(Download PDF) Techniques of Functional Analysis For Differential and Integral Equations 1St Edition Paul Sacks Full Chapter PDF

Techniques of Functional Analysis for
Differential and Integral Equations 1st

Edition Paul Sacks
Visit to download the full and correct content document:
https://ebookmass.com/product/techniques-of-functional-analysis-for-differential-and-i
ntegral-equations-1st-edition-paul-sacks/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Advanced Differential Equations 1st Edition Youssef

Raffoul
https://ebookmass.com/product/advanced-differential-
equations-1st-edition-youssef-raffoul/
Schaum's Outline of Differential Equations, Fifth

Edition Bronson
https://ebookmass.com/product/schaums-outline-of-differential-
equations-fifth-edition-bronson/
Transforms and Partial Differential Equations T.

Veerarajan
https://ebookmass.com/product/transforms-and-partial-
differential-equations-t-veerarajan/
Proper orthogonal decomposition methods for partial

differential equations Chen
https://ebookmass.com/product/proper-orthogonal-decomposition-
methods-for-partial-differential-equations-chen/
Boundary value problems for systems of differential,
difference and fractional equations : positive
solutions 1st Edition Johnny Henderson
https://ebookmass.com/product/boundary-value-problems-for-
systems-of-differential-difference-and-fractional-equations-
positive-solutions-1st-edition-johnny-henderson/
Mechanics and Physics of Structured Media : Asymptotic

and Integral Equations Methods of Leonid Filshtinsky
1st Edition Igor Andrianov
https://ebookmass.com/product/mechanics-and-physics-of-
structured-media-asymptotic-and-integral-equations-methods-of-
leonid-filshtinsky-1st-edition-igor-andrianov/
Differential Equations and Boundary Value Problems:

Computing and Modeling (Edwards, Penney & Calvis,
Differential Equations: Computing and Modeling Series)
5th Edition, (Ebook PDF)
https://ebookmass.com/product/differential-equations-and-
boundary-value-problems-computing-and-modeling-edwards-penney-
calvis-differential-equations-computing-and-modeling-series-5th-
edition-ebook-pdf/
Integral Equations for Real-Life Multiscale

Electromagnetic Problems (Electromagnetic Waves)
Francesca Vipiana
https://ebookmass.com/product/integral-equations-for-real-life-
multiscale-electromagnetic-problems-electromagnetic-waves-
francesca-vipiana/
Linear Algebra and Partial Differential Equations T

Veerarajan
https://ebookmass.com/product/linear-algebra-and-partial-
differential-equations-t-veerarajan/
Techniques of Functional
Analysis for Differential and
Integral Equations
Mathematics in Science and
Engineering
Techniques of
Functional Analysis for
Differential and
Integral Equations
Paul Sacks
Department of Mathematics, Iowa State University,
Ames, IA, United States
Series Editor
Goong Chen
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1800, San Diego, CA 92101-4495, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
© 2017 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, further information about the Publisher’s permissions policies and our arrangements
with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency,
can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described herein. In
using such information or methods they should be mindful of their own safety and the safety of
others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of any methods, products,
instructions, or ideas contained in the material herein.
Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library
ISBN: 978-0-12-811426-1
For information on all Academic Press publications

visit our website at https://www.elsevier.com/books-and-journals
Publisher: Candice Janco

Acquisition Editor: Graham Nisbet
Editorial Project Manager: Susan Ikeda
Production Project Manager: Paul Prasad Chandramohan
Cover Designer: Alan Studholme
Typeset by SPi Global, India

Contents
Preface ix
1. Some Basic Discussion of Differential and Integral

Equations 1
1.1 Ordinary Differential Equations
1 .1 .1 Initial Value Problems 2
1.1.2 Boundary Value Problems 4
1.1 .3 Some Exactly Solvable Cases 4
1.2 Integral Equations 5
1.3 Partial Differential Equations 8
1 .3.1 First Order PDEs and the Method of Characteristics 9
1.3.2 Second Order Problems in ]R2 12
1.3.3 Further Discussion of Model Problems 15
1.3.4 Standard Problems and Side Conditions 20
1.4 Well-Posed and III-Posed Problems 22
1.5 Exercises 23
2. Vector Spaces 29
2.1 Axioms of a Vector Space 29
2.2 linear Independence and Bases 31
2.3 linear Transformations of a Vector Space 32
2.4 Exercises 33
3. Metric Spaces 35
3.1 Axioms of a Metric Space 35
3.2 Topological Concepts 37
3.3 Functions on Metric Spaces and Continuity 40
3.4 Compactness and Optimization 42
3.5 Contraction Mapping Theorem 45
3.6 Exercises 48
vi Contents
4. Banach Spaces 51
4.1 Axioms of a Normed linear Space 51
4.2 Infinite Series 53
4.3 linear Operators and Functionals 54
4.4 Contraction Mappings in a Banach Space 56
4.5 Exercises 57
5. Hilbert Spaces 59
5.1 Axioms of an Inner Product Space 59

5.2 Norm in a Hilbert Space 60
5.3 Orthogonality 62
5.4 P roj ec ti o n s 63
5.5 Gram-Schmidt Method 65
5.6 Bessel's Inequality and Infinite Orthogonal Sequences 66
5.7 Characterization of a Basis of a Hilbert Space 67
5.8 Isomorphisms of a Hilbert Space 69
5.9 Exercises 71
6. Distribution Spaces 75
6.1 The Space of Test Functions 76
6.2 T he Space of Distributions 76
6.3 Algebra and Calculus With Distributi ons 80
6.3.1 Multiplication of Distributions 80
6.3.2 Convergence of Distributions 81
6.3.3 Derivative of a Distribution 83
6.4 Convolution and Distributions 88
6.5 Exercises 92
7. Fourier Analysis 95
7.1 Fourier Series in One Space Dimension 95
7.2 Alternative Forms of Fourier Series 100
7.3 More Abou t C onver gen ce of Fourier Series 101
7.4 The Fourier Transform on ]RN 104
7.5 Further Properties of the Fourier Transform 107
7.6 Fourier Series of Distributions 111
7.7 Fourier Transforms of Distributions 114
7.8 Exercises 119
8. Distributions and Differential Equations 125
8.1 Weak Derivatives and Sobolev Spaces 125

8.2 Differential Equations in '0' 127
8.3 Fundamental Solutions 130
Contents vii
8.4 Fundamental Solutions and the Fourier Transform 132

8.5 Fundamental Solutions for Some Important PDEs 136
8.6 Exercises 139
9. Linear Operators 141
9.1 linear Mappings Between Banach Spaces 141

9.2 Examples of linear Operators 142
9.3 linear Operator Equations 148
9.4 The Adjoint Operator 149
9.5 Examples of Adjoints 151
9.6 Conditions for Solvability of linear Operator Equations 153
9.7 Fredholm Operators and the Fredholm Alternative 154
9.8 Convergence of Operators 155
9.9 Exercises 156
10. Unbounded Operators 159

10.1 General Aspects of Unbounded linear Operators 159
10.2 The Adjoint of an Unbounded linear Operator 162
10.3 Extensions of Symmetric Operators 166
10.4 Exercises 168
11. Spectrum of an Operator 171

11.1 Resolvent and Spectrum of a linear Operator 171
11.2 Examples of Operators and Their Spectra 174
11.3 Properties of Spectra 177
11.4 Exercises 180
12. Compact Operators 1 83

12.1 Compact Operators 183
12.2 Riesz-Schauder Theory 188
12.3 The Case of Self-Adjoint Compact Operators 192
12.4 Some Properties of Eigenvalues 198
12.5 Singular Value Decomposition and Normal Operators 200
12.6 Exercises 201
13. Spectra and Green's Functions for Differential

Operators 205
13.1 Green's Functions for Second-Order ODEs 205

13.2 Adjoint Problems 209
13.3 Sturm-liouville Theory 212
13.4 laplacian With Homogeneous Dirichlet Boundary Conditions 215
13.5 Exercises 221
viii Contents
14. Further Study of Integral Equations 227

14.1 Singular Integral Operators 227
14.2 Layer Potentials 230
14.3 Convolution Equations 235
14.4 Wiener-Hopf Technique 236
14.5 Exercises 239
15. Variational Methods 243
15.1 The Dirichlet Quotient 243

15.2 Eigenvalue Approximation 247
15.3 The Euler-Lagrange Equation 249
15.4 Variational Methods for Elliptic Boundary Value Problems 250
15.5 Other Problems in the Calculus of Variations 254
15.6 The Existence of Minimizers 258
15.7 Calculus in Banach Spaces 259
15.8 Exercises 263
16. Weak Solutions of Partial Differential Equations 269

16.1 Lax Milgram Theorem
- 269
16.2 More Function Spaces 275
16.3 Galerkin's Method 280
16.4 Introduction to Linear Semigroup Theory 283
16.5 Exercises 291
Appendix 295
A.1 Lebesgue Measure and the Lebesgue Integral 295
A.2 Inequalities 298
A.3 Integration by Parts 301
A.4 Spherical Coordinates in JRN 302
Bibliography 305
Index 307
Preface
The purpose of this book is to provide a textbook option for a course in

modern methods of applied mathematics suitable for first year graduate students
in Mathematics and Applied Mathematics, as well as for students in other
science and engineering disciplines with good undergraduate-level mathematics
preparation. While the term “applied mathematics” has a very broad meaning,
the scope of this textbook is much more limited, namely to present techniques
of mathematical analysis which have been found to be particularly useful
in understanding certain kinds of mathematical problems which commonly
occur in the scientific and technological disciplines, especially physics and
engineering. These methods, which are often regarded as belonging to the realm
of functional analysis, have been motivated most specifically in connection with
the study of ordinary differential equations, partial differential equations, and
integral equations. The mathematical modeling of physical phenomena typically
involves one or more of these types of equations, and insight into the physical
phenomenon itself may result from a deep understanding of the underlying
mathematical properties which the models possess. All concepts and techniques
discussed in this book are ultimately of interest because of their relevance for
the study of these three general types of problems. Of course there is a great deal
of beautiful mathematics, which has grown out of these ideas, and so intrinsic
mathematical motivation cannot be denied or ignored.
The background the student will obtain by studying this material is sufficient
preparation for more advanced graduate-level courses in differential equations
and numerical analysis, as well as more specialized topics such as fluid
dynamics. The presentation avoids overly technical material from measure
and function theory, while at the same time maintaining a high standard of
mathematical rigor. This is accomplished by including careful presentation of
the statements of such results whose proofs are beyond the scope of the text,
and precise references to other resources where complete proofs can be found.
As the book is meant as textbook for a course, which is typically the basis for
a written qualifying examination for PhD students, a great many examples and
exercises are given.
The topics presented in this book have served as the basis for a two-semester
course for first year graduate students at Iowa State University. A normal
division is that Chapters 1–8 are covered in the first semester, and Chapters 9–16
in the second. A partial exception is Chapter 14, all of whose topics are optional
in relation to later material, and so may be easily omitted.
ix
x Preface
I would like to thank past and present friends and colleagues for many
helpful discussions and suggestions about the teaching of Applied Mathematics
at this level and points of detail in earlier drafts of this book: Tuncay Aktosun,
Jim Evans, Scott Hansen, Fritz Keinert, Howard Levine, Gary Lieberman,
Hailiang Liu, Tasos Matzavinos, Hien Nguyen, John Schotland, Mike Smiley,
Pablo Stinga and Jue Yan. All mistakes and shortcomings are of course my own
doing.
Paul Sacks
October 28, 2016
Chapter 1
Some Basic Discussion of

Differential and Integral
Equations
In this chapter we will discuss “standard problems” in the theory of ordinary

differential equations (ODEs), integral equations, and partial differential equa-
tions (PDEs). The techniques developed in this book are all meant to have
some relevance for one or more of these kinds of problems, so it seems best
to start with some awareness of exactly what the problems are. In each case
there are some relatively elementary methods, which the reader may well have
seen before, or which rely only on simple calculus considerations, which we
will review. At the same time we establish terminology and notations, and begin
to get some sense of the ways in which problems are classified.
1.1 ORDINARY DIFFERENTIAL EQUATIONS

An nth order ordinary differential equation for an unknown function u = u(t)
on an interval (a, b) ⊂ R is any equation of the form
F(t, u, u , u , . . . , u(n) ) = 0 (1.1.1)
where we use the usual notations u , u , . . .
for derivatives of order 1, 2, . . .
and also u(n) for derivative of order n. Unless otherwise stated, we will assume
that the ODE can be solved for the highest derivative, that is, written in
the form
u(n) = f (t, u, u , . . . , u(n−1) ) (1.1.2)
For the purpose of this discussion, a solution of either equation will mean a real
valued function on (a, b) possessing continuous derivatives up through order
n, and for which the equation is satisfied at every point of (a, b). While it is
easy to write down ODEs in the form (1.1.1) without any solutions (e.g., (u )2 +
u2 + 1 = 0), we will see that ODEs of the type (1.1.2) essentially always have
solutions, subject to some very minimal assumptions on f .
Techniques of Functional Analysis for Differential and Integral Equations

http://dx.doi.org/10.1016/B978-0-12-811426-1.00001-5
© 2017 Elsevier Inc. All rights reserved. 1
2 Techniques of Functional Analysis for Differential and Integral Equations
The ODE is linear if it can be written as

n
aj (t)u(j) (t) = g(t) (1.1.3)
j=0
for some coefficients a0 , . . . , an , g, and homogeneous if also g(t) ≡ 0. It

is common to use operator notation for derivatives, especially in the linear
case. Set
d
D= (1.1.4)
dt
so that u = Du, u = D(Du) = D2 u, etc., in which case Eq. (1.1.3) may be
given as

n
Lu := aj (t)Dj u = g(t) (1.1.5)
j=0
By standard calculus properties L is a linear operator, meaning that
L(c1 u1 + c2 u2 ) = c1 Lu1 + c2 Lu2 (1.1.6)
for any scalars c1 , c2 and any n times differentiable functions u1 , u2 .

An ODE normally has infinitely many solutions—the collection of all
solutions is called the general solution of the given ODE.
Example 1.1. By elementary calculus considerations, the simple ODE u = 0
has general solution u(t) = c, where c is an arbitrary constant. Likewise u = u
has the general solution u(t) = cet and u = 2 has the general solution u(t) =
t2 + c1 t + c2 , where c1 , c2 are arbitrary constants. 2
1.1.1 Initial Value Problems

The general solution of an nth order ODE typically contains exactly n arbitrary
constants, whose values may be then chosen so that the solution satisfies n
additional, or side, conditions. The most common kind of side conditions of
interest for an ODE are initial conditions,
u(j) (t0 ) = γj j = 0, 1, . . . , n − 1 (1.1.7)
where t0 is a given point in (a, b) and γ0 , . . . , γn−1 are given constants. Thus
we are prescribing the value of the solution and its derivatives up through order
n − 1 at the point t0 . The problem of solving Eq. (1.1.2) together with the initial
conditions (1.1.7) is called an initial value problem (IVP). It is a very important
fact that under fairly unrestrictive hypotheses a unique solution exists. In stating
conditions on f , we regard it as a function f = f (t, y1 , . . . , yn ) defined on some
domain in Rn+1 .
Some Basic Discussion of Differential and Integral Equations Chapter | 1 3
Theorem 1.1. Assume that

∂f ∂f
f, ,..., (1.1.8)
∂y1 ∂yn
are defined and continuous in a neighborhood of the point (t0 , γ0 , . . . , γn−1 ) ∈
Rn+1 . Then there exists > 0 such that the IVP (1.1.2), (1.1.7) has a unique
solution on the interval (t0 − , t0 + ).
A proof of this theorem may be found in standard ODE textbooks (see,
e.g., [4] or [7]). A slightly weaker version of this theorem will be proved in
Section 3.5. As will be discussed there, the condition of continuity of the partial
derivatives of f with respect to each of the variables yi can actually be replaced
by the weaker assumption that f is Lipschitz continuous with respect to each of
these variables. If we assume only that f is continuous in a neighborhood of the
point (t0 , γ0 , . . . , γn−1 ) then it can be proved that at least one solution exists,
but it may not be unique, see Exercise 1.3. Similar results are valid for systems
of ODEs.
It should also be emphasized that the theorem asserts a local existence
property, that is, only in some sufficiently small interval centered at t0 . It has to
be this way, first of all, since the assumptions on f are made only in the vicinity
of (t0 , γ0 , . . . , γn−1 ). But even if the continuity properties of f were assumed to
hold throughout Rn+1 , then as the following example shows, it would still only
be possible to prove that a solution exists for points t close enough to t0 .
Example 1.2. Consider the first order IVP
u = u2 u(0) = γ (1.1.9)
for which the assumptions of Theorem 1.1 hold for any γ . It may be checked
that the solution of this problem is
γ
u(t) = (1.1.10)
1 − γt
which is only a valid solution for t < 1

γ, which can be arbitrarily small. 2
With more restrictions on f it may be possible to show that the solution exists
on any interval containing t0 , in which case we would say that the solution exists
globally. This is the case, for example, for the linear ODE (1.1.3).
Whenever the conditions of Theorem 1.1 hold, the set of all possible solu-
tions may be regarded as being parametrized by the n constants γ0 , . . . , γn−1 ,
so that as mentioned above, the general solution will contain exactly n arbitrary
parameters. In the special case of the linear equation (1.1.3) it can be shown that
the general solution may be given as

n
u(t) = cj uj (t) + up (t) (1.1.11)
j=1
where up is any particular solution of Eq. (1.1.3), and u1 , . . . , un are any n linearly
independent solutions of the corresponding homogeneous equation Lu = 0. Any
such set of functions u1 , . . . , un is also called a fundamental set for Lu = 0.
Example 1.3. If Lu = u + u then by direct substitution we see that u1 (t) =
sin t, u2 (t) = cos t are solutions, and they are clearly linearly independent. Thus
{sin t, cos t} is a fundamental set for Lu = 0 and u(t) = c1 sin t + c2 cos t is
the general solution of Lu = 0. For the inhomogeneous ODE u + u = et one
may check that up (t) = 12 et is a particular solution, so the general solution is
u(t) = c1 sin t + c2 cos t + 12 et . 2
1.1.2 Boundary Value Problems

For an ODE of degree n ≥ 2 it may be of interest to impose side conditions
at more than one point, typically the endpoints of the interval of interest. We
will then refer to the side conditions as boundary conditions and the problem of
solving the ODE subject to the given boundary conditions as a boundary value
problem (BVP). Since the general solution still contains n parameters, we still
expect to be able to impose a total of n side conditions. However we can see
from simple examples that the situation with regard to existence and uniqueness
in such BVPs is much less clear than for IVPs.
Example 1.4. Consider the BVP
u + u = 0 0 < t < π u(0) = 0 u(π ) = 1 (1.1.12)
Starting from the general solution u(t) = c1 sin t + c2 cos t, the two boundary
conditions lead to u(0) = c2 = 0 and u(π ) = c2 = 1. Since these are
inconsistent, the BVP has no solution. 2
Example 1.5. For the BVP
u + u = 0 0 < t < π u(0) = 0 u(π ) = 0 (1.1.13)
we have solutions u(t) = C sin t for any constant C, that is, the BVP has
infinitely many solutions. 2
The topic of BVPs will be studied in much more detail in Chapter 13.
1.1.3 Some Exactly Solvable Cases

Let us finally review explicit solution methods for some commonly occurring
types of ODEs.
● For the first order linear ODE
u + p(t)u = q(t) (1.1.14)
define the so-called integrating factor ρ(t) = eP(t) where P is any function
satisfying P = p. Multiplying the equation through by ρ we then get the
equivalent equation
(ρu) = ρq (1.1.15)
so if we pick Q such that Q = ρq, the general solution may be given as
Q(t) + C
u(t) = (1.1.16)
ρ(t)
● For the linear homogeneous constant coefficient ODE
n
Lu = aj u(j) = 0 (1.1.17)
j=0
if we look for solutions in the form u(t) = eλt then by direct substitution
we find that u is a solution provided λ is a root of the corresponding
characteristic polynomial
n
P(λ) = aj λj (1.1.18)
j=0
We therefore obtain as many linearly independent solutions as there are

distinct roots of P. If this number is less than n, then we may seek further
solutions of the form teλt , t2 eλt , . . . , until a total of n linearly independent
solutions have been found. In the case of complex roots, equivalent expres-
sions in terms of trigonometric functions are often used in place of complex
exponentials.
● Finally, closely related to the previous case, is the so-called Cauchy-Euler
type equation

n
Lu = (t − t0 )j aj u(j) = 0 (1.1.19)
j=0
for some constants a0 , . . . , an . In this case we look for solutions in the form
u(t) = (t−t0 )λ with λ to be found. Substituting into Eq. (1.1.19) we will find
again an nth order polynomial whose roots determine the possible values of
λ. The interested reader may refer to any standard undergraduate level ODE
book for the additional considerations which arise in the case of complex or
repeated roots.
1.2 INTEGRAL EQUATIONS

In this section we discuss the basic set-up for the study of linear integral
equations. See, for example, [16, 22] as general references in the classical theory
of integral equations. Let ⊂ RN be an open set, K a given function on ×
and set

Tu(x) = K(x, y)u(y) dy (1.2.20)

Here the function K is called the kernel of the integral operator T, which is
linear since Eq. (1.1.6) obviously holds.
A class of associated integral equations is then

λu(x) − K(x, y)u(y) dy = f (x) x ∈ (1.2.21)

for some scalar λ and given function f in some appropriate class. If λ = 0
then Eq. (1.2.21) is said to be a first kind integral equation, otherwise it is
second kind. Let us consider some simple examples which may be studied by
elementary means.
Example 1.6. Let = (0, 1) ⊂ R and K(x, y) ≡ 1. The corresponding first
kind integral equation is therefore
1
− u(y) dy = f (x) 0 < x < 1 (1.2.22)
0
For simplicity here we will assume that f is a continuous function. The left-
hand side is independent of x, thus a solution can exist only if f (x) is a constant
function. When f is constant, on the other hand, infinitely many solutions will
exist, since we just need to find any u with the given definite integral.
For the corresponding second kind equation,
1
λu(x) − u(y) dy = f (x) (1.2.23)
0
a solution, if one exists, must have the specific form u(x) = (f (x) + C)/λ
for some constant C. Substituting into the equation then gives, after obvious
algebra, that
1
C(λ − 1) = f (y) dy (1.2.24)
0
Thus, for any continuous function f and λ = 0, 1, there exists a unique solution
of the integral equation, namely
1
f (x) f (y) dy
u(x) = + 0 (1.2.25)
λ λ(λ − 1)
In the remaining case that λ = 1, it is immediate from Eq. (1.2.24) that a solution
1
can exist only if 0 f (y) dy = 0, in which case u(x) = f (x) + C is a solution for
any choice of C. 2
This very simple example already exhibits features which turn out to be
common to a much larger class of integral equations of this general type.
These are
● The first kind integral equation will require much more restrictive conditions
on f in order for a solution to exist.
● For most λ = 0 the second kind integral equation has a unique solution for
any f .
● There may exist a few exceptional values of λ for which either existence or
uniqueness fails in the corresponding second kind equation.
All of these points will be elaborated and made precise in Chapter 12.
Example 1.7. Let = (0, 1) and
x
Tu(x) = u(y) dy (1.2.26)
0
corresponding to the kernel

1 y<x
K(x, y) = (1.2.27)
0 x≤y
The corresponding integral equation may then be written as
x
λu(x) − u(y) dy = f (x) (1.2.28)
0
This is the prototype of an integral operator of so-called Volterra type, see
Definition 1.1 below.
In the first kind case, λ = 0, we see that f (0) = 0 is a necessary condition
for solvability, in which case the solution is u(x) = −f (x), provided that f is
differentiable in some suitable sense. For λ = 0 we note that differentiation of
Eq. (1.2.28) with respect to x gives
1 f (x)
u − u= (1.2.29)
λ λ
This is an ODE of the type (1.1.14), and so may be solved by the method given
there. The result, after some obvious algebraic manipulation, is

ex/λ 1 x (x−y)/λ
u(x) = f (0) + e f (y) dy (1.2.30)
λ λ 0
Note, however, that by an integration by parts, this formula is seen to be
equivalent to
x
f (x) 1
u(x) = + 2 e(x−y)/λ f (y) dy (1.2.31)
λ λ 0
Observe that Eq. (1.2.30) seems to require differentiability of f even though

Eq. (1.2.31) does not, thus Eq. (1.2.31) would be the preferred solution formula.
It may be verified directly by substitution that Eq. (1.2.31) is a valid solution of
Eq. (1.2.28) for all λ = 0, assuming only that f is continuous on [0, 1]. 2
Concerning the two simple integral equations just discussed, there are again
some features which will turn out to be generally true.
● For the first kind equation, there are fewer restrictions on f needed for
solvability in the Volterra case (1.2.28) than in the non-Volterra case
(1.2.23).
● There are no exceptional values λ = 0 in the Volterra case, that is, a unique
solution exists for every λ = 0 and every continuous f .
Finally let us mention some of the more important ways in which integral
operators, or the corresponding integral equations, are classified:
Definition 1.1. The kernel K(x, y) is called
● symmetric if K(x, y) = K(y, x)
● Volterra type if N = 1 and K(x, y) = 0 for x > y or x < y
● convolution type if K(x,y) = F(x − y) for some function F
● Hilbert-Schmidt type if × |K(x, y)|2 dxdy < ∞
● singular if K(x, y) is unbounded on ×
Important examples of integral operators, some of which will receive much
more attention later in the book, are the Fourier transform

1
Tu(x) = e−ix·y u(y) dy (1.2.32)
(2π )N/2 RN
the Laplace transform
∞
Tu(x) = e−xy u(y) dy (1.2.33)
0
the Hilbert transform
∞
1 u(y)
Tu(x) = dy (1.2.34)
π −∞ x−y
and the Abel operator
x u(y)
Tu(x) = √ dy (1.2.35)
0 x−y
1.3 PARTIAL DIFFERENTIAL EQUATIONS

An mth order partial differential equation (PDE) for an unknown function u =
u(x) on a domain ⊂ RN is any equation of the form
F(x, {Dα u}|α|≤m ) = 0 (1.3.36)
Here we are using the so-called multiindex notation for partial derivatives which
works as follows. A multiindex is vector of nonnegative integers
α = (α1 , α2 , . . . , αN ) αi ∈ {0, 1, . . . } (1.3.37)

In terms of α we define

N
|α| = αi (1.3.38)
i=1
the order of α, and
∂ |α| u
Dα u = (1.3.39)
∂xα11 ∂xα22 . . . ∂xαNN
the corresponding α derivative of u. For later use it is also convenient to define
the factorial of a multiindex
α! = α1 !α2 ! . . . αN ! (1.3.40)
The PDE (1.3.36) is linear if it can be written as

Lu(x) = aα (x)Dα u(x) = g(x) (1.3.41)
|α|≤m
for some coefficient functions aα .
1.3.1 First Order PDEs and the Method of Characteristics

Let us start with the simplest possible example.
Example 1.8. When N = 2 and m = 1 consider
∂u
=0 (1.3.42)
∂x1
By elementary calculus considerations it is clear that u is a solution if and only
if u is independent of x1 , that is,
u(x1 , x2 ) = f (x2 ) (1.3.43)
for some function f . This is then the general solution of the given PDE, which
we note contains an arbitrary function f . 2
Example 1.9. Next consider, again for N = 2, m = 1, the PDE
∂u ∂u
a +b =0 (1.3.44)
∂x1 ∂x2
where a, b are fixed constants, at least one of which is not zero. The equation
amounts precisely to the condition that u has directional derivative 0 in the
direction θ = a, b , so u is constant along any line parallel to θ. This in turn
leads to the conclusion that u(x1 , x2 ) = f (ax2 − bx1 ) for some arbitrary function
f , which at least for the moment would seem to need to be differentiable. 2
The collection of lines parallel to θ , that is, lines ax2 − bx1 = C

obviously play a special role in the previous example, they are the so-called
characteristics, or characteristic curves associated to this particular PDE. The
general concept of characteristic curve will now be described for the case of a
first order linear PDE in two independent variables (with a temporary change of
notation)
a(x, y)ux + b(x, y)uy = c(x, y) (1.3.45)
Consider the associated ODE system
dx dy
= a(x, y) = b(x, y) (1.3.46)
dt dt
and suppose we have some solution pair x = x(t), y = y(t) which we regard as
a parametrically given curve in the (x, y) plane. Any such curve is then defined
to be a characteristic curve for Eq. (1.3.45). The key observation now is that if
u(x, y) is a differentiable solution of Eq. (1.3.45) then
d
u(x(t), y(t)) = a(x(t), y(t))ux (x(t), y(t)) + b(x(t), y(t))uy (x(t), y(t))
dt
= c(x(t), y(t)) (1.3.47)
so that u satisfies a certain first order ODE along any characteristic curve. For
example, if c(x, y) ≡ 0 then, as in the previous example, any solution of the
PDE is constant along any characteristic curve.
We now use this property to construct solutions of Eq. (1.3.45). Let ⊂ R2
be some curve, which we assume can be parametrized as
x = f (s), y = g(s), s0 < s < s1 (1.3.48)
The Cauchy problem for Eq. (1.3.45) consists in finding a solution of

Eq. (1.3.45) with values prescribed on , that is,
u(f (s), g(s)) = h(s) s0 < s < s1 (1.3.49)
for some given function h. Assuming for the moment that such a solution u
exists, let x(t, s), y(t, s) be the characteristic curve passing through (f (s), g(s)) ∈
when t = 0, that is,

∂t = a(x, y) x(0, s) = f (s)
∂x
∂y (1.3.50)
∂t = b(x, y) y(0, s) = g(s)
We must then have

∂
u(x(t, s), y(t, s)) = c(x(t, s), y(t, s)) u(x(0, s), y(0, s)) = h(s) (1.3.51)
∂t
This is a first order IVP in t, depending on s as a parameter, which is guaranteed
to have a solution at least for |t| < for some > 0, provided that c is
continuously differentiable. The three relations x = x(t, s), y = y(t, s), z =

u(x(t, s), y(t, s)) generally amounts to the parametric description of a surface
in R3 containing . If we can eliminate the parameters s, t to obtain the surface
in nonparametric form z = u(x, y) then u is the sought after solution of the
Cauchy problem.
Example 1.10. Let denote the x axis and let us solve
xux + uy = 1 (1.3.52)
with u = h on . Introducing f (s) = s, g(s) = 0 as the parametrization of , we

must then solve
⎧
⎪
⎨ ∂t = x
∂x
x(0, s) = s
∂y
∂t = 1 y(0, s) = 0 (1.3.53)
⎪
⎩∂
∂t u(x(t, s), y(t, s)) = 1 u(0, s) = h(s)
We then easily obtain

x(t, s) = set y(t, s) = t u(x(t, s), y(t, s)) = t + h(s) (1.3.54)
and eliminating t, s yields the solution formula
u(x, y) = y + h(xe−y ) (1.3.55)
The characteristics in this case are the curves x = set , y = t for fixed s, or
x = sey in nonparametric form. Note here that the solution is defined throughout
the x, y plane even though nothing in the preceding discussion guarantees that.
Since h has not been otherwise prescribed we may also regard Eq. (1.3.55)
as the general solution of Eq. (1.3.52), again containing one arbitrary
function. 2
The attentive reader may already realize that this procedure cannot work
in all cases, as is made clear by the following consideration: if c ≡ 0 and is
itself a characteristic curve, then the solution on would have to simultaneously
be equal to the given function h and to be constant, so that no solution can
exist except possibly in the case that h is a constant function. From another,
more general, point of view we must eliminate the parameters s, t by inverting
the relations x = x(s, t), y = y(s, t) to obtain s, t in terms of x, y, at least
near . According to the inverse function theorem this should require that the
Jacobian matrix
∂x ∂y
a(f (s), g(s)) b(f (s), g(s))
∂t ∂t = (1.3.56)
∂x
∂s
∂y
∂s
f (s) g (s)
t=0
be nonsingular for all s. Equivalently the direction f , g should not be parallel

to a, b , and since a, b must be tangent to the characteristic curve, this
amounts to the requirement that itself should have a noncharacteristic tangent
direction at every point. We say that is noncharacteristic for the PDE (1.3.45)
when this condition holds.
The following precise theorem can be established, see, for example, Chapter
1 of [19], or Chapter 3 of [11]. The proof amounts to showing that the method
of constructing a solution just described can be made rigorous under the stated
assumptions.
Theorem 1.2. Let ⊂ R2 be a continuously differentiable curve, which is
noncharacteristic for Eq. (1.3.45), h a continuously differentiable function on
, and let a, b, c be continuously differentiable functions in a neighborhood of
. Then there exists a unique continuously differentiable function u(x, y) defined
in a neighborhood of which is a solution of Eq. (1.3.45).
The method of characteristics is capable of a considerable amount of
generalization, in particular to first order PDEs in any number of independent
variables, and to fully nonlinear first PDEs, see the references just mentioned
previously.
1.3.2 Second Order Problems in R2

In order to better understand what can be known about solutions of a potentially
complicated looking PDE, one natural approach is to try to obtain another
equation which is equivalent to the original one, but is somehow simpler in
structure. We illustrate with the following special type of second order PDE in
two independent variables x, y:
Auxx + Buxy + Cuyy = 0 (1.3.57)
where A, B, C are real constants, not all zero. Consider introducing new coordi-
nates ξ , η by means of a linear change of variable
ξ = αx + βy η = γ x + δy (1.3.58)
with αδ − βγ = 0, so that the transformation is invertible. Our goal is to make a

good choice of α, β, γ , δ so as to achieve a simpler looking, but equivalent PDE
to study.
Given any PDE and any change of coordinates, we obtain the expression
for the PDE in the new coordinate system by straightforward application of the
chain rule. In the case at hand we have
∂u ∂u ∂ξ ∂u ∂η ∂u ∂u
= + =α +γ (1.3.59)
∂x ∂ξ ∂x ∂η ∂x ∂ξ ∂η
2
∂ u ∂ ∂ ∂u ∂u
= α + γ α + γ
∂x2 ∂ξ ∂η ∂ξ ∂η
2u
(1.3.60)
2∂ ∂ 2u 2
2∂ u
=α + 2αγ + γ
∂ξ 2 ∂ξ ∂η ∂η2
with similar expressions for uxy and uyy . Substituting into Eq. (1.3.57) the
resulting PDE is
auξ ξ + buξ η + cuηη = 0 (1.3.61)
where
a = α 2 A + αβB + β 2 C (1.3.62)
b = 2αγ A + (αδ + βγ )B + 2βδC (1.3.63)
c = γ 2 A + γ δB + δ 2 C (1.3.64)
We now seek to make special choices of α, β, γ , δ to achieve as simple a form
as possible for the transformed PDE (1.3.61).
Suppose first that B2 − 4AC > 0, so that there exist two real and distinct
roots r1 , r2 of Ar2 + Br + C = 0. If α, β, γ , δ are chosen so that
α γ
= r1 = r2 (1.3.65)
β δ
then a = c = 0, and αδ − βγ = 0, so that the transformed PDE is simply, after
division by a constant, uξ η = 0. The general solution of this second order PDE
is easily obtained: uξ must be a function of ξ alone, so integrating with respect
to ξ and observing that the “constant of integration” could be any function of η,
we get
u(ξ , η) = F(ξ ) + G(η) (1.3.66)
for any differentiable functions F, G. Finally reverting to the original coordinate
system, the result is
u(x, y) = F(αx + βy) + G(γ x + δy) (1.3.67)
an expression for the general solution containing two arbitrary functions.
The lines αx + βy = C, γ x + δy = C are called the characteristics for
Eq. (1.3.57). Characteristics are an important concept for this and some more
general second order PDEs, but they don’t play as central a role as in the first
order case.
Example 1.11. For the PDE
uxx − uyy = 0 (1.3.68)
we have − 4AC > 0 and the roots r satisfy − 1 = 0. We may then choose,
B2 r2
for example, α = β = γ = 1, δ = −1, to get the general solution
u(x, y) = F(x + y) + G(x − y) (1.3.69)
2
Next assume that B2 − 4AC = 0. If either of A or C is 0, then so is B, in
which case the PDE already has the form uξ ξ = 0 or uηη = 0, say the first of
these without loss of generality. Otherwise, choose
B
α=− β=1 γ =1 δ=0 (1.3.70)
2A
to obtain a = b = 0, c = A, so that the transformed PDE in all cases may be
taken to be uξ ξ = 0.
Finally, if B2 − 4AC < 0 then A = 0 must hold, and we may choose
2A −B
α=√ β=√ γ =0 δ=1 (1.3.71)
4AC − B 2 4AC − B2
in which case the transformed equation is
uξ ξ + uηη = 0 (1.3.72)
We have therefore established that any PDE of the type (1.3.57) can be
transformed, by means of a linear change of variables, to one of the three simple
types,
uξ η = 0 uξ ξ = 0 uξ ξ + uηη = 0 (1.3.73)
each of which then leads to a prototype for a certain larger class of PDEs. If we
allow lower order terms
Auxx + Buxy + Cuyy + Dux + Euy + Fu = 0 (1.3.74)
then after the transformation (1.3.58) it is clear that the lower order terms remain
as lower order terms. Thus any PDE of the type (1.3.74) is, up to a change of
coordinates, one of the three types (1.3.73), up to lower order terms, and only
the value of the discriminant B2 − 4AC needs to be known to determine which
of the three types is obtained.
The preceding discussion motivates the following classification: The PDE
(1.3.74) is said to be:
● hyperbolic if B2 − 4AC > 0
● parabolic if B2 − 4AC = 0
● elliptic if B2 − 4AC < 0
The terminology comes from an obvious analogy with conic sections, that is, the
solution set of Ax2 + Bxy + Cy2 + Dx + Ey + F = 0 is respectively a hyperbola,
parabola, or ellipse (or a degenerate case) according as B2 − 4AC is positive,
zero, or negative.
We can also allow the coefficients A, B, . . . , G to be variable functions of x, y,
and in this case the classification is done pointwise, so the type can change. An
important example of this phenomenon is the so-called Tricomi equation (see,
e.g., Chapter 12 of [14])
uxx − xuyy = 0 (1.3.75)
which is hyperbolic for x > 0 and elliptic for x < 0. One might refer to the
equation as being parabolic for x = 0 but generally speaking we do not do this,
since it is not really meaningful to speak of a PDE being satisfied in a set without
interior points.
The preceding discussion is special to the case of N = 2 independent
variables—in the case of N ≥ 3 there is no such complete classification. As
we will see there are still PDEs referred to as being hyperbolic, parabolic, or
elliptic, but there are others which are not of any of these types, although these
tend to be of less physical importance.
1.3.3 Further Discussion of Model Problems

According to the previous discussion, we should focus our attention on a
representative problem for each of the three types, since then we will also gain
considerable information about other problems of the given type.
Wave Equation
For the hyperbolic case we consider the wave equation
utt − c2 uxx = 0 (1.3.76)
where c > 0 is a constant. Here we have changed the name of the variable y
to t, following the usual convention of regarding u = u(x, t) as depending on a
“space” variable x and “time” variable t. This PDE arises in the simplest model
of wave propagation in one space dimension, where u represents, for example,
the displacement of a vibrating medium from its equilibrium position, and c is
the wave speed.
Following the procedure outlined at the beginning of this section, an
appropriate change of coordinates is ξ = x + ct, η = x − ct, and we obtain
the expression, also known as d’Alembert’s formula, for the general solution,
u(x, t) = F(x + ct) + G(x − ct) (1.3.77)
for arbitrary twice differentiable functions F, G. The general solution may be
viewed as the superposition of two waves of fixed shape, moving to the right
and to the left with speed c.
The IVP for the wave equation consists in solving Eq. (1.3.76) for x ∈ R and
t > 0 subject to the side conditions
u(x, 0) = f (x) ut (x, 0) = g(x) x∈R (1.3.78)
where f , g represent the initial displacement and initial velocity of the vibrating
medium. This problem may be completely and explicitly solved by means of
d’Alembert’s formula. Setting t = 0 and using the prescribed side conditions,
we must have
F(x) + G(x) = f (x) c(F (x) − G (x)) = g(x) x∈R (1.3.79)
x
Integrating the second relation gives F(x) − G(x) = 1c 0 g(s) ds + C for some
constant C, and combining with the first relation yields

1 1 x
F(x) = f (x) + g(s) ds + C
2 c 0
(1.3.80)
1 1 x
G(x) = f (x) − g(s) ds − C
2 c 0
Substituting into Eq. (1.3.77) and doing some obvious simplification we obtain

1 1 x+ct
u(x, t) = (f (x + ct) + f (x − ct)) + g(s) ds (1.3.81)
2 2c x−ct
We remark that a general solution formula like Eq. (1.3.77) can be given for
any PDE which is exactly transformable to uξ η = 0, that is to say, any hyperbolic
PDE of the form (1.3.57), but once lower order terms are allowed such a simple
solution method is no longer available. For example, the so-called Klein-Gordon
equation utt − uxx + u = 0 may be transformed to uξ η + 4u = 0 which
unfortunately cannot be solved in so transparent a form. Thus the d’Alembert
solution method, while very useful when applicable, is limited in its scope.
Heat Equation
Another elementary method, which may be used in a wide variety of situations,
is the separation of variables technique. We illustrate with the case of the initial
and boundary value problem
ut = uxx 0<x<1 t>0 (1.3.82)
u(0, t) = u(1, t) = 0 t>0 (1.3.83)
u(x, 0) = f (x) 0<x<1 (1.3.84)
Here Eq. (1.3.82) is the heat equation, a parabolic equation modeling, for
example, the temperature in a one-dimensional medium u = u(x, t) as a function
of location x and time t, Eq. (1.3.83) are the boundary conditions, stating that
the temperature is held at temperature zero at the two boundary points x = 0
and x = 1 for all t, and Eq. (1.3.84) represents the initial condition, that is, that
the initial temperature distribution is given by the prescribed function f (x).
We begin by ignoring the initial condition and otherwise looking for special
solutions of the form u(x, t) = φ(t)ψ(x). Obviously u = 0 is such a solution,
but cannot be of any help in eventually solving the full stated problem, so we
insist that neither of φ or ψ is the zero function. Inserting into Eq. (1.3.82) we
obtain immediately that
φ (t)ψ(x) = φ(t)ψ (x) (1.3.85)
must hold, or equivalently
φ (t) ψ (x)
= (1.3.86)
φ(t) ψ(x)
Since the left side depends on t alone and the right side on x alone, it must be that
both sides are equal to a common constant which we denote by −λ (without yet
at this point ruling out the possibility that λ itself is negative or even complex).
We have therefore obtained ODEs for φ and ψ
φ (t) + λφ(t) = 0 ψ (x) + λψ(x) = 0 (1.3.87)
linked via the separation constant λ. Next, from the boundary condition (1.3.83)
we get φ(t)ψ(0) = φ(t)ψ(1) = 0, and since φ is nonzero we must have ψ(0) =
ψ(1) = 0.
The ODE and side conditions for ψ, namely
ψ (x) + λψ(x) = 0 0 < x < 1 ψ(0) = ψ(1) = 0 (1.3.88)
is the simplest example of a so-called Sturm-Liouville problem, a topic which
will be studied in detail in Chapter 13, but this particular case can be handled
by elementary considerations. We emphasize that our goal is to find nonzero
solutions of Eq. (1.3.88), along with the values of λ these correspond to, and as
we will see, only certain values of λ will be possible.
Considering first the case that λ > 0, the general solution of the ODE is
√ √
ψ(x) = c1 sin λx + c2 cos λx (1.3.89)
The first boundary

√ condition ψ(0) = 0 implies that c2 = 0 while the second
gives c1 sin λ √ = 0. √have c1 = 0, since otherwise ψ = 0,
We are not allowed to
so instead sin λ = 0 must hold, that is, λ = π , 2π , . . . . Thus we have found
one collection of solutions of Eq. (1.3.88), which we denote ψk (x) = sin kπ x,
k = 1, 2, . . . . Since they were found under the assumption that λ > 0, we
should next consider other possibilities, but it turns out that we have already
found all√possible solutions of Eq. (1.3.88). For example, if we suppose λ < 0
and k = −λ then to solve Eq. (1.3.88) we must have ψ(x) = c1 ekx + c2 e−kx .
From the boundary conditions
c1 + c2 = 0 c1 ek + c2 e−k = 0 (1.3.90)
we see that the unique solution is c1 = c2 = 0 for any k > 0. Likewise we can
check that ψ = 0 is the only possible solution for k = 0 and for nonreal k.
For each allowed value of λ we obviously have the corresponding function
φ(t) = e−λt , so that
uk (x, t) = e−k
2π 2t
sin kπx k = 1, 2, . . . (1.3.91)
represents, aside from multiplicative constants, all possible product solutions of
Eqs. (1.3.82), (1.3.83).
To complete the solution of the initial and boundary value problem, we
observe that any sum ∞ k=1 ck uk (x, t) is also a solution of Eqs. (1.3.82), (1.3.83)
as long as ck → 0 sufficiently rapidly, and we try to choose the coefficients ck
to achieve the initial condition (1.3.84). This amounts to the requirement that
∞

f (x) = ck sin kπx (1.3.92)
k=1
hold. For any f for which such a sine series representation is valid, we then have
the solution of the given PDE problem
∞

ck e−k
2π 2t
u(x, t) = sin kπ x (1.3.93)
k=1
The question then becomes to characterize this set of f ’s in some more straight-
forward way, and this is done, among many other things, within the theory
of Fourier series, which will be discussed in Chapter 7. Roughly speaking the
conclusion will be that essentially any reasonable function can be represented
this way, but there are many aspects to this, including elaboration of the precise
sense in which the series converges. One other fact concerning this series which
we can easily anticipate at this point, is a formula for the coefficient ck : If we
assume that Eq. (1.3.92) holds, we can multiply both sides by sin mπ x for some
integer m and integrate with respect to x over (0, 1), to obtain
1 1
cm
f (x) sin mπx dx = cm sin2 mπx dx = (1.3.94)
0 0 2
1
since 0 sin kπx sin mπ x dx = 0 for k = m. Thus, if f is representable by a sine
series, there is only one possibility for the kth coefficient, namely
1
ck = 2 f (x) sin kπx dx (1.3.95)
0
Laplace Equation
Finally we discuss a model problem of elliptic type,
uxx + uyy = 0 x2 + y2 < 1 (1.3.96)
u(x, y) = f (x, y) x2 + y2 = 1 (1.3.97)
where f is a given function. The PDE in Eq. (1.3.96) is known as Laplace’s
∂2 ∂2
equation, and is commonly written as u = 0 where = ∂x 2 + ∂y2 is the
Laplace operator, or Laplacian. A function satisfying Laplace’s equation in some
set is said to be a harmonic function on that set, thus we are solving the bound-
ary value problem of finding a harmonic function in the unit disk x2 + y2 < 1
subject to a prescribed boundary condition on the boundary of the disk.
One should immediately recognize that it would be natural here to make use
of polar coordinates (r, θ ), where according to the usual calculus notations,

y
r = x2 + y2 tan θ = x = r cos θ y = r sin θ (1.3.98)
x
and we regard u = u(r, θ ) and f = f (θ ).
To begin we need to find the expression for Laplace’s equation in polar

coordinates. Again this is a straightforward calculation with the chain rule, for
example,
∂u ∂u ∂r ∂u ∂θ
= + (1.3.99)
∂x ∂r ∂x ∂θ ∂x
x ∂u y ∂u
= − 2 (1.3.100)
x + y ∂r
2 2 x + y ∂θ
2
∂u sin θ ∂u
= cos θ − (1.3.101)
∂r r ∂θ
∂u
and similar expressions for ∂y and the second derivatives. The end result is
1 1
uxx + uyy = urr + ur + 2 uθθ = 0 (1.3.102)
r r
We may now try separation of variables, looking for solutions in the special
product form u(r, θ ) = R(r)(θ ). Substituting into Eq. (1.3.102) and dividing
by R gives
R (r) R (r) (θ )
r2 +r =− (1.3.103)
R(r) R(r) (θ )
so both sides must be equal to a common constant λ. Therefore R and must
be nonzero solutions of
+ λ = 0 r2 R + rR − λR = 0 (1.3.104)
Next it is necessary to recognize that there are two “hidden” side conditions
which we must make use of. The first of these is that must be 2π periodic,
since otherwise it would not be possible to express the solution u in terms of
the original variables x, y in an unambiguous way. We can make this explicit by
requiring
(0) = (2π ) (0) = (2π ) (1.3.105)
As in the case of Eq. (1.3.88) we can search for allowable values of λ by

considering the various cases λ > 0, λ < 0, etc. The outcome is that nontrivial
solutions exist precisely if λ = k2 , k = 0, 1, 2, . . . , with corresponding solutions
being, up to multiplicative constant,

1 k=0
ψk (x) = (1.3.106)
sin kx or cos kx k = 1, 2, . . .
If one is willing to use the complex valued solutions, we could replace

sin kx, cos kx by e±ikx for k = 1, 2, . . . .
With λ determined we must next solve the corresponding R equation,
r2 R + rR − k2 R = 0 (1.3.107)
which is of the Cauchy-Euler type (1.1.19). The general solution is

c1 + c2 log r k = 0
R(r) = (1.3.108)
c1 rk + c2 r−k k = 1, 2, . . .
and here we encounter the second hidden condition: the solution R should not be
singular at the origin, since otherwise the PDE would not be satisfied throughout
the unit disk. Thus we should choose c2 = 0 in each case, leaving R(r) = rk , k =
0, 1, . . . .
Summarizing, we have found all possible product solutions R(r)(θ ) of
Eq. (1.3.96), and they are
1, rk sin kθ , rk cos kθ k = 1, 2, . . . (1.3.109)
up to constant multiples. Any sum of such terms is also a solution of

Eq. (1.3.96), so we may seek a solution of Eqs. (1.3.96), (1.3.97) in the form
∞

u(r, θ ) = a0 + ak rk cos kθ + bk rk sin kθ (1.3.110)
k=1
The coefficients must then be determined from the requirement that

∞

f (θ ) = a0 + ak cos kθ + bk sin kθ (1.3.111)
k=1
This is another problem in the theory of Fourier series, which is very similar to
that associated with Eq. (1.3.92), and again will be studied in detail in Chapter 7.
Exact formulas for the coefficients in terms of f may be given, as in Eq. (1.3.95),
see Exercise 1.19.
1.3.4 Standard Problems and Side Conditions

Let us now formulate a number of typical PDE problems which will recur
throughout this book, and which are for the most part variants of the model
problems discussed in the previous section. Let be some domain in RN and
let ∂ denote the boundary of . For any sufficiently differentiable function u,
the Laplacian of u is defined to be

N
∂ 2u
u = (1.3.112)
k=1
∂xk2
● The PDE
u = h x ∈ (1.3.113)
is Poisson’s equation, or Laplace’s equation in the special case that h = 0.
It is regarded as being of elliptic type, by analogy with the N = 2 case
discussed in the previous section, or on account of a more general definition
of ellipticity which will be given in Chapter 8. The most common type of
side conditions associated with this PDE are
– Dirichlet, or first kind, boundary conditions
u(x) = g(x) x ∈ ∂ (1.3.114)
– Neumann, or second kind, boundary conditions
∂u
(x) = g(x) x ∈ ∂ (1.3.115)
∂n
∂n (x) = (∇u · n)(x) is the directional derivative in the direction of
where ∂u
the outward normal direction n(x) for x ∈ ∂.
– Robin, or third kind, boundary conditions
∂u
(x) + σ (x)u(x) = g(x) x ∈ ∂ (1.3.116)
∂n
for some given function σ .
● The PDE
u + λu = h x ∈ (1.3.117)
where λ is some constant, is the Helmholtz equation, also of elliptic type.
The three types of boundary condition mentioned for the Poisson equation
may also be imposed in this case.
● The PDE
ut = u x∈ t>0 (1.3.118)
is the heat equation and is of parabolic type. Here u = u(x, t), where
x is regarded as a spatial variable and t a time variable. By convention,
the Laplacian acts only with respect to the N spatial variables x1 , . . . , xN .
Appropriate side conditions for determining a solution of the heat equation
are an initial condition
u(x, 0) = f (x) x ∈ (1.3.119)
and boundary conditions of the Dirichlet, Neumann, or Robin type men-
tioned above. The only needed modification is that the functions involved
may be allowed to depend on t, for example, the Dirichlet boundary
condition for the heat equation is
u(x, t) = g(x, t) x ∈ ∂ t > 0 (1.3.120)
and similarly for the other two types.
● The PDE
utt = u x∈ t>0 (1.3.121)
is the wave equation and is of hyperbolic type. Since it is second order in t
it is natural that there be two initial conditions, usually given as
u(x, 0) = f (x) ut (x, 0) = g(x) x∈ (1.3.122)
Suitable boundary conditions for the wave equation are precisely the same
as for the heat equation.
● Finally, the PDE
iut = u x ∈ RN t>0 (1.3.123)
is the Schrödinger equation. Even when N = 1 it does not fall under the
classification
√ scheme of Section 1.3.2 because of the complex coefficient
i = −1. It is nevertheless one of the fundamental PDEs of mathematical
physics, and we will have some things to say about it in later chapters.
The spatial domain here is taken to be all of RN rather than a subset
because this is by far the most common situation and the only one which
will arise in this book. Since there is no spatial boundary, the only needed
side condition is an initial condition for u, u(x, 0) = f (x), as in the heat
equation case.
1.4 WELL-POSED AND ILL-POSED PROBLEMS

All of the PDEs and associated side conditions discussed in the previous section
turn out to be natural, in the sense that they lead to what are called well-
posed problems, a somewhat imprecise concept which we explain next. Roughly
speaking a problem is well-posed if
● A solution exists.
● The solution is unique.
● The solution depends continuously on the data.
Here by “data” we mean any of the ingredients of the problem which we might
imagine being changed, to obtain another problem of the same general type. For
example, in the Dirichlet problem for the Poisson equation
u = f x∈ u = 0 x ∈ ∂ (1.4.124)
the term f = f (x) would be regarded as the given data. The idea of continuous
dependence is that if a “small” change is made in the data, then the resulting
solution should also undergo only a small change. For such a notion to be made
precise, it is necessary to have some specific idea in mind of how we would
measure the magnitude of a change in f . As we shall see, there may be many
natural ways to do so, and no precise statement about well-posedness can be
given until such choices are made. In fact, even the existence and uniqueness
requirements, which may seem more clear cut, may also turn out to require
much clarification in terms of what the exact meaning of “solution” is.
A problem which is not well-posed is called ill-posed. A classical problem
in which ill-posedness can be easily recognized is Hadamard’s example:
uxx + uyy = 0 −∞<x<∞ y>0 (1.4.125)
u(x, 0) = 0 uy (x, 0) = g(x) −∞<x<∞ (1.4.126)
Note that it is not of one of the standard types mentioned above.
If g(x) = α sin kx for some α, k > 0 then a corresponding solution is
sin kx ky
u(x, y) = α e (1.4.127)
k
This is known to be the unique solution, but notice that a change in α (i.e., of the
data g) of size implies a corresponding change in the solution for, say, y = 1 of
size ek . Since k can be arbitrarily large, it follows that the problem is ill-posed,
that is, small changes in the data do not necessarily lead to small changes in the
solution.
Note that in this example if we change the PDE from uxx + uyy = 0 to
uxx − uyy = 0 then (aside from the name of a variable) we have precisely the
problem (1.3.76), (1.3.78), which from the explicit solution (1.3.81) may be seen
to be well-posed under any reasonable interpretation. This serves to emphasize
that some care must be taken in recognizing what are the “correct” side
conditions for a given PDE. Other interesting examples of ill-posed problems
are given in Exercises 1.23 and 1.26, see also [26].
1.5 EXERCISES
1.1. Find a fundamental set and the general solution of u + u + u = 0.
1.2. Let L = aD2 + bD + c (a = 0) be a constant coefficient second order
linear differential operator, and let p(λ) = aλ2 + bλ + c be the associated
characteristic polynomial. If λ1 , λ2 are the roots of p, show that we can
express the operator L as L = a(D − λ1 )(D − λ2 ). Use this factorization
to obtain the general solution of Lu = 0 in the case of repeated roots,
λ1 = λ2 .
√
1.3. Show that the solution of the IVP y = 3 y, y(0) = 0 is not unique.
(Hint: y(t) = 0 is one solution, find another one.) Why doesn’t this
contradict the assertion in Theorem 1.1 about unique solvability of
the IVP?
1.4. Solve the IVP for the Cauchy-Euler equation
(t + 1)2 u + 4(t + 1)u − 10u = 0 u(1) = 2 u (1) = −1

1.5. Consider the integral equation

1
λu(x) − K(x, y)u(y) dy = g(x)
0
for the kernel
x2
K(x, y) =
1 + y3
(a) For what values of λ ∈ C does there exist a unique solution for any
function g which is continuous on [0, 1]?
(b) Find the solution set of the equation for all λ ∈ C and continuous
functions g. (Hint: For λ = 0 any solution must have the form u(x) =
− g(x)
λ + Cx for some constant C.)
2
1
1.6. Find a kernel K(x, y) such that u(x) = 0 K(x, y)f (y) dy is the unique
solution of
u + u = f (x) u(0) = u (0) = 0
(Hint: Review the variation of parameters method in any undergraduate
ODE textbook.)
1.7. If f ∈ C([0, 1]),

y(x − 1) 0 < y < x < 1
K(x, y) =
x(y − 1) 0 < x < y < 1
and
1
u(x) = K(x, y)f (y) dy
0
show that
u = f 0<x<1 u(0) = u(1) = 0
1.8. For each of the integral operators in Eqs. (1.2.26) and (1.2.32)–(1.2.35),
discuss the classification(s) of the corresponding kernel, according to
Definition 1.1.
1.9. Find the general solution of (1 + x2 )ux + uy = 0. Sketch some of the
characteristic curves.
1.10. The general solution in Example 1.10 was found by solving the cor-
responding Cauchy problem with being the x axis. But the general
solution should not actually depend on any specific choice of . Show
that the same general solution is found if instead we take to be the y
axis.
1.11. Find the solution of
u(0, y) = e−y
2
yux + xuy = 1
Discuss why the solution you find is only valid for |y| ≥ |x|.
1.12. The method of characteristics developed in Section 1.3.1 for the linear
PDE (1.3.45) can be easily extended to the so-called semilinear equation
a(x, y)ux + b(x, y)uy = c(x, y, u) (1.5.128)
We simply replace Eq. (1.3.47) by
d
u(x(t), y(t)) = c(x(t), y(t), u(x(t), y(t))) (1.5.129)
dt
which is still an ODE along a characteristic. With this in mind, solve
1
ux + xuy + u2 = 0 u(0, y) = y>0 (1.5.130)
y
1.13. Find the general solution of uxx − 4uxy + 3uyy = 0.

1.14. Find the regions of the xy plane where the PDE
yuxx − 2uxy + xuyy − 3ux + u = 0
is elliptic, parabolic, and hyperbolic.
1.15. Find a solution formula for the half line wave equation problem
utt − c2 uxx = 0 x>0 t>0 (1.5.131)
u(0, t) = h(t) t>0 (1.5.132)
u(x, 0) = f (x) x > 0 (1.5.133)
ut (x, 0) = g(x) x > 0 (1.5.134)
Note where the solution coincides with Eq. (1.3.81) and explain why this
should be expected.
1.16. Complete the details of verifying Eq. (1.3.102).
1.17. If u is a twice differentiable function on RN depending only on r = |x|,
show that
N−1
u = urr + ur
r
(Spherical coordinates in RN are reviewed in Section A.4, but the details
of the angular variables are not needed for this calculation. Start by
x
showing that ∂x∂u
j
= u (r) rj .)
1.18. Verify in detail that there are no nontrivial solutions of Eq. (1.3.88) for
nonreal λ ∈ C.
1.19. Assuming that Eq. (1.3.111) is valid, find the coefficients ak , bk in terms
of f . (Hint: multiply the equation by sin mθ or cos mθ and integrate from
0 to 2π.)
1.20. In the two-dimensional case, solutions of Laplace’s equation u = 0
may also be found by means of analytic function theory. Recall that if
z = x + iy then a function f (z) is analytic in an open set if f (z) exists
at every point of . If we think of f = u + iv and u, v as functions of x, y
then u = u(x, y), v = v(x, y) must satisfy the Cauchy-Riemann equations

ux = vy , uy = −vx . Show in this case that u, v are also solutions of
Laplace’s equation. Find u, v if f (z) = z3 and f (z) = ez .
1.21. Find all of the product solutions u(x, t) = φ(t)ψ(x) that you can which
satisfy the damped wave equation
utt + αut = uxx 0<x<π t>0
and the boundary conditions
u(0, t) = ux (π, t) = 0 t>0
Here α > 0 is the damping constant. What is the significance of the
condition α < 1?
1.22. Show that any solution of the wave equation utt − uxx = 0 has the “four
point property”
u(x, t) + u(x + h − k, t + h + k) = u(x + h, t + h) + u(x − k, t + k)
for any h, k. (Suggestion: Use d’Alembert’s formula.)
1.23. In the Dirichlet problem for the wave equation
utt − uxx = 0 0<x<1 0<t<1
u(0, t) = u(1, t) = 0 0<t<1
u(x, 0) = 0 u(x, 1) = f (x) 0<x<1
show that neither existence nor uniqueness holds. (Hint: For the nonex-
istence part, use Exercise 1.22 to find an f for which no solution exists.)
1.24. Let be the rectangle [0, a] × [0, b] in R2 . Find all possible product
solutions
u(x, y, t) = φ(t)ψ(x)ζ (y)
satisfying
ut − u = 0 (x, y) ∈ t>0
u(x, y, t) = 0 (x, y) ∈ ∂ t > 0
1.25. Find a solution of the Dirichlet problem for u = u(x, y) in the unit disc
= {(x, y) : x2 + y2 < 1},
u = 1 (x, y) ∈ u(x, y) = 0 (x, y) ∈ ∂
(Suggestion: look for a solution in the form u = u(r) and recall
Eq. 1.3.102.)
1.26. The problem
ut = uxx 0<x<1 t<T (1.5.135)
u(0, t) = u(1, t) = 0 t>0 (1.5.136)
u(x, T) = f (x) 0<x<1 (1.5.137)
is sometimes called a final value problem for the heat equation.

(a) Show that this problem is ill-posed.
(b) Show that this problem is equivalent to Eqs. (1.3.82)–(1.3.84) except
with the heat equation (1.3.82) replaced by the backward heat
equation ut = −uxx .
Chapter 2
Vector Spaces
We will be working frequently with function spaces which are themselves
special cases of more abstract spaces. Most such spaces which are of interest
to us have both linear structure and metric structure. This means that given any
two elements of the space it is meaningful to speak of (i) a linear combination
of the elements, and (ii) the distance between the two elements. These two kinds
of concepts are abstracted in the definitions of vector space and metric space. In
this chapter we focus on the first of these aspects.
2.1 AXIOMS OF A VECTOR SPACE

Definition 2.1. A vector space is a set X such that whenever x, y ∈ X and λ is a
scalar we have x + y ∈ X and λx ∈ X, and for which the following axioms hold.
[V1] x + y = y + x for all x, y ∈ X
[V2] (x + y) + z = x + (y + z) for all x, y, z ∈ X
[V3] There exists an element 0 ∈ X such that x + 0 = x for all x ∈ X
[V4] For every x ∈ X there exists an element −x ∈ X such that x + (−x) = 0
[V5] λ(x + y) = λx + λy for all x, y ∈ X and any scalar λ
[V6] (λ + μ)x = λx + μx for any x ∈ X and any scalars λ, μ
[V7] λ(μx) = (λμ)x for any x ∈ X and any scalars λ, μ
[V8] 1x = x for any x ∈ X
Here the field of scalars may be either the real numbers R or the complex
numbers C, and we may refer to X as a real or complex vector space accordingly,
if a distinction needs to be made.
if x1 , . . . , xm ∈ X and λ1 , . . . , λm are
By an obvious induction argument,
scalars, then the linear combination m j=1 λj xj must also be an element of X.
Example 2.1. Ordinary N-dimensional Euclidean space
RN := {x = (x1 , x2 , . . . , xN ) : xj ∈ R}
is a real vector space with the usual operations of vector addition and scalar
multiplication,

http://dx.doi.org/10.1016/B978-0-12-811426-1.00002-7
(x1 , x2 , . . . , xN ) + (y1 , y2 , . . . , yN ) = (x1 + y1 , x2 + y2 , . . . , xN + yN )

λ(x1 , x2 , . . . , xN ) = (λx1 , λx2 , . . . , λxN ) λ ∈ R
If we allow the components xj as well as the scalars λ to be complex, we obtain
instead the complex vector space CN . 2
Example 2.2. If E ⊂ RN , let
C(E) = {f : E → R : f is continuous at x for every x ∈ E}
denote the set of real valued continuous functions on E. Clearly C(E) is a
real vector space with the ordinary operations of function addition and scalar
multiplication
(f + g)(x) = f (x) + g(x) (λf )(x) = λf (x) λ ∈ R
If we allow the range space in the definition of C(E) to be C then C(E) becomes
a complex vector space.
Spaces of differentiable functions likewise may be naturally regarded as
vector spaces, for example,
Cm (E) = {f : Dα f ∈ C(E), |α| ≤ m}

and
C∞ (E) = {f : Dα f ∈ C(E), for all α}
2
Example 2.3. If 0 < p < ∞ and E is a measurable subset of RN , the space
Lp (E) is defined to be the set of measurable functions f : E → R or f : E → C
such that

|f (x)|p dx < ∞ (2.1.1)
E
Here the integral is defined in the Lebesgue sense. The reader unfamiliar with
measure theory and Lebesgue integration should consult a standard textbook
such as [30, 32], or see a brief summary in Section A.1.
We may now verify that Lp (E) is vector space for any 0 < p < ∞. To
see this we use the known fact that if f , g are measurable then so are f + g
and λf for any scalar λ, and the numerical inequality (a + b)p ≤ Cp (ap + bp )
holds for a, b ≥ 0, where Cp = max (2p−1 , 1). It follows from these facts that
f + g ∈ Lp (E) whenever f , g ∈ Lp (E) and checking the remaining axioms is
routine.
The related space L∞ (E) is defined as the set of measurable functions f
for which
ess supx∈E |f (x)| < ∞ (2.1.2)
Here M = ess supx∈E |f (x)| (the essential supremum of |f |) if |f (x)| ≤ M a.e.
and there is no smaller constant with this property. We leave the verification of
the vector space axioms as an exercise. 2
Vector Spaces Chapter | 2 31
Definition 2.2. If X is a vector space, a subset M ⊂ X is a subspace of X if

(i) x + y ∈ M whenever x, y ∈ M
(ii) λx ∈ M whenever x ∈ M and λ is a scalar
That is to say, a subspace is a subset of X which is closed under formation

of linear combinations. Clearly a subspace of a vector space is itself a vector
space.
Example 2.4. The subset M = {x ∈ RN : xj = 0} is a subspace of RN for any
fixed j. 2
Example 2.5. If E ⊂ RN then C∞ (E) is a subspace of Cm (E) for any m, which
in turn is a subspace of C(E). 2
Example 2.6. If X is any vector space and S ⊂ X, then span(S), the span of
S, is the set of all finite linear combinations of elements of S. That is to say,
x ∈ span(S) if there exist scalars λ1 , λ2 , . . . λm and elements x1 , . . . xm ∈ S
such that
m
x= λj xj (2.1.3)
j=1
It is obviously a subspace of X, also sometimes referred to as the subspace
generated by S, or the linear envelope of S. 2
Example 2.7. If we take X = C([a, b]) and fj (x) = xj−1 for j = 1, 2, . . . then
the subspace generated by {fj }N+1
j=1 is PN , the vector space of polynomials of
degree less than or equal to N. Likewise, the subspace generated by {fj }∞
j=1 is P,
the vector space of all polynomials. 2
2.2 LINEAR INDEPENDENCE AND BASES

Definition 2.3. We say that S ⊂ X is linearly independent if whenever
x1 , . . . , xm ∈ S, λ1 , . . . , λm are scalars and m
j=1 λ j xj = 0 then λ1 = λ2 =
· · · = λm = 0. Otherwise S is linearly dependent.
Equivalently, S is linearly dependent if it is possible to express at least one
of its elements as a linear combination of the remaining ones. In particular any
set containing the zero element is linearly dependent.
Definition 2.4. We say that S ⊂ X is a basis of X if for any x ∈ X there
m unique scalars λ1 , λ2 , . . . , λm and elements x1 , . . . , xm ∈ S such that x =
exists
j=1 λj xj .
The following characterization of a basis is then immediate:
Proposition 2.1. S ⊂ X is a basis of X if and only if S is linearly independent
and span(S) = X.
It is important to emphasize that in this definition of basis it is required that
every x ∈ X be expressible as a finite linear combination of the basis elements.
This notion of basis will be inadequate for later purposes, and will be replaced
by one which allows infinite sums, but this cannot be done until a meaning of
convergence is available. The notion of basis in Definition 2.4 is called a Hamel
basis if a distinction is necessary.
Definition 2.5. We say that dim (X), the dimension of X, is m if there exist
m linearly independent vectors in X but any collection of m + 1 elements of
X is linearly dependent. If there exists m linearly independent vectors for any
positive integer m, then we say dim (X) = ∞.
Proposition 2.2. The elements {x1 , x2 , . . . , xm } form a basis for
span ({x1 , x2 , . . . , xm }) if and only if they are linearly independent.
Proposition 2.3. The dimension of X is the number of vectors in any basis of X.

The proof of both of these propositions is left for the exercises.
Example 2.8. RN or CN has dimension N. We will denote by ej the standard unit
vector with a one in the jth position and zero elsewhere. Then {e1 , e2 , . . . , eN }
will be referred to as the standard basis for either RN or CN . 2
Example 2.9. In the vector space C([a, b]) the elements fj (t) = tj−1 are
linearly independent (see Exercise 2.6), so that the dimension is ∞, as is the
dimension of the subspace P. Also evidently the subspace PN is of dimension
N + 1. 2
Example 2.10. The set of solutions of the ordinary differential equation
u + u = 0 is precisely the set of linear combinations u(t) = λ1 sin t + λ2 cos t.
Since sin t, cos t are linearly independent functions, they form a basis for this
two-dimensional space. 2
The following is an interesting result, although not of great practical signif-
icance for us. Its proof, which is not at all obvious in the infinite dimensional
case, relies on the Axiom of Choice and will not be given here.
Theorem 2.1. Every vector space has a basis.
2.3 LINEAR TRANSFORMATIONS OF A VECTOR SPACE

If X and Y are vector spaces, a mapping T : X −→ Y is called linear if
T(λ1 x1 + λ2 x2 ) = λ1 T(x1 ) + λ2 T(x2 ) (2.3.4)
for all x1 , x2 ∈ X and all scalars λ1 , λ2 . Such a linear transformation is uniquely
determined on all of X by its action on any basis ofX, that is, if S = {xα }α∈A
is a basis of X and yα = T(xα ), then for any x = m j=1 λj xαj we have T(x) =
m
j=1 λ j yαj .
For a linear mapping it is common to omit parentheses and write Tx instead
of T(x), and we will always do so if it does not cause any confusion.
Vector Spaces Chapter | 2 33
If T is such a linear mapping and X and Y are both of finite dimension,

let us choose bases {x1 , x2 , . . . , xm }, {y1 , y2 , . . . , yn } of X, Y,
respectively. For
1 ≤ j ≤ m there must exist unique scalars akj such that Txj = nk=1 akj yk and it
follows that
m n m
x= λj xj ⇒ Tx = μk yk where μk = akj λj (2.3.5)
j=1 k=1 j=1
m
For a given basis {x1 , x2 , . . . , xm } of X, if x = j=1 λj xj we say that
λ1 , λ2 , . . . , λm are the coordinates of x with respect to the given basis. The
n × m matrix A = [akj ] thus maps the coordinates of x with respect to the basis
{x1 , x2 , . . . , xm } to the coordinates of Tx with respect to the basis {y1 , y2 , . . . , yn },
and therefore encodes all information about the linear mapping T.
If T : X −→ Y is linear, one-to-one and onto then we say T is an
isomorphism between X and Y, and the vector spaces X and Y are isomorphic
whenever there exists an isomorphism between them. If T is such an isomor-
phism, and S is a basis of X then it is easy to check that the image set T(S) is a
basis of Y. In particular, any two isomorphic vector spaces have the same finite
dimension or are both infinite dimensional.
For any linear mapping T : X → Y we define the kernel, or null space,
of T as
N(T) = {x ∈ X : Tx = 0} (2.3.6)
and the range of T as
R(T) = {y ∈ Y : y = Tx for some x ∈ X} (2.3.7)
It is immediate that N(T) and R(T) are subspaces of X, Y, respectively, and T
is an isomorphism precisely if N(T) = {0} and R(T) = Y. If X = Y = RN or
CN , we learn in linear algebra that these two conditions are equivalent, but this
is false in general.
2.4 EXERCISES
2.1. Using only the vector space axioms, show that the zero element in [V3]
is unique.
2.2. Prove Propositions 2.2 and 2.3.
2.3. Show that the intersection of any family of subspaces of a vector space
is also a subspace. Is the same true for the union of subspaces?
2.4. Show that Mm×n , the set of m × n matrices, with the usual definitions of
addition and scalar multiplication, is a vector space of dimension mn.
Show that the subset of symmetric matrices n × n matrices forms a
subspace of Mn×n . What is its dimension?
2.5. Under what conditions on a measurable set E ⊂ RN and p ∈ (0, ∞] will
it be true that C(E) is a subspace of Lp (E)? Under what conditions is
Lp (E) a subspace of Lq (E)?
2.6. Let uj (t) = tλj where λ1 , . . . , λn are arbitrary unequal real numbers.
Show that {u1 . . . un } are linearlyindependent functions on any interval
n
(a, b) ⊂ R. (Suggestion: If j=1 αj t
λj ≡ 0, divide by tλ1 and
differentiate.)
2.7. A side condition for a differential equation is homogeneous if when-
ever two functions satisfy the side condition then so does any linear
combination of the two functions. For example, the Dirichlet type
boundary condition u α = 0 for x ∈ ∂ is homogeneous. Now let
Lu = |α|≤m aα (x)D u denote any linear differential operator. Show
that the set of functions satisfying Lu = 0 and any homogeneous side
conditions is a vector space.
2.8. Consider the differential equation u + u = 0 on the interval (0, π ).
What is the dimension of the vector space of solutions which satisfy
the homogeneous boundary conditions (a) u(0) = u(π ), and (b) u(0) =
u(π ) = 0. Repeat the question if the interval (0, π ) is replaced by (0, 1)
and (0, 2π ).
2.9. Let Df = f for any differentiable function f on R. For any N ≥ 0 show
that D : PN → PN is linear and find its null space and range.
2.10. If X and Y are vector spaces, then the Cartesian product of X and Y, is
defined as the set of ordered pairs
X × Y = {(x, y) : x ∈ X, y ∈ Y} (2.4.8)
Addition and scalar multiplication on X × Y are defined in the natural
way,
(x, y) + ( y) = (x +
x, x, y +
y) λ(x, y) = (λx, λy) (2.4.9)
(a) Show that X × Y is a vector space.
(b) Show that R × R is isomorphic to R2 .
2.11. If X, Y are vector spaces of the same finite dimension, show X and Y are
isomorphic.
2.12. Show that Lp (0, 1) and Lp (a, b) are isomorphic, for any a, b ∈ R and
p ∈ (0, ∞].
Chapter 3
Metric Spaces
3.1 AXIOMS OF A METRIC SPACE

The idea of a metric space is that of a set on which some natural notion of
distance may be defined. The formal definition is as follows.
Definition 3.1. A metric space is a pair (X, d) where X is a set and d is a real
valued mapping on X × X, such that the following axioms hold.
[M1] d(x, y) ≥ 0 for all x, y ∈ X
[M2] d(x, y) = 0 if and only if x = y
[M3] d(x, y) = d(y, x) for all x, y ∈ X
[M4] d(x, y) ≤ d(x, z) + d(z, y) for all x, y, z ∈ X.
Here d is the metric on X, that is, d(x, y) is regarded as the distance from x
to y. Axiom [M4] is known as the triangle inequality. Although strictly speaking
the metric space is the pair (X, d) it is a common practice to refer to X itself as
being the metric space, if the metric d is understood from context. But as we will
see in examples it is often possible to assign different metrics to the same set X.
If (X, d) is a metric space and Y ⊂ X then it is clear that (Y, d) is also a
metric space, and in this case we say that Y inherits the metric of X.
Example 3.1. If X = RN then there are many choices of d for which (RN , d) is
a metric space. The most familiar is the ordinary Euclidean distance
⎛ ⎞1/2
N
d(x, y) = ⎝ |xj − yj |2 ⎠ (3.1.1)
j=1
In general we may define

⎛ ⎞1/p

N
dp (x, y) = ⎝ |xj − yj |p ⎠ 1≤p<∞ (3.1.2)
j=1
and
d∞ (x, y) = max (|x1 − y1 |, |x2 − y2 |, . . . , |xn − yn |) (3.1.3)
http://dx.doi.org/10.1016/B978-0-12-811426-1.00003-9
The verification that (Rn , dp ) is a metric space for 1 ≤ p ≤ ∞ is left to the

exercises—the triangle inequality is the only nontrivial step. The same family
of metrics may be used with X = CN . 2
Example 3.2. To assign a metric to C(E), the vector space of continuous
functions on E, more specific assumptions must be made about E. If we assume,
for example, that E is a closed and bounded1 subset of RN we may set
d∞ (f , g) = max |f (x) − g(x)| (3.1.4)
x∈E
so that d(f , g) is always finite by virtue of the well known theorem that a
continuous function achieves its maximum on such a set. Other possibilities are
1/p
dp (f , g) = |f (x) − g(x)|p dx 1≤p<∞ (3.1.5)
E
Note the analogy with the definition of dp in the case of RN or CN .

For more arbitrary sets E there is in general no natural metric for C(E). For
example, if E is an open set, none of the metrics dp can be used since there is no
reason why dp (f , g) should be finite for f , g ∈ C(E).
As in the case of vector spaces, some spaces of differentiable functions may
also be made into metric spaces. For this we will assume a bit more about E,
namely that E is the closure of a bounded open set O ⊂ RN , and in this case
will say that Dα f ∈ C(E) if the function Dα f defined in the usual pointwise
sense on O has a continuous extension to E. We then can define
Cm (E) = {f : Dα f ∈ C(E) whenever |α| ≤ m} (3.1.6)
with metric
d(f , g) = max max |Dα (f − g)(x)| (3.1.7)
|α|≤m x∈E
which may be easily checked to satisfy [M1]–[M4].

We cannot define a metric on C∞ (E) in the obvious way just by letting
m → ∞ in the above definition, since there is no reason why the resulting
maximum over m in Eq. (3.1.7) will be finite, even if f ∈ Cm (E) for every m.
See however Exercise 3.18. 2
Example 3.3. Recall that if E is a measurable subset of RN , we have defined
corresponding vector spaces Lp (E) for 0 < p ≤ ∞. To endow them with metric
space structure let
1/p
dp (f , g) = |f (x) − g(x)|p dx (3.1.8)
E
1. That is, E is compact in RN . Compactness is discussed in more detail below, and we avoid using
the term until then.
Another random document with
no related content on Scribd:
appropriate powers of the mind are weakened, while its moral faculty
is at the same time, and by the same means, greatly deteriorated.
An extraordinary but deplorable instance of this kind was exhibited

to the world by the justly celebrated astronomer Lalande, in his own
conduct and character, towards the concluding part of a long life.
These are so well portrayed in the very interesting Letters on France
and England, published in The American Review of History and
Politicks, that the writer of the present memoirs cannot forbear
presenting to his reader the following extract from Letter III.
“Lalande, if not the most profound and original, was certainly the
most learned astronomer of France, and the principal benefactor of
the science to which he was so passionately devoted. He was
remarkable for the most egregious vanity, and for the broadest
eccentricities of character, and almost equally eminent for the most
noble virtues of the heart. By a very singular perversion of intellect,
he became a professed atheist, about the commencement of the
revolution; pronounced, in the year 1793, in the Pantheon, a
discourse against the existence of a God, with the red cap upon his
head; and displayed, on this subject, the most absolute insanity,
during the rest of his life. This monstrous infatuation betrayed him
into the most whimsical acts of extravagance, and particularly into
the publication of a Dictionary of Atheists, in which he enregistered
not only many of “the illustrious dead,” but a great number of his
cotemporaries, and among these, some of the principal dignitaries of
the empire.
“This circumstance led to an occurrence in the Institute, which that

body will not soon forget. At an extraordinary sitting of all the
classes, convoked for the purpose, when Lalande was present, a
letter from the Emperor was announced and read aloud, in which it
was declared, that Mr. Lalande had fallen into a state of dotage, and
was forbidden to publish, thereafter, any thing under his own name.
The old astronomer rose very solemnly, bowed low, and replied, that
he would certainly obey the orders of his majesty. His atheistical
absurdities deserved, no doubt, to be repressed; but, besides the
singularity of this form of interdiction, there was an unnecessary
degree of severity in it, as the end might have been attained without
so public a humiliation. Lalande was notoriously superannuated, and
not therefore a fit object for this species of punishment. Some
consideration, moreover, was due to his many private virtues, to his
rank in the scientific world, and to the large additions which he had
made to the stock of human knowledge. His atheistical opinions
arose, not from any moral depravity, but from a positive alienation of
mind on religious topics. He was not the less conspicuous for the
most disinterested generosity; for warm feelings of humanity; for the
gentleness of his manners; for the soundness of his opinions on
questions of science, and for a certain magnanimity with regard to
the merits of his rivals and detractors. The extravagance of his
opinions and his manners during his dotage, rendered him an object
of almost universal derision in Paris, and subjected him to the most
cruel and indecent mockery. It became fashionable, even among
those who had derived their knowledge from his lessons and
experienced his bounty, to depreciate his merits both as an
astronomer and as a man. Lalande had the misfortune of living to
see a maxim verified in his own regard, which has been exemplified
in every age and country, that some disciples may become superiour
to their masters. But he was, nevertheless, at all times among the
luminaries of science; and to him astronomy was indebted for more
substantial and unremitted services, than to any one of his
cotemporaries.”
This very Mr. Lalande, in the preface to the third edition of his
inestimable work entitled Astronomie, published at Paris so late as
the year 1792, shews, that astronomy furnishes most powerful
proofs of the being of a God. Yet this same man, in one year after,
when in his “dotage,” with a mind enfeebled by age, and corrupted
by the delusions of the new philosophy of his countrymen, became
an object of “derision,” and of “mockery,” even among Frenchmen;
for his absurdities, and his endeavours to set himself up as a
champion of atheism! Is it necessary to furnish the rational part of
mankind with a more striking, and at the same time a more
lamentable proof, of the deleterious effects produced by those
illusions, which, under the assumed name of “Philosophy,” have
been conjured up by some modern Theorists and Political
Speculators? Certainly, it is not. The instance, here adduced, may
stand as a monument of the folly and depravity of the Philosophy of
the Gallican School.
335. “If,” (says a late anonymous writer,) “from the advantages of
sound learning to the state, we turn to its influence on the characters
of individuals, we will find its effects to be no less striking. We will
find, that although, without much learning, man may become useful
and respectable, yet that he cannot, without it, become polished,
enlightened and great; he cannot ascend to that grade in the scale of
his Creator’s works, to which his powers are intended to exalt him. If
to this rule, a Franklin, a Rittenhouse, and a Washington present
exceptions, they are to regarded as mere exceptions, and therefore
do not amount to an infraction of the rule. They were prodigies;
which necessarily implies a departure from, and an ascendency over
common principles.” See an Account of Dickinson College, Carlisle,
in the Port Folio, for March, 1811; supposed to be written by
Professor Cooper.
336. Three years before Dr. Rush expressed these opinions, so

generally unfavourable to classical learning and an academic
education, he seems to have thought these necessary qualifications
for a physician at least. In his Lecture on the Character of Dr.
Sydenham, delivered in Dec. 1793, is this passage: “From the short
records of his life, which have been published by the different editors
of his works, it appears that his education in academical learning and
medicine, was perfectly regular. He became a scholar at Oxford, and
a doctor of medicine at the university of Cambridge. I mention these
facts,” adds our learned Professor, “in order to refute an opinion
which has been introduced by some lazy and illiterate practitioners of
physic, that he was indebted wholly to intuition for all his knowledge
of medicine. Men may become wise and distinguished by meditation
or observation, in the science of morals and religion; but education
and study are absolutely necessary to constitute a great physician.”
With all due deference to the abilities and judgment of the

Professor, the Memorialist presumes, that if “education and study are
absolutely necessary to constitute a great physician,” they are
equally requisite in the formation of a great astronomer: because a
knowledge of geometry and optics can no more be attained by
intuition, than that of anatomy and the materia medica; yet these
sciences are, respectively, indispensable in the formation of the two
characters, to which they severally relate.
Still, adds Dr. Rush, “It is true Dr. Sydenham did not adopt, or
follow, the errors of the schools in which he had been educated; but,
by knowing them thoroughly, he was able, more easily, to examine
and refute them.” Here, then, is an admission, that even an intimate
knowledge of such errors is eminently useful, by enabling a man of a
sound and cultivated mind to refute them: for, the refutation of
existing errors, affords a most important aid to the advancement of
true science.
Sydenham, it appears, received his collegiate education at both

the English universities. It may not therefore be improper, on this
occasion, to introduce a quotation from an invaluable elementary
work;[336a] in order to shew, what was the opinion entertained by a
learned and distinguished German, of the English Universities—on
the models of which, the higher seminaries of learning in the United
States are formed. “Of all the Universities of Europe,” says Baron
Bielfeld, “those of Oxford and Cambridge in England appear at
present to approach the nearest to perfection: The great men they
produce, are a better proof than any other argument.[336b] We could
wish,” adds this highly enlightened foreigner, “always to see an
university a real city of learning, a place consecrated entirely to the
muses and their disciples; that the Greek and Latin languages were
there predominant; and that every thing were banished from thence,
which could cause the least dissipation in those who devote
themselves to letters.” “The man who confines himself to his
closet,”—says our author, in another place,—“is but rarely visited by
the sciences, the arts and the belles lettres: to acquire their intimate
acquaintance, he must seek them in those places where Minerva,
Pallas, Apollo, and the Muses, have fixed their residence. Emulation,
that strong impulse in the career of all our pursuits, should constantly
attend the man of letters from his early youth to the last period of his
life; in the school, at college, at the university, in those employments
to which his knowledge may lead him, or in those academies of
science to which he may be admitted. Emulation is an animating
faculty, that results from society: and few there are, to whom nature
has given a genius sufficiently strong to attain an extensive erudition
in solitude; who are provided with wings that can bear them, without
guides, without models, without companions or supports, to the lofty
regions of the empyrean.”
336a. The Elements of Universal Erudition, containing an analytical

abridgment of the Sciences, Polite Arts, and Belles Lettres; by Baron Bielfeld. In
three 8vo. volumes; translated from a Berlin edition, by W. Hooper, M. D. and
printed in London, in the year 1770.
336b. The three great Universities of England and Ireland enjoy the right, in
addition to many other important privileges, of sending, each, two members to
represent them in parliament, Would to heaven! that there were something like a
representation of the interests of learning and science, in the legislative bodies of
our own country.
337. Bacon (the celebrated Viscount of St. Albans and Baron of

Verulam) published his great philosophical work, the Novum
Organum, in the year 1620. The learned and sagacious professor
Cooper remarks, that “Lord Bacon” (whom the honourable Mr.
Walpole considers as the Prophet of the Arts, which Newton came
to reveal,) “was the first among the moderns, who pointed out the
way by which real knowledge was to be obtained, and turned the
minds of the learned from playing tricks with syllogisms, and the
legerdemain of words without ideas; and taught them to rest theory
upon the basis of experiment alone.” See the Introductory Lecture of
Thomas Cooper, Esq. Professor of Chemistry at Carlisle College,
Pennsylvania.
338. The Greek and Latin are called by way of pre-eminence, the
learned languages. Baron Bielfeld enumerates the advantages
resulting from a knowledge of the former; among which he notices
that important one, of its enabling us more readily and clearly to
comprehend the meaning of that almost boundless list of terms in
the arts and sciences, used in modern languages and styled
technical, which are either altogether Grecian, or derived from that
language. He then makes this remark: “From all that has been said,
it is apparent how much utility attends the study of the Greek tongue;
and how much reason the English have, for applying themselves to
it, from their early youth.” “But,” observes this learned and
discriminating writer, “that which has given the Latin an advantage
over the Greek itself, that has rendered it indispensable to every
man of letters, and has made it the basis of erudition, is, that during
the middle age, and in general in all modern times, the learned of all
Europe have made it their common and universal language; so that
the Latin forms, if we may use the expression, the natural language
of the sciences.” Elem. of Univ. Erud.
339. Although Mr. T. Cooper (before quoted) admits, that the “strict
adherence to the syllogistic mode of reasoning,” that which he calls
“playing tricks with syllogisms,” together with “the legerdemain of
words without ideas,” was carried much too far by some late
metaphysical writers of eminence; yet he is of opinion, that “in
modern times, this invention of Aristotle is abandoned more than it
deserves to be: For,” continues Mr. Cooper, “no man can so skilfully
analyse the argument of another, as one who is well acquainted with
the rules of scholastic logic, and accustomed to apply them. Good
reasoners there are and will be, who know nothing of these rules, but
better reasoners who do.”
Mr. Cooper doubts, whether metaphysical lectures should be

delivered, at all, in colleges; but thinks, that if metaphysics were to
be there taught, the writings of Beattie, Oswald and Gregory, would
be unworthy of notice. Much as the Writer of these Memoirs respects
the talents and ingenuity of the learned Professor of Chemistry, he
can by no means concur in this opinion: and he regrets, that he feels
himself obliged to differ still more widely, from a gentleman of such
acknowledged abilities, respecting the propriety of his
recommending to youth the study of the works of Hobbes, Leibnitz
and Collins.
Now, what the complexion and tendency of the tenets of Hobbes,

Leibnitz, and other philosophers of the same class are, may be
learnt from the following passages, translated from a French work,
entitled, “De la Philosophie de la Nature, ou Traité de Morale pour
l’Epece Humaine, tire de la Philosophie et fonde sur la Nature;” a
work which, though anonymous in respect to its author, had passed
through three editions in the year 1777. The writer thus says:
“Of what importance to me are the names of Carneades, of

Lysander, of Hobbes, and the author of The System of Nature,
names unhappily celebrated, which the apostle of the moral
indifference of human actions alleges in favour of this atrocious
extravagance?” (the doctrines of Fatality, Moral Scepticism, &c.)
“Carneades was an arrogant Pyrrhonian, who doubted of every
thing, excepting the superiority of his own logic. Hobbes had the
audacity to write a book against the everlasting truths of geometry.
Lysander, the enemy of the liberty of Sparta, and the corrupter of the
oracles of Delos and Ammon, was one of those spirits of spleen and
filth, who strive to acquire a name by reducing wickedness to a
system. As for the anonymous Writer, whose licentious pen vents so
much blasphemy on Nature, in disavowing the existence of God, he
has purchased the right to deny that of Morality. He is equally silly
with Salmonius, in braving the thunderbolt destined to stifle the
stings of conscience.” Speaking of Leibnitz, in another place, this
French Moralist observes, that “the Philosopher of Leipsick made of
the soul a monad, and explained all the phænomena of its union with
matter by a pre-established harmony. One portion of Europe
believed him; because he set up a new system! and what is it but a
metaphysical theory, without system?” And again: “What names
have we to oppose to those of Descartes, Leibnitz, Pascal and
Malbranch? The suffrage of Newton, alone, is sufficient to crush
their Materialism; if, in the humble materials for the examination of
human reason, the suffrage of one great man is competent to
balance a syllogism.”
340. The professorships, all well supported and endowed, which

are established at Oxford and Cambridge, (and, probably, there are
similar institutions in the universities of Scotland and Ireland,) are in
the following departments of literature and science: viz. Divinity,
Hebrew, Arabic, Greek, Modern Languages, History (general,)
Modern History, Civil Law, Common Law, Physic, Anatomy, Botany,
Chemistry, Natural Philosophy, Experimental Philosophy, Astronomy,
Mathematics, Geometry, Moral Philosophy, Casuistry, Music.
341. See the editorial review, in that work, of an “Historical Report

upon the progress of History and ancient Literature, since the year
1789, and upon their actual condition,” &c. vol. iii. No 1.
342. In the year 1789, Dr. Rittenhouse translated from the German
of Mr. Lessing, director of the theatre at Hamburg, a tragedy called
Lucia Sampson; which translation was printed; in the same year, by
Mr. Charles Cist, of Philadelphia. In the preface to it, the translator
says:—“This translation was attempted at the request of a friend;
and the many virtuous sentiments and excellent lessons of morality it
contains, will apologize for its being offered to the public. To young
ladies it may afford useful instruction, and will, from the nature of the
distress, be particularly useful to them: an elegant writer well
acquainted with the human heart, has observed, that the affection of
a father to his daughter unites extreme sensibility with the utmost
delicacy; and this sentiment is, no doubt, in a great degree
reciprocal.”
343. See Dr. Rush’s Eulog. on Ritt.
344. The memorialist undertakes to say, on the authority of his

father (the late Rev. Mr. Barton,) that our philosopher was sufficiently
well versed in the Latin, to have read Newton’s Principia in that
language, besides studying it in his native tongue: and further, that,
although he was very imperfectly acquainted with the grammatical
construction of the Greek language, he had so far familiarized
himself to a knowledge of its written characters and words, as
enabled him to consult a lexicon; which he frequently did, for the
purpose of ascertaining the true etymology of many of those
technical terms, derived from the Greek, that are in common use in
our language, particularly in relation to his favourite sciences.
345. In Hill’s Life of Dr. Barrow, it is remarked, that this great

Mathematician (as well as learned Divine) “was always addicted to
poetry, and very much valued that part of it which consists of
description.” In like manner, Dr. Rittenhouse delighted in poetic
effusions of genius and science. His Eulogist observes, that “the
muse of Thomson charmed him most:” indeed, an astronomer, and a
man of virtue and taste, could not but be charmed by the chaste and
glowing descriptions of that fascinating poet, blended, as they are,
with philosophical reflections. Our philosopher, however, greatly
admired Milton also: so that these two celebrated votaries of the
muses seemed to be his favourites. Why should not these partialities
of Rittenhouse be noticed?—when similar observations have been
made respecting the characters of other men, eminent in science;
as, for example, that the favourite author of Erasmus and the
younger Scaliger, was Terence, and that Grotius was an admirer of
Terence, Lucan and Horace.
346. Nec lusisse pudet, is an observation which has, in particular

instances, been applied to the occasional conduct and disposition of
some of the wisest, best, and even gravest characters. Dr. Warton, in
remarking on this line of Mr. Pope, viz.
“Unthought-of frailties cheat us in the wise,”—
says; “Who could imagine that Locke was fond of romances; that
Newton once studied astrology; that Dr. Clarke valued himself for his
agility, and frequently amused himself, in a private room of his
house, in leaping over the tables and chairs; and that our author
himself (Mr Pope) was a great epicure.”
In our own country, the sage Franklin abounded in anecdote and

humour, and thought it not unwise to recreate his mind, at times, with
the game of chess: the conversation of Judge Hopkinson was
replete with sprightly wit, and he admired well written novels of no
immoral tendency; as did also the late Judge Wilson: the illustrious
Washington, in his earlier years, enjoyed the pleasures of the festive
board, in the society of men of understanding and worth: and no man
delighted more in cheerful conversation, and in reading works of
fancy and taste, than the philosophic Rittenhouse. The almost
universal tendency, in persons of all classes, to an occasional
playfulness of temper, even in cases which may sometimes be
considered as bordering on weakness, has given the force of a
maxim to the observation of the latinists—Nemo omnibus horis sapit.
Indeed, as a biographer of the celebrated Dr. Clarke has remarked,
“to be capable of drawing amusement from trivial circumstances,
indicates a heart at ease, and may generally be regarded as the
concomitant of virtue.”
348. Sir Isaac Newton, it is well known, was thoroughly persuaded

of the Truth of Revelation: yet he did not escape the imputation of
being an Arian, Mr. Whiston having represented him as such. It is
equally a matter of notoriety, that similar opinions have prevailed
respecting Dr. Rittenhouse’s religious creed: nay, further, that doubts
were entertained by some, whether he believed at all in the
fundamental principles of the Christian religion. In one instance,
indeed, a virulent party-writer[348a] had the hardiness, one might say
folly, to proclaim him an “Atheist!” The publication in which this false
and shameful accusation was made, appeared about the time of Dr.
Rittenhouse’s death, and, it is believed, shortly after that event.
As a Biographer of such a man as Rittenhouse, the Author of

these Memoirs would do great injustice to his memory, did he not lay
before his readers, in a full and undisguised manner, that sort of
testimony concerning our Philosopher’s religious sentiments, which it
is presumed will eradicate every doubt or suspicion, that has
heretofore existed in the minds of some, on the subject. He is aware
of the influence, which the opinions of eminently wise and good men
(or, of such sentiments as are sometimes attributed to them,) have,
in their operation on society; and, in every point of view, he fully
estimates the importance of representing them to the world, in a
strict conformity to truth.
These considerations have induced the Memorialist to devote a

larger portion of his work to an elucidation of Dr. Rittenhouse’s real
opinions on the all-important subject of Religion, than he should
have thought proper, under other circumstances, to appropriate to
that part of his character.
Under these impressions, then, the Memorialist could not think it

consistent with his duty, to withhold from the public a letter
addressed to him by the Rev. Mr. Cathcart, a clergyman of much
respectability and pastor of a presbyterian congregation in the
borough of York. This letter (which will be found in the Appendix)
contains what may be fairly deemed conclusive evidence, even if
such had been before wanting, that Dr. Rittenhouse was “a firm
Believer in Christianity.” Bishop White had communicated to the
Memorialist, in conversation, the interesting facts stated in Mr.
Cathcart’s letter; the knowledge of which, the Bishop had derived,
verbally, from that gentleman: his letter was written in answer to one
which the Memorialist addressed to him, on the occasion, at the
instance of the Right Rev. Prelate.
348a. Mr. William Cobbett.
349. “Astronomy, like the Christian religion, if you will allow me the
comparison,” said our Philosopher, “has a much greater influence on
our knowledge in general, and perhaps on our manners too, than is
commonly imagined. Though but few men are its particular votaries,
yet the light it affords is universally diffused among us; and it is
difficult for us to divest ourselves of its influence so far, as to frame
any competent idea of what would be our situation without it.” See
Ritt. Orat.
In another part of his Oration, is this passage: “Our Religion

teaches us what Philosophy could not have taught: and we ought to
admire, with reverence, the great things it has pleased Divine
Providence to perform, beyond the ordinary course of nature, for
man, who is, undoubtedly, the most noble inhabitant of this globe:”
&c.
And in addition to these sentiments, uttered and published by our

Philosopher himself, let the testimony of Dr. Rush, who had long and
intimately known him, be quoted from the learned Professor’s
Eulogium. “He believed in the Christian Revelation,” says the Doctor:
and then subjoins, “Of this he gave many proofs; not only in the
conformity of his life to the precepts of the Gospel, but in his letters
and conversation.”
350. The Rev. Ashbel Green, D. D. This gentleman succeeded the

Rev. Dr. Sproat, an aged clergyman, of amiable disposition and
unaffected piety, for whose character our Philosopher entertained a
great esteem, and, during the latter part of whose ministry in that
church, he first attended it. Dr. Green has lately become President of
the College of New-Jersey, in consequence of the resignation of the
learned and eloquent Samuel Stanhope Smith, D. D.
351. The following extract of a letter, which Professor Rusk was so

obliging as to address to the writer of these memoirs, in the spring of
the year 1812, in answer to some questions proposed by the
memorialist, favours the presumption, that our philosopher in some
points dissented from the opinions of very respectable Calvinistic
Divines, on the subject of religion. “I understood from the Rev. Dr.
Green,” says the learned Professor, that his late colleague, the Rev.
Dr. Sproat, had informed him, that in a visit he once paid to Dr.
Rittenhouse, they were led accidentally to converse upon a religious
subject, on which they held different opinions. Dr. Sproat, in
defending his opinions, quoted several texts of scripture; but
observed, after doing so; “Perhaps, Mr. Rittenhouse, you do not
admit of the validity of arguments derived from the bible.” “Pardon
me, Sir,” (said Mr. Rittenhouse,) “I admit the divine authority of the
contents of that book.” Another fact stated by Dr. Rush, at the same
time, and which was also communicated to the memorialist, by a
very near and dear friend of the deceased, is thus related by the
Doctor: “His late worthy companion, Mrs. Rittenhouse, informed me,
that the last sourse from whence he derived intellectual and moral
pleasure, was Dr. Price’s excellent sermon upon the Goodness of
God, which she read to him, at his request, on the two successive
days before he died.” It may not be thought unworthy of being
remarked on this occasion, that Mr. T. Dobson, of Philadelphia,
republished Price’s Sermons, in the year 1788, and that Mrs.
Rittenhouse’s name appears in the list of subscribers to that edition.
In Dr. Rush’s letter, just quoted, he introduces the subject in these

terms. “In answer to your question, relative to the religious opinions
of your late uncle and my excellent friend, Dr. Rittenhouse, I am
happy in being able to inform you, that I have no doubt of his having
been a sincere believer in the most essential doctrines of the
Christian religion: the ground upon which I formed this opinion, were
derived not only from many incidental remarks in its favour, that fell
from him in our conversations upon other subjects, but from the
testimony of persons upon whose correctness I have the fullest
reliance.”
Upon the whole it appears, that although our philosopher was,

most probably, not strictly Calvinistical in his religious creed, he was
nevertheless a pious man, and a sincere Christian in the
fundamental articles of his faith.
352. Dr. Rittenhouse had no more faith in the notion entertained by

some visionary men, of the attainment of the perfection of virtue, in
this life, than he had in the fantastic opinion, maintained also by
some, of the perfectibility of human reason. He supposed that we are
capable, by a progressive “enlargement of our faculties,” to “advance
towards the perfection of the Divinity;” not like those pretenders to
philosophy, who, as Mr. Voltaire expresses it, “took it into their
heads, by the example of Descartes, to put themselves into God’s
place, and create a world with a word!” Our philosopher knew, that
pure virtue and perfect reason do not belong to human nature.
353. Dr. Rush’s Eulogium.
354. Ibid.
355. This quotation and the other passages, before which inverted
commas are placed in the margin, in the two last paragraphs of the
text, are extracted from Dr. Rittenhouse’s Oration.
356. Dr. Thomas Newton, Bishop of Bristol.
357. A late learned philosopher and eloquent divine, after

adverting to the irrational and infatuated notions of men of the class
above referred to, contrasted with doctrines founded in truth, and the
awful gloom, destitute of every ray of consolation, that must
necessarily accompany their reflections upon their own principles,
addresses to them this short but serious invocation: “When these
things are fairly weighed, as in nature they exist, I call on you, nay I
challenge you, ye boasting philosophists! to comfort yourselves, and
be easy under your dreary doctrine, or notion of being safe after
death, in a state of annihilation or future nothingness! I call on you,
ye wise Illuminati! of upstart name, to weigh these things seriously;
and try whether you can comfort yourselves, and remain easy, in
considering, and striving to make others consider, Death, as only an
“everlasting Sleep,” from which they will never be awakened, nor
their ashes disturbed!” See Sermon V. in The Works of William
Smith, D. D. late Provost of the College and Academy of
Philadelphia.
In no instance have the impious and absurd doctrines of the

“Philosophists” and the “Illuminati,” of our times, been carried to such
a height of extravagance, as by the revolutionists of modern France.
These infatuated people undertook, in the year 1793, to abolish by
Law, a Futurity of Existence; having then decreed, that no such state
existed! They also decreed, that in every cemetery there should be
erected a figure representing Sleep, pointing towards the tombs; and
this Sleep of Death, the decree declared to be eternal!! It is to this
sort of wickedness and folly that an allusion is made, in the foregoing
quotation; as well as in the following lines, copied from the Pursuits
of Literature:
“Systems which laugh to scorn th’avenging rod,

And hurl defiance at the throne of God;
Shake pestilence abroad with madd’ning sweep,
And grant no pause—but everlasting Sleep!”
358.
“Let others creep by timid steps, and slow,
On plain experience, lay foundations low;
By common sense, to common knowledge bred;
And lost to nature’s cause through nature led:
All-seeing in thy mists, we want no guide,
Mother of Arrogance and source of pride!
We nobly take the high priori road,
And reason downward, till we doubt of God.”
Pope’s Dunciad, b. IV. l. 455.
The following observation, in the form of a note, is referred to, from

the lines above quoted, in a work which contains that extract, viz.
“Those, who, from the effects in this visible world, deduce the eternal
power and Godhead of the first cause, though they cannot attain to
an adequate idea of the Deity, yet discover so much of him, as
enables them to see the end of their creation and the means of their
happiness: whereas they who take “the high priori road,” as Hobbes,
Spinoza, Descartes, and some better reasoners, for one that goes
right, ten lose themselves in “mists,” or ramble after visions, which
deprive them of all sight of their end, and mislead them in the choice
of wrong means.”
Mr. Pope had put the above poetical lines into the mouth of one of
his Dunces, when addressing himself to the goddess Dullness. And
as the great Dr. Samuel Clarke had previously endeavoured to shew,
[358a]
that the Being of a God may be demonstrated by arguments
deduced â priori, the Doctor conceived himself to be struck at,
among those “better reasoners” alluded to, in the note above
mentioned.
358a. In his work entitled, “A Discourse concerning the Being and Attributes of
a God, the Obligations of Natural Religion, and the truth and certainty of the
Christian Revelation; in answer to Mr. Hobbes, Spinoza, the Author of the Oracles
of Reason, and other deniers of natural and revealed Religion.”
359. See Ritt. Orat.
360. Ibid.
361. “Other systems of Philosophy have ever found it necessary to
conceal their weakness and inconsistency, under the veil of
unintelligible terms and phrases, to which no two mortals, perhaps,
ever affixed the same meaning. But the philosophy of Newton
disdains to make use of such subterfuges; it is not reduced to the
necessity of using them, because it pretends not to be of nature’s
privy council, or to have access to her most inscrutable mysteries;
but, to attend carefully to her works, to discover the immediate
causes of visible effects, to trace those causes to others more
general and simple, advancing by slow and sure steps towards the
Great First Cause of all things.” Ritt. Orat.
362. Dr. Morse, the Geographer.
363. See his Life of George Washington.
364. See the Eulogium on Rittenhouse.
365. The names ordinarily used to distinguish things, do not

always truly denote the nature of the things they are designed to
signify: and it is very evident, that any misapplication of a name, to
which a specific meaning has been appropriated, cannot alter or
otherwise affect the essence or inherent quality of the thing itself to
which it is wrongly applied.
A nation may be a republic, notwithstanding its chief executive

magistrate be denominated a king. A kingly government may be
essentially republican, provided the people be governed by known
laws, and their king be limited in his prerogative, by the constitution
of the state; not such a monarch as is vested with uncontrouled
power. In this sense, the British government may, as some modern
writers have shewn, be called a commonwealth, or republic: and
under a similar impression, Sir Thomas Smith, even in the reign of
so rigid a prince as Henry VIII. wrote his book De Republicâ
Anglicanâ. The republic of Poland was long governed by elective
kings; and Shakespeare, (nay, even the leveller Godwin,[365a])
appears to have considered Monarch, King and President, as
synonymous terms.
365a. The Memorialist can truly say, with the author of the Pursuits of
Literature:—“I have given some attention to Mr. Godwin’s work on Political Justice,
as conceiving it to be the code of improved modern ethics, morality, and
legislation. I confess I looked not for the Republic of Plato, or even for the Oceana
of Harrington; but for something different from them all. I looked, indeed, for a
superstructure raised on the revolutionary ground of Equality, watered with the
Guillotine; and such I found it.” See Pursuits of Literature, Dial. the third, note p. of
the seventh Lond. edit.
366. “It belongs to monarchies,” says Dr. Rush, “to limit the
business of government to a privileged order of men.” See Eulog.
367. See Ritt. Orat. before the Am. Philos. Soc. in 1775.
368. See the ordaining clause of the Constitution of the United

States.
369. Mr. Pope was not singular in the opinion here expressed: one
of the most illustrious legislators and best practical statesmen the
world has ever known, appears to have entertained the same
sentiment, when he penned the following passages: they are
extracted from the Frame of Government originally designed by
William Penn, for Pennsylvania: published in the year 1682.
“Any government is free to the people under it (whatever be the

frame,) where the laws rule and the people are a party to those laws;
and more than this is tyranny, oligarchy, and confusion.”
“There is hardly one frame of government in the world so ill

designed by its first founders, that, in good hands, would not do well
enough; and story tells us, the best, in ill ones, can do nothing that is
great or good.” “I know,” continues Penn, “some say, Let us have
good laws, and no matter for the men that execute them: but let
them consider, that though good laws do well, good men do better:
for good laws may want good men, and be abolished or evaded by ill
men; but good men will never want good laws, nor suffer ill ones. It is
here, good laws have some awe upon ill ministers; but that is where
they have not power to escape or abolish them, and the people are
generally wise and good: but a loose and depraved people (which is
to be the question) love laws and an administration like themselves.
That, therefore, which makes a good constitution, must keep it, viz.
men of wisdom and virtue; qualities that, because they descend not
with worldly inheritances, must be carefully propagated by a virtuous
education of youth.”
371. About the middle of January, 1813, the Memorialist passed a

very pleasant evening, in company with an agreeable party of
friends, at the house of Dr. Rush. Among various subjects, which
were then discussed with much ingenuity and good humour,
Redhefer’s pretended discovery of what is called the Perpetual
Motion, a thing which had then, very recently, attracted a good deal
of the public attention, was brought upon the tapes: when Dr. Rush,
addressing himself to the Writer, who had just expressed his opinion
decidedly against the projector’s theory, as being utterly incompatible
with established principles of physics and well-known laws of the
material world, said quite emphatically; “Sir, I entirely agree with you:
and let me observe, there are four things, concerning which I have
always been completely sceptical, as I am sure your good uncle[371a]
also was; that is to say, the perfectibility of human reason; the
possibility of transmuting base metals into silver and gold; a
panacea, in the healing art; and a power, in any mortal, to give
perpetuity of motion to matter.” These were, substantially, the
sentiments expressed by Dr. Rush, on the occasion; and the Writer
believes he is pretty accurate in his recollection of the very words
which the Doctor used.
371a. Dr. Rittenhouse.
372. See Notes on Virginia.
373. All the boundary-lines, mentioned above, were determined by

astronomical observations. The manner in which the work was
performed, with an account of the instruments used on those
occasions, will be found in the fourth volume of the Transactions of
the American Philosophical Society. Some of Dr. Rittenhouse’s
associates, in those arduous undertakings, were men of high
reputation in the same departments of science; but his talents were
principally relied on.
374. See Notes on Virginia.
375. Ibid.
376. It will, perhaps, have occurred to the reader, that besides

such of the WORKS of Dr. Rittenhouse, as are referred to in the text, in
some of which, the blended effects of genius, philosophical science
and mechanical skill, were equally conspicuous, he put the Mint into
operation. In the language of his worthy successor in the direction of
that institution, “his lofty and correct mind, capable alike of ascending
to the sublimest heights of science, and of condescending to
regulate the minute movements of mechanical machinery, organized
the Mint, and created the workmen and the apparatus.” His agency
in directing the construction, and arranging the operative
departments, of this important establishment, though less indicative
of extraordinary mechanical genius than many of his other works,
was nevertheless an arduous undertaking: it was conducted, as Mr.
De Saussure very justly observed, “amidst complicated difficulties,
from which the most persevering minds might have shrunk without
dishonour.”
377. Dr. Rush, in his Eulogium on Rittenhouse, has introduced a

short invocation, which aptly applies in this place: it is in these
words; “Come, and learn by his example to be good, as well as
great. His virtues furnish the most shining models for your imitation;
for they were never obscured by a single cloud of weakness or vice.”
378. Mr. Chief Justice Marshall makes an observation, in

reference to General Washington, which applies with equal force to
Dr. Rittenhouse. “To estimate rightly his worth, we must contemplate
his difficulties: we must examine the means placed in his hands, and
the use he made of those means.” Pref. to Marshall’s Life of
Washington.

(Download PDF) Techniques of Functional Analysis For Differential and Integral Equations 1St Edition Paul Sacks Full Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Download PDF) Techniques of Functional Analysis For Differential and Integral Equations 1St Edition Paul Sacks Full Chapter PDF

Uploaded by

Copyright:

Available Formats

Techniques of Functional Analysis for

Differential and Integral Equations 1st

Advanced Differential Equations 1st Edition Youssef

Schaum's Outline of Differential Equations, Fifth

Transforms and Partial Differential Equations T.

Proper orthogonal decomposition methods for partial

Mechanics and Physics of Structured Media : Asymptotic

Differential Equations and Boundary Value Problems:

Integral Equations for Real-Life Multiscale

Linear Algebra and Partial Differential Equations T

© 2017 Elsevier Inc. All rights reserved.

Library of Congress Cataloging-in-Publication Data

British Library Cataloguing-in-Publication Data

For information on all Academic Press publications

Publisher: Candice Janco

Typeset by SPi Global, India

1. Some Basic Discussion of Differential and Integral

5.1 Axioms of an Inner Product Space 59

8. Distributions and Differential Equations 125

8.1 Weak Derivatives and Sobolev Spaces 125

8.4 Fundamental Solutions and the Fourier Transform 132

9. Linear Operators 141

9.1 linear Mappings Between Banach Spaces 141

10. Unbounded Operators 159

11. Spectrum of an Operator 171

12. Compact Operators 1 83

13. Spectra and Green's Functions for Differential

13.1 Green's Functions for Second-Order ODEs 205

14. Further Study of Integral Equations 227

15. Variational Methods 243

15.1 The Dirichlet Quotient 243

16. Weak Solutions of Partial Differential Equations 269

The purpose of this book is to provide a textbook option for a course in

Some Basic Discussion of

In this chapter we will discuss “standard problems” in the theory of ordinary

1.1 ORDINARY DIFFERENTIAL EQUATIONS

Techniques of Functional Analysis for Differential and Integral Equations

The ODE is linear if it can be written as

for some coefficients a0 , . . . , an , g, and homogeneous if also g(t) ≡ 0. It

By standard calculus properties L is a linear operator, meaning that

L(c1 u1 + c2 u2 ) = c1 Lu1 + c2 Lu2 (1.1.6)

for any scalars c1 , c2 and any n times differentiable functions u1 , u2 .

1.1.1 Initial Value Problems

u(j) (t0 ) = γj j = 0, 1, . . . , n − 1 (1.1.7)

Theorem 1.1. Assume that

which is only a valid solution for t < 1

1.1.2 Boundary Value Problems

1.1.3 Some Exactly Solvable Cases

We therefore obtain as many linearly independent solutions as there are

1.2 INTEGRAL EQUATIONS

Observe that Eq. (1.2.30) seems to require differentiability of f even though

1.3 PARTIAL DIFFERENTIAL EQUATIONS

α = (α1 , α2 , . . . , αN ) αi ∈ {0, 1, . . . } (1.3.37)

for some coefficient functions aα .

1.3.1 First Order PDEs and the Method of Characteristics

u(x1 , x2 ) = f (x2 ) (1.3.43)

The collection of lines parallel to θ , that is, lines ax2 − bx1 = C

x = f (s), y = g(s), s0 < s < s1 (1.3.48)

The Cauchy problem for Eq. (1.3.45) consists in finding a solution of

We must then have

continuously differentiable. The three relations x = x(t, s), y = y(t, s), z =

Example 1.10. Let denote the x axis and let us solve

with u = h on . Introducing f (s) = s, g(s) = 0 as the parametrization of , we

with αδ − βγ = 0, so that the transformation is invertible. Our goal is to make a

+ λ = 0 r2 R + rR − λR = 0 (1.3.104)

(0) = (2π ) (0) = (2π ) (1.3.105)