Visit https://ebooknice.

com to download the full version and

explore more ebooks

(Ebook) Power Exhaust in Fusion Plasmas by

Wojciech Fundamenski ISBN 9780521851718,
0521851718

_____ Click the link below to download _____

https://ebooknice.com/product/power-exhaust-in-fusion-
plasmas-1806942

Explore and download more ebooks at ebooknice.com

Here are some suggested products you might be interested in.
Click the link to download

(Ebook) Introduction to Plasmas and Plasma Dynamics. With Reviews of

Applications in Space Propulsion, Magnetic Fusion and Space Physics by
Tang, Haibin; York, Thomas M ISBN 9780128016619, 9780128018002,
0128016612, 0128018003
https://ebooknice.com/product/introduction-to-plasmas-and-plasma-
dynamics-with-reviews-of-applications-in-space-propulsion-magnetic-
fusion-and-space-physics-5432510

(Ebook) Biota Grow 2C gather 2C cook by Loucas, Jason; Viles, James

ISBN 9781459699816, 9781743365571, 9781925268492, 1459699815,
1743365578, 1925268497

https://ebooknice.com/product/biota-grow-2c-gather-2c-cook-6661374

(Ebook) The Interaction of High Power Lasers with Plasmas by S.

Ellezer ISBN 9780750307475, 0750307471

https://ebooknice.com/product/the-interaction-of-high-power-lasers-
with-plasmas-989404

(Ebook) Whistleblowing : In Defense of Proper Action by Wojciech W.

Gasparski; Wojciech W. Gasparski ISBN 9781412844239, 1412844231

https://ebooknice.com/product/whistleblowing-in-defense-of-proper-
action-51418522
(Ebook) The Interaction of High-Power Lasers with Plasmas by Shalom
Eliezer (Author) ISBN 9780429141362, 9780750307475, 9781420033380,
042914136X, 0750307471, 1420033387

https://ebooknice.com/product/the-interaction-of-high-power-lasers-
with-plasmas-12051636

(Ebook) Life in Fusion ISBN 9781608202379, 1608202372

https://ebooknice.com/product/life-in-fusion-7976088

(Ebook) Semiconductor Industry: Wafer Fab Exhaust Management by J.

Michael Sherer ISBN 9781574447200, 1574447203

https://ebooknice.com/product/semiconductor-industry-wafer-fab-
exhaust-management-1385750

(Ebook) Aspects of Anomalous Transport in Plasmas (Series in Plasma

Physics) by Radu Balescu ISBN 0750310308

https://ebooknice.com/product/aspects-of-anomalous-transport-in-
plasmas-series-in-plasma-physics-2169382

(Ebook) Kappa Distributions. Theory and Applications in Plasmas by

George Livadiotis (Eds.) ISBN 9780128046395

https://ebooknice.com/product/kappa-distributions-theory-and-
applications-in-plasmas-6614398
Power Exhaust in Fusion Plasmas 1st Edition Wojciech
Fundamenski Digital Instant Download
Author(s): Wojciech Fundamenski
ISBN(s): 9780521851718, 0521851718
Edition: 1
File Details: PDF, 5.02 MB
Year: 2009
Language: english
This page intentionally left blank
 POWER EXHAUST IN FUSION PLASMAS

Nuclear fusion research is entering a new phase, in which power exhaust will play
a vital role. This book presents a comprehensive and up-to-date summary of this
emerging field of research in fusion plasmas, focusing on the leading tokamak
concept.
Emphasis is placed on rigorous theoretical development, supplemented by
numerical simulations, which are used to explain and quantify a range of exper-
imental observations. The text offers a self-contained introduction to power
exhaust, and deals in detail with both edge plasma turbulence and edge local-
ized modes, providing the necessary background to understand these important,
yet complicated phenomena.
Combining an in-depth overview with an instructive development of concepts,
this is an invaluable resource for academic researchers and graduate students in
plasma physics.

W O J C I E C H F U N DA M E N S K I is the leader of the Exhaust Physics Task Force at

the Joint European Torus (JET), where he pursues research into edge plasma physics
and particle / power exhaust. He is also a Visiting Lecturer in Plasma Physics at
Imperial College, London, and a committee member of the Plasma Physics Group
of the Institute of Physics.
POWE R EXHAU S T IN FUSION
PLASMAS

WO J C I E C H F U N DA M E N S K I
Culham Science Centre
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore,
São Paulo, Delhi, Dubai, Tokyo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521851718

© W. Fundamenski 2010

This publication is in copyright. Subject to statutory exception and to the

provision of relevant collective licensing agreements, no reproduction of any part
may take place without the written permission of Cambridge University Press.
First published in print format 2009

ISBN-13 978-0-511-76979-5 eBook (NetLibrary)

ISBN-13 978-0-521-85171-8 Hardback

Cambridge University Press has no responsibility for the persistence or accuracy

of urls for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
dla Oli
 Contents

Preface page xi
1 Introduction 1
1.1 Fusion reactor operating criteria 2
1.2 Plasma stability limits on fusion reactor performance 6
1.3 Power exhaust limits on fusion reactor performance 8
1.4 Chapter summary 11
1.5 Units and notation 14
1.6 Further reading 15
2 Magnetized plasma physics 16
2.1 What is a plasma? 16
2.1.1 Plasma parameter 17
2.1.2 Magnetization parameter 18
2.2 Charged particle motion 19
2.2.1 Guiding centre drifts 20
2.2.2 Canonical (angle-action) variables 27
2.3 Kinetic description 33
2.3.1 Phase space conservation laws 34
2.3.2 Guiding centre kinetic theory 36
2.4 Fluid description 44
2.4.1 Co-ordinate space conservation laws 45
2.4.2 Guiding centre fluid theory 50
2.5 The relation between MHD- and drift-ordered dynamics 72
2.6 Further reading 73
3 Magnetized plasma equilibrium 74
3.1 Magnetic geometry and flux co-ordinates 75
3.2 Plasma current in MHD equilibrium 84
3.2.1 Hamada co-ordinates 86
3.2.2 Symmetry co-ordinates 88

vii
viii Contents

3.3 Large aspect ratio, toroidal equilibrium 92

3.3.1 General screw pinch 92
3.3.2 Cylindrical tokamak 95
3.3.3 Large aspect ratio (small ) tokamak 95
3.4 Further reading 100
4 Magnetized plasma stability 101
4.1 Hydrodynamic waves and instabilities 101
4.2 MHD waves and instabilities 107
4.2.1 Ideal MHD waves in a uniform plasma 107
4.2.2 MHD waves and instabilities in a stratified plasma 109
4.2.3 Ideal MHD waves and instabilities in a confined plasma 109
4.2.4 Ideal MHD waves and instabilities in a general screw pinch 115
4.2.5 Flute-reduced MHD 117
4.2.6 Non-homogeneous shear Alfvén waves 122
4.2.7 Current-driven ideal MHD instabilities: kink modes 123
4.2.8 Pressure-driven ideal MHD instabilities: ballooning modes 127
4.2.9 Resistive MHD instabilities: tearing modes 142
4.3 Drift-waves and instabilities 151
4.4 Kinetic waves and instabilities 157
4.5 Further reading 161
5 Collisional transport in magnetized plasmas 162
5.1 Collisional transport in a neutral gas 163
5.1.1 Maxwell–Boltzmann collision operator 163
5.1.2 Chapman–Enskog expansion 166
5.1.3 Fokker–Planck collision operator 170
5.2 Charged particle collisions in a plasma 172
5.2.1 Coulomb collision operator 172
5.2.2 Test particle dynamics in a plasma 178
5.2.3 Collisional momentum exchange 179
5.2.4 Collisional energy (heat) exchange 182
5.3 Collisional transport in a plasma 184
5.3.1 Collisional transport in an unmagnetized plasma 184
5.3.2 Collisional transport in a cylindrical plasma 188
5.3.3 Collisional transport in a toroidal plasma 200
5.4 Further reading 219
6 Turbulent transport in magnetized plasmas 220
6.1 Hydrodynamic turbulence 220
6.1.1 Transition to turbulence in hydrodynamics 222
6.1.2 HD turbulence in 3D 224
6.1.3 HD turbulence in 2D 239
 Contents ix

6.2 MHD turbulence 243

6.2.1 MHD turbulence in 3D 245
6.2.2 MHD turbulence in 2D 251
6.3 DHD turbulence 252
6.3.1 Drift-fluid turbulence 253
6.3.2 Gyro-fluid turbulence 275
6.3.3 Drift-kinetic and gyro-kinetic turbulence 280
6.4 Comparison of collisional and turbulent diffusivities 283
6.5 Further reading 285
7 Tokamak plasma boundary and power exhaust 286
7.1 The scrape-off layer (SOL) 287
7.1.1 Plasma–surface interactions 287
7.1.2 Plasma–neutral interactions 295
7.1.3 SOL geometry: limiter, divertor and ergodic SOL 300
7.1.4 SOL equilibrium, stability and transport 307
7.1.5 SOL modelling approaches 318
7.2 L-mode power exhaust: edge-SOL turbulence 322
7.2.1 Experimental observations 323
7.2.2 Numerical simulations 328
7.3 H-mode power exhaust: edge localized modes (ELMs) 353
7.3.1 Edge transport barrier 353
7.3.2 Power exhaust in between ELMs 367
7.3.3 Power exhaust during ELMs 376
7.3.4 Power exhaust control techniques 388
7.4 Further reading 394
8 Outlook: power exhaust in fusion reactors 395
8.1 ITER 395
8.2 DEMO 401
8.3 PROTO and beyond 403
8.4 Further reading 404
Appendix A Maxwellian distribution 405
Appendix B Curvilinear co-ordinates 407
References 410
Index 426
 Preface

Power exhaust, by which we mean the safe removal of power from a burning
plasma, is an essential requirement for the successful operation of any fusion
reactor. Specifically, plasma thermal energy must be conveyed across the first
wall without undue damage to plasma facing components (divertor and limiter
tiles) by heat load related plasma–surface interactions (ablation, melting, erosion).
Unlike other ‘technological’ problems related to fusion reactor design, e.g. tritium
retention in plasma facing materials, neutron damage to structural components or
non-inductive current drive, power exhaust is intimately linked to plasma confine-
ment and thus a perennial concern for any fusion reactor. While only a minor issue
in existing tokamaks, it will be critical for ITER (the next step plasma-burning
experiment) and even more so for DEMO (the demonstration fusion power plant).
Even non-burning, superconducting machines, such as EAST, KSTAR, JT60-SA,
W7-X, etc. will be forced to tackle this problem due to their long pulse capabilities.
This monograph is an attempt to draw a unified and up-to-date picture of power
exhaust in fusion plasmas, focusing primarily on the leading tokamak concept.
Emphasis is placed on rigorous theoretical development, supplemented by numer-
ical simulations when appropriate, which are then employed to explain and model
a range of experimental observations. The objective is not just to provide the reader
with a reliable map of the conquered territory and a guided tour over its many hills
and valleys,1 but also to supply him or her with the tools necessary to embark
on independent, and hopefully fruitful, journeys into the uncharted regions, the
white spaces on the map, la terra incognita. In this respect, the book is aimed both
at graduate students of magnetically confined plasmas and at researchers already
working in the field wishing to develop a deeper understanding of plasma exhaust
physics – a quickly emerging area of fusion research.

1 This function being well served by regular review articles appearing in topical journals.

xi
xii Preface

Broadly speaking, the text is organized into two parts. The first (Chapters 2
to 4) is dedicated to developing the theoretical framework necessary to describe the
equilibrium and stability properties of magnetically confined plasmas, the second
(Chapters 5 to 8) deals with plasma transport phenomena necessary to understand
power exhaust in real experiments. After a brief examination of charged parti-
cle motion, the two basic orderings of plasma dynamics (MHD and drift) are
introduced and the corresponding guiding centre kinetic and fluid equations are
derived. These are then used to investigate the equilibrium, stability and transport
properties of magnetically confined plasmas. Energy transport in the radial, dia-
magnetic and parallel directions due to collisional (classical and neoclassical) and
turbulent (drift-Alfvén and interchange) processes is examined with special empha-
sis on plasma turbulence in the boundary (edge) plasma and the scrape-off layer
(SOL). Next, the relevant experimental results from tokamaks and the modelling
approaches typically used to interpret these results are reviewed. Finally, the tools
developed hereto are applied collectively to study power exhaust in low and high
confinement regime plasmas in tokamaks, in particular to edge / SOL turbulence
and edge localized modes (ELMs).
The idea for this book originates with my early inroads into power exhaust on
JET and owes much to the difficulties I encountered in finding relevant material
in the topical literature. During this period I was, and indeed still am, fortu-
nate enough to benefit from the vibrant scientific environment of the Culham
Science Centre. I thus feel highly indebted to my many colleagues and friends
at JET and in other labs around the world, without whom this project would
certainly not have succeeded. In particular, I would like to thank A. Alonso,
P. Andrew, N. Asakura, M. Beurskens, J. Boedo, S. Brezinsek, D. Campbell,
C. S. Chang, A. Chankin, J. Connor, G. Corrigan, D. Coster, G. Counsell, T. Eich,
S. K. Erents, M. Fenstermacher, O. E. Garcia, B. Gonçalves, P. Helander, T. Hender,
C. Hidalgo, G. Huysmann, S. Jachmich, A. Kirk, S. Krashenninikov, A. Kukushkin,
B. LaBombard, B. Lipschultz, S. Lisgo, A. Loarte, G. F. Matthews, G. McCracken,
W. Morris, D. Moulton, V. Naulin, A. Nielsen, V. Philipps, R. A. Pitts, J. Rapp,
R. Schneider, B. Scott, S. Sipilä, P. C. Stangeby, M. Tokar, D. Tskhakaya,
S. Wiesen, M. Wischmeyer, G. S. Xu, R. Zagórski and S. Zweben. I would also
like to thank UKAEA, EFDA-JET, EPSRC and Imperial College, London, for their
support and the many research opportunities which they supplied.
 1
Introduction

‘Faced with something unusual our thought should not be “What

next?” but “Why?”. By answering the second of these questions we can
answer the first. And this, in brief, is the scientific method.’
Roger Scruton (c. 1990)

By definition, all exothermal reactors, including any fusion reactor one may envis-
age (tokamak, stellarator, etc.), produce both energy and spent reactants, or ash.1
In order for the reactor to operate in steady-state, (i) fresh fuel must be added at
the rate at which it is consumed, (ii) this fuel must be heated, ideally by the reac-
tions themselves, (iii) fuel must be confined, by whatever means are available, for
sufficiently long to allow the exothermic processes to continue, (iv) the energy and
ash must be removed from the system at the rate at which they are created, (v) the
impurities released from the reactor walls must likewise be removed at the rate at
which they are produced, and (vi) the reactor itself, primarily its walls, must not
be damaged by all the exhaust processes. Translating the above to a D–T burn-
ing tokamak, conditions (i)–(iii) may be labelled loosely as the ignition criteria,
and conditions (iv)–(vi) as the exhaust criteria. Taken together they constitute the
criteria of mutual compatibility between the burning plasma and first wall mate-
rials/components. Since the ignition criteria speak primarily to the central (core)
plasma, while the exhaust criteria refer to the boundary (edge) plasma, and since
the two regions are coupled by largely self-governing plasma transport processes,
it is the exhaust criteria which determine the optimum reactor performance for a
given reactor design. In the following, we introduce the basic concepts of fusion
reactor operation, including the stability and exhaust limits on reactor performance.

1 This applies both to chemical reactors, such as a candle or a steam engine, and nuclear reactors, such as a star
or a fission power plant. It is equally true for all fusion reactors, irrespective of whether the reacting fuel is
confined by gravity, as in the Sun, by magnetic fields and electric currents, as in a tokamak or stellarator, or by
the inertia of the ions themselves, as in the violent implosion of a hydrogen ice pellet after it is heated by lasers
or heavy ion beams.

1
2 Introduction

1.1 Fusion reactor operating criteria

Let us consider the ignition and exhaust criteria for a magnetically confined
fusion (MCF) reactor, operated either in steady-state or in successive pulse cycles;
although we restrict the discussion to MCF, most of the following remarks apply
equally well to inertially confined fusion (ICF). There are four fusion reactions of
interest for energy production (Krane, 1988),

D + T → He4 (3.52 MeV) + n (14.06 MeV), (1.1)

D + D → T (1.01 MeV) + p (3.03 MeV), (1.2)
D + D → He (0.82 MeV) + n (2.45 MeV),
3
(1.3)
D + He → He (3.67 MeV) + p (14.67 MeV),
3 4
(1.4)

where D and T represent the two isotopes of hydrogen: deuterium (D ≡ H2 ) and

tritium (T ≡ H3 ). In all four cases, the strong Coulomb repulsion of the positively
charged nuclei implies that the fusion cross-sections σ are only significant at ion
energies above 10 keV, e.g. at 100 keV, σ DT ∼ 5 barn, σ D D ∼ σ D H e3 ∼ 0.01 barn.
In thermonuclear fusion, the supra-thermal particles in the tail of the Maxwellian
distribution are responsible for most of the fusion reactions. Since the average reac-
tion rate σ v is largest for (1.1), especially for Ti < 100 keV, e.g. at Ti = 10 keV,
σ v DT ∼ 10−22 m3 /s, while σ v D D ∼ 10−24 m3 /s, a mixture of D and T is the
preferred fuel for future fusion reactors, including ITER and DEMO. At keV tem-
peratures, the atoms of hydrogen (for which the ionization potential is only 4 eV),
as well as those of most low and medium Z elements, become fully ionized and the
neutral gas mixture is transformed into an ion–electron plasma.
We now return to our six reactor criteria, the first two of which state that the
D–T fuel burned in reaction (1.1) must be replenished, criterion (i), and heated to
the operating reactor temperature, criterion (ii). In practice, (i) is achieved either by
gas puffing or ice pellet injection, although neither of these methods is capable of
delivering the fuel directly to the plasma core, i.e. the hot central region where the
thermonuclear burn is active; instead the fresh fuel is deposited (ablated/ionized) in
the edge plasma, and only reaches the core by a relatively slow diffusion process.
In contrast, the steady flow of power required by (ii) is delivered directly to the core
plasma either by external heating, e.g. by neutral beams or radio waves resonant
with the gyration frequencies of ions and electrons, or by the charged fusion prod-
ucts, such as the 3.5 MeV alpha particle in (1.1) or the 14.7 MeV proton in (1.4),
which are trapped by the magnetic fields.2

2 Neutrons released in fusion reactions do not interact with, and thus cannot heat, the plasma. The same is true
for photons released as bremsstrahlung and synchrotron radiation.
 1.1 Fusion reactor operating criteria 3

Particle and energy confinement of a thermonuclear plasma, criterion (iii), have

been the central focus of MCF research over the past 50 years. To appreciate the
difficulties posed by this task, recall that energy break-even,
Q DT ≡ PDT /Pheat > 1, Q α ≡ Pα /Pheat = 0.2Q DT , (1.5)
where Q DT is the energy multiplication factor, defined as the ratio of fusion and
auxiliary heating powers, was only approached recently. Since 80% of the energy
released in (1.1) appears as the kinetic energy of neutrons and is thus promptly lost
from the plasma, the fusion reactions can only be self-sustaining when Q α  1.
Note that PDT and Pα may be evaluated as

E DT
PDT = 5Pα = n 2 σ v DT dx, E DT = 5Eα = 17.58 MeV (1.6)
4
where E DT is the energy released per fusion reaction, n =  n D + n T is the particle
density, σ v DT is the fusion reaction cross-section and dx is a volume integral
over the plasma.
The slow progress towards Q α > 1 can be ultimately traced to one of the great
unsolved problems of classical physics, namely fluid turbulence. Indeed, much of
the success of MCF can be ascribed to the basic dimensional scaling: volume/area
∼ size, and thus to the building of ever bigger, and more expensive, devices, specif-
ically the toroidal, axis-symmetric, inductively driven tokamaks, see Fig. 1.1. It is
thus no coincidence that Q DT ∼ 1 was finally approached in the largest present day

Inner poloidal field coils

Poloidal magnetic field
Outer poloidal field coils

Resulting helical magnetic field Toroidal field coils

Plasma electric current Toroidal magnetic field

Fig. 1.1. Schematic representation of a tokamak. 

c EFDA-JET.
4 Introduction

tokamak, namely the Joint European Torus (JET), with Q α > 1 expected in ITER.
That future event may be viewed as the watershed beyond which issues related
to energy confinement, criterion (iii), will be increasingly overshadowed by those
related to power exhaust, criteria (iv)–(vi). This tendency, which first emerged in
the technological challenges encountered during the conceptual, and later engi-
neering, design of ITER, is also evident in a new generation of super-conducting,
actively cooled machines.
Since both fusion power and ash (He4 ) are generated in the plasma centre, their
exhaust mechanisms are partly related. Thus, power is removed from an MCF
plasma by three channels: (a) by neutrons released in the fusion reaction itself; (b)
by photons emitted during bremsstrahlung, synchrotron and (hydrogenic or impu-
rity) line radiation; and (c) by kinetic energy of the ions and electrons which are
transported across the magnetic field largely by turbulent plasma motions; in con-
trast, fusion ash is removed from the core plasma only by turbulent advection.
Assuming that bremsstrahlung is the dominant mode of radiation in the hot plasma
core, we may approximate the steady-state power balance for a burning fusion
plasma as
Pheat + Pα = (1 + Q α )Pheat ≈ Pbr + Ptr = Ploss , (1.7)
where the left-hand side represents the auxiliary (Pheat ) and α particle (Pα ) heating,
and the right-hand side the total losses due to bremsstrahlung (Pbr ) and plasma
transport by convection and conduction (Ptr ),
 
Pbr = αbr n T dx,
2 1/2
W = 3 nT dx ≡ Ptr τ E ≡ Ploss τ E∗ , (1.8)

where n is the particle density, T is the plasma temperature and τ E and τ E∗ are the
thermal energy confinement times associated with plasma transport and transport
+ radiation, respectively. Hence Q α may be estimated as
 −1
Pα Ploss Pα
Qα ≡ = −1 = . (1.9)
Pheat Pα Ploss − Pα
The limit of vanishing heating power, which is equivalent to Q α → ∞, can thus
be expressed as Pα = Ploss . Substituting from (1.6) and (1.8), we obtain the
corresponding ignition criterion,
12T
Pα ≥ Ploss ⇒ nτ E > > 1.5 × 1020 m−3 s, (1.10)
E DT σ v DT
where n and T represent volume-average values and the final expression represents
a minimum value near T ≈ 30 keV. In the keV temperature range, one finds that
σ v DT ∝ T 2 and the above result simplifies further to
nT τ E = pτ E > 3 × 1021 m−3 keV s ≈ 5 bar · s. (1.11)
 1.1 Fusion reactor operating criteria 5

The numerical values in (1.10) and (1.11) assume flat radial profiles of n and T ;
for peaked profiles, these values are somewhat higher.
Let us next consider a corresponding condition for a fusion reactor, in which
all power leaving the plasma is converted to electricity with an efficiency ηe and
then used to heat the plasma with efficiency ηh . Defining η = ηe ηh , for which one
expects values in the range 0.2–0.4, the requirement for net energy production may
be written as,
η(P f us + Ploss ) > Ploss , P f us = PDT + PLi , (1.12)
where the additional power PLi refers to the energy released in the breeder blanket
by the reaction (1.14), see below. Substituting from (1.6) and (1.8) leads to the
celebrated Lawson’s criterion (Lawson, 1957),
 −1
η E DT
nτ E > 3T σ v DT − αbr T 1/2
∼ 3 × 1019 m−3 s, (1.13)
1−η 4
where the final expression was evaluated near T ≈ 30 keV and η = 1/3. The
plasma ignition criterion (1.10) is equivalent to (1.13) with η = 0.136.3
Power exhaust channels (a)–(c) lead to three different types of heat loads on the
first wall and require three different power removal systems: (a) the neutron energy
is deposited volumetrically in a neutron-absorbing envelope surrounding the first
wall, ideally a breeder blanket, employing the reactions,
Li6 + n → T + He4 + 4.8 MeV, (1.14)
Li + n → T + He + n − 2.5 MeV,
7 4
(1.15)
to breed tritium fuel from solid lithium, (b) the photon energy generates a fairly
uniform surface heat load on first wall components, and (c) the plasma thermal
energy is convected and conducted along the magnetic field lines to dedicated heat
load bearing tiles. In each case, the power deposited on, or absorbed in, the vessel
wall must be removed by an active coolant loop. Moreover, the effective heat load
must not exceed some limit imposed by thermo-mechanical constraints. This in
turn limits the energy flow crossing the outer boundary of the plasma in each of the
three channels, i.e.
P⊥σ /A p ≡ q⊥σ < q⊥σ
exh
, σ ∈ {n, γ , tr }, (1.16)
where A p is the plasma area. In practice, the last of these conditions imposes
the most severe constraints on plasma operation, e.g. for ITER, the time-averaged
power loads on plasma facing components (PFCs) are limited to ∼10 MW/m2 and

3 To demonstrate this, it suffices to insert P = P

α loss in (1.12), which yields η = Pα /(Pα + PDT + PLi ) =
3.52/(3.52 + 17.58 + 4.8) = 0.136.
6 Introduction

transient energy loads to ∼0.5 MJ/m2 in ∼250 µs.4 Consequently, in the rest of the
book we will focus on channel (c) above, i.e. the exhaust of fusion energy (and
to a lesser extent, of fusion ash) by plasma transport processes. As motivation for
this investigation, we first compare the limits on fusion reactor performance, which
for simplicity we assume to be a tokamak, imposed by plasma stability and power
exhaust.

1.2 Plasma stability limits on fusion reactor performance

Let us first assess the limits imposed by global (MHD, magneto-hydrodynamic)
plasma stability requirements, which will be derived in Section 4.2 and summa-
rized in Table 4.1.5 Expressing the fusion power density in terms of the toroidal
beta,
PDT /V = 14 E DT n 2 σ v DT a ∝  p 2 a ∝ βT2 B04 , (1.17)
where βT is given by (3.19) and ·a is the average over the plasma volume,
V = a dx, and noting that the toroidal magnetic field (on axis) is limited by tech-
nological constraints to roughly B0max ∼ 5−10 T, we find that the MHD pressure
limits determine the maximum fusion power density and hence the reactor cost.6
In order for the burning plasma equilibrium to be MHD stable, the MHD beta limit
βTmhd , as given in Table 4.1, must exceed the minimum beta required for ignition
ign
βT , which may be inferred from (1.11).
βT ∝ B0−2 τ E−1 ,
ign
βTmhd ∝ a κa /q0 qa . (1.18)
Here we defined the inverse aspect ratio, a = a/R0 , where a and R0 are the minor
and major radii of the torus, the elongation κa = A p /πa 2 where A p is the cross-
sectional area of the plasma, and the safety factors on axis (r = 0) and at the edge
of the plasma (r = a), q0 and qa , given by (3.18).
On the other hand, βT < βTmhd amounts to inefficient use of the ‘expensive’
toroidal magnetic field, which is optimally used only for βT ≈ βTmhd . Hence, the
ign
condition βT ≈ βTmhd determines the size aign of the smallest reactor able to ignite
for given field B0 , inverse aspect ratio a , elongation κa , etc.
To evaluate aign , we need to estimate the energy confinement time τ E (1.8), e.g.
we may assume that radial transport is purely diffusive, so that τ E ≈ a 2 /χ⊥ , where

4 This value should not be confused with q exh , which refers to the power flux crossing the last closed flux surface,
⊥tr
see Chapter 7.
5 Here we will anticipate some of the definitions which will be made formally in Chapters 2 to 4.
6 Since the reactor capital cost is roughly proportional to the plasma volume V, being driven largely by the cost
associated with super-conducting poloidal coils, one finds that the cost of electricity it generates scales inversely
with PDT /V, i.e. an economical reactor should be as small as possible to generate a desired power output in
MWe. Hence, the power density (1.17) may be interpreted as the financial figure of merit for a fusion reactor.
 1.2 Plasma stability limits on fusion reactor performance 7

χ⊥ is the average radial heat diffusivity.7 Anticipating the results of Chapter 6, we

write down the generic expression,
√
B0 τ E ∝ (qa ρ∗ )−x ∝ (qa ρti /a)−x ∝ (qa T /a B0 )−x , (1.19)

which states that B0 τ E scales inversely with the product of the safety factor qa ,
or q∗ (4.110), and the normalized, toroidal gyro-radius ρ ∗ ≡ ρti /a; here x = 3
corresponds to the gyro-Bohm scaling and x = 2 to the Bohm scaling.
ign
Inserting (1.19) into (1.11), one finds a scaling of βT with reactor size,
√
βT ∝ [qa T /a B0 ]x /B0 ∝ a −x B0−(x+1) T x/2 qax .
ign
(1.20)
ign
Since, the minimum beta needed for ignition, βT , decreases sharply with size,
whereas the maximum beta imposed by MHD stability, βTmhd , is size independent,
ign
we find that ignition is always possible for large enough plasmas.8 Equating βT ,
(1.20), and βTmhd , (1.18), gives the minimum ignition radius,

aign ∝ (qa /B0 )1+1/x T 1/2 (q0 /εa κa )1/x , (1.21)

−3/2 −4/3
which decreases with toroidal field as B0 (x = 2) and B0 (x = 3).
The plasma volume corresponding to (1.21) is found to scale as

κa /εa ∝ (qa /B0 )3(1+1/x) T 3/2 q0 εa−(1+3/x) κa1−3/x .

3/x
Vign ∝ aign
3
(1.22)

Since Vign increases with qa , q0 and T , and decreases with B0 , a and κa (although
the κa dependence vanishes for x = 3), we would like to minimize (maximize)
the former (latter) parameters. This can be done by (i) fixing q0 ≈ q0mhd ≈ 1 and
qa ≈ qamhd ≈ 2 − 3 at their MHD stability limits, (ii) choosing T ∼ 10−30 keV,
which, although below the maximum of the fusion cross-section, σ v DT , min-
imizes (1.10), and (iii) setting the axial field at B0max ∼ 5 T. Moreover, (1.22)
strongly favours small aspect ratios (a ∼ 1) and weakly favours elongated
plasma shapes (κa > 1), provided x < 3. The upper limit on a and lower
limit on R0 are imposed by the requirement for a neutron heat shield on the
inner solenoid;9 the upper limit on κa is imposed by an axis-symmetric (n = 0)
vertical displacement MHD instability, which becomes increasingly acute for elon-
gated poloidal cross-sections. For instance, for ITER, whose aim is to achieve
Q DT = 10, or Q α = 2, and produce 500 MW of fusion power, the above

7 This scaling expresses the easily verified fact that larger objects take longer too cool.
8 However, as will be shown in Section 7.1, power exhaust considerations impose an upper limit on a cost
effective reactor, i.e. one with βT ≈ βTmhd .
9 Some designs dispense with this requirement by envisioning a replaceable central column, thus allowing a
much smaller R0 and thus a larger inverse aspect ratio a .
8 Introduction

parameters were carefully optimized with respect to cost and performance to

yield

a = 2.0 m, B0 = 5.3 T, qa = 3.0, κa = 1.7, a = 0.33, (1.23)

where qa = 3 represents a plasma current of IT = 15 MA, see Fig. 8.1.

1.3 Power exhaust limits on fusion reactor performance

The limits on fusion reactor performance imposed by plasma stability, as derived
above, should be compared with those imposed by power exhaust (1.16). Let us
assume that the radial energy flow at the last closed flux surface (LCFS) is limited
to some value q⊥exh , which can be written as

Ptr /A p < q⊥ ,
exh
Ap = dS⊥ ≈ 4π 2 a R0 κa ≈ 4π 2 a 2 κa /a . (1.24)
LC F S

Defining 0 < f α ≡ Q α /(1 + Q α ) < 1, it follows from (1.8) that

Pα Eα n 2 σ v DT a  p 2 a τ E∗
fα = = ∗ ∝ ∝  pa τ E∗ ∝ βT B02 τ E∗ . (1.25)
Ploss 12 pa /τ E  pa
We next consider the low radiation limit (Pbr  Ptr ) for which τ E∗ ≈ τ E and
Ploss ≈ Ptr . In this case, we can eliminate τ E in Ptr using (1.25),

Ptr /V ∝ βT B02 /τ E ∝ βT2 B04 / f α , Pα /V ∝ βT2 B04 . (1.26)

Dividing (1.24) by (1.26) introduces the volume to surface ratio V/A p ≈ a/2,
which increases linearly with size. This imposes an upper, power exhaust limit
aexh on the reactor size for given values of q⊥exh , plasma pressure p ∝ βT B02 , energy
confinement τ E and/or level of ignition f α , see Fig. 1.2,
 2
a < aexh ∝ q⊥exh τ E /βT B02 ∝ q⊥exh f α / βT B02 . (1.27)

For ignited ( f α = 1) and marginally MHD stable βT ≈ βTmhd plasmas,

aign ≤ a ≤ aexh
mhd
∝ q⊥exh (q0 q∗ /a κa )2 B0−4 ∝ q⊥exh (qa /a κa )2 B0−4 , (1.28)

where aign is given by (1.21) and we set q0 = 1 and q∗ = qa . When q⊥exh is

ign  2
sufficiently large, i.e. when q⊥exh > q⊥ ∝ aign βTmhd B0 , the maximum βT power
exhaust radius exceeds the ignition radius, aexh mhd
> aign , and (1.24) is satisfied
automatically in the range of minor radii given by (1.28). This range of optimal a
is reduced as q⊥exh decreases, eventually prohibiting fusion burn at MHD marginal
ign  mhd
stability when q⊥ < q⊥ aexh < aign .
exh
 1.3 Power exhaust limits on fusion reactor performance 9
1.4 beta MHD
beta ign Bohm
beta ign gyro-Bohm
1.2 beta exh

1.0

0.8

0.6

0.4

0.2

JG08.360-2c
0
0.5 1.0 1.5 2.0 2.5 3.0
Normalized miror radius, a/aign

Fig. 1.2. Beta limits (ignition, stability and exhaust) vs. reactor size.

To estimate the minimum ignition radius in that case, we equate the minimum
ign
beta needed for ignition βT (1.20), with the maximum beta allowed by the power
exhaust limit, which follows from (1.27) with f α = 1,
 1/2
βTexh ∝ q⊥exh /a B04 . (1.29)
This yields an estimate of the power exhaust limited ignition radius,
 −y/2 x y x y/2 (1−x)y
exh
aign ∝ q⊥exh qa T B0 , y = 1/(x − 1/2). (1.30)

Since βT ∝ a −x decays faster than βTexh ∝ a −1/2 , it is possible to achieve ignition

ign

for any value of q⊥exh , by increasing the size of the reactor; this is reflected in the
 −y/2
weak, inverse scaling aignexh
∝ q⊥exh , with the exponent being equal to −1/3
for x = 2 and −1/5 for x = 3. In short, power exhaust imposes the minimum
ign
reactor size only when q⊥exh < q⊥ ,

amin = max aign , aign
exh
. (1.31)
 1/2
The corresponding toroidal beta, βTmin ∝ q⊥exh /amin B02 , is smaller than βTmhd
thus reducing the fusion power density and the cost effectiveness of the reactor.
The determination of q⊥exh as a function of plasma and field quantities is the chief
task of both experimental and numerical power exhaust studies. It is also one of the
main incentives for investigating transport processes in the plasma boundary and
the ultimate goal of the theoretical development, and the accompanying discussion,
in the rest of the book.
10 Introduction

Let us summarize the above findings. Since the exhaust limits provide the bound-
ary conditions for the plasma thermodynamic quantities, they effectively determine
the maximum achievable fusion gain, Q α , for a given reactor design (RD), by
which we mean a set of hardware including magnetic coils, heating, fuelling and
current drive systems, vacuum vessel and mechanical support, cooling and pump-
ing systems, and last, but not least, the plasma facing components (PFCs), i.e. the
first wall armour against plasma fluxes. This relation may be expressed as
Q α = Q α (P S, R D), α (R D) = max[Q α (P S, R D)|P S],
Q max (1.32)
where plasma scenario (PS) refers to a combination of plasma shape, magnetic
field, current profile, heating and fuelling methods, etc., i.e. to the way in which
the given reactor design is utilized within each plasma discharge. Thus, the issue
of compatibility or integration between the ignition and exhaust criteria, and specif-
ically between plasma scenarios and PFCs, is really one of optimization of the PS
for a given RD with respect to Q α .10
Since the fusion power density is roughly proportional to the square of the cen-
tral fuel ion plasma pressure, (1.6), while the plasma density is limited to roughly
the Greenwald density, see (7.50), this optimization amounts to maximizing the
ion temperature, Ti , and minimizing the effective charge, Z e f f , in the centre of the
plasma column. In the absence of internal transport barriers, e.g. in the inductive or
baseline tokamak plasma scenario, the ion temperature gradient (ITG) is set by the
threshold for the ITG drift-wave turbulence (Garbet and Waltz, 1998). Hence, the
central ion temperature is a linear function of the edge, or pedestal, temperature,
T ped , e.g. in ITER, it is predicted that T ped ∼ 4 keV is necessary to achieve the
desired fusion gain factor, Q α ∼ 2 (Doyle et al., 2007).11 The impact of any given
PFC limit on the reactor performance can then be quantified as

α (P FC)/Q α (∞),
ζ (P FC) = 1 − Q max max
(1.33)

α (P FC) is the maximum fusion gain factor for a specified PFC limit,
where Q max
α (∞) the same factor without any limit
i.e. (1.32) with PFC in place of RD, and Q max
on PFC plasma loads, or some previously chosen reference limit value. One can
recast (1.33) in terms of density and energy confinement degradation by estimating
Q α ∝ pτ E ∝ ( f GW H98 )z , with z ∼ 2−3,
ζ (P FC) = 1 − [ f GW (P FC)/ f GW (∞)]z [H98 (P FC)/H98 (∞)]z , (1.34)

10 It is worth noting that the very terms ‘compatibility’ and ‘integration’ reflect the historical disconnection
between the tasks of investigating, on the one hand, the plasma equilibrium, stability and transport, and, on
the other, its particle and power exhaust properties. Such a disconnection is of course absent in a real plasma
where the core and edge regions form an integrated whole.
11 Whether such high edge plasma temperatures are compatible with the desired lifetime of the divertor and
limiter PFCs remains a matter of active research.
 1.4 Chapter summary 11

where f GW = na /n GW is average density normalized by n GW and H98 =

98,y
τ E /τ E is the energy confinement time normalized by the ITER98(y,2) scaling;
the values required by the ITER reference scenario are f GW ∼ 0.85 and H98 ∼ 1.
Similar impact factors could be obtained for other limits, e.g. tritium retention, dust
inventory, etc. Finally, an alternative measure of reactor performance, e.g. price of
electricity, neutron fluence, bootstrap fraction, etc., could be used instead of Q α to
defined the impact factor.
In short, terms such as ‘integration’ and ‘compatibility’ of plasma scenarios and
PFCs are not binary signifiers, but qualitative ones, and refer to the optimization
of the former subject to the latter, thus determining the maximum Q α for a given
reactor design. Put another way, PFC limits carry a certain ‘cost’ in terms of the
reduction in Q α with respect to its unconstrained value, i.e. if arbitrarily high heat
loads on divertor and main chamber PFCs were permitted. It will be our task in the
following chapters to develop the tools necessary to translate the limits on plasma
heat loads onto PFCs into limits on plasma flows across the last closed flux surface.

1.4 Chapter summary

The remainder of this book is divided into seven chapters, six of comparable
lengths, the last somewhat shorter. Their content is briefly summarized below.
Chapter 2 represents a self-contained and rigorous introduction to the theory
of magnetized plasma physics. Starting from general considerations of collec-
tive behaviour, Section 2.1, the basic plasma parameters are introduced. Next,
Section 2.2, the motion of charged particles in both uniform and non-uniform,
constant and time-varying magnetic fields is discussed. The notion of the guiding
centre and its drift are formulated, as are the two fundamental orderings (MHD and
drift) of magnetized plasma dynamics. Angle-action variables and adiabatic invari-
ants for a moving charged particle are derived by introducing and exploiting the
Lagrange–Hamilton formalism of classical mechanics. The ensemble of charged
particles is then studied in the six-dimensional phase space using the techniques
of non-equilibrium statistical mechanics, Section 2.3, culminating in the derivation
of the three (MHD, drift and gyro) guiding centre kinetic theories. In Section 2.4,
the kinetic equations are integrated over velocity space to yield corresponding fluid
equations, in particular the MHD and DHD models.
Chapter 3 addresses the topic of magnetized plasma equilibrium. Beginning
with general considerations, it is shown that toroidal topology and the existence of
magnetic flux surfaces is a natural requirement for effective plasma confinement,
Section 3.1. MHD equilibrium of the plasma is then discussed in magnetic flux
co-ordinates, the Grad–Shafranov equation for axis-symmetric equilibria is derived
and the origin of the parallel return current in the quasi-neutrality constraint,
12 Introduction

∇ · J = 0, is demonstrated, Section 3.2. Finally, the above formalism is applied

to the somewhat idealized case of a large aspect ratio, circular torus, Section 3.3.
Chapter 4 investigates confined plasma stability, or rather the various waves
to which the plasma can act as the medium and the various instabilities into
which these may develop. Starting with a discussion of Rayleigh–Taylor and
Kelvin–Helmholtz instabilities in hydrodynamics, Section 4.1, the bulk of the chap-
ter, Section 4.2, is devoted to ideal MHD waves in a magnetized plasma and
instabilities in a magnetically confined plasma. The latter are divided into ideal
MHD instabilities, specifically the current-driven (kink) modes and pressure-driven
(interchange and ballooning) modes, and resistive MHD instabilities known as
tearing modes. The various limits on the current and pressure within the plasma
imposed by MHD instabilities are discussed. The final two sections briefly intro-
duce plasma waves and instabilities in the drift ordering, Section 4.3, and in the
kinetic approximation, Section 4.4, both of which play an important role in the
micro-instabilities which drive plasma turbulence.
Chapter 5 deals with diffusive transport of mass, momentum and energy due
to binary Coulomb collisions between charged particles which make up a plasma.
Section 5.1 introduces the basic concepts of kinetic theory of gases by deriving the
Maxwell–Boltzmann and Fokker–Planck collision operators for binary collisions
between neutral particles. These concepts are then extended to binary Coulomb
collisions between the ions and electrons, which are characterized by a cubic
dependence of the collision rates on the relative particle velocity, Section 5.2.
Finally, the resulting transport processes in unmagnetized and magnetized plas-
mas are discussed in Section 5.3. The latter are divided according to the plasma
geometry into collisional transport in cylindrical plasmas, known as classical trans-
port, Section 5.3.2, and in toroidal plasmas, known as neoclassical transport,
Section 5.3.3.
Chapter 6 turns our attention to cross-field transport of mass, momentum and
energy due to turbulent advection. As before, it begins with an exposition of tur-
bulent flows in hydrodynamics, Section 6.1, including transition from laminar to
turbulent state and separate treatments of turbulent flows in two and three dimen-
sions. The basic concepts and theories of turbulent flow analysis are presented: the
mean field description and the Reynolds stress, the spectral cascades of energy and
enstrophy, the Kolmogorov spectrum and its non-Gaussian variants intermittency
modifications, and the statistical description of turbulent flows in terms of proba-
bility distribution functions of fluctuating quantities. Special emphasis is placed on
the inverse energy cascade and its relation in increased intermittency. These con-
cepts are then extended to magneto-hydrodynamics (MHD), where the discussion
is again divided into two- and three-dimensional MHD turbulence, Section 6.2.
Finally, the same conceptual framework is applied to drift-hydrodynamics (DHD),
 1.4 Chapter summary 13

Section 6.3, with the bulk of the discussion dedicated to plasma turbulence in the
drift-fluid (DF) and gyro-fluid (GF) approximations, specifically the drift-Alfvén
and interchange dynamics; the extension to drift-kinetic (DK) and gyro-kinetic
(GK) approximations is touched upon briefly.
Chapter 7 brings the thread of our argument home to our starting point, and
applies all the newly developed formalism to the problem of power exhaust in
fusion plasmas. To this end, the plasma boundary of magnetically confined plasma,
the so-called scrape-off layer (SOL), is first described in Section 7.1, including its
equilibrium, stability and transport properties. The difference, in all these aspects,
between the closed field line, or the edge region, and the open field line, or scrape-
off layer region, are outlined in some detail. This is supplemented by a discussion
of plasma–surface and plasma–neutral interactions, both vital to plasma dynam-
ics in the SOL. The various SOL geometries are compared, with an emphasis on
the divertor SOL and the phenomenon of divertor plasma detachment, and lead-
ing edge plasma modelling approaches are compared. In Section 7.2, edge-SOL
transport in low confinement (L-mode) plasmas is analyzed. The most recent and
relevant experimental observations are reviewed, including observation of radial
propagation of plasma filaments (blobs), and are then broadly explained using
numerical simulations of edge-SOL turbulence with collisional, (neo)classical dis-
sipation. In Section 7.3, power exhaust in high confinement (H-mode) plasmas is
addressed. Starting with a discussion of edge transport barrier dynamics and its
quasi-periodic relaxation by edge localized modes (ELMs), power exhaust during
and in between ELMs is treated separately. The onset of Type-I ELMs is shown to
depend on the combination of pressure (ballooning) and current (peeling) driven
modes, while that of Type-III ELMs to depend on some other non-ideal, resistive
MHD or DHD, instability. Once again, a combination of collisional and turbu-
lent effects is evident in the inter-ELM phase of the ELM, while ELM exhaust
is characterized by plasma filamentary structures. The fraction of ELM energy
reaching the main chamber wall is shown to depend on the combination of radial
filament motion (dominated by interchange drive) and parallel losses (initially
kinetic free-streaming, later fluid convection) to the divertor targets. The techniques
for controlling inter-ELM and ELM power exhaust are then reviewed, including
extrinsic impurity seeding, pellet ELM pacing and resonant magnetic perturbation.
Finally, in Chapter 8, the book concludes with a discussion and power exhaust
issues in future fusion reactors, which are divided into the next step burning plasma
experiment, ITER, and the demonstration power plant, DEMO. To illustrate the
issues and apply the physics basis developed in Chapters 1–7, power balance in
ITER is analyzed in detail in Section 8.1. The challenges likely to be faced in
the design and construction of DEMO are outlined in Section 8.2, including the
complete avoidance of ELMs and disruptions, and the need for radiative fractions
14 Introduction

approaching unity. On the positive note, the opportunities offered by the prospect
of advanced fusion fuel cycles and direct energy conversion are highlighted.
Each chapter concludes with a section dedicated to further reading, containing
a brief selection of recommended texts related to the material discussed in the
chapter. Finally, the book is supplemented by two appendices dealing with the
Maxwellian distribution and curvilinear co-ordinates.

1.5 Units and notation

Unless otherwise stated, S I (mks) units are used throughout the text, with the
exception of temperatures, which are expressed in eV . This choice seems better
suited to many practical problems, in which comparison with experimental data
is essential. To convert any SI (mks) expression to Gaussian (cgs) units, simply
replace 0 by 1/4π, μ0 by 4π/c2 and B by B/c. Readers wishing to operate with
other unit system may wish to consult the conversion tables in the appendix of
Jackson (1975).
A few comments on mathemetical notation are in order. For the sake of brevity,
the following shorthand is introduced for differential operators,
∂ ∂
∂t ≡ , ∂x ≡ ≡ ∇, (1.35)
∂t ∂x
or more generally for any scalar η and vector z,
d ∂ ∂
dη ≡ , ∂η ≡ , ∂z ≡ ≡ ∇z . (1.36)
dη ∂η ∂z
This notation has the advantage of treating the derivative as an operator, such that
higher-order derivatives can be denoted by algebraic powers,
dk ∂k ∂k
dkη ≡ , ∂ηk ≡ , ∂zk ≡ ≡ ∇zk . (1.37)
dk η ∂kη ∂kz
The traditional notation will only be used when the denominator contains sub-
scripts, e.g. dψT /dψ P rather than dψ P ψT . Otherwise, the shorthand notation will
be preferred in both equations and in-line expressions. Finally, special symbols
are used for exact derivatives with respect to time (Ȧ ≡ dt A) and flux surface
radius (A ≡ dA/dr f ), which becomes A ≡ dψ A in symmetry co-ordinates and
A ≡ dr A in small  tokamak co-ordinates.
The symbol ⊥ is used in the literature to denote both the direction perpendic-
ular to the magnetic field and that normal to the magnetic flux surface; the two
conventions are frequently used side-by-side, with the meaning inferred from the
context. To prevent confusion, we use bold font, ⊥, when referring to the earlier,
 1.6 Further reading 15

gyro-tropic convention, and regular font, ⊥, when referring to the latter, natural
magnetic co-ordinates convention.
To avoid confusion with co-variant components of a vector, denoted by lower
case subscripts, e.g. Bζ , Bθ for co-variant components of B, we employ upper case
subscripts in the sense of the usual convention in physics, i.e. to denote simple
vector components of B. Hence, in cylindrical geometry, BT and B P represent
the axial and poloidal components, while in toroidal geometry they represent the
toroidal and poloidal components.
Finally, two notations are used side-by-side for the thermal velocity, namely,
vts = Ts /m s and vT2 s = 2Ts /m s , differing by the factor of 2 appearing in the
2

nominator.

1.6 Further reading

The study of fusion plasmas requires frequent contact with other fields of theoreti-
cal physics, applied mathematics and engineering, for which a list of good sources
is useful. When in doubt on any of these topics, the reader may wish to consult the
following references.
Classical mechanics: Goldstein (1980); Arnold (1989). Quantum mechanics:
Cohen-Tannoudji et al. (1977). Thermodynamics and statistical physics: Reif
(1965); Callen (1985); Huang (1987). Electrodynamics: Jackson (1975). Nuclear
physics: Krane (1988). Atomic physics: Haken and Wolf (1987); Cohen-Tannoudji
et al. (1992). Solid state physics: Kittel (1995). Chemistry: Pauling (1960, 1970).
Astrophysics: Padmanabhan (2000). All other aspects of theoretical physics:
Landau and Lifschitz (1960); Feynmann et al. (1963). Applied mathematics: Morse
and Feschbach (1953); Smirnov (1964); Boas (1983); Ross (1985); Zwillinger
(1989); Arfken and Weber (2001). Materials: Budynas (1977); Popov (1978);
Ashby and Jones (2005). Electric circuits: Dorf (1989). Thermal circuits: Wood
(1982). Control systems: Van de Vegte (1990). Plasma diagnostics: Hutchin-
son (2002). Nuclear engineering: Benedict et al. (1981); Lamarsh (1983). MCF:
Miyamoto (1989); Goldston and Rutherford (1995); Wesson (2004). ICF: Motz
(1979); Hora (1981); Kruer (1988).
 2
Magnetized plasma physics

‘Physics and other natural sciences are successful because physical

phenomena associated with each range of energy and other parameters
are explainable to a good, if not perfect, accuracy by an appropriate,
self-consistent theory.’
E. H. Lieb (c. 1990)

The physics of plasmas is no exception. We thus begin our journey by introduc-

ing the basic concepts, key results and essential theories pertaining to magnetized
plasmas, which will underlie the more specialized material found in subsequent
chapters. Our discussion will follow the excellent texts by Hazeltine and Meiss
(1992) and Hazeltine and Waelbroeck (2004).

2.1 What is a plasma?

The ancient-Greek word πλασ μα (plasma) originates with the verb πλασ σ ω
(plasso): to shape, mould, and signifies a pliant, malleable object (hence plas-
tic, plasticine, etc.). The term plasma entered the physics lexicon when used by
Tonks and Langmuir to describe the macroscopic behaviour of a glow discharge.
More generally, it may be used to denote any multi-constituent system dominated
by collective (as opposed to binary) interactions. This information-based, systems
theory definition includes the traditional electromagnetic plasma of ions and elec-
trons, the gravitational plasma of stars in a galaxy or galaxies in a cluster, and
the quark–gluon plasma of elementary particles in a heavy ion collision. It may
also be extended beyond the domain of physics, to such fields as finance, semi-
otics, sociology and cognitive science. One may thus speak of a market plasma
in which commodity prices interact via market player transactions, a semantic
plasma in which word or sign meanings interact via syntactical references in their
respective languages, or a neural plasma in which neurons are inter-linked by mul-
tiple pathways in a network. In each case, the trajectory of a single constituent is

16
 2.1 What is a plasma? 17

dominated by the combined effect of many simultaneous interactions, rather than

by a series of discrete encounters. This defining characteristic may be quantified
in terms of the plasma parameter, , which measures the average number of con-
stituents creating the effective force field acting on any given member. A system is
said to be in a plasma state when   1, which applies to 99% of the universe, as
well as to consciousness, language and society.

2.1.1 Plasma parameter

While the systems theory approach provides valuable insights into the nature of
collective behaviour, it is too broad for the purpose of this book. Henceforth, we
will restrict the term plasma to mean a gas of ions and electrons satisfying the
electromagnetic version of the plasma criterion,  p  1. We expect this criterion
to hold due to the long-range effect of the Coulomb potential, ϕ(r ) ∝ r −1 . To
obtain  p , we begin by noting that, barring relativistic and quantum corrections,
an ion–electron gas may be described by Newton’s and Maxwell’s equations. Sta-
tistical mechanics may then be applied, in analogy with the case of an ideal gas, to
define two intensive, thermodynamic variables for each plasma species, s = (i, e),
namely: particle density, n s and kinetic temperature, Ts = 13 m s v 2 v , where m s is
the particle mass and ·v denotes an average over all particle velocities at a given
point in space. A two-species plasma may thus be described by six parameters:
two densities, n s , two thermal speeds, vts = (Ts /m s )1/2 and two magnitudes of
the Coulomb interaction, es2 /0 m s , es being the species charge. Note that global
quasi-neutrality requires n ≈ n i ≈ n e /Z , where Z = |ei /ee |, but allows Ti and Te
to differ substantially. Since electrons are much lighter than ions (m i /m e > 103 ),
they are also much more mobile, that is
vte /vti ∼ (m i /m e )1/2  1. (2.1)
Consequently, relativistic effects may be neglected provided that
vte /c < 0.05. (2.2)
We can use these six parameters to construct the basic length and time scales of
plasma physics, which should be small compared to the system size, L, and the
observation time, τ . We thus obtain the plasma frequency, ω ps , which describes
electrostatic oscillations of species s,
ω2ps = n s es2 /0 m s , (2.3)
the Debye length, λ Ds , which measures the exponential shielding of a test charge
by the rearrangement of free charges,
ϕ(r ) ∝ r −1 exp(−r/λ Ds ), (2.4)
18 Magnetized plasma physics

and hence the maximum extent of collective interactions,

 1/2
λ Ds = vts /ω ps = 0 Ts /n s es2 , (2.5)

and the classical distance of closest approach in a Coulomb collision, rcs ,

rcs = es2 /4π0 Ts . (2.6)

−1/3
Comparing this distance with the average inter-particle spacing, rds = n s , the
plasma, or weak-coupling, criterion may be restated as
√
 ps = λ Ds /rcs = (rds /rcs )3/2 / 4π = 4πn s λ3Ds  1. (2.7)

It requires that a Debye sphere (a sphere of radius λ Ds ) contain a large number of

particles, such that the combined effect of frequent grazing collisions far exceeds
that of rare close encounters; (2.7) is amply satisfied by most ion–electron gases
and all fusion plasmas. When it is violated, particles become strongly coupled by
binary interactions and cease to be a plasma in the sense of systems theory. Strong
coupling generally applies to a degenerate quantum gas, e.g. electrons in a metal,
quantum effects being negligible only when the thermal electron deBroglie length,
λ Be , is less than rde ,
λ Be /rde = n 1/3
e h̄/m e vte < 1. (2.8)

In contrast, (2.8) is satisfied in all fusion plasmas, so that quantum effects

may be relegated to nuclear, atomic and molecular interactions, and treated by
corresponding transition probabilities and collisional cross-sections.

2.1.2 Magnetization parameter

The presence of an ambient magnetic field B introduces additional length and time
scales, namely the gyro-frequency,

s = es B/m s , (2.9)

and the (thermal) gyro-radius of species s,

ρs (v) = v⊥ / s , ρts = ρs (vts ) = vts / s , (2.10)

which arise due to gyration of charged particles around the magnetic lines of force.
A plasma is said to be magnetized when ρts is small in relation to the system size or,
more stringently, to the (smallest) scale length of the electromagnetic force fields,
L = min(L B , L E ), L −1
B = ∇ ln B, etc.,

δs = ρts /L = ωts / s  1, ωts ≡ vts /L , (2.11)

 2.2 Charged particle motion 19

where ωts is the local transit frequency and δs is known as the magnetization
parameter. The number of particles in a gyro-sphere (a sphere of radius ρts ),
namely nρts3 , is an alternative, though less frequently used, measure of magnetiza-
tion. Since e  i and ρti  ρte due to the mass difference, (2.11) is equivalent
to δi  1. This is generally true for fusion plasmas, but can be violated in regions
of strong transverse gradients, see Section 7.3.1.
Particle gyration generates plasma currents which in turn produce a plasma mag-
netic field opposite in direction to the ambient field B, i.e. a diamagnetic field. The
relative strength of this back-reaction is measured by the ratio βs of the kinetic
( ps ≡ n s Ts ) and magnetic (B 2 /2μ0 ) pressures,

βs = 2μ0 ps /B 2 , β= βs = 2μ0 p/B 2 . (2.12)

When the plasma beta, β, is small, as is the case in most fusion plasmas, the
ambient magnetic field is merely modulated by plasma dynamics.
It can be shown that the three dimensionless parameters s (2.7), δs (2.11) and
βs (2.12) for each plasma species s, or any combination thereof, are sufficient to
completely describe Maxwell’s and Newton’s equations, which characterize the
dynamics of a magnetized plasma. One such combination, widely used in confined
plasma research, is the triplet of ρs∗ , νs∗ and βs ,

ρs∗ ≡ ρts /a, νs∗ ≡ L/λs ≈ L/vts τs ≈ νs /ωts , (2.13)

where a is the toroidal minor radius of the plasma, νs = τs−1 is the collision fre-
quency of species s, λs is their collisional mean free path and νs∗ is known as the
collisionality parameter, typically defined in terms of the parallel gradient length
or the parallel connection length of the magnetic field L  .

2.2 Charged particle motion

All macroscopic phenomena have microscopic origins. This key insight of modern
physics pertains to all systems in Nature, including ion–electron plasmas. Conse-
quently, the origin of all plasma behaviour lies in the motions of charged particles
and their interactions with ambient and self-generated force fields. Provided (2.2)
and (2.8) hold, the trajectory of each plasma particle follows Newton’s equations
of motion,
dt x = v, dt (m s v) = m s as = Fs + es v × B, (2.14)

where Fs = es E + m s g is the sum of electric and gravitational forces acting on a

particle (the latter represents any generic charge independent force) and the s index
on x and v was omitted for brevity.
20 Magnetized plasma physics

2.2.1 Guiding centre drifts

If the force fields are constant in time and uniform in space, (2.14) may be solved
by transforming to a frame moving at the particle drift velocity,1
Fs × b E×b g×b
vGC⊥ ≡ U⊥ ≡ = + = v E + vgs , (2.15)
es B B s
where b = B/B is a unit vector in the direction of B. The electric drift v E is
independent of mass and charge, while the gravitational drift vgs is both mass and
charge dependent. The former causes electrons and (all) ions to drift normal to both
E and b, resulting in flow of mass but not charge, while the latter causes opposite
charges to drift in opposite directions, normal to both g and b, producing a net flow
of both mass and charge.
At this stage, we write x and v as a sum of gyrating and non-gyrating parts,
denoted by lower and upper case r and u, respectively,
x = r + R, v = u + U. (2.16)
In addition, we introduce the gyro-tropic notation, in which any vector A is
decomposed into projections parallel () and perpendicular (⊥) to b,
A = A + A⊥ = A b + b × (A × b), A = A · b. (2.17)
Thus, the  and ⊥ velocities in the drifting frame are U and u ≡ u⊥ , respec-
tively. The former is unaffected by the magnetic force, es v × B, whereas the latter
describes gyration around the magnetic field lines.
The solution to (2.14) is a combination of free-streaming parallel to b, gyration
perpendicular to b and slow drift normal to both b and Fs ,
u⊥ = u ⊥ (ê2 cos γ − ê3 sin γ ) = u ⊥ êu , u ⊥ = ρs s , u = 0 (2.18)

U = v = v0 + Fs t/m s b, U⊥ = (2.15). (2.19)
Here γ = s t + γ0 is the gyro-phase and (b, ê2 , ê3 ) form a right-handed basis.
Integration of (2.18)–(2.19) yields a drifting helical orbit,
r = −1
s b × u⊥ = ρs (ê2 sin γ + ê3 cos γ ) = ρs êr , (2.20)

R = v0 t + Fs t 2 /2m s b + U⊥ t. (2.21)
Note that (2.14) takes on a particularly simple form in the (b, êr , êu ) basis,
Fs
dt x = v, dt v = as = − ρs 2s êr , (2.22)
ms

1 Since a particle drift occurs along with motion of its guiding centre, see below, we denote the drift velocity by
the subscript GC.
 2.2 Charged particle motion 21

in which bb + êu êu + êr êr = I, êu = ∂γ êr , êr = −∂γ êu and
 
2 êu êu dγ = 2 êr êr dγ = êr êu + γ (êu êu + êr êr ) + const. (2.23)

To summarize, the particle gyrates with radius ρs and speed u ⊥ = |u⊥ | = ρs s

around its centre of gyration, known as the guiding centre (GC). The GC-position,
xGC , and velocity, vGC , defined as gyro-averages of x and v,

1
R ≡ xGC ≡ xγ ≡ xdγ , U ≡ vGC ≡ vγ = dt R, (2.24)
2π
are independent of γ . In magnetized plasmas, the GC-description greatly simplifies
charged particle dynamics, as we will see presently.
The above results presuppose stationary and uniform force fields. If the fields
are either time-dependent or non-uniform, a moving particle will experience
forces which change during the course of its orbit. Provided the force fields vary
adiabatically, i.e. the forces change little during one gyro-period,

 ≡ max (ρs /L , 1/τ s ) ∼ O(δs )  1, (2.25)

where L = min(L B , L E ) with L −1 B = ∇ ln B, etc. and τ ≡ min(τ B , τ E ) with

−1
τ B = ∂t ln B, etc. are the length and time scales of this variation, then the slow
GC-drift may be calculated using a multiple time scale expansion (Bogoliubov
and Mitropolski, 1961). This technique exploits the large difference in temporal
scales, evident in (2.25), to treat the gyro-phase γ and the time t as independent
variables. The equations of motion (2.14) are then expanded in the small parameter
, averaged over the rapid gyration using ·γ = ·dγ /2π, and solved order by
order in . Specifically, the gyration radius r(R, U, t, γ ) and velocity u(R, U, t, γ )
are expanded in powers of , denoted by subscripts,

x = R + r 0 + r1 + · · · , v = U + u 0 + u1 + · · · (2.26)

The time derivative of the gyro-phase is similarly expanded, dt γ = ω−1 +ω0 +· · · ,

taking note that, to lowest order, dt γ ∼ s ∼ O( −1 ). The assumed periodicity
requires the averages of the gyrating quantities to vanish, rγ = uγ = 0, which
sets the solubility criteria at each order.
Substituting (2.26) into (2.14), equating terms of same order in , and satisfying
the solubility criteria, yields evolution equations for U and γ , with dt R = U being
satisfied to all levels of accuracy. Thus, to zeroth order,
e
ω∂γ u − s u × b = (E + U0 × B), (2.27)
m
22 Magnetized plasma physics

where the order-subscript is only retained in U. The average of (2.27) yields the
zeroth-order GC-drift, with all terms evaluated at R,
U0 = U0 + U0⊥ = U0 b + v E . (2.28)
The perpendicular component is just the electric drift, U0⊥ = v E , as expected for a
stationary, uniform force field seen by the particle to lowest order in . Integrating
(2.27) we recover the gyration velocity u = u⊥ = ∂γ r, (2.18) with γ = s t +γ0 ,
and the corresponding gyro-radius r, (2.20).
Gyro-averaging the next-order equation of motion, we find
mdt U0 = e[U1 × B + E  b + u × B(x)γ ], (2.29)
where the last term must be evaluated over the gyro-orbit, x = R + r. This gyro-
average can be found by Taylor expanding B(R + r), with the result
mu 2⊥ mu 2⊥
eu × B(x)γ = − ∇ B = −μ∇ B, μ= , (2.30)
2B 2B
where m = −μb is the magnetic moment of the gyration evaluated at R. Note
that m always opposes the magnetic field, i.e. it is diamagnetic, and that μ/m is
independent of both mass and charge. The term −μ∇ B is known as the mirror
force for reasons that will become clear shortly.
Combining (2.28)–(2.30), we find
mdt U0 = mdt (U0 b + v E ) = e[U1 × B + E  b − μ∇ B]. (2.31)
Dot product with b yields the evolution of U0 , while cross product results in the
first-order perpendicular velocity U1⊥ ,
e μ
dt U0 = E  + b · ∇ B − b · dt v E , (2.32)
m m
b μ  b μ 
U1⊥ = × dt U0 + ∇ B = × U0 dt b + dt v E + ∇ B . (2.33)
 m  m
The first term in (2.33) is linked to the zeroth-order GC-acceleration, dt U0 , the
second term to the mirror force, −μ∇ B. The two terms, known as the inertial and
magnetic drifts, can be expressed in the gravitational form, vgs = g × b/ s (2.15),
with g = Fs /m replaced by −dt U0 and −μ∇ B/m, respectively. The g-form offers
useful insight into the origin of both drifts, namely the variation of the gyro-radius
during the course of gyration.
The inertial drift consists of two parts, made explicit in (2.33), which reflect
changes to U0 and U0⊥ = v E . Since b × bdt U0 = 0, the parallel component of
the inertial drift depends only on the change in b perceived by the guiding centre
(expressed by a zeroth-order advective derivative),
dt b = (∂t + U0 · ∇)b = ∂t b + v E · ∇b + U0 b · ∇b. (2.34)
 2.2 Charged particle motion 23

Parallel motion being generally much faster than any perpendicular drift (U0 
v E ), the last term usually dominates. It describes the centripetal acceleration felt
by the guiding centre moving along a curved magnetic field line. Introducing the
magnetic curvature vector, κ ≡ b · ∇b,2 this curvature drift can also be written in
gravitational form with g = −κU0 2
. Using b · ∇b + b × ∇ × b = 0 and Ampere’s
law, ∇ × B = μ0 J, we find
 
∇⊥ B μ0 J × b μ0 B2
κ ≡ ∇ b = −b × (∇ × b) ≈ + ≈ 2 ∇⊥ p + . (2.35)
B B B 2μ0
Finally, changes to U0⊥ = V E give rise to the polarization drift,
b b E×b
vp = × dt v E = × (∂t + U0 · ∇) , (2.36)
  B
where the variation is once again evaluated at the moving guiding centre.
We can now combine (2.28) and (2.33) to write the GC-velocity, vGC = U =
U0 + U1 + O( 2 ) to first order in ,
g×b
U = U b + v E + + O( 2 ), −1
s ∼ O(), (2.37)
s
μ
−g = (∂t b + v E · ∇b + κU )U + dt v E + ∇ B. (2.38)
m
In the typical case of stationary B and weak, electrostatic E = −∇ϕ, the first,
second and fourth terms in (2.38) may be omitted, leaving3
 
μ u 2⊥ ∇ B
− g ≈ κU + ∇ B ≈ U +
2 2
. (2.39)
m 2 B
The second form applies only to low-β, current free (J×B ≈ 0) plasmas, for which
(2.35) reduces to κ ≈ ∇⊥ ln B = ∇ ln B. For higher β, we find
 
b u 2⊥ ∇ B U2 (∇ × B)⊥
U⊥ = v E + × U  + 2
+ + O( 2 ). (2.40)
 2 B  B
The relative strength of the transverse electric field, E ⊥ , and hence of the elec-
tric drift, v E = E ⊥ /B, compared to the ion thermal speed, vti , gives rise to
two ordering schemes for magnetized plasma dynamics. These are traditionally
referred to by the names of the resulting dynamical equations, irrespective of

2 The curvature of some vector field is defined as the variation of its unit vector along the lines of force. Hence,
κ ≡ b · ∇b = ∇ b represents the variation of b along the magnetic field lines.
3 The first term in this ‘negative’ gravity is simply the centripetal acceleration of a particle travelling along a
curved magnetic field line, and is thus known as the curvature drift. The second term is related to the non-
uniformity in the the field magnitude and thus the variation of the gyro-radius; it is thus known as the gradient
B, or simply ∇ B, drift.
24 Magnetized plasma physics

whether these are formulated in the kinetic or fluid descriptions. One thus speaks
of magneto-hydrodynamic (MHD) ordering,
v E /vti ∼ 1, ω ∼ ωti ∼ δi i , δi  1, (2.41)
and of drift ordering,
v E /vti ∼ δi  1, ω ∼ δi ωti ∼ δi2 i  ωti , (2.42)
where ω is the typical frequency of dynamical evolution being considered and all
expressions are written for ions on account of δi  δe . In both cases, the remaining
GC-drifts are small compared to vti , vgs /vti ∼ δi .
Consider the relative strength of the electrostatic, Eϕ , and inductive, E A ,
contributions to the electric field and hence the electric drift,
E = Eϕ + E A = −∇ϕ − ∂t A, (2.43)
vE = vϕE + v EA = (∇ϕ + ∂t A) × (b/B), (2.44)
where ϕ is the scalar and A the vector potential, defined by B = ∇ × A. Hence vϕE
arises due to a perpendicular gradient of ϕ and v EA due to the time variation of A⊥
and hence of B = B = |B|. The magnitude of these two contributions to v E may
be estimated as
ϕ
v E = ∇⊥ ϕ/B ∼ ϕ/L ⊥ B, (2.45)
v EA = ∂t A⊥ /B ∼ L ⊥ ∂t B/B ∼ L ⊥ ω B ∼ L ⊥ ω  
B ( B/B). (2.46)
Under the MHD ordering, we have ω   
B = ∂t B/ B ∼ ωti so that v E ∼ vti B/B,
A 
ϕ
which is typically larger than v E , e.g. assuming that eϕ is comparable to Te , we find
v ϕE ∼ Te /L ⊥ eB ∼ δi vti which is a factor δi /( 
B/B) smaller than v EA . The exception
is the important case of so-called flute-reduced MHD, which corresponds to a low-
beta plasma confined in a strong, externally imposed field, which evolves primarily
by perpendicular (flute-like) displacements, see Section 4.2.5. In this case, parallel
field fluctuations become

B/B ∼ μ0 
p /B 2 ∼ (
p / p)β ∼ δi β, v EA /v ϕE ∼ ( 
B/B)/δi ∼ β  1. (2.47)
A similar result is found under the drift ordering, (2.42), when the magnetic field
is constrained to evolve so slowly that the inductive contribution is negligible,
especially in the case of magnetically confined plasmas,

B ∼ δi ωti ,
ω v EA ∼ δi vti ( 
B/B), v EA /v ϕE ∼ 
B/B  1. (2.48)
In short, for most fusion plasma applications, including power exhaust, one may
assume that v E ≈ ∇ϕ × (b/B) irrespective of the ordering scheme.
The particle kinetic energy, K = 12 mv 2 , evolves according with dt K = v · F =
ev·E, obtained by taking a dot product of (2.14) with v; the magnetic force, ev×B,
 2.2 Charged particle motion 25

being normal to v, cannot modify K. Here and below, dt is taken along the GC-
trajectory, so that dt = ∂t + U · ∇. Expanding v 2 = v · v, with v = U + u given by
(2.26), and gyro-averaging, yields
m 
dt K = dt U2 + u 2⊥ + v 2E = eU · E + eu · E(x)γ , (2.49)
2
where, as in (2.29), we find contributions from both GC-motion and gyration.
The integral over the gyro-orbit, evaluated analogous to that in (2.29), gives the
electromagnetic work, −μb · ∇ × E = μ∂t B. Hence,

dt K = eU · E + μ∂t B. (2.50)

The rate of change of the total energy, defined as

1 m 2
E = K + eϕ = mv 2 + eϕ = u + v 2E + U2 + eϕ, (2.51)
2 2 ⊥
can now be evaluated using (2.50) and (2.43),

dt E = μdt B + edt ϕ + eU · E = μ∂t B + e∂t ϕ − eU · ∂t A. (2.52)

The total energy is evidently conserved in stationary fields. Using (2.32) and (2.33)
to eliminate U2 and v 2E = v E · v E from (2.51), we find that the magnetic moment
is conserved to first order in ,
 2
mu ⊥
d t μ = dt = O(). (2.53)
2B
In Section 2.2.2, we will see that μ is the lowest-order approximation to an adia-
batic invariant, conserved to all orders in . We can thus treat 12 mu 2⊥ = μB as an
internal energy of the guiding centre,
m 2
E = K + eϕ = v + U2 + μB + eϕ, (2.54)
2 E
and express the parallel GC-velocity as a function of E and μ,

2
U (R, E, μ) = ± (E − μB − eϕ) − v 2E , (2.55)
m
where all quantities are now evaluated at the GC-position, R.
Consider the typical case of stationary B and weak, electrostatic E = −∇ϕ, for
which the last term in (2.55) may be omitted. To express the GC-velocity (2.37)
in terms of the conserved quantities E and μ, we eliminate ∇ B and ∇ϕ using the
gradient of (2.54), mU ∇U = −μ∇ B − e∇ϕ, to find
  
b μ e 1
U = U b + × κU + ∇ B + ∇ϕ = U b + ∇ × (U b) . (2.56)
2
 m m 
26 Magnetized plasma physics

Introducing the pseudo-magnetic field, B∗ and vector potential, A∗ ,

m
B∗ = ∇ × A∗ = B + ∇ × (U b), A∗ = A + mU b/e, (2.57)
e
allows (2.56) to be expressed in a particularly simple form, U = U B∗ /B, i.e. the
GC is seen to follow the pseudo-magnetic field lines. Together, (2.55) and (2.56)
give U(R, E, μ) to order . More generally,
   
1 1 u 2⊥ (∇ × B)
vGC = U = U b + ∇ × (U b) − U −
2
, (2.58)
  2 B
where the additional terms represent the parallel drift velocity, vGC = (μ/m)
(∇ × B) , which will be derived in Section 2.3.2, (2.114), and the high-β
perpendicular drift, U2 /  (∇ × B)⊥ , already obtained in (2.40).
The origin of the label mirror force to denote −μ∇ B should now clear. As the
guiding centre moves into regions of stronger B field, the perpendicular energy
1
2
mu 2⊥ = μB increases to satisfy dt μ ≈ 0. Since total energy is also conserved
(dt E ≈ 0), parallel energy 12 mU2 must be reduced and the guiding centre decel-
erates. At the turning point R , defined by U (E, μ, R ) = 0, or E = μB(R ), the
GC is reflected and the direction of U reverses. Defining the velocity pitch angle
as α = tan−1 (u ⊥ /U ), we see that

sin2 α = u 2⊥ /v 2 = μB/E, (2.59)

and α increases with B to a maximum of π/2 at the turning point. Since both E
and μ are constants of motion, reflection occurs when

sin2 α(R) = B(R)/B(R ), (2.60)

which defines the loss cone in U − u ⊥ velocity space; this cone becomes narrower
with increasing B(R ), reducing to a ray as B(R ) → ∞.4
Consider a gas of charged particles in a (periodic) magnetic well with a mini-
mum and maximum field, Bmin and Bmax , at locations Rmin and Rmax lying on the
same field line. Assuming an isotropic velocity distribution at Rmin , the fraction of
particles trapped in the well follows from (2.60),

α(Rmin ) 2 Bmin
fT = 1 − = 1 − sin−1 . (2.61)
π/2 π Bmax

In an infinite well, the loss cone closes and all particles are trapped.

4 Note the analogy with celestial mechanics: only a particle (body) with zero magnetic moment μ (angular
momentum ) in the region of vanishing magnetic (gravitational) field, can enter the strong field region; all
orbits with finite μ () are reflected to satisfy dt E = dt μ = 0 (dt  = 0).
 2.2 Charged particle motion 27

Finally, we evaluate the rates of change of E, μ and γ over the exact particle
trajectory, i.e. dt = ∂t +v·∇, which we will need in Section 2.3.2. The calculation is
simplified by adopting the (b, êr , êu ) basis, defined by (2.18) and (2.20). Taking dot
products of (2.22) with v, v⊥ and êr , and negotiating some rather lengthy algebra,
we obtain

dt E = es ∂t ϕ − es v · ∂t A, (2.62)
Bdt μ = −μdt B − m s v v⊥ · dt b + es v⊥ · E, (2.63)
v es
dt γ =  − êr · dt êu − êr · dt b − êr · E. (2.64)
v⊥ m s v⊥
Here we made use of ê2 · dt ê2 = ê3 · dt ê3 = ê3 · dt ê2 + ê2 · dt ê3 = 0, and

êr · dt v = v êr · dt b, êr · dt v⊥ = v⊥ ê3 · dt ê2 − v⊥ dt γ . (2.65)

Assuming ∂t ∼ δs s and v E /vt ∼ δs , all right-hand side terms in (2.62)–(2.64) are

small compared to the gyration term s . Gyro-averaging with these assumptions,
we recover the conservation of E and μ to order δs ,
μ
dt Eγ ≈ es ∂t ϕ − es v · ∂t A, dt μγ ≈ − ∂t B, dt γ γ ≈ s . (2.66)
B

2.2.2 Canonical (angle-action) variables

The analysis presented above was based on the expansion of Newton’s equations
(2.14) in the small parameter  (2.25). This procedure, while formally correct, has
several limitations: it becomes unwieldy at higher orders in ; fares poorly in com-
plicated geometries; and is ill-suited to exploiting the symmetry properties that
may exist in a given plasma configuration.
To overcome these problems, we resort to the Lagrange–Hamilton formulation
of classical mechanics, an elegant and powerful meta-theory forming the skele-
ton of modern physics. Its cornerstone is the Principle of Least Action, which
statesthat a dynamical system evolves in such a way as to minimize the action,
S = L(q, q̇, t)dt, where the Lagrangian, L = K − V, defined as the difference
between the kinetic and potential energies, is a function of only the generalized
co-ordinates q = (q1 , q2 , . . .), velocities q̇ = (q˙1 , q˙2 , . . .) and time. Expressed as
a variational principle,

δS = δ L(q, q̇, t)dt = 0, (2.67)

it states that the phase path taken by the system in the phase space formed by q
and q̇ minimizes the path integral of L, i.e. the action. Performing the variation we
obtain Lagrange’s equations of motion,
28 Magnetized plasma physics

dt q = p, dt p = F = ∂q L, (2.68)

where p are the generalized momenta and F the generalized forces. This shows
that dynamical invariants originate in the symmetry properties of the Lagrangian:
if L is independent of co-ordinate qk , the corresponding momentum pk becomes a
constant of motion. For example, consider a closed system of n (self-)interacting
1 2
particles: q j = x j , q̇ j = v j , L = mv j − V(x1 , . . . , xn ). Homogeneity of
2 1 2
time (∂t L = 0) results in conservation of energy, E = mv + V, homogeneity
 2 j
of space (∇L = 0) in conservation of momentum, mv j , and isotropy of space

(∂φ L = ∂θ L = 0) in conservation of angular momentum, x j × p j , (Landau and
Lifschitz, 1960, Vol. I).
To illustrate that Newton’s and Lagrange’s equations are indeed equivalent, let
us consider a single charged particle5 moving in an electromagnetic field, for which
Newton’s equations are given by (2.14). In this case, q = x, q̇ = v and L = 12 mv2 +
eA · v − eϕ, where A is the vector potential, B = ∇ × A. Inserting these into (2.68),
we recover (2.14), with the generalized momentum p = mv + eA. If the magnetic
field is symmetric in some q j , then p j will be conserved. For instance, in cylindrical
geometry (r, θ, z), longitudal symmetry (∂z = 0) requires pz = m ż + e A z =
const, while poloidal symmetry (∂θ = 0), results in pθ = mr 2 θ̇ + er Aθ = const.
Similarly, in toroidal geometry (r, θ, ζ ), axis-symmetry (∂ζ = 0) implies

pζ = m R 2 ζ̇ + e R Aζ = R(mvζ + e Aζ ) = const. (2.69)

It is worth noting that the gyro-average of pζ , which represents the generalized

toroidal momentum of the guiding centre, is likewise conserved,

 pζ  B = R(mbζ v + e Aζ ) = m Rbζ v − eψ P ≈ const ≈ pζ , (2.70)

where we wrote the toroidal GC velocity as vζ ≈ bζ v and introduced the poloidal

flux function, ψ P = −R Aζ , see Chapter 3.
Conservation of generalized momenta in response to system symmetries moti-
vates a tranformation from the phase space defined by q and q̇ to that formed by q
and p. This is accomplished by means of a Legendre transformation, and results in
Hamilton’s formulation of classical mechanics. Thus, the Lagrangian L(q, q̇, t) is
replaced by the Hamiltonian H(q, p, t) = p · q − L, and (2.68) by Hamilton’s (or
canonical) equations of motion,

dt q = ∂p H, dt p = −∂q H, (2.71)

5 For simplicity we drop the species index, so that m = m , e = e , x = x , v = v , etc.

s s s s
 2.2 Charged particle motion 29

which may also be derived from (2.67) written as δ p · dq − Hdt = 0. The action
S evolves according to the related Hamilton–Jacobi equation,
∂q S = p, ∂t S = −H, dt S = L ⇒ ∂t S + H(q, ∂q S) = 0. (2.72)
The physical meaning of the Hamiltonian becomes evident when H is evaluated

for a given system: e.g. our self-interacting particles, H = mv j · v j − L =
1 2
2
mv j + V or a charged particle in an electromagnetic field, H = v · (mv +
eA) − L = 12 mv 2 + eϕ. In both cases, the Hamiltonian is simply the total energy of
the system, E = K + V. Conservation of energy may thus be obtained by taking the
total time derivative of H, and using (2.71) to eliminate dt q and dt p, which yields
dt H = ∂t H. More generally, the evolution of any quantity A may be expressed as
dt A = ∂t A + {H, A} = ∂t A + ∂p H · ∂q A − ∂q H · ∂p A, (2.73)
where we introduced the commutator, or Poisson bracket, which should be familiar
from quantum mechanics. If A does not explicitly depend on time, it will be a
constant of motion (dt A = 0) provided it commutes with H, that is if {H, A} =
0. It can be shown (Poisson’s theorem) that a Poisson bracket of two constants of
motion is likewise a constant of motion.
The principal advantage of (2.71) over (2.68) is the conjugate symmetry of the
canonical variables q and p. This symmetry permits a wide range of canonical
transformations q, p → q (q, p, t), p (q, p, t), where the new variables satisfy
(2.71) with some new H as well as the conjugate constraints,
        
qk , ql = 0, pk , pl = 0, qk , pl = δkl . (2.74)
The relations between the new and old quantities may be derived from the generat-
ing function of the transformation F, defined as dF = p·dq−p ·dq +(H −H)dt.
In particular, H = H + ∂t F so that the Hamiltonian remains constant if F does
not explicitly
 depend on time. In that case, the phase space volume, measured by
the integral dqdp, is conserved.
Consider a dynamical system with n degrees of freedom. Its phase space may be
described as a 2n-dimensional differentiable manifold and its evolution according
with (2.71) may be viewed as a series of infinitesimal canonical transformations
within that manifold, with −S playing the role of the generating function, F.
Therefore, provided the system conserves energy (∂t H = 0), its flow in phase
space is ‘incompressible’,

∂q dt q + ∂p dt p = 0 ⇔ dt dqdp = 0. (2.75)

This result, expressed above in differential and integral forms, is known as Liou-
ville’s theorem. It applies to any conjugate pair qk , pk , as well as the entire system
30 Magnetized plasma physics

trajectory q, p. If the system is dissipative (∂t H < 0), its phase space flow is
‘compressed’ at the rate of energy loss.
Liouville’s theorem may also be expressed in path integral form as

dt J = dt p · dq − Hdt = dt Ldt = 0, (2.76)

C(t) C(t)

where C(t) is any closed path satisfying the dynamical equations (2.71). The inte-
grals Jk , one for each conjugate pair qk , pk , are known as Poincaré invariants and
are evidently conserved.6
When n is large, the system is best described in terms of the phase space density
of states, F(q,
 p). Since individual trajectories satisfy (2.75), their sum must do
likewise, dt Fs dqdp = 0, and Liouville’s theorem becomes

dt F = ∂t F + {H, F} = 0. (2.77)

To fully exploit
 (2.76), we effect a canonical transformation, (q, p) → (ϑ, I)
using F = p · dq as the generating function. The quantities ϑk = ∂Ik F and
Ik = Jk /2π, known as canonical, or angle-action, variables, form the simplest
canonically conjugate pairs for a bounded, conservative system. Their Hamilton’s
equations (2.71) are found to describe multiple rotations with angles, ϑk , constant
frequencies, ωk , and constant radii, Ik ,

dt ϑk = ∂Ik H = ωk (Ik ) = const, dt Ik = −∂ϑk H = 0. (2.78)

For a system with n degrees of freedom, (2.78) define an n-torus, i.e. an n-

dimensional hyper-surface formed by concentric tori with k-axial radii Ik = const
and angles ϑk ; the system evolves on the n-torus by constant rotation around each
of its axes. Determining all the angle-action variables is thus formally analogous to
an eigenvalue decomposition, with the final solution formed by a linear superposi-
tion of oscillating modes. Louiville’s theorem (2.76) takes on a particularly simple
form in the angle-action representation, Lk dϑk = Jk = 2πIk = const, showing
a formal identity, aside from a numerical constant, between action variables and
Poincaré invariants.
Consider a system whose Lagrangian depends on several parameters λ j (Landau
and Lifschitz, 1960). If dt λ j = 0, all the action variables are conserved and the
system evolves according to (2.78). If λ j vary with time, the action S is modified by
terms involving λ j , namely (∂λ j H)q, p dλ j =−(∂λ j L)q,q̇ dλ j . Effecting a canonical
transformation using F = S0 (q, E; λ j ) = pdq, the new Hamiltonian becomes

6 Their existence can also be deduced from the complete integral of (2.72), which contains one integration
constant ck for each independent variable qk : S = S(t, q, c) + c0 . Provided the system is non-degenerate,
solving (2.72) by separation of variables yields the required constants of motion, ck = Jk = const.
 2.2 Charged particle motion 31

H = H +  j dt λ j with  j = (∂λ j S0 )q,I . The canonical equations (2.78) are
thus replaced by,

dt ϑk (λ j ) = ∂Ik H = ωk (Ik ; λ j ) + ∂Ik  j ϑ,λ j dt λ j , (2.79)

dt Ik (λ j ) = −∂ϑk H = − ∂ϑk  j I ,λ j dt λ j . (2.80)

Let us assume that λ j vary adiabatically with time, in the sense of (2.25), i.e. that
dt ln λ j =  × min(ωk ), where   1 and ωk = (∂Ik H)λ j are the unperturbed
angular frequencies. We see from (2.80) that the evolution of the perturbed action
variables Ik (λ j ) now involves a product of fast oscillation, represented by the peri-
odic coefficients (∂ϑk  j )I ,λ j , and slow variation at dt λ j ∼ O(). We can thus
average (2.80) over the fast rotation using ·ϑk = ·dϑk /2π , taking dt λ j outside
the integral, with the result
 
dt Ik = dt Ik  ≈ − ∂λ j ∂ϑk S0 ϑ dt λ j ≈ 0. (2.81)
k

The averaged action variables, which represent the Poincaré invariants, Jk = 2πIk
averaged over the fast rotation, are therefore adiabatically invariant. Treating ϑk
as complex variables, it can be shown that Ik  are indeed conserved to all orders
in  as exp(−1/).7
We next subject the system to small, sinusoidal perturbations λ j ∼ exp(iωλ j t).
If the pitch of the helical field lines covering the n-torus has a ‘radial’ shear
dωk /dIk > 0, then dynamical modes resonant with the perturbations (ωk = mωλ j ,
where m is an integer) will be excited, in what is known as parametric resonance.
As a result, the associated toroidal surfaces break up into m helical islands, whose
O-points and X-points remain at the ‘radial’ location of the unperturbed surface,
and which are separated from the smooth nested tori by a separatrix surface.
What happens when the strength of the perturbation is increased? This ques-
tion was formally answered by a celebrated Kolmogorov–Arnold–Moser (KAM)
theorem, a good account of which may be found in Arnold (1989). The answer
is two-fold: (i) the identity of the islands is preserved, even as their ‘radial’
width increases; and (ii) chaotic volumes appear close to the X-points and expand
along with the strength of the perturbation. Eventually, the chaotic region fills
the entire volume of the helical island and the unperturbed resonant surface is
destroyed, as illustrated in Fig. 2.1. In Chapter 3 we will see that flux surfaces
in magnetically confined plasmas are realizations of such abstract canonical tori in
three-dimensional space.
Returning to the discussion of charged particle gyration (Section 2.2.1), we recall
that all fields were assumed to vary adiabatically (  1). We could thus proceed

7 This result is the first ‘dividend’ of our ‘investment’ in the Hamilton–Lagrange formalism.
Exploring the Variety of Random
Documents with Different Content
 My very dear Friend,

M AN appoints, but an all-wise, all-gracious God disappoints. Dear

Mr. H――y guessed right. This day sevennight we weighed
anchor, and sailed, though very slowly, as far as Fair-Lee. But for
near five days we have been tossed by violent gales, and last night,
through infinite mercy, cast anchor off Dungenness and New-
Romney. The new sailors have been quite sick, but are now almost
recovered. I have felt very little, comparatively speaking, and have
been able to read, &c. &c. Had I known of having such a handy
Steward, I might have spared one hand; but what is, is best. In
God’s due time, the winds will have a commission to carry us on.
Satan doth not like this voyage. Pray on, pray on, my very dear
friends, and never fear. All shall work together for good to those who
love God. I wish you had advertised against the publisher of my last
sermon. It is not verbatim as I delivered it. In some places, he
makes me to speak false concord, and even nonsense. In others, the
sense and connection are destroyed, by the injudicious disjointed
paragraphs; and the whole is entirely unfit for the public review ¹.
But we must suffer by the false zeal of professing friends, as well as
by the inveterate malice of public avowed enemies. If one sentence
is blessed to the conviction and edification of any single individual, I
care not what becomes of my character, though there is no occasion
of bringing ourselves into needless contempt. I write this at a
venture; we see a boat approaching, and hope it will reach us. We
have put back: O that it may quicken friends to pray us forwards.
We are like a man of war that hath been out on a short cruise, and
then returned into harbour. What awaits us we know not. It is not fit
we should. God bless you all, my dear, very dear friends. I remember
your times of meeting at both ends of the town. Sea is sea, land is
land. The God whom we serve, is God of both. To his never-failing
mercy I commend you. Continue to do the same for, my very dear
friend,

Less than the least of all,

 G. W.

Off New-Romney, September 28.

S TILL we are prisoners. But blessed be God, prisoners of hope! In

God’s due time, the word of command will be, Go forward. I am
sadly off for want of white biscuit. But God will supply every want.
The boat is going, that came off with some eatables. Adieu. Tender
love to all. Blessed be God, all is well! Cease not to pray for, my very
dear friend,

Less than the least of all,

G. W.
¹ The very same censures are too justly applicable to the
volume of sermons, now published by Mr. Gurney, as Mr.
Whitefield’s.――Relying on the accuracy of the shorthand-
writer, and on the fidelity of the reverend gentleman who
was to revise the sermons, and on their joint regard for
the memory of Mr. Whitefield, his executors did agree to
recommend the intended publication, and promote its
sale; for which, a consideration was to be paid by Mr.
Gurney, to be applied by them, according to the tenor of
Mr. Whitefield’s Will. But on their receiving nine of the
sermons, (worked off) to their great surprise, they found
themselves, after perusal, unable to authenticate them,
either as to language or sentiments; therefore judged
them utterly unfit for publication, and told Mr. Gurney,
that on no consideration whatever, could they recommend
them to the public. The executors are extremely
concerned on Mr. Gurney’s account, as well as for the
character of their late worthy friend; and now wish that
they had not rested with the hearing only one half-sheet
read to them, but had insisted on seeing the whole
Manuscript, and every sheet from the press. However,
though Mr. Gurney ought to have stopped the press when
first applied to, and although the agreement was never
signed by either of the parties, the executors have
repeatedly offered, that, besides chearfully renouncing all
advantages, they will pay whatever expences Mr. Gurney
hath been at in the affair, and so take and burn the whole
impression; as otherwise the purchasers must be
deceived, and the name of the deceased sorely wounded.

LETTER MCCCCXLI.
To Mr. R―― K――n.

On board the Friendship, Captain Ball, in Five-

fathom Hole, about ten miles from Charles-Town,
 South-Carolina. November 30, 1769.

My dear Friend,

W E have had a long, and in several respects a trying passage.

Particulars expect by the packet, which the pilot says will sail
in two or three days. This day week we first saw land; came over
the bar the 28th instant, and should have been at Charles-Town that
evening, but our ship was too light to obey the helm. We had the
mortification of seeing ten sailing in before us, and we ever since left
in jeopardy. Surely Satan foresees some signal good attending this
voyage. In the midst of all, blessed be God, we have had plenty of
outward things; and I am in better health than at the end of any
voyage I have made for some years. Mr. Smith hath really behaved
well, and been very handy and attentive. The same may be said of
Mr. Winter. We have been like the three children in the fiery furnace.
But the Son of God hath been, and is (O amazing grace!) still with
us. Please to remember us to all concerned. Hoping soon to write
from on shore, and most earnestly praying, that grace, mercy and
peace, may be multiplied upon you all, I must hasten to subscribe
my old but true name, “The chief of sinners, less than the least of all
saints,” but

Your affectionate, obliged friend, and willing servant to all,

G. W.

Charles-Town, December 1.

B LESSED be God, a pilot-boat came yesterday along-side, and

brought us hither in safety to our unspeakable comfort in the
evening. Our reception as hearty or heartier than ever. Grace! grace!
The ship is not yet come up. Blessed be God, I am brave and well,
and am to preach this afternoon. Praise Him whose mercy endureth
for ever! We have been delivered out of great jeopardy. You will say
so, when you receive further particulars from, my dear friend,
 Less than the least of all,

G. W.

LETTER MCCCCXLII.
To Miss H――y.

Charles-Town, South-Carolina,
December 6, 1769.

Dear Madam,

S HALL I promise, and not perform? God forbid! You have one of
my first letters since our arrival. The long passage made shore
more agreeable. Miss H――y knows how to apply such an account.
All we meet with here, will be sweetly overruled to render heaven,
and a sight of Jesus in the heaven of heavens, more delightful. I am
in hopes, by this last week’s preaching, that some South-Carolina
souls are beginning to look heavenwards. Grace! grace! In a day or
two, God willing, we shall move to Bethesda. Mr. W――t is come to
meet me, and tells me all is in great forwardness there. Ere long the
top-stone of a building not made with hands will be brought forth.
How many of your beloved family will join the shout, crying, Grace,
grace unto it! Most cordial and due respects attend them all. God be
praised, heaven is in sight. Jesus is our pilot: he will steer us safe
over every bar; even over the last bar, Death. Then will we sing,

All our sorrows left below,

And earth exchang’d for heav’n.

O how good is it to bear the yoke in our youth! A glorious
preparative for a comfortable old age. That you and all your dear
relatives, may increase with all the increase of God, is the earnest
prayer of, dear Madam,

Less than the least of all,

G. W.

LETTER MCCCCXLIII.
To Mr. R―― K――n.

Charles-Town, December 9, 1769.

My very dear Friend,

I MMEDIATELY on our arrival, I sent you a letter by way of

Liverpool, and then promised you a particular and very explicit
letter by this packet: but it must be deferred a few days. So much
company crowds in, that together with my preaching every other
day, &c. &c. I have scarce the least leisure. Blessed be God, I have
already met with some fruits of my feeble labours in this place. An
earnest, I hope, of good things to come. To-morrow, I set off by
water to Georgia, the roads being almost impassable by land. Mr.
Wright is come to go with me, and acquaints me that all is in great
forwardness at Bethesda. How I am directed in respect to that
institution, you shall know hereafter. And glory be to a never-failing
God, an hereafter is drawing on apace, when we shall sing,

All our sorrows left below,

And earth exchang’d for heav’n.

I hope you and yours are helped to possess your souls in patience.
God bless and reward you! Grace, mercy, and peace be multiplied
upon you and yours! All join in sending due and cordial respects. O
give thanks unto the Lord, for his mercy endureth for ever. Brethren,
pray for us. Tender love to all. Your letter to me by Anderson, I have
not yet received. Blessed be God, I am in health. Grace! grace! The
packet is about to be closed. Other ships are almost ready to sail. By
one or all expect to hear again from, my very dear friend,

Less than the least of all,

G. W.

LETTER MCCCCXLIV.
To Mr. R―― K――n.

Savannah, December 24, 1769.

My very dear Friend,

T HE bearer B―― F―― is the steward of the ship in which I came

over: a very handy useful man; to whom in a great measure I
owe the chief comfort of my voyage, as to eatables. He hath had
convictions at times, and longs to live ashore. If you can serve him,
do. I write this at my old friends Mr. Habersham’s. I am to preach
here this morning, and to-morrow, and purpose in a few days to pay
a visit to Charles-Town. Blessed be God, all things are in a most
promising way. But I am obliged to leave Mr. W――t behind, for the
work’s-sake. Mr. Smith goes with me. He is attentive, hath behaved
well, and been useful in the house. Never was I blessed with so
many proper industrious workmen and helpers before. Grace! grace!
Next Wednesday I am fifty-five years old. God be merciful to me a
sinner, a sinner, a sinner! As such, continue to pray, my dear steady
friend, for,

Less than the least of all,

G. W.

LETTER MCCCCXLV.
To Mr. S―― S――.

Bethesda, January 11, 1770.

Dear Sir,

C AN I forget my dear, very dear old steady friend? rather let my

right hand forget her cunning. How are you? Still afflicted? still
in pain? still made to possess wearisome nights, and wearisome
days? Well, all will be over soon; soon, yea very soon shall we sing,

All our sorrows left below,

And earth exchang’d for heav’n.

This prospect gives songs in the night; this makes Georgia and
Bethesda to more than smile: and indeed you and yours would smile
too, were you to see what a lasting foundation is laying for the
support and education of many yet unborn. All admire the work
already done. In a few months the top-stone, I trust, will be brought
forth, with shouting, Grace! grace! In the mean while I must range
northward. I know who will follow me with their prayers; even you
and yours, whom I dearly love, and whom I salute much in the Lord.
God bless you all, for all kindnesses conferred upon, my very dear
friends,

Less than the least of all,

G. W.

If I thought you did not, or would not use your globes, I would
beg them for our infant library. The increase of this colony is almost
incredible. Real good, I trust, is doing; and a blessed door is opening
for Mr. W――’s usefulness. Blessed be God! Blessed be God!
 LETTER MCCCCXLVI.
To Mrs. H――e.

Bethesda, January 11, 1770.

Dear Mrs. H――e,

M Y last to you left me just arrived at Charles-Town. This leaves

me an old inhabitant of, or rather a worthless sojourner at
Bethesda. Both, I hope, will find the worthy Mrs. H――e and her
daughter enjoying thriving souls in healthy bodies. Every thing here
exceeds my most sanguine expectations. I am almost tempted to
say, “It is good for us to be here.” But all must give way to gospel
ranging: Divine employ!

For this let men revile my name,

I’d shun no cross, I’d fear no shame:

All hail, reproach――

I hope London friends meet with enough of this. It is bad, more than
bad, when the offence of the cross ceaseth. This cannot be, till we
cease to be crucified to the world, and the world crucified to us: and
when that is the case, things are very bad. As Mrs. H――s’s heart
and house are opened to so many ministers of a despised Jesus, she
must expect a double share. Not only reproach from the world, but
judgings and censures from the narrow-hearted bigotted part of the
Church. But she hath counted the cost; she knows in whom she hath
believed, and who will be her exceeding great reward. God will not
have us take up with any thing short of himself. Leaving you to cry,
Grace! grace! with ten thousand thanks for all unmerited favours, I
must hasten to subscribe myself

Your most obliged and ready servant, in our common Lord,

G. W.

LETTER MCCCCXLVII.
To Mr. R―― K――n.

Charles-Town, February 10, 1770.

My dear Friend,

L AST night a kind Providence brought me hither, where I received

your kind letter, dated November 2d. Blessed be God, that all
things go on so well at London! You reason well about ――. I
entirely agree with you in sentiment concerning that matter. Through
infinite mercy, this leaves me enjoying a greater share of bodily
health than I have known for many years. I am now enabled to
preach almost every day, and my poor feeble labours seem not to be
in vain in the Lord. Blessed be God, all things are in great
forwardness at Bethesda. I have conversed with the G――r in the
most explicit manner, more than once, concerning an act of
assembly, for the establishment of the intended Orphan-house
College. He most readily consents. I have shewn him a draught,
which he much approves of; and all will be finished at my return
from the northward. In the mean while, the buildings will be carried
on. As two ministers from the New-Jersies, and Rhode-Island, have
been soliciting benefactions for their respective colleges, no
applications of that nature can be made here: but the Lord will
provide! My eyes wait upon Him, from whom all temporal and
spiritual salvations come. Since my being in Charles-Town, I have
shewn the draught to some persons of great eminence and
influence. They highly approve of it, and willingly consent to be
some of the wardens: near twenty are to be of Georgia, and about
six of this place; one of Philadelphia, one of New-York, one of
Boston, three of Edinburgh, two of Glasgow, and six of London.
Those of Georgia and South-Carolina, are to be qualified; the others
to be only honorary corresponding wardens. I have therefore taken
the freedom of nominating ******; and as my name is to be
annihilated, they may accept the trust without expecting much
trouble, or suffering contempt for being connected with me. This, I
think, is the chief of the plan: more particulars that may occur,
together with the draught of the charter, you may expect hereafter.
In the mean while, cease not to pray for, my dear steady friend,

Less than the least of all,

G. W.

LETTER MCCCCXLVIII.
To Mr. R―― K――n.

Charles-Town, February 22, 1770.

My very dear Friend,

N O letters by the packet, or another ship that hath brought in

above five hundred from London! As I hear Captain Rainier is
bound for Savannah, I hope at my return to Bethesda to find a letter
there. Your last, dated November 2, was immediately answered. Mr.
B――s will accept my most grateful acknowledgments for his kind
present of maps, charts, &c. In a few months, I hope, all will be
compleated. But what may these few months produce? Lord Jesus,
prepare us for whatever thou hast prepared for us, and give peace
in our time, for thine infinite mercy’s-sake! You must expect another
draught soon. God be praised for that saying, “It is more blessed to
give than to receive.” You would be pleased to see with what
attention people hear the word preached. I have been in Charles-
Town near a fortnight, am to preach at a neighbouring country
parish church next Sunday, and hope to see Georgia the week
following. Perhaps I may sail from thence to the northward, and
perhaps embark from hence. Lord Jesus, direct my goings in thy way!
I am blessed with bodily health, and am enabled to go on my way
rejoicing. Grace! grace! Join in shouting those blessed words. I
wrote by one Captain Watt, who was to sail from Georgia this week.
In that, you will find something concerning my late visit to, and
public entertainment at Bethesda. You see how often I pester you
with letters. I can only add, that you may tell all, I am happier than
words can express: which I take, in a great measure, to be owing to
the prayers of my dear English friends, which are daily put up for,
and I hope daily returned by, an unworthy worm. Remember me
most affectionately to all, and expect to hear again very speedily
from, my very dear, dear friend,

Less than the least of all,

G. W.

LETTER MCCCCXLIX.
To Mr. B――n.

Charles-Town, February 27, 1770.

My dear Mr. B――n,

I OWE you an answer to your kind letter. Blessed be God, I can
send you good news from a far country! All things at Bethesda
go on quite well. My bodily health is upon the advance, and the
word, I trust, runs and is glorified. At present, my intended plan
about returning continues the same: but all depends on news from
home. Strange! that none could write a line or two by so many
ships. Only one letter have I received from Mr. K――n since my
arrival. Next week, God willing, I return to Georgia, and soon after I
purpose to go to the northward. I know who will follow me with their
prayers: they will avail much. The Lord Jesus be with all your spirits! I
suppose you heard from Bethesda by Captain Anderson. Mr. Wright
is the main spring with regard to the buildings, and all the other
wheels move orderly and well. Praise the Lord, O my soul! O this
pilgrim way of life! To me it is life indeed. No nestling, no nestling,
my dear Mr. B――n, on this side eternity. This is not our rest. Ere
long we shall sing,

All our sorrows left below,

And earth exchang’d for heav’n.

Leaving you to add Hallelujah, and sending most hearty greetings to

your whole self, and all enquiring friends, I must hasten to subscribe
myself, my dear Mr. B――n,

Less than the least of all,

G. W.

March 4.

Pray tell Mr. K――n, that I hope to write to him in a few days
from Georgia. To-morrow, God willing, I return thither. I trust
substantial good hath been done here. Grace! grace!
 LETTER MCCCCL.
To Mrs. H――e.

Charles-Town, March 4, 1770.

Dear Mrs. H――e,

A LTHOUGH at such a distance, I cannot forget Mrs. H――e and

her daughter, and all their works of faith and labours of love. I
doubt not, but this will find them on the full stretch for heaven, and
as usual abounding in the work of the Lord. It leaves me almost
ready to return to Bethesda, from a place where, I trust, the word
hath run and been glorified. Matters are now drawing near to a
wished-for close. All things have succeeded beyond my most
sanguine expectation. I expect to come according to the appointed
time. But future things belong to Him who orders all things well.
Through mercy I enjoy more bodily health than for many years last
past. You will join in crying, Grace! grace! Next month, I purpose
moving to the northward. As Mr. W――t is the main spring at the
Orphan-house, I must leave him behind. Mr. Smith is with me: he
behaves well, and is diligent and attentive. You will be so good as to
remember me to all, as they come in your way. I hope my good old
Mrs. Eades, at Tottenham-Court, is well. That she and all may ripen
for heaven every day and every hour, is the earnest prayer of, dear
Mrs. H――e,

Less than the least of all,

G. W.
 LETTER MCCCCLI.
To Mr. R―― K――n.

Savannah, March 11, 1770.

My very dear worthy Friend,

B LESSED be God, the good wine seemed to be kept till the last at
Charles-Town. Last Thursday I returned, and found all well at
Bethesda. I am come to town to preach this morning, though
somewhat fatigued with being on the water three nights: upon the
whole, however, I am better in health than I have been for many
years. Praise the Lord, O my soul! I have been sadly disappointed in
receiving no letters by the Charles-Town packet. All knew that I was
to be in these parts, only till the ensuing Lady-day: then I purpose to
set off for the northward. I drew at Charles-Town for ――l. perhaps
may draw again soon. Expect more particulars in a few days. This is
waited for. God bless you all! Dearly beloved in the Lord, pray for us.
Time is scarce allowed me to subscribe myself, my very dear friend,

Less than the least of all,

G. W.

LETTER MCCCCLII.
To the Same.

Bethesda, April 6, 1770.

My very dear Sir,

I AM waiting here for a brig that is to carry me northward, and for a
letter and news from England. Your last was dated, November 2:
several months have intervened. I now almost despair of hearing
from you again, till my arrival at Boston. But I hope that you and all
remember us more frequently than you write. You are daily
remembered at a throne of grace. How glad would many be to see
our Goshen, our Bethel, our Bethesda! Never did I enjoy such
domestic peace, comfort, and joy during my whole pilgrimage. It is
unspeakable, it is full of glory. Peace, peace unutterable attends our
paths, and a pleasing prospect of increasing, useful prosperity is
continually rising to our view. I have lately taken six poor children,
and, God willing, purpose to add greatly to their number. Dear Mr.
D――n and his wife are to sail the beginning of next month in the
Britannia, Captain Dean, bound for Portsmouth. We part with great
respect. Fain would I retain such an old tried disinterested friend in
the service of the sanctuary, and near my person. But what scheme
to pursue, I know not, being so uncertain as to the path which I
shall be called to take. A few months will determine: perhaps a few
weeks. More particulars expect by the Britannia. In the mean while,
I can only recommend you all to the blessed Jesus, and the word of
his grace, and entreat the continuance of your prayers in behalf of,
my very dear friend,

Less than the least of all,

G. W.

LETTER MCCCCLIII.
To the Same.

Bethesda, April 16, 1770.

 My very dear worthy Friend,

H ALLELUJAH! Praise the Lord! The books and letters both by Ball
and Sunbury, are come safe. You have done quite right. Our
Lord must choose his own means to bring about his own purpose.
Mr. Smith (the clerk) was much rejoiced by receiving a letter. Poor
Mr. Jacob W――t, an honest industrious creature, was as much
dejected by receiving none. If Mr. G――s had added a line or two to
his present, it would have been doubly acceptable. Next week, God
willing, we sail for Philadelphia. I shall leave letters behind me to
come by Mr. D――n. All is well, all more than well here! Never, never
did I enjoy such an æra of domestic peace and happiness. I have
taken in about ten orphans. Prizes! prizes! Hallelujah! Join, my very
dear friends, join in praising Him whose mercy endureth for ever. If
possible, I shall write a line to the Welch brethren. They have
sustained a loss indeed, in the death of Mr. Howell Davies. God
sanctify it! Surely my turn will come by and by. But I must away to
Savannah. Real good, I trust, is doing there. The ship that brings
this, is expected to sail to-morrow. I have desired Mr. W――r to send
you the particulars of our voyage. All send due respects. God bless
you! God reward you! Cease not to add to my obligations, by
continuing to pray for, my very dear worthy friend,

Less than the least of all,

G. W.

LETTER MCCCCLIV.
To the Same.

Bethesda, April 20, 1770.

 My very dear Sir,

T O my very great joy, a few days ago I received your kind letters
with all the papers. Give peace in our time, O Lord! We enjoy a
little heaven upon earth here. With regret I go northward, as far as
Philadelphia at least, next Monday. Though I am persuaded, as the
house is now altered, I should be cooler here, during the summer’s
heat, than at any other place I know of, where I used to go. I should
be glad to treat you with some of the produce of our colony, which is
much earlier than yours. The audits, &c. sent with this, be pleased
to communicate to all my real friends. You have certainly determined
quite right in a late affair. Every thing concurs to shew me, that
Bethesda affairs must go on as yet in their old channel. A few
months may open strange scenes. O for a spirit of love and
moderation on all sides, and on both sides the water! I wish some
books might be procured for our infant library. But more of this in
my next. Letters may now be sent by way of Boston, New York, and
Philadelphia. I should be glad to hear often, if it be but a line. In all
probability I shall not return hither till November. Was ever any man
blest with such a sett of skilful, peaceful, laborious helpers! O
Bethesda, my Bethel, my Peniel! My happiness is inconceivable. A
few hundreds, besides what is already devoted, would finish all. I do
not in the least doubt. I have had nine or ten prizes lately. You know
what I mean. Nine or ten orphans have been lately taken in.
Hallelujah! hallelujah! Let Chapel, Tabernacle, heaven, and earth,
rebound with Hallelujah! I can no more. My heart is too big to speak
or add more, than my old name,

Less than the least of all,

G. W.

LETTER MCCCCLV.
 To Mr. and Mrs. S――n.

Bethesda, April 21, 1770.

My dear Friends,

L ONG before now, I hope you have found, that I have not
forgotten you or your labours of love. This comes to inform you,
that the Father of mercies hath not forgotten to be gracious to the
chief of sinners, and less than the least of all saints. On the contrary,
he daily loads us with his benefits. Bethesda is a place, that the Lord
doth and will bless. Dear Mr. D――n and his wife, will inform you of
particulars. Among other things, they will tell you of our new chapel.
I have sent for sundries for its use and completion. O help me to
praise Him, whose loving kindness is better than life! I hope your
daughter grows in grace, and will become like unto one of the
polished corners of the temple. That root and branch may increase
with all the increase of God, most earnestly prays, my very dear
friends,

Less than the least of all,

G. W.

LETTER MCCCCLVI.
To Mrs. H――e.

Bethesda, April 21, 1770.

Dear Mrs. H――e,

N O such good news yet. Less than the least of all, is not drowned
to this very day. Perhaps he may live to see his London friends
in England, or at Bethesda. How would many rejoice to be in such a
peaceful, commodious, and comfortable habitation! I cannot tell you
half. Blessed be God, I was never better, at this season of the year, in
bodily health; never more comfortable in my soul. Grace! grace!
Hallelujah! Praise the Lord! Mr. D――n, a faithful disinterested friend,
will acquaint you with particulars. He hath often heard me speak of
Mrs. H――e and her daughter Phebe. Still employed in the old way. I
am glad Mr. D――cy is under your roof. Put it all down to the old
account. God bless you all! I am sure you pray for me at London. All
join in sending cordial respects. Happy Bethesda! Help, help in
praising Him, whole mercy superaboundeth to, dear Mrs. H――e,

Less than the least of all,

G. W.

LETTER MCCCCLVII.
To Mr. S―― S――.

Bethesda, April 21, 1770.

Dear Sir,

A LTHOUGH I have scarce time to turn round, being just setting off
for Philadelphia, yet I must drop a few lines to my old
invariable friend. As I hear nothing to the contrary, I suppose he is
yet in this dying world. Well! so that we die daily to ourselves and
the world, all is well, and shall end well. This I am persuaded is your
happy case. In some degree, I trust, it is mine. Would you think it?
My bodily strength seems to be renewed, and every thing at
Bethesda is in a most promising way. Dear Mr. D――n, the bearer of
this, must be referred to for particulars. Never did I spend such a
comfortable domestic winter, as the last. Never was a man blessed
with a better set of skilful, peaceful, laborious helpers. All is of grace,
with which, that you, your dear yoke-fellow, and other connections,
may be filled brimful, is the hearty prayer of, my very dear Sir,

Less than the least of all,

G. W.

LETTER MCCCCLVIII.
To Mr. K――n.

Savannah, April 24, 1770, Five in the morning.

My very dear Mr. K――n,

I AM just going into the boat, in order to embark for Philadelphia. I

hope the good wine was kept to the last, on Sunday. Mr. D――n
and his wife are to sail in about a fortnight. He is an honest
creature, and an excellent accomptant. I have written strongly in his
behalf. He will bring a large packet, and is to have ―― pounds of
you, which I have given him as a present. I have also drawn on you
for £.――; perhaps shall draw no more for some time. This will prove
a blessed year for me at the day of judgment. Hallelujah! Come
Lord, come! Mr. Robert W――t hath herewith sent you a power of
attorney, begging you would settle his affairs in Essex. He is worthy,
for whom you should do this. A quiet, ingenious, good creature, and
his wife an excellent mistress of the family. Such a set of helpers I
never met with. They will go on with the buildings, while I take my
gospel range to the northward. It is for thee, O Jesus, even for thee,
thou never-failing Bethesda’s God! But I can no more at present.
Hoping to write again soon from Philadelphia, and praying that all
may increase with all the increase of God, I must hasten to subscribe
myself, my dear, dear Sir,

Less than the least of all,

G. W.

LETTER MCCCCLIX.
To Mr. R―― K――n.

Philadelphia, May 9, 1770.

My very dear Friend,

T HIS leaves me a two days inhabitant of Philadelphia. I embarked

at Savannah, in the Georgia packet, on the 24th ultima and
arrived here the 6th instant. The evening following, I was enabled to
preach to a large auditory, and am to repeat the delightful task this
evening. Pulpits, hearts, and affections, seem to be as open and
enlarged towards me, as ever. Praise the Lord, O our souls! Whilst I
am writing, perhaps Mr. D――n and his wife are ready to sail from
Savannah. By them you will receive a large packet concerning
Bethesda. All is well, blessed be God, all is more than well there. As
yet I have my old plan in view, to travel in these northern parts all
summer, and return late in the fall to Georgia. All the letters and
packets came safe. I believe you had best write by the New-York
packet. But letters directed either to New-York, Boston, or this place,
will reach or be sent to me. Through infinite mercy, I still continue in
good health, and more and more in love every day with a pilgrim
life. God bless you, and all my dear friends and hearers in the great
metropolis. I know they pray for me. They are never forgotten day
or night. That all may increase with all the increase of God, is the
continual cry of, my very dear steady friend,

Less than the least of all,

G. W.

LETTER MCCCCLX.
To the Same.

Philadelphia, May 24, 1770.

My very dear Friend,

I WROTE to you by the last New-York packet, as well as by Mr.

D――n, who was to sail from Savannah the 10th instant. I have
now been here near three weeks, and in about a week more I
purpose to set off for New-York in my way to Boston. A wide and
effectual door, I trust, hath been opened in this city. People of all
ranks flock as much as ever. Impressions are made on many, and I
trust they will abide. To all the episcopal churches, as well as most
of the other places of worship, I have free access. My bodily health
is preserved, and notwithstanding I preach twice on the Lord’s-day,
and three or four times a week besides, yet I am rather better than I
have been for many years. This is the Lord’s doing. To this long-
suffering, never-failing Lord, be all the glory! Be pleased to excuse
my enlarging. Whilst I am itinerating, little leisure will be allowed for
writing. The New-York packet will be the best opportunity for you.
The particulars herein mentioned, be so good as to send by the first
opportunity. I can no more at this time. God willing, you will soon
have another line from, my very dear steady friend,

Less than the least of all,

 G. W.

LETTER MCCCCLXI.
To the Same.

Philadelphia, June 14, 1770.

My very dear steady Friend,

T HIS leaves me just returned from a hundred and fifty miles

circuit, in which, blessed be God! I have been enabled to
preach every day. So many new as well as old doors are open, and
so many invitations sent from various quarters, that I know not
which way to turn myself. However, at present I am bound to New-
York, and so on further northward. Help me to praise Him whose
mercy endureth for ever. As yet I am enabled to ride and travel
chearfully; the heat not greater than yours in England. Expect to
hear further, as we go along. The ship I find is going. Tender love to
all. Cease not to stir up all to persevere in praying for, my very dear
friend,

Less than the least of all,

G. W.

LETTER MCCCCLXII.
To the Same.

New-York, June 30, 1770.

 My very dear Friend,

I HAVE been here just a week. Have been enabled to preach four
times, and am to repeat the delightful task this evening.
Congregations are rather larger than ever. You will see by the
inclosed packet, what numerous invitations from every quarter I am
daily receiving. Blessed be God, I have been strengthened to
itinerate and preach daily for some time. Next week I purpose to go
to Albany. From thence, perhaps, to the Onoida Indians. There is to
be a very large Indian congress; Mr. Kirkland accompanies me. He is
a truly christian minister, and missionary. Every thing possible should
be done to strengthen his hands and his heart. I shall write, God
willing, at my return. The letters dated February 22, with the packet
of papers, I have received here from Charles-Town. The New-York
packet is the surest conveyance. Perhaps I may not see Georgia till
Christmas. As yet, I keep to my intended plan, in respect to my
returning. Lord Jesus, direct my goings in thy way! The heat begins
now to be a little intense; but through mercy I am enabled to bear
up bravely. What a God do we serve! By this time I hope Mr. D――n
and his wife are arrived. Hearty love to them, and to all who find it
in their hearts to pray for, and enquire after, my very dear Sir,

Less than the least of all,

G. W.

LETTER MCCCCLXIII.
To Mr. R―― K――n.

New-York, July 29, 1770.

My very dear Friend,

S INCE my last, and during this month, I have been above a five
hundred miles circuit, and have been enabled to preach and
travel through the heat every day. The congregations have been
very large, attentive, and affected, particularly at Albany,
Schenecdady, Great Barrington, Norfolk, Salisbury, Sharon,
Smithfield, Powkeepsy, Fishkill, New Rumburt, New Windsor, and
Peckshill. Last night I returned hither, and hope to set out for Boston
in two or three days. O what a new scene of usefulness is opening in
various parts of this new world! All fresh work, where I have been.
The divine influence hath been as at the first. Invitations croud upon
me both from ministers and people, from many, many quarters. A
very peculiar providence led me lately to a place, where a horse-
stealer was executed. Thousands attended. The poor criminal had
sent me several letters, hearing I was in the country. The Sheriff
allowed him to come and hear a sermon under an adjacent tree.
Solemn, solemn! After being by himself about an hour, I walked half
a mile with him to the gallows. His heart had been softened before
my first visit. He seemed full of solid divine consolations. An
instructive walk. I went up with him into the cart. He gave a short
exhortation. I then stood upon the coffin, added, I trust, a word in
season, prayed, gave the blessing, and took my leave. Effectual
good, I hope, was done to the hearers and spectators. Grace! grace!
But I must not enlarge. The Ship is going, and I keep at home to
write this. O that you had only dropped a line by the New-York
packet! That is convenient for all parts of the continent. My next
may be from Boston. Pray excuse me to all; for travelling and
preaching entirely prevent my writing as I would. All are continually
remembered by, my very dear friend,

Less than the least of all,

G. W.
 LETTER MCCCCLXIV.
To Mr. W――t.

Boston, September 17, 1770.

Dear Mr. W――t,

I AM afraid, as Mr. E――n mentioned your writing, that your letter

hath miscarried. But, blessed be God! I find all was well; only I
want to know what things are wanted, that I might order them from
Philadelphia, by Captain Souder. Fain would I contrive to come by
him, but people are so importunate for my stay in these parts, that I
fear it will be impracticable. Lord Jesus, direct my goings in thy way!
He will, he will! My God will supply all my wants, according to the
riches of his grace in Christ Jesus. By a letter, received last night from
Mr. W――y, of July 5, I find that Mr. D――n was arrived, Anderson
sailed, and that all orders would be immediately complied with. Two
or three evenings ago, I was taken in the night with a violent lax,
attended with reaching and shivering, so that I was obliged to return
from Newbury, &c. &c.; but, through infinite mercy, I am restored,
and to-morrow morning hope to begin to begin again. Never was the
word received with greater eagerness than now. All opposition
seems as it were for a while to cease. I find God’s time is the best.
The season is critical as to outward circumstances. But when forts
are given up, the Lord Jesus can appoint salvation for walls and for
bulwarks; he hath promised to be a wall of fire round about his
people. This comforts me concerning Bethesda, though we should
have a Spanish war. You will be pleased to hear I never was carried
through the summer’s heat so well; I hope it hath been so with you,
and all my family. Hoping, ere long, to see you, I must hasten to
subscribe myself, my dear Mr. W――t,

Yours, &c. &c. &c.

 G. W.

LETTER MCCCCLXV.
To Mr. R―― K――n.

Portsmouth, New Hampshire,

September 23, 1770.

My very dear Friend,

Y OUR letters, of May 2 and 22, came to hand. New-York packet is

always the surest and most centrical medium of conveyance.
Before I left Boston, on Friday afternoon, I left a large packet in the
hands of a young man, who promised to deliver it to you safely. You
and Mr. H――y may peruse all, and communicate what you think
proper. By this time I thought to be moving southward. But never
was greater importunity used to detain me longer in these northern
parts. Poor New-England is much to be pitied; Boston people most
of all. How falsely misrepresented! What a mercy, that our christian
charter cannot be dissolved! Blessed be God for an unchangeable
Jesus! You will see, by the many invitations, what a door is opened
for preaching his everlasting gospel. I was so ill on Friday, that I
could not preach, though thousands were waiting to hear. Well, the
day of release will shortly come, ¹ but it does not seem yet; for, by
riding sixty miles, I am better, and hope to preach here to-morrow. I
trust, my blessed Master will accept of these poor efforts to serve
him. O for a warm heart; O to stand fast in the faith, to quit
ourselves like men, and be strong! May this be the happy experience
of you and yours! I suppose letters are gone for me, in Anderson, to
Georgia. If spared so long, I expect to see it about Christmas. Still
pray and praise. I am so poorly, and so engaged when able to
preach, that this must apologize for not writing to more friends. It is
quite impracticable. Hoping to see all dear friends about the time
proposed, and earnestly desiring a continued interest in all your
prayers, I must hasten to subscribe myself, my dear, very dear Sir,

Less than the least of all,

G. W.

¹ Mr. Whitefield died the 30th.

End of the Letters.

Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebooknice.com