. Show that limtan nz = −i uniformly for z in any compact subset of G = ¦z : Im z > 0¦.

Solution. Not available.

Exercise 13. (a) Show that if f is analytic on an open set containing the disk

¯

B(a; R) then

¦ f (a)¦

2

≤

1

πR

2

_

2π

0

_

R

0

¦ f (a + re

iθ

)¦

2

r dr dθ.

(b) Let G be a region and let M be a ﬁxed positive constant. Let / be the family of all functions f in H(G)

such that

G

¦ f (z)¦

2

dx dy ≤ M. Show that / is normal.

Solution. a) Let

¯

B(a; R) ⊂ G and f : G → C be analytic. Let γ(t) = a + re

it

for 0 ≤ t ≤ 2π, then by

Cauchy’s Integral Formula we have

g(a) =

1

2πi

_

γ

g(w)

w − a

dw =

1

2π

_

2π

0

f (a + re

it

) dt.

Thus, letting g(z) = f (z)

2

, we obtain

f

2

(a) =

1

2π

_

2π

0

_

f (a + re

it

)

_

2

dt.

Multiply by r, we get

f

2

(a)r =

1

2π

_

2π

0

_

f (a + re

iθ

)

_

2

r dθ (7.7)

and then integrate with respect to r yields

_

R

0

f

2

(a)r dr = f

2

(a)

R

2

2

=

(7.7)

_

R

0

1

2π

_

2π

0

_

f (a + re

iθ

)

_

2

r dθ dr =

1

2π

_

2π

0

_

R

0

_

f (a + re

iθ

)

_

2

r dr dθ

which implies

f

2

(a) =

1

πR

2

_

2π

0

_

R

0

_

f (a + re

iθ

)

_

2

r dr dθ.

So

¦ f (a)¦

2

≤

1

πR

2

_

2π

0

_

R

0

¸

¸

¸ f (a + re

iθ

)

¸

¸

¸

2

r dr dθ.

97

b) Let G be a region and M be a ﬁxed constant. Let / be the family of all functions f in H(G) such that

G

¦ f (z)¦

2

dx dy ≤ M. (7.8)

To show that / is normal.

According to Montel’s Theorem, it suﬃces to show that / is locally bounded, that is ∃ r > 0 such that

sup¦¦ f (z)¦ : ¦z − a¦ < r, f ∈ /¦ < ∞(p. 153). We will prove this fact by contradiction:

Assume / is not locally bounded. Then there is a compact set K ⊂ G such that sup¦¦ f (z)¦ : z ∈ K, f ∈ /¦ =

∞. That is, there is a sequence ¦ f

n

¦ in / such that sup¦¦ f

n

(z)¦ : z ∈ K¦ ≥ n. Therefore, ∃ z

m

∈ K such that

¦ f

n

(z

m

)¦ ≥

n

2

. (7.9)

Every sequence has a convergence subsequence, say ¦z

n

k

¦, with z

n

k

→ z

0

∈ K since K is compact. WLOG,

we write ¦z

n

k

¦ = ¦z

n

¦. Clearly z

0

∈ G since z

0

∈ K (K ⊂ G), hence ∃ R > 0 such that B(z

0

; R) ⊂ G (since

G is open). If we pick m large enough, we can get z

m

∈ B(z

0

;

R

2

) (a picture might help). In addition, we can

ﬁnd an

ˆ

R > 0 such that B(z

m

;

ˆ

R) ⊂ B(z

0

;

R

2

) because B(z

0

;

R

2

) is open. Apply part a) to the ball B(z

m

;

ˆ

R) to

obtain:

¦ f

n

(z

m

)¦

2

≤

1

π

ˆ

R

2

_

2π

0

_

ˆ

R

0

¸

¸

¸ f

n

(z

m

+ re

iθ

)

¸

¸

¸

2

r dr dθ ≤

B(z

m

;

ˆ

R)⊂G

1

π

ˆ

R

2

G

¦ f

n

(z)¦

2

dx dy ≤

(7.8)

1

π

ˆ

R

3

M < ∞.

So

¦ f

n

(z

m

)¦ ≤

_

M

π

1

ˆ

R

< ∞.

But by (7.9) we have

¦ f

n

(z

m

)¦ ≥

n

2

.

This gives a contradiction if we let n → ∞. Thus, / is locally bounded and therefore normal.

7.3 Spaces of meromorphic functions

Exercise 1. Prove Proposition 3.3.

Solution. a) Let a ∈ C and r > 0. To show: There is a number δ > 0 such that B

∞

(a; δ) ⊂ B(a; r). Let

R > 0 such that B(a; r) ⊂ B(0; R) and B

∞

(a; δ) ⊂ B(0; R).

Claim: Pick

δ =

2r

√

1 + R

2

_

1 + ¦a¦

2

to obtain

B

∞

(a; δ) ⊂ B(a; R).

Proof of the claim: Let z ∈ B

∞

(a; δ). This is equivalent to

2¦z − a¦

_

1 + ¦z¦

2

_

1 + ¦a¦

2

< δ

⇐⇒

2¦z − a¦

_

1 + ¦z¦

2

_

1 + ¦a¦

2

<

2r

√

1 + R

2

_

1 + ¦a¦

2

⇐⇒ ¦z − a¦ <

r

√

1 + R

2

_

1 + ¦z¦

2

98

which implies (z ∈ B

∞

(a; δ) ⇒ z ∈ B(0; R), that is ¦z¦ < R)

¦z − a¦ <

r

√

1 + R

2

√

1 + R

2

⇐⇒ ¦z − a¦ < r.

So z ∈ B(a; r). Therefore

B

∞

(a; δ) ⊂ B(a; r).

b) Let δ > 0 and a ∈ C. To show: There is a number r > 0 such that B(a; r) ⊂ B

∞

(a; δ).

Claim: Pick r <

δ

2

to obtain

B(a; r) ⊂ B

∞

(a; δ).

Proof of the claim: Let z ∈ B(a; r). This is equivalent to

¦z − a¦ < r

⇒

r<

δ

2

¦z − a¦ <

δ

2

⇐⇒ ¦z − a¦ <

δ

2

1

⇒

1≤

√

1+¦z¦

2

√

1+¦a¦

2

¦z − a¦ <

δ

2

_

1 + ¦z¦

2

_

1 + ¦a¦

2

⇐⇒

2¦z − a¦

_

1 + ¦z¦

2

_

1 + ¦a¦

2

< δ

⇐⇒ d(z, a) < δ.

So z ∈ B

∞

(a; δ). Therefore,

B(a; r) ⊂ B

∞

(a; δ).

c) Let δ > 0. To show: There is a compact set K ⊂ C such that C

∞

− K ⊂ B

∞

(∞; δ).

Claim: Choose K = B(0; r) where r >

_

4

δ

2

− 1 to obtain

C

∞

− K ⊂ B

∞

(∞; δ).

(Clearly K is compact and r > 0 since δ ≤ 2.)

99

Proof of the claim:

C

∞

− K = C

∞

− B(0; r)

= ¦C − B(0; r)¦ ∪ ¦∞¦

= ¦C − ¦z ∈ C : ¦z¦ ≤ r¦¦ ∪ ¦∞¦

⊂

r>

_

4

δ

2

−1

_

¸

¸

_

¸

¸

_

C −

_

¸

¸

_

¸

¸

_

z ∈ C : ¦z¦ ≤

_

4

δ

2

− 1

_

¸

¸

_

¸

¸

_

_

¸

¸

_

¸

¸

_

∪ ¦∞¦

=

_

¸

¸

_

¸

¸

_

z ∈ C : ¦z¦ >

_

4

δ

2

− 1

_

¸

¸

_

¸

¸

_

∪ ¦∞¦

=

_

¸

¸

_

¸

¸

_

z ∈ C :

_

4

δ

2

− 1 < ¦z¦

_

¸

¸

_

¸

¸

_

∪ ¦∞¦

=

_

¸

¸

_

¸

¸

_

z ∈ C :

2

_

1 + ¦z¦

2

< δ

_

¸

¸

_

¸

¸

_

∪ ¦∞¦

= ¦z ∈ C : d(∞, z) < δ¦ ∪ ¦∞¦

= B

∞

(∞; δ).

Therefore,

C

∞

− K ⊂ B

∞

(∞; δ).

d) Let K ⊂ C where K is compact. To show: There is a number δ > 0 such that B

∞

(∞; δ) ⊂ C

∞

− K. Let

B(0; r) such that B(0; r) ⊃ K where r > 0. Clearly

C

∞

− B(0; r) ⊂ C

∞

− K.

So it suﬃces to show: There is a number δ > 0 such that

B

∞

(∞; δ) ⊂ C

∞

− B(0; r)

for a given r > 0.

Claim: Choose δ ≤

2

√

r

2

+1

to obtain

B

∞

(∞; δ) ⊂ C

∞

− B(0; r)

and hence

B

∞

(∞; δ) ⊂ C

∞

− K.

Proof of the claim:

B

∞

(∞; δ) = ¦z ∈ C : d(∞, z) < δ¦ ∪ ¦∞¦

=

_

¸

¸

_

¸

¸

_

z ∈ C :

2

_

1 + ¦z¦

2

< δ

_

¸

¸

_

¸

¸

_

∪ ¦∞¦

⊂

δ≤

2

√

r

2

+1

_

¸

¸

_

¸

¸

_

z ∈ C :

2

_

1 + ¦z¦

2

<

2

√

r

2

+ 1

_

¸

¸

_

¸

¸

_

∪ ¦∞¦

= ¦z ∈ C : r ≤ ¦z¦¦¦ ∪ ¦∞¦

= ¦C ∪ ¦z ∈ C : ¦z¦ < r¦ ∪ ¦∞¦

= C

∞

− ¦z ∈ C : ¦z¦ < r¦

= C

∞

− B(0; r).

100

Therefore,

B

∞

(∞; δ) ⊂ C

∞

− B(0; r).

Exercise 2. Show that if / ⊂ M(G) is a normal family in C(G, C

∞

) then µ(/) is locally bounded.

Solution. Not available.

7.4 The Riemann Mapping Theorem

Exercise 1. Let G and Ω be open sets in the plane and let f : G → Ω be a continuous function which

is one-one, onto, and such that f

−1

: Ω → G is also continuous (a homeomorphism). Suppose ¦z

n

¦ is

a sequence in G which converges to a point z in ∂G; also suppose that w = lim f (z

n

) exists. Prove that

w ∈ ∂Ω.

Solution. Not available.

Exercise 2. (a) Let G be a region, let a ∈ G and suppose that f : (G − ¦a¦) → C is an analytic function

such that f (G − ¦a¦) = Ω is bounded. Show that f has a removable singularity at z = a. If f is one-one,

show that f (a) ∈ ∂Ω.

(b) Show that there is no one-one analytic function which maps G = ¦z : 0 < ¦z¦ < 1¦ onto an annulus

Ω = ¦z : r < ¦z¦ < R¦ where r > 0.

Solution. Not available.

Exercise 3. Let G be a simply connected region which is not the whole plane and suppose that ¯ z ∈ G

whenever z ∈ G. Let a ∈ G ∩ R and suppose that f : G → D = ¦z : ¦z¦ < 1¦ is a one-one analytic function

with f (a) = 0, f

·

(a) > 0 and f (G) = D. Let G

+

= ¦z ∈ G : Im z > 0¦. Show that f (G

+

) must lie entirely

above or entirely below the real axis.

Solution. Not available.

Exercise 4. Find an analytic function f which maps ¦z : ¦z¦ < 1, Re z > 0¦ onto B(0; 1) in a one-one fashion.

Solution. Not available.

Exercise 5. Let f be analytic on G = ¦z : Re z > 0¦, one-one, with Re f (z) > 0 for all z in G, and f (a) = a

for some real number a. Show that ¦ f

·

(a)¦ ≤ 1.

Solution. Clearly G is a simply connected region (not the whole plane). Let a ∈ G. Then, there is a unique

analytic function g : G → D having the properties

a) g(a) = 0 and g

·

(a) > 0

b) g is 1-1

c) g(G) = D

(by the Riemann Mapping Theorem). Since, we have f : G

1−1

−−−−−−→ G (since Re f (z) > 0) and g :

1−1

−−−−−−→

onto

D so g

−1

: D

1−1

−−−−−−→

onto

G we can deﬁne h(z) = g( f (g

−1

(z))). Clearly h(z) is analytic and one-

one, since f and g are analytic and one-one, and we have

h(D) = D

by construction. We have

h(0) = g( f (g

−1

(0))) =

g

−1

(0)=a

g( f (a)) =

f (a)=a

g(a) = 0.

101

We also have

¦h(z)¦ ≤ 1 ∀z ∈ D

since h : D → D. Thus, the hypothesis of Schwarz’s Lemma are satisﬁed, and hence, we get

¦h

·

(0)¦ ≤ 1.

We have

h

·

(z) =

_

g( f (g

−1

(z)))

_

·

= g

·

( f (g

−1

(z))) f

·

(g

−1

(z))

_

g

−1

(z)

_

·

= g

·

( f (g

−1

(z))) f

·

(g

−1

(z))

1

g

·

(g

−1

(z))

where the last step follows from Proposition 2.20 provided g

·

(g

−1

(z)) 0. So

h

·

(0) = g

·

( f (g

−1

(0))) f

·

(g

−1

(0))

1

g

·

(g

−1

(0))

= g

·

( f (a)) f

·

(a)

1

g

·

(a)

= g

·

(a) f

·

(a)

1

g

·

(a)

= f

·

(a)

(g

·

(a) > 0 by assumption). Therefore

¦h

·

(0)¦ = ¦ f

·

(a)¦ ≤ 1.

Thus, we have shown that

¦ f

·

(a)¦ ≤ 1.

Exercise 6. Let G

1

and G

2

be simply connected regions neither of which is the whole plane. Let f be a

one-one analytic mapping of G

1

onto G

2

. Let a ∈ G

1

and put α = f (a). Prove that for any one-one analytic

map h of G

1

into G

2

with h(a) = α it follows that ¦h

·

(a)¦ ≤ ¦ f

·

(a)¦. Suppose h is not assumed to be one-one;

what can be said?

Solution. Deﬁne the function F(z) = f

−1

(h(z)). This is well-deﬁned since f : G

1

1−1

−−−−−−→

onto

G

2

and h :

G

1

1−1

−−−−−−→ G

2

. Clearly F(z) is analytic since f and h are analytic, F(z) is one-one and F : G

1

→ G

1

by

construction. We have

F(a) = f

−1

(h(a)) = f

−1

(α) = a.

Thus by Exercise 5, we get

¦F

·

(a)¦ ≤ 1.

We have

F

·

(z) =

_

f

−1

(h(z))

_

·

= f

·

(h(z))h

·

(z) =

1

f

·

( f

−1

(h(z)))

h

·

(z)

where the last step follows by Proposition 2.20 provided f

·

( f

−1

(h(z))) 0. So

F

·

(a) =

1

f

·

( f

−1

(h(a)))

h

·

(a) =

h

·

(a)

f

·

( f

−1

(α))

=

h

·

(a)

f

·

(a)

which is well-deﬁnes since f

·

(a) 0 by assumption (f is one-one). Since

¦F

·

(a)¦ =

¦h

·

(a)¦

¦ f

·

(a)¦

≤ 1

102

gives

¦h

·

(a)¦ ≤ ¦ f

·

(a)¦.

If h is not one-one, then we get the same result, since I do not need the assumption that F(z) has to be

one-one for exercise 5.

Exercise 7. Let G be a simply connected region and suppose that G is not the whole plane. Let ∆ = ¦ξ :

¦ξ¦ < 1¦ and suppose that f is an analytic, one-one map of G onto ∆ with f (a) = 0 and f

·

(a) > 0 for some

point a in G. Let g be any other analytic, one-one map of G onto ∆ and express g in terms of f .

Solution. Let a ∈ G such that f (a) = 0 and f

·

(a) > 0. Since g is one-one mapping from G onto ∆, we

clearly have g(a) = α, where α ∈ ∆ (α does not have to be zero). Next, deﬁne

g

1

(z) =

z − α

1 − ¯ αz

.

Clearly, g

1

is a Möbius transformation mapping ∆ to ∆ and in addition g

1

is one-one and analytic. Now

deﬁne

g

2

(z) = g

1

(g(z)).

Then, g

2

is one-one and analytic and maps G onto ∆. In addition, we have

g

2

(a) = g

1

(g(a)) = g

1

(α) = 0. (7.10)

But g

·

2

(a) need not necessarily satisfy g

·

2

(a) > 0. However, g

·

2

(a) 0 ∀z ∈ G since g

2

is one-one. Therefore,

deﬁne the rotation

g

3

(z) = e

iθ

z

for θ ∈ [0, 2π). Clearly g

3

maps ∆ onto ∆. g

3

is one-one and analytic. Finally, let

h(z) = g

3

(g

2

(z)).

Also h is one-one, analytic and maps G onto ∆. We have

h(a) = g

3

(g

2

(a)) =

(7.10)

g

3

(0) = 0.

And,

h

·

(z) = (g

3

(g

2

(z)))

·

= (g

3

(g

1

(g(z))))

·

= g

·

3

(g

1

(g(z)))g

·

1

(g(z))g

·

(z)

= e

iθ

(1 − ¯ αg(z)) 1 − (g(z) − α) (−¯ α)

(1 − ¯ αg(z))

2

g

·

(z).

So,

h

·

(a) = e

iθ

(1 − ¯ αα) − 0

(1 − ¯ αα)

2

g

·

(a) = e

iθ

1 − ¦α¦

2

(1 − ¦α¦

2

)

2

g

·

(a) = e

iθ

1

1 − ¦α¦

2

g

·

(a).

Clearly, we can choose a θ ∈ [0, 2π) such that h

·

(a) > 0 (depending on the sign of g

·

(a). Recall g

·

(a) 0

since g is one-one and α ∈ ∆, so

1

1−¦α¦

2

> 0). For example: If g

·

(a) < 0, then pick θ = π, so e

iθ

= −1 or

θ = −Arg(g

·

(a)) will work, too.

By uniqueness of the map f ( f satisﬁes the properties of the Riemann Mapping Theorem), we have

f (z) = h(z)

103

since h satisﬁes the properties of the Riemann Mapping Theorem, too. Thus,

f (z) = h(z) = g

3

(g

2

(z)) = e

iθ

g

2

(z) = e

iθ

g

1

(g(z)) = e

iθ

g(z) − α

1 − ¯ αg(z)

.

So

1 − ¯ αg(z) = e

iθ

g(z) − αe

iθ

⇐⇒ f (z) − ¯ αg(z) f (z) = e

iθ

g(z) − αe

iθ

⇐⇒ (e

iθ

+ ¯ αf (z))g(z) = f (z) + αe

iθ

⇐⇒ g(z) =

f (z) + αe

iθ

e

iθ

+ ¯ αf (z)

.

Exercise 8. Let r

1

, r

2

, R

1

, R

2

, be positive numbers such that R

1

/r

1

= R

2

/r

2

; show that ann(0; r

1

, R

1

) and

ann(0; r

2

, R

2

) are conformally equivalent.

Solution. The function f : C → C; f (z) =

r

2

r

1

z is analytic and it maps the annulus ann(0; r

1

, R

1

) bijectively

into ann(0; r

2

, R

2

). The inner boundary ¦z : ¦z¦ = r

1

¦ of the domain is mapped to the inner boundary of

ann(0; r

2

, R

2

). For the outer boundary note that by assumption R

1

=

R

2

r

1

r

2

, so if ¦z¦ = R

1

then ¦ f (z)¦ =

r

2

r

1

R

1

=

R

2

. The inverse function is f

−1

(z) =

r

1

r

2

z, an analytic and non-constant function also. Therefore the result is

proved.

Exercise 9. Show that there is an analytic function f deﬁned on G =ann(0; 0, 1) such that f

·

never vanishes

and f (G) = B(0; 1).

Solution. Not available.

7.5 The Weierstrass Factorization Theorem

Exercise 1. Show that

(1 + z

n

) converges absolutely iﬀ

(1 + ¦z

n

¦) converges.

Solution. Assume Re(z

n

) > −1. Then the product

(1 + z

n

) converges absolutely iﬀ

_

∞

k=1

z

n

converges

absolutely (by Corollary 5.6 p. 166 applied to w

n

= 1 + z

n

).

_

∞

k=1

z

n

converges absolutely iﬀ

_

∞

n=1

¦z

n

¦

converges (by deﬁnition).

Claim:

_

∞

n=1

¦z

n

¦ converges iﬀ

_

∞

n=1

log(1 + ¦z

n

¦) converges.

Then

_

∞

n=1

log(1+¦z

n

¦) converges iﬀ

(1+¦z

n

¦) converges (by Proposition 5.2 p.165 applied to w

n

= 1+¦z

n

¦).

Thus, ones we have proved the claim, we obtain:

(1 + z

n

) converges absolutely iﬀ

(1 + ¦z

n

¦) converges.

Proof of the claim: Let x

n

= ¦z

n

¦ for convenience. Clearly x

n

≥ 0 ∀n ∈ N.

⇒: Assume

_

¦z

n

¦ converges, that is

_

x

n

converges. Therefore, we have x

n

→ 0. So given > 0, we have

0 ≤ log(1 + ¦x

n

¦) ≤ (1 + )x

n

for suﬃciently large n where the last inequality follows since lim

z→0

log(1+z)

z

= 1 by Exercise 2. By the

comparison test,

_

log(1 + x

n

) converges, that is

_

log(1 + ¦z

n

¦) converges.

⇐: Assume

_

log(1 + ¦z

n

¦) converges, that is

_

log(1 + x

n

) converges. Therefore, we have log(1 + x

n

) → 0

and therefore x

n

→ 0. So given > 0, we have

(1 − )x

n

≤ log(1 + x

n

)

for suﬃciently large n where the last inequality follows since lim

z→0

log(1+z)

z

= 1 by Exercise 2. By the

comparison test,

_

x

n

converges, that is

_

¦z

n

¦ converges.

Therefore,

_

∞

n=1

¦z

n

¦ converges iﬀ

_

∞

n=1

log(1 + ¦z

n

¦) converges.

104

Exercise 2. Prove that lim

z→0

log(1+z)

z

= 1.

Solution. First note that,

f (z) =

log(1 + z)

z

has a removable singularity at z = 0, since

lim

z→0

z f (z) = lim

z→0

z

log(1 + z)

z

= lim

z→0

log(1 + z) = log(1) = 0

(by Theorem 1.2 Chapter 5 p. 103). We know

log(1 + z) =

p.165

∞

k=1

(−1)

k+1

z

k

k

if ¦z¦ < 1. Therefore,

f (z) =

log(1 + z)

z

=

_

∞

k=1

(−1)

k+1 z

k

k

z

=

∞

k=1

(−1)

k+1

z

k−1

k

= 1 +

∞

k=2

(−1)

k+2

z

k−1

k

= 1 +

∞

k=1

(−1)

k

z

k

k + 1

where the sum is an analytic function. Hence,

lim

z→0

f (z) = lim

z→0

_

¸

¸

¸

¸

¸

_

1 +

∞

k=1

(−1)

k

z

k

k + 1

_

¸

¸

¸

¸

¸

_

= 1 +

∞

k=1

(−1)

k

(lim

z→0

z)

k

k + 1

= 1 +

∞

k=1

0 = 1

and therefore

lim

z→0

log(1 + z)

z

= 1.

Exercise 3. Let f and g be analytic functions on a region G and show that there are analytic functions

f

1

, g

1

, and h on G such that f (z) = h(z) f

1

(z) and g(z) = h(z)g

1

(z) for all z in G; and f

1

and g

1

have no

common zeros.

Solution. Let f and g be analytic functions on the region G. Let ¦a

j

¦ be the zeros of f with multiplicity n

j

.

Let ¦b

j

¦ be the zeros of g with multiplicity ˜ n

j

.

If f and g have no common zeros, let h = 1. Then f

1

= f and g

1

= g. Clearly f

1

, g

1

and h are analytic on

G such that f (z) = h(z) f

1

(z) and g(z) = h(z)g

1

(z) for all z in G and f

1

and g

1

have no common zeros.

Otherwise let ¦z

j

¦ be the set of common zeros of f and g with multiplicity m

j

= min(n

j

, ˜ n

j

). Then ¦z

j

¦ form

a sequence of distinct points in G with no limit points on G (since the zeros are isolated). Then there is an

analytic function h deﬁned on G whose only zeros are at the points z

j

with multiplicity m

j

(by Theorem 5.15

p. 170). Let f

1

=

f

h

and g

1

=

g

h

, then clearly f

1

and g

1

are analytic functions on G such that f (z) = h(z) f

1

(z)

and g(z) = h(z)g

1

(z) for all z in G. (since f

1

and g

1

have removable singularities at the z

j

’s by construction;

the multiplicity of z

j

’s in f and g is greater or equal than the multiplicity of z

j

’s in h). In addition, we have

that f

1

and g

1

have no common zeros by construction.

Exercise 4. (a) Let 0 < ¦a¦ < 1 and ¦z¦ ≤ r < 1; show that

¸

¸

¸

¸

¸

a + ¦a¦z

(1 − ¯ az)a

¸

¸

¸

¸

¸

≤

1 + r

1 − r

(b) Let ¦a

n

¦ be a sequence of complex numbers with 0 < ¦a

n

¦ < 1 and

_

(1 −¦a

n

¦) < ∞. Show that the inﬁnite

product

B(z) =

∞

_

n=1

¦a

n

¦

a

n

_

a

n

− z

1 − ¯ a

n

z

_

105

converges in H(B(0; 1)) and that ¦B(z)¦ ≤ 1. What are the zeros of B? (B(z) is called a Blaschke Product.)

(c) Find a sequence ¦a

n

¦ in B(0; 1) such that

_

(1 − ¦a

n

¦) < ∞and every number e

iθ

is a limit point of ¦a

n

¦.

Solution. Not available.

Exercise 5. Discuss the convergence of the inﬁnite product

∞

n=1

1

n

p

for p > 0.

Solution. Not available.

Exercise 6. Discuss the convergence of the inﬁnite products

_

1 +

i

n

_

and

¸

¸

¸1 +

i

n

¸

¸

¸.

Solution. First consider

∞

n=1

_

1 +

i

n

_

. We claim that this product does not converge.

Let z

n

:= 1 +

i

n

= r

n

e

iθ

n

with r

n

=

_

1 +

1

n

2

and θ = arctan

_

1

n

_

. If there is z = re

iθ

such that

∞

n=1

z

n

= z,

then

∞

n=1

r

n

= r and

_

∞

n=1

θ

n

= θ.

Since

arctan(x)

x

≥

1

2

for all values x ∈ (0, 1] the estimate

∞

n=1

θ

n

=

∞

n=1

arctan

_

1

n

_

≥

1

2

∞

n

1

1

n

is valid. The right hand side does not converge, therefore the angles do not converge and there is no z ∈ C

with the desired properties (note that r

n

> 1 for all n, thus z 0).

The second inﬁnite product is a product of real numbers and z

n

= ¦1 +

i

n

¦ =

_

1 +

1

n

2

. The estimate

1 +

1

n

2

≤

_

1 +

1

n

2

_

2

implies that 1 ≤ z

n

≤ 1 +

1

n

2

. Thus for any integer N ∈ N

N

_

n=1

1 ≤

N

_

n=1

z

n

≤

N

_

n=1

_

1 +

1

n

2

_

where the last term is absolutely convergent by an example in class. Therefore the sequence of the partial

products

N

n=1

z

n

is an increasing, bounded sequence that converges in the real numbers. With log(z

n

) > 0

for all n, the comparison with the sum of logarithms (Proposition 5.4) gives absolute convergence.

Exercise 7. Show that

∞

n=2

_

1 −

1

n

2

_

=

1

2

.

Solution. Consider

P

n

=

n

_

k=2

_

1 −

1

k

2

_

=

n

_

k=2

_

1 −

1

k

_ _

1 +

1

k

_

=

n

_

k=2

_

k − 1

k

_ _

k + 1

k

_

= exp

_

¸

¸

_

¸

¸

_

log

n

_

k=2

_

k − 1

k

_ _

k + 1

k

_

_

¸

¸

_

¸

¸

_

= exp

_

¸

¸

_

¸

¸

_

n

k=2

log

__

k − 1

k

_ _

k + 1

k

__

_

¸

¸

_

¸

¸

_

= exp

_

¸

¸

_

¸

¸

_

n

k=2

_

log(k − 1) − log(k)

_

_

¸

¸

_

¸

¸

_

+ exp

_

¸

¸

_

¸

¸

_

n

k=2

_

log(k + 1) − log(k)

_

_

¸

¸

_

¸

¸

_

= exp

__

log(1) − log(n)

__

+ exp

__

−log(2) + log(n + 1)

__

= exp

__

−log(2n) + log(n + 1)

__

= exp

_

log

_

n + 1

2n

__

=

n + 1

2n

.

So,

∞

_

n=2

_

1 −

1

n

2

_

= lim

n→∞

P

n

= lim

n→∞

n + 1

2n

=

1

2

.

106

Exercise 8. For which values of z do the products

(1 − z

n

) and

(1 + z

2n

) converge? Is there an open

set G such that the product converges uniformly on each compact subset of G? If so, give the largest such

open.

Solution. Consider the inﬁnite product

∞

n=1

(1 − z

n

). By deﬁnition and Corollary 5.6 this converges abso-

lutely if and only if

_

∞

n=1

−z

n

converges absolutely. From Calculus it is known that the convergence is also

uniform for ¦z¦ ≤ r < 1 with r ∈ [0, 1).

If z is an n

th

root of unity for a positive integer n, then one of the factors in the inﬁnite product becomes 0

and the product converges to 0, but not absolutely according to Deﬁnition 5.5. If there is no n such that z

is n

th

root of unity then no factor of the inﬁnite product equals zero and there is a subsequence ¦(1 − z

n

k

¦

k

such that ¦1 − z

n

k

¦ > 1.5. We conclude that

∞

n=1

(1 + z

n

) does not converge.

With the denseness of

_

n∈N

¦z ∈ C : z is n

th

root of unity¦ in the unit sphere it follows that the maximal set G

such that the convergence is uniform on compact subsets is the unit disk.

The behavior of

∞

n=1

(1 + z

2n

) is almost the same as the previous one. Substitute w := z

2

, then the conver-

gence is uniform for ¦w¦ ≤ r < 1, so ¦z¦ ≤ sqrtr < 1. Also if ¦z¦ -and hence ¦w¦- is greater than 1, then the

inﬁnite product diverges. Similar to the previous case one factor equals 0 if z is a 2n

th

root of −1.

The maximal region G such that the convergence is uniform on compact sets is again the open unit disk.

Exercise 9. Use Theorem 5.15 to show there is an analytic function f on D = ¦z : ¦z¦ < 1¦ which is not

analytic on any open set G which properly contains D.

Solution. Not available.

Exercise 10. Suppose G is an open set and ¦ f

n

¦ is a sequence in H(G) such that f (z) =

f

n

(z) converges

in H(G). (a) Show that

∞

k=1

_

¸

¸

¸

¸

¸

¸

_

f

·

k

(z)

_

nk

f

n

(z)

_

¸

¸

¸

¸

¸

¸

_

converges in H(G) and equals f

·

(z). (b) Assume that f is not the identically zero function and let K be a

compact subset of G such that f (z) 0 for all z in K. Show that

f

·

(z)

f (z)

=

∞

n=1

f

·

n

(z)

f

n

(z)

and the convergence is uniform over K.

Solution. Not available.

Exercise 11. A subset _ of H(G), G a region, is an ideal iﬀ: (i) f and g in _ implies af + bg is in _ for

all complex numbers a and b; (ii) f in _ and g any function in H(G) implies f g is in _. _ is called a

proper ideal if _ (0) and _ H(G); _ is a maximal ideal if _ is a proper ideal and whenever [ is an

ideal with _ ⊂ [ then either [ = _ or [ = H(G); _ is a prime ideal if whenever f and g ∈ H(G) and

f g ∈ _ then either f ∈ _ or g ∈ _. If f ∈ H(G) let ¸( f ) be the set of zeros of f counted according to

their multiplicity. So ¸((z − a)

3

) = ¦a, a, a¦. If S ⊂ H(G) then ¸(S) = ∩¦¸( f ) : f ∈ S¦, where the zeros

are again counted according to their multiplicity. So if S = ¦(z − a)

3

(z − b), (z − a)

2

¦ then ¸(S) = ¦a, a¦.

(a) If f and g ∈ H(G) then f divides g (in symbols, f ¦g if there is an h in H(G) such that g = f h. Show that

f ¦g iﬀ ¸ ⊂ ¸(g).

(b) If S ⊂ H(G) and S contains a non-zero function then f is a greatest common divisor of S if: (i) f ¦g for

each g in S and (ii) whenever h¦g for each g in S, h¦ f . In symbols, f = g.c.d.S. Prove that f = g.c.d.S. iﬀ

¸( f ) = ¸(S) and show that each non-empty subset of H(G) has a g.c.d.

(c) If A ⊂ G let _(a) = ¦ f ∈ H(G) : ¸( f ) ⊃ A¦. Show that _(A) is a closed ideal in H(G) and _(A) = (0)

107

iﬀ A has a limit point in G.

(d) Let a ∈ G and _ = _(¦a¦). Show that _ is a maximal ideal.

(e) Show that every maximal ideal in H(G) is a prime ideal.

(f) Give an example of an ideal which is not a prime ideal.

Solution. Not available.

Exercise 12. Find an entire function f such that f (n + in) = 0 for every integer n (positive, negative or

zero). Give the most elementary example possible (i.e., choose the p

n

to be as small as possible).

Solution. Not available.

Exercise 13. Find an entire function f such that f (m+ in) = 0 for all possible integers m, n. Find the most

elementary solution possible.

Solution. Not available.

7.6 Factorization of the sine function

Exercise 1. Show that cos πz =

∞

n=1

_

1 −

4z

2

(2n−1)

2

_

.

Solution. We know by the double-angle identity of sine sin(2z) = 2 sin(z) cos(z) (this is proved easily by

using the deﬁnition) or sin(2πz) = 2 sin(πz) cos(πz). Since, we know sin(πz) = πz

∞

n=1

_

1 −

z

2

n

2

_

, we obtain

sin(2πz) = 2 sin(πz) cos(πz)

⇐⇒ 2πz

∞

_

n=1

_

1 −

4z

2

n

2

_

= 2πz

∞

_

n=1

_

1 −

z

2

n

2

_

cos(πz)

⇐⇒ 2πz

∞

_

m=1

_

1 −

4z

2

(2m)

2

_

∞

_

m=1

_

1 −

4z

2

(2m − 1)

2

_

= 2πz

∞

_

n=1

_

1 −

z

2

n

2

_

cos(πz)

where the last statement follows by splitting the product into a product of the even and odd terms (rear-

rangement of the terms is allowed). Hence

2πz

∞

_

n=1

_

1 −

4z

2

(2n)

2

_

∞

_

n=1

_

1 −

4z

2

(2n − 1)

2

_

= 2πz

∞

_

n=1

_

1 −

z

2

n

2

_

cos(πz)

⇐⇒ 2πz

∞

_

n=1

_

1 −

z

2

n

2

_

∞

_

n=1

_

1 −

4z

2

(2n − 1)

2

_

= 2πz

∞

_

n=1

_

1 −

z

2

n

2

_

cos(πz).

Thus,

cos(πz) =

∞

_

n=1

_

1 −

4z

2

(2n − 1)

2

_

.

Exercise 2. Find a factorization for sinh z and cosh z.

108

Solution. We know

sinh(z) =

e

z

− e

−z

2

=

e

−i(iz)

− e

−(−i)(iz)

2

=

−e

i(iz)

+ e

−i(iz)

2

= i

−e

i(iz)

+ e

−i(iz)

2i

= −i

e

i(iz)

− e

−i(iz)

2i

= −i sin(iz)

= −i(iz)

∞

_

n=1

_

1 −

(iz)

2

n

2

π

2

_

= z

∞

_

n=1

_

1 +

z

2

n

2

π

2

_

since by p. 175 Equation 6.2

sin(πz) = πz

∞

_

n=1

_

1 −

z

2

n

2

_

⇒ sin(z) = z

∞

_

n=1

_

1 −

z

2

n

2

π

2

_

.

Therefore,

sinh(z) = z

∞

_

n=1

_

1 +

z

2

n

2

π

2

_

.

We have

cosh(z) =

e

z

+ e

−z

2

=

e

−i(iz)

+ e

−(−i)(iz)

2

=

e

i(iz)

+ e

−i(iz)

2

= cos(iz)

=

∞

_

n=1

_

1 −

4(iz)

2

(2n − 1)

2

π

2

_

=

∞

_

n=1

_

1 +

4z

2

(2n − 1)

2

π

2

_

since by Exercise 1 p. 176

cos(πz) =

∞

_

n=1

_

1 −

4z

2

(2n − 1)

2

_

⇒ cos(z) =

∞

_

n=1

_

1 −

4z

2

(2n − 1)

2

π

2

_

.

Therefore,

cosh(z) =

∞

_

n=1

_

1 +

4z

2

(2n − 1)

2

π

2

_

.

Exercise 3. Find a factorization of the function cos

_

πz

4

_

− sin

_

πz

4

_

.

Solution. Not available.

Exercise 4. Prove Wallis’s formula:

π

2

=

∞

n=1

(2n)

2

(2n−1)(2n+1)

.

Solution. Not available.

109

7.7 The gamma function

Exercise 1. Show that 0 < γ < 1. (An approximation to γ is .57722. It is unknown whether γ is rational or

irrational.)

Solution. Not available.

Exercise 2. Show that Γ(z)Γ(1 − z) = π csc πz for z not an integer. Deduce from this that Γ(

1

2

) =

√

π.

Solution. Assume z is not an integer, then by Gauss’s Formula p. 178 we get

Γ(z)Γ(1 − z) = lim

n→∞

n!n

z

z(z + 1) (z + n)

lim

m→∞

m!m

1−z

(1 − z)(2 − z) (m + 1 − z)

= lim

n→∞

n!n!n

z+1−z

z(1 + z)(1 − z)(2 + z)(2 − z) (n + z)(n − z)(n + 1 − z)

= lim

n→∞

(n!)

2

n

z(1

2

− z

2

)(2

2

− z

2

) (n

2

− z

2

)(n + 1 − z)

= lim

n→∞

(n!)

2

n

z(n!n!)

_

1 −

z

2

1

2

_ _

1 −

z

2

2

2

_ _

1 −

z

2

2

3

_

_

1 −

z

2

n

2

_

(n + 1 − z)

= lim

n→∞

1

z

_

1 −

z

2

1

2

_ _

1 −

z

2

2

2

_ _

1 −

z

2

3

2

_

_

1 −

z

2

n

2

_

n+1−z

n

=

1

z lim

n→∞

_

1 −

z

2

1

2

_ _

1 −

z

2

2

2

_ _

1 −

z

2

3

2

_

_

1 −

z

2

n

2

_

lim

n→∞

n + 1 − z

n

.,,.

=1

=

1

z

∞

n=1

_

1 −

z

2

n

2

_ =

π

πz

∞

n=1

_

1 −

z

2

n

2

_

=

π

sin(πz)

= π csc(πz)

where the step before the last step follows by p. 175 Equation 6.2. So

Γ(z)Γ(1 − z) = π csc(πz)

for z not an integer. Now, let z =

1

2

, then

Γ

_

1

2

_

Γ

_

1 −

1

2

_

= π csc

_

π

2

_

which implies

Γ

_

1

2

_

=

√

π 1 =

√

π.

Exercise 3. Show:

√

πΓ(2z) = 2

2z−1

Γ(z)Γ(z +

1

2

). (Hint: Consider the function Γ(z)Γ(z +

1

2

)Γ(2z)

−1

.)

Solution. Following the hint deﬁne a function f on its domain G = ¦z ∈ C : z (Z − N)¦ by

f (z) =

2

z−1

√

π

Γ

_

z

2

_

Γ

_

z

2

+

1

2

_

.

110

Let x be a positive real number. It suﬃces to show that f (x) satisﬁes the Bohr-Mollerup Theorem. By the

logarithmic convexity of Γ is follows that

log f (x) = (x − 1) log 2 − log

√

π + log Γ

_

x

2

_

+ log Γ

_

x

2

+

1

2

_

is a convex function. Also f satisﬁes the functional equation of the Gamma-function:

f (x + 1) =

2

x

√

π

Γ

_

x

2

+

1

2

_

Γ

_

x

2

+ 1

_

=

2

x

√

π

Γ

_

x

2

+

1

2

_

x

2

Γ

_

x

2

_

=

x2

x−1

√

π

Γ

_

x

2

+

1

2

_

Γ

_

x

2

_

= x f (x).

Lastly, also f (1) =

Γ(

1

2

)

√

π

= 1; thus by Theorem 7.13, f agrees with Γ on the positive real numbers. With the

Identity Theorem then f agrees with Γ on the whole domain G.

Exercise 4. Show that log Γ(z) is deﬁned for z in C − (−∞, 0] and that

log Γ(z) = −log z − γz −

∞

n=1

_

log

_

1 +

z

n

_

−

z

n

_

.

Solution. Not available.

Exercise 5. Let f be analytic on the right half plane Re z > 0 and satisfy: f (1) = 1, f (z + 1) = z f (z), and

lim

n→∞

f (z+n)

n

z

f (n)

= 1 for all z. Show that f = Γ.

Solution. Not available.

Exercise 6. Show that

Γ(z) =

∞

n=0

(−1)

n

n!(z + n)

+

_

∞

1

e

−t

t

z−1

dt

for z 0, −1, −2, . . . (not for Re z > 0 alone).

Solution. Write Γ(z) = Φ(z) + Ψ(z) where

Φ(z) =

_

1

0

e

−t

t

z−1

dt

and

Ψ(z) =

_

∞

1

e

−t

t

z−1

dt. (7.11)

We can write

Φ(z) =

_

1

0

e

−t

t

z−1

dt =

_

1

0

∞

n=0

(−t)

n

n!

t

z−1

dt =

∞

n=0

(−1)

n

n!

_

1

0

t

n

t

z−1

dt

=

∞

n=0

(−1)

n

n!

_

t

n+z

n + z

_

1

0

=

∞

n=0

(−1)

n

n!(z + n)

111

Thus,

Γ(z) =

∞

n=0

(−1)

n

n!(z + n)

+

_

∞

1

e

−t

t

z−1

dt (7.12)

Claim 1: Γ(z) given by (7.12) is the analytic continuation of (7.11), that is Γ(z) given by (7.12) is deﬁned

for all z ∈ C − ¦0, −1, −2, . . .¦.

Proof of Claim 1: We know from the book that Ψ(z) is analytic for Re (z) > 0.

Claim 2: Ψ(z) is analytic for Re (z) ≤ 0. Thus Ψ(z) is analytic on C.

Proof of Claim 2: Assume Re (z) ≤ 0. Then

¦t

z−1

¦ = t

Re(z)−1

≤

t∈[1,∞),Re(z)≤0

t

−1

.

But since e

−

1

2

t

t

Re (z)−1

→ 0 as t → ∞, there exists a constant C > 0 such that t

Re(z)−1

≤ Ce

1

2

when t ≥ 1.

Hence, we have

¦e

−t

t

z−1

¦ ≤ ¦e

−t

¦ ¦t

z−1

¦ = e

−t

t

Re(z)−1

≤ e

−t

Ce

1

2

t

= Ce

−

1

2

t

and therefore Ce

−

1

2

t

is integrable on (1, ∞). By Fubini’s Theorem for any ¦γ¦ ⊂ G = ¦z : Re(z) ≤ 0¦,

_

γ

_

∞

1

e

−t

t

z−1

dt dz =

_

∞

1

_

γ

e

−t

t

z−1

dz dt = 0

which implies

_

∞

1

e

−t

t

z−1

dt ∈ H(G).

In summary,

Ψ(z) =

_

∞

1

e

−t

t

z−1

dt ∈ H(C).

Thus, Claim 2 is proved.

It remains to show that

Φ(z) =

∞

n=0

(−1)

n

n!(z + n)

is analytic on C − ¦0, −1, −2, . . .¦. Note that Φ(z) is uniformly and absolutely convergent as a series in any

closed domain which contains none of the points 0, −1, −2, . . . and thus provides the analytic continuation

of Φ(z). Since

∞

n=0

(−1)

n

n!(z + n)

+

_

∞

1

e

−t

t

z−1

dt

is analytic and we know (Theorem 7.15 p. 180)

Γ(z) =

_

∞

0

e

−t

t

z−1

dt

for Re(z) > 0, we get

Γ(z) =

∞

n=0

(−1)

n

n!(z + n)

+

_

∞

1

e

−t

t

z−1

dt

is the analytic continuation of (7.11) for z ∈ C − ¦0, −1, −2, . . .¦.

112

Exercise 7. Show that

_

∞

0

sin(t

2

) dt =

_

∞

0

cos(t

2

) =

1

2

_

1

2

π.

Solution. We have shown that Γ

_

1

2

_

=

√

π (see Exercise 2 p. 185). Thus

√

π = Γ

_

1

2

_

=

_

∞

0

e

−t

t

−

1

2

dt = 2a

_

∞

0

e

−a

2

t

2

dt

(because of Theorem 7.15 p. 180). The last step involves integrating a complex integral. Hence, we have

_

∞

0

e

−a

2

t

2

dt =

√

π

2a

. (7.13)

Let a =

1−i

√

2

, then a

2

=

1−2i+i

2

2

= −i and therefore

_

∞

0

e

−

_

1−i

√

2

_

2

t

2

dt =

_

∞

0

e

−(−i)t

2

dt =

_

∞

0

e

it

2

dt

=

_

∞

0

cos(t

2

) + i sin(t

2

) dt =

_

∞

0

cos(t

2

) dt + i

_

∞

0

sin(t

2

) dt

=

(7.13)

√

π

2

1−i

√

2

=

√

π

√

2

1 + i

(1 − i)(1 + i)

=

1

2

_

1

2

π(1 + i) =

1

2

_

1

2

π + i

1

2

_

1

2

π.

Thus,

_

∞

0

cos(t

2

) dt + i

_

∞

0

sin(t

2

) dt =

1

2

_

1

2

π + i

1

2

_

1

2

π

which implies

_

∞

0

cos(t

2

) dt =

1

2

_

1

2

π

and

_

∞

0

sin(t

2

) =

1

2

_

1

2

π.

Exercise 8. Let u > 0 and v > 0 and express Γ(u)Γ(v) as a double integral over the ﬁrst quadrant of the

plane. By changing to polar coordinates show that

Γ(u)Γ(v) = 2Γ(u + v)

_ π

2

0

(cos θ)

2u−1

(sin θ)

2v−1

dθ.

The function

B(u, v) =

Γ(u)Γ(v)

Γ(u + v)

is called the beta function. By changes of variables show that

B(u, v) =

_

1

0

t

u−1

(1 − t)

v−1

dt =

_

∞

0

t

u−1

(1 + t)

u+v

dt

Can this be generalized to the case when u and v are complex numbers with positive real part?

113

Solution. Let u > 0 and v > 0, then by Theorem 7.15 p. 180 and three changes of variables (indicated by

“CoV”) we obtain

Γ(u)Γ(v) =

_

∞

0

e

−s

s

u−1

ds

_

∞

0

e

−t

t

v−1

dt

=

CoV:s=x

2

,t=y

2

2

_

∞

0

e

−x

2

x

2u−2

x dx 2

_

∞

0

e

−y

2

y

2v−2

y dy

= 4

_

∞

0

_

∞

0

e

−x

2

−y

2

x

2u−1

y

2v−1

dx dy

=

CoV:x=r cos(θ),y=r sin θ

4

_

π/2

0

_

∞

0

e

−r

2

r

2u−1

(cos θ)

2u−1

r

2v−1

(sin θ)

2v−1

r dr dθ

= 4

_

π/2

0

_

∞

0

e

−r

2

r

2u+2v−1

(cos θ)

2u−1

(sin θ)

2v−1

r dr dθ

= 4

_

∞

0

e

−r

2

r

2u+2v−1

dr

_

π/2

0

(cos θ)

2u−1

(sin θ)

2v−1

dθ

=

CoV:t=r

2

4

1

2

_

∞

0

e

−t

t

u+v−1

dt

_

π/2

0

(cos θ)

2u−1

(sin θ)

2v−1

dθ

= 2Γ(u + v)

_

π/2

0

(cos θ)

2u−1

(sin θ)

2v−1

dθ.

This implies

B(u, v) =

Γ(u)Γ(v)

Γ(u + v)

= 2

_

π/2

0

(cos θ)

2u−1

(sin θ)

2v−1

dθ.

Next, we show

B(u, v) =

_

1

0

t

u−1

(1 − t)

v−1

dt.

We have

B(u, v) =

Γ(u)Γ(v)

Γ(u + v)

= 2

_

π/2

0

(cos θ)

2u−1

(sin θ)

2v−1

dθ

=

CoV:t=cos

2

θ

2

_

1

0

_

√

t

_

2u−2

_ √

1 − t

_

2v−2

dt

=

_

1

0

t

u−1

(1 − t)

v−1

dt.

Hence,

B(u, v) =

_

1

0

t

u−1

(1 − t)

v−1

dt.

Finally, we show

B(u, v) =

_

∞

0

t

u−1

(1 + t)

u+v

dt.

114

We have

B(u, v) =

_

1

0

t

u−1

(1 − t)

v−1

dt

=

CoV:s=

t

1−t

_

∞

0

_

s

1 + s

_

u−1

_

1 −

s

1 + s

_

v−1

(1 + s)

−2

ds

=

_

∞

0

s

u−1

1

(1 + s)

u−1

1

(1 + s)

v−1

(1 + s)

−2

ds

=

_

∞

0

s

u−1

1

(1 + s)

u+v−2

(1 + s)

−2

ds

=

_

∞

0

s

u−1

(1 + s)

u+v

ds.

Hence,

B(u, v) =

_

∞

0

t

u−1

(1 + t)

u+v

dt.

Yes, the beta function can be generalized to the case when u and v are complex numbers with positive real

part. We have seen that

B(u, v) =

_

∞

0

t

u−1

(1 + t)

u+v

dt = 2

_

π/2

0

(cos θ)

2u−1

(sin θ)

2v−1

dθ

for u > 0 and v > 0. Thus it remains to show that

B(u, v) =

_

∞

0

t

u−1

(1 + t)

u+v

dt ∈ H(G × G)

where G × G ⊂ C × C with G = ¦z : Re(z) > 0¦.

Let Re(u) > 0 and Re(v) > 0.

_

∞

0

t

u−1

(1 + t)

u+v

dt =

_

1

0

t

u−1

(1 + t)

u+v

dt

.,,.

I

+

_

∞

1

t

u−1

(1 + t)

u+v

dt

.,,.

II

.

I) We have

_

1

0

¦t

u−1

(1 + t)

−u−v

¦ dt ≤

_

1

0

¦t

u−1

¦ ¦(1 + t)

−u−v

¦ dt

≤

_

1

0

t

Re(u)−1

(1 + t)

−Re(u)−Re(v)

dt

≤

_

1

0

t

Re(u)−1

< ∞

where the last step follows since Re(u) > 0 and the previous one since Re(u) > 0 and Re(v) > 0 and

t ∈ (0, 1). This implies ¦t

u−1

(1 + t)

−u−v

¦ is integrable on (0, 1). By Fubini’s Theorem for any ¦γ¦ ⊂ G × G

_

γ

_

1

0

t

u−1

(1 + t)

u+v

dt dz =

_

1

0

_

γ

t

u−1

(1 + t)

u+v

dz dt = 0

115

which implies

_

1

0

t

u−1

(1 + t)

u+v

dt ∈ H(G × G).

II) We have

_

∞

1

¦t

u−1

(1 + t)

−u−v

¦ dt ≤ lim

n→∞

_

n

1

¦t

u−1

(1 + t)

−u−v

¦ dt

≤ lim

n→∞

_

n

1

¦t

u−1

¦ ¦(1 + t)

−u−v

¦ dt

= lim

n→∞

_

n

1

t

Re(u)−1

(1 + t)

−Re(u)−Re(v)

dt

= lim

n→∞

_

n

1

t

Re(u)−1

1

t

Re(u)+Re(v)

_

1

t

+ 1

_

Re(u)+Re(v)

dt

= lim

n→∞

_

n

1

t

−1−Re(u)

1

_

1

t

+ 1

_

Re(u)+Re(v)

dt

≤ lim

n→∞

_

n

1

1

t

1+Re(v)

dt < ∞

where the last step follows since Re(v) > 0 and the previous one since Re(u) + Re(v) > 0, t ≥ 1 and

therefore

1

(

1

t

+1)

Re(u)+Re(v)

≤ 1. This implies ¦t

u−1

(1 + t)

−u−v

¦ is integrable on (1, ∞). By Fubini’s Theorem for

any ¦γ¦ ⊂ G × G

_

γ

_

∞

1

t

u−1

(1 + t)

u+v

dt dz =

_

∞

1

_

γ

t

u−1

(1 + t)

u+v

dz dt = 0

which implies

_

∞

1

t

u−1

(1 + t)

u+v

dt ∈ H(G × G).

In summary

_

∞

0

t

u−1

(1 + t)

u+v

dt ∈ H(G × G).

Exercise 9. Let α

n

be the volume of the ball of radius one in R

n

(n ≥ 1). Prove by induction and iterated

integrals that

α

n

= 2α

n−1

_

1

0

(1 − t

2

)

(n−1)/2

dt

Solution. Deﬁne the n-dimensional ball with radius r by

B

n

(r) = ¦x ∈ R

n

: ¦x¦ ≤ r¦

and deﬁne the volume of B

n

(r) by V

n

(r). Clearly, V

1

(r) = 2r, V

2

(r) = πr

2

, V

3

(r) =

4

3

πr

3

and therefore

V

1

(1) = 2, V

2

(1) and V

3

=

4

3

π.

Claim: For arbitrary n we have

V

n

(r) = r

n

V

n

(1)

116

which implies

V

n

(r) = r

n

V

n

(1) =

(def)

r

n

α

n

(7.14)

Assume the claim is true, then we write

B

n

(1) = ¦x ∈ R

n

: ¦x¦ ≤ 1¦ = ¦(x

1

, x

2

, . . . , x

n−1

, t) ∈ R

n

: x

2

1

+ x

2

2

+ . . . + x

2

n−1

+ t

2

≤ 1¦.

For a ﬁxed t ∈ [−1, 1] we have the balls

B

n−1

_ √

1 − t

2

_

= ¦(x

1

, x

2

, . . . , x

n−1

) ∈ R

n−1

: x

2

1

+ x

2

2

+ . . . + x

2

n−1

≤ 1¦

and by (7.14) we obtain

V

n−1

_ √

1 − t

2

_

=

_ √

1 − t

2

_

n−1

V

n−1

(1) =

_

1 − t

2

_ n−1

2

α

n−1

.

Clearly, the ball B

n

(1) is the union of all disjoint balls B

n−1

_ √

1 − t

2

_

as t varies over [−1, 1]. Hence

α

n

= V

n

(1) =

_

1

−1

_

1 − t

2

_ n−1

2

α

n−1

dt = 2α

n−1

_

1

0

_

1 − t

2

_ n−1

2

dt

which proves the formula above.

Finally, we prove the claim V

n

(r) = r

n

V

n

(1). We have by using a symmetry argument

V

n

(r) =

_

r

−r

_

√

r

2

−x

2

1

−

√

r

2

−x

2

1

_

√

r

2

−x

2

1

−x

2

2

−

√

r

2

−x

2

1

−x

2

2

_

√

r

2

−

_

n−1

i=1

x

2

i

−

√

r

2

−

_

n−1

i=1

x

2

i

1 dx

n

dx

n−1

dx

2

dx

1

= 2

n

_

r

0

_

√

r

2

−x

2

1

0

_

√

r

2

−x

2

1

−x

2

2

0

_

√

r

2

−

_

n−1

i=1

x

2

i

0

1 dx

n

dx

n−1

dx

2

dx

1

=

CoV:x

i

=ry

i

2

n

_

r

0

_

√

1−y

2

1

0

_

√

1−y

2

1

−y

2

2

0

_

√

1−

_

n−1

i=1

y

2

i

0

1 r

n

.,,.

Jacobian

dy

n

dy

n−1

dy

2

dy

1

= r

n

_

1

−1

_

√

1−y

2

1

−

√

1−y

2

1

_

√

1−y

2

1

−y

2

2

−

√

1−y

2

1

−y

2

2

_

√

1−

_

n−1

i=1

y

2

i

−

√

1−

_

n−1

i=1

y

2

i

1 dy

n

dy

n−1

dy

2

dy

1

= r

n

V

n

(1).

This implies V

n

(r) = r

n

V

n

(1). (Note that this is kind of obvious, since we scale the unit ball in R

n

by r in

each dimension to obtain the ball with radius r in R

n

). We could have proved the formula by mathematical

induction. Using spherical coordinates simpliﬁes the computation if preferred.

Exercise 10. Show that

α

n

=

π

n/2

(n/2)Γ(n/2)

where α

n

is deﬁned in problem 9. Show that if n = 2k, k ≥ 1, then α

n

= π

k

/k!

Solution. Not available.

117

Exercise 11. The Gaussian psi function is deﬁned by

Ψ(z) =

Γ

·

(z)

Γ(z)

(a) Show that Ψ is meromorphic in C with simple poles at z = 0, −1, . . . and Res(Ψ; −n) = −1 for n ≥ 0.

(b) Show that Ψ(1) = −γ.

(c) Show that Ψ(z + 1) − Ψ(z) =

1

z

.

(d) Show that Ψ(z) − Ψ(1 − z) = −π cot πz.

(e) State and prove a characterization of Ψ analogous to the Bohr-Mollerup Theorem.

Solution. Not available.

7.8 The Riemann zeta function

Exercise 1. Let ξ(z) = z(z − l)π

−

1

2

z

ζ(z)Γ(

1

2

z) and show that ξ is an entire function which satisﬁes the

functional equation ξ(z) = ξ(1 − z).

Solution. The function ξ is analytic on C − (¦−2k : k ∈ N

0

¦ ∪ ¦1¦) as a product of analytic functions. The

factors z, z − 1 and π

−

1

2

z

are entire. The zeta-function has a simple pole at z = 1, hence lim

z→1

(z − 1)ζ(z)

exists in C and the function ξ is well-deﬁned at z = 1.

The function Γ

_

1

2

z

_

has simple poles at the non-positive even integers. For z = 0, lim

z→0

zΓ

_

1

2

z

_

∈ C. For

negative even integers z = 2k, −k ∈ N the poles of Gamma coincide with the simple zeros of the zeta-

function, hence lim

z→2k

ζ(z)Γ

_

1

2

z

_

exists in C also. Therefore the function ξ is analytic at each point.

The functional equation of ξ is related to the functional equation of ζ (8.3, p. 192), evaluated at 1 − z, i.e.

ζ(1 − z) = 2(2π)

−z

Γ(z)ζ(z) sin

_

1

2

π(1 − z)

_

. (7.15)

Furthermore the proof uses variants of problems 1, 2 of the homework set 5, in particular

Γ

_

1

2

−

z

2

_

Γ

_

1

2

+

z

2

_

= π csc

_

1

2

π(1 − z)

_

(7.16)

√

πΓ(z) = 2

z−1

Γ

_

z

2

_

Γ

_

1

2

+

z

2

_

. (7.17)

118

Then the following chain of equalities holds.

ξ(1 − z) = (1 − z)(−z)π

−

1

2

+

z

2

ζ(1 − z)Γ

_

1

2

−

z

2

_

(7.15)

= z(z − 1)π

−

1

2

+

z

2

Γ

_

1

2

−

z

2

_

2

1−z

π

−z

ζ(z)Γ(z) sin

_

1

2

π(1 − z)

_

= ξ(z)π

−

1

2

2

1−z

Γ(z)Γ

_

1

2

−

z

2

_

Γ

_

z

2

_ sin

_

1

2

π(1 − z)

_

(7.16)

= ξ(z)π

−

1

2

2

1−z

Γ(z)

Γ

_

z

2

_

π csc

_

1

2

π(1 − z)

_

Γ

_

1

2

+

z

2

_ sin

_

1

2

π(1 − z)

_

= ξ(z)π

1

2

2

1−z

Γ(z)

Γ

_

z

2

_

Γ

_

1

2

+

z

2

_

(7.17)

= ξ(z)π

1

2

2

1−z

1

Γ

_

1

2

+

z

2

_

2

z−1

π

1

2

Γ

_

1

2

+

z

2

_

= ξ(z).

Exercise 2. Use Theorem 8.17 to prove that

_

p

−1

n

= ∞. Notice that this implies that there are an inﬁnite

number of primes.

Solution. Euler’s Theorem asserts that for real values x > 1, ζ(x) =

∞

n=1

_

1

1−p

−x

n

_

, where p

n

are the prime

numbers. The left hand side is unbounded as x ¸ 1 and so is the right hand side.

Now the argument goes as follows. Seeking contradiction suppose that

_

∞

n=1

1

p

n

is ﬁnite. All summands

are positive, therefore the sum converges absolutely. By Proposition 5.4, p.165, also

_

∞

n=1

log

_

1 −

1

p

n

_

converges absolutely. But this implies that

−

∞

n=1

log

_

1 −

1

p

n

_

= log

∞

_

n=1

_

1

1 − p

n

_

= log ζ(1) < ∞

which gives the desired contradiction because the log(ζ(z)) has a simple pole at z = 1.

Exercise 3. Prove that ζ

2

(z) =

_

∞

n=1

d(n)

n

z

for Re z > 1, where d(n) is the number of divisors of n.

Solution. Not available.

Exercise 4. Prove that ζ(z)ζ(z − 1) =

_

∞

n=1

σ(n)

n

z

, for Re z > 1, where σ(n) is the sum of the divisors of n.

Solution. Not available.

Exercise 5. Prove that

ζ(z−1)

ζ(z)

=

_

∞

n=1

ϕ(n)

n

z

for Re z > 1, where ϕ(n) is the number of integers less than n and

which are relatively prime to n.

Solution. Not available.

Exercise 6. Prove that

1

ζ(z)

=

_

∞

n=1

µ(n)

n

z

for Re z > 1, where µ(n) is deﬁned as follows. Let n = p

k

1

1

p

k

2

2

p

k

m

m

be the factorization of n into a product of primes p

1

, . . . , p

m

and suppose that these primes are distinct. Let

µ(1) = 1; if k

1

= . . . = k

m

= 1 then let µ(n) = (−1)

m

; otherwise let µ(n) = 0.

Solution. Not available.

119

Exercise 7. Prove that

ζ

·

(z)

ζ(z)

= −

_

∞

n=1

Λ(n)

n

z

for Re z > 1, where Λ(n) = log p if n = p

m

for some prime p and

m ≥ 1; and Λ(n) = 0 otherwise.

Solution. Not available.

Exercise 8. (a) Let η(z) = ζ

·

(z)/ζ(z) for Re z > 1 and show that lim

z→z

0

(z − z

0

)η(z) is always an integer for

Re z

0

≥ 1. Characterize the point z

0

(in its relation to ζ) in terms of the sign of this integer.

(b) Show that for > 0

Re η(1 + + it) = −

∞

n=1

Λ(n)n

−1(1+)

cos(t log n)

where Λ(n) is deﬁned in Exercise 7.

(c) Show that for all > 0,

3Re η(1 + ) + 4Re η(1 + + it) + Re η(1 + + 2it) ≤ 0.

(d) Show that ζ(z) 0 if Re z = 1 (or 0).

Solution. Not available.

120

Chapter 8

Runge’s Theorem

8.1 Runge’s Theorem

Exercise 1. Prove Corollary 1.14 if it is only assumed that E

−

meets each component of C

∞

− G.

Solution. Not available.

Exercise 3. Let G be the open unit disk B(0; 1) and let K = ¦z :

1

4

≤ ¦z¦ ≤

3

4

¦. Show that there is a function

f analytic on some open subset G

1

containing K which cannot be approximated on K by functions in H(G).

Remarks. The next two problems are concerned with the following question. Given a compact set K

contained in an open set G

1

⊂ G, can functions in H(G

1

) be approximated on K by functions in H(G)?

Exercise 2 says that for an arbitrary choice of K, G, and G

1

this is not true. Exercise 4 below gives criteria

for a ﬁxed K and G such that this can be done for any G

1

. Exercise 3 is a lemma which is useful in proving

Exercise 4.

Solution. Not available.

Exercise 3. Let K be a compact subset of the open set G and suppose that any bounded component D of

G − K has D

−

∩ ∂G . Then every component of C

∞

− K contains a component of C

∞

− G.

Solution. Not available.

Exercise 4. Let K be a compact subset of the open set G; then the following are equivalent:

(a) If f is analytic in a neighborhood of K and > 0 then there is a g in H(G) with ¦ f (z) − g(z)¦ < for all

z in K;

(b) If D is a bounded component of G − K then D

−

∩ ∂G ;

(c) If z is any point in G − K then there is a function f in H(G) with

¦ f (z)¦ > sup¦¦ f (w)¦ : w in K¦.

Solution. Not available.

Exercise 5. Can you interpret part (c) of Exercise 4 in terms of

ˆ

K?

Solution. Not available.

121

Exercise 6. Let K be a compact subset of the region G and deﬁne

ˆ

K

G

= ¦z ∈ G : ¦ f (z)¦ ≤ j f j

K

for all f in

H(G)¦.

(a) Show that if C

∞

− G is connected then

ˆ

K

G

=

ˆ

K.

(b) Show that d(K, C − G) = d(

ˆ

K

G

, C − G).

(c) Show that

ˆ

K

G

⊂ the convex hull of K ≡ the intersection of all convex subsets of C which contain K.

(d) If

ˆ

K

G

⊂ G

1

⊂ G and G

1

is open then for every g in H(G

1

) and > 0 there is a function f in H(G) such

that ¦ f (z) − g(z)¦ < for all z in

ˆ

K

G

. (Hint: see Exercise 4.)

(e)

ˆ

K

G

= the union of K and all bounded components of G − K whose closure does not intersect ∂G.

Solution. Not available.

8.2 Simple connectedness

Exercise 1. The set G = ¦re

it

: −∞ < t < 0 and 1 + e

t

< r < 1 + 2e

t

¦ is called a cornucopia. Show that G

is simply connected. Let K = G

−

; is int

ˆ

K connected?

Solution. he result follows once it is established that every closed piecewise smooth curve in G is 0-

homotopic. This requires the following subclaim: let z ∈ G then the argument of z is uniquely deﬁned.

Suppose otherwise then there are real r

1

, r

2

, t

1

, t

2

such that z = r

1

e

it

1

= r

2

e

it

2

. Suppose WLOG that t

1

< t

2

<

0 then there is a positive integer k such that t

1

= 2kπ = t

2

and therefore r

2

> 1+e

t

2

= 1+e

t

1

e

kπ

> 1+2e

t

1

> r

1

contradicting the fact that ¦z¦ = r

1

= r

2

.

Deﬁne

R : (−∞, 0) →G; R(t) = 1 +

3

2

e

t

.

Let γ : [0, 1] → G be an arbitrary closed piecewise smooth curve in G with radius function ρ(s) and

argument function τ(s) such that

γ(s) = ρ(s)e

iτ(s)

,

and deﬁne

Γ : [0, 1] × [0, 1] →G,

Γ(s, 0) = γ(s),

Γ(s, 1) = R(τ(s)),

Γ(0, u) = Γ(1, u) ∀ u ∈ [0, 1].

Also deﬁne a function

Θ : [0, 1] × [0, 1] →G,

Θ(s, 0) = R(τ(s)),

Θ(s, 1) = R

_

τ

_

1

2

__

,

Θ(0, u) = Θ(1, u) ∀ u ∈ [0, 1].

The function Θ ◦ Γ satisﬁes Deﬁnition IV 6.1 (p.88), the curve γ is homotopic to the point R

_

τ

_

1

2

__

and by

Deﬁnition IV 6.14 (p.93) the set G is simply connected.

Exercise 2. If K is polynomially convex, showthat the components of the interior of K are simply connected.

Solution. Not available.

122

8.3 Mittag-Leﬄer’s Theorem

Exercise 1. Let G be a region and let ¦a

n

¦ and ¦b

m

¦ be two sequences of distinct points in G without limit

points in G such that an a

n

b

m

for all n, m. Let S

n

(z) be a singular part at a

n

and let p

m

be a positive

integer. Show that there is a meromorphic function f on G whose only poles and zeros are ¦a

n

¦ and ¦b

m

¦

respectively, the singular part at z = a

n

is S

n

(z), and z = b

m

is a zero of multiplicity p

m

.

Solution. Let G be a region and let ¦b

m

¦ be a sequence of distinct points in G with no limit point in G;

and let ¦p

m

¦ be a sequence of integers. By Theorem 5.15 p.170 there is an analytic function g deﬁned on G

whose only zeros are at the points b

m

; furthermore, b

m

is a zero of g of multiplicity p

m

.

Since g ∈ H(g) and ¦a

n

¦ ∈ G, g has a Taylor series in a neighborhood B(a

n

; R

n

) of each a

n

, that is

g

n

(z) =

∞

k=0

α

k

(z − a

n

)

k

∈ B(a

n

; R

n

)

where α

k

=

1

k!

g

(k)

(a

n

). Goal: Try to use this series to create a singular part r

n

(z) at a

n

such that r

n

(z)g

n

(z) =

s

n

(z) or

r

n

(z)

∞

k=0

α

k

(z − a

n

)

k

=

m

n

j=1

A

jn

(z − a

n

)

j

⇐⇒

m

n

j=1

A

jn

(z − a

n

)

j

=

∞

k=0

α

k

r

n

(z)(z − a

n

)

k

.

Claim:

r

n

(z) =

m

n

j=1

B

jk

(z − a

n

)

−j−k

(8.1)

works.

Proof of the claim:

∞

k=0

α

k

r

n

(z)(z − a

n

)

k

=

∞

k=0

α

k

m

n

j=1

B

jk

(z − a

n

)

−j−k

(z − a

n

)

k

=

∞

k=0

α

k

m

n

j=1

B

jk

(z − a

n

)

−j

=

m

n

j=1

(z − a

n

)

−j

∞

k=0

α

k

B

jk

=

m

n

j=1

A

jn

(z − a

n

)

j

where the last step follows by choosing

∞

k=0

α

k

B

jk

= A

jn

.

Since G is a region, G is open. Let ¦a

n

¦ be a sequence of distinct points without a limit point in G and such

that a

n

b

m

for all n, m. Let ¦r

n

(z)¦ be the sequence of rational functions given by

r

n

(z) =

m

n

j=1

B

jk

(z − a

n

)

j+k

(see (8.1)). By Mittag-Leﬄer’s Theorem, there is a meromorphic function h on G whose poles are exactly

the points ¦a

n

¦ and such that the singular part of h at a

n

is r

n

(z).

Set f = g h. Then by construction f is the meromorphic function on G whose only poles and zeros are ¦a

n

¦

and ¦b

m

¦ respectively, the singular part at z = a

n

is S

n

(z), and z = b

m

is a zero of multiplicity p

m

. (Note that

the zeros do not cancel the poles since by assumption a

n

b

m

∀ n, m).

123

Exercise 2. Let ¦a

n

¦ be a sequence of points in the plane such that ¦a

n

¦ → ∞, and let ¦b

n

¦ be an arbitrary

sequence of complex numbers.

(a) Show that if integers ¦k

n

¦ can be chosen such that

∞

k=n

_

r

a

n

_

k

n

b

n

a

n

(3.4)

converges absolutely for all r > 0 then

∞

k=n

_

r

a

n

_

k

n

b

n

z − a

n

(3.5)

converges in M(C) to a function f with poles at each point z = a

n

.

(b) Show that if limsup ¦b

n

¦ < ∞then (3.4) converges absolutely if k

n

= n for all n.

(c) Show that if there is an integer k such that the series

∞

n=1

b

n

a

k+1

n

(3.6)

converges absolutely, then (3.4) converges absolutely if k

n

= k for all n.

(d) Suppose there is an r > 0 such that ¦a

n

− a

m

¦ ≥ r for all n m. Show that

_

¦a

n

¦

−3

< ∞. In particular,

if the sequence ¦b

n

¦ is bounded then the series (3.6) with k = 2 converges absolutely. (This is somewhat

involved and the reader may prefer to prove part (f) directly since this is the only application.)

(e) Show that if the series (3.5) converges in M(C) to a meromorphic function f then

f (z) =

∞

n=1

_

¸

¸

¸

¸

¸

_

b

n

z − a

n

+

b

n

a

n

_

¸

¸

_

¸

¸

_

1 +

_

z

a

n

_

+ . . . +

_

z

a

n

_

k

n

−1

_

¸

¸

_

¸

¸

_

_

¸

¸

¸

¸

¸

_

(f) Let ω and ω

·

be two complex numbers such that Im(ω

·

/ω) 0. Using the previous parts of this exercise

show that the series

ζ(z) =

1

z

+

·

_

1

z − w

+

1

w

+

z

w

2

_

,

where the sum is over all w = 2nω+2n

·

ω for n, n = 0, ±1, ±2, . . . but not w = 0, is convergent in M(C) to a

meromorphic function ζ with simple poles at the points 2nω+2n

·

ω

·

. This function is called the Weierstrass

zeta function.

(g) Let /(z) = −ζ

·

(z); / is called the Weierstrass pe function. Show that

/(z) =

1

z

2

+

·

_

1

(z − w)

2

−

1

w

2

_

where the sum is over the same w as in part (f). Also show that

/(z) = /(z + 2nω + 2n

·

ω

·

)

for all integers n and n

·

. That is, / is doubly periodic with periods 2ω and 2ω

·

.

Solution. Not available.

124

Exercise 3. This exercise shows how to deduce Weierstrass’s Theorem for the plane (Theorem VII. 5.12)

from Mittag-Leﬄer’s Theorem.

(a) Deduce from Exercises 2(a) and 2(b) that for any sequence ¦a

n

¦ in C with lima

n

= ∞ and a

n

0 there

is a sequence of integers ¦k

n

¦ such that

h(z) =

∞

n=1

_

¸

¸

¸

¸

¸

_

1

z − a

n

+

1

a

n

+

1

a

n

_

z

a

n

_

+ . . . +

1

a

n

_

z

a

n

_

k

n

−1

_

¸

¸

¸

¸

¸

_

is a meromorphic function on C with simple poles at a

1

, a

2

, . . .. The remainder of the proof consists of

showing that there is a function f such that h = f

·

/ f . This function f will then have the appropriate zeros.

(b) Let z be an arbitrary but ﬁxed point in C− ¦a

1

, a

2

, . . .¦. Show that if γ

1

and γ

2

are any rectiﬁable curves

in C − ¦a

1

, a

2

, . . .¦ from 0 to z and h is the function obtained in part (a), then there is an integer m such that

_

γ

1

h −

_

γ

2

h = 2πim.

(c) Again let h be the meromorphic function from part (a). Prove that for z a

1

, a

2

, . . . and γ any rectiﬁable

curve in C − ¦a

1

, a

2

, . . .¦,

f (z) = exp

__

γ

h

_

deﬁnes an analytic function on C − ¦a

1

, a

2

, . . .¦ with f

·

/ f = h. (That is, the value of f (z) is independent of

the curve γ and the resulting function f is analytic.

(d) Suppose that z ∈ ¦a

1

, a

2

, . . .¦; show that z is a removable singularity of the function f deﬁned in part

(c). Furthermore, show that f (z) = 0 and that the multiplicity of this zero equals the number of times that z

appears in the sequence ¦a

1

, a

2

, . . .¦.

(e) Show that

f (z) =

∞

_

n=1

_

1 −

z

a

n

_

exp

_

¸

¸

¸

¸

¸

_

z

a

n

+

1

2

_

z

a

n

_

2

+ . . . +

1

2

_

z

a

n

_

k

n

_

¸

¸

¸

¸

¸

_

(3.7)

Remark. We could have skipped parts (b), (c), and (d) and gone directly from (a) to (e). However this

would have meant that we must show that (3.7) converges in H(C) and it could hardly be classiﬁed as a new

proof. The steps outlined in parts (a) through (d) give a proof of Weierstrass’s Theorem without introducing

inﬁnite products.

Solution. Not available.

Exercise 4. This exercise assumes a knowledge of the terminology and results of Exercise VII. 5.11.

(a) Deﬁne two functions f and g in H(G) to be relatively prime (in symbols, ( f , g) = 1) if the only common

divisors of f and g are non-vanishing functions in H(G). Show that ( f , g) = 1 iﬀ ¸( f ) ∩ ¸(g) = .

(b) If ( f , g) = 1, show that there are functions f

1

, g

1

in H(G) such that f f

1

+gg

1

= 1 (Hint: Show that there

is a meromorphic function ϕ on G such that f

1

= ϕg ∈ H(G) and g¦(1 − f f

1

).)

(c) Let f

1

, . . . , f

n

∈ H(G) and g = g.c.d ¦ f

1

, . . . , f

n

¦. Show that there are functions ϕ

1

, . . . , ϕ

n

in H(G) such

that g = ϕ

1

f

1

+ . . . + ϕ

n

f

n

. (Hint: Use (b) and induction.)

(d) If ¦_

α

¦ is a collection of ideals in H(G), show that _ =

_

α

_

α

is also an ideal. If S ⊂ H(G) then let

_ = ∩¦S : S is an ideal of H(G) and S ⊂ _¦. Prove that _ is the smallest ideal in H(G) that contains S

and _ = ¦ϕ

1

f

1

+ . . . + ϕ

n

f

n

: ϕ

k

∈ H(G), f

k

∈ S for 1 ≤ k ≤ n¦. _ is called the ideal generated by S and

is denoted by _ = (S). If S is ﬁnite then (S) is called a ﬁnitely generated ideal. If S = ¦ f ¦ for a single

function f then ( f ) is called a principal ideal.

(e) Show that every ﬁnitely generated ideal in H(G) is a principal ideal.

125

(f) An ideal _ is called a ﬁxed ideal if ¸(_) ; otherwise it is called a free ideal. Prove that if _ = (S)

then ¸(_) = ¸(S) and that a proper principal ideal is ﬁxed.

(g) Let f

n

(z) = sin(2

−n

z) for all n ≥ 0 and let _ = (¦ f

1

, f

2

, . . .¦). Show that _ is a ﬁxed ideal in H(C) which

is not a principal ideal.

(h) Let _ be a ﬁxed ideal and prove that there is an f in H(G) with ¸( f ) = ¸(_) and _ ⊂ ( f ). Also show

that _ = ( f ) if _ is ﬁnitely generated.

(i) Let Abe a maximal ideal that is ﬁxed. Show that there is a point a in G such that A= ((z − a)).

(j) Let ¦a

n

¦ be a sequence of distinct points in G with no limit point in G. Let _ = ¦ f ∈ H(G) : f (a

n

) = 0

for all but a ﬁnite number of the a

n

¦. Show that _ is a proper free ideal in H(G).

(k) If _ is a free ideal show that for any ﬁnite subset S of _, ¸(S) . Use this to show that _ can

contain no polynomials.

l) Let _ be a proper free ideal; then _ is a maximal ideal iﬀ whenever g ∈ H(G) and ¸(g) ∩ ¸( f ) for

all f in _ then g ∈ _.

Solution. Not available.

Exercise 5. Let G be a region and let ¦a

n

¦ be a sequence of distinct points in G with no limit point in G.

For each integer n ≥ 1 choose integers k

n

≥ 0 and constants A

(k)

n

, 0 ≤ k ≤ k

n

. Show that there is an analytic

function f on G such that f

(k)

(a

n

) = k!A

(k)

n

. (Hint: Let g be an analytic function on G with a zero at a

n

of

multiplicity k

n

+ 1. Let h be a meromorphic function on G with poles at each an of order k

n

+ 1 and with

singular part S

n

(z). Choose the S

n

so that f = gh has the desired property.)

Solution. Not available.

Exercise 6. Find a meromorphic function with poles of order 2 at 1,

√

2,

√

3, . . . such that the residue at

each pole is 0 and lim(z −

√

n)

2

f (z) = 1 for all n.

Solution. We claim that the function

f (z) =

∞

k=1

3

√

nz

2

− 2z

3

n

3/2

(z −

√

n)

2

has the desired properties. f has poles of order 2 at z =

√

n, n ∈ N (and no other poles) and it is analytic

for every z ∈ C that is not one of the poles. Moreover f converges in M(C). To see this let r > 0 and choose

N ∈ N so large that

√

n > 2r for all n ≥ N. Then for ¦z¦ ≤ r and n ∈ N

¦

3

√

nz

2

− 2z

3

n

3/2

(z −

√

n)

2

¦ ≤

r

2

(3

√

n +

√

n)

n

3/2

(

√

n − r)

2

≤

16r

2

n

2

=: M

n

.

This is the majorant independent of z for the Weierstrass’ Criterion.

Next ﬁx m ∈ N and consider the limit in equation lim

z→

√

n

(z −

√

n)

2

f (z) = 1,

lim

z→

√

m

(z −

√

m)

2

f (z) = lim

z→

√

m

(z −

√

m)

2

∞

n=1

3

√

nz

2

− 2z

3

n

3/2

(z −

√

n)

2

= lim

z→

√

m

m−1

n=1

(z −

√

m)(3

√

nz

2

− 2z

3

)

n

3/2

(z −

√

n)

2

+

3m

3/2

− 2m

3/2

m

3/2

+ lim

z→

√

m

∞

m+1

(z −

√

n)(3

√

nz

2

− 2z

3

)

n

3/2

(z −

√

n)

2

= 0 + 1 + 0 = 1.

126

Lastly we have to check that the residuals are 0 at the poles. By Proposition V 2.4, p.113, for a function

g(z) := (z −

√

m)

2

f (z), it has to be veriﬁed that g

·

(z)¦

z=

√

m

= 0.

First ﬁnd the derivative of f as

f

·

(z) =

∞

n=1

6

√

nz

2

− 6nz − 2z

3

n

3/2

(z −

√

m)

3

then compute the residuum at z =

√

m

Res( f ,

√

m) = g

·

(z)¦

z=

√

m

= 2(z −

√

m) f (z) + (z −

√

m)

2

f

·

(z)

¸

¸

¸

z=

√

m

=

_

¸

¸

¸

¸

¸

_

∞

n=1

2(z −

√

m)(3

√

nz

2

− 2z

3

)

n

3/2

(z −

√

n)

2

+

∞

n=1

(z −

√

m)

2

(6

√

nz

2

− 6nz − 2z

3

)

n

3/2

(z −

√

m)

3

_

¸

¸

¸

¸

¸

_

z=

√

m

=

_

¸

¸

¸

¸

¸

_

∞

n=1

_

2(z −

_

(n))(3

√

nz

2

− 2z

3

) + (z −

√

m)(6

√

nz

2

) − 6nz − 2z

3

)

_

z −

√

m

n

3/2

(z −

√

n)

3

_

z=

√

m

If n m and z =

√

m, the last factor equals zero, in the case n = m, (z −

√

m)

2

cancels directly and we are

left with

12

√

m − 2z

3

− 6mz

n

3/2

(z −

√

m)

¸

¸

¸

¸

¸

¸

z=m

=

−6z(z −

√

m)

n

3/2

¸

¸

¸

¸

¸

¸

z=m

= 0.

The residuum vanishes for all

√

m, m ∈ N and therefore the function considered in this exercise has all the

required properties.

127

Chapter 9

Analytic Continuation and Riemann

Surfaces

9.1 Schwarz Reﬂection Principle

Exercise 1. Let γ be a simple closed rectiﬁable curve with the property that there is a point a such that for

all z on γ the line segment [a, z] intersects ¦γ¦ only at z; i.e. [a, z] ∩ ¦γ¦ = ¦z¦. Deﬁne a point w to be inside

γ if [a, w] ∩ ¦γ¦ = and let G be the collection of all points that are inside γ.

(a) Show that G is a region and G

−

= G ∪ ¦γ¦.

(b) Let f : G

−

→ C be a continuous function such that f is analytic on G. Show that

_

γ

f = 0.

(c) Show that n(γ; z) = ±1 if z is inside γ and n(γ; z) = 0 if z G

−

.

Remarks. It is not necessary to assume that γ has such a point a as above; each part of this exercise remains

true if γ is only assumed to be a simple closed rectiﬁable curve. Of course, we must deﬁne what is meant

by the inside of γ. This is diﬃcult to obtain. The fact that a simple closed curve divides the plane into two

pieces (an inside and an outside) is the content of the Jordan Curve Theorem. This is a very deep result of

topology.

Solution. Not available.

Exercise 2. Let G be a region in the plane that does not contain zero and let G

∗

be the set of all points z

such that there is a point w in G where z and w are symmetric with respect to the circle ¦ξ¦ = 1. (See III.

3.17.)

(a) Show that G

∗

= ¦z : (1/¯ z) ∈ G¦.

(b) If f : G → C is analytic, deﬁne f

∗

: G

∗

→ C by f

∗

(z) = f (1/¯ z). Show that f

∗

is analytic.

(c) Suppose that G = G

∗

and f is an analytic function deﬁned on G such that f (z) is real for z in G with

¦z¦ = 1. Show that f = f

∗

.

(d) Formulate and prove a version of the Schwarz Reﬂection Principle where the circle ¦ξ¦ = 1 replaces R.

Do the same thing for an arbitrary circle.

Solution. Not available.

Exercise 3. Let G, G

+

, G

−

, G

0

be as in the statement of the Schwarz Reﬂection Principle and let f :

G

+

∪ G

0

→ C

∞

be a continuous function such that f is meromorphic on G

+

. Also suppose that for x in G

0

f (x) ∈ R. Show that there is a meromorphic function g : G → C

∞

such that g(z) = f (z) for z in G

+

∪G

0

. Is

it possible to allow f to assume the value ∞on G

0

?

128

Solution. Not available.

9.2 Analytic Continuation Along a Path

Exercise 1. The collection ¦D

0

, D

1

, . . . , D

n

¦ of open disks is called a chain of disks if D

j−1

∩ D

j

for

1 ≤ j ≤ n. If ¦( f

j

, D

j

) : 0 ≤ j ≤ n¦ is a collection of function elements such that ¦D

0

, D

1

, . . . , D

n

¦ is a chain

of disks and f

j−1

(z) = f

j

(z) for z in D

j−1

∩ D

j

, 1 ≤ j ≤ n; then ¦( f

j

, D

j

) : 0 ≤ j ≤ n¦ is called an analytic

continuation along a chain of disks. We say that ( f

n

, D

n

) is obtained by an analytic continuation of ( f

0

, D

0

)

along a chain of disks.

(a) Let ¦( f

j

, D

j

) : 0 ≤ j ≤ n¦ be an analytic continuation along a chain of disks and let a and b be the centers

of the disks D

0

and D

n

respectively. Show that there is a path γ from a to b and an analytic continuation

¦(g

t

, B

t

)¦ along γ such that ¦γ¦ ⊂

_

n

j=0

D

j

, [ f

0

]

a

= [g

0

]

a

and [ f

n

]

b

= [g

1

]

b

.

(b) Conversely, let ¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ be an analytic continuation along a path γ : [0, 1] → C and let

a = γ(0), b = γ(1). Show that there is an analytic continuation along a chain of disks ¦(g

j

, B

j

) : 0 ≤ j ≤ n¦

such that ¦γ¦ ⊂

_

n

j=0

B

j

, [ f

0

]

a

= [g

0

]

a

and [ f

1

]

b

= [g

n

]

b

.

Solution. Not available.

Exercise 2. Let D

0

= B(1; 1) and let f

0

be the restriction of the principal branch of

√

z to D

0

. Let γ(t) =

exp(2πit) and σ(t) = exp(4πit) for 0 ≤ t ≤ 1.

(a) Find an analytic continuation ¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ of ( f

0

, D

0

) along γ and show that [ f

1

]

1

= [−f

0

]

1

.

(b) Find an analytic continuation ¦(g

t

, B

t

) : 0 ≤ t ≤ 1¦ of ( f

0

, D

0

) along σ and show that [g

1

]

1

= [g

0

]

1

.

Solution. Not available.

Exercise 3. Let f be an entire function, D

0

= B(0; 1), and let γ be a path from 0 to b. Show that if

¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ is a continuation of ( f , D

0

) along γ then f

1

(z) = f (z) for all z in D

1

(This exercise is

rather easy; it is actually an exercise in the use of the terminology.)

Solution. Not available.

Exercise 4. Let γ : [0, 1] → C be a path and let ¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ be an analytic continuation along γ.

Show that ¦( f

·

t

, D

t

) : 0 ≤ t ≤ 1¦ is also a continuation along γ.

Solution. Not available.

Exercise 5. Suppose γ : [0, 1] → C is a closed path with γ(0) = γ(1) = a and let ¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ be

an analytic continuation along γ such that [ f

1

]

a

= [ f

·

0

]

a

and f

0

0. What can be said about ( f

0

, D

0

)?

Solution. Not available.

Exercise 6. Let D

0

= B(1; 1) and let f

0

be the restriction to D

0

of the principal branch of the logarithm.

For an integer n let γ(t) = exp(2πint), 0 ≤ t ≤ 1. Find a continuation ¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ along γ of

( f

0

, D

0

) and show that [ f

1

]

1

= [ f

0

+ 2πin].

Solution. Not available.

Exercise 7. Let γ : [0, 1] → C be a path and let ¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ be an analytic continuation along γ.

Suppose G is a region such that f

t

(D

t

) ⊂ G for all t, and suppose there is an analytic function h : G → C

such that h( f

0

(z)) = z for all z in D

0

. Show that h( f

t

(z)) = z for all z in D

t

and for all t. Hint: Show that

T = ¦t : h( f

t

(z)) = z for all z in D

t

¦ is both open and closed in [0, 1].

129

Solution. Not available.

Exercise 8. Let γ : [0, 1] → C be a path with γ(0) = 1 and γ(t) 0 for any t. Suppose that ¦( f

t

, D

t

) : 0 ≤

t ≤ 1¦ is an analytic continuation of f

0

(z) = log z. Show that each f

t

is a branch of the logarithm.

Solution. Not available.

9.3 Monodromy Theorem

Exercise 1. Prove that the set T deﬁned in the proof of Lemma 3.2 is closed.

Solution. Not available.

Exercise 2. Let ( f , D) be a function element and let a ∈ D. If γ : [0, 1] → C is a path with γ(0) = a and

γ(1) = b and ¦( f

t

, D

t

) : 0 ≤ t ≤ 1¦ is an analytic continuation of ( f , D) along γ, let R(t) be the radius of

convergence of the power series expansion of f

t

at z = γ(t).

(a) Show that R(t) is independent of the choice of continuation. That is, if a second continuation ¦(g

t

, B

t

)¦

along γ is given with [g

0

]

a

= [ f ]

a

and r(t) is the radius of convergence of the power series expansion of g

t

about z = γ(t) then r(t) = R(t) for all t.

(b) Suppose that D = B(1; 1), f is the restriction of the principal branch of the logarithm to D, and

γ(t) = 1 + at for 0 ≤ t ≤ 1 and a > 0. Find R(t).

(c) Let ( f , D) be as in part (b), let 0 < a < 1 and let γ(t) = (1 − at) exp(2πit) for 0 ≤ t ≤ 1. Find R(t).

(d) For each of the functions R(t) obtained in parts (b) and (c), ﬁnd min¦R(t) : 0 ≤ t ≤ 1¦ as a function of a

and examine the behavior of this function as a → ∞or a → 0.

Solution. Not available.

Exercise 3. Let Γ : [0, 1] × [0, 1] → C be a continuous function such that Γ(0, u) = a, Γ(1, u) = b for

all u. Let γ

u

(t) = Γ(t, u) and suppose that ¦( f

t,u

, D

t,u

) : 0 ≤ t ≤ 1¦ is an analytic continuation along γ

u

such that [ f

0,u

]

a

= [ f

0,v

]

a

for all u and v in [0, 1]. Let R(t, u) be the radius of convergence of the power

series expansion of f

t,u

about z = Γ(t, u). Show that either R(t, u) ≡ ∞ or R : [0, 1] × [0, 1] → (0, ∞) is a

continuous function.

Solution. Not available.

Exercise 4. Use Exercise 3 to give a second proof of the Monodromy Theorem.

Solution. Not available.

9.4 Topological Spaces and Neighborhood Systems

Exercise 1. Prove the propositions which were stated in this section without proof.

Solution. Not available.

Exercise 2. Let (X, 7) and (Ω, S) be topological spaces and let Y ⊂ X. Show that if f : X → Ω is a

continuous function then the restriction of f to Y is a continuous function of (Y, 7

Y

) into (Ω, S).

Solution. Not available.

130

Exercise 3. Let X and Ω be sets and let ¦A

x

: x ∈ X¦ and ¦A

ω

: ω ∈ Ω¦ be neighborhood systems and let

7 and S be the induced topologies on X and Ω respectively.

(a) Show that a function f : X → Ω is continuous iﬀ when x ∈ X and ω = f (x), for each ∆ in A

ω

there is

a U in A

x

such that f (U) ⊂ ∆.

(b) Let X = Ω = C and let A

z

= A

z

= ¦B(z; ) : > 0¦ for each z in C. Interpret part (a) of this exercise

for this particular situation.

Solution. Not available.

Exercise 4. Adopt the notation of Exercise 3. Show that a function f : X → Ω is open iﬀ for each x in X

and U in A

x

there is a set ∆ in A

ω

(where ω = f (x)) such that ∆ ⊂ f (U).

Solution. Not available.

Exercise 5. Adopt the notation of Exercise 3. Let Y ⊂ X and deﬁne 1

Y

= ¦Y ∩ U : U ∈ A

Y

¦ for each y in

Y. Show that ¦1

Y

: y ∈ Y¦ is a neighborhood system for Y and the topology it induces on Y is 7

Y

.

Solution. Not available.

Exercise 6. Adopt the notation of Exercise 3. For each point (x, ω) in X × Ω let

1

(x,ω)

= ¦U × ∆ : U ∈ A

x

, ∆ ∈ A

ω

¦

(a) Show that ¦1

(x,ω)

: (x, ω) ∈ X×Ω¦ is a neighborhood system on X×Ωand let / be the induced topology

on X × Ω.

(b) If U ∈ 7 and ∆ ∈ S, call the set U × ∆ an open rectangle. Prove that a set is in / iﬀ it is the union of

open rectangles.

(c) Deﬁne p

1

: X × Ω → X and p

2

: X × Ω → X by p

1

(x, ω) = x and p

2

(x, ω) = ω. Show that

p

1

and p

2

are open continuous maps. Furthermore if (Z, F) is a topological space show that a function

f : (Z, F) → (X × Ω, /) is continuous iﬀ p

1

◦ f : Z → X and p

2

◦ f : Z →Ω are continuous.

Solution. Not available.

9.5 The Sheaf of Germs of Analytic Functions on an Open Set

Exercise 1. Deﬁne F : S(C) → C by F(z, [ f ]

z

) = f (z) and show that F is continuous.

Solution. Not available.

Exercise 2. Let / be the complete analytic function obtained from the principal branch of the logarithm

and let G = C − ¦0¦. If D is an open subset of G and f : D → C is a branch of the logarithm show that

[ f ]

a

∈ / for all a in D. Conversely, if ( f , D) is a function element such that [ f ]

a

∈ / for some a in D, show

that f : D → C is a branch of the logarithm. (Hint: Use Exercise 2.8.)

Solution. Not available.

Exercise 3. Let G = C − ¦0¦, let / be the complete analytic function obtained from the principal branch

of the logarithm, and let (F, ρ) be the Riemann surface of / (so that G is the base of F). Show that F is

homeomorphic to the graph Γ = ¦(z, e

z

) : z ∈ G¦ considered as a subset of C × C. (Use the map h : F → Γ

deﬁned by h(z, [ f ]

z

) = ( f (z), z) and use Exercise 2.) State and prove an analogous result for branches of

z

1/n

.

131

Solution. Not available.

Exercise 4. Consider the sheaf S(C), let B = ¦z : ¦z − 1¦ < 1¦, let l be the principal branch of the logarithm

deﬁned on B, and let l

1

(z) = l(z) + 2πi for all z in B. (a) Let D = ¦z : ¦z¦ < 1¦ and show that (

1

2

, [l] 1

2

) and

(

1

2

, [l

1

] 1

2

) belong to the same component of ρ

−1

(D). (b) Find two disjoint open subsets of S(C) each of which

contains one of the points (

1

2

, [l] 1

2

) and (

1

2

, [l

1

] 1

2

).

Solution. Not available.

9.6 Analytic Manifolds

Exercise 1. Show that an analytic manifold is locally compact. That is, prove that if a ∈ X and U is an

open neighborhood of a then there is an open neighborhood V of a such that V

−

⊂ U and V

−

is compact.

Solution. Not available.

Exercise 2. Which of the following are analytic manifolds? What is its analytic structure if it is a manifold?

(a) A cone in R

3

. (b) ¦(x

1

, x

2

, x

3

) ∈ R

3

: x

2

1

+ x

2

2

+ x

2

3

= 1 or x

2

1

+ x

2

2

> 1 and x

3

= 0¦.

Solution. Not available.

Exercise 3. The following is a generalization of Proposition 6.3(b). Let (X, Φ) be an analytic manifold, let

Ω be a topological space, and suppose there is a continuous function h of X onto Ω that is locally one-one

(that is, if x ∈ X there is an open set U such that x ∈ U and h is one-one on U). If (U, ϕ) ∈ Φ and h is

one-one on U let ∆ = h(U) and let ψ : ∆ → C be deﬁned by ψ(ω) = ϕ◦ (h/U)

−1

(ω). Let Ψ be the collection

of all such pairs (∆, ψ). Prove that (Ω, Ψ) is an analytic manifold and h is an analytic function from X to Ω.

Solution. Not available.

Exercise 4. Let T = ¦z : ¦z¦ = 1¦ × ¦z : ¦z¦ = 1¦; then T is a torus. (This torus is homeomorphic to the

usual hollow doughnut in R

3

.) If ω and ω

·

are complex numbers such that Im (ω/ω

·

) 0 then ω and

ω

·

, considered as elements of the vector space C over R, are linearly independent. So each z in C can be

uniquely represented as z = tω+t

·

ω

·

; t, t

·

in R. Deﬁne h : C → T by h(tω+tω

·

) = (e

2πit

, e

2πit

·

). Show that h

induces an analytic structure on T. (Use Exercise 3.) (b) If ω, ω

·

and ζ, ζ

·

are two pairs of complex numbers

such that Im (ω/ω

·

) 0 and Im (ζ/ζ

·

) 0, deﬁne σ(sζ+s

·

ζ

·

) = (e

2πis

, e

2πis

·

) and τ(tω+t

·

ω

·

) = (e

2πit

, e

2πit

·

).

Let G = ¦tω + t

·

ω

·

: 0 < t < 1, 0 < t

·

< 1¦ and Ω = ¦sζ + s

·

ζ

·

: 0 < s < 1, 0 < s

·

< 1¦; show that both σ

and τ are one-one on G and Ω respectively. (Both G and Ω are the interiors of parallelograms.) If Φ

τ

and

Φ

σ

are the analytic structures induced on T by τ and σ respectively, and if the identity map of (T, Φ

τ

) into

(T, Φ

σ

) is analytic then show that the function f : G → Ω deﬁned by f = σ

−1

◦ τ is analytic. (To say that

the identity map of (T, Φ

τ

) into (T, Φ

σ

) is analytic is to say that Φ

τ

and Φ

σ

are equivalent structures.) (c)

Let ω = 1, ω

·

= i, ζ = 1, ζ = α where Im α 0; deﬁne σ, τ, G, Ω and f as in part (b). Show that Φ

τ

and

Φ

σ

are equivalent analytic structures if and only if α = i. (Hint: Use the Cauchy-Riemann equations.) (d)

Can you generalize part (c)? Conjecture a generalization?

Solution. Not available.

Exercise 5. (a) Let f be a meromorphic function deﬁned on C and suppose f has two independent periods

ω and ω

·

. That is, f (z) = f (z +nω+n

·

ω

·

) for all z in C and all integers n and n

·

, and Im (ω/ω

·

) 0. Using

the notation of Exercise 4(a) show that there is an analytic function F : T → C

∞

such that f = F ◦ h. (For

an example of a meromorphic function with two independent periods see Exercise VIII. 4.2(g).)

(b) Prove that there is no non-constant entire function with two independent periods.

132

Solution. Not available.

Exercise 6. Show that an analytic surface is arcwise connected.

Solution. Not available.

Exercise 7. Suppose that f : C

∞

→ C

∞

is an analytic function.

(a) Show that either f ≡ ∞or f

−1

(∞) is a ﬁnite set.

(b) If f ∞, let a

1

, . . . , a

n

be the points in C where f takes on the value ∞. Show that there are polynomials

p

0

, p

1

, . . . , p

n

such that

f (z) = p

0

(z) +

n

k=1

p

k

_

1

z − a

k

_

for z in C.

(c) If f is one-one, show that either f (z) = az + b (some a, b in C) or f (z) =

a

z−c

+ b (some a, b, c in C).

Solution. Not available.

Exercise 8. Furnish the details of the discussion of the surface for

√

z at the end of this section.

Solution. Not available.

Exercise 9. Let G =

_

z : −

π

2

< Re z <

π

2

_

and deﬁne f : G → C by f (z) = sin z. Give a discussion for f

similar to the discussion of

√

z at the end of this section.

Solution. Not available.

9.7 Covering spaces

Exercise 1. Suppose that (X, ρ) is a covering space of Ω and (Ω, π) is a covering space of Y; prove that

(X, π ◦ ρ) is also a covering space of Y.

Solution. Not available.

Exercise 2. Let (X, ρ) and (Y, σ) be covering spaces of Ω and Λ respectively. Deﬁne ρ×σ : X×Y →Ω×Λ

by (ρ × σ)(x, y) = (ρ(x), σ(y)) and show that (X × Y, ρ × σ) is a covering space of Ω× Λ.

Solution. Not available.

Exercise 3. Let (Ω, ψ) be an analytic manifold and let (X, ρ) be a covering space of Ω. Show that there is

an analytic structure Φ on X such that ρ is an analytic function from (X, Φ) to (Ω, ψ).

Solution. Not available.

Exercise 4. Let (X, ρ) be a covering space of Ω and let ω ∈ Ω. Show that each component of ρ

−1

(ω)

consists of a single point and ρ

−1

(ω)) has no limit points in X.

Solution. Not available.

Exercise 5. Let Ω be a pathwise connected space and let (X, ρ) be a covering space of Ω. If ω

1

and ω

2

are

points in Ω, show that ρ

−1

(ω

1

) and ρ

−1

(ω

2

) have the same cardinality. (Hint: Let γ be a path in Ω from ω

1

to ω

2

; if x

1

∈ ρ

−1

(ω

1

) and ˜ γ is the lifting of γ with initial point x

1

= ˜ γ(0), let f (x

1

) = ˜ γ(1). Show that f is a

one-one map of ρ

−1

(ω

1

) onto ρ

−1

(ω

2

).)

133

Solution. Not available.

Exercise 6. In this exercise all spaces are regions in the plane.

(a) Let (G, f ) be a covering of Ω and suppose that f is analytic; show that if Ω is simply connected then

f is one-one. (Hint: If f (z

1

) = f (z

2

) let γ be a path in G from z

1

to z

2

and consider a certain analytic

continuation along f ◦ γ; apply the Monodromy Theorem.)

(b) Suppose that (G

1

, f

1

) and (G

2

, f

2

) are coverings of the region Ω such that both f

1

and f

2

are analytic.

Show that if G

1

is simply connected then there is an analytic function f : G

1

→ G

2

such that (G

1

, f ) is a

covering of G

2

and f

1

= f

2

◦ f . That is the diagram is commutative.

(c) Let (G

1

, f

1

), (G

2

, f

2

) and Ω be as in part (b) and, in addition, assume that both G

1

and G

2

are simply

connected. Show that there is a one-one analytic function mapping G

1

onto G

2

.

Solution. Not available.

Exercise 7. Let G and Ω be regions in the plane and suppose that f : G → Ω is an analytic function

such that (G, f ) is a covering space of Ω. Show that for every region Ω

1

contained in Ω which is simply

connected there is an analytic function g : Ω

1

→G such that f (g

1

(ω)) = ω for all ω in Ω

1

.

Solution. Not available.

Exercise 8. What is a simply connected covering space of the ﬁgure eight?

Solution. Not available.

Exercise 9. Give two nonhomeomorphic covering spaces of the ﬁgure eight that are not simply connected.

Solution. Not available.

Exercise 10. Prove that the closed curve in Exercise IV.6.8 is not homotopic to zero in the doubly punctured

plane.

Solution. Not available.

134

Chapter 10

Harmonic Functions

10.1 Basic properties of harmonic functions

Exercise 1. Show that if u is harmonic then so are u

x

=

∂u

∂x

and u

y

=

∂u

∂y

.

Solution. Let u be harmonic, that is

u

xx

+ u

yy

= 0. (10.1)

Since u : G → R is harmonic, u is inﬁnitely times diﬀerentiable by Proposition 1.3 p. 252. Thus, the third

partial derivatives of u exist and are continuous (therefore u

xyy

= u

yyx

and u

yxx

= u

xxy

). Now, we show that

u

x

is harmonic, that is

(u

x

)

xx

+ (u

x

)

yy

= 0.

(u

x

)

xx

+ (u

x

)

yy

= u

xxx

+ u

xyy

= u

xxx

+ u

yyx

= (u

xx

+ u

yy

)

x

=

(10.1)

0

x

= 0.

Thus, u

x

is harmonic. Finally, we show that u

y

is harmonic, that is

(u

y

)

xx

+ (u

y

)

yy

= 0.

(u

y

)

xx

+ (u

y

)

yy

= u

yxx

+ u

yyy

= u

xxy

+ u

yyy

= (u

xx

+ u

yy

)

y

=

(10.1)

0

y

= 0.

Thus, u

y

is harmonic.

Exercise 2. If u is harmonic, show that f = u

x

− iu

y

is analytic.

Solution. Since, u is harmonic, we have

u

xx

+ u

yy

= 0. (10.2)

In addition, u is inﬁnitely times diﬀerentiable by Proposition 1.3 p. 252. Thus, all the partials are continuous

(therefore u

xy

= u

yx

).

To show: f = u

x

− iu

y

is analytic which is equivalent to showing

a) Re( f ) = u

x

and Im( f ) = −u

y

are harmonic and

b) u

x

and −u

y

have to satisfy the Cauchy Riemann equations (by Theorem III 2.29).

135

a) By the previous exercise, we have seen that u

x

and u

y

are harmonic. Therefore, −u

y

is harmonic, too.

Thus, part a) is proved.

b) We have to show that the Cauchy-Riemann equations are satisﬁed, that is

(u

x

)

x

= −(u

y

)

y

and (u

x

)

y

= −(−u

y

)

x

.

The ﬁrst is true by (10.2) since

(u

x

)

x

= −(u

y

)

y

⇐⇒ u

xx

+ u

yy

= 0.

The second is true since u

xy

= u

yx

. Therefore

(u

x

)

y

= −(−u

y

)

x

⇐⇒ u

xy

= u

yx

⇐⇒ u

xy

= u

yx

.

In summary, f = u

x

− iu

y

is analytic provided u is harmonic.

Exercise 3. Let p(x, y) =

_

n

k,l=0

a

kl

x

k

y

l

for all x, y in R.

Show that p is harmonic iﬀ:

(a) k(k − 1)a

k,l−2

+ l(l − 1)a

k−2,l

= 0 for 2 ≤ k, l ≤ n;

(b) a

n−1,l

= a

n,l

= 0 for 2 ≤ l ≤ n;

(c) a

k,n−1

= a

k,n

= 0 for 2 ≤ k ≤ n.

Solution. Not available.

Exercise 4. Prove that a harmonic function is an open map. (Hint: Use the fact that the connected subsets

of R are intervals.)

Solution. Not available.

Exercise 5. If f is analytic on G and f (z) 0 for any z show that u = log ¦ f ¦ is harmonic on G.

Solution. Not available.

Exercise 6. Let u be harmonic in G and suppose

¯

B(a; R) ⊂ G. Show that

u(a) =

1

πR

2

¯

B(a;R)

u(x, y) dx dy.

Solution. Let D be a disk such that

¯

B(a; R) ⊂ D ⊂ G. There exist an f ∈ H(D) such that Re( f ) = u on D.

From Cauchy’s Integral Formula, we have

f (a) =

1

2πi

_

γ

f (w)

w − a

dw =

1

2π

_

2π

0

f (a + re

iθ

) dθ

where γ(t) = a + re

it

, 0 ≤ t ≤ 2π. So

f (a) =

1

2π

_

2π

0

f (a + re

iθ

) dθ.

Multiply by r, we get

f (a)r =

1

2π

_

2π

0

f (a + re

iθ

)r dθ

136

and then integrate with respect to r yields

_

R

0

f (a)r dr =

_

R

0

1

2π

_

2π

0

f (a + re

iθ

)r dθ dr

⇐⇒ f (a)

r

2

2

=

1

2π

_

R

0

_

2π

0

f (a + re

iθ

)r dθ dr

⇐⇒ f (a) =

1

πR

2

_

2π

0

_

R

0

f (a + re

iθ

)r dr dθ

⇐⇒ f (a) =

1

πR

2

¯

B(a;R)

f (x, y) dx dy.

where the last step follows by changing from polar coordinates to rectangular coordinates. Now

u(a) = Re( f (a)) =

1

πR

2

¯

B(a;R)

Re( f (x, y)) dx dy =

1

πR

2

¯

B(a;R)

u(x, y) dx dy.

Thus

u(a) =

1

πR

2

¯

B(a;R)

u(x, y) dx dy.

Exercise 7. For ¦z¦ < 1 let

u(z) = Im

_

¸

¸

¸

¸

¸

_

_

1 + z

1 − z

_

2

_

¸

¸

¸

¸

¸

_

.

Show that u is harmonic and lim

r→1

− u(re

iθ

) = 0 for all θ. Does this violate Theorem 1.7? Why?

Solution. Let D = ¦z ∈ C : ¦z¦ < 1¦. Let f (z) =

_

1+z

1−z

_

2

. Clearly, f ∈ H(D) since z = 1 is not in D. By

Theorem III 2.29 we obtain that v = Re( f ) and u = Im( f ) are harmonic. Therefore

u(z) = Im

_

¸

¸

¸

¸

¸

_

_

1 + z

1 − z

_

2

_

¸

¸

¸

¸

¸

_

is harmonic. Next, we will show

lim

r→1

−

u(re

iθ

) = 0

137

for all θ. We have

u(re

iθ

) = Im

_

¸

¸

¸

¸

¸

_

_

1 + re

iθ

1 − re

iθ

_

2

_

¸

¸

¸

¸

¸

_

= Im

_

¸

¸

¸

¸

¸

_

_

1 + re

iθ

1 − re

iθ

_

2

_

1 − re

iθ

1 − re

iθ

_

2

_

¸

¸

¸

¸

¸

_

= Im

_

¸

¸

¸

¸

¸

_

_

(1 + re

iθ

)(1 − re

iθ

)

(1 − re

iθ

)(1 − re

iθ

)

_

2

_

¸

¸

¸

¸

¸

_

= Im

_

(1 + re

iθ

− re

−iθ

− r

2

)

2

(1 − re

iθ

− re

−iθ

+ r

2

)

2

_

= Im

_

(1 + 2ri sin θ − r

2

)

2

(1 − 2r cos θ + r

2

)

2

_

= Im

_

((1 − r

2

) + 2ri sin θ)

2

(1 − 2r cos θ + r

2

)

2

_

= Im

_

(1 − r

2

)

2

+ 4ri sin θ(1 − r

2

) − 4r

2

sin

2

θ

(1 − 2r cos θ + r

2

)

2

_

=

4r(1 − r

2

) sin θ

(1 − 2r cos θ + r

2

)

2

.

Note that the denominator is zero iﬀ r = 1 and θ = 2πk, k ∈ Z ⇐⇒ z = 1.

If we assume, that r = 1 and θ 2πk, k ∈ Z, then

lim

r→1

−

u(re

iθ

) = u(e

iθ

) =

4 1 (1 − 1

2

) sin θ

(1 − 2 1 cos θ + 1

2

)

2

=

0

2

2

(1 − cos θ)

2

= 0.

If θ = 2πk, k ∈ Z, then

lim

r→1

−

u(re

2πk

) = lim

r→1

−

4r(1 − r

2

) sin(2πk)

(1 − 2r cos(2πk) + r

2

)

2

= lim

r→1

−

0

(1 − r)

4

= 0.

Thus,

lim

r→1

−

u(re

iθ

) = 0.

This does not violate Theorem 1.7. The reason is the following: If we pick u = u and v = 0, then the

condition

limsup

z→a

u(z) ≤ 0

does not hold ∀a ∈ ∂D.

Claim: limsup

z→1

u(z) > 0.

Proof of the claim:

u(z) = u(x, y) = Im

_

¸

¸

¸

¸

¸

_

_

1 + x + iy

1 − x − iy

_

2

_

¸

¸

¸

¸

¸

_

=

4y(1 − x

2

− y

2

)

(1 − 2x + x

2

+ y

2

)

2

.

If we ﬁx x = 1, then

u(1, y) =

4y(−y

2

)

y

4

=

−4y

3

y

4

= −

4

y

∀y 0.

But −

4

y

→ ∞as y → 0

−

.

138

Exercise 8. Let u : G → R be a function with continuous second partial derivatives and deﬁne U(r, θ) =

u(r cos θ, r sin θ).

(a) Show that

r

2

_

∂

2

u

∂x

2

+

∂

2

u

∂y

2

_

= r

2

∂

2

U

∂r

2

+ r

∂U

∂r

+

∂

2

U

∂θ

2

= r

∂

∂r

_

r

∂U

∂r

_

+

∂

2

U

∂θ

2

.

So if 0 G then u is harmonic iﬀ

r

∂

∂r

_

r

∂U

∂r

_

+

∂

2

U

∂θ

2

= 0.

(b) Let u have the property that it depends only on ¦z¦ and not arg z. That is, u(z) = ϕ(¦z¦). Show that u is

harmonic iﬀ u(z) = a log ¦z¦ + b for some constants a and b.

Solution. Not available.

Exercise 9. Let u : G → R be harmonic and let A = ¦z ∈ G : u

x

(z) = u

y

(z) = 0¦; that is, A is the set of zeros

of the gradient of u. Can A have a limit point in G?

Solution. Let u be harmonic and not constant. By Exercise 1 p. 255, we know that u

x

and u

y

are harmonic.

By exercise 2 p. 255, we know that f = u

x

− iu

y

is analytic. Deﬁne B = ¦z ∈ G : f (z) = 0¦ = ¦z ∈ G :

u

x

(z) − iu

y

(z) = 0¦. Clearly

A = ¦z ∈ G : u

x

(z) = u

y

(z) = 0¦ = B

since u

x

(z) − iu

y

(z) = 0 iﬀ u

x

(z) = u

y

(z) = 0 (u

x

(z) and u

y

(z) are real). Since f is analytic, the zeros are

isolated and therefore B cannot have a limit point in G (Corollary 3.10 p. 79). This implies that A cannot

have a limit point in G.

If u is constant, then A = B = G and thus A can have a limit point in G. (In this case f ≡ 0 and by Theorem

3.7 p. 78 A has a limit point in G).

Exercise 10. State and prove a Schwarz Reﬂection Principle for harmonic functions.

Solution. Not available.

Exercise 11. Deduce the Maximum Principle for analytic functions from Theorem 1.6.

Solution. Not available.

10.2 Harmonic functions on a disk

Exercise 1. Let D = ¦z : ¦z¦ < 1¦ and suppose that f : D

−

→ C is a continuous function such that both Re f

and Im f are harmonic. Show that

f (re

iθ

) =

1

2π

_

π

−π

f (e

it

)P

r

(θ − t) dt

for all re

iθ

in D. Using Deﬁnition 2.1 show that f is analytic on D iﬀ

_

π

−π

f (e

it

)e

int

dt = 0

for all n ≥ 1.

139

Solution. Let D = ¦z : ¦z¦ < 1¦ and suppose f = u + iv : D → C is a continuous function such that u and v

are harmonic. Clearly u and v are continuous functions mapping D to R. By Corollary 2.9 we have

u(re

iθ

) =

1

2π

_

π

−π

P

r

(θ − t)u(e

it

) dt for 0 ≤ r < 1 and all θ (10.3)

and

v(re

iθ

) =

1

2π

_

π

−π

P

r

(θ − t)v(e

it

) dt for 0 ≤ r < 1 and all θ. (10.4)

Hence,

f (re

iθ

) = u(re

iθ

) + iv(re

iθ

) =

(10.3),(10.4)

1

2π

_

π

−π

P

r

(θ − t)u(e

it

) dt + i

1

2π

_

π

−π

P

r

(θ − t)v(e

it

) dt

=

1

2π

_

π

−π

P

r

(θ − t)

_

u(e

it

) + iv(r

iθ

)

_

dt =

1

2π

_

π

−π

P

r

(θ − t) f (e

it

) dt

for 0 ≤ r < 1 and all θ. Thus, we have proved

f (re

iθ

) =

1

2π

_

π

−π

f (e

it

)P

r

(θ − t) dt

for all re

iθ

in D.

Next, we will prove

f is analytic ⇐⇒

_

π

−π

f (e

iθ

)e

int

dt = 0

for all n ≥ 1. ⇒: Let f be analytic on D and assume n ≥ 1 (n integer). Deﬁne g(z) = f (z)z

n−1

. Clearly g is

analytic on D. Let γ = re

it

, −π ≤ t ≤ π, 0 < r < 1, then by Cauchy’s Theorem

_

γ

g(z) dz = 0

⇐⇒

_

γ

f (z)z

n−1

dz = 0

⇐⇒

_

π

−π

f (re

it

)ire

it

e

it(n−1)

r

n−1

dt = 0

⇐⇒ i

_

π

−π

f (re

it

)e

int

r

n

dt = 0

⇐⇒

_

π

−π

f (re

it

)e

int

r

n

dt = 0

which implies

lim

r→1

−

_

π

−π

f (re

it

)e

int

r

n

dt = 0 ⇒

by Ex 3a p. 262

_

π

−π

f (e

it

)e

int

dt = 0 ∀n ≥ 1.

140

⇐: Assume

_

π

−π

f (re

it

)e

int

r

n

dt = 0 for all n ≥ 1. We have for all z ∈ D

f (z) = f (re

iθ

) =

Part 1

1

2π

_

π

−π

f (e

it

)P

r

(θ − t) dt

=

1

2π

_

π

−π

f (e

it

)

∞

n=−∞

r

¦n¦

e

in(θ−t)

dt

=

1

2π

∞

n=−∞

r

¦n¦

e

inθ

_

π

−π

f (e

it

)e

−int

dt

=

1

2π

−1

n=−∞

r

¦n¦

e

inθ

_

π

−π

f (e

it

)e

−int

dt

.,,.

=0 by assumption

+

1

2π

∞

n=0

r

n

e

inθ

_

π

−π

f (e

it

)e

−int

dt

=

1

2π

∞

n=0

z

n

_

π

−π

f (e

it

)e

−int

dt.

Now, for any closed rectiﬁable curve γ in D we get

_

γ

f (z) dz =

_

γ

1

2π

∞

n=0

z

n

_

π

−π

f (e

it

)e

−int

dt dz =

1

2π

∞

n=0

_

γ

z

n

dz

.,,.

=0

_

π

−π

f (e

it

)e

−int

dt

since z

n

is analytic for n ≥ 0.

Thus f is analytic on D.

Exercise 2. In the statement of Theorem2.4 suppose that f is piecewise continuous on ∂D. Is the conclusion

of the theorem still valid? If not, what parts of the conclusion remain true?

Solution. Not available.

Exercise 3. Let D = ¦z : ¦z¦ < 1¦, T = ∂D = ¦z : ¦z¦ = 1¦

(a) Show that if g : D

−

→ C is a continuous function and g

r

: T → C is deﬁned by g

r

(z) = g(rz) then

g

r

(z) → g(z) uniformly for z in T as r → 1

−

.

(b) If f : T → C is a continuous function deﬁne

˜

f : D

−

→ C by

˜

f (z) = f (z) for z in T and

˜

f (re

iθ

) =

1

2π

_

π

−π

f (e

it

)P

r

(θ − t) dt

(So Re

˜

f and Im

˜

f are harmonic in D). Deﬁne

˜

f

r

: T → C by

˜

f

r

(z) =

˜

f (rz). Show that for each r < 1 there

is a sequence ¦p

n

, (z, ¯ z)¦ of polynomials in z and ¯ z such that p

n

(z, ¯ z) →

˜

f

r

(z) uniformly for z in T. (Hint: Use

Deﬁnition 2.1.)

(c) Weierstrass approximation theorem for T. If f : T → C is a continuous function then there is a

sequence ¦p

n

(z, ¯ z)¦ of polynomials in z and ¯ z such that p

n

(z, ¯ z) → f (z) uniformly for z in T.

(d) Suppose g : [0, 1] → C is a continuous function such that g(0) = g(1). Use part (c) to show that there is

a sequence ¦p

n

¦ of polynomials such that p

n

(t) → g(t) uniformly for t in [0, 1].

(e) Weierstrass approximation theorem for [0,1]. If g : [0, 1] → C is a continuous function then there is a

sequence ¦p

n

¦ of polynomials such that p

n

(t) → g(t) uniformly for t in [0, 1]. (Hint: Apply part (d) to the

function g(t) + (1 − t)g(1) + tg(0).)

(f) Show that if the function g in part (e) is real valued then the polynomials can be chosen with real

coeﬃcients.

141

Solution. Not available.

Exercise 4. Let G be a simply connected region and let Γ be its closure in C

∞

; ∂

∞

G = Γ − G. Suppose

there is a homeomorphism ϕ of Γ onto D

−

(D = ¦z : ¦z¦ < 1¦) such that ϕ is analytic on G.

(a) Show that ϕ(G) = D and ϕ(∂

∞

G) = ∂D.

(b) Show that if f : ∂

∞

G → R is a continuous function then there is a continuous function u : Γ → R such

that u(z) = f (z) for z in ∂

∞

G and u is harmonic in G.

(c) Suppose that the function f in part (b) is not assumed to be continuous at ∞. Show that there is a

continuous function u : G

−

→ R such that u(z) = f (z) for z in ∂G and u is harmonic in G (see Exercise 2).

Solution. Not available.

Exercise 5. Let G be an open set, a ∈ G, and G

0

= G − ¦a¦. Suppose that u is a harmonic function on G

0

such that lim

z→a

u(z) exists and is equal to A. Show that if U : G → R is deﬁned by U(z) = u(z) for z a

and U(a) = A then U is harmonic on G.

Solution. Not available.

Exercise 6. Let f : ¦z : Re z = 0¦ → R be a bounded continuous function and deﬁne u : ¦z : Re z > 0¦ → R

by

u(x + iy) =

1

π

_

∞

−∞

x f (it)

x

2

+ (y − t)

2

dt.

Show that u is a bounded harmonic function on the right half plane such that for c in R, f (ic) = lim

z→ic

u(z).

Solution. Not available.

Exercise 7. Let D = ¦z : ¦z¦ < l¦ and suppose f : ∂D → R is continuous except for a jump discontinuity at

z = 1. Deﬁne u : D →

¯

R by (2.5). Show that u is harmonic. Let v be a harmonic conjugate of u. What can

you say about the behavior of v(r) as r → 1

−

? What about v(re

iθ

) as r → 1

−

and θ → 0?

Solution. Not available.

10.3 Subharmonic and superharmonic functions

Exercise 1. Which of the following functions are subharmonic? superharmonic? harmonic? neither

subharmonic nor superharmonic? (a) ϕ(x, y) = x

2

+ y

2

; (b) ϕ(x, y) = x

2

− y

2

; (c) ϕ(x, y) = x

2

+ y; (d)

ϕ(x, y) = x

2

− y; (e) ϕ(x, y) = x + y

2

; (f) ϕ(x, y) = x − y

2

.

Solution. Note that

_

π

−π

sin θ dθ = 0,

_

π

−π

cos θ dθ = 0,

_

π

−π

sin

2

θ dθ = π and

_

π

−π

cos

2

θ dθ = π. For all

a = (α, β) ∈ C and any r > 0 we have

a)

ϕ(a + re

iθ

) = ϕ(α + r cos θ, β + r sin θ) = (α + r cos θ)

2

+ (β + r sin θ)

2

= α

2

+ 2αr cos θ + r

2

cos

2

θ + β

2

+ 2βr sin θ + r

2

sin

2

θ

= α

2

+ β

2

+ 2αr cos θ + 2βr sin θ + r

2

.

Thus

1

2π

_

π

−π

ϕ(a + re

iθ

) dθ = α

2

+ β

2

+

αr

π

_

π

−π

cos θ dθ

.,,.

=0

+

βr

π

_

π

−π

sin θ dθ

.,,.

=0

+r

2

= α

2

+ β

2

+ r

2

≥ α

2

+ β

2

= ϕ(a).

142

Hence, ϕ ∈ Subhar(G).

b)

ϕ(a + re

iθ

) = ϕ(α + r cos θ, β + r sin θ) = (α + r cos θ)

2

− (β + r sin θ)

2

= α

2

+ 2αr cos θ + r

2

cos

2

θ − β

2

− 2βr sin θ − r

2

sin

2

θ.

Thus

1

2π

_

π

−π

ϕ(a + re

iθ

) dθ = α

2

− β

2

+

αr

π

_

π

−π

cos θ dθ

.,,.

=0

+

r

2

2π

_

π

−π

cos

2

θ dθ

.,,.

=π

−

βr

π

_

π

−π

sin θ dθ

.,,.

=0

−

r

2

2π

_

π

−π

sin

2

θ dθ

.,,.

=π

= α

2

− β

2

+

r

2

2

−

r

2

2

= α

2

− β

2

= ϕ(a).

Hence, ϕ ∈ Har(G), therefore also ϕ ∈ Subhar(G) and ϕ ∈ Superhar(G).

c)

ϕ(a + re

iθ

) = ϕ(α + r cos θ, β + r sin θ) = (α + r cos θ)

2

+ (β + r sin θ)

= α

2

+ 2αr cos θ + r

2

cos

2

θ + β + r sin θ.

Thus

1

2π

_

π

−π

ϕ(a + re

iθ

) dθ = α

2

+ β +

αr

π

_

π

−π

cos θ dθ

.,,.

=0

+

r

2

2π

_

π

−π

cos

2

θ dθ

.,,.

=π

+

r

2π

_

π

−π

sin θ dθ

.,,.

=0

= α

2

+ β +

r

2

2

≥ α

2

+ β = ϕ(a).

Hence, ϕ ∈ Subhar(G).

d)

ϕ(a + re

iθ

) = ϕ(α + r cos θ, β + r sin θ) = (α + r cos θ)

2

− (β + r sin θ)

= α

2

+ 2αr cos θ + r

2

cos

2

θ − β − r sin θ.

Thus

1

2π

_

π

−π

ϕ(a + re

iθ

) dθ = α

2

− β +

αr

π

_

π

−π

cos θ dθ

.,,.

=0

+

r

2

2π

_

π

−π

cos

2

θ dθ

.,,.

=π

−

r

2π

_

π

−π

sin θ dθ

.,,.

=0

= α

2

− β +

r

2

2

≥ α

2

− β = ϕ(a).

Hence, ϕ ∈ Subhar(G).

e)

ϕ(a + re

iθ

) = ϕ(α + r cos θ, β + r sin θ) = (α + r cos θ) + (β + r sin θ)

2

= α + r cos θ + β

2

+ 2βr sin θ + r

2

sin

2

θ.

143

Thus

1

2π

_

π

−π

ϕ(a + re

iθ

) dθ = α + β

2

+

r

2π

_

π

−π

cos θ dθ

.,,.

=0

+

βr

π

_

π

−π

sin θ dθ

.,,.

=0

+

r

2

2π

_

π

−π

sin

2

θ dθ

.,,.

=π

= α + β

2

+

r

2

2

≥ α + β

2

= ϕ(a).

Hence, ϕ ∈ Subhar(G).

f)

ϕ(a + re

iθ

) = ϕ(α + r cos θ, β + r sin θ) = (α + r cos θ) − (β + r sin θ)

2

= α + r cos θ − β

2

− 2βr sin θ − r

2

sin

2

θ.

Thus

1

2π

_

π

−π

ϕ(a + re

iθ

) dθ = α − β

2

+

r

2π

_

π

−π

cos θ dθ

.,,.

=0

−

βr

π

_

π

−π

sin θ dθ

.,,.

=0

−

r

2

2π

_

π

−π

sin

2

θ dθ

.,,.

=π

= α − β

2

−

r

2

2

≤ α − β

2

= ϕ(a).

Hence, ϕ ∈ Superhar(G).

Exercise 2. Let Subhar(G) and Superhar(G) denote, respectively, the sets of subharmonic and superhar-

monic functions on G.

(a) Show that Subhar(G) and Superhar(G) are closed subsets of C(G; R).

(b) Does a version of Harnack’s Theorem hold for subharmonic and superharmonic functions?

Solution. Not available.

Exercise 3. (This exercise is diﬃcult.) If G is a region and if f : ∂

∞

G → R is a continuous function let

u

f

be the Perron Function associated with f . This deﬁnes a map T : (∂

∞

G; R) → Har(G) by T( f ) = u

f

.

Prove:

(a) T is linear (i.e., T(a

1

f

1

+ a

2

f

2

) = a

1

T( f

1

) + a

2

T( f

2

)).

(b) T is positive (i.e., if f (a) ≥ 0 for all a in ∂

∞

G then T( f )(z) ≥ 0 for all z in G).

(c) T is continuous. Moreover, if ¦ f

n

¦ is a sequence in C(∂

∞

G; R) such that f

n

→ f uniformly then T( f

n

) →

T( f ) uniformly on G.

(d) If the Dirichlet Problem can be solved for G then T is one-one. Is the converse true?

Solution. Not available.

Exercise 4. In the hypothesis of Theorem 3.11, suppose only that f is a bounded function on ∂

∞

G; prove

that the conclusion remains valid. (This is useful if G is an unbounded region and g is a bounded continuous

function on ∂G. If we deﬁne f : ∂

∞

G → R by f (z) = g(z) for z in ∂G and f (∞) = 0 then the conclusion

of Theorem 3.11 remains valid. Of course there is no reason to expect that the harmonic function will have

predictable behavior near ∞ — we could have assigned any value to f (∞). However, the behavior near

points of ∂G can be studied with hope of success.)

Solution. Not available.

Exercise 5. Show that the requirement that G

1

is bounded in Corollary 3.5 is necessary.

144

Solution. Not available.

Exercise 6. If f : G → Ω is analytic and ϕ : Ω → R is subharmonic, show that ϕ ◦ f is subharmonic if f

is one-one. What happens if f

·

(z) 0 for all z in G?

Solution. Clearly f is continuous, since f : G → Ω is analytic and ϕ is continuous, since ϕ : Ω → R is

subharmonic. Deﬁne u = ϕ ◦ f , then u : G → R is continuous. If we can show that for every bounded

region G

1

such that

¯

G

1

⊂ G and for every continuous function u

1

:

¯

G

1

→ R that is harmonic in G

1

and

satisﬁes u(z) ≤ u

1

(z) for z in ∂G

1

, we have

u(z) ≤ u

1

for z in G

1

, then u is subharmonic (by Corollary 3.5 p. 265).

Hence, let G

1

be a bounded region G

1

such that

¯

G

1

⊂ G and let u

1

:

¯

G

1

→ R that is harmonic in G

1

and

satisﬁes u(z) ≤ u

1

(z) for z on ∂G

1

. We have to show u(z) ≤ u

1

(z) ∀z ∈ G

1

.

Since f is one-one, we know that f

−1

exists and is analytic. So deﬁne

ϕ

1

= u

1

◦ f

−1

:

¯

Ω → R

with f

−1

:

¯

Ω

1

→

¯

G

1

. Clearly ϕ

1

is harmonic (harmonic composite analytic is harmonic). Since ϕ : Ω → R

is subharmonic, we can use Theorem 3.4 to obtain: For every region Ω

1

contained in Ωand every harmonic

function ˜ ϕ on Ω

1

, ϕ−˜ ϕ satisﬁes the MaximumPrinciple on Ω

1

. In fact ϕ−ϕ

1

satisﬁes the MaximumPrinciple

on Ω

1

. By assumption we have

u(z) ≤ u

1

(z) ∀z ∈ G

1

which is the same as

u(z) = ϕ( f (z)) ≤ u

1

(z) = ϕ

1

( f (z)) ∀z ∈ ∂G

1

since f maps G

1

onto Ω

1

, and thus ∂G

1

onto ∂Ω

1

(we choose this by assumption). Thus,

ϕ( f (z)) − ϕ

1

( f (z)) ≤ 0 ∀z ∈ ∂G

1

.

Since ϕ − ϕ

1

satisﬁes the Maximum Principle on Ω

1

, this implies

ϕ( f (z)) − ϕ

1

( f (z)) ≤ 0 ∀z ∈ G

1

,

that is

u(z) − u

1

(z) ≤ 0 ∀z ∈ G

1

and we are done. Therefore, u = ϕ ◦ f is subharmonic on G.

If f

·

(z) 0 for all z in G, then the previous result holds locally for some neighborhood in G. ( f is lo-

cally one-one and onto and f

−1

is analytic and hence u = ϕ ◦ f is locally subharmonic). Here is the claim:

Let f : G → C be analytic, let z

0

∈ A, and let f

·

(z

0

) 0. Then there is a neighborhood U of z

0

and a

neighborhood V of w

0

= f (z

0

) such that f : U → V is one-one and onto and f

−1

: V → U is analytic.

Proof of the claim: f (z) − w

0

has a simple zero at z

0

since f

·

(z

0

) 0. We can use Theorem 7.4 p. 98 to

ﬁnd > 0 and δ > 0 such that each w with ¦w − w

0

¦ < δ has exactly one pre-image z with ¦z − z

0

¦ < . Let

V = ¦w¦ ¦w − w

0

¦ < δ¦ and let U be the inverse image of V under the map f restricted to ¦z¦ ¦z − z

0

¦ < ¦.

By Theorem 7.4 p. 98, f maps U one-one onto V. Since f is continuous, U is a neighborhood of z

0

. By the

Open Mapping Theorem (Theorem 7.5 p. 99), f = ( f

−1

)

−1

is an open map, so f

−1

is continuous from V to

U. To ﬁnally show that it is analytic, we can use Proposition 3.7 p. 125. We have

f

−1

(w) =

1

2πi

_

¦z−z

0

¦=

z f

·

(w)

f (z) − w

dz.

Note, if we assume that ϕ is twice diﬀerentiable, then the result holds globally.

145

10.4 The Dirichlet Problem

Exercise 1. Let G = B(0; 1) and ﬁnd a barrier for G at each point of the boundary.

Solution. Not available.

Exercise 2. Let G = C − (∞, 0] and construct a barrier for each point of ∂

∞

G.

Solution. Not available.

Exercise 3. Let G be a region and a a point in ∂

∞

G such that there is a harmonic function u : G → R with

lim

z→a

u(z) = 0 and liminf

z→w

u(z) > 0 for all w in ∂

∞

G, w a. Show that there is a barrier for G at a.

Solution. Not available.

Exercise 4. This exercise asks for an easier proof of a special case of Theorem 4.9. Let G be a bounded

region and let a ∈ ∂G such that there is a point b with [a, b] ∩ G

−

= ¦a¦. Show that G has a barrier at a.

(Hint: Consider the transformation (z − a)(z − b)

1

.)

Solution. Not available.

10.5 Green’s Function

Exercise 1. (a) Let G be a simply connected region, let a ∈ G, and let f : G → D = ¦z : ¦z¦ < 1¦ be a

one-one analytic function such that f (G) = D and f (a) = 0. Show that the Green’s Function on G with

singularity at a is g

a

(z) = −log ¦ f (z)¦.

(b) Find the Green’s Functions for each of the following regions: (i) G = C−(∞, 0]; (ii) G = ¦z : Re z > 0¦;

(iii) G = ¦z : 0 < Im z < 2π¦.

Solution. Solution to part a):

Let G be simply connected, let a ∈ G, and let f : G → D = ¦z : ¦z¦ < 1¦ be a one-one analytic function such

that f (G) = D and f (a) = 0.

To show: g

a

(z) = −log ¦ f (z)¦ is a Green’s Function on G with singularity at a. Clearly, g

a

(z) has a singu-

larity if ¦ f (z)¦ = 0. This happens if f (z) = 0. By assumption f (z) = 0 iﬀ z = a. (Note that there is no other

point z ∈ G such that f (z) = 0 since f is assumed to be one-one). This implies g

a

(z) has a singularity at a.

It remains to verify a), b) and c) of Deﬁnition 5.1.

a) g

a

is harmonic in G − ¦a¦. This follows directly by Exercise 5 p. 255 in G − ¦a¦.

b) G(z) = g

a

(z)+log ¦z−a¦ is harmonic in a disk about a. Let the disk be B(a; r). Clearly g

a

(z) = −log ¦ f (z)¦ is

harmonic on B(a; r)−¦a¦. By Exercise 5 p. 255, we can argue again that log ¦z−a¦ is harmonic on B(a; r)−¦a¦

(choose f (z) = z − a and f (z) = 0 iﬀ z = a). Hence G(z) = g

a

(z) + log ¦z − a¦ = −log ¦ f (z)¦ + log ¦z − a¦ is

harmonic on B(a; r) − ¦a¦ as a sum of two harmonic functions. But

G(z) = −log ¦ f (z)¦ + log ¦z − a¦ = −log

_

¦ f (z)¦

¦z − a¦

_

= −log

_

¦z − a¦ ¦

˜

f (z)¦

¦z − a¦

_

= −log ¦

˜

f (z)¦

where f (z) = (z − a)

˜

f (z) by assumption and

˜

f (z) is analytic on G with no zero. So G(z) has a removable

singularity at a and therefore G(z) is harmonic on B(a; r).

c) lim

z→w

g

a

(z) = 0 for each w in ∂

∞

G. We have ¦ f (z)¦ = 1 ∀z ∈ ∂

∞

G by assumption ( f (∂

∞

G) = ∂D). Hence

lim

z→w

g

a

(z) = lim

w→a

−log ¦ f (w)¦ = −log 1 = 0

146

for each a in ∂

∞

G. This implies g

a

(z) = −log ¦ f (z)¦ is a Green’s Function on G with singularity at a.

Solution to part b) ii):

Clearly, G is a simply connected region (not the whole plane). If we can ﬁnd a map, say h : G → D, then

the Green’s Function, say g

a

, is given by

g

a

(z) = −log ¦h(z)¦

by Part 1 a). Note that h has to satisfy the following assumptions to use Part 1 a):

a) h : G → D one-one and analytic

b) h(a) = 0

c) h(G) = D.

The existence and uniqueness of this function h with the assumptions a), b) and c) is guaranteed by the

Riemann Mapping Theorem p. 160 Theorem 4.2. It remains to ﬁnd h:

Claim: h : G → D given by

h(z) = f (g(z))

where

g(z) =

z − 1

z + 1

and f (z) =

z − g(a)

1 − g(a)z

works. Thus, the Green’s Function for the region G = ¦z : Re(z) > 0¦ is given by g

a

(z) = −log ¦h(z)¦ where

h(z) is given by the Claim.

Proof of the Claim:

We will invoke the Orientation Principle (p. 53) to ﬁnd g : G → D where g is one-one and analytic. Then

g(z) =

z − 1

z + 1

(see p. 53 in the book for the derivation).

From p. 162 we have seen that the Möbius Transformation (which is one-one and analytic)

f (z) =

z − g(a)

1 − g(a)z

maps D onto D such that f (g(a)) = 0. Hence h(z) = f (g(z)) maps G onto D. In addition h is one-one,

analytic and h(a) = 0 and therefore satisﬁes a), b) and c).

Exercise 2. Let g

a

be the Green’s Function on a region G with singularity at z = a. Prove that if ψ is a

positive superharmonic function on G − ¦a¦ with liminf

z→a

[ψ(z) + log ¦z − a¦] > −∞, then g

a

(z) ≤ ψ(z) for

z a.

Solution. Not available.

Exercise 3. This exercise gives a proof of the Riemann Mapping Theorem where it is assumed that if G is

a simply connected region, G C, then: (i) C

∞

− G is connected, (ii) every harmonic function on G has a

harmonic conjugate, (iii) if a G then a branch of log(z − a) can be deﬁned.

(a) Let G be a bounded simply connected region and let a ∈ G; prove that there is a Green’s Function g

a

on

G with singularity at a. Let u(z) = g

a

(z) + log ¦z − a¦ and let v be the harmonic conjugate of u. If ϕ = u + iv

let f (z) = e

iα

(z − a)e

−ϕ(z)

for a real number α. (So f is analytic in G.) Prove that ¦ f (z)¦ = exp(−g

a

(z))

and that lim

z→w

¦ f (z)¦ = 1 for each w in ∂G (Compare this with Exercise 1). Prove that for 0 < r < 1,

C

r

= ¦z : ¦ f (z)¦ = r¦ consists of a ﬁnite number of simple closed curves in G (see Exercise VI.1.3). Let G

r

be a component of ¦z : ¦ f (z)¦ < r¦ and apply Rouche’s Theorem to get that f (z) = 0 and f (z) − w

0

= 0,

147

¦w

0

¦ < r, have the same number of solutions in G

r

. Prove that f is one-one on G

r

. From here conclude that

f (G) = D = ¦z : ¦z¦ < 1¦ and f

·

(a) > 0, for a suitable choice of α.

(b) Let G be a simply connected region with G C, but assume that G is unbounded and 0, ∞ ∈ ∂

∞

G. Let l

be a branch of log z on G, a ∈ G, and α = l(a). Show that l is one-one on G and l(z) α + 2πi for any z in

G. Prove that ϕ(z) = [l(z) − α − 2πi]

−1

is a conformal map of G onto a bounded simply connected region in

the plane. (Show that l omits all values in a neighborhood of α + 2πi.)

(c) Combine parts (a) and (b) to prove the Riemann Mapping Theorem.

Solution. Not available.

Exercise 4. (a) Let G be a region such that ∂G = γ is a simple continuously diﬀerentiable closed curve. If

f : ∂G → R is continuous and g(z, a) = g

a

(z) is the Green’s Function on G with singularity at a, show that

h(a) =

_

γ

f (z)

∂g

∂n

(z, a) ds (5.5)

is a formula for the solution of the Dirichlet Problem with boundary values f ; where

∂g

∂n

is the derivative of

g in the direction of the outward normal to γ and ds indicates that the integral is with respect to arc length.

(Note: these concepts are not discussed in this book but the formula is suﬃciently interesting so as to merit

presentation.) (Hint: Apply Green’s formula

Ω

[u∆v − v∆u] dx dy =

_

∂Ω

_

u

∂v

∂n

− v

∂u

∂n

_

ds

with Ω = G − ¦z : ¦z − a¦ ≤ δ¦, δ < d(a, ¦γ¦), u = h, v = g

a

(z) = g(z, a).)

(b) Show that if G = ¦z : ¦z¦ < 1¦ then (5.5) reduces to equation (2.5).

Solution. Not available.

148

Chapter 11

Entire Functions

11.1 Jensen’s Formula

Exercise 1. In the hypothesis of Jensen’s Formula, do not suppose that f (0) 0. Show that if f has a zero

at z = 0 of multiplicity m then

log

¸

¸

¸

¸

¸

¸

f

(m)

(0)

m!

¸

¸

¸

¸

¸

¸

+ mlog r = −

n

k=1

log

_

r

¦a

k

¦

_

+

1

2π

_

2π

0

log ¦ f (re

iθ

)¦ dθ.

Solution. We have

r(z−a

k

)

r

2

−¯ a

k

z

maps B(0; r) onto itself and takes the boundary to the boundary. Let

F(z) = f (z)

r

m

z

m

n

_

k=1

r

2

− ¯ a

k

z

r(z − a

k

)

.

Then F ∈ H(G) and F has no zeros in B(0; r) and ¦F(z)¦ = ¦ f (z)¦ if ¦z¦ = r. Thus, by a known result, we have

log ¦F(0)¦ =

1

2π

_

2π

0

log ¦F(re

iθ

)¦ dθ =

1

2π

_

2π

0

log ¦ f (re

iθ

)¦ dθ.

Also

g(z) ≡

f (z)

z

m

=

_

∞

i=0

f

(i)

(0)z

i

i!

z

m

=

_

∞

i=m

f

(i)

(0)z

i

i!

z

m

=

∞

i=m

f

(i)

(0)z

i−m

i!

=

f

(m)

(0)

m!

z

0

+

f

(m+1)

(0)

(m + 1)!

z

1

+ . . .

( f has a zero at z = 0 of multiplicity m) which implies g(0) =

f

(m)

(0)

m!

. Thus,

F(0) = g(0)r

m

n

_

k=1

_

−

r

a

k

_

.

149

Therefore

¦F(0)¦ = ¦g(0)¦r

m

n

_

k=1

r

¦a

k

¦

implies

log ¦F(0)¦ = log ¦g(0)¦ + mlog r +

n

k=1

log

r

¦a

k

¦

= log

¸

¸

¸

¸

¸

¸

f

(m)

(0)

m!

¸

¸

¸

¸

¸

¸

+ mlog r +

n

k=1

log

r

¦a

k

¦

.

Hence,

log ¦F(0)¦ = log

¸

¸

¸

¸

¸

¸

f

(m)

(0)

m!

¸

¸

¸

¸

¸

¸

+ mlog r +

n

k=1

log

r

¦a

k

¦

=

1

2π

_

2π

0

log ¦ f (re

iθ

)¦ dθ

which is equivalent to

log

¸

¸

¸

¸

¸

¸

f

(m)

(0)

m!

¸

¸

¸

¸

¸

¸

+ mlog r = −

n

k=1

log

r

¦a

k

¦

+

1

2π

_

2π

0

log ¦ f (re

iθ

)¦ dθ

which proves the assertion.

Exercise 2. Let f be an entire function, M(r) = sup¦¦ f (re

iθ

)¦ : 0 ≤ θ ≤ 2π¦, n(r) = the number of zeros of f

in B(0; r) counted according to multiplicity. Suppose that f (0) = 1 and show that n(r) log 2 ≤ log M(2r).

Solution. Since f is an entire function, it is analytic on

¯

B(0; 2r). Suppose a

1

, . . . , a

n

are the zeros of f in

B(0; r) repeated according to multiplicity and b

1

, . . . , b

m

are the zeros of f in B(0; 2r) − B(0; r). That is,

a

1

, . . . , a

n

, b

1

, . . . , b

m

are the zeros of f in B(0; 2r) repeated according to multiplicity. Since f (0) = 1 0,

we have by Jensen’s formula

0 = log 1 = log ¦ f (0)¦ = −

n

k=1

log

_

2r

¦a

k

¦

_

−

m

k=1

log

_

2r

¦b

k

¦

_

+

1

2π

_

2π

0

log ¦ f (2re

iθ

)¦ dθ

which implies

n

k=1

log

_

2r

¦a

k

¦

_

= −

m

k=1

log

_

2r

¦b

k

¦

_

.,,.

≥ 0 since r ≤ ¦b

k

¦ < 2r

+

1

2π

_

2π

0

log ¦ f (2re

iθ

)¦ dθ

≤

1

2π

_

2π

0

log ¦ f (2re

iθ

)¦ dθ

≤

Def of M(r)

1

2π

log ¦M(2r)¦

_

2π

0

dθ = log ¦M(2r)¦.

Hence,

n

k=1

log

_

2r

¦a

k

¦

_

≤ log ¦M(2r)¦.

But

n

k=1

log

_

2r

¦a

k

¦

_

=

n(r)

k=1

log

_

2r

¦a

k

¦

_

= log

n(r)

_

k=1

_

2r

¦a

k

¦

_

= log

_

¸

¸

¸

¸

¸

¸

_

2

n

n(r)

_

k=1

r

¦a

k

¦

_

¸

¸

¸

¸

¸

¸

_

= log

_

2

n(r)

_

+ log

n(r)

_

k=1

2r

¦a

k

¦

≥ n(r) log(2)

150

since 0 < ¦a

k

¦ < r implies

r

¦a

k

¦

> 1 ∀k which means

n(r)

k=1

r

¦a

k

¦

> 1 so the logarithm is greater than zero. Thus,

n(r) log(2) ≤ log ¦M(2r)¦

which proves the assertion.

Exercise 3. In Jensen’s Formula do not suppose that f is analytic in a region containing

¯

B(0; r) but only

that f is meromorphic with no pole at z = 0. Evaluate

1

2π

_

2π

0

log ¦ f (re

iθ

)¦ dθ.

Solution. Not available.

Exercise 4. (a) Using the notation of Exercise 2, prove that

_

r

0

n(t)

t

dt =

n

k=1

log

_

r

¦a

k

¦

_

where a

1

, . . . , a

n

are the zeros of f in B(0; r).

(b) Let f be meromorphic without a pole at z = 0 and let n(r) be the number of zeros of f in B(0; r) minus

the number of poles (each counted according to multiplicity). Evaluate

_

r

0

n(t)

t

dt.

Solution. Not available.

Exercise 5. Let D = B(0; 1) and suppose that f : D → C is an analytic function which is bounded.

(a) If ¦a

n

¦ are the non-zero zeros of f in D counted according to multiplicity, prove that

_

(1 − ¦a

n

¦) < ∞.

(Hint: Use Proposition VII. 5.4).

(b) If f has a zero at z = 0 of multiplicity m ≥ 0, prove that f (z) = z

m

B(z) exp(−g(z)) where B is a Blaschke

Product (Exercise VII. 5.4) and g is an analytic function on D with Re g(z) ≥ −log M (M = sup¦¦ f (z)¦ :

¦z¦ < 1¦).

Solution. Not available.

11.2 The genus and order of an entire function

Exercise 1. Let f (z) =

_

c

n

z

n

be an entire function of ﬁnite genus µ; prove that

lim

n→∞

c

n

(n!)

1/(µ+1)

= 0.

(Hint: Use Cauchy’s Estimate.)

Solution. Deﬁne M(r) = max¦ f (re

iθ

) : 0 ≤ θ ≤ 2π¦ = max

¦z¦=r

¦ f (z)¦. For suﬃciently large r = ¦z¦, we have

M(r) < e

αr

µ+1

(11.1)

by Theorem 2.6 p. 283. By Cauchy’s estimate p. 73, we have

¦c

n

¦ ≤

M(r)

r

n

151

where c

n

=

f

(n)

(0)

n!

. Hence,

¦c

n

¦ ≤

M(r)

r

n

<

(11.1)

e

αr

µ+1

. (11.2)

Deﬁne, r = n

1

µ+1

λ, where λ is a positive constant, and choose λ > e for convenience. Then (11.2) can be

written in the form

¦c

n

¦ <

e

αnλ

µ+1

(n

n

)

1

µ+1

λ

n

⇐⇒ (n

n

)

1

µ+1

¦c

n

¦ <

e

αnλ

µ+1

λ

n

=

_

¸

¸

¸

¸

_

e

αλ

µ+1

λ

_

¸

¸

¸

¸

_

n

.

Now, if we choose α such that αλ

µ+1

= 1, the last inequality becomes

(n

n

)

1

µ+1

¦c

n

¦ <

_

e

λ

_

n

<

for any > 0, provided n is taken large enough, n > N say. Since n! < n

n

, for n ≥ 2, we have

(n!)

1

µ+1

¦c

n

¦ <

for n > N or

lim

n→∞

c

n

(n!)

1/(µ+1)

= 0.

Exercise 2. Let f

1

and f

2

be entire functions of ﬁnite orders λ

1

and λ

2

respectively. Show that f = f

1

+ f

2

has ﬁnite order λ and λ ≤ max(λ

1

, λ

2

). Show that λ = max(λ

1

, λ

2

) if λ

1

λ

2

and give an example which

shows that λ < max(λ

1

, λ

2

) with f 0.

Solution. Not available.

Exercise 3. Suppose f is an entire function and A, B, a are positive constants such that there is a r

0

with

¦ f (z)¦ ≤ exp(A¦z¦

a

+ B) for ¦z¦ > r

0

. Show that f is of ﬁnite order ≤ a.

Solution. Not available.

Exercise 4. Prove that if f is an entire function of order λ then f

·

also has order λ.

Solution. Let f be an entire function of order λ, then f has power series

f (z) =

∞

n=0

c

n

z

n

.

By Exercise 5 e) p. 286 we have

α = liminf

n→∞

−log ¦c

n

¦

n log n

=

1

λ

.

Diﬀerentiating f yields

f

·

(z) =

∞

n=1

nc

n

z

n−1

=

∞

n=0

(n + 1)c

n+1

.,,.

d

n

z

n

.

152

Now, let λ

·

be the order of f

·

, then

1

λ

·

= liminf

n→∞

−log ¦d

n

¦

n log n

= liminf

n→∞

−log ¦(n + 1)c

n+1

¦

n log n

= liminf

n→∞

−log(n + 1) − log ¦c

n+1

¦

n log n

= liminf

n→∞

_

−log(n + 1)

n log n

−

log ¦c

n+1

¦

n log n

(n + 1) log(n + 1)

(n + 1) log(n + 1)

_

= liminf

n→∞

_

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

_

−log(n + 1)

n log n

.,,.

→0

+

−log ¦c

n+1

¦

(n + 1) log(n + 1)

.,,.

→

1

λ

n + 1

n

.,,.

→1

log(n + 1)

log(n)

.,,.

→1

_

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

_

=

1

λ

.

Hence, λ

·

= λ and therefore f

·

also has order λ.

Exercise 5. Let f (z) =

_

c

n

z

n

be an entire function and deﬁne the number α by

α = liminf

n→∞

−log ¦c

n

¦

n log n

(a) Show that if f has ﬁnite order then α > 0. (Hint: If the order of f is λ and β > λ show that ¦c

n

¦ ≤

r

−n

exp(r

β

) for suﬃciently large r, and ﬁnd the maximum value of this expression.)

(b) Suppose that 0 < α < ∞ and show that for any > 0, < α, there is an integer p such that ¦c

n

¦

1/n

<

n

−(α−)

for n > p. Conclude that for ¦z¦ = r > 1 there is a constant A such that

¦ f (z)¦ < Ar

p

+

∞

n=1

_

r

n

α−

_

n

(c) Let p be as in part (b) and let N be the largest integer ≤ (2r)

1/(α−)

. Take r suﬃciently large so that

N > p and show that

∞

n=N+1

_

r

n

α−

_

n

< 1 and

N

n=p+1

_

r

n

α−

_

n

< Bexp

_

(2r)

1/(α−)

_

where B is a constant which does not depend on r.

(d) Use parts (b) and (c) to show that if 0 < α < ∞then f has ﬁnite order λ and λ ≤ α

−1

.

(e) Prove that f is of ﬁnite order iﬀ α > 0, and if f has order λ then λ = α

−1

.

Solution. Not available.

Exercise 6. Find the order of each of the following functions: (a) sin z; (b) cos z; (c) cosh

√

z; (d)

_

∞

n=1

n

−an

z

n

where a > 0. (Hint: For part (d) use Exercise 5.)

Solution. c) The order is λ =

1

2

. Note that sinh

2

z = cosh

2

z −1 and that cos x is bounded for real values x,

so

¦ cos z¦

2

= cos

2

x + sinh

2

y ≤ cosh

2

y ≤ cosh

2

¦z¦.

153

Thus

¦ cos z¦ ≤ cosh ¦z¦ =

1

2

(e

¦z¦

+ e

−¦z¦

) ≤

1

2

e

¦z¦

≤ e

¦z¦

,

and therefore

¦ cos

√

z¦ ≤ exp(¦z¦

1/2

).

This implies that the order λ ≤

1

2

. Next we want to show that λ =

1

2

. Seeking contradiction assume that

there is > 0 and r > 0 such that

¦ cos

√

z¦ ≤ exp(¦z¦

1

2

−

) (11.3)

for all ¦z¦ > r.

Let z = −n

2

for an n ∈ N with n > max¦

√

r, 4, 2

1/(2)

¦ then in particular (i) ¦z¦ > r, (ii) −ln 2 + n >

n

2

, and

(iii)

1

2

>

1

n

2

and

¦ f (z)¦ = ¦ cos

√

z¦ =

1

2

¦e

n

+ e

−n

¦ >

1

2

e

n

= e

−log 2+n

>

(ii)

e

n

2

>

(iii)

exp(n

−2

n) = exp(n

2(

1

2

−)

) = exp(¦z¦

1

2

−

).

This shows that no > 0 can be found that satisﬁes equation (11.3). This implies λ =

1

2

.

Exercise 7. Let f

1

and f

2

be entire functions of ﬁnite order λ

1

, λ

2

; show that f = f

1

f

2

has ﬁnite order

λ ≤ max(λ

1

, λ

2

).

Solution. Deﬁne M(r) = max

¦z¦=r

¦ f (z)¦, M

1

(r) = max

¦z¦=r

¦ f

1

(z)¦ and M

2

(r) = max

¦z¦=r

¦ f

2

(z)¦. Since f

1

and

f

2

are entire functions of ﬁnite order λ

1

and λ

2

, respectively, we have for > 0,

M

1

(r) < e

r

λ

1

+/2

(11.4)

and

M

2

(r) < e

r

λ

2

+/2

(11.5)

for suﬃciently large r = ¦z¦ by Proposition 2.14. Hence

M(r) = max

¦z¦=r

¦ f (z)¦ = max

¦z¦=r

¦ f

1

(z) f

2

(z)¦

≤ max

¦z¦=r

¦ f

1

(z)¦ max

¦z¦=r

¦ f

2

(z)¦

= M

1

(r)M

2

(r) <

by (11.4),(11.5)

e

r

λ

1

+/2

e

r

λ

2

+/2

= e

r

λ

1

+/2

+r

λ

2

+/2

≤

λ≤max(λ

1

,λ

2

)

e

2r

λ+/2

< e

r

λ+

provided r is suﬃciently large. Hence, the order of f (z) = f

1

(z) f

2

(z) has ﬁnite order λ ≤ max(λ

1

, λ

2

).

Exercise 8. Let ¦a

n

¦ be a sequence of non-zero complex numbers. Let ρ = inf¦a :

_

¦a

n

¦

−a

< ∞¦; the

number called the exponent of convergence of ¦a

n

¦.

(a) If f is an entire function of rank p then the exponent of convergence ρ of the non-zero zeros of f satisﬁes:

p ≤ ρ ≤ p + 1.

(b) If ρ

·

= the exponent of convergence of ¦a

n

¦ then for every > 0,

_

¦a

n

¦

−(ρ+)

< ∞and

_

¦a

n

¦

−(ρ−)

= ∞.

(c) Let f be an entire function of order λ and let ¦a

1

, a

2

, . . .¦ be the non-zero zeros of f counted according

to multiplicity. If ρ is the exponent of convergence of ¦a

n

¦ prove that ρ ≤ λ. (Hint: See the proof of (3.5) in

the next section.)

154

(d) Let P(z) =

∞

n=1

E

p

(z/a

n

) be a canonical product of rank p, and let ρ be the exponent of convergence of

¦a

n

¦. Prove that the order of P is ρ. (Hint: If λ is the order of P, ρ ≤ λ; assume that ¦a

l

¦ ≤ ¦a

2

¦ ≤ . . . and ﬁx

z, ¦z¦ > 1. Choose N such that ¦a

n

¦ ≤ 2¦z¦ if n ≤ N and ¦a

n

¦ > 2¦z¦ if n ≤ N + 1. Treating the cases ρ < p + 1

and ρ = p + 1 separately, use (2.7) to show that for some > 0

∞

n=N+1

log

¸

¸

¸

¸

¸

¸

E

p

_

z

a

n

_

¸

¸

¸

¸

¸

¸

< A¦z¦

ρ+

.

Prove that for ¦z¦ ≥

1

2

, log ¦E

p

(z)¦ < B¦z¦

p

where B is a constant independent of z. Use this to prove that

N

n=1

log

¸

¸

¸

¸

¸

¸

E

p

_

z

a

n

_

¸

¸

¸

¸

¸

¸

< C¦z¦

ρ+

for some constant C independent of z.)

Solution. Not available.

Exercise 9. Find the order of the following entire functions:

(a)

f (z) =

∞

_

n=1

(1 − a

n

z), 0 < ¦a¦ < 1

b)

f (z) =

∞

_

n=1

_

1 −

z

n!

_

c)

f (z) = [Γ(z)]

−1

.

Solution. b) Considering g(z) :=

∞

m=1

_

1 −

z

n!

_

one notes that g(z) is already in form of the canonical

product with the entire function in the exponent being constantly zero and p = 0. Hence also the genus

µ = 0. The simple zeros of g(z) are at z = n! and from Exercise 8a) it follows that λ ≤ 1. In fact we will

show that λ = 0. It suﬃces to show that the exponent of convergence ρ = 0 and to employ Exercise 8d).

Let m ≤ N, let n > 4m then

n! ≥ n (n − 1) 2

.,,.

≥n

(n − 2) 3

.,,.

≥n

. . . (n − (2k − 1)) 2m

.,,.

≥n

.

Each product of consecutive numbers is of the form (n − (k − 1))k, for 1 ≤ k ≤ 2m. By choice of n,

k

n

≤

1

2

and n − k + 1 > 2m which justiﬁes the individual estimates by n. In total this yields n! > n

2m

for n large, or

equivalently

∞

n=4m

(n!)

−

1

m

≤

∞

n=4m

n

−2

and the sumon the left converges for every positive integer m. We conclude that ρ = inf¦m : the left sum converges¦ =

0. By Part d) of Exercise 8 we conclude that the order of g(z) is zero.

155

11.3 Hadamard Factorization Theorem

Exercise 1. Let f be analytic in a region G and suppose that f is not identically zero. Let G

0

= G − ¦z :

f (z) = 0¦ and deﬁne h : G

0

→ R by h(z) = log ¦ f (z)¦. Show that

∂h

∂x

− i

∂h

∂y

=

f

·

f

on G

0

.

Solution. Let f be analytic in a region G and suppose that f is not identically zero. Let G

0

= G − ¦z :

f (z) = 0¦, then h : G

0

→ R given by h(z) = log ¦ f (z)¦ is well deﬁned as well as

f

·

f

is well deﬁned on G

0

.

Let f = u(x, y) + iv(x, y) = u + iv. Since f is analytic, the Cauchy-Riemann (C-R) equations u

x

= v

y

and

u

y

= −v

x

are satisﬁed. We have by p. 41 Equation 2.22 and 2.23

f

·

= u

x

+ iv

x

and f

·

= −iu

y

+ v

y

and thus

2f

·

= u

x

+ iv

x

− iu

y

+ v

y

implies

f

·

=

1

2

_

u

x

+ iv

x

− iu

y

+ v

y

_

=

(C-R)

1

2

_

u

x

− iu

y

− iu

y

+ u

x

_

=

1

2

_

2u

x

− 2iu

y

_

= u

x

− iu

y

.

Therefore,

f

·

f

=

u

x

− iu

y

u + iv

. (11.6)

Next, we calculate

∂h

∂x

= h

x

and

∂h

∂y

= h

y

where h(z) = log ¦ f (z)¦ =

1

2

log(u

2

+ v

2

). Using the chain rule, we

get

h

x

= h

u

u

x

+ h

v

v

x

=

u

u

2

+ v

2

u

x

+

v

u

2

+ v

2

v

x

and

h

y

= h

u

u

y

+ h

v

v

y

=

u

u

2

+ v

2

u

y

+

v

u

2

+ v

2

v

y

and hence

∂h

∂x

− i

∂h

∂y

= h

x

− ih

y

=

u

u

2

+ v

2

u

x

+

v

u

2

+ v

2

v

x

− i

u

u

2

+ v

2

u

y

− i

v

u

2

+ v

2

v

y

=

u

u

2

+ v

2

(u

x

− iu

y

) +

v

u

2

+ v

2

(v

x

− iv

y

)

=

(C-R)

u

u

2

+ v

2

(u

x

− iu

y

) +

v

u

2

+ v

2

(−u

y

− iu

x

)

=

u

u

2

+ v

2

(u

x

− iu

y

) −

iv

u

2

+ v

2

(u

x

− iu

y

)

=

u − iv

u

2

+ v

2

(u

x

− iu

y

)

=

u − iv

(u − iv)(u + iv)

(u

x

− iu

y

)

=

u

x

− iu

y

u + iv

.

156

Compare this with (11.6) and we obtain the assertion

∂h

∂x

− i

∂h

∂y

=

f

·

f

.

Exercise 2. Refer to Exercise 2.7 and show that if λ

1

λ

2

then λ = max(λ

1

; λ

2

).

Solution. Not available.

Exercise 3. (a) Let f and g be entire functions of ﬁnite order λ and suppose that f (a

n

) = g(a

n

) for a

sequence ¦a

n

¦ such that

_

¦a

n

¦

−(λ+1)

= ∞. Show that f = g.

(b) Use Exercise 2.8 to show that if f , g and ¦a

n

¦ are as in part (a) with

_

¦a

n

¦

−(λ+)

= ∞ for some > 0

then f = g.

(c) Find all entire functions f of ﬁnite order such that f (log n) = n.

(d) Give an example of an entire function with zeros ¦log 2, log 3, . . .¦ and no other zeros.

Solution. a) Since f and g are entire functions of ﬁnite order λ, also F = f −g is an entire function of ﬁnite

order λ (see Exercise 2 p. 286). Suppose that f (a

n

) = g(a

n

) for a sequence ¦a

n

¦ such that

_

¦a

n

¦

−(λ+1)

= ∞.

Assume F 0 ( f g), we will derive a contradiction. Since F 0, the sequence ¦a

n

¦ are the zeros of F,

because

F(a

n

) = f (a

n

) − g(a

n

) = 0

since f (a

n

) = g(a

n

) by assumption. By Hadamard’s Factorization Theorem (p. 289), F has ﬁnite genus

µ ≤ λ since F ∈ H(C) with ﬁnite order λ. Since by deﬁnition µ = max(p, q) where p is the rank of F, we

certainly have

p ≤ λ.

By the deﬁnition of the rank (p. 281 Deﬁnition 2.1), we have

∞

n=1

¦a

n

¦

−(p+1)

< ∞,

where a

n

’s are the zeros of F. Therefore by the comparison test, we also have

∞

n=1

¦a

n

¦

−(λ+1)

≤

p≤λ

∞

n=1

¦a

n

¦

−(p+1)

< ∞

so

_

∞

n=1

¦a

n

¦

−(λ+1)

< ∞ contradicting the assumption

_

∞

n=1

¦a

n

¦

−(λ+1)

= ∞. Hence F ≡ 0, and therefore we

obtain f = g.

b) Since f and g are entire functions of ﬁnite order λ, we also have F = f − g is an entire function of ﬁnite

order λ (see Exercise 2 p. 286). Suppose that f (a

n

) = g(a

n

) for a sequence ¦a

n

¦ such that

_

¦a

n

¦

−(λ+)

= ∞.

Assume F 0 ( f g), we will derive a contradiction. Since F 0, the sequence ¦a

n

¦ are the zeros of F,

because

F(a

n

) = f (a

n

) − g(a

n

) = 0

since f (a

n

) = g(a

n

) by assumption. Let ρ be the exponent of convergence of ¦a

n

¦, then

ρ ≤ λ

by Exercise 8 c) p. 286-287. Hence, according to Exercise 8 b) p. 286-287, we have

∞

n=1

¦a

n

¦

−(ρ+)

< ∞.

157

Therefore by the comparison test, we also have

∞

n=1

¦a

n

¦

−(λ+)

≤

ρ≤λ

∞

n=1

¦a

n

¦

−(ρ+)

< ∞.

Hence

_

∞

n=1

¦a

n

¦

−(λ+)

< ∞ contradicting the fact

_

∞

n=1

¦a

n

¦

−(λ+)

= ∞. Thus, F ≡ 0 and therefore we get the

assertion f = g.

158

Chapter 12

The Range of an Analytic Function

12.1 Bloch’s Theorem

Exercise 1. Examine the proof of Bloch’s Theorem to prove that L ≥ 1/24.

Solution. Mimic the proof of Block’s Theorem up to the application of Schwarz’s Lemma, i.e. set K(r) =

max¦¦ f

·

(z)¦¦¦z¦ = r¦, h(r) = (1 − r)K(r), r

0

= sup¦r : h(r) = 1¦. Choose a with ¦a¦ = r

0

, ¦ f

·

(a)¦ = K(r

0

) =

1

1−r

0

.

Also choose ρ

0

=

1

2

(1 − r

0

) and infer that for all z ∈ B(a, ρ)

¦ f

·

(z)¦ ≤

1

ρ

0

, and ¦ f

·

(z) − f

·

(a)¦ <

3¦z − a¦

2ρ

2

0

.

For z ∈ B(0, ρ

0

) deﬁne g(z) = f (z + a) − f (a). By convexity of B(a, ρ

0

) the line segment γ = [a, a + z] is

completely contained in B(a, ρ

0

). It follows that

¦g(z)¦ = ¦

_

γ f

·

(w) dw¦ ≤

1

ρ

0

¦z¦ ≤ 1 =: M.

The function g is not necessarily injective on B(0, ρ

0

) but the problem only asks for an estimate on the

Landau constant for which injectivity is not required. In this setting Lemma 1.2 gives

g(B(0, ρ

0

)) ⊃ B

_

0,

ρ

2

¦g

·

(0)¦

2

6M

_

= B

_

0,

1

4 6

_

= B

_

0,

1

24

_

.

This statement rewritten for f yields

f (B(a, ρ

0

)) ⊃ B

_

f (a),

1

24

_

.

Exercise 2. Suppose that in the statement of Bloch’s Theorem it is only assumed that f is analytic on D.

What conclusion can be drawn? (Hint: Consider the functions f

s

(z) = s

−1

f (sz), 0 < s < 1.) Do the same

for Proposition 1.10.

Solution. Not available.

159

12.2 The Little Picard Theorem

Exercise 1. Show that if f is a meromorphic function on C such that C

∞

− f (C) has at least three points

then f is a constant. (Hint: What if ∞ f (C)?)

Solution. First consider an entire function f that misses three points in C

∞

, one on which is ∞. Then there

are two distinct numbers a, b ∈ C that are not in the image of f . By Little Picard’s Theorem the function f

must be a constant.

Next consider a meromorphic function on C that is not entire. Hence ∞ ∈ f (C) and by assumption there

are distinct a, b, c ∈ C − f (C). Deﬁne a function

g(z) :=

1

f (z) − a

, z ∈ C

then g is entire because f (z) − a 0 for all z ∈ C and

1

b−a

,

1

c−a

∈ C − g(C). Again using Little Picard’s

Theorem conclude that g is a constant function. It cannot be the zero-function, because f was assumed

to be meromorphic, hence f (z) ∞ for at least some z ∈ C. Thus g is a nonzero constant function and

f (z) = g(z)

−1

+ a is also a constant function and the result is established.

Exercise 2. For each integer n ≥ 1 determine all meromorphic functions f and g on C with a pole at ∞

such that f

n

+ g

n

= 1.

Solution. Not available.

12.3 Schottky’s Theorem

No exercises are assigned in this section.

12.4 The Great Picard Theorem

Exercise 1. Let f be analytic in G = B(0; R) − ¦0¦ and discuss all possible values of the integral

1

2πi

_

γ

f

·

(z)

f (z) − a

dz

where γ is the circle ¦z¦ = r < R and a is any complex number. If it is assumed that this integral takes on

certain values for certain numbers a, does this imply anything about the nature of the singularity at z = 0?

Solution. Not available.

Exercise 2. Show that if f is a one-one entire function then f (z) = az +b for some constants a and b, a 0.

Solution. Not available.

Exercise 3. Prove that the closure of the set / in Theorem 4.1 equals / together with the constant functions

∞, 0, and 1.

Solution. Not available.

160

PREFACE

Most of the exercises I solved were assigned homeworks in the graduate courses Math 713 and Math 714 ”Complex Analysis I and II” at the University of Wisconsin – Milwaukee taught by Professor Dashan Fan in Fall 2008 and Spring 2009. The solutions manual is intented for all students taking a graduate level Complex Analysis course. Students can check their answers to homework problems assigned from the excellent book “Functions of One Complex Variable I”, Second Edition by John B. Conway. Furthermore students can prepare for quizzes, tests, exams and ﬁnal exams by solving additional exercises and check their results. Maybe students even study for preliminary exams for their doctoral studies. However, I have to warn you not to copy straight of this book and turn in your homework, because this would violate the purpose of homeworks. Of course, that is up to you. A I strongly encourage you to send me solutions that are still missing to kleefeld@tu-cottbus.de (LTEXpreferred but not mandatory) in order to complete this solutions manual. Think about the contribution you will give to other students. If you ﬁnd typing errors or mathematical mistakes pop an email to kleefeld@tu-cottbus.de. The recent version of this solutions manual can be found at http://www.math.tu-cottbus.de/INSTITUT/kleefeld/Files/Solution.html. The goal of this project is to give solutions to all of the 452 exercises.

CONTRIBUTION

I thank (without special order) Christopher T. Alvin Martin J. Michael for contributions to this book.

2

Contents

1 The Complex Number System 1.1 The real numbers . . . . . . . . . . . . . . . . . . 1.2 The ﬁeld of complex numbers . . . . . . . . . . . 1.3 The complex plane . . . . . . . . . . . . . . . . . 1.4 Polar representation and roots of complex numbers 1.5 Lines and half planes in the complex plane . . . . . 1.6 The extended plane and its spherical representation 2 Metric Spaces and the Topology of C 2.1 Deﬁnitions and examples of metric spaces 2.2 Connectedness . . . . . . . . . . . . . . 2.3 Sequences and completeness . . . . . . . 2.4 Compactness . . . . . . . . . . . . . . . 2.5 Continuity . . . . . . . . . . . . . . . . . 2.6 Uniform convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 1 5 5 7 7 9 9 12 13 15 17 20 21 21 24 30 40 40 46 56 58 59 61 64 65 66 66 73 80

3

Elementary Properties and Examples of Analytic Functions 3.1 Power series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Analytic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Analytic functions as mappings. Möbius transformations . . . . . . . . . . . . . . . . . . Complex Integration 4.1 Riemann-Stieltjes integrals . . . . . . . . . . . . . . . . . . . . . . . 4.2 Power series representation of analytic functions . . . . . . . . . . . . 4.3 Zeros of an analytic function . . . . . . . . . . . . . . . . . . . . . . 4.4 The index of a closed curve . . . . . . . . . . . . . . . . . . . . . . . 4.5 Cauchy’s Theorem and Integral Formula . . . . . . . . . . . . . . . . 4.6 The homotopic version of Cauchy’s Theorem and simple connectivity 4.7 Counting zeros; the Open Mapping Theorem . . . . . . . . . . . . . 4.8 Goursat’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4

5

Singularities 5.1 Classiﬁcation of singularities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 The Argument Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3

. . . . . . . . . . . . . 122 8. . . . . . . . .3 Monodromy Theorem . . . . . . . . . . . . . 8 Runge’s Theorem 121 8. . . . . . . . . . . . . . . . . . . . . . . .6 Factorization of the sine function . . . . . . . . . . . . . . . . . . . . .2 Spaces of analytic functions . . . . . . . . . . . . . . . . 7. . . . . . . . . . . . . . . . . . . . . . . . 9. . . . . . . . . . . . . . . . . . . . . . 11 Entire Functions 149 11. . 9. . . . . .5 Green’s Function . . . . . . . . . . . . . . . . . . 9. . . . . . . .4 The Great Picard Theorem . . . . . . . 7. . . . . . . . . . . . . . . . . .2 The genus and order of an entire function . . .1 Schwarz Reﬂection Principle . . . . . . . . . . .2 Harmonic functions on a disk . . . . . . . . . . 4 . . . . .2 Simple connectedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 The gamma function . . . . . .3 Convex functions and Hadamard’s Three Circles Theorem 6. . . . . . . . . . . . . . . . . . . . . . . . . 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. . . . . . . . Ω) . . . . . . . . . . . . . . . . . . . . . . . . 151 11. . .3 Subharmonic and superharmonic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. . . . . . . . . . . . . . . 9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 The Little Picard Theorem . . . . .1 Jensen’s Formula . . . . . . . 123 Analytic Continuation and Riemann Surfaces 9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Runge’s Theorem . . . . . . . . . . . . . .1 The space of continuous functions C(G. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 11. . . . . . . . . . . . . . . . . . . . . . 7. 121 8. . . . . . . . .5 The Sheaf of Germs of Analytic Functions on an Open Set 9.6 Analytic Manifolds . . . . . . . . . . . .1 Basic properties of harmonic functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Mittag-Leﬄer’s Theorem . . . . . . . . . . . . . . . .8 The Riemann zeta function . . . . . . . . . . .1 Bloch’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. . . .4 The Dirichlet Problem . . . . . . . . . . . . . . . . . . . . . 12. . . . . . . . . . . . . . . . . 159 159 160 160 160 . . . 7. . . . . . . . . . . 7. . . . . . . . . . . . . . . . . . . . . . . . 82 82 84 86 88 91 91 95 98 101 104 108 110 118 7 Compactness and Convergence in the Space of Analytic Functions 7. . . . . . . . . . . . . . . . . . . . . . . . .5 The Weierstrass Factorization Theorem . . . . . . .4 The Phragmén-Lindelöf Theorem . . . . . . . 10. . . . . . . . . . .3 Hadamard Factorization Theorem .3 Schottky’s Theorem . . . . . . . . . . . . . . . . . . . . 12. . . . . . . . . . . . . . . . . . . . . . .3 Spaces of meromorphic functions . . . . . . . . . . . . . . . . . .2 Analytic Continuation Along a Path . .2 Schwarz’s Lemma . 6.4 The Riemann Mapping Theorem . . . . . . . . . . . . . . .7 Covering spaces . . . . . . . . . . . . . . . . . . 6. . . . . . . .4 Topological Spaces and Neighborhood Systems . . . 7. . . . . . . . . . . . . . . . 156 12 The Range of an Analytic Function 12. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 The Maximum Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 The Maximum Modulus Theorem 6. . . . . . . . 7. . . . . 128 128 129 130 130 131 132 133 135 135 139 142 146 146 9 10 Harmonic Functions 10. . . . . . . . . . . . . . . . . . .

i .1 The real numbers No exercises are assigned in this section. (a ∈ R). Then a) √ 6 −1 − i 3 n 1 + i .2 The ﬁeld of complex numbers Exercise 1. z . √ 2 2 1 z 1 Im z Re for 2 ≤ n ≤ 8. z z+a 7i + 1 2 n Solution. .Chapter 1 The Complex Number System 1. x Re(z) = 2 +y |z|2 y Im(z) = − 2 =− 2 2 x +y |z| = x2 b) z−a z+a z−a Im z+a Re c) Re z3 Im z3 = x3 − 3xy2 = Re(z)3 − 3Re(z)Im(z)2 1 = = |z|2 − a2 x2 + y2 − a2 = 2 2 + 2ax + a2 +y |z| + 2aRe(z) + a2 2ya 2Im(z)a = 2 2 + y2 + 2xa + a2 x |z| + 2aRe(z) + a2 x2 = 3x2 y − y3 = 3Re(z)2 Im(z) − Im(z)3 . Let z = x + iy. 1. Find the real and imaginary parts of the following: √ 3 1 z−a 3 3 + 5i −1 + i 3 . .

d) 3 + 5i 7i + 1 3 + 5i Im 7i + 1 Re e) √ −1 + i 3 3 Re 2 √ −1 + i 3 3 Im 2 √ −1 − i 3 6 Re 2 √ −1 − i 3 6 Im 2 0, 1, −1, 0, 1, −1, = 1 19 25 8 = − 25 =

= 0

f)

= 1

= 0

g)

Re (in )

=

n is odd n ∈ {4k : k ∈ Z} n ∈ {2 + 4k : k ∈ Z} n is even n ∈ {1 + 4k : k ∈ Z} n ∈ {3 + 4k : k ∈ Z}

Im (in )

=

2

h) 0, √ 2 − , 2 −1, √ 2 − 2 , 0, √ 2 , 2 1, 1, √ 2 , 2 0, √ 2 − 2 , −1, √ 2 − , 2 0, n=2 n=3 n=4 n=5 n=6 n=7 n=8 n=2 n=3 n=4 n=5 n=6 n=7 n=8

Re

1+i √ 2

n

=

Im

1+i √ 2

n

=

Exercise 2. Find the absolute value and conjugate of each of the following: 3−i i −2 + i; −3; (2 + i)(4 + 3i); √ ; (1 + i)6 ; i17 . ; i+3 2 + 3i Solution. It is easy to calculate: a) z = −2 + i, b) z = −3, c) z = (2 + i)(4 + 3i) = 5 + 10i, d 3−i z= √ , 2 + 3i z= f) z = (1 + i)6 = −8i, g) i17 = i, |z| = 1, z = −i ¯ Exercise 3. Show that z is a real number if and only if z = z. ¯ |z| = 8, z = 8i ¯ 1 3 i = + i, i + 3 10 10 |z| = |z| = 3, z = −3 ¯ z = 5 − 10i ¯ √ |z| = 5 5, 1 √ 110, 11 |z| = 1 √ 10, 10 |z| = √ 5,

z = −2 − i ¯

e)

3+i z= √ ¯ 2 − 3i z= ¯ 3 1 − i 10 10

3

Solution. Let z = x + iy. ⇒: If z is a real number, then z = x (y = 0). This implies z = x and therefore z = z. ¯ ¯ ⇔: If z = z, then we must have x + iy = x − iy for all x, y ∈ R. This implies y = −y which is true if y = 0 and ¯ therefore z = x. This means that z is a real number. Exercise 4. If z and w are complex numbers, prove the following equations: |z + w|2 |z − w|2 = |z|2 + 2Re(zw) + |w|2 . ¯ = |z|2 − 2Re(zw) + |w|2 . ¯

|z + w|2

+ |z − w|2 = 2 |z|2 + |w|2 .

¯ Solution. We can easily verify that z = z. Thus |z + w|2 z ¯ z ¯ z ¯ = (z + w)(z + w) = (z + w)(¯ + w) = z¯ + zw + w¯ + ww 2 2 2 2 ¯ = |z| + |w| + zw + zw = |z| + |w| + zw + zw ¯ ¯ ¯ ¯ zw + zw ¯ ¯ = |z|2 + |w|2 + zw + zw = |z|2 + |w|2 + 2 ¯ ¯ 2 = |z|2 + |w|2 + 2Re(zw) = |z|2 + 2Re(zw) + |w|2 . ¯ ¯ z ¯ z ¯ z ¯ = (z − w)(z − w) = (z − w)(¯ − w) = z¯ − zw − w¯ + ww 2 2 2 2 ¯ = |z| + |w| − zw − zw = |z| + |w| − zw − zw ¯ ¯ ¯ ¯ zw + zw ¯ ¯ ¯ = |z|2 + |w|2 − zw − zw = |z|2 + |w|2 − 2 ¯ 2 = |z|2 + |w|2 − 2Re(zw) = |z|2 − 2Re(zw) + |w|2 . ¯ ¯ = |z|2 + Re(zw) + |w|2 + |z|2 − Re(zw) + |w|2 ¯ ¯ 2 2 2 2 2 = |z| + |w| + |z| + |w| = 2|z| + 2|w|2 = 2 |z|2 + |w|2 .

|z − w|2

|z + w|2 + |z − w|2

Exercise 5. Use induction to prove that for z = z1 + . . . + zn ; w = w1 w2 . . . wn : |w| = |w1 | . . . |wn |; z = z1 + . . . + zn ; w = w1 . . . wn . ¯ ¯ ¯ ¯ ¯ ¯ Solution. Not available. z Exercise 6. Let R(z) be a rational function of z. Show that R(z) = R(¯) if all the coeﬃcients in R(z) are real. Solution. Let R(z) be a rational function of z, that is R(z) = an zn + an−1 zn−1 + . . . a0 bm zm + bm−1 zm−1 + . . . b0

where n, m are nonnegative integers. Let all coeﬃcients of R(z) be real, that is a0 , a1 , . . . , an , b0 , b1 , . . . , bm ∈ R. Then R(z) = = an zn + an−1 zn−1 + . . . a0 an zn + an−1 zn−1 + . . . a0 = m+b m−1 + . . . b bm z m−1 z 0 bm zm + bm−1 zm−1 + . . . b0 an zn + an−1 zn−1 + . . . a0 bm zm + bm−1 zm−1 + . . . b0 4 = an zn + an−1 zn−1 + . . . a0 ¯ ¯ = R(¯). z bm zm + bm−1 zm−1 + . . . b0 ¯ ¯

3 The complex plane Exercise 1. respectively.4 Polar representation and roots of complex numbers Exercise 1. Hence r = 1 and θ = 2kπ with k ∈ 6 {−3. therefore r6 cis(6θ) = 1. to the origin while |z − w| is the distance between z and w. we get zw · ¯ w w = |w|2 · z w ≥ 0 if w 0. using a rotation of the plane. The following table gives a list of principle values of arguments and the corresponding value of the root of the equation z6 = 1. w w. this is zw ≥ 0. Now let a be any complex number and. Not available. 2}. −2. Start with z6 = 1 and z = rcis(θ). Then t ≥ 0 and z = tw. Considering the construction of the implied triangle. −1. Find the sixth roots of unity. Describe the set of points z satisfying |z − a| − |z + a| = 2c for every possible choice of a and c. 0. it is necessary and suﬃcient that ⇐⇒ ||z| − |w|| = |z − w| (|z| − |w|)2 = |z − w|2 ⇐⇒ ⇐⇒ ⇐⇒ ⇐⇒ |z||w| = Re(zw) ¯ |zw| = Re(zw) ¯ ¯ (|z| − |w|)2 = |z|2 − 2Re(zw) + |w|2 ¯ 2 2 2 |z| − 2|z||w| + |w| = |z| − 2Re(zw) + |w|2 ¯ Exercise 2. θ0 = 0 z0 = 1 θ1 = π z1 = cis( π ) 3 3 2π z2 = cis( 2π ) θ2 = 3 3 θ3 = π z3 = cis(π) = −1 θ4 = −2π z4 = cis( −2π ) 3 3 z5 = cis( −π ) θ5 = −π 3 3 5 . Solution.4) and give necessary and suﬃcient conditions for equality. Then ||z| − |w|| = ||z − w + w| − |w|| ≤ ||z − w| + |w| − |w|| = ||z − w|| = |z − w| Notice that |z| and |w| is the distance from z and w. Solution. Prove (3. l ≤ n. If 1. in order to guarantee equality. for which zl 0. Equivalently.3) if and only if zk /zl ≥ 0 for any integers k and l. Not available. Show that equality occurs in (3. Let z and w be complex numbers.1. Exercise 3. 1. Multiplying this by ¯ z 1 2 z t = w = |w|2 · |w| · w . 1 ≤ k. Let a ∈ R and c > 0 be ﬁxed. Solution. Solution. describe the locus of points satisfying the above equation.

. . a. Calculate the following: a) the square roots of i b) the cube roots of i c) the square roots of √ 3 + 3i √ Solution.Exercise 2. . the second roots of z are given by √ √ 4 4 108 12 . Then r = |z| = 3 + 32 = 12 and α = tan−1 √ are given by 2 r cos α+2kπ + i sin α+2kπ where k = 0. cosn−1 θ sin θ − cosn−3 θ sin3 θ + . 1. Exercise 4. . the 2 distinct roots of z 3 + 2kπ + i sin 2 π 3 + 2kπ . n = cosn θ − = n 1 n 2 cosn−2 θ sin2 θ + n 3 cosn−4 θ sin4 θ − ..− 2 2 . a2 . in rectangular form. k!(n − k)! n 4 and compare the real and imaginary parts of each side of de Moivre’s formula to obtain the formulas: cos nθ sin nθ Solution. Use the binomial equation n (a + b)n = k=0 n n−k k a b. A primitive nth root of unity is a complex number a such that 1. Show that if a and b are primitive nth and mth roots of unity. an−1 are distinct nth roots of unity. 2 2 and − √ 4 √ 4 108 12 . then ab is a kth root of unity for some integer k.. n n √ √ 4 z = 12 cos π 3 3 √ 3 = π . Not available. c) The square roots of 3 + 3i. Let z = cis 2π for an integer n ≥ 2. √ √ √ 2 Let z = 3 + 3i. So. the square roots of z. Show that 1 + z + .. Exercise 3. + zn−1 = 0. . Not available. 6 6 2 So. What is the smallest value of k? What can be said if a and b are nonprimitive roots of unity? Solution. 6 . zk . 2 Therefore. Exercise 5. are given by √ √ √ 4 4 1 z0 = 12 cos π + i sin π = 12 23 + 2 i 6 6 √ √ √ 4 4 z1 = 12 cos 7π + i sin 7π = 12 − 23 − 1 i . respectively. . k where n k = n! . Speciﬁcally. .

r > 0. zn−1 can be calculated as z −1 z−1 . Not available. 2 . Exercise 2. Not available. It follows that 1 + z + z + . Similarly. Therefore. 3. then |z1 | = 0 and therefore 2y |z|2 − 1 2x . z. Solution. . Show that ϕ(t) = cis t is a group homomorphism of the additive group R onto the multiplicative group T = {z : |z| = 1}. Thus Re(zn ) = rn cos(nθ) ≥ 0 implies cos(nθ) ≥ 0 for all n. . 1. since r = 0 and Re(zn ) = 0. 7 . z2 ..}. 2 |z|2 + 1 |z| + 1 |z| + 1 Φ(z1 ) = (0. Find necessary and suﬃcient conditions in terms of β that Lβ be tangent to C at a. give the corresponding point of S : 0. show that z is a non-negative real number. 1 + i. If we choose n = k + 1. π/2). Not available. . let a = c + rcis α and put Lβ = z : Im z−a =0 b where b = cisβ. 0. Solution.Solution. Then 2π − π/k ≤ θ < 2π − π/(k + 1) for some k ∈ {2. Let n be an arbitrary but ﬁxed positive integer and let z ∈ C and Re(zn ) ≥ 0. Exercise 7. z = 2π n n th n n j−1 j=1 z n cis 2π n = = cis n · = cis(2π) = 1. This implies θ = 0 as we will show next. we have k π ≤ nθ < (k + 1)π k If z = 0. 3. So.6 The extended plane and its spherical representation Exercise 1.7) and (6. Clearly. then we are done. we have Re(zn ) = rn cos(nθ) ≥ 0. 1.. Solution. Exercise 6. assume z which is impossible since cos(nθ) ≥ 0.5 Lines and half planes in the complex plane Exercise 1. Since zn = rn (cos(nθ) + i sin(nθ)). 3 + 2i. . we can derive a contradiction if we assume θ ∈ (3π/2. . 3π/2]. The summation of the ﬁnite geometric sequence 1. Let C be the circle {z : |z − c| = r}. n We want to show that z is an n root of unity. For each of the following points in C. Solution. Solution. . then there exists a k ∈ {2. We have Φ(z) = If z1 = 0. . If π θ ∈ (0. then r > 0. θ [π/2. using de Moivre’s formula. .8). . −1). If z ∈ C and Re(zn ) ≤ 0 for every positive integer n.} such that k+1 ≤ θ < π . Give the details in the derivation of (6. 2π). . 0. . + z 2 n−1 = zn −1 z−1 = 1−1 z−1 = 0 as required.

0. Find the coordinates of W in terms of the coordinates of Z and Z . y2 + 1 y2 + 1 corresponds to the imaginary axes in C. −1) on the sphere S . then |z2 | = 2 and therefore Φ(z2 ) = 2 2 1 . 7. 0. 0. Exercise 5. Let Λ be a circle lying in S . Otherwise. . 7 7 7 on the sphere S . β3 ) is a vector orthogonal to P and l is some real number. Solution. 8 . 1 on the sphere S . √ If z2 = 1 + i. Thus Φ(z) = Therefore the set x2 2x x2 − 1 |x ∈ R ⊂ S . Solution. x2 . Solution.Thus. then z = x which implies |z|2 = x2 . z2 = 1 + i corresponds to the point 2 . y2 2y y2 − 1 |y ∈ R ⊂ S . Not available. . 3 . That means the unit circle x2 + z2 = 1 lying in the xz−plane corresponds to the real axes in C. Recall from analytic geometry that P = {(x1 . Therefore the set 0. 3 3 √ If z3 = 3 + 2i. . Λ projects onto a circle in C. Let W be the point on S corresponding to z + z . then z = iy which implies |z|2 = y2 . If z is on the imaginary axes. Exercise 3. Then there is a unique plane P in R3 such that P ∩ S = Λ. That means the unit circle y2 + z2 = 1 lying in the yz−plane corresponds to the imaginary axes in C. 2 +1 x +1 x2 − 1 2x . Use this information to show that if Λ contains the point N then its projection on C is a straight line. Let Z and Z be points on S corresponding to z and z respectively. 7 3 2 6 . Thus Φ(z) = 0. Which subsets of S correspond to the real and imaginary axes in C. It can be assumed that β2 +β2 +β2 = 1 2 3 1. If z is on the real axes. z3 = 3 + 2i corresponds to the point 3 2 6 7. then |z3 | = 13 and therefore Φ(z3 ) = Thus. 2 x2 + 1 x +1 corresponds to the real axes in C. Exercise 4. β2 . Not available. . 3 3 3 2 Thus. . z1 = 0 corresponds to the point (0. . + 1 y2 + 1 2y y2 − 1 . x3 ) : x1 β1 + x2 β2 + x3 β3 = l} where (β1 .

(b) B := {z ∈ C : z = x + iy. To show that C cannot be closed. Not available. (c) C := {z ∈ C : zn = 1 for some integer n ≥ 1} Claim: C is neither closed nor open. Example (1. y = 0} (the real axis) then B(z. Solution. then Imz complement is open. (e) {z ∈ C : z is real and 0 ≤ z ≤ 1}? Solution.1 Deﬁnitions and examples of metric spaces Exercise 1. εz ) ⊂ C − B. Which of the following subsets of C are open and which are closed: (a) {z : |z| < 1}. 9 .Hence B is closed since its ε ε For any real x and any ε > 0 the point x + i 2 ∈ B(x. Exercise 2. describe B(x. (d) {z ∈ C is real and 0 ≤ z < 1}. Thus B is not open. 0.2)-(1. εz ) ⊂ A is open and therefore A = z∈A B(z. Set εz = |Imz| 2 .6) is. then B(z. indeed. otherwise A and C − A were both closed and open sets yet C is connected. Now zm = 1 implies that sin(mx2π) = 0 which in turn means that mx ∈ Z contradicting the choice x ∈ R − Q. Also. Now ﬁx x ∈ R − Q and let {xn }n be a rational sequence that converges to x. Let m be any natural number.6) is the only one likely to give any diﬃculty. Show that each of the examples of metric spaces given in (1. We have (a) A := {z ∈ C : |z| < 1} Let z ∈ A and set εz = 1−|z| . We have constructed a sequence of elements of C that converges to a point that is not element of C. note that for cis p q 2π p q ∈ C since ∈ Q with p ∈ Z and q ∈ N the number z := zq = cis(p2π) = cos(p2π) + i sin(p2π) = 1. (c) {z : zn = 1 for some integer n ≥ 1}. εz ) is open also. A 2 cannot be closed.Chapter 2 Metric Spaces and the Topology of C 2. a metric space. (b) the real axis. r) for each of these examples. Hence C is not closed. ε) but x + i 2 ∈ C − R. C is not open because if zn = 1 then z = rcis(θ) with r = 1 and any ε-ball around z would contain an ε element z := (1 + 4 )cis(θ). Let z ∈ C − B.

z ) = [(1+|z|22|z−z | |2 )]1/2 = [(1+|z |2|z −z| 2 )]1/2 = d(z . Not available. z ∈ C. z ∈ C is a )(1+|z metric space. z ) = [(1+|z|22|z−z | |2 )]1/2 for z. d(z.1) . Exercise 5. Not available. d(∞. Also d(z. z ∈ C (the denominator is always positive). x) + d(x.1) and it is not closed because zn := 1 − sequence in D that converges to a point outside of D. |Imy| ) is completely contained in the complement of E. Solution. 2 We have d(z. We have d(z. Exercise 4. we have d(z. then B(z. z ). 2 Exercise 3. )(1+|z 2 Obviously. z). 6. Exercise 7. d) is any metric space show that every open ball is.7) and (I. Solution. To show (C∞ . Prove that a set G ⊂ X is open if and only if X − G is closed. d) with d(z. Solution. d(∞. we need the following equation (z − z )(1 + x x) = (z − x)(1 + xz ) + (x − z )(1 + xz). If (X. ∞) = (1+|z|2 )1/2 ≥ 0 for all z ∈ C. an open set. d) where d is given (I. First. Prove Proposition 1. in fact.1) and it is closed because its complement is open: If z E and z real. Also. 0) = 0 iﬀ z = ∞. x) + d(x. 1 n is a (e) E := {z ∈ C : z is real and 0 ≤ z ≤ 1} This set E is not open by the observation in 2.(d) D := {z ∈ C : z is real and 0 ≤ z < 1} The set cannot be open by the argument given in 2. Thus. z ) = [(1+|z|22|z−z | |2 )]1/2 ≥ 0 for all z. Not available. z ) = 0 iﬀ z = z . z ) in general. Not available. iii) d(z. z) by the symmetry of d(z. The ﬁrst inequality will be shown next. Solution. i) d(z.|x−1|} ) is contained in the complement of E. Therefore. Give the details of the proof of (1. ¯ ¯ ¯ 10 (2. z ) = [(1+|z|22|z−z | |2 )]1/2 = 0 iﬀ z = z . z ) for all x. z. ∞) = (1+|z|2 )1/2 = 2 )(1+|z| )(1+|z d(∞. ∞) = )(1+|z 2 limz→∞ (1+|z|2 )1/2 = 0.9c).11. ii) d(z. z ) ≤ d(z. z ) = 2|z − z | [(1 + |z|2 )(1 + |z |2 )]1/2 2|z − x| 2|x − z | ≤ + 2 )(1 + |x|2 )]1/2 [(1 + |z| [(1 + |x|2 )(1 + |z |2 )]1/2 = d(z. z ) = d(z . min{|x|. z) for all z.8) is a metric space. We have 2|z − z | = 0 iﬀ z = z . 6. z ∈ C. show that every closed ball is a closed set. d(z. z ) ≥ 0. iv) d(z. 2 Solution. if z is imaginary then 2 B(z. Show that (C∞ . z ∈ C and d(z. ∞) = (1+|z|2 )1/2 . Since |z−z | ≥ 0. Exercise 6.

2) where the last inequality follows by ⇔ (1 + xz )(1 + x¯ ) ≤ (1 + |x|2 )(1 + |z |2 ) ¯ z and ⇔ (1 + xz)(1 + x¯) ≤ (1 + |x|2 )(1 + |z|2 ) ¯ z since (1 + xz)(1 + x¯) ≤ (1 + |x|2 )(1 + |z|2 ) ¯ z ⇔ xz + x¯ ≤ |x|2 + |y|2 ¯ z |z + xz| ≤ (1 + |x|2 )1/2 (1 + |z|2 )1/2 ¯ ⇔ 2Re(x¯) ≤ |x|2 + |y|2 z which is true by Exercise 4 part 2 on page 3. there exists an −ball BY (p : ) ⊂ G1 . show that G ∩ Y is open in (Y. let G be open in (X. Multiplying this by (1 + gives the assertion above. d). Let G1 be an open set in Y. ¯ Using (2. d) and Y ⊂ X.2) by (1 + |x|2 )1/2 yields |z − z |(1 + |x|2 )1/2 ≤ |z − x|(1 + |z |2 )1/2 + |x − z |(1 + |z|2 )1/2 . there exists an > 0 such that BX (p. 11 . Thus. d). ) ⊂ G. |x|2 )1/2 (1 2 + |z |2 )1/2 (1 + |z|2 )1/2 Exercise 8. Then p ∈ G and since G is open. pick an arbitrary point p ∈ G1 . In order to show that G1 is open in (Y. Then. Solution. But then BY (p. Thus G1 is open in Y (Proposition 1. ) = BX (p. d).1) and taking norms gives |(z − z )(1 + x x)| ¯ = |(z − z )(1 + |x|2 )| = |(z − z )|(1 + |x|2 ) = |(z − x)(1 + xz ) + (x − z )(1 + xz)| ¯ ¯ 2 1/2 2 1/2 ≤ |z − x|(1 + |x| ) (1 + |z | ) + |x − z |(1 + |x|2 )1/2 (1 + |z|2 )1/2 |z + xz | ≤ (1 + |x|2 )1/2 (1 + |z |2 )1/2 ¯ (2. ) ∩ Y ⊂ G ∩ Y = G1 which proves that p is an interior point of G1 in the metric d.which follows from a simple computation (z − x)(1 + xz ) + (x − z )(1 + xz) = z − x + xz z + x xz + x − z + x xz − xz z ¯ ¯ ¯ ¯ ¯ ¯ = z + x xz − z − x xz ¯ ¯ = (z − z )(1 + x x). for every p ∈ G1 . Suppose G ⊂ X is open. Set G1 = G ∩ Y. dividing (2.13a). Conversely. Let (X. d) be a metric space and Y ⊂ X. show that if G1 ⊂ Y is open in (Y. there is an open set G ⊂ X such that G1 = G ∩ Y.

12 . Let x ∈ A and y ∈ B and deﬁne the function f (t) : [0. what are its 1 components: (a) X = {z : |z| ≤ 1} ∪ {z : |z − 2| < 1}.3 are open. Show that the sets S and T in the proof of Theorem 2.13. Not available. 1] to X with f (0) = x and f (1) = y (this function f is called the path from x to y). 1] → X by 1 (1 − 3t)x + 3tRe(x). Show that {cis k : k a non-negative integer } is dense in T = {z ∈ C : |z| = 1}. ) ∩ Y = G ∩ Y where G = p∈G1 BX (p. 0 ≤ θ ≤ ∞}? Solution. Exercise 10. Exercise 2. (a) Show that a set A ⊂ R is an interval iﬀ for any two points a and b in A with a < b. (c) X = C − (A ∩ B) where A = [0. (b) Use part (a) to show that if a set A ⊂ R is connected then it is an interval. ) = p∈G1 BX (p. if X is not connected. ) ∩ Y = p∈G1 BX (p. 2 3 < t ≤ 1. Since we can write BY (p. the interval [a. a) Deﬁne X = {z : |z| ≤ 1} ∪ {z : |z − 2| < 1} := A ∪ B. Exercise 9. 2. Next. Recall that a space is path-connected if for any two points x and y there exists a continuous function f from the interval [0. Solution. 3 < t ≤ 3 (3 − 3t)Re(y) + (3t − 2)y. Solution. ) is open in (X. (b) X = [0.2 Connectedness Exercise 1. Prove Proposition 1. For which values of θ is {cis(kθ) : k a non-negative integer } dense in T? Solution. The purpose of this exercise is to show that a connected subset of R is an interval. 0≤t≤ 3 1 2 f (t) = (2 − 3t)Re(x) + (3t − 1)Re(y). Not available. It suﬃces to show that X is path-connected. Not available. Do Exercise 8 with “closed” in place of “open.Thus G1 = p∈G1 BY (p. Exercise 3. Not available. we get G1 = p∈G1 BY (p. (Proposition 1. d). ) is open). Not available. Obviously A is path-connected and B is path-connected. Which of the following subsets X of C are connected. ). Exercise 11. we will show that X is path-connected. b] ⊂ A. ∞) and B = {z = r cis θ : r = θ. Solution.9c since each BX (p. 1) ∩ 1 + n : n ≥ 1 . ) = BX (p.” Solution. ) ∩ Y.

) must 13 .6. 1). −2). Thus. { 3 }. Because x A. “⇐”: Assume S contains all its limit points. Show that if F ⊂ X is closed and connected then for every pair of points a. {1 + n }. But.13f) that for every > 0. zn in F with z0 = a. Solution. a) A set is closed iﬀ it contains all its limit points. By assumption x is not a limit point and hence there exists an open −ball around x. Hence. for each C ∈ C either C ⊂ A or C ⊂ B. Solution. Suppose x A. either D = A and B = ∅ or D = B and A = ∅. we assume A ∪ B is disjoint and thus E ∪ F is disjoint which contradicts the assumption E ∩ F ∅. If {D j : j ∈ J} is a collection of connected subsets of X and if for each j and k in J we have D j ∩ Dk then D = {D j : j ∈ J} is connected. 3 4 1 { 2 }. we could write D as the disjoint union A ∪ B where A and B are nonempty subsets of X. b) If A ⊂ X. that does not intersect S . . zk ) < for 1 ≤ k ≤ n. ¯ “⊆”: Let x ∈ A.4. ) ⊂ S c and therefore S c is open. Exercise 5. 2πk − 2π ≤ θ < 2πk}. Therefore X is not connected. then there exist E. D= j∈J Dj is connected. B(x. Let A ⊂ X be a set. In addition. Let D = j∈J D j and C = {D j : j ∈ J}. 2. Exercise 4. . Not available. since f (1/3) = limt→ 1 − f (t) = limt→ 1 + f (t) = Re(x) ∈ X and 3 3 f (2/3) = limt→ 2 − f (t) = limt→ 2 + f (t) = Re(y) ∈ X. . Is the hypothesis that F be closed needed? If F is a set which satisﬁes this property then F is not necessarily connected. (Proposition 1. ) ∩ A ∅. So B(x. Since ¯ x ∈ A. Let x ∈ S c . b in F and each > 0 there are points z0 . since there is no way to connect (2.13f). z1 . we have (Proposition 1. B(x. “⇒”: Let S be closed and x be a limit point. all members of C are contained in either A or all B. If not. {2}. .. We claim x ∈ S .}. then A− = A ∪ {x : x is a limit point of A} := A ∪ A .13f again) which contradicts the fact that x is a limit point of S . even if F is closed. . Solution. then obviously x ∈ A ∪ A . Thus. Therefore. 1) and (1. k ∈ {1. F ∈ C such that E ⊂ A and F ⊂ B. S c (open) would be an open neighborhood of x. c) X is not connected. 3. Both contradicting the fact that A. If not. The components are [0. 3 b) There is no way to connect {2} and { 2 }. Give an example to illustrate this. We have to show that S is closed or that S c is open. B(x.This function is obviously continuous. . B are assumed to be nonempty. . Prove Proposition 3. If x ∈ A. . Therefore 3 3 X is path-connected and hence X is connected. ) ∩ S = ∅ (negation of Proposition 1. We want to show x ∈ A ∪ A . ) such that B(x. Prove the following generalization of Lemma 2. The k − th component is given by C − {A ∩ B} where A = [2πk − 2π. If D is connected. zn = b and d(zk−1 . . we have f (0) = x and f (1) = y. . 2. We have C ⊂ A ∀C ∈ C or C ⊂ B ∀C ∈ C. Let S ⊂ X be a set.3 Sequences and completeness Exercise 1. 2πk) and B = {z = r cis θ : r = θ.

m ≥ N . z) = = 2 4 |zn |2 − zn z − zn z + |z|2 ¯ ¯ 4 |zn |2 − 2Re(zn z) + |z|2 ¯ 4 1− 2Re(zn z) ¯ |zn |2 |z|2 |zn |2 1+|z|2 |zn |2 1 + |zn |2 1 + |z|2 1 + |z|2 |zn |2 + 1 + |z|2 + 1 + |z|2 + if |zn | 0 = 4 0. Then d (zn . “⊇”: To show A ∪ A ⊆ A− . Exercise 4. xn ) < n which implies xn → x (see Proof of Proposition 3. xm ) ≤ d(xn . Let zn . ∞) → 0 implies 1+|zn | 1 + |zn |2 → ∞ which is equivalent to |zn | → ∞. Furnish the details of the proof of Proposition 3. That is. Thus d(xn . then d(zn . Let x ∈ A ∪ A . x) + d(x. Then there exists {xn } ⊂ A with limn→∞ xn = x. then d(zn . z) 0 for any z ∈ C. First assume that |zn − z| → 0. Now.2). d) is a Cauchy sequence. 14 2 + 2 = . so x ∈ A ∪ A . z be points in C and let d be the metric on C∞ . To see the converse. given n > N and d(x. Let {xn } be a convergent sequence with limit x. Show that every convergent sequence in (X. let d(zn . n ) ∩ A. z) → 1+|z|2 d(zn . > 0 ∃N such that d(xn . In particular. for every integer n there is a point xn in B(x.1 intersect A in a point diﬀerent from x. If x ∈ A. Then x ∈ A . z) → 0. ) ∩ A ∅ since {xn } ⊂ A. xm ) < 2 if m > N. Show that |zn − z| → 0 if and only if d(zn . 1 Thus d(x. Then clearly also 1 + |zn |2 → ∞ and therefore d(zn . z) → 0. d). Solution. ∞) = √ 2 2 → 0. also |zn | → ∞. Solution. Solution.8. Hence convergence in d-norm implies convergence in |·|-norm for numbers zn . Not available. It follows. Exercise 2. ∀ > 0 B(x. z ∈ C. d(zn . The last thing to show is that if zn → ∞ in (C∞ . In fact we will show that if |zn | → ∞.13f we get x ∈ A− . xm ) < and therefore {xn } is a Cauchy sequence. z) → 0 or equivalently d2 (zn . We need to show that if zn → z in the d-norm. Not available. This shows that the denominator remains bounded as in particular if |zn | → ∞. then |zn | ∞ because otherwise the denominator grows without bounds. By Proposition 1. then x ∈ A− (A ⊂ A− ). Also show that if |zn | → ∞ then {zn } is Cauchy in C∞ . (Must {zn } converge in C∞ ?) Solution. z) → 0 and therefore the numerator 2|zn − z| → 0. z) = 2 (1 + |zn |2 )(1 + |z|2 ) |zn − z| → 0 since the denominator (1 + |zn |2 )(1 + |z|2 ) ≥ 1 is bounded below away from 0. x) < if Exercise 5. 2 ∀n. Exercise 3. Next assume that |zn | → ∞. assume x ∈ A but not in A. But d(zn . Show that diam A = diam A− .

[ f (x) − g(x)]2 dx. Solution. the rationals. . ﬁx nk0 > M + N. ∀n ≥ N. but that {xn } is a Cauchy sequence in (R. the metric space (X. Not available. there is a x such that xnk → x as k → ∞. q) = n k=1 Solution. qn ] and show that diam R = d(p. . q1 ] × .Exercise 6. 1 Example 2: Let X = (0. R is compact. xnk0 ) + d(xnk0 . Solution. . x) < Thus.y∈R 15 . Consider Solution. Suppose {xn } is a Cauchy sequence and {xnk } is a subsequence that is convergent. Therefore. f. Give three examples of non complete metric spaces. √ ∈ X. 2 diam(R) = sup d(x. . By deﬁnition 1 2 (qk − pk ) . y) = |x − y|. x) ≤ d(xn . xn → x as n → ∞ and therefore {xn } is convergent. the metric space is not complete. x) < 2 ∀nk ≥ M since xnk → x as k → ∞.4. (Hint: Take inspiration from C∞ . Exercise 8. Thus. The sequence deﬁned by x1 = 1. The sequence { n } is Cauchy. . 2. y xn+1 = x2n + x1n is a Cauchy sequence of rational numbers. the metric space is not complete. We have to show that xn → x as n → x. Let R = [p1 . x) → 0. 1 n < x ≤1 1 It is obvious that the limit function f is discontinuous. Example 3: Let X = Q. y ∈ X. x∈R. so we have diam(R) = max d(x. m ≥ N since {xn } is Cauchy and ∃M ∈ N such that d(xnk . g) = −1 the Cauchy sequence 0. pn ) and q = (q1 . Hence. xm ) < 2 ∀n. y). d) is not complete.y∈R Obviously. Now. Let p = (p1 . y). Then we have ∃N ∈ N such that d(xn .) Solution. . which is not in the space. y) = |x − y|. The limit 2 is not a rational number. Exercise 2. d) when |xn | → ∞. then d(xn . Exercise 7. Put a metric d on R such that |xn − x| → 0 if and only if d(xn . Show that {xn } must be convergent. x. . 1] and the metric d( f.4 Compactness Exercise 1. qn ) be points in Rn with pk < qk for each k. with metric d(x. Not available. x∈R. −1 ≤ x ≤ 0 1 fn (x) = nx. . x. 2 + 2 = . × [pn . Finish the proof of Proposition 4. 0 < x ≤ n . but converges to 0. Let > 0. . 1] with metric d(x. g ∈ X. Since {xnk } is convergent. 1. . Example 1: Let X = C[−1.

B(x. . Let F = [a1 . . Let X be the set of all bounded sequences of complex numbers. use Exercise 2 to show that there are rectangles R1 . . . Show that for each x in X and ¯ > 0. n. Not available. . × [an . . . j i=1 j=1 which is a ﬁnite union and therefore K is compact. Thus the maximum distance is obtained. Exercise 5. . .3) we obtain n n ∀i = 1. . y) = x∈R. {xn } ∈ X iﬀ sup{|xn | : n ≥ 1} < ∞. . . ). by (2. . Solution. Show that the closure of a totally bounded set is totally bounded. . . pi ≤ xi ≤ qi and pi ≤ yi ≤ qi for all i = 1. Let K = n i=1 Ki be a ﬁnite union of compact sets. Then clearly. yn ) ∈ R. that is Ki ⊂ Since each Ki is compact.3) d(x. . . Thus n Gi. q). . n. deﬁne d(x. . . Exercise 4. That is. n. Let {Gλ }λ∈Γ be an open cover of K. We also have (yi − xi )2 ≤ (qi − pi )2 . . Therefore. Exercise 3. . Solution. n. {Gλ }λ∈Γ is an open cover of each Ki . . . Ki ⊂ for each i = 1. bn ] ⊂ R and let > 0. Rm such that F = m Rk and diam Rk < for each k. that is K⊂ Gλ . so n diam(R) = sup d(x. Show that the union of a ﬁnite number of compact sets is compact. . . ) is not totally bounded although it is complete. n. .Let x = (x1 . xn ) ∈ R and y = (y1 . . . . 0. . . y) = i=1 (yi − xi )2 ≤ i=1 (qi − pi )2 = d(p. . j j=1 n ki K= j=1 Ki ⊂ Gi. . . . Not available. y) = sup{|xn − yn | : n ≥ 1}. . . n. Exercise 6. λ∈Γ ∀i = 1. . Note that we get equality if for example xi = pi and yi = qi for all i = 1. If xk ∈ Rk then it follows that k=1 Rk ⊂ B(xk . (2. (Hint: you might have an easier time of it if you ﬁrst show that you can assume x = (0.y∈R n i=1 (qi − pi )2 = d(p.). q). . 16 .) Solution. ki Gλ . λ∈Γ Of course. i = 1. If x = {xn } and y = {yn }. b1 ] × . .

k=1 n ¯ S ⊆ B(xk . y ∈ X. f (y)) = | f (x) − f (y)| < 1 whenever d(x. y) = (M + N)d(x. 2 + 2. . 17 . We have ρ( f (x) + g(x). Not available. d) be a given metric space and let S ⊂ X be totally bounded. let f.9 d) p. f (y) + g(y)) = | f (x) + g(x) − f (y) − g(y)| = | f (x) − f (y) + g(x) − g(y)| ≤ | f (x) − f (y)| + |g(x) − g(y)| ≤ 1 2 1 > 0 ∃δ1 > 0 such that ρ( f (x). f (y) + g(y)) ≤ Kd(x. ∀x. Solution. Since S is totally bounded. g(y)) = |g(x) − g(y)| ≤ Nd(x. Prove Proposition 5. The distance in C is ρ(x. y) < δ2 . y ∈ X. gives the desired result B(xk . y ∈ X. . that is. Solution. that is. /2). y) = |x − y|. . y ∈ X Since there is a constant K = M + N > 0 such that ρ( f (x) + g(x). ∀x. Now. Let (X. y). Exercise 2.5 Continuity Exercise 1. Let f : X → C be Lipschitz. we have that f + g is Lipschitz. Show that if f and g are uniformly continuous (Lipschitz) functions from X into C then so is f + g. that is. f (y)) = | f (x) − f (y)| ≤ Md(x. k=1 2.2. xn ∈ S such that n S ⊆ Taking the closure. 22 that there exist a ﬁnite number of points x1 . y) + Nd(x. we have by Theorem 4.Solution. y) < δ1 whenever d(x. ∀x. there is a constant N > 0 such that ρ(g(x). y). there is a constant M > 0 such that ρ( f (x). g be both uniformly continuous. f (y) + g(y)) = | f (x) + g(x) − f (y) − g(y)| = | f (x) − f (y) + g(x) − g(y)| ≤ | f (x) − f (y)| + |g(x) − g(y)| ≤ Md(x. The distance in X is d(x. y). . ∀x. Let > 0. and g : X → C be Lipschitz. y). f (y)) = | f (x) − f (y)| < > 0 ∃δ2 > 0 such that ρ( f (x). ∀ and ∀ We have ρ( f (x) + g(x). y). ).

y ∈ X. y ∈ X. δ2 ). f (y)) = | f (x) − f (y)| < > 0 ∃δ2 > 0 such that ρ( f (x). y) < min(δ1 . y) < δ2 . δ2 ). y) ∀x. Show that if f and g are bounded uniformly continuous (Lipschitz) functions from X into C then so is f g. there exists a constant K = M1 N2 + M2 N1 > 0 with ρ( f (x)g(x). f (y)) = | f (x) − f (y)| < ∀x ∈ X ∀x ∈ X 1 2 whenever d(x. choosing = δ = min(δ1 . Solution. y) < δ1 and d(x. there exists a constant M = M1 M2 with | f (x)g(x)| ≤ M ∀x ∈ X. So. y ∈ X and let g be Lipschitz. y) + M2 N1 d(x. So. 1 + 2 and Exercise 3. y ∈ X.whenever d(x. f (y)g(y)) ≤ Kd(x. that is. y) = (M1 N2 + M2 N1 )d(x. there exists a constant M1 > 0 with | f (x)| < M1 for all x ∈ X and let g be bounded. ρ( f (x)g(x). Obviously f g is bounded. that is. We say that f : X → C is bounded if there is a constant M > 0 with | f (x)| ≤ M for all x in X. f (y)g(y)) = | f (x)g(x) − f (y)g(y)| = = | f (x)g(x) − f (x)g(y) + f (x)g(y) − f (y)g(y)| ≤ ≤ ∀x. Thus. because | f (x)g(x)| ≤ | f (x)| |g(x)| ≤ M1 · M2 ∀x ∈ X. 18 . that is. y) ∀x. we have shown that ∀ > 0 ∃δ > 0 such that ρ( f (x). there exists a constant M2 > 0 with |g(x)| < M2 for all x ∈ X. y) so f g is Lipschitz and bounded. let f and g be both bounded and uniformly continuous. there exists a constant N1 > 0 such that ρ( f (x). that is. whenever d(x. that is ∃ ∃ M1 such that | f (x)| ≤ M1 M2 such that |g(x)| ≤ M2 1 2 ∀x. Let f be bounded. f (y)) = | f (x) − f (y)| < whenever d(x. y) < δ. ∀ ∀ > 0 ∃δ1 > 0 such that ρ( f (x). Thus f + g is uniformly continuous. y) Now. Now. y) < δ2 . Let f be Lipschitz. f (y)) ≤ N1 d(x. g(y)) ≤ N1 d(x. there exists a constant N2 > 0 such that ρ(g(x). ≤ | f (x)g(x) − f (x)g(y)| + | f (x)g(y) − f (y)g(y)| | f (x)| |g(x) − g(y)| + | f (x)| |g(y) − g(y)| M1 N2 d(x. y) < δ1 whenever d(x. that is.

δ2 ). Use Theorems 5. Thus. Exercise 5. xm ) < δ which tells us that { f (xn )} is a Cauchy sequence in Ω.17 to show that there is a polygon Q ⊂ G from a to b which is composed of line segments which are parallel to either the real or imaginary axes. Therefore f (x) = uniformly continuous. we have. Is the composition of two uniformly continuous (Lipschitz) functions again uniformly continuous (Lipschitz)? Solution. d). for every > 0 there exists δ > 0 such that ρ( f (x). 1). f (xm )) < ∀n. that is. 1). Let G be an open subset of C and let P be a polygon in G from a to b.) Solution. where D is dense in (X. But n 1 { f (xn )} = { f ( n )} = {n} is obviously not Cauchy. We have ρ( f (x)g(x). Not available. show that if {xn } is a Cauchy sequence in X then { f (xn )} is a Cauchy sequence in Ω. ρ) is uniformly continuous. But then. m ≥ N. Note that f (x) = 1 is not uniformly continuous on (0. whenever d(x. Recall the deﬁnition of a dense set (1. WLOG assume δ < 1 since the interval (0. 1 x cannot be Exercise 6. by the uniform continuity. Assume f : X → Ω is uniformly continuous. If {xn } is a Cauchy sequence in X. d) → (Ω. δ2 ). m ≥ N whenever d(xn . Then there is no δ > 0 such that | f (x) − f (y)| < 1 whenever |x − y| < δ. Solution. Is this still true if we only assume that f is continuous? (Prove or give a counterexample. x y xy δ/2· δ δ that is no matter what δ < 1 we choose. Let y = x + δ/2 and set x = δ/2. Exercise 7. Not available.14). f (y)) < if d(x. 1) is considered. y) < δ. the statement is not true. we always obtain | f (x) − f (y)| > 1. Here is a counterexample: Let f (x) = 1 which is continuous x on (0.15 and 5. Suppose f : X → Ω is uniformly continuous. y) < δ. for every 1 > 0 there exists N ∈ N such that d(xn . since we have already shown that f g is bounded. y) < min(δ1 . Assume there exists such a δ.It remains to verify that f g is uniformly continuous. Exercise 4. x To see that pick = 1. we have that ρ( f (xn ). So choosing = M1 2 + M2 1 and δ = min(δ1 . If f is continuous. 19 . Use Exercise 5 to show that there is a uniformly continuous function g : X → Ω with g(x) = f (x) for every x in D. Suppose that Ω is a complete metric space and that f : (D. then δ/2 1 1 1 y−x | f (x) − f (y)| = − = = = > 1. f (y)g(y)) = | f (x)g(x) − f (y)g(y)| = = | f (x)g(x) − f (x)g(y) + f (x)g(y) − f (y)g(y)| ≤ | f (x)g(x) − f (x)g(y)| + | f (x)g(y) − f (y)g(y)| ≤ | f (x)| |g(x) − g(y)| + | f (x)| |g(y) − g(y)| ≤ M1 2 + M2 1 . we have ∀ > 0 ∃δ > 0 such that | f (x)g(x) − f (y)g(y)| < whenever d(x. f g is uniformly continuous and bounded. The sequence xn = 1 is apparently convergent and therefore a Cauchy sequence in X. xm ) < 1 for all n.

b in X. Use Lebesgue’s Covering Lemma (4. δ) ⊂ B(z. f (z)) + ρ( f (z). More precisely. Suppose f : X → Ω is continuous and X is compact.8) to give another proof of Theorem 5. Solution. Let { fn } be a sequence of uniformly continuous functions from (X. d) into (Ω. . If sup Mn = ∞. x. f (z)) ≤ ρ( f (x).9 p. 22). Solution.5. p) and let D be a dense subset of X. Suppose (X. z1 . and d(zk−l . zn = b. zk ) < for 1 ≤ k ≤ n. d) is a compact metric space having the property that for every > 0 and for any points a. 21 to obtain a δ > 0 such that x ∈ X implies that B(x. Exercise 9. show that f is a Lipschitz function. f (y)) < and hence f is uniformly continuous on X. 20 . Since f is continuous. ρ) and suppose that f = u − − lim fn exists. zn in X with z0 = a. Since X is by assumption compact (it is also sequentially compact as stated in Theorem 4. X= B(x. In addition. show that f may fail to be Lipschitz.Solution. f (y)) < /2 whenever d(x.8 p. d) to (Ω. If each fn is a Lipschitz function with constant Mn and sup Mn < ∞. . 2 + 2 = 2.15. Exercise 8. Not available. Then (X. Not available. (Hint: Use Theorem 5.6 Uniform convergence Exercise 1. Prove that f is uniformly continuous. we can use Lebesgue’s Covering Lemma 4. . Let > 0. Prove the following converse to Exercise 2. d) is connected. Not available. we have for all x ∈ X there is a δ x > 0 such that ρ( f (x). δz ) and therefore ρ( f (x). Let f and g be continuous functions from (X. Prove that if f (x) = g(x) for x in D then f = g. δz ) for some z ∈ X. Exercise 10. To show f is uniformly continuous. y) < δ x . δ x ) x∈X is an open cover of X.17. there are points z0 . Solution. y ∈ B(z. Not available.) Solution. . Use this to show that the function g obtained in Exercise 6 is unique.

5. n→∞ Since this holds for any r < lim inf n→∞ an and s < lim inf n→∞ bn . Solution.6. Prove Proposition 1. implies that r + s ≥ lim sup(an + bn ). Solution. implies that r + s ≤ lim inf (an + bn ). Exercise 2. 21 .1 Power series Exercise 1. This however. Show that lim inf an ≤ lim sup an for any sequence in R. Exercise 3. Prove that lim sup(an +bn ) ≤ lim sup an +lim sup bn and lim inf(an +bn ) ≥ lim inf an +lim inf bn for bounded sequences of real numbers {an } and {bn }. Not available. there are only ﬁnitely many by deﬁnition). Give the details of the proof of Proposition 1.Chapter 3 Elementary Properties and Examples of Analytic Functions 3. n→∞ n→∞ n→∞ Exercise 4. Then r + s < an + bn for all but ﬁnitely many n’s. we have lim inf (an + bn ) ≥ lim inf an + lim inf bn . there are only ﬁnitely many by deﬁnition). Let r > lim supn→∞ an (we know there are only ﬁnitely many by deﬁnition) and let s > lim supn→∞ (same here. n→∞ Since this holds for any r > lim supn→∞ an and s > lim supn→∞ bn . Solution. Then r + s > an + bn for all but ﬁnitely many n’s. Not available. This however. n→∞ n→∞ n→∞ Let r < lim inf n→∞ an (we know there are only ﬁnitely many by deﬁnition) and let s < lim inf n→∞ (same here. we have lim sup(an + bn ) ≤ lim sup an + lim sup bn .

R = |a| ∞. (d) ∞ zn! . But the inﬁmum of all these s’s is M. R = lim 2 1. n→∞ n→∞ Exercise 5. This holds for all s > M. If {an } is a convergent sequence in R and a = lim an . This shows that a − ≤ lim inf an ≤ a + n→∞ =:m and a − ≤ lim sup an ≤ a + . Suppose that {an } is a convergent sequence in R with limit a = limn→∞ an . n→∞ n→∞ ∞ n=0 Exercise 6. Then by deﬁnition. . n=0 n=0 n=0 Solution. a ∈ C. |a| = 1 . Let m = lim inf n→∞ an and bn = inf{an . |a| < 1 22 2 . We also have. Therefore m ≤ M which is lim inf an ≤ lim sup an .Solution. |a| > 1 1 = lim 2n+1 = 1. . Let M = lim supn→∞ an . Solution. a ∈ C. Find the radius of convergence for each of the following power series: (a) 2 (b) ∞ an zn . we obtain that an < s for inﬁnitely many n’s which implies that bn < s for all n and hence lim supn→∞ bn = m < s. so b) In this case. a=0 an 1 an bn = lim (n+1)2 = lim n2 +2n+1 = lim 2n+1 n→∞ a n→∞ a n→∞ a n→∞ bn+1 0. k→∞ k→∞ lim sup |bk |1/k = lim sup |ak |1/k = lim sup |a| = |a|. Since > 0 is arbitrary. Then. k an integer 0. we have: ∀ > 0 ∃N > 0 such that ∀n ≥ N. bn = an where a ∈ C. by deﬁnition of the lim supn→∞ an = M. This means that all but ﬁnitely many an ’s are ≤ a + and ≥ a − . a ∈ C. show that a = lim inf an = lim sup an . that is a − ≤ an ≤ a + . n→∞ we obtain lim inf an = lim sup an = a. we obtain m = M and further n→∞ 0≤ M−m≤2 . k→∞ Therefore. n→∞ |a| ∞. an zn = ∞ k=0 bk zn with bk = ak . a) We have ∞ n=0 an zn . n→∞ =:M By the previous Exercise 4. (c) ∞ kn zn . Hence. . an+1 . Take any s > M.}. a − ≤ lim inf an ≤ lim sup an ≤ a + . R = 1/|a|. we also have a− ≤m≤ M ≤a+ . 2 a 0 . we have |an − a| ≤ .

bn = kn . n n n . k→∞ k=0 k=1 k = n!. k = n(n + 1) . 1 The ﬁrst inequality is immediate from the fact that n ≥ 1 and hence n n(n+1) ≥ 1. n > 1 otherwise lim sup |ak |1/k = lim sup |1|1/k! = 1. ak = 1. = k 1 R= |k| − k .) Solution. n ∈ N. k integer d) We can write ∞ n! n=0 z = ∞ k=0 Thus. k < 0. Exercise 7. 0. k > 0. ak zk where 0.c) Now. k→∞ Therefore 1/R = 1 which implies R = 1. Show that the radius of convergence of the power series ∞ n=1 (−1)n n(n+1) z n is 1.t. Rewrite the power series in standard form. k is an integer 0. | (−1)n |= n and n(n+1) 1 √ ≤1 n n(n+1) | (−1)n |= n 23 n(n+1) 1 1 √ ≥ √ . then n ∞ ∞ (−1) (−1)n n(n+1) k z = ak z with ak = n 0 n n=1 n=1 if ∃ n ∈ N s. −1. n→∞ So 1 1. else To ﬁnd the radius of convergence we use the root criterion and therefore need the estimates √ √ 1 ≤ n(n+1) n ≤ n n for n ∈ N. We have n→∞ n→∞ R = lim sup |bn |1/n = lim sup |kn |1/n = lim sup |k| = |k|. For the second inequality note that ⇔ ⇔ Using this one obtains n(n+1) n ≤ nn+1 1 1 n n(n+1) ≤ n n √ √ n(n+1) n ≤ nn . k integer . (Hint: The nth coeﬃcient of this series is not (−1)n /n. and discuss convergence for z = 1. 2. and i.

1} (−1)n n(n+1) .0) in the complex plane C in any manner. 2.e. 1. 2. 3} n n=1 n=1 Deﬁne the partial sums S k := k n=0 ck . l. Now let z = i.) Take the path along the real axes. hence d). 3. We claim that the following chain of inequalities holds a) b) c) d) 0 ≤ S 4k+3 < S 4k < S 4k+4 < S 4k+2 < S 4k+1 ≤ 1. i. hence a) (4k + 1)(4k + 2)(4k + 3) 1 1 S 4k+4 − S 4k = c4k+3 + c4k+4 = − + < 0. The expression in(n+1) will always be real. Therefore the power series converges also in the case of z = i. Now that |cn | 0 the diﬀerence between S 4k+l and S 4k+m . We have f (z + h) − f (z) h = = ¯ ¯ |z + h|2 − |z|2 (z + h)(¯ + h) − z¯ z¯ + h¯ + zh + hh − z¯ z ¯ z z z z = = h h h ¯ h z + h + z =: D. i. note that S 4k+3 − S 4k = c4k+1 + c4k+2 + c4k+3 = − 16k2 + 8k − 1 < 0. m ∈ 0.Vague memories of calculus classes tell me that n √ n n → 1. If z = 1 the series reduces to ∞ (−1) which converges with the Leibniz Criterion. Relation b) is obvious and so are the upper bound c1 = 1 and the non-negativity constraint. hence c) 4k + 3 4k + 4 S 4k+2 − S 4k+1 = c4k+2 < 0. Solution. 3 tends to zero. that is y = 0. h h∈C h→0 If the limit of D exists.e. y) approach the origin (0. The derivative of f at z is given by f (z) = lim provided the limit exist. 1. all subsequences converge to the same limit. l ∈ {0. Show that f (z) = |z|2 = x2 + y2 has a derivative only at the origin. so if the series converges at z = i. We also note that formally ∞ ∞ 1 ifn mod 4 ∈ {0.2 Analytic functions Exercise 1. 3} describe bounded and monotone subsequences that converge to some point. Then h = h and thus D=z+h+z ¯ h =z+h+z ¯ h 24 . n=1 n If z = −1 we note that the exponents n(n + 1) are always even integers and therefore the series is the same as in the previous case of z = 1. it converges to a real number. R = 1. To verify this. We remark that {S 4k+l }k≥1 . ¯ ¯ h f (z + h) − f (z) . ¯ 1. i = cn with n 1 − n if n mod 4 ∈ {2. it may be found by letting the point h = (x. thus 1 R = lim sup n(n+1) √ an = 1.

. It remains to show that the limit of D exists at z = 0. 0. since for all n ∈ N. Hence. In summary. lim sup an bn ≥ n→∞ k→∞ n→∞ n→∞ n→∞ k→∞ k→∞ lim sup an bn . Docs this remain true if the requirement of positivity is dropped? Solution. n→∞ n→∞ n→∞ Hence. n→∞ In both cases. ¯ ¯ ¯ if the limit of D exists. Since lim bnk = b. ¯ Because of the uniqueness of the limit of D. −4 . Show that lim n1/n = 1. −3.) Take the path along the imaginary axes. − 1 . n→∞ n→∞ ab = 0 1 = lim sup an bn . 0. its value has to be z + z. Hence lim sup an bn = ∞. Applying the inequality we have established replacing bn with n→∞ bn b with an bn : lim sup an = lim sup n→∞ n→∞ and an replaced 1 1 (an bn ) ≥ lim sup an bn . and 2 3 4 note 0 = b = lim bn < ∞. . n→∞ 1 bn consider lim 1 1 = . lim ank bnk = ∞. its value has to be z − z. − 1 . {ank bnk } is a subsequence of an bn that converges to ab. lim ank bnk = ab. Now. since lim bn > 0. ¯ ¯ 2. n→∞ n→∞ It follows that ab = lim sup an bn as required. we must have z + z = z − z ⇐⇒ z = −z ⇐⇒ z = 0. . . lim sup an bn ≥ b lim sup an . Also let an = 0. 25 . let a = lim sup an = ∞. Since z = 0. and note lim sup an = a = 0.and therefore. So Now. Exercise 2. if the limit of D exists. Then there exists a monotonic subsequence {ank } of {an } that converges ab ≤ to a. the function f (z) = |z|2 = x2 + y2 has a derivative only at the origin with value 0. bn b n→∞ Rearranging. Prove that if bn . we have that D = h and thus the limit of D is 0. In this second case. And. Then h = −h and thus D=z−h−z ¯ h =z−h−z ¯ h and therefore. . 0. Then there exists a subsequence {ank } of {an } such that lim ank = a > 0. In this case. that is x = 0. an are real and positive and 0 < b = lim bn < ∞. an > 0 and bn > 0. n→∞ n→∞ b lim sup an . Let a = lim sup an < ∞. −2. a = lim sup an then ab = lim sup(an bn ). n→∞ Exercise 3. lim sup an bn ≤ b lim sup an . we have established that ab ≤ lim sup an bn . if the limit of D exists. n→∞ Now. Let bn = 0. consider the case where we drop the positivity requirement. − 1 . 0.

Derive formulas (2. Thus. 2i 1 + 2k πi . Solution. {z : ez = −i}. n→∞ Now. Solution.14). Solution. {z : cos z = 0} = 1 +k π . Then lim f (x) = 0 by L’Hopital’s Rule. {z : cos z = 0}. Also let a = n1/n . {z : ez = −1}. lim = lim f (n) = 0 also. We have cos(z) = sin(z) = 26 eiz + e−iz 2 eiz − e−iz . Solution. 2 {z : ez = −1} = {(1 + 2k) πi} . We have by deﬁnition (cos z) = and similarly (sin z) = 1 iz e − e−iz 2i = i iz 1 iz e + e−iz = e + e−iz = cos z. Then a = n1/n ⇐⇒ log a = log n1/n log n ⇐⇒ log a = n log n ⇐⇒ lim log a = lim n→∞ n→∞ n log(x) log n . So n→∞ n→∞ n n→∞ x 1/n lim a = lim n = 1. let f (x) = n→∞ Exercise 4. Show that (cos z) = − sin z and (sin z) = cos z. Using the deﬁnition we obtain {z : ez = i} = {z : ez = −i} = and {z : sin z = 0} = {kπ} where k ∈ Z. Exercise 7. Let n ∈ N. 2 1 − + 2k πi . Exercise 6. 2i 2 1 iz e + e−iz 2 = i iz 1 iz e − e−iz = − e − e−iz = − sin z 2 2i Exercise 5.Solution. Not available. Prove formulas for cos(z + w) and sin(z + w). {z : sin z = 0}. Describe the following sets: {z : ez = i}. 2 .

Suppose that f : G → C is a branch of the logarithm and that n is an integer. g (ω) 0 for any ω. Not available. Solution. Proof: sin(z) cos(w) + cos(z) sin(w) = = = = = Exercise 8. 2. Let G and Ω be open in C and suppose f and h are functions deﬁned on G. Prove that if zn → z then θn → θ and rn → r. Solution. −π < θ < π and Log(z) = ln(r) + iθ. Exercise 12. Claim: cos(z) cos(w) − sin(z) sin(w) = cos(z + w). Prove that zn = exp(n f (z)) for all z in G. θn < π. Exercise 10. We know that we can write z = reiθ 0. Suppose that g and h are analytic. Exercise 9. Prove the following generalization of Proposition 2. Claim: sin(z) cos(w) + cos(z) sin(w) = sin(z + w). Not available.1. where is this function deﬁned and analytic? 1 1 1 1 1 1 1 θ θ z 2 = eLogz 2 = e 2 Logz = e 2 (lnr+iθ) = e 2 lnr e 2 θi = e 2 lnr cos + i sin 2 2 27 . Exercise 11. Proof: cos(z) cos(w) − sin(z) sin(w) = = = = = eiz + e−iz eiw + e−iw eiz − e−iz eiw − e−iw − 2 2 2i 2i 1 iz iw 1 iz iw e e + e−iz e−iw + e e + e−iz e−iw 4 4 1 iz iw 1 −iz −iw e e + e e 2 2 1 iz iw e e + e−iz e−iw 2 cos(z + w). Deﬁne tan z = Solution. z ∈ G = C − {z : z ≤ 0} and zn = rn eiθn . Thus . Show that f is analytic. that f is continuous. Suppose that zn . h is one-one. sin z cos z . Show that the real part of the function z 2 is always positive. Solution. Solution. z = reiθ where −π < θ. Give a formula for f (z).20. Not available. g : Ω → C and suppose that f (G) ⊂ Ω. 1 eiz − e−iz eiw + e−iw eiz + e−iz eiw − e−iw + 2i 2 2 2i 1 iz iw 1 iz iw e e − e−iz e−iw + e e − e−iz e−iw 4i 4i 1 iz iw 1 −iz −iw e e − e e 2i 2 1 iz iw e e − e−iz e−iw 2i sin(z + w). and that they satisfy h(z) = g( f (z)) for z in G. Not available.

then log(z) = Log(z) + 2kπi for some k ∈ Z. Find all analytic functions f : G → C such that z = ( f (z))n for all z ∈ G. Re(z) = e 2 lnr cos 1 θ > 0. . Therefore. Since G is connected and f : G → C is diﬀerentiable with f (z) = 0 ∀z ∈ G. 28 . we have that u and v have continuous partial derivatives. ∂u ∂v = ∂x ∂y Since f (z) is real ∀z ∈ G. Hence f (z) = 0.1) we obtain ∂u ∂u = =0 ∂x ∂y and thus u (z) = 0 (see reasoning of equation 2. They correspond to the n distinct powers of the expression ζ = e2πi/n . Exercise 14. we have and by (3. Thus. v are real-valued functions.Hence. First of all. where k = 0. Let Log(z) be the principal branch. Suppose f : G → C is analytic and that G is connected. Show that if f (z) is real for all z in G then f is constant. that is f is continuously diﬀerentiable (Deﬁnition 2.22 and 2. 2 1 since e 2 lnr > 0 and cos 1 θ 2 > 0 since −π < θ < π. . Let G = C − {z ∈ R : z ≤ 0} and let n be a positive integer. Solution. So. We know that the latter factor are the n-th roots of unity and depend only on k and n. v satisfy the Cauchy-Riemann equations. we can write f : G → C as f (z) = u(z) + iv(z) where u. the real part of the function z 2 is always positive. Exercise 15. .3). ∂y ∂x (3.29. we can write f (z) = z1/n = elog(z)/n = e(Log(z)+2kπi)/n = eLog(z)/n · e2kπi/n . this implies and ∂u ∂v =− . this implies that u. n − 1 and therefore they are all constant multiples of each other.1) v(z) ≡ 0 ∂v ∂v = =0 ∂x ∂y and therefore f (z) = u(z). That is. Exercise 13. determine the set A. . we have that f is constant. By Theorem 2. For r > 0 1et A = ω : ω = exp 1 z where 0 < |z| < r . Thus. Since f : G → C is analytic. the branches of z1/n on the set U are given by f (z) = ζ k · eLog(z)/n . Solution.23 on page 41). since v(z) = 0.

Let z1 . If the restrictions on the zk are removed. Using polar coordinates we can write w = |w|eiθ . Find an open connected set G ⊂ C and two continuous functions f and g deﬁned on G such that f (z)2 = g(z)2 = 1 − z2 for all z in G. Can you make G maximal? Are f and g analytic? Solution. Consider ﬁrst the case where n = 2. . y) = u(x. But note arg(z1 z2 ) = arg z1 + arg z2 implies that −π < arg z1 + arg z2 < π. zn be complex numbers such that Re zk > 0 and Re(zl . . . r Exercise 16. the equation ez = w has a solution z ∈ T . . . We have to ﬁnd a complex number z = x + iy such that x2 + y2 > 1 and r e x eiy = w = |w|eiθ . . y) = −v(x. Not available. x y v∗ (x. Claim: The image of A is C − {0} (thus not depending on r). we need to show that for w 0. . Suppose that f (G) ⊂ G and g(G) ⊂ G and prove that both f ◦ g and g ◦ f are branches of zab . Let f : G → C and g : G → C be branches of za and zb respectively. zk ) > 0 for 1 ≤ k ≤ n. z2 . Exercise 17. Using x = log |w| and k 0 such that (log |w|)2 + (θ + 2kπ)2 > 1 . . . Not available. Deﬁne the set S = {z : 0 < |z| < r} where r > 0. −y). zn ∈ C such that Re(z j ) > 0 and Re(z1 · · · z j ) > 0 for 1 ≤ j ≤ n. Let z = x + iy and let f (z) = u(x. y). Hence. Let z1 . −y). .Solution. y) where u∗ (x. . Then − π < arg z1 < π . y) = −v x (x. In addition the Cauchy-Riemann Equations u x = vy and uy = −v x are satisﬁed. then we found z = x + iy. is also analytic. y) = −uy (x. 2 2 2 Now. zk ) = log((z1 · · · zk−1 )zk ) = log(z1 · · · zk−1 ) + log zk by the case when n = 2 = log z1 + log z2 + . −y) ¯ and v∗ (x. Give the principal branch of Solution. y) = u x (x. . recall log(z1 z2 ) = ln |z1 z2 | + i arg(z1 z2 ) = ln |z1 | + ln |z2 | + i arg(z1 ) + i arg(z2 ) = log(z1 ) + log(z2 ) Assume this formula is true for n = k − 1. −y) = vy (x. r To ﬁnd the image of A is the same as ﬁnding the image of T under ez . By assumption f is analytic and therefore u and v have continuous partial derivatives. y) = vy (x. . Let z1 . . Let z1 . −y) and v∗ (x. x y x y y x Exercise 20. . y) + iv∗ (x. Exercise 18. where log z is the principal branch of the logarithm. u∗ (x. deﬁned by f ∗ (z) = f ( x). Not available. we have f ∗ (z) = u∗ (x. We want e x = |w| and y = θ + 2kπ. does the formula remain valid? Solution. −y). Then log(z1 . we have u∗ (x. Exercise 19. . for some k ∈ Z. zk ∈ C as above. −y) and therefore u∗ = v∗ and u∗ = −v∗ so f ∗ is analytic. + log zk−1 + log zk by the case when n = k − 1 29 √ 1 − z. The image of this set under 1/z is clearly T = z: 1 < |z| . Since f ∗ (z) = f ( x). . If f : G → C is analytic prove that ¯ f ∗ : G∗ → C. Show that log(z1 . . To proof the claim. . − π < 2 2 2 arg z2 < π and − π < arg(z1 z2 ) < π . ¯ Solution. Show that f g is a branch of za+b and f /g is a branch of za−b . Let G be a region and deﬁne G∗ = {z : z ∈ G}. . Solution. . . z2 ∈ C as above. The proof will be by induction. y) + iv(x. zn ) = log z1 + . + log zn . −y) = v x (x. .

this is true for all n such that zi . If the restriction on the zk are removed. The image of {z = x + iy : x < 0. Solution. −π < y < π} = {e x eiy : x < 0. 2 the formula is invalid. −π < y < π} = {r(cos(y) + i sin(y)) : 0 < r < 1. Exercise 21. Therefore its only diﬀerence to the principal branch is 2πik for some k ∈ Z. Since r varies between 0 and 1. we get that the image is a solid circle with radius 1. −π < y < π} under the exponential function is given by {e x+iy : x < 0. Then restricting f to G gives us a branch of the log ˆ on G. 30 . π) and log(z1 z2 ) = Thus log z1 + log z2 = ln 5 + ln 5 + i( 3π ) = ln 5 + i( 3π ). Find the image of {z : Re z < 0. −π < y < π}. ˆ Assume f (z) is a branch of the logarithm deﬁned on G. Do exercise 1 for the set {z : |Im z| < π/2}. z2 = −2 + i. |Im z| < π} = {z = x + iy : x < 0. We will prove the statement above by contradiction. thus we have derived a contradiction 3. Deﬁne the subset G of G by G = C − {z ∈ R : z ≤ 0}.Hence. −π < y < π} = {e x (cos(y) + i sin(y)) : x < 0. We use the notation Log for the principal ˆ that is part of the log on G. 1. 1 ≤ i ≤ n. Since f is analytic in G. does the formula remain valid? Consider the complex numbers z1 = −1 + 2i. Then z1 z2 = 0 − 5i. log(z1 ) = ln |z1 | + i arg z1 log(z2 ) = ln |z2 | + i arg z2 √ √ (−π. Exercise 2. Log(z) = log |z| + i arg(z) where arg(z) ∈ (−π. We have {z : Re z < 0. Hence. it is continuous at −1. where the boundary does not belong to it. Clearly. |Im z| < π} under the exponential function. Möbius transformations Exercise 1. This yields f (z) = log |z| + i arg(z) + 2πik ˆ where z ∈ G. π). satisfy the restrictions. and z1 z2 do not meet the restrictions as stated above. Now. Prove that there is no branch of the logarithm deﬁned on G = C − {0}. 0) centered at (0. then r(cos(y) + i sin(y)) describes a circle with radius r without the point (−r. but note that 3π 2 2 2 ln 5 + i(− π ). z1 . −π < y < π}. (Hint: Suppose such a branch exists and compare this with the principal branch.3 Analytic functions as mappings. the negative x-axis does not belong to it and the origin does not belong to is. z2 . If 0 < r < 1 is ﬁxed. Thus log z1 + log z2 log(z1 z2 ) where log z is the principal branch of the logarithm. 0).) ˆ ˆ Solution. But we can check that this is not the case.

c To get a translation S (z) = z + b. c = 0. Exercise 5. To ﬁnd a ﬁxed point. we have that b = 0. 0) lying in the right half plane not touching the imaginary axis. To see that S (∞) = a = 0 = ∞. In this case z + b = z. 1 + i.) Solution. Since r varies between 0 and inﬁnity. b = 1. 1. In general. − < y < } 2 2 2 2 π π = {r(cos(y) + i sin(y)) : r > 0. we have to z ﬁnd all z such that S (z) = z. we have az + b . Not available. then r(cos(y) + i sin(y)) describes a half circle with radius r centered at (0. − {e x+iy : x ∈ R. 1.3) (3. a translation and the inversion on C∞ .4) 31 . 1. ∞. a since a· ∞ = ∞(a > 0) or S (∞) = a = 0 = ∞. We have π π < y < }. 2 2 If r > 0 is ﬁxed. Also z = ∞ is a ﬁxed point.Solution. To ﬁnd a ﬁxed point. 0). Evaluate the following cross ratios: (a) (7 + i. if z2 . 1. − π π π π < y < } = {e x (cos(y) + i sin(y)) : x ∈ R. 1−0 z − z3 z2 − z3 / . 1 − i. (Hint: use polar coordinates. c To get a inversion S (z) = 1 . 0. z2 − z3 (3. Solution. In this case 1 = z which is equivalent to z2 = 1 and thus z = 1 and z = −1 are z ﬁxed points.4) we get (7 + i. In this case az = z. 0. − < y < }. we have that a = 0. i. − π < y < π } under the exponential function is given by 2 2 {z : |Im z| < π/2} = {z = x + iy : x ∈ R. Exercise 4. if z4 = ∞. ∞) = 7+i−0 = 7 + i. We have S (z) = S (z) = S (z) = a) By (3. which is true if z = ∞. we get that the image is the right half plane without the imaginary axis. 1 + i) (c) (0. Solution. To ﬁnd a ﬁxed point. we have to ﬁnd all z such that S (z) = z. we have that a = 1. −1) (d) (i − 1. Exercise 3. S (z) = Exercise 6. if z2 = ∞ z − z4 z − z3 . 2 2 The image of {z = x + iy : x ∈ R. z3 . Solution. c = 0. we have 1 to ﬁnd all z such that S (z) = z. Discuss the mapping properties of zn and z1/n for n ≥ 2.2) (3. d = 1 and a > 0. z4 ∈ C z − z4 z2 − z4 z − z3 . Not available. Discuss the mapping properties of cos z and sin z. d = 1 and b ∈ R. ∞) (b) (2. Obviously z = 0 is a ﬁxed point. Find the ﬁxed points of a dilation. cz + d To get a dilation S (z) = az. c = 1 and d = 0.

then T (z)T (z) = 1. b. Not available.b) By (3. i−1−0 i − 1 i + 1 −2 ﬁnd z2 . 32 . Hence. z3 = − b and a show that T (R∞ ) = R∞ iﬀ we can choose a. ﬁnd necessary and suﬃcient conditions that T (Γ) = Γ where Γ is the unit circle Solution. Exercise 9. we have the two conditions ¯ ¯ ab − cd = 0 and |a|2 + |b|2 = |c|2 + |d|2 . If T z = az+b cz+d . i. c. c Exercise 8. c. z2 .3) we get (i − 1. Solution. 1 + i. z3 . z4 ). z a c ¯ This is the same as the equation z¯ − 1 = 0 if z b¯ − dc a ¯ and a¯ − c¯ a c = 1 ¯ ¯ ab − cd = 0 = 0 ¯ ¯ bb − dd = −1 which is equivalent to |a|2 − |c|2 = 1 and 1 = |d|2 − |b|2 . z4 (in terms of a. Set z2 = a c d−b a−c . 1 + i) = c) By (3. 47.2) we get (0. z2 . z3 .2) we get (2. b. 1−i−1 1 −i 2 1+i 1+i 2−1 / = / = =2 =2 = 1 + i. 0) = Exercise 7. T −1 (0) = − b and T −1 (∞) = − d . ∞. −1) = d) By (3. 0+1 1+1 2 1−i1+i 2 i−1−1−i −2 i + 1 −2 = = (i + 1) = 1 + i. 1. The inverse of T is given by as shown on p. 2 − 1 − i 1 − i − 1 − i 1 − i −2i 1 − i (1 − i)(i + 1) 2 1−i 2 1+i 2 0−i 1−i / = −i/ = −i = −i (1 + i) = −i − i2 = 1 − i. az+b cz+d . 1. d) such that T z = (z. We want if z¯ = 1. z3 . 1 − i. If T z = {z : |z| = 1}. z4 ). We have T −1 (1) = z4 = − d to obtain T z = (z. d to be real numbers. T −1 (z) = dz − b −cz + a Solution. z T (z)T (z) = 1 ⇐⇒ ⇐⇒ ⇐⇒ az + b az + b (az + b)(¯ z + b) a¯ ¯ = 1 ⇐⇒ =1 cz + d cz + d (cz + d)(¯ z + d) c¯ ¯ ¯ ¯ ¯ az¯ z + b¯ z + azb + bb = cz¯ z + dcz + dcz + dd a¯ a¯ c¯ ¯¯ ¯ ¯ ¯ ¯ a ¯ z¯(a¯ − c¯ ) + z(ab − cd) + z(b¯ + dc) − dd = 0. If T z = az+b cz+d d−b a−c .

then This mapping transforms |z| = 1 into |T (z)| = 1. αz − 1 ¯ So |K| = 1 implies K = eiθ for some real θ. Take an α ∈ D = {z : |z| < 1} such that T (α) = 0. for some θ. Exercise 11. Its symmetric point with respect to the unit circle is 1 α∗ = α ¯ since R2 z∗ − a = z−a ¯ ¯ (see page 51) with a = 0 and R = 1 (the unit circle).a ¯ ¯ ¯ ¯ ¯ ¯¯ ¯ These are the suﬃcient conditions. |λ|2 = λ. Let c = λb. ω4 ) = (z. we are going to ¯ iθ choose the constant K in such a way that |T (z0 )| = 1 where z0 = e . show that if ω2 . Thus T looks like T (z) = K z−α αz − 1 ¯ 1 where K is a constant. Finally. ω3 . ω4 are also in Γ then equation (3. and so the form of the Möbius transformation is az + b . ω3 . Let D = {z : |z| < 1} and ﬁnd all Möbius transformations T such that T (D) = D. Therefore T (α∗ ) = ∞.) 33 . ¯ 2 2 2 2 Insert this into |c| + |d| = |a| + |b| yields |λ|2 |b|2 + and therefore we can take |λ| = 1. ¯ λ(bz + a) ¯ where |λ| = 1 where |λ| = 1. ¯ bz + a ¯ Take λ = e−iθ . T (z) = λ ¯ bz + a ¯ T (z) = eiθ az + b . We arrive at T (z) = eiθ for some real θ. ω2 . That is. z4 . then ab − cd = 0 yields ab − λbd = 0 iﬀ a = λd iﬀ d = λ . Show that the deﬁnition of symmetry (3. (Hint: Use Exercise 8. (It is easy to check that T (α) = 0 and T (α∗ ) = T ( α ) = ∞). Exercise 10. ω3 .17) does not depend on the choice of points z2 . Solution. T (z) = or ¯ az + b . ω4 ). z3 . = |K| = |K| −iθ α − eiθ iθ − 1| iθ |· |α − e−iθ | α−e ¯ |αe ¯ |e ¯ z−α . ω2 . We have T (z0 ) = K and therefore 1 = |T (z0 )| = |K| eiθ − α αeiθ − 1 ¯ eiθ − α e−iθ − α ¯ eiθ − α e−iθ − α ¯ |eiθ − α| = |K|.18) is satisﬁed iﬀ (z∗ . Then 1 ¯ λ |a|2 = |a|2 + |b|2 .

z1 z2 z1 z2 z1 z2 With the assumption 0 < |zi | < 1. To see this. ∞). hence f (z) is injective. 2 Applying the exponential function to this region yields the right half plane exp(S (T (G))) = {x + iy : x > 0}. a. 1]. b = 0 cannot be in the range of f . Give a discussion of the mapping f (z) = 2 (z + 1/z). Therefore points of the form a + ib. Solution. (Hint: ﬁrst try (z − a)−1 .Solution.5) sin θ. The function f (z) = 1 z + 1 = z 2z can be deﬁned for all z ∈ C − {0} and therefore also in the 2 z punctured disk 0 < |z| < 1. If cos θ 0 and sin θ 0 then cos θ − sin θ = 1. Prove Theorem 3. it is possible to send this region to any region bounded by two parallel lines we want. is always nonzero and we conclude that z1 = z2 . +1 Solution. a ∈ [−1. Exercise 14. In fact from formulas (3.) Solution. b ∈ R. then 2 0 = f (z1 ) − f (z2 ) = z2 + 1 z2 + 1 z2 z2 + z2 − z1 z2 − z1 (z1 z2 − 1)(z1 − z2 ) 2 1 − 2 = 1 = .5) that the graph of f (reiθ ) is a hyperbola and it degenerates to rays if z is purely real or imaginary. Hence. 34 . To see that in this domain the function is injective. For all other points the equations (3. deﬁne the Möbius transformation T (z) = (z − a)−1 which sends the region G between two lines.5) can be solved for r and θ uniquely (after restricting the argument to [−π. choose S (z) = cz + d where |c| = 1 such that S (T (G)) = x + iy : 0 < y < π . Not available. In the case θ ∈ {(2k + 1)π | k ∈ Z} the graph of f in dependence on r is on the imaginary axis and for θ ∈ {2kπ | k ∈ Z} the graph of f (reiθ ) is a 2 b 2 either (−∞. 1). Not available. Exercise 12.4. 2 the factor z1 z2 −1. If f (z) = w has imaginary part b = 0 then sin θ = 0 and |cos θ| = 1. If we ﬁx the argument θ and let r vary in (0. Given any value of r ∈ (0. i = 1. Afterward applying a rotation followed by a translation. z2 be two numbers in the domain of f with the same image. Let G be the region between the two circles and map G conformally onto the open unit disk. the graph of f (reiθ ) as a function of θ looks like an ellipse. r r For the real and imaginary part of w the following equations must hold a b = = 1 2 1 2 r+ r− 1 r 1 r cos θ (3. 1) it follows from from equation (3. π)). 1 Exercise 13. The range of the function is C − {z ∈ C | z ∈ [−1. 1] and Im z = 0}. Suppose that one circle is contained inside another and that they are tangent at the point a. Using the hint.5) we see that a 2 1 1 2 (r+ r ) + b 2 1 1 2 (1− r ) = 1. f (z) = f (reiθ ) = 1 1 1 reiθ + e−iθ = 2 r 2 r+ 1 1 cos θ + i r − sin θ . write z = reiθ in polar coordinates and let f (z) = w = a + ib. let z1 . −1) or (1.

too. Thus. h(z) is the angle depicted in the ﬁgure. M maps C to Y and F to V. Let −∞ < a < b < ∞ and put Mz = z−ib . Determine which of the regions A. the Möbius transformation z−1 z+1 maps the right half plane onto the unit disk (see page 53). Can f be one-one? Solution. and if h(z) = Im [log Mz] then 0 < h(z) < π for Re z > 0. B. Doing some simpliﬁcations. Let x. Deﬁne the lines L1 = {z : Im z = b}. Thus we conclude that C or F goes to either V or Y. y be small positive real numbers such that the poin z = x + iy + ia ∈ E. W. Not available. are mapped by M onto the regions U. (b) Show that log(z − ic) is deﬁned for Re z > 0 and any real number c. also prove that |Im log(z − ic)| < π 2 if Re z > 0. Not available. . C. Z in Figure 2. b. Therefore we conclude that M maps E to U and B to X. (a) Show that log(Mz) is deﬁned for all z except z = ic. D. Y.Finally. Show that f is constant. we obtain that M maps A to Z and D to W. Let G be a region and suppose that f : G → C is analytic such that f (G) is a subset of a circle. Let us ﬁnd out. Exercise 17. L2 = {z : Im z = a} and L3 = {z : Re z = 0}. X. Let a. (c) Let h be as in (a) and prove that h(z) = Im [log(z − ia) − log(z − ib)]. X and Y which touch 0. Therefore the region B. We easily see that we have M(ia) = 0. Exercise 19. Because B and C meet at the line ia. V. z−ia Exercise 18. Similarly we have M(ib) = ∞ and therefore the region B and E which touch the line ib are mapped somehow to the region U and X which touch ∞. By a similar argument. V. E and F which touch the line ia are mapped somehow to the region U. C. F in Figure 1. E. Hence the function f deﬁned by R(exp(S (T (z)))) maps G onto D and is the desired conformal mapping ( f is a composition of conformal mappings). we obtain c e z−a +d − 1 f (z) = c +d e z−a + 1 where the constants c and d will depend on the circle location. a ≤ c ≤ b. Exercise 16. Solution. the imaginary part of Mz is a positive number multiplied by x(b − a) and therefore also positive. (d) Show that b dt = i[log(z − ib) − log(z − ia)] z − it a (Hint: Use the Fundamental Theorem of Calculus. Not available.) (e) Combine (c) and (d) to get that b h(x + iy) = a y−b y−a x − arctan dt = arctan x x x2 + (y − t)2 35 (f) Interpret part (e) geometrically and show that for Re z > 0. Can you map the open unit disk conformally onto {z : 0 < |z| < 1}? Solution. Hence. we conclude that X and M(C) do. Map G = C − {z : −1 ≤ z ≤ 1} onto the open unit disk by an analytic function f . Solution. R(z) = Exercise 15. and M be as in Exercise 18 and let log be the principal branch of the logarithm.

36 β b β ⇒ = ⇒ β = λb. Solution. δ d dλ α a α ⇒ = ⇒ α = λa. ⇐: Let λ αz+β γz+δ . λ 0 or δ c γ δ Insert (3. S (1) = T (1) and S (∞) = T (∞) which is equivalent to β b = = λ1 ⇐⇒ b = λ1 d. β = λ1 δ. γ c cλ = cλ = dλ.8) c γ Insert b = λ1 d.8) to obtain a = c Therefore. then λ3 c + λ1 d λ3 γ + λ1 δ = c+d γ+δ λ3 cγ + λ1 dγ + λ3 cδ + λ1 dδ = λ3 γc + λ1 δc + λ3 γd + λ1 δd λ1 dγ + λ3 cδ − λ1 δc − λ3 γd = 0 λ1 (dγ − δc) − λ3 (dγ − δc) = 0 (λ1 − λ3 )(dγ − δc) = 0. (3. then from (3. Not available.7). a = λ3 c and α = λ3 γ into (3.10) into (3. dγ − δc = 0 ⇐⇒ d = γ ⇐⇒ γ = d := λ. b If λ1 − λ3 = 0.7) c+d γ+δ α a = = λ3 ⇐⇒ a = λ3 c.9) (3.Solution.6) and (3. either λ1 − λ3 = 0 or dγ − δc = 0.10) into (3. c δ Thus.6) d δ α+β a+b = (3. we have α = λa. (3. α = λ3 γ. γ = λc. β = λb. Let S z = az+b and T z = cz+d that α = λa. we get λ3 = a = d = λ1 which implies ad − bc = 0 a contradiction c (Not possible. otherwise we do not have a Möbius transformation). Then S (0) = T (0). show that S = T iﬀ there is a non zero complex number λ such 0 be a complex number such that α = λa β = λb γ = λc δ = λd. Then T (z) = αz + β λaz + λb λ(az + b) (az + b) = = = = S (z). γ = λc.6) to obtain b = d Insert (3. ⇐⇒ ⇐⇒ ⇐⇒ ⇐⇒ Thus.10) . β = λb. β = λ1 δ (3. δ = λd.9) and (3. ⇒: Let S = T .7).9) and (3. γz + δ λcz + λd λ(cz + d) (cz + d) Thus. S = T . Exercise 20. δ = λd. that is S (z) = T (z).

Exercise 23. Similar. the second step since S S −1 = id and the last step by T z1 = z1 since z1 is a ﬁxed point of T . We have S −1 T S (S −1 z1 ) = S −1 T S S −1 z1 = S −1 T z1 = S −1 z1 . The composition of Möbius transformation is a Möbius transformation. To show S −1 T S (S −1 z2 ) = S −1 z2 . To show T S = S T . we obtain MT M −1 (0) = 0 and MT M −1 (∞) = ∞ and therefore is also a dilation. Not available. thus (MT M −1 )(MS M −1 ) = (MS M −1 )(MT M −1 ) ⇒ MT S M −1 = MS T M −1 ⇒ T S = S T. Exercise 24. Let T be a Möbius transformation. . (a) Show that a Möbius transformation has 0 and ∞ as its only ﬁxed points iﬀ it is a dilation. Note that S −1 T S is a well deﬁned Möbius transformation. but not the identity.Exercise 21. Solution. and S −1 z2 . Solution. Let T be a Möbius transformation with ﬁxed points z1 and z2 . Show that a Möbius transformation T satisﬁes T (0) = ∞ and T (∞) = 0 iﬀ T z = az−1 for some a in C. Show that a Möbius transformation S commutes with T if S and T have the same ﬁxed points. but not the identity. To show S −1 T S (S −1 z1 ) = S −1 z1 .) Solution. Let M be a Möbius transformation with M(z1 ) = 0 and M(z2 ) = ∞. Exercise 22. since S and T are Möbius transformation. where the ﬁrst step follows by the associativity of compositions. 37 MS M −1 (∞) = MS M −1 Mz2 = MS z2 = Mz2 = ∞. It is easy to check that dilations commute (deﬁne C(z) = az. b > 0. then CD(z) = abz = baz = DC(z)). where the ﬁrst step follows by the associativity of compositions. We have S −1 T S (S −1 z2 ) = S −1 T S S −1 z2 = S −1 T z2 = S −1 z2 . (Hint: Use Exercises 21 and 22. T the identity. If S is a Möbius transformation show that S −1 T S has ﬁxed points S −1 z1 . Solution. the second step since S S −1 = id and the last step by T z2 = z2 since z2 is a ﬁxed point of T . Then MS M −1 (0) = MS M −1 Mz1 = MS z1 = Mz1 = 0 and Thus MS M −1 is a dilation by exercise 22 a) since MS M −1 has 0 and ∞ as its only ﬁxed points. a > 0 and D(z) = bz. (b) Show that a Möbius transformation has ∞ as its only ﬁxed point iﬀ it is a translation. T id. : Suppose T and S have two ﬁxed points. Let T and S have the same ﬁxed points. Not available. say z1 and z2 .

then CD(z) = z + a + b = z + b + a = DC(z)). A simple calculation shows that this is true. Not available. show that S U2 is a group under matrix 38 . G is a simple group if the only normal subgroups of G are {I} (I = the identity of G) and G itself. Show that the image of S L2 (C) under ϕ is all of M. Solution. Discuss the mapping properties of (1 − z)i . Thus MS M −1 is a translation by exercise 22 b) since MS M −1 has ∞ as its only ﬁxed point. Exercise 29. b) Restricting ϕ to S L2 (C) still yields a surjective map since for any matrix A ∈ GL2 (C) both A and 1 the modiﬁcation M = √det A A have the same image and the modiﬁcation matrix M has by construction determinant 1. Not available. (b) If S U2 is the set of all unitary matrices with determinant 1. Solution. 26. Solution. To ﬁnd the kernel of the group homomorphism we have to ﬁnd all z such that az+b = z. What part of the kernel of ϕ is in S L2 (C)? Solution. Exercise 25. Exercise 27. (a) Let GL2 (C) = all invertible 2 × 2 matrices with entries in C and let M be the group a b = az+b . Let M be a Möbius transformation with M(z) = ∞. (b) Let S L2 (C) be the subgroup of GL2 (C) consisting of all matrices of determinant 1. Exercise 28. Show that ϕ is a group of Möbius transformations.: Suppose T and S have one ﬁxed points. The kernel of the restriction is simply N ∩ S L2 (C) = {±I}. Find all the abelian subgroups of the group of Möbius transformations. Deﬁne ϕ : GL2 (C) → M by ϕ cz+d c d homomorphism of GL2 (C) onto M. It is easy to check that translations commute (deﬁne C(z) = z + 1 and D(z) = z + b. Not available. Find the kernel of ϕ.β = ¯ αz − β βz + α ¯ and let U = {uα. Exercise 26. Prove that the group M of Möbius transformations is a simple group. we obtain MT M −1 (∞) = ∞ and therefore is also a translation.β : |α|2 + |β|2 = 1} (a) Show that U is a group under composition. say z. Therefore. a) We have to check that if A= a b c d and B= a b ˆ ˆ c d ˆ ˆ then ϕ(AB) = ϕ(A) ◦ ϕ(B). If G is a group and N is a subgroup then N is said to be a normal subgroup of G if S −1 T S ∈ N whenever T ∈ N and S ∈ G. b > 0. Then MS M −1 (∞) = MS M −1 Mz = MS z = Mz = ∞. the kernel is given by N = ker(ϕ) = {λI : λ ∈ C× }. For complex numbers α and β with |α|2 + |β|2 = 1 uα. This is equivalent to az + b = cz2 + dz and by cz+d comparing coeﬃcients we obtain b = c = 0 and a = d. Similar. Note that the kernel is a normal subgroup of GL2 (C). thus (MT M −1 )(MS M −1 ) = (MS M −1 )(MT M −1 ) ⇒ MT S M −1 = MS T M −1 ⇒ T S = S T.

For uα. onto Hl .β is an isomorphism of the group S U2 onto U. and u → T u ¯ is an injective homomorphism of U into the group of invertible linear transformations of Hl . (b) Find all Möbius transformations S (z) that map D onto D and such that f (S (z)) = f (z) when |z| < 1. 3 .multiplication and that for each A in S U2 there are unique complex numbers α and β with |α|2 + |β|2 = 1 and α β A= ¯ ¯ −β α α β ¯ ¯ → uα. Not available.β = u in U deﬁne T u : Hl → Hl . Not available. . . What is its kernel? −β α (l) (d) If l ∈ {0. Exercise 30. For |z| < l deﬁne f (z) by 1+z f (z) = exp −i log i 1−z 1/2 (a) Show that f maps D = {z : |z| < 1} conformally onto an annulus G. 39 . 2 2 (l) (l) (l) by (T u f )(z) = (βz + α)2l f (u(z)).} let Hl = all the polynomials of degree ≤ 2l. 1 . 1. (c) Show that Solution. . Solution. Show that T u is an invertible linear transformation on Hl .

P) ≥ v(γ. . Q). Solution. . Q) ≥ v(γ. Let P = {a = t0 < t1 < . Solution. then αγ + βσ is of bounded variation and 40 . .1 Riemann-Stieltjes integrals Exercise 1.Chapter 4 Complex Integration 4. < tm = b}. v(γ. For a given partition of [a. β ∈ C. + [γ(t2 ) − γ(t1 )] + [γ(t1 ) − γ(t0 )] = [γ(b) − γ(tm−1 )] + [γ(tm−1 ) − γ(tm−2 )] + . It follows that γ is of bounded variation for any partition P. P) ≤ v(γ. |γ(tn ) − γ(tn−1 )| = γ(tn ) − γ(tn−1 ) ≥ 0. . . b] → C is also of bounded variation and α. Prove Proposition 1. Let γ : [a. for all n ∈ N. whenever P ⊂ Q. . . < tm = b}. b]. Then. . Q \ P) ≥ |γ(b) − γ(a)|. b] → R be a monotone. Then. (b) If σ : [a. b]. non-decreasing function. The smallest partition set for Q \ P contains the trivial partition composed of {a. P) : P a partition of [a. m n=1 |γ(tn ) − γ(tn−1 )| = [γ(tm ) − γ(tm−1 )] + [γ(tm−1 ) − γ(tm−2 )] + . b]} = V(γ) = γ(b) − γ(a). Using the case deﬁned above. Suppose Q = P ∪ {x}. . P) = sup{v(γ. Let γ : [a.2. γ(b) − γ(a) < ∞. Therefore. + [γ(t2 ) − γ(t1 )] + [γ(t1 ) − γ(a)] = γ(b) − γ(a) Since γ(a). Show that γ is of bounded variation and V(γ) = γ(b) − γ(a). γ(b) ∈ R. Let γ : [a. (a) If P and Q are partitions of [a. b}. Exercise 2. Proof will be by induction. Q) Now consider the general case on the number of elements of Q \ P. where tk < x < tk+1 . P) = i k+1 |γ(ti ) − γ(ti−1 )| + |γ(tk+1 ) − γ(tk )| |γ(ti ) − γ(ti−1 )| + |γ(x) − γ(tk )| + |γ(tk+1 ) − γ(x)| ≤ i k+1 = v(γ. . any partition of [a. b] and P ⊂ Q then v(γ. b] → C be of bounded variation. we have v(γ. Let P = {a = t0 < t1 < . And also v(γ. {a = t0 < t1 < . b] → R be non decreasing. γ(tn ) ≥ γ(tn−1 ). < tm = b}.

Prove Proposition 1. such that v(γ. Exercise 3. Give a rough sketch of the trace of γ. Then. . b] → C is a Lipschitz function then γ is of bounded variation. Let γ(t) = exp((−1 + i)/t−1 ) for 0 < t ≤ 1 and γ(0) = 0. Exercise 6. Let P = {a = t0 < t1 < . < tm = b} be a partition. Exercise 4. P) : P a partition of [a. b] ⊂ R.7. P) = k=1 |σ(tk ) − σ(tk−1 )| ≤ Mσ . P) ≤ |α|sup{v(γ. 41 . m v(αγ + βσ. Since γ and σ are of bounded variation for [a. Not available. P) : P a partition of [a. Solution. Not available. . Show that γ is a rectiﬁable path and ﬁnd V(γ). P) + |β|v(σ. Solution. Show that if γ : [a. Exercise 5.V(αγ + βσ) ≤ |α|V(γ) + |β|V(σ). b]} + |β|sup{v(σ. b]} The result is two-fold.8 (Use induction). Solution. respectively. P) = k=1 m |(αγ + βσ)(tk ) − (αγ + βσ)(tk−1 )| |αγ(tk ) + βσ(tk ) − αγ(tk−1 ) − βσ(tk−1 )| |(αγ(tk ) − αγ(tk−1 )) + (βσ(tk ) − βσ(tk−1 ))| |αγ(tk ) − αγ(tk−1 )| + |βσ(tk ) − βσ(tk−1 )| m = k=1 m = k=1 m ≤ = k=1 m k=1 m |αγ(tk ) − αγ(tk−1 )| + k=1 m |βσ(tk ) − βσ(tk−1 )| |β| |(σ(tk ) − σ(tk−1 ))| |σ(tk ) − σ(tk−1 )| = k=1 |α| |(γ(tk ) − γ(tk−1 ))| + m k=1 m = |α| k=1 |γ(tk ) − γ(tk−1 )| + |β| k=1 = |α|V(γ) + |β|V(σ) <∞ = |α|v(γ. m there exists constants Mγ > 0 and Mσ > 0. P) = m k=1 |γ(tk ) − γ(tk−1 )| ≤ Mγ and v(γ. Not available. V(αγ + βσ) ≤ |α|V(γ) + |β|V(σ) and it follows αγ + βσ is of bounded variation by |α|V(γ) + |β|V(σ). Prove Proposition 1.

Sketch this path. . m m ∀x. 3 3 Since σ is composed of multiple lines. since there is a constant M > 0 such that for any partition P = {a = t0 < t1 < . i] and [1. i]. 0 and γ(0) = 0. Hence. we have v(γ. b] → C be Lipschitz. b]. Express γ and σ as paths and calculate f and σ f where f (z) = |z|2 . that is ∃ a constant C > 0 such that |γ(x) − γ(y)| ≤ C|x − y|. Let γ : [a. 1 γ |z|2 dz = 0 ((1 − t)2 + t2 )(−1 + i)dt 1 0 = (−1 + i) = (−1 + i) (2t2 − 2t + 1)dt 1 0 2 3 2 t −t +t 3 2 −1+1 = (−1 + i) 3 2 2 =− +i . P) = k=1 |γ(tk ) − γ(tk−1 )| ≤ k=1 C|tk − tk−1 | = C k=1 |tk − tk−1 | = C(b − a) =: M > 0. Solution. γ Solution. for [a. the function γ : [a. Exercise 8. b]. 1 + i. b] m v(γ. γ2 (t) = (1 − t) + i : 42 . Therefore. b] ⊂ R. Show that γ : [0. is of bounded variation. b] → C. < tm = b} of [a. y ∈ [a. < tm = b} of [a. . m Then. For γ we have γ(t) = (1 − t) + it for 0 ≤ t ≤ 1. the path is given by {γ1 (t) = 1 + it : 0 ≤ t ≤ 1. So γ (t) = −1 + i. Not available. Let γ and σ be the two polygons [1. . is a path but is Exercise 7. 1] → C.Solution. P) = k=1 |γ(tk ) − γ(tk−1 )| ≤ M. for any partition P = {a = t0 < t1 < . . deﬁned by γ(t) = t + it sin 1 for t t not rectiﬁable.

Deﬁne γ : [0. Show that γ 1 dz = 2πin. Exercise 10. Clearly. 1 + i. Thus 2π 2π z−1 dz = γ 0 e−int ineint dt = 0 in dt = in(2π − 0) = 2πin. γ(t) = eint is continuous and smooth on [0. zn dz for every integer n. 1 − i]. γ(t) = eint is continuous and smooth on [0. 0 γ 0 Case 2: n −1 2π 2π zn dz γ = 0 eint ieit dt = i 0 ei(n+1)t dt = t ei(n+1)t i(n + 1) 2π 0 2π 1 1 = ei(n+1)t = ei(n+1)2π − 1 0 n+1 n+1 1 1 [cos((n + 1)2π) − i sin((n + 1)2π)1] = [1 − i· 0 − 1] = n+1 n+1 = 0. Find 1 γ z dz. z Solution. f = γ γ1 f+ γ2 f |z|2 dz 1 0 1 = γ1 1 |z|2 dz + γ2 = 0 (t2 + 1)(i)dt + 1 ((1 − t)2 + 1)(−1)dt (t2 − 2t + 2)dt 1 0 =i 0 (t2 + 1)dt − 1 0 t3 +t =i 3 t3 − t2 + 2t − 3 0 4 1 = i− −1+2 3 3 4 4 = i− 3 3 4 = (−1 + i) 3 Exercise 9. 2π] (It is the unit circle). Let γ be the closed polygon [1 − i. Deﬁne γ(t) = eit for 0 ≤ t ≤ 2π and ﬁnd γ Solution. or zero). −1 + i. Clearly. −1 − i. 2π]. Case 1: n = −1 2π 2π z−1 dz = e−it ieit dt = i dt = 2πi. 43 . Therefore. Exercise 11.0 ≤ t ≤ 1}. So γ1 (t) = i and γ2 (t) = −1. negative. 2π] → C by γ(t) = exp(int) where n is some integer (positive.

then |I(r) − 0| = |I(r)| = eiz dz = z π 0 γ eire rieit dt = i reit it π 0 eire dt ≤ it π 0 eire it dt. π 0 eire it π dt = 0 e−r sin(t) dt = 0 δ1 e−r sin(t) dt + π−δ2 δ1 π e−r sin(t) dt + π−δ2 e−r sin(t) dt for δ1 > 0 and δ2 > 0. π] → C is deﬁned by γ(t) = reit . = −1 1 = −1 1 1 −t + i + t + i −1 + it − 1 − it dt + i dt −1 (t + i)(t − i) −1 (1 + it)(−1 + it) 1 1 1 1 dt − 2i dt = 2i 2+1 2 −1 −1 − t −1 t 1 1 1 1 = 2i dt + 2i dt 2+1 t t2 + 1 −1 −1 1 π π 1 − − dt = 4i [arctan(z)]1 = 4i = 4i −1 4 4 t2 + 1 −1 −1 −1 1 1 1 i dt + dt + i dt + 1 + it t+i −1 + it 1 1 1 −1 1 1 1 dt + i dt + − t+i t−i −1 + it −1 1 + it 1 −1 1 dt t−i = = 4i π = 2πi. recall that |ez | = eRe(z) (p. = −1 + it. Let I(r) = eiz γ z dz where γ : [0. Since e−r sin(t) is continuous. ∀r > 0. Deﬁne γ1 (t) γ2 (t) γ3 (t) γ4 (t) Thus γ = 1 + it. = t − i. Similar. Solution. so δ1 > 0 can be chosen such that δ1 0 e−r sin(t) dt < 3 . 2 Exercise 12. Now. so it it eire = eRe(ire ) = eRe(ir cos(t)+ir·i sin(t)) = eRe(ir cos(t)−r sin(t)) = e−r sin(t) .13). then |I(r)| < . To show limr→∞ I(r) = 0 which is equivalent to ∀ > 0. Show that limr→∞ I(r) = 0. 44 . = t + i. ∃n ∈ N such that if r > N. 2. Let > 0.Solution. 38 eq. 1 dz z 1 t from − 1 to 1 t from 1 to − 1 t from 1 to − 1 t from − 1 to 1 → → → → γ1 (t) γ2 (t) γ3 (t) γ4 (t) = i = 1 = i = 1. ∀r > 0. δ2 > 0 can be chosen such that π e−r sin(t) dt < π−δ2 3 . Hence.

Prove Proposition 1. Hence. Solution. b] → [c. Solution. ϕ(b) = d then ϕ is one-one iﬀ ϕ is strictly increasing. Let h = F1 − F2 for all z ∈ G. Not available. (z2 − 1)−1 dz. 1] → C. Then for all z ∈ G. say c: h = c. Solution. Exercise 17. Let γ = 1 + eit for 0 ≤ t ≤ 2π and ﬁnd Solution. Exercise 13. r→∞ 1 lim I(r) = 0. Exercise 21. then h is a constant. Show that if F1 and F2 are primitives for f : G → C and G is open end connected then there is a constant c such that F1 (z) = c + F2 (z) for each z in G. 45 γ γ (z2 − 1)−1 dz. Not available. Exercise 16. Let γ = 2eit for −π ≤ t ≤ π and ﬁnd Solution. Solution.16 is an equivalence relation. Exercise 15.17. h (z) = F1 (z) − F2 (z) = f (z) − f (z) = 0 Since G is open and connected with h (z) = 0 for all z ∈ G. Show that the relation in Deﬁnition 1. Not available. Prove that if ϕ : [a. Exercise 18. Find γ z− 2 dz where: (a) γ is the upper half of the unit circle from +1 to −1: (b) γ is the lower half of the unit circle from +1 to −1. Not available. b] → C is a path then there is an equivalent path σ : [0. Exercise 20. Exercise 19. So h = F1 − F2 = c and F1 (z) = F2 (z) + c for all z ∈ G. . Not available. Recall then that F1 = f = F2 . d] is continuous and ϕ(a) = c. Exercise 14. Show that if γ and σ are equivalent rectiﬁable paths then V(γ) = V(σ). Not available. Solution. Not available. Let F1 and F2 be primitives for f : G → C. Solution. Solution.Finally. Show that if γ : [a. because of the Lebesgue’s dominated convergence Theorem π−δ2 r→∞ lim e−r sin(t) dt = π−δ2 δ1 δ1 π−δ2 r→∞ r→∞ lim e−r sin(t) dt we have lim e−r sin(t) dt < δ1 3 . Not available.

γ f g = f (b)g(b) − f (a)g(a) − f g. z2 ∈ C.s rule (this exercise will be frequently used in the later sections. Prove the following analogue of Leibniz. γ 4. z) dw.18 we obtain ( f g) = ( f g)(b) − ( f g)(a) = f (b)g(b) − f (a)g(a). Exercise 2. γ (z − a)−n dz = F(z2 ) − F(z1 ) = 0. Since z1 = z2 . So. z) dw then g is continuous. ∂z Solution.2 Power series representation of analytic functions Exercise 1. Then f has the antiderivative F(z) = 1−n F (z) = f (z) where n ≥ 2 (note F is undeﬁned for n = 1). (z − a)1−n on G. z) in {γ} × G and is continuous then g is analytic and g (z) = γ ∂ϕ (w.) Let G be an open set and let γ be a rectiﬁable curve in C.7 (a) f (γ(t))g (γ(t)) + f (γ(t))g(γ(t)) dγ(t) a = (def) fg + f g = γ γ ( f g) Since ( f g) is continuous ( f and g are analytic) by Theorem 1. Let f. Then b b fg + γ γ f g = (def) f (γ(t))g (γ(t)) dγ(t) + a b a f (γ(t))g(γ(t)) dγ(t) = Prop.2) is continuous. If ∂ϕ ∂z exists for each (w. Then γ f g = f (b)g(b) f (a)g(a) − γ f g. Not available. Let f and g be analytic in G and let γ be a rectiﬁable curve from a to b in G. γ G. Prove the following integration by parts formula. Now. Not available. Show that the function deﬁned by (2. Exercise 23.Exercise 22. Show that for n ≥ 2. Solution. let γ be a closed rectiﬁable curve in G with endpoints z1 . Solution. Let f (z) = (z − a)−n for all z ∈ G. F(z1 ) = F(z2 ). Suppose that ϕ : {y} × G → C is a continuous function and deﬁne g : G → C by g(z) = γ ϕ(w. Let γ be a closed rectiﬁable curve in an open set G and a (z − a)−n dz = 0. Since Solution. 1. g be analytic in G and let γ be a rectiﬁable curve from a to b in G. γ Hence. 46 .

From the deﬁnition of numbers with rational exponents za = exp(a log z) we see that the square root cannot be analytic on all of C.) (b) Use Abel’s Theorem to prove that log 2 = 1 − 1 + 1 − . Give the power series expansion of log z about z = i and ﬁnd its radius of convergence. Use Exercise 2 to show that ϕ(w) dw g(z) = w−z γ is analytic on C − {γ} and g(n) (z) = n! γ ϕ(w) dw. The convergence radius is |i| = 1. Solution. Exercise 4. 2 3 Solution. Let a = i. Suppose that γ is a rectiﬁable curve in C and ϕ is deﬁned and continuous on {γ}. (a) Prove Abel’s Theorem: Let an converges to A. (w − z)n+1 Solution. .. Give the power series expansion of z about z = 1 and ﬁnd its radius of convergence. then 1 1 1 = = z a+z−a a 1 a+z−a a = 1 1 1 = a 1 + z−a a a ∞ n=0 (−1)n (z − a)n = an ∞ n=0 (−1)n (z − a)n . f (1) = an = n! n!2n k=0 and with these an . Exercise 5. Prove that an (z − a)n have radius of convergence 1 and suppose that lim an rn = A.Exercise 3. Solution. 47 . Since the logarithm is undeﬁned at zero. f (z) = √ z= ∞ n=0 an (z − 1)z = ∞ k=0 1 n!2n n−1 k=0 (1 − 2k)(z − 1)n . the radius of convergence is 1. Not available. . taking the ﬁrst derivatives we see that f (n) (z) = 2−n n−1 (1 − k=0 1 2k)z−n+ 2 and therefore n−1 1 1 (n) . Assume a ∈ C is not zero. n since Log(i) = π i. r→1− (Hint: Find a summation formula which is the analogue of integration by parts. 2 √ Exercise 6. then integrate term by term yields π log(z) = i + 2 ∞ n=0 π (z − i)n+1 + i − 2 (n + 1)in+1 i2 n ∞ n=0 π in+1 (z − i)n+1 = i − n+z 2 ∞ n=1 (1 + iz)n . an+1 where the radius of convergence is |a|. In fact. Not available. no matter which branch we consider.

Clearly f (z) is ¯ analytic on C (p. 0) in C). n = 2 and f (z) = sin(z). log(z) dz = 0. z2 ¯ b) Let a = a. Use a Möbuis transformation to show that Proposition 2. r = r. Not available. 2 Solution. 1 ) where n ≥ 0. Now use Corollary 2. γ (d) γ log z dz. zn 1 γ(t) = 1 + eit . n = 1.13. Then f (z) = cos(z) and f (z) = − sin(z). Clearly f (z) is analytic on C and B(a. zn γ Exercise 8. r) ⊂ C. z−a γ γ γ c) Let a = 0. and f (z) = 1. (c) γ(t) = eit . Now use Corollary 2. γ is a closed rectiﬁable curve (γ(t) is a circle with 2 radius r = 1/2 and center (1. 0 ≤ t ≤ 2π. r = 1. 2 γ z (b) γ dz . Therefore. f (0) = 1! 2πi γ 1 eiz dz ⇐⇒ i = 2πi (z − 0)2 γ eiz dz ⇐⇒ z2 γ eiz dz = −2π. f (z) = eiz and f (z) = ieiz . γ(t) = eit . 0 ≤ t ≤ 2π. Hence. Solution. n = 0. f (0) (a) = 0! 2πi 1 1 dz ⇐⇒ 1 = 2πi (z − a)1 1 dz ⇐⇒ z−a 1 dz = 2πi. z3 log(z) zn d) We have that f (z) = is analytic in the disk B(1. 1) ⊂ C. a) Let a = 0. 38) and B(0. z−a sin(z) dz.13. 1 ). f 2 2 has a primitive and therefore γ f = 0. Use the results of this section to evaluate the following integrals: (a) eiz dz. 0 ≤ t ≤ 2π and n ≥ 0.Exercise 7. r = 1. by Proposition 2. 1) ⊂ C. Obviously γ(t) = 1 + 1 eit ⊆ B(1. 0 ≤ t ≤ 2π. 48 .15 holds if the disk B(a. R) is replaced by a half plane. Clearly f (z) is analytic on C and ¯ B(0.15. Now use Corollary 2.13. z3 γ(t) = a + reit . Furthermore. f (0) = 2! 2πi γ sin(z) dz (z − 0)3 ⇐⇒ ⇐⇒ − sin(0) = γ 1 πi γ sin(z) 1 dz ⇐⇒ 0 = πi z3 γ sin(z) dz z3 sin(z) dz = 0.

zn γ (b) dz γ z− 1 n 2 where n is a positive integer and γ(t) = 1 + eit . m odd dz = 2 . a) Let a = 0. n = m − 1 and ¯ Clearly f (z) is analytic on C and B(0.Exercise 9. r = 1. z2 + 1 sin z dz where γ(t) = eit . 0 ≤ t ≤ 2π. m even γ ez − e−z zm γ 0 ez − e−z dz = 4πi m z (m > 0). . Solution. . n = m − 1 and (m − 1)! 2πi (m − 1)! 2πi ez − e−z dz m γ (z − 0) 0 . Now use Corollary 2. 0 ≤ t ≤ 2π. m odd . r = 1. (m−1)! f (z) = 1 f (z) = 0 f (z) = 0 . 0 ≤ t ≤ 2π. (Hint: expand (z2 + 1)−1 by means of partial fractions).13. 49 . 2 (c) dz where γ(t) = 2eit . . (m−1) f (z) = ez − (−1)m+1 e−z . m (z − 1) 2 f (z) = ez − e−z f (z) = ez + e−z f (z) = ez − e−z . z γ (d) γ (e) γ z1/m 1 dz where γ(t) = 1 + eit . m even . 1) ⊂ C. . f (m−1) (0) = ⇐⇒ ⇐⇒ b) Let a = 1/2. 0 ≤ t ≤ 2π. (m−1) f (z) = 0. Evaluate the following integrals: (a) ez − e−z dz where n is a positive integer and γ(t) = eit . 0 ≤ t ≤ 2π.

Now use Corollary 2. 2 f (m−1) 1 (m − 1)! = 2 2πi 1 γ 1 m 2 m dz z− 1 2 1 . n = 0 and f (z) = sin(z).6. 1 2i . use Proposition 2. .2) Thus 1 From (4. 1) ⊂ C. 2) ⊂ C. Let a = 0. z γ γ γ . Let a = 0. r = 2 and f (w) = 1. γ(t) = 2eit .13. z + i 2i 2i So γ dz = 0. r = 1. For the second integral on the right side. γ(t) = 2eit . Clearly f (z) is analytic on C. ¯ 0 ≤ t ≤ 2π.1). Then f (z) = 1 2πi γ f (w) 1 dw ⇐⇒ 1 = w+i 2πi γ 1 dw ⇐⇒ w+i γ 1 dz = 2πi. Clearly f (z) is analytic on C and B(0. use Proposition 2. we obtain 2πi dz = 0 z− . Then f (z) = 1 2πi γ f (w) 1 dw ⇐⇒ 1 = w−i 2πi γ 1 dw ⇐⇒ w−i γ 1 dz = 2πi.2) yields −2Bi = 1 and therefore B = − 2i and A = γ dz 1 = z2 + 1 2i γ dz 1 − z − i 2i γ dz . We also have B(0. z+i For the ﬁrst integral on the right side. z+i where |i| < 2 ⇐⇒ 1 < 2. We also have B(0.6.1) (4. Hence γ dz 1 = z2 + 1 2i γ dz 1 − z − i 2i γ dz 1 1 = 2πi − 2πi = π − π = 0. z−i where | − i − 0| < 2 ⇐⇒ 1 < 2. Now use Corollary 2. m > 1 γ 1 ⇐⇒ ⇐⇒ (m − 1)! 2πi 1 γ z− 1 m 2 c) Using partial fraction decomposition. m = 1 dz = 0 . z2 + 1 ¯ d) Let a = 0. (4.m > 1 1 B Az + Ai + Bz − Bi A ! + = = z2 + 1 z − i z + i z2 + 1 which implies A+B = 0 Ai − Bi = 1. r = 2 and f (w) = 1. ¯ 0 ≤ t ≤ 2π.13. 1) ⊂ C.m = 1 . Clearly f (z) is analytic on C. 2) ⊂ C. f (0) (0) = 0! 2πi sin(z) 1 dz ⇐⇒ 0 = 1 2πi (z − 0) 50 sin(z) dz ⇐⇒ z sin(z) dz = 0. we get A = −B inserted into (4.¯ Clearly f (z) is analytic on C and B( 1 .

. we use Proposition 2.6 with a = 0. m m m ¯ Clearly f (z) is analytic on C and B(1..· m m z1/m dz (z − 1)m γ (1 − im). n = m − 1 and f (z) = z1/m 1 1/m−1 f (z) = z m 1 1 − 1 z1/m−2 f (z) = m m .6. note γ(t) = reit dz = 0. Evaluate 2 < r < ∞. 1 1 1 (m−1) f (z) = − 1 ·.. Now. 2) and γ = 0 + reit with 0 < r < 2. note z−2i is 1 dz = 0. = z z − 2i z + 2i z(z2 + 4) 1 z2 + 1 1 1 1 3 3 1 dz = dz + dz + dz... analytic on B(0. 2) and γ for 0 < r < 2 is in B(0. r = 1/2. . Thus. by Proposition 2. It follows.· − (m − 2) z1/m−(m−1) . for 2 < r < ∞. z − 2i γ 1 1 z2 + 1 π dz = 2πi = i. 2 < r < ∞. 2). Now use Corollary 2. 1 ) ⊂ C. 2 f (m−1) (1) = ⇐⇒ ⇐⇒ ⇐⇒ Exercise 10. z2 +1 γ z(z2 +4) dz where γ(t) = reit . By Proposition 2. and γ(t) = 0 + reit for 51 . Similarly. Now.. Solution. dz = 4 γ z 8 γ z − 2i 8 γ z + 2i z(z2 + 4) γ γ z 1 2πi f (0) = 2πi where f ≡ 1 as f ≡ 1 is analytic on B(0. .15. 2 + 4) 4 2 γ z + 2i γ z(z so γ (t) = ireit . Calculating the integrals in turn.· − (m − 2) · 1 = m m m 2πi γ z1/m dz (z − 1)m γ 2πi 1 1 − m z1/m dz = (z − 1)m (m − 1)! m m z1/m 2πi dz = (z − 1)m (m − 1)!mm−1 m−2 i=0 1 − 2m 1 − (m − 1)m ·.13.e) Let a = 1. 0 < r < 2 and z2 + 1 . for all possible values of r. 1 dz = z 2π 0 γ ireit dt = i reit 2π dt = 2πi. 0 For the second integral. Then f (z) can be decomposed using partial fractions as z(z2 + 4) f (z) = 3/8 3/8 z2 + 1 1/4 + + . 0 ≤ t ≤ 2π. f (w) = 1. It follows for 0 < r < 2. Let f (z) = (m − 1)! 2πi γ (m − 1)! 1 1 1 − 1 ·.

Clearly f (z) is analytic on C and B(0. . E4 . 52 . Find the domain of analyticity of f (z) = 1 1 + iz log .e. What is the radius of convergence of this series? Use the fact that 1 = cos z sec z to show that E2n − 2n 2n − 2 E2n−2 + 2n 2n − 4 E2n−4 + · · · + (−1)n−1 2n 2 E2 + (−1)n = 0. f (2i) = It follows that. Similarly. show that tan f (z) = z (i. . Show that f (z) = (Hint: see Exercise III. E4 . 1 z2 + 1 1 1 3 3 dz = dz + dz + 2 + 4) 4 γ z 8 γ z − 2i 8 z(z 3 3 1 = (2πi) + (2πi) + (2πi) 4 8 8 = 2πi 1 dz z + 2i 1 2πi γ f (w) dw ⇐⇒ w − 2i γ 1 dz = 2πi. Solution. f is a branch of arctan z). Evaluate E2 . Not available. f (−2i) = 1 2πi γ f (w) f (w) 1 dw ⇐⇒ 1 = dw w + 2i 2πi γ w + 2i 1 ⇐⇒ dz = 2πi z + 2i γ if 2 < r < ∞. 2i 1 − iz also.¯ 0 ≤ t ≤ 2π. (2k)! (for some E2 . Exercise 12. Show that sec z = 1 + ∞ k=1 ∞ k=0 (−1)k z2k+1 .19. These numbers are called Euler’s constants. .) Solution. E6 . 2 < r < ∞) ⊂ C. E4 . (E10 = 50521 and E12 = 2702765). It is easily seen that sec z = 1 + ∞ k=1 E2k 2k z . · · · . 2k + 1 for |z| < 1 E2k 2k z (2k)! for some constants E2 . E8 . 3. z − 2i γ γ Exercise 11. Then.)..

. 2 Since 1 = cos z sec z. Calculate B2 . 0. The Bernoulli numbers Bn are deﬁned by Bn = lim We determine the largest R such that z = ez − 1 ∞ k=0 z→∞ dn ez z −1 . Consider f (z) = ∞ k=0 ak k z k! be its power series expansion about zero. E6 = 61 and E8 = 1285.Finally. 2π) and the negative indices disappear because the singularity at the origin is a removable one. The uniqueness of the Laurent expression together with the 53 The function f (z) = ez z has a removable singularity at z = 0. What is the radius of convergence? Show that 0 = a0 + n+1 1 a1 + · · · + n+1 n an . dzn n = 0. 1. Find the series expansion of f (z) = ez z and let −1 about zero and determine its radius of convergence. The numbers 2 B2n = (−1)n−1 a2n are called the Bernoulli numbers for n ≥ 1. r) of 0. Using the fact that f (z) + 1 z is an even function show that ak = 0 for k odd and k > 1. (−1)i−k 1= (2k)!(2i − 2k)! i=0 k=0 (−1)n−k E2k i=0 2n 2k =0 which is equivalent to E2n − 2n 2n − 2 E2n−2 + 2n 2n − 4 ez −1 z E2n−4 + · · · + (−1)n−1 2n 2 E2 + (−1)n = 0. B4 . we can multiply the series for sec z and cos z together to obtain ∞ n 2i E2k z . we have E2 = 1. Solution. Bk k z k! for 0 < |z| < R. Therefore f (z) has a Laurent series expansion on the punctured disk ann(0. The radius of convergence is the closest 1 distance from 0 to the nearest non-analytic point which is precisely π . · · · . 1 if x = 0 . we can compare the coeﬃcients of z2i and multiply by (2i)! to obtain the formula n 1 1 since sec(0) = cos(0) = 1 = 1 and sec z = cos z is an even function. E4 = 5. . Bl0 . k ∈ Z − {0}. 2. because it can be extended to 0 by an −1 analytic function g on a small neighborhood B(0. Exercise 13. . Furthermore the function f has essential singularities exactly at z = 2πk. In addition. setting zz if 0 < |z| < r g(z) := e −1 .

Taking the limits. 2π). yet in other terms. 0. 0. f (z) + z = 2 2(ez − 1) 2(ez − 1) 1 −z(ez + 1) −z(1 + ez ) z(1 + ez ) 1 f (−z) − z = −z = = = f (z) + z. p.8. Consider the equation ez z = ∞ Bk zk for k=0 k! −1 (ez −1)2 z ∈ ann(0. 54 . 2 e −1 2(1 − ez ) 2(ez − 1) 2 ∞ k=2 ∞ k=2 1 With the power series expansion of f and recalling that B1 = − 2 this is the same as saying 1+ or. (n + 1)! m=0 m ∞ n=1 n 0= zn n+1 Bm (n + 1)! m=0 m n as well as t’ for all nonzero z ∈ ann(0. then 1= = ∞ l=1 ∞ n ∞ ∞ ∞ Bk Bk zl zl−1 = l! k=0 k! l=0 (l + 1)! k=0 k! zn−m Bm m z m! (n + 1 − m)! n=0 m=0 ∞ n=0 ∞ n=0 = zn Bm m!(n + 1 − m)! m=0 n n = zn n+1 Bm (n + 1)! m=0 m ∞ n=1 = B0 + Since B0 = 0 conclude that zn n+1 Bm . B1 = limz→0 (1−z)e −1 = − 2 . z= l! k=0 k! l=1 Divide by z 0. First we compute B0 = limz→0 ez z = 1 and note that the result in question does not hold for B0 . This is equivalent to ∞ ∞ zl Bk k z . It is true that 2z + (ez − 1)z z(ez + 1) 1 = . Bk (−z)k = 1 + k! Bk k (z) k! Bk = −Bk = 0 for k odd . 1 This prove the ﬁrst part of the statement. 2π). one only needs to show that f (z) + 2 z is an even function. Also −1 z 1 compute B1 . This can only happen if the inner summation equals zero for all n ∈ N. 72) give the desired result.formula for the coeﬃcients in the expansion (Theorem 2. For the second claim. k > 1.

We obtain f (2iz) = cos(z) 2iz(e2iz + 1) iz(eiz + e−iz ) = =z = z cot(z). we replace z by 2iz of the power series expansion for f (z) to obtain z cot(z) = 1 − Multiplying the formula cot 2z = 1 2 4n B2n 2n z . expressing the coeﬃcients in terms of Bernoulli numbers. Also B3 = B5 = B7 = 0 by 2 the second statement of part b). To ﬁnd the power series expansion of tan z about z = 0. 4n B2n 2n z − 1 − (2n)! tan(z) = ∞ n=1 ∞ n=1 42n B2n 2n z = (2n)! ∞ n=1 4n (4n − 1)B2n 2n z . (Hint: use Exercise 13 and the formula cot 2z = 1 cot z − 1 tan z. Find the power series expansion of tan z about z = 0. k n+1 k=0 This gives 1 3 1 1 B2 = − (B0 + 3B1 ) = − 1 − = . 9 30 Exercise 14. It remains to compute B2 .In part b) the values of B0 = 1 and B1 = − 1 have been computed already. (2n)! 4n (4n − 1)B2n 2n−1 z . B4 . B6 and B8 . (2n)! 55 .) 2 2 Solution. 5 30 1 1 B6 = − (B0 + 7B1 + 21B2 + 35B4 ) = . The formula involving binomial coeﬃcients in the part above can be solved for Bn and n−1 1 n+1 Bn = − Bk . we replace z by 2iz in the function f (z) = z(ez + 1) 2(ez − 1) which we obtained in the previous problem. 3 3 2 6 1 1 B4 = − (B0 + 5B1 + 10B2 ) = − . iz − e−iz 2iz − 1) e sin(z) 2(e ∞ n=1 Next. 7 42 and 1 1 B8 = − (B0 + 9B1 + 36B2 + 126B4 + 84B6 ) = − . (2n)! cot z − 1 tan z with 2z gives 2 z tan z = z cot z − 2z cot 2z and therefore z tan z = 1 − ∞ n=1 Hence.

Solution. Show that f is a polynomial of degree ≤ n. Hence. Not available. Let G be a region and suppose that f : G → C is analytic and a ∈ G such that | f (a)| ≤ | f (z)| for all z in G. R ⊆ H.73 k! 2πi γ f (z) k! dz ≤ 2π zk+1 γ | f (z)| |dz|. |γ(t)| = |reit | = r > R.13 p. f (x) = e x . we get k! 2π γ | f (z)| |dz| |z|k+1 ≤ = |z|=r = k! k!M M|z|n |dz| = |z|n−k−1 |dz| 2π γ |z|k+1 2π γ k!M n−k−1 k!M n−k−1 r r 2πr |dz| = 2π 2π γ 1 k!Mrn−k = k!M k−n → 0 r as r → ∞ since k ≥ n + 1 ⇐⇒ k − n ≥ 1. Not available. Hence. | f (k) (0)| = 0 ∀k ≥ n + 1 and thus n f (z) = k=0 f (k) (0) k z. f (z) ≡ g(z). 56 . Clearly 0 is a limit point of H and 0 ∈ G. Solution. Let r > R and let γ(t) = reit . Exercise 2. So. We have to show that f (k) (0) = 0 for k ≥ n + 1. So | f (z)| ≤ M|z|n . Exercise 4. Note that γ is a closed and rectiﬁable curve. |z|k+1 Since r > R. that f has a power series expansion ∞ f (k) (0) ak zk . Solution. Then | f (k) (0)| = Corollary 2. Then for all x ∈ R. Now let g be an entire function such that g(x) = e x for all x ∈ R. Prove that ez+a = ez ea by applying Corollary 3. an R > 0. k! a polynomial of degree less or equal n.3 Zeros of an analytic function Exercise 1.8. Now. Solution.4. Let f (z) = ez . Find all entire functions f such that f (x) = e x for x in R. Show that either f (a) = 0 or f is constant. by Corollary 3.7. Exercise 6. Let f be an entire function and suppose there is a constant M. we have by Proposition 3. Exercise 3. 0 ≤ t ≤ 2π. So f (z) = ez is the set of all entire functions f such that f (x) = e x for all x ∈ R. and an integer n ≥ 1 such that | f (z)| ≤ M|z|n for |z| > R. Solution. Prove that cos(a + b) = cos a cos b − sin a sin b by applying Corollary 3. set H = {z ∈ G : f (z) = g(z)}. Exercise 5. Since f is an entire function. Not available.3 p. that is f ∈ A(C). z ∈ {γ}.8. whereak = f (z) = k! k=0 with inﬁnite radius of convergence. then we are done.8. 77. Give an example to show that G must be assumed to be connected in Theorem 3. 0 ≤ t ≤ 2π. Clearly f is analytic on C since it is entire.

Show that either f ≡ 0 or g ≡ 0. then (since f is continuous. . R). 0 ∀z ∈ G. Thus. then by the same reasoning. we conclude: Either f (a) = 0 or f is constant. (def) ∀z ∈ C. R). Deﬁne 1 . Let U : C → R be a harmonic function such that U(z) ≥ 0 for all z in C. Not available. For all z ∈ C. Re( f (z)) ≥ 0 ∀z ∈ C by assumption. So g(z) = f (z) − c ≡ 0 on G. | f (z)| (4. Similar. then f (z) =same constant for all z ∈ G.11 p. R) ⊂ G. f (a) satisﬁes (4. Let G be a region and let f and g be analytic functions on G such that f (z)g(z) = 0 for all z in G. ∀z ∈ G.Solution. 2. (4. Hence g(z) = 0 ∀z ∈ B(a. Let f : G → C be analytic. 57 . There exists an analytic function f : C → C such that U(z) = Re( f (z)). R) to fulﬁll f (z)g(z) = 0 ∀z ∈ B(a. too.. g(n) (a) = 0 for all n = 0. then g(z) ∈ A(G). ∀z ∈ G. . we obtain f ≡ 0. Proof: Suppose f (z) = c in B(a. g ∈ A(G) such that f (z)g(z) = 0. It follows that U is constant. Let G be a region and let f. Exercise 7. R). . and therefore g is constant according to Liouville’s Theorem 3. 77 and hence f (z) is constant.3) Obviously. deﬁne g(z) = e− f (z) ∈ A(C). According to the Maximum Modulus Theorem 3. Deﬁne g(z) = f (z) − c. Solution. Thus. prove that U is constant. 1. Exercise 8. where G is a region.3) | f (a)| (def) ∀z ∈ G. Hence. since U : C → R is assumed to be harmonic. assume g(z) f ≡ 0 or g ≡ 0. Assume f (z) 0. g ∈ A(G) (since f (a) 0 ∀a ∈ G) and |g(z)| = (def) 1 1 ≤ = |g(a)|. Now. Give an elementary proof of the Maximum Modulus Theorem for polynomials. Thus. 79.4 p. Let a ∈ G such that | f (a)| ≤ | f (z)| ∀z ∈ G. where the last step follows since Re( f (z)) ≥ 0 by assumption. Solution. g(z) = f (z) Clearly. we conclude: Either Exercise 9. because f ∈ A(G)) there is an R > 0 such that f (z) 0 whenever z ∈ B(a. It is easy to verify the following claim: If f (z) =constant in B(a. g is constant and thus f is constant. So g ≡ 0. suppose f (a) 0 ∀a ∈ G. Solution. Then |g(z)| = |e− f (z) | = eRe(− f (z)) = e−Re( f (z)) ≤ 1. Let a ∈ G such that f (a) 0. g(z) is bounded and an entire function.3) since 0 ≤ | f (z)| ∀z ∈ G is true. By the claim above we obtain that g(z) = 0. ∀z ∈ G.

77) where a1 . Exercise 2. p(z) Solution. an are the zeros of p(z). a2 )2πi + . + γ (z − an )−1 dz n = n(γ. Not available. . p(z) Exercise 4. Not available. Thus p (z) dz = p(z) = γ c γ n k=1 n l=1 (z l k − al ) γ c n k=1 (z 1 1 1 + + .3. 81 γ(t) = Reit . . . γ p (z) dz = 2πin. 0 ≤ t ≤ 1 is a closed and rectiﬁable curve and a1 . Show that if f and g are analytic functions on a region G such that f¯g is analytic then either f is constant or g ≡ 0.. we can write n p(z) = c k=1 (z − ak ) for some constant c (see Corollary 3. Solution. Let p(z) be a polynomial of degree n and let R > 0 be suﬃciently large so that p never vanishes in {z : |z| ≥ R}. .4 The index of a closed curve Exercise 1. Solution. ai ) = 1 ∀1 ≤ i ≤ n. In addition. . . an {γ} since R > 0 suﬃciently large so that p never vanishes in {z : |z| ≥ R}.6 p. Since p(z) is a polynomial of degree n. Solution. . we have n(γ. 4. 58 . a) = k. . show that γ p (z) dz = 2πin. . Exercise 3. Prove Proposition 4. . 0 ≤ t ≤ 2π or γ(t) = Re2πit .2 p. Solution. Show that there is an integer k such that γ z−1 dz = log r + iθ + 2πik. a1 )2πi + n(γ. Hence. since by Deﬁnition 4.Exercise 10. + n(γ. Fix w = reiθ 0 and let γ be a rectiﬁable path in C − {0} from 1 to w. Give an example of a closed rectiﬁable curve γ in C such that for any integer k there is a point a {γ} with n(γ. If γ(t) = Reit .. By the product rule we obtain n n n n p (z) = c k=1 (z − ak ) l=1 l k (z − al ) = c k=1 l=1 l k (z − al ). . Not available. 0 ≤ t ≤ 2π. . an )2πi = k=1 2πi = 2πin. Not available. + dz z − a1 z − a2 z − an (z − a1 )−1 dz + γ − ak ) dz = γ (z − a2 )−1 dz + .

2πi γ z − a γ Exercise 5. where a {γ}. γ2 .82.4. γ2 .9 p. R) by Theorem 4.4 p. Suppose f : G → C is analytic and deﬁne ϕ : G × G → C by ϕ(z. w) = 0. G = B(0. g(z) = z−a f (a). Not available.86 if we let G = B(a. w) = 0 ∀w ∈ C\B(a. Let γ be a closed rectiﬁable curve in C and a {γ}. Assume the conditions of Cauchy’s Theorem. {γ} is bounded. Since γ is a closed rectiﬁable curve in C. a) f (a). w) = 0 for all w in C − G. Show that for n ≥ 2 γ (z − a)−n dz = 0. γ3 be curves whose traces are |z−1| = 1. Not available. Solution. . z a . f (z) − f (a) dz = 0 z−a f (z) f (a) dz − dz = 0 γ z−a γ z−a f (z) 1 dz = f (a) dz γ z−a γ z−a f (z) dz = f (a)n(γ. that is: Let G be an open subset of the plane and f : G → C an analytic function. R). Deﬁne f (z)− f (a) . 3)−(B+ ∪ B− ). and |z| = 2. let a ∈ G. Further γ is a closed and rectiﬁable curve in G such that n(γ. w)+n(γ2 .6. Solution. R) the open set and f : G → C given by f (z) = 1 is analytic. w)+n(γ3 . g ∈ A(G).7 c) ⇒ b) p. Exercise 2. a)2πi γ z−a 1 f (z) dz = n(γ. Let B± = B(±1. w) is analytic. Now. R). Hence C\B(a. 59 ∀w ∈ C\G.2 p. 4. z) = f (z). (Check the power series of f (z) like in the proof of Theorem 3. z → ϕ(z.81 ⇒ ⇒ This is exactly Cauchy’s Integral Formula. and γ3 orientations such that n(γ1 . then γ f = 0. Let γ1 . 2 |z+1| = 1. Therefore n(γ. R) belongs to the unbounded component of B(a. Solution. Solution. Show that the Integral Formula follows from Cauchy’s Theorem. 1 ). Solution. we obtain g=0 γ a γ ⇒ ⇒ ⇒ Def. w) = [ f (z) − f (w)](z − w)−1 if z w and ϕ(z.5 Cauchy’s Theorem and Integral Formula Exercise 1. 78) By Cauchy’s Theorem. z=a Clearly. Give γ1 . w) = 0 ∀w ∈ G − C. Give the details of the proof of Theorem 5. respectively. Not available. If γ is a closed rectiﬁable curve in G such that n(γ. Prove that ϕ is continuous and for each ﬁxed w. ¯ Exercise 3. So there exists R > 0 such that {γ} ⊂ B(a. We can apply Corollary 5. Exercise 4.

then for a = 1 in G\{γ} f (n−1) (1)n(γ. 82. that n(T .14 p. Let f be analytic on D = B(0. Let G be a region and suppose fn : G → C is analytic for each n ≥ 1. Solution. Solution. 85 with m = 1. Find γ 1!· 1 = 1. Let γ(t) = 1 + eit for 0 ≤ t ≤ 2π.4) . w) = 0 ∀w ∈ C\G and γ is a closed rectiﬁable curve in G. We also have. Let G = B(1. 60 (4. Hence γ (z − a)−n dz = 0. R = 1. 82). Then | f (0)| ≤ So | f (0)| ≤ 1. 1) and | f (z)| ≤ 1 for all z in B(0. R)\{γ} 1 (n − 1)! dz 2πi (z − a)n γ (n − 1)! 1 ⇒ 0· n(γ. We can directly apply Cauchy’s estimate 2. we can apply Corollary 5. we have fn = 0. for n ≥ 2.4 p. Solution. a) = dz 2πi (z − a)n γ f (n−1) (a)n(γ. Show | f (0)| ≤ 1. Exercise 7. Then T is closed and rectiﬁable. z−1 zn (n − 1)! (n − 1)! dz = 2πi (z − 1)n 2πi γ n z dz z−1 γ z z−1 n dz since f (n−1) (1) = n! and n(γ. Let G be a region and suppose fn : G → C is analytic for each n ≥ 1. that us fn is continuous for each n ≥ 1.4 p.9 with G and f deﬁned above such that n(γ. w) = 0 ∀w ∈ C\G. M = 1 and N = 1. 73 with a = 0. since w is an element of the unbounded component of G (Theorem 4. Suppose that { fn } converges uniformly to a function f : G → C Show that f is analytic.then for a in B(a. R) (open) where 1 < R < ∞ and let f : G → C be given by f (z) = zn which is analytic on G. a) = since f (n−1) (a) = 0 if n ≥ 2. Let T be an arbitrary triangular path in G. w) = 0 ∀w ∈ C\G since C\G belongs to the unbounded component of G and γ is a closed and rectiﬁable curve in C. By Cauchy’s Theorem (First Version). 1) and suppose | f (z)| ≤ 1 for |z| < 1. a) = ⇒ n! = ⇒ γ (n − 1)! 2πi γ z n dz = 2πin. 11 z n z−1 dz for all positive integers n. By Theorem 4. Exercise 6. p. n(γ. Now. 1) (since |z| < 1). that is by assumption f is analytic on B(0. a) = 1. T ∀n ∈ N. Exercise 8.

since we suppose fn : G → C is analytic for each n ≥ 1 and T is closed and rectiﬁable. (This exercise was taken from a paper by C. 4. Summary: Since T f = 0 for every triangular path in C and f is continuous. Monthly. it suﬃce to show that for every triangular path T in C. By Cauchy’s Theorem T f = 0 over the translated triangle and let → 0 since f is continuous). Solution. we get f = lim T n→∞ f = lim 0 = 0.) 61 . but whose limit is the given T . 29). σ2 : [0. Let {T n } be a sequence of triangles that are not intersecting I. since we can ﬁnd an open neighborhood G containing T (is closed and rectiﬁable) such that f ∈ A(G) (by assumption) and n(T . We also have f = lim T n→∞ T fn = lim 0 = 0 (4. 390–391).4) n→∞ The ﬁrst equality follows by Lemma 2. Finally. There are several cases to consider: Case 1: A triangle does not intersect I. Case 5: A triangle T contains parts of I (also here a sketch might help). Show that if f : C → C is a continuous function such that f is analytic oﬀ [−1. Then. 1] then f is an entire function.4 p. T f = 0 again. w) = 0 ∀w ∈ C\G (by Theorem 4. Use Cauchy’s Integral Formula to prove the Cayley–Hamilton Theorem: If A is an n × n matrix over C and f (z) = det(z − A) is the characteristic polynomial of A then f (A) = 0. (Procedure: Translate the triangle by ± i depending if it lies above or below the x-axis. Math. σ2 (t) ≡ b. that is f is an entire function. Not available. that T f = 0. then γ ∼ σ2 . we can apply Morera’s Theorem 5. Show that if γ is closed rectiﬁable curve in G and γ ∼ σ1 . 71 since T is closed and rectiﬁable and since each { fn } ∈ A(G). we get f is continuous (Theorem 6. Again apply Cauchy’s Theorem to obtain T f = 0. (Hint: connect a and b by a curve. we can decompose the triangle into 5 parts. 82). A.7 p. Like in case 2 we can translate the triangle by ± We can also argue this way: Let G be an open neighborhood containing the triangle T . we have fn is continuous for each fn and by the uniform convergence to a function f .7 and the second one by case 1. McCarthy. Solution. In this instance. Case 2: A triangle touches I exactly at one point P. Exercise 10. since we already assume f is continuous. we get by Morera’s Theorem: f ∈ A(C). we have T f = 0. since f is continuous. Let G be a region and let σ1 . 1] → G be the constant curves σ1 (t) ≡ a. 82 (1975). we obtain by Cauchy’s Theorem. Exercise 9.10 p. Thus. By Morera’s Theorem. Amer. Two triangles are of the kind explained in case 3 and one is of the kind explained in case 2 (a sketch might help). This single point of intersection P is a removable singularity. Then by the continuity of f . In this case. we can always decompose the triangle T in three parts. T f = 0. Hence. Case 4: A triangle T is cut into 2 parts by I. 86 to obtain that f is analytic in G.6 The homotopic version of Cauchy’s Theorem and simple connectivity Exercise 1. Tn n→∞ where the ﬁrst equality follows by Lemma 2. Two triangles are of the kind explained in case 3 and the other three triangles are of the kind explained in case 2. since f is continuous and γ f = 0 for every triangular path T in G.1 p. Case 3: One edge of the triangle touches I.

−iπ) = 0 − 1 = −1. t) = tb + (1 − t)a = Γ(1. If we can show that σ1 ∼ σ2 . 81 γ 1 1 1 dz − dz z−i 2i γ z + i γ 1 1 1 1 π dz − π dz 2πi γ z − i 2i γ z + i π (n(γ. = since n(γ. now we show σ1 ∼ σ2 . Exercise 4. then we get γ ∼ σ2 .1 then the curve γ0 (t) = e2πit . Using partial fraction decomposition yields dz z2 + π2 = = Def 4. for 0≤s≤1 So. Evaluate the integral dz γ z2 +1 where γ(θ) = 2| cos 2θ|eiθ for 0 ≤ θ ≤ 2π. 0) = σ1 (s) = a and σ1 ∼ σ2 ⇒ γ ∼ σ2 . 81 γ 1 1 1 1 dz − dz 2πi γ z − iπ 2πi γ z + iπ n(γ. 3. Let Γ(s. Let G = C − {0} and show that every closed curve in G is homotopic to a closed curve whose trace is contained in {z : |z| = 1}.11 gives an equivalence relation on C. i) = n(γ. t) for all t” from Deﬁnition6. 1) = σ2 (s) = b. (it is constant with respect to s and a line with respect to t). In addition. So σ1 ∼ σ2 . t) = tb + (1 − t)a.2 p. 1] → G such that and Γ(s. Hence γ dz = 0. Solution. and Γ(0. −i) since i and −i are contained in the region generated by γ. A sketch reveals that the zero −πi is inside the region and the zero πi is outside the region. Let C = all rectiﬁable curves in G joining a to b and show that Deﬁnition 6. we surely have that σ1 and σ2 are closed and rectiﬁable curves in G. is homotopic to the constant curve γ1 (t) ≡ 1 in the region G = C − {0}.Solution. Show that if we remove the requirement “Γ(0. that is: γ ∼ σ1 Γ(s. 0 ≤ t ≤ 1. t) ∀0 ≤ t ≤ 1. we have to ﬁnd a continuous function Γ : [0. −i)) 1 2i π· 0 = 0. Exercise 3. Evaluate Solution.2 p. Assume γ is a closed and rectiﬁable curve in G and γ ∼ σ1 . Using partial fraction decompositions gives dz z2 + 1 = = = Def 4. Clearly Γ is a continuous function. γ z2 +π2 Exercise 6. t) for 0 ≤ t ≤ 1. 1] × [0. Solution. Solution. i) − n(γ. Not available. Exercise 2. 1) = b ∀0 ≤ s ≤ 1 and Γ(0. that is. 0) = a ∀0 ≤ s ≤ 1 and Γ(s. t) = Γ(1. A sketch shows that the two zeros of z2 + 1 (they are ±i) are inside of the closed and rectiﬁable curve γ (The region looks like a clover with four leaves ). Not available. since ∼ is an equivalence relation. Let γ(θ) = θeiθ for 0 ≤ θ ≤ 2π and γ(θ) = 4π − θ for 2π ≤ θ ≤ 4π. 62 . Since σ1 (t) ≡ a and σ2 (t) ≡ b. Solution. Exercise 5. Not available. t) = Γ(1. it satisﬁes Γ(s. iπ) − n(γ. z2 + 1 dz .

Solution. Exercise 8. Suppose γ0 ∼ γ1 . a b. 0]. Let γ be the path depicted on p. two integrals will be zero and we will have another 2 pair of non-closed paths. 2 + π2 γ z i 1 Exercise 7. Let G be a region and let γ0 and γ1 be two closed smooth curves in G. They also meet at the rightmost crossing point. and Γ satisﬁes (6. b}. iπ) = 0 and −iπ is contained in the region. z) = 0 for all z not in G without γ being homotopic to zero. 1] → Z is continuous. Exercise 9. Also suppose that γt (s) = Γ(s. Can you prove it?) Notice that this example shows that it is possible to have a closed curve γ in a region such that n(γ. If we integrate over the path around a 1 is equivalent to evaluate γ −γ z−a dz where γ1 (t) = a + reit .9 p. 2+2i. and let γ be the curve in the ﬁgure below.10 is false. The ﬁrst pair begins at the leftmost crossing pair and goes around a in opposite direction. Hence.2). Evaluate dz where γ is one of the curves depicted below. dz . 2 2 Find γ f . If w ∈ C − G deﬁne h(t) = n(γt . 0 ≤ t ≤ π and γ2 (t) = a + reit . (a) Show that n(γ. π ≤ t ≤ 2π for 1 2 some r > 0. dz = −1.since iπ is not contained in the region generated by γ. −iπ) = 1. so n(γ. i) − 2πin(γ. (Notice that the word is “convince” and not “prove”. Exercise 10. (Justify your answer. t) is smooth for each t. (b) Convince yourself that γ is not homotopic to zero. 86 we have ∀a ∈ G − {γ} γ 1 f (z) dz = 2πi f (k) (a)n(γ. so n(γ. The second pair begins at the middle crossing pair and goes around b in opposite direction. 2i. k+1 k! (z − a) 63 . a) = n(γ. Not available. We can write it as a sum of 6 paths. Therefore. w) and show that h : [0. Solution. Let G = C − {a. 2. Solution. Two of them will be closed and have a and b in their unbounded component. It is easily seen that we have 1 dz = z−a 1 dz = πi z−a γ1 γ2 and therefore n(γ. b) = 0. 96. Using partial fraction decomposition and the deﬁnition of the winding number we obtain 1 dz = 1 + z2 2i 1 dz − z−1 γ e −e z4 z −z γ γ γ 1 1 dz = (2πin(γ. Let f (z) = [(z− 2 −i)· (z−1− 3 i)· (z−1− 2 )· (z− 3 −i)]−1 and let γ be the polygon [0. Let γ be a closed rectiﬁable curve in C not passing through ±i. Not available. z+i 2i Exercise 11. −i)) = π (n(γ. γ 1+z2 where γ is any closed rectiﬁable curve in C not passing Solution. Using Corollary 5. the converse to Corollary 6. b) = 0. That is. −i)) . a) = 0 and similarly we obtain n(γ.) Solution. a) . i) − n(γ. Find all possible values of through ±i.

(m − (n − k))! 64 n = 1. Exercise 2. b) and c). . w) = 0 for all w ∈ C − G and f : G → C is analytic. we obtain n g(a) 0 (p. there exists g ∈ A(B(0. . Not available. we can write f (z) = = ∞ k=0 ∞ k=0 0. We obtain the same result for part c). ∂G) < 1 r} ⊂ H. Set r = d({γ}. Let f be analytic in B(a. .7 Counting zeros. . Solution. Not available. R)). R)). We have f (3) (0) = e0 + e−0 = 2 and using the formula above for k = 3 we obtain 1 2πi ez − e−z dz = 2πi· 2n(γ. By deﬁnition. . In our case a = 0 and γ satisﬁes the above assumptions for exercise a). 2.where γ is a closed rectiﬁable curve in G (open) such that n(γ. Show that if f : G → C is analytic and γ is a rectiﬁable curve in G then f ◦ γ is also a rectiﬁable curve. (b) Use part (a) 2 to show that if f : G → C is analytic then f (z) = α has at most a ﬁnite number of solutions z such that n(γ. f (n) (z) = k=0 n k m! (z − a)m−(n−k) g(k) (z). Let f (z) = ez − e−z which is clearly analytic on any region G. R)) such that f (z) = (z − a)m g(z). Let G be open and suppose that γ is a closed rectiﬁable curve in G such that γ ≈ 0. Hence. the Open Mapping Theorem Exercise 1. 97). z) 0. a is a zero of multiplicity m. Then after a lengthy calculation. Solution. . We have n(γ. m. 0) = 1 for part a) and thus the result is is 4πi . Solution. . . m − 1 (we also have f (a) = 0) and f (m) (a) f ∈ A(B(a. 0) = n(γ. . f (m) 0 where g ∈ A(B(a. R) and suppose that f (a) = 0. (a) Show that {z : d(z. = f (a) = 0 and f (a) (a) 0. . . Since ak (z − a)k where ∞ k=m ak = f (k) (a) and |z − a| < R k! f (k) (a) (z − a)k = k! ∞ k=0 f (k) (a) (z − a)k k! = since (z − a)m g(z) = (z − a)m f (k) (a) (z − a)k−m k! g(z) g(a) 0. ⇐: Let f (n) (a) = 0 for all n = 1. ⇒: Suppose a is a zero of multiplicity m ≥ 1 of f . z) = 0}. 6 3 z4 γ Hence. Exercise 3. 2. Show that a is a zero of multiplicity m iﬀ f (m−1) (a) = . ∂G) and H = {z ∈ C : n(γ. 0). 3 2πi 3 . 0) = 1 for part b) and thus the result 4. n(γ.

8 Goursat’s Theorem No exercises are assigned in this section.4 p. ) ⊂ G such that f (z1 ) = ξ = f (z2 ) contradicting that f is 1 − 1. f (n) (a) = 0 for n = 1. Hence. If k = 0 and m = n. is one-one and onto. . Then. 2. Assume f (a) = 0 for some a ∈ G. Solution. ). Not available. then f (m) (a) = f (n) (a) = m!g(a) 0 since g(a) 0 by assumption. Then f (a) = α where α is a constant. m − 1 and f (n) (a) 0 for n = m. Exercise 7. Solution. Let P : C → R be deﬁned by P(z) = Re z. So f (z) 0 for any z in G. we can ﬁnd at least two distinct points z1 . ).If z = a. Hence. 65 . the equation f (z) = ξ has exactly m simple roots in B(a. Exercise 6. Since f (z) = ξ has exactly m (our case m ≥ 2) simple roots in B(a. Exercise 5. Exercise 4. z2 ∈ B(a. Show that f is an open map iﬀ f is a closed map. 98 to argue that there is an > 0 and δ > 0 such that for 0 < |ξ − a| < δ. show that P is an open map but is not a closed map. we can use Theorem 7. Then F(a) = f (a) − α = α − α = 0 and F (a) = f (a) = 0 by assumption.) Solution. Suppose that f : G → C is analytic and one-one.2 to give another proof of the Fundamental Theorem of Algebra.) Solution. Not available. . then clearly f (n) (a) = 0 for m − (n − k) > 0 and f (n) (a) 0 for m − (n − k) = 0. Not available. Use Theorem 7. (A function f is a closed map if it takes closed sets onto closed sets. (Hint: Consider the set F = {z : Im z = (Re z)−1 and Re z 0}. F has a as a root of multiplicity m ≥ 2. Let X and Ω be metric spaces and suppose that f : X → Ω. Also f (z) − α has a zero of order m ≥ 2 at z = a. . R) and α = f (a). . since f is analytic in B(a. show that f (z) 0 for any z in G. Deﬁne F(z) = f (z) − α. 4.

f (z) = z b) cos z . if it is a pole ﬁnd the singular part. if it is a removable singularity deﬁne f (0) so that f is analytic at z = 0.Chapter 5 Singularities 5. e) f (z) = f) f (z) = g) f (z) = h) i) 1 f (z) = z sin . f (z) = z c) cos z − 1 f (z) = .1 Classiﬁcation of singularities Exercise 1. z−1 log(z + 1) . Determine its nature. a) sin z . z(z − 1) cos(z−1 ) . z z2 + 1 . Each of the following functions f has an isolated singularity at z = 0. z d) f (z) = exp(z−1 ). z2 f (z) = (1 − ez )−1 . if it is an essential singularity determine f ({z : 0 < |z| < δ}) for arbitrarily small values of δ. 66 .

38) ∞ z2k (−1)k cos(z) = (2k)! k=0 and therefore cos(z) = z cos z z ∞ k=0 = 1. since lim (z − 0) sin(z) sin(z) = lim z = lim sin(z) = 0.j) 1 f (z) = zn sin . since lim (z − 0) cos(z) − 1 cos(z) − 1 = lim z = lim cos(z) − 1 = 1 − 1 = 0. Now partial fraction decomposition of z(z−1) yields z −z 1 2 z+1 =− + z(z − 1) z z−1 so f (z) = 1 − 1 2 + z z−1 67 . z=0 sin(z) z z→∞ If we deﬁne.103. z = 0 is a removable singularity. A simple computation using L’Hospital’s rule yields limz→0 b) We have (see p. z→∞ z→∞ z z cos(z)−1 . z→∞ z→∞ z z sin(z) . z = 0 is a removable singularity. then f is analytic at z = 0. we simplify z 2+1 = 1 + z(z−1) . z 0 .2 p. f (z) = z 1. f (z) = z 1. (2k)! z Thus has a simple pole at z = 0. z Solution. (−1)k+1 zk = k (−1)k+1 zk−2 1 1 = − + k z 2 (−1)k+1 zk−2 k and hence f (z) = log(1+z) has a pole of order 1 at z = 0 and the singular part is 1 (by Equation 1. c) According to Theorem 1.8 p.2 p.103. a) According to Theorem 1. A simple computation using L’Hospital’s rule yields limz→0 e) We know ∞ zk (−1)k+1 . 2 z+1 z+1 g) First. z=0 cos(z)−1 z z→∞ If we deﬁne. 105). log(1 + z) = k k=1 So we can write 1 log(1 + z) = 2 z2 z ∞ k=1 ∞ k=1 ∞ k=3 = 0. z 0 .7 and z z2 Deﬁnition 1. then f is analytic at z = 0. (−1)k 1 z2k−1 = + analytic part.

(b) ann(0. (0 < |z| < 1).and hence f (z) = z 2+1 has poles of order 1 at z = 0 and z = 1 and the singular part is z −z 1. 300). a) We have 0 < |z| < 1 and using partial fraction decomposition yields f (z) = = = = 1 1 1 −1 1 1 = = + z(z − 1)(z − 2) z (z − 1)(z − 2) z z − 1 z − 2 ∞ ∞ 1 1 z n 1 1 1 zn − 1 − = z z 1 − z 2 1 − 2 |z|<1and |z/2|<1 z n=0 2 n=0 2 ∞ ∞ ∞ n+1 n+2 1 n 1 − 1 z n−1 n−1 z = z − 2 2 n=−1 n=0 n=0 ∞ n=−1 1 − 2−n−2 zn . ∞). . 0. b) We have 1 < |z| < 2 and thus 1 f (z) = z(z − 1)(z − 2) 1 1 1 1 1 − − = z z 1− 1 21− z = 1 − z n=0 −2 n=−∞ ∞ n+2 − ∞ n=0 1 2 1 1 − = z z |1/z|<1and |z/2|<1 z 2 n+1 ∞ n=0 ∞ n=0 1 z 1 2 n 1 − 2 n+1 ∞ n=0 z 2 zn−1 = ∞ n n=0 −z−n−2 − zn−1 = −zn − ∞ n=−1 1 2 n+2 zn . δ)] is either C or C\{a} for one a ∈ C by the Great Picard Theorem (p. Let f (z) = z(z−1)(z−2) . 1). Not available.16). give the Laurent Expansion of f (z) in each of the following annuli: (a) ann(0.17) from (1. (z2 +z+1)(z−1)2 Solution. Solution. 105). Exercise 2. δ)] = C by the Casorati-Weierstrass Theorem p. f ({z : 0 < |z| < δ}) = f [ann(0. δ)] = C. 0. 2). Exercise 3. 1 Exercise 4. In this case. Give the partial fraction expansion of r(z) = Solution. i) We know (see p. 0. Give the details of the derivation of (1. Not available. z2 +1 . 2. 68 (1 < |z| < 2). 1. 109.8 p. 38) ∞ z2k+1 (−1)k sin(z) = (2k + 1)! k=0 so 1 z = f (z) = z sin = z (−1)k z (2k + 1)! k=0 1 z ∞ 1 2k+1 ∞ k=0 2 2 z−1 − 1 z (by Equation (−1) k 1 2k z (2k + 1)! =1− 1 1 1 + − + .7 and Deﬁnition 1. 3!z2 5!z4 7!z6 Therefore f (z) = z sin has an essential singularity at z = 0 (it is neither a pole nor a removable singularity by Corollary 1. But f ({z : 0 < |z| < δ}) = f [ann(0.18c). (c) ann(0. 0.. We have f ({z : 0 < |z| < δ}) = f [ann(0..

19) or (1.2 p. Solution. 0. π 2 Exercise 5. Exercise 6. Solution.2)) and its extended value is also zero by (5. Let f have an isolated singularity at z = a and suppose f (z) 0.c) We have 2 < |z| < ∞ and thus f (z) = = 1 z(z − 1)(z − 2) 1 1 1 1 1 − + z z 1− 1 z 1− z ∞ n=0 2 z = − 1 z n+2 + ∞ 2n n=0 1 z = |1/z|<1and |2/z|<1 n+2 ∞ n=0 = (2n − 1)z−n−2 1 1 − z z ∞ n=0 1 z n 1 + z ∞ n=0 2 z n = −2 n=−∞ (2−n−2 − 1)zn . there exist an integer n > 0 such that lim(z − a)n f (z) = 0 z→a (5.1) and lim(z − a)n+1 f (z) = 0 z→a (5. Hence. Deﬁne zn = π + nπ. that is limz→a (z − a) s f (z) = 0 for some s in R. lim |z − a| | f (z)| = ∞. z→a s 0.2) (In fact we are free to choose any positive integer greater than s). Show that f (z) = tan z is analytic in C except for simple poles at z = n. for each integer sin(z) Solution. Hence s>n−k s<n−k s = n − k. We know tan(z) = cos(z) and (cos(z)) = − sin(z). Thus (z − a)n f (z) has a zero of ﬁnite order k at z − a. Show that if either (1. Then it holds obviously for any greater s’s. 103) applied to (5. tan(z) has a simple 2 1 pole at each zn with residue −1 and with singular part − z−zn at each zn . Determine the singular part of f at each of these poles. Hence. z→a z→a z→a h(a) 0. s>m s<m 69 . Thus. s s−n n s−n+k lim(z − a) f (z) = lim(z − a) (z − a) f (z) = lim(z − a) h(z) = ±∞. Let limz→a |z − a| s | f (z)| = 0 hold. So (z − a)n f (z) = (z − a)k h(z) where h(z) is analytic at a and h(a) 0. + nπ.20) holds if s < m. Not available. (z − a)n f (z) thus has a removable singularity at z = a (by Theorem 1.19) holds if s > m and (1.2). Now.20) holds for some s in R then there is an integer m such that (1. Exercise 7. If f : G → C is analytic except for poles show that the poles of f cannot have a limit point in G. (2 < |z| < ∞).

z→a Recall limz→a (z − a) s f (z) = 0 if s > −m so it is true of s = 0 and s = 1.2 p.4 p. Hence s>n+l s<n+l s = n + l. where m = n + l. 103) with the extended value equal to f (a) = limz→a f (z) 0. s > m s lim |z − a| | f (z)| = ∞. z→a s s>0 .105 and (5. lim |z − a| s | f (z)| = ∞. (b) m < 0 iﬀ z = a is a removable singularity and f has a zero at z = a of order −m. Solution. s = 0 = m .4) But m = 0. f (z) has a removable singularity (by Theorem 1.1)) and (z − a)n f (z) has a pole of ﬁnite order l at z = a and so (z − a)n f (z) = (z − a)−l k(z) where k(z) is analytic at a and k(a) 0 (Proposition 1. and m be as in Exercise 7. Then limz→a (z − a) f (z) = 0 by Theorem 1. z→a (5. f (z) has a removable singularity at z = a and also the extended function has the value limz→a f (z) = 0 at z = a (p. a) ⇒: Let m = 0.105). z→a z→a z→a k(a) 0. Now (z − a)n f (z) thus has a pole at z = a (by Deﬁnition 1. ⇐: Let z = a be a removable singularity and f (a) = limz→a f (z) 0. z→a 0. (c) m > 0 iﬀ r = a is a pole of f of order m. Exercise 8. z→a s>m s<m 0. we obtain that there is an integer m such that 0. since Thus using s = 1. b) ⇒: Let the algebraic order be deﬁned at z = a and be equal to −m (m > 0).103. we have 0. s s−n+n s−n−l lim(z − a) f (z) = lim(z − a) f (z) = lim(z − a) k(z) = ±∞. s < m . 103 Theorem 70 . Then by Exercise 7. Thus.3) (In fact we are free to choose any negative integer less than s). Let f . we have lim f (z) = 0 z→a and lim(z − a) f (z) = 0. a.3 p.where m = n − k. s<0 which follows by (5. lim(z − a) f (z) = ∞. Show: (a) m = 0 iﬀ z = a is a removable singularity and f (a) 0. By Exercise 7. Then there exists an integer n < 0 so that lim |z − a|n | f (z)| = ∞ z→a (5. Now let limz→a |z − a| s | f (z)| = ∞ for some s in R. Then by Exercise 7. s>0=m 2 lim(z − a) f (z) = 0. 0.4). Thus.2 p.

Assume (1. s < −m z→a z→a by Exercise 7. Again by Problem 7 and 8. δ > 0 and h(a) 0.19) or (1.19) holds if s > m and (1. that is lim(z − a)m+1 f (z) = 0 z→a and lim(z − a)m f (z) z→a 0 by Exercise 7. ⇒: We prove this direction by proving the contrapositive (P → Q ⇐⇒ ¬Q ⇒ ¬P). we have h(z) = (z − a)m f (z) has a removable singularity at z = a with h(a) 0.3 p. Then lim(z − a) s f (z) = 0 z→a m for s > m by Exercise 7.1. Suppose that f has an essential singularity at z = a.20) holds for any real number s. Then 0.2). then either m = 0 or m < 0.20) holds. Thus f (z) has a pole of order m at z = a. Then for z a.3 p. Hence z = a is not an essential singularity (by Deﬁnition 1. ⇐: Also this direction by is being proved by showing the contrapositive.20) hold for some real number s. If c ∈ C and > 0 are given then for each δ > 0 there is a number α. Exercise 10. 71 . z→a z→a So lim |z − a| | f (z)| = lim |z − a| z→a z→a s s−m by Exercise 7. δ). in either case there exists an m such that (1. m s>m s<m Exercise 9. Then by Exercise 7.105). (If z = a is removable. the algebraic order of f at z = a is equal to −m (m > 0). such that f (z) = α has inﬁnitely many solutions in B(a. if z = a is a pole. Therefore. |c − α| < . Therefore f (z) = (z − a) h(z) has a zero of order m at z = a. Hence. z = a is either a removable singularity or a pole. A function f has an essential singularity at z = a iﬀ neither (1. Now h(z) := (z − a)−m f (z) has a removable singularity at z = a (since limz→a (z − a)−m+1 f (z) = 0) but h(z) has extended value h(a) = lim(z − a)−m f (z) 0 z→a (by Exercise 7). 0.105). Thus limz→a (z − a)m f (z) = g(a) 0 and lim(z − a)m+1 f (z) = lim(z − a)g(z) = 0. Solution.19) nor (1. ⇐: Let f (z) have a pole of order m at z = a.20) holds if s < m. there exists an integer m such that (1. Prove the following strengthened version of the Casorati–Weierstrass Theorem. then m > 0 by Exercise 8). δ)) and h(a) 0. ⇐: Assume f (z) has a removable singularity at z = a and let z = a be a zero of order m > 0. Then by Exercise 8. Then either z = a is a removable singularity or a pole (by Deﬁnition 1. m > 0 is the algebraic order of f at z = a. |z − a| | f (z)| = ∞. Assume that f has an isolated singularity at z = a which is not an essential singularity.19) or (1. δ)) and g(a) 0. c) ⇒: Let the algebraic order be m > 0. Then f (z) = (z−a)−m g(z) where g(z) ∈ A(B(a. Now f (z) = (z − a)−m h(z) where h ∈ A(B(a. s > −m s s+m lim |z − a| | f (z)| = lim |z − a| |h(z)| = ∞. Then f (z) = (z − a)m h(z) where h(z) is analytic in |z − a| < δ.

0. and g(z) = 0 (t − z)−1 dt are Exercise 16. (d) Characterize those rational functions which have a pole of order m at inﬁnity. a removable singularity. Exercise 15. a pole. a) = 0 for all a in C − G. (a) Let λ ∈ C and show that exp for 0 < |z| < ∞. Give the Laurent series development of f (z) = exp Solution. Solution. Not available. Not available. Show that if G | f (x + iy)|2 dx dy < ∞ then f has a removable singularity at z = a. Not available. 0 ∞ n=1 π 0 1 1 λ z− 2 z 1 π = b0 + bn zn + (−1)n zn cos(nt − λ sin t) dt. Show that for any ω in C there is a sequence {zn } in G with a = lim zn and ω = lim f (zn ). Determine the regions in which the functions f (z) = sin 1 z analytic. show exp for 0 < |z| < ∞. Not available. Exercise 11. where for n ≥ 0 an = (b) Similarly. or an essential singularity at inﬁnity if f (z−1 ) has. where bn = Solution. Let G = {z : 0 < |z| < 1} and let f : G → C be analytic. respectively. What is γ f ? Why? Solution. 72 −1 1 1 z . what can be said about the nature of the singularity at z = a? Solution. Let f be analytic in G = {z : 0 < |z − a| < r} except that there is a sequence of poles {an } in G with an → a. Exercise 17. . If f has a pole at ∞ then the order of the pole is the order of the pole of f (z−1 ) at z = 0. or an essential singularity at z = 0. Exercise 13. Not available. Can you generalize this result? 1 1 λ z+ 2 z 1 π π = a0 + ∞ n=1 an zn + 1 zn eλ cos t cos nt dt. Exercise 14. (b) Prove that an entire function has a pole at inﬁnity of order m iﬀ it is a polynomial of degree m. Let R > 0 and G = {z : |z| > R}. (c) Characterize those rational functions which have a removable singularity at inﬁnity. a pole. Exercise 12. Not available. Suppose that γ is a closed rectiﬁable curve in G such that n(γ. Solution. Not available. Do they have any isolated singularities? Do they have any singularities that are not isolated? Solution. Not available. (a) Prove that an entire function has a removable singularity at inﬁnity iﬀ it is a constant. a function f : G → C has a removable singularity. Let f be analytic in the region G =ann(a.Solution. R). Suppose that p > 0 and G | f (x + iy)| p dx dy < ∞.

5.2 Residues

Exercise 1. Calculate the following integrals: a)

0 ∞

x2 dx x4 + x2 + 1 cos x − 1 dx x2

b)

0

∞

c)

0

π

d)

**cos 2θ dθ where a2 < 1 1 − 2a cos θ + a2
**

π 0

**dθ where a > 1. (a + cos θ)2
**

∞ −∞

**Solution. Note that
**

0 x2 since x4 +x2 +1 is even. z2 f (z) = z4 +z2 +1 has two

∞

x2 dx 1 = x4 + x2 + 1 2

x2 dx x4 + x2 + 1

**Let γR = [−R, R] ∪ CR where CR = Reit , 0 ≤ t ≤ π and R 0 (draw a sketch). 2π iπ 3 , a = ei 3 enclosed by γ . By the residual Theorem, we have simple poles at a1 = e 2 R
**

R

**2πi(Res( f, a1 ) + Res( f, a2 )) =
**

γR

f =

−R

x4

x2 dx + + x2 + 1

2

CR

z4

z2 dz. + z2 + 1

Thus,

∞ −∞

x2 dx = − lim 4 + x2 + 1 R→∞ x

CR

z2 Res( f, ak ). dz + 2πi 4 + z2 + 1 z k=1 |z|2 |dz| ≤ + z2 + 1| |z|2 |dz| − |z|2 − 1

We have f (z) dz =

CR CR

z4

z2 dz ≤ + z2 + 1

2 CR

CR

z4

z2 dz ≤ + z2 + 1

3

CR

=

R R4 − R2 − 1

|dz| =

πR → 0 as R → ∞. R4 − R2 − 1

2π π z2 1 = ei 3 limπ (z − ei 3 ) 4 i3 + z2 + 1 z + z2 + 1 z→e 2π π

|z4

CR

|z|4

**Next, we calculate Res( f, a1 ). Res( f, ei 3 )
**

π

=

z→ei 3

2π

limπ (z − ei 3 )

π

z4

= ei 3 limπ = =

ei 3 1 ei 3 = = 3 + 2z iπ i 3π i 2π 4e 3 + 2e 3 4e 3 + 2 z→ei 3 4z √ √ 1 1 1 1 + 2 3i 1 1 2 + 2 3i = 2 √ = √ + √ 1 1 4(− 2 + 2 3i) + 2 2 3i 4 3i 4 1 1 − √ i. 4 4 3 73

**Finally, we calculate Res( f, a2 ). Res( f, ei 3 )
**

2π

=

2π z→ei 3 4π

lim (z − ei 3 )

2π

z4

4π 2π z2 1 = ei 3 lim (z − ei 3 ) 4 2+1 2π +z z + z2 + 1 z→ei 3 4π 2π

**ei 3 1 ei 3 = = 4π 2π 6π 2π 4z3 + 2z 4ei 3 + 2 4ei 3 + 2ei 3 z→ei 3 √ √ 1 1 − 1 + 2 3i − 1 + 2 3i 1 1 2 = = 2 √ = √ − √ 1 1 4 4(− 2 − 2 3i) + 2 −2 3i 4 3i 1 1 = − − √ i. 4 4 3 = ei 3 lim So 2πi
**

k=1 2

Res( f, ak ) = 2πi

1 1 1 π π 1 1 = 2πi − √ i = − √ i2 = √ . − √ i+ √ − 4 4 3 4 4 3i 2 3 3 3

∞ 0

Hence,

x4

x2 dx 1 π = √ . 2+1 2 3 +x

and γ = [−R, −r] + γr + [r, R] + γR where γR = Reit , 0 ≤ t ≤ π and γr = re−it , 0 ≤ t ≤ π b) Let f (z) = and 0 < r < R (A sketch might help to see that γ is a closed and rectiﬁable curve). Clearly γ f (z) dz = 0 since f is analytic on the region enclosed by γ. Also f (z) dz =

γ −r −R −r −R −r −R

eiz −1 z2

eix − 1 dx + x2 eix − 1 dx = x2

R

R r

eix − 1 dx + x2 e−ix − 1 dx. x2

f+

γr γR

f.

We have

R r

So eix − 1 dx + x2

R r

eix − 1 dx x2

=

e−ix + e−ix − 2 dx = 2 x2 r R cos(x) − 1 dx = 2 x2 r eiz − 1 1 dz − 2 2 z

r

−ix −ix R e +e 2 x2

−1

dx

Therefore,

R r

cos(x) − 1 1 dx = − 2 2 x

γR

γr

eiz − 1 dz z2 eiz − 1 dz z2

**which implies that
**

∞ 0

1 cos(x) − 1 dx = lim − R→∞ 2 x2

γR

eiz − 1 1 dz − lim r→0 2 z2

γr

The ﬁrst term on the right hand side is zero, since eiz − 1 dz z2 ≤ ≤ 1 |eiz − 1| |dz| ≤ 2 |dz| + |eiz | |dz| 2 |z| R γR γR γR 2π 1 (πR + πR) = → 0 as R → ∞. 2 R R 74

γR

**The second term on the right hand side is π, since lim
**

r→0 γr

**1 eiz − 1 dz = − 2πi Res( f (z), 0) = −πi2 = π 2 2 z part and therefore Res( f (z), 0) = i). Hence,
**

∞

( f (z) can be written as

i z +analytic

0

1 π 1 cos(x) − 1 dx = − · 0 − · π = − . 2 2 2 2 x

π

c) If a = 0, then

π 0

Assume a

0. We can write

π 0

cos 2θ dθ = 1 − 2a cos θ + a2

2π 0 dz iz .

cos 2θ dθ =

0

1 1 [sin 2θ]π = [0 − 0] = 0. 0 2 2

2π 0 e2θi +e2θi 2 eθi +eθi 2a 2

cos 2θ dθ 1 = 1 − 2a cos θ + a2 2

cos 2θ dθ 1 = 1 − 2a cos θ + a2 2

z2 +

1 z2

1−

+ a2

dθ.

**Let z = eiθ . Then dz = iz dθ =⇒ dθ = 1 2
**

2π 0 e2θi +e2θi 2 eθi +eθi 2a 2

Hence,

2

1−

+ a2

dθ

=

1 2

|z|=1

1−a z+

|z|=1

= −

1 4ia

z +1 z2 z2 −

1+a2 a z

1 z 4

1 dz = 2 iz 4i +a +1

|z|=1

z4 + 1 dz z2 z − az2 − a + a2 z z4 + 1

|z|=1

dz = −

1 4ia

z2 (z − a) z −

1 a

dz

**By the Residue Theorem and the fact that |a| < 1, since a2 < 1 z4 + 1
**

|z|=1

z2 (z − a) z −

1 a

dz = 2πi Res f ( f, 0) + Res( f, a)

where f (z) =

z4 +1 . z2 (z−a)(z− 1 ) a

**It is easily obtained d 2 z4 + 1 Res( f, 0) = lim z 2 z→0 dz z (z − a) z − = lim
**

z→0

1 a

1 −4z3 z2 − az − a z + 1 − (z4 − 1) 2z − a −

= lim d z→0 dz

2

z4 + 1 (z − a) z −

1 a 1 a

=a+

z2 − az − 1 z + 1 a z4 + 1 z2 (z − a) z − dz = 2πi a +

1 a

1 a

**and Res( f, a) = lim(z − a)
**

z→a

= lim

z4 + 1 z−

1 a

=

a4 + 1 a2 a −

1 a

=

z→a z2

a4 + 1 . a(a2 − 1)

So

|z|=1

z4 + 1 z2 (z − a) z −

π 0 1 a

2a4 4πia3 1 a4 + 1 = 2πi 2 =− . + 2 − 1) a a(a a(a = 1) 1 − a2

Finally,

a2 1 cos 2θ dθ 4πia3 =π =− . · − 4ia 1 − 2a cos θ + a2 1 − a2 1 − a2 75

3 3 z→ai (z + ai) (2ai) 4a i z→ai lim R −R dx + (x2 + a2 )2 f (z) dz CR where the latter integral on the right hand side is zero since 1 dz (z2 + a2 )2 ≤ = 1 1 1 |dz| ≤ |dz| ≤ 2 |z2 + a2 |2 (|z|2 − |a|2 )2 (R − a2 )2 CR CR πR → 0 as R → ∞. e) π log x dx = − . 0 ≤ t ≤ π. 2 )2 +a 4a ∞ a > 0. 4 (1 + x2 )2 ∞ 0 f) dx π = . 1 + x2 2 if 0 < a < 1. So we have by the Residue Theorem f (z) dz = 2πi Res( f. Verify the following equations: a) ∞ 0 (x2 π dx = 3. 1 Solution.Exercise 2. z1 ) γR where Res( f. b) 0 (log x)3 dx = 0. g) ∞ −∞ 2π eax π dx = . 4 (1 + x2 )2 π dθ = . R] + CR where CR = Reit . x 1+e sin aπ 4π h) 0 log sin2 2θ dθ = 4 0 log sin θ dθ = −4π log 2. 1 + x2 if a > 0. (R2 − a2 )2 76 |dz| CR CR . 1 2 a + sin θ 2 [a(a + 1)] 2 ∞ 0 d) π/2 0 if a > 0. c) 0 ∞ π(a + 1)e−a cos ax dx = . z1 ) = = Also 2πi π 1 = = 4a3 i 2a3 d d 1 1 = lim (z − ai)2 z→ai dz (z + ai)2 dz (z − ai)2 (z + ai)2 −2 2 1 lim =− = 3 . a) Let f (z) = (z2 +a2 )2 and γR = [−R. So f has z1 = ai as a pole of order 2 enclosed by γR .

let u = t2 . 0 ∞ (x2 1 dx = + a2 )2 2 ∞ −∞ (x2 1 dx = lim + a2 )2 2 R→∞ it R −R 1 π π dx = = . g) First of all substitute u = e x . R] + CR where CR = Re . so ∞ −∞ eax dx = 1 + ex ∞ 0 ua du = 1+u u t2a−2 t dt = 2 1 + t2 ∞ 0 ∞ 0 ua−1 du. 1 + t2 Now. (x2 + a2 )2 2 2a3 4a3 eiaz c) Let f (z) = (1+z2 )2 enclosed by γR . f (z) has a simple pole i enclosed by γ. then 0 ∞ ua−1 du = 2 1+u ∞ −∞ ∞ 0 ∞ 0 Hence. 1+u t2a−1 dt. θ = arg(z). 0 ≤ t ≤ π. Hence ∞ 0 1 cos ax dx = 2 (1 + x2 )2 ∞ −∞ cos ax 1 dx = lim 2 R→∞ (1 + x2 )2 R −R cos ax 1 π(a + 1)e−a π(a + 1)e−a = dx = 2 )2 2 2 4 (1 + x provided a > 0. By the Residue 2 77 2a−1 . So f has z1 = i as a pole of order 2 the Residue Theorem. (R2 − 1)2 CR (|z|2 1 |dz| − 1)2 since |eiaz | = e−Im(az) ≤ 1 since Im(az) ≥ 0 (a > 0 by assumption). Deﬁne log z on G = {z ∈ C : z 0. −r] + γr + [r.Hence. z1 ) = lim d d eiaz eiaz = lim (z − i)2 z→i dz z→i dz (z + i)2 (z − i)2 (z + i)2 2 iaz iaz (z + i)iaeiaz − 2eiaz (z + i) iae − 2(z + i)e = lim = lim z→i z→i (z + i)4 (z + i)3 2i2 aei a − 2ei (2i)3 2 2 a = Also 2πi = −2ae−a − 2e−a (a + 1)e−a = . By and γR = [−R. R] + γR (0 < r < R). eax dx = 2 1 + ex x2a−1 dx. we have f (z) dz = 2π Res( f. z1 ) γR where Res( f. − π < 2 1+z arg(z) < 3π } with log(z) = log |z| + iθ. −8i 4i R (a + 1)e−a π(a + 1)e−a = = 4i 2 f (x) dx + −R CR f (z) dz where the latter integral on the right hand side is zero since f (z) dz CR = = eiaz |eiaz | dz ≤ |dz| ≤ 2 2 22 CR (1 + z ) CR |1 + z | 1 Rπ → 0 as R → ∞. 1 + x2 Let f (z) = z 2 and γ = [−R.

1 + x2 e(2a−1)(log |−x|+iπ) dx 1 + x2 e(2a−1) log x dx 1 + x2 z2a−1 dz ≤ 1 + z2 γR |z|2a−1 R2a−1 πR2a |dz| ≤ 2 πR = 2 → 0 as R → ∞ |1 + z2 | R −1 R −1 πr2a → 0 as r → 0 r2 − 1 since a < 1. R i2a−1 z2a−1 = 2πi (z − i)(z + i) 2i π π −r −R f (x) dx + γr I f (z) dz + r II f (x) dx + γR III f (z) dz . −πieaπi = (1 − e2aπi ) and so 0 ∞ 0 γ f (z) dz = −e2aπi ∞ 0 f (x) dx + 0 + 0 ∞ f (x) dx + 0 = (1 − e2aπi ) ∞ 0 f (x) dx. ∞ f (x) dx f (x) dx = ∞ −∞ Therefore. II) Similar to IV) γr f (z) dz ≤ since a > 0. . Find all possible values of Solution. Not available. we get −πieaπi = Thus. if r → 0 and R → ∞. −iπ iπ iπ π −πieaπi = = = = . 78 γ exp z−1 dz where γ is any closed curve not passing through z = 0.Theorem f (z) dz γ = 2πRes( f. Exercise 3. 1 − e2aπi e−aπi − eaπi eaπi − e−aπi 2i sin(aπ) 2 sin(aπ) eax dx = 2 1 + ex ∞ 0 π π x2a−1 dx = 2 = 2 sin(aπ) sin(aπ) 1 + x2 provided 0 < a < 1. IV I) −r −R R R f (x) dx = r f (−x) dx = r R (2a−1)iπ r R (−x)2a−1 dx = 1 + x2 R r R r = e = −e IV) f (z) dz = γR γR 2aπi r e(2a−1) log x dx = −e2aπi 1 + x2 x2a−1 dx. Hence. i) = 2πi lim(z − i) z→i = πi Also f (z) dz = γ 2a−1 = πe (2a−1) log i = πe(2a−1)(0+i 2 ) = πeaπi e−i 2 = −πieaπi .

Not available. . Not available. Solution. Exercise 9. . a). an . an such that γ ≈ 0 in G. −n − 1 + ni. Use methods similar to those of Exercises 6 and 8 to show that 1 π = + sin πa a for a an integer. Not available. Not available. n + 1 − ni. . Use Exercise 4 to show that if G is a region and f is analytic in G except for simple poles at a1 . deduce that π2 1 = 2 sin πa n=−∞ (a + n)2 (Hint: Use the fact that for z = x + iy. Show that limn→∞ π(z2 − a2 )−1 cot πz dz = 0. 1 + a ∞ n=1 γ ∞ n=0 ∞ 1 (2n + 1)2 π(z2 − a2 )−1 cot πz dz for a 2a a2 − n2 Solution. . . Let γ be the polygonal path deﬁned in Exercise 6 and evaluate an integer. ak )g(ak )Res( f . Let γ be the rectangular path [n + 2 + ni. Solution. 79 . Exercise 5. 1 1 Exercise 6. . ∞ n=1 2(−1)n a a2 − n2 Solution. | cos z|2 = cos2 x + sinh2 y and | sin z|2 = sin2 x + sinh2 y to show that | cot πz| ≤ 2 for z on γ if n is suﬃciently large. Suppose that f has a simple pole at z = a and let g be analytic in an open set containing a. Show that limn→∞ γ π(z + a)−2 cot πz dz = 0 and. and consequently π cot πa = for a an integer. . Exercise 8. and if g is analytic in G then 1 2πi n fg = γ k=1 n(γ. by using the ﬁrst part. Not available.Exercise 4. Use Exercise 6 to deduce that π2 = 8 Solution. Exercise 7. ak ) for any closed rectiﬁable curve γ not passing through a1 . −n − 2 − ni. Not available.) Solution. Show that Res( f g. n + i + ni] and 2 2 evaluate the integral γ π(z + a)−2 cot πz dz for a an integer. . a) = g(a)Res( f .

m+n+1 2πi γ z otherwise Solution. . Solution. Not available. if m = 2p + n. Exercise 12. R) with f (0) = 0. ¯ Exercise 3. Not available. Res( f . 1) and satisﬁes | f (z)| < 1 for |z| = l. Deﬁne g : B(0. Put z f (z) dz f (z) − ω where γ is the circle |z| = R. In Exercise 1. p ≥ 0 1 = 0. 80 . . for suﬃciently large R. Show that g is analytic and discuss the properties of g. Exercise 11. Equivalently. If f is meromorphic on G and f˜ : G → C∞ is deﬁned by f˜(z) = ∞ when z is a pole of f and f˜(z) = f (z) otherwise. Solution. ¯ Exercise 2. Suppose f is analytic on B(0. 0 ≤ t ≤ 2pi. show that f˜ is continuous. am . Use this result to give another proof of Liouville’s Theorem. Not available.) What can you say if f has inﬁnitely many isolated singularities? Solution. ∞) = − 2πi f when it γ(t) = Re . Find the number of solutions (counting multiplicities) of the equation f (z) = zn where n is an integer larger than or equal to 1.) Solution. Show that Res( f . Prove Theorem 3. Solution. Show that (±1) p (n+2p)! (z2 ± 1)m dz p!(n+p)! . Exercise 4.3 The Argument Principle Exercise 1.6. ∞) = − ∞ k=1 Res( f. Not available. 5. 0 ≤ t ≤ 2π. ∞) is deﬁned as the residue of −z−2 f (z−1 ) at z = 0. .Exercise 10. Not available. Not available. Let f be an entire function and let a. 1 (Res( f . Solution. consider an and bn as functions of the parameter λ and use Exercise 10 to compute power series expansions for an (λ) and bn (λ).12. (bn (λ) is called a Bessel function. ak ). If γ(t) = Reit . a2 . evaluate γ [(z − a)(z − b)]−1 f (z) dz. ρ) → C by g(ω) = 1 2πi γ 0 and f (z) 0 for 0 < |z| ≤ R. Not available. Exercise 13. f (0) ρ = min{| f (z)| : |z| = R} > 0. Let f be analytic in the plane except for isolated singularities at a1 . Let f be analytic in B(0. Let γ be the circle |z| = 1 and let m and n be non-negative integers. b ∈ C such that |a| < R and |b| < R. .

what can you say? ¯ Solution. that is. If we choose f (z) = z. Not available. Show that this solution must be real. Let λ > 1 and show that the equation λ − z − e−z = 0 has exactly one solution in the half plane {z : Re z > 0}. If | f (z)| < 1 for |z| = 1. Not available. is every meromorphic function on G the quotient of two analytic functions on G? Alternately. H(G) is a commutative ring with no zero divisors. show that neither the poles nor the zeros of f have a limit point in G. 1). h are analytic in a neighborhood of B(0. By Roché’s Theorem. Let g(z) = f (z) − z and h(z) = z. Exercise 6. meromorphic functions are analogues of rational functions. Others reserve the term “analytic” for what many call the complete analytic function. h have no poles.) Show that H(G) is an integral domain. Not available. Let f be analytic in a neighborhood of D = B(0. Clearly g. g(z) = f (z) − z has exactly one zero inside the unit circle. But Pg = Ph = 0. we get that . is a ﬁeld. 1) since f is analytic. Let G be a region and let H(G) denote the set of all analytic functions on G. Exercise 7. Some authors call a diﬀerentiable function holomorphic and call functions analytic if they have a power series expansion about each point of their domain. 81 Since h(z) = z has only one zero (z = 0) inside the unit circle. Solution. If we choose f (z) = 1. (The letter “H” stands for holomorphic. Exercise 5. Not available. Show that M(G). Exercise 8. We have said that analytic functions are like polynomials. Not available. Let γ = {z : |z| = 1}. Solution. ¯ Exercise 10. What happens to the solution as λ → 1? Solution. Is a non-constant meromorphic function on a region G an open mapping of G into C? Is it an open mapping of G into C∞ ? Solution. Solution. We have |g(z) + h(z)| = | f (z) − z + z| = | f (z)| < 1 = |z| = |h(z)| ≤ |g(z)| + |h(z)| on γ since | f (z)| < 1 for |z| = 1. show that there is a unique z with |z| < 1 and f (z) = z. then | f (z)| ≤ 1 for |z| = 1 and therefore we do not have any ﬁxed points. Note that g. similarly. If | f (z)| ≤ 1 for |z| = 1. Exercise 9. we get Zg − Pg = Zh − Ph . State and prove a more general version of Rouche’s Theorem for curves other than circles in G. the meromorphic functions on G.Solution. Let f be meromorphic on the region G and not constant. which we will not describe here. there is a unique z such that f (z) − z = 0 ⇐⇒ f (z) = z with |z| < 1. then | f (z)| ≤ 1 for |z| = 1 and therefore we have inﬁnitely many ﬁxed points. The question arises. is M(G) the quotient ﬁeld of H(G)? The answer is yes but some additional theory will be required before this answer can be proved. so Zg = Zh (inside the unit circle). Not available. Hence.

Therefore. then either f has a zero in G or | f | assumes its minimum value on ∂G. Show that if there is a constant c ≥ 0 such that | f (z)| = c for all z on the boundary of G then either f is a constant function or f has a zero in G. Exercise 3. (a) Let f be entire and non-constant. Solution. so f has to have a zero in G. According to the Maximum Modulus Theorem (Version 2). Let G be a bounded region and suppose f is continuous on G− and analytic on G.6. (6. ¯ ∀z ∈ G. we get ¯ max{| f (z)| : z ∈ G} = max{| f (z)| : z ∈ ∂G} = c.) ¯ ¯ ¯ ¯ Solution. Assume there is a constant c ≥ 0 such that | f (z)| = c for all z ∈ ∂G. (b) Let p be a polynomial and show that each component of {z : |p(z)| < c} contains a zero of p. then a ∈ G and we can write G = Ai where Ai are the components of G. then | f | assumes its minimum value on ∂G and we are done. Exercise 2. Otherwise.1 The Maximum Principle Exercise 1. So | f (z)| ≤ c. that is a ∈ Ai for some i.6 either f is constant of f has a zero in G. if a ∂G. Hence ∃ a ∈ G such that | f (a)| ≤ | f (z)| ∀z ∈ G. But f was assumed to be non-constant. If f is a non-constant analytic function on a bounded open set G and is continuous on G− .1) then by Exercise IV 3.1) ¯ Since f ∈ C(G) which implies | f | ∈ C(G) and hence there exists an a ∈ G such that | f (a)| ≤ | f (z)| ≤ c. (Hint: Use 82 . (6. (See Exercise IV. so we can use Exercise IV 3. 3.6 which yields either f (a) = 0 or f is constant. Prove the following Minimum Principle. For any positive real number c show that the closure of {z : | f (z)| < c} is the set {z : | f (z)| ≤ c}.Chapter 6 The Maximum Modulus Theorem 6. But each Ai is a region. Since f ∈ C(G) we have | f | ∈ C(G). either f has a zero in G or | f | assumes its minimum value on ∂G. If a ∈ ∂G.

¯ Exercise 5. Then we can use the Maximum Modulus Theorem (Version 3). . . Not available. 83 . Discuss the behavior of these paths as c → ∞. zn are the zeros of f 3 a in B(0. Solution. Solution. n k=1 ak = a1 a2 . and M is a constant such that whenever z is on ∂∞G and {zn } is a sequence in G with z = lim zn we have lim sup | f (zn )| ≤ M. Show that unless f is a constant. Solution. for each z in G. |λ| = 1. A(r) is a strictly increasing function of r. Not available. R) with | f (z)| ≤ M for |z| ≤ R and | f (0)| = a > 0. Let f be analytic in the disk B(0. f : G → C is analytic. sup{|p(z) − z−1 | : z ∈ A} ≥ This says that z−1 is not the uniform limit of polynomials on A. Instead of showing that lim sup | f (zn )| ≤ M ⇒ lim sup | f (z)| ≤ M. Not available.Exercise 2.) (c) If p is a polynomial and c > 0 show that {z : |p(z)| = c} is the union of a ﬁnite number of closed paths. that is continuous. 1 R) is less than or equal to log 2 log M . Exercise 8. Show that if neither f nor g vanishes in B(0. . We need to show lim sup | f (z)| ≤ M z→a ∀u ∈ ∂∞G. . z→a So assume lim supz→a | f (z)| > M. Exercise 4. R) then there is a constant λ. Show that | f (z)| ≤ M. . . such that f = λg. 1 R). Show that the 1 number of zeros of f in B(0. z→a we show the contrapositive. Solution. we have f (zn ) gets arbitrarily close to f (z). (Notation: Solution. 1− g(z) = f (z) z k=1 k > 0 such and note that g(0) = f (0).) ¯ Exercise 6. Suppose G is a region. Suppose that both f and g are analytic on B(0. Show that there is a positive number that for each polynomial p. Let f be analytic on B(0. consider the function 3 n −1 z . Solution. Hint: If z1 . Not available. Let 0 < r < R and put A = {z : r ≤ |z| ≤ R}. Not available. But this implies lim supz→a | f (zn )| > M since zn → z as z → ∞. that is zn gets arbitrarily close to z and since f is analytic. an . R) with | f (z)| = |g(z)| for |z| = R. R) and for 0 ≤ r < R deﬁne A(r) = max{Re f (z) : |z| = r}. that is lim sup | f (z)| > M ⇒ lim sup | f (zn )| > M. Exercise 7.

It is easy to check that g(0) = 0 since g(0) = We also have |g(z)| = | f (z) − a| ≤1 |1 − a f (z)| ¯ f (0) − a a−a = = 0. Assume there exist an analytic function f : D → D with f ( 2 ) = 3 . we have a = 1 2 and α = 3 . f (a) = α}. (Can you see it?) Now. Does there exist an analytic function f : D → D with f ( 2 ) = 1 Solution. 1 + |a| |z| 1 − |a| |z| 3 4 1 2 and f ( 2 ) = 3 ? Set a = f (0) to obtain the desired result. apply Schwarz Lemma. we have a Eα = { f ∈ A(D) : | f (z)| ≤ 1. We claim that |g(z)| ≤ 1 and g(0) = 0. 1 Exercise 2. We know. Suppose | f (z)| ≤ 1 for |z| < 1 and f is a non-constant analytic function. ⇐⇒ ⇒ ⇒ ⇒ |g(z)| ≤ |z| for |z| < 1 | f (z) − a| ≤ |z|· |1 − a f (z)| ¯ ⇒ || f (z)| − |a|| ≤ | f (z) − a| ≤ |z|· |1 − a f (z)| ≤ |z| + |z|· |a|· | f (z)| ¯ || f (z)| − |a|| ≤ |z| + |z|· |a|· | f (z)| −|z| − |z|· |a|· | f (z)| ≤ | f (z)| − |a| ≤ |z| + |z|· |a|· | f (z)| |a| − |z| |a| + |z| ≤ | f (z)| ≤ . Solution. prove that | f (0)| + |z| | f (0)| − |z| ≤ | f (z)| ≤ 1 + | f (0)| |z| 1 − | f (0)| |z| for |z| < 1.6 which is not possible.2 Schwarz’s Lemma Exercise 1. 16 4 12 ¯ = 0. The answer is no. 1 − a f (0) 1 − aa ¯ ¯ since | f (z)| ≤ 1. In our case. According 4 to the deﬁnition. 84 .583. By considering the function g : D → D deﬁned by f (z) − a g(z) = 1 − a f (z) ¯ where a = f (0). 4 f (a) ≤ and therefore we must have f But f ( 1 ) = 2 2 3 1 − |α|2 1 − |a|2 1 2 ≤ 1− 1− 3 2 4 1 2 2 = 7 16 3 4 = 7 3 7 ¯ = = 0.6.

Can f (0) = 2 ? 2 Solution. Solution. Suppose f is analytic in some region containing B(0. R) and let M(r) = max{| f (z)| : |z| = r}. Suppose f is analytic in a region containing B(0. . (b) By using an appropriate Möbius transformation. Not available.) Solution. and 85 . Let f be analytic in D = {z : |z| < 1} and suppose that | f (z)| ≤ M for all z in D. 0. ﬁnd a formula for f . ﬁnd such an f . then for 0 < r < R. Is there an analytic function f on B(0. Not available. . 1 2. ¯ Exercise 6. 1) and | f (z)| = 1 where |z| = 1. Suppose f : D → C satisﬁes Re f (z) ≥ 0 for all z in D and suppose that f is analytic. Not available. and f (0) = Meia (z1 z2 . . 1). ¯ Exercise 7. M(r) ≤ R+r [A(R) + | f (0)|] R−r 1 − |z| 1 + |z| 1 + |z| 1 − |z| (Hint: First consider the case where f (0) = 0 and examine the function g(z) = f (Rz)[2A(R) + f (Rz)]−1 for |z| < 1. each zk Solution. (Hint: First consider the case where f has no zeros in B(0. f also satisﬁes | f (z)| ≥ (Hint: Use part (a)). if A(r) ≥ 0. Exercise 5. zn ). 1) such that | f (z)| < 1 for |z| < 1.Exercise 3. Not available. 1) and | f (z)| = 1 when |z| = 1. (a) If f (zk ) = 0 for 1 ≤ k ≤ n show that n |z − zk | | f (z)| ≤ M |1 − zk z| ¯ k=1 for |z| < 1.) Solution. A(r) = max{Re f (z) : |z| = r}. Not available. Not available. . Is it unique? 4 Solution. f (0) = f (0) = 3 ? If so. ¯ Exercise 4. (b) If f (zk ) = 0 for 1 ≤ k ≤ n. (a) Show that Re f (z) > 0 for all z in D. What can be said if f (0) 1? (c) Show that f (0) = 1. Suppose that 1 1 f has a zero at z = 4 (1 + i) and a double zero at z = 1 . apply Schwarz’s Lemma to prove that if f (0) = 1 then | f (z)| ≤ for |z| < 1. Find a formula for f . Exercise 8. Prove Caratheodory’s Inequality whose statement is as follows: Let f be analytic on B(0.

x + δ] ⊂ (a. b) choose δ > 0 such that [x − δ. Let f : [a. we have proved (6. Since f is convex on (a. we get (rearrange terms and add/subtract a term) t f (u) − t f (s) − s f (u) + s f (s) − u f (t) − u f (s) + s f (t) − s f (s) ≥ 0 (6.3). Let a < s < t < u < b. we get (t − s)( f (u) − f (s)) − (u − s)( f (t) − f (s)) ≥ 0 f (t) − f (s) f (u) − f (s) ≤ . we get u−t t−s f (t) ≤ f (s) + f (u) u−s u−s which implies (t − s) f (u) + (u − t) f (s) − (u − s) f (t) ≥ 0 which is equivalent to t f (u) − s f (u) + u f (s) − t f (s) − u f (t) + s f (t) ≥ 0.4) 86 . we have that f (x)−δ (x−δ) (z − x) → 0 and f (x+δ)− f (x) (z − x) → 0. u−s u−t Thus. from (6. From (6. Solution. Since f is convex on (a. 1). Not available. Claim: f (x) − f (x − δ) f (z) − f (x) f (x + δ) − f (x) ≤ ≤ δ z−x δ If the claim is true. we have f (u) − f (s) f (u) − f (t) f (t) − f (s) ≤ ≤ t−s u−s u−t (6. δ δ f Taking the limit z → x.4) is equivalent to f (x + δ) − f (x) f (x) − f (x − δ) (z − x) ≤ f (z) − f (x) ≤ (z − x). b) and if a < s < t < u < b. Show that if f : (a.3 Convex functions and Hadamard’s Three Circles Theorem Exercise 1. δ that is |z − x| < δ ⇒ | f (z) − f (x)| < which shows that f is continuous. b] → R and suppose that f (x) > 0 for all x and that f has a continuous second derivative. t−s u−s ⇐⇒ ⇐⇒ u f (u) − s f (u) − u f (t) + s f (t) − u f (u) + u f (s) + t f (u) − t f (s) ≥ 0 (u − s)( f (u) − f (t)) − (u − t)( f (u) − f (s)) ≥ 0 f (u) − f (s) f (u) − f (t) ≤ . (6. (6. b). Does this remain true if f is deﬁned on the closed interval [a. b) → R is convex then f is continuous. Exercise 2.3). t = λs + (1 − λ)u with λ = u−s ∈ (0. Given x ∈ (a.2). Thus f (z) − f (x) → 0. ∀z ∈ (x − δ. b]? Solution.3) ⇐⇒ ⇐⇒ Similar.2) u−t which we show now. b). x + δ).6. Show that f is logarithmically convex iﬀ f (x) f (x) − [ f (x)]2 ≥ 0 for all x.

Now. Solution.4). b]! Here is a counterexample: Deﬁne f : [a.2) applied to x. u−x y−u Interpret these conditions geometrically. Solution. Exercise 4. z. b] → R by 0. Exercise 5. The proof uses (6. Show that a function f : [a. First. x) and apply the second inequality in (6. but f is not continuous at x = b (but the function is continuous on (a.2) applied to the three points x − δ. Applying the outer inequality in (6.4) follows from the outer inequality in (6. Not available.2) to the three points z < x < x + δ gives f (x) − f (z) f (x + δ) − f (x) f (z) − f (x) f (x + δ) − f (x) ≤ ⇐⇒ ≤ x−z x+δ−x z−x δ which gives the second inequality in (6. f is convex on [a. x−x+δ z−x δ z−x The second inequality in (6. 87 .4).2) gives f (x) − f (x − δ) f (z) − f (x) ≤ δ z−x which gives the ﬁrst inequality in (6. f (x) − f (x − δ) f (z) − f (x) f (x) − f (x − δ) f (z) − f (x) ≤ ⇐⇒ ≤ . x = b Clearly. consider a point z ∈ (x − δ. f (y) y 1 f (y)− f (x) f (u)− f (x) (b) a ≤ x < u < y ≤ b gives u−x ≤ y−x . Supply the details in the proof of Hadamard’s Three Circle Theorem.Thus it remains to show the claim. f (z) − f (x) f (x + δ) − f (x) f (z) − f (x) f (x + δ) − f (x) ≤ ⇐⇒ ≤ . f (x) = 1. x + δ. consider the case x < z < x + δ. (c) a ≤ x < u < y ≤ b gives f (u)− f (x) ≤ f (y)− f (u) . Not available. b] → R is convex iﬀ any of the following equivalent conditions is satisﬁed: f (u) u 1 (a) a ≤ x < u < y ≤ b gives det f (x) x 1 ≥ 0. Solution. The statement is not true if f is deﬁned on the closed interval [a.2). Exercise 3. we have proved the claim. z−x x+δ−x z−x δ Thus. the ﬁrst inequality in (6. Not available. x ∈ [a. z. b]. x. b) . Give necessary and suﬃcient conditions on the function f such that equality occurs in the conclusion of Hadamard’s Three Circle Theorem. b)).4) follows from the ﬁrst inequality in (6. Then.

.4 The Phragmén-Lindelöf Theorem Exercise 1. n. the requirement that G be simply connected is not necessary. Not available. . k=1 ≤ | f (z)| k=1 η −η κk k κk k |ϕk (z)| = | f (z)| ∀z ∈ G. . Prove Hardy’s Theorem: If f is analytic on B(0. ϕn on G that never vanish and ∂∞G = A ∪ B1 ∪ . . Let f be analytic in ann(0.1 to regions G with the property that for each z in ∂∞G there is a sphere V in C∞ centered at z such that V ∩ G is simply connected. Solution. . n (Corollary IV 6.) 0 Solution. .¯ Exercise 6. so ϕ may depend on r. Hence. then F is analytic on G and n n −η |ϕk (z)|ηk κk k |ϕk (z)|≤κk ∀k |F(z)| ≤ | f (z)| Hence. Not available. . . ϕn ∈ A(G).(b) Mκ . . Hence. (Note that r is ﬁxed. lim supz→b | f (z)||ϕk (z)|ηk ≤ M ∀ k = 1. R) and not constant then I(r) = 1 2π 2π 0 | f (reiθ )| dθ is strictly increasing and log I(r) is a convex function of log r. Extend Theorem 4. . . deﬁne I2 (r) = 1 2π 2π 0 | f (reiθ )|2 dθ. |ϕk (z)| ≤ κk ∀ z in G for each k = 1. . M. . Exercise 2. Not available. . ϕ2 . . . . z→a k a∈A a ∈ Bk ∀k . Solution. 2π] → C such that ϕ(θ) f (reiθ ) = | f (reiθ )| and consider the function F(z) = 2π f (zeiθ )ϕ(θ) dθ. . R2 ) and not identically zero. . Also because G is simply connected.17). Deﬁne F : G → C by n F(z) = f (z) k=1 −η gk (z)κk k . . R1 . gk (z) = exp{ηk log ϕk (z)} is an analytic branch of ϕk (z)ηk for ηk > 0 and |gk (z)| = |ϕk (z)|ηk ∀ k = 1. In the statement of the Phragmén-Lindelöf Theorem. R1 < r < R2 . 88 . Exercise 7. Solution. . Give some examples of regions that are not simply connected but have this property and some which don’t. . Further a) for every a in A. bounded and ϕk 0 by assumption. 6. We have ϕ1 . Show that log I2 (r) is a convex function of log r. . In Theorem 4.1 suppose there are bounded analytic functions ϕ1 . ∪ Bn such that condition (a) is satisﬁed and condition (b) is also satisﬁed for each ϕk and Bk . there is an analytic branch of log ϕk (z) on G for each k = 1. . . n. . n. lim supz→a | f (z)| ≤ M and b) for every b in Bk and ηk > 0. Hint: If 0 < r1 < r < r2 ﬁnd a continuous function ϕ : [0. lim sup | f (z)| = −ηk (a). Prove that | f (z)| ≤ M for all z in G.

Letting ηk → 0 ∀k = 1. | f (z)| = |F(z)| n k=1 η |gk (z)|−1 κk k ≤ + n k=1 κk ϕk (z) ηk −η −η max{M. G is assumed to be an open set. Assume it would be the case. . .5) implies lim sup | f (z)| ≤ M z→a for all a ∈ ∂∞G. . (6. n gives that | f (z)| ≤ M for all z in G.6) If G would be a region together with the claim would give | f (z)| ≤ M ∀z ∈ G by the Maximum Modulus Theorem III.8). Also. . n→∞ ∀z ∈ ∂∞G (6. Mκn n } ∀z ∈ G. Mκ1 1 . Let f : G → C be analytic and suppose M is a constant such that lim sup | f (zn )| ≤ M for each sequence {zn } in G which converges to a point in ∂∞G. (See Exercise 1. Mκn n } for all z in G and for all ηk > 0. .6). Solution. Show that | f (z)| ≤ M. Proof of the claim: Assume not (6. Mκ1 1 . . Hence. Can we take a = 1 above? Solution. Let G = {z : |Im z| < 1 π} and suppose f : G → C is analytic and lim supz→w | f (z)| ≤ M for w 2 in ∂G. Show that | f (z)| ≤ M for all z in G. we can apply the Maximum Modulus Theorem III to each component once we have shown the claim and additionally we have to show that the boundary of G is the same as taking the boundaries of the components: Clearly. Not available. Exercise 4. Examine exp(exp z) to see that this is the best possible growth condition. . So each component is a region. However. . the boundary of G equals the boundary of all its components. we have −η −η |F(z)| ≤ max{M. since the components of G add up to G. suppose A < ∞ and a < 1 can be found such that | f (z)| < exp[A exp(a|Re z|)] for all z in G. But this would imply that it must be connected to another component of G contradicting the fact a component is a maximally connected subset. So it remains to show that the boundary of the component does not include more than the boundary of G.By the Maximum Modulus Theorem III. Then there needs to be a boundary of a component lying inside G (since every component is lying inside G). . we have that the boundary of the components must include the boundary of G. that is ∀M > 0 ∃a ∈ ∂∞G : lim sup | f (z)| > M z→a which is equivalent to ∀M > 0 ∃a ∈ ∂∞G : lim+ sup{| f (z)| : z ∈ G ∩ B(a. Claim: Let M be a constant such that lim supn→∞ | f (zn )| ≤ M for each sequence {zn } in G with z = lim zn . Exercise 3. . It remains to show the claim which we will do by showing the contrapositive. r)} > M r→0 89 . Thus. But every open set can be written as a union of components (components are open and connected). Therefore. . .

suppose that for every δ. Then limn→∞ αn > M + δ. Not available. . Then. rn )}. Not available. 2 r→∞ Show that | f (z)| ≤ M for all z in G. 90 . Exercise 6. ∃ yn ∈ G ∩ B(a. Let f : G → C be analytic and suppose that G is bounded. Taking lim sup in this inequality yields lim sup | f (yn )| ≥ M + δ. Solution. Hence (6. Next. clearly ∃δ > 0 such that limr→0+ sup{| f (z)| : z ∈ G ∩ B(a. Thus. Show that if limz→z0 |z − z0 | | f (z)| = 0 for every > 0 then | f (z)| ≤ M for every z in ∂G.6). Exercise 5. w z0 . 0 < δ < 1.by deﬁnition. (Hint: If a G. lim sup{exp(− /r)| f (reiθ )| : |θ| < 1 π} = 0. Solution. Exercise 7. consider the sequence αn = sup{| f (z)| : z ∈ F ∩ B(a. there is a constant P such that | f (z)| ≤ P exp(|z|1−δ ). n→∞ that is lim sup | f (yn )| ≥ M n→∞ where limn→∞ yn = a. Let G = {z : Re z > 0} and let f : G → C be an analytic function with lim supz→w | f (z)| ≤ M for w in ∂G. Fix z0 in ∂G and suppose that lim supz→w | f (z)| ≤ M for w in ∂G. therefore limn→∞ αn > M + δ implies αn > M + δ ∀n. In addition. we have limn→∞ rn = 0. which gives “not (6. Let {zn } be a sequence with limn→∞ = a and let rn = 2|zn − a|. r)} > M + δ.5) implies (6. {αn } is a nonincreasing sequence. Obviously. rn ) such that | f (yn )| > M + δ. Not available.) Solution. and also suppose that for every > 0. consider ϕ(z) = (z − z0 )(z − a)−1 . Let G = {z : Re z > 0} and let f : G → C be analytic such that f (1) = 0 and such that lim supz→w | f (z)| ≤ M for w in ∂G. Prove that | f (z)| ≤ M Hint: Consider f (z) 1+z 1−z (1 − x)2 + y2 (1 + x)2 + y2 1 2 .5)”. Also.

u) ≤ µ(s. t) 1 + d(t. t) = 1+d(s.1 The space of continuous functions C(G. t) + µ(t. There is 91 . To show that the metrics d and µ are equivalent on S . < min . Let therefore s. It remains to show the triangle inequality. t) 1 + d(t. u) 1 + d(s. (7. µ(t. µ(s. u) ≤ 1 + d(s. Ω) Exercise 1. d(t. t. d) and let x ∈ O. t) d(t. t) + µ(t.1) t The function f (t) = 1+t is a strictly increasing. The metric d is symmetric and therefore µ is also.5 (Hint: Study the function f (t) = t 1+t for t > −1. The metric d satisﬁes the triangle inequality d(s. u) Together with equation (7. t) + d(t. u)} and by nonnegativity of µ also µ(s. u) d(s. u) = d(s. ∞). t).t) Solution. u). u) = µ(s. Also µ(s. u) 1 + d(s.t) For s. t) = 0 and since d is a metric it follows that s = t. u) 1 + d(s. u)}.t) is well deﬁned in the nonnegative real numbers because d is a metric. d(t. u) ≤ max{µ(s. then also µ(s. concave function on the interval [0. t). u) = + ≤ + 1 + d(s. u)} < d(s. Prove Lemma 1. t) = 0 if and only if d(s. u) 1 + d(s. To show that µ : S × S → R. t) + d(t.t) is a metric we proof that µ satisﬁes the conditions of a metric function.) d(s.Chapter 7 Compactness and Convergence in the Space of Analytic Functions 7. t). Thus if d(s. t ∈ S the value µ(s. u) ≤ max{d(s. 1 + d(s. u) ≤ d(s. u) d(s. u ∈ S . u) d(s. let O be on open set in (S . If instead max{d(s. u) then 1 1 1 . which gives the triangle inequality. u). t) d(t.1) we have µ(s. d(s. t) = 1+d(s. u).

r + . but not on x ∈ S . xn ) < ε for m. n j=−n n an intersection of closed set. d) are open in (S . (d) G is an inﬁnite strip. ε) ⊂ O.18. Supply the omitted details in the proof of Proposition 1. Kn .2 for each of the following choices of G: (a) G is an open disk. (c) G is the plane with n pairwise disjoint closed disks removed. then Bµ (x. To see that it is Cauchy in (S . xn ) < δ whenever m. n ≥ N and therefore also µ(xm . µ) and δ is such that x ∈ Bµ (x. δ) ⊂ Bd (x. Since {xn }n is Cauchy in (S . d). a) G is an open disk. Since this can be done for any element x ∈ O. Depending on the size of r. . Similarly. not violating the condition Kn ⊂ int Kn+1 . R. d) there is N ∈ N such that d(xm . r − 1 n where a ball of negative radius is to be understood as the empty set. r). R − . The opposite statement holds with a similar argument. 1) such that Bd (x. (e) G = C − Z. open sets in (S . Assume that {xn }n is a Cauchy sequence in (S . Here we can deﬁne Kn = {z ∈ C : |z| ≤ n} ∩ 1 B ( j. The fact that exactly the open sets in (S . if δ O is an open set in (S . δ). a ﬁnite number of Kn may be empty. hence Kn is compact. µ) leads to the statement about Cauchy sequences. let ε > 0 be arbitrary but ﬁxed. (b) G is an open annulus. µ) and vice-versa. Exercise 2. Set Kn = B a. y ∈ R|y| ≤ 1 − 1 . δ) ⊂ Bµ (x. ε) ⊂ Bµ (x. Find the sets Kn obtained in Proposition 1. n=1 Exercise 3. δ) ⊂ O then choose ε > 0 with ε ≤ 1−δ which implies Bd (x. 0 ≤ r < R < ∞ so that G = ann(a. R). |x| ≤ n} ∩ z = x + iy ∈ C : x. r. Let a ∈ C and r > 0 such that G = B(a. n ∞ n=1 Although G is unbounded each Kn is bounded and closed and hence compact and G = e) G = C − Z. For this open set G deﬁne Kn = {z = x + iy ∈ C : x. ε). one of which is bounded. n=1 b) G is an open annulus. Deﬁne 1 1 Kn = ann a. n n again with the interpretation that an annulus with inner radius larger than the outer radius is the empty set. d) G = {z ∈ C : |Imz| < 1}. and obviously G = ∞ .ε ε ∈ (0. Choose δ positive such that δ < 1+ε . d) are open in (S . Let a ∈ C be the center of the annulus with radii r. µ). Solution. ε) and δ depends only on ε. n ≥ N. y ∈ R. Also Kn ⊂ int Kn+1 by deﬁnition and G = ∞ . By the above there is a δ small enough that Bd (x. 92 C .

Show that fn → f . Because of the monotonicity of gn . If f ∈ C(G. Since fn is equicontiuous at each point of G. Let F be a subset of a metric space (X. (Dini’s Theorem) Consider C(G. Not available. we also get d( f (x). fn (z) ≤ fn+1 (z) for all z in G) and lim fn (z) = f (z) for all z in G where f ∈ C(G. δ) cover G. Solution. (7. let M = maxi Ni . . Not available.2) by the pointwise convergence ( f (z) = lim fn (z) ∀z ∈ G by assumption). Let { fn } ⊂ C(G. there is an open neighborhood B(x. we have that gn (y) < ∀ n ≥ M in every open neighborhood B(xi . (7.2 we can ﬁnd a sequence {Kn } of compact subsets of G such that ∞ Kn . since f and fn are continuous by assumption. For any y ∈ B(x. r) we have gN (y) < . gn converges pointwise to zero. Not available. Note that the values of N can change the chosen center x of each those neighborhoods. These open neighborhoods will cover G. we can ﬁnd ﬁnitely many points y1 .2 we can ﬁnd a sequence {Kn } of compact subsets of G such that ∞ Kn . By Proposition 1. r). Deﬁne gn = f − fn ∀n. By Proposition 1. Ω) and suppose that { fn } is equicontinuous at each point of G. for every value in G. n=1 WLOG we can assume that G is compact. ∀n ≥ M ∀x ∈ G : | fn (x) − f (x)| < . Thus. . that is ∀ > 0 ∃N ∈ N ∀x ∈ G : 0 ≤ gN (x) < 2 . r). r) around x (with r < ∞). As in the previous exercise. n=1 WLOG we can assume that G is compact. Show lim fn (zn ) = f (z). we can ﬁnd ﬁnitely many x1 . r) cover G for every value of . f (y)) < 3 whenever |x − y| < δ. Since gN is monotonically decreasing.Solution. Show that F is totally bounded. Ω) and f (z) = lim fn (z) for each z then show that fn → f . Since gN is continuous. we surely get gn (y) < for every n ≥ N and for every y ∈ B(x. r) : gN (y) < . Exercise 6. . Ω) which converges to f and {zn } is a sequence in G which converges to a point z in G. Hence. yk ∈ G such that the open neighborhoods B(yi . . Solution. . Since fn is monotonically increasing. Since G is compact. say Ni . gn is monotonically decreasing. .3) 3 whenever |x − y| < δ. that is. Solution. Solution. Exercise 5. fn (y)) < Taking the limit as n → ∞. xk such that B(xi . R) which is monotonically increasing (i. . which shows fn → f . that is ∀ > 0 ∀x ∈ G ∃N ∈ N ∃B(x. r) : ∀y ∈ B(x.. Since fn converges to f . Exercise 4. .e. Exercise 7. since G is 93 . R). R) and suppose that { fn } is a sequence in C(G. Suppose { fn } is a sequence in C(G. d) such that F − is compact. we have that ∀ > 0 ∃δ > 0 : d( fn (x). Therefore. ∀ > 0 ∃M ∈ N. Thus. Clearly gn is continuous ∀n.

(a) Let f be analytic on B(0. If fn (z) = n=0 that fn → f in C(G. k.3). R) and let f be analytic on G with Laurent series development f (z) = k fn (z) = n k=−∞ ak z and show that fn → f in C(G. show an zn . (b) Let G = ann(0.5) and the computations from above we have for all w ∈ ∂B(z0 . With the function g(z) deﬁned in equation (7. r))| ≤ M. min{ 2 .4) 3 3 3 ak zk . Finally. f (x)) ≤ d( fn (x). Then for this ﬁxed z and for any n ∈ N n n | fn (z)| = | k=0 ak zk | ≤ k=0 |ak ||z|k = g(|z|) < ∞. ∞ n k=0 < + + = (7. f (yi )) < 3 ∀n ≥ N ∀i = 1. C).4) 3 ∀n ≥ Ni . Exercise 8.3) and { fn } ⊂ H(G) thus also the limit function f ∈ H(G). 0. we have d( fn (yi ). R) and let f (z) = ∞ an zn for |z| < R.(7. { fn (z) : fn ∈ F } is bounded and therefore has compact closure. This result is proved with the Arzela-Ascoli Theorem.compact. Next. . { fn } itself is an analytic Cauchy sequence in C(G). . r). Put ∞ n=−∞ Solution. r) ⊂ G. Since f is analytic and { fn } is the sequence of partial sums of f . This shows fn → f . . a) Let G = B(0.(7. f (yi )) + d( f (yi ). k. fn (yi )) + d( fn (yi ). f (x)) where we choose yi such that |x − yi | < δ. As a last step recall that H(G) is complete as a subspace of C(G) (Corollary 2. Thus for a given z ∈ G. which shows that fn → f in C(G). The boundary of the disk with radius r is compact and so is g(∂B(z0 . . we want to show equicontinuity at each point of G. Hence every subsequence converges to the same limit. Let z ∈ G and deﬁne r := we have |z|+R 2 (< g(z) := k=0 |ak |zk . r) and for all n ∈ N | fn (w)| ≤ g(|w|). R). . Thus there is a positive number M such that |g(∂B(z0 . To be more precise: For every point x in G. . Let r > 0 be such that B(z0 . . Fix z0 ∈ G and let ε be positive. (7.2). for an arbitrary x ∈ G and n ≥ N we have d( fn (x). (7. r fn (w)(z − z0 ) dw| (w − z)(w − z0 ) By Arzela-Ascoli the sequence { fn } is normal in C(G) so there is a subsequence { fnk } that converges in C(G). we can ﬁnd an open neighborhood B(yi . C). Let F = { fn }n . δ) for some i = 1. . With the given function f deﬁne an analytic function g : G → C.5) R) then z ∈ B(0. then we have d( fn (yi ). r εr Choose δ ∈ 0. Since fn → f pointwise. 94 . f (yi )) < Set N = maxi Ni . r)). 2M } and let |z − z0 | < δ then as in the proof of Montel’s theorem the following estimate holds | fn (z) − fn (z0 )| = | ≤ 1 2πi |w−z0 |=r 2M δ ≤ ε.

Solution. Not available.8) already f = g on G which gives a contradiction on the set K and the point z0 because | fnm (znm ) − f (znm )| → |g(z0 ) − f (z0 )| ≥ ε. Solution. and suppose that f : [a. A set F ⊂ C(G. The sequence { fn } is locally bounded. Exercise 4. Suppose that this integral does converge uniformly on compact subsets of G and that for each t in (a. and again by Montel’s Theorem now applied to the subsequence and again the completeness of H(G) there is an analytic function g such that fnm → g in H(G). let a ∈ R. Deﬁne the integral F(z) = a f (t. Corollary 3. Hence we can conclude that fn → f in H(G). z) F (k) (z) = dt ∂zk a Solution. Let G be a region. (b) For every > 0 there is a number c > 0 such that {c f : f ∈ F } ⊂ B(0. fn (z) → f (z) uniformly for z in {γ}. (b) If F ∈ C(G. On the set A of points of pointwise convergence fn (z) → f and fn (z) → g. Solution. Exercise 5. C) is locally Lipschitz then F is equicontinuous at each point of G. so in particular the subsequence { fnm } with indices corresponding to {znm } is. Prove Vitali’s Theorem: If G is a region and { fn } ⊂ H(G) is locally bounded and f ∈ H(G) that has the property that A = {z ∈ G : lim fn (z) = f (z)} has a limit point in G then fn → f . f (t. Show that if F ⊂ H(G) is normal then F = { f : f ∈ F } is also normal. . Is the converse true? Can you add something to the hypothesis that F is normal to insure that F is normal? 95 b . z) dt exists uniformly for z in any compact subset of G. Exercise 6. Let f. Not available. Show that for a set F ⊂ H(G) the following are equivalent conditions: (a) F is normal. say {znm } that converges to a point z0 ∈ K. Not available. In other words there must be an ε > 0 so that for all n ∈ N there is zn ∈ K with | fn (zn ) − f (zn )| ≥ ε.2 Spaces of analytic functions Exercise 1. Now G is a region and A has a limit point in G so by the Identity Theorem (Chapter 4. Exercise 2. The proof of Montel’s Theorem can be broken up into the following sequence of deﬁnitions and propositions: (a) Deﬁnition. so there is a subsequence { fnk } ⊂ { fn } that converges to f in H(G). z) dt to be uniformly convergent on compact subsets of G if limb→∞ a f (t. (c) If F ⊂ H(G) is locally bounded then F is locally Lipschitz. ∞] × G → C is a continuous func∞ tion. Exercise 3.7. . f2 . Solution. be elements of H(G) and show that fn → f iﬀ for each closed rectiﬁable curve γ in G. . Not available. C) is locally Lipschitz if for each a in G there are constants M and r > 0 such that | f (z) − f (a)| ≤ M|z − a| for all f in F and |z − a| < r. ∞). Suppose that fn f in H(G) then there must be a compact set K ⊂ G the convergence is not uniform on K. ) (here B(0. By compactness of K extract a subsequence of {zn }. f1 . The sequence { fn } is a locally bounded sequence in H(G) and by Montel’s Theorem then { fn } is normal. Prove that F is analytic and ∞ k ∂ f (t. · ) is analytic on G. ) is the ball in H(G) with center at 0 and radius ).

the sequence {gn } converges to g(z) = f (z) − f (z0 ) on the open connected set G\{z0 }. 0 < r < 1. Choose an arbitrary point z0 ∈ G and deﬁne the sequence gn (z) = fn (z) − fn (z0 ). (7.10 b) yields that { fn } converges to f ∀ z ∈ D. Exercise 7. Assume { fn } ⊂ H(G) is a sequence of one-one functions which converge to f where G is a region (open and connected). we 96 . 1) and for 0 < r < 1 let γr (t) = re2πit . Not available. we have | f (z) − fn (z)| |dz| < 2π δ. Exercise 10. (7.0≤t≤1 sup | f (z) − fn (z)| z∈γr z∈γr γr |dz| ∀0 < r < 1. By the assumption. Exercise 8. Clearly. 0 < r < 1. If G is a region. we have f (a) − fn (a) = Thus. Let K be an arbitrary compact subset r of D = B(0. Suppose f is not the constant function. Then γr | f (z) − fn (z)| |dz| ≤ = γr =re2πit . z−a 2π δ | f (z) − fn (z)| |dz| < = . Not available. In addition. 1) and choose 0 < r < 1 such that K ⊂ B(0. γr | f (z) − fn (z)| |dz| → 0 as n → ∞ for each r. Not available. ⇐: Assume γ | f (z) − fn (z)| |dz| → 0 as n → ∞ for each r. | f (a) − fn (a)| ≤ 1 2π ∂B 1 2πi ∂B f (z) − fn (z) dz. Solution.Solution. Let D = B(0. then by Cauchy’s Integral formula. But supz∈γr | f (z) − fn (z)| → as n → ∞ since limn→∞ fn (z) = f (z) ∀ z ∈ D. 0 < r < 1. sup | f (z) − fn (z)|· 2πr. Since it converges uniformly on all compact subsets of D. Let D = {z : |z| √ 1} and show that F ⊂ H(D) is normal iﬀ there is a sequence {Mn } of positive < constants such that lim sup n Mn ≤ 1 and if f (z) = ∞ an zn is in F then |an | ≤ Mn for all n. r) and denote the shortest distance between an arbitrary point in K and an arbitrary point on ∂B by δ > 0. Therefore. fn (a) → f (z) uniformly ∀a ∈ K. Suppose F is normal in H(G) and Ω is open in C such that f (G) ⊂ Ω for every f in F . Solution. show that either f is one-one or f is a constant function. 0 ≤ t ≤ 1. Let { fn } ⊂ H(G) be a sequence of one-one functions which converge to f . Then. which is again a region. Show that a sequence { fn } in H(D) converges to f iﬀ γ | f (r) − fn (z)| |dz| → 0 as n → ∞ for each r. Show that if g is analytic on Ω and is bounded on bounded sets then {g ◦ f : f ∈ F } is normal. Proposition 1. Exercise 9. r Solution.6) ∂B Let a ∈ K.6) 2πδ ∂B | f (z) − fn (z)| 1 |dz| ≤ |z − a| 2πδ Hence. we have to show that f is one-one. 0 Solution. ⇒: Assume limn→∞ fn (z) = f (z) ∀z ∈ D.

Since g(z) = f (z) − f (z0 ) is not identically zero on G\{z0 }. we get f 2 (a)r = and then integrate with respect to r yields R 1 2π f (a + reit ) 0 2 dt. Suppose that { fn } is a sequence in H(G). (a) Show that if f is analytic on an open set containing the disk B(a.7) f 2 (a)r dr = f 2 (a) 0 R2 = 2 (7. R) ⊂ G and f : G → C be analytic. since fn are assumed to be one-one functions. Therefore. 1 2π 2π 0 f (a + reiθ ) r dθ 2 (7. the sequence {gn } satisﬁes the conditions of the Corollary 2. ¯ Exercise 13. letting g(z) = f (z)2 . we obtain f 2 (a) = Multiply by r. f is a non-constant function. Show that lim tan nz = −i uniformly for z in any compact subset of G = {z : Im z > 0}. we have shown that f is one-one on G. Exercise 12.have that gn never vanishes on G\{z0 }. R) then | f (a)|2 ≤ 1 πR2 2π 0 0 R | f (a + reiθ )|2 r dr dθ. (ii) fn (an ) = α for suﬃciently large n. and fn → f in H(G). a) Let B(a. we must have g(z) = f (z) − f (z0 ) has no zeros in G\{z0 }. Solution. ¯ Solution. Not available. Show that F is normal. Let a ∈ G and α = f (a). Solution. (b) Let G be a region and let M be a ﬁxed positive constant. either g ≡ 0 or g never vanishes (g has no zeros). where the region is G\{z0 }. Exercise 11. Since z0 was an arbitrary point in G. Thus. Not available.6 resulting from Hurwitz’s Theorem. 97 2 . R 0 2 f (a + reiθ ) r dr dθ. Let γ(t) = a + reit for 0 ≤ t ≤ 2π. Let F be the family of all functions f in H(G) such that G | f (z)|2 dx dy ≤ M. show that there is a sequence {an } in G such that: (i) a = lim an . 0 γ Thus. This implies f (z) f (z0 ) ∀z ∈ G\{z0 }. then by Cauchy’s Integral Formula we have g(a) = 1 2πi g(w) 1 dw = w−a 2π 2π 2π f (a + reit ) dt.7) R 0 1 2π 2π 0 f (a + reiθ ) r dθ dr = 2 1 2π 2π 0 0 R f (a + reiθ ) r dr dθ 2 which implies f 2 (a) = So | f (a)|2 ≤ 1 πR2 0 1 πR2 2π 0 2π 0 R f (a + reiθ ) r dr dθ.

2 (7. f ∈ F } = ∞. that is ∃ r > 0 such that sup{| f (z)| : |z − a| < r. This is equivalent to 1 + |z|2 1 + |a|2 2|z − a| 1+ |z|2 |z − a| < √ 1+ r |a|2 2|z − a| <δ 2r < √ 2 1 + |a|2 1+R 1 + |z|2 ⇐⇒ ⇐⇒ 1+ R2 98 . Prove Proposition 3. | fn (zm )| ≥ 7. Apply part a) to the ball B(zm . In addition.8) πR3 n . a) Let a ∈ C and r > 0. That is. r) ⊂ B(0. According to Montel’s Theorem. Let F be the family of all functions f in H(G) such that G | f (z)|2 dx dy ≤ M. δ). R ) (a picture might help). To show: There is a number δ > 0 such that B∞ (a. δ) ⊂ B(a.8) To show that F is normal. Proof of the claim: Let z ∈ B∞ (a. Solution. ˆ π R 1 ˆ πR2 2π 0 0 ˆ R fn (zm + reiθ ) r dr dθ 2 ˆ B(zm . say {znk }. We will prove this fact by contradiction: Assume F is not locally bounded. Then there is a compact set K ⊂ G such that sup{| f (z)| : z ∈ K. ∃ zm ∈ K such that | fn (zm )| ≥ n .9) we have M1 < ∞. R) ⊂ B(z0 . R). Let R > 0 such that B(a. R) to 2 2 obtain: | fn (zm )|2 ≤ So | fn (zm )| ≤ But by (7. we can get zm ∈ B(z0 . δ) ⊂ B(0. we write {znk } = {zn }. F is locally bounded and therefore normal. R) and B∞ (a. 2 This gives a contradiction if we let n → ∞.R)⊂G ≤ 1 ˆ πR2 G 1 M < ∞. with znk → z0 ∈ K since K is compact. it suﬃces to show that F is locally bounded. WLOG. 153). R) ⊂ G (since G is open). we can 2 ˆ ˆ ˆ ﬁnd an R > 0 such that B(zm . R ) is open. δ) ⊂ B(a. Therefore. Thus. f ∈ F } < ∞ (p. | fn (z)|2 dx dy ≤ ˆ (7.9) Every sequence has a convergence subsequence.3 Spaces of meromorphic functions Exercise 1. Claim: Pick 2r δ= √ 1 + R2 1 + |a|2 to obtain B∞ (a. If we pick m large enough. R). there is a sequence { fn } in F such that sup{| fn (z)| : z ∈ K} ≥ n.3. Clearly z0 ∈ G since z0 ∈ K (K ⊂ G). r). R ) because B(z0 . (7. hence ∃ R > 0 such that B(z0 .b) Let G be a region and M be a ﬁxed constant.

r) ⊂ B∞ (a. a) < δ. δ Claim: Pick r < 2 to obtain B(a. r) ⊂ B∞ (a. To show: There is a number r > 0 such that B(a. Therefore |z − a| < r. r 1 + R2 √ 1 + R2 b) Let δ > 0 and a ∈ C.which implies (z ∈ B∞ (a. B∞ (a. r). δ). δ). r). This is equivalent to ⇒ |z − a| < r δ |z − a| < 2 δ |z − a| < · 1 2 δ 1 + |z|2 1 + |a|2 |z − a| < 2 1 + |z|2 1 + |a|2 d(z. δ) ⊂ B(a. δ). that is |z| < R) |z − a| < √ ⇐⇒ So z ∈ B(a. r) ⊂ B∞ (a. Therefore. δ). Claim: Choose K = B(0. (Clearly K is compact and r > 0 since δ ≤ 2.) 99 . 2|z − a| <δ c) Let δ > 0. δ). To show: There is a compact set K ⊂ C such that C∞ − K ⊂ B∞ (∞. δ) ⇒ z ∈ B(0. 4 δ2 δ r< 2 ⇐⇒ 1≤ √ ⇒√ 1+|z|2 1+|a|2 ⇐⇒ ⇐⇒ So z ∈ B∞ (a. r). B(a. Proof of the claim: Let z ∈ B(a. δ). R). r) where r > − 1 to obtain C∞ − K ⊂ B∞ (∞.

z) < δ} ∪ {∞} 2 < δ ∪ {∞} z ∈ C : 2 1 + |z| 2 2 < √ z ∈ C : ∪ {∞} 2 + 1 2 r 1 + |z| {z ∈ C : r ≤ |z|}} ∪ {∞} {C ∪ {z ∈ C : |z| < r} ∪ {∞} C∞ − {z ∈ C : |z| < r} C∞ − B(0. δ) = = {z ∈ C : d(∞. r) for a given r > 0. d) Let K ⊂ C where K is compact. δ) ⊂ C∞ − B(0. So it suﬃces to show: There is a number δ > 0 such that B∞ (∞. δ) ⊂ C∞ − B(0. Proof of the claim: B∞ (∞. z) < δ} ∪ {∞} B∞ (∞. To show: There is a number δ > 0 such that B∞ (∞. r) such that B(0. δ). Claim: Choose δ ≤ and hence B∞ (∞. r) ⊃ K where r > 0. r) ⊂ C∞ − K. to obtain B∞ (∞. r)} ∪ {∞} {C − {z ∈ C : |z| ≤ r}} ∪ {∞} 4 − 1 ∪ {∞} C − z ∈ C : |z| ≤ 2 δ r> 4 δ2 −1 = = = = Therefore. r) δ≤ √ 2 ⊂ r2 +1 = = = = . r). δ) ⊂ C∞ − K. 4 − 1 ∪ {∞} z ∈ C : |z| > 2 δ 4 − 1 < |z| ∪ {∞} z ∈ C : 2 δ 2 < δ ∪ {∞} z ∈ C : 2 1 + |z| {z ∈ C : d(∞.Proof of the claim: C∞ − K = = = ⊂ = C∞ − B(0. Clearly C∞ − B(0. δ). δ) ⊂ C∞ − K. r) {C − B(0. 100 √2 r2 +1 C∞ − K ⊂ B∞ (∞. Let B(0.

Not available. Show that | f (a)| ≤ 1. Suppose {zn } is a sequence in G which converges to a point z in ∂G. and such that f −1 : Ω → G is also continuous (a homeomorphism). Not available. Not available. Not available. Exercise 3. If f is one-one. Solution. there is a unique analytic function g : G → D having the properties a) g(a) = 0 and g (a) > 0 b) g is 1-1 c) g(G) = D (by the Riemann Mapping Theorem). δ) ⊂ C∞ − B(0. and we have h(D) = D by construction. Show that if F ⊂ M(G) is a normal family in C(G. (a) Let G be a region. Exercise 5. r). Solution. let a ∈ G and suppose that f : (G − {a}) → C is an analytic function such that f (G − {a}) = Ω is bounded. (b) Show that there is no one-one analytic function which maps G = {z : 0 < |z| < 1} onto an annulus Ω = {z : r < |z| < R} where r > 0. onto. Clearly G is a simply connected region (not the whole plane). Re z > 0} onto B(0. Let G and Ω be open sets in the plane and let f : G → Ω be a continuous function which is one-one. Solution.4 The Riemann Mapping Theorem Exercise 1. show that f (a) ∈ ∂Ω. Solution. Then. 7. also suppose that w = lim f (zn ) exists. 1) in a one-one fashion. Exercise 2. since f and g are analytic and one-one. B∞ (∞. f (a)=a 101 . we have f : G −− −→ G (since Re f (z) > 0) and g : −−− −− −→ D so g−1 : D −− −→ G we can deﬁne h(z) = g( f (g−1 (z))). Solution. Show that f has a removable singularity at z = a. C∞ ) then µ(F ) is locally bounded. Exercise 2. Let G be a simply connected region which is not the whole plane and suppose that z ∈ G ¯ whenever z ∈ G. Let a ∈ G. Prove that w ∈ ∂Ω. Since. We have h(0) = g( f (g−1 (0))) = g−1 (0)=a 1−1 1−1 1−1 g( f (a)) = g(a) = 0. Show that f (G+ ) must lie entirely above or entirely below the real axis. Let G+ = {z ∈ G : Im z > 0}. Find an analytic function f which maps {z : |z| < 1.Therefore. Solution. f (a) > 0 and f (G) = D. Let a ∈ G ∩ R and suppose that f : G → D = {z : |z| < 1} is a one-one analytic function with f (a) = 0. with Re f (z) > 0 for all z in G. Clearly h(z) is analytic and one−−− −−− onto onto one. Exercise 4. one-one. Not available. and f (a) = a for some real number a. Let f be analytic on G = {z : Re z > 0}.

the hypothesis of Schwarz’s Lemma are satisﬁed. Prove that for any one-one analytic map h of G1 into G2 with h(a) = α it follows that |h (a)| ≤ | f (a)|. Thus. Let f be a one-one analytic mapping of G1 onto G2 . Let a ∈ G1 and put α = f (a). Suppose h is not assumed to be one-one. So = g ( f (g−1 (z))) f (g−1 (z)) where the last step follows from Proposition 2. We have F (z) = f −1 (h(z)) = f (h(z))h (z) = 1 f ( f −1 (h(z))) h (z) 1−1 where the last step follows by Proposition 2. F(z) is one-one and F : G1 → G1 by −−− construction.We also have |h(z)| ≤ 1 ∀z ∈ D since h : D → D. This is well-deﬁned since f : G1 −− −→ G2 and h : −−− onto 1−1 G1 −− −→ G2 . Therefore Thus. Since |F (a)| = |h (a)| ≤1 | f (a)| 102 . Let G1 and G2 be simply connected regions neither of which is the whole plane. Deﬁne the function F(z) = f −1 (h(z)). we have shown that | f (a)| ≤ 1. We have F(a) = f −1 (h(a)) = f −1 (α) = a.20 provided g (g−1 (z)) h (0) = g ( f (g−1 (0))) f (g−1 (0)) 1 (g−1 (0)) g 1 1 = g (a) f (a) = f (a) = g ( f (a)) f (a) g (a) g (a) |h (0)| = | f (a)| ≤ 1.20 provided f ( f −1 (h(z))) F (a) = which is well-deﬁnes since f (a) 0. what can be said? Solution. (g (a) > 0 by assumption). So h (a) h (a) 1 h (a) = = f (a) f ( f −1 (h(a))) f ( f −1 (α)) 0 by assumption (f is one-one). we get |h (0)| ≤ 1. Clearly F(z) is analytic since f and h are analytic. We have h (z) = g( f (g−1 (z))) = g ( f (g−1 (z))) f (g−1 (z)) g−1 (z) 1 g (g−1 (z)) 0. Exercise 6. and hence. we get |F (a)| ≤ 1. Thus by Exercise 5.

Also h is one-one. (7. for θ ∈ [0. analytic and maps G onto ∆. we have f (z) = h(z) 103 . too. Let G be a simply connected region and suppose that G is not the whole plane. Since g is one-one mapping from G onto ∆. g2 is one-one and analytic and maps G onto ∆. Next. g1 is a Möbius transformation mapping ∆ to ∆ and in addition g1 is one-one and analytic. 2π). let h(z) = g3 (g2 (z)). Let a ∈ G such that f (a) = 0 and f (a) > 0.10) 0 ∀z ∈ G since g2 is one-one. By uniqueness of the map f ( f satisﬁes the properties of the Riemann Mapping Theorem). Recall g (a) 0 1 since g is one-one and α ∈ ∆. Finally. where α ∈ ∆ (α does not have to be zero). If h is not one-one. then we get the same result. so eiθ = −1 or θ = −Arg(g (a)) will work. so 1−|α|2 > 0). h (z) = (g3 (g2 (z))) = (g3 (g1 (g(z)))) = g3 (g1 (g(z)))g1 (g(z))g (z) (1 − αg(z))· 1 − (g(z) − α)· (−α) ¯ ¯ g (z). Solution. deﬁne g1 (z) = z−α . 1 − αz ¯ Clearly. We have h(a) = g3 (g2 (a)) And. 2 2 )2 (1 − αα) ¯ (1 − |α| 1 − |α|2 = g3 (0) = 0. Exercise 7. = eiθ 2 (1 − αg(z)) ¯ So. However. But g2 (a) need not necessarily satisfy g2 (a) > 0. Let ∆ = {ξ : |ξ| < 1} and suppose that f is an analytic.10) Clearly. g2 (a) deﬁne the rotation g3 (z) = eiθ z (7. For example: If g (a) < 0. Then. we clearly have g(a) = α. since I do not need the assumption that F(z) has to be one-one for exercise 5. 2π) such that h (a) > 0 (depending on the sign of g (a). one-one map of G onto ∆ with f (a) = 0 and f (a) > 0 for some point a in G. h (a) = eiθ 1 − |α|2 1 (1 − αα) − 0 ¯ g (a) = eiθ g (a) = eiθ g (a). Therefore. In addition. then pick θ = π. one-one map of G onto ∆ and express g in terms of f .gives |h (a)| ≤ | f (a)|. Let g be any other analytic. Now deﬁne g2 (z) = g1 (g(z)). g3 is one-one and analytic. we can choose a θ ∈ [0. we have g2 (a) = g1 (g(a)) = g1 (α) = 0. Clearly g3 maps ∆ onto ∆.

Then the product (1 + zn ) converges absolutely iﬀ ∞ zn converges k=1 absolutely (by Corollary 5. Not available. ones we have proved the claim. = 1 by Exercise 2. n=1 n=1 104 log(1+z) z (1 + |zn |) converges. that is log(1 + |zn |) converges. 0. The function f : C → C. r2 . The inner boundary {z : |z| = r1 } of the domain is mapped to the inner boundary of 2 2 ann(0. R1 ) bijectively r1 into ann(0. n=1 n=1 Then ∞ log(1+|zn |) converges iﬀ (1+|zn |) converges (by Proposition 5. For the outer boundary note that by assumption R1 = Rr2r1 . ∞ zn converges absolutely iﬀ ∞ |zn | n=1 k=1 converges (by deﬁnition).165 applied to wn = 1+|zn |). f (z) = h(z) = g3 (g2 (z)) = eiθ g2 (z) = eiθ g1 (g(z)) = eiθ So 1 − αg(z) = eiθ g(z) − αeiθ ¯ f (z) − αg(z) f (z) = eiθ g(z) − αeiθ ¯ (eiθ + α f (z))g(z) = f (z) + αeiθ ¯ f (z) + αeiθ g(z) = iθ . that is log(1 + xn ) converges. 7. we obtain: (1 + zn ) converges absolutely iﬀ (1 + |zn |) converges. 166 applied to wn = 1 + zn ). we have log(1 + xn ) → 0 and therefore xn → 0. Therefore the result is proved.since h satisﬁes the properties of the Riemann Mapping Theorem. we have xn → 0. So given > 0.2 p. ∞ |zn | converges iﬀ ∞ log(1 + |zn |) converges. r1 . that is xn converges. Show that (1 + zn ) converges absolutely iﬀ Solution. Therefore.6 p. r2 . R2 ). an analytic and non-constant function also. xn converges. 1 − αg(z) ¯ ⇐⇒ ⇐⇒ ⇐⇒ Exercise 8. Therefore. By the . Proof of the claim: Let xn = |zn | for convenience.5 The Weierstrass Factorization Theorem Exercise 1. Clearly xn ≥ 0 ∀n ∈ N. By the z comparison test. Let r1 . log(1 + xn ) converges. we have 0 ≤ log(1 + |xn |) ≤ (1 + )xn for suﬃciently large n where the last inequality follows since limz→0 log(1+z) = 1 by Exercise 2. 1). Assume Re(zn ) > −1. r1 . be positive numbers such that R1 /r1 = R2 /r2 . e + α f (z) ¯ g(z) − α . ⇒: Assume |zn | converges. show that ann(0. So given > 0. f (z) = r2 z is analytic and it maps the annulus ann(0. Solution. Exercise 9. 1) such that f never vanishes and f (G) = B(0. The inverse function is f −1 (z) = r2 z. that is |zn | converges. R2 ). Thus. R1 . Solution. we have (1 − )xn ≤ log(1 + xn ) for suﬃciently large n where the last inequality follows since limz→0 comparison test. Therefore. so if |z| = R1 then | f (z)| = rRr1 = 1 r1 R2 . r2 . too. R2 . Claim: ∞ |zn | converges iﬀ ∞ log(1 + |zn |) converges. Show that there is an analytic function f deﬁned on G =ann(0. ⇐: Assume log(1 + |zn |) converges. R1 ) and ann(0. r2 . R2 ) are conformally equivalent. n=1 Thus.

Therefore. Exercise 4. ˜ If f and g have no common zeros.Exercise 2. z Solution.165 (−1)k+1 zk k z = ∞ k=1 (−1)k+1 zk−1 =1+ k ∞ k=2 (−1)k+2 zk−1 =1+ k ∞ k=1 (−1)k zk k+1 and therefore where the sum is an analytic function. and f1 and g1 have no common zeros.15 f p. show that a + |a|z 1+r ≤ (1 − az)a ¯ 1−r (b) Let {an } be a sequence of complex numbers with 0 < |an | < 1 and product ∞ |an | an − z B(z) = an 1 − an z ¯ n=1 105 (1 − |an |) < ∞. let h = 1. then clearly f1 and g1 are analytic functions on G such that f (z) = h(z) f1 (z) h and g(z) = h(z)g1 (z) for all z in G. (a) Let 0 < |a| < 1 and |z| ≤ r < 1.2 Chapter 5 p. f (z) = log(1 + z) = z ∞ k+1 zk k=1 (−1) k p. Let {a j } be the zeros of f with multiplicity n j . Let f and g be analytic functions on a region G and show that there are analytic functions f1 . 103). g1 and h are analytic on G such that f (z) = h(z) f1 (z) and g(z) = h(z)g1 (z) for all z in G and f1 and g1 have no common zeros. First note that. Let f and g be analytic functions on the region G. Then {z j } form ˜ a sequence of distinct points in G with no limit points on G (since the zeros are isolated). (since f1 and g1 have removable singularities at the z j ’s by construction. ∞ k =1+ 1 + k z lim f (z) = lim (−1) z→0 z→0 k + 1 k=1 ∞ k=1 (−1)k (limz→0 z)k =1+ k+1 ∞ k=1 0=1 log(1 + z) = 1. Then there is an analytic function h deﬁned on G whose only zeros are at the points z j with multiplicity m j (by Theorem 5. z Exercise 3. Show that the inﬁnite . since lim z f (z) = lim z z→0 z→0 log(1 + z) z log(1 + z) = lim log(1 + z) = log(1) = 0 z→0 z ∞ k=1 (by Theorem 1. In addition. Let {b j } be the zeros of g with multiplicity n j . We know log(1 + z) = if |z| < 1. Clearly f1 . Hence. limz→0 Solution. Otherwise let {z j } be the set of common zeros of f and g with multiplicity m j = min(n j . f (z) = has a removable singularity at z = 0. Let f1 = h and g1 = g . g1 . Prove that limz→0 log(1+z) = 1. we have that f1 and g1 have no common zeros by construction. Then f1 = f and g1 = g. n j ). the multiplicity of z j ’s in f and g is greater or equal than the multiplicity of z j ’s in h). and h on G such that f (z) = h(z) f1 (z) and g(z) = h(z)g1 (z) for all z in G. 170).

1 n i 1 Let zn := 1 + n = rn eiθn with rn = 1 + n2 and θ = arctan then ∞ rn = r and ∞ θn = θ. 1) such that (1 − |an |) < ∞ and every number eiθ is a limit point of {an }. n2 Thus for any integer N ∈ N N N 1≤ n=1 zn ≤ 1+ n=1 1 n2 where the last term is absolutely convergent by an example in class. = exp − log(2n) + log(n + 1) = exp log 2n 2n So. Discuss the convergence of the inﬁnite product Solution. θn = ∞ n=1 arctan 1 1 ≥ n 2 ∞ n1 1 n is valid. Show that Solution. thus z 0). i n i n 1+ and 1+ . 1] the estimate x 2 ∞ n=1 .converges in H(B(0. i The second inﬁnite product is a product of real numbers and zn = |1 + n | = 1+ 1 . = lim Pn = lim n→∞ n→∞ 2n 2 n2 106 . If there is z = reiθ such that ∞ n=1 zn = z.) (c) Find a sequence {an } in B(0. therefore the angles do not converge and there is no z ∈ C with the desired properties (note that rn > 1 for all n. bounded sequence that converges in the real numbers. Not available. n2 The estimate 1+ 1 n2 ≤ 1+ 1 2 n2 implies that 1 ≤ zn ≤ 1 + N n=1 1 . 1+ i n . Therefore the sequence of the partial N products n=1 zn is an increasing. Discuss the convergence of the inﬁnite products Solution. We claim that this product does not converge. Consider Pn k−1 k+1 1 = k k k k=2 k=2 k=2 n n k−1 k+1 k−1 k+1 = exp log = exp log k k k k k=2 k=2 n n = exp log(k − 1) − log(k) + exp log(k + 1) − log(k) = 1− 1 = k2 1− 1 k 1+ k=2 k=2 n n n ∞ n=2 1− 1 n2 1 = 2. Exercise 5. Solution. The right hand side does not converge.4) gives absolute convergence. = exp log(1) − log(n) + exp − log(2) + log(n + 1) n+1 n+1 = . With log(zn ) > 0 for all n. Exercise 7. ∞ n=2 1− 1 n+1 1 = . First consider ∞ n=1 ∞ 1 n=1 n p for p > 0. n=1 n=1 Since arctan(x) ≥ 1 for all values x ∈ (0. Exercise 6. the comparison with the sum of logarithms (Proposition 5. Not available. 1)) and that |B(z)| ≤ 1. What are the zeros of B? (B(z) is called a Blaschke Product.

The behavior of ∞ (1 + z2n ) is almost the same as the previous one. We conclude that ∞ (1 + zn ) does not converge. Consider the inﬁnite product ∞ (1 − zn ). (a) Show that ∞ f (z) fn (z) k k=1 n k ∞ n=1 fn (z) converges converges in H(G) and equals f (z). f |g if there is an h in H(G) such that g = f h. In symbols. If f ∈ H(G) let Z( f ) be the set of zeros of f counted according to their multiplicity. then the convern=1 gence is uniform for |w| ≤ r < 1. then the inﬁnite product diverges. Not available.S.c. For which values of z do the products (1 − zn ) and (1 + z2n ) converge? Is there an open set G such that the product converges uniformly on each compact subset of G? If so. (c) If A ⊂ G let J(a) = { f ∈ H(G) : Z( f ) ⊃ A}. Also if |z| -and hence |w|. (b) Assume that f is not the identically zero function and let K be a compact subset of G such that f (z) 0 for all z in K. 1). Use Theorem 5. Show that f |g iﬀ Z ⊂ Z(g).Exercise 8.d. J is a prime ideal if whenever f and g ∈ H(G) and f g ∈ J then either f ∈ J or g ∈ J. Exercise 10. give the largest such open. Exercise 11. f = g. Substitute w := z2 . Similar to the previous case one factor equals 0 if z is a 2nth root of −1.15 to show there is an analytic function f on D = {z : |z| < 1} which is not analytic on any open set G which properly contains D. If there is no n such that z is nth root of unity then no factor of the inﬁnite product equals zero and there is a subsequence {(1 − znk }k such that |1 − znk | > 1. a. where the zeros are again counted according to their multiplicity. (z − a)2 } then Z(S) = {a. (ii) f in J and g any function in H(G) implies f g is in J. The maximal region G such that the convergence is uniform on compact sets is again the open unit disk.6 this converges abson=1 lutely if and only if ∞ −zn converges absolutely. (b) If S ⊂ H(G) and S contains a non-zero function then f is a greatest common divisor of S if: (i) f |g for each g in S and (ii) whenever h|g for each g in S. From Calculus it is known that the convergence is also n=1 uniform for |z| ≤ r < 1 with r ∈ [0.5. Show that f (z) = f (z) and the convergence is uniform over K. h| f . Solution. but not absolutely according to Deﬁnition 5. So if S = {(z − a)3 (z − b). (a) If f and g ∈ H(G) then f divides g (in symbols. So Z((z − a)3 ) = {a. a}. J is a maximal ideal if J is a proper ideal and whenever L is an ideal with J ⊂ L then either L = J or L = H(G). Show that J(A) is a closed ideal in H(G) and J(A) = (0) 107 fn (z) fn (z) . so |z| ≤ sqrtr < 1. iﬀ Z( f ) = Z(S) and show that each non-empty subset of H(G) has a g. is an ideal iﬀ: (i) f and g in J implies a f + bg is in J for all complex numbers a and b. Solution.is greater than 1. G a region. A subset J of H(G).5.d. then one of the factors in the inﬁnite product becomes 0 and the product converges to 0.c. J is called a proper ideal if J (0) and J H(G). Not available. If z is an nth root of unity for a positive integer n. Suppose G is an open set and { fn } is a sequence in H(G) such that f (z) = in H(G). If S ⊂ H(G) then Z(S) = ∩{Z( f ) : f ∈ S}.c. Exercise 9.d. By deﬁnition and Corollary 5. a}. Solution.S. Prove that f = g. n=1 With the denseness of n∈N {z ∈ C : z is nth root of unity} in the unit sphere it follows that the maximal set G such that the convergence is uniform on compact subsets is the unit disk.

n. We know by the double-angle identity of sine sin(2z) = 2 sin(z) cos(z) (this is proved easily by z2 using the deﬁnition) or sin(2πz) = 2 sin(πz) cos(πz).. Exercise 12.6 Factorization of the sine function Exercise 1. Find an entire function f such that f (n + in) = 0 for every integer n (positive. Solution. Find an entire function f such that f (m + in) = 0 for all possible integers m. Since.e. Solution. (2n − 1)2 Exercise 2. Not available. Find the most elementary solution possible. cos(πz) = 2πz ∞ n=1 ∞ n=1 1− 4z2 (2n)2 ∞ n=1 ∞ n=1 1− 4z2 z2 = 2πz 1 − 2 cos(πz) (2n − 1)2 n n=1 ∞ ∞ z2 1− 2 n 1− ∞ n=1 4z2 z2 = 2πz 1 − 2 cos(πz). Exercise 13. we know sin(πz) = πz ∞ 1 − n2 . Hence 2πz ⇐⇒ Thus. 108 . (f) Give an example of an ideal which is not a prime ideal. Find a factorization for sinh z and cosh z. (e) Show that every maximal ideal in H(G) is a prime ideal. Show that cos πz = ∞ n=1 1− 4z2 (2n−1)2 .iﬀ A has a limit point in G. Give the most elementary example possible (i. negative or zero). Not available. (d) Let a ∈ G and J = J({a}). (2n − 1)2 n n=1 1− 4z2 . choose the pn to be as small as possible). Show that J is a maximal ideal. 7. Solution. Not available. we obtain n=1 sin(2πz) = 2 sin(πz) cos(πz) ⇐⇒ ⇐⇒ 2πz 2πz ∞ n=1 ∞ m=1 1− 1− 4z2 z2 = 2πz 1 − 2 cos(πz) n2 n n=1 4z2 (2m)2 ∞ m=1 ∞ 1− 4z2 z2 = 2πz 1 − 2 cos(πz) (2m − 1)2 n n=1 ∞ where the last statement follows by splitting the product into a product of the even and odd terms (rearrangement of the terms is allowed). Solution.

ez + e−z e−i(iz) + e−(−i)(iz) = 2 2 ei(iz) + e−i(iz) = cos(iz) 2 ∞ ∞ 4z2 4(iz)2 = 1+ 1− (2n − 1)2 π2 (2n − 1)2 π2 n=1 n=1 1− 4z2 ⇒ cos(z) = (2n − 1)2 ∞ n=1 ∞ n=1 1− 4z2 . Not available. 176 cos(πz) = Therefore. 2 n nπ n=1 ∞ n=1 ∞ 1+ z2 n2 π2 . Exercise 4. 175 Equation 6. Find a factorization of the function cos Solution. (2n − 1)2 π2 1+ πz 4 4z2 .Solution. (2n − 1)2 π2 − sin πz 4 Exercise 3. = (2n)2 ∞ n=1 (2n−1)(2n+1) . We know sinh(z) = ez − e−z e−i(iz) − e−(−i)(iz) = 2 2 −ei(iz) + e−i(iz) −ei(iz) + e−i(iz) =i = 2 2i ei(iz) − e−i(iz) = −i = −i sin(iz) 2i ∞ ∞ (iz)2 z2 = −i(iz) 1− 2 2 =z 1+ 2 2 nπ nπ n=1 n=1 since by p. π 2 .2 sin(πz) = πz Therefore. 109 . cosh(z) = ∞ n=1 ∞ n=1 1− z2 z2 ⇒ sin(z) = z 1− 2 2 . sinh(z) = z We have cosh(z) = = = since by Exercise 1 p. Not available. Prove Wallis’s formula: Solution.

7 The gamma function Exercise 1.57722. Show: √ 1 1 πΓ(2z) = 22z−1 Γ(z)Γ(z + 2 ). then Γ which implies π 1 1 Γ 1− = π csc 2 2 2 √ √ 1 = π· 1 = π. (An approximation to γ is .2. Exercise 2.) Solution. lim 1 z 1− z2 12 1− z2 22 1− z2 22 z2 32 1 1− ··· 1 − z2 32 z2 n2 n+1−z n 1− = πz ··· 1 − z2 lim n2 n→∞ n+1−z n =1 π ∞ n=1 π = π csc(πz) sin(πz) 1− z2 n2 1− z2 n2 where the step before the last step follows by p. Now. So Γ(z)Γ(1 − z) = π csc(πz) 1 for z not an integer. (Hint: Consider the function Γ(z)Γ(z + 2 )Γ(2z)−1 . Show that 0 < γ < 1. Assume z is not an integer. Deduce from this that Γ( 1 ) = 2 Solution. Not available. then by Gauss’s Formula p. It is unknown whether γ is rational or irrational. Show that Γ(z)Γ(1 − z) = π csc πz for z not an integer. Following the hint deﬁne a function f on its domain G = {z ∈ C : z 2z−1 z z 1 . let z = 2 . 2 Γ Exercise 3. 175 Equation 6. + Γ f (z) = √ Γ 2 2 π 2 110 .) (Z − N)} by Solution. 178 we get Γ(z)Γ(1 − z) = n!nz m!m1−z lim n→∞ z(z + 1) · · · (z + n) m→∞ (1 − z)(2 − z) · · · (m + 1 − z) n!n!nz+1−z = lim n→∞ z(1 + z)(1 − z)(2 + z)(2 − z) · · · (n + z)(n − z)(n + 1 − z) (n!)2 n = lim n→∞ z(12 − z2 )(22 − z2 ) · · · (n2 − z2 )(n + 1 − z) (n!)2 n = lim 2 2 z z2 z2 n→∞ z(n!n!) 1 − z 1 − 22 1 − 23 · · · 1 − n2 (n + 1 − z) 12 lim = = z limn→∞ 1 − = z = 1 ∞ n=1 z2 12 n→∞ √ π.7.

Write Γ(z) = Φ(z) + Ψ(z) where Φ(z) = 0 e−t tz−1 dt and Ψ(z) = 1 ∞ e−t tz−1 dt. . Show that Γ(z) = for z ∞ n=0 ∞ n=1 log 1 + z z − . and f limn→∞ n(z+n) = 1 for all z. Exercise 4. 0] and that log Γ(z) = − log z − γz − Solution. Show that log Γ(z) is deﬁned for z in C − (−∞. −2. With the Identity Theorem then f agrees with Γ on the whole domain G. Also f satisﬁes the functional equation of the Gamma-function: x x 1 2x Γ +1 + f (x + 1) = √ Γ 2 2 2 π 2x x 1 x x = √ Γ + Γ π 2 2 2 2 x x2 x−1 x 1 Γ + = √ Γ 2 2 2 π = x f (x). It suﬃces to show that f (x) satisﬁes the Bohr-Mollerup Theorem. f agrees with Γ on the positive real numbers. By the logarithmic convexity of Γ is follows that log f (x) = (x − 1) log 2 − log √ x x 1 π + log Γ + + log Γ 2 2 2 is a convex function.Let x be a positive real number. (not for Re z > 0 alone). Let f be analytic on the right half plane Re z > 0 and satisfy: f (1) = 1. Γ( 1 ) Lastly. (7. z f (n) Solution. 1 Solution. Not available. Exercise 5. Show that f = Γ. f (z + 1) = z f (z). . Exercise 6. Not available. n n (−1)n + n!(z + n) ∞ 1 e−t tz−1 dt 0.11) We can write 1 Φ(z) = 0 ∞ n=0 e−t tz−1 dt = 0 1 ∞ n=0 ∞ (−t)n z−1 t dt = n! (−1) n!(z + n) n ∞ n=0 (−1)n n! 1 tn tz−1 dt 0 = (−1) n! n t n+z n+z 1 0 = n=0 111 . also f (1) = √2π = 1. −1. .13. thus by Theorem 7.

Thus, Γ(z) =

∞ n=0

(−1)n + n!(z + n)

∞ 1

e−t tz−1 dt

(7.12)

Claim 1: Γ(z) given by (7.12) is the analytic continuation of (7.11), that is Γ(z) given by (7.12) is deﬁned for all z ∈ C − {0, −1, −2, . . .}. Proof of Claim 1: We know from the book that Ψ(z) is analytic for Re (z) > 0. Claim 2: Ψ(z) is analytic for Re (z) ≤ 0. Thus Ψ(z) is analytic on C. Proof of Claim 2: Assume Re (z) ≤ 0. Then |tz−1 | = tRe(z)−1 ≤ t−1 . t∈[1,∞),Re(z)≤0

1 1 But since e− 2 t tRe (z)−1 → 0 as t → ∞, there exists a constant C > 0 such that tRe(z)−1 ≤ Ce 2 when t ≥ 1. Hence, we have 1 1 |e−t tz−1 | ≤ |e−t |· |tz−1 | = e−t tRe(z)−1 ≤ e−t Ce 2 t = Ce− 2 t

**and therefore Ce− 2 t is integrable on (1, ∞). By Fubini’s Theorem for any {γ} ⊂ G = {z : Re(z) ≤ 0},
**

∞ γ 1

1

e−t tz−1 dt dz =

1

∞ γ

e−t tz−1 dz dt = 0

which implies

1

∞

e−t tz−1 dt ∈ H(G).

∞ 1

In summary, Ψ(z) = Thus, Claim 2 is proved. It remains to show that

e−t tz−1 dt ∈ H(C).

Φ(z) =

∞ n=0

(−1)n n!(z + n)

is analytic on C − {0, −1, −2, . . .}. Note that Φ(z) is uniformly and absolutely convergent as a series in any closed domain which contains none of the points 0, −1, −2, . . . and thus provides the analytic continuation of Φ(z). Since ∞ ∞ (−1)n + e−t tz−1 dt n!(z + n) 1 n=0 is analytic and we know (Theorem 7.15 p. 180) Γ(z) =

0 ∞

e−t tz−1 dt

for Re(z) > 0, we get Γ(z) =

∞ n=0

(−1)n + n!(z + n)

∞ 1

e−t tz−1 dt

is the analytic continuation of (7.11) for z ∈ C − {0, −1, −2, . . .}. 112

**Exercise 7. Show that
**

0

∞

sin(t2 ) dt =

0

∞

cos(t2 ) =

1 2

1 π. 2

Solution. We have shown that Γ

1 2

**√ = π (see Exercise 2 p. 185). Thus
**

∞ 0

√ 1 π=Γ = 2

e−t t− 2 dt = 2a

0

1

∞

e−a

2 2

t

dt

(because of Theorem 7.15 p. 180). The last step involves integrating a complex integral. Hence, we have √ ∞ π 2 2 . (7.13) e−a t dt = 2a 0 Let a =

∞ 0 1−i √ , 2 −

then a2 =

2

1−2i+i2 2

= −i and therefore

∞ 0 ∞ 0

e

1−i √ 2

t2

dt

= = = (7.13)

e−(−i)t dt =

0 2

2

∞

eit dt

∞ 0

2

cos(t ) + i sin(t2 ) dt =

cos(t2 ) dt + i

0

∞

sin(t2 ) dt 1 1 π+i 2 2 1 π. 2

√ √ π π 1 1+i = = √ 1−i 2 √2 2 (1 − i)(1 + i) 2

∞ 0 ∞ 0

1 1 π(1 + i) = 2 2

Thus,

∞ 0

cos(t2 ) dt + i

sin(t2 ) dt =

1 2 1 π 2 1 π. 2

1 1 π+i 2 2

1 π 2

which implies cos(t2 ) dt = 1 2 1 2

and

0

∞

sin(t2 ) =

Exercise 8. Let u > 0 and v > 0 and express Γ(u)Γ(v) as a double integral over the ﬁrst quadrant of the plane. By changing to polar coordinates show that

π 2

Γ(u)Γ(v) = 2Γ(u + v)

0

(cos θ)2u−1 (sin θ)2v−1 dθ.

The function B(u, v) = Γ(u)Γ(v) Γ(u + v)

**is called the beta function. By changes of variables show that
**

1

B(u, v) =

0

tu−1 (1 − t)v−1 dt =

∞ 0

tu−1 dt (1 + t)u+v

Can this be generalized to the case when u and v are complex numbers with positive real part? 113

**Solution. Let u > 0 and v > 0, then by Theorem 7.15 p. 180 and three changes of variables (indicated by “CoV”) we obtain Γ(u)Γ(v) =
**

0 ∞

e−s su−1 ds

0 ∞

2

∞

e−t tv−1 dt

∞ 0

=

CoV:s=x2 ,t=y2

2

0

**e−x x2u−2 x dx· 2
**

∞ 0

e−y y2v−2 y dy dx dy

2

= =

CoV:x=r cos(θ),y=r sin θ

4

0

∞ π/2

e−x

2

−y2 2u−1 2v−1

x

y

4

0 π/2 0

∞ ∞ 0

e−r r2u−1 (cos θ)2u−1 r2v−1 (sin θ)2v−1 r dr dθ e−r r2u+2v−1 (cos θ)2u−1 (sin θ)2v−1 r dr dθ

π/2

2

2

= = =

CoV:t=r2

4

0

4

0

∞

e

−r2 2u+2v−1

r

dr

0 π/2

**(cos θ)2u−1 (sin θ)2v−1 dθ (cos θ)2u−1 (sin θ)2v−1 dθ
**

0

1 4 2

∞ 0

e−t tu+v−1 dt

π/2

= This implies B(u, v) = Next, we show

2Γ(u + v)

0

(cos θ)2u−1 (sin θ)2v−1 dθ.

Γ(u)Γ(v) =2 Γ(u + v)

1

π/2

**(cos θ)2u−1 (sin θ)2v−1 dθ.
**

0

B(u, v) =

0

tu−1 (1 − t)v−1 dt.

We have B(u, v) = =

CoV:t=cos2 θ π/2 Γ(u)Γ(v) =2 (cos θ)2u−1 (sin θ)2v−1 dθ Γ(u + v) 0 1 √ 2v−2 2u−2 √ t 1−t dt 2 0 1

=

0

tu−1 (1 − t)v−1 dt.

1

Hence, B(u, v) =

0

tu−1 (1 − t)v−1 dt.

∞ 0

Finally, we show B(u, v) =

tu−1 dt. (1 + t)u+v

114

This implies |tu−1 (1 + t)−u−v | is integrable on (0. Thus it remains to show that B(u. u+v 0 (1 + s) ∞ 0 B(u. By Fubini’s Theorem for any {γ} ⊂ G × G 1 γ 0 tu−1 dt dz = (1 + t)u+v 1 0 γ tu−1 dz dt = 0 (1 + t)u+v 115 . v) = 0 ∞ tu−1 dt = 2 (1 + t)u+v π/2 (cos θ)2u−1 (sin θ)2v−1 dθ 0 for u > 0 and v > 0.We have 1 B(u. v) = tu−1 dt. v) = 0 ∞ tu−1 dt ∈ H(G × G) (1 + t)u+v where G × G ⊂ C × C with G = {z : Re(z) > 0}. (1 + t)u+v Yes. Let Re(u) > 0 and Re(v) > 0. 1). ∞ 0 tu−1 dt = (1 + t)u+v 1 0 tu−1 dt + (1 + t)u+v I ∞ 1 tu−1 dt . (1 + t)u+v II I) We have 1 0 1 |tu−1 (1 + t)−u−v | dt ≤ ≤ ≤ 0 1 0 1 0 |tu−1 |· |(1 + t)−u−v | dt tRe(u)−1 (1 + t)−Re(u)−Re(v) dt tRe(u)−1 < ∞ where the last step follows since Re(u) > 0 and the previous one since Re(u) > 0 and Re(v) > 0 and t ∈ (0. the beta function can be generalized to the case when u and v are complex numbers with positive real part. v) = 0 tu−1 (1 − t)v−1 dt ∞ = t CoV:s= 1−t 0 ∞ s 1+s su−1 u−1 1− s 1+s v−1 (1 + s)−2 ds = = = Hence. 1 1 (1 + s)−2 ds (1 + s)u−1 (1 + s)v−1 0 ∞ 1 su−1 (1 + s)−2 ds (1 + s)u+v−2 0 ∞ su−1 ds. We have seen that B(u. 1).

V2 (1) and V3 = 4 π.which implies 1 0 tu−1 dt ∈ H(G × G). V2 (r) = πr2 . V3 (r) = V1 (1) = 2. Let αn be the volume of the ball of radius one in Rn (n ≥ 1). ∞). 3 Claim: For arbitrary n we have Vn (r) = rn Vn (1) 116 4 3 3 πr and therefore . By Fubini’s Theorem for Re(u)+Re(v) ≤ 1. (1 + t)u+v In summary 0 ∞ Exercise 9. (1 + t)u+v n II) We have ∞ 1 |tu−1 (1 + t)−u−v | dt ≤ ≤ = = n→∞ lim 1 n |tu−1 (1 + t)−u−v | dt |tu−1 |· |(1 + t)−u−v | dt n n→∞ lim 1 n→∞ lim tRe(u)−1 (1 + t)−Re(u)−Re(v) dt tRe(u)−1 tRe(u)+Re(v) 1 t 1 n n→∞ lim 1 +1 Re(u)+Re(v) dt 1 n = n→∞ lim t−1−Re(u) 1 t 1 n +1 1 dt Re(u)+Re(v) ≤ n→∞ lim 1 t1+Re(v) 1 dt < ∞ where the last step follows since Re(v) > 0 and the previous one since Re(u) + Re(v) > 0. (1 + t)u+v tu−1 dt ∈ H(G × G). V1 (r) = 2r. Clearly. This implies |t (1 + t) 1 ( t +1) any {γ} ⊂ G × G ∞ γ 1 tu−1 dt dz = (1 + t)u+v ∞ 1 ∞ 1 γ tu−1 dz dt = 0 (1 + t)u+v which implies tu−1 dt ∈ H(G × G). Deﬁne the n-dimensional ball with radius r by Bn (r) = {x ∈ Rn : |x| ≤ r} and deﬁne the volume of Bn (r) by Vn (r). Prove by induction and iterated integrals that 1 αn = 2αn−1 0 (1 − t2 )(n−1)/2 dt Solution. t ≥ 1 and 1 u−1 −u−v therefore | is integrable on (1.

xn−1 . . Hence 1 αn = Vn (1) = −1 1 − t2 n−1 2 αn−1 dt = 2αn−1 0 1 − t2 n−1 2 dt which proves the formula above. we prove the claim Vn (r) = rn Vn (1). . x2 .which implies Vn (r) = rn Vn (1) = rn αn (def) (7. We have by using a symmetry argument √ 2 n−1 2 √2 2 √2 2 2 r Vn (r) = −r r = = CoV:xi =ryi 2n 0 √ 2 − r2 −x1 √2 0 r −x1 r −x1 √ 2 2 ··· − r2 −x1 −x2 √2 2 2 2 r −x1 −x2 0 r −x1 −x2 1 dxn dxn−1 · · · dx2 dx1 √ − r2 − n−1 xi2 i=1 √ 2 n−1 2 r − i=1 r − i=1 xi xi r 2 n 0 1 0 √ √ 1−y2 1 0 √ ··· 1−y2 −y2 1 2 √ 0 0 1 dxn dxn−1 · · · dx2 dx1 1· rn Jacobian dyn dyn−1 · · · dy2 dy1 1− n−1 2 i=1 yi ··· ··· = = rn rn Vn (1). 117 . We could have proved the formula by mathematical induction. . k ≥ 1. then we write 2 2 2 Bn (1) = {x ∈ Rn : |x| ≤ 1} = {(x1 . Not available. . For a ﬁxed t ∈ [−1. t) ∈ Rn : x1 + x2 + . . Clearly. . xn−1 ) ∈ Rn−1 : x1 + x2 + . (Note that this is kind of obvious. Using spherical coordinates simpliﬁes the computation if preferred. Finally. + xn−1 ≤ 1} and by (7.14) we obtain Vn−1 √ 1 − t2 = √ 1 − t2 n−1 Vn−1 (1) = 1 − t2 n−1 2 αn−1 . Show that if n = 2k. . the ball Bn (1) is the union of all disjoint balls Bn−1 1 √ 1 − t2 as t varies over [−1. then αn = πk /k! Solution. since we scale the unit ball in Rn by r in each dimension to obtain the ball with radius r in Rn ). . . −1 − 1−y2 1 − √ √ √ 1−y2 −y2 1 2 √ − 1− n−1 2 i=1 yi n−1 2 i=1 yi 1−y2 1 1−y2 −y2 2 1 √ 1− 1 dyn dyn−1 · · · dy2 dy1 This implies Vn (r) = rn Vn (1). Show that αn = πn/2 (n/2)Γ(n/2) where αn is deﬁned in problem 9. 1] we have the balls √ 2 2 2 Bn−1 1 − t2 = {(x1 . . . Exercise 10. . 1]. + xn−1 + t2 ≤ 1}. x2 .14) Assume the claim is true.

i. . 1 1 The function Γ 2 z has simple poles at the non-positive even integers. p. z (d) Show that Ψ(z) − Ψ(1 − z) = −π cot πz. (c) Show that Ψ(z + 1) − Ψ(z) = 1 . 192). (b) Show that Ψ(1) = −γ. Solution. The zeta-function has a simple pole at z = 1. Not available. 2 (7. 2 of the homework set 5. (e) State and prove a characterization of Ψ analogous to the Bohr-Mollerup Theorem. The 1 factors z. For negative even integers z = 2k. evaluated at 1 − z. + Γ 2 2 2 (7. −n) = −1 for n ≥ 0. in particular Γ 1 z 1 z 1 − + Γ = π csc π(1 − z) 2 2 2 2 2 √ 1 z z πΓ(z) = 2z−1 Γ . limz→0 zΓ 2 z ∈ C. −1. −k ∈ N the poles of Gamma coincide with the simple zeros of the zeta1 function. The function ξ is analytic on C − ({−2k : k ∈ N0 } ∪ {1}) as a product of analytic functions.e.8 The Riemann zeta function 1 Exercise 1. hence limz→2k ζ(z)Γ 2 z exists in C also. . The functional equation of ξ is related to the functional equation of ζ (8. For z = 0. hence limz→1 (z − 1)ζ(z) exists in C and the function ξ is well-deﬁned at z = 1. z − 1 and π− 2 z are entire.16) (7.17) 118 . 7. Therefore the function ξ is analytic at each point. The Gaussian psi function is deﬁned by Ψ(z) = Γ (z) Γ(z) (a) Show that Ψ is meromorphic in C with simple poles at z = 0. ζ(1 − z) = 2(2π)−z Γ(z)ζ(z) sin 1 π(1 − z) . 1 Solution.15) Furthermore the proof uses variants of problems 1. .3. Let ξ(z) = z(z − l)π− 2 z ζ(z)Γ( 2 z) and show that ξ is an entire function which satisﬁes the functional equation ξ(z) = ξ(1 − z).Exercise 11. and Res(Ψ.

.17 to prove that number of primes. therefore the sum converges absolutely. . Let n = pk1 pk2 · · · pkm m n=1 nz 1 2 be the factorization of n into a product of primes p1 .16) = ξ(z)π− 2 21−z z Γ 2 Γ 1 2 π(1 1 2 + z 2 = ξ(z)π 2 21−z Γ 1 Γ(z) z 2 Γ 1 2 1 2 + z 2 z 2 1 (7. . otherwise let µ(n) = 0. Not available. where d(n) is the number of divisors of n. where pn are the prime n=1 n numbers. where µ(n) is deﬁned as follows. Exercise 5.4. if k1 = .Then the following chain of equalities holds. Prove that ζ(z) = ∞ µ(n) for Re z > 1. Exercise 3. 119 . Not available. ∞ ϕ(n) n=1 nz ∞ σ(n) n=1 nz . . Seeking contradiction suppose that ∞ pn is ﬁnite. Prove that ζ 2 (z) = Solution. also converges absolutely. ξ(1 − z) = (1 − z)(−z)π− 2 + 2 ζ(1 − z)Γ 1 z 1 z − 2 2 1 z 1 z 1−z −z 1 (7. By Proposition 5. Not available. Exercise 4. p. 1 Exercise 6.17) = ξ(z)π 2 21−z 1 Γ + 2z−1 π 1 2 Γ 1 z + 2 2 = ξ(z). Solution. . Let µ(1) = 1. Euler’s Theorem asserts that for real values x > 1. Notice that this implies that there are an inﬁnite n 1 Solution. Use Theorem 8.15) = z(z − 1)π− 2 + 2 Γ − 2 π ζ(z)Γ(z) sin π(1 − z) 2 2 2 = ξ(z)π 2 − 1 1−z 2 Γ(z)Γ Γ 1 2 z 2 − z 2 sin 1 π(1 − z) 2 − z) sin 1 π(1 − z) 2 1 Γ(z) π csc (7. ∞ d(n) n=1 nz for Re z > 1. Exercise 2. The left hand side is unbounded as x 1 and so is the right hand side. 1 Now the argument goes as follows. where σ(n) is the sum of the divisors of n. where ϕ(n) is the number of integers less than n and Solution. Not available. p−1 = ∞. All summands n=1 are positive. But this implies that − ∞ n=1 ∞ n=1 log 1 − 1 pn log 1 − 1 1 = log = log ζ(1) < ∞ pn 1 − pn n=1 ∞ which gives the desired contradiction because the log(ζ(z)) has a simple pole at z = 1. Prove that ζ(z−1) = ζ(z) which are relatively prime to n. for Re z > 1. = km = 1 then let µ(n) = (−1)m .165. for Re z > 1. ζ(x) = ∞ 1−p−x . Prove that ζ(z)ζ(z − 1) = Solution. pm and suppose that these primes are distinct. .

(z) Exercise 7. Prove that ζζ(z) = − m ≥ 1; and Λ(n) = 0 otherwise.

∞ Λ(n) n=1 nz

for Re z > 1, where Λ(n) = log p if n = pm for some prime p and

Solution. Not available. Exercise 8. (a) Let η(z) = ζ (z)/ζ(z) for Re z > 1 and show that limz→z0 (z − z0 )η(z) is always an integer for Re z0 ≥ 1. Characterize the point z0 (in its relation to ζ) in terms of the sign of this integer. (b) Show that for > 0 Re η(1 + + it) = − where Λ(n) is deﬁned in Exercise 7. (c) Show that for all > 0, 3Re η(1 + ) + 4Re η(1 + + it) + Re η(1 + + 2it) ≤ 0. (d) Show that ζ(z) 0 if Re z = 1 (or 0).

∞

Λ(n)n−1(1+ ) cos(t log n)

n=1

Solution. Not available.

120

Chapter 8

Runge’s Theorem

8.1 Runge’s Theorem

Exercise 1. Prove Corollary 1.14 if it is only assumed that E − meets each component of C∞ − G. Solution. Not available. Exercise 3. Let G be the open unit disk B(0; 1) and let K = {z : 1 ≤ |z| ≤ 3 }. Show that there is a function 4 4 f analytic on some open subset G1 containing K which cannot be approximated on K by functions in H(G). Remarks. The next two problems are concerned with the following question. Given a compact set K contained in an open set G1 ⊂ G, can functions in H(G1 ) be approximated on K by functions in H(G)? Exercise 2 says that for an arbitrary choice of K, G, and G1 this is not true. Exercise 4 below gives criteria for a ﬁxed K and G such that this can be done for any G1 . Exercise 3 is a lemma which is useful in proving Exercise 4. Solution. Not available. Exercise 3. Let K be a compact subset of the open set G and suppose that any bounded component D of G − K has D− ∩ ∂G . Then every component of C∞ − K contains a component of C∞ − G. Solution. Not available. Exercise 4. Let K be a compact subset of the open set G; then the following are equivalent: (a) If f is analytic in a neighborhood of K and > 0 then there is a g in H(G) with | f (z) − g(z)| < for all z in K; (b) If D is a bounded component of G − K then D− ∩ ∂G ; (c) If z is any point in G − K then there is a function f in H(G) with | f (z)| > sup{| f (w)| : w in K}. Solution. Not available. ˆ Exercise 5. Can you interpret part (c) of Exercise 4 in terms of K? Solution. Not available.

121

ˆ Exercise 6. Let K be a compact subset of the region G and deﬁne KG = {z ∈ G : | f (z)| ≤ f K for all f in H(G)}. ˆ ˆ (a) Show that if C∞ − G is connected then KG = K. ˆ G , C − G). (b) Show that d(K, C − G) = d(K ˆ (c) Show that KG ⊂ the convex hull of K ≡ the intersection of all convex subsets of C which contain K. ˆ G ⊂ G1 ⊂ G and G1 is open then for every g in H(G1 ) and > 0 there is a function f in H(G) such (d) If K ˆ that | f (z) − g(z)| < for all z in KG . (Hint: see Exercise 4.) ˆ G = the union of K and all bounded components of G − K whose closure does not intersect ∂G. (e) K Solution. Not available.

8.2 Simple connectedness

Exercise 1. The set G = {reit : −∞ < t < 0 and 1 + et < r < 1 + 2et } is called a cornucopia. Show that G ˆ is simply connected. Let K = G− ; is intK connected? Solution. he result follows once it is established that every closed piecewise smooth curve in G is 0homotopic. This requires the following subclaim: let z ∈ G then the argument of z is uniquely deﬁned. Suppose otherwise then there are real r1 , r2 , t1 , t2 such that z = r1 eit1 = r2 eit2 . Suppose WLOG that t1 < t2 < 0 then there is a positive integer k such that t1 = 2kπ = t2 and therefore r2 > 1+et2 = 1+et1 ekπ > 1+2et1 > r1 contradicting the fact that |z| = r1 = r2 . Deﬁne 3 R : (−∞, 0) → G; R(t) = 1 + et . 2 Let γ : [0, 1] → G be an arbitrary closed piecewise smooth curve in G with radius function ρ(s) and argument function τ(s) such that γ(s) = ρ(s)eiτ(s) , and deﬁne Γ : [0, 1] × [0, 1] → G, Γ(s, 0) = γ(s), Γ(s, 1) = R(τ(s)), Γ(0, u) = Γ(1, u) Also deﬁne a function Θ : [0, 1] × [0, 1] → G, 1 2 , ∀ u ∈ [0, 1].

1 2

∀ u ∈ [0, 1].

Θ(s, 0) = R(τ(s)), Θ(s, 1) = R τ

Θ(0, u) = Θ(1, u)

The function Θ ◦ Γ satisﬁes Deﬁnition IV 6.1 (p.88), the curve γ is homotopic to the point R τ Deﬁnition IV 6.14 (p.93) the set G is simply connected.

and by

Exercise 2. If K is polynomially convex, show that the components of the interior of K are simply connected. Solution. Not available. 122

1)). and z = bm is a zero of multiplicity pm . Let S n (z) be a singular part at an and let pm be a positive integer. rn (z) = j=1 B jk (z − an )− j−k (8. that is gn (z) = where αk = sn (z) or 1 (k) k! g (an ). and let {pm } be a sequence of integers. Let G be a region and let {bm } be a sequence of distinct points in G with no limit point in G.3 Mittag-Leﬄer’s Theorem Exercise 1. 123 .8. m. (Note that the zeros do not cancel the poles since by assumption an bm ∀ n. g has a Taylor series in a neighborhood B(an . Since g ∈ H(g) and {an } ∈ G. By Mittag-Leﬄer’s Theorem. Solution. Let {an } be a sequence of distinct points without a limit point in G and such that an bm for all n. Set f = g· h. ∞ k=0 ∞ k=0 αk (z − an )k ∈ B(an . furthermore.170 there is an analytic function g deﬁned on G whose only zeros are at the points bm . Rn ) of each an .15 p. Then by construction f is the meromorphic function on G whose only poles and zeros are {an } and {bm } respectively. and z = bm is a zero of multiplicity pm . the singular part at z = an is S n (z). Proof of the claim: ∞ k=0 αk rn (z)(z − an )k = ∞ k=0 mn αk j=1 mn B jk (z − an )− j−k (z − an )k B jk (z − an )− j ∞ k=0 mn = ∞ k=0 mn αk j=1 = j=1 (z − an )− j αk B jk = j=1 A jn (z − an ) j where the last step follows by choosing ∞ k=0 αk B jk = A jn . By Theorem 5.1) works. Let G be a region and let {an } and {bm } be two sequences of distinct points in G without limit points in G such that an an bm for all n. the singular part at z = an is S n (z). G is open. there is a meromorphic function h on G whose poles are exactly the points {an } and such that the singular part of h at an is rn (z). Rn ) Goal: Try to use this series to create a singular part rn (z) at an such that rn (z)gn (z) = mn rn (z) Claim: αk (z − an )k = j=1 A jn ⇐⇒ (z − an ) j mn mn j=1 A jn = (z − an ) j ∞ k=0 αk rn (z)(z − an )k . Show that there is a meromorphic function f on G whose only poles and zeros are {an } and {bm } respectively. m. Let {rn (z)} be the sequence of rational functions given by mn rn (z) = j=1 B jk (z − an ) j+k (see (8. bm is a zero of g of multiplicity pm . Since G is a region. m).

(b) Show that if lim sup |bn | < ∞ then (3. . (d) Suppose there is an r > 0 such that |an − am | ≥ r for all n m. but not w = 0. In particular. Show that |an |−3 < ∞. + a n n n n n=1 where the sum is over the same w as in part (f).) (e) Show that if the series (3.4) converges absolutely for all r > 0 then ∞ k=n r an kn bn z − an (3. n = 0. is convergent in M(C) to a meromorphic function ζ with simple poles at the points 2nω + 2n ω . . P is called the Weierstrass pe function. . P is doubly periodic with periods 2ω and 2ω . (This is somewhat involved and the reader may prefer to prove part (f) directly since this is the only application.5) converges in M(C) to a meromorphic function f then ∞ k −1 bn z n z bn f (z) = z − a + a 1 + a + . That is.Exercise 2. Show that P(z) = 1 + z2 1 1 − 2 2 (z − w) w converges absolutely.4) converges absolutely if kn = k for all n. if the sequence {bn } is bounded then the series (3. then (3. Not available.4) converges absolutely if kn = n for all n. Also show that P(z) = P(z + 2nω + 2n ω ) for all integers n and n . (a) Show that if integers {kn } can be chosen such that ∞ k=n r an kn bn an (3. (g) Let P(z) = −ζ (z). Solution. ±1. This function is called the Weierstrass zeta function.5) converges in M(C) to a function f with poles at each point z = an . (c) Show that if there is an integer k such that the series ∞ n=1 bn ak+1 n (3.6) with k = 2 converges absolutely.6) (f) Let ω and ω be two complex numbers such that Im(ω /ω) 0. . . Let {an } be a sequence of points in the plane such that |an | → ∞. ±2. 124 . and let {bn } be an arbitrary sequence of complex numbers. Using the previous parts of this exercise show that the series 1 1 z 1 . + + ζ(z) = + z z − w w w2 where the sum is over all w = 2nω + 2n ω for n.

is a meromorphic function on C with simple poles at a1 . Show that if γ1 and γ2 are any rectiﬁable curves in C − {a1 . . (a) Deﬁne two functions f and g in H(G) to be relatively prime (in symbols. If S ⊂ H(G) then let J = ∩{S : S is an ideal of H(G) and S ⊂ J}. fk ∈ S for 1 ≤ k ≤ n}. then there is an integer m such that γ1 Exercise 3. fn }. a2 . g) = 1. (a) Deduce from Exercises 2(a) and 2(b) that for any sequence {an } in C with lim an = ∞ and an 0 there is a sequence of integers {kn } such that ∞ k −1 1 1 z 1 z n 1 h(z) = z − a + a + a a + . . Prove that for z curve in C − {a1 . + a a n n n n n n n=1 h− h = 2πim.) (c) Let f1 . fn ∈ H(G) and g = g. show that z is a removable singularity of the function f deﬁned in part (c).}.. f (z) = exp γ a1 . . . . a2 . a2 .} with f / f = h.7) converges in H(C) and it could hardly be classiﬁed as a new proof. . If S is ﬁnite then (S) is called a ﬁnitely generated ideal. . and (d) and gone directly from (a) to (e). . Furthermore. (c). . . This function f will then have the appropriate zeros. (That is. Prove that J is the smallest ideal in H(G) that contains S and J = {ϕ1 f1 + . This exercise shows how to deduce Weierstrass’s Theorem for the plane (Theorem VII. . show that f (z) = 0 and that the multiplicity of this zero equals the number of times that z appears in the sequence {a1 . show that J = α Jα is also an ideal. The remainder of the proof consists of showing that there is a function f such that h = f / f . . a2 . + ϕn fn : ϕk ∈ H(G). .}. ( f. a2 . The steps outlined in parts (a) through (d) give a proof of Weierstrass’s Theorem without introducing inﬁnite products. (d) Suppose that z ∈ {a1 . . .c.. . . . (e) Show that ∞ 2 kn z z + 1 z + . . Show that there are functions ϕ1 . and γ any rectiﬁable h deﬁnes an analytic function on C − {a1 . . . ϕn in H(G) such that g = ϕ1 f1 + . 5. (Hint: Use (b) and induction. . a2 . (e) Show that every ﬁnitely generated ideal in H(G) is a principal ideal. . . Not available. g1 in H(G) such that f f1 + gg1 = 1 (Hint: Show that there is a meromorphic function ϕ on G such that f1 = ϕg ∈ H(G) and g|(1 − f f1 ). . Exercise 4. . .}.) (d) If {Jα } is a collection of ideals in H(G). J is called the ideal generated by S and is denoted by J = (S). 125 .12) from Mittag-Leﬄer’s Theorem. g) = 1) if the only common divisors of f and g are non-vanishing functions in H(G).}. Show that ( f. .d { f1 . . . . . . We could have skipped parts (b). + 1 z (3. . + ϕn fn .11. . . Solution. . If S = { f } for a single function f then ( f ) is called a principal ideal. . the value of f (z) is independent of the curve γ and the resulting function f is analytic. (b) If ( f.. (b) Let z be an arbitrary but ﬁxed point in C − {a1 . This exercise assumes a knowledge of the terminology and results of Exercise VII. . . a2 . a2 . . However this would have meant that we must show that (3. γ2 (c) Again let h be the meromorphic function from part (a).} from 0 to z and h is the function obtained in part (a). 5.7) f (z) = exp 1− a an 2 an 2 an n n=1 Remark. . show that there are functions f1 . g) = 1 iﬀ Z( f ) ∩ Z(g) = .

3.(f) An ideal J is called a ﬁxed ideal if Z(J) . Exercise 5. f has poles of order 2 at z = n. (k) For each integer n ≥ 1 choose integers kn ≥ 0 and constants An . √ √ 2 √ 2 ∞ 3 nz2 − 2z3 lim (z − m) f (z) = lim (z − m) √ √ √ z→ m z→ m n3/2 (z − n)2 n=1 = lim √ √ √ m)(3 nz2 − 2z3 ) 3m3/2 − 2m3/2 + √ m3/2 z→ m n3/2 (z − n)2 n=1 √ √ ∞ (z − n)(3 nz2 − 2z3 ) + lim √ √ z→ m n3/2 (z − n)2 m+1 m−1 (z − = 0 + 1 + 0 = 1. Not available. (g) Let fn (z) = sin(2−n z) for all n ≥ 0 and let J = ({ f1 . Exercise 6. such that the residue at This is the majorant independent of z for the Weierstrass’ Criterion. Not available. Show that J is a ﬁxed ideal in H(C) which is not a principal ideal. √ each pole is 0 and lim(z − n)2 f (z) = 1 for all n. (j) Let {an } be a sequence of distinct points in G with no limit point in G. Then for |z| ≤ r and n ∈ N √ √ √ 3 nz2 − 2z3 r2 (3 n + n) 16r2 | 3/2 ≤ 2 =: Mn . . . . Z(S) . Choose the S n so that f = gh has the desired property. Solution. Let h be a meromorphic function on G with poles at each an of order kn + 1 and with singular part S n (z). Prove that if J = (S) then Z(J) = Z(S) and that a proper principal ideal is ﬁxed. Show that J is a proper free ideal in H(G). We claim that the function √ 3 nz2 − 2z3 √ n3/2 (z − n)2 k=1 √ has the desired properties. Show that there is an analytic (k) (k) function f on G such that f (an ) = k!An . Find a meromorphic function with poles of order 2 at 1. √ 2 | ≤ 3/2 √ n n (z − n) n ( n − r)2 f (z) = ∞ √ √ 2. Use this to show that J can contain no polynomials. Also show that J = ( f ) if J is ﬁnitely generated. Let J = { f ∈ H(G) : f (an ) = 0 for all but a ﬁnite number of the an }. (i) Let M be a maximal ideal that is ﬁxed. . 126 . . Moreover f converges in M(C). Show that there is a point a in G such that M = ((z − a)). (k) If J is a free ideal show that for any ﬁnite subset S of J.) Solution.}). otherwise it is called a free ideal. To see this let r > 0 and choose √ N ∈ N so large that n > 2r for all n ≥ N. 0 ≤ k ≤ kn . Solution. (h) Let J be a ﬁxed ideal and prove that there is an f in H(G) with Z( f ) = Z(J) and J ⊂ ( f ). (Hint: Let g be an analytic function on G with a zero at an of multiplicity kn + 1. l) Let J be a proper free ideal. n ∈ N (and no other poles) and it is analytic for every z ∈ C that is not one of the poles. then J is a maximal ideal iﬀ whenever g ∈ H(G) and Z(g) ∩ Z( f ) for all f in J then g ∈ J. . f2 . Let G be a region and let {an } be a sequence of distinct points in G with no limit point in G. √ Next ﬁx m ∈ N and consider the limit in equation limz→ √n (z − n)2 f (z) = 1.

By Proposition V 2. = √ n3/2 n3/2 (z − m) z=m z=m √ The residuum vanishes for all m.4.113. in the case n = m. (z − m)2 cancels directly and we are √ z− m · 3/2 √ n (z − n)3 127 . First ﬁnd the derivative of f as √ ∞ 6 nz2 − 6nz − 2z3 f (z) = √ n3/2 (z − m)3 n=1 √ then compute the residuum at z = m √ √ √ Res( f. m ∈ N and therefore the function considered in this exercise has all the required properties. m) = g (z)|z= √m = 2(z − m) f (z) + (z − m)2 f (z) z= √m = = ∞ n=1 2(z − ∞ n=1 + ∞ n=1 √ √ m)(3 nz2 − 2z3 ) √ n3/2 (z − n)2 √ √ (z − m)2 (6 nz2 − 6nz − 2z3 ) √ 3 3/2 (z − n m) √ z= m 2(z − √ √ √ (n))(3 nz2 − 2z3 ) + (z − m)(6 nz2 ) − 6nz − 2z3 ) √ z= m √ √ 12 m − 2z3 − 6mz −6z(z − m) = 0. p. If n m and z = left with √ √ m. for a function √ g(z) := (z − m)2 f (z).Lastly we have to check that the residuals are 0 at the poles. the last factor equals zero. it has to be veriﬁed that g (z)|z= √m = 0.

z) = 0 if z G− . Of course. (c) Suppose that G = G∗ and f is an analytic function deﬁned on G such that f (z) is real for z in G with |z| = 1. G− . The fact that a simple closed curve divides the plane into two pieces (an inside and an outside) is the content of the Jordan Curve Theorem. Not available. Show that γ f = 0. G0 be as in the statement of the Schwarz Reﬂection Principle and let f : G+ ∪ G0 → C∞ be a continuous function such that f is meromorphic on G+ . (d) Formulate and prove a version of the Schwarz Reﬂection Principle where the circle |ξ| = 1 replaces R. z] intersects {γ} only at z. i.Chapter 9 Analytic Continuation and Riemann Surfaces 9. Not available. (a) Show that G is a region and G− = G ∪ {γ}. Show that f = f ∗ . z) = ±1 if z is inside γ and n(γ. 3.17. deﬁne f ∗ : G∗ → C by f ∗ (z) = f (1/¯). It is not necessary to assume that γ has such a point a as above.e. Let G. Exercise 2. z z (b) If f : G → C is analytic. Let γ be a simple closed rectiﬁable curve with the property that there is a point a such that for all z on γ the line segment [a. G+ . [a. Let G be a region in the plane that does not contain zero and let G∗ be the set of all points z such that there is a point w in G where z and w are symmetric with respect to the circle |ξ| = 1. each part of this exercise remains true if γ is only assumed to be a simple closed rectiﬁable curve. Show that there is a meromorphic function g : G → C∞ such that g(z) = f (z) for z in G+ ∪ G0 . z] ∩ {γ} = {z}. Solution. Do the same thing for an arbitrary circle. This is a very deep result of topology.1 Schwarz Reﬂection Principle Exercise 1. we must deﬁne what is meant by the inside of γ. (c) Show that n(γ. Also suppose that for x in G0 f (x) ∈ R. Solution. Exercise 3.) (a) Show that G∗ = {z : (1/¯) ∈ G}. (See III. Remarks. Show that f ∗ is analytic. Deﬁne a point w to be inside γ if [a. w] ∩ {γ} = and let G be the collection of all points that are inside γ. This is diﬃcult to obtain. (b) Let f : G− → C be a continuous function such that f is analytic on G. Is it possible to allow f to assume the value ∞ on G0 ? 128 .

Exercise 4. Dt ) : 0 ≤ t ≤ 1} along γ of ( f0 . then {( f j . Show that h( ft (z)) = z for all z in Dt and for all t. . D j ) : 0 ≤ j ≤ n} is called an analytic continuation along a chain of disks. j=0 Solution. and let γ be a path from 0 to b. and suppose there is an analytic function h : G → C such that h( f0 (z)) = z for all z in D0 . Bt ) : 0 ≤ t ≤ 1} of ( f0 . Suppose G is a region such that ft (Dt ) ⊂ G for all t. . Let γ(t) = exp(2πit) and σ(t) = exp(4πit) for 0 ≤ t ≤ 1. . What can be said about ( f0 .2 Analytic Continuation Along a Path Exercise 1. D0 ) along γ and show that [ f1 ]1 = [− f0 ]1 . 1) and let f0 be the restriction to D0 of the principal branch of the logarithm. . 1] → C is a closed path with γ(0) = γ(1) = a and let {( ft . Dn } is a chain of disks and f j−1 (z) = f j (z) for z in D j−1 ∩ D j . let {( ft . Not available. Not available. Find a continuation {( ft . Not available. D0 )? Solution. Dt ) : 0 ≤ t ≤ 1} is also a continuation along γ. Dt ) : 0 ≤ t ≤ 1} be an analytic continuation along a path γ : [0. . Exercise 6. j=0 (b) Conversely. b = γ(1). D0 ) along a chain of disks. (b) Find an analytic continuation {(gt . [ f0 ]a = [g0 ]a and [ fn ]b = [g1 ]b . Let D0 = B(1. D0 ) and show that [ f1 ]1 = [ f0 + 2πin]. Exercise 5. D0 ) along σ and show that [g1 ]1 = [g0 ]1 . Solution. 1) and let f0 be the restriction of the principal branch of z to D0 . . D j ) : 0 ≤ j ≤ n} be an analytic continuation along a chain of disks and let a and b be the centers of the disks D0 and Dn respectively. Let γ : [0. Dn ) is obtained by an analytic continuation of ( f0 . 129 .Solution. Not available. Show that there is an analytic continuation along a chain of disks {(g j . D0 = B(0. Let γ : [0. Dt ) : 0 ≤ t ≤ 1} be an analytic continuation along γ. 1). D j ) : 0 ≤ j ≤ n} is a collection of function elements such that {D0 . Not available. 1] → C be a path and let {( ft . D1 . We say that ( fn . Not available. If {( f j . it is actually an exercise in the use of the terminology. Show that there is a path γ from a to b and an analytic continuation {(gt . Show that if {( ft . Solution. Dt ) : 0 ≤ t ≤ 1} of ( f0 . The collection {D0 . . Dt ) : 0 ≤ t ≤ 1} is a continuation of ( f. . 9. √ Exercise 2. 1] → C and let a = γ(0). (a) Find an analytic continuation {( ft . Suppose γ : [0.) Solution. Exercise 7. Show that {( ft . For an integer n let γ(t) = exp(2πint). Hint: Show that T = {t : h( ft (z)) = z for all z in Dt } is both open and closed in [0. Dn } of open disks is called a chain of disks if D j−1 ∩ D j for 1 ≤ j ≤ n. Solution. Let f be an entire function. 1]. Dt ) : 0 ≤ t ≤ 1} be an analytic continuation along γ. Bt )} along γ such that {γ} ⊂ n D j . 1 ≤ j ≤ n. Dt ) : 0 ≤ t ≤ 1} be an analytic continuation along γ such that [ f1 ]a = [ f0 ]a and f0 0. Let D0 = B(1. Not available. (a) Let {( f j . D1 . 0 ≤ t ≤ 1. 1] → C be a path and let {( ft . [ f0 ]a = [g0 ]a and [ f1 ]b = [gn ]b . Exercise 3. D0 ) along γ then f1 (z) = f (z) for all z in D1 (This exercise is rather easy. B j ) : 0 ≤ j ≤ n} such that {γ} ⊂ n B j .

Exercise 4. Solution.2 is closed. Show that if f : X → Ω is a continuous function then the restriction of f to Y is a continuous function of (Y. Exercise 8. S). and γ(t) = 1 + at for 0 ≤ t ≤ 1 and a > 0. 130 . 1] × [0. Not available. Solution. Not available. u) and suppose that {( ft. TY ) into (Ω.u .Solution. Let γu (t) = Γ(t. Not available. Solution. Bt )} along γ is given with [g0 ]a = [ f ]a and r(t) is the radius of convergence of the power series expansion of gt about z = γ(t) then r(t) = R(t) for all t. Show that either R(t. Not available. (a) Show that R(t) is independent of the choice of continuation. 1]. Find R(t). 1). Exercise 3. Let R(t. u) = a. (c) Let ( f. Solution. let R(t) be the radius of convergence of the power series expansion of ft at z = γ(t). Not available. u) be the radius of convergence of the power series expansion of ft. Dt ) : 0 ≤ t ≤ 1} is an analytic continuation of ( f. Prove the propositions which were stated in this section without proof. ∞) is a continuous function. (b) Suppose that D = B(1. S) be topological spaces and let Y ⊂ X. Dt ) : 0 ≤ t ≤ 1} is an analytic continuation of f0 (z) = log z. T ) and (Ω. Let γ : [0. (d) For each of the functions R(t) obtained in parts (b) and (c).4 Topological Spaces and Neighborhood Systems Exercise 1. 1] → C be a path with γ(0) = 1 and γ(t) 0 for any t.u ) : 0 ≤ t ≤ 1} is an analytic continuation along γu such that [ f0. Γ(1. Solution. Solution.3 Monodromy Theorem Exercise 1. Let ( f. Let Γ : [0. 9. If γ : [0. let 0 < a < 1 and let γ(t) = (1 − at) exp(2πit) for 0 ≤ t ≤ 1. Exercise 2. Show that each ft is a branch of the logarithm. u).u ]a = [ f0. f is the restriction of the principal branch of the logarithm to D.v ]a for all u and v in [0. Suppose that {( ft . Let (X. Prove that the set T deﬁned in the proof of Lemma 3. Solution. u) = b for all u. D) be as in part (b). if a second continuation {(gt . Dt. 1] → C be a continuous function such that Γ(0.u about z = Γ(t. 1] → C is a path with γ(0) = a and γ(1) = b and {( ft . D) along γ. Not available. Not available. That is. Find R(t). 1] × [0. 1] → (0. ﬁnd min{R(t) : 0 ≤ t ≤ 1} as a function of a and examine the behavior of this function as a → ∞ or a → 0. Use Exercise 3 to give a second proof of the Monodromy Theorem. Exercise 2. u) ≡ ∞ or R : [0. D) be a function element and let a ∈ D. Not available. 9.

Exercise 4. Not available. Solution. Let Y ⊂ X and deﬁne UY = {Y ∩ U : U ∈ NY } for each y in Y. Not available. (b) If U ∈ T and ∆ ∈ S. ez ) : z ∈ G} considered as a subset of C × C.) State and prove an analogous result for branches of z1/n . Exercise 5. [ f ]z ) = ( f (z). (c) Deﬁne p1 : X × Ω → X and p2 : X × Ω → X by p1 (x. Adopt the notation of Exercise 3. Exercise 2. if ( f. (Use the map h : R → Γ deﬁned by h(z. Solution. ω) in X × Ω let U(x. ∆ ∈ Mω } (a) Show that {U(x. Not available. R) is a topological space show that a function f : (Z. Solution. Not available. Adopt the notation of Exercise 3.Exercise 3. ρ) be the Riemann surface of F (so that G is the base of R). (b) Let X = Ω = C and let Nz = Mz = {B(z. Let X and Ω be sets and let {N x : x ∈ X} and {Mω : ω ∈ Ω} be neighborhood systems and let T and S be the induced topologies on X and Ω respectively. Interpret part (a) of this exercise for this particular situation. If D is an open subset of G and f : D → C is a branch of the logarithm show that [ f ]a ∈ F for all a in D. Show that {UY : y ∈ Y} is a neighborhood system for Y and the topology it induces on Y is TY . R) → (X × Ω. Show that a function f : X → Ω is open iﬀ for each x in X and U in N x there is a set ∆ in Mω (where ω = f (x)) such that ∆ ⊂ f (U). Deﬁne F : S(C) → C by F(z. ω) ∈ X × Ω} is a neighborhood system on X × Ω and let P be the induced topology on X × Ω.ω) : (x. let F be the complete analytic function obtained from the principal branch of the logarithm. 131 . Show that p1 and p2 are open continuous maps. ) : > 0} for each z in C. (Hint: Use Exercise 2.ω) = {U × ∆ : U ∈ N x . Solution. Prove that a set is in P iﬀ it is the union of open rectangles. call the set U × ∆ an open rectangle. Solution.5 The Sheaf of Germs of Analytic Functions on an Open Set Exercise 1. for each ∆ in Mω there is a U in N x such that f (U) ⊂ ∆. show that f : D → C is a branch of the logarithm. Exercise 3. P) is continuous iﬀ p1 ◦ f : Z → X and p2 ◦ f : Z → Ω are continuous.) Solution. Conversely.8. 9. Let F be the complete analytic function obtained from the principal branch of the logarithm and let G = C − {0}. ω) = ω. Adopt the notation of Exercise 3. (a) Show that a function f : X → Ω is continuous iﬀ when x ∈ X and ω = f (x). Let G = C − {0}. Not available. and let (R. z) and use Exercise 2. For each point (x. Show that R is homeomorphic to the graph Γ = {(z. [ f ]z ) = f (z) and show that F is continuous. Exercise 6. ω) = x and p2 (x. Furthermore if (Z. D) is a function element such that [ f ]a ∈ F for some a in D. Not available.

Exercise 2. and Im (ω/ω ) 0.) (b) If ω. are linearly independent. ζ = 1. deﬁne σ. then T is a torus.) If ω and ω are complex numbers such that Im (ω/ω ) 0 then ω and ω . let l be the principal branch of the logarithm 1 1 deﬁned on B. t. Φ) be an analytic manifold. x3 ) ∈ R3 : x1 + x2 + x3 = 1 or x1 + x2 > 1 and x3 = 0}. If (U. Let G = {tω + t ω : 0 < t < 1. Show that an analytic manifold is locally compact. ψ). Φσ ) is analytic is to say that Φτ and Φσ are equivalent structures.6 Analytic Manifolds Exercise 1. Not available. Which of the following are analytic manifolds? What is its analytic structure if it is a manifold? 2 2 2 2 2 (a) A cone in R3 . prove that if a ∈ X and U is an open neighborhood of a then there is an open neighborhood V of a such that V − ⊂ U and V − is compact. Exercise 5. τ. Prove that (Ω. Not available. and if the identity map of (T. 0 < s < 1}. (For an example of a meromorphic function with two independent periods see Exercise VIII. Exercise 4. The following is a generalization of Proposition 6. e2πit ). [l] 2 ) and 1 −1 ( 2 . Let T = {z : |z| = 1} × {z : |z| = 1}. e2πit ). (Hint: Use the Cauchy-Riemann equations. Let (X. t in R.) (b) Prove that there is no non-constant entire function with two independent periods. G. Ψ) is an analytic manifold and h is an analytic function from X to Ω. 4. 0 < t < 1} and Ω = {sζ + s ζ : 0 < s < 1. let B = {z : |z − 1| < 1}. Solution. if x ∈ X there is an open set U such that x ∈ U and h is one-one on U). deﬁne σ(sζ + s ζ ) = (e2πis . That is.2(g). Not available.Solution. ζ = α where Im α 0. (Both G and Ω are the interiors of parallelograms. 9. Φτ ) into (T.) (c) Let ω = 1. f (z) = f (z + nω + n ω ) for all z in C and all integers n and n . (Use Exercise 3. Solution. [l1 ] 1 ) belong to the same component of ρ (D). Exercise 4. (b) {(x1 . 132 . [l] 2 ) and ( 2 . Ω and f as in part (b). Φσ ) is analytic then show that the function f : G → Ω deﬁned by f = σ−1 ◦ τ is analytic. Let Ψ be the collection of all such pairs (∆. Not available. (a) Let f be a meromorphic function deﬁned on C and suppose f has two independent periods ω and ω . (a) Let D = {z : |z| < 1} and show that ( 2 . and suppose there is a continuous function h of X onto Ω that is locally one-one (that is. Exercise 3. ω and ζ. ζ are two pairs of complex numbers such that Im (ω/ω ) 0 and Im (ζ/ζ ) 0. [l1 ] 2 ). Φτ ) into (T. x2 . Not available.) If Φτ and Φσ are the analytic structures induced on T by τ and σ respectively. Show that Φτ and Φσ are equivalent analytic structures if and only if α = i. and let l1 (z) = l(z) + 2πi for all z in B. (To say that the identity map of (T. considered as elements of the vector space C over R. Deﬁne h : C → T by h(tω + tω ) = (e2πit .3(b). Show that h induces an analytic structure on T . e2πis ) and τ(tω+t ω ) = (e2πit . So each z in C can be uniquely represented as z = tω + t ω . Not available. Solution.) (d) Can you generalize part (c)? Conjecture a generalization? Solution. (b) Find two disjoint open subsets of S(C) each of which 2 1 1 1 1 contains one of the points ( 2 . Consider the sheaf S(C). show that both σ and τ are one-one on G and Ω respectively. That is. (This torus is homeomorphic to the usual hollow doughnut in R3 . Solution. Using the notation of Exercise 4(a) show that there is an analytic function F : T → C∞ such that f = F ◦ h. ω = i. ϕ) ∈ Φ and h is one-one on U let ∆ = h(U) and let ψ : ∆ → C be deﬁned by ψ(ω) = ϕ ◦ (h/U)−1 (ω). let Ω be a topological space.

Show that there are polynomials p0 . (c) If f is one-one. Solution. . ρ × σ) is a covering space of Ω × Λ. Show that each component of ρ−1 (ω) consists of a single point and ρ−1 (ω)) has no limit points in X. Let (X. Not available. prove that (X. Solution. an be the points in C where f takes on the value ∞. (a) Show that either f ≡ ∞ or f −1 (∞) is a ﬁnite set. Not available. ρ) is a covering space of Ω and (Ω. Exercise 7. pn such that n 1 pk f (z) = p0 (z) + z − ak k=1 for z in C. p1 . Not available.) 133 .7 Covering spaces Exercise 1. b in C) or f (z) = Solution. c in C). Exercise 2. show that either f (z) = az + b (some a. . Not available. Show that there is an analytic structure Φ on X such that ρ is an analytic function from (X. Not available. Let (Ω. Exercise 8. Exercise 6. 9. y) = (ρ(x). . (b) If f ∞. . Exercise 3. Solution. Solution. √ z at the end of this section. Show that an analytic surface is arcwise connected. show that ρ−1 (ω1 ) and ρ−1 (ω2 ) have the same cardinality. Not available. Let (X. Not available. a z−c + b (some a. (Hint: Let γ be a path in Ω from ω1 to ω2 . Φ) to (Ω. Show that f is a ˜ ˜ ˜ one-one map of ρ−1 (ω1 ) onto ρ−1 (ω2 ). σ) be covering spaces of Ω and Λ respectively. Give a discussion for f 2 2 √ similar to the discussion of z at the end of this section. Deﬁne ρ × σ : X × Y → Ω × Λ by (ρ × σ)(x. Exercise 5. let a1 . ψ) be an analytic manifold and let (X. . . Not available. Exercise 9. ρ) be a covering space of Ω and let ω ∈ Ω. σ(y)) and show that (X × Y.Solution. Exercise 4. Furnish the details of the discussion of the surface for Solution. ρ) be a covering space of Ω. Solution. Suppose that f : C∞ → C∞ is an analytic function. Not available. . ρ) be a covering space of Ω. if x1 ∈ ρ−1 (ω1 ) and γ is the lifting of γ with initial point x1 = γ(0). π) is a covering space of Y. Suppose that (X. π ◦ ρ) is also a covering space of Y. . b. ρ) and (Y. If ω1 and ω2 are points in Ω. ψ). let f (x1 ) = γ(1). Let G = z : − π < Re z < π and deﬁne f : G → C by f (z) = sin z. Solution. Let Ω be a pathwise connected space and let (X.

Exercise 6. Exercise 10. Not available. f2 ) and Ω be as in part (b) and. Not available. f1 ). show that if Ω is simply connected then f is one-one. In this exercise all spaces are regions in the plane. Solution. apply the Monodromy Theorem. (G2 . Solution. Let G and Ω be regions in the plane and suppose that f : G → Ω is an analytic function such that (G. 134 . (c) Let (G1 . Show that if G1 is simply connected then there is an analytic function f : G1 → G2 such that (G1 . Exercise 8. That is the diagram is commutative. Not available. Exercise 9. Not available. Prove that the closed curve in Exercise IV. What is a simply connected covering space of the ﬁgure eight? Solution. assume that both G1 and G2 are simply connected. f ) is a covering of G2 and f1 = f2 ◦ f .6. (Hint: If f (z1 ) = f (z2 ) let γ be a path in G from z1 to z2 and consider a certain analytic continuation along f ◦ γ. Show that there is a one-one analytic function mapping G1 onto G2 .Solution. f ) is a covering space of Ω. f1 ) and (G2 . Give two nonhomeomorphic covering spaces of the ﬁgure eight that are not simply connected. Not available. Solution. in addition. Show that for every region Ω1 contained in Ω which is simply connected there is an analytic function g : Ω1 → G such that f (g1 (ω)) = ω for all ω in Ω1 . f ) be a covering of Ω and suppose that f is analytic. Solution. (a) Let (G. Not available. f2 ) are coverings of the region Ω such that both f1 and f2 are analytic. Exercise 7.8 is not homotopic to zero in the doubly punctured plane.) (b) Suppose that (G1 .

that is (uy ) xx + (uy )yy = 0. Solution. 135 .1) In addition. (10. (u x ) xx + (u x )yy = u xxx + u xyy = u xxx + uyyx = (u xx + uyy ) x Thus. Now. To show: f = u x − iuy is analytic which is equivalent to showing a) Re( f ) = u x and Im( f ) = −uy are harmonic and b) u x and −uy have to satisfy the Cauchy Riemann equations (by Theorem III 2. Thus.1 Basic properties of harmonic functions Exercise 1. u is harmonic. (10. (10. we show that u x is harmonic. u is inﬁnitely times diﬀerentiable by Proposition 1.29). Thus.3 p. we have u xx + uyy = 0. u x is harmonic. Since u : G → R is harmonic. Let u be harmonic. 252. all the partials are continuous (therefore u xy = uyx ).3 p. uy is harmonic. (uy ) xx + (uy )yy = uyxx + uyyy = u xxy + uyyy = (u xx + uyy )y Thus. the third partial derivatives of u exist and are continuous (therefore u xyy = uyyx and uyxx = u xxy ).1) = 0 x = 0.2) = 0y = 0. Finally. Since. Exercise 2.Chapter 10 Harmonic Functions 10. u is inﬁnitely times diﬀerentiable by Proposition 1. that is (u x ) xx + (u x )yy = 0. If u is harmonic. that is u xx + uyy = 0. 252.1) ∂u ∂x and uy = ∂u ∂y . (10. Show that if u is harmonic then so are u x = Solution. we show that uy is harmonic. show that f = u x − iuy is analytic.

Not available.l=0 Show that p is harmonic iﬀ: (a) k(k − 1)ak. Prove that a harmonic function is an open map. So f (a) = Multiply by r. we have f (a) = 1 2πi f (w) 1 dw = w−a 2π 2π f (a + reiθ ) dθ 0 γ where γ(t) = a + reit . (Hint: Use the fact that the connected subsets of R are intervals. There exist an f ∈ H(D) such that Re( f ) = u on D. If f is analytic on G and f (z) Solution. y) = n akl xk yl for all x.l = 0 for 2 ≤ k. part a) is proved. (c) ak. 0 for any z show that u = log | f | is harmonic on G. The second is true since u xy = uyx . R) ⊂ D ⊂ G. Not available.l−2 + l(l − 1)ak−2.) Solution. 0 f (a + reiθ )r dθ 0 136 . y) dx dy. we get f (a)r = 1 2π 2π 1 2π 2π f (a + reiθ ) dθ.l = 0 for 2 ≤ l ≤ n. Exercise 5. we have seen that u x and uy are harmonic. Exercise 3. l ≤ n. From Cauchy’s Integral Formula. Let D be a disk such that B(a. Let p(x. ¯ Solution. Therefore.a) By the previous exercise. Not available. Therefore (u x )y = −(−uy ) x ⇐⇒ u xy = uyx ⇐⇒ u xy = uyx . f = u x − iuy is analytic provided u is harmonic.R) and (u x )y = −(−uy ) x . Let u be harmonic in G and suppose B(a. ¯ Exercise 6.n = 0 for 2 ≤ k ≤ n. R) ⊂ G. Solution. −uy is harmonic. Thus.2) since (u x ) x = −(uy )y ⇐⇒ u xx + uyy = 0. y in R. Exercise 4. Show that u(a) = 1 πR2 u(x. that is (u x ) x = −(uy )y The ﬁrst is true by (10. b) We have to show that the Cauchy-Riemann equations are satisﬁed.n−1 = ak. (b) an−1. ¯ B(a. too.l = an. k. 0 ≤ t ≤ 2π. In summary.

Therefore 1 + z 2 u(z) = Im 1−z is harmonic. ¯ B(a. For |z| < 1 let 1 πR2 Re( f (x. Next.7? Why? Solution. y)) dx dy = ¯ B(a. y) dx dy. lim u(reiθ ) = 0 137 .and then integrate with respect to r yields R R f (a)r dr = 0 0 R 0 2π 0 1 2π 2π 2π f (a + reiθ )r dθ dr 0 ⇐⇒ ⇐⇒ ⇐⇒ f (a) r2 1 = 2 2π 1 f (a) = πR2 1 f (a) = πR2 f (a + reiθ )r dθ dr 0 R f (a + reiθ )r dr dθ 0 f (x. y) dx dy. ¯ B(a. Clearly. f ∈ H(D) since z = 1 is not in D. Let f (z) = 1+z . y) dx dy. ¯ B(a. By 1−z Theorem III 2.29 we obtain that v = Re( f ) and u = Im( f ) are harmonic.R) Show that u is harmonic and limr→1− u(reiθ ) = 0 for all θ. we will show r→1− 2 1+z u(z) = Im 1−z 2 .R) 1 πR2 u(x.R) 1 πR2 u(x.R) where the last step follows by changing from polar coordinates to rectangular coordinates. Now u(a) = Re( f (a)) = Thus u(a) = Exercise 7. Does this violate Theorem 1. Let D = {z ∈ C : |z| < 1}.

This does not violate Theorem 1. 2 )2 r→1 (1 − r)4 (1 − 2r cos(2πk) + r r→1− Thus. k ∈ Z.for all θ. then 1 + x + iy u(z) = u(x. 4y(−y2 ) −4y3 4 = 4 =− y y4 y But − 4 → ∞ as y → 0− . k ∈ Z ⇐⇒ z = 1. that r = 1 and θ 2πk. y 138 . (1 − 2r cos θ + r2 )2 Note that the denominator is zero iﬀ r = 1 and θ = 2πk. We have 1 + reiθ 2 u(re ) = Im 1 − reiθ 1 + reiθ 2 1 − reiθ = Im 1 − reiθ 1 − reiθ (1 + reiθ )(1 − reiθ ) = Im (1 − reiθ )(1 − reiθ ) iθ = Im (1 + reiθ − re−iθ − r2 )2 (1 − reiθ − re−iθ + r2 )2 (1 + 2ri sin θ − r2 )2 = Im (1 − 2r cos θ + r2 )2 ((1 − r2 ) + 2ri sin θ)2 = Im (1 − 2r cos θ + r2 )2 = Im = 2 2 (1 − r2 )2 + 4ri sin θ(1 − r2 ) − 4r2 sin2 θ (1 − 2r cos θ + r2 )2 4r(1 − r2 ) sin θ . then r→1 lim− u(re2πk ) = lim− r→1 4r(1 − r2 ) sin(2πk) 0 = lim− = 0. lim u(reiθ ) = 0. y) = 2 2 2 = 4y(1 − x − y ) .7. y) = Im 1 − x − iy u(1. 2 )2 (1 − 2· 1· cos θ + 1 2 (1 − cos θ)2 If θ = 2πk. (1 − 2x + x2 + y2 )2 ∀y 0. k ∈ Z. Claim: lim supz→1 u(z) > 0. then the condition lim sup u(z) ≤ 0 z→a does not hold ∀a ∈ ∂D. Proof of the claim: If we ﬁx x = 1. then r→1 lim− u(reiθ ) = u(eiθ ) = 4· 1· (1 − 12 ) sin θ 0 = 2 = 0. The reason is the following: If we pick u = u and v = 0. If we assume.

State and prove a Schwarz Reﬂection Principle for harmonic functions. Using Deﬁnition 2. θ) = u(r cos θ. 255. By exercise 2 p. By Exercise 1 p. Show that f (reiθ ) = iθ 1 2π π −π f (eit )Pr (θ − t) dt for all re in D. Can A have a limit point in G? Solution. Let D = {z : |z| < 1} and suppose that f : D− → C is a continuous function such that both Re f and Im f are harmonic. Show that u is harmonic iﬀ u(z) = a log |z| + b for some constants a and b. Solution.Exercise 8. r sin θ). then A = B = G and thus A can have a limit point in G. we know that f = u x − iuy is analytic. Since f is analytic. 78 A has a limit point in G). we know that u x and uy are harmonic. Not available. Clearly A = {z ∈ G : u x (z) = uy (z) = 0} = B since u x (z) − iuy (z) = 0 iﬀ u x (z) = uy (z) = 0 (u x (z) and uy (z) are real). 10. Solution. Not available. the zeros are isolated and therefore B cannot have a limit point in G (Corollary 3. ∂r ∂r ∂θ (b) Let u have the property that it depends only on |z| and not arg z. That is. Not available. that is.2 Harmonic functions on a disk Exercise 1. 255. u(z) = ϕ(|z|). 79). Let u : G → R be a function with continuous second partial derivatives and deﬁne U(r. Let u be harmonic and not constant. Exercise 9. Let u : G → R be harmonic and let A = {z ∈ G : u x (z) = uy (z) = 0}. 139 . (a) Show that ∂2 U ∂U ∂2 U ∂ ∂U ∂2 U ∂2 u ∂2 u + 2 = r2 2 + r r + 2. Deﬁne B = {z ∈ G : f (z) = 0} = {z ∈ G : u x (z) − iuy (z) = 0}.7 p. Exercise 10. A is the set of zeros of the gradient of u. (In this case f ≡ 0 and by Theorem 3. Exercise 11. Deduce the Maximum Principle for analytic functions from Theorem 1.10 p. If u is constant. This implies that A cannot have a limit point in G.1 show that f is analytic on D iﬀ π f (eit )eint dt = 0 −π for all n ≥ 1. Solution.6. + 2 =r r2 2 ∂r ∂r ∂r ∂x ∂y ∂r ∂θ ∂θ So if 0 G then u is harmonic iﬀ r ∂ ∂U ∂2 U r + 2 = 0.

By Corollary 2. 262 ⇒ −π f (eit )eint dt = 0 ∀n ≥ 1. Let D = {z : |z| < 1} and suppose f = u + iv : D → C is a continuous function such that u and v are harmonic. 140 . ⇒: Let f be analytic on D and assume n ≥ 1 (n integer). we will prove f is analytic ⇐⇒ 1 2π π −π f (eit )Pr (θ − t) dt π f (eiθ )eint dt = 0 −π for all n ≥ 1. f (reiθ ) = u(reiθ ) + iv(reiθ ) = 1 2π π −π 1 2π π −π Pr (θ − t)u(eit ) dt for 0 ≤ r < 1 and all θ (10.9 we have u(reiθ ) = and v(reiθ ) = Hence. 0 < r < 1. Let γ = reit .4) 1 = 2π (10. Clearly u and v are continuous functions mapping D to R. Next. (10.(10.Solution.3) 1 2π π −π Pr (θ − t)v(eit ) dt for 0 ≤ r < 1 and all θ.3). −π ≤ t ≤ π. Thus. we have proved f (reiθ ) = for all reiθ in D. Deﬁne g(z) = f (z)zn−1 .4) π −π Pr (θ − t)u(eit ) dt + i 1 2π π −π 1 2π π −π Pr (θ − t)v(eit ) dt Pr (θ − t) u(eit ) + iv(riθ ) dt = Pr (θ − t) f (eit ) dt for 0 ≤ r < 1 and all θ. Clearly g is analytic on D. then by Cauchy’s Theorem g(z) dz = 0 γ ⇐⇒ ⇐⇒ ⇐⇒ ⇐⇒ which implies r→1 π f (z)zn−1 dz = 0 γ π f (reit )ireit eit(n−1) rn−1 dt = 0 −π π i −π π f (reit )eint rn dt = 0 f (reit )eint rn dt = 0 −π π lim− f (reit )eint rn dt = 0 −π by Ex 3a p.

) (f) Show that if the function g in part (e) is real valued then the polynomials can be chosen with real coeﬃcients. Thus f is analytic on D. z) → f˜ (z) uniformly for z in T .1]. ¯ ¯ ¯ (d) Suppose g : [0. (b) If f : T → C is a continuous function deﬁne f˜ : D− → C by f˜(z) = f (z) for z in T and 1 f˜(reiθ ) = 2π π −π n f (eit )Pr (θ − t) dt (So Re f˜ and Im f˜ are harmonic in D). Not available.1. z)} of polynomials in z and z such that pn (z. Let D = {z : |z| < 1}. 1] → C is a continuous function then there is a sequence {pn } of polynomials such that pn (t) → g(t) uniformly for t in [0. Exercise 3. z)} of polynomials in z and z such that pn (z. what parts of the conclusion remain true? Solution. If f : T → C is a continuous function then there is a sequence {pn (z. Is the conclusion of the theorem still valid? If not.⇐: Assume π −π f (reit )eint rn dt = 0 for all n ≥ 1. If g : [0. Now.) (c) Weierstrass approximation theorem for T.4 suppose that f is piecewise continuous on ∂D. (Hint: Apply part (d) to the function g(t) + (1 − t)g(1) + tg(0). In the statement of Theorem 2. 1]. 1]. (z. (Hint: Use ¯ ¯ ¯ r Deﬁnition 2. We have for all z ∈ D f (z) = = f (reiθ ) = Part 1 1 2π π −π ∞ f (eit )Pr (θ − t) dt r|n| ein(θ−t) dt f (eit )e−int dt 1 2π π f (eit ) ∞ −π n=−∞ π −π π = 1 r|n| einθ 2π n=−∞ 1 r|n| einθ 2π n=−∞ 1 2π ∞ n=0 π −1 = f (eit )e−int dt + −π =0 by assumption 1 2π ∞ n=0 π rn einθ −π f (eit )e−int dt = zn −π f (eit )e−int dt. Exercise 2. Use part (c) to show that there is a sequence {pn } of polynomials such that pn (t) → g(t) uniformly for t in [0. 141 . for any closed rectiﬁable curve γ in D we get f (z) dz = γ γ 1 2π ∞ n=0 π zn −π f (eit )e−int dt dz = 1 2π ∞ n=0 γ π zn dz −π =0 f (eit )e−int dt since z is analytic for n ≥ 0. z) → f (z) uniformly for z in T . Show that for each r < 1 there r r is a sequence {pn . (e) Weierstrass approximation theorem for [0. Deﬁne f˜ : T → C by f˜ (z) = f˜(rz). 1] → C is a continuous function such that g(0) = g(1). T = ∂D = {z : |z| = 1} (a) Show that if g : D− → C is a continuous function and gr : T → C is deﬁned by gr (z) = g(rz) then gr (z) → g(z) uniformly for z in T as r → 1− .

a ∈ G.3 Subharmonic and superharmonic functions Exercise 1. Solution. (a) Show that ϕ(G) = D and ϕ(∂∞G) = ∂D. 142 . Not available. f (ic) = limz→ic u(z). y) = x + y2 . a = (α. Note that −π sin θ dθ = 0. Solution. Suppose that u is a harmonic function on G0 such that limz→a u(z) exists and is equal to A. Exercise 4. β + r sin θ) = (α + r cos θ)2 + (β + r sin θ)2 = α2 + 2αr cos θ + r2 cos2 θ + β2 + 2βr sin θ + r2 sin2 θ = α2 + β2 + 2αr cos θ + 2βr sin θ + r2 . Let G be an open set. (b) Show that if f : ∂∞G → R is a continuous function then there is a continuous function u : Γ → R such that u(z) = f (z) for z in ∂∞G and u is harmonic in G. u(x + iy) = 2 + (y − t)2 π −∞ x Show that u is a bounded harmonic function on the right half plane such that for c in R. 10.Solution. What can you say about the behavior of v(r) as r → 1− ? What about v(reiθ ) as r → 1− and θ → 0? Solution. π −π sin2 θ dθ = π and π −π cos2 θ dθ = π.5). Not available. Thus 1 2π π π π ϕ(a + reiθ ) dθ −π = α2 + β2 + αr π cos θ dθ + −π =0 βr π sin θ dθ +r2 −π =0 = α2 + β2 + r2 ≥ α2 + β2 = ϕ(a). (b) ϕ(x. Suppose there is a homeomorphism ϕ of Γ onto D− (D = {z : |z| < 1}) such that ϕ is analytic on G. Which of the following functions are subharmonic? superharmonic? harmonic? neither subharmonic nor superharmonic? (a) ϕ(x. For all ϕ(a + reiθ ) = ϕ(α + r cos θ. (c) Suppose that the function f in part (b) is not assumed to be continuous at ∞. Exercise 6. (f) ϕ(x. ∂∞G = Γ − G. y) = x2 − y. β) ∈ C and any r > 0 we have a) π π −π cos θ dθ = 0. Let v be a harmonic conjugate of u. Let D = {z : |z| < l} and suppose f : ∂D → R is continuous except for a jump discontinuity at ¯ z = 1. Deﬁne u : D → R by (2. Let f : {z : Re z = 0} → R be a bounded continuous function and deﬁne u : {z : Re z > 0} → R by x f (it) 1 ∞ dt. y) = x2 + y2 . y) = x2 + y. y) = x − y2 . Not available. Solution. Not available. Solution. Exercise 7. y) = x2 − y2 . (d) ϕ(x. (e) ϕ(x. Show that if U : G → R is deﬁned by U(z) = u(z) for z a and U(a) = A then U is harmonic on G. Show that there is a continuous function u : G− → R such that u(z) = f (z) for z in ∂G and u is harmonic in G (see Exercise 2). and G0 = G − {a}. Show that u is harmonic. Exercise 5. (c) ϕ(x. Let G be a simply connected region and let Γ be its closure in C∞ . Not available.

β + r sin θ) = (α + r cos θ)2 − (β + r sin θ) = α2 + 2αr cos θ + r2 cos2 θ − β − r sin θ. β + r sin θ) = (α + r cos θ)2 − (β + r sin θ)2 = α2 + 2αr cos θ + r2 cos2 θ − β2 − 2βr sin θ − r2 sin2 θ. therefore also ϕ ∈ Subhar(G) and ϕ ∈ Superhar(G). b) ϕ(a + reiθ ) = ϕ(α + r cos θ. 2 Hence. β + r sin θ) = (α + r cos θ)2 + (β + r sin θ) = α2 + 2αr cos θ + r2 cos2 θ + β + r sin θ. e) ϕ(a + reiθ ) = ϕ(α + r cos θ. Thus 1 2π π ϕ(a + reiθ ) dθ −π = α2 + β + αr π 2 π cos θ dθ + −π =0 r2 2π π cos2 θ dθ + −π =π r 2π π sin θ dθ −π =0 = α2 + β + Hence. ϕ ∈ Subhar(G). ϕ ∈ Subhar(G). β + r sin θ) = (α + r cos θ) + (β + r sin θ)2 = α + r cos θ + β2 + 2βr sin θ + r2 sin2 θ. d) ϕ(a + reiθ ) r ≥ α2 + β = ϕ(a). π Thus 1 2π π ϕ(a + reiθ ) dθ −π = α2 − β + = α2 − β + αr π 2 cos θ dθ + −π =0 r2 2π π −π cos2 θ dθ − =π r 2π π sin θ dθ −π =0 r ≥ α2 − β = ϕ(a).Hence. ϕ ∈ Har(G). ϕ ∈ Subhar(G). 143 . 2 = ϕ(α + r cos θ. Thus 1 2π π π ϕ(a + reiθ ) dθ −π = α2 − β2 + r2 2π π αr π cos θ dθ + −π =0 r2 2π π −π cos2 θ dθ − =π βr π π sin θ dθ −π =0 − r2 r2 sin2 θ dθ = α2 − β2 + − = α2 − β2 = ϕ(a). 2 2 −π =π Hence. c) ϕ(a + reiθ ) = ϕ(α + r cos θ.

r 2π π −π Thus 1 2π π ϕ(a + reiθ ) dθ −π = α − β2 + cos θ dθ − =0 βr π π −π sin θ dθ − =0 r2 2π π sin2 θ dθ −π =π r2 = α − β2 − ≤ α − β2 = ϕ(a). (This exercise is diﬃcult.Thus 1 2π π ϕ(a + reiθ ) dθ −π = α + β2 + r 2π π cos θ dθ + −π =0 βr π π sin θ dθ + −π =0 r2 2π π sin2 θ dθ −π =π r2 ≥ α + β2 = ϕ(a). (b) T is positive (i. if { fn } is a sequence in C(∂∞G. T (a1 f1 + a2 f2 ) = a1 T ( f1 ) + a2 T ( f2 )). = α + β2 + 2 Hence. However.e. (d) If the Dirichlet Problem can be solved for G then T is one-one. Not available. Exercise 4. Let Subhar(G) and Superhar(G) denote. ϕ ∈ Superhar(G).11 remains valid. β + r sin θ) = (α + r cos θ) − (β + r sin θ)2 = α + r cos θ − β2 − 2βr sin θ − r2 sin2 θ.) Solution. prove that the conclusion remains valid. suppose only that f is a bounded function on ∂∞G. Exercise 5. Not available. (This is useful if G is an unbounded region and g is a bounded continuous function on ∂G. Is the converse true? Solution. f) ϕ(a + reiθ ) = ϕ(α + r cos θ. the sets of subharmonic and superharmonic functions on G. In the hypothesis of Theorem 3. the behavior near points of ∂G can be studied with hope of success. Not available. If we deﬁne f : ∂∞G → R by f (z) = g(z) for z in ∂G and f (∞) = 0 then the conclusion of Theorem 3. Prove: (a) T is linear (i. (c) T is continuous. Exercise 3.11. (a) Show that Subhar(G) and Superhar(G) are closed subsets of C(G. 144 . (b) Does a version of Harnack’s Theorem hold for subharmonic and superharmonic functions? Solution. R)..5 is necessary. This deﬁnes a map T : (∂∞G. Exercise 2. R) such that fn → f uniformly then T ( fn ) → T ( f ) uniformly on G. R) → Har(G) by T ( f ) = u f .. Of course there is no reason to expect that the harmonic function will have predictable behavior near ∞ — we could have assigned any value to f (∞). ϕ ∈ Subhar(G). Moreover. Show that the requirement that G1 is bounded in Corollary 3.e. 2 Hence. if f (a) ≥ 0 for all a in ∂∞G then T ( f )(z) ≥ 0 for all z in G). respectively.) If G is a region and if f : ∂∞G → R is a continuous function let u f be the Perron Function associated with f .

To ﬁnally show that it is analytic. If we can show that for every bounded ¯ ¯ region G1 such that G1 ⊂ G and for every continuous function u1 : G1 → R that is harmonic in G1 and satisﬁes u(z) ≤ u1 (z) for z in ∂G1 . let z0 ∈ A. Clearly f is continuous. If f (z) 0 for all z in G. since ϕ : Ω → R is subharmonic. What happens if f (z) 0 for all z in G? Solution. u = ϕ ◦ f is subharmonic on G.4 p.4 to obtain: For every region Ω1 contained in Ω and every harmonic function ϕ on Ω1 . Not available.4 p.7 p. We can use Theorem 7. Since f is continuous. f = ( f −1 )−1 is an open map. Since ϕ : Ω → R is subharmonic. By assumption we have u(z) ≤ u1 (z) ∀z ∈ G1 which is the same as u(z) = ϕ( f (z)) ≤ u1 (z) = ϕ1 ( f (z)) ∀z ∈ ∂G1 since f maps G1 onto Ω1 . ϕ( f (z)) − ϕ1 ( f (z)) ≤ 0 ∀z ∈ ∂G1 . 99).5 p. 145 . Clearly ϕ1 is harmonic (harmonic composite analytic is harmonic). and thus ∂G1 onto ∂Ω1 (we choose this by assumption). this implies ϕ( f (z)) − ϕ1 ( f (z)) ≤ 0 that is and we are done. we know that f −1 exists and is analytic. We have to show u(z) ≤ u1 (z) ∀z ∈ G1 . U is a neighborhood of z0 . ¯ ¯ Hence. f (z) − w |z−z0 |= Note. then the previous result holds locally for some neighborhood in G. Deﬁne u = ϕ ◦ f . By the Open Mapping Theorem (Theorem 7. then u is subharmonic (by Corollary 3. ( f is locally one-one and onto and f −1 is analytic and hence u = ϕ ◦ f is locally subharmonic). Then there is a neighborhood U of z0 and a neighborhood V of w0 = f (z0 ) such that f : U → V is one-one and onto and f −1 : V → U is analytic. let G1 be a bounded region G1 such that G1 ⊂ G and let u1 : G1 → R that is harmonic in G1 and satisﬁes u(z) ≤ u1 (z) for z on ∂G1 . Exercise 6. If f : G → Ω is analytic and ϕ : Ω → R is subharmonic. and let f (z0 ) 0. Thus. since f : G → Ω is analytic and ϕ is continuous. then u : G → R is continuous. Let V = {w| |w − w0 | < δ} and let U be the inverse image of V under the map f restricted to {z| |z − z0 | < }. so f −1 is continuous from V to U. In fact ϕ−ϕ1 satisﬁes the Maximum Principle ˜ ˜ on Ω1 . if we assume that ϕ is twice diﬀerentiable. then the result holds globally. ϕ−ϕ satisﬁes the Maximum Principle on Ω1 . show that ϕ ◦ f is subharmonic if f is one-one. f maps U one-one onto V. Since f is one-one. 98 to ﬁnd > 0 and δ > 0 such that each w with |w − w0 | < δ has exactly one pre-image z with |z − z0 | < . we have u(z) ≤ u1 for z in G1 . 125. We have f −1 (w) = 1 2πi z f (w) dz. 265). we can use Theorem 3. u(z) − u1 (z) ≤ 0 ∀z ∈ G1 ∀z ∈ G1 . we can use Proposition 3. 98. Since ϕ − ϕ1 satisﬁes the Maximum Principle on Ω1 . By Theorem 7. Therefore.5 p. Proof of the claim: f (z) − w0 has a simple zero at z0 since f (z0 ) 0.Solution. So deﬁne ¯ ϕ1 = u1 ◦ f −1 : Ω → R ¯ ¯ with f −1 : Ω1 → G1 . Here is the claim: Let f : G → C be analytic.

Let G = B(0. Not available. r). 255 in G − {a}. r) − {a} as a sum of two harmonic functions. w a. This exercise asks for an easier proof of a special case of Theorem 4. we can argue again that log |z−a| is harmonic on B(a. (ii) G = {z : Re z > 0}. (Note that there is no other point z ∈ G such that f (z) = 0 since f is assumed to be one-one). Not available. Hence z→w lim ga (z) = lim − log | f (w)| = − log 1 = 0 w→a 146 . By assumption f (z) = 0 iﬀ z = a. Show that G has a barrier at a. (a) Let G be a simply connected region. Solution. Show that the Green’s Function on G with singularity at a is ga (z) = − log | f (z)|. Not available. let a ∈ G. Let the disk be B(a. ga (z) has a singularity if | f (z)| = 0. r)−{a}.4 The Dirichlet Problem Exercise 1.5 Green’s Function Exercise 1. and let f : G → D = {z : |z| < 1} be a one-one analytic function such that f (G) = D and f (a) = 0. r)−{a} (choose f (z) = z − a and f (z) = 0 iﬀ z = a). So G(z) has a removable singularity at a and therefore G(z) is harmonic on B(a. (Hint: Consider the transformation (z − a)(z − b)1 .10. Exercise 4. We have | f (z)| = 1 ∀z ∈ ∂∞G by assumption ( f (∂∞G) = ∂D). Solution. Clearly ga (z) = − log | f (z)| is harmonic on B(a. This happens if f (z) = 0. Let G = C − (∞. Solution. b) and c) of Deﬁnition 5. Not available. let a ∈ G. r). Solution to part a): Let G be simply connected. 10. Solution. Hence G(z) = ga (z) + log |z − a| = − log | f (z)| + log |z − a| is harmonic on B(a. Exercise 2. and let f : G → D = {z : |z| < 1} be a one-one analytic function such that f (G) = D and f (a) = 0.9. This follows directly by Exercise 5 p. c) limz→w ga (z) = 0 for each w in ∂∞G. 0] and construct a barrier for each point of ∂∞G. Let G be a bounded region and let a ∈ ∂G such that there is a point b with [a. To show: ga (z) = − log | f (z)| is a Green’s Function on G with singularity at a. Let G be a region and a a point in ∂∞G such that there is a harmonic function u : G → R with limz→a u(z) = 0 and lim inf z→w u(z) > 0 for all w in ∂∞G. b) G(z) = ga (z)+log |z−a| is harmonic in a disk about a. Show that there is a barrier for G at a. Exercise 3. 0]. (b) Find the Green’s Functions for each of the following regions: (i) G = C − (∞. It remains to verify a). But G(z) = − log | f (z)| + log |z − a| = − log |z − a|· | f˜(z)| | f (z)| = − log = − log | f˜(z)| |z − a| |z − a| where f (z) = (z − a) f˜(z) by assumption and f˜(z) is analytic on G with no zero. b] ∩ G− = {a}. Clearly. By Exercise 5 p.) Solution. 1) and ﬁnd a barrier for G at each point of the boundary. 255. (iii) G = {z : 0 < Im z < 2π}. This implies ga (z) has a singularity at a. a) ga is harmonic in G − {a}.1.

Let u(z) = ga (z) + log |z − a| and let v be the harmonic conjugate of u. In addition h is one-one. If we can ﬁnd a map. say h : G → D. analytic and h(a) = 0 and therefore satisﬁes a). 160 Theorem 4. is given by ga (z) = − log |h(z)| by Part 1 a). Proof of the Claim: We will invoke the Orientation Principle (p. Exercise 3. (iii) if a G then a branch of log(z − a) can be deﬁned.1. Exercise 2. then the Green’s Function. 147 . then ga (z) ≤ ψ(z) for z a. Let ga be the Green’s Function on a region G with singularity at z = a. Hence h(z) = f (g(z)) maps G onto D. Let Gr be a component of {z : | f (z)| < r} and apply Rouche’s Theorem to get that f (z) = 0 and f (z) − w0 = 0. This implies ga (z) = − log | f (z)| is a Green’s Function on G with singularity at a. b) and c) is guaranteed by the Riemann Mapping Theorem p. G is a simply connected region (not the whole plane). (ii) every harmonic function on G has a harmonic conjugate. 53 in the book for the derivation).3). G C. (So f is analytic in G. Solution. Note that h has to satisfy the following assumptions to use Part 1 a): a) h : G → D one-one and analytic b) h(a) = 0 c) h(G) = D. prove that there is a Green’s Function ga on G with singularity at a. Not available. Prove that if ψ is a positive superharmonic function on G − {a} with lim inf z→a [ψ(z) + log |z − a|] > −∞. Thus. 162 we have seen that the Möbius Transformation (which is one-one and analytic) f (z) = z − g(a) 1 − g(a)z maps D onto D such that f (g(a)) = 0. From p. If ϕ = u + iv let f (z) = eiα (z − a)e−ϕ(z) for a real number α. say ga . Cr = {z : | f (z)| = r} consists of a ﬁnite number of simple closed curves in G (see Exercise VI.2.for each a in ∂∞G. (a) Let G be a bounded simply connected region and let a ∈ G. 53) to ﬁnd g : G → D where g is one-one and analytic. This exercise gives a proof of the Riemann Mapping Theorem where it is assumed that if G is a simply connected region.) Prove that | f (z)| = exp(−ga (z)) and that limz→w | f (z)| = 1 for each w in ∂G (Compare this with Exercise 1). Prove that for 0 < r < 1. It remains to ﬁnd h: Claim: h : G → D given by h(z) = f (g(z)) where g(z) = z − g(a) z−1 and f (z) = z+1 1 − g(a)z works. b) and c). The existence and uniqueness of this function h with the assumptions a). Solution to part b) ii): Clearly. the Green’s Function for the region G = {z : Re(z) > 0} is given by ga (z) = − log |h(z)| where h(z) is given by the Claim. Then g(z) = z−1 z+1 (see p. then: (i) C∞ − G is connected.

Solution. a ∈ G. Not available.5) is a formula for the solution of the Dirichlet Problem with boundary values f . 148 . and α = l(a).5) reduces to equation (2. δ < d(a. for a suitable choice of α. a) = ga (z) is the Green’s Function on G with singularity at a. v = ga (z) = g(z. ∞ ∈ ∂∞G. (Note: these concepts are not discussed in this book but the formula is suﬃciently interesting so as to merit presentation. (Show that l omits all values in a neighborhood of α + 2πi. Solution. u = h. (b) Let G be a simply connected region with G C. If f : ∂G → R is continuous and g(z.) (Hint: Apply Green’s formula [u∆v − v∆u] dx dy = u ∂Ω Ω ∂v ∂u −v ds ∂n ∂n with Ω = G − {z : |z − a| ≤ δ}. Prove that f is one-one on Gr . have the same number of solutions in Gr .5).) (c) Combine parts (a) and (b) to prove the Riemann Mapping Theorem. From here conclude that f (G) = D = {z : |z| < 1} and f (a) > 0. Show that l is one-one on G and l(z) α + 2πi for any z in G. where ∂g is the derivative of ∂n g in the direction of the outward normal to γ and ds indicates that the integral is with respect to arc length. Prove that ϕ(z) = [l(z) − α − 2πi]−1 is a conformal map of G onto a bounded simply connected region in the plane. a) ds ∂n (5. Let l be a branch of log z on G. show that h(a) = γ f (z) ∂g (z.|w0 | < r. a). {γ}).) (b) Show that if G = {z : |z| < 1} then (5. but assume that G is unbounded and 0. Exercise 4. (a) Let G be a region such that ∂G = γ is a simple continuously diﬀerentiable closed curve. Not available.

r(z−ak ) r2 −¯ k z a maps B(0. Let rm F(z) = f (z) m z n k=1 r2 − ak z ¯ .Chapter 11 Entire Functions 11. r) onto itself and takes the boundary to the boundary. we have log |F(0)| = Also g(z) ≡ = = f (z) = zm ∞ f (i) (0)zi i=0 i! zm 1 2π 2π 0 log |F(reiθ )| dθ = 1 2π 2π 0 log | f (reiθ )| dθ. i! m! (m + 1)! f (m) (0) m! .1 Jensen’s Formula Exercise 1. ak 149 .. r(z − ak ) Then F ∈ H(G) and F has no zeros in B(0. ∞ f (i) (0)zi i=m i! zm ∞ i=m f (m) (0) 0 f (m+1) (0) 1 f (i) (0)zi−m = z + z + .. In the hypothesis of Jensen’s Formula. Show that if f has a zero log | f (reiθ )| dθ. Thus. by a known result. ( f has a zero at z = 0 of multiplicity m) which implies g(0) = n Thus. r) and |F(z)| = | f (z)| if |z| = r. F(0) = g(0)rm k=1 − r . We have f (m) (0) r 1 + m log r = − log + m! |ak | 2π k=1 n 2π 0 0. do not suppose that f (0) at z = 0 of multiplicity m then log Solution.

. 2π 0 ≥ 0 since r ≤ |bk | < 2r log k=1 2r ≤ log |M(2r)|. That is. n 2π 1 log | f (2reiθ )| dθ 2π 0 2π 1 log |M(2r)| dθ = log |M(2r)|. . Since f is an entire function. . . 2r). an . log |F(0)| = log which is equivalent to log log k=1 r f (m) (0) = log + m log r + |ak | m! n log k=1 r . 2r) − B(0. Let f be an entire function. Suppose that f (0) = 1 and show that n(r) log 2 ≤ log M(2r). . n(r) = the number of zeros of f in B(0. it is analytic on B(0. . an are the zeros of f in B(0. bm are the zeros of f in B(0. . Exercise 2. ¯ Solution. . . . a1 .Therefore n |F(0)| = |g(0)|rm implies n k=1 r |ak | n log |F(0)| = log |g(0)| + m log r + Hence. Suppose a1 . b1 . r) counted according to multiplicity. . |ak | f (m) (0) + m log r + m! n log k=1 r 1 = |ak | 2π 2π 0 2π 0 log | f (reiθ )| dθ 1 f (m) (0) r + + m log r = − log m! |ak | 2π k=1 log | f (reiθ )| dθ which proves the assertion. M(r) = sup{| f (reiθ )| : 0 ≤ θ ≤ 2π}. Since f (0) = 1 0. we have by Jensen’s formula n 0 = log 1 = log | f (0)| = − which implies 2r log |ak | k=1 n log k=1 2r − |ak | m m log k=1 2r 1 + |bk | 2π 1 2π 2π 0 log | f (2reiθ )| dθ = − log k=1 2r |bk | 2π 0 + log | f (2reiθ )| dθ ≤ ≤ Def of M(r) Hence. . r). r) repeated according to multiplicity and b1 . . . . bm are the zeros of f in B(0. |ak | n(r) But n log k=1 2r |ak | = ≥ n(r) log(2) 2r 2r = log |ak | |ak | k=1 k=1 n(r) n(r) n r 2r 2 = log 2n(r) + log = log |ak | |ak | k=1 k=1 log 150 n(r) . 2r) repeated according to multiplicity. .

Let f (z) = cn zn be an entire function of ﬁnite genus µ. n(t) dt = t n log k=1 r |ak | where a1 . For suﬃciently large r = |z|. (Hint: Use Cauchy’s Estimate. Evaluate r 0 n(t) dt. Let D = B(0. prove that f (z) = zm B(z) exp(−g(z)) where B is a Blaschke Product (Exercise VII. prove that n→∞ lim cn (n!)1/(µ+1) = 0. In Jensen’s Formula do not suppose that f is analytic in a region containing B(0.1) by Theorem 2. . 283. . (a) If {an } are the non-zero zeros of f in D counted according to multiplicity. Not available. 1) and suppose that f : D → C is an analytic function which is bounded. Exercise 4. Not available. By Cauchy’s estimate p. . prove that (1 − |an |) < ∞. Exercise 5. r) minus the number of poles (each counted according to multiplicity). 5. Solution. Deﬁne M(r) = max{ f (reiθ ) : 0 ≤ θ ≤ 2π} = max|z|=r | f (z)|.2 The genus and order of an entire function Exercise 1. t Solution. Not available. 5. we have M(r) < eαr µ+1 (11. Thus. (Hint: Use Proposition VII. ¯ Exercise 3. 11. Evaluate 1 2π Solution. (a) Using the notation of Exercise 2. n(r) log(2) ≤ log |M(2r)| which proves the assertion.) Solution. (b) If f has a zero at z = 0 of multiplicity m ≥ 0. r). (b) Let f be meromorphic without a pole at z = 0 and let n(r) be the number of zeros of f in B(0. we have |cn | ≤ M(r) rn 151 . prove that r 0 2π 0 log | f (reiθ )| dθ.4) and g is an analytic function on D with Re g(z) ≥ − log M (M = sup{| f (z)| : |z| < 1}). 73. r) but only that f is meromorphic with no pole at z = 0.since 0 < |ak | < r implies r |ak | > 1 ∀k which means n(r) r k=1 |ak | > 1 so the logarithm is greater than zero.4). an are the zeros of f in B(0. .6 p.

⇐⇒ (n ) |cn | < |cn | < = 1 n λ λ n ) µ+1 λn (n e λ n 1 < for any > 0. and choose λ > e for convenience. n > N say. |cn | ≤ M(r) µ+1 < eαr . − log |cn | 1 = . provided n is taken large enough. Exercise 2. Since n! < nn . B. Hence. dn 152 . Let f be an entire function of order λ. we have (n!) µ+1 |cn | < for n > N or n→∞ 1 lim cn (n!)1/(µ+1) = 0. if we choose α such that αλµ+1 = 1. n r (11. Solution. r = n µ+1 λ. Then (11. Show that f = f1 + f2 has ﬁnite order λ and λ ≤ max(λ1 . where λ is a positive constant. λ2 ) with f 0. λ2 ). Prove that if f is an entire function of order λ then f also has order λ. 286 we have α = lim inf n→∞ ∞ n=0 cn zn . Suppose f is an entire function and A. for n ≥ 2. Exercise 3. a are positive constants such that there is a r0 with | f (z)| ≤ exp(A|z|a + B) for |z| > r0 .where cn = f (n) (0) n! . the last inequality becomes (nn ) µ+1 |cn | < 1 Deﬁne. Exercise 4.2) Now. Solution. Not available. Not available. Show that λ = max(λ1 . Let f1 and f2 be entire functions of ﬁnite orders λ1 and λ2 respectively. Solution.1) (11. Show that f is of ﬁnite order ≤ a. then f has power series f (z) = By Exercise 5 e) p. n log n λ ∞ n=0 Diﬀerentiating f yields f (z) = ∞ n=1 ncn zn−1 = (n + 1)cn+1 zn . λ2 ) if λ1 λ2 and give an example which shows that λ < max(λ1 .2) can be written in the form αλµ+1 n µ+1 µ+1 e 1 eαnλ eαnλ n µ+1 .

) (b) Suppose that 0 < α < ∞ and show that for any > 0. Note that sinh2 z = cosh2 z − 1 and that cos x is bounded for real values x. Take r suﬃciently large so that N > p and show that ∞ n=N+1 r nα− n N < 1 and n=p+1 r nα− n < B exp (2r)1/(α− ) where B is a constant which does not depend on r. (c) cosh ∞ −an n z where a > 0. 153 . Exercise 5. so | cos z|2 = cos2 x + sinh2 y ≤ cosh2 y ≤ cosh2 |z|. (e) Prove that f is of ﬁnite order iﬀ α > 0. (b) cos z. (d) 1 Solution. there is an integer p such that |cn |1/n < n−(α− ) for n > p. Conclude that for |z| = r > 1 there is a constant A such that | f (z)| < Ar p + ∞ n=1 r nα− n (c) Let p be as in part (b) and let N be the largest integer ≤ (2r)1/(α− ) . Let f (z) = cn zn be an entire function and deﬁne the number α by α = lim inf n→∞ − log |cn | n log n (a) Show that if f has ﬁnite order then α > 0. Solution. let λ be the order of f . (d) Use parts (b) and (c) to show that if 0 < α < ∞ then f has ﬁnite order λ and λ ≤ α−1 .Now. < α. λ = λ and therefore f also has order λ. λ Hence. and ﬁnd the maximum value of this expression. c) The order is λ = 2 .) n=1 n √ z. (Hint: For part (d) use Exercise 5. Not available. then 1 λ − log |dn | − log |(n + 1)cn+1 | = lim inf n→∞ n log n n log n − log(n + 1) − log |cn+1 | = lim inf n→∞ n log n − log(n + 1) log |cn+1 | (n + 1) log(n + 1) = lim inf − · n→∞ n log n n log n (n + 1) log(n + 1) − log(n + 1) − log |cn+1 | n + 1 log(n + 1) + · · = lim inf n log n n→∞ (n + 1) log(n + 1) n log(n) →1 1 = lim inf n→∞ →0 →λ →1 = 1 . Find the order of each of the following functions: (a) sin z. and if f has order λ then λ = α−1 . (Hint: If the order of f is λ and β > λ show that |cn | ≤ r−n exp(rβ ) for suﬃciently large r. Exercise 6.

4). If ρ is the exponent of convergence of {an } prove that ρ ≤ λ. we have for > 0. (c) Let f be an entire function of order λ and let {a1 . Solution. Next we want to show that λ = 1 . λ2 .Thus | cos z| ≤ cosh |z| = and therefore | cos 1 1 |z| (e + e−|z| ) ≤ e|z| ≤ e|z| . Exercise 8.(11.5) in the next section. 4. show that f = f1 f2 has ﬁnite order λ ≤ max(λ1 . Deﬁne M(r) = max|z|=r | f (z)|. Hence. for all |z| > r. 2 Exercise 7. M1 (r) = max|z|=r | f1 (z)| and M2 (r) = max|z|=r | f2 (z)|. Since f1 and f2 are entire functions of ﬁnite order λ1 and λ2 . (ii) (iii) 2 2 This shows that no > 0 can be found that satisﬁes equation (11. .5) for suﬃciently large r = |z| by Proposition 2.14. Seeking contradiction assume that 2 there is > 0 and r > 0 such that √ 1 | cos z| ≤ exp(|z| 2 − ) (11.λ2 ) ≤ e2r λ+ /2 < er λ+ provided r is suﬃciently large. |an |−(ρ+ ) < ∞ and |an |−(ρ− ) = ∞. Let ρ = inf{a : |an |−a < ∞}. the number called the exponent of convergence of {an }.4) (11. M1 (r) < er and M2 (r) < er λ2 + /2 λ1 + /2 (11. . a2 . and 1 1 (iii) 2 > n2 and | f (z)| = | cos √ n 1 1 1 1 z| = |en + e−n | > en = e− log 2+n > e 2 > exp(n−2 n) = exp(n2( 2 − ) ) = exp(|z| 2 − ). This implies λ = 1 . Let f1 and f2 be entire functions of ﬁnite order λ1 .) 154 . λ2 ).3). Hence M(r) = max | f (z)| = max | f1 (z)· f2 (z)| ≤ max | f1 (z)|· max | f2 (z)| |z|=r |z|=r |z|=r |z|=r = M1 (r)M2 (r) λ1 + /2 +rλ2 + /2 < by (11. Let {an } be a sequence of non-zero complex numbers. (Hint: See the proof of (3. 21/(2 ) } then in particular (i) |z| > r.} be the non-zero zeros of f counted according to multiplicity. 2 2 1 This implies that the order λ ≤ 2 . λ2 ). respectively.3) √ z| ≤ exp(|z|1/2 ). (ii) − ln 2 + n > 2 . √ n Let z = −n2 for an n ∈ N with n > max{ r.5) er λ1 + /2 er λ2 + /2 = er λ≤max(λ1 . (a) If f is an entire function of rank p then the exponent of convergence ρ of the non-zero zeros of f satisﬁes: p ≤ ρ ≤ p + 1. (b) If ρ = the exponent of convergence of {an } then for every > 0. the order of f (z) = f1 (z)· f2 (z) has ﬁnite order λ ≤ max(λ1 . .

(d) Let P(z) = ∞ E p (z/an ) be a canonical product of rank p. ≥n ≥n ≥n k Each product of consecutive numbers is of the form (n − (k − 1))k.7) to show that for some > 0 ∞ n=N+1 log E p z an < A|z|ρ+ . In fact we will show that λ = 0. for 1 ≤ k ≤ 2m. assume that |al | ≤ |a2 | ≤ . ρ ≤ λ. n ≤ 1 2 2m and n − k + 1 > 2m which justiﬁes the individual estimates by n. . ∞ n=1 0 < |a| < 1 z n! 1− n! ≥ n· (n − 1)· 2· (n − 2)· 3 . We conclude that ρ = inf{m : the left sum converges} = 0.) Solution. Prove that the order of P is ρ. It suﬃces to show that the exponent of convergence ρ = 0 and to employ Exercise 8d). Not available. By Part d) of Exercise 8 we conclude that the order of g(z) is zero. use (2. By choice of n. z Solution. Choose N such that |an | ≤ 2|z| if n ≤ N and |an | > 2|z| if n ≤ N + 1. Find the order of the following entire functions: (a) f (z) = b) f (z) = c) f (z) = [Γ(z)]−1 . Let m ≤ N. (n − (2k − 1))· 2m . . (Hint: If λ is the order of P. . and let ρ be the exponent of convergence of n=1 {an }. |z| > 1. b) Considering g(z) := ∞ 1 − n! one notes that g(z) is already in form of the canonical m=1 product with the entire function in the exponent being constantly zero and p = 0. Exercise 9. Hence also the genus µ = 0. or equivalently ∞ n=4m (n!)− m ≤ 1 ∞ n−2 n=4m and the sum on the left converges for every positive integer m. Treating the cases ρ < p + 1 and ρ = p + 1 separately. Use this to prove that N log E p n=1 z an < C|z|ρ+ for some constant C independent of z. . 155 . 1 Prove that for |z| ≥ 2 . log |E p (z)| < B|z| p where B is a constant independent of z. In total this yields n! > n for n large. let n > 4m then ∞ n=1 (1 − an z). The simple zeros of g(z) are at z = n! and from Exercise 8a) it follows that λ ≤ 1. and ﬁx z.

Show that ∂h − i ∂h = ff on G0 . we h x = hu u x + hv v x = and hy = hu uy + hv vy = and hence ∂h ∂h −i ∂x ∂y = = = = (C-R) u v ux + 2 vx u2 + v2 u + v2 u2 v u u + 2 vy 2 y +v u + v2 h x − ihy u v u v ux + 2 vx − i 2 uy − i 2 vy u2 + v2 u + v2 u + v2 u + v2 v u (u − iuy ) + 2 (v x − ivy ) 2 + v2 x u u + v2 v u (u x − iuy ) + 2 (−uy − iu x ) u2 + v2 u + v2 u iv (u x − iuy ) − 2 (u x − iuy ) u2 + v2 u + v2 u − iv (u x − iuy ) u2 + v2 u − iv (u x − iuy ) (u − iv)(u + iv) u x − iuy . Let f = u(x.11.23 f = u x + iv x and thus 2 f = u x + iv x − iuy + vy implies f = Therefore. We have by p. Let G0 = G − {z : f (z) = 0}. Since f is analytic.22 and 2. f u + iv Next. Let G0 = G − {z : f (z) = 0} and deﬁne h : G0 → R by h(z) = log | f (z)|.6) 1 2 = h x and ∂h ∂y = hy where h(z) = log | f (z)| = log(u2 + v2 ). then h : G0 → R given by h(z) = log | f (z)| is well deﬁned as well as ff is well deﬁned on G0 . the Cauchy-Riemann (C-R) equations u x = vy and uy = −v x are satisﬁed. y) = u + iv. u + iv 156 = = = = . we calculate get ∂h ∂x and f = −iuy + vy 1 1 1 u x + iv x − iuy + vy = u x − iuy − iuy + u x = 2u x − 2iuy = u x − iuy . Let f be analytic in a region G and suppose that f is not identically zero. u x − iuy f = . Using the chain rule.3 Hadamard Factorization Theorem Exercise 1. Let f be analytic in a region G and suppose that f is not identically zero. (C-R) 2 2 2 (11. y) + iv(x. ∂x ∂y Solution. 41 Equation 2.

also F = f − g is an entire function of ﬁnite order λ (see Exercise 2 p. . we also have ∞ n=1 |an |−(λ+1) ≤ ∞ n=1 p≤λ |an |−(p+1) < ∞ so ∞ |an |−(λ+1) < ∞ contradicting the assumption ∞ |an |−(λ+1) = ∞. Assume F 0 ( f g). Show that f = g. Refer to Exercise 2. we have ∞ n=1 |an |−(ρ+ ) < ∞. . because F(an ) = f (an ) − g(an ) = 0 since f (an ) = g(an ) by assumption. (c) Find all entire functions f of ﬁnite order such that f (log n) = n. Suppose that f (an ) = g(an ) for a sequence {an } such that |an |−(λ+1) = ∞. q) where p is the rank of F. 289). Since F 0. 286).} and no other zeros. the sequence {an } are the zeros of F. ∂x ∂y f Exercise 2. we will derive a contradiction. according to Exercise 8 b) p. the sequence {an } are the zeros of F. because F(an ) = f (an ) − g(an ) = 0 since f (an ) = g(an ) by assumption. then ρ≤λ by Exercise 8 c) p. By the deﬁnition of the rank (p. Since by deﬁnition µ = max(p. Exercise 3. we have ∞ n=1 λ2 then λ = max(λ1 . By Hadamard’s Factorization Theorem (p. Solution. where an ’s are the zeros of F. b) Since f and g are entire functions of ﬁnite order λ. Assume F 0 ( f g). 286). 286-287. 281 Deﬁnition 2. Since F 0. Suppose that f (an ) = g(an ) for a sequence {an } such that |an |−(λ+ ) = ∞. |an |−(p+1) < ∞. we also have F = f − g is an entire function of ﬁnite order λ (see Exercise 2 p.7 and show that if λ1 Solution. 286-287.8 to show that if f .1). we will derive a contradiction. 157 . (a) Let f and g be entire functions of ﬁnite order λ and suppose that f (an ) = g(an ) for a sequence {an } such that |an |−(λ+1) = ∞. Hence F ≡ 0. Therefore by the comparison test. we certainly have p ≤ λ. Hence. . (b) Use Exercise 2. (d) Give an example of an entire function with zeros {log 2. g and {an } are as in part (a) with |an |−(λ+ ) = ∞ for some > 0 then f = g. a) Since f and g are entire functions of ﬁnite order λ. and therefore we n=1 n=1 obtain f = g. Let ρ be the exponent of convergence of {an }. λ2 ). Not available.Compare this with (11. log 3.6) and we obtain the assertion ∂h ∂h f −i = . F has ﬁnite genus µ ≤ λ since F ∈ H(C) with ﬁnite order λ.

∞ n=1 |an |−(λ+ ) = ∞. Thus. we also have ∞ n=1 |an |−(λ+ ) ρ≤λ ≤ ∞ n=1 |an |−(ρ+ ) < ∞. F ≡ 0 and therefore we get the 158 . Hence ∞ |an |−(λ+ ) < ∞ contradicting the fact n=1 assertion f = g.Therefore by the comparison test.

. Choose a with |a| = r0 . In this setting Lemma 1. ρ) 2 | f (z)| ≤ 3|z − a| 1 . It follows that |g(z)| = | γ f (w) dw| ≤ 1 |z| ≤ 1 =: M. ρ0 )) ⊃ B f (a). Solution. ρ0 2ρ2 0 For z ∈ B(0. Solution. h(r) = (1 − r)K(r). ρ0 The function g is not necessarily injective on B(0. = B 0. ρ0 )) ⊃ B 0. Examine the proof of Bloch’s Theorem to prove that L ≥ 1/24. ρ0 ) the line segment γ = [a. Also choose ρ0 = 1 (1 − r0 ) and infer that for all z ∈ B(a. Suppose that in the statement of Bloch’s Theorem it is only assumed that f is analytic on D. set K(r) = 1 max{| f (z)|||z| = r}. 24 ρ2 |g (0)|2 1 1 = B 0. i.) Do the same for Proposition 1. This statement rewritten for f yields f (B(a. Not available. ρ0 ) but the problem only asks for an estimate on the Landau constant for which injectivity is not required. 1 . r0 = sup{r : h(r) = 1}.10. a + z] is completely contained in B(a. and | f (z) − f (a)| < . 159 .e. 6M 4·6 24 Exercise 2. ρ0 ) deﬁne g(z) = f (z + a) − f (a). | f (a)| = K(r0 ) = 1−r0 .1 Bloch’s Theorem Exercise 1.2 gives g(B(0. Mimic the proof of Block’s Theorem up to the application of Schwarz’s Lemma. By convexity of B(a. ρ0 ). What conclusion can be drawn? (Hint: Consider the functions f s (z) = s−1 f (sz).Chapter 12 The Range of an Analytic Function 12. 0 < s < 1.

1 equals F together with the constant functions ∞. 12. 0. Not available. b. b ∈ C that are not in the image of f . Thus g is a nonzero constant function and f (z) = g(z)−1 + a is also a constant function and the result is established.12. hence f (z) ∞ for at least some z ∈ C.4 The Great Picard Theorem Exercise 1. 12. Not available. does this imply anything about the nature of the singularity at z = 0? Solution. Show that if f is a one-one entire function then f (z) = az + b for some constants a and b. (Hint: What if ∞ f (C)?) Solution.3 Schottky’s Theorem No exercises are assigned in this section. It cannot be the zero-function. For each integer n ≥ 1 determine all meromorphic functions f and g on C with a pole at ∞ such that f n + gn = 1. a Solution. R) − {0} and discuss all possible values of the integral 1 2πi γ f (z) dz f (z) − a where γ is the circle |z| = r < R and a is any complex number. 160 .2 The Little Picard Theorem Exercise 1. First consider an entire function f that misses three points in C∞ . one on which is ∞. because f was assumed to be meromorphic. Deﬁne a function g(z) := 1 . Not available. Not available. f (z) − a z∈C 1 1 then g is entire because f (z) − a 0 for all z ∈ C and b−a . Prove that the closure of the set F in Theorem 4. Exercise 2. Exercise 3. If it is assumed that this integral takes on certain values for certain numbers a. Solution. Again using Little Picard’s Theorem conclude that g is a constant function. 0. and 1. Exercise 2. Then there are two distinct numbers a. Show that if f is a meromorphic function on C such that C∞ − f (C) has at least three points then f is a constant. Hence ∞ ∈ f (C) and by assumption there are distinct a. Solution. Next consider a meromorphic function on C that is not entire. By Little Picard’s Theorem the function f must be a constant. Let f be analytic in G = B(0. c−a ∈ C − g(C). c ∈ C − f (C).

- MeasureTheory UCLA
- Foundation of Mathematical Economics Solutions
- Differential Equations
- _A Short Course on the Lebesgue Integral and Measure Theory
- folland2
- Homework #8, Sec 12.3 and 13.1
- Problems-and-Solutions-For-Complex-Analysis
- Measure Theory Notes - David Simms
- Solutions
- Willard - General Topology (Solutions)
- 275 Differential Equations Student Solution Manual1[1]
- Solutions Functions of One Complex Variable I, Second Edition by John B. Conway
- Calculus Salas
- Solution Measure Integration
- (GTM 011) John B Conway-Functions of One Complex Variable (1978)
- Oxford Notes on Limits and Cont
- Homework #9, Sec 13.2 and Sec 13.3
- Student Solutions Manual to Accompany Boyce Elementary Differential Equations and Boundary Value Problems
- 58670038 Answers Munkres
- Solutions to Apostol
- Functions of One Complex Variable
- Dummit Solutions
- Applied. Econometrcisstatapdf
- Topics in Analysis(part II).pdf
- FirstCourseGR_notes_on_Schutz2009.pdf
- Solutions manual Apostol.pdf
- Spivak Calculus of Manifolds Solutions
- Problems and Solutions in Mathematics, Li - Ta- Sien - World Scientific
- Solucionario Tom m Apostol - Analisis Matematico 2a Edicion (Editorial Reverte)
- Abstract Algebra I. N. Herstein p265

- Hinton - A New Era of Thought
- Hologram World
- UT Dallas Syllabus for huas6394.501.11f taught by Thomas Riccio (txr033000)
- NASA Facts Gravitational Biology Facility and Centrifuge Facility on the International Space Station
- The Universal Religion through Science
- The Futures of the City Region
- tmp5D76.tmp
- Relativity Gravity Probe Press Kit
- Hinton - The Fourth Dimension
- tmp909F.tmp
- tmpFB7B.tmp
- The Universal Bible
- tmpA038
- The Edinburgh Lectures on Mental Science by Troward, Thomas, 1847-1916
- tmp18AE.tmp
- International Programs May 1973
- Wormhole
- The Unified Theory of Physics
- Artspace _ Alternative Space (Weekend 3, October 26-27)
- UT Dallas Syllabus for phys5392.501.07s taught by Wolfgang Rindler (rindler)
- Hinton - Scientific Romances (1)
- The Unified Theory of Physics 1
- The Unified Thoery of Physics 2
- A Series of Lessons in Gnani Yoga by Ramacharaka, Yogi [pseud.], 1862-1932
- tmpC01E.tmp
- NASA Facts Goddard Space Flight Center
- Four-Dimensional Vistas by Bragdon, Claude Fayette, 1866-1946
- tmp50E7.tmp
- The Unified Thoery of Physics
- The Concept of NatureThe Tarner Lectures Delivered in Trinity College, November 1919 by Whitehead, Alfred North, 1861-1947

Sign up to vote on this title

UsefulNot usefulRead Free for 30 Days

Cancel anytime.

Close Dialog## Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

Close Dialog## This title now requires a credit

Use one of your book credits to continue reading from where you left off, or restart the preview.

Loading