Download the full version of the ebook at

https://ebookgate.com

Statistical Mechanics of Nonequilibrium

Liquids Cambridge 2008 2nd Edition Denis J.
Evans

https://ebookgate.com/product/statistical-
mechanics-of-nonequilibrium-liquids-
cambridge-2008-2nd-edition-denis-j-evans/

Explore and download more ebook at https://ebookgate.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Statistical mechanics of membranes and surfaces 2nd

Edition David Nelson

https://ebookgate.com/product/statistical-mechanics-of-membranes-and-
surfaces-2nd-edition-david-nelson/

ebookgate.com

Nonequilibrium Statistical Physics Linear Irreversible

Processes Oxford Graduate Texts 1st Edition Pottier

https://ebookgate.com/product/nonequilibrium-statistical-physics-
linear-irreversible-processes-oxford-graduate-texts-1st-edition-
pottier/
ebookgate.com

Thermodynamics and Statistical Mechanics An Integrated

Approach 1st Edition Robert J. Hardy

https://ebookgate.com/product/thermodynamics-and-statistical-
mechanics-an-integrated-approach-1st-edition-robert-j-hardy/

ebookgate.com

Nonequilibrium Statistical Physics Linear Irreversible

Processes Oxford Graduate Texts Noëlle Pottier

https://ebookgate.com/product/nonequilibrium-statistical-physics-
linear-irreversible-processes-oxford-graduate-texts-noelle-pottier/

ebookgate.com
Thermodynamic Formalism The Mathematical Structure of
Equilibrium Statistical Mechanics 2nd Edition David Ruelle

https://ebookgate.com/product/thermodynamic-formalism-the-
mathematical-structure-of-equilibrium-statistical-mechanics-2nd-
edition-david-ruelle/
ebookgate.com

An introduction to thermodynamics and statistical

mechanics 2nd Edition Stowe K.

https://ebookgate.com/product/an-introduction-to-thermodynamics-and-
statistical-mechanics-2nd-edition-stowe-k/

ebookgate.com

Quantum statistical mechanics 1st Edition William C.

Schieve

https://ebookgate.com/product/quantum-statistical-mechanics-1st-
edition-william-c-schieve/

ebookgate.com

Fundamentals of classical statistical thermodynamics

dissipation relaxation and fluctuation theorems 1st
Edition Evans
https://ebookgate.com/product/fundamentals-of-classical-statistical-
thermodynamics-dissipation-relaxation-and-fluctuation-theorems-1st-
edition-evans/
ebookgate.com

Thatcher and Thatcherism The Making of the Contemporary

World 2nd Edition Eric J. Evans

https://ebookgate.com/product/thatcher-and-thatcherism-the-making-of-
the-contemporary-world-2nd-edition-eric-j-evans/

ebookgate.com
This page intentionally left blank
 S TAT I S T I C A L ME C H A N I C S O F
N O N E Q U I L I B R I U M LI Q U I D S
SECOND EDITION

In recent years the interaction between dynamical systems theory and nonequilibrium
statistical mechanics has been enormous. The discovery of fluctuation theorems as a
fundamental structure common to almost all nonequilibrium systems, and the connec-
tions with the free-energy calculation methods of Jarzynski and Crooks, have excited
both theorists and experimentalists. This book charts the development and theoretical
analysis of molecular dynamics as applied to equilibrium and nonequilibrium
systems.
Substantially updated and revised, this book is designed both for experts in the
field and beginning graduate students of physics. It connects molecular-dynamics
simulation with mathematical theory to understand nonequilibrium steady states.
It also provides a link between the atomic, nano, and macro worlds, showing
how these length scales relate. The book ends with an introduction to the use of
nonequilibrium statistical mechanics to justify a thermodynamic treatment of none-
quilibrium steady states, and gives a direction to further avenues of exploration.

DENIS J. EVANS is Professor of Theoretical Chemistry at the Australian National

University (ANU), Dean of the Research School of Chemistry and Convenor of
the ANU College of Science. He has won several prizes, including the Moyal
Medal from Macquarie University for distinguished contributions to mathematics,
physics or statistics, the Centenary Medal from the Australian Government, and the
H. G. Smith Memorial Medal from the Royal Australian Chemical Institute.

GARY MORRISS is Associate Professor and Undergraduate Director in the School of

Physics at the University of New South Wales, Australia. He is a Fellow of the
Institute of Physics and a member of the American Physical Society. His research
areas include nonequilibrium statistical mechanics and dynamical systems.
S TATI S T I C A L M E C H A N I C S O F
N O N E Q U I L I B R I U M LI Q U I D S

SECOND EDITION

D E N I S J. E VA N S
Research School of Chemistry, Australian National University,
Canberra

G A RY MO R R I S S
School of Physics, University of New South Wales, Sydney
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521857918

© D. Evans and G. Morris 2008

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

relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
First published in print format 2008

ISBN-13 978-0-511-39864-3 eBook (EBL)

ISBN-13 978-0-521-85791-8 hardback

Cambridge University Press has no responsibility for the persistence or accuracy of urls
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
 Contents

Preface to the second edition page ix

Preface to the first edition xi

1 Introduction 1

2 Linear irreversible thermodynamics 11

2.1 The conservation equations 11
2.2 Entropy production 17
2.3 Curie’s theorem 20
2.4 Non-markovian constitutive relations: viscoelasticity 27

3 The microscopic connection 33

3.1 Classical mechanics 33
3.2 Phase space 44
3.3 Distribution functions and the liouville equation 46
3.4 Ergodicity, mixing, and Lyapunov exponents 53
3.5 Equilibrium time-correlation functions 59
3.6 Operator identities 62
3.7 The Irving-Kirkwood procedure 66
3.8 Instantaneous microscopic representation of fluxes 68
3.9 Microscopic representation of the temperature 77

4 The Green–kubo relations 79

4.1 The Langevin equation 79
4.2 Mori–Zwanzig theory 83
4.3 Shear viscosity 87
4.4 Green–Kubo relations for Navier–Stokes transport coefficients 92

5 Linear-response theory 95
5.1 Adiabatic linear response theory 95
5.2 Thermostats and equilibrium distribution functions 100

v
vi Contents

5.3 Isothermal linear response theory 111

5.4 The equivalence of thermostatted linear responses 116

6 Computer simulation algorithms 119

6.1 Introduction 119
6.2 Self diffusion 125
6.3 Couette flow and shear viscosity 130
6.4 Thermostatting shear flows 142
6.5 Elongational flows 146
6.6 Thermal conductivity 150
6.7 Norton ensemble methods 152
6.8 Constant-pressure ensembles 156
6.9 Constant stress ensembles 160

7 Nonlinear response theory 167

7.1 Kubo’s form for the nonlinear response 167
7.2 Kawasaki distribution function 169
7.3 The transient time-correlation function formalism 173
7.4 Trajectory mappings 177
7.5 Differential response functions 185
7.6 The van Kampen objection to linear response theory 193
7.7 Time-dependent response theory 200

8 Dynamical stability 209

8.1 Introduction 209
8.2 Chaotic dynamical systems 211
8.3 The characterization of chaos 221
8.4 Chaos in planar couette flow 230
8.5 Conjugate pairing of Lyapunov exponents 238
8.6 Periodic orbit measures 244
8.7 Positivity of transport coefficients 255

9 Nonequilibrium fluctuations 259

9.1 Introduction 259
9.2 The specific heat 260
9.3 The compressibility and isobaric specific heat 265
9.4 The fluctuation theorem 267
9.5 Gallavotti and Cohen fluctuation theorem 274
9.6 The Jarzynski equality 277
 Contents vii

9.7 The Crooks relation 280

9.8 Experimental verification 282

10 Thermodynamics of steady states 283

10.1 The thermodynamic temperature 283
10.2 Green’s expansion for the entropy 292
10.3 Prospects 299

References 301

Index 309
 Preface to the second edition

Since 1990, when the first edition appeared, there has been a significant advance in
the development of nonequilibrium systems. The centerpiece of the first edition
was the nonequilibrium molecular-dynamics methods and their theoretical analysis,
the connections between linear and nonlinear response theory, and the design of the
simulation methods. This is now a mature field with only one significant addition,
which is the new method for elongational flows.
Chapter 10 in the first edition was called “Towards a thermodynamics of steady
states.” This contained an introduction to deterministic chaotic systems. The
second edition has the same title for Chapter 10, but the contents are now comple-
tely different. The application of the ideas of modern dynamical-systems theory to
nonequilibrium systems has grown enormously with all of Chapter 8 devoted to
this. However, this still constitutes the barest of introductions with whole books
(Gaspard, 1998; Dorfman, 1999; Ott, 2002; and Sprott, 2003) devoted to this
theme. The theoretical advances in this area are some of the biggest. The develop-
ment of methods to study the time evolution using periodic orbits, and the use of
periodic orbits to develop SRB measures for nonequilibrium systems are exciting
steps forward.
Based on the dynamical properties, Lyapunov exponents in particular, there have
been great strides made in the development of the study of fluctuations in nonequi-
librium systems. The fluctuation theorems, and methods for calculating free-energy
differences using nonequilibrium paths, have dominated conferences for the last
6–7 years. The additional fact that these can be tested in real (rather than computer)
experiments and used to measure free-energy differences in the unfolding of
biological molecules will have a large impact.
Thanks are due to many people. Customarily we thank our wives, and remark
that they are still the same! To those that have led the development of statistical
mechanics and inspired and mentored those that followed we owe a great debt:
Eddie Cohen, Bob Dorfman, Siegfried Hess, Christian Gruber, and many others.

ix
x Preface to the second edition

A number of people were especially generous with their time for which we are very
grateful, in particular, Peter Daivis, Carl Dettmann, Tooru Taniguchi, and Debra
Bernhardt (Searles). Special thanks also to Billy Todd, Tom Hunt, Lamberto
Rondoni, Chris Angstmann, David Kruss, Anthony Whelan, Dean Robinson,
David Monaghan and Tony Chung. I am very pleased to acknowledge the contri-
bution of Ian Watson, who as an undergraduate student read and learnt from this
book, in the process discovering (and helping to correct) many of its faults. As
ever, Eddie Cohen was a frustrating inspiration to us all!
I feel the pull of the white ship at the Grey Havens in the long firth of Lune.
 Preface to the first edition

During the 1980s there have been many new developments regarding the
nonequilibrium statistical mechanics of dense classical systems. These develop-
ments have had a major impact on the computer simulation methods used to
model nonequilibrium fluids. Some of these new algorithms are discussed in the
recent book by Allen and Tildesley (1987), Computer Simulation of Liquids.
However, that book was never intended to provide a detailed statistical mechanical
backdrop to the new computer algorithms. As the authors commented in their
preface, their main purpose was to provide a working knowledge of computer
simulation techniques. The present volume is, in part, an attempt to provide a ped-
agogical discussion of the statistical mechanical environment of these algorithms.
There is a symbiotic relationship between nonequilibrium statistical mechanics
on the one hand and the theory and practice of computer simulation on the other.
Sometimes, the initiative for progress has been with the pragmatic requirements
of computer simulation and at other times, the initiative has been with the funda-
mental theory of nonequilibrium processes. Although progress has been rapid,
the number of participants who have been involved in the exposition and develop-
ment, rather than with application, has been relatively small.
The formal theory is often illustrated with examples involving shear flow in
liquids. Since a central theme of this volume is the nonlinear response of
systems, this book could be described as a text on theoretical rheology. However
our choice of rheology as a test-bed for theory is merely a reflection of personal
interest. The statistical mechanical theory that is outlined in this book is capable
of far wider application.
All but two pages of this book are concerned with atomic rather than molecular
fluids. This restriction is one of economy. The main purpose of this text is best
served by choosing simple applications.
Many people deserve thanks for their help in developing and writing this book.
Firstly we must thank our wives, Val and Jan, for putting up with our absences, our

xi
xii Preface to the first edition

irritability, and our exhaustion. We would also like to thank Dr. David MacGowan
for reading sections of the manuscript. Thanks must also go to Mrs. Marie
Lawrence for help with indexing. Finally special thanks must go to Professors
Cohen, Hanley, and Hoover for incessant argument and interest.
 1
Introduction

Mechanics provides a complete microscopic description of the state of a system.

When the equations of motion are combined with initial conditions and boundary
conditions, the subsequent time evolution of a classical system can be predicted. In
systems with more than just a few degrees of freedom such an exercise is imposs-
ible. There is simply no practical way of measuring the initial microscopic state of,
for example, a glass of water, at some instant in time. In any case, even if this was
possible we could not then solve the equations of motion for a coupled system of
1023 molecules.
In spite of our inability to fully describe the microstate of a glass of water, we are
all aware of useful macroscopic descriptions for such systems. Thermodynamics
provides a theoretical framework for correlating the equilibrium properties of
such systems. If the system is not at equilibrium, fluid mechanics is capable of pre-
dicting the macroscopic nonequilibrium behaviour of the system. In order for these
macroscopic approaches to be useful, their laws must be supplemented, not only
with a specification of the appropriate boundary conditions, but with the values
of thermophysical constants such as equation-of-state data and transport coeffi-
cients. These values cannot be predicted by macroscopic theory. Historically this
data has been supplied by experiments. One of the tasks of statistical mechanics
is to predict these parameters from knowledge of the interactions of the system’s
constituent molecules. This then is a major purpose for statistical mechanics.
How well have we progressed?
Equilibrium classical statistical mechanics is relatively well developed. The
basic ground rules – Gibbsian ensemble theory – have been known for the best
part of a century (Gibbs, 1902). The development of electronic computers in the
1950s provided unambiguous tests of the theory of simple liquids leading to a con-
sequently rapid development of integral equation and perturbation treatments of
liquids (Barker and Henderson, 1976). With the possible exceptions of phase equi-
libria and interfacial phenomena (Rowlinson and Widom, 1982) one could say that

1
2 Introduction

the equilibrium statistical mechanics of atomic fluids is a solved problem. Much of

the emphasis has moved to molecular, even macromolecular, liquids.
The nonequilibrium statistical mechanics of dilute atomic gases – kinetic
theory – is, likewise, essentially complete (Ferziger and Kaper, 1972). However,
attempts to extend kinetic theory to higher densities have been fraught with
severe difficulties. One might have imagined being able to develop a power-
series expansion of the transport coefficients in much the same way that one
expands the equilibrium equation of state in the virial series. Dorfman and
Cohen (1965; 1972) proved that such an expansion does not exist. The Navier–
Stokes transport coefficients are nonanalytic functions of density.
It was at about this time that computer simulations began to have an impact on
the field. In a celebrated paper, Kubo (1957) showed that linear transport coeffi-
cients could be calculated from a knowledge of the equilibrium fluctuations in
the flux associated with the particular transport coefficient. For example the
shear viscosity η, is defined as the ratio of the shear stress, −Pxy, to the shear
rate, ∂ux /∂y ; γ:

Pxy ;
hg: ð1:1Þ

The Kubo relation predicts that the limiting, small shear rate, viscosity, is given by:
ð1
h ¼ bV ds kPxy ð0ÞPxy ðsÞl, ð1:2Þ
0

where b is the reciprocal of the absolute temperature T, multiplied by Boltzmann’s

constant kB, V is the system volume and the angle brackets denote an equilibrium
ensemble average. The viscosity is then the infinite time integral of the equilibrium,
autocorrelation function of the shear stress. Similar relations are valid for the other
Navier–Stokes transport coefficients such as the self diffusion coefficient, the
thermal conductivity, and the bulk viscosity (see Chapter 4).
Alder and Wainwright (1956) were the first to use computer simulations to
compute the transport coefficients of atomic fluids. What they found was unex-
pected. It was believed that at sufficiently long time, equilibrium autocorrelation
functions should decay exponentially. Alder and Wainwright discovered that in
two-dimensional systems, the velocity autocorrelation function which determines
the self-diffusion coefficient, only decays as t −1. Since the diffusion coefficient is
thought to be the integral of this function, we were forced to the reluctant con-
clusion that the self diffusion coefficient does not exist for two-dimensional
systems. It is presently believed that each of the Navier–Stokes transport
coefficients diverge in two dimensions (Pomeau and Resibois, 1975).
 Introduction 3

This does not mean that two-dimensional fluids are infinitely resistant to shear
flow. Rather, it means that the Newtonian constitutive relation Equation (1.1) is
an inappropriate definition of viscosity in two dimensions. There is no linear
regime close to equilibrium where Newton’s law (Equation 1.1), is valid. It is
thought that at small strain rates, Pxy % γlogγ. If this were the case then the limiting
value of the shear viscosity limγ→0 −∂Pxy/∂γ would be infinite. All this presupposes
that steady laminar shear flow is stable in two dimensions. This would be an
entirely natural presumption on the basis of our three-dimensional experience.
However there is some evidence that even this assumption may be wrong (Evans
and Morriss, 1983b). Recent computer simulation data suggests that in two dimen-
sions, laminar flow may be unstable at small strain rates.
In three dimensions the situation is better. The Navier–Stokes transport coeffi-
cients appear to exist. However the nonlinear Burnett coefficients, higher-order
terms in the Taylor series expansion of the shear stress in powers of the strain
rate (Section 2.3, Section 9.5), are thought to diverge (Kawasaki and Gunton,
1973). These divergences are sometimes summarized in Dorfman’s Lemma
(Zwanzig, 1982): all relevant fluxes are nonanalytic functions of all relevant vari-
ables! The transport coefficients are thought to be nonanalytic functions of density,
frequency, and the magnitude of the driving thermodynamic force, the strain rate, or
the temperature gradient etc.
In this book we will discuss the framework of nonequilibrium statistical mech-
anics. We will not discuss in detail, the practical results that have been obtained.
Rather we seek to derive a nonequilibrium analog of the Gibbsian basis for equili-
brium statistical mechanics. At equilibrium we have a number of idealizations
which serve as standard models for experimental systems. Among these are the
well-known microcanonical, canonical, and grand canonical ensembles. The real
system of interest will not correspond exactly to any one particular ensemble,
but such models furnish useful and reliable information about the experimental
system. We have become so accustomed to mapping each real experiment
onto its nearest Gibbsian ensemble that we sometimes forget that the canonical
ensemble, for example, does not exist in Nature. It is an idealization.
A nonequilibrium system can be modeled as a perturbed equilibrium ensemble;
we will therefore need to add the perturbing field to the statistical mechanical
description. The perturbing field does work on the system – this prevents the
system from relaxing to equilibrium. This work is converted to heat, and the
heat must be removed in order to obtain a well-defined steady state. Therefore ther-
mostats will also need to be included in our statistical mechanical models. A major
theme of this book is the development of a set of idealized nonequilibrium systems
which can play the same role in nonequilibrium statistical mechanics as the
Gibbsian ensembles play at equilibrium.
4 Introduction

After a brief discussion of linear irreversible thermodynamics in Chapter 2, we

address the Liouville equation in Chapter 3. The Liouville equation is the funda-
mental vehicle of nonequilibrium statistical mechanics. We introduce its formal sol-
ution using mathematical operators called propagators (Section 3.3). In Chapter 3,
we also outline the procedures by which we identify statistical mechanical
expressions for the basic field variables of hydrodynamics.
After this background in both macroscopic and microscopic theory we go on to
derive the Green–Kubo relations for linear transport coefficients in Chapter 4 and
the basic results of linear response theory in Chapter 5. The Green–Kubo relations
derived in Chapter 4 relate thermal transport coefficients, such as the Navier–
Stokes transport coefficients, to equilibrium fluctuations. Thermal transport
processes are driven by boundary conditions. The expressions derived in Chapter 5
relate mechanical transport coefficients to equilibrium fluctuations. A mechanical
transport process is one that is driven by a perturbing external field which actually
changes the mechanical equations of motion for the system. In Chapter 5 we show
how the thermostatted linear mechanical response of many body systems is related
to equilibrium fluctuations.
In Chapter 6 we exploit similarities in the fluctuation formulae for the mechani-
cal and the thermal response, by deriving computer simulation algorithms for cal-
culating the linear Navier–Stokes transport coefficients. Although the algorithms
are designed to calculate linear thermal-transport coefficients, they employ mech-
anical methods. The validity of these algorithms is proved using thermostatted
linear-response theory (Chapter 5) and the knowledge of the Green–Kubo relations
provided in Chapter 4.
A diagrammatic summary of some of the common algorithms used to compute
shear viscosity is given in Figure 1.1. The Green–Kubo method simply consists of
simulating an equilibrium fluid under periodic boundary conditions and making the
appropriate analysis of the time-dependent stress fluctuations using Equation (1.2).
Gosling et al. (1973) proposed performing a nonequilibrium simulation of a system
subject to a sinusoidal transverse force. Monitoring the field-induced velocity
profile and extrapolating the results to infinite wavelength, the viscosity can be cal-
culated. Hoover and Ashurst (1975), used external reservoirs of particles to induce
a nearly planar shear in a model fluid. In the reservoir technique, the viscosity is
calculated by measuring the average ratio of the shear stress to the strain rate, in
the bulk of the fluid, away from the reservoir regions. The presence of the reservoir
regions gives rise to significant inhomogeneities in the thermodynamic properties
of the fluid and in the strain rate in particular. This leads to obvious difficulties
in the calculation of the shear viscosity. Lees and Edwards (1972), showed that
if one used “sliding brick” periodic boundary conditions, one could induce
planar Couette flow in a simulation. The so-called Lees–Edwards periodic
 Introduction 5

Green–Kubo Sinusoidal transverse force

Momentum reservoirs Homogeneous shear

Figure 1.1 Methods of determining the shear viscosity

boundary conditions enable one to perform homogeneous simulations of shear flow

in which the low Reynolds-number velocity profile is linear.
With the exception of the Green–Kubo method, these simulation methods all
involve nonequilibrium simulations. The Green–Kubo technique is useful in that
all linear transport coefficients can, in principle, be calculated from a single simu-
lation. It is restricted though, to only calculating linear transport coefficients. The
nonequilibrium methods, on the other hand, provide information about the non-
linear as well as the linear response of systems. They therefore provide a direct
link with rheology.
The use of nonequilibrium computer simulation algorithms, so-called
nonequilibrium molecular dynamics (NEMD), leads inevitably to the question of
the large field, nonlinear response. Indeed the calculation of linear transport coeffi-
cients using NEMD proceeds by calculating the nonlinear response and extrapolat-
ing the results to zero field. One of our main aims will be to derive a number of
nonlinear generalizations of the Kubo relations which give an exact framework
within which one can calculate and characterize transport processes far from equi-
librium (Chapter 7). Because of the divergences alluded to above, the nonlinear
theory cannot rely on power-series expansions about the equilibrium state. A
major system of interest is the nonequilibrium steady state. Theory enables us to
relate the nonlinear transport coefficients and mechanical quantities, like the
6 Introduction

internal energy or the pressure, to transient fluctuations in the thermodynamic flux

which generates the nonequilibrium steady state (Chapter 7). We derive the transi-
ent time correlation function (TTCF, Section 7.3) and the Kawasaki representations
(Section 7.2) of the thermostatted nonlinear response. These results are exact and
do not require the nonlinear response to be an analytic function of the perturbing
fields. The theory also enables one to calculate specific heats, thermal-expansion
coefficients and compressibilities from knowledge of steady-state fluctuations
(Chapter 9). After we have discussed the nonlinear response, we present a resol-
ution of the van Kampen objection to linear response theory and to the Kubo
relations in Chapter 7.
An innovation in our theory is the use of reversible equations of motion which
incorporate a deterministic thermostat (Section 3.1). This innovation was motivated
by the needs imposed by nonequilibrium computer simulation. If one wants to use
any of the nonequilibrium methods depicted in Figure 1.1 to calculate the shear vis-
cosity, one needs a thermostat to achieve a reliable steady-state average. It is not
clear how to calculate the viscosity of a fluid whose temperature and pressure
are increasing in time.
The first deterministic thermostat, the so-called Gaussian thermostat, was
independently and simultaneously developed by Hoover and Evans (Hoover et al.,
1982) and Evans (1983a). It permitted homogeneous simulations of nonequilibrium
steady states using molecular-dynamics techniques. Hitherto molecular dynamics
had involved solving Newton’s equations for systems of interacting particles. As
work was performed on such a system in order to drive it away from equilibrium,
the system inevitably heated with the irreversible conversion of work into heat.
Hoover and Evans showed that if such a system evolved under their
thermostatted equations of motion, the so-called Gaussian isokinetic equations of
motion, the dissipative heat could be removed by a thermostatting force which
was part of the equations of motion themselves. Now, computer simulators had
been simulating nonequilibrium steady states for some years, but in the past the
dissipative heat was removed by simple ad hoc rescaling of the second moment
of the appropriate velocity. The significance of the Gaussian isokinetic equations
of motion was that since the thermostatting was part of the equations of motion
it could be analyzed theoretically using response theory. Earlier ad hoc rescaling
or Andersen’s stochastic thermostat (Andersen, 1980), could not be so easily ana-
lyzed. In Chapter 5 we prove that while the adiabatic (i.e. unthermostatted) linear
response of a system can be calculated as the integral of an unthermostatted
(i.e. Newtonian) equilibrium time-correlation function, the thermostatted linear
response is related to the corresponding thermostatted equilibrium time-correlation
function. These results are quite new and can be proved only because the
thermostatting mechanism is reversible and deterministic.
Another random document with
no related content on Scribd:
The Project Gutenberg eBook of The viaduct
murder
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: The viaduct murder

Author: Ronald Arbuthnott Knox

Release date: January 1, 2024 [eBook #72585]

Most recently updated: February 4, 2024

Language: English

Original publication: New York, NY: Jacobsen Publishing Company,

Inc, 1926

Credits: Brian Raiter

*** START OF THE PROJECT GUTENBERG EBOOK THE VIADUCT

MURDER ***
THE VIADUCT MURDER

by

RONALD A. KNOX
DEDICATED BY COMMAND

TO

TONY WILSON
 “In the matter of information, above all regard with suspicion that
which seems probable. Begin always by believing what seems
incredible.”
Gaboriau, Monsieur Lecoq.
 Contents
I. The Paston Oatvile Dormy-house
II. In the Rough
III. Piecing it Together
IV. Endless Clues
V. On the Railway
VI. The Movements of Mr. Davenant
VII. Carmichael’s Account of it
VIII. The Inquest, and a Fresh Clue
IX. The Animated Picture
X. In which a Book is more communicative than a Lady
XI. A Funeral and a Vigil
XII. A Search with Piano Accompaniment
XIII. The Man in the Passage
XIV. A Chase, ending with a Surprise
XV. Gordon takes the Opportunity to Philosophize
XVI. Reeves promises to do his Best
XVII. By which Train?
XVIII. The Holmes Method
XIX. Mordaunt Reeves talks to Himself
XX. Proof at Last
XXI. The Test
XXII. In the Fog
XXIII. Marryatt breaks the Pledge
XXIV. Gordon offers the Consolation of Philosophy
XXV. The Dull Facts
 Chapter I.
The Paston Oatvile Dormy-house
Nothing is ever wasted. The death of the animal fertilizes the
vegetable world: bees swarm in the disused pillar-box; sooner or
later, somebody will find a use for the munition-factories. And the
old country-seats of feudal England, that bask among their figured
terraces, frowning at the ignoble tourist down secular avenues and
thrusting back the high-road he travels by into respectful detours—
these too, although the family have long since decided that it is too
expensive to live there, and the agents smile at the idea of letting
them like one humouring a child, have their place in the hero-
tenanted England of to-day. The house itself may be condemned to
the scrap-heap, but you can always make a golf-course out of the
Park. Acres, that for centuries have scorned the weight of the
plough, have their stubborn glebe broken with the niblick, and over-
populated greens recall the softness and the trimness of earlier
lawns. Ghosts of an earlier day will walk there, perhaps, but you can
always play through them.
Paston Oatvile (distrust the author whose second paragraph does
not come to ground in the particular) seemed to have been specially
adapted by an inscrutable Providence for such a niche in the scheme
of things. The huge Italianate building which the fifteenth Lord
Oatvile raised as a monument to his greatness (he sold judiciously
early out of the South Sea Company) took fire in the nineties of last
century and burned for a whole night; the help given by the local
fire brigade was energetic rather than considerate, and Achelous
completed the havoc which Vulcan had begun. It stands even now,
an indecent skeleton, papered rooms and carved mantelpieces
confronting you shamefacedly, like the inside of a doll’s house whose
curtain-wall has swung back on the hinge. What secrets that ball-
room, those powder-closets must have witnessed in the days of an
earlier gallantry, when the stuccoed façade still performed its
discreet office! Poor rooms, they will never know any more secrets
now. The garden, too, became involved in the contagion of decay:
weeds have overgrown its paved walks, and neglected balustrades
have crumbled; a few of the hardier flowers still spring there, but
half-smothered in rank grass, shabby-genteel survivors of an ancien
régime. For the family never attempted to rebuild; they prudently
retired to the old Manor at the other end of the park, a little brick
and timber paradise which had served the family for a century and a
half as dower-house. In time, even this reduced splendour was
judged too expensive, and the family sold.
No need, then, to mourn for Paston Oatvile; the sanctities of its
manorial soil will be as interminable as golf. An enterprising club,
seconded by an accommodating railway, has invested its rural
solitude with an air of suburbanity; it is only an hour’s journey from
London, and the distance could be covered in three-quarters of the
time if the club were less exclusive. Bungalows, each fitted with its
own garage, and cottages that contain billiard-rooms have sprung up
in the neighbourhood; thirty or forty of these, all rough-cast and red
tiles, conceal by a series of ingenious dissimilarities their
indebtedness to the brain of a single architect. In the middle of
these—the cathedral, the town hall, the market-place around which
all their activities centre—stands the dower-house of the Oatviles,
the dormy-house of to-day. The committee have built on to it largely
in what is understood to be the same style, and indeed, the new
part is undeniably brick and timber, though in wet weather the
timber is apt to warp and fall off. It is not only a club-house, of
course, it is also an expensive hotel—if we may call it an hotel, and
not rather a monastic settlement; for the inhabitants of these
pleasant rooms all live for one end—golf: twice daily they go round
the course, with all the leisurely solemnity of Benedictines reciting
their office, and every night they meet in corona to discuss the
mysteries of their religion.
Which reminds me that I have forgotten to mention the village
Church. There is still a village, that straggles mysteriously, like so
many English villages, in the form of a hollow square. In the old
days, the Church interposed itself between the village and the Great
House, a kind of mercy-seat through which the squire could be
appeased upon occasion. Though much older than the Park or the
fortunes of the Oatvile family, it had acquired, from its enclosed
position, the air of a parasitical institution, an undergrowth of
Protestant feudalism. To-day, it somehow strikes the eye as a by-
product of the golfing industry; people who ask the way to it (and
they are rare) are directed to the fifteenth green; the service on
Sunday is at half-past nine, so as to allow for the improbable chance
of anybody wanting to fortify himself for the morning round by
divine worship; the sexton will caddy for you except on the
afternoon of a funeral. Conformably with this, the incumbent of the
parish, who is to figure in this story, was a golfing parson presented
by an absentee squire to a living which offered few material
attractions. He had managed to let the parsonage, which was more
than twenty minutes’ walk from the first tee, and lived in the dormy-
house permanently; arguing, not without reason, that it was the
centre of all the life there was in the parish. If you are disposed to
take a look at him, you have only to open the smoking-room door;
there he sits, this October afternoon of rain and fog, with three
equally weatherbound companions, a foursome in potentia.
He was a man now approaching middle age, a bachelor and
unambitious. You would say that he had a clerical face—is that
clerical face a mark of predestination, or does it develop by natural
mimicry?—but the enthusiasm which it registered was, it is to be
feared, principally directed towards one object, and that object a
game. He was mild-mannered, and had been known to keep his
temper successfully in the most trying circumstances, even at the
ninth; no oath was ever heard to escape his lips, though his
invariable phrase, “What tam I doing?” was held by some to have a
relish of perdition in it. The other three were acquaintances of his,
as acquaintance goes at Paston Oatvile, where you know
everybody’s handicap, nobody’s politics or religion. One of them,
indeed, Alexander Gordon in nature and in name, could hardly be
known otherwise than by his handicap, for in politics, in religion, in
every subject that could form a digression from the normal
conversation of the dormy-house, his point of view was entirely
undistinguished and British to the last degree. He was not, like the
others, a permanent inmate, but was on a holiday visit to his more
interesting friend, Mordaunt Reeves.
Reeves was a permanent inmate, more by force of circumstances
than from any natural indolence. He had left school at the beginning
of the War, and had been incapacitated for active service by an
extreme short-sightedness which gave his face a penetrating, not to
say a peering, look. Work had been found for him easily enough in
an outlying department of the War Office, and he was perhaps a
little too fond of beginning his sentences with, “When I was in the
Military Intelligence.” The picture which the words conjured up to
the uninitiated was that of Mordaunt Reeves concealed behind the
arras with a revolver at half-cock, overhearing the confidential
discussions of German super-spies. Actually, his business had been
to stroll into a very uncomfortable office at half-past nine in the
morning, where a docket of newspaper cuttings, forwarded from
another department, awaited him. Singling out some particularly
fire-eating utterance of a Glasgow shop-steward, he would have it
typed out and put in a jacket; then he would scrawl across it: “Can
something be done about this? Please initial”—and so the document
would be caught up in that vast maelstrom of unregarded jackets
that circulated aimlessly through the sub-departments of Whitehall.
An orphan, with a comfortable income, he had found himself unable
to settle down to ordinary employment on the outbreak of peace. He
had put several romantic advertisements into the daily papers,
indicating his readiness to undertake any mysterious commissions
that might call for the services of an “active, intelligent young man,
with a turn for the adventurous”: but the supply of amateur
adventurers was at the time well ahead of the demand, and there
was no response. In despair, he had betaken himself to Paston
Oatvile, and even his ill-wishers admitted that his game was
improving.
That Mr. Carmichael, the fourth member of the party, had been a
don you knew as soon as he opened his mouth. There was that
precision in his utterances, that benignity in his eye, that spontaneity
in his willingness to impart information, that no other profession
breeds. A perpetual fountain of interesting small-talk, he unnerved
his audience with a sense of intellectual repletion which was worse
than boredom. Not that he talked the “shop” of the learned: his
subject had been Greek archæology; his talk was of county families,
of travels in the Near East, of the processes by which fountain-pens
are manufactured, of county families again. He was over sixty—he,
alone of the party, was married, and lived in one of the bungalows
with a colourless wife, who seemed to have been withered by long
exposure to the sirocco of his conversation: at the moment she was
absent, and he was lodging in the dormy-house like the rest. It must
be confessed that his fellow-members shunned him, but he was
useful upon occasion as a last court of appeal on any matter of fact;
it was he who could remember what year it was the bull got loose
on the links, and what ball the Open Championship was won with
three years back.
Marryatt (that was the clergyman, yes; I see you are a proper
reader for a detective-story) rose once more and took a good look at
the weather. The fog was lifting, but the rain still fell pitilessly. “I
shouldn’t wonder,” he said, “if we had some showers before
nightfall.”
“It’s a curious thing,” said Carmichael, “that the early Basque
poets always speak of the night not as falling but as rising. I
suppose they had a right to look at it that way. Now, for myself
———”
Marryatt, fortunately, knew him well enough to interrupt him. “It’s
the sort of afternoon,” he said darkly, “on which one wants to
murder somebody, just to relieve one’s feelings.”
“You would be wrong,” said Reeves. “Think of the footmarks you’d
be bound to leave behind you in mud like this. You would be caught
in no time.”
 “Ah, you’ve been reading The Mystery of the Green Thumb. But
tell me, how many murderers have really been discovered by their
footprints? The bootmakers have conspired to make the human race
believe that there are only about half a dozen different sizes of feet,
and we all have to cram ourselves into horrible boots of one uniform
pattern, imported by the gross from America. What does Holmes do
next?”
“Well, you see,” put in Gordon, “the detectives in the book always
have the luck. The murderer generally has a wooden leg, and that
doesn’t take much tracing. The trouble in real life is the way
murderers go about unamputated. And then there’s the left-handed
men, how conveniently they come in! I tried detection once on an
old pipe, and I could show you from the way the side of it was
charred that the owner of it was right-handed. But there are so
many right-handed people.”
“In most cases,” said Carmichael, “it’s only nerves that make
people think they’re left-handed. A more extraordinary thing is the
matter of parting the hair. Everybody is predestined from birth to
part his hair on one particular side; but most of the people who
ought to part their hair on the right do it on the left instead, because
that’s easier when you’re right-handed.”
“I think you’re wrong in principle, Gordon,” said Reeves.
“Everybody in the world has his little peculiarities, which would give
him away to the eye of a trained detective. You, for example, are the
most normal specimen, if I may say so, of the human race. Yet I
know which of those whisky glasses on the mantelpiece is yours,
though they’re empty.”
“Well, which?” asked Gordon, interested.
“The one in the middle,” said Reeves. “It’s pushed farther away
from the edge: you, like the careful soul you are, instinctively took
more precaution against its being brushed off. Aren’t I right?”
“To tell the truth, I can’t for the life of me remember. But there,
you see, you’re talking of somebody you know. None of us are
murderers, at least, I hope not. If you were trying to detect a
murderer you’d never been introduced to, you wouldn’t know what
to look out for.”
 “Try it on,” suggested Marryatt. “You know, the Holmes stunt,
deducing things from the bowler hat, and from Watson’s brother’s
watch. Try that umbrella over there, whatever it’s doing here; what
will you deduce from it?”
“I should deduce that it had been raining recently,” put in Gordon
with great seriousness.
“As a matter of fact,” said Reeves, turning the umbrella this way
and that, “an umbrella’s a very difficult thing to get any clues out of.”
“I’m glad of that,” said Carmichael, “because——”
“—only this one,” continued Reeves, ignoring him, “happens to be
rather interesting. Anybody could see that it’s pretty new, yet the
ferrule at the end of it is nearly worn through, which shows it’s been
used a lot. Inference; that it’s used by somebody who doesn’t keep
his umbrella for days like this, but uses it as a walking-stick.
Therefore it belongs to old Brotherhood; he’s the only man I know in
this club who always carries one.”
“You see,” said Carmichael, “that’s the sort of thing that happens
in real life. As I was just going to say, I brought in that umbrella
myself. I took it by mistake from a complete stranger in the Tube.”
Mordaunt Reeves laughed a little sourly. “Well,” he said, “the
principle holds, anyhow. Everything tells a story, if you are careful
not to theorize beyond your data.”
“I’m afraid,” said Gordon, “I must be one of Nature’s Watsons. I
prefer to leave things where they lie, and let people tell me the
story.”
“There you are wrong,” protested Reeves. “People can never tell
you a story without putting their own colour upon it—that is the
difficulty of getting evidence in real life. There, I grant you, the
detective stories are unreal: they always represent witnesses as
giving the facts with complete accuracy, and in language of the
author’s own choosing. Somebody bursts into the room, and says,
‘The body of a well-dressed man in middle-life has been found four
yards away from the north end of the shrubbery. There are marks of
violence about the person of the deceased’—just like a reporter’s
account of an inquest. But in real life he would say, ‘Good God! A
man’s shot himself on the lawn’—leaping at once, you see, from
observation to inference.”
“Journalism,” explained Carmichael, “makes havoc of all our
detective stories. What is journalism? It is the effort to make all the
facts of life correspond, whether they will or no, to about two
hundred ready-made phrases. Head-lines are especially destructive—
you will have noticed for yourselves how the modern head-line
aspires to be a series of nouns, with no other parts of speech in
attendance. I mean, the phrase, ‘She went into the garden to cut a
cabbage-leaf to make an apple pie’ becomes ‘Apple-pie fraud
cabbage-leaf hunt,’ and ‘What; no soap! So he died’ becomes ‘Soap-
shortage fatality sequel.’ Under this treatment, all the nuances of
atmosphere and of motive disappear; we figure the truth by trying
to make it fit into a formula.”
“I agree with you about inference,” said Marryatt, disregarding
Carmichael’s last remark—one always did disregard Carmichael’s last
remark. “But think how much of one’s knowledge of other people is
really inference. What do we really know about one another down
here? Fellow-passengers on the stream of life, that’s all we are. Take
old Brotherhood, whom you were mentioning just now. We know
that he has some sort of business in London, but we’ve no idea
what. We know that he comes down here every night in the week
from Monday onwards, and then from Saturday to Monday he
disappears—how do you know what he does with himself during the
week-ends? Or take young Davenant down at the Hatcheries; he
turns up there every Saturday evening, and does his two rounds on
the Sunday, and then on Monday he’s off again into the Ewigkeit.
What do we really know about him?”
“I should have thought you knew all you wanted to about
Brotherhood,” chuckled Reeves. “Hasn’t he taken to disproving the
existence of God on Wednesday evenings on the village green?”
Marryatt flushed slightly. “After all,” he said, “what does that
amount to? You might as well say I know Davenant’s a Roman
Catholic. But all I know is that once in a way he goes over on a
Sunday to Paston Bridge—the priest there knows something about
him, I suppose, but he wouldn’t tell you.”
 “I had a very extraordinary experience once,” said Carmichael, “in
Albania. I had to translate the confession of a dying man into French
for the sake of a priest who didn’t know the local language. The
priest told me afterwards I was bound not to disclose to anybody
what I’d heard.”
“He didn’t know you, anyhow, Carmichael,” suggested Reeves.
“As a matter of fact, I’ve never mentioned what he said to
anybody, though it was sufficiently curious.”
“It’s impossible,” resumed Reeves, “not to make inferences; the
mistake is to depend on them. In ordinary life, you have to take
risks; you have to sit down in the barber’s chair, although you know
it is just as easy for him to cut your throat as to shave you. But in
detection one should take no chances, give no one the benefit of the
doubt. Half of the undetected crimes in the world are due to our
reluctance to suspect anybody.”
“But surely,” urged Marryatt, “you would allow character to go for
something? I was a schoolmaster once, and while one knew the little
beasts were capable of almost anything, one did clear some people
of suspicion on mere character.”
“But there again,” argued Gordon, “you knew them very well.”
“Not really,” said Marryatt. “A perpetual war of mutual deception is
kept up between schoolmasters and schoolboys. One trusted, I
think, one’s unconscious impressions most.”
“If I were a detective,” persisted Reeves, “I would suspect my own
father or mother as soon as anybody else. I would follow up every
clue, blinding myself deliberately to the thought, where does this
point to?”
“Then you would be unreasonable,” said Gordon. “In the old days,
when the answers of sums used always to come out to an integral
number—I believe it’s all different now—one was wiser than that. If
you could see that the answer was going to involve two-thirds of a
policeman, you argued at once that you were on the wrong track;
you started again, and suspected your working.”
“But real life,” retorted Reeves, “doesn’t always work out to a
simple answer. And if the policeman who’s in charge of a case
argues as you’re arguing, he’s only himself to blame for it if he gets
trisected by the criminal before he’s finished.”
“At least you must respect the principle of Cui bono?”
“It’s extraordinary,” began Carmichael, “how many people make
the old mistake about the meaning of——”
“Cui bono is the worst offender of the lot,” said Mordaunt Reeves
cheerfully. “Look at those two boys in America who murdered
another boy just to find out what it felt like.”
“But that was pathological.”
“And how many crimes aren’t pathological, if it comes to that?”
“I was on Holy Island once for a month,” said Carmichael, “and
would you believe it, there was a man there that was sick if he ever
caught sight of a dog? Sick, positively.”
“What do you think it really feels like,” asked Marryatt, “to have
murdered a man? I mean, murderers in general always seem to lose
their heads when the thing is actually done, and give themselves
away somehow. But one would have thought, if the thing is planned
with proper deliberation, one’s feeling would be that things were
working out according to plan, and the next thing was to get clear—
above all things, to see plenty of people, and to behave quite
naturally in company.”
“Why that?” asked Gordon.
“To establish your alibi. People are often careless about that.”
“By the way,” asked Carmichael, “did you bring a paper with you
down from London? I’m interested to see the verdict in that
Stanesby case. The young fellow is connected, I hear, with the
Stanesbys of Martington.”
“Afraid I left London at three, and that’s too early for anything but
betting tips. I say, you fellows, it’s stopped raining.”
 Chapter II.
In the Rough
The view from the third tee was one which even a golfer might
pause to admire. Let the Wordsworthian say what he will, railways
ennoble our landscape; they give to our unassuming valleys a hint of
motive and destination. More especially, a main line with four tracks
pillowed on a sweep of tall embankment, that cannot cross a
meandering country stream without a stilt-walk upon vast columns
of enduring granite, captivates, if not the eye, at least the
imagination. Such was the railway that stretched far into the
distance, paralleling the course of your drive on the right: such was
the great viaduct, some hundred feet ahead of you, that spanned
laboriously, over four giant arches, the little river Gudgeon, most
insignificant of streams. Shallow and narrow it ran, fringed by
willow-herb and meadow-sweet, a paddling-place for cows and for
unoccupied caddies. Here and there it threw out a patch of osiers—
one in particular, that nestled at the foot of the railway-arches, was
especially dreaded by golfers. In front, just visible above the railway
where it receded northwards, were the thatched and tiled roofs of
Paston Whitchurch, the next station down the line. To the right lay
the old house, in its melancholy grandeur, behind it the village and
church of Paston Oatvile. A superb avenue of elms connected the old
house with the road between the two villages. The sun had newly
come out, showing grass the greener and earth the browner for the
late rain; elemental scents of turf and furrow greeted its restoration.
It may be doubted whether Mordaunt Reeves was particularly
sensitive to such influences; if he was, it may have been this
distraction which made him slice his drive. The ball dwindled down
the gradual slope towards the river; cleared in a couple of bounds
the tussocks of thick grass that dotted the little valley, and buried
itself at last in the osier bed at the foot of the arches. Gordon and he
—they were partners—set out at once to retrieve it, distrusting the
efforts of an inefficient caddie, who was nearer the spot. It was only
a closer view that showed how well-chosen a lair was this for a golf-
ball hard pressed in the chase. The ground was all tussocks of rank
grass, with hidden runlets that made islands of them; stubborn little
shoots of willow arrested the searching club. They might have spent
a full half-hour in vain scrutiny, had not Reeves’s eye lighted
suddenly on something he never looked to find there, a darker patch
among the surrounding green, close to the foot of the first arch. It
showed the outlines of a man.
A dog sleeps on the alert, with the visible threat of waking at any
moment. A man’s sleep is like the sleep of the horse; it imitates
death. Reeves’s first idea was that this man who lay so still must be
a tramp who had strayed off the London high-road, and was taking
his siesta in the lee of the viaduct. Then a gleam of more than
military intelligence assured him that on such an afternoon of
downpour a man composing himself to sleep would have been under
the arch, not by the side of it. “Hello!” he shouted uneasily to
Gordon, “looks as if there was something wrong here.” Together
they approached the prostrate body; it lay face downwards, and
there was no movement of life. The thrill of distaste with which
healthy nature shrinks from the sight of dissolution seized both of
them. Gordon had served three years in the army, and had seen
death; yet it was always death tricked out in the sacrificial garb of
khaki; there was something different about death in a town-coat and
striped grey trousers—it was a discord in the clear weather. The sun
seemed to lose a shade of its brightness. Together they bent, and
turned the body over, only to relinquish it again by a common
instinct. Not only did the lolling head tell them that here the
architecture of the human frame had been unknit; the face had
disappeared, battered unrecognizably by some terrible and
prolonged friction. They looked upwards, and knew at once that the
sloping buttress of the arch, all of rough granite, must have
intercepted a fatal fall, and added to its horror. Little about the head
could be distinguished except closely-cut grey hair.
“Poor devil,” said Gordon huskily. “Down from the line, I suppose.”
“I say,” said Reeves, “we mustn’t let the caddie see this. Send him
across to fetch the other two.” Marryatt and Carmichael were now
close behind them, and came up almost immediately.
“Is there somebody dead?” asked Marryatt. “I say, how awful.” He
kept on walking up and down as if thoroughly unnerved, repeating
to himself, “How awful.” Carmichael, for once, was dumb. It was a
new voice that summed up the situation, in the words, “ ’E’s got ’is
properly, ain’t ’e?” and they turned round to find the caddie
obviously enjoying a new sensation.
“Look here, we must move this somehow,” suggested Gordon.
“What about the tool-house under that arch?”
“I’m not quite sure I could lift it,” said Reeves.
“That’s all right, sir,” said the caddie, “I’ll whistle across to Ginger;
in the scouts ’e was; they teach ’em what to do with bodies and
that. ’Ere, Ginger!” and as his fellow-caddie approached, “Bloke fell
off of the railway-line and smashed hisself up something cruel.”
Ginger whistled: “Dead, is ’e?” “Not half ’e ain’t; shamming, that’s
what ’e is; go and ’ave a look at ’im.”
Ginger satisfied his curiosity on the point; and these two cold-
blooded young persons proceeded to hoist the body on to an
ingenious arrangement of sticks, and so carried it off, under
Gordon’s directions, to the tool-house.
As the spell of the uncanny presence was removed, Reeves’s
horrified embarrassment ebbed from him a little, and left him with
the sense that he ought to take command of the proceedings.
“Where’s Beazly likely to be?” he asked—Beazly was the doctor.
“He went out in the rain,” said Marryatt; “I should say he’d be
about the tenth or eleventh by now. Look here, I’ll nip across and
get him,” and in a moment he was running across the fairway.
“Seemed glad to get away,” said Reeves; “well, it’s too late for
visiting the sick, and too soon for burying the dead. Carmichael,
you’re looking a bit on edge, too; would you mind going across to
Paston Whitchurch station and ’phoning up the police? Binver, I
suppose, is the nearest place to get a bobby from. You will? Good.”
And as Carmichael too made off, “Look here, Gordon, what are we
going to do about it? I’ve got the feeling that there’s something
wrong here. What do you say to doing a bit of detective work on our
own—or are you feeling rotten?”
“Oh, I’m feeling all right,” said Gordon, “only what about the
police? Won’t they want to look through the man’s things first? It
would be awkward if we put ourselves on the wrong side of the law.
Funny thing, I’ve no idea whether there’s any law against searching
a dead body; yet, if there isn’t, how do the police ever get their
clues?”
“Oh, rot, the police can’t be here for a good half-hour, and Beazly
won’t mind if he comes along. Let’s take a bit of a look round,
anyhow. He fell off the arch, and smashed up his face against the
buttress, that looks pretty clear. Now, did he fall off the line, or off a
train?”
“If you ask me, I should say he fell off the parapet. I’ve noticed,
sometimes, what a long way it really is from the door of one’s
carriage to the parapet—a man falling from a carriage would never
reach the edge.”
“Ah,” said Reeves, looking up, “but you’re imagining the train
stationary. He would be hurled forward some way by the impetus, if
he jumped off a moving train. And I should say he could have
started falling down that bank to the right, just before the parapet
begins. He’d roll forwards and sideways, if you see what I mean, till
he got to where the stonework begins, up there, and then, plop.”
“I dare say you’re right. Anyhow, we’d better be quick and look at
the body.”
As they went towards the tool-house, Reeves gave a sudden
exclamation. “By Jove, his hat! And it’s—let’s see—I should say
fifteen yards to the north of the body. Now why?”
“How do you mean?”
“There was no wind this afternoon. If his hat fell with him, it
would lie with him. If it lies a dozen yards away, that looks as if—as
if it was thrown after him. The considerate fellow-passenger hardly
does that, does he?”
“You mean there’s been dirty work?”
“I mean it looks as if there’d been dirty work. Now for the tool-
shed.”
To search a dead body is not an easy performance, unless you are
in a hurry and have got to do it. Gordon did most of the work, and
Reeves checked his results for him. The pockets contained a
handkerchief, marked with the name “Masterman,” a cigarette-case,
of a common pattern, containing a cigarette of a brand smoked by
every second man in the neighbourhood, a half-empty box of
matches, a pipe and an empty pouch, two florins, a letter and a
business communication both addressed to S. Brotherhood, Esq.,
and a watch and chain. They also found, written on the back of the
letter, a pencilled list of goods, as if to remind a man of his shopping
needs.
“It’s a queer thing,” said Reeves, “that watch; because he’s got
one on his wrist too. How many people, I wonder, carry a stomach-
watch as well as a wrist-watch? It’s stopped, I suppose?”
“Blessed if it isn’t going! An hour fast, apparently, but going. Good
advertisement for the makers, what?”
“But the wrist-watch?”
“That’s stopped.”
“When?”
“Six minutes to five.”
“What did I say about trains? The 4.50 from Paston Oatvile would
be just passing here at six minutes to five. How’s that for
deduction?”
“Looks all right, anyhow. And, by Gad, here’s a third single from
town to Paston Whitchurch. Is to-day the sixteenth? Yes, then that’s
quite on the square. Now, stand by while I see if his clothes are
marked.”
But neither coat nor shirt, neither collar nor trousers bore any
mark of ownership. The suit was from Messrs. Watkins in New
Oxford Street, the shirt and collar were of a brand which it would be
mere advertisement to mention. During all this time, Reeves was
making a transcript of the three documents, not without a certain
sense of intrusion upon a dead man’s confidence. As Gordon began
to look into one of the boots, Reeves gave a whisper of warning, and
a policeman (for they have motorcycles even in the police force)
came into distant view. Panic seized the forces of Baker Street, and
(forgetting that they had a perfect right to be in charge of the dead
man’s body) they resumed, very shamefacedly, their search for the
lost ball. It seemed incongruous somehow, to be worrying about a
golf-ball—ought there to be a local rule about what happened if you
found a corpse on the links? Certainly the game had been
abandoned, and the caddies, to their great regret, sent back with
the clubs.
“Good evening, gentlemen,” said the policeman, eyeing them
narrowly. It was not that he suspected them or anybody of anything;
he merely sized them up by force of habit to see whether they were
the kind of people you touched your hat to or the kind of people you
told to move on. The scrutiny being favourable, he allowed them to
slash about in the undergrowth and watch, with ill-concealed
curiosity, the official proceedings of Scotland Yard.
Scotland Yard did very much what they had done, only with a
splendidly irrelevant thoroughness. Not only the destination, class,
and date of the ticket had to be registered in the notebook, but its
price—there even seemed to be a moment’s hesitation about the
Company’s regulations on the back. Nor did the names of the
cigarette-importer and the collar-maker go unrecorded; both
watchmakers, the post-marks on the correspondence, the date on
the florins—nothing escaped this man. Tired of waiting for the
doctor and the inevitable ambulance, Gordon and Reeves abandoned
the truant ball, and made their way thoughtfully to the dormy-house.
Wilson, the club gossip, met them at the entrance. “Heard about
old Brotherhood?” he asked, and went on, before they had time to
gasp: “He’s gone bankrupt; heard it to-day in the City.”
“Really?” said Reeves. “Come and have a drink.” But if he thought
that he too had the telling of a story, he was mistaken; the door
opened on a well-known voice:
 “Yes, sliced his drive badly, did Reeves. A curious thing, that,—you
‘slice’ a ball in golf and you ‘cut’ a ball at cricket, and it’s the same
action in either case, and yet it’s nothing whatever to do with the
motion of cutting a cake. What was I saying? Oh yes. Right against
the viaduct—did you ever see the big viaduct they’ve got at Welwyn?
A finer one than ours, even—he found . . .”
Which made it evident that Mr. Carmichael was telling, in his own
way, the story of the day’s adventure.
 Chapter III.
Piecing it Together
If the general accommodation at the Paston Oatvile dormy-house
cannot be described as cloistral, it must be admitted that the rooms
in it where you can claim privacy are not much better than cells.
Mordaunt Reeves, however, had done something to turn his
apartments into a civilized dwelling-place; there were pictures which
did not illustrate wings, and books devoted to other subjects than
the multitudinous possibilities of error in playing golf. Gordon and he
had each a comfortable arm-chair, each a corner of the fire-place to
flick his cigarette-ash into, when they met that evening to talk over
the possibilities of the situation as it had hitherto developed.
“Everybody,” said Reeves, “if you notice, has already started
treating an assumption as if it were a fact. They all say it was
Brotherhood we found lying there; they all say he committed suicide
because he had just gone bankrupt. Now, as a matter of fact, we
don’t know that it was Brotherhood at all. He has not been heard of,
but there hasn’t been much time to hear of him; and nothing is
more probable than that a man who has gone bankrupt should skip
without leaving any traces.”
“Yes, but somebody’s dead; you’ve got to find a gap somewhere in
the ranks of Society to match our corpus.”
“Still, that’s mere negative arguing. And there are several points
that tell against its being Brotherhood. In the first place, that ticket.
Brotherhood goes up and down every day; do you mean to tell me
he hasn’t got a season? Second point, if it was Brotherhood there’s
an odd coincidence—he died within ten minutes’ walk of his own
bungalow; why there, any more than anywhere else on the line?”
“It’s a coincidence that Brotherhood should be killed so near his
own bungalow. But the murder, whether we like it or not, has been
committed just there, so I don’t see why it shouldn’t be him as much
as anybody else. However, go on.”
“Third point, the handkerchief. Why should Brotherhood be
carrying somebody else’s handkerchief?”
“If it comes to that, why should somebody else be carrying
Brotherhood’s correspondence?”
“Oh, Brotherhood is mixed up in it somehow right enough. We
shall see. Next point to be considered, was it accident, suicide, or
murder?”
“You can cut out accident, surely. That would be a coincidence—
somebody carrying Brotherhood’s letter to fall out of the train by
mere accident just where Brotherhood lives.”
“Very well, for the present we’ll ask Murder or Suicide? Now, I’ve
several arguments against suicide. First, as I told you, the hat. He
wasn’t alone when he fell out of the carriage, or who threw the hat
after him?”
“There was no mark in the hat, was there?”
“Only the maker’s; that’s the irritating thing about this business.
Hats, collars, shirts, people buy them at a moment’s notice and pay
cash for them, so there’s no record in the books. And watches—of
course you don’t have a watch sent, you take it with you, to save
the danger of carriage by post. I’ll try all those tradesmen if the
worst comes to the worst; probably the police have already; but I
bet nothing comes of it.”
“What’s your next argument against suicide?”
“The ticket. That extra four bob would have got him a first instead
of a third. Now, a man who means to commit suicide doesn’t want
four bob, but he does want to be alone.”
“But the suicide might have been an impulse at the last moment.”
“I don’t believe it. The place where he fell was just the one place
about here where he was bound to kill himself, not merely maim
himself. That looks like preparation.”