Professional Documents
Culture Documents
Enumerative Combinatorics Volume 2 Second Edition Richard P Stanley 3 full chapter pdf docx
Enumerative Combinatorics Volume 2 Second Edition Richard P Stanley 3 full chapter pdf docx
Enumerative Combinatorics Volume 2 Second Edition Richard P Stanley 3 full chapter pdf docx
https://ebookmass.com/product/enumerative-combinatorics-
volume-2-second-edition-richard-p-stanley-3/
https://ebookmass.com/product/enumerative-combinatorics-
volume-2-second-edition-richard-p-stanley-2/
https://ebookmass.com/product/fetal-and-neonatal-
physiology-2-volume-set-6th-edition-richard-a-polin/
https://ebookmass.com/product/principles-of-bone-
biology-2-volume-set-4th-edition-edition-john-p-bilezikian/
(eBook PDF) America: The Essential Learning Edition
(Second Edition) (Vol. Volume 2) 2nd Edition
https://ebookmass.com/product/ebook-pdf-america-the-essential-
learning-edition-second-edition-vol-volume-2-2nd-edition/
https://ebookmass.com/product/grays-basic-anatomy-second-edition-
richard-lee-drake/
https://ebookmass.com/product/handbook-of-applied-hydrology-
second-edition-vijay-p-singh/
https://ebookmass.com/product/benzels-spine-surgery-2-volume-set-
techniques-complication-avoidance-and-management-5th-edition-
michael-p-steinmetz-md-editor/
https://ebookmass.com/product/big-chance-cowboy-teri-anne-
stanley-stanley/
C A M B R I D G E S T U D I E S I N A DVA N C E D M AT H E M AT I C S 2 0 8
Editorial Board
J. B E RTO I N, B. B O L L O B Á S , W. F U LTO N, B. K R A , I . M O E R D I J K ,
C . P R A E G E R , P. S A R NA K , B. S I M O N, B. TOTA RO
ENUMERATIVE COMBINATORICS
Volume 2
Richard Stanley’s two-volume basic introduction to enumerative combinatorics has
become the standard guide to the topic for students and experts alike. This thoroughly
revised second edition of volume two covers the composition of generating functions,
in particular the exponential formula and the Lagrange inversion formula, labelled and
unlabelled trees, algebraic, D-finite, and noncommutative generating functions, and
symmetric functions. The chapter on symmetric functions provides the only availa-
ble treatment of this subject suitable for an introductory graduate course and focusing
on combinatorics, especially the Robinson–Schensted–Knuth algorithm. An appendix
by Sergey Fomin covers some deeper aspects of symmetric functions, including jeu de
taquin and the Littlewood–Richardson rule. The exercises in the book play a vital role in
developing the material, and this second edition features over 400 exercises, including
159 new exercises on symmetric functions, all with solutions or references to solutions.
RICHARD P. STANLEY
Massachusetts Institute of Technology
With an Appendix by
SERGEY FOMIN
University of Michigan
Shaftesbury Road, Cambridge CB2 8EA, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467
www.cambridge.org
Information on this title: www.cambridge.org/9781009262491
DOI: 10.1017/9781009262538
© Cambridge University Press 1997, 2012
© Richard P. Stanley 2024
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press & Assessment.
First edition published by Wadsworth 1986
First Cambridge edition 1997
Second edition (Volume 1) published 2012
Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
A Cataloging-in-Publication data record for this book is available from the Library of
Congress.
ISBN 978-1-009-26249-1 Hardback
ISBN 978-1-009-26248-4 Paperback
Cambridge University Press & Assessment 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.
to Atsuko, Kenneth, Sharon, Beth, Bennett, and Cole
Contents
ix
x Contents
The primary difference between this second edition and the first is the addi-
tion of 159 new problems for Chapter 7 (symmetric functions), almost all
with solutions. They appear in the Chapter 7 sections entitled “Supplementary
Exercises” and “Supplementary Solutions.” These new problems are further
evidence of the amazing fecundity of the theory of symmetric functions. Philip
Hall was well aware of the future potential of symmetric functions when he
wrote in a 1955 letter:∗
. . . whenever in mathematics you meet with partitions, you have only to turn over the
stone or lift up the bark, and you will, almost infallibly, find symmetric functions under-
neath. More precisely, if we have a class of mathematical objects which in a natural and
significant way can be placed in one-to-one correspondence with the partitions, we must
expect the internal structure of these objects and their relations to one another to involve
sooner or later . . . the algebra of symmetric functions.
∗ See mathshistory.st-andrews.ac.uk/Biographies/Hall.
xiii
xiv Preface to Second Edition
as in the first edition. However, some equation numbers and figure numbers, as
well as almost all the citation numbers, have been changed. The index has been
upgraded; in particular, every reference to a person is a separate subentry.
Innumerable persons have contributed to the new Chapter 7 problems. They
are acknowledged in the problem solutions. I would also like to thank Kaitlin
Leach at Cambridge University Press for her excellent editorial assistance and
to Suresh Kumar for his TEX support.
Preface
xv
xvi Preface
X xn
F(x) = f (n) ,
n!
n≥0
X xn
G(x) = g(n) .
n!
n≥0
X xn
Ef (x) = f (n) .
n!
n≥0
1
2 Trees and the Composition of Generating Functions
where X is a finite set, and where (S,T) ranges over all weak ordered partitions
of X into two blocks, that is, S ∩ T = ∅ and S ∪ T = X . Then
Eh (x) = Ef (x)Eg (x). (5.2)
5.1.2 Example. Given an n-element set X , let h(n) be the number of ways to
split X into two subsets S and T with S ∪T = X , S ∩T = ∅; and then to linearly
order the elements of S and to choose a subset of T. There are f (k) = k! ways
to linearly order a k-element set, and g(k) = 2k ways to choose a subset of a
k-element set. Hence
X xn X xn X xn
n
h(n) = n! 2
n! n! n!
n≥0 n≥0 n≥0
e2x
= .
1−x
1 All references to Volume 1 are to the second edition.
5.1 The Exponential Formula 3
where (T1 , . . . , Tk ) ranges over all weak ordered partitions of S into k blocks,
that is, T1 , . . . , Tk are subsets of S satisfying: (i) Ti ∩ Tj = ∅ if i 6= j, and (ii)
T1 ∪ · · · ∪ Tk = S. Then
k
Y
Eh (x) = Efi (x).
i=1
We are now able to give the main result of this section, which explains
the combinatorial significance of the composition of exponential generating
functions.
Proof. Suppose #S = n, and let hk (n) denote the right-hand side of (5.3) for
fixed k. Since B1 , . . . , Bk are nonempty they are all distinct, so there are k! ways
of linearly ordering them. Thus by Proposition 5.1.3,
g(k)
Ehk (x) = Ef (x)k . (5.3)
k!
Summing (5.3) over all k ≥ 1 yields the desired result.
6 3 4 2 8
5.1.5 Example. Let h(n) be the number of ways for n persons to form into
nonempty lines, and then to arrange these lines in a circular order. Figure 5.1
shows one such arrangement of nine persons. There are f ( j) = j! ways to
linearly order j persons, and g(k) = (k − 1)! ways to circularly order k ≥ 1
lines. Thus
X xn x
Ef (x) = n! = ,
n! 1−x
n≥1
X xn
Eg (x) = 1 + (n − 1)! = 1 + log(1 − x)−1 ,
n!
n≥1
so
Eh (x) = Eg (Ef (x))
−1
x
= 1 + log 1 −
1−x
X xn
= 1+ (2n − 1)(n − 1)! ,
n!
n≥1
whence h(n) = (2n− 1)(n − 1)!. Naturally such a simple answer demands a
simple combinatorial proof. Namely, arrange the n persons in a circle in (n−1)!
5.1 The Exponential Formula 5
6 4
3 2
1 8
ways. In each of the n spaces between two persons, either do or do not draw a
bar, except that at least one bar must be drawn. There are thus 2n − 1 choices
for the bars. Between two consecutive bars (or a bar and itself if there is only
one bar) read the persons in clockwise order to obtain their order in line. See
Figure 5.2 for this method of representing Figure 5.1.
The most common use of Theorem 5.1.4 is the case where g(k) = 1
for all k. In combinatorial terms, a structure is put together from “con-
nected” components, but no additional structure is placed on the components
themselves.
X
h(#S) = f (#B1 )f (#B2 ) · · · f (#Bk ), #S > 0, (5.4)
π={B1 ,...,Bk }∈5(S)
h(0) = 1.
Then
Eh (x) = exp Ef (x). (5.5)
Let us say a brief word about the computational aspects of equation (5.5).
If the function f (n) is given, then one can use (5.4) to compute h(n). However,
there is a much more efficient way to compute h(n) from f (n) (and conversely).
n
X n
h(n + 1) = h(k)f (n + 1 − k), (5.6)
k
k=0
n
X n
f (n + 1) = h(n + 1) − h(k)f (n + 1 − k). (5.7)
k
k=1
Now equate coefficients of xn /n! on both sides of (5.8) to obtain (5.6). (It
is also easy to give a combinatorial proof of (5.6).) Equation (5.7) is just a
rearrangement of (5.6).
Proof. Since there are (j − 1)! ways to cyclically order a j-set, equation (5.9)
may be written
X
h(#S) = [(#B1 − 1)!f (#B1 )] · · · [(#Bk − 1)!f (#Bk )]g(k),
π={B1 ,...,Bk }∈5(S)
5.1 The Exponential Formula 7
so by Theorem 5.1.4,
X xn
Eh (x) = Eg (n − 1)!f (n)
n!
n≥1
. X xn
= Eg f (n)
n
n≥1
Proof. Let P and Q be locally finite posets, and let u ∈ I(P, K), v ∈ I(Q, K).
Define u × v ∈ I(P × Q, K) by
It follows from Lemma 5.1.10 that the set M(5) = M(5, K) of multipli-
cative functions on 5 forms a semigroup under convolution. In fact, M(5)
is even a monoid (= semigroup with identity), since the identity function
δ = (δ1 , δ2 , . . . ) is multiplicative with δ(n) = δ1n . (CAVEAT: M(5) is not
closed under addition!)
X xn
φ( f ) = Ef (x) = f (n) .
n!
n≥1
Since (5.13) agrees with (5.3), the proof follows from Theorem 5.1.4.
The next result follows from Theorem 5.1.11 in the same way that Prop-
osition 3.18.5 follows from Theorem 3.18.4 (using Proposition 5.4.1), so the
proof is omitted. (A direct proof avoiding Theorem 5.1.11 can also be given.)
If f = ( f1 , f2 , . . . ) where fn ∈ I(5n , K) and each fn−1 exists in I(5n , K), then
we write f −1 = ( f1−1 , f2−1 , . . . ).
Eδ (x) = x,
so by Theorem 5.1.11
[exp Eµ (x)] − 1 = x
⇒ Eµ (x) = log(1 + x)
X xn
= (−1)n−1 (n − 1)!
n!
n≥1
⇒ µ(n) = (−1)n−1 (n − 1)!.
Thus we have another derivation of the Möbius function of 5n (equa-
tion (3.37)).
5.1.14 Example. Let h(n) be the number of ways to partition the set [n], and
then partition each block into blocks of odd cardinality. We are asking for the
number of chains 0̂ ≤ π ≤ σ ≤ 1̂ in 5n such that all block sizes of π are odd.
Define f ∈ M(5) by
1, n odd
f (n) =
0, n even.
10 Trees and the Composition of Generating Functions
Then
Proof. Easy.
5.2.1 Example. The number of graphs (without loops or multiple edges) on
an n-element vertex set S is clearly 2(2) . (Each of the n2 pairs of vertices may
n
or may not be joined by an edge.) Let c(#S) = c(n) be the number of connected
graphs on the vertex set S. Since a graph on S is obtained by choosing a parti-
tion π of S and then placing a connected graph on each block of π, we see that
equation (5.5) holds for h(n) = 2(2) and f (n) = c(n). Hence by Corollary 5.1.6,
n
X n xn
Eh (x) = 2(2)
n!
n≥0
= exp Ec (x)
X xn
= exp c(n) .
n!
n≥1
Equivalently,
xn X n xn
2(2) .
X
c(n) = log (5.22)
n! n!
n≥1 n≥0
Note that the generating functions Eh (x) and Ec (x) have zero radius of
convergence; nevertheless, they still have combinatorial meaning.
Of course there is nothing special about graphs in the above example. If, for
instance, h(n) is the number of posets (or digraphs, topologies, triangle-free
graphs, . . . ) on an n-set and c(n) is the number of connected posets (digraphs,
topologies, triangle-free graphs, . . . ) on an n-set, then the fundamental rela-
tion Eh (x) = exp Ec (x) continues to hold. In some cases (such as graphs
and digraphs) we have an explicit formula for h(n), but this is an incidental
“bonus.”
12 Trees and the Composition of Generating Functions
5.2.2 Example. Suppose we are interested in not just the number of connected
graphs on an n-element vertex set, but rather the number of such graphs with
exactly k components. Let ck (n) denote this number, and define
XX xn
F(x, t) = ck (n)tk . (5.23)
n!
n≥0 k≥0
There are two ways to obtain this generating function from Theorem 5.1.4 and
Corollary 5.1.6. We can either set f (n) = c(n) and g(k) = tk in (3), or set
f (n) = c(n)t in (5.5). In either case we have
X
h(n) = ck (n)tk .
k≥0
Thus
X xn
F(x, t) = exp t c(n)
n!
n≥1 t
X n xn
= 2(2) .
n!
n≥0
Again the same reasoning works equally as well for posets, digraphs, topolo-
gies, . . . In general, if Eh (x) is the exponential generating function for the total
number of structures on an n-set (where of course each structure is a unique
disjoint union of connected components), then Eh (x)t also keeps track of the
number of components, as in (5.23). Equivalently, if h(n) is the number of
structures on an n-set and ck (n) the number with k components, then
X
tk Eck (x) = Eh (x)t
k≥0
= exp tEc1 (x)
X Ec (x)k
= tk 1 , (5.24)
k!
k≥0
so
1 1
Eck (x) = Ec1 (x)k = (log Eh (x))k ,
k! k!
where we set ck (0) = δ0k and h(0) = 1. In particular, if h(n) = n! (the number
of permutations of an n-set) then Eh (x) = (1 − x)−1 while ck (n) = c(n, k), the
number of permutations of an n-set with k cycles. In other words, c(n, k) is a
signless Stirling number of the first kind (see Chapter 1.3); and we get
X xn 1 h ik
c(n, k) = log(1 − x)−1 . (5.25)
n! k!
n≥0
5.2 Applications of the Exponential Formula 13
Let us give one further “direct” application, concocted for the elegance of the
final answer.
5.2.4 Example. Let B(n) = B1 (n) denote the nth Bell number, that is, B(n) =
#5n (Chapter 1.9). Setting each f (i) = 1 in (5.4), we obtain
X xn
EB (x) = B(n) = exp(ex − 1).
n!
n≥0
(See equation (1.94f).) Now let B2 (n) be the number of ways to partition an n-
set S, and then partition each block. Equivalently, B2 (n) is the number of chains
0̂ ≤ π1 ≤ π2 ≤ 1̂ in 5n . Putting each f (i) = B(i) in (5.4), or equivalently using
Theorem 5.1.11 to compute φ(ζ 3 ), we obtain
X xn
B2 (n) = exp[exp(ex − 1) − 1].
n!
n≥0
14 Trees and the Composition of Generating Functions
123456789
14689 257
3
1
469 8 2 5 7
49
4 9
In other words, writing F h−1i (x) for the compositional inverse of F(x) = ax +
bx2 + · · · where a 6 = 0, that is,
we have
It does not seem possible to obtain a simpler result. In particular, in Section 5.4
we will discuss methods for computing the coefficients of compositional
inverses, but these methods don’t seem to yield anything interesting when
F(x) = 1 + 2x − ex . For some enumeration problems closely related to total
partitions, see Exercises 26 and 40.
5.2.6 Example. Consider the variation of the preceding example where each
non-singleton block must be partitioned into exactly two blocks. Call such a
procedure a binary total partition of S, and denote the number of them by
b(#S). As with total partitions, set b(0) = 0. The tree representing a binary
total partition is a complete (unordered) binary tree, as illustrated in Figure 5.4.
(Thus b(n) is just the number of (unordered) complete binary trees with n
labelled endpoints.) It now follows from Theorem 5.1.4 (with g(k) = δ2k )
or just by (5.3) (with k = 2 and g(2) = 1), in a similar way to how we obtained
(5.26), that
1
Eb (x)2 = Eb (x) − x. (5.28)
2
16 Trees and the Composition of Generating Functions
whence
b(n) = 1 · 3 · 5 · · · (2n − 3).
π = {{1, 4}, {2, 9}, {3, 10}, {5, 7}, {6, 8}, {11, 12}}.
We leave the reader to check (not entirely trivial) that this procedure yields the
desired bijection.
5.2 Applications of the Exponential Formula 17
Certain problems involving symmetric matrices are well-suited for use of the
exponential formula. (Analogous results for arbitrary matrices are discussed in
Section 5.5.) The basic idea is that a symmetric matrix A = (auv ) whose rows
and columns are indexed by a set V may be identified with a graph G = GA
on the vertex set V , with an edge uv connecting u and v labelled by auv . (If
auv = 0, then we simply omit the edge uv, rather than labelling it by 0. More
generally, if auv ∈ P then it is often convenient to draw auv (unlabelled) edges
between u and v.) Sometimes the connected components of GA have a simple
structure, so that the exponential formula can be used to enumerate all the
graphs (or matrices).
X xn x 2 1 X x n 1 X xn
Sn (2) = exp x + + (n − 1)! + n!
n! 2! 2 n! 2 n!
n≥0 n≥3 n≥2
x2 1 X xn 1 X n
= exp + + x
4 2 n 2
n≥1 n≥1
2
x 1 x
= exp + log(1 − x)−1 +
4 2 2(1 − x)
2
−1/2 x x
= (1 − x) exp + .
4 2(1 − x)
1
Sn+1 = (2n + 1)Sn − (n)2 Sn−1 − (n)2 Sn−2 + (n)3 Sn−3 , n ≥ 0.
2
5.2.8 Example. Suppose that in the previous example A must be a 0–1 matrix
(i.e., the entry two is not allowed). Now the components of GA of type (a) or
18 Trees and the Composition of Generating Functions
(b) above are not allowed. If we let Sn∗ (2) denote the number of such matrices,
it follows that
X xn x2 X xn
Sn∗ (2) = e−x− 2 Sn (2)
n! n!
n≥0 n≥0
x2
−1/2 x
= (1 − x) exp −x − + .
4 2(1 − x)
As a further variation, suppose we again allow two as an entry, but that tr A = 0
(i.e., all main diagonal entries are zero). Now the components of GA cannot
have loops, so are of type (b) or (c). Hence, letting Tn (2) be the number of such
matrices, we have
X xn x2 1 X x n
Tn (2) = exp + (n − 1)!
n! 2! 2 n!
n≥0 n≥3
x x2
= (1 − x) −1/2 exp − + .
2 4
Similarly, if Tn∗ (2) denotes the number of traceless symmetric n×n 0-1 matrices
with line sum 2, then
X x n 1 X x n
Tn∗ (2) = exp (n − 1)!
n! 2 n!
n≥0 n≥3
x x2
−1/2
= (1 − x) exp − − . (5.29)
2 4
The recurrence relations for Sn∗ (2), Tn (2), and Tn∗ (2) turn out to be (using the
technique of Exercise 24(c))
det X3 = x11 x22 x33 + 2x12 x23 x13 − x213 x22 − x11 x223 − x212 x33 .
Hence L(3) = 5, since the above sum has five terms. One might ask whether
we should count a monomial which does arise in the expansion of det Xn but
whose coefficient because of cancellation turns out to be zero. But we will soon
see that no cancellation is possible; all occurrences of a given monomial have
the same sign. Suppose now that π ∈ Sn . Define
We conclude this section with some examples where it is more natural to use
the permutation version of the exponential formula (Corollary 5.1.9).
20 Trees and the Composition of Generating Functions
in the variables t = (t1 , t2 , . . . ). We call Z(π) the cycle index (or cycle indicator
or cycle monomial) of π. Define the cycle index or cycle index polynomial (or
cycle indicator, etc.) Z(Sn ) (also denoted ZSn (t), PSn (t), Cyc(Sn , t), etc.) of
Sn by
1 X
Z(Sn ) = Z(Sn , t) = Z(π ).
n!
π∈Sn
(In Chapter 7.7.24 we will consider the cycle index of other permutation
groups.) It is sometimes more convenient to work with the augmented cycle
index
X
Z̃(Sn ) = n! Z(Sn ) = Z(π).
π ∈Sn
For instance
Z̃(S1 ) = t1
Z̃(S2 ) = t12 + t2
Z̃(S3 ) = t13 + 3t1 t2 + 2t3
Z̃(S4 ) = t14 + 6t12 t2 + 8t1 t3 + 3t22 + 6t4 .
er (n) = Z̃(Sn )
1, d|r
td =
0, d6 |r.
5.2 Applications of the Exponential Formula 21
It follows that
X xn X xd
er (n) = exp td
1, d|r
n! d t =
n≥0 d≥1 d
0, d6 |r
X xd
= exp . (5.31)
d
d|r
This is the same generating function encountered way back in equation (1.11).
Now we are able to understand its combinatorial significance more clearly.
There is a surprising connection between (a) Corollary 5.1.9, (b) the rela-
tionship between linear and circular words obtained in Proposition 4.7.13, and
(c) the bijection π 7 → π̂ discussed in Chapter 1.3 between permutations writ-
ten as products of cycles and as words. Basically, such a connection arises from
a formula of the type
X xn X xn
n! f (n) = exp (n − 1)! g(n)
n! n!
n≥0 n≥1
XX xn
= tkn
n
k n≥1
X xn
= pn (t) , (5.34)
n
n≥1
22 Trees and the Composition of Generating Functions
X
where pn (t) = tkn . On the other hand, it is clear from (5.33) that
k
X
F(x) = hn (t)xn , (5.35)
n≥0
where hn (t) is the sum of all monomials of degree n in t = (t1 , t2 , · · · ), that is,
t1a1 t2a2 · · · .
X
hn (t) =
a1 +a2 +···=n
ai ∈N
(In Chapter 7 we will analyze the symmetric functions pn (t) and hn (t), as well
as many others, in much greater depth.) From (5.34) and (5.35) we conclude
X xn X xn
n! hn (t) = exp pn (t) . (5.36)
n! n
n≥0 n≥1
first impose some linear ordering on the tj ’s, say t1 < t2 < · · · . For fixed j, take
all the cycles C of π with weight tj#C and write their standard representation (in
the sense of Proposition 1.3.1), that is, the largest element of each cycle is writ-
ten first in the cycle, and the cycles are written left-to-right in increasing order
of their largest elements. Remove the parentheses from this standard represen-
tation, obtaining a word wj . Finally set φ(π ) = (w, ta ), where w = w1 w2 · · ·
(juxtaposition of words). For instance, suppose π is the weighted permutation
(19) (82) (3) (547) (6)
π=
vv uu v uuu u
where u < v. The cycles weighted by u’s and v’s, respectively, have standard
form
(6)(754)(82) (weight u6 )
(3)(91) (weight v3 ).
Hence
w1 = 675482, w2 = 391,
φ(π) = (675482391, u6 v3 ).
It is easy to check that φ is a bijection. Given (w, ta ), the monomial ta deter-
|w |
mines the words wj with their weights tj j . Each word wj then corresponds to
|C|
a collection of cycles C (with weight tj ) using the inverse of the bijection
π 7 → π̂ of Chapter 1.3.
A similar argument leads to a direct combinatorial proof of equation (4.41);
see Exercise 21.
Figure 5.6 A plane tree built from the subtrees of its root
where r(n) as above is the number of rooted trees on the vertex set [n] (with
r(0) = 0). Then y = xey , or equivalently (since x = ye−y ),
In the functional equation y = xey of Proposition 5.3.1, substitute xey for the
occurrence of y on the right-hand side to obtain
y
xe
y = xe
5.3 Enumeration of trees 25
y
xe
xe
y = xe .
·
·
x·
xe
xe
y = xe . (5.41)
where rk (n) in the number of rooted trees on [n] of length ≤ k. For instance,
X xn
y1 = xex = n ,
n!
n≥1
f (n) = number of free forests (i.e., disjoint unions of free trees) on [n]
We set t(0) = 0, f (0) = 1, p(0) = 1. Also write T(x) = Et (x), F(x) = Ef (x),
and P(x) = Ep (x). It is easy to verify the following relations:
[n]
5.3.2 Proposition. We have pS (n) = knn−k−1 for any S ∈ k . Thus
n n−k−1 n − 1 n−k
pk (n) = k n = n
k k−1
n−1
r(n) = n
t(n) = nn−2
p(n) = (n + 1)n−1 .
The alien was still making his meaningless signals when Blake
brought in a pair of earphones and clamped them on MacIlheny's
head. A throat mike around his neck completed the communication
circuit. "Can you hear me, Mr. President?" MacIlheny asked.
"Yes. Your man Blake explained everything to me."
"Got any advice?"
"Not yet. Let's see what happens. By the way, I've given the
impression to the rest of the world that it was through your efforts that
Number Four avoided crashing; I don't think we'd better let this leak
out just yet."
"Right. Meantime, I'm going to try to capture that lad."
"How?" asked the President.
"Invite him into the ship and bring him back with it."
"All right," said the President, "but be careful."
"He's given up," said Blake, gesturing toward the screen.
The alien had given up his incomprehensible gesticulating and stood
with his odd arms folded in an uncomfortable-looking knot.
"Major," said MacIlheny, "open the cargo hold."
The officer looked puzzled, but did as he was told. After all, the
President himself had ordered him to obey MacIlheny. He touched a
button on one side of the control panel. After four or five seconds, a
light came on above it, indicating that the cargo hold of the RJ-37 was
open. The alien evidently saw the door swing inward; he hesitated for
a moment, then went around to the side of the ship, out of range of
the TV camera.
But he didn't go inside immediately. MacIlheny hadn't expected him
to; the alien couldn't be that stupid. After perhaps half a minute, the
alien figure reappeared and strode deliberately back to his own ship.
He opened a port in the side and disappeared within.
Then, quite suddenly, the screen went blank.
"What happened?" snapped MacIlheny.
Blake, who had been watching the beam control instruments, said: "I
don't know how he's done it, but he's managed to jam our radio
beam! We're not getting any signal through!"
The President's voice crackled in MacIlheny's ears.
"Fitz! Detonate that bomb! We can't take any chances!"
MacIlheny half grinned. "Major," he said, "set off the H-bomb."
The major pressed a red button on the control panel.
Twenty minutes later, the screen came on again, showing the same
scene as before. No one was surprised. By then, reports had come in
that the satellite was still visible, still in its orbit. The H-bomb had
failed to go off; the signal had never reached the detonation device.
The alien was standing in front of the camera, holding a large piece of
mechanism in his hands. On Earth, the thing would have been almost
too heavy to lift, but the gravitational pull of Satellite Number Four
was almost negligible.
"He's got the H-bomb!"
MacIlheny recognized the President's voice in his ears.
The alien bowed toward the camera, then straightened and went
back to his own ship. He clambered up the side of it with magnetic
soles as easily as he had walked on the hull of the space station.
Near the end of his ship, he opened a small door in the hull. Within
was utter blackness.
"Here's the way I see it, Mr. President," said MacIlheny several hours
later. "When he cracked up by accidentally plowing into Number Four,
something happened to his energy supply. Maybe he was already
low, I don't know. Anyway, he was out of fuel."
"What do you think he used for fuel?"
"The most efficient there is," said MacIlheny. "Pure energy. Imagine
some sort of force field that will let energy in, but won't let it out. It
would be dead black on the outside, just like that whatever-it-was in
the alien's ship. He just set off the H-bomb inside that field; what little
radiation did get out made the field look gray—and that's a damned
small loss in comparison with the total energy of that bomb."
"You know, Fitz, I'm going to have a hell of a job explaining where that
bomb went," said the President.
"Yeah, but we've got his gun or whatever in exchange."
"But how do you know our technicians will be able to figure it out?"
"I think they will," MacIlheny said. "Their technology must be similar to
ours or he wouldn't have been able to figure out how to fire the jets on
the satellite or how to set off that bomb. He wouldn't have even
known what the bomb was unless he was familiar with something
similar. And he wouldn't have been able to blank out our controls
unless he had a good idea of how they operated. They may be a little
ahead of us, but not too much, and I'll bet we have some things they
haven't."
"The trouble is," the President said worriedly, "that we don't know
where he came from. He knows where we are, but we don't have any
idea where his home planet is."
"That's true. On the other hand, we know something about his
physical characteristics, while he doesn't know anything about ours.
For instance, I doubt if he'd be happy here on Earth; judging by the
helmet he wore, he can't stand too much light. He had it polarized
almost black. Probably comes from a planet with a dim, red sun."
"Well, Fitz, when they do come, I hope it's for trade and not for war."
MacIlheny grinned. "It won't be war. Don't you remember? We've
started trading already!"
*** END OF THE PROJECT GUTENBERG EBOOK BEYOND OUR
CONTROL ***
Updated editions will replace the previous one—the old editions will
be renamed.
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside the
United States, check the laws of your country in addition to the terms
of this agreement before downloading, copying, displaying,
performing, distributing or creating derivative works based on this
work or any other Project Gutenberg™ work. The Foundation makes
no representations concerning the copyright status of any work in
any country other than the United States.
• You pay a royalty fee of 20% of the gross profits you derive from
the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.F.
1.F.4. Except for the limited right of replacement or refund set forth in
paragraph 1.F.3, this work is provided to you ‘AS-IS’, WITH NO
OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.
Please check the Project Gutenberg web pages for current donation
methods and addresses. Donations are accepted in a number of
other ways including checks, online payments and credit card
donations. To donate, please visit: www.gutenberg.org/donate.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.