Download the full version of the ebook at

https://ebookfinal.com

An Introduction to Modern Variational

Techniques in Mechanics and Engineering 1st
Edition B. D. Vujanovic

https://ebookfinal.com/download/an-introduction-
to-modern-variational-techniques-in-mechanics-and-
engineering-1st-edition-b-d-vujanovic/

Explore and download more ebook at https://ebookfinal.com

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

An introduction to genetic engineering 3ed Edition Nicholl

D.

https://ebookfinal.com/download/an-introduction-to-genetic-
engineering-3ed-edition-nicholl-d/

ebookfinal.com

Modern trends in structural and solid mechanics 3 Non

deterministic mechanics 1st Edition J. D. Kaplunov

https://ebookfinal.com/download/modern-trends-in-structural-and-solid-
mechanics-3-non-deterministic-mechanics-1st-edition-j-d-kaplunov/

ebookfinal.com

Materials Science and Engineering An Introduction 10th

edition. Edition William D. Callister

https://ebookfinal.com/download/materials-science-and-engineering-an-
introduction-10th-edition-edition-william-d-callister/

ebookfinal.com

An Introduction to Continuum Mechanics 2ed. Edition Reddy

J.N.

https://ebookfinal.com/download/an-introduction-to-continuum-
mechanics-2ed-edition-reddy-j-n/

ebookfinal.com
An Introduction to Structural Mechanics 2001st Edition
Paul Smith

https://ebookfinal.com/download/an-introduction-to-structural-
mechanics-2001st-edition-paul-smith/

ebookfinal.com

Engineering Materials 2 Third Edition An Introduction to

Microstructures Processing and Design D R H Jones

https://ebookfinal.com/download/engineering-materials-2-third-edition-
an-introduction-to-microstructures-processing-and-design-d-r-h-jones/

ebookfinal.com

Sets of Finite Perimeter and Geometric Variational

Problems An Introduction to Geometric Measure Theory 1st
Edition Francesco Maggi
https://ebookfinal.com/download/sets-of-finite-perimeter-and-
geometric-variational-problems-an-introduction-to-geometric-measure-
theory-1st-edition-francesco-maggi/
ebookfinal.com

An Introduction to Modern Astrophysics 2nd Edition Carroll

https://ebookfinal.com/download/an-introduction-to-modern-
astrophysics-2nd-edition-carroll/

ebookfinal.com

An Introduction to Rehabilitation Engineering 1st Edition

Rory A Cooper

https://ebookfinal.com/download/an-introduction-to-rehabilitation-
engineering-1st-edition-rory-a-cooper/

ebookfinal.com
An Introduction to Modern Variational Techniques in
Mechanics and Engineering 1st Edition B. D. Vujanovic
Digital Instant Download
Author(s): B. D. Vujanovic, T. M. Atanackovic (auth.)
ISBN(s): 9781461264675, 1461264677
Edition: 1
File Details: PDF, 14.17 MB
Year: 2004
Language: english
B.D. Vujanovic
T.M. Atanackovic

An Introduction to
Modem Variational Techniques
in Mechanics and Engineering

Springer Science+Business Media, LLC

B.D. Vujanovic T.M. Atanackovic
University of Novi Sad University of Novi Sad
Faculty of Technical Sciences Faculty of Technical Sciences
21121 Novi Sad 21121 Novi Sad
Serbia and Montenegro Serbia and Montenegro

Library of Congress CataIoging-in-Publication Data

Vujanovic, B. D. (Bozidar D.). 1930-

An introduction to modem variational techniques in mechanics and engineering I Bozidar
D. Vujanovic, Teodor M. Atanackovic.
p. cm.
Includes bibliographical references and index.
ISBN 978-1-4612-6467-5 ISBN 978-0-8176-8162-3 (eBook)
DOI 10.1007/978-0-8176-8162-3
1. Variational principles. 2. Mechanics, Analytic. 1. Atanackovic, Teodor M., 1945- II.
Title.

QA808.V852003
53 1'.01'5 l-dc22 2003062781

AMS Subject Classifications: 70H25, 70H30, 49-99, 74G65

ISBN 978-1-4612-6467-5 Printed on acid-free paper.

<1:12004 Springer Science+Business Media New York

Originally published by Birkhăuser Boston in 2004
Softcover reprint of the hardcover 1st edition 2004

AlI rights reserved. This work may noI be translated or copied in whole or in part without the written
permission of the publisher. Springer Science+Business Media, LLC.
except for brief excerpts in connection with reviews or scholarly
analysis. Use in connection with any form of information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed is
forbidden.
The use in this publication of trade names, trademarks, service marks and similar terms, even if they
are not identified as such, is not to be taken as an expression of opinion as to whether or not they are
subject to proprietary rights.

987 6 543 2 1 SPIN \0951811

www.birkhauser-science.com
Contents
Preface ix

I Differential Variational Principles of Mechanics

1 The Elements of Analytical M echanics Expressed Using the

Lagrange-D 'Alembert Differential Variational Principle 3
1.1 Introducti on . 3
1.2 Different ial Equ ations of Motion in Cartesian Coordinates 3
1.2.1 Free Dyn amic al Syst ems . 3
1.2.2 Constrained Motion. Lagrangian Equat ions with
Mu ltipliers 4
1.3 An Invar iant Form of Dynamics, the Lagrange-D'Alemb ert
Differential Vari ati onal Principle for Holonomic Dynamical
Systems 9
1.3.1 The Principle . . . 9
1.3.2 Generalized Coord inates and T heir Var iati ons . 11
1.3.3 The Lagrange-D'Al embert Variation al Principle
Expressed in Terms of Generalized Coo rdinates,
Central Lagrangian Equ ations. 13
1.4 Euler-Lagrangian Equati ons . . 16
1.4.1 The Structure of th e Kinetic En ergy. Explicit Form
of Euler- Lagrangian Eq uations 18
1.4.2 Tw o Important Cons ervation Laws of the
Euler-Lagrangian Equati ons: Momentum and J acobi
Conservation Laws 22
1.4.3 On t he Disturbed Motio n and Geomet ric St ability
of t he Scleronomic Potential Dynamical Syst ems 28
1.5 A Bri ef Out line of the Nonho lonom ic Dyn am ical Systems 33
1.6 Some Other Forms of th e Equations of Motio n 44
1.6.1 The Gibbs-Appell Eq uations: Holonomic Dynamical
Syst ems 44
1.6.2 T he Gibbs-Appell Equ ations : Non holonom ic
Dynamical Syst ems . 49
1.6.3 Kane's Equations . 52
1.7 Nielsen and Mangerone-Deleanu Differential Equations . 55
1.8 Hamilton 's Canonical Differential Equations of Motion . 59
vi Contents

1.9 Canonical Transformations . 64

1.10 Poisson Brackets, the Conditions of Canonicity
of a Given Transformation . . . . . . . . . . . . 69

2 The Hamilton-Jacobi Method of Integration of Canonical

Equations 73
2.1 Introduction. . . . . . . . . . . . .. .. . .. . .. . 73
2.2 The Hamilton-Jacobi Partial Differenti al Equ ation . 73
2.3 Some Appl ications of the Hamilton-Jacobi Method . 77
2.3.1 Linearly Damped Oscillator . . . . . . . . . . 77
2.3.2 Simple Harmonic Oscillator . . . . . . . . . . 78
2.3.3 The Case When a Particular Solution of the Riccati
Equation Is Available 80
2.4 The Oscillatory Motion with Two Degrees of Freedom . . . 84
2.5 Application of the Hamilton-Jacobi Method to the Study of
Rheolinear Oscillations . . . . . . . . . . . . . . . . . . . . . 96
2.6 A Conjugate Approach to Hamilton-Jacobi Theory. The Case
of Rheolinear Systems . . . . . . . . . . . . . . . . . . . 111
2.7 Quadratic Conservation Laws of Rheolinear Dynamical
Systems with Two Degrees of Freedom. . . . . . . . . . 116
2.7.1 An Alternative Form of th e Quad ratic Conservation
Law . . . . . . . 122
2.7.2 Some Examples. . . . . . . . . . . . . . . . . . . . . 124

3 Transformation Properties of Lagrange-D'Alembert Vari-

ational Principle: Conservation Laws of Nonconservative
Dynamical Systems 131
3.1 Introduction . . . . . .. .. . . . . .. . . . .. .. .. . . . 131
3.2 Simultaneous and Nonsimultaneous Virtual Displacements
(Variat ions) , Infinitesim al Transform at ions . . . . . . . . . . 132
3.3 A Transformation of the Lagrange-D'Alembert Principle. . 135
3.4 Th e Cond itions for the Exist ence of a Conserved Quantity
of the Given Dynamical System . . . . . . . . . . . . . 136
3.5 Th e Generalized Killing Equations . . . . . . . . . . . 138
3.6 The Basic Noeth er Ident ity and Int egrating Factors of
Equ at ions of Motion . . . . . . . . . . . . . . . . . . . 141
3.7 Quadrat ic Conservation Laws of Euler's Equ ation. . . 142
3.8 Quadratic Conservation Laws of the Scleronomic Duffing
Oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . 146
3.9 Conservation Laws of the Arbitrary Degree of a Purely
Dissipat ive Dynamical Syst em . . . . . . . . . . . . 149
3.10 Polynomial Conservation Laws of th e Generalized
Emden-Fowler Equ ation . . . . . . . . . . . . . . . 152
 Contents vii

4 A Field Method Suitable for Application in Conservative

and Nonconservative Mechanics 159
4.1 Introduction .. . . . . . . . . . .. .. . . . . . . .. . . 159
4.2 The Field Concept and Its Partial Differential Equation 160
4.2.1 The Bundle of Conservation Laws 161
4.2.2 The Initial Value Problems . . . . . . . . 161
4.3 A Non-Hamiltonian Rhconomic Syst em .. .. . 164
4.4 Some Ex amples with Many-Degrees-of-Freedom
Dynamical Systems. . . . . . . . . . . . . . . . . 166
4.4.1 Projectile Motion with Linear Air Resistance 166
4.4.2 Application of the Field Method to Nonholonomic
Dynamical Systems . . . . . . . . . . . . . . . . . . 170
4.5 Nonlinear Analysis . . . . . . . . . . . . . . . . . . . . . . 173
4.6 Conservation Laws and Reduction to Qu adr atures of the
Generalized Time-D ependent Duffing Equation 179
4.6.1 The Case of Arbitrary Q's (P = 0) . . . . . . . . . . 185
4.6.2 The Case Q3 = 0 (E = 0) . . . . . . . . . . . . . . . 186
4.6.3 Reduction to Quadratures by Means of the Hamilton-
J acobi Method . . . . . . . . . . . . . 187
4.6.4 The Case When a Particular Solution
of the RiccatiEquation Is Available . 192

II The Hamiltonian Integral Variational

Principle
5 The Hamiltonian Variational Principle and Its
Applications 197
5.1 Introduction .. . . . .. . .. . . . .. .. .. . . . . . . . . 197
5.2 The Simplest Form of the Hamiltonian Variational Principle 198
5.3 The Hamiltonian Principle for Nonconservative Force Field 205
5.4 The Fun ctional Containing the Higher Order Derivatives . 206
5.5 The Functional Depending upon Several Independent
Variables 207

6 Variable End Points, Natural Boundary Conditions, Bolza

Problems 215
6.1 Introduction . . . .. .. . . . .. . . .. . . . . . . 215
6.2 Time Interval (to, td Specified, qi (to), qi (td Free. 215
6.3 The Problem of Bolza . . . . . . . . . . . . . . . . 221
6.4 Unspecified Initial and Terminal T ime, Variable End Points 226
6.5 Jacobi's Form of the Variational Principle Describing the
Paths of Conservative Dynamical Syst ems . . . . . . . . . . 232
6.6 Piecewise Continuous Extremals. The Weierstrass-Erdmann
Corner Conditions . . . . . . . . . . . . . . . . . . . . . . . 236
viii Contents

1 Constrained Problems 241

7.1 Introduction. . . . . . . . .. . . . 241
7.2 Isoperimetric Constraints . . . . . 241
7.3 Algebraic (Holonomic) Constraints 246
7.4 Differential Equ ations Constraints 248
7.5 The Simplest Form of Hamilton's Variational Principle as a
Problem of Optimal Control Theory . . . . . . . . . . . . . 250
7.6 Continuous Optimal Control Problems 252
7.7 Optimal Control Problems with Unspecified Terminal Time 257

8 Variational Principles for Elastic Rods and Columns 263

8.1 Introduction . . . .. . . . . . . . . . . . . . . .. . .. . 263
8.2 The Column with Concentrated Force at the End . . . . 264
8.3 Rod with Compressible Axis and the Influence of Shear Stresses
on the Deformation. 270
8.4 Rotating Rod . . . . . . . . . . . . . . . . . . . . . . . 272
8.4.1 Bernoulli-Euler Theory . . . . . . . . . . . . . 272
8.4.2 Rotating Rod with Shear and Compressibility:
A Director Theory . . . . . . . . . . 277
8.5 Rod Loaded by a Force and a Torque . . . . 282
8.6 Optimal Shape of a Simply Supported Rod
(Lagrange's Problem) 289
8.7 Optimal Shape of a Rod Loaded by Distributed Follower
Force. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
8.8 Optimal Shape of the Rotating Rod .. . . . . . . . . . . . 303
8.9 Optimal Shape of a Rod Loaded by a Force and a Torque . 321
8.10 Variational Principle for Small Deformation Imposed on Large
Deformation of a Rod .. . . . . . . . . . . . . . . . . . . . 327

Bibliography 333

Index 343
Preface
This book is devoted to the basic variational principles of mechani cs: the
Lagr ange-D 'Alemb ert different ial var iationa l principle and t he Hamil ton
integral variationa l principle. These two variational principles form t he
main subject of conte mpora ry ana lyt ical mechanics, and from th em the
whole colossal corpus of classical dynamics can be deductively derived as
a par t of physical theory.
In recent year s student s and resear chers of engineering and physics have
begun to realize the utility of variational principles and the vast possi-
bilities that they offer, and have applied th em as a powerful tool for the
st udy of linear and nonlinear problems in conservative and nonc onservative
dyn amic al systems.
The present book has evolved from a series of lectures to graduate stu-
dents and resear chers in engineering given by the aut hors at t he Depart-
ment of Mechani cs at the University of Novi Sad Serbi a , and numerous
foreign univ ersities.
The object ive of t he aut hors has been to acquaint t he reader with t he
wide possibilities to apply vari ati onal principles in num erou s problems of
cont emporary an alytical mechanics, for example, th e Noether theory for
findin g conservatio n laws of conservat ive and non conservative dynamical
syste ms, application of the Hamilton-Jacobi method and the field method
suitable for non conservative dynamical systems, the variational approach to
the mod ern op timal control theory, the application of variationa l methods
to sta bility and determining th e optima l sh ap e in th e elastic rod theory,
among others.
In order to reach a level of pr actical effectiveness , numerou s concrete ex-
amples are solved in orde r to clarify the vitality of t he t heory. It is hop ed
t ha t this book will be useful as a text in graduat e and senior und ergraduate
cours es wit h an emphas is on mechanics and/or applied ma thematics and
in graduate engineering courses. The exposition is intended to be sugges-
ti ve rather than (mathematically) rigorous . Thus, the mathematical level
has been kept as elementary as possible. Each chapter starts from widely
understood principles and brings the reader to th e forefront of the topic
in a logical way. An important part of the material present ed was already
publish ed by the a ut hors of th is book in th e numerous papers printed in
the current liter ature for th e last 10 or so years, and th e reader is dire cted
to t hese sources at the proper places in the text .
The book is divided int o two parts. The first part contains four chapters .
In t he first chapter we consider the basic form s of the Lagrange-D 'Alemb ert
x P reface

principle in the form of the central Lagrangian equa tion, Euler-Lagrangian

different ial equations of motion for holonomic and nonholonomic dynamical
sys te ms, the Hamilton ca nonical equations, canonical transformations, and
Poisson's brackets.
The second chapter is devoted to the Hamilton-Jacobi method of inte-
gration of canonical equa t ions. Special attention is paid to the an alysis of
rh eolinear oscillations a nd quadratic conservat ion laws of rh eoline ar sys-
te ms wit h two degrees of freedom .
In the t hird chapter we study methods of obtaining the conservation
laws of conserva t ive and non conserva tive dyn amical systems by means of
Noether theory. The necessary condit ions for th e existence of conserva-
tion laws are obtained by st udy ing the invariant prop erties of the central
Lagr angi an equation with respect to the infinitesimal tr ansformations of
generalized coordinates and tim e, in the presence of the gauge fun ction.
Generally, the generators of space and time transformations ar e supposed
to dep end upon time, generalized coordinates, and generalized velocities.
In the fourth chapte r, we consid er a field method suitable for applications
in conservative and nonconservative dynamics. The essence of the method
is the supposit ion th at one component of the momentum vector can be
repr esented as a field function dep ending on time, generalized coordina te s,
and the rest of th e components of generalized momenta.
The second par t of t he book is devot ed to t he Hamil ton integr al vari a-
tional principle, and its various applica tions. It cont ains four chapte rs.
The fifth chapter is the introductory character for this part.
The sixth chapter cont ains the variational problems subject to natu-
ral boundary conditions, vari able end points, the Bolza problem, and the
J acobi form of the variation al principle describing the trajectories of con-
serva t ive dynamical systems.
Chapter 7 discusses constrained problems and the variational approach
to opt ima l control theory. The various specified and natural boundar y con-
ditions are dis cussed in detail.
Ch apter 8 contains applicat ions of vari ational methods to t he problems
of elastic rod theory. The vari at ional methods are used to estimate t he
critical load of elastic columns, to det erm ine post critical shape (t he sha pe
aft er buckling) , and to det erm ine th e optimal shap e of elast ic rods and
columns. By optimal we mean rods of minimal mass (volume) for specified
buckling load .
We are grateful to Prof. Dragan Spasic and assistant Mrs . Br anislava
Novakovic for helping us in preparing the manuscript.
The book is gratefully dedicated to our children and gr andchildren : Mil-
ica, Dragutin, Milena, Djordje Mihajlo and Bozidar (B.D.V.) and J elena
and Milica (T. M.A.)
B. D. Vujan ovic
Novi Sad , June 2003 T. M. At anackovic
 An Introduction to
Modem Variational Techniques
in Mechanics and Engineering
 Part I

Differential Variational
Principles of Mechanics
Chapter 1

The Elements of Analytical

Mechanics Expressed Using
the Lagrange-D'Alembert
Differential Variational
Principle

1.1 Introduction
The t ext mat erial of the pr esent chapte r is design ed to be a more or less self-
cont ained introduction to analyt ical mechanics expressed in an invar iant form
t hat is not connected to any privileged coordinate syste m . To accomplish this
goal we turn first to the Lagr ange-D'Alemb ert differential variational principle,
whose applica tions ar e very wide and encompa ss holonomic and nonh olonomi c
dyn amical syste ms and also conservative and purely nonconservative systems as
well. The elements of this part of contemporary analyt ical mechanics in fact,
constitute the content of this chapter.

1.2 Differential Equations of Motion

in Cartesian Coordinates
1.2.1 Free Dynamical Systems
We comm ence our considerations by rega rding t he simplest dyn amical system
consist ing of N materi al particles th at are completely free to move in a Cartesian
inertial coordinat e syste m Oxyz. Let us denote the position vector of the i th
particle by r i=xiel + Yie2 + Zie3, where el , e2, and e3 are the unit vectors of

B. D. Vujanovic, et al., An Introduction to

Modern Variational Techniques in Mechanics and Engineering
© Birkhäuser Boston 2004
4 Chapter 1. The Elements of Analy tical Mechanics

t he axes Ox , ay, and Oz , resp ectively. In every problem of particl e dynamics

the act ive (applied), impressed forces F , = Fxiel + F yie2 + Fzie3, acting on
the ith par t icle should be given in adva nce. Generally, t hese forces are the
functi ons of time t , positi on vect ors ri , and velocity vectors Vi = dr;/dt = r i =
:tie l + Yie 2 + Zie 3, namely

Fi = F i(t ,rl , ..., r N, v I, .." VN) , i = 1, ..., N . (1.2.1)

Denoting by a, = ~ r;/dt2 = r i = Xie l + Yie2 + Zie3 the acceleration vector of

th e i t h particle, and applying the second Newton's law of motion, we arrive at
the following simultaneous syst em of differential equations of motion of a free
dynamical system:

m iri = F, (t,rI, ..., r N, VI, ..., VN) , i = 1, ..., N , (1.2 .2)

where mi denotes the mass of the ith particle.

For the motion of a free dynamical system, Newton's law supplies all the
dynamical information that we need . Namel y, the problem of find ing the motion
of every particle of th e dynam ical syste m r, = r, (t ) , i = 1, ..., N, is reduced to
that of integrati on of a set of N , vectorial (or 3N sca lar) differenti al equat ions
of t he second order (1.2.2). If we are able to integrate t he syste m (1.2.2) , we
find t he positions r, of each particle at time t if the values r, (to) and Vi (to) ar e
prescribed in advan ce at th e initial moment t = to.

1.2.2 Constrained Motion. Lagrangian Equations with

Multipliers
Frequently, the particles of a dynamical syst em are not complete ly free to move
in t he physical sp ace, but are rather forced to be in permanent contact with
some mat eri al obj ect s th at can be describ ed in a mathematic al form (for exam-
ple, fixed or moving surfaces, curves, et c.). Such limitations to the freedom of
motion are known as constraints, and they are specified by certain geomet rical
or kinematical relations.
Constraints may be classified in various ways, I and we shall use here the
simplest, but very important type of const raints named holonomic constraints,
which are of purely geomet rical charac ter and can be expressed as

f s(t , XI ,Yl ,ZI, ..., XN, YN,ZN) = 0, s = l, ...,k, wherek <3N. (1.2 .3)

The explicit dep endence on tim e in these relations means that physically the
const raints are in motion. Such constraints are usu ally referr ed to as rheonomic
or nonstationary, in cont rast to the cases when they are fixed in space or scle-
romic or stationary const ra ints, that is, they do not depend on time t explicitly,
namely 8fs/&t = O. It is to be not ed t ha t the case k = 3N is not of any int erest
IThe reader can find a rat her ex haust ive classificatio n of co nstraints as , for example, non -
holon om ic cons tmints. bilatera l, un ilateral, etc.• in t he monogr ap hs of P a rs [84], Sant illi [951.
a nd Papastavri dis [821. P a pastav rid is has also co nsi de red se rvo const raints.
1.2. Differential Equations of Motion 5

since we could solve the complete system (1.2.3) and find all 3N coordinates
Xi, Yi, Zi, (i = 1, ..., N) as functions of time t, which means that the motion of
the dynamical system is given in advance.
If the dynamical system is completely mobile without restraints, all 3N co-
ordinates Xl,Yl,Zl, .. . ,XN,YN,ZN can vary separately, and such a dynamical
system is said to have 3N degrees of freedom. Naturally, the existence of con-
straints reduces the number of independent coordinates. In fact, we can use
the equations of constraints to eliminate as many coordinates as there are con-
straints. This would bring the number of coordinates down to the number of
degrees of freedom. Namely, we can eliminate k of 3N coordinates from (1.2.3)
and express them as functions of 3N - k independent coordinates. Then, it is
said that the dynamical system has 3N - k degrees of freedom. Consequently,
the minimum number of the geometrical parameters that uniquely determine
the position of the dynamical system at each moment of time is known as the
number of degrees of freedom. Also, we tacitly assume that the constraints
(1.2.3) are independent, that is, that they have been reduced to the least pos-
sible number, which implies that the functions fI, ..., fk are not connected by
a relation B (fI, ..., fk) = O. In many practical situations the elimination of the
k redundant coordinates can be tedious or difficult, and there are advantages
in retaining more coordinates than the number of degrees of freedom. We will
pursue this possibility in the next paragraph.
Since the particles of the dynamical system are compelled to be in permanent
contact with the given constraints, we have to suppose that, as the result of
interaction between the particles and constraints, there are forces of constraints
R, = Rxiel + Ryie2 + Rzie3 , i = 1, ..., N , acting on the particles. The differential
equations of motion in the presence of holonomic constraints are

miri = F i (t, rI, ... ,rN, vI, ..., VN) + Ri . (1.2.4)

In contrast to the active forces Fi' which are fully specified, the forces of con-
straint are not furnished a priori. They are among the unknowns of the prob-
lem and must be obtained from the solution we seek. On the other hand, it
is easy to see that the problem posed by 3N differential equations (1.2.4) does
not constitute a sufficient set of equations for finding 3N unknown coordinates
Xl, YI, Zl, .. . , XN , YN, ZN as functions of time and 3N unknown orthogonal pro-
jections of the reaction forces R x i, Ryi,Rzi, i = 1, ..., N .
In order to establish a consistent problem we have to introduce some addi-
tional assumptions about the character offorces of constraints R;. It is sufficient
to require that the constraints are smooth, that is, that the reaction forces R,
are directed toward the normal of the hypersurfaces (1.2.3) and the magnitudes
of reaction forces are not limited. In addition, it is of vital importance to un-
derline some kinematical properties of the constraints.
Let us differentiate the expression (1.2.3) totally with respect to time

~(afs
L !l X,.. +!l
afs., afs .) afs_O
Y, +!l Z, + !l - , s = 1, ... . k, (1.2.5)
i=l UXi UYi UZi ut
6 Chapter 1. The Elements of Analyt ical Mechanics

T he velocity vector of the ith particle Vi = Xie l +Yie2+ ,iie 3 satisfies this relation
by all possible velocities t hat th e dyn ami cal system might have. However , t he
velocity vector that is compat ible sim ultaneously wit h (1.2.4) and (1.2.5) will be
referr ed to as th e actual velocity vector. Equivalentl y, t he actual displacement
vector of th e i th pa rticle dri = Vidt = r idt = dXie l + dYie2 + dZie3, which
satisfies at t he sam e time the different ial expression
N
~ (Of' o f. o f.
~ -0dx, + -,:;-dYi + -,:;-dzi
) of.
+ -;;-dt = 0, s = I , ..., k, (1.2.6)
i=1 Xi UYi UZi ut

and th e differential equations of moti on (1.2.4) is said to be the actual displace-

m ent vector of th e it h particle.
Together with the act ual velocity and act ual displac ement vect ors, we shall
also introduce a new kind of infinitesimal displacement, usually referred as vir-
tual displacement or simply variation, which we denot e by

(1.2.7)

This kind of displacement is int roduced in such a way t hat of smoot h constraints
and notwithst anding of scleronomic or rh eonomi c syst ems, t he relations
N

L (-oBx,f . UXi + -ooY'f. OYi + -Ofoz .' OZi )

i= l t
J:

1. t
= 0, s = l , ..., k, (1.2.8)

are satisfied for t he arbit ra ry values of th e vector (1.2.7) at t he given instant

of tim e t . T hese displace ments are called virtu al to distin guish them from th e
actual dis placements dr i occurring in the tim e interval dt . They are the displace-
ments that would be poss ible at t he constraints (1.2.3) if they were petrified in
t he form t hat they have at the inst ant t . Not e also t hat t he virtual displace-
ments do not sat isfy the differential equations of motion , and they have purely
geomet rica l significance since they are not influenced by th e forces acting on
the par ti cles. Comparing (1.2.8) and (1.2.6) it is evident t ha t for t he case of
rheonomic systems (a f . I at i- 0) t he act ual displacement vecto r of the ith par-
ti cle dr, and the corres ponding virtual displacement vector ori do not coincide .
Moreover , if the constra ints are scleronomic (of.lot = 0) , it follows that both
vect ors are belonging to the same class of displacements. Nevertheless, even
in th e case of scleronomic syste ms we will make distinctions between these two
classes of displacements.
'Ve now restrict our selves to th e dynamical syste ms for which the total vir-
tu al work of th e forces of constraints defined as oA = 2:;:'1 R, . Ori is zero :
N N
oA = L R; . ori = L (RxioXi + R yioYi + R zioZi ) = 0, (1.2.9)
i=1 i=1
where we used · to de note a sca lar product of vecto rs. The condition (1.2.9) is
one of t he most imp ort ant properties t hat is fulfilled for t he case in which the
1.2. Differential Equations of Motion 7

constraints are smooth and the system is holonomic. Namely, the particles are
compelled to move on the constraints (surfaces, curves, et c.) and the reaction
forces are perpendicular to those surfaces, while the virtual displacement must
be tangent to them, hence the total virtual work vanishes . It should be stressed,
that for the rheonomic systems, the total work done by the forces of constraints
on the actual displacements dr, is not zero:
N
ER i · dr, =f O. (1.2.10)
i=l

Comparing conditions (1.2.8) and (1.2.9), we conclude that the reaction forces
can be expressed in terms of k multipliers >'5 (t) in th e following way:

k et, k et,
R Xi = 2:>'5-8. ' ti; = 2: >'5 a--:- ' i = 1, ..., N ,
5=1 X, 5=1 Y,
(1.2.11)

where the multipliers >'5 (t) are related to the magnitude of the forces of con-
straints. Therefore, the 3N differential equations (1.2.4),

(1.2.12)

become

mi X i

miZi (1.2.13)

These differential equations should be considered together with k equations of

constraints (1.2.3) :

(1.2.14)

Note that 3N differential equations (1.2.13) and k equations of constraints

(1.2.14) form 3N+k equa tions with 3N+k unknowns: 3N unknowns X l , y}, Z}, .. . ,
X N , YN, ZN and k unknown multipliers >'1 , ..., >'k . They express the equations of
motion for the general, holonomi c dyn amical system in a simple and conceiv-
able form . They are known as the Lagrangian equation of the first kind . Note
that equations (1.2.13), (1.2.14) are not quite useful in practical applicat ions,
since the pro cess of finding th e solution requires simultaneous treatment of the
Cartesian coordinates Xi, Yi , Zi , and also the Lagrangian multipliers >'5' In the
8 Chapter 1. The Elements of Analytical Mechanics

text that follows, we will transform these differential equations into a different
form in which the Ai do not appear.
As a simple illustration of the pre ceding theory, consider the triangular prism
with inclination a th at moves with constant speed V on a horizontal plane Ox.
A particle of mass m slides down the smooth inclined face AB under the force of
gravity. The particle in motion is in contact with the moving constraint, which
represents the straight line AB moving parallel to the right, as shown in Figure
1.2.1, with constant velocity V.

a B

x
Figure 1.2.1

The equa tion of constraint is therefore

kx +y - k (Vt + a) = 0, (1.2.15)

where BC = a and tan a = k. This equation represents a particular case of a

holonomi c, rheonomic constraint of the type (1.2.14), for s = 1. The particle has
one degree of freedom since the projections of the virtual displacement vector
are connected by the relation 8y + k8x = O.
From Figure 1.2.1 it is seen that the force of constraint R is permanently
normal to the moving constraint AB and the virtual displacement vector 8r
is orthogonal to R at each moment of time . It is also seen that the actual
displacement dr is not orthogonal to R, and the work done by this force on the
actual displacement dr is not zero.
The Lagrangian equations with multipliers (1.2.13) are of the form

.. .. \ aj \
mx = \ aj
A-
k
ax = >. ' my = -mg + A ay = A - mg, (1.2.16)

where 9 is gravitational acceleration.

From (1.2.15) it follows that, ki = -jj , and combining this with (1.2.16) we
easily find the components of the acceleration vector and multiplier A:

= M ' y·· -- - ~ - N, mg
i = -.!5JL
1 + k2 1 + k2 - A = 1 + k2 · (1.2.17)

If the initial conditions are given in the form

x (0) = 0, y (0) = h, x (0) = V, Y(0) = 0, (1.2.18)
1.3. An Invariant Form of Dynamics 9

we find the motion of the particle

x
Mt 2
= -2- + Vt, y=h-- .
Ne (1.2.19)
2
With (1.2.19) we can verify that the constraint (1.2.15) is identically satisfied.
According to (1.2.11), the projections of the reaction force Rare
k
Rx = k): = mg 1 + k 2 ' (1.2.20)

and the total force of constraint is

1
R = ( R2x + R2)
y
1/2
= mg 1/2' (1.2.21)
(1 + k 2 )

1.3 An Invariant Form of Dynamics, the

Lagrange-D'Alembert Differential
Variational Principle for Holonomic
Dynamical Systems
The differential equations of motion of holonomic dynamical systems can be
described in a variety of mutually different forms, depending upon the coordi-
nate system we employ. In the previous section we have derived the differential
equations of motion in Cartesian coordinates. As mentioned already, the equa-
tions on the form (1.2.13) are not generally feasible when working with the
system with many degrees of freedom. Namely, in order to find 3N - k in-
dependent coordinates as functions of time, we must solve 3N + k equations
consisting of 3N Cartesian coordinates Xi, Yi, Zi and as many Lagrangian mul-
tipliers As, (s = 1, ..., k) as the number of holonomic constraints figuring in the
system.

1.3.1 The Principle

In this section we introduce a single invariant expression known as the Lagrange-
D'Alembert differential variational principle from which, due to its generality,
we can derive analytical mechanics as a part of physical theory independently
of any coordinate system we use, which is free of the unknown Lagrangian
multipliers As and reaction forces Ri.
Let us consider N differential equations of motion (1.2.4) together with the
equations of constraints (1.2.3):

mii\ = F, (t, rI, ..., rN , VI, ..., VN) + R;, i = 1, ...,N, (1.3.1)
and

(1.3.2)
10 Chapter 1. The Elements of An alytical Mechanics

where the forces of constrai nts It; sa tisfy the condition of ort hogona lity (1.2.9):
N
LIt; · lir; =0. (1.3.3)
;= 1

Multiplying (1.3.1) by th e virt ua l displacement vector (i.e., by forming a scal ar

product ) and summing over i, we immediately arrive at the following equa-
tion which , toget her with the const raints equations, expresses the Lagrange-
D'Alemb ert variational principle:
N
L (m;r; - F ;) . 8r; = 0, (1.3.4)
i= l

Is (t , X I, YI , ZI, ... , XN, YN , Z N ) = 0, S = 1, ..., k. (1.3 .5)

Or , written in coordinate form , th e equation (1.3.4) reads

N
L [(mi x; - Fr;) lix; + (m;y; - Fy;) 8y; + (m;z; - Fz ;) liz;] = 0. (1.3.6)
;=1

The significance of the Lagrange-D 'Alembert principle can be summa rized

in the following few remark s:
(i) This principle is form ulated as a scalar product , which is one of th e most
fund am ental invar iant s used in physics and geometry, since t he scalar product
does not depend on t he coordinate system used bu t exclusively on t he vect ors
themselves.
(ii) By postulating scalar invar ian t (1.3.4) we actually replaced N vectorial
differential equations of motion (1.3.1) by a single scalar equa tion.
(iii) The differ enti al expression (1.3.4) cont ains t he to tal work of active forces
L:~I F; -br , and the unknown forces of const raint s It; do not figure int o it . One
of t he imp ortant advant ages of t he Lagrange-D'Al embert principle is the fact
t hat t he act ive forces ente ring into equation (1.3.4) are not limited in structure.
Nam ely, they can be pot ential or purely nonconservative forces.
Besides classical mechanic s, th e Lagr ange-D 'Alembert variational principle
can be employed as a st arting point in different br anches of physics that ar e
not intimately conne cted with classical mechanics. Nam ely, in some sense , it
plays a unifying concept in physics due to its invariance and also du e to the
st ruc t ural similarity of many physical manifestations with the models of classical
mechani cs. At this point it is ofi nte rest to invoke t he rem ark of W. Heisenberg
[53, p. 49], that "t he concept of classical physics will always rem ain the basis
for any exac t and ob jective science ."
For example M. A. Biot [24) extended the applications of th e Lagrange-
D'Alemb ert prin ciple to nonlinear nonstationary heat conduct ion processes, and
V. V. Dobronravov [39] applied this pr inciple to the elect ric machine theory, to
mention just two exa mp les.
1.3. An Invariant Form of Dynamics 11

1.3.2 Generalized Coordinates and Their Variations

Since the Lagr ange-D' Alembert principle is invariant with respect to the arbi-
trary coordinate system we can, instead of the Cartesian coordinates used so
far , introduce coordinates of more gener al type. Namely, we shall introduce new
geometrical parameters qI (t), ... , qn (t), whose number is equal to the number of
degrees of freedom , that is, n = 3N - k (where N denotes the numb er of particles
of the system and k is the number of holonomic constraint s). These parameters
are known as the generalized ind ependent coordinates and they uniquely deter-
mine the configuration of the dynamical system at the given moment of time.
By the term independent we understand th at the set qI , ..., qn is the minimal
number of coordinates that are potentially able to specify the position of the
dynamical system.
From the definition of the generalized coordinates, it follows that they must
satisfy the following two requirements. First, the position vectors of each par-
ticl e must be uniquely express ed in terms of generalized coordinates qs (t) , s =
1, ... , n, and time t,
ri=ri(t ,qI , ..., qn ) , i=l , ..., N , n=3N-k , (1.3.7)
or
Xi =Xi (t, q1, ..., qn ) , Yi=Yi(t,qr, ... , qn ), Zi=Zi(t ,q1 , ... , qn ) . (1.3 .8)
Second, the equations of constraint (1.2.3) must be satisfied identically by
the equations (1.3.8), which means that the sets of independent coordinates
qs , s = 1, ..., n , contain the constraint conditions implicitly in the transforma-
tion condition (1.3.7) . Therefore
is (t , ... , Xi (t, q1, ..., qn) , Yi (t, qI , ... , qn) , z, (t , q1, ... , qn) , ...) == O.

i = l, ..., N s = l, ...,k. (1.3.9)

To illustrate this, we turn to the example considered in the previous section
and shown in Figure 1.2.1. Let us introduce as the generalized coord inate the
distance AM = q. The position of the particle in terms of this coordinate is
X=Vt+qcosex , y=h-qsinex. (1.3.10)
Entering with this into the constraint equation (1.2.15), kx+y-k (Vt + a) = 0,
we verify that the identity of the typ e (1.3.9) is satisfied for k = tan ex and
h = a tan ex .
By using (1.3.7), we can tr ansform the Lagrange-D'Alembert principle in
terms of the generalized coordina tes. Differentiating (1.3.7) totally with respect
to tim e we find the following expression for the velocity vector of the ith particle
in terms of generalized coordinates qi and generalized velocities tj;:
. dri n ori. ori
r, = Vi = -;It = L -a
qs + -at' i = 1, ..., N; (1.3.11)
s=1 qs
12 Chapter 1. The Elements of Analytical Mechanics

that is, the velocity vector is the linear function of the generalized velocities.
Since the quantities ari/at and ari/aq5 depend only on the generalized co-
ordinates and time, it is easy to verify that the following two functional relations
are valid :

i = 1, ...,N, s = 1, ... , n , (1.3.12)

and
d ari aVi
-- = - , i = 1, ... ,N. (1.3 .13)
dt aq5 Bq,
The virtual displacement vector (or variation) of the ith particle, according to
(1.3.7), is of the form
n ari
8ri = L a8q5'
5=1 q5
i = 1, ...,N, (1.3 .14)

where the variations of the generalized coordinates are denoted by the symbol
8q5 '
As mentioned previously, the variation as a differential operator does not
produce any infinitesimal change upon time, that is, 8t = O. The variation of
velocity (1.3 .11) is found to be

(1.3 .15)

At the same time, the total time derivative of (1.3.14) (after simple interchange
of dummy indices) is of the form
d n " a2 . n a· d " a 2.
dt 8ri = LL a ~. 45 8qp + L
art dt (8q5) +L a 5~t8q5 ' (1.3 .16)
5=1 p=1 q5 qp 5=1 qs s=1 q
The difference between the last two expressions gives

8ri - :t 8ri = i: ~;:

s=1
(84s - ~ (8qs») . (1.3.17)

It is to be noted that in the classical variational calculus and integral varia-

tional principles of Hamiltonian type, it is generally accepted that the following
commutative rules are valid:

84s - :t (8q5) = 0; (1.3.18)

that is, the variation of velocity is equal to the velocity of variation. However,
in the realm of the differential variational principles and especially in the for-
mulation of the Lagrange-D'Alembert differential principle, the commutative
rules (1.3 .18) are not obligatory. Readers can find a very exhaustive discussion
concerning the commutative and noncommutative rules in analytical mechanics
in [68], [751, and [82] .
1.3. An Invariant Form of Dynamics 13

1.3.3 The Lagrange-D'Alembert Variational Principle

Expressed in Terms of Generalized Coordinates,
Central Lagrangian Equations
Since by introducing the independent generalized coordinates the constraint
equat ions are elimina te d (see (1.3.9)) , we substitute (1.3.14) into (1.3.4) and
permute the sign of summat ion:

n [N ..
~ ~m or o·- -~Fo .- 6q =0
ori N ori]
(1.3.19)
00"0 0 '0 s •
s =1 i= 1 q. i=1 q.
Transforming the first term in the following way and using relations (1.3.12)
and (1.3.13), one has
N
N •. ori
Lmiri ' - !:.- ~ m 'r o. ori
_ ~ nui» !!:.- ori
i= 1 oq. dt 0 " 0 q. 0i=1 " dto q.

d ~ . ori ~ . ori
dt 0 mir i ' q. a:- - 0 i=1
mir i '
q. a' (1.3.20)

Let us introduce the kinet ic energy of the dyn amica l system in terms of the
generalized coordinates

rN = r N(t,q,q)
wher e q = {ql , ..., qn} , q = {(1I , ..., qn} . Differentiating this function partially with
respect to generalized velocit ies q. and generalized coordinates q. we have

(1.3.22)

Thus, the equa tion (1.3.20) reads

N
Lmiri ' ~ri = dd ~T _ ~, s = 1, ...,n. (1.3.23)
i=1 oq, t vq. oq,

The second term in the brackets of equation (1.3.19) is usually term ed the
genemlized force of the corresponding genera lized coord inate q., s = 1, ..., n :

Q. =~ ori ~ (FO
0Fi . -,;- = 0
Xi 0Yi OZi)
xi-';- + Fyi-';- + F zi-';- , s = 1, ... , n .
i=1 vq . i=1 oq; vq. vq.
(1.3.24)
14 Chapter 1. The Elements of Analytical Mechanics

Since, according to (1.2.1), the active forces F, are functions of position ri,
velocities Vi, and time t, and in accordance with (1.3.7) and (1.3.11), these
vectors depend upon q's and q'e. Using the introduced notation, the virtual
work of the active forces and its transformation to the generalized coordinates
can be represented as

N n

OA(F) = 2:Fi ' <Sri = 2: Qs<Sqs. (1.3.25)

i=l s=l
From this, it follows that a generalized force can be interpreted as a coefficient
of the independent variations of the generalized coordinates <Sq1, ..., <Sqn in the
expression of the virtual work of the applied (impressed) forces.
Generally, the generalized forces are functions of the time t generalized co-
ordinates qs and generalized velocities qs:

Qs = Qs (t, q1, ..., qn, 41, ...,4n) , s = 1, ..., n, (1.3.26)

and they belong to the class of purely nonconservative forces.

As the case of special interest let the active forces F, = Fxie1 +Fyie2 +Fzie3
(i = 1, ..., N) not depend on the velocities Xi, Yi, Zi but are functions of Xi, Yi, z,
and the time t. It can happen at the same time that this type of forces is derivable
from a single scalar function usually referred to as the potential function

(1.3.27)

in the following way:

all all all

FXi=-~,
UXi
F yi = -~,
UYi
FZi=-~ '
oz;
(1.3.28)

In this case the generalized forces according to (1.3.7) and (1.3.24) can be written

Qs = _ (all OXi + all 0Yi + all OZi) , i = 1, ...,N, s = 1, ... ,n,

OXi Bq, 0Yi oqs OZi Bq,
(1.3.29)

which is actually the same expression for the partial derivative of the function
- I1 (t , Xl> Y1, Zl, ..., XN,YN, ZN) with respect to qs:

Qs
all
= --, (1.3.30)
oqs

where the potential function is expressed in terms of qs by means of the relation

(1.3.8) , that is,

(1.3.31)
1.3. An Invariant Form of Dynamics 15

Let us suppose that t he dynamical syste m is subjected to n nonp otential

forces of t he typ e (1.3.26) and also to n potential forces (1.3.30). The expression
(1.3.19) becomes

t (!:...~ - or -
5=1 dt oqs oqs
Qs + OIT ) Sq,
8qs
= O. (1.3.32)

Since, according to (1.3.31), t he pote nt ial forces do not depend upon the gen-
era lized velocities qs th e last equation can be written in the form

(1.3.33)

where we introduced a new function L known as th e Lagrangian function or

kinetic potential or simply Lagrangian defined as th e difference between the
kinetic energy and potenti al function

L = L (t ,q], ..., qn,ql, ..., qn) = T (t ,q], ..., qn,ql , ..., qn) - IT (t ,ql , ..., qn) .
(1.3.34)

T he scalar equation (1.3.33) plays t he fund amen t al role in analytical mechanics

and is usu ally termed the central Lagrangian equation.
Note t hat t he cent ral Lagrangian equat ion can be transformed in a way
t ha t can be of interest for add it iona l considerat ions. In what follows we will
aba ndon th e commutat ive rules, "t he variat ion of generalized velocit y is equal
to the velocity of the genera lized variat ion" shown in equation (1.3.18).
Let us calculate the variation of t he Lagrangian function (1.3.34):

6L = i:
5=1
( oOL Sq, + o8~ 6qs + 8o~ (6qs)" -
qs qs qs
o8~qs (6qs)" ) • (1.3.35)

where we added and subtracted t he term (oL j oqs) (6qs)" . By using the iden-
ti ty (oL j oqs) (6qs)" = 1; [(oLj8qs) (6qs)] -1; (oLj8qs) Sq, th e last equa tion
becomes

6L = ~[(8L d 8qs
~ Oq5 - dt OL) Sq, + dtd (8qs
8L ) OL (d .)]
Sq, + 8qs dt Sq, - 6qs .
(1.3.36)

Employing the central Lagrangian equa tion (1.3.35) we find

(1.3.37)

Let us introduce t he generalized momentum vector Ps defined by the equat ion

oL
Ps = oqs ' s = 1, ..., n . (1.3.38)
16 Chapter 1. The Elements of Analytical Mechanics

Note that this (covariant) vector will play a very important role in the subse-
quent text. Finally, with (1.3.38), the equation (1.3.37) becomes
d n n n
dt ~Ps8qs = 8L + ~ Qs8qs + ~Ps [(8qs)" - 8qs]. (1.3.39)
s=1 s=1 s=1
This is the second form of the central Lagrangian equation. According to our
best knowledge the form of this equation was first published by Lurie [68, p.
257] and Neimark and Fufaev [75, p. 133J.
It is of interest to note that the variational equation (1.3.39) can be success-
fully employed if the commutativity rule (1.3.18) is accepted. For this case we
have

(1.3.40)

It is of importance to note that the derivation of the differential equations of mo-

tion do not depend on whether we accept the commutative or noncommutative
variational rules .

1.4 Euler-Lagrangian Equations

The central Lagrangian equation (1.3.33) is valid for an arbitrary set of gener-
alized coordinates qI (t) , ..., qn (t) for which we suppose that they are mutually
independent. Consequently, the corresponding virtual displacements 8QI, ..., 8qn
are mutually independent and (1.3.33) can only be satisfied if the following n
equations hold:

5!.. 88T _ [)T + 8IT _ Qs = 0, s = 1, ..., n . (1.4 .1)

dt qs 8qs 8qs
These equations are called the Euler-Lagrangian equations, or frequently La-
grangian equations of the second kind. They are valid for all holonomic dynam-
ical systems. Equations (1.4.1) are the ordinary differential equations of second
order with respect to the generalized coordinates qs. To show this, we write
(1.4.1) in the explicit form, taking into account that kinetic energy, according
to (1.3.21) , is a function of t,qs, and qs:

(1.4.2)

where the potential function IT and generalized forces Qs have the structure
given by (1.3.21) and (1.3.24), respectively.
Note that we have used the generally accepted summation convention, which
means that whenever an index occurs two times in a term, it is implied that the
terms are to be summed over all possible values of the index. Thus, in (1.4.2)
1.4. Euler-Lagrangian Equations 17

t he summation is perform ed wit h respect to t he dummy index m. For the rest

of this book the summation convention will be perm anently assu med (applied).
From the syste m of equations (1.4.2) it follows t hat t he generalized accelera t ions
qs ente r in these equations linearly. We will suppose that t he equations (1.4.2)
are solvable wit h resp ect to t he genera lized accelerations, and to do so, t he
following det erminant must be different from zero:

(1.4.3)

It is easy to see that if we int egrat e th e Eul er-Lagrangian equat ions of motion
and find the generalized coord inates as functions of time ql (t ) , ..., qn (t ), then
the moti on of the dyn ami cal syste m is complete ly determined. Ind eed , ent ering
with qs (t) into (1.3 .5) we det ermine th e motion of each particle r i = r, (t) , i =
1, ..., N , and the reaction forces follow from (1.3.1) as R , = m ifi (t) --:" F i. 2 From
the pr evious analysis it follows tha t th e Euler-Lagrangian equat ions play the
cent ra l role in the study of moti on of holonomic dynamical syste ms, since they do
not contain redundant coord inates, th ey are free of equat ions of constra ints, and
they are ind ependent of unkn own Lagrangi an multipliers and reacti on forces. In
gene ra l, t he Euler-Lagrangian equations can be considered as t he corne rst one
of t he whole of ana lytical dynamics, and according t o Pars [84, p. 76] t hey
are "right ly regarded as one of t he outstanding intellectu al achievements of
manki nd ."
For the sake of completeness , note that t he Eu ler-Lagrangian equat ions
(1.4 .1) can be written in t he form
d 8L 8L
dt 8qs - 8qs = Qs, s = 1, ...,11, (1.4.4)

which follows dir ectl y from (1.3.33). Here th e Lagrangian function L is defined
as t he difference between th e kineti c energy and potential funct ion, as indicated
by (1.3.34). It is of special interest to not e that in th e case for which t he genera l-
ized nonconservative forces are equal to zero, Qs = 0, that is, the case in which
all active forces acting on t he dynamical system are of t he potent ial charac ter ,
the Euler-Lagran gian equations of motion depend solely on one function L :

.!!:.- 8L _ 8L _ 0
dt 8qs Bq, - ,
s = 1, ..., n . (1.4.5)

We ca ll this typ e of dyn ami cal system th e Lagrangian dyn amical system. The
fact t hat the Lagr an gian function depends explicitly on th e tim e as indic ated
by (1.3.34) signifies that the dynamic al syste m is subjected to th e rheonomic
constraints given by equations (1.2.3). However , if t he Lagrangian func tion is
form ed as the difference betwe en t he kineti c energy and potenti al function which
does not depend explicit ly upon ti me t, t hat is, if

(1.4.6)
2U nder li ned indices shou ld n ot be summed .
18 Ch ap ter 1. The Element s of An alytical Mechanics

which we call the potential eneryy function, and also if the kinet ic energy does
not depend on time t , t hen we call th e dynamical syste m characterized by t he
Lagrangi an

(1.4.7)
the conservative dynam ical system. Not e that the Lagr angi an dyn amic al system
whose Lagrangian function is given by L(qll " " qn, (h , ...,iln, t) Le.BL / Ot f. 0 is
also nonconservative since the tot al energy is not conserved . This fact will be
discu ssed in the proc eeding paragraphs.

1.4.1 The Structure of the Kinetic Energy. Explicit Form

of Euler-Lagrangian Equations
In this section we will briefly consider the structure of the kinetic energy of a
holonomi c dynamical syste m in terms of the generalized coordinates and the
structure of the Euler-Lagrangian equations whose form considerably depends
up on th e struct ure of th e kinetic energy.
As mentioned previously, the total kinetic energy of a dynamical syst em is
defined as
1 N
T=2 L m ;v r. (1.4.8)
;= ]

At t he sa me t ime t he velocity of th e ith particle in term s of th e generalized

coordinates qs and generalized velocities qs is, according to (1.3.11),
Br; . Br;
v; = oqsqs + 7ft ' i = 1, ... , N , s = 1, ... ,n. (1.4.9)

Enterin g with thi s into (1.4.8), th e kinet ic energy becomes

T = To + T] + T2, (1.4.10)
where

To ~m;(~;) 2 =F(t ,ql , ...,qn), i= I, ..., N , (1.4.11)

N
~ Br; Br; .
~m ·_·-q
;=] t Bq, ot s
K s(t ,q] , ..., qn) q.. s=I , ... ,n , (1.4 .12)
1 s ( t ,q] , ...,qn) qkqs,
2ak . . k , s= 1, ... ,n, (1.4.13)

where in t he last expression t he coefficients ask (t , q], ..., qn) are given by

Br; Br;
ask = aks = L m ;-Bq,
N

;= ]
.- .
8 qk
(1.4 .14)
1.4. Eul er-Lagr angi an Equ ations 19

It is evident from t he structure of th e equa tions (1.4.11) and (1.4.12) that, if

t he dyn amical system is scleronom ic, that is, ori!ot = 0, and if the constraints
are not moving constraints, t he t ime does not occur explicit ly in t he coefficient s
ask and To and T 1 are equa l to zero. In this case t he kinetic energy becomes
T = T2 = ~akS (ql' ..., qn ) qkqs. (1.4.15)

From this expression it is seen t hat th e kineti c energy for th e scleronomic dy-
nam ical systems repr esent s a homogeneous quadrat ic form with respect to th e
genera lized velocities qs .
Tak ing into account the form of the kinetic ener gy (1.4.11)-(1.4.13) let us
write th e Euler-Lagr angian equations. First, we calculate the following expres-
sion:

(1.4.16 )

Not ing that

[km , sJ qkqs , (1.4.17)

where t he symbol

! ( oakS oams _ oakm) = [m k ,sI ,

[k m , s 1 = 2 !> +!> !> (104.18)
oq-; u qk oq,
denot es the Kristoffel s ym bol of the firs t kind. The differenti al equations of
moti on (1.4.1) can now be written in the from

Q _ an _ of _ oKs _ Daks qk
s Bq, Elqs Elt at '
k, s,m 1, ... ,n, (1.4 .19)

which are t he explicit form of the Euler-Lagrangian equat ions we have been
seeking.
As mentioned pr eviously, t he generalized accelerati ons ente r into (1.4.19)
linearl y, an d equat ions (1.4.19) can be solved with respect to lis since according
to the eq uation (1.4.3) t he matrix A with elements ams is not singular; that is,

o2T I= det lams I -I O. (1.4.20)

det
Ioqmoqs
20 Chapter 1. The Elements of Analytical Mechanics

Since the matrix A = lams] is a square nonsingular matrix, let us multiply equa-
tion (1.4.19) by inverse matrix A-I = [a:;r1] and, taking into account that

for k i= r,
-1
aksa sr = Dkr = {O1 t:
lor k -_ r, (1.4.21)

where summation with respect to repeated indices goes from 1 to n , we find

..
qr + a,r
-1 [k ] ..
rn, s qkqm = a,r
-1 (Q s -
all of oKs / oak,.)
oqs - oqs - -rit - -at qk .
(1422)
. .

If we introduce the Kristoffel symbols 0/ the second kind by the relation

(1.4.23)
the equation (1.4.22) becomes

..
q; + rrkmqkqm
" _ -1
- asr
(Q _0 (IlBq,+ F) _ oKat s _ oaks
s
. )
at qk· (2 )
lA. 4

It is to be noted that the Kristoffel symbols of the first and second kind , (1.4.18)
and (1.4.23), play an important role in Riemannian geometry and geometrical
interpretation of classical mechanics. For an elaborate discussion of the appli-
cations of Riemannian geometry in classical mechanics, see, for example, [106]
and [71].
In the case of scleronomic dyn amical systems, the equations of motion (1.4.19)
and (1.4.20) are reduced to the simpler forms

..
aksqk+ [krn,s ] qkqm=
. . all'
Qs - 8 s= 1, ... ,n, (1.4.25)
qs
or
.. rr"
qr + kmqkqm
- 1
= a,r (Q s -
0 (Iloq,+ F)) ' r = 1, ... ,n, (1.4.26)

which are the explicit form of the Euler-Lagrangian equations of motion.

At thi s point it is of interest to note that as far as the Lagrangian dynamical
systems are concerned, the prescription for finding the Lagrangian function
L = T - Il is a very important and reliable way to determine the corresponding
Lagrangian function L. However, it is also often possible to find some alternative
Lagrangian functions besides those formed by this rule . This point can be made
by noting that th e given Lagrangian function L (t, ql, ..., qn, (it,..., <in) can always
be replaced by a new Lagrangian L * in the following way:

L*=cL + :/(t,qI, ... ,qn), (1.4.27)

where c is an arbit rary nonzero constant and / is a function depending on time

t and the generalized coordinates qj , ...,qn. The Euler-Lagrangian differential
equations formed by means of the Lagrangian function L, and L * are the same

(1.4.28)
1.4. Euler-Lagrangian Equations 21

Note that the function f (t, ql , ..., qn) figuring in equation (1.4.27) is frequently
referred to as the gauge function.
Moreover, in many instances it is possible to find a Lagrangian function L *
satisfying the relation (1.4.28) , where L* is not formed by the rule (1.4.27). As
an example, consider the simple harmonic oscillator whose differential equation
of motion is

ij + w2 q = 0 (w = given constant parameter) . (1.4.29)

Obviously, the rule L =T - IT is equivalent to

(1.4.30)

It is easy to verify that the time-dependent Lagrangian function of the form

L = ~ (q + qwtanwt)2 (1.4.31)

will generate the corr ect differential equation (1.4.29) . For w = 1 it is also
demonstrated in [92] and [96] that the following two Lagrangian functions,

L1 = ~ arct an ( ~) - ~ In [q2 (1 + :~) ] (1.4.32)

and

q
1
L 2 = 2" (qcost - qsint) In .
(qcost - q sin
.
q sm t + q cos t
t) , (1.4.33)

will also generate the sam e differential equation (1.4.29) and all three La-
grangians (1.4.31)-(1.4.33) are not formed by the rule L = T - IT or the pre-
scription (1.4.27) .
The fact that the form of the Lagrangian functions in dynamics is not unique
raises the question of finding functions L for a given holonomic dynamical sys-
tem whose differential equat ions of motion are given in advance. This important
problem, usually referred to as th e inverse Lagrangian problem , was first con-
sidered by Helmholtz in 1887 and later studied by numerous aut hors (see, for
example, [95] , [122], and [27]).
Finally, it is of interest to note that the Euler-Lagrangian equations are form
invariant with respect to any one-to-one (i.e., punctual) transformation of two
syste ms of gener alized coordinates; the "old" generalized coordinates q}, ... , qn
are tr ansformed to the "new" generalized coordinates QI, ..., Qn by

qi =qi(t,Ql , ...,Qn) , i= 1, ... ,n. (1.4.34)

Let us consider the given Lagrangian function L (t, ql , ..., qn,ql , ..., qn) . By using
(1.4.34) one has

(1.4.35)
22 Chapter 1. The Elements of Analytical Mechanics

For the analysis that follows we will need the following two identities, which are
similar to (1.3.12), (1.3.13):

aqi d aqi
(1.4.36)
aQj - dt aQ)'

Now we have
aL' aL aqj aL aqj
(1.4.37)
aQi = aqj aQi = aqj aQi ·
Hence ,

daL' ( daL) aqj et. (d 8qj)

dt 8Qi = dt 8qj 8Qi + 8qj dt 8Qi .
We also have

From th e last two equ ations, we have

(1.4.38)

Since the determinant det (8qj/8Qi) i- 0, the invariance of the Euler-Lagrangian

equat ions with respect to th e point transformation (1.4.34) follows.

1.4.2 Two Important Conservation Laws of the Euler-

Lagrangian Equations: Momentum and Jacobi
Conservation Laws

(i) Momentum Integral

The Euler-Lagr angian equations

d 8L 8L
----=0, s=l , ...,n , (1.4.39)
dt aqs Bq;

have a very suitable form for finding conservation laws (or first int egrals) of
th e dynamical systems whose behavior can be completely described by the La-
grangian function L(t, qi , ... , qn, qI, ..., qn). By the term conservation laws or first
integrals we underst and some specific functional relations between physical and
geometrical parameters figuring in dynamical systems, which are satisfied iden-
tic ally du e to the differential equations of motion of th e dynamical system in
question. The existe nce of conservat ion laws can considerably simplify the inte-
gration of the differential equat ions of motion. Before discussing two important
1.4. Euler-Lagrangian Equations 23

conservation laws that appear frequently in analytical dynamics, we introduce

the generalized momenta Ps by the relation

8L
Ps = Bq,' s = 1, ..., n . (1.4.40)

It can happen very often that there are generalized coordinates that do not
occur in the Lagrangian function L, although their time derivatives (generalized
velocities) do . Such coordinates are usually referred to as ignorable or cyclic
coordinates.
Let qj be an ignorable coordinate, where j is a fixed particular integer. In
this case 8L/8qj = 0, and from the equation (1.4.39) for s = i , it follows that
the momentum or cyclic integral

8L
Pj = -8' = const. (1.4.41)
qj

It is important to note that when L does not depend on qj (j-fixed), it is

invariant under the translation in the jth coordinate; that is,

7jj = qj + C, (C = an arbitrary constant). (1.4.42)

Note also that the existence of a momentum integral is a privilege of a particular

coordinate system in which the motion is studied.
(ii) The Jacobi Conservation Law. Energy Integral
To obtain the second, very important conservation law for the Euler-Lagrangian
equations (1.4.5), we form the total time derivative of the Lagrangian function
L and combine the result with the differential equations (1.4.5). Thus we have

(1.4.43)

From this, we find

d (8L , ) 8L (1.4.44)
dt 8qs qs - L = -fit'
If the Lagrangian function L does not depend explicitly upon time, that is,
8L/8t = 0, we arrive at the Jacobi conservation law (first integral)

8L.qs - L
-8' = E = const. (1.4.45)
qs
If the dynamical system can be represented by means of the kinetic and
potential energy, then the Lagrangian function is of the form

L =T-fI. (1.4.46)
24 Chapter 1. The Elements of Analytical Mechanics

Supposing that the kinetic energy is of the form given by equation (1.4.15),
T = (1/2) aijqiqj, and the potential energy according to (1.4.6) depends only
on qs, we have

aT.
?i7qi
..
= aijqiqj = 2T = ?i7qi,
aL. (1.4.47)
vqi v~

and the Jacobi conservation law (1.4.33) becomes

T + IT = E = const., (1.4.48)
which is a familiar expression for the conservation of the total mechanical energy:
the sum of the kinetic and potential energy is constant during the motion of the
system if the Lagrangian function is given in the form (1.4.46) and if it does not
depend explicitly upon time, namely aLlot = O. It can also be stated that the
existence of the energy integral is a consequence of the time invariance of the
Lagrangian function L with respect to the time translation

t=t + B, (B = an arbitrary constant). (1.4.49)

Example 1.4.1. Two masses on a string. As an illustration of the foregoing

theory, consider two masses ml and m2 that are connected by a weightless string
of length l. The mass ml can move freely on a horizontal plane, while the string
can move frictionless through a small hole 0 in the plane so that the mass m2
moves vertically. To begin, let at t = 0, the distance of the mass ml from the
hole 0 be r = ro, and the initial velocity vo of this particle be normal to the
string while the mass m2 does not move. The system has two degrees of freedom
and we select as the generalized coordinates the polar coordinates of the mass
mI, r, and 'P as, indicated in Figure 1.4.1.

ll-r
m1

Figure 1.4.1

At the initial moment t = 0, r = ro, and r (0) = O. The kinetic and potential
energy of the system are

T 21 [(ml + m2 ) r.2 + ml r 2'P. 2] ,

IT -m2g (l - r) m2gr + const. (1.4.50)
lA . Euler-Lagrangian Equations 25

The Lagrangian function is given by

L ="21 [( ml +m2 ) r·2 +mlr2'P. 2] - m 2g r. (1.4.51)

It is obvious that the generalized coordinate 'P is ignorable and the momentum
(cyclic) integral reads

P<p = et. 2· C
o<p = mlr 'P = = const. (1.4.52)

Since the Lagrangian function does not depend explicitly on time t, we have
also the energy integral T + IT = E, which in our case becomes

(1.4.53)

Since our dynamical system has two degrees of freedom, we can base our analysis
on two conservation laws (1.4.52) and (1.4.53), ignoring the Euler-Lagrangian
equations of motion.
At t = 0 the constant C is found to be C = mlrOvO, and from (1.4.52) it
follows that
. rovo
'P = -;:2 ' (1.4.54)

The total energy E is

(1.4.55)

Entering with (1.4.54) and (1.4.55) into (1.4.53) we find after simple calculation
the following relation depending only upon variable r :

(1.4.56)

The roots of the quadratic expression with respect to r in the square brackets
are found to be

(1.4.57)

Therefore, rl > 0, r2 < 0, and the particle ml will oscillate between the bound-
ary circles "o and rl. The equation (1.4.56) can be written in the form

2m2g 1
f2 = 2 (ro - r) (r - rl) (r - r2) = <I> (r). (1.4.58)
ml +m2r
26 Ch apter 1. The Elements of Analytical Mechanics

From th is equation it follows that <I> (r) must be positive during moti on . Since
the root r 2 is negative, th e expression

(1.4.59)

is always positiv e. The right-hand side of (1.4.58) wiIl be positive in the following
two cases:

(a ) rO<r< rl; (b) rl < r < ro . (1.4.60)

The trajectories of th e particles ml for the cases (a) and (b) are depicted in
Figure 1.4.2a and b.

.. . G
......•.........•.•.....

<~:~- - -
A
o
\.. \\ r
" .....
.................................

Figure 1.4.2

It is clear that the case (a) is the motion of the particle ml for which , at the
initial moment t = 0, the cent rifugal force of ml is greater than the weight of
the particle m 2, t ha t is,

(1.4.61)

and th e radius ro wiIl start to increase from the initial point A if

(1.4 .62)

Similarly, the motion of ml wiIl describe th e trajectory represented in Figure

1.4.2b if at t = 0 the weight of the particle m 2 is greater than t he centrifugal
force of t he pa rt icle ml , so one has the following condit ion for the initial velocity
vo:

(1.4.63)

and t he distan ce of the particle m l will st art to decrease from the initi al dist ance
corr espond ing to point A . Finally it may happen t hat the cent rifuga l force of
lA . Euler-Lagrangian Equations 27

the mass m1 is equal to the weight m29 of the particle m2, and we have the
motion of m1 along the circle of the radius ro for which '

(1.4.64)

Let us briefly describe the motion for the case (a). During the initial increasing
period from A to B the angular velocity <p given by (1.4.54) will decrease;
the velocity and centrifugal force are going to decrease also until the point m1
reaches the point B, where the weight m19 is equal to the centrifugal force. In
the subsequent motion from B to C the regime of motion is going to be the
opposite.
From equation (1.4 .64) it follows that
dr
dt=±--· (1.4.65)
J<I> (r)'
since <I> (r) > 0 we have to take the plus sign when r is increasing and the minus
sign when r is decreasing. For the case (a) the radius is increasing from ro to
r1 in accordance with the equation

dr
L
T
t- (1.4.66)
- TO J<I> (r)'

The time from A to B (Figure 1.4.2a) is therefore

T1
dr
t =
1
lTO J<I> (r)'
(104.67)

while the time needed to pass from the extreme point B to C is

TO
r J<I>(r) =
t2 =- l
r1 J<I>(r)
dr
= iTo
dr
1
t ·
(1.4.68)

Therefore, the period of this oscillatory motion from the point A to B and back
to C is finally

(104.69)

and the amplitude of the oscillation of the mass Tn1 is obviously

a = r1 - rO. (1.4.70)
3The circular motion of the p article ml and stability of the motion for which (1.4.64) is
satisfied is considered in [681, pp . 635-638.
28 Chapter 1. The Elements of Analytical Mechanics

The integral figuring in (1.4.69) is the elliptic type, and the integration can be
accomplished for given ro, r}, and r2.
It is of interest to note that for this type of oscillatory motion we can use
two equivalent formulas for the approximate period proposed by Pars ([84, p.
10]):

Tapprox = 1r (r1 - ro) 1/2 [(d<P)-1/2

dr _ - (d<P)-1/2]
dr _
T-TO r-Tt
(1.4.71)

or

Tapprox = 1r [A (ro) + A (rdl , (1.4.72)

where

A (r) = [lit (r)r 1/ 2 . (1.4 .73)

For the case (a), that is, ro < r < r1 ,

(1.4.74)

with
A = 2m2g (1.4.75)
m1+ m2
Finally, the second generalized coordinate <p can be found from the conservation
law (1.4.54)

(1.4.76)

1.4.3 On the Disturbed Motion and Geometric Stability

of the Scleronomic Potential Dynamical Systems
In this section we shall briefly discuss the problem of stability in the geomet-
rical sense , which is based upon Synge's famous work, "On the Geometry of
Dynamics" [1061 . Since the exposition that follows is based strictly on the
tensor calculus, we suppose that the reader is familiar with this mathematical
discipline, although a broad knowledge of this subject is not essential.
As we have seen in section 1.4.1 the kinetic energy of a scleronomic dynam-
ical system is given by equation (1.4.15): 2T = aijl'jil'}j. We suppose that the
configuration of the dynamical system is specified by n generalized coordinates,
which we now denote by (qi, ..., qi) . These coordinates define the so-called con-
figuration space Vn and introduce the metric of the configuration space in the
form

(1.4.77)
lA. Euler-Lagrangian Equations 29

which is a basic invariant of Vn . Note that the quantity aij (q1, ..., qj) represents
the covariant metric tensor in the configuration space Vn . Note also that the
quantity ds introduced by (1.4.77) is referred to as the basic line element in
Vn and configuration space defined in this way is also the Riemannian space.
Together with the covariant metric tensor aij we can also introduce the con-
travariant metric tensor aij (which is denoted by aij1 in equation (1.4.26)), and
we have that aika kj = 5{ , where 5{ denotes the Kronecker delta symbol and
is equal to unity or zero when i and j are equal or different, respectively. Let
us consider an arbitrary scleronomic dynamical system subject to generalized
forces Qi (ql, ..., qn) . Using the tensorial notation we can write the differential
equations of motion of this system in two different forms (already indicated
by (1.4.25) and (1.4.26)) . In the covariant form the differential equations of
motion are (nonconservative forces are equal to zero , Qs = 0, in (1.4.25) and
Qi = -8IT/8qi )
aiih + [jm,i]tfti'" = Qi; S = 1, ... ,n. (1.4.78)
As demonstrated earlier in this section (see (1.4.19)), these equations can be
derived from the corresponding Euler-Lagrangian equations.
The equations of motion can also be written in the contravariant form
;l+r;kljjljk=Qi=aijQj, r=l, ..., n . (1.4.79)
Note that the symbols rim, j] and r;k
denote the Kristoffel symbols of the first
and second kind, respectively, introduced earlier in this section.
After this introduction in the tensor notations, we now consider two mo-
tions: the undisturbed motion qi (t) and the infinitesimal disturbed motion
xi (t). These two motions take place along the neighboring curves in the con-
figuration space Vn . We suppose that the undisturbed motion qi (t) satisfies
identically the differential equations (1.4.79). We call these trajectories, accord-
ing to Synge, the natural trajectories C. Let qr, r = 1, ..., n, be a point P of
C and qr + z" the coordinates of the corresponding (simultaneous) point P* of
disturbed natural trajectory C., z" being infinitesimally small. We shall call the
vector with contravariant components x T the disturbance vector. If the distur-
bance vector between simultaneous configurations remains permanently small,
we say that the undisturbed motion is stable in the kinematical sense .
In order to obtain the differential equations of the disturbed motion, we
substitute the generalized coordinates qi by qi + xi in (1.4.79) and calculate the
corresponding Kristoffel symbols and generalized forces in the disturbed region
as
(104.80)
where the asterisk denotes the quantities that should be calculated at p •. There-
fore, expanding r.A and Q. i and retaining the first powers of small quantities
only, we have

r .jki = r ijk + X m 8x8m rijk' (1.4.81)

30 Chapter 1. The Elements of Analytical Mechanics

Thus, the equations of the disturbances are

··i k;j'm 8 r i 2ri 'j . k _ j8Qi

x +X'1 q 8x k jm + jkq X - X 8x j' (1.4.82)

which are usually employed in solving concrete dynamical problems.

As demonstrated by Synge the equations of disturbed motion written in
tensorial form shed light on the geometrical structure of the disturbed motion
theory,"
Let us introduce the so called absolute derivatives of the disturbance vector.
The first absolute derivative of this vector is given as

(a)
o xi . i r i . j k
6t =x + jkq X (1.4.83)

where :t.. (.)

( )

denotes the absolute derivative with respect to time formed with

respect to metric tensor aij. Differentiation with respect to time leads to

(a) ,
d 0 x'
dt Ot
X + riJ'kqooj Xk + riJ'kqii :Xk + -
ooi 8 ri'k'l;j q'm Xk
8q-
m J

+ (x + r Jk q.j xk) rimqr

m i .r' (1.4.84)

Substituting ijj from (1.4.79) in this expression we obtain the second absolute
derivative of the disturbance vector in the form [68, p. 624]

x.. i + q' j q·k (

88qk rijm + riks r-jm -
r:mr rjkr )
i · ' ·k i k'
+2rjkiflx +rjkQ xJ. (1.4.85)

Substituting Xi from this equation into the equation of disturbed motion (1.4 .82) ,
we have after simple manipulations the following equation of disturbed motion
in the contravariant form [106, p. 79J, [68, p. 625J:

(1.4.86)

4Concerning the geometrical interpretation of (1.4 .82) which follows, Synge wrote [106 ,
p. 78]. "The use of the tensorial notation is of greatest importance. The appearance of the
Riemannian curvature tensor makes it difficult to believe that similar results could be obtained
without the use of this method ."
lA. Euler-Lagrangian Equations 31

Note that the expression in square brackets represents the Riemann-Kristoffel

curvat ure tensor

8 ri
R kjm j -- 8qj 8 ri rs ri rims rsjk· (1.4.87)
mk - 8qm jk + mk js -

Therefore, the equation of disturbed motion in the contravariant form becomes

(a)
2 i
o&2 + Rkmji m
X .j ' k i xr Qi
X q q = X Vj , (1.4.88)

where
i 8Qi i k
\JjQ = -8. + rjkQ
qJ
(1.4.89)

denotes the covariant derivative of the generalized force.

It is convenient to write the equations of disturbed motion (1.4.88) in eo-
variant form
(a)
2 i
air o&2
X
+ R kmjrq·k·qj X m - X
m'<7
vm
Qr = 0 , (1.4.90)

where Rkmjr denotes the covariant form of the Riemann-Kristoffel tensor Rprmn
= a ps R rmn", The explicit form of the Riemann-Kristoffel tensor is
1 8 2a 8 2a 8 2a 8 2a
R rsmn - --
28 8
rn
-+ - sm
8qr aqn
- rm
- -aqSaqn
--- -sn-
aqr aqm
qs qm

+a pq ([rn ,p] [sm,q]- [rm ,p] [sn,q]) . (1.4.91)

Note that for the case of the potential forces, that is, Qi = -aITj8qi, where
IT = IT (qI ,..., c") is the potential energy, the equations (1.4.90) become
(a)

air
2 i
sot + R kmjrq·kq·j
2
X
X
m
= -X m ( a
2
IT
8q m8qr -
r srm 8qS
aIT)
= -X m'<7v m (8IT)
8qr .
(1.4.92)

We shall call the differential equations of disturbed motion (1.4.88), (1.4.90),

and (1.4.92) the Synge disturbed equations, which are all written in tensorial
(Le., invariant) form .
The Synge disturbed equations form a basis for studying the stability of
motion by geometrical means. In many practical situations the geometrical
method has been used with advantage in various problems of mechanics and
optimal control theory. The reason is that the Liapunov method of stability
analysis, in spite of the fact that it is the strongest and most full method, is
not invariant in all coordinate systems. Namely, the same motion can be stable
Random documents with unrelated
content Scribd suggests to you:
 Both fleets lay to during the night, to repair damages, and day-
break discovered them on opposite tacks, each in line-of-battle
ahead.
The Spanish had the weather gage, and still possessed eighteen
or twenty effective sail-of-the-line, but they made no attempt to
renew the action. Probably some of their ships were not in condition
to fight. The great Santa Trinidada was nearly out of sight to
leeward, in tow of a frigate. As it was necessary to keep the British
fleet together, Sir John Jervis sent no vessels in chase of her.
The whole Spanish line was standing to the northward, while the
British fleet, which—including the Colossus and Culloden, neither of
which was fit to take a place in the line—could muster but fourteen
ships-of-the-line, then took their four prizes and the Captain in tow,
and very slowly made their way southward.
The damage sustained from the contest by the British ships was
not so great as might have been expected, from the severity of the
contest. The only ship of theirs dismasted was the Captain, which
ship also suffered much in the hull.
The Colossus and Culloden were both very much cut up, and the
latter had suffered especially in the hull, and was very leaky. She
had only one carronade dismounted, however, and two first and two
second deck guns.
The loss of life among the British was comparatively small. Except
in the cases of the Colossus and Egmont, those ships which suffered
most in hull and rigging had most killed and wounded. The total for
the fleet was 73 killed and 227 wounded. Of course, these were only
the badly wounded; for it was not the custom, in those days, to
report the slightly wounded. It is, therefore, fair to consider the total
as about 400; an amazingly small number, considering the nature of
the action.
According to the Spanish accounts, ten of their ships, besides
those crippled, suffered materially, but not more than half of them
showed any signs of being at all crippled. The Santa Trinidada,
Soberano, Principe d’Asturias and Conde de Regla were very much
damaged.
 The damages of the prizes are better known. All four ships had
lost masts, and all were so hulled as to be very leaky. The San
Nicolas was badly on fire, but her captors extinguished it. Their loss
in killed and wounded amounted to about 1000.
The detached and confused state of the Spanish fleet at the
beginning of the battle, and the consequent partial and irregular
manner in which their ships came into action, would render any
statement of comparative force, by comparing the totals on each
side, very unfair.
It would be correct to say that the British line consisted of fifteen
ships-of-the-line, and the Spanish line (if it could be called so) of
twenty-five, and afterwards of twenty-seven, ships-of-the-line.
The Santissima Trinidada was a monster in size. She was built in
Havana, in 1769, as a 112-gun ship, except that she had greater
beam than was usual with that class. Some time about 1796 her
quarter-deck and forecastle were formed into a whole deck,
barricades built along her gangways, with ports in them, and she
was made into a flush four-decker, but was not really much superior
in force to the three-decked 112s.
The most striking feature in this victory is the boldness of the
attack. Another commander might have paused before running into
the midst of twenty-five sail with fifteen. If he had paused to weigh
the chances, the separated ships would have closed, and the
Spanish line then have been too compact to be attacked with hopes
of success.
Sir John Jervis, relying upon the character of his force, and
viewing with a general’s eye the loose and disordered state of his
enemy’s line, resolved to profit by it, attacked promptly, and
conquered. It cannot be said that he broke the Spanish line, for
there was no line to be broken. He simply chose the proper moment
for advance, had a leader who never flinched or fell back, and he
had all about him those who were emulous to follow so bright an
example.
On the other hand, the bold front he put on was calculated to sink
the hearts of those among the Spanish fleet who had little
experience of naval warfare. The Spanish fleet was not only in
confusion at the outset, but continued to be so; and some of their
ships undoubtedly fired into their comrades, while they were so
huddled together that if a shot missed one it was sure to strike
another of them.
Then the British were better sailors, and repaired damages more
quickly; and to many of them the battle was more like a rattling
game than a grim matter of life and death and national renown.
It is reported that the Captain actually expended all her shot in
this action, and when grape was needed for her 32-lb carronades,
used 7-lb shot as a substitute.
This at a short distance must have caused great execution.
When the Spanish Admiral at last formed his scattered divisions
into line, he found the British in equal, if not better, alignment; and
each side then drew off, the one to lament, the other to exult, over
the events of the day.
The Spanish were never accused of a lack of courage, either by
sea or on land, and their discomfiture appears to have been caused
principally by the worthlessness of the crews which manned their
ships. These were composed of pressed landsmen, and soldiers of
new levies, with a very few seamen in each ship. It has been
reported that these “poor panic-stricken wretches,” when called
upon to go aloft, to repair the damaged rigging, fell upon their
knees, and cried out that they preferred being sacrificed on the spot
to performing a duty where death seemed inevitable from more than
one cause. The numerical superiority of their guns seemed little in
their favor, for some of the San Josef’s were found with their
tompions in, on the side which had been engaged, after the battle
was over. Indeed, the numbers on board some of the Spanish ships
seem to have been rather a detriment to them.
A rather prejudiced writer says that if eight of their twenty-five
ships had been left at Carthagena, and had the five or six hundred
seamen they probably contained been substituted for twice that
number of raw hands, taken from the remaining seventeen ships,
the latter would probably have made a better stand; and the victory,
if achieved at all, have been at the expense of a much greater
number of lives in the British fleet. Whatever the fault of the crews,
the officers fought well. “Upon the whole, the victory off Cape St.
Vincent, although from its consequences pre-eminently great, from
its results, dispassionately considered, cannot be pronounced in an
equal degree glorious.”
At about 3 p.m. of the 16th the British fleet and the prizes
anchored in Lagos Bay. Here the Spanish prisoners, numbering
about 3000 men, were landed; and, a receipt being given by the
proper authority, were allowed to remain.
On the 23d, after riding out a gale of wind with much difficulty, it
blowing dead on shore, Sir John Jervis sailed, and in five days the
whole were in safety in Lisbon. It was remarked that the prizes,
under jury-masts, beat all the English ships in working into the
Tagus.
Great congratulations and celebrations took place at Lisbon, for
the Portuguese had every reason to rejoice at this victory, while in
England the news was met with immense enthusiasm. Sir John
Jervis was created a peer of Great Britain, by the title of Baron Jervis
of Meaford, and Earl of St. Vincent; with a pension of £3000 per
annum. Vice-Admiral Thompson and Admiral Parker were created
Baronets, and Vice-Admiral the Hon. William Waldegrave was
appointed to a lucrative post abroad.
Commodore Nelson, who had so often proved in his own person
that the danger of a bold enterprise required only to be met to be
overcome, was not mentioned in Sir John Jervis’ despatches, but
received the insignia of the Bath, and the freedom of the City of
London.
Thanks of Parliament were voted to the fleet, and gold medals
were given to all the flag-officers and captains, as on similar
occasions. The four Spanish prizes were commissioned, and retained
in service on the Lisbon station.
The gale which had assailed the British fleet in Lagos Bay caught
the remainder of Admiral Cordova’s fleet at sea. It dispersed his
ships, and prevented them from reaching Cadiz until March. Among
them was the huge Santissima Trinidada, which, being so much
injured by shot, was least able to stand bad weather.
 On the morning of the 28th of February, as she was striving to
regain the coast, the English frigate Terpsichore appeared in sight,
to the westward. Her captain knew of the battle, and divined, at
once, that the four-decker must be the Sta. Trinidada. He instantly
cleared for action, and bore down upon her, and began engaging, so
manœuvring that he kept clear of her broadside. The great ship had,
therefore, only her chasers with which to chastise the temerity of
her pigmy foe. The frigate kept her company until March 2d, doing
her considerable damage, and receiving some in return.
On that date twelve sail of Spanish men-of-war appeared, and the
Terpsichore hauled up for the Mediterranean.
Several ships from England joined the fleet, and the Admiral
cruised off Cadiz, with twenty-one sail-of-the-line, blockading
twenty-six Spanish ships, and the latter did not again appear at sea
that year.
Admiral Cordova, and his two divisional flag-officers, Montlez and
Merino, together with eleven captains, were brought before a council
of war, to answer for their conduct in the battle. Nothing appears to
have come of it, for the personal gallantry of the officers was beyond
all dispute.
One fact is certain, that a Spanish three-decker, bearing a Vice-
Admiral’s flag, did her best to cut through the line, between the
Victory and Egmont.
In cases of this kind the officers are too frequently made the
scapegoats of a blundering Administration.
 ENGLISH FLEET IN CANARY ISLANDS. A. D. 1797.

n addition to the blockade and bombardment of

Cadiz, with which Lord St. Vincent was carrying on
an active warfare against the Spaniards, he
detached two expeditions against Santa Cruz, in
the Canary Islands, in the more important of which
Rear-Admiral Nelson was not only repulsed, but
lost his arm, a model of which is still among the
trophies and ex-votos to be seen in the Cathedral
of that place.
On May 28th, 1797, Captain Hallowell, of the
Royal Navy, in command of the Lively frigate, with
the Minerve frigate in company, stood into the bay
of Santa Cruz, Teneriffe, and discovered at anchor
in the road an armed brig, which, as the frigate
approached, hoisted French colors.
The two commanders deeming it practicable to cut her out, the
boats of the frigates were next day manned, and placed under the
orders of Lieut. Thos. Masterman Hardy (who afterwards much
distinguished himself, and became an Admiral). At about half-past
two in the afternoon, Hardy, with three other naval lieutenants, and
one of the Lieutenants of Marines, in the boats of the Lively, and two
lieutenants of the Minerve, with her boats, and their respective
crews, made a very resolute attack upon the brig, as she lay at
anchor, and, in the face of a smart fire of musketry, boarded, and
almost immediately carried her.
This alarmed the town, and a heavy fire of musketry and artillery
was opened upon the brig, not only from every battery, but from a
large ship which lay in the road.
The lightness of the wind at the time retarded the weighing of the
brig’s anchor, and then made it necessary for the boats to take the
brig in tow. During nearly an hour an unremitting fire was kept up
from the shore and ship. At length, at a little before four o’clock,
they succeeded in getting the vessel out of gun-shot. She was the
French national brig, Mutine, mounting fourteen guns, twelve of
them long 6-pounders, and the remaining two brass 36-pounder
carronades.
She had on board 113 men, the rest of her ship’s company, with
her captain, being on shore at the time of the capture.
Hardy, in effecting this handsome capture, did not lose a man, but
had fifteen wounded.
The Mutine was a remarkably fine brig, and was put in
commission by Earl St. Vincent; and the command of her given to
the officer in command of the party that cut her out. Lord St.
Vincent set an example which was not followed by all other
commanders in-chief, in those stirring times. “He appointed, and
gave out that he would always appoint, to the command of any of
the enemy’s armed vessels the senior lieutenant of the party that
captured her.” This “win her wear her” plan was a better way to
multiply Nelsons, than by filling up the vacancies with the oftener
high-born than deserving gentlemen sent out by the Admiralty.”
So ended the first small and successful expedition. Let us now
look at the second. This was of a much more serious character.
The rumored arrival at Santa Cruz, on her way to Cadiz, of a
richly-freighted Manilla ship, the Principe d’Asturias, and the
represented vulnerability of the town to a well conducted sea attack,
induced Earl St. Vincent to attempt another enterprise.
Accordingly, on the 15th of July, 1797, his lordship detached upon
that service a squadron of three sail-of-the-line, the Theseus,
Culloden, and Zealous, 74s; the Seahorse, Emerald, and Terpsichore,
frigates, the Fox, 10-gun cutter, and a mortar boat. The whole were
under the orders of Rear-admiral Nelson, in the Theseus.
In about five days the squadron arrived off the island. Every
arrangement that sound judgment could devise having been
completed, two hundred seamen and marines from each of the line-
of-battle ships, and one hundred from each of the three frigates,
exclusive of commissioned officers and servants, and a small
detachment of Royal artillery, the whole together amounting to
about 1050 men, were placed under the command of Captain
Trowbridge, of the Culloden. Each captain, under his direction,
commanded the detachment of seamen from his own ship; and
Captain Thomas Oldfield, of the marines, as senior marine officer,
the entire detachment from that corps.
On the night of the 20th of July the three frigates, accompanied
by the cutter and mortar-boat, and most of the boats of the
squadron, stood in close to the land, to debark the shore party.
A strong gale in the offing, and a strong current against them,
near the shore, prevented them from reaching the intended point of
debarkation. At about half-past three on the morning of the 22d the
squadron bore up for Santa Cruz, and soon after daylight was joined
by the frigates and small craft. The unavoidable appearance of the
latter off the coast gave the islanders the very warning it was so
desirable, for the success of the expedition, they should not have. A
consultation of the principal officers of the squadron now took place,
and decided that an attack should be made on the heights
immediately over the fort at the northeast part of the bay; and then,
from that commanding position, to storm and carry the fort itself. At
nine o’clock on the night of the 22d the frigates anchored inshore,
off the east end of the town, and landed their men; but the latter
finding the heights too strongly guarded to be attempted, re-
embarked in the course of the night, without loss. The three line-of-
battle ships had meanwhile kept under way, to batter the fort, by
way of diversion; but, owing to calms and contrary currents, were
unable to approach nearer than three miles.
Nelson, not being one to abandon an enterprise until after a stout
struggle to accomplish it, resolved to give his seasoned men a
chance at the Santa Cruz garrison as soon as possible. On the 24th
the 50-gun ship Leander joined the squadron, having been sent to
reinforce it, by Lord St. Vincent. Her captain had considerable
experience as a cruiser in those parts, and his local knowledge was
therefore valuable; while the additional force was very acceptable
and added to the hopes of the attacking party.
 On the afternoon of the 24th, at five o’clock, everything being in
readiness, and secrecy no longer possible, the whole squadron
anchored to the northeastward of the town: the line-of-battle ships
about six miles off, and the frigates much nearer. At eleven o’clock at
night, about 700 seamen and marines embarked in the boats of the
squadron, 180 more in the Fox cutter, and about 75 on board a large
boat that had just been captured; numbering altogether, with the
small detachment of Royal artillery, about eleven hundred men. The
different detachments of seamen, under the immediate command of
their respective captains, the marines under Captain Oldfield, the
artillery under Lieut. Baynes, and the whole force under command of
the Rear-Admiral, in person, then pushed off for the shore.
Every precaution had been taken to keep the boats together, in
order that the attack might be simultaneous; but the rough state of
the weather, and the extreme darkness of the night, rendered it
almost impossible for them to keep each other within sight or
hearing. At about half-past one in the morning, the Fox cutter, with
the Admiral’s boat, those of Captains Fremantle and Bowen, and one
or two others, reached, undiscovered, within half gunshot of the
head of the Mole, when, suddenly, the alarm bells on shore began to
ring, and a fire was opened by many pieces of artillery and by
infantry stationed along the shore.
Two shots raked the Fox, and another struck her between wind
and water; so that she sank instantly. Of those on board no less
than ninety-seven were lost; and among them her commander,
Lieut. Gibson.
Another shot struck Rear-Admiral Nelson on the elbow, just as he
was drawing his sword and stepping out of his boat. The wound
completely disabled him, and he was carried back to his ship at
once. Another shot sank the boat in which Captain Bowen was about
approaching the Mole, and seven or eight seamen of her crew
perished.
In spite of this very spirited and determined opposition, the British
effected a landing, and carried the Mole, although it was defended
by about three hundred men and six 24-pounders. Having spiked
these guns, the English were about to advance, when a heavy fire of
musketry and grape-shot from the citadel and from the houses at
the Mole head began to mow them down by scores. Captain Bowen,
of the Terpsichore, and his first lieutenant were almost immediately
killed, and the whole party which landed then were either killed or
wounded.
Meanwhile, Captain Trowbridge, of the Culloden, being unable to
hit the Mole, the spot appointed for landing, pushed on shore under
a battery close to the battery to the southward of the citadel.
Captain Waller, of the Emerald, and a few boats with him, landed
at the same time, but the surf was so high that many of the boats
put back; and all that did not were filled with water, which spoiled
the ammunition in the men’s pouches.
Captain Trowbridge advanced as soon as he had collected a few
men, accompanied by Captain Waller. They reached the great square
of the town, the appointed rendezvous, in hopes of there meeting
the Admiral and the rest of the landing party; but we have seen
already how these were disposed of.
Captain Trowbridge now sent a sergeant, accompanied by two
citizens of the place, to summon the citadel to surrender. No answer
was returned, and the sergeant is supposed to have been killed on
the way. As the scaling ladders which had been brought were lost in
the surf, there was no way of storming the citadel, and after waiting
there an hour, Trowbridge went to join Captains Hood and Miller,
who, with a small body of men, had landed to the southwest. At
daybreak it was found that Trowbridge was in command of about
three hundred and forty survivors, consisting of marines, pikemen,
and seamen with small arms. Having procured a small quantity of
ammunition from some Spanish prisoners whom they had taken,
Trowbridge resolved to try what could be done with the citadel
without ladders, and then found that the streets were commanded
by field-pieces, while an overwhelming force was approaching them
by every avenue. The boats being all stove, there was no possibility
of getting any reinforcements; they were short of ammunition, and
their provisions had been lost in the boats.
Trowbridge now sent Captain Hood, with a flag of truce, to the
Governor, expressing a determination to burn the town if the
Spanish forces advanced, and proposing terms of capitulation, to the
following effect: that the British should be allowed to re-embark,
with their arms, taking their own boats, if saved, and if not, to be
provided with others. And Captain Trowbridge engaged, in case of
compliance, that the ships then before the town should not molest
it, nor attack any one of the Canary Islands.
The Governor, Don Juan Antonio Guttierez, received Captain Hood
and his message, being considerably astonished at receiving such a
proposal from men whom he considered already in his power.
Nevertheless, he accorded the terms, and Trowbridge marched to
the Mole head, where he and his officers and men embarked, in
boats furnished by the Spaniards.
The Governor supplied each of the retreating invaders with a
ration of bread and wine, and directed that the British wounded
should be received into the hospital. He, moreover, sent word to
Admiral Nelson that he was at liberty to send on shore for, and
purchase, fresh provisions.
This was a most disastrous defeat for Nelson, independent of the
melancholy loss of life, which was almost as great as in the battle off
Cape St. Vincent.
 BATTLE OF CAMPERDOWN. 11TH OCTOBER, A. D.
1797.

ord Viscount Duncan, who won the decisive naval

battle of Camperdown, under rather extraordinary
conditions, was born, as simple Adam Duncan, in
Dundee, Scotland, in 1731; so that he was a
veteran, as well in years as in service, when he
gained the victory for which he will always be
remembered.
As a Lieutenant he had served in the expedition
to America, in “the French war;” being in the fleet
which brought Braddock over to meet well earned
defeat, as well as death. He was afterwards
distinguished in the attack upon Belleisle, and in
the capture of Havana. In the war of 1778 he was
actively employed under Rodney. At the first battle of St. Vincent he
was in command of a ship; the first to engage and capture a 70-gun
ship.
After participating in many other actions of importance he was
made a Rear-Admiral in 1759, a Vice-Admiral in 1793, the rank he
held at Camperdown, and finally became full Admiral in 1799.
He was a man of great and unaffected piety, and excited the
wonder and admiration of the Dutch Admiral, when a prisoner on
board his flag-ship, after Camperdown, by summoning his ship’s
company, and then going down on his knees and thanking God for
the mercy vouchsafed them.
Admiral Duncan had, in 1797, the command of the North Sea
English fleet. But that fleet had been so thinned by the secession of
the disaffected ships which took part in the great mutiny of the
English fleet, in that year—called the “Mutiny of the Nore,” and the
“Mutiny at Spithead”—that, towards the end of May, he found
himself at sea with only his own ship (the Venerable, 74) and the
Adamant, 50.
It is necessary here to touch upon the causes which gave rise to a
mutiny which has forever remained a disgrace to the Lords of the
British Admiralty, and to the officers of the fleet serving under them
at that time and for a long time before.
Avoiding any speculations or reflections, we will simply quote from
a well known writer on naval affairs, Admiral Ekins, of the British
Navy, who, quoting another writer in respect to the state of the
British Navy about that time says, “in 1796 and the following years,
after the naval force became so much expanded, the seamen were
exceedingly deteriorated by the introduction of a large mass of Irish
rebels, and the sweepings of all the gaols in England, on the home
station; and by as large an introduction of foreigners on the stations
abroad.”
This writer seems to intend to say, as he goes on, that the Irish,
many of whom had filled offices of some kind at home, had, by
plausible ways, acted with great influence on the minds of the British
seamen whom they found on board their ships, and who were
certainly suffering, at the time, from very oppressive regulations and
fraudulent practices.
 ENGLISH FLEET, OFF TENERIFFE.

“These men entirely overturned the whole discipline and

constitutional temperament of the navy. An honest zeal was changed
for gloomy discontent: grievances were magnified into oppressions,
and the man who had cheerfully executed his subordinate duties, in
what he as yet considered his proper sphere, now aimed at an
equality with, or superiority to, his own respected superiors. Thus
arose the mutiny.”
“After the mutiny, numbers of the Irish were sent to foreign
stations, as a punishment, and disseminated the same spirit.”
The poor Irish! They have for many generations fought the battles
of the English, and of several other nations, but their case seems
more unsettled than before. Without a permanent contingent from
Irish recruits England would be badly off.
To continue with our quotation, “Patrick Little, who was Secretary
to Parker, the leader of the mutiny of the Nore, had been an
attorney in Dublin. He was sent to the West Indies, and, in a few
months, was accused of fostering mutiny there. He was not
convicted of the full offence charged, but was sent to receive six
hundred lashes, did receive two hundred and fifty, and is said to
have died, soon after, of the ‘prevailing fever.’
“The ships in the Mediterranean in 1797-’98-’99, were so short of
men that foreigners of all descriptions were received; and I have
often heard it stated that the fleet could not have gone to sea at
times, if a certain commissioner at Lisbon, about that period, had
not assumed the post of head of Police in that Metropolis, and made,
occasionally, clean sweep of all individuals on the quays and
adjacent streets, who were sent indiscriminately on board the British
fleet; from whence none returned who were serviceable.”
This British Admiral proceeds to quote, “if the battle of the Nile
had not been fought under the directing skill of such a chief, and
under all the effects of surprise, I have heard Sir Thomas Louis
declare that the result might have been very different. As it was, the
defence was much more obstinate than is generally imagined, and
much more protracted.” (In America we have been used to read
English accounts of the battles of those days, because they were
written in our own language.) He goes on to say, “I have understood
it was certainly not the superiority of the crews which prevailed. The
Vanguard was wretchedly manned; and but for the assistance of the
Minotaur, which I saw acknowledged in Lord Nelson’s handwriting,
her fate would have been precarious.”
These remarks are from high English cotemporary authority, we
must remember.
In a note, Admiral Ekins says, “at the conclusion of the war in
1802, the Victorious, of 74 guns, returned to Europe after serving a
considerable time in the East Indies; but, being in a bad state, from
length of service, reached no further than Lisbon. She was there
broken up. Part of her crew were put on board the Amazon, to be
taken to England to receive their wages and return to their homes.
But, unfortunately for them, poor fellows, before they arrived at
Spithead, war had again declared itself, and they learned, with tears
streaming from their eyes, that they were to be detained to serve
another war. They remained nine or ten years in the Amazon, and
were then distributed (the Amazon being worn out) to other ships. A
few of them were afterwards killed serving in the boats of the
Bacchante, in the Mediterranean. Perhaps the whole, certainly the
greatest part, of these men were originally impressed against their
will.”
These are only some authentic instances of the state of the
personnel of the British Navy at this time; and the wonder is that the
officers did so well with such material. Men were often nine or ten
years without setting foot on shore.
And now to return to Admiral Duncan and his operations. Having,
as we have said, been left with only the Venerable, his flag-ship (a
name which reminds us of H. M. S. Pinafore), and the Adamant, he
nevertheless proceeded to his station, off the Texel, to watch the
Dutch, with whom they were then at war.
In the Texel lay at anchor the Dutch fleet of fifteen sail of the line
(including 56s), under the command of Vice Admiral De Winter.
In order to detain the latter in port until a reinforcement should
arrive, Admiral Duncan caused repeated signals to be made, as if to
the main body of his fleet in the offing. This stratagem, it was
supposed, had the desired effect. At length, about the middle of
June, several line-of-battle-ships, in detached portions, joined the
British Admiral, and the two fleets were again placed on an equal
footing.
The Venerable, having been nearly five months at sea, and during
a part of the time exposed to very boisterous weather, was in want
of almost every description of stores. Others of the ships had also
suffered by the recent gales of wind, and were short of provisions.
Thus circumstanced, the Admiral, on the 3d of October, put into
Yarmouth roads, to refit and re-victual, leaving off the Dutch coast a
small squadron of observation, under the orders of Captain Trollope,
of the Russell.
Early on the morning of October 9th an armed lugger, hired as a
despatch vessel, came into the back of Yarmouth sands, with the
signal flying for an enemy.
After great bustle and hurried preparations, Admiral Duncan put to
sea, a little before noon, with eleven sail-of-the-line. With a fair wind
he steered straight for his old station. On the following day three
more ships joined him; so that he had seven 74s, and seven 64s,
and two 50-gun frigates. There were also the Beaulieu, 40; the
Circe, 28; and the Martin, sloop.
On the afternoon of the 11th the advanced ships were near
enough to count twenty-two sail of square-rigged vessels, chiefly
merchantmen, at anchor in the Texel.
Admiral Duncan, having received from Captain Trollope
information of what course the enemy’s fleet was steering, now
stood along shore to the southward.
At about seven on the following morning, the Russell, Adamant
and Beaulieu were made out in the southwest, bearing at their mast-
heads the signals for an enemy in sight, to leeward; and at about
half-past eight a strange fleet, consisting of twenty-one ships and
four brigs, made its appearance in that quarter.
The Dutch fleet consisted of four 74s, seven 64s, four 50s and two
44-gun ships, with two 32-gun frigates, two corvettes, four brig-
sloops, and two advice-boats. Some accounts give more ships than
this. Probably there were more.
These vessels, under the command of Vice-Admiral De Winter, had
quitted the Texel at ten o’clock on the morning of the 10th of
October, with a light breeze at about east by north. On the night of
that day, the wind being then southwest, Captain Trollope’s
squadron was discovered by them, to windward, and immediately
chased; but the Dutch ships, being dull sailers, did not come near
him. The Dutch fleet then stretched out toward the Meuse flat,
where Admiral De Winter expected to be joined by a 64-gun ship.
Not meeting her, he stood on to the westward, followed, or rather,
as the wind was, preceded, by Captain Trollope’s squadron.
The wind continued westerly during the three succeeding days,
and prevented the Dutch fleet from getting abreast of Lowestoffe
until the evening of the 10th. The extreme darkness of that night
induced Admiral De Winter to detach a few of his best sailing ships,
in hopes that they would be enabled, by daybreak, to get to
windward of, and capture or chase away, Captain Trollope’s
squadron, which had followed them with great pertinacity. Just as
the ships had made sail for that purpose some friendly merchant
ships came into the fleet, and informed Admiral De Winter that the
English fleet was within thirty miles of him, in the north-northwest,
and steering east by south. The detached ships were instantly
recalled; and the Dutch fleet, as soon as formed in compact order,
edged away, with the wind northwest, towards Camperdown, the
appointed place of rendezvous.
At daylight on the 11th the Dutch fleet was about thirty miles off
the village of Scheveningen, in loose order, and speaking a friendly
convoy, from which additional information was obtained.
At this time the English squadron of observation was seen to
windward, with numerous signals flying, which convinced Admiral De
Winter that the English fleet was in sight. He accordingly ordered his
ships to their stations, and to facilitate the junction of the ships most
to leeward, stood towards the land. The Wykerdens bearing east,
about twenty miles off, the Dutch fleet hauled to the wind, on the
starboard tack, and shortly afterwards discovered Admiral Duncan’s
fleet in the north-northwest. The Dutch fleet then tacked, and, as
soon as a close line was formed in the direction of northeast and
southwest, the Dutch ships, throwing their main top-sails aback,
resolutely awaited the approach of the British fleet.
Owing chiefly to inequality in point of sailing among the British
ships, their fleet, when that of the Dutch appeared in sight, was in
very loose order. To enable the dull sailers to take their proper
stations, Admiral Duncan, at about eleven A.M., brought to, on the
starboard tack; but soon afterwards observing that the Dutch ships
were drawing fast inshore, he made signal for each ship to engage
her opponent in the enemy’s line; then to bear up; and, lastly, for
the van to attack the enemy’s rear. At about half-past eleven, the
centre of the Dutch line then bearing southeast, distant four or five
miles, the British fleet bore down, but, owing to some of the ships
not yet being up, in no regular order of battle. Some were stretching
across to get into their stations; others seemed in doubt where to
go; and others, again, were pushing for the thickest of the enemy,
without regard to stations.
 A little before noon Admiral Duncan made signal that he should
pass through the enemy’s line and engage him to leeward. This
signal appears to have been kept flying but a short time, and the
weather was so thick that the ships generally did not make it out. It
was replaced by one for close action, which was kept flying for an
hour and a half; till, indeed, it was shot away. About half-past twelve
Vice-Admiral Onslow, whose ship, the Monarch, was leading the
advanced or port division of the British fleet, cut through the Dutch
line, between the Haerlem, 64, and the Jupiter, 74; pouring into
each, in passing, a well-directed broadside. Then the Monarch,
leaving the Haerlem to the Powerful, which followed her, luffed up
close alongside the Jupiter, and these two ships became warmly
engaged. The Jupiter carried the flag of Vice-Admiral Reyntjes. The
rounding to of the Monarch afforded the Dutch Monnikendam frigate
and Atalanta brig, which were in shore and in the rear, an
opportunity to rake the English ship several times; and the very
plucky little brig, in particular, did not retire until she had been much
damaged by the Monarch’s shot. It was supposed she had been
sunk by the 74, but she arrived safely, after the battle, in a Dutch
port. The remaining ships of the English port division, especially the
Monmouth, 64, and the Russell, 74, were soon in action with the
Dutch rear-ships; among the last of which to surrender was the
Jupiter, 74, the first to be engaged.
About twenty minutes after the Monarch, with Vice-Admiral
Onslow’s flag, had broken the Dutch line, Duncan’s flag-ship, the
Venerable, frustrated in an attempt to pass astern of the Vryheid,
74, De Winter’s flag-ship, by the great promptness of the States-
General, 74, in closing the interval, ran under the stern of the latter,
and soon compelled her to bear up; and the Triumph, the
Venerable’s second astern, found herself closely engaged with the
Wassenaer, the second astern to the States-General. Meanwhile the
Venerable had ranged up close on the lee side of her first intended
antagonist, the Vryheid, with whom, on the other side, the Ardent
was also warmly engaged, and in front, the Belford, as she cut
through the line astern of the Dutch Gelykheid, 64. The Dutch ships
Brutus, 74, Rear-Admiral Bloys, and the Leyden, 64, and Mars, 44,
not being pressed upon by opponents, advanced to the succor of
their closely beset Admiral, and did considerable damage to the
Venerable, as well as the Ardent, and others of the British van ships.
Just at this critical period the Hercules, 64, which ship had caught
fire on the poop, bore up and fell out of line, drifting down very near
the Venerable.
Although, to the surprise of every one, the Dutch crew managed
to extinguish the flames, yet, having thrown overboard their powder,
they were obliged to surrender the ship, which had already had her
mizzen-mast shot away, to the first opponent which challenged her.
The serious damages which the Venerable had sustained obliged her
to haul off and wear round on the starboard tack. Seeing this, the
Triumph, which had compelled the Wassenaer to strike, approached
to help finish the Vryheid; but that very gallant ship still made a
good defence. At length, after being pounded at by the Venerable,
Triumph, Ardent and Director, her three masts fell over the side, and
disabled her starboard guns, when the overmatched but heroic
Vryheid dropped out of the line of battle, an ungovernable hulk, and
struck her colors.
A curious incident occurred in regard to the Wassenaer, 64, which,
we have just seen, was compelled by the English Triumph, 74, to
strike her colors, and fall out of the line. One of the Dutch brigs
followed her, and fired at her, persistently, until she re-hoisted her
colors. The Russel, 74, soon came up, however, and compelled the
unfortunate Wassenaer again to strike to her. With the surrender of
Admiral De Winter’s ship the action ceased, and the English found
themselves in possession of the Vryheid and Jupiter, 74s, Devries,
Gelykheid, Haerlem, Hercules and Wassenaer, 64s, Alkmaar and
Delft, 50s, and the frigates Monnikendam and Ambuscade. The first
of these frigates had been engaged by the Monmouth, 64, and was
finally taken possession of by the Beaulieu, a 40-gun frigate of the
English.
The Dutch van ship, the Beschermer, 50, dreading, very naturally,
so strong an opponent as the Lancaster, 64, had early wore, and
fallen out of the line. Her example was followed, with much less
reason, by several of the other Dutch ships, which, although seen
making off, could not be pursued, on account of the nearness of the
land, and the shallowness of the water. The Venerable at this time
sounded, and found only nine fathoms, and the shore, under their
lee, which was that between Camperdown and Egmont, and about
thirty miles northwest of Amsterdam, was only about five miles off.
The British ships now hastened to secure their prizes, so that they
might, before nightfall, work clear of this dangerous coast.
The appearance of the victorious British fleet was very different
from that which generally presented itself after a battle with the
French or the Spaniards. Not a single lower mast, not even a
topmast, in the British fleet, was shot away. Nor were the sails and
rigging of the latter very much damaged.
It was at the hulls of their adversaries that the sturdy Dutchmen
had directed their shot, and they did not fire until they were so near
that no shot could well miss. All the English ships had shot sticking
in their sides; many were pierced by them in all directions, and some
of them had such dangerous wounds between wind and water that
their pumps had to be kept going briskly. The Ardent had received
about one hundred round shot in her hull; the Belliqueux, Belford,
Venerable and Monarch had nearly as many. But the latter ship was
so untouched aloft, that when her top-sail sheets, which had been
shot away, were spliced and hauled home, no one looking at her
from a little distance would have believed she had been in action.
With such fire, directed almost exclusively at the hulls, even by the
feeble guns of that day, the loss of men could not be otherwise than
severe. The British loss was 203 killed, and 622 wounded.
The captured ships were all either dismasted outright, or so
injured in their masts that most of the latter fell as soon as the wind
and sea, during the passage to England, began to act powerfully
upon them. The Dutch ships’ hulls were also terribly cut up, and
were so damaged as mostly to be brought into port to be exhibited
as trophies, and then broken up.
Their loss was proportionately severe. The Dutch Vice-Admiral and
the two Rear-Admirals were all wounded. Vice-Admiral Reyntjes died
in London soon after, not of his wound, but of a chronic disease.
Captain Holland, of the Wassenaer, was killed early in the action,
which may account, partially, for her not holding out longer. Admiral
De Winter’s captain, Van Rossem, had his thigh carried away by a
round shot, and died almost immediately.
Many other Dutch officers were killed and wounded, and their
loss, including that on board the Monnikendam frigate, which was
not in the line, was 540 killed, and 620 wounded.
The actual force of the two fleets in this battle was, according to
English accounts—not always very reliable at that time—
British. Dutch.
Ships 16 16
Guns 1,150 1,034
Agg. weight of metal, lbs. 11,501 9,857
Crews 8,221 7,175
Size, tons 23,601 20,937

It is fair to say that the Dutch had several frigates and brigs
abreast the intervals in their line, which did good service, raking the
English ships as they came through and luffed up to leeward of the
enemy’s line.
As it was, Admiral Duncan met and fought the Dutch fleet before a
98- and two 74-gun ships which De Winter had expected could join
him.
Admiral De Winter, in his official report of the action, attributed his
failure to four causes: first, the superiority of the British in large
ships; secondly, their having been together at sea for many weeks,
and hence well accustomed to work together; thirdly, the advantage
of the attack, and fourthly, the early retreat of some of his ships,
and the bad sailing of some of the others. He also expressed his
belief that, if his signals had been obeyed as promptly as Admiral
Duncan’s were, some of the English fleet would have been brought
into the Texel, instead of the Dutch ships going to England. His
statement about the English ships being so long together was not
altogether correct. Captain Williamson, of the English ship Agincourt,
64, was court-martialed for his conduct in this action. He was
accused of disobedience of signals and failure to go into action; and
also, on a second charge, of cowardice or disaffection. The first
charge was found proved, but not the second, and Williamson
received a very severe sentence. It was proved on this trial that
some of Admiral Duncan’s fleet did not know other ships in the same
fleet. In the great fleet actions of those days, between ships-of-the-
line, it was not customary for frigates and smaller ships to fire, or to
be fired at, unless they provoked it; and the Dutch frigates,
corvettes and brigs formed in this action a second line, and fought
well. The Dutch were, indeed, an enemy not to be despised, and
Admiral Duncan did full justice to the determined way in which most
of them fought.
Scarcely was the British fleet, with its prizes, pointed to the
westward, when a gale of wind came on, which scattered and
endangered the whole of them. The injured masts fell, and the
vessels leaked through shot-holes which in any ordinary weather
would have been above the reach of the water.
On the 13th, the Delft, 50, a prize, exhibited a board with the
words chalked on it “The ship is sinking.” Assistance was sent, and
most of the men removed; but several of the prize crew and many of
the prisoners perished in her, so quickly did she founder.
The Monnikendam frigate was wrecked on a shoal; but all on
board were saved; and the Ambuscade frigate, being driven on the
Dutch coast, was recaptured. One by one the rest of the scattered
fleet and the prizes reached English ports.
Admiral Duncan was made a peer, and Vice-Admiral Onslow a
baronet, for this action. Gold medals were presented to the flag-
officers and captains; and the thanks of Parliament were voted to
the fleet.
We often realize more of the real spirit of a fight from private
accounts and comments than from the official reports; and we,
therefore, add a few remarks and anecdotes from such sources. In
the first place, the promptitude and decision of Admiral Duncan on
meeting the Dutch fleet is especially to be noticed. “The British
Admiral soon perceived that if he waited to form his line (the enemy
drawing fast in with the land) there would be no action.” He,
therefore, hoisted the signal to make all sail, break the line, and
engage the enemy to leeward; and for close action, which last signal
flew until it was shot away. This signal could not be mistaken, and,
coupled with the gallant Admiral’s example, superseded all former
ones.
If further proof of the superior efficacy of such a mode of attack
be wanting, it is to be found not only in the declaration of the brave
Dutch Admiral, but also in the testimony of Lord Nelson, who,
although not acquainted with Lord Duncan, wrote to him, after the
battle of the Nile, to tell him how “he had profited by his example.”
The Dutch Admiral De Winter said, “Your not waiting to form a line
ruined me; if I had got nearer to the shore, and you had attacked, I
should probably have drawn both fleets on, and it would have been
a victory to me, being on my own coast.”
It is a fact that many of the vessels of Admiral Duncan’s fleet were
intended for Indiamen, and not so stoutly built as men-of-war
usually are; and many of his ships were in bad condition, and had
not had time to complete their stores when called away from
Yarmouth Roads to encounter the enemy.
Among other incidents of this action, it is recorded that, when the
main-top-gallant mast of the Venerable was shot away, a seaman
named Crawford went aloft with another flag, and hammer and
nails, and nailed the flag to the topmast-head.
Had Duncan’s fleet been of as good material as that of Lord St.
Vincent, it is probable that every Dutch ship would have been taken.
When the action ceased the English fleet were in only nine fathoms
of water, and a severe gale was nearly upon them; and the wonder
is that they saved themselves and so many of their prizes, in their
battered condition.
Captain Inglis, of the Belliqueux, of 64 guns, owing either to a
long absence from active service, or an inaptitude to the subject,
sometimes apparent in sea officers, had neglected to make himself a
competent master of the signal-book, and on the morning of the day
of the battle, when it became necessary to act with promptitude in
obedience to signals, found himself more puzzled than enlightened
by it, and, throwing it with contempt upon the deck, exclaimed, in
broad Scotch: “D—n me, up wi’ the hellum, and gang intil the middle
o’t!”
 In this manner he bravely anticipated the remedy in such cases
provided by Nelson, who, in his celebrated “Memorandum,” observes
that, “when a captain should be at a loss he cannot do very wrong if
he lay his ship alongside of the enemy.”
In strict conformity with this doctrine the Belliqueux got herself
very roughly treated by the van of the Dutch fleet.

BATTLE OF THE NILE.—FRENCH FLAG-SHIP L’ORIENT, 120 GUNS, ON FIRE.

 BATTLE OF THE NILE. 1ST. AUGUST, 1798.

his battle is called by the French Aboukir, the name

of the bay in which it took place, and it is really a
more proper name for the action, as only a small
mouth of the Nile opened into the bay.
Beside the great naval action, Aboukir has given
its name to a bloody and decisive land battle,
which took place July 25th, 1799, between the
French and a Turkish army. We may dispose of the
latter briefly before taking up the more important
sea fight, although in point of time the latter
precedes it a year.
Bonaparte having learned of the landing of a
Turkish army of 18,000 infantry at Aboukir, advanced to attack them,
at the head of only about 6000 men. The Turks, who were mostly
Janissaries, had a very considerable force of artillery, and were in
part commanded by English officers. Being strongly intrenched at
the village of Aboukir, they should have beaten off the French force
easily; but, at the word of command from Bonaparte, Generals
D’Estaing, Murat and Lannes attacked the entrenchments with
desperate courage, and, after a terrible fight, which lasted some
hours, the Turks were fairly driven into the sea. Thousands of bodies
floated upon the bay, which the year before had borne the corpses
of so many French sailors, who had perished from gun-shot or by
fire. Perhaps for the first time in the history of modern warfare, an
army was entirely destroyed.
It was on this occasion that Kleber, at the close of the fighting,
seized Bonaparte in his arms, and embracing him, exclaimed:
“General, you are the greatest man in the world!”
A year previous to the event just recorded, while Bonaparte was
occupied in organizing his new conquest of Egypt, fortune was
preparing for him one of the most terrible reverses which the French
arms had ever met, by sea or by land.
What must have made it harder for him to bear was, that when
leaving Alexandria to go to Cairo he had very strongly recommended
Admiral Brueys, who commanded the fleet which had brought him to
Egypt, not to remain at the anchorage of Aboukir, where the English
could, he thought, take him at a disadvantage. In fact, Napoleon’s
military mind foresaw just what afterwards happened.
Brueys at first thought of taking his fleet to Corfu, but lost
precious time in waiting for news from Cairo, and this delay brought
on the disaster which had a very important influence in moulding the
destiny, not only of Egypt, but of the whole of Europe.
Learning of the departure of a large body of troops, and of a
strong fleet, from Toulon, but in entire ignorance of the object of
their expedition, Nelson, after vainly seeking for them in the
Archipelago, in the Adriatic, at Naples, and on the coasts of Sicily, at
last learned with certainty that they had effected a landing in Egypt.
He made all sail at once for Alexandria, determined to fight the
French fleet the moment it was found, and wherever it might be. He
found it at Aboukir bay, just to the eastward of Alexandria, on the
1st of August, 1798; and we shall now give a general sketch of what
ensued, and after that the particulars of this important action—from
both French and English sources.
Although it was nearly six o’clock in the evening when the French
fleet was discovered, Nelson resolved to attack immediately.
Admiral Brueys’ fleet was moored in the bay, which forms a pretty
regular semicircle, and had arranged his thirteen ships-of-the-line in
a curved line, parallel with the shore; having upon his left, or
western flank, a little island, called also Aboukir.
Thinking it impossible that a ship-of-the-line could pass between
this island and the last ship of his line, to take him in the rear, he
contented himself with establishing upon the island a battery of
twelve or fourteen guns; thinking, indeed, that part of his position so
little liable to attack that he placed his worst vessels there.
But with an adversary like Nelson, most formidable, not only for
the brilliancy of his conceptions, but for the skillful audacity with
which he carried them out, the precautions which under ordinary
circumstances would have been sufficient proved of no avail.
The British fleet comprised the same number of line-of-battle-
ships as the French, but the latter had more smaller ships.
The British Admiral advanced intrepidly to the attack; a portion of
his ships taking a course between the French line and the coast. The
Culloden, the leading English ship, ran upon a shoal, and stuck fast;
but, although her batteries were thus thrown out of the engagement
which followed, her mishap piloted the others in. The Goliath, the
Audacious, the Theseus and the Orion succeeded in passing inside
the French line; penetrating as far as the Tonnant, which was the
eighth of the French line, and thus engaged the French centre and
left.
The rest of the English fleet advanced outside the French line, and
so put the left and centre of it between two fires.
The battle was a terrible one, especially at the French centre,
where the French Admiral’s ship, L’Orient, was stationed. The
Bellerophon, one of Nelson’s best ships, was dismasted, terribly cut
up, and obliged to haul off; and other English ships so damaged that
they were obliged to withdraw.
In spite of the success of Nelson’s grand manœuvre, Brueys still
had some chance of success, if the orders which he gave to his
right, or eastern, wing had been carried out. But Admiral Villeneuve,
who commanded there, did not make out Brueys’ signals, and
remained in his position, at anchor, instead of getting under way,
and doubling upon the English outside line, which would have thus
put the latter, in their turn, between two fires.
Nelson’s ready mind had foreseen this danger; but Villeneuve, who
was to lose another even more important battle at Trafalgar, lacked
the instinctive resolution which causes a second in command, under
such circumstances, to hasten to the relief of his chief, without
formal orders.
Like Grouchy at Waterloo, he heard and saw the cannonade which
was destroying the centre and left of the French line, without
coming to the rescue; and while that part of the French fleet was
performing prodigies of valor to uphold the honor of their flag,
Villeneuve escaped, with four ships-of-the-line, thinking himself
praiseworthy in saving them from the fate of the rest.
The unfortunate Brueys, though wounded, would not leave the
deck. “An Admiral ought to die giving his orders,” he is reported, on
good authority, to have said. Not long after this speech another shot
killed him. The brave Captain Dupetit-Thouars had both legs carried
away, but, like the Admiral, would not leave the deck, but remained
there, taking snuff, and coolly directing operations, until another
shot struck and killed him.
In fact, acts of heroism were performed by many of the officers
and men on both sides.
About eleven o’clock at night the Orient, a huge and magnificent
vessel, blew up, with a terrible explosion. By this time all the French
vessels were destroyed or rendered worthless, except the four
carried off by Villeneuve, and Nelson’s fleet was in no condition to
pursue them.
Such, in brief, was the celebrated battle of Aboukir, or the Nile, the
most disastrous the French navy had ever fought, and the military
consequences of which were of such immense importance. It shut
up the French and their army in Egypt, and abandoned them to their
own resources.
France lost, and England gained, ascendancy in the Levant, and
what was worse, it destroyed the morale of the French navy—the
effects being seen for years, and especially at Trafalgar.
And now we will proceed to give a more detailed account of the
action.
Nelson’s fleet arrived off Alexandria on the morning of the 1st of
August, at about 10 o’clock. They found there a forest of masts—
belonging to transports and troopships, but few men-of-war. The
harbor did not permit of the entrance of such large ships as
composed the French line. The two British look-out ships, the
Alexander and Swiftsure, also found the French flag flying on the
forts and walls.
About noon the Zealous, which ship had been looking further to
the eastward (just as the Pharos tower of Alexandria bore south-
southwest, distant about 20 miles), signaled that sixteen ships-of-
the-line lay at anchor, in line of battle, in a bay upon her port bow.
The British fleet instantly hauled up, steering to the eastward,
under top-gallant sails, with a fine breeze from the northward and
westward. These ships were in good discipline, and it did not take
them long to clear for action.
Let us now turn to the French fleet, which they were soon to
encounter. On the 1st of July, Admiral Brueys, with his fleet, brought
to off the old port of Alexandria, and at once learned that a British
squadron had been looking for him there. On hearing this, General
Bonaparte desired to be landed, and the Admiral at once proceeded
to disembark the General and 6000 men, in a creek near Marabout
Castle, about six miles from the city of Alexandria.
Between the 1st and the 6th of July all the troops, with their
baggage, were landed; and six vessels, armed en flute, went into
Alexandria harbor, to protect the transports. As the ships-of-the-line
drew too much water to enter, Admiral Brueys, with three frigates
and 13 sail-of-the-line, stood off and proceeded to Aboukir bay,
about 15 miles to the eastward of Alexandria. Reaching the bay, he
anchored his ships very judiciously, in line ahead, about one hundred
and sixty yards (Engl.) from each other, with the van-ship close to a
shoal in the northwest, and the whole of the line just outside a four-
fathom bank. It was thus considered that an enemy could not turn
either flank.
The French ships, beginning at the van, were ranged in the
following order: Guerrier, Conquérant, Spartiate, Aquilon, Peuple-
Souvérain (all 74’s), the Franklin, 80, Rear-Admiral Blanquet, second
in command; Orient, 120, (formerly called the Sans Culotte, and the
flag-ship of Admiral Brueys, Rear-Admiral Gauteaume and Captain
Casa-Bianca); next the Tonnant, 80; the Heureux, 74; the Mercure,
74; the Guillaume Tell, 80, and the Généreux and Timoléon, both
74’s.
Having thus moored his fleet in a strong position, the French
Admiral awaited the issue of General Bonaparte’s operations on
shore.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookfinal.com