Math Horizons 02 2016

Circles, Circles,
Everywhere | 8
Pi Day: Silly or Inspiring? | 16

Reconstructing Utopia | 18
Published by the Mathematical Association of America : : February 2016

Flow Free on a Torus
Brian Kronenthal and Wing Hong Tony Wong
T
hese puzzles are motivated by the it as a doughnut). So a path can go o one edge of the
smartphone game Flow Free. The board and reenter at the corresponding square on the
objective is to build pipes connecting opposite edge.
pairs of dots of the same color and We give a solution to the rst puzzle below.
shape. The rules are Solutions to the rest can be found at maa.org/
1. pipes can run horizontally or vertically, mathhorizons/supplemental.htm. Q
2. only one pipe can enter each square,
3. a pipe cannot cross itself, and Brian Kronenthal and Wing Hong Tony Wong are
4. pipes must pass through every square on the assistant professors of mathematics at Kutztown
board. University of Pennsylvania. They both study
In this variation, we play the game on a torus. combinatorics, a branch of mathematics that counts.
Imagine the opposite edges of the board are glued Email: kronenthal@kutztown.edu
togetherblue glued to blue and yellow glued to Email: wong@kutztown.edu
yellow. The resulting shape would be a torus (imagine http://dx.doi.org/10.4169/mathhorizons.23.3.2
Puzzle #2.
Puzzle #1. Solution to puzzle #1.
Puzzle #3.
Puzzle #5.
Puzzle #4.
2 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

Contents
ABOUT THE COVER
In (Always) Room for One More Christopher Ennis proves that for certain p-values,
it is possible to ll a circle with circles of area An = C n p , placed randomly, one
at a time (for n = 1,2,3, ...). The artwork on the cover, created by John Shier, shows
the random placement of 1,000 circles with a p-value greater than Enniss bound.
Computational tests show that for this exponent (p = 1.46) if the rst few circles are
placed successfully, then the algorithm will not terminate.
8 (Always) Room for One More

Christopher Ennis proves that a random placement of circles
inside a larger circle will never terminate.
2 Flow Free on a Torus
Play a popular smart phone app on a doughnut.
Brian Kronenthal and Wing Hong Tony Wong
5 All You Need to Be a Mathematician
Katharine Merow chats with University of Chicago mathematician Amie Wilkinson.
12 Run, Hero, Run!
A poem inspired by the binary numbers zero to seven. Mike Naylor
13 Radical Dash
Students participated in an exciting scavenger hunt at MAA MathFest. Alexandra Branscombe
16 Pi Day, Pro and Con

Nicholas Fiori writes that there are better ways to celebrate than memorizing digits and eating
pie, while Luke Anderson argues that Pi Day can hook reluctant math students.
18 Minimizing Utopia
Andrew Simoson re-creates the island Utopia from Thomas Mores description.
22 DO THE MATH!
Tunnel-Cube
Create a mathematical sculpture out of playing cards. George Hart
24 A Pi Day of the Century Every Year

Missed Pi Day of the Century last year? It is not too late to celebrate. Cornelia A. Van Cott
28 THE BOOKSHELF:
Matt Davis reviews The Magic of Math: Solving for x and Figuring Out Why, by Arthur Benjamin;
Jeb Collins reviews Genius at Play: The Curious Mind of John Horton Conway, by Siobhan Roberts.
30 THE PLAYGROUND:
The Math Horizons problem section, edited by Gary Gordon
34 AFTERMATH:
The Law of the Broken Futon
A missing piece in your mathematical education can hinder future learning. Ben Orlin
www.maa.org/mathhorizons : : Math Horizons : : February 2016 3

Math Horizons is a vibrant and accessible forum for practitioners, students,
educators, and enthusiasts of mathematics, dedicated to exploring the
folklore, characters, and current happenings in mathematical culture.
EDITOR EDITORIAL BOARD

David Richeson
Ezra Brown Angie Hodge
Dickinson College
Virginia Tech University of Nebraska Omaha
MANAGING EDITOR
Nathan Carter David Kung
Lois M. Baron
Bentley University St. Marys College of Maryland
ART DIRECTOR
Mary Anne Mulligan Marc Chamberland Rachel Levy
Grinnell College Harvey Mudd College
PROBLEMS EDITOR
Gary Gordon Timothy Chartier Colm Mulcahy
Lafayette College Davidson College Spelman College
FB: facebook.com/MAAMathHorizons Alissa S. Crans Pamela Pierce

Twitter: @MathHorizons Loyola Marymount University College of Wooster
Michael Dor Jennifer Quinn
Brigham Young University University of Washington Tacoma
Math Horizons (ISSN 1072-4117) is published
four times a year (September, November, William Dunham Amy Shell-Gellasch
February, and April) by the Mathematical Bryn Mawr College Montgomery College
Association of America, 1529 Eighteenth Dana Ernst Allan Rossman
Street, NW, Washington, DC 20036. Northern Arizona University Cal Poly San Luis Obispo
February 2016, Volume XXIII, Issue 3. Brie Finegold Steven Strogatz
Periodicals postage paid at Washington, Rincon Research Corporation Cornell University
D.C., and additional mailing offices. All
Jerey Forrester Talitha Washington
MAA members have free access to the Dickinson College Howard University
electronic edition of Math Horizons and
Deanna Haunsperger Jim Wiseman
can opt for a print subscription at a rate
Carleton College Agnes Scott College
of $20 annually. Annual subscription rates
are $54 for nonmembers (print) and $51 for
libraries (electronic via JSTOR). Bulk print Instructions for authors: Math Horizons is intended primarily for
undergraduates interested in mathematics. Thus, while we especially value
subscriptions sent to a single address are
and desire to publish high-quality exposition of beautiful mathematics, we also
encouraged. The minimum order is 20 cop-
wish to publish lively articles about the culture of mathematics. We interpret
ies ($220); additional subscriptions can be this quite broadlywe welcome stories of mathematical people, the history of
ordered in units of 10. an idea or circle of ideas, applications, ction, folklore, traditions, institutions,
Subscription inquiries humor, puzzles, games, book reviews, student math club activities, and career
opportunities and advice.
Email: maaservice@maa.org
Manuscripts may be submitted electronically to Editor David Richeson
Web: www.maa.org (mathhorizons@maa.org). If submitting by mail, please send a hard copy to David
Richeson, Dickinson College, Department of Mathematics and Computer Science,
Fax: 240-396-5647
P.O. Box 1773, Carlisle, PA 17013-2896.
Call: 800-331-1622 or 301-617-7800
Write: Math Horizons, MAA Service Center, POSTMASTER: Send address changes to Math Horizons, MAA Service
Center, P.O. Box 90973, Washington, DC 20090-1112.
P.O. Box 90973, Washington, DC 20090-1112
Advertising information
Call: 202-387-5200
Printed in the United States of America.
Copyright 2016
Mathematical Association of America MATHEMATICAL ASSOCIATION OF AMERICA

All You Need
to Be a Mathematician
A Conversation with Amie Wilkinson
Katharine Merow
A
mie Wilkinson is a fellow of the American solution, which was
Mathematical Society (AMS) and a pro- rather revolutionary,
fessor of mathematics at the University was throw away
of Chicago, where she studies ergodic the equations,
theory and smooth dynamical systems. throw away the
Wilkinsons work garnered her the 2011 Satter Prize, goal of writing
awarded every other year by the AMS to recognize an down exact innite
outstanding contribution to mathematics research by a series solutions, and
woman in the previous six years. instead analyze the
I know Wilkinson through her husband, fellow qualitative properties
University of Chicago mathematician Benson Farb, who of his solution.
had my father for a high school math teacher. Wilkinson What Poincar
and I chatted by phone in October 2015. This interview discovered through
has been edited for length and clarity. some very clever
geometric and
Katharine Merow: When someone on an airplane or at a
topological
family function asks what you do, what do you say?
techniques was that
Amie Wilkinson: I usually take a historical approach. I
you can nd within
say, I study something called dynamical systems. A Jessica Wynne
Amie Wilkinson. the conguration of
dynamical system is a space and a set of ruleslike an
three bodieslike
algorithmthat tells you how to move around points
the sun, Jupiter, and an asteroidchaotic motion.
in the space. You take this set of rules and your space,
Highly unpredictable evolution of a system over time.
and you apply the rules and move the points. You apply
The nal thing I say is that dynamical systems is a
the same rules again. You move the points again. You
eld thats rooted in describing solutions to fundamental
repeatedly apply this xed set of rules, and you see
physical problems, but that the math is so powerful
what happens. A fundamental example of a dynamical
that it became an independent eld of mathematics
system is where your space is a set of congurations of
with fundamental questions completely unrelated to
some physical system and the set of rules that you apply
the initial application. And also these techniques have
comes from the laws of physics.
proved to be very useful in elds completely outside of
The birth of dynamical systems was the work of Henri
dynamical systems, such as number theory, geometry,
Poincar, who studied a variety of physical problems.
topology, and so on.
One that got quite a bit of attention was his prize-
winning entry in a contest sponsored by King Oscar KM: You work in a historically male-dominated eld.
of Sweden in 1872, which was to describe the motion How has the experience of being a woman in math
of three heavenly bodies by solving the associated changed over the years?
dierential equations from physics. The contest was to AW: Well, I think the men have changed. I think the
write the solution to these equations in the form of an male mathematicians areon the wholeway, way,
innite series. What Poincar did in his prize-winning way, way more aware of the impact of what they say.

Theyre much more aware of the potential diculties KM: I read the commencement speech you gave at the
that women face. So I think that mathematics is a much Berkeley math department graduation in 2014. Can you
more welcoming place to be a woman. talkas you did in that speechabout procrastination?
But the number of women in math at the senior levels AW: One of the most important things about growing
has just budged only the smallest amount. I mean, up for me has been coming to terms with my natural
you could say its up by a factor of innity, but thats tendency to procrastinate. And turning those tendencies
because there were zero women in many math depart- into an advantage. I heard it said about a famous
ments. The number of women at the tenured level in mathematician that he seems to
math research departments is still miserably low. procrastinate quite a bit, but in fact
I didnt really have many role models when I was what hes doing is running a lot of
younger. I remember when I got pregnant I was terried things on background. And when I
that I was going to have to leave math. And that was heard that, I realized thats what
only 1998. I think young women now at least have I do as well. If you are the type of
some role models they can look up to and say, Yeah, person who puts things o, it might
you know what, I can be like a normal person and be a not be because youre lazy. It might
mathematician. be because you need to dwell on
You now have to take this little bit of role model stu things internally before they emerge
thats there and magnify it by a factor of 100, and then in their right form.
you would approach sort of true equality, the ideal. The KM: You also told the Berkeley
ideal is that however many women who are talented in crowd that youd taken up sewing.
mathematics and who enjoy mathematics, they become Danny Calegari
A dress
What interests you about sewing?
mathematicians. And I think that number should really
Wilkinson made AW: Picking up sewing was
approach 50 percent of mathematicians. from fabric particularly fascinating because Id
featuring a always thought of myself as someone
KM: How do you balance being a research mathemati-
fractal pattern who has a really strong 3D intuition.
cian with being a mother of two? by Danny
AW: I struggled very hard with my daughter to do Calegari. I think I am someone who has a very
everything: write papers, teach, take care of my kid, strong 3D intuition. But when I took
and do everything kind of at 100 percent. And I did not up sewing, I found myself doing things like sewing a
succeed. I put up a very tough front, so many people torus with a hole in it and then trying to turn it inside
did not realize that I felt like I was drowning. out without realizing that was not possible. Or not
When my son camewhich was ve years laterI being careful and sewing a Mbius band instead of a
was already tenured, and I really didnt feel any cylinder. And realizing that when you add a temporal
pressure. I took it easy. I went with the pregnancy. I element to your spatial intuitionlike rst Im going
kept doing the math, but I didnt put pressure on myself to sew this to this, then this to this, then Im going to
to prove anything in particular. And after he was born, turn this inside out, and then sew this to thisit taps a
I was so excited to get back to math that I jumped right whole other geometric side of your brain that you didnt
back into it as soon as I could. know existed. So I enjoy sewing because I enjoy that
How do I juggle? It slowed me down. I dont think geometric challenge.
thats necessarily a bad thing. Ive focused in my KM: I heard that you and Benson came up with some
research on quality over quantity. And Ive had some of the mathematics displayed in one of the Simpsons
really good collaborators, too, which can be very helpful Halloween episodes. Howd you get that gig?
when youre not 100 percent there when you have small AW: Through my Harvard roommate I met a bunch
children. Ive had to be pretty rm sometimes about of people who wrote for the Lampoon, and a bunch of
setting my priorities, not feeling guilty about not be- those people went on to write for television. One of
ing there for my family always. And, you know what? them, David Cohen, was a computer science graduate
Theyve survived. Theyre doing really well. So that was student at Berkeley, a couple years ahead of me, and
an important thing to learn. You dont have to be the we were friends. At some point he left graduate school
perfect parent; your kids can still turn out great. to write for The Simpsons. This was still when I was a

met. Theyre much more accustomed to meeting
peoples needs than to demanding things.
And Bensonbeing a stereotypical guy, or
maybe a sort of aggressive form of a stereotypical
guyis really good at asking for what he needs.
He sees me as sort of an extension of himself, as
married couples tend to do, and so he found it
unacceptable if I did not demand things. Pushing
for recognition, pushing for promotions, and so
on. Pushing for raises. He was kind of behind the
scenes not letting me give up. He had a tremen-
dous impact on my career just in this very simple
way of not letting me be passive.
Jessica Wynne KM: You didnt know upon graduation from

Amie Wilkinson (second from right) with her husband, Harvard that you would become a mathematician
mathematician Benson Farb, and their two children. or even that you wanted to. How do you gauge
graduate student, and he sent me an email saying that as an undergrad, as a beginning grad studentwhether
youre mathematician material?
he was writing a segment for the Simpsons Halloween
AW: For the year after I graduated, I worked for a con-
show and he was thinking of putting some little
sulting rm doing actuarial consulting. What I learned
math jokes in the background. That was right when
from that year away from academia is that all you need
Fermats last theorem had just been proved by Wiles.
to be a mathematician is the desire to live in the world
I think I was talking about it with Benson, and we
of mathematics. All you need is a little circuit in your
were thinking of ideas. One of our suggestions was to
brain that produces pleasure when you engage with
put a counterexample to Fermats last theorem. Dave
mathematics. If mathematics brings you pleasure, you
really liked that idea. He wrote a computer programI
have what it takes to be a mathematician.
mean this is total brilliant Dave Cohen stuto nd
At Harvard I looked around and I didnt see very
near counterexamples so that if you typed it into your
manydidnt see anywomen professors. I couldnt
calculator, it would look like its a counterexample. visualize myself being one of these guys. And I basically
And he came up with this near counterexample thought I wasnt a nerd enough, wasnt smart enough,
which he then included in and so on. Well, all it took was a year o where I was no
the episode. longer immersed in mathematics to realize how much I
KM: Whats it like being married to a fellow missed that. And I went back to grad school with the at-
mathematician? Competitive? Supportive? titude that I was just going to take it one day at a time.
Collaborative? As long as I was getting pleasure out of it, Id stick with
it. I was going to see how far I could go with just appre-
AW: Its very supportive. It is not collaborative. Such
ciating how fortunate I was to be able to be doing that.
as, we would never write a paper together. We tried
It turns out that that can take you all the way to being
once, and it was very adversarial. Thats the best I
a full professor at the University of Chicago!
can describe it. Our styles were so dierent we really So thats my feeling. I think its entirely a question of
clashed. We talk about math with each other all the emotion and drive and pleasure and not how quick you
time. We help each other giving talks, writing papers. are or how well you do on tests or anything like that. I
We give each other lots of feedback. dont think thats the important thing. Q
One thing Benson did for me that I think was really
Katharine Merow is a freelance writer and editor living
crucial in my success as a mathematician is he gave me
in the metropolitan Washington, D.C., area.
a spine. Not to make too-broad generalizations about
Email: kmerow@gmail.com
women, but I think its demonstrated that women
are much less comfortable asking that their needs be http://dx.doi.org/10.4169/mathhorizons.23.3.5

Figure 1. Five thousand disjoint circular discs.
(Always) Room
Christopher Ennis
for One More
I
n 2010, John Shier, a physicist who had worked That this can be done follows from a fact taught in
most of his career in the semiconductor industry rst-year calculus: The so-called p-series,
and who has a longstanding interest in computer-
1
generated art, wrote a simple computer program A= n
n=1
p
,
for the random placement of shapes into a bounded
region of the plane [1]. In gure 1 we see 5,000 circular converges when The value of this series as a
discs in a rectangular region. function of the exponent p is called the (real) Riemann
He and others used this algorithm with a wide variety zeta function and is denoted So, if we choose
of shapes to create many beautiful pieces. The algo- the sum of the areas equals A.
rithm worked awlessly, but he had no proof that it However, just because the sum of the areas equals the
always would. It just did. In this article we prove that total frame area, it does not guarantee that there will be
Shiers algorithm works for certain special shapes. a suciently large block of unused space somewhere
in the frame for the algorithm to squeeze in the next
Shiers Algorithm smaller shape in the sequence. Perhaps after many
Shier required the areas of the shapes to decrease ac- placements, the remaining area is so diusely spread
cording to a power law with a xed exponent throughout the frame that there is no room for the next
The algorithm ran- shape.
domly places the shape with area A1 into the frame, Shier found that for certain values p, this does not
and then each smaller shape is to be placed in sequence, happen: The algorithm appears to run forever! He has
disjoint from the previously placed shapes. The area of to manually terminate the program after some specied
the initial shape is chosen so that the (innite) sum of number of placements or after some percentage ll has
the areas equals the area of the enclosing frame, A: been achieved.

A= A .
n=1
n Shier and others have used his algorithm (and have
even extended it to three dimensions) to produce a

)LJXUH$VDPSOLQJRI-RKQ6KLHUV
artwork. 2(a), left, 2(b), above.
2(c) above, 2(d) and 2(e), right, 2(f)

below.
considerable Mathematically this is equivalent to the frame being

body of graphic a torus. In gure 2(b) the shape, the word YES, is a
art of captivat- union of other shapes, as is the frame (NO). In gure
ing intricacy. 2(c) all shapes are squares, but the white squares are
The coding oriented as diamonds and the black squares have sides
skills required to parallel to the frames sides. Despite the random algo-
implement the rithm, an emergent self-organization is clearly apparent;
algorithm with diamonds tend to cluster together, as do squares. As
simple shapes gure 2(d) demonstrates, there is no requirement that
like circles and the shapes be aligned consistently. Here each square is
rectangles are given a random rotation. Shapes with same orientations
well within the reach of beginning programmers. are colored the same. Nor do the shapes need to be
Figure 2 shows a selection of Shiers artistic output. similar to each other, as gure 2(e) demonstrates. The
(For more such art, see john-art.com or paulbourke.net/ blobs have boundaries that are dierent polar curves,
randomtile.) but their areas decrease according to the power law. In
Notice that in gure 2(a) the shapes are annular gure 2(f) the shape boundaries are a union of four cir-
rings. The frame boundaries are identied; a shape cular arcs. The white uncovered space is clearly visible
that runs o one side, reappears on the other side. in this beautiful lace-like example.

Shiers conjecture is that given any frame region and I0
any shape there exists an interval of possible exponents,
I1 I5 I4 I2 I3
(1,P), for which the algorithm could run forever, lling
more and more of the available space.
[ ]
But is it true? Shier and Bourke present a detailed sta- 0 1
tistical analysis of the algorithm indicating that it is true Figure 3: One-dimensional version.
[2]. Thousands of computer runs, some involving up to 36
hours of processor time, resulting in upwards of a million
disjoint placements suggest that the algorithm can run
forever. But these merely serve to strengthen the empirical
[ ]
evidence for the conjecture. A proof is required! 0 1
Figure 4. An extended interval and the inner band.
The One-Dimensional Conjecture
We begin with a one-dimensional version of the conjec-
total length possible for the black region, which would
ture. It is not as visually interesting as the two-dimen-
sional version, but the proof is good preparation for occur only if all extended intervals and the inner band
the proof for two-dimensional shapes. In this version are mutually disjoint (or overlapped at their end-
we randomly ll the interval [0,1] with a sequence of points). It is more likely that some of these caps will
disjoint intervals whose lengths decrease according to a have nontrivial intersections. Any such overlap would
power law (see gure 3). increase the amount of black region available for the
Let be our exponent, and let In denote next placement.
the length of the nth interval. As described previously, We now prove this inequality. Observe that
let and so that k1 k1
1 I n = 1 I k (I n + Ik)
n=0 n=1
k1
We shall prove the one-dimensional conjecture using = 1 kI k I
n=1
n.
induction. We can clearly place the rst interval in
[0,1]. Now assume that the rst intervals have
been placed. We must show that there is sucient Now, because I n = I 1 / n p and we have
room to t the next interval, of length Ik. k1 k1
I1 I1 I1
First, extend each of the
interval caps of length
intervals with
adjoined at each end
1 kI k
n=1
In = n
n=1
p

k p1
n
n=1
p
such that the length of the nth extended interval is

1 I1
Also, adjoin an interval of length = I1 n
n=k
p

k p1
.
to each interior end of [0,1]; we call this the inner band
and denote its length (see gure 4). The p-series is notoriously unsummable. We know
Dene the black region as the set of points inside that it converges (for but we do not have a
[0,1] but outside the inner band and all extended closed form for the sum. One of Leonhard Eulers rst
intervals. Observe that if we place the kth interval so mathematical accomplishments (in 1735) was to sum
that its center is in the black region, then it will be in the series for even values of p, but the exact sums for
[0,1] and will be disjoint from the previously placed other p-values are unknown. However, we can use the
intervals. Thus, to prove the conjecture it suces to integral test from calculus to obtain bounds for the
show that the black region is nonempty. This will sum (see gure 5). In particular,
follow if we prove that
Note that the quantity on the left is the smallest It follows that

1 I1 I1 I1
I1 n
n=k
p

k p1
>
(p 1)k p1

k p1
I 1(2 p)
=
(p 1)k p1
> 0,
because
Although the above argument settles our version of
the one-dimensional conjecture, the astute reader will
wonder if the conjecture holds for . It can be
shown the algorithm will not terminate when
...
but for it may. In fact, there is a nice formula
for the smallest k that can cause the algorithm to ter-
minate for a given [3].
Figure 5. The integral test gives a lower
bound for the p-series.
The Two-Dimensional Conjecture
We now prove Shiers conjecture for the case that k1
the shapes are circular discs, as in gure 1, except we B rn2 2rk r n 2r0rk (k 2) rk2 .
assume a circular boundary frame. The proof can be n=k n=1
extended to two-dimensional shapes considerably more As the areas obey the correspond-
general than discs, but the details are somewhat tech- ing radii must satisfy Moreover,
nical. These will appear elsewhere. ( p ) = A / A1 = r0 2 / r12 , so it follows that
Suppose the area of the circular frame is A = r0 2 . r0 = r1 ( p ). Substituting these expressions we obtain
Let where P is a value that we will k1
2 (p)

1 2 1 k 2
B r12 .
specify later. As before, let the area of the rst disc
be and the areas of subsequent discs be

n=k n
p

k p/2 n
n=1
p/2

k p/2

k p
Let rn denote the radius of the nth disc. Here we have three summations that need
Suppose discs have been placed disjointly in estimating, and for all three we can use the integral as
the framing circle. We need to show there is space to we did in the one-dimensional case. In particular, the
place the kth disc. rst summation is greater than
Extend each of the discs with a circumferential
band of width rk; that is, the nth disc becomes a disc
of area Also, extend inwardly the
the second summation is less than
boundary circle with a circumferential inner band of

width rk so that it has area A 0 = r0 2 (r0 rk )2 .
Dene the black region as all points inside the fram-
ing circle but outside the inner band and the extended and
discs. Note that if we place the kth disc so that the cen-
ter is in the black region, then it will be in the framing
circle and disjoint from previously placed discs (see g- When all of these bounds are substituted for their
ure 6 or view an animation at http://bit.ly/1RmAyBf). corresponding terms, we obtain
Let B be the area of the black region. Because
it suces to show that the quantity on the right is Because and it follows that
positive. Using the area formula for the circles (and and
some algebra), we obtain Thus,

Acknowledgments
The author thanks John Shier for freely sharing his
insight, voluminous data, and the beautiful graphics in
gures 1 and 2; and Barry Cipra for numerous helpful
Figure 6. An extended
discussions, his suggestions, and his encouragement.
disc, the inner band,
black region, and the
next-to-be-placed disc.
Further Reading
[1] J. Shier, Filling space with random fractal non-
overlapping simple shapes, Hyperseeing (Summer,
2011) 131140 (proceedings of ISAMA 11, June 1317,
2011).
[2] J. Shier and P. Bourke, An algorithm for random
fractal lling of space, Computer Graphics Forum 32
no. 8 (2013) 8997.
Observe that the rst factor of this expression is pos- [3] Barry Cipra, private communications. Q
itive for and for the second factor is ap-
Christopher Ennis teaches at Normandale Community
proximately 1. Thus In particular, we can use College. He buys many mathematics books, but because
the second factor to nd an interval of p-values, (1,P), he has less and less space available to t them into his
independent of k, for which Using a graphing bookshelves, he is buying smaller books.
calculator, we nd that . This completes Email: christopher.ennis@normandale.edu
the proof of Shiers conjecture for circular discs. http://dx.doi.org/10.4169/mathhorizons.23.3.8
Run, Hero, Run! Mike Naylor
Hero, hero, hero

Hero, hero . . . run!
Hero, run hero.
Hero run run!
Run, hero hero
Mike Naylor creates and Run, hero, run!
teaches math through art,
Run run hero.
poetry, music, and literature. He
is a founder and artistic director
Run run run!
of Amborneset Math Creativity
Center in Norway.
Email: abacaba@gmail.com
http://dx.doi.org/10.4169/mathhorizons.23.3.12
Radical
Dash
Figure 1. Fibonacci bunnies.
Alexandra Branscombe
T
wo students raced up the escalators to the posted pictures of themselves completing the tasks.
hotel lobby, then dashed over to a man The organizers kicked o the Dash by posting the rst
with a small silver dot stuck on his name clues to their account on Wednesday night (now
tag. Breathless, they gleefully introduced @maaradicaldash).
themselves, then all three posed with Kalene Ireland, Sean Ingimarson, Rebekah Bright,
their thumbs up for a photo. No, it was not a celebrity and Garrett Jones were all rst-time MAA MathFest
sighting, but something almost as rare: a roving piece attendees from Slippery Rock University. They bunched
of the mathematical scavenger hunt taking place amid together to complete their rstand easiestchallenge:
MAA MathFest. take a group sele of their team, The Means, the
For each annual MAA meeting, the Committee Medians, and the Codes.
on Undergraduate Student Activities (CUSAC) or-
ganizes special events, such as student presentations Math Scavenger Hunt
and a student problem-solving competition. But for Every day of the meeting, the Dash teams must
the 2015 MAA centennial celebration in Washington, decipher a new code, solve clues that lead to booths
D.C., CUSAC introduced something very dierent: the in the exhibit hall, and complete a physical math
Radical Dash. activity. I want students to actually dash; there needs
Ready to Dash
The Radical Dash is like the Amazing Race for
mathematics students. On the rst night of MAA
MathFest, more than 100 undergraduates crowded
around the registration table to sign up for the Radical
Dash, clumping together in teams of three to ve
students.
Radical Dash organizer Stacey Muir communicated
to the contestants through the photo-sharing social
media platform Instagram. Every day up to ve
clues were released, including a code to break, a
mathematical brainteaser, targets to nd throughout
the meeting, creative math artwork to fashion, and
math to nd in everyday objects. Student teams Figure 2. The Means. From left to right: Kalene Ireland,
received points for completing challenges when they Rebekah Bright, Sean Ingimarson, and Garrett Jones.

to be a physical
aspect to it, said
Jennifer Bergner,
a professor in the
Department of
Mathematics and
Computer Science
at Salisbury
University.
Wikipedia The MAA 100
Figure 3. A book themed Radical
embedding of a graph
Dash took months
requiring three pages.
of logistical
planning, said Bergner. She and the rest of the team
Julie Barnes, Janine Janoski, Jerey Liebner, Lisa
Marano, and Stacey Muirhad to build a mathematical Figure 4. Jennifer Bergner.
scavenger hunt: create clues to physical locations at
build community and get dierent groups talking that
the meeting, encrypt those clues, and embed them into
wouldnt normally talk, she said.
math problems.
That is how she and her fellow organizers came up
The Means Break Codes with the idea: While competing for points, students
The rst day of problem solving got o to a rocky get to interact with exhibitors, meet Project NExT
start for The Means. The four teammates had divvied fellows (the mathematicians with the dots on their
up the work so each person could tackle a dierent badges), and solve math problems in a creative
part of the clue. Jones was at ease solving a linear wayall while attending a national conference. Who
algebra problemuntil the very end, when he realized doesnt like a challenge? Some of us are nerds. We
the problem had been incorrectly dictated to him by a just like to tackle fun math problems! said Bergner.
teammate. He went back to the beginning. This year, those fun math problems included
Meanwhile, Bright and Ireland had just decoded a geometrical art project created by the Bridges
the rst part of another clue. But to solve the rest, Organizationan organization that promotes
connections between mathematics and art. The teams
they needed a cipher to decode the second encrypted
were given rulers and scissors to create their own
message. They both bent over their cell phones while
hat pattern, named after a mosaic pattern with hat-
they read o names of ciphers that might crack
shaped tiles discovered in a 14th-century Iranian bowl
the code. Bright and Ireland admitted they had
(see archive.bridgesmathart.org/2010/bridges2010-531.
never tried code breaking before, so this was a new
pdf).
challenge. So why now? I like problem solving! said
On Friday, dashers learned how to turn a sudoku
Bright.
board into a mathematical grapheach cell is
Amid the hours of lectures and presentations at the a vertex and two vertices are connected if the
conference, the problem solving was a welcome break corresponding cells must contain dierent values.
for the day, agreed the team. It is nice to go to talks, Robin Blankenship, a professor at Morehead State
but I want to do math too, said Bright. University, then gave the students a glimpse of the
That was the same thought that inspired Bergner to book embedding problem. In this case, the task is to
create the Radical Dash for her MD-DC-VA section. place the 16 vertices for a four-by-four sudoku board
Students are at a math conferencethey already on the spine of a book and to draw the 56 edges on
like mathematics, she said. She realized that this the fewest number of book pages so that no edges
idea could be applied at the national level and help cross (as in gure 3).
students walk away with more knowledge about the Each day of the meeting, the Radical Dash hashtag
mathematical community. It is such a natural way to (#MathFestDash15) grew as students posted

creative Instagram post was called: The Means, the
Medians, and the Codes for their Fibonacci origami
bunnies.
The team leapt up and cheered. The joyful winners
went to the front to receive their prizes: books on
graph theory and Alan Turing. We came into this not
expecting to actually learn anything, but now we have
a really good foundation in [code breaking], said Jones
after the awards were announced.
Ireland said that they also got a sense of the MAA
community that other attendees had been talking
about. She explained that one afternoon during the
)LJXUH7KH0HDQVKROGXSWKHLUMXGJHVFKRLFH competition, they worked with another Radical Dash
awards. From left to right: Garrett Jones, Sean team to solve one of the harder problems.
Ingimarson, Kalene Ireland, and Rebekah Bright. Jones added, Even though we are all in a
competition, we are all here to do math, and if we can
activities on Instagram. Teams posed with a member
help each other do more math, it is a win. Q
wearing an Association for Women in Mathematics
T-shirt, found mathematical images at national
Alexandra Branscombe is a sta writer for the MAA.
monuments, made patterns with their sushi, and even
Email: abranscombe@maa.org
recorded a math song in front of the White House.
Fibonacci Bunnies
On the last day of the meeting, The Means sat
together in the middle of the MAA Ice Cream Social.
Scooping up their ice cream, they talked excitedly
about their Friday night Instagram posttheir nal L.A. Math
chance to rack up points in the competition. Romance, Crime, and
The teams were prompted to create an image Mathematics in the
celebrating the MAAs centennial anniversary, so The City of Angels
Means got especially creative: Their nal image (see James D. Stein
gure 1) showed 12 white origami bunnies arranged
to show ve terms of the Fibonacci sequence: 1, 1, 2,
Cloth $24.95
3, 5. (Fibonaccis original math problem was about
reproducing rabbits.) Underneath it was a hint and a
cipher encrypted into a coded message.
The hint is: You get the key with the square root
I love this book. With attention to scene and character,
of the centennial number, its as easy as zero one one
L.A. Math weaves together the detective genre with
two three. And the cipher is Happy one-hundredth mathematics in entertaining and enjoyable ways.
birthday MAA, said Ireland. Oscar Fernandez, author of Everyday Calculus
Jones said, We were really buckling down on being
creative, so that meant sitting down for an hour A fun collection of short stories whose mysteries are
folding origami bunnies to try to scrape out a few resolved through classical mathematical conundrums. . . .
Entertaining both in and out of the classroom.
extra points.
Arthur Benjamin, author of The Magic of Math
The awards ceremony began, and the Radical Dash
judges stood on a platform at the center of the room
to announce the winning teams. One by one, the high-
scoring teams were called forward, receiving their prize See our E-Books at press.princeton.edu
T-shirts. Finally, one of the judges choices for most

PI DAY
An Irrational Holiday
Nicholas Fiori
S
tudents across the United compared to drinking milkshakes (or
States celebrate Pi Day on shaking spears?) at a Shakespeare
March 14. The connection festival.
between pi and this date Suppose you do want to throw a party
involves a whirlwind of for pi. First of all, dont worry about
contorted units: Pi is the ratio of the what day it is. The mystic, alluring ratio
circumference to the diameter of a is exciting any day of the year. Feel free
circle, and pi rounded to the nearest to indulge in long decimal approxima-
hundredth is 3.14. So we celebrate Pi tions of pi, but focus on the substance of
Day on 3/14 (even though a year doesnt the number. These estimates have a long
have 10 months and a month doesnt have history that features many clever mathemati-
100 days). Some folks got very exact last year and cal arguments. Approximations of pi date back to the
celebrated on 3/14/15 at 9:26:53. Babylonians, who placed it between 3 and 3 1/8. Later,
This may irk mathematicians, who tend to shun Archimedes inscribed and circumscribed polygons in
such coincidences and convolutions of units, but circles to obtain the now-familiar bounds of and
maybe any excuse to get people excited about math Isaac Newton and others used innite series
is a good one. Right? Id like to believe so. However, to arrive at accurate approximations even faster.
I have been to many Pi Day celebrations, and Im Suppose you want your festival to play o of pis mys-
skeptical of the mathematical messages that are often tery and charm. Leave the pie puns at the doorthe
sent. long and interesting history of pi is much more gripping.
Pi Day celebrations tend to focus on pis string of For instance, did you know that in 1897 the Indiana
essentially random digits. On one Pi Day I witnessed House of Representatives passed an absurd bill stating
a student recite over 250 digits in a pi digit recitation that
contest. Ive seen attendees read poems and perform What about pis tendency to show up in some of the
songs that play o of pis mystique and the innitude most unexpected places? Why does pi also appear in
of its digits, but allude only vaguely to its meaning. the formula for the area of a circle and in the volume
My love for you is innite, starts one poem. and surface area formulas of a sphere? How could it be
Other activities ignore that pi is a number at all. that this series of rational numbers converges to pi?
Some focus on puns on its namepeople eat pie, do
word games and puzzles with phrases like cow pie
and pie in the sky. They play pi bingo, where pis
role is merely as a letter in the Greek alphabet. So, please celebrate pi. Its mixture of simplicity and
These approaches are distinctly nonmathematical. depth make it one of the most astonishing wonders of
Mathematicians rarely pay attention to pis actual the mathematical world. But always celebrate with sub-
decimal digits. We approximate irrational numbers stance. Dont hold it up as a vague mathematical deity.
only when there is a physical or mathematical rea- Get to know the guest of honor. And againthere is no
son to obtain further accuracy. Arguments such as need to save your party supplies (compasses, straight-
Johann Lamberts 1768 proof that pi is irrational are edges, and so on) for March 14! Q
more important than endless attempts to enumerate
Nicholas Fiori is the chair of the mathematics depart-
its digits.
ment at Saint Anns School in Brooklyn, New York, a
The rampant attention paid to the name pi is
school that does not celebrate Pi Day. Illustrations by
also a problematic. Mathematicians dont care much
Elettra.
about the names given to ideashence such dry
Email: nori@saintannsny.org
names as x and y (and pi)its what you do with
them that matters. Eating pies on Pi Day could be http://dx.doi.org/10.4169/mathhorizons.23.3.16

PI DAY
Learn, Dream, Play

Luke Anderson
I
magine youre a fourth-grader, bright Day can and should shake us loose from the
and eager to learn, but math is not serious grind, even if for a day, and help
your favorite subject. Its not that us recall why we fell in love with math in
you arent good at ityou just the rst place.
havent felt a spark, or felt any- For example, a Pi Day history lesson
thing, really, while studying it. Your can remind us that math is surpris-
teacher announces that next Monday is ingly interconnected. How many of us
something called Pi Day. are aware, as Leonhard Euler was in the
Well be honoring the number pi, she 1700s, that
gushes, with a lesson from our textbook
about its mathematical properties. Well learn
some exciting facts!
Its probably a safe bet that this version of Pi Day The number is also a marvel of randomness and inn-
will do little to excite you or your classmates who ity, and Pi Day gives us space to explore these elusive
arent already in love with math. What if, instead, concepts. There are many roadside attractions along the
your teacher told you that next Monday would be lled endless journey through pis digits, like its six consecu-
with stories, contests, songs, and round snacks that all tive nines at digit 762 known as Feynmans point or
somehow pointed you back to lessons about a basic but the exclusive club of self-locating strings (a whimsi-
mysterious concept that has enchanted and eluded some cal cousin to the search for ever-larger primes). And
of the worlds sharpest minds for thousands of years? nally, whod have guessed that a faux battle between
This is not just a hypothetical situation for me, and two competing holidays, Pi Day (March 14) and Tau
its not just about fourth-graders. I have spent the last Day (June 28), would spark a valid debate about which
two decades speaking to audiences of all ages at Pi Day constant, or 2, is more fundamental?
events around the countryfrom preschoolers to under- Ill be the rst to confess that Pi Day, like any other
graduates to senior citizens, in venues ranging from a math holiday, is wholly invented and arbitrary. (To
Manhattan hedge fund to a suburban Colorado YMCA. wit, a reporter once asked me to invent a holiday for
Whatever the setting, I walk into the room knowing July 16, 2011, and I dubbed it Square-Prime Sandwich
that not everyone in the audience is on the edge of his Day. See if you can gure out why.)
or her seat, leaning into math and enjoying it purely on But like so many things in life, Pi Day is what we
its own merits. make of it. It has the potential to light up the eyes of
For those folks, I tell stories from the rich and dra- those who might otherwise miss their calling as mem-
matic human history of the numberlives devoted to bers of tomorrows generation of creatively quantitative
grinding out a few new digits, clashes in courtrooms minds. It gives todays math scholars permission and in-
and Congress about the numbers value and signi- spiration to wander to far-ung and trivial places. And
cance, and modern mavericks who think theyve proven while were there, our brains infused with pie sugars
that were all wrong. And if thats not enough, we sing and catchy song lyrics, we just might make new connec-
pi songs and swap feats of digit memory. On what other tions about things that really matter. Have a wonderful
day of the year can math be made so inviting to so Pi Day, and save me a slice! Q
many?
You may be thinking, Its nice that Pi Day entertains Luke Anderson is founder of TeachPi.org and a leading
the average public, but isnt it just a distraction to real voice in pis potential to spread enthusiasm for math.
math people? Although Pi Days noblest purpose is to He wouldnt get it if you asked him for his digits.
awaken the senses of would-be math lovers from outside Email: contact@teachpi.org
our ranks, its also here to tickle our own mathemati-
cal senses. For teachers and learners at all levels, Pi http://dx.doi.org/10.4169/mathhorizons.23.3.17

Beinecke Rare Book & Manuscript Library, Yale University
Minimizing Utopia
Andrew Simoson
T
his year is the 500th anniversary of Sir as in Euclid, not Euler)meaning a good placeand a
Thomas Mores Utopia. Utopia, Mores natural mathematical problem:
vision of an ideal society based on rea- What is the approximate greatest lower bound area of
son and equality rather than power and
Utopia?
wealth, means no place. Yet his itemized
physical parameters for the island of Utopia punningly Utopia is Mores improved take on Platos Republic.
suggests a Eutopia (with the rst syllable pronounced And its appearance in turn prompted other utopian
fantasies, such as Samuel Johnsons Rasselas and
Figure 1, above. Woodcut of the island of Utopia from Voltaires Candide and the dystopias of Aldous
the 1516 edition (colored by Simoson). Huxleys Brave New World and George Orwells 1984.

Platos work describes a philosopher kings triumph
over the mysterious realm of Atlantis. Johnson and
Voltaire place their utopian kingdoms in inaccessible
valleys of Abyssinia and Peru, respectively. And both
Huxley and Orwell place theirs in a depersonalized all-
too-accessible-and-inevitable futuristic London.
But More describes his fantasy island with curious
detail in the rst few pages of Book II of Utopia:
z Utopia consists of 54 city-states, each separated
from the nearest neighbor by at least 24 miles.
z Each city-state is square-like with side lengths of
at least 20 miles.
z Utopia has a circular-like harbor with a mouth
of 11 miles so as to make a perimeter of about
500 miles.
z Each cross section (parallel to a midline of the
island through the harbor mouth) of Utopia is
about 200 miles, except near the extremes where
it collapses to zero.
z The capital city, located in the center of the
island, lies about 60 miles from the harbor and
140 miles from the opposite coast.
z A channel 15 miles long separates Utopia from
the mainland.
z Utopia is crescent-like, with the horns enclosing
the central harbor.
In some ways, these description clues are camouage
for Henry VIIIs England (including Wales) of 53 coun-
ties and London isolated from the continent by the
English Channel. Nevertheless, as a celebration of the Peace Palace Library, The Hague
Figure 2. Woodcut of Utopia from the 1717 edition.
fth centennial of Utopia, we attempt to approximate
the inmum for Utopias possible area based on this
list of physical properties.
The Shape of Utopia

Artists have tried to use Mores instructions to lay out
the island of Utopia. Figure 1 is a woodcut of Utopia
from the rst edition of the book (1516). Figure 2
shows the circular harbor of Utopia from a 1717 edi-
tion. And gure 3, from a 1704 edition, is an attempt
to show the city-states, but it has room for only 28.
Our goal is to produce a map of the island that has
all 54 city-states. To achieve this goal, we make a few
assumptions. (See gure 4.)
z The outer coast of Utopia is an ellipse para-
meterized by
z The inner harbor coast is almost an osculating Beinecke Rare Book & Manuscript Library, Yale University
circle of radius d positioned on the right-hand )LJXUH8WRSLDVFLW\VWDWHVDZRRGFXWIURPWKH
side of (Such a circle is tangent to the ellipse edition displaying 28 cities.

Length of Cross Section
midline
y
p d q b
)LJXUH$SRVVLEOHVKDSHDQGVL]HIRU8WRSLDZLWKWKH Distance from Midline
capital city at the centroid of the island. )LJXUH&URVVVHFWLRQVDUHQHDUO\PLOHVIRU

and has the same radius of curvature.) Instead,
to ensure that the inner harbor has a perimeter Numerically solving this system of ve equations
of 500 miles, we exchange the osculating circle gives
for an ellipse of semi-axial lengths c and d and
where c slightly exceeds d. With these parameter values, gure 5 shows that
z The capital city is located at the centroid of the Utopias cross sectional length is nearly 200 miles at a
island. distance y from Utopias midline for and
z Each city-state has a central downtown disk it plunges to zero as y goes from about 95 to b.
region of radius one mile. Thus the distance Since each city-state is at least 20 miles along
between the centers of any two city-states is each side, each city must be centered at least 10
at least 26 miles (while the minimum distance miles inland from the coast, a boundary marked by
between neighboring downtowns remains at least the dotted curves along Utopias shoreline in g-
24 miles). ure 4. The capital city at this regions centroid is
Using the techniques found in the parametric marked as a red circle. The area for this island is
equations chapter of almost any calculus text, we uncannily near the area of
modern-day England (excluding Wales), 50,346
compute the radius of curvature of ellipse at para-
Before placing the remaining 53 city-states as dots
meter t to be Thus,
within Utopia, we insert twin reservoirs on the is-
the radius of the osculating circle to at
land (see gure 6). The centroids of the reservoirs lie
is which we take as the semi-
along and are symmetric to the vertical line through
axial height of the elliptical inner harbor . Since
the capital city so that the cross sectional spike in
the major axis of the island has length 2a as well as
By symmetry, Utopias centroid is for some

nonnegative number s. Since the distance from the
capital city to the harbor is 60 miles and the cen-
ters of and are (0,0) and (100,0), respectively,
Think of the inner harbor as having negative area

centered at (100,0) and the outer ellipse as
having positive area centered at (0,0). Therefore,
Finally, the perimeter of the harbor is 500 miles, so

)LJXUH$EHWWHU8WRSLDVL]HDQGVKDSHZLWKDOOFLW\
VWDWHVDQGWZLQUHVHUYRLUV

Further Reading
Length of Cross Section
[1] B. R. Goodey, Mapping Utopia: a comment on

the geography of Sir Thomas More, Geographical
Review 60 (1970) 1530.
[2] W. A. McClung, Designing Utopia, Moreana 31
nos. 118119 (June 1994) 928.
[3] Sir Thomas More, Utopia: written in Latin by
Sir Thomas More and done into English by Ralph
y Robynson, Heritage Press, New York, 1935.Q
p q b
Andy Simoson teaches math at a small college in
Distance from Midline the Appalachian Mountains. This past summer, he
)LJXUH&URVVVHFWLRQVDUHQHDUO\PLOHVIRU happened to page through Utopia in a library across
the street from a hospital, awaiting the birth of his rst
grandchild, Rebekah Lynnand was surprised to nd
gure 5 is attened to a nearly constant 200 miles on
Mores 500-year-old mathematical puzzle on the size
the interval (the cross sectional widths of
and shape of Utopia Island.
the island in gure 6 are 201 miles and 200 miles at
Email: ajsimoso@king.edu
and respectively).
After this engineering feat, the capital city yet
remains at the centroid. We use Cavalieris principle
to conclude that the area of each reservoir is ap-
proximately 1,297 So our revised inmum for
Utopias area is 45,096
Figure 7 shows that the cross sectional width of
this modied island is indeed nearly 200 miles for
and tails to zero afterward. Figure 6 shows
a distribution of 54 city-states, stacked like two-
dimensional cannonballs or oranges: two rows of seven
dots, three rows of six dots, four rows of ve dots,
and one blue dot each along the northern and south-
ern Utopian shores.
Other Interpretations
Can Mores clues be interpreted dierently? Certainly.
For example, Goodey [1] interprets Utopias 200-mile
cross sections as being perpendicular to the midline of
the island so as to reach a crescent with wide horns,
and he ultimately concludes that Mores island clues
cannot be taken literally. McClung [2] more or less
dismisses mapping eorts [from Mores clues] as fu-
tile. McClung appears to have misinterpreted Mores
clues as giving an island perimeter of 500 miles and a
diameter of 200 miles.
One way to improve upon our approach is to dredge
coves along the northern and southern coasts to
achieve a lesser area that still satises each of Mores
clues. Perhaps the reader may engineer a solution of
greater elegance.

DO THE MATH!
Tunnel-Cube
George Hart
T
he Tunnel-Cube is a sculpture puzzle made
from playing cards. It is a 10-centimeter
cube that, when displayed standing on a
corner as shown in gure 1, is 15 centime-
ters tall. Be warned that constructing it
is trickier than it looksit requires some care to avoid
ripping or damaging the cards. Figure 3. A triangular tunnel passes from any corner to
the opposite corner.
Assembly Instructions
The puzzle is made from 12 identical playing cards, each
slotted, folded, and assembled in the same manner.
Step 1. Cut 10 slits in each card following the tem-
plate of gure 4. (Printable templates are available at
maa.org/mathhorizons/supplemental.htm.) Be sure the
cards are face up as shown, so the slots on the diagonal
do not go through the letter or digit of the card value.
You can use scissors to cut each card, but for quantity
production, I use a scroll saw with a spiral blade to cut
through a deck at a time.
Figure 1. The Tunnel-Cube is made from 12 cards. Step 2. Fold each card along
the partially cut diagonal, as
When properly assembled, there are six square tun- shown in gure 5. Dont rip
nels leading to the center, like the one shown in gure anything. The picture side is
2. Each square tunnel starts in the center of a face, inside, leaving the back visible
passes through the center of the cube, and comes out from both sides. A thin metal
in the opposite face. The sides of the tunnel are rotated ruler makes this easyline
45 degrees relative to the sides of the cube. There are the diagonal over the edge and
also eight triangular tunnels leading to the center, as in push down with your thumb.
gure 3. Each starts in a corner, goes through the center Crease well, then let it spring
of the cube, and continues out to the opposite corner. Figure 4. The
template: Cut on the
open naturally to roughly a
So the center of the cube is a complex arrangement of 60-degree angle. (If you happen
solid lines and fold
intersecting tunnels, with 14 ways to exit. on the dotted line. to have a table or counter with
a sharp edge, that also works
well for creasing.) No additional bending or creasing is
necessary.
Step 3. Slide the end of one card onto the two short
slots in the side of another card, as shown in gure 6.
It takes only a slight ex of the card, then it locks into
place. It is easier if you rst connect the diagonal slot,
which is slightly longer. The lengths work out so the
edges of the two cards cross as shown. Keep in mind
how basic this is if you get frustrated later, because
every connection in this puzzle is exactly like this.
Step 4. This is the step where some people make a
Figure 2. A square opening is in the center of each face. mistake, so look carefully at gure 7. First, notice that

Figure 8.
Join two
cycles of
three to
make the
UVWVTXDUH
tunnel.
visualize how there will be an identical square tunnel in

the centers of the surrounding faces. Add the remaining
Figure 5, above. six cards one at a time, being sure to make the square
Fold the face to tunnels four-sided and not three-sided. Inserting the last
itself along the card is tricky because four connections need to be made.
diagonal.
Figure 6, right.
So be patient, and be careful not to bend or rip the cards.
Every connection Dont get upset if the sculpture falls apart during con-
is the same as struction. Cards can be slippery, but everything holds
this. together securely when the 12th card is inserted. When
you have nished, look on all sides and make sure all
the slots are completely engaged.
More
two corners of each card have a digit or letter while the
If you want to understand what you built and think
other two corners have a diagonal slot. Then observe about the underlying mathematical ideas, you might
how two of the digit (or letter) corners are near each rst nd all the symmetry axes. The four-fold and
other (the letters A here). You want to add a third card three-fold axes are easy because they go through the
to join each of the rst two in a kind of three-way cycle, tunnels, but can you nd the six two-fold symmetry
so the backs of the three digit corners make a triangular axes?
tunnel like the one in gure 7. Then think about the set of planes in which the
Step 5. Make a second three-card module, just like cards lie. Twelve folded cards could dene 24 planes
in space, but in fact there are only 12 planes, because
the rst. Figure 7 shows two views of the identical
each triangular half of any card is coplanar with an-
three-card structure.
other half card. What polyhedral shape denes these 12
Step 6. Join the two three-part modules to create a planes? (It is the intersection of 12 innite half-spaces
square tunnel as in gure 8. You need to make two con- dened by these planes.) Notice the 12 planes come in
nections to complete the square. six parallel pairs, so there are only six normal direc-
Step 7. At this stage, you have completed one square tions; characterize them with respect to the cube. The
face of the cube and parts of the four surrounding faces. cards automatically adjust to make everything work out
Look at the spiral structure of the completed square and without your having to know the exact fold angle; what
is the fold angle?
Here is a video showing the assembly process: youtube.
com/watch?v=kacnO5Ykfuk. Q
Figure 7.
Make George Hart is a research professor at Stony Brook
cycles University. His mathematical sculptures can be seen at
of three, many universities around the world. For examples of his
which will
be the
work, see http://georgehart.com.
corners of Email: george@georgehart.com
the cube. http://dx.doi.org/10.4169/mathhorizons.23.3.22

Every Year
I
Cornelia A. Van Cott
n 2015, Pi Day celebrations reached new
heights of enthusiasm. The digits of the
datewhen written in the American style
(3/14/15)matched the first five digits of
As this wont happen again until 2115, Figure 2.
Traveling
math enthusiasts dubbed this day the Pi Day of the from point
Century and celebrated appropriately. Those who A to point B
were ultra-enthusiastic celebrated the Pi Instant on using taxicab
3/14/15 at precisely 9:26:53.58979..., a moment in distance.
time that includes all the digits of (see [3]).
As we face future Pi Days, which no longer coincide to be the length of the line segment
with ve digits of we may see a dip in enthusiasm. AB. In other words, we measured distance as the
Our purpose here is to bring good news: We will be crow ies. The formula for this Euclidean distance is
able to celebrate a Pi Day of the Century again! Living
until 2115 would admittedly be a stretch, but that is Determining the length of a curvesuch as the
not the plan we have in mind. Our idea isof course circumference of a circleis trickier. To do so, we rst
a mathematical one. approximate the length by picking a nite number of
points on the curve, and summing the
Where Does Come From? distances from one point to the next (see gure 1),
A circle, as we well know, is the set of points equidis- The arc length
tant from a given pointthe center of the circle. The of the curve is the supremum over all such values
ratio of the circumference of a circle to its diameter that is, it is the smallest number that is greater
does not depend on the circle. This ratio is than or equal to the lengths of all these polygonal
To compute approximations.
we must be Computing the circumference of a circle in this way
P5 P4
able to mea- P3 is dicult. Through thousands of years of work
sure distances (which has involved everything from geometric argu-
and lengths P6 P2
ments to throwing breadsticks on tiles to harnessing
of curves in the power of supercomputers), mathematicians have
P7
the plane. computed trillions of decimal digits of
P1 = P13
Traditionally,
we dened P8 P12
Trading in Crows for Taxicabs
the distance Euclidean distance is a reasonable way to measure
between P9 P11 distance if youre a crow or if you own a helicopter,
P10 but there are innitely many other ways to measure
two points
Figure 1. A polygonal approximation distance.
and of a Euclidean circle. One alternative to Euclidean distance is the so-called

taxicab dis- 1 the ratio of circumference to diameter is
tance. Imagine So we write Yes, thats correct. Using taxicab
we are travel- distance, the corresponding -value is simply 4.
ing in a taxi- The fact that is 4 may be disappointing. Not
cab in a city only is rational, it is an integer! There is no mys-
whose streets tery, no need for supercomputing power, and no fun in
form a rect- having a -recitation contest. Nevertheless, our in-
angular grid. 1 terest is piqued. If we can change our distance formula
To get from and get circles to be diamond shaped and
point to point, then what other strange things are possible?
we may drive
only along Distance (Re)dened
the streets Before we begin cooking up more distance functions,
in the grid. Figure 3. The taxicab unit circle. we must introduce some notation and denitions. The
By analogy, most concise way to express a distance function is to
to measure the taxicab distance from to dene the distance from each point to the
in the plane, we add the horizontal and origin (0,0), which we denote and which we call
vertical distances between them (see gure 2). So, the norm of A.
.
We dene the distance between the points A and
Lets adopt the taxicab distance formula and watch
B as . One can check that the
the domino eect of changes that ensue. First, circles
Euclidean distance function arises from the norm
change shape. For example, the unit circle is the set of
and the taxicab distance func-
points (x,y) that are taxicab distance of 1 from (0,0).
tion arises from the norm .
Thus, the circle is given by the equation
Before moving forward, we mention a useful fact
and is diamond shaped (see gure 3).
about norms: For any norm the arc length of the
What is the circumference of this taxicab circle? One
line segment AB is . We observed this for the
is tempted to say that it is but thats if we com-
taxicab distance, and, with a brief argument, one can
pute the answer using Euclidean distance.
prove it for any norm.
According to the denition of arc length, we must
rst take a collection of points on the circle and sum
the taxicab distances from one point to the next. Geometry to the Rescue
Lets focus on the part of the unit circle in the rst Finding norms other than the ones we have already
quadrant. Something wonderfully convenient happens. mentioned might feel like stumbling around in the
No matter what points we take, the sum of the taxi- dark. Fortunately, there is a visual way to characterize
cab distances is always 2 (see gure 4)! Thus, the arc all such norms.
length of the unit circle in the rst quadrant is 2, and
the circumference is 8. Theorem 1. Norms are in one-to-one corre-
So, the ratio of the circumference to the diameter is spondence with closed, bounded, convex sets that
In fact, for any taxicab circle of radius r, contain a Euclidean disk around the origin and
are symmetric about the origin. Moreover,
1 1 1 such a set K is the unit disk for the associ-
ated norm .
Thats a mouthful! Lets look closely

at the parts of this theorem. That K is
closed means that it contains its boundary.
1 1 1 Bounded means that it does not extend to
Figure 4. The length of the line segment connecting (1,0) and (0,1), innity. Convex means that if A and B are
using taxicab distance, is 2. points in K, then the segment AB is in K. To

say that K is sym-
metric about the Now we have found three dierent -values: 3, 4,
origin means that if and the usual We could continue with this
(x,y) is in K, then P step-by-step work, but at this rate we will never nish,
is in K. for, as the following theorem states, there are innitely
For instance, many possible -values!
theorem 1 rules out
K
the possibility that Theorem 2. Let be a norm with unit
we could come up disk K, and let denote the associated -value.
with a norm where 1)
circles are hearts or Figure 5. The set K determines a
norm in which AK =| r | .
2) if, and only if, K is an ane
stars (a disappoint- regular hexagon (the image of a regular
ment, to be sure), since these shapes are not convex. hexagon under an ane transformation).
Similarly, triangles will never arise as circles under any 3) if, and only if, K is a parallelogram.
norm, since they are not symmetric about the origin. Moreover, for any a with there exists a
However, regular 2n-gons, ellipses, and a host of other set K such that
shapes do work.
The one-to-one correspondence goes as follows. In 1932, Stanisaw Goab proved parts (1)(3) of this
Given a norm the associated set K is the norms theorem (see [2] and also [1] and [4]). The last part of
unit disk (the lled-in unit circle). Conversely, suppose the theorem can be justied succinctly here. Consider
K satises all the required properties. Then we dene a the six-sided polygon Kt in gure 7, where
norm as follows. Every point A can be expressed This set satises the requirements of theorem 1, so Kt
as where P is a point on the boundary of K. determines a norm that we denote by . Since the
In that case, dene (see gure 5). boundary of Kt is polygonal, we can compute the cir-
For example, the regular hexagon H in gure 6 satis- cumference (and hence by summing the distances
es all the conditions of theorem 1. So it is the unit
between the vertices. Thus,
circle of some norm . We can compute the circum-
ference of this unit circle with respect to . By our
previous discussion, the arc length of the line segment
with endpoints A = ( 23 , 21 ) and B = ( 3 , 1 ) is the
2 2
distance between A and B,
The second equality holds because (0,1) lies on the
unit circle (the boundary of H). Similarly, every
side of the hexagon has length 1 with respect to this
norm. So, the circumference of the unit circle is 6, and The penultimate equality follows from the fact that
(1,0), and (0,1) lie on the unit circle, so
1 their norm is 1. Since we can realize all
values in the interval [3,4] as -values.
When Can I Eat Pie?

With our new perspective on distances, circles, and
circumferences, it is apparent that Pi Day celebrations
1 do not have to be conned to March 14. Generalized
-values begin at 3 and range past 3.31, so we can
celebrate Pi Day on any day in March. Moreover, if the
reader missed out on the Pi Day of the Century (or the
Pi Instant) in 2015, there is no cause for disappoint-
Figure 6. The hexagon H determines a norm ment. For instance, if we want to celebrate a Pi Instant
with H = 3. at 12:37:15 on March 25, 2016, we produce a norm

1
Figure 7. The
polygon Kt
determines a norm
with K t = 3 +t.
such that (let

0.2516123715 for the set in gure 7), and the
party is on!
Finally, many of our friends outside the United
States have previously been unable to celebrate Pi
Day because their convention for writing dates (with
the day rst and month second) would place Pi Day
on the 31st of April (31/4), a day that does not exist.
However, with this new perspective, our international
friends can celebrate on the 30th of any monthexcept
February, of course.
In general, changing norms produces not only new
-values, but an entirely new geometry on the plane.
This area of study is called Minkowski geometry. Can
you foresee geometric concepts other than circles that
depend on the norm? See [4] for a full discussion of
Minkowski geometry both on the plane and in higher
dimensions.
Further Reading
[1] J. Duncan, D. Luecking, and C. M. McGregor.
On the values of pi for norms on College Math
Journal 35 no. 2 (March 2004) 8492.
[2] S. Goab, Quelques problmes mtriques de la
gometrie de Minkowski, Trav. lAcad. Mines Cracovie
6 (1932) 179.
[3] J. Rosenthal, Pi Instant, Math Horizons 22 no. 3
(February 2015).
[4] A. C. Thompson, Minkowski Geometry,
Encyclopedia of Mathematics and its Applications,
Cambridge Univ. Press, Cambridge, 1996.Q
Cornelia Van Cott is an associate professor of math-

ematics at the University of San Francisco. She enjoys
thinking about geometry and topology, speaking about
math to children and college students, and celebrating Pi
Day.
Email: cvancott@usfca.edu

THE BOOKSHELF
Pancho will, my love,
The Magic of Math: pick up my new mover
Ginger is a mnemonic
Solving for x and for remembering the
rst 24 digits of pi. And
Figuring Out Why of course, the book also
pays close attention to
Reviewed by Matt Davis some more advanced
topics. My favorite is
A
t first glance, the worlds of magic and
The Magic of i and e,
mathematics seem ill suited for one
which starts with the
another. In a typical magic trick, the
performer manipulates some ordinary denition of a complex
objects in seemingly unremarkable number and eventually
convinces the reader of
ways, until suddenly the surprising outcome is
the truth of Leonhard
revealed. Then the viewer is left to wonder how
Eulers beautiful equation,
the trick was donea secret the magician is honor-
Benjamins style leans toward showmanship rather
bound never to reveal.
than a more academic style that one would nd in a
On the other hand, mathematicians have a tendency
lecture or a textbook. His rst goal is usually to amaze
to announce their tricks beforehand, although they
his readers with a surprising pattern or an elegant
prefer the term theorem. Moreover, most math pro-
trick, and he saves the messy details of why it works
fessors cannot resist the urge to explain how a theo-
for later. Thats an eective approach for a book in-
rem works (although my students occasionally wish I
tended to be accessible to a general audience, although
could!).
naturally, those details are all present for the more
Arthur Benjamin is skilled at blending these two
advanced reader who wants them, usually in asides set
worlds. He has performed his unique Mathemagics
o from the main text.
shows across the country, entertaining audiences in
This theatrical approach means that the pacing of
college lecture halls, TED conferences, and even on
the book is decidedly brisk, which keeps the readers on
The Colbert Report. So its no surprise that Benjamin
their toes. (How else would it be possible to introduce
brings some magical air to his most recent book, The
most of a course in dierential calculus in 23 pages?)
Magic of Math: Solving for x and Figuring Out Why
But thats entirely in keeping with the theme of the
(Basic Books, 2015).
book. Watching a magician repeatedly make an object
The book is a fast-paced tour of 12 broad topics
disappear and reappear leaves us delightfully disorient-
ranging from simple arithmetic to the subtleties of in-
ed. Benjamins approach is similar, whether its page
nite sums, with room along the way for formal subjects
after page of hidden patterns in the Fibonacci num-
such as dierential calculus and recreational math-
bers, or an exploration of all those facts about innite
ematics favorites that include tilings and the Fibonacci
series that just dont quite line up with our intuition.
numbers. Benjamin approaches all these subjects with
The book oers a dizzying array of mathematical
the goal of emphasizing the wonder and magic inher-
delights. But here, once the mathemagician has n-
ent in them, while still giving the reader a sense of the
ished his sleight-of-hand, we have the luxury of peeking
mathematics hiding up the magicians sleeve.
behind the curtain to understand how all those tricks
Although some chapters contain material that will be
are done. Q
well known to any serious math student, every chapter
covers enough ground to surprise most readers with a
Matt Davis is an assistant professor of mathematics
hidden gem. For example, while chapter 1, The Magic
at Muskingum University in New Concord, Ohio. He
of Arithmetic, might sound elementary, it oers magic enjoys board gaming and bad math jokes.
tricks to amaze and deceive your friends, and focuses Email: mattd@muskingum.edu
on techniques of mental arithmetic that Benjamin uses
in his shows to perform lightning calculations.
Similarly, the chapter The Magic of Pi includes an
explanation of how the nonsense sentence My turtle http://dx.doi.org/10.4169/mathhorizons.23.3.28

In addition to the
Genius at Play: The anecdotal diversions, the
biography is interrupted
Curious Mind of by descriptions of serious
mathematics, including
John Horton Conway Conways search for a
universal two-dimension-
Reviewed by Jeb Collins al cellular automaton,
I
t is hard to explain what mathematicians do his discovery of the
and how we think. The general population sees surreal numbers (from
mathematiciansespecially the geniuses of games), and his love of
the fieldas otherworldly figures who are lost symmetries.
in an abstract realm that outsiders can never Other mathematical
penetrate. topics are less impor-
In Genius at Play (Bloomsbury, 2015), Siobhan tant, but no less inter-
Roberts portrays the brilliant mathematician John esting, such as how to determine the day of the week
Horton Conway as a normal human whom we can given any particular datea favorite party trick of
relate to. He enjoys parties, friends, and games just as Conways. Many of the mathematical explanations are
much as, if not more than, anyone else. She brings this direct quotes from Conway and, in typical Conway
mathematical genius down to earth and shows that he fashion, are boiled down to their simplest form.
is a man with a great passion and ability for math. Roberts excellently explains Conways attraction to
This book is challenging to classify. Obviously, it is a mathematics and, in doing so, shows why mathema-
biography about John Conway, inventor of the famous, ticians love math. We see his love of the simplicity
or infamous if you talk to Conway, Game of Life (he of mathematics, his excitement at the surprises that
resents that the Game of Life is his best-known schol- occur when examining a problem, and his desire to
arly contribution). It roughly follows the linear time- understand the why, not just the what, of a theo-
line of Conways lifefrom a student at Cambridge, rem. We see his pursuit of simpler proofs and explana-
through his many mathematical discoveries as a profes- tions so that he can better understand a problem. By
sor, to the present where he is retired from Princeton describing Conways fascination with mathematics,
University. Roberts reveals what draws all mathematicians to their
subject.
But, possibly due to the subjects own ability to be
Genius at Play is a genuine page-turner. The nar-
easily distracted, the narrative jumps around consider-
rative of Conways life jumps around so much, the
ably. The tone of the book is informal and conversa-
reader is never sure what will come next. The stories,
tionalreading it is like listening to two friends chat-
anecdotes, and humorous comments from Conway
ting. The author narrates the story, but direct quotes
keep things interesting. His personality shines through,
from Conway are scattered throughout, giving the feel-
and after reading Genius at Play, the reader feels
ing of a live interview. Moreover, Roberts inserts her-
as though she has met the man. More importantly,
self into the storywriting about her struggles manag-
though, the reader can better understand the strange
ing Conways creative memory and distractibility.
and elusive creature that is the mathematician. Q
Conway seems to have more stories than Schehera-
zade. For example, as a student he read about a he-
Jeb Collins is a strange and elusive assistant professor
reditary skill involving tongue gymnastics. After three
of mathematics at West Texas A&M University.
hours in front of a mirror, he discovered he was one of
Email: jcollins@wtamu.edu
the rare individuals who could perform all four tongue
tricks.
My favorite diversions are when Conway shows o
his knowledge of etymology. He frequently interrupts
his stories to give an esoteric description of the origin
of a word (esoteric, by the way, derives its meaning
from Pythagorass exclusive teaching methods). http://dx.doi.org/10.4169/mathhorizons.23.3.29

Brida Anderson/Flickr
THE PLAYGROUND
Welcome to the shown in gure 1.) The process continues ad innitum.
Playground. Playground What percentage of the square will eventually be con-
rules are posted on page tained inside one of the circles?
33, except for the most
important one: Have fun!
THE SANDBOX
In this section, we highlight problems that anyone can
play with, regardless of mathematical background. But
just because these problems are easy to approach doesnt
Figure 1. A square
necessarily mean that they are easy to solve! with circles
inscribed in the
Problem 334. Call a positive integer n good if corners.
the sum of n consecutive integers could be a perfect
square, and bad otherwise. For example, 3 is good
because In Square Sums, you are
THE JUNGLE GYM
asked to nd all bad numbers. Any type of problem may appear in the Jungle Gym
climb on!
Problem 335. This problem comes to us from
Mehtaab Sawhney of Commack High School. In Triple Problem 337. In a calculus class we learn that the
Inequality, Mehtaab asks you to prove the following. alternating harmonic series
Let n be a positive integer and let (a1,a2,,an), converges to ln(2). In Alternating Madness, we
(b1,b2,,bn), and (c1,c2,,cn) be three nite sequences consider three related alternating series.
of nonnegative real numbers. Prove that 1) Show that
converges, and nd the exact value of the sum.

2) Does
where the sums are over all possible i, j, and k (each

ranging from 1 to n), including repeated subscript converge or diverge?
values. 3) How about
THE ZIP-LINE
This section oers problems with connections to ar- For part 2, each new block has one more term than the
ticles that appear in this issue. Not all of the problems previous block. For part 3, each new block has twice as
in this section require you to read the corresponding many terms as the previous one.
articles, but doing so can never hurt, of course.
THE CAROUSEL
Problem 336. In (Always) Room for One More, OLDIES BUT GOODIES
(page 8) Christopher Ennis places disjoint circles in a
In this section, we present an old problem that we like
circle. In Round Pegs in a Square Hole, circles are
so much, we thought it deserved another go-round. Try
placed in a square as follows. First, we inscribe a circle
this, but be carefulold equipment can be dangerous.
in the square. At step n, we add four circles in the
Answers appear at the end of the column.
four corners so that each circle added is tangent to two
sides of the square and to a circle that was added in You are given a cube of cheese (or tofu, for our
step (The rst three steps in this procedure are vegan readers) and a sharp knife. What is the largest

number of pieces one may decompose the cube using Problem 327. A two-person game is played as fol-
n straight cuts? You may not rearrange the pieces lows. A rectangular box with dimensions
between cuts! is lled with unit cubes, where
each unit cube has a unique (x,y,z)-coordinate in the
box. Alice selects a cube (a,b,c) and removes it; this
SEPTEMBER WRAP-UP also removes all the cubes in the three lines through
Problem 326 came to us from Alan Loper and this cube parallel to the edges of the box. Then Bob
Greg Oman. In One-Sum, they asked you to does the same thing, choosing a unit cube from what
nd all positive integers n such that the base 10 remains, removing it and everything else along the
three lines determined by his cube. The players alter-
representation of consists of all ones.
nate turns, each time choosing a cube as above.
We received solutions from Brian D. Beasley
Assume that each player takes as many cubes as
(Presbyterian College), C. J. Dowd (San Francisco
possible each turn. There is no winner; the players
University High School), Dmitry Fleischman, Chris
hug at the end. Grabbing Cubes asked three
Jones (Horace Mann School), Henry Ricardo (New
questions.
York Math Circle), Randy Schwartz (Schoolcraft
1) How long will this game last?
College), Ryan Suby (North Central College), and the
2) How many cubes will each player end up with?
Northwestern University Problem Solving Group.
3) Note that 2,015 is the product of three distinct
The only solution is There were two
odd primes. Will I still be editing this column the
distinct ways the solvers attacked this problem.
next year this happens?
All solvers noted that, if the equation is true, then We received solutions from Taylor Absell (Twin
One group of solutions Rivers PMP Solving Team), Jamie Curry (North
used the quadratic equation to solve for n. They then Central College), and Dmitry Fleischman.
showed (using a variety of techniques) that n is not All solvers determined the game will last exactly 65
rational unless In particular, this solution moves. Its clear this is sucient; removing 65 cubes
follows from the fact that the number is with addresses (a,b,*) where and
never a perfect square. Brian D. Beasley and Henry will remove all the cubes. Necessity follows from show-
Ricardo mentioned that this result follows from John ing that the algorithm for removing the maximum
H. Jaromas paper Triangular RepunitThere Is But number of cubes at each step takes 65 steps.
1, Czechoslovak Mathematical Journal 60 (December The algorithm employed by all the solvers removed
2010) 10751077. 47 cubes for each of the rst ve steps, 45 cubes for
The second solution used the following equation: each of the next ve, then 43 for each of the next
Then one can argue that three, and so on, as in the table. Jamie points out
we must have consecutive integers and that the number removed drops by 2 whenever the
where one is a power of 2 and the other is a power step number passes a new multiple of 5 or 13. The
of 5. This works only when although this also number of cubes removed at each step of the algo-
takes some analysis. We omit the remaining details. rithm is given in table 1.
Steps 15 610 1113 1415 1620 2125 26 2730
Number
of Cubes 47 45 43 41 39 37 35 33
Removed
Steps 3639 40 4145 4650 5152 5355 5660 6165
Number
of Cubes 29 27 25 23 21 19 17 15
Removed
Table 1. The number of cubes removed at each step of the algorithm.

There was a small discrepancy between the total these primes in RN as and now set
number Taylor and Jamie gave to each player and the This nishes the problem.
number Dmitry gave: Taylor and Jamie had Alice with
1,023 and Bob with 992, while Dmitry had 1,024 for Problem 329. For each positive integer n, draw
Alice and 991 for Bob. We leave this discrepancy to the quarter-circle in the rst quadrant of
the interested reader. the xy-plane. Let a n be the number of positive integer
Finally, note that 2035 is the next year that lattice points on this quarter-circle, that is, points (a,b)
is the product of three distinct odd primes where both a and b are positive integers.
I will not be editing this column in In Supernova, you had two tasks:
2035. I promise. 1) Show that for innitely many k.
2) Show that for innitely many k.
Problem 328. In PrimeTime, you were asked to Extra credit: Can you say anything about the
show that there is a prime number p and an innite number of times for
sequence of positive integers so that C. J. Dowd (San Francisco University High School)
every term in the sequence and the Northwestern University Problem Solving
is prime. Group solved the entire problem, including the extra
We received solutions from Ethan Acklesberg (Bard credit. Dmitry Fleischman solved parts 1 and 2, and
College at Simons Rock), Dmitry Fleischman, Taylor also did the case. We also received solutions
Spino (North Central College), and the Northwestern to parts 1 and 2 from Kalleigh Kuehl (North Central
University Problem Solving Group. College) and Charlie Michel (Ashland University), and
Three of the solutions used Dirichlets theorem a solution to part 1 from Christopher Havens (Twin
on primes in arithmetic progressions, but, as Rivers PMP Solving Team) and Randy Schwartz
Ethan Ackleberg noticed, all you need is the (Schoolcraft School).
pigeonhole principle. For let For part 1, its easiest to consider k-values congruent
be the positive integers to 3 mod 4. Note that for any integer m, m2 is
congruent to n modulo 2,015. Since there are an congruent to 0 or 1 mod 4. Then, if there
innite number of primes, one of these classes, say are no integer solutions to Thus
RN, must include an innite number of primes. Write for innitely many k.
krungchingpixs/123RF

For part 2, let m be a positive integer and let
Then if we have a solution to
To see that there are no other solutions,
lets count the number of times 2 divides each side of
the equation. If both x and y are even and then
2 will divide an even number of times, but 2
divides k a total of times. If both x and y are
odd, then mod 4, so 2 divides
exactly once. Since 2 divides k at least 3 times, we
conclude there are no solutions when We
conclude has an innite number of solutions.
For the extra credit problem, we show that
has an innite number of solutions for all
Both C. J. Dowd and the Northwestern group used
a head wins. Let p(q,r) be the probability that Alice
Jacobis 2-square theorem that gives the exact number
wins. Now assume that q and r are chosen randomly
of solutions to Applying this theorem,
and independently from a uniform distribution on [0,1].
we rst x n and i, and then set Then
In Self-Referential Probability, they ask if there is
will have exactly n solutions; so
a value x such that: x equals the probability (as q and
for innitely many k.
r vary) that Solutions to this problem are
CAROUSEL SOLUTION now due May 31, 2016.
We also provide an update on problem 325 from the
Let an be the maximum number of pieces produced April 2015 Playground, Menger Madness. In that
by n cuts. The sequence begins 1, 2, 4, 8, 15. problem, you were asked to nd the shortest path
This is a good example of a sequence that will fake you joining opposite corners of the Sierpinski carpet and
out if you stop before the Menger sponge. In the November 2015 issue, we
Slicing the block with n cuts creates a maximum reported that the shortest path for the Menger sponge
of regions. The general formula, found by readers had length However,
due to Schlai (1901), gives the maximum number of the problem proposers, Ethan Berkove and Derek
regions n hyperplanes can decompose Let a n,d Smith, have recently found innitely many paths,
be the number of regions determined by n cuts in d all of a certain length less than 2, and have proved
dimensions. Then that no other path can have a shorter length. More
information about the paths and their method of proof
Heres a sketch of the proof of this beautiful formula. are forthcoming. Q
First use a geometric argument to prove the recursion http://dx.doi.org/10.4169/mathhorizons.23.3.30
relation George Plya helps his
students discover this relation in the lovely lm Let Us
Teach Guessing (MAA, 1966). Then the result follows Submission & Contact Information
by induction. Playground features problems for students at the
undergraduate and (challenging) high school levels. All
problems and/or solutions should be submitted to Gary
CLEANING UP Gordon, Mathematics Department, Lafayette College,
Easton, PA 18042. Electronic submissions (PDF format
There was a typo in the statement of Problem 331 in preferred) may also be sent to gordong@lafayette.edu. Please
the November 2015 issue. Here is the correct statement include your name, email address, and school affiliation,
of the entire problem, which is due to Steven J. Miller and indicate if you are a student. If a problem has multiple
(Williams College) and Stan Wagon (Macalester parts, solutions for individual parts will be accepted. Unless
College). Alice has a coin that comes up heads with otherwise stated, problems have been solved by their
probability q; Bob has another coin, with heads proposers.
appearing with probability r. They alternate tossing The deadline for submitting solutions to problems in this
the coins with Alice going rst. The rst person to get issue is May 31, 2016.

AFTERMATH
The Law of the Broken Futon
Ben Orlin
T
ry asking random adults about their math satisfying the equation
education. They refer to it like some sort then youre a broken futon.
of NCAA tournament. Everybody gets Youre missing a piece upon
eliminated, and its only a question of how which future learning will
long you can stay in the game. I couldnt crucially depend. Quadratics
handle algebra signies a rst-round knockout. I will haunt you; the sine
stopped at multivariable calculus means Hey, I didnt curve will never make sense;
win, but Im proud of making it to the nal four. and youll bail after cal-
Its as if each of us has a mathematical ceiling, a culus, consoling yourself,
cognitive breaking point, beyond which we can never Well, at least my ceiling
advance. was higher than some.
But theres a new orthodoxy among teachers, an ac- Why not wait to add
cepted wisdom that just about anyone can learn just the missing piece later,
about anything. It takes grit, eort, and good instruc- when its actually needed?
tion. But eventually, you can bust through any ceiling. Because thats much harder.
I love that optimism, that populism. But if theres no In the intervening years,
such thing as ceilings, then what do students keep thud- you develop shortcuts that
ding their heads against? do the job, but warp the
Is there any way to bridge this canyon-wide gap in frame. Youll need to un-
views? learn these workarounds Ben Orlin
I believe there is: the Law of the Broken Futon. bending the futon back into its original shapebefore
In college, my roommates and I bought a lightly used you can proceed.
futon. Carrying it up the stairs, we heard a crack. A Once under way, damage is hellishly dicult to
little metallic bar had snapped o. The futon seemed undo.
newe couldnt even tell where the piece had come This, I believe, is the ceiling so many students expe-
fromso we simply shrugged it o. rience in high school and early college. Its not some
After a week, the futon had begun to sag. Did it inherent limitation of their neurology. Its something we
always look like this? we wondered. create. We create it by prizing right answers over deep
A month later, it was embarrassingly droopy. Its cur- reasoning. We create it by saying, Only clever people
vature dumped all sitters into one central pig-pile. will get it; everyone else just needs to be able to do it.
And by the end of the semester, it had collapsed in a We create it by saying, in word or in deed, Its OK not
heap on the dorm room oor. to understand. Just follow these steps and check your
Now, Ikea furniture is the fruit y of the living room: answer in the back.
notoriously short-lived. There was undoubtedly a ceiling We may succeed in getting the futon up the stairs.
on our futons lifespan, perhaps three or four years. But But something is lost in the process. Moving forward
this one survived barely eight months. without key understandings is like marching into battle
In hindsight, its obvious that the broken piece was without replacement ammo. You may re o a few
crucial. The futon seemed ne without it. But day by rounds, but by the time you realize something is miss-
day, butt by butt, weight pressed down on structures ing, itll be too late to recover.
never meant to bear the load alone. The framework A student who can answer questions without under-
warped. The futons internal clock was silently ticking standing them is a student with an expiration date. Q
down toward an inevitable failure. Ben Orlin is a teacher in Birmingham, England. His
And, sadly, so it is in math class. blog is Math with Bad Drawings.
Say youre acing eighth grade. You can graph lines, Email: ben.orlin@gmail.com
compute slopes, specify points. But if youre missing one
vital understandingthat these graphs are the x-y pairs http://dx.doi.org/10.4169/mathhorizons.23.3.34

FASTER FORWARD TO
WEB EXPERIENCES THAT
KNOWS NO BOUNDS.
With organizations under more pressure than ever to provide an instantaneous, personalized and secure
online experience on any device, we need to re-think not only our content, but how its delivered.
From content delivery to transaction security and fending off cyber attacks, Akamai provides solutions for
the worlds most recognizable brands - in categories like entertainment, sports, gaming, nance, retail,
software, social media, education and others.
We helped NBC Sports Digital deliver gold during the 2014 Winter Olympics.
We enabled RedHat to expand their global reach.
Weve managed Black Friday web trafc for the biggest retailers on the planet.
LET AKAMAI HELP YOU MOVE FASTER FORWARD.

AKAMAI.COM
Media Delivery

Math Horizons 02 2016

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Math Horizons 02 2016

Uploaded by

Copyright:

Available Formats

Circles, Circles,

Pi Day: Silly or Inspiring? | 16

Published by the Mathematical Association of America : : February 2016

Puzzle #1. Solution to puzzle #1.

2 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

8 (Always) Room for One More

16 Pi Day, Pro and Con

24 A Pi Day of the Century Every Year

www.maa.org/mathhorizons : : Math Horizons : : February 2016 3

EDITOR EDITORIAL BOARD

FB: facebook.com/MAAMathHorizons Alissa S. Crans Pamela Pierce

4 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

www.maa.org/mathhorizons : : Math Horizons : : February 2016 5

6 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

Jessica Wynne KM: You didnt know upon graduation from

www.maa.org/mathhorizons : : Math Horizons : : February 2016 7

8 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

2(c) above, 2(d) and 2(e), right, 2(f)

considerable Mathematically this is equivalent to the frame being

www.maa.org/mathhorizons : : Math Horizons : : February 2016 9

such that the length of the nth extended interval is

10 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

www.maa.org/mathhorizons : : Math Horizons : : February 2016 11

Run, Hero, Run! Mike Naylor

Hero, hero, hero

www.maa.org/mathhorizons : : Math Horizons : : February 2016 13

14 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

www.maa.org/mathhorizons : : Math Horizons : : February 2016 15

16 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

Learn, Dream, Play

www.maa.org/mathhorizons : : Math Horizons : : February 2016 17

18 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

The Shape of Utopia

www.maa.org/mathhorizons : : Math Horizons : : February 2016 19

By symmetry, Utopias centroid is for some

Think of the inner harbor as having negative area

Finally, the perimeter of the harbor is 500 miles, so

20 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

[1] B. R. Goodey, Mapping Utopia: a comment on

www.maa.org/mathhorizons : : Math Horizons : : February 2016 21

22 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

visualize how there will be an identical square tunnel in

www.maa.org/mathhorizons : : Math Horizons : : February 2016 23

24 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

Thats a mouthful! Lets look closely

www.maa.org/mathhorizons : : Math Horizons : : February 2016 25

When Can I Eat Pie?

26 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

such that (let

Cornelia Van Cott is an associate professor of math-

www.maa.org/mathhorizons : : Math Horizons : : February 2016 27

28 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

www.maa.org/mathhorizons : : Math Horizons : : February 2016 29

converges, and nd the exact value of the sum.

where the sums are over all possible i, j, and k (each

30 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

Steps 15 610 1113 1415 1620 2125 26 2730

Steps 3639 40 4145 4650 5152 5355 5660 6165

www.maa.org/mathhorizons : : Math Horizons : : February 2016 31

32 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

www.maa.org/mathhorizons : : Math Horizons : : February 2016 33

34 February 2016 : : Math Horizons : : www.maa.org/mathhorizons

We enabled RedHat to expand their global reach.

LET AKAMAI HELP YOU MOVE FASTER FORWARD.