Download the full version of the ebook at ebookname.

com

Ordinary Differential Equations Applications

Models and Computing 1st Edition Charles Roberts

https://ebookname.com/product/ordinary-differential-
equations-applications-models-and-computing-1st-edition-
charles-roberts/

OR CLICK BUTTON

DOWNLOAD EBOOK

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

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

Handbook of Differential Equations Ordinary Differential

Equations 1st Edition A. Canada

https://ebookname.com/product/handbook-of-differential-equations-
ordinary-differential-equations-1st-edition-a-canada/

ebookname.com

Handbook of Differential Equations Ordinary Differential

Equations Volume 2 1st Edition A. Cañada

https://ebookname.com/product/handbook-of-differential-equations-
ordinary-differential-equations-volume-2-1st-edition-a-canada/

ebookname.com

Handbook of Differential Equations Ordinary Differential

Equations Volume 2 1st Edition A. Cañada

https://ebookname.com/product/handbook-of-differential-equations-
ordinary-differential-equations-volume-2-1st-edition-a-canada-2/

ebookname.com

Teaching ESL Composition Purpose Process and Practice 2nd

Edition Dana R. Ferris

https://ebookname.com/product/teaching-esl-composition-purpose-
process-and-practice-2nd-edition-dana-r-ferris/

ebookname.com
Cultures of Disaster Society and Natural Hazards in the
Philippines 1st Edition Greg Bankoff

https://ebookname.com/product/cultures-of-disaster-society-and-
natural-hazards-in-the-philippines-1st-edition-greg-bankoff/

ebookname.com

The Tao of Dreaming_A Holistic Approach to Dreams Health

and Healing Rosemary Ellen Guiley

https://ebookname.com/product/the-tao-of-dreaming_a-holistic-approach-
to-dreams-health-and-healing-rosemary-ellen-guiley/

ebookname.com

Migration Education and Change Routledge Research in

Education 1st Edition S. Luchtenberg

https://ebookname.com/product/migration-education-and-change-
routledge-research-in-education-1st-edition-s-luchtenberg/

ebookname.com

iPhone and iPad App 24 Hour Trainer 1st Edition Abhishek

Mishra

https://ebookname.com/product/iphone-and-ipad-app-24-hour-trainer-1st-
edition-abhishek-mishra/

ebookname.com

Strategic Studies A Reader 1st Edition Thomas Mahnken

https://ebookname.com/product/strategic-studies-a-reader-1st-edition-
thomas-mahnken/

ebookname.com
A Watershed Year Anatomy of the Iowa Floods of 2008 Bur
Oak Book 1st Edition Edition Cornelia F. Mutel

https://ebookname.com/product/a-watershed-year-anatomy-of-the-iowa-
floods-of-2008-bur-oak-book-1st-edition-edition-cornelia-f-mutel/

ebookname.com
 ORDINARY
DIFFERENTIAL
EQUATIONS
Applications, Models, and Computing

© 2010 by Taylor and Francis Group, LLC

K11006_FM.indd 1 3/4/10 12:15:23 PM

 TEXTBOOKS in MATHEMATICS
Series Editor: Denny Gulick

PUBLISHED TITLES

ABSTRACT ALGEBRA: AN INTERACTIVE APPROACH

William Paulsen
COMPLEX VARIABLES: A PHYSICAL APPROACH WITH APPLICATIONS AND MATLAB®
Steven G. Krantz
ESSENTIALS OF TOPOLOGY WITH APPLICATIONS
Steven G. Krantz
INTRODUCTION TO ABSTRACT ALGEBRA
Jonathan D. H. Smith
INTRODUCTION TO MATHEMATICAL PROOFS: A TRANSITION
Charles E. Roberts, Jr.
INTRODUCTION TO PROBABILITY WITH MATHEMATICA®, SECOND EDITION
Kevin J. Hastings
LINEAR ALBEBRA: A FIRST COURSE WITH APPLICATIONS
Larry E. Knop
LINEAR AND NONLINEAR PROGRAMMING WITH MAPLE™: AN INTERACTIVE, APPLICATIONS-BASED
APPROACH
Paul E. Fishback
MATHEMATICAL AND EXPERIMENTAL MODELING OF PHYSICAL AND BIOLOGICAL PROCESSES
H. T. Banks and H. T. Tran
ORDINARY DIFFERENTIAL EQUATIONS: APPLICATIONS, MODELS, AND COMPUTING
Charles E. Roberts, Jr.

© 2010 by Taylor and Francis Group, LLC

K11006_FM.indd 2 3/4/10 12:15:23 PM

 TEXTBOOKS in MATHEMATICS

ORDINARY
DIFFERENTIAL
EQUATIONS
Applications, Models, and Computing

Charles E. Roberts, Jr.

Indiana State University
Terre Haute, U.S.A.

© 2010 by Taylor and Francis Group, LLC

K11006_FM.indd 3 3/4/10 12:15:23 PM

 MATLAB® is a trademark of The MathWorks, Inc. and is used with permission. The MathWorks does not
warrant the accuracy of the text or exercises in this book. This book’s use or discussion of MATLAB® soft-
ware or related products does not constitute endorsement or sponsorship by The MathWorks of a particular
pedagogical approach or particular use of the MATLAB® software.

Chapman & Hall/CRC

Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2010 by Taylor and Francis Group, LLC

Chapman & Hall/CRC is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed in the United States of America on acid-free paper

10 9 8 7 6 5 4 3 2 1

International Standard Book Number: 978-1-4398-1908-1 (Hardback)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts
have been made to publish reliable data and information, but the author and publisher cannot assume
responsibility for the validity of all materials or the consequences of their use. The authors and publishers
have attempted to trace the copyright holders of all material reproduced in this publication and apologize to
copyright holders if permission to publish in this form has not been obtained. If any copyright material has
not been acknowledged please write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmit-
ted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented,
including photocopying, microfilming, and recording, or in any information storage or retrieval system,
without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.
com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and
registration for a variety of users. For organizations that have been granted a photocopy license by the CCC,
a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used
only for identification and explanation without intent to infringe.

Library of Congress Cataloging‑in‑Publication Data

Roberts, Charles E., 1942-

Ordinary differential equations : applications, models, and computing / Charles
Roberts.
p. cm. -- (Textbooks in mathematics)
Includes bibliographical references and index.
ISBN 978-1-4398-1908-1 (hardcover : alk. paper) 1. Differential equations. I. Title. II.
Series.

QA372.R753 2010
515’.352--dc22 2009052757

Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com

© 2010 by Taylor and Francis Group, LLC

K11006_FM.indd 4 3/4/10 12:15:24 PM

 v

To the memory of my mother and father, Evelyn and Charles

And to those who “light up my life,”

my wife, Imogene

my children, Eric and Natalie

and my grandson, Tristan

© 2010 by Taylor and Francis Group, LLC

 Contents

Preface xi
1 Introduction 1
1.1 Historical Prologue . . . . . . . . . . . . . . . . . . . 1
1.2 Definitions and Terminology . . . . . . . . . . . . . . . 5
1.3 Solutions and Problems . . . . . . . . . . . . . . . . . 17
1.4 A Nobel Prize Winning Application . . . . . . . . . . . . 29

2 The Initial Value Problem y0 = f (x, y); y(c) = d 33

2.1 Direction Fields . . . . . . . . . . . . . . . . . . . . . 34
2.2 Fundamental Theorems . . . . . . . . . . . . . . . . . 42
2.3 Solution of Simple First-Order Differential Equations . . . . 55
2.3.1 Solution of y0 = g(x) . . . . . . . . . . . . . . . . 55
2.3.2 Solution of the Separable Equation y0 = g(x)/h(y) . . . 58
2.3.3 Solution of the Linear Equation y0 = a(x)y + b(x) . . . 64
2.4 Numerical Solution . . . . . . . . . . . . . . . . . . . 73
2.4.1 Single-Step Methods . . . . . . . . . . . . . . . . 78
2.4.1.1 Taylor Series Method . . . . . . . . . . . . 78
2.4.1.2 Runge-Kutta Methods . . . . . . . . . . . 86
2.4.2 Multistep Methods . . . . . . . . . . . . . . . . . 97
2.4.2.1 Adams-Bashforth Methods . . . . . . . . . . 97
2.4.2.2 Nystrom Methods . . . . . . . . . . . . . . 100
2.4.2.3 Adams-Moulton Methods . . . . . . . . . . . 101
2.4.3 Predictor-Corrector Methods . . . . . . . . . . . . . 102
2.4.4 Pitfalls of Numerical Methods . . . . . . . . . . . . 107

vii

© 2010 by Taylor and Francis Group, LLC

 viii

3 Applications of the Initial Value Problem

y0 = f (x, y); y(c) = d 117

3.1 Calculus Revisited . . . . . . . . . . . . . . . . . . . 117

3.2 Learning Theory Models . . . . . . . . . . . . . . . . . 129

3.3 Population Models . . . . . . . . . . . . . . . . . . . 131

3.4 Simple Epidemic Models . . . . . . . . . . . . . . . . . 138

3.5 Falling Bodies . . . . . . . . . . . . . . . . . . . . . 142

3.6 Mixture Problems . . . . . . . . . . . . . . . . . . . 146

3.7 Curves of Pursuit . . . . . . . . . . . . . . . . . . . . 153

3.8 Chemical Reactions . . . . . . . . . . . . . . . . . . . 157

4 N-th Order Linear Differential Equations 163

4.1 Basic Theory . . . . . . . . . . . . . . . . . . . . . . 164

4.2 Roots of Polynomials . . . . . . . . . . . . . . . . . . 188

4.3 Homogeneous Linear Equations with Constant Coefficients . . 200

4.4 Nonhomogeneous Linear Equations with Constant Coefficients 211

4.5 Initial Value Problems . . . . . . . . . . . . . . . . . . 219

5 The Laplace Transform Method 223

5.1 The Laplace Transform and Its Properties . . . . . . . . . 223

5.2 Using the Laplace Transform and Its Inverse

to Solve Initial Value Problems . . . . . . . . . . . . . . 242

5.3 Convolution and the Laplace Transform . . . . . . . . . . 250

5.4 The Unit Function and Time-Delay Function . . . . . . . 257

5.5 Impulse Function . . . . . . . . . . . . . . . . . . . 267

© 2010 by Taylor and Francis Group, LLC

 ix

6 Applications of Linear Differential Equations

with Constant Coefficients 275
6.1 Second-Order Differential Equations . . . . . . . . . . . 275
6.1.1 Free Motion . . . . . . . . . . . . . . . . . . . 281
6.1.1.1 Free Undamped Motion . . . . . . . . . . 282
6.1.1.2 Free Damped Motion . . . . . . . . . . . 283
6.1.2 Forced Motion . . . . . . . . . . . . . . . . . 292
6.1.2.1 Undamped Forced Motion . . . . . . . . . 292
6.1.2.2 Damped Forced Motion . . . . . . . . . . 293
6.2 Higher Order Differential Equations . . . . . . . . . . . 297

7 Systems of First-Order Differential Equations 313

8 Linear Systems of First-Order Differential Equations 335

8.1 Matrices and Vectors . . . . . . . . . . . . . . . . . 335

8.2 Eigenvalues and Eigenvectors . . . . . . . . . . . . . . 349
8.3 Linear Systems with Constant Coefficients . . . . . . . . 363

9 Applications of Linear Systems with Constant Coefficients 383

9.1 Coupled Spring-Mass Systems . . . . . . . . . . . . . . 383

9.2 Pendulum Systems . . . . . . . . . . . . . . . . . . 390
9.3 The Path of an Electron . . . . . . . . . . . . . . . . 392
9.4 Mixture Problems . . . . . . . . . . . . . . . . . . . 397

10 Applications of Systems of Equations 407

10.1 Richardson’s Arms Race Model . . . . . . . . . . . . . 407

10.2 Phase-Plane Portraits . . . . . . . . . . . . . . . . . 416
10.3 Modified Richardson’s Arms Race Models . . . . . . . . 433
10.4 Lanchester’s Combat Models . . . . . . . . . . . . . . 445
10.5 Models for Interacting Species . . . . . . . . . . . . . 452
10.6 Epidemics . . . . . . . . . . . . . . . . . . . . . . 470

© 2010 by Taylor and Francis Group, LLC

 x

10.7 Pendulums . . . . . . . . . . . . . . . . . . . . . 480

10.8 Duffing’s Equation . . . . . . . . . . . . . . . . . . 489
10.9 Van der Pol’s Equation . . . . . . . . . . . . . . . . 490
10.10 Mixture Problems . . . . . . . . . . . . . . . . . . 491
10.11 The Restricted Three-Body Problem . . . . . . . . . . 494

Appendix A CSODE User’s Guide 499

Appendix B PORTRAIT User’s Guide 525

Appendix C Laplace Transforms 537

Answers to Selected Exercises 539

References 571

Index 575

© 2010 by Taylor and Francis Group, LLC

 Preface

Ordinary Differential Equations: Applications, Models, and Computing is

an introductory level textbook for undergraduate students majoring in math-
ematics, applied mathematics, computer science, one of the various fields of
engineering, or one of the physical or social sciences. During the past century,
the manner in which solutions to differential equations have been calculated
has changed dramatically. We have advanced from paper and pencil solution,
to calculator and programmable calculator solution, to high speed computer
calculation. Yet, in the past fifty years there has been very little change in
the topics taught in an introductory differential equations course, in the or-
der in which the topics are taught, or in the methods by which the topics
are taught. The “age of computing” is upon us and we need to develop new
courses and new methods for teaching differential equations. This text is an
attempt to facilitate some changes. It is designed for instructors who wish
to bring the computer into the classroom and emphasize and integrate the
use of computers in the teaching of differential equations. In the traditional
curriculum, students study few nonlinear differential equations and almost no
nonlinear systems due to the difficulty or impossibility of computing explicit
solutions manually. The theory associated with nonlinear systems may be
considered advanced, but generating a numerical solution with a computer
and interpreting that solution is fairly elementary. The computer has put the
study of nonlinear systems well within our grasp.
The word “computing” appears in the title of this book because many
examples and exercises require the use of some computer software to solve a
problem. Consequently, the reader needs to have computer software available
which can perform, at least, the following functions:

1. Graph a given function on a specified rectangle.

2. Graph the direction field of the first-order differential equation y0 =

f(x, y) in a specified rectangle.

3. Solve the first-order initial value problem y0 = f(x, y); y(c) = d on an

interval [a, b] which contains c.

4. Find all roots of a polynomial with complex coefficients.

5. Calculate the eigenvalues and eigenvectors of a real n × n matrix A

where 2 ≤ n ≤ 6.

xi

© 2010 by Taylor and Francis Group, LLC

 xii

6. Solve on an interval [a, b] a vector initial value problem consisting of

a system of n first-order differential equations and n initial conditions
where 2 ≤ n ≤ 6.

Many computer software packages are readily available which include these
features and usually many additional features. Three of the best known and
most widely used packages are MAPLE, Mathematica, and MATLAB r . In
general, each instructor already has his or her own favorite differential equa-
tions software package or combination of packages. For this reason, the text
was written to be independent of any particular software package. The soft-
ware we used to generate solutions and many of the graphs for the examples
as well as the answers to the selected exercises which appear at the end of this
text is contained on the computer disc which accompanies the text. Complete
instructions for running this software are contained in Appendices A and B.
It is assumed the reader has completed calculus at least up to and including
the concept of partial derivatives and knows how to add, subtract, multiply,
and divide complex numbers. Concepts with which the reader may not already
be familiar are introduced and explained to the degree necessary for use within
the text at the location where the concept is first used.
Students who enroll in ordinary differential equations courses normally do
so for only one or two semesters as an undergraduate. In addition, few of these
students ever enroll in a numerical analysis course. However, most students
who complete a differential equations course find employment in business,
industry, or government and will use a computer and numerical methods to
solve mathematical problems almost exclusively. Consequently, one objective
of this text is to solve ordinary differential equations in the same way they are
solved in many professions—by computer. Thus, the single most useful and
distinguishing feature of this text is the use of computer software through-
out the entire text to numerically solve various types of ordinary differential
equations. Prior to generating a numerical solution, applicable theory must be
considered; therefore, we state (but usually do not prove) existence, unique-
ness, and continuation theorems for initial value problems at various points in
the text. Numerical case studies illustrate the possible pitfalls of computing
a numerical solution without first considering the appropriate theory.
Differential equations are an important tool in constructing mathemati-
cal models for physical phenomena. Throughout the text, we show how to
numerically solve many interesting mathematical models—such as popula-
tion growth models, epidemic models, mixture problems, curves of pursuit,
the Richardson’s arms race model, Lanchester’s combat models, Volterra-
Lotka prey-predator models, pendulum problems, and the restricted three-
body problem. When feasible we develop models entirely within separate
sections. This gives the instructor more flexibility in selecting the material
to be covered in the course. We hope to enrich and enliven the study of
differential equations by including several biographical sketches and historical

© 2010 by Taylor and Francis Group, LLC

 xiii

comments. In this text, we also attempt to provide an even balance between

theory, computer solution, and application.
We recommend that the core of a one quarter or one semester course consist
of material from chapters 1, 2, 4, 5, 7, 8, and sections 10.1, 10.2, and 10.3. The
remainder of material covered in the course should come from the applications
and models in chapters 3, 6, 9, and sections 10.4 through 10.11. The selection
of applications and models to be included in the course will depend on the
time available, on the intent of the course, on the student audience, and, of
course, on the preferences of the instructor. The following is a summary of
the material to be found in each chapter.
In Chapter 1 we present a very brief history of the development of calculus
and differential equations. We introduce essential definitions and terminology.
And we define and discuss initial value problems and boundary value prob-
lems.
In Chapter 2 we discuss in detail the first-order initial value problem
y0 = f(x, y); y(c) = d. First, we define the direction field for the differential
equation y0 = f(x, y), we discuss the significance of the direction field, and we
show how to use a computer program to produce a graph of the direction field.
Next, we state a fundamental existence theorem, a fundamental existence and
uniqueness theorem, and a continuation theorem for the initial value problem.
We show how to apply these theorems to a variety of initial value problems
and we illustrate and emphasize the importance of these theorems. Then we
discuss how to find explicit solutions of simple first-order differential equations
such as separable equations and linear equations. Next, we present simple
applications of linear first-order differential equations. Finally, we present
some of the simpler single-step, multistep, and predictor-corrector methods
for computing a numerical approximation to the solution of an initial value
problem. We explain how to use a computer program to generate approximate,
numerical solutions to initial value problems. We illustrate and interpret
the various kinds of results which the computer produces. Furthermore, we
reiterate the importance of performing a thorough mathematical analysis,
which includes applying the fundamental theorems to the problem, prior to
generating a numerical solution.
In Chapter 3 we consider a variety of applications of the initial value prob-
lem y0 = f(x, y); y(c) = d. First, in the section Calculus Revisited, we show
that the solution to the special initial value problem y0 = f(x); y(a) = 0 on the
Rb
interval [a, b] is equivalent to the definite integral a f(x) dx. Then we show
how to use computer software to calculate an approximation to the definite
Rb
integral a f(x) dx. This will allow one to numerically solve many problems
from calculus. In the sections Learning Theory Models, Population Models,
Simple Epidemic Models, Falling Bodies, Mixture Problems, Curves of Pur-
suit, and Chemical Reactions we examine physical problems from a number

© 2010 by Taylor and Francis Group, LLC

 xiv

of diverse disciplines which can be written as initial value problems and then
solved using numerical integration software. At the end of the chapter, we
present additional applications.
In Chapter 4 we discuss the basic theory for n-th order linear differential
equations. We present a history of the attempts of mathematicians to find
roots of polynomials. Then we illustrate how to use computer software to
approximate the roots of polynomials numerically. Next, we show how to find
the general solution of an n-th order homogeneous linear differential equation
with constant coefficients by finding the roots of an n-th degree polynomial.
Finally, we indicate how to find the general solution of a nonhomogeneous
linear differential equation with constant coefficients using the method of un-
determined coefficients.
In Chapter 5 we define the Laplace transform and examine its properties.
Next, we show how to solve homogeneous and nonhomogeneous linear dif-
ferential equations with constant coefficients and their corresponding initial
value problems using the Laplace transform method. Then we define the con-
volution of two functions and prove the convolution theorem. Finally, we show
how to solve nonhomogeneous linear differential equations with constant coef-
ficients in which the nonhomogeneity is a discontinuous function, a time-delay
function, or an impulse function.
In Chapter 6 we examine several linear differential equations with constant
coefficients which arise in the study of various physical and electrical systems.
In Chapter 7 we define a system of first-order differential equations. We
state a fundamental existence and uniqueness theorem and a continuation
theorem for the system initial value problem. Then, we show how to apply
these theorems to several initial value problems. Next, we show how to rewrite
an n-th order differential equation as an equivalent system of n first-order
equations.
In Chapter 8 we discuss linear systems of first-order differential equations.
First, we introduce matrix notation and terminology, we review fundamental
facts from matrix theory and linear algebra, and we discuss some compu-
tational techniques. Next, we define the concepts of eigenvalues and eigen-
vectors of a constant matrix, we show how to manually compute eigenvalues
and eigenvectors, and we illustrate how to use computer software to calcu-
late eigenvalues and eigenvectors. We show how to write a system of linear
first-order differential equations with constant coefficients using matrix-vector
notation, we state existence and representation theorems regarding the gen-
eral solution of both homogeneous and nonhomogeneous linear systems, and
we show how to write the general solution in terms of eigenvalues and eigen-
vectors.
In Chapter 9 we examine a few linear systems with constant coefficients
which arise in various physical systems such as coupled spring-mass systems,
pendulum systems, the path of an electron, and mixture problems.

© 2010 by Taylor and Francis Group, LLC

 xv

In Chapter 10 we present techniques for determining the behavior of so-

lutions to systems of first-order differential equations without first finding
the solutions. To this end, we define and discuss equilibrium points (criti-
cal points), various types of stability and instability, and phase-plane graphs.
Next, we show how to use computer software to solve systems of first-order
differential equations numerically, how to graph the solution components and
how to produce phase-plane graphs. We also state stability theorems for sys-
tems of first-order differential equations. Throughout this chapter we develop
and discuss a wide variety of models and applications which can be written
as vector initial value problems and then solved numerically.
Although not the primary purpose for which this book was intended, it
may also be used as (1) a text for a second course in ordinary differential
equations for students who have taken a traditional course previously, (2) a
text for a laboratory course, and (3) a supplement to the differential equations
component of a numerical analysis course.
Comments on Our Computer Software No prior knowledge of computers
or of any particular programming language is required to use our computer
software. Furthermore, no programming can be done. The user simply selects
a program to perform a particular task and enters the appropriate data. Then
the user interacts with the program by selecting options to be executed. The
user only needs to know the acceptable formats for entering numerical data
and the appropriate syntax for entering functions. The computer disc included
with this text contains two main programs.
The first program, CSODE, includes the six subprograms: GRAPH, DIR-
FIELD, SOLVEIVP, POLYRTS, EIGEN, and SOLVESYS. The subprogram
GRAPH graphs a function y = f(x) on a specified rectangle in the xy-plane.
The instructional purposes of this program are to teach the user how to en-
ter functions into programs properly and how to interact with programs.
Of course, GRAPH may be used to graph explicit solutions of differential
equations and view their behavior. The subprogram DIRFIELD graphs the
direction field of the first-order differential equation y0 = f(x, y) on a spec-
ified rectangle. The output of DIRFIELD permits the user to “see” where
the differential equation is and is not defined, where solutions increase and
decrease, and where extreme values occur. Sometimes asymptotic behavior
of the solutions can be determined also. SOLVEIVP solves the scalar first-
order initial value problem y0 = f(x, y); y(c) = d on an interval [a, b] where
c ∈ [a, b]. The solution values yi at 1001 equally spaced points xi ∈ [a, b]
may be viewed or plotted, with or without the associated direction field, on
a rectangle specified by the user. The subprogram POLYRTS calculates the
roots of a polynomial with complex coefficients of degree less than or equal
to ten. EIGEN calculates the eigenvalues and associated eigenvectors of an
n × n matrix with real coefficients where 2 ≤ n ≤ 6. The sixth subprogram
SOLVESYS solves the vector initial value problem y0 = f (x, y); y(c) = d on
the interval [a, b] where c ∈ [a, b] and the vector has from two to six com-

© 2010 by Taylor and Francis Group, LLC

 xvi

ponents. The user may view the solution values on the interval [a, b], may
graph any subset of solution components on any subinterval of [a, b], and may
view a phase-plane graph of any solution component versus any other solu-
tion component on any specified rectangle. Complete details for using the
six subprograms GRAPH, DIRFIELD, SOLVEIVP, POLYRTS, EIGEN, and
SOLVESYS appear in Appendix A.
The second program, PORTRAIT, solves the two component, autonomous
initial value problem
dy1
= f1 (y1 , y2); y1 (ci ) = d1i
dx
dy2
= f2 (y1 , y2); y2 (ci ) = d2i
dx
on the interval [ai, bi] where ci ∈ [ai, bi] for 1 ≤ i ≤ 10. After the solution
of an initial value problem has been calculated, the user may elect (i) to
print the solution components of any initial value problem already solved,
(ii) to graph any subset of the solution components previously solved in a
rectangle, (iii) to produce a phase-plane portrait of any pair of initial value
problems already solved on any rectangle, (iv) to rerun the most recent initial
value problem using a different interval of integration or initial conditions, or
(v) to input the initial conditions for the next initial value problem to be
solved. Complete details for using PORTRAIT appear in Appendix B.
The numerical integration procedure which is employed in the programs
SOLVEIVP, SOLVESYS, and PORTRAIT is a variable order, variable step-
size, multistep, Adams predictor-corrector method. The order is selected by
the program and varies from order one to order twelve. At each step, the step-
size is selected so that the maximum of the relative error and the absolute
error remains less than 10−12.
Classroom Environment We believe our “computer laboratory” is the
best environment in which to teach an introductory differential equations
course which emphasizes the use of computers. The center half of our labo-
ratory has long tables at which students sit during the lecture portion of the
class—which some days may be the entire class period. At the front of the
classroom are two “white boards” on which lecture notes can be written and
a “smartboard.” Also at the front of the room is a microcomputer which is
linked to a projector. The instructor can demonstrate how to run programs
from the microcomputer and discuss the output of the programs on the smart-
board. The outside two quarters of the computer laboratory has more than
forty microcomputers with the programs CSODE and PORTRAIT on the
desktops. When the students are seated at the microcomputers, they can run
the same program the instructor is running and compare the display on their
monitors with the display projected onto the smartboard. When the students
are working on their own, the instructor can walk around the classroom and
assist the students individually.

© 2010 by Taylor and Francis Group, LLC

Exploring the Variety of Random
Documents with Different Content
 III.—VAN DYCK, Cornelius van der Geest. National Gallery, London.

And now, if we stop to consider the great men in the arts, we shall
invariably find that each one of them is marked by some quality of
universal significance. There is something about them all that
overleaps the provincial, the accidental, the small, and the trifling.
They disregard in a measure the local truths and aim at the general
truths—at things essentially true for all humanity. Our Shakespeares
and Platos and St. Pauls survey the world from mountain tops. From
these vantage points their perspective is far-reaching, their view of
the world expansive. They see and grasp the essentials, the basic
elements, the foundations of things. It is this, for one thing, that
makes the art of Titian so superlatively great. What wonderful men
and women people his pantheon! What types they are of manhood
and womanhood! What embodiments of loftiness, dignity, and
nobility! And are they not universally admired? No matter what a
man’s nationality, he cannot choose but be interested in “The Man
with a Glove” or the “Charles V.” at Madrid. There is something in
them of that truth seen from mountain heights which every one will
recognize as the nobler part of his little valley-world.
Just so with the art of Rembrandt. His type is essentially of the
Low Countries; his costumes, landscapes, light-and-shade, and
methods are all localized in Holland. But a sadder painter you cannot
find in all the reach of painting. His emotional nature had been wrung
by trial and suffering and his sympathies were with the down-trodden
and the grief-stricken. There never was a painter who painted so
much of sorrow in the faces of his people as Rembrandt. The “Christ
at Emmaus” is, in form and figure, only a poor emaciated Amsterdam
Jew; but in emotional truth it is the one Christ of all painting. That
face appeals to Christian, Mahometan, Jew, and infidel alike, not
because of its divinity but because of its intense humanity. Should
we bring up the names of the other great masters of painting we
should find that each one of them is remarkable for some quality of
universal significance—Michael Angelo for his great command of
form, Rubens for his great splendor of effect, Velasquez for his
sense of vitality in the physical presence (Plate 13), Raphael for his
unity and his harmony.
The great men are remarkable for their breadth—the wide angle of
their vision. They see, not differently from others, but they see more.
Yet it is only a point of view, a limited outlook, and not by any means
the total sum of truth. The report of nature made by man, which we
have defined as “truth,” is always a report of some sort whoever
makes it. The difference between the great minds and the small
ones consists in what is seen and reported. A Rousseau who sees
and tells of the solidity of the earth, the volume of the forest, the
great luminous expanse of the sky, does not think to tell everything
that may be in the landscape. He sees the great truths, those truths
that are of universal permanence in all landscape, and emphasizes
them at the expense of the smaller details. A man of narrower vision
would perhaps overlook the sky and earth, and fail to see the forest
for the trees. He might centre all his interest in blades of grass, in
dew-drops and spider-webs and opening buttercups—the infinitely
little things in the landscape.
In portraiture men like Gerard Dou and Denner emphasize the
small skin-facts of a man’s face with such minute workmanship that
you may study them with a magnifying-glass. You will never see
anything like this in the portraits by Titian, Rembrandt, Velasquez,
Van Dyck (Plates 3, 13, 18). They waste no time on small truths.
They are intent upon giving the large physical presence, not the
petty deformities of the epidermis.
Again in drawing a hand and arm you will observe that men like
Gérôme give every curve and break of light along the arm, every
accidental contortion of muscle, every wrinkle and twist of flesh; but
somehow, when all these features are put down, the arm fails to live,
fails to move. It is a petrified arm. For an opposite statement of truth
look at the arm of Millet’s “Sower.” There is nothing absolute or
minute about the drawing. The arm is generalized, summarized,
synthetized as it were. The wrinkles in it are not apparent, the
covering of it is vague, the hand is not articulated in the muscles or
even definite in the drawing of the fingers. In short the whole arm
and hand are cut down to a few elementary lines, so that they
appear to the uninitiated somewhat sketchy and peremptory. But
looked at for those qualities which Millet thought more important than
surface texture, looked at for bulk, mass, weight, motion—
particularly motion—and there is a larger view apparent. The arm
and hand certainly have motion and life. And these are precisely
what Gérôme’s arm and hand have not. Can it not be claimed then
that the truth of life and motion is a greater truth than the truth of
momentary rigidity? Is it not a fact that Millet has seized upon a
general and universal truth characteristic of all arms and hands—that
is, the truth of life and movement—whereas Gérôme has seized
upon an accidental truth of light-and-shade which may be something
local and peculiar to that one hand and arm?

IV.—MILLET, The Gleaners. Louvre, Paris.

If one shows us a snap-photograph of breaking waves, what do

we see if not the highest and most brittle wave the camera man
could catch? Does this give us a general or a particular truth of the
sea? Do waves stand rigidly in air, petrified from base to crest, as we
see in the photograph, or do they roll and keep on rolling indefinitely
and ceaselessly? Does not the very essence of truth about a wave
lie in its restless heave and toss, its breaking and reforming, its
eternal indefiniteness of form? How many sea pictures have we seen
with every wave in place—pounded into place like hammered steel—
with every facet shining like a mirror, and not a possibility of motion
in anything? Perhaps we have rather enjoyed them and fancied, in
crossing the ocean, that the waves looked like that. Perhaps they
did; perhaps we were content to see only the small truths of the
ocean; but a study of the marines by Courbet, Manet, and Monet
may convince us that there are larger truths of the ocean than those
relating to its mirror-like sparkles—larger truths in the ocean’s depth,
power, and its restless, ceaseless motion. These painters have
discarded small things on the surface of the water, as Frans Hals the
small spots on a man’s face, in order to give the sense of form back
of it (Plate 22).
In the same way you will often find painters discarding the exact
drawing of objects such as wood or cloth or stone or metal in order
that they may give the weight, the elasticity, or the density of these
objects. A feather or a leaf may be an epitome of floating, dancing
lightness, but if you draw its complete anatomy and paint all its
surface texture you will have something that is as heavy as wrought
iron. It does not follow either, because Desgoffes gives us the sheen
and flash of brasses, china, and satins, that he has told all or the
most vital truths about those articles. Vollon may paint the same
things in a fuller manner, showing us something of structural
character which is just as important and just as true as surface
appearance. Moreover, the broader method leaves something to
implication and suggestion, where the other method buries under an
accumulation of fact.
Please note the word “suggestion,” for it is by suggestion that the
greatest truths of art are brought home to us. The realist, whom we
have been hastily considering, does not care for this method of
approach. He is bent upon realization. He is analytical in his
statement of each and every fact and makes a full report. All painters
do this in some degree during the early stages of their career, but as
they advance in years and experience there is a tendency to a
broader treatment, a return to the simple line of the child, to the
synthesis of a Millet, as shown in the arms, hands, and backs of the
women in “The Gleaners” (Plate 4), to the implication and suggestion
of a Corot, as shown in the sky of the “Biblis.” Facts are summarized.
A mere charcoal outline drawn by Degas gives us the reliefs,
proportions, weight, and bulk of a human figure; a shadow with
Giorgione or Rembrandt sums up the series of facts beneath it, and
becomes suggestive by its very mystery and uncertainty; a blended
blur of color by Whistler may bring to mind a heaving wave in mid-
ocean better than all the drawn and tinted and “realized” waves of all
the realists.
It is not the heaping of fact upon fact that flashes the truth upon us
—at least not in art, though it may in logic or in law. Indeed, the
accumulation of evidence often confuses. It is common studio
experience that a sketch of a picture is frequently better than the
picture itself. The attempt to “finish” (that is, to put in all the details
and minutiæ) makes it dull and unsuggestive. The unfinished
marbles of Michael Angelo, do they really suffer much by being
unfinished? I have sometimes thought that the figure of “Day” in the
Medici Chapel gained by its incompleteness—that it was better than
the “Night” upon the opposite side of the tomb because the
sculptor’s intention is perfectly obvious and yet the spectator’s
imagination is not stifled. There, like a fallen god, he lies, half
embedded in his matrix of stone with a suggestion of mighty power,
never so strongly felt in any other marble in this world. The lack of
finish, the mystery, the uncertainty, help on the imagination. One may
fancy, as many have done, that the figure symbolizes the loss of
Florentine freedom, and that the grand captive, with his massive
brow and sunken eyes, half-rises wearily to view the morning light
shining for him in vain. And again one may imagine he is a new
Prometheus bound to the rock; one of the Gigantes; or perhaps a
conquered Titan lying along the hills of Tartarus in the drear twilight,
brooding in melancholy silence over the loss of Olympus. To
whatever the mind may conjure up regarding the figure, the element
of reserved strength will lend assistance. Cut the captive from his
bed of stone and the strength falls short, lacking the foil of
resistance; finish the marble and an existent fact precludes the
possibility of wide imagination.
 The great English master of art, how well he knew what to leave
out! The lovers Lorenzo and Jessica are in the still, evening air, and
with what consummate skill Shakespeare paints the landscape with
that one suggestive line:

“How sweet the moonlight sleeps upon this bank.”

Not a word about the trees or grasses or ponds or meadows; not a

word about the stillness of the night, the hushed winds, and the
shining stars; but do you not see them all? Do they not rise up before
your eyes as by magic? Your realist would have put us to sleep with
dreary descriptions of grass and groves and glittering dew-drops
instead of the moonlight. And Shakespeare himself might have
written a volume of description and still not roused us to his meaning
so quickly as with that one suggestive line. The value of the sign in
art, whether it be pictorial, sculptural, or literary, lies in its suggestive
quality; and the “Sower” of Millet, the “Day” of Michael Angelo, and
the moonlight of Shakespeare are merely so many suggestive signs.
 V.—PAOLO VERONESE, Marriage in Cana. Louvre, Paris.

Thus far our inquiry has extended no farther than the truth of
nature—the truth of appearance as shown in realistic art. But there
are other truths with which the picture has to do that perhaps call for
a moment’s consideration. The truth of history for which the public
contends so valiantly need not detain us long. That Paolo Veronese
and his contemporaries chose to garb the sacred characters of the
“Marriage in Cana” (Plate 5) or “Moses saved from the Nile” (Plate
23), in Venetian costume, is matter of small importance. And it is of
still less importance whether Christ and the Apostles show the
Semitic cast of countenance or not. The intense reverence for local
and ethnographical truth possessed by the Holman Hunts and Alma-
Tademas of the art world would seem somewhat misplaced. No
matter what care is bestowed upon the archæology, there is always
something not quite true to the fact. And moreover, all art in all times
has pictured its own race, costume, and country. It would not be
worth much unless it did. The marble gods of Greece are all Greek,
the painted Madonnas of Italy are all Italian. How otherwise would
you have it? Marlowe’s Mephistopheles talks English, and Goethe’s
Mephistopheles talks German. What language should they talk?
When art deals with the past it translates it into the present. It could
not possibly do otherwise. No Anglo-Saxon could feel, think, or work
like a Greek, simply because he is an Anglo-Saxon.
There is another truth of far more consequence than historical
accuracy, and that is the truth of art. This comes in here opportunely
enough, for art-truth is produced by the suggestive method of
dealing with facts which I have just been illustrating. The method is
absolutely essential to all strong work in all departments. It is usually
known in painting as the “law of sacrifice”; and you will find it in
literature under the name of “dramatic force.” We should never have
had such characters as Faust and Macbeth had all the other
characters in the plays been treated with an importance equal to that
of the heroes. Hamlet is an elevated Hamlet simply because the
other characters are subordinate characters, just as Corot’s light is
light, because everything else in the picture is sacrificed to it. There
is no quarrel with truth to nature in this truth to art. Great art seldom
falsifies, but it always selects, emphasizing some features and
subordinating other features. It usually gives the large truths and
merely implies the small ones. Millet in his “Sower” has no notion of
telling you more than a few prominent facts about the man and his
work. He shows a peasant, working under the shadow of a hill,
working late in the evening, swinging and sowing with rhythmic
motion of foot, hand, arm and body. It is matter of no importance
whether he wears linen or woollen or cotton, whether his blouse has
buttons upon it or not, whether his face is clean or not. The all-
pervading truth of the picture lies in the swinging form of the sower,
and to keep your attention upon that he omits everything else. The
figure is but a suggestion, a something that stands as an equivalent
for that man whom Millet thought should be recognized for his
patience and fortitude of spirit, his nobility and dignity in the hard
labor of life, his fine pictorial qualities as seen against the
background of his native heath. That is the ulterior meaning which he
would show us. The sign is true to the great truth of a sower, the
meaning is true within the limits of pictorial creation, and finally the
recording of it is true to the truth of art.
This method of procedure, wherein suggestion becomes such an
important factor, implies two people in the work of art rather than
one. The spectator must do his part as well as the artist. The latter
suggests, the former takes up the suggestion and builds upon it.
When Velasquez painted Christ on the cross, hanging there alone in
the night, the head bowed forward on the breast, and the long dark
hair tailing over the face and half covering it, he did not think to
obliterate the face—to take it out of the picture completely. He knew
very well that the imagination of the spectator would go behind the
veil and picture that face more vividly than he could paint it. What
painter ever yet produced a wholly satisfactory face of Christ?
Velasquez was wise in leaving it to the imagination of the spectator.
How wise he was you can perhaps gather by contrasting his “Christ
on the Cross” with the same subject by Léon Bonnat—one of the
noblest of the latter-day realists. Bonnat simply took a dead body
from the morgue and hung it upon a cross in the court-yard of the
École de Médecine, and painted it exactly as he saw it. But it is not
Christ; it is the dead body he took from the morgue. There is strain of
arm and leg and torso, the anatomy is wrenched, the muscles are
contorted, the veins are swollen. But there is no suggestion of
anything that had been noble or exalted in the living. In fact there is
not a suggestion of any kind. Everything is told and the spectator’s
imagination is not called upon. Realism has been pushed into the
last ditch, and yet has produced only a sign standing for Christ on
the cross, and not the real thing—a sign which, in gaining an
elaborate truth to fact, has lost its truth to art and its power of
suggestion.
 VI.—CARPACCIO, St. Ursula and Prince of England (detail). Academy, Venice.

We may as well conclude then, without further illustration, that the

exact portrayal of nature known as realism falls somewhat short of
its mark. It may report and report, but it cannot realize. Light, air,
hills, mountains, human beings and their habitations cannot be
reproduced, but they may be translated through the medium of
pigment and thus rendered intelligible to us. You may translate them
“realistically” or you may translate them suggestively, but in either
case it is the translation that you will have, and not the original. Each
art—music, poetry, painting—has its peculiar method of translation,
and we have called the result in each case a sign—a convention
which we have agreed to recognize as meaning thus and so; but of
course the signs in painting are not quite so arbitrary as in language
or chemistry. The painting of a wave certainly looks more like a wave
than the word “water,” or the symbol H2O. The sign has a certain
resemblance to the original which gives a reason for the existence of
realism and also adds to the confusion of those who would spin a
theory of art; but the resemblance should not mislead us. The sign is
still a sign, though in the one case it is representative and in the
other case symbolic. Its meaning has not changed in any way. The
all-seeing eye of Osiris is not like those speaking eyes in Van Dyck’s
portrait of “Cornelius Van der Geest” (Plate 3). One is more
conventional than the other, but both are conventions.
It is not necessary that we should deny value to this realistic art,
even though we do not wholly accept it. The very endeavor to make
the work faithful to the original in every detail, though it may hurt its
deeper sentiment, cannot but result in good workmanship; and that
in itself is always acceptable and pleasurable. Indeed, bald realism,
with nothing else back of it, is seldom seen in art. The man, the
material, and the method are inextricably mixed together, so that the
product always has more or less individuality about it, or is
decorative in form or color, or expresses some thought or feeling of
the painter, or stands for something in subject. In any event the well-
made sign—even as a sign—is not to be scorned. We shall see
hereafter how it is distorted by the personal element, how it is twisted
by the imagination, how it is warped by the decorative instinct; but
we are not to forget at any time that it is but a symbol, merely a
means of suggesting reality, and not reality itself.
 CHAPTER II

INDIVIDUALITY OR THE PERSONAL ELEMENT

The fact that “the report about nature” which we have called “truth”
varies with the reporter is of vital importance to us in comprehending
the measure of exactness in the result. It is something that must be
reckoned with in every thought, deed, and utterance, for its presence
is potent in all human endeavor. Two astronomers, to use the
accepted illustration, taking the time of the passage of a series of
stars over the same meridian, will not precisely agree in their
arithmetical results. However accurate, unbiassed, and mechanical
in action they may seek to be, it happens that one takes the time
earlier or later than the other. Consequently there is always a
variation in the product, which has to be rectified by adding a
constant. This is what is called the personal equation—a something
we have heard about in literature and art as well as in science.
Perhaps you may remember that in the writing class of our youth
when the motto, “Evil communications corrupt good manners,” was
given us as an example to copy, we all wrote the motto, and we all
tried to follow the exact form of the copper-plate pattern before us;
but somehow our performances differed one from another. In some
the letters were large, in others they were small; the angle was
flatter, the line was firmer, or the shading heavier. We used to think it
merely a matter of practice, and fancied if we kept at it long enough
we could ultimately write exactly like the copper-plate pattern. But I
wonder if we thought quite correctly about that. Certainly there are
thousands of people who have been writing all their lives and have
had practice enough, but these are the ones that show the most
marked variations from the model. Each one writes in a manner
peculiarly his own. And these handwritings that vary so radically
interest us very much. We see all sorts of striking peculiarities in
them suggestive of their authors, and we even have so-called
scientists who read character out of them, or into them, I will not say
which. The cause of the variation is not far to seek. It is the personal
element appearing in the work and influencing it. If we would get the
same result in all handwritings we must eliminate the personal
element or, if you please, reckon with the personal equation.
This quality which creates the variance in handwriting is met with
even more positively in painting. For painting is, after all, only an
elaborated picture-writing, more flexible, perhaps, than letter-writing,
and, therefore, more easily bent by a personality; but in the main
influenced by the same principles as regards the variation of the
characters. We all write the letter “A” and they are all “A’s,” but each
is different from the other, just as all landscape painters paint hills
and trees and they are all hills and trees, yet each is different again.
If three painters, say Turner, Rousseau, and Claude Monet, could be
brought together and induced, each for himself, to paint a given tree,
there can be no doubt that all three of the paintings would represent
the tree and be true enough representations into the bargain; but
they would not be at all like one another. The Turner would
undoubtedly give the height, the branching outline, the grace and
grandeur of the tree; but in flattened form, perhaps in silhouette
against a yellow evening sky. In any event and under any
circumstances we may be sure that it would be a Turnerian tree. And
the Rousseau would be correspondingly true to Rousseau’s peculiar
point of view. It would probably have an emphasis of mass and
volume; it would be as deep through as broad across, it would be
firm in its rooting, massive in its trunk and branches, heavy in its
foliage, rich in its coloring. But Claude Monet, painting the same tree,
would not see the things that appealed to Turner and Rousseau, or if
he did he would disregard them. He would overlook form and line
and body, perhaps lose them entirely in studying the sunlight falling
upon the foliage, in painting the colored reflections cast by sky and
ground and water, in surrounding the tree with colored air and giving
it a setting in an atmospheric envelope. Undoubtedly we should be
able to recognize the original tree in any one of the three counterfeit
presentments. Each would differ from the other and yet no one of
them be false. There would be three different truths about the one
tree—three different phases of the one fact. And undoubtedly we
should be able to say just which painter painted each picture. How?
Because we should recognize in each the point of view peculiar to its
maker—we should recognize the individuality of the painter.
If we consider this same tree as part of a landscape—consider it in
connection with foreground, background, and sky—we shall see that
the chance for the display of individuality is even greater. The choice
of the painter as to how the tree shall be seen determines at the very
start the character of the representation. If it is placed in the
foreground and spreads in a pattern of branches and leaves high up
against the sky, we have one phase of tree-truth, one kind of picture
which may perhaps resemble, in a way, the work of Harpignies, if it is
placed in the middle distance, a shadowy form against a pale
morning sky, with a feeling of heavy air and rising mists, we have
another phase of truth, something which may represent Corot; if it is
seen in the far background against a yellow twilight sky, tall, dark,
motionless, we have still another phase of truth which may stand for
Daubigny. Any change in the position of the tree, any change in
foreground or sky-line, in light or reflection or atmosphere, would
represent a new angle of vision and hence a new truth. And the
preference of the painter for any particular phase of the
manifestation, any particular truth, would exhibit what we have called
his individuality.[2]
2. This matter of personality and choice is well illustrated by Mr. La Farge in his
“Considerations on Painting.” He says (p. 71): “I remember myself, years
ago, sketching with two well-known men, artists who were great friends,
great cronies, asking each other all the time how to do this and how to do
that! but absolutely different in the texture of their minds and in the result that
they wished to obtain, so far as the pictures and drawings by which they
were well known to the public are concerned.
 “What we made, or rather, I should say, what we wished to note, was
merely a memorandum of the passing effect upon the hills that lay before us.
We had no idea of expressing ourselves or of studying in any way the
subject for any future use. We merely had the intention to note this affair
rapidly, and we had all used the same words to express to each other what
we liked in it. There were big clouds rolling over hills, sky clearing above,
dots of trees and water and meadow land below; and the ground fell away
suddenly before us. Well, our three sketches were in the first place different
in shape; either from our physical differences, or from a habit of drawing
certain shapes of a picture, which itself usually indicates—as you know or
ought to know—whether we are looking far or near. Two were oblong, but of
different proportions; one was more nearly a square: the distance taken into
the right and left was smaller in the latter case, and, on the contrary, the
height up and down—that is to say, the portion of land beneath and the
portion of sky above—was greater. In each picture the distance bore a
different relation to the foreground. In each picture the clouds were treated
with different precision and different attention. In one picture the open sky
was the main intention of the picture. In two pictures the upper sky was of no
consequence—it was the clouds and the mountains that were insisted upon.
The drawing was the same—that is to say, the general make of things—but
each man had involuntarily looked upon what was most interesting to him in
the whole sight; and though the whole sight was what he meant to represent,
he had unconsciously preferred a beauty or interest different from what his
neighbors liked.
“The color of each painting was different—the vivacity of colors and tone,
the distinctness of each part in relation to the whole; and each picture would
have been recognized anywhere as a specimen of work by each one of us,
characteristic of our names. And we spent on the whole affair perhaps twenty
minutes. I wish you to understand again that we each thought and felt as
though we had been photographing the matter before us. We had not the
first desire of expressing ourselves, and I think would have been very much
worried had we not felt that each one was true to nature. Of course there is
no absolute nature, as with each slight shifting of the eye, involuntarily we
focus more or less distinctly some part to the prejudice of others. And not
only would this result have been the same if we had gone on painting, but
had we made a drawing, had we made a careful representation or rapid note
of what we saw by lines (that is to say, by an abstraction of the edges of the
surfaces that we saw), anyone could have told the names of the men who
had done it.”
 This preference for a peculiar point of view crops out very early in
the painter’s life. The students in an art class, drawing from the living
model on the platform, and each one striving to follow that model
literally, all show it. The sketches indicate by the placing of the figure
upon the paper, the size of the figure, the height or depth of the
shadows, the clearness or vagueness of the outline, that the
personal element—individuality—is present, influencing and
practically dominating the work of everyone in the class-room. And
this, too, in charcoal work, where the color problem is eliminated.
Moreover, there are features of these charcoal sketches, aside from
mere technique, that are equally interesting as indicative of the
peculiar temperament behind the pencil. You cannot fail to be struck
with the mood or spirit that creeps into each one of the drawings. On
one paper the model looks pleasant, almost jovial, on another he will
appear sad-faced or morose, on another, romantic as you might
conceive a Wagner hero, or classic and insipid like a Canova marble,
and on still another, gross, brutal, or perhaps foolish-looking. It is not
possible that the model could exhibit all these different moods. The
variation is not in him. He presents the same stolid, tired front
common to all models; the mood is added to him by the personality
holding the charcoal.
 VII.—BELLINI, Madonna and Saints. S. M. dei Frari, Venice.

We see the same variation among the works of older people—full-

fledged artists, in the world of art. Nowhere is it more apparent than
in the portrait, the one thing which might be thought to call for the
elimination of the painter and a close fidelity to the facts of the
original. But such is the power of preference that the painter almost
invariably emphasizes certain features at the expense of others less
interesting to him; or such is the warp of the vision that certain
qualities appear abnormal, certain prominences appear unduly
accentuated. There are portraits of the Duchess of Devonshire and
of Mrs. Siddons (Plate 19) by both Reynolds and Gainsborough, but
how very different they are! With Reynolds both of the characters are
healthy, robust, good-natured, somewhat loud and stormy; with
Gainsborough they are both delicate, subdued, refined, even
melancholy. And think of the portraits in the Louvre of Francis I. by
different hands, where only a slight thread of resemblance holds
them together; or, better still, the portraits of Napoleon I., painted by
the classic painters of his reign who believed in the utter effacement
of the artist in favor of the facts before him. How very different in
form, feature, mood, and character Napoleon appears in each
picture. He is classic; he is romantic; he is thin, fat, amiable, moody,
fiery, dreamy. David, Delaroche, Gros, no matter what their theories
in art, could not keep themselves out of the representation. All that
any one of them could do was to give his individual impression of the
model before him. Necessarily each was tinctured by a predilection
or a bias. It could not have been otherwise.
What is the cause of the variation in results to be seen in the
portrait? Why, for instance, do the photographs of Queen Victoria
show substantially the same thing, while the portraits of her by
painters show different things? Because the cameras are all made of
practically the same material, have the same sensitiveness, and
receive light in the same way; whereas men are not made of the
same material, have not the same sensitiveness, and receive
varying degrees of light according to their lucidity or absorbent
power, which is sometimes called genius. No two people are
fashioned precisely after the same pattern. They vary in intellectual,
emotional, and physical make-up. And let a painter strive as he may
to record an exact fact before him, he cannot escape the action of
his inherent faculties. These may be brighter, clearer, keener, than
those of other painters, or they may be duller and feebler; but at
least they are different, and he must use what nature has given him.
He was equipped originally to see with his own eyes, think with his
own brain, and work with his own hands. Is it not very apparent then
that the eye may warp the vision and report peculiarly to the brain,
which in turn may tell the hand to work thus and so? And the result in
art is what? Why, the individual view of one man; or nature passed
through the alembic of that man’s personality.[3]
3. “Our eyes, our ears, our sense of smell, of taste, differing from one person to
another, create as many truths as there are men upon earth. And our minds,
taking instructions from these organs, so diversely impressed, understand,
analyse, judge, as if each of us belonged to a different race. Each one of us,
therefore, forms for himself an illusion of the world; and the writer (the
painter, too) has no other mission than to reproduce faithfully this illusion,