Download the full version of the ebook at ebookname.

com

Convex Functional Analysis 1st Edition Andrew J.

Kurdila

https://ebookname.com/product/convex-functional-
analysis-1st-edition-andrew-j-kurdila/

OR CLICK BUTTON

DOWNLOAD EBOOK

Download more ebook instantly today at https://ebookname.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Applied Functional Analysis 2nd Edition J. Tinsley Oden

https://ebookname.com/product/applied-functional-analysis-2nd-edition-
j-tinsley-oden/

ebookname.com

Convex analysis and optimization in Hadamard spaces

Miroslav Bacak

https://ebookname.com/product/convex-analysis-and-optimization-in-
hadamard-spaces-miroslav-bacak/

ebookname.com

Functional Analysis Vagn Lundsgaard Hansen

https://ebookname.com/product/functional-analysis-vagn-lundsgaard-
hansen/

ebookname.com

Contemporary Studies in Ethnography 1st Edition Alfonso

Ivarra

https://ebookname.com/product/contemporary-studies-in-ethnography-1st-
edition-alfonso-ivarra/

ebookname.com
Automotive Electricity Electric Drive 1st Edition Joseph
Beretta

https://ebookname.com/product/automotive-electricity-electric-
drive-1st-edition-joseph-beretta/

ebookname.com

Leadership Six Studies in World Strategy 1st Edition Henry

Kissinger

https://ebookname.com/product/leadership-six-studies-in-world-
strategy-1st-edition-henry-kissinger/

ebookname.com

Professional Microsoft SQL Server 2012 Administration 1.

Auflage Edition Adam Jorgensen

https://ebookname.com/product/professional-microsoft-sql-
server-2012-administration-1-auflage-edition-adam-jorgensen/

ebookname.com

Richard the Lionheart Jean Flori

https://ebookname.com/product/richard-the-lionheart-jean-flori/

ebookname.com

Past and Present National Identity and the British

Historical Film James Chapman

https://ebookname.com/product/past-and-present-national-identity-and-
the-british-historical-film-james-chapman/

ebookname.com
Foliations Geometry and Dynamics 1st Edition Pawel Walczak

https://ebookname.com/product/foliations-geometry-and-dynamics-1st-
edition-pawel-walczak/

ebookname.com
Andrew J. Kurdila
Michael Zabarankin

Convex
Functional Analysis

Birkhauser Verlag
Basel • Boston • Berlin
Authors:

Andrew J. Kurdila Michael Zabarankin

Department of Mechanical and Department of Mathematical Sciences
Aerospace Engineering Stevens Institute of Technology
University of Florida Castle Point on Hudson
Gainesville, FL 32611-6250 Hoboken, NJ 07030
USA USA
ajk@mae.ufl.edu mzabaran@stevens.edu

2000 Mathematics Subject Classification 46N10, 49J15, 49J20, 49J27,

49J40, 49J50, 49K15, 49K20, 49K27, 49N90, 65K10, 90C25, 93C20, 93C25

A CIP catalogue record for this book is available from the Library of
Congress, Washington D.C., USA

Bibliographic information published by Die Deutsche Bibliothek

Die Deutsche Bibliothek lists this publication in the Deutsche
Nationalbibliografie; detailed bibliographic data is available in the Internet
at <http://dnb.ddb.de>.

ISBN 3-7643-2198-9 Birkhäuser Verlag, Basel – Boston – Berlin

This work is subject to copyright. All rights are reserved, whether the whole
or part of the material is concerned, specifically the rights of translation,
reprinting, re-use of illustrations, recitation, broadcasting, reproduction on
microfilms or in other ways, and storage in data banks. For any kind of use
permission of the copyright owner must be obtained.

© 2005 Birkhäuser Verlag, P.O. Box 133, CH-4010 Basel, Switzerland

Part of Springer Science+Business Media
Printed on acid-free paper produced of chlorine-free pulp. TCF f
Printed in Germany
ISBN-10: 3-7643-2198-9
ISBN-13: 978-3-7643-2198-7

987654321
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

1 Classical Abstract Spaces in Functional Analysis

1.1 Introduction and Notation . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Topological Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.1 Convergence in Topological Spaces . . . . . . . . . . . . . . 13
1.2.2 Continuity of Functions on Topological Spaces . . . . . . . 15
1.2.3 Weak Topology . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.2.4 Compactness of Sets in Topological Spaces . . . . . . . . . 19
1.3 Metric Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.3.1 Convergence and Continuity in Metric Spaces . . . . . . . . 21
1.3.2 Closed and Dense Sets in Metric Spaces . . . . . . . . . . . 23
1.3.3 Complete Metric Spaces . . . . . . . . . . . . . . . . . . . . 23
1.3.4 The Baire Category Theorem . . . . . . . . . . . . . . . . . 25
1.3.5 Compactness of Sets in Metric Spaces . . . . . . . . . . . . 27
1.3.6 Equicontinuous Functions on Metric Spaces . . . . . . . . . 30
1.3.7 The Arzela-Ascoli Theorem . . . . . . . . . . . . . . . . . . 33
1.3.8 Hölder’s and Minkowski’s Inequalities . . . . . . . . . . . . 35
1.4 Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
1.5 Normed Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.5.1 Basic Definitions . . . . . . . . . . . . . . . . . . . . . . . . 45
1.5.2 Examples of Normed Vector Spaces . . . . . . . . . . . . . 46
1.6 Space of Lebesgue Measurable Functions . . . . . . . . . . . . . . . 52
1.6.1 Introduction to Measure Theory . . . . . . . . . . . . . . . 52
1.6.2 Lebesgue Integral . . . . . . . . . . . . . . . . . . . . . . . . 54
1.6.3 Measurable Functions . . . . . . . . . . . . . . . . . . . . . 57
1.7 Hilbert Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

2 Linear Functionals and Linear Operators

2.1 Fundamental Theorems of Analysis . . . . . . . . . . . . . . . . . . 65
2.1.1 Hahn-Banach Theorem . . . . . . . . . . . . . . . . . . . . 65
2.1.2 Uniform Boundedness Theorem . . . . . . . . . . . . . . . . 69
2.1.3 The Open Mapping Theorem . . . . . . . . . . . . . . . . . 71
2.2 Dual Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
vi Contents

2.3 The Weak Topology . . . . . . . . . . . . . . . . . . . . . . . . . . 79

2.4 The Weak∗ Topology . . . . . . . . . . . . . . . . . . . . . . . . . . 80
2.5 Signed Measures and Topology . . . . . . . . . . . . . . . . . . . . 88
2.6 Riesz’s Representation Theorem . . . . . . . . . . . . . . . . . . . . 91
2.6.1 Space of Lebesgue Measurable Functions . . . . . . . . . . . 91
2.6.2 Hilbert Spaces . . . . . . . . . . . . . . . . . . . . . . . . . 94
2.7 Closed Operators on Hilbert Spaces . . . . . . . . . . . . . . . . . 95
2.8 Adjoint Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2.9 Gelfand Triples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
2.10 Bilinear Mappings . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

3 Common Function Spaces in Applications

3.1 The Lp Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
3.2 Sobolev Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
3.2.1 Distributional Derivatives . . . . . . . . . . . . . . . . . . . 114
3.2.2 Sobolev Spaces, Integer Order . . . . . . . . . . . . . . . . . 117
3.2.3 Sobolev Spaces, Fractional Order . . . . . . . . . . . . . . . 118
3.2.4 Trace Theorems . . . . . . . . . . . . . . . . . . . . . . . . 122
3.2.5 The Poincaré Inequality . . . . . . . . . . . . . . . . . . . . 123
3.3 Banach Space Valued Functions . . . . . . . . . . . . . . . . . . . . 126
3.3.1 Bochner Integrals

. . . . . . . . . . . . . . . . . . . . . . . . 126
3.3.2 The Space Lp (0,
), X . . . . .
T . . . . . . . . . . . . . . . 131
p,q
3.3.3 The Space W (0, T ), X . . . . . . . . . . . . . . . . . . 133

4 Differential Calculus in Normed Vector Spaces

4.1 Differentiability of Functionals . . . . . . . . . . . . . . . . . . . . 137
4.1.1 Gateaux Differentiability . . . . . . . . . . . . . . . . . . . 137
4.1.2 Fréchet Differentiability . . . . . . . . . . . . . . . . . . . . 139
4.2 Classical Examples of Differentiable Operators . . . . . . . . . . . 143

5 Minimization of Functionals
5.1 The Weierstrass Theorem . . . . . . . . . . . . . . . . . . . . . . . 161
5.2 Elementary Calculus . . . . . . . . . . . . . . . . . . . . . . . . . . 163
5.3 Minimization of Differentiable Functionals . . . . . . . . . . . . . . 165
5.4 Equality Constrained Smooth Functionals . . . . . . . . . . . . . . 166
5.5 Fréchet Differentiable Implicit Functionals . . . . . . . . . . . . . . 171

6 Convex Functionals
6.1 Characterization of Convexity . . . . . . . . . . . . . . . . . . . . . 177
6.2 Gateaux Differentiable Convex Functionals . . . . . . . . . . . . . 180
6.3 Convex Programming in Rn . . . . . . . . . . . . . . . . . . . . . . 183
6.4 Ordered Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.4.1 Positive Cones, Negative Cones, and Orderings . . . . . . . 189
6.4.2 Orderings on Sobolev Spaces . . . . . . . . . . . . . . . . . 191
Contents vii

6.5 Convex Programming in Ordered Vector Spaces . . . . . . . . . . . 193

6.6 Gateaux Differentiable Functionals on Ordered Vector Spaces . . . 199

7 Lower Semicontinuous Functionals

7.1 Characterization of Lower Semicontinuity . . . . . . . . . . . . . . 205
7.2 Lower Semicontinuous Functionals and Convexity . . . . . . . . . . 208
7.2.1 Banach Theorem for Lower Semicontinuous Functionals . . 208
7.2.2 Gateaux Differentiability . . . . . . . . . . . . . . . . . . . 210
7.2.3 Lower Semicontinuity in Weak Topologies . . . . . . . . . . 210
7.3 The Generalized Weierstrass Theorem . . . . . . . . . . . . . . . . 212
7.3.1 Compactness in Weak Topologies . . . . . . . . . . . . . . . 213
7.3.2 Bounded Constraint Sets . . . . . . . . . . . . . . . . . . . 215
7.3.3 Unbounded Constraint Sets . . . . . . . . . . . . . . . . . . 215
7.3.4 Constraint Sets on Ordered Vector Spaces . . . . . . . . . . 217

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
List of Figures

1.1 Illustrations of notions of lim sup xn and lim inf xn . . . . . . . . . 4

n→∞ n→∞
1.2 Blurred Sierpinski Gasket, Various N . . . . . . . . . . . . . . . . . 7
1.3 Refinements of a Topological Space . . . . . . . . . . . . . . . . . . 8
1.4 Simple Function and Sequence of Fourier Approximations . . . . . 10
1.5 Nested Refinements . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.6 Path of Triangular Refinements . . . . . . . . . . . . . . . . . . . . 14
1.7 Convergent Sequence, Schematic Representation . . . . . . . . . . 22
1.8 Cauchy Sequence, Schematic Representation . . . . . . . . . . . . . 24
1.9 Embedding of Incomplete Metric Space X in Complete Metric Space
Y via Isometry i . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.10 Young’s Inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

2.1 Schematic Diagram of Dilation of Open Ball . . . . . . . . . . . . . 72

2.2 C(X), its Dual and Bidual . . . . . . . . . . . . . . . . . . . . . . . 91
2.3 Difference adjacent basis functions . . . . . . . . . . . . . . . . . . 91
2.4 Basis function for Theorem 2.6.1 . . . . . . . . . . . . . . . . . . . 92

3.1 L1 (Ω), its Dual and Bidual . . . . . . . . . . . . . . . . . . . . . . 113

3.2 φx Diffeomorphic from Ox to B1 (0) . . . . . . . . . . . . . . . . . . 120

6.1 Convex and Nonconvex Sets . . . . . . . . . . . . . . . . . . . . . . 175

6.2 Graph of a Function . . . . . . . . . . . . . . . . . . . . . . . . . . 178
6.3 Lower Sections of the Graph . . . . . . . . . . . . . . . . . . . . . . 178
6.4 Convex and Nonconvex Functionals . . . . . . . . . . . . . . . . . . 179
6.5 Epigraph of a Functional . . . . . . . . . . . . . . . . . . . . . . . . 179
6.6 Separation of Sets via a Hyperplane . . . . . . . . . . . . . . . . . 180
6.7 Examples of Cones in R2 . . . . . . . . . . . . . . . . . . . . . . . . 189
6.8 Open and Closed Cones . . . . . . . . . . . . . . . . . . . . . . . . 190
6.9 Ordering Associated with a Convex Cone . . . . . . . . . . . . . . 191

7.1 A function that is lower semicontinuous at x0 but not lower semi-

continuous at x1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Preface

Overview of Book
This book evolved over a period of years as the authors taught classes in varia-
tional calculus and applied functional analysis to graduate students in engineering
and mathematics. The book has likewise been influenced by the authors’ research
programs that have relied on the application of functional analytic principles to
problems in variational calculus, mechanics and control theory.
One of the most difficult tasks in preparing to utilize functional, convex, and
set-valued analysis in practical problems in engineering and physics is the intimi-
dating number of definitions, lemmas, theorems and propositions that constitute
the foundations of functional analysis. It cannot be overemphasized that functional
analysis can be a powerful tool for analyzing practical problems in mechanics and
physics. However, many academicians and researchers spend their lifetime study-
ing abstract mathematics. It is a demanding field that requires discipline and
devotion. It is a trite analogy that mathematics can be viewed as a pyramid of
knowledge, that builds layer upon layer as more mathematical structure is put in
place. The difficulty lies in the fact that an engineer or scientist typically would
like to start somewhere “above the base” of the pyramid. Engineers and scientists
are not as concerned, generally speaking, with the subtleties of deriving theorems
axiomatically. Rather, they are interested in gaining a working knowledge of the
applicability of the theory to their field of interest.
The content and structure of the book reflects the sometimes conflicting
requirements of researchers or students who have formal training in either engi-
neering or applied mathematics. Typically, before taking this course, those trained
within an engineering discipline might have a working knowledge of fundamental
topics in mechanics or control theory. Engineering students may be perfectly com-
fortable with the notion of the stress distribution in an elastic continuum, or the
velocity field in an incompressible flow. The formulation of the equations govern-
ing the static equilibrium of elastic bodies, or the structure of the Navier-Stokes
Equations for incompressible flow, are often familiar to them. This is usually not
the case for first year graduate students trained in applied mathematics. Rather,
these students will have some familiarity with real analysis or functional analy-
sis. The fundamental theorems of analysis including the Open Mapping Theorem,
the Hahn-Banach Theorem, and the Closed Graph Theorem will constitute the
foundations of their training in many cases.
xii Preface

Coupled with this essential disparity in the training to which graduate stu-
dents in these two disciplines are exposed, it is a fact that formulations and so-
lutions of modern problems in control and mechanics are couched in functional
analytic terms. This trend is pervasive.
Thus, the goal of the present text is admittedly ambitious. This text seeks
to synthesize topics from abstract analysis with enough recent problems in control
theory and mechanics to provide students from both disciplines with a working
knowledge of functional analysis.

Organization

This work consists of two volumes. The primary thrust of this series is a discussion
of how convex analysis, as a specific subtopic in functional analysis, has served to
unify approaches in numerous problems in mechanics and control theory. Every
attempt has been made to make the series self-contained.
The first book in this series is dedicated to the fundamentals of convex func-
tional analysis. It presents those aspects of functional analysis that are used in
various applications to mechanics and control theory. The purpose of the first vol-
ume is essentially two-fold. On one hand, we wish to provide a bare minimum of the
theory required to understand the principles of functional, convex and set-valued
analysis. We want this presentation to be accessible to those with little advanced
graduate mathematics, which makes it a formidable task indeed. For this reason,
there are numerous examples and diagrams to provide as intuitive an explanation
of the principles as possible. The interested reader is, of course, referred to the
numerous excellent texts that present a complete treatment of the theory. On the
other hand, we would like to provide a concise summary of definitions and theo-
rems, even for those with a background in graduate mathematics, so that the text
is relatively self-contained.
The second book in the series discusses the application of functional analytic
principles to contact problems in mechanics, shape optimization problems, con-
trol of distributed parameter systems, identification problems in mechanics and
control of incompressible flow. While this list of applications is impressive, it is
hardly exhaustive. The second volume also overviews recent problems that can
be addressed with extensions of convex analysis. These applications include the
homogenization of steady state heat conduction equations, approximation theory
in identification problems and nonconvex variational problems in mechanics.
The first volume is organized as follows. Chapter 1 begins with a brief over-
view of topological spaces, and quickly focuses on metric topologies in particular.
Two of the most fundamental theorems included in this chapter are the Arzela-
Ascoli Theorem and Baire Category Theorem. Next, a brief discussion of normed
vector spaces follows. Section 1.4 presents the foundations of measure and integra-
tion theory, while Section 1.7 introduces Hilbert Spaces.
Preface xiii

Chapter 2 includes a tutorial on bounded linear operator and weak topologies.

The chapter presents the Hahn-Banach Theorem, Uniform Boundedness Theorem,
Closed Graph Theorem, Riesz’s Representation Theorem and common construc-
tions such as Gelfand Triples.
Chapter 3 is particularly important to applications: it includes descriptions
of common abstract spaces encountered in practice. A special emphasis is given to
a discussion of Sobolev Spaces.
Chapter 4 discusses the most common notions of differentiability for func-
tionals: Gateaux and Fréchet differentiability. It also presents classical examples
of differentiable functionals.
Chapter 5 introduces Lagrange multiplier techniques to characterize extrema
of optimization problems. The chapter includes a discussion of certain specialized
techniques for studying Fréchet differentiable functionals and equality constrained
problems.
Chapter 6 represents the fundamentals of classical convex analysis in func-
tional analysis. Convex functionals and sets are defined. The relationship between
continuity, convexity and differentiability is outlined. The chapter discusses formal,
“engineering techniques” for solving certain optimal control problems. Multiplier
methods are derived for convex functionals on ordered vector spaces.
Chapter 7 introduces the important notion of lower semicontinuity and intro-
duces the class of lower semicontinuous functionals. The crucial interplay of semi-
continuity and compactness in weak topologies is discussed. The chapter presents
several versions of the Generalized Weierstrass Theorem formulated for lower semi-
continuous functionals.

Acknowledgements
The authors would like to thank Professor William Hager of the Department of
Mathematics, Professor Panos M. Pardalos of the Department of Industrial and
Systems Engineering and Professor R. Tyrrel Rockafellar of the Department of
Mathematics at the University of Washington for their insight, advice and com-
ments on portions of the manuscript. The authors would also like to thank the
numerous project and contract officers who have had the foresight to support vari-
ous research projects by the authors. Many of the examples discussed in this book
are related to or extracted from research carried out under their guidance. Their
support has resulted in some of the examples discussed in the text. In particular,
the authors would like to acknowledge the generous support of Dr. Kam Ng of
the Office of Naval Research, Dr. Walt Silva of NASA Langley Research Center,
Dr. Marty Brenner of the Dryden Flight Research Center, and Dr. Gary Anderson
of the Army Research Office. Dr. Clifford Rhoades, the director of Physical and
Mathematical Sciences of the Air Force Office of Scientific Research, has had a
profound influence on the development of this text. Both authors have been fortu-
nate in that they have prepared substantial tracts of the text while working under
xiv Preface

AFOSR sponsorship. Furthermore, the authors extends their warmest gratitude to

Dr. Pasquale Storza, Director of the University of Florida Graduate Engineering
and Research Center (UF-GERC). This book was assembled in its final form while
the authors were visiting scholars at UF-GERC.
A large portion of this text was typeset by Maryum Ahmed, Rob Eick, Kristin
Fitzpatrick, and Joel Steward, who have our thanks for their careful attention to
detail and patience with the authors during the editorial process.
Both authors express their deepest gratitude to their wives and families. The
preparation of this book was carried out with their sustained encouragement and
support, despite the burden it sometimes put on the families.

Andrew J. Kurdila and Michael Zabarankin

Gainesville, Florida
April 24, 2004
Chapter 1

Classical Abstract Spaces

in Functional Analysis

1.1 Introduction and Notation

Set Theory Notation
Throughout this text, sets are typically denoted by capital letters

A, B, C

while elements of sets are denoted by lower case letters

a, b, c.

The membership of an element a in a set A is written as

a∈A

while we write
a∈
/A
if the element a is not in the set A. A set A is sometimes described explicitly by
listing its elements
A = {a1 , a2 , . . . , an }.
More frequently, set builder notation is employed wherein we write
 
A = a ∈ X : D(a) .

That is, A is comprised of all elements a in some well-defined, universal set X such
that the description, or property, D(a) is true. For example, if we want to define
the set consisting of the positive integers we can write

N = {i ∈ R : i is an integer and i > 0}.

2 Chapter 1. Classical Abstract Spaces in Functional Analysis

As usual, we define the union of sets A and B as all the elements that are a member
of A or B:
A ∪ B = {c : c ∈ A or c ∈ B}.
The intersection of A and B is the collection of elements that are members of both
A and B
A ∩ B = {c : c ∈ A and c ∈ B}.
A set B is a subset of A if every element of B is also an element of A
B⊆A ⇐⇒ (x ∈ B ⇒ x ∈ A) .
If B is a subset of A, the complement of B in A, denoted B̃, consists of all the
elements of A that are not elements of B. We write
B̃ = {x ∈ A : x ∈
/ B}.
Sometimes, when we want to emphasize that we refer to the complement of B in
the particular set A, we write
A\B = {x ∈ A : x ∈
/ B}.
Two sets are equal if they are comprised of the same elements. That is, A = B if
and only if
A ⊆ B and B ⊆ A.
A set B is a proper subset of A if and only if
B ⊆ A and B = A.
On many occasions, we employ set theoretic operations on families of sets. Suppose
F is a collection of sets. Then we define

F
{x : x ∈ F for some F ∈ F }

F
{x : x ∈ F for all F ∈ F }.
Specific sets occur so frequently in this text that we have reserved designa-
tions for them.

N The set of positive integers

Z The set of integers
Q The set of rational numbers
R The set of real numbers
C The set of complex numbers

By a sequence in the set X, we mean a mapping either from N to X or from

Z to X. In this book, sequences will be denoted as
{xk }k∈N or {xm }m∈Z
depending on the desired index set.
1.1. Introduction and Notation 3

Real Variables
When considering sets consisting of real numbers, conventional notation is em-
ployed for the open, closed and half-open intervals

[a, b] = {x ∈ R : a ≤ x ≤ b}
(a, b) = {x ∈ R : a < x < b}
[a, b) = {x ∈ R : a ≤ x < b}.

Of particular importance throughout the text are the infimum, supremum, limit
superior and limit inferior of a set of real numbers. If a set of real numbers A ⊆ R
is bounded from above, there is a smallest real number ā ∈ R such that

x ≤ ā

for all x ∈ A. The number ā is the supremum of the set A ⊆ R. This relationship
is written

ā = sup{x : x ∈ A}
= sup x.
x∈A

The supremum of a set A ⊆ R is also sometimes referred to as the least upper

bound, or l.u.b., of A. Likewise, if the set A is bounded below, there is a greatest
real number a ∈ R such that
a≤x
for all x ∈ A. The number a is the infimum of A

a = inf{x : x ∈ A}
= inf x.
x∈A

The infimum of a set A ⊆ R is also sometimes referred to as the greatest lower

bound, or g.l.b., of A. Let {xn }n∈N ⊆ R be some sequence of real numbers. We
define the limit superior, or lim sup, of the sequence to be
 
lim sup xn = inf sup xn .
n→∞ k∈N n≥k

The limit inferior, or lim inf, of the sequence is defined to be

 
lim inf xn = sup inf xn .
n→∞ k∈N n≥k

The following example illustrates notions of lim sup xn and lim inf xn .
n→∞ n→∞
4 Chapter 1. Classical Abstract Spaces in Functional Analysis

Example 1.1.1. Let a sequence {xn }n∈N be defined as

1
xn = 1+ cos πn.
n

This sequence is depicted in Figure 1.1. We have

sup xn = 1 + k1 for k = 2j j∈N

n≥k

inf xn = − 1+ k 1
for k = 2j − 1 j ∈ N.
n≥k

1
xn xn

1  CosΠn
n
4

2 lim sup xn
1
n

n
1 2 3 4 5 6
lim inf xn
1
-2 n

-4

Figure 1.1: Illustrations of notions of lim sup xn and lim inf xn

n→∞ n→∞

By definition of lim sup xn and lim inf xn , we obtain

n→∞ n→∞

1
lim sup = inf 1+ =1
k→∞ j→∞ 2j
1
lim inf = sup − 1 + = −1.
k→∞ j→∞ 2j − 1

This example also makes clear that the lim sup, or lim inf, of a sequence of real
numbers are simply the largest, or smallest, accumulation points of the sequence,
respectively. Note that if a sequence {xn }n∈N converges, then

lim sup xn = lim inf xn .

n→∞ n→∞
1.2. Topological Spaces 5

For f : R → R the notation f (x) = O(x) means

f (x)
lim sup =c
x→0 x

for some constant c = 0. We write f (x) = o(x) whenever

f (x)
lim sup = 0.
x→0 x

A sequence of real numbers {xk }k∈N is said to converge to its limit x∗ if for
any  > 0 there exists n
such that for any k > n
, we have |xk − x∗ | < , or in
terms of “δ − ” notation

∀  > 0 ∃ n
∈ N : k > n
=⇒ |xk − x∗ | < .

A function f (x) is said to be continuous at point x0 if for any  > 0 there exists
δ > 0 such that for any x ∈ (x0 − δ, x0 + δ), we have |f (x) − f (x0 )| < , i.e.,

∀ > 0 ∃δ > 0 : x ∈ (x0 − δ, x0 + δ) =⇒ |f (x) − f (x0 )| < .

1.2 Topological Spaces

The previous section discussed notions of infimum, supremum, convergence and
continuity for sequence of real numbers and functions of real variables. But can
we generalize these notions for other abstract mathematical objects? A large part
of mathematical analysis has been devoted to giving a precise definition to the
notion of “convergence.” In conventional language the word “convergence” has
many connotations. We say that a hurricane is “converging on the coast,” or that
“negotiators are converging” in discussions. When asked, most graduate students
would have little difficulty in defining a convergent sequence of numbers. Consid-
erable difficulty arises in defining the convergence of more abstract mathematical
objects, such as those, for instance, that represent the states of physical systems.
For example, when can we say that a sequence of functions converges “as a whole”
to another function? To answer to this question we should specify what we mean
by “convergence of a function as a whole.” This intuitive notion is hardly a rig-
orous mathematical definition. We should also know whether a considered set of
functions is “rich enough” to contain the limit of the sequence. All these matters
lead us to determining a structure of abstract spaces under consideration.
For instance, the set of real numbers when plotted may intuitively be per-
ceived as a continuous, infinitely long line. But this graphical description does not
determine R completely. In fact, if we also plot the set of rational numbers Q,
then it also “looks” like a continuous, infinitely long line. However, from classical
√
analysis, we know that not√ every point on the real line belongs to Q. Indeed, 2
is such an example, i.e., 2 ∈ / Q. Moreover, if we consider the sequence of rational
Exploring the Variety of Random
Documents with Different Content
 It was a long time before the sullen answer came:

"Always was afraid of him."

"Afraid of me?"

"No fear," said Arnie, with a laugh. "Not much. But I always have been
feared of Bert Mestaer."

"What?" Otis gave a kind of jump into the air. "Mestaer? That? Mestaer
coming the county gentleman? And I never knew it! You little devil, why
didn't you tell me?"

"Thought you'd know him."

Otis stood still in the dusty lane. He went from red to pale, and almost
to purple again.

"If he thinks he's done with me, he makes a mistake, that's all," he
growled; and he snarled like a wild beast.

CHAPTER XXXV

CONFESSION
"Well, I have played and lost. But that is best.
Where was my right to win and keep such glory?
Now I will let the book of living rest.
Closed is my story."
—ALICE HERBERT.
 Hubert's motor traversed the distance from the sports to Ilbersdale at a
rate so far beyond the police limit as to make limits seem ridiculous.

On his way he encountered Mr. Cooper, pushing his bicycle up hill. The
seething excitement which had gripped him that afternoon had not yet
expended itself; and he pulled up.

"Good evening, Vicar. Just a word! I hope you have not been to the
Burmesters with any of those awful lies about your niece that Otis was
getting off his chest in that field?"

The vicar assumed his most stony aspect. His cold eye said eloquently,
"Beware!" Aloud his reply was: "I fear I do not understand you, Captain
Brooke."

"Sorry; it's my slang makes me difficult to follow. I drop into it when

I'm excited. You see, I happen to know all about Otis, and I've just been
enlightening the bystanders a bit. There isn't enough of him left to make a
War Office clerk! They've hissed him off the ground, General Ayres has
washed his hands of him, and I've given him twelve hours to choose
between a libel action and a written apology."

"It is a—surely—a somewhat extraordinary proceeding on your part to

talk of libel actions on behalf of a young lady who has her own relatives to
protect her," said the vicar, fastening, in the whirl of his mind, upon a
breach of conventionalities which he might legitimately resent.

"Her own relatives didn't seem to me to be doing much protecting this

afternoon," observed Bert drily. "However, Mrs. Cooper and your daughters
will be able to tell you all about it. They saw the brute knocked out of
time."

The vicar began to be anxious to get home.

"I fancy you are under a misapprehension," he said. "I have just been to
Ilbersdale to correct a misunderstanding. I own I was disturbed this
afternoon to find that Mr. and Mrs. Helston had allowed the engagement
between Mr. Burmester and my niece to take place, without informing the
bridegroom's relatives of the serious family disabilities of the bride. I was
anxious to assure Lady Burmester that, had the affair rested with me, I
should have been quite frank; but that I naturally imagined that Miss
Lutwyche's adopted parents had supplied all the facts."

"That was very thoughtful of you. But if you were, as you say, in
possession of the facts, how is it that you did not contradict the horrible
misstatements made by Otis up at the field?"

The vicar grew still more stony.

"I have never been made acquainted with the exact truth concerning my
niece's injuries at the time of her father's death."

"Injuries?" echoed Bert. "Injuries, indeed! But I've seen her righted. She
was the darling of the Dale before; she'll be its idol now! Did you hear them
cheering her?"

The vicar stood amazed. "Cheering my niece?"

Hubert laughed mischievously.

"And who may you be, to have the intimate knowledge which I lack
concerning this young lady?" inquired the vicar.

"Hubert Brooke, late Captain in Lacy's Brigade," laughed Bert, as he

drove away.

Mr. Cooper pursued his road, in much wrath and discomfiture. His
reception at Ilbersdale had affronted him, his encounter with Bert
bewildered him. He remounted the bicycle which he was pushing at the
time of his meeting, and rode home with what speed he might.

By the Vicarage gate two men were awaiting him—Otis and the
unattractive Boer boy. Evidently they meant to speak with him.

Otis approached with the easy confidence and winning smile that he
could assume at will. He begged pardon for troubling, but he had
unthinkingly mixed himself up in what looked to him like a local scandal of
formidable dimensions, and he had come to the vicar for advice. Mr.
Cooper's anger was not altogether proof against the insidious appeal. He
was used to being ignored and left out of things, and, to one whose own
idea of his own importance, both socially and parochially, was enormous,
the way this man approached him was a salve to a wound always more or
less smarting.

After a short parley, understanding that Otis had material facts to

communicate, he invited him in.

Mrs. Cooper and her daughters were at tea; and there was a flutter of
consternation among them for they had seen the exit of Otis from the
Fransdale sports. Mrs. Cooper became unspeakably coy, blushing like a
girl, and dismissing her brood, with their tea half done, on the flimsiest of
pretexts.

"This is not surprisin', Mrs. Cooper," said Otis sadly. "I was a stranger
up there, and nobody knew me. It was my word against that of a bad and
dangerous man, who is sailin' among you all under false colours. These
young ladies heard him givin' me the lie up there. In justice, I should like to
have them hear what I want to tell you now. You may have heard Mr.
Mayne or Miss Lutwyche talk of the man Mestaer?"

"Yes, yes," gasped Madeline, at the door. "She had one letter from him
after she got here—you remember, mother, the letter she would not show
you. She said he wanted to marry her!"

"She wouldn't show you his letter?" slowly said Otis, standing by the
table, and turning his hat round in his hand as if on the point of taking
leave. "Has she ever told you that he goes now under the name of Brooke?"

"What's that you say?" sharply asked the vicar.

"He's well disguised," replied Otis. "He bluffed me, I own it. But Arnie
here, he knew him from the first; didn't you, boy?"

Arnold looked at the three tall, full-blown girls, blushed admiringly and
assented.
 Surprise deprived them all of speech.

"Now, I'm told," said Otis, "that this fine Captain Brooke is buildin'
himself a house, and that Miss Lutwyche is his—architect." He gave a little
chuckle, "Excuse me: I really got to laff," he drawled humorously. "The
idea of him an' his architect is a bit too thick—eh?"

"Miss Lutwyche is duly qualified," began the vicar, in his stateliest

manner.

"Do I doubt it? No, sir! But I hear she has been stayin' down in the
shires with him pretty near all summer, gettin' this house ready while her
lover's in Russia. Now, I couldn't help just wonderin'—we really couldn't
help it, Arnie and me—whether young Burmester knows that Brooke's her
old lover."

"Why, do you think she knows?" cried Theo excitedly.

Otis bent on her the sliest, most waggish look, and slowly closed one
eye.

"Dear young ladies, you live in Arcadia," he said. "You remind me of

three hedge-roses; an' you're doubtless as simple as you're sweet. But your
cousin, Miss Lutwyche, she wasn't born yesterday, you know. She knows a
thing or two, you may take my word for that."

The vicar was silent, struggling with mortification. That day he had
broken through his lifelong rule to do nothing hurriedly. He had gone
straight from hearing Otis's revelations to be first with Lady Burmester. He
felt sure that what was said must ultimately come to her ears. He thought
his duty was plain.

But if he had only waited! If he had only gone to Millie, armed with this
fact! If he could have charged her with knowing who Bert was, and
concealing her knowledge, how differently things might have gone!

He looked at his wife, who seemed to be still blushing. She rose from
table.
 "As you say, Mr. Otis," said she, with archness which was unutterably
comic upon her middle-aged, substantial personality, "my dear girls are
very unsophisticated. They have been carefully brought up, as English girls
usually are. I will leave you to discuss this serious matter with Mr. Cooper,
and take them away. Come, my darlings."

* * * * * * * *

Meanwhile Bert drove straight to the Grange, inquired for Lance, and
found him alone in the smoking-room, sunk in profound gloom and a large
arm-chair.

"Burmester," he said abruptly, "I've come to talk to you—to tell you

something that will perhaps sever our friendship for ever. Confession is
good, they say; but I funk mine." He sank down in the opposite chair,
drawing out his cigar-case. "I funk it; make it as easy as you can, old man."

Lance was not smoking. He lifted a haggard young face from the depths
of his chair.

"Sorry," he said nervously, "but fact is, I'm feeling a bit off—
preoccupied. I must own I'm not in a sympathetic mood."

"It's about that—same thing. My confession touches the spot," said

Hubert.

"What are you talking about?"

"You are upset because you find scandal busy with the name of ... your
... the girl you love. And because you feel she hasn't been open with you.
You don't doubt her, but you feel there are things you should have known,
which she has kept back. Is that so?"

"That's precisely it," said Lance hurriedly. "I oughtn't to talk to you
about it—about her. But there must be some kind of understanding between
me and her if—if things are to go on. I feel a brute, to talk like this, but I am
all abroad, so to speak. We have had a very unpleasant scene here. Old
Cooper turned up, and said there were wild rumours flying about, on the
authority of those who claimed to have known her in Africa, to the effect
that she, Melicent, had got out of her bedroom window and gone with a
man called Mestaer, and that she had been in his house three or four weeks.
He said he came to Melicent for an authoritative contradiction. He wished
to be able to refute the story; thought he had a right to ask for the exact
facts."

He leaned forward, running his hands up through his hair.

Hubert sat very still. "And what did she say?"

"... Said it was true."

There was a pause. "Not as if she meant it?" asked Hubert tentatively.

"She was very angry. She looked splendid. She said that it was at her
uncle's own request that she had kept silence—that when she first came to
England she was anxious to tell him everything, but was forbidden to
mention the subject. If we wanted to know the truth we could write to the
Bishop: he knew. Then she got up and took her leave, and went off with the
Helstons. Of course I know this is a cock-and-bull story; but I feel ... I
ought to have been told."

"Yes," agreed his friend, "you ought to have been told."

"One thing I do wish," said Lance, clenching his fists, "that I had that
man Mestaer here to strangle."

"Well, if all our wishes could be as easily granted," said Hubert. "I'm
Mestaer."

Lance bounded from his seat, then sank back, as red as fire.

"Is this a time for your rotting?"

"No rotting here. I told you it would mean the breaking of our
friendship very likely. I am Hubert Mestaer. I took the name of Brooke
because it was English, and my mother's, and I wished to live in England
and be English. May I go on, or are you too angry to hear me?"
 Lance rose to his feet again. He stared blankly for a minute or two, then
his eyes suddenly blazed.

"You're Mestaer! Good God! Then you're the man that knows! You can
tell me ... what happened that night!"

"Yes; I can tell you: and I will. Mayne knows, she knows, I know.
Nobody else."

A shiver ran through Lancelot; he seemed on the brink of a hundred

questions; he choked them back.

"Speak, can't you?" he said.

Hubert spoke. He told his story from the beginning, making Melicent's
attitude towards himself throughout quite clear. He did not dwell on his own
feelings, but made it no secret that he had come to England solely in the
hope of being able to obtain her regard.

Lancelot listened to it all, as to information respecting some girl whom

he had never known.

"That she should have undergone all this, and never told me a word!"

"I can see where her difficulty came in," said Bert "Before she engaged
herself to you, she had guessed who I am. That altered everything. If you
can see what I mean, it turned the past into the present. She could not speak
to you of Mestaer without adding that he was here, in England, under
another name. That would have been giving me away—"

"Why couldn't she warn you that she should speak?"

"She never confessed to me that she knew. She tried to avoid intimacy."

"You ought to have told me yourself!" cried Lance.

"Well, until she engaged herself to you, it was certainly no concern of

yours," said Hubert bluntly. "Do you suppose that what you are feeling now
is anything like as bad as what I felt about you, when I heard you had
carried off the only thing that made life worth living to me?"

Lance paced the room restlessly.

"Is that still the same?" he cried. "Do you still care about her?"

"It's chronic," said Bert calmly. "There's only one woman in my world.
She might have Boer relations on every bush for aught I should care.
Nothing she could do, nothing anybody could say of her, would make any
difference to me."

"But—then—when she's my wife?" stammered Lance.

"When she's—your wife I shall never see her any more," said Bert
quietly. "It wouldn't be safe."

"Safe? No! But am I safe now?" cried the young man bitterly. "I don't
understand. What is the situation at this moment between you and her?"

Bert folded his arms tight, hunching himself together as if to keep

control over his temper.

"That's a question, surely, that you must ask her to answer," he said, in a
colourless voice.

Lance uttered an exclamation of rage.

"You ought to know without asking," went on Hubert presently. "Does

she love you? Surely you must know that, If she does ... I'm out of it, you
see."

Lancelot paused in his pacing. He leaned against the window-frame

staring out Hubert had touched the weak spot. He knew that he had
persuaded Millie into the engagement, had ever since continued to assure
her that she was happy, or that, if not, she certainly would be. He knew that,
were he sure of her love, distrust would be impossible to him. He was not
sure. He did distrust her. He was madly, wildly jealous of Hubert. Crossing
to where he sat, he seized his shoulder, shaking him violently.
 "When she promised to marry me, did she know who you are?"

"Yes."

"Then it's all right! It must be! She said she liked me better than
anybody else."

"If she said so, it was true."

"She's—she's not like most girls, you see. She's a cold nature—"

"Is she?"

Hubert closed his eyes, thinking of the lips that had clung to his, the
eyes that had looked into his, the hands that had trembled beneath his, as
they stood together in the chalky pit He got up suddenly: he had had about
as much as he could stand.

At the moment a footman entered, with a note on a salver.

"From her," said Lance, very white, as the man left the room.

"Breaking it off," said Hubert, relighting his cold cigar with a shaking
hand.

Lance read it

"Just so. She declines to give any kind of explanation of the statements
made by Mr. Cooper. She prefers to consider the engagement at an end." He
stood silent a moment, the note crumpled in his hand. "I'll go to her," he
said unsteadily. "I've simply got to have it out with her! When she hears that
I know—that you have told me"—he was half-way to the door. Then he
stopped, as if choked. "When I think that I have never known all this! When
I think that I have been shut out from her confidence, and that you—you—
have known all the time! When I think that I've been away in Russia and
you two, with this common memory between you, have been together! Day
after day! Over that confounded house-building! I feel that I have good
ground to consider myself hardly used."
 Hubert turned slowly round. He was so white that Lance considered him
attentively.

"Why have you told me now?" he cried. "Why?"

"Only because it couldn't be helped," returned Hubert, in a hard voice.

"And, but for this scoundrel turning up, she would have married me
without a word! Brooke, I can't stand it! No man could! She's right, it had
better be broken off."

"Steady on!" said Bert, getting out his words with difficulty. "Listen a
moment! She may be offering you your freedom because she believes you
desire it. She is—very proud. She may think this miserable tittle-tattle has
shaken your faith in her, and she offers you your way out. What you have to
discover, or so it seems to me, is the cost to herself. Does she want to be
free?" He took out his handkerchief and wiped his forehead. "Everyone
round here will say you have treated her badly if it's broken off now. That
doesn't matter if it isn't true. But make sure—make sure, for God's sake!"

Lance stared at him.

"You're a queer chap. You must want the engagement to be off—that

would give you your chance. Yet you send me to her!"

Bert shrugged his shoulders.

"Well," he said, with a half laugh, "I've discovered to-day that I've been
mistaken all these years. Ever since I was four-and-twenty I have believed
that most of everything on this earth I desired—her. Now I find there is
something that I desire more—her happiness. If you're the man to make her
happy, in God's name go and do it."
 CHAPTER XXXVI

WHAT CHANCED UPON THE MOORS

"What can I give thee back, O liberal
And princely giver, who hast brought the gold
And purple of thine heart, unstained, untold,
And laid them on the outside of the wall
For such as I to take or leave withal?"
—SONNETS FROM THE PORTUGUESE.

Lancelot had gone to Glen Royd.

Bert found that he could not stay indoors. He wandered out, through the
gardens, into the long carriage drive which ran upon the side of the ravine,
among the pines, with a sheer ascent on one hand, and a sheer dip on the
other, down to where the trout stream rushed over its noisy bed.

He paced along the road, dark beneath the over-arching trees, till he
came to the lodge and gate at the end of the wood. There was no sign of
Lance returning, so he walked on, and bending to the left, came, after a
climb, out upon the highroad to Fransdale. Nobody was about. He sat down
upon a grassy hummock, watching the pretty horned sheep cropping the
short grass about him in the fragrant evening. He was so still that the sheep
grazed nearer and nearer.

He was thinking that man is strangely in the grip of circumstance.

Since that night of parting at Clunbury, he had been through deep

waters. Carol, who had been so long excluded, was summoned at last. The
final knock-out had been too serious for the champion to bear alone.

He had laid all the circumstances before the Bishop, and told him that
he would follow his advice as to the course now to be pursued—whether he
should let things go on, or make one more effort to show Melicent what he
believed to be the truth, to induce her to break her engagement.
 In the light of what he now heard—namely, that Melicent had almost at
once recognised Bert—Mayne had little difficulty in falling in with Bert's
opinion as to her reason for engaging herself to Lance. His disapproval of
the said proceeding was so grave that he felt, and said, that he really
thought it would be better for Bert to have nothing more to do with her.

"Incredible selfishness and duplicity," he mercilessly called it at first, till

warned by Bert's rising anger that strictures upon Melicent would merely
have the effect of drying up the flow of confidence.

"I began the duplicity," said Bert doggedly. "I don't see what she could
do but follow suit."

"Does that excuse your further duplicity," came the answering thrust,
"in proceeding to make love to her while Lance was in Russia?"

"I'm hanged if I see what the d—" A sudden pause. "I quite fail to see
what other course I could have taken. They were to be married when he
came back. I had got to show her before then that the thing couldn't be
done. And I succeeded, within a hair's-breadth. If it hadn't been for my d—!
Beg pardon! if it hadn't been for my—unfortunate temper, I should have
succeeded."

"From what you tell me," the Bishop opined, "it really seems as if she
does like you best, but as if her pride would not allow her to give way. The
question is—Has your violence destroyed your chance finally? I think you
ought to find out."

"You do?—you do? You think I might have another try?"

"Well, are you to be trusted to keep yourself in check? You know of old
that when you lose your temper you have no chance at all with Millie,
because she never loses hers."

Hubert grinned. "She did the other night, though."

"Do you think she said things that she will be ashamed of when she
thinks them over?"
 "Yes, I do."

"Then I think that should give you a hold of a new kind over her. If you
can only manage to put her in the wrong, old man, and be magnanimous
and forgive her—see?"

Bert's admiration was open and glowing.

"You have a genius! You ought to marry, Bishop—you ought really."

"My advice is," finally said the mentor, "that you go at once to
Fransdale, and see how the land lies. See what frame of mind Millie is in. If
she is scornful and gay, and wrapped up in Lance and her marriage, your
course will be more difficult. If she shows you, by word or look, that she
thinks she behaved ill, or wounded you, or desires your good opinion, then
to my thinking you have a chance that you ought to take. You have now
been disciplined by failure; you should have learnt something; and Melicent
also must be wiser. For if she has any feeling at all—which, as you know, I
always took leave to doubt—she must have suffered keenly during these
last few months."

It was with the ring of this advice in his ears that Bert had hurled
himself and his motor through England, and arrived at Fransdale, where in
his rage he had vowed never to set foot more.

At the first sight of Melicent he believed that he had done right to come.

And behold, an hour later, all things were changed; the chance
appearance of Otis had, as it were, altered the entire situation. Bert was no
longer suppliant, but defender of Melicent's name against all comers—even
against the man she was to marry.

It seemed to him that, whether she ever came to care for himself or no,
she must break with Lance; for, knowing her as he did, he was sure that
now she must feel that, if the engagement went on, full confession of all
that had passed was imperative; and that seemed impossible.
 Suspense grew and mounted in him till he felt desperate; yet still he sat
there, with a kind of charmed stillness, while the quiet-coloured end of
evening slowly merged in twilight.

It was growing dark when at last he saw figures moving along the path
that led up on the other side of the ridge from Glen Royd—two figures,
indistinct at first in the dusk, then clearer. It was Lancelot and Melicent
walking together. Bert felt dizzy.

Then he had lost! They were reconciled! They moved slowly along, and
he saw that Lance was pushing her bicycle.

He rose and obliterated himself hastily behind a craggy boulder.

They both turned into the road leading to the carriage drive of the
Grange, which was also a short cut to the lower parts of Fransdale. They
passed in complete silence, and he watched them along the white track of
road until they were lost in the shadows of the wood.

A light glimmered out in the lodge window. It was the only sign of
human life within his ken.

Lance must be bringing her back to dine, and intending to cycle home
with her afterwards. He lit a match and looked at his watch. A quarter to
eight. He could not meet them. He realised that he must have time.

He had done what he believed to be right—told the truth, as far as he

thought it permissible, to wipe away from his conscience the stain of his
treachery to Lance. And the result was that they were reconciled, and he
was left in outer darkness. He knew that he had not expected it.

His misery was too great for him to reflect at present. He could not tell
what he should do. He plunged down the hill-side with no thought at first
but to walk, to get away, to move fast, to fight down some overwhelming
bitterness of darkness which was clutching him.

At the bottom of the valley by the mill he turned and hastened up stream
by the fishing-path among the thick trees. Careless of his direction, he
walked on until he was at the head of Ilbersdale, and had emerged from the
woods upon the open, broken moor that lay at the feet of Fransdale. There
he lifted up his eyes, and saw far above him, perched upon the very verge of
the precipice, the lights of Ilberston Church and Vicarage.

He thought of Mr. Hall. It occurred to him to wonder whether this man,

whose personality had impressed him, as it impressed most people, would
have help or counsel for him. Anyhow, to breast that hill was something to
attempt—something that chimed in with his mood just then.

It was choir practice night, and the church door being open, the sound of
the sweet Cleveshire voices floated out over the uplands, and made
articulate the beauty of the night.

Hubert pushed his way hurriedly, yet not fast, because heedlessly, over
the broken ground, with which he was not familiar. There were short cuts,
but in the dusk he did not find them. Several times he was brought up short
by hollow ravines or boggy ground, and it was long before he struck the
road that leads up to the steep verge.

He had been walking for nearly two hours when at last he found himself
at the top. The full moon was up, and was flooding the moors with silver.
The prospect was grand. On the horizon line the Three Howes stood up
black against the radiance—the prehistoric burial-place of forgotten chiefs.
At his back the white crosses that mark the resting-places of the Dalesmen
glimmered among the neatly shorn grass of the churchyard.

He sat upon the low wall, gazing out over the silent waste. The church
was in darkness now, and closed; the village beyond showed but few lights.
The lamp over the Vicarage door beamed steadily upon the night, and
showed a lady's bicycle leaning against the Vicarage garden wall. Surely the
little brown basket on the handle-bars was familiar to him? Surely he had
seen that same machine leaning against the trees of the plantation at Lone
Ash?

He sprang from his seat and went up close. It was Melicent's bicycle.
Then she herself was within! She had not gone to dine with the Burmesters;
she had come straight up here to Mr. Hall. A wave of excitement passed
over Hubert. Should he go in, and let the priest hear both sides of the
question? What had happened—what had passed between her and Lance? If
he had not played the coward, and run away, he would have known by now.
As he hesitated, the Vicarage door opened. He saw Mr. Hall stand in the
light, with the girl beside him. For the second time that night he drew back
and hid himself.

"Don't ride the steep bit to-night," he heard Mr. Hall say, as he lit her
lamp.

"I know every inch of the way; it's really quite safe," was the
characteristic response.

"I shall feel more comfortable if you promise me. It is late for you to be
returning alone, but I cannot come with you; I must go on to poor old
Martha Hirst."

Millie's little laugh sounded sad. "You needn't be nervous about me on

these moors at night."

"I don't think I need, or I would not let you go; but the moon is glorious.
Good-night, and God bless you!"

She mounted, and rode swiftly away, past the church, along the little
level bit of road that came before the steep dip over the mountain-side. The
brief dialogue had decided Bert. Mr. Hall was not at leisure, and Millie was
riding home late, alone. His place was to follow her. He had ascertained that
there was a footpath which was far shorter than the windings she must go
down with her cycle. If she were going to walk the steep bit, he thought he
could overtake her when she dismounted to walk up the next ascent.

At the lower end of the steep hill, if you followed the road, you came to
a stone bridge; and here the Ilba flowed more silently, and deep pools
harboured many a fat trout. Trees arched over the road, growing by the
water-side; and under them were inky shadows.

Melicent's lamp gleamed brightly, but not bright enough to show the
wire fastened across the road. She was riding fast, with the impetus of the
long hill just negotiated, and she checked herself with difficulty as the
figure of a man detached itself from the shadow, waving his arms and
crying, "Stop! Danger! You'll fall!"

"What is the matter?" she cried, putting on her brakes and just managing
to alight "You!"—she stopped short, recognising Amurrica.

"We were gettin' ready for someone else," said Amurrica drily. "This is
an unexpected pleasure. Are you ridin' alone?"

"You see that I am."

"Our friend's gone home another way then, seemin'ly. But as you're
here, let's make the most of it. Give us a kiss, little Millie."

"Don't be a cad, Amurrica!" said Millie, with a most unlooked-for

gentleness. "I don't know why you stopped me, but I am glad you did, for
there are things I want to say to you. Is my brother here?"

"Yes," said Arnie, slouching out from the gloom.

"Amurrica," said Millie earnestly, "first of everything, I want to beg

your pardon. You were very cruel to me in the old days, but I have been
shown to-day that it was my fault. I was hard and insolent. If I had been a
different kind of girl, perhaps you'd not have wanted to injure me?"

Amurrica stood staring. Was this Millie? "What yer givin' us?" he
growled.

"I want to say I am sorry," said the girl steadily. "I was hard and insolent
to you again to-day. I provoked you to try and do me harm. But I—didn't
know you had Arnie with you. I—I remember Arnie when he was a dear
little curly-headed baby. I never was good to him. I was always—
disagreeable. Arnie, I—am—so—sorry! I want to say—forgive me!"

Her voice broke. She turned her head away and drew out a
handkerchief. Amurrica was stricken dumb. That Millie could humble
herself—that Millie could cry—these incidents had seemed to him utterly
out of the range of the things that happen. He had nothing to say. Arnie
giggled awkwardly.

"Amurrica," said Millie earnestly, laying a hand on his sleeve, "you

would have been a better man if you had known better women. I am one
that helped to make you worse, because I never appealed to the good in
you. There was only one of us who did the right thing all through; and that
was Bert. He saved me then, and to-day he has saved me again. He has
done more for me this day than I could ever tell anybody. Oh, Amurrica, we
ought to be so ashamed of ourselves—you and I!"

Amurrica, during this remarkable interview, had been like one bereft of
his usual faculties.

"Well, I'm d—d!" he said at last. "What kind of palaver's this? Mestaer's
playin' his own hand, same as I am—him an' his bloomin' millions!
Thought I didn't know him! Thought he was safe, did he? Bless his kind
heart, he'll find out that I'm goin' to get even with him—if not one way, then
another!"

A thought went like lightning through Millie's brain.

"Are you waiting here for him?"

They did not answer.

"What makes you think he will come this way?"

"He went up there," said Arnie.

"Up there? Up that hill? He's not there now; I've just come from there."

"Hist!" said Otis.

They all heard a footstep, clear in the night stillness, swinging down the
hill at a steady run.

"If it is he, Amurrica, now is your time to make it up," urged Millie, and
her heart began to beat faster, and sweet, wild thoughts surged up within her
at the thought that she was hearkening her lover's approaching feet.

"I'll make it up, no fear!" was the muttered reply, as Otis, who was
standing behind her, gripped her firmly by both elbows, pinioning her in his
strong hold, and backing into the deepest shadow on the farther side of the
bridge, under the trees. "Hold her!" he gasped to Arnie; "hold tight, we've
only a minute!"

The inky darkness, rendered blacker by contrast with the white wash of
moonlight on the road in front, held the struggling group invisible. Had
Millie had an inkling of her captor's plan she would have screamed, but
intent upon her peace-making desires, she still wished to try gentle
methods. Before she realised his intentions, Otis had rammed a
handkerchief forcibly against her mouth, and swiftly wound the feather boa
she wore round and round her head, forming a most excellent impromptu
gag. He was reckless now, and cared only for his revenge, consequences
had faded out of sight. Millie, sensible in a flash of her own helplessness
and Bert's danger, fought with all her strength.

The light, firm steps came on fast. They were round the corner. Hubert
hastened in the moon's full radiance to the darkness where the trap lay for
him. Just before he reached the fatal spot, a sound came to the trained ear of
the scout—a muffled, indeterminate sound, which was not running water,
nor the sound of feet upon a hard road.

Full in the light he stood, a brave target for a bullet; and even as he
paused, before he had drawn a breath, there was the report of a revolver, a
cry of some kind, a sound of scuffling, a splashing as of someone wading in
water; and silence.

He stood bewildered. The idea that somebody had tried to shoot him
never suggested itself. He thought it must be poachers, though the report
did not sound like a rifle shot, and there were no woods quite near. He at
once started to run on and see what had happened; and at once tripped and
fell, caught by the unseen wire.

Having fallen with some impetus, he came to the ground heavily; and
regained his feet with quite a new impression of some danger imminent,