Textbook Auxiliary Polynomials in Number Theory 1St Edition Masser Ebook All Chapter PDF

Auxiliary polynomials in number theory
1st Edition Masser

Visit to download the full and correct content document:
https://textbookfull.com/product/auxiliary-polynomials-in-number-theory-1st-edition-m
asser/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Number Theory 1st Edition Róbert Freud
https://textbookfull.com/product/number-theory-1st-edition-
robert-freud/
Invitation to Number Theory 1st Edition Ore
https://textbookfull.com/product/invitation-to-number-theory-1st-
edition-ore/
Statistical independence in probability analysis number

theory Kac
https://textbookfull.com/product/statistical-independence-in-
probability-analysis-number-theory-kac/
Quaternion Algebras in Number Theory Spring 2016

Kimball Martin
https://textbookfull.com/product/quaternion-algebras-in-number-
theory-spring-2016-kimball-martin/
Algebraic Number Theory - A Brief Introduction 1st
Edition J.S. Chahal
https://textbookfull.com/product/algebraic-number-theory-a-brief-
introduction-1st-edition-j-s-chahal/
Invitation to Number Theory 2nd Edition Oystein Ore
https://textbookfull.com/product/invitation-to-number-theory-2nd-
edition-oystein-ore/
Number Theory Revealed An Introduction Andrew Granville
https://textbookfull.com/product/number-theory-revealed-an-
introduction-andrew-granville/
Research Directions in Number Theory Women in Numbers

IV Jennifer S. Balakrishnan
https://textbookfull.com/product/research-directions-in-number-
theory-women-in-numbers-iv-jennifer-s-balakrishnan/
Anomalies in Net Present Value, Returns and

Polynomials, and Regret Theory in Decision-Making 1st
Edition Michael C. I. Nwogugu (Auth.)
https://textbookfull.com/product/anomalies-in-net-present-value-
returns-and-polynomials-and-regret-theory-in-decision-making-1st-
edition-michael-c-i-nwogugu-auth/
C A M B R I D G E T R AC T S I N M AT H E M AT I C S
General Editors
B. BOLLOB ÁS, W. FULTON, A. KATOK, F. KIRWAN,

P. SARNAK, B. SIMON, B. TOTARO
207 Auxiliary Polynomials in Number Theory

C A M B R I D G E T R AC T S I N M AT H E M AT I C S
GENERAL EDITORS
B. BOLLOBÁS, W. FULTON, A. KATOK, F. KIRWAN,
P. SARNAK, B. SIMON, B. TOTARO
A complete list of books in the series can be found at www.cambridge.org/mathematics.

Recent titles include the following:
171. Orbifolds and Stringy Topology. By A. Adem, J. Leida, and Y. Ruan
172. Rigid Cohomology. By B. LE Stum
173. Enumeration of Finite Groups. By S. R. Blackburn, P. M. Neumann, and
G. Venkataraman
174. Forcing Idealized. By J. Zapletal
175. The Large Sieve and its Applications. By E. Kowalski
176. The Monster Group and Majorana Involutions. By A. A. Ivanov
177. A Higher-Dimensional Sieve Method. By H. G. Diamond, H. Halberstam, and W. F.
Galway
178. Analysis in Positive Characteristic. By A. N. Kochubei
179. Dynamics of Linear Operators. By F. BAYART and É. Matheron
180. Synthetic Geometry of Manifolds. By A. Kock
181. Totally Positive Matrices. By A. Pinkus
182. Nonlinear Markov Processes and Kinetic Equations. By V. N. Kolokoltsov
183. Period Domains over Finite and p-adic Fields. By J.-F. Dat, S. Orlik, and M. Rapoport
184. Algebraic Theories. By J. Adámek, J. Rosický, and E. M. Vitale
185. Rigidity in Higher Rank Abelian Group Actions I: Introduction and Cocycle Problem. By
A. Katok and V. NIŢIC̆Ă
186. Dimensions, Embeddings, and Attractors. By J. C. Robinson
187. Convexity: An Analytic Viewpoint. By B. Simon
188. Modern Approaches to the Invariant Subspace Problem. By I. CHALENDAR and
J. R. Partington
189. Nonlinear Perron-Frobenius Theory. By B. LEMMENS and R. Nussbaum
190. Jordan Structures in Geometry and Analysis. By C.-H. Chu
191. Malliavin Calculus for Lévy Processes and Infinite-Dimensional Brownian Motion.
By H. Osswald
192. Normal Approximations with Malliavin Calculus. By I. NOURDIN and G. Peccati
193. Distribution Modulo One and Diophantine Approximation. By Y. Bugeaud
194. Mathematics of Two-Dimensional Turbulence. By S. KUKSIN and A. Shirikyan
195. A Universal Construction for Groups Acting Freely on Real Trees. By I. CHISWELL and
T. MÜLLER
196. The Theory of Hardy’s Z-Function. By A. Ivić
197. Induced Representations of Locally Compact Groups. By E. KANIUTH and K. F. Taylor
198. Topics in Critical Point Theory. By K. PERERA and M. Schechter
199. Combinatorics of Minuscule Representations. By R. M. Green
200. Singularities of the Minimal Model Program. By J. Kollár
201. Coherence in Three-Dimensional Category Theory. By N. Gurski
202. Canonical Ramsey Theory on Polish Spaces. By V. Kanovei, M. Sabok, and J. Zapletal
203. A Primer on the Dirichlet Space. By O. EL-Fallah, K. Kellay, J. Mashreghi, and
T. Ransford
204. Group Cohomology and Algebraic Cycles. By B. Totaro
205. Ridge Functions. By A. Pinkus
206. Probability on Real Lie Algebras. By U. FRANZ and N. Privault
207. Auxiliary Polynomials in Number Theory. By D. Masser
Auxiliary Polynomials in Number Theory
DAV I D M A S S E R
University of Basle, Switzerland
University Printing House, Cambridge CB2 8BS, United Kingdom
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781107061576
© David Masser 2016
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 published 2016
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloguing in Publication Data
Names: Masser, David William, 1948–
Title: Auxiliary polynomials in number theory / David Masser, Universitat Basel, Switzerland.
Description: Cambridge : Cambridge University Press, 2016. | Series: Cambridge tracts
in mathematics ; 207 | Includes bibliographical references and index.
Identifiers: LCCN 2015050947 | ISBN 9781107061576 (Hardback : alk. paper)
Subjects: LCSH: Number theory. | Polynomials.
Classification: LCC QA241 .M395 2016 | DDC 512.7/4–dc23 LC record
available at http://lccn.loc.gov/2015050947
ISBN 978-1-107-06157-6 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy
of URLs for external or third-party Internet Web sites referred to in this publication
and does not guarantee that any content on such Web sites is, or will remain,
accurate or appropriate.
Contents
Introduction page vii
1 Prologue 1
2 Irrationality I 7
3 Irrationality II – Mahler’s Method 20
4 Diophantine equations – Runge’s Method 30
5 Irreducibility 50
6 Elliptic curves – Stepanov’s Method 64
7 Exponential sums 76
8 Irrationality measures I – Mahler 88
9 Integer-valued entire functions I – Pólya 101
10 Integer-valued entire functions II – Gramain 111
11 Transcendence I – Mahler 123
12 Irrationality measures II – Thue 133
13 Transcendence II – Hermite–Lindemann 158
14 Heights 166
15 Equidistribution – Bilu 193
16 Height lower bounds – Dobrowolski 200
17 Height upper bounds 212
18 Counting – Bombieri–Pila 218
19 Transcendence III – Gelfond–Schneider–Lang 228
v
vi Contents
20 Elliptic functions 243

21 Modular functions 279
22 Algebraic independence 292
Appendix: Néron’s square root 312
References 334
Index 342
Introduction
Ever since it was invented, arguably by Runge, the method of auxiliary poly-
nomials has been vital to (and of unreasonable effectiveness in) the modern
development of key aspects of number theory. The aim of this book is to give
an account of the method in many of its forms, focusing almost exclusively on
those polynomials which cannot be written down explicitly.
I well remember (standing in Heffers bookshop Cambridge around 1970)
reading about this method in the foreword to Lang’s book on transcendental
numbers, and experiencing disbelief that anything so far-fetched could work at
all. So I will not attempt any explanation at this point.
Instead, I (or from now on, the authorial we) treat the method as the union
of its examples, and there is no shortage of these.
Here is the plan of this book (Mike Tyson said that everyone has a plan until
you punch them in the face – then they don’t have a plan). The general strategy
is to present in each chapter an application of the method to a different sort of
problem, often the simplest in its area. Then at the end of each chapter we give
a brief account of subsequent developments in the area.
We start with a short Prologue (Chapter 1) where we show that the basic idea
can be used in rather simple situations which have nothing to do with number
theory.
Then in Chapter 2 we commence our diophantine considerations with a
discussion of irrationality. We quickly dispose of the number e by the standard
truncation argument and we show also that e is not a quadratic irrational.
Here we meet a small problem, which can be rather quickly solved; how-
ever, it is typical of the problems that arise in later applications and in some
examples its solution can be distinctly non-trivial. Thus Roth, in showing that
irrational algebraic numbers cannot be approximated to within an order of
q−2− by rationals p/q, had to solve such a problem. The solution is called
vii
viii Introduction
Roth’s Lemma, and it was certainly one of the achievements that gained him
a Fields Medal. We do not prove Roth’s Theorem here but we do treat Thue’s
Method in Chapter 12.
We postpone to Chapter 13 a proof that
∞
αk 1 1 1
eα = = 1 + α + α2 + α3 + α 4 + · · ·
k! 2 6 24
k=0
is irrational and even transcendental for all rational and even algebraic α = 0,
as this requires elements of algebraic number theory. Truncation gives only
irrationality
√ and that essentially only for α = 1, 2, 4 (and slightly surprisingly
α = 2).
In the same Chapter 2 we construct our first auxiliary polynomial with a
diophantine purpose: namely, to show that for any rational α = 0 the classical
series
∞
αk 1 1 1
= 1 + α + α2 + α3 + α4 + · · ·
2k(k−1)/2 2 8 64
k=0
is irrational. This is somewhat related to theta functions. Although it converges

quite rapidly, the speed is also insufficient for mere truncation. The result
itself is not so fundamental, but it provides a good introduction to the use
of auxiliary polynomials; that used here is probably the simplest of its kind,
and we calculate a few examples. One needs also some elementary complex
analysis, which will be much developed later on.
In Chapter 3 we then progress to the similar but more elaborate Mahler’s
Method, still sticking just to irrationality; the results here are historically
important and they led to the solution of the Mahler–Manin Conjecture and
then to Nesterenko’s Theorem on the algebraic independence of π and eπ . The
irrationality here will be generalized to full-blooded transcendence in a later
chapter. Here we treat just
∞
k
α 2 = α + α 2 + α 4 + α 8 + α 16 + · · ·
k=0
for every rational α with 0 < |α| < 1. The proof is quite similar to that
in Chapter 2 but a little more analysis is needed, and further the auxiliary
polynomial is more complicated, in fact already of a fairly typical sort; still we
calculate some more examples. Mahler’s Method has been greatly developed
and some recent applications refer to the famous Mandelbrot set. It also
√ played
a transient role in proving that the decimal digits of numbers like 2 cannot
be generated by a finite automaton.
Introduction ix
In Chapter 4 we prove that certain diophantine equations in two variables

have at most finitely many solutions, using the auxiliary polynomial pioneered
by Runge. The method enables all solutions to be found in principle. A typical
example is that there are at most finitely many integers x, y with
x(x3 − 2y3 ) = y.
Or, coming from Cassels’s well-known result on the Catalan Equation recently
solved completely by Mihăilescu, there are at most finitely many integers x, y
with x5 − y7 = 1 provided y is not divisible by 5 (we do not prove Mihăilescu’s
Theorem here). Of course equations like
x3 − 2y3 = m
for fixed m are more natural, and these will be considered in Chapter 12. For
the proofs here, we need to know that the large complex solutions are given
by Puiseux (or better Laurent) series. It seems that this is not so easy to find
in the literature, especially regarding the crucial convergence properties, so we
provide quite a few details.
Then in Chapter 5 we prove some results similar to the classical Hilbert
Irreducibility Theorem, usually abbreviated to HIT, by using the machinery of
the preceding chapter. They are not so general as HIT, but when they do work,
they deliver more information. The results were first found by Sprindzhuk also
using auxiliary polynomials, but in a more elaborate way. Nowadays this sort
of thing can be done with heights machinery, but that is not so elementary. A
typical example, related to that of the previous chapter, is that there are at most
finitely many integers y such that the polynomial
X(X 3 − 2y3 ) − y
in Q[X] is reducible over the rationals, and in principle these can all be found.
A literal application of HIT would show only that there are infinitely many
rational y such that the polynomial is not reducible. So sometimes we get a
Strong Hilbert Irreducibility Theorem; but we refrain from abbreviating this.
Here we need resultants; these can be found almost anywhere, but because we
use them frequently in this book we provide a self-contained account.
In Chapter 6 we jump to a different topic. We prove that the number N of
points modulo a prime p on an affine elliptic curve satisfies
√
|N − p| ≤ 8 p;
√
this is slightly weaker than the classical result of Hasse involving 2 p. The
proof uses the simplest non-trivial example of the auxiliary polynomial intro-
duced in a surprising way by Stepanov in 1969; here we attempt to motivate the
x Introduction
proof with the help of some easier intermediate results. Not even the definition
of elliptic curve is needed here, let alone any properties. Thus all we do is count
the integer solutions (x, y) modulo p of an equation y2 = x3 + ax2 + bx + c
modulo p. There are many generalizations (and Schmidt wrote an entire book
about them) but none has quite the same appeal. With rather little extra effort
one can treat y2 = x5 + · · · and worse; in the geometric context this is far from
simple because it would involve curves of genus 2 and worse.
In Chapter 7 we make another jump which seems even bigger, to the topic
of exponential sums. The best known is Gauss’s
p
2π ik2
exp ,
p
k=1
also for prime p, whose absolute value p1/2 is much smaller than the number of
its terms. One of these sums, due to Heilbronn, resisted for some time all efforts
to prove its smallness until Heath-Brown in 1996 achieved this. His beautiful
proof imitated Stepanov’s auxiliary polynomial in a kind of analytic context
involving a logarithm-like function. Some arguments had been anticipated by
Mitkin in 1992. We give the details. Specifically
p

2π ikp
exp ≤ 4p11/12 .
p2
k=1
As far as I know, these arguments have not been developed very far since then,
despite some interesting features involving differential equations.
In Chapter 3 we proved the irrationality of the values μ of Mahler’s series
at non-zero rationals. Thus the quantity |μ − p/q| is positive for all integers p
and q ≥ 1. A natural question is: “How small can this quantity get?” Indeed
with an algebraic irrational in place of μ this question is fundamental in the
theory of diophantine equations, as we will see in Chapter 12. Our answer
in Chapter 8 requires refining the arguments of Chapter 3. There are two key
steps. One is a “zero estimate” asserting that not too many things can vanish;
such estimates play a major role in more recent developments. The other, more
classical, is an estimate for the coefficients of the auxiliary polynomial; this
involves the famous Siegel Lemma, which will be used over and over again
in the sequel. We also make a simple application of the maximum modulus
principle for analytic functions. This too will be used frequently later, under
the popular name of the Schwarz Lemma. In this way we will prove that there
exist c = c(μ) > 0 and κ = κ(μ) such that

μ − p ≥ c .
q qκ
Introduction xi

For example with μ = ∞ 2k
k=0 (2/3) we can take κ = 77.
There is a famous result of Pólya on entire functions mapping the natural
numbers to the rational integers; this may have influenced Gelfond in his pio-
neering work on the transcendence of αβ (see Chapter 19). Pólya’s original
proof used interpolation formulae and gave the best possible constant. Much
later Waldschmidt gave a version by auxiliary polynomials, which sadly gives
a worse constant. The proof is nevertheless illuminating; it needs binomial
coefficients to avoid factorials, one of the key ideas in Thue’s famous proof
(see Chapter 12). More precisely, we show in Chapter 9 that an entire function
f with
f (0), f (1), f (2), . . . , f (n), . . .
all in Z must be a polynomial if |f (z)| grows of order at most C |z| for a certain
C > 1. Pólya could take any C < 2; and the standard example 2z shows that
nothing better is possible. Or reformulated: if a non-polynomial entire function
f has this growth, then at least one of f (0), f (1), f (2), . . . must be non-integral.
Gelfond’s step from non-integrality to transcendence needed many more ideas,
all of which will be developed in this book.
The rather natural generalization to the Gaussian integers G = Z + Zi with
f mapping G into itself also played a similar historical role; for example it
probably directly inspired Gelfond’s proof of the transcendence of eπ . But the
best possible constant did not appear until a relatively recent paper of Gramain;
paradoxically enough, his proof involves an auxiliary polynomial (or better an
auxiliary function). More precisely, f itself must be a polynomial if |f (z)| now
grows of order at most C|z| for a certain C > 1. Gelfond considered this
2
problem too, and obtained the notorious value

π
C = exp < 1 + 10−45
2(1 + exp(164/π ))2
(modestly not mentioned in his book). In the late 1970s, I obtained a constant,
extremely difficult to compute, which later turned out to be about 1.181; and
π
I conjectured that the best possible constant was exp( 2e ) about 1.782. This
Gramain proved, and so do we in Chapter 10.
In Chapter 11 we present our first transcendence result. We extend Mahler’s

Method in Chapter 3 to prove the transcendence of his ∞ 2k
k=0 α for all alge-
braic α with 0 < |α| < 1. That is apparently how he tested his recovery while
convalescing at home from an illness. No more ideas are needed, but to go
beyond irrationality requires some rudimentary notion of “size” of an algebraic
number, with some sort of “Liouville estimate”. This sort of technicality is
fundamental to all transcendence proofs. The concept will be developed later
xii Introduction
into the more sophisticated “height”, which will then be studied for its own
sake, for example with reference to Lehmer’s Question of 1933 in connexion
with factorization problems.
At last in Chapter 12 we prove the famous Thue improvement of Liouville’s
classical result. The proofs here start getting more elaborate, and another key
element is dealing with the dangerously heavy factorials that threaten to sink
the method; however this problem has been solved in Chapter 9. Yet another
feature is a simple form of zero estimate. These have proved crucial in later
developments involving Roth, Schmidt, Schlickewei, and others. More pre-
cisely, given any algebraic number α of degree d ≥ 3 and any κ > d2 + 1, we
show that there is a positive constant c = c(α, κ) such that

α − p ≥ c
q qκ
for all integers p and q ≥ 1. The Liouville result was for κ = d, and the
later Roth estimate was for any κ > 2. Here we try to break the proof into
molecules, and we also speculate on how Thue may have arrived at his proof;
there are interesting connexions with Newton’s Method in numerical analysis
and later improvements by Halley and others. We also give the applications to
diophantine equations. Here we encounter the uncomfortable phenomenon of
ineffectivity for the first time.
Then in Chapter 13, using the machinery of the previous chapter, we
prove the Hermite–Lindemann result on the transcendence of the values of
the exponential function at algebraic numbers; thus eα is transcendental for
every algebraic α = 0. Our proof is a kind of ad hoc development of the
auxiliary polynomial techniques introduced so far; we have by now illustrated
so many of these techniques that several proofs are available. We choose the
one most suited for generalization to the Schneider–Lang Theorem later on in
Chapter 19.
Chapter 14 is where we develop the size in Chapter 11 to the absolute height
H(α) ≥ 1 or the logarithmic version h(α) = log H(α) ≥ 0. This is rather easy
to define, but to establish properties like H(α 2 ) = H(α)2 , we need quite a
bit of algebraic number theory, and we will sketch the details. The motivation
is two-fold: first, the results of the next two chapters are about heights per
se, and second, the proof of the later Schneider–Lang result then becomes
fairly streamlined. We also give a version of the Siegel Lemma in the heights
language. This requires essentially defining the height of a vector (α1 , . . . , αn )
of algebraic numbers. To break the monotony, we prove on the way some easy
results on lower and upper bounds for heights that have led to some lively
modern developments.
Introduction xiii
Then in Chapter 15 we prove Bilu’s Theorem on the distribution of the con-

jugates of an algebraic number, using an auxiliary polynomial due to Mignotte
as well as the Siegel Lemma from the previous chapter. As a matter of fact,
our version is completely explicit numerically. But there is a problem: this
explicitness is based on the Erdős–Turán Theorem, and there seems to be no
easy proof of that. So at this point the book is definitely not self-contained;
however we find this didactically permissible, as the present chapter serves
as a natural springboard for the next one, and Bilu’s Theorem is not further
used in the book. More precisely, if α is an algebraic number of degree d and
absolute logarithmic height h, we show that the number n of its conjugates in
any sector of angle θ based at the origin satisfies

n − θ d ≤ 24(d 2/3 (log 2d)1/3 + dh1/3 ).
2π
That n is asymptotically 2πθ
d as h → 0 is the main content of Bilu’s result
(which is expressed more felicitously in terms of weak approximation).
Then in Chapter 16 using the machinery developed in the previous chapter,
we prove up to logarithms the famous Dobrowolski Theorem, which is to this
day the best approach to the classical Lehmer Question, using essentially the
original auxiliary polynomial. The result is exceptionally useful and, as far as
I know, none of the applications actually need the logarithms. Providing the
best known logarithms is an exercise on the Prime Number Theorem, which is
carried out in several books. Thus we prove here that for any κ > 1 there is a
positive constant c = c(κ) such that every non-zero algebraic α = 0 of degree
d which is not a root of unity satisfies
c
h(α) ≥ κ .
d
Admittedly there are quicker proofs without auxiliary polynomials, but these
don’t generalize to the higher dimensional results such as the Amoroso–David
Theorem that are very important today in diophantine geometry.
In Chapter 17 we restore some symmetry by giving a non-trivial height
upper bound. This concerns the algebraic numbers α with α n + (1 − α)n = 1
for some integer n ≥ 2. In a relatively recent investigation connected with irre-
ducibility, Beukers showed that H(α) ≤ 216. His proof used hypergeometric
functions. Using instead the powerful method of auxiliary polynomials, we get
H(α) ≤ 10120 (in the style of Stephen Leacock “ten years ago the deficit on my
farm was about a hundred dollars; but by well-designed capital expenditure,
by drainage and greater attention to detail, I have got it into the thousands”).
However this method generalizes considerably, as current work of Amoroso,
Zannier and the author shows.
xiv Introduction
In Chapter 18 we use some of the ideas developed so far to give a general-

ization to algebraic points of the 1989 Bombieri–Pila Theorem on counting
rational points on analytic curves. The original proof, although not funda-
mentally different from ours, is based on identities related to the confluent
Lagrange Interpolation Formulae and not on an auxiliary polynomial. Such
counting results (usually in higher dimensions) are nowadays being applied to
prove a variety of special cases of the general Zilber–Pink Conjectures about
unlikely intersections. We will prove something implying the following. Let f
be a transcendental function analytic on an open set containing the real interval
[0, 1]. Then for any > 0 there exists c = c(f , ) such that, for every positive
integer n, at most cn of the values

1 2
f (0), f ,f , . . . , f (1)
n n
are in Z/n. This vaguely resembles the reformulation of Pólya’s Theorem.

Then in Chapter 19 we prove the famous Schneider–Lang Theorem, which
includes Hermite–Lindemann in Chapter 13 as well as several other things
involving elliptic and abelian functions. Thanks to the preceding chapters the
proof is now reasonably smooth. It is a natural climax to the book; however
the next chapter follows on quite naturally, and so does the one after that.
Thus we prove the Gelfond–Schneider Theorem on the transcendence of αβ =
exp(β log α) whenever√α = 0 and irrational β are algebraic, which includes
the transcendence of 2 2 as specified by Hilbert in his Seventh Problem. A key
technical trick is the use of “large radius” in the Schwarz Lemma.
In Chapter 20 we systematically consider the elliptic analogues, motivated
partly by the need to prove the transcendence of integrals like

1 dX 5 (X − 8)dX
√ , √ .
0 X − X3 4 X 3 − 7X + 6
The results involve a Weierstrass function ℘ (z) with invariants g2 , g3 that are
themselves algebraic; the analogue of Hermite–Lindemann then asserts the
transcendence of ℘ (α) for any algebraic α = 0. Already the elliptic ana-
logue of Gelfond–Schneider has consequences for the modular function j(τ )
defined on the upper half-plane: namely that j(α) is transcendental whenever
α is algebraic but not quadratic; this we postpone to the next chapter. But as
Schneider discovered, there are several other interesting consequences; and
even he overlooked one of them. This chapter is the longest in the book, due to
our supplying the main details for the proofs of most of these consequences. It
might get shorter if we could use facts about commutative group varieties, but
Introduction xv
that would introduce too much algebraic geometry not in the elementary spirit
of the book.
In 1969 Mahler conjectured that the alternative modular function
J(q) = q−1 + 744 + 196884q + 21493760q2 + 864299970q3

+20245856256q4 + · · ·
(the pattern of coefficients is quite self-evident), defined for all q in the unit
disc, takes transcendental values at non-zero algebraic q. In 1996 Barré, Diaz,
Gramain and Philippon proved this using an auxiliary polynomial directly
on J itself. This was not only the first auxiliary polynomial of its kind (as
Schneider had wanted many years ago in his Second Problem), but it led soon
afterwards to Nesterenko’s unexpected breakthrough implying the algebraic
independence of π and eπ . This is a most attractive area where aspects of
elliptic, modular and exponential functions blend into each other. We give
a proof of Mahler’s Conjecture in Chapter 21, after deducing the analogous
result for j(τ ) = J(e2π iτ ) from the results of the previous chapter.
Up to now we never discussed problems of algebraic independence. Maybe
the reader’s curiosity for this topic has been awakened at the end of the pre-
vious chapter, and now she gets a classical example. After the famous Linde-
mann–Weierstrass result (not covered in this book), which was generalized to
E-functions by Siegel and Shidlovsky (likewise not here), the most spectacular
2
was the algebraic independence of α β and α β , for algebraic α = 0 and cubic
β, due to Gelfond in 1949. But as a lot of the machinery is already available,
our proof in Chapter 22 will not be too long. Here too one needs “large radius”.
Finally in an Appendix we prove exotic height results like

3ξ − 4 ξ 3 + 3ξ + 4 + 8
h ≤ h(ξ ) + 10000( h(ξ ) + 1)
ξ2
where a crude estimate would give at least 2h(ξ ) on the right-hand side. Indeed
if we replace 8 in the numerator by 7 this is unavoidable. The square root here,
traditionally associated with the quadratic nature of Néron–Tate heights on
abelian varieties, is actually needed.
Let us mention here yet another use for auxiliary polynomials: to show
that certain algebraic numbers arising from commutative group varieties have
“large degree”. It is well-known that the root of unity e2π i/n has degree φ(n)
the Euler φ-function, and also that for any θ < 1 there is a positive con-
stant c, of course effectively computable, with φ(n) ≥ cnθ . By a famous
result of Serre the elliptic analogue has any θ < 2, but only recently has
this been made effective, in an elaborate proof involving, among other things,
xvi Introduction
isogeny estimates. Using an auxiliary polynomial directly, in the functions

℘ (z), ℘ (Nz) which are “almost algebraically independent”, one can quickly
obtain an effective lower bound for any θ < 1. Furthermore this method works
also for abelian varieties, where the analogue of Serre’s Theorem is still not
yet fully known. The resulting estimates have recently been very useful in
problems of unlikely intersections. We omitted any detailed account, first for
lack of space and second because one needs more theory, such as Néron–Tate
heights. See Masser (1977) and also Appendix D of Zannier (2012). However
in Exercise 14.92 we sketch how the lower bound cn/ log n can be obtained in
the cyclotomic case.
The reader will observe that the auxiliary polynomial usually operates in a
proof by contradiction. So this book is mostly about things that don’t exist!
With Woody Allen we may hate reality but it’s still the best place to get a
decent steak. Or we may think of the Cape Town telephone company error
message “the number that you have called does not exist”.
The pleasant task of collecting together all these applications of auxiliary
polynomials has resulted in some features that may not be familiar to all
experts.
Thus I am not sure if Theorem 5.1 in Chapter 5 appears explicitly in the
literature. In Chapter 6 the warm-up before the proof of Theorem 6.1 may not
have appeared before in this form. In Chapter 7 the proof of Lemma 7.3 is
new, although it proceeds on well-known general principles. The (rather easy)
estimate (8.7) of Chapter 8 is probably new. In Chapter 10 the Proposition 10.4
might possibly be useful in other contexts. In Chapter 12 it is indeed I who must
accept full responsibility for the attempt to explain the proof of Theorem 12.1
in terms of numerical analysis; also the Proposition 12.2, although known to
some experts, may not have appeared explicitly before. Our explicit estimate
in Theorem 15.2 of Chapter 15 could be new, although its shape is fairly well-
known. Some of the preliminary discussion in Chapter 16 may not be familiar.
The method introduced in Chapter 17 is new, due to Amoroso, Zannier and
myself. In Chapter 18 the main result Theorem 18.2 for Q(i) is not in the
published work that I have seen, although here too its shape for Q is fairly well-
known. Lemma 20.7 from Chapter 20 might look familiar, but it is not; also
some of the details towards the end of this chapter have never appeared in print,
although this may well be due to the alternative approach, more conceptual to
some, through group varieties. In particular the proof of Theorem 20.11 might
well be a “desperately-needed gap in the literature”. Here also the (again rather
easy) remark about the gamma function is new. And in Chapter 21 the Lemma
21.8 enables us to avoid an appeal to certain estimates for coefficients of mod-
ular transformation polynomials, whose (non-classical) proofs are somewhat
Introduction xvii
elaborate. In Chapter 22 the proof of Proposition 22.5 is a small variation of

a proof that I have seen. Finally in the Appendix the Theorem A.1, although
presented only for a particular example, is also new, arising from the above
work of Amoroso, Zannier and myself.
What are the prerequisites for a happy reading of this book? The first thirteen
chapters could be understood by third-year university students or good second-
year students (and indeed in 2013/2014 they were – and I thank this class,
especially Gabriel Dill, who examined with a fine-toothed comb the first ten,
although I may well have invented new mistakes during revision). The proofs
are elementary (but that does not always imply that they are easy). Here there
are elements of algebra such as the concept of transcendence, the fact that
Z[X] is a unique factorization domain, or the integral closure of a ring R in
a larger ring S (which I like to denote by RS ); elements of analysis such as
order of vanishing, Cauchy’s Theorem or the Maximum Modulus Principle;
and elements of algebraic number theory such as field embeddings, conjugates
or rings of integers. There is a jump at Chapter 14, where we need slightly
more advanced algebraic number theory, which we explain without full proofs,
freely using concepts like prime ideals and valuations. This enables us to
get all the way to Chapter 20, where we then need some theory of elliptic
functions, which again we explain without full proofs. Similarly in Chapter 21
we need some theory of modular functions. Finally in Chapter 22 we need a
bit about transcendence degree. By contrast in the Appendix, although it has
a considerable whiff of diophantine geometry, we develop from scratch the
rudiments of algebraic curve theory that we need. And oh yes, it will be good
to bear in mind that our
N = {1, 2, 3, . . .}
does not contain 0 as it might in some other cultures. But Z, Q, R, C and the
fields Fp = {0, 1, . . . , p − 1} are unambiguous.
And what about the exercises? These are at the end of each chapter, where
they are divided into two parts by a starred line. Those above the line need only
the prerequisites above and what’s in the book so far, and they are essentially
what were fed to students as homework reinforcing the lectures. Those below
the line go further, and sometimes need extra knowledge; they are of varying
levels of difficulty, sometimes hinted at. Concerning the lectures themselves,
there are many possibilities; for example I covered Chapters 1, 2, 3, 4, 5
together with some algebraic number theory in a single semester, then followed
up with Chapters 6, 8, 9, 11, 12, 13 and more algebraic number theory, and so
was able to start a third semester with Chapter 14 in detail, then Chapters 15,
16, 18, 19 and bits of Chapter 20.
xviii Introduction
There is also a bibliography, but this has no pretence of being comprehen-

sive. Instead I have tried to restrict it to books, especially those that give a good
overview of the subsequent development of some of the topics treated here; but
I have also included some key original papers.
I conjecture, but have no time to prove, that every mathematics book with at
least 100 pages contains at least one misprint (possibly apart from those that
have gone through several editions – however in a 2008 seminar we did find
a mistake in Landau’s “Elementary Number Theory” (Chelsea 1958), despite
the author, according to Littlewood, reading proof sheets seven times, once
for each sort of error – curiously we could not find it again later, this “Lost
Mistake”). Boas has a conjecture that is shockingly stronger, and (continued
p. 94).
The book you are now reading is certainly no countexerample, and I apolo-
gize in advance for my misprints, howlers and blunders (and my King Charles’s
Head of continued fractions). In fact I was once thanked in print by a non-
English author for “teaching him mistakes”. I hope to be able to pass on these
skills to my readers.
And also to convey to them the joys of “doing transcendence” rather than
merely “doing mathematics”.
I gladly express my great gratitude to David Tranah of Cambridge University
Press, for his warm initial encouragement to write the book, for his gentle
reminders about actually writing it, and, once I gave in and started in earnest,
for his regular enquiries about its progress and his rapid and detailed answers
to my many questions.
1
Prologue
In this chapter we give a couple of examples where the method of auxiliary

polynomials is used for problems that have no diophantine character. Thus we
are not following Sam Goldwyn’s advice to start with an earthquake and work
up to a climax.
Here is maybe one of the simplest examples.
There is an old chestnut which often turns up in problem-solving sessions:
given a polynomial F in a variable X, can one always multiply it by a non-zero
polynomial to get a product involving only powers X p for p prime?
For example with F = X 100 + 1 we have X 3 F = X 103 + X 3 . But what about
F = X 100 + X 3 ? Here multiplying by some P = aX d will not do. However
(X 111 − X 14 )F = X 11 (X 100 − X 3 )F = X 11 (X 200 − X 6 ) = X 211 − X 17 .
At first sight it appears to be a difficult problem about primes, possibly in

arithmetic progressions. So what about F = X 1000 + X 100 + X 3 ?
Let us consider multiplying F by some

L
P= pi X i (1.1)
i=0
for unknown L and undetermined coefficients pi (not necessarily primes, but

they might be). Then PF has degree at most L + 1000, and its coefficients are
linear forms in the pi . We would like to eliminate the terms X n for n not prime
with 0 ≤ n ≤ L + 1000. There are
L + 1001 − π(L + 1000)

of these, where π(x) = p≤x 1 is the standard prime-counting function. If
we equate the corresponding coefficients to zero, then we get a system of
1
L + 1001 − π(L + 1000) homogeneous linear equations in the L + 1 unknowns

pi . By linear algebra this system is solvable non-trivially, provided
L + 1 > L + 1001 − π(L + 1000);
that is, π(L + 1000) > 1000. Every schoolgirl knows that there are infinitely
many prime numbers, so π(x) tends to infinity with x and there exists such an
L; for example L = 6927 (with Maple). So the answer is yes for this F; the
trouble of course is that we have to solve 6927 equations in 6928 unknowns to
get P explicitly.
The reader may now see first that this works for any F, and second that the
primes are irrelevant, in the sense that we may demand only powers X m in PF
with m in any prescribed infinite set; for example the elements of the sequence
4, 27, 3125, 823543, . . . of all m = pp .
This is perhaps the simplest application of the method of auxiliary polyno-
mials.
Here is a second example.
Consider the expressions
x = t2 + t, y = t2 + 1. (1.2)
How can we eliminate t? Common sense, or a general consideration of tran-
scendence degree, shows that there must be an algebraic relation between x, y
not involving t. And indeed a moment’s thought gives
x2 − 2xy + y2 + 2x − 3y + 2 = 0. (1.3)
But what about
x = t3 + t, y = t4 + t?
We could solve the first equation by radicals for t, and then substitute into the
second equation, and finally somehow clear the radicals. We get
x4 + 3x3 − 4x2 y − y3 + 3x2 − 5xy + 2y2 + 2x − 2y = 0 (1.4)
but I confess that here I just used Maple to calculate the resultant (see Chap-
ter 5) of
t3 + t − x, t4 + t − y
with respect to t.
What about
x = t1948 + t666 + 1, y = t1291 + t163 + t? (1.5)
Here the degrees are my year of birth and the (traditional) year of birth of
the earliest part of the Swiss Federation, where I first gave these lectures
Prologue 3
(in Basle, after earlier attempts in Ann Arbor, Constance, Hong Kong,
Heraklion and Vienna).
Maple doesn’t respond for 47 seconds; and then gives an incomprehensible
error message (the resultant that you are seeking does not exist). How can we
find this relation P(x, y) = 0?
Let us write
L L
P(X, Y) = pij X i Y j (1.6)
i=0 j=0
for unknown L and undetermined coefficients pij which are presumably inte-
gers as in (1.3) and (1.4). If we substitute (1.5) into P(x, y), then we obtain a
polynomial in t of degree at most 3239L. Its coefficients are linear forms in the
pij . If we equate these coefficients to zero, then we get a system of 3239L + 1
homogeneous linear equations in the (L + 1)2 unknowns pij . By linear algebra
this system is solvable non-trivially, provided
(L + 1)2 > 3239L + 1
for example if L = 3238.
This proves something: namely that there is a non-trivial relation of degree
at most 3238 in each variable. There may well be (L + 1)2 = 10491121 terms
in the relation, which accounts for Maple’s chickening out. And the trouble
for anyone, of course, is that we now have to solve 10487883 equations in
10491121 unknowns.
The reader may see first that this works for any two polynomials F, G in t
instead of (1.5), and second that it generalizes to more variables; for example
between any three polynomials in two variables there is a non-trivial alge-
braic relation (as would follow more simply by consideration of transcendence
degree).
This example is perhaps more typical of those to follow in these pages. After
the substitution (1.5) we may regard the function P(x, y) as having a large order
of vanishing at t = 0; so large, indeed, that it must vanish identically.
In both examples the goal is practically the auxiliary polynomial itself; so it
is hardly “auxiliary”.
We will see many more and subtler applications in this book (sadly not
Siegel’s Theorem (Siegel, 1955) about functions on compact manifolds, which
is a very sophisticated generalization of the second example – see however
Lemma 20.4 and Exercise 20.67). But before we start, let us ask: since there
appear to be so many terms in the relation connecting (1.5), what are the coef-
ficients like? If we normalize them to be integers, how big are they? We note
that the Cramér formulae for solving linear equations involve determinants
whose size is the number n of unknowns. Such determinants already have

n! terms, so their values are likely to be somewhat larger. Thus it would be
surprising if the integers in our relation were substantially less than the factorial
10491121!. And it can be seen that some of the entries of the determinants are
almost as large as 36476 (as in Exercise 1.17). This means that we could expect
some coefficients in P to have thirty thousand million (American thirty billion,
Swiss dreissig Milliarden) decimal digits (see however Exercise 8.13). So it is
doubtful if P could ever be expressed explicitly.
Exercises
1.1 Show that there is P = 0 in C[X] such that P(X)(X 1000 + X 100 + X 3 ) has
n2
the form Nn=0 an X .
1.2 Let F be in C[X] with degree at most D. Show that there is P = 0 in C[X]
n2
with degree at most D2 − D such that PF has the form N n=0 an X .
1.3 Let F be in C[X]. Find P = 0 in C[X] such that PF has the form
N 2n
n=0 an X .
1.4 Let F be in C[t] with degree at most D ≥ 1 and let G be in C[t] with
degree at most E ≥ 1 . Show that there is P = 0 in C[X, Y] with degree at most
D + E − 1 in each variable such that P(F, G) = 0.
degree at most E ≥ 1. Show that there is P = 0 in C[X, Y] with degree at most
E in X and degree at most DE − E + 1 in Y such that P(F, G) = 0.
1.6 Let t, u be independent variables, and let F, G, H be in C[t, u]. Show that
there is P = 0 in C[X, Y, Z] such that P(F, G, H) = 0.
∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗∗
1.7 Show that there is an absolute constant c (that is, not depending on any
parameters) with the following property. Let F be in C[X] with degree at most
D ≥ 2. Then there is P = 0 in C[X] with degree at most cD log D such that PF

has the form p ap X p , where p runs over the set of primes.
1.8 Let F be in C[X]. Find P = 0 in C[X] such that PF has the form
N 3n
n=0 an X .
1.9 Let F be in Fp [X]. Show that there is P = 0 in Fp [X] such that G = PF

satisfies G(X1 + X2 ) = G(X1 ) + G(X2 ) in Fp [X1 , X2 ].
Prologue 5
degree at most E ≥ 1. Show that there is P = 0 in C[X, Y] with degree at most
E in X and degree at most D in Y such that P(F, G) = 0 [Hint: resultants].
1.11 Let P = 0 in C[X, Y] be such that P(t1948 + t666 + 1, t1291 + t163 + t) =
0. Show that P has degree at least 1291 in X and degree at least 1948 in Y
(compare Exercise 5.7).
1.12 Can one essentially improve the D2 − D in Exercise 1.2? I don’t know.
1.13 Let F, G be in C(t) (rational functions). Show that there is P = 0 in
C[X, Y] such that P(F, G) = 0.
1.14 Let
(t2 − t + 1)3 (t2 + t + 1)3
F = 256 , G = 256 .
t2 (t − 1)2 t2 (t + 1)2
Show that P(F, G) = 0 for
P = X 3 Y − 2X 2 Y 2 + XY 3 − 1728(X 3 + Y 3 ) + 1216(X 2 Y + XY 2 )
+3538944(X 2 + Y 2 ) − 2752512XY − 2415919104(X + Y) + 549755813888.

(This is related to the simultaneous complex multiplication of two different
elliptic curves and also to the André–Oort Conjecture – see Exercise 21.19. Of
course the present exercise and the following are illustrations of Littlewood’s
Principle that “All identities are trivial (once they have been written down by
someone else)” as quoted by Cassels.)
1.15 Let
F = tu(t10 + 11t5 u5 − u10 ), G = −t20 − u20 + 228(t15 u5 − t5 u15 ) − 494t10 u10 ,
H = t30 + u30 + 522(t25 u5 − t5 u25 ) − 10005(t20 u10 + t10 u20 ).

Show that G3 + H 2 = 1728F 5 . (This is related to the icosahedron – see Klein,
1956.)
1.16 Let
u3 u2 v u(tuv − 3u3 − 4v2 )
F = 1728 , G = −1728 , H = −288 ,
u3 − v2 u3 − v2 u3 − v2
3t2 u2 v − 18tu4 − 24tuv2 + 95u3 v + 16v3
K = −24 .
u3 − v2
Show that
2F 2 (F − 1728)2 GK − 3F2 (F − 1728)2 H 2 + (F 2 − 1968F + 2654208)G4 = 0.
(This is related to the differential equation for the modular function – see
Exercise 21.15.)
1.17 With L = 3238 show that (t1948 + t666 + 1)L (t1291 + t163 + t)L has a
coefficient at least
32L
> 103082 .
L +L+1
2
1.18 Show that there is P = 0 in Q[X] for Exercise 1.1.

1.19 Show that there is P = 0 in Z[X] for Exercise 1.1.
1.20 Show that there is P = 0 in Z[X, Y] for Exercise 1.10.
1.21 Find P = 0 in Z[X, Y] with P(t + i, t − i) = 0.
1.22 If x, y are in C with (1.3), must there exist t in C with (1.2)?
1.23 What about Exercise 1.22 with C replaced by R, Q, Z, Fp ?
2
Irrationality I
The main application of the method of auxiliary polynomials is in diophantine

approximation and transcendence. But before these topics comes irrationality:
one seeks to prove that a given number is not in Q. One of the earliest examples
is of course
∞
1 1 1 1 1 1 1
e = = + + + + ··· = 1 + 1 + + + ··· .
k! 0! 1! 2! 3! 2 6
k=0
It is proved in any elementary text on number theory that e is irrational;

the proof is based on the rapid convergence of the series together with the
reasonable behaviour of the denominators. We give a proof nevertheless.
Consider the truncation

n ∞

1 1
fn = e − =
k! k!
k=0 k=n+1
for n = 0, 1, 2, . . .. The first term on the extreme right-hand side dominates

and indeed
∞
1 1 1 2
= 1+ + ··· < .
k! (n + 1)! n+2 (n + 1)!
k=n+1
Thus 0 < fn < 2/(n + 1)! and

2
0 < |n!fn | < . (2.1)
n+1
Now if s is a denominator for the rational e, then multiplying by s and
making n tend to infinity gives a contradiction to the so-called Fundamental
Theorem of Transcendence that every non-zero integer has absolute value at
least 1.
7
The proof is slightly easier for the alternating series

∞
(−1)k 1 1
e−1 = = 1−1+ − + ··· ,
k! 2 6
k=0
because we no longer need the dominance.

The proofs extend to give the linear independence over Q of 1, e, e−1 , which
amounts to the fact that e cannot be quadratic over Q. In particular e2 is
irrational. But there is a minor snag. We assume that r + se + te−1 = 0 for
integers r, s, t not all zero, and then for

n
n ∞
∞

1 (−1)k 1 (−1)k
fn = r + s +t = −s −t ,
k! k! k! k!
k=0 k=0 k=n+1 k=n+1
we get
2 2
|n!fn | ≤ |s| + |t| .
n+1 n+1
Hence the n!fn are integers tending to zero as n tends to infinity. But we no
longer know that these integers are non-zero as in (2.1).
In fact it is not too hard to show that
fn = fn+1 = fn+2 = 0 (2.2)
is impossible for any n. Namely,
(n + 1)!(fn+1 − fn ) = s + (−1)n+1 t
and so
(n + 2)!(fn+2 − fn+1 ) = s − (−1)n+1 t.
Thus (2.2) would imply s = t = 0 so r = 0 too, a contradiction.

Here the problem makes its first appearance but is relatively harmless; how-
ever in later chapters we will see it getting more and more dangerous.

But as soon as we consider e2 = ∞ k
k=0 2 /k! directly, some other difficulties
arise. The convergence is practically just as fast, but after truncating at k = n
and multiplying by a denominator n!, we get a term 2n+1 /(n + 1) which no
longer tends to zero.
The proof can be fixed by calculating the power of 2 dividing n! to yield a
smaller denominator; this involves restricting n to a special form like 2m . Such
a trick can be extended to give the linear independence over Q of 1, e2 , e−2
(see Exercise 2.13); in particular e4 is irrational. But it is amusing (I learnt it
Irrationality I 9
from the wonderful book of Conway √ and Guy (1996), p. 253) that these ideas
also give the irrationality of λ = e 2 via
2l ∞
1
λ+ = 2
λ (2l)!
l=0
(see Exercise 2.4).

Actually I know of no such simple proof that e3 is irrational, although this is
1
not difficult to establish by considering 0 e3t tn (1−t)n dt (see Exercise 2.14). It
can also be done with auxiliary polynomials of the type mentioned in Chapter
1, most efficiently by introducing derivatives as in the differential equation
(ez ) = ez . However some extra arithmetic and analytic machinery is needed.
This will be introduced step by step in the following chapters. By these means
we will show in Chapter 13 that
∞
αk
eα =
k!
k=0
is irrational and even transcendental for any rational and even algebraic α = 0.
In the present chapter we will consider the series
∞
zk 1 1 1
= 1 + z + z2 + z3 + z4 + · · · , (2.3)
2k(k−1)/2 2 8 64
k=0
which similarly converges for all real and even complex z to a function f (z).
It converges faster than the series of ez , so let us see what truncation of f (α)
gives.
Let us start with α = a/b, for simplicity taking a ≥ 1, b ≥ 1 in Z. The
truncations are

n ∞

αk αk
fn = f (α) − = .
2k(k−1)/2 2k(k−1)/2
k=0 k=n+1
Again there is domination on the extreme right, and if n is large enough we get
α n+1 2an+1
|fn | ≤ 2 = . (2.4)
2n(n+1)/2 2n(n+1)/2 bn+1
Taking into account a common denominator 2n(n−1)/2 bn , we deduce
2an+1 2a a n
|2n(n−1)/2 bn fn | ≤ = .
2n b b 2
The proof works if the estimate tends to zero as n tends to infinity (assuming
we can rule out the snag fn = 0 as in (2.2) above). Unfortunately this is the
case only for a = 1.
Another random document with
no related content on Scribd:
CHAPTER VII.
A STRANGE FANCY.
“What was that?” asked Mabin with a shiver.

She and Rudolph had both turned instinctively toward the spot
from which the rustling noise had come.
“A cat, most likely,” answered Rudolph.
But Mabin shook her head.
“I saw something,” whispered she. “It was not a cat, it was not an
animal at all; it was a man.”
Perhaps Rudolph had his suspicions, for he expressed no
surprise. Before he could answer her they heard the crackling and
rustling again, but at a little distance. The intruder was making his
way through the shrubbery.
“Won’t you find out who it is?” whispered the girl again.
Rudolph hesitated.
“Perhaps I know,” said he shortly. “But if you wish, of course I can
make sure.”
Then, with evident reluctance, and taking no pains to go
noiselessly, he followed the intruder through the bushes, and was in
time to catch a glimpse of him as he disappeared over a part of the
fence that was in a broken-down condition. Rudolph did not attempt
to continue his pursuit, but contented himself with waiting until he
heard the side gate in the garden wall of “Stone House” swing back
into its place with a loud creaking noise. Then he went back to
Mabin. She was standing where he had left her, on the broad gravel
path under the faded laburnum. The shadows were very deep under
the trees by this time, and in the half-light her young face, with its
small, delicate features, its dreamy, thoughtful eyes, full of the
wonder at the world of the very young, looked so pretty that for the
moment Rudolph forgot the errand on which he had been sent, and
approached her with no thought of anything but the beauty and the
sweetness of her face.
She, all unconscious of this, woke him into recollection with one
abrupt word: “Well?”
“Oh!” almost stammered he, “it was as I thought, the same person
that I saw watching before.”
“And he went into our garden. I heard the gate,” said Mabin with
excitement. “It must be this Mr. Banks. Oh, who do you think he is?
What do you think he has come for?”
Rudolph was silent. Even to the least curious mind the
circumstances surrounding both him and Mrs. Dale could not seem
other than mysterious. If he were a detective, and he certainly did
not look like one, surely he would not go to work in this extravagant
manner, by renting a large and expensive house merely for the
purpose of watching his next-door neighbor. Neither, it might be
supposed, would he set to work in such a clumsy fashion as to be
caught making his investigations at the very outset. Rudolph felt that
the whole affair was a mystery to which he could not pretend to have
the shadow of a clew. He confessed this to Mabin.
“I wish,” he went on, in a gentle tone, “that I had known something
of this before your father went away.”
“Why?” asked Mabin in surprise, and with something like revolt in
her tone.
“Because I should have told him something, just enough at any
rate to have made him take you away with him.”
Mabin was for a moment dumb with surprise.
“What,” she stammered at last, “after all your talk about my being
right to stand by my friend?”
“Even after all that,” assented Rudolph with decision. “The matter
is getting too serious,” he went on gravely. “I am afraid myself of
what may be going to happen.”
“Then,” retorted the girl, “for all your talk about meanness being
excusable in a girl, I can be a better friend than you.”
Rudolph smiled.
“Ah,” said he, “you forget that with you it is only a question of your
friendship for Mrs. Dale. Now I have to think of both of you.”
“You need not trouble yourself about me, I assure you.”
“But that is just what I must do, madam, even at the risk of your
eternal displeasure,” said Rudolph, with a mock-heroic air which
concealed real anxiety. “You are not only daring enough, you are too
daring where your heart is concerned, and it is the business of your
friends to see that you do not suffer for your generosity.” He spoke
with so much quiet decision that Mabin was impressed and rather
frightened, and it was with a sudden drop from haughtiness to
meekness that she then asked:
“What are you going to do, then?”
Rudolph hesitated.
“What I should like to do,” said he, “is to take you to my mother’s
——”
Mabin almost screamed.
“You won’t do that,” she said quietly, with her lips very tightly
closed.
“She would be very kind to you,” suggested Rudolph gently,
pleadingly.
He knew the prospect was not an enticing one, but he was not so
quick as the girl to see all its disadvantages.
“And don’t you see that it would set them all saying the most
dreadful things about poor Mrs. Dale, if I were to leave her suddenly
like that? I shouldn’t think of such a thing. It would be cruel as well as
cowardly. She would never be able to stay in Stone after that.”
“I don’t think she will be able to stay in any case,” said Rudolph
gloomily. “If she is persecuted by this spy on the one hand, and by
the old woman on the other, it isn’t likely that she will be able to stay
here long.”
A new idea flashed suddenly into Mabin’s mind and then quickly
found expression:
“Do you suppose,” she asked, “that this man, this Mr. Banks, is
paid by the old lady to spy upon Mrs. Dale? The old lady must be
very rich, I think, and she is eccentric evidently.”
But Rudolph was inclined to think this idea far-fetched. From what
he had seen of the mysterious spy he had come to quite another
conclusion, one that at present he did not care to communicate to
Mabin, for fear of alarming her unnecessarily.
“Of course it is possible that the man may be a paid detective,”
admitted he doubtfully, “but there was nothing of the cut of the ex-
detective about your Mr. Banks. And now,” went on Rudolph, who
found Mabin herself a more interesting mystery than the unknown
man, “let us forget all about him for a little while, and go up to the old
seat where the trees leave off, before it gets too dark for us to see
the sea. You remember the old seat, and how we used to trespass to
get at it, don’t you?”
Mabin blushed a little. She remembered the old seat very well; an
old broken-down bench supported on the stumps of a couple of
felled trees, just on the edge of the plantation belonging to “The
Towers.” Being conveniently near both to “Stone House” and the
Vicarage, the children of both houses had established, in those far-
off years which Rudolph was recalling, a right to tread down the old
fence at that particular point, and to hold wonderful picnics of
butterscotch and sour apples.
“We won’t go up there now,” she said, with a sudden demureness
which contrasted strongly with the eagerness she had shown while
discussing the persecution of Mrs. Dale. “It’s getting dark, and rather
cold, I think, and besides, I hope by this time that Mrs. Dale may be
ready to see us again.”
Rudolph felt snubbed. The girl’s manner was so precise, so stiff,
that it was impossible for him to understand that her sudden
primness was only a relapse into her ferocious girlish modesty. He
followed her without a word toward the house, and there just inside
the portico they saw the slight figure in black looking like a pathetic
vision in the gloaming, with its white, tear-stained face and slender
little jewelled hands.
“Well?” said Mrs. Dale. And her voice was hoarse and broken. “I
have been waiting here for you, wondering where you had gone. I
had almost begun to think,” she went on, with assumed playfulness,
which did not hide the fact that her fear had been real, “that you had
run away from me altogether.”
Mabin lost her awkwardness, her stiffness, her shy, girlish reserve
in an instant; moved by strong pity and affection, she took the two
steps which brought her under the portico, and stooping, flung her
arms round the little figure.
“You didn’t—really?” she whispered hoarsely. “Oh, I hope not, I
hope not!”
Mrs. Dale could not answer. But Mabin felt her frame quiver from
head to foot, and heard the sound of a stifled sob. Rudolph stepped
noiselessly out into the garden again.
“My dear, my dear child,” murmured Mrs. Dale, when she had
recovered some of her self-possession by a strong effort, “you would
have been quite justified if you had gone. But I am glad, oh, so glad,
that you have waited for me to drive you away.”
“You won’t do that!” cried Mabin, starting back, and seeing with
surprise in the fair, blue-eyed face an expression of strong
resolution. “After pretending you were so glad to have me!”
“It was no pretence, believe me!” said Mrs. Dale with a sad little
smile. “But I have got to send you away all the same. It would not be
right to keep you here, now that I see the persecution I am to be
subjected to still.” And her blue eyes flashed angrily as she spoke.
But the next moment her face changed again, and she added
quickly, “I have deserved it all. More than all. I am not complaining of
that; I have no right to complain. Only—she might have spared you. I
should have done you no harm; you would have learnt no evil from
me, wicked as I am.”
The girl interrupted her, with a frightened face, and speaking in an
eager whisper:
“Oh, hush, hush! You are not wicked. It is dreadful to hear you say
such things! I will not let you say them. You have the kindest heart in
the world; if you have ever done wrong, you are sorry, bitterly sorry.
Wicked people are never sorry. Let me stay with you and comfort
you if I can, by showing you how happy it makes me to be with you!”
Mrs. Dale shook her head. She did not, however, repeat in words
her resolve that Mabin must go, though the girl guessed by the
expression of her face that her mind was made up on the subject.
They stood silently looking out at the soft beauties of the twilight,
the greens as they melted into grays blending in such a tender
harmony of color that the sight seemed to supply a balm, through
tear-dimmed eyes, to their heavy hearts; the scent of the roses came
to them across the broad space of gravel, too, mingled with the
pleasantly acrid perfume of the limes.
Rudolph’s step, as he took advantage of the silence to thrust
himself again upon the notice of the ladies, startled them both.
“Now you’ve spoilt it all!” cried Mrs. Dale, in a tone which was
meant to be one of light-hearted pleasantry, but which betrayed too
plainly the difficulty she had in assuming it. “The garden looked like a
fairy picture till you rushed in and ruined the perspective. Aren’t you
going to apologize?”
“No. The picture wanted human interest, so I painted myself into
the canvas, just to satisfy your artistic susceptibilities. I am sorry to
find you so ungrateful. I hope you, Mabin, have more appreciation?”
But the girl’s eyes were full of tears, and not being used to this
light strain of talk, she could not answer, except by a few mumbled
words which had neither sense nor coherence. Mrs. Dale put up her
hand—she had to stretch it up a long way—and smoothed the girl’s
pretty brown hair.
“Don’t tease her,” she said softly. “Mr. Bonnington, I mustn’t ask
you to dine with us, but I would if I might.”
“And why mustn’t you?” asked Rudolph.
“Well, because, in the miserably equivocal position I am in, it
would be a pleasure—if I may take it for granted that it would be a
pleasure to you, as it would certainly be to me—dearly bought. The
Vicar would strongly disapprove; your mother would be shocked
beyond measure.”
“But I shouldn’t mind that, I assure you. I’ve shocked my mother
and excited the disapproval of my father so often that they don’t
expect anything else from me. Besides, I am afraid you flatter
yourself too much in believing that you have such an enviable
peculiarity; if you were to issue invitations to the whole parish to a
garden party, or a dinner, or anything you liked, I’m afraid you would
be disappointed to find that everybody would come.”
“Perhaps they would think there was safety in numbers, and that,
fortified by the presence of everybody else, they could gaze at the
monster in security!” suggested Mrs. Dale with a smile.
“In the mean time how much nearer have I got to get to inviting
myself to dinner this evening?” said Rudolph, with a subdued voice
and a meek manner.
“Ah, well, for Mabin’s sake then, I spare you the humiliation and
invite you myself. You shall stay to amuse her, since I am afraid she
would find me a very dreary companion.”
“Indeed I shouldn’t,” cried Mabin, blushing deeply, and speaking
with as much energy as if the presence of Rudolph were an injury. “I
should like nothing better than an evening alone with you.”
Rudolph drew a deep sigh, and even Mrs. Dale could not suppress
a smile at the girl’s unconscious gaucherie. When Mabin realized
what a stupid thing she had said, she was of course too much
ashamed of herself to laugh at her clumsy words, and fell, instead,
into a stiff silence which the others found it impossible to make her
break except by demure monosyllabic answers.
When they went into the dining-room, therefore, the evening did
not promise to be a lively one. Mrs. Dale seemed to find it impossible
to shake off the effects of the visit of her persecutor. Rudolph was
oppressed by fears for both the ladies, and by doubts whether his
presence there was not an indiscretion which would make matters
worse for both of them. While Mabin, perplexed and troubled by a
score of unaccustomed sensations, was the most silent, the most
distressed of all.
Daylight was still streaming in from the West as they took their
seats at the table in the dingily furnished room. Mrs. Dale gave a
little shudder as she glanced from the “furnished house” knives to
the commonplace dinner service.
“Ah!” she said, “it is not like this that I used to entertain my friends.
My little dinners had quite a reputation—once!”
Then, as if she felt that these regrets were worse than vain, she
turned the subject abruptly, while a spasm of pain for the moment
convulsed her face.
Rudolph on his side was sorry she had mentioned the “little
dinners.” They suggested a past life in which there had been
something more than frivolity; something with which he would have
dissociated Mrs. Dale if he could. But innocent Mabin, wishing to say
something, brought the conversation back to the point it had left.
“But why can’t you have pretty dinners now, if you like to?”
Mrs. Dale’s fair face grew whiter as she answered gently:
“I will tell you—presently—some day—why I don’t have anything
pretty or nice about me now.”
And Mabin, feeling that she had touched a painful chord, became
more silent than ever.
Perhaps it was her sudden subsidence into absolute gloom which
caused the other two to make a great effort to restore something like
animation to the talk. And being both young, and of naturally high
spirits, they succeeded so well that before the meal which had begun
so solemnly was over, Mrs. Dale and Rudolph were talking and
laughing as if there had never been a shadow upon either of their
lives. At first they made brave attempts to drag Mabin into the
conversation. But as these efforts were in vain, it naturally ended in
her being left out of the gayety, and in her sitting entrenched in a
gloomy silence of her own.
And when dinner was over, and they all went into the little
adjoining room which Mrs. Dale called her “den,” it was quite natural
that Mrs. Dale should sit down at the piano, in the good-natured wish
to leave the young people to entertain each other; and equally
natural that Rudolph, on finding that Mabin had nothing to say to
him, and that she was particularly frigid in her manner, should go
over to the piano, and by coaxing Mrs. Dale to sing him his favorite
songs and then hers, should continue the brisk flirtation begun at the
dinner-table.
Mabin had brought it all upon herself, and she tried to persuade
herself that it was quite right and natural, and that she did not mind.
And when Rudolph was gone, and she was alone with her hostess,
she succeeded in persuading her that she had not felt neglected, but
had enjoyed the merriment she had refused to share.
But when she got upstairs into her pretty bedroom, after bidding
Mrs. Dale good-night, she had the greatest difficulty in keeping back
the tears which were dangerously near her proud eyes.
She did not care for Rudolph, of course not; she wanted him to fall
in love with Mrs. Dale, if indeed he had not already done so, and
marry her and console her for all her troubles, and stop the
persecution of “the cat.” But somehow this hope, this wish, did not
give her all the unselfish satisfaction it ought to have done.
And Mabin, wondering what had happened to take the prettiness
out of the room and the pleasure out of her acquaintance with Mrs.
Dale, fell asleep with her heart heavy and full of nameless grief.
She woke with a start to find a white figure standing motionless in
the middle of the room. Mabin sprang up in bed and rubbed her
eyes. Was she awake? Or was she only dreaming that the body of a
dead woman, stiff, rigid, but in an upright position, was standing like
a marble statue between the bed and the nearest window?
She leaped out of bed, and, not without uncanny fears, touched
the statuesque figure.
“Mrs. Dale!” she almost shrieked, as the great eyes suddenly
turned and fixed a blank, wild gaze upon her face. “Oh, what has
happened? What is the matter?”
The figure, which, in white night garments, had looked so unlike
the black-robed widow that she had not recognized it, trembled from
head to foot. The lips parted, but at first no word escaped them. At
last with a strong effort she uttered these words:
“Let me stay here. Let me sit in this arm-chair till morning. Oh, I
will not hurt you, or frighten you. But if I go back I shall go mad! This
house is haunted, haunted! I have seen——”
A hoarse rattle in her throat seized her, threatened to choke her.
With one wild glance round, peering into the corners of the room,
she flung herself on the floor, and buried her face in the chair.
CHAPTER VIII.
A HAUNTED HOUSE.
Mabin was taken so thoroughly by surprise, on seeing the wild

self-abandonment of her unhappy companion, that for a few minutes
she stood staring at the crouching figure on the floor like one only
half-awake.
Was this really Mrs. Dale, this haggard, panting creature with the
hoarse voice, the twitching hands, the wide eyes full of unspeakable
terror. Mabin’s sympathy was ready, but at first she did not dare to
offer it. Such terrible anguish, such paralyzing fear, as that from
which the miserable woman was suffering, was something surely
beyond her poor powers of comfort! And even as the girl advanced
timidly a step nearer to her grief-stricken friend, there flashed into
her mind the horrible question: What must this secret be which was
locked in the widow’s breast, that could throw her into such
paroxysms of abject terror? For, not unnaturally, Mabin came to the
conclusion that the vision which had alarmed Mrs. Dale was one of
the results of the remorse from which she owned that she was
suffering.
“Don’t! Don’t sob like that! You will make yourself ill; you will
indeed. There is nothing, there is nobody here to frighten you,” said
the girl at last, drawing a little closer to the crouching figure, but not
yet daring to touch her, or to speak in a tone louder than a whisper.
At the first sound of her voice, Mrs. Dale had started, and raised
her head quickly, turning to the girl’s view a face so much altered, so
drawn, so old-looking, that she hardly recognized the features of the
lovely widow. Then, when the voice ceased, she glanced round the
room again, with the same hunted, anxious look as before.
“Nothing—nobody to frighten me!” she repeated in a shaking
voice. “No, of course not, of course not. How silly I have been! I am
afraid I frightened you, dear,—with my dreams, my silly fancies!”
She struggled, as if worn out and exhausted by her emotion, to
gain her feet. Timidly, gently, Mabin helped her to rise.
“I’m very glad I was here,” answered Mabin, in kindly tones that
sent a shiver of grateful recognition through her agitated companion.
“Do you feel better now?”
“Yes, oh, yes, I am all right. I am not ill. I am so much ashamed of
myself for disturbing you. I don’t know how to apologize,” answered
Mrs. Dale, trying bravely to speak in her usual tone, but glancing at
the door and then back to the windows as she uttered the words: “It
must have been a dream, of course, that frightened me.”
And then, quite suddenly, she broke down again, and slipping from
the supporting arm of her young companion, she threw herself into
the wicker arm-chair, and burst into a passion of tears. Uncertain
what to do, Mabin, in her sympathy and kindness, did exactly the
right thing. She drew another chair besides the wicker one, sat down
in it, and putting her right arm round Mrs. Dale’s shoulder, and
holding the poor lady’s trembling fingers in her own, remained in
perfect silence until the first ebullition of violent grief had passed
away.
“I shall never forget your kindness, child, never,” said Mrs. Dale,
when, as suddenly as it had begun, her passion of tears ended. “You
have saved me from going mad—yes, mad. I—I must leave you now,
or you won’t get any rest.”
She rose as she spoke; but Mabin saw that the panic of terror
which had been upon her at her entrance was regaining its hold
upon her as she approached the door. With her fingers on the handle
she stopped, and seemed once more to grow rigid with fear.
Mabin was by her side in an instant.
“Stay here,” she said. “You will have the dream again perhaps, if
you go away by yourself.”
At these words a shiver ran through Mrs. Dale, and she faltered.
“It must have been that gloomy room!” she said at last in a
whisper. “And the effect of her visit! But it will kill me if it comes
again!” Suddenly she turned to Mabin. “May I lie on the sofa until the
morning?” she asked piteously. “I won’t disturb you. I feel as if I
should be safe from—it—in here with you?”
The wistful pleading in her eyes brought the tears to Mabin’s.
“Of course you must stay,” she cried heartily. “And I do hope you
will get to sleep, and not have any more dreams.”
Very quietly Mrs. Dale lay down on the couch between the
windows, and drawing the sofa blanket over her, and refusing any
other covering, she closed her eyes. Mabin knew that this apparent
tranquillity was assumed only, and she placed herself on the bed in
such a position that she could watch her friend, while appearing to
be herself asleep.
Before many minutes had passed, she saw, from between half-
closed eyelashes, that Mrs. Dale was sitting up, and bending her
head in a listening attitude. And presently the slender figure with its
white dressing-gown slipped softly off the sofa, and hurried on tiptoe
across the floor to the door. There it knelt down and listened again.
And after a few minutes Mrs. Dale turned the key in the lock and
crept back, not to the couch, but to the arm-chair.
Mabin shut her eyes and tried to disentangle the knot of strange
ideas that filled her brain:
What was the nature of the secret which weighed on the
conscience of Mrs. Dale? Why was she kept in luxury by the very
woman who tried to make her life unbearable, to cut her off from
every human friend? What was the strange tie between the hard,
elderly woman and the impulsive, volatile young one? What was the
vision which had caused her so much distress? And, above all, why,
if it was only a vision, did she try to keep it away by locking the door?
And why—and why—? More questions surged up into her tired
brain; but Mabin forgot them as they rose. She fell asleep.
When she awoke in the morning it was to find that some one was
knocking at the door, and then she heard the housemaid’s voice
announcing that it was eight o’clock. She sprang up, and looked
toward the sofa, but there was no one but herself in the room.
Surely, she thought, the strange visit of the night must have been
a dream? The rug on the sofa was neatly folded, just as it had been
when she came up to bed last night. Not a sign was to be seen of
any intrusion during the night.
Even when she went downstairs and met Mrs. Dale in the hall,
there was little to tell of the experience of the hours of darkness.
Perhaps the pretty widow looked a little paler than usual, but in every
other respect she was the same airy, impulsive creature, now
smiling, now looking sad, as she had been before the dreadful visit
of the lady whom irreverent Mabin called “the cat.”
It was not indeed until breakfast was over and they had gone out
into the garden to cut some flowers while the dew was on them, that
either of the ladies made any reference to the events of the night.
Mrs. Dale, with one daintily shod foot in a flower-bed, was
stretching out her hands toward a bush of sweet-peas, when, without
turning her head, she said:
“I am in great trouble about you, Mabin.”
“Are you? Why, Mrs. Dale?”
“I don’t quite know what to do with you. If I send you to Mrs.
Bonnington, I shall have to tell some shocking tarradiddle about the
drains having come up, or the roof having given way, and she will be
sure to find me out and to pry, and to give both of us what the old
women call ‘much unpleasantness.’ And if I send you on to Geneva,
I don’t know whether they will be glad to see you when you arrive.”
“And I’m sure they won’t,” said Mabin heartily. “And there is one
other objection to sending me anywhere, and that is that I won’t go.”
Mrs. Dale dropped her sweet-peas, and turned round. Her eyes
were full of sudden tears.
“Nonsense, child!” she said sharply, but in a querulous tone which
betrayed her emotion, “nonsense! It was decided yesterday
afternoon that you were to go. You know it was.”
“You decided that I was to go. I didn’t. And—” instinctively she
dropped her voice—“And something that happened last night—in the
night, made me decide not to go. There!”
“But, my dear——”
“No, Mrs. Dale, I’m not to be ‘got round.’ You’ve chosen to take me
upon your shoulders, so now you must just keep me. Ha, ha! You
didn’t know I had so much determination, did you?”
But Mrs. Dale could scarcely speak. Now for the first time that
morning Mabin realized that the scene of the night had really taken
place, for the emotion aroused by this little bit of talk had brought
back into Mrs. Dale’s blue eyes a faint reproduction of the wild terror
she had shown when she came to the girl’s room. When she had
recovered her voice, the lady in black, pale, hoarse, shaken with her
agitation, stammered out these words:
“My dear girl, it is beautiful of you to offer to stay. But I cannot let
you. You ought never to have come. I was mad, wicked to let you
come; and my madness and my wickedness I must bear alone.”
How strange these words seemed in the broad daylight, Mabin
thought! By the weak glimmer of the night-light Mrs. Dale’s wild looks
and words had seemed fantastic, weird. But the broad sunlight
seemed to give the nameless horror which hung about the poor little
lady in black a reality as vivid as it was painful. But with this feeling
there came also into the heart of the young girl a great tenderness
toward the suffering woman, who was haunted by the shadow of her
own past. So she smiled, and with a pretty, half-shy look in her eyes,
said:
“You told me I saved you from going mad. So I mean to stay. And I
mean to sleep in the same room with you, so that you shall not be
frightened any more.”
Mrs. Dale shook her head.
“I can’t let you do that,” said she. “I don’t sleep very well, and
sometimes I start up and cry out. I should frighten you.”
“Then we will exchange rooms,” said Mabin.
By the look of joy and relief which flashed over Mrs. Dale’s face at
this suggestion, Mabin saw that she had conquered.
“But—won’t you be afraid?” asked the widow in a troubled voice.
“What! Of a ghost, a vision? Or of having bad dreams? No, not a
bit.”
Mrs. Dale glanced gratefully at the young face, with its look of
robust Philistine scorn of phantoms.
“It is a temptation,” she murmured. “For, after all, I know, I know
that it was only a dream, a horrible dream. And there is no fear that
the dream will come to you.”
“And if it did,” retorted Mabin stoutly, “it wouldn’t frighten me. I’m
too ‘stodgy;’ I have no imagination.”
Mrs. Dale smiled sadly.
“You are right,” she said. “If you did have the same dream, it would
have no terrors for you. Your conscience is clear.”
“And my digestion good,” added Mabin lightly, as she picked up
the fallen flowers and put them in her basket.
There was no doubt that her refusal to go had taken a load of
melancholy from the shoulders of her hostess, who sent the young
girl out for a walk as soon as the gathering of the flowers was over,
and charged her not to go far enough to tire her still weak ankle.
Mabin, with a book and a sunshade, sauntered slowly down the
hill to the nearest gap in the cliff, and went down the steep descent
to the sands. This was no paradise of nursemaids and babies, but a
solitary nook beloved by quiet maiden ladies and sentimental
couples. With rash disregard of the danger of sitting under a chalk
cliff, Mabin found a seat on a rock worn smooth by the sea, opened
her book and began—not to read.
The circumstances to which she found herself were far too
interesting for her to be able to give herself up to the milder
excitements of fiction. She sat with her open book on her lap and her
eyes staring out at the sea, which was vividly blue in the strong
sunshine, when she became suddenly conscious of a footstep she
knew in her immediate neighborhood.
Although she affected to be surprised when Rudolph appeared
before her, she had known that he was approaching, and her heart
began to beat very fast. He looked down at her between the spikes
of her sunshade, pretending to be afraid to speak to her.
“Good-morning,” said she at last.
“I was wondering whether I dared say the same thing!”
“Dared?”
“Yes. After your treatment of me last night, I felt nervous.”
“My treatment of you! What treatment?”
“Why were you so unkind? Or mustn’t I ask why?”
“You may ask, of course. But I can’t give you any answer, because
I didn’t know that I was unkind.”
“I wish I could believe that.”
“Well, if you won’t believe it, I have nothing to say.”
Rudolph was silent a few minutes. Then with a burst of explosive
energy, he made up his mind.
“No!” he cried so loudly that Mabin started, and threw himself
down on the sand beside her, “I will not be daunted. I will encase
myself in double snub-proof armor plates, and I will try to teach her
that to be dignified it is not necessary to be unkind—and—yes, I will
say it—absolutely rude.”
Mabin became crimson, and the tears started to her eyes. She
had not meant to be rude, but undoubtedly her behavior had laid her
open to this accusation.
“I am stupid, clumsy; I am rude without meaning it,” she said in a
tone of such excessive humility and penitence that it was impossible
to doubt her sincerity. “I am very sorry. But you shouldn’t take any
notice of what I do or say. Nobody does at home. When I am more
awkward and tiresome than usual, they always say: ‘Oh, it’s only
Mabin!’ And then nobody minds.”
“Ah, well, I can’t quite feel like that—that it’s only Mabin. When one
likes a person, and wants to be good friends, very good friends with
that person, just as one used to be when that person and one’s self
were little things in short frocks and knickerbockers, it is very
disheartening to find that person so determined to be—er—to be—er
—so reserved that when one sits beside that person as I did last
night, you know, she will only let one see so much of her right ear as
to practically turn her back to one!”
“I didn’t!”
“You did though. And it is what you were doing again just now until
the horror of hearing the truth made you turn around to fly at me! You
did turn your back upon me last night, Miss Rose, and you hurt my
feelings.”
“Indeed, you did not seem to be hurt. You seemed to be enjoying
yourself very much!”
“Well, so I was in a way. But I should have enjoyed myself much
more if you had been as nice as you were in the garden.”
Mabin heaved a deep sigh.
“It’s no use expecting me to be nice,” she said in a voice of
despair, “I can only manage it so very seldom.”
“Well, could you hold out some signal, such as by wearing a
particular flower, or color, or some especial knot of ribbon, to let one
know when one may speak to you without being snubbed?”
“No, I couldn’t,” retorted Mabin with great fierceness, but with a
twinkle of fun in her gray eyes, which gave greater hopes than her
words did. “It is of no use for me to promise more than I can perform.
You had much better look upon me as a decidedly disagreeable
person, with rare moments of proper behavior.”
“Proper behavior, then, means niceness? I’m glad you think it
proper to be nice to me!” said Rudolph. “I perceive that I’ve lighted
upon one of the ‘rare moments,’ and I’m going to take advantage of
it,” he added, as he came a little nearer to her, and looked up in her
face with a glance of amusement and admiration which made her
blush a little. “I’m going to make you talk to me, and amuse me, as
you were told to do last night.”
“No! It was you who were told to amuse me!”
“Was it? Well, we’ll take it in turn then. Do you remember how I
taught you cricket?”
“Yes, oh yes.”
“And what a rage you used to be in when you were caught out?”
“Yes,” answered Mabin, “I remember; but I don’t want to talk about
cricket. I want to tell you something. Mrs. Dale has a fancy that ‘The
Towers’ is haunted.”
And she related the adventure of the previous night, and her
intention of changing rooms with her hostess.
Rudolph listened gravely, and there was a pause when she had
finished before he made any comment. Then he said abruptly:
“You are not nervous, are you, Mabin? I know you used to have no
end of pluck.”
“Well, I haven’t any less than I ever had.”
“Well, if you do change rooms, you have got to be prepared to see
the ghost yourself.”
“You make me feel rather—rather creepy! What do you really think
I shall see?”
“A face at the window probably. The face of the spy from your
house. What else can she have seen?”
Mabin considered a moment.
“I’ll risk it!” she cried at last. “I shan’t go to bed at all. I shall sit up
and watch.”
“I wish you would. We should find out something if you had the
strength of mind to do that.”
Not without a wild beating of the heart, Mabin undertook the task
of holding the strange night-watch, without saying a word to Mrs.
Dale of her intention.
“She thinks she only had a dreadful dream, you know,” said Mabin.
“Well,” replied Rudolph, “I want to know what sort of dream you
will have.”
He had to admire the courage she showed in undertaking a task
which was, as she expressed it, “rather shuddery,” but when he left
her at the gate of “The Towers” she was still steadfast in her
intention.
It was not until after dinner that evening that Mrs. Dale introduced
the young girl to the apartment she was to occupy that night. Mabin
was astonished at its dinginess, its gloominess, contrasting so
strongly as they did with the fresh prettiness of the room which had
been prepared for herself.
It was a large square room, with a mouldy old-fashioned wall-
paper, on which unnatural pink roses climbed up a succession of thin
hop poles. The pictures were groups of trees, done with the pencil in
the woolly early Victorian manner, and stiff bouquets, in water-color,
of conventional early Victorian flowers. The bed, which was hung
with green curtains, occupied an undue space; and Mabin felt that, in
the weird circumstances of her tenancy of the room, she would have
died rather than sleep in that funereal erection.
When Mrs. Dale had kissed her and bade her good-night, after
receiving Mabin’s assurance that she did not feel in the least
nervous, the young girl felt a strong inclination to follow her friend out
of the room, and to implore her to find her some other sleeping-
place.
By a valiant effort, however, she conquered this weakness, and
made a careful survey of her surroundings. In the first place, the
windows and their fastenings had to be examined. They were
carefully secured, and were both so high above the ground that it
would have been impossible for an intruder to reach them without a
ladder.
There were three doors; and at first Mabin was inclined to regard
this as a disquieting circumstance. But on finding that two of them
were unused, locked, and without a key, and that there was a bolt on
the door by which she had entered, she began to feel more at ease.
Exchanging her frock for a dressing-gown, and providing herself
with a book, she placed herself in an arm-chair which stood near the
fireplace, which, although shabby, was sufficiently comfortable, and,
putting her candles on a small table beside her, settled herself to
read. Her book was a novel of an excellent type, not too clever to be
charming, not so commonplace as to be dull. Much to her own
surprise, she got interested, and forgot, or almost forgot, the vague
fears which kept her in the arm-chair instead of in bed.
She was in the very heart of the book, and her candles had burnt
low in their sockets, when a sound, a very slight sound, behind her
back, caused the blood almost to freeze in her veins.
It was a soft, stealthy tread.
Looking round, half paralyzed with terror, she saw that the door
was ajar, and that creeping softly round toward the inside handle
was the long, thin hand of a man.
CHAPTER IX.
A PICTURE.
In the ordinary course of things, it would have been natural for

Mabin to conclude, on seeing a man’s hand inside her door in the
middle of the night, that the intruder was a burglar. But her mind had
been rendered more clear, her perceptions more acute, by the
stimulating mystery which she had been for the past two days trying
to solve.
Instead, therefore, of screaming, or stretching out her hand to the
old-fashioned bell-rope which hung by the fireside at a little distance
from her right hand, she waited, watched, and listened. Apparently
she had unconsciously made some slight noise as she turned in her
chair to look behind her, for the intruder, whoever he was, instead of
entering, waited and listened also.
There was a pause; and then the hand which had crept so
stealthily round the door was slowly and quietly withdrawn. Mabin,
fascinated, watched the long, bloodless fingers as they gradually
disappeared from her sight; and was sufficiently self-possessed to
observe that the hand was that of a gentleman. And upon this
discovery there sprang up in her mind a strong curiosity to see the
face of the intruder.
Even while she felt the last remains of fear give place to courage
and vivid interest, Mabin, with her wits all sharpened with
excitement, wondered at the change in herself. She sprang lightly to
her feet, and with the intention of taking him by surprise, ran lightly
round on the tips of her toes toward the door. But the candles,
flickering in the draught caused by her light hanging sleeves, caused
the shadows on the dingy rose-covered wall to dance and quiver.
The mysterious visitor, as much on the alert as the girl was, closed
the door softly between her and himself.
Mabin, however, sprang forward and seized the door-handle. She
heard the sound of rapid footsteps on the other side, and for one
moment she hesitated to go in pursuit. With the clearness of intellect

Textbook Auxiliary Polynomials in Number Theory 1St Edition Masser Ebook All Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Textbook Auxiliary Polynomials in Number Theory 1St Edition Masser Ebook All Chapter PDF

Uploaded by

Copyright:

Available Formats

Auxiliary polynomials in number theory

1st Edition Masser

Number Theory 1st Edition Róbert Freud

Invitation to Number Theory 1st Edition Ore

Statistical independence in probability analysis number

Quaternion Algebras in Number Theory Spring 2016

Invitation to Number Theory 2nd Edition Oystein Ore

Number Theory Revealed An Introduction Andrew Granville

Research Directions in Number Theory Women in Numbers

Anomalies in Net Present Value, Returns and

B. BOLLOB ÁS, W. FULTON, A. KATOK, F. KIRWAN,

207 Auxiliary Polynomials in Number Theory

A complete list of books in the series can be found at www.cambridge.org/mathematics.

Cambridge University Press is part of the University of Cambridge.

Introduction page vii

20 Elliptic functions 243

is irrational. This is somewhat related to theta functions. Although it converges

In Chapter 4 we prove that certain diophantine equations in two variables

problem too, and obtained the notorious value

Then in Chapter 15 we prove Bilu’s Theorem on the distribution of the con-

In Chapter 18 we use some of the ideas developed so far to give a general-

are in Z/n. This vaguely resembles the reformulation of Pólya’s Theorem.

J(q) = q−1 + 744 + 196884q + 21493760q2 + 864299970q3

isogeny estimates. Using an auxiliary polynomial directly, in the functions

elaborate. In Chapter 22 the proof of Proposition 22.5 is a small variation of

There is also a bibliography, but this has no pretence of being comprehen-

In this chapter we give a couple of examples where the method of auxiliary

(X 111 − X 14 )F = X 11 (X 100 − X 3 )F = X 11 (X 200 − X 6 ) = X 211 − X 17 .

At first sight it appears to be a difficult problem about primes, possibly in

for unknown L and undetermined coefficients pi (not necessarily primes, but

L + 1001 − π(L + 1000)

L + 1001 − π(L + 1000) homogeneous linear equations in the L + 1 unknowns

whose size is the number n of unknowns. Such determinants already have

1.9 Let F be in Fp [X]. Show that there is P = 0 in Fp [X] such that G = PF

+3538944(X 2 + Y 2 ) − 2752512XY − 2415919104(X + Y) + 549755813888.

H = t30 + u30 + 522(t25 u5 − t5 u25 ) − 10005(t20 u10 + t10 u20 ).

1.18 Show that there is P = 0 in Q[X] for Exercise 1.1.

The main application of the method of auxiliary polynomials is in diophantine

It is proved in any elementary text on number theory that e is irrational;

for n = 0, 1, 2, . . .. The first term on the extreme right-hand side dominates

Thus 0 < fn < 2/(n + 1)! and

The proof is slightly easier for the alternating series

because we no longer need the dominance.

fn = fn+1 = fn+2 = 0 (2.2)

is impossible for any n. Namely,

Thus (2.2) would imply s = t = 0 so r = 0 too, a contradiction.

(see Exercise 2.4).

“What was that?” asked Mabin with a shiver.

Mabin was taken so thoroughly by surprise, on seeing the wild

In the ordinary course of things, it would have been natural for

You might also like