Combinatorial Physics Combinatorics Quantum Field Theory and Quantum Gravity Models Adrian Tanasa Full Chapter PDF
Combinatorial Physics Combinatorics Quantum Field Theory and Quantum Gravity Models Adrian Tanasa Full Chapter PDF
Combinatorial Physics Combinatorics Quantum Field Theory and Quantum Gravity Models Adrian Tanasa Full Chapter PDF
https://ebookmass.com/product/introduction-to-quantum-field-
theory-with-applications-to-quantum-gravity-1st-edition-iosif-l-
buchbinder/
https://ebookmass.com/product/quantum-field-theory-and-critical-
phenomena-5th-edition-jean-zinn-justin/
https://ebookmass.com/product/quantum-space-loop-quantum-gravity-
and-the-search-for-the-structure-of-space-time-and-the-universe-
baggott/
https://ebookmass.com/product/philosophy-beyond-spacetime-
implications-from-quantum-gravity-christian-wuthrich/
Quantum Communication, Quantum Networks, and Quantum
Sensing Ivan Djordjevic
https://ebookmass.com/product/quantum-communication-quantum-
networks-and-quantum-sensing-ivan-djordjevic/
https://ebookmass.com/product/quantum-physics-of-semiconductor-
materials-and-devices-debdeep-jena/
https://ebookmass.com/product/quantum-bullsht-how-to-ruin-your-
life-with-advice-from-quantum-physics-chris-ferrie/
https://ebookmass.com/product/ideas-of-quantum-chemistry-
volume-1-from-quantum-physics-to-chemistry-3rd-edition-lucjan-
piela/
https://ebookmass.com/product/field-theoretic-simulations-in-
soft-matter-and-quantum-fluids-glenn-h-fredrickson/
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
C O M B I N AT O R I A L P H Y S I C S
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
Combinatorial Physics
Combinatorics, quantum field theory, and quantum
gravity models
Adrian Tanasa
University of Bordeaux, France
1
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Adrian Tanasa 2021
The moral rights of the author have been asserted
First Edition published in 2021
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2021932400
ISBN 978–0–19–289549–3
DOI: 10.1093/oso/9780192895493.001.0001
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
Contents
1 Introduction 1
viii Contents
Contents ix
x Contents
13.3 Combinatorial analysis of the general term of the large N expansion 219
13.3.1 Dipoles, chains, schemes, and all that 220
13.3.2 Generating functions, asymptotic enumeration,
and dominant schemes 226
13.4 The double-scaling limit 230
13.4.1 The two-point function 231
13.4.2 The four-point function 232
13.4.3 The 2r-point function 232
13.5 Selected further reading 233
3
14 Random tensor models—the O(N ) -invariant model 234
14.1 General model and large N expansion 234
14.2 Quartic model, large N expansion 241
14.2.1 Large N expansion: LO 242
14.2.2 NLO 247
14.3 General quartic model: Critical behaviour 248
14.3.1 Explicit counting of melonic graphs 248
14.3.2 Diagrammatic equations, LO and NLO 252
14.3.3 Singularity analysis 253
14.3.4 Critical exponents 256
14.4 Selected further reading 259
15 The Sachdev–Ye–Kitaev (SYK) holographic model 260
15.1 Definition of the SYK model: Its Feynman graphs 261
15.2 Diagrammatic proof of the large N melonic dominance 264
15.3 The coloured SYK model 271
15.3.1 Definition of the model, real, and complex versions 271
15.3.2 Diagrammatics of the real and complex model 272
15.3.3 More on the coloured SYK Feynman graphs 282
15.3.4 Non-Gaussian disorder average in the complex model 284
15.4 Selected further reading 290
16 SYK-like tensor models 291
16.1 The Gurau–Witten model and its diagrammatics 292
16.1.1 Two-point functions: LO, NLO, and so on 293
16.1.2 Four-point function: LO, NLO, and so on 295
16.2 The O(N )3 -invariant SYK-like tensor model 300
16.3 The MO SYK-like tensor model 303
16.4 Relating MO graphs to O(N )3 -invariant graphs 304
16.5 Diagrammatic techniques for O(N )3 -invariant graphs 306
16.5.1 Two-edge-cuts 306
16.5.2 Dipole removals 307
16.5.3 Dipole insertions 309
16.5.4 Chains of dipoles 310
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
Contents xi
Bibliography 383
Index 395
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
1
Introduction
The interplay between combinatorics and theoretical physics is a recent trend which
appears to us as particularly natural, since the unfolding of new ideas in physics is
often tied to the development of combinatorial methods, and, conversely, problems
in combinatorics have been successfully tackled using methods inspired by theoretical
physics. A lot of problems in physics are thus revealed to be enumerative. On the other
hand, problems in combinatorics can be solved in an elegant way using theoretical
physics-inspired techniques. We can thus speak nowadays of an emerging domain of
Combinatorial Physics.
The interference between these two disciplines is moreover an interference of multiple
facets. Thus, its most known manifestation (both to combinatorialists and theoretical
physicists) has so far been the one between combinatorics and statistical physics,
or combinatorics and integrable systems, as statistical physics relies on an accurate
counting of the various states or configurations of a physical system.
However, combinatorics and theoretical physics interact in various other ways. One
of these interactions is the one between combinatorics and quantum mechanics, because
combinatorial tools can be used here for a better mathematical understanding of the
algebras underlying quantum mechanics.
In this book, we mainly focus on yet another type of these multiple interactions
between combinatorics and theoretical physics, the one between combinatorics and
quantum field theory (QFT). We estimate that combinatorics is built into the mathe-
matical formulation of QFT. This stems initially from the fact that the most popular
tool of QFT is perturbation theory in the coupling constant of the model, which means
that one considers Feynman graphs, with appropriate combinatorial weights, in order to
encode the physical information of the respective system. Moreover, one elegant way of
expressing the Feynman integrals associated with these graphs is to use the Kirchhoff–
Symanzik polynomials of the parametric representation, polynomials which can be
proven to be related to some multi-variate version of the celebrated Tutte polynomial
of combinatorics. A particularly elegant way to prove this is to use the Grassmann
development of the determinants and Pfaffians involved in these computations. Let us
emphasize here that this Grassmann development uses Grassmann calculus, which were
developed by physicists to express fermionic QFT. Grassmann calculus is further used in
this book to give a simple proof of the celebrated Lingström–Gessel–Viennot (for graphs
Combinatorial Physics: Combinatorics, Quantum Field Theory, and Quantum Gravity Models. Adrian Tanasa,
Oxford University Press (2021). © Adrian Tanasa. DOI: 10.1093/oso/9780192895493.003.0001
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
2 Introduction
with cycles) and to further generalize some identities, initially proved by Stembridge, in
the same context of graphs with cycles.
The so-called 0−dimensional QFT (called by some authors, combinatorial QFT), or
more precisely the use of the intermediate field method in this setting, allows to establish a
theorem concerning partial elimination of variables in the celebrated Jacobian conjecture
(which concerns the global invertibility of polynomial systems).
Moreover, analytic combinatorial techniques are used on a regular basis in QFT
computations. Thus, the propagator of any scalar model can be represented using the
heat kernel. The Mellin transform technique can also be used in order to rapidly prove
the meromorphy of Feynman integrands. The saddle point method is frequently used to
tame the divergent behaviour of these integrals.
Last but not least, renormalization in QFT (which one can say lies at the very heart of
QFT) has a highly non-trivial combinatorial core, and this has been recently presented
in a combinatorial Hopf algebra form—the Connes–Kreimer Hopf algebra. Related to this,
the Hochschild cohomology of this combinatorial Hopf algebra can be used to express
the combinatorics of the Dyson–Schwinger equation (DSE) as a simple power series in
some appropriate insertion operator of Feynman graphs.
All these combinatorial techniques (analytic or algebraic) generalize to more involved
QFT models. Thus, non-commutative QFT (that is, QFT on a non-commutative space-
time) also possesses most of these combinatorial properties. First, the graphs used in
QFT are uplifted to ribbon graphs (or combinatorial maps). Furthermore, one can still
use the heat kernel for propagators of the theories, but in order to have renormalizable
models, one needs to use a more involved special function, the Mehler kernel or some
non-trivial modification of the heat kernel. Moreover, the Mellin transform technique can
again be used, as in the case of commutative QFT. The corresponding non-commutative
Kirchhoff–Symanzik polynomials are proven to be a limit of a multi-variate version of
the Bollobás–Riordan polynomial (which is a natural generalization for ribbon graphs
of the universal Tutte polynomial). Finally, algebraic combinatorial techniques can also
be used in non-commutative QFT. The corresponding combinatorial Connes–Kreimer
Hopf algebra of ribbon Feynman graphs can be defined and related to non-commutative
renormalization. Furthermore, the appropriate Hochschild cohomology then describes
the combinatorics of the DSE of these models.
Non-commutative QFT can also be seen as a special case of the celebrated matrix
models. Following this line of reasoning, one can naturally generalize random matrix
models to random tensor models,
The combinatorics of tensor models per se is extremely involved. One cannot just use
the genus to characterize the so-called large N expansion, N being the size of the matrix
resp. of the tensor. It is worth emphasizing here that the large N expansion is, from a
combinatorial point of view, a certain asymptotic expansion (corresponding to the limit
N → ∞).
However, in order to make the combinatorics simpler, several QFT-inspired simpli-
fications of tensor models can be proposed. The first two such simplications were the
coloured model and the multi-orientable models. For both of these models, one can
implement the large N expansion and the double-scaling mechanism, which, in the case
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
Introduction 3
of matrix models, are very important mathematical physics tools. Several other models
(based on U (N ) and then O(N ) models have also been studied.
Feynman graphs associated to tensor models, through the celebrated QFT perturba-
tive expansions, are called tensor graphs and can be seen as a natural 3D generalization of
maps or of ribbon graphs. The dominant term of the large N expansion of the previously
mentioned tensor models are the so-called melonic graphs, which are, from a graph
theoretical point of view, a particular case of series-parallel graphs.
The large N expansion is controlled in the 2D case, the matrix model case, by the
genus of the corresponding combinatorial maps. In dimension higher than two, there
is no direct analogue of the genus. Nevertheless, the tensor asymptotic expansion in N
is controlled by an integer, called the degree, which is defined as the half-sum of the
non-orientable genus of ribbon graphs canonically embedded in a tensor graph (called
the jackets of the respective tensor graphs). The degree is thus a half integer, naturally
generalizing the 2D notion of genus for tensor models.
In order to study the general term of the large N expansion of various such tensor
models, we extensively use, in this book, various graph theoretical and enumerative
combinatorics techniques to perform their enumeration and we establish which are the
dominant configurations of a given degree.
It is worth emphasizing here that tensor models have recently been proven by
Witten to be related to the celebrated holographic Sachdev–Ye–Kitaev (SYK) quantum
mechanical model. This comes from the fact that, in the so-called large N expansion
(N being in the case of the SYK model the total number of fermions of the model),
both types of models are dominated by the melonic graphs. The large N expansion
for various SYK-like tensor models is then studied using again graph theoretical and
enumerative combinatorics techniques. These techniques allow us to asymptotically
enumerate Feynman graphs of various SYK-like tensor models.
The book is organized as follows. In Chapter 2, we present some notions of graph
theory that will be useful in the rest of the book. It is worth emphasizing that graph
theorists and theoretical physicists adopt, unfortunately, different terminologies. We
present here both terminologies, such that a sort of dictionary between these two
communities can be established. We then extend the notion of graph to that of maps (or
of ribbon graphs). Moreover, graph polynomials encoding these structures (the Tutte
polynomial for graphs and the Bollobás–Riordan polynomial for ribbon graphs) are
presented.
In Chapter 3, we briefly exhibit the mathematical formalism of QFT, which, as
mentioned previously, has a non-trivial combinatorial backbone. The QFT setting can
be understood as a quantum description of particles and their interactions, a description
which is also compatible with Einstein’s theory of special relativity. Within the framework
of elementary particle physics (or high energy physics), QFT led to the Standard
Model of Elementary Particle Physics, which is the physical theory tested with the best
accuracy by collider experiments. Moreover, the QFT formalism successfully applies
to statistical physics, condensed matter physics, and so on. We show in this chapter
how Feynman graphs appear through the so-called QFT perturbative expansion, how
Feynman integrals are associated to Feynman graphs, and how these integrals can be
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
4 Introduction
Introduction 5
6 Introduction
2
Graphs, ribbon graphs,
and polynomials
In this chapter, we present some notions of graph theory that will be useful in the
rest of this book. Let us emphasize that graph theorists and quantum field theorists
adopt, unfortunately, different terminologies. We present both here, such that a sort of
dictionary between these two communities may be established.
We then extend the notion of graphs to that of maps (or of ribbon graphs). Moreover,
graph polynomials encoding these structures (the Tutte polynomial for graphs and the
Bollobás–Riordan polynomial for ribbon graphs) are presented.
In this chapter, we follow the original article (Thomas Krajewski et al. 2010) and the
review article (Adrian Tanasa 2012).
Definition 2.1.1 A graph Γ is defined as a set of vertices V and of edges E together with an
incidence relationship between them.
Notice that we allow multi-edges and self-loops (see definition 2.1.2 4), but still use
the term ‘graph’ (and not ‘pseudograph’).
The number of vertices and edges in a graph are also noted V and E for simplicity,
since our context prevents confusion.
One needs to emphasize that in QFT a supplementary type of edge exists, external
edges. These edges are only hooked to one of the vertices of the graph, the other end
of the edge being ‘free’ (see Fig. 2.1 for an example of such a graph, with four external
edges). In elementary particle physics, these external edges are related to the observables
in some experiments.
Combinatorial Physics: Combinatorics, Quantum Field Theory, and Quantum Gravity Models. Adrian Tanasa,
Oxford University Press (2021). © Adrian Tanasa. DOI: 10.1093/oso/9780192895493.003.0002
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
f3
f1 e2
e3 e4
f2 e1
f4
Figure 2.1 A Φ4 graph, with four internal edges and four external edges
Definition 2.1.2
1. The number of edges at a vertex is called the degree of the respective vertex ( field
theorists refer to this as the coordination number of the respective vertex).
2. An edge whose removal increases the number of connected components of the
respective graph is called a bridge ( field theorists refer to this as a 1-particle
reducible edge).
3. A connected subset of equal number of edges and of vertices which cannot be
disconnected by removing any of the edges is called a cycle ( field theorists refer
to this as a loop).
4. An edge which connects a vertex to itself is called a self-loop ( field theorists refer to
this as a tadpole edge).
5. An edge which is neither a bridge nor a self-loop is called regular.
6. An edge which is not a self-loop is called semi-regular.
7. A graph with no cycles is called a forest.
8. A connected forest is called a tree.
9. A two-tree is a spanning tree without one of its edges.
10. The rank of a subgraph A is defined as
Remark 2.1.3 For a connected graph, the nullity defined previously represents the number of
independent circuits.
In QFT, one often uses the term (number of) loops to denote (the number of)
independent loops.
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
2 2
1 3 1
4 4
Figure 2.2 An example of a graph (with seven edges and six external edges). We chose a spanning tree
and labelled its edges by 1, . . . , 4. We then chose the edge 3 to be removed. The set {1, 2, 4} is a two-tree;
one has two connected components (the first one formed by the edges 1 and 2 and the second one formed
by edge 4). The external edges are attached to one of these two connected components
Remark 2.1.4 A two-tree generates two connected components on the respective graph. Let us
also note that a two-tree can be defined as a spanning forest with two connected components (in
this way, no relation with a tree is given).
Let us illustrate this in Fig. 2.2. One can define two natural operations for an arbitrary
edge e of some graph Γ:
1. the deletion, which leads to a graph noted Γ − e;
2. the contraction, which leads to a graph noted Γ/e. This operation identifies the
two vertices v1 and v2 at the ends of e into a new vertex v12 , attributing all the
edges attached to v1 and v2 to v12 ; finally, the contraction operation removes e.
For an illustration of these two operations, one can refer to Fig. 2.3, where these
operations are iterated until one reaches terminal forms namely graphs formed only
of self-loops and bridges.
Let us now give a first definition of the Tutte polynomial:
TΓ (x, y) := (x − 1)r(E)−r(A) (y − 1)n(A) . (2.3)
A⊂E
e2
e3 e4
e1
G–e2
G/e2
e3 e1
e1 e3
G–e3 e4 e4
G/e4
G/e3 G–e3
e1 e4 e1 e1
e3
e1 e4 e3
G–e1 G/e1
e3 e3
Figure 2.3 The deletion/contraction of some graph. One is left with various possibilities (here five) of
terminal forms (that is, graphs with only bridges or self-loops)
This property of the Tutte polynomial is often used as its definition, if one completes
it by giving the form of the Tutte polynomial on terminal forms:
TΓ (x, y) := xm y n , (2.5)
Similarly, one can prove that the multi-variate Tutte polynomial (2.6) satisfies the
deletion/contraction relation, for any edge e. The definition of the polynomial on the
terminal forms (graphs with v isolated vertices) is
One can prove (through direct inspection) the relation between the Tutte polynomial
(2.3) and its multi-variate counterpart (2.6):
q −V ZΓ (q, β) |βe =y−1,q=(x−1)(y−1) = (x − 1)k(E)−V TΓ (x, y). (2.8)
It is this version of the multi-variate Tutte polynomial (2.6) that we use in this book to
prove the relation with the parametric representation of Feynman integrals in QFT (see
Chapter 6.6).
Putting aside the Tutte polynomial, several graph polynomials have been defined and
extensively studied in the literature. As we have seen previously, the Tutte polynomial
is a two-variable polynomial. It has one-variable specializations, such as the chromatic
polynomial or the flow polynomial.
The chromatic polynomial is a graph polynomial PΓ (k) (k ∈ N ) which counts the
number of distinct ways to colour the graph Γ with k or fewer colours, colourings being
counted as distinct even if they differ only by permutation of colours. For a connected
graph, this polynomial is related to the Tutte polynomial (2.3) by the relation
In order to define the flow polynomial, we need a finite abelian group G. One
can arbitrarily choose an orientation for each edge of the graph Γ, the result being
independent of this choice (the same type of situation appears when computing Feynman
integrals, see next section). A G−flow on Γ is a mapping
ψ : E → G, (2.10)
One has:
Definition 2.2.1 A ribbon graph Γ is an orientable surface with its boundary represented
as the union of closed disks, also called vertices, and ribbons also called edges, such that:
the disks and ribbons intersect in disjoint line segments, each such line segment lying on the
boundary of precisely one disk and one ribbon and finally, every ribbon containing two such
line segments.
Note that, as in the case of graphs, this definition can be extended by adding a new
type of edge (not with two line segments, see the previous definition) such that external
edges are allowed (see previous subsection). Examples of such graphs are given in
Figs. 2.4 and 2.5.
Let us also mention that ribbon graphs can be defined as graphs equipped with a
cyclic ordering of the incidence edges at each vertex or as graphs embedded in surfaces
(the latter was actually the mathematical object on which B. Bollobás and O. Riordan
defined their generalization of the Tutte polynomial in Bollobás and Riordan (2001) and
Bollobás and Riordan (2002), see following section).
An interesting connection between the Tutte polynomial of graphs and combinatorial maps
was proven in Bernardi (2008). He gave a characterization of the Tutte polynomial of
graphs which is different to the initial one given by Tutte (which required some choice
of linear order on the edge set, in order to write the Tutte polynomial as a function of
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
Figure 2.4 An example of a ribbon graph with one vertex, one internal edge, and two external edges
Figure 2.5 An example of a ribbon graph with two vertices, three internal edges, and two external edges
spanning trees). O. Bernardi proved that the Tutte polynomial can also be written as the
generating function of spanning trees counted with the help of a cyclic order of the edges
around each vertex.
Let us now give the following definition:
For example, the graph of Fig. 2.4 has two faces, while the one of Fig. 2.5 has a single
face. If we glue disks along the faces we obtain a closed Riemann surface whose genus is
also called the genus of the graph.
Definition 2.2.3 The ribbon graph is called planar if it has vanishing genus.
For example, the graph of Fig. 2.4 is planar while the one of Fig. 2.5 is non-planar (it
has genus 1).
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
Definition 2.2.4 A planar ribbon graph is called regular if the number of faces broken by
external edges is equal to 1.
For example, the graph of Fig. 2.4 is planar irregular, while the one of Fig. 2.5 is
regular (it has only one face, which is also broken by both of the external edges).
Remark 2.2.5 Planar regular ribbon graphs are also known in combinatorics literature as
outer maps.
Note that we have denoted by F (H) the number of components of the boundary
of the respective subgraph H (the number of faces). The supplementary variable z is
required to keep track of the additional topological information.
Similar to the Tutte polynomial, the Bollobás–Riordan polynomial also obeys a
deletion/contraction relation (see Fig. 2.6 for an example).
e1
e2 f2
f1
e3
G/e2
G–e2
e1
e1
f1 f2 f2
f1
e3
e3
G/e1
G–e1
e3
e3
f1 f2 f2
f1
RΓ = xRΓ/e (2.14)
The situation is also analogous to that of the Tutte polynomial, in the sense that,
defining the Bolobás–Riordan polynomial on terminal forms transforms the property
(2.13) into a definition. On these terminal forms (that is, graphs with one vertex), the
polynomial is defined as:
RΓ ( y, z) := y E(H) z 2g(H) , (2.15)
H⊂Γ
3
Quantum field theory (QFT)—built-in
combinatorics
Combinatorial Physics: Combinatorics, Quantum Field Theory, and Quantum Gravity Models. Adrian Tanasa,
Oxford University Press (2021). © Adrian Tanasa. DOI: 10.1093/oso/9780192895493.003.0003
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
18 QFT—built-in combinatorics
Φ : RD → K, (3.1)
t := ıxD . (3.2)
This is known as a Wick rotation from Euclidean to Minkowskian space-time; the field
Φ(x) then describes an (elementary) particle. In this book, we only work with the
Euclidean signature.
A QFT model is defined by means of a functional integral representation of the partition
function Z ; from this partition function, Green functions (or Schwinger functions, see the
following section) can then be obtained (by functional derivation).
Let us now explain what we mean by all these notions. One first needs to define the
action, which, from a mathematical point of view, is a functional in the field φ(x). For the
Φ4 model previously given, the action is written:
4 2
∂ 1
D
S[Φ(x)] = d x Φ(x) + m2 Φ2 (x) + V [Φ(x)], (3.3)
μ=1
∂xμ 2
where the parameters m and λ are referred to as the mass and the coupling constant
respectively. Moreover, the interaction potential is written:
λ 4
V [Φ(x)] = Φ (x). (3.4)
4!
Let us now introduce the notion of functional integration as the product ofintegrals at
each space point x (up to some irrelevant normalization factor): Dφ(x) := x dΦ(x).
This infinite multiplication of Lebesgue measure is mathematically ill-defined; for the
way in which to deal with this, the interested reader may refer for example to Manfred
Salmhofer’s book (1999). Another solution for this problem can be found, in Vincent
Rivasseau’s review article (2002) or Rivasseau (1992), where the quadratic part of the
action is sent inside the measure.
Following Razvan Gurau, Vincent Rivasseau, and Alessandro Sfondrini., let us now
present a yet different approach for this issue, which can be considered a different
starting point for defining a QFT model. Probability measures are characterized by the
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
Z := DΦ(x)e−S[Φ(x)] . (3.5)
1
G(N ) (x1 , . . . , xN ) := Dφ(x)Φ(x1 ) . . . Φ(xN )e−S[Φ] . (3.6)
Z
One can prove that only correlation functions with an even number of fields N are
non-vanishing (see any of the mentioned textbooks for details). For N = 2 one refers
to G(2) (x1 , x2 ) as the two-point function. Furthermore, when the coupling constant λ is
(2)
also taken to be vanishing, the two-point function G(2) (x1 , x2 ) is denoted by G0 (x1 , x2 )
and is referred to as the free two-point function (or the free propagator). On a more general
basis, the theory taken at vanishing value of the coupling constant λ is referred to as the
free theory.
Another important result one can prove using appropriate combinatorics is that any
N -point function can be expressed as a sum of (N − 1)!! different products of N/2 propagators
(2)
G0 :
(N −1)!! N/2
(N )
(2)
G0 (x1 , . . . , xN ) = S0 (xπi (2j−1) , xπi (2j) ). (3.8)
i=1 j=1
20 QFT—built-in combinatorics
Translation-invariance
One can explicitly check (see again any of the mentioned textbooks for details) that
the propagator is invariant under arbitrary translations a ∈ RD of the arguments of the
field Φ(x)
x → x + a. (3.9)
(2) (2)
G0 (x1 , x2 ) = G0 (x1 − x2 ). (3.10)
This translation-invariance property further generalizes to the level of the free N -point
function:
(N ) (N )
G0 (x1 , . . . , xN ) = G0 (x1 − xN , . . . , xN −1 − xN , 0). (3.11)
This can be proved using the Wick expansion (3.8) combined with the translation-
invariance property (3.10) of the two-point function.
∞
(N ) 1 (N )
G (x1 , . . . , xN ) = G (x1 , . . . , xN ). (3.12)
Z p=0 p
∞
Z= Zp , (3.13)
p=0
One can prove (see again any of the indicated textbooks) the following formula for
(N )
the coefficient Gp (x1 , . . . , xN ):
p
1 −λ
G(N )
p (x1 , . . . , xN ) = d D z1 . . . dD zp (3.14)
p! 4!
(N +4p)
G0 (z1 , z1 , z1 , z1 , . . . , zp , zp , zp , zp , x1 , . . . , xN ). (3.15)
This formula expresses the N -point function as a function of the free-field N -point
function. Furthermore, one can now use (3.8) in order to Wick-expand these free-field
N -point functions into sums over products of propagators S0 .
Working all these formulae leads to the following result:
G(N )
p (x1 , . . . , xN )
p N
1 −λ
= dD z1 . . . dD zp dD y1 . . . dD y4p+N δ (D) (y4p+ − x )
p! 4!
=1
p
δ (D) (y4k−3 − zk )δ (D) (y4k−2 − zk )δ (D) (y4k−1 − zk )δ (D) (y4k − zk )
k=1
4p+N
(4p+N −1)!!
2
G0 (yπ(4p+N
i
)
(2j − 1), yπ(4p+N
i
)
(2j)). (3.16)
i=1 j=1
Each of the products in the sum (3.16) can be depicted by a Feynman graph. Let us
give a few more explanations on this issue. The N external points (to which the external
edges hook) are
y4p+1 = x1 , . . . , y4p+N = xN
(through the δ -functions in the first line of equation (3.16)). One has a free propagator
G0 (y, y ) connecting any of the points of the graph. The order p in perturbation theory
is nothing but the number of vertices of the graph. For each such vertex one has four
δ -functions (the second line in equation (3.16)); this represents the locality of the
interaction. These 4p variables (equal by groups of four, each group for one of the
vertices of the Feynman graph) are integrated against; they are the internal points. For
each such vertex one then has a final integration over the point z . Let us recall that each
such vertex comes with a coupling constant λ.
We have thus seen that integrals can be associated with these graphs; this is done
through Feynman rules.
An N -point (Feynman) graph is a Feynman graph associated with the respective
N -point function.
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
22 QFT—built-in combinatorics
Thus, a four-point Feynman graph is given in Fig. 2.1 (it has four external edges).
Following Hagen Kleinert and Verena Schulte-Frohlinde (2001), let us now give a few
more definitions that are useful in the combinatorics of QFT:
This multiplicity can be computed using the following considerations. The symmetry
factor of the Φ4 interaction is 4!. This can be seen in the fact that there are 4! ways of
labelling the four incoming/outgoing edges of a vertex. (One can check this analytically by
performing the integrations over y1 , . . . , y4 ). Using the Feynman rules introduced earlier,
the associated integral writes
dD y1 . . . dD y4 δ (D) (y1 − z) . . . δ (D) (y4 − z)G0 (y1 , ȳ 1 ) . . . G0 (y4 , ȳ 4 )
where we have denoted by ȳ 1 , . . . , ȳ 4 the four external points to which the four propaga-
tors hook to (previously mentioned). There are indeed 4! ways of labelling the y1 , . . . , y4
variables inside the Feynman integrand of (3.17) which leave the result unchanged.
Nevertheless, this vertex factor can be reduced in the following cases:
1. The self-contraction of a vertex—two of the y points contract to each other. The
integral is written:
dD y1 . . . dD y4 δ (D) (y1 − z) . . . δ (D) (y4 − z)G0 (y1 , y2 )G0 (y3 , ȳ 3 )G0 (y4 , ȳ 4 )
This is the QFT tadpole (or (self-)loop in graph theory, see Definition 2.1.2 4).
Let us denote by S the number of tadpoles in the respective graph.
2. The double connection of a vertex—two of the y variables (say y1 and y2 ) contract
to the same point ȳ 1 . The integral thus is written
dD y1 . . . dD y4 δ (D) (y1 − z) . . . δ (D) (y4 − z)S0 (y1 , ȳ 1 )S0 (y2 , ȳ 1 )S0 (y3 , ȳ 3 )S0 (y4 , ȳ 4 )
The permutation y1 , y2 is irrelevant and thus the appropriate factor is 4!/2. Let us
denote by D the number of double connections in a graph.
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
3. The triple connection—three of the y variables contract to the same point; the
appropriate factor is 4!/3!. Let us denote by T the number of triple connections in
a graph.
4. The fourfold connection—all four of the y variables contract to the same point; the
appropriate factor is 1. Let us denote by F the number of fourfold connections in
a graph.
Another important combinatorial notion is the one of identical vertex permutations
(IVP). It corresponds to vertices that can be interchanged in the Feynman integrand
such that the final result is not affected. We denote the number of such IVP by NIVP .
One can now prove the following proposition:
Note that the entities appearing on the right-hand side (number of vertices, number of tadpoles
and so on) refer to the respective graph Γ. Nevertheless, in order to simplify the notations, we
have not indexed them with Γ.
Definition 3.2.3 The weight factor WΓ of a Feynman graph Γ with p vertices (or at order p
in perturbation theory) is defined by:
MΓ
WΓ := . (3.21)
(4!)p p!
where we have denoted by Ṽ int [Φ̃] the Fourier transform of the interaction poten-
tial (3.4).
As in position space, one has perturbative expansion in powers of the coupling
constant λ; this leads to the Feynman graphs described in the previous section. One
associates some orientation and some momentum to any edge (external or internal).
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
24 QFT—built-in combinatorics
Each such momentum is a four-vector (it lives on R4 ). The Feynman rules that allow
one to associate a Feynman integral with such a graph consist of:
• conservation of momenta—the sum of incoming momenta at a vertex must be equal
to the sum of outgoing ones. This is realized with a δ−function.
• Integration of the internal momenta against their propagators. For each such edge
of momentum p, one has a propagator
1
C(p) = . (3.23)
p 2 + m2
4
4
3 4 1
λ d p ei 2 + m2
i=1
p
i=1 ei
δ(pf1 + pf2 − pe1 − pe2 )δ(pe1 + pe3 − pe4 − pf4 )δ(pe2 − pe3 + pe4 − pf3 ). (3.24)
We invite the interested reader to find out the orientation of the edges which were chosen
for such a Feynman integral to occur. Because of the presence of the three δ−functions,
the number of remaining integrals is equal to the number of independent cycles of the
Feynman graph (the number of (independent) loops, in QFT terminology, see again
Definition 2.1.2 3). This is a property which is always valid in QFT.
In the end, the Feynman integral (3.24) leads to a logarithmic divergence in the high-
energy regime (|p| → ∞) of the internal momenta (the so-called ultraviolet regime). The
appearance of such divergences is actually a very frequent phenomena in QFT; it is the
renormalization process which, when possible, takes care of these infinities in a highly
non-trivial way (see Section 3.6).
Note that, in the case presented here, the mass m prevents the integral from being
divergent in the infrared regime (that is, |p| → 0) as well. We will come back to these
divergence issues in Section 3.6.
Inserting these formulae into the general expression of a Feynman integral allows to
integrate out (through Gaussian integrations) the internal momenta pe .
The Feynman integral then writes
∞ E
e−VΓ (pext ,α)/UΓ (α)
(e−m
2
α
φ(Γ) = dα ), (3.26)
0 UΓ (α)D/2
=1
where U (α), and respectively V (pext , α), are polynomials in the set of parameters α, and
in the set of external momenta pext and the set of parameters α respectively. They are
called the Kirchoff–Symanzik polynomials. Their exact expression is proven to depend
only on the structure of the respective graphs
UΓ (α) = α , (3.27)
T ∈T
and
VΓ (pext , α) = α ( pi ) 2 ,
T2 ∈T2 i∈E(T2 )
where we have denoted by T a tree of the graph and by T2 —a two tree which, as already
stated in the previous section (see remark 2.1.4), separates the graph in two connected
components. We have denoted by E(T2 ) one of the connected components thus obtained.
Remark 3.4.1 By momentum conservation, the total momenta of one of these connected
components (for example E(T2 )) is equal to that of the other connected component.
UΓ (α) = α3 α4 + α2 α4 + α2 α3 + α1 α3 + α1 α4 , (3.28)
VΓ (pext , α) = (pf1 + pf2 )2 α1 α2 (α3 + α4 ) + p2f4 α1 α3 α4 + p2f3 α2 α3 α4 .
Remark 3.4.2 The parametric representation (3.26) of a Feynman integral has a trivial
dependence of the space-time dimension D, which is now only a parameter. (This is of particular
importance for the implementation of the dimensional renormalization scheme, see Subsection
3.6.5).
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
26 QFT—built-in combinatorics
where j runs over the set of spanning trees and k over the set of the two-trees,
0 if the line belongs to the tree j
uj = (3.30)
1 otherwise
and
0 if the line belongs to the two-tree k
vk = (3.31)
1 otherwise.
One thus sees the integral representation of the heat kernel. This allows to establish
a connection between the scalar propagator and the Gaussian probability distribution
of a Brownian path going from x to y in time a. Moreover, this heat-kernel form
of the Euclidean propagator permits the Schwinger functions GN to satisfy the so-
called Osterwalder–Schrader axioms which allow the analytic continuation in Minkowski
space—see, for example, Vincent Rivasseau’s course (2012).
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
For the sake of completeness, we also mention that new relationships between
hypergeometric functions have recently been found by evaluating Feynman integrals in
Kniehl and Tarasov (2011).
28 QFT—built-in combinatorics
Z. Wang (2010a). We also refer the reader to the last section of this chapter for some
considerations on DSE, whose solutions also represent a non-perturbative QFT result.
The first step in a renormalization process is the regularization procedure, which
renders the Feynman integrals finite. The main idea of such a regularization procedure
is to introduce a regularization parameter, such that the divergences of the Feynman
integrals appear as singularities in this new parameter.
Several regularization schemes exist in the literature. Following again Hagen Kleinert
and Verena Schulte-Frohlinde (2001), we give here a short list of the main regularization
schemes:
• momentum space cut-off. The idea behind this procedure is to introduce a cut-
off Λ such that the integrations on |p| are carried out only until Λ. The Feynman
integrals are now convergent, but they are of course divergent in the limit Λ → ∞.
• Pauli–Villars regularization (1949). One changes the propagator in the following
way:
1 1 1
→ 2 − . (3.33)
p 2 + m2 p + m2 p 2 + M 2
The modified propagator decreases faster within the limit |p| → ∞; the role of the
momentum space cut-off Λ is now played by the parameter M .
• analytic regularization Speer (1969). One has the following change of propagator:
1 1
→ 2 , (3.34)
p 2 + m2 (p + m2 )−z
where z ∈ C with Re(z) being large enough to make the Feynman integrals
converge. The result is then analytically continued to a region around the physical
value z = 1—the divergences now appear as poles for z = 1.
• dimensional renormalization Hooft (1972, 1973). The idea behind this approach
is to allow the space-time dimension D in Feynman integrals to be a complex
number. This scheme is extensively used nowadays in elementary particle physics
computations. (Its main interest for physicists comes from the fact it naturally
preserves gauge invariance.) Moreover, it is interesting to mention here that a
concrete geometric meaning of this renormalization scheme can be exhibited (see
section 1.19 of Alain Connes and Matilde Marcolli (2008)).
The multi-scale analysis is more complex than all these, because it also takes into
account the energy scale on which the internal propagators reside. Thus, the propagators
are ‘sliced’. This is an appropriate tool for implementing the constructive renormal-
ization (see previous mention). For more details on the multi-scale analysis refer to
Subsection 3.6.3.
The Polchinski flow equation method (see, for example, Manfred Salmhofer
(1999)) is yet another way of performing renormalization.
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
It is, in our opinion, particularly interesting to mention that the Polchinski flow
equation idea is also used in combinatorics to prove a result of E. M. Wright’s which
expresses the generating function of connected graphs under certain conditions (fixed
excess). In this proof, one needs to choose an arbitrary edge in such a connected graph
and then remove it. Two possibilities for this chosen edge can then appear (see, for
example, Proposition II.6 of Philippe Flajolet and Robert Sedgewick (2008)), just as
in the QFT Polchinski renormalization.
Finally, let us also mention that such a QFT-inspired Polchinski flow equation
technique was recently used in a proof given by T. Krajewski for Postnikov’s hook length
formula (2012) (the Postnikov hook length formula gives the number obtained when one
sums over all plane binary trees of a given order n on the product over v of (1 + 1/hv ),
v being a vertex, and hv the hook length, i.e. the number of vertices below v ).
All the divergences of a renormalizable model come from insertion of these Feynman
graphs into ‘larger’ Feynman graphs. Furthermore, only a finite class of graphs should
be primitively divergent, if one wants to have a renormalizable model.
Let us now give the following definition:
Definition 3.6.2 The superficial degree of divergence ω , is the difference between the
total number of powers in internal momenta between the denominator and the nominator.
Thus, for the Feynman integral (3.24), after integration using the δ−function, one
has two independent momenta in the denominator (each of them being a four vector)
and four momenta, each of them squared) in the nominator. One thus has
ω = 4 × 2 − 2 × 4 = 0. (3.35)
The power counting theorem gives the superficial degree of divergence for a general
Feynman graph (of a given QFT model). For the Φ4 model, the power counting theorem
is the following:
ω = N − 4. (3.36)
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
30 QFT—built-in combinatorics
A direct way of proving this theorem is through a rescaling of the parameters α in the
parametric representation (3.26) of the Feynman integrand.
Example 3.6.4 The Φ4 graph of Fig. 2.1 has four external edges; thus, its superficial degree
of divergence is equal to 0. This corresponds to a logarithmic divergence in an ultraviolet
momentum space cut-off.
In order for a QFT model to be renormalizable, its superficial degree should not
depend on the internal structure (number of edges, loops etc.) of the graph.
Let us also give the following definition:
Remark 3.6.6 A primitively divergent graph is a superficially divergent graph with no sub-
divergences.
3.6.2 Locality
The principle of locality states that the counterterms one requires to subtract the
divergence of some Feynman integral are of the same type as the terms already presented
in the bare (non-renormalized) action. These terms are local, hence the name of locality.
This principle can be most easily understood when computing Feynman integrals in
position space (since space locality appears the most clearly in ... position space). Thus,
for the example of a one-loop four-point graph of Fig. 3.1, (graph with two vertices,
localized in two points x and y in position space), the Feynman integral is divergent
when integrating on the sector:
x ∼ y. (3.37)
The rest of the integration domain does not lead to a divergence. One thus needs to
subtract a counterterm which corresponds exactly to the local part of the Feynman
integral—a local counterterm. This can be illustrated as is done in Fig. 3.2.
x y
Figure 3.1 A four-point one-loop Feynman graph. It has two vertices, localized, in position space, in x
and y
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
x y x∼y
= + convergent.
Figure 3.2 Illustration of the local subtraction of the counterterm for the one-loop graph of Fig. 3.1.
The divergent part of the integral corresponds to the region of the integration domain where x ∼ y (the
locality principle)
More concretely, writing the Feynman integral to renormalize gives (up to irrelevant
normalization factors)
φ(Γ) = dxdyC(x, y)2 C(x, ȳ 1 )C(x, ȳ 2 )C(y, ȳ 3 )C(y, ȳ 4 )Φ(x)2 Φ(y)2 , (3.38)
where we have denoted the four external points that the external edges hook to by
ȳ 1 , . . . , ȳ 4 . Let us now leave aside for the moment the fields (for simplicity) and focus on
the propagators of the formula (3.38). As previously discussed, we will subtract a term
corresponding to the region x ∼ y . This can be achieved using the expansion formula
1
f (t) = f (0) + dtf (t). (3.39)
0
where we have completed the formula with the products of the corresponding external
fields. One can prove that the second term in the right-hand side of (3.40) is finite. The
first term in the sum correspond to the divergent part of the integral—it will be subtracted
(the counterterm),
4
τΓ φ(Γ) = dx (Φ(x)) dyC(x, y)2 . (3.41)
This term is of the desired local-like form of the term present in the action. This term
can thus be reabsorbed in a redefinition of the coupling constant λ).
The same type of redefinition of the other ‘constants’ appearing in the action can be
done in the case of two-point graphs.
OUP CORRECTED PROOF – FINAL, 16/2/2021, SPi
32 QFT—built-in combinatorics
For a more general Feynman graph (not necessary a one-loop one), the locality of
the counterterms can be seen as sending all the external edges to the same point (again, in
position space).
We denote by Γ/γ the cograph obtained by shrinking the (divergent) subgraph γ
inside the Feynman graph Γ:
Remark 3.6.8 When shrinking a two-point subgraph, one should take into consideration
whether the respective divergence contributes to the renormalization of the mass or of the wave
function (the two parameters corresponding to the two terms in the quadratic part of the action).
Nevertheless, this does not play an important part at a combinatorial level.
p
C(x − y) = Ci (x − y), (3.43)
i=0
where
M −2(i−1)
(x−y)2 da
e−m
2
Ci (x − y) = a− 4a , (3.44)
M −2i aD/2
and
∞
(x−y)2 da
e−m
2
C0 (x − y) = a− 4a . (3.45)
1 aD/2
Another random document with
no related content on Scribd:
To see these friends once again Hilary had no time to waste. As he
made his way along the sandy road with the stars palpitating whitely
in the sky above the heavy forest, which rose so high on either hand
as to seem almost to touch them, this deep, narrow passage looked
when the perspective held a straight line to rising ground, ending in
the sidereal coruscations, like the veritable way to the stars, sought
by every ambitious wight since the days of the Cæsars. Hilary had
never heard an allusion to that royal road, but as he walked along
with a buoyant, steady step, his hat in his hand that the breeze might
cool his hot brow and blow backward his long masses of fair hair, he
followed indeed an upward path in the sentiments that quickened his
pulses, for he was resolved upon duty and thinking high thoughts
that should materialize in fine deeds. He was to do and dare! He
would be useful and faithful and brave—brave! He had a reverence
for the quality of courage—not for the sake of its emulous display,
but for the spirit of all nobly valiant deeds. He had rejoiced in the
very expression of the captain’s eyes—so true and tried! He, too,
would meet the coming years fairly. The raw recruit could see his
way to the stars at the end of that mountain vista.
But it seemed a poor preparation for all this when he awoke the
inmates of the Noakes cabin, for it was past midnight, with the news
that he had “jined the cavalry” and was to march at peep of dawn
with Bertley’s squadron. It is true that the elders crowded around him
half dressed only, so hastily had they been roused, and expressed
surprise, congratulations, and regrets in one inconsistent breath, and
old Mrs. Dite, Delia’s grandmother, bestowed on him a woolen
comforter which she had knitted for him, and for which,
improvidently, it being now near summer, he cared less than for the
turmoil of excitement and interest they had manifested in his
preferment, for he felt every inch a man and a soldier, and they
respectfully seemed to defer to his new pretensions. Delia, however,
the most unaccountable of girls—and girls are always unaccountable
—put her arm over her eyes as she stood beside the mantelpiece,
beneath which the embers had been stirred into a blaze for light, and
turned her face to the wall and burst into tears. She wept with so
much vehemence that her long plait of black hair hanging down her
back swayed from side to side of her shoulders as she shook her
head to and fro in the extremity of her woe. When the elders
remonstrated with her, and declared this was no occasion for sorrow,
she only lifted her tear-stained face for a moment to say in
justification that she believed that bullets were too small to be
dodged with any success when they were flying round
promiscuously. And in the midst of the volley of laughter which this
evoked from the old people, Hilary’s voice rang out indignantly, “An’ I
ain’t no hand ter dodge bullets, nuther.”
“That’s jes’ what I am a-crying about,” replied Delia, to the
mischievous delight of the elders.
Thus the farewell to his old friends was not very exhilarating to
Hilary. Delia did not even at the last unveil her face or change her
attitude against the wall. To shake hands he was obliged to pull her
hand from her eyes by way of over her head, and in this maneuver
he was moved to notice how much taller he had lately grown. Her
hand was very limp and cold and wet with her tears—so wet that he
had to wipe these tears from his own hand on the brim of his hat on
his homeward way.
And when, as in sudden enchantment, darkness became light and
night developed into dawn, when color renewed the landscape, and
the dull sky grew red as if flushed with sudden triumph, and the black
mountains turned royally purple in the distance and tenderly green
nearer at hand, and the waters of the river leaped and flashed like a
live thing, as with an actual joy in existence, and the great fiery sun,
full of vital yellow flame, flared up over the eastern horizon, the
squadron, with jingling spurs and clanking sabers, with carbines and
canteens rattling, with the trumpet now and again sending forth
those elated, joyous martial strains, so sweet and yet so proud, rode
forth into a new day, and Hilary Knox, among the troopers and
gallantly mounted, rode forth into a new life.
The bivouac fires glowed for a while, then fell to smoldering and
died, leaving but a gray ash to tell of their presence here. Day by day
the eagles in the great bare pine tree on the high rock at the summit
of the mountain looked for Hilary to visit his point of observation and
stir their hearts with fear and wrath. Time and again the male bird
might have been seen to circle about at the usual hour for the boy’s
coming; first with apprehension lest his absence was too good to be
true, then, with the courage of immunity undisciplined by fear,
screaming and flouncing as if to challenge this apparition of
quondam terror. Now and then the pair seemed to argue and
collogue together upon the mystery of his non-appearance, and to
chuffily compare notes, and seek to classify their impressions of this
singular specimen of the animal kingdom. Perhaps, tabulated, their
conclusions might stand thus: Genus, boy; habits, noisy; diet,
omnivorous; element, mischief; uses, undiscovered and
undiscoverable.
Long, long after the eagles had forgotten the intruder, after their
brood, the two ill-feathered nestlings, had taken strongly to wing,
after their nest, a mass of loose, but well collocated sticks and grass,
had given way to the beat of the rain and the blasts of the wind, did
Hilary’s mother wearily gaze from the heights where the mountain
cabin was perched down upon the curves of the valley road along
which she had seen him riding away with that glittering train, and
sigh and let her knitting fall from her nerveless hands, and wonder
what would the manner of his home-coming be, or whether the future
held at all a home-coming for him.
And her many sighs kept her heart sick and turned her hair very
white.
CHAPTER II
It was a wonderful period of mental development for this wild young
creature of the woods, when Hilary received in his sudden transition
to the “valley kentry” his first adequate impressions of civilization. He
learned that the world is wide; he beheld the triumphs of military
science; he acquiesced in the fixed distinctions of rank, since he
must needs concede the finer grades of capacity. But courage, the
inherent, inimitable endowment, he recognized as the soul of
heroism, and in all the arrogance of elation he became conscious
that he possessed it. This it was that opened his stolid mind to the
allurements of ambition. He rejoiced in an aspiration.
He was brave. That was his identity—his essential vitality! Was he
ignorant, poor, the butt of the campfire jokes, because of his
simplicity in the wide world’s ways, slothful, slow, wild, and turbulent?
He took heed of none of this! He was the bravest of the brave—and
all the command knew it!
With an exultant heart he realized that Captain Bertley was aware of
the fact, and often took account of it in laying his plans. The regiment
of which this squadron was a part belonged to one of those brigades
of light cavalry whose utility was chiefly in quick movements, in
harassing an enemy’s march, in following up and hanging on his
retreat, and sometimes in making swift forced marches, appearing
unexpectedly in distant localities far from the main body and adding
the element of surprise to a sudden and furious onslaught. Often
Hilary was among a few picked men sent out to reconnoiter, or as
the rear-guard when the little band was retreating before a superior
force and it was necessary to fight and flee alternately. It was now
and again in these skirmishes that he had the opportunity to show
his pluck and his strength and his cool head and his ready hand.
More than once he had been the bearer of dispatches of great
importance sent by him alone, disguised in citizen’s dress and his
destination a long way off. Thus did the captain commanding the
squadron demonstrate his confidence in the boy’s fidelity and
courage and resource. For his ready wit in an emergency was hardly
less than his courage.
“What did you do, then, with the Colonel’s letter that you were to
deliver at brigade head-quarters?” asked the Captain in much
agitation, but with a voice like thunder and a flashing eye, when one
day Hilary returned from a fruitless expedition, with his finger in his
mouth, so to speak, and a tale of having encountered Federal
scouts, who had stopped and questioned him, and finally after
suspiciously searching him, had turned him loose, believing him
nothing more than he seemed—a peaceful, ignorant country boy.
Hilary glanced ruefully down at the hat that he swung in his hand,
then with anxious deprecation at the Captain, whose face as he
stood beside his horse, ready to mount, had flushed deeply red,
either because of the reflection of the sunset clouds massed in the
west or because of the recollection that he had earnestly
recommended the boy to his superior officer, for this dangerous
mission, and thus felt peculiarly responsible; for the letter had
contained details relating to the Colonel’s orders from brigade
headquarters, his numbers, and other matters, the knowledge of
which in the enemy’s hands might precipitate his capture, together
with all the detachment.
“It’s gone, sir,” mumbled Hilary, the picture of despair; “I never
knowed what ter do, so—”
“So you let them have that letter—when I had told you how important
it was!”
“I don’t see how it could have been helped, since the boy was
searched,” said Captain Blake, the junior captain of the squadron,
who was standing by. “I am glad he came back to let us know.”
“That’s why I done what I done,” eagerly explained Hilary. “I—I—eat
it.”
“All of it?” cried Captain Bertley, with a flash of relief.
“Yes, sir, I swallowed it all bodaciously—just ez soon ez I seen ’em
a-kemin’ dustin’ along the road.”
“Well done, Baby Bunting!” cried the senior officer, for thus was
Hilary distinguished among the troopers on account of his tender
years.
The gruff Captain Blake laughed delightedly, a hoarse, discordant
demonstration, much like the chuckling of a rusty old crow. He
seemed to think it a good joke, and Hilary knew that he, too, was
vastly relieved to have saved from the enemy such important
information.
“Pretty bitter pill, eh?”
“Naw, sir,” said Hilary, his eyes twinkling as he swung his hat in his
hand, for he could never be truly military out of his uniform; “it war
like eatin’ a yard medjure of mustard plaster, bein’ stiff ter swaller an’
somehow goin’ agin the grain.”
The senior captain gravely commended his presence of mind, and
said he would remember this and his many other good services. As
he dismissed the young trooper and still standing, holding a sheet of
paper against his saddle, began to write a report of the fate of the
letter that had so threatened the capture of the whole command,
Hilary overheard Captain Blake say in his bluff, extravagant way,
“That boy ought to be promoted.”
“What?” said Captain Bertley, glancing back over his shoulder with
the pencil in his hand. “Baby Bunting with a command!”
Despite the ridicule of the idea Hilary’s heart swelled within him as
he strolled away, for he cared only to deserve the promotion and the
confidence shown him, even if on account of his extreme youth and
presumable irresponsibility he was debarred from receiving it.
He could not have said why he was not resentful of being called
“Baby Bunting” by Captain Bertley. He felt it was in the nature of a
courteous condescension that the officer should comment on the
inadequacy of his age and the discrepancy between his limited
powers and his valuable deeds—almost as a jesting token of
affection, kindly meant and kindly received. But the name fell upon
his ear often with a far different significance; the camp cry “Bye, oh,
Baby Bunting,” was intended to goad him to such a degree of anger
as should make him the sport of the groups around the bivouac fire.
The chief instigator of this effort was a big, brutal cavalryman, by
name Jack Bixby. He had a long, red beard; long, reddish hair; small,
twinkling, dark eyes, and a powerfully built, sinewy, well-compacted
figure. He was superficially considered jolly and genial, for few of his
careless companions were observant enough of moral phenomena
or sufficiently students of human nature to take note of the fact that
there was always a spice of ill-humor in his mirth. Malice or jealousy
or grudging or a mean spirit of derision pervaded his merriment. He
found great joy in ridiculing a raw country boy, whose lack of
knowledge of the world’s ways laid him liable to many mistakes and
misconceptions, and at first Hilary’s credulity in the big lies told him
by Jack Bixby and his simplicity in acting upon them exposed him to
the laughter of the whole troop. This was checked in one instance,
however; having been instructed that it was an accepted detail of the
observances of a soldier, Hilary was induced to advance with great
ceremony one day, and duly saluting ask Captain Bertley how he
found his health. The officer was standing on ground somewhat
elevated above the site of the camp, in full gray uniform, a field-glass
in his hand, his splendid charger at his shoulder, the reins thrown
over his arm. The humble “Baby Bunting” approaching this august
military object, and presuming to ask after the commanding officer’s
health, was in full view of a hundred or more startled and amazed
veterans.
But Captain Bertley had seen and known much of this world and its
ways. He instantly recognized the incident as a bit of malicious play
upon the simplicity of the new recruit, and he took due note, too, of
his own dignity. He realized how to balk the one and to support the
other. He accepted the unusual and absurd demonstration
concerning his health by saying simply that he was quite well, and
then he kept the boy standing in conversation as to the state of a
certain ford some distance up the river, with which Hilary was
acquainted, having been of a scouting party which had been sent in
that direction the previous day. The staring military spectators, their
attention previously bespoken by Bixby, saw naught especial in the
interview, the boy apparently having been summoned thither by
order of the officer to make a report or give information, and thus the
joke, attenuated to microscopic proportions, failed of effect. It had,
however, very sufficient efficacy in recoil. Before dismissing Hilary
the Captain asked how he had chanced to accost him in the manner
with which he had approached him, and the boy in guilelessly
detailing the circumstance, before he was admonished as to his
credulous folly, betrayed Bixby as the perpetrator of the pleasantry at
his expense, and what was far more serious at the expense of the
officer. Jack Bixby, dull enough, as malicious people often are, or
they would not otherwise let their malice appear—for they are not
frank—did not see it in that light until he suddenly found himself
under arrest and then required to mount the “wooden horse” for
several weary hours.
“You’ll be hung up by the thumbs next time, my rooster,” said the
sergeant, as he carried the sentence into effect. “The Cap’n ain’t so
mighty partial to your record, no hows. He asked me if you hadn’t
served with Whingan’s rangers, ez be no better’n bushwhackers, an’
ye know he is mighty partic’lar ’bout keepin’ up the tone an’ spirit o’
the men.”
Hilary, contradictorily enough, lost all sense of injury and shame in
sorrow that he should have divulged Bixby’s agency in the matter
and brought this disaster upon the trooper, who perhaps had only
intended a little diversion, and had neither the good taste nor the
good sense to perceive its offensiveness to the officer. Bixby had
served in a band generally reputed bushwhackers, who did little
more than plunder both sides, and in which discipline was
necessarily slight. And thus after this episode they were better
friends than before. True, in the days of dearth, for these men must
needs starve as well as fight, when only rations of corn were served
out, which the soldiers parched and ate by the fire, and which were
so scanty that a strict watch was kept to prevent certain of them from
robbing their own horses, on the condition and speed of which their
very lives depended, Hilary, as in honor bound, being detailed for
this duty, reported his greedy comrade, but in view of the half-
famished condition of the troops Bixby’s punishment was light, and
the incident did not break off their outward semblance of friendship,
although one may be sure Bixby kept account of it.
So the years went—those wild years of hard riding and hard fighting;
sleeping on the ground under the open skies whether cloudy or clear
—it was months after it was all over before Hilary could accustom
himself to sleep in a bed; roused by the note of the trumpet,
sometimes while the stars were yet white in the dark heavens, with
no token of dawn save a great translucent, tremulous planet
heralding the morn, and that wild, sweet, exultant strain of reveille,
so romantic, so stirring, that it might seem as if it had floated down,
proclaiming the day, from that splendid vanguard of the sun. So they
went—those wild years, all at once over.
The end came on a hard-contested field, albeit only a thousand or so
were engaged on either side. The squadron, in one of those wild
reckless assaults of cavalry against artillery, for which the
Confederate horse were famous in this campaign, had gone to the
attack straight up a hill, while the muzzles of the big, black guns sent
forth smoke and roar, scarcely less frightful than the bombs which
were bursting among the horses and men riding directly at the
battery. It was hard to hold the horses. Often they swerved and
faltered, and sought to turn back. Each time Captain Bertley, with
drawn sword, reformed the line, encouraging the men and urging
them to the almost impossible task anew. At it they went once more,
in face of shot and shell. Now and again Hilary, riding in the rear
rank, with his saber at “the raise,” heard a sharp, singing sibilance,
which he knew was a minie-ball, whizzing close to his ear, and he
realized that infantry was there a little to one side supporting the
battery. The rush, the turmoil, the blare of the trumpets sounding “the
charge,” the clamor of galloping hoofs, of shouting men, the roar of
cannon, the swift panorama of moving objects before the eye, the
ever-quickening speed, and the tremendous sensation of flying
through the air like a projectile—it was all like some wild tempest,
some mad conflict of the elements. And suddenly Hilary became
aware that he was flying through the air without any will of his own.
The horse had taken the bit between his teeth, and maddened by the
noise, the frenzy of the fight, was rushing on he knew not whither, his
head stretched out, his eyes starting, straight up the hill unmindful of
the trumpet now sounding the recall and the heavy pull of the boy on
the curb. Hilary was far away in advance of the others when the line
wheeled. A few more impetuous bounds and plunges, and he was
carried in among the Federal guns, mechanically slashing at the
gunners with his saber, until one of the men, with a well-directed
blow, knocked him off his horse with the long, heavy sponge-staff.
So it was that Hilary was captured. He surrendered to the man with
the sponge-staff, for the others were busily limbering up the guns;
they were to take position on a new site—one less exposed to attack
and very commanding. They had more than they wanted in Hilary.
He realized that as he was on his way to the rear under guard. The
engagement was practically at an end, and the successful Federals
were keenly eager to pursue the retreating force and secure all the
fruits of victory. To be hampered with the disposition of prisoners at
such a moment was hardly wise, when an active pursuit might cut off
the whole command. Therefore the few already taken, who were
more or less wounded, were temporarily paroled in a neighboring
hamlet, and Hilary, the war in effect concluded for him—for the
parole was a pledge to remain within the lines and report at stated
intervals to the party granting it—found himself looking out over a
broad white turnpike in a flat country, down which a cloud of dust
was all that could be seen of the body of cavalry so lately contending
for every inch of ground.
Now and again a series of white puffs of smoke from amidst the
hillocks on the west told that the battery of the Federals was shelling
the woods which their enemy had succeeded in gaining, the shells
hurtling high above the heads of their own infantry marching forward
resolutely, secure in the fact of being too close for damage.
Presently the battery became silent. Their vanguard was getting
within range of their own guns, and a second move was in order. The
boy watched the flying artillery scurrying across the plain, as he
struck down a “dirt-road” which intersected the turnpike, and soon he
noticed the puffs of white smoke from another coign of vantage and
the bursting of shells still further away.
“Them dogs barkin’ again! Waal, I’m glad ter be wide o’ thar mark,”
said a familiar voice at his elbow; the speaker was Bixby, a paroled
prisoner, too, having been captured further down the hill during the
general retreat.
Hilary was not ill-pleased to see him at first, especially as something
presently happened which made him solicitous for the advice and
guidance of an older head than his own. By one of the vicissitudes of
war victory suddenly deserted the winning side, and presently here
was the erstwhile successful party in full retreat, swarming over the
flat country, the battery scurrying along the turnpike with two of its
guns missing, captured as they barked with their mouths wide open,
so to speak. The hurrying crash and noisy rout went past like the
phantasmagoria of a dream, and these two prisoners were presently
left quite outside the Federal lines by no act or volition of their own,
and yet apparently far enough from Bertley’s squadron, for the
pursuit was not pressed, both parties having had for the nonce
enough of each other. The first object of the two troopers was to
procure food of which they stood sadly in need. They set forth to find
the nearest farmhouse, Hilary on his own horse, which in the
confusion had not been taken from him when he was disarmed, and
Bixby easily caught and mounted a riderless steed that had been in
the engagement, but was now cropping the wayside grass.
A thousand times that day Hilary wished, as they went on their
journey together, that he had never seen this man again. All Jack
Bixby’s methods were false, and it revolted Hilary, educated to a
simple but strict code of morals, to seem to share in his lies and his
dubious devices to avoid giving a true account of themselves. In fact
their progress was menaced with some danger. Having little to
distinguish them as soldiers, for the gray cloth uniform in many
instances had given place to the butternut jeans, the habitual garb of
the poorer classes of the country, they could be mistaken for
citizens, peacefully pursuing some rustic vocation, and this
impression Bixby sought to impose on every party who questioned
them. He feared to meet the Federals, because of their paroles,
which showed them to be prisoners and yet out of the lines, and he
thought this broken pledge might subject them to the penalty of
being strung up by the neck.
“That air tale ’bout our bein’ in the lines an’ the lines shrinkin’ till we
got out o’ ’em ain’t goin’ ter go down with no sech brash fellers,” he
argued with some reason, for the probabilities seemed against them.
And now he dreaded an encounter with Union men, non-combatants,
for the same reason. He slipped off his boot at one time and hid the
paper under the sole of his foot. “Ef we-uns war ter be sarched they
wouldn’t look thar, mos’ likely.” And finally when they reached the
house of an aged farmer, who with partisan cordiality welcomed and
fed them, declaring that although he was too old to fight he could
thus help on the southern cause, Bixby took advantage of his host’s
short absence from the dining-room to strike a match which he
discovered in a candlestick on the mantel piece, for the season was
too warm for fires, and lighting the candle he held the parole in the
flame till the paper was reduced to a cinder; then he hastily
extinguished the candle.
When once more on the road, however, Bixby regretted his decision.
For aught he knew they were still within the Federal lines. The Union
troops had doubtless been reinforced, for they were making a point
of holding this region at all hazards. He was a fool he said to have
burnt his parole—it was his protection. If he were taken now by
troops not in the extreme activities of resisting a spirited cavalry
attack, who had time to make his capture good, and means of
transportation handy, he would be sent off to Camp Chase or some
other prison, and shut up there till the crack of doom, whereas his
parole rendered him for the time practically free.
“Why didn’t you keep me from doin’ it, Hil’ry?”
“Why, I baiged an’ baiged an’ besought ye ’fore we went in the house
ter do nothin’ ter the paper,” said Hilary, wearied and excited and
even alarmed by his companion’s vacillations, so wild with fear had
Bixby become. “I wunk at ye when the old man’s back was turned. I
even tried ter snatch the paper whenst ye put yer boot-toe on the
aidge of a piece of it on the ha’thstone an’ helt it down till it war
bu’nt.”
“I war a fool,” said Bixby, gloomily. “I wish I hed it hyar now.”
“I tole ye,” said Hilary, for he had spent the day in urging the fair and
open policy, let come what might of it, “I tole ye ez I war a-goin’ ter
show my parole ter the fust man ez halts me, an’ ef I be out’n the
lines, an’ he won’t believe my tale, let him take it out on me
howsumdever the law o’ sech doin’s ’pears. Nobody could expec’ me
ter set an’ starve on that hillside till sech time ez the Fed’rals throw
out thar line agin.”
“I wisht I hed my parole agin,” said Bixby, more moodily still.
Down the road before them suddenly they saw a dust, and a steely
glitter—not so strong a reflection, however, as marching infantry
throws out. A squad of cavalry was approaching at a steady pace.
Jack Bixby’s first idea was flight; this the condition of the jaded horse
rendered impolitic. Then he thought of concealment—in vain. On
either hand the level, plowed fields afforded not the slightest bush as
a shield. The only thicket in sight was alongside the road and now in
line with the approaching party whom it so shadowed that it was
impossible to judge by uniform or accoutrements to which army they
belonged.
“Hil’ry,” said Jack Bixby, “let’s stick ter the country-jake story; I’ll say
that I be a farmer round hyar somewhar, an’ pretend that you air my
son. That’ll go down with any party.”
“I be goin’ ter tell the truth myself, an’ show my parole, whoever they
be; that’s the right thing,” said Hilary, stoutly.
“But I ain’t got no parole,” quavered Bixby.
“Tell the truth an’ I’ll bear ye out,” said Hilary. “Tell ’em that thar be so
many parties—Feds an’ Confeds an’ Union men an’ bushwhackers,
an’ we-uns got by accident out’n the lines an’ ye took alarm an’
deestroyed yer parole. I’ll bear ye out an’ take my oath on it; an’ ye
know the old man war remarkin’ on them cinders on the aidge o’ the
mantel shelf an’ ha’thstone ez we left the house.”
“Hil’ry,” said Bixby, as with a sudden bright idea—anything but the
truth seemed hopeful to him—“I’ll tell ye. I’ll take yer parole an’ claim
it ez mine, an’ pretend that ye air my son—non-combatant, jes’ a
boy, ez ye air.”
“But it’s got my name on it. It’s a-parolin’ of me,” said Hilary, “an’ I
ain’t no non-combatant.”
“But I’ll claim your name; I’ll be Hil’ry Knox, an’ tall ez ye air, yer face
shows ye ain’t nuthin’ but a boy. Nobody wouldn’t disbelieve it.”
“I won’t do it! I won’t put off a lie on ’em! I hev fought an’ fought an’
I’ll take the consekences o’ what I done—all the consekences o’
hevin’ fought. I am Hilary Knox, an’ I be plumb pledged by my word
of honor. But I’ll bear ye out in the fac’s, an’ thar’s nuthin’ ter doubt in
the fac’s—they air full reasonable.”
He had taken the paper out of his ragged breast-pocket to have it in
readiness to present to the advance guard, who had perceived them
and had quickened the pace for the purpose of halting them.
Perhaps Bixby had no intention, save, by sleight-of-hand, to possess
himself of the paper. Perhaps he thought that having it in his power
the boy would hardly dare to contradict the story he had sketched
and the name he intended to claim as the owner of the parole; if
Hilary should protest he could say his son was weak-minded, an
imbecile, a lunatic. He made a sudden lunge from the saddle and a
more sudden snatch at the paper. But the boy’s strong hand held it
fast. Jack Bixby hardly noted the surprise, the indignation, the
reproach in Hilary’s face—almost an expression of grief—as he
turned it toward him. With the determination that had seized him to
possess the paper, Bixby struck the boy’s wrist and knuckles a
series of sharp, brutal blows with the back of a strong bowie-knife,
which had been concealed in his boot-leg at the surrender. They
palsied the clutch of the boy’s left hand. But as the quivering fingers
opened, Hilary caught the falling paper with his right hand.
“Let go, let go!” cried Jack Bixby in a frenzy; “else I’ll let you hev the
blade—there, then!—take the aidge—ez keen ez a razor!”
The steel descended again and again, and as the boy was half
dragged out of the saddle the blood poured down upon the parole. It
would have been hard to say then what name was there!
A sudden shout rang out from down the road. The approaching men
had observed the altercation, and mending their pace, came on at a
swift gallop.
With not a glance at them, Jack Bixby turned his horse short around
and fled as fast as the animal could go, striking out of the road and
into the woods as soon as he reached the timbered land.
Poor “Baby Bunting,” dragged out of his saddle, fell down in the road
beneath his horse’s hoofs, and all covered with white dust and red
blood there he lay very still till the cavalrymen came up and found
him.
For this was what they called him—“Poor Baby Bunting!” They were
a small reconnoitering party of his own comrades, and it was with a
hearty good will that they pursued Jack Bixby who fled, as from his
enemies, through the brush. Perhaps his enemies would have been
gentler with him than his quondam friends could they only have laid
hands on him, for they all loved “Baby Bunting” for his brave spirit
and his little simplicities and his hearty good-comradeship. Hilary
recognized none of them. He only had a vague idea of Captain
Bertley’s face with a grave anxiety and a deep pity upon it as the
officer gazed down at him when he was borne past on the stretcher
to the field hospital where his right arm was taken off by the surgeon.
He was treated as kindly as possible, for the remembrance of his
gallant spirit as well as humanity’s sake, and when at last he was
discharged from the more permanent hospital to which he had been
removed he realized that he had indeed done with war and fine
deeds of valiance, and he set out to return home, tramping the weary
way to the mountain and his mother.
After that fateful day, when maimed and wan and woebegone he
came forth from the hospital and journeyed out from among the
camps and flags and big guns and all the armaments of war, thrice
splendid to his backward gaze, it seemed to him that he had left
there more than was visible—that noble identity of valor for which he
had revered himself.
For he found as he went a strange quaking in his heart. It was an
alien thing, and he strove to repudiate it, and ached with helpless
despair. When he came into unfamiliar regions, and a sudden clatter
upon the lonely country road would herald the approach of mounted
strangers, halting him, the convulsive start of his maimed right arm
with the instinct to seize his weapons and the sense of being
defenseless utterly would so unnerve him that he would give a
disjointed account of himself, with hang-dog look and faltering words.
And more than once he was seized and roughly handled and
dragged to headquarters to show his papers and be at last passed
on by the authorities.
He began to say to himself that his courage was in his cavalry pistol.
“Before God!” he cried, “me an’ my right arm an’ my weepon air like
saltpetre an’ charcoal an’ sulphur—no ’count apart. An’ tergether
they mean gunpowder!”
And doubly bereaved, he had come in sight of home.
But his mother fell upon his neck with joy, and the neighbors
gathered to meet him. The splendors of the Indian summer were
deepening upon the mountains, with gorgeous fantasies of color,
with errant winds harping æolian numbers in the pines, with a
translucent purple haze and a great red sun, and the hunter’s moon,
most luminous. The solemnity and peace stole in upon his heart, and
revived within him that cherished sense of home, so potent with the
mountaineer, and in some wise he was consoled.
Yet he hardly paused. In this lighter mood he went on to the
settlement, eager that the news of his coming should not precede
him.
There was the bridge to cross and the rocky ascent, and at the
summit stood the first log cabin of the scattered little hamlet. From
the porch, overgrown with hop vines, he heard the whir of a spinning
wheel. He saw the girl who stood beside it before she noticed the
sound of his step. Then she turned, staring at him with startled
recognition, despite all the changes wrought in the past two years. “It
air me,” he said, jocosely.
From his hollow eyes and sunken cheeks and wan smile her gaze
fell upon his empty sleeve. She suddenly threw her arm across her
face. “I—I—can’t abide ter look at ye!” she faltered, with a gush of
tears.
He stood dumfounded for a moment.
“Durn it!” he cried. “I can’t abide ter look at myself!”
And with a bitter laugh he turned on his heel.
He would not be reconciled later. The wound she had unintentionally
dealt him rankled long. He said Delia Noakes was a sensible girl.
Plenty of brave fellows would come home from the war, hale and
hearty and with two good arms, better men in every way, in mind and
body and heart and soul, for the stern experiences they were
enduring so stanchly. The crop of sweethearts promised to be
indeed particularly fine, and there was no use in wasting politeness
on a fellow with whom she used to play before either of them could
walk, but whose arm was gone now, through no glorious deed
wrought for his country, for which he had intended to do all such
service as a man’s right arm might compass, but because he was a
fool, and had made a friend of a malevolent scoundrel, who had
nearly taken his life, but had only—worse luck—taken his right arm!
And besides he had seen enough of the world in his wanderings to
know that it behooves people to look to the future and means of
support. He had learned what it was to be hungry, he had learned
what it was to lack. He was no longer the brave and warlike man-at-
arms, “Baby Bunting.” He had no vocation, no possibility of a future
of usefulness; he could not hold a gun or a plow or an ax, and Delia
doubtless thought he would not be able to provide for her. And “dead
shot” though he had been he could not now defend himself, he
declared bitterly, much less her.
CHAPTER III
It was the last month of the year, and the month was waning. The
winds had rifled the woods and the sere leaves all had fallen. Yet still
a bright after-thought of the autumnal sunshine glowed along the
mountain spurs, for the tardy winter loitered on the way, and the
silver rime that lay on the black frost-grapes melted at a beam.
“The weather hev been powerful onseasonable an’ onreasonable, ter
my mind,” said old Jonas Scruggs, accepting a rickety chair in his
neighbor’s porch. “’Tain’t healthy.”
“Waal, ’tain’t goin’ ter last,” rejoined Mrs. Knox, from the doorway,
where she sat with her knitting. “’Twar jes’ ter-day I seen my old gray
cat run up that thar saplin’ an’ hang by her claws with her head
down’ards. An’ I hev always knowed ez that air a sure sign of a
change.”
Presently she added, “The fire air treadin’ snow now.”
She glanced over her shoulder at the deep chimney-place, where a
dull wood fire was sputtering fitfully with a sound that suggested
footfalls crunching on a crust of snow.
“I dunno ez I need be a-hankerin’ fur a change in the weather,
cornsiderin’ the rheumatiz in my shoulder ez I kerried around with
me ez a constancy las’ winter,” remarked Jonas Scruggs, pre-
empting a grievance in any event.
“Thar’s the wild geese a-sailin’ south,” Hilary said, in a low,
melancholy drawl, as he smoked his pipe, lounging idly on the step
of the porch.
His mother laid her knitting in her lap and gazed over her spectacles
into the concave vault of the sky, so vast as seen from the vantage
ground of the little log cabin on the mountain’s brow. Bending to the
dark, wooded ranges encircling the horizon, it seemed of a
crystalline transparency and of wonderful gradations of color. The
broad blue stretches overhead merged into a delicate green of
exquisite purity, and thence issued a suffusion of the faintest saffron
in which flakes of orange burned like living fire. A jutting spur
intercepted the sight of the sinking sun, and with its dazzling disk
thus screened, upon the brilliant west might be descried the familiar
microscopic angle speeding toward the south. A vague clamor
floated downward.
“Them fowels, sure enough!” she said. “Sence I war a gal I hev
knowed ’em by thar flyin’ always in that thar peaked p’int.”
“They keep thar alignment ez reg’lar,” said her son suddenly, “jes’
like we-uns hev ter do in the army. They hev actially got thar
markers. Look at ’em dress thar ranks! An’ thar’s even a sergeant-
major standin’ out ez stiff an’ percise—see him! Thar! Column
forward! Guide left! March!” he cried delightedly.
“I ’lowed, Hilary, ez ye bed in an’ about bed enough o’ the army,” said
the guest, bluntly.
Hilary’s face changed. But for some such reminder he sometimes
forgot that missing right hand. He made no answer, his moody eyes
fastened on that aërial marshaling along the vast plain of the sunset.
His right arm was gone, and the stump dangled helplessly with its
superfluity of brown jeans sleeve bound about it.
“Now that air a true word!” exclaimed Mrs. Knox, “only Hil’ry won’t
hev it so. I ’lows ter him ez he los’ his arm through jinin’ the Confed’
army, an’ he ’lows ’twar gittin’ in a fight with one o’ his own
comrades.”
Jonas Scruggs glanced keenly at her from under his bushy, grizzled
eyebrows, his lips solemnly puckered, and his stubbly pointed chin
resting on his knotty hands, which were clasped upon his stout stick.
He had the dispassionate, pondering aspect of an umpire, which
seemed to invite the cheerful submission of differences.
“Ye knows I war fur the Union, an’ so war his dad,” she continued.
“My old man had been ailin’ ennyhows, but this hyar talk o’ bustin’ up
the Union—why, it jes’ fairly harried him inter his grave. An’ I ’lowed
ez Hil’ry would be fur the Union, too, like everybody in the mountings
ez hed good sense. But when a critter-company o’ Confeds rid up
the mounting one day Hil’ry he talked with some of ’em, an’ he war
stubborn ever after. An’ so he jined the critter-company.”
She fell suddenly silent, and taking up her needles knitted a row or
two, her absorbed eyes, kindling with retrospection, fixed on the far
horizon, for Mrs. Knox was in a position to enjoy the melancholy
pleasures of a true prophet of evil, and although she had never
specifically forewarned Hilary of the precise nature of the disaster
that had ensued upon his enlistment, she had sought to defer and
prevent it, and at last had consented only because she felt she must.
She had her own secret satisfaction that the result was no worse; it
lacked much of the ghastly horrors that she had foreboded—death
itself, or the terrible uncertainty of hoping against hope, and fearing
the uttermost dread that must needs abide with those to whom the
“missing” are dear. Never now could the fact be worse, and thus she
could reconcile herself, and talk of it with a certain relish of finality, as
of a chapter of intense and painful interest but closed forever.
The old man nodded his head with deliberative gravity until she
recommenced, when he relapsed into motionless attention.
“An’ Hil’ry fought in a heap o’ battles, and got shot a time or two, an’
war laid up in the horspital, an’ kem out cured, an’ fought agin. An’
one day he got inter a quar’l with one o’ his bes’ frien’s. They war jes’
funnin’ afust, an’ Hil’ry hit him harder’n he liked, an’ he got mad, an’
bein’ a horseback he kicked Hil’ry. An’ Hil’ry jumped on him ez
suddint ez a painter, ter pull him out’n his saddle an’ drub him. Hil’ry
never drawed no shootin’ irons nor nuthin’, an’ warn’t expectin’ ter
hurt him serious. But this hyar Jack Bixby he war full o’ liquor an’
fury; he started his horse a-gallopin’, an’ ez Hil’ry hung on ter the
saddle he drawed his bowie-knife an’ slashed Hil’ry’s arm ez war
holdin’ ter him agin an’ agin, till they war both soakin’ in blood, an’ at
last Hil’ry drapped. An’ the arm fevered, an’ the surgeon tuk it off. An’
so Hil’ry hed his discharge gin him, sence the Confeds hed no mo’
use fur him. An’ he walked home, two hunderd mile, he say.”