Electromagnetic Radiation Richard Freeman Full Chapter

Electromagnetic radiation Richard
Freeman
Visit to download the full and correct content document:
https://ebookmass.com/product/electromagnetic-radiation-richard-freeman/
ELECTROMAGNETIC RADIATION
Electromagnetic Radiation
Richard Freeman, James King,

and Gregory Lafyatis
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
© Richard Freeman, James King, Gregory Lafyatis 2019
The moral rights of the authors have been asserted
First 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: 2018953425
ISBN 978–0–19–872650–0
DOI: 10.1093/oso/9780198726500.001.0001
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.
Contents
Part I Introductory Foundations

1 Essentials of Electricity and Magnetism 3
1.1 Maxwell’s static equations in vacuum 3
1.1.1 Electrostatic equations 4
1.1.2 Magnetostatic equations 5
1.1.3 Lorentz force 6
1.2 Maxwell’s static equations in matter 6
1.2.1 Response of material to fields 7
1.2.2 Bound charges and currents 9
1.2.3 Macroscopic fields 10
1.2.4 Polarizability and Susceptibility 11
1.2.5 The canonical constitutive relations 13
1.2.6 Electric fields and free charges in materials 13
1.3 Energy of static charge and current configurations 14
1.3.1 Electrostatic field energy 14
1.3.2 Magnetic field energy 16
1.4 Maxwell’s dynamic equations in vacuum 18
1.4.1 Faraday’s contribution 19
1.4.2 Conservation of charge and the continuity equation 20
1.4.3 Maxwell’s contribution 21
1.5 Maxwell’s dynamic equations in matter 22
1.5.1 Origin of material currents 22
1.6 Plane wave propagation in vacuum 24
1.6.1 Polarization of plane waves 26
1.7 E&M propagation within simple media 29
1.8 Electromagnetic conservation laws 30
1.8.1 Energy density 30
1.8.2 Poynting’s Theorem 31
1.8.3 Linear momentum density 31
1.8.4 Maxwell stress tensor 33
1.9 Radiation in vacuum 34
1.9.1 Field amplitude as a function of distance from the source 35
1.9.2 Decoupling of radiation fields from the source 35
1.9.3 Illustration of coupled and decoupled fields from an accelerated charge 36
Exercises 38
1.10 Discussions 39
vi Contents
2 The Potentials 43
2.1 The magnetic and electric fields in terms of potentials 43
2.2 Gauge considerations 44
2.3 The wave equations prescribing the potentials using the Lorenz gauge 45
2.4 Retarded time 46
2.4.1 Potentials with retarded time 48
2.5 Moments of the retarded potential 49
2.5.1 Potential zones 49
2.5.2 General expansion of the retarded potential 51
Exercises 55
2.6 Discussions 55
Part II Origins of Radiation Fields

3 General Relations between Fields and Sources 63
3.1 Relating retarded potentials to observable fields 63
3.1.1 Spatial derivatives of retarded potentials 65
3.2 Jefimenko’s equations from the retarded potentials 67
3.3 Graphical representation of transverse fields arising from acceleration 69
3.4 Jefimenko’s equations without regard to retarded potentials:
Green Functions 71
3.4.1 Field characteristics 74
3.4.2 Example: fields directly from Jefimenko’s equations 75
Exercises 79
3.5 Discussions 80
4 Fields in Terms of the Multipole Moments of the Source 85
4.1 Multipole radiation using Jefimenko’s equations 85
4.1.1 Approximate spatial dependence 85
4.1.2 Radiation from zeroth order moments 87
4.1.3 Radiation from first order moments 89
4.2 Multipole radiation from the scalar expansion of the vector potential 91
4.2.1 Fields from an electric dipole moment 92
4.2.2 Fields from magnetic dipole moment 94
4.2.3 Fields from electric quadrupole moment 98
4.3 Power radiated in terms of multipole moments of the source 99
4.3.1 Power radiated by electric dipole moment 99
4.3.2 Power radiated by magnetic dipole moment 100
4.3.3 Power radiated by electric quadrupole moment 101
Exercises 103
4.4 Discussions 106
Contents vii
Part III Electromagnetism and Special Relativity

5 Introduction to Special Relativity 113
5.1 Historical introduction–1666 to 1905 115
5.1.1 The nature of space and time 115
5.1.2 The nature of light 117
5.1.3 Michelson–Morley experiments 121
5.2 Einstein and the Lorentz transformation 121
5.2.1 Einstein’s approach 122
5.2.2 The Lorentz transformation: covariance among inertial frames 125
5.3 The invariant interval and the geometry of space-time 130
5.3.1 Minkowski space-time diagrams 131
5.3.2 Physical consequences of special relativity 135
5.4 Vector space concepts 139
5.4.1 Contravariant and covariant vectors 141
5.4.2 The metric tensor 148
5.4.3 Generation of other 4-vectors and 4-tensors 149
5.5 Some important general 4-vectors 150
5.5.1 The 4-gradient operator 151
5.5.2 The 4-vector velocity 153
5.5.3 The 4-vector momentum 154
5.5.4 The 4-vector force 157
5.6 Some important “E&M” 4-vectors 159
5.6.1 The 4-wavevector 159
5.6.2 The 4-current density 161
5.6.3 The 4-potential (in Lorenz Gauge) 161
5.7 Other covariant and invariant quantities 162
5.7.1 The angular momentum 4-tensor 162
5.7.2 Space-time volume 163
5.7.3 Space-time delta function 164
5.8 Summary of 4-vector results 164
5.9 Maxwell’s equations and special relativity 165
5.9.1 Manifest covariance of Maxwell’s equations 165
5.9.2 The electromagnetic field tensor 166
5.9.3 Simple field transformation examples 169
5.10 The Einstein stress-energy tensor 173
Exercises 175
5.11 Discussions 176
6 Radiation from Charges Moving at Relativistic Velocities 184
6.1 Lienard–Wiechert potentials 185
6.1.1 Derivation by integral transform 187
6.1.2 Derivation by geometric construction 188
viii Contents
6.2 Radiation fields from a single charge undergoing acceleration 190

6.2.1 Moving charge general field characteristics 195
6.3 Power radiated from an accelerated charge 196
6.3.1 Low velocities and classical Larmor’s formula 197
6.3.2 Radiated power for relativistic particles 198
6.4 Acceleration parallel and perpendicular to velocity 200
6.4.1 Angular distribution for acceleration to velocity 200
6.4.2 Angular distribution for acceleration ⊥ to velocity 202
6.4.3 Total radiated power for acceleration and ⊥ to velocity 203
6.5 Spectral distribution of radiation from an accelerated charge 205
6.6 Synchrotron radiation 209
6.7 Fields from a single charge moving with constant velocity 214
6.7.1 Parametrization of the fields 218
6.7.2 Spectral energy density of the fields 220
6.7.3 Number of photons associated with fields of a passing charge 222
6.8 Bremsstrahlung 223
Exercises 227
6.9 Discussions 228
7 Relativistic Electrodynamics 229
7.1 Dynamics using action principles: Lagrangian and Hamiltonian mechanics 229
7.1.1 Concept of action 230
7.2 Relativistic mechanics of single point-like particles 234
7.2.1 The relativistic mechanics of a free particle 234
7.2.2 Free particle canonical 4-momentum 236
7.2.3 Free particle angular momentum 4-tensor 237
7.2.4 A charged particle in an external electromagnetic field 239
7.3 The action principle description of the electromagnetic field 243
7.3.1 Equations of motion 245
7.3.2 Lagrangian density function 247
7.3.3 Recovery of Maxwell’s equations 249
7.3.4 Gauge invariance 250
7.3.5 The Proca Lagrangian 252
7.4 The Hamiltonian density and canonical stress-energy tensor 254
7.4.1 From the Maxwell stress tensor to the 4D stress-energy tensor 254
7.4.2 Hamiltonian density: the “00” canonical stress-energy tensor component 255
7.4.3 Canonical stress-energy tensor and conservation laws 256
7.4.4 Canonical electromagnetic stress-energy tensor 257
7.4.5 Symmetric electromagnetic stress-energy tensor 258
7.4.6 Angular momentum density of fields 259
7.4.7 Electromagnetic stress-energy tensor including source terms 261
Exercises 261
7.5 Discussions 263
Contents ix
8 Field Reactions to Moving Charges 267

8.1 Electromagnetic field masses 268
8.2 Field reaction as a self-force 269
8.2.1 Lorentz calculation of the self-force 270
8.2.2 Some qualitative arguments for the self-force 274
8.3 Abraham–Lorentz formula and the equations of motion 276
8.3.1 The equations of motion 278
8.3.2 Landau–Lifshitz approximation 282
8.3.3 Characteristic time 283
8.4 The 4/3 problem, instability, and relativity 284
8.5 Infinite mass of the Abraham–Lorentz model 291
Exercises 294
8.6 Discussions 296
Part IV Radiation in Materials

9 Properties of Electromagnetic Radiation in Materials 303
9.1 Polarization, magnetization, and current density 304
9.2 A practical convention for material response 306
9.4 Frequency dependence 310
9.4.1 ω → ∞ 310
9.4.2 ω → 0 312
9.4.3 Plane waves versus diffusion 313
9.4.4 Transient response in a conductor 316
9.4.5 Temporal wave-packet 317
9.4.6 Group velocity versus phase velocity 319
9.4.7 Pulse broadening 320
9.5 Plane waves at interfaces 322
9.5.1 Boundaries 322
9.5.2 Fresnel transmission and reflection amplitude coefficients 325
9.5.3 Total internal reflection 328
9.5.4 Fresnel transmission and reflection intensity coefficients 332
9.5.5 Fresnel transmission and reflection: vacuum/material interface 333
9.6 Some practical applications 335
9.6.1 The two-surface problem 335
9.6.2 Lossy dielectrics and metals 338
9.7 Frequency and time domain polarization response to the fields 339
9.7.1 Example 342
9.8 Kramers–Kronig relationships 343
9.9 Measuring the response of matter to fields 346
9.9.1 Measuring the optical constants of a material 347
x Contents
9.9.2 Single frequency measurements 348

9.9.3 Spectral measurements 355
Exercises 358
10 Models of Electromagnetic Response of Materials 366
10.1 Classical models of Drude and Lorentz 366
10.1.1 The Drude model of free electrons 368
10.1.2 The lorentz model of bound electrons 370
10.1.3 The combined model: Lorentz–Drude 374
10.1.4 Lorentz and Drude model response functions 375
10.2 Lorentz insulators 376
10.2.1 Multiple binding frequencies 383
10.3 Drude metals and plasmas 384
10.4 Measuring the Lorentz–drude response of matter to fields 388
10.4.1 Single frequency measurements 388
10.4.2 Dual polarization Fresnel reflectivity measurement 389
10.4.3 Broadband measurements and response models 391
Exercises 394
11 Scattering of Electromagnetic Radiation in Materials 398
11.1 Scattering 398
11.2 Scattering by dielectric small particles 402
11.2.1 Scattering by a free electron: Thomson scattering 404
11.2.2 Scattering by a harmonically bound electron 405
11.2.3 Scattering near resonance 407
11.2.4 Plasmon resonance 408
11.3 Integral equations, the Born approximation and optical theorem 411
11.3.1 Scalar theory 413
11.3.2 Vector theory 417
11.4 Partial wave analysis 423
11.4.1 Scalar theory 424
11.4.2 Vector partial wave analysis 429
11.4.3 Solution of scattering from a homogeneous sphere: Mie scattering 438
11.5 Some results 443
11.5.1 The long wavelength limit 443
11.5.2 Scattering off dielectric spheres: water droplets 445
Exercises 455
12 Diffraction and the Propagation of Light 467
12.1 Diffraction 467
12.2 Geometric optics and the eikonal equation 470
12.3 Kirchhoff’s diffraction theory 471
Contents xi
12.3.1 Kirchhoff’s integral theorem 471

12.3.2 Kirchhoff’s diffraction theory: boundary conditions 473
12.3.3 Alternate boundary conditions: Rayleigh–Sommerfeld diffraction 475
12.3.4 Babinet’s principle 479
12.3.5 Fresnel approximation 481
12.3.6 Fraunhofer (far-field) diffraction 483
12.3.7 Fresnel diffraction of rectangular slit: the near-field 485
12.4 The angular spectrum representation 488
12.4.1 Gaussian beams 491
12.4.2 Fourier optics (far-field) 495
12.4.3 Tight focusing of fields 496
12.4.4 Diffraction limits on microscopy 506
Exercises 514
13 Radiation Fields in Constrained Environments 523
13.1 Constrained environments 523
13.2 Mode counting: the density of electromagnetic modes in space 527
13.3 Thermal radiation 530
13.4 Casimir forces 532
13.5 Spontaneous emission: the Einstein A and B coefficients 537
13.6 Microwave cavities 540
13.7 Microwave waveguides 543
13.7.1 General features of waveguides 543
13.7.2 Rectangular conducting waveguides 545
13.7.3 Transmission lines and coaxial cables: TEM modes 546
13.8 One-dimensional optical waveguides: the ray optic picture 548
13.8.1 The three-layer planar waveguide: the wave solutions of Maxwell’s equations 553
13.8.2 Fiber optics: the step-index circular waveguide 557
13.8.3 Higher order modes, single mode fibers, and dispersion 562
13.9 Photonic crystals 565
Exercises 571
A Vector Multipole Expansion of the Fields 583
A.1 Vector spherical harmonics 583
A.1.1 VSH expansion of general radiation fields 584
A.2 Multipole expansion of electromagnetic radiation 584
A.2.1 Non-homogeneous field wave equations 584
A.2.2 VSH expansion of the field wave equations 585
A.2.3 Parity considerations 587
A.2.4 Multipole expansion in a source-free region 588
A.3 Multipole radiation: energy and angular momentum 589
A.3.1 Energy density and the Poynting vector 589
xii Contents
A.3.2 Momentum density and angular momentum density 591

A.4 Multipole fields from vector harmonic expansion 594
A.4.1 Multipole expansion including sources 594
A.4.2 The small source approximation: near and far zones 598
References 613
Index 617
Part I
Introductory Foundations
Essentials of Electricity
and Magnetism 1
1.1 Maxwell’s static
equations in vacuum 3
• Review of Maxwell’s steady-state equations in vacuum 1.2 Maxwell’s static

equations in matter 6
• Modifications of Maxwell’s steady-state equations in the pres-
1.3 Energy of static
ence of matter: electric and magnetic polarization charge and
• Generalization of Maxwell’s equations in the presence of time current configurations 14
varying sources leading to a causal unification of fields in the 1.4 Maxwell’s dynamic
form of additional sources equations
in vacuum 18
• Origin of electromagnetic radiation directly from time-
1.5 Maxwell’s dynamic
dependent Maxwell’s equations and the response of materials equations
to electromagnetic radiation in matter 22
• Electromagnetic conservation laws, including electromag- 1.6 Plane wave propagation

in vacuum 24
netic energy, momentum and angular momentum
1.7 E&M propagation
within simple media 29
1.8 Electromagnetic
conservation laws 30
1.1 Maxwell’s static equations 1.9 Radiation in vacuum 34
in vacuum 1.10 Discussions 39
Maxwell’s equations are the foundational equations of classical elec-

tromagnetic phenomena. They are comprised of four 1st order linear
partial differential equations and are essentially statements that define
the electric and magnetic vector fields (e.g., specify their divergence
and curl) in terms of specific boundary conditions and electric charge
and current distributions. The mathematical origins of Maxwell’s
equations can be found in the basic inverse square laws of electro-
statics and magnetostatics, which were mainly formulated in the late
eighteenth through early nineteenth centuries, but far from exclusively,
through the observations and work of Coulomb, Ampere, Biot, and
Savart. When coupled with Faraday’s concept of a field and the
general mathematical theorems of Gauss, Laplace, and Poisson, we
begin to see the formal modern description of electric and magnetic
phenomena–at least for steady-state conditions in vacuum.
Electromagnetic Radiation. Richard Freeman, James King, Gregory Lafyatis,

Oxford University Press (2019). © Richard Freeman, James King, Gregory Lafyatis.
DOI: 10.1093/oso/9780198726500.001.0001
4 1 Essentials of Electricity and Magnetism
1.1.1 Electrostatic equations

The integral form of the law of electrostatics or Coulomb’s law is:

1 ρ ( r o − r)
r ) (
E (
ro) = dV (1.1)
4π εo r o − r|3
|
where the position vectors r and ro refer to the source and observer
locations, respectively, and ρ is the (static) charge density. We note
from this equation a number of important features: first, the static
electric field falls off as the inverse square of the distance to the
observer and is proportional to the charge density; second, a contri-
bution to the field at ro due to an element of charge ρ (
r ) dV will point
along R = ro − r, the direction from source to observer, with a polarity
dependent on the charge sign; and third, the field E ( r o ) is a linear
vector superposition of contributions from charge elements integrated
over all space, independent of time. If we now look at the divergence
taken with respect to ro of this field,

1 ρ (
ro)
∇ · E (
ro) = ρ (
r ) δ (
r o − r) dV = (1.2)
εo εo
where we have used ∇ · (R̂/R2 ) = 4π δ(R). This is the differential form

of what is known as Gauss’ law and is equivalent to Coulomb’s law.
It is the first of Maxwell’s equations. In this form, we see that any
divergence in the field is local to and proportional to the charge density.
In its integral form, which can be directly obtained from Eq. 1.2 using
the divergence theorem, it states that the integral of the E field over
an arbitrary closed surface is equal to the charge enclosed within that
surface divided by εo . Noting that R/R 3 = −∇(1/R), Eq. 1.1 can be
rewritten as a gradient

−1 ρ (r)
E (
ro) = ∇ dV = −∇φ (
ro) (1.3)
4π εo |
r o − r|
where φ (r o ) is the scalar potential. Because in general the curl of a

gradient vanishes, it follows from Eq. 1.3 that
∇ × E (
ro) = 0 (1.4)
which is the second of the two electrostatic Maxwell’s equations and

states that electrostatic fields are irrotational (curlless) everywhere.
The integral form of Eq. 1.4, which can be directly obtained using
Stokes’s theorem, states that the line integral of the E field around an
arbitrary closed curve is zero, thus confirming its status as a gradient.
1.1 Maxwell’s static equations in vacuum 5
Because the curl of a gradient is always zero, Eqs. 1.2 and 1.4 can
be more compactly expressed in terms of φ as ∇ 2 φ = ρ/εo (Poisson’s
equation) or ∇ 2 φ = 0 (Laplace’s equation) in charge free regions.
1.1.2 Magnetostatic equations

The integral form of the law of magnetostatics or the Biot–Savart
law is:

μo r ) × (
J ( r o − r)
B (
ro) = dV (1.5)
4π r o − r|3
|
where, as before, the distance and direction from a source element to

the observation point is represented by R = ro − r but the source is
now a distribution of steady-state current density elements, J ( r ) dV ,
each contributing to B ( r o ) an amount proportional to J (
r ), in the
direction, J (
r ) × (
r o − r), given by the right-hand rule. Also, as with
the electric field, the static magnetic field (due to each current element)
falls off as the inverse square of the distance. Following, analogously,
the electrostatic development of Section 1.1.1 to obtain a differential
form, we consider the curl of B

μo ro − r
∇ × B (
ro) = ∇ × J (
r)× dV (1.6)
4π r o − r|3
|
which, with some manipulation (see Discussion 1.1), can be written

∇ × B (
r o ) = μo J ( r o − r) dV = μo J (
r ) δ ( r) (1.7)
where, again, we have used ∇ · (R̂/R2 ) = 4π δ(R). This is the differen-

tial form of what is known as Ampere’s law and is equivalent to the
Biot–Savart law. It is the third of Maxwell’s equations. The integral
form of Eq. 1.7, which can be directly obtained using Stokes’s theorem,
states that the line integral of the B field around an arbitrary closed
curve is equal to the current enclosed by that curve multiplied by μo . 1 Expand the divergence of the integrand
To obtain the fourth differential form of Maxwell’s equations under of Eq. 1.6
steady-state conditions in the absence of matter, we take the divergence
ro − r ro − r
∇· J (
r )× = · ∇×J (
r)
of Eq. 1.5 to obtain1 r o − r|3
| r o − r|3
|

ro − r
−J (r)· ∇ ×
∇ · B (
ro) = 0 (1.8) r o − r|3
|
where the first term vanishes because the J

which is the second of the two magnetostatic Maxwell’s equations is not a function of the observer coordinates.

and states that magnetostatic fields are solenoidal (divergenceless) For the second term, we again note that RR3 =
everywhere. The integral form of Eq. 1.8, which can be directly −∇ R1 and the curl of a gradient vanishes.
obtained using the divergence theorem, states that the surface integral
of the normal component of B field around an arbitrary closed surface
is zero. Because B has no divergence value, it can be written as the curl
This vector field, known as the “vector potential”,
of another field, A.
is analogous to the scalar potential, φ, encountered in electrostatics.
So, continuing in close analogy with electrostatics, we are tempted to
write Eqs. 1.7 and 1.8 in terms of a single second order differential
equation of the potential such as the Poisson or Laplace equations.
Thus, we note that much like writing E as −∇φ automatically satisfies
∇ × E = 0 and turns ∇ · E = ρ/εo into the Poisson equation, writing B
as ∇ × A automatically
satisfies
∇ · B = 0 and turns ∇ × B = μo J into
∇ × ∇ × A = ∇ ∇ · A − ∇ 2 A = μo J. This is a more compact way
of expressing Eqs. 1.7 and 1.8. In summary, Maxwell’s equations for
steady state and in the absence of matter are:
∇ · E = ρ/εo (1.9)
∇ × E = 0 (1.10)
∇ × B = μo J (1.11)
∇ · B = 0 (1.12)
1.1.3 Lorentz force

The effects of magnetic and electric fields on a charge q were given
their modern form by Lorentz in 1892, building on the work of Thom-
son’s (1881)2 and Heaviside’s (1889)3 extrapolations of Maxwell’s
exposition of his equations (1865):
F = q[E + v × B]
(1.13)
This description of the total force on a charge in the presence of

external fields E and B has been so well verified experimentally, even
2 Thomson, J. J. On the electric and
for charge velocities approaching the speed of light, that it is used as
an empirical definition of E and B at any space-time point when q, F,
v
magnetic effects produced by the motion
of electrified bodies. Philosophical Maga-
are known. 4
zine, 11, 229–249, https://doi.org/10.1080/
14786448108627008 (1881).
3 Heaviside, Oliver. On the electromag-
netic effects due to the motion of electri-

fication through a dielectric. Philosophical 1.2 Maxwell’s static equations in matter
Magazine 324 (April 1889).
4 The relativistic formulation of Eq. 1.13
is the same, with the proviso that the force is Within matter, where there are charges that respond to external fields
related to the velocity by by moving freely, or charges bound to other charged objects that
d orient or displace in response to external fields, Maxwell’s equations
F = (γmv)

dt become exceedingly difficult to solve exactly. It is useful then, when
working with fields in matter, to divide the problem conceptually into

microscopic and macroscopic fields with the microscopic fields, in a
sense, being the true yet practically intractable fields in all their grainy
detail, while the macroscopic fields are spatial and temporal averages
of the micro-fields over regions and times that are microscopically
large yet macroscopically small. In this subsection it will be shown that
the response of matter to applied fields generally results in so-called
“bound” sources of charge and current density and for materials with
a component of free electrons, an additional source of “free” current.
While this will modify the two inhomogenous Maxwell’s equations, it
will, in the steady-state case, leave unaffected the two homogeneous
equations. As a consequence of this, we can immediately see that
the macroscopic versions of the two homogeneous equations will be
identical to the microscopic versions. That is,
∇ × E = 0 and ∇ · B = 0 (1.14)
1.2.1 Response of material to fields

Polarization, either electric P (
r ) or magnetic M (
r ), is defined macro-
scopically as dipole moment per unit volume and its existence within
a material is a result of the local alignment of atomic or molecular
electric p or magnetic m dipole moments within a macroscopically
small but microscopically large volume about the evaluation point, r.
This alignment can be permanently frozen into the material as in the
case of ferromagnets and the less often encountered electric analogs
known as electrets. Alignment of dipoles resulting in polarization is,
however, more commonly a response to the presence of electric and
magnetic fields. Two basic types of dipole response have been found:
Either pre-existing dipole moments are rotated into alignment by the
fields or dipole moments are induced by the applied fields within
the material. A well known example of the first type of response to
electric fields occurs within water because the positive and negative
charge centers of the “polar” H2 O molecule are intrinsically separate.
Similarly, the pre-existing magnetic atomic dipoles (due to unpaired
electrons) within paramagnetic materials will align with an applied
magnetic field. The second type of response in which dipole moments
are induced occurs in all materials but is most noticeable within “non-
polar” materials devoid of pre-existing dipoles. A classical picture of
such a material response to an electric field is that of neutral atoms with
initially overlapping positive (nuclear) and negative (electron) charge
centers that, upon application of the field, get stretched in opposite
directions to the mechanical limits of their bonds, thus forming electric
dipole moments. The induction of magnetic dipole moments by a
magnetic field is known as diamagnetism. In this case, there is no

stretching but rather currents within atomic or molecular structures
are induced via Faradays law (Section 1.4.1) resulting in an anti-
alignment of the dipoles to the field. Macroscopic polarization in terms
of the microscopic electric dipole moments is given as:
r ) = N (
P( r ) < p(
r) > (1.15)
where < p(r ) > is the average of all the electric dipole moments in
a macroscopically small but microscopically large volume centered
at the location r, and N (
r ) is the number of such objects per unit
volume. In the case of magnetically active materials, whether there is an
orientation of magnetic objects, or induced currents centered around
atoms or molecules, the corresponding expression is the generation of
a macroscopic magnetic polarization or “magnetization”:
r ) = N (
M( r) < m
(
r) > (1.16)
where in the same way < m (

r ) > is the average magnetic dipole
centered at the location r and N ( r ) is the number of objects per unit
volume at that location.
The most general instance of material response to a field is not
linear, isotropic or homogeneous and therefore requires a non-linear,
spatially dependent tensor for its mathematical description. In the
present case, we will initially assume a simpler material that is linear
and isotropic but not necessarily homogeneous. In this case, for
example, the average dipole moment and the polarizing electric field
Ep within the material are related by a coefficient, α, known as the
polarizability:
< p( r ) Ep

r ) > = εo α ( (1.17)
where εo is generally included for later convenience. Combining

Eq. 1.17 with Eq. 1.15 gives us an expression for the polarization in
terms of the polarizing field:
r ) = εo N (
P( r ) Ep
r ) α ( (1.18)
Now, it is an easily overlooked but important point that the macro-

scopic applied field amplitude, E, within the material is not necessarily
the average amplitude of the polarizing field, Ep , felt by the atoms
and molecules in the matter. Indeed, both field amplitudes are average
values, however, while the applied field results from macroscopically
averaged surface and volume charge densities external to the material,
the polarizing field refers specifically to volumes local to the atoms and
molecules and so additionally takes into account the fields of all nearby
dipoles. The source of these additional local fields are represented
in the form of bound charges that, along with bound currents, are
discussed next.
1.2.2 Bound charges and currents

The presence of polarization or magnetization in dielectric and mag-
netically active materials is characterized by the existence of “bound”
charges and currents. For the case of polarized material, this can be
shown by considering the potential at a point ro due to all the
electric dipole moments within a non-uniformly polarized material in
a volume, V

1 R̂ · P (
r)
(ro ) = dV (1.19)
4π εo R 2
V
where P (
r ) dV = d p is a macroscopically small but microscopically
large element of dipole moment within the material and as usual, r is
a source point and R = ro − r is the vector pointing from the source to
the observation point. Through the use of integration by parts and the
divergence theorem, this equation can be re-expressed as5

1 P (
r ) · n̂ 1 −∇ · P (
r)
(ro ) = da + dV (1.20)
4π εo R 4π εo R
S V
where S is the surface bounding the material and n̂ is the outward

surface normal at points of integral evaluation. The numerators times
their respective differentials have the form and units of elements of
charge so that we equate them to surface charge and volume charge
densities. That is,
P · n̂ = σb and − ∇ · P = ρb
Physically, this is not hard to visualize. For example, to see the 1
5Noting RR̂2 = ∇ R1 and P · ∇ =
latter equivalence, consider a small, yet still macroscopic volume R
P
∇· R − ∇·RP , we have
within the non-uniformly polarized material and in this region let
there be generally positive divergence. Consider just the x direction.
1 P (
r)
(ro ) = ∇· dV
Because a larger polarization means more positive charge displaced 4π εo R
V
in the positive x direction and more negative charge displaced in
1 −∇ · P (
r)
the negative x direction, a positive divergence means that more pos- + dV
4π εo R
itive charges are pushed out of the right bounding surface than V
negative charges are pushed out of the left boundary, thus leaving a and then use the divergence theorem on the
net negative charge density. first term.
By similar considerations of the vector potential at a point ro due to

all the magnetic dipole moments within a non-uniformly magnetized
material it can be shown that
(
M (
r ) × n̂ = K b and ∇ × M r ) = Jb
which is to say that the curl of the magnetization ∇ × M (

r ) in a
magnetizable material can be identified with a real, yet bound, current
density, Jb , within the volume and M (
r ) × n̂ can be identified as a

bound surface current density, K b , on the surface.6
The physical interpretation of this can be seen by imagining a
uniformly magnetized material in which each little magnetic dipole
has an associated current loop. Because of the uniformity, all the
neighboring dipole current loops cancel out. However, at surfaces not
perpendicular to M, or equivalently, where M (
r ) × n̂ = 0, there are
missing neighbors and so there is net bound current on the surface.
On the other hand, if the magnetization is not uniform in such a
way that ∇ × M (
r ) = 0 within the volume, then the magnitude of M

varies in a direction perpendicular to M so neighboring dipoles do not
completely cancel and again there is a net current in the direction of
∇ ×M (
r ).
1.2.3 Macroscopic fields

We have seen that the presence of polarization and magnetization
within matter is equivalent to a distribution of “bound” charge and
current densities as given by
ρb = −∇ · P (1.21)

Jb = ∇ × M (1.22)
In terms of the total (free plus bound) charge and current densities,
the two inhomogeneous Maxwell’s equations in a material possessing
both a polarization and a magnetization can now be written:
εo ∇ · E = ρf + ρb = ρf − ∇ · P
6 The vector potential, at a field point r ,
o (1.23)
resulting from a superposition of all the little
magnetic dipole moments within a volume is: 1 = Jf + Jb = Jf + ∇ × M

(∇ × B) (1.24)
(
μo
1 M r ) × R̂
A (
ro ) = dV
4π εo c2 R
v If we now combine divergence and curl terms to get
Using the identity R̂
=∇ 1
R , and integrating
∇ · εo E + P = ρf
R2
by parts, (1.25)

r )×n̂ ∇ ×M (
r) 1 = Jf
A (ro ) = μo M( μo
da + 4π dV ∇× B−M (1.26)
4π
s
R
v
R μo
we can then define two new macroscopic fields
D = εo E + P (1.27)
1
H = B−M (1.28)
μo
So that in terms of these new fields, which represent the fundamental

fields plus polarization and magnetization effects due to the
material, the steady-state macroscopic Maxwell’s equations then
simplify to:
= ρf
∇ ·D (1.29)
∇ × E = 0 (1.30)
∇ · B = 0 (1.31)
∇ × H = Jf (1.32)
It is important to emphasize that while both D and H have only

the free charges and currents as sources, both of these quantities are
just convenient constructs introduced to permit a compact method
of accounting for the response of the material to the fundamental
fields, E and B.
It is also important to keep in mind that while P and

M, like E and B, are macroscopically averaged vector fields within
matter (i.e., they are the same type of mathematical object), they differ
substantially in that E and B represent the fundamental fields in the
purest ethereal sense as envisioned by Faraday, while P and M are
essentially representations of charge and current distributions within
matter. From this perspective, we can see the conceptual difference
between, for example, the two equations ∇ · E = ρ/εo and −∇ · P = ρb :
We read the first equation as “a collection of charge (any charge)
acts as a source of the electric field” while the second equation reads
“a collection of charge (bound charge) results from a distortion of
charge distribution in a material” Implicit in these statements is that
in the first case the source somehow “causes” or at least accompanies
the field but the two things are not physically the same whereas in the
second case the charge distribution is equivalent to a distortion in P.

Finally, D and H , while often treated more like fields akin to E and B,
are composite vector fields that are part pure field and part material
response.
1.2.4 Polarizability and Susceptibility

Earlier, in our discussion of the response of matter to electric fields, we
obtained an expression (Eq. 1.18) for the polarization P in terms of
the polarizability α and the polarizing field, Ep . We further noted that
the polarizing field, Ep , felt by the atoms and molecules in the matter,
was not generally of the same amplitude as the macroscopic applied
within a material and this was said to be due to the
field amplitude, E,
specific accounting, by Ep , of fields from other nearby dipoles. As can
be imagined, this difference is density dependent and in fact, for the
case of gases, the material is tenuous enough that we can approximate
However, this is not the case for denser liquids and solids
Ep E.
and we would therefore like to find the relation between these two
field amplitudes so we can then write an expression relating the
macroscopic applied field amplitude, E, to the polarization P in terms
of the microscopic polarizability, α. This constant of proportionality
is known as the (DC) electric susceptibility, χe = χe (α), and can be
seen as the macroscopic equivalent to the microscopic polarizability,
α. The electric susceptibility is thus defined by
r ) = εo χe E
P( (1.33)
where, for example, in the case of gases, this connection is trivial:

χe = N ( r ) α (
r ). An analogous relation exists that expresses the mag-
netization response, M, of a magnetic material to the macroscopic
field, H . The magnetic susceptibility, χm , is thus similarly defined as,
r ) = χm H
M( (1.34)
For the electric field case, to see how the difference between the
applied and polarizing fields (E and Ep ) comes about, we divide our
treatment of the material into two regions: (a) a macroscopically small
but microscopically large spherical cavity, centered on the point in
question, in which we must account for the specific charge config-
urations of the surrounding atoms and molecules and (b) all the rest
of the material outside the cavity that we can safely treat as smooth
and macroscopically averaged. Let us express the relation between the
two average fields as Ep = E + E. We first note that if region (a) were
to be treated like (b), smooth and macroscopically averaged, then we
would essentially be eliminating any reference to specific fields, which
is required for the evaluation of Ep , and our result would yield E = 0.
So, E is what we get when we replace the field resulting from a
smooth and macroscopically averaged treatment of region (a) with a
more detailed and accurate treatment of the region. Two results are
important: (a) It is a well-known result from electrostatics in matter7
that the electric field within a dielectric sphere of uniform polarization
P is also uniform and given by E = −P/3ε o and (b) it can be shown
that in material lattices of sufficient symmetry, the total contribution
7 Jackson, J. D., Classical Electrodynamics, to the electric field at a given lattice point due to atoms at all nearby
3rd edition, Wiley, New York (1999). lattice points (i.e., within the small cavity region in our problem)
goes to zero.8 Now, with the additional assumptions that within the
small cavity region (a) the polarization is constant and the material is
sufficiently symmetric, these results can be used to write E = P/3ε
o.
Then letting
Ep = E + P/3ε
o (1.35)
the substitution of this result into Eq. 1.18, with some rearranging,
yields the polarization in terms of the applied field:
r ) = εo Nα ( r) r ) = εo χe ( r)
P( E( r ) E( (1.36)
1 − 3 N (
1
r ) α (
r)
which, as mentioned, gives the macroscopic susceptibility (χe ) in terms

of the microscopic polarizabilty (α) for dense materials. This is known
as the Clausius–Mossotti equation.
1.2.5 The canonical constitutive relations

Note that Eqs. 1.27 and 1.28 make no assumption about whether
the polarization or magnetization is frozen into the material or is, for
example, a linear response to an applied field. If we consider the latter
case, then following from P = εo χe E of the previous section along with
the analogous result of M = χm H for the macroscopic magnetization
response, Eqs. 1.27 and 1.28 lead to9
= εE
D (1.37)
B = μH (1.38)
r ) = εo (1 + χe ) and μ = μ (
in which ε = ε ( r ) = μo (1 + χm ) are the
permittivity and permeability, respectively, of the material (see Dis-
cussion 1.2).
1.2.6 Electric fields and free charges 8 Purcell, E.M. and Morin, D.J. Electricity
in materials and Magnetism, 3rd edition, Cambridge Uni-

versity Press, 2013.
9 The macroscopic derived fields are
If a material has free charges there is a further potential relation given by Eqs. 1.27 and 1.28 as
between the free current densities discussed before and the applied
= εo E + P and H = 1 B − M
D
electric field: μo
For linear responses (P = εo χe E and M =

Jf = σ E (1.39)
χm H ) to an applied field, these become
where σ is the conductivity of the material. The introduction of con- D = εo (1 + χe ) E = ε E
ductivity here is properly a “constitutive” relation, because the concept B = μo (1 + χm ) H = μH

is inherently macroscopic. The conductivity, as expressed in Eq. 1.39,

is a relationship that essentially says that in a macroscopic region of
the material under consideration, a free current is associated with an
applied electric field, and the magnitude of the current depends upon
macroscopic parameters of the material, in this case the resistivity ρ.
For materials with inherently large resistivity, the conductivity (σ =
1/ρ) is small enough that an applied field can exist in the material with
no excitation of a free current. On the other hand, if the resistivity is
extremely low (in a superconductor, e.g.,), there can be no equilibrium
applied field for then there would be extremely large current flow. This
idea is perhaps best understood by considering the free current to be
the flow of individual free charges. In a material with extremely low
resistivity, the free charges will move to cancel out the applied field;
that is, the current described by Eq. 1.39 will be extremely high until
the free electrons have arranged themselves to electrostatically cancel
the applied field.
1.3 Energy of static charge and

current configurations
1.3.1 Electrostatic field energy
The simplest starting point for the calculation of the field energy
arising from a static placement of charges is to consider a collection of
charges, qi . The electrostatic energy of the i th charge is given by
Eis = qi ϕis
where ϕis = nj ϕijs is the summed electrostatic potential from all the
other charges, qj , ( j = i), evaluated at the position of qi . The total
energy of n assembled charges (so-called “configuration energy”)
is then
n
1
Es = qi ϕis (1.40)
2
i
where the factor of 1/2 arises in this summation because we have

essentially counted the potential energy from each charge pair twice.
Also, we have been careful to specify that Eq. 1.40 is only the energy
of assembling the charges (relative to their being infinitely separated);
that is, the energy to assemble the individual charges themselves is not
included. To calculate the total electrostatic energy of a charge dis-

tribution, we proceed formally by restating Eq. 1.40 for a continuous
charge distribution

1
Ecs = ρ ϕcs dV (1.41)
2
If we note that the charge density is given at each location by ρ =

εo ∇ · E and that the electric field within the volume is given at each
location by E = −∇ϕcs , then with some manipulation,10 Eq. 1.40 can
be written

εo
Ecs = E 2 dV (1.42)
2
To make contact with the expression in Eq. 1.40, now imagine the
continuous charge distribution to be made up of a large collection
of individual charges and that at any location in the volume the total
electric field is given by11
E (
r) = Ej (
r) (1.43)
j 10 Express ρ in terms of E in Eq. 1.41

then use the vector identity ∇ · E ϕcs =
where the vectors Ej (
r ) are the Coulomb fields of each individual
∇ϕcs + ∇ · ϕcs E and then ∇ϕcs = −E;
−E·
charge evaluated at the field point of interest.12 Then, Eq. 1.42 can use the divergence theorem and note that the
be written product of the electric field and the potential
tends to zero faster than 1/R2 .
11 This argument follows that of Panofsky
Ecs = Eos + EIs and Phillips (Panofsky, W. K. H. and Phillips,
M., Classical Electricity and Magnetism. 2nd
where the first term does not depend upon the relative position of the edition, Addison-Wesley, 1962.).
12 Here and in what follows, the number
charges under consideration and is given by13 : of individual charges is considered to be so
very large so that the idea of an approximate
εo continuous charge distribution is reasonable.
Eos = Ei2 (
r ) dV (1.44) 13 If the sum of the fields, Eq. 1.43, is
2
i inserted into Eq. 1.42, we have
⎛ ⎞

while the second one does, and is given by: εo
Ecs = Ei · ⎝ Ej ⎠ dV
2
i j
εo
EIs = r ) · Ej (
(1 − δi, j )Ei ( r ) dV (1.45) which can be separated into a sum in which
2
i, j i = j,
εo
Eos = Ej2 dV
2
which clearly vanishes for j = i. We can analyze EIs further by using j
the relation and a sum in which i = j,

εo
EIs = (1 − δij )Ei · Ej dV .
Ej = −∇ϕjs (
r) 2
i, j
where ϕjs (
r ) is the potential at the field point of interest due to the j th
charge. Then Eq. 1.45 can be written14

εo
EIs =− (1 − δij )(∇ · (E i · ϕj ) − ϕj ∇ · E i ) dV
s s
(1.46)
2
i, j
Now the volume is occupied by point charges, ρ( r ) = i qi δ(

r − ri ) so
that ∇ · Ei (
r ) = (qi /εo )δ(
r − ri ). Inserting this into Eq. 1.46 and using
the divergence theorem while noting that the product of the field times
the potential goes to zero faster than R−2 , yields

1
EIs = (1 − δij )qi r ) δ(
ϕjs ( r − ri )dV
2
i, j

If we note that ϕjs ( r ) δ(
r − ri )dV = ϕijs , then the portion of the
electrostatic energy that depends upon the arrangement of the finite
point charges expressed in terms of the fields in Eq. 1.42 is given by
n
1 1
EIs = (1 − δij )qi ϕijs = qi ϕis
2 2
i,j i
which is in agreement with the “configuration energy” we noted in

Eq. 1.40. The meaning of this result is that when the total electrostatic
energy of a collection of charges is calculated by the sum of the
electrostatic field energies, there is a term that does not depend on the
relative positions and one that does. The one that does is exactly equal
to the energy we would have calculated by assuming the energy was
contained in the charges as they are brought together. The portion of
the electrostatic energy calculated in Eq. 1.44 is evidently the energy
associated with creating the finite charges themselves. It is not possible
to determine in any meaningful manner whether the electrostatic
energy is “in the fields” or “inherent in the charges.”
1.3.2 Magnetic field energy

In this section, we will show that in the same way the electrostatic
energy of a system of charges can be represented as a volume integral
of the product of charge density and electric potential (Eq. 1.41),
the magnetic energy of a system of currents can be represented
by a volume integral of the scalar product of current density and
) = f ∇ ·
14 Use the vector identity ∇ · (f V the vector potential. In the case of currents, unlike for charges and
V + V · ∇f . electrostatic energy, the reversible work of assembling the system does
not include bringing the components in from infinity. Rather, the final
magnetic energy of the system can be obtained by starting at an initial
situation in which all the currents are zero and ramping up to the final
system values.
As we saw, for a system of charges the potential at the location of
charge i due to all the other charges is φis = nj φijs where ϕijs is linearly
proportional to the value of charge qj . Likewise, for a system of current
loops, the magnetic flux passing through the i th current loop due to
all the other loops is φis = nj φijs where φijs is linearly proportional to
current in the j th loop, Ij . If the j th current increases by dIj in time dt,
then the associated flux through the i th current loop increases and a
back emf, Vij = −dφijs /dt, is induced by Faraday’s law of induction. To
maintain the current during this time, the external source must then
provide an equal and opposite emf, −Vij . If an amount of charge dqi
has passed through the source during this time then the work done
by the external source for the i th current loop as a result of a current
change in the j th loop is,
dWij = −Vij dqi = Ii dφijs
And if the currents are now all increased to their final values and we
sum over j all the flux contributions to the i th current loop, we find
that the magnetic energy of the i th current loop is given by
n
Bis = Wij = Ii φis
j
where φis = nj φijs is the magnetic flux due to all the other currents, Ij ,
(j = i), passing through the current loop of Ii . The total energy of the
n fully energized current loops is then
n
1
Bis = Ii φis (1.47)
2
i
where, again, the factor of 1/2 arises in this summation because we

have essentially counted the energy from each current pair twice.
To calculate the total magnetic energy of a current distribution, we
proceed by rewriting the total flux in the i th loop, φis , in terms of the
r i ) = Ai . We know that
vector potential A(

φis = ∇ × Ai · d a = Ai · ds
S C
and thus Eq. 1.47 can be written,
n
1
Bis = Ai · Ii ds
2 C
i
so the closed line integral sums the contributions from all the current
elements of the i th loop and the sum over i then combines the
contributions from all current loops. Finally, if we re-express the
current element in terms of current density and a volume element,
, we can formally restate Eq. 1.47 for a continuous current
Ids ⇒ JdV
distribution,

1
Bcs = J · A dV (1.48)
2
which is the magnetic analogue to Eq. 1.41 for electrostatic energy.

In the forms of Eqs. 1.41 and 1.48, the configuration energy is
emphasized to reside at the location of the charges and currents. How-
ever, it is often useful to express electric and magnetic configuration
energies in forms that emphasize the fields themselves as carriers of
this energy. We have already obtained this form for electrostatic energy
(Eq. 1.42). We now obtain this “field” form for magnetic energy in an
analogous way.
Writing the current density at each location as J = 1/μo ∇ × B
and noting that the magnetic field within the volume is given at each
location by B = ∇ × A,
we obtain–in much the same way as for the
electrostatic case,15

1
Bcs = B2 dV
2μo
which is the magnetic energy counterpart to Eq. 1.42 for electrostatic

energy of continuous distributions.
1.4 Maxwell’s dynamic equations

15 Express J in terms
of B in Eq.1.48; use
in vacuum
the vector identity ∇ × B · A = ∇ × A ·

B − ∇ · A × B and ∇ × A = B; use the
To this point, we have limited our consideration to steady-state sources
divergence theorem and note that the cross
of charge and current. In the context of radiation, for which the sources
product of the magnetic induction and
the vector potential tends to zero faster are necessarily changing in time, this consideration is mostly periph-
than 1/R2 . eral but has been included to serve as a review. Thus, in a practical
Fig. 1.1 Faraday’s apparatus show-

ing electromagnetic induction from
G 1892 text book on Magnetism and
Electricity by A.W. Poyser. Battery
B
on the right powers the coil that is
inserted into or removed from the
standing coil. During the insertion or
removal, the galvanometer connected
to the standing coil responds. (By
J. Lambert [Public domain], via
Wikimedia Commons).
sense, this section and the next on the time-dependent Maxwell’s

equations is the starting point for our discussion of radiation.16
1.4.1 Faraday’s contribution

Historically, it was a series of experiments carried out and reported by
Faraday (1831) that provided the first quantitative results on the rela-
tion between time-varying electric and magnetic fields. Specifically, he
studied circuits placed in temporally and/or spatially varying magnetic
fields and noticed that a current was induced when the magnetic flux
enclosed by the circuit changed in time. He observed three distinct
cases:
(1) Change in flux due to a circuit moving through a spatially

varying magnetic field.
(2) Change in flux due to a spatially varying magnetic field moving
across a stationary circuit (see Fig. 1.1).
(3) Change in flux due to a spatially constant, temporally varying 16 Generally, in this book, when dis-
magnetic field within a stationary circuit. cussing the fields and sources in vacuum,
we will omit the subscripts, “b” and “f ”,
for bound and free, since all charges are
where, with the hindsight of special relativity and its consequences necessarily free and there is no need for a
for the transformation of E and B fields, we can immediately see the distinction.
equivalence of cases 1 and 2. However, prior to special relativity, case 1

current, which was understood to result from the J × B magnetic force,
was distinct from cases 2 and 3 current for which there was no moving
charge and therefore no magnetic force. So, it was the great insight of
Faraday to identify the shared experience of a changing flux within the
three cases and then to conclude that this changing flux induced an
electric field that acted on the charges to generate current in the wire–
a first step toward both the causal (classical) and actual (relativistic)
unification of the E and B fields. In differential form, Faraday’s law is
expressed (see Discussion 1.3):
∂ B
∇ × E = − (1.49)
∂t
which we see is the first time-dependent modification to the steady-

state vacuum Maxwell’s equations. Note that the induced electric field
is independent of the existence of a circuit and that the current in the
wire is circumstantial in comparison to Eq. 1.49.
1.4.2 Conservation of charge and the

continuity equation
Electric charge is locally conserved, and in a volume V where the
current inflow and outflow are not equal, this conservation of charge
requires there to be a build-up or or depletion of charge (assuming no
additional charge sources or sinks). More specifically, the total current,
I , flowing through the surface of the volume is exactly the negative of
the rate of change of the total charge Q within a volume,
dQ
I =− (1.50)
dt

Noting that I = J · d A and Q = ρdV , where A is the surface of vol-
ume V , the divergence theorem then immediately yields the continuity
equation for charge and current densities
∂ρ
∇ · J + =0 (1.51)
∂t
For example, an overall outflow of current in a region is represented

by a positive value for ∇ · J and this term is exactly balanced by a
corresponding depletion of charge within that region expressed as a
negative value for ∂ρ/∂t.
1.4.3 Maxwell’s contribution

For the final time-dependent modification to Maxwell’s equations,
the stage is set by considering an inconsistency that arises within
them when dealing with time-varying charge and current densities.
As we know from the mathematics of vector calculus, the divergence
of the curl of a vector field vanishes. In particular, within Maxwell’s
equations, the divergence of the right-hand side terms of the curl
equations must vanish. In electrostatics, this requirement is satisfied
trivially since ∇ × E = 0, and in magnetostatics the currents are steady
state by definition so that Ampere’s law is satisfied: ∇ · (∇ × B) =
μo ∇ · J = 0. Neither is there a problem with Faraday’s law (Eq. 1.49)

since ∇ · (∂ B/∂t) = 0. On the other hand, if we consider
= ∂/∂t(∇ · B)
the case of a time-varying (non-steady-state) current then there will
be regions in which the rates of inflow and outflow of charge are
not equal and therefore we will have a situation where ∇ · J = 0 and
thus an inconsistency within Maxwell’s equations: ∇ · (∇ × B) = 0 =
It was this inconsistency in Ampere’s law that was so brilliantly
μo ∇ · J.
resolved in 1865 by Maxwell and for this revolutionary work, which
essentially completed the classical theory of electrodynamics, the set
of equations were named in honor of Maxwell.
So, how did Maxwell resolve the inconsistency? It has to do with
local conservation of charge and the resulting continuity equation
(Eq. 1.51) as discussed previously. With the use of Gauss’ law or
Maxwell’s 1st equation, Eq. 1.9, the charge density can be expressed
as a divergence and Eq. 1.51 can be rewritten as,
∂ E
∇ · J + εo =0 (1.52)
∂t
And it was Maxwell’s realization that the zero divergence term in

parentheses, which properly reduces to ∇ · J = 0 for steady-state con-
ditions, is the correct term to be equated to the curl of B for the
general case of time-varying charges and currents. Ampere’s law is
thus modified to
∂ E
∇ × B = μo J + εo = μo J + μo Jd (1.53)
∂t
So now we have, for time dependent cases, an additional source for

the B field - the time rate of change of the E field. Maxwell named this
term the “displacement current”. However, it hardly resembles any
kind of traditional current, and, in fact, although it was derived for a

situation in which it represented a changing charge density necessary
to compensate for the divergence of an actual current, its existence
within the equation is more general than that. Indeed, it need not
accompany any actual current and it alone can act as a source of B.
It is this point that greatly adds to the symmetry of the Maxwell’s
equations–Faradays law says that a changing magnetic field induces an
electric field and now the addition of Maxwell’s displacement current
says that a changing electric field induces a magnetic field. The (now
complete) Maxwell’s equations are
ρ
∇ · E = (1.54)
εo
∂ B
∇ × E = − (1.55)
∂t
∇ · B = 0 (1.56)
1 ∂ E
∇ × B = μo J + (1.57)
c2 ∂t
√
where c = ( εo μo )−1 . Faraday’s and Maxwell’s contributions have
shown us that the fields can result not only from charge and current
densities but also from changing fields. The implications of this are
profound. We will see that this result led not only to the unification of
electric and magnetic phenomena but also to an explanation of light
and radiation as self-propagating waves of electric and magnetic fields.
Furthermore, as will be detailed in Chapter 5, it was the prediction of
an absolute speed of light in the face of the intuitive notion of absolute
space and time, which led to the development of special relativity.
These equations are universally valid both in vacuum and in matter
and–for reasons that will be explained in the next section–are known
as the “microscopic” Maxwell’s equations.
1.5 Maxwell’s dynamic equations

in matter
1.5.1 Origin of material currents
Within matter, the time-dependent Maxwell’s equations (Eqs. 1.54–
1.57) are microscopically correct but, as described in Section 1.2.2,
a macroscopic description is more useful. In that section, we saw
that polarization and magnetization (macroscopic concepts) within
matter are equivalent to distributions of “bound” charge and current
1.5 Maxwell’s dynamic equations in matter 23
densities as given by Eqs. 1.21 and 1.22. We then expressed the two
steady-state inhomogeneous Maxwell’s equations within a material
in terms of these “bound” contributions to the charge and current
densities (Eqs. 1.23 and 1.24). Now, in the time-dependent case, we
expect that the polarization and magnetization and thus the bound
charge and current densities, in so much as they are responding to the
time-varying fields, are also varying in time. That is, we now expect
that −∂/∂t(∇ · P) = ∂ Jb /∂t = 0. So, at
= ∂ρb /∂t = 0 and ∂/∂t(∇ × M)
first sight, if we allow Eqs. 1.23 and 1.24 to vary with time, there
appears to be no change to the equations–the sources vary and the
fields follow. Closer inspection, however, reveals that a time varying
polarization also represents another type of bound current density (in
within the material.
addition to the previously discussed Jb = ∇ × M)
This general “polarization current” source term is written as (see
Discussion 1.4):
∂ P
Jp = (1.58)
∂t
Thus, combining the current density source terms due to the material
responses of magnetization and polarization (Eqs. 1.22 and 1.58) with
the Maxwell time-dependent equation in vacuo (Eq. 1.57), we obtain
the 4th Maxwell time-dependent equation including material response,

+ ∂ P + εo ∂ E
∇ × B = μo Jf + ∇ × M (1.59)
∂t ∂t
where we can define a total material current density, Jt : (see

Discussion 1.5)

+ ∂P
Jt = Jf + ∇ × M (1.60)
∂t
so that Eq. 1.59 can be written more compactly as
∂ E
∇ × B = μo Jt + εo
∂t
= εo E + P and
Alternatively, through the use of the derived fields D

H = B/μo − M as defined before, Eq. 1.59 can be arranged into the
macroscopic form

∂D
∇ × H = Jf + (1.61)
∂t
so that the remaining equations may now be expressed through the

collection of Eqs. 1.29, 1.55, 1.56, and 1.61.
= ρf
∇ ·D (1.62)
∂ B
∇ × E = − (1.63)
∂t

∇ ·B = 0 (1.64)

∂D
∇ × H = Jf + (1.65)
∂t
which is a macroscopic representation of the fields including the

effects of material response alone (∇ · P and ∇ × M),
the effects of

time dependence alone (∂ B/∂t
and εo ∂ E/∂t) and the combined effect

of material response and time dependence (∂ P/∂t).
1.6 Plane wave propagation in vacuum

In the absence of charge and current (ρ = 0 and J = 0), the time
dependent Maxwell’s equations for a vacuum, Eqs. 1.54–1.57, are
given as
∇ · E = 0 (1.66)
∂ B
∇ × E = − (1.67)
∂t
∇ · B = 0 (1.68)
1 ∂ E
∇ × B = 2 (1.69)
c ∂t
If we take the curl of Eq. 1.67,
=− ∂
∇ × (∇ × E) (∇ × B) (1.70)
∂t
then expansion of the left-hand side and use of Eq. 1.66 yields
− ∇ 2 E = −∇ 2 E = − ∂ (∇ × B)
∇(∇ · E) (1.71)
∂t
Finally, substitution of Eq. 1.69 results in a wave equation for the

electric field,
1 ∂ 2 E
∇ 2 E − =0 (1.72)
c2 ∂t 2
And, in the exact same way, Eqs. 1.69, 1.68, and 1.67 can be used
to obtain an identical wave equation for the magnetic field, B. The
solution to such an equation, a plane wave, is of infinite extent and
of constant value along the planes transverse to its propagation and
so is not practically realizable (A more practical solution obtained
by imposing a transverse cylindrical constraint will be discussed later
in Chapter 12). In the direction of propagation (say, z), the general
solution to Eq. 1.72 has the form
E (z, t) = E+ (z − ct) + E− (z + ct) (1.73)
where both E+ (t = 0) and E− (t = 0) can be any general function of z.

For t > 0, E+ propagates toward positive z and E− propagates toward
negative z both with a speed c = (εo μo )−1/2 . In contrast to this general
solution, the simplest (and in some sense the most fundamental)
solution to Eq. 1.72 are the harmonic (monochromatic) transverse
electric plane waves. Indeed, not surprisingly, they form the basis set
by which any finite electric transverse wave, of the general form given
by Eq. 1.73, can be represented. Such harmonic solutions are, by
definition, of the form17 F ( r o )exp(−iωt) so that
r o , t) = F(
∂ E ∂ B
= −iωE and = −iωB (1.74)
∂t ∂t
while the vacuum plane wave equations for E (Eq. 1.72) and B reduce
to the Helmholtz equations

ω2 ω2
∇ + 2 E(
2
r o ) = 0 and ∇ + 2 B(
2
ro) = 0 (1.75)
c c
to which the solutions are of the form F ( r o ) = F + ω

o exp(i c n̂ · ro ) +
F − ω
o exp(−i c n̂ · ro ). Thus, solutions to Eq. 1.75 have required that we
set E and B spatial dependence proportional to exp(±i ωc n̂ · ro ) and so,

for example, a harmonic solution of Eq. 1.72 can be represented as
E ( r o )e−iωt = E+
r o , t) = E( −
o exp[i(kn̂ · ro − ωt)] + E o exp[−i(kn̂ · ro + ωt)]
(1.76)
in which we have introduced the wave vector k = kn̂ = (ω/c)n̂ that

points along the wave propagation direction. The physical field is the
real part of the expression in Eq. 1.76. Notice in Eq. 1.76 that if
we pull k out of the parentheses, we are left with the arguments
(n̂ · ro ± ωk t) = (n̂ · ro ± ct) thus obtaining the form of Eq. 1.73 with E+
o 17 The physical fields are represented by
moving in the direction of n̂ and E− o moving opposite the direction of n̂. the real parts of these wave functions.
If we consider only the positive-going solutions, we can express the

harmonic field solutions as
r o , t) = Eo exp[i(k · ro − ωt)]

E ( (1.77)
r o , t) = B o exp[i(k · ro − ωt)]

B ( (1.78)
Inserting these harmonic solutions, the two Maxwell curl equations

(Eqs. 1.67 and 1.69) become18
k × E = ωB and k × B = −ωε0 μ0 E.

(1.79)
both of which individually indicate that E and B are orthogonal and,

when taken together, indicate that k is orthogonal to both fields. And
B,
so the curl equations tell us E, and k are mutually orthogonal.19
18 For example, inserting 1.77 and 1.78
into Eq. 1.67, Finally, putting Eqs. 1.79 together with Eqs. 1.77 and 1.78,

∇ × Eo exp[i(k · ro − ωt)] E( = Eo exp(i k · ro − iωt)
r o , t) = −c(n̂ × B) (1.82)
∂
=−
∂t
Bo exp[i(k · ro − ωt)] r o , t) = 1 (n̂ × E)
B( = B o exp(i k · ro − iωt) (1.83)
c
noting that k · ro = kx xo + ky yo + kz zo , we
consider just the x component of the left side where we have used k = kn̂ = (ω/c)n̂ and εo μo = c−2 . And so, from
∂Ez ∂Ey
=
these equations we see that the field amplitudes are related by |E|
∇ × E = −
x ∂yo ∂zo c|B|. In summary, the transport of radiation in vacuum is a transverse
= iky Ez − ikz Ey

plane wave, moving at the speed of light in the direction k with E,
= i k × E
B and k mutually orthogonal and with the field amplitudes related
= c|B|.
Figure 1.2 is a representation of the properties of
x
by |E|
combining with the x component of the right
side
a plane wave.

i k × E = iωBx
x
1.6.1 Polarization of plane waves
which for all components
k × E = ωB According to the discussion before, E is subject to the condition

and in a like manner that it must be perpendicular to the direction of travel (n̂ · E = 0).
A moment’s reflection on this reveals that there must be two inde-

k × B = −ωε0 μ0 E.
pendent solutions for E corresponding to the two dimensions of the
19 Further confirmation that both E and
B must be perpendicular to k, the direction transverse plane. That is, given any arbitrarily chosen orthogonal pair
of wave travel, is found by inserting the har- of directions ê1 and ê2 = n̂×ê1 within a plane defined by this condition,
monic solutions (Eqs. 1.77 and 1.78) into the
two independent solutions for the wave equation would then be two
divergence equations (Eqs. 1.66 and 1.68).
We get electric vectors, E1 and E2 , of arbitrary magnitude pointed along
ê1 and ê2 , respectively. For example, for a wave propagating in the
∇ · E = k · E = 0 (1.80)
positive z direction, ê1 and ê2 could be taken to be in the x and y
∇ · B = k · B = 0 (1.81) directions. Specifically, the two solutions would read
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
– –
–
– –
– –
– –
– –
– –
– –
– – Fig. 1.2 Representation of a plane
– – wave. The electric field vectors
–
– are shown. The magnetic field (not
–
shown) is perpendicular to the electric
field and the direction of propagation.
The “sheets” of constant phase extend
infinitely in the directions transverse
to the propagation ([Public domain],
via Wikimedia Commons).
E1 = E1 exp(i k · ro − iωt − iα) ê1 (1.84)

E2 = E2 exp(i k · ro − iωt − iα2 ) ê2 (1.85)
where α and α2 are independent arbitrary phases set by the initial

conditions. In what follows, E1 and E2 are taken to be real and α
is assumed to be zero (if, necessary, this may be accomplished by
redefining the time origin). The total electric vector E (
r o , t) is then
the vector sum of 1.84 and 1.85:
r o , t) = E1 exp(i k · ro − iωt) ê1 + E2 exp(i k · ro − iωt − iα2 ) ê2

E (
(1.86)
→ Eo , since E 2 = E 2 + E 2 we can parametrize the ampli-

Defining |E| 0 1 2
tudes by the normalization of the field:
E1 E2
cos θ = and sin θ = (1.87)
Eo Eo
Then Eq. 1.86, may be rewritten:
r o , t) = Eo exp(i k · ro − iωt) (cos θ ê1 + sin θe−iα2 ê2 )

E( (1.88)
we define the unit polarization vector for E as ê⊥ :
ê⊥ = cos θ ê1 + sin θe−iα2 ê2 (1.89)
so that the wave is represented by
r o , t) = Eo exp(i k · ro − iωt) ê⊥

E( (1.90)
The physical field is found by taking the real part of this expression.
The direction of the electric field is given by the direction of

Re cos θ ê1 + sin θe−iα2 ê2 exp(i k · ro − iωt) .
We consider three cases:

Case 1: If the two components are in phase, α2 = 0. Then, for a
particular position and time:

Re cos θ ê1 + sin θ e−iα2 ê2 exp(i k · ro − iωt)

= cos θ ê1 + sin θ ê2 cos(i k · ro − iωt) (1.91)
The field is linearly polarized at an angle θ with respect to ê1 and its
amplitude varies in time and space.
Case 2: If the two components are equal in magnitude, θ = 45o ,
and are out of phase by 90o , e−iα2 = i, the wave has circular
polarization:

Re cos θ ê1 + sin θ e−iα2 ê2 exp(i k · ro − iωt)
1
= √ cos(k · ro − ωt)ê1 ± sin(k · ro − ωt)ê2 (1.92)
2
The magnitude of the field is everywhere the same but it rotates

in time at a fixed point in space and in space for a fixed time. For
example, taking the plus sign in the right-hand side of Eq. 1.90, for
t = 0, the electric field’s direction starting at the origin for increasing
distance along the wave’s direction of propagation, k · ro , increases and
the electric field direction rotates from ê1 (at the origin) toward ê2 and
beyond. This is “right circular polarization.” See Fig. 1.3. Note that
a person standing at a fixed location would observe a field rotating
clockwise at the wave frequency as the wave passed. Similarly, taking
the minus sign in the right-hand side expression produces left circular
polarization.
Fig. 1.3 The orientation of the total

perpendicular E field as it propagates
for circular polarization. This partic-
ular wave appears to be left handed to
the observer, right handed as viewed
from the source ([Public domain]. via
Wikimedia Commons).
Case 3: The general case is when neither of these two conditions

hold and this leads to elliptical polarization. Here, the electric field’s
direction rotates and its amplitude oscillates at the wave frequency.
1.7 E&M propagation within

simple media
As shown in Section 1.6 for a vacuum, the microscopic time depen-
dent Maxwell’s equations can be used to obtain wave equations for
E and B. In this section we will show that the macroscopic time
dependent Maxwell’s equations, Eqs. 1.62–1.65, are similarly used
to obtain the general equations for electromagnetic propagation in
material. In the following, we will assume the polarization P, mag-
and current density J all have the following simplifying
netization M,
properties:
(1) They respond linearly to applied fields (linearity)

(2) Their response is independent of the direction of the applied
fields (isotropy)
(3) Their response is independent of position and time
(homogeneity)
If we make the further assumption that ε, μ and σ are independent

= ε E
of frequency, then we can use the constitutive relationships D

and B = μH (Eqs. 1.37 and 1.38) and Ohm’s law J f = σ E (Eq. 1.39)
for general time-dependent fields (see Discussion 1.6) Finally, if we
assume that our medium has no regions of non-zero charge density,20
then the usual manipulation of Eqs. 1.62–1.65 yields:
∂ 2 E ∂ E
∇ 2 E = με + μσ (1.93)
∂t 2 ∂t
which is the general time-dependent, second order partial differential

equation for electric fields within linear, isotropic, homogeneous, and
frequency-independent material in the absence of free charge. In a
completely analogous way, we can obtain the identical equation for B,
∂ 2 B ∂ B
∇ 2 B = με + μσ (1.94)
∂t 2 ∂t
In Chapter 9, we will revisit electromagnetic propagation in matter

more systematically and in greater detail.
1.8 Electromagnetic conservation laws

1.8.1 Energy density
Maxwell’s equations give a prescription for how the total energy in
the electromagnetic field can change. If we add together the results of
taking the dot product of H with both sides of Eq. 1.63 and the dot
product of E with both sides of Eq. 1.6521
= E · Jf + E · D
E · (∇ × H ) − H · (∇ × E) ˙ + H · B˙ (1.95)
∇ · (E × H ) = −E · J − E · D
f ˙ − H · B˙ (1.96)
In the following, we continue to restrict ourselves to materials in

which the susceptibility properties are all linear, isotropic and homo-
geneous in the applied fields, and crucially, are also independent of
the frequency, at least in the neighborhood of the frequency spread
in the applied field. Then, if the surface element of the surface A
surrounding the volume V is designated d A, and we use the divergence
20 Note that the conductivity σ of the
theorem,
material is independent of the existence of
ρf so that we have here a situation in which
∂ 1
ρf = 0 and Jf = σ E = 0. (E × H ) · d A = − E · Jf dV − [E · D + H · B]dV

21 Make use of the vector identity:
S ∂t 2
= b · (∇ × a)
∇ · (a × b)
− a · (∇ × b). (1.97)
1.8 Electromagnetic conservation laws 31
If we now make the reasonable physical identifications of rates:

(E × H ) · d A =
S
energy flow through surface A of volume V

E · Jf dV =
electric field work on charges in volume V

∂ 1
[E · D + H · B]dV =
∂t 2
electromagnetic energy change in volume V
With these identifications, Eq. 1.97 expresses the conservation of
energy: In words, the energy flow into the volume through the surface
“A” is equal to the work done by the fields on the sources within the
volume, plus the increase in the field energy within the volume.
1.8.2 Poynting’s Theorem

Taking the differentials of Eq. 1.97 yields Poynting’s Theorem, a point
by point relationship between the field energy density, flow of energy,
and the work performed by the fields on charges:
∂
U + ∇ · S + E · Jf = 0 (1.98)
∂t
where U = 12 [E · D + H · B]
is the total electromagnetic field energy
density. Here, S = (E × H ), known as the Poynting vector, is the
flow of electromagnetic energy per unit area (i.e., intensity), and
E · Jf is the work per unit volume the electric field does on the
free charges in the volume. Equations 1.97 and 1.98 express the
macroscopic and microscopic conservation of energy, respectively,
Just as with Maxwell’s equations, Eq. 1.98 applies to each point in
space, at any given time. That is, the change of the energy density is
constrained at each space-time point by this microscopic relationship
(see Discussion 1.7).
1.8.3 Linear momentum density

We seek to construct an equation that expresses the conservation of
momentum in analogy to the conservation of energy. From our discus-
sion before, we have noted that the term E · Jfree is the rate of work done
by the electromagnetic fields on the free charges, and is thus the rate
of change of the mechanical energy of the free charges in the medium.
This statement is a concept readily adopted by considering the
current to be the sum of individual particles of charge q moving with

Then, writing E as the total energy in the volume (electro-
velocity v.
magnetic fields plus mechanical movement) (see Discussion 1.8):

d
E=− S · d A (1.99)
dt S

Physically, this equation states that when the influx of energy |S|
is inward on the surface A, the total energy, mechanical plus field,
within the volume surrounded by A increases with time in a prescribed
manner. Likewise, the analogous expression in the case of momentum
is expected to be of the form,

d
Ltotal = Pfields · d A
dt S
where Ltotal is the total linear momentum of the fields and the charges,
and Pfields is the flow of electromagnetic momentum density per unit
area. Here there are two momenta to consider: One is the momentum
associated with the fields, and the other is the mechanical momentum
of the charges. The expression for the mechanical momentum is
obtained, assuming only forces from electromagnetic fields on a single
charge, from
d pm
= total electromagnetic force on the charged particle
dt
= q(E + v × B)
(1.100)
Note that while the B field cannot do work on a charge, it can affect
a charge’s mechanical momentum, depending on the direction of the
charge’s velocity with respect to the field.
We anticipate the expression for the momentum of the fields by
noting the expression for the electromagnetic field energy density:
1 = 1 (E × H ) · k̂
U = [E · D
+ H · B]
2 c
where k̂ is perpendicular to both E and B and points in the direction of

energy flow. Because classical time-varying fields are a representation
of photons, they must share their energy/momentum relation, that is,
Pphoton = (Ephoton )/c, so the electromagnetic field momentum density is
U 1
P= = 2 (E × H ) · k̂ (1.101)
c c
Another random document with
no related content on Scribd:
“Robespierre has triumphed over the others, and he has
had Hébert, Vincent, etc., arrested and guillotined.
Robespierre had declared himself anxious to stop the flow of
blood ...; he had spoken up for the prisoners in the Temple.
Fresh letters are arriving here. It is certain, I think, that my
wife has not yet been charged with anything, or even
suspected of anything in regard to the prisoners.”
The event was inopportune. Cormier had just decided to leave
London for the coast, where he was to receive certain information
and to take counsel with his agents. Now his plans were all upset.
He would have to postpone the journey and redouble his
precautions.
At the end of five days there was ground for taking a hopeful view
of things. There was every reason to believe that Mme. Cormier’s
arrest would not have any grave results.
“What annoys me most,” writes Cormier to Lady Atkyns on
March 28, “is the fact that the news had got back to Paris,
with commentaries which may do harm both to my wife and to
our affairs.”
As a matter of fact, Peltier and d’Auerweck hastened, on hearing
of what had happened, to convey their sympathy to their friend, and,
like true journalists, spread the tidings in every direction, thus
intensifying Cormier’s uneasiness.
“But I must only try and put aside this anxiety,” he
continues, “as I have so many others. I have not yet started; I
shall not start before Monday or Tuesday, because I must wait
for replies from Dieppe, which cannot arrive before Sunday or
Monday. Have no fears; my courage will not fail me—indeed,
at present it is taking the shape of a feeling of rage, which I
am trying to keep down. You will have learnt from the public
prints that the statement has gone out that the King has been
carried off to the army of the Prince of Saxe-Coburg. This
false report has troubled me a good deal. I don’t want
attention to be directed that way just now, especially as
something has happened which would increase our
confidence—something which I cannot at present confide to
paper. Do not exert yourself too much, madame; do not
measure your efforts by your courage. Your friends beg this of
you.”
In all these letters of the Breton magistrate there is a real ring of
sincerity. The admiration he feels for this interesting woman resolves
itself into a whole-hearted devotion to her cause, and if, later, her
large fortune and her generosity seem to have too large a part in
Cormier’s thoughts and too great an influence upon his actions, at
least he must be credited with absolute frankness throughout.
The death of Sir Edward Atkyns on March 27, 1794, gave Cormier
an opportunity for expressing his sympathy with the widow, and of
enlarging still further upon his feelings. The scant mention made of
Sir Edward, indeed, in the correspondence of this little circle
suggests that the relations between husband and wife must have
become perceptibly colder of late. It is probable that the baronet
looked with disfavour upon his wife’s schemes and the heavy outlay
they entailed.
“A score of times,” writes Cormier, “I have taken pen in
hand this morning to express to you the intense interest with
which I have learnt of the sad event which occurred, and as
often my courage has failed me. Truly you have been the
victim of many misfortunes. Will the Fates never have done
pursuing you? You must only make use of the great qualities
Providence has given you to bear up against what has
befallen. Your courage is exceptional. Make the most of a
quality which is rare with men, but rarer still in women. As for
me, I vow I shall not give in under my misfortune, and shall
not be put off by any perils.... I have not started yet, and shall
not start to-morrow, not having yet received the letters I was
expecting. If they come to-morrow, I shall start on Thursday.
So that this delay may not cause you anxiety, I may mention
that in the last letters which have come to me, he who left last
... asks me not to start until I heard again from him. He has
not been beyond D(ieppe), and the others have returned from
P(aris) to take counsel with him—I don’t know on what.”
These last words show that something was already happening on
the Breton coast, and that it was desired to send news of interest to
Cormier. But the departure postponed so often was still
impracticable, and Cormier began to lose patience.
“I am still kept here,” he writes. “It is becoming incensing. I
feel as though I were being chained up, but prudence and
common sense keep me quiet. I get news regularly from
D(ieppe). I have just received a third letter enjoining me to
make no movement until they give me the word, and insisting
that the success of our project and the safety of him who is so
precious to us depend upon this. I don’t understand, however,
their not telling us why and how.... I have lost patience, and
have sent one of these gentlemen.[72] (That is not the same
as myself.) I am afraid that Hamelin may really have been
killed; I can’t make it out at all.”
Who was Hamelin? It is difficult to guess. It is difficult to identify a
great many of the individuals of whom there is question in these
letters, and who are designated by borrowed names. The most
elementary prudence called for absolute secrecy concerning the
names of the agents who were working for our committee, and
although the messages were carried by the most trustworthy
emissaries, it was always possible that one of them might be
arrested en route. This doubles our difficulty in clearing up the
imbroglio, and enhances a mystery already sufficiently troublesome.
Failing Mme. Cormier, who was still under arrest, and whose
absence had been making itself felt more and more, another
arrangement had been made for securing news from Paris. At what
expense? Heaven knows! But once again money had set tongues
going and procured the needed help. Cormier, coming back to the
question of his departure, writes again (April 14, 1794) to his friend
to tell her of the messages he has sent from England:—
“I shall not start until this evening,” he tells her. “You can
guess why. I have just despatched two messengers. Things
are moving, but very slowly. However, let us not lose heart. If
we go slowly we go all the more surely, and every day
achieve something which helps to advance our schemes and
to keep us in security. Therefore do not be impatient.”
The weeks passed by, and that fateful day “9th Thermidor,” which
was to bring with it such a bouleversement in Paris, was drawing
nigh. At the Temple there had been no change—the Dauphin was
still sequestrated from the outside world.
On May 11, 1794, Robespierre visited the prison, and had a brief
interview with Marie-Therèse, but we have no information as to what
happened.
The 9th Thermidor arrives and throws the dictator down from his
pedestal, thereby proclaiming the end of his reign of terror. General
Barras, invested with the command of the armed forces within the
city, begins to take an important part in the management of affairs.
One of his first acts, it will be remembered, after he had triumphed
over Robespierre’s party, was to go to the prison of the Temple, on
July the 28th, accompanied by his brilliant staff, bedecked with gold.
The miserable aspect of the child after being shut up for months
caused the general to take immediate steps, and by his order of July
29, 1794, a special guardian, chosen by himself, named Laurent, a
native of Martinique, was brought to the prison, there to be entrusted
with the sole care for nearly five months of the young Capet.
A careful study of the documents bearing upon this period of the
captivity of the Dauphin makes it quite clear that in the hands of his
new guardian he was looked after in a fashion which contrasted
strongly with the previous neglect, and that he soon became
attached to Laurent, who proved himself good-natured, kind, and
even affectionate in his attitude towards his charge. If strange things
came about in the Temple at that time, we may be certain that
Laurent knew about them, and we may assume that Barras was the
prime mover in all that happened.
It is impossible, as we have said before, to recapitulate all the
arguments which tend to bring home to the general some complicity
in the fate of Louis XVII., and which implicate a large number of
persons, most of them people of influence in the world of the
Convention. Other writers, notably M. Henri Provins,[73] have done
this so conscientiously and thoroughly that there is no need for us to
attempt it. We may content ourselves with making public a series of
documents and newly ascertained matters, the gist of which bears
out exactly all that we knew already of Laurent’s conduct at the
Temple. Lady Atkyns and her friends could not have done without
him. It is true that his name never appears in their communications,
for reasons already given, but the striking connection between the
events within the prison walls and their effects in London upon the
Royalist Committee proves beyond doubt the relations subsisting
between them. Between the lines of these documents we get to
understand what Cormier meant by “new combinations.” Lady
Atkyns has been at pains to say it herself in one of her notes which
she used to make upon her correspondence, and which often serve
to explain her actions.
In his anxiety about the future, did Cormier entertain fears lest all
remembrance of his heroine’s devotion would vanish with her if by
some mischance her enterprise should fail, or if she herself should
lose her life? Who knows? However that may be, it is the case that
on August 1, 1794, he had two statements drawn up (the text of
which, unluckily, is not forthcoming), in which Lady Atkyns recorded
all that she had achieved down to that date for the safety of those
who were so dear to her.
“These records are to my knowledge the absolute truth,”
attested Cormier at the foot of the deposition, “and I declare
that ever since I first knew Lady Atkyns, she has always
shown the same purity of principles, and that all she has here
stated is true in every particular.”
These documents were to have been handed over for
preservation, with a number of others, to a solicitor or some
trustworthy person in London.
Meanwhile, renewed efforts were being made to bring about a
good service of news to the Continent and Paris. As time passed,
Lady Atkyns’ friends realized more and more that it would have been
madness to proceed with a regular attempt at a sudden rescue in the
actual conditions of things. In truth, the calm which had followed the
9th Thermidor, and which gave Paris time to take breath, was
making itself felt within the Temple. Laurent’s nomination was
evidence of this. Any attempt to act at once would have been sheer
folly. What was to be done was to “get at” those who had any kind of
influence within the Temple or without, whilst taking care not to let
too many people into the secret of the enterprise. Here, again,
unluckily, the wise secretiveness of all their papers prevents us from
ascertaining any names. Those who were tempted by Lady Atkyns’
gold to compromise themselves in any way, took too many
precautions against being found out.
Lady Atkyns, however, was not idle. Two sailing vessels were
continually plying between different points on the French coast. A
third, which she had recently purchased, had orders to keep close to
land between Nantes and La Rochelle, ready at any moment to
receive the Dauphin.[74]
The cost of keeping these three ships was considerable, and Lady
Atkyns had great difficulty in providing the money. She was in the
hands of agents whose services, indispensable to her, could be
depended upon only so long as the sums they demanded were
forthcoming. We can imagine the feelings of anxiety and
despondency with which she must have read the following letter from
Cormier. What answer was she to make to him? (The person to
whom she had applied for financial help appears on several
occasions in their correspondence under the designation of “le diable
noir.”)
“Your diable noir’s reply is very little consolation to me,”
writes Cormier; “he has promised and postponed so often.
For Heaven’s sake, see to it that he does not promise us this
time also to no purpose!... I gather that you were to have two
definite replies to-day—I shall be in Purgatory until five
o’clock. Mon Dieu! Mon Dieu! I wonder what you will send me,
or rather what you will be able to send me? Our own courage
alone does not suffice—we have to keep up the courage of
others, and they are losing heart. Worst of all, there is that
avaricious Jew of a captain! We are absolutely dependent
upon him. If we lost him where should we get another to take
his place? I beg of you, in the name of the one you know, to
do all you possibly can, to exert all your resources, to prevent
his having to leave me empty-handed.”
And to excuse the ultimatum-like tone of his letter, Cormier adds—
“Forgive the urgent persistent style in which I write! But
when one is writing about business matters and matters of
this importance, one has to forget one is writing to a woman—
especially when it is a question of a Lady Atkyns, who is
different from the rest of her sex.”
The occasions for entering into communication with their agents
on the Continent are more propitious now than ever, but many efforts
are frustrated owing to the sharp watch which is kept along the
coast.
“They have tried eleven times to land since Saturday last,”
writes Cormier, “and failed every time. There were always
either people in sight or else there were transports sailing
from Havre to Dieppe or from Dieppe to Saint-Valery, etc., etc.
There has been a lot going on evidently, for signals have
been given on fifteen or twenty different occasions. That
shows how important it is to effect a landing. They returned
simply to make this fact known to me, and went back again
without coming on shore—except the captain, who came for
an hour and who is positive they have something to hand over
to him. I believe this myself, for I learn also this morning that
the Government boat which plies along the coast of Brittany
has made thirty vain attempts during the last three weeks.”
We can imagine the mental condition of poor Lady Atkyns on
receiving letter after letter in this strain. She no longer goes away
from London at this period, feeling too remote in the country from the
centre of news. She stays either at the Royal Hotel or else with
friends at 17, Park Lane. Here it is that she receives Cormier, Frotté,
Peltier. When there is a long interval between their visits her fears
grow apace. What would she not give to take an active part herself in
the enterprise! “No messenger arrived—no news, therefore, from
France,” that is the message that comes to her only too often. And
Cormier writes, full of excuses for his persistent appeals—
“Forgive my tone,” he writes. “I apologize a thousand times
for being such a worry to you, but I can’t help it in regard to so
important a matter, calling for so much energy and hurry. You
have voluntarily abandoned the position ensured you by your
sex and great advantages in order to play the rôle of a great
and high-minded statesman. There are discomforts and
disadvantages attached to this new estate, and it is my
misfortune to have to bring this home to you. I can but
console myself with the thought of your goodness and of the
great cause which we have embraced and which is the
subject of all our anxieties. May God prosper it, and may it
bring you glory and me happiness!”
In the mouth of any one but Cormier these protestations would
arouse one’s distrust; but what we already know of him, and what we
are to learn presently of his later conduct, serve to reassure us in
regard to him.
In spite of all his good will, however, Cormier is constantly being
interrupted in his work. Now it is the health of his son, Achille, which
disquiets him, now he is a prey to terrible attacks of gout which will
give him no rest.
“I have been bent double for two nights and a day,” he
writes to his friend on September 1, 1794, “without being able
to change my position. It takes four persons to move this
great body of mine. I am a little more free from pain at
present, and I take up my pen at the earliest possible moment
to send you this explanation of my silence.”
It is at this moment that Louis de Frotté, who has been a little in
the background, comes again to the front of the stage. Since his
arrival in London, the young officer, without neglecting the society of
the Royalist Committee, has been spending most of his time in the
offices of the English Government, endeavouring to impress upon
Windham “the desirability of carrying out his ideas, and the ease with
which they may be brought to fruit, as he has made up his mind to
devote himself to them.” One project he has specially at heart, that of
receiving some kind of official mission from the Government which
will enable him to land in Normandy with adequate powers and to
give new life there to the Royalist insurrection. Should he succeed,
the help he “would thus obtain would lead to the execution of our
cherished plans,” he writes to Lady Atkyns, and she will reap at last
“all the honour that will be due to the generous sacrifices that she
has made.”
But in his interview with the Minister he does not think it necessary
to speak of their relations with the Temple. This secret is too
important for him to confide it to any one. “Too many people know it
already.” These words, hinting a delicate reproach, are meant,
perhaps, to put his fair friend upon her guard. Perhaps they mean
more than that. Read in the light of subsequent letters from the
young émigré, they serve as a key to his private feelings—to his
dislike at having to share her confidence with so many others, and to
his jealousy later of the man who has so large a place in her heart.
These feelings, still slight, soon become more marked, and presently
we find that they are reciprocated.
For the time being, however, both Frotté and Cormier worked with
the same ardour at their allotted tasks. Frotté, proceeding with his
negotiation with Windham, counted now upon support from Puisaye,
his famous compatriot recently come to England. Cormier writes to
her to report that, despite apparent dilatoriness, their agents have
not been inactive.
“I have received letters through the captain,” he tells her on
October 1, 1794, “which satisfy me, brief as they are. Here is
what they have to tell me: ‘Be at ease in your mind; they
imagine they are working for themselves, and really they are
working for us, and we shall have the profit. Be patient and
don’t lose trust.’ The captain had orders to return at once to-
day, but he will not start until to-night or to-morrow morning,
and we have news by the packet-boats meanwhile that order
reigns in Paris.”
Day after day passed by, bringing new reports, none of them
positive, of the death of the little Dauphin. Lady Atkyns knew not
what to make of the situation. Presently—eight days after the last—
there came another letter from Cormier, to reassure her.
“I have great faith in your judgment,” he declares, “and your
presentiments are almost always right, but I really do not think
that you have ground for disquiet now. Three agents of ours
at the Temple are either at work silently or else they are in
hiding. All we know for certain is that they have not been
guillotined, as they have not been mentioned in any of the
lists.”
His wife was still unfortunately detained, but there was prospect of
her being shortly at liberty, and then she would write to him. If the
agents had taken it upon themselves to modify their project—the one
thing that was to be feared—they could not possibly have succeeded
in sending particulars yet of this. But an explanation of the mystery
was soon to be forthcoming.
“The Dauphin is not to be got out by main force or in a balloon,”
Cormier had once written. Any attempt at carrying him off under the
very nose of his warders and of the delegates of the Commune
would have been madness. All idea of such a rescue had long been
put aside. How, then, was the matter to be dealt with? By such
means as circumstances might dictate—by finding a substitute for
the young prisoner, a mute who should play the rôle until an
occasion should offer for smuggling away the real Dauphin,
concealed meanwhile somewhere in the upper chambers of the
Tower. Mme. Atkyns did not herself approve of this plan.
“I was strongly opposed to it,” she notes at the foot of a
letter from Cormier dated June 3, 1795, “as I pointed out to
my friends that it might have an undesirable result, and that
those who were being entrusted with the carrying off of the
Dauphin, after getting the money, might declare afterwards
that he had not been got out of the Temple.”
She saw reason to fear that at the last moment she would be done
out of the recompense of all her efforts, and that the Royal child
would not be entrusted to her care.
However, it was clear that once the plan was agreed upon it was
necessary in order to carry it out to secure the help of the gaoler
Laurent, who had had the Dauphin under his charge during the last
four months. Laurent’s complicity may be traced through the
documents bearing upon the whole episode.
Let us examine first of all Laurent’s own famous letters, the first of
which, dated November 7, 1794, synchronizes with the events we
have been following.
It is well known that only copies of these letters are in existence—
the originals have never been discovered. They were published first
in a book which appeared in 1835, Le Véritable Duc de Normandie,
the work of an adherent of the pretender, Nauendorff, Bourbon-
Leblanc, whose real name was Gabriel de Bourbon-Russet, dit
Leblanc. From the fact of the originals being missing, the authenticity
of these letters has long been a matter for debate. A close
examination of them, side by side with all the other documents upon
which we have come in the course of our researches, results, we
think, in justifying our belief in their genuineness.
Cormier, then, was not mistaken in supposing that his agents had
modified their plan. The letter in which he confided his suspicion to
Lady Atkyns was dated October 8, 1794. On the last day of the same
month he wrote to her again:—
“I have to thank you cordially for your kind letter of
yesterday. I have had no time to answer it properly, not
because of the gout, for that has left me. In fact, my mind is
so fully occupied that I have no time to trouble about any kind
of malady, and am, in fact, at my wits’ end with excitement.
However, I must just send you this brief note in haste (for it is
just post time) to bid you not merely be at rest but to rejoice! I
am able to assure you positively that the Master and his
belongings are saved! There is no doubt about it. But say
nothing of this, keep it absolutely secret, do not let it be
suspected even by your bearing. Moreover, nothing will
happen to-day, or to-morrow, or the day after, nor for more
than a month, but I am quite sure of what I say, and I was
never more at my ease in my own mind. I can give you no
details now, and can only tell you all when we meet; but you
can share my feeling of security. I am glad to say I have good
news of my wife, but I must continue to keep a sharp look out
all round me.”
This letter evidently alludes to what had happened at the Temple.
The young Dauphin, we may conclude, was halfway on his road to
liberty. Lodged in the garrets of the Temple tower, and with the little
mute as his substitute down below, he was not yet out of peril. But
an important step had been taken towards the ultimate goal.
It seemed clear that Laurent, l’homme de Barras, was having a
share in this, and had at least rendered possible the execution of the
project. The letter which he wrote eight days later to a general,
whose identity has never been established, bore out exactly what
Cormier had said; here it is:—
“General,
“Your letter of the 6th came too late, for your first plan had
been carried out already—there was no time to lose. To-
morrow a new warder is to enter upon his duties—a
Republican named Gommier, a good fellow from what B——
tells me, but I have no confidence in such people. I shall find it
very difficult to convey food to our P——. But I shall take care
of him; you need not be anxious. The assassins have been
duped, and the new municipal people have no idea that the
little mute has been substituted for the Dauphin. The thing to
be done now is to get him out of this cursed tower—but how?
B—— tells me he cannot do anything on account of the way
he is watched. If there were to be a long delay I should be
uneasy about his health, for there is not much air in his
oubliette—the bon Dieu would not find him there if he were
not almighty! He has promised me to die rather than betray
himself, and I have reason to believe that he would. His sister
knows nothing; I thought it prudent to pass the little mute off
on her as her real brother. Meanwhile, this poor little fellow
seems quite happy, and plays his part so well, all
unconsciously, that the new guard is convinced that he is
merely refusing to speak. So there is no danger. Please send
back our faithful messenger to me, as I have need of your
help. Follow the advice he will convey to you orally, for that is
the only way to our success.
“The Temple Tower, November 7, 1794.”
The contents of this letter, taken together with its date, accord in a
remarkable way with Cormier’s communication to Lady Atkyns.
There is another striking argument in favour of the authenticity of
Laurent’s letters. When they were produced by the pretender
Nauendorff, they were for the most part in complete contradiction to
all that was known of the Dauphin’s captivity and the testimonies of
those connected with it. Certain facts to which they made allusion
were known to nobody. Thus Laurent states clearly on November 7
that a new warder—whom he calls Gommier instead of Gomin—is to
come to the Temple next day and to be associated with him. Now, in
1835, when this letter was published, what was known of Gomin?
Next to nothing, and the little that was known did not tally with
Laurent’s statements. Simeon Despreaux, author of a book entitled
“Louis XVIII.,” published in 1817, did not even know of Gomin’s
existence. Gomin himself made a formal declaration before the
magistrates that he entered the Temple about July 27, 1794, before
Laurent was there at all. Many years later it was found, on examining
all the documents referring to the Temple that were kept in the
National Archives, that Laurent’s statements were quite correct.
Some days after this letter to Lady Atkyns, Cormier informed
Frotté of the great news, in the course of a visit paid him by the
latter.
“I know all about it,” he said, according to Frotté’s account
of the interview afterwards in a letter to Lady Atkyns,
“because they could do nothing without me; but everything is
now ready, and I give you my word that the King and France
are saved. All the necessary steps have been taken. I can tell
you no more.... Do not question me, don’t try to go further into
the matter. Already I have told you more than I had any right
to, and from Mr. Pitt down to myself there is now no one who
knows more about it than you do. So I beg of you to keep it
absolutely to yourself.”
From November 8, then, Laurent is no longer sole guardian of the
young Prince. His duties are henceforth shared with Gomin. What
kind of relations subsisted between the two? It is hard to say, for it is
even more difficult to find out the truth about the Temple during the
subsequent months than during those which went before.
We find one innovation introduced during these months which is
worth noting. It is no longer the delegates of the Commune who have
to pay the daily visit to the prison, but the representatives of the
Comités Civils of the forty-eight divisions of Paris. Now, among all
those who visited the Dauphin none left any record, with one
exception, to which we shall come presently. All that we can learn
from Gomin’s own statements, so often contradictory, is that
throughout the period the child placed under his care uttered no
word. The warder takes no further notice of this strange conduct,
Laurent having satisfied him that if the Dauphin will not open his
mouth it is because of the infamous deposition against his mother
that he was made to sign. It is unnecessary to point out how
improbable was this explanation, the Dauphin’s examination having
taken place on October 6, 1793, and Laurent not having come to the
Temple until July 29, 1794. Gomin, however, asked no further
questions, and Laurent experiencing no further anxiety in regard to
him, sought what means he could of bringing about the desired end.
Six weeks pass, however, without further progress, and then on
November 5 Laurent hears, to his great satisfaction, that his master
has become a member of the Committee of Public Safety. This new
office would surely enable the general to carry out his plan and
relieve the anxious guardian from the heavy responsibility lying on
his shoulders.
It was, therefore, not without surprise that on December 19
Laurent and Gomin saw three Commissioners of the Committee of
Public Safety make their way into the prison and up the stairway of
the Tower to the Dauphin’s cell. These three visitors—Harmand la
Meuse, Matthieu, and Reverchon—asked to see the Dauphin, so
that they might question him and satisfy themselves as to the way in
which he was kept under supervision. At a time when there were so
many rumours current about the Temple, and when rescues were
openly talked about, when every day brought forth some new
sensational report, it was only natural that the Convention, in order to
silence these rumours and calm public opinion, should institute an
official inspection of the prison in this way.
In a work which he published twenty years later, Harmand de la
Meuse tells us all that we know of this visit, and of the impression
made upon the delegates by the little mute ushered into their
presence. Suffice it here to record that this narrative (written with an
eye to the good graces of Louis XVIII.) makes it quite clear that it
was a mute whom they saw, and that all efforts to extract replies
were quite in vain.
Harmand repeats the explanation of this persistent silence which
had been furnished by Laurent. He ignores the fact that the Dauphin
had talked with the Simons, had been interviewed by Barras, and
had been heard to speak on several other occasions.
Assuredly, Harmand and his colleagues—his narrative allows it to
be seen on every page—very soon realized that they were not in the
presence of the Dauphin. This is proved by the fact that, despite the
very distinct terms of the resolution of the Committee entrusting them
with this mission, and the object of which was to dispel the rumours
current in Paris, “they decided they would make no public report, but
would confine themselves to a secret record of their experience to
the Committee itself.”
However natural and intelligible all this may have been to those
who knew what was in the mind of the Convention and the
exigencies of the situation at this period, to Laurent it was a matter of
stupefaction. Barras had sent him no warning, and his position was
getting more and more difficult, for his colleague, who had, of
course, to be taken into his confidence, was beginning to be nervous
about participating any further in the intrigue, and might betray him
any day. At last he loses patience, and expresses himself as follows
to his friend the unknown general:—
“I have just received your letter. Alas, your request is
impossible. It was easy enough to get the ‘victim’ upstairs, but
to get him down again is for the moment impossible, for so
sharp a watch is being kept and I am afraid of being betrayed.
The Committee of Public Safety sent those monsters Matthieu
and Reverchon, as you know, to establish the fact that our
mute is really the son of Louis XVI. General, what does it all
mean? I don’t know what to make of B——’s conduct. He
talks now of getting rid of our mute and replacing him by
another boy who is ill. Were you aware of this? Is it not a trap
of some kind. I am getting very much alarmed, for great care
is being taken not to let any one into the prison of our mute,
lest the substitution should become known, for if any one
examined him they would discover that he was deaf from
birth, and in consequence naturally mute. But to substitute
some one else for him! The new substitute will talk, and will
do both for our half-rescued P—— and for myself with him.
Please send back our messenger at once with your written
reply.
“The Temple Tower, February 5, 1795.”
Let us note the date of this letter—February 5. Therefore the visit
referred to must have taken place before February 5. Now, Eckard,
one of the earliest biographers of the Dauphin, having in the first
edition of his book made the date December 2, 1794, altered it
afterwards to February 13, 1795. De Beauchesne makes it February
27. Chantelauze, February 26.
On referring to the original documents at our disposal, however,
we find that Laurent’s letter is borne out. In his book, Le dernier roi
legitime de France, M. Provins shows that the visit must have taken
place between November 5, 1794, and January 4, 1795, as it was
only during this period that the three delegates were all members of
the Committee. A recent discovery of documents in the National
Archives establishes the fact that it took place on December 19,
1794.
FOOTNOTES:
[66] G. Lenôtre, Vielles Maisons, Vieux Papiers, 2nd series.
[67] A curious plan of this house is to be found at the
Bibliothèque Nationale, Print Department, Paris topography, the
Madeleine quarter.
[68] The decree of divorce of Marie-Anne-Suzanne-Rosalie
Butler, forty-nine years old, born at La Rochelle, resident in Paris,
Rue Basse, section des Piques, daughter of Jean-Baptiste Butler
and of Suzanne Bonfils; and Yves-Jean-François-Marie Cormier,
aged fifty-six, born at Rennes, department d’Ile-et-Vilaine, son of
the late Yves-Gilles Cormier and of Marie-Anne-Françoise
Egasse.
[69] V. Delaporte, article already quoted, Études, October,
1893, p. 265.
[70] Unpublished Papers of Lady Atkyns.
[71] Note in Lady Atkyns’ own handwriting at the end of a letter
of Cormier’s, dated March 24, 1794.
[72] M. M. de Corbin (note on the letter in Lady Atkyns’
handwriting).
[73] Henri Provins, Le dernier roi légitime de France, Paris,
1889, 2 vols.
[74] Note in Lady Atkyns’ handwriting at the foot of a letter from
Cormier, dated June 3, 1795.
CHAPTER V
THE MYSTERY OF THE TEMPLE (continued)
Meanwhile the feelings of jealousy and suspicion which had

sprung up between Cormier, still Lady Atkyns’s principal lieutenant
and confidant, and the Chevalier de Frotté were becoming more and
more marked. At the beginning of October, 1794, Cormier learns of a
correspondence in progress between Lady Atkyns and a person
whom he imagines to be his rival (but who turns out to be merely the
“little baron”), and his ill-humour breaks out in the form of
reproaches.
“Chance has willed that I should become acquainted with
the fact that some one has been getting up a correspondence
with you,” he writes to Lady Atkyns, “in such a way as to
prevent me from hearing of it.... You will admit that I am
justified in assuming there are reasons why this
correspondence is being kept secret from me.”
But he proceeds to assure Lady Atkyns that she still retains all his
admiration and respect, and to protest that he only acquaints her
with the discovery that he has made because of his attachment to
her. Filled with mistrust of Frotté, Cormier withholds from him
particulars as to the progress of affairs at the Temple, and only
vouchsafes his information now and again in vague terms. “I refused
to give Frotté the names of the agents,” he wrote to Lady Atkyns
some months later. “Please remember that. I shall always be proud
of that.”
It is not astonishing that Frotté should show some surprise at the
way in which he was being treated, though he was prevented by
other causes of annoyance—his failure to get any satisfaction out of
the British Government and the repeated postponements of his
departure—from taking his position in this respect too much to heart.
Lady Atkyns herself was keeping him at a distance at this time and
avoiding him when she came to London. When he asks for an
interview, she refuses on the pretext of her widowed state and public
opinion.
“I wished to avoid seeing or writing M. de Frotté,” she
herself records at a later period, “as I was not in a position to
talk to him about the means being taken for the rescue of the
King.”
However, on the eve of setting out from England into the unknown,
the Chevalier makes one more effort to see her.
“You do not write to me,” he begins his letter (December 27,
1794), “and I should be angry with you if I could be angry with
any one, now that I have all my wishes fulfilled. In three days
everything has changed, and I have nothing more to ask for in
England. The longed-for moment has come. P[uisaye] wants
me. I go with him, and all my requests are granted. We start
on Thursday at latest. It is important that I should see you. I
beg of you to set out at once and spend twenty-four hours
here, but without any one knowing of your journey, lest its
object should be suspected. Try to be here by Monday
evening, and let me know where I could see you.”
This time the appeal was too strong to be resisted. It was in the
depths of winter, and the letter arrived at Ketteringham in the
evening; but Lady Atkyns hired a post-chaise at once, and set out a
few hours later, and travelled all night in stormy weather to London,
arriving there in the morning. She seems, however, to have resisted
the temptation to let Frotté into the secret of the Temple doings.
Perhaps she had a presentiment that the Chevalier, for all his
protestations of fidelity now, would fall away later and pass into the
camp of some other pretendant to the throne.
We have spoken already of the endless intrigues which were
being hatched round the British Government by the hordes of
émigrés and broken-down exiles from the Continent. For these
gentry, mostly penniless and forced to beg their livelihood, no
resource was too base by which they could get into favour with the
Ministers. Besides scheming in a thousand different fashions against
the common enemy, the Revolution, they stuck at nothing in their
efforts to throw suspicion upon each other. The little court which had
gathered round the Comte D’Artois on the Continent was also a
hotbed of plots and schemes, the influence of which made itself felt
in London. Every one spied on every one else.
In the midst of this world of intruders a sort of industrial association
came into being in the course of the year 1794, for the purpose of
inundating France with false paper-money. It was hoped that in this
way a severe blow would be dealt at the hated Jacobins and their
friends. These nefarious proceedings soon became known, and
called forth the indignation of some of the better class of émigrés,
among them the honest Cormier.
His position among his compatriots was not at this time of the
best. They had no love for this man of firm character, faithful to his
principles and incapable of lending his countenance to such doings.
He himself soon came to realize this.
“One doesn’t know whom to trust,” he wrote to Lady Atkyns.
“I am sure some one has furnished the Government with a
long report upon my projects. I am on the track of the man
who I think is guilty. There is no reason for you to be anxious
on the subject. I shall soon know what has been done, and
both the traitor and the Government shall be outwitted.”
About this time a flood of memorials of all sorts poured in by
mysterious channels upon the British Government, maintaining that
“the general desire of the French was for a change in the ruling
family.” Cormier discovered that they all were traceable to the same
source, and we find him declaring energetically that “the
blasphemous scoundrels” who were responsible for them all
belonged to one clique.
His indignation, in which he found few sympathizers, made him a
number of enemies, and the disfavour with which he was already
regarded in French circles soon changed into downright hatred. The
fact that he denounced the false paper-money to the British
Government—and not in vain—was a cause of special bitterness

Electromagnetic Radiation Richard Freeman 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

Electromagnetic Radiation Richard Freeman Full Chapter

Uploaded by

Copyright:

Available Formats

Electromagnetic radiation Richard

Richard Freeman, James King,

Part I Introductory Foundations

Part II Origins of Radiation Fields

Part III Electromagnetism and Special Relativity

6.2 Radiation fields from a single charge undergoing acceleration 190

8 Field Reactions to Moving Charges 267

Part IV Radiation in Materials

9.9.2 Single frequency measurements 348

12.3.1 Kirchhoff’s integral theorem 471

A.3.2 Momentum density and angular momentum density 591

• Review of Maxwell’s steady-state equations in vacuum 1.2 Maxwell’s static

• Electromagnetic conservation laws, including electromag- 1.6 Plane wave propagation

Maxwell’s equations are the foundational equations of classical elec-

Electromagnetic Radiation. Richard Freeman, James King, Gregory Lafyatis,

1.1.1 Electrostatic equations

where we have used ∇ · (R̂/R2 ) = 4π δ(R). This is the differential form

where φ (r o ) is the scalar potential. Because in general the curl of a

which is the second of the two electrostatic Maxwell’s equations and

1.1.2 Magnetostatic equations

where, as before, the distance and direction from a source element to

which, with some manipulation (see Discussion 1.1), can be written

where, again, we have used ∇ · (R̂/R2 ) = 4π δ(R). This is the differen-

where the first term vanishes because the J

1.1.3 Lorentz force

This description of the total force on a charge in the presence of

netic effects due to the motion of electri-

working with fields in matter, to divide the problem conceptually into

1.2.1 Response of material to fields

magnetic field is known as diamagnetism. In this case, there is no

where in the same way < m (

< p( r ) Ep

where εo is generally included for later convenience. Combining

Now, it is an easily overlooked but important point that the macro-

1.2.2 Bound charges and currents

where S is the surface bounding the material and n̂ is the outward

By similar considerations of the vector potential at a point ro due to

which is to say that the curl of the magnetization ∇ × M (

1.2.3 Macroscopic fields

we can then define two new macroscopic fields

So that in terms of these new fields, which represent the fundamental

It is important to emphasize that while both D and H have only

1.2.4 Polarizability and Susceptibility

where, for example, in the case of gases, this connection is trivial:

which, as mentioned, gives the macroscopic susceptibility (χe ) in terms

1.2.5 The canonical constitutive relations

in materials and Magnetism, 3rd edition, Cambridge Uni-

For linear responses (P = εo χe E and M =

where σ is the conductivity of the material. The introduction of con- D = εo (1 + χe ) E = ε E

ductivity here is properly a “constitutive” relation, because the concept B = μo (1 + χm ) H = μH

is inherently macroscopic. The conductivity, as expressed in Eq. 1.39,

1.3 Energy of static charge and

where the factor of 1/2 arises in this summation because we have

included. To calculate the total electrostatic energy of a charge dis-

If we note that the charge density is given at each location by ρ =

the relation and a sum in which i = j,

Now the volume is occupied by point charges, ρ( r ) = i qi δ(

which is in agreement with the “configuration energy” we noted in

1.3.2 Magnetic field energy

dWij = −Vij dqi = Ii dφijs

where, again, the factor of 1/2 arises in this summation because we

and thus Eq. 1.47 can be written,

which is the magnetic analogue to Eq. 1.41 for electrostatic energy.