Visit https://ebookfinal.

com to download the full version and

explore more ebook

Advanced Solid State Physics 2nd Edition Phillips

_____ Click the link below to download _____

https://ebookfinal.com/download/advanced-solid-state-
physics-2nd-edition-phillips/

Explore and download more ebook at ebookfinal.com

Here are some recommended products that might interest you.
You can download now and explore!

Elements of Solid State Physics 4th Edition J. P.

Srivastava

https://ebookfinal.com/download/elements-of-solid-state-physics-4th-
edition-j-p-srivastava/

ebookfinal.com

Solid State Physics Structure and Properties of Materials

Second Edition M.A. Wahab

https://ebookfinal.com/download/solid-state-physics-structure-and-
properties-of-materials-second-edition-m-a-wahab/

ebookfinal.com

Mesoscopic Electronics in Solid State Nanostructures 2nd

Edition Thomas Heinzel

https://ebookfinal.com/download/mesoscopic-electronics-in-solid-state-
nanostructures-2nd-edition-thomas-heinzel/

ebookfinal.com

Solid State Shear Pulverization 1st Edition Klementina

Khait

https://ebookfinal.com/download/solid-state-shear-pulverization-1st-
edition-klementina-khait/

ebookfinal.com
Advanced University Physics 2nd Edition Mircea S. Rogalski
(Author)

https://ebookfinal.com/download/advanced-university-physics-2nd-
edition-mircea-s-rogalski-author/

ebookfinal.com

Surface Modification by Solid State Processing 1st Edition

R. Miranda

https://ebookfinal.com/download/surface-modification-by-solid-state-
processing-1st-edition-r-miranda/

ebookfinal.com

The Oxford Solid State Basics 1st Edition Steven H. Simon

https://ebookfinal.com/download/the-oxford-solid-state-basics-1st-
edition-steven-h-simon/

ebookfinal.com

Robust Control System Design Advanced State Space

Techniques 2nd Edition V.N. Tsytovich

https://ebookfinal.com/download/robust-control-system-design-advanced-
state-space-techniques-2nd-edition-v-n-tsytovich/

ebookfinal.com

Robust Control System Design Advanced State Space

Techniques 2nd Edition Aman Ullah

https://ebookfinal.com/download/robust-control-system-design-advanced-
state-space-techniques-2nd-edition-aman-ullah/

ebookfinal.com
Advanced Solid State Physics 2nd Edition Phillips Digital
Instant Download
Author(s): Phillips, Philip
ISBN(s): 9780521194907, 0521194903
Edition: 2
File Details: PDF, 27.16 MB
Year: 2012
Language: english
\

Ad時間組 Solid St祖臨時y$i 岱

Second Edition

PHI Ll P PHIL Ll PS
University of IIlinois at Urbana-Champaign

m晶~CA悶BRIDGE
~，;，臨'iJ UNIVERSITY PRESS
 CAMBRIDGE UNIVERSITY PRESS
Cambridge , New York, Me1bourne , Madri d, Cape Town ,
Singapore, São Pau10 , De1hi , Mexico City
Cambridge University Press
The Edinburgh Bui1ding, Cambridge CB2 8RU, UK
Published in the United States of Am erica by Cambridge University Press , New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521194907

@ P. Phillips 2012

This pub1ication is in copyright. Subject to statutory exception

and to the provisions of re1evant collective 1icensing agreements ,
no reproduction of any part may take p1ace without the written
pe口nission of Cambridge University Press.

First edition pub1ished by Westview Press , a member ofthe Perseus Books Group , 2002
Second edition pub1ished by Cambridge University Press 2012

Printed in the United Kingdom at the University Press , Cambridge

A catalog record for this publicα tion is α:vailable from the British Libra.門

Library of Congress Catalogi時 inP他/ication data

Philli阱，Philip (Phi1ip W)
Advanced solid state physics / Philip Phillips. - 2nd ed.
p. cm.
Includes index.
ISBN 978趟。“ 521-19490-7 (hardback)
1. Solid-state physics. 1. Tit1e.
QC176.P46 2012
530.4'1 - dc23 2011042607

ISBN 978-0-521-19490-7 Hardback

Additiona1 resources for this publication at www.cambridge.org/so1idstate

Cambridge University Press has no responsibility for the persistence or

accuracy ofURLs for external or third-party internet websites referred to
in this publication , and does not guarantee that any content on such
websites 芯， or will remain , accurate or appropriate.
To Orestes and Angeliki
 “ The scientists of today think deeply instead of clearly. One must be sane to think clearly,
but one can think deeply and be quite insane."
Nikola Tesla, July 1934.
 Pr，φ ce page Xl

11479
1 Introduction
1.1 Spontaneously broken symmet可
1.2 Tracking broken symme仕y: order parameter
1.3 Beyond broken symmet可
References

2 Non-interacting electron gas 10

Problems 15

667023

1A-1--7
3 Born-Oppenheimer approximation
3.1 Basic Hamiltonian
3.2 Adiabatic approximation
3.3 Tight七inding approximation

』勻，“句，中
Prob1em
References

4.A 峙-LU
吋缸
4 Second quantization

J 呵，中 ?M
4.1 Bosons
4.2 Fermions

今卻司 3
勻/AVAV
4.3 Fermion operators
Problems
司3

References
司3
1111

5 Hartree-Fock approximation
司3

5.1 Non-interacting limit

司3
司JF3rhufo

5.2 Hartree-Fock approximation

司3

5.3 Diagrams
司3

Problem
司3

References
334
行/OOAV

6 Interacting electron gas

6.1 Uniform electron gas
6.2 Hartree-Fock excitation spectrum

V11
VIII Contents

-O缸 m 叫
fa札

(mh

hys
2889

而且可 A且可 A峙
n Vd O rI m ,4.• a

σb
.-E

戶UV

戶iH

唔，
V

ρlH

戶uv
ri

口3
EU

A
、 dwnb

EA
且
甘心們
NhR
nb
白白
Md

EA--
、

冶
3
11400

，J
local 冊 agnetic

、 ququfO勻/勻/勻
7 moments in metals
7.1 Local moments: phenomenology
7.2 Impurity density of states
7.3 Green functions
7 .4 Friedel's sum rule and local moments
Summary
Appendix to Chapter 7: Luttinger's theorem

J
勻/勻/
Problems
References

02374

OOOOOOOOAYAYAυAυnυ'111
8 Quenching of local moments: the Kondo problem
8.1 The Kondo Hamiltonian
8.2 Why is J negative?
8.3 Scattering and the resistivity minimum
8 .4 Electron-impuri句T scattering ampli個des
8.5 Kondo temperature
•••

8.6 Poor Man 's scaling

A
唔，
Summary

EA4EEA
Appendix to Chapter 8: the Schrieffer-Wolfftransformation

唔，
Problems

EA
唔，
References

EA
9 5creening and plasmons 115
9.1 Thomas-Fermi screening 115
9.2 凹的ma oscillations and collective coordinates 117
9.3 Linear response theory 122
9 .4 Dielectric response function 127
9.5 Kubo formula: electrical conductivity 137
9.6 Stopping power of a plasma 140
Summary 143
Problems 144
References 145

10 Bosonization 146
10.1 Luttinger liquid 146
10.2 Bosonization ofLuttinger model 151
10.3 Pair binding: can electrons do it alone? 160
10 .4 Excitation spectrum 162
~ ~~~

Summary 167
Problems 167
References 168

11 Electron-Iattice interactions 169

11.1 Harmonic chain 169
11.2 Acoustic phonons 171
11.3 Electron-phonon interaction 172
1 1.4 Ultrasonic attenuation 176
1 1. 5 Elec仕ical conduction 178
Summary 187
Problems 187
References 188

12 Superconductivity in metals 189

12.1 Superconductivi句: phenomenology 189
12.2 Electron-phonon effective interaction 197
12.3 Model interaction 199
12 .4 Cooper pairs 201
12.5 Fermi liquid theory 205
12.6 Pair amplitude 216
12.7 BCS ground state 221
12.8 Pair fluctuations 224
12.9 Ground state energy 226
12.10 Critical magnetic field 229
12.11 Energy gap 231
12.12 Quasi-particle excitations 233
12.13 Thermodynamics 237
12.14 Experimental applications 241
12.15 Josephson tunneling 253
Summary 255
Problems 255
References 257

13 Disorder: localization and exceptions 258

13.1 Primer on localization 259
13.2 Re制rn probability: localization criterion 261
13.3 Weak localization 264
13 .4 Scaling theory 269
13.5 Exceptions to localization 275
Summary 285
Problems 286
References 287
 Contents

14 Quantum phase transitions 289

14.1 Quantum rotor mode1 291
14.2 Sca1ing 293
14.3 Mean-fie1d solution 296
14.4 Landau-Ginsburg theory 302
14.5 Transport properties 306
14.6 Experiments 308
14.7 Sca1ing and T-1inear resistivity 310
Prob1ems 314
References 314

15 Quantum Hall and other topological states 317

15 .1 Wh叫 is the quantum Hall e:ffect? 317
15.2 Landau 1eve1s 320
15.3 The ro1e of disorder 324
15 .4 Currents at the edge 326
15.5 Topo1ogical insu1ators 330
15.6 Laughlin 1iquid 343
Summary 349
Prob1ems 349
References 350

16 Electrons at strong coupling: Mottness 353

16.1 Band insu1ator 354
16.2 Mott's prob1em 357
16.3 Much ado about zeros: Luttinger surface 363
16 .4 Beyond the atomic limit: Heisenberg versus Slater 368
16.5 Dynamica1 spectra1 weight transfer 380
16.6 Epilogue: 1 = 2 - 1 396
Prob1ems 397
References 398

lndex 400
 In writing the second edition of this text, 1 have tried to accomplish three things. First,
correct all the typos in the first edition. This has turned out to be somewhat harder than 1
had anticipated. While 1 am certain my proofreaders and 1 corrected all mistakes we could
fin d, that might not have been su血cient. As there will undoubtedly be a second printing,
simply email me any errors you might find at [email protected]. Second, include all the
material that should have been in the first edition but that 1 had given up on writing. This
includes Green functions , Luttinger's theorem, renormalization of short-range interactions
for Fermi liquids , and symmetry. In keeping with this being a physics rather than a technique
or mathematics tract, these subjects are interwoven wherever they are first needed. For
examp扭， the section on Green functions is in Chapter 7 where the An derson impurity
problem is treated. For completeness , Luttinger's theorem is also presented in the same
chapter but in an appendix. Third, include new material that re:flects the fast-moving pace
of lï = 1 research in condensed matter physics. Here 1 made a judgement based on what 1
anticipate students would find most useful. Since there are no texts that present the pedagogy
oftopological insulators (though some excellent review articles exist) and Mott insulators ,
1 chose to focus on those topics. In writing the topological insulator section, 1 have tried
to stick to the formulations that require the fewest definitions and new concepts since
the physics of these systems is inherently simple. Regarding the Mott problem, 1 present
what 1 think is non-controversial but not written down anywhere in a single manuscript.
Chapter 16 starts with the band insulator in which the rigid-band picture is valid and then
demonstrates that the physics of the Mott problem stands apart because no such rigid-band
picture applies. While tomes have been written about rigid-band m。由Is ， no text deals
with the breakdown ofthe rigid-band picture in strongly correlated electron problems. The
cuprate problem is discussed in this context. 1 had also intended to write a chapter on
quantum computing and extend the discussion in Chapter 14 to include the Bose-Hubbard
mode l. However, including such topics would have pushed the page count well over 600
pages , thereby making the book unwieldy. Further, such topics are not, in my estimation,
particularly suited to a core second-semester graduate class but rather to a more specialized
course. Perhaps 1 will think di在erently in a few years.
1 have benefitted from much input in the final editing of the current manuscript. Babak
Serac可 eh， Juan Jottar, and Taylor Hughes offered invaluable critiques of the topological
insulator section. Wei-Cheng Lee , Mohammad Edalati , and Taylor Hughes also read the
Mott chapter and caught several typos and inaccuracies. 1 also thank Mohammad for reading
and correcting the chapter on symmetries and Robert Leigh for his characteristically level-
headed and incisive remarks on strong coupling physics and symmetry. Wade deGottardi
offered numerous suggestions on the bosonization chapter. While 1 received emails from
XI
XII Preface

several students around the world detailing the typographical errors they have caught, 1
would especially like to thank Wei Han who found two key typos in two figures from the
first edition. Many thanks to Taylor Hughes for redrawing these figures. The duty of proof-
reading fell on my research group and other members ofthe ICMT gtoup at Urbana whose
arms are still recovering from the non-adiabatic distortions 1 applied to them. These include
Wei-Cheng Lee , Mohammad Edalati, Seungmin Hong , Wade deGottardi , Rodrigo Garrido ,
and Ki aran Dave. In addition, at the proof stage , Ki aran Dave , Ka Wai Lo , and Huihuo
Zheng read the entire manuscript and corrected it assiduously, in their relentless drive to
eliminate all typographical errors. 1 would like to thank Matthew Feickert for converting the
LaTeX files to the Cambridge style and for spotting several typographical errors along the
way, and the Cambridge staff, Mike Nugent, Simon Capelin, Claire Poole , Abigail Jones
and Frances Nex for their dedication to this project. Early influences without which this
book might not have been possible include my high school English teacher, Duane Kusler,
who encouraged me to write and my twin sisters Andi and Lyndi from whom 1 learned
many math tricks. My endearing thanks go to my family for their support and calming
presence.
Solid state physics grew out of applications of quantum mechanics to the problem of electron
conduction in solids. This seemingly simple problem defied solution because the presence
of an ion at each lattice site seemed to present an obvious impediment to conduction.
How the electrons avoid the ions was thus the basic question. Although the answer to this
question is well known, it does serve to illuminate the very essence of solid state physics:
there is organization in the many. Each electron adjusts its wavelength to take advantage of
the periodicity of the lattice. In the absence of impurities , conduction is perfect. Hence , by
understanding this simple fact that periodicity implies perfect conduction, it became clear
that the experimentally observed resistivity in a metal came not from electrons running
into each ofthe ions but rather from dirt (disorder) , thermal effects mediated by dynamical
motion of the ions , or electron-electron interactions. This book examines each of these
effects with an eye for identi有Ting underlying organizing principles that simplify the physics
of such interactions.

1.1 Sp曲目t晶晶體申明sly broke目 sy臨臨的時

The search for organizing principles that help simpli身 the physics of many-body systems
is at the heart of modern solid state or, more generally, condensed matter physics. One such
to01 is symme仕y. Consider the simple case ofpermutation symme仕y typically taught in a
first class in quantum mechanics. This symmetry was introduced into quantum mechanics by
W Heisenberg in the context of the indistinguishability of identical particles. The permuta-
tion group has a finite number of elements and hence is associated with a discrete symmetry.
Permutation symmetry allows us to classi冉T fundamental particles into two groups. Bosons
are even with respect to interchange of two particles and fermions odd. This symmetry can
be generalized to include a non-integer phase when two particles are interchange d，的 we
will see in the context of the fractional quantum Hall effect.
To a 1arge extent, the symmetries that are most relevant in condensed matter systems are
typical1y continuous , for example rotationa1 symmetry. Spontaneously breaking a continu-
ous symmetry has a fundamental consequence. For example , the existence of phonons in
a solid or spin waves in a magnet fOllOWS from the spontaneous breaking of a continuous
symmetry. By spontaneous, we mean without the application of an external field. A periodic
arrangement of ions in a crysta1 breaks continuous translational and rotational symmetry.
Such spontaneous breaking of a continuous symmetry by the very existence of the lattice
is necessarily accompanied by a massless spinless bosonic excitation. That such massless
2 Introduction

spinless bosons , known as Nambu-Goldstone bosons (G1961; N1960) , necessarily accom-

pany the breaking of a continuous symmetry can easily be deduced from the following
considerations. We consider a system with a Lagrangian
1
1i

唔，
/,

乙 =T-V((jJ)，
z

EA
,/
•••
.

、、
consisting of a kinetic energy, T , and a potential energy, V ((jJ ), where we are allowing for
(jJ to be a complex function. The claim that such a system is invariant under a symmetry
operation is captured by

V( (jJ) = V( (jJ + E8 (jJ), (l. 2)

where E8 (jJ is the generator ofthe symmetry operation. Here εis an infinitesimal. We have
assumed for the moment that 8(jJ is independent of space. To illustrate what is meant by
this identity, consider a potential of the form V ((jJ) = EO I 利 2. This potential is invariant
under transformations of the form 月h → 4忱的. Let () be a small quantity completely inde-
pendent of space. Then we can expand the exponential and retain only the :first-order term.
Consequently, (jJ•(jJ (1 + i()) and we identi有TE8 (jJ asiθ (jJ; th的 is ，石= () and 8φ = i(jJ.
This symmetry, known as U (1), is present in models that preserve charge conservation.
Expansion of V ((jJ ) to linear order in εimplies that

8V
8(jJ-;:-;- 0, (1.3)
8(jJ

assuming that the symme仕y is intact. N ow assume explicitly that the symmetry is broken
such that V → V((jJ。十 χ) ， where (jJo minimizes the potential and χcannot be written as
a generator of a symmetry operation as in Eq. (1 .2). Since the potential has a minimum, it
makes sense to expand

V(你 +χ) = 1
V ((jJ o) + ~ X 2 一一 I
"a 2 v I = V( (jJ o) + 一 χ 2m 2 ， (l. 4)
2/
l
a
(jJ 2 I <þ=<þ。

truncating at the restoring term at second order. The second term, which can be used to
de :fine the mass (m) in a standard harmonic expansion, is i址lerently positive semi闖 definite
since we have expanded about the minimum. With this equation in hand, we differentiate
Eq. (1.3) ,
2
a8 (jJ 8V . '" a v
一一一一一 +8 (jJ 一一才 =0 (1 .5)
a(jJ 8(jJ , a(jJ 2
with respect toφThe :first term vanishes when evaluated at the minimum, implying that

a2 VI
(1 .6)
8(jJ a(jJ 2! <þ=<þo
 1.1 Spontaneously broken symmetry

must identically vanish for any variation of cþ in the broken symme仕y state. Since 8cþ is
non-zero , Eq. (1 .6) is satisfied only if the second-order-derivative term vanishes or equiv-
alently if m 2 = O. That 站， the mass vanishes. This is Goldstone's theorem (G1961). A
zero mode exists for each generator of a continuously broken symme仕y. As a result ofthis
theorem, symmetry occupies a central place in all areas ofphysics , in particular particle and
condensed matter physics. Typically, the massless bosons th叫 arise in condensed matter
systems represent collective excitations ofthe entire many-body system. 1n fl自由， phonons
are purely longitudinal and arise from spontaneous breaking of Galilean invariance. 1n
solids , phonons are both transverse and longitudinal , though with no simple correspon-
dence with the spontaneous breaking of Galilean，仕anslational， and rotational symme仕y.
1n magnets , spin waves or magnons are the collective gapless excitations that emerge from
the spontaneous breaking of rotational symmet可﹒
We can of course relax the cons仕aint that () be independent of space. 1n so doing ,
we can entertain what happens under local rather than global (() independent of space)
transformations. While our analysis on the potential energy remains the same , the kinetic
energy,

T → jw) 州十七CÞl2 (圳))2 (1. 7)

does acquire a new term describing the spatial variation of the phase. 1f the U (1) symme仕y
is not broken by this transformation, then the second term must vanish. Demanding that

aμθ 止。 (1. 8)

requires that () be spatially homogeneous for the symmetηto be preserved. As a result , a

consequence of breaking the continuous U (1) symmetry is that () must be spatially non-
uniform. This is the situation in a superconductor. 1n fact , the current inside a superconductor
arises entirely 仕om the spatial variation of the phase , as can be seen from the quantum
mechanical equation for the current,

jμ= 竺1m叭μ 1/r=~主 1 ，6. 1 2 äμ() ， (1.9)

m m

if we interpret ψas the wavefunction for the superconducting state; that 站， ψ= b. e iθ .We
will see in the chapter on superconductivity precisely how this state of affairs arises. We will
interpret ψas the order parameter of a superconducting state. While the Bardeen-Cooper-
Sc趾ieffer (BCS) theory of superconductivity was certainly not formulated as an example
of a broken continuous symmetry , this is the basic principle that underlies this theory. 1n
fact , the key ingredients of superconductivity, charge 2e carriers and a supercurrent, all
follow from breaking U (1) symme仕y.
Massless bosons that emerge from broken symmetry typically generate new unexpected
physics. For example , phonons mediate pairing between electrons , thereby driving the onset
of superconductivity in metals such as Hg and more complicated systems , for example
MgB 2 . However, strict rules determine how such N ambu-Goldstone bosons can affect
4 Introduction

any system. As shown by Adler (A1965) , the interactions induced by massless bosons
arising from the breaking of a continuous symme仕y must be proportional to the transferred
momentum. More formally, interactions mediated by the exchange of a N ambu-Goldstone
boson can only obtain through derivative couplings. Consequently, the interaction vanishes
for zero exchanged momentum. This principle impliesthat the electron-phonon interaction
which mediates pairing in elemental superconductors is inherently dynamical in nature. We
will 而且有T this important principle in the context of the elec仕on-phonon coupling through
an explicit derivation. Hence , entirely 企om the existence of a lattice , phonons and the kinds
of interactions they mediate can be easily deduced.

1.2 T時 cki 臨 9 broken 5y臨醋的時;自rder pa時帥的er

The idea of an order parameter is another powerful concept in condensed matter physics.
Order parameters track broken sy虹rme仕y. That is , they are nonωzero in the broken sy虹立ne仕y
phase and zero otherwise. Consider a ferromagne t. Locally each spin can point along any
direction. This is the case at high temperature in which no symmetry is broken. In a phase
transition controlled by thermal fluctuations , typically it is the high-temperature phase
that has the higher symmetry. To quanti方 the order in a collection of spins , we sum the
z-component of each of the spin operators ,

../' --

Aυ
Il

、，/
M=jz(Sj) ,

、、
scaled by the number of spins , N. Here Sf is the z-compone凶 of the spin of the atom on
site i and the angle brackets indicate a thermal average over the states of the system. M is
the magnetization. At high temperature before any symme仕y is broken, the magnetization
is identically zero. At sufficiently low temperatur郎， the spins order and the magnetization
acquires a non-zero value. Consider iron for which the Curie or ordering temperature is
1340 K. It turns out that most parts of a block of iron below the magnetization tempera仇lre
have vanishing magnetization. This state of affairs obtains because the magnetization is
in general a function of space. As a result, a block of iron does not break the symmetry
uniformly. In fact , the actual magnetization in bulk magnets is not acquired spontaneously
but rather by some external means to align all of the individual magnetic domains. At the
boundary of a domain, the magnetization changes sign, creating a domain wall. Typical
domain sizes in iron are roughly 300 ions. Placing a chunk of Fe in a magnetic field will
orient all of the domains in the same direction, a state of affairs that will persist long after the
field is turned off. This is important since the re-oriented domain state does not constitute
a minimum energy state of the system. The domains lock into place by becoming pinned
to defects. One would expect then that as the magnetizing field is varie d, the magnetization
would not change continuously but by discontinuous jumps as domain walls de-pin from
defects. This is the essence of the Barkhaussen effect, the tiny discontinuous jumps the
magnetization makes in the presence of an external magnetic field and ultimately the reason
why the magnetization curve in a ferromagnet exhibits hysteresis as depicted in Fig. 1.1.
5 1.2 Tracking broken symmetry: order parameter 、

80

40

(2E
。
nu

1)這
-40

一2.5 0 2.5 5
H(kOe)

Hysteresis curve of the magnetization as a function of the applied field for a CoPtCrB thin film. Multilayer Co/Pt is used
in memory storage. Reprinted from Carpenter et al. , Phys. Re v. B72 , 052410 (2005).

That 白， ramping the field on and off is path dependent determined by which domains de-pin
sequentially.
What is crucial in the magnetic system is that locally there.are two degrees of freedom
for each of the spins. At high temperatu時， both states are accessible. At su血ciently low
temperatu間， one of the spin states is selected. Such state selectivity can be modeled with a
double-well potential ofthe form
--

、、‘'，'，
咀EI
f'a 4EEi

V(M)=-LM2+lyM4 , ••

、、、
2 4
where M is the magnetization and αand y are positive. The minima of this potential occur
at
厄hv

(1. 12)

Both of the minima are accessible at high temperature and no magnetization is possible.
Our choice ofα> 0 ensures that deviations of M away from M土 cost energy. As a result ,
the minimum energy of V is not zero but rather the non-zero value of 的 = -a 2 j4y. As it
stands , our theory is completely symmetrical with respect to the change M • -M. Surely
the physics cannot change if we were to recast our theory by shifting the scalar field M by
a constant such that M • M+ + cþ (x). The new potential
/ 1 3 吋\吋
V' 己的十 1\ 一一α
2 ~ 十一
2 yM~ ) cþ2
r H~+ J I
+ yM+cþ3 + ~ycþ4 (1 .13)

no longer looks symmetrical in terms of the new field cþ. Why? What we have done by
expanding around one of the minima of V is to hide the symmetry. Essentially we have
broken the symmetry by setting the magnetization to M+ . In the broken symmetry phase,
up and down spins are no longer equivalent. The field M functions as the measure of the
magnetic order. M is the order parameter. Unlike the old potential which was minimized by
6 Introduction

V(ψ)

伊1

Potential corresponding to the complex 5日 lar 有 eld <p. The minima correspond to circles satis句ing
<p i + <pi = α jy.
a non-zero value of M , the minima of V' take place at ø(x) = O. This corresponds to a 仕ue
vacuum. In classical models for ferromagnets , the magnetization turns on continuously,

MαIT - Tc 尸， (1. 14)

at a non-zero temperatu間 ， Tc. It acquires its maximum value at T = O. The exponentαlS
the critical exponent for the turn回 on of the order parameter.
A slightly more complicated example of broken symme仕y occurs when we modi命 our
potential,
V(伊)=一αψ*ψ + y(伊*伊 )2 ， (1. 15)
叫low :D臼
or cωC
a ∞O蚓ex 叫叫el旭
dψ: 吉 (ψ
仙1+ 圳 Fo
叮rα>0 削 y > 0, 仕伽
t1t
is i1lustrated in Fig. 1.2. The corresponding Lagrangian takes the form

乙 =•: M )川 … ( 叭 +竹叫
y州(圳
Our Lagrangian has the 剖 g 10
吋 ba
叫1 叮 m時et缸ry ♀ψ9 →♀ψ9冗e i8θ ， where e is a constant. As a result of
s ym虹
妙

this symme旬， the minima of V now take on a circle of vah闊別i的ing 討+后 =α 仰，

as illustrated in Fig. 1.2. That 站， there are infinitely many saddle points as a result of the
continuous global symme甘y. As before with the single scalar field for the magnetization,
we can expand about the circular minima by definingψ 口(，冬 + f(x) + ig(x)). That is,
、/ Y
we break the symme仕y by hand.Because at the minima 9月 and 的 are not independent , this
仕ansformation is not a simple translation of <P l and 仙 separately. As a result, the quadratic
term essentially has only one degree of 企eedom. We can interpret this as a vanishing
of the mass for one of the scalar fields , in line with Goldstone's theorem that a massless
mode must emerge upon the breaking of a continuous symmetry. Such a solution in which a
complex field acquires a non-zero value is the heart ofthe superconducting transition. What
the Bardeen-Cooper-Schrieffer solution laid plain is that the phenomenological Landau-
Ginzburg treatment in terms of a complex order parameter acquiring a non間zero value in
the superconducting state has a microscopic basis in the electron-phonon interaction. Such
an interaction mediates pairing and the order parameter for the superconducting state is a
product of an amplitude for pair formation times e iθ (r) ， where e is the phase of the pair
field.
7 1.3 Beyond broken symmetry

Energy-Ievel diagram for the ground sta妞， Eo ， and first excited state，日， as a function of a coupling constant, 9. The
crossing at 9c signals a phase transition between the ground and the 莉的t excited state.

Our statement that the low-temperature phase 句pically has lower symmetry is only 仕的
classically. There are many examples of symme仕y breaking at T = 0 that have nothing
to do with thermal fluctuations. Such phase transitions are governed by fluctuations of the
vacuum, that is the uncertain句T principle. In general , such phase transitions are governed
by a transition among the quantum mechanical states of a many-body system simply by
changing some system parameter. Consider a 耳amiltonian H (g) , where g is a coupling
constant. A typical energy-level diagram for this system as a function of g is depicted
in Fig. 1.3. If, as a function of g , the first圓 excited and the ground states cross , a phase
transition obtains to the first吋xcited state. For the transition to be continuous , we must have
that 3Ej3g = o. These types ofsituation are discussed explicitly in Chapter 14.

1..3

Despite the utility of symmetry in classifying collective phenomena, physics is replete with
examples of transitions between states of matter that share the same symmetry but are ,
nonetheless , distinct. An obvious example is the liquid-gas transition or the formation of
the fractional quantum Hall state. However, the particular examples we focus on here , which
typi命 the physics of s甘ong coupling , are those in which the formation of some kind of
bound state is the distinguishing feature. Consider, for example , the vulcanization or cross-
linking transition in rubber. In the un-vulcanized state , rubber is a viscous liquid in which
long-chain monomers move independently. Cross-linking between neighboring monomers ,
resulting in the formation of a highly entangled enmeshed amorphous state , defines the
vulcanization transition. Although the monomers are localized in the vulcanized state , they
are randomly located. Consequently, there are no Bragg peaks. Nonetheless , one can define
an appropriate order parameter (GCZ 1996) which reflects the fact that at t = 0 and t = ∞，
the configuration of the monomer strands in the liquid changes while it is essentially static
in the amorphous state. The resulting resilience and emergent static modulus of rubber
both arise from the effective gluing together of the monomers. In high-energy physics ,
mesons or bound states of quarks are the propagating degrees of freedom at low energy
in nuclei. They arise without the breaking of any continuous symmetry. In problems more
8 Introduction

relevant for this text , a magnetic impurity in a non-magnetic host forms a bound state
with all the conduction electrons below a characteristic temperature , once again without
breaking any symmetry or even inducing a phase transition. The formation of such bound
states is the essence of the Kondo problem which stands as one of the key triumphs of
the renormalization group principle. As we will show 企om a systematic integration of
the high-energy degrees of 企eedom， a bound state emerges because the coupling constant
between the impurity and the host electrons diverges. Hence , at low temperatures it is not
correct to think of the magnetic and conduction electron degrees of freedom separately.
A new entity emerges at low energies that is not present in the starting ultraviolet (UV)-
complete Lagrangian, a characteristic feature of strong-coupling systems. Although new
degrees of freedom are present at strong coupling which are absent in the weakly coupled or
high-temperature regime , the two states are still adiabatically connected in that by varying
the system parameters , one can go smoothly from one phase to the other. Nonetheless ,
the phases are quite distinct. They possess different degrees of freedo血， and no unified
description exists of such systems in terms of a single entity. Bound-state formation is a
standard paradigm in strong coupling physics and，的 we will see , the Mott problem, an
insulating state in a partially filled band, is no exception.
Another key example is Fermi liquid theory. The primary tenet ofthis theory is that the
excited states of a metal stand in a one-to-one correspondence with those of a non“ interacting
electron gas. The interactions in a metal are of course non-zero. However, they are strongly
screened and can be treated as essentially short-ranged. The Landau (L1957) assertion is
that all such short-ranged repulsive interactions do not destroy the sharpness ofthe electron
excitations in the non-interacting electron gas. That this state of affairs obtains is perhaps
one of the most remarkable principles in nature. Why can the short-range interactions be
ignored in a metal? The answer lies in a fundamental renormalization principle which
we present in Chapter 12. The key to solving any many-body problem is to identify the
propagating degrees of freedom. Identifying that the propagating degrees of freedom are
"single electrons with a dispersion relation given by p2/2m in an interacting electron gas
is highly non-trivial. In fact , it cannot be deduced directly from the Hamiltonian. Some
further fact is needed. That further fact is the existence of a Fermi surface. As we will see ,
the fundamental principle that makes Fermi liquid theory work is that a Fermi surface is
remarkably resilient to short-range repulsive interactions. We will demonstrate explicitly
that all renormalizations (PI992; SM1991; BG1990) arising from such interactions are
towards the Fermi surface. As a result, such interactions can effectively be integrated out,
leaving behind dressed electrons or quasi-particles , thereby justifying the key Landau tenet
(L1957) that the low-energy electronic excitation spec仕um of a metal is identical to that
of a non-interacting Fermi gas. Consequently, breaking Fermi liquid theory in dimensions
greater than or equal to two is notoriously difficult. In one spatial dimension, interactions
are always relevant, as will be seen, and a new state of matter arises , termed a Luttinger
liquid, in which spin and charge move but with different velocities. In higher dimensions ,
the problem is open and stands as the key outstanding problem in solid state physics.
As Fermi liquid theory made the BCS theory of the superconducting state possible in
that it cleanly identified the propagating degrees of freedom , a similar identification of the
propagating degrees of freedom in the normal state of the copper-oxide high-temperature
9 References

superconductors is necessary to know what pairs up to form the superconducting conden-

sate. This prob1em is particu1arly di且cult as the parent materia1s are all antiferromagnetic
Mott insu1ators. Some ofthe agreed-upon physics of this remarkab1e prob1em and a forward-
1eaning perspective are discussed in the fina1 chapter of this book.

Refere臨時S

[A1965] S. L. Ad1er, Phys. Rev. B 137, 1022 (1965); Phys. Rev. 139, B 1638 (1 965).
[BG 1990] G. Benefatlo and G. Gallavot位，J. S，的t. Phys. 59 , 541 (1 990).
[GCZ1996] P. M. Go1dba肘， H. E. Castillo, A. Zippelius, Adv. Phys. 45, 393 (1996).
[G1961] J. Go1dstone , Nuovo Cimento 19 , 154 (1961).
[N1960] Y. Nambu, Phys. Rω~ 117, 648 (1 960).
[L1957] L. D. Landau, Sov. Phys. JETP 3, 920 (1957); 5, 101 (1 957); 870 (1 959).
[P1992] J. Polchins姐， arXiv:hep-th/9210046.
[S孔11991] R. Shankar, Physica A177, 530 (1 991).
 At the close of the previous chapter, we noted that one can understand the elementary
properties of metals in terms of non-interacting electrons and phonons. For example , the
low-temperature specifìc heat of a metal is the sum of a term linear in the tempera仙詞 ， T ，
from the electrons and a term proportional to T 3 企 om the phonons. This result follows
from a non-interacting particle picture. The electrical conductivity limited by non-magnetic
impurity scattering is also well described by a non-interacting electron gas. In addition,
from a knowledge of single-electron band theor耳 one can discern qualitatively the diι
ferences between metals , insulators , and semiconductors. The remarkable success of the
non-interacting model is paradoxical because electrons and ions are strongly interacting
both with themselves and with one another. Along with its successes , the non-interacting
picture has colossal shortcomings , most notably its inability to describe old problems such
as cohesive energies , superconductivity, magnetism, and newer phenomena such as doped
Mott insulators , the Kondo problem, and the fractional quantum Hall effects. We fìrst review
the physics ofthe non-interacting electron gas. It is only after we develop methodology for
dealing with electron interactions than we can lay plain the reasons why the non-interacting
model works so well.
Electrons in metals are quantum mechanical particles with spin fï/2 obeying Fermi-Dirac
statistics. The Hamil扯tωonian ofa 剖 s ingle elect仕ron 臼
1 Sp2/ρ2mwher臼e 富 i臼st由
he electron momentu
(仰
opμer叫前呦叫
O r) and m 伽
the 耐
e圳
lec
e臼忱
ct
叫訓
t仕r昀
肋
∞
O nmas臼s. Its eigenstates are plane waves ofthe form eip .r / 1ï /.JV
times a spinor which speci在es the electron spin projection on a convenient axis (usually 2),
先σ/2 where σ: 士 1; here V is the system volume. The Hamiltonian (operator) for N such
non-interacting electrons ,
N 合至

H=L 氛， (2 .1)

is simply the sum of the kinetic energies of the individual particles. In this case , the
eigenstates are products ofthe occupied single-particle plane-wave states. Each plane-wave
state can be occupied at most by one electron of a given spin. We label these eigenstates
by the distribution ft肌tion fpa , which is 1 if the single-particle momentum-spin st前e
is occupied and 0 otherwise. In the ground st駒， the lowest N/2 single且particle st前的
are doubly occupied with electrons of opposite spin. Consequently, in the ground state
(temperature T 口 0) ， the distribution function is

fpσ= e(μ。一 p2/2m) ， (2.2)

where e(x) is the Heaviside function , e(x > 0) 1, and 0 otherwise. Here μo is the
zero-temperature electron chemical potential , which in this case is simply the Fermi energy,
10
11 Non-interacting electron gas

Li Na K Rb Cs

rs 3.25 3.93 4.86 5.2 5.62

the energy of the highest occ叩 ied state , p~/2 肌 where PF is the e1e∞ct
叫仕ronFe叮rm
盯
ml吐i 玟lOmen
前肌叫
tu
um
叫
The Fermi 紀t empe叮ra拍枷仇仙1虹re 丹
1全 eq
伊ua
站ls μ o/kB .
In terms of fpσ ， the tota1 number of e1ectrons is given by

N=工九σ (2.3)
p，σ

In the ground state, we can rep1ace the sum by an integra1 and find the e1ectron density at
T =0:
N (T = 0) 2" '"1 [PF dp p~
巴= - ,- -/ = - ') . = 2 I 一一一-71-TT (2 .4)
v v 行于
V"--PF
o (2π fï)j 2 fïj3n

The average interparticle spacing is essentially the radius ，凡， of a sphere containing a sing1e
e1ectron,

4π r~
一一一-=-ne = 1 (2.5)
3
Thus , from Eq. (2 .4), the sca1e for the interparticle separation is

re = 問 1/3tzl 位 (2.6)

which is on the order ofthe 1attice spacing. For re 勾 1 rA , we find that the Fermi ve10city
VF = PF/m ~ !ï/m几何 108 c m/s 勾 c /300 , where c is the speed of 1ight. (Re1ativistic
effects are generally not important for the motion of e1ectrons in the ground state.) It is
conventiona1 to work with the dimensionless ratio rs - 几/旬， where αo =妒 /me2 is
the Bohr radius; this quantity provides a measure of the e1ectron density. The dense limit
corresponds to rs << 1 and the di1ute regime to rs >> 1. In meta1s , rs varies between 2 and 6.
Listed in Tab1e 2.1 are va1ues of rs for the a1ka1i meta1s. Cesium is , in fact , the most di1ute
of all meta1s. It is this 1arge va1ue of rs that is responsib1e for the inhomogeneities in the
density of Cs. In Chapter 5 , we will discuss further physics associated with 1arge 門， such
as the eventua1 formation of a Wigner crystal. We can a1so use Eq. (2 .4) to solve for the
zero-temperature chemica1 potentia1
主之2

μ0=(3 的 2/3 已':__

Lm
n; /3 (2.7)

ForNa， μo = 3.1 eV; typically in a meta1 , f-L o ranges between 1 and 5 eV.
12 Non-interacting electron gas

The total energy of the system is the sum over the occupied states weighted by the
single-particle energies ，句 = p2/2m:

E=L 石山σ (2.8)

p，σ

At T = 0 , the energy is given by

Eo = .10mπ
~L
p2 V
r= -;Nμ。
') ,_'J
3
(2.9)
2ñ} 5

From the thermodynamic relation, P = 一 (ðE/ðV )r， N ， we find the pressure in the ground
state , Po = 2μone/5. This quantity is of order 106 atm and arises entirely from the exclusion
principle between the particles.
At finite temperature , we de曲時 the distribution function which ranges between 0 and 1
and measures the average occupation of the single “ particle states. For a system in equilib-
rium at chemical potentialμ ，

!pσ- eß(E p 一 μ )+1 (2 .l 0)

is the Fermi-Dirac distribution, where ß = 1/ kB T. The Fermi-Dirac distribution function

maximizes the entropy at a given energy and electron number. In general , the entropy, S, is
given by the log of the number of microscopic states W ,

S = kB lnW, (2.11)

consistent with the macroscopic thermodynamic state of the system. For the electron
problem, the microscopic states are indexed by a momentum p with occupancy 0 or 1. The
distribution function !pcJ' however, is a smooth function over all the momentum states. To
construct this function , we group the momentum states into cells , each cell containing gi
momentum states and ni particles. Because each cell contains gi states ,

Lgi'''=L (2.12)
p，σ

For the ith cell , the number of distinct ways of distributing ni particles in gi states is given
by the combinatoric factor wi = gi! /ni! (gi - ni) 1. Applying Stirling's approximation

lnN! ~ N (l nN - 1) (2 .1 3)

to 阱， we obtain
gi-ni
ln Wi 記 -ni ln 立一 (gi - ni) ln 一一一一 (2 .1 4)
gi gi

-gd 生 ln 全+ ( 1 一叫 ln ( 1 一叫| (2 .1 5)
Lgi gi \ g i/ \ g i/ J
where ni/ gi is the fraction of states occupied in cell number i. In fact , ni/ gi = fi is the
smooth distribution function we seek. If we substitute this expression into the equation for
13 Non-interacting electron gas

ln 阱， we recover the familiar result for the entropy,

S= -kBL“σlnfpσ+ (l -fpσ) ln (l - fpσ)] ， (2.16)

where we have converted the sum over cells to a sum over spin and momentum states by
using Eq. (2 .1 2). To obtain the distribution function , we maximize the entropy subject to
the constraint that the particle number and energy are :fixed. Extremizing S - E j T + μNjT
with respect to fp σ ， 附
we n
1 岳缸

0= ε句pcJ
恥σ 一→
μ一k
站BT
川 ln (協
4l 一 1) , (2 .1 7)
which implies 由前 fpσis the Fermi distribution function in Eq. (2.10)
Let us now calculate the heat capacity from the thermodynamic relationship
/ âS\
Cv =T ( --;::;;;) . (2.18)
\âTJv
To compute the temperature derivative ofthe entropy, we consider the general variation of
the entropy,

8S = -2kB γ 8fp ln 一企一 =2如 γ 8fp 生二主 (2 .1 9)

γ t' 1 一九 γ t' kBT
/ ap 石p 一 μ
= 2VkB JI 一一一τ 叫一一一，
(2π {ïy.'> J t' kBT
(2.20)

with respect to the distribution fi肌tion fpσ ﹒ We can simplify this expression further by
introducing the single-particle density of states per unit volume

N(ε) = 2 {
J
/~ d~，
(2π 的.'>
'l 8 (在一句)
l"
(2.21)

1 r 外 dv
- "，τ I pL-!-8(在一句 )d句 (2.22)
π L.{ï.'> J '- d在 P t'/ l"

mv
N(Ep ) =τ云. (2.23)
aπ "'n-'

The key point here is that the single開particle density of states is a linear function of
momentum. An equivalent way of writing this quantity is N (句) = (dpjd句 )(p2 jπ2 先3).If
we rewrite 8S in terms ofthe density of states ,

8S=; f 蚓削(Ep)(E p 一仇 。立的

we obtain a single integral that can be evaluated using the Sommerfeld expansion. While
this expansion is standard, we will review it.

Sommerfeld expansion
Consider an integral of the form

1= f 的州ε) (2.25)
14 Non-interacting electron gas

where h(ε) is any smooth function and f (ε) = 1/(e ß (在一μ) + 1). We integrate 1 by p缸ts:

1= f 州州 k (2 .2 6)

where

H=- l' 你) (2.27)

Because f'(ε) is strong1y peaked at the chemica1 potential , we can expand H in a Tay10r
series around ε=μ，

的μ)+ 刊) (的J (2.28)

which gives us a series of integra1s of the form

令 =-1∞ (ε 一川ε (2.29)

to calcu1ate. For j = 0, Lo = - f(∞) + f(O) 立 1. In the remaining integra1s , we can

一
rep1ace the lower 1imit with 一∞. Lettingx =戶 (ε 一 μ) ， we have

d …
l-F
fil
LJ JJ 'u
一­

(2 .3 0)

Since the integrand is odd in x for j odd, only the even j' s survive. The first severa1 va1ues
of Lj are

Lo = 1, (2 .3 1)
一2
L 2 = 了(如 T)2 ， (2 .3 2)
'ï 一 4
L4 = 二~(kBT)4. (2 .3 3)
15
Consequently, we systematically obtain the series expansion for 1:
fμ7 月

1= I
Jo
r h(E)dε+ 主(如 T)2h' (μ)+LEhrfhfff(μ)+
6 ，~/ ,,/ 360
(2.34)

For functions h independent oftemperature , the first term in this expansion is independent
of temperature and hence , at fixed μ，

f “f州 (2.35)

is the 1eading term when fp is varied as a function of temperature.

Now 1et us return to the calcu1ation ofthe low-temperature entropy. We want to calcu1ate
the temperature variation of the entropy at fixed particle number. We first calculate at fixed
μand then show that the change in chemica1 potentia1 with temperature for fixed particle
number can be neglected here. Ifwe now substitute Eq. (2.35) into Eq. (2.24) , we find that
15 Problems

the variation of the entropy per unit volume ,

以 -3
局
N T

一-
pAU
qu CLEI CO
(2 .3 6)

is a constant independent of temperature. The heat capaci可 per unit volume ,

= T (;三) =主如何)T， (2.37)

\ 8T ) v 3 D

scales as a linear function of temperature. This contribution arises entirely from the con-
duction electrons. The contribution per electron is

Cv 'J( 2k~mT 丌 2L T
(2.38)
ne = fi = 2κB 冗

Comparing this result with the classical heat capaci旬， 3kB /2 , we find that the quantum
mechanical value is smaller by a factor of 'J( 2T /3TF . In a metal, only a fraction of the
electrons are at the Fermi level. The ratio T / 丹 defines this fraction of electrons within
kB T of the Fermi energy. The further an electron is below the Fermi level , the smaller is its
contribution to the heat capacity.
We may use the Sommerfeld expansion to show that the chemical potential as a function
of density and temperature is given, at low T , by

/π2 ! T \2 、
μ(慌 ， T) = μ (n e ， 0) I1
\
一~ 1,
12 \ TF } I
, :" , (2.39)

where μ(慌， 0) =句﹒ The proof of this result is similar to that used to derive the en仕opy，
and we leave the derivation as an exercise (Problem 2.2). From Eq. (2.39) we see that the
first correction to the chemical potential at fixed ne is of order 戶， and hence this correction
can be ignored in calculating the low間temperature entropy, Eq. (2.36).

Pr串bl會酷$

2.1 Evaluate the integral

f∞ ex

Lj = J一∞ XJ(ex+1)2dx

Show in particular that only the integrals for even js survive.

2.2 U se the Sommerfeld expansion to compute the first temperature correction, Eq. (2.39) ,
的 the chemical potential of a Fermi gas as a function of the density.
 In this chapter, we develop the basic framework to see how elec仕on-electron (e-e), electron-
ion (e-i) , and ion-ion (卜i) interactions affect the properties of solids. We fust show that the
elec仕on and ion degrees of freedom can be decoupled. Such a separation arises because
ions and electrons have vastly different velocities in a solid; rough趴 the electron velocity
is 1000 times larger than the ion velocity. As a consequence , the electrons view the ions as
providing a static background in which they move. This physical pic側記 is at the heart of
the Born-Oppenheimer approximation (B01927).

3月 Basic Ha 臨時to詞 ia間

It useful at the outset to consider the total Hamiltonian to understand how to separate the
electronic from the ionic motion. We denote the position and momentum of the ith ion by
Ri and P i , respectively, and that of the jth electron by rj and Pj . We assume that all the
ions have mass M and nuclear charge Ze. The total Hamiltonian is then

H=L 五 +EJz+
P;
一~L
,'" P] , (Ze)2
j ij|Rif-RIγ|

+勻|月 1 抖f「 MZ|JRI| (3.1)

This Hamiltonian does not include extemal electric or magnetic fields or magnetic inter-
actions among the constituents. The last three terms are the i-i , e-e , and e-i interactions ,
respectively. To progress with this Hamiltonian, we divide the electrons into two groups
the core and the valence or conduction electrons - depending on the degree to which they
are bound. This separation is useful because core electrons move with the nuclei. Conduc-
tion or valence electrons 甘ansport throughout the solid. With this separation in mind, we
obtain

H= 丈丟 +ZLUi(|Ri-RI|)

心 jjfIZdelec|山JPi仲
16
17 3.2 Adiabatic approximation

。 o --E-- B 一----0

---ò--
。 \v 。

。 。 。
lon lattice with spacing a. Each ion oscillates in a harmonic well with a deviation from its home position that is small
relative to the ion spacing. The deviation is roughly ö rv (m /M) 1/4 a rv 1 。一切， where mis the electron mass and M
the mass of the ions. It is for this reasonthat the ions 日 n be treated essentially as 有 xed relative to the electronic
degrees of freedom. That the ions comprise a fixed , almost rigid background for the electron motion is the essence of
the Born-Oppenheimer approximation.

as the partitioned Hamiltonian. In Eq. (3.2) , Vi,i and Vei represent the effective potential
l

between the ions and the valence electrons with the ions , respectively. The energy of the
core electrons is Ecore. In the example of Na, the total Z is 口， with the orbital filling ls2
2S2 2p 6 3s. There is only one unpaired valence electron. The effective charge ofthe ion in
this representation is Z = 1.

3.2 Adiabatic approxi 閥割io自

To simpli勾T Eq. (3.2) , we separate the nuclear motion from that of the electrons , a sep-
aration that makes sense because the ions are much more massive than the electrons.
Typically m/M rv 1/2000 to 1/500000. The small parameter characterizing the expansion
is (m/M) 1/4. We now show that the ion velocity is related to the Fermi velocity by the ratio
(m/M)3 j 4. As a result, the ions can be treated as essentially static relative to the electrons.
This allows us to solve for the electron motion assuming 且rst that the ions are fixed at their
equilibrium positions. The effects of the ion motion can be treated as a perturbation; that
is , the electrons a吐just adiabatically to the ion motion. From the perspective of the ions , the
rapid motion of the electrons creates an overall average electron potential which they fee l.
This separation is the essence of the Bom-Oppenheimer approximation.
To understand the relative orders of magnitude of the ionic and elec仕on velocities and
energies , we assume , for the sake of the argument, that ions individually move in harmonic
wells (see Fig. 3.1) ofthe form 九sc 口 Mω2 R 2 /2 , where R measures the deviation of an ion
企om its home (or equilibrium) position. Consider displacing an ion by a lattice spacing,
α. The energy required to do so , rvλdω2α2/2 ， is essentially that required to distort the
electron wavefunction, and thus the energy is of order fî2 /2ma 2 , which in tum is of the
order of the electron kinetic energ耳片 /2m. Thus Mw 2 a 2 /2 rv p~/2m. Solving for 的 we
find that

ωrv (m/M)1/2 土 (3.3)

m α中
18 Born … Oppenheimer approximation

However, for an ion in a harmonic well , P2/2M = 缸。 /2 ， or equivalently, the square of the
ion velocity is ñw/M. Combining this result with those for ω， we see that

Vion rv (mjM)3/4 vF rv (1 0- 2 to 10-3)vF . (3 .4)

Let us estimate how far the ions move 企om their equilibrium positions. For a displacement
8 ， Mω28 2 /2 "-' ñw/2. Substituting ωrv (m/M) 1/2 ñ/ma 2, we find an ion displacement,

8 rvα (m/M)1/4 rv 10一勻， (3.5)

which is negligible ，的 illustrated in Fig. 3.1. As far as the electrons are conceme d, the ions
are static. We can calculate the role of the electronic degrees of freedom by developing a
perturbation series in the small quantity 8/αor (m / M) 1/4. N ote also that because P2/2M rv
他)/2 rv 句 (m/M)1/2 <<句， the ion kinetic energy is relatively small.
The formal development of the Bom-Oppenheimer approximation begins with the as-
sumption that the full wavefunction is a function of the many-electron positions r 三 {rj}
and ionic positions R 三 {Ri} ， and can be expanded as

中 (r， R) = L φn (R) \IIe ,n (r , (3.6)

n

where the \IIe， n 付， R) (indexed by n) are the solutions to the electron-ion problem for a
fixed set of ion positions R. The ionic wavefunctions ， φn (R) , on the one hand, describe
the amplitude for the ions to be found at positions R; on the other hand they can be re-
garded as expansion coe:fficients of the electronic wave加lctions. The \IIe ,n form a complete
orthonormal set. Consequently,

f 叫n (r , R) 中e，m (r , R) =川) = 8nm (3.7)

Coupled with the orthogonality condition on the nuclear wavefunctions ,

f dR<Þ~帥m(R) = 8nm (3.8)

the complete electron-ion wavefunction is normalize d,

(\11付， R) I 中 (r ， R)) = 1.

We determine each of the expansion coe :fficients from the equations of motion obeyed by
φ n and \11巴 ， n' To procee d, we rewrite Eq. (3.2) 的

H= Ti+ 丸 +Vii+ 九e+ 九i + Ecore , (3.9)

19 3.2 Adiabatic approximation

where there is a one-to-one correspondence between the terms in (3.9) and those in (3.2).
The eigenvalue equation for (3.9) is

(1i +丸+仇+九e 十九i + Ecore) \Þ = E \Þ (3.10)

(1i +只i + Ecore) 中+主 φn(足十几十九i) 中e，n(r ， R) =凹 (3.11)
n

We simplify this equation by noting that 1'e +九e + Vei on1y operates on the electron part
of the product wavefunction. Let Ee ,n(R) be the energy of the e1ectron system for a fixed
set of nuclear coordinates. As a consequence , the nuclear and the e1ectronic eigenvalue
equatlOns ,

主(1i +只i + Ecore + Ee ,n - E) φ品，n = 0 , (3.12)

n

and

(1'e +九e+ 九i) 中e， n(r ， R) = E，巴， n(R) 中e， n ， (3 .1 3)

can be separated.
Let us mu1tiply the nuclear eigenvalue equation by \Þ斗= (emlr , R) and integrate:

平 j 叫m 叫 1i<Þn(R叫什) + (只i + Ecore + Ee m (R) 一恥m (R) =。 ,

(3.14)

Because the matrix element (en Ij:-'ù (R) Iem) involves purely algebraic operators , it has only
diagonal elements. The kinetic energy term, however, has off-diagona1 elements; explicitly,

別的|品問

+叫叫
2川(v
可叭Rii

Because 苑
V孟 acωt 臼S 閃1us肘
u站
si
S位i的
如
ve
吻O∞nφ俄恥叫
nλ1(R
酌很酌
) in
叫 伽
t1he
旭e fir削 term 血
in
削t伽
蚓
he
1昀el帥g叫，
叭 帥
the
昀e 附u址枷
l
element 臼
i s pur
昀 el妙
y diagonal. For the moment , we ignore the last two terms in the kinetic
energy matrix e1ement and obtain

L (1i + Jíii 十丸。re + Ee n (R))φn (R) =

, L ιφn(R) ， (3.16)
n n

or, equivalently,

[1i +只i+Ec的+丸，n (R) ]φI1 (R) =丸 φI1 (R) ， (3.17)

的 the eigenvalue equation for the nuclear degrees of freedom. Equations (3.13) and (3.17)
are the principa1 results in the Bom-Oppenheimer method. In the nuclear eigenvalue
equation, E，巴， 11 (R) serves as the effective nuclear potentia1 that results when the electronic
degrees of freedom are integrated out. The solutions to (3.17) will describe the phonon
modes of the ions.
20 Born-Oppenheimer approximation

To justify this 仕eatment ofthe ion kinetic energy term, we analyze the relative magnitude
of the three contributions in Eq. (3.15). We first need a reasonably accurate form for
the nuclear wavefunctions. For our purposes , the harmonic approximation we made in
conjunction with the derivation of the ion velocity is adequate. In this case ,
φ n '"'-J e…Mω(R_RO )2/ 2h ， (3 .l 8)
where R O is the equilibrium position ofthe ion. Consequently,
民2 '1 ñ2 (Mω\ 2 / m \ 1/2
一u
2且![ R;
V K; <Þ "n ,n '"'-J一:-::1 一;::--8 I <Þ nWe.n
. Wev ," 2M \ ñ . J - " - v , H
'"'-J 1-:-:: 1
\M J 句 φ nWe.n. (3 .l 9)

Consider now the second term in Eq. (3.15). The inverse length scale on which the electron
wavefunctions change is VR '"'-J 1/α. As a consequence ,
先2 民2Mω8 / m \ 3/4
一;-; V R ; <Þ n ,n '"'-J一一一一-\丸， n 趴'"'-J I 一)
2且![ n , " . V R,; We ,"
u v 2M ñ a , " " \M J 在F <Þ n We ,n
v
(3.20)

and
民2 吋民21m
一一
2M φV主
- fl . K; W戶..一一一恥
- ", fl
"φ.. '"'-J M
2M α2 - 'V, rt - fl
-Evφ fl"， Wρm
- 'v月 -l_" -
(3.21)

As is evident, the largest contribution to the nuclear kinetic ener喀gy matrix element arises
f企kOm ViiI? 丸 趴趴n. The prima
φ 叮ry reason why 吐thi
由臼S叫 m
叭i臼st由
hat
叫前tg叫i白er刮
a lts of φ n 侃
exceed 由
those of ψ削
by a f:臼ac叫to叮r of (σJ且M/mη1)1/4. Consequently, dropping the last two terms in Eq. (3 .l 5) incurs a
negligible error on the order of (m/M) 1/4.
To lowest order in m/M , we then neglect the ion kinetic energy and assume R = R O•
The ground electronic wavefunction is We， n=O 付， R 0 ); its corresponding energy, E巴， n=o(RO ) ，
obeys the Schrödinger equation
O
[丸+九巴十九i(r - R )] 中e， n=O (r , R O
= Ee , n=O (R ) 中e， n=O 付， R
O O
) ). (3.22)
The subsequent ground間state nuclear wavefunctions and energies can be found with the
effective potential E巴 ， 11口。 (RO ). If deviations about the equilibrium ion positions are consid-
ered within a simple harmonic oscillator model , the total energy of each low-energy state
is given by

En = Ee.n(Ro) + E叫民 (RO ) + :L ñ州 q + ~) +叫armonic terms 叫

q 、，

3.3 Ti會輸-bi 聞組 i 時晶 ppr甜甜臨銅牌時

From our preceding discussion , the effective electron problem is

H已= Te +Vee 十 J!ion (r) , (3.24)

where

几n(r) =丈九i(rj - Rn (3.25)

l ,}
 nrnrAUVA m+tnunH

γ-2
+t inunHJHMnHnH3
EHH

、〈

、〈
nu3

、d

、d
V目，
J
J
21

Here Vï on (r) is the potential felt by the electrons produced by the ions in their equilibrium
positions. To a good approximation, this potential is periodic. Let us group all the one-body
terms together as he(r) 口五十Vïon (r). The reduced electronic Hamiltonian is then
H巳= he -lγ Vee ﹒ (3.26)
It is this Hamiltonian that we will primarily discuss in the remainder of this book.
A closer look at the one-body part of our Hamiltonian is warranted. In the previous
chapter, we 仕eated the electrons in a metal as if they were free particles not tethered to
any ions. Of course this is wrong. Electrons know about the ions and interact strongly with
them. Why then is it possible to regard the electrons as being free? It turns out that had
we started with the localized picture advanced by Wannier, we would still end up with the
same answer. The solution to the one-body part of the Hamiltonian is described by a set of
periodic Bloch waves ,
的叭的= e ik 叮nk (r) , (3.27)
for the nth band with momentum k. Here Cnk (r) is a function that has the periodicity ofthe
lattice. Since the Bloch functions are complete , we can form a state localized on a particular
lattice site R ,

峙， r)=hze叫nk(r) (3.28)

www
的
mm1-Nl-No句
對dR
白 wy

由
BH
w川
叫衍三

叫川
位m

咕"的

mR 'm mzJ
J-3

nitR e huky
r hd k x u
a'!k
叫到

(+ s
閃出

fk
eWAKm
R R

UM
PAC a i
MAnRK
9urAf--

a'

e
C U ohH

e
4

T且
可且口、 ujλ

: L

口b
4
UUiW 4
1 .I ri
•••
f
qu
rr

••• •••
r ffd

ZMZMn的
fij-e

…
r- *mk
d

一一一一一…

r r
r

/s 、、
，切

、，

••
12
、

2
仰

司3
/'

、tf
QVJ
E，

'
Z
、、、

as can be seen from a direct calculation. The statement that Wannier states form an or闖
thonormal basis is somewhat problematic. If we were to rewrite he in terms of the Wannier
states ,

he = 芝 IR')(實 IheIR)(酌， (3.30)

R. R'

by using the completeness relationship on the state vectors associated with the Wannier
states , IR) , we would have to conclude that there are no off-diagonal matrix elements of
our Hamiltonian. That is , if the Wannier states on two neighboring sites do not have any
overlap , any matrix element with the Hamiltonian should be zero. The statement that the
Wannier basis forms an orthornomal basis in practice means that beyond nearest neighbors ,
22 Born … Oppenheimer approximation

the wavefunctions have no appreciable amplitude. This is the tight-binding approximation

and the Hamiltonian that results is termed the tight-binding (TB) Hamiltonian. Restricting
ourselves to nearest neighbors , we define the parameters in our TB model as

(RlheIR) =. Eo (3.31)

and the nearest回neighbor matrix element as

(RlheIR+ 8) 告 -t (3.32)

where 0 is a unit vector from site R to the nearest neighbors. Our Hamiltonian simplifies to

H
TB
= -t L IR)(R+ 81 +Eo L IR)(RI (3.33)
RδR

Regardless of the form of the underlying lattice , the tight-binding model can be solved
exactly by Fourier transforming,

IR) = ~L 戶 Rlk). (3 划
v ~. k

The resultant Hamiltonian,

HTB=L 烈的 Ik)(kl ， (3.35)

k

depends on a single parameter, namely the band structure

E(k) = 此 - t Leikð (3.36)

The energy band, E (k) , de宜nes the energy of a non-扛lteracting particle with momentum k
and hence is the TB analog of the free-particle dispersion. The TB model can be viewed as
the lattice analog of the free-parti cI e problem.

3.1 This problem concems the band s仕uc加re of a single sheet of carbon, graphene (see
Fig. 3.2). Nearest-neighbor atoms are connected by the set ofvectors

01 = ~ (1 ，叫， 02 = ~ (1 ,-v'3), 03 =一α(1 ， 0) (3.37)

and the primitive lattice vectors are

1 = ~ (3 ,v'3) , a戶~ (3 ,-v'3), (3.38)

23 References

kx

Graphene lattice. The panel on the left shows the primitive lattice vecto時， ai， and nearest-neighbor lattice vectors ð i . ,
The labels Aand Brefer to the two kinds of carbon atoms in graphene. The panel on the rig 忱的 ows the first Brillouin
zone spanned by the lattice vectors bi .

where a = 1.42A , the nearest-neighbor carbon-carbon spacing. The first Brillouin

zone shown in Fig. 3.2 is spanned by the reciproca1 vectors
2π 1 ，--\ 27τ I ,--\
b1 = ~ ( 1, v3 ) , b2 =γ( 1, -v3 ) . (3.39)
5a \ I 5α \ I

The comers of the first Brillouin zone are 10cated at

K= 三 (1，吉)， K' =~: (1 一步) (3 .4 0)

You are to (a) write down the tight-binding Hamiltonian for this system and s01ve
it for the energy bands , (b) determine the va1ues (there are two) of k for which the
energy dispersion vanishes , (c) show that the dispersion can be written as
E(q) = 土 hVFlq l， (3 .4 1)

where VF = 3t /2 17，α~ 106 m/s and q is the deviation 仕om the va1ue of the momentum
at the zero crossing , and (d) show that the effective Hami1tonian near the zero crossings
can be written as
+o
。-

/it--\
oix oiy \ka--/ LA
“
﹒唔，

H --- LAU --
A

σ
們U

EI EI
1
OA
X
-4EA
OA
vd

3 TA 4gLLU
CIA vh OA mHmh
，1品

hM=-
戶LV

.們

+oOAY
、
O

/iz--\
\KEEF/
﹒唔，A

OA
x + oiy
恥九
一一

σ
、‘，/

/'‘\
們U

EA OA
U

., OA EI X x VJ
oix -A Y
一

for the K point.

臨睡f智體臨時$

[B01927] M. Born and J. R. Oppenheimer, Ann. d. Physik 8唔， 457 (1927).

[WI937] G. H. Wannier, Phys. Rev. 52 , 191 (1937).
 To solve many-particle problems , it is expedient to introduce the language of second
quantization. This approach does not add new conceptual baggage to quantum mechanics.
Rather, it provides a convenient book-keeping method for dealing with many且particle
states. As a result, we will be brief in the formal development of this technique. For a more
complete development, see , for example , Baym (B1969). We develop this approach for
fermions , that is , particles that are antisymmetric with respect to interchange of any two of
their coordinates , and for bosons , particles for which such interchanges do not incur any
sign change in the many-particle wavefunction.

4.1 8050臨$

Consider the many-body state of a boson system containing n identical particles:

伊o(r)=(叭， r2 尸. ., rnln). (4.1)

For states containing different numbers of particles , the orthogonality condition

(nlm) = ðnm (4.2)

holds. As in the case of the harmonic oscillator, we define the operator ao such that

aoln) = Jn ln - 1). (4.3)

That is , the operatorαo annihilates a single particle from the 泣的|的 and produces the
corresponding n 一 1 particle state. The a再joint operation in Eq. (4.2) suggests that

(nla6 = (n - 11 Jn. (4 .4)

As a consequence , the matrix element

(nla6ln - 1) = ~/ii(n 一 lln - 1) = Jn (4.5)

An equivalent statement ofthe result in Eq. (4.5) is that

α61 卜 1) = Jn ln) (4.6)

The ope肌r a6 acts 。叫le In - 1) 問ticle st瓶 produci叩

24
25 4.1 Bosons

The choice of the factor of .Jïi will now become 由ar when we consider the ope則or
丸 =αμ。丸 is aH位血

祠的=吻。 In) = ab ,J]ï ln - 1) = nln) , (4.7)

counts the number of particles in 泣的 In). Consequently, (nINoln) = n, and N o is ca11ed
the number operator. Consider now the creation of the state 1 的 from the vacuum state 10)
It fo11ows from Eqs. (4 .l) through (4.7) that

(a~)n
In) = 一社汁。) . (4.8)
、/n!

執f扭
e see then that the construction of an n-particle state of bosons 臼
is directly ana
站logous to
伽叫臼叫臼
for creatir叭a
缸n口
mo

the commutation relation

{旬 ， abJ = 1 (4.9)

Tos叫ow 帥 comesabouω的d叮伽 operation [α叫 -GM巾). FromEqs. (4 .1)一(4.7)

we have found that

[α叫 - abao]ln) = (n + 1 - n)ln) = In) , (4.10)

which implies th前 [ao ， ab] = 1

Let us now generalize to a boson state of the form 11旬 ， nl ， 的， . . .), where ni represents the
occupation number for the boson state i. Application ofthe boson rules outlined previously
yields

afl... ， 的，...) =布 1 .. .科 -1 ，...)

afajl. .. , nf , .. . , nj , ...) = -Jlië"，Jlïj I... ， 的一 1 ， .. . ，的一 1 ，.. .). (4.11)

Analogous relationships hold for the creation operators. The generalized commutation
relations are

[αj ， a~J = ðjf ,

[α川f] 口 [α) ，叫= 0 (4.12)

A final note on the application ofproducts ofboson operators to the state Ino , . . . , nj , . . .)
is“吋
s缸
innord位 B
i 加ωa
∞ 削吋
u
弘沾臼s鵬
u 閻
可切
吋b
e 加
枷Oωso
岫咖 肌
ns
1臼sc
∞ omml師1Í臼
e， 咖
op
阱
pera
叮昀
阿叫
ra
剖t世lOn
a 叫
1

the same 臼削叫， 控 1rre恥仙f吧eωoft昀 也

h旭盯
eωOI吋r oft血
h時盯
eωO叩pe叫or昀Sαdj 組dα4J ASWeWi山l1 s悅恤 i臼S
not the case with fermi昀ons
旭s.
 r、J
nc nunHEHunu- nunH

ι
4EL?L
nH

、d
pz

、d
Hu ..
26

、
4.2 Fer臨 i臨時$

Let us assume now that the particles occupying the state Ino) are fermions. We are restricted
by the Pau1i exclusion princip1e that no = 0 , 1. Ana1ogous1y,

aol1) = 10) ,
必 10) = 11) (4.13)

Note a1叫hatα010) =吋 1) = 0 , beca闊的 particle can be annihi1a削 from 伽 vaωm

and each state can have at most one particle. To illustrate the inherent antisymmetry of
fermions , consider the state Ino ， 叭， . . .). The sign convention for the application of creation
and annihi1ation operators to this reference state is

一 1 (一 1) 的 1... ，的一 1 ， 1 ，的+1 ，.. .), 昀

nj = 0,
吵Jhi
α4 而
，J
n叫j六件川一 (4.14)

一 i扒(←一 1)戶ηj 1 . . .仆， 的

nj一
-l ， O ， 叫
n j+
刊1 ， .…. .)，的= 1,
αjl... ， nj-1 ， n川仙) = lò , 的i =0 , (4.15)

where TJj is the number of occupied states to the 1eft of state j.

Let \Þ represent an occupation number wavefunction that contains the state j but not the
state l. Let's also assume j > l. Also let 可j and 可 f represent the number of occupied states
immediately to the left of j and l , respectively. By the ru1es in Eqs. (4.14) and (4 .1 5) , we
have that

α;Gj 中 = (-1) 可川 1 . . . , nf-1 , 1, . . . , n川 0 , . . .) (4.16)

的 α: 中= (一 1)η卅 j+11... ， nf 一 1 ， 1, . . . , n j -1 , 0 , . . .). (4 .1 7)

If we now add (4.16) and (4.17) , we obtain

[a1aj + αja1] 中=(一 1)η卅 e [1 一 1] 1. . . , nf-1 , 1, . . . , n川 0 ， .. .), (4.18)

。 r equivalently
{a1 ， aj}=a1的 + a ja1 = 0 for j 利. (4 .1 9)

It is straigh自m仙 o show 伽 if j = l in Eq. (4.16) ， ψj+ αja) = 1. The gen叫

anticommutation relation is

{a) , ad = ojf (4.20)

The remainder of the anticommutation relations , namely

{α川e} = {a) , a1} = 0, (4.21)

27 4.3 Fermion operators

can be derived in an analogous fashion. The final quantity we should define is the number
operator,
方= La}aj (4.22)
j

As in the boso叭楓 Gjdj CO削S 伽 nun蜘 of particles in 馴ej

4.3 Fer帥 io臨。瞬時臨時

We now tum to the task of writing one- and two-body fermion operators in their second
quantized form. A straightforward way of accomplishing this is to consider their action on
a fully antisymmetrized many-particle state. Consider 何10 single-particle states <!>I and <fJ2.
The normalized wavefunction that is antisymmetric with respect to particle interchange is

(r1 , r21 n 1 月)=主[</>叫做 (2) 一 </>1 (2)仇 (1)] . (4 幻)

、/之

This state can be constructed from the determinant of </>1 and 仇:

D7= 土 1</>1 (1)仇(1月 (4.24)

4 而 1</>1 (2) 仇 (2)1

Dirac (D1929; D1958) and Slater (S1929) showed that the general rule for constructing an
antisymmetric wavefunction out of n single-particle states is

Dn = (叭， r2 ， . ..1 丸，你. . .)口一再 11 </>1 . . . </>心， (4.25)

'、/n!
where 11 11 represents the determinant. Dn contains all antisymmetrized permut前ions of
the orbital set φ1 . . . </>n and hence may be written as

Dn戶=叫寸(仕
川
r
-v 叫
n !ι7

where P is the permutation operator. The inherent advantage in using second-quantized

notation is that a general many-particle fermionic state can be written compactly as

I n 1 ， 呃， • • .) = ai a1. . . 10) (4.27)

Complete antisymmetry under particle interchange is built into this many-body state as
a result of the anticommuting property of the fermion oper泣的﹒ Note there is no -JnT
normalization factor. In first quantization, however, an explicit -JnT factor appears , because
particles are placed in particular single-particle states and all possible permutations are
summed over. In second quantization, no labels are attached to the particles.
Let us now return to our initial goal. Consider the one-body operator H 1 . The matrix
element of a one-body operator between two many-particle states is non-zero only if the two
Exploring the Variety of Random
Documents with Different Content
 “What—has—prevented—him?” I roared.
“Look out—your van will be in the ditch.”
And turning quickly I was just in time to pull the tiresome brute of a
horse, who never could be left to himself an instant, straight again.
I walked on shrugging my shoulders. Menzies-Legh was without any
doubt as ill-conditioned a specimen of manhood as I have ever come across.
At the four crossroads beyond Brede, on the party’s pausing as usual to
argue over the signpost while Fate, with Frogs’ Hole Farm up her sleeve,
laughed in the background, I laid my hand on Jellaby’s arm—its thinness
quite made me jump—and said, “Where is Lord Sigismund?”
“Gone home, I believe, with his father.”
“Why is he not coming back?”
“He’s prevented.”
“But by what? Is he ill?”
“Oh, no. He’s just—just prevented, you know.”
And Jellaby slipped his arm out of my grasp and went to stare with the
others up at the signpost.
On the road we finally decided to take, while they were all clustering
round the labourer I have mentioned who directed us to the deserted farm, I
approached Frau von Eckthum who stood on the outer fringe of the cluster,
and said in the gentler voice I instinctively used when speaking to her, “I
hear Lord Sigismund is not coming back.”
Gently as my voice was, it yet made her start; she generally did start
when spoken to, being unusually (it adds to her attractiveness) highly strung.
(“She doesn’t when I speak to her,” said Edelgard, on my commenting to
her on this characteristic.
“My dear, you are merely another woman,” I replied—somewhat sharply,
for Edelgard is really often unendurably obtuse.)
“I hear Lord Sigismund is not coming back,” I said, then, very gently, to
the tender lady.
“Oh?” said she.
For the first time I could have wished a wider range of speech.
“He has been prevented, I hear.”
“Oh?”
 “Do you know what has prevented him?”
She looked at me and then at the others absorbed by the labourer with a
funny little look (altogether feminine) of helplessness, though it could not of
course have been that; then, adding another letter but not unfortunately
another word to her vocabulary, she said “No”—or rather “N-n-n-o,” for she
hesitated.
And up bustled Jellaby as I was about to press my inquiries, and taking
me by the elbow (the familiarity of this sort of person!) led me aside to
overwhelm me with voluble directions as to the turnings to Frogs’ Hole
Farm.
Well, it was undoubtedly a blow to find by far the most interesting and
amiable member of the party (with the exception of Frau von Eckthum)
gone, and gone without a word, without an explanation, a farewell, or a
regret. It was Lord Sigismund’s presence, the presence of one so
unquestionably of my own social standing, of one whose relations could all
bear any amount of scrutiny and were not like Edelgard’s Aunt Bockhügel
(of whom perhaps more presently) a dark and doubtful spot round which
conversation had to make careful détours—it was undoubtedly,
 Gentle as my voice was, it yet made her start

I say, Lord Sigismund who had given the expedition its decent air of
being just an aristocratic whim, stamped it, marked it, raised it altogether
above mere appearances. He was a Christian gentleman; more, he was the
only one of the party who could cook. Were we, then, to be thrown for future
sustenance entirely on Jellaby’s porridge?
That afternoon, dining in the mud of the deserted farmyard, we had
sausages; a dinner that had only been served once before, and which was a
sign in itself that the kitchen resources were strained. I have already
described how Jellaby cooked sausages, goading them round and round the
pan, prodding them, pursuing them, giving them no rest in which to turn
brown quietly—as foolish a way with a sausage as ever I have seen. For the
second time during the tour we ate them pink, filling up as best we might
with potatoes, a practice we had got quite used to, though to you, my
hearers, who only know potatoes as an adjunct, it will seem a pitiable state
of things. So it was; but when one is hungry to the point of starvation a hot
potato is an attractive object, and two hot potatoes are exactly doubly so.
Anyhow my respect for them has increased tenfold since my holiday, and I
insist now on their being eaten in much larger quantities than they used to be
in our kitchen, for do I not know how thoroughly they fill? And servants
quarrel if they have too much meat.
“That is poor food for a man like you, Baron,” said Menzies-Legh,
suddenly addressing me from the other end of the table.
He had been watching me industriously scraping—picture, my friends,
Baron von Ottringel thus reduced—scraping, I say, the last remnants of the
potatoes out of the saucepan after the ladies had gone, accompanied by
Jellaby, to begin washing up.
It was so long since he had spoken to me of his own accord that I paused
in my scraping to stare at him. Then, with my natural readiness at that sort of
thing, I drew his attention to his bad manners earlier in the afternoon by
baldly answering “Eh?”
“I wonder you stand it,” he said, taking no notice of the little lesson.
“Pray will you tell me how it is to be helped?” I inquired. “Roast goose
does not, I have observed, grow on the hedges in your country.” (This, I felt,
was an excellent retort.)
“But it flourishes in London and other big towns,” said he—a foolish
thing to say to a man sitting in the back yard of Frogs’ Hole Farm. “Have a
cigarette,” he added; and he pushed his case toward me.
I lit one, slightly surprised at the change for the better in his behaviour,
and he got up and came and sat on the vacant camp-stool beside me.
“Hunger,” said I, continuing the conversation, “is the best sauce, and as I
am constantly hungry it follows that I cannot complain of not having enough
sauce. In fact, I am beginning to feel that gipsying is a very health-giving
pursuit.”
 “Damp—damp,” said Menzies-Legh, shaking his head and screwing up
his mouth in a disapproval that astonished me.
“What?” I said. “It may be a little damp if the weather is damp, but one
must get used to hardships.”
“Only to find,” said he, “that one’s constitution has been undermined.”
“What?” said I, unable to understand this change of attitude.
“Undermined for life,” said he, impressively.
“My dear sir, I have heard you myself, under the most adverse
circumstances, repeatedly remark that it was healthy and jolly.”
“My dear Baron,” said he, “I am not like you. Neither Jellaby, nor I, nor
Browne either, for that matter, has your physique. We are physically,
compared to you—to be quite frank—mere weeds.”
“Oh, come now, my dear sir, I cannot permit you—you undervalue—of
slighter build, perhaps, but hardly——”
“It is true. Weeds. Mere weeds. And my point is that we, accordingly, are
not nearly so likely as you are to suffer in the long run from the privations
and exposure of a bad-weather holiday like this.”
“Well now, you must pardon me if I entirely fail to see——”
“Why, my dear Baron, it’s as plain as daylight. Our constitutions will not
be undermined for the shatteringly good reason that we have none to
undermine.”
My hearers will agree that, logically, the position was incontrovertible,
and yet I doubted.
Observing my silence, and probably guessing its cause, he took up an
empty glass and poured some tea into it from the teapot at which Frau von
Eckthum had been slaking her thirst in spite of my warnings (I had, alas, no
right to forbid) that so much tea drinking would make her still more liable to
start when suddenly addressed.
“Look here,” said he.
I looked.
“You can see this tea.”
“Certainly.”
“Clear, isn’t it? A beautiful clear brown. A tribute to the spring water
here. You can see the house and all its windows through it, it is so perfectly
transparent.”
 And he held it up, and shutting one eye stared through it with the other.
“Well?” I inquired.
“Well, now look at this.”
And he took another glass and set it beside the first one, and poured both
tea and milk into it.
“Look there,” he said.
I looked.
“Jellaby,” said he.
I stared.
Then he took another glass, and poured both tea and milk into it, setting it
in a line with the first two.
“Browne,” said he.
I stared.
Then he took a fourth glass, and filled it in the same manner as the
second and third and placed it at the end of the line.
“Myself,” said he.
I stared.
“Can you see through either of those three?” he asked, tapping them one
after the other.
“No,” said I.
“Now if I put a little more milk into them”—he did—“it makes no
difference. They were muddy and thick before, and they remain muddy and
thick. But”—and he held the milk jug impressively over the first glass—“if I
put the least drop into this one”—he did—“see how visible it is. The
admirable clearness is instantaneously dimmed. The pollution spreads at
once. The entire glass, owing to that single drop, is altered, muddied,
ruined.”
“Well?” I inquired, as he paused and stared hard at me.
“Well?” said he. “Do you not see?”
“See what?” said I.
“My point. It’s as clear as the first glass was before I put milk into it. The
first glass, my dear Baron, is you, with your sound and perfect constitution.”
I bowed.
“Your splendid health.”
 I bowed.
“Your magnificent physique.”
I bowed.
“The other three are myself, and Jellaby, and Browne.”
He paused.
“And the drop of milk,” he said slowly, “is the caravan tour.”
I was confounded; and you, my hearers, will admit that I had every
reason to be. Here was an example of what is rightly called irresistible logic,
and a reasonable man dare not refuse, once he recognizes it, to bow in
silence. Yet I felt very well. I said I did, after a pause during which I was
realizing how unassailable Menzies-Legh’s position was, and endeavouring
to reconcile its unassailableness with my own healthful sensations.
“You can’t get away from facts,” he answered. “There they are.”
And he indicated with his cigarette the four glasses and the milk jug.
“But,” I repeated, “except for a natural foot-soreness I undoubtedly do
feel very well.”
“My dear Baron, it is obvious beyond all argument that the more
absolutely well a person is the more easily he must be affected by the
smallest upset, by the smallest variation in the environment to which he has
got accustomed. Paradox, which plays so large a part in all truths, is rampant
here. Those in perfect health are nearer than anybody else to being seriously
ill. To keep well you must never be quite so.”
He paused.
“When,” he continued, seeing that I said nothing, “we began caravaning
we could not know how persistently cold and wet it was going to be, but
now that we do I must say I feel the responsibility of having persuaded you
—or of my sister-in-law’s having persuaded you—to join us.”
“But I feel very well,” I repeated.
“And so you will, up to the moment when you do not.”
Of course that was true.
“Rheumatism, now,” he said, shaking his head; “I greatly fear
rheumatism for you in the coming winter. And rheumatism once it gets hold
of a man doesn’t leave him till it has ravaged each separate organ, including,
as everybody knows, that principal organ of all, the heart.”
 This was gloomy talk, and yet the man was right. The idea that a holiday,
a thing planned and looked forward to with so much pleasure, was to end by
ravaging my organs did not lighten the leaden atmosphere that surrounded
and weighed upon Frogs’ Hole Farm.
“I cannot alter the weather,” I said at last—irritably, for I felt ruffled.
“No. But I wouldn’t risk it for too long if I were you,” said he.
“Why, I have paid for a month,” I exclaimed, surprised that he should
overlook this clinching fact.
“That, set against an impaired constitution, is a very inconsiderable
trifle,” said he.
“Not inconsiderable at all,” said I sharply.
“Money is money, and I am not one to throw it away. And what about the
van? You cannot abandon an entire van at a great distance from the place it
belongs to.”
“Oh,” said he quickly, “we would see to that.”
I got up, for the sight of the glasses full of what I was forced to
acknowledge was symbolic truth irritated me. The one representing myself,
into which he had put but one drop of milk, was miserably discoloured. I did
not like to think of such discolouration being my probable portion, and yet
having paid for a month’s caravaning what could I do?
The afternoon was chilly and very damp, and I buttoned my wraps
carefully about my throat. Menzies-Legh watched me.
“Well,” said he, getting up and looking first at me and then at the glasses
and then at me again, “what do you think of doing, Baron?”
“Going for a little stroll,” I said.
And I went.
 CHAPTER XVII

T HIS was a singular conversation.

I passed round the back of the house and along a footpath I found
there, turning it over in my mind. Less than ever did I like Menzies-
Legh. In spite of the compliments about my physique I liked him less than
ever. And how very annoying it is when a person you do not like is right; bad
enough if you do like him, but intolerable if you do not. As I proceeded
along the footpath with my eyes on the ground I saw at every step those four
glasses of tea, particularly my one, the one that sparkled so brilliantly at first
and was afterward so easily ruined. Absorbed in this contemplation I did not
notice whither my steps were tending till I was pulled up suddenly by a
church door. The path had led me to that, and then, as I saw, skirted along a
fringe of tombstones to a gate in a wall beyond which appeared the
chimneys of what was no doubt the parsonage.
The church door was open, and I went in—for I was tired, and here were
pews; ruffled, and here was peace. The droning of a voice led me to
conclude (rightly) that a service was in progress, for I had learned by this
time that in England the churches constantly burst out into services,
regardless of the sort of day it is—whether, I mean, it is a Sunday or not. I
entered, and selecting a pew with a red cushion along its seat and a
comfortable footstool sat down.
The pastor was reading the Scriptures out of a Bible supported, according
to the unaccountable British custom, on the back of a Prussian eagle. This
prophetic bird—the first swallow, as it were, of that summer which I trust
will not long be delayed, when Luther’s translation will rest on its back and
be read aloud by a German pastor to a congregation forced to understand by
the simple methods we bring to bear on our Polish (also acquired) subjects—
eyed me with a human intelligence. We eyed each other, in fact, as old
friends might who meet after troublous experiences in an alien land.
Except for this bird, who seemed to me quite human in his expression of
alert sympathy, the pastor and I were alone in the building; and I sat there
marvelling at the wasteful folly that pays a man to read and pray daily to a
set of empty pews. Ought he not rather to stay at home and keep an eye on
his wife? To do, indeed, anything sooner than conduct a service which
nobody evidently wants? I call it heathenism; I call it idolatry; and so would
any other plain man who heard and saw empty pews, things of wood and
cushions, being addressed as brethren, and dearly beloved ones into the
bargain.
When he had done at the eagle he crossed over to another place and
began reciting something else; but very soon, after only a few words, he
stopped dead and looked at me.
I wondered why, for I had not done anything. Even, however, with that
innocence of conscience in the background, it does make a man
uncomfortable when a pastor will not go on but fixes his eyes on you sitting
harmless in your pew, and I found myself unable to return his gaze. The
eagle was staring at me with a startling expression of comprehension, almost
as if he too were thinking that a pastor officiating has such an undoubted
advantage over the persons in the pews that it is cowardice to use it. My
discomfort increased considerably when I saw the pastor descend from his
place and bear down on me, his eyes still fixing me, his white clothing
fluttering out behind him. What, I asked myself greatly perturbed, could the
creature possibly want? I soon found out, for thrusting an open Prayer-book
toward me he pointed to a verse of what appeared to be a poem, and
whispered:
“Will you kindly stand up and take your part in the service?”
Even had I known how, surely I had no part nor lot in such a form of
worship.
“Sir,” I said, not heeding the outstretched book, but feeling about in my
breast-pocket, “permit me to present you with my card. You will then see
——”
He, however, in his turn refused to heed the outstretched card. He did not
so much as look at it.
“I cannot oblige you to,” he whispered, as though our conversation were
unfit for the eagle’s ears; and leaving the open book on the little shelf in the
front of the pew he strode back again to his place and resumed his reading,
doing what he called my part as well as his own with a severity of voice and
manner ill-suited to one presumably addressing the liebe Gott.
Well, being there and very comfortable I did not see why I should go. I
was behaving quite inoffensively, sitting still and holding my tongue, and the
comfort of being in a building with no fresh air in it was greater than you,
my friends, who only know fresh air at intervals and in properly limited
quantities, will be able to understand. So I stayed till the end, till he, after a
profusion of prayers, got up from his knees and walked away into some
obscure portion of the church where I could no longer observe his
movements, and then, not desiring to meet him, I sought the path that had
led me thither and hurriedly descended the hill to our melancholy camp.
Once I thought I heard footsteps behind me and I hastened mine, getting as
quickly round a bend that would conceal me from any one following me as a
tired man could manage, and it was not till I had reached and climbed into
the Elsa that I felt really safe.
The three caravans were as usual drawn up in a parallel line with mine in
the middle, and their door ends facing the farm. To be in the middle is a most
awkward situation, for you cannot speak the least word of caution (or
forgiveness, as the case may be) to your wife without running grave risk of
being overheard. Often I used carefully to shut all the windows and draw the
door curtain, hoping thus to obtain a greater freedom of speech, though this
was of little use with the Ilsa and the Ailsa on either side, their windows
open, and perhaps a group of caravaners sitting on the ground immediately
beneath.
My wife was mending, and did not look up when I came in. How
differently she behaved at home. She not only used to look up when I came
in, she got up, and got up quickly too, hastening at the first sound of my
return to meet me in the passage, and greeting me with the smiles of a dutiful
and accordingly contented wife.
Shutting the Elsa’s windows I drew her attention to this.
“But there isn’t a passage,” said she, still with her head bent over a sock.
Really Edelgard should take care to be specially feminine, for she
certainly will never shine on the strength of her brains.
“Dear wife,” I began—and then the complete futility of trying to thresh
any single subject out in that airy, sound-carrying dwelling stopped me. I sat
down on the yellow box instead, and remarked that I was extremely fatigued.
“So am I,” said she.
“My feet ache so,” I said, “that I fear there may be something serious the
matter with them.”
“So do mine,” said she.
 This, I may observe, was a new and irritating habit she had got into:
whatever I complained of in the way of unaccountable symptoms in divers
portions of my frame, instead of sympathizing and suggesting remedies she
said hers (whatever it was) did it too.
“Your feet cannot possibly,” said I, “be in the terrible condition mine are
in. In the first place mine are bigger, and accordingly afford more scope for
disorders. I have shooting pains in them resembling neuralgia, and no doubt
traceable to some nervous source.”
“So have I,” said she.
“I think bathing might do them good,” I said, determined not to become
angry. “Will you get me some hot water, please?”
“Why?” said she.
She had never said such a thing to me before. I could only gaze at her in a
profound surprise.
“Why?” I repeated at length, keeping studiously calm. “What an
extraordinary question. I could give you a thousand reasons if I chose, such
as that I desire to bathe them; that hot water—rather luckily for itself—has
no feet, and therefore has to be fetched; and that a wife has to do as she is
told. But I will, my dear Edelgard, confine myself to the counter inquiry, and
ask why not?”
“I, too, my dear Otto,” said she—and she spoke with great composure,
her head bent over her mending, “could give you a thousand answers to that
if I chose, such as that I desire to get this sock finished—yours, by the way;
that I have walked exactly as far as you have; that I see no reason why you
should not, as there are no servants here, fetch your own hot water; and that
your wishing or not wishing to bathe your feet has really, if you come to
think of it, nothing to do with me. But I will confine myself just to saying
that I prefer not to go.”
It can be imagined with what feelings—not mixed but unmitigated—I
listened to this. And after five years! Five years of patience and guidance.
“Is this my Edelgard?” I managed to say, recovering speech enough for
those four words but otherwise struck dumb.
“Your Edelgard?” she repeated musingly as she continued to mend, and
not even looking at me. “Your boots, your handkerchief, your gloves, your
socks—yes——”
I confess I could not follow, and could only listen amazed.
 “But not your Edelgard. At least, not more than you are my Otto.”
“But—my boots?” I repeated, really dazed.
“Yes,” she said, folding up the finished sock, “they really are yours. Your
property. But you should not suppose that I am a kind of living boot, made to
be trodden on. I, my dear Otto, am a human being, and no human being is
another human being’s property.”
A flash of light illuminated my brain. “Jellaby!” I cried.
“Hullo?” was the immediate answer from outside. “Want me, Baron?”
“No, no! No, no! No, NO!” I cried leaping up and dragging the door
curtain to, as though that could possibly deaden our conversation. “He has
been infecting you,” I continued, in a whisper so much charged with
indignation that it hissed, “with his poisonous——”
Then I recollected that he could probably hear every word, and muttering
an imprecation on caravans I relapsed on to the yellow box and said with
forced calm as I scrutinized her face:
“Dear wife, you have no idea how exactly you resemble your Aunt
Bockhügel when you put on that expression.”
For the first time this failed to have an effect. Up to then to be told she
looked like her Aunt Bockhügel had always brought her back with a jerk to
smiles; even if she had to wrench a smile into position she did so, for the
Aunt Bockhügel is the sore point in Edelgard’s family, the spot, the smudge
across its brightness, the excrescence on its tree, the canker in its bud, the
worm destroying its fruit, the night frost paralyzing its blossoms. She cannot
be suppressed. She cannot be explained. Everybody knows she is there. She
was one of the reasons that made me walk about my room the whole of the
night before I proposed marriage to Edelgard, a prey to doubts as to how far
a man may go in recklessness in the matter of the aunts he fastens upon his
possible children. The Ottringels can show no such relatives; at least there is
one, but she looms almost equal to the rest owing to the mirage created by
fogs of antiquity and distance. But Edelgard’s aunt is contemporary and
conspicuous. Of a vulgar soul at her very birth, as soon as she came of age
she deliberately left the ranks of the nobility and united herself to a dentist.
We go there to be treated for toothache, because they take us (owing to the
relationship) on unusually favourable terms; otherwise we do not know
them. There is however an undoubted resemblance to Edelgard in her less
pleasant moods, a thickened, heavier, and older Edelgard, and my wife, well
aware of it (for I help her to check it as much as possible by pointing it out
whenever it occurs) has been on each occasion eager to readjust her features
without loss of time. On this one she was not. Nay, she relaxed still more,
and into a profounder likeness.
“It’s true,” she said, not even looking at me but staring out of the
window; “it’s true about the boots.”
“Aunt Bockhügel! Aunt Bockhügel!” I cried softly, clapping my hands.
She actually took no notice, but continued to stare abstractedly out of the
window; and feeling how impossible it was to talk really naturally to her
with Jellaby just outside, I chose the better part and with a movement I could
not wholly suppress of impatience got up and left her.
Jellaby, as I suspected, was sitting on the ground leaning against one of
our wheels as though it were a wheel belonging to his precious community
and not ours, hired and paid for. Was it possible that he selected this wheel
out of the twelve he could have chosen from because it was my wife’s
wheel?
“Do you want anything?” he asked, looking up and taking his pipe out of
his mouth; and I just had enough self-control to shake my head and hurry on,
for I felt if I had stopped I would have fallen upon him and rattled him about
as a terrier rattles a rat.
But what terrible things caravans are when you have to share one with a
person with whom you have reason to be angry! Of all their sides this is
beyond doubt the worst; worse than when the rain comes in on to your bed,
worse than when the wind threatens to blow them over during the night, or
half of them sinks into the mud and has to be dug out laboriously in the
morning. It may be imagined with what feelings I wandered forth into the
chill evening, homeless, bearing as I felt a strong resemblance to that
Biblical dove which was driven forth from the shelter of the ark and had no
idea what to do next. Of course I was not going to fetch the hot water and
return with it, as it were (to pursue my simile), in my beak. Every husband
throughout Germany will understand the impossibility of doing that—picture
Edelgard’s triumph if I had! Yet I could not at the end of a laborious day
wander indefinitely out-of-doors; besides, I might meet the pastor.
The rest of the party were apparently in their caravans, judging from the
streams of conversation issuing forth, and there was no one but old James
reclining on a sack in the corner of a distant shed to offer me the solace of
companionship. With a sudden mounting to my head of a mighty wave of
indignation and determination not to be shut out of my own caravan, I turned
and quickly retraced my steps.
“Hullo, Baron,” said Jellaby, still propped against my wheel. “Had
enough of it already?”
“More than enough of some things,” I said, eyeing him meaningly as I
made my way, much impeded by my mackintosh, up the ladder at an oblique
angle (it never could or would stand straight) against our door.
“For instance?” he inquired.
“I am unwell,” I answered shortly, evading a quarrel—for why should I
allow myself to be angered by a wisp like that?—and entering the Elsa drew
the curtain sharply to on his expressions of conventional regret.
Edelgard had not changed her position. She did not look up.
I pulled off my outer garments and flung them on the floor, and sitting
down with emphasis on the yellow box unlaced and kicked off my boots and
pulled off my stockings.
Edelgard raised her head and fixed her eyes on me with a careful
imitation of surprise.
“What is it, Otto?” she said. “Have you been invited out to dine?”
I suppose she considered this amusing, but of course it was not, and I
jerked myself free of my braces without answering.
“Won’t you tell me what it is?” she asked again.
For all answer I crawled into my berth and pulled the coverings up to my
ears and turned my face to the wall; for indeed I was at the end both of my
patience and my strength. I had had two days’ running full of disagreeable
incidents, and Menzies-Legh’s fatal drop of milk seemed at last to have
fallen into the brightness of my original strong tea. I ached enough to make
his prophesied rheumatism a very near peril, and was not at all sure as I lay
there that it had not already begun its work upon me, beginning it with an
alarming promise of system and thoroughness at the very beginning, i. e., my
feet.
“Poor Otto,” said Edelgard, getting up and laying her hand on my
forehead; adding, after a moment, “It is nice and cool.”
“Cool? I should think so,” said I shivering. “I am frozen.”
 She got a rug out of the yellow box and laid it over me, tucking in the
side.
“So tired?” she said presently, as she tidied up my clothes.
“Ill,” I murmured.
“What is it?”
“Oh, leave me, leave me. You do not really care. Leave me.”
At this she paused in her occupation to gaze, I fancy, at my back as I lay
resolutely turned away.
“It is very early to go to bed,” she said after a while.
“Not when a man is ill.”
“It isn’t seven yet.”
“Oh, do not, I beg you, argue with me. If you cannot have sympathy you
can at least leave me. It is all I ask.”
This silenced her, and she moved about the van more careful not to sway
it, so that presently I was able to fall into an exhausted sleep.
How long this lasted I could not on suddenly waking tell, but everything
had grown dark and Edelgard, as I could hear, was asleep above me.
Something had wrenched me out of the depths of slumber in which I was
sunk and had brought me up again with a jerk to that surface known to us as
sentient life. You are aware, my friends, being also living beings with all the
experiences connected with such a condition behind you, you are aware what
such a jerking is. It seems to be a series of flashes. The first flash reminds
you (with an immense shock) that you are not as you for one comfortable
instant supposed in your own safe familiar bed at home; the second brings
back the impression of the loneliness and weirdness of Frogs’ Hole Farm (or
its, in your case, local equivalent) that you received while yet it was day; the
third makes you realize with a clutching at your heart that something
happened before you woke up, and that something is presently going to
happen again. You lie awake waiting for it, and the entire surface of your
body becomes as you wait uniformly damp. The sound of a person breathing
regularly in the apartment does but emphasize your loneliness. I confess I
was unable to reach out for matches and strike a light, unable to do anything
under that strong impression that something had happened except remain
motionless beneath the bed-coverings. This was no shame to me, my friends.
Face me with cannon, and I have the courage of any man living, but place
me on the edge of the supernatural and I can only stay beneath the
bedclothes and grow most lamentably damp. Such a thin skin of wood
divided me from the night outside. Any one could push back the window
standing out there; any one ordinarily tall would then have his head and
shoulders practically inside the caravan. And there was no dog to warn us or
to frighten such a wretch away. And all my money was beneath my mattress,
the worst place possible to put it in if what you want is not to be personally
disturbed. What was it I had heard? What was it that called me up from the
depths of unconsciousness? As the moments passed—and except for
Edelgard’s regular breathing there was only an awful emptiness and absence
of sound—I tried to persuade myself it was just the sausages having been so
pink at dinner; and the tenseness of my terror had begun slowly to relax
when I was smitten stark again—and by what, my friends? By the tuning of a
violin.
Now consider, you who frequent concerts and see nothing disturbing in
this sound, consider our situation. Consider the remoteness from the
highway of Frogs’ Hole Farm; how you had, in order to reach it, to follow
the prolonged convolutions of a lane; how you must then come by a cart
track along the edge of a hop-field; how the house lay alone and empty in a
hollow, deserted, forlorn, untidy, out of repair. Consider further that none of
our party had brought a violin and none, to judge from the absence in their
conversation of any allusions to such an instrument, played on it. No one
knows who has not heard one tuned under the above conditions the
blankness of the horror it can strike into one’s heart. I listened, stiff with
fear. It was tuned with a care and at a length that convinced me that the spirit
turning its knobs must be of a quite unusual musical talent, possessed of an
acutely sensitive ear. How came it that no one else heard it? Was it possible
—I curdled at the thought—that only myself of the party had been chosen by
the powers at work for this ghastly privilege? When the thing broke into a
wild dance, and a great and rhythmical stamping of feet began apparently
quite near and yet equally apparently on boards, I was seized with a panic
that relaxed my stiffness into action and enabled me to thump the underneath
of Edelgard’s mattress with both my fists, and thump and thump with a
desperate vigour that did at last rouse her.
Being half asleep she was more true to my careful training than when
perfectly awake, and on hearing my shouts she unhesitatingly tumbled out of
her berth and leaning into mine asked me with some anxiety what the matter
was.
 “The matter? Do you not hear?” I said, clutching her arm with one hand
and holding up the other to enjoin silence.
She woke up entirely.
“Why, what in the world——” she said. Then pulling a window curtain
aside she peeped out. “There’s only the Ailsa there,” she said, “dark and
quiet. And only the Ilsa here,” she added, peeping through the opposite
curtain, “dark and quiet.”
I looked at her, marvelling at the want of imagination in women that
renders it possible for them to go on being stolid in the presence of what
seemed undoubtedly the supernatural. Unconsciously this stolidity, however,
made me feel more like myself; but when on her going to the door and
unbolting it and looking out she made an exclamation and hastily shut it
again, I sank back on my pillow once more hors de combat, so great was the
shock. Face me, I say, with cannon, and I can do anything but expect nothing
of me if it is ghosts.
“Otto,” she whispered, holding the door, “come and look.”
I could not speak.
“Get up and come and look,” she whispered again.
Well my friends I had to, or lose forever my moral hold of and headship
over her. Besides, I was drawn somehow to the fatal door. How I got out of
my berth and along the cold floor of the caravan to the end I cannot
conceive. I was obliged to help myself along, I remember, by sliding my
hand over the surface of the yellow box. I muttered, I remember, “I am ill—I
am ill,” and truly never did a man feel more so. And when I got to the door
and looked through the crack she opened, what did I see?
I saw the whole of the lower windows of the farmhouse ablaze with
candles.
 CHAPTER XVIII

M Y hearers will I hope appreciate the frankness with which I show them
all my sides, good and bad. I do so with my eyes open, aware that some
of you may very possibly think less well of me for having been, for
instance, such a prey to supernatural dread. Allow me, however, to point out
that if you do you are wrong. You suffer from a confusion of thought. And I
will show you why. My wife, you will have noticed, had on the occasion
described few or no fears. Did this prove courage? Certainly not. It merely
proved the thicker spiritual skin of woman. Quite without that finer
sensibility that has made men able to produce works of genius while women
have been able only to produce (a merely mechanical process) young, she
felt nothing apparently but a bovine surprise. Clearly, if you have no
imagination neither can you have any fears. A dead man is not frightened.
An almost dead man does not care much either. The less dead a man is the
more do possible combinations suggest themselves to him. It is imagination
and sensibility, or the want of them, that removes you further or brings you
nearer to the animals. Consequently (I trust I am being followed?) when
imagination and sensibility are busiest, as they were during those moments I
lay waiting and listening in my berth, you reach the highest point of
aloofness from the superiority to the brute creation; your vitality is at its
greatest; you are, in a word, if I may be permitted to coin an epigram, least
dead. Therefore, my friends, it is plain that at that very moment when you
(possibly) may have thought I was showing my weakest side I was doing the
exact opposite, and you will not, having intelligently followed the argument,
say at the end of it as my poor little wife did, “But how?”
I do not wish, however, to leave you longer under the impression that the
deserted farmhouse was haunted. It may have been of course, but it was not
on that night of last August. What was happening was that a party from the
parsonage—a holiday party of young and rather inclined to be noisy people,
which had overflowed the bounds of the accommodation there—was
utilizing the long, empty front room as an impromptu (I believe that is the
expression) ball-room. The farm belonged to the pastor—observe the fatness
of these British ecclesiastics—and it was the practice of his family during
the holidays to come down sometimes in the evening and dance in it. All this
I found out after Edelgard had dressed and gone across to see for herself
what the lights and stamping meant. She insisted on doing so in spite of my
warnings, and came back after a long interval to tell me the above, her face
flushed and her eyes bright, for she had seized the opportunity, regardless of
what I might be feeling waiting alone, to dance too.
“You danced too?” I exclaimed.
“Do come, Otto. It is such fun,” said she.
“With whom did you dance, may I inquire?” I asked, for the thought of
the Baroness von Ottringel dancing with the first comer in a foreign farm
was of course most disagreeable to me.
“Mr. Jellaby,” said she. “Do come.”
“Jellaby? What is he doing there?”
“Dancing. And so is everybody. They are all there. That’s why their
caravans were so quiet. Do come.”
And she ran out again, a childishly eager expression on her face, into the
night.
“Edelgard!” I called.
But though she must have heard me she did not come back.
Relieved, puzzled, vexed, and curious together, I did get up and dress,
and on lighting a candle and looking at my watch I was astonished to find
that it was only a quarter to ten. For a moment I could not credit my eyes,
and I shook the watch and held it to my ears, but it was going, as steadily as
usual, and all I could do was to reflect as I dressed on what may happen to
you if you go to bed and to sleep at seven o’clock.
And how soundly I must have done it. But of course I was unusually
weary, and not feeling at all well. Two hours’ excellent sleep, however, had
done wonders for me so great are my recuperative powers, and I must say I
could not help smiling as I crossed the yard and went up to the house at the
remembrance of Menzies-Legh’s glass of tea. He would not see much milk
about me now, thought I, as I strode, giving my moustache ends a final
upward push and guided by the music, into the room in which they were
dancing.
The dance came to an end as I entered, and a sudden hush seemed to fall
upon the company. It was composed of boys and young girls attired in
evening garments next to which the clothes of the caravaners, weather-
beaten children of the road, looked odd and grimy indeed. The tender lady, it
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookfinal.com