Mathematics and Computer Science III Algorithms

Trees Combinatorics and Probabilities Trends in

Mathematics 1st Edition by Michael Drmota,
Philippe Flajolet, Danièle Gardy, Bernhard
Gittenberger ISBN 3034896204 9783034896207
https://ebookball.com/product/mathematics-and-computer-science-
download
iii-algorithms-trees-combinatorics-and-probabilities-trends-in-
mathematics-1st-edition-by-michael-drmota-philippe-flajolet-
dania-le-gardy-bernhard-gittenberger-isbn-30/

Instantly Access and Download Textbook at https://ebookball.com

 Get Your Digital Files Instantly: PDF, ePub, MOBI and More
Quick Digital Downloads: PDF, ePub, MOBI and Other Formats

Computer Science with MATHEMATICA Theory and Practice for Science

Mathematics and Engineering 1st Edition by Roman Maeder ISBN
1595720596 9780521663953

https://ebookball.com/product/computer-science-with-mathematica-
theory-and-practice-for-science-mathematics-and-engineering-1st-
edition-by-roman-maeder-isbn-1595720596-9780521663953-19788/

Discrete Mathematics and Its Applications With Combinatorics and Graph

Theory 7th Edition by Kenneth Rosen 0070681880 9780070681880

https://ebookball.com/product/discrete-mathematics-and-its-
applications-with-combinatorics-and-graph-theory-7th-edition-by-
kenneth-rosen-0070681880-9780070681880-17214/

An Introduction to the Analysis of Algorithms 2nd Edition by Robert

Sedgewick, Philippe Flajolet 9780321905758

https://ebookball.com/product/an-introduction-to-the-analysis-of-
algorithms-2nd-edition-by-robert-sedgewick-philippe-
flajolet-9780321905758-15270/

Operations Research Engineering and Cyber Security Trends in Applied

Mathematics and Technology 1st Edition by Nicholas Daras, Themistocles
Rassias ISBN 3319846671 9783319846675

https://ebookball.com/product/operations-research-engineering-
and-cyber-security-trends-in-applied-mathematics-and-
technology-1st-edition-by-nicholas-daras-themistocles-rassias-
isbn-3319846671-9783319846675-17006/
Mathematics and Statistics for Financial Risk Management 1st Edition
by Michael Miller ISBN 1118170628 9781118170625

https://ebookball.com/product/mathematics-and-statistics-for-
financial-risk-management-1st-edition-by-michael-miller-
isbn-1118170628-9781118170625-24836/

Undergraduate Texts in Mathematics 1st Edition by LarrySmith ISBN

https://ebookball.com/product/undergraduate-texts-in-
mathematics-1st-edition-by-larrysmith-isbn-16404/

Fete of Combinatorics and Computer Science 1st Edition by Gyula OH

Katona, Alexander Schrijver, Tamas Szonyi ISBN 3642135803
9783642135804

https://ebookball.com/product/fete-of-combinatorics-and-computer-
science-1st-edition-by-gyula-oh-katona-alexander-schrijver-tamas-
szonyi-isbn-3642135803-9783642135804-9650/

Handbook of Design Research Methods in Education Innovations in

Science Technology Engineering and Mathematics Learning and Teaching
1st Edition by Anthony Kelly, Richard Lesh, John Baek ISBN 0805860592
9780805860597
https://ebookball.com/product/handbook-of-design-research-
methods-in-education-innovations-in-science-technology-
engineering-and-mathematics-learning-and-teaching-1st-edition-by-
anthony-kelly-richard-lesh-john-baek-isbn-0805860592/

(Ebook PDF) Handbook of Design Research Methods in Education

Innovations in Science Technology Engineering and Mathematics Learning
and Teaching 1st edition by Anthony Kelly, Richard Lesh, John Baek
9781315759593 full chapters
https://ebookball.com/product/ebook-pdf-handbook-of-design-
research-methods-in-education-innovations-in-science-technology-
engineering-and-mathematics-learning-and-teaching-1st-edition-by-
anthony-kelly-richard-lesh-john-baek-97813/
Mathematics and Computer Science III Algorithms, Trees,
Combinatorics and Probabilities Michael Drmota Philippe Flajolet
Daniele Gardy Bernhard Gittenberger Editors P dip p id Birkhauser
Verlag Basel • Boston • Berlin

TRENDS IN MATHEMATICS Trends in Mathematics is a series

devoted to the publication of volumes arising from con- conferences
and lecture series focusing on a particular topic from any area of
mathematics. Its aim is to make current developments available to
the community as rapidly as possible without compromise to quality
and to archive these for reference. Proposals for volumes can be
sent to the Mathematics Editor at either Birkhauser Verlag P.O. Box
133 CH-4010 Basel Switzerland or Birkhauser Boston Inc. 675
Massachusetts Avenue Cambridge, MA 02139 USA Material
submitted for publication must be screened and prepared as follows:
All contributions should undergo a reviewing process similar to that
carried out by journals and be checked for correct use of language
which, as a rule, is English. Articles without proofs, or which do not
contain any significantly new results, should be rejected. High
quality survey papers, however, are welcome. We expect the
organizers to deliver manuscripts in a form that is essentially ready
for direct reproduction. Any version of TEX is acceptable, but the
entire collection of files must be in one particular dialect of TEX and
unified according to simple instructions available from Birk-
Birkhauser. Furthermore, in order to guarantee the timely
appearance of the proceedings it is essential that the final version of
the entire material be submitted no later than one year after the
con- conference. The total number of pages should not exceed 350.
The first-mentioned author of each article will receive 25 free
offprints. To the participants of the congress the book will be offered
at a special rate.

Foreword These are the Proceedings of the International Colloquium

of Mathematics and Computer Science held at the Vienna University
of Technology, September 13-17, 2004. This colloquium is the third
one in a now regularly established series following the first two
venues in September 2000 and September 2002 in Ver- Versailles.
The present issue is centered around Combinatorics and Random
Struc- Structures, Graph Theory, Analysis of Algorithms, Trees,
Probability, Combinatorial Stochastic Processes, and Applications. It
contains invited papers, contributed papers (lectures) and short
communications (posters). The contributions have been carefully
reviewed for their scientific quality and originality by the Scientific
Committee chaired by Michael Drmota (Vienna Uni- University of
Technology, Austria) and composed of Brigitte Chauvin (Universite
de Versailles, Prance), Luc Devroye (McGill University, Canada),
Daniele Gardy (Uni- (Universite de Versailles, Prance), Philippe
Flajolet (INRIA Rocquencourt, Prance), Michal Karonski (Adam
Mickiewicz University, Poland), Abdelkader Mokkadem (Universite de
Versailles, Prance), Helmut Prodinger (University of Witwatersrand,
South Africa), J. Michael Steele (University of Pennsylvania,
Philadelphia, USA), Brigitte Vallee (Universite de Caen, Prance). We
thank them and all anonymous referees for their impressive work.
We also thank the invited speakers: Jean Bertoin (Universite Paris
VI, Prance), Mireille Bousquet-Melou (Universite Bordeaux 1,
Prance), Hsien-Kuei Hwang (Academia Sinica, Taiwan), Svante
Janson (Uppsala University, Sweden), Christian Krattenthaler
(Universite Lyon, Prance), Jean-Prangois Marckert (Uni- (Universite
de Versailles, Prance), Boris Pittel (The Ohio State University, USA),
Si- Simon Tavare (University of Southern California, USA), the
authors of submitted papers and posters, and the participants for
their contribution to the success of the conference. Finally, we
express our acknowledgements to the Institute of Discrete Math-
Mathematics and Geometry, the Vienna University of Technology, the
Federal Ministry for Education, Science, and Culture, the City of
Vienna, the Austrian Research Society (OFG), the Austrian
Mathematical Society (OMG), the Goedel-Society, and the Bank
Austria-Creditanstalt for providing generous financial and material
support. The Organizing Committee Bernhard Gittenberger Thomas
Klausner Alois Panholzer
Editors: Michael Drmota Vienna University of Technology Institute of
Discrete Mathematics and Geometry Wiedner Hauptstrasse 8-10
1040Wien Austria e-mail: [email protected] Daniele
Gardy Universite de Versailles-St-Quentin PRISM Batiment Descartes
45 avenue des Etats-Unis 78035 Versailles Cedex France e-mail:
[email protected] Philippe Flajolet INRIA Rocquencourt 78153 Le
Chesnay France e-mail: [email protected] Bernhard
Gittenberger Vienna University of Technology Institute of Discrete
Mathematics and Geometry Wiedner Hauptstrasse 8-10 1040Wien
Austria e-mail: [email protected] 2000 Mathematical
Subject Classification 05-XX, 60C05, 60Gxx, 68P30, 68Q25, 68Rxx,
68W20, 68W40, 90B15 A CIP catalogue record for this book is
available from the Library of Congress, Washington D.C., USA
Bibliographic information published by Die Deutsche Bibliothek Die
Deutsche Bibliothek lists this publication in the Deutsche
Nationalbibliografie; detailed bibliographic data is available in the
Internet at http://dnb.ddb.de ISBN 3-7643-7128-5 Birkhauser
Verlag, Basel - Boston - Berlin The logo on the cover is a binary
search tree in which the directions of child nodes alternate between
horizontal and vertical, and the edge lengths decrease as 1 over the
square root of 2. The tree is a Weyl tree, which means that it is a
binary search tree constructed from a Weyl sequence, i.e., a
sequence (na) mod 1, n = 1,2 where a is an irrational real number.
The PostScript drawing was generated by Michel Dekking and Peter
van der Wai from the Technical University of Delft. This work is
subject to copyright. All rights are reserved, whether the whole or
part of the material is concerned, specifically the rights of
translation, reprinting, re-use of illustrations, recitation,
broadcasting, reproduction on microfilms or in other ways, and
storage in data banks. For any kind of use permission of the
copyright owner must be obtained. © 2004 Birkhauser Verlag, P.O.
Box 133, CH-4010 Basel, Switzerland Part of Springer
Science+Business Media Printed on acid-free paper produced from
chlorine-free pulp. TCF °° Printed in Germany ISBN 3-7643-7128-5
987654321 www.birkhauser.ch
Preface These colloquium proceedings address problems at the
interface between Mathematics and Computer Science, with special
emphasis on discrete probabilis- probabilistic models and their
relation to algorithms. Combinatorial and probabilistic prop-
properties of random graphs random trees, combinatorial stochastic
processes (such as random walks) as well as branching processes
and related topics in probability are central. Applications are to be
found in the analysis of algorithms and data structures, the major
application field, but also in statistical theory, information theory, and
mathematical logic. This colloquium is the third one in a now
regularly established series, following the first two venues in
September 2000 and September 2002 in Versailles. The book
features a collection of original refereed contributions: contributed
papers (lectures) and short communications (posters), supplemented
by more detailed articles written by invited speakers (and
coauthors): Jean Bertoin (and Christina Goldschmidt), Svante
Janson, and Boris Pittel (and Alan Frieze). During the final
preparation of this volume we received the sad news that Rainer
Kemp (from Frankfurt, Germany) has passed away. Rainer Kemp was
one of the founding fathers of the Analysis of Algorithms, a main
topic of our conference. His book Fundamentals of the average case
analysis of particular algorithms, Wiley, 1984, was one of the first
books on this subject and had a considerable influence on the
development of the field. He was organizer of several meetings on
this subject and served the scientific community with many other
duties. But first of all we lost a good friend and colleague.
Combinatorics and Random Structures. The starting point of many
studies of random discrete models is combinatorics, which often
provides us with exact representations in terms of counting
generating functions that can also be used for a probabilistic study.
Sylvie Corteel, Guy Louchard, and Robin Pemantle work on the
common distribution of intervals in pairs of permutations. Next Sylvie
Cor- Corteel, Jeremy Lovejoy, and Ae Ja Yee provide generating
functions for generalized Frobenius partitions. Luca Ferrari, Renzo
Pinzani, and Simone Rinaldi present some results on integer
partitions. Toufik Mansour considers generating functions for 321-
avoiding permutations. Eugenijus Manstavicius proves an iterated
loga- logarithm law for the cycle lengths of a random permutation.
Martin Rubey provides a sufficient condition for transcendence of
geaerating functions of walks on the slit plane. Michael Schlosser
finds some curious g-series expansions, and Klaus Simon presents
relations between the numbers of partitions and the divisor
functions. Graph Theory. Graphs are a basic object in discrete
mathematics. They are widely used in applications, and algorithms
on graphs as well as theoretical ques- questions on graphs have
been "modern topics" of research in mathematics and com-
computer science for several decades. Mindaugas Bloznelis uses
Hoeffding decompo- decomposition to prove asymptotic normality of
subgraph count statistics. Robert Cori, Arnaud Dartois, and
Dominique Rossin compute so-called "avalanche polynomi-
polynomials" for certain families of graphs. The invited paper by Alan
Frieze and Boris Pittel gives a detailed analysis on perfect matchings
in random graphs with pre- prescribed minimal degree. Omer
Gimenez and Marc Noy provide very tight estimates

viii Preface for the growth constant of labelled planar graphs. Finally,
Stavros D. Nikolopou- los and Charis Papadopoulos present an
algorithm for determining the number of spanning trees in P4-
reducible graphs. Analysis of Algorithms. This field was created by
Donald E. Knuth and is concerned with accurate estimates of
complexity parameters of algorithms and aims at predicting the
behavior of a given algorithm. Javiera Barrera and Chris- Christian
Paroissin consider specific search cost in random binary search trees.
Monia Bellalouna, Salma Souissi, and Bernard Ycart analyze
probabilistic bin packing problems. Pawel Hitczenko, Jeremy
Johnson, and Hung-Jen Huang consider al- algorithms for computing
the Walsh-Hadamard transform. Tamur Ali Khan and Ralph Neininger
analyze the performance of the randomized algorithm to evalu-
evaluate Boolean decision trees proposed by Smir, in particular they
consider the worst case input and provide limit laws and tail
estimated. Next, Shuji Kijima and To- momi Matsui propose a
polynomial time perfect sampling algorithm for two-rowed
contingency tables. Conrado Martinez and Xavier Molinero combine
two genera- generation algorithms to obtain a new efficient
algorithm for the generation of unlabelled cycles. Finally, Yuriy A.
Reznik and Anatoly V. Anisimov suggest the use of tries for universal
data compression. Trees. Trees are perhaps the most important
structure in computer sci- science. They appear as data structures
and are used in various algorithms such as data compression. David
Auber, Jean-Philippe Domenger, Maylis Delest, Philippe Duchon, and
Jean-Marc Fedou present an extension of Strahler numbers to rooted
plane trees. Julien Fayolle analyzes mean size and external path
length of a suffix tree that is related to the LZ'77 data compression
algorithm. Eric Fekete considers two different kinds of external
nodes in binary search trees and describes the evo- evolution of this
process in terms of martingales. The invited paper by Svante Janson
offers an analysis of the number of records in a complete binary tree
or equiva- lently the number of random cutting to eliminate a
complete binary tree. Interest- Interestingly the distribution is, after
normalization, asymptotically a periodic function in log n — log log n,
where n is the size of the tree. Mehri Javanian and Mohammad Q.
Vahidi-Asl consider multidimensional interval trees. Anne Micheli and
Dominique Rossin describe a specific distance between unlabelled
ordered trees, that is based on deletions and insertions of edges.
Katherine Morris determines grand averages on some parameters in
monotonically labelled tree structures. Tatiana Myllari proves local
central limit theorems for the number of vertices of a given
outdegree in a Galton-Watson forest. And finally, Alois Panholzer
gives a precise analysis of the cost distribution for destroying
recursive trees in the case of toll functions of polynomial growth.
Probability. Probabilistic methods get more and more important is
the analy- analysis of discrete structures: random graphs, random
trees, average case analysis of al- algorithms etc. Margaret Archibald
addresses the question of the probability that the maximum in a
geometrically distributed sample occurs in the first d positions of a
word. The invited paper by Jean Bertoin and Christina Goldschmidt
describes the duality between a fragmentation associated to certain
Dirichlet distributions and a natural random coagulation. This gives
rise to an application to the genealogy of Yule processes. Mykola S.
Bratiychuck considers semi-Markov walks in queue- ing and risk
theory. Amke Caliebe characterizes fixed points of linear stochastic
fixed point equations as mixtures of infinitely divisible distributions.
Peter Jagers and Uwe Rosier describe a systematic approach to find
solutions of stochastic fixed points involving the maximum. Arnold
Knopfmacher and Helmut Prodinger

Preface ix provide central limit theorems for the number of descents

in samples of geomet- geometric random variables. Alain Rouault
proves a law of large numbers and describes a new large deviation
phenomenon for cascades. Christiane Takacs investigates
partitioning properties of piecewise constant eigenvectors of
matrices describing the mutual positions of points. Vladimir Vatutin
and Elena Dyakonova consider branching processes in random
environment, find asymptotics of the survival prob- probabilities and
prove a Yaglom type limit theorem. Finally, Vladimir Vatutin and
Valentin Topchii study the joint distribution of the number of
individuals at the origin and outside the origin on a continuous time
random walk on the integers. Combinatorial Stochastic Processes.
Random walks are the most prominent representatives of
combinatorial stochastic processes. They play a central role in the
interplay between combinatorics and probability. Enrica Duchi and
Gilles Schaeffer consider a model of particles jumping on a row of
cells with general boundary conditions where the stationary
distribution is not uniform. Guy Fay- olle and Cyril Furtlehner study
stochastic deformations of sample paths of random walks. Johannes
Fehrenbach and Ludger Riischendorf show that a Markov chain that
is naturally defined on the Eulerian orientation of planar graph
converges to uniform distribution. Alexander Gnedin considers
regenerative composition struc- structures. Jean Mairesse and
Frederic Matheus study transient nearest neighbor ran- random
walks on groups with a finite set of generators and compute various
char- characteristics such as the drift and the entropy. Finally,
Philippe Marchal gives a fractal construction of nested, stable
regenerative sets and studies the associated inhomogeneous
fragmentation process. Applications. Random combinatorics interacts
with many other areas of sci- science. Eda Cesaratto and Brigitte
Vallee consider numeration schemes, defined in terms of dynamical
systems and determine the Hausdorff dimension of sets of reals
which obey some constraints on their digits. Adriana Climescu -
Haulica deals with large deviation analysis of space-time Trellis
codes. David Coupier, Agnes Desol- neux, and Bernard Ycart provide
a zero-one law for first order logic on random images. Nadia
Creignou and Herve Daude study threshold phenomena for random
generalized satisfyability problems. Guy Fayolle, Vadim Malyshev,
and Serguei Pirogov introduce new models of energy redistribution
in stochastic chemical ki- kinetics with several molecule types and
energy parameters. Laszlo Gyorfi discusses Chernoff type large
deviations of Hellinger distance on partitions. Nadia Lalam and
Christine Jacob address the problem of estimating the offspring
mean for a general class of size-dependent branching processes.
Malgorzata and Wlodzimierz Moczurad deal with the problem of
decidability of simple brick codes. And finally, Joel Ratsaby
generalizes Sauer's Lemma to finite VC-dimension classes of binary
valued functions. Altogether papers assembled in this volume offer
snapshots of current re- research. At the same time, they illustrate
the numerous ramifications of the the- theory of random discrete
structures throughout mathematics and computer science. Many of
them, in particular invited lectures, include carefully crafted surveys
of their field. We thus hope that the book may serve both as a
reference text and as a smooth introduction to many fascinating
aspects of this melting pot of continuous and discrete mathematics.
Michael Drmota Philippe Flajolet Daniele Gardy Bernhard
Gittenberger

Contents PART I. Combinatorics and Random Structures Common

Intervals of Permutations Sylvie Corteel, Guy Louchard, and Robin
Pemantle 3 Overpartitions and Generating Functions for Generalized
Frobenius Partitions Sylvie Corteel, Jeremy Lovejoy, and Ae Ja Yee
15 Enumerative Results on Integer Partitions Using the ECO Method
Luca Ferrari, Renzo Pinzani, and Simone Rinaldi 25 321-Avoiding
Permutations and Chebyshev Polynomials Toufik Mansour 37 Iterated
Logarithm Laws and the Cycle Lengths of a Random Permutation
Eugenijus Manstavicius 39 Transcendence of Generating Functions of
Walks on the Slit Plane Martin Rubey 49 Some Curious Extensions of
the Classical Beta Integral Evaluation Michael Schlosser 59 Divisor
Functions and Pentagonal Numbers Klaus Simon 69 PART II. Graph
Theory On Combinatorial Hoeflfding Decomposition and Asymptotic
Normality of Subgraph Count Statistics Mindaugas Bloznelis 73
Avalanche Polynomials of Some Families of Graphs Robert Cori,
Arnaud Dartois, and Dominique Rossin 81 Perfect Matchings in
Random Graphs with Prescribed Minimal Degree Alan Frieze and
Boris Pittel 95 Estimating the Growth Constant of Labelled Planar
Graphs Omer Gimenez and Marc Noy 133 The Number of Spanning
Trees in P4-Reducible Graphs Stavros D. Nikolopoulos and Charis
Papadopoulos 141

xii Contents PART III. Analysis of Algorithms On the Stationary

Search Cost for the Move-to-Root Rule with Random Weights Javiera
Barrera and Christian Paroissin 147 Average-Case Analysis for the
Probabilistic Bin Packing Problem Monia Bellalouna, Salma Souissi,
and Bernard Ycart 149 Distribution of WHT Recurrences Pawel
Hitczenko, Jeremy R. Johnson, and Hung-Jen Huang 161
Probabilistic Analysis for Randomized Game Tree Evaluation Tamur
Ali Khan and Ralph Neininger 163 Polynomial Time Perfect Sampling
Algorithm for Two-Rowed Contingency Tables Shuji Kijima and
Tomomi Matsui 175 An Efficient Generic Algorithm for the Generation
of Unlabelled Cycles Conrado Martinez and Xavier Molinero 187
Using Tries for Universal Data Compression Yuriy A. Reznik and
Anatoly V. Anisimov 199 PART IV. Trees New Strahler Numbers for
Rooted Plane Trees David Auber, Jean-Philippe Domenger, Maylis
Delest, Philippe Duchon, and Jean-Marc Fedou 203 An Average-Case
Analysis of Basic Parameters of the Suffix Tree Julien Fayolle 217
Arms and Feet Nodes Level Polynomial hi Binary Search Trees Eric
Fekete 229 Random Records and Cuttings hi Complete Binary Trees
Svante Janson 241 Multidimensional Interval Trees Mehri Javanian
and Mohammad Q. Vahidi-Asl 255 Edit Distance between Unlabelled
Ordered Trees Anne Micheli and Dominique Rossin 257 On
Parameters in Monotonically Labelled Trees Katherine Morris 261

Contents Number of Vertices of a Given Outdegree in a Galton-

Watson Forest Tatiana Myllari 265 Destruction of Recursive Trees
Alois Panholzer 267 PART V. Probability Restrictions on the Position
of the Maximum/Minimum in a Geometrically Distributed Sample
Margaret Archibald 283 Dual Random Fragmentation and
Coagulation and an Application to the Genealogy of Yule Processes
Jean Bertoin and Christina Goldschmidt 295 Semi-Markov Walks in
Queueing and Risk Theory Mykola S. Bratiychuk 309 Representation
of Fixed Points of a Smoothing Transformation Amke Caliebe 311
Stochastic Fixed Points for the Maximum Peter Jagers and Uwe
Rosier 325 The Number of Descents in Samples of Geometric
Random Variables Arnold Knopfmacher and Helmut Prodinger 339
Large Deviations for Cascades and Cascades of Large Deviations
Alain Rouault 351 Partitioning with Piecewise Constant Eigenvectors
Christiane Takacs 363 Yaglom Type Limit Theorem for Branching
Processes in Random Environment Vladimir Vatutin and Elena
Dyakonova 375 Two-Dimensional Limit Theorem for a Critical
Catalytic Branching Random Walk Valentin Topchii and Vladimir
Vatutin 387 PART VI. Combinatorial Stochastic Processes A
Combinatorial Approach to Jumping Particles II: General Boundary
Conditions Enrica Duchi and Gilles Schaeffer 399

xiv Contents Stochastic Deformations of Sample Paths of Random

Walks and Exclusion Models Guy Fayolle and Cyril Furtlehner 415 A
Markov Chain Algorithm for Eulerian Orientations of Planar Triangular
Graphs Johannes Fehrenbach and Ludger Riischendorf 429
Regenerative Composition Structures: Characterisation and
Asymptotics of Block Counts Alexander Gnedin 441 Random Walks
on Groups With a Tree-Like Cayley Graph Jean Mairesse and Frederic
Matheus 445 Nested Regenerative Sets and Their Associated
Fragmentation Process Philippe Marchal 461 PART VII. Applications
Real Numbers with Bounded Digit Averages Eda Cesaratto and
Brigitte Vallee 473 Large Deviation Analysis of Space-Time Trellis
Codes Adriana Climescu-Haulica 491 A Zero-One Law for First-Order
Logic on Random Images David Coupier, Agnes Desolneux, and
Bernard Ycart 495 Coarse and Sharp Transitions for Random
Generalized Satisfyability Problems Nadia Creignou and Herve Daude
507 Stochastic Chemical Kinetics with Energy Parameters Guy
Fayolle, Vadim Malyshev, and Serguei Pirogov 517 Large Deviations
of Hellinger Distance on Partitions Laszlo Gyorfi 531 Estimation of
the Offspring Mean for a General Class of Size-Dependent Branching
Processes. Application to Quantitative Polymerase Chain Reaction
Nadia Lalam and Christine Jacob 539 Decidability of Simple Brick
Codes Malgorzata Moczurad and Wlodzimierz Moczurad 541

Contents xv A Constrained Version of Sauer's Lemma Joel Ratsaby

543 Index 553 Author Index 555

Part I Combinatorics and Random Structures

Trends in Mathematics, © 2004 Birkhauser Verlag Basel/Switzerland

Common Intervals of Permutations Sylvie Corteel, Guy Louchard,
and Robin Pemantle ABSTRACT: An interval of a permutation is a
consecutive substring con- consisting of consecutive symbols. For
example, 4536 is an interval in the permutation 71453682. These
arise in genetic applications. For the applications, it makes sense to
generalise so as to allow gaps of bounded size 5 — 1, both in the
locations and the symbols. For example, 4527 has gaps bounded by
1 (since 3 and 6 are missing) and is therefore a 5-interval of
4***4*5*27**** for 5 = 2. After analysing the distribution of the
number of intervals of a uniform ran- random permutation, we study
the number of 2-intervals. This is exponentially large, but tightly
clustered around its mean. Perhaps surprisingly, the quenched and
an- annealed means are the same. Our analysis is via a multivariate
generating function enumerating pairs of potential 2-intervals by size
and intersection size. 1. Introduction Let [n] denote the set {1,2,...
,n}. We are interested in counting the common intervals of a pair of
permutations. To be precise, if Ga and Gb are two permu-
permutations of [n], we are interested in counting the pairs of
intervals (/, J) for which Ga{I) = Gb{J)- It is equivalent to count
intervals / for which G~^1Ga{I) is also an interval. Accordingly, we
define Definition 1.1. The interval I := [i, i + k — 1] C [n] is called
an interval of the permutation G ifG~1(I) is an interval, that is, if
there is aj such that The proper intervals are those whose lengths
are at least 2 and at most n — 1. Here and throughout, we use
vector notation for permutations rather than cycle notation, so that
(cti, ... ,an) denotes the permutation i <—> at rather than the
permutation consisting of a single n-cycle. Example: Let G be the
permutation C,1,2,4,5). Then the proper intervals of G are[l,2],[4,5],
[l,3]and[l,4]. When G is a random variable, uniformly distributed
over all permutations of [n], let Xk denote the number of of intervals
of length k of G and let X = ^2k -^k denote the number of intervals
of G. Uno et al [18] compute EXk; the following easy proposition is
proved at the end of this section. Proposition 1.2. As n —> 00, the
distribution of X converges to a Poisson with mean 2. The number of
intervals, or runs of a permutation, was studied in the forties by
Kaplansky [12] and Wolfowitz [19, 20] from a statistical point of
view. See also [14]. Recently several algorithms were designed to
efficiently enumerate all common intervals of permutations [10, 18]
and their time complexity is O(n + K) where n is the size of the
permutation and K the number of intervals. These algorithms were

4 Sylvie Corteel, Guy Louchard, and Robin Pemantle designed

because common intervals have several applications. They relate to
the consecutive arrangement problem [7]. Genetic algorithms for
sequencing problems are based on common intervals [13, 15]. In
bioinformatics [4, 5, 9, 10, 11], genomes of prokaryotes can be
modelled as a permutation of genes. A common interval is then a set
of orthologous genes that appear consecutively, possibly in different
orders, in two genomes. Therefore common intervals can be used to
detect groups of genes that are functionally associated [10, 11]. As
the annotation of genomes is not perfect, the notion of consecutivity
in intervals needs to be relaxed. A notion of gene teams was defined
in [6], where a gene team is a maximal set of orthologous genes,
possibly occurring in different orders in the two species, but
separated in each case by gaps that do not exceed a fixed threshold,
5. To study these, we consider a generalisation of intervals, namely
^-intervals (the previous case corresponds to 5 = 1). Definition 1.3.
The set I C [n] is called a ^-interval of [n] of length k if I is a set of
integers {i1,. ¦ ¦ ,ik} with 1 < ir+i — ir < 5 for each I < r < k — 1.
We call I a S-interval of length k of G if both I and G~1(I) are S-
intervals. Proper 6-intervals are again those of cardinality at least 2
and at most n — 1. Example: G = C,1,2,4,5) possesses the 2—
intervals: {1,2},{1,3},{1,2,3},{2,3},{1,2,3,4},{1,3,4},{2,3,4},
{2,4,5},{2,4},{2,3,5},{2,3,4,5},{1,3,4,5},{1,2,4,5},{1,2,3,5} In [6] a
polynomial time enumeration algorithm for gene teams is presented.
Our notion of 5-intervals removes the maximality constraint, whence
the number of these may grow exponentially and it is natural to
enumerate asymptotically rather than enumerating exactly. The main
purpose of this note is to investigate the asymptotic properties of X\.
', where this denotes the number of S—intervals of length A; of a
uniformly chosen random permutation of [n], and of the total
number X^ := Ylk-^k °^ ^-intervals of a random permutation. We
are interested in all S > 1 but in the present manuscript we examine
only the case 6 = 2. To reduce the number of superscripts, we let Y
and Yk denote X^> and X^ respectively. The number X^ of ^-
intervals when S > 1 behaves very differently from X. Whereas X is
0A) as n —> oo, with all the contributions coming from short
intervals, there will typically be many ^-intervals. In fact a thumbnail
computation produces numbers a^ in the unit interval @:2 ~
0.57939) such that for k ~ an and a > otk, the random variable X^
will be typically exponentially large: the number of ^-intervals of [n]
of size k grows exponentially, the probability of G~l of one of these
also being a ^-interval decays exponentially, and the growth
overcomes the decay when a > a^. Seeing that X^ grows
exponentially in n, it is natural to look at the rescaled quantity n~1
logX(<5). In the next section we compute the annealed mean,
namely n~l logEX^). The term "annealed" means that we first take
an expectation over the (uniform) measure on permutations. The
more interesting quantities are the quenched quantities, which refer
to the typical, rather than the mean behaviour of X^K Often one has
a so-called lottery effect, meaning that the mean of a quantity X
comes primarily from an exponentially small number of values that
are expo- exponentially larger than the median value, and that
consequently, ElogX < log EX. For example, when there is a
Gaussian limit law, n-1/2(logX — n/i) —> N@,cr2),

Common intervals of permutations 5 then one will typically have a

lottery effect. Perhaps surprisingly in light of the discussion in
section 4, there is no lottery effect. Our main result, Theorem 4.1
below, is that for 5 =_2, we have E(X^J = O(EX^S)J. This shows
that as n —* oo, the sequence X := X^ /EX^ is tight. The mean of
X^ is computed in the next section, with the remaining sections
devoted to the computation of the second moment. We start in
Section 2 with arguments in the case 5 = 1. 2. Intervals Recall that
Xk denotes the number of of intervals of length k of G and that X =
J2k Xk denotes the number of intervals of G. Uno et al [18]
computed Although this was not explicitly stated in [18], it is not
hard to show Propo- Proposition 1.2. Proof of Proposition 1.2:
Letting X' := Y^Zl ^' we see () so X' —*¦ 0 in probability as n —>
oo. Thus it suffices to show that X2 converges to a Poisson of mean
2. Kaplansky proves this in [12]. A modern approach is to use the
Poisson approximation machinery first developed by Chen and Stein,
and put in an explicit and usable form in [1]. Given k ? [n — 1], let
Ak be the event that G~l{k,k + 1} is an interval. Let Bk = {k - 1, k, k
+ 1} n [n - 1]. Write pk for F(Ak) and pk,i for F(Ak D Ai). Theorem 1
of [2] shows that the total variation distance between X and a
Poisson with mean J2kpk is bounded by the sum of three quantities,
namely Yl Yl Pk'1 k k^jeBk 63 := J2\( k Bounding 61 and b2 by
O(l/n) is straightforward. The same bound on 63 follows from the
identity \E(lAk-Pk\*)\ = 2 sup[IP(# n Ak) -pk?{H)} together with a
coupling argument obtaining fXAk from fx by switching the values of
G~l onk and G(j), and on k + 1 and G(j + 1) for a uniformly chosen
j. ? More generally, one may consider the distribution of Xk. k)-n(n-
l)...(n-k + 2)
6 Sylvie Corteel, Guy Louchard, and Robin Pemantle Set n-1 k=2 We
see that, as n —> oo, the dominant terms of E(X) are given by k =
0A) and k = n — 0A). Indeed, by Stirling, and setting k = an + 1,
we easily derive nda B) which goes to zero exponentially as n goes
to infinity for fixed e. We obtain -,vs n 8 36 228 E(X) ~ 2 + - + — +
_ + ... n n2 n3 As an example of Xk behaviour, let us now turn to X$
and compute E2 -.= E (X|). Combinatorial analysis leads to
asymptotic series expressions for the moments of Xk and for the
probabilities P(X& = j)- For example, n n* n nz and n . n2 n n2 " " '
n" Details are given in [8]. We turn now to 2-intervals. 3. The mean
number of 2-intervals Let AT (A;, n) denote the number of 2-
intervals that are subsets of [n] and have cardi- cardinality A;. We
will take advantage of the uniformity of G~l. For each of the N(k,n)
2-intervals of cardinality A;, its inverse image under G is uniformly
distributed on A>subsets of [n]. Therefore, the probability is exactly
Ar(A;,n)/(^) for any given 2-interval of cardinality A;, that its inverse
image under G is again a 2-interval. Thus N(k,nf = 77a• \k) To
evaluate N(k,n), note that there may be anywhere from 0 to min{A;
— l,n — k} "holes" in a 2-interval, where a hole is an element not in
the 2-interval but between its endpoints. These may be enumerated
by the following procedure. Pick a starting position r with l<r<n — k
— i + 1 and let r be the least element of the 2-interval. Choose any
sequence with i occurrences of the word "skip" and k - 1 - i
occurrences of the word "no-skip". If the first word in the sequence
is "no-skip" then r + 1 is the next element of the 2-interval; if the
first word is "skip" then r + 2 is the next element. Continue in this
manner until the sequence is used up. This method of enumeration
makes it clear that EYk = 77T\= 7m \k) \k)

Common intervals of permutations 7 The sum may be evaluated

exactly for k/n < 1/2 and asymptotically for k/n > 1/2 leading to fc-
2n/3 N(k,n) k—1 x "|2 : if : if € (-00,00), E) Bfc-l-nJ 1 (k-l Cfc-l-2nJJ
U- fc-2n/3 where ^ is the expectation of the positive part of Z + x
and Z is a standard normal. Via equation D), these asymptotics for
N(k,n) lead directly to asymptotics for EYfc. To obtain asymptotics
for EY, we then sum over &, using a saddle point approximation. The
only significant terms are near k = a*n, where a* will be determined
shortly but is evidently greater than 2/3. We therefore use E) with n
—> 00 and a := A;/n to obtain 2 N(k,ny 2k- k k ' 2k-n 2>k-2n -1 F)
A(a)n ~1/2 where Ba - 1 y ' - Ca - 2L y 27ra(l - a) ' Fi(a) := 3aloga-
(l-a)log(l-a) - 2Ba - l)logBa - 1). The maximum of Fi occurs in [2/3,1]
where F[(a) vanishes, which occurs when a = a*zz .7840013296
solves 17a4 - 33a3 + 24a2 - 8a + 1 = 0. A saddle point
approximation now gives us n-l n k=2 k=3n/4 2tt -KM =
exp(nFi(a*))< 2.4253 [email protected]). 4. Counting pairs of 2-
intervals We will now use the second moment method to show that
Y/EY is tight as n by establishing: Theorem 4.1. Er2 = 0(ErJ. 00

8 Sylvie Corteel, Guy Louchard, and Robin Pemantle Just as the

mean of Y may be obtained from a saddle point analysis of EYk near
k/n = a*, we expect the second moment of Y to be dominated by
terms EY"fc2 with k near some a**. Because we have seen from
numerical data that the quenched and annealed behaviour are the
same, we expect to find, and do find, that a** = a*. Again we will
take advantage of symmetry. This time, if / and /' are 2- intervals,
we will need to know the cardinality of their intersection before we
can say the probability that G~l{I) and G~l(I ) are both 2-intervals.
We therefore define N{k,k'',n,«) to be the number of pairs of 2-
intervals (/,/') of [n] with |/| = k,\I'\ = k' and \I D I'\ = k. For the
computation of EYk2 we will want to specialise to the case k = k', so
we denote N(k,n,K) := N(k,k,n,K). Our computations will now be
analogous to the computation Fi and its argmax, a*. Specifically,
letting a := - , p ¦= — , p ¦= - , n n n we will find a rate function
rate(o;, /?, p) such that N(k, k',n, k) ~
Ao(a,C,p)n~3/2exp(nrate(o;,/?,p)) G) for all parameter values in a
range containing the dominant contributions to the second moment
of Y. To obtain the analogue of F) for second moments, we need the
rate function for total number of pairs of subsets A and B of [n] with
\A\ = k,\B\ = kf and \A D B\ = k. The total number is exactly f n \ n\
\n,k - n,k' - n,n - k - k' + kJ n\(k - «)!(&' - «)!(n - k - k' + «)! '
whence for any pair of sets of respective cardinalities k and k' whose
intersection has cardinality k < k, the probability that their union is a
specific pair of sets is the reciprocal, P, of this. Let ent denote the
exponential rate limn logP for this probability: ent := plogp+(a-
p)log(a-p) + (/3-p)log(/3-p) + (l-a-P+p)log(l-a-/3+p). (8) We may
now evaluate the rate function F2(a,0,p) := lim N(k,k',n,KJP(k,k',n,K)
(9) = 2 • rate + ent for the expected number of pairs of 2-intervals
of respective sizes k and k' with overlap n. Since Er2 = J2kk, EYkYk>
= Ysk^* ,KN (k,k' ,n, kJ P{k,k' ,n, k) is the sum of polynomially
many summands, it follows that the exponential order of EY2 is the
same as the order of the largest summand, namely -> sup F2(a,P,p)
:= A**. A0) Furthermore, from the inequality EYkYk> < \{EYk2 +
EYk2,), we see that the max- maximum of EYfcYfc' (for fixed n) can
only occur when k' — k, and therefore A** = sup F2(a, a, p). A1)
a,p Next, consider how N(k,k',n,K) varies with k for fixed (k,k',n). In
other words, enumerate pairs of 2-intervals of fixed sizes k and k'
according to the size

Common intervals of permutations 9 of their intersection, k. Observe

that J2K N(k, k',n, k) counts all pairs of 2-intervals of sizes k and k',
so that Y^N(k,k',n,K) = N(k,n)N(K,n). K Since the number of
summands is linear in n, we have at the exponential level that sup
rate(a:, /5, p) = u(a)u{C) A2) p where u(a) = \\mn~lN(k, n) as n —
> oo with k/n —> a. We will show: Lemma 4.2. For a and 0 in a
neighbourhood ofa**, this supremum occurs at p = a- f3. Observe
that -ent(a,/?, •) has a maximum at a ¦ C as well, where it is given
by the expression ent(a,0,a-0) = h(a) + h(P). A3) Here, h(x) —
a;log(a;) + A - a;)log(l - a;) is the usual entropy function. Both rate
and ent are smooth, so we see that the function F2 = 2 • rate + ent,
viewed as a function of the third argument, has a critical point at p
= a ¦ C as well. The next lemma, requiring a four-variable generating
function and considerable computation, is: Lemma 4.3. The critical
point (a,/5, a ¦ /3) is in fact a maximum for F2. Corollary 4.4. Proof.
It follows from A2) and A3) that sup F2(a,0,p) = F2(a,0,a-0) = 2 •
rate(a,/?, a-/?) + ent(a,/?,<*•/?) = 2u(a) + 2u(P) + h(a) Taking the
maximum over a and C then gives the result. ? Theorem 4.1 now
follows directly from Corollary 4.4 and: Lemma 4.5. 1. There is a
neighbourhood of (a*, a*, p) for which N(k, k', n, KJP(k, &', n,«) ~
A(a,/5, p)n~3/2 exp(nF2(a,/5, p)) uniformly as n —> 00 with (k/n,
k'/n, K/n) —> (a, /5, p). 2. The Hessian of F2 here is non-
degenerate, whence , k', n, nJP(k, k', n, k) ~ ^^ We will prove
Lemmas 4.2, 4.3 and 4.5 in the next section by means of the
aforementioned four-variable generating function.

10 Sylvie Corteel, Guy Louchard, and Robin Pemantle 5. The

generating function Recall that N(k,k',K,n) denotes the number of
pairs of 2-intervals (/,/') of [n] with |/| = k, \I'\ = k! and |7 n I'\ = k.
Define the generating function F(ui,v,2,t,z):= k,k' ,K,n so that
N(k,K,n) = [u$u%tKzn]F. Given a pair of subsets of [n], the positions
1 through n may be divided into three parts: 1. An initial sequence
of positions before the first common position; 2. A common position
followed by zero or more segments of the form: a sequence of
positions not common to either set, in such a way that no two
positions in a row are absent from either set, followed by a common
position; 3. A final sequence of positions after the last common
position. We give generating functions for the three parts separately.
To enumerate the second of the three parts, note that each segment
between common positions can be one of six possible classes of
configuration. We list these here, along with the factor contributed
by such a step to the generating function. (a) Empty. fa = 1. (b) A
single position, which can belong to either set or neither, but not
both. /& = z{\ + ui + u2). (c) A positive number of pairs (j,j + 1)
where j G / \ /' and j + 1 G /' \ /. fc = z2uiu2/(l - z2uiu2). (c') A
positive number of pairs (j,j + 1) where j ? I'\I and j + 1 G / \ /'. fc>
= fc- (d) the same as (c) but there is a single position in I \I' at the
end. fd = zuifc- (d') the same as (c') but there is a single position in
I' \I at the end. fd' = zu2fc' • The generating function for an
arbitrary sequence of these is f = ?!^ A4) 1 - ZUlU2t{fa + fb + B +
Z(U1 + U2)fc) ' V ^ To enumerate the first (or last) of the three
parts, we first write down the generating function F6(u,z) := —-—r—
1 - A + z)uz for that part of a 2-interval between its first and last
point. By symmetry, parts 1 and 3 have the same generating
function, which is equal to 1/A — z) times the generating function g
for the segment to the right of the last common position but to the
left of the last position in /U /'. We may write g as the sum of
several cases. (e) Empty. ge = 1. (/) A position in neither set,
followed by a non-empty string of positions, each of which is in
neither set or /, with no two in a row not in /. g/ = zFe(ui,z). (/') A
position in neither set, followed by a non-empty string of positions,
each of which is in neither set or /', with no two in a row not in /'.
gj> = zF§(u2, z). (g) A string of pairs as in case (c) above, followed
by a nonempty sequence as in case (i). gg = F6(ui,z)/A-z2uiu2). (g')
A string of pairs as in case (c') above, followed by a nonempty
sequence as in case (f). gg, = FG(u2,z)/(l - z2uiu2). (h) The same as
(g) except with a position in /' \ / in the beginning. gh = zu2F(;
(ui,z)/(l - z2u\u2).

Common intervals of permutations 11 (h!) The same as (g') except

with a position in / \ /' in the beginning. gh> = zuiF6(u2,z)/(l -
z2uiu2)- Summing, we see that the factor from the first and last
parts is 9 = z 1 + zF6(uu z) + zF6{u2, z) + '- 1 - Z \ 1 - ZZU\U2 and
finally, 9 - ZlU\U2 F(Ul,u2,t,z)=fg2. A5) 6. Proofs of lemmas There is
only room for sketches here. Details are given in [8]. Let V be the
variety where the rational function F blows up, namely the union of
the varieties where the denominators of / and g vanish. It simplifies
the computation to change variables to r := ztu\u2, u — zu\ and v
— zu2. The [k,k\ n, k] coefficient of F now becomes the [k — k, k' —
k, n — k — k' + k, k] coefficient of the function F(x, y, z, r). Thus 1 1
~ — logfna, n/3,n, np]F = — log[n^, riv, nS, np]F n n when a = fj,
+ p, /3 = is + p, p = a + C + S — 1 and F is F under the change of
variables. In the new variables, j r{l-xy) 1 — xy — tA + x + y + xy +
zxy — z) We let h denote the denominator of this expression. Then V
is the union of the variety Vh where h vanishes, together with the
varieties where the functions g± := 1 - z, g2 := 1 - xy, g3 := 1 - A +
z)x and g4 := 1 - A + z)y vanish. Let Dom denote the domain of
convergence for the power series for rp r:=(k,k',n,K) Let log D
denote the logarithmic domain, which is always convex: logD := {log
|xi|,... ,log \x4\) : x G Dom} . Asymptotics of ar in the direction r/|r|
-» s are "controlled" by corresponding points x(s) e V as described in
the following paragraph (see [8] for details). First, it is shown in [16,
Theorem 6.3] that for each vector s ? (E+L, there is a point x(s) € V
with positive real coordinates such that x € dDom and such that the
normal to the support hyperplane to logD at logx is parallel to s. It
then follows from [16, Theorem 3.5], provided a certain function has
a non- degenerate Hessian, that if x(s) lies in some smooth
component of V for all s in some neighbourhood K, then there is a
function A such that the coefficients of v = V arxr r: = (k,k',n,K)

12 Sylvie Corteel, Guy Louchard, and Robin Pemantle are uniformly

asymptotically approximated on K by ar ~ A(^-, ^, rf)rCV2 exp(-r •
x(r)). The key computational step necessary to make use of these
results is to verify that the point x(r) is in the part Vh of the pole set
V that corresponds to the vanishing of h. Combinatorially, the factor
/ in F enumerates pairs of subsets for which the first and last
common positions are 1 and n respectively. The other divisors
correspond to the cases where one or both intervals begin at G(n) or
end at n — @(n). We now give an argument that the pairs of sets
beginning at 1 and ending at n count "nearly all" pairs, meaning that
the rate n~1 logiV is the same. Let N(k, k', k, n) denote the number
of pairs of 2-intervals in which both the subsets contains 1 and n.
Suppose, for a given (k, kf, k, n) that N(k, k', An, k) > N(k, kf, n, k)
A6) for some fixed A < 1 and arbitrarily large n (this can happen, for
example when ^ = a — p is small). Then N(k,k',\n,KJP(k,k',\n,K) >
N(k,k',n,KJP(k,kf,Xn,K) > N(k,k',n,KJP(k,k',n,K) by an exponential
factor. But supapp F2(a,0,p) > 0, so if this supremum is achieved at
a = k/n, C = k'/n, p = n/n, then the inequality would be reversed.
Thus, in a neighbourhood of where the supremum of F2 is achieved,
it is not possible for A6) to hold. Consequently, in this
neighbourhood the coefficients of / are, on the exponential scale, as
large as those of F, whence the minimal point occurs on Vh- Proof of
Lemma 4.2: Choose a pair of 2-intervals of sizes k and k' uniformly
from pairs where both contain 1 and n. Define the random variable
Zktk',n to be n~1 times the cardinality of the intersection of these
two sets. If Zk,k>,n —> 7 in probability as n —> co with k/n —> a
and k'Jn —> C, then the supremum of rate(a,/3, •) occurs at 7. But
we see from the construction of the generating function that the
indicator functions of the two sets form two independent Markov
chains, both constrained to go from 1 to n in a given number of
steps. This implies asymptotically independent statistics, meaning
that the limit in probability of Zk,k',n is kk'/n. ? Proof of Lemma 4.3:
We argue here only what we need, namely the lemma for P = a and
a in a neighbourhood of a*. We have seen that rate and —ent both
have maxima at (a*,a*,a^), so this is a matter of checking the
second derivative of 2 • rate + ent for negative definiteness and
then checking that this local maximum is in fact a global maximum
for F2. To do this, we evaluate the point x controlling asymptotics of
F2(a, a, p). Using Maple, we solve for x e Vh and (x]dh/dx\, X2

Common intervals of permutations 13 parallel to s for which a =

si/s3,0 = s2/s:i, p = s4fs3. We get 2a-l-R Uq = Vq = — ; z = Ba - 1
- RN 0 462 -66 + 26R + 46a - 6a + 3 + 4a2 - R ' 462 -66 + 26R +
46a - 6a + 3 + 4a2 - R _ T° ~ 26a-6-6R-4a + 4a2 + q + l-2aR '
here, 8 — l — 2a + p and R:= \/8a2 - 12a + 5 + 88a -88 + 462 .
We take two derivatives and evaluate at p = a2, obtaining a rational
function of a. Doubling and adding the second derivative of ent, we
may verify negativity of the second derivative in p for all a > 2/3.
This establishes a local maximum at p = a2. We have already seen
that the global maximum of F2(a,0,p) occurs at 0 = a. It is
straightforward to check numerically that there is no maximum in a
set bounded away from the diagonal 0 = a exceeding the value of
F2(a,a,a2) at a = a*. ? Proof of Lemma 4.5: For this we need to
compute the solution x(s) in all four variables, rather than just when
0 = a. The result is rather messy and may be found in [8], In a
neighbourhood of (a*,a*,a1) we may use rigorous numerical
estimates to verify that the Hessian is non-degenerate. This
establishes the second assertion of the lemma. The first follows from
[16, Theorem 3.5]. ? References [1] R. Arratia, L. Goldstein and L.
Gordon. Two moments suffice for Poisson approxi- approximation:
the Chen-Stein method. Annals of Probability, 17:9-25, 1989. [2] R.
Arratia, L. Goldstein and L. Gordon. Poisson approximation and the
Chen-Stein method. Statistical Science, 5:403-424, 1990. [3] Y.
Baryshnikov and R. Pemantle. Manuscript in preparation. [4] A.
Bergeron and J. Stoye, On the Similarity of Sets of Permutations and
Its Applica- Applications to Genome Comparison, COCOON 2003,
Lecture Notes in Computer Science, 2697: 68-79, B003). [5] A.
Bergeron, S. Heber and J. Stoye, Common intervals and sorting by
reversals: a marriage of necessity. Proceedings of ECCB 2002: 54-63,
B002). [6] A. Bergeron, S. Corteel and M. Raffinot: The Algorithmic
of Gene Teams. WABI 2002, Lecture Notes in Computer Science,
2452: 464-476, B002). [7] K. S. Booth and G. S. Lueker, Testing for
the Consecutive Ones Property, Interval Graphs, and Graph Planarity
Using PQ-Tree Algorithms. J. Comput. Syst. Sci. 13C): 335-379 A976)
[8] S. Corteel, G. Louchard and R. Pemantle. Common intervals in
permutations (unabridged version),
http://www.math.upenn.edu/"pemantle/louchard/version040122.ps
[9] G. Didier, Common Intervals of Two Sequences. WABI 2003,
Lecture Notes in Com- Computer Science, 2812: 17-24 B003).

14 Sylvie Corteel, Guy Louchard, and Robin Pemantle [10] S. Heber

and J. Stoye, Algorithms for Finding Gene Clusters. WABI 2001,
Lecture Notes in Computer Science, 2149: 252-263, B001). [11] S.
Heber and J. Stoye, Finding All Common Intervals of k Permutations.
CPM 2001, Lecture Notes in Computer Science, 2089: 207-218,
B001). [12] I. Kaplansky. The asymptotic distributions of runs of
consecutive elements. Annals of Mathematical Statistics, 16:200-203,
1945. [13] S. Kobayashi, I. Ono and M. Yamamura, An Efficient
Genetic Algorithm for Job Shop Scheduling Problems. ICGA 1995:
506-511, A995). [14] V.K. Kolchin, A.S. Sevastyanov, and P.C.
Chistiakov. Random Allocations. Wiley, 1978. [15] H. Miihlenbein, M.
Gorges-Schleuter, and O. Kramer. Evolution algorithms in com-
combinatorial optimization.Parallel Comput., 7:65-85, A988). [16] R.
Pemantle and M. Wilson Asymptotics of multivariate sequences, part
I: smooth points of the singular variety. J. Comb. Theory, Series A,
97:129-161, 2001. [17] R. Pemantle and M. Wilson Asymptotics of
multivariate sequences, part II: multiple points of the singular
variety. Combinatorics, Probability and Computing, to appear. [18] T.
Uno and M. Yagiura, Fast Algorithms to Enumerate All Common
Intervals of Two Permutations. Algorithmica 26B): 290-309 B000).
[19] J. Wolfowitz. Additive partition functions and a class of
statistical hypotheses. An- Annals of Mathematical Statistics, 13:247-
279, 1942. [20] J. Wolfowitz. Note on runs of consecutive elements.
Annals of Mathematical Statis- Statistics, 15:97-98, 1944. Sylvie
Corteel CNRS PRiSM, Universite de Versailles Saint-Quentin, 45
Avenue des Etats-Unis, 78035 Versailles Prance email:
[email protected] Guy Louchard Universite Libre de Bruxelles,
Departement d'Informatique, CP 212, Boulevard du Triomphe, B-
1050 Bruxelles, Belgium email:[email protected] Robin Pemantle
Department of Mathematics, University of Pennsylvania, 209 S. 33rd
Street, Philadelphia, PA 19104 USA, Supported by NSF Grant # DMS-
0103635 email:[email protected]

Trends in Mathematics, © 2004 Birkhauser Verlag Basel/Switzerland

Overpartitions and Generating Functions for Generalized Frobenius
Partitions Sylvie Corteel, Jeremy Lovejoy, and Ae Ja Yee ABSTRACT:
Generalized Frobenius partitions, or F-partitions, have re- recently
played an important role in several combinatorial investigations of
basic hypergeometric series identities. The goal of this paper is to
use the framework of these investigations to interpret families of
infinite products as generating func- functions for F-partitions. We
employ q-series identities and bijective combinatorics. 1.
Introduction Let Pa,b{u) denote the number of generalized
Frobenius partitions of n, i.e., the number of two-rowed arrays, (
Oi,O2,...,Om \ ^ bub2,...,bm )' (!) in which the top (bottom) row is a
partition from a set A (B), and such that ZXa; + bi) + m = n[2]. The
classical example is the case Po,D(n), where D is the set of partitions
into distinct non-negative parts. Frobenius' observed that these
objects are in one-to-one correspondence with the ordinary
partitions of n, giving oo oo ^ E ^.° ("><?" =1171—-V <2> n=0
n=l V H ' Andrews [2] later made an extensive study of two infinite
families of im- impart it ions that begin with Pd,d(w). He replaced D
by Dk or Cfe, the set of par- partitions where parts repeat at most k
times and the set of partitions into distinct parts with k colors,
respectively. The generating functions are multiple theta series,
which in three known cases can be written as an infinite product. ,
C) n=o Vi pDz,D3{n)q = fo3.o6N2 ' D) n=0 W/ooW >y /oo °° / 2\2
En / \ n \~Q'iQ Joo /k\ ^ca,ca(n)9 = ( . f , • E) n=0 w/ooW) y /oo
Here we have employed the standard notation oo (,a1,...,aj)oo :=
(oi,... .Ojj^oo := JJA -aigfe)---(l - ajqk). F)

16 Sylvie Corteel, Jeremy Lovejoy, and Ae Ja Yee While Andrews'

families subsequently received quite a bit of attention [10, 13, 15,
16, 18, 20], other types of Probenius partitions have recently been
turning up as novel interpretations for some infinite products that
figure prominently in basic hypergeometric series identities [6, 7, 8,
21]. The combinatorial setting here is that of overpartitions, which
are partitions wherein the first occurrence of a part may be
overlined. Let O denote the set of overpartitions into non-negative
parts. Then it turns out that ^ G) m,n>0 W°° and
Po,o(t>rn,n)a<bmqn = {aq\b*)o°. (8) Here PD,o(m, n) denotes the
number of F-partitions counted by Pd,o{^) that have m non-
overlined parts in the bottom row, and Po,o(t,m,n) denotes the
number of objects counted by Pn,o{n) that have I non-overlined
parts in the top row and m non-overlined parts in the bottom row.
Since a thorough combinatorial understanding of G) and (8) has
been so useful, we give in this paper a variety of other infinite
product generating functions for F-partitions and begin to study
them using bijective combinatorics. The first goal is to use restricted
overpartitions and a useful property of the 1^1 summation (see
Lemma 2.2) to embed some of B) - (8) in families of infinite
products that generate F-partitions. Theorem 1.1. Let Ok be the set
of overpartitions where the non-overlined parts occur less than k
times. Let Pofc)ofc(m,n) (resp. Pofc ,0G71,71),) be the number of F-
partitions counted by Pnk,ok{n) (resp. Pnkjn(n)) wherein the
number of overlined parts on the top minus the number of overlined
parts on the bottom is m. Then Pn n (m n)bman - ( D q - Notice that
in both instances the case k —> 00 is the case a = 1/6 of (8), while
the case k = 1 of (9) is Probenius' example B), the case b = 1, k = 2
of (9) is Andrews' E), and the case k — 1 of A0) is G). Our next
object is to exhibit more families like those above, but where the
base cases are none of B) - (8). We use the notation AB for the set
of vector partitions (Xa,^b) E A x B, and Dk for the set of partitions
into non-negative parts where each part occurs 0 or A; times.
Theorem 1.2. E n=0 n — \~q'°° (nk- ak\ (n2k •

Overpartitions and Frobenius partitions 17 Then, by employing more

general ^-series identities, we find generating func- functions with
more parameters, like Theorems 1.3-1.6 below. The first two contain
the k = 1 case of A1) and the case k = 2 of A0), respectively.
Theorem 1.3. Let PD,OD(m,n) be the number of F-partitions counted
by PD,OD{n) that have m parts in Xp . Then m,n>0 (9; 9)oo
Theorem 1.4. Let O2 denote the number of overpartitions in O
where the non- overlined parts repeat an even number of times. Let
POO2D(?,m,n) denote the number ofF-symbols counted by
Po,o2D(n) where ? is the number of non-overlined parts in the top
row minus the number of parts in XD and m is the number of non-
overlined parts in the bottom row. Then ^^^"ff^^. A4) o The next
theorem also contains the case k = 2 of A0). We are concerned here
with D, which denotes the set of overpartitions into distinct parts
such that parts have to differ by at least two if the bigger is
overlined and 0 does not occur. These overpartitions have recently
arisen in a number of works [5, 14, 17]. Theorem 1.5. Let
PoD,o{^im->n) den°te the number of overpartitions counted by
PoD,o(n) such that ? is the number of non-overlined parts in Xo plus
the number of overlined parts in Xq and such that m is the number
of non-overlined parts on the bottom minus the number of parts in
Xq. Then P - (?m n)aebman - (~a9^ The last example contains (8)
and deals with O', the set of overpartitions in O that have no 0.
Theorem 1.6. Let Poo>,o(^^^rn^n) denote the number of F-
partitions counted by Poo',o{n) where k is the number of non-
overlined parts in Xo plus the number of overlined parts in Xo>, ? is
the number of non-overlined parts in the bottom row, and m is the
number of parts in Xo> ¦ Then E WM,m,n)aW=<^^. A6) k,?,m,n>0
Finally, we give bijective proofs for some of the generating functions
above. We are able to establish E), A3), and the case k = 2 of A0) in
this way. 2. Recollections and Proofs Given a set A of partitions we
denote by PA(n,k) the number of partitions of n from the set A
having k parts. We recall from [2] that

18 Sylvie Corteel, Jeremy Lovejoy, and Ae Ja Yee Lemma 2.1. The

generating function for Frobenius partitions is given by z-k, A7) n=0
n,k n,k where [zk]Y^Anzn = Ak. We assume enough familiarity with
the elementary theory of partitions and overpartitions [1, 8] that we
can state generating functions for simple PA.(n,k) without
explanation. The following is the key lemma mentioned in the
introduc- introduction. Lemma 2.2. // ^(-639,-1/63H0^ „,__
(-69,-9/6H then (-6g, -9/6) 00 fc Proof. Let H{q) = [z°]F(z,q)G(z,q),
with F(z,q) = E^i(?)^' and G(z,q) = If Akj(q) = Aj(qk), then
{z°]F(z,q)G(zk,qk) = The proof is finished when we apply the above
observation to Substituting a = 1/6 and z = 62 in the 1^1
summation, °° )n _ (9,069,-39,-1/3) Ey-L/ajnyazq) _ yq, auq, -zq, -
i/^qq . . (—hn}~ (— hn —an. n.zn. hiz\ ' n= —00 ^ we have Then
Vl)oo and the lemma follows. Above we have introduced the
notation

Overpartitions and Probenius partitions 19 Proof of Theorem 1.1. For

the first part, take G(z,q) = (zq, 1/2H0 in Lemma 2 2 By B), H(q) =
(«)oo/(-&9)oo(-9/&)oo. Then -H{qk) (-^)oo(-9/6)oo(9fe;9fe)c
Similarly, take G(z,q) = B9H0 for the second part. ? In the following
we will use the \ipi summation B1) or one of its corollaries for the
first step of each proof. Proof of Theorem 1.2. For the first part, take
G(z,q) = (-1/2,29H0 in the case b = 1 of Lemma 2.2. Then vtt \ (g)L
r oi (-*g> 9H0(-^-1; 9H0(-^-1; 9H0 H{q) = (=^M F1!^ r 0 -959H0
n? (-9;9)oo v^ 9 (9-,9)oo ^0( (9;9)oo(-9;92 by the g-binomial
theorem, For the second part we again apply the case b = 1 of
Lemma 2.2, this time with G(z,q) = (z-1,zq)oo. Then , h2 [z\{zq;
(-9,9)?o 0l V n _ q 00 (9H0 y^ 9 (9H0 (-9)^(9,94;95)oo'

20 Sylvie Corteel, Jeremy Lovejoy, and Ae Ja Yee by the first Rogers-

Ramanuj an identity, Eq 1 Tn\~ = (n n^-n$\ ' D Proof of Theorem
1.3. m,n>0 (9; q)oo = i~^ci)ooi~vq^q2)oo (q; 9H0 the final
equality being the case q = q2,z = —q/a, and a —> 00 of B4). D
Proof of Theorem 1.4. (q), (q, abq)^ ^ (-bq, q)n )(-q/a,-
ab2q2;q2)oo (q)oc(q,a2b2q2;q2)c by the ^-Kummer identity, (a,bM-
q/b)n y ^ (q,aq/b)n Proof of Theorem 1.5. - It m n)J {-l/b)n{b/z)n f,
(-aq)nqn(n+1)/2(z/bT {-aq)n ^ (q)n {q, abq)oo ^ {-aq, -bq)oo{-
q/b; r,q2)oc

Overpartitions and Frobenius partitions 21 the final equality being

Lebesgue's identity which is the case b —* oo and a = —lib of B6). ?
Proof of Theorem 1.6. kAm,n>o {azq,b/z,czq)oo {-l/b)n{b/zY ^ (-
aq)n(czqy ^ [q)n {-\/b)n{bcq)n (-aq, -bq, -cq) {q^abq^bcq)^ by
the ^-binomial theorem B4). ? It should be noted that Theorems 1.4
- 1.6 have /^-generalizations like The- Theorems 1.1 and 1.2 but the
combinatorial definition of the F-partitions are less palatable. 3.
Bijections In this section we establish some partitions that explains
some of the first cases. Bijection for E) We will here give a
combinatorial proof of VP q (n)qn= (~9;9)°°(~^2)^ n=0 \?i?j<ju
We start with a F-partition and add one to each entry of the first
row. Let k be the number of overlined parts in the first row minus
the number of overlined parts of the second row. Suppose without
loss of generality that k > 0. Then we split the F-partition into two F-
partitions : one that contains the overlined parts and one that
contains the non-overlined parts. Apply Wright's bijection (see [20])
to both F-partitions and get two ordinary partitions and two triangles
(k, k — 1,..., 1) and (k — 1,..., 1). We keep the first partition, which
gives l/(q; q)<x> and the odd parts of the second partition, which
gives l/(q; q2)oo- Then we divide the even parts of the second
partition by 2. To the left of these parts we put k, k - 1,..., 1. By
applying the reverse of Wright's bijection, we obtain two partitions
into distinct parts. We multiply by two and decrease by 1 the parts
of the first partition and we multiply by 2 and increase by 1 the parts
of the second partition. We get two partitions into distinct odd parts,
which is generated by (—q;q2J- It is easy to check that the weight is
preserved and that every step is reversible. Bijection for A0). Now let
us prove combinatorially that This proof is inspired by some ideas of
[21]. We start with the Frobenius partition and add one to each
entry of the first row. The top row is an overpartition in
22 Sylvie Corteel, Jeremy Lovejoy, and Ae Ja Yee top not overlined n
parts bottom overlined at least n bottom non overlined top overlined
k parts OOuOO Qverpartition into odd parts Overpartition Figure 1.
Bijection for 10 O2. Therefore the non-over lined parts form a
partition into distinct parts. Let us suppose that this top row has n
non-overlined parts and k overlined parts. We separate these into
partitions a and j3. The bottom row is an overpartition into n + k
non-negative parts. Using algorithm Z [4, 22], we can decompose it
into a partition 8 into parts at most n + k and a partition 7 into non-
negative distinct parts less than n + k. We take the parts of 7 that
are less than n, conjugate them, and add them to a to create an
overpartition into distinct parts (a part is overlined if the difference
with the previous part is at least 2). Then we change a into an
overpartition into odd parts generalizing Sylvester's bijection [19].
First we take off 2[(ra—1)/2J +2-i from the ith part and change it to
77 a B[(n - 1)/2J + 1) x \n/2] rectangle. Then we look at the
conjugate of what is left. Every odd part is inserted in 77 and every
even part 2i is changed to two i parts that are inserted in the
conjugate of 77. The overlines follow. Note that this part is a
bijective proof of Lebesgue's identity. Now we take the parts of 7
that are equal to or greater than n. If the part n+i — 1 occurs we
add it to the ith part of C and take off the over line. After sorting,
that creates an overpartition where the non-overlined parts are
greater than n + k. We add then 5 and get an overpartition, which is
generated by {—q\ q)oo/(q', #)oo- For example, we start with ( ^
6^| *' ^2_' *'°Q'0Q Y We get a = G,5,4,3,1), C = (9,5,2,1), 7 =
G,6,4,1) and 5 = F,4,4). Then we apply the mapping and get 77 =
(fl, 9,5,1,1) and // = A1,9,9,6,4,4,1). See Fig 1. Bijection for A3) We
will here give a combinatorial proof of n=0 We start with a Frobenius
partition and add one to each entry of the first row. The bottom row
is made of an overpartition j3 and a partition into distinct parts a.
Let n be the number of parts of a. We apply a generalization of
Wright's bijection to the top row and E. We draw the Ferrers diagram
of the top row and we shift the ith part by i - 1. We draw the Ferrers
diagram of the non overlined parts of /3 and put it at the left of the
diagram of the top row. We draw the Ferrers diagram of the
overlined parts of /3, conjugate it and put it at the bottom right of
the diagram of the top row, as shown on the left of Figure 2. Then
we break the diagram into two parts: the left and the right of the
largest overlined part. On the right we get an ordinary partition,
which gives l/(q; q)oo and on the left we get a partition A into
distinct parts where all the parts from 1 ton occur.

Overpartitions and Probenius partitions 23 bottom not overlined

bottom overlined distinct all parts from 1 to n occur distinct ordinary
ordinary bottom distinct n parts distinct n parts distinct n odd parts
Figure 2. Bijection for A3) Let i be the index of the smallest even
part in a. Then we take off i from A and add it to the conjugate of a.
We do that until a has only odd parts. We finally get A a partition
into distinct parts, which is generated by (—q; q)^ and a a partition
into distinct odd parts, which is generated by {—q\ q)^. Each step is
easily reversible. A1 10 ft 7 ?1 A *¦? 0 1 \ C4 3 3 2 01 'G 4 2 11 )
4,3,3,3,3,2,2), A = (9,8,4,2) and a = (9,5,3,1). See Fig 2. Bijection
for A5). A combinatorial proof of we can be easily done mixing the
combinatorial proof of the i-0i summation of [21] and a
combinatorial proof of the Lebesgue's identity for example the one
used in the previous bijection. Bijection for A6). A combinatorial
proof of n=0 can be easily done mixing the combinatorial proof of
the i^i summation of [21] and a combinatorial proof of the q-
binomial identity (see for example [12]). Acknowledgments We
would like to thank Bruce Berndt for making [15] available to us.
References [1] G.E. Andrews, The Theory of Partitions, Cambridge
University Press, Cambridge, 1998. [2] G.E. Andrews, Generalized
Probenius partitions, Mem. Amer. Math. Soc. 49 A984), no. 301.

24 Luca Ferrari, Renzo Pinzani, and Simone Rinaldi [3] G.E. Andrews,
g-Series: Their Development and Application in Analysis, Number
Theory, Combinatorics, Physics, and Computer Algebra, CBMS 66,
American Math- Mathematical Society, Providence, 1986. [4] G.E.
Andrews and D.M. Bressoud, Identities in combinatorics, III: Further
aspects of ordered set sorting, Discrete Math. 49 A984), 223-236.
[5] D. Corson, D. Favero, K. Liesinger, and S. Zubairy, Characters and
g-series in Q(\/2), preprint. [6] S. Corteel, Particle seas and basic
hypergeometric series, Adv. Appl. Math., 31 B003), 199-214. [7] S.
Corteel and J. Lovejoy, Frobenius partitions and the combinatorics of
Ramanujan's iV>i summation, J. Combin. Theory Ser. A 97 B002),
177-183. [8] S. Corteel and J. Lovejoy, Overpartitions, Trans. Amer.
Math. Soc, to appear. [9] N.J. Fine, Basic Hypergeometric Series and
Applications, American Mathematical Society, Providence, RI, 1988.
[10] F. Garvan, Partition congruences and generalizations of Dyson's
rank, PhD Thesis, Penn State, 1986. [11] G. Gasper and M. Rahman,
Basic Hypergeometric Series, Cambridge University Press,
Cambridge, 1990. [12] J.T. Joichi and D. Stanton, Bijective proofs of
basic hypergeometric series identities, Pacific J. Math. 127 A987),
103-120. [13] L.W. Kolitsch, Some analytic and arithmetic properties
of generalized Frobenius partitions, PhD thesis, Penn State, 1985.
[14] J. Lovejoy, Gordon's theorem for overpartitions, J. Combin. Th.
Ser. A 103 B003), 393-401. [15] Padmavathamma, Studies in
generalized Frobenius partitions, Ph. D. Thesis, Univ. of Mysore,
1985. [16] J. Propp, Some variants of Ferrers diagrams. J. Combin.
Theory Ser. A 52 A989), no. 1, 98-128. [17] J.P.O. Santos and D.V.
Sills, g-Pell sequences and two identities of V.A. Lebesgue, Disc.
Math. 257 B002), 125 - 143. [18] J. Sellers, New congruences for
generalized Frobenius partitions with two or three colors, Discrete
Math. 131 A994), 367-373. [19] J.J. Sylvester, A construtive theory
of partitions in three acts, an interact and an exodion, in Collected
Math. Papers, vol. 4, pp. 1-83, Cambridge Univ. Press, London and
New York, 1912; reprinted by Chelsea, New York, 1974. [20] A.J.
Yee, Combinatorial proofs of generating function identities for F-
partitions, J. Combin. Theory Ser. A 102 B003), 217-228. [21] A.J.
Yee, Combinatorial proofs of Ramanujan's iV'i summation and the g-
Gauss summation, preprint. [22] D. Zeilberger, A g-Foata proof of
the g-Saalschutz identity, European J. Combin. 8 A987), 461-463.
Sylvie Corteel CNRS, PRiSM, Email: [email protected] Jeremy
Lovejoy CNRS, LaBRI, Email: [email protected] Ae Ja Yee Pennsylvania
State University, Email: [email protected]
Trends in Mathematics, © 2004 Birkhauser Verlag Basel/Switzerland
Emimerative Results on Integer Partitions Using the ECO Method
Luca Ferrari, Renzo Pinzani, and Simone Rinaldi ABSTRACT: In this
paper we apply the ECO method to the study of some enumerative
properties of integer partitions. In so doing, we both give an original
description of some known constructions regarding partitions and
propose some results, especially in the context of generalized hook
partitions (i.e., partitions whose Ferrers diagrams fit inside of a
suitable hook shape). 1. Introduction The main goal of this work is
to present an alternative approach to the study of enumerative
properties of integer partitions. Namely, we are going to apply the
ECO1 method in order to effectively construct integer partitions.
After briefly recalling the basics of the ECO method, with particular
emphasis on those notions we will need in the sequel, we give a
general ECO construction for partitions. We then modify such a
construction to deal with restricted classes of partitions. Next we
propose a second (less classical) construction of partitions and make
use of it to propose a bijective approach to Lecture Hall partitions,
leading to a challenging open question. Finally, we state some new
results on generalized hook partitions, also proposing an open
problem on partitions fitting into the intersection of two hooks
having different shapes. Our notations will be quite classical. In the
partition A = {pi,-.-,Pi) the parts are in decreasing order, that is p\ >
P2 > • • • > Pi- If A is a partition of ra, then we write A h n or |A| =
n. The whole class of integer partitions will be denoted T. 2.
Background We call ECO system a purely formal system of the kind
n-{ (k)^(ei(k))(e2(k))---(ek(k)) ' (^ where a, k,ei(k),..., ek(k) are
positive integers, (a) is called the axiom of the ECO system; the
second row is just a shortcut in order to express a set of productions
of the denoted form, for k running over a (finite or infinite) set of
positive integers. Thus, for example, the following are typical
examples of ECO systems: f \ I B) W : \ xWe recall that ECO stands
for Enumeration of Combinatorial Objects.

26 Luca Ferrari, Renzo Pinzani, and Simone Rinaldi Here we tacitly

assume that the ECO systems (*) and (**) have, as labels, all the
positive integers (greater than 2), whereas the ECO system (***)
contains the only labels A) and B). Among the many ways of
interpreting an ECO system, a well-known and pic- pictorial one is
that of drawing its generating tree, which is, by definition, the
infinite, rooted, labelled tree whose root is labelled (a) (the axiom)
and such that every node labelled (k) produces exactly k sons,
labelled respectively (ei(k)),..., (ek(k)). For instance, the generating
tree related to the ECO system (*) looks as follows: 4) E) Figure 1.
The first 4 levels of the generating tree of (*). From an enumerative
point of view, one of the main information encoded by an ECO
system is the numerical sequence (/n)nsN such that fn is the num-
number of nodes at level n in the generating tree related to the ECO
system (the concept of level in a generating tree like that in Fig. 1 is
straightforward). Actu- Actually, it often happens that ECO systems
appear as coding of special combinatorial constructions, in which the
objects of a given class are recursively constructed by performing a
sort of local expansion on them. This kind of combinatorial construc-
construction is the core of the ECO method, for which we refer the
reader to the paper [BDLPP] introducing it. Throughout our work we
will be mainly concerned with some modifications of the classical
concept of ECO system (or succession rule, as it is often called)
given above. The first generalization we need is that of mixed
succession rule. We start by introducing the concept of mixed
generating tree: it is a rooted, labelled tree whose edges can have
different "lengths" (whereas in the classical case the length of each
edge is equal to 1). The lengthened level (briefly, level) of a node in
a mixed generating tree is defined as the sum of the lengths of the
edges connecting the root to that node. Now consider a set of
production rules (instead of a single one, as we do in a classical ECO
system) and make it act on every label, each rule producing its sons
at a different level. What we get is precisely a mixed generating tree
where a set of h (possibly different) production rules is involved:
every node produces sets of sons

Integer partitions with the ECO method 27 at h different successive

levels, each set according to one of the defining succession rule. A
simple example of a mixed succession rule is the following. Consider
the two production rules: (k)(k + l) , (k) ¦* A)B) ¦ ¦ ¦ (k- One can
define the following mixed succession rule: A) (*) - B)C) ¦ ¦ ¦ (k)(k +
1) . ? A)B) ¦ ¦ ¦ (k - l)(k) The above formal system must be
interpreted as follows: in its mixed gener- generating tree a node at
level n produces 2 sets of sons, one set at level n + 2 (according to
the first production rule) and the other at level n + 3 (according to
the second production rule). From the point of view of the ECO
method, it is not difficult to see that the above mixed succession rule
arises from a natural construction of generalized Motzkin paths using
an horizontal step of length 3. The theory of mixed succession rules
has not yet been developed in its full generality, but some specific
cases have been considered. Here we only mention the paper
[FPPR], in which a particular kind of mixed succession rules (called
jumping succession rules) is systematically studied. Another
generalization of the concept of succession rule we will need in the
sequel is the notion of ECO system with multiple labels. More
precisely, it is an ECO system in which the labels are not simply
positive integers, but rather couples (more generally, ^-tuples) of
positive integers. Also this concept has occasionally surfaced in
previous works (see, for instance, [GPP]), but the theory behind it
remains largely undeveloped. 3. A general succession rule for integer
partitions In this section we will rediscover a classical construction of
integer partitions giving it an ECO interpretation. For any k € N, k ^
0, consider the following mixed succession rule: (k) (h) i A) 4 B) ¦ C)
If we denote by Fk(x) the generating function of the numerical
sequence associated with the rule fU, we easily have: For a generic
k, a look at the generating tree of the rule suggests the following
equality: Fk{x) = 1 + xF^x) + x2F2(x) + ... + xkFk(x),

28 Luca Ferrari, Renzo Pinzani, and Simone Rinaldi whence, using a

simple induction argument: k If we let k tend to infinity, we obtain
the following generating function: F(x) = lim Fk(x) = J k—+oo E)
Finding an ECO system ft associated with the generating function
F(x) is possible, provided that we allow the use of the symbol oo as
a possible label. In fact, we get for ft the following expression: (oo)
CO - (i) - B) , F) - CO where the unique node labelled (oo) (which is
the axiom of the system) produces, by convention, a node labelled
(i) at level i, for every i > 1 (so that it has an infinite number of
sons). The first levels of the generating tree of ft are given in Fig. 2.
D) A) B) Figure 2. The first 4 levels of the generating tree of Q.
Using a terminology taken from [FPPR], we can informally say that
Q, = -xxjQjfc. The rule Q enumerates precisely integer partitions,
since F(x) is the well-known generating function of them. Every
integer partition A = (pi,- ¦ ¦ ,pi) has a simple graphical
representation in terms of Ferrers diagrams made of I rows such
that the i-ih row is made of pi cells. There is an ECO construction of
Ferrers diagrams of partitions encoded by Q; it can be described as
follows: i) the empty object (Ferrers diagram without cells,
corresponding to the void partition of 0) has label (oo) and produces
a single row having h cells at each level h (corresponding to the
partition (h) of h);

Integer partitions with the ECO method 29 ii) a Ferrers diagram

representing a partition of n, whose bottom row is made of h cells,
has label (h) and produces h Ferrers diagrams representing
partitions of n + 1,..., n + h, respectively, and having labels A),...,
(h), as it is shown in Fig. 3. C) A) m D.3.1) D.3) B) C) D.3.2) m
D.3.3) Figure 3. The ECO construction for partitions. An obvious
consequence of this construction is that, for any k € N, the rule Clk
enumerates partitions A = (pi,... ,pi) where each part pi is less than
or equal to k. 4. A generalization and some examples The previous
rule for integer partitions can be suitably generalized. Suppose to
consider only those partitions using a prescribed type of parts. More
precisely, a sufficiently general setting consists of taking partitions
each of whose parts depends on the previous one. So, given a
partition A = (pi,...pi) \~ n, we require that Pi e II(pi_i) C N (i.e., the
set II(pi_i) only depends on pi-i). In particular, we define II(oo) to
be the (possibly infinite) set of allowed "maximum parts" for the
partitions we are taking into consideration. Therefore, a construction
for the above described class of partitions can be obtained by
suitably modifying the general ECO construction for partitions given
above, so getting to the following succession rule: @0) (oo) -w G)
(k) 31 3k (**) In the above ECO system the set {(an)}n is the set of
labels of the parts belonging to the set II(oo) = {in}n, whereas, if
(k) is the label of a part p of a partition, then {(fci),... (kk)} is the
set of labels of the parts belonging to the set

30 Luca Ferrari, Renzo Pinzani, and Simone Rinaldi = {ji,. --,jk}-

Some interesting examples fall into this framework and can be
obtained by specializing the quite general succession rule in G). 4.1.
Partitions into distinct parts and partitions into odd parts Let D be
the set of partitions into distinct parts and 0 the set of partitions into
odd parts. The two classes are easily proved to be enumerated by
the same generating function: X' (8) The ECO construction
previously proposed for integer partitions can be slightly modified in
order to describe an ECO construction for the classes 0 and V. A
construction for D.: Partitions of integers into distinct parts are
repre- represented by Ferrers diagrams whose rows all have
different lengths. The ECO construction is the following: i) the empty
partition of 0 has label (oo) and produces a single row made of h
cells at each level h (corresponding to the partition (h) of h); so
n(oo) = N* = N \ {0}; ii) a Ferrers diagram representing a partition
of n, whose bottom row is made of h+1 cells, is labelled (h) and
produces h Ferrers diagrams rep- representing partitions of n+1,...
n+h and having labels @),..., {h-1), respectively, as it is shown in
Fig. 4; therefore, U(h+1) = {1,2,..., h}. B) @) E,4,1) E,4) A) E,4,2)
B) E,4,3) FIGURE 4. The ECO construction for partitions into distinct
parts. The succession rule associated with this construction is the
following: (oo) (h) ± @) A) (h-1) 0)

Integer partitions with the ECO method 31 In the generating tree of

ft23, a node labelled @) produces no son. The generating function
for 0,® can be obtained in an analogous way to that for the
generating function of ft, getting again the series Ft, (x) =
Ili>oA+a;i) in (8). A construction for 0.: An ECO construction for the
class of partitions into odd parts can be obtained by specializing the
above described general setting to the case in which II(oo) is the set
of odd positive integers and, for every odd p, U(p) = {q \ q odd, q <
p} (see Fig. 5). C) A) E,5,1) E,5) B) E,5,3) C) E,5,5) Figure 5. The
ECO construction for partitions into odd parts. The associated
succession rule is: @0) (h) - A) - B) • 2h^ (h) Also in this case it is
easy tb determine the generating function by an inductive argument,
so obtaining F0(x) — fLx) l-gWi > ^ ^ is we^~ known. This last
example is particularly nice, since II(oo) is the set of all possible
parts appearing in the ECO system (odd positive integers) and II(p)
is just the subset of II(oo) whose elements are less than or equal to
p. More generally, if II(oo) = {an}n and II(on) = {ak \ k < n}, the
usual inductive argument leads to the generating function This is the
classical result concerning the enumeration of partitions whose parts
belong to a fixed set.

32 Luca Ferrari, Renzo Pinzani, and Simone Rinaldi 5. An alternative

ECO system for integer partitions The ECO approach suggests an
alternative construction for integer partitions. Starting from a given
partition A = (p1,... ,pt) of n, we define two new partitions of n+1
and n+pi, respectively, which are precisely Ai = (p\ + l,p2, • • • ,pi)
I- n+1 and X2 = (pi,pi,p2, ¦ ¦ ¦ ,Pi) \~ n+pi. Fig. 5 graphically
describes this construction on Ferrers diagrams. D) E) E,2,1) D,2,1)
D) D,4,2,1) rule: Figure 6. An alternative ECO construction for
integer partitions. Such a construction is immediately seen to be
associated with the succession A) (h) A0) ¦ (h+1) . * (h) Rule f2* in
A0) does not satisfy the consistency principle typical of ECO systems,
i.e. labels do not denote the number of sons; in particular, each
node in the generating tree of f2* produces exactly two sons.
According to the construction suggested by Fig. E), each label (h)
corresponds to a partition whose maximum part is h. 6. Some
applications and open problems 6.1. Lecture Hall partitions The
theory of Lecture Hall partitions has been initiated in [BME1], which
is the basic article we refer the reader to concerning this topic. Only
in this section, we change our notation for partitions: if A = (pi,...
,pi) then we assume that pi < p2 < ¦ ¦ ¦ < pi. For k > 1, let ?& be
the following set of partitions (having possibly some empty parts):
0<S- <f <-< The elements of ?/. are called Lecture Hall partitions of
length k. We also denote by T>k the set of all partitions of Lk with
empty parts removed. For example, the partition B,3) of 5 belongs to
D3 but not to D2. It is clear that, for any given A, there exists a
minimum k such that AeD^: this will be called the minimum length
of A as a Lecture Hall partition. The concept of Lecture Hall partition
allows to give a finite version of the well-known result (due to Euler)
that the number of

Integer partitions with the ECO method 33 partitions of n into odd

parts equals the number of partitions of n into distinct parts. More
precisely, in [BME1] it is shown that the number of partitions of n
into odd parts less than or equal to 2k - 1 is equal to the number of
Lecture Hall partitions of length k of n. The first bijective proof of (a
refined version of) this result appears in [BME2]; however, the
authors themselves admit that such a proof finds its origin in the
algebraic context of Coxeter groups. Some bijective proofs have
been recently given in [E, Y]. The present approach to integer
partitions suggests a possible way to find a new natural bijection
proving the Lecture Hall theorem in a purely combinatorial way. The
idea is to give two distinct combinatorial interpretations to the
generating tree associated with a given ECO construction. A possible
ECO construction for 0 can be obtained by suitably modifying the
one given in section 5 for unrestricted partitions. It is not difficult to
see that the associated succession rule is the following: \0 (h + 2)
(h) A1) It is clear that the set of labels of this ECO system is the set
of odd positive integers. The first levels of its generating tree are the
following: Figure 7. The first levels of the generating tree of the rule
Q,®. We conjecture that the ECO system in A1) provides a
construction also for partitions into distinct parts. This would lead to
a presumably new bijection between 0 and T> from which it would
be immediate to deduce an explicit bijection proving the Lecture Hall
theorem. Indeed, in the conjectured interpretation of the above
generating tree, it seems natural to think that the label of a partition
is strictly related to its (minimum) length as a Lecture Hall partition:
more precisely, a node labelled 2k — 1 represents a Lecture Hall
partition of minimum length k. Concerning this problem, we cite the
remarkable paper [P], where the author introduces a truly nice
setting to deal with bijective questions on partitions which could be
useful in this context.

34 Luca Ferrari, Renzo Pinzani, and Simone Rinaldi 6.2. Generalized

Hook partitions In the paper [BR], Berele and Regev show how the
representation theory of Lie superalgebras heavily relies upon the
knowledge of the combinatorics of partitions fitting inside of a hook
shaped figure (we will briefly call them generalized hook partitions).
The above article has been followed by many others, such as [Rl,
R2], and the study of this kind of partitions is still object of
investigations in combinatorics and algebra. G,5,3,2,1,1) Figure 8. A
hook partition of shape B,3). Let "Kh,k be the set of generalized
hook partitions of shape (h, k), that is, by definition, the set of all
partitions that fit inside a hook shape of k rows and h columns. Our
aim is to restrict the general ECO construction for partitions to the
set !Kh,k- This will lead us to the determination of the generating
function of "Kh,k- We consider two disjoint subsets of J-Ch,k- i)
Partitions having j < k parts; a partition in this set has label (l,j),
where / is the number of cells in the last row of its Ferrers diagram.
The ECO construction applied to such a partition works exactly like
in the general case, leading to the following production: 1 I A2)
Observe that A2) is a production with double labels, since we have
to take into account also the number of parts. However, if j = k — 1,
we choose to delete the second label, for reasons that will be
explained below. ii) All the remaining partitions. In this case,
because of the hook shape con- constraint, the ECO construction is
made by adding to a Ferrers diagram only rows having at most h
cells. Since the last production described in i) produces simple labels,
here we can avoid the use of multiple labels. The associated
production is then the following: @ A) (m). A3) where m = min (l,h).

Integer partitions with the ECO method 35 Let Hh.k(x)be the

generating function for the class "Khtk. For any label (l,j), with j < k,
denoting by H^l\x) the generating function of the ECO system
having (l,j) as axiom and the above ones as production rules, we can
deduce the following recursion: <f (*) = 1 + xH^+l\x) + ... +
xlHili+1\x). A4) Analogously, if j > k we have: ?(*) + • • • +
*mH$(x)- A5) Starting from this general setting we are able to
compute the generating function of generalized hook partitions of
any fixed shape. For instance if we consider hook partitions of shape
B,3), we get to the following succession rule: (oo,0) (U) A,2) A) @
leading to the generating function: #2,3 (x) = H3>2(x) 1 A,2) A,2)
A) @ A) A) B) A6) 1 + xH xc More generally, we have: Hh>k(x) = x
h+1 X 1-x A7) Observe that the equality Hh,k(x) = Hk,h(x), which is
immediate by a com- combinatorial point of view, is by no means
obvious from an algebraic one. We point out here that the problem
of determining the generating functions Hh,k(x) was previously
considered in [OZ], where the authors find an explicit formula for
them. However, their expression involves rather complicated
quantities, whereas formula A7) is quite easy to read. It would be
instead interesting to extend this result to the case of partitions
fitting inside of the intersection of two hooks of different shapes.
Also this problem arises in the study of representation of Lie

36 Luca Ferrari, Renzo Pinzani, and Simone Rinaldi superalgebras, in

connection with the module decomposition of supersymmetric power
of matrices [S], and it is still open. References [BDLPP] E. Barcucci,
A. Del Lungo, E. Pergola, R. Pinzani, ECO: A methodology for the
enumeration of combinatorial objects, J. Differ. Equations Appl. 5
A999) 435-490. [BR] A. Berele, A. Regev, Hook Young diagrams with
applications to combinatorics and to representations of Lie
superalgebras, Adv. in Math. 64 A987) 118-175. [BMEl] M. Bousquet-
Melou, K. Eriksson, Lecture Hall partitions, Ramanujan J. 1 A997)
101-111. [BME2] M. Bousquet-Melou, K. Eriksson, A refinement of
the Lecture Hall theorem, J. Combin. Theory Ser. A 86 A999) 63-84.
[DFR] E. Deutsch, L. Ferrari, S. Rinaldi, Production matrices,
(submitted). [E] N. Eriksen, A simple bisection between Lecture Hall
partitions and partitions into odd integers, proceedings of FPSAC
2002, Melbourne. [FPPR] L. Ferrari, E. Pergola, R. Pinzani, S. Rinaldi,
Jumping succession rules and their generating functions, Discrete
Math. 271 B003) 29-50. [FP] L. Ferrari, R. Pinzani, A linear operator
approach to succession rules, Linear Algebra Appl. 348 B002) 231-
246. [GPP] O. Guibert, E. Pergola, R. Pinzani, Vexillary involutions
are enumerated by Motzkin numbers, Ann. Comb. 5 B001) 153-174.
[OZ] R. C. Orellana, M. Zabrocki, Some remarks on the characters of
the general Lie superalgebra, arXiv.math.CO/0008152vl, 2000. [P] I.
Pak, Partition identities and geometric bisections, Proc. Amer. Math.
Soc, to appear. [Rl] J. B. Remmel, The combinatorics of (k, I)-hook
Schur functions, Combinatorics and Algebra (Boulder, Colorado,
1983), 253-287, Contemp. Math., 34, Amer. Math. Soc, Providence,
RI, 1984. [R2] J. B. Remmel, A bijective proof of a factorization
theorem for (k, I)-hook Schur functions, Linear and Multilinear
Algebra 28 A990) 119-154. [S] T. Seeman private communication,
2003. [Y] A.J.Yee On the refined lecture hall theorem, Discrete Math.
248 B002) 293-298. Luca Ferrari Dipartimento di Scienze
Matematiche ed Informatiche, Pian dei Mantellini, 44, 53100, Siena,
Italy [email protected] Simone Rinaldi Dipartimento di Scienze
Matematiche ed Informatiche, Pian dei Mantellini, 44, 53100, Siena,
Italy [email protected] Renzo Pinzani Dipartimento di Sistemi e
Informatica, via Lombroso 6/17, 50135 Firenze, Italy
[email protected]

Trends in Mathematics, © 2004 Birkhauser Verlag Basel/Switzerland

321-A voiding Permutations and Chebyshev Polynomials Toufik
Mansour ABSTRACT: In [6] it was shown that the generating
function for the number of permutations on n letters avoiding both
321 and (d + l)(d + 2)... kl2 ...d is given by 2t^(t) for all k > 2, 2 <
d + 1 < k, where Um is the mth Chebyshev polynomial of the
second kind and t = ^j • In this paper we present three different
classes of 321-avoiding permutations which are enumerated by this
generating function. Let a?Sn and r e Sk be two permutations. Then
a contains r if there exists a subsequence 1 < i\ < %2 < • • • < h <
n such that (a^,..., aik) is order- isomorphic to r; in such a context r
is usually called a pattern; a avoids r, or is r-avoiding, if a does not
contain such a subsequence. The set of all r-avoiding permutations
in Sn is denoted by Sn(r). For a collection of patterns T, a avoids T if
a avoids all t eT; the corresponding subset of Sn is denoted by
Sn(T). While the case of permutations avoiding a single pattern has
attracted much attention, the case of multiple pattern avoidance
remains less investigated. In particular, it is natural to consider
permutations avoiding pairs of patterns Ti, t2. This problem was
solved completely for ri,r2 G S3 (see [8]), for t\ G 63 and r2 G S4
(see [10]). Several recent papers [1, 2, 4, 5, 6, 7] deal with the case
t\ e S3, T2 G Sk for various pairs Ti,T2, e.g. in [1] it was found by
using transfer matrices that the generating function for the number
of permutations in 5nC21, [k, k]) is given by where Um(cos6) =
sin(m + 1H/sin0 is the mth Chebyshev polynomial of the second
kind and [d, k] = d(d + 1)... fcl2... (d - 1). Later, in [6] Mansour and
Vainshtein proved a natural generalization for this theorem. Theorem
1. For any k > 2 and 2 < d + 1 < k, the generating function for the
number of permutations in 5nC21, [d + 1, k]) is given by Rk(x).
Recently, Mansour and Stankova [3] presented an exact enumeration
for the case 321-fc-gon-avoiding permutations in Sn which
generalizes the methods in [9] and [6]. In particular they proved the
following result: Theorem 2. For any k > 4 and 2 < d < k - 2, the
generating function for the number of permutations in 5nC21, {d +
l){d + 2) • • • (k - I)lfc23 • • • d) is given by Rk{x).

38 Toufik Mansour Let us define there patterns: ad k = d(d + 2)(d +

3)... kl2... (d - l)(d + 1), /3dk=d(d + 2)(d + 3)...(k- I)lfc23...(d ld,k
= d(d + 2)(d + 4)...kl2...(d- l)(d + l)(d + 3). The main theorem of
the paper is formulated as follows. Theorem 3. (i) Let k > 4 and 2 <
d < k—2. Then the generating function for the number of
permutations which avoid both 321 and ad,k is given by Rk(x). (ii)
Let k > 6 and 3 < d < k - 3. Then the generating function for the
number of permutations which avoid both 321 and fid,k is given by
Rk(x). (iii) Let k > 6 and 2 < d < k — 4. Then the generating
function for the number of permutations which avoid both 321 and
7^ is given by Rk(x). Our proof of the Theorem 3 is based on finding
a recursion for the numbers in question by purely analytical means.
In particular, we generalize the methods and extend the results in
[1, 3, 6]. In spite of the paradigm formulated in [2], that any
enumeration problem leading to Chebyshev polynomials is related to
Dyck paths, it would be tempting to find a proof that exploits such a
relation. References [1] T. Chow and J. West, Forbidden
subsequences and Chebyshev polynomials, Discr. Math. 204 A999)
119-128. [2] C. Krattenthaler, Permutations with restricted patterns
and Dyck paths, Adv. in Applied Math. 27 B001) 510-530. [3] T.
Mansour and Z. Stankova, 321-polygon-avoiding permutations and
Chebyshev polynomials, Elect. J. Combin. 9:2 B003) #R5. [4] T.
Mansour and A. Vainshtein, Restricted permutations, continued
fractions, and Chebyshev polynomials, Elect. J. Combin. 7 B000)
#R17. [5] T. Mansour and A. Vainshtein, Restricted 132-avoiding
permutations, Adv. Appl. Math. 126 B001) 258-269. [6] T. Mansour
and A. Vainshtein, Layered restrictions and Chebyshev polynomials,
Ann. of Combin. 5 B001) 451-458. [7] T. Mansour and A. Vainshtein,
Restricted permutations and Chebyshev polynomials, Sem. Lothar.
de Combin. 47 B002) Article B47c. [8] R. Simion, F.W. Schmidt,
Restricted Permutations, Europ. J. Combin. 6 A985) 383-406. [9] Z.
Stankova and J. West, Explicit enumeration of 321-hexagon-avoiding
permuta- permutations, Disc. Math., to appear. [10] J. West,
Generating trees and forbidden subsequences, Discr. Math. 157
A996) 363- 372. Toufik Mansour Department of Mathematics,
University of Haifa, 31905 Haifa, Israel [email protected]

Trends in Mathematics, © 2004 Birkhauser Verlag Basel/Switzerland

Iterated Logarithm Laws and the Cycle Lengths of a Random
Permutation Eugenijus Manstavicius ABSTRACT: We are concerned
with the iterated logarithm laws for map- mappings defined on the
symmetric group. For the sequences of the cycle lengths and the
different cycle lengths appearing in the decomposition of a random
permuta- permutation, such laws provide asymptotical formulas valid
uniformly in a wide region for the sequence parameter. The main
results are analogues to Feller's and Strassen's theorems proved for
partial sums of independent random variables. 1. Weak convergence
of distributions Let <x G Sn be an arbitrary permutation and a = ki-
• • kw, w = w(a), A) be its unique up to the order expression by the
product of the independent cycles k. Denote by un(...) = (n\)~l\{(r
G Sn : ... }| the uniform probability measure on Sn. In what follows
we assume that n —> oo. Despite to the long list of asymptotic
results (see [1], [2], [7-10], [12-15], and other publications), we
examine the ordered statistics 1 < Ji(°) <••¦< Jw{°) <n and l<M<r)
<---<ja(<r)<n, s = s(a), consisting of the cycle lengths and the
different cycle lengths appearing in decom- decomposition A). In this
remark, from these two very close cases we always choose that
requiring less technical details. Let us remind a few of classical
results. In 1942 Goncharov [8] (see also [9]) found the limiting
distribution, say, Fk(x), of Jw-k(a)/n for arbitrary fixed k > 0. In
particular, for the longest cycle, we have Vn(Jw{°) < xn) -+ FQ{x) if
0 < x < 1 and F0(x) = 1 if x > 1. Here and in what follows 1{-}
denotes the indicator function. The moments of Fk(x) were
calculated in [13]. As it is stated in [9] (see also [10]), un (log Jm <
xy/m + m) -+ $(x) provided that m = alogn + o(^logn) and 0 < a <
1. Here and in what follows $(ar) denotes the distribution function of
the standard normal law.

40 Eugenijus Manstavicius In 1972 A.M. Vershik and A.A. Schmidt

[14] announced (all details were furnished in [15]) several results on
the asymptotic distribution of the random element (Jw((r)/n,«/«,_!
(a)/n,... Ji(<r)/n,0,...) under the measure un in the simplex T :=
{(xi,x2, ¦ ¦ •) : 1 > Xi > x2 > ¦ ¦ • > 0, X\ + x2 H = 1}. Applying
continuous mappings of F to other spaces, they derived some fairly
in- interesting corollaries. In particular, they proved that lim lim un
(\og( Jw-i(a)/n) < x\Ti - i t-toon-too \ Our questions are simple:
What asymptotic results on {Jm(<?)} and {jm(o")} are valid
uniformly in m e [fln,^n]> where [an,bn] is a fortiori given
subinterval of [l,n]? What type of con- convergence can give such
information? First, we indicate an idea that the functional limit
theorems can give some approach to the problem. Let kj(a), 1 < j <
n, denote the number of cycles of length j in A). By the functional
limit theorem for additive functions on the symmetric group [4] or
[3], the processes Wn(a;t):=^-^(w(a;nt)-t\ogn), w(a;y) := ^hj(a),
te[0,l], weakly converge to the standard Brownian motion W(t).
Taking the maximum functional, we obtain un ( max \Wr(o~,t)\ < x
) -> P ( max \W(t)\ < x) Vefoi] " / ye[o,i] / The maximum under
consideration can be attained at the points t = log Jm(a) /logn, 1 <
m < w or at t = 1. Applying this together with w(a; Jm(a)) = m, we
derive / \ vn max|log Jm(«r) — m\ < Xyflogn, \w(a) -logn| < a;vlogn
= V(x) +o(l). \m<w J C) Further, observing that by B) un (\ log
Jw(a) - logn\ > e^logn) . = 1 - FQ(exp{e^logn}) for each e > 0, we
can drop the second event in the frequency of equality C). So, we
obtain the following result. Theorem 1.1. We have vn
Imax|logJm(o-)-m| <Xy/\ogn) = V(x) + o(l). \rn<w J D) Most
probably, the last assertion could also be derived from the aforemen-
aforementioned results proved in [14] and [15].

Iterated logarithm laws 41 2. Strong convergence The next result

related to our question involves stronger convergence than the weak
convergence of distributions. Actually, we examine an analog of the
convergence with probability one. To give a sense for this notion on
a sequence of probability spaces, we have to consider convergence
in distribution of the "tails'' of sequences of random variables. In
[12] we proved the following result. Theorem 2.1. We have hm hm
vn max — L > i i x \ _ n E) m-oon-oo U \n^<m<8 Bmloglog mI/2 ~
+ ) ~ V ' and lim lim vn ( min ni-+oo n_+oo -m BmloglogmI/2 -b
<d)=l F) for each b G [-1,1] and 5 > 0. In connection to D), E), and
F), it is worth to recall that tho mean values of w = w(a) and 5 =
s(a) are asymptotically equivalent to log n. The estimate of
U°gJm(o") - m\ following from E) still has the error 5Bmloglog
toI//2. In this direction we now have the following improvement.
Denote Lu = logmax{«, e} = L\u,..., L^u = L(Lk-iu) for u € R- For 5
> 0 and k > 2, set Emk(l ±5)= Urn (L2m + -L3m + L4m + ¦ • • + A
± 5)Lkm\ J . Theorem 2.2. For arbitrary 0 < 5 < 1 and k > 2, we
have ,. f— ( |logjm((T)-m| \ hm hm vn max ° \ ' ' > 1 = 0 ni-+oon-
+oo \n\<m<s pkrnyi.-\-Oj J and lim M» n1-+oon_+oo ^ ( j Thus,
answering our question we may say that "for almost all a € Sn"
\logjm(o-)-m\<f3km(l+6) uniformly in m, nx < m < s(a), where m -
> oo arbitrarily slowly. This assertion is sharp in the sense that we
can not change 5 by -6. The numbers jm{a) are just the success
epochs of the following partial sum function ;?>l}, 1 < V < n. By
virtue of s(a;jm(a)) = m but using a bit complicated calculations,
one can show that Theorem 2.2 is equivalent to the next result. For
6 > 0 and k > 3, set \\ 1//2 7fcm(l ±S)= B{Lm) [L3m + -L4m +
L5m + • • • + A =' " '

42 Eugenijus Manstavicius Theorem 2.3. Let 0 < 5 < 1 and k > 3 be

arbitrary. We have i- r-— + v v— ( |s(<r;m) — logm| \ lim hm i/+ :=
lim hm vn max •L-^-J—?- ^—- > 1 I = 0 m-+oon-+oo 1 ni-+oon-
+oo yni<m<n TfcroA ~l~ Oj I and ,. ,. _ ,. ,. / \s(a;m) — logm| \
hm hm un „ := hm hm i/n max ——-—; ^—¦ > 1 = 1. ni °° n—+oo
ni~+o° n—+oo yni<m<n 'y^M —o) J A proof will be given in the
last section. 3. Strassen's law of iterated logarithm Adopting our
experience obtained in probabilistic number theory [11], we can pro-
propose further generalization of the results of Section 2. We now
deal with sequences of functions defined on the symmetric group.
Set Sm{<T;t) = — The trajectories of these random processes lay in
the space D = D[0,1] of the right continuous functions on [0,1]
having left-hand limits at each point. We assume that D is endowed
with the Skorokhod metrics. We also introduce the linearized version
Gm(a;t) of Sm(a;t). For ti := (log/)/logm and t € [ti,ti+\), we set
Gm((r;t) = Sm(a;ti) + (Sm(a;ti+i) — Sm(a;ti)) —, ti+i — ti where 1
< / < m — 1. Our problem now is to describe the set of limit curves
of {Sm(a; t)} and {Gm(<T; t)} "for almost all a e Sn" in the sense of
Section 2. Again, to overcome some obstacles appearing in the
sequence of probability spaces, we have to introduce more general
notions. Let (U, d) be a separable metric space and Y, Y\,..., Yn be
U-valued random elements all defined on the probability space {fin,
Jn,Pn}, CP = {Pn}, where n = 1,2,.... Let, as usual, d(X, A) = inf
{d(X, Z) : Z eA}, where A C U. We say that Ym converges to Y CP-
almost surely (CP — a.s.) if for each e > 0 lim lim Pn f max d(Ym,Y)
> e) = 0 n\—+oo n—+oo \ni<m<n J Thus, a compact set A C U
such that, for each e > 0 and each X E A, lim lim Pn max d(Ym,A) >
e ) = 0 G) \—+oo n—+oo \ni<m<n I and lim lim Pn ( min d(Ym,X)
< s ) = 1 (8) ni-+oon_+oo \ni<m<n J may be called a cluster set of
the sequence {Ym} 7 - a.s.. In what follows we denote the relations
G) and (8) by Ym^A (y-a.s.). (9) Denoting (8) by Ym> —> X (CP -
a.s.) we have in mind that m! can be a random increasing
subsequence.

Iterated logarithm laws 43 Let C = C[0,1] be the space of

continuous functions on the interval [0,1] endowed with the
supremum distance /?(.,.). We recall that the Strassen set X agrees
with the set of absolutely continuous functions g such that g@) — 0
and / (g'(t)Jdt<l. o These definitions are applicable to {Gm(a, •)}
and {Sm(a, •)}, where 1 < m < n. For the family of probabilities, we
now take v := {un}. Theorem 3.1. We have Gm(v,-)^3C (v-a.s.) A0)
in the space C and Sm(a;-)^X (u-a.s.) A1) in the space D. The idea
of the proof is similar to that of Theorem 2.3. The details will be
exposed in a forthcoming paper. Applying the same argument as in
the last section, we can verify that max sup \Sm(er;t)-Gm(<r;t)\> e)
=o(l) <m<n[0\] J for every e > 0 as n —> oo and n\ —> oo. So,
A0) and A1) are equivalent. The following lemma is very useful in
various applications of the last result. Lemma 3.2. Let (V,d) and
(V\,di) be separable metric spaces and let f : U —> Ui be a
continuous map into Ui. If A is compact subset of"V and Ym <—>¦ A
(CP —a.s.) in (U,d), then f(Ym) <--*¦ f(A) (CP - a.s.) in the second
space ) By virtue of this lemma, Theorem 3.1 implies Theorem 2.3
which assertion can now be rewritten as Going along this path, we
can list more consequences. Corollary 3.3. The following relations
hold (u — a.s.): • (Gm(a; 1/2), Gm(a; 1)) -> ? := {(u, v) : u2 + (v -
uf < 1/2} ; (/)[//} • if m' is the subsequence for which Gm/(<r;l/2)
—> \2/2, then we have Gm'(<?;¦) ->9i, where _ (ty/2, if 0<t< 1/2,
9l[t)-\V2/2, if 1/2 <«< 1; • ifm' is the subsequence for which
Gm'(<T; 1/2) —> 1/2 and Gm/(a; 1) —> 0, then we have Gm>(o-,
¦) —> g<i, where n\ - /*» # ° ^ * - 1/2' teW-\l-t, if 1/2 <t<l. These
sophisticated examples of continuous functional / can be found in
[6]. Using the fact that the cycle lengths jm(a) are the counts of the
partial sums s(a;m), as in Section 2, we could convert the relations
of this Corollary to that for jm(a). Nevertheless, we now prefer the
functional form of doing that.
44 Eugenijus Manstavicius First, we introduce a new sequence of
processes. Set j(cr; «) = lifO<«<l and j{a;u) = j[u](a) if 1 < u < s =
s(a). Define Theorem 3.4. We have Vm(<r;-)^X {u-a.s.) in the space
D. To derive this claim from Theorem 3.1, we can use the
generalized inverses. Let Do denote the subspace of D consisting of
nonnegative nondecreasing func- functions. For X G Do, we define
X~x G Do by X-1 (t) = inf {u G [0,1] : X(u) > t} with the agreement
X'1^) = 1 for X(l) <t<l.A useful auxiliary result for the proof has
been provided by W. Vervaat [16]. Lemma 3.5. Let Xm G Do and 5m
be a sequence of positive numbers, 5m —> 0. If 9 = 9(t) ? C, then
as m —> oo the following relations are equivalent and sup sup
Xm(t)-t ~9(t) 9{t) : t€[0,l]j-0 For the proof of Theorem 3.4, it
suffices to apply A1) and Lemma 3.5 with g ex, Xm = Xm{a,t) = s(a,
m')/ log m, and 5m = BL3m)/logmI' . In the style of Corollary of
Theorem 3.1, we have, for instance, (u-a.s). To check the second
relation here, we can use the two-dimensional convergence pointed
out in the first item of Corollary 3.3. The last two relations show
some fascinating symmetry between the behaviour of log j(o~, tm)
and log (j(a; m)/j(a; A - t)m)) at the point t = 1/2. We have some
other observations of a similar symmetry. In what general forms can
such phenomenon appear? 4. Proof of Theorem 2.3 The main
probabilistic ingredient is found in the classical paper of W. Feller [5].
Let Xn, n > 1 be independent random variables (r.vs), EXn = 0, EX%
< oo, and B2n := EX\ + ¦ ¦ ¦ + EX% -> oo. Here as previously n ->
oo. Denote Yn = Lemma 4.1. Let C > 0 be a constant and \n be a
positive increasing sequence. In addition, assume that the r.vs Xn,
n> 1, satisfy \Xn\<CB2JXi

Iterated logarithm laws 45 with probability one. If the series A2) n>\
converges, then P (Yn > XnBn infinitely often) = 0. If the series A2)
diverges, then P (Yn > XnBn infinitely often) = 1. We will need just a
corollary for independent Bernoulli r.vs ?,n,n>l, such that P(?n = 1)
= pn = 1 - P(?n = 0). We reformulate it for the sum (n := ?i H 1- Cn
in a slightly modified form. Corollary 4.2. Let the Bernoulli r.vs satisfy
the condition pn = 1/n + O(l/n1+e), where s > 0 is arbitrary. Then,
for every 0 < 5 < 1 and k>2>, we have and ^J^P+^.-^lhn^J^pf
max ICm logm| > lim lim P := lim lim P max ni-+oon-KX) ni-ioon-
+oo \ — 0) = 0 = 1. Our approach originated in probabilistic number
theory (see [11]) has two steps. The first one is based upon the
following lemma. Lemma 4.3 (Fundamental Lemma). There exist a
probability space {fi,^,P} and independent Poisson r.vs Zj, EZj = 1/j,
j > 1, such that \un((k1(a),...,kr(a))eA)-P((Z1,...,Zr)eA)\<Cr/n with
an absolute constant C > 0 uniformly in A C Z+r for each 1 < r < n.
The assertion of this lemma follows from the Feller coupling. Much
more precise estimate of this total variation distance is proved in [2].
It is also shown that the distance does not vanish when the
condition r = o(n) is not satisfied. So, as in the probabilistic number
theory (see [11]) Fundamental lemma allows to deal with
"truncated" up to r additive functions. The remainder appearing in
this procedure can be estimated by the following inequality obtained
in author's paper [12]. Lemma 4.4. Let {hj(k)}, k > 0 and j > 1, be a
two-dimensional array of real numbers such that hj(O) = 0. If Zj, j
>1, are the Poisson r.vs as in Lemma 4.3, then for arbitrary x > 0,
bm G R, and 1 < r < m < n, we have vn\ max r<j<n > x I < 32e2P
I max r<j<n > x/3 . In particular, for arbitrary dr > ¦ ¦ ¦ > dn > 0
and aj G R, j > 1, 288e2 J un max dm V r<m< E «aw - E t >-l< ,2

46 Eugenijus Manstavicius Proof of Theorem 2.3. We start from the

observation which actually explains why the results on the
sequences {w(a; m)} and {s(a; m)} are so close. This further leads
to the similarity of estimates valid for {Jm(a)} and {jm{<r)}. If e >
0 and a positive sequence tp(n) / oo are arbitrary, then by Lemma
4.4 we have •=^n[ max \ni<m<n 1 \s(a;m) - w(a; m)\ > e) J 32e2P
max ip(m) -l ni<m<n > 1}) >e/3 < 32e2P \j<n as n —* oo and ni
—* oo. Take r = r(n) = n/Ln > ni and some arbitrary e > 0. By
virtue of the last estimate and Lemma 4.4, we obtain vn ( max
7fcm(l + 5) |s(<r;m) -logml > 1 ) \r<m<n I \ — — / <!/„( max
7fcm(l+5)~1|«;(cT;m)-logm| > 1/2) +vrn(l/2) \r<m<n J Hence,
with such a choice of r we have + ^ f \s(o-;m) -logm\ \ . Kin < "n
max ^ ' > 1 I + o( By the fundamental Lemma this may be rewritten
as v^ n < P max 7fcm(l + 5)~ 1 \ ni<m<r For the Bernoulli r.vs ?j
:= l{Zj > 1}, j > 1, we may apply Corollary 4.2. It yields the first
assertion of Theorem 2.3. To prove the second one, it suffices to
observe that \s(a;m) — logmj \ni<"i<r 7fcm(l-t>) / and again to use
the corollary with ?j = ^-. Theorem 2.3 is proved. References [1] R.
Arratia and S. Tavare, A992) Limit theorems for combinatorial
structures via discrete process approximations, Random Structures
and Algorithms 3, 321-345. [2] R. Arratia, A. Barbour, and S. Tavare,
Logarithmic Combinatorial Structures: a Probabilistic Approach, EMS
Monographs in Math., The EMS Publishing House, Zurich, 2003.

Iterated logarithm laws 47 [3] G.J. Babu and E. Manstavicius, A999)

Brownian motion for random permutations, Sankhya A: The Indian J.
of Statistics 61, No. 3, 312-327. [4] J.M. DeLaurentis and B.G. Pittel,
Random permutations and the Brownian motion, Pacific J. Math. 119
A985), No. 2, 287-301. [5] W. Feller, The general form of the so-
called law of the iterated logarithm, Trans. AMS 54 A943), 373-402.
[6] S. D. Freedman, Brownian Motion and Diffusion, Holden-day, San
Francisco, 1971. [7] V.L. Goncharov, On the distribution of cycles in
permutations, Dokl. Acad. Nauk SSSR 35 A942), 9, 299-301
(Russian). [8] V.L. Goncharov, Some facts from combinatorics, Izv.
Akad. Nauk SSSR, Ser. Mat. 8 A944), 3-48 (Russian); On the field of
combinatory analysis, Transl. AMS 19 A962), 1-46. [9] V.F. Kolchin, A
problem of the allocation of particles in cells and cycles of random
permutations, Teor. Veroyatnost. i Primenen. 16 A971), No. 1, 67-82
(Russian). [10] V.F. Kolchin, Random Mappings, Optimization
Software, Inc. New York, 1986. [11] E. Manstavicius, Functional
approach in the divisor distribution problems, Ada Math. Hungarica
66 A995), No. 3, 343-359. [12] E. Manstavicius, The law of iterated
logarithm for random permutations, Lith. Math. J. 38 A998), No. 2,
160-171. [13] L.A. Shepp and S.P. Lloyd, Ordered cycle lengths in a
random permutation, Trans. AMS 121 A966), 340-357. [14] A.M.
Vershik and A.A. Schmidt, Symmetric groups of high order, Dokl.
Acad. Nauk SSSR 206 A972), No. 2, 269-272 (Russian). [15] A.M.
Vershik and A.A. Schmidt, Limit measures that arise in the
asymptotic theory of symmetric groups I, Teor. Veroyatnost. i
Primenen. 22 A977), No. 1, 72-88; II, ibid. 23 A978), No. 1, 42-54
(Russian). [16] W. Vervaat, Success epochs in Bernoulli trials with
applications in number theory, Math. Centrum, Amsterdam, 1972.
Eugenijus Manstavicius Vilnius University, Department of
Mathematics and Informatics [email protected]

Trends in Mathematics. © 2004 Birkhauser Verlag Basel/Switzerland

Transcendence of Generating Functions of Walks on the Slit Plane
Martin Rubey ABSTRACT: Consider a single walker on the slit plane,
that is, the square grid 1? without its negative x-axis, who starts at
the origin and takes his steps from a given set 6. Mireille Bousquet-
Melou conjectured that - excluding pathological cases - the
generating function counting the number of possible walks is
algebraic if and only if the walker cannot cross the negative x-axis
without touching it. In this paper we prove a special case of her
conjecture. 1. Introduction Let 6 - the set of steps - be a finite
subset of 1?. A walk on the slit plane is a sequence @,0) = wQ,
Wi,..., wn of points in 1?, such that the difference of two consecutive
points W{+i — Wi belongs to the set of steps 6 and none of the
points but the first lie on the half-line {(x, 0) : x < 0}. An example
for such a walk with set of steps 6 = {(-1, -2), (-1,1), (-1, 2), A, -2),
A,1), A, 2)} is shown in Figure 1. Recall that a generating function
F(t) = J2n>o fntn ls algebraic, if there is a nontrivial polynomial P in
two variables, such that P(F(t),t) = 0. Otherwise, it is
transcendental. In [2] Mireille Bousquet-Melou conjectured the
following: Conjecture 1.1. Consider the generating function for walks
in the slit plane with a given set of steps &, counted according to
their length and their end-coordinates: S(x, y;t) = ?
*lengthwxx'finalwyy^ w. W walk on the slit plane starting at the
origin with steps in & Suppose that the set of steps is not
degenerated and thus all four quadrants of the plane can be reached
by some walk, and that the greatest common divisor of the vertical
parts of the steps is equal to one. Then this generating function is
algebraic in t, if and only if the height of any step is at most one. In
fact, she proved one part of this conjecture in Section 7 of the above
paper, namely, that walks with steps that have height at most one
have an algebraic generating function. Furthermore, in Section 8 she
proved for one family of step- sets that the corresponding
Other documents randomly have
different content
 “I guess so, son,” said Pete, rising rather weakly to his feet, for
the blow the tree had struck him, while it had not broken the skin,
had been a stunning one.
“You see,” he went on, “they got a good start of us and should
have reached the high ground afore the water hit.”
“That’s so,” agreed Jack, “and I can see now that the water did
not rise so very high. It was its speed and anger that made it
terrible.”
“Wonder how far that blamed old tree carried us,” said Pete,
rather anxiously. “It’s just curred to me that if we don’t connect with
the stock and some grub pretty quick, we’ll be in a bad fix.”
He gazed about him as he spoke. On every side stretched
monotonous plains covered with the same gray-green brush as the
savannah amidst which they had camped the night before. But the
question in Pete’s mind was whether or not it was the same plain or
another altogether on which they stood.
But fortunately for them, for they were not in the mood or
condition to stand hardship long, they were not destined to remain
long in doubt as to the whereabouts of their companions. While they
were gazing anxiously into the distance Jack’s keen eye suddenly
detected a sharp flash off to the eastward. It was as if the sun had
glinted for an instant on a bit of sharply cut diamond. The flash was
as bright as a sudden ray of fire. The next instant it was seen no
more. But a second later it flashed up again. This time the glitter
was to be seen for a longer interval.
“What on airth is it?” gasped Pete, to whom Jack had indicated
the phenomenon.
“Wait one moment and maybe I can tell you if it is what I hope,”
cried Jack in an excited tone. With burning eyes he watched the
distant point of light flashing and twinkling like a vanishing and
reappearing star.
“Hooray!” he cried suddenly, “it’s all right! It’s Ralph and the rest
and they are all safe. But they don’t know yet where we are.”
 Pete gazed at the boy as if he suspected that the stress of the
night might have turned his mind.
“Anything else you kin see off thar?” he asked sardonically.
“Nothing but that they say the horses are all right, and that if we
see their signals we are to send up a smoke column,” replied Jack
calmly, his countenance all aglow.
“Look hyar, Jack Merrill, I promised your father ter take care of
yer,” said Pete sternly, “an’ I don’t want ter take back a raving
loonertick to him. What’s all this mean?”
“That Ralph is signalling with a bit of mirror,—heliographing, they
call it in the army,” cried Jack, with a merry laugh, which rather
discomfited Pete.
“Wall, that may be, too,” he admitted grudgingly, “thar sun would
catch it and make it flash. But how under ther etarnal stars kin you
tell what he’s saying?”
“Simple enough,” rejoined Jack; “he was making the flashes long
and short,—using the Morse telegraph code, in fact. You know we
had a cadet corps at Stonefell to which we both belonged. Field
signalling and heliographing was part of our camping instruction, but
I guess neither of us ever dreamed it would come in handy in such a
way as this. That certainly was a bully idea of Ralph’s. He knew if we
were any place around we would see the flashes and be able to read
them, whereas we couldn’t have sighted them in the tall brush so
easily and might have missed them altogether.”
“Wall, what air we goin’ ter do now?” asked Pete, rather
apathetically.
“Do? Why, light a fire, of course. Then they’ll see the smoke
column and come over to us with grub and the ponies.”
“Hum,” snorted Pete. “Got any matches?”
“Why, no. Haven’t you?”
“Nary a one.”
“Phew!” whistled Jack. “Now we are in a fix for certain. What can
we do?”
 “Keep your shirt—or what’s left of it—on, son, you’ll need it,” said
Pete slowly, a smile overspreading his sun-bronzed features, “thar’s
more ways of killing cats than choking ’em ter death with superfine
cream. Likewise thar’s more ways of lighting a fire than by using
parlor matches.”
Jack watched Pete wonderingly as he took out his knife in silence
and strode off to the tree. He found a dead branch and whittling off
the wet outside bark soon reached the dry interior. This done, he cut
the wood down to a stick about two feet long and a little thicker
than a stout lead pencil. Then he hacked away at some more of the
dry wood till he had a small flat bit of thoroughly dry timber. In this
he excavated a small hole to fit the point of the pencil-like stick.
“Now git me some dry twigs from that brush yonder,” he directed
Jack, who had been gazing on these preparations with much interest
and a dawning perception of what the old plainsman was going to
do.
By the time Jack was back with the twigs,—the dryest he could
find,—Pete had scraped off a lot of sawdust-like whittlings and piled
them about the hole he had dug out. Then taking the pencil-like
stick between his palms, he inserted its lower end in the hole,
carefully heaped the sawdust stuff about it, and began rotating it
slowly at first and then fast.
All at once a smell of burning wood permeated the air. From the
sawdust a tiny puff of blue smoke rolled up. Suddenly it broke into
flame.
“Now the twigs! Quick!” cried Pete, and as Jack gave him the dry
bits of stick he piled them on the blazing punk-wood, blowing
cautiously at the flame. In ten minutes he had a roaring fire. But the
old plainsman’s work wasn’t finished yet. He began hacking green
branches from the tree and piling them on top of his blaze.
Instantly a pillar of dun-colored, smoke, thick and greasy, rolled
upward into the still air.
Pete took off his leather coat and threw it over the smoking pyre,
smothering the column of vapor.
 “Now then, son,” he said, with the faintest trace of triumph in his
voice, “yer see that this here hell-io-what-you-may-call ’em, ain’t
ther only trick in the plainsman’s bag. By raising and lowering that
coat you kin talk in your Remorse thing as long as you like.”
“Pete, I take off my hat to you,” exclaimed Jack, feeling ashamed
of the rather superior manner he had assumed when talking of the
heliograph a while before.
“That’s all right, son. But take it frum yer Uncle Dudley that we
none of us know everything. Thar’s things you kin larn from an
Injun, jus’ as I larned how ter git that fire a-goin’.”
Kneeling by the smoldering smoke-pile, Jack raised and lowered
the coat at long and short intervals, forming a species of smoke
telegraphy easily readable by anyone who understood the Morse
code.
An hour of anxious waiting followed and then upon the scene
galloped at top speed the rest of the adventurers bearing with them
some food, scanty but welcome, and best of all, the ponies and one
rifle.
 CHAPTER IX.

THE LONE RANCHO.

Well, that was an odd meal, that refection of water-soaked

biscuit and canned corned beef, with flood water as a beverage.
Perhaps in all the adventures of the Border Boys, when in after years
they came to recall them, no scene stood out quite so strikingly.
For one thing, Coyote Pete alone, of the party, possessed any
sort of wardrobe. The professor was clad in his “barber pole”
pajamas. Ralph boasted a shirt and Walt Phelps possessed the same
with the addition of a pair of socks, which latter hardly fulfilled
requirements so far as a covering for his nether limbs was
concerned.
From time to time the Border Boys had to look at each other and
burst out laughing. Only the professor viewed the matter in a serious
light.
“Suppose we should meet some ladies,” he asked indignantly.
“Reckon thar ain’t many of ’em hereabouts,” ventured Coyote,
spreading a big slice of beef on a bit of soggy bread. “The burros is
ther only representatives of the gentle sex fer a good many miles, I
opinion.”
The burros, relieved of their packs, which had been swept away,
wagged their ears appreciatively at this, and continued browsing on
the short, coarse grass which grew in patches here and there, and
which the boys were delighted to see seemed also to be palatable to
the horses.
Ralph and the others had already related how the terrified
animals had been recaptured without difficulty early that day. In
fact, a circumstance which has often been noted was their good
fortune, namely, that panic-stricken horses in lonely, wild countries,
will actually seek human companionship,—provided, of course, that
they have already been domesticated. As for the burros, their loud
“hee-haws” had resounded all night.
Ralph also explained how the idea of the mirror heliograph came
to him. The lad who, as has been explained, was a bit of a dandy,
was horrified to discover the abbreviated state of his wardrobe. But
a search of his shirt pocket revealed his pocket-mirror with its folding
brush and comb fittings. The railroad king’s son had at once set to
work to make himself presentable about the head at least, and was
combing his hair neatly and wondering how Jack and Pete had fared,
when the sun caught the mirror and it flashed blindingly into his
eyes. This gave him the idea of flashing it in all directions in the
hope that the others, if within sight, would catch its glint. Then
came the happy thought of telegraphing with the bit of glass by
alternately covering and uncovering it. The idea had met with the
warm approval of the professor and Walt Phelps, although, perhaps,
even they had not been over sanguine of results.
“Well,” said Jack at length, after the events of the night and the
following incidents had been discussed and re-discussed, “what are
we going to do now?”
“Get clothes,” cried Ralph, without an instant’s hesitation,
regarding his bare legs disparagingly.
“By all means, yes,” agreed the professor.
Coyote Pete grinned.
“Jack,” said he, “will you be so kind as ter step ter the telephone
and tell the Blue Front Store to send up a few samples of men’s
furnishings?”
All but the professor burst into a roar of laughter at this sally.
 “At any rate,” suggested Walt Phelps, “we’re not likely to get held
up.”
“Not so sure about that,” said the professor, “I have the money
belt containing most of our finances around my waist. I always sleep
with it there.”
“Hooray!” shouted the boys, who, up to that moment had not
once thought of the important question of finances. It struck them
now with sobering force.
“By George!” cried Jack, “if it hadn’t been for your foresight,
professor, we might have been penniless as well as wardrobeless.”
“That’s right,” agreed Coyote Pete, “and ther chance that you’d
stand of being helped out by the greasers would be about ther same
as a snowflake ’ud have on a red-hot cook stove.”
“My idea is to lose no time in striking out for a town or village
where we can get some clothes, even if they are only Mexican
garments,” announced Jack.
“And food, too,” put in Walt Phelps, who liked to get his three
meals a day, “we’ll be on starvation diet if we don’t stock up on
that.”
After more discussion it was agreed to follow up the dry bed of
the river, as the professor’s map showed a small village some
distance up a stream which, though unnamed on the map, seemed
to be the one on whose banks they now were. This decision
reached, no time was lost in mounting. There was no saddling to be
done, for the saddles had been swept off with most of the rest of
their outfit.
“If you ever catch me camping in the dry bed of a river again you
are welcome to hang me to a sour apple tree,” grumbled Walt
Phelps, as he mounted.
“I reckon I’m ter blame fer it all,” volunteered Coyote Pete, “but I
never thought as how that far-off storm would affect us in the
plains. That must have bin a jim-dandy of a cloudburst.”
“I’d hate to have been any closer to it than we were,” laughed
Jack. “If we had been, we’d have been going yet, I imagine.”
 “I heard of a cloudburst once that did some good, though,”
struck in Pete; “ther thing happened to a friend of mine in Californy.
He wuz a miner, Jefferson Blunt by name.
“Wall, sir, Jeff had struck such all-fired bad luck up on the
Stanislaus River that he’d about concluded to pull out for other
regions when, all of a sudden, one night up came a storm, and in
the middle of it there come the all-firedest cloudburst that Jeff had
ever heard of. It picked up his cabin and floated Jeff off down the
river, a-going like a blue streak. He thought every minute that he’d
hear Gabriel’s trumpet and see ther golden stairs, but that little old
cabin was well built and watertight, and it floated like a boat.
“It must hev been hours, Jeff says, afore he felt ther thing give a
bump and stop. As soon as he dared he opened ther door and
peeked out. He wuz in a part uv ther country he’d never seen. It
was all cliffs and big trees and very imposing, and ther like of that,—
that ‘imposing’ is Jeff’s word.
“Wall, Jeff he steps out of his sea-going shack and looks about
him, and ther first thing he sees is a big streak of ore just a-glitter
with gold and stuck, like a band of yaller ribbon along ther cliff face
above his head.
“Jeff had bin so unlucky that first he thinks it’s jes’ fool’s gold and
not the real article. But he soon convinces himself thet he’s struck it
rich at last. Wall, ter make a long story short, Jeff files a claim and in
a few y’ars is a rich man, and what d’ye s’pose he called ther mine?”
“‘The Cloud Burst,’ of course!” cried Jack.
“How’d yer guess it?” asked Pete. “But yer right, and thet’s ther
only cloudburst I ever hearn’ of, thet brought anybody any luck.”
“Personally, if I could find a pair of trousers,” wailed the
professor, “I should esteem their possession almost above even such
a lucky discovery as you have related.”
“I think I’d trade it right now for a porter-house steak and
trimmings, brown gravy and green corn, and——”
“See here,” put in Ralph, with assumed indignation, “if you don’t
shut up I’ll, I’ll——”
 “Go right home,” chuckled Walt teasingly; “you’d be a fine sight
in that rig. I’ll bet the folks back east would have you put in the
calaboose.”
But by noon the gay spirits of the boys were considerably toned
down. No sign of a town had yet come in sight and they were all
hot, hungry and tired. The odd procession, with the burros tagging
along behind, looked disconsolate enough as it followed the
windings of the river. The shallow aftermath of the flood steamed
and simmered under the hot sun, sending up unpleasant odors,—yet
they had to drink it or go without.
By way of cheering the party up, Coyote Pete began to sing—or
rather wail—in the high-pitched voice affected by cow-punchers
singing to their cattle:
“O-ho-wa-hay da-own upon the Su-wahanee River,
Fa-har, fa-har a-way——”

But before he could begin the next line Ralph struck in with:
“There’s where our pants are floating ever;
There’s where they’re gone to stay!”

In the general roar of laughter which followed, the “grouch”

which had settled down on the tired wayfarers vanished like the
spring snow under a burst of sunlight.
With a shout the boys, their troubles forgotten in an outburst of
that good nature that makes the whole world kin, plunged forward,
their shirt tails flying.
“Yip-yip-ye-ee!”
The joyous yell filled the air. And then it broke off into a real
cheer, for, on surmounting the summit of a small eminence, they saw
below them, not more than a mile off, a small adobe house of
unusual type, for it had two stories. It was surrounded by a grove of
green willows which delighted the eye tired by the endless gray-
green stretches of grease-wood savannahs.
 Even the dignified professor joined in the enthusiasm, and in a
minute a cavalcade was bearing down on the place at breakneck
speed. As they neared it in a thunder of hoofs and a cloud of yellow
dust, a door opened and the figure of a gaunt Mexican, with long,
shaggy, black hair hanging straight and lank to his shoulders,
stepped out. His next move halted the leaders of the party abruptly.
He jerked a long-barreled rifle to his shoulder and pointed it
threateningly.
“Mira rurales!” he yelled to some one within the house.
“No rurales! Americanos!” cried Coyote Pete.
The effect was magical. The man’s startled air changed, and with
a sheepish smile he stepped forward as Jack and Ralph, who were in
advance, drew rein.
“What did he mean by rurales, I wonder?” asked Ralph of Jack in
a low tone as the others loped up.
“Why, rurales are a species of police. Rangers, they are called
sometimes. They are wild chaps, mostly recruited from the ranks of
brigands and highwaymen. The government pays them a high figure
to be good and keep law and order.”
“But this man seemed to fear them.”
“Maybe he has reason to. But we can’t be particular. At any rate,
we are a strong enough party to look after our own hands. But see,
here comes his wife. I guess, after all, he is nothing more unlawful
than a cattle rancher in a small way, who perhaps, once-in-a-while
takes an unbranded calf or two from his neighbor’s estates.”
The woman who joined the man, who by this time had set down
the rifle, was a stout, slatternly-looking creature in a greasy cotton
wrapper. She shot out a few rapid words in a low voice to the other,
who replied in equally low tones. So far as Jack, who was closest,
could judge, the woman seemed to be protesting against something,
and the man stilling her objections.
Coyote Pete as spokesman now advanced, and in Spanish asked
if they could obtain lodging and refreshment for themselves and
their stock.
 CHAPTER X.

AFTER MIDNIGHT.

To their astonishment, the man seemed to hesitate. They had

judged from the poverty-stricken look of his place and belongings
that he would jump at the chance to make some money easily. But it
seemed that this was not the case.
While the fellow still hesitated, glancing covertly at the
newcomers, the professor did a foolish thing. He exhibited his
money belt and tapping it made it give forth the suggestive jingle of
coins. Coyote Pete’s expression grew angry for a moment, but he
checked his chagrin at the professor’s foolish move.
But the exhibition of the party’s financial solidity seemed to have
decided the ill-favored Mexican and his wife, for after some more
parley, which somehow appeared to Jack to be merely for form’s
sake, they agreed to shelter the party and their stock at two dollars
each, Mexican, which is equivalent to one dollar of our money.
“Cheap enough,” said Jack, as ten minutes later they turned their
stock loose in the corral and watched them attack with wholesome
appetites the hay stack in the center of the enclosure.
“May be dear enough before we get through,” thought Coyote
Pete to himself.
He refrained from mentioning his mistrustful feeling to the
others, however, as, after all, the Mexicans might be honest enough
folks even if his impressions were otherwise.
After a wash-up in a small creek which flowed at the back of the
place, the adventurers were quite ready to sit down to a smoking
meal of frijoles (beans fried with red peppers) and eggs cooked in
the Mexican style. Some thin red wine was served with the meal, but
as none of the party had any use for alcoholic beverages in any
form, they were content to wash it down with water from the great
stone olla,—or water cooler which hung under the broad eaves of
the veranda.
Jack had an uneasy sense that they were being scrutinized as
they ate, by some unseen pair of eyes, and once looking up quickly
he caught, or thought he did, a glimpse of the woman’s print gown
slipping from a shuttered window. Jack was not a boy to make a
mountain out of a mole hill, though, and concluded that, in all
probability, the woman, if she had been looking at them, had been
merely curious at the advent of so many strangers.
The rest of the afternoon, for it was late when they concluded
their meal, was passed in chatting and lounging about under the
trees. Nobody felt inclined for more strenuous occupations. The
professor, however, having obtained some old canvas, succeeded in
fashioning a rough pair of trousers. They were short and shapeless,
and his legs stuck out oddly from them like the drumsticks of a fowl,
but they were better than nothing, he thought. As for the boys, they
had bought some baggy garments of the Mexican type from the lone
rancher, which would have to last them till they reached the nearest
town. This, they were informed, was Santa Anita, and was not more
than ten miles distant.
An early start being determined on, they sought their beds soon
after supper, which consisted of the same fare as the other meal
with the addition of some greasy pancakes. Jack ate some of these,
not caring for a second dose of the peppery beans and a short time
after felt, as he expressed it to himself, “as if a cannon ball were in
his midst.”
Perhaps this accounts for his wakefulness, for he found it
impossible to sleep after they had all turned in, in one large room,—
or, rather, garret,—which formed the second floor. The others flung
themselves on the straw, which served for beds, with the lassitude
of complete exhaustion, but Jack lay awake, with the pancakes on
his chest like a leaden weight. At length he fell into an uneasy
slumber, from which he awakened a short time later with a start and
a queer feeling that something in which they were vitally interested
was going forward.
His first vague feelings rapidly crystallized into more definite
shape as, from the yard outside, he could now distinctly hear the
trampling of horses’ hoofs. There seemed to be several of them, to
judge by the noise.
Moonlight was streaming into the garret through an unglazed
opening in the adobe wall, and holding his watch in the rays, Jack
saw that it was half an hour after midnight.
“Queer time to receive visitors,” he thought to himself.
At the same time he was conscious of an overwhelming curiosity
to ascertain who and what the midnight arrivals could be. The boy
had noticed a door in the wall of the garret when they first entered it
that evening, and from his previous inspection of the exterior of the
house he had formed an idea that it opened upon the top landing of
an outside stairway. They had been conducted to the garret,
however, by a ladder leading from the room below.
As well as he could judge, the noise came from the opposite side
of the house to that on which the door was situated, so there did
not seem to be much chance of detection in slipping out of the door,
down the outside stairway and, from some point of vantage, seeing
what all the racket might portend. There was one possible difficulty
in the way, and that was that the door might be locked. But it
proved to be unlatched, and Jack, swinging it open, after he had
partially dressed, found himself, as he had surmised he would, on a
landing or platform at the top of an outside flight of stairs.
In his bare feet, for he had not paused to put on shoes, he
slipped as noiselessly as possible down the stairway and presently
found himself in the yard. The moonlight cast black and white
patterns of the overhanging willows on the ground, but a brief
inspection convinced Jack that there was no human being astir but
himself on that side of the house.
As he reached the ground he could distinctly hear the voice of
the slatternly woman crying out:—
“Hush!” to the new arrivals.
The voices which had been loud at first were instantly lowered,
and he could hear the riders, whoever they were, addressing
quieting remarks to their horses.
“Well, I’m going to see what all this means, if it’s the last thing I
do,” said Jack to himself, and suiting the action to the word he glided
rapidly along in the shadow of the wall till he reached the corner of
the house. There was a low outbuilding there, which might at one
time have been used as a pigstye. This was just what Jack wanted.
He placed both hands on the top bar of the little enclosure outside
the pen-like erection, and the next instant had vaulted lightly over
and was inside the little shack. The boards of which it was composed
were interspersed by wide cracks, and applying his eye to one of
these the Border Boy commanded a fine view of the moonlit yard at
the end of the house.
As he had expected, it was full of riders, one of whom was
mounted on an animal which somehow seemed familiar to the boy.
He with difficulty suppressed a cry of astonishment, as the next
instant the rider emerged into the moonlight, and Jack saw that he
was none other than Black Ramon. The others, he now recognized
as men he had seen in the camp on that adventurous morning
following the delivery of the warning letter.
But Jack had not much time to meditate on all this, for he
suddenly became aware that Ramon was riding behind the cantle of
his saddle, and that lying across the saddle itself was a human
figure. A second later the boy made out that it was the senseless
form of a woman that the outlaw chief was carrying before him.
Hardly had he made this discovery before the woman and the
man of the lone ranch came forward and lifted the inanimate form
from the back of the black horse of the Border scourge. As they did
so a mantilla of elaborate workmanship which covered her face, fell
from it, disclosing her marble-like features, as pale as death. Jack
then saw that she was young and very beautiful. As the girl was
lifted by the lone rancheros, her consciousness returned, and
opening her eyes she began to pour out a flood of Spanish. Jack,
like most boys bred along the border, had a working knowledge of
the language, and it didn’t take him long to gather that she was
promising rich rewards, estates, anything to her captors if they
would release her and restore her to her parents.
But Ramon’s rejoinder was a hoarse laugh. He informed the girl
that he meant to exact a heavy ransom from her father for her
freedom, and that if it were not forthcoming he would make her his
own wife.
An astonishing change came over the girl at these words. From a
pleading, terror-stricken maiden, she became a fine figure of scorn.
Drawing herself up proudly, she exclaimed with blazing eyes:—
“I would die before such a thing happened. My father will find
you out and punish you like the wicked men you are.”
“Colonel Don Alverado will never find Black Ramon or see his
daughter again if a hundred thousand pesos are not forthcoming
before the end of the week,” was the rejoinder.
In speaking these last words Ramon had unconsciously raised his
voice, and the rancheros, with faces full of alarm, stepped forward.
“Hush! for heaven’s sake not so loud!” the woman exclaimed,
“there are several Gringoes in the house!”
Ramon’s face grew black.
“Gringoes!” he snarled, “what do you mean by admitting the
Yankee pigs when I have paid you well for the use of your house?”
“But they are here only for the night and are sound asleep,”
protested the male ranchero. “Depend on it, they will not interfere.
They are pressing on toward Santa Anita to-morrow at dawn.”
“And then, too, they have a belt full of money, Senor Ramon,”
whined the woman, “there is no reason why your excellent self
should not have it. We had that idea in our head when we consented
to let them stop here.”
“Oh, so that’s the reason you suddenly became willing to let us
stop,” thought Jack in his hiding place.
But Ramon was now leaning forward with a sudden expression of
keen interest.
“These Gringoes, old woman,” he asked, “tell me, are they three
boys, a tough-looking, long-legged man with a yellow moustache,
and a spectacled old man?”
“Si, senor,” was the rejoinder.
“Santa Maria,” exclaimed Ramon, “here is good fortune. It is
those Border Boys and their companions delivered into our hands for
the plucking. You did well to let them stop here, senora. They are all
asleep, you say?”
“Si. It is but a few minutes ago that my man crept up the ladder
and peered into the garret in which they are sleeping. They are all
snoring like the Yankee pigs they are.”
“Bueno. We will attend to them shortly,” was the rejoinder; “but
now to dispose of the girl. Have you a room in which we can confine
her?”
“Yes, in the small room at the other end of the house. It was
formerly used as a wine room and is without windows, except a
small one at the top for ventilation. It has a strong door, too, for
when we grew vines and made wine, thieves used to visit us, ill
fortune light upon them.”
“That’s a queer sort of morality,” thought Jack, “for if I ever saw
or heard of a precious band of rascals, these are surely they. That
poor senorita! We must devise some way of aiding her to escape,
but what can we do? I guess I’ll sneak back now while they are busy
elsewhere and wake up the others, for if I’m not mistaken we are
going to have a tough fight on our hands before very many
minutes.”
As Jack cautiously slipped back by the way he had come, he saw
the senorita being led away into the house, proudly disdaining to
parley further with her captors.
“There’s a girl in a thousand,” thought Jack to himself, “no
hysterics or uproar about her. We’ve just got to help her out of the
clutches of those ruffians.”
 CHAPTER XI.

TRAPPED!

Cautiously awakening his companions one by one, Jack told them

of his adventures while in the pig pen.
“The scoundrels!” exclaimed the professor, “we must act at once.”
“Now hold your horses,” drawled Coyote Pete in the easy tone he
always adopted when danger was near, “it ain’t our move yet. If I
ain’t very much mistaken we’ll have all the action we want in a very
short time, too. As a first step I’d suggest we bar that door yonder,—
the one that Jack sneaked out of—I see it’s got a good big latch on
the inside. In that way we’ll head off an attack frum thar, an’ we’ll
only have the trap door from below to look after.”
The heavy bar being noiselessly placed in its hasps, Pete outlined
his further plans.
“They’ll figger we are asleep,” he said, “but it ain’t likely they’ll
jump us till they’ve sent someone up to make sure. It’s our play then
ter git back on the straw and all snore as natural as possible.”
“What then?” asked Walt Phelps in rather an alarmed tone.
“We’ve only got one rifle.”
“That’s so, consarn it,” grunted Pete, “wall, we’ll hev ter do ther
best we can an’—hush, hyar comes the advance guard now!”
In the room below they could hear cautious footsteps. Evidently
Ramon had lost no time in hatching out his plans.
 “Lie down, everybody, and sham sleep as hard as yer can,”
ordered Pete in a low, tense whisper, “our lives may depend on it.”
The order was obeyed none too soon, for before many seconds
had passed they could hear the creaking of the ladder as someone
mounted it. Presently, from one half-closed eye, Jack perceived a
head poked upward through the trap in the floor. By the light which
streamed up from below he saw that it was the cranium of the red-
headed man whom he was pretty sure was the author of the
warning message which had been carried into their camp.
The man stood still as a statue for perhaps five minutes. During
the tense moments Jack’s heart beat as if it would break through his
ribs. It was not fear, but intense excitement that thrilled him. The
moment was at hand when they would be engaged in a desperate
game against terrible odds. What would be the result?
Having apparently satisfied himself that they all slept soundly, the
scout of the outlaws descended once more, the ladder creaking
under his weight.
“It’s goin’ ter come in a few minutes, now,” whispered Pete,
rousing himself, “gimme the rifle, Walt. How many cartridges is in
it?”
“Five,” was the disheartening reply.
“An’ we ain’t got another one between us,” moaned Pete. “Wall, it
can’t be helped, as the hawk said to ther chicken when he carried
her of, leavin’ her numerous family behind. Now, I’m going ter git
right by this here opening and the first head that pokes through it
gits a crack. We’ll save the cartridges for an emergency.”
“An emergency!” exclaimed Ralph, thinking that if ever there was
an emergency the present situation had already arrived at that
stage.
They could now hear whispers below, and worse still, the
ominous click and slide of repeating rifles being got in readiness for
use.
“There’s some old furniture piled in that corner,” exclaimed Jack
suddenly, “couldn’t we use it to block the trap with?”
 “A good idea when the worst of it comes,” assented Pete, “but
we’ve got ter keep ther trap open so as to disable as many as
possible before we have to come to close quarters.”
The next ten minutes,—for though it seemed like the same
number of hours, it was not in reality any more,—was the most
painful period the boys ever recalled having put in. From the room
below came furtive sounds, but they were so soft and infrequent
that it looked as if the main body must have withdrawn further to
discuss the attack.
“Say, let’s rush them. I can’t stand this any longer.”
It was Ralph who spoke, but Coyote laid a restraining hand on
his arm.
“Easy, lad, easy,” he admonished in a low breath, almost in the
lad’s ear, “it won’t be long before they start tuning up for the
performance, and it ain’t goin’ ter be a funeral march for us neither.”
As he spoke, Pete “clubbed” their solitary rifle, holding it by the
barrel. At the same instant a door beneath quietly opened and
closed, and the next minute the ladder creaked as a foot was placed
upon it.
“Up with you, Miguel,” they heard Ramon whisper, “here’s the
knife. Remember the money belt is on the old man. Jose, you follow
him closely, and Migullo, you come after. That is all it is safe to trust
on the ladder at one time. I myself will come later.”
“The cowardly greaser,” breathed Coyote, with one of his
increasingly frequent lapses into plain English, “I guess he’ll feel less
like climbing than ever when he sees what’s going to happen to the
first arrival. It’s a good thing for us they can’t come but one at a
time. In that way they’ll have no chance of rushing us.”
As he finished speaking the boys felt the peculiar thrill that
comes before the enactment of some exciting deed. A black head
poked itself cautiously through the trap and as it did so Coyote
raised his rifle stock, swung it, and brought it down with crushing
force on the head of the intruding wretch. He fell backward with a
crash, and landed in a heap in the room below. Under ordinary
circumstances, not one of the Border Boys would have stood for
such drastic measures. But they knew that now it was their life or
the Mexican’s. Nevertheless they felt relieved as they heard the
fellow stagger to his feet and begin cursing in picturesque Mexican.
“Diablo! The fiend himself is in those Gringoes,” he raved, “I
think they have broken every bone in my body.”
“More fool you, for not being more cautious,” growled Ramon,
and then, raising his voice, he shouted up in English:
“It will be of no use to you to resist. I have a superior force and
if you injure another of my men when I do get you it will go hard
with you. Surrender and give me the money and no harm will come
to you with the exception of Jack Merrill. I mean to deal with him as
I choose.”
“When you get him, you dog,” shouted Coyote Pete, “which won’t
be yet or for a long time to come,—ah! you would, would you!”
As he spoke, the cow-puncher had projected his head
thoughtlessly over the edge of the trap door. A bullet aimed to kill,
which, however, whizzed harmlessly by his ear, was the result. The
missile sang through the air and buried itself in one of the rafters.
“We’ll give you all you want of that directly,” hailed Coyote Pete,
essaying what is sometimes called “a bluff,” “we have plenty of rifles
and ammunition, and we can use them, too, so bring on your next
man.”
“You shall smart for this, you Gringo pig,” cried Ramon from
below. Evidently the complete failure of his first attack and Coyote’s
bantering tone had driven him beside himself with fury.
“Oh, I’m a smart fellow, anyhow,” chuckled Coyote Pete, “come
on. One cigar for every head I crack. That’s the way they do it at the
county fair with the Jolly Nigger Dodger, and I don’t know as you
greasers have anything on him.”
“Rush up and bring them down out of that!” screamed Ramon
furiously. But the sharp lesson they had just had seemed to hold the
Mexicans in check. Evidently the Gringoes above were not to be
trifled with. Ramon strode up and down the room stamping and
raging and biting his nails. Altogether he was in a fit of black Latin
rage which is not so very different from the tantrums we
occasionally find in our own nurseries.
“Why not come up yourself, Ramon?” was Coyote’s next thrust.
“If your head is burning with such blazing thoughts it must need
ventilating.”
But the Mexican, wisely enough perhaps, did not reply. Instead,
he called down the men from the ladder, seeing, in spite of his rage,
that it was useless to waste his followers in that fashion.
“We’d better bottle up the trap door now,” said Pete, as the
voices below became more inaudible. “Get that old furniture, boys,
and we’ll make things snug.”
“Here’s an old table top that might fit over the hole,” said Jack,
bringing the article in question, “it’ll just fit too, and it’s solid
mahogany.”
“Just the thing, boy. Now quickly bring all the stuff you can to
pile on it.”
“Say, there’s a pile of big stones over here where the chimney
goes through,” reported Ralph presently, “how would those do for
weights?”
“Fine. Bring them right along. Your Uncle Dudley will pile them.”
One would have said from the cow-puncher’s boisterous spirits
that he was in perfect security instead of a situation the danger of
which he, perhaps, more fully realized than any of his companions,
comparatively inexperienced as they were.
One by one the lads carried the big stones over and they were
piled on the table top.
“That will do,” said Coyote at length, “they’ll never get that up
unless they use dynamite.”
“What do you suppose they’ll do now?” wondered Jack as, the
work over, they sat down about the newly covered hole.
“Try rushing that back door, most likely. Suppose you take a peek
out of the window. It gives a view of the steps and it’s too small for
the varmint ter git through.”
 The small aperture, mentioned before, was quite high up in the
wall, but, hoisted up by Ralph and Walt, Jack was able to rest his
elbows on the sill and peer out. He did so cautiously, which was just
as well, for, as the astute cow-puncher had surmised, the next attack
must come from the back door. So much was evidenced by a view of
the steps which were covered with dark forms advancing stealthily.
“We’ll give ’em another surprise party,” announced Pete when he
had heard his young lieutenant’s report. “Jack, take the rifle while I
guard the trap. There’s a chance they may try to rush the two places
at once. Aim through the keyhole, and when you think it time to, let
’em have it. Don’t be scared of hurting them. Remember it’s our lives
or theirs.”
Feeling a bit squeamish, but far too good a soldier to attempt to
disobey orders, or even question them, Jack did as he was directed.
Placing the muzzle of the rifle to the keyhole he waited with beating
heart the first signal that their enemies had ascended the stairway
and were actually on the balcony outside the door.
He had not long to wait. Presently there came a scuffling,
scratching sound without, as the Mexicans fumbled about the door,
evidently feeling for a latch of some sort. With a hasty prayer that he
might not inflict a mortal wound, Jack awaited the right moment, as
he judged it, and fired.
There was instantly a loud yell of pain from without.
“Good for you, boy,” grunted old Pete grimly “you brung him
down.”
 CHAPTER XII.

THE GRINGOES MOVE.

From without the door there now came shouts of baffled rage.
The Mexicans were finding out, as their kind has done before, that a
party of brave Americans is more than a match for twice their
number in a fight. Moreover, thanks mainly to Jack’s presence of
mind in slipping out of the house and performing scout work, our
party was strongly entrenched. The door was stout, and the iron bar
within solid. There was no apparent way of forcing an entrance by
battering it down, for the landing was too small to use a “ram”
effectually.
“Hooray, we’ve got ’em beaten!” cried Ralph thoughtlessly.
Coyote flashed a scornful eye on him.
“Beaten!” he scoffed, “we ain’t got ’em beaten till we’re out of
this place and miles on our way. Why, if they kain’t do anything else
they kin starve us out if they want to.”
“That’s so,” assented Ralph sorrowfully, and then with a violent
twist of spirits, “I guess we’re goners.”
“There, go galloping off the reservation agin,” struck in Pete; “we
ain’t goners yit by a long shot, but we’ve got a powerful lot of work
afore us, as the government said when they tackled digging that
Panama Canal.”
 All now became silent once more, or at least the boys could hear
nothing. Evidently the Mexicans had withdrawn for a council of war.
“This time they’ll be in dead earnest,” opined the cow-puncher,
“so keep a smart eye open for ’em everywhere.”
Hanging breathlessly on the least sound, the besieged party
waited for the first sign of the coming attack. It was a long time in
making itself manifest, and when it did, it was for a moment
puzzling enough. It came in the form of a noise from above.
“Somebody’s on the roof!” exclaimed Pete. “The foxy varmints! I
wonder they didn’t think of that before.”
The roof of the lonely rancho was flat, and soon they could hear
several footsteps on it as their besiegers paced about.
“What are they going to do?” asked Ralph in a puzzled tone.
“Not hard to guess,” rejoined the professor, “cut a hole in it, I
guess, and then they’ll have us completely at their mercy.”
“If we let them,” said Jack, “but why not try to escape by the
trap, while they are busy on the roof?”
“That might be a good idea if it warn’t likely that they have the
foot of the ladder guarded, or most probably have taken it down,”
said Coyote Pete; “no, you’ll have to guess agin, Jack. Think uv
something new and original.”
“I might say try that door, but I guess that’s guarded, too.”
“Not a doubt of it,” was the reply.
“Tell you what we’ll do,” exclaimed Jack suddenly, struck with an
inspiration, “we’ll try the walls. There may be a secret passage or a
concealed window in them some place.”
The cow-puncher laughed.
“This ain’t a story book, son, and I never heard of such things
outside of one. Lady Gwendolens in real life come out by the fire
escape more often than by the old secret passage or the haunted
wing.”
Undismayed, however, Jack set about his task. He was in the
midst of it, and had met with no success,—not that he had seriously
hoped for any,—when a sudden sound pierced the darkened garret.
The noise was that of axes cutting into the roof.
As Jack listened a slight shudder ran through him. From that
point of vantage the outlaws could shoot them down as they wished,
and there would not be much chance of using their four remaining
shots in return. By this time Jack had reached the spot by the big
stone chimney from which they had taken the stone used to weight
the table above the trap door.
With a rather vague idea of using some more of the stones as
weapons, he started pulling down the remaining loose ones. He had
been at this work but a few minutes when he gave a sudden cry of
triumph.
“Look! Boys! Look here!” he cried, amazedly.
They scurried to his side to find him pointing into a black,
yawning mouth, evidently intended originally for a fireplace but left
unfinished, as the stones they had used now testified.
“It’s big enough to swallow a horse almost,” cried Ralph.
“It’s big enough to save our lives, maybe,” grunted Pete, “but
maybe it’s only a blind lead, and may come out nowhere. In that
case a fellow at the bottom of a well would be better off than the
chap in there, for ther’d be no way of gitting out uv that chimney
once you got in, and—Jumping Jupiter! Come back, boy!”
But it was too late. While Coyote Pete had been talking, Jack had
slipped into the fireplace, and clutching the rough sides of the
chimney had taken the daring drop.
The others listened above in breathless anxiety, and then, to
their infinite relief, a voice trickled up to them from the depths.
“It’s all right, boys! Come on, but take it easy, for I knocked all
the skin off my shins in my hurry.”
The blows on the roof were by this time becoming louder, and
they could distinctly hear the sound of splintering wood as the axe
blades cut into it.
“They’ll hev pecked through that in ten minutes, now,” said Pete,
getting over to one side of the fireplace, “come on, boys. Be on your
way.”
But the boys insisted on the professor going first, now that they
knew the drop was safe enough. Not without misgivings, to which he
was too brave to give utterance. Professor Wintergreen, scientist and
writer, cast himself into that black hole in the garret of the lonely
rancho. An instant later, after a prodigious scraping and bumping,
word came up that he, too, was safe. Ralph and Walt came next, the
former softly humming:—
“I don’t know where I’m goin’, but I’m on my way.”
Coyote Pete came last; and now we shall follow the party, leaving
the Mexicans still hacking away at the roof. It is a trip worth taking,
too, for at the bottom of the chimney an astonishing condition of
things prevailed.
The smoke duct led not into a cellar or into a blind hole, but
instead, Jack, on alighting, had found himself, soot covered and
scratched and torn, in a large open fireplace in a small room. As he
made his sensational entrance there was a sudden sharp scream
from a corner of the room and a female figure clad in white sprang
up.
For an instant a dreadful fear that he had alighted in some sort
of a trap flashed into Jack’s mind. But the next instant he realized
that the alarmed girl was none other than the senorita, and that the
room into which he had fallen was the one selected as her prison.
“Hush, senorita!” exclaimed the boy, as soon as he had given the
signal to his comrades above that all was well, “do not fear me. I am
not one of your enemies but a friend, an American. My companions
are with me,—er—er—that is, they will be.”
“Oh, senor!” cried the girl in English, “what a dreadful fright you
gave me. You—you, if you will excuse me, you are so black. I
suppose it’s the soot in the chimney.”
Jack could hardly refrain from smiling, as, for the first time, he
bethought himself of the alarming figure he must present.
“I’m not as black as I’m painted, senorita, really, I’m not. Nor are
these two new arrivals chimney sweeps, but young American
gentlemen,” he added with a sweeping bow, as Walt Phelps and
Ralph popped out of the chimney. “Allow me to present myself. I am
Jack Merrill, and these are my friends, Walt Phelps, of New Mexico,
and Ralph Stetson, of New York. Not forgetting,” he added merrily,
as the professor straightened up from an instinctive brushing of his
clothes, “our instructor and—er—er—chaperone, Professor
Wintergreen, of Stonefell College, and,” as the other member of the
party appeared, “Mister Peter de Peyster, of the Merrill Ranch.”
“At your service, miss,” said Coyote Pete with a low, sweeping
bow and a deep flourish of his sombrero, to which even in his fall he
had clung.
“Oh, I feel safer now,” cried the girl delightedly, “but,” and she
clasped her hands, “Madre de Dios, what I have passed through! I
was summoned to my garden this evening by a decoy message, that
one of the good sisters at the convent wished to see me. I had
hardly set foot on the path when I was seized and carried off!”
“The rest of your story we know, senorita,” said Jack earnestly.
“You know it?” repeated the girl in an amazed tone, “but, senor, I
do not understand.”
“I will explain later,” said Jack, “at least, we all hope to have the
pleasure of doing so. I may add that I overheard the ruffians, your
captors, discussing the matter while I was hiding in a pig pen.”
The senorita’s large dark eyes grew larger than ever at this. She
began to think Jack a very peculiar young person to come sliding
down chimneys into rooms and to choose to eavesdrop on brigands
from pig pens. But she made no comment, and the talk at once
turned to the subject of escape.
The door of the room was of oak, barred and bolted on the
outside, and impregnable. But the window, high up in the wall
though it was, appeared to be just about large enough to squeeze
through, ample enough even for Coyote Pete, who was the largest of
the party.
“Reckon we can reach it by putting this chair on that table
yonder,” declared Pete, “but we’ll have ter look slippy, for those
Welcome to Our Bookstore - The Ultimate Destination for Book Lovers
Are you passionate about books and eager to explore new worlds of
knowledge? At our website, we offer a vast collection of books that
cater to every interest and age group. From classic literature to
specialized publications, self-help books, and children’s stories, we
have it all! Each book is a gateway to new adventures, helping you
expand your knowledge and nourish your soul
Experience Convenient and Enjoyable Book Shopping Our website is more
than just an online bookstore—it’s a bridge connecting readers to the
timeless values of culture and wisdom. With a sleek and user-friendly
interface and a smart search system, you can find your favorite books
quickly and easily. Enjoy special promotions, fast home delivery, and
a seamless shopping experience that saves you time and enhances your
love for reading.
Let us accompany you on the journey of exploring knowledge and
personal growth!

ebookball.com