Download the full version of the textbook now at textbookfull.

com

Basic Real Analysis 2nd Edition Anthony W.

Knapp

https://textbookfull.com/product/basic-real-
analysis-2nd-edition-anthony-w-knapp/

Explore and download more textbook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Basic Real Analysis 2nd Edition Anthony W. Knapp

https://textbookfull.com/product/basic-real-analysis-2nd-edition-
anthony-w-knapp-2/

textbookfull.com

Advanced Real Analysis Anthony W Knapp

https://textbookfull.com/product/advanced-real-analysis-anthony-w-
knapp/

textbookfull.com

Basic Algebra Anthony W. Knapp

https://textbookfull.com/product/basic-algebra-anthony-w-knapp/

textbookfull.com

Western Mainstream Media and the Ukraine Crisis A Study in

Conflict Propaganda 1st Edition Oliver Boyd-Barrett

https://textbookfull.com/product/western-mainstream-media-and-the-
ukraine-crisis-a-study-in-conflict-propaganda-1st-edition-oliver-boyd-
barrett/
textbookfull.com
The OECD’s Historical Rise in Education: The Formation of
a Global Governing Complex Christian Ydesen

https://textbookfull.com/product/the-oecds-historical-rise-in-
education-the-formation-of-a-global-governing-complex-christian-
ydesen/
textbookfull.com

ALLEN PHYSICS HANDBOOK FOR NEET 2019th Edition Allen

https://textbookfull.com/product/allen-physics-handbook-for-
neet-2019th-edition-allen/

textbookfull.com

Talent Management in Small and Medium Enterprises Context

Practices and Outcomes 1st Edition Aleksy Pocztowski

https://textbookfull.com/product/talent-management-in-small-and-
medium-enterprises-context-practices-and-outcomes-1st-edition-aleksy-
pocztowski/
textbookfull.com

Security and Trust Management 14th International Workshop

STM 2018 Barcelona Spain September 6 7 2018 Proceedings
Sokratis K. Katsikas
https://textbookfull.com/product/security-and-trust-management-14th-
international-workshop-stm-2018-barcelona-spain-
september-6-7-2018-proceedings-sokratis-k-katsikas/
textbookfull.com

Pedophilia Empire Satan Sodomy The Deep State Chapter 21

BBC Pedo Ring s Deadly Silencing of Those Who Knew Too
Much The Jill Dando Murder and a Dozen Suspicious BBC
Deaths Joachim Hagopian
https://textbookfull.com/product/pedophilia-empire-satan-sodomy-the-
deep-state-chapter-21-bbc-pedo-ring-s-deadly-silencing-of-those-who-
knew-too-much-the-jill-dando-murder-and-a-dozen-suspicious-bbc-deaths-
joachim-hagopian/
textbookfull.com
Higher Education Institutions in the EU: Between
Competition and Public Service 1st Edition Andrea Gideon
(Auth.)
https://textbookfull.com/product/higher-education-institutions-in-the-
eu-between-competition-and-public-service-1st-edition-andrea-gideon-
auth/
textbookfull.com
Basic Real Analysis
Digital Second Editions
By Anthony W. Knapp
Basic Algebra
Advanced Algebra
Basic Real Analysis,
with an appendix “Elementary Complex Analysis”
Advanced Real Analysis
Anthony W. Knapp

Basic Real Analysis

With an Appendix “Elementary Complex Analysis”

Along with a Companion Volume Advanced Real Analysis

Digital Second Edition, 2016

Published by the Author

East Setauket, New York
Anthony W. Knapp
81 Upper Sheep Pasture Road
East Setauket, N.Y. 11733–1729, U.S.A.
Email to: [email protected]
Homepage: www.math.stonybrook.edu/∼aknapp

Title: Basic Real Analysis, with an appendix “Elementary Complex Analysis”

Cover: An instance of the Rising Sun Lemma in Section VII.1.
Mathematics Subject Classification (2010): 28–01, 26–01, 42–01, 54–01, 34–01, 30–01, 32–01.
First Edition, ISBN-13 978-0-8176-3250-2
c
"2005 Anthony W. Knapp
Published by Birkhäuser Boston
Digital Second Edition, not to be sold, no ISBN
c
"2016 Anthony W. Knapp
Published by the author
All rights reserved. This file is a digital second edition of the above named book. The text, images,
and other data contained in this file, which is in portable document format (PDF), are proprietary to
the author, and the author retains all rights, including copyright, in them. The use in this file of trade
names, trademarks, service marks, and similar items, even if they are not identified as such, is not
to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
All rights to print media for the first edition of this book have been licensed to Birkhäuser Boston,
c/o Springer Science+Business Media Inc., 233 Spring Street, New York, NY 10013, USA, and
this organization and its successor licensees may have certain rights concerning print media for the
digital second edition. The author has retained all rights worldwide concerning digital media for
both the first edition and the digital second edition.
The file is made available for limited noncommercial use for purposes of education, scholarship, and
research, and for these purposes only, or for fair use as understood in the United States copyright law.
Users may freely download this file for their own use and may store it, post it online, and transmit it
digitally for purposes of education, scholarship, and research. They may not convert it from PDF to
any other format (e.g., EPUB), they may not edit it, and they may not do reverse engineering with it.
In transmitting the file to others or posting it online, users must charge no fee, nor may they include
the file in any collection of files for which a fee is charged. Any exception to these rules requires
written permission from the author.
Except as provided by fair use provisions of the United States copyright law, no extracts or quotations
from this file may be used that do not consist of whole pages unless permission has been granted by
the author (and by Birkhäuser Boston if appropriate).
The permission granted for use of the whole file and the prohibition against charging fees extend to
any partial file that contains only whole pages from this file, except that the copyright notice on this
page must be included in any partial file that does not consist exclusively of the front cover page.
Such a partial file shall not be included in any derivative work unless permission has been granted
by the author (and by Birkhäuser Boston if appropriate).
Inquiries concerning print copies of either edition should be directed to Springer Science+Business
Media Inc.

iv
 To Susan

and

To My Children, Sarah and William,

and

To My Real-Analysis Teachers:

Salomon Bochner, William Feller, Hillel Furstenberg,

Harish-Chandra, Sigurdur Helgason, John Kemeny,
John Lamperti, Hazleton Mirkil, Edward Nelson,
Laurie Snell, Elias Stein, Richard Williamson
 CONTENTS

Contents of Advanced Real Analysis xi

Dependence Among Chapters xii
Preface to the Second Edition xiii
Preface to the First Edition xv
List of Figures xviii
Acknowledgments xix
Guide for the Reader xxi
Standard Notation xxv
I. THEORY OF CALCULUS IN ONE REAL VARIABLE 1
1. Review of Real Numbers, Sequences, Continuity 2
2. Interchange of Limits 13
3. Uniform Convergence 15
4. Riemann Integral 26
5. Complex-Valued Functions 41
6. Taylor’s Theorem with Integral Remainder 43
7. Power Series and Special Functions 45
8. Summability 54
9. Weierstrass Approximation Theorem 59
10. Fourier Series 62
11. Problems 78
II. METRIC SPACES 83
1. Definition and Examples 84
2. Open Sets and Closed Sets 92
3. Continuous Functions 96
4. Sequences and Convergence 98
5. Subspaces and Products 103
6. Properties of Metric Spaces 106
7. Compactness and Completeness 109
8. Connectedness 116
9. Baire Category Theorem 118
10. Properties of C(S) for Compact Metric S 122
11. Completion 128
12. Problems 131
vii
viii Contents

III. THEORY OF CALCULUS IN SEVERAL REAL VARIABLES 136

1. Operator Norm 136
2. Nonlinear Functions and Differentiation 140
3. Vector-Valued Partial Derivatives and Riemann Integrals 147
4. Exponential of a Matrix 149
5. Partitions of Unity 152
6. Inverse and Implicit Function Theorems 153
7. Definition and Properties of Riemann Integral 162
8. Riemann Integrable Functions 167
9. Fubini’s Theorem for the Riemann Integral 171
10. Change of Variables for the Riemann Integral 173
11. Arc Length and Integrals with Respect to Arc Length 181
12. Line Integrals and Conservative Vector Fields 194
13. Green’s Theorem in the Plane 203
14. Problems 212

IV. THEORY OF ORDINARY DIFFERENTIAL EQUATIONS

AND SYSTEMS 218
1. Qualitative Features and Examples 218
2. Existence and Uniqueness 222
3. Dependence on Initial Conditions and Parameters 229
4. Integral Curves 234
5. Linear Equations and Systems, Wronskian 236
6. Homogeneous Equations with Constant Coefficients 243
7. Homogeneous Systems with Constant Coefficients 246
8. Series Solutions in the Second-Order Linear Case 253
9. Problems 261

V. LEBESGUE MEASURE AND ABSTRACT

MEASURE THEORY 267
1. Measures and Examples 267
2. Measurable Functions 274
3. Lebesgue Integral 277
4. Properties of the Integral 281
5. Proof of the Extension Theorem 289
6. Completion of a Measure Space 298
7. Fubini’s Theorem for the Lebesgue Integral 301
8. Integration of Complex-Valued and Vector-Valued Functions 310
9. L 1 , L 2 , L ∞ , and Normed Linear Spaces 315
10. Arc Length and Lebesgue Integration 325
11. Problems 327
 Contents ix

VI. MEASURE THEORY FOR EUCLIDEAN SPACE 334

1. Lebesgue Measure and Other Borel Measures 335
2. Convolution 344
3. Borel Measures on Open Sets 352
4. Comparison of Riemann and Lebesgue Integrals 356
5. Change of Variables for the Lebesgue Integral 358
6. Hardy–Littlewood Maximal Theorem 365
7. Fourier Series and the Riesz–Fischer Theorem 372
8. Stieltjes Measures on the Line 377
9. Fourier Series and the Dirichlet–Jordan Theorem 384
10. Distribution Functions 388
11. Problems 390

VII. DIFFERENTIATION OF LEBESGUE INTEGRALS

ON THE LINE 395
1. Differentiation of Monotone Functions 395
2. Absolute Continuity, Singular Measures, and
Lebesgue Decomposition 402
3. Problems 408

VIII. FOURIER TRANSFORM IN EUCLIDEAN SPACE 411

1. Elementary Properties 411
2. Fourier Transform on L 1 , Inversion Formula 415
3. Fourier Transform on L 2 , Plancherel Formula 419
4. Schwartz Space 422
5. Poisson Summation Formula 427
6. Poisson Integral Formula 430
7. Hilbert Transform 435
8. Problems 442

IX. L p SPACES 448

1. Inequalities and Completeness 448
2. Convolution Involving L p 456
3. Jordan and Hahn Decompositions 458
4. Radon–Nikodym Theorem 459
5. Continuous Linear Functionals on L p 463
6. Riesz–Thorin Convexity Theorem 466
7. Marcinkiewicz Interpolation Theorem 476
8. Problems 484
x Contents

X. TOPOLOGICAL SPACES 490

1. Open Sets and Constructions of Topologies 490
2. Properties of Topological Spaces 496
3. Compactness and Local Compactness 500
4. Product Spaces and the Tychonoff Product Theorem 507
5. Sequences and Nets 512
6. Quotient Spaces 520
7. Urysohn’s Lemma 523
8. Metrization in the Separable Case 525
9. Ascoli–Arzelà and Stone–Weierstrass Theorems 526
10. Problems 529

XI. INTEGRATION ON LOCALLY COMPACT SPACES 534

1. Setting 534
2. Riesz Representation Theorem 539
3. Regular Borel Measures 553
4. Dual to Space of Finite Signed Measures 558
5. Problems 566

XII. HILBERT AND BANACH SPACES 570

1. Definitions and Examples 570
2. Geometry of Hilbert Space 576
3. Bounded Linear Operators on Hilbert Spaces 585
4. Hahn–Banach Theorem 587
5. Uniform Boundedness Theorem 593
6. Interior Mapping Principle 595
7. Problems 599

APPENDIX A. BACKGROUND TOPICS 603

A1. Sets and Functions 603
A2. Mean Value Theorem and Some Consequences 610
A3. Inverse Function Theorem in One Variable 611
A4. Complex Numbers 613
A5. Classical Schwarz Inequality 614
A6. Equivalence Relations 614
A7. Linear Transformations, Matrices, and Determinants 615
A8. Factorization and Roots of Polynomials 618
A9. Partial Orderings and Zorn’s Lemma 623
A10. Cardinality 627
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
 Contents xi

APPENDIX B. ELEMENTARY COMPLEX ANALYSIS 631

B1. Complex Derivative and Analytic Functions 632
B2. Complex Line Integrals 636
B3. Goursat’s Lemma and the Cauchy Integral Theorem 641
B4. Cauchy Integral Formula 648
B5. Taylor’s Theorem 654
B6. Local Properties of Analytic Functions 656
B7. Logarithms and Winding Numbers 660
B8. Operations on Taylor Series 665
B9. Argument Principle 669
B10. Residue Theorem 673
B11. Evaluation of Definite Integrals 675
B12. Global Theorems in Simply Connected Regions 681
B13. Global Theorems in General Regions 694
B14. Laurent Series 696
B15. Holomorphic Functions of Several Variables 701
B16. Problems 704
Hints for Solutions of Problems 715
Selected References 793
Index of Notation 795
Index 799

CONTENTS OF ADVANCED REAL ANALYSIS

I. Introduction to Boundary-Value Problems

II. Compact Self-Adjoint Operators
III. Topics in Euclidean Fourier Analysis
IV. Topics in Functional Analysis
V. Distributions
VI. Compact and Locally Compact Groups
VII. Aspects of Partial Differential Equations
VIII. Analysis on Manifolds
IX. Foundations of Probability
X. Introduction to Wavelets
 DEPENDENCE AMONG CHAPTERS

Below is a chart of the main lines of dependence of chapters on prior chapters.

The dashed lines indicate helpful motivation but no logical dependence. Apart
from that, particular examples may make use of information from earlier chapters
that is not indicated by the chart. Appendix B is used in problems at the ends of
Chapters IV, VI, and VIII, and it is used also in Section IX.6.

I, II, III in order

V IV

VI

VIII VII X

IX

IX.6 XI

XII

xii
 PREFACE TO THE SECOND EDITION

In the years since publication of the first edition of Basic Real Analysis, many read-
ers have reacted to the book by sending comments, suggestions, and corrections.
People appreciated the overall comprehensive nature of the book, associating this
feature in part with the large number of problems that develop so many sidelights
and applications of the theory. Some people wondered whether a way might
be found for a revision to include some minimal treatment of Stokes’s Theorem
and complex analysis, despite the reservations I expressed in the original preface
about including these topics.
Along with the general comments and specific suggestions were corrections,
well over a hundred in all, that needed to be addressed in any revision. Many of
the corrections were of minor matters, yet readers should not have to cope with
errors along with new material. Fortunately no results in the first edition needed
to be deleted or seriously modified, and additional results and problems could be
included without renumbering.
For the first edition, the author granted a publishing license to Birkhäuser
Boston that was limited to print media, leaving the question of electronic publi-
cation unresolved. A major change with the second edition is that the question of
electronic publication has now been resolved, and a PDF file, called the “digital
second edition,” is being made freely available to everyone worldwide for personal
use. This file may be downloaded from the author’s own Web page and from
elsewhere.
The main changes to the first edition of Basic Real Analysis are as follows:
• A careful treatment of arc length, line integrals, and Green’s Theorem for
the plane has been added at the end of Chapter III. These aspects of Stokes’s
Theorem can be handled by the same kinds of techniques of real analysis
as in the first edition. Treatment of aspects of Stokes’s Theorem in higher
dimensions would require a great deal more geometry, for reasons given in
Section III.13, and that more general treatment has not been included.
• The core of a first course in complex analysis has been included as Appendix B.
Emphasis is on those aspects of elementary complex analysis that are useful
as tools in real analysis. The appendix includes more than 80 problems, and
some standard topics in complex analysis are developed in these problems. The
treatment assumes parts of Chapters I–III as a prerequisite. How the appendix
fits into the plan of the book is explained in the Guide for the Reader.
xiii
xiv Preface to the Second Edition

• A new section in Chapter IX proves and applies the Riesz–Thorin Convexity

Theorem, a fundamental result about L p spaces that takes advantage of ele-
mentary complex analysis.
• About 20 problems have been added at the ends of Chapters I–XII. Chiefly these
are of three kinds: some illustrate the new topics of arc length, line integrals,
and Green’s Theorem; some make use of elementary complex analysis as
in Appendix B to shed further light on results and problems in the various
chapters; and some relate to the topic of Banach spaces in Chapter XII .
• The corrections sent by readers and by reviewers have been made. The most
significant such correction was a revision to the proof of Zorn’s Lemma, the
earlier proof having had a gap.
The material in Appendix B is designed as the text of part of a first course in
complex analysis. I taught such a course myself on one occasion. A course in
complex analysis invariably begins with some preliminary material, and that can
be taken from Chapters I to III; details appear in the Guide to the Reader. Appendix
B forms the core of the course, dealing with results having an analytic flavor,
including the part of the theory due to Cauchy. The topic of conformal mapping,
which has a more geometric flavor, has been omitted, and some instructors might
feel obliged to include something on this topic in the course. Appendix B states
the Riemann Mapping Theorem at one point but does not prove it; all the tools
needed for its proof, however, are present in the appendix and its problems. Often
an instructor will end a first course in complex analysis with material on infinite
series and products of functions, or of aspects of the theory of special functions,
or on analytic continuation. Supplementary notes on any such topics would be
necessary.
It was Benjamin Levitt, Birkhäuser mathematics editor in New York, who
encouraged the writing of this second edition, who made a number of suggestions
about pursuing it, and who passed along comments from several anonymous
referees about the strengths and weaknesses of the book. I am especially grateful
to those readers who have sent me comments over the years. Many of the
corrections that were made were kindly sent to me either by S. H. Kim of South
Korea or by Jacques Larochelle of Canada. The correction to the proof of Theorem
1.35 was kindly sent by Glenn Jia of China. The long correction to the proof of
Zorn’s Lemma resulted from a discussion with Qiu Ruyue. The typesetting was
done by the program Textures using AMS-TEX, and the figures were drawn with
Mathematica.
Just as with the first edition, I invite corrections and other comments from
readers. For as long as I am able, I plan to point to a list of known corrections
from my own homepage, www.math.stonybrook.edu/∼aknapp.
A. W. KNAPP
February 2016
 PREFACE TO THE FIRST EDITION

This book and its companion volume, Advanced Real Analysis, systematically
develop concepts and tools in real analysis that are vital to every mathematician,
whether pure or applied, aspiring or established. The two books together contain
what the young mathematician needs to know about real analysis in order to
communicate well with colleagues in all branches of mathematics.
The books are written as textbooks, and their primary audience is students who
are learning the material for the first time and who are planning a career in which
they will use advanced mathematics professionally. Much of the material in the
books corresponds to normal course work. Nevertheless, it is often the case that
core mathematics curricula, time-limited as they are, do not include all the topics
that one might like. Thus the book includes important topics that may be skipped
in required courses but that the professional mathematician will ultimately want
to learn by self-study.
The content of the required courses at each university reflects expectations of
what students need before beginning specialized study and work on a thesis. These
expectations vary from country to country and from university to university. Even
so, there seems to be a rough consensus about what mathematics a plenary lecturer
at a broad international or national meeting may take as known by the audience.
The tables of contents of the two books represent my own understanding of what
that degree of knowledge is for real analysis today.
Key topics and features of Basic Real Analysis are as follows:
• Early chapters treat the fundamentals of real variables, sequences and series
of functions, the theory of Fourier series for the Riemann integral, metric
spaces, and the theoretical underpinnings of multivariable calculus and ordi-
nary differential equations.
• Subsequent chapters develop the Lebesgue theory in Euclidean and abstract
spaces, Fourier series and the Fourier transform for the Lebesgue integral,
point-set topology, measure theory in locally compact Hausdorff spaces, and
the basics of Hilbert and Banach spaces.
• The subjects of Fourier series and harmonic functions are used as recurring
motivation for a number of theoretical developments.
• The development proceeds from the particular to the general, often introducing
examples well before a theory that incorporates them.
xv
xvi Preface to the First Edition

• More than 300 problems at the ends of chapters illuminate aspects of the
text, develop related topics, and point to additional applications. A separate
55-page section “Hints for Solutions of Problems” at the end of the book gives
detailed hints for most of the problems, together with complete solutions for
many.
Beyond a standard calculus sequence in one and several variables, the most
important prerequisite for using Basic Real Analysis is that the reader already
know what a proof is, how to read a proof, and how to write a proof. This
knowledge typically is obtained from honors calculus courses, or from a course
in linear algebra, or from a first junior-senior course in real variables. In addition,
it is assumed that the reader is comfortable with a modest amount of linear algebra,
including row reduction of matrices, vector spaces and bases, and the associated
geometry. A passing acquaintance with the notions of group, subgroup, and
quotient is helpful as well.
Chapters I–IV are appropriate for a single rigorous real-variables course and
may be used in either of two ways. For students who have learned about proofs
from honors calculus or linear algebra, these chapters offer a full treatment of real
variables, leaving out only the more familiar parts near the beginning—such as
elementary manipulations with limits, convergence tests for infinite series with
positive scalar terms, and routine facts about continuity and differentiability. For
students who have learned about proofs from a first junior-senior course in real
variables, these chapters are appropriate for a second such course that begins with
Riemann integration and sequences and series of functions; in this case the first
section of Chapter I will be a review of some of the more difficult foundational
theorems, and the course can conclude with an introduction to the Lebesgue
integral from Chapter V if time permits.
Chapters V through XII treat Lebesgue integration in various settings, as well
as introductions to the Euclidean Fourier transform and to functional analysis.
Typically this material is taught at the graduate level in the United States, fre-
quently in one of three ways: The first way does Lebesgue integration in Euclidean
and abstract settings and goes on to consider the Euclidean Fourier transform in
some detail; this corresponds to Chapters V–VIII. A second way does Lebesgue
integration in Euclidean and abstract settings, treats L p spaces and integration on
locally compact Hausdorff spaces, and concludes with an introduction to Hilbert
and Banach spaces; this corresponds to Chapters V–VII, part of IX, and XI–XII.
A third way combines an introduction to the Lebesgue integral and the Euclidean
Fourier transform with some of the subject of partial differential equations; this
corresponds to some portion of Chapters V–VI and VIII, followed by chapters
from the companion volume, Advanced Real Analysis.
In my own teaching, I have most often built one course around Chapters I–IV
and another around Chapters V–VII, part of IX, and XI–XII. I have normally
 Preface to the First Edition xvii

assigned the easier sections of Chapters II and X as outside reading, indicating

the date when the lectures would begin to use that material.
More detailed information about how the book may be used with courses may
be deduced from the chart “Dependence among Chapters” on page xiv and the
section “Guide to the Reader” on pages xv–xvii.
The problems at the ends of chapters are an important part of the book. Some
of them are really theorems, some are examples showing the degree to which
hypotheses can be stretched, and a few are just exercises. The reader gets no
indication which problems are of which type, nor of which ones are relatively
easy. Each problem can be solved with tools developed up to that point in the
book, plus any additional prerequisites that are noted.
Two omissions from the pair of books are of note. One is any treatment of
Stokes’s Theorem and differential forms. Although there is some advantage,
when studying these topics, in having the Lebesgue integral available and in
having developed an attitude that integration can be defined by means of suitable
linear functionals, the topic of Stokes’s Theorem seems to fit better in a book
about geometry and topology, rather than in a book about real analysis.
The other omission concerns the use of complex analysis. It is tempting to try
to combine real analysis and complex analysis into a single subject, but my own
experience is that this combination does not work well at the level of Basic Real
Analysis, only at the level of Advanced Real Analysis.
Almost all of the mathematics in the two books is at least forty years old, and I
make no claim that any result is new. The books are a distillation of lecture notes
from a 35-year period of my own learning and teaching. Sometimes a problem at
the end of a chapter or an approach to the exposition may not be a standard one,
but no attempt has been made to identify such problems and approaches. In the
reverse direction it is possible that my early lecture notes have directly quoted
some source without proper attribution. As an attempt to rectify any difficulties
of this kind, I have included a section of “Acknowledgments” on pages xix–xx
of this volume to identify the main sources, as far as I can reconstruct them, for
those original lecture notes.
I am grateful to Ann Kostant and Steven Krantz for encouraging this project and
for making many suggestions about pursuing it, and to Susan Knapp and David
Kramer for helping with the readability. The typesetting was by AMS-TEX, and
the figures were drawn with Mathematica.
I invite corrections and other comments from readers. I plan to maintain a list
of known corrections on my own Web page.
A. W. KNAPP
May 2005
 LIST OF FIGURES

1.1. Approximate identity 60

1.2. Fourier series of sawtooth function 66
1.3. Dirichlet kernel 70
2.1. An open set centered at the origin in the hedgehog space 89
2.2. Open ball contained in an intersection of two open balls 93
3.1. Polygonal approximation for estimating arc length 182
3.2. Cycloid 193
3.3. A piecewise C 1 curve that retraces part of itself 200
3.4. Green’s Theorem for an annulus 206
3.5. Failure of inscribed triangles to give a useful notion of
surface area 210
4.1. Graphs of solutions of some first-order ordinary differential
equations 220
4.2. Integral curve of a vector field 234
4.3. Graph of Bessel function J0 (t) 260
6.1. Construction of a Cantor function F 381
7.1. Rising Sun Lemma 396
9.1. Geometric description of points in the Riesz–Thorin
Convexity Theorem 468
B.1. First bisection of the rectangle R in Goursat’s Lemma 642
B.2. Handling an exceptional point in Goursat’s Lemma 644
B.3. Reduction steps in the proof of Theorem B.9 645
R
B.4. Computation of ∞ (≥ − z)−1 d≥ over a standard circle
when z is not the center 647
B.5. Construction of a cycle ∞ and a point a with n(∞ , a) = 1 684
B.6. Two polygonal paths σ1 and σ2 from z 0 to z in the region U 686
B.7. Boundary cycle: (a) as the difference of two cycles,
(b) as the sum of three cycles 696

xviii
 ACKNOWLEDGMENTS

The author acknowledges the sources below as the main ones he used in preparing
the lectures from which this book evolved. Any residual unattributed direct
quotations in the book are likely to be from these.
The descriptions below have been abbreviated. Full descriptions of the books
and Stone article may be found in the section “Selected References” at the
end of the book. The item “Feller’s Functional Analysis” refers to lectures by
William Feller at Princeton University for Fall 1962 and Spring 1963, and the
item “Nelson’s Probability” refers to lectures by Edward Nelson at Princeton
University for Spring 1963.
This list is not to be confused with a list of recommended present-day reading
for these topics; newer books deserve attention.
CHAPTER I. Rudin’s Principles of Mathematical Analysis, Zygmund’s
Trigonometric Series.
CHAPTER II. Feller’s Functional Analysis, Kelley’s General Topology,
Stone’s “A generalized Weierstrass approximation theorem.”
CHAPTER III. Rudin’s Principles of Mathematical Analysis, Spivak’s
Calculus on Manifolds.
CHAPTER IV. Coddington–Levinson’s Theory of Ordinary Differential
Equations, Kaplan’s Ordinary Differential Equations.
CHAPTER V. Halmos’s Measure Theory, Rudin’s Principles of Mathematical
Analysis.
CHAPTER VI. Rudin’s Principles of Mathematical Analysis, Rudin’s Real and
Complex Analysis, Saks’s Theory of the Integral, Spivak’s Calculus on Manifolds,
Stein–Weiss’s Introduction to Fourier Analysis on Euclidean Spaces.
CHAPTER VII. Riesz–Nagy’s Functional Analysis, Zygmund’s Trigonometric
Series.
CHAPTER VIII. Stein’s Singular Integrals and Differentiability Properties of
Functions, Stein–Weiss’s Introduction to Fourier Analysis on Euclidean Spaces.
CHAPTER IX. Dunford–Schwartz’s Linear Operators, Feller’s Functional
Analysis, Halmos’s Measure Theory, Saks’s Theory of the Integral, Stein’s
Singular Integrals and Differentiability Properties of Functions.
xix
xx Acknowledgments

CHAPTER X. Kelley’s General Topology, Nelson’s Probability.

CHAPTER XI. Feller’s Functional Analysis, Halmos’s Measure Theory,
Nelson’s Probability.
CHAPTER XII. Dunford–Schwartz’s Linear Operators, Feller’s Functional
Analysis, Riesz–Nagy’s Functional Analysis.
APPENDIX A. For Sections 1, 9, 10: Dunford–Schwartz’s Linear Operators,
Hayden–Kennison’s Zermelo–Fraenkel Set Theory, Kelley’s General Topology.
APPENDIX B. Ahlfors, Complex Analysis, Gunning’s Introduction to Holo-
morphic Functions of Several Variables.
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
 GUIDE FOR THE READER

This section is intended to help the reader find out what parts of each chapter are
most important and how the chapters are interrelated. Further information of this
kind is contained in the abstracts that begin each of the chapters.
The book pays attention to certain recurring themes in real analysis, allowing
a person to see how these themes arise in increasingly sophisticated ways. Ex-
amples are the role of interchanges of limits in theorems, the need for certain
explicit formulas in the foundations of subject areas, the role of compactness and
completeness in existence theorems, and the approach of handling nice functions
first and then passing to general functions.
All of these themes are introduced in Chapter I, and already at that stage they
interact in subtle ways. For example, a natural investigation of interchanges of
limits in Sections 2–3 leads to the discovery of Ascoli’s Theorem, which is a
fundamental compactness tool for proving existence results. Ascoli’s Theorem
is proved by the “Cantor diagonal process,” which has other applications to
compactness questions and does not get fully explained until Chapter X. The
consequence is that, no matter where in the book a reader plans to start, everyone
will be helped by at least leafing through Chapter I.
The remainder of this section is an overview of individual chapters and groups
of chapters.
Chapter I. Every section of this chapter plays a role in setting up matters
for later chapters. No knowledge of metric spaces is assumed anywhere in the
chapter. Section 1 will be a review for anyone who has already had a course in real-
variable theory; the section shows how compactness and completeness address
all the difficult theorems whose proofs are often skipped in calculus. Section 2
begins the development of real-variable theory at the point of sequences and series
of functions. It contains interchange results that turn out to be special cases of
the main theorems of Chapter V. Sections 8–9 introduce the approach of handling
nice functions before general functions, and Section 10 introduces Fourier series,
which provided a great deal of motivation historically for the development of real
analysis and are used in this book in that same way. Fourier series are somewhat
limited in the setting of Chapter I because one encounters no class of functions,
other than infinitely differentiable ones, that corresponds exactly to some class of
Fourier coefficients; as a result Fourier series, with Riemann integration in use,
xxi
xxii Guide for the Reader

are not particularly useful for constructing new functions from old ones. This
defect will be fixed with the aid of the Lebesgue integral in Chapter VI.
Chapter II. Now that continuity and convergence have been addressed on
the line, this chapter establishes a framework for these questions in higher-
dimensional Euclidean space and other settings. There is no point in ad hoc
definitions for each setting, and metric spaces handle many such settings at once.
Chapter X later will enlarge the framework from metric spaces to “topological
spaces.” Sections 1–6 of Chapter II are routine. Section 7, on compactness
and completeness, is the core. The Baire Category Theorem in Section 9 is not
used outside of Chapter II until Chapter XII, and it may therefore be skipped
temporarily. Section 10 contains the Stone–Weierstrass Theorem, which is a
fundamental approximation tool. Section 11 is used in some of the problems but
is not otherwise used in the book.
Chapter III. This chapter does for the several-variable theory what Chapter I
has done for the one-variable theory. The main results are the Inverse and Implicit
Function Theorems in Section 6 and the change-of-variables formula for multiple
integrals in Section 10. The change-of-variables formula has to be regarded as
only a preliminary version, since what it directly accomplishes for the change
to polar coordinates still needs supplementing; this difficulty will be repaired in
Chapter VI with the aid of the Lebesgue integral. Section 4, on exponentials of
matrices, may be skipped if linear systems of ordinary differential equations are
going to be skipped in Chapter IV. Sections 11–13 contain a careful treatment
of arc length, line integrals, and Green’s Theorem for the plane. These sections
emphasize properties of parametrized curves that are unchanged when the curve
is reparametrized; length is an example. An important point to bear in mind is
that two curves are always reparametrizations of each other if they have the same
image in the plane and they are both traced out in one-one fashion. This theory
is tidier if carried out in the context of Lebesgue integration, but its placement in
the text soon after Riemann integration is traditional. The difficulty with using
Riemann integrals arises already in the standard proof of Green’s Theorem for
a circle, which parametrizes each quarter of the circle twice, once with y in
terms of x and once with x in terms of y. The problem is that in each of these
parametrizations, the derivative of the one variable with respect to the other is
unbounded, and thus arc length is not given by a Riemann integral. Some of
the problems at the end of the chapter introduce harmonic functions; harmonic
functions will be combined with Fourier series in problems in later chapters to
motivate and illustrate some of the development.
Chapter IV provides theoretical underpinnings for the material in a traditional
undergraduate course in ordinary differential equations. Nothing later in the book
is logically dependent on Chapter IV; however, Chapter XII includes a discussion
of orthogonal systems of functions, and the examples of these that arise in Chapter
 Guide for the Reader xxiii

IV are helpful as motivation. Some people shy away from differential equations
and might wish to treat Chapter IV only lightly, or perhaps not at all. The subject
is nevertheless of great importance, and Chapter IV is the beginning of it. A
minimal treatment of Chapter IV might involve Sections 1–2 and Section 8, all
of which visibly continue the themes begun in Chapter I.
Chapters V–VI treat the core of measure theory—including the basic conver-
gence theorems for integrals, the development of Lebesgue measure in one and
several variables, Fubini’s Theorem, the metric spaces L 1 and L 2 and L ∞ , and
the use of maximal theorems for getting at differentiation of integrals and other
theorems concerning almost-everywhere convergence. In Chapter V Lebesgue
measure in one dimension is introduced right away, so that one immediately has
the most important example at hand. The fundamental Extension Theorem for
getting measures to be defined on σ -rings and σ -algebras is stated when needed but
is proved only after the basic convergence theorems for integrals have been proved;
the proof in Sections 5–6 may be skipped on first reading. Section 7, on Fubini’s
Theorem, is a powerful result about interchange of integrals. At the same time
that it justifies interchange, it also constructs a “double integral”; consequently
the section prepares the way for the construction in Chapter VI of n-dimensional
Lebesgue measure from 1-dimensional Lebesgue measure. Section 10 introduces
normed linear spaces along with the examples of L 1 and L 2 and L ∞ , and it goes
on to establish some properties of all normed linear spaces. Chapter VI fleshes
out measure theory as it applies to Euclidean space in more than one dimension.
Of special note is the Lebesgue-integration version in Section 5 of the change-
of-variables formula for multiple integrals and the Riesz–Fischer Theorem in
Section 7. The latter characterizes square-integrable periodic functions by their
Fourier coefficients and makes the subject of Fourier series useful in constructing
functions. Differentiation of integrals in approached in Section 6 of Chapter VI
as a problem of estimating finiteness of a quantity, rather than its smallness; the
device is the Hardy–Littlewood Maximal Theorem, and the approach becomes a
routine way of approaching almost-everywhere convergence theorems. Sections
8–10 are of somewhat less importance and may be omitted if time is short;
Section 10 is applied only in Section IX.6.
Chapters VII–IX are continuations of measure theory that are largely indepen-
dent of each other. Chapter VII contains the traditional proof of the differentiation
of integrals on the line via differentiation of monotone functions. No later chapter
is logically dependent on Chapter VII; the material is included only because of its
historical importance and its usefulness as motivation for the Radon–Nikodym
Theorem in Chapter IX. Chapter VIII is an introduction to the Fourier transform
in Euclidean space. Its core consists of the first four sections, and the rest may be
considered as optional if Section IX.6 is to be omitted. Chapter IX concerns L p
spaces for 1 ≤ p ≤ ∞; only Section 6 makes use of material from Chapter VIII.
xxiv Guide for the Reader

Chapter X develops, at the latest possible time in the book, the necessary part
of point-set topology that goes beyond metric spaces. Emphasis is on product
and quotient spaces, and on Urysohn’s Lemma concerning the construction of
real-valued functions on normal spaces.
Chapter XI contains one more continuation of measure theory, namely special
features of measures on locally compact Hausdorff spaces. It provides an example
beyond L p spaces in which one can usefully identify the dual of a particular
normed linear space. These chapters depend on Chapter X and on the first five
sections of Chapter IX but do not depend on Chapters VII–VIII.
Chapter XII is a brief introduction to functional analysis, particularly to Hilbert
spaces, Banach spaces, and linear operators on them. The main topics are the
geometry of Hilbert space and the three main theorems about Banach spaces.
Appendix B is the core of a first course in complex analysis. The prerequisites
from real analysis for reading this appendix consist of Sections 1–7 of Chapter I,
Section 1–8 of Chapter II, and Sections 1–3, 5–6, and 11–12 of Chapter III;
Section 6 of Chapter III is used only lightly. According to the plan of the book,
it is possible to read the text of Chapters I–XII without using any of Appendix B,
but results of Appendix B are applied in problems at the end of Chapters IV,
VI, and VIII, as well as in one spot in Section IX.6, in order to illustrate the
interplay between real analysis and complex analysis. The problems at the end
of Appendix B are extensive and are of particular importance, since the topics of
linear fractional transformations, normal families, and the relationship between
harmonic functions and analytic functions are developed there and not otherwise
in the book.
 STANDARD NOTATION

Item Meaning
#S or |S| number of elements in S
∅ empty set
{x ∈ E | P} the set of x in E such that P holds
Ec complement of the set E
SE ∪ F, ET∩ F, E − F union, intersection, difference of sets
α Eα , α Eα union, intersection of the sets E α
E ⊆ F, E ⊇ F E is contained in F, E contains F
E × F, ×s∈S X s products of sets
(a1 , . . . , an ), {a1 , . . . , an } ordered n-tuple, unordered n-tuple
f : E →Ø F, x 7→ f (x) function, effect of function
f ◦ g, f Ø E composition of f following g, restriction to E
f ( · , y) the function x 7→ f (x, y)
f (E), f −1 (E) direct and inverse image of a set
°δi j¢ Kronecker delta: 1 if i = j, 0 if i 6= j
n
k binomial coefficient
n positive, n negative n > 0, n < 0
Z, Q, R, C integers, rationals, reals, complex numbers
max
P (and Q similarly min) maximum of finite subset of a totally ordered set
or sum or product, possibly with a limit operation
countable finite or in one-one correspondence with Z
[x] greatest integer ≤ x if x is real
Re z, Im z real and imaginary parts of complex z
z̄ complex conjugate of z
|z| absolute value of z
1 multiplicative identity
1 or I identity matrix or operator
dim V dimension of vector space
Rn , Cn spaces of column vectors
det A determinant of A
Atr transpose of A
diag(a1 , . . . , an ) diagonal matrix
∼
= is isomorphic to, is equivalent to

xxv
Random documents with unrelated
content Scribd suggests to you:
adornantur[431].» Ces décorations ont dû être faites sur le passage
de Clovis; or, entre l'évêché et le baptistère, il n'y avait pas de place
suffisante pour qu'un cortège pût se déployer; on n'avait qu'un faible
espace à franchir, et l'on ne rencontrait sur son chemin ni ces rues,
ni ces églises qui avaient pris une si brillante parure. Au contraire, si
l'on admet que le roi des Francs, avec sa suite, est parti de son
palais, de la domus regia de la porte Basée, en suivant, pour gagner
la cathédrale, la grande rue qui conduisait jusqu'au centre de la cité,
alors tout s'explique, la cérémonie s'accomplit d'une façon très
naturelle, la procession peut être admise et n'est plus l'objet
d'aucune discussion[432].
[427] «Eundi via ad baptisterium a domo regia præparatur, velisque atque cortinis
depictis ex utraque parte prætenditur et desuper adumbratur. Plateæ sternuntur et
ecclesiæ componuntur... Sicque, præcedentibus sacrosanctis evangeliis et
crucibus, cum ymnis et canticis spiritalibus atque letaniis, sanctorumque
nominibus acclamatis, sanctus pontifex, manum tenens regis, a domo regia pergit
ad baptisterium, subsequente regina et populo.» Ch. iv, 62, AA. SS. Boll., octobre,
t. i, p. 146. Flodoard n'a fait que copier ce passage, Hist., l. i, ch. xiii.
[428] Notice sur le Baptême de Clovis, par M. le chanoine Cerf (1891), p. 6 et suiv.
[429] H. Schrörs, Hinkmar Erzbischof von Reims, p. 448.
[430] Pour ce qui concerne le récit de la cérémonie; quant au lieu de la résidence
de Clovis et à l'oratoire de Saint-Pierre, il paraît, comme nous l'avons dit, s'inspirer
de traditions locales.
[431] Hist. Francorum, l. ii, ch. xxxi.
[432] Le P. Jubaru, l. cit., p. 316-317.
Nous avons attribué plus haut à l'église de la porte Basée, et non à
la chapelle du palais, le legs fait dans le grand testament de saint
Remi «ecclesiæ Sancti Petri infra urbem.» C'est à elle aussi que
nous assignons le legs de trois sous d'or fait au septième siècle par
l'évêque Sonnace «ad basilicam Sancti Petri in civitate[433],» et le
don de l'évêque Landon à l'église «Sancti Petri ad cortem[434].»
Cette cortis est bien incontestablement la curtis dominica nommée
dans le grand testament.
[433] Flodoard, Hist., l. ii, ch. v, Mon. Germ., t. xiii, p. 454.
[434] Ibid., l. ii, ch. vi, Mon. Germ., t. xiii, p. 455.—L'évêque Sonnace, mourut le 20
octobre 631, et Landon le 14 mars 649.
Il faut aussi sans doute identifier avec cette église l'«ecclesia Sancti
Petri quæ est infra muros urbis Remensis» de la Vie de sainte
Clotilde[435]. Peut-être cependant, à la date assez tardive où écrivait
l'auteur de cette vie, s'était-il déjà produit avec l'église Saint-Pierre-
le-Vieil une confusion que nous verrons prendre corps à une époque
plus avancée du moyen âge.
[435] Mon. Germ., Scriptores rerum merovingicarum, t. ii; Rec. des hist. de la
France, t. iii, p. 401.
La multiplicité des églises et des chapelles consacrées à saint
Pierre, qui existaient jadis à Reims, en rend souvent la distinction
très difficile. Ainsi, quand l'auteur de la Vita sancti Gildardi,
composée vers le dixième siècle et récemment mise en lumière par
les Bollandistes[436], parle de la «basilica Sancti Petri quæ nunc
dicitur ad palatium», nous ne saurions dire au juste quel édifice il a
en vue. En raison de la date de ce texte, nous inclinons à croire qu'il
s'agit ici de la chapelle du palais de l'archevêché.
[436] Analecta Boll., t. viii, p. 397.
A la fin de la période carolingienne, il s'est produit une opinion qui
voulait associer au récit du baptême de Clovis le souvenir d'une
ancienne église dédiée à saint Pierre. Elle n'a aucune valeur
traditionnelle et est née d'une méprise qui s'est manifestée
postérieurement à Hincmar. Ni Grégoire de Tours ni Hincmar ne
laissent supposer que Clovis ait été baptisé dans une basilique de
Saint-Pierre. Hincmar nous représente seulement, ainsi que nous
l'avons vu, Clovis, à la veille de son baptême, conférant avec saint
Remi dans l'«oratorium Sancti Petri», contigu à ses appartements.
Ce passage a été la source de toute l'erreur. On a retenu
vaguement, un peu plus tard, ce nom de saint Pierre; on en a
exagéré la portée, et l'on en a fait à tort l'application au lieu du
baptême de Clovis. Et l'auteur de la Vie de sainte Clotilde, par
exemple, est venu nous dire que la pieuse reine avait une grande
prédilection pour l'église de Saint-Pierre, parce que son époux y
avait reçu la grâce du baptême: «Hanc itaque ecclesiam cunctis
diebus quibus advixit, multum dilexit et excoluit, pro eo quod vel
suus rex Ludovicus in ea sancti baptismatis gratiam accepit[437].» Il
se fait ici évidemment l'écho, non d'une tradition sérieuse, mais
d'une conjecture erronée. Au reste, cette Vie de sainte Clotilde n'est
qu'une compilation sans caractère original, rédigée vers le dixième
siècle[438]. Un autre ouvrage, qui est à peu près du même temps et
n'a pas plus d'autorité au point de vue historique, la Vie de saint
Gildard, semble placer aussi la cérémonie du baptême dans la
«basilica Sancti Petri[439].» On aurait tort d'attribuer quelque
importance à ces deux textes; ils ne prouvent rien, sinon qu'il s'était
produit sur ce point, au dixième siècle, une croyance absolument
fausse.
[437] Ibid.
[438] B. Krusch, Script. rerum merov., t. II, p. 341.
[439] «... In civitatem Remorum venientes, in basilica Sancti Petri, quæ nunc
dicitur ad palatium, missas celebraverunt, et ea quæ Dei sunt agentes, beatus
Remedius regem baptizavit, et de sacro fonte illum beatus Medardus suscepit.»
Analecta Boll., t. VIII, p. 397.
L'idée du baptême de Clovis dans l'église de Saint-Pierre une fois
admise, il s'est formé,—et cela dès le moyen âge,—un courant
d'opinion en faveur de l'église paroissiale de Saint-Pierre-le-Vieil. Un
chanoine de Reims, du dix-septième siècle, Pierre Cocquault, dans
un vaste recueil historique dont le manuscrit est aujourd'hui
conservé à la bibliothèque de cette ville, nous révèle à ce sujet un
détail assez curieux. En l'année 1486, les paroissiens de Saint-
Pierre-le-Vieil faisaient courir le bruit que Clovis avait été baptisé
dans leur église. «Le 22 novembre, ajoute notre chroniqueur, leur fut
imposé silence comme estant chose non véritable, car Clovis fut
baptisé à l'église de Reims.» Et il fait observer, en s'appuyant sur le
vocable de saint Pierre, conservé de son temps à la chapelle basse
de l'archevêché, que l'oratorium Sancti Petri, indiqué par Hincmar,
était dans le palais de l'évêque et à proximité de l'église
cathédrale[440].
[440] «Les parrochians de l'église de Saint Pierre le Vielle de Reims faissoient
courir un bruict contre toutes apparences de vérité, que la Sainte Ampoule avoit
esté aultrefois en ceste paroisse, et que Clovis, premier roy de France chrestien, y
avoit esté baptissé et coronné roy de France. Le 22 novembre leur fut imposé
silence comme estant chose non véritable, car Clovis fut baptissé à l'église de
Reims, et en ce lieu la Sainte Ampoule y fut apportée à saint Remy.» Chronique
de Pierre Cocquault, t. IV, fol. 75 vº.
Ainsi tout ce que l'on a dit de Saint-Pierre-le-Vieil, à propos du
baptême de Clovis, est inexact, et l'on doit, en la question, mettre
cette église complètement à l'écart. Nous ignorons, du reste,
entièrement son origine et le temps de sa fondation. L'épithète de
Vieil (Sancti Petri Veteris) lui a été appliquée de bonne heure: on la
trouve dès le douzième siècle[441]; mais la vieillesse d'un monument
est une chose relative, et l'on se tromperait peut-être en assignant à
notre église une date trop reculée. En tout cas, nous ne voyons dans
Flodoard aucune mention qui puisse lui être rattachée avec
certitude. Les plus anciens documents qui la concernent ne nous
permettent pas de remonter au delà du douzième siècle. En 1172,
on y établit une confrérie, dite de Saint-Pierre-aux-Clercs, dont les
titres originaux furent brûlés en 1330, dans un grand incendie qui
consuma plusieurs maisons de la ville[442]. Par suite de cet
événement, la série des pièces comprenant l'ancien chartrier de
l'église Saint-Pierre ne s'ouvre plus qu'au quatorzième siècle, et
encore les pièces de cette dernière date sont-elles rares, car ce
fonds, tel qu'il existe maintenant aux archives de Reims, offre bien
des lacunes. Les matériaux dont nous disposons sont donc
insuffisants pour reconstituer toute l'histoire de cette paroisse, et
surtout pour éclaircir le mystère de son origine.
[441] Ordinaire de l'église de Reims du douzième siècle, Ul. Chevalier,
Bibliothèque liturgique, t. VII, p. 298-299; cf. une charte du 4 février 1259, citée
dans Varin, Archives administratives de Reims, t. I, p. 788.
[442] Archives de Reims, fonds de la paroisse Saint-Pierre, Inventaire des titres et
papiers de la confrérie du Saint-Nom-de-Jésus et de Saint-Pierre-aux-Clercs,
1724, p. 9 à 11.

Nous savons qu'on a parlé aussi d'une prétendue fondation, faite par
saint Remi en l'église Saint-Pierre-le-Vieil; mais c'est une simple
conjecture, sans aucun fondement, ainsi que Marlot l'a fort bien vu
en son histoire de Reims: «On tient, dit-il, que cette église servit
autrefois d'un monastère où saint Remy logea quarante vefves, dont
il est parlé en la vie de saint Thierry, et qu'elle devint paroisse,
lorsque ces vefves furent transférées à Sainte-Agnès; mais... Floard
ne dit rien de tout cela[443].» Flodoard, effectivement, garde sur ce
point un silence complet, et la Vie de saint Thierry ne dit rien non
plus qui autorise cette supposition. Nous sommes encore en
présence d'une de ces fausses légendes dont on a encombré les
histoires locales, et qu'il appartient à la critique d'éliminer.
[443] Histoire de la ville, cité et université de Reims, t. I, p. 689.
Pour en revenir à Clovis, il est certain qu'aucune église de Saint-
Pierre n'a été témoin de son baptême, et que les traditions
invoquées en faveur de cette opinion n'ont rien d'historique. Ainsi
s'écroulent par la base toutes les raisons accumulées pour
démontrer que la cérémonie a eu lieu dans un baptistère situé près
de l'ancienne cathédrale, dédiée aux Apôtres, et devenue plus tard
l'église Saint-Symphorien[444]. Cette opinion s'appuie surtout sur les
passages précédemment cités des Vies de saint Gildard et de sainte
Clotilde; c'est là un étai bien fragile, sur lequel on ne peut se reposer
en sécurité. On pourrait observer au surplus que le vocable des
Apôtres n'est pas tout à fait identique au vocable de saint Pierre;
mais à quoi bon, puisqu'il ne doit plus être question ici de saint
Pierre lui-même?
[444] Voy. la notice de M. le chanoine Cerf sur le Baptême de Clovis, p. 16 et suiv.
Clovis n'a pas été baptisé davantage dans l'église de Saint-Martin de
Reims, ainsi que l'a supposé Adrien de Valois[445], pour expliquer
une allusion de la lettre de saint Nizier, dont nous avons parlé plus
haut, et d'après laquelle Clovis, décidé à embrasser la foi
chrétienne, se serait rendu «ad limina domini Martini[446]». Cette
expression ne peut assurément désigner autre chose que la
basilique de Saint-Martin de Tours, qui reçut, en effet, une visite
solennelle du roi des Francs[447].
[445] AA. SS. Boll., octobre, t. I, p. 82. Cf. Krusch, Zwei Heiligenleben des Jonas
von Susa, p. 443.
[446] Voy. ci-dessus.
[447] En l'année 508, au retour de sa campagne contre les Visigoths, Grégoire de
Tours, Hist. Francorum, l. II, chap. xxxvii et xxxviii.
Ainsi ces diverses solutions doivent être écartées, et Clovis, suivant
toute vraisemblance, a reçu le baptême dans un baptistère attenant
à la cathédrale qui existait de son temps, à celle que saint Nicaise
avait bâtie en l'honneur de la sainte Vierge[448]. Il n'y avait alors sans
doute à Reims, comme dans les autres villes épiscopales, qu'un seul
baptistère, où l'évêque administrait le sacrement à des époques
déterminées[449]. C'est bien là le templum baptisterii, désigné par
Grégoire de Tours dans son récit de la conversion de Clovis[450].
Toutes les présomptions sont en faveur de cette assertion; pour la
combattre, il faudrait avoir des preuves, or on n'en découvre nulle
part.
[448] Il ne serait pas impossible, à la rigueur, que l'on ait conservé alors un
baptistère dépendant de la cathédrale antérieure, celle qui était dédiée aux
Apôtres; mais il était plus naturel qu'en construisant une nouvelle cathédrale, au
commencement du cinquième siècle, on lui eût annexé un nouveau baptistère.
[449] Martigny, Dict. des antiquités chrétiennes, p. 74.
[450] Hist. Francorum, t. II, chap. xxxi.
Du reste, cette opinion avait déjà cours au neuvième siècle, Louis le
Pieux, dans un diplôme donné à l'archevêque Ebbon, entre les
années 817 et 825[451], pour lui permettre d'employer les pierres des
murs de Reims à la reconstruction de la cathédrale, rappelle que
Clovis, son prédécesseur, a été dans cette église régénéré par le
baptême[452]. Nous n'insistons pas, bien entendu, sur un
témoignage aussi tardif, et nous ne lui attribuons aucune force
probante; nous nous bornons à reconnaître que, malgré les
divergences qui allaient bientôt se manifester, la vérité historique
avait dès lors reçu une sorte de consécration officielle.
[451] Telle est la date assignée par Sickel, Acta Karolin., II, p. 150 et 330.
[452] «... Metropolis urbis sancta mater nostra ecclesia, in honore sanctæ
semperque virginis ac [Dei] genitricis Mariæ consecrata,... in qua, auctore Deo et
cooperante sancto Remigio, gens nostra Francorum, cum æquivoco nostro rege
ejusdem gentis, sacri fontis baptismate ablui... promeruit.» Flodoard, Hist., l. II,
chap. xix, Mon. Germ., t. XIII, p. 469.
Il nous reste à rechercher en quel endroit au juste s'élevait le
baptistère. Deux textes peuvent nous fournir quelques indices à ce
sujet. On lit dans une continuation de la chronique de Flodoard que
l'archevêque Adalbéron fit détruire, en l'année 976, un ouvrage muni
d'arcades, qui était voisin des portes de l'église de Notre-Dame de
Reims, et près duquel se trouvait un autel dédié au Saint Sauveur, et
des fonts d'un admirable travail: «Destruxit Adalbero arcuatum opus
quod erat secus valvas ecclesiæ Sanctæ Mariæ Remensis, supra
quod altare Sancti Salvatoris habebatur, et fontes miro opere erant
positi.» Ce passage semble bien s'appliquer à un baptistère primitif,
construction isolée, située en dehors et à proximité de l'entrée de
l'église, et telle a été l'interprétation adoptée par Marlot[453]. Mais
Richer, en rapportant le même fait dans sa chronique, se sert de
termes assez obscurs, qui viennent compliquer un peu la question. Il
nous parle d'arcades élevées qui s'avançaient depuis l'entrée
jusqu'au quart environ de la basilique entière, et que l'archevêque fit
démolir pour donner à celle-ci plus d'ampleur: «Fornices qui ab
ecclesiæ introitu per quartam pene totius basilicæ partem eminenti
structura distendebantur, penitus diruit. Unde et ampliori receptaculo
et digniore scemate tota ecclesia decorata est[454].» On pourrait
croire, en lisant ces lignes, qu'il s'agit d'une construction intérieure
qui encombrait l'église, d'une tribune peut-être, ainsi que le pensait
Jules Quicherat[455]. A vrai dire, on ne se représente guère ce que
pouvait être une semblable disposition, et l'explication est en somme
peu satisfaisante. Le P. Jubaru, dans l'article déjà cité[456], émet à ce
sujet d'autres vues qui nous semblent fort justes, et qui concilient
très bien les données fournies par nos deux chroniqueurs. Nous
croyons qu'il a eu le mérite de découvrir la vraie solution du
problème.
[453] Metr. Remens. hist., t. I, p. 160.
[454] L. III, ch. xxii.
[455] Mélanges d'archéologie, moyen âge, p. 133.
[456] P. 301 à 310.
La cathédrale bâtie par saint Nicaise, celle qu'a vue Clovis, avait des
dimensions restreintes dont on peut se faire aujourd'hui encore une
idée assez exacte. On a conjecturé non sans raison que l'autel, situé
dans l'abside, devait être à la place qu'occupe maintenant le maître-
autel de la cathédrale actuelle[457]. La tradition a gardé aussi un
souvenir précis de l'endroit où s'ouvrait la porte de la basilique.
C'était là, sur le seuil même, que saint Nicaise avait été massacré
par les Vandales, et le lieu de son martyre était resté l'objet d'une
vénération non interrompue à travers les âges. Ce lieu correspond à
la sixième travée de notre cathédrale à partir du portail; au treizième
siècle, il était indiqué par un petit monument commémoratif; une
dalle de marbre le désigne de nos jours à la piété des fidèles. Le P.
Jubaru pense que la basilique primitive, suivant l'usage du temps,
était précédée d'un atrium, parvis carré entouré de portiques; au
milieu de ce parvis ou sur l'un des côtés s'élevait l'édicule du
baptistère[458]. D'après lui, l'église reconstruite par Ebbon et
achevée par Hincmar au neuvième siècle, aurait été prolongée vers
le chœur, mais la façade n'aurait pas changé de place, et l'atrium
ancien, ainsi que le baptistère, aurait été respecté. Leur destruction
a été l'œuvre d'Adalbéron; l'arcuatum opus, l'ouvrage garni
d'arcades qu'il démolit, doit s'entendre des galeries cintrées du
portique qui régnait autour du parvis. Avec ce portique, il supprima le
baptistère qui renfermait l'autel du Saint Sauveur et les fonts, sans
doute richement décorés de marbres et de mosaïques, dont on
admirait le beau travail. La préposition supra, employée ici par le
continuateur de Flodoard, n'a pas évidemment son sens habituel; on
ne comprend pas comment l'autel et les fonts auraient pu être
superposés à l'arcuatum opus. Supra, dans le latin du moyen âge,
indique souvent la juxtaposition, le voisinage immédiat; c'est ainsi
qu'on doit l'interpréter dans notre texte; il exprime la contiguïté du
baptistère aux arcades de l'atrium[459].
[457] Tourneur, Description historique et archéologique de N.-D. de Reims (1889),
p. 94.
[458] P. 304 et 308. C'est la disposition qu'a conservée jusqu'à nos jours l'antique
basilique de Parenzo en Istrie.
[459] P. 308.
Ces arcades, Adalbéron les sacrifia pour augmenter de ce côté la
nef de la cathédrale et la rendre plus imposante,—«ampliori
receptaculo decorata.» Elles commençaient alors près de l'entrée de
l'église, et se développaient sur le quart environ de la longueur totale
de la basilique, c'est-à-dire en y comprenant le parvis. Telles
devaient être, en effet, à peu près les dimensions de cet atrium.
Ainsi s'explique le texte de Richer qui devient plus intelligible, si on
en fait l'application, non pas uniquement au vaisseau intérieur de
l'église, mais en même temps à la place close qui la précédait au
dehors.
De l'hypothèse que nous venons d'exposer à la suite du P. Jubaru, il
résulte que l'emplacement de l'ancien atrium de l'église
contemporaine de Clovis peut être représenté dans la cathédrale
actuelle par une surface carrée qui s'étendrait au milieu de la nef, à
partir de la dalle qui rappelle le martyre de saint Nicaise. C'est dans
cet espace restreint, mais en un point indéterminé, que s'élevait le
baptistère de Clovis. On connaît donc, à quelques mètres près, ce
lieu mémorable, auquel s'attachent de si grands souvenirs. Peut-on
espérer encore davantage et compter sur une découverte imprévue
ou sur d'heureuses fouilles qui nous montreraient les substructions
du vénérable édifice? Le sol de la cathédrale a été si remanié que
nous n'osons prédire cette joie aux archéologues de l'avenir.
L. Demaison.
TABLE DES NOMS PROPRES
 CITÉS DANS LE SECOND VOLUME
A
Abbon, abbé de Fleury-sur-Loire, 242.
Abra, fille de saint Hilaire de Poitiers, 75.
Abraham, patriarche, 163.
Abraham (saint), abbé à Clermont-Ferrand, 46.
Adalbéron, archevêque de Reims, 311, 312, 313.
Adelfius (saint), évêque de Poitiers, 161.
Adour (l'), fleuve, 82.
Ægidius (le comte), général romain, 29, 30, 31.
Aétius, général romain, 2, 5, 29.
Afrique, 52, 53, 221, v. aussi Vandales.
Agaune, v. Saint-Maurice-en-Valais.
Agde, 54, 83, 84, 102, 135, 138.
Agen, 83.
Ahun, v. Idunum.
Aimoin, chroniqueur, 216 n., 241, 242, 243, 283.
Aire, 83, 139.
Aisne (l'), rivière, 290, 291.
Aix-en-Provence, 42.
Ajax, émissaire arien, 41.
Alamans (les), 3, 6, 122, 198, 201, 235, 268, 272, 273, 278, 280,
281, 291.
Alaric II, roi des Visigoths, 19, 21, 25, 48, 50, 54, 61, 62, 64, 66, 71,
72, 73, 75, 76, 77, 78, 87 n., 99, 105, 113, 181, 192, 208, 258, 273,
279.
Albi, 53.
Albigeois (l'), 85.
Alboflède, sœur de Clovis, 202, 272.
Alboïn, roi des Lombards, 277.
Alcimus Ecdicius Avitus, v. Avitus (saint), évêque de Vienne.
Alcuin, 290, n.
Allemagne, 2, 120.
Alpaïde, fille de Louis le Pieux, 303.
Alpes (les), 36, 87, 98, 108, 109.
Alpes Cottiennes (les), 112.
Amalaric, roi des Visigoths, 77, 86, 113 (où on a imprimé par erreur
Amalaric II), 208.
Amales (la famille des), 109, 116.
Amand (saint), 181, n., 183, n.
Amboise, 58, 70.
Amiens, 140, 242.
Ammien Marcellin, 74.
Ampsivariens (les), 121.
Anastase II, pape, 275.
Anastase, empereur romain, 59, 61, 93, 94, 107, n.
Andély, 206.
Angers, 235, 236.
Angoulême, 82, 86, n., 88, 90, 95, 238.
Annales Burgondes, 236.
Annales d'Angers, 235.
Antée, 227.
Anthémius, empereur romain, 31.
Antonin le Pieux, empereur romain, 27.
Apollinaire, comte de Clermont-Ferrand, 78, 238.
Apt, 110.
Aptonius (saint), évêque d'Angoulême, 89.
Apulie (l'), 106.
Aquitaine (l'), 25, 30, 40, 47, 49, 51, 52, 53, 56, 58, 60, 69, 74, 82,
95, 97, 98, 116, 117, 128, 132, 136, 162, 166, 178, 179, 180, 182,
183, 191, 194, 205, 227, 235, 238, 255, 272.
Aquitaine (la première), 136.
Aquitaine (la deuxième), 28.
Aquitaine (les deux), 100, 132.
Aquitains (les), 47, 53, 96, 97.
Arcachon, 83, n.
Arcadius, patricien d'Auvergne, 209, 210, 211, n.
Aredius, conseiller du roi Gondebaud, 17, 19, 248.
Arius, hérésiarque, 41.
Arles, 6, 27, 29, 35, 36, 51, 54, 99, 100, 101, 102, 108, 109, 110,
111, 113, 114, 133, 135, 136, 137.
Arlésiens (les), 104, 112.
Arnoul (saint), de Tours, 156, 244.
Arras, 140, 176.
Arvandus, ancien préfet du prétoire, 6.
Arvernes (les), 35, 238, v. aussi Clermontois.
Ataulf, roi des Visigoths, 37.
Athanase (saint), évêque d'Alexandrie, 176.
Atlantique (l'), 84, 85.
Attigny, 291, n.
Attila, roi des Huns, 3, 29, 63.
Auch, 39, 43, 45, 83, 139, 164, 169, 180.
Aurélien, conseiller légendaire de Clovis, 156.
Ausone, 26, 27.
Austrasie (l'), 205.
Auvergne (l'), 6, 32, 34, 35, 39, 44, 46, 70, 85, 88, 205, 208.
Auxerre, 22, 24, 159, 206, 248, 249.
Auzance (l'), rivière, 72, 73.
Avignon, 7, 16, 17, 18, 25, 110, 113, 115, 236, 288.
Avitus (saint), évêque de Clermont-Ferrand, 234.
Avitus de Micy (saint), abbé, 211.
Avitus, empereur romain, 29, 32, 210.
Avitus (saint), évêque de Vienne en Dauphiné, 8, 9, 10, 11, 12, 19,
22, 48, 51, 67, 69, 135, 203, 273, 278, 280, 282.
Aymeri de Peyrac, abbé de Moissac, 181, 183, n., 203.
B
Baralle, abbaye, 176, 178, 179.
Barcelone, 87, n., 115, 208.
Basée (la porte), à Reims, 300, 301, 302, 303, 304, 305, 306.
Balther, hagiographe, 254.
Basile, évêque d'Aix-en-Provence, 42.
Basine, mère de Clovis, 200.
Bataves (l'Ile des), 121.
Bathilde, reine des Francs, 201, 206, 207.
Bayonne, 82.
Bazas, 83, 139.
Beaucaire, 99, 103, 104, 105.
Beauvais, 140.
Beda le Vénérable, chroniqueur anglo-saxon, 263.
Bégon, gendre de Louis le Pieux, 303, 304.
Bélisaire, général romain, 81, n.
Belgique (la première), 140.
Belgique (la seconde), 140, 173, 224.
Bénarn, 83, 84, 139, cf. Lescar.
Berry (le), 31.
Bertold, moine de Micy, 261.
Besançon, 7.
Bethléem (Sainte-Marie de), abbaye en Gâtinais, 186, 187.
Bièvre (la), rivière, 196.
Blodesindus, abbé de Blois, 246.
Blois, v. saint Dié.
Boèce, philosophe romain, 199.
Boivre (la), rivière, 74.
Bonaparte (la rue), à Paris, 192.
Bonneuil, 192.
Bordeaux, 27, 28, 36, 37, 45, 51, 52, 74, 82, 85, 88, 98, 116, 128,
136, 137, 141, 152.
Bordelais (le), 83, 95.
Boulogne-sur-Mer, 140.
Bourges, 88, 259, 260.
Bourgogne (la), 15, 186, 205, 207.
Bretons (les), 31.
Brice, évêque de Nantes, 170.
Brioude, 46.
Brotonne (la forêt de), 212.
Bructères (les), 121.
Brunehaut, reine des Francs, 209.
Buch, 139.
Bucianum, 174.
Buchonie (la), forêt, 123, 124, 125, 128.
Burgondes (les), 2, 3, 4, 5, 6, 7, 8, 10, 11, 12, 21, 22, 24, 25, 31, 32,
36, 41, 52, 56, 59, 60, 62, 69, 70, 80, 87, 98, 100, 101, 102, 105,
107, 109, 110, 111, 114, 115, 142, 221, 248.
Burgondie, 2, 5, 7, 12, 16, 22, 25, 42, 69, 85, 88, 207, 236.
Byzance, 5, 8, 58, 60, 61, 63, 106, 107, 109, 199.
Byzantins (les), 59, 107.
C
Camargue (Ile de la), 100.
Cambrai, 140, 176, 177, 178, 235.
Campine (la), 227.
Camulogène, roi des Parisii, 192.
Carcassonne, 81 n., 84 n., 99, 114.
Carthage, 52, 81 n.
Cassiodore, ministre de Théodoric-le-Grand, 272.
Caudebec, 212.
Cavaillon, 110.
Celsus, Gallo-Romain du vie siècle, 159.
Célestin, irlandais, 265, 266.
Céneret (le camp de), 73, 75, 76.
Cervon, abbaye, 248.
Césaire (saint), évêque d'Arles, 51, 52, 101, 102, 103, 104, 111, 114,
133, 135, 136, 137, 138, 244, 245.
Césarie (sainte), sœur de saint Césaire d'Arles, 102, 244.
Césars (les), 116.
Cévennes (les), 88.
Châlons-sur-Marne, 140.
Chalon-sur-Saône, 52.
Chararic, roi franc, 118.
Charbonnière (la forêt), 156, 271.
Charlemagne, 83, 130, 170, n., 179, 198, 203, 245, 271.
Charles le Chauve, roi de France, 246.
Charles le Simple, roi de France, 170, n.
Charles Martel, duc des Francs, 301.
Charroux, abbaye, 303, n.
Chartres, 246, 259, 283.
Château-Landon, 266, 267.
Châtellerault, 72.
Chattes (les), 121.
Chavigny, 172.
Chelles, abbaye, 192.
Childebert Ier, roi des Francs, 157, n., 192, 205, 208, 209, 210, 212.
Childéric Ier, 129, 157, 202, 221, 235, 242, 250.
Chilpéric, roi des Burgondes, 5, 13, 14.
Chlodéric, roi des Ripuaires, 69, 77, 124, 125, 126, 127, 128.
Choisy, 196.
Chroniques de Saint-Denis (les), 242.
Cité (la) de Paris, 209.
Clain (le), rivière, 72, 73, 74.
Claudius, prêtre gallo-romain, 159, 160.
Clermont-Ferrand, 6, 9, 31, 33, 35, 74, 86, 138, n., 234.
Clermontois (les), 238.
Clichy, 192.
Clodion, roi des Francs, 157, 226, 242.
Clodoald ou Cloud, prince mérovingien, 207, 211.
Clodomir, roi des Francs, 205, 207, 209, 210, 211.
Clotaire Ier, roi des Francs, 157, n., 170, n., 205, 209, 210, 212, 213,
256, 284.
Clotilde (sainte), reine des Francs, 13, 14, 25, 129, 164, 167, 180,
186, 194, 196, 201, 202, 205, 206, 209, 210, 211, 212, 213, 237,
238, 245, 257, 277, 279, 281, 291, 294, n., 306, 307, 308, 310.
Clotilde, fille de Clovis, 208.
Clotsinde, femme du roi Alboïn, 277, 278, 280.
Clovis, 1, 2, 10, 11, 14, 15, 17, 19, 20, 21, 22, 23, 24, 25, 26, 43, 49,
50, 53, 55, 56, 57, 58, 61, 62, 63, 65, 66, 67, 68, 69, 70, 71, 72, 73,
75, 77, 78, 81, 82, 83, 84, 85, 88, 89, 90, 92, 93, 94, 95, 97, 105,
115, 116, 118, 121, 122, 123, 124, 125, 126, 127, 128, 130, 131,
132, 133, 134, 135, 136, 138, n., 139, 151, 152, 153, 155, 156, 157,
159, 160, 162, 164, 166, 167, 170, 171, 173, 175, 176, 178, 179,
180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 191, 192, 194,
195, 196, 199, 200, 201, 202, 203, 204, 205, 206, 208, 212, 213,
214, 215, 216, 217, 218, 219, 220, 221, 224, 226, 229, 230, 233,
234, 235, 236, 237, 238, 239, 241, 242, 243, 244, 245, 246, 247,
249, 255, 256, 257, 258, 260, 261, 262, 265, 266, 267, 268, 269,
270, 271, 272, 273, 274, 275, 277, 278, 279, 280, 281, 282, 283,
284, 287, 289, 290, 291, 292, 293, 294, 297, 299, 305, 306, 307,
308, 309, 310, 311, 312, 314.
Clovis II, 170 n., 181 n., 242.
Cocy, 173.
Coffin (saint), 186.
Cologne, 43, 69, 77, 119, 120, 121, 122, 123, 124, 125, 126, 130,
140, 235.
Comminges, v. Saint-Bertrand de Comminges.
Conques (abbaye Sainte-Foi de), 179.
Constance Chlore, empereur romain, 192.
Constantin le Grand, empereur romain, 74, 100, 138, n., 143.
Constantinople, 59, 67, 106.
Continuatio Prosperi Havniensis, 282.
Corniche (la), 108.
Côte-d'or (la), 187.
Cottiennes (les Alpes), v. Alpes.
Couserans, 83, 139.
Cousin (le), rivière, 22, n., 248.
Crocus, évêque de Nîmes, 45.
Cure (la), rivière, 22, 24, 58, 248.
Cyprien, évêque de Toulon, 244.
Cyprien, évêque de Bordeaux, 136, 137, 141, 152.
D
Dani, v. Normands.
Dauphiné (le), 147.
Dax, 83, 139.
De Gestis regum Francorum, v. Aimoin.
Déols, 31.
Diane, déesse, 194.
Dié (saint), solitaire de Blois, 246.
Dijon, 7, 15, 16.
Dioclétien, empereur romain, 52, 247.
Dispargum, 227.
Dodilon, évêque de Cambrai, 177.
Dorat (le), abbaye, 183.
Durance (la), rivière, 36, 98, 108, 110, 112, 115.
E
Eauze, 45, 83, 139.
Ebbon, archevêque de Reims, 295, 311, 313.
Ecdicius, fils de l'empereur Avitus, 32, 33, 34, 35, 40, 78 n.
Ecdicius, v. Avitus (saint), de Vienne.
Edwin, roi de Northumbrie, 200.
Église franque (l'), 153, 223.
Eleuthère (saint), évêque de Tournai, 164, 246, 247.
Emma ou Emmia, prétendue fille de Clovis, 205 n.
Eonius, évêque d'Arles, 111.
Epaone, 135, 138 n.
Epineuil, 192.
Epiphane (saint), évêque de Pavie, 133.
Epitome, v. Frédégaire.