Download the full version of the ebook at ebookfinal.

com

Functional Methods in Quantum Field Theory and

Statistical Physics First Edition Vasiliev

https://ebookfinal.com/download/functional-methods-in-
quantum-field-theory-and-statistical-physics-first-edition-
vasiliev/

OR CLICK BUTTON

DOWNLOAD EBOOK

Download more ebook instantly today at https://ebookfinal.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Quantum field theory in condensed matter physics 2nd

Edition Alexei M. Tsvelik

https://ebookfinal.com/download/quantum-field-theory-in-condensed-
matter-physics-2nd-edition-alexei-m-tsvelik/

ebookfinal.com

Mathematical methods of many body quantum field theory 1st

Edition Detlef Lehmann

https://ebookfinal.com/download/mathematical-methods-of-many-body-
quantum-field-theory-1st-edition-detlef-lehmann/

ebookfinal.com

Instantons and Large N An Introduction to Non Perturbative

Methods in Quantum Field Theory Marcos Marino

https://ebookfinal.com/download/instantons-and-large-n-an-
introduction-to-non-perturbative-methods-in-quantum-field-theory-
marcos-marino/
ebookfinal.com

Quantum Field Theory From Operators to Path Integrals

Physics Textbook 1st Edition Kerson Huang

https://ebookfinal.com/download/quantum-field-theory-from-operators-
to-path-integrals-physics-textbook-1st-edition-kerson-huang/

ebookfinal.com
Conformal Field Theory New Non perturbative Methods In
String And Field Theory Yavuz Nutku

https://ebookfinal.com/download/conformal-field-theory-new-non-
perturbative-methods-in-string-and-field-theory-yavuz-nutku/

ebookfinal.com

Six Quantum Pieces A First Course in Quantum Physics

Valerio Scarani

https://ebookfinal.com/download/six-quantum-pieces-a-first-course-in-
quantum-physics-valerio-scarani/

ebookfinal.com

Quantum Field Theory in a Nutshell Second Edition A. Zee

https://ebookfinal.com/download/quantum-field-theory-in-a-nutshell-
second-edition-a-zee/

ebookfinal.com

Quantum Field Theory 1st Edition Michael V. Sadovskii

https://ebookfinal.com/download/quantum-field-theory-1st-edition-
michael-v-sadovskii/

ebookfinal.com

Introduction to Quantum Field Theory Second Edition

Roberto Casalbuoni

https://ebookfinal.com/download/introduction-to-quantum-field-theory-
second-edition-roberto-casalbuoni/

ebookfinal.com
Functional Methods in Quantum Field
Theory and Statistical Physics
C RC Press
Taylor & Fra ncis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 1998 by Taylor & Fra ncis Group, LLC

C RC Press is an imprint of Taylor & Francis Group, an In forma business

No claim to original U.S. Government wo rks

This book contains information obtained from authentic and hi ghly regarded sources. Reaso nable efforts have bee n made to
publish reliable data and information, but th e author and publisher ca nnot assume res ponsibility fo r the va lidity of all materials
or the co nsequ ences of th eir use. The authors an d publishers have attempted to trace the copyr ight holders of all material repro·
duced in this publication and apologize to copyright holders if permission to publish in this form has not been obta ined. If any
copyright material has not bee n ac knowledged please write and let us kn ow so we may rec tify in any future reprint.

Except as permi tted under U.S. Copyri ght Law, no part of t his book may be reprinted, reproduced, transmi tted, or utilized in any
form by any electronic, mechanical, or other means, now known or hereafter inve nted, including photocopying, microfilmin g,
and recording, or in any informat io n storage or ret rieval system, wit hout written permission from the publishers.

For permission to photocopy or use material elec tronic ally from this wo rk, please access www.copy ri ght. co m (http://www.co py·
right.com /) or co ntac t the Co pyright Clea rance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400.
CCC is a not· for· profit organi zation th at provides licenses and registration for a va riet y of use rs. For organizations that have been
granted a photocopy license by the CCC, a separate sys tem of pay ment has been arranged.

Trademark Notice: Product or corporate names may be trademark s or registered trademarks, and are used only for identifica·
tion and explanation wit hout intent to infringe.
Visit the Taylor & Francis We b site at
http: //www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com
 CONTENTS
Foreword xi

Preface xiii

Chapter 1 THE BASIC FORMALISM OF FIELD THEORY

1. 1 Fields and Products
1.1.1 Canonical quantization 1
1.1.2 The classical free theory 3
1.1.3 Anticommuting fields 4
1.1.4 The normal-ordered product of free-field operators 7
1.2 Functional Formulations of Wick's Theorem 9
1.2.1 Wick's theorem for a simple product 9
1.2.2 The Sym-product and the T-product 12
1.2.3 Wick's theorem for symmetric products 15
1.2.4 Reduction formulas for operator functionals 18
1.2.5 The Wick and Dyson T-products 21
1.3 The S·Matrix and Green Functions 23
1.3.1 Definitions 23
1.3.2 Transformation to the interaction picture
in the evolution operator 25
1.3.3 Transformation to the interaction picture for
the Green functions 27
1.3.4 Interactions containing time derivatives of the field 31
1.3.5 Generating functionals for the S-matrix and Green
functions 34
1.4 Graphs 37
1.4.1 Pertuibation theory 37
1.4.2 Some concepts from graph theory 39
1.4.3 Symmetry coefficients 39
1.4.4 Recursion relation for the symmetry coefficients 41
1.4.5 Transformation to Mayer graphs for the exponential
interaction 42
1.4.6 Graphs for the Yukawa-type interaction 43
1.4.7 Graphs for a pair interaction 48
1.4.8 Connectedness of the logarithm of R(<p) 49
1.4.9 Graphs for the Green functions 50
1.5 Unitarity of the S Matrix 53
1.5.1 The conjugation operation 53
1.5.2 Formal unitarity of the off-shell S matrix 55
v
vi CONTENTS

1.6 Functional Integrals 58

1.6.1 Gaussian integrals 58
1.6.2 Integrals on a Grassmann algebra 61
Gaussian integrals on a Grassmann algebra
~ .6.3 63
1.6.4 Gaussian integrals in field theory 65
1.6.5 Representations of generating functionals
for the S-matrix and Green functions by
functional integrals . 68
1.6.6 The stationary-phase method 70
1.6.7 The Dominicis-Englert theorem 73
1.7 Equations in Variational Derivatives 75
1.7.1 The Schwinger equations 75
1.7.2 Linear equations for connected Green functions 77
1.7.3 General method of deriving the equations 78
1.7.4 Iteration solution of the equations 82
1.8 One-Irreducible Green Functions 84
1.8.1 Definitions 84
1.8.2 The equations of motion fQr r 87
1.8.3 Iteration solution of the equations and proof
of !-irreducibility 88
1.9 Renormalization Transformations 91
1.10 Anomalous Green Functions and Spontaneous
Symmetry Breaking 94

Chapter 2 SPECIFIC SYSTEMS 97

2.1 Quantum Mechanics 97
2.1.1 The oscillator 97
2.1.2 The free particle 99
2.2 Nonrelativistic Field Theory 104
2.2.1 The quantum Bose and Fermi gases 108
2.2.2 The atom 109
2.2.3 Electrons in solids and phonons 110
2.3 Relativistic Field Theory 111
2.4 Integral Representations of the Transition Amplitude 114
2.5 The Space E(A) for Various Systems 120
2.6 Functional Integrals Over Phase Space ·122

Chapter 3 THE MASSLESS YANG-MILLS FIELD 127

3.1 Quantization of the Yang-Mills Field 127
3.1.1 The classical theory 127
 CONTENTS vii

3.1 .2 A general recipe for quantization 128

3.1.3 Perturbation theory for gauges nB + c =0 131
3.1.4 The generalized Feynman gauge 134
3.1 .5 The S-matrix generating functional 134
3.2 Gauge Invariance 136
3.2.1 The Ward-Slavnov identities 136
3.2.2 Transversality and gauge invariance of
the S-matrix in electrodynamics 137
3.2.3 Transversality and gauge invariance of the
on-shell S-matrix for the Yang-Mills field 139

Chapter 4 EUCLIDEAN FIELD THEORY 141

4.1 The Euclidean Rotation 141
4.1.1 Definitions 141
4.1.2 The formal Euclidean rotation of the action
functional 142
4.1.3 Euclidean rotation of the Green functions 143
4.1.4 Properties of the field (j) and the action Se ({j)) 145
4.1.5 Rotation of the Lorentz group into 0 4 146
4.1.6 Examples 148
4.2 Functional Integral Representations 150
4.3 Convexity Properties 154
4.3.1 Quasiprobabilistic theories 154
4.3.2 Convexity and spectral representations 156

Chapter 5 STATISTICAL PHYSICS 159

5.1 The Quantum Statistics of Field Systems 159
5.1.1 Definitions 159
5.1.2 The free theory 161
5.1.3 The average of an operator in normal-ordered form 162
5 .1.4 Diagrammatic representation of the partition function
and the Green functions 165
5.1.5 Periodic extensions of the Green functions 166
5.1.6 Representations by functional integrals 169
5.1.7 The zero-temperature limit 170
5.1.8 The Feynman-Kac formula 171
5.1.9 Convexity properties 172
5.1. 10 Convexity of the logarithm of the partition function 174
5.1 .11 Representation of the partition function of the
free theory by a functional integral 176
viii CONTENTS

5.2 Lattice Spin Systems 180

5.2.1 The Ising model 180
5.2.2 The Heisenberg quantum ferromagnet 182
5.3 The Nonideal Classical Gas 185
5.3.1 A gas with two-body forces 185
5.3.2 A gas with many-body forces 188

Chapter 6 VARIATIONAL METHODS AND FUNCTIONAL

LEGENDRE TRANSFORMS 191
6.1 Phase Transitions 191
6.1.1 Introduction 191
6.1.2 Transformation to the variational problem
in thermodynamics 192
6.1.3 The infinite-volume limit 197
6.1.4 Singular and critical points 199
6.1.5 Description of pha&e transitions 201
6.1.6 Critical and Goldstone fluctuations 204
6.2 Legendre Transformations of the Generating
Functional of Connected Green Functions 207
6.2.1 Functional formulations of the variational
principle 207
6.2.2 The equations of motion in connected
variables 213
6.2.3 The equations of motion in !-irreducible
variables 220
6.2.4 Linear equations and their general solutions 224
6.2.5 Iteration solution of the equations 229
6.2.6. The second Legendre transform 232
6.2.7 The self-consistent field approximation 235
6.2.8 The third Legendre tran~form 238
6.2.9 The fourth transform 243
6.2.10 Stationarity equations, renormalization, and
parquet graphs 249
6.2.11 Symmetry properties of the complete Legendre
transform and the "spontaneous interaction" 252
6.2.12 The ground-state energy 256
6.2.13 Stability and convexity properties of
functional Legendre transforms 258
6.3 Legendre Transforms of the Logarithm of
the S-Matrix Generating Functional 261
 CONTENTS ix

6.3.1 Definitions and general properti~s 261

6.3.2 The classical nonideal gas and the virial expansion 266
6.3.3 The Ising model 269
6.3.4 Analysis of nonstar graphs for the Ising model 272
6.3.5 The second Legendre transform for the classical gas 278
6.3.6 The stationarity equations and the self-consistent
field approximation 284

Appendix 1 NONSTATIONARY PERTURBATION

THEORY FOR A DISCRETE LEVEL 289

Appendix 2 GRAPHS AND SYMMETRY COEFFICIENTS 301

REFERENCES 305

SUBJECT INDEX 309

Foreword

I am pleased to present to English-speaking physicists this classic of the

Russian physics literature, published more than 20 years ago in the USSR.
The original Russian edition was very popular throughout the former USSR
and Eastern Europe: the initial printing of 3000 copies sold out in one
month and then the book became rare. However, it has remained almost
completely unknown in the West until a few years ago, when it arrived with
the recent wave of Russian physicists. Indeed, it is thanks to one of these
physicists, Dr Alexander Bochkarev, that it was first brought to my atten-
tion.
Functional Methods in Quantum Field Theory and Statistical Physics is
primarily a textbook designed for training young physicists for research in
theoretical physics, or for theorists desiring to broaden their technical skills.
The techniques it describes remain as important now as they were 20 years
ago and continue to be extended into various subfields of, for example, con-
densed-matter physics. Indeed, this text contains the most complete exposi-
tion of the functional variational technique and Legendre transforms still to
be found anywhere.
Finally, in this English edition the author has corrected some misprints,
and has made a few changes and additions in order to reflect the recent
progress in the field.

Patricia A. Millard
University ofPittsburgh, USA

xi
Preface

This book is based on a revised and expanded series of lectures given by the
author over several years to fifth-year students specializing in theoretical
physics at the Physics Faculty of St Petersburg State University (formerly
Leningrad State University). It is intended for a well prepared reader famil-
iar with the basics of quantum field theory and statistical physics.
The book is devoted not to physics as such, but to certain techniques used
in physics. The term "functional methods" in the title refers to a set of tech-
nical methods which allow the quantum-mechanical operator formalism to
be interpreted into the language of classical objects: nonlinear functionals.
This language is very well suited to describing the expressions which
typically arise in field theory, for example, Wick's theorem. In addition, it is
universal and can be used in widely differing areas of theoretical physics:
quantum field theory, Euclidean theory, the quantum statistics of field and
spin systems, and the statistics of the classical nonideal gas.
In Chapter 1 we systematically discuss the mathematical apparatus of
ordinary (pseudo-Euclidean) quantum field theory, and in the following
chapters this formalism is made specific and generalized to the case of
Euclidean theories and statistical physics. The short Chapter 3 is devoted to
the theory of the massless Yang-Mills field, and the final Chapter 6, which
makes up nearly a third of the book, is devoted to variational methods and
functional Legendre transformations.
The material of Chapters 1, 2, 4, and 5, except for that in certain sections,
is not new, but it cannot all be found in a single place. The material pre-
sented in these chapters would most accurately be thought of as the "folk-
lore" of modem quantum field theory. Of course, this folklore did not
appear spontaneously~ it was created by the labor of many people, and I
hav~ not attempted to give any detailed list of references.
This book consists of six chapters divided into sections, most (but not all)
of which are in tum divided into subsections. Throughout the book, parts
of the text are referred to by giving the chapter, section, and, if applicable,
subSection number. The displayed equations are numbered sequentially

xiii
xiv Preface

within each chapter by the chapter and equation number. The system of
units with n = c = 1 is used everywhere.
It is my pleasant task to thank everyone who helped me with this book:
L.D. Faddeev and V.A. Franke, who read the manuscript and made anum-
ber of useful comments which were taken into account in the final version
of the text; and A.G. Basuev, N.M. Bogolyubov, A.K. Kazanskii,
A.V. Kuz'menko, Yu.M. Pis'mak, and R.A. Radzhabov, who provided
essential assistance in the preparation of the original text. I am also grateful
to all my colleagues and the graduate students in theoretical nuclear and ele-
mentary particle physics at the Physics Faculty of St Petersburg State Uni-
versity for many useful discussions about the topics dealt with here. But
above all, the publication of the manuscript can be attributed to my wife,
who created the necessary working conditions. It is to her that I dedicate
this book.

I am pleased that this English translation is now available to a wider com-

munity of readers, and hope that it will prove useful to physicists studying
quantum field technique.
 Chapter 1

THE BASIC FORMALISM OF FIELD THEORY

1.1 Fields and Products

1.1.1 Canonical quantization
In classical physics the dynamics of a system with a finite number of degrees
of freedom are determined by the Lagrangian !£, which is given as a function
=
of the generalized coordinates q ={ql···qnl and velocities 4; a,q;. The
classical equations of motion are obtained by requiring that the action
functional S =Jdt P(t) be stationary.
The classical system is usually quantized using the canonical formalism,
i.e., in the language of coordinates and momenta. The momenta Pi conjugate
to the coordinates q; are given by Pi= apJaq; .If these relations can be solved
uniquely for the velocities, it is said that the Lagrangian is nondegenerate. In
this case the classical Hamiltonian .Jf= I:.p;4;- Pcan be viewed as a known
function of the coordinates and momenta.
In going to the quantum theory the coordinates and momenta are realized
as linear Hermitian operators qi and pi acting in some Hilbert space and
satisfying the canonical commutation relations ?JsPm- Pm?Js = iosm" The
quantum analog of a classical observable F(p, q) is the operator F (p, q) ; in
particular, the quantum energy operator, the Hamiltonian H, is by definition
.Jf(p, q). These rules for going from the classical theory to the quantum
theory are the well known recipe for canonical quantization.
A function of noncommuting operators .Jf(p, q) is not defined uniquely
even when the natural additional requirement of Hermiticity is imposed. First
of all, there is an arbitrariness in the ordering of the factors p and q, i.e., an
arbitrariness in the choice of how the classical Hamiltonian .Jf(p, q) is written.
For example, the identical classical functions p 2 + q 2 and (p + iq)(p - iq)
correspond to different operators when quantized. This implies that, in
general, a given classical system corresponds to many different quantum
systems, and the choice of any particular one is always made on the basis of
an auxiliary condition.
In addition to this "purely algebraic" ambiguity, there is also an ambiguity
in the choice of realization of the operators p and q: the commutation
2 A. N. VASILIEV

relations do not detennine it uniquely, but up to a unitary equivalence (the

well known uniqueness theorem of von Neumann pertains only to the Weyl
form of the commutation relations). The realization must be chosen on the
basis of additional considerations regarding the nature of the canonical
variables p and q: whether they are variables like Cartesian coordinates and
momenta or variables like angles and angular momenta and so on. For the
actual systems usually considered there is some natural realization which is
always what is chosen. In the case of Cartesian coordinates and momenta this
is the well known representation in terms of multiplication and differentiation
operators.
The recipe for canonical quantization, with all these caveats, is easily
generalized to systems with an infinite number of degrees of freedom, the
generalized coordinates of which. as a rule, are described as a "field" q(x) or
a finite set of such fields. The argument x plays the role of a continuous index
numbering the coordinates. The Lagrangian will now be not a function, but a
functional of the coordinates q(t, x) and velocities 4 (t, x) on the surface
t = const. In the definition of the momentum the ordinary derivative is
a
replaced by the variational derivative, and the in the commutation relations
a
is replaced by the function [1]:

q(x) p (x') - p(s'') q(x) = ;a (x- x'). (1.1)

In this book the term "field" will be used to refer to the notation for the
generalized coordinates of any ·system, independently of whether we are
discussing a true field theory or a system with a finite number of degrees of
freedom. In particular, the particle coordinate in mechanics is treated as a
field depending only on time. To economize on notation and make the formu-
las more uniform we shall often use universal notation, in which the entire set
of generalized coordinates is described by a unified field <p(x) =<p(t, x). Here
the argument x denotes the set of continuous and discrete variables (indices),
except for time, on which the field depends. For complex quantities, cp and cp+
= =
are assumed to be different components of a single field {cp cp 1, cp+ CJ>2),
and the index distinguishing these components is included in the argument x.
The symbol Jdx ... will denote integration over all continuous components
and summation over all discrete components of x, and the symbol (x- x')a
a
will denote the product of functions for all continuous components and
a a
Kronecker symbols for all discrete components of X. Finally, (x- x') =
a{t- ()li(x- x') and I dx .. .=Jdt Jdx .. ., where, unless stated otherwise, the
integration over time is always understood to run from -oo to +oo.
In this notation the commutation relation (1.1) holds for all systems.
 FUNCflONAL METHODS 3

1.1.2 The classical free theory

In this book the term "theory" is used in various ways: in a broad sense
(quantum field theory, relativistic theory) and in a narrow sense, to refer to a
particular system or class of systems with certain dynamics [the free theory,
the A.q>4 theory, the theory with action ( ... ), and so on]. For example, the Ising
model is a particular theory in this sense. (This terminology has become
common in the specialized literature, since it makes it possible to avoid
repetitions of constructions like "the system with interaction Lagrangian
M~>"'' and so on.)
Systems for which the Lagrangian is a quadratic form of the field or its
derivatives of finite order will be referred to as free. For such Lagrangians the
classical equation of motion is linear and can be written as

Kq>=O, (1.2)

where K is some linear operator on a set of fields q>, the explicit form of which
is determined by the Lagrangian. It is often convenient to take K to be a linear
integral operator

[Kq>] (x) = Jdx'K(x,x')q>(x') (1.3)

with suitably chosen kernel K(x, x').

We shall term a linear operator K as t-local if it is an ordinary differential
operator of finite order in the variable t. We shall term afunctional F(t, q>) as
t-local if it depends only on the field and its derivatives of finite order at a
fixed instant of time t. Any Lagrangian is an example of a t-local functional,
and the t-locality of the linear operator Kin (1.2) follows from the t-locality
o
of the Lagrangian. For at-local operator the kernel K(x, x') contains (t- ()
and finite-order derivatives of it, so the integration over the time ( in (1.3) is
removed.
The free action corresponding to (1.3) can formally be written as a
quadratic form:

0
S (q>) = ~q>Kq>=~Jfdxdx'q>(x)K(x,x')q>(x'), (1.4)

but actually to go from the usual expression s0 = Jdtf£0 (t) to the form (1.4)
it is necessary to integrate by parts and discard the surface terms. For
=
example, for a free particle in one-dimensional space q> = q(t), P 0 mq 212,
K = -ma7,
and K(t, () = -ma7o(t- ().The expressions 2S0 _= mldtq 2 and
4 A. N. VASILIEV

2S0=-m Jdt q q obviously differ by a surface term. In a true field theory, in

going to the form (1.4) it is usually necessary to integrate by parts not only
with respect to time, but also with respect to the spatial arguments x.
Equation (1.4) can be identified as the free action only when the latter is
treated as a functional on a space of fields which fall off sufficiently rapidly,
and the kernel Kin this case can without loss of generality be assumed to be
symmetric: K = KT, where KT is the transposed operator, the kernel of which
is obtained by interchanging the arguments of the original kernel
[KT(x, x') =K(x', x)]. This definition of transposition is also applicable to
differential operators, since they can be represented by integral operators with
8-like kernels, and it is equivalent to the following simple rule for
transposition of the first derivative: dT =-d.
=
The equation K KT expresses the following: the operator K in (1.2)
obtained from the requirement that the free action be stationary will always
be a symmetric linear operator on a space of fields which fall off sufficiently
rapidly.
Let us be more precise about what we mean by symmetry of the kernel for
a complex field q>, q>+. The form q>+L<p with arbitrary kernel L can always be
written as the half-sum [q>+Lq> + q>LTq>+]/2, which in universal notation
q> = ci> 1, q>+ =Cl>z takes the form

(1.5)

By definition, the transposition of a block operator is the transposition of

the corresponding matrix accompanied by the transposition of each of its
blocks. Clearly, the matrix kernel K in (1.5) is symmetric in this sense,
independently of the properties of the kernel L.

1.1.3 Anticommutingfie/ds
In quantum theory there are also systems whose operators corresponding to
canonical variables satisfy not commutation relations (1.1), but anti-
commutation relations:

q(x)p(x') +p(x')q(x) = iS(x-x').

Such relations arise in the theory of identical particles obeying Fermi
statistics. Historically, they were arrived at via the well known procedure of
second quantization, but they can also be viewed as the natural recipe for the
ordinary "first" quantization of not completely ordinary classical systems, the
 FUNCTIONAL METHODS 5

generalized coordinates of which are not simple functions but rather anti-
commuting functions. Here we shall give only the small amount of informa-
tion on anticommuting quantities that we shall need in this text (more details
can be found in [2]).
In the language of mathematics, a set of a finite number of pairwise anti-
=
commuting objects* cp cp 1.. •q>n forp1s a set of generators of a finite-
dimensional Grassmann algebra, and the algebra itself is defined as the set of
all polynomials constructed from cp1. . . q>n. The number of independent
monomials q>; 1 . .. q>;k is finite, because, first of all, any two monomials
differing only by interchange of the factors coincide up to a sign and, second,
no monomial can contain any of the generators cp; twice: by permutations the
two identical factors cp; could be put next to each other and then it would be
possible to use the equation cp? = 0 following from the anticommutativity of
the generators. From this it follows that the monomial of highest degree is the
product of all the generators cp;. The total number of all independent
monomials, including unity, is easily found to be 2n, i.e., the Grassmann
algebra with n generators is a finite-dimensional space of dimension 2n and
its general element can be written as a linear combination

t< cp) =to+ D1 u> q>i + l:t2 (i, k) q>iq>k + .. · (1.6)

i i<k

with arbitrary numerical coefficients fk (il .. . ik).

Ordinary functions of the type exp cp; are understood as series which, as a
rule, are truncated (for example, exp cp; =1 + cp;).
All odd monomials anticommute with each other, and every even
monomial commutes with any other one.
For a Grassmann algebra we introduce the concept of derivatives with
.....
respect to cp;, with a dl.stinction made between "left" derivatives l 'd/dcp;) and
"right" derivatives (dldcp;). The action of the left derivative 'dl'dcp; on an
arbitrary monomial can be determined as follows. If the given monomial does
not contain the factor cp;. the result is zero, while if it does contain cp; (only
once) this factor must be moved to the position farthest to the left by means
of permutations and then deleted, or (equivalently) simply deleted followed
by multiplication by the sign factor ± 1 according to whether the number of

• In this book we shall often encounter functions and functionals of many variables,
and for compactness of notation we shall not write out the ellipsis when listing argu-
ments [in, for example, "functional of the variables cp1••• cp." or W.(x1••.x.)]. In those
relatively rare cases where we are speaking of a product of quantities rather than a list
this will be clear from the context or will be indicated by the term "monomial" or
"product."
6 A. N. VASILIEV

permutations needed to move <p; to the extreme left position is even or odd.
The right derivative is defined analogously, except that here <p; must be .
moved to the position farthest to the right. The derivatives are different,
because the parities of the permutations to the left and right are in general
different. It is easily checked that the left and right derivatives commute with
each other, and that identical derivatives anticommute.
When this language is generalized to the case of a field <p(x) the argument x
is treated as a continuous index. The field cp represents a set of anticommuting
generators of the Grassmann algebra (i.e., cp(x)cp(x') = -cp(x')<p(x)], and
functionals F(cp) are elements of this algebra. We introduce the left and right
variational derivatives with respect to cp(x), each of which acts on an
individual factor cp (x') according to the usual rule:

-+ +-
ocp(x')/~cp(x) = ocp(x')/~cp(x) = ~(x-x') . (1.7)

If the differentiated field is separated from the symbol 0/0<p by other

factors, it must first be moved by permutations to one of the extreme positions
(depending on which derivative occurs), and then differentiated according to
the rule (1.7). Variational derivatives possess the same commutation
properties as ordinary derivatives.
Let us list some useful formulas which we shall need later; their proof is left
to the reader.
For any functional F(<p) and any even n

From this, taking into account the anticommutativity of like derivatives, we

obtain

-+ -+
= -~/ocp (x)
f- f-
~/~cp (x) . ~/~cp (x') . o/ocp (x') . (1.9)

Now let cp(x) and A(x) be a pair of anticommuting (including with each
other) fields. Then

...
exp (-cpA) F(~/ocp) exp (cpA)
...
= F(A +~/~cp), (1.10)

...
exp(Ao/ocp)F(cp} = F(cp+A) . (1.11)
 FUNCOONAL METHODS 7

Here and below we use the abbreviated notation

<pA= Jdx<p(x)A(x), ....

Aolo<p= ....
fdxA (x)o/o<p(x).

We note that the operation o/O<p anticommutes with the factor A owing to
the anticommutativity of A and <p. From (1.10) we have

....
F(o/o<p)expq>A = F(A)expq>A. (1.12)

Equations (1.10)-(1.12) are direct generalizations of the usual rules for

taking variational derivatives. The fact that these generalizations involve the
left derivative is a consequence of the notation we have used: all operations
are placed to the left of the functional and act sequentially-first the nearest
one acts, then the next nearest, and so on.
In what follows systems with ordinary fields will be termed bosonic and
those with anticommuting fields will be termed fermionic. The free equation
of motion for fermionic fields also has the form (1.2), but the operator K, the
kernel of the form (1.4), will now be antisymmetric rather than symmetric
owing to the anticommutativity of the factors <p(x)<p (x'). For brevity we shall
term the kernel K symmetric in both cases and write K = xKT, always having
in mind the symmetry consistent with the statistics. Here and below
throughout the book x =1 for bosons and x =-1 for fermions. For a complex
fermionic field (and in real models the fermionic fields are always complex),
the symmetry of the kernel is manifested, as in the bosonic case, only when
universal notation is used:

Integrals on the Grassmann algebra will be studied in Sec. 1.6.2.

1.1.4 The normal-ordered product offree-field operators

The following discussion pertains equally to relativistic quantum field theory,

to nonrelativistic quantum mechanics in the second-quantized representation,
and, finally, to the ordinary quantum mechanics of systems with a finite
number of degrees of freedom. In all these cases the procedure for canonical
quantization of the free theory is well known, and we shall not dwell on it.
Additional information about various specific systems can be found in
Chap. 2, and here we shall restrict ouselves to the following remarks.
8 A. N. VASILIEV

The time dependence of the operators of a quantized free field in the

interaction representation, which will henceforth be referred to simply as
free-field operators, is determined by the quantum-mechanical evolution law
<jl(x) =<P(t,x) = exp(iB0 t)<jl(O,x)exp(-iB0 t),where <jl(O, x) is the
field operator in the Schrooinger picture and Ho is the free Hamiltonian. At
the same time the operator <P (x) always satisfies the classical equation of
motion (1.2). The field operators at different points (anti)commute with each
other to give a c-number.
Let us now turn · to the definition of the concept of the nomu:zl-ordered
product. We assume that the free-field operator cp (x) can be written as
the sum of two terms a(X) + h (X) such that a(x) a(x') =X a(x') a(x) ,
b (x) b (x') = x b(x') b(x) , and

a(x) b (x') -xb (x') a(x) = n (x, x')' (1.13)

where x = ± 1 depending on the statistics and n(x, x') is some c-number

function, henceforth referred to as the simple contraction of the field <p.
We shall say that the product of several factors a (x;) and b (yk) is in
normal-ordered form if in this product any of the factors b is located to the
left of any of the factors a. The normal-ordered product of the factors
a(x 1) ••• a(xn), b(y 1) ••• b(Ym) arranged in any order is defined as the
expression

(1.14)

in which £p is a sign factor, always equal to 1 for a bosonic field. For a

=
fermionic field ep 1 if the original product is reduced to the product in
(1.14) by an even number of permutations of then+ m factors a and b, and
£p =-1 if the number of needed permutations is odd.
Denoting the normal-ordered product as N, we can write

where p denotes an arbitrary permutation of the factors a and b.

This definition can be extended by linearity to any polynomial forms
a
constructed from the operators and b' i.e.,
 FUNCfiONAL METHODS 9

N{Ln ... } = LN{n ... },N{o:n ... } = o:N{fl ... }.

N{o:} =o:, (1.15)

where n ... denotes any product of the factors and b and 0: is an arbitrary
a
constant (a c-number).
The rules (1.15) define the symbol N as a linear operator on a set of poly-
nomial forms constructed from a and b; each such form 'P is uniquely
associated with a form J:' = NJ:. However, it does not follow from this that
N can be understood as a linear operation on a set of operators. There exist
polynomials which are equal as operators, and from the operator equation
J: 1 = J:2 it does not follow that NJ: 1 =NJ:2• A counter-example is the
operator equation (1.13), for which

N[a(x)b(x') -xb(x')a(x)] = O#n(x,x') = N{n(x,x')}.

This shows that in the symbol N J; it is in general not possible to replace the
operator 'P by the operator 'P' equal to it.
Representing each factor of the product <P ( x 1) ... <P ( xn) as a sum a + b
and using the rules (1.15), we thereby define the symbol
N [ <P (x 1) •.. <P<xn)] . In particular, in this manner we obtain

N [ <P (x) <P (x')] = <P (x) <P (x') - n (x, x') , (1.16)

where n is the same function as in (1.13).

From the definition it is clear that theN-product is symmetric, i.e., the fields
inside the N-product behave as classical objects-they commute in the case
of bosons and anticommute in the case of fermions:
N{P[<P(x 1) ... <P(xn)]} = = EpN[<P(x 1) ... <P(xn)]. Here P is an
arbitrary permutation of the factors <P (x;) and Ep has the usual meaning.

1.2 Functional Formulations of Wick's Theorem

1.2.1 Wick's theorem for a simple product

The well known (see, for example, [1]) Wick's theorem specifies the rule for
reducing the product <P<x 1) ... <P<xn) to normal-ordered form. For n 1, 2 =
we have
10 A. N. VASILIEV

4> (x) = N <j>(x), <i>(x)<i>(x') = N[<j>(x)<i>(x') +n(x,x 1

)]. (1.17)

The first of these equations is obvious, and the second is the same as (1.16).
We shall give a proof by induction of the validity of the following rule for
reduction to the N-form:

<P<x.) ... <j>(xn) = ~{ n (~ + :-n;-)"'·

i<k 'P; cpk
(x.) .. . cpn (xn>} 1
...
.

(1.18)

where the symbol! , as below, stands for 'PI = <pz = .. . = <9n = 4>. In this and
the following expr~~~ions <p(x) stands for the classical analog of 4> (x) , i.e.,
the ordinary classical field for bosons and the anticommuting field for
fermions. The fields <p1... <pn on the right-hand side of (1.18) are treated as
independent functional arguments, and the differential quadratic forms are
defined as ·

0 0_ 0 I 0
~n~ = Jldxdx ~(
1
) n(x,x) 0 ( 1) ·
u<p; u<pk ucp; X cpk X

For i < k and s < m we have

(1.19)

For anticommuting fields all the variational derivatives in (1.18) will be

assumed to be right ones. Equation (1.19) is then valid for both types of
statistics.
The objects inside the curly brackets in (1.18) are classical: they are
ordinary functionals and their variational derivatives. As explicitly indicated
in (1.18), all the classical fields <p; remaining after differentiation must be
replaced by the quantum operator 4>.
Turning now to the proof of the reduction formula (1.18), we note that it
coincides with (1.17) for n =1,2. We shall assume that it holds for any number
k ~ n of factors 4>, and shall prove that it is also valid when the number of
factors is n + 1. Using the more compact notation <j>(x;) =4>;. we write the
last factor 4> n + 1 as the sum an+ 1 + hn + 1 • Using the commutation relations
for the operators a and h' we obtain
 FUNCTIONAL METHODS 11

n
+ L ekn (xk' xn + 1) [ (j>1 ... <Pn1 k' (1.20)
k=1

where [ <1>1..• <Pn1 k denotes the product (j> 1.•. <Pn without the factor <P k• and e
and Et are sign factors. For bosons they are always equal to 1, and for
=
fennions e (-l)n and ek (-1)n-k.=
The factors a and b in the first two terms on the right-hand side of (1.20)
are arranged in the correct ("normal") order and, using the rule (1.18) for the
n factors (j>, the sum of these two terms can be written as

(1.21)

where g.'n is the reduction operator for n factors:

(1.22)

and the symbol! has the same meaning as in (1.18).

The expression..in the curly brackets in (1.21) is an operator inside the
N-product. Using the symmetry of the latter, the factor bn + 1 can be moved
to the extreme right-hand position. Here it is important that all the monomials
in the operator polynomial g'ncp 1 •.• cpnl have the same parity, since the
fields are contracted in pairs. From this 1tfollows that the permutation of this
polynomial with bn + 1 gives rise to an additional (for fennions) sign factor e
which is exactly the same as in (1.21), so that the operators a and b are
grouped together in the extreme right-hand position to form the field operator
cpn + 1 . This proves that Eq. (1.21) can be written as N { g.'ncp 1 .•. cpn + 1 } I .
We recall that the reduction operator g.'n does not contain a derivative with
respect to cpn + 1•
Let us now consider the sum over k on the right-hand side of (1.20). Using
Eq. (1.19) and the reduction rule (1.18) for the product [ <P1... (j>n] k
containing n - 1 factors, we write this sum as

(1.23)
12 A. N. VASILIEV

We have written the reduction operator for [ <P1. .. <Pn1 kin the same form
(1.22) as for the operator with n factors, because the extra terms containing
derivatives with respect to <Pk do not contribute, as the field <Pk does not appear
in the differentiated expression.
It is now necessary to return the factors cpk and cpn + 1 to their natural
location. In the fermionic case, taking both derivatives to be right ones, we
have

The differentiation with respect to cpn + 1 from the right does not introduce a
sign factor, while the next differentiation with respect to cpk from the right
gives a factor (-1 )n- k equal to the parity of the permutation of cpk to the right.
The sign factorek in (1.23) is also equal to (-l)n-k, which allows us to write
(1.23) as

Adding to this the expression obtained earlier for the sum of the first two
terms on the right-hand side of (1.20), we conclude that the complete
reduction operator for n + 1 factors is written as

(the equality :: is valid because the differentiated expression is linear in

t>·
cpn + Comparing this with the definition of the operator .'1'n in (1.22), we
find that the operator given above is .'1'n + 1, and thus Eq. (1.18) is proved by
induction.
We again note that in the fermionic case all variational derivatives are
assumed to be right ones.

1.2.2 The Sym-product and the T-product

There are also other "products" in which the field operators behave like
classical objects. First there is the ordinary symmetrized product of field
operators:
 FUNCTIONAL METHODS 13

(1.24)

The summation runs over all n! permutations P of the operators <P, and Ep
has the usual meaning. From this definition and from (1.17) we find

Sym [ <P (x) <P (x')) = N [ <j>(x) <P (x')) + ns (x, x') ,
where ns is the symmetric part of the simple contraction. n:

(1.25)

The time-ordered product or T-product plays a very important role in field

theory:

T[<j>(x 1) ... <j>(xn)] = I,epP{9(t 1... tn)<$(x 1) ... <$(xn)}. (1.26)

p

Here and throughout this book we define 9 tt' =9 ( t - t') and

=n 9(tk-tk+1).
n-1
e(t .... tn) (1.27)
k=l

The summation in ( 1.26) runs over all simultaneous permutations of the

factors <j> (x;) and times I; in the 9 function.
Equation (1.26) is uniquely (and symmetrically) defined only when all the
times t 1... tn are different, and when some or all of the arguments t; are the
same an auxiliary condition is needed. For this we require that the
Sym-product and the T-product be equal on the surface t =const. * This is
equivalent to supplementing the definition of the 9 function (1.27) on its
surfaces of discontinuity (the surfaces on which some subgroups of the
arguments t; coincide) by means of complete symmetrization relative to
permutations of coincident arguments. In particular, for t = ( we take
e,,, e,.,
= = 112. With this auxiliary condition the T-product remains
symmetric also when any subgroups of time arguments coincide.

• Sometimes theN-product rather than the Sym-product is used here; see, for example,
[2].
14 A. N. VASILIEV

For the field to the "zeroth" power the symbols T and Sym are defined using
= =
the usual auxiliary condition Sym 1 T 1 1. We note that Sym and T, like
N, are not actually linear operators on an operator space: neither TF 1 = TF2
nor Sym J; 1 = Sym J;2 follows from the operator equation 1< 1 = J;2.
The time-ordered contraction or propagator of the field cp is the c-number
function A(x, x') defined by the relation

T[<jl(x)cp(x')] = N[cp(x)<jl(x')] +A(x,x'). (1.28)

From the definition (1.26) and Eq. (1.17) we obtain

A(x,x') = 9 11,n(x,x') + x91, 1n(x', x). (1.29)

=
We can write this compactly as A etfn + x9r,nT. By convention, for
coincident times the contraction A is defined by the auxiliary condition

A(x,x')l 1 =t' = ~[n(x,x') +xn(x',x)]l 1 =t' = ns(x,x')lt=t' · (1.30)

=
In contrast to n, the contraction A is always symmetric, i.e., A xAT. The
contractions have another important property: the simple contraction n
satisfies the free equation (1.2) in each of its arguments, while the time-
ordered contraction (1.29) is always the Green function (up to a factor l) of
this equation•:

Kn = KnT = 0, KA = i. (1.31)

The left-hand sides are understood as products of linear operators, and the
i on the right-hand side is a multiple of the unit operator [the kernel of the
latter is o(x-x')]. We shall not present the general proof ofEqs. (1.31), since
they are easily verified for all the specific systems considered in the next
chapter. A special case is that of a massless vector field, which is studied
separately in Chap. 3.

• Often iA instead of A is written in (1.28), and then the equation KA = i takes the form
KA = l. In this notation A= K-1 is the true Green function (without the l). We prefer
the notation (1.28), because then the contraction A= iK- 1 enters as a whole into all the
expressions, but for brevity we shall refer to it simply as the Green function, under-
standing that this means ''up to a factor of i".
 FUNCfiONAL METHODS 15

1.2.3 Wick's theorem for symmetric products

The statement of Wick's theorem is simplified considerably if the symme-
trized product of field operators rather than the simple product appears on the
left-hand side.
Let us first consider the Sym-product. Reducing each term of the sum
(1.24) to normal form using the rule (1.18), we obtain

Sym [.p (x 1) ... .p (xn)] =

= N{ n (1 + ~~<I>; n /' )sym [<pi (xl) · ··<~>n (xn)]} I

i<k q>k
(1 .32)

The symmetrization on the right-hand side is done with respect to all permu-
tations of x 1•• •Xn while preserving the order of the functional arguments
<l>t· ··<l>n· For example,

The fields q>; in (1.32) are classical objects, and by permutations of the
factors q>; we can return the arguments x; to the original order x 1•• •Xn in each
term of the sum on the right-hand side of ( 1.32). The extra sign factor arising
in the <I>; permutations obviously coincides with the factor £p in the definition
(1.24) of the symmetrization operation, and these two factors cancel each
other out. From this we see that the symmetrization on the right-hand side of
(1.32) can also be understood as symmetrization of the bosonic type (all terms
with plus sign) relative to permutations of the functional arguments q> 1•• ·<l>n
while preserving the order x 1•••xn. In the example of (1.33),

We stress the fact that the sign on the right-hand side of this equation is
independent of the statistics.
Let us now tum to the reduction operator and expand each of the differ-
ential quadratic forms into parts which are symmetric and antisymmetric
under permutation of the indices:

(1.34)
16 A. N. VASILIEV

Using the facts that (i) the differentiated functional is proved to be even under
any permutation of the functional arguments <p;. (ii) it is linear in each of the
arguments <p;. and (iii) after differentiation all the <p; are assumed to be equal
to the same operator q,, we conclude that the odd parts of any of the forms
(1.34) do not contribute to the final expression and can be discarded.
Therefore, each factor of the reduction operator can be written as

The replacement of this by an exponential is justified because the extra higher

powers of derivatives do not contribute since the differentiated functional is
linear in each of the arguments <p;. The full reduction operator in ( 1.32) is the
product of i < k operators (1.35):

(1.36)

Again using the linearity of the differentiated functional in <p; , we can add
diagonal terms with i =k to the argument of the exponential in (1.36), after
which the exponent takes the form of a "perfect square":

(1.37)

If we change from <p 1.. ·<lln to new functional arguments, the average
<p =(<pi +...+ <!ln)ln and the differences <p;- <p; + 1, the form (1.37) reduces to
the second derivative with respect to <p:

o 0<1>; o o
-=I,-.-
. o<p . o<p o<p.
= I,- .
. o<p. '
(1.38)
I I I I

After performing the differentiation we must replace all the <p; by q,. Clearly,
it is possible to set <p 1 =<p2 =... = <lln = <p even before the differentiation,
because the form (1.38) does not contain derivatives with respect to the
differences <p;- <p; + 1• This means that Wick's theorem ( 1.32) can be rewritten
as follows:
 FUNCTIONAL METHODS 17

Sym[cp(x 1) ... cp(xn)] o o

= N{exp(-21 F"""nF""")q>(x 1) ... q>(xn)}
I •·
oq> oq> ~=~

(1.39)

The symbol Sym on the right-hand side has been omitted because the
product q>(x 1) ••• q>(xn> is automatically symmetric. We also note that the
contraction n in (1.39) can be replaced by its symmetric part (1.25), because
the kernel of the differential quadratic form is automatically symmetrized
owing to the commutativity of the derivatives in the bosonic case and the anti-
commutativity of like derivatives (in this case, right ones) in the fermionic
case:

(1.40)

Equation (1.39) is better than (1.18) in the sense that the reduction operator
in (1.39) is universal, i.e., it is independent of the type of expression reduced.
Let us now turn to the Tproduct. From the definition (1.26) and Eq. (1.18)
we have

T [ cp(x 1) •.• cp(xn)] = (1.41)

= N{ n (1 + ;-n;-)L£pP
i<k <i>; q>k
[9 (tl"""tn) q>l (xl) ... q>n (xn)]} I .
p .. .

The summation on the right-hand side runs over all simultaneous permuta-
tions of the arguments X1..• Xn and times t 1... tn in the 9 function with the order
of the functional arguments <p 1... <i>n preserved. It is important to note that in
each term of the sum over permutations the field with the higher index always
has the smaller time, i.e., from i < k it follows that the time argument t; of the
field q>; is larger than that of the field <i>k· This follows directly from the
definition of the T-product and (1.18): the time arguments of the fields
decrease in going from left to right in all the terms of the sum ( 1.26), and the
indices of the fields in (1.18) increase in·going from left to right.
Let us first assume that among the times t 1.. . tn there are none which
coincide. It is then clear from the above discussion that in all the differential
fonns (0/0<p;)n(O/O<pk) of the reduction operator in (1.41) it is possible to
replace the contraction n by ~. using the fact that for t > ( these two
contractions coincide. Then, using the symmetry of the contraction ~. the
18 A. N. VASILIEV

form (&~q>;)t\(&~cpk) can be symmetrized relative to interchange of the

indices:

Arguing as in the proof of the preceding theorem, the reduction operator

can be represented as the exponential of a "perfect square," after which all the
fields cp; are set equal even before differentiation. Then the right-hand side of
(1.41) takes the form

It should be noted that the product of classical fields is symmetric, i.e.,

=
EpP[q>(xl) ...cp(xn)] cp(xl) ... cp(xJ, and so the sum over Pin (1.42) is

since the sum of all permutations of the e function is unity. Therefore,

T[ <j}(x 1) •.. <P<xn) 1=

=N {expG · ~~t\s~]q>(x 1 ) ••. q>(xn)} lcp=IP· (1.43)

We have derived this expression assuming that all of the times t 1.. . tn are
different, but it can easily be checked that it is true also in the general case: if
the contraction t\ in (1.43) for coincident times is given by (1.30), the
right-hand side of (1.43) defines the T-product when arbitrary groups of time
arguments coincide, just as in the preceding section.
Equation (1.43) is the functional statement of the Wick time-ordering
theorem. It was obtained in this form by Hori [3].

1.2.4 Reduction formulas for operator functionals

Expressions of the form

 FUNCfiONAL METIIODS 19

co

F (<jl) = I, J. .. Jdx1 ••• dxn Fn (x 1 ••• xn) <P (x1) ••. <jl(xn), (1.44)
n=O

which are specified by a set of their, possibly generalized, coefficient

functions Fn(x 1•••xn), will be called operator functionals. We shall term a
functional symmetric if all its coefficient functions are symmetric. As always,
· here we mean the symmetry consistent with the statistics. A symmetric
operator functional can be associated with a classical functional, replacing the
free-field operator <P in (1.44) by the classical argument <p. The symmetric
functions Fn are uniquely determined by the functional F(<p). We also note
that for a symmetric functional F ( <P) = Sym F ( <P) .
To avoid misunderstandings, we again note that an operator functional
( 1.44) is specified by its coefficient functions, and not by the operator F ( <P).
The functions Fn uniquely determine the operator F ( <P) , but the converse is
not true, because the free field <P is not a completely arbitrary argument, but
one of the solutions of (1.2). The set of symmetric functions Fn is in
one-to-one correspondence with the classical functional F(<p), and to uniquely
determine the operator F ( <P) it is sufficient to know the functional F( <p) on
the mass shell, i.e., on the set of all solutions of the free equation (1.2). The .
concepts "operator" and "operator functional" are not identical.
Equations (1.39) and ( 1.43) of the preceding section also can be generalized
directly to operator functionals owing to the universality of the corresponding
reduction operators. This is the real virtue of universality. The same is true for
t-local functionals (see the definition in Sec. 1.1.2), if they depend only on the
field on the surface t =const, and not on its time derivatives. The point is that
we have taken special care to supplement the definition of the T-product of
fields for coincident times, but we have not yet done this for the analogous
expressions involving time derivatives of the fields (see the next section).
Excluding such functionals from consideration for now, we rewrite the
reduction formulas (1.39) and (1.43) as

SymF(<p) = Nexp
A ~ ~ JF{<p) I • ,
[12 · SnS (1.45)
<p <p <P=<P

TF(<p) A
= N exp [1- · -~. d~-
2 ~<p ~<p
J F(<p) I
<P =<P
. (1.46)

From this we easily find the expressions for the inverse transformations and
all possible combined transformations. For example,
20 A. N. VASILIEV

NF(f9)
.
A
= Symexp [ 0 OJ
--1 · - n - F(f9)
2 Of9 Of9 ,=~
I• (1.47)

Starting from the general statement of Wick's theorem (1.18) and arguing
exactly as in the derivation of (1.39), we obtain the following expression
generalizing (1.45) to the case of the product of several symmetric operator
functionals:

lln [SymF;(<i>)] [1 I_-n-·+

= N {exp -
00
i=I 2 i Of9; Of9;

(1.48)

Here and below the noncommuting factors on the left-hand side are assumed
to be arranged in increasing order:

n
ll [Sym F; (</>)] = Sym F 1 (<!>) · ... · Sym Fn (<!>) .
i=1

The diagonal terms of the quadratic form of the derivatives in (1.48)

generate contractions inside the cofactors F; which reduce them to normal-
ordered form, and the nondiagonal terms give contractions between different
cofactors.
Equation (1.48) has an obvious generalization to the case where several of
the factors on the left-hand side are written in N-ordered, rather than
Sym-ordered, form:

n [.<IF;<<~>)]=
n

i =1

(1.49)
 FUNCfiONAL METHODS 21

Here d stands for Sym or N, and the summation in the diagonal terms of the
quadratic form runs over only those fields which correspond to factors in the
Sym-ordered form. If all the factors are written in N-ordered form, there are
no diagonal terms.
If a product of operator functionals appears inside some kind of symmetric
product it is automatically symmetrized, and the usual expressions (1.45) and
(1.46) can be used. For example,

(1.50)

However, it is sometimes convenient to explicitly split the contractions

between different cofactors. This can be done by rewriting (1.50) as follows:

n
T{ fl F; (<P) } =
i =I

(1.51)

The transformation from this_representation to the previous one is accom-

plished using the usual trick based on (1.38).
By means of the simple Wick theorem (1.18), any (not necessarily symmet-
ric) operator functional F( <P) can be reduced to the form NF(NJ( <P ), and then
(1.45) and (1.46) can be used to make the replacements N-..+Sym-+T. The
equations F( <P) =NF(NJ( <P) =Sym F(Sym)( <!>) =TFm( <!>) define, in terms of
the original F( <P ), the symmetric (by condition) functionals F(N)(<p),
F(Sym)(<p), and Fm(<p), which we shall respectively refer to below as the
N-form, the Sym-form, and the T-form of the original F(<!>) , These three
functionals are related to each other by expressions following from (1.45)-
(1 .47). We note that for any F( <P ), by definition F ( <P) = Sym F(Sym)< <P) =
F(Sym)( <!>) owing to the assumed symmetry of F(Sym)( <!> ), but the equation
=
F(<p) F(Sym)(<p) holds only for symmetric F(<!>).

1.2.5 The Wick and Dyson T-products

In this section we shall discuss the question of defining the T-product in those
cases where the expression contains derivatives of the field with respect to the
argument x, in particular, derivatives with respect to time. We define
22 A. N. VASILIEV

T[!l!l (xl) ... !l!n (xn)<P (xl) ... <(l(xn) l =

=gll (xl) ... g)n (xn) T[.P(xl) ... .p(xn) l • (1.52)

where !ll;(x;) are arbitrary differential operators acting on the arguments x;.
The T-product defined with this auxiliary condition is called the Wick
product, distinguished from the Dyson T-product, which will be discussed
below.
The expression inside the T-product on the left-hand side of (1.52) can be
formally represented as an operator functional (1.44) with a singular kernel:

(1.53)

We recall that the transpose of the differential operator !ll(x) is given by the
rule aT= -'d. If we assume that (1.53) is an ordinary operator functional and
write

Tj. .. jdyl ... dynF (yt .. ·Yn) <P (yl) ... .p (yn) =

=f ... fdyl ... dynF (yt .. ·Yn) T[ <(l{yl) ... <(~ (yn)] •

the right-hand side of this equation for the singular kernel (1.53) is exactly the
right-hand side of (1.52). In other words, the rule (1.52) is equivalent to
extending the reduction formula (1.46) to the case of functionals with any
singular kernel, including to t-local functionals with time derivatives of the
field. We note that an equation analogous to (1.52) is always satisfied for the
N- and Sym-products.
Let us now define the Dyson T-product. Let li';(t;) be arbitrary. time-
dependent operators (other arguments, if they exist, are treated as fixed para-
meters), and each of the F; be either a bosonic or a fermionic operator, but
not a mixture of the two types. The Dyson T-product of these operators is
defined as

in which the summation runs over all simultaneous permutations of the times
t; in the 9 function and the operators Ji';(t;) as a whole, and the sign factor Ep
 FUNCfiONAL METHODS 23

is defined, as usual, as the parity of the permutation of fermion factors. In the

special case where each of the I<'; is a free-field operator cp(x;). the definition
(1.54) coincides with the usual Wick T-product (1.26). However, in those
cases where the F; contain time derivatives of the field, the Dyson T-product
(1.54) will not coincide with the Wick T-product defined by (1.52).
For example, let l<'t(t) = <jl(x) and 1<'2 (t') = 1,<jl (x'). Then
o

while

T [ <jl (x) ()1.<jl (x')] = o1,T[<!> (x) <i> (x')] =

These two expressions clearly differ, because the differentiation with

respect to (in the second produces additional terms containing derivatives of
9 functions.

1.3 The S Matrix and Green Functions

1.3.1 Definitions
Let us now consider interacting quantum systems, for which H Ho + V, =
where H and Ho are respectively the full and the free Hamiltonians and V is
the interaction Hamiltonian.
The dynamics of the quantum system can be described by a unitary
evolution operator U('t1, 'tz) in the interaction picture. This operator is the
solution of the equation ioU('t1, 't2)10't1 = V('t1)U('t 1;tz), in which V('t) is the
interaction Hamiltonian in the interaction picture: V('t) = exp (illo't) x
V exp (-illo't). The solution of the equation of motion for U with the initial
condition U('t, 't) =1 is written as

= I. <-i> nf. .. fdr 1•• •drne u1••• tn> v <r1>... v (tn>. (t.55)
n=O
24 A. N. VASILIEV

The integration over each timet; runs from t2 to t 1•

By definition, the limit of the evolution operator U(t 1,t2) for t 1 ---+ oo and
=
t2 ~ -oo, U U(+oo, -oo), is called the S-matrix operator U.
Mathematically, U(t 1,t2) is a well defined unitary operator when Ho and H
are Hermitian (not simply symmetric) operators. For translationally invariant
systems, in particular, relativistic ones, as a rule the operator H does not have
this property. However, even if the operatorU(t 1 ,t~ is well defined, the limit
U = U(oo, -oo) may not exist in a mathematically rigorous sense. The
necessary condition for this limit to exist is that the spectra of the operators
Ho and H coincide, and as a rule this condition is not satisfied. Therefore, in
most cases of practical importance the operators that we consider are only
certain formal constructions from which it is nevertheless possible to extract
all the needed physical information about the system.
Let us now return to Eq. (1.55). After symmetrizing the general term ofthe
series with respect to all permutations of the times t 1••• tn, we express it in
terms of the Dyson T-product of the operators V(t;):

f ... fdt .... dtne u.... tn) v (tl) ... v (tn) =

=-\-n. f .. . fdt 1 .. . dtnTD [V (t1) ... V (tn)], (1.56)

and the entire series (1.55) can then be written as a Dyson T-ordered
exponential:

(1.57)

Equation (1.56) is true only when V(t) is a bosonic operator, which is what
we shall always assume, although in practice one encounters cases where this
is not so. For example, the interaction operator of a two-level system with
quantized electromagnetic field is written in the form a+b + b+a, where b+ and
bare the fermionic raising and lowering operators of the two-level system and
a+ and a are bosonic operators. The interaction in this case is a fermionic
operator and the Volterra series ( 1.55) does not reduce to the Dyson T-ordered
exponential.
An important object of study in field theories are the Green functions of a
field, which are defined as

(1.58)
 FUNCfiONAL METHODS 25

Here I0) denotes the ground state of the full Hamiltonian, the "physical
vacuum," which is assumed to be nondegenerate and normalized to unity;
cPH (x) is the field operator in the Heisenberg picture <PH (t, x) =
= exp (iHt) <P (0, x) exp ( -iHt). The symbol T0 is defined in the
preceding section.

1.3.2 Transformation to the interaction picture in the evolution operator

In field theories the operator V(t) is given explicitly in the form of a certain
t-local operator functional. In this and in the following sections it is assumed
that the operator functional V(t) does not contain time derivatives of the field.
We shall refer to such functionals as simple.
Any operator functional can be written inN-ordered form, after which it is
possible, if desired, to transform to the Sym-ordered form using Eq. (1.47). In
the discussion which follows, a special role will be played by the symmetric
functional 7/(t, <P>. which is the Sym-fonn of the interaction operator: V(t) =
=
Y(t, <P> Sym Y(t, <P>· By assumption, r(t, cp) is a simple t-local functional.*
The quantity

'tl

Sv ('tl' 't 2 ;cp) =- f dtY(t, cp) (1.59)

't2

will be called the interaction functional in the interval 't2 ::;; t $ 'tt for a given
quantum interaction V. The interaction functional on the entire time axis will
be denoted Sv(cp).
Let us immediately emphasize the fact that Sv is treated simply as some
classical image of the given quantum interaction, and not as a characteristic
of the classical system whose canonical quantization leads to the quantum
theory in question. The functional Sv determines the operator V uniquely,
while the canonical quantization in general is not unique.
We can transform to the interaction picture in (1.57) on the basis of the
following statement: for any set t1 (11) ... tn (tn) of simple t-local operator
functionals,

• As noted in Sec. 1.2.4, a classical functional corresponds to a quantum operator only

up to an arbitrary addition which vanishes on the mass shell, i.e., on the set of solutions
of the equation K19 = 0. However, for ordinary K such an addition must necessarily
contain time derivatives of the field, so in the class of simple functionals the Sym-
form of Vis determined uniquely.
26 A. N. VASILIEV

The symbol Ton the right-hand side denotes the Wick T-product, in which
all operators F; are assumed to be represented by their Sym-forms:
I<'; (t;} = Sym-¥; ( cp) = -¥;< cp ). The need for this representation arises from
our choice of definition of the T-product for coincident times (Sec. 1.2.2}.
To prove ( 1.60), we reduce both sides of this equation toN-form and verify
that the resulting expressions are identical.
The right-hand side of (1.60) is reduced toN-form using the rule (1.50):

T[l<'t (t 1) ••• 1<'" (t,.)] =

=N {expG. o~a:cpJ1i (cp) ... fF, (<p)} I (1.61)

In order to reduce the left-hand side of (1.60) to N-form, we use the

definition (1.54), and by means of the rule (1.48) we reduce each term in the
sum over permutations to N-form. As a re'sult, we arrive at the expression

N{exp[!:r,~n~+ L
2 ; ocp; ocp; i<tocp; Jx
ocpJ ::-n °
x:r,epP[9(t1 •. . t,.}~1 (cp 1 ) ••• ~"(<p")] lj , (1.62)
p . ..

in which the summation runs over all simultaneous permutations of the times
t; in the a function and the subscripts of the functionals fij while preserving
the order of the functional arguments cp 1••• cp"; £pis determined, as usual, by
the parity of the permutation of the fermionic factors.
The further arguments are nearly identical to the proof of the Wick time-
ordering theorem ( 1.43). The main consideration is that in all the terms of the
sum over permutations in (1.62) the arguments cp; are ordered with index
increasing from left to right, and this ordering exactly coincides with the
ordering of decreasing times in the T-product. Therefore, in the off-diagonal
terms of the quadratic form of the derivatives the contraction n can be
replaced by a, followed by symmetrization in i r;!k exactly as was done in
the proof of Eq. (1.43):

(1.63)
 FUNCTIONAL METIIODS 27

The diagonal terms of the quadratic form of the derivatives were

unimportant in the proof of (1.43) because they did not contribute. Now they
are important, but in these terms it is also possible to replace n by a using,
ftrSt, the fact that the kernel of the form (&'&p;)n(Ol&p;) is automatically
symmetrized and, second, the fact that these operators act on simple t-local
functionals and therefore the results involve only the values of n8 on the
surface t= (, where, according to (1.30), n8 and d coincide. This proves that
the quadratic form of the derivatives in (1.62) can be written as a perfect
square like (1.38); then, repeating word for word the corresponding part of the
proof of Wick's theorem (1.43) we obtain the desired result.
Using Eq. (1.60) for the case where each of the fl; ( t;) is the interaction
operator V(t;). we represent the evolution operator and the S matrix as a Wick
T-ordered exponential, which can then be reduced to normal-ordered form
using the rule (1.46):

U (tl' t 2) = Texp iSv (tl' t 2 ;<P) =

=Nexp (
1 a a exp iSv(tl't2 ;<p)l =cf (1.64)
20<pao<p) 111

In these equations Sv(t1, t 2; <p) is the classical interaction functional (1.59),

the Sym-form of the quantum interaction.

1.3.3 Transformation to the interaction picture for the Green functions

Let us consider the transformation from the Dyson T-product to the Wick
T-product in the Green functions (1.58), assuming as before that the inter-
action operator does not contain time derivatives of the field.
Let fl; ( t;) be certain simple t-local operator functionals and f;iH ( t;) be
the same operators in the Heisenberg picture:

iHt -iH0 t iH t -iHt

f;iH (t) =e e 1<"; (t) e 0
e = U (0, t) fl; (t) U (t, 0). (1.65)

Here 1<"; ( t) is the operator in the interaction picture and U is the evolution
operator (1.55). By definition,

Using Eq. (1.65) and the group property U(t1, t 2)U(t2, t3) =U(t1, t3) for the
evolution operator, we obtain
28 A. N. VASILIEV

PIH (tl) ... PnH (tn) =

=U (0, t 1) j;l (t1) U (tp t2) P2 (t2) ... Pn (tn) U (tn, 0). (1.66)

Let t 1 and 't2 be arbitrary numbers such that 't2 S t; S 'tt for any of the times
=
t; in the product (1.66). Using the equatiQns U(O, t 1) U(O, 'tt)U(t1, t 1) and
U(tn, 0) =U{tn• 't2)U('t2, 0), we obtain

where

(1.67)

The summation in the last expression runs over all simultaneous permuta-
tions of the operators P; (t;) as a whole and the times t; in the 9 function and
in the evolution operators U.
We shall prove that the right-hand side of (1.67) is

(1.68)

where T denotes the Wick T-product, Sv is the interaction functional ( 1.59),

and ~(<p) are symmetric functionals corresponding to the Sym-form of the
operators P; (t;) in the interaction picture: 1<'; {t;) =
Sym ~ ( <P) = (
~ <P ).
The equality of (1.68) and (1.67) is proved just as for the analogous
equation (1.60}, namely, the right-hand sides of (1.67) and (1.68) are reduced
toN-form and it is checked that the results coincide.
The general term in (1.67) is the product of n operators P; (t;) = ~ (<P)
and n + 1 evolution operators U, each of which can be written inN-form using
(1.64). Each term of the sum over permutations (1.67) can be reduced to
N-form using (1.49}, introducing 2n + 1 independent arguments <p; which will
be arranged in the reduction formula with the index increasing from left to
right. The argument of the exponential of the complete reduction operator
will look like:
 FUNCITONAL METHODS 29

(1.69)

and in the reduced functional each of the operators 'P; must be replaced by .¥;
and each of the operators U by exp iSv for the corresponding time interval. In
the form (1.69) the first sum (with the prime) runs over then+ 1 fields <p;
which are the arguments of the classical interaction functionals Sv, and the
factor exp ( 1/2 "i:.' •••) is the contribution of all the reduction operators ( 1.64)
for the evolution operators U. The second group of diagonal terms (the sum
with two primes) is the contribution of the reduction operators to theN-form
of all the symmetric functionals .¥;; the t-locality of these functionals together
with Eq. (1.30) allows n to be replaced by.!\ in these terms, exactly as in the
proof of (1.60). As far as the off-diagonal terms in (1.69) are concerned, the
usual arguments according to which the field with highest index has the
smallest time remain valid also for the right-hand side of (1.67), because the
time arguments of the fields in the functional U(t1, t2) are located between It
and t2 (here the assumption t 2 S t; S t 1 is important). This allows us to
substitute (1.63) in the off-diagonal terms of Eq. (1.69), so that the latter
becomes a perfect square and all the fields <p; can, as usual, be set equal even
before differentiation. The product of the factors exp iSv for all the evolution
operators in (1.67) is then combined to form the complete exponential for the
interval t 2 S t St1; the sum over permutations, as in the proof of(l.43),leads
to the substitution 9--+ 1, and in the end we arrive at the right-hand side of
(1.68).
The statement we have proved can be written as follows:

To ['PIH (tl) .. . lt'nH (tn) 1=

=U (0, t 1) T['Pl (t1) ... lt'n (tn) exp iSv (t 1, t 2;<P)] U (t2, 0) . (1.70)

We note that in deriving (1.45), (1.46), (1.60), and (1.70) we have essen-
tially used only the fact that fort> (the contractions .!\(x, x') and n(x, x')
=
coincide, and for t (they are related by (1.30). In the chapters which follow
we shall study Euclidean field theory and quantum statistics. In these theories
the contractions .!\ and n will be different, but the functional form of the
= =
r~uction operators and the equalities .!\ n for t > ( and .!\ ns for t will=(
be preserved. Therefore, the proof of Eqs. (1.45), (1.46), (1.60), and (1.70)
remains valid also for these theories.
30 A. N. VASILIEV

Let us now consider the expectation value of the operator ( 1.70) in the true
ground state I0). We assume that the following asymptotic expressions of
nonstationary perturbation theory for a discrete nondegenerate level are well
known: for t 1 -+ oo and t 2 -+ -oo·

Here I 0) is the ground state of the free Hamiltonian, which is assumed to be

nondegenerate, and a and~ are phase factors. The proof ofEq. (1.71) and the
generalization to the case of a degenerate level are given in Appendix 1.
For t 1 -+ oo and t 2 -+ -oo we have

The meaning of the phase factor ~*(t 1 )<x(t2) is easily found from (1.71):
p(tl)<X*('tz)= (0 IU ('tl' t 2) I0 ). Therefore, for t 1 -+ oo, t 2 -+ -oo

0 ITD [tlH (tl) .. . tnH (tn)] I0)=

(OIT{/i'I(t 1) .. . tn(tn) exp iSv(cp)} IO)

= (1.73)
(OITexp iSv(cp) lo}

where Sv is the interaction functional (1.59) on the entire time axis.*

The denominator on the right-hand side of (1.73) is the vacuum expectation
value of the S matrix U =U(oo, -oo ). This quantity determines fl£0 , the energy
shift of the ground state when the interaction is switched on, because for any
nondegenerate discrete level for t 1 - t 2 -+ oo

The proof of this asymptotic expression is given in Appendix 1. The

statement (1.74) as applied to the S matrix will be written as (OIUIO) =
J
= exp [ -itJ.E0 dt] .
The numerator on the right-hand side of (1.73) is called the full Green
function of the operators tiH• while the quotient (1.73) is called the Green

• The S-matrix operator is sometimes written in this expression instead of exp iS.(~).
This notation is careless: strictly speaking, "T-operator" has no meaning at all, in con-
trast to "T-operator functional" (cf. the remark in Sec. 1.2.1).
 FUNCTIONAL METHODS 31

function without vacuum loops. The meaning of this terminology will become
clear later. The quantities

(1.75)

are by definition the full Green functions of the field <PH· The "zeroth-order
function" Go= (OI TexpiSV (cp) I0) is a number, the vacuum expectation
value of the S matrix. The functions

Hn (xi .. . xn) = (0 IT0 [cpH (xi) .. .cpH (xn)] I 0) = G01Gn (xi . . .xn)
(1.76)

defined by (1.58) can now be refered to as the Green functions without

vacuum loops of the field <PH·

1.3.4 Interactions containing time derivatives of the field

In obtaining Eqs. (1.60) and (1.70) reducing Dyson T-products to Wick
products, we have assumed that there are no time derivatives of the field in
the interaction operator V(t). Now we consider the general case where V(t) is
at-local functional depending not only on the field cp (x) itself, but also on
derivatives <Pn =<Y,'<P of finite order. The general case reduces immediately
to the one we have studied if each of the derivatives cp n is assumed to be an
independent field and we define the matrix of contractions of the set of fields
cp n in terms of the Dyson T-product:

The symbol Tin this expression can also be understood as the ordinary Wick
T-product of the independent fields cp n• thereby treating ( 1.77) as a special
case of the general definition (1.28). This means that in the language of a
system of independent fields cp n• all the standard reduction formulas remain
valid (the index n, if desired, can be included in the argument x). In particular,
it is possible to use Eqs. (1.60) and (1.70), since in the language of the fields
q>n the interaction functional Sv('POt q>1, ••• ), which is the Sym-form of the
quantum interaction, d~s not contain time derivatives of the fields.
Therefore,
32 A. N. VASILIEV

(1.78)

The symbol I indicates that after the differentiation each of the classical
fields ({); must 't)e replaced by the quantum field <Pi·
The resulting theory can also be formulated in the language of a singl~ field
q> ='Po by changing the interaction functional in a certain way. To do this, we
isolate from the contrac.tion matrix A.m. the "Wick part" l:i'nm:

ti' (x x')
nm '
=ant amt' ti (x' x') .' (1.79)

The extra piece anm is t-local, i.e., it contains the factor ~(t - () and
finite-order derivatives of it. It is also clear that a 00 =0.
Let us now define the functional ~('Po. fPl> .. .) as

exp iS~ (q>O' q>l' ... ) =

= exp (~I.~~ anm~ ~ ) exp iS" (q>0 , fPp ... ) • (1.80)

nm 'Pn 'Pm .

The S-matrix functional inN-form (1.78) is expressed in terms of~ as

U = Nexp(~L~~ ti'nm~~
nm 'Pn 'Pm
)exp iS~(cp0, q>l' ... ) I =cP·.
q1.
(1.81)
I I

The final thing to note is that .this expression can be viewed as the reduction
formula of the ordinary Wick T-exponential

U = N exp ( - · !:.~ lis:.~ ) exp

1
i~ff (q>) I (1.82)
2 uq> ucp =cp
A

ql

for the interaction

(1.83)

In fact, let <Pn =Kn <P and ti'n, =KnKml:i =Knl:iK!r where Kn are arbitrary
linear operators (for us Kn =if,). We have
 FUNcnONAL METHODS 33

~ ' ~ ~ T 0
!ll=~-6. -=~-KA.K-=
~om nmom ~om n mo<p
nm Tn Tm nm Tn m

=l(K!t)A(K~-!-)·
nm <pn <l>m

i.e., the differential form reduces to a perfect square. From this we find

since

The quantity 0/0<pn on the right-hand side is understood as the partial

derivative with respect to <!>n =Kn<i>• and 0/0<p on the left-hand side is the "total
derivative." These arguments prove Eq. (1.82).
A closed expression for the effective interaction (1.83) can be obtained
when the original interaction functional Sv(<p0,<p1, . . . ) has sufficiently simple
(linear or quadratic) dependence on the derivatives <l>n• n <!: 1. We note that the
effective interaction defined by (1.80) and (1.83) wiii always have
Lagrangian form, i.e., it will be an integral over time of some t-local
functional, a Lagrangian. But this Lagrangian is, in general, complex rather
than real, as in the case usually encountered (see Sec. 1.5.2 regarding the
unitarity of the S-matrix).
It should also be noted that for most cases of practical importance the
scheme we have described of constructing the effective interaction is too
general. Actually, the free-field operator from which the interaction is
constructed satisfies a t-local equation of motion (1.2). For ordinary
Lagrangians this equation is either first- or second-order in the time deriv-
ative (see Chap. 2 for more details). In the first-order equation the derivative
<P 1 is expressed in terms of the field <P itself or its derivatives with respect to
other arguments, and in this case it can be assumed without loss of generality
that the interaction does not contain any time derivatives of the field at all. In
the second-order equation the second derivative of the field with respect to
time can be eliminated in exactly the same way, and it can be assumed that
any interaction is expressed only in terms of the field operator and its first
derivative with respect to time.
 34 A. N. VASILIEV

Up to now we have considered only the quantum theory, assuming V to be

some given operator functional whose origin we are not interested in. Now let
us say a few words about the quantization of classical systems whose inter-
action Lagrangian contains time derivatives of a field. Without loss of gener-
ality, it can be stated that the Lagrangian depends only on first derivatives
with respect to time, i.e., velocities. If the Lagrangian is nondegenerate (the
special features of the quantization of systems with degenerate Lagrangians
are described in, for example, [4]), it can be used to uniquely determine the
classical interaction Hamiltonian r as a function of coordinates and
momenta. In quantization in the SchrOdinger picture, q and p are replaced by
operators, with different arrangements of noncommuting factors
corresponding to different operators V = Y( q, p), i.e., different "quantization
recipes" (see the example in Sec. 2.6). The operator realization of the coordi-
nates and momenta in the Schrooinger picture can always be taken to be the
same as for the free theory. Then in going to the interaction picture the time
dependence of these operators will be determined by the equations of the free
theory, so that the momentum operators in the interaction picture can be
expressed in terms of time derivatives of the coordinate operators (of the free
field). As a result, we obtain the explicit representation of the interaction
operator V(t) in the form of at-local operator functional of the free field, and
the further study of the quantum theory proceeds as described above.
Different versions of the quantization corresponding to different operators
V(t) will be associated with different functionals ~ff. since this functional
determines V(t) uniquely. In particular cases the choice of quantization recipe
is limited by various additional considerations, for example, the requirement
of relativistic invariance in relativistic theories.

1.3.5 Generating functionals for the S-matrix and Green functions

The functional

(1.84)

representating the S-matrix operator in normal-ordered form will be referred

to as the S-matrix generating functional. For an interaction involving time
derivatives of the fields, Sv in this expression should be understood as the
_ effective interaction studied in the preceding section. By definition, the
S-matrix operator is NR(<P). The field <P satisfies the free equation (1.2),
from which we see that to define the S matrix as an operator it is sufficient to
know the functional R(<p) restricted to the set of <p satisfying (1.2). This
 FUNcnONAL METHODS 35

restricted functional will be referred to as the on-shell S-matrixfunctional, in

contrast to the functional (1.84), which provides the off-shell representation
of the S matrix.
Let us also define the generating functional for the full Green functions
(1.75):

G(A) = ;k _!_G
n! n
(iA)n=
n=O

(1.85)

The argument of the functional A(x) is a classical object of the same nature
as the field <p(x), (.~ .• an ordinary function for bosons and an anticonum.iting
function for fermions. In the latter case only functions Gn with even n are
nonzero.
Using Eqs. (1.75) and (1.76), we obtain

All the notation is standard and needs no explanation.

Usually theN-product is defined such that ( Ol N [cp (x 1) ••• cp (xn) Jl 0)= 0
for any n :2: 1. Then for an arbitrary functional

(OINF(cp) lo) = F(O). (1.87)

Assuming this to be true, from (1.86) and (1.46) we obtain

G(A) = exp G. o~Ao~] exp i(Sv(<p) +<pA]'<P=O· (1.88)

By convention, in the fermionic case the derivatives in the differential

reduction operators are assumed to be right ones. This is convenient for
writing out the reduction formulas (1.45) and (1.46), which tum out to be
identical for the two statistics. But here we are planning to use Eqs.
(1.10)-(1.12), which requires transforming to left derivatives; we see from
( 1.9) that this leads only to a change of sign in the exponent of the reduction
operator. After this replacement, using (1.12) we find the functional (1.88) for
the free theory:
Exploring the Variety of Random
Documents with Different Content
sure, but it pointed very strongly to his guilt. You doubtless read the
result in the papers?”
“Yes, I followed the case,” she answered. “And I’m sure you did the
best you could.”
His solemnity would have been amusing at any other time. He
clearly had no idea that she had learned of his duplicity in taking
money from her for the defense of a Corrigan who was in no manner
related to her.
“I assume,” he said, “that no steps have yet been taken to offer for
probate any will Mr. Farley may have left. I had hoped to see you
first; this accounts for my visit to-day. I thought it best to see you
before going to Mr. Thurston. Mr. Joseph C. Thurston was, I believe,
Mr. Farley’s attorney?”
“Yes. He was one of papa’s best friends and he had charge of his
affairs as far back as I can remember.”
“An excellent man. There’s no better lawyer in the State,” Harlowe
responded heartily. “But I occasionally find it best to deal directly
with a client. We lawyers, you know, are sometimes unwisely
obstinate, and lead our clients into unnecessary trouble. As you are
the person chiefly concerned in this matter, I came directly to you. I
did this because in that former matter you were so quick to see the
justice of my—er—request.”
Her amazement at his effrontery almost equalled her curiosity as to
what lay behind his deliberate approaches.
“It is generally known that Mr. Farley was a man of violent temper,”
he went on. “Some of his old friends on the river remember him
well, and you may never have known—and I am sorry to be obliged
to mention so unpleasant a fact—that his mother died insane. That
is a matter of record, of course. The malady from which Mr. Farley
suffered for many years is one that frequently affects the mind. No
doubt living with him here, as you did, you noticed at times that he
behaved oddly—didn’t conduct himself quite normally?”
Remembering Eaton’s instructions she acquiesced without offering
any comment. His designs, she now assumed, were not personal to
herself, but directed against Farley’s estate.
“I represent two cousins of Mr. Farley’s who live in my county. Very
worthy men they are; you may have heard Mr. Farley speak of
them.”
“Yes; I knew about them. I sent them telegrams advising them of his
death.”
“That was very thoughtful on your part, Miss Farley, and they
appreciate it. But by reason of their poverty they were unable to
attend the funeral. They asked me to thank you for thinking of them.
Several times during the past twenty years Mr. Farley had advanced
them small sums of money—an indication of his kindly feeling
toward them.”
“I didn’t know of that; but it was like papa.”
“In case Mr. Farley left a will, it is my duty to inform you, that you
may have time for reflection before taking up the matter with Mr.
Thurston, that we are prepared to attack it on the ground of Mr.
Farley’s mental unsoundness. I assume, of course, that Mr. Farley
made a handsome provision for you, but quite possibly he
overlooked the natural expectations of his own kinsfolk.”
She merely nodded, thinking it unnecessary to impart information
while he continued to show his hand so openly.
“You have probably understood, Miss Farley, that in case your foster-
father died intestate, that is to say, without leaving a will in proper
form, you would, as his heir, be entitled to the whole of his
property.”
“Yes; I think I have heard that,” she answered uneasily.
The cold-blooded fashion in which he had stated his purpose to
contest the will on the ground of Farley’s insanity had shocked her.
Copeland had suggested the same thing, but it was a preposterous
pretension that Timothy Farley’s mind had been affected by his long
illness. Even the assertion that his mother had been a victim of
mental disorder, plausibly as he had stated it, would hardly stand
against the fact that Farley’s faculties to the very end had been
unusually clear and alert.
“In case there should be no will,” Harlowe continued, “your rights
would rest, of course, upon your adoption. It would have to be
proved that it was done in accordance with law. The statutes are
specific as to the requirements. I’m sorry, very sorry indeed, my dear
Miss Farley, that in your case the law was not strictly complied with.”
“I don’t know what you mean; I don’t understand you!” she faltered.
“Please don’t be alarmed,” he went on, with a reassuring smile. “I’m
sure that everything can be arranged satisfactorily; I am not here to
threaten you—please remember that; I merely want you to
understand my case.”
“But my father never dreamed of anything of that kind,” she gasped;
“it’s impossible—why, he would never have made a mistake in so
serious a matter.”
“Unfortunately, we are all liable to err, Miss Farley,” he answered,
with a grotesque affectation of benevolence. “And I regret to say
that in this case the error is undeniable. What Mr. Farley’s intentions
were is one thing; what was actually done to make you his child in
law is another. We need not go into that. It is a legal question that
Mr. Thurston will understand readily; the more so, perhaps,” he
added with faint irony, “because he was not himself guilty of the
error, not being Mr. Farley’s attorney at the time the adoption was
attempted.”
The room swayed and she grasped the arms of her chair to steady
herself. The man’s story was plausible, and he spoke with an easy
confidence. All Farley’s deliberation about the disposal of his
property would go for naught; her victory over the temptation to
destroy his wills had been futile!
“Please don’t misunderstand me, Miss Farley,” the man was saying.
“My clients have no wish to deprive you wholly of participation in the
estate. And we should deplore litigation. In coming to you now, I
merely wish to prepare you, so that you may consider the case in all
its aspects before taking it up with your lawyer. No doubt a
satisfactory settlement can be arranged, without going into court. I
believe that is all. Henceforth I can’t with propriety deal directly with
you, but must meet your counsel. I assume, however, that he will
not wholly ignore your natural wish to—er—arrange a settlement
satisfactory to all parties.”...
The door had hardly closed upon him before she was at the
telephone calling Eaton, and in half an hour he was at the house.
Harlowe’s words had so bitten into her memory that she was able to
repeat them almost verbatim. Eaton listened with his usual
composure. It might have seemed from his manner that he was
more interested in Nan herself than in her recital. She betrayed no
excitement, but described the interview colorlessly as though
speaking of matters that did not wholly concern her. When she
concluded Eaton chuckled softly.
“You’re taking it nobly,” were his first words; “I’m proud of you! You
see, I had expected something of the sort—prepared for it, in fact,
right after this fellow got that thousand dollars out of you. He’s
crafty, shrewd, unscrupulous. But you have nothing to worry over.
He came to you first and at the earliest possible moment in the hope
of frightening you as he did before, hoping that you’d persuade
Thurston to settle with him. As for Farley’s incompetence to make a
will, that’s all rubbish! His mother suffered from senile dementia—no
symptoms until she was nearly ninety. Every business man in town
would laugh at the idea that Tim Farley wasn’t sane. He was just a
little bit saner than most men. His occasional fits of anger were only
the expression of his vigorous personality; wholly characteristic;
nothing in that for Harlowe to hang a case on.
“But this point about the adoption is more serious. When I was
down there watching Harlowe defend the man he pretended to you
—but to nobody else—was your brother, I looked up those adoption
proceedings, out of sheer vulgar curiosity. The law provides that
adoption proceedings shall be brought in the county where the child
resides, and that the parents appear in court and consent. Your
parents were dead, and Mr. Farley’s petition was filed in this county
after you had been a member of his household for fully two years.
“I seriously debated mentioning these points to Thurston, after my
visit down there, but on reflection decided against it. Contrary to the
common assumption the law is not an ass—not altogether! I can’t
imagine the courts countenancing an effort to set aside this adoption
on so flimsy a pretext. Mr. Farley not only complied with the law to
the best of his belief, but let the world in general understand that he
looked on you as his child and heir.”
“That’s what every one believed, of course,” Nan murmured.
“I dare say there’s a will,” Eaton continued. “Thurston may have to
defend that—but you may rely on him. I have already made an
appointment to meet him at luncheon to turn over to him all my
data. I’ll say to you in all sincerity that I don’t see the slightest cause
for uneasiness. If there’s a valid will, that settles the adoption line of
attack, though this man may go the length of trying to annul it on
the insanity plea, merely to tie up the estate until you pay something
to these cousins to get rid of him.”
“There is a will; there are a number of them, I think,” said Nan
soberly.
“Mr. Farley told you about them—let you know what he was doing?”
“No; he never spoke of them, except in general terms. I used to see
him hiding them; once one dropped out of his dressing-gown.” She
hesitated; then added quickly: “I read that one before putting it
back. I know I shouldn’t have done it, but I did—as I’ve done a good
many things these last two years I shouldn’t!”
“Don’t be so hard on yourself! It was quite natural for you to look at
it.”
“The night he died,” she went on breathlessly, “he had been looking
at a number of wills he kept hidden in mamma’s old sewing-table. I
put them back in the drawer. I suppose Mr. Thurston will ask for
them when he comes.”
“Yes; he should see all such papers. You must tell him everything
you know that relates to them.”
“I almost burnt them all up last night,” she exclaimed in a strange,
hard tone. “That one I read made me angry. I thought it niggardly
and unjust. And—some one told me”—in her eagerness to make her
confession complete she nearly blurted out Copeland’s name—“that
if there should be no will I’d inherit everything. And last night I
fought that out. And it was a hard fight; it was horrible! But for once
in my life I got a grip on myself. You may remember saying to me,
‘Don’t wobble.’ Well, I wobbled till I was dizzy—but I wobbled right!
And now that that’s over, I believe—though I’m afraid to say it aloud
—that I’m a different sort of a girl some way. I hope so; I mean to
be very, very different.”
“You poor, dear, little Nan,” he said softly. “I’m proud of you—but not
very much surprised!”
“But you see it doesn’t count, anyhow,” she said, smiling, pleased
and touched by his praise. “If there’s a will, it’s bad; if there isn’t,
I’m not to be considered!”
“Don’t belittle your victory by measuring it against mere money. As
for those purely business matters, they’ll be attended to. You’re not
going to be thrown out on the world just yet.”
“I shouldn’t cry—not now—if it came to that! Now that I know what
they mean, I think I rather like these little wars that go on inside of
us. But I tell you it was good to see the daylight this morning and
know I could pass a mirror and not be afraid of my own face!”
“It is rather nicer that way; much nicer,” he said, with his rare smile.
“I’m glad you told me this. I see that I don’t need to worry about
you any more.”
“You haven’t really been doing that?”
“At times, at times, my dear Nan,” he said, looking at her quizzically,
“you’ve brought me to the verge of insomnia!”
 CHAPTER XX
THE COPELAND-FARLEY CELLAR

At twelve o’clock on the night of Nan’s prolonged struggle, Jerry,

having walked to the station with a traveling man of his
acquaintance, paused at the door of Copeland-Farley, hesitated a
moment, and then let himself in. He whistled a warning to the
watchman, as was his habit when making night visits to the
establishment. Hearing no response, he assumed that the man was
off on his rounds and would reach the lower floor shortly.
He opened his desk and busied himself with some memoranda he
had made from the books that afternoon. There was no denying that
the house was in a bad way; the one hundred thousand dollars of
notes carried by the Western National matured the next day, and in
addition to these obligations the Company was seriously behind in its
merchandise accounts.
A quarter of an hour passed, and the watchman made no sign. Jerry
closed his desk, walked back to the elevator-shaft, and shouted the
man’s name. From the dark recesses of the cellar came sounds as of
some one running, followed by a stumble and fall. He called again,
more loudly, but receiving no response, he ran to the stairway,
flashed on the lights, and hurried down.
His suspicions were aroused at once by a heap of refuse,
surmounted by half a dozen empty boxes, piled about the wooden
framework of the elevator-shaft.
The room where oils, paints, ethers, acids and other highly
inflammable or explosive stock was stored was shut off from the
remainder of the cellar by an iron door that had been pushed open.
As he darted in and turned on the lights, he heard some one
stealthily moving in the farther end of the room.
Seizing a fire-extinguisher he bawled the watchman’s name again
and plunged in among the barrels. A trail of straw indicated that the
same hand that had piled the combustibles against the shaft had
carried similar materials into the dangerous precincts of the oil room.
In a moment he came upon a barrel of benzine surrounded with
kindling.
He decided against calling for help. No harm had yet been done, and
it was best to capture the guilty person and deal with him quietly if
possible. He kicked the litter away from the barrel and waited. In a
moment a slight noise attracted his attention, and at the same
instant a shadow vanished behind an upright cask. He waited for the
shadow to reappear, advancing cautiously down the aisle with his
eyes on the cask.
“Come out o’ that!” he called.
A foot scraped on the cement floor and definitely marked the cask as
the incendiary’s hiding-place. He jumped upon a barrel, leaped from
it to the cask, and flung himself upon a man crouched behind it.
They went down together with Jerry’s hand clutching the captive’s
throat.
“Good God!” he gasped, as he found himself gazing into Copeland’s
eyes.
The breath had been knocked out of Billy and he lay still, panting
hard. His right hand clenched a revolver.
“Give me that thing!”
Jerry wrenched it from Copeland’s convulsive clutch, thrust it into his
coat pocket, and stood erect.
“I’m very sorry, sir,” he said.
“Damn’ near shootin’ you, Jerry,” drawled Copeland, sitting up and
passing his hand slowly across his face; “damn’ near! Gimme your
hand.”
Jerry drew him to his feet. Copeland rested heavily on the cask and
looked his employee over with a slow, bewildered stare.
“Might ’a’ known I couldn’t pull ’er off! Always some damn’ fool like
you buttin’ into my blizness. ’S my blizness! Goin’ do what I damn’
please with my blizness. Burn whole damn’ thing down ’f want to.
I’m incenjy—what you call ’m?—incenjyary,—what you call ’m—
pyromaniac. Go to jail and pen’tenshary firs’ thing I know.”
“Not this time,” said Jerry sternly. “I’m going to take you home.”
“Home? Whersh that?” asked Copeland, grinning foolishly.
“Well, I guess a Turkish bath would be better. Where’s Galloway?”
“Gall’way’s good fellow; reli’ble watchman. Wife’s sick; sent him
home with my comp’ments. Told ’im I’d take full reshponshibility.”
“You didn’t expect to collect the insurance on that story, did you?
You must have a low opinion of the adjusters. I’ll fire Galloway to-
morrow for leaving you here in this shape.”
“Not on yer life y’ won’t! Silly old man didn’t know I wuz loaded.
Came on me sud’ly—very sud’ly. Only had slix slocktails—no; thass
wrong; thass all wrong. You know what I mean. Effect unusual—
mos’ unusual. Just a few small drinks at club. Guess I can’t carry
liquor’s graceful-ly as I used to. Billy Copeland’s no good any more.
Want lie down. Good place on floor. Nice bed right here, Jerry.
Lemme go t’ sleep.”
He grasped the edge of the cask more firmly and bent his head to
look down at the heap of straw he had been planting round it when
Amidon interrupted him.
“Not much I won’t! But before we skip I’ve got to clean up this
trash. Steady, now; come along!”
He seized Copeland’s arm and forced him to the stairway, where he
left him huddled on the bottom step.
“No respec’ for head of house; no respec’ whatever,” Copeland
muttered.
Jerry bade him remain quiet, and began carrying the straw and
boxes back to the packing-room. He swept the floor clean, and when
he was satisfied that no telltale trace remained he got Copeland to
the counting-room and telephoned for a taxi.
“Goin’ to be busted to-morrow; clean smash. You made awful
mistake, Jeremiah, in not lessing—no, not lesting me burn ’er up.
Insurance’d help out consid’ble. Need new building, anyhow.”
“I guess we don’t need it that bad,” remarked Jerry, rolling a
cigarette. He called the police station and asked for the loan of an
officer to do watchman duty for the remainder of the night; and this
accomplished he considered his further duty to his befuddled
employer.
Now that the calamity had been averted, his anger abated.
Copeland’s condition mitigated somewhat the hideousness of the
crime he was about to commit. Only his desperate financial situation
could have prompted him to attempt to fire the building. Jerry’s
silence and unusual gravity seemed to trouble Copeland.
“Guess you’re dis’pointed in your boss, Jeremiah. Don’ blame you.
Drunken fool—damn’ fool—incenjy-ary; no end bad lot.”
“Put your hat on straight and forget it,” remarked Jerry.
He telephoned to Gaylord, an athletic trainer who conducted a
Turkish bath, and told him to prepare for a customer. He knew
Gaylord well, and when they reached his place Jerry bade him stew
the gin out of Copeland and be sure to have him ready for business
in the morning. While Copeland was in the bath, Jerry tried all the
apparatus in the gymnasium and relieved his feelings by putting on
the gloves with Gaylord’s assistant. After all the arts of the
establishment had been exercised upon Copeland and he was
disposed of for the night, Jerry went to bed....
In the morning Gaylord put the finishing touches on his patient and
turned him out as good as new. It had occurred to Amidon that
Copeland might decide to avoid the store that day. He was relieved
when he announced, after they had shared Gaylord’s breakfast, that
he would walk to the office with him.
“Guess I’ll give the boys a jar by showing up early,” he remarked.
It was a clear, bracing morning, and Copeland set a brisk pace. He
was stubbornly silent and made no reference to the night’s affair
until they reached the heart of the city. Then he stopped suddenly
and laid his hand on Jerry’s arm.
“Jerry, I never meant to do that; for God’s sake, don’t believe I did!”
he broke out hoarsely. “I was troubled about the business, and some
other things had worried me lately. I took too many drinks—and I’d
never meant to drink again! I wouldn’t have tried that sober—I
wouldn’t have had the nerve!”
“It was the drink, of course,” Jerry assented. “It’s all over now. You’d
better forget it; I’m going to!”
“I wish to God I could forget it!”
Copeland shrugged his shoulders impatiently, then drew himself
erect and walked on more quickly. Jerry cheerfully changed the
subject, and when they were near the store dived into an alley that
led to the rear door of Copeland-Farley to avoid appearing before
the clerks in Copeland’s company.
Copeland remained in his room all morning, summoning the auditor
from time to time to ask for various data. He called Jerry once and
bade him make every effort to find Kinney by telephone. Kinney was
in New York; had been there for a week. Copeland smiled
sardonically at this news.
“All right. I knew he’d been away, but the fool said he’d be back to-
day,” he said spitefully. “That’s all!”
At two o’clock he put a bundle of papers into his pocket and walked
toward the Western National. The bookkeepers exchanged
meaningful glances and Jerry imagined that even the truckmen
loading freight appeared depressed. Copeland’s desperation had
been expressed vividly enough in his drunken attempt to burn the
store. And now, if the Western National refused to extend his loans,
Copeland-Farley might cease to exist. Jerry’s usual nonchalance left
him. He failed to seize a chance to “land” on a drummer from a New
York perfumery house who was teasing him for the latest news of
Main Street....
At three o’clock Eaton called Jerry on the telephone.
“I want to see Copeland; please call me the minute he comes in,”
said the lawyer.
Shortly before four Copeland came back and walked directly to his
office. There was another exchange of glances along the
accountants’ desks, where the clerks bent with affected diligence
over their books.
The auditor was summoned again, carried a book into Copeland’s
room, and reappeared instantly. The air was tense. It was a source
of relief to Jerry to hear Eaton’s voice as he reported Copeland’s
return.
“Watch him,” said the lawyer, with his usual calmness; “and don’t let
him leave the store.”
As Jerry nervously watched the door for Eaton’s appearance, Louis
M. Eichberg, of Corbin & Eichberg, entered and asked for Copeland.
The bookkeepers exchanged glances again and bent over their
ledgers with renewed zeal. The door of the private office closed
upon Eichberg. It snapped shut sharply—ominously, Jerry thought.
 CHAPTER XXI
A SOLVENT HOUSE

“I’ve bought in your stock,” Eichberg was saying to Copeland. “You

put up fourteen hundred and eighty-five shares with the Western
National and I’ve bought ’em in at private sale under your collateral
agreement. As I understand it there are fifteen shares held by
employees to qualify as directors. I guess there won’t be any trouble
about them, and we’ll let ’em stand for the present.”
“Those men paid for their stock and you have no right to touch it,”
said Copeland. “The stock in this company has an actual value of
two hundred dollars a share—”
“Rubbish! Your capital’s shrunk till you can’t see it any more.”
“Don’t you believe it! The house was never as sound as it is to-day. I
hope you don’t think I’m going to stand by and let the Western sell
me out on a small loan in this high-handed fashion! It’s a frame-up,
a conspiracy to clean me out. I’ve still got a majority of the stock,
and I’ll give you a run for your money before you get through with
me!”
“Keep your temper, Copeland! I don’t like doing this, but it’s better
for me to have the business than to let it peter out, the way it’s
doing. I’ll even say that after we consolidate I’ll be glad to make a
place for you in the house.”
“Oh, you needn’t trouble!” returned Copeland hotly. “You’re not
going to get rid of me so easy!”
“All right! Just how much stock do you think you’ve got?” asked
Eichberg with a faint ironic smile.
“I’ve got fifteen hundred shares; the bank understood that when I
refused their demand for a majority,” Copeland replied, frowning
over the stock-ledger.
“That shows how much you know about your own business! There’s
twenty shares out of your half that I’ve been trying to lay my hands
on for two months. It was a deal Farley made the last year he was
down here with a Fort Wayne jobber named Reynolds that he
bought out after your father died. I know because we tried to buy
up Reynolds ourselves, but old Uncle Tim went us one better. There
wasn’t much to the business, but the good-will was worth something
and Farley let Reynolds have twenty shares just to beat us out of the
sale. Farley had sense! When Reynolds died his executor sold the
stock to somebody here. Foreman handled it, but he won’t tell me
who he sold to. I know you didn’t get it! Foreman says he spent a
month last summer lookin’ for you to give you a chance to buy the
stock, but he couldn’t get hold of you. You were always off sportin’
with Kinney!”
Copeland had forgotten about the Reynolds shares. He mentally
cursed Farley for not reminding him of them; Farley had never dealt
squarely with him! Very likely he had personally told Eichberg and
the Western National of the Reynolds shares. It was galling to be
obliged to learn from Eichberg things he should have known himself.
He had flattered himself that in persuading the bank to accept
fourteen hundred and eighty-five shares as collateral instead of the
majority for which demand had been made at first, he had shown
his business sagacity; but evidently Eichberg had known of the
Reynolds shares all along.
“I don’t intend that what’s left of this business shall go to the bad,”
said Eichberg. “Either you come to terms, and let ’em know outside
that we’ve arranged a merger in a friendly way, or I’ll call up my
lawyer and tell him to apply for a receiver.”
Outside, the interested and anxious clerks and stenographers, cold
with excitement, watched their associate, Mr. Jeremiah Amidon, who
was inviting the wrath of the gods by knocking upon Copeland’s
door. When he entered in response to an angry bellow, they
expected to see him reappear instantly, possibly at the end of
William B. Copeland’s foot. To their chagrin Amidon remained in the
private office for some time; and they judged from the sudden quiet
that followed his disappearance that he was exerting a calming
influence upon Copeland and his visitor....
“I beg your pardon,” Jerry remarked while Copeland and Eichberg
glared at him.
To Copeland the sight of Jerry was an unwelcome reminder of the
previous night. His remorse over his effort to burn the store
vanished; if it hadn’t been for this meddlesome cub he wouldn’t now
be entertaining Eichberg in his office!
“Well, what does the boy want?” demanded Eichberg, when
Copeland found it impossible to express his wrath at Jerry’s
intrusion.
Eichberg knew Jerry perfectly well; everybody in the street knew
Jerry! And it was the basest insult to refer to him as the boy.
“Excuse me, Mr. Eichberg! I just wanted to hand a memorandum to
Mr. Copeland.”
He drew from his pocket the certificate he had purchased from
Foreman, and handed it to Copeland, who snatched it from him with
an angry snarl.
“Where did you get this?” he asked faintly after a glance at the
paper.
“Oh, it just blew in my way early in the fall. I never bothered to get
a new certificate, but I’ll turn it in right now.”
He pulled out a fountain pen, removed the cap deliberately, and
wrote his name in the blank space above the executor’s
endorsement. This done, he brushed an imaginary speck from his
cuff, as he had seen Eaton do, and went out, closing the door softly.
“Well, here’s the answer, Eichberg,” said Copeland, with affected
nonchalance; “here are those Reynolds shares.”
“How did that damn’ little fool get this?” demanded Eichberg, after a
careful scrutiny of the certificate and endorsements.
“Oh, he’s a useful little damn’ fool! He’s always picking up
something,” replied Copeland coolly.
“I suppose it was all set up,” Eichberg sneered. “Why didn’t you
come right out and say you had that stock, and save my time? It’s
worth something if yours ain’t! You’ll either sell me that stock or I’ll
have the court throw you out. It’s up to you!”
“I told you the truth about these shares,” said Copeland, whose
good humor was returning. “I’m ashamed to say I’d clean forgotten
them; but you see stock never figured much in our corporation; it’s
always been a sort of family affair. I have no idea where Amidon got
Reynolds’s shares—that’s straight! He’s always doing something he
isn’t paid for. And you see it isn’t quite so easy to clean me out. But I
take off my hat to you; you’re a business man!”
Hope had risen in him. In spite of his futile efforts to tide over the
crisis there was still the remote chance that Kinney, who always
seemed able to borrow all he wanted for his own purposes, might
extend a helping hand. His change of manner had its effect on
Eichberg.
“The stock doesn’t cut any ice,” he fumed. “I’m not goin’ to have a
hundred thousand dollars in a concern that’s losin’ money like this
one! That statement you showed the bank was rotten! You ain’t got
any credit; and you know mighty well you can’t go on here. You’ll
either come to terms or I’ll get a receiver to-morrow. That’s all there
is of that!”
He clapped on his hat and turned to the door just as it opened upon
Eaton.
“I’ll look in again in the morning, Copeland,” said Eichberg in a loud
tone. “You just think over that matter, and I guess you’ll see it my
way.”
“Never put off till to-morrow what you can do to-day,” remarked
Eaton, projecting himself into the office. “I’ll close the door if you
don’t mind, Copeland. And, Mr. Eichberg, please wait a moment.”
“If you’re his lawyer, you don’t want me here. I’ve said all I’ve got to
say to Copeland,” Eichberg answered. But he waited, glowering at
Eaton, who removed his overcoat, placed it carefully on a chair, and
began drawing off his gloves.
“Mr. Eichberg, they told me a moment ago at the Western National
that certain stock held as collateral for maturing Copeland-Farley
notes had been bought by you. Is that true?”
“That’s correct! I guess it was all regular,” Eichberg snapped.
“We’ll come to that presently. You have now in your possession
through that purchase fourteen hundred and eighty-five shares of
stock?”
“Right!” ejaculated Eichberg loudly.
Eaton raised his hand, glanced intently at the palm, and then, with
one of his familiar tricks, bent his gaze directly upon Eichberg.
“Being a competitor of Copeland-Farley and a director of the bank,
you have naturally—quite naturally—thought it would be a good
investment to own a large block of the stock? And it undoubtedly
occurred to you that a combination of Copeland-Farley with Corbin &
Eichberg would be highly advantageous? In fact, you thought you
had more stock than Copeland owns, and that you could come in
here and discharge him like a drayman!”
“That’s my business! You haven’t explained yet how you come to be
buttin’ in here.”
“Presently—presently!” replied Eaton soothingly.
His calm demeanor and refusal to lift his voice further infuriated
Eichberg, who breathed hard for a moment, then pointed a stubby
forefinger at the lawyer as his wrath found utterance.
“Copeland-Farley’s ruined—busted! If you’ll take a look at their last
statement you’ll see they can’t pull out!”
“You anticipate me,” replied Eaton gently. “The fact is I had meant to
buy that stock myself, but the bank’s haste to turn it over to you has
spoiled that. I was annoyed—greatly annoyed—when I found awhile
ago that the stock had been sold—sold, in violation of the stipulation
—on the bank’s usual form—that three days’ grace were to be given
to the debtor to release his collateral. I don’t believe the Comptroller
would like that. I shall consider seriously bringing it to his attention.”
“What good would three days have done him?” cried Eichberg. “The
sooner he’s put out the better. His accounts payable are goin’ to
bring his general creditors down on him in a few days! Don’t you
suppose I know? Haven’t they been telegraphin’ me from all over
the country for months askin’ about this house?”
“And, of course,” said Eaton softly, “you did all you could to protect
your competitor—neighborly feeling, and that sort of thing. Well, it
will be a great relief to you to know that those accounts will be paid
to-morrow—just as soon as the exchange window of your piratical
bank is opened. There’s a hundred thousand dollars to the credit of
Copeland-Farley over there right now. I know, because I went in a
quarter of an hour ago and made the deposit. This house is solvent
—absolutely solvent. Moreover, Copeland’s stock in the Kinney Ivory
Cement Company is now marketable. I take some pride in that fact
myself—immodestly, I dare say, and yet—I am only human!”
He drew a telegram from his waistcoat pocket and handed it to
Copeland.
“That patent case was decided to-day—in favor of Kinney. Copeland,
I congratulate you!”
Copeland read the message, and looked dully from Eaton to
Eichberg. He was roused by Eichberg, who had no difficulty in
expressing his emotions.
“You fool,” he shouted, shaking his fist in Eaton’s face. “If you’re
tellin’ the truth, what do you mean to do about my stock?”
Eaton was drawing on his gloves without haste. His face expressed
the mildest surprise at Eichberg’s perturbation.
“My dear Mr. Eichberg, you were in such a rush to buy the Western’s
collateral that I’m surprised that you should trouble me—a casual
acquaintance—with such a question.”
“It’s a cheat; it’s a swindle! If there’s any law for this—”
He flung out of the office and tramped heavily to the front door,
while the clerks, worn with the many agitations of the day, stared
after him mutely.
“In the morning,” Eaton was saying to Copeland, “I’ll have fuller
details of the decision, but there’s no doubt about it—we’ve won on
every point. Allow me to congratulate you!”
Copeland half rose to take his proffered hand; then with a groan he
sank back and buried his face in his hands.
 CHAPTER XXII
NULL AND VOID

“Those documents have a familiar look,” remarked Thurston with a

smile as Nan placed the packet of wills on the table beside him in
the Farley parlor. “Mr. Farley was hard to please; I’ve learned a lot
about will-writing just from studying the different schemes he
proposed from time to time.”
Nan described the manner in which she had found the wills on the
night of Farley’s death.
“He was evidently troubled about them and got out of bed to look
them over. This one, that I found lying open on the table, is torn
across as though he had begun to destroy it when the end came.”
“Very likely that was his intention,” Thurston replied. “I had just
written a new will for him, but it wasn’t signed—not unless he
executed it that same afternoon. Perhaps you know about that?”
“No one was here, I’m sure,” said Nan, after a moment’s
consideration. “The nurse was off duty; she left for the evening at
four o’clock, and I’m sure the servants weren’t in his room. I carried
up his dinner tray myself.”
“It’s hardly possible he had signed that last will. I was always
present on such occasions and I got the witnesses. When I called
now and then with a couple of his friends, or telephoned for them,
there was a will to be signed. You probably understood that.”
He began opening the papers, glancing quickly at the last sheet of
each will, and turning them face down on the table. The torn one he
scrutinized more carefully, and returned to it for further examination
when he had disposed of the others. Nan watched him nervously. He
was a small, slight man of sixty, with a stiff gray mustache and a
sharp, rasping voice. It would not have been easy to deceive
Thurston if she had destroyed the wills; she could never have gone
through with it!
She felt that she had touched with her finger-tips the far horizons
and knew at last something of the meaning of life. She had
subjected herself to pitiless self-analysis and stood convicted in her
own conscience of vanity, selfishness, and hardness. The recollection
of her gay adventures with the Kinneys and her affair with Copeland
had become a hideous nightmare. Not only was she ashamed of her
dallying with Billy, but she accused herself of having exerted a
baneful influence upon him. In all likelihood he would never have
sunk so low as to propose the destruction of Farley’s will but for his
infatuation for her.
Farley’s death had in itself exercised a chastening effect upon her.
She was conscious of trying to see herself with his eyes and fortify
herself with something of the stern righteousness that made him, in
the retrospect, a noble and inspiring figure. The upturned faces at
the Settlement haunted her; there was a work for her to do in the
world if only she could lay her hands upon it! In this new mood the
life of ease which money would secure weighed little against self-
dependence and service. Money had ceased to be an important
integer in her calculations.
Having concluded his examination of the papers, the lawyer lifted his
head with an impatient jerk, then sighed, and began smoothing the
open sheets into a neat pile.
“Those wills are worthless, Miss Farley,—not one of them can be
probated. The testator’s signatures and the names of the witnesses
have been scratched out!”
In proof of his statement he extended one of the wills, pointing to
the heavy cross-crosses at the bottom of the sheet.
“You have no idea when he did this—you weren’t present, I
suppose?”
“No; he used to do his writing at the table where he hid the wills. He
occasionally wrote a letter or a check there; but I never saw him
open the table. I never knew of that inner compartment till the night
he died.”
“Oh, I know that table very well; he had shown me the hidden
drawer and explained how to open it. But this is most unfortunate,
deplorable! I kept in touch with his doctor about his condition and
feared something like this might happen. And he dreaded it himself
—was afraid he might die some time without leaving just the will he
had determined to make. I account for all the wills I wrote for him
but the last. The last time I was here I brought a new will, which I
don’t find among these. Are you sure you haven’t overlooked it?”
She was quite sure of it, but after she had described in minute detail
the events of the last afternoon of Farley’s life, to confirm her
statement that no one who could have acted as witness had visited
Farley, she took the lawyer upstairs to examine the table for himself.
They broadened the scope of the search, but without success.
“For the present I think it best for you not to read those wills,” he
said, when they had returned to the parlor. “They represent Mr.
Farley’s changes of feeling in regard to many things—including
yourself. A little later I shall be glad to submit them to you. The
important thing just now is the threat of this man Harlowe to attack
your rights under the adoption. Mr. Eaton and I have already
discussed that. Now that we’re pretty sure there’s no will, this may
give us some trouble, but with characteristic thoroughness Mr. Eaton
has prepared for just this emergency. His reasons for not telling me
earlier about these things are sound enough—his fear of disturbing
Mr. Farley unnecessarily. He would undoubtedly have wanted a
proceeding brought to correct the adoption, but that could only have
advertised the error, and Mr. Farley might have died before we
finished it. Still, if I had known I should have taken care that he
didn’t die intestate. But from what Mr. Eaton tells me, this man is all
primed to attack any will that might have been left, on the ground of
Mr. Farley’s mental incapacity—which is ludicrous, of course. There
was never a saner man; and yet his eccentricities might be
magnified before a jury—you never can tell. On the whole, Mr.
Eaton’s silence was justified. But our next step must be carefully
considered. In the mean time—”
He paced the floor, considering means of relieving her anxiety.
“Of course, while these things are pending we shall arrange for your
maintenance, on the old basis, in this house. No one can pretend
that Mr. Farley didn’t have every intention of providing for you
generously. It’s only fair to tell you this, that even when he seemed
to waver at times he never cut your legacy below a hundred
thousand dollars; and I know he regretted the comparative
meagerness of that—tripled the amount in the very next will he
made! You need have no fears, Miss Farley,” he went on reassuringly.
“But you are entitled to your own counsel; it’s only right that I
should say this to you immediately; and I suggest that you ask Mr.
Eaton to represent you. I hope you will confer with him at once.”
He bowed with old-fashioned formality. He was more troubled than
he cared to have Nan know, and her silence disconcerted him. But
her face expressed neither disappointment nor alarm. She stood
erect by the table, an intent look in her eyes. Not wishing to leave
her weighed down by the uncertainties of her future, he said briskly:
—
“You mustn’t bother yourself about these matters, Miss Farley. In the
end you will find yourself a rich woman. So—”
He waved his hand as the preliminary to a quick exit, but she called
him back. He did not like being called back; now, he thought, there
would be the tears he dreaded.
“You don’t understand,” she said quietly. “I ought to have made it
clear in the first place, but I didn’t know just how—or when—to say
it. I can’t—I will not take any of Mr. Farley’s money—not even if the
law should give it to me.”
He looked at her with the mute appeal of the deaf when they fail to
catch a meaning.
“Really, Miss Farley—”
“I won’t take one cent of Mr. Farley’s money,” Nan repeated firmly.
“I can’t blame you for being disappointed—for resenting what may
appear to be a lack of consideration on his part for your comfort—”
“Oh, it isn’t that! I wouldn’t have you think that! I’m sure he meant
to do what was right—what was generous! You don’t know how glad
I am that our last day together was a happy one—we had never
been on better terms. It’s not that I have any unkind feeling toward
papa; it’s all myself. The Farleys were only too kind to me. I went my
own way and it made me selfish—and pretty hard, too, I’m afraid.
Papa knew it; and you know yourself how little he trusted me. And
he was right about me: I didn’t deserve his confidence. But I’m
going to begin all over again, as I couldn’t if I began fighting for this
money. I can see now that money can’t make me happy. I’m going
to work; I’m going to stop living, as I always have, just for myself:
I’m going—I’m going to think about the rest of the folks a lot!”
“The folks?” repeated Thurston feebly. “What folks?”
“Oh, everybody! The down-and-outers—girls like me who get a bad
start or make mistakes!”
Thurston’s brows worked convulsively. He had been prepared for
anything but this.
“Do I—do I understand you to mean that, even if this estate could
be turned over to you to-morrow, you’d decline to receive it? It can’t
be possible—”
“Yes; that’s what I mean!” she cried eagerly. “I’ve thought it all out
and have made up my mind about it. I don’t want to be considered
in anything that has to do with papa’s property.”
“But, my dear child, you can’t—you can’t abandon your claims in any
such fashion! It’s my duty—I owe it to my friend and client to see
that his wishes are fulfilled. Why—”
“Well,” she persisted, “between all those wills you can’t tell what he
wanted—only that I was a great problem to him. I caused him a
great deal of unnecessary worry and heartache. I hope this isn’t
going to cause you any trouble—” And she smiled in spite of herself
at his consternation, as indicated by the twitching of his brows. And
there was, she realized, something absurd to her cool statement to a
hard-headed lawyer that she renounced claims whose validity he
was in duty bound to support. The situation was too much for him;
he must escape as quickly as possible from this young woman who
brushed away a fairly tangible fortune as a waiter clears away bread
crumbs.
“Really, Miss Farley—” he began; but, thinking of nothing further to
say, he backed awkwardly into the hall.
She helped him into his coat and opened the street door. He hurried
off without saying good-bye, clasping Timothy Farley’s wills tightly
under his arm.
A light snow was falling; Nan stood on the steps and lifted her hot
face to the fluttering flakes. She watched Thurston until he turned
the corner and then went to the telephone.
In a moment she was connected with Mrs. Copeland at the farm. “I
want a job,” she was saying in a cheerful tone; “yes, that’s it—a
chance to work. You told me the other day you needed some one to
look after your business at the market-house. I’m applying for the
job. Oh, no! I’m not fooling; I want that place! Well, I want to see
you, too; I’ll be out early in the morning!”
 CHAPTER XXIII
IN TRUST

“Copeland Farm Products” in blue letters against a white background

swung over Nan’s head on Tuesday, Thursday, and Saturday
mornings in the city market-house. On those days she left Mrs.
Copeland’s farm at five o’clock with the day’s offerings and by six the
stand was in order.
An endless, jostling throng surged by, and every sale she effected,
every negotiation for the future delivery of an order, had all the joy
of an adventure. Her immediate neighbors were a big-fisted German
gardener and a black-eyed Italian girl who sold fruits and
vegetables. When business lagged, the German chaffed her about
her wares or condoled with her when some frugal marketer priced
her butter, sniffed, and departed. Nan commanded a meager
knowledge of Italian and flung a phrase at her dark-eyed neighbor
now and then in the spirit of comradeship which the place
encouraged. She liked her “job.” She assured herself that she had
never had so much fun in all her life, and that never again would she
eat the bread of idleness.
But it had not proved so easy as she imagined it would be to slip out
of her old life into the new. If she had left the Farley house preceded
by a brass band and had marched round the monument and the
length of Washington Street before taking her place in the market,
her flight could hardly have attracted more attention.
The town buzzed. The newspapers neglected no phase of Nan’s
affairs, nor did they overlook her as she stood behind the counter
dispensing “Copeland Farm Products.” She was surprised and vexed
by her sudden notoriety. A newspaper photographer snapped her, in
her white sweater and blue-and-white tam o’shanter, passing eggs
over the counter. The portrait bore the caption, “Miss Nancy Farley in
a New Rôle,” and was supplemented by text adorned with such sub-
headings as “Renounces her Fortune” and “Throws Away a Million
Dollars.” To be thus heralded was preposterous; she had merely
gone to work for reasons that were, in any view of the matter, her
own private affair. But public sentiment was astonishingly friendly;
even those who had looked askance at her high flights with the
Kinney crowd said it was an outrage that Farley had failed to provide
for her decently.
Fanny, thinking at first it was only a joke, a flare of temperament
(references to her temperament had begun to pall upon Nan!), had
welcomed Nan to her house and given her charge of the market-
stand; but it was not without difficulty that she persuaded the girl to
occupy her guest-room and share her meals.
“You’d better scold me when I make mistakes, for if I find I don’t
suit I’ll fire myself,” Nan declared. “And if I have to leave you, I’ll go
to clerking in a department store. I just mention this so you won’t be
too polite. This isn’t any grandstand play, you see; I’m serious for
the first time in my life!”
It was certain, at any rate, that Copeland Farm Products were sold
with amazing ease. When it became known that Nan Farley had
become Mrs. Copeland’s representative “on market,” there was lively
competition for the privilege of purchasing those same “products.”
Fanny complained ruefully that the jellies, jams, and pickles created
by the young women in her industrial house would be exhausted
before Christmas and that nothing would remain to sell but butter
and eggs. Nan suggested orange marmalade and a cake-baking
department to keep the girls at work during the winter, and on the
off days she set herself to planning the preparation of these
“specialties.” Mrs. Farley’s cooking lessons had not gone for naught;
Nan could bake a cake in which there was no trace of “sadness,” and
after some experiments with jumbles and sand-tarts she sold her
first output in an hour and opened a waiting list.
Mrs. Copeland told Eaton at the end of the second week that she
had never known the real Nan till now. There was no questioning
the girl’s sincerity; she had cut loose from her old life, relinquished
all hope of participating in Farley’s fortune, and addressed herself
zealously to the business of supporting herself. She became
immediately the idol of the half-dozen young women in the old
farmhouse, who thought her an immensely “romantic” figure and
marveled at her industry and resourcefulness.
“Splendid! Give her all the room she wants,” Eaton urged Mrs.
Copeland. “She’s only finding herself; we’ll have the Nan she was
meant to be the first thing we know.”
“I didn’t know all these nice church-going people would come to
condole with me, or I’d have left town,” Nan confided to Fanny.
“These women who wouldn’t let their daughters associate with me a
year ago can’t buy enough eggs now to show how much they
sympathize with me. If they don’t keep away, I’m going to raise the
price of their eggs, and that will break their hearts—and the eggs!
But do you know,” she went on gravely, “I’ve never been so happy in
my life as I am now! And I wouldn’t have anybody think it was out
of pique, or with any unkind feeling toward papa,”—tears shone in
her eyes as the word slipped from her tongue,—“but I tell you
nobody ever could have made a nice, polite girl out of me. I was
bound to get into scrapes as long as I hadn’t anything really to do
but fill in time between manicuring and hair-washing dates. There’s
a whole lot in that old saying about making a silk purse out of a
sow’s ear: it can’t be did!”
“If you talk that way,” Fanny laughed, “I shall turn you out of my
house. I don’t want you to think I approve of what you’re doing. I’m
letting you do it because I’m scared not to!”
“You’d better be—for if you hadn’t taken me in, I should have gone
on the stage,—honestly, I should,—in vaudeville, most likely, doing
monologues right between the jugglers and the trained seals.”...
On Tuesdays, Thursdays, and Saturdays Mr. Jeremiah A. Amidon
found it convenient to visit the market-house as early as seven-thirty
(in spite of pressing duties at the store), to make sure, he said, that
Nan, and the farmhand who drove her in and helped arrange the
stock, had safely passed all the railroad crossings on the way to
town. Jerry was a consoling person and unobtrusively thoughtful and
helpful. And in his way he was almost as keen as Eaton. Jerry did
not require explanations, and nothing is so wholly satisfactory as a
friend who understands without being told.
“Little girl, if your eggs are guaranteed under the Pure Food Act, I’ll
take one—the large size.”
“You’ll find the hard-boiled eggs at the lunch counter in the next
aisle, little boy,” Nan answered. “How is John Cecil?”
“Working himself to death. You’ve driven him to it!”
“I hope you two are not abusing me; how about it?”
“No; not vocally. Cecil’s shut up in his office every night, getting
ready to clean up those cousins of Farley’s down on the river, but he
doesn’t say anything. Look here, Nan, we’ve got a line of cold cream
and other toilet marvels—stuff you could handle here as a side line.
Let us send you up a bunch to put next to that pink jelly. It’s high
grade and we’ll make it to you at the right price.”
“Not on your life, Jerry. Drugs and hand-made country butter can’t
associate. You’d better run down to your own little shop now and go
to work.”
After his morning inspection he was likely to reappear at lunch time,
to see her for a moment before she left for the farm; and he
assisted in balancing her cash when she confessed that it wouldn’t
“gee.” His pride in her was enormous; he was satisfied that there
was no other girl to compare with her.
Jerry’s admiration was so obviously genuine and supported by so
deep an awe and reverence that no girl could have helped liking it.
And Jerry was unfailingly amusing; his airs and graces, his attempts