Introduction To Modern Dynamics Chaos Networks Space and Time 2Nd Edition David D Nolte Full Chapter

Introduction to Modern Dynamics:
Chaos, Networks, Space, and Time 2nd

Edition David D. Nolte
Visit to download the full and correct content document:
https://ebookmass.com/product/introduction-to-modern-dynamics-chaos-networks-sp
ace-and-time-2nd-edition-david-d-nolte/
INTRODUCTION TO MODERN DYNAMICS
Introduction to Modern Dynamics
Chaos, Networks, Space and Time
Second Edition
David D. Nolte
Purdue University
1
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© David D. Nolte 2019
The moral rights of the author have been asserted
First Edition published in 2015
Second Edition published in 2019
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2019945041
ISBN 978–0–19–884462–4 (hbk.)
ISBN 978–0–19–884463–1 (pbk.)
DOI: 10.1093/oso/9780198844624.001.0001
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
Preface to the Second Edition
Introduction to Modern Dynamics: Chaos, Networks, Space and Time (2015) is part
of an emerging effort in physics education to update the undergraduate physics
curriculum. Conventional junior-level mechanics courses have overlooked many
modern dynamics topics that physics majors will use in their careers: nonlinearity,
chaos, network theory, econophysics, game theory, neural nets, geodesic geometry,
among others. These are the topics at the forefront of physics that drive
high-tech businesses and start-ups where more than half of physicists are
employed. The first edition of Introduction to Modern Dynamics contributed
to this effort by introducing these topics in a coherent program that emphasized
common geometric properties across a wide range of dynamical systems.
The second edition of Introduction to Modern Dynamics continues that trend
by expanding chapters to including additional material and topics. It rearranges
several of the introductory chapters for improved logical flow and expands them
to add new subject matter. The second edition also has additional homework
problems.
New or expanded topics in the second edition include
• Lagrangian applications
• Lagrange’s undetermined multipliers
• Action-angle variables and conserved quantities
• The virial theorem
• Non-autonomous flows
• A new chapter on Hamiltonian chaos
• Rational resonances
• Synchronization of chaos
• Diffusion and epidemics on networks
• Replicator dynamics
• Game theory
• An extensively expanded chapter on economic dynamics
The goal of the second edition of Introduction to Modern Dynamics is to

strengthen the sections on conventional topics (which students need for the GRE
physics subject test), making it an ideal textbook for broader adoption at the junior
vi Preface to the Second Edition
level, while continuing the program of updating topics and approaches that are
relevant for the roles that physicists will play in the twenty-first century.
The historical development of modern dynamics is described in Galileo
Unbound: A Path Across Life,the Universe and Everything, by D. D. Nolte, published
by Oxford University Press (2018).
Preface: The Best Parts of Physics
The best parts of physics are the last topics that our students ever see. These
are the exciting new frontiers of nonlinear and complex systems that are at
the forefront of university research and are the basis of many of our high-
tech businesses. Topics such as traffic on the World Wide Web, the spread of
epidemics through globally mobile populations, or the synchronization of global
economies are governed by universal principles just as profound as Newton’s
Laws. Nonetheless, the conventional university physics curriculum reserves most
of these topics for advanced graduate study. Two justifications are given for this
situation: first, that the mathematical tools needed to understand these topics are
beyond the skill set of undergraduate students, and second, that these are specialty
topics with no common theme and little overlap.
Introduction to Modern Dynamics: Chaos, Networks, Space and Time dispels these
myths. The structure of this book combines the three main topics of modern
dynamics—chaos theory, dynamics on complex networks and the geometry of
dynamical spaces—into a coherent framework. By taking a geometric view of
physics, concentrating on the time evolution of physical systems as trajectories
through abstract spaces, these topics share a common and simple mathematical
language with which any student can gain a unified physical intuition. Given the
growing importance of complex dynamical systems in many areas of science and
technology, this text provides students with an up-to-date foundation for their
future careers.
While pursuing this aim, Introduction to Modern Dynamics embeds the topics
of modern dynamics—chaos, synchronization, network theory, neural networks,
evolutionary change, econophysics, and relativity—within the context of tradi-
tional approaches to physics founded on the stationarity principles of variational
calculus and Lagrangian and Hamiltonian physics. As the physics student explores
the wide range of modern dynamics in this text, the fundamental tools that are
needed for a physicist’s career in quantitative science are provided, including
topics the student needs to know for the Graduate Record Examination (GRE).
The goal of this textbook is to modernize the teaching of junior-level dynamics,
responsive to a changing employment landscape, while retaining the core tradi-
tions and common language of dynamics texts.
viii Preface: The Best Parts of Physics
A unifying concept: geometry and dynamics

Instructors or students may wonder how an introductory textbook can contain
topics, under the same book cover, on econophysics and evolution as well as the
physics of black holes. However, it is not the physics of black holes that matters,
rather it is the description of general dynamical spaces that is important and the
understanding that can be gained of the geometric aspects of trajectories governed
by the properties of these spaces. All changing systems, whether in biology or
economics or computer science or photons in orbit around a black hole, are
understood as trajectories in abstract dynamical spaces.
Newton takes a back seat in this text. He will always be at the heart of dynamics,
but the modern emphasis has shifted away from F = ma to a newer perspective
where Newton’s Laws are special cases of broader concepts. There are economic
forces and forces of natural selection that are just as real as the force of gravity
on point particles. For that matter, even the force of gravity recedes into the
background as force-free motion in curved space-time takes the fore.
Unlike Newton, Hamilton and Lagrange retain their positions here. The vari-
ational principle and the minimization of dynamical quantities are core concepts
in dynamics. Minimization of the action integral provides trajectories in real
space, and minimization of metric distances provides trajectories—geodesics—
in dynamical spaces. Conservation laws arise naturally from Lagrangians, and
energy conservation enables simplifications using Hamiltonian dynamics. Space
and geometry are almost synonymous in this context. Defining the space of a
dynamical system takes first importance, and the geometry of the dynamical space
then determines the set of all trajectories that can exist in it.
A common tool: dynamical flows and the

ODE solver
A mathematical flow is a set of first-order differential equations that are solved
using as many initial values as there are variables, which defines the dimensionality
of the dynamical space. Mathematical flows are one of the foundation stones that
appears continually throughout this textbook. Nearly all of the subjects explored
here—from evolving viruses to orbital dynamics—can be captured as a flow.
Therefore, a common tool used throughout this text is the numerical solution
of the ordinary differential equation (ODE). Computers can be both a boon and
a bane to the modern physics student. On the one hand, the easy availability of
ODE solvers makes even the most obscure equations easy to simulate numerically,
enabling any student to plot a phase plane portrait that contains all manner of
behavior. On the other hand, physical insight and analytical understanding of
complex behavior tend to suffer from the computer-game nature of simulators.
Therefore, this textbook places a strong emphasis on analysis, and on behavior
Preface: The Best Parts of Physics ix
under limiting conditions, with the goal to reduce a problem to a few simple
principles, while making use of computer simulations to capture both the whole
picture as well as the details of system behavior.
Traditional junior-level physics: how to use

this book
All the traditional topics of junior-level physics are here. From the simplest
description of the harmonic oscillator, through Lagrangian and Hamiltonian
physics, to rigid body motion and orbital dynamics—the core topics of advanced
undergraduate physics are retained and are interspersed throughout this textbook.
What’s simple in complex systems?

The traditional topics of mechanics are integrated into the broader view of modern
dynamics that draws from the theory of complex systems. The range of subject
matter encompassed by complex systems is immense, and a comprehensive
coverage of this topic is outside the scope of this book. However, there is still
a surprisingly wide range of complex behavior that can be captured using the
simple concept that the geometry of a dynamic space dictates the set of all
possible trajectories in that space. Therefore, simple analysis of the associated
flows provides many intuitive insights into the origins of complex behavior. The
special topics covered in this textbook are:
• Chaos theory (Chapter 4)
Much of nonlinear dynamics can be understood through linearization of the

flow equations (equations of motion) around special fixed points. Visualizing the
dynamics of multi-parameter systems within multidimensional spaces is made
simpler by concepts such as the Poincaré section, strange attractors that have fractal
geometry, and iterative maps.
• Synchronization (Chapter 6)
The nonlinear synchronization of two or more oscillators is a starting point for

understanding more complex systems. As the whole can be greater than the sum
of the parts, global properties often emerge from local interactions among the
parts. Synchronization of oscillators is surprisingly common and robust, leading
to frequency-entrainment, phase-locking, and fractional resonance that allow small
perturbations to control large networks of interacting systems.
x Preface: The Best Parts of Physics
• Network theory (Chapter 7)
Everywhere we look today, we see networks. The ones we interact with daily
are social networks and related networks on the World Wide Web. In this
chapter, individual nodes are joined into networks of various geometries, such as
small-world networks and scale-free networks. The diffusion of disease across these
networks is explored, and the synchronization of Poincaré phase oscillators can
induce a Kuramoto transition to complete synchronicity.
• Evolutionary dynamics (Chapter 8)
Some of the earliest explorations of nonlinear dynamics came from studies of pop-
ulation dynamics. In a modern context, populations are governed by evolutionary
pressures and by genetics. Topics such as viral mutation and spread, as well as the
evolution of species within a fitness landscape, are understood as simple balances
within quasispecies equations.
• Neural networks (Chapter 9)
Perhaps the most complex of all networks is the brain. This chapter starts with the
single neuron, which is a limit-cycle oscillator that can show interesting bistability
and bifurcations. When neurons are placed into simple neural networks, such as
perceptrons or feedforward networks, they can do simple tasks after training by error
back-propagation. The complexity of the tasks increases with the complexity of
the networks, and recurrent networks, like the Hopfield neural net, can perform
associated memory operations that challenge even the human mind.
• Econophysics (Chapter 10)
A most baffling complex system that influences our daily activities, as well as
the trajectory of our careers, is the economy in the large and the small. The
dynamics of microeconomics determines what and why we buy, while the dynamics
of macroeconomics drives entire nations up and down economic swings. These
forces can be (partially) understood in terms of nonlinear dynamics and flows
in economic spaces. Business cycles and the diffusion of prices on the stock market
are no less understandable than evolutionary dynamics (Chapter 8) or network
dynamics (Chapter 7), and indeed draw closely from those topics.
• Geodesic motion (Chapter 11)
This chapter is the bridge between the preceding chapters on complex systems
and the succeeding chapters on relativity theory (both special and general). This
is where the geometry of space is first fully defined in terms of a metric tensor, and
where trajectories through a dynamical space are discovered to be paths of force-
Preface: The Best Parts of Physics xi
free motion. The geodesic equation (a geodesic flow) supersedes Newton’s Second
Law as the fundamental equation of motion that can be used to define the path of
masses through potential landscapes and the path of light through space-time.
• Special relativity (Chapter 12)
In addition to traditional topics of Lorentz transformations and mass-energy equiv-

alence, this chapter presents the broader view of trajectories through Minkowski
space-time whose geometric properties are defined by the Minkowski metric.
Relativistic forces and noninertial (accelerating) frames connect to the next
chapter that generalizes all relativistic behavior.
• General relativity (Chapter 13)
The physics of gravitation, more than any other topic, benefits from the over-
arching theme developed throughout this book—that the geometry of a space
defines the properties of all trajectories within that space. Indeed, in this geometric
view of physics, Newton’s force of gravity disappears and is replaced by force-
free geodesics through warped space-time. Mercury’s orbit around the Sun,
and trajectories of light past black holes, are elements of geodesic flows whose
properties are easily understood using the tools developed in Chapter 4 and
expanded upon throughout this textbook.
Acknowledgments
I gratefully acknowledge the many helpful discussions with my colleagues

Ephraim Fischbach, Andrew Hirsch, Sherwin Love, and Hisao Nakanishi during
the preparation of this book. Special thanks to my family, Laura and Nicholas, for
putting up with my “hobby” for so many years, and also for their encouragement
and moral support. I also thank the editors at Oxford University Press for help in
preparing the manuscript and especially Sonke Adlung for helping me realize my
vision.
Part I
Geometric Mechanics
Traditional approaches to the mechanics of particles tend to focus on individual

trajectories. In contrast, modern dynamics takes a global view of dynamical
behavior by studying the set of all possible trajectories of a system. Modern
dynamics furthermore studies properties in dynamical spaces that carry names
like state space, phase space,and space–time. Dynamical spaces can be highly abstract
and can have high dimensionality. This initial part of the book introduces the
mathematical tools necessary to study the geometry of dynamical spaces and the
resulting dynamical behavior within those spaces. Central to modern dynamics
is Hamilton’s Principle of Stationary Action as the prototypical minimization
principle that underlies much of dynamics. This approach will lead ultimately
(in Part III) to the geodesic equation of general relativity, in which matter
warps Minkowski space (space–time), and trajectories execute force-free motion
through that space.
Physics and Geometry
1
1.1 State space and dynamical
flows 4
1.2 Coordinate representation of
dynamical systems 10
1.3 Coordinate transformations 15
1.4 Uniformly rotating frames 25
1.5 Rigid-body motion 32
1.6 Summary 48
1.7 Bibliography 48
1.8 Homework problems 49
Foucault’s Pendulum in the Pantheon in Paris
Modern dynamics, like classical dynamics, is concerned with trajectories through

space—the descriptions of trajectories (kinematics) and the causes of trajectories
(dynamics). However, unlike classical mechanics, which emphasizes motions of
physical masses and the forces acting on them, modern dynamics generalizes the
notion of trajectories to encompass a broad range of time-varying behavior that
goes beyond material particles to include animal species in ecosystems, market
prices in economies, and virus spread on connected networks. The spaces that
these trajectories inhabit are abstract, and can have a high number of dimensions.
These generalized spaces may not have Euclidean geometry, and may be curved
like the surface of a sphere or space–time warped by gravity. The central object
of interest in dynamics is the evolving state of a system. The state description of
a system must be unambiguous, meaning that the next state to develop in time is
Introduction to Modern Dynamics. Second Edition. David D. Nolte, Oxford University Press (2019).
© David D. Nolte. DOI: 10.1093/oso/9780198844624.001.0001
4 Introduction to Modern Dynamics
uniquely determined by the current state. This is called deterministic dynamics,

which includes deterministic nonlinear dynamics for which chaotic trajectories
may have an apparent randomness to their character.
This chapter lays the foundation for the description of dynamical systems that
move continuously from state to state. Families of trajectories, called dynamical
flows, are the fundamental elements of interest; they are the field lines of dynamics.
These field lines are to deterministic dynamics what electric and magnetic field
lines are to electromagnetism. One key difference is that there is only one set of
Maxwell’s equations, while every nonlinear dynamical system has its own set of
equations, providing a nearly limitless number of possibilities for us to study.
This chapter begins by introducing general ideas of trajectories as the set of all
possible curves defined by dynamical flows in state space. To define trajectories,
we will establish notation to help us describe high-dimensional, abstract, and
possibly curved spaces. This is accomplished through the use of matrix (actually
tensor) indices that look strange at first to a student familiar only with vectors,
but which are convenient devices for keeping track of multiple coordinates. The
next step constructs coordinate transformations from one coordinate system
to another. For instance, a central question in modern dynamics is how two
observers, one in each system, describe the common phenomena that they
observe. The physics must be invariant to the choice of coordinate frame, but the
descriptions can differ widely.
1.1 State space and dynamical flows

Configuration space is defined by the spatial coordinates needed to describe a
dynamical system. The path the system takes through configuration space is its
trajectory. Each point on the trajectory captures the successive configurations of
the system as it evolves in time. However, knowing the current configuration of
the system does not guarantee that the next configuration can be defined. For
instance, the trajectory can loop back and cross itself. The velocity vector that
pointed one direction at the earlier time can point in a different direction at a later
time. Therefore, a velocity vector must be attached to each configuration to define
how it will evolve next.
1.1.1 State space

By adding velocities, associated with each of the coordinates, to the configuration
space, a new expanded space, called state space, is created. For a given initial
condition, there is only a single system trajectory through this multidimensional
1 See A. E. Jackson, Perspectives of Non-
space, and each point on the trajectory uniquely defines the next state of the
linear Dynamics (Cambridge University system.1 This trajectory in state space can cross itself only at points where all the
Press, 1989). velocities vanish, otherwise the future state of the system would not be unique.
Physics and Geometry 5
Example 1.1 State space of the damped one-dimensional harmonic oscillator
The damped harmonic oscillator in one coordinate has the single second-order ordinary differential equation2
mẍ + γ ẋ + kx = 0 (1.1)
where m is the mass of the particle, γ is the drag coefficient, and k is the spring constant. Any set of second-order
time-dependent ordinary differential equations (e.g., Newton’s second law) can be written as a larger set of first-order
equations. For instance, the single second-order equation (1.1) can be rewritten as two first-order equations
ẋ = v
(1.2)
mv̇ + γ v + kx = 0
It is conventional to write these with a single time derivative on the left as
ẋ = v
(1.3)
v̇ = −2βv − ω02 x
in the two variables (x, v) with β = γ /2m and ω02 = k/m. State space for this system of equations consists of two
coordinate axes in the two variables (x, v), and the right-hand side of the equations are expressed using only the same
two variables.
To solve this equation, assume a solution in the form of a complex exponential evolving in time with an angular
frequency ω as (see Appendix A.1)
x(t) = Xeiωt (1.4)

Insert this expression into Eq. (1.1) to yield
−mω2 Xeiωt + iωγ Xeiωt + kXeiωt = 0 (1.5)

with the characteristic equation
0 = mω2 − iωγ − k
= ω2 − i2ωβ − ω02 (1.6)
where the damping parameter is β = γ /2m, and the resonant angular frequency is given by ω02 = k/m. The solution
of the quadratic equation (1.6) is

ω = iβ ± ω02 − β 2 (1.7)
Using this expression for the angular frequency in the assumed solution (1.4) gives

x(t) = X1 exp (−βt) exp i ω02 − β 2 t + X2 exp (−βt) exp −i ω02 − β 2 t (1.8)
Consider the initial values x(0) = A and ẋ(0) = 0; then the two initial conditions impose the values
2 The “dot” notation stands for a time derivative: ẋ = dx/dt and ẍ = d 2 x/dt 2 . It is a
modern remnant of Newton’s fluxion notation.
Example 1.1 continued
⎛ ⎞
A ⎝ ω0 − β − iβ ⎠
2 2
X1 =
2 ω02 − β 2
⎛ ⎞ (1.9)
A ⎝ ω0 − β + iβ ⎠
2 2
X2 = = X1∗
2 ω2 − β 20
The final solution is ⎡ ⎤

β
x(t) = A exp (−βt) ⎣cos ω02 − β 2 t + sin ω02 − β 2 t ⎦ (1.10)
ω02 − β 2
which is plotted in Fig. 1.1(a) for the case where the initial displacement is a maximum and the initial speed is zero.
The oscillator “rings down” with the exponential
decay constant β. The angular frequency of the ring-down is not
equal to ω0 , but is reduced to the value ω0 − β . Hence, the damping decreases the frequency of the oscillator from
2 2
its natural resonant frequency. A system trajectory in state space starts at an initial condition (x0 , v0 ), and uniquely
traces the time evolution of the system as a curve in the state space. In Fig. 1.1(b), only one trajectory (stream line) is
drawn, but streamlines fill the state space, although they never cross, except at singular points where all velocities vanish.
Streamlines are the field lines of the vector field. Much of the study of modern dynamics is the study of the geometric
properties of the vector field (tangents to the streamlines) and field lines associated with a defined set of flow equations.
(a) (b)
Configuration space State space
1
1.0
A=1
Envelope function β = 0.05
exp(–βt) ω0 = 0.5
0.5
Position
Velocity
0 0
–0.5
–1 –1.0
–1 –0.5 0 0.5 1
0 20 40 60 80 100
Time Position
Figure 1.1 Trajectories of the damped harmonic oscillator. (a) Configuration position versus time. (b) State space, every point
of which has a tangent vector associated with it. Streamlines are the field lines of the vector field and are dense. Only a single
streamline is shown.
1.1.2 Dynamical flows

This book works with a general form of sets of dynamical equations called a
dynamical flow. The flow for a system of N variables is defined as
dq1
= F1 (q1 , q2 , . . . , qN ; t)
dt
dq2
= F2 (q1 , q2 , . . . , qN ; t)
dt (1.11)
..
.
dqN
= FN (q1 , q2 , . . . , qN ; t)
dt
or, more succinctly,
dqa
= Fa (qa ; t) (1.12)
dt
which is a system of N simultaneous equations, where the vector function Fa is

a function of the time-varying coordinates of the position vector. If Fa is not an
explicit function of time, then the system is autonomous, with an N-dimensional
state space. On the other hand, if Fa is an explicit function of time, then the system
is non-autonomous, with an (N + 1)-dimensional state space (space plus time) The
solution of the system of equations (1.12) is a set of trajectories qa (t) through the
state space.
In this book, the phrase configuration space is reserved for the dynamics of
systems of massive particles (with second-order time derivatives as in Examples
1.1 and 1.2). The dimension of the state space for particle systems is even-
dimensional because there is a velocity for each coordinate. However, for general
dynamical flows, the dimension of the state space can be even or odd. For
dynamical flows, state space and configuration space are the same thing, and the
phrase state space will be used.
Example 1.2 An autonomous oscillator
Systems that exhibit self-sustained oscillation, known as autonomous oscillators, are central to many of the topics
of nonlinear dynamics. For instance, an ordinary pendulum clock, driven by mechanical weights, is an autonomous
oscillator with a natural oscillation frequency that is sustained by gravity. One possible description of an autonomous
oscillator is given by the dynamical flow equations

ẋ = ωy + ωx 1 − x2 − y2
(1.13)
ẏ = −ωx + ωy 1 − x2 − y2
continued
where ω is an angular frequency. The (x, y) state-space trajectories of this system are spirals that relax to the unit circle as
they approach a dynamic equilibrium, shown in Fig. 1.2. Without the second terms on the right-hand side, this is simply
an undamped harmonic oscillator. Examples and problems involving autonomous oscillators will recur throughout this
book in Chapters 4 (Chaos), 6 (Synchronization), 7 (Networks), 8 (Evolutionary Dynamics), 9 (Neurodynamics) and
10 (Economic Dynamics).
Figure 1.2 Flow lines of an autonomous oscillator with a limit cycle. All trajectories converge on the limit cycle.
Example 1.3 Undamped point-mass pendulum
The undamped point-mass pendulum is composed of a point mass m on a massless rigid rod of length L. It has a
two-dimensional state-space dynamics in the space (θ , ω) described by
θ̇ = ω
g (1.14)
ω̇ = − sin θ
L
The state-space trajectories can be obtained by integrating these equations using a nonlinear ODE solver. Alternatively,
the state-space trajectories can be obtained analytically if there are constants of the motion. For instance, because the
pendulum is undamped and conservative, the total energy of the system is a constant for a given initial condition,
1
mL2 ω2 + mgL (1 − cos θ )
E= (1.15)
2
referenced to the bottom of the motion in configuration space. If the maximum angle of the pendulum for a given
trajectory is θ0 , then
E = mgL (1 − cos θ0 ) (1.16)
and
1
mL2 ω2 = mgL (cos θ − cos θ0 ) (1.17)
2
which is solved for the instantaneous angular velocity ω as

ω (θ ) = ±ω0 2 (cos θ − cos θ0 ) (1.18)
These are oscillatory motions for θ0 < π. For larger energies, the motion is rotational (also known as libration). The
solutions in this case are

ω (θ ) = ±ω0 2 (cos θ0 − cos θ ) (1.19)
where cos θ 0 is not a physical angle, but is an effective parameter describing the total energy as
E
cos θ0 = 1 + (1.20)
mgL
The (θ, ω) state-space trajectories of the undamped point-mass pendulum are shown in Fig. 1.3. When the state space
pertains to a conservative system, it is also called phase space. Conservative systems are Hamiltonian systems and are
described in Chapter 3.
10
Open orbits
5
Momentum
0 Separatrix
–5
Closed orbits
–10
–5 0 5
Angle
Figure 1.3 State space of the undamped point-mass pendulum. The configuration space is one-dimensional along the angle
θ . Closed orbits (oscillation) are separated from open orbits (rotation) by a curve known as a separatrix.
Example 1.4 A three-variable harmonic oscillator
As an example of an odd-dimensional state space, consider the three-dimensional flow

ẋ = ω0 (y − z)
ẏ = ω0 (z − x) (1.21)
ż = ω0 (x − y)
This mathematical model is equivalent to a three-variable linear oscillator with no dissipation. To solve this flow, assume
a solution in the form of a complex exponential in time evolving with an angular frequency ω as x(t) = Xeiωt . Insert
this expression into Eq. (1.21) to yield
iωX = ω0 (Y − Z)
iωY = ω0 (Z − X) (1.22)
iωZ = ω0 (X − Y )
Solve the secular determinant for the angular frequency ω:

iω −ω0 ω0

ω0 iω −ω0 = iω −ω2 + ω02 + ω0 ω0 iω − ω02 + ω0 ω02 + ω0 iω = 0
(1.23)
− ω0 ω0 iω

ω −ω2 + 3ω02 = 0
The solutions for ω are
√
ω = 0, ± 3ω0 (1.24)
The solutions, for any initial condition, are three sinusoids with identical amplitudes and frequencies, but with relative
phases that differ by ±2π/3. A dynamical system like this is not equivalent to modeling a particle with inertia. It is a
dynamical flow with a state-space dimension equal to three that might model the behavior of an economic system, or
an ecological balance among three species, or a coupled set of neurons. In the study of modern dynamical systems,
the emphasis moves away from particles acted on by forces and becomes more abstract, but also more general and
versatile.
This example has what is called “neutral stability.” This means that even a slight perturbation of this system may
cause the oscillations to either decay to zero or to grow without bound. In Chapter 4, a stability analysis will identify
this system as a “center.” This oscillatory system is not a robust system, because a small change in parameter can
cause a major change in its qualitative behavior. However, there are types of self-sustained oscillations that are robust,
maintaining steady oscillatory behavior even as parameters, and even dissipation, change. These are autonomous
oscillators and are invariably nonlinear oscillators.
1.2 Coordinate representation of dynamical

systems
Although physics must be independent of any coordinate frame, the description
of what we see does depend on which frame we are viewing it from. Therefore, it
often will be convenient to view the same physics from different perspectives. For
this reason, we need to find transformation laws that convert the description from
one frame to another.
1.2.1 Coordinate notation and configuration space

The position of a free particle in three-dimensional (3D) space is specified by
three values that conventionally can be assigned the Cartesian coordinate values
x(t), y(t), and z(t). These coordinates define the instantaneous configuration
of the system. If a second particle is added, then there are three additional
coordinates, and the configuration space of the system is now six-dimensional.
Rather than specifying three new coordinate names, such as u(t), v(t), or w(t),
it is more convenient to use a notation that is extended easily to any number of
dimensions. Index notion accomplishes this by having the index span across all
the coordinate values.
Vector components throughout this text will be denoted with a superscript. For
instance, the position vector of a free particle in 3D Euclidean space is a 3-tuple
of values
⎛ ⎞
x1
⎜ 2⎟
x = ⎝x ⎠ (1.25)
x3
Vectors are represented by column matrices (which is the meaning of the super-
scripts here3 ). It is important to remember that these superscripts are not
“powers.” A coordinate component raised to an nth power will be expressed
as (xa )n . For N free particles, a single 3N-dimensional position vector defines
the instantaneous configuration of the system. To abbreviate the coordinate
description, one can use the notation

x = xa a = 1, . . . , 3N (1.26)
where the curly brackets denote the full set of coordinates. An even shorter, and
more common, notation for a vector is simply
xa (1.27)
where the full set a = 1, . . ., 3N is implied. Cases where only a single coordinate
is intended will be clear from the context. The position coordinates develop in
3 The superscript is a part of the nota-
time as
tion for tensors and manifolds in which
vectors differ from another type of com-
xa (t) (1.28) ponent called a covector that is denoted
by a subscript. In Cartesian coordinates,
a superscript denotes a column vector
which describes a trajectory of the system in its 3N-dimensional configuration and a subscript denotes a row vector (see
space. Appendix A.3).
1.2.2 Trajectories in 3D configuration space

A trajectory is a set of position coordinate values that vary continuously with
a single parameter and define a smooth curve in the configuration space. For
instance,
xa = xa (t) or xa = xa (s) (1.29)
where t is the time and s is the path length along the trajectory. Once the trajectory
of a point has been defined within its configuration space, it is helpful to define
properties of the trajectory, like the tangent to the curve and the normal. The
velocity vector is tangent to the path. For a single particle in 3D, this would be
⎛ ⎞
dx1 (s)
⎜ ds ⎟
⎜ ⎟
⎜ dx2 (s) ⎟ ds
=⎜
v(s) ⎜
⎟
⎟ (1.30)
⎜ ds ⎟ dt
⎝ dx3 (s) ⎠
ds
where the ds/dt term is simply the speed of the particle. In the simplified index
notation, this is
ds dxa (s)
va (s) =
dt ds
ds
= Ta (1.31)
dt
where T a is a unit tangent vector in the direction of the velocity:
dxa
Ta = (1.32)
ds
Each point on the trajectory has an associated tangent vector. In addition to the
tangent vector, another important vector property of a trajectory is the normal to
the trajectory, defined by
dT a
= κN a (1.33)
ds
where N a is the unit vector normal to the curve, and the curvature of the
trajectory is
1
κ= (1.34)
R
where R is the radius of curvature at the specified point on the trajectory.

The parameterization of a trajectory in terms of its path length s is often
a more “natural” way of describing the trajectory, especially under coordinate
transformations. For instance, in special relativity, time is no longer an absolute
parameter, because it is transformed in a manner similar to position.Then it is
possible to define a path length interval ds2

in space–time that remains invariant
under Lorentz transformation (see Chapter 12) and hence can be used to specify
the path through space–time.4
Example 1.5 Parabolic trajectory in a gravitational field
This is a familiar problem that goes back to freshman physics. However, it

is seen here in a slightly different light. Consider a particle in a constant
gravitational field thrown with initial velocity v0 in the x direction. The
mathematical description of this motion is
dx
= v0
dt (1.35)
dy
= −gt
dt
with the solution, for initial conditions x = 0, y = 0, ẋ = v0 , ẏ = 0,
x = v0 t
1 (1.36)
y = − gt 2
2
giving the spatial trajectory
1 g 2
y=− x (1.37)
2 v20
The speed of the particle is
2 2 2
ds dx dy
= +
dt dt dt
= v20 − 2gy
= v20 + g 2 t 2 = v2 (1.38)
with the arc length element

ds = v20 + g 2 t 2 dt

g 2 x2
= 1+ dx (1.39)
v40
and the tangent vector components
dx 1
T1 = =
ds g 2 x2
1+
v40
gx (1.40)
− 4 More generally, the invariant squared
dy v20
T2 = = path length interval ds2 is an essential part
ds g 2 x2 of the metric description of the geometry
1+ of space–time and other dynamical spaces,
v40 and is a key aspect of geodesic motion for
bodies moving through those spaces (see
continued
Chapter 11).
The trajectory and its tangent vector are described as functions of position—
a geometric curve rather than an explicit function of time. While the results
for this familiar problem may look unfamiliar, it is similar to the description
of trajectories in special relativity, or to geodesic trajectories near gravitating
bodies in space–time that will be treated in later chapters.
1.2.3 Generalized coordinates

The configuration coordinates considered so far have been Cartesian coordinates
(x, y, z). However, there are abstract coordinates, called generalized coordinates,
that may be more easily employed to solve dynamical problems. Generalized
coordinates arise in different ways. They may be dictated by the symmetry of
the problem, like polar coordinates for circular motion. They may be defined
by constraints on the physical system, like a particle constrained to move on a
surface. Or they may be defined by coupling (functional dependence) between
the coordinates of a multicomponent system, leading to generalized coordinates
known as normal modes. Generalized coordinates are often denoted by q’s.
They may be described in terms of other coordinates, for instance Cartesian
coordinates, as

qa = qa xb , t
(1.41)
xb = xb (qa , t)
where the transformations associated with each index may have different func-
tional forms and do not need to be linear functions of their arguments. The
generalized coordinates do not need to have the dimension of length, and each can
have different units. However, it is required that the transformation be invertible
(one-to-one).
Generalized coordinates can be used to simplify the description of the motions
of complex systems composed of large numbers of particles. If there are N par-
ticles, each with three coordinates, then the total dimension of the configuration
space is 3N and there is a dense set of system trajectories that thread their way
through this configuration space. However, often there are constraints on the
physical system, such as the requirement that particles be constrained to reside
on a physical surface such as the surface of a sphere. In this case, there are
equations that connect two or more of the coordinates. If there are K equa-
tions of constraints, then the number of independent generalized coordinates is
3N − K and the motion occurs on a (3N – K)-dimensional hypersurface within
the configuration space. This hypersurface is called a manifold. In principle, it
is possible to find the 3N – K generalized coordinates that span this manifold,
and the manifold becomes the new configuration space spanned by the 3N – K
generalized coordinates. Furthermore, some of the generalized coordinates may
not participate in the dynamics. These are called ignorable coordinates (also known
as cyclic coordinates), and they arise owing to symmetries in the configuration space
plus constraints, and are associated with conserved quantities. The dimensionality
of the dynamical manifold on which the system trajectory resides is further
reduced by each of these conserved quantities. Ultimately, after all the conserved
quantities and all the constraints have been accounted for, the manifold that
contains the system trajectory may have a dimension much smaller than the
dimension of the original Cartesian configuration space.
Example 1.6 Bead sliding on a frictionless helical wire
Consider a bead sliding without friction on a helical wire with no gravity. The trajectory is defined in 3D Cartesian
coordinates by
x(t) = R cos ωt
y(t) = R sin ωt (1.42)
z(t) = vz t
parameterized by time t. There are two constraints
x2 + y 2 = R 2
(1.43)
z = aθ
where a is the pitch of the helix and θ = ωt. These constraints reduce the 3D dynamics to 1D motion (3 – 2 = 1), and
the 1D trajectory has a single generalized coordinate

q(t) = t R2 ω2 + v2z (1.44)
which is also equal to the path length s. The speed of the particle is a constant and is

ṡ = R2 ω2 + v2z (1.45)
1.3 Coordinate transformations

For a general coordinate transformation, the original Cartesian coordinates x, y,
and z are related to coordinates q1 , q2 , and q3 by the functions

x = x q1 , q2 , q3

y = y q1 , q2 , q3 (1.46)

z=z q1 , q2 , q3
These equations can be inverted to yield
q1 = q1 (x, y, z)
q2 = q2 (x, y, z) (1.47)
q3 = q3 (x, y, z)
which may be generalized coordinates that are chosen to simplify the equations
of motion of a dynamical system.
1.3.1 Jacobian matrix

The Jacobian matrix of the transformation is defined from the coordinate trans-
formations (and inverse transformations) as
⎛ ⎞ ⎛ ⎞
∂x ∂x ∂x ∂q1 ∂q1 ∂q1
⎜ ∂q1 ∂q2 ∂q3 ⎟ ⎜ ∂x ∂z ⎟
⎜ ⎟ ⎜ ∂y ⎟
⎜ ∂y ∂y ⎟ ⎜ 2 ⎟
⎜ ∂y ⎟ ⎜ ∂q ∂q2 ∂q2 ⎟
J=⎜ 1 ⎟ J −1 =⎜ ⎟ (1.48)
⎜ ∂q ∂q2 ∂q3 ⎟ ⎜ ∂x ∂y ∂z ⎟
⎜ ⎟ ⎜ 3 ⎟
⎝ ∂z ∂z ∂z ⎠ ⎝ ∂q ∂q3 ∂q ⎠
3
∂q1 ∂q2 ∂q3 ∂x ∂y ∂z
The determinant |J| is called the Jacobian.

The Jacobian matrix requires two indices to define its individual elements, just
as a vector required one index. Because the Jacobian matrix is generated using
derivatives, an index notation that distinguishes between the differential vector in
the numerator relative to the differential vector in the denominator is
row index
a
Jba = ∂x (1.49)
∂qb
column index
where the superscript and subscript relate to xa and qb , respectively. The super-
script is called a contravariant index, and the subscript is called a covariant index.
One way to remember this nomenclature is that “co” goes “below.” The covariant
index refers to the columns of the matrix, and the contravariant index refers to the
rows. Column vectors have contravariant indices because they have multiple rows,
while row vectors have covariant indices because they have multiple columns. Row
vectors are also known as covariant vectors, or covectors.
When transforming between Cartesian and generalized coordinates, an
infinitesimal transformation is expressed as
∂xa
dxa = dqb = Jba dqb (1.50)
∂qb
b b
where the Jacobian matrix Jba can depend on position. If the transformation is
linear, then the Jacobian matrix is a constant. The operation of the Jacobian matrix
on the generalized coordinates generates a new column vector dxa .
Rather than always expressing the summation explicitly, there is a common
convention, known as the Einstein summation convention, in which the summation
symbol is dropped and a repeated index—one above and one below—implies
summation:

Einstein summation convention xa = Λab qb ≡ Λab qb (1.51)
b
where the “surviving” index—a—is the non-repeated index. Note that Λba is a
linear transformation. For example, in three dimensions, this is
x1 = Λ11 q1 + Λ12 q2 + Λ13 q3

x2 = Λ21 q1 + Λ22 q2 + Λ23 q3 (1.52)
x3 = Λ31 q1 + Λ32 q2 + Λ33 q3
which is recognizable in matrix multiplication as

⎛ ⎞ ⎛ ⎞⎛ ⎞
x1 Λ11 Λ12 Λ13 q1
⎜ 2⎟ ⎜ 2 ⎟ ⎜ 2⎟
⎝x ⎠ = ⎝Λ1 Λ22 2
Λ3 ⎠ ⎝q ⎠ (1.53)
x3 Λ31 Λ32 Λ33 q3
and is simplified to
xa = Λab qb (1.54)
with the Einstein repeated-index summation. The Einstein summation convention

is also convenient when defining the inner (or “dot”) product between two vectors.
For instance,
A · B = Ax Bx + Ay By + Az Bz
= Aa Ba (1.55)
and the implicit summation over the repeated indices produces a scalar quantity
from the two vector quantities. The inner product in matrix notation multiplies a
column vector from the left by a row vector.
The Jacobian matrix and its uses are recurring themes in modern dynamics.
Its uses go beyond simple coordinate transformations, and it appears any time a
nonlinear system is “linearized” around fixed points to perform stability analysis
(Chapter 4). The eigenvalues of the Jacobian matrix define how rapidly nearby
initial conditions diverge (called Lyapunov exponents—Chapters 4 and 9). The
determinant of the Jacobian matrix is the coefficient relating area and volume
changes (Chapter 11), and is used to prove which processes conserve volumes in
phase space or Minkowski space (Chapters 3 and 12).
1.3.2 Metric spaces and basis vectors

In Cartesian coordinates, basis vectors are the familiar unit vectors x̂, ŷ, ẑ
directed along the coordinate axes. In the case of generalized coordinates, basis
vectors need to be defined through their relationship to the original Cartesian
coordinates. In two dimensions, the differential transformation between coordi-
nates xa and qb is expressed as
∂x 1 ∂x
dx = dq + 2 dq2
∂q1 ∂q
(1.56)
∂y 1 ∂y
dy = dq + 2 dq2
∂q1 ∂q
which is written in matrix form as

⎛ ∂x ∂x ⎞

dx ⎜ ∂q1 ∂q2 ⎟ dq1
=⎜
⎝ ∂y
⎟ (1.57)
dy ∂y ⎠ dq2
∂q1 ∂q2
The square matrix is the Jacobian matrix of the transformation. Transposing this
expression gives
⎛ ⎞
∂x ∂y
⎜ 1
∂q ∂q1 ⎟
(dx dy) = dq1 dq2 ⎜⎝ ∂x
⎟
∂y ⎠
∂q2 ∂q2

e 1
q
= dq dq
1 2
(1.58)
eq2
where the rows of the matrix have become basis vectors (covectors)

∂x ∂y
eq1 =
∂q1 ∂q1
(1.59)
∂x ∂y
eq2 =
∂q2 ∂q2
Basis vectors are used to express elements of the Cartesian vectors in terms of the
curvilinear coordinates as
d x = eqa dqa (1.60)

where the Einstein summation is implied. Vectors describe the properties of

a physical system that cannot depend on the coordinate frame that is chosen:
changing coordinate systems cannot change the physics of the system! Therefore,
vectors such as d x are coordinate-free expressions. Vector components, on the
other hand, do depend on the choice of coordinate system used to express them.
Basis vectors are attached to the coordinate system, defining the elementary com-
ponents along which a vector is decomposed. Basis vectors can vary depending on
their location within the coordinate system, and are not necessarily unit vectors.
A vector can be expressed in terms of basis vectors as
A = A1 e1 + A2 e2 + A3 e3

(1.61)
= Ab eb
which shows how vectors and covectors combine.
Example 1.7 Cylindrical coordinates
The coordinate transformations describing Cartesian coordinates in terms of polar coordinates are
x = r cos θ
y = r sin θ (1.62)
z=z
The basis row vectors for the composition of Cartesian components in terms of the curvilinear coordinates are

∂x ∂y ∂z
er = = (cos θ sin θ 0)
∂r ∂r ∂r

∂x ∂y ∂z
eθ = = (−r sin θ r cos θ 0) (1.63)
∂θ ∂θ ∂θ

∂x ∂y ∂z
ez = = (0 0 1)
∂z ∂z ∂z
1.3.3 Metric tensor

The path length element is a quadratic form that is expressed in terms of
generalized coordinates as
ds2 = gab dqa dqb (1.64)
(Einstein summation implied), where gab is called the metric tensor. To find
an explicit expression for the metric tensor, given a coordinate transformation,
consider the differential transformation between coordinates xa and qb ,
∂xa b
dxa = dq (1.65)
∂qb
which is written out explicitly for three dimensions as
∂x 1 ∂x ∂x
dx = dq + 2 dq2 + 3 dq3
∂q1 ∂q ∂q
∂y 1 ∂y ∂y
dy = dq + 2 dq2 + 3 dq3 (1.66)
∂q1 ∂q ∂q
∂z 1 ∂z ∂z
dz = 1
dq + 2 dq2 + 3 dq3
∂q ∂q ∂q
The square of the first line is

∂x 1 ∂x ∂x 3 2
(dx)2 = dq + 2 dq2 + dq
∂q1 ∂q ∂q3
∂x ∂x 2 ∂x ∂x 2 2 ∂x ∂x 2
= 1 1 dq1 + 2 2
dq + 3 3 dq3
∂q ∂q ∂q ∂q ∂q ∂q
∂x ∂x 1 2 ∂x ∂x ∂x ∂x
+2 1 2
dq dq + 2 1 3 dq1 dq3 + 2 2 3 dq2 dq3 (1.67)
∂q ∂q ∂q ∂q ∂q ∂q
with similar expressions for dy and dz. These squares are added (in quadrature)
to give the squared line element
ds2 = dx2 + dy2 + dz2 (1.68)
which leads to a new expression
ds2 = g11 dq1 dq1 + g12 dq1 dq2 + g13 dq1 dq3
+ g21 dq2 dq1 + g22 dq2 dq2 + g23 dq2 dq3
+ g31 dq3 dq1 + g32 dq3 dq2 + g33 dq3 dq3
= gab dqa dqb (1.69)
in terms of the metric tensor gab . Collecting the coefficients of each of the dqa dqb
terms, and equating ds2 to the right-hand side, yields
∂x ∂x ∂y ∂y ∂z ∂z
g11 = + 1 1 + 1 1
∂q1 ∂q1 ∂q ∂q ∂q ∂q
∂x ∂x ∂y ∂y ∂z ∂z (1.70)
g12 = 1 2
+ 1 2 + 1 2
∂q ∂q ∂q ∂q ∂q ∂q
...
with the general expression
∂x ∂x ∂y ∂y ∂z ∂z
gab = + a b + a b (1.71)
∂qa ∂qb ∂q ∂q ∂q ∂q
Cylindrical coordinates (r, θ, z) x r cos θ

xa = y = r sin θ
z z
z-axis
ds2 = dr 2 + r 2dθ2 + dz 2
r
P (r, θ, z)
O
y-axis
θ
x-axis
Spherical coordinates (r, θ, φ)

x r sinθ cosφ
z-axis
xa = y = r sinθ sinφ
P (r, θ, φ) z r cosθ
θ
ds2 = dr 2 + r 2dθ2 + r 2 sin 2 θdφ2
r z
O y-axis
φ
x-axis
Figure 1.4 Cylindrical and spherical
coordinate systems with line elements
ds2 .
for each element of the metric tensor. Alternatively, one can begin with
ds2 = d r · d r

= ea dqa · eb dqb
= (
ea · eb ) dqa dqb (1.72)
and the assignment can be made directly for
gab = ea · eb (1.73)
where the metric tensor elements are the inner products of basis vectors.
The metric tensor will be described extensively in Chapter 11 as a fundamental

element used in the description of general geometries that go beyond ordinary
Cartesian coordinates. The metric tensor is central to the description of space–
time in special relativity and warped space–time in general relativity.
1.3.4 Two-dimensional rotations

A common class of coordinate transformation consists of rotations. The rotation
matrix is an operator that operates on the components of a vector to express them
in terms of a new rotated set of coordinate axes. The components of a vector Aa
are transformed as

Ab = Rba Aa (1.74)

where Rba is the rotation matrix and Aa are the components of the vector as
viewed in the unprimed frame. For a 2D coordinate frame O that has been rotated
clockwise by an angle θ relative to the unprimed frame O, the rotation matrix that
transforms the vector components (described with respect to the new frame) is

cos θ − sin θ
2D rotation matrix Rba = . (1.75)
sin θ cos θ
It is important to keep a clear distinction between basis vectors (like x̂ and

ŷ) that point along the coordinate axes and the components of the vector A
projected onto these axes. If the basis vectors (and hence the coordinate frame)
are rotated clockwise, then the vector components, as seen from the transformed
coordinate frame, appear to have rotated counter-clockwise. This is shown in Fig.
1.5. Therefore, basis vectors are rotated through the inverse rotation matrix. The
transformation of the basis vectors is therefore
eb = Rab ea (1.76)
which is the inverse of Eq. (1.74):

−1 cos θ sin θ
Rba = Rab = (1.77)
− sin θ cos θ

The inverse transformation Rba has the primed index below, while in the forward

transformation of vector components, Rba , the primed index is above.
A vector quantity is expressed in terms of the basis vectors as

A = Aa ea = Ab eb (1.78)
y
y

Ab = Rba Aa
Ay
Ay
Ax cos θ –sin θ Ax
=
A Ay sin θ cos θ Ay
êy
êx
êy Ax x Figure 1.5 Rotated coordinate axes

Ax
θ through the transformation Rab . The
êx vector A remains the same—only the
description of the vector (vector compo-
x
nents projected onto the axes) changes.
which is independent of the coordinate system—vectors are invariant quantities

(elements of reality) that exist independently of their coordinate description. This
invariance of vectors is described explicitly through the derivation

A = Aa ea

= Aa Rba Rab ea

= Aa Rba Rab ea
= Ab eb (1.79)

where the quantity Rba Rab is the identity matrix.
1.3.5 Three-dimensional rotations of coordinate

frames
Three-dimensional rotations of coordinate axes can be constructed as successive
2D rotations applied around different axes. Three angles are required to express
an arbitrary 3D rotation, and the general rotation matrix can be expressed as

Rad = Rcd (ψ) Rbc (θ ) Rab (φ) (1.80)
where each rotation is applied around a different axis. When applied to a basis
vector ea , this produces the successive transformations

eb = Rab (φ) ea

ec = Rbc (φ) eb (1.81)
c
ed = Rd (φ) ec
where the original primed frame is rotated into the double-primed frame, then the
z
. double-primed frame is rotated into the triple-primed frame, which is rotated into
φ the unprimed frame, which is the resultant frame of the 3D rotation. Although
there is no unique choice for the rotation axes, one conventional choice known
z = z . y as Euler angles uses a rotation by φ around the z axis, then by θ around the x
Ψ θ
axis, and finally by ψ around the z axis (Fig. 1.6). The rotation matrices for this
choice are
⎛ ⎞
y cos φ − sin φ 0
ψ x a ⎜ ⎟
Rz (φ) = Z = ⎝ sin φ
b cos φ 0⎠ (1.82)
x φ
. 0 0 1
θ ⎛ ⎞
x = x 1 0 0
⎜ ⎟
Line of Nodes
Rx (θ ) = Xcb = ⎝0 cos θ − sin θ ⎠ (1.83)
Figure 1.6 Euler angles for a 3D rota- 0 sin θ cos θ
tion. The original primed axes are first ⎛ ⎞
cos ψ − sin ψ 0
rotated by φ around the z-axis (fixed- ⎜ ⎟
frame z-axis), then by θ around the Rz (ψ) = Zdc = ⎝ sin ψ cos ψ 0⎠ (1.84)
x-axis (also known as the line of nodes), 0 0 1
and finally by ψ around the z-axis (body
z-axis). Euler angles are important for describing spinning or rotating systems in terms
of angular velocities. The angular velocities in the body frame are
⎛ ⎞ ⎛ ⎞⎛ ⎞
ω1 sin θ sin ψ cos ψ 0 φ̇
⎜ 2⎟ ⎜ ⎟⎜ ⎟
⎝ω ⎠ = ⎝ sin θ cos ψ − sin ψ 0⎠ ⎝θ̇ ⎠ (1.85)
ω3 cos θ 0 1 ψ̇
with individual components in the body frame being
ω1 = φ̇ sin θ sin ψ + θ̇ cos ψ

ω2 = φ̇ sin θ cos ψ − θ̇ sin ψ (1.86)
ω3 = φ̇ cos θ + ψ̇
These expressions will be useful when solving problems of rotating or tumbling

objects.
The Euler angles are a natural choice for a 3D rotation to describe the
complicated motions of spinning tops (see Section 1.5). However, the choice is
not unique, and a different choice for the rotation matrix in 3D can be used when
there is a single rotation axis and rotation angle θ . For a defined rotation axis given
by a unit vector ûa and a rotation angle θ , the rotation matrix is

3D rotation matrix Rab = Iba cos θ + Sba sin θ + Tba (1 − cos θ) (1.87)

where Iba = δba is the identity matrix, and the other matrices are
⎛ ⎞
0 −uz uy
a ⎜ ⎟
Sb = ⎝ uz 0 −ux ⎠
− uy ux 0
⎛ ⎞ (1.88)
ux ux ux uy ux uz
⎜ ⎟
Tba = ⎝uy ux uy uy uy uz ⎠
uz ux uz uy uz uz

with ua being the Cartesian components of the unit vector. The matrix Tba is the

tensor product of the unit vector with itself, denoted in vector notation as û ⊗ û .

The matrix Sba is a skew-symmetric matrix constructed from the unit vector and

is denoted in vector notation as the operator û× for the cross product. The
structure of the skew-symmetric matrix reflects the geometry of rotations in 3D
space. It is this intrinsic property of 3-space that is the origin of physics equations
containing cross products, such as definitions of angular momentum and torque
as well as equations that depend on the moments of inertia, which are encountered
later in this chapter.
1.4 Uniformly rotating frames

A uniformly rotating frame is an important example of a non-inertial frame. In
this case, acceleration is not constant, which leads to fictitious forces such as the
centrifugal and Coriolis forces.
Consider two frames: one fixed and one rotating. These could be, for instance,
a laboratory frame and a rotating body in the laboratory, as in Fig. 1.7. The fixed
frame has primed coordinates, and the rotating frame has unprimed coordinates.
The position vector in the fixed lab frame is

r = xa êa = x êx + y êy + z êz (1.89)
The position vector in the rotating frame is
r = xa êa = xêx + yêy + zêz (1.90)
relative to the origin of the rotating frame. The primed position vector is then
r = R + r (1.91)
x3
P
r
x3
x2
r

R
x1
x2
Figure 1.7 Coordinates for a rotating

frame. The body frame is the unprimed
frame. The lab, or fixed, frame is primed. x1
Taking the time derivative gives
˙ + d xa ê
r˙ = R a
dt
=R˙ + ẋa ê + xa ê˙
a a
˙ + r˙ + xa ê˙
=R (1.92)
a
(the Einstein summation convention on the repeated index is assumed), where the
last term is a non-inertial term because the basis vectors of the rotating frame are
changing in time.
To obtain the time derivative of the basis vectors, consider an infinitesimal
rotation transformation that operates on the basis vectors of the body frame

d êa
êb = Rab êa = δba êa + dt (1.93)
dt

where the infinitesimal rotation matrix Rba from Eq. (1.87) is expressed to lowest
order in dθ = ωdt as
⎛ ⎞ ⎛ ⎞
1 0 0 0 −ωz ωy
⎜ ⎟ ⎜ ⎟
Rab ≈ ⎝0 1 0⎠ − ⎝ ω z 0 −ωx ⎠ dt (1.94)
0 0 1 − ωy ωx 0
5 Cartesian vector components can be

where the ωa are the Cartesian components of the angular velocity vector ω
denoted with subscripts, but they are not
to be confused with the general covectors along the axes.5 Therefore, the time derivatives of the basis vectors of the body
that are defined in Chapter 11. frame are
d êx
= ωz êy − ωy êz
dt
d êy
= ωz êx + ωy êz (1.95) ωz
dt
d êz
= ωy êx − ωx êy
dt
eˆz
The rotation of the basis vectors by the different components ωa is shown in ωy
Fig. 1.8. eˆx eˆy
Using Eq. (1.95) to express the non-inertial term in Eq. (1.92) gives
xa ê˙a = xωz êy − xωy êz ωx

+ y −ωz êx + y ωx êz Figure 1.8 Angular velocities related to
+ zωy êx − zωx êy (1.96) changes in the basis vectors.
Combining terms gives

xa ê˙a = êx ωy z − ωz y

− êy ωx z − ωz x

+ êz ωx y − ωy x (1.97)
where the result is recognized as the cross product
xa ê˙a = ω
× r (1.98)
Cross products occur routinely in the physics of rotating frames and rotating
bodies, and are efficiently expressed in vector notion, which will be used through
most of the remainder of the chapter instead of the index notation.6 By using Eq.
(1.98) in Eq. (1.92), the fixed and rotating velocities are related by
vf = V + vr + ω
× r (1.99)
the time rate of change in the fixed

This result is general, and, for any vector Q,
frame is related to the time rate of change in the rotating frame as

dQ
dQ
= +ω
×Q (1.100)
dt fixed dt rotating
= ω,
As an example, consider the case Q 6 Vector cross products arise from the
wedge product A ∧ B of Hermann Grass-

dω
dω mann (1844), introduced in The Theory of
= +ω
×ω
(1.101)
dt fixed dt rotating Linear Extension, a New Branch of Mathe-
matics.
where the last term is clearly zero. Therefore,
˙ f = ω
ω ˙ r (1.102)
proving that angular accelerations are observed to be the same, just as linear
accelerations are the same when transforming between inertial frames. This
equality is because the rotating frame is in constant angular motion.
As a second, and more important example, take the time derivative of
Eq. (1.99). This is

d vf d V d vr ˙ × r + ω d r
= + +
ω × (1.103)
dt fixed dt fixed dt fixed dt fixed
The second term on the right is expanded using Eq. (1.100) as

d vr d v
= +ω
× vr (1.104)
The fourth term in Eq. (1.103) becomes

d r d r
×
ω =
ω × +ω
× (ω
× r)
=ω
× vr + ω
× (ω
× r) (1.105)
The acceleration in the fixed frame is then
¨ + a + ω
af = R ˙ × r + ω
× (ω
× r) + 2ω
× vr (1.106)
r
For a particle of mass m, Newton’s second law is
¨ + ma + mω
F f = mR ˙ + r + mω
× (ω
× r) + 2mω
× vr (1.107)
r
which is the force in the fixed frame.

Therefore, in the rotating frame, there is an effective force
¨ − mω
F eff = mar = Ff − mR ˙ × r − mω
× (ω
× r) − 2mω
× vr (1.108)
The first two terms on the right are the fixed-frame forces. The third term is the
effect of the angular acceleration of the spinning frame. The fourth term is the
centrifugal force, and the last term is the Coriolis force. The centrifugal and the
Coriolis forces are called fictitious forces. They are only apparent in the rotating
frame because the rotating frame is not inertial.
Center-of-mass
Centrifugal force
acceleration
.
¨
mar = Ff – mR – mω × r – mω × (ω × r) – 2mω × νr
External force Angular Coriolis force Figure 1.9 Effective force in a frame
acceleration rotating with angular velocity ω.
1.4.1 Motion relative to the Earth

For a particle subject to the Earth’s gravitational field (Fig. 1.10), the effective
force experienced by the particle is
gmax
¨ − mω
F eff = F ext + mg0 − mR ˙ × r − mω
× (ω
× r) − 2mω
× vr (1.109) g0
g
The fourth term is related to the deceleration of the Earth and is negligible. The
−mω × ω × R
( )
centrifugal term is re-expressed as

¨ = ω
R × ω × R gmin
=ω ˙
×R (1.110)
The effective force is then

Figure 1.10 Geometry for motion rela-
F eff = F ext + mg0 − mω
× ω × r + R − 2mω
× vr (1.111) tive to the Earth.
and redefining the effective gravitational acceleration through

geff = g0 − ω
× ω × r + R (1.112)
gives
F eff = F ext + mgeff − 2mω

× vr (1.113)
This last equation adds the centrifugal contribution to the measured gravitational
acceleration. The last term in Eq. (1.113), −2mω× vr , is the Coriolis force that has
important consequences for weather patterns on Earth, and hence has a powerful
effect on the Earth’s climate (Fig. 1.11). It is also a sizeable effect for artillery
projectiles. On the other hand, it plays a negligible role in the motion of whirlpools
in bathtubs.
Figure 1.11 Dramatic example of cyc-

lone motion in the Northern Hemisphere
for a low-pressure center between Green-
land and Iceland.
1.4.2 Foucault’s pendulum

z
Foucault’s pendulum is one of the most dramatic demonstrations of the rotation
of the Earth. It also provides a direct measure of latitude.
A simple pendulum (Fig. 1.12) swinging at a latitude λ will precess during the
day as a consequence of the Coriolis force. The acceleration in the rotating Earth

frame is
1
T
y ar = g + T − 2ω
× vr (1.114)
m
mg
where the components of the tension and angular velocity are
x
x
Tx = −T ωx = −ω cos λ

Figure 1.12 Geometry of a Foucault y
pendulum of mass m attached to a mass- Ty = −T ωy = 0 (1.115)

less string of length supporting a ten- z
Tz = −T ωz = ω sin λ
sion T.
The cross product is

⎛ ⎞⎛ ⎞
0 −ωz ωy vx
⎜ z ⎟⎜ ⎟
× vr = ⎝ ω
ω 0 −ω ⎠ ⎝vy ⎠
x
− ωy ωx 0 vz
⎛ ⎞⎛ ⎞
0 −ω sin λ 0 vx
⎜ ⎟⎜ ⎟
= ⎝ω sin λ 0 ω cos λ⎠ ⎝vy ⎠
0 −ω cos λ 0 0
⎛ ⎞
−ωẏ sin λ
⎜ ⎟
= ⎝ ωẋ sin λ ⎠ (1.116)
− ωẏ cos λ
and the acceleration in the rotating frame is
Tx
arx = ẍ = + 2ẏω sin λ
m (1.117)
Ty
ary = ÿ = − − 2ẋω sin λ
m
leading to the coupled equations
ẍ + ω02 x = 2ωz ẏ
T g (1.118)
ÿ + ω02 y = −2ωz ẋ where ω02 = ≈
m
The coupled equations are added in quadrature (See Appendix A.3) to yield
(ẍ + i ÿ) + ω02 (x + iy) = −2iωz (ẋ + i ẏ) (1.119)
This is converted into a single second-order equation through substitution of the

variable q = x + iy to give
q̈ + 2iωz q̇ + ω02 q = 0 (1.120)
which is the equation of a harmonic oscillator with imaginary damping. The

solution is
⎡ ⎤
i ωz2 +ω02 t −i ωz2 +ω02 t
q(t) = e−iωz t ⎣Ae + Be ⎦ (1.121)
For a typical pendulum, ω0 >> ωz and the solution simplifies to

q(t) = e−iωz t Aeiω0 t + Be−iω0 t (1.122)
Another random document with
no related content on Scribd:
the next evening, if the weather permitted. This programme had
been drawn up by my husband and myself two or three weeks
before Mr. Quinton’s arrival, but it has since come to light that the
Jubraj suspected us of treachery in asking him to arrange and be
present at these nautches.
NATIVES OF THE MANIPUR HILLS.

We had never seen so many people in the Residency at once as
there were that Sunday night at dinner—fifteen in all. I felt rather
forlorn, being the only lady present, and wished that I had even one
familiar spirit in the shape of another woman to keep me company.
The band was very much appreciated, and everything seemed
very bright and cheery. No thought of evil troubled any of us, for little
we knew that it was the last evening we were to spend in peace
there all together.
The next day Mr. Melville was to leave us. I had tried to persuade
him to stay longer, as he had only been two days with us before, and
had seen nothing of the place; but his time was precious, and he had
his work to do in a great many other places, so we could not get him
to alter his arrangements. He agreed to compromise matters by
remaining until the afternoon of the following day, the 23rd, and
about eleven the party broke up and retired to rest.
CHAPTER XIV.
Up early on the eventful morning—The Jubraj does not attend the Durbar—Visit of
Mr. Grimwood to the Jubraj—Finds him in high fever—Matters assume a
serious aspect—Thoroughfares deserted—Terrific thunderstorm—Our
servants take French leave—My ayah deserts—Melancholy thoughts—Lovely
moonlight night—A Manipuri arrives to spy out our doings—The night before
the outbreak—Attack on the Residency—Capture of the Jubraj’s house—
Anxiety about Lieutenant Brackenbury—Stray bullets find their billet in the
Residency—Attack gets hot, and big guns play on the Residency—We have
to take to the cellars—The Regent invites Mr. Quinton to an interview.
It is now some time since the events took place which I am

recording here, and not one vestige of the past remains to help me in
my work, not a single scrap of writing or note of any kind; yet the
smallest detail of those few terrible days is engraved so indelibly
upon my memory that it seems but the occurrence of yesterday, and
I need no reference to help me in my description of a catastrophe
which almost outrivals some of the horrors of the Mutiny.
We were all up early on the morning of the 23rd. The durbar was
fixed to take place at eight, and the rooms had to be prepared for the
ceremony. But when eight o’clock came, it brought only a message
from the palace, saying that the Jubraj was too ill to leave his house,
and therefore the regent had not come; so the red cloth arranged for
his reception was put away, and a consultation took place between
the Chief and my husband as to what the next move should be.
It was decided to make one last attempt to get the princes to
attend, and then if that failed, other measures were to be resorted to.
But twelve o’clock brought no better results, and about four my
husband was sent to the palace to see the Jubraj, and convey to him
personally the orders of the Government, and use all his influence to
persuade the prince to give himself up quietly, telling him at the
same time that the proposed banishment was not to last for ever, but
that it would depend chiefly on his good behaviour, and eventually, at
the death of his brother, the regent, he (the Jubraj) would be allowed
to return to Manipur, and ascend the throne as Maharajah. It was a
veritable hornets’ nest into which my husband ventured that
afternoon, accompanied only by his friend Mr. Simpson. He would
not take even a single orderly with him, knowing in what an excited
state the whole palace was at that time. It was crowded with Sepoys,
collected, the regent told him, for the review which we had desired to
witness. I got very anxious about them both when more than an hour
had passed and they had not returned, but when my husband did
come back I knew at a glance that his mission had failed. He said
the Jubraj was certainly very unwell. He had had some difficulty at
first in persuading the prince to come and see him at all, but after
finding out from his people that the two Sahibs had come quite
alone, the Jubraj had himself carried down to see them in a litter.
The exertion caused him to faint, and my husband said that there
was no doubt as to his illness, and that he found him in high fever.
Shortly before this visit to the palace took place, Mr. Melville
started off on the first stage of his journey to Kohima. He intended
travelling as far as Sengmai that day, a distance of ten miles. My
husband tried to make him reconsider his decision to go, and I
added my persuasions to his. We did not like the look of things at all,
and how matters would end was, to say the least, very uncertain.
I remember so well our all standing on the steps of the Residency
that afternoon, watching the coolies collecting Mr. Melville’s luggage,
and begging him to remain even one day longer with us, for fear of
anything going wrong; and I remember equally well his answer:
‘Thank you very much, Mrs. Grimwood,’ he said, ‘but do not fear for
me. I am not important enough to be captured by these Manipuris. I
shall get on all right, never fear’; and in a few more minutes he had
left us. But the thoughts of his going away like that, alone, without
any guard to protect him, troubled me more than once that
afternoon, and I could not get it out of my head. Matters assumed a
serious aspect indeed when my husband returned about six o’clock
from the palace with the news that he had been unsuccessful with
regard to the Jubraj. There was only one way, then, out of the
difficulty, and that was to place the affair in the hands of the military,
and apply force where persuasion had failed. It was a council of war,
indeed, and everything seemed to combine to fill me with sensations
of dread for what was going to happen. I could not feel the
excitement that took possession of the men when the chances of a
probable fight became known. Such an idea filled me with alarm and
horror. The place had a deserted look about it, and on the principal
road, as a rule crowded with people at that hour, not a soul was
visible.
The clouds had been gathering up all the afternoon, and about
seven o’clock a terrific thunderstorm occurred, and darkness set in,
which was only lit up now and then by brilliant flashes of lightning. I
busied myself about the house, where I found a state of confusion
reigning. A number of the servants had taken French leave and
departed, scenting danger.
My old ayah was among the first to go. She had been with me four
years, and had followed me about faithfully till now; but at the first
sign of danger she packed up her belongings and went off. I
wondered where she had taken refuge, for she had a good many
enemies, and was not a native of Manipur, so had no home or
relations in the place to whom she could fly for protection. I felt her
desertion very much. She was only a native, but she was at any rate
a woman, and better than no one in a case like that. However, there
was no good to be got out of thinking over her departure, and I had
as much as I could do as it was to keep the other servants up to the
mark, and get them to understand that dinner that evening would
have to be gone through the same as usual. Mr. Quinton and three
of the others amused themselves by playing whist until dinner-time,
as, of course, going out was an impossibility; and I went to the
kitchen to superintend the arrangements there, and to make
preparations for the next day, as I knew that if there were fighting
going on, I should be left without a single servant, and so resolved to
get as much work out of them while they were there as was possible.
We made a quantity of soup that night, as I thought it would be
useful, and cooked some fowls to provide us with something to eat
the next day in case of accidents. And then we had dinner.
No one seemed inclined to speak much that evening. With me
conversation was almost an impossibility, and the rest were too
excited about the morrow to be able to talk and laugh as they had
done the day before. It was a relief to me when dinner was over. I felt
nervous and low-spirited, and very lonely, quite out of place amongst
those men whose profession it was to fight, and who were longing
for the next morning.
Thoughts of England and of all whom I loved there, flocked
through my mind, and I wondered what they would say if they could
see us then, and know the possible danger that threatened us and
our home. My husband was troubled at the thoughts of my being in
the place at such a time, and he blamed himself for having agreed to
my staying, though I had done so of my own free will. Even then we
did not dream of any really serious ending. We expected that the
Jubraj would fight well—in fact, the officers and Sepoys were hoping
that the resistance would be strong, and my husband was afraid that
the house might get knocked about, and some of our property
destroyed; but serious alarm for our own safety never entered our
heads. This was the night of the 23rd, the date that we had originally
fixed for the Manipuri nautch to take place; but under the
circumstances we did not think it likely that the girls would come. Mr.
Brackenbury amused us by singing comic songs, accompanying
himself on his banjo after dinner, and all went to bed early, as
everyone had to be up at three the next morning.
It was a lovely moonlight night, and my husband and I walked up
and down in the garden for some time after our guests had gone. I
felt restless and unhappy, but he did his best to reassure me and
make me believe that we should all be perfectly safe. Just before we
were preparing to go in, the sentry challenged at the gate, and
appeared a few minutes afterwards with a Manipuri, who had been
sent from the palace to inquire whether we wished to have the
nautch or not, saying at the same time that the girls who were to
dance were waiting outside in the road if we wanted them. Of course
we told him it was much too late to think of such a thing at that time,
and the man left the place. We believed that he had really been sent
to spy out the land, and find out what preparations, if any, we had
been making. If that were his mission, he must have been seriously
disappointed, for the whole place, and everyone in it, was wrapt in
slumber, with the exception of my husband and myself, and we very
often walked about the grounds late on moonlight nights, so there
was nothing unusual in our doing so on this occasion. There were a
few extra sentries on guard, but chiefly at the back of the house, and
the presence of the Chief Commissioner was quite sufficient cause
for a larger guard than usual.
At last we turned in too, one to sleep as unconcernedly as ever,
knowing not that it was his last night upon earth; the other to lie
awake, listening to the hours as they were struck out on the gong
down at the quarter-guard, and wonder what the ending of the next
day would bring. I never closed my eyes all through the watches of
that night—the last I was destined to spend in Manipur—and when
three o’clock came I woke my husband, and told him that the hour
had come when we were all to get up, and the work of day
commenced.
It was a bitterly cold morning, and quite dark. I dressed quickly,
putting on a warm, tight-fitting winter dress. We had a sort of scratch
breakfast of eggs and bread-and-butter about 3.30 a.m.; but most of
the officers had theirs at the camp, and started from there, so that I
did not know when they actually left to commence the attack. My
husband accompanied Colonel Skene, much to my distress, as I
thought he would have stayed at the Residency, being a civilian; but
he seemed just as keen on going as the others, and I had to make
the best of it. Mr. Quinton, Mr. Cossins, and I all went off to the
telegraph-office, which was situated at the end of the drive, about
three hundred yards from the Residency. It was well built and fairly
strong, the basement being made of stone, and there was a similar
building on the opposite side, which my husband used as an office
for himself, the lower half of which contained the treasury. Here we
took up our position, going up into the telegraph-office first to send a
message to the Government of India, giving details of all that had
occurred up to date. Mr. Cossins, who was acting as secretary to the
Chief Commissioner, brought the telegram down, and while we were
waiting and watching the Baboo[15] despatch it, we heard the first
shot fired in the palace, which was followed up quickly by others, and
we knew then that the fight had begun.
By this time the dawn was breaking, and streaks of daylight were
dispelling the darkness around. It seemed difficult to me to realize
what was really taking place. I had heard firing in the palace so often
that it seemed almost impossible to understand that a sterner game
was being played now, and one which was to cost both sides so
dear. Only half the telegram had been sent, when we were startled
by the sudden advent of a bullet through the office window at our
elbows. It crashed through the glass, breaking it to pieces, and went
into the wall opposite. My heart went to my mouth with fright, and I
left the place with considerable rapidity, taking up a more secure
position below, where I was fairly protected by the stone basement of
the building. Mr. Cossins occupied his time in taking several journeys
up to the house, where he mounted to the roof to discover whether
he could see what was going on inside the palace wall; but it was
impossible, as the Jubraj’s house, which we knew was being
attacked, was some distance off, and hidden from us by intervening
buildings. It was situated near the outer wall of the palace, and our
men seemed to have taken a very short time in getting up to it.
The whole palace was fortified. Five walls surrounded the
Maharajah’s enclosure. The outer of these was much broken, and of
no great height; but the inner ones were very strong, built of brick
and supplied with bastions, and they surrounded the inner palace on
all four sides. On three sides of the outer wall was a canal, very deep
and wide. It was here that the great boat-races took place every
year, and the water was always kept weeded and clean for those
events. The whole citadel was built with a view of resisting attack in
the time before Burmah was annexed, when armies of raiders used
to come down upon Manipur with hostile intent; and it was a place
which could easily be held against an attacking force, provided big
guns were not brought to bear upon it. The Manipuris were well
armed, and supplied with ammunition. The Maharajah had four
mountain guns which had been presented to his father by our
Government in return for services which he had rendered in times
gone by. The Jubraj understood perfectly how to work these guns.
We had seen him fire them himself for our amusement on an
occasion already described, and we knew he would be perfectly
cognizant of their powers of destruction when the opportunity
occurred to bring them into play against us. Of course we, who were
left at the Residency, did not know what was going on round the
Jubraj’s house, where all the firing seemed to come from. From time
to time stray bullets came over our heads where we sat down at the
telegraph-office. I thought it was very exciting then, and the little
Ghoorkas, who had remained to keep guard over the place, were
constantly running out on to the road in front of our entrance-gate, to
see whether they could discover what was happening. They did not
like being inactive at all.
About half-past ten my husband returned, and came to the
treasury to get out some of the reserve ammunition which had been
stored there. He only stayed a few minutes, talking to me before
rejoining Colonel Skene. He told me that the Jubraj’s house had
been captured after a good fight, and that our men were in
possession of it, and the principal gateway besides, and had taken a
good many prisoners. I asked if anyone had been hurt, and he said
there were grave rumours about Lieutenant Brackenbury. No one
seemed to be certain of his whereabouts, while some affirmed that
he had been wounded, and others that he had been killed. We were
very anxious about him, but my husband said that it was all
uncertain, and he might be perfectly safe all the time; and of course
we hoped he was all right.
About twelve Mr. Quinton and I went up to the house, but long
before going he had made another attempt to get the telegram to the
Government of India despatched, and had found that the wires had
been cut on all sides, so all hope of communication from that source
was abandoned. We were rather hungry by twelve, as no one had
eaten much at the hasty repast at three in the morning, and we were
very glad of some hot tea and sandwiches now. I went on a voyage
of discovery round the house. One or two servants still remained, but
they seemed very frightened, and were saying many prayers to their
gods for their safety. A stray bullet or so had hit the walls of the
house, knocking off some of the plaster, but otherwise everything
looked the same as usual.
SKETCH MAP OF MANIPUR.
We returned to the office in about an hour, after I had seen that all
the preparations for lunch were made. The cook had departed, but
the bearer and I between us managed to get things ready in a
fashion. I took a book to read with me, and busied myself in that
manner until, about one o’clock, Colonel Skene and some of his
officers, with my husband, returned from the scene of action. Our
first inquiries were for Mr. Brackenbury, and then it became evident
that something serious had happened to him, and all our fears were
aroused. After that, things seemed to assume quite a different
aspect, for the officers were all talking so gravely together, and did
not seem quite satisfied with the way things were going.
However, we went back to the Residency to get something to eat.
All had returned with the exception of Mr. Simpson and Captain
Butcher, who were still at the Jubraj’s house, and Mr. Brackenbury,
whose exact whereabouts were unknown. We had commenced
lunch, when my husband asked me if I would give orders that some
food should be sent to the two officers who were not able to leave
their posts, and I went away to a little room adjoining the dining-room
and commenced cutting sandwiches for them, as the servants had
disappeared, and one had to get everything for one’s self or go
without.
I had been busily engaged for about ten minutes, when I heard a
sound which filled me with alarm, and a bullet crashed through the
window above my head. It frightened me more than the one at the
telegraph-office had done, and I dropped my knife, left the
sandwiches as they were, and rushed into the dining-room. All the
officers meanwhile had gone out, and had found that the Manipuris
had crept round to the back of the Residency and commenced an
attack upon us, using as cover the Naga village which lay between
our grounds and the river. This was a clever move on their part, and
it was some time before the troops could drive them back, as most of
our men were engaged in holding the posts inside the palace
captured early in the morning, and this left only a small guard for the
Residency, treasury offices, and Sepoys’ camp. Eventually our party
set fire to the Naga village, and drove the Manipuris out. Bullets had
made their way through the window-panes and doors of the dining-
room, and had smashed some of the breakfast-things and the glass
on the sideboard. It was difficult to find out the most secure place in
the house, as the firing was hot in the front of the Residency by this
time, and the walls, being only lath and plaster, were little or no
protection.
My husband suggested my descending to the cellars, which were
under the house and built of stone; but I did not like the idea, and
remembered how scornful I had been when we had talked over
matters weeks before, and he had joked about the snug corner he
would make ready for me in the basement of the house. So I made
up my mind to remain above-board, so to speak, until the worst
came to the worst. It was heart-rending to see the work of
destruction which was proceeding in the different rooms meanwhile.
The windows were broken, and every now and then bullets crashed
into the rooms, smashing different things—first a picture, then a
vase, then a photograph. All my beloved household gods seemed
coming to grief under my very eyes, and I was powerless to save
them. We did try to collect some of the most valuable of our
belongings together and put them away in a heap in the durbar
room, which at that time had escaped with only one broken pane; but
it was dangerous work going into the front rooms to remove them, for
as the afternoon went on the firing became hotter, and bullets rained
into the house at every second.
It must have been about half-past four that the big guns began to
be played against us. It had been found necessary to concentrate
the whole of our force on the Residency and out-buildings, such as
the treasury and offices, and this entailed abandoning all the
positions captured in the early part of the day inside the outer wall of
the palace, and bringing all the men together. The wounded had to
be recovered from all directions and conveyed to the hospital, which
was some distance from the Residency.
Lieutenant Brackenbury had been discovered lying on the bank of
the river which flowed north of the palace, where he had fallen
shortly after the attack was made early in the morning. He had
mistaken the direction, having got the wrong side of the wall near the
Jubraj’s house, from which point he had been exposed to a heavy
fire from the enemy. It was only a marvel that he was still alive when
eventually discovered, for he had remained where he fell the whole
of that day, and the Manipuris had never ceased firing at him as he
lay. When his exact whereabouts did become known, it was a difficult
and dangerous task to remove him. Efforts had been made by some
of the Sepoys to drag him away, and a native officer had been
mortally wounded in the attempt. At last, about four o’clock in the
afternoon, he was rescued and brought into the hospital, and it was
found that he had received terrible injuries, being wounded in several
places.
The sound of the first shell which whizzed over the Residency
made me speechless with terror. I had heard the boom of the guns in
the morning, and knew that they had been used to try and drive
Captain Butcher’s party out of the Jubraj’s house, which had been
captured; but they had sounded some distance off, and I had not
realized how terrible they could be until they were turned against our
own house.
The cellars were by this time unavoidable. My husband told me
that we should have to make some sort of rough hospital in one of
them, as the Residency hospital, where the wounded had been
taken, was built of plaster and would not be bullet-proof; so we set to
work to get blankets and sheets down from the house, and
everything we thought might be useful.
Meanwhile shells were doing dreadful damage over our heads,
and we were afraid they might set fire to the thatch and force us out
of our temporary shelter. Luckily most of them went over the house
into the garden at the back, where they could not do such serious
damage; but the noise the guns made, added to the other firing,
which had never ceased, was deafening.
There was not the slightest doubt by this time that our position was
about as bad as it could very well be. I seemed paralyzed with fear,
and it was only by forcing myself to do something, and never thinking
or imagining for one moment what the end of it all might be, that I
kept my senses sufficiently to be able to make an effort to help the
rest. I heard that the wounded were to be brought up to the house
immediately, as the hospital was getting too hot for them to remain in
it. Poor fellows! they had endured so much as it was in getting there,
that it seemed very hard to be obliged to move them again so soon,
and take them up to the Residency.
There were a good many of us in the cellar by this time—Mr.
Quinton, Colonel Skene, my husband and myself, Mr. Cossins, and
Mr. Gurdon. It was about seven o’clock, and a lovely evening. The
sun was just setting, and the red glow of the sky seemed to
illuminate the landscape around and the faces of the colonel and my
husband as they stood in the doorway talking together in low tones.
It was no difficult matter to read what was written on both their faces,
and I did not dare ask what was going to happen.
At last my husband came and told me that we were to leave the
Residency, and try and find our way to Cachar. It seemed worse to
me to think of going out of the house than to remain there; but
whatever was to take place had to be at once, and there was no time
to spend in giving way to the terrible fear which possessed me.
However, a further consultation was held, and it was decided to
make a truce with the regent, and put an end to hostilities by coming
to some terms with him. A letter was written, which the Chief
Commissioner signed. It ran as follows:
‘On what condition will you cease firing on us, and give us
time to communicate with the Viceroy, and repair the
telegraph?’
While this letter was being written, the colonel had ordered our
buglers to sound the ‘cease fire,’ which they did at once; but it was
some time before the Manipuris followed suit. At last their guns
ceased, and all was quiet. Then my husband went out with the letter,
and called a Manipuri off the wall to take it to the Jubraj. The man
went away with it, and my husband returned to the Residency.
FRANK ST. CLAIR GRIMWOOD.
FROM A PHOTOGRAPH BY VANDYK.
LONDON: RICHARD BENTLEY & SON, 1891

Some minutes later a message came to say that the regent
wished to see Mr. Quinton and talk over matters with him; and this
message was followed by a letter written in Bengali, which contained
an acknowledgment of the Chief’s letter, and a proposal to the effect
that we should surrender our arms if the Manipuris agreed to cease
firing. There was some discussion about the translation of part of this
letter, and Mr. Quinton proposed that the Jubraj should be called
upon to explain the meaning of the passage in question, and asked
whether it would be possible to see him.
Meanwhile the Chief Commissioner’s party, consisting of himself,
Colonel Skene, Mr. Cossins, Lieutenant Simpson, and my husband,
had gone down to the office at our entrance gate, and waited there
while the regent’s letter was being translated. Mr. Simpson had gone
of his own accord, as he wished to accompany my husband, and I
had begged to be allowed to go with him too; but he said I was safer
where I was, and bade me good-bye, telling me to keep a brave
heart, that the firing was at an end, and peace about to be restored;
and he told me to try and get some rest, as I looked so tired. I never
saw him again.
CHAPTER XV.
Mr. Brackenbury—Scenes in the little cellar—Destruction of our home—Another
moonlight night with a difference—Reopening of the attack on the Residency
—Death of Mr. Brackenbury—Preparations to escape.
I remained where he had left me, alone for some minutes, though
some of the officers were standing just outside the door of the cellar
where I was sitting. It seemed so hard that I could not go with my
husband. I feared being left alone without him, and felt very lonely
and broken-hearted among so many men, mostly strangers to me. I
knew, too, that they would look upon me as an extra burden, and
wish me very far away.
I was roused to action by the doctor, who had taken advantage of
the truce to get his wounded brought up from the hospital to the
house, and had come up first to see what kind of a place could be
got ready. I showed him the cellars, for there were several, which all
communicated with each other, and formed the entire basement of
the house.
Shortly afterwards the Kahars[16] arrived, carrying poor Mr.
Brackenbury on a mattress, and the others followed fast, so that the
small cellar was very soon quite full of men lying side by side on the
stone floor. The blankets and sheets that we had already collected
were very useful, and I made several journeys up to the house, and
gathered up every kind of covering from every direction, and all the
pillows I could find. A little cooking-stove proved of great service. I
fixed it securely upon a table in one corner which I reserved for
cooking operations.
The soup we had made on the previous day was in great request.
Fortunately there was a large quantity of it, to which I added the
contents of five or six tins which I found in the store-room. Milk was
the difficulty. All the cows were out in the grounds, and many of them
had strayed away altogether and we could not get any milk from
them, so were obliged to fall back on condensed milk, of which we
also had several tins.
Some of the men were terribly wounded, but poor Mr. Brackenbury
was by far the worst. His legs and arms were all broken, and he had
several other injuries besides. It seemed a marvel that he was still
alive and fully conscious to all that was going on around him. The
doctor attended to him first of all, and had bound up his broken
limbs, and done as much as possible to alleviate his sufferings; but it
was a terrible sight to see the poor lad in such agony, and be so
powerless to lessen it in any way. He was very thirsty, and drank a
good deal of soup and milk, but we could not get him into a
comfortable position. One minute he would lie down, and the next
beg to be lifted up; and every now and then his ankle would
commence bleeding, and cause him agony to have it bound up
afresh. His face was gray and drawn, and damp dews collected on
his forehead from the great pain he was suffering.
That scene will never be forgotten—the little cellar with a low roof,
and the faces of the wounded lying together on the floor. We did not
dare have a bright light for fear of attracting attention to that
particular spot, and the doctor did his work with one dim lantern.
Such work as it was, too! Every now and then he asked me to go
outside for a few moments while the dead were removed to give a
little more space for the living.
There were some terrible sights in the cellar that night—I pray I
may never see any more like them; but being able to help the doctor
was a great blessing to me, as it occupied my attention, and gave
me no time to think of all the terrible events of the day, and the wreck
of our pretty home. I was very weary, too—in fact, we all were—and
when at about half-past ten I asked everybody to come and get
some sort of a dinner, they seemed much more inclined to go to
sleep, and no one ate much.
The dinner was not inviting, but it was the best that could be got
under the circumstances, for I had had to do it all myself. One or two
of the servants still remained, but they cowered down in corners of
the house, and refused to move out or help me at all. Perhaps had
we known that it was our last meal for nearly forty-eight hours, we
should have taken care to make the most of it; but no thought of
what was coming entered our minds, and long before the melancholy
meal was ended most of the officers were dozing, and I felt as
though I could sleep for a week without waking.
We all separated after dinner about the house. I went back to the
hospital for a little, and found the doctor wanted more milk, so I
returned to the dining-room, where I was joined by Captain Boileau,
and we sat there for some time mixing the condensed milk with
water, and filling bottles with it, which I took downstairs. It was
quieter there than it had been. The wounded had all been attended
to, and most of them had fallen asleep. Even Mr. Brackenbury was
dozing, and seemed a little easier, and only one man was crying out
and moaning, and he was mortally wounded in the head. So finding I
could do no more there, I went upstairs again, resolving, if possible,
to go to my room and lie down for a little while and sleep, for I was
very tired.
I went sorrowfully through our once pretty drawing-room, where
everything was now in the wildest confusion, and saw all the
destruction which had overtaken my most cherished possessions.
There are those who imagine that in a case like this a woman’s
resource would be tears; but I felt I could not weep then. I was
overwhelmed at the terrible fate which had come upon us, and too
stunned to realize and bewail our misfortunes.
Perhaps the great weariness which overcame me may have
helped me to look passively on my surroundings, and I walked
through the house as one in a dream, longing only to get to some
haven of rest, where I could forget the misery of it all in sleep.
I wended my way to the bedroom through a small office of my
husband’s, but when I reached the door I found it would not open,
and discovered that part of the roof had fallen in, caused by the
bursting of a shell. So I gave up the idea of seeking rest there, and
retired to the veranda.
I went down the steps and stood outside in the moonlight for a few
minutes. It was a lovely night, clear and bright as day! One could
scarcely imagine a more peaceful scene. The house had been
greatly damaged, but that was not apparent in the moonlight, and the
front had escaped the shells which had gone through the roof and
burst all round at the back. The roses and heliotrope smelt heavy in
the night air, and a cricket or two chirped merrily as usual in the
creepers on the walls.
I thought of the night before, and of how my husband and I had
walked together up and down in the moonlight, talking of what the
day was to bring, and how little he had thought of such a terrible
ending; and I remembered that poor lad lying wounded in the cellar
below now, who only twenty-four hours ago had been the life and
soul of the party, singing comic songs with his banjo, and looking
forward eagerly to the chances of fighting that might be his when the
morning came.
I wondered where my husband was, and why they had been away
so long. They would be hungry and tired, I thought, and might have
waited to arrange matters till the next day, as they had apparently
been successful in restoring peace. I had an idea of wandering as far
as the gate to see whether the party was visible, but on second
thoughts I went back into the veranda, and resolved to wait there
until my husband should return.
There was one of the officers asleep in a chair close to me, and I
was about to follow his example, when Captain Boileau came out,
and I went to him and asked him if he would mind going down to the
gate and finding out whether he could hear or see anything of the
Chief Commissioner’s party, and if he came across any of them to
say I wanted my husband. He went off at once, and I fell into a doze
in the chair.
It was about twelve o’clock at this time. I do not know how long I
had been asleep, when I was awaked suddenly by hearing the
deafening boom of the big guns again, and knew then that it was not
to be peace.
For a few seconds I could not stir. Terror seemed to have seized
hold of me, and my limbs refused to move; but in a minute I
recovered, and ran through the house down to the cellar again,

Introduction To Modern Dynamics Chaos Networks Space and Time 2Nd Edition David D Nolte Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Introduction To Modern Dynamics Chaos Networks Space and Time 2Nd Edition David D Nolte Full Chapter

Uploaded by

Copyright:

Available Formats

Introduction to Modern Dynamics:

Chaos, Networks, Space, and Time 2nd

The goal of the second edition of Introduction to Modern Dynamics is to

A unifying concept: geometry and dynamics

A common tool: dynamical flows and the

Traditional junior-level physics: how to use

What’s simple in complex systems?

• Chaos theory (Chapter 4)

Much of nonlinear dynamics can be understood through linearization of the

The nonlinear synchronization of two or more oscillators is a starting point for

• Network theory (Chapter 7)

• Evolutionary dynamics (Chapter 8)

• Neural networks (Chapter 9)

• Econophysics (Chapter 10)

• Geodesic motion (Chapter 11)

• Special relativity (Chapter 12)

In addition to traditional topics of Lorentz transformations and mass-energy equiv-

• General relativity (Chapter 13)

I gratefully acknowledge the many helpful discussions with my colleagues

Traditional approaches to the mechanics of particles tend to focus on individual

Foucault’s Pendulum in the Pantheon in Paris

Modern dynamics, like classical dynamics, is concerned with trajectories through

uniquely determined by the current state. This is called deterministic dynamics,

1.1 State space and dynamical flows

1.1.1 State space

Example 1.1 State space of the damped one-dimensional harmonic oscillator

x(t) = Xeiωt (1.4)

−mω2 Xeiωt + iωγ Xeiωt + kXeiωt = 0 (1.5)

Example 1.1 continued

The final solution is ⎡ ⎤

1.1.2 Dynamical flows

or, more succinctly,

which is a system of N simultaneous equations, where the vector function Fa is

Example 1.2 An autonomous oscillator

Example 1.2 continued

Example 1.3 Undamped point-mass pendulum

Example 1.3 continued

Example 1.4 A three-variable harmonic oscillator

As an example of an odd-dimensional state space, consider the three-dimensional flow

1.2 Coordinate representation of dynamical

1.2.1 Coordinate notation and configuration space

1.2.2 Trajectories in 3D configuration space

xa = xa (t) or xa = xa (s) (1.29)

where T a is a unit tangent vector in the direction of the velocity:

where R is the radius of curvature at the specified point on the trajectory.

possible to define a path length interval ds2

Example 1.5 Parabolic trajectory in a gravitational field

This is a familiar problem that goes back to freshman physics. However, it

Example 1.5 continued

1.2.3 Generalized coordinates

Example 1.6 Bead sliding on a frictionless helical wire

1.3 Coordinate transformations

These equations can be inverted to yield

1.3.1 Jacobian matrix

∂q1 ∂q2 ∂q3 ∂x ∂y ∂z

The determinant |J| is called the Jacobian.

x1 = Λ11 q1 + Λ12 q2 + Λ13 q3

which is recognizable in matrix multiplication as

with the Einstein repeated-index summation. The Einstein summation convention

1.3.2 Metric spaces and basis vectors

which is written in matrix form as

d x = eqa dqa (1.60)

where the Einstein summation is implied. Vectors describe the properties of

eb = Rab ea (1.76)

wedge product A ∧ B of Hermann Grass-