Topics In Complex Analysis Dan Romik download

https://ebookbell.com/product/topics-in-complex-analysis-dan-
romik-51439654

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

Topics In Complex Analysis 1st Edition Joel L Schiff

https://ebookbell.com/product/topics-in-complex-analysis-1st-edition-
joel-l-schiff-46597782

Topics In Complex Analysis And Operator Theory Oscar Blasco Jose A

Bonet

https://ebookbell.com/product/topics-in-complex-analysis-and-operator-
theory-oscar-blasco-jose-a-bonet-6702456

Selected Topics In Complex Analysis The S Ya Khavinson Memorial Volume

Operator Theory Advances And Applications 1st Edition Vladimir Ya
Eiderman Editor

https://ebookbell.com/product/selected-topics-in-complex-analysis-the-
s-ya-khavinson-memorial-volume-operator-theory-advances-and-
applications-1st-edition-vladimir-ya-eiderman-editor-2002094

Bergman Spaces And Related Topics In Complex Analysis Proceedings Of A

Conference In Honor Of Boris Korenblums 80th Birthday November 2023
2003 Hadenmalm Zhu Borichev Ed

https://ebookbell.com/product/bergman-spaces-and-related-topics-in-
complex-analysis-proceedings-of-a-conference-in-honor-of-boris-
korenblums-80th-birthday-november-2023-2003-hadenmalm-zhu-borichev-
ed-4764956
Current Topics In Pure And Computational Complex Analysis 1st Edition
Santosh Joshi

https://ebookbell.com/product/current-topics-in-pure-and-
computational-complex-analysis-1st-edition-santosh-joshi-4976312

A Course In Complex Analysis From Basic Results To Advanced Topics

2012th Edition Wolfgang Fischer

https://ebookbell.com/product/a-course-in-complex-analysis-from-basic-
results-to-advanced-topics-2012th-edition-wolfgang-fischer-2363322

Topics In Contemporary Differential Geometry Complex Analysis And

Mathematical Physics Proceedings Of The 8th International Workshop On
Complex And Infomatics Bulgaria 2126 August Stancho Dimiev

https://ebookbell.com/product/topics-in-contemporary-differential-
geometry-complex-analysis-and-mathematical-physics-proceedings-of-
the-8th-international-workshop-on-complex-and-infomatics-
bulgaria-2126-august-stancho-dimiev-4445784

Topics In Several Complex Variables First Usauzbekistan Conference

Analysis And Matematical Physics May 2023 2014 California State
University Fullerton Ca Zair Ibragimov

https://ebookbell.com/product/topics-in-several-complex-variables-
first-usauzbekistan-conference-analysis-and-matematical-physics-
may-2023-2014-california-state-university-fullerton-ca-zair-
ibragimov-6703884

Classical Topics In Complex Function Theory 1st Edition Reinhold

Remmert

https://ebookbell.com/product/classical-topics-in-complex-function-
theory-1st-edition-reinhold-remmert-60695968
Dan Romik
Topics in Complex Analysis
Also of Interest
Topics in Complex Analysis
Joel L. Schiff, 2022
ISBN 978-3-11-075769-9, e-ISBN (PDF) 978-3-11-075782-8
in: De Gruyter Studies in Mathematics
ISSN 0179-0986

Complex Analysis
Theory and Applications
Teodor Bulboacǎ, Santosh B. Joshi, Pranay Goswami, 2019
ISBN 978-3-11-065782-1, e-ISBN (PDF) 978-3-11-065786-9

Elementary Operator Theory

Marat V. Markin, 2020
ISBN 978-3-11-060096-4, e-ISBN (PDF) 978-3-11-060098-8

Real Analysis
Measure and Integration
Marat V. Markin, 2019
ISBN 978-3-11-060097-1, e-ISBN (PDF) 978-3-11-060099-5

Elementary Functional Analysis

Marat V. Markin, 2018
ISBN 978-3-11-061391-9, e-ISBN (PDF) 978-3-11-061403-9
Dan Romik

Topics in Complex
Analysis

�
Mathematics Subject Classification 2020
Primary: 30-01, 11-01; Secondary: 52C07, 52C17

Author
Prof. Dan Romik
Department of Mathematics
University of California
One Shields Ave
Davis CA 95616
USA
[email protected]

ISBN 978-3-11-079678-0
e-ISBN (PDF) 978-3-11-079681-0
e-ISBN (EPUB) 978-3-11-079688-9
DOI https://doi.org/10.1515/9783110796810

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0

International License. For details go to http://creativecommons.org/licenses/by-nc-nd/4.0/.

Library of Congress Control Number: 2023935854

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available on the Internet at http://dnb.dnb.de.

© 2023 the author(s), published by Walter de Gruyter GmbH, Berlin/Boston. The book is published open
access at www.degruyter.com.
Cover image: Guy Kindler and Dan Romik
Typesetting: VTeX UAB, Lithuania
Printing and binding: CPI books GmbH, Leck

www.degruyter.com
�
To Abigail
Contents
Preface � XI

0 Prerequisites and notation � 1

0.1 Prerequisites � 1
0.2 Notation � 1
Exercises for Chapter 0 � 3

1 Basic theory � 4
1.1 Motivation: why study complex analysis? � 4
1.2 The fundamental theorem of algebra � 9
1.3 Holomorphicity, conformality, and the Cauchy–Riemann equations � 12
1.4 Additional consequences of the Cauchy–Riemann equations � 18
1.5 Power series � 20
1.6 Contour integrals � 22
1.7 The Cauchy, Goursat, and Morera theorems � 28
1.8 Simply connected regions and the general version of Cauchy’s
theorem � 32
1.9 Consequences of Cauchy’s theorem � 36
1.10 Zeros, poles, and the residue theorem � 46
1.11 Meromorphic functions, holomorphicity at ∞, and the Riemann
sphere � 50
1.12 Classification of singularities and the Casorati–Weierstrass theorem � 52
1.13 The argument principle and Rouché’s theorem � 53
1.14 The open mapping theorem and maximum modulus principle � 57
1.15 The logarithm function � 58
1.16 The local behavior of holomorphic functions � 60
1.17 Infinite products and the product representation of the sine function � 63
1.18 Laurent series � 68
Exercises for Chapter 1 � 71

2 The prime number theorem � 82

2.1 Motivation: analytic number theory and the distribution of prime
numbers � 82
2.2 The Euler gamma function � 83
2.3 The Riemann zeta function: definition and basic properties � 89
2.4 A theorem on the zeros of the Riemann zeta function � 97
2.5 Proof of the prime number theorem � 99
Exercises for Chapter 2 � 110
VIII � Contents

3 Conformal mapping � 118

3.1 Motivation: classifying complex regions up to conformal equivalence � 118
3.2 First singleton conformal equivalence class: the complex plane � 121
3.3 Second singleton conformal equivalence class: the Riemann sphere � 123
3.4 The Riemann mapping theorem � 124
3.5 The unit disc and its automorphisms � 126
3.6 The upper half-plane and its automorphisms � 129
3.7 The Riemann mapping theorem: a more precise formulation � 131
3.8 Proof of the Riemann mapping theorem, part I: technical background � 132
3.9 Proof of the Riemann mapping theorem, part II: the main
construction � 137
3.10 Annuli and doubly connected regions � 140
Exercises for Chapter 3 � 145

4 Elliptic functions � 146

4.1 Motivation: elliptic curves � 146
4.2 Doubly periodic functions � 149
4.3 Poles and zeros; the order of a doubly periodic function � 151
4.4 Construction of the Weierstrass ℘-function � 154
4.5 Eisenstein series and the Laurent expansion of ℘(z) � 158
4.6 The differential equation satisfied by ℘(z) � 159
4.7 A recurrence relation for the Eisenstein series � 160
4.8 Half-periods; factorization of the associated cubic � 161
4.9 ℘(z) and ℘′ (z) generate all doubly periodic functions � 163
4.10 ℘(z) as a conformal map for rectangles � 165
4.11 The discriminant of a cubic polynomial � 168
4.12 The discriminant of a lattice � 170
4.13 The J-invariant of a lattice � 170
4.14 The modular variable τ: from elliptic functions to elliptic modular
functions � 171
4.15 The classification problem for complex tori � 172
4.16 Equivalence between complex tori and elliptic curves � 177
Exercises for Chapter 4 � 179

5 Modular forms � 182

5.1 Motivation: functions of lattices � 182
5.2 The modular group Γ = PSL(2, ℤ) � 184
5.3 The modular group as a group of Möbius transformations � 185
5.4 The fundamental domain and the modular surface ℍ/Γ � 186
5.5 The classification problem for complex tori, part II � 190
5.6 The point at i∞, premodular forms, and their Fourier expansions � 191
5.7 Fourier expansions and number-theoretic identities � 194
 Contents � IX

5.8 Modular functions � 199

5.9 Klein’s J-invariant � 205
5.10 The J-invariant as a conformal map � 208
5.11 The classification problem for complex tori, part III � 209
5.12 Modular forms � 209
5.13 Examples of modular forms � 214
5.14 Infinite products for modular forms � 218
Exercises for Chapter 5 � 228

6 Sphere packing in 8 dimensions � 233

6.1 Motivation: the sphere packing problem in d dimensions � 233
6.2 A high-level overview of the proof � 236
6.3 Preparation: some remarks on Fourier eigenfunctions � 237
6.4 The (+1)-Fourier eigenfunction � 239
6.5 The (−1)-Fourier eigenfunction � 250
6.6 A modular form inequality � 256
6.7 Proof of Theorem 6.1 � 263
Exercises for Chapter 6 � 265

A Appendix: Background on sphere packings � 267

A.1 Sphere packings and their densities � 267
A.2 Lattices and lattice packings � 268
A.3 Periodic sphere packings � 268
A.4 Lattice covolume � 269
A.5 Dual lattices � 269
A.6 The Poisson summation formula for lattices � 270
A.7 Construction of the lattice E8 � 271
A.8 The Cohn–Elkies sphere packing bounds � 276
A.9 Magic functions � 278
A.10 Radial functions and their Fourier transforms � 279
A.11 Structural properties of E8 magic functions � 281
A.12 Summary � 284
Exercises for Appendix A � 286

Bibliography � 289

Web Bibliography � 291

Index � 293
Preface
This book covers the basic theory of complex analysis and a selection of advanced top-
ics. It evolved out of lecture notes from two quarter-long graduate classes that I taught
several times at the University of California, Davis in 2016–2021. The book is primarily
aimed at graduate students, advanced undergraduate students, and postgraduate math-
ematics researchers. It is suited for self-study or as a primary reference material for
approximately two semester-long graduate-level university courses.
The advanced topics covered in Chapters 2–5 are classical and are discussed in many
other places. It is my hope that my own exposition advances the pedagogy of the subject,
if only ever so slightly, by simplifying the explanations, logical arguments, notation, etc,
as much as it has been within my power to do.
The last chapter, Chapter 6, is more modern in content and covers Maryna Via-
zovska’s spectacular application of modular forms to the solution of the sphere pack-
ing problem in dimension 8. Published in 2016, this work was until now only accessi-
ble to learn about from the primary literature [71] and from a few expository papers
[12, 13, 20, 52]. The detailed exposition of Viazovska’s work in Chapter 6, and the ac-
companying Appendix A covering the relevant background material on sphere packing,
should be useful to students and researchers wishing to get up to speed about these
beautiful recent developments, which are at the forefront of much ongoing research.
The choice of topics you will find in this work is idiosyncratic and reflects my own
mathematical taste, interests, and biases. I make no claim that they are the most impor-
tant parts of the vast theory that is complex analysis; only that they are beautiful, that
they relate to many topics and theories that are of interest to a broad section of pure
mathematicians, and that they are, broadly speaking, a fine set of mathematical ideas,
one could devote one’s time to studying and thinking about. I hope some readers will
agree.
I am grateful to Guy Kindler for help with the book cover design and to Christopher
Alexander, Jennifer Brown, Brynn Caddel, Keith Conrad, Bo Long, Anthony Nguyen, Jian-
ping Pan, and Brad Velasquez for helpful comments on versions of the lecture notes the
book evolved from.

Davis Dan Romik

March 2023

Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative
Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
https://doi.org/10.1515/9783110796810-201
0 Prerequisites and notation
0.1 Prerequisites
This book assumes knowledge of the following subjects, roughly at the level covered by
advanced undergraduate courses in the United States:
– Real analysis and multivariable calculus
– Topology of ℝn (mostly for n = 2)
– Complex numbers and their basic properties
– The transcendental functions ez , sin z, cos z of a complex variable

In a few places, some familiarity with Fourier analysis is needed to fully understand
the material. Specifically, in Chapter 2 the Poisson summation formula is derived from
basic properties of Fourier series, and this is used to prove some of the fundamental
properties of the Riemann zeta function. Chapter 6 and Appendix A assume knowledge
of the Fourier transform in ℝn and its basic properties.
Starting in Chapter 3, and increasingly in Chapter 5, knowledge of the basic language
of group theory may be needed to fully understand some of the topics being discussed.
No results from group theory are used beyond the definition of a quotient group.

0.2 Notation
The following notation is used throughout the book.
– ℝ — the real numbers
– ℂ — the complex numbers
– ℤ — the integers
– i — the imaginary unit
– Re(z) — the real part of a complex number z
– Im(z) — the imaginary part of a complex number z
– z — the complex conjugate of a complex number z
– |z| — the modulus of a complex number z
– arg z — the argument of a complex number z
– DR (z) — the open disc of radius R centered at z
– D≤R (z) — the closed disc of radius R centered at z
– CR (z) — the circle of radius R centered at z
– cl(E) — the topological closure of a set E ⊂ ℂ
– 𝔻 — the open unit disc D1 (0)
– ℍ — the upper half-plane: {z ∈ ℂ : Im(z) > 0}
– Ω — a complex region (open and connected subset of ℂ)

Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative
Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
https://doi.org/10.1515/9783110796810-001
2 � 0 Prerequisites and notation

Big-O notation and asymptotic equality. In a few places, the standard big-O notation
is used. Formally, the statement “F = O(G),” where F, G are complex-valued quantities
that depend on one or more variables, means that |F| ≤ C|G| when the variable or vari-
ables in question range over some specified set of values (usually a neighborhood of
some limiting point). Big-O expressions can also be combined in various ways in formu-
las, e. g., “f (t) = O(e−t ) + O(t 2 ) as t → ∞” means that f (t) can be expressed as a sum of
two quantities of the forms O(e−t ) and O(t 2 ), respectively, as t → ∞.
The statement F ∼ G (read as “F is asymptotically equal to G”) means that F/G
converges to 1 in some limiting sense, which is either specified explicitly or inferred
from the context. For example,

sin(x) ∼ x as x → 0

states an asymptotic equality, as does

(2n)! 4n
2
∼ as n → ∞.
(n!) √πn
 Exercises for Chapter 0 � 3

Exercises for Chapter 0

0.1 Important formulas. Below there is a list of basic formulas in complex analysis.
Review each of them, making sure that you understand what it says and why it is
true; that is, if it is a theorem, then prove it, or if it is a definition, then make sure
you understand it.

In the formulas below, a, b, c, d, t, x, y denote arbitrary real numbers, and w, z de-

note arbitrary complex numbers.
a. i2 = −1 l. |wz| = |w| ⋅ |z|
b. (a + bi)(c + di) m. ||w| − |z|| ≤ |w + z| ≤ |w| + |z|
= (ac − bd) + (ad + bc)i n. ex+iy = ex (cos(y) + i sin(y))
c. 1i = −i o. |ez | = eRe(z)
d. z = Re(z) + i Im(z) p. |ez | ≤ e|z|
e. z = Re(z) − i Im(z) q. eit = cos(t) + i sin(t)
z+z
f. Re(z) = 2 r. |eit | = 1
it
z−z
s. cos(t) = e +e
−it
g. Im(z) = 2i 2
h. |z|2 = zz
it
t. sin(t) = e −e
−it

1 z 2i
i. z = |z|2 u. eπi = −1
1
j. x+iy = xx−iy
2 +y2 v. e±πi/2 = ±i
k. w ⋅ z = w ⋅ z w. e2πi = 1

0.2 Reminder of basic analysis concepts. Remind yourself of the definitions of the
following terms in real and complex analysis and the topology of ℂ, referring to
other textbooks or online sources if necessary.
a. real part j. bounded set
b. imaginary part k. compact set
c. complex conjugate l. region
d. modulus m. convergent sequence
e. argument n. Cauchy sequence
f. open set (in ℂ) o. limit point
g. closed set p. accumulation point
q. continuous function
h. closure
i. connected set
1 Basic theory
What is unpleasant here, and indeed directly to be objected to, is the use of complex numbers. ψ is
surely fundamentally a real function.

Erwin Schrödinger, June 6, 1926 letter to Hendrik Lorentz

1.1 Motivation: why study complex analysis?

This book is about complex analysis, the area of mathematics that studies holomorphic
functions of a complex variable and their properties. Although this may sound a bit
specialized, there are (at least) two excellent reasons why all mathematicians should
learn about complex analysis. First, it is, in my humble opinion, one of the most beautiful
areas of mathematics. One way of putting it is that complex analysis seems to have a very
high ratio of theorems to definitions (i. e., a very low “entropy”): you get a lot more as
“output” than you put in as “input.”
The second reason is that complex analysis and, more generally, complex numbers,
have a large number of applications in both the pure mathematics and applied math-
ematics senses of the word. Moreover, many of these applications are to problems that
a priori look like they ought to have little to do with complex numbers. Here are a few
examples, including some that will be discussed later in the book:
– Solving polynomial equations. In 1545, the Italian thinker Gerolamo Cardano pub-
lished the famous formula for solving cubic equations, after learning of the solution
found earlier by Scipione del Ferro. Historically, this appears to have been the first
problem in mathematics to be solved using complex numbers. One surprising aspect
of Cardano’s formula is that it sometimes requires taking operations in the complex
plane as an intermediate step to get to the final answer, even when the cubic equation
being solved has only real roots.
– Proving asymptotic formulas. A well-known approximation to the factorial func-
tion n! is given by Stirling’s formula, which states that the behavior of the factorial
function for large values of n is given by
n
n
n! ∼ √2πn( ) (1.1)
e

(using the notation of Section 0.2). Another famous asymptotic formula is the Hardy–
Ramanujan formula, which states that the number p(n) of integer partitions of n
behaves for large n like

1 π√2n/3
p(n) ∼ e . (1.2)
4√3n

A standard approach to proving these types of results uses complex analysis, as dis-
cussed, for example, in [28].

Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative
Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
https://doi.org/10.1515/9783110796810-002
 1.1 Motivation: why study complex analysis? � 5

– Counting prime numbers. Let π(n) denote the number of primes less than or equal
to n. This function is known as the the prime-counting function. The prime num-
ber theorem states that
n
π(n) ∼ as n → ∞.
log n

This is one of the most celebrated asymptotic formulas (and, indeed, one of the most
famous theorems) in mathematics. Because it deals with prime numbers, it stands
apart from the more general class of asymptotic formulas, such as (1.1)–(1.2) men-
tioned above, and its proof requires more specialized techniques. A standard path to
a proof of the prime number theorem goes through complex analysis, and this is the
subject of Chapter 2.
– Evaluation of complicated definite integrals. Complex analysis offers a set of tech-
niques for evaluating definite integrals that are difficult or impossible to derive using
standard calculus methods. An example is the integral

∞
√π
∫ sin(t 2 ) dt =
2√2
0

(known as one of the Fresnel integrals). See Exercise 1.47 at the end of this chapter
for additional examples.
– Solving partial differential equations. Complex-analytic techniques are very use-
ful for solving several kinds of partial differential equation, particularly those arising
in various applied physics problems in hydrodynamics, heat conduction, electrostat-
ics, and more.
– Analyzing alternating current electrical networks. Electrical engineers learn that
the usefulness of Ohm’s law can be greatly extended by generalizing the notion of
electrical resistance to that of electrical impedance, a complex-valued quantity.
Complex analysis also has many other important applications in electrical engineer-
ing, signal processing, and control theory.
– Solution of the sphere packing problem in 8 and 24 dimensions. It was proved in
2016 that the optimal densities for packing unit spheres in 8 and 24 dimensions are
π4 12

384
and π12! , respectively. The proofs make use of complex analysis in a fundamental
way. The proof for the case of 8 dimensions is presented in Chapter 6.
– Applications in probability and combinatorics. Over the last few decades, com-
plex analysis has been applied in spectacular ways to prove asymptotic results
in probability and combinatorics. One such application is a proof of the Cardy–
Smirnov formula in percolation theory, which answers the following question:
consider a parallelogram-shaped section of cells in the honeycomb lattice with m
rows of cells, each containing n cells. Each cell is colored either black or white ac-
cording to the outcome of a fair coin toss, independently of all other cells (Fig. 1.1(a)).
6 � 1 Basic theory

Figure 1.1: (a) Percolation on a honeycomb: the Cardy–Smirnov formula gives the asymptotic probability
of a left-to-right crossing event. In this sample configuration, a left-to-right crossing has occurred, as illus-
trated by the trail of red dots representing one possible crossing path. (b) A self-avoiding walk of length 45
on the hexagonal lattice.

A left-to-right crossing event is the event that we can find a contiguous path of
white-colored cells connecting the left edge of the parallelogram to the right edge.
What is the asymptotic probability of this event in the limit as the side lengths of
the parallelogram grow to infinity but its shape tends toward a parallelogram with
a fixed aspect ratio?
Specifically, let P(m, n) denote the probability of a left-to-right crossing event. Cardy
conjectured [10] and Smirnov proved [64] the following result.

Theorem 1.1 (Cardy–Smirnov formula). As m, n → ∞ with the aspect ratio m/n con-
verging to a fixed value λ ∈ (0, ∞), the probabilities P(m, n) have the limiting behavior

P(m, n) 󳨀󳨀󳨀󳨀󳨀→ Φ(λ)

m,n→∞
m/n→λ

for an explicit function Φ(λ).

A detailed account of Smirnov’s proof can be found in [34, 73]. The function Φ(λ) is
most naturally defined as a certain geometric invariant associated with the parallel-
ogram with corners 0, 1, ( 1+2 3i )λ, and ( 1+2 3i )λ + 1 and can be written down explicitly
√ √

in terms of modular forms [43] and other special functions from complex analysis.
A second example of a recent application of complex analysis to probability and com-
binatorics is the evaluation of the connective constant of the hexagonal lattice. Let
cn denote the number of self-avoiding walks of length n in the hexagonal lattice that
start at the origin; that is, hexagonal lattice paths that do not intersect themselves;
see Fig. 1.1(b). Without the condition of the path being self-avoiding, the number of
such paths would be exactly equal to 3n . The sequence (cn )∞ n=1 , with initial values
1, 3, 6, 12, 24, 48, 90, 174, 336, . . . [W1], is much more mysterious, and its rate of growth
(as well as the rates of growth of similar sequences associated with the square lattice
 1.1 Motivation: why study complex analysis? � 7

and other natural lattices) have been the subject of much study. From general consid-
erations it is fairly easy to see that the sequence grows roughly exponentially, that is,
there exists a constant μ > 0 such that cn1/n → μ as n → ∞. The constant μ is known
as the connective constant of the hexagonal lattice. Nienhuis [51] conjectured in 1982
and Duminil-Copin and Smirnov [65] proved in 2010 the following remarkable result
concerning the value of μ.

Theorem 1.2 (Duminil-Copin–Smirnov theorem). The connective constant of self-avoi-

ding walks in the hexagonal lattice is equal to √2 + √2 ≈ 1.84776, that is, the numbers
cn satisfy

lim c1/n = √2 + √2.

n→∞ n

– Running the universe. Nature uses complex numbers in the fundamental laws of
physics, Schrödinger’s equation and quantum field theory. This is not a mere math-
ematical convenience or sleight-of-hand, but appears to be a built-in feature of the
very equations describing our physical universe. Why? No one knows.1 (But it is a
fun topic for debate; see, e. g., [42], [W2], [W3].)
– Conformal maps. A conformal map is a mapping from one planar region to another
that preserves angles. This notion, which comes up in purely geometric applications
where the algebraic or analytic structure of complex numbers seems irrelevant, are
in fact deeply tied to complex analysis. Conformal maps were used by the Dutch artist
M. C. Escher (though he had no formal mathematical training) to create amazing art
and used by others to better understand, and even to improve on, Escher’s work. See
Fig. 1.2 and [21, 59] for more on the connection of Escher’s work to mathematics. We
discuss conformal maps in detail in Chapter 3.
– Proving number-theoretic identities. Lagrange proved in 1770 a classic result in
number theory, which states that every positive integer can be represented as a sum
of four squares of integers. Jacobi later proved a more precise fact: if we denote by
r4 (n) the number of distinct ways in which a positive integer n can be represented as
a sum of four squares (with different orderings counting as distinct), then we have
the remarkable identity

r4 (n) = 8 ∑ d. (1.3)
d | n, 4 ∤ d

(In words: eight times the sum of divisors of n that are not divisible by 4.) This beau-
tiful identity and many others like it with a number-theoretic flavor can be proved

1 Schrödinger himself appeared dissatisfied with the idea that his equation uses complex numbers to
describe physical reality. See the epigraph at the beginning of this chapter and [42] for further discussion.
8 � 1 Basic theory

Figure 1.2: The photo is only available in the printed edition.

using complex analysis; see Chapter 5 (and Exercise 5.21 at the end of that chapter
for the particular application to proving (1.3)).
– Complex dynamics. Iteration of complex-analytic maps can be used to generate
beautiful fractals with remarkable properties. A famous example is the iconic Man-
delbrot set (Fig. 1.3) defined as the set of complex numbers c ∈ ℂ for which the
sequence of functional iterates fc(n) (0) of the map fc (z) = z2 + c starting from the
point z = 0 remains bounded.

This has been just a short and necessarily very incomplete survey on the importance
of complex analysis. There are many other intriguing applications and connections of
complex analysis to other areas of mathematics.
In the next section, I will begin our journey into the subject by proving a famous
theorem about polynomials over the complex numbers.
 1.2 The fundamental theorem of algebra � 9

Figure 1.3: (a) The Mandelbrot set. (b) Magnified details of a small region.

1.2 The fundamental theorem of algebra

One of the most famous results about complex numbers is the fundamental theorem of
algebra. Although the statement of the theorem is indeed very fundamental to algebra,
most of its known proofs rely on complex analysis in an essential way. Looking at a few
of these proofs seems like a fitting place to start our journey into the theory.

Theorem 1.3 (Fundamental theorem of algebra). Every nonconstant polynomial

p(z) = an zn + an−1 zn−1 + ⋅ ⋅ ⋅ + a0 (n ≥ 1) (1.4)

with complex coefficients has a complex root.

The fundamental theorem of algebra is a striking and subtle result and has many
beautiful proofs. I will show you three of them.

First proof: analytic proof. Let p(z) be as in (1.4), and consider where |p(z)| attains its
infimum.
First, note that the infimum cannot be attained as |z| → ∞, since

󵄨󵄨 󵄨 n 󵄨 −1 −2 −n 󵄨
󵄨󵄨p(z)󵄨󵄨󵄨 = |z| ⋅ (󵄨󵄨󵄨an + an−1 z + an−2 z + ⋅ ⋅ ⋅ + a0 z 󵄨󵄨󵄨)

and, in particular,

|p(z)|
lim = |an |, (1.5)
|z|→∞ |z|n
10 � 1 Basic theory

so for large |z|, it is guaranteed that |p(z)| ≥ |p(0)| = |a0 |. Now fix some radius R > 0 for
which |z| > R implies |p(z)| ≥ |a0 |, and choose a complex number z0 in the disc DR (0)
for which |p(z0 )| = min|z|≤R |p(z)|. (The minimum exists because p(z) is a continuous
function on the disc.) We then have that

󵄨 󵄨 󵄨 󵄨 󵄨 󵄨 󵄨 󵄨
m0 := inf 󵄨󵄨󵄨p(z)󵄨󵄨󵄨 = inf 󵄨󵄨󵄨p(z)󵄨󵄨󵄨 = min󵄨󵄨󵄨p(z)󵄨󵄨󵄨 = 󵄨󵄨󵄨p(z0 )󵄨󵄨󵄨.
z∈ℂ |z|≤R |z|≤R

Denote w0 = p(z0 ), so that m0 = |w0 |. We now claim that m0 = 0. Indeed, assume by

contradiction that this is not the case. The idea is now to examine the local behavior of
p(z) around z0 . Expanding p(z) in powers of z − z0 , we can write

n
p(z) = w0 + ∑ cj (z − z0 )j
j=1

for some complex coefficients c1 , . . . , cn . This can also be written as

p(z) = w0 + ck (z − z0 )k + ⋅ ⋅ ⋅ + cn (z − z0 )n , (1.6)

where we denote by k the minimal positive index for which cj ≠ 0. Now imagine starting
at the initial point z = z0 and then making a small perturbation away from z0 in the
direction of some unit vector eiθ . We estimate the way that such a perturbation affects
the value p(z). Expansion (1.6) gives

p(z0 + reiθ ) = w0 + ck r k eikθ + ck+1 r k+1 ei(k+1)θ + ⋅ ⋅ ⋅ + cn r n einθ . (1.7)

When r (the magnitude of the perturbation) is very small, the power r k dominates the
other terms r j with k < j ≤ n; that is, (1.7) can be rewritten as

p(z0 + reiθ ) = w0 + r k (ck eikθ + ck+1 rei(k+1)θ + ⋅ ⋅ ⋅ + cn r n−k einθ )

= w0 + ck r k eikθ (1 + g(r, θ)), (1.8)

where we denote
n cj
g(r, θ) = ∑ r j−k ei(j−k)θ .
c
j=k+1 k

Note that g(r, θ) satisfies a bound of the form

󵄨󵄨 󵄨
󵄨󵄨g(r, θ)󵄨󵄨󵄨 < Ar (1.9)

for all r ∈ [0, 1] and some constant A > 0.

To reach a contradiction, we now choose θ, the angle of the perturbation, to be such
that the vector ck r k eikθ “points in the opposite direction” from w0 , that is, such that
 1.2 The fundamental theorem of algebra � 11

ck r k eikθ
∈ (−∞, 0).
w0

This is clearly possible: take θ = k1 (arg w0 − arg(ck ) + π). The idea in doing this is that for
this choice of θ, the expression w0 + ck r k eikθ that forms the dominant term in (1.8) will
have a smaller magnitude than w0 if r is chosen small enough.
To make this precise, choose a number r ∈ [0, 1] smaller than the minimum of the
two numbers 1/(2A) (where A is the constant in (1.9)) and (|w0 |/|ck |)1/k . This choice en-
sures the two inequalities

󵄨󵄨 k ikθ 󵄨󵄨 󵄨 󵄨 1
󵄨󵄨ck r e 󵄨󵄨 < |w0 | and 󵄨󵄨󵄨g(r, θ)󵄨󵄨󵄨 < .
2

With those choices for θ and r, we have that

󵄨󵄨 iθ 󵄨 󵄨 k ikθ 󵄨 󵄨 k ikθ 󵄨 󵄨 k 󵄨
󵄨󵄨p(z0 + re )󵄨󵄨󵄨 = 󵄨󵄨󵄨w0 + ck r e (1 + g(r, θ))󵄨󵄨󵄨 ≤ 󵄨󵄨󵄨w0 + ck r e 󵄨󵄨󵄨 + 󵄨󵄨󵄨ck r g(r, θ)󵄨󵄨󵄨
1
= |w0 | − |ck |r k + |ck |r k 󵄨󵄨󵄨g(r, θ)󵄨󵄨󵄨 < |w0 | − |ck |r k < |w0 | = 󵄨󵄨󵄨p(z0 )󵄨󵄨󵄨.
󵄨 󵄨 󵄨 󵄨
2

This is in contradiction to the defining property of z0 and completes the proof.

Second proof: topological proof. If the constant coefficient a0 = p(0) of p(z) is equal to 0,
then we are done, since 0 is a complex root of p(z). Otherwise, consider the image under
p of the circle |z| = r. Note that, on the one hand, for sufficiently small values of r, the
image is contained in a neighborhood of w0 , so it cannot “go around” the origin.
On the other hand, for r very large, we have

an−1 −1 −iθ a
p(reiθ ) = an r n einθ (1 + r e + ⋅ ⋅ ⋅ + 0 r −n e−inθ )
an an
= an r n einθ (1 + h(r, θ)),

where h(r, θ) is a function that satisfies limr→∞ h(r, θ) = 0 (uniformly in θ). As θ goes
from 0 to 2π, this is a closed curve that goes around the origin n times (in an approxi-
mately circular path, which becomes closer and closer to a circle as r → ∞).
As we gradually increase r from 0 to a very large number, to transition from a curve
that does not go around the origin to a curve that goes around the origin n times, there
has to be a value of r for which the curve crosses 0. This means that the circle |z| = r
contains a point z such that p(z) = 0, which was the claim.

The argument presented in the topological proof is imprecise. It can be made rigor-
ous in a couple of ways—one way we will see a bit later is using Rouché’s theorem (see
Section 1.13 and Exercise 1.30 at the end of the chapter). The difficulty of making these
sorts of arguments precise, in spite of their appealing intuitive nature, gives a hint as to
the importance of subtle topological arguments in complex analysis.
12 � 1 Basic theory

As another remark, the topological proof should be compared to the standard calcu-
lus proof that any odd-degree polynomial over the reals has a real root. That argument
is also “topological”, although much more elementary.

Third proof: typical textbook proof (or: “hocus-pocus” proof). This is a one-liner of a
proof that assumes some complex analysis knowledge. Recall that an entire function is
a function f : ℂ → ℂ that is everywhere holomorphic. Recall the well-known Liouville’s
theorem, which states that any bounded entire function is constant.
Assuming this result (which we will prove in Section 1.9), if p(z) is a polynomial
with no root, then 1/p(z) is an entire function. Moreover, it is bounded, since our earlier
observation (1.5) implies that lim|z|→∞ 1/p(z) = 0. By Liouville’s theorem it follows that
1/p(z) is a constant, which then has to be 0, leading to a contradiction.

To summarize this section, we saw three proofs of the fundamental theorem of al-
gebra. They are all beautiful—the “hocus-pocus” proof certainly packs a punch, which is
why it is a favorite of complex analysis textbooks—but personally I like the first one best
since it is fully rigorous while being completely elementary and not requiring the use
of either Cauchy’s theorem or any of its consequences, or of subtle topological concepts.
Moreover, it employs a “local” argument based on understanding how a polynomial be-
haves locally, where by contrast the other two proofs can be characterized as “global.”
It is a general principle in mathematical analysis (that has analogies in other areas of
mathematics, such as number theory and graph theory) that local arguments are con-
ceptually easier than global ones.

Suggested exercises for Section 1.2. 1.1, 1.2.

1.3 Holomorphicity, conformality, and the Cauchy–Riemann

equations

In this section, we begin to build the theory in a systematic way by laying its most basic
cornerstone, the definition of holomorphicity, along with some of the useful ways to
think about this fundamental concept.

1.3.1 Definition of holomorphicity

A function f (z) of a complex variable is called holomorphic at z0 if the limit

f (z0 + h) − f (z0 )
lim (1.10)
h→0 h
 1.3 Holomorphicity and the Cauchy–Riemann equations � 13

exists. In this case, we denote this limit by f ′ (z0 ) and call it the derivative of f at z0 . A
function of a complex variable defined on all of the complex plane that is everywhere
holomorphic is called an entire function.
The terms analytic, differentiable, and complex-differentiable are synonyms for
“holomorphic.” Some books will make a somewhat pedantic distinction between “ana-
lytic” and “holomorphic” as two distinct concepts that are defined in a priori different
ways but are then shown to be equivalent soon afterward, at which point the distinction
ceases to have any real importance. In this book, we do not follow that approach.
The following are basic properties of complex derivatives.

Lemma 1.4. Under appropriate assumptions (see Exercise 1.4), we have the relations

(f + g)′ (z) = f ′ (z) + g ′ (z), (1.11)

′ ′ ′
(fg) (z) = f (z)g(z) + f (z)g (z), (1.12)
′
1 f (z) ′
( ) =− , (1.13)
f f (z)2
′
f f ′ (z)g(z) − f (z)g ′ (z)
( ) = , (1.14)
g g(z)2
(f ∘ g)′ (z) = f ′ (g(z))g ′ (z). (1.15)

Proof. Exercise 1.4.

The concept of the derivative in complex analysis is clearly at the heart of the sub-
ject, and there are several helpful ways to think about its meaning. Assume that f (z)
is holomorphic at z0 . In the discussion below, we make the further assumption that
f ′ (z0 ) ≠ 0.

1.3.2 First interpretation of holomorphicity: local geometric behavior

If we write the polar decomposition f ′ (z0 ) = reiθ of the derivative, then for points z that
are close to z0 , we will have the approximate equality

f (z) − f (z0 )
≈ f ′ (z0 ) = reiθ
z − z0

or, equivalently,

f (z) ≈ f (z0 ) + reiθ (z − z0 ) + [lower-order terms],

where “lower-order terms” refers to a quantity that is much smaller in magnitude that
|z − z0 | when z is close to z0 . Geometrically, this means that to compute f (z), we start
from f (z0 ) and move by a vector that results by taking the displacement vector z − z0 ,
14 � 1 Basic theory

rotating it by an angle of θ, and then scaling it by a factor of r (which corresponds to a

magnification if r > 1, a shrinking if 0 < r < 1, or no scaling if r = 1). This idea can be
summarized by the slogan:

Holomorphic functions behave locally as a rotation composed with a scaling.

The local behavior of analytic functions in the case f ′ (z) = 0 is more subtle; see Sec-
tion 1.16.

1.3.3 Second interpretation of holomorphicity: the Cauchy–Riemann equations

Next, we interpret holomorphicity from the point of view of real analysis. Remembering
that complex numbers are vectors that have real and imaginary components, we can
denote z = x +iy, where x and y are the real and imaginary parts of the complex number
z, and f = u + iv, where u and v are real-valued functions of z (or, equivalently, of x and
y) that return the real and imaginary parts, respectively, of f . Now if f is holomorphic
at z, then the limit (1.10) exists as a complex limit, that is, independently of the way h
approaches 0 as a complex number. In particular, we can evaluate the limit in two ways
by considering two specific ways of letting h approach 0, as a pure real number or as a
pure imaginary number. For the first of those possibilities, we have

f (z + h) − f (z)
f ′ (z) = lim
h→0 h
u(x + h + iy) − u(x + iy) v(x + h + iy) − v(x + iy)
= lim +i
h→0, h∈ℝ h h
𝜕u 𝜕v
= +i .
𝜕x 𝜕x

Similarly, for the second method of approaching 0, we get that

f (z + h) − f (z)
f ′ (z) = lim
h→0 h
u(x + h + iy) − u(x + iy) v(x + h + iy) − v(x + iy)
= lim +i
h→0, h∈iℝ h h
u(x + iy + ih) − u(x + iy) v(x + iy + ih) − v(x + iy)
= lim +i
h→0, h∈ℝ ih ih
𝜕u 𝜕v 𝜕v 𝜕u
= −i −i⋅i = −i .
𝜕y 𝜕y 𝜕y 𝜕y

Since these limits are equal, by equating their real and imaginary parts we get a cele-
brated system of partial differential equations, the Cauchy–Riemann equations:

𝜕u 𝜕v 𝜕v 𝜕u
= , =− . (1.16)
𝜕x 𝜕y 𝜕x 𝜕y
 1.3 Holomorphicity and the Cauchy–Riemann equations � 15

We have proved that if f is holomorphic at z = x +iy, then the components u and v of

f satisfy the Cauchy–Riemann equations (1.16). A kind of converse to this is also true but
requires additional assumptions. Assume that f = u + iv is continuously differentiable
at z = x + iy (in the sense that each of u and v is a continuously differentiable function
of x, y as defined in ordinary real analysis) and satisfies the Cauchy–Riemann equations
there. This implies that f has a differential at z; that is, in the notation of vector calculus,
if we denote f , z, and Δz as the column vectors

u x h
f = ( ), z = ( ), Δz = ( 1 ) ,
v y h2

then we have
𝜕u 𝜕u
u(z) h
) ( 1)
𝜕x 𝜕y
f (z + Δz) = ( ) + ( 𝜕v + E(h1 , h2 ),
v(z) 𝜕v h2
𝜕x 𝜕y

where E(Δz) is a function of Δz that satisfies

|E(Δz)|
lim = 0.
Δz→0 |Δz|

Now by the assumption that the Cauchy–Riemann equations hold, we also have

𝜕u 𝜕u
h
𝜕u
h + 𝜕u
𝜕x 1
h
𝜕y 2
) ( 1)
𝜕x 𝜕y
( 𝜕v = ( 𝜕u 𝜕u ) ,
𝜕v h2 − 𝜕y h1 + 𝜕x h2
𝜕x 𝜕y

which is the vector calculus notation for the complex number

𝜕u 𝜕u 𝜕u 𝜕u
( − i )(h1 + ih2 ) = ( − i )Δz.
𝜕x 𝜕y 𝜕x 𝜕y

So we have shown that (again, in complex analysis notation)

f (z + Δz) − f (z) 𝜕u 𝜕u E(Δz) 𝜕u 𝜕u

lim = lim ( −i + )= −i .
Δz→0 Δz Δz→0 𝜕x 𝜕y Δz 𝜕x 𝜕y

This proves that f is holomorphic at z with derivative given by f ′ (z) = 𝜕u

𝜕x
− i 𝜕u
𝜕y
. We
summarize the above discussion with the following proposition.

Proposition 1.5 (Cauchy–Riemann equations). Let f = u + iv be a function of a complex

variable z with real and imaginary parts u and v, respectively. If f is holomorphic at z,
then the Cauchy–Riemann equations (1.16) are satisfied at z. Conversely, if equations (1.16)
are satisfied at z and if u and v are continuously differentiable functions at z, then f is
holomorphic at z.
16 � 1 Basic theory

1.3.4 Third interpretation of holomorphicity: conformal maps

Going back to a more geometric way of thinking about holomorphicity, a further inter-
pretation of the meaning of this property is that holomorphic functions are conformal
mappings where their derivatives do not vanish. More precisely, assume as before that
f (z) is holomorphic at z0 and f ′ (z0 ) ≠ 0. Let γ1 , γ2 : (a, b) → ℂ be two differentiable
parameterized planar curves defined on some interval (a, b) containing 0, such that
γ1 (0) = γ2 (0) = z0 . The tangent vectors to the curves γ1 and γ2 at z0 are the complex
numbers v1 and v2 defined by

v1 = γ1′ (0), v2 = γ2′ (0). (1.17)

Similarly, the tangent vectors to the curves f ∘ γ1 and f ∘ γ2 at f (z0 ) are

w1 = (f ∘ γ1 )′ (0) w2 = (f ∘ γ2 )′ (0),

which, by a version of the chain rule from vector calculus adapted to complex-analytic
notation (Exercise 1.6), can be rewritten as

w1 = f ′ (γ1 (0))γ1′ (0) = f ′ (z0 )γ1′ (0), (1.18)

′
w2 = f (γ2 (0))γ2′ (0) =f ′
(z0 )γ2′ (0). (1.19)

It follows that we can write the inner products (in the ordinary sense of planar vector
geometry) between the complex number pairs v1 , v2 and w1 , w2 as

⟨v1 , v2 ⟩ = Re(v1 v2 ),
⟨w1 , w2 ⟩ = Re(w1 w2 ) = Re((f ′ (z0 )γ1′ (0))(f ′ (z0 )γ2′ (0)))
󵄨 󵄨2
= f ′ (z0 )f ′ (z0 ) Re(v1 v2 ) = 󵄨󵄨󵄨f ′ (z)󵄨󵄨󵄨 ⟨v1 , v2 ⟩. (1.20)

If we denote by θ and φ the angle between v1 , v2 and the angle between w1 , w2 , respec-
tively, we then get using (1.17)–(1.20) that

⟨w1 , w2 ⟩ |f ′ (z0 )|2 ⟨v1 , v2 ⟩ ⟨v , v ⟩

cos φ = = ′ ′
= 1 2 = cos θ.
|w1 | |w2 | |f (z0 )v1 | |f (z0 )v2 | |v1 | |v2 |

So we have shown that under the assumption that f ′ (z0 ) ≠ 0, the function f (z) maps two
curves meeting at an angle θ at z0 to two curves that meet at the same angle at f (z0 ). A
function with this property is said to be conformal at z0 ; see Fig. 1.4.
We can also prove that, under additional assumptions, the converse to the fact that
holomorphicity with a nonvanishing derivative implies conformality also holds, making
holomorphicity and conformality into nearly equivalent concepts. An important addi-
tional condition is that the conformal map needs to be orientation-preserving; this
 1.3 Holomorphicity and the Cauchy–Riemann equations � 17

Figure 1.4: A conformal map f preserves the angle between curves crossing at a point: θ = φ.

condition can be seen to be necessary by considering the map f (z) = z, which is con-
formal but not holomorphic. Recall from vector calculus that for a differentiable vector
planar map f : U → ℝ2 (where U is some open set in ℝ2 ), the Jacobian matrix of f is
the matrix of partial derivatives,

𝜕u 𝜕u
𝜕x 𝜕y
Jf = ( 𝜕v 𝜕v
). (1.21)
𝜕x 𝜕y

If det Jf > 0, then we say that f preserves orientation.

Theorem 1.6. If f = u + iv is holomorphic at z0 and f ′ (z0 ) ≠ 0, then f is conformal at

z0 . Conversely, if f is conformal at z0 , continuously differentiable at z0 in the real analysis
sense, and preserves orientation at z0 , then f is holomorphic at z0 .

The first claim of the theorem was already proved above. The converse direction is
proved with the help of the Cauchy–Riemann equations. First, we will need the following
simple lemma about linear transformations in the plane.

Lemma 1.7 (Linear conformal maps). Assume that A = ( ac db ) is a 2 × 2 real matrix. The
following are equivalent:
(a) A preserves orientation (that is, det A > 0) and is a linear conformal map, that is,
satisfies

⟨Aw1 , Aw2 ⟩ ⟨w1 , w2 ⟩

= (w1 , w2 ∈ ℝ2 \ {(0, 0)}). (1.22)
|Aw1 | |Aw2 | |w1 | |w2 |

(b) A takes the form

a b
A=( ) for some a, b ∈ ℝ with a2 + b2 > 0. (1.23)
−b a
18 � 1 Basic theory

(c) A takes the form

cos θ − sin θ
A = r( ) for some r > 0 and θ ∈ ℝ.
sin θ cos θ

(That is, geometrically A acts by a rotation followed by a scaling.)

Proof that (a) 󳨐⇒ (b). Note that both columns of A are nonzero vectors by the assump-
tion that det A > 0. Now applying assumption (1.22) with w1 = (1, 0)⊤ , w2 = (0, 1)⊤ yields
that (a, c) ⊥ (b, d), so that we must have

(b, d) = κ(−c, a) (1.24)

for some κ ∈ ℝ\{0}. On the other hand, applying (1.22) with w1 = (1, 1)⊤ and w2 = (1, −1)⊤
yields that (a + b, c + d) ⊥ (a − b, c − d), which is easily seen to be equivalent to a2 + c2 =
b2 + d 2 . When combined with (1.24), this implies that κ = ±1. So A is of one of the two
a c
forms ( ac −c
a ) or ( c −a ). Finally, the assumption that det A > 0 means that it is the first of
those two possibilities that must occur.

Proof of the implications (b) ⇐⇒ (c) and (b) 󳨐⇒ (a). This is left as an exercise (Exer-
cise 1.7).

Proof of Theorem 1.6. Assume that f is conformal, continuously differentiable, and

orientation-preserving at z0 . Let γ : (a, b) → ℂ be a differentiable parameterized planar
curve with 0 ∈ (a, b), γ(0) = z0 , and tangent vector v = γ′ (0) at z0 . By standard prop-
erties of differentiable planar maps the tangent vector of f ∘ γ at f (w0 ) is Jf (z0 )v (that
is, the Jacobian matrix of f at z0 acting as a linear map on the vector v, interpreted as
a column vector). This means that f is conformal at z0 if and only if the matrix Jf (z0 ) is
a linear conformal map in the sense of satisfying condition (1.22) in Lemma 1.7(a). Now
adding the knowledge that f is orientation-preserving at z0 , the equivalence stated in
the lemma implies that Jf (z0 ) must be of the form given on the right-hand side of (1.23).
Comparing that form with (1.21), we see that this precisely means that f satisfies the
Cauchy–Riemann equations at z0 . This means that the converse part of Proposition 1.5
applies, and we conclude that f is holomorphic at z0 , as claimed.

Suggested exercises for Section 1.3. 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 1.10, 1.11.

1.4 Additional consequences of the Cauchy–Riemann equations

In the previous section, we saw that the Cauchy–Riemann equations can be used to
prove the near-equivalence between holomorphicity with a nonvanishing derivative
and conformality. Another curious consequence of the Cauchy–Riemann equations,
which gives an alternative geometric picture to that of conformality, is that holomor-
 1.4 Additional consequences of the Cauchy–Riemann equations � 19

phicity implies the orthogonality of the level curves of u and of v. That is, if f = u + iv is
holomorphic, then

⟨∇u, ∇v⟩ = ⟨ (ux , uy ), (vx , vy ) ⟩ = ux vx + uy vy = vy vx − vx vy = 0.

Since ∇u (resp., ∇v) is orthogonal to the level curve {u = c} (resp., the level curve {v = d}),
this proves that the level curves {u = c} and {v = d} meet at right angles whenever they
intersect.
Yet another important and remarkable consequence of the Cauchy–Riemann equa-
tions is that, at least under mild smoothness assumptions (which, as we will see later, can
be removed) in addition to holomorphicity, u and v are harmonic functions. Assume
that f is holomorphic at z and is twice continuously differentiable (in the real analysis
sense) there. Then

𝜕2 u 𝜕2 u 𝜕 𝜕u 𝜕 𝜕u
+ = ( )+ ( )
𝜕x 2 𝜕y2 𝜕x 𝜕x 𝜕y 𝜕y
𝜕 𝜕v 𝜕 𝜕v 𝜕2 v 𝜕2 v
= ( )− ( )= − = 0,
𝜕x 𝜕y 𝜕y 𝜕x 𝜕x𝜕y 𝜕y𝜕x

i. e., u satisfies Laplace’s equation

△u = 0,
2 2
where △ = 𝜕x𝜕
2 + 𝜕y2 is the two-dimensional Laplacian operator. A function that satisfies
𝜕

this equation is called a harmonic function. Similarly (check), v also satisfies

𝜕2 v 𝜕2 v
△v = + = 0.
𝜕x 2 𝜕y2

So we have shown that u and v are harmonic functions. This fact is an important con-
nection between complex analysis, real analysis, and the theory of partial differential
equations.
We will later see that the assumption of f being twice continuously differentiable is
unnecessary, but proving this requires more advanced ideas (see Theorem 1.30 in Sec-
tion 1.9).
A final remark related to holomorphicity and the Cauchy–Riemann equations is the
observation that if f = u + iv is holomorphic, then its Jacobian matrix is given by

ux uy 󵄨2
) = ux vy − uy vx = ux2 + v2x = |ux + ivx | = 󵄨󵄨󵄨f ′ (z)󵄨󵄨󵄨 .
󵄨
Jf = det ( (1.25)
vx vy

This can also be understood geometrically—spend a moment thinking what the geomet-
ric interpretation is.

Suggested exercises for Section 1.4. 1.12.

20 � 1 Basic theory

1.5 Power series

Until now we have not discussed any specific examples of functions of a complex vari-
able. Of course, there are the standard functions that you probably encountered already
in your undergraduate studies: polynomials, rational functions, ez , the trigonometric
functions, etc. Aside from these examples, it would be useful to have a general way
to construct a large family of functions. Of course, there is such a way: power series,
which—nonobviously—turn out to be essentially as general a family of functions as one
could hope for.
To make things precise, a power series is a function of a complex variable z defined
by
∞
f (z) = ∑ an (z − z0 )n , (1.26)
n=0

where z0 ∈ ℂ, and (an )∞

n=0 is a sequence of complex numbers. This function is defined
wherever the respective series converges.
For which values of z does this formula make sense? Define the number R ∈ [0, ∞]
as
−1
R = (lim sup |an |1/n ) ,
n→∞

which we refer to as the radius of convergence of the power series. Its significance is
explained in the following simple result.

Lemma 1.8. 1. The series (1.26) converges absolutely if |z − z0 | < R.

2. The series (1.26) diverges for all z satisfying |z − z0 | > R.

Proof. We assume that 0 < R < ∞; the edge cases R = 0 and R = ∞ are left as an
exercise (Exercise 1.13). The defining property of R is that for all ϵ > 0, we have that
|an | < ( R1 + ϵ)n if n is large enough, and R is the maximal number with that property. Let
z ∈ DR (0). Since |z| < R, we have |z|( R1 + ϵ) < 1 for some fixed ϵ > 0 chosen small enough.
This implies that for all n > N (for some large enough N that depends on ϵ),

∞ ∞ n
1
∑ 󵄨󵄨󵄨an zn 󵄨󵄨󵄨 < ∑ [( + ϵ)|z|] ,
󵄨 󵄨
n=N n=N
R

so the series is dominated by a convergent geometric series and hence converges.

Conversely, if |z| > R, then |z|( R1 − ϵ) > 1 for some small enough fixed ϵ > 0. Taking a
1 nk
subsequence (ank )∞ k=1 for which |ank | > ( R − ϵ) for all k (such a subsequence exists by
the definition of R), we see that
nk
󵄨󵄨 n 󵄨 1
󵄨󵄨ank z k 󵄨󵄨󵄨 ≥ [|z|( − ϵ)] > 1,
R
 1.5 Power series � 21

that is, the power series (1.26) contains infinitely many terms with modulus > 1 and
hence diverges.
Another important property of power series is given in the following theorem.

Theorem 1.9 (Power series are holomorphic). Power series are holomorphic functions in
the interior of their disc of convergence and can be differentiated termwise there; that is,
the derivative of the infinite series is equal to the series of the derivatives.

Proof. Denote
∞
f (z) = ∑ an zn = SN (z) + RN (z), where
n=0
N ∞
SN (z) = ∑ an zn , RN (z) = ∑ an zn ,
n=0 n=N+1

and let
∞
g(z) = ∑ nan zn−1 .
n=1

The claim is that f is differentiable on the disc of convergence and that its derivative is
the power series g. Since n1/n → 1 as n → ∞, it is easy to see that f (z) and g(z) have the
f (z +h)−f (z )
same radius of convergence. Fix z0 with |z0 | < r < R. We wish to show that 0 h 0
converges to g(z0 ) as h → 0. Observe that

f (z0 + h) − f (z0 ) S (z + h) − SN (z0 )

− g(z0 ) = ( N 0 − SN′ (z0 ))
h h
R (z + h) − RN (z0 )
+ N 0 + (SN′ (z0 ) − g(z0 )). (1.27)
h
In this last expression, the first term converges to 0 as h → 0 for any fixed N. To bound
the second term, fix some ϵ > 0, and assume that |h| < r, and moreover that |h| is small
enough so that |z0 + h| < r. Now make use of the algebraic identity

pn − qn = (p − q)(pn−1 + pn−2 q + ⋅ ⋅ ⋅ + pqn−2 + qn−1 )

to get that
󵄨󵄨 R (z + h) − R (z ) 󵄨󵄨 ∞ 󵄨󵄨 (z + h)n − zn 󵄨󵄨
󵄨󵄨 N 0 N 0 󵄨󵄨 󵄨 0 0 󵄨󵄨
󵄨󵄨 󵄨󵄨 ≤ ∑ |an | 󵄨󵄨󵄨 󵄨󵄨
󵄨󵄨 h 󵄨󵄨 n=N+1 󵄨󵄨 h 󵄨󵄨
∞ 󵄨󵄨 h ∑n−1 hk (z + h)n−1−k 󵄨󵄨
󵄨 k=0 0 󵄨󵄨
= ∑ |an | 󵄨󵄨󵄨 󵄨󵄨
n=N+1
󵄨󵄨 h 󵄨󵄨
∞
≤ ∑ |an |nr n−1 .
n=N+1
22 � 1 Basic theory

The last expression in this chain of inequalities is the tail of an absolutely convergent
series, so it can be made < ϵ be taking N large enough (before taking the limit as h → 0).
Third, we have the limit SN′ (z0 ) → g(z0 ) as N → ∞, so we can choose N large
enough so that |SN′ (z0 ) − g(z0 )| < ϵ. Having thus chosen N, we get finally from (1.27) and
the above estimates that
󵄨󵄨 f (z + h) − f (z ) 󵄨󵄨
lim sup󵄨󵄨󵄨 0 0
󵄨 󵄨
− g(z0 )󵄨󵄨󵄨 ≤ 0 + ϵ + ϵ = 2ϵ.
h→0 󵄨 󵄨 h 󵄨󵄨

f (z0 +h)−f (z0 )

Since ϵ was an arbitrary positive number, this shows that h
→ g(z0 ) as h → 0,
as claimed.

Corollary 1.10. Holomorphic functions defined as power series are differentiable (in the
complex-analytic sense) infinitely many times in the disc of convergence.
n
Corollary 1.11. For a power series g(z) = ∑∞
n=0 an (z − z0 ) with positive radius of conver-
gence, we have

g (n) (z0 )
an = . (1.28)
n!

In other words, g(z) satisfies Taylor’s formula

∞
g (n) (z0 )
g(z) = ∑ (z − z0 )n .
n=0
n!

Suggested exercises for Section 1.5. 1.13, 1.14, 1.15, 1.16.

1.6 Contour integrals

We now introduce contour integrals, which are another fundamental building block
of the theory.
Contour integrals, like many other types of integrals, take as input a function to be
integrated and a “thing” (or “place”) over which the function is integrated. In the case of
contour integrals, the “thing” is a contour, which is (for our current purposes at least) a
kind of planar curve. We start by developing some terminology to discuss such objects.
A parameterized curve is a continuous function γ : [a, b] → ℂ. The value γ(a) is called
the starting point, and γ(b) is called the ending point (both a, b together are referred
to as the endpoints). Two curves γ1 : [a, b] → ℂ, γ2 : [c, d] → ℂ are called equivalent,
denoted γ1 ∼ γ2 , if γ2 (t) = γ1 (I(t)) where I : [c, d] → [a, b] is a continuous, one-to-one,
onto, increasing function. A curve γ is called simple if it does not intersect itself, that is,
if γ is injective. It is called closed if γ(a) = γ(b).
 1.6 Contour integrals � 23

What we will refer to as a curve is, formally speaking, an equivalence class of pa-
rameterized curves with respect to the equivalence relation defined above. We also use
the word contour as a synonym for curve.
In practice, we will usually refer to parameterized curves simply as “curves,” which
is the usual abuse of terminology that one sees in various places in mathematics, in
which one blurs the distinction between equivalence classes and their members, re-
membering that various definitions, notation, and proof arguments need to “respect
the equivalence” in the sense that they do not depend of the choice of a member. (As a
meta exercise, try to think of other examples of this phenomenon you might have en-
countered in your studies.)
For our present context of developing the theory of complex analysis, we will as-
sume that all our curves are piecewise continuously differentiable. More generally, we
can assume them to be rectifiable, but we will not bother to develop that theory. There
are yet more general contexts in which allowing curves to be merely continuous is ben-
eficial (and indeed some of the ideas we will develop in a complex-analytic context can
be carried over to that more general setting), but we will not pursue such distractions
either.
You probably encountered curves and parameterized curves in your earlier studies
of multivariate calculus, where they were used to define the notion of line integrals
of vector and scalar fields. Recall that there are two types of line integrals, which are
referred to as line integrals of the first and second kind. The line integral of the first
kind of a scalar (usually real-valued) function u(z) over a curve γ is defined as

n
∫ u(z) ds = lim ∑ u(zj )Δsj , (1.29)
max Δsj →0
γ j j=1

where the limit is a limit of Riemann sums with respect to a family of tagged partitions of
the interval [a, b] over which the curve γ is defined as the norm of the partitions shrinks
to 0. Such a partition consists of partition points

a = t0 < t1 < ⋅ ⋅ ⋅ < tn = b,

and each partition subinterval [tj−1 , tj ] is “tagged” or marked with an arbitrary point τj
chosen from the subinterval. Given this partition, we denote zj = γ(τj ), and the symbols
Δsj refer to finite line elements, namely Δsj = |zj − zj−1 |. This notation gives meaning to
the right-hand side of (1.29).
The line integral of the second kind is defined for a vector field F = (P, Q) (using
the more traditional notation from calculus; in the complex analysis context, we would
regard this object as the complex-valued function F = P + iQ) by

n
∫ F ⋅ ds = ∫ P dx + Q dy = lim ∑ P(zj )Δxj + Q(zj )Δyj ,
max Δsj →0
γ γ j j=1
24 � 1 Basic theory

where the numbers zj are associated with the tagged partition as above, and

xj = Re(zj ), yj = Im(zj ), Δxj = xj − xj−1 , Δyj = yj − yj−1 .

It is well known from calculus that line integrals can be expressed in terms of ordi-
nary (single-variable) Riemann integrals. Take a couple of minutes to remind yourself
of why the following formulas are true (assuming that all the functions involved are
piecewise continuously differentiable):

b
󵄨 󵄨
∫ u(z) ds = ∫ u(γ(t))󵄨󵄨󵄨γ′ (t)󵄨󵄨󵄨 dt, (1.30)
γ a
b

∫ F ⋅ ds = ∫ F(γ(t)) ⋅ γ′ (t) dt. (1.31)

γ a

(In (1.31), “⋅” refers to the dot product of vectors in the plane.)
As a further reminder, the basic result known as the fundamental theorem of cal-
culus for line integrals states that if F = ∇u, then

∫ F ⋅ ds = u(γ(b)) − u(γ(a)).
γ

We are now ready to define contour integrals and arc length integrals, which are
the complex-analytic analogues of line integrals of the first and second kinds (and are
defined in terms of those integrals). For a function f = u + iv of a complex variable z and
a curve γ, the contour integral ∫γ f (z) dz (in words: the integral of f over the curve γ) is
defined, loosely speaking, as the line integral of the second kind “∫γ (u + iv)(dx + idy)”.
More precisely, expanding this product of a complex number and a complex differential
and separating into real and imaginary components, this definition becomes

∫ f (z) dz = (∫ u dx − v dy) + i(∫ v dx + u dy), (1.32)

γ γ γ

that is, the complex number whose real part is the line integral of F ⋅ ds and whose
imaginary part is the line integral of G⋅ds, where F and G are the vector fields F = (u, −v)
and G = (v, u). Appealing to (1.31), you can check easily that the contour integral can be
evaluated explicitly as the ordinary Riemann integral

∫ f (z) dz = ∫ f (γ(t))γ′ (t) dt. (1.33)

γ a

Similarly, the arc length integral is defined as

 1.6 Contour integrals � 25

∫ f (z) |dz| = ∫ f (z) ds = ∫ u ds + i ∫ v ds, (1.34)

γ γ γ γ

which is simply a line integral of the first kind in which the integrand is complex-valued.
If γ is a closed curve, then we denote the contour integral as ∮γ f (z) dz, and similarly
∮γ f (z) |dz| for the arc length integral.
A particular case of an arc length integral is the length of the curve, denoted len(γ)
and defined as the integral of the constant function 1:

b
󵄨 󵄨
len(γ) = ∫ |dz| = ∫󵄨󵄨󵄨γ′ (t)󵄨󵄨󵄨 dt.
γ a

As mentioned above, our convention of mildly abusing terminology puts on us the

burden of having to remember to check that these definitions do not depend on the
parameterization of the curve. Indeed, if γ1 ∼ γ2 are representatives of the same equiv-
alence class of parameterized curves, that is, γ2 (t) = γ1 (I(t)) for some nicely behaved
function, then using a standard change of variables in single-variable integrals, we see
that

d d

∫ f (z) dz = ∫ f (γ2 (t))γ2′ (t)dt = ∫ f (γ1 (I(t)))(γ1 ∘ I)′ (t) dt

γ2 c c
d b

= ∫ f (γ1 (I(t)))γ1′ (I(t))I ′ (t) dt = ∫ f (γ1 (τ))γ1′ (τ) dτ = ∫ f (z) dz. (1.35)
c a γ1

The analogous verification in the case of arc length integrals is left as an exercise
(Exercise 1.17).
Contour integrals have many surprising properties, but the ones on the following
list of basic properties are not of the surprising kind.

Proposition 1.12 (properties of contour integrals). Contour integrals satisfy the following
properties:
(a) Linearity as an operator on functions: for functions f (z), g(z) and complex numbers
α, β, we have

∫(αf (z) + βg(z)) dz = α ∫ f (z) dz + β ∫ g(z) dz.

γ γ γ

(b) Linearity as an operator on curves: if a contour Γ is a “composition” of two contours

γ1 and γ2 (in a sense that is easy to define graphically but tedious to write down pre-
cisely), then
26 � 1 Basic theory

∫ f (z) dz = ∫ f (z) dz + ∫ f (z) dz.

Γ γ1 γ2

Similarly, if γ2 is the “reverse” contour of γ1 , then

∫ f (z) dz = − ∫ f (z) dz.

γ2 γ1

(c) Triangle inequality:

󵄨󵄨 󵄨󵄨
󵄨󵄨 󵄨 󵄨 󵄨 󵄨 󵄨
󵄨󵄨∫ f (z) dz󵄨󵄨󵄨 ≤ ∫󵄨󵄨󵄨f (z)󵄨󵄨󵄨 |dz| ≤ len(γ) ⋅ sup󵄨󵄨󵄨f (z)󵄨󵄨󵄨.
󵄨󵄨 󵄨󵄨 z∈γ
γ

Proof. Exercise 1.18.

Contour integrals have their own version of the fundamental theorem of calculus.

Theorem 1.13 (The fundamental theorem of calculus for contour integrals). If γ is a curve
connecting two points w1 and w2 in a region Ω on which a function F is holomorphic, then

∫ F ′ (z) dz = F(w2 ) − F(w1 ).

γ

Equivalently, the theorem says that to compute a general contour integral ∫γ f (z) dz,
we try to find a primitive of f , that is, a holomorphic function F such that F ′ (z) = f (z)
on all of Ω. (A term synonymous with “primitive” is antiderivative.) If we found such a
primitive, then the contour integral ∫γ f (z) dz is given by F(w2 ) − F(w1 ).

Proof. For smooth curves, an easy application of the chain rule gives

b b

∫ F ′ (z) dz = ∫ F ′ (γ(t))γ′ (t) dt = ∫(F ∘ γ)′ (t) dt = (F ∘ γ)(t)|t=b

t=a
γ a a

= F(γ(b)) − F(γ(a)) = F(w2 ) − F(w1 ).

For piecewise smooth curves, this is a trivial extension that is left to the reader.

Many of our discussions of contour integrals will involve the behavior of integrals
over closed contours and the interplay between the properties of such integrals and
integrals over general contours. As an example of this interplay, the above result has an
easy—but important—consequence for integrals over closed contours.

Corollary 1.14. If f = F ′ where F is holomorphic on a region Ω—that is, f has a

primitive—then for any closed contour γ in Ω, we have

∮ f (z) dz = 0.
γ
 1.6 Contour integrals � 27

This last result has the following partial converse.

Proposition 1.15. If f : Ω → ℂ is a continuous function on a region Ω such that

∮ f (z) dz = 0
γ

for any closed contour in Ω, then f has a primitive.

Proof. Fix some z0 ∈ Ω. For any z ∈ Ω, there is some curve γ(z0 , z) connecting z0 and
z (since Ω is connected and open, hence pathwise-connected—a standard exercise in
topology). Moreover, it is also not hard to see that the curve can be assumed to be piece-
wise differentiable. Define

F(z) = ∫ f (w) dw. (1.36)

γ(z0 ,z)

By the assumption this integral does not depend on which curve γ(z0 , z) connecting z0
and z was chosen, so F(z) is well-defined. We now claim that F is holomorphic and its
derivative is equal to f . To see this, note that if h is a complex number such that z+h ∈ Ω,
then

F(z + h) − F(z)
− f (z)
h
1
= ( ∫ f (w) dw − ∫ f (w) dw) − f (z)
h
γ(z0 ,z+h) γ(z0 ,z)
1 1
= ∫ f (w) dw − f (z) = ∫ (f (w) − f (z)) dw, (1.37)
h h
γ(z,z+h) γ(z,z+h)

where γ(z, z + h) denotes a curve in Ω connecting z and z + h. When |h| is sufficiently

small so that the disc Dh (z) is contained in Ω, we can take γ(z, z + h) as the straight line
segment connecting z and z + h. For such h, we get that

󵄨󵄨 F(z + h) − F(z) 󵄨󵄨 1
󵄨󵄨 󵄨 󵄨 󵄨
󵄨󵄨 − f (z)󵄨󵄨󵄨 ≤ len(γ(z, z + h)) sup 󵄨󵄨󵄨f (w) − f (z)󵄨󵄨󵄨
󵄨󵄨 h 󵄨󵄨 h w∈Dh (z)
󵄨 󵄨
= sup 󵄨󵄨󵄨f (w) − f (z)󵄨󵄨󵄨 󳨀󳨀󳨀󳨀→ 0
w∈Dh (z) h→0

by the continuity of f .

Lemma 1.16. If f is holomorphic on Ω and f ′ ≡ 0, then f is a constant.

Proof. Fix some z0 ∈ Ω. For any z ∈ Ω, as we discussed above, there is a path γ(z0 , z)
connecting z0 and z. Then
28 � 1 Basic theory

f (z) − f (z0 ) = ∫ f ′ (w) dw = 0,

γ(z0 ,z)

and hence f (z) ≡ f (z0 ), that is, f is constant.

Suggested exercises for Section 1.6. 1.17, 1.18.

1.7 The Cauchy, Goursat, and Morera theorems

One of the central results in complex analysis is Cauchy’s theorem.

Theorem 1.17 (Cauchy’s theorem.). If f is a holomorphic function on a simply connected

region Ω, then for any closed curve in Ω, we have

∮ f (z) dz = 0.
γ

The challenges facing us are as follows: first, to prove Cauchy’s theorem for curves
and regions that are relatively simple (where we do not have to deal with subtle topolog-
ical considerations); second, to define what “simply connected” means; third, to extend
the theorem to the most general setting. This is done in the next section.
Two other theorems closely related to Cauchy’s theorem are Goursat’s theorem, a
relatively easy particular case of Cauchy’s theorem, and Morera’s theorem, which is a
kind of converse to Cauchy’s theorem.

Theorem 1.18 (Goursat’s theorem). If f is holomorphic on a region Ω, T is a triangle con-

tained in Ω, and 𝜕T is the boundary of T (considered as a curve in the usual sense), then

∮ f (z) dz = 0. (1.38)
𝜕T

Theorem 1.19 (Morera’s theorem). If f : Ω → ℂ is a continuous function on a region Ω

such that

∮ f (z) dz = 0
γ

for any closed contour in Ω, then f is holomorphic on Ω.

Morera’s theorem is proved in Section 1.9.

Proof of Goursat’s theorem. The proof can be summarized with a slogan “localize the
damage.” Namely, try to translate a global statement about the integral around the tri-
angle to a local statement about behavior near a specific point inside the triangle, which
would become manageable since we have a good understanding of the local behavior of
 1.7 The Cauchy, Goursat, and Morera theorems � 29

a holomorphic function near a point. If something goes wrong with the global integral,
then something has to go wrong at the local level, and we will show that cannot happen.
(Although technically the proof is not a proof by contradiction, conceptually I find this
a helpful way to think about it).
The idea can be made more precise using triangle subdivision. Specifically, let T (0) =
T, and define a hierarchy of subdivided triangles:

order 0 triangle: T (0) ,

order 1 triangles: Tj(1) , 1 ≤ j ≤ 4,
(2)
order 2 triangles: Tj,k , 1 ≤ j, k ≤ 4,
(3)
order 3 triangles: Tj,k,ℓ , 1 ≤ j, k, ℓ ≤ 4,
..
.
order n triangles: Tj(n)
,...,j
, 1 ≤ j1 , . . . , jn ≤ 4.
1 n

..
.

Here the triangles Tj(n)

,...,j
for jn = 1, 2, 3, 4 are obtained by subdividing the order-(n − 1)
1 n

triangle Tj(n−1)
,...,j
into 4 subtriangles whose vertices are the vertices and/or edge bisectors
1 n−1

of Tj(n−1)
,...,j
; see Fig. 1.5.
1 n−1

Figure 1.5: The triangle T = T (0) and the first few steps in its hierarchy of subdivided triangles.
30 � 1 Basic theory

Now, given the way this subdivision was done, it is clear that we have the relation

4
∮ f (z) dz = ∑ ∮ f (z) dz
jn =1
𝜕Tj(n−1) 𝜕Tj(n),...,j
1 ,...,jn−1 1 n

(where 𝜕Tj(n) refers as before to the boundary of the triangle Tj(n) , considered as a
1 ,...,jn 1 ,...,jn
curve oriented in the positive sense) due to cancelation along the internal edges, and
hence
4
∮ f (z) dz = ∑ ∮ f (z) dz.
j1 ,...,jn =1
𝜕T (0) 𝜕Tj(n),...,j
1 n

So the contour integral around the boundary of the original triangle is equal to the sum
of the integrals around all 4n triangles at the nth subdivision level. Now a key obser-
vation is that one of these integrals has to have a modulus that is at least as big as the
average, that is, there exists an n-tuple j(n) = (j1(n) , . . . , jn(n) ) ∈ {1, 2, 3, 4}n for which

󵄨󵄨 󵄨󵄨 4 󵄨󵄨 󵄨󵄨 󵄨 󵄨󵄨
󵄨󵄨 󵄨 󵄨 󵄨 n 󵄨󵄨 󵄨
󵄨󵄨 ∮ f (z) dz󵄨󵄨󵄨 ≤ ∑ 󵄨󵄨󵄨 ∮ f (z) dz󵄨󵄨󵄨 ≤ 4 󵄨󵄨󵄨 ∮ f (z) dz󵄨󵄨󵄨. (1.39)
󵄨󵄨 󵄨󵄨 󵄨󵄨 󵄨󵄨 󵄨󵄨 󵄨󵄨
(0) j1 ,...,jn =1 (n) (n)
𝜕T 𝜕Tj 𝜕Tj(n)
1 ,...,jn

Moreover, we can choose j(n) inductively in such a way that the triangles Tj(n)
(n)
are nested,
that is, Tj(n)
(n)
⊂ Tj(n−1)
(n−1)
for n ≥ 1, or, equivalently, j(n) = (j1(n−1) , . . . , jn−1
(n−1)
, k) for some
1 ≤ k ≤ 4. To make this happen, choose a value of k for which | ∮𝜕T (n) f (z) dz| is
(j(n−1),k)
greater than or equal to the average

1 4 󵄨󵄨󵄨󵄨 󵄨󵄨
󵄨
∑󵄨 ∮ f (z) dz󵄨󵄨󵄨,
4 d=1󵄨󵄨󵄨 󵄨󵄨
(n)
𝜕T(j(n−1),d)

which in turn can be seen (by induction) to be greater than or equal to

󵄨 󵄨󵄨
1 󵄨󵄨󵄨󵄨 4 󵄨󵄨 1 󵄨󵄨󵄨 󵄨󵄨 1
󵄨
󵄨∑ ∮ f (z) dz󵄨󵄨󵄨 = 󵄨󵄨󵄨 ∮ f (z) dz󵄨󵄨󵄨 ≥ ⋅ 4−(n−1) ∮ f (z) dz,
4 󵄨󵄨󵄨󵄨d=1 󵄨󵄨 4 󵄨󵄨
󵄨
󵄨󵄨 4
(n)
𝜕T(j(n−1),d) (n−1)
𝜕Tj(n−1) 𝜕T

thereby justifying (1.39).

We now claim that the sequence of nested triangles Tj(n)
(n)
shrinks to a single point,
that is, we have
∞
(n)
⋂ Tj(n) = {z0 }
n=0
 1.7 The Cauchy, Goursat, and Morera theorems � 31

for some point z0 ∈ T. Indeed, the diameter of the triangles goes to 0 as n → ∞, so

certainly there cannot be two distinct points in the intersection. On the other hand, the
triangles Tj(n)
(n)
are all compact, and the finite intersections ⋂Nn=0 Tj(n)
(n)
are nonempty, so by
the standard finite intersection property of compact sets the full intersection ⋂∞
n=0 Tj(n)
(n)

is also nonempty.
Having defined z0 , write f (z) for z near z0 as

f (z) = f (z0 ) + f ′ (z0 )(z − z0 ) + ψ(z)(z − z0 ),

where
f (z) − f (z0 )
ψ(z) = − f ′ (z0 ).
z − z0

The holomorphicity of f at z0 implies that ψ(z) → 0 as z → z0 . Denote by d (n) the diam-

eter of Tj(n)
(n)
and by p(n) its perimeter. Each subdivision shrinks both the diameter and
perimeter by a factor of 2, so we have

d (n) = 2−n d (0) , p(n) = 2−n p(0) .

It follows that
󵄨󵄨 󵄨󵄨 󵄨󵄨 󵄨󵄨
󵄨󵄨 󵄨 󵄨 ′ 󵄨
󵄨󵄨 ∫ f (z) dz󵄨󵄨󵄨 = 󵄨󵄨󵄨 ∫ (f (z0 ) + f (z0 )(z − z0 ) + ψ(z)(z − z0 )) dz󵄨󵄨󵄨
󵄨󵄨 󵄨󵄨 󵄨󵄨 󵄨󵄨
(n) (n)
𝜕Tj(n) 𝜕Tj(n)
󵄨󵄨 󵄨󵄨
󵄨 󵄨 󵄨 󵄨
= 󵄨󵄨󵄨 ∫ ψ(z)(z − z0 ) dz󵄨󵄨󵄨 ≤ p(n) d (n) sup 󵄨󵄨󵄨ψ(z)󵄨󵄨󵄨
󵄨󵄨 󵄨󵄨
(n) z∈T (n)
j(n)
𝜕Tj(n)
󵄨 󵄨
= 4−n p(0) d (0) sup 󵄨󵄨󵄨ψ(z)󵄨󵄨󵄨.
z∈Tj(n)
(n)

This estimate allows us to finish, since combining it with (1.39), we get that
󵄨󵄨 󵄨󵄨
󵄨󵄨 󵄨 (0) (0) 󵄨 󵄨
󵄨󵄨 ∫ f (z) dz󵄨󵄨󵄨 ≤ p d sup 󵄨󵄨󵄨ψ(z)󵄨󵄨󵄨 󳨀󳨀󳨀󳨀󳨀→ 0,
󵄨󵄨 󵄨󵄨 n→∞
(0) z∈T (n)
j(n)
𝜕T

which establishes (1.38).

The next few results illustrate how Goursat’s theorem, for all its apparent simplicity,
can be used to quickly derive even stronger versions of Cauchy’s theorem, gradually
building up our knowledge toward the general version that will be proved in the next
section.

Corollary 1.20 (Goursat’s theorem for rectangles). Theorem 1.18 is also true when we re-
place the word “triangle” with “rectangle.”
32 � 1 Basic theory

Proof. Obviously, a rectangle can be decomposed as the union of two triangles, with the
contour integral around the rectangle being the sum of the integrals around the two
triangles due to cancelation of the integrals going in both directions along the diagonal.

Corollary 1.21 (existence of a primitive for a holomorphic function on a disc). If f is holo-

morphic on a disc D, then f = F ′ for some holomorphic function F on D.

Proof. The claim is identical to Proposition 1.15, but with a different set of assumptions.
In fact, the proof of that proposition can be easily adapted to prove the existence of a
primitive in the current setting. Specifically, we again define the purported primitive F
for f using (1.36), but this time using a particular choice of path γ(z0 , z) connecting z0
and z, namely, we take γ(z0 , z) to be the straight line segment from z0 to z.
We now claim that with this definition, for h small in magnitude (so that z + h is
still in the disc D), the chain of equalities (1.37) still holds, where in this chain, we also
interpret γ(z, z + h) as the straight line segment connecting z and z + h. If we can show
this, then the rest of the proof carries through as before. Now, upon inspection of (1.37),
we see that the first and third equalities still hold trivially; it is only the middle equality
that needs to be explained. This equality can be rewritten as

∫ f (w) dw + ∫ f (w) dw − ∫ f (w) dw = 0,

γ(z0 ,z) γ(z,z+h) γ(z0 ,z+h)

a relationship between the contour integrals of f along the three straight line segments
γ(z0 , z), γ(z0 , z + h), and γ(z, z + h). This is simply the statement that the contour integral
along the boundary of the triangle with vertices z0 , z, and z + h is 0, which follows from
Goursat’s theorem.

Theorem 1.22 (Cauchy’s theorem for a disc). If f is holomorphic on a disc, then ∮γ f dz = 0

for any closed contour γ in the disc.

Proof. By Corollary 1.21, f has a primitive, so Corollary 1.14 implies the claimed conse-
quence.

1.8 Simply connected regions and the general version of Cauchy’s

theorem
We now develop the additional concepts required to formulate and prove the general
version of Cauchy’s theorem. A key notion is that of homotopy of curves. Given a region
Ω ⊂ ℂ, two parameterized curves γ1 , γ2 : [0, 1] → Ω (assumed for simplicity of notation
to be defined on [0, 1]) are said to be homotopic (with fixed endpoints) if γ1 (0) = γ2 (0),
γ1 (1) = γ2 (1), and there exists a function F : [0, 1] × [0, 1] → Ω such that
i) F is continuous.
 1.8 Simply connected regions and Cauchy’s theorem � 33

ii) F(0, t) = γ1 (t) for all t ∈ [0, 1].

iii) F(1, t) = γ2 (t) for all t ∈ [0, 1].
iv) F(s, 0) = γ1 (0) for all s ∈ [0, 1].
v) F(s, 1) = γ1 (1) for all s ∈ [0, 1].

The map F is called a homotopy between γ1 and γ2 . Intuitively, for each s ∈ [0, 1], the
function Fs : t 󳨃→ F(s, t) defines a curve connecting the two endpoints γ1 (0) and γ1 (1). As
s grows from 0 to 1, this family of curves transitions in a continuous way between the
curve γ1 and γ2 with the endpoints being fixed in place; see Fig. 1.6.

Figure 1.6: A homotopy between two curves γ1 and γ2 , visualized as a one-parameter family of curves
t 󳨃→ Fs (t) that interpolate continuously between γ1 and γ2 , with the endpoints staying fixed.

A common alternative way to define the notion of homotopy of curves is for closed
curves, where the endpoints are not fixed, but the homotopy must keep the curves closed
as it is deforming them. The definition of a simply connected region then becomes a
region in which any two closed curves are homotopic. It is not hard to show that those
two definitions are equivalent.
It is easy (but recommended!) to check that the relation of being homotopic is an
equivalence relation; see Exercise 1.19.
Next, we define the notion of a simply connected region. A region Ω is called simply
connected if any two curves γ1 , γ2 in Ω with the same endpoints are homotopic. Note that
this is a topological property (in the sense that it is preserved under homeomorphism).
The complex plane, the unit disc, and any region homeomorphic to the unit disc are
simply connected regions (Exercise 1.20).

Theorem 1.23. If f is a holomorphic function on a region Ω, and γ0 ,γ1 are two curves on
Ω with the same endpoints that are homotopic, then

∫ f (z) dz = ∫ f (z) dz.

γ0 γ1

Proof. As with the proof of Goursat’s theorem in the previous section, this proof is based
on the idea of reducing the global statement about the equality of the two contour in-
34 � 1 Basic theory

tegrals into a local statement. Denote by F : [0, 1] × [0, 1] → Ω the homotopy between
γ0 and γ1 , and for any s ∈ [0, 1], denote by γs : [0, 1] → ℂ the curve γs (t) = F(s, t). The
strategy of the proof is to show that there are values 0 = s0 < s1 < s2 < ⋅ ⋅ ⋅ < sn = 1 such
that

∫ f (z) dz = ∫ f (z) dz = ⋅ ⋅ ⋅ = ∫ f (z) dz = ∫ f (z) dz.

γs0 γs1 γsn−1 γsn

In fact, we can take sk = k/n for 0 ≤ k ≤ n with large n; we will define n more precisely
below. Fix 1 ≤ k ≤ n. To prove the equality between the two integrals ∫γ f (z) dz and
sk−1

∫γ f (z) dz, we decompose each of the two integrals into a sum of integrals over small
sk
pieces of the contours γsk−1 and γsk by writing them as

m
∫ f (z) dz = ∑ ∫ f (z) dz, (1.40)
γsk−1 j=1 γ
sk−1 |[tj−1 ,tj ])

m
∫ f (z) dz = ∑ ∫ f (z) dz. (1.41)
γsk j=1 γ
sk |[tj−1 ,tj ]

Here γsk |[tj−1 ,tj ] denotes the restriction of the contour γ to the interval [tj−1 , tj ], where tj
denotes some sequence of points 0 = t0 < t1 < ⋅ ⋅ ⋅ < tn = 1 partitioning [0, 1] into
subintervals [tj−1 , tj ]. We will show at the end of the proof that the partition tj = j/n for
0 ≤ j ≤ n − 1, where n is large (and is the same n that was used for the definition of
sk above), works well for our purposes. Specifically, we will show that with the way we
defined sk and tj above and with n taken sufficiently large, the following assumption is
satisfied: for all 1 ≤ k, j ≤ n, there exists an open disc Dk,j ⊂ Ω containing the two curve
segments γsk−1 |[tj−1 ,tj ] and γsk |[tj−1 ,tj ] .
Under this assumption, to prove that the two integrals (1.40)–(1.41) are equal, it suf-
fices to prove that for any 1 ≤ j ≤ n, we have the equality

∫ f (z) dz = ∫ f (z) dz (1.42)

γsk−1 |[tj−1 ,tj ]) γsk |[tj−1 ,tj ]

between the integrals over the small subcontours.

For each 0 ≤ j ≤ n, let ηk,j denote a straight line segment (considered as a param-
eterized curve) from γsk−1 (tj ) to γsk (tj ), and for each 1 ≤ j ≤ m, let Γk,j denote the closed
curve γsk−1 ([tj−1 , tj ]) + ηk,j − γsk ([tj−1 , tj ]) − ηk,j−1 (in words: the concatenation of the four
curves γsk−1 ([tj−1 , tj ]), ηk,j , “the reverse of γsk ([tj−1 , tj ]),” and “the reverse of ηk,j−1 ”). By
the assumption on the disc Dk,j the curve Γk,j is contained in Dk,j . Therefore by Cauchy’s
theorem for discs (Theorem 1.22) we have
 1.8 Simply connected regions and Cauchy’s theorem � 35

∮ f (z) dz = 0,
Γk,j

or, more explicitly,

∫ f (z) dz − ∫ f (z) dz = ∫ f (z) dz − ∫ f (z) dz.

γsk−1 |[tj−1 ,tj ] γsk |[tj−1 ,tj ] ηk,j−1 ηk,j

Summing this relation over j and recalling (1.40)–(1.41), we get that

m
∫ f (z) dz − ∫ f (z) dz = ∑( ∫ f (z) dz − ∫ f (z) dz)
γsk−1 γsk j=1 ηk,j−1 ηk,j

= ∫ f (z) dz − ∫ f (z) dz = 0.
ηk,0 ηk,m

(Here, in the next-to-last step the sum is telescoping, and in the last step, we note that
ηk,0 and ηk,m are both degenerate curves, each of which simply stays at a single point.)
This is precisely equality (1.42) we wanted.
It remains to justify the assumption about the discs Dk,j . This is done as follows.
First, since the set A = F([0, 1] × [0, 1]) is compact, it is easy to see (for example, using
the Heine–Borel property) that there exists a number ϵ > 0 such that the discs Dϵ (z) are
contained in Ω for all z ∈ A. Second, since F is continuous, and hence also uniformly
continuous, on [0, 1] × [0, 1], there exists a number δ > 0 such that for any 0 ≤ s, t ≤ 1
with |s − s′ | + |t − t ′ | < δ, we have

󵄨󵄨 ′ ′ 󵄨 󵄨 ′ ′ 󵄨
󵄨󵄨γs (t ) − γs (t)󵄨󵄨󵄨 = 󵄨󵄨󵄨F(s , t ) − F(s, t)󵄨󵄨󵄨 < ϵ.

Let n be an integer larger than 2/δ, and let sk = k/n and tj = j/n as before. We define the
discs Dk,j by Dk,j = Dϵ (γsk−1 (tj−1 )) and claim that they satisfy our assumption. Indeed, if
t ∈ [tj−1 , tj ], then |t−tj−1 | ≤ 1/n < δ/2, so |γsk−1 (t)−γsk−1 (tj−1 )| < ϵ. This shows that the curve
segment γsk−1 |[tj−1 ,tj ] is contained in Dk,j . Similarly, |t − tj−1 | + |sk − sk−1 | ≤ 1/n + 1/n < δ, so
|γsk (t) − γsk−1 (tj−1 )| < ϵ, that is, the curve segment γsk |[tj−1 ,tj ] is also contained in Dk,j . This
proves that our assumption about the discs Dk,j is satisfied and finishes the proof.

Theorem 1.24 (Cauchy’s theorem, general version). If f is holomorphic on a simply con-

nected region Ω, then for any closed curve in Ω, we have

∮ f (z) dz = 0.
γ

Proof. Assume without loss of generality that γ is parameterized as a curve on [0, 1].
Then it can be thought of as the concatenation of two curves γ1 and −γ2 , where γ1 =
36 � 1 Basic theory

γ|[0,1/2] , and γ2 is the “reverse” of the curve γ|[1/2,1] . Note that γ1 and γ2 have the same
endpoints. By Theorem 1.23 we have

∫ f (z) dz = ∫ f (z) dz = ∫ f (z) dz − ∫ f (z) dz = 0.

γ γ1 −γ2 γ1 γ2

Combining Theorem 1.24 with Proposition 1.15, we get the following result.

Corollary 1.25. Any holomorphic function on a simply connected region has a primitive.

One subtle issue that is glossed over in many complex analysis textbooks is the ques-
tion of how to recognize when a region is simply connected. In many practical situations,
it is easy to recognize or at least accept as intuitively plausible, that the region under dis-
cussion is homeomorphic to a disc, which of course implies the property of being simply
connected. This informal style of reasoning will be sufficient for our needs in this book.
For those readers who prefer a higher level of rigor, we cite without proof the following
result from topology.

Theorem 1.26. Given any simple closed curve γ in the plane, there is a region Ω such that:
1. Ω is bounded;
2. Ω is the unique connected component of ℂ \ γ that is bounded;
3. Ω is homeomorphic to a disc.

Because of the second property of Ω given in the theorem, Ω is usually referred to

as “the region enclosed by γ.”
Theorem 1.26 is a version of the Jordan–Schoenflies theorem, which in turn is
a strengthened version of the Jordan curve theorem. These results have elementary
proofs that do not require complex analysis; see [9, 69] and [W6] for additional discus-
sion and references. A planar curve that is simple and closed is often referred to as a
Jordan curve.

Suggested exercises for Section 1.8. 1.19, 1.20, 1.21, 1.22.

1.9 Consequences of Cauchy’s theorem

Theorem 1.27 (Cauchy’s integral formula). If f is holomorphic on a region Ω containing
the closed disc D≤R (z0 ), then

{ f (z) if z ∈ DR (z0 ),
1 f (w) {
{
∮ dw = {0 if z ∈ Ω \ D≤R (D), (1.43)
2πi w−z {
{
{undefined if z ∈ CR (z0 )
CR (z0 )

Proof. The case where z ∈ Ω \ D≤R (D) is covered by Cauchy’s theorem in a disc, since in
that case the function w 󳨃→ f (w)/(w−z) is holomorphic in an open set containing D≤R (D).
 1.9 Consequences of Cauchy’s theorem � 37

Figure 1.7: The keyhole contour Γϵ,δ .

It remains to deal with the case z ∈ DR (z0 ). In this case, denote Fz (w) = f (w)/(w − z). The
idea is now to consider instead the integral

f (w)
∮ Fz (w) dw = ∮ dw,
w−z
Γϵ,δ Γϵ,δ

where Γϵ,δ is a so-called keyhole contour, namely a contour comprising a large circular
arc around z0 that is a subset of the circle CR (z0 ), and another smaller circular arc of
radius ϵ centered at z, with two straight line segments connecting the two circular arcs
to form a closed curve, such that the width of the “neck” of the keyhole is δ. (Here ϵ and
δ are two small positive parameters; think of ϵ as being small and of δ as being much
smaller than ϵ.) See Fig. 1.7. Note that the function Fz (w) is holomorphic inside the region
enclosed by Γϵ,δ . Moreover, this region is clearly homeomorphic to a disc and so is simply
connected. Therefore Cauchy’s theorem gives that

∮ Fz (w) dw = 0.
Γϵ,δ

We now take the limit of this equation as δ → 0. The two parts of the integral along the
“neck” of the contour Γϵ,δ cancel out in the limit because Fz is continuous, and hence
uniformly continuous, on the compact set D≤R (z0 ) \ Dϵ (z). So we can conclude that

∮ Fz (w) dw = ∮ Fz (w) dw. (1.44)

CR (z0 ) Cϵ (z)

The next and final step is to take the limit as ϵ → 0 of the right-hand side of this equation.
Write
38 � 1 Basic theory

f (w) − f (z) 1
Fz (w) = + f (z) ⋅ . (1.45)
w−z w−z

Integrating each of these two terms separately, for the first term, we have

󵄨󵄨
󵄨󵄨 f (w) − f (z) 󵄨󵄨
󵄨 |f (w) − f (z)|
󵄨󵄨 ∮ dw󵄨󵄨󵄨 ≤ 2πϵ ⋅ sup
󵄨󵄨 w−z 󵄨󵄨 |w−z|=ϵ ϵ
Cϵ (z)
󵄨 󵄨
= 2π sup 󵄨󵄨󵄨f (w) − f (z)󵄨󵄨󵄨 󳨀󳨀󳨀→ 0 (1.46)
|w−z|=ϵ ϵ→0

by the continuity of f ; and for the second term,

1 1
∮ f (z) ⋅ dw = f (z) ∮ dw = 2πif (z) (1.47)
w−z w−z
Cϵ (z) Cϵ (z)

(by a standard calculation; see Exercise 1.21). Combining (1.44) and (1.47) gives that
1
∮C (z ) 2πi Fz (w) dw = f (z), which was the formula to be proved.
R 0

An important particular case of (1.43) is the one in which z = z0 . Cauchy’s integral

formula gives in this case that

2π
1 dw 1
f (z) = ∮ f (w) = ∫ f (z + Reit )dt.
2π i(w − z) 2π
CR (z0 ) 0

In other words, we have proved the following result.

Theorem 1.28 (Mean value property for holomorphic functions). If f is holomorphic on a

region Ω containing the closed disc D≤R (z0 ), then the value f (z0 ) is equal to the average of
the values of f around the circle CR (z0 ).

Considering what the mean value property means for the real and imaginary parts
of f = u +iv, which are harmonic functions, we see that they in turn also satisfy a similar
mean value property:

2π
1
u(x, y) = ∫ u(x + R cos t, y + R sin t) dt. (1.48)
2π
0

In fact, (1.48) holds for all harmonic functions and is a result known as the mean value
property for harmonic functions. This result is proved in many textbooks using meth-
ods from real analysis or partial differential equations. Alternatively, it can be derived
from the above considerations by proving that every harmonic function in a disc is the
real part of a holomorphic function.
Random documents with unrelated
content Scribd suggests to you:
—¿Usted ha tomado el billete?—me preguntó la señora mayor.
—No; todavía, no.
—¿Quiere usted acompañarnos?
—Con mucho gusto.
Nos metimos los tres en un café que acababan de abrir.
La señora mayor tenía unos cincuenta o cincuenta y cinco años, y
llevaba tocas de viuda; la otra era una muchacha, pálida e
insignificante, de unos veintitrés años.
La señora hablaba con un acento nervioso y asustado; la señorita
estaba como apabullada.
Cuando pasó el tiempo necesario nos acercamos al despacho de
diligencias. Esperamos a que prepararan el Cuco, y entramos en él
las dos señoras y un capitán de la gendarmería de Bayona, que
había ido a Pau a recibir órdenes, y yo.
El capitán y yo hablamos. La señora mayor no hacía mas que saltar
en el asiento, de impaciencia. El Cuco marchaba perfectamente, con
un movimiento suave.
En los diferentes puntos que mudaban los caballos se presentaban
los gendarmes y preguntaban invariablemente si no iban españoles.
—Point d'espagnols—decía el capitán—. Dos damas francesas, un
señor inglés y un capitán de la gendarmería real.
—Perdón, mi capitán—decían los gendarmes, haciendo el saludo
militar.
—¿Por qué preguntan siempre si van españoles?—dije yo.
—Es que se teme que haya por aquí agentes españoles
revolucionarios—contestó al capitán.
Llegamos a Orthez por la mañana. El capitán y yo ofrecimos a las
señoras nuestra compañía, y como ellas aceptaron, fuimos hasta su
casa. El capitán dió el brazo a la mayor, y yo a la muchacha.
Llegamos delante de la puerta de la verja de una magnífica posesión
y nos despedimos de las señoras. El capitán fué hacia un lado y yo
hacia el contrario. Avancé un poco paralelamente a la verja, que era
más larga de lo que yo me figuraba, y al volver vi que las dos
mujeres estaban todavía a la entrada.
—¿No les oyen?—les pregunté—. ¿Quieren ustedes que yo llame?
—No, no—dijeron las dos, asustadas.
—Lo que ustedes quieran—y me preparé a seguir.
—¿Podría usted hacernos un favor?—me preguntó la señora con su
voz trágica.
—Sí, con mucho gusto.
—Querríamos entrar en el parque sin que nos viera el portero.
—No sé la manera.
—Hay una puerta chiquita, cerrada con solo un cerrojo, aquí, a un
lado.
—¿Y desde fuera cómo la va usted a abrir?
—No, desde fuera ya sé que no. ¿Usted no sería capaz de escalar
esta verja?
—¡Escalar la verja! ¿Y si le ven a uno?
—No. No se levanta en la casa nadie hasta muy tarde.
—Bueno; avísenme ustedes si aparece alguien.
Sin más dejé mi fardelillo en el suelo, escalé la verja, bajé por el otro
lado, corrí hacia la puerta pequeña y abrí el cerrojo. Las dos mujeres
entraron en el jardín y yo salí al camino.
Al pasar de nuevo por delante de la puerta de la verja estaba la
señora aguardando y me dijo:
—Quiero darle a usted una explicación y hablar con usted. Venga
usted cuando se haga de noche a esta verja.
—Sí, señora, vendré.
Me fuí a una fonda con la imaginación un poco excitada, y de noche
me presenté en la verja. Al poco rato llegó la señora. Me habló
durante más de una hora con un tono inquieto, lleno de angustia, y
me contó, atropelladamente, una porción de cosas.
Aquella dama era pariente y al mismo tiempo señora de compañía
de la muchacha joven que había venido con ella en el coche. Se
llamaba madama Domesan. La muchacha, Gabriela de Beaumont;
por lo que me dijo, vivía con su padre, su tío y una señora amiga de
su padre, Enriqueta Sarrazin, que se había hecho dueña de la casa
de tal manera, que los tenía presos a todos, sin dejarles salir de allí.
El día anterior esta señora había marchado del castillo, y
aprovechando su salida, Gabriela y ella habían ido a Pau a hablar
con un pariente y a explicarle la situación en que se encontraban,
pero no le habían visto.
En la casa, la Enriqueta Sarrazin mandaba como dueña, y había
dispuesto casar a su hijo, que era un perturbado, con Gabriela, y
estaba aislando a la familia de Beaumont de sus amigos y parientes,
de tal manera, que ya nadie entraba en la casa. En los planes le
ayudaba un cura del pueblo.
Después de todos estos datos, madama Domesan me dijo que si yo
tenía valor y energía para ello, que me presentara al día siguiente en
el castillo preguntando por el vizconde Beaumont de Lomagne; que
le dijera que llegaba de Londres y que era aficionado a los árboles y
a las plantas exóticas, y que quería ver el parque y el invernadero, y
me hiciera amigo de él.
Me sugestionaron los relatos de aquella dama y prometí seguir la
aventura.
Al día siguiente, al mediodía, me presenté en el castillo y llamé
tirando de la cadena.
Salió a abrir un portero viejo, con una gran librea; le di mi tarjeta, y
esperé.
Poco después se abrió la verja, y el criado me dijo que pasara.
Comencé a marchar por una avenida enarenada. Al final de ésta se
veía un edificio grande, pesado, de piedra, con varias torres de
pizarra adornadas con veletas.
A un lado y a otro había árboles centenarios, altísimos, y delante de
la fachada del castillo, un estanque oval, de agua profunda y
obscura, a cuyo alrededor las hojas caídas en muchos años
formaban como un marco de plata.
Este estanque parecía un espejo negro que reflejase el cielo a través
del follaje de los árboles. Bordeando el estanque nos acercamos al
castillo, y entramos en un gran zaguán, que parecía una cripta, con
el suelo, las paredes y el techo de piedra. Subimos la ancha
escalera, pasamos un salón grande como un museo y fuimos a un
gabinete elegante, pero también triste, en donde había dos viejos
momificados sentados el uno frente al otro, la señora y la señorita
del coche y madama Sarrazin, una mujer de cara juanetuda, de ojos
claros y pelo blanco.
El vizconde me saludó amablemente. Era un hombre alto, encorvado
y pálido, con un aire de temor y de cansancio.
Vestía un traje del tiempo del Imperio, y al andar parecía arrastrarse.
Su hermano, el caballero de Maslac, era un vejestorio del tipo más
completo del antiguo régimen; llevaba calzones de terciopelo de
color, medias de seda, casaca y coleta. Iba perfumado, pintado, con
colorcitos en las mejillas y en los labios; los dientes, postizos, y
peluca. Usaba constantemente un lente y una tabaquera; en los
dedos, anillos, y dijes, y, al levantarse de la butaca, se apoyaba en
un bastón con puño de oro.
La señorita Gabriela y madama Domesan me saludaron
amablemente, y la señora Sarrazin apenas se dignó mirarme.
El vizconde de Beaumont, que tenía la manía de la botánica, me
mostró el parque y el invernadero de su castillo.
El parque era tristísimo; parecía que habían querido darle un aire
lúgubre, haciendo que los árboles gigantescos estuvieran tan cerca
uno de otro que, paseando por las sendas, no se veía el cielo.
El estanque reflejaba las nubes como una pupila desesperada y
sombría.
El vizconde me enseñó la antigua torre de los Beaumont, con sus
baluartes y sus argollas, que daban al río y servían para atar las
gabarras.
Después de ver sus plantas extrañas, dije que tenía que marcharme;
pero el vizconde me rogó varias veces que me quedara a cenar y a
dormir. Como este era mi objeto, me quedé allá.
La cena fué siniestra. El vizconde miraba a un lado y a otro, como
poseído por el mayor espanto; el caballero de Maslac, con sus
adobes y sus dijes, parecía una momia desenterrada.
No se habló en la mesa mas que de genealogías, y únicamente el
vizconde interrumpía esta conversación para disertar acerca de
botánica. Después de cenar jugaron una partida de cartas entre los
dos viejos, la Sarrazin y Gabriela, y madama Domesan me indicó que
fuera a la biblioteca, donde hablaríamos.
Efectivamente, fuí a ella y hablamos largamente. Me dijo, de una
manera nerviosa y perentoria, que yo, que había sido simpático al
vizconde, debía entrar en la casa y luchar contra la influencia de
madama Sarrazin, que les dominaba a todos.
Después me contó, con su tono dramático, la historia de un
muchacho que había galanteado largo tiempo a Gabriela, y a quien
se había encontrado ahogado en el río, y de un hombre misterioso
que aparecía de cuando en cuando en las proximidades del castillo.
Luego me habló de su vida y de su familia.
Me dijo que ella procedía del secretario de Felipe II, Antonio Pérez.
—Al evadirse Antonio Pérez de la cárcel de la Inquisición de
Zaragoza—me contó—, se refugió en el Bearn y fué protegido por
Enrique IV y por Margarita de Valois. Antonio Pérez tuvo amores con
una señora de Orthez, y su hijo se estableció aquí definitivamente, y
de él procedo yo.
Siguió la señora Domesan contando una serie de relatos de crímenes
y de sucesos extraños donde aparecían asesinos, misterios,
fantasmas, y llegué a pensar si aquella mujer estaría un poco
perturbada, y sería, sin proponérselo, una especie de Anna Radcliffe
gascona. Por lo menos era un folletín de muchas entregas.
Al pasar a la alcoba que me destinaron, que era inmensa y obscura,
no pude dormir. Toda la noche la pasé pensando en ahogados y
muertos.
Al día siguiente comprendí que aquellas grandezas no eran para mí,
y, sin despedirme de nadie, con el pretexto de dar un paseo, me
marché del castillo y no volví.

XIV.
EN LA DILIGENCIA
Corrí con mi morral al sitio donde salían las diligencias y tomé un
asiento para Bayona.
Me encontré con el mismo capitán de la gendarmería con quien
había ido a Orthez. Nos saludamos y nos dimos nuestros nombres.
Me dijo que se llamaba Montmartin, y me invitó a tomar una copa de
coñac.
En la diligencia iba mucha gente que subía y bajaba en los pueblos
pequeños, llevando cestas y encargos, y un comerciante bayonés,
con su mujer y dos hijas.
Una de ellas, por lo que contó su madre, tenía una voz preciosa, y
había obtenido un gran éxito cantando la Cavatina «Una voce poco
fá», del Barbero de Sevilla, en una de las casas del gran mundo de
Orthez.
La otra señorita poseía, según su madre, grandes conocimientos
literarios e históricos, y sabía el inglés y el español. Había sido muy
galanteada por un joven oficial de la guarnición de Orthez, llamado
Alfredo de Vigni, que había escrito para ella una poesía preciosa.
A pesar de hablar yo bastante mal el francés («¡Celibataire, une
boutaille de cercueil!), quedé un poco mejor que el capitán de la
gendarmería, pues éste consideraba que ante las señoras debía
tomar una aptitud rígida, como si estuviera en actos de servicio.
Quizá influían en su tiesura las frecuentes libaciones, pues
aprovechaba todas las paradas para intoxicarse cuanto podía.
Con este combustible se reveló en él su fondo de francés, y dijo que
Napoleón era un grande hombre, a quien los ingleses habían hecho
perecer miserablemente. Habló también de la batalla de Orthez, en
que Wéllington, con el ejército aliado, había batido al mariscal Soult,
y se deshizo en insultos contra el vencedor de Waterloo.
El comerciante bayonés y su familia parecían desolados al oír esto, y
me miraban como pidiéndome mil perdones.
El capitán vió que yo no me daba por aludido, se calmó, se hizo
amigo mío y amenizó el viaje con algunos cuentos de cuerpo de
guardia.
A la tardecita llegamos a Bayona, y, pasado el puente sobre el Adour,
el sargento del puesto de la gendarmería preguntó si no había
viajeros españoles. El capitán Montmartin dijo que no, y seguimos
adelante hasta la plaza de Armas.
El capitán sintió no sé por qué un vago impulso de simpatía o de
remordimiento al despedirse de mí, quizá por haber hablado mal de
los ingleses, y me invitó a cenar con él al café del Comercio tan
insistentemente, que tuve que aceptar.
Estaba el café, envuelto en una nube de humo, atestado de oficiales
de la guarnición. Se hablaba a gritos.
En una mesa había un grupo de tenientes y suboficiales, y uno de
ellos leía un libro que acababa de publicarse, de un tal Paul de Kock,
llamado Gustavo el calavera. Los que escuchaban se reían a
carcajadas. El capitán Montmartin y yo nos acercamos al grupo, y
aunque yo apenas oía la lectura, contagiado por la risa de todos,
acabé por reírme.
Mareado y algo intoxicado me despedí de Montmartin y me fuí a la
fonda. Al día siguiente me levanté temprano y salí a la calle. Vi
muchos grupos de españoles que me dijeron eran realistas, y entre
ellos un cura y un fraile, el uno con su gran sombrero de teja y el
otro con su cerquillo.
Los dos tiraban al blanco con carabina y tenían una magnífica
puntería.
—Son soldados de la Fe—me dijo un francés que debía ser realista
entusiasta.
—No cabe duda que con esa puntería—le contesté yo—han de ganar
muchas almas para el cielo.
 XV.
MARY LA DE BIRIATU

En la fonda de Bayona me dijeron que podía ir a San Juan de Luz a

caballo en un cacolet. No sabía lo que era esto, que resultó un
artefacto que en castellano llaman jamugas.
Llegué a San Juan de Luz en mi cacolet; dejé el morralillo en un
fonducho de la salida del pueblo y fuí a estirarme las piernas hacia la
playa.
Me sorprendió un chubasco y entré en un café pequeño y me senté
delante de una ventana con cristales, y estuve contemplando cómo
chocaban las gotas de agua en la tierra, y las nubes que corrían por
el cielo.
Al terminar el chaparrón volví al fonducho de la salida del pueblo e
hice mis preparativos para entrar al día siguiente en España.
Estaba sentado en la mesa y estudiando un mapa cuando entró una
muchacha a preguntarme si quería cenar. Al verla, me pareció que el
cuarto se iluminaba; tan bonita era.
—¿Usted me va a servir la cena?—le dije.
—Sí.
—No creí poder ser tan feliz.
Ella se rió. Yo la contemplé embobado. Tenía unos ojos claros azul
verdosos, una boca burlona y un cuerpo ligero y fuerte al mismo
tiempo. Era un fruto del Norte dorado por el sol del mediodía.
Le pregunté cómo se llamaba y me dijo que Mary; le volví a
preguntar de dónde era y me contestó que de Biriatu, un pueblecillo
pequeño asentado en un cerro próximo al Bidasoa.
—Voy a quedarme aquí—le dije—para poder verla a usted muchos
días.
—No podrá ser—contestó ella.
—¿Por qué?
—Porque me marcho a Biriatu mañana.
—Iré yo a Biriatu.
—Es igual; no me verá usted.
—¿Tendrá usted novio?
—No.
—¿Pero tendrá usted muchos pretendientes?
—No; tampoco.
—¿Cómo puede ser eso, siendo tan bonita?
—No opinan todos como usted—me replicó riendo.
—Eso es imposible—exclamé—. ¿Es que los hombres de este país no
tienen ojos? ¿Es que son como esos peces de los lagos sin luz, que
son ciegos? ¿Es que tienen alguna membrana nictitante perpetua?
¿Es que...?
Mary la de Biriatu iba y venía trayendo platos, haciendo poco caso
de mis frases.
Cuando se acabó la cena le dije que ya que no podía verla quería
marcharme al amanecer y que me diera la cuenta.
Me la trajo y quise darle de propina un luis de oro.
—No, no—me dijo—; guárdese usted su moneda de oro. No la
quiero.
—¡Pero, si yo no la pido nada a cambio!
—Es igual; no la quiero. Le hará a usted más falta que a mí. ¡Adiós!
Buenas noches.

J. H. Thompson dice, al llegar aquí, que se metió en su cuarto y

sacando lápiz y papel escribió una poesía en inglés en honor de la
muchacha que encontró en la fonda. La tal poesía es una
españolada poco seria que no nos puede agradar a las personas
sensatas, y si la traducimos y la copiamos es, más que para otra
cosa, para demostrar la extravagancia de los extranjeros cuando se
ocupan de España. Dice así la canción traducida al pie de la letra:

«A Mary la de Biriatu:

»Tienes los ojos azul verde claros, Mary la de Biriatu, como

las olas del mar; tienes la boca burlona y fresca y el cuerpo
ágil y armónico como el de una diosa. Cuando te veo
marchar de aquí para allá, mi corazón tiembla y siente el
mismo sobresalto que si fuera una pieza de porcelana de
Sèvres en manos de una criada cerril, o la copa más fina de
cristal de Bohemia entre los dedos de un chico atolondrado.
»Eres amable, Mary la de Biriatu, y, sin embargo, eres
cruel. Tienes la crueldad de la fuerza, que no sospecha la
debilidad ajena; tienes la exactitud del teorema
matemático, que es un tormento para la inteligencia
obscura; eres soberbia como la Naturaleza, y yo soy
humilde como una cosa humana.
»¡Si tú quisieras!, yo saldría de mí mismo como un dragón
de su agujero, y sería el hombre más turbulento y más
dionisíaco de la tierra. Pero no, no lo sería; lo soy ya.
»Me he transfigurado, y las furias anidan en mi corazón. Ya
no soy un inglés pesado y grueso; soy andaluz y tengo
sangre mora en las venas; tengo garras como las águilas y
colmillos agudos como los tigres. Ya no diseco fieras, las
mato; ya no discuto con los hombres, los domino.
»Ven conmigo, Mary, Mary la de Biriatu. Yo te llevaré en mi
caballo cordobés, desde el Pirineo a Sierra Nevada y
reposaremos al pie de las palmeras de Andalucía al son de
las castañuelas y las guitarras.
»Si quieres que sea contrabandista, Mary, me haré
contrabandista; si quieres que sea salteador de caminos, lo
seré sin miedo e imitaré al bandido generoso,

el que a los ricos robaba

y a los pobres protegía.

»Para mí no habrá más leyes que tu capricho, Mary, Mary la

de Biriatu; para mí no habrá más cielo azul que el azul
verdoso de tus ojos. Con el trabuco al brazo, montado en
mi jaca torda, seré una exhalación. Yo me escabulliré de
entre las manos de la justicia y haré llorar de rabia a los
alguaciles, y a los alcaldes, y a los corchetes de la Santa
Hermandad.
»¿Hay que desafiar al rey, a la Inquisición, a los ángeles, a
los demonios?
»Aquí estoy yo. Yo robaré las alhajas de la Virgen para
adornar tu garganta y te daré la Biblia de Lutero para que
con sus hojas hagas papillotes.
 »Y cuando el mundo entero esté retemblando con mi gloria
como una caldera de vapor, y mis hazañas sean cantadas
por los ciegos, tú, con tu mantilla de casco y una peineta
de concha; yo, con el calzón corto y mi capa andaluza,
iremos los dos del brazo a la corrida.
»Ven conmigo, Mary, Mary la de Biriatu. Mira que soy capaz
de todo por ti. Mira que si no te pierdes al mismo Robin
Hood con calañés».

Esta es la absurda e insensata poesía que J. H. Thompson dedicó a

la muchacha de la fonda de San Juan de Luz, donde estuvo
hospedado, y que ha desagradado profundamente a varias personas
respetables que la han leído.

XVI.
LA VENTA DE INZOLAS

Después de descargar mi corazón en estos versos me tendí en la

cama, me quedé dormido, y por la mañana, al amanecer, me levanté
y salí de casa.
—Veremos lo que nos reserva la suerte—me dije.
Anduve una legua antes de que saliera el sol, y me senté al pie de
un árbol y saqué del bolsillo mi mapa de España. Estaba publicado
en Londres, en 1808, por la casa John Stockdale de Piccadilly, y
debió de servir para las tropas de Wéllington que iban a la Península.
—Como no tengo objeto—murmuré—, seguiré el meridiano. El mito
de mi tío el comandante Cox y el meridiano serían mis directrices.
Decidí pasar uno o dos meses en el país vasco, medio año en
Castilla, e ir a parar a Andalucía. Estaba enfrascado en la
observación del mapa cuando pasó una chiquilla que se me quedó
mirando.
Me levanté y la pregunté:
—¿Este es el camino de Navarra?
—Sí.
La muchacha iba hasta un caserío llamado Herburu, y yo fuí con ella.
Encontré a un aduanero francés a quien le dije me indicara el
camino de España. Me miró con desconfianza y me mostró un
sendero.
Siguiéndolo, llegué a un bosque bastante cerrado, con una venta, la
venta de Inzola. Estaba en territorio español. Pedí en la venta que
me pusieran algo de comer, y con un gran trozo de pan, de chorizo y
de queso y una botella de vino, me senté en la hierba, en un prado.
Brillaban las margaritas y las flores del brezo; una serpentaria
mostraba su mazorca roja entre lo verde. Corría allá un vientecillo
del mar fresco y agradable; el cielo estaba muy azul; en Francia se
veía la llanura y la costa; hacia España, un laberinto de montes
ceñudos y sombríos. Unos grillos amenizaban la soledad y un cuco
lanzaba su voz irónica entre los árboles.
Devoré mis provisiones, y después dirigí un toast elocuente a la vieja
España de Don Quijote, y del Cid, y de San Ignacio de Loyola. Añadí
a Loyola, para probarme a mí mismo, que este Amadís de Gaula,
católico y papista, no sólo no irritaba mis sentimientos de
protestante de raza, sino que veía en él un hermoso manantial de
energía y de tesón.
Después de este toast hice mi segunda libación brindando por las
damas españolas, los caballeros, las majas, los toreadores, los
gitanos, los corchetes, los alguaciles y los alcaldes, y, sobre todo, por
la bella entre las bellas, Mary la de Biriatu. Como me quedaba más
vino en la botella y no era desagradable, tuve que brindar por el
mar, por el cielo azul, y hasta por la Cosa en sí, y me quedé un
momento dormido.

SEGUNDA PARTE

DEL PIRINEO A MADRID

I.
LOS PLACERES DEL CAMPO

Cuando yo leía de chico las descripciones de los placeres campestres

—dice J. H. Thompson—, me parecían una de las cosas más insulsas
y más tontas del mundo. Es extraño cómo la retórica, a fuerza de
repetir las mismas frases, llega a borrar todo sentido de la realidad.
Los placeres campestres en las páginas de los escritores bucólicos
del siglo XVII y XVIII han sido siempre placeres amables y sociales;
se ve que para estos escritores la Naturaleza estaba representada
por un parque bien cuidado, como para Fenelón la gruta de Calipso
era uno de los subterráneos del jardín de Versalles. Los placeres
campestres en la pintura han sido también tan sosos, tan
amanerados, como los descriptos por los poetas.
Al llegar a vivir en el campo por primera vez, nunca recordé las
descripciones que había leído en la infancia, ni los cuadros de los
pintores. No me acordé jamás de Galatea, ni de Amarilis, ni de
Thirsis, ni de Nemoroso; todas estas amables personificaciones no
salieron del estante que les corresponde en el armario de la
guardarropía poética para presentarse a mi imaginación. Me
desdeñaron tanto como les desdeñaba yo a ellas.
Al acercarme al campo, la Naturaleza, en vez de una impresión
amable, pastoril y bucólica, me dió una sensación ruda y me habló
con una voz áspera y discordante.
El viento y la lluvia, el murmullo de los árboles en el follaje y el
rumor del arroyo, el caminar por entre las altas hierbas o por el claro
del bosque me produjeron una sorpresa.
Tuve también otras sorpresas y descubrimientos. Uno de éstos fué
encender hogueras.
Pocas cosas me han parecido tan sugestivas. ¡Hacer fuego al borde
de un camino y ver cómo chisporrotean las hierbas secas,
serpentean las llamas y se desparrama el humo por el aire! ¡Qué
gran placer! ¡Qué eterna admiración!
Siempre parece un espectáculo nuevo, como si guardara uno en el
fondo del alma el asombro del hombre primitivo, descubridor del
fuego al ver levantarse las llamas en el aire.
Este es uno de los grandes placeres tristes y melancólicos del
campo. Mirar la llama de la hoguera, ver el humo que mancha las
claridades del crepúsculo, mientras las estrellas comienzan a
presentarse en el cielo...
Hoy, al pensar en ello, siento melancolía, la melancolía del
enamorado de la Naturaleza unida a la melancolía del reumático.
 II.
ERLAIZ EL PANADERO

Después de mis libaciones dejé la venta de Inzola y comencé a

marchar hacia Vera. Enfrente tenía un enmarañamiento de montañas
fragosas y obscuras, de crestas y de barrancos.
Por toda la zona pirenaica vasconavarra ocurre lo mismo: lo trágico y
fosco ha quedado para España; lo sonriente y amable, para Francia.
A pesar de esto, el espíritu de los vascos de un lado y otro de la
frontera ha quedado el mismo; la misma seriedad, el mismo gusto
por los trajes negros, el mismo aire de desilusión.
Parece que este pequeño pueblo tiene la conciencia vaga de su
desaparición, de su absorción por los de alrededor, y le queda la
tristeza y el orgullo de los pueblos viejos que se hunden sin dejar
apenas rastro de su existencia.
Bajaba despacio de la venta de Inzola a Vera del Bidasoa cuando oí
a lo lejos el ruido de una carreta. ¡Cómo chirriaba! Tan pronto se la
oía como se perdía su sonido, como volvía a aparecer. Estas carretas
vascas tienen las ruedas de madera de una sola pieza y sujetas al
eje, lo que hace el rozamiento muy grande.
Preguntaba unos días después a los campesinos en Vera por qué
hacían así las carretas, al menos por qué no daban sebo a los ejes, y
uno me dijo que con aquel chirrido áspero se divertían los bueyes, y
otro, que así no había que avisar a nadie del paso de la carreta,
porque el chirrido de las ruedas avisaba solo.
Iba bajando al fondo de un arroyo, a cuyo borde se veían varios
caseríos, cuando me encontré con un viejo que marchaba seguido
de su perro. Era un hombre afeitado, encorvado, con un perfil de
cuervo.
Entablé conversación con él y, después de someterme a un
interrogatorio, me dijo que andaba buscando minas.
En el interrogatorio tuve que decir quién era y a qué venía a España,
y eché mano del mito Cox y de la herencia, y expliqué mis planes.
El mismo individuo me preguntó qué pensaba hacer en Vera; le dije
que pasaría allí un día nada más y seguiría adelante.
—¿Tiene usted posada?
—No.
—Pues yo le llevaré a casa de un paisano amigo mío, que le
hospedará barato.
Llegamos a uno de los barrios del pueblo al anochecer. En lo hondo
de un valle se veían unas cuantas casas viejas en fila, envueltas en
la niebla; el humo salía de las chimeneas en ligeras columnas azules.
El viejo y yo recorrimos una calle larga, pasamos por cerca de la
iglesia y salimos a la carretera, a orilla del Bidasoa, y en una casa
con una tienda nos detuvimos.
A la puerta estaba Erlaiz, el panadero; hablaba con un herrador de
una fragua próxima. El panadero, un hombre bajo, cuadrado, picado
de viruelas, de cara fosca y ceñuda, explicaba algo al herrador,
hombre grueso, panzudo, con una sonrisa llena de malicia.
El panadero nos recibió ásperamente, al viejo y a mí, y a una
muchacha que estaba en la tienda le dijo que me llevara a una
habitación.
Crucé la tienda, subí a un cuarto pintado de verde, me lavé y eché
un vistazo al pueblo. El vasco es indiferente y un tanto hostil al
extranjero; aunque se le hable en español, si le ven a uno extraño,
le miran con desconfianza y con suspicacia. La gente a quien
pregunté algo, en vez de responderme dándome los datos que les
pedía, me contestaban preguntándome a qué venía y qué pensaba
hacer.
Estos vascos recelosos suponen que se les tiende un lazo al hacerles
la pregunta más sencilla.
Pensé que no estaría muchas horas en el pueblo.
A la hora de cenar volví a mi posada de casa del panadero, y me
hicieron pasar a un comedor, en donde se hallaban el buscador de
minas, que había encontrado en el monte, Erlaiz y un militar.
El panadero, mi patrón, cambiado por completo de aspecto, se
mostraba sonriente y amable. Me indicaron mi sitio en la mesa, y
nos pusimos a cenar.
El viaje me había abierto el apetito, y di un ataque formidable a los
platos, al pan y al vino. Los demás no se quedaron atrás. Después
de cenar trajeron café y licores, y nos pusimos a hablar y a cantar.
Yo no he visto compadres más alegres que aquéllos.
El militar, guerrillero con Mina en la guerra de la Independencia,
contó sus hechos de armas, y el panadero habló de sus aventuras en
tierra de Castilla.
Los dos estuvieron a cuál más exagerados.
Estos buenos vascos, cuando se lanzan a ello, son un tanto
fanfarrones, como los escoceses de Walter Scott, o como los
gascones. Al oírles a ellos, cualquier encuentro de cincuenta
hombres contra otros cincuenta es una batalla de Austerlitz; una
aldea con cuatro casas viejas, una Florencia y un granero con una
torre es el Louvre o el Kremlin.
Después de las hazañas del militar y del panadero, el viejo buscador
de minas, que se llamaba Bidarraín, nos dió lecciones de botánica y
de mineralogía popular, mezcladas con algunas supersticiones.
A las doce y media, rendido de sueño, me fuí a la cama, dormí de un
tirón hasta las diez, y, al despertar, pensé si la cena de la noche
habría sido una realidad o una fantasía.
Me vestí, bajé a la tienda de Erlaiz y me lo encontré displicente y
malhumorado.
—Ahí ha venido ese viejo Bidarraín a preguntar por usted—me dijo
—. En la huerta debe estar.
Tomé el café con leche que me sirvió la sobrina de Erlaiz, salí a la
huerta y encontré al viejo buscador de minas.
Me preguntó si quería dar un paseo con él, le dije que sí y echamos
a andar. Bidarraín me mostró varias muestras de mineral, y
hablamos de mineralogía y de botánica. Luego le pregunté qué clase
de hombre era Erlaiz, el panadero, pues me parecía de genio
mudable.
—Es buena persona—me dijo—, pero muy violento y muy terco.
Cuando se le pone una cosa en la cabeza no hay quien le pueda
convencer de lo contrario. Le hemos querido persuadir el teniente
Leguía y yo de que es una barbaridad que ponga cepos en el
Bidasoa para los salmones en época de veda; pues los pone, y
aunque viniese el obispo y se lo pidiera de rodillas, los seguiría
poniendo.
Bidarraín contó otros detalles de la barbarie del panadero. Llegamos
a mi posada; el buscador de minas se marchó y yo entré en la
tienda. Pasé a la tahona, y vi a dos viejas que amasaban los panes
en una artesa, mientras Erlaiz trabajaba con la pala en el horno.
Murgui, la sobrina del panadero, me sirvió la comida; entablé
conversación con esta muchacha, y le pregunté qué clase de hombre
era Bidarraín. Me dijo que pasaba por hombre rico; que tenía minas
de plata y de oro.
También le pregunté a Murgui acerca del teniente Leguía, y, por lo
que contó, deduje que a éste le consideraban como el enemigo del
pueblo.
Por la tarde volvió a presentarse Bidarraín y me llevó a una
huertecilla contigua al cementerio, donde se hallaban enterrados dos
oficiales ingleses, muertos en el pueblo al pasar los aliados el
Bidasoa, en 1813.
Después fuimos hasta Lesaca, villa donde tuvo lord Wéllington su
cuartel general.
Bidarraín debió hablar al panadero de mis conocimientos
mineralógicos, porque Erlaiz, por la noche, me preguntó si era
ingeniero. Le dije que no, y él pareció no creerme. Me preguntó
también si tendría algún inconveniente en ver unas minas algo
lejanas. Le contesté que ninguno. Dispusimos hacer la expedición al
día siguiente. El panadero, contento, trajo la guitarra y estuvo
cantando. Cantaba de una manera bárbara y graciosa. Cuando
terminó, me fuí a mi cuarto y estuve un rato en la ventana mirando
las estrellas, oyendo el rumor del río y el canto de un sapo (bufo
músicus), que entretenía su soledad con sus notas.

III.

EL PARADOR DE SUMBILLA

Bidarraín y Erlaiz me llevaron varias veces a ver sus minas. Estaban

empeñados los dos en que yo entendía mucho de minería; pero que,
por razones especiales, no lo quería confesar.
Erlaiz y Bidarraín me pidieron que les escribiera varias cartas en
francés y en inglés, y cuando yo indiqué al panadero me hiciera la
cuenta, me dijo que no le debía nada.
El teniente Leguía pensaba marchar a Elizondo con unos cuantos
hombres de su partida, y yo quedé en acompañarle y seguir después
a Pamplona.
Con este motivo se decidió obsequiarnos a los dos con una cena de
despedida en las Ventas de Yanci.
Eramos los comensales, además del panadero, Leguía, Bidarraín y
yo; dos milicianos nacionales, sargento y cabo de la partida de
Leguía, y un liberal de Vera, que gastaba antiparras de plata, a quien
llamaban Laubeguicua (el de los cuatro ojos).
Fuimos todos paseando a las Ventas de Yanci, que distan una legua
y media de Vera; nos sentamos a beber sidra, y se llamó al ventero
y a la ventera y se les sometió a un grave interrogatorio.
Erlaiz, Bidarraín y el sargento de milicianos dieron a la consulta una
importancia sacerdotal.
—Vamos a ver, ¿qué podemos comer?—preguntó Erlaiz.
—Si quieren ustedes un cordero, ya lo asaremos—dijo la ventera,
cantando al hablar.
—Bueno, un cordero. ¿Que más?
—Ya tenemos también buenas truchas.
—¿Truchas? No está mal. ¿Que más?
—Pollos también ya tenemos.
—¿Pollos? Bueno. ¿Qué más?
—Jamón bueno ya pondremos.
Así siguió la ventera explicando las provisiones que tenía, siempre
empleando esta fórmula de ya tenemos o ya pondremos. Este ya, de
aire germánico, me chocaba verlo empleado a todo pasto.
Después de consultarse con la mirada Bidarraín, Erlaiz y el sargento
de nacionales, decidieron, de común acuerdo, que pusieran todo lo
que hubiese para no engañarse.
Dispuesta la cena, seguimos bebiendo, hasta que nos dijeron que la
mesa estaba puesta.
Al sargento de los milicianos, hombre alto, de vientre piriforme, se le
encandilaron los ojos, y frotándose las manos de gusto exclamó:
—¡Pien, pien! Una puena cena. Esto es lo que me gusta. Puen
cordero, puenas truchas, puen pollo y puen vino. ¡A comerr! ¡A
comerr!
Comimos como buitres y bebimos hasta quedar mareados, lo que
me dió una idea bastante pobre de la sobriedad de los vascos; se
habló con entusiasmo de Mina, Riego y el Empecinado; con rabia, de
las correrías que hacían por Navarra Juanito el de la Rochapea y don
Santos Ladrón, y se cantó el Himno de Riego, a pesar de que el
ventero y su mujer suplicaron que callásemos, porque les
comprometíamos.
Salimos de las Ventas de Yanci a media noche; los de Vera se
marcharon a su pueblo, y Leguía, con sus dos milicianos y yo,
seguimos hasta Sumbilla.
Nos detuvimos en el parador de San Tiburcio. El sargento del vientre
piriforme me dijo ingenuamente que con el paseo se le había abierto
el apetito, y que iba a mandar que le hicieran unas sopas de ajo. Le
miré con asombro y me fuí a acostar.
Al despertarme por la mañana supe que Leguía había partido con
sus milicianos camino da Santesteban, dejándome una carta para un
amigo suyo de Pamplona.
Como no tenía prisa y hacía calor, dejé la marcha hasta que cayera
el sol. Estaba en el portal del parador de San Tiburcio cuando se
acercó un carro grande, tirado por siete mulas.
El arriero fué soltando sus animales, llamándolos uno a uno y
llevándolos a la cuadra. La Morena, la Montesina, la Capitana, la
Coronela, la Bonita, el Vigilante y la Leona fueron despacio al
pesebre, donde primero se les dió de beber.
El posadero me preguntó:
—¿No va usted a Pamplona?
—Sí.
—Pues si quiere usted, puede usted ir con este arriero.
—¿No hay inconveniente?...
—Ninguno.
El arriero se llamaba Mandashay, y era un hombre de unos treinta y
cinco a cuarenta años, rubio, con unos ojos que parecían de cristal
azul.
Me advirtió que si quería ir con él saldríamos a la mañana siguiente;
le dije que tendría mucho gusto en marchar en su compañía, y le
convidé a un vaso de vino. Quedamos de acuerdo; yo me fuí a
acostar, y al amanecer me llamaron.
La mañana estaba fresca; había una niebla espesa que prometía un
día de calor. Mandashay sacó sus mulas y echamos a andar camino
de Almandoz.
—Cuando se canse usted puede tenderse en la galera—me dijo
Mandashay.
—No, no me canso tan fácilmente.
La galera española es un carro grande, de cuatro ruedas, tirado por
una larga recua de mulas. En Navarra y en Castilla la Vieja se ven
con frecuencia estas galeras; en Castilla la Nueva abunda más el
carromato, que también llaman carro catalán.
El viaje a pie detrás de un carro tiene sus encantos. El que aproximó
de un modo ideológico la galera carro a la galera barco, dándole el
mismo nombre, no estaba equivocado. El parecido de estos dos
medios de comunicación salta a la vista. El barco es una casa que
flota, como el carro es una casa que rueda. El carretero tiene algo
de marino: es un hombre que pasa y no se detiene, que lleva una
ruta, que vive en un mundo de soledad.
Mandashay era un hombre muy interesante y ameno. Cada rincón
del camino le recordaba una historia. Aquí habían salido a robar a un
rico unos enmascarados; allá había vivido una muchacha de cabeza
loca que trajo revueltos a todos los jóvenes de los contornos.
En algunos momentos Mandashay se agarraba a la galga, y en otros
tiraba de la brida del macho de varas, gritando: ¡Eup! ¡Eup! o
¡ueschqué! ¡ueschqué!
Charlando llegamos a Almandoz y seguimos subiendo una cuesta
hasta el alto de Velate. El cielo estaba azul y el sol calentaba de
firme. No hacía mucho calor porque íbamos ya a bastante altura y
corría aire fresco.
A media tarde cruzamos un bosque, que me pareció debía servir
para los misterios de los druidas, y fuimos a parar a las Ventas
Quemadas, en lo más alto del puerto.

IV.
PAMPLONA

Salimos de Ventas Quemadas por la mañana, y emprendimos la

marcha hacia la vertiente del Ebro.
El paisaje había cambiado en absoluto. El cielo se mostraba más
azul; el campo, más seco; en los altos corrían pequeños caballos de
grandes colas y triscaban las cabras y los corderos; abajo
resplandecían los campos de trigo y alguno que otro viñedo.
Al comenzar a descender hacia la cuenca del Ebro, me pareció que
empezaba España; todo tomaba a mis ojos un carácter más triste y
más serio: veía pueblos taciturnos, casas de paredes grises, árboles
cubiertos de polvo.
Nos alejamos de la altura a medida que avanzábamos, y fuimos
bajando hacia el llano. En los trigales brillaban las amapolas como
gotas de sangre y los grillos nos ensordecían con sus chirridos.
Dormimos en Villaba, y al día siguiente entraba yo en Pamplona. Me
despedí de Mandashay y fuí a parar a una posada de la calle de la
Curia. Saqué la carta del teniente Leguía; era para un capitán de
ejército llamado Iriarte. Me presenté a él, me acogió con amabilidad
y me invitó a comer.
Durante la comida le hablé del mito Cox, y de cómo esperaba
recoger una pequeña fortuna. En tanto, le dije, me hallaba dispuesto
a trabajar en lo que se me presentase.
—Y usted, ¿qué sabe hacer?—me preguntó Iriarte.
—Sé francés y, naturalmente, inglés.
—Sí; quizá esto le pueda servir de algo.
—También tengo nociones de botánica.
—¿Botánica? No creo que haya aquí nadie que se ocupe de eso. A
no ser algún herbolario.
—Pues éstos son mis conocimientos. También sé disecar animales—
añadí con resignación.
—¡Hombre! Eso quizá nos sirva. Aquí hay un profesor que todos los
pajarracos y alimañas que le dan los envía a Francia a disecarlos, lo
que le cuesta mucho dinero.
—Voy a verle.
—Sí, iremos juntos.
Fuimos, efectivamente; hablamos con él, y yo me comprometí a
restaurarle algunos animales y a disecarle de nuevo otros, por el
sueldo de seis pesetas al día, mientras durara el trabajo.
Disequé para aquel señor un caimán, un águila, un cisne y varios
otros bicharracos.
El capitán Iriarte me recomendó una casa de huéspedes de la plaza
del Castillo y me trasladé a ella.
Mi vida en Pamplona, mientras tuve trabajo, fué muy agradable. Por
la mañana y por la tarde trabajaba, y al anochecer paseaba por los
alrededores, y cuando no tenía tiempo de sobra iba a la Taconera.
Allí se reunían los aristócratas y los burgueses, los militares, las
señoritas, los chicos y los curas, y algunos días de fiesta, por la
noche, se ponían unos farolillos de papel colgados de los árboles.
Yo cultivaba mucho el mirador de la Taconera, un sitio bonito, desde
donde se ven los pueblos de la cuenca de Pamplona.
Daba con prudencia la vuelta a las murallas. Conocía la Ciudadela y
los baluartes: el de la Reina, el de Redín, el de Labrit, el de los
Canónigos, el de Gonzaga; las cinco puertas y la poterna de la
Tejería.
Tenía algunos amigos, porque el capitán Iriarte me presentó a varios
de sus compañeros, militares liberales.
Por entonces, entre los militares y los milicianos de Pamplona, había
gran hostilidad; los militares se manifestaban anticlericales,
partidarios de la Constitución; en cambio, los milicianos eran
fervientes católicos y monárquicos, y armaban trifulcas gritando:
«¡Viva Dios!» No parecía sino que tenían miedo de que lo mandasen
matar los liberales. Yo, como extranjero, no daba mi opinión acerca
de estas cuestiones.
En la casa de huéspedes donde fuí por recomendación de Iriarte,
eran todos perfectamente reaccionarios, comenzado por la dueña,
doña Saturnina, señora vieja, nariguda y charlatana.
Doña Saturnina era un producto clásico de las ciudades levíticas;
tenía la adoración por el aristócrata, por el cura, por el Don Juan
provinciano; hablaba con ternura de los mozos calaveras
alborotadores, que iban a los toros, bebían vino hasta
emborracharse y hacían de cuando en cuando alguna canallada y
luego iban a confesarse con aire hipócrita y santurrón al
confesonario del cura que pasaba por más severo, que generalmente
era el penitenciario de la catedral.
Al principio de estar en casa de doña Saturnina creí que se podría
bromear con las costumbres del pueblo, e hice algunos chistes
acerca de ese cartel que se pone en ciertos días en las iglesias de
España: «Hoy se sacan ánimas del Purgatorio»; pero pronto vi que
en Pamplona las bromas de este género tenían sus peligros.
Doña Saturnina la patrona y un sobrino suyo me espiaron y hasta
me siguieron un domingo para ver si iba a misa. En vista de que no
frecuentaba la iglesia, doña Saturnina me interpeló con valor:
—Dígame usted, ¿usted no es católico?—me dijo.
—No, señora—le contesté.
—¿Pues qué es usted? ¿Protestante?
—Sí; soy de una clase de secta que se llama de los agnósticos, que
supone que no se sabe nada de nada.
—¿Pero usted no cree en la Virgen y en los santos?
—Los de mi secta creemos más bien en la substancia única, y
practicamos el culto del nuestro señor el Yo, y de nuestra señora de
la Cosa en Sí.
A esto dijo mi patrona que esta virgen sería muy importante; pero
que los milagros de la Virgen del Camino eran mayores, porque se la
había visto elevarse en el aire y ponerse en un madero que hay en la
iglesia de San Cernín de Pamplona, a lo cual yo repliqué diciendo
que bien podía la ley de la gravitación, inventada por mi paisano
Newton, ser una costumbre o una rutina de la Naturaleza, y de
nuestro espíritu, y que, como dijo atrevidamente Protágoras, todas
las cosas son verdaderas, y que el hombre es la medida de todas las
cosas, de las que existen como existentes y de las que no existen
como no existentes.
Doña Saturnina me preguntó si este Protágoras era algún santo; yo
la dije que si no lo era, podía haberlo sido.
Entonces, doña Saturnina me recomendó que me convirtiese al
catolicismo, y yo la tranquilicé diciendo que estudiaría la cuestión.
En España y en pueblos como Pamplona todavía hay un gran
atractivo en ser incrédulo. ¿Cuánto durará esto? Al paso que vamos,
ya poco. Cien años; doscientos años. Nada, una miseria.
La verdad es que, con el progreso, se priva al hombre libre de los
grandes encantos y emociones de ser perseguido.
¿Qué vale un incrédulo, un librepensador en un país donde todo el
mundo puede serlo impunemente? Nada. En cambio, en plena
persecución, ¡qué delicia! Tener el libro prohibido bien guardado,
leerlo a escondidas, burlarse por dentro de todas las ceremonias y
mojigangas, y escapar de las tramas de esta red con la cual el
despotismo judaico-cristiano ha intentado envolver al mundo.
¡Admirable cosa!
Los incrédulos debíamos protestar de la lenidad actual, que nos priva
de una de nuestras mayores satisfacciones...
 V.
LOS CABALLEROS

En la casa de huéspedes de doña Saturnina conocí a varias

personas, gentes pintorescas, cuya vida sabía luego por la misma
patrona.
Uno de los fijos en la casa era un señor alto, moreno, de pelo
blanco, vestido con traje obscuro, y que se paseaba por los arcos de
la plaza del Castillo ataviado con un sombrero de copa cubierto de
hule y una levita larga, y que cuando hacía fresco se ponía una
esclavina azul sobre los hombros.
—¿Quién es este señor?—le pregunté a la patrona.
Doña Saturnina me dió tres o cuatro nombres de estos compuestos
y largos que usan los españoles.
—¿Y este caballero no trabaja?—pregunté yo.
—No. ¡Ca!
—¿Es rico?
—Poca cosa.
—¿Es de buena familia?
—Ya lo creo. ¡Es un Pérez de Cascante! Es de los caballeros de Olite.
Este caballero tenía un amigo que le acompañaba en sus paseos, el
señor Sánchez de Peralta.
Doña Saturnina me explicó la genealogía de ambos.
—Ninguno de los dos ha trabajado nunca—me decía la patrona con
entusiasmo—. Son caballeros.
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com