More Monthly Problem

Gems
More Monthly Problem Gems is a sequel to Monthly Problem Gems (CRC Press,
2021). This book covers a broader range of math problems. In addition to analysis
problems, problems from number theory, combinatorics, algebra, and geometry
are included.

The book offers problems to promote creative techniques for problem-solving

and undergraduate research. Each problem is selected for its natural charm, the
connection with an authentic mathematical experience, originating from the in-
genious work of professionals, and ready developments, all into well-shaped re-
sults of broader interest.

Each problem provides either a novel application of a familiar theorem or a lively

discussion of multiple solutions. Special attention is paid to informal exploration
of the essential assumptions, suggestive heuristic considerations, and roots of the
motivations of the problem.

This text then presents a new type of problem-solving. It will challenge and stimu-
late math problem-solvers at varying degrees of proficiency. Since the selected
problem gems contain sophisticated ideas and connect to important current re-
search, this book is also geared toward graduate students in math and engineering.

Many of the problems in this book were originally offered in The American Math‑
ematical Monthly.

Hongwei Chen earned his PhD from North Carolina State University in 1991.
He is currently a professor of mathematics at Christopher Newport University.
He has published more than 60 research papers in analysis and partial differential
equations. Dr. Chen also authored Monthly Problem Gems and Classical Analy‑
sis: An Approach through Problems published by CRC Press and Excursions in
Classical Analysis published by the Mathematical Association of America.

Roberto Tauraso holds a PhD in mathematics from Scuola Normale Superiore,

Pisa, Italy. He is currently a professor of mathematical analysis at Tor Vergata
University of Rome. His research interests include complex analysis, number
theory, and combinatorics. He has published over 60 research papers in various
mathematical journals.
More Monthly Problem
Gems

Hongwei Chen and Roberto Tauraso

Designed cover image: Roberto Tauraso
First edition published 2026
by CRC Press
2385 NW Executive Center Drive, Suite 320, Boca Raton, FL 33431
and by CRC Press
4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

CRC Press is an imprint of Taylor & Francis Group, LLC

© 2026 Hongwei Chen and Roberto Tauraso

Reasonable efforts have been made to publish reliable data and information, but the author and pub-
lisher cannot assume responsibility for the validity of all materials or the consequences of their use.
The authors and publishers have attempted to trace the copyright holders of all material reproduced
in this publication and apologize to copyright holders if permission to publish in this form has not
been obtained. If any copyright material has not been acknowledged, please write and let us know so
we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information stor-
age or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, access www.copyright.com or
contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-
750-8400. For works that are not available on CCC, please contact [email protected]

Trademark notice: Product or corporate names may be trademarks or registered trademarks and are
used only for identification and explanation without intent to infringe.

Library of Congress Cataloging‑in‑Publication Data

Names: Chen, Hongwei, author. | Tauraso, Roberto, author.
Title: More monthly problem gems / Hongwei Chen and Roberto Tauraso.
Description: First edition. | Boca Raton, FL ; Abingdon, Oxon : CRC Press,
2026. | Includes bibliographical references and index.
Identifiers: LCCN 2025020685 (print) | LCCN 2025020686 (ebook) | ISBN
9781041000150 (hardback) | ISBN 9781041003212 (paperback) | ISBN
9781003607809 (ebook)
Subjects: LCSH: Mathematical analysis--Problems, exercises, etc.
Classification: LCC QA301 .C437 2026 (print) | LCC QA301 (ebook) | DDC
510--dc23/eng/20250625
LC record available at https://lccn.loc.gov/2025020685
LC ebook record available at https://lccn.loc.gov/2025020686

ISBN: 978-1-041-00015-0 (hbk)

ISBN: 978-1-041-00321-2 (pbk)
ISBN: 978-1-003-60780-9 (ebk)

DOI: 10.1201/9781003607809

Typeset in CMR10 font

by KnowledgeWorks Global Ltd.

Publisher’s note: This book has been prepared from camera-ready copy provided by the authors.
Contents

Preface ix

1 Analysis 1
1.1 Rolling ellipse . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 A limit involving a geometric mean of roots of factorials . . . . 5
1.3 A Lobachevsky-type integral . . . . . . . . . . . . . . . . . . . 10
1.4 A quadratic series with the tail of ζ(2) . . . . . . . . . . . . . . 16
1.5 A trigonometric logarithmic integral . . . . . . . . . . . . . . . 21
1.6 An interesting series with a product of two central binomial
coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.7 A log-tangent integral . . . . . . . . . . . . . . . . . . . . . . . 30
1.8 An arccosine integral . . . . . . . . . . . . . . . . . . . . . . . . 34
1.9 Two norms involving function moments . . . . . . . . . . . . . 36
1.10 Reciprocal power sums of real roots for a complex function . . 41
1.11 A monotonic and convex quadratic spline . . . . . . . . . . . . 46
1.12 An estimate of a series involving the absolute value of sine . . . 51
1.13 An integral with many solutions . . . . . . . . . . . . . . . . . 56
1.14 An inequality on the Wallis integrals . . . . . . . . . . . . . . . 61
1.15 A sequence defined by inequalities . . . . . . . . . . . . . . . . 64

2 Identities 69
2.1 Two identities concerning Fibonacci and Lucas numbers . . . . 69
2.2 Summing Catalan numbers . . . . . . . . . . . . . . . . . . . . 74
2.3 Perfect squares from an arithmetic function . . . . . . . . . . . 78
2.4 Identity related to the parity constant . . . . . . . . . . . . . . 82
2.5 Wrestle with Jacobi’s triple product . . . . . . . . . . . . . . . 86
2.6 A Gaussian q-binomial identity . . . . . . . . . . . . . . . . . . 90
2.7 A rational function identity . . . . . . . . . . . . . . . . . . . . 93
2.8 An identity associated with the zeros of z n + 1 . . . . . . . . . 95
2.9 An identity struck by the elementary symmetric functions . . . 100
2.10 A sum of products of truncated binomial expansions . . . . . . 103
2.11 An identity with alternating weighted binomial coefficients . . . 106
2.12 An identity with the generalized binomial coefficients . . . . . . 110

v
vi Contents

3 Geometry 115
3.1 A line perpendicular to the Newton line . . . . . . . . . . . . . 115
3.2 About two cyclic quadrilaterals . . . . . . . . . . . . . . . . . . 118
3.3 Concurrency of three lines . . . . . . . . . . . . . . . . . . . . . 123
3.4 A triangle inscribed in another triangle . . . . . . . . . . . . . . 126
3.5 A surprising bisection . . . . . . . . . . . . . . . . . . . . . . . 128
3.6 A tangent to a circle . . . . . . . . . . . . . . . . . . . . . . . . 130
3.7 Collinearity of three points . . . . . . . . . . . . . . . . . . . . 133
3.8 Another property of the Nagel point . . . . . . . . . . . . . . . 136
3.9 An inequality involving the Fermat point of a triangle . . . . . 139
3.10 Supplementary pairs of Heronian triangles . . . . . . . . . . . . 142

4 Combinatorics 145
4.1 An equality for integer partitions . . . . . . . . . . . . . . . . . 145
4.2 A recurrence involving integer partitions . . . . . . . . . . . . . 148
4.3 Another equality for integer partitions . . . . . . . . . . . . . . 151
4.4 Tiling a board with cuts . . . . . . . . . . . . . . . . . . . . . . 154
4.5 About the tilings of a 2 × n strip with squares, dominoes, and
trominoes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
4.6 Fault-free tilings of 3-by-n strip . . . . . . . . . . . . . . . . . . 161
4.7 A sum involving a p-root of unity . . . . . . . . . . . . . . . . . 165
4.8 Sparse binary representation of an integer . . . . . . . . . . . . 168
4.9 Subsets with equal sums of powers . . . . . . . . . . . . . . . . 171
4.10 Equality of two sums of reciprocals . . . . . . . . . . . . . . . . 174
4.11 A monochromatic pentagons with given area . . . . . . . . . . 178
4.12 Coloring a graph . . . . . . . . . . . . . . . . . . . . . . . . . . 180
4.13 Arranging coins along a line . . . . . . . . . . . . . . . . . . . . 182
4.14 Averaging the number of fixed points of permutations . . . . . 186
4.15 A one-sided inverse involving Motzkin numbers . . . . . . . . . 189

5 Number Theory 193

5.1 Divisibility of a central binomial sum . . . . . . . . . . . . . . . 193
5.2 A congruence implied by Wolstenholme’s theorem . . . . . . . 197
5.3 A congruence of a multiple sum involving Fibonacci numbers . 201
5.4 A congruence for a product of quadratic forms . . . . . . . . . 204
5.5 Powers of a prime dividing a product of binomial coefficients . 208
5.6 Divisibility of coefficients of powers of polynomials . . . . . . . 211
5.7 A Diophantine equation with powers . . . . . . . . . . . . . . . 213
5.8 Multiples with small digits . . . . . . . . . . . . . . . . . . . . . 216
5.9 Another property of the taxicab number 1729 . . . . . . . . . . 219
5.10 An application of prime density . . . . . . . . . . . . . . . . . . 222
5.11 The asymptotic behavior of a sum . . . . . . . . . . . . . . . . 225
5.12 A family of sums with logarithmic powers . . . . . . . . . . . . 227
5.13 Another application of Bertrand’s postulate . . . . . . . . . . . 231
5.14 A property of the product of consecutive primes . . . . . . . . 234
Contents vii

5.15 A congruence for the integer part of a power of a cosine . . . . 236

5.16 A variant of the Collatz map . . . . . . . . . . . . . . . . . . . 240

6 Potpourri 243
6.1 Fixed point of the distance to the boundary . . . . . . . . . . . 243
6.2 A variant alternating series with the floor function . . . . . . . 246
6.3 A real analytic function with no zeros . . . . . . . . . . . . . . 250
6.4 Finding a special half-line . . . . . . . . . . . . . . . . . . . . . 254
6.5 Squares of palindromes . . . . . . . . . . . . . . . . . . . . . . . 256
6.6 Counting some strange mappings . . . . . . . . . . . . . . . . . 259
6.7 Comparing the coefficients of two generating functions . . . . . 262
6.8 A vector sum of modulus at least 1 . . . . . . . . . . . . . . . . 267
6.9 Counting equilateral triangles in hypercubes . . . . . . . . . . . 270
6.10 Counting rectangles with prime area . . . . . . . . . . . . . . . 274
6.11 Expected number of throws of an n-sided die . . . . . . . . . . 278
6.12 Removing tiles . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
6.13 Writing a Gaussian integer as sum of powers of 1 + i . . . . . . 287
6.14 Enumerating the positive rationals . . . . . . . . . . . . . . . . 291

List of Problems 295

Bibliography 301

Index 309
Preface

This book, a much-anticipated sequel to the first author’s Monthly Problem

Gems (MPG), features a new collection of 82 monthly problems and solutions
in areas such as analysis, geometry, combinatorics, and number theory. Since
2017, both authors have started exchanging monthly problem solutions and
shared enjoyment of problem-solving, and developed a mutual respect, which
led to the realization of this book.
We will adopt the same format as MPG:
1. State the numbered monthly problem with its proposer(s).
2. Address the heuristic consideration of the solution. A careful discussion
of the challenges and related mathematical ideas needed to solve the
problem, focusing on mathematical insight and intuition to make sure
the reader is on board.
3. Present one or more solutions based on the heuristics. Most are either
our published solution in the Monthly or a solution we submitted to
the Monthly, which differs from the featured one, aiming to train and
develop typical inventive approaches.
4. Reveal some possible generalizations of the problem to promote further
research. The materials vary in their degree of rigor and sophistication
to accommodate readers with diverse interests. Each problem comes
with a short, carefully selected list of additional problems for the most
engaged readers. Some extensions may encourage deeper explorations
and lead to valuable mathematics.
We hope, with some confidence, that this book conveys our deep appreciation
for the Monthly problems.
This book is organized into six chapters as follows:
Chapter 1. Analysis. This chapter collects 15 monthly problems in analysis.
These problems are concerned with roulette curves, special types of in-
tegrals, interesting series associated with famous numbers, power sums
of real roots of a complex function, extrema of functional moments, and
monotonic convex splines.

ix
x Preface

Chapter 2. Identities. We prove 12 identities which range from analysis

and number theory to combinatorics and complex variables. Among the
approaches used in the proofs, we will meet Wilf’s “snake oil method”,
Euler’s number-theoretic functions, q-series, integer partitions, elemen-
tary symmetric functions, Jacobi’s triple product identity, formal power
series, and differential operators.
Chapter 3. Geometry. Here the reader finds 10 problems exploring different
results in Euclidean geometry, such as Anne’s theorem, Napoleon’s theo-
rem, Heron’s formula, Brianchon’s theorem, and Newton’s quadrilateral
theorem. Also, several notable points of a triangle, including the Nagel
point and Fermat point, show some unexpected properties.
Chapter 4. Combinatorics. We gathered 15 problems spanning topics from
integer partitions and tilings to applications of the pigeonhole principle.
Various methods have been used to unravel the interactions between
combinatorics and other branches of mathematics, especially number
theory. A particular focus is placed on recursions, generating functions,
enumeration, arithmetics properties of combinatorial quantities, permu-
tations, and graph theory.
Chapter 5. Number Theory. This chapter presents 16 monthly problems,
which concentrate on the central area of number theory. Questions
here run into divisibility, multiplication functions, congruence, quadratic
forms, quadratic residues, and sums of squares. During the journey,
we will revisit many important results such as Euler’s criterion, Gauss
quadratic reciprocity law, Legendre formula, Lucas theorem, and Wol-
stenholme’s theorem.
Chapter 6. Potpourri. We selected 14 problems, which range from geome-
try and number theory to combinatorics and probability. The solutions
touch on various topics, including asymptotic estimates of sums involv-
ing the floor function, characterization of palindrome numbers, test for
existence of real zeros, enumerations of rationals, algorithmic puzzles,
and Markov chains.
We try to present solutions in a coherent and captivating style, including
some historical accounts, putting the problems in a broader mathematical con-
text, and stimulating further inquiry. Some problems have concise solutions,
while others may be quite complicated. Occasionally, Mathematica will help
with solutions. Some problems are even related to the Riemann hypothesis
and “3n + 1 conjecture”, affirming once again the vitality of monthly problems
as a catalyst for research activity. In several cases, a solution requires a result
that might not be well-known, so we provided all the details. For problems
related to the authors’ research, we have given either their extensions or ref-
erences for further study. We hope this motivates undergraduates to find out
Preface xi

“what comes after calculus and linear algebra” and also drives professional
mathematicians in their everyday work.
Most of the ideas involved in the solutions should be comprehensible to
the good undergraduate math student. The chapters and sections are laid
out independently, so feel free to jump into any section and sample whatever
tempts you. We are excited and delighted to work with monthly problems.
We hope there is much here to please you. As problems in each new issue of
the Monthly arrive, the reader may regard this book as a starter set, acquire
a jumping-off point to new ideas, and extend one’s profile of problems and
solutions.
We are pleased to acknowledge those individuals who made this book pos-
sible. It couldn’t have happened without the encouragement of Senior Editor
Robert Ross of CRC Press and the valuable feedback from the anonymous
reviewers. Thanks to the CRC Press staff for their help in making this book
a reality and for all the assistance they offered. We express our deep appreci-
ation to all individuals who contributed to the Monthly problems, from whom
we have learned so much. The exceptional cover artwork by Adriano Tauraso
has given this book its unique character, and for that, we are truly grateful.
The first author also offers a special thanks to Professor Brian Bradie, who
has carefully read the part of the earlier draft of the manuscript and corrected
many errors. Lastly, we are indebted to our families for their love and constant
support.
Naturally, all possible errors are our own responsibility. Comments, cor-
rections, and suggestions from readers are always welcome. Please email at
[email protected] or [email protected]. Thank you in advance.
Chapter 1

Analysis

In this chapter, we select 15 Monthly problems for analysis. These problems

are related to roulette curves, special types of integrals, interesting series asso-
ciated with famous numbers, power sums of real roots of a complex function,
extrema of functional moments, and monotonic convex splines. We try to
present solutions in a cohesive and engaging manner, including some histori-
cal accounts of ideas and contexts. Some problems may touch upon current
research topics. While some have concise solutions, others can be quite com-
plicated. Occasionally, solutions are provided with the help of Mathematica.
In several cases, a solution relies on a less well-known result, so we have in-
cluded full details for clarity.

1.1 Rolling ellipse

Problem 12476 (Proposed by G. Fera, 131(6), 2024). Let C be one arch of
the elliptic cycloid generated by the ellipse x2 + 14 (y − 2)2 = 1. That is, let
C be the curve traced by the vertex at the origin as the ellipse rolls without
slipping along the x-axis for one revolution (see Figure 1.1). What is the area
under C and above the x-axis?
Discussion.
The statement of the problem naturally suggests we revisit the cycloid, the
path of one fixed point on the circumference of a circle as it rolls along a
straight line with a constant angular velocity.
If we place the circle of radius a with the center initially at (0, a) in the
x-y plane with the origin O, then the center moves to C = (at, a) after the
−−→ −−→ −−→
circle has turned t radian. Let P = (x(t), y(t)). Then r = OP = OC − P C,
and it follows that the parametric equation of the cycloid is given by

r(t) = (x(t), y(t)) = (at, a) − (a sin t, a cos t) = (at − a sin t, a − a cos t).

DOI: 10.1201/9781003607809-1 1
2 Analysis

Figure 1.1: Elliptic cycloid with rolling ellipse

y
(x(t), y(t))

(at, a)
t

Figure 1.2: Cycloid with rolling circle

Here we see that the position point P relative to the center is given by
(d sin t, d cos t), where d is the distance between the point P and the center.
Moreover, x′ (t) = y(t).
In general, when one curve C rolls on another curve without slipping, the
path of one fixed point on C is called a roulette. In 2011, F. Kuczmarski proved
that x′ (t) = y(t) is not a coincidence for the cycloid only. Indeed, in [62], he
established the following general Roulette Lemma: Let (x, y) = (f (t), g(t)) be
a parametric equation of a roulette rolling on the x-axis. Then f ′ (t) = g(t).
Thus, based on the above observations, for the proposed problem, it suffices
to find the y-coordinate for the center of the rolling ellipse.
Solution.
We show a more general result: If the ellipse is given by x2 /a2 +(y −b)2 /b2 = 1
with a < b, then the area under C and above the x-axis is (a2 + 2b2 )π. Hence,
the answer to the proposed problem is 9π.
To this end, we first derive a parametric equation for C. Let θ be the
angle between the y-axis and the major axis of the rolling ellipse. To simplify,
Rolling ellipse 3
η

G θ
ξ
θ

Figure 1.3: Parameterizing the center of rolling ellipse

we rotate the entire diagram so that the ellipse’s major axis becomes vertical,
and introduce a new Cartesian coordinate system (ξ, η) with the origin at the
ellipse’s center G, as shown in Figure 1.3.
The equation of the ellipse now is
ξ2 η2
2
+ 2 = 1. (1.1)
a b
Suppose the ellipse is tangent to the x-axis at P = (ξP , ηP ). Let the in-
tersection point of the vertical line from G to the x-axis be Q. Then Q =
(y sin θ, −y cos θ), where y = |GQ|. In the ξ-η plane, the slope of the x-axis is
tan θ. By the slope-intercept form, the equation of the x-axis is
y
η = ξ tan θ − .
cos θ
Plugging this into (1.1) yields a quadratic equation in ξ:
(a2 sin2 θ + b2 cos2 θ)ξ 2 − (2a2 y sin θ)ξ + a2 (y 2 − b2 cos2 θ) = 0. (1.2)
Since the tangency at P implies that (1.2) must have double roots, the dis-
criminant of (1.2) yields
p
y = a2 sin2 θ + b2 cos2 θ,
which is the y-coordinate of the center of the ellipse in the x-y plane. Hence,
the y-coordinate of the vertex of the ellipse in the x-y plane is y − b cos θ.
4 Analysis

Applying the Roulette Lemma, we find the parametric equation of C as

Z θ !
(x, y) = y(t) dt, y(θ) .
0

where p
y(θ) = a2 sin2 θ + b2 cos2 θ − b cos θ.
Hence, the area under C and above the x-axis is
Z 2π Z 2π p 2
y dx = a2 sin2 θ + b2 cos2 θ − b cos θ dθ
0 0
Z 2π  p 
= a2 sin2 θ + 2b2 cos2 θ − 2b cos θ a2 sin2 θ + b2 cos2 θ dθ
0
Z 2π q
= (a2 + 2b2 )π − 2b cos θ b2 − (b2 − a2 ) sin2 θ dθ = (a2 + 2b2 )π.
0


Remark. It would be interesting to find the x-coordinate of the parametric
equation for the center of the rolling ellipse directly without using the Roulette
Lemma. In terms of Figure 1.3, we have
x(θ) = the arclength of the ellipse from 0 to θ + |P Q|.
This will force us to grind out the elliptical integral.
The idea of roulettes was first introduced by W. H. Besant in 1870 and
later popularized by S. Wagon and his square-wheeled tricycle (see [55] and
the more recent [58]). They have used a geometrical description: the road as
a roulette curve and the wheel as a pedal curve. For example, the catenary,
which is usually given as the solution to a hanging flexible cable, now is
the locus of the focus of a parabola which rolls on the x-axis. Moreover, J.
Steiner established two theorems relating the area and arclength of a roulette
to those of a corresponding pedal. For more details, please refer again to F.
Kuczmarski’s paper [62].
Additional problems for practice.
1. Let the parabola y = x2 /4 roll without slipping along the x-axis. Show
the parametric equation of the roulette traced by the focus is given by
(x, y) = (ln(sec θ + tan θ), sec θ).

2. Find the locus of the vertex of the parabola y = x2 as it rolls along the
x-axis.
3. Putnam 1974-A5. Consider the two mutually tangent parabolas y = x2
and y = −x2 . [These have foci at (0, 1/4) and (0, −1/4)), and directrices
y = −1/4 and y = 1/4, respectively.] The upper parabola rolls without
slipping around the fixed lower parabola. Find the locus of the focus of
the moving parabola.
A limit involving a geometric mean of roots of factorials 5

4. Refer to Figure 1.3, show that

|a2 − b2 | sin 2θ
|P Q| = p .
2 a2 sin2 θ + b2 cos2 θ

5. Problem 10254 (Proposed by E. Ehrhart, 99(8), 1992). The curve traced

out by a fixed point of a closed convex curve as that curve rolls without
slipping along a second curve will be called a “roulette”. Let S be the
area of one arch of a roulette traced out by an ellipse of area s rolling
on a straight line. Prove or disprove that S ≥ 3s, with equality only if
the ellipse is a circle.
The featured solution is due to M. Klamkin, a prolific proposer and
editor of professionally-challenging mathematical problems. His solution
cited the Steiner theorems that appeared in [103, p. 201-203].

1.2 A limit involving a geometric mean of roots of

factorials
Problem 12360 (Proposed by D. M. Bătinetu and N. Stanciu, 129(10), 2022).
Evaluate
(n + 1)2 n2
lim − ,
n→∞ xn+1 xn
p
n
√ √ √
where xn = 2! 3 3! · · · n n!.
Discussion.
Pattern recognition sometimes plays a key role in problem solving. First ask
yourself if you’ve seen a similar problem before, then revisit it to understand
how it works, and try to figure out which “trick” in the old problem can
be applied to your current problem. Our first solution illustrates how the
arguments of Monthly 11935 [30, pp. 21-23] can be used to solve this problem
effortlessly.
On the other hand, by the Stirling approximation
√
n! = 2πn(n/e)n (1 + O(1/n)),

we have √   
n ln(2πn) 1
ln n! = ln(n) − 1 + +O .
2n n2
The proposed problem naturally suggests establishing an asymptotic approx-
imation for their average:
1  √  √ 
3
 √ 
n
ln xn = ln 2! + ln 3! + · · · + ln n! .
n
6 Analysis

Indeed, we will prove that for some constant C,

 
ln2 n ln(2πe) ln(n) C ln(n)
ln xn = ln(n) − 2 + + + +O . (1.3)
4n 2n n n2

This, together with the Stolz-Cesàro theorem, leads us to a second solution.

Solution I.
We show that the limit is e2 . To this end, we establish two facts first.
F1 . limn→∞ xnn = e12 .
By the Stolz-Cesàro theorem, we have
Pn √k
k=2 ln( k!) − n ln(n)
x 
n
lim ln = lim
n→∞ n n→∞
 n 
SC p
n+1
= lim ln( (n + 1)!) − (n + 1) ln(n + 1) + n ln(n)
n→∞
p !  !
n+1
(n + 1)! 1
= lim ln − n ln 1 +
n→∞ n+1 n
= −1 − 1 = −2,

which is equivalent to F1 . Here we have used the well-known fact that

√
n
n! 1
lim = .
n→∞ n e
F2 . limn→∞ (xn+1 − xn ) = e12 .
p
Since xn+1 n n+1
n+1 = xn (n + 1)!, as n → ∞, using F1 , we have
p 1
! n+1 p 1
! n+1
xn+1 (n + 1)!
n+1 n+1
(n + 1)! n + 1 n
= = · · → 1.
xn xn n+1 n xn

Similarly, as n → ∞, we have
 n p p
xn+1 xn+1 1 n+1
(n + 1)! n+1
(n + 1)! n + 1 1
= n+1 = = · → · e2 = e.
xn xnn xn+1 xn+1 n+1 xn+1 e

Hence, using these facts and F1 , as n → ∞, we get

 
xn+1
xn − 1
 n
xn xn+1 1 1
xn+1 − xn = ·    · ln → 2 ·1 · 1 = 2.
n ln 1 + n+1 − 1
x xn e e
xn
A limit involving a geometric mean of roots of factorials 7

Finally, using F1 and F2 , we find that

(n + 1)2 n2 (n + 1)2 xn − n2 xn+1
lim − = lim
n→∞ xn+1 xn n→∞ xn xn+1
 
n(n + 1) (n + 1)xn nxn+1
= lim −
n→∞ xn xn+1 n n+1
 
n(n + 1) xn xn+1
= lim + + (xn − xn+1 )
n→∞ xn xn+1 n n+1
 
1 1 1
= e4 + 2 − 2 = e2
e2 e e
as claimed. 
Solution II.
To establish the asymptotic approximation (1.3), we first collect several facts
that all follow from the Taylor expansion − ln(1 − x) = x + x2 /2 + O(x3 ) as
x → 0:
 
1 1 1
ln(n) − ln(n − 1) = + 2 + O ;
n 2n n3
 
1 1
n ln(n) − (n − 1) ln(n − 1) = ln(n) + 1 − +O ;
2n n2
 
2 ln(n) ln(n) 1
ln2 (n) − ln2 (n − 1) = + 2
+ O ;
n n n2
 
ln(n) ln(n − 1) ln(n) 1
− =− 2 +O .
n n−1 n n2
Next, let
ln2 (n) ln(2πe) ln(n)
an = n ln(n) − 2n + + , and
4 2
n
X ln(k!)
bn = n ln(xn ) − an = − an .
k
k=1

Applying the facts above, we have

 
ln(2πn) ln(n) 1
an − an−1 = ln(n) − 1 + + +O .
2n 4n2 n2
Furthermore, by the Stirling approximation
 
ln(2πn) 1
ln(n!) = n ln(n) − n + +O ,
2n n
we obtain
 
ln(n!) ln(n) 1
bn − bn−1 = − (an − an−1 ) = − +O .
n 4n2 n2
8 Analysis

P∞
This implies that n=1 (bn − bn−1 ) converges, and so limn→∞ bn exists.
Let C = limn→∞ bn and cn = an + C. Then n ln xn − cn = bn − C → 0 as
n → ∞. Applying the Stolz-Cesàro theorem yields
cn
ln(xn ) − n bn − C bn − bn−1
lim ln(n)
= lim ln(n)
= lim
n→∞ n→∞ n→∞ ln(n) − ln(n−1)
n2 n n n−1
ln(n) 1


− 4n2 +O n2 1
= lim
= .
n→∞ − ln(n) +O 1 4
n2 n2

This proves (1.3) as claimed. Now, from (1.3),

  
n ln2 (n) ln(2πe) ln(n) C ln(n)
xn = exp + + +O .
e2 4n 2n n n2

Using e−x = 1 − x + O(x2 ) as x → 0, we have

  4 
n2 ln2 (n) ln(2πe) ln(n) C ln (n)
= ne2 1 − − − +O
xn 4n 2n n n2
  
2 ln2 (n) ln(2πe) ln(n) ln4 n
=e n− − −C +O .
4 2 n

Thus,

(n + 1)2 n2 2 ln2 (n + 1) − ln2 (n)
lim − = lim e 1 −
n→∞ xn+1 xn n→∞ 4
 
ln(2πe)(ln(n + 1) − ln(n)) ln4 (n)
− +O = e2 .
2 n


Remark. If we assume the proposed limit exists, we can determine the value
of the limit by using the Stolz-Cesàro theorem and F1 only. Indeed, we have
2 2
(n+1) n
(n + 1)2 n2 xn+1 − xn
lim − = lim
n→∞ xn+1 xn n→∞ n + 1 − n
n2
xn n
= lim = lim = e2 .
n→∞ n n→∞ xn
(1.3) provides an exquisite estimate, which is of interest in its own right.
Numerically, we have C = 1.54836 . . ., but, we can’t determine the value of C
in a closed form.
The approaches used in the above solutions can be applied to a large class
of sequences.
A limit involving a geometric mean of roots of factorials 9

Additional problems for practice.

q p
√
1. Does the sequence xn = 1! 2! · · · n! converge?
2. For p ≥ 1, let
 p 2p  √ 2p 
n
n+1
(n + 1)! n!
xn (p) = n1−p 
 
− .
(n + 1)p np

Find limn→∞ xn (p).

xn+1
3. Let xn be a positive sequence and p ≥ 0. If limn→∞ np+1 ·xn = q > 0,
show that √ √
n+1 x
n+1 −
n x
n
lim p
= (p + 1)qe−(p+1) .
n→∞ n
Let xn = n! and p = 0. This yields the classical limit
 p √
n
 1
lim n+1 (n + 1)! − n! = .
n→∞ e

4. Given a positive sequence xn , define (x1 )! = x1 , (xn+1 )! = xn+1 · (xn )!

for all n ≥ 1. For p ≥ 0, if limn→∞ xn+1
np
−xn
= q > 0, show that

(xn+1 )! q
lim = .
n→∞ np+1 (xn )! p+1

5. Problem 12129 (Proposed by H. Ohtsuka, 126(7), 2019). Compute

v s
u r
u q
t √
2 + 2 + 2 + ···+ 2 − 2 + ···

where the sequence of signs consists of n − 1 plus signs followed by a

minus sign and repeats with period n.
6. Problem 12510 (Proposed by H. Ohtsuka, 132(2), 2025). For c > 0 and
n ≥ 1, let s r q
√
Rn (c) = c + c + c + · · · + c

with n radicals. Evaluate

n Y
n
X 1
lim .
n→∞ Rj (c)
k=1 j=k
10 Analysis

1.3 A Lobachevsky-type integral

Problem 12351 (Proposed by S. Stewart, 129(9), 2022). Evaluate
Z ∞
ln(cos2 x) sin3 x
dx.
0 x3 (1 + 2 cos2 x)

Discussion.
We begin our analysis by getting some feel for the proposed integral in two
distinct directions. First, when routine methods such as substitution and in-
tegration by parts don’t shed any insight, we naturally try to apply R. Feyn-
man’s “a different box of tools” — the parametric differentiation. Introduce
Z ∞
ln(1 − p sin2 x) sin3 x
I(p) := dx.
0 x3 (3 − 2 sin2 x)

The proposed integral is simply I(1). For p ∈ [0, 1), using partial fractions
and the geometric series, we have
Z ∞
1 sin5 x
I ′ (p) = − 2 2 dx
0 (1 − p sin x)(3 − 2 sin x) x3
Z ∞ 2/3 p
!
3p−2 3p−2 sin5 x
= − dx
0 1 − 32 sin2 x (1 − p2 sin2 x) x3
∞  Z ∞
1 X 2n n sin2n+3 x
= − p dx, (1.4)
3p − 2 n=0 3n 0 x3

where the interchange of the summation and integration is justified since the
series is uniformly convergent. Our first solution will begin with (1.4).
Second, the sinc function sinc(x) := sinx x in the integrand reminds us of
Lobachevsky’s formula
Z ∞ Z ∞ Z π/2
sin2 x sin x
f (x) dx = f (x) dx = f (x) dx, (1.5)
0 x2 0 x 0

where f (x) is a continuous function with f (x ± π) = f (x) for all x ≥ 0. This

formula is quite unique because it uses the sinc function as a weight instead of
the usual exponential function that appears in the Laplace transform. Letting
f (x) = 1 yields the classical Dirichlet formula:
Z ∞ Z ∞
sin2 x sin x π
dx = dx = .
0 x2 0 x 2
Now, let us consider
ln(cos2 x)
f (x) = .
1 + 2 cos2 x
A Lobachevsky-type integral 11

Clearly, f is continuous and satisfies f (x ± π) = f (x) for all x ≥ 0. So the

proposed integral becomes the Lobachevsky-type:
Z ∞
sin3 x
I := f (x) dx.
0 x3
Since Z ∞
sin3 x 3π π
dx = 6= ,
0 x3 8 2
Lobachevsky’s formula in the form of (1.5) no longer holds for the cubic power
of sinc function. On the other hand, taking advantage of the π-periodic nature
of f , we decompose the interval [0, ∞) into intervals of length of π/2. Then
Z ∞ ∞ Z (n+1)π/2
sin3 x X sin3 x
f (x) dx = f (x) dx.
0 x3 n=0 nπ/2
x3

So we can also proceed with the proposed problem by first establishing

Z ∞ Z π/2  
sin3 x 1 2
f (x) dx = 1 − sin x f (x) dx. (1.6)
0 x3 0 2
Our second solution will be based on (1.6).
Solution I.
Applying integration by parts twice to the integral in (1.4), we find
Z ∞ Z
sin2n+3 x 2n + 3 ∞ sin2n+2 x cos x
dx = dx
0 x3 2 0 x2
Z
(2n + 3)(2n + 2) ∞ sin2n+1 x
= dx
2 0 x
Z
(2n + 3)2 ∞ sin2n+3 x
− dx.
2 0 x
Since n  
2n+1 1 X k 2n + 1
sin x= n (−1) sin(2k + 1)x
4 n−k
k=0
and Z ∞
sin(αx) π
dx = for every α > 0,
0 x 2
we find
Z ∞ n  Z ∞
sin2n+1 x 1 X k 2n + 1 sin(2k + 1)x
dx = n (−1) dx
0 x 4 n−k 0 x
k=0
n    
1 X 2n + 1 π π 2n
= n (−1)k = 2n+1 .
4 n−k 2 2 n
k=0
12 Analysis

This yields Z   
∞
sin2n+3 x π/4 1/2 2n
3
dx = n 1+ .
0 x 4 n+1 n
By (1.4), we obtain
π f (2/3) − f (p)
I ′ (p) = ,
4 3p − 2
where
∞   ∞   
X 2n  p n 1X 1 2n p n
f (p) = + , for |x| < 1.
n=0
n 4 2 n=0 n + 1 n 4

Applying the binomial series of (1 − p)−1/2 and its integration, we find

1 1
f (p) = √ + √ .
1−p 1+ 1−p

Since I(0) = 0, we finally obtain

Z 1 Z 1
f (2/3) − f (p)
π
I(1) = I(0) + I ′ (p) dp = dp
0 0 4 3p − 2
Z 1 
π 1 1 p
=− +√ dt (use t = 1 − p)
4 0 t+1 3t + 1
√ !
π 3 √
=− ln 2 + ln( 3 + 1) .
4 3


Solution II.
To prove (1.6), we split the right hand side sum into two parts by the parity
of n to get
∞ Z kπ+ π ∞ Z kπ
X 2 sin3 x X sin3 x
I= 3
f (x) dx + f (x) dx. (1.7)
kπ x kπ− π x3
k=0 k=1 2

Let x = kπ + t. Then
Z kπ+ π2 Z π/2
sin3 x k sin3 t
f (x) dx = (−1) f (t) dt.
kπ x3 0 (kπ + t)3

Similarly, let x = kπ − t. Then

Z kπ Z π/2
sin3 x sin3 t
f (x) dx = (−1)k−1 f (t) dt.
kπ− π
2
x3 0 (kπ − t)3
A Lobachevsky-type integral 13

Using these two integrals and (1.7), we have

Z π/2 ∞ Z π/2  
sin3 t X
k sin3 t sin3 t
I= f (t) dt + (−1) + f (t) dt
0 t3 (kπ + t)3 (t − kπ)3
k=1 0
Z π/2 " ∞  #
1 X 1 1
= sin3 t 3 + (−1)k + f (t) dt.
0 t (t + kπ)3 (t − kπ)3
k=1

Here the interchange of the order of the summation and integration is justified
by the uniform convergence of the series. Next, recall that if x 6= nπ for n ∈ Z,
then
∞  
1 1 X 1 1
= + (−1)k + . (1.8)
sin x x x − kπ x + kπ
k=1

Differentiating (1.8) twice term-by-term yields

∞  
2 1 2 X 1 1
3 − = 3+ (−1)k 2 + . (1.9)
sin x sin x x (x − kπ)3 (x + kπ)3
k=1

Finally, applying (1.9) obtains

Z π/2   Z π/2  
31 2 1 1 2
I= sin t
2 3 − sin t f (t) dt = 1 − sin t f (t) dt.
2
0 sin t 0

This proves (1.6) as claimed.

Using the formula [45, 4.385(3)]
Z π/2  
ln cos x π b
2 dx = ln for a, b > 0,
0
2 2 2
a cos x + b sin x 2ab a+b

after letting
ln(cos2 x)
f (x) =
1 + 2 cos2 x
we find
Z π/2 Z π/2  
ln cos x π 1
f (x) dx = 2 dx = √ ln √ .
0 0 3 cos2 x + sin2 x 3 3+1
14 Analysis

Hence,
Z ∞ Z π/2
ln(cos2 x) sin3 x 1
dx = (1 + cos2 x)f (x) dx
0 x3 (1 + 2 cos2 x) 0 2
Z Z
1 π/2 1 π/2
= f (x) dx + ln(cos x) dx
4 0 2
  0  
π 1 1 1
= √ ln √ + · − π ln 2
4 3 3+1 2 2
   
π 1 1
= √ ln √ − ln 2
4 3 3+1
√ !
π 3 √
=− ln( 3 + 1) + ln 2 .
4 3


Remark. The proof of (1.6) has its own interest: it serves as a demonstra-
tion of how the proof extends Lobachevsky’s formula (1.5) to higher powers.
Indeed, similar to the proof of (1.6), using (1.8) and its derivatives, we have
Z ∞ Z π/2  
sin4 x 2 2
f (x) dx = 1 − sin x f (x) dx,
0 x4 0 3
Z ∞ Z π/2  
sin5 x 5 2 1 4
f (x) dx = 1 − sin x + sin x f (x) dx.
0 x5 0 6 4
Recently, by using higher derivatives of the partial fraction expansion of csc x
and their derivative polynomials, H. Chen has found Lobachevsky’s formula
in explicit form for all integer powers (see [33]). This provides a good example
of how problem-solving can lead to a publication.
Additional problems for practice.
1. Show that
Z π/2  
ln cos x π b
2 dx = ln for a, b > 0,
0
2 2 2
a cos x + b sin x 2ab a+b

2. Problem 11423 (Proposed by G. Minton, 116(3), 2009). ShowR ∞ that if n

n
and m are positive integers with n ≥ m and n−m even, then 0 sin x
xm dx
is a rational multiple of π.
3. Problem 12375 (Proposed by H. Chen, 130(2), 2023). Let
Z ∞  n
2 2 1
In = 1 − x sin dx.
0 x
Monthly 12288 [2021, 956] asked for a proof that I2 = π5 . Prove that In
is a rational multiple of π whenever n is a positive integer.
A Lobachevsky-type integral 15

4. Let n and m be positive integers with n > m ≥ 2 and n − m odd.

Evaluate Z ∞
sinn x
dx
0 xm
in a closed form.
5. Show that Z ∞
arctan(sin x) π √
dx = ln(1 + 2).
0 x 2
6. In Lobachevsky’s formula (1.5), if the condition f (x + nπ) = f (x) is
replaced by f (−x) = f (x) and f (x + nπ) = (−1)n f (x), show that
Z ∞ Z π/2
sin x
f (x) dx = f (x) cos x dx.
0 x 0

Using this result proves that

Z ∞  
sin(tan x) π 1
dx = 1− .
0 x 2 e

7. Research Project. Find a Lobachevsky’s formula for all positive integer

powers and f with arbitrary period T via Fourier analysis.
Recall some standard notations from Fourier analysis. Let f, g ∈ L1 (R).
We define Z ∞
F (ξ) = f (x)e−2πixξ dx
−∞

as the Fourier transform of f and

Z ∞
(f ∗ g)(x) = f (y)g(x − y) dy
−∞

as the convolution of f and g. Furthermore, let T > 0. Define

L1 [−T /2, T /2] = {f (x) ; f is T -periodic and absolutely integrable}.

For the set of such functions, the Fourier coefficient is defined by

Z
ˆ 1 T /2 −2πi nx
f (n) = e T f (x) dx, n ∈ Z.
T −T /2

(a) (Parseval Formula). Let f ∈ L1 [−1/2, 1/2] and g ∈ L1 (R). If

supp(g) ∈ [−M, M ] for some positive constant M and is bounded
variation in the neighborhoods of all n ∈ Z with |n| ≤ M , show
that
Z ∞ X g(n− ) + g(n+ )
f (x)ĝ(x) dx = fˆ(n) .
−∞ 2
n∈Z, |n|/T ≤M
16 Analysis

(b) Define the normalized cardinal sine on R by

 sin(πx)
πx , x =6 0,
Sinc(x) =
1, x = 0.

Find F −1 (Sinc(x)).
(c) Let B(x) = F −1 (Sinc(x)). If f ∈ L1 [−1/2, 1/2] and k is a positive
integer, show that
Z ∞ X n
Sinck (x)f (x) dx = fˆ(n) B ∗k ,
−∞ T
|n|/T ≤k/2

where B ∗k = B
| ∗ B {z
∗ · · · ∗ B}. In particular, if k = 1 and 0 < T < 2,
k times
then
Z ∞ Z T /2 Z T /2
Sinc(x)f (x) dx = B ∗ (0) f (x) dx = f (x) dx.
−∞ −T /2 −T /2

1.4 A quadratic series with the tail of ζ(2)

Problem 12287 (Proposed by O. Furdui and A. Sı̂ntămărian, 128(10), 2021).
Prove  !2 
∞ ∞
X X 1 1 3 1 3
n
2
−  = − ζ(2) + ζ(3).
n=1
k n 2 2 2
k=n

Discussion.P
∞
Let Tn2 := k=n 1/k 2 , the tail of ζ(2). The summand of the series nTn2 − 1/n
simply admits no special series that shed any insight. Naturally, we try to
transform the summand into an equivalent but simpler form in a meaningful
way. This can be accomplished by two different ways. The first method is
based on the standard reformulate approach – Abel’s summation formula,
which is analogous to integration by parts. This technique has been used
in Solution I of Monthly 11810 [30, pp. 71-72]. The second method is to
decompose the summand into the form
1
nTn2 − = tn − tn−1 + sn ,
n
where sn is a known convergent series. This idea originated from Monthly
E3352 [84]: Show that
∞
X 1 e
4 + n2 + 1)
= .
n=0
n!(n 2
A quadratic series with the tail of ζ(2) 17

Here the series itself is not telescoping, but

 
1 1 n n−1 1
= − + .
n!(n4 + n2 + 1) 2 (n + 1)!(n2 + n + 1) n!(n2 − n + 1) (n + 1)!
Thus, the original series becomes the sum of a telescoping series and a well-
known series.
Solution I.
Let the proposed series be S. For n ≥ 1, let
∞
!2
X 1
an = n, bn = ,
k2
k=n
Pn
and An = k=1 ak . Then An = n(n + 1)/2 and
∞
!2 ∞
!2
1 X 1 X 1
bn − bn+1 = + −
n2 k2 k2
k=n+1 k=n+1
∞
1 2 X 1
= + 2 .
n4 n k2
k=n+1

Applying Abel’s summation formula:

∞
X ∞
X
an bn = lim An bn+1 + An (bn − bn+1 ),
n→∞
n=1 n=1

we have
∞ ∞
! !
1 X 1 1 2 X 1 1
S = lim n(n + 1)bn+1 + n(n + 1) + −
n→∞ 2
n=1
2 n4 n2 k2 n
k=n+1
∞ ∞
!
1 1 X n+1 X 1 1
= lim n(n + 1)bn+1 + (ζ(2) + ζ(3)) + − .
n→∞ 2 2 n=1
n k2 n
k=n+1
(1.10)
By the Stolz-Cesàro theorem,, we obtain
P∞ 2
k=n 1/k − 1/n −1/n2 − 1/(n + 1) + 1/n 1
lim 2
= lim 2 2
= ,
n→∞ 1/n n→∞ 1/(n + 1) − 1/n 2
and so
∞
X 1 1 1
= + 2 + o(1/n2 ) (1.11)
k2 n 2n
k=n
and  2
1 1 1 1
bn = + 2 + o(1/n2 ) = 2 + 3 + o(1/n3 ).
n 2n n n
18 Analysis

Therefore,
1 1
n(n + 1)bn+1 = .
lim
2 n→∞ 2
Reversing the order of summation yields
∞ ∞ ∞ k−1 ∞ ∞
X 1 X 1 X 1 X1 X 1 X 1
2
= 2
= H
2 k−1
= Hk = ζ(3),
n=1
n k k n=1
n k (k + 1)2
k=n+1 k=2 k=2 k=1

where the well-known Euler’s identity is used in the last equality.

Applying Abel’s summation formula again with an = 1 and
∞
X 1 1
bn = 2
− ,
k n
k=n+1

in view of (1.11), we have

∞ ∞
X 1 X 1 1 1 1
= − 2 = − 2 + o(1/n2 ),
k2 k2 n n 2n
k=n+1 k=n

and
∞ ∞
! ∞  
X X 1 1 X 1 1 1
2
− = lim nbn+1 + n − +
n=1
k n n→∞
n=1
(n + 1)2 n n+1
k=n+1
  X ∞
1 1
= lim − + o(1/n) − = 1 − ζ(2).
n→∞ 2n n=1
(n + 1)2

In summary, by (1.10), we find that

1 1 3 1 3
S= + (ζ(2) + ζ(3)) + ζ(3) + 1 − ζ(2) = − ζ(2) + ζ(3)
2 2 2 2 2
as claimed. 
Solution II.
We begin with
1
nTn2 − = tn − tn−1 + sn , t0 = 0, (1.12)
n
2
where tn and sn for n ≥ 1 are to be determined. Let tn = αn Tn+1 + βn Tn+1 ,
2
where αn and βn will be chosen later. Since Tn+1 = Tn − 1/n , we have
 2  
1 1
tn − tn−1 = αn Tn − 2 + βn Tn − 2 − αn−1 Tn2 − βn−1 Tn
n n
2α n αn βn
= (αn − αn−1 )Tn2 − 2 Tn + 4 + (βn − βn−1 )Tn − 2 .
n n n
A quadratic series with the tail of ζ(2) 19

To match the coefficient of Tn2 in (1.12), we set α0 = 0 and αn − αn−1 = n

for n ≥ 1. Then αn = n(n + 1)/2 and
   
2 1 n+1 n + 1 βn 1
tn − tn−1 = nTn − + βn − βn−1 − Tn + − 2 + .
n n 2n3 n n

Comparing with (1.12) again, we set β0 = 0 and

Pnβn = βn−1 + (n + 1)/n for
n ≥ 1. We now have βn = n + Hn , where Hn = k=1 1/k is the nth harmonic
number. Hence,
 
1 1 1 Hn
tn − tn−1 = nTn2 − + + − .
n 2n2 2n3 n2

Summing this identity yields

N   N N N
X
2 1 X Hn 1X 1 1X 1
nTn − = tN + − − .
n=1
n n=1
n2 2 n=1 n2 2 n=1 n2
P∞
Letting N → ∞ and using Euler’s identity n=1 Hn /n2 = 2ζ(3), we find
∞  
X 1 1 1
nTn2 − = lim tN + 2ζ(3) − ζ(2) − ζ(3).
n=1
n N →∞ 2 2

Moreover, by the definition of tn and the selected αn and βn , we now have

N (N + 1) 2
tN = TN +1 + (N + HN )TN +1 .
2
Applying (1.11) yields

N (N + 1) 2 1
lim TN +1 = and lim N TN +1 = 1.
N →∞ 2 2 N →∞

Together with limN →∞ HN TN +1 = 0 since HN = ln(N ) + γ + O(1/N ), we

finally obtain the claimed result. 
Remark. When the summand of the series involves binomial coefficients, fac-
torials, and products of rational functions, there is a very good chance the
series can be solved by the powerful method of Wilf-Zeilberger via telescop-
ing. For example, returning to Monthly E3352, we rewrite the series in the
equivalent form
∞  
X 1 1
− = 0.
n=0
n!(n4 + n2 + 1) 2n!
The Wilf-Zeilberger algorithm decomposes the summand into bn+1 − bn with

n2
bn =
2 n!(n2 − n + 1)
20 Analysis

which immediately implies that the above sum gives zero. See [80] for more
examples of this kind. Similarly, we can rewrite this monthly problem as
 !2 
∞ ∞
X X 1 2 1 3
2n
2
− + 2 − 3  = 3.
n=1
k n n n
k=n

Can Wilf-Zeilberger algorithm handle this kind of summand?

Additional problems for practice.
P∞
1. Let ζ(z) = n=1 1/nz be the Riemann zeta function. Show that
∞ ∞
!2
X X 1 5
(a) 2
= 3ζ(3) − ζ(4).
n=1 k=n+1
k 2
∞ ∞
! 2
X 1 X 1
(b) = 5ζ(2)ζ(3) − 9ζ(5).
n=1
n k2
k=n+1
∞ ∞
!
X Hn X 1 7
(c) 3
= 2ζ(2)ζ(3) − ζ(5).
n=1
n k 2
k=n+1

2. Problem 11805 (Proposed by G. Glebov and S. Fraser, 121(10), 2014).

(a) Show that
∞ ∞ √
X (−1)k X (−1)k 5π 3 3
+ = ,
(3k + 1)3 (3k + 2)3 243
k=0 k=0
∞ k ∞
X (−1) X (−1)k 13
− = ζ(3).
(3k + 1)3 (3k + 2)3 18
k=0 k=0

(b) Prove that

Z 1 ∞
9 (ln x)2 18 X (−1)k
ζ(3) = dx − .
13 0 x3 + 1 13 (3k + 2)3
k=0

3. Problem 12045 (Proposed by O. Furdui and A. Sı̂ntămărian, 125(5),

2018). Prove that the series
∞ ∞
! !
X
n−1
X 1
(−1) n −1
n=1
k2
k=n+1
2
converges to π /16 − (1 + ln 2)/2.
4. Problem 12091 (Proposed by C. I. Vălean, 125(2), 2019). Prove
∞ X
∞ X
∞
X i!j!k!
2 (Hi+j+k − Hk ) = ζ(3)
i=1 j=1 k=1
ij(i + j + k)!

where Hk is the kth harmonic number.

A trigonometric logarithmic integral 21

5. Problem 12215 (Proposed by O. Furdui and A. Sı̂ntămărian, 127(9),

2020). Calculate
∞   
X 1 1 1 1
+ + + · · · − .
n=1
n2 (n + 2)2 (n + 4)2 2n

6. Problem 12222 (Proposed by R. Tauraso, 127(10), 2020). Prove

∞ ∞
X (−1)k X 1 13ζ(3)
=− .
k2 n2n 24
k=1 n=k

7. Problem 12254 (Proposed by C. Lupu (USA) and T. Lupu, 128(5),

2021). Prove
∞ n
!
X (−1)n X 1 3π
= ln(2) − G
n=0
2n + 1 n+k 8
k=1
P∞
where G is Catalan’s constant k=0 (−1)k /(2k + 1)2 .
8. Problem 12494 (Proposed by J. Santmyer, 131(9), 2024). Prove

∞ X ∞ 2ζ(3) if r = 0,
X 1
2 2
= 1 2 (2)

n=1 m=1
m n + mn + rmn  Hr + Hr if r is a positive integer.
r

1.5 A trigonometric logarithmic integral

Problem 12274 (Proposed by R. Tauraso, 128(8), 2021). Evaluate
Z 1   2
arctan x 2x
ln dx.
0 1 + x2 1 − x2

Discussion.
Let the proposed integral be I. Using the substitution x = tan(θ/2), we have
Z
1 π/2
I= θ ln2 (tan θ) dθ. (1.13)
4 0
Beginning with this integral, we will present twoR solutions. The first solution
1
transforms (1.13) into the well-known integral 0 ln2 x/(1 + x2 ) dx, and the
second solution applies the Fourier series of ln(tan θ).
Solution I.
The substitution u = tan θ in (1.13) readily leads to
Z
1 ∞ arctan u ln2 u
I= du.
4 0 1 + u2
22 Analysis

We rewrite this integral as

Z 1 Z ∞ 
1 arctan u ln2 u arctan u ln2 u
I= du + du .
4 0 1 + u2 1 1 + u2

Letting u → 1/u in the second integral yields

Z 1 Z 1 
1 arctan u ln2 u arctan(1/u) ln2 u
I= du + du .
4 0 1 + u2 0 1 + u2

Since
π
arctan u + arctan(1/u) = ,
2
we finally find
Z 1 ∞ Z 1
π ln2 u πX
I= du = (−1)n
u2n ln2 u du
8 0 1 + u2 8 n=0 0
∞
π X (−1)n π π3 π4
= = · = .
4 n=0
(2n + 1)3 4 32 128


Solution II.
Recall that
∞
X cos(2nθ)
ln(sin θ) = − ln 2 − ,
n=1
n
∞
X (−1)n cos(2nθ)
ln(cos θ) = − ln 2 − .
n=1
n

From which we obtain

∞
X cos(2(2n − 1)θ)
ln(tan θ) = −2 , for θ ∈ (0, π/2).
n=1
2n − 1

Plugging this into (1.13), we have

Z ∞
!2
π/2 X cos(2(2n − 1)θ)
I= θ dθ
0 n=1
2n − 1
Z ∞
!
π/2
X cos2 (2(2n − 1)θ)
= θ dθ
0 n=1
(2n − 1)2
Z π/2 ∞ n−1
!
X X cos(2(2n − 1)θ) cos(2(2m − 1)θ)
+2 θ dθ.
0 n=2 m=1
(2n − 1)(2m − 1)
A trigonometric logarithmic integral 23

Therefore
∞ Z π/2
X 1
I= θ cos2 (2(2n − 1)θ) dθ
n=1
(2n − 1)2 0
n−1 Z π/2
X 1
+2 θ cos(2(2n − 1)θ) cos(2(2m − 1)θ) dθ.
m=1
(2n − 1)(2m − 1) 0

Using trigonometric product to sum formulas and then integrating by parts,

we find ( π2
Z π/2
if n = m,
θ cos(2nθ) cos(2mθ) dθ = 16
0 0 if n 6= m.
Hence,
∞  
X 1 π2 π2 1 π4
I= · = ζ(2) − ζ(2) = .
n=1
(2n − 1)2 16 16 4 128

Remark. Along the same lines in Solution I, we can derive a more general
formula
Z 1   2n
arctan x 2x (2n)!
2
ln 2
dx = πβ(2n + 1),
0 1+x 1−x 8
P∞ (−1)n
where β(z) = n=0 (2n+1)z is the Dirichlet beta function.
Additional problems for practice.
1. CMJ Problem 1261 (Proposed by B. Bradie, 55(1), 2024). Evaluate the
following integral for real α > −2:
Z ∞
arctan x
dx.
0 1 + αx + x2

2. Problem 12054 (Proposed by C. I. Vălean, 125(6), 2018). Prove that

Z 1  
arctan x 1 + x2 π3
ln dx = .
0 x (1 − x)2 16

For more challenging integrals of the same author see [100].

3. Problem 12158 (Proposed by H. Grandmontagne, 127(1), 2020). Prove
Z 1
(ln x)2 arctan x 21πζ(3) π 2 G π 3 ln 2
dx = − − ,
0 1+x 64 24 32
P∞ n
where G = n=0 (−1) /(2n + 1)2 is Catalan’s constant.
24 Analysis

4. Problem 12221 (Proposed by N. Batir, 127(10), 2017). Prove

Z 1
ln(x6 + 1) π ln(6)
2 +1
dx = − 3G,
0 x 2
P∞
where G is the Catalan’s constant n=0 (−1)n /(2n + 1)2 .
5. Problem 12256 (Proposed by P. Bracken, 128(5), 2021). Prove
Z 1
ln(1 + x) ln(1 − x) 5
dx = − ζ(3).
0 x 8

6. Problem 12501 (Proposed by H. Grandmontagne, 131(10), 2024). Prove

Z ∞
(ln(x/(1 + x)))4 ln(x3 (1 + x)17 )
dx = −240 ζ 2 (3),
0 1+x
P
where ζ(3) is Apéry’s constant ∞ 3
n=1 1/n .
P∞ (−1)n
7. Let β(z) = n=0 (2n+1)z be the Dirichlet beta function (Note that
β(2) = G). Prove
Z 1
ln x arctan x 7ζ(3) β(2)π
(a) 2
dx = − ,
0 1 + x 16 4
Z 1
ln x(arctan x)2 β(4) β(2)π 2
(b) dx = − ,
0 1 + x2 2 16
Z 1
x arctan x(ln x)2 7πζ(3) π 3 ln 2
(c) 2
dx = + β(4) − ,
0 1+x 64 16
Z 1 √
ln(x2 + x + 1) π β(2)
(d) 2
dx = ln(2 + 3) − .
0 1+x 6 3

1.6 An interesting series with a product of two

central binomial coefficients
Problem 12381 (Proposed by H. Chen, 130(3), 2023). Prove
∞


2n 4n
X
n 2n 1 √ √ 
= 1 − 2 + ln(1 + 2) .
n=1
64n (2n − 1) π

Discussion.
The key idea which makes the problem feasible is that we can build up the
desired summand step by step. We begin with the well-known generating
An interesting series with a product of two central binomial coefficients 25
 2n


function of the sequence n n≥0
∞  
X 2n n 1
x = √ . (1.14)
n=0
n 1 − 4x
With some manipulation, then we apply the Wallis integral formula
Z π/2  
π 2n
sin2n x dx = 2n+1 (1.15)
0 2 n


to associate with the factor 4n
2n .

Solution.
Using the substitution 4x → x2 in (1.14) yields
∞ 2n


X 1
1+ n
n
x2n = √ .
n=1
4 1 − x2
Rearranging this as
∞ 2n


X 1 1
n
x2n−2 = √ − 2,
n=1
4n x2 1 − x2 x
then integrating with respect to x gives
∞ 2n

√
X
n 2n−1 1 − 1 − x2
x = ,
n=1
4n (2n − 1) x
and so

∞
X 2n p
n
x2n = 1 − 1 − x2 .
n=1
4n (2n − 1)
2
Let x = sin t. Then
∞ 2n


X p p
n
sin4n t = 1 − 1 − sin4 t = 1 − | cos t| 1 + sin2 t.
n=1
4n (2n − 1)
Integrating this series with respect to t from 0 to π/2 and then using the
Wallis integral formula (1.15) yields
∞ 2n

Z π/2 ∞


2n 4n
X
n 4n π X n 2n
n (2n − 1)
sin t dt = n (2n − 1)
n=1
4 0 2 n=1
64
Z π/2 p
= (1 − cos t 1 + sin2 t) dt
0
Z 1p
π
= − 1 + u2 du (use u = sin t)
2 0
π 1 √ √ 
= − 2 + ln(1 + 2) ,
2 2
which is equivalent to the claimed identity. 
26 Analysis

Remark. Similarly, by applying the Wallis integral formula with odd expo-
nent,
Z 1
16k
sin4k+1 t dt =
, for all k ∈ N,
0 (4k + 1) 4k
2k
we can evaluate the series involving the ratio of the central binomial coeffi-
cients. We illustrate the idea by showing that
∞

X 4n n 2n
n 1 √
4n

= (5 2 − 4).
n=1
(2n − 1)2 (4n + 1) 2n 9

Integrating the first series in the above Solution gives

∞ 2n


X
n
n (2n + 1)
x2n+1 = arcsin x.
n=0
4

Applying the substitution x → x2 and then shifting the index n−1 → n yields
∞

X n 2n 1
n
n (2n − 1)2
x4n−2 = arcsin(x2 ),
n=1
4 2

and so ∞

X n 2n 1
n
n (2n − 1)2
x4n+1 = x3 arcsin(x2 ).
n=1
4 2
Let x = sin t. Then
∞

X n 2n 1
n
n (2n − 1)2
sin4n+1 t = sin3 t arcsin(sin2 t).
n=1
4 2

Integrating with respect to t from 0 to π/2 and using the Wallis integral
formula, we find that
∞

4n n 2n
Z π/2
X 1
n

= sin3 t arcsin(sin2 t) dt
(2n − 1)2 (4n + 1) 4n 2
n=1 2n 0
Z
1 1 u
= √ arcsin u du (use u = sin2 t)
4 0 1−u
Z  
1 1 2 √
= arcsin u d − (2 + u) 1 − u
4 0 3
Z 1
1 √ du
= (2 + u) 1 − u √
6 0 1 − u2
Z 1
1 2+u 1 √
= √ du = (5 2 − 4),
6 0 1+u 9
as claimed.
An interesting series with a product of two central binomial coefficients 27

It is worth taking a brief look at the background of the term “interesting

series”. In [65] marking the 60th anniversary of his first contribution to the
Monthly, D. H. Lehmer studied two classes of series:
∞   ∞
X 2n X an
an and 2n .


n=1
n n=1 n

In his sense, the adjective “interesting” means that the sum of the series can
be expressed in terms of known constants.
By manipulating the known power series through specialization, differentia-
tion and integration, he found numerous intriguing results. Stimulated by his
work, searching for interesting series associated with central binomial coef-
ficients, Catalan numbers and harmonic numbers has become a very active
research topic. Investigators have developed many different approaches in-
cluding generating functions, Gauss hypergeometric functions and the Fourier-
Legendre series. For example, by using Fourier-Legendre series expansions,
The first author [32] derived many interesting series including these elegant
Ramanujan-like formulas:
∞  2 √ 2
X (−1)n 2n 2Γ (1/4)
n
= ,
n=0
16 n 4π 3/2
∞  3
X (−1)n (4n + 1) 2n 8
n (n + 1)(1/2 − n)
= ,
n=0
16 n π
∞  4
X (4n + 1) 2n 128
n (n + 1)2 (1/2 − n)2
= 2.
n=0
256 n 3π

For more details on these kind of series, please refer also to [25, 31, 67] and
the references listed therein. The interested reader is encouraged to pursue
new results in this topic.
Additional problems for practice.
P
1. Let Hn be the n-th harmonic number, and let hn = nk=1 1/(2k − 1).
For |x| < 1/4, show that
∞    √ 
X 2n n 2 1 + 1 − 4x
(a) Hn x = √ ln √ .
n=1
n 1 − 4x 2 1 − 4x
∞  
X 2n 1 √
(b) hn xn = − √ ln( 1 − 4x).
n=1
n 1 − 4x
Pn
2. Let Hn (2) = k=1 1/k 2 . For |x| < 2, show that
∞

2n 2n+1 n−1
n x 1 1
X X
n
= arcsin3 (x/2).
n=0
16 (2n + 1) (2k + 1)2 3
k=0
28 Analysis

Use this result to prove

∞

2n  
X
n 1 2n arcsin2 (x/2)
H 2n (2) − H n (2) x = √ .
n=1
16n 4 4 − x2

Pn
3. Let Hn (2) = k=1 1/k 2 . Prove
2n 2
∞

X
n H2n 6 ln 2 − 2
(a) n
= .
n=1
16 (2n − 1) π
2n 2
∞

X
n Hn 12 − 16 ln 2
(b) = .
n=1
16 (2n − 1)2
n π
2n 2
∞

X
n Hn (2) π 8
(c) n
=4− − .
n=0
16 (2n − 1) 3 π
2n 2
∞

X
n Hn (2) 2π 32
(d) = −12 + + .
n=0
16 (2n − 1)2
n 3 π
√
4. For |x| ≤ 4, let p2 = ( x2 + 16 − 4)/x. Show that
∞
X (−1)n+1 2n

4n

x = 8 arctanh2 (p) − arctan2 (p) .
n=1
n2 2n

Use this formula to derive

∞
X 16n 2 2
√
4n = π − 4 ln ( 2 − 1).


2
n 2n
n=1

Hint: Begin with the power series of arcsin2 x, then use the substitution
x → xi.
5. Prove
∞

2n 4n


X
n 2n 4 √
(a) ln(1 + 2).
=
n=0
64n (2n +
1) π
∞

2n 4n

√ √
X
n 2n 4 2 + 16 ln(1 + 2)
(b) = .
n=0
64n (2n + 3) 15π

6. Problem 12051 (Proposed by P. Ribeiro, 125(6), 2018). Prove

∞  
X 2n 1 π3 π
n (2n + 1)3
= + ln2 2.
n=0
n 4 48 4
An interesting series with a product of two central binomial coefficients 29

7. Problem 12094 (Proposed by P. F. Refolio, 126(2), 2019). Prove

2n 2
∞

X
n 48 − 32G
= 16(ln 2 − 1) + ,
n=0
16n (n + 1)3 π
P∞ k
where G = k=0 (−1) /(2k + 1)2 is Catalan’s constant.
8. Problem 12180 (Proposed by P. F. Refolio, 127(4), 2020). Prove
∞ 4n 2

√ 2
X
2n 2 2C π 1/2
= − 3/2 + √ ,
n=0
28n (2n + 1) π π 2 C2
R∞
where C = 0 t−1/4 e−t = Γ(3/4).
9. Problem 12337 (Proposed by H. Ohtsuka, 129(7), 2022). For k ∈
{0, 1, 2}, let
X (−4)n 2n−1
Sk = ,
2n + 1 n
where the sum is taken over all nonnegative integers n that are congruent
to k modulo 3.
Prove
√
ln(1 + 2) π
(a) S0 = √ + ;
3 2 6
√ √
ln(1 + 2) ln(2 + 3) π
(b) S1 = √ − √ − ; and
3 2 2 3 12
√ √
ln(1 + 2) ln(2 + 3) π
(c) S2 = √ + √ − .
3 2 2 3 12
10. Research Project – Lehmer’s limit. In [65], D. H. Lehmer studied the
series
X nk
Sk (x) :=
n
2n x .
n=1 n

In particular, recursively, he found that S1 (2) = π+3, S2 (2) = 7π/2+11,

. . . , and in general

Sk (2) = R1 (k)π + R2 (k),

where both R1 (k) and R2 (k) are rational numbers. Based on his calcu-
lations up to k = 10, he observed that the ratio R2 (k)/R1 (k) provided
an approximation to π. In [37], J. Dyson, N. E. Frankel, and M. L.
Glasser applied the Gauss hypergeometric function 2 F1 combined with
30 Analysis

some other advanced tools to find Sk (x) in non-recursive closed form

and eventually proved that

R2 (k)
lim = π.
k→∞ R1 (k)

Can we have an elementary solution to this limit?

1.7 A log-tangent integral

Problem 12317 (Proposed by S. Stewart, 129(4), 2022). Prove
Z π/2
sin(4x) ζ(3)
dx = −14 2 ,
0 ln(tan x) π

where ζ(3) is Apéry’s constant.

Discussion.
The usual integration techniques don’t work on this integral in the current
form. So we transform it by substitution. Since

sin(4x) = 2 sin(2x) cos(2x) = 4 sin x cos x(1 − 2 sin2 x),

by using the substitution u = tan x, then t = u2 , we obtain

Z π/2 Z ∞ Z ∞
sin(4x) 4u(1 − u2 ) t−1
I := dx = du = −4 dt
0 ln(tan x) 0 ln u(1 + u2 )3 0 ln t(1 + t)3
(1.16)

The challenge now comes from the ln t in the integrand. In the following,
we present two solutions and demonstrate how to remove ln t by parametric
integration and parametric differentiation, respectively.
Solution I – by parametric integration
To this end, observe that Z 1
t−1
tp dp = .
0 ln t
We have
Z ∞ Z 1 Z 1 Z ∞ 
tp tp
I = −4 dpdt = −4 dt dp.
0 0 (1 + t)3 0 0 (1 + t)3
References
M. Aigner , Moving into the Desert with Fibonacci, Math. Mag., 70 :11–21, 1997.
M. Aigner and G. Ziegler , Proofs from THE BOOK (6th edition), Berlin, Springer-Verlag, 2018.
A. V. Akopyan and A. A. Zaslawsky , Geometry of conics, Mathematical World 26 , American
Mathematical Society, Providence RI, 2007.
O. J. Alabi and G. Dresden , Fault-free tilings of the 3×n rectangle with squares and dominos, J.
Integer Seq., 24 , Article 21.1.2, 2021.
J.-P. Allouche and J. Shallit , The ubiquitous Prouhet-Thue-Morse sequence, in Ding, C. (ed.) et
al., Sequences and their applications. Proceedings of the international conference, SETA ‘98,
Singapore, December 14–17, 1998. Springer Series in Discrete Mathematics and Theoretical
Computer Science. 1-16, 1999.
J.-P. Allouche and J. Shallit , Automatic Sequences, Cambridge, 2003.
C. Alsina and R. B. Nelsen , Charming proofs. A journey into elegant mathematics The Dolciani
Mathematical Expositions 42 , The Mathematical Association of America, Washington D.C.,
2010.
T. Amdeberhan and R. Tauraso , Congruences for sums of MacMahon's q-Catalan polynomials,
Bull. Aust. Math. Soc., First published online, 2024.
T. Amdeberhan , G. E. Andrews and R. Tauraso , Extensions of MacMahon's sums of divisors,
Res. Math. Sci., 11 , Article 8, 2024.
G. E. Andrews , A simple proof of Jacobi's triple product identity, Proc. Amer. Math. Soc., 16
:333–334, 1965.
G. E. Andrews , Theory of partitions, Cambridge University Press, 1998.
G. E. Andrews , q-Analogs of the binomial coefficient congruences of Babbage, Wolstenholme
and Glaisher, Discrete Math., 204 :15–25, 1999.
G. E. Andrews and K. Eriksson , Integer Partitions, Cambridge University Press, 2004.
G. E. Andrews and M. Merca , The truncated pentagonal number theorem, J. Comb. Theory,
Ser. A, 119 :1639–1643, 2012.
T. M. Apostol , Introduction to Analytic Number Theory, Springer, 1976.
M. Beresin , E. Levine and J. Winn , A Chessboard Coloring Problem, Coll. Math. J., 20
:106–114, 1989.
E. R. Berlekamp , J. H. Conway and R. K. Guy , Winning Ways, Academic Press, 1982.
B. C. Berndt , Number Theory in the Spirit of Ramanujan, American Mathematical Society,
Providence RI, 2006.
F. Beukers , F. Luca and F. Oort , Power Values of Divisor Sums, Amer. Math. Monthly, 119
:373–380, 2012.
R. Blecksmith , M. McCallum and J. L. Selfridge , 3-smooth representations of integers, Amer.
Math. Monthly, 105 :529–543, 1998.
B. Bollobás , The Art of Mathematics. Coffee Time in Memphis, Cambridge, Cambridge
University Press, 2006.
B. Bollobás , The Art of Mathematics - Take Two. Tea Time in Cambridge, Cambridge,
Cambridge University Press, 2022.
G. M. Boros and V. H. Moll , Irresistible Integrals, Cambridge, Cambridge University Press,
2004.
P. B. Borwein , The Prouhet-Tarry-Escott problem, Computational Excursions in Analysis and
Number Theory, CMS Books in Mathematics, Springer-Verlag, 2002.
K. N. Boyadzhiev , Series with central binomial coefficients, Catalan numbers, and harmonic
numbers, J. Integer Seq., 15 , Article 12.1.7, 2012.
D. M. Burton , Elementary Number Theory, McGraw Hill, 2010.
N. Calkin and H. S. Wilf , Recounting the rationals, Amer. Math. Monthly, 107 :360–363, 2000.
E. Chen , Euclidean Geometry in Mathematical Olympiads, The Mathematical Association of
America, Washington D.C., 2016.
H. Chen , Excursions in Classical Analysis: Pathways to Advanced Problem Solving and
Undergraduate Research, The Mathematical Association of America, Washington D.C., 2010.
H. Chen , Monthly Problem Gems, CRC Press, 2022.
H. Chen , Interesting Series Associated with Central Binomial Coefficients, Catalan Numbers
and Harmonic Numbers, J. Integer Seq., 19, Article 16.1.5, 2016.
H. Chen , Interesting Ramanujan-Like Series Associated with Powers of Central Binomial
Coefficients, J. Integer Seq., 25 , Article 22.1.8, 2022.
H. Chen , Another extension of Lobachevsky's formula, Elem. der Math, 78 :93–100, 2023.
H. S. M. Coxeter and S. L. Greitzer , Geometry Revisited, 19 , Mathematical Association of
America, 1967.
B. Csákány and R. Juhász , The Solitaire Army reinspected, Math. Mag., 73 :354–362, 2000.
R. Donaghey and L. W. Shapiro , Motzkin numbers, J. Comb. Theory, Ser. A, 23 :291–301,
1977.
J. Dyson , N. E. Frankel , and M. L. Glasser , Lehmer's Interesting Series, Amer. Math. Monthly,
120 :116–130, 2013.
H. M. Edwards , Riemann's Zeta Function, Dover Publications, Inc., New York, 1974.
B. Farhi , An identity involving the least common multiple of binomial coefficients and its
application, Amer. Math. Monthly, 116 :836–839, 2009.
P. Flajolet and R. Sedgewick , Analytic Combinatorics, Cambridge University Press, 2009.
M. Gardner , Hexaflexagons, Probability Paradoxes, and the Tower of Hanoi, Cambridge
University Press and Mathematical Association of America, 2008.
I. M. Gessel , Wolstenholme Revisited, Amer. Math. Monthly, 105 :657–658, 1998.
W. Gilbert , Fractal geometry derived from complex bases, Math. Intell., 4 :78–86, 1982.
H. W. Gould , Combinatorial Identities, published by the author, revised edition, 1972.
I. Gradshteyn and I. Ryzhik , Table of Integrals, Series, and Product, 8th edition, edited by A.
Jeffrey and D. Zwillinger , Academic Press, New York, 2007.
R. L. Graham , Fault-free tilings of rectangles, in D. A. Klarner , ed., The Mathematical Gardner:
A Collection in Honor of Martin Gardner, Wadsworth, 120–126, 1981.
R. L. Graham , D. E. Knuth and O. Patashnik , Concrete Mathematics, 2nd edition, Addison-
Wesley, 1994.
A. Granville , Zaphod Beeblebrox's brain and the fifty-ninth row of Pascal's triangle, Amer. Math.
Monthly, 99 :318–331, 1992.
A. Granville , Arithmetic properties of binomial coefficients. I. Binomial coefficients modulo prime
powers, in Organic Mathematics (Burnaby, BC, 1995), CMS Conf. Proc., vol. 20, American
Mathematical Society, Providence RI, 253–275, 1997.
A. Granville , Smooth numbers: computational number theory and beyond, Buhler, J. P. (ed.) et
al., Algorithmic Number Theory. Lattices, Number Fields, Curves and Cryptography, Cambridge
University Press, Mathematical Sciences Research Institute Publications, 44 :267–323, 2008.
B. Green and T. Tao , On sets defining few ordinary lines, Discrete Comput. Geom., 59
:409–468, 2013.
C. M. Grinstead and J. L. Snell , Introduction to Probability, Second Revised edition, American
Mathematical Society, 2012.
L. Gualà , S. Leucci , E. Natale , and R. Tauraso , Large peg-army maneuvers, Proceedings of
the 8th International Conference on Fun with Algorithms (FUN'16), vol. 49 of LIPIcs , 1–15,
2016
L.-S. Hahn , Complex Numbers and Geometry, The Mathematical Association of America,
Washington D.C., 1994.
L. Hall and S. Wagon , Roads and wheels, Math. Mag., 65 :283–301, 1992.
G. H. Hardy and E. M. Wright , An Introduction to the Theory of Numbers, 6th edn., Oxford
University Press, 2008.
H. Iwaniec , Lectures on the Riemann Zeta Function, vol. 62 of University Lecture Series,
American Mathematical Society, Providence RI, 2014.
A. Jacquemot , T. Randall-Page , A. Slavík and S. Wagon , A rolling square bridge: reimagining
the wheel, Math. Intell., 46 :171–182, 2024.
J. H. Jaroma , On expanding 4/n into three Egyptian fractions, Crux Math., 30 :36–37, 2014.
C. Jeffrey C. (ed.), The Ultimate Challenge. The 3x+1 Problem , American Mathematical
Society, Providence RI, 2010.
V. Klee and S. Wagon , Old and new unsolved problems in plane geometry and number theory,
The Dolciani Mathematical Expositions, 11 , Mathematical Association of America, 1991.
F. Kuczmarski , Roads and Wheels, Roulettes and Pedals, Amer. Math. Monthly, 118 :479–496,
2011.
E. Landau , Handbuch der Lehre von der Verteilung der Primzahlen, vol. 2, Taubner, Leipzig,
1909.
D. H. Lehmer , Two nonexistence theorems on partitions, Bull. Amer. Math. Soc., 52 :538–544,
1946.
D. H. Lehmer , Interesting series involving the central binomial coefficient, Amer. Math. Monthly,
92 :449–457, 1985.
A. Levitin and M. Levitin , Algorithmic Puzzles, Oxford University Press, 2011.
C. Li and W. Chu , Infinite series concerning harmonic numbers and quintic central binomial
coefficients, Bull. Aust. Math. Soc., 109 :225–241, 2024.
J.-C. Liu , On a congruence involving q-Catalan numbers, C. R. Math. Acad. Sci. Paris, 358
:211–215, 2020.
L. Lovász , Combinatorial Problems and Exercises, 2nd edn., AMS Chelsea Publishing,
Providence, RI, 2007.
M. Marden , Geometry of Polynomials, Mathematical Surveys and Monographs. Vol. 3 (2nd
ed.), American Mathematical Society, Providence, RI, 1966.
J. Mason , L. Burton and K. Stacey , Thinking Mathematically, Pearson Education, 2010.
S. Mattarei and R. Tauraso , Congruences of multiple sums involving sequences invariant under
the binomial transform, J. Integer Seq., 13 , Article 10.5.1, 2010.
S. Mattarei and R. Tauraso , Congruences for central binomial sums and finite polylogarithms,
J. Number Theory, 133 :131–157, 2013.
M. Mehrabi and K. Andersen , Problem 12318, Problems and Solutions, Amer. Math. Monthly,
131 :173–174, 2024.
J. Nagura , On the interval containing at least one prime number, Proc. Japan Acad., 28
:177–181, 1952.
I. Niven , Integers of quadratic fields as sums of squares, Trans. Amer. Math. Soc., 48
:405–417, 1940.
S. Northshield , Stern's diatomic sequence 0,1,1,2,1,3,2,3,1,4… Amer. Math. Monthly, 117
:581–598, 2010.
H. Ohtsuka and Li Zhou , Problem 12361, Problems and Solutions, Amer. Math. Monthly, 131
:634, 2024.
E. Passow and J. Roulier , Monotone and Convex Spline Interpolation, SIAM Journal on
Numerical Analysis, 14 :904–909, 1977.
M. Petkovsek , H. Wilf , and D. Zeilberger , How to do monthly problems with your computer,
Amer. Math. Monthly, 104 :506–519, 1997.
K. H. Pilehrood , T. Pilehrood , R. Tauraso New properties of multiple harmonic sums modulo p
and p-analogues of Leshchiner's series, Trans. Am. Math. Soc., 366 :3131–3159, 2014.
G. W. Reitwiesner , Binary arithmetic, Advances in Computers, Academic Press, 1 :231–308,
1960.
L. Schumaker , On Shape Preserving Quadratic Spline Interpolation, SIAM Journal on
Numerical Analysis, 20 :854–864, 1983.
G. I. Senum and S.-J. Bang , Problem E3352, Problems and Solutions, Amer. Math. Monthly,
98 :369–370, 1991.
J. Silverman , Taxicabs and Sums of Two Cubes, Amer. Math. Monthly, 100 :331–340, 1993.
N. J. A. Sloane , The On-Line Encyclopedia of Integer Sequences© (OEIS©), Available at
http://oeis.org
A. Soifer , The Mathematical Coloring Book. Mathematics of Coloring and the Colorful Life of Its
Creators, Springer, 2009.
R. P. Stanley , Enumerative Combinatorics, Vol. 1., 2nd edn., Camb. Stud. Adv. Math., vol. 49,
Cambridge University Press, 2012.
R. P. Stanley , Catalan Numbers, Cambridge University Press, 2015.
R. P. Stanley , Conversational Problem Solving, American Mathematical Society, Providence,
RI, 2020.
R. P. Stanley and H. Kwong , Problem 12113, Problems and Solutions, Amer. Math. Monthly,
131 :89–90, 2021.
E. M. Stein and R. Shakarchi , Complex Analysis, Princeton University Press, 2003.
Z.-W. Sun and R. Tauraso , On some new congruences for binomial coefficients, Int. J. Number
Theory, 7 :645–662, 2011.
L. Szalay , On the diophantine equation (2 n –1)(3 n –1=x 2 ), Publ. Math. Debrecen, 57 :1–9,
2000.
T. Tao , Solving Mathematical Problems - A Personal Perspective, Oxford University Press,
2006.
R. Tauraso , A new domino tiling sequence, J. Integer Seq., 7 , Article 4.2.3, 2004.
R. Tauraso , q-Analogs of some congruences involving Catalan numbers, Adv. Appl. Math., 48
:603–614, 2012.
I. Tomescu , Problems in Combinatorics and Graph Theory, Wiley-Interscience, New York,
1985.
L. Tóth , Linear Combinations of Dirichlet Series Associated with the Thue-Morse Sequence,
Integers, 22 , A98, 2022.
C. I. Vălean , (Almost) Impossible Integrals, Sums, and Series, Problem Books in Mathematics,
Springer, 2019.
J. G. Wendel , Note on the Gamma Function, Amer. Math. Monthly, 55 :563–564, 1948.
H. Wilf , Generatingfunctionology, A. K. Peters, Ltd., 1994. Available at
http://www.math.upenn.edu/~wilf/DownldGF.html
B. Williamson An Elementary Treatise on The Integral Calculus, Longmans, Green (London),
1888.
D. R. Woods and D. Robbins , Problem E2692, Problems and Solutions, Amer. Math. Monthly,
85 :394–395, 1979.
I. M. Yaglom and V. G. Boltyanskii , Convex Figures, Rinehart and Winston, New York, 1961.
P. Yiu , Heronian triangles are lattice triangles, Amer. Math. Monthly, 108 :261–263, 2001.
J. Zhao , Bernoulli numbers, Wolstenholme's theorem, and p 5 variations of Lucas' theorem, J.
Number Theory, 123 :18–26, 2007.