Visit https://ebookultra.

com to download the full version and

explore more ebooks

Combinatorics topics techniques algorithms 1.

ed., transferred to digital print., Nachdr.
Edition Cameron

_____ Click the link below to download _____

https://ebookultra.com/download/combinatorics-topics-
techniques-algorithms-1-ed-transferred-to-digital-
print-nachdr-edition-cameron/

Explore and download more ebooks at ebookultra.com

Here are some suggested products you might be interested in.
Click the link to download

Responsibility and control a theory of moral

responsibility 1. pbk. ed., digital print., Nachdr. der
Ausg. Cambridge 200 Edition Fischer
https://ebookultra.com/download/responsibility-and-control-a-theory-
of-moral-responsibility-1-pbk-ed-digital-print-nachdr-der-ausg-
cambridge-200-edition-fischer/

SAT II Success MATH 1C and 2C 2002 Peterson s SAT II

Success Peterson'S

https://ebookultra.com/download/sat-ii-success-
math-1c-and-2c-2002-peterson-s-sat-ii-success-petersons/

Enumerative Combinatorics Volume 2 1, Digital Printing

2004 Edition Richard P. Stanley

https://ebookultra.com/download/enumerative-combinatorics-
volume-2-1-digital-printing-2004-edition-richard-p-stanley/

English in the Southern United States 1. paperback ed.,

Nachdr. Edition Nagle

https://ebookultra.com/download/english-in-the-southern-united-
states-1-paperback-ed-nachdr-edition-nagle/
Caffeine for the creative mind 250 exercises to wake up
your brain 1. ed., Nachdr. Edition Mumaw

https://ebookultra.com/download/caffeine-for-the-creative-
mind-250-exercises-to-wake-up-your-brain-1-ed-nachdr-edition-mumaw/

Made by hand furniture projects from the unplugged

woodshop 1. ed., [Nachdr.] Edition Fidgen

https://ebookultra.com/download/made-by-hand-furniture-projects-from-
the-unplugged-woodshop-1-ed-nachdr-edition-fidgen/

Core Servlets and JavaServer Pages Vol 1 Core Technologies

2. ed., Nachdr Edition Hall

https://ebookultra.com/download/core-servlets-and-javaserver-pages-
vol-1-core-technologies-2-ed-nachdr-edition-hall/

Digital Techniques for Wideband Receivers 3rd ed Edition

Cheng

https://ebookultra.com/download/digital-techniques-for-wideband-
receivers-3rd-ed-edition-cheng/

Metallographic Polishing by Mechanical Methods 4th ed.,

digital print. 2010 Edition Leonard E. Samuels

https://ebookultra.com/download/metallographic-polishing-by-
mechanical-methods-4th-ed-digital-print-2010-edition-leonard-e-
samuels/
Combinatorics topics techniques algorithms 1. ed.,
transferred to digital print., Nachdr. Edition Cameron
Digital Instant Download
Author(s): Cameron, Peter J
ISBN(s): 9780521457613, 0521457610
Edition: 1. ed., transferred to digital print., Nachdr.
File Details: PDF, 32.22 MB
Year: 2010
Language: english
Combinatorics is a subject of increasing importance, owing to its
links with computer science, statistics and algebra. This is a textbook
aimed at second-year undergraduates to beginning graduates. It
stresses common techniques (such as generating functions and
recursive construction) which underlie the great variety of subject
matter, and the fact that a constructive or algorithmic proof is more
valuable than an existence proof.
The book is divided into two parts, the second at a higher level and
with a wider range than the first. Historical notes are included and
give a wider perspective on die subject. More advanced topics are
given as projects, and there are a number of exercises, some with
solutions given.
Combinatorics:
Topics, Techniques, Algorithms
COMBINATORICS:
TOPICS, TECHNIQUES, ALGORITHMS

PETER J. CAMERON

Queen Mary & Westfield College, London

CAMBRIDGE
UNIVERSITY PRESS
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, São Paulo, Delhi, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by
Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521451338
© Cambridge University Press 1994
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.
First edition published 1994
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
isbn 978-0-521-45133-8 Hardback
isbn 978-0-521-45761-3 Paperback
Cambridge University Press has no responsibility for the persistence or
accuracy of URLs for external or third-party internet websites referred to in
this publication, and does not guarantee that any content on such websites is,
or will remain, accurate or appropriate. Information regarding prices, travel
timetables, and other factual information given in this work is correct at
the time of first printing but Cambridge University Press does not guarantee
the accuracy of such information thereafter.
Contents

Preface ix
1. What is Combinatorics? 1
Sample problems — How to use this book — What you need to know — Exercises

2. On numbers and counting 7

Natural numbers and arithmetic — Induction — Some useful functions — Orders
of magnitude — Different ways of counting — Double counting — Appendix on
set notation — Exercises
3. Subsets, partitions, permutations 21
Subsets — Subsets of fixed size — The Binomial Theorem and Pascal's Triangle
— Project: Congruences of binomial coefficients — Permutations — Estimates for
factorials — Selections — Equivalence and order — Project: Finite topologies —
Project: Cayley's Theorem on trees — Bell numbers — Generating combinatorial
objects — Exercises
4. Recurrence relations and generating functions 49
Fibonacci numbers — Aside on formal power series — Linear recurrence relations
with constant coefficients — Derangements and involutions — Catalan and Bell
numbers — Computing solutions to recurrence relations — Project: Finite fields
and QUICKSORT — Exercises
5. The Principle of Inclusion and Exclusion 75
PIE — A generalisation — Stirling numbers — Project: Stirling numbers and
exponentials — Even and odd permutations — Exercises
6. Latin squares and SDRs 87
Latin squares — Systems of distinct representatives — How many Latin squares?
— Quasigroups — Project: Quasigroups and groups — Orthogonal Latin squares
— Exercises
7. Extremal set theory 99
Intersecting families — Sperner families — The De Bruijn-Erdos Theorem —
Project: Regular families — Exercises
8. Steiner triple systems 107
Steiner systems — A direct construction — A recursive construction — Packing and
covering — Project: Some special Steiner triple systems — Project: Tournaments
and Kirkman's schoolgirls — Exercises
9. Finite geometry 123
Linear algebra over finite fields — Gaussian coefficients — Projective geometry —
Axioms for projective geometry — Projective planes — Other kinds of geometry
— Project: Coordinates and configurations — Project: Proof of the Bruck-Ryser
Theorem — Appendix: Finite fields — Exercises
10. Ramsey's Theorem 147
The Pigeonhole Principle — Some special cases — Ramsey's Theorem — Bounds
for Ramsey numbers — Applications — The infinite version — Exercises
viii Contents

11. Graphs 159

Definitions — Trees and forests — Minimal spanning trees — Eulerian graphs
— Hamiltonian graphs — Project: Gray codes — The Travelling Salesman —
Digraphs — Networks — Menger, Konig and Hall — Diameter and girth —
Project: Moore graphs — Exercises
12. Posets, lattices and matroids 187
Posets and lattices — Linear extensions of a poset — Distributive lattices —
Aside on propositional logic — Chains and antichains — Products and dimension
— The Mobius function of a poset — Matroids — Project: Arrow's Theorem —
Exercises
13. More on partitions and permutations 209
Partitions, diagrams and conjugacy classes — Euler's Pentagonal Numbers
Theorem — Project: Jacobi's Identity — Tableaux — Symmetric polynomials —
Exercises
14. Automorphism groups and permutation groups 225
Three definitions of a group — Examples of groups — Orbits and transitivity —
The Schreier-Sims algorithm — Primitivity and multiple transitivity — Examples
— Project: Cayley digraphs and Frucht's Theorem — Exercises
15. Enumeration under group action 245
The Orbit-counting Lemma — An application — Cycle index — Examples —
Direct and wreath products — Stirling numbers revisited — Project: Cycle index
and symmetric functions — Exercises
16. Designs 257
Definitions and examples — To repeat or not to repeat — Fisher's Inequality —
Designs from finite geometry — Small designs — Project: Hadamard matrices —
Exercises
17. Error-correcting codes 271
Finding out a liar — Definitions — Probabilistic considerations — Some bounds
— Linear codes; Hamming codes — Perfect codes — Linear codes and projective
spaces — Exercises
18. Graph colourings 291
More on bipartite graphs — Vertex colourings — Project: Brooks' Theorem —
Perfect graphs — Edge colourings — Topological graph theory — Project: The
Five-colour Theorem — Exercises
19. The infinite 307
Counting infinite sets — Konig's Infinity Lemma — Posets and Zorn's Lemma —
Ramsey theory — Systems of distinct representatives — Free constructions — The
random graph — Exercises
20. Where to from here? 325
Computational complexity — Some graph-theoretic topics — Computer software
— Unsolved problems — Further reading
Answers to selected exercises 339
Bibliography 343
Index 347
Preface

Ive got to work the E qwations and the low cations

Ive got to comb the nations of it.
Russell Hoban, Riddley Walker (1980)

We have not begun to understand the relationship between combinatorics and

conceptual mathematics.
J. Dieudonne, A Panorama of Pure Mathematics (1982)

If anything at all can be deduced from the two quotations at the top of this page,
perhaps it is this: Combinatorics is an essential part of the human spirit; but it is
a difficult subject for the abstract, axiomatising Bourbaki school of mathematics to
comprehend. Nevertheless, the advent of computers and electronic communications
have made it a more important subject than ever.
This is a textbook on combinatorics. It's based on my experience of more than
twenty years of research and, more specifically, on teaching a course at Queen Mary
and Westfield College, University of London, since 1986. The book presupposes
some mathematical knowledge. The first part (Chapters 2-11) could be studied by
a second-year British undergraduate; but I hope that more advanced students will
find something interesting here too (especially in the Projects, which may be skipped
without much loss by beginners). The second half (Chapters 12-20) is in a more
condensed style, more suited to postgraduate students.
I am grateful to many colleagues, friends and students for all kinds of contribu-
tions, some of which are acknowledged in the text; and to Neill Cameron, for the
illustration on p. 128.
I have not provided a table of dependencies between chapters. Everything is
connected; but combinatorics is, by nature, broad rather than deep. The more
important connections are indicated at the start of the chapters.
Peter J. Cameron
17 March 1994
1. What is Combinatorics?
Combinatorics is the slums of topology.
J. H. C. Whitehead (attr.)1

I have to admit that he was not bad at combinatorial analysis — a branch,

however, that even then I considered to be dried up.
Stanislaw Lem, His Master's Voice (1968)

Combinatorics is special. Most mathematical topics which can be covered in a

lecture course build towards a single, well-defined goal, such as Cauchy's Theorem
or the Prime Number Theorem. Even if such a clear goal doesn't exist, there is
a sharp focus (finite groups, perhaps, or non-parametric statistics). By contrast,
combinatorics appears to be a collection of unrelated puzzles chosen at random.
Two factors contribute to this. First, combinatorics is broad rather than deep.
Its tentacles stretch into virtually all corners of mathematics. Second, it is about
techniques rather than results. As in a net,2 threads run through the entire con-
struction, appearing unexpectedly far from where we last saw them. A treatment of
combinatorics which neglects this is bound to give a superficial impression.
This feature makes the teacher's job harder. Reading, or lecturing, is inherently
one-dimensional. If we follow one thread, we miss the essential interconnectedness
of the subject.
I have attempted to meet this difficulty by various devices. Each chapter begins
with a list of topics, techniques, and algorithms considered in the chapter, and
cross-references to other chapters. Also, some of the material is set in smaller
type and can be regarded as optional. This usually includes a 'project' involving a
more difficult proof or construction (where the arguments may only be sketched,
requiring extra work by the reader). These projects could be used for presentations
by students. Finally, the book is divided into two parts; the second part treats topics
in greater depth, and the pace hots up a bit (though, I hope, not at the expense of
intelligibility).
As just noted, there are algorithms scattered throughout the book. These are not
computer programs, but descriptions in English of how a computation is performed.
I hope that they can be turned into computer programs or subroutines by readers
with programming experience. The point is that an explicit construction of an object
usually tells us more than a non-constructive existence proof. (Examples will be
given to illustrate this.) An algorithm resembles a theorem in that it requires a proof
(not of the algorithm itself, but of the fact that it does what is claimed of it).
1
This attribution is due to Graham Higman, who revised Whitehead's definition to 'Combinatorics
is the mews of algebra.'
2
'Net Anything reticulated or decussated at equal distances, with interstices between the intersec-
tions.' Samuel Johnson, Dictionary of the English Language (1775).
2 1. What is Combinatorics?

But what is combinatorics? Why should you read further?

Combinatorics could be described as the art of arranging objects according
to specified rules. We want to know, first, whether a particular arrangement is
possible at all, and if so, in how many different ways it can be done. If the rules
are simple (like picking a cricket team from a class of schoolboys), the existence
of an arrangement is clear, and we concentrate on the counting problem. But for
more involved rules, it may not be clear whether the arrangement is possible at all.
Examples are Kirkman's schoolgirls and Euler's officers, described below.
Sample problems
In this section, I will give four examples of combinatorial questions chosen to
illustrate the nature of the subject. Each of these will be discussed later in the book.
Derangements

Given n letters and n addressed envelopes, in how many ways can

the letters be placed in the envelopes so that no letter is in the
correct envelope?
DISCUSSION. The total number of ways of putting the letters in the envelopes is the
number of permutations of n objects,3 which is n! (factorial n). We will see that
the fraction of these which are all incorrectly addressed is very close to 1/e, where
e = 2.71828... is the base of natural logarithms — a surprising result at first sight.
In fact, the exact number of ways of mis-addressing all the letters is the nearest
integer to n\/e (see Exercise 1).
Kirkman's schoolgirls

Fifteen schoolgirls walk each day in five groups of three. Arrange

the girls9 walks for a week so that, in that time, each pair of girls
walks together in a group just once.
DISCUSSION. If it is possible at all, seven days will be required. For any given
girl must walk once with each of the other fourteen; and each day she walks with
two others. However, showing that the walks are actually possible requires more
argument. The question was posed and solved by Kirkman in 1847. The same
question could be asked for other numbers of girls (see Exercise 2). Only in 1967
did Ray-Chaudhuri and Wilson show that solutions exist for any number of girls
congruent to 3 modulo 6.
Eider's officers

Thirty-six officers are given, belonging to six regiments and holding

six ranks (so that each combination of rank and regiment cor-
responds to just one officer). Can the officers be paraded in a
6 x 6 array so that, in any line (row or column) of the array, each
regiment and each rank occurs precisely once?
3
Permutations will be described in Chapter 3.
 How to use this book 3

DISCUSSION. Euler posed this problem in 1782; he believed that the answer was 'no'.
This was not proved until 1900, by Tarry. Again, the problem can be generalised, to
n2 officers, where the number of regiments, ranks, rows and columns is n (we assume
n > 1) — see Exercise 3. There is no solution for n = 2. Euler knew solutions for
all n not congruent to 2 modulo 4, and guessed that there was no solution for n = 2
(mod 4). However, he was wrong about that. Bose, Shrikhande and Parker showed
in 1960 that there is a solution for all n except n = 2 and n = 6.

A Ramsey game

This two-player game requires a sheet of paper and pencils of two

colours, say red and blue. Six points on the paper are chosen, with
no three in line. Now the players take a pencil each, and take turns
drawing a line connecting two of the chosen points. The first player
to complete a triangle of her own colour loses. (Only triangles with
vertices at the chosen points count.)
Can the game ever result in a draw?
DISCUSSION. We'll see that a draw is not possible; one or other player will be forced
to create a triangle. Ramsey proved a wide generalisation of this fact. His theorem
is sometimes stated in the form 'Complete disorder is impossible.'

How to use this book

1. The book is divided into two parts: Chapters 2-11 and Chapters 12-20. In the
second part, along with some new material, we revisit many of the topics from the
first part and treat them from a more advanced viewpoint; also, as I mentioned
earlier, the pace is a little faster in the second part. In any case, a first course can be
devised using only the first part of the book. (The second-third year undergraduate
course at Queen Mary and Westfield College includes a selection of material from
Chapters 3 (Sections 3.1, 3.2, 3.3, 3.5, 3.7, 3.11, 3.12), 4 (Sections 4.1, 4.3, 4.4, 4.5), 5,
6, 7, 8 and 10; other courses treat material from Chapters 9, 11, 14-17.)
2. Chapter 3 plays a special role. The material here is central to combinatorics:
subsets, partitions, and permutations of finite sets. Within the other chapters, you are
encouraged to dabble, taking or leaving sections as you choose; but I recommend
reading all of Chapter 3 (except perhaps the Projects, see below).
3. A number of sections are designated as Projects. These are to be regarded as
less central and possibly more difficult than the others. The word suggests that
they could be worked through by individuals outside class time, and then made the
subject of presentations to the class.
4. Each chapter after this one begins with a box containing 'topics, techniques,
algorithms and cross-references'. This is designed to give you some indication of the
scope of the chapter. Roughly speaking, topics are specific results or constructions;
techniques are of wider applicability, indicating general methods which may be
illustrated in specific cases in the chapter; algorithms are self-explanatory; and
4 1. What is Combinatorics?

cross-references pinpoint at least some occurrences of the material in other chapters.

These are usually backward references, but the multidimensional nature of the
subject means that this is not always so. You should use these as pointers to places
where you might find help if you are stuck on something. The index can also be
used for this purpose.
5. The exercises are a mixed bunch; but, by and large, I have tended to avoid
'drill' and give more substantial problems. You will certainly learn more if you work
conscientiously through them. But I have tried not to assume that you have done all
the problems. When (as often happens) the result of an exercise is needed in a later
chapter, I have usually supplied a proof (or, failing that, a hint). Indeed, hints are
strewn liberally through the exercises, and some example solutions are given (rather
more briefly than I would expect from students!) at the end of the book.
6. The last chapter does two jobs. First, it treats (somewhat sketchily) some further
topics not mentioned earlier; second, it gives pointers to further reading in various
parts of combinatorics. I have included a small collection of unsolved problems
here, to indicate the sort of thing that research in combinatorics might involve. But
beware: these problems are unsolved; this means that somebody has given some
thought to them and failed to solve them, so they are probably more difficult than
the exercises in other chapters.
7. The numbering is as follows. Chapter A is divided into sections, of which a
typical one is Section A.B. Within a section, theorems (and similar statements such
as propositions, lemmas, corollaries, facts, algorithms, and numbered equations)
have numbers of the form A.B.C. On the other hand, diagrams are just numbered
within the chapter, as A.D, for example; and exercises are typically referred to
as 'exercise E of Chapter A'. Some theorems or facts are displayed in a box for
easy reference. But don't read too much into the difference between displayed and
undisplayed theorems, or between theorems and propositions; it's a matter of taste,
and consistency is not really possible.
8. An important part of combinatorics today is the algorithmic side: I can prove that
some object exists; how do I construct it? I have described algorithms for a wide
range of constructions. No knowledge of computers or programming languages is
assumed. The description of the algorithms makes use of words like 'While ...',
'Repeat . . . until ...', and so on. These are to be interpreted as having their usual
English meaning. Of course, this meaning has been taken over by programming
languages; if you are fluent in Pascal, you will I hope find my descriptions quite
congenial. If you are a competent programmer and have access to a computer, you
are advised at several places to implement these algorithms.

What you need to know

The mathematical results that I use are listed here. You don't need everything all
at once; the more advanced parts of algebra, for example, are only required later
in the book, so you could study algebra and combinatorics at the same time. If
all else fails, I have tried to arrange things so that you can take on trust what you
don't know. Topics in square brackets are treated in the book, but you may feel the
 Exercises 5

need of more explanation from a course or textbook in that subject. As you see,
combinatorics connects with all of mathematics; you will see material from many
other areas being used here.
• Basic pure mathematics: Sets and functions, ordered n-tuples and cartesian
products; integers, factorisation, modular arithmetic; [equivalence and order
relations].
• Linear algebra: Vector spaces, subspaces; linear transformations, matrices; row
operations, row space; eigenvalues of real symmetric matrices.
• Abstract algebra: [Elementary group theory; finite fields].
• Number theory: [Quadratic residues; two and four squares theorems].
• Analysis: Basic operations (limits, differentiation, etc.); [power series].4
• Topology: [Definition of metric and topological space; surfaces; Jordan curve
theorem].
• Probability: Basic concepts (for finite spaces only) [except in Chapter 19].
• Set theory: See Chapter 19.

Exercises
1. For n — 3,4,5, calculate the number of ways of putting n letters into their
envelopes so that every letter is incorrectly addressed. Calculate the ratio of this
number to n! in each case.
2. Solve Kirkman's problem for nine schoolgirls, walking for four days.
3. Solve Euler's problem for nine, sixteen and twenty-five officers. Show that no
solution is possible for four officers.
4. Test the assertion that the Ramsey game cannot end in a draw by playing it with
a friend. Try to develop heuristic rules for successful play.

4
As will be explained in Section 4.2, our treatment of power series is formal and does not involve
questions of convergence.
2. On numbers and counting

One of them is all alone and ever more shall be so

Two of them are lily-white boys all clothed all in green Oh
Three of them are strangers o'er the wide world they are rangers
Four it is the Dilly Hour when blooms the Gilly Flower
Five it is the Dilly Bird that's seldom seen but heard
Six it is the ferryman in the boat that o'er the River floats Oh
Seven are the Seven Stars in the Sky, the Shining Stars be Seven Oh
Eight it is the Morning's break when all the World's awake Oh
Nine it is the pale Moonshine, the Shining Moon is Nine Oh
Ten Forgives all kinds of Sin, from Ten begin again Oh
English traditional folksong
from Bob Stewart, Where is Saint George? (1977)

TOPICS: Natural numbers and their representation; induction; use-

ful functions; rates of growth; counting labelled and unlabelled
structures; Handshaking Lemma
TECHNIQUES: Induction; double counting
ALGORITHMS: Odometer Principle; [Russian peasant multiplication]
CROSS-REFERENCES:

This chapter is about counting. In some sense, it is crucial to what follows, since
counting is so basic in combinatorics. But this material is part of mathematical
culture, so you will probably have seen most of it before.

2.1. Natural numbers and arithmetic

Kronecker is often quoted as saying about mathematics, 'God made the integers;
the rest is the work of man.' He was referring to the natural numbers (or counting
numbers), which are older than the earliest archaeological evidence. (Zero and the
negative numbers are much more recent, having been invented (or discovered) in
historical time.)1 Since much of combinatorics is concerned with counting, the
natural numbers have special significance for us.

1
See Georges Ifrah, From One to Zero: A Universal History of Numbers (1985), for an account of
the development of numbers and their representation.
8 2. On numbers and counting

As each new class of numbers was added to the mathematical repertoire, it was
given a name reflecting the prejudice against its members, or the 'old' numbers were
given a friendly, reassuring name. Thus, zero and negative integers are contrasted
with the 'natural' positive integers. Later, quotients of integers were 'rational', as
opposed to the 'irrational' square root of 2; and later still, all numbers rational and
irrational were regarded as 'real', while the square root of —1 was 'imaginary' (and
its friends were 'complex').

The natural numbers are the first mathematical construct with which we become
familiar. Small children recite the names of the first few natural numbers in the same
way that they might chant a nursery rhyme or playground jingle. This gives them
the concept that the numbers come in a sequence. They grasp this in a sophisticated
way. The rhyme2

One, two,
Missed a few,
Ninety-nine,
A hundred

expresses confidence that the sequence of numbers stretches at least up to 100, and
that the speaker could fill in the gap if pressed.

Order or progression is thus the most basic property of the natural numbers.3
How is this expressed mathematically? First we must stop to consider how natural
numbers are represented. The simplest way to represent the number n is by a
sequence of n identical marks. This is probably the earliest scheme mankind
adopted. It is well adapted for tallying: to move from one number to the next,
simply add one more mark. However, large numbers are not easily recognisable.
After various refinements (ranging from grouping the marks in sets of five to the
complexities of Roman numerals), positional notation was finally adopted.
This involves the choice of a base b (an integer greater than 1), and b digits (dis-
tinguishable symbols for the integers 0,1,2,. . . , 6 — 1). (Early attempts at positional
notation were bedevilled because the need for a symbol for zero was not recognised.)
Now any natural number N is represented by a finite string of digits. Logically the
string is read from right to left; so we write it as xn-i .. .Xixo, where each X{ is
one of our digits. By convention, the leftmost digit is never zero. The algorithm for
advancing to the next number is called the Odometer Principle. It is based on the
principle of trading in b counters in place i for a single counter in place i + 1, and
should be readily understood by anyone who has watched the odometer (or mileage
gauge) of a car.

2
I have heard the feminist version of this: 'One, two, Mrs. Few, ...'
3
'The operations of arithmetic are based on the tacit assumption that we can always pass from any
number to its successor, and this is the essence of the ordinal concept.' Tobias Dantzig, Number: the
Language of Science (1930).
 2.1. Natural numbers and arithmetic

(2.1.1) Odometer Principle

to find the successor of a natural number to base b
Start by considering the rightmost digit.
• If the digit we are considering is not 6 — 1 , then replace it by
the next digit in order, and terminate the algorithm.
• If we are considering a blank space (to the left of all the digits),
then write in it the digit 1, and terminate the algorithm.
• If neither of the above holds, we are considering the digit 6—1.
Replace it with the digit 0, move one place left, and return to
the first bullet point.

For example, if the base 6 is 2 and the digits are 0 and 1, the algorithm (starting
with 1) generates successively 10, 11, 100, 101, 110, . . . .
Now it can be proved by induction that the string xn-\ . . . X^XQ represents the
positive integer
xn-ibn~l H h xxb + x0
(see Exercise 2).
Often the number 0 is included as a natural number. (This is most usually done
by logicians, who like to generate the whole number system out of zero, or nothing.
But it conflicts with our childhood experience: I have never heard a child say
'nought, one, two, ... '4, and we don't count that way.) This is done by modifying
our representation so that the digit 0 represents the number 0. This is the one
allowed exception to the rule that the left-most digit cannot be 0; the alternative,
representing 0 by a blank space, would be confusing.
The odometer of a car actually works slightly differently. It works with a fixed
number of digits which are initially all zero, so that the 'blank space' case of the
algorithm cannot arise. If there are k digits, then the integers 0,... , bk — 1 are
generated in turn, and then the odometer returns to 0 and the process repeats.
Now that we have a representation of positive integers, and understand how to
move to the next integer, we should explore the arithmetic operations (ambition,
distraction, uglification and derision).5 Algorithms for these are taught in primary
school.6 I will not consider the details here. It is a good exercise to program
a computer to perform these algorithms7, or to investigate how many elementary
4
A possible exception occurs when one child has been appointed to be first, and another wishes to
claim precedence, as in 'Zero the hero'. But this is closer to the historical than the logical approach.
5
Lewis Carroll, Alice's Adventures in Wonderland (1865).
6
These algorithms were known to the Babylonians in 1700 B.C.
7
Most programming languages specify the 'maximum integer' to be something like 32767 or
2147483647. Often, the answer to a counting problem will be much larger than this. To find it by
computer, you may have to write routines for arithmetic operations on integers with many digits. If
you need to do this, write your routines so that you can re-use them!
10 2. On numbers and counting

operations are required to add or multiply two ra-digit numbers (where elementary
operations might consist of referring to one's memory of the multiplication tables,
or writing down a digit).
2.2. Induction
Induction is a very powerful principle for proving assertions about the natural
numbers. It is applied in various different forms, some of which are described in
this section. We also see that it is a consequence of our most basic intuition about
the natural numbers.
The Principle of Induction asserts the following:

(2.2.1) Principle of Induction

Let P(n) be a proposition or assertion about the natural number n.
Suppose that P(l) is true. Suppose also that7 if P(n) is true, then
P(n + 1) is also true. Then P(n) is true for all natural numbers n.

Why is this true? As we saw, the basic property of the natural numbers,
recognised even by children, is that we can count up to any natural number n
starting from 1 (given sufficient patience!) Now, with the assumptions of the
Principle, P(l) is true, so P(2) is true, so (miss a few here) so P(n — 1) is true, so
P{n) is true.
As this argument suggests, if you are reading a mathematical argument, and the
author puts in a few dots or the words 'and so on', there is probably a proof by
induction hiding there. Consider, for example, the function / satisfying /(I) = 2
and f(n + 1) = 2/(n) for all natural numbers n. Then
/(2) = 4 = 2 2 ,/(3) = 8 = 23, ... /(n)=2».
The dots hide a proof by induction. Let P(n) be the assertion that f(n) = 2 n . Then
P(l) holds; and, assuming that P(n) holds, we have
P(n + 1) = 2P(n) = 2 • 2n = 2 n + \
so P(n + 1) also holds. So the Principle of Induction justifies the conclusion. The
point is that very simple arguments by induction can be written out with three
dots in place of the detailed verification, but this verification could be supplied if
necessary. We'll see more examples of this later.
Now I give some alternative forms of the Principle of Induction and justify their
equivalence. The first one is transparent. Suppose that P(n) is an assertion, for
which we know that P(27) is true, and that if P(n) holds then so does P(n + 1).
Then we conclude that P(n) holds for all n > 27. (To prove this formally, let Q{n)
be the assertion that P(n + 26) is true, and verify the hypotheses of the Principle of
Induction for Q(n).)
For the next variation, let P{n) be a proposition about natural numbers. Suppose
that, for every natural number n, if P(m) holds for all natural numbers m less than
 2.3. Some useful functions 11

n, then P(n) holds. Can we conclude that P(n) holds for all n? On the face of it,
this seems a much stronger principle, since the hypothesis is much weaker. (Instead
of having to prove P(n) from just the information that P(n — 1) holds, we may
assume the truth of P(m) for all smaller m.) But it is true, and it follows from the
Principle as previously stated.
We let Q(n) be the statement 'P(ra) holds for all m < n\ Now it is clear that
Q(n + 1) implies P(n), so we will have succeeded if we can prove that Q(n) holds
for all n. We prove this by induction.
First, Q(l) holds: for there are no natural numbers less than 1, so the assertion
P holds for all of them (vacuously).
Now suppose that Q(n) holds. That is, P(ra) holds for all m < n. By assumption,
P(n) also holds. Now P(m) holds for all m < n + 1 (since the numbers less than
n + 1 are just n and the numbers less than n) 8. In other words, Q(n + 1) holds.
Now the Principle of Induction shows that Q(n) holds for all n.
The final re-formulation gives us the technique of 'Proof by Minimal Counterex-
ample'. Suppose that P(n) is a proposition such that it is not true that P(n) holds
for all natural numbers n. Then there is a least natural number n for which P(n) is
false; in other words, P(m) is true for all m < n but P(n) is false. For suppose that
no such n exists; then the truth of P(m) for all m < n entails the truth of -P(n),
and as we have seen, this suffices to show that P(n) is true for all n, contrary to
assumption.
This argument shows that any non-empty set of natural numbers contains a
minimal element. (If S is the set, let P(n) be the assertion n ^ S.)
2.3. Some useful functions
I assume that you are familiar with common functions like polynomials, the function
\x\ (the absolute value or modulus), etc.
Floor and Ceiling. The floor of a real number x, written [x\ > is the greatest integer
not exceeding x. In other words, [^J is the integer m such that m < x < m + l. li x
is an integer, then \x\ = x. This function is sometimes written [x\\ but the notation
\x\ suggests 'rounding down'. It is the number of the floor of a building on which
x would be found, if the height of x above the ground is measured in units of the
distance between floors. (The British system of floor numbering is used, so that the
ground floor is number 0.)
The ceiling is as you would probably expect: \x~\ is the smallest integer not less
than x. So, if x is not an integer, then \x] = [x\ + 1; if x is an integer, its floor and
ceiling are equal. In any case, you can check that

Factorial. The factorial function is defined on positive integers by the rule that n! is
the product of all the integers from 1 to n inclusive. It satisfies the condition

8
Let p be an integer less than n + 1. Then p < n oi p = n OT p > n; and the last case is impossible,
since there is no integer between n and n + 1.
12 2. On numbers and counting

for n > 1. In fact, we can consistently define 0! = 1; then (*) holds for all n > 0.
In fact, the conditions 0! = 1 and (*) actually define n\ for all natural numbers n.
(This is proved by induction: 0! is defined; if n! is defined then so is (n + 1)!; so n\
is defined for all n.)
Exponential and logarithm. These two functions are familiar from elementary
calculus. We will often use the power series expansions of them. The equation
oo n 2
e* - V" — - 1 + Z + — +
71=0

is valid for all real numbers x. On the other hand, the function log re can't be
expanded as a series of powers of z, since log 0 is undefined. Instead, we have

71=1

which is valid for all x with \x\ < 1 (and in fact also for x = 1).
The exponential function grows more rapidly than any power of x] this means
that ex > xc for all sufficiently large x (depending on c). In fact, for x > (c + 1)!, we
have
xc+1

On the other hand, the logarithm function grows more slowly than any power of x.
We will often write exp(z) instead of ex.

2.4. Orders of magnitude

People use the phrase 'the combinatorial explosion' to describe a counting function
which grows very rapidly. This is a common phenomenon, and it means that, while
we may be able to give a complete description of all the objects being counted
for small values of the parameter, soon there will be far too many for this to be
possible, and maybe even far too many for an exact count; we may have to make do
with fairly rough estimates for the counting function. I will consider now what such
rough estimates might look like. In this section, some results from later chapters will
be anticipated. If you are unfamiliar with these, take them on trust until we meet
them formally.
Let X be a set with n elements, say X = {1,2,... ,n}. The number of subsets
of X is 2 n . This is the most familiar example of an exponential function, or function
with exponential growth. A function / which has (precisely) exponential growth has
the property that
f(n + 1) = cf{n)
for some c > 1. (If c = 1, the function is constant; if c < 1, then f(n) —> 0 as n —> oo.
In these cases, the term 'exponential growth! is not really appropriate! 9 ) A function
9
Economists define a recession as a period when the exponential constant for the GDP is less than
1.004. Sometimes you have to run in order to stand still.
 2.4. Orders of magnitude 13

/ satisfying the above equation is given by f(n) = acn, where a is a constant (and
is equal to the value of /(0)).
We also say that a function / has 'exponential growth' if it is roughly the same
size as an exponential function. So the function f(n) = 2n + n has exponential
growth, since the term n is dwarfed by 2 n for large n. Formally, the function / is
said to have exponential growth if f(n)lln tends to a limit c > 1 as n —> oo. This
means that, for any positive number e, f(n) lies between (c — e)n and (c + e)n for all
sufficiently large n. The number c is called the exponential constant for / .
Of course, a function may grow more slowly than exponentially. Examples
include
• polynomial growth with degree c, like the function f(n) = nc;
• fractional exponential growth with exponent c, like the function enC, where 0 <
c< 1.
These functions arise in real combinatorial counting problems, as we will see.
But many functions grow faster than exponentially. Here are two examples.
The number of permutations of the set X is equal to n\ = n(n — 1 ) . . . 1, the
product of the integers from 1 t o n inclusive. We have
2n~l <n\<nn-\
because (ignoring the factor 1) there are n — 1 factors, each lying between 2 and
n. In fact it is easy to see that the growth is not exponential. We will find better
estimates in the next chapter.
Now let V(X), the power set of X, denote the set of all subsets of X. We will
be considering subsets of V(X), under the name families of sets. How many families
of sets are there? Clearly the number is 22". This number grows much faster than
exponentially, and much faster than the factorial function. A function like this is
called a double exponential.
For comparing the magnitudes of functions like these, it is often helpful to
consider the logarithm of the function, rather than the function itself. The logarithm
of an exponential function is a (roughly) linear function. The logarithm of n! is
fairly well approximated by nlogn; and the logarithm of a double exponential is
exponential. Other possibilities are functions whose logarithms are polynomial.
Of course, this is only the beginning of a hierarchy of growth rates; but for the
most part we won't have to consider anything worse than a double exponential.
In connection with growth rates, there is a convenient analytic notation. We
write O(f(n)) (read 'big Oh /(n)') to mean a (possibly unknown) function g(n) such
that, for all sufficiently large n, \g(n)\ < cf(n) for some constant c. This is typically
used in the form

where (f> is a combinatorial counting function and F: f are analytic functions where
/ grows more slowly than F] this has the interpretation that the order of magnitude
of <j) is similar to that of F. For example, in Section 3.6, we show that
log n! = n log n — n + O(log n).
14 2. On numbers and counting

We write o(/(n)) (and say 'small oh /(ft)') to man a function g(n) such that
g{n)l f(n) —> 0 as n —> oo; that is, g is of smaller order of magnitude than / .
There are several variants. For example, Ct is the opposite of 0; that is, fi(/(n))
is a function g(n) with |#(^)| > cf(n) for some constant c > 0. Also, g(n) ~ / ( n )
means that both <?(n) = O(f(n)) and #(n) = fi(/(n)) hold: roughly, / and # have
the same order of magnitude apart from a constant factor.

2.5. Different ways of counting

In combinatorics (unlike real life10), when we are asked to count something, there
are very many different answers which can be regarded as correct. Consider the
simple problem of choosing three items from a set of five. Before we can work out
the right answer, the problem must be specified more precisely. Are the objects in
the set identical (five electrons, say, or five red billiard balls), or all different (the
ace, two, three, four, and five of spades, for example)? Does the order of selection
matter? (That is, do we just put in a hand and pull three objects out, or do we draw
them one at a time and record the order?) And are we allowed to choose the same
object more than once (say, by recording the result of each draw and returning the
object to the urn), or not? There are various intermediate cases, like making words
using the letters of a given word, where a letter may be repeated but not more often
than it occurs in the original word.

Almost always, we assume that the objects are distinguishable, like the five
spade cards. Under this assumption, the problem will be solved under the four
possible combinations of the other assumptions in Chapter 3. What if they are
indistinguishable? In this case, there is obviously only one way to select three red
billiard balls from a set of five: any three red billiard balls are identical to any other
three.
i

What difference does indistinguishability make? If the underlying objects are

distinguishable, we can assume that they carry labels bearing the numbers 1 , 2 , . . . , n.
In this case, we say that the configurations we are counting are labelled. If the n
underlying objects are indistinguishable, we are counting unlabelled things. An
example will illustrate the difference.

Suppose that we are interested in n towns; some pairs of towns are joined by
a direct road, others not. We are not concerned with the geographical locations,
only in whether the towns are connected or not. (This is described by the structure
known as a graph.11 See Chapter 11 for more about graphs.) Figure 1 shows the
eight labelled graphs for n = 3. If the towns are indistinguishable, then the second,

10
According to folklore, it is impossible to count the Rollright Stones consistently.
11
This usage of the term is quite different from the sense in the phrase 'the graph of y — sin a?'.
Some people distinguish the two meanings by different pronunciation, with a short a for the sense
used here.
 2.6. Double counting 15

A A /. A A A A A
23 23 23 23 23 23 23 23

Fig. 2.1. Graphs on three vertices

third and fifth graphs are identical, as are the fourth, sixth and seventh. So there
are just four unlabelled graphs with n = 3.
In general, let f(n) and g(n) denote the numbers of labelled and unlabelled
configurations, respectively, with n underlying objects. Then two labelled configura-
tions will be regarded as identical as unlabelled configurations if and only if there
is a permutation of { 1 , 2 , . . . , n } which carries one to the other. (For example, the
cyclic permutation 1 —i > 2 h-> 3 — i > 1 carries the second graph in Fig. 1 to the fifth.)
So at most n! labelled configurations collapse into a single unlabelled one, and we
have
f(n)/n\ < g(n) < f(n).
Now there are two possibilities for the 'order of magnitude' behaviour.
If f(n) grows much more rapidly than n!, then the left and right hand sides
of this equation are not so very far apart, and we have a reasonable estimate for
g(n). For example, we saw that there are 22" families of subsets of the n-element set
X. The number of permutations is insignificant by comparison, so it doesn't matter
very much whether the elements of X are distinguishable or not, that is, whether we
count labelled or unlabelled families.
But if this doesn't occur, then more care is needed. There are just 2 n subsets of
the n-element set X, and this function grows more slowly than n!. In this case, we
can count unlabelled sets another way. If all elements of X are indistinguishable,
then the only thing we can tell about a subset of X is its cardinality; two subsets
containing the same number of elements are equivalent under a permutation. So the
number of unlabelled subsets is n + 1, since the cardinality of a subset can take any
one of the n + 1 values 0,1, 2 , . . . , ra.
This theme can be refined, using the concepts of permutation group and cycle
index. These are more advanced topics, and will be treated in Part 2 (see Chapter 15).

2.6. Double counting

We come now to a deceptively simple but enormously important counting principle:

If the same set is counted in two different ways, the answers are the
same.

This is analogous to finding the sum of all the entries in a matrix by adding the row
totals, and then checking the calculation by adding the column totals.
The principle is best illustrated by applications (of which there will be many
later) — here is one:
16 2. On numbers and counting

(2.6.1) Handshaking Lemma

At a convention, the number of delegates who shake hands an odd
number of times is even.

To show this, let D\,..., Dn be the delegates. We apply double counting to the
set of ordered pairs (Di,Dj) for which D{ and Dj shake hands with each other at
the convention. Let X{ be the number of times that D{ shakes hands, and y the
total number of handshakes that occur. On the one hand, the number of pairs is
YZ=i xii since for each D{ the number of choices of Dj is equal to X{. On the other
hand, each handshake gives rise to two pairs (£),-,Dj) and (Dj,D{)] so the total is
2y. Thus

But, if the sum of n numbers is even, then evenly many of the numbers are odd.
(If we add an odd number of odd numbers and any number of even numbers, the
answer will be odd.)
The double counting principle is usually applied to counting ordered pairs.
For lovers of formalism, here is a general result, which encapsulates most of the
applications we will make of it.

(2.6.2) Proposition. Let A = {«i,..., am} and B = {&i,..., bn) be sets. Let S be a
subset of A x B. Suppose that, for i — 1,..., m, the element at is the first component
of X{ pairs in S, while, for j = 1,..., n, the element bj is the second component of
yj pairs in S. Then
m n

t=l j=l

Often it happens that X{ is constant (say x) and yj is also constant (say y). Then
we have
mx = ny.

2.7. Appendix on set notation

The basic notation for sets is listed here. If A and B are sets, then we write x G A
if x is an element of A, x $ A otherwise. Also
|-A| (the cardinality of A) is the number of elements in A]
A U B (the union) is the set of elements in A or B (or both);
An B (the intersection) is the set of elements in both A and B\
A\B (the difference) is the set of elements in A but not Z?;
AAB (the symmetric difference) is the set of elements in just one of the two sets;
A C B if every element of A belongs to B;
A = B if A and B have exactly the same elements.
 2.8. Exercises 17

So, for example,

AAB = (A\B)U(B\A) = (A\JB)\(An B),

The notation {x : P} means the set of all elements x having property P. So, for
example,
AU B = {x : x e A oi x <E B}.
Similarly, {x, y} is the set consisting of the elements x and y only. It is sometimes
called an unordered pair, since {x,y} = {y,x}. By contrast, the ordered pair (x,y)
has the property that (x:y) — (it, v) if and only if x = u and y — v. This is familiar
from Cartesian coordinates of points in the Euclidean plane.
The Cartesian product A x B is the set of all ordered pairs (a, b), with a G A
and b £ B. Similarly for more than two factors. For example, we write An for the
set of ordered n-tuples of elements of A, for any positive integer n. We have
\AxB\ = \A\-\B\,
\An\ = \A\n.
Until last century, a function was something described by a formula (typically a
polynomial or a power series); it was the ambiguity in this definition which led to
the modern version. A function f from A to B is a subset of A x B with the property
that, for any a G A, there is a unique 6 G B such that (a, b) G / . If (a, b) G / , we
write / ( a ) = 6.12 Usually there is a rule for calculating 6 = f(a) from a, but this is
not part of the definition.
If A = {ai, a<i, • • •, an}, then any function f : A —> B can be specified by giving
the n-tuple of values (/(«i), f ^ ) , • • • ? /(«n))- Thus the number of functions from
A to B is |B|'^L Motivated by this, the set of functions from A to B is sometimes
written BA, so that \BA\ = |J3|W.
The power set V(A) is the set of all subsets of A. Any subset X of A is specified
by its characteristic function, the function fx '. A —» {0,1} defined by
1 if a G X;

(Two subsets are equal if and only if their characteristic functions are equal.) So
there are as many subsets of A as there are functions from A to {0,1}; that is,

2.8. Exercises
1. Criticise the following proof that 1 is the largest natural number.
Let n be the largest natural number, and suppose than n ^ 1. Tien
n > 1, and so n2 > n; thus n is not the largest natural number.
12
This definition is very familiar, despite appearances. You probably visualise 'the function y — x2' in
terms of its graph in the Euclidean plane with coordinates (x, y); and the graph consists of precisely
those ordered pairs (x,y) for which y = x2. In other words, the graph is the function!
18 2. On numbers and counting

2. Prove by induction that the Odometer Principle with base b does indeed give the
representation xn-\ . . . X\x$ for the natural number

3. (a) Prove by induction that

for n > 1. (You may use the fact that (1 + \)n < e for all n.)
)
(b) Use the arithmetic-geometric mean inequality to show that n! < (ZL2~^)n f° r
13

n > 1, and deduce that

for n > 1.
4. (a) Prove that log x grows more slowly than xc for any positive number c.
(b) Prove that, for any c, d > 1, we have cx > xd for all sufficiently large x.
5. (a) We saw that there are 223 = 256 labelled families of subsets of a 3-set. How
many unlabelled families are there?
(b) Prove that the number F(n) of unlabelled families of subsets of an n-set
satisfies log2 F(n) = T + 0(n log n).
6. Verity that the numbers of graphs are given in Table 1 for n < 5.

n 2 3 4 5
labelled 2 8 64 1024
unlabelled 2 4 11 34
Table 2.1. Graphs
7. Suppose that an urn contains four balls with different colours. In how many
ways can three balls be chosen? As in the text, we may be interested in the order
of choice, or not; and we may return balls to the urn, allowing repetitions, or not.
Verify the results of Table 2.
order order
important unimportant
repetition &A <\r\
b 4 Z[)
allowed

repetition <\A A
not allowed
Table 2.2. Selections
8. A Boolean function takes n arguments, each of which can have the value TRUE
or FALSE. The function takes the value TRUE or FALSE for each choice of values of
its arguments. Prove that there are 22" different Boolean functions. Why is this the
same as the number of families of sets?
13
The arithmetic-geometric mean inequality states that the arithmetic mean of a list of positive
numbers is greater than or equal to their geometric mean, with equality only if all the numbers are
equal. Can you prove it? (HINT: DO the special case when all but one of the numbers are equal by
calculus, and then the general case by induction.)
 2.8. Exercises 19

9. Logicians define a natural number to be the set of all its predecessors: so 3 is the
set {0,1,2}. Why do they have to start counting at 0?
10. A function / has polynomial growth of degree d if there exist positive real numbers
a and b such that and < f(n) < bnd for all sufficiently large n. Suppose that / has
polynomial growth, and g has exponential growth with exponential constant greater
than 1 (as defined in the text). Prove that f(n) < g(n) for all sufficiently large n. If
f(n) = 106n106 and g{n) = (1.000001)n, how large is 'sufficiently large'?
11. Let B be a set of subsets of the set {1,2, . . . , u } , containing exactly b sets.
Suppose that
• every set in B contains exactly k elements;
• for i = 1,2,..., v, the element i is contained in exactly r members of B.
Prove that bk = vr.
Give an example of such a system, with v = 6, k = 3, b = 4, r = 2.
12. The 'Russian peasant algorithm' for multiplying two natural numbers m and n
works as follows.14

(2.7.3) Russian peasant multiplication

to multiply two natural numbers m and n
Write m and n at the head of two columns,
REPEAT the sequence
• halve the last number in the first column (discarding the re-
mainder) and write it under this number;
• double the last number in the second column and write it under
this number;
UNTIL the last number in the first column is 1.
For each even number in the first column delete the adjacent
entry in the second column. Now add the remaining numbers in
the second column. Their sum is the answer.

For example, to calculate 18 x 37:

18 £?
9 74
4 148
2 206
1 592

666
Table 2.3. Multiplication
PROBLEMS, (i) Prove that this method gives the right answer.
14
No tables needed, except two times!
20 2. On numbers and counting

(ii) What is the connection with the primary school method of long multiplication?
HINT FOR (i) AND (ii): Express m (and n) to the base 2.
(iii) Suppose we change the algorithm by squaring (instead of doubling) the numbers
in the second column, and, in the last step, multiplying (rather than adding)
the undeleted numbers. Prove that the number calculated is nm. How many
multiplications does this method require?
13. According to the Buddha,

Scholars speak in sixteen ways of the state of the soul after death.
They say that it has form or is formless; has and has not form,
or neither has nor has not form; it is Unite or infinite; or both or
neither; it has one mode of consciousness or several; has limited
consciousness or infinite; is happy or miserable; or both or neither.

How many different possible descriptions of the state of the soul after death do you
recognise here?
14. The library of Babel15 consists of interconecting hexagonal rooms. Each room
contains twenty shelves, with thirty-five books of uniform format on each shelf.
A book has four hundred and ten pages, with forty lines to a page, and eighty
characters on a line, taken from an alphabet of twenty-five orthographical symbols
(twenty-two letters, comma, period and space). Assuming that one copy of every
possible book is kept in the library, how many rooms are there?
15. COMPUTER PROJECT. Develop a suite of subroutines for performing arithmetic on
integers of arbitrary size, regarded as strings of digits. (You should deal with input
and output, arithmetic operations — note that division should return a quotient
and a remainder — and comparisons. You might continue with exponentiation and
factorials, as well as various combinatorial functions to be defined later.)

Jorge Luis Borges, Labyrinths (1964).

3. Subsets, partitions, permutations

The emphasis on mathematical methods seems to be shifted more towards

combinatorics and set theory — and away from the algorithm of differential
equations which dominates mathematical physics.
J. von Neumann & 0. Morganstern,
Theory of Games and Economic Behaviour (1944).

The process is directed always towards analysing and separating the material
into a collection of discrete counters, with which the detached intellect can
make, observe and enjoy a series of abstract, detailed, artificial patterns of
words and images (you may be reminded of the New Criticism)...
Elizabeth Sewell, 'Lewis Carroll and T. S. Eliot as Nonsense Poets'
in Neville Braybrooke (ed.), T. S. Eliot (1958).

TOPICS: Subsets, binomial coefficients, Pascal's Triangle, Binomial

Theorem; [congruences of binomial coefficients]; permutations, or-
dered and unordered selections, cycle decomposition of a permuta-
tion; estimates for the factorial function; relations; [finite topolo-
gies; counting trees]; partitions, Bell numbers
TECHNIQUES: Binomial coefficient identities; use of double counting;
estimates via integration
ALGORITHMS: Sequential and recursive generation of combinatorial
objects
CROSS-REFERENCES: Odometer Principle; double counting (Chap-
ter 2); recurrence relations (Chapter 4)

This chapter is about the central topic of 'classical' combinatorics, what is often
referred to as 'Permutations and Combinations'. Given a set with n elements, how
many ways can we choose a selection of its elements, with or without respect to the
order of selection, or divide it up into subsets? We'll define the various numbers
involved, and prove some of their properties; but these echo through subsequent
chapters.
22 3. Subsets, partitions, permutations

3.1. Subsets
How many subsets does a set of n elements have?
The number of subsets is 2n. There are several different ways to see this. Perhaps
most easily, for each of the n elements of the set, there are two choices in building
a subset (viz., put the element in, or leave it out); all combinations of these choices
are possible, giving a total of 2 n .
Implicitly, this argument sets up a bijection between the subsets of a set X and
the functions from X to {0,1}. The function fy corresponding to the subset Y is
defined by the rule
/ 1 if x £ Y

Conversely, a function / corresponds to the set Y — {x G X : f(x) = 1}. The

function fy is called the characteristic function or indicator function of Y.
If X = { 0 , 1 , . . . , n — 1}, then we can represent a function / : X —> {0,1} by
the n-tuple (/(0), / ( I ) , . . . , f(n — 1)) of its values. Thus subsets of X correspond to
n-tuples of zeros and ones.
We can take this one step further, and regard the n-tuple as the base 2 repre-
sentation of an integer

as described in Chapter 2. Each n-tuple corresponds to a unique integer; the smallest

is 0 (corresponding to the empty set), and the largest is 2n~1 + . . . + 2 + l = 2n — 1
(corresponding to the whole set X), and every integer between represents a unique
subset. So the number of subsets is equal to the number of integers between 0 and
2n - 1 (inclusive), namely 2 n .
Note that this method gives a convenient numbering of the subsets of the set
{0,...,n — 1}: the fcth subset Xk corresponds to the integerfc,where 0 < k < 2n — 1.
The set Xk is easily recovered by writing k to base 2. The numbering has some
further virtues. For example, the set Xk depends only on k, and not on the particular
value of n used; replacing n by a larger value doesn't change it. So we get a unique
set Xk of non-negative integers corresponding to each non-negative integer k. For
another nice property, see Exercise 2.
Yet another proof of the formula for the number F(n) of subsets of an n-set is
obtained by noting that we can find all subsets of { 1 , . . . , n + 1} by taking all subsets
of { 1 , . . . ,ra} and extending each in the two possible ways — either do nothing,
or add the element n + 1. So F(n + 1) = 2F(n). This is a recurrence relation, by
which the value of F is determined by its values on smaller arguments. Recurrence
relations form the subject of the next chapter.

3.2. Subsets of fixed size

Let n and k be non-negative integers, with 0 < k < n. The binomial coefficient \Jj is
denned to be the number offc-elementsubsets of a set of n elements. (The number
obviously doesn't depend on which n-element set we use.) This number is often
 3.2. Subsets of fixed size 23

written as nCk, and is read 'n choose k\ It is called a binomial coefficient (for reasons
to be elaborated later).

(3.2.1) Formula for binomial coefficients

n\ _ n(n - 1)... (n - k + 1) _ n!
k) Jb(Jfc-l)...! k\(n-k)\

Note that (jj) = 1 (the empty set) and (*) = 1 (the whole set) — the proposed
formula is correct in these cases, in view of the convention that 0! = 1 (see
Section 2.3).
As suggested by the name, we prove this by counting choices. Given a set X of
n elements, in how many ways can we choose a set of k of them? Clearly there are
n possible choices for the 'first' element, (n — 1) choices for the 'second', ... , and
(n — k + 1) choices for the 'fcth'; in total, n(n - 1)... (n - k +1). But we put the terms
'first', 'second', etc., in quotes because a subset has no distinguished first, second, ...
element. In other words, if the same k elements were chosen in a different order, the
same subset would result. So we must divide this number by the number of orders
in which the k elements could have been chosen. Arguing exactly as before, there
are k choices for which one is 'first', (k — 1) for which is 'second', and so on. Division
gives the middle expression in the box. Now the third expression is equal to the
second because n(n — 1)... (n — k + 1) = n\/(n — &)!; the denominator cancels all
the factors from n — k on in the numerator.
Once we have a formula, there are two possible ways to prove assertions or
identities about binomial coefficients. There is a combinatorial proof, arguing from
the definition (we will interpret u j as the number of ways of choosing a team of k
players from a class of n pupils); and there is an algebraic proof, from the formula.
We give a few simple ones.

(3.2.2) Fact.
fn
) =(
) \n — k
FIRST PROOF. Choosing a team of k from a class of n is equivalent to choosing the
n — k people to leave out.
SECOND PROOF. It's obvious from the last formula in the box.

(3.2.3) Fact.

-CO—
FIRST PROOF. We choose a team of k and designate one team member as captain.
There are Cf) possible teams and, for each team, there are k choices for the captain.
24 3. Subsets, partitions, permutations

Alternatively, we could choose the captain first (in n possible ways), and then the
remainder of the team (k — 1 from the remaining n — 1 class members).
Note that this is an application of the 'double counting' principle described in
Section 2.6.
SECOND PROOF. Try it yourself!
You will find that the SECOND PROOFS above probably come more naturally to
you. For this reason, I'll concentrate on the combinatorial style of proof for the next
couple of results. Remember that the algebraic proof is not always appropriate or
even possible — sometimes we won't have a formula for the numbers in question,
or the formula is too complex. (See the discussion of Stirling numbers in Section 5.3
for examples of this.)

(3.2.4) Fact.

PROOF. We have a class of n +1 pupils, one of whom is somehow 'distinguished', and

wish to pick a team of k. We could either include the distinguished pupil (in which
case we must choose the other k — 1 team members from the remaining n pupils),
or leave him out (when we have to choose the whole team from the remaining n).

(3.2.5) Fact.

PROOF. This one is easy — there are 2n subsets altogether (of arbitrary size).

(3.2.6) Fact.
n n\2 (2n
k=0 \ /v / \ n

PROOF. The right-hand side is the number of ways of picking a team of n from a
class of 2n. Now suppose that, of the 2n pupils, n are girls and n are boys. In how
manyy ways can we pick a team of k girls and n — k boys? Obviously this number is
t l i c h is e
(fc) ((n-j
) <l ual t o (fc)2) b y F a c t 3-2-2- T h e r e s u l t n o w follows.
The definition of the binomial coefficient (jj actually makes sense for any non-
negative integers n and k: if k > n, then there are no fc-subsets of an n-set, and
rM = 0. The (first) formula gives the right answer, since if k > n then one of the
factors in the numerator is zero. (This cannot be assumed, since the argument we
gave is only valid if k < n.) However, the second formula makes no sense (unless,
very dubiously, we assume that the factorial of a negative integer is infinite!).
Facts 3.2.2-4 above remain valid with this more general interpretation. (You
should check this.)
Sometimes it is convenient to widen the definition still further. For example, if
k < 0, we should define u ) = 0, in order that Fact 3.2.2 should hold in general. We'll
 3.3. The Binomial Theorem and Pascal's Triangle 25

see in Chapter 4 that it is possible to relax the requirement that n is a non-negative

integer even further. The most general definition, using the formula, works for any
real number n and any integer k: we set

l)...(n-*+l)
k[ itk> U,
if k < 0.
3.3. The Binomial Theorem and Pascal's Triangle
Fact 3.2.5 above can be generalised to the celebrated Binomial Theorem.1 A binomial
is a polynomial with two terms; the Binomial Theorem states that, if a power of a
binomial is expanded, the coefficients in the resulting polynomial are the binomial
coefficients (from which, obviously, they get their name).

(3.3.1) Binomial Theorem

FIRST PROOF. It's clear that (1 + t)n is a polynomial in t of degree n. To find the
coefficient of tk, consider the product
(1 + *)(1 + t)... (1 + t) (n factors).
The expansion is obtained by choosing either 1 or t from each factor in all possible
ways, multiplying the chosen terms, and adding all the results. A term tk is obtained
when t is chosen from k of the factors, and 1 from the other n — k factors. There
are u j ways of choosing these k factors; so the coefficient of tk is m , as claimed.
SECOND PROOF. The theorem can be proved by induction on n. It is trivially true
for n = 0. Assuming the result for n, we have

the coefficient of tk on the right is U ^ j + \Jj (the first term coming from tk~x • t
and the second from tk • 1); and

k-l
by Fact 3.2.4.
1
Proved by Sir Isaac Newton in about 1666.
26 3. Subsets, partitions, permutations

The Binomial Theorem allows the possibility of completely different proofs of

properties of binomial coefficients, some of which are quite difficult to prove in other
ways. Here are a couple of examples. First, a proof of Fact 3.2.3.
Differentiate the Binomial Theorem with respect to t:

The coefficients of tk x
on the left and right of this equation are ^U_j) and k(Jj
respectively.
(3.3.2) Fact. For n > 0, the numbers of subsets of an n-set of even and of odd
cardinality are equal (viz., 2n~1).
PROOF. Put t = — 1 in the Binomial Theorem to obtain
n

hence
In
2
0<k<n
k odd

But the two sides of this equation are just the numbers of subsets of even, resp. odd,
cardinality.
If n is odd, then k is even if and only if n — k is odd; so complementation sets up
a bijection between the subsets of even and odd size, proving the result. However, in
general, a different argument is required. The map X H-> XA{n} (that is, if n £ X,
then remove it; otherwise put n into X) is a bijection on subsets of { 1 , . . . , n} which
changes the cardinality by 1, and hence reverses the parity; so there are equally
many sets of either parity.
The argument can be refined to calculate the number of sets whose size lies in
any particular congruence class. I illustrate by calculating the number of sets of size
divisible by 4. I assume that n is a multiple of 8. (The answer takes different forms
depending on the congruence class of n mod 8.)
(3.3.3) Proposition. If n is a multiple of 8, then the number of sets of size divisible
by 4 is 2 n " 2 + 2<n"2)/2.
For example, if n = 8, the number of such sets is Qj + (4) + (g) = 26 + 23.
PROOF. We let A be the required number, and B the number of sets whose size is
congruent to 2 (mod 4). By Fact 3.3.2, A + B = 2n~1.
Now substitute t = i in the Binomial Theorem. Note that 1 + i = \/2e17r/4, and
so (since n is a multiple of 8), (1 + i) n = 2 n/2 . Thus
2
2"/ - f
k=o
Take the real part of the right-hand side, noting that i* = l,i, — 1, — i according as
k — 0,1,2 or 3 (mod 4). We obtain A — B — 2 n / 2 . From this and the expression for
A + B above, we obtain the value of A (and that of B).
 3.4. Project: Congruences of binomial coefficients 27

REMARK. By taking the imaginary part of the equation, we find the numbers of sets
with size congruent to 1, or to 3, mod 4.
The binomial coefficients are often written out in the form of a triangular array,
known as Pascal's Triangle.2
1
1 1
1 2 1
1 3 1 1
1 4 6 4 1
1 5 10 10 5 1
1 6 15 20 15 6 1
1 7 21 35 35 21 7 1
1 8 28 56 70 56 28 7 1
Thus, u j is the A;th element in the n th row, where both the rows an the elements
in them are numbered starting at zero. Fact 3.2.4 shows that each internal element
of the triangle is the sum of the two elements above it (i.e., above and to the left
and right). Moreover, the borders of the triangle are filled with the number 1 (since
(o) = (n) = -O" With these two rules, it is very easy to continue the triangle as far
as necessary. This suggests that Pascal's Triangle is an efficient tool for calculating
binomial coefficients. (See Exercise 7.)

3.4. Project: Congruences of binomial coefficients

A popular school project is to examine the patterns formed by the entries of Pascal's
Triangle modulo a prime. For example, the first eight rows mod 2 are as follows:
1
1 1
1 0 1
1 1 1 1
1 0 0 0 1
1 1 0 0 1 1
1 0 1 0 1 0 1
1 1 1 1 1 1 1 1
T
If T consists of the first 2 n rows, then the first 2 n + 1 rows look like T ° T .
Thus the pattern has a 'self-similarity' of the kind more usually associated with
fractals than with combinatorics! A similar pattern holds for congruence modulo
other primes, except that the copies of T are multiplied by the entries of the p-rowed
Pascal triangle.

2
Not surprisingly, this object was known long before Pascal. I owe to Robin Wilson the information
that it appears in the works of the Majorcan theologian Ramon Llull (1232-1316). Llull also
gives tables of combinations and mechanical devices for generating them, complete graphs, trees,
etc. However, combinatorics for him was only a tool in his logical system, and logic was firmly
subservient to theology. In his first major work, a commentary on Al-Ghazali, he says, 'We will speak
briefly of Logic, since we should speak of God.'
28 3. Subsets, partitions, permutations

The mathematical formulation is remarkably simple. It was discovered by Lucas

in the nineteenth century.

(3.4.1) Lucas9 Theorem

Let p be prime, and let m = ao + a LP + . . . + akl n = b0 -\-blP +
bkpk> where 0 < a t , b{ <p for i = 0 , . . . , * - 1. Then

NOTE. We assume here the usual conventions for binomial coefficients, in particular, (J) = 0 if a < b.
PROOF. It suffices to show that, if m = cp + a and n = dp + 6, where 0 < a, 6 < p, then

0 - 0 (;)<—>•
For a = ao, 6 = &o, and c = ai + ... + a*;/?* x, d — b\ + ... + bkPk *; and then induction finishes the
job.
This assertion can be proved directly, but there is a short proof using the Binomial Theorem.
The key is the fact that, if p is prime, then

(l + t)p = l+tp (modp).

This is because each binomial coefficient (^), for l < i < p — 1 , i s a multiple of p, so all intermediate
terms in the Binomial Theorem vanish mod p. (For (^) = p\/i\{p — i)!, and p divides the numerator
but not the denominator.) Thus (congruence mod p):

Since 0 < a, b < p, the only way to obtain a term in tn = tdP+b in this expression is to take the
term i = d in the first sum and the term j = b in the second; this gives

(:)•©(;)«-"»•
as required.

3.5. Permutations
There are two ways of regarding a permutation, which I will call 'active' and
'passive'. Let X be a finite set. A permutation of X, in the active sense, is a
one-to-one mapping from X to itself. For the passive sense, we assume that there
is a natural ordering of the elements of X, say {zi,£ 2 ,... ,x n }. (For example, X
 3.5. Permutations 29

might be {1,2,..., n}.) Then the passive representation of the permutation n is the
ordered n-tuple (TT(XI), T T ^ ) , • • • ? ^(^n))-3
In the preceding paragraph, I wrote ir(x) for the result of applying the function
7T to the element x. However, in the algebraic theory of permutations, we often have
to compose permutations, i.e., apply one and then the other. In order that the result
of applying first TTI and then ?r2 can be called 7ri7r2, it is more natural to denote the
image of x under TT as xn. Then

which looks like a kind of associative law!4

As is (I hope) familiar to you, the set of all permutations of { 1 , . . . , n}, equipped
with the operation of composition, is a group. It is known as the symmetric group
of degree n, denoted by Sn (or sometimes Sym(n)). The symmetric groups form one
of the oldest and best-loved families of groups.
From now on, we take X = {1,2,..., n).
A permutation n can be represented in so-called two-line notation as
1 0 n

ITT 2TT . . . TITT

The top row of this symbol can be in any order, as long as XTT is directly under x
for all x. If the top row is in natural order, then the bottom row is the passive form
of the permutation.

(3.5.1) Proposition. Tie number of permutations of an n-set is n\.

PROOF. Take the top row of the two-rowed symbol to be (1 2 ... n). Then there
are n choices for the first element in the bottom row; n — 1 choices for the second
(anything except the first chosen element); and so on.
Note that this formula is correct when n = 0: the only permutation of the empty
set is the 'empty function'.
There is another, shorter, representation of a permutation, the cycle form. A
cycle, or cyclic permutation, is a permutation of a set X which maps

where # i , . . . , z n are all the elements of X in some order. It is represented as

(#i X2 ... xn) (not to be confused with the passive form of a permutation!) The
cycle is not unique: we can start at any point, so (x{ ... xn X\ ... z t _i) represents
the same cycle.
3
In the nineteenth century, it was more usual to refer to a passive permutation as a permutation,
synonymous with 'rearrangement'. An active permutation was called a substitution.
4
We say that permutations act on the right if they compose according to this rule.
30 3. Subsets, partitions, permutations

(3.5.2) Proposition. Any permutation can be written as tie composition of cycles on

pairwise disjoint subsets. The representation is unique, apart from the order of the
factors, and the starting-points of the cycles.

The proof of this theorem is algorithmic. Let TT be a permutation of X.

(3.5.3) Decomposition into disjoint cycles

WHILE tAere is a point of X not yet assigned to a cycle,
• choose any such point x;
• let m be the least positive integer such that X7rm = x;
• construct the cycle (x XTT . . . XTT171"1).
RETURN tAe product of all cycles constructed.

PROOF. In the algorithm, we use the notation 7rm for the composition of m copies of
7T. We first have to show that the construction makes sense, that is, (x XTT . . . XTT171'1)
really is a cycle. This could only fail if the sequence of elements contains a repetition.
But, if XTT1 = XTT\ where 0 < i < j < ra, then (because TT is one-to-one) it holds that
x = XTT^~%\ but this contradicts the choice of m as least integer such that XTT171 = x.
Next, we establish that the cycles use disjoint sets of points. Suppose that XTT1 =
yir\ and suppose that x is chosen before y. If yTTm = y, then #7rz+m~J = yTTm = y,
contradicting the fact that y (when chosen) doesn't already lie in a cycle.
It is clear that any point of X lies in one of the chosen cycles. Finally, the
composition of all these cycles is equal to TT. For, given a point z, there is a unique
y and i such that z — y-K1. Then the cycle containing y agrees with ?r in mapping z
to i/7rt+1, and all the other cycles have no effect on z.
EXAMPLE. The permutation (* \ \ \ \ !j), in cycle notation, is (1 3 4)(2 6)(5). This
is just one of 36 different expressions: there are 3! = 6 ways to order the three cycles,
and 3 - 2 - 1 = 6 choices of starting points.

3.6. Estimates for factorials

Since many kinds of combinatorial objects (for example, binomial coefficients) can
be expressed in terms of factorials, it is often important to know roughly how large
n\ is. In Exercise 3 of the last chapter, upper and lower bounds were found by ad
hoc methods. In this section, a more systematic approach will yield better estimates.
I will prove:

(3.6.1) Theorem.

nlog n — n + 1 < log n\ < nlog n — n + (log(n + 1) + 2 — log 2).

From this, it follows that

log n! = n log n — n + 0(log n).

 3.6. Estimates for factorials 31

This is weaker than an asymptotic estimate for n\ itself: the exponentials of the
upper and lower bounds are e(n/e) n and |(n -f I)e 2 (n/e) n , which differ by a factor
of (n + l)e/2. A more precise estimate (not proved here) is:

(3.6.2) Stirling's Formula

PROOF OF THEOREM. The main tool is shown in the pictures of Fig. 3.1. Since

= logx

V
12 3 n—1 n 12 3 n—1 n n+1
(a) (b)
Fig. 3.1. Sums and integrals

y = logx is an increasing function of x for all positive x (its derivative, 1/x, is

positive), the tops of the rectangles in Fig. 3.1(a) all lie above the curve y = logs,
and those in Fig. 3.1 (b) lie below the curve. In other words,
n
rn rn+\
I log x dx < ^ log z < / log x dx.

The term in the middle is log n!. So

n l o g n - n + 1 <logn! < (n + l)log(n + 1) - (n + 1) -21og2 + 2.
The lower bound is exactly what is needed. For the upper bound, note that

log(n + 1) - log n = / — < -,

Jn X n
so nlog(n + 1) < nlog n + 1. Combining this with the upper bound, we obtain
log n\ < log(n + 1) + n log n - n + 2 log 2 - 2.
If you are interested, you could regard the proof of Stirling's Formula as a
project.5 A lower bound only slightly weaker than Stirling's is given in Exercise 11.
Exercise 12 gives an example of the use of Stirling's Formula to estimate a
binomial coefficient. A weaker result can be obtained much more easily:
5
An accessible proof can be found in Alan Slomson, Introduction to Combinatorics (1991).
32 3. Subsets, partitions, permutations

(3.6.3) Proposition.
22V(2n + 1) < M < 22n.

PROOF. Immediate from the fact that the 2n + 1 binomial coefficients (2nJ, for
i = 0 , . . . , 2n, have sum 22n, and the middle one is the largest.

3.7. Selections
In how many ways can one select k objects from a set of size n?
The answer differs according to the terms of the problem, as we saw in Chapter 2.
Specifically, is the order in which the objects are chosen significant (a permutation)
or not (a combination)'! and is the same object permitted to feature more than once
in the selection, or not? (The term 'permutation' is used in a more general sense
than in the last section: this is what might more accurately be called a 'partial
permutation'.)

(3.7.1) Theorem. Tie number of selections of k objects from a set of n objects is

given by the following table:

Permutations and combinations

Order significant Order not significant

Repetitions k fn + k -1
n k
allowed \
Repetitions , .,x , 7 , 1\ ln
not allowed n(n - 1) • • • (n - k + 1) ^

PROOF. For the column 'order significant', these are straightforward. If repetitions
are allowed, there are n choices for each of the k objects; if repetitions are not
allowed, there are n choices for the first, n — 1 for the second, n — k + 1 for the kth.
For 'order not significant', if repetitions are not allowed, we are counting the
^-subsets of an rc-set, which we already know how to do. The final entry is a bit
harder.

(3.7.2) Lemma. The number of choices of k objects from n with repetitions allowed
and order not significant is equal to the number of ways of choosing n non-negative
integers whose sum is k.

PROOF. Given a choice of k objects from the set cti,..., an, let xt be the number of
times that the object az gets chosen. Then X{ > 0, Y^=i xi ~ k- Conversely, given
( x i , . . . , x n ), form a selection by choosing object at just X{ times.
 3.7. Selections 33

(3.7.3) Lemma. The numoer

number ol
of n-tupi
n-tuples of non-negative integers xi,...,xn with
Xi + ... + xn = k is Un+k-l\ _ (n+k-l\
- i j - I k )-
PROOF. Consider the following correspondence. Put n + k — 1 spaces in a row,
and fill n — 1 of them with markers. Let X\ be the number of spaces before the
first marker; X{ the number of spaces between the (i — l) s t and ith marker, for
2 < i < n — 1; and xn the number of spaces after the n th marker. Then X{ > 0,
Y, Xi = (n + k — 1) — (n — 1) = k. Conversely, given x i , . . . , x n , put markers after x\
spaces, after x2 more spaces, ..., after xn_i more spaces (so that xn spaces remain).
EXAMPLE. Suppose that n = 3, k = 4. The pattern of spaces and markers

• • 13 • M •
corresponds to the values Xi = 2, x2 = 1, x3 = 1. Conversely, the values (xi, x2, x3) =
(0,0,4) correspond to the pattern

Now the number of ways of choosing the positions of the markers is fn*^j =
' ~ 1 ), as claimed.
REMARK. Using the extended definition of binomial coefficients, the number of
selections with repetitions allowed and order not significant can be written

\n—k

A common puzzle is to find as many words as possible which can be formed

from the letters of a given word. Of course, the crucial feature of this problem
is that the words formed should belong to some given human language (i.e., they
should be found in a standard dictionary). There are two possible strategies for this
problem. We could either form all potential words (all permutations of whatever
length), and look each one up in the dictionary; or go through the entire dictionary,
and check whether each word uses a subset of the given letters. In order to decide
which strategy is more efficient, we need to answer a theoretical question (how many
permutations are there?) and some practical ones (how many words are there in the
dictionary, and how fast can we look them up?)
We will solve a special case of the theoretical question. Assume that the n given
letters are all distinct. We will call any ordered selection without repetition from
these letters a word (without judging its legality — note in particular that we include
the 'empty word' with no letters, which doesn't appear in any dictionary6).

(3.7.4) Proposition. The number of ordered selections without repetition from a set
ofn objects is [e • n!J, where e is the base of natural logarithms.

6
If it did, how would you look it up?
34 3. Subsets, partitions, permutations

PROOF. The number f(n) in question is just

jfc=0 **' k=0 k-

From the familiar Taylor series for e37, we see that

k=o k'

So
. ... i i
+

1 1
+
^TT (n + 1)2 + •' •

so / ( n ) = |_e • n\\.
If the allowed letters contain repetitions, the problem is harder. It is possible to
derive a general formula; but it is probably easier to argue ad hoc in a particular
case, as the next example shows.
EXAMPLE. HOW many words can be made from the letters of the word FLEECE?
We count words according to the number of occurrences of the letter E. If there
is at most one E, we can invoke the previous result: there are 24 + 24 + 12 + 4 + 1 = 65
such words (including the empty word). If there are two Es, let us imagine first that
they are distinguishable; then there are 2 + 3 • 6 + 3 • 24 + 120 = 212 possibilities. (For
example, with four letters altogether, we choose two of the remaining three letters
in \Vj = 3 ways, and arrange the resulting four in 4! = 24 ways.) Since the two
Es are in fact indistinguishable, we have to halve this number, giving 106 words.
Finally, with three distinguishable Es, there would be 6 + 3 • 24 + 3 • 120 + 720 = 1158
possibilities, and so there are 1158/6 = 193 words of this form. So the total is
65 + 106 + 193 =• 364 words.

3.8. Equivalence and order

A relation on a set X is normally regarded as a property which may or may not
hold between any two given elements of X. Typical examples are 'equal', 'less than',
'divides', etc. The definition comes as a surprise at first: a relation on X is a subset
of X2 (the set of ordered pairs of elements of X). What is the connection? Of
course, a relation in the familiar sense is completely determined by the set of pairs
which satisfy it; and conversely, given any set of pairs, we could imagine a property
which was true for those pairs and false for all others.
This dual interpretation causes a small problem of notation. In general, if
R C X2 is a relation, we could write x Ry to have the same meaning as (x,y) G R.
This is consistent with the usual notations x = y, x < y, x\y, etc. But we don't
reverse the procedure and write (x,y) G =, (x,y) G <, etc.!
Another Random Scribd Document
with Unrelated Content
Helaman found the Nephite forces, numbering about six thousand
warriors, in a somewhat deplorable condition. The Lamanites, in the
strength of greatly superior numbers, had captured the cities of
Manti, Cumeni, Zeezrom and Antiparah, and held possession of the
country round about. These cities had not been taken without much
bloodshed on both sides. The Nephites especially had lost large
numbers in prisoners, who were generally put to death by their
captors, except the superior officers, who were sent to the land of
Nephi. Antipus, the Nephite commander, was locked up in the city of
Judea, where, dispirited and weakened by excessive toil and
fighting, his troops were making a desperate and painful effort to
fortify the city. The arrival of Helaman and his corps brought hope
and joy again to their hearts, and renewed vigor to their endeavors.
King Ammoron, learning that reinforcements had reached the
defenders of Judea, ordered all active operations to be suspended
for a season. This suspension was most providential for the soldiers
of Antipus, as it gave them time to finish the work of fortifying the
beleaguered city, and also to recruit their health and energies. By
the commencement of the following year the works of defense were
completed, and the Nephites became anxious for the onslaught they
had so greatly dreaded a few months previous. But they were
disappointed. The Lamanites did not feel sufficiently strong to renew
aggressive movements. They contented themselves with occupying
the Nephite cities they had already captured. In the second month
of this year (B. C. 65) a convoy of provisions and two thousand
additional warriors arrived from the land of Zarahemla. The Nephites
in the city of Judea were now ten thousand strong, with abundant
provisions, and they were anxious for a forward movement in order,
if possible, to retake some of their cities in the hands of the enemy.
Antipus and Helaman resolved on a ruse to entice the Lamanites
from behind their fortifications. It was decided that Helaman and his
command should march out of Judea with the apparent intention of
carrying supplies to one of the cities in the hands of the Nephites,
that was built near the seashore. In executing this manœuvre, they
purposely passed at no great distance from the city of Antiparah, in
which was stationed the most numerous of the Lamanite armies, in
the hope that the Lamanites would notice that their numbers were
few, and thus be led to attack them. The stratagem proved
successful. The garrison of Antiparah issued forth in pursuit of
Helaman, who, with all haste, retreated into the wilderness
northward, his intent being to draw his pursuers as far as possible
from Antiparah. When the Lamanites had started in pursuit of
Helaman, Antipus, with a considerable portion of his army, marched
out of the city of Judea and fell in the Lamanites' rear. The retreat
soon became a race. The Lamanites crowded forward with all
possible expedition in the endeavor to reach Helaman before Antipus
caught them. Helaman, on the other hand, used his utmost energy
to keep out of their clutches. Neither of the three bodies turned to
the right or to the left, but kept straight on in the effort to out-march
their foes. Night came and went, and on the morrow the double
pursuit was still kept up. Another night fell, but neither dare turn
from its course.
On the third morning the race for life and victory was again
renewed, but before long the Lamanites, concluding they could not
overtake Helaman, suddenly stopped, and awaited the coming of
Antipus and his weary soldiers, whom they unexpectedly attacked
with great fury, slew Antipus and several of his captains, threw the
Nephite troops into great confusion and forced them to commence a
retreat.
In the meantime, Helaman discovered that he was no longer
pursued, and not knowing the reason, was in doubt what course to
take. He called a hasty council of war, at which it was determined to
return at once, and risk the chances of being caught in a trap by the
crafty Lamanites.
The statement which Helaman makes regarding the conduct of his
young soldiers at this council is very interesting. After he had
explained the situation to them, he inquired, What say ye, my sons,
will ye go against them in battle? Without hesitancy they answered
in the affirmative, saying: Father, behold our God is with us, and he
will not suffer that we shall fall; then let us go forth; we would not
slay our brethren if they would let us alone; therefore let us go lest
they should overpower the army of Antipus. Here Helaman remarks:
Now they never had fought, yet they did not fear death; and they
did think more of the liberty of their fathers than they did upon their
lives; yea, they had been taught by their mothers that if they did not
doubt that God would deliver them. And they rehearsed unto me the
words of their mothers, saying, We do not doubt our mothers knew
it.
Helaman and his sons arrived none too soon on the field of battle.
The soldiers of Antipus were already fleeing before their more
numerous foes, but the valor and impetuosity of the youthful
Ammonites was irresistible. They fell on the Lamanite rear with a
daring and miraculous strength possessed only by men who put their
whole trust in God. Thus attacked in the rear, the Lamanites
immediately halted, changed front, and threw their whole force
against the Ammonites. The surviving officers of Antipus' army,
finding that Helaman had come to their rescue, stopped the retreat,
re-organized their scattered bands, and renewed the attack. The
Lamanites were compelled to succumb; they could not resist the
desperate courage of the Nephites that was driving them in at both
front and rear. Their legions all surrendered, and, by Helaman's
orders, were sent as prisoners of war to Zarahemla.
And what about the young warriors of Ammon? So great was their
faith, so potent its workings, that when, after the battle, Helaman
called the roll of his youthful heroes, not one was missing. The faith
sown by their mothers' words had borne fruit—they were all
preserved. To their undaunted prowess, for they fought as if with the
strength of God, the Nephites unhesitatingly accorded the glory of
the day.
Still the hardly contested war continued. Six thousand men, with
provisions, reached Helaman from Zarahemla and the regions round
about (B. C. 63), besides sixty more young Ammonites who had
grown sufficiently vigorous to assume the hardships of military life.
The city of Cumeni shortly afterwards surrendered through the want
of provisions, their supplies having been continuously cut off by
Helaman's troops. This surrender threw so many prisoners on the
hands of the Nephites that they were unable to guard or feed them.
An officer named Gid, with a sufficient force, was detailed to convey
them to Zarahemla, but on their way, passing near to an invading
body of Lamanites, the prisoners made a desperate attempt to
escape. A few succeeded in getting away, but the greater number
were slain by their guard. Gid and his command returned to
headquarters, as it proved, just in time, for the Lamanites had made
a sudden and unexpected attack at Cumeni, and but for Gid's timely
arrival the Nephite forces would probably have received a severe
defeat. As it was, defeat was turned to victory by their coming.
In this desperate battle every one of the young Ammonites was
wounded, but not one was slain. According to the promise made to
them they were preserved by the marvelous power of God.
 CHAPTER XXXIV.

THE RELIEF OF MANTI—THE OVERTHROW OF THE KING MEN—

PACHUS SLAIN—THE STRUGGLE AT MORONI—TEANCUM SLAYS
AMMORON, BUT AT THE COST OF HIS OWN LIFE—TEANCUM'S
NOBLE CHARACTER.
(ALMA CHAP. 58 TO 62.)
FTER the battle at Cumeni, the Lamanites retreated eastward to
A Manti, which was situated on the upper waters of the Sidon. Nor
was it for several months that this city could be taken, as owing
to internal dissensions at the Nephite capital, and the attempts on
the part of some of the people to overthrow the republic and
establish a monarchy, Pahoran, the chief judge, was unable to
supply the necessary provisions and reinforcements.
In this strait Helaman and his fellow officers called on the Lord in
fervent prayer, which was not unanswered. They received
assurances of deliverance and victory. These blessed assurances
inspired fresh faith and infused renewed courage in the war-weary
hearts of those not given to the love of carnage. Fired with the
determination, by God's grace, to conquer, they entered on a
campaign against the city of Manti, which, by strategy, they captured
before the end of the year (B. C. 63). The moral effect of this victory
was so great that the Lamanites retreated into the wilderness,
evacuating the whole of the territory on the west, but unfortunately
taking with them, as prisoners, many women and children.
For more than a year Moroni could not send the needed help to
Helaman. The rebels in Zarahemla had driven the chief judge out of
the city, and he had taken refuge in Gideon. From there he wrote to
Moroni to come to his assistance, which that officer did at the
earliest possible moment, leaving the armies in the north-east under
the command of Lehi and Teancum. As he advanced he rallied the
people on his line of march to the defense of the liberties of the
republic, and was so successful that, after having joined the chief
judge, Pahoran, he succeeded in overthrowing the "king men,"
killing their leader, Pachus, and completely crushing the rebellion.
This being accomplished, he sent 6,000 men with the necessary
provisions to reinforce Helaman (B. C. 61).
The campaign during this year, along the Atlantic coast, was a
decisive one. At last the soldiers of Ammoron were driven out of
Omner, Morianton, Gid, Lehi, Nephihah, and every other Nephite city
on that sea-board, except the outlying one, called Moroni, where the
whole of the invading host was massed for a final desperate stand,
and around which Moroni, by hurried and lengthened marches, had
concentrated his warriors.
It was the night before an expected decisive battle, and the Nephite
officers and soldiers were too fatigued to either devise stratagems or
execute them. Teancum alone was in a condition of unrest. He
remembered with intense bitterness all the bloodshed, woes,
hardships, famine, etc., that had been brought about in this great
and lasting war between the two races, which he rightly attributed
to the infamous ambition of Amalickiah and Ammoron. He reflected
how he had slain the former, and determined that as he had slain
Amalickiah, so should Ammoron fall. In his anger he stole forth into
the enemy's camp, let himself over the walls of the city, sought out
the king's tent, and when he had found the object of his search, he
cast a javelin at him, which pierced him near the heart. But, unlike
Amalickiah, Ammoron's death was not instantaneous. He had time to
awaken his servant before he passed away. The alarm was given,
the guards started in pursuit; Teancum was overtaken, caught and
slain. On the morrow Moroni attacked the Lamanites, defeated them
with great slaughter, captured the city, and drove them entirely out
of Nephite territory. (B. C. 61).
The writer of the Book of Alma records: When Lehi and Moroni knew
that Teancum was dead, they were exceedingly sorrowful; for
behold, he had been a man who had fought valiantly for his country;
yea, a true friend to liberty, and he had suffered very many
exceeding sore afflictions. But behold, he was dead, and had gone
the way of all the earth.
In that glorious galaxy of patriot-priests, or warrior-prophets, call
them which we may, to whose stern integrity, inspired valor, and
unflinching virtue the Nephite republic, in the days of which we
write, owed so much of its stability and was so greatly indebted for
its perpetuity, Teancum shines among the brightest. View him from
whatever point we please, there is no mistaking the man. His ardent
disposition, his fiery impetuosity, his zealous patriotism, his
undaunted courage, his love of liberty, his entire disinterestedness,
shine forth in every action. Indeed, we might almost call him rash,
so little did he consider his personal safety when he thought the
good of his country required the sacrifice.
In picturing the heroes of those days, Teancum looms up before us
almost as a Hotspur or Murat. In our mind's eye we can see him
charging the solid phalanxes of the Lamanites, rushing at full speed
towards the enemy several lengths ahead of his line of battle; his
commanding presence inspiring confidence, his unwavering voice
ringing out the word of command, his bright armor shining in the
sun, and his hair streaming from beneath his helmet, as, regardless
of all save the liberties of his country, he falls upon the thickest of
the foe, seeking out their chief captains, that by their death an end
may possibly be put to the horrors of war. Thus we find him slaying
with his own hand, at different times, Morianton, Amalickiah and
Ammoron. In fact, it is quite noticeable that in nearly all the great
battles of this age, the Nephites appear to have made it a
conspicuous part of their policy to slay the commander of the
opposing hosts. So fell Amlici, Morianton, Jacob, Coriantumr and
others.
 CHAPTER XXXV.

PEACE ONCE MORE—THE RESULTS OF THE WAR—THE LABORS OF

HELAMAN—SHIBLON RECEIVES THE RECORDS—HAGOTH, THE
SHIP-BUILDER—ANOTHER WAR—MORONIHAH—PAHORAN'S
DEATH—CONTENTION REGARDING THE CHIEF JUDGESHIP—
PAANCHI's REBELLION—THE GADIANTON BANDS—
ASSASSINATION OF PAHORAN II.—ANOTHER LAMANITE
INVASION.
(ALMA CHAP. 62 TO HELAMAN CHAP. 1.)
N THE next year after the capture of the city of Moroni peace was
I established in all the land; not a Lamanite warrior remained on
Nephite soil. Then Pahoran returned to his judgment seat, and
Helaman recommenced his labors in the ministry.
The long-continued and savage war just closed had brought various
evils to the church. In many parts of the land it may be said to have
been disorganized. The occupancy of so many of the Nephite cities
by the unbelieving Lamanites had produced numerous demoralizing
effects. Murders, contentions, dissensions and all manner of iniquity
had become rife, and the hearts of the people had grown hardened.
Yet not altogether so, for there were some who acknowledged the
hand of the Lord in all their afflictions. These humbled themselves in
the depths of humility; and because of the prayers of the righteous
the people were spared.
Such was the state of affairs when Helaman went forth to call the
people to repentance and set the church in order. In this blessed
work he had much success, and with the help of his brethren he
again established the Church of God throughout all the land. These
labors he continued until the time of his death, and his joy therein
was greatly increased by the continued faithfulness of the people.
They, notwithstanding their abundant prosperity, which, as ever,
followed their repentance, remained humble, fervent in prayer and
diligent in well-doing. Such was the happy condition of the people of
Nephi when Helaman died (B. C. 57), he having survived his
illustrious father sixteen years. Shiblon, at the death of his brother,
took possession of the sacred things that had been delivered unto
Helaman by Alma, and held them for four years.
The next year (B. C. 56) the valiant Moroni, one of the greatest and
most virtuous of God's sons, passed away from this state of
mortality to the glories of eternity, at the early age of forty-three
years. Some time before his death he had given the chief command
of the armies of the Nephites to his son, Moronihah, who, from the
history of later years, we judge to have been a worthy son of so
illustrious a sire.
The four years that Shiblon held the plates are principally
noteworthy for the commencement of Nephite emigration to the
northern continent. It was during this period that Hagoth established
his ship-building yards on the Pacific, near the land Bountiful. It is
probable that ships were built by the Nephites before Hagoth's time,
but he being an exceedingly expert mechanic, constructed much
larger ones than had hitherto been built, and thus inaugurated a
new feature in Nephite colonization. [7]
When Shiblon died he committed the records to the care of
Helaman, the son of his brother Helaman. The history of the
Nephites and Lamanites still continued a history of wars. In the
same year that Shiblon died, the Lamanites again raised a numerous
army and went down against their traditional foes. The campaign
was a short one. Moronihah, the son of Moroni, inflicted a signal
blow upon their advancing legions, and drove them back to their
own lands. Their loss in this deservedly ill-fated expedition was
great.
Still this blood-thirsty race never seemed to gain experience by the
things it suffered. This, no doubt, arose to a great extent from the
continued irritation kept up by the wily apostates, who had much
private spleen to gratify in the sufferings of the Nephites, and who
held no particular love or respect for their credulous dupes and cat's-
paws, the Lamanites.
It was in the year B. C. 53 that Helaman took charge of the sacred
plates, etc. In the next year Pahoran, the chief judge, died, which
event gave rise to serious contention amongst the Nephite people.
Three of his sons, named Pahoran, Pacumeni and Paanchi, were
ambitious to fill the exalted position left vacant by their father's
death. Each had his adherents and following, but, according to the
national law, the matter was decided by the voice of the people, and
Pahoran was chosen.
Pacumeni assented to the decision of the citizens, but Paanchi
attempted to raise a rebellion, for which crime he was arrested, tried
by the law, and condemned to death. Still the more wicked part of
the community supported his unlawful claims. These determined to
slay Pahoran, which resolve they carried into effect, and the chief
judge was slain by an assassin named Kishkumen. [8] This foul
murder was committed while the chief magistrate was sitting in the
judgment seat administering the law, but through the connivance of
the murderer's associates in iniquity he escaped.
These lawless men bound themselves together by a secret oath and
covenant, that they would never divulge who was the murderer of
Pahoran, and they swore, by the most horrible oaths, one to
another, to conceal each other's crimes, to aid and sustain each
other in their villainies, and to carry out the designs and directions of
their leaders. Over this band of conspirators, assassins and robbers,
Gadianton stood as the head.
The next year after Pahoran's assassination, the Lamanites invaded
the lands of the Nephites. The Lamanite armies were commanded by
a Nephite dissenter named Coriantumr. He was a descendant of
Zarahemla, therefore, presumedly, of the tribe of Judah. He
determined on new and venturesome tactics, and caused his forces
to make an unexpected dash through the Nephite territory. The
Nephites everywhere gave way before them. They marched through
the center of the country, ravaging its most populous and richest
districts. Before the astonished Nephites could collect their armies
the enemy had assaulted and captured their beautiful and strongly
fortified capital, and for the first time the savage soldiery of Laman
held possession of the towers, temples and palaces of Zarahemla.
On this occasion the chief judge, Pacumeni, was slain. Intoxicated
with his uninterrupted successes, the Lamanite general crowded yet
further north, neglecting to keep up his line of communication in the
rear.
Coriantumr's hope was to obtain possession of the narrow isthmus
which was the key to both continents. In this he failed. The Nephite
commander first checked his progress northward, and then cut off
his retreat. In a fierce battle that followed he was killed, his armies
surrendered, and the remnants hastened ingloriously home,
Moronihah, the Nephite commander, magnanimously permitting
them to return unmolested. (B. C. 51.)

FOOTNOTES:
[7] These ships of Hagoth carried many colonies to the land northward;
as it was their custom to take one load of emigrants and when they had
disembarked, to return for another. Some of these vessels were
eventually lost; that is, the ships and their passengers never reached
their destination. It is supposed by many that a part of them were
carried out to mid-ocean by storms and probably wrecked; and that the
survivors found safety and shelter on some of the islands of the Pacific
Ocean. In this way, it is suggested, the Hawaiian, Samoan and other
islands were first peopled.
[8] Pacumeni was chosen to succeed Pahoran as Chief Judge.
 CHAPTER XXXVI.

PACUMENI SLAIN—HELAMAN CHOSEN CHIEF JUDGE—THE

CONSPIRACY TO SLAY HIM—KISHKUMEN KILLED—THE
PROSPERITY OF THE NEPHITES UNDER HELAMAN.
(HELAMAN CHAP. 1 TO 3.)
S PACUMENI, the chief judge, had been slain at the capture of
A Zarahemla, no sooner was the war over than an election took
place to fill his vacant seat. The choice fell upon Helaman, the
more righteous of the people providentially being still in the majority.
Helaman being a God-fearing, just man, his election was very
distasteful to the Gadianton band and its sympathizers. They
resolved to slay him as they had before slain the younger Pahoran,
and place Gadianton on the judgment seat in his stead. To
accomplish this the same vile instrument was chosen—Kishkumen.
But the protecting hand of the great Jehovah was over and around
about Helaman, and he preserved him from the assassin's knife. A
servant of Helaman, possibly a detective commissioned in such time
of peril to watch the movements of the dangerous classes, by
disguise became acquainted with the doings of the robber band, and
of their intentions toward his master.
As Kishkumen was on his way to fulfil his bloody work, this servant,
whose name is not recorded, met him, and gave him one of their
secret signs. This admitted him into the confidence of the assassin,
who explained his errand, and asked to be conducted privately into
the judgment hall, where Helaman was then sitting in the
performance of his duties. This was agreed upon; the two proceeded
to where the murderer expected to find his victim. The strategy of
the servant disarmed his suspicions, he was off his guard. At the
opportune moment the servant stabbed Kishkumen, and so adroitly
did he perform his work, that the robber fell dead without a groan.
The servant immediately ran to the judgment hall, and informed
Helaman of all that he had heard, seen and done. Without delay,
orders were issued for the arrest of the band, but its members,
finding that Kishkumen did not return, and fearing he had miscarried
in his unholy work, under the guidance of their leader fled
precipitately into the wilderness by a secret way, and, in the depths
of its luxuriant vegetation, hid in a place where they could not be
found. (B. C. 50.)
The succeeding years were of peculiar prosperity, though not of
great righteousness, amongst the Nephite people. They spread out
and colonized in every direction. Many thousands emigrated to the
northern continent, among them great numbers of Ammonites.
Numerous new cities were built, and old ones repaired; ship building
was largely carried on, and the arts and manufactures encouraged.
Temples, tabernacles and sanctuaries were erected in great
numbers; in fact, the people spread out and covered both continents
north and south, east and west. The sacred historian states that he
has not recorded one hundredth part of the doings of the people—
their wickedness and righteousness, their wars and contentions,
their peace and prosperity; but many records were kept, upon which
the history of these things were engraved, and all that is necessary
for the world's good will be brought to light in heaven's own time.
The annals of the remainder of Helaman's rule are very short. In the
years B. C. 45 and 44 there were many contentions in the land, but
in the latter portion of the succeeding year they measurably ceased,
and tens of thousands were baptized unto repentance. So great was
the prosperity of the church at this time that even the priesthood
were surprised thereat, and at the multiplicity of blessings that were
poured out upon the people. This happy state of affairs continued
until the death of Helaman, though somewhat marred by the
increasing pride and vanity that long-continued prosperity had
begotten in the hearts of many of the Christians.
Helaman himself was a righteous man: He did observe to keep the
judgments, and the statutes, and the commandments of God; and
he did do that which was right in the sight of God continually, and he
did walk after ways of his father, insomuch that he did prosper in the
land. So writes the historian of Helaman; what more can be said of
any man?
Helaman had two sons to whom he gave the names of Nephi and
Lehi, to remind them, when they heard their own names called, of
the faith and goodness of their great ancestors, who, by God's
direction, led their fathers to the promised land. When Helaman died
he was succeeded by his son Nephi.
 CHAPTER XXXVII.

THE SONS OF HELAMAN—NEPHI's RIGHTEOUS RULE—THE

LAMANITES AGAIN INVADE ZARAHEMLA—THEY DRIVE THE
NEPHITES INTO THE NORTHERN CONTINENT—THE
MINISTRATIONS OF NEPHI AND LEHI—THE MANIFESTATIONS
OF GOD'S POWER IN THE CITY OF NEPHI—AMINADAB—THE
CONVERSION OF THE LAMANITES—UNIVERSAL PEACE.
(HELAMAN CHAP. 4 AND 5.)
N NEPHI we have one of the greatest prophets that ever trod the
I earth, or to whom the God of our salvation revealed his glorious
will. He lived during the greater portion of the first century before
Christ, and disappeared from the knowledge of mankind but a short
time before the advent of the Messiah as a babe in Bethlehem. He is
first referred to in the Book of Mormon (B. C. 44) as the elder of
Helaman's two sons, Lehi being the younger. These two brothers
appear to have been inseparable during their lives. They are nearly
always mentioned as associated in the great and oft-times perilous
labors of the ministry undertaken for the salvation of either Nephites
or Lamanites. We have no information with regard to the time of
Nephi's birth, but when his father died, in the year B. C. 39, he
succeeded him as chief judge, the duties of which office he filled
with wisdom and justice for about nine years, when owing to the
wickedness of the people, he resigned that office, and Cezoram was
chosen by the people in his stead (B. C. 30).
The years that Nephi judged his people are some of the darkest in
Nephite history. Owing to their great pride and iniquity, the Lord left
them to themselves, and they became weak like unto the Lamanites,
man for man. When war was declared, the latter, being much the
more numerous, carried everything before them. In vain the
Nephites struggled for their homes and their liberties. They were
forced back by the hordes of the Lamanites from city to city, from
land to land. Manti, Gideon, Cumeni, Moroni, and even Zarahemla
fell. Nor did the war end when the blood-thirsty Lamanites held high
carnival in the midst of its towers and palaces. Onward swept the
invading host; backward fled the defenders of the commonwealth,
and backward they continued until every town and city, every tower
and fort, from Melek to Moroni, from Manti to Bountiful, were filled
with the savage, half-disciplined, dark-skinned warriors of Laman.
Not a place could be found in the whole southern continent where
the soldiers of the Nephites successfully held their ground.
Zarahemla, with its hallowed associations, its glorious temples,
where the daily sacrifice was unceasingly offered, its proud palaces,
its luxurious homes, its courts of justice, where the chief judge sat in
the magnificence of almost kingly authority to administer the law—
this their queen city, the seat of their government, the centre of their
civilization, the home of their highest priesthood, was in the hands
of their merciless, vandal-like foes. Nor had the danger stopped;
with hurried hands the Nephites built a line of defense across the
Isthmus of Panama from sea to sea, for the unnumbered hosts of
their conquerors were still pushing forward. This line of fortifications
was effectual; it stopped the roll of the barbaric tide northward, and
the Lamanite commanders rested with the possession of a continent.
In this war the Nephite dissenters took active part against their
white brethren, and to this fact, in part, may be attributed the
sudden success that shone on the Lamanite arms. But little by little
in succeeding years the half repentant Nephites regained their lost
ground, until (B. C. 31) the most northerly half of their possessions
had again fallen into their hands; but because of their only partial
repentance, their leaders had not strength to lead them further, and
Zarahemla still remained in the hands of the warriors of Laman.
When Nephi retired from the judgment seat it was with the intention
of devoting his entire time to the preaching of the gospel. He
associated his brother Lehi with him, and commencing at Bountiful,
he journeyed and preached throughout all the land southward in the
possession of the Nephites. From thence the two brothers passed
onwards to Zarahemla, where they found many Nephite dissenters,
to whom they proclaimed the word of God in great power. Numbers
of these confessed their sins, were baptized unto repentance, and
immediately returned to their brethren to repair, if possible, the
wrongs they had done, and make such restitution as lay in their
power.
Numbers of the Lamanites also received the truth gladly, insomuch
that eight thousand of that race were baptized in Zarahemla and the
regions round about.
From Zarahemla the prophets proceeded to the Lamanite capital in
the land of Nephi, where yet mightier power attended them. The
voice of God from heaven sustained their testimony; angels
ministered to the people who assembled to see them; neither
prisons, nor chains, nor bonds could restrain or hold them, and they
accomplished an ever blessed and marvelous work amongst the
benighted children of Laman (B. C. 30). God's power was manifested
at these times in mercy to the darkened condition of the minds of
the Lamanites, when only extraordinary manifestations of his divine
goodness could reach their hearts. They had no records to which
they could appeal, and all their traditions were opposed to the Holy
Being whose message of eternal joy the Nephite prophets bore.
Thus in their weakness they were strengthened by signs and
wonders which a people better educated in the things of God could
with but ill grace claim.
The story of the ministration of Nephi and Lehi in the land of Nephi
is of the deepest interest. When they reached its chief city they were
thrust into that same prison into which Ammon and his companions
were cast by the guards of king Limhi. Here they were kept with
little or no food for a number of days. At the end of this time the
officers of the Lamanites went to the prison with the intention of
slaying the two brothers. But to their intense surprise the Lamanites
found them encircled about as if by fire. At this strange spectacle
fear fell upon the officers. They dared not touch the two prisoners
lest they should be burned. Yet when they saw that Nephi and Lehi
were not consumed their hearts took courage, though they still
stood as if struck dumb with amazement.
At this point the two brethren stood forward and began to explain
that what was seen was manifested that the spectators might learn
that no one could harm them, and that they were the servants of
the Most High, and his all-powerful arm shielded them. Nor was this
all: a sudden earthquake shook the ground, the prison walls tottered
to their foundations, a pall of thick darkness covered all whom
curiosity or other motives had gathered to the prison. The unburning
flame, the tottering walls, the quivering earth, the impenetrable
cloud of blackness, all conspired to fill the hearts of the Lamanites
with solemn fear and awful dread. They realized the almighty power
of God; they were filled with the sense of their own abject
insignificance. A voice, the voice of One whom they knew not,
sounded in their affrighted ears. Once and again, yea, a third time,
and each time that the voice came it was followed by the trembling
of the earth and the shaking of the prison walls. All nature quivered
at the presence of the Majesty on High, whilst the heavy, palpable,
impenetrable darkness still enshrouded them.
From above the voice descended; it was outside the cloud; its tones
came not to their quaking hearts with the roar of the pealing
thunder; nor was it like the tumultuous flow of angry waters; but a
still voice of perfect mildness, almost a whisper, that pierced to their
inmost souls. That voice was the voice of the mighty God of Jacob,
and he called upon all those who heard him to repent, and to do his
servants no hurt. With the third repetition of this command were
added marvelous words of salvation that cannot be uttered by men.
And because of the darkness that enveloped them, and the fearful
dread that filled their hearts, none dared to move. Fear,
astonishment, apprehension of what was to come, had riveted each
to the spot on which he stood.
Among the crowd was a Nephite dissenter, an apostate from the true
church, named Aminadab. This man, happening to turn his face in
the direction in which the two disciples stood, beheld that their faces
shone with a glorious light, and that they were conversing with some
one who appeared to be above them, for their eyes were turned
heavenward. Aminadab drew the attention of those who surrounded
him to this glorious appearance, and the spell that bound them was
sufficiently removed to enable them to turn towards the prisoners
and to become witnesses of the fact also. What do all these things
mean? they anxiously inquired. They do converse with the angels of
God, answered Aminadab. What shall we do that this cloud of
darkness may be removed? was their next question. You must
repent and cry unto the Voice, even until ye shall have faith in
Christ, he replied. They did cry unto God with all the energy that
their terrifying surroundings inspired, and so continued to supplicate
until the cloud was dispersed. Then, to their great surprise, they
discovered that they also were entombed in a pillar of living fire. Yet
this fire did not hurt them, it did not singe their garments, it did not
consume the prison walls, but their terror was swept away, and they
were filled with a joy that was unspeakable, for the Holy Spirit of
God filled their souls, and they broke forth in marvelous words of
praise and rejoicing. Again a pleasant, searching whisper reached
their gladdened ears. It said unto them, Peace, peace be unto you
because of your faith in my Well-beloved, who was from the
foundation of the world. Now there were about 300 souls who heard
and saw these things, and they cast up their eyes unto heaven,
which was opened to their vision, and holy angels came down and
ministered unto them.
The tidings of this glorious appearing were quickly spread near and
far in the lands where the Lamanites dwelt. So powerful was the
testimony, and so great were the evidences, that the major portion
of the people believed, repented and obeyed the gospel. Then, like
all true saints, they manifested the sincerity of their repentance by
works of restitution; they laid down their weapons of war, they cast
aside their false traditions, their hatred gave place to love, and they
restored to the Nephites Zarahemla and the other lands which they
had taken from them (B. C. 30).
So great was the reformation in their character that the Lamanites
soon exceeded the Nephites in their faith and good works.
Extraordinary as it may appear, instead of Nephite missionaries
visiting the Lamanites, Lamanite missionaries were soon ministering
the precious truths of the gospel among the Nephites. Then a
universal peace, such as had never before been known since the
division of the two races, extended over the whole land. Indeed,
from this time the history of the two nations, to a great extent,
becomes one. Together they worshiped the Lord, together they rose
and sank, together they battled with the assassin hosts of
Gadianton, together they triumphed over those desperadoes, and
together they sought refuge in one vast body when there was no
safety but in massing the people in one land, together the more
unrighteous portions of both races were destroyed at the crucifixion
of the Savior, and together the more righteous ones witnessed his
appearing, listened to his words, received his law, and became
members of his holy church. Henceforth, for generations, they were
no more of Nephi, no more of Laman, no more of Jacob, no more of
Ishmael—all were of Christ.
 CHAPTER XXXVIII.

GROWTH OF EVIL AMONGST THE NEPHITES—THE INCREASE OF

THE GADIANTON ROBBERS—NEPHI'S ANNOUNCEMENT OF THE
MURDER OF THE CHIEF JUDGE—THE DISCOVERY—NEPHI
ARRESTED—HE IS PROVEN INNOCENT—GOD'S COVENANT
WITH HIM—INCREASE OF INIQUITY—A TERRIBLE FAMINE—THE
WELCOME RAIN—THE TREND TO DEATH.
(HELAMAN CHAP. 6 to 11.)
HE GOODLY reign of universal peace, to which we referred in our
T last chapter, brought stability, stability developed wealth, wealth
engendered pride, pride gave birth to numerous sins, to be
followed by contentions, dissensions, and then wars. These evils
begat sorrow, sorrow softened the hearts of the people to
repentance, repentance was followed by the blessing of God, which
again brought peace, prosperity and, by-and-by, riches. At this era of
Nephite national life, this is the one eternal round which their
inspired historians are compelled to chronicle. Within four short
years of the happy time of universal peace we have just referred to,
the riches of the world had induced stubbornness and rebellion
towards God, combined with the insane desire to rob, plunder and
murder their fellow-men. If there ever were a people swift to do evil,
it was the Nephites of this generation. In the year B. C. 26,
Cezoram, the chief judge, was murdered by an unknown hand, as he
sat on the judgment seat, and his son, who succeeded him, suffered
in like manner within the year. The Gadianton robbers grew in
strength, both in numbers and influence, and were actually fostered
amongst the Nephites, while the more righteous Lamanites utterly
destroyed all they found within their borders. The one people
dwindled in unbelief, the other grew in grace and in the power of
God's divine Spirit.
Nephi, who had gone to the northern continent, tarried there until
the year B. C. 23, when, his teachings and prophecies having been
rejected by its inhabitants, he returned in sorrow to Zarahemla; but
he found no comfort there. The Gadianton robbers filled the
judgment seats, and perverted the law to their own avarice and lust.
The life, the property, the liberty, the virtue of righteous men and
women were counted but things of naught, their playthings or their
spoil.
Nephi's house in Zarahemla was situated on one of the principal
thoroughfares. It led to the chief market-place. In his garden, near
the highway, he built a tower whither it was his wont to repair for
prayer. On one occasion, shortly after his return from the north, he
became so deeply concerned because of the iniquities of the people,
that in earnest supplication to the Lord he raised his voice so high
that he was heard by the passers by in the street below. A listening
crowd soon gathered, and when the prophet had ended his
devotions and became aware of their presence, he commenced to
teach them. His words were not sugar-coated, to adapt them to the
tastes of his congregation. To the contrary, he boldly rebuked their
sins, their murders, and their secret wickedness; at the same time,
in the love of the gospel, he entreated and plead with them to
amend their lives and do better. He also warned them of the terrible
judgments that would fall upon them if they did not turn from their
sinful ways.
Towards the conclusion of his address, Nephi surprised his hearers
by stating that the chief judge, Seezoram, had been murdered by his
brother, Seantum, who was anxious to obtain the chief judgeship
himself. Both these men were members of the vile band who owned
Gadianton as their chief.
The people did not believe Nephi's statement, so five incredulous
men ran to the judgment hall to find out the truth of the matter.
When they reached there they discovered Seezoram lying dead in a
pool of blood near the judgment seat. The five messengers were so
overcome with fear at this awful sight that they fell to the earth.
Soon after, other citizens who had not heard Nephi came in. Finding
the dead judge and the five men all there they concluded that the
latter must be the murderers, who, by some manifestation of the
power of heaven, had been prevented from leaving the scene of
their shameful deed. The officers therefore took the five and cast
them into prison.
When the wicked learned that Nephi's words had proven true, they
charged him with being an accomplice. They did not believe in
revelation from God, so argued that Nephi must have had a hand in
the murder or he could not have known anything about it. He was
therefore taken and bound and brought before the multitude. Then
they cross-examined him, abused him, and finally offered him
money to confess that he had employed some one to commit the
dreadful crime. They were anxious to bring reproach and trouble
upon him that they might have an excuse for not believing his words
nor heeding his teachings.
Nephi to establish his innocence sent his accusers to the house of
Seantum, and instructed them what to say. He further told them
how the fratricide would act; how he would acquit him (Nephi) of all
complicity in the murder, assert his own innocence, until shown
stains of blood on his cloak, and then, overwhelmed with terror, he
would confess.
The people went and followed Nephi's instructions, and all that he
had told them came to pass; for according to his words Seantum did
at first deny, and according to the words he did afterwards confess;
acknowledging also that Nephi knew nothing of the matter, without
it had been revealed to him of God.
Some of the citizens now acknowledged that Nephi was a prophet,
others declared that he was a god, whilst many remained hardened
in their sins. So violent became the contention that the people
gathered in excited crowds upon the streets, wrangling and
disputing about the events of the past two days, and in their
excitement they entirely forgot Nephi, and left him standing alone in
the street.
With a sorrowful heart he wended his way homeward; but before he
reached there, the voice of the Lord came to him with many words
of comfort and commendation. As with others of his servants, the
Lord made a covenant with him, that he would bless him forever;
that whatsoever he bound on earth should be bound in heaven, and
whatsoever he loosed on earth should be loosed in heaven; that he
should have power over the elements to bless and to curse; to smite
the earth with famine, and pestilence, and destruction.
Notwithstanding the many proofs the ungodly Nephites had that
Nephi was a true prophet, they continued to reject his teachings.
They persecuted him, and even went so far as to seek his life. But
he was conveyed out of their midst by the power of God, and
ministered among other peoples.
The general character of the Nephites continued to grow worse and
worse. The Gadianton robbers grew stronger and stronger. For a few
years there was increasing commotion, disunion and bloodshed. At
last, wearied at beholding so much misery and contention, Nephi
prayed that the Lord would not suffer the people to be destroyed by
the sword, but rather let a famine desolate the land, and,
peradventure, bring the people to an understanding of their awful
condition, and cause them to humble themselves and repent. The
Holy One heard and answered his petition, the heavens became as
brass over the land, the rains ceased, the earth dried up, the crops
failed, the people perished for want of food.
Two years passed (B. C. 19 and 18) and the third came, and still the
refreshing rain was withheld (B. C. 17). During this year the people,
humbled by their sufferings, turned towards the Lord. They
endeavored to root out iniquity from their midst. They destroyed the
Gadianton robber bands, and established the government on a more
righteous foundation. Nephi, observing the change in their conduct
and feelings, interceded with the Lord in their behalf. His prayers
were answered, the welcome rain descended on the parched-up soil,
and a bounteous harvest once more crowned the labors of the
husbandman (B. C. 16).
The repentant people now regarded Nephi in his true light; they
revered him as a great prophet, and for a few short years they
listened to his teachings. While they did so they prospered. But the
leaven of unrighteousness had too thoroughly permeated the
national life for their faithfulness to God to be of long duration. Two,
three, or perhaps half a dozen years they would maintain their
integrity, and then corruption would seethe, the vile would snatch
the reins of government, the good would be oppressed, and
contention and war, with all their horrors, would again reign
supreme. Thus it was after the three years of famine. For two years
there was peace, in the third there began to be much strife (B. C.
13), in the next, the Gadianton bands reappeared, and carried havoc
amongst their more peaceable fellow-countrymen. Going on, year by
year they grew in iniquity and ripened for destruction. For many
years Nephi strove to stem the tide of vice. At times partial success
rewarded his unceasing efforts, and he had joy in the baptism of
some honest souls. But the great bulk of the people had rejected the
gospel, they had no love for its holy principles, and were unfit for its
blessings.
 CHAPTER XXXIX.

SAMUEL THE LAMANITE—HIS MISSION AND PROPHECIES—THE

VAIN ATTEMPT TO DESTROY HIM—HE RETURNS TO HIS OWN
COUNTRY.
(HELAMAN CHAP. 13 TO 16.)
E COME now to the days of Samuel the Lamanite (B. C. 6).
W Without any previous reference to him, he appears suddenly in
the foreground of ancient American history, bearing a weighty
and solemn message; a messenger of God's displeasure, he stands a
Jonah to the Nephites. That message is faithfully delivered; then he
disappears forever from our sight.
The condition of society in the days of Samuel was somewhat
peculiar. The Nephites and Lamanites had, so far as righteousness is
concerned, to a great extent changed places. The former were
puffed up with worldly pride, were full of vain boastings, envyings,
strifes, malice, persecutions, murders and all manner of iniquities.
They cast out, they stoned, they slew the servants of God, while
they encouraged, exalted and rewarded the false teachers who
flattered them in their vileness and sung in their ears the siren's
song of "all is well." They reveled in all the luxury that the fatness of
the land brought forth; they were ostentatious in the use of gold and
silver and precious things; but their hearts never turned in
thankfulness to the great Giver of all these bounties. The majority of
the Lamanites, on the contrary, walked circumspectly before God;
they were full of faith and integrity, were zealous in the work of
converting their fellows, and kept the commandments, statutes and
judgments of the Lord according to the law of Moses.
Such was the condition of affairs when the Lamanite prophet Samuel
appeared among the sin-stained citizens of Zarahemla, and for many
days preached repentance in their midst. Their eyes were blind and
their ears were deaf, sin filled their souls, and in their anger they
cast him out. But the work of his mission was not yet accomplished.
As he was preparing to return to his own country, a holy angel
visited him and proclaimed the voice of the Lord. That voice
commanded that he should turn back and prophesy to the people of
Zarahemla the things that should come into his heart.
He returned to the city, but was refused admission at its gates. The
iniquitous dwellers therein had no desire to have their peace
disturbed by the voice of divine threatenings. But the prophet had
the word of the Lord burning within him, and could not be
restrained. He mounted the walls of the city, and from this
conspicuous vantage ground, with out-stretched hands and loud
voice, he proclaimed to the wicked the unwelcome tidings of their
coming destruction. Many listened to his proclamation, some few
were pricked in their hearts, repented of their evil deeds, and sought
the prophet Nephi, that they might be baptized. Others were angry,
they gathered up the stones in the roadway and hurled them at
Samuel; they drew forth their bows and shot arrows at him. But to
no effect; the protecting power of the Holy Spirit was around him,
and he could not be harmed.
When some beheld how wonderfully the prophet was preserved, it
was a testimony to them that God was with him, and they also
sought Nephi, confessing their sins. But the great body of the
populace grew more enraged at the want of success that attended
their murderous efforts. They called upon their captains to seize and
bind him. They cried out, He hath a devil, and it is by this power he
is preserved; take the fellow, bind him, and away with him!
Following the wild satanic cry of the multitude, the officers of the
law endeavored to arrest Samuel. But he cast himself down from the
wall of the city and fled out of the lands of the Nephites into his own
country. There he preached and prophesied among his own people;
but among the people of Nephi he was never heard of more.
The prophecies of Samuel are among the most wonderful recorded
in holy writ. He especially foretold many things regarding the life and
death of our Savior, and concerning the future destiny of his people,
and of the Nephites.
With regard to the birth of the Redeemer he said:
Behold, I give unto you a sign; for five years more cometh, and
behold, then cometh the Son of God, to redeem all those who shall
believe on his name.
And behold, this will I give unto you for a sign at the time of his
coming; for behold, there shall be great lights in heaven, insomuch
that in the night before he cometh there shall be no darkness,
insomuch that it shall appear unto man as if it was day.
Therefore there shall be one day and a night, and a day, as if it were
one day, and there were no night; and this shall be unto you for a
sign; for ye shall know of the rising of the sun, and also of its
setting; therefore they shall know of a surety that there shall be two
days and a night; nevertheless the night shall not be darkened; and
it shall be the night before he is born.
And behold there shall a new star arise, such an one as ye never
have beheld; and this also shall be a sign unto you.
And behold this is not all, there shall be many signs and wonders in
heaven.
And it shall come to pass that ye shall all be amazed and wonder,
insomuch that ye shall fall to the earth.
Regarding the death of the Lord Jesus he declared:
But behold, as I said unto you concerning another sign, a sign of his
death, behold, in that day that he shall suffer death, the sun shall be
darkened and refuse to give his light unto you; and also the moon,
and the stars; and there shall be no light upon the face of this land,
even from the time that he shall suffer death, for the space of three
days, to the time that he shall rise again from the dead.
Yea, at the time that he shall yield up the ghost, there shall be
thunderings and lightnings for the space of many hours, and the
earth shall shake and tremble, and the rocks which are upon the
face of this earth; which are both above the earth and beneath,
which ye know at this time are solid, or the more part of it is one
solid mass, shall be broken up;
Yea, they shall be rent in twain, and shall ever after be found in
seams and in cracks, and in broken fragments upon the face of the
whole earth; yea, both above the earth and beneath.
And behold there shall be great tempests, and there shall be many
mountains laid low, like unto a valley, and there shall be many
places, which are now called valleys, which shall become mountains,
whose height thereof is great.
And many highways shall be broken up, and many cities shall
become desolate.
And many graves shall be opened, and shall yield up many of their
dead; and many saints shall appear unto many.
And behold thus hath the angel spoken unto me for he said unto
me, that there should be thunderings and lightnings for the space of
many hours:
And he said unto me that while the thunder and the lightning lasted,
and the tempest, that these things should be, and that darkness
should cover the face of the whole earth for the space of three days.
And the angel said unto me, that many shall see greater things than
these, to the intent that they might believe that these signs and
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com