PI in the SKY

ISSUE 13, FALL 2009

IN THIS ISSUE:
THE BOARD GAMING GENIUS: REINER KNIZIA
ELEVATOR RIDES IN PURGATORY
HEXAGONAL CODES
THE ABORIGINAL GAME OF LAHAL

Pacific Institute for the
Mathematical Sciences
On the Cover
The Egyptian game Senet, one of the oldest board games, dates
back as early as 3500 BCE. This board, from the Brooklyn Mu-
seum, is inscribed with the name, Amunhotep III.
Editorial Board
John Bowman (University of Alberta)
Tel: (780) 492-0532, E-mail: bowman@math.ualberta.ca
Murray Bremner (University of Saskatchewan)
Tel: (306) 966-6122, E-mail: bremner@math.usask.ca
John Campbell (Archbishop MacDonald High, Edmonton)
Tel: (780) 441-6000, E-mail: jcdotcalm@shaw.ca
Florin Diacu (University of Victoria)
Tel: (250) 721-6330 , E-mail: diacu@math.uvic.ca
Sharon Friesen (Galileo Educational Network, Calgary)
Tel: (403) 220-8942 , E-mail: sfriesen@ucalgary.ca
Gordon Hamilton (Masters Academy and College, Calgary)
Tel: (403) 242-7034, E-mail: gamesbygord@gmail.com
Klaus Hoechsmann (University of British Columbia)
Tel: (604) 822-3782, E-mail: hoek@pims.math.ca
Dragos Hrimiuc (University of Alberta)
Tel: (780) 492-3532, E-mail: dhrimiuc@ualberta.ca
Michael Lamoureux (University of Calgary)
Tel: (403) 220-8214, E-mail: mikel@math.ucalgary.ca
Mark MacLean (University of British Columbia)
Tel: (604) 822-5552, E-mail: maclean@math.ubc.ca
Patrick Maidorn (University of Regina)
Tel: (306) 585-4013, E-mail: Patrick.Maidorn@uregina.ca
Anthony Quas (University of Victoria)
Tel: (250) 721-7463, E-mail: aquas@uvic.ca
Wendy Swonnell (Greater Victoria School District)
Tel: (250) 477-9706, E-mail: wswonnell@shaw.ca
Editorial Coordinator
Adam Wojtowicz
(604) 822-0402, E-mail: media@pims.math.ca
Contact Information
Pi in the Sky
PIMS University of Victoria Site Office, SSM Building Room A418b
PO Box 3060 STN CSC, 3800 Finnerty Road, Victoria, BC, V8W 3R4
T/(250) 472-4271 F/(250) 721-8958, E-mail: pims@math.uvic.ca
Significant funding
for Pi in the Sky is
provided by INNOVATION AND SCIENCE
Pi in the Sky is a publication of the Pacific Institute for the Mathematical
Sciences (PIMS). PIMS is supported by the Natural Sciences and Engineering
Research Council of Canada, the Province of Alberta, the Province of British
Columbia, the Province of Saskatchewan, Simon Fraser University, the University
of Alberta, the University of British Columbia, the University of Calgary, the
University of Lethbridge, the University of Regina, the University of Victoria and
the University of Washington.
Pi in the Sky is aimed primarily at high school students and teachers, with
the main goal of providing a cultural context/landscape for mathematics.
It has a natural extension to junior high school students and undergradu-
ates, and articles may also put curriculum topics in a different perspec-
tive.
Submission Information
For details on submitting articles for our next edition of Pi in the
Sky, please see:
http://www.pims.math.ca/resources/publications/pi-sky
Editorial: David Leeming, Managing Editor, Pi in the Sky
We welcome our readers to another issue of Pi in the Sky. In addition
to being on-line at www.pims.math.ca/pi , we have about 1200 subscrib-
ers in 56 countries.
Once again, we were faced with the challenge of trying to publish
articles that can be understood by capable high school and College-level
students. In this context, we are constrained by the articles we receive
from authors. This issue contains articles on mathematical games, ge-
ometry, face recognition, the Laws of Physics and elevator rides (in pur-
gatory!). We also have a book review written by high school students.
In this issue, we announce the winners of the Math Challenge posed in
Issue #11 (Spring 2008).
One of our Editors, Volker Runde, has resigned. Volker has served the
magazine well for a number of years and we thank him for his dedicated
service to Pi in the Sky. We welcome Murray Bremner, University of
Saskatchewan, to the Editorial Board.
This is my last issue as Managing Editor of Pi in the Sky. I have thor-
oughly enjoyed my five years in the position. I would like to acknowl-
edge the support of my Editorial Board for reviewing articles submitted
and for their efforts in proofreading drafts of pending issues. I leave the
magazine in the capable hands of Anthony Quas, University of Victoria,
who will become Managing Editor of Pi in the Sky starting with Issue
#14.
Table of Contents
The Aboriginal Game of Lahal
by Peter J. Dukes and David J. Leeming ...................................... 1
Brad Pitt gets Angelina Jolie and I get ... Math?
by P. H. Trinh ............................................................................... 4
Hexagonal Codes
by Anthony Quas and Peter J. Dukes .......................................... 7
Snell on the Beach
by Klaus Hoechsmann ............................................................... 13
Genius Board Game Inventor: Reiner Knizia
by Gordon Hamilton ................................................................. 15
Higher Derivative Equations and the Laws of Physics
by Neil Barnaby ......................................................................... 18
Heaven to Hell: Elevator Rides in Purgatory
By Gordon Hamilton ................................................................. 22
Book Reviews ............................................................................ 24
Sacred Mathematics: Japanese Temple Geometry
Reviewed by Tom Archibald
Uncle Petros and Goldbach's Conjecture
Reviewed by Clara Park
Pi in the Sky Math Challenges ................................................... 27

Peter J. Dukes
A Simple Version of Lahal
David J. Leeming1
In [1], the authors state that
\Mathematics is a subject where
Aboriginal students can feel particularly isolated
and alienated." There are, however, some examples
of Aboriginal games and activities that provide
opportunities for exploration of the underlying
mathematical ideas. One such game is Lahal, also
known as `the bones game.' It is usually played by
two teams and each team could have anywhere
from three to twenty members. The game consists
of bones (typically two male, two female) and
sticks (six or more). An excellent description of
this and other versions of Lahal is contained in
[1].
Here, we will present a simple version of Lahal
for two players. Each player begins with six sticks
and one bone. They takes turns with one player
(the hider) hiding the bone in one of his hands
and the other player (the pointer) guessing which
hand contains the bone. If the pointer guesses
correctly he gets a stick from his opponent. If he
guesses incorrectly, the hider gets a stick from his
opponent. The game continues until one of the
players has all twelve sticks. Naturally, we assume
a 50% probability that the pointer can correctly
guess the hand containing the bone.
The outcomes of the first six rounds of this
game are shown in Table 1. Specifically, the entry
in the kth column of round n indicates the number
of possible game sequences which arrive at the
outcome of one player holding k sticks. It is easy
to compute each successive row in the table as it
is the sum of the entries to the right and left in
1
Both authors are with the Department of Mathematics and Statistics,
University of Victoria, Victoria, B.C., CANADA V8W 3R4
e-mail: dukes@uvic.ca, leemingd@uvic.ca
the row above. So, for instance, we have, in round
5 (row five - we count the first row as row zero)
the entry 10 in column k=5. This is the sum of the
numbers 4 and 6 above it in row 4. This corresponds
to ten (5, 7) stick arrangements, coming from four
(4, 8) arrangements and six (6, 6) arrangements in
the previous round.
The array of numbers in Table 1 is known as
Pascal's Triangle. Later, we will describe another
way to find the entries in Pascal's Triangle using
By
binomial coefficients. Note that the sum of all the
and entries in each row is a power of 2, more precisely,
in row n the sum of the entries is 2n. To calculate
the probability of holding k sticks after n rounds,
n ≤ 6, we divide the entry in the kth position by
2 n.
For example, after six rounds, since there are
6
2 =64 different ways to arrive at one of the seven
outcomes, so the probability of winning all six
sticks from the pointer (or losing all six sticks to
the pointer) is 1/64 and the probability that each
player still holds six sticks is 20/64 = 5/16.
Note that it takes an even number of turns to
win a game. (Why?) After six rounds, the hider
could be holding 0, 2, 4, 6, 8, 10 or 12 sticks. After
six rounds, the game may be over, with the hider
winning or losing. In case the game is not over,
we cannot rely entirely on Pascal's Triangle any
longer to calculate the probabilities of a win or a
loss. Let us assume now that the game proceeds
to eight or more rounds. The number of ways of
reaching one of the outcomes after seven, eight,
nine or ten rounds is shown in Table 2 below the
line.
To calculate the probability of a particular
outcome, for example,
that each player has
six sticks after eight
rounds, we find the
sum of the numbers in
row eight which will
be the denominator
of the fraction we use.
That sum is 248, so the
probability that each
player has six sticks is
70/248 = 0.282. Using
the same reasoning (you
can fill in the steps)
http://brownsharpie.courtneygibbons.org/
1
 is 70/248 = 0.282. Using the same reasoning
(you can fill in the steps) we see that after ten The answers to these questions requires prob-
read ‘n choose k’) give us rounds,
the n + 1 theentries in y
probability that each player has six ing deeper into the mathematics behind this sim-
the nth row of Pascal’s
we see that after sticksTriangle. A
ten rounds,very useful
the = 0.267. Calculate the proba- may We be will
useful to you
identity
is 252/944 ple game. pursue theseinandreading
other ques-
�n �inn calculating
 �n+1  theeach
binomial coefficients
isprobability that bility player
k + k+1 = k+1 . See if you �can
that each
prove has
this player has six sticks after twelve what follows.
tions in Section 2 of this article.

six sticks
identity using theisfactorial rounds,
252/944form k =
n
ork!(n−k)!
= after. fourteen rounds. What is hap-
0.267.
n!
If you First,
are familiar
let us with
examinebinomial coefficients,
factorials.
�n
Calculate
From this, we seethe k pening
actually to
that probability these
counts the probabilities
that as the number of then you
Formay wish to go directly to Section
n, the 2. If
10

number of k-element subsetsrounds

of an n-element set.
any non-negative integer
each player has six sticks increases?
after you are not, then this brief introduction
number n factorial whose symbol may be
For this
�n reason, there is a standard convention
twelve
that rounds, or after We could fourteen
k is deemed equal to zero if k > n or
continue this process to find the usefuliston!you
is in readingbywhat
defined 0! =follows.
1, 1! = 1,
< 0. Pascal’s
krounds. What probabilities
Triangleisgives fortothe number of sticks held by the
us a triangular
happening First,
n! = n(n - 1) . . . (2)(1), For
let us examine factorials. so any
for non-
display of the nonzero binomial
these probabilitieshider
Table 1).
as coefficients
the the(see
andnumber pointer after any even number of negative integer n, the number n factorial whose
instance 4! = 24, 9! = 362880,
of rounds increases? rounds. However, there are some obvious ques- symbol is n! is defined by 0! = 1, 1! = 1,
and so on.
tions that any mathematician reading this would n! = n(n − 1) . . . (2)(1), so for instance 4! = 24,
2 We Analysis
could and Generalizations
continue this process
be asking by now. 9! = 362 For
880, and so on.
nonnegative integers n
to find the probabilities for the 14 x
It is helpful to draw a diagram which depicts a Forand k, 0 ≤ kintegers
nonnegative ≤ n, the n and k, 0 ≤
binomial � k ≤ n,
number
game such asof sticks
lahal. Call held
the twoby
1. Is the Xhider
players
there aand
formula for computing
Figure the number
1: Various minimal the binomial
lattice paths coefficients,
coefficients, denotedby
denoted by nk (and
, and associate
Yand them with the
the pointer x- and y-axes
after any ofeven a Figure 1: Various minimal lattice paths
Cartesian grid. Suppose instead of sticks trading (and‘nread
read choose `n choose
k’) givek') usgive
the usn +the1 entries i
number
back and forthofthat rounds.
we simplyHowever,
keep track of there
how 2 in‘n y th y
read ‘n choose on k’) give the
or below the ynn
us line =+ + 1(Inentries
1 entries
x. read
lahal, in
the
this the
choose rowrow
nthnk’) give ofusPascal's
of Pascal’s the n + Triangle.
Triangle.
1 entriesA veryin usefu
many times Yobvious
are some has won, versus how many
questions that times
any mathematician
X has won. Each point (a, b) inthe the nth row of
Cartesian Pascal’s
means Triangle.
one player is alwaysAAwinning
very useful
very useful
throughout
the identity
nthidentity �inn calculating
in calculating
row of Pascal’s  �Triangle.  the Abinomial
the binomial
very usefulcoefficient
reading this would be asking by the game.) Counting the number of subdiago- �n
now. n+1
plane, where a and b are nonnegative identity �inn calculating
� integers, nal �minimal the binomial coefficients
coefficients identity
lattice paths is a famous �problem
is is  + �k+1 =the
 �innkcalculating k+1 . See
binomial Seeif coefficients
you
if canyouprove th
�n
corresponds
1. Is thereto a wins for playerisX
a formula fornand+ b wins
computing = n+1
the . See
number
k+1 in combinatorics. if you can
can
It turns k’) prove
prove this
n
isusnumber
this +identity
identity n+1
+ 1 =entries
k+1 .the
usingusing inSeefactorial
ify you can
the form
factorial prove
form this n!
k = k!(n−k)
k
for player Y . There have been a+b rounds played read ‘n choose
k+1 �nout that the
 give kthe nk+1 � � 
ofthe
paths on orsticks
below line y = = x, from (0, 0) tousing the factorial form n
has eitheridentity or using factorial form .. From
n! n
sooffar,
sequences resulting ingained
a given lost number the ofnth identity From this, we
we see seethat =
that k k actually n!
.
counts th
�nrow of Pascal’s Triangle. A very
and each player k k!(n−k)! this,useful actually
k!(n−k)!
(n, n), equals the nth Catalan number, or �n
|afor
− b|each
sticks.player?
In this model, the From‘equilibrium’
this, we seeidentity that in actually
calculating
counts counts the
From
the the
binomialnumber
this,
number of
10 wecoefficients
see
of that
k-element
k-element subsets
actually
subsets of an
counts
of anthe
n-element 10se
state is the line y = x, where each player has the �n k� n 1 �2n  k
2. Will the game number
almost of
surely k-element
end isinsubsets
a +
finite
C = of an= n+1
n-element
n-element. . See
numberset.
if
set. youFor
of
For this
can
this
k-elementreason,
prove this
subsets
reason, there
there of isan
isan-element
astandard
standard conventio
set.
same number of sticks. k n k+1n + 1 k+1 n �n�n n!
For this reason, there is a standard
identity using the convention
For
factorial form this that
reason,
= is
theredeemed
is a equal
standard to zero if
convention k > n o
Each gameof
number of turns?
lahal unfolds as a ‘path’ �nin the convention
�  �n k is
that k!(n−k)! . equal to zero if k >
k deemed
Cartesian grid, from (0, 0) to (a, that b), the is
cur- deemed
See theequal
excellent
Fromto this,to zero
reference
we n if
[2]kfor>
seeorthethat <that
more n or
details k < 0.
k is deemed
Pascal’s
counts equal Triangle
to zero gives
if k > n triangula
us a or
k actually thegives
rent 3. If not,
situation. is along
Moves therethese a k
limiting
paths are on value
minimal latticethepaths and
kCatalan0. Pascal's
num-
Triangle 10 us a triangular
k < 0. Pascal’s Triangle gives us a triangular
k < 0. display
Pascal’sof the
Trianglenonzerogives binomial
us a coefficients
triangular (se
either east or north, according to whether X bers. number of k-element
probabilities? display subsets
of the of an nonzero
n-element set.
binomial coefficients (see
display of the nonzero
Let’sFor binomial
this coefficients (see
display Table
of the 1).
nonzero binomial coefficients (see
�n kreason, there 1). is aa general
standard convention
or Y (respectively) wins a given round. Such return to lahal. TableSuppose in
paths,The which never ‘backtrack’
answers to these Table
are called1).
questionsmini- requires
lahal game
that with
probing is deemed equal toafter
sticks per player, Table
that 1).
zero if k > n or
mal �paths. Just as in our early discussion, there 2n turns one kplayer has k + r sticks, and the
deeper
a+b
 into the mathematics behind this
k <Analysis and Generalizations
simple
0. Pascal’s
2 Analysis
are a paths from (0, 0) to (a, b). This is ath other has k − r sticks, 0 < r < k. Triangle
and Generalization gives us a triangular
game. We will pursue these
entry in the (a + b)th row of Pascal’s Triangle,
2 and other Thedisplay
number ofof
questions
Analysis and Generalizations thethisnonzero
in
ways can occurbinomial
2 Analysis and Generalizations is in one-to-coefficients (see
where
Section indexing
2 ofentries
thisand rows begins with 0. one correspondence
article. Table 1). with the number of minimal
It is helpful It
to isdraw
helpful to draw awhich
a diagram diagram whicha depicts
depicts
Figure 1 shows various minimal paths from (0, 0) lattice paths from (0, 0) to (n+r, n−r) which do 14 x
If you are familiar with binomial
to (14, 10). It is helpful to draw a diagram
coefficients,
not intersect which
game depicts
It
such
either of the lines y = x−k and y = is asa game
helpful
Lahal. such
to draw
Call as alahal.
the diagram
two Call the
which
players two players
depicts
X and a X an
One player wins when the game’s
then you may wish to game path such
go reaches
directlyas xlahal.
+ k. Callthethe
to Like
the two Y,
Catalan
next players
restriction game
X
and associate andthis
above, Y ,them
such and associate
as lahal.
withCall thethem
the withplayers
x- two
and the x- and
y-axes X ay-axes of
of and
a ‘boundary’, which is a certainYdiagonal line. is another
, and associate
2 Analysis and Generalizations
problem
them with theonx-lattice
and paths,
y-axes now
of
Y , grid. awith Cartesian
and associate Figure grid.
them 1: Suppose
Various
with the instead
minimal
and y-axesof sticks
lattice a tradin
of paths
Section. Cartesian Suppose instead of x- sticks trading
We must nowIfdiscuss
you these
are not, thenfurther.
boundaries this brief two introduction
linear boundaries. See Figure 3 for n = 12,
Cartesian grid. Suppose
A minimal lattice path is subdiagonal if it lies r = 2, It
instead of sticks trading
k =is6.helpful to draw a diagram backback
Cartesian and
whichand
grid. forth that
Suppose
forth
depicts that we
a insteadwesimply
of keep
sticks
simply track of how
trading
keep
0 1 4
back2 3
5and forth
6 7that 8we simply
9 10 11 track 12 of many times has won,keep versus how many time
game such askeep lahal. Call backhow
the and
two
track onforth
players
oforhow that we
Yand
Xmany
below thesimplytimes
line y Y=trackhas
x. won, of how
(In lahal,ontho
Start many times 1 3Y hasYwon, , andversus
associate howthemmany withtimes
many thetimes
versus X has
and won.
has
Y y-axes
how Each
won,
of a versus point (a,
how
Figure many
b) in the
times
1: Various Cartesia
minim
x-means onemanyplayer times
is always X has winning won. throughou
mea
Round 1 X has 1 won. Each point (a, grid.
1 Cartesian b) in Suppose
the Cartesian has
Xinstead plane,
won.
ofpoint where
Each
sticks ainand
point
trading b)are
(a, bCartesian nonnegative
in the Cartesian integer
Each the game.)(a, b) Counting
the the number plane, of subdiagthe
Round 2 plane,
1 where
2 a and 1b are
back andnonnegative
forth that we integers,
plane,
simply corresponds
where
keep a and to a wins
are
blattice for playerintegers,
nonnegative X and b win
where nala track
minimal
and bofare
how nonnegativepaths is a famous
integers, proble
ynal=
Round 3 1 corresponds
3 to3a winsmanyfor player
1times Y hasX and wins for
won,bcorresponds
versus how player
to a Ywins
many . Therefor on haveor been
player below
Xout and
a+b the line playe
brounds
wins
in combinatorics.
corresponds totimesa wins It
means turns
for playerthat the
X numb
in co
Round 4 1 for
4 player Y 6 . ThereX4have has been
won.1 a+b Eachrounds
pointplayed
for(a,player so
in Yfar,
.
the and
There each
have
Cartesian player
been hasone
a+b roundsplayer
either is always
gained
played or los
b)
and |a of
b −paths
wins on
for orplayerbelow
the line
Y. y = Counting
There
game.) x,havefrom (0, of0)p
Round 5 1 5 so far,10 and each 10 player
plane,has either
5 where gained
a1 and or
so lost
b are far, and
nonnegative
(n,
b| sticks.
each playerthe
integers,
equals
In this
hasnth eithermodel,
Catalangained number, lostorthe n
theor‘equilibrium
been n),
a+bis the rounds played so each
far,lattice
and paths (n, n
Round 6 1 6 |a − b| sticks.
15 20 In corresponds
this
15 model, 6 to thea ‘equilibrium’
1 for
wins |a −player state
b| sticks.X and line
In bthis
wins =nal
ymodel, minimal
x, where
the ‘equilibrium’ player has th is
each sameplayernumber has either in eachgained 
combinatorics. 
or lost It turns o
state is the line y = forx, whereYeach
player player
.toThere havehas
statetheisa+b
been therounds
line y= x,ofwhere
played sticks. 1 player 2n has the
Table 1: Pascal’s Triangle: sticks versus rounds, 0 6 |a-b| Eachsticks.
ofgame In C =
of
this paths
of lahalnunfolds on
model, or . below
the line y
same number of sticks. sameeither
number sticks. + 1 nas a ‘path’ in th
n
so far, and each player has gained or lost
Each game of lahal `equilibrium'
Cartesian state
grid,unfolds (n,
is
from (0, the
n), equals
line
as a0)‘path’ the
y=x, nth Catala
to (a,inb),thethe cur
|a −unfolds
b| sticks. as aIn‘path’ in Each
this model, the the game of lahal
‘equilibrium’
1 2 3 Cartesian
4 5 grid,
6 from (0, 0) to (a, b),
7state is8 the 9line y10= x,11wherethe where
cur-
Cartesian rent each
See
each player situation.
the
grid, fromplayer
excellent
has the Moves
has the along
same
(0, 0) to (a, b),[2]
reference these
number
for more
the paths
cur- ar
detai
Se

1
rent situation. Moves same numberalong these pathsrentare either
on minimal
ofsituation.
sticks. eastMoves
or north,
lattice alongpathsaccording theto
and paths
these n whether
CCatalan
are on
= nummX
Round 5 1 5 10 10 5 of sticks.1 n +Suc 1
either east or north,Each according to whethereither X or
bers.
east Y or (respectively)
north, according wins a
to given
whether round.X bers
Round 6 6 15 20 15 game of 6lahal unfolds Each as a ‘path’ gamein of theLahal unfolds as a
or Y (respectively) wins
Cartesian 21 a given round. or Such
grid, from 6(0, 0)`path' Y paths,
Let’s
(respectively)
to (a, in which
b), the return never
wins ‘backtrack’
to alahal.
given Suppose
round. are Suchcalled min
in reference
a gener Le
Round 7 6 21 35 35 the cur- Cartesian See the excellent
grid, from
paths, which neverrent ‘backtrack’ are called
situation. 27Moves along mini-
paths, mal
lahal
which
these paths.
game
never
paths Just
with as
‘backtrack’
 b), are k in our
sticks early
areper discussion,
player,
called mini- that ther
aft
laha
Round 8 27 56 70 56 (0, 0) �a+b
to (a, the on minimal
current lattice paths an
situation.
mal �paths. Just as in our early discussion,
either east83or north,27according there
mal paths.are
2n turns
Just paths
asone
in from
player
our early (0,
has 0) to+ (a,
discussion,
k r sticks, Thisand
b).there is
2nat th
t
Round 9 27 83 126 126 �a+b toalong
Moves whether
a theseXpaths bers.are either east
a+b
are a paths from (0,
or Y 209 0) to (a,
(respectively) b). This is
are ath entry
other
paths in
has the
from
k −(a(0,
r +sticks,
0)b)thto row
0
(a, <b).of
r Pascal’s
This
< k. is ath Triangle
othe
Round 10 110 209 252 110 wins aorgiven a north, round. Such toLet’s
according whetherreturnXtoorlahal. Su
entry in the (a + b)th paths, row of Pascal’s
which entry inwhere
Triangle,
never ‘backtrack’ the
are The (aindexing
number
called + b)th
mini- ofentries
row ways and
this
of Pascal’s
lahal game rows
can occurbegins
Triangle,
with inwith
kissticks
one-t pe0
Th
where
Table 2: sticks indexing
versus rounds, entries and rows
5 to paths.
mal 10 Just begins with
as in our where
early Figure
0. indexing
discussion, 1 shows
one correspondence
entries various
thereand2nrows withminimal
turns the
begins
one paths
numberwith
player from
of onek0
(0,
minim
0.has
�  2
Figure 1 shows various are minimal
a+b
paths paths
from from (0,to
(0,Figure
0) 1to
0) (a, (14,This
lattice
shows
b). 10).
paths
various from
is ath (0,
minimal
other 0) paths
to
has(n+r,
kfrom
− rn−r) (0, 0)which
sticks, < rd
0latti
a
 same number boundaries of sticks. outcomeCdelivers
n =
g the binomial coefficients other, no n +h
matter
. See if you can prove this Each game of lahal unfolds as a ‘path’ in the
�   ∞            
ctorial form nk = n! 2n
. 2n Cartesian
2n By results
grid, 2nin [3]
from (0, 0)the to number
(a,2n b),ofthe suchcur- paths2nis See the excellent refere
�  k!(n−k)! + + −  Moves along − − are on . lattice paths
at nk actually i=−∞ counts nthe + 12i10 n + 12i + 1 n +rent 12i + situation.
2  ∞ n + 12i + 6  n these 
+ 12i + paths
7 n+ 12iminimal
+ 8References
2n 2n
subsets of an n-element set. Figure 2: f(n), the total numbereither east or north, according
of non-ending sequences n + rwins +after
2ik 2n−rounds
to whether X bers.
+ rounds
nround. r + (2i
.
+ 1)k Let’s[1]return
e is a standard convention or Y (respectively)
i=−∞ a given Such T. Pelton, L. F
to lahal.
Y (respectively)
equal to zero if k > n or Figure
wins a 2:
givenf (n), the
round. totalSuch number
paths,paths, of
whichnon-ending
Note never that sequences
‘backtrack’
the sum after
are
is 2n
called
finite, rounds
mini-
since lahal
almost game
all Exploring
with
terms k prob
stick
Note that the sum is finite, since almost all terms
angle gives which us never
a triangular `backtrack' are called minimalmal �paths.
paths.  zero.Just
are zero. as in our early discussion, there 2n turns one Vector, BCAM
player has
a+b are
Just as
o binomial coefficients (see yin our early discussion, there are are paths from (0, 0) to (a, b). This is ath other has k − r sticks, 0
a In our The
Incase, casecase,
our wer= 6we
apply ends apply
this theformula
game,
this formulaand
for kone = forcannot
6, k=6,[2] R. and
Brualdi, In
paths from (0, 0) to (a, b). This is a entry th entry in the in the (a + b)th row of Pascal’s Triangle, The number of ways th
and rr=0, explicitly
= 0, 1, ...apply .., 5. ,to5the toformula
obtain obtain in thisdistribution
the distribution
the case. for for
ed.thePrentice H
th
(a+b) row of Pascal's Triangle, where indexing where indexing entries and rows begins with 0. one correspondence with
the 2nth There
2nth round. are
round. For 2 2n possible
Forinstance,
instance, sequences
thethe number for
number 2n coin
of of game
entries and rows begins with 0. Figure 1Figure shows 1 shows various minimal paths from (0, 0) lattice paths
flips
gamesequences (or wins/losses
sequences which result
which by the
result lahal
inintwotwo pointer).
players
players with Some
with [3]6 S. from
Dua,(0,
sticks S.0)Kh
t
nd Generalizations to (14, 10). not intersect either of the
various minimal paths from (0, 0) to (14, 10). 6 sticks number
each each (i.e. f (n)
(i.e.r= r of
= these
0)
0)game’s aftersequences
after 2n 2n== 20 result in
20 rounds
rounds isis the fied random w
One player wins when the path reaches x + k. Like the Catalan
a diagram  which
One player 
10depicts
∞ 
 a
wins 2n
when
 
 the  2n
game's
 
path  
reaches 2n
14 not
game   x 2n being
  over  by the 2nthround.
  2n 
 This reflecting
2n
 barr
∞
2n 2n + 2n + a ‘boundary’,2n which
20 is a
2n20 certain 20 diagonal 2n line. is another problem
169–175. on l
f (n) − can−be found−− using
−16n +the =12i formula
175 − above, and .
Call the atwo players X
`boundary,' + and which n + is 12ia certain + n + 12i diagonal
+1 − line.
We n +must We
12i + now 2 discuss−10 n + 12i these 4+ 6boundaries +066.
further. 7 . two n + linear
12i + 8boundaries. S
n + 12i i=−∞n + 12i +Figure 1 n +Various
1: 12i + 2minimal n +lattice
12i +is6paths shown n+ 12i +
fully in 7Figuren + 2. 12i (Who+ 8 would have
m withi=−∞ the x- and
must now y-axes
discuss of a these boundaries further. A minimal lattice path is subdiagonal if it lies r = 2, k[4]= Pascal’s 6. Trian
ose instead of sticks trading The The thought
behavior behavior that
aftersuch an aodd
after ansimplenumber
odd game2nleads
number +2n1 to+of1such of rounds
A minimal lattice path is subdiagonal if it lies en.wikipedia
we simply keep track of how Figure 2: f (n), the total number rounds a non-ending
of
can long
can bebe formula!?)
determined
determined sequences The from probability
from round
afterround 2n rounds 2n
2nof using
the
using gamethe `sum
Figure
on or below the lineony or 2: f (n), the x. (Innumber
total
= below the
Lahal, lineofthisynon-ending
=means x. the (Insequences
lahal,
ending this
on after
or 2n
before rounds
the 2nth round 3
is
on, versus how many times ‘sum rule’ which generates
rule' which generates Tables 1 and 2. Tables 1 and 2.
oint (a,y b) oneinplayer
the Cartesianis yalwaysmeans winning onethroughout
player is always the game.) winning throughout
Counting the number the game.) Counting theminimal
of subdiagonal number ofrsubdiago- TheThe case case rP= rr == 661ends
ends
− 2−2n game,
thef (n).game,and andone one cannot
cannot
b are nonnegative
  integers,     The case   = 6 ends  the game, and one cannot 4
∞
2n paths 2n nal minimal
2n lattice pathsexplicitly
in2ncombinatorics. is 2na famous explicitly
problem
explicitly 2n apply apply thethe formulaininthis
formula this case.
case.
ns for player lattice X and + iswins
a famous+ problem − − apply
− the formula . in
2nthis case.
n + 12i n +b 12i + 1 inncombinatorics.
+ 12i + 2 n + 12i It +turns
6 nout
+ 12ithat + 7 the This
n+ actually
number
12iThere
+ 8 tends
are 2 to 1
possibleas n sequences
→ ∞. Therefore,
for 2n coin
ave beenIta+b
i=−∞
turnsrounds outplayedthat the number of paths on or below There are 2There 2n possible sequences
are 22n possible for 2n coin
sequences for 2n coin
of paths on or below line flips y14= (or x, a
th from (0,
game flips of
0) (or to lahal (almost surely)
wins/losses by the lahal pointer). Some will end.
er has either line y =
gained
 2: for x, from (0, 0)
lost total number of  to (n, n), equals the n x wins/losses
flips (or by the
wins/losses lahal pointer).
by the Some
Lahal pointer). Some
 ∞ Figure (n),the (n, n), equals  nth
the
non-ending Catalan
sequences after 2n rounds This
number, number
orof is rather
 f (n)intuitive.of these   Ifsequences
a fair 
coinresult
is tossed in the
is model, the ‘equilibrium’ 2n 2n Catalan 2nnumber, number f 2n
(n) these 2n
sequences result in 2nthe
+ + − or number repeatedly, −f(n) youofshouldthese sequences
expect
−by theeventually result . in
a runtheThis
game
, where y10 each player n +has 12i
10
the n + 12i + 1 n + 12i + game 2 not n + being12i
not +game 6over not
being byn +being
over the12i
by +
2nth
over
the 7 2n thn + 12i
round. round.
2nth
This 8round.
+This f(n)
i=−∞ Figure 3: Two minimal lattice The case rpaths=16 ends2n the game, of
within and12 one
f (n)(or
cannot evenbeafound
can million) using headsthe in a row.above,
formula This and be
can
s. C
explicitly
n = apply thef (n)
formula can. in be
this found
foundcase. using
using the
the the formula
formula above, and
boundaries n+1 n . outcome is 2n shown delivers
coin fully inwin Figureforabove,
one
2. player and over
(Who is shown
would thehave fully
l unfolds as a ‘path’ in the There are 22n possibleis shown fully
sequences forin Figure 2. (Who would have
Figure 2: f (n), the total flips (or number
wins/losses of the lahal in
non-ending
by other, Figure no
thought
sequences
pointer). Some 2.that
matter (Who
after how2nwould
such many
a simple
rounds have
sticks gamethought
are in
leadsplay! that
to suchsuch
m (0, 0) to (a, b), the cur- See the excellent See freference
the thought
these[2]
(n) ofexcellent forthatmore
a such
simple details athe simple game leads toa such
By results in [3] the number
number
of being
suchover paths
sequences
is 2nth round. a long formula!?) The probability of the gameThe
result in game leads to such long formula!?)
ves along these paths are on minimal reference lattice
game notpaths a long
and
[2] for bythe
more theformula!?)
Catalan num- The
This probability of the game
10
ending
probability onoforthe before game the ending
2nth round on isor before the
y
, according to whether ∞  X bers. 2n
 f (n)
detailscan be
on
2n
found endingThe
using
minimal  the on or References
formula
case
2n
before
above,
th r = 6
round
the
and
ends
is
2nth theroundgame,isand one cannot
is shown fully in Figure 2. (Who would have
wins a given round. Such Let’s − lattice explicitly
. game a apply the formula in this case.
2ik return tor lahal. Suppose in leads general −2n
n+r+ thought
n+ paths
that
+ (2i +and
such a1)k the
simple to such −2n P r = 1 − 2 f (n).
backtrack’ are called mini- lahal game Catalan
i=−∞ with sticks
a long kformula!?) per
The There
player,
probability [1]thatP
ofT.
are r
the Pelton,
=
2
after
2n
game1 − 2 L.
possible Francis-Pelton
f (n).
sequences for and 2nL.coinWilliams,
numbers.
ending on or before the 2nth round isExploring probability with aboriginal games,
our early discussion, Note that therethe2n sumturns is finite, one since
player has kall+terms
almost flips
r sticks, (orThiswins/losses
This
and theactually
actually bytendsthe
tends lahal
toto1pointer).
as n →Some
1 as Therefore, a
∞.. Therefore,
Let's returnThis actually to tends
Vector, toBCAMT,
1 as n →Spring ∞. Therefore,2006.
(0, 0) to (a, b). are Thiszero. is ath other has k − r sticks, 0P r<= r1 − <number
2k. f (n).
−2n
game f (n)
a gameofofLahal these
of lahal sequences
(almost(almost result
surely)
surely) inwill
will the end.
end.
Lahal. Suppose a game in 14 aof xlahal (almost surely) will end.
h row of Pascal’s Triangle,
In our case, we apply this The number of ways
This formula
14 x this can
for toThis game
occur
k 1=as 6, not
is
[2] in being
R. This
one-to-
Brualdi,over
is by
rather the
Introductory 2nth
intuitive. round.
If a fair
Combinatorics, Thiscoin is tossed
10
generalactually tends
Lahal game is
n→ rather
∞.This intuitive.
Therefore,is rather If a intuitive.
fair coin is a fair coin 4th
Iftossed is tossed
es and rows begins and with r = 0,0.1, .one . . , 5correspondence
to obtain a game with
theofdistribution the number
lahal (almost (n)
for can
fsurely) be
of end.
will minimal
ed. found
repeatedly,
Prentice using you theshould
Hall, formula
New Jersey,above,
expect 2004. and
eventually a run
14 x
with k sticks repeatedly,per you
repeatedly, should expect
you eventually
should expect a run
eventually a run of
us minimal paths
Figure3: 3: Two from (0,
Figure
theminimal 0)
2nth round. 3: lattice
Two
lattice lattice
paths paths
minimal from
For instance, the number(0,
This lattice
is 0)
rather to (n+r,
intuitive.
paths is shown
If a fair coin
within fully
of
which is 12 in
do
tossed(or Figure
of even a million) heads in a row. This
n−r) even 2.
a (Who
million) would
heads have
in a row. This
Figure Two minimal paths within
repeatedly, you of 12
should (or
expect 12 (or
eventually even
a run a million) heads in a row. This outcome
boundaries notwhichintersect player,
either of that
thea million) after
lines with thought
= x−k [3] S.
that Dua,
outcome
and yThissuch = S.
a Khadilkar
simple
delivers game
the and
win leads
forK. Sen,
to
one suchA
player modi-
over the
3: Twogame sequences result
of 12in(ortwo evenplayers
within boundaries y heads
boundaries
Figure minimal lattice paths within outcome delivers
in a row.the
delivers winwin
the for one for player
one over the
player over the
hen the game’s path
boundaries 6 sticks reaches each (i.e. x + k. r =Like 2n
0) after turns
the Catalan
outcome 2ndelivers one
= 20 the wina for
player
restriction
rounds long above,
isonematter
player fied
formula!?)
other,
over random
thisthenoThe walk
matter in the
probability
how manypresence
of the sticks ofare
game in play! no
partially other,
other, no matter how how many many sticks sticks are in
are play!
in play!
is a certainhas kdiagonal+ r sticks, By line. and
 results is  another
the
in [3] other
 theproblem
other,
has
number kof-lattice
on
no matter rhowsticks,
such ending
many paths,
paths
stickson arereflecting
is now or with barriers.
before
in play! the 2nth J. roundAppl.is Prob. 13 (1976),
ByBy results in
0 < r < k.further.  two
results
these boundaries in [3] the [3]
number 20
the number
of such 20
pathsof issuch 20 paths is 169–175.
− linear − boundaries. = 175 066. See Figure 3 for n = 12,
ReferencesReferences P r = 1 − 2−2nReferences
∞     ∞10 4 k= 6. References
16   f (n).
path is subdiagonal
∞
The number 2n if it 
lies r
2n = 2n
2, 2n
2n− of ways this. 2n can−occur is in one-to- . [4] Pascal’s Triangle on Wikipedia: http://
n + r + 2ik n + r− +n(2i++r1)k + 2ikodd n + .r L. +2n (2i++ 11)k
n The+ r +behavior after
n + ran
[1] T.number
Pelton, Francis-Pelton of actually
and L.[1] Williams,
T. Pelton, L. Francis-Pelton and L. Williams,
i=−∞
one correspondence 2ik
i=−∞ with + (2inumber
the +Exploring
1)k of minimal
[1] T.
probability ThisPelton,
with aboriginalL.
[1] tends
Francis-Pelton
T.
games,Pelton, to 1L.
en.wikipedia.org/wiki/Pascal’s asFrancis-Pelton
n →L.∞.
and Therefore,
Williams, and triangle
L. Williams,
rounds
is finite,can since3 be determined from round 2nwhich using
i=−∞
Note that the sum
lattice paths from almost
(0, all terms
0) to (n + r,
Vector, n -
BCAMT, r) a
Spring game of lahal Exploring
(almost probability
surely) will with
end. aboriginal games,
are zero. Note that the sumgenerates
is finite, since 14 almost
Exploring
all2.terms2006. probability
Exploring with
probability aboriginal with games, aboriginal games,
Note do thatnot the
the ‘sum
sum
intersect is rule’
finite,
either which
sinceof almost
the linesall Tables
terms
y = x 1x- andk and This is rather Vector,
intuitive. BCAMT, If a Spring
fair coin 2006.
is tossed
In our case, we apply arethis zero.formula for k = 6, [2] R. Brualdi, Introductory Vector, BCAMT,BCAMT,
Vector,
Combinatorics, 4th Spring 2006. Spring 2006.
areryzero.
and == 0, 1,x. .+. , 5k.to obtain
Like the
the distribution
Catalan forrestriction
ed. Prentice Hall, New
above, repeatedly,
this Jersey, 2004. you should expect eventually a run
the In2nthour
In our the case, we apply this formula for k = 6, [2] [2]R. R.Brualdi,
Brualdi, Introductory
Introductory Combinatorics, 4th4th
Figure case,
round. Forwe
3: problem
Two apply
instance, this formula
number of for k = within 6, [2] R. ofandBrualdi,
124K. (orSen,even Introductory
a million) Combinatorics,
heads in a Jersey,
row. Combinatorics,
4th This
gameis another
sequences which andresult =minimal
r in 0,on
two 1, . . ,lattice
.lattice
players 5 topaths,
with [3] paths
obtain
S. Dua,now with two
theS. distribution
Khadilkar for ed.
A modi- Prentice Hall, New 2004.
and boundaries
r = 0, 1, . . . , 5 to obtain the distribution for walk ined. ed.delivers
Prentice Prentice
Hall, New Hall,
Jersey, New Jersey, 2004.
2004.
linear
6 sticks each (i.e. r =the
boundaries. 0) after2nth 2nSee=round. Figure
20 rounds For fied random
is 3 instance,
for n = the 12, numberr=outcome
the2,
of of partiallythe win for one player over the
presence
the 2nth  round.
   For  instance, the number reflectingof barriers. J.other,Appl. Prob. no [3] 13 (1976),
matter
[3]S.S.Dua, how S.
Dua, many
S.Khadilkar sticks are
Khadilkar and in K.play!
K. Sen,
Sen, A A modified
modi-
k = 20 6. 20 game 20 sequences which result 169–175.in two[3] playersS. Dua, with S. Khadilkar and K. Sen, A modi-
and
game 10 sequences
− −which=result 175 066.in two players with fied random walk in the presence of partially
By results
results
4 in [3]
6 sticks 16 the number
each (i.e. r =of0) such
after paths
2n =is20 is rounds
fied randomrandom
is walkwalk in the inpresence
the presence of partially of partially reflecting
6 sticks By each (i.e. in r =[3] 0)theafternumber 2n = 20 rounds
[4]of such
Pascal’s is
paths
Triangle on Wikipedia: http://
reflecting barriers. J. Appl. Prob. 13 (1976),
The behavior after an
    odd 
number  2n + 1 
of    reflecting
en.wikipedia.org/wiki/Pascal’s References barriers.
barriers.
triangleJ. J. Appl. Prob. 13 (1976), 169-175.
Appl. Prob. 13 (1976),
rounds can ∞ 
be determined 2n 20 2n using
from round 20 2n 20 169–175.
20 20 20 − − = 175 . 066. 169–175. [4]
the ‘sum rule’ which
n generates
− + r4 + 2ik 10 −1 and
−Tables =n 2.4175
+ r 066.
+ (2i16+ 1)k
Pascal's Triangle on Wikipedia:
i=−∞10 16 [1] T. Pelton, L. Francis-Pelton
[4] Pascal’s Triangle on Wikipedia: http://
http://en.wikipedia.org/wiki/Pascal's and L. Williams, triangle
The behavior after 4an odd number [4] 2n Pascal’s 1 of Triangle
+Exploring on Wikipedia:
probability with aboriginal http:// games, triangle
The Note behavior after an odd number 2n + 1 of
that the sum is finite, since almost all terms en.wikipedia.org/wiki/Pascal’s en.wikipedia.org/wiki/Pascal’s
rounds can be determined from round 2n using Vector,
3 BCAMT, Spring 2006. triangle
rounds are canzero.be determined from round 2n using
the ‘sum rule’ which generates Tables 1 and 2.
 Who in the Land is
Fairest of All?
MyHeritage.com is an internet-
based company that offers
By you the chance to see which
P. H. Trinh celebrity you most resemble.
Oxford Remember how in Snow
University White, the queen has a

Seventh Grade Blues magical mirror which provides

her with uninhibited flattery? This is the same,
Back in the seventh grade, one of the girls told but like, tons better.
me I looked like Keanu Reeves. Seriously, I was After a free signup, you upload a large-ish jpeg
hanging upside-down on the jungle gym, minding of your mug, then let the software crank away.
my own business, and she just walked over and My personal resemblance results were: Brad Pitt
told me like it was no biggie. (71%), Keanu Reeves (63%), Luke Perry (63%),
Then she giggled like a moron and ran away. and Matt Damon (63%).
Now while this singular moment of brazen Brad Pitt? Really? Matt Damon? Really? Who
flattery would become the highlight of wouldathunk? But y'know, as I gaze into the
my paste-eating academic career, I was mirror... well... yes, I see it now. Definitely. We're
also torn. On one hand, I wondered how practically brothers!
anyone could confuse Keanu (black How does it all work? Is this actual science or
shades, gothic trench coat, totally just deceptive flattery? To understand how facial
awesome) with me (pubescent, recognition works, we're going to have to delve
angst-ridden, gawky)? Was this into the mathematics behind the algorithm.
all some awfully cruel and How might
sadistic joke girls liked to we go about Recognizing Faces
play on unsuspecting boys? identifying that
person from a Suppose we were given someone's
On the other hand, maybe
large database picture. How might we go about
| maybe she was on to
of faces? identifying that person from a large
something. Maybe somewhere
database of faces?
| somehow, behind all that
bad acne and ruffled hair, my hidden One way we can go about
One way we
Keanu-like features beckoned faintly, it is by identifying the
can go about it
like some distant lighthouse characteristics of the subject
A more is by identifying
obscured by fog. | perhaps the person has
computationally the characteristics
small lips, or a pointed chin,
Today, however, thanks to efficient way would of the subject
be to examine these or distinctive eyes. From here,
the latest advances in facial
faces as a statistical we then consult the database,
recognition, I no longer have to
whole rather than as going from picture to picture, each
wonder: she was right.
the sum of their time isolating the features of the faces and
parts.
4
 checking for a match.
Whilst this might work, it
As would also be a lot of work;
stored algorithms would need
in a to be defined to analyse
computer, a each desired feature. FFigure
IGURE 2: Faces can be
2: FACES CANidentified as coordinates
BE IDENTIFIED AS in a
higher
COORDINATES IN A dimensional plane.
HIGHER DIMENSIONAL
picture is nothing Imagine having to do
PLANE.
Now more than a great this for each of the as well as a direction. Moreover, they can be
big grid of dots thousands of faces
in the (or pixels). Like vectors
added, subtracted, representing
multiplied, force or motion,
and manipulated
that stream pass the
abstract like most other mathematical quantitiesa `magnitude’
these new ‘face vectors’ have | the
theory of gates at a football match (an overall brightness), as well as a ‘direction’.
only difference is that they inhabit some higher-
linear algebra, or in a busy airport. Moreover, they can be added, subtracted,
dimensional face space, rather than the two or
these grids of A more computationally multiplied, and manipulated like most other
three dimensional physical
mathematical world we
quantities liveonly
– the in. difference
pixels are called efficient way would be to examine
is that they inhabit some higher-dimensional
‘vectors’. these faces as a statistical whole What’s
face Your
space, Eigenface
rather than Basis?
the two or three-
rather than as the sum of their dimensional physical world we live in.
parts. This is similar to the difference However, face spaces are complicated
between identifying a city by its landmarks affairs | they're high dimensional boxes
and identifying the same city by the density stuffed with a large number of faces, each face
of its roads, the clusters and heights of its containing thousands of pixels.
buildings, its downtown areas, and so on. It would thus be foolish to try and
...we look to compare each face pixel by pixel; instead
A Picture is Worth a Thousand construct a small
we look to construct a small group
group of pictures
Digits representing the general of pictures representing the general
facial patterns of the facial patterns of the database. This
Snap! But what are pictures, database. This small but small but crucial group is called the
really? crucial group is called eigenface basis.
the eigenface basis.
As stored in a computer, a picture is Think of how, when we analyse
nothing more than the motion of a ball flying through the
aFIGURE
great1: Abig grid of What’s Your
air, Eigenface
we break Basis?
the motion into its horizontal
dots (or pixels). If the picture and vertical components. These two components
PICTURE IS
NOTHING
is However,
MORE each pixel
greyscale, is Face Spaces are complicated affairs
THAN A LARGE — they’re
provide
high
a fundamental
dimensional
basis capable of describing
boxes stuffed with
associated with a number any arbitrary motion.
GRID OF NUMBERS.
from a large
0 to 255 representing itsnumber of faces, each face containing
thousands of pixels.
Similarly, once the eigenface basis is found using
brightness, from pitch black
(0) to pure white (255). linear algebra, each face in the database can then
It would thus be expressed to
be foolish try and
using compare
certain percentages of each of
Now in the abstracteach
theoryface pixel by pixel; instead we look to
of linear algebra,constructthese a small group of pictures
grids of pixels are representing
called the general facial patterns of the
vectors. You've probablydatabase. This small but crucial group is called
the eigenface basis.
encountered vectors before
in Physics class and in fact,
these `face vectors' are quite
similar.
Like vectors representing
force or motion, these new `face
vectors' have a magnitude
Figure 1: A picture is nothing
(an overall brightness),
more than a large grid of
Now in numbers.
the abstract theory of linear algebra, 5 http://brownsharpie.courtneygibbons.org/
these grids of pixels are called ‘vectors’.
 database. This small but crucial group is called
the eigenface basis.
FIG
eigenface components. DAT
FAC
Now the problem of CAN
recognising the subject CON
becomes as simple as finding EIG
the shortest distance (or AFT
In optimal
closest match) between FAC
conditions (with
our subject and the SCR
good lighting,
faces in the database, DEC
, a representative
DIF
. database, front-facing a process aided
pictures, etc.), a simple PER
n enormously by the
eigenface routine might EAC
’ FIGURE
Figure 3: VECTORS
3: Vectors can beCAN BE DECOMPOSED
decomposed into INTO fact that each face (FA
produce accurate
BASIS ELEMENTS basis. elements. readings of up to 90%.
is now represented AT&
by only a handful of
the of
Think building
how, when blocks. For example,
we analyse the motion of eigenface components.
a we
ball may
flying say thatthea air,
through picture is the
we break
composed
motion into ofits 10% of the and
horizontal first vertical The Future and You
components.
eigenface, These25% two of thecomponents
second, provide a
Unfortunately,
fundamental
4% of the basisthird,capable
and soofon.describing realany
life is never But really, just how
arbitrary motion. that simple, and
accurate are these eigenface
The beauty of this one must contend
treatment algorithms?
Similarly, once is that even
the eigenface basis iswith a multitude of
found
in linear
using a large database,
algebra, each face each ‘noisy’ factors.
in the database In optimal conditions
unique
can then facebe can be expressed
expressed using certain (with good lighting, a Figure 4: A database A Problemof faces
of Dist
percentages
very simply of each of theits
using building blocks. For
eigenface can be used to construct an
representative database,
example, we may We say no thatlonger
a picture eigenface basis. Afterwards,
decomposition. haveis to express front-facing pictures, etc.), Now imagine each face in
, composed of 10% of the first eigenface, 25%
a face under scrutiny can be
each face using thousands of pixels; now, like a simple eigenface routine decomposed into represented in different
terms of
’ ofathesimple
second,recipe
4% of the third, and so on.
. in which the eigenfaces are the might produce accurate percentages, percentagesakin to coordinate
of each
key ingredients, the entire database can be readings of up to 90%. higher-dimensional
eigenface (Faces courtesy plane.
of O
, The beauty of this treatment is that even in a (which may or may
AT&T Cambridge) not lie in
r reconstructed as it was before.
large database, each unique face can be Unfortunately, real life is then projected onto this plane
e expressed very simply using its eigenface
l never that simple, and one in terms of the eigenface comp
decomposition. We no longer have to express must contend with a multitude of `noisy' factors.
- each face usingAthousands
Problem of Distance
of pixels; now, like These include variance in pose (person facing
a simple recipe in which the eigenfaces are the
at an angle), obstructions (sunglasses or other
key Now
ingredients,
imagine the entire
each database
face incanthe be database,
reconstructed people), resolution, lighting, and so on. Despite
representedas in it was
termsbefore.
of its eigenface percentages,
this, however, the science of facial recognition has
akin to coordinates lying in some higher-
steadily improved to the point where today, it is
dimensional plane. Our test subject (which may
becoming a standard for many military, security,
or may not lie in the database) is then projected
and commercial applications.
onto this plane by expressing it in terms of the

FIGURE 5: FINDING THE CLOSE

DOWN TO CALCULATING THE SH
DISTANCE BETWEEN POINTS.

6 http://xkcd.com/
 in terms of the eigenface components. drunken, hand-wavy speculation that
resemble Brad Pitt and Keannu Reeves.
really, really do. It’s backed up with scienc
But I digress. You see, the whole point of FURTHER READING
and everything.
this article was that the inner workings of
facial recognition is nothing but science and [1] M. Turk and A. Pentland. Eigenfaces
mathematics. That's right. It's not half- FURTHER
for recognition. Journal of READING
Cognitive
drunken, hand-wavy speculation that Neuroscience (1991) 3 (1), 71-86.
[1] M. Turk and A. Pentland. Eigenfaces for
I resemble Brad Pitt and Keanu recognition.
[2] W, Zhao Journal
andofR.
Cognitive Neuroscience
Chellappa.
Reeves. I really, really do. It's (1991) 3 (1), 71–86.
Face Processing: Advanced Modeling
backed up with science and and Methods, Academic Press, 2006.
everything. [2] W, Zhao and R. Chellappa. Face Processing:
Advanced Modeling and Methods, Academic
Press, 2006.
FFigure 5: Finding the closest match boils down to
IGURE 5: FINDING THE CLOSEST MATCH BOILS
calculating the shortest distance between points.
DOWN TO CALCULATING THE SHORTEST
DISTANCE BETWEEN POINTS.

By
The idea for this article Anthony Quas
arose from a simple magic and
Peter
trick based on error-correcting UniversityJ. Dukes
of
codes. This activity has been Victoria
used in Math Mania, an outreach
event in which students and staff members at the
University of Victoria visit local elementary schools
for hands-on math activities with students.
In the trick, the young student is presented
with a rectangular array of cups, some face up
and some face down. While the magician looks
away, the subject is asked to invert (i.e. flip) any
cup. Looking back at the cups, the magician then
is able to determine which
cup the student inverted,
without any memorizing!
To do this trick, the cups
are initially put into a
configuration in which
each row and each column
contains an even number
of cups that are the right
way up, as in Figure 1.
After a cup is inverted,
the column in which it sits
contains an odd number
of cups the right way up
Figure 1: Rectangular array of (there is either one more
cups used in the magic trick.
7
or one less than before); and similarly the row in
which it sits contains an odd number of cups the
right way up. This allows the magician to figure
out which row and column the cup that the student
inverted is in. Of course this is enough to figure
out which cup was inverted.
In this article, we will discuss the connections
between this magic trick and error-correcting codes,
and investigate an extension where hexagonal
arrays replace the rectangular array of cups.
It is shown that the hexagonal pattern
corresponds to a more robust code (and hence a
flashier magic trick!) More on this later; we now
comment on codes and their uses.
In a basic model of a communication system
there is a source that produces some sort of data
that is to be sent to a receiver. A typical first step
in this process is for the data to be transformed
into binary digits (bits - 0s and 1s), with strings
of several bits representing one piece of data. If we
want to use strings of k bits to represent pieces of
data then there are 2k different strings. This means
that if there are N different pieces of data, then
we need to choose k with 2k≥N so that all of the
possibilities can be encoded. For example, if the
data that we want to transmit are letters of the
alphabet, then since 24 < 26 ≤ 25 we need to use
 upted bit. The minimum distance of a code is important with
respect to error detection and correction since it is the
0 1 1 0 0 1 1 0
minimum number of errors that must occur in order to
1 0 1 0 1 0 1 0 strings of at least 5 bits to the error can be fixed).
transform one codeword into another. A code C  can
0 0 0 0 0 0 0 0 encode the letters of the
detect all combinations Now of lett us returnerrors
or fewer to the magic
if and onlytrick with cups.
1 0 0 1 1 0 1 1 alphabet. We can calculate
if the minimum Insteaddistanceof of thinking
C is at least of cups
t + 1.theTo right
see way up and
1 0 1 0 1 0 1 0
the number of bits required the wrong way up, we write a 1 for a cup the right
this, note that if we start from a codeword and make
1 1 1 1 1 1 1 1
to be able to encode N way and a 0 for a cup the wrong way up. With
t errors or less, then we cannot end up at another
different possibilities as this view point, the student is presented with an
codeword. This means that we can detect that an er-
gure 2: corresponding
Figure array of bits,
2: Corresponding arraywith[log
one N]flip.(where [x] means array of bits with an even number of 1s in each
2 ror in transmission has taken place. In this case we
of bits, with one flip x rounded up to the next row and column. Let's say that a row or column
say that C is t-error-detecting. Even though you can
rve that 2 errors cannot alwaysinteger). be corrected identify that anwith erroran has taken
even place,ofit 1s
number may not the
(cups be right way up)
s procedure. If Ifwe
thecode
magician seesin
letters two odd rows
binary (`A' = possible
00001,to fix theis error.
called balanced. Inverting the cup corresponds
d two odd `B'columns
= 00010, c , c′,`C'
he or= she is notetc.),
00011 sure then Wewe will illustratetoerror detectiona with
corrupting singlea simple
data ex-bit, thus making
positions (r  , run
often c) andinto (r′,problems
c′) were flipped,
if there or is aample
noisy related
Error to the ‘cup
some game’.
row and Consider
column starting
unbalanced. Noting
s (r , c′)channel
and (r′, where
c) weresome flipped. We wonder
of the bits maywith get all 32 words of length these5.unbalanced
This would lines be sufficient
is a way of correcting
correction
, using acorrupted.
hexagonal array of cups (where
For instance, consider stray there to encode the English letters, plus some punctuation.
the error. Translating the configuration of
means that
ections to check
radio in rather thana 2),
interference, it is possible
scratch on a CD For single error detection, we can 1append a check 1s,at we obtain the
and bit
you know there Figure into 0s
ately correct two (or maybe more) errors. Be- the end of each word,
ROM, etc. Notice that the strings was an error and you left array in Figure 2. Flipping creating words of length 6. The the bit in
cussing the cup trick and hexagonal arrange- rule for the check bit is that the resulting weight must
00110 00001 01100 01100 and 00110 can work out what row 4, column 3 leads to the right array
rther, we give a brief introduction to the theory be even. For instance, 00101 becomes 001010 after
01001 01100 01100 differ in a single bit was there before the in Figure 2. The unbalanced row 4 and
-correcting codes.
but represent the words `FALL' andthe check errorbitoccurred
(zero) is appended.columnAs3 another
identifyexample,
the corrupted
`FILL' respectively. 10101 becomes 101011 after
bit. the check bit (one) is
nology and facts for codes appended. We will think about the code you get con-
rd of length Error-detecting
n is a finite sequence and oferror-correcting
bits v = sisting of all words arising Observe that 2 errors cannot
this way. In this code, there
 n, where each bit b i , is either 0 or 1. A code is are still 32 words, each of length 6, and the with
codes are designed to give a way to always be corrected distance
this
words. The wordsencode belonging datatoso somethatcode we too
if not many bits procedure. If the magician sees two Codes
between any pair of different words is at least two.
ng about are usually called
get codewords
corrupted, it is. possible
We will to(Verify
detectthis
that odd rows
for yourself: whyr; can’t
r' andtwo two odd columns
codewords, with of
concerned with codes having codewords all of check bits included, be at distance one?) Consequently,
there has been an error in transmission c; c' he or she is not sure whether
this type
e length n. This Erroris theor length
(evenof better)
the code.to correct this it.
codeNote positions
is 1-error (r; c)Toand
detecting. see (r';
that c')
it iswere
not er- are used in
Hamming weight detection of a word , written
that verror detectionwt (v),means that you flipped,
ror-correcting, imagine that or positions
we receive (r;thec')codeword
and
umber of times the digit 1 occurs
thereinhas v. been
For an assigning ISBN
means that know error, but
001101. There (r';
is noc)way were flipped.
to tell whetherWeitwonder
was a cor-
numbers to
e, wt (110101) = 4 and wt (00000) you can't = 0. tell what the errorofwas
ruption 101101,whether,
011101,using 000101, a 001001,
hexagonal 001111
there has been books
and v be words of the same lengtherror
whereas n. The or 001100.
correction means array of cups (where there are 3
an error, but you
ng distance between them, denoted that youd (u, knowv) isthat there was directions to check in rather than
can't tell what the
an error and you can work out 2 just
Addition modulo 2),means
1
it is0 + possible to0 +accurately
1 = 1 + 0 = 1 and 0=1+1=0 correct
error was what was there before the error two (or maybe more) errors. Before discussing the

occurred. cup trick and hexagonal arrangement further, we
Doing this requires that we don't use give a brief introduction to the theory of error-
the minimum possible number of bits to represent correcting codes.
the data. The aim is to design error correcting
Terminology and facts for codes
codes which are simple to use, don't involve adding
too many extra bits to the strings, and are able A word of length n is a finite sequence of bits
to detect and correct most commonly occurring v = b1b2... bn, where each bit bi, is either 0 or 1.
errors. A code is a set of words. The words belonging to
Codes of this type are used in assigning ISBN some code we are talking about are usually called
numbers to books (so that if someone makes a codewords. We will only be concerned with codes
mistake with one of the digits then computers having codewords all of the same length n. This is
can detect the error rather than bringing up the the length of the code.
wrong book), recording data on CDs (so that if The Hamming weight of a word v, written
the CD gets some dirt on it then the music can wt(v), is the number of times the digit 1
still be read) and in computer memory (so that if
an electronic glitch corrupts a memory location,
8
occurs in v. For example, wt(110101) = 4 and with all 32 words of length 5. This would
wt(00000) = 0. be sufficient to encode the English
Let u and v be words of the letters, plus some punctuation. For
The single error detection, we can
same length n. The Hamming
minimum append a check bit at the end
distance between them, denoted
distance of of each word, creating words
d(u, v) is the number of positions in
a linear code of length 6. The rule for the
which u and v disagree.
C is the smallest check bit is that the resulting
If u and v are words The weight of a non- weight must be even. For
of length n, we define minimum zero codeword instance, 00101 becomes
u+ v to be the word obtained distance of a in C. 001010 after the check bit (zero)
by componentwise addition code is important is appended. As another example,
modulo1 2. For example, with respect to error 10101 becomes 101011 after the check
01101 + 11001 = 10100. detection and correction bit (one) is appended. We will think about the
since it is the minimum
You can check that number of errors that must code you get consisting of all words arising this
d(u,v) = wt(u + v), occur in order to transform way. In this code, there are still 32 words, each
because u + v has a one codeword into of length 6, and the distance between any pair
1 in precisely those another. of different words is at least two. (Verify this for
positions in which u yourself: why can't two codewords, with check
and v differ. For example, bits included, be at distance one?) Consequently,
d(01101, 11001) = 2 = wt(01101 + 11001) = this code is 1-error detecting. To see that it is
wt(10100). not error-correcting, imagine that we receive the
It is not hard to see that Hamming distance codeword 001101. There is no way to tell whether
is symmetric, i.e. d(u, v) = d(v,u), and it was a corruption of 101101, 011101, 000101,
obeys the so-called triangle inequality: 001001, 001111 or 001100.
d(u,w) ≤ d(u,v) + d(v,w). We use this later A stronger condition is that a code C be such
without commenting on it - see if you can that whenever a codeword is subjected
spot where. For a code C having at least two to a combination of t or fewer errors,
codewords, the minimum distance of C is the we can guarantee to correctly recover
smallest of the numbers d(v,w) over all pairs the codeword. For this, we require
v,w of distinct codewords in C. For example, let the minimum distance of C to be at
C = {0000, 1010, 0111}. The minimum distance of least 2t + 1. The idea is that if t The
C is 2. (or fewer) positions change in v, rule for
the check
The minimum distance of a code is important the resulting `noise corrupted
bit is that the
with respect to error detection and correction since word' v’ is still closer to v than
resulting weight
it is the minimum number of errors that must occur to any other codeword. (The
in order to transform one codeword into another. distance from v to v’ is at most must be even.
A code C can detect all combinations of t or fewer t so the distance from v’ to any
errors if and only if the minimum distance of C is other codeword is at least t +
at least t + 1. To see this, note that if we start 1.) The way to decode a word that is
from a codeword and make t errors or less, then we received across a noisy channel is just to replace
cannot end up at another codeword. This means it by the closest codeword. As we've just seen,
that we can detect that an error in transmission has provided the code has minimum distance at least
taken place. In this case we say that C is t-error- 2t + 1 and no more than t bits change, we are
detecting. Even though you can identify that an guaranteed to get back to the original word.
error has taken place, it may not be possible to fix Thinking back to the cup game, we think of all
the error. the rectangular arrays of 0s and 1s with an even
We will illustrate error detection with a simple
example related to the `cup game'. Consider starting Addition modulo 2 just means 0 + 1 = 1 + 0 = 1 and 0 + 0 = 1 + 1 = 0
1

9

http://twistedpencil.com/
number of 1s in each row and column as the set
of codewords in a code (we could write them out
as a long string instead of as a rectangular array
if we wanted). We pointed out before that the
rectangular cup code is 1-error-correcting. It turns
out that the minimum distance of the code is 4
(we will see this below) so that the code is 3-error-
detecting and 1-error-correcting.
Linear Codes
A code C is called a linear code if u + v is in
C whenever both u and v are in C. Observe that
u and v are allowed to be the same! In this case,
u + u = 000 ... 0 is in C. Therefore, linear codes
always contain the zero word, which we usually
abbreviate simply as 0. Structure like this makes
linear codes an important branch of coding theory.
For example, rather than checking all pairs of
words to find the minimum distance, things are a
bit simpler for linear codes.
Theorem 1: The minimum distance of a
linear code C is the smallest weight of a
non-zero codeword in C.
Our strategy to prove this will be the following:
if we let m stand for the minimum distance of
the code and w stand for the smallest weight of
a non-zero codeword, we will show first that m ≤
w and secondly that w ≤ m. Of course this proves
that they are equal. This method for showing that
two numbers are equal (prove that the first is no
bigger than the second; then prove that the second
is no bigger than the first) is used all the time in
higher mathematics.
Proof. Let u be a word in C of weight w. Since
10
d(0,u) = w and 0 and u are both in the code, we
see that the minimum distance of the code is less
than or equal to w (i.e. m ≤ w).
On the other hand, since m is the minimum
distance, there must be words v and w in
the code with d(v, w) = m. Since the code is
linear, we must have v + w d C. We now have
d(v, w) = wt(v + w) from earlier. However,
we also have d(0, v + w) = wt(v + w) so that
d(0, v + w) = m. It follows that C contains a
codeword (v + w) with weight m so that the
minimum weight of a non-zero codeword is less
than or equal to m (i.e. w ≤ m). This shows that
w = m.
It turns out that the rectangular cup code that
we described before is a linear code. We can see
this as follows: suppose u and v are two codewords
(so that each has an even number of 1s in each
row and column). Let's set w = u + v and let's
think about any row of w. Suppose that u had 2s
1s in the row and v had 2t 1s in the row. Let's
also assume that there are exactly r places where
both u and v have 1s (r can be an even number
or an odd number). Now, thinking about how
binary addition works, we see that w has a 1 in
a spot exactly when u had a 1 there and v had
a 0 or vice versa. There are 2s - r places in the
row where u has a 1 and v has a 0, and 2t - r
places v had a 1 and u has a 0. This means that
the total number of 1s in the chosen row in w is
(2s - r) + (2t - r) = 2(s + t - r) - an even number.
This argument works for every row and every
column so that we see that w is in the code. It
follows that the code is linear. We now use this
together with Theorem 1 to work out the minimum
distance of the code. This is the same thing as the
minimum weight of a non-zero element.
To be a non-zero element, it must contain a
row with at least two 1s in (remember that rows
have to contain an even number of 1s). The
columns containing the 1s need an even number
of 1s in them, so there must be at least one further
1 in each of these columns. This shows that the
minimum weight must be at least 4. In fact the
minimum weight is exactly 4 as we can think of
the configuration with 1s in all four corners and
0s everywhere else. This proves that the minimum
distance is 4 as we claimed before.
Every linear code has a dimension, which
we now describe. Given a set of For example, our earlier code
words S, the linear span of S, C = {000, 101, 110, 011} has
span(S) is defined to be the set of information rate 2/3. At one
all words that can be obtained as extreme, we have the code {000 ...
a sum of elements of S (counting 0,111...1} with one basis element,
the 0 word as the sum of none of Figure 3: 2-hexagon with associated and information rate 1/n. At
the elements of S). binary word the other extreme, we have the
Consider the following process code consisting of all words of
applied to a linear code C. First, set S = Ø. length n, having rate n/n = 1. But the first code
We now keep repeating has distance n (and can detect n- 1 errors), while
the following step until the second code has no error detection at all, since
span(S) = C: Notice that its minimum distance is 1. Both of these codes
span(S) is a subset of C. If are rather useless. Coding theory aims for nice
span(S) ≠ C we add compromises between rate and distance.
any element of C that For further reading on coding theory, consult
is not in span(S) to S. the web or reference [1].
When we have finished
4 Hex-arrays and correcting two errors
doing this we end up with
a set S with span(S) = C.
Figure 4: 5-hexagon By an n-hexagon, we mean a pattern of cups
The set S is then called a
(or data positions) arranged so that each interior
basis for C.
cup is surrounded by six neighbouring cups forming
For example, suppose our code is a regular hexagon, and the boundary cups trace
C = {000, 101, 110, 011}. Start with S = Ø out a regular hexagon with n cups per side. For
and add 110 to S. Now span(S) = {000, 110}. instance, a 1-hexagon is just a single cup, and a
We add 011 to S so that S = {110, 011}. Now 2-hexagon has seven cups as in Figure 3 (a).
span(S) = {000, 110, 011, 101} = C so that
The translation from cups to binary words is
{110, 011} is a basis for C. Observe that all four
illustrated in Figures 3 (b) and (c). The array for
words in C are found by summing either none, one
a 5-hexagon is shown in Figure 4.
of, or both of these basis words. In general, the
sum of any subset of the basis words gives a unique By the way, it is a fun exercise to get a formula
codeword in C. for the number of cups in an n-hexagon. Notice
that there are several
Theorem 2: Suppose there are k codewords geometric differences from
in a basis for C. Then C has 2k codewords. the rectangular array.
Corollary: Every basis for C has the same First, the rows go in three
number of codewords. In other words, the order in directions: horizontal,
which words are crossed off in the process does not northwest, and northeast.
affect the number of remaining elements. Also, the number of cups
This common number of basis elements for a in a line is not constant:
linear code is what we mean by its dimension. it depends on how close
If a linear code with dimension k and length n the line is to the centre. Figure 5: A weight 6 codeword
is used, we say it has information rate k/n. However, we can still insist
Thus, in the Example above, C has a dimension that a line is balanced if
2. This is because if the dimension of the code is the number of cups in it which are the right way
k then there are 2k codewords which means that up is even.
sending a single codeword carries the same amount Going back to the magic trick, this time on a
of information as k bits. Since the actual length of hexagonal array, the argument we gave earlier
the word is n bits the information rate is k useful shows that the set of codewords (the arrangements
bits per n bits sent. that are balanced in each of the three directions)

11
 is a linear code. In fact, the but quite a good way to do
dimension of this code is it is to put the cups in an
the number of cups in an (n-1)-hexagon in any
(n-1)-hexagon — see below arrangement at all. It turns
for more on this. More out that if you then try to
importantly though, we extend the configuration
want to find the minimum to an n-hexagon, there's
Figure 6: A 3-hexagon of cups, distance in the code. By always exactly one way Figure 7: Two flips on different
balanced in every line
Theorem 1, this is the same to do it. Another way to lines.
as the minimum weight of a non-zero configuration, proceed is to start with all
balanced in every line. cups down, and repeatedly perform six flips as in
One possible configuration is shown in Figure Figure 5. (There is also a way to wrap around the
5. Check that all lines are balanced, so that the boundary by flipping eight cups).
minimum weight in this code is no more than 6. Next you should invite your victim to turn over
On the other hand, we can see that the minimum any two cups. As the magician you have to figure
weight is at least 6 as follows. In order to be a non- out which two cups were inverted. Look along the
zero configuration, there must be a horizontal line horizontal and diagonal rows and see which rows
which is not all 0s. If there is just one horizontal have odd numbers of upturned cups.
line that is not all 0s then we're in trouble because There are 2 basic possibilities. In the first
a diagonal line that goes through one of the 1s case (see Figure 7) there are two rows in each
would have just a single 1 on it (an odd number). of the three directions that have odd numbers of
This shows that there must be at least two upturned cups. In this case there will be exactly
horizontal lines with 1s on them. Similarly for the two spots where three lines cross. These spots are
northwest and northeast directions. If there are 3 where the cups were inverted.
or more lines in one of the directions with 1s on
then since each line contains at least two 1s, we'd In the second case (refer to Figure 8) two
have at least six 1s in total. cups were inverted on
the same line so that
The only possibility to rule out is that there are this line ended up
exactly two lines in each of the directions and each with an even number
line has exactly two 1s so that the total number of upturned cups. In
of 1s in the configuration is 4. Considering the this case there are two
two diagonal directions first, the only possibility directions that have two
would be that the 1s form a parallelogram with unbalanced lines and in
1s at the corners and edges parallel to the two the third direction all
diagonal directions. It is then easy to see that the Figure 8: Two flips in the same
the lines are balanced. line.
top and bottom of the parallelogram are in lines The unbalanced lines
with a single 1. taken together trace out a parallelogram. One of
This shows that the minimum weight is at least 6 the diagonals of the parallelogram is in the third
and so the minimum distance of the code is exactly direction. The two corners of the parallelogram on
6. According to our observations from before, this this line are the cups which were inverted.
means that we can detect that up to 5 errors have
been made (but not which errors they are), but since REFERENCES:
6 ≥ 2 x 2 + 1, we can correct any 2 errors!
Pless, V. 1982, Introduction to the Theory of
Finally we should do what a good magician
Error-Correcting Code, John Wiley & Sons Inc.
should never do and tell you how to perform
the magic trick in this case. First you need to
set up a hexagonal array of cups so that in each
direction there are an even number of cups the
right way up. This isn't quite as easy as it sounds
12
 By
Carole (C) was sitting on the beach Klaus
answer is: very little. For now, let us just say that
Hoechsmann
watching her little brother Ivan University of Grant was ogling Judy through his binoculars. This
(I) play in the water. It was only British Columbia whole story could be told entirely without them,
knee-deep, and she was not worried except that Judy stubbornly insisted on being part
about him at all. He always of this excursion, and as she
enjoyed clowning around, C went so did Grant. We shall
grimacing and gesticulating. say more about them at the
When he suddenly cried end of this tale.
\Help!" Carole just waved Fred's real name was
at him, but when his head Fernando, and at school they
disappeared under water and D H G
called him \Ferd the nerd"
stayed there, she jumped up F
because he liked mathematics.
and ran to help. When Carole had told him
Fred (F) was impressed to about Snell's Law, his first
see her veer toward Henry I reaction was to pull out a
(H), almost stepping on piece of paper and scribble.
Donna (D), instead of going \It's easy, Carole," he said,
J
straight for Ivan (see Figure \.. but, hm, yes, I am using
1). But, Carole was a certified derivatives. There's got to be
life guard and knew that she another way: that Dutchman
would get there faster by this Snell died in 1626 — long
little detour, and Fred knew Figure 1 before Newton was even born."
why this was so. She was used to that kind of stunt from him, but
Since she ran twice as fast on the sand as through this much detail astonished her. \How come you
the splashing water, she had learned that the cosine remember that?" she wanted to know. \Ten years
of the angle FDC should be twice as big as that of after Shakespeare," was his answer.
HDI. The general rule was The next day, he was back with two neat
that these cosines should drawings, the first of which is shown here in Figure
be in the same ratio as 2. That was his style: only rarely did he use algebra
the respective speeds. The with Carole because he knew it did not convince
instructor had said that her. She had taken an \interdisciplinary" course,
this was just a fact, called where she learned to speak about derivatives and
Snell's Law, and that you integrals but not to work with them. Fred spent
could not understand it hours with her using reams of graph paper, until
unless you knew Calculus. one day she exclaimed: \You mean, the steeper the
The attentive reader curve, the faster it's moving away from the x-axis?
might wonder what Grant And that steepness, is the derivative? Wow!"
(G) and Judy (J) are doing Their conversation about the new diagram was
in the picture, and the rather private, so we'll paraphrase it. If you go
Figure 2
13
 back to the beach picture and imagine
D moving from F to H, you'll easily
see that CD would be increasing
while DI would be decreasing. As
The
you slide your pencil-tip from left
general to right along the horizontal
rule was
line in Figure 3, its distance
that these from the upper green area is
cosines should
increasing (like CD), while
be in the same ratio its distance from the lower
as the respective gray area is decreasing (like
speeds. DI). As this drawing is not
to scale anyway, we might as
well regard the lower distance as
representing not DI itself but mDI, where m is
a positive numerical factor reflecting the ratios
of the speeds on sand and through water. Then
CD + mDI would be represented by the vertical
distance between the green and gray regions,
right?
That distance is minimal at the place
where they would just barely touch if you
slid them together vertically. \Yes," said
Carole, \the red points would be kissing."
\And the chaperone could separate them with
one straight thrust of her cane," Fred added.
C \They'd share the same tangent line, you mean,"
Carole sighed, \(you are so romantic,
Fred) — if you don't wish to slide respectively, what is the total time it takes her to
them together, let us say, it's reach Ivan? What is the meaning of the factor m?
How does it show up in the segment FG?
where their boundaries have the same steepness."
Back to the beach scene. As D’ moves toward D,
the triangle FD’C gradually morphs into FDC, and
we wish to compare the growth of the hypotenuse
(CD’ to CD) with that of the side (FD’ to FD). If
we mark D” on CD so that CD” always equals CD’
in length (i.e. the green triangle remains isosceles
throughout), the changes in hypotenuse and side
are represented by D”D and D’D, respectively.
Now, as D’ slides toward D (in synchronisation
with its twin D”), the green base angles approach
90 degrees, and the quotient D”D/D’D approaches
— hold your breath — the cosine of the included
angle FDC.
So, now you have all the ingredients to roll
your own proof of Snell's Law. But how Grant
and Judy fit into this ? Well, it so happens
that in issue Number 7 [1] some graphics expert
thought it would be pretty if the whole picture
were inside a circle, with Donna in the centre
and Ivan on the rim (where Judy is now). With
CD = DI of equal length, CD + mDI was constant,
and the argument was sunk. But who needs a
theory when you have facts?
Exercise
If v and w are Carole's speed on land and water,

References
D‫״‬
[1] Judith V. Grabiner, How Looking for the
Best Explanations Revealed the Properties of
F D′ D Light, Pi in the Sky, September 2003, 20-22.
Figure 3

http://xkcd.com/
14
 Interview
by Gordon
Hamilton Reiner - I want you to imagine a Reiner Knizia has published hundreds
of board games including eleven in
triangle with the points labeled the top 100 on
\Entertainment Businessman," www.boardgamegeek.com
\Creative Artist" and \Scientific Researcher" "Genius
is 1%
(researching \What is fun?") Where would inspiration Number 2, Pages 375-394]. Do you
and 99%
you place yourself? perspiration.” I think games are under-utilized in
Where are you on mean mathematics classrooms?
Least on the Scientific Researcher side… Thomas Edison’s 99 and 1
and probably in the middle between the
Scale? is probably a It is difficult to say. I think
Ente ssman
Busin

little bit extreme...

Entertainment Businessman and the I think you actually
you can teach math without
rtain

need both. You need to using games at all, so it's not

e

Creative Artist - because it takes both. have a good mind and you
men

I think this is one of the main success need to have the capability to
a necessity; there are many
do things, but then it very much different methods of exciting
t

criteria in game design: that you can depends on discipline and

handle both. There are lots of artists out the willingness to put in the kids and making them
there who are very good artistically, but
work to produce a really want to learn. I think the
good product.
they are chaotic and don't get themselves important thing in learning
organized… and if you organize yourself, is that you have a positive
but have no artistry (and game design emotional experience; If the emotions
is an art) then you are also kind of are not there the learning is very flat.
empty handed. Getting both of them If you can create positive emotions
combined in your life is the big in the learning process you get by
challenge. far the best results – and this is
where games can help a
Artists are sometimes called Scien
t
Rese ific lot. If the learning
upon to protest with their art. Do arche
r content is almost
you ever feel the urge to protest in the background, almost unnoticeably
through your art? presented, then I think we have an ideal situation.
The honest answer is The learning needs to become fun. I think games
tis ive
Ar reat

probably `No.' I think are an ideal tool for this – they can be used much
t
C

games are there to entertain
people, and to bring enjoyment to people. I think
there is no mission beyond that. However there are
nuances: We have educational games where you
engage kids emotionally so that they are open to
learning.
Last year, a randomized controlled trial found
some games improved mathematical ability [Child
Development, March/April 2008, Volume 79,
15
more widely than today, but they are not the only
tool.
Reiner, to avoid becoming biased, you don't
play other people's games.
You're right - When you design games you have
to take lots and lots of minute decisions. All these
decisions have to be taken, and if you already
know a decision from somebody else, it is much
 more difficult to think of a new one. OnceAnd one of the great mentors was my professor
you know how a light bulb works – this is
with whom I did my thesis. He was a professor in
“In
order
the way to do lights – and if you don't
Germany and he was a professor in America, and
to write, know you might come up with a very he was working in Analysis so naturally I drifted
you have to different solution… towards this. I learned a lot from him. With
live”
Ernest As a Masters and PhD student respect to mathematics, but also how to approach
Hemingway
did you also protect your mind from problems generally. I could probably have worked
bias? in many fields, but I knew him from very early
stages in my studies so we kept working together.
No I did not do that in mathematics I did my diploma with him and then my doctoral
– there were many things to learn – and I had thesis with him.
great fun in learning them… and I didn't try to
re-invent the wheel – so I was quite happy to take Apart from your mentor, who are you standing
in the knowledge from other people and because in on the shoulders of – or is game design too young
mathematics it is not about solving the problem for that?
in a different way – once it has been solved it's In game design I probably don't stand on
solved – it's about finding new knowledge – and anybody's shoulders. I think I have very much
the more you know, the more new knowledge you developed my own style and learned by doing and
can create. I think game design is a very peculiar profession in
Let's go back to some of your high school a way because you can't really study it.
experiences with math. Can you think of an If you can't study it, and you don't play other
instance where you thought \Mathematics is really people's games - how do you stay fresh?
beautiful!" My teacher always said: \If you don't read your
I remember looking at mastermind, and working daily newspaper – you will always be uneducated
out an algorithm: What do you need to ask, and and stupid." I have always been uneducated and
when you have the answers what do you need to stupid because I've never read it regularly in my
do to actually solve it? I tried to work out the life. (laugh-out-loud) I have a different philosophy.
complete algorithm and therefore prove that you I'm not interested in day to day news which is
could always do it in so many steps. I applied that relevant today to talk to people and next week
approach to quite a number of logic puzzles so is totally irrelevant. I prefer to read books rather
I was always fascinated by combinatorics – how than newspapers because they have long term
many different ways are there to do this or that. relevance.
The most fascinating questions in mathematicsIt is very important to take in the world.
for the general public are probably coming from
Hemingway said so nicely \In order to write you
number theory: \Okay, so I understand what a have to live" and the same is true for game design.
prime number is. So how do I know that there Games are a mirror of our times. We have a very
are an infinite number of prime numbers? How hectic life and a small attention span - so the
many pairs of prime numbers [x, x+2] are there?"
dynamic of the game has to be much much greater
You can ask very simple number questions which
today than 50 years ago. Half an hour playing time
are unsolved in mathematics and people can really
is almost too long. There are a lot of things that
understand. That is rare – most of the questions in
are influenced by our times and I need to take
mathematics today are complicated and you first
these in if I am to remain successful. So, as well as
need years to understand the question. reading, I stay in contact with people - particularly
I was perplexed that you chose the area of with the younger age group. I learn what the kids
mathematics that you did. You just named like and then I build that into my design process.
combinatorics which was where I would have You`re quoted as saying \Other people steal my
pegged you. Why not? ideas before I've ever had them."
I think – apart from finding a mission in life Yes–absolutely!
it is extremely important to find good mentors.
16
 What happens if I put
Is this a big motivator? area? Well then our some holes
Yes it is, because I do believe there normal understanding
in there?

are lots of games in the universe (like of area starts to fail,

If you
don’t read
your daily
newspaper - you
mathematics) and I believe that lots but if the process still
will always be of these games should be mine, and leads to a result we
belong to me (I'm kind of smiling when still have something
uneducated and
stupid
The Board Gaming Genius:
I say this)... and the panic always grabs tangible to talk about.

Reiner Knizia
me when I walk around a fair and I look What I worked on in
at games – and I say \hmmm – yes – that's one my thesis was a very
I should have developed – that's the one I should general mathematical
PhD Mathematics
have found - that should have been mine, I'm too process (which we

Ente essman
Busin
late there and somebody else stole it." I mean this call the integral) to calculate the \area" for very
Interview by Gordon Hamilton
is meant to be funny, but there is a sense of urgency very crazy shapes where the normal processes that

rtainm
that overcomes me in these fairs... calculate area no longer work.
to imagine aYou did it ;)

en t
There are the right times Reiner
for- Icertain
want youthings triangle
when the technology becomes withripe.
the points labelled “Entertainment
For example, One last quote: \Life offers so many good
Businessman”, “Creative Artist” and “Scientific
I will soon have a game on the market that works choices, you can never take them all, and that's
Researcher” (researching “What is fun?”)
with transparent cards... I was very
Where anxious
would to get
you place good, because it makes life so rich." That's by
yourself?
it out quickly... I might have been the first one to Reiner Knizia...
Leastcards,
design a game with transparent on thebutScientific
it tookResearcher side…
Absolutely – that is one of the absolute
and probably in the middle
a while to get it to market and now I see one or statements between the
Entertainment Businessman and the Creative– I don't want to be bored – I think I
two other products coming Artist
out which means I'm I think Reiner Knizia has publis
- because it takes both.can onlythisdo
is one or two things in life reallyboard
properly.
games including
only second or third. It's this
oneurgency.
of the main When thecriteria
success Therein game
are many other routes I could have100 taken:
on www.boardg
time is ripe you need to do design:
it andthat youyou
then canneed
handle Iboth. There being a board member, and running
enjoyed
to be fast. are lots of artists out there who are very
financial companies, and doing mathematical
good artistically, but they are chaotic and
You haven't asked medon’t thegetmost difficult
themselves organized… and – I enjoyed all that, and it was a great
research
if you
question, which was to explain myorganize yourself,
PhD thesis to but have no
satisfaction, but I don't think in the end it was
a high school audience. artistry (and game design is an art)my
thenmission.
you are also kind of empty handed.
Other
Well - I subsequently looked at your
Getting boththesis, people in As a young person, you need to find your
of them combined
and thought that it would your life istough
be too the big achallenge. my
steal mission in life and then to have the courage
question! ...but go for it if you wish! ideas before to follow it. I only found my mission when
Artists are sometimesI’ve everupon
called had
I got to 40, and I jumped and said \now I
...There is this funny thingto protestcalled anart. Dothem
with their you ever
integral that we use to calculate do games full time" – I've never regretted it
feel thetheurgearea of through
to protest
certain shapes. Of course we yourall
art?know very nice and I've found my niche. I know I miss out on
formulas that tell us how to calculate the area of almost everything in life,Last
butyear,
as ayou
randomized
say \you controlled
can trial f
e some games improved mathematical ab
specific shapes: The area ofThe honest answer is
a rectangle... length ahave tiv ist anything, but youDevelopment,
can't haveMarch/April
everything."
2008, Volume
probably “No”. I think
re ItsArtlots of choices we have Pages take – and
to375-394]. if think
Do you thesegames ar
times width; the area of triangle...
games are lengththere totimesC
height divided by two. It gets more people,
difficultandwhen are rich choices and in
you mathematics
have lots ofclassrooms?
alternatives
entertain
you want to calculate the area to bringofenjoyment but then
a circle to– people. I thinkit's a is
there great life.
no mission
It is difficult to say. I think you can teac
beyond
there is a method there as well – andthat.
thenHowever
you ask,there are nuances:
And We reflects
this also have into games. A good game
using games at all, so it’s not a necessity
educational games where you engage kids emotionally
\What happens if I put some holes in there? Can I for me is one that givesdifferent me rich methods
choices.of exciting
I do notkids and ma
so that they are open to learning.
to learn. I think
still have a process to calculate the area?" If I have want to find the least bad option. I want to have the important thing in lea
twenty holes can I do it? If I have a positive emotional
so many good options. I want experience;
are not there the learning is very flat. If
have a thousand holes can topositive
sit there biting my finger
emotions in the learning proces
I do it? Yes – because I can nails
the best andresults
hoping thatthis
– and when
is where ga
still count them, but what if itlot.comes
If the learning
round to content is almost in t
me that
I have uncountably many? almost is
option unnoticeably
still there.presented, then I th
ideal situation. The learning needs to bec
Can I still calculate the
games are an ideal tool for this – they ca
more widely than today, but they are not
 Higher Derivative
Higher Derivative Equations Equations andofthe
and the Laws Laws of Physics
Physics
Neil Barnaby Equations and the Laws of Physics Neil Barnaby
Higher
Higher
Canadian Derivative
Derivative
Institute for Equations
Theoretical and the
Astrophysics, Laws ofof Physics
University Toronto
Canadian Institute for Theoretical Astrophysics, University of Toronto
60 St. George 60 St. George
Street, Toronto, Street,
ON M5SToronto,
1A7, ON M5S 1A7, Canada
Canada
Neil Barnaby
Email: barnaby@cita.utoronto.ca
Neil Barnaby
Email: barnaby@cita.utoronto.ca
Canadian Institute for Theoretical
Canadian Institute for Theoretical Astrophysics, Astrophysics, University
University of Toronto
of Toronto
Higher Derivative Equations and
60 St.
60 Februarythe
George
St. George Laws of
February
Street,
26,Street, Physics
26, 2009
Toronto, ON M5S
2009 Toronto, ON M5S 1A7, Canada 1A7, Canada
Email: barnaby@cita.utoronto.ca
Email: barnaby@cita.utoronto.ca
Neil Barnaby
The dynamical laws of physics at
are February
described 26,time
2009
Canadian Institute for
The dynamical lawsTheoretical Astrophysics,
of physics are described time tofisat
February
University
any any
26, 2009
Toronto
By
described byt is
thedescribed
function by
x= the function x =
by
60 St. George
by differential differential
Asequations.
Street, Toronto,
equations. ON M5S
a theoreticalAs x(t).
a1A7, Canadaisx(t).
theoretical
Velocity Velocity
theNeil
rate is the of
of change rate of change of position
position
physicist,
Email:
physicist, a big part a big part of the job
barnaby@cita.utoronto.ca
of the is solving those
job is solving those with respect to time with respect
Barnaby to time
equations
The The indynamical
dynamical various laws laws
settings ofand,
of physics physics are described
sometimes,
are described Canadian
at anyat any
timetime t is described
t is described by the by function
the function x =x =
equations in various
The dynamical settings
laws of and,
physics sometimes,are described by dx
by up differential equations. Institute for
dx Velocity
coming up withcoming by
newdifferential February
equations with that new 26,
equations. 2009de-Asthat
equations
might aAsmight
theoretical de- x(t).x(t).
a theoretical
v =Velocity , is the vis=rate
the rate , change
of of change of (2)
of position position (2)
differential equations.
scribe physicist,
nature As ina asome
big theoretical
part of
unexplored the physicist,
job is
regime. solving For those Theoretical
with dt respect to timedt (2)
physicist,
scribe nature in some unexplored regime. For a big part of the job is solving those with
Astrophysicsrespect to time
a big part ofme, theinventing
job is solving
equations in various
new those
theories equations
settings
ofand,physicsand, in sometimes,
is the and andis the acceleration is the rate of
me, inventing new equations
theoriesin of various
physics settings
is the andsometimes,
acceleration is acceleration
the rate of change rate
of the
dx dx
of change of the
variousofsettings and,
coming sometimes,
up with atnew coming
equations up with
that might the functionchange
de- x = v of the velocity
dynamical
most lawsexcitingphysics most
partcoming
of exciting
are thedescribed
upjob. with partOfnew of the
any job.
equations
course, one Of
time iscourse,
tthat described
velocity might onede-byvelocity = v =2, dt , (2) (2)
new equations thatscribe might nature describe
in some nature
unexplored in some regime. For dv dt d x
erentialcan’t just startcan’t
equations. As
scribe
writingajustnature startinequations
theoretical
down writing
some downhap- equations
x(t).unexplored
Velocity isregime.
the rate hap-
For of change dv of position2
d x a= = (3) . (3)
unexplored regime.
hazardly me, For
and me,
inventing
expect inventing
new
these to new
theories play theories
of
some physics
role in is a
the= and = .
acceleration is thedt2 rate of change of the(3)
st, a big part of the
hazardly and expect job is solving
me, theseinventing those
to play new with
some respect
theories to time
role inof physics is the anddtacceleration dt2 isdtthe rate of change of the
of
ns in nature. physics
various Theresettings is the
nature.
and, most
mostThere exciting
sometimes, exciting
are a part
lotpart
of of the
of
criteria thejob. job.
that Of we Ofcourse,
must one
Using velocity
(3) wethe could replace the
aremosta lotexciting
of criteria
can’t just
part that
start
of we themust
writing
job. Of
down Using course,(3) we
equations
onecould
dxUsing hap-
velocity
replace
(3) we could acceleration,
replace dvdthe a, acceleration,
2 x dacceleration,
2x
a,
a, in
up with course,
new one
equations use
can't
can’tto
that guide
just
might
just our
start
start de- attempts
writing
writing
use to guide our attempts to come up with new in (1) by adtsecond to
down come
down up
equations with
v = new
hap- , in (1) by a second
(2) dv
derivative
a = = of position
. with
naturephysical
inequations
sometheories. physical
unexplored hazardly
theories. and expect
For these
example, to
anyplay some
candidate (1)
role by derivative
in a second of position
derivative
a= with
=dtof position . 2
2 dtdo something
(3) (3)
with respect
Forregime.
haphazardly
hazardly andand
example, For
expectanyexpect these
candidate these
to play some
respect role
to we in
time. respectWeWe to
cancan time.
also We
do do dt can
something also
dt
for a nature.
new isThere
law of aarelotaof
physics lotmust of criteria isthat tomust time. also something thesimilar with the
ventingfornew
toa playtheories
new some
law ofof physics
role
nature.
physics inTherenature.
must the
are be and
There
studied in be
acceleration
are
criteria that
similar
studiedwethe
with
in
rate
must the of
similar
Using
force,
Using
change with
(3)
F .
(3)
ofthe
we
The
we
the could
force,
could
particular F
replace.replace
The
formtheparticular acceleration,
acceleration, form a, a,
detail use
to to
make guide sure our attempts
that its to come
predictions up with
don’t new force,
in F.
(1) byThe a particular
second form
derivative of F depends on
xcitingdetail
part
a lot of the
to of
make job.
criteria
sure Of
use to course,
guide
that
that itswe ouronemust velocity
attempts
predictions usedon’t to
tocome up with new
In F depends 2oninthe
of of F depends
(1)system
by a second on
under the system
derivative
considera- of position withwith
under of position
considera-
ust start writing ourdown conflict physical
equations with theories.
what
hap- is For example,
already known any
about dvcandidate
na-d xtion. the In system
respect
general under
to the
time. force consideration.
We can
acting also do In general
something
guide
conflict with what physical
attemptsis already theories.
to come
known For
up aboutexample,
with na- any
this
tion.a= candidate
In studied
general
= 2 the respect
. force toacting
time.
(3)on We somecan body alsoon dosome body
something
toture. for
aIn a
this newarticle,law of
one physics
of the must
things
article, be
I wantdt toin in the force
similar acting on some body might depend
y andture.
expect these
newInphysical
this article,play
for some
theories.
one new role
of thelaw
For in
of physics
example,
things I want must
toone be studied
might
of the depend dton might
similar
the withwith
depend
position, onx force,
the the
(for
force,
position,
F . The
example
F . xThe (forparticular
particular example form form
Thereaccomplish
are a lot of accomplish
criteria detail
that we to
ismustmake
to give sure
you that
a senseits predictions
of what it’s don’trestoring on
of F the position,
depends
force on x
the (for
system example,
under restoring
considera-
any candidate is to detail
give foryoutoa make
conflict
anewsense
with
sure
lawofUsing
what
that
of (3)
what its
it’s
is already
predictions
we
things could
I want replace
restoring
known
don’t
forcethe
about of ofaacceleration,
na-
Fspring),
As(fortion.
depends or of a,athe
on spring),
system
velocity, or under
on theconsidera-
velocity,
uide our attempts
likephysics
to try to to like
come to
conflict
develop uptry withto
new theories develop
new
what is
in new
already
(1) theories
by to known
accomplish
a second
of physics. v (for example the of physics.
about
is na-
derivative v of
tion. force
example
position In
In general of
with a
general
the spring),
the
viscous force or
drag on
acting
as the
a on
bodyvelocity,
some is bodyv
must be ture.
studied Incan in detail
this article, onetoup of
give the
you athings
senseequa- I want a viscous
to drag the as force
a body acting
is on some body
l theories. For
Before example,
we sure
can try Before
ture.any In
to come we
candidate
this try
article,
up withrespect to come
one
new equa-of the with
things
to time. new I want to
We cantheoretical moved
also
might might
do depend(for
through depend
something example on
molasses).
on means the
the
the position, position,
viscous
This meansx
drag (for as example
that, a body
to make that its
accomplish predictions is to give of whatmoved
you a it’s like tothrough
sense of what molasses).
it’s This that, x (for example
ew law of
tions physics tions
must
need toaccomplish
we conflict be we
understand need
studied to
isthe in understand
to existing
give youlaws
similar the
a try
withsense existing
tothe offorce,
develop what laws
Fit’s even
. physicist,
The restoring
awithout
particular
restoring is form
force force
moved
specifying of
of a spring), a spring),
through
the or or
physical on
onmolasses). the
the system, velocity,
velocity,This
don't with likewhat is already
to Perhaps
try to develop new even
theories without
ofsystem
physics. specifying the physical system,
o make sure
of known that
physics. Perhaps its of physics.
predictions
like to
the try todon’t
most famous develop the
of new
F most newfamous
theories
depends theories
on
example we can write Newton’sof example
of
physics.
the big part
we of
undercan v
the (for
write example
considera-
v (for example Newton’s
means the
that, law
thedifferential viscous
in
even true drag
without
viscous drag as a body as
differentiala body is is
specifying
about nature.
Before we Incanthis try to come physics.
up with job isequa-
new solving those law in true
with iswhat is already is Newton’s
known
Newton’s law Before we can trytion. about law na- to come up with
In general new
the equa-
force actingform moved
as
on some through
bodymolasses). molasses). This means that,
article, one of the tionsthingswe I
need wantto to
understand formthe as equationsmoved
existing laws in various � d2the
through
x� physical1
� system,
dx
This
� means we can that,write
n this article, one of thetions things weI needwant to to understand
might depend the existing
on thesettings laws
position, 2 even
x 1(for
d x sometimes, without
example = specifying
dxNewton's the the
law physical
in. true system,
differential
lish is toaccomplish
give you a is F = ma ,what (1) and,even without specifying F x, physical system,
(4)
ofto give you aPerhaps
sense of it's
sense of
Fphysics.
= physics.
ofma what , Perhaps
it’s the (1)most famous example
coming up =withwe F can
new x, write
dt 2 . m
Newton’s (4)
dt
law in true differential
the most famous
restoring force of a spring), example dt orweonm
2 canthewrite
velocity,
dtNewton’s
form as law in true differential
like tonew
try to develop try theories
toisdevelop
Newton’s new
isofNewton’s
physics.lawtheories
lawv (forofexample physics.the viscous equations that form
drag
For any
might as
as a body
given is 2 of F �we can
choice � find� a unique
that tells us how a body of mass,
m, accelerates For any given m, accelerates form
choice as
of in
F we can
thattry
e we can tells
to us howup
come a body
with of mass,
new equa-
describe nature
d 2 x find
d x1a unique 1 dx � dx
Before we can
in applied try
responseforce, to come
to anFapplied moved up
Fma with
through
force,
= ,ma F,solution new molasses).
. Sometimes somethe
once
This
solution
unexplored meansonce
(1)initial conditions that,
the initial
= =F =
2x(0)
conditions
x,Fx0x, dt . x(0) . = x(4) 0 (4) (4)
e needinto response
understand
equations
to an
we the
when need existing
we to
are laws
understand
first
. FSometimes
introduced
=
even withoutthe to specifying
existing
equation (1)
(1)
it regime.
theand physical
v(0) = system,
v 0 are2 dt
dt provided. m m dt that only two
Note
when wethe
ics. Perhaps aremost
first famous
introduced example to equation (1) it and v(0) = v0 are provided. Note that only two
laws of physics.
is made
that that
Perhaps
to
tells tells
look
us how us
the
like how
amostawe
body a can
body
simplefamous write
of mass, of mass,
algebraic Newton’s
example
m,
initial m,
relation,
accelerates law in
accelerates
conditions Fortrue
initial
areany
For Forgiven
any
differential
conditions
given
needed any are
choice given
choice
corresponding of F of
needed choice
weF to
weof
can canF find
corresponding
find awe acan
uniqueto find
unique
is made to look like a simple algebraic
on’s law relation,
form as �a (4) unique solution
solution once the initial conditions
initial x(0)
= x 0 x0
conditions =
really is but
butis itNewton's alaw itinreally
indifferential
responseresponse to aanto
isequation! an applied
differential
applied Toforce,see
force,
equation!
d
F .To
F2 .x Sometimes
the fact 1
Sometimes
see
that the
dx is� fact
solution
second that
once (4)
order the inonce
is second
initial the
derivatives order
conditions in derivatives
x(0) 2
.and v(0) v0= v0provided.
are
= vprovided. Note
that that ).onlythattwo
when we firstare first introduced to equation d
this, suppose
F = ma that , this,
when thesuppose
we are
position that thethe
(1)x introduced
of position
massto xequation
(itofcontains
2
the mass
= (1) (1)x(0)
F it(1)(it
x, it=contains
and
derivatives x0no
v(0) and
=higher v(0)
derivatives
are
(4) that no 2 are
higherprovided.
0 ). Note that only
d Note
dt2 two
is made to look like a simple dtalgebraic m relation, dt
onlyinitial initial
two initial conditions
conditions dt2 needed corresponding to
are are needed corresponding
is made to look like a simple algebraic relation, conditions are needed corresponding to
ls us howthata body tells us how
of mass, m, but a itbody really
accelerates of ismass,
but it really is a differential equation! a
For any m,given
differential accelerates
equation!
choice of F
To see To
we
to1 the see
can
the fact the
find fact
a
fact that that
unique
that (4) (4)
(4)isissecond is
secondsecondorder order
order in derivatives
in derivatives
in derivatives
in applied
response toFthis,
an applied
suppose force, theF.position 1
Sometimes d2
onse to an force,this, . suppose
Sometimes thatthat solution
the position once xthe of xinitial
the of mass
the(it mass
conditions
contains (it x(0)
(it contains contains =
derivativesx0derivatives
derivatives nonohigher no higher
higher thatthat
that d2
dt2
). dt 2 ).
when we are first introduced to equation
e are first introduced to equation (1) it and v(0) = v0 are provided. Note that only two (1) it
Newton's law to is not the only example of a
to look islikemade
a simple to algebraic
look likerelation, a simpleinitial algebraic
conditions relation,are needed corresponding
1 1
eally is but it reallyequation!
a differential is a differential
To see theequation! fact that (4) Tois see second order in derivatives in physics. There are the
differential equation
ppose that this,thesuppose
position that x of the themass position x of the
(it contains derivativesmass noNavier-Stokeshigher that dt d2equations that describe the dynamics
2 ).

at any time t is described by the function of fluids, the Maxwell equations that describe the
x = x(t). Velocity is the rate1 of change of position laws of electro-magnetism, the diffusion equation
with respect to time that describes the propagation of heat through a

18
namics of fluids, the Maxwell equations that ten librium equation position (5) by as x(t). As long as x(t)
an approximation is not
because
describe the laws of electro-magnetism, the dif- at larger displacements there will be nonlinear too large, then the restoring force is propor-
fusion equation that describes the propagation corrections tional to thelike displacement
αx3 , βx5 , etc. Newton’s law
of heat through a medium, the Einstein equa- for this system now F ≈takes −k x(t) the ,simple form (5) (5)
tions that describe how space-time curves in
where k is a where d x k 2that
constant
2 is acharacterizes constant that the characterizes Figure 1: A the plot of the
response to matter or energy, the Schrodinger + ω of x(Don’t
= 0 , spring. (6) let the minus sign
stiffness of the stiffness
spring.
2 the let the (Don't
minus
equation that describes
Newton’s Newton’s law theis lawquantum
not isthe not mechan-
onlythe example
only
Newton’s
Newton’s
example of law alawof isis not
sign a in indt
not the equation
the
this
sign
only example
only
in equation
this
example
equation bother ofofyou,
bother
aa sign sign
ityou, in this
in
simply it
thisre-
simply
equation bother
equation
re-
bother you, you, iti
Newton’s law is not the only example of
differential
differential a 1sign in
equation
equation this inin this
equation
physics.
physics. bother �
There
There you, bother
are
are it
the
the simplyyou,
minds
minds re- it us
us thatsimply
Oftenthat inthe reminds
physics
force one
is woul
always
ical state differential
of a microscopic
differential equation system,
equation ...in The
in physics. physics. list There
There are There
where the weisminds
are
a caveat
have
the us
defined
minds
here
that that
usω the =
that
youforce could
k/m
the
obtain
isin
force always
order higher
is pulling
to
always the the
pulling the
force is always
differential equation in physics. There are the
Navier-Stokes
Navier-Stokes order minds equations us
equations
equations us
that
by that
the
taking
that
that the
force
derivatives
describe
describe force
is always
or
theis by
the0.) always pulling
decoupling
dy- We
dy- mass
mass pulling
the total
back
back energy
the mass
towards xx =
towards of a
back system.
= 0.) 0.) We Th
We
could go on andNavier-Stokes
Navier-Stokes on and equations on! equations that describe that the
make
describe dy- the
the mass
equation
dy- back
mass towards
lookback simpler.towards x to = The x =general
0.) have
We writ-
have writ-
a set of equations. towards I’m referring
x=0.) only
= 0.)WeWe thehave most fun-
written of energy:
equation kinetic
(5) asand pote
INavier-Stokes
want to namicsdraw equations
of
namicsyour fluids, attentionthat describe
the Maxwell
ofMaxwell
fluids, to
the a Maxwell
common the
namics
namics
equations
dy- of ofmass
solution
damental
equations that fluids,
fluids, back
often the
the
formulation.
that
towards
equation
equation
Maxwell
Maxwell
ten (6)(5)
So
equation
x equations
I’m
equations
as (5) an as
is approximation
definitely
that
that
approximation
an
not
haveten ten writ-
trying to because
approximation
equation
equation (5) as
(5)
because
as an approximat
an approximat
namics of fluids, the equations that
describe ten
the equation
laws an
oflarger (5) as
approximation
electro-magnetism, an because
the dif- because
atat ergy
larger
larger is generically
displacements
displacements given
there will by bb
will
feature ofdescribe all these the
describe different
laws the equations.
oflaws
electro-magnetism,
of Indescribe
electro-magnetism, all claim thethe dif- laws
that
the at of
higher
dif-
electro-magnetism,
order
at equations are the
displacements
larger displacements there
never dif- will
useful
there
at larger
bein nonlinear
will be
displacements
nonlinear
there
describe the laws of electro-magnetism, the dif-physics, at larger displacements v
there itwill be nonlinear 3
3 55
fusion equation that =there xdescribes will be
the 3nonlinear
propagation corrections
corrections like
like αx αxx ,, βx βxx,, etc. etc. N
3 5
known theories
fusionfusion of physics,
equation that
equation thedescribes
underlying
that describesthefusion
dy-
propagation equation
theforpropagation
only
x(t)that that
corrections describes
where
0 cos(ωt)
corrections
they
like the
+ appear
αx propagation
0
5 , ,etc.
like βx 5 asetc.
is
αx3,equa-
sin(ωt) ,equations.
βx 5 , corrections
.a stand-in
Newton’s
(7) Newton’s
etc. law like
law takes 1 Ne 2
fusion equation that describes the propagation of heat
of heat through corrections
through
a more a
fundamental
a medium, like
medium,
etc.system αx set
Newton's
3 , βx
the of
the now ω
Einstein
lower-order
Einstein law equa- Newton’s
for this for
forsystemlaw
this
this system
system now now now
takes takesK =the
thethe2 simp mv
simp
namical equations of heatofthrough contain
heat a
through at
medium,most
a medium, two
the time
Einstein
the Einsteinequa- for
equa- this forsignthis system takes
now the
takes simple
the simple form
of heat through
derivatives whenthat aNewton’s
written medium, in describe
terms law
the isEinstein
not the tions
tions only
equa- that
that example
fordescribe
this
describe of simple
system ahow
how now intakes
space-time
space-time
form this the equation
curvessimple
curves bother
in
in form you, itform simply re-
tions tions describe
that howofspace-time
fundamen-
how space-time The
curves solution
in
curves in (7) describes 2
a function that os-
dd22xxthe 22
tal tions
degrees that of freedom. differential
describe 1how It’s equation
space-time
kind of incurves
surprising physics.
response
response in There toto matter are the
matter orwith
or minds
energy,
energy, us
d xthat
the
the 2the
Schrodinger
Schrodinger
d 2x just force is always 2 pulling + +ωω xx = = 00 ,,
response to matter or energy, the Schrodingercillates in time 2 frequency ω, the be-
response response
todifferentNavier-Stokes
matter or to matter
energy, or Schrodinger
equations
the energy, that the Schrodinger
describe the dy- mass d x back 22 + ω x+=ω02 x, = 0 ,
towards x = 0.)(6)We havedt
(6) dt 22
(6)
writ-
that so many equation equations
that describes that describe
the quantum equation
equation haviour
mechan- that
that describes
describes
we expect the
the a quantum
+
quantum
mass ωdt x on= dt0mechan-
the
2,
mechan- end of a (6)
equation that equation
namicsregimes
describes that
ofthe describes
fluids,quantum the the
Maxwell
mechan-quantum equations mechan- that ten dt2equation (5) as list � approximation
an because ��
nature in ical
very different
state of a microscopic all share
system, ical
ical
http://brownsharpie.courtneygibbons.org/
such ... state
statespring!
The ofof
list aa microscopic
microscopic
This function system,
system,
is ...
illustrated ... TheThe inlist Figure �
where
where we
we have
have defined
defined ω ω == k/m
k/m
ical state of a laws
microscopic system, ...onThe where
list we have defined � ω = ωk/m in order to
ical a describe
state of mathematicalmicroscopic the system, of electro-magnetism,
...In The could
could listgo gofor on
where and
and thewe on
on dif-where
and
and
have at
where on!
on!
definedwe wehave
larger have =massdefined
ωdisplacements
definedk/m in = there
order will
k/mtothe in
in order
beequation
nonlinear
order totolookmake
medium,
a fundamental could the go
could on Einstein
and
goand on
on on! and
and on equations
property. on! anddescribes fact,
on! I wantthat 1 describe the special
make the case where
equation look is
simpler. released
3 , βx5 , The
make
make the
general equation look simpler.
simpler. T
could
this how go
commonality on and
I wantIisto
fusionon equation
fundamental that that ittois I want the propagation
toto
make draw
draw the the
your
your make
equation equation
corrections
attention
attention thelook equation
to
to look
like
aa
simpler. simpler.
αx
common
common look The simpler.The
generaletc.
solution general
Newton’s
The of solution
general
equation law (6) of
is
space-time curves
sodraw your inattention
response to matter
from
a common restor so that ofv0equation
solution = 0. Exercise: (6) is (6) check
solution of equation (6) is
Ioverlooked.
oftenenergy,want to draw of want
heat
At your first
to draw
through
attention
glance
your to attention
a this
medium,a seems
common
feature
feature the to a all
Einstein common
ofsolution
of all equa-
these
these of equation for this
solution
different
different
equation ofsystem
(6)
equations.
(6) isisandnow
equation
equations. InIn all takes
all is the simple form
feature ofSchrodinger
thefeature all Ä these different
of all describe
these equation equations.
different that
equations.that In
describes (7)
all really
In of allphysics,does solve (6) also that x0 , vv0
likefeature of all these tions that
different equations. how space-time
In
known
known all theories
theories curves of in
physics, the
the underlying
underlying vdy-
dy- x(t) = xx00cos(ωt)
cos(ωt)+ + 0 sin(ωsin(
0 . x(t) =
a remarkable conspiracy; is there some- 0 v
the known
quantum theoriesmechanical of physics, state the underlying
of a v
microscopic
0 appear
dy- correctly x(t) to = be x interpreted
cos(ωt) v d 2+x as initial
sin(ωt) (7) (7)
known theories known
response
of theories
physics, to of physics,
matter
the or
underlying energy,thedy-
namical
namical
underlying
the Schrodinger
equations
equations
dy-
x(t) =
contain
contain x cos(ωt)
atatx(t) 0
most
most =+ 0 cos(ωt)
xtwo0
two sin(ωt)time
time+ω ω +2
. x = sin(ωt)
0(7) , . (7)(6) ωω
thing wrong namicalwith
system,
namicalhigher-order
equations ... The
equations equations
contain list at most
could
contain that
atgo twoon
most positiontime
andtwo and
on
time velocity. 0
ω dt2 ω
namical equations equation contain that atdescribes
most twothe quantum
time
derivatives
derivatives when
when mechan- The
written
written solution
interms
in terms of
of (7) describes
fundamen-
fundamen- aThe
The function
solution that (7) oscillates
describes aa functi funct
prevents them from
derivatives
and playing
on!
derivatives when a
when fundamental
written in
written terms in role
of
terms fundamen- of fundamen- The solution
The (7)
solution describes
(7) describes a function a �solution
that os-
function that
(7) describes
os-
derivatives
in physics? when ical state
written of
in a microscopic
terms of
1 It’s kindtalfundamen- system,
tal degrees
degrees of The ... The
solution
of freedom.
freedom. list
in time (7)
11
It’s
where describes
It’s kind
with
kindwe ofof
have a function
surprising
frequency
surprising
defined ωω,cillates that
,cillatesos-
justk/m
=just in time
the
in time
in behaviour
order with
with frequency
we
tofrequency ω,ω,
tal degrees tal of freedom.
degrees of
1 on freedom. 1 It’sof kind surprising
ofmany surprising cillates in
cillates time with
infrequency
time frequency
with frequency ω,the just be- the be- a mass on th
tal degrees of could
freedom.
I want go It’s
to and kind
draw on and
of
your on!
surprising
attention
that
that soso many cillates to a in
different
different time
expect with
equations
equations
make the that
that
equation ω,
describe
describe just
look thehaviour
simpler. be- we
The expect
general
To answer thatthis sothatquestion,
many so let’s
different
many consider
equations
different an that
equations ex- describe that describe haviour we aexpect
haviour
mass on
we expecta massthe end
a on of
mass theaonspring!
haviour end the
weofend This
a of function
expect
a
a mass on th
that so many common I
different want to draw
equations
feature your
that
of attention
describe
allnaturethese
nature to
indifferent
in a
haviour
very common
very different we
different expectregimes
regimes
solution
is illustrated a massall
allin
of on
share
share
equation
Figure the
such
such (6) end
1 for is of
spring!
spring!the a This function
This function isis illustrate
illustrate
plicit example. naturenatureSuppose
in veryinwe veryfocus
different on
regimes
different a mass all
regimes mshare all sharesuch spring!
such This function
spring! This functionis illustrated isinillustrated in special
Figure in
case where
Figure
nature in very
In feature
different
equations. of all
regimes these
In all
all different
share
known aa such equations.
fundamental
theories
fundamental spring! In
of Thisall
mathematical
mathematical function property.is
property. illustrated InIn fact,
fact, 1 Figure
1 for
for the
the special
special case
case where
where mass
mass
bouncing aaround fundamental on a spring.
a fundamental mathematical We denote
mathematical property. the property. In fact,In fact, 1 formass the isthe
special
1 fundamental
for
released case from
special where
case
restmass
where
sovis that
0 released
mass
vis0=0.
released
a fundamental all known
mathematical
physics, theories the of
property.physics,
underlying In
this
this the
fact, underlying
commonality
commonality
dynamical1 for the dy-
is special
is so case
fundamental where
x(t) = x mass
that
that cos(ωt)
itit is
is
is released
+from
from sin(ωt)
rest so . that (7)
v = 0. Exer
ω rest so that v00 = 0. Exer
displacement this of the commonality
commonality mass away
is so from thefundamental
fundamental equi- that itthatis it from sorest sorest that vthat 0
0 = v0. =Exercise: check
knownthis
this commonality namical isequations
soAs equations
fundamental is so contain that at isoverlooked.
it most from two resttimeis so from
that v = so 0. 0.that Exercise:
check check
librium position
often
theories by x(t).
overlooked.
often overlooked. long
At as
firstcontain
x(t)glance
At this is
firstin
often
often
notatthis
glance
overlooked.
mostseems
this
two that
seems
AtAt(7)that
first
first
really glance
glance
0
does
(7)solution
really
this
this
solve
does
Exercise:
seems
0
seems
(6)
solve
Exercise
and that
(6) also
and
(7)
(7)that really
really
x 0
alsocorrectly,
that
does solve
does
x0os-
solve (6) (6) and
, to be interpret
and aa
often overlooked. derivatives At first when glancewritten like
like terms
seems a a of that
remarkable
remarkable fundamen-
(7) really The
conspiracy;
conspiracy; does solveisis (6)
there
there (7)and describes
some-
some- also that
v v a xfunction
appear
appear , that
correctly to be interpret
too large,oflike then the
a like
physics, restoring
remarkable timeconspiracy;
derivatives
force is propor-1 isIt’s when
there written
some- v0 appear correctly to be interpreted 0 0 0
as initial
like to
tional a remarkable
thethe
taladegrees
displacement
underlying
remarkable
conspiracy; of freedom. conspiracy;
is there some-
thing
thing
is wrong
kind there
wrong of
vthat some-
0 surprising
appear
with
with v0cillates
correctly
higher-order
higher-order appear
Check tothatincorrectly
be time
equations
equations interpreted
(7) with tothat
that
really
be asinterpreted
frequency initialsolve
position
position
does ω,and and as velocity.
just initial
theand
velocity.
(6) be- also
thing wrong
thing with
wrong in with terms
higher-orderhigher-order ofequations fundamental
equations position
that and
position velocity.
and velocity.
thing wrong that
with so many
higher-order different
equations equations that
preventsthem that
position
them describe and
fromthat haviour
velocity.
playing we expect a mass on the end of a
dynamical preventsequations
F them
prevents
≈ −k from
them playing
degrees
x(t) , from playing
prevents
aoffundamental
freedom.
(5)
a fundamental
1role from
It's role
playing x0a,afundamental
vfundamental
0
appear correctly role
role to be interpreted as
prevents
contain at mostnature
them from
two playing in verya fundamental
different regimes role all share such spring! This function is illustrated in Figure
in physics?
in physics?
in physics? inawhen
physics? kind mathematical
of surprising that In so initial position and velocity.
intime kderivatives
physics? fundamental Toproperty. thisfact, 1 for
where
written
is To a constant
in answer
terms of
that
this characterizes
question,
many let’s
different
theTo
consider answer
answer
an
equations Figure
ex- this question,
question,
1: A plot ofthe
let’s
let’s the special
consider
consider solution an
an case ex-where mass is released
ex-
(7).
stiffness To answer
of this
the example.To
this answer
commonality
question,
spring. (Don’t this
let’s question,
is
consider
letwethe so minuslet’s
fundamental
an
plicit consider
ex- example. an
that ex-it is
Supposefrom we focus rest on
focus so aathat
on mass v0m m= 0. Exercise: check
fundamentalplicit degrees Suppose focus plicit
on a example.
mass m Suppose we mass
plicit example. plicit
often example.
Suppose that
overlooked.
we describe
Suppose
focus At
on awe
firstnature
massfocus
bouncingglance
m oninaaroundthisvery
mass m
seems on a that
spring. (7) really
We denote does solve
the (6) andlike alsoto that x0 , the
1
Thereof freedom.
isbouncing
a caveat around
here that on could
you a spring.obtain Webouncing
higher denote Often around
the on a Often
in physics spring.
one would inWephysics denote
like to one the
knowwould the know
bouncing bouncing
around like ona different
a around
remarkable
spring. on regimes
We a denote
spring.
conspiracy; all We
the
displacement share
is denote
there such the
some-
of the
theoftotal mass v appear
away from from correctly to be interpreted as initial
displacement displacement total of
energy mass 0 away
a system. energy There of the athe
are equi-
equi-
system. two kinds There are two kinds of
a of the of mass
order equations by taking derivatives or byaway decoupling from the equi-
displacement displacement
ofthing
the wrong
mass fundamental
away the
with mass
higher-order
from theaway equi-
librium from
mathematical equations the equi-
position that
by x(t). position
As long and as velocity.
x(t) is not
a set of equations.
librium I’m referring
position only
by x(t). to the most librium
fun- of position
energy: by x(t).
kinetic As
and
not is not energy: kinetic and potential. long as
potential. x(t) is not
Kinetic en- Kinetic energy is
property.In
librium
damental position
formulation. librium
fact, Sothis
prevents
by position
x(t).
I’m commonality
themAs long
definitely by As
from x(t).
not
long
asplaying
x(t)As
istoo
trying
aslong
so
too
x(t)asisx(t)
istoafundamental
fundamental
not
large, then therole restoring force is propor-
too large, then the restoring forceinis large,
ergy then
propor- the generically
is generically restoring given force bygiven is propor-
by
claim
too that
that ithigher
large, is often
thentoo
order
inthe large,
equations
overlooked.
physics? thenare
restoring thenever
At
force restoring
first useful
is glance
propor-
tional force this
to is
the propor-
seems displacement
physics, only tional
that whereto the theydisplacement
appear it is as a
tional
stand-in
to the displacement
tional thetional
like atoremarkable Totoanswer
displacement the displacement
conspiracy; this question, is there let’ssomething
consider an ex- 1
for a more fundamental set of lower-order equations. For our simple FF ≈
harmonic
≈ −k
−k Kx(t)
x(t) ,mv 2 .
= ,oscillator the poten- (5)
(5) (8)restoring force (5) with (8) our mod
wrong with higher-order plicit F
example. ≈ −k x(t)
Suppose , we focus on (5) a mass Form our 2
simple harmonic oscillator the poten- restoring force (5
F ≈ −k x(t) F,equations ≈ −k x(t) that , tial
(5) For prevents
energy our (5) harmonic oscillator the poten-
issimple law restoring
(12) we find force (5) with our
themwhere fromwhere bouncing
playing
k is ak constant around
a fundamental on a where
spring.
that characterizes wherein kphysics?
role Wek is a
denote
isthe constantthe For
tial our
energy
that
a constant that characterizes the simple
is
characterizes harmonic the oscillator
Figure 1:
Figure the
1: A potential
A plotlaw
plot of (12)
of the we
the solutiofind
solutio
where k is a constant is a constant
that ofcharacterizes that tial
characterizes
2 from
the energy the is Figure 1:
Figure A plot
1: A of plotthe solution
of the (7). law (12)
solution (7). d4 x d2 x we find
stiffness displacement
of the spring. the
(Don’t masslet stiffness
away
stiffness thelet of
minus of the theFigure
the spring.
equi-
spring. 1:= A1(Don’t
Uenergy
(Don’tis2 xof2let
plot let
.thethe
the
solution minus
minus (7).
To answer
stiffness of thelibrium this
stiffness question,
of
spring.position the let's
spring.
(Don’tbyletx(t). consider
(Don’t
the As minus an explicit
the minus mω 1 2 2
(9) g 4 + 2 + ω 2dx4 x=
long as
1 There is a caveatOften
1 x(t) is not
herethat
2
inyou
that 1 could
physics
you could U one = mω
would x
higherlike to . Often
Often
know in
inthe physics
(9)
physics dt one 4 xdt
done would
would like t
d2 xglike
example. 1
Suppose
There 1is a caveat we focus
here thaton youa couldmass There
m
obtain is a caveat here
bouncing
higher Often in 2obtain
obtain
2physics 2 higher
one would like to know the dtω4
1
There too here
is aequations
caveat There large,is a
that then
caveat the restoring
here
youderivatives
could that
obtain you
order could
higher force is by
obtain
equations Often propor-
higher
by in
taking
U
physics =
derivatives onemω or
x
would . like
by decoupling
decoupling to know total (9)
total the energy
energy of g
of aa system.
system. (9)
+ There
There + ar
ar
order by taking order
or by The equations
total
decoupling taking
total
energy derivatives
energy
is given 2of a
by or by
system.
the sum There of the are two
This kinds
equation is dt
not 4 much dt 2 hard
around
order equations on order
aby spring.
tional equations
taking to We
the
derivatives denote
bydisplacement
taking or by the
derivativesa
decoupling
a setdisplacement
set oforequations.
of bytotal decoupling
equations. energy
I’m
I’m The
total
of
referringa
total
referring
energy
system.
only
onlyenergy toto of
There
the
the
amost
is
system.
most
given are fun-
fun- two
by
There
ofof kinds
the
are two
energy:
sum
energy: of the
kinds
kinetic
kinetic and
This
and potential.
equation
potential. is
a set of equations. I’m referring only to the
a setaway of equations. I’m referring only
most fun-and
potential
toThethe of kinetic
energy:
The ofI’m total kinetic
energies
energy: energy andispotential.
kinetic given
and by
potential. Kinetic
the (6). sum The en-general
of
Kinetic theen- potential
solution looksh
aof set the mass
of damental
equations. I’m fromSoonly
referring
formulation. the
I’mto equilibrium
the most
definitely damental
damental
notfun- of most
position totalfun-
formulation.
energy:
formulation.
trying to energy So
So
potential I’m
kinetic isdefinitely
given
definitely
and kinetic
and by notthe
potential.
not trying
trying sum
energies to
Kinetic
to ofergy theen-
ergy is
is This equation
generically
generically given
(6).
given isThenotby
by much
general
damental damentalSo
formulation. formulation.
I’m F So ≈not −ktrying
I’m x(t)
definitely
claim , that
that not trying
higher ergy to
order (5)andis generically
kinetic
equations
ergy isgiven areenergies
generically given
never useful by
useful
givenininby
by x(t). claim As that long higher x(t)definitely
as higher
order is not
equations tooareclaim
large,
never to then higher
potential
useful ergy in theis order
and
generically equations
kinetic
1 appear
are
energies never
1ititisby (6).
x(t) The = general
A+ cos(ω solution Blo
claim that physics,
claim
higher only
where
thatequations
order
that k where
is a
order
they are
constant
equations
never
appear it physics,
useful
physics,
thatis as
areinnever
a only
only
E
stand-in
characterizes = useful
that
thatK +where
where in they
Uthe = they mv appear
2
+ mω is asas2 aa2stand-in
x1 stand-in
. (10)
1 11 +22t)=+ A
x(t)
+
restoring
physics, only force
physics,
that is only
where proportional
they that
appearwhere it isto
theyas the
appear
a for displacement
stand-in
for a a it is as
more
more a stand-in set
fundamental
fundamental set 2
E of
of = KFigure+
lower-order
lower-order 2
1 U
K = 1: Amv
equations.
equations. plot22 1 of the 2solution
+ . mω x 2
. (8) (7).
(10) KK = =cos(ω mv
(10)mv t).. + B+
for a more fundamental set of lower-order equations. 1 1 2K2 =2 mv 2
. (8) +
x(t) (8) A
= − A2+ 2 cos(ω − + t) + −
for a more fundamental forstiffness
a moreset of the spring.
fundamental
of lower-order set equations. (Don’t letequations.
of lower-order the
E =minus K + U = Kmv =2+ mv 2mω . x 2. 2 (10) + A−
To check that this really 2 2
isthat 2
thephysics right really expres- +the A− cos(ω t) +
Tocheck checkOften in one would likeright to
weknow
thisthis is right
the
where expression,
have defined the −freq
1
There is a caveat here that you could obtain higher To that really is 22the expres-
There is a caveatorder here that you could obtain higher sion,
orderTo notice 2 that E defined
total by
energy (10) is conserved.
of a right system. There are two kinds where we have d
2equations notice that EisE defined by (10) is conserved. That�is,
1
equations by taking derivatives or bycheckdecouplingthat 2 this
sion, notice really that the
defined by expres-
(10) is conserved.
by taking derivatives a setorofbyequations.
decouplingI’m a set of equations. That I’m is, if
refer- you take the
of defined derivative
energy: kinetic you’ll find where we have1 defined the f
referring only sion,to the most fun-
notice that
if you E take the by (10)
derivative is andconserved. potential.
you'llyou’ll findKinetic that en- �
ring only to the most fundamental formulation.
damental formulation. So I’m definitely So I’m definitely
thatThat not That is, if
not trying to ergy is generically given by you take the derivative ωfind
± = √ 1 ± 1 − 1
trying to claim that higher order equations is, if you take the derivative you’ll find 2g 1 ω± = � � √
claim that higher order are never useful
equations are neverin phys- usefulthat in dE ω± = √ 1±
ics, only that where they appear
physics, only that it is where
as a stand-in
they appear for a more that
it isfunda-
as a stand-in dt
=0 dE (11) (11)
2gfour arbi
mental set of lower-order equations. dE = 0 1 2 The solution (11) contains
for a more fundamental set of lower-order equations. = 0 in time. dt K = 2 mv(11) .
A± ,The rather The
B± solution
(8) solution con
than two becaus
implying that E is a constant dt Exer- contains four ta
19 implying that E is a constant in time.
order Exer-
in time Aderivatives.
± , B± rather T
cise: check that E really is conserved by taking A , B± rather than two
implying that cise: E is a constant
2 check that E really in time. Exer-
is conserved could by±be takingrelated orderto the timebec
in initial cd
 Fortotal
our simple energy harmonic =x
oscillator 0 . the thing poten- wrong (11)
(9)restoring
with force g(5)contains
with our +modified Newton’s
The is U given = by mω
2dt the sum of the
This Thehigher-order
equation solutionis
1 not dt
+
4much
equations
2 dt
2fundamental
harder ω that
2 four arbitrary
x
to=solve 0position
. than constants and velocity.
(13)
d4 x d2 x
tial
potential energy
1
� is �
and kinetic energies prevents them law
(6). The (12)
from U we
playing
=
general find mω a
solutionx . looks role
like (9) g + 2 + ω2x = 0 .
ω± = The √implying 1 ± that1 − E 4ω is 2ag constant
. (15) in time. Exer- A± , B± 2rather than two because (12) is fourth dt4 dt
2g total energy is given by theinsum physics? of the This order equation
ind4time is not much harder
derivatives.
2x These to solve than
constants
cise: check that 1 E1 really 1 is conserved by takingx(t) = isgA x cos(ω d
2
xmω. 2 x2 . in To
2 answer
(9)total this (6).question,
The + let’s + t)
consider ++to B xan sin(ω
ex- + t) This
potential
Eimplying + Uand =Ukinetic
that E is +aenergies
constant time. begeneral
don't solution
need looks
go like
through that equation is not much harder to
The energy given 2+
= Kderivative
the
=mv 2mω
(10) (6). could related+ by to the ωsum = of0 .the
2 the initial conditions x(0),
(13)
olution contains 2 of 2(10) constants
four arbitrary 2and using equation plicit example.
potential andv(0), Suppose +etc.
kinetic dt
we
4
Aeffort focus
− cos(ω
energies dt on
t)make
to−don’t +Ba mass sin(ω
− our m − t) (6).The
,(14) The general solution looks like
Now, let’s see what would 1 happen =But Awe need to point.
go through t) that
± rather than two because 1 Exercise
(12) is
2 fourth 2 if we around
2bouncing de-
This x(t)
on
equation a spring.
is not +We cos(ω
much denote
+ t) + the
harder B + sin(ω
to solve +than
ToThecheck total
cided
Ethat =energy
K this
to modify
is=given
+ Ureally mvthe by+ the
2 isconstants
Newton’s
mωsum x
2rightbydisplacement
laws expres- of
. the(10) effort to solution
adding some 1+ make our 1from
(14) The
point. is plotted solution (14) in is
in time derivatives.
potential and kinetic These energies where
(6).of we
the
The have
mass
general defined2A
away solution the 2frequencies
cos(ω the2t) +
looks B
equi-
like sin(ω t) x(t)
,(14) = A+ cos(ω+ t) + B+ sin(
sion, notice
higher that E defined
derivative terms. by (10) This is conserved.
might sound E =
likeK + U = Figure
mv + − 2mω with
plotted in Figure 2 with gω = 0.01 and aarep- x− . 2 (10)
− and
−
be related to the initial conditions isisx(0), librium position bythe 2
x(t).representative
Aschoice 2
long asof
That To is, Check
check
a funnyif you thing that
that
take this E
the
to
really
doreally
derivative
(after
conserved
the
all,you’llright
we find
know
by
expres- taking
that resentative 1 defined
� �x(t) initial is not
choice of initial
conditions. Com-+ A− cos(ω− t) + B− sin(
tc. But derivative
we don’t need to 1 go through 1 that too large, then where
x(t) ω
the we
=
=
restoring
√ have
A cos(ω
1
force± t)
is 1the
+ B 4ω
propor-
− frequencies
2sin(ω
g . t)(15)
that sion,E =notice K +U of=(10)
that Emv and
defined2
+ using by (10)
mω 2
xequation
2 is
. conserved.
To (6). that
check
(10) pare ±
thisthis really +
plotistothe
conditions. +
Figure +
right 1.expres- +
verify
�t)Compare thiswhere plot
Newton’s law works perfectly 2g Exercise:
to makeThat our point.
is, if The
you dE 2solution
take the (14) isfine!)
2derivative tional
you’ll
but
sion, to it’s
the
find
notice
adisplacement
that
thatE(14) +defined A cos(ω
by 1 equation
(10) + B sin(ω t) ,(14) we have defined the frequenci
worthwhile exercise.
2 = 0 Often, we can learn
(11) a lot ω tosolves
=
−
Figure
√
− 1.1is±conserved.
� (13). 2
−
1 − 4ω
−
g . (15)
d in Figure 2 with gω dt = 0.01 and a rep- Theif solution contains
± four arbitrary constants
To that
check that this really is nature
the right That
expres- is, you F ≈ take the
x(t)derivative
, 2g you’ll (5)find 11: A plot of�
�
−k Figure
ative about
choiceNow, why
of initial let's the see laws
what
conditions. of
dE would Com- take
happen the if Aform
we where
± , decided
B± ratherwe have thandefined two becausethe frequencies (12) is fourth ω± = √ 1 ± the 1 − 4ω 2 g
implying
sion,theynoticethatdo E is
that
by Eadefined
imaginingconstant = in
by
how 0(10) time.
things Exer-
is conserved.
would thatbe
where ksome (11)
isdif-
a constant that For
characterizes Exercise
our simple
theconstants harmonic oscillator solution
2gthe (7).
poten- restor
plotto
hiscise: to modify
Figure Newton's
1. Exercise: laws
dtconserved verifybyby adding order higher
The solution
in time dE contains
derivatives. four These arbitrary Figure that
constants
noting 1: Athe plot of the
total energy solution (10) (7
That check is, that
if you E really
take the is derivative taking
you’ll find 1 law wi
�
ferent if they took some other form.
stiffness So, let’s
of the spring. (Don’t tial
= 0 let the minus(11) � energy is2 (1
14)the solves equation
derivative
implying that (13).
terms.
E is aThisconstant might in sound
time. like
Exer- could a Afunny
be± , related
Bω± ± rather
=√
dt to than
the ±two 1because
1initial − 4ω g(12)
conditions . x(0), is
The be fourth
(15) solution contains
conserved four arbitrary
derivative
that of (10) and using
boldly consider replacing (4) by the1 following order in timeequation (6). Verify
2g that (14) solvesOften equation in (13).one would like to kni
physics
if the motion of x(t)
thing
cise: to
check do that(afterEdE all,
really we
is know
conserved that by
There Newton's
taking
implying v(0),
is a etc.
caveat
that law
E But
here
is we
that
a don’t
you
constant derivatives.
couldneedin obtainto
time. gohigher These
through constants
A
thatby ,
1 B
equationrather thanrather two because (12
Exer- .a (13) than (6).
Now,equationlet’s see what would ± ±
= 0 happen if order we de- (11) could be related to the initial U order
total
conditions = energy mω
x(0),
2 2
xof system. There(9) are tw
cidedthe derivative
to modify dt
of (10)laws
Newton’s works
and by using perfectly
adding equation equations
fine!)
cise:(6).
some effort
The
butby to taking
it'smake
solutiona derivatives
our
containspoint. or by
four The decoupling
solution
arbitrary
check that E really is conserved by taking a 2new expression for the kinetic e (14)
constants is in time derivatives. These
a set of equations.
A±v(0), I’m referring
etc. But only
we to
don’t the most
2 need fun-
to(12)go ofisa energy:
andthrough kinetic and potential. Kine
implying Now,that let’s see
xis awhat
dE4terms. x worthwhile
2This
dconstant would 1 inhappen time. exercise.
if
the we plotted
de-
derivative , Bin
Often, ±we
of Figure
rather
(10) I’mand
2To
than with
usingtwo
see gωequation= 0.01
because
what's wrong
(6). could
rep- withthat
fourth
derivation beequation
related
is pretty to(13),
the initial
we This
complicated condi ae
� �
higher derivative might sounddx like
Exer-
damental formulation. So definitely The not trying energy
total to is given by the sum of the
cided to g
modify + Newton’s = lawsF x,
by adding , some(12) effort
resentative
order in to make
choice
time of our
initial
derivatives. point. The
conditions.
These ergy
solution v(0),
Com-
constants is generically
(14) etc. is But we given
don’t by need to go th
funnycheck
a cise: thingthat todtdo E really
4 dt iscan
(after 2 mlearn
conserved
all, we know adt
by lot
claim taking
that about
Now,
that higherwhyorder
let’s seethe what need
equations would tohappen
are compute
never
potential
if2 we
useful
andthe
in de-energy.
kinetic
a lot of It theoretical
energies turns outmachinery in our (6). beyoT
thehigher
Newton’s derivativelaw derivative
worksof (10) terms.
andlaws
perfectly This
usingfine!) might
ofequation
nature
but sound
physics,cided
it’s aonly
take
(6). pare
like
to theplotted
could
that
modifythis
where
form plot
beNewton’sin to
related
they Figure
Figure
appear
higher tolaws it2 iswith
the by 1.
initial
derivativeasadding gω
aExercise:
stand-in =some
theory0.01
conditions and
effort
verify x(0),
equation a rep-
school tolevel.
make Let
(9) still ourme point.
1 simply
gives the The tell solut
you
a where
funny g is
thing a newto constant
do (after of
all, nature
we forknowahigher
morethat
having di- resentative
fundamental
thatv(0),(14) etc. set
solves
But of we choice
lower-order
equation
don’t of needinitial
equations.
(13). to conditions.
go through plotted
thatCom- in Figure K =2 with mv 2gω . 2 = 0.01
worthwhile
Now, let’s exercise.
see what Often, would we happen
they candolearn byif aweimagining
lotde- derivative terms.
how potential energy, This might sound like
however, 1the simple2 1 expression
2 2 2 (8) � x(
0 pare E 1. = solution
KExercise:
+ U =(14) mv is +valid. mω x of . initial (10) conditio
about
cided mensions
Newton’s
whyto modifythe law of
laws seconds
works
Newton’sof nature squared.
perfectly
lawstake
things by fine!) Ifthe
adding
would we form atake
butsome
be it’s
funny =aeffort
gdifferent
thing tothis
to ifmake
do for plotour
(after toall,
thepoint.
Figure
kinetic weThe know
energy that is no 2
verify
resentative
longer 2 choice
The dx d3 x
easiest mg d2
that (14) solves equation 2 (13). pare rep-this K plot = to mgFigure3 1. − Exerc
they do
higher then
worthwhile
by obviously
imagining
derivative exercise. this
how reproduces
terms. Often,
things
This
theymight we can
would
took the be
sound
some usual
learn
Newton’s
dif-
like
other lotplotted
aforce law
form.
in
works
So,
Figure
Figure
notingperfectly
way
2Awith
2:thatto plotseefine!)
the
gωof
that
total
= 0.01
but
the is it’s
solution
energyby 2and
anoting
(10)
a(14).
willthat no the
longer total 1850 dta Russian
dtenergy 2 dm
law
about (4).why For the g sufficiently
laws of naturesmall we
take expect
the
worthwhileformthatresentative
exercise. choice
Often, ofwe no To
initial
can checklearn that
conditions. a be this that
lot conservedCom-really(14) is the
solves right
equation expres- (13).
2
ferent
a funny if they thing took to somedo (after other form.
all, we know So,consider
let’s
that replacing waswhere
� �
let's boldly be (10)conserved will if the
sion, longermotion thatofE x(t) byif(10)
is described theismotion mgradski dx of able
boldly equation
they
Newton’s dolaw
consider by(12) willperfectly
imagining
replacing
works give how
(4) basically
by things
the
fine!) theabout
would
following
but samebeadif-
it’s pre-
whyparethe this
Tolaws plotof
see tonature
what’s Figurewrong take1.notice Exercise:
with
the equation
form defined
verify
(13), we + conserved. .
Figure 2: A plot (4) by the following equation by equation
x(t) is (13)
described rather by than
equation (6). So
(13) we need
rather eral
than
2 theorem
dt find [1]. O
(6).
equation dictions
ferent
worthwhile as of
if exercise.
they (4). the So (12) seems like a decent
tookOften, some we other can form.
learn So,
theya lot that
let’sby
do (14) solves
need
imagining to compute
how equation
things That (13).
thewould is, ifbeyou
energy. dif- take
turnsthe
Itkinetic outderivative
in The you’ll
solution
gure 2: boldly
A (14).
plot consider
of the solution (14). a new expression
So derivative
we need that atheoryfor the
new expression energy.
forstillthe kinetic energy. laws
fundamental
about starting
why the point laws for ofour
replacing first
nature (4) attempt
by
take the
the at modify-
following
ferent
form if they our tookhigher some other form. So, equation
let’s (9)
Using this new expression for kin
4 derivation
The is pretty iscomplicated
derivation pretty complicated dEand involves andder differential
involves eq
d2 xof nature.
� �
they dog dthe
ing
equation byximagining
laws
+ =
1 things
how F x,
dx would be
, boldly
(12) dif- consider gives the
replacing
a lot potential
of (4) by
theoretical energy,
the however,
following
machinery beyondthe
the
dt
=sim-
total 0 energy
the high E =
kineticK (11)
+U will
energy The be
if s
see what’s wrong 4 with 2 equation (13), we (12) a lot (8) of for theoretical kineticmachinery beyond (13) the to high
ferent To if dt make
they dtlife
took even
some m other simpler, dt�form. let’s�So, focus
equation
let’s again ple expression
school level. Let the
me simply energy
tell you
you is no
use
the equation
answer: rem describe
precludes th
the
to compute the
our energy. dIt2 xturns 1 out indx Using the longer valid. school level. Letway me tosimply Exer- A± , B
of tell is you justthe as itanswer:
on consider d4 xreplacing
mass-on-a-spring example. The implying
easiest that Eseeisthat a constant by in time.
boldly
where g is + a new = (4) F by
constant x,the following
of , nature (12) having x(t), should
tive be. order
equations pla
gher derivative
where g is a new theory 4 equation
dtconstant dt2 of m (9) still
nature having
dt di- 4
d x d x 2 1
�
dx check
cise:
�
that E really � 2 is�conserved
2 by taking
equation
dimensions ofhowever,
seconds squared. g + = F x, dx, d 3 x (12) mg Aside
d x from being very
physics. complica
Moreove
themensions
potential ofenergy,
seconds squared.the If we sim- takeIfgwe = take
0 =04 then 2
dt dt m K =the dtmg derivative−of (10) and using equation (6).
something deeplycessful (16) abou
troubling
could
obviously this reproduces the usual force 3 (4).
law dt3 see 2what
dt let’s 2
dtwould law of phye
v(0),
pression where
(8)
then obviously for g is
the a new
kinetic
d xthisdreproduces
4 2 x constant
energy
1
� of is nature
no
thedxusual force
� having di- Figure 2: A plot of theNow, solution (14). happen if we de-
g + = F x, , (12)
� �2 (16): the kinetic to energydate can
has be
invol
effort ne
valid.
law (4).For
mensions
The easiest
For gsufficiently
dt
ofway seconds
4sufficientlydt2
to see small
squared.
thatwe
small
m
we
is If expect
by
expect
dt
we take where
that that
g =g 0isequation
a new constant of nature cidedm having
to dx di- Newton’s laws by adding some
modify
+ . gardless of the (16)sign of g we can
derivatives. alwa
equationthen(12)
(12) obviously
will givegive
will thisbasically
basically reproduces the the same the usual
same pre- forceTo
predictions
mensions ofassee (4).
seconds
what’s
Figure squared.
wrong
2: A plot If we
higher
with take
2 derivative
of equation
the dtg =(13),
solution 0 terms.
(14). we This might sound like plotte
some function x(t) that makes
Ostrogradksi’s this
resent
dictions
wherelaw g(12)
So (4).
as aFor
is(4). seems
newSo g constant
sufficiently
like seems
(12) a of small
decent
nature we
like starting expect
ahaving thendi-
decent that
obviously
pointneedfor to our this reproduces
compute the energy. athe funny usual
Itthing forcetoout
turns doin(afterFigure all,
2: A weplot know of thatsolution
the
Using this new expression less
for than
kinetic zero. This
energy very deep. is unlike any s
startingequation
first
mensions point
attempt (12)
for our
of seconds will
atfirst giveattempt
modifying
squared. basically If we atthe
the same
modify-
laws
take law
g= 0pre-
of(4). ourFor
nature. gTo
higher seeUsing
sufficiently what’s
derivative this
smallwrongnew
theory we
Newton’s expression
with
expect
equation equation
law that works
(9) forstill kinetic
(13), we energy
perfectly fine!) also but it’s a pare t
the total energy = E=K+U we
will arebeinused toifin physics.
conserved erwisewith At first
have looke g(
ingthen
thedictions
laws ofasnature.
obviously (4). So (12) seems
this reproduces the usual like equation
aforce
decent gives need
(12) the will
Figure tothe
potential
2:
total
compute
give Abasically
plot
energy
energy, the
of the
worthwhile
the
E same
energy.
however, K +U
solution Itthe
pre- will
turns
exercise.
(14). sim- be out
To conserved
Often,see what’s
we can wrong
learn a lot that
equatio
To make life even simpler, let's focus again on youif you use equation (13) at all
tothe obvious how to make sense of
To starting
n- restoring
law (4). force
make Forlife point
g(5) for
with
sufficiently
even our
simpler,ourfirst modified
small let’s attempt
we focus at
Newton’s
expect again modify-
dictions our
that pleasexpression(4). higher So (8)use
(12) forequation
derivative
seemsthe about theory
like
kinetic (13)
why to describe
equation
a decent the
energy isneed
laws (9)
no ofdescribe
to dynamics
still compute
nature takethetheand itform
theenergy. also serves It tu
our mass-on-a-spring example. Using the restoring netic energies. In physics, negative
oten- law
on
restoring
restoring
ing
(12)
equation
our the
we(12) force
laws
findwith
mass-on-a-spring
force (5)
of (5)
will give our
with
nature. our
basically modified
example.Newton’s
modified the Using Newton’s
samethe starting
pre- longer gives
point
Tovalid.for theof
our
see what’s dynamics
x(t),
potential
first
The easiest just
attempt
wrong way as
theyof it
energy,
withto x(t),
should
at
do just
however,
modify-
by be.
imagining
see that(13),
equation as it
the
our bywe are usually interpreted asnas
is ergies should
sim-
higher
how be.
thingsderivative
would ify the
theory
be laws
dif- of
equatio
force
lawTo (12) (5)
make we with
find our
lifeSoeven modified
(12)simpler,
Newton's
let’s afocus ing the law
again laws(12)
ple we
oftoexpression
nature. Aside (8) from for
ferentbeing
theifbeing very
kinetic
they complicated,
energy
took givessome isthe no
otherthere’s
potential isenergy,
notlet’s widelyhowever app
dictions
law (12) we find as (4). seems like decent need compute Aside the energy.
from It turns very out in
complicated,
instability. 2 form. Physically
So,
find d 4x d 2x something deeply troubling about equation Often, this The
physicists. is un
) on our
starting g mass-on-a-spring
point for
+4 our + first
2 ω2x example.
attempt
= 0 . at modify- Using To3make
(13) theourlonger
life
higher even valid.simpler,
derivative
there's The easiest let’s
boldly
theory
something way
focus consider
equation todeeply
again seereplacing
that
(9)plestill is by
expression
troubling (4) by (8) thewe for the kinetic en
following
can
4 d2 xdt2 d x wenegative!
know that a mass
theorem bobbing is so aro
(9)ing the laws 4
d x of dt g dnature.
x + 2 2 + ω x = 0 . on our 2
(13)mass-on-a-spring
gives the potential (16):
about
the kinetic
example. equation
energy,
equation
energy
Using
however, (16):
can
thethe belonger
the sim- Re-
valid. The easiest
kinetic learn a lot way to fun
see
g 4 + 2 +ω x=0 . dt 4 dt (13) (13)gardless of the sign of g energywe can is end
always of a spring
imagineabout is a completely
forwhygranted
the stab
that th
e ThisTo make
equation dt life is dt even
not much simpler, harder let’s focus than
to solve again ple 3 expression (8) for
energy can thebe kinetic
negative! Regardless no important
It’s � laws�of to be
nature clear
take about e
f the onThis
our equation
mass-on-a-spring is not much harder
example. to
Usingsolvethe than longerthan some
valid. of function
Thethe easiest x(t)
way that
to
4
dsee makes
x that 2 this
d xis by1 expression dx ond order in deriv
(6). The
This general
equation solution is not looks much like harder to solve sign This of is we g + 3 = F x, the form , they do
(12) by
This equation
(6). Theis general not much harderlooks
solution to solve likethan less than zero. 4 can
dtunlike dtany 2thealwaysproblem
system
m that with equation
dt to how why.(13) is
ask things
imagining
(6). The (6).
generalThe general
solution solution
looks like looks like imagine some function
we are used to in physics. At first glace it’s not x(t) perfectlythat well-defined
would beYou mathematica
might find
different
x(t) = A + cos(ω+ t) + B+ sin(ω+ t) 3
) x(t) = A+ cos(ω+ t) + B+ sin(ω+ t) (14) makes this
where expression
g
at all obvious how to make sense of negative ki-is a new less However,
constant than of it isn’t
nature if they an took
having
first acceptable
somedi-
attempt tolawm
(10)x(t) = +A+A − cos(ω
cos(ω + t)−+ t)B++B − sin(ω
sin(ω + t)− t) ,(14) because it predicts othersomething
form. - na
+ A− cos(ω− t) + B− sin(ω− t) ,(14) zero. This mensions is unlike
of seconds any
netic energies. In physics, negative kinetic en- unsuccessful. But system
squared. If we take g = 0
s- + A − cos(ω − t) + B − sin(ω − t) ,(14) ergiesthatareweusually then
are obviously
used
interpreted to in tive
thisphysics.
as kinetic
reproduces
signaling Atenergy
the
an usual - that
new forceis not comp
theories of phF
pres- where we have
where we defined
have the frequencies
defined the frequencies
d. where we have defined the frequencies 2
first glace Physically
instability. law (4). For
it's not at this g sufficiently
all obvious the world
is unacceptable: small we
how to make see
we around
expect
patible us.
that
sense with natua
So we
rved.
d where we have defined 1 the�
� frequencies
discard our first attempt to modif
�
weofknow negative that equation mass(12)
a kinetic will give
energies.
bobbing around basically
In physics,
on the the perform
same
negative pre-any To
new
find ω± = √� 1 1 ± 1 � − 4ω 2 g . (15) (15) law and regard equation (13) as
ω1± = √ 2g � 1 ± 21 − 4ω g . 2 (15) end of a spring dictions is a as (4).
completely
kinetic energies are usually interpreted as signaling So (12)
stable seems
system. like a
are decent
still usefulneed bea
) ω ± = √ 1 2g
± 1 − 4ω g . (15) starting point for our But firstwhat about
attempt more
at
what ture attempts. modify- complicated our Ith
The The
(11) solution solution 2g
contains containsfour arbitrary four arbitrary constants constants A±, It’s an important
instability. to be
2 clear about
Physically exactly
this is unacceptable: we
What
ing the laws of nature. This is a hole to Ostrograd if we added even more deriv
gives
- Asolution
±, B B±
The solution
ratherthan
± rather
contains
than two two four
because
because arbitrary
(12) (12) is fourth order inthe problem with equation (13) is.
constants
is fourth
The Atime , Btime contains
rather four
than These arbitrary
two because constants
(12) iscould fourthbe relatedperfectly well-defined mathematical To make life even simpler,
2 let’s
didn’t really to
Weequation. focus
seeadd again
this higher pleder
instability exm
gA±order
Exer- ±in
, B±order
rather
±derivatives.
than derivatives.
two because
constants
These
(12) is constants
fourth
in initial
time
2
However,We didn’t it on really
isn’t our anmass-on-a-spring
see this instabilityin
acceptable manifest
lawour of example.
toy itself
model
physicsin our Using
toy
because model
without wethe didn’t longer
running inclui
aking
ordercouldintobetimethe
related to derivatives.
derivatives. conditions
the initial
These
These v(0),
x(0),
conditionsconstants
constants
x(0), etc. But we because we didn’t include nonlinearterms terms like like αx x33 ininequation
equation (5).(5). it requ
could be related to needthe initial conditions x(0), because it predicts something - namely nega- However,
(6).v(0),
e-could be etc.
related But wethe
to don’t initial to go through
conditions x(0),thatthat
v(0), etc. But we don’t need to go through tive kinetic energy - that is not compatible with mathematical 3 obj
eev(0),
de-
effort
etc. to
But make we our point.
don’t need to Thego solution that
through (14) is 20
effort the world we see around us. So we are forced to with infinitely m
some plotted in to Figuremake2 our withpoint. gω 2 =The 0.01solution and a rep- (14) is
 know that a mass bobbing around on experiments. Failed theories are still useful
the end of a spring is a completely because they help to guide future attempts.
stable system. It turns out that there is a loophole to
Failed
It's important to be clear Ostrogradski's theorem that allows us theories
In
physics, about exactly what the problem to add higher derivatives to the laws are still
negative of nature without running into trouble useful because
kinetic
with equation (13) is. This they help to
is a perfectly well-defined with instabilities. However, it requires guide future
energies are
usually interpreted mathematical equation. recourse to some exotic mathematical attempts.
as signaling an However, it isn't an objects (for example equations with
instability. Physically this is infinitely many derivatives) and a complete
unacceptable: we know that acceptable law of physics
a mass bobbing around because it predicts discussion is beyond the scope of this article.
on the end of a spring something - namely negative
is a completely stable References
system. kinetic energy - that is not
compatible with the world we
[1] M. Ostrogradksi, Mem. Ac. St.
see around us. So we are forced
Petersbourg VI 4 385 (1850).
to discard our first attempt to modify Newton's
law and regard equation (13) as a dead end. But
what about more complicated equations? What if
we added even more derivatives? In 1850 a Ostrogradski
Russian mathematician named Ostrogradski showed that
was able to prove an amazingly general any fundamental
laws of physics
theorem [1]. Ostrogradski showed that any involving N-th order
fundamental laws of physics involving differential equations
nth order differential equations will This will lead to negative
powerful kinetic energy if
lead to negative kinetic energy if N > 2.
theorem
n > 2. This powerful theorem precludes the
precludes the possibility of higher possibility of higher
derivative equations playing any derivative equations
playing any fundamental
fundamental role in physics.
role in physics. Moreover, it
Moreover, it explains why every explains why every successful
successful law of physics that law of physics that has been
has been discovered to date has discovered to date has
involved no more than
involved no more than two time two time derivatives
derivatives.
Ostrogradksi's theorem is very
general and also very deep. It explains what might
otherwise have looked like a strange coincidence
and it also serves as a guide for trying to modify
the laws of nature. However, this theorem is not
widely appreciated by most practicing physicists.
The reason, I think, is that this theorem is so
fundamental. Most of us take it for granted that the
laws of nature must be second order in derivatives
without ever stopping to ask why.
You might find it a little depressing that our
first attempt to modify the laws of physics was
unsuccessful. But don't despair! Indeed, many new
theories of physics are found to be incompatible with
nature even without the need to perform any new
http://brownsharpie.courtneygibbons.org/

21
 Domenico di Michelino, La
Divina Commedia di Dante
(Dante and the Divine Comedy).
Fresco in the nave of the Duomo
of Florence, Italy - the terrace
purgatorio lies over his right
shoulder

Since his death in 1968, George likely to be on any floor.

By Gamow has been living in purgatory • It goes at a constant speed when moving, and
Gordon in a seven floor terraced apartment
Hamilton when it has no requests, it just waits where it is.
building with penthouse garden
Masters Academy
on top. Scummy tenants reside on • When the elevator is going past a floor, souls
& College
the lowest floor, but the residents like George can jump on or off without altering
become increasingly sophisticated as the final destination of the elevator.
one moves up. George's apartment is on In his 1958 book, Puzzle Math, George showed
the third floor. the impulsive answer of ½ is wrong. He argued
Each day George wakes up, dresses, eats, In that the elevator's last stop before visiting his
locks-up, walks down the corridor to the
Julian floor was equally likely to be 1 or 2 (below),
Havil’s
elevator and impetuously pushes both the or 4, 5, 6 or 7 (above) so
“Impossible?”
up and down buttons. When the elevator (reviewed in the elevator is twice as Paradiso
arrives – he compulsively steps in –
Volume 12) George’s likely to have descended
original calculations
before checking whether it's ascending from a sweet smelling
are extended to include
from the nicotine miasma of the lower multiple elevators. Although floor than to have 7
he does not state it, Julian’s ascended from the
floors or descending from the perfumed
fragrances of the upper floors.
multiple elevators are
stench below. That 6
assumed stationary
when George pushes was great news. It 5
If it's ascending from the lower floors, the buttons. meant that George
George gets a nasty elevator ride, and is off could expect to have a 4
to a bad start in the day. bad elevator ride only 1/3rd of 3 George
Gamow’s
What goes up must come down - so the first the time.
impulse is to think that George is going to have Sure enough, when George 2
a bad start to the day about ½ the time. arrived in purgatory in 1968,
1
Is that right or wrong? Before we try to before the population explosion,
solve the problem, we need to understand he confirmed that his day-to-
exactly how the elevator operates: day elevator rides were a nasty
experience only about 1/3rd of the Inferno
“...the • The elevator goes time.
elevator up and down picking up But over the last few years,
almost always and dropping off souls George has realized that something is
appeared as it was on different floors. The wonderfully wrong. He is now getting
going down, and so
the journey [up to see elevator's destination bad elevator experiences a lot less
his friend] would be is always randomly frequently than 1 in 3 and his luck
a frustrating ‘down Julian Havil determined - is especially good during morning
and then up’...” in “Impossible?” so when the rush-hour. Find an explanation for
(reviewed in Pi in
elevator is at George's statistical joy. (Spoiler alert:
the Sky - #12)
rest, it is equally answer on the next page.) George Gamov
1904 – 1968
22
 George finally figures out that before, when he pushed an average lifespan), the transcen-
the buttons, the elevator was at rest. The probability dental administration decided last
was 1/3rd of a bad elevator experience. Now, dur- month to scrap the1outdated Paradiso
ter-
ingThe only thing
rush-hour, the that is different
elevators now when
are moving compared to
George raced landscaping
the transcendental administration that Dante vis-
1968 is that there are more souls after the population
pushes the buttons. Does that make a difference? The decided last month ited,toandscrap
replace
theit with a modern
explosion. The with
elevator travels onlya thing
constantdifferent
speed, between
so all we rush-
need outdated terraced landscaping that that has each
apartment complex
hour and non-rush-hour is that there are more
to do is to find all possible origin-destination pairs and Dante visited, andsoul allocated
replace to their own unique
it with
souls using
find out howthe
much elevator
of eachatis rush-hour. Using
below George’s these
floor. floor.
a modern apartment complex that
two clues, George finally figures out that before, Earlier
has each soul allocated this week,
to their own George learned
George
that he will be p
allocated floor p
Gamow’s
when he pushed the buttons, the elevator was at unique floor.
rest. The probability was 1/3rd of a bad elevator where 0≤p≤1. He quickly figured
Earlier this week,
outGeorge
that hislearned
probability of having a
experience. Now, during rush-hour, the elevators that he will be allocated
smelly elevator inpthe morning is:
floor
are moving when George pushes the buttons. Does where 0 ≤ p ≤ 1. He quickly figured
that make a difference? The elevator travels with out that his probability of having a 0 Inferno

•p if he presses the buttons when the elevator is

a constant speed, so all we need to do is to find all smelly elevator in the morning is:
possible origin-destination pairs and find out how
much of each is below George's floor. • pstationary.
if he presses the buttons when the elevator
is stationary.
The total length beneath George •
•
p22(3-2p) if he presses the buttons when the
p (3-2p) if he presses the buttons when the
Gamow's is 16 floor heights out of a total of elevator is moving. George used integration to
21 + 15 + 10 + 6 + 3 + 1 = 56 floor heights. That elevator is moving. George used integration to find
find this, but he might also have used the limit
is, the probability that a moving elevator this,ofbut he might also have used the limit of the
the ratio above:
Paradiso is below George when he presses the ratio above:
buttons is 2/7 which is less than 1/3. probabilityProbability
(elevator below)(elevator below)
That is why, as more and more souls (pF - 1) (pF) (pF+1) + (F - pF) (pF - 1) (pF)
7
moved in and started increasing = lim
F→∞ 6 2
elevator usage – George got
The total length beneath George Gamow’s is 16 floor (F - 1) (F) (F + 1)
heights out of a total of luckier and luckier
21+15+10+6+3+1 = 56 asfloor
his
probability
heights. That is, the probability of elevator
that a moving getting 6
is below George when he presses a smelly
the buttons elevator
is 2/7 (pF-1) (pF) (pF+1) + 3 (F-pF) (pF-1) (pF)
3 whichGeorge
dropped from
is less than 1/3. That is why, as more and more
Gamow’s 1/3 = lim
to 2/7.
souls moved in and started increasing elevator usage
F→∞
(F - 1) (F) (F + 1)
– George got luckier and luckier as In his probability
general, of if
1 getting a smelly elevator droppedGeorge
from 1/3istoon 2/7.floor
= (pF) 3
+ 3(F - pF) (pF)2
In general, if George is on floorff inina building with F
a building (F)3
floors: The total length of possiblewithfloor-to-floor
F com-
floors:
Inferno
mutes is a tetrahedral number: (F-1)(F)(F+1)/6 and the
The total length = p2 (3-2p)
total length of these below George’s floor is another
of possible floor-
tetrahedral number (f-1)(f)(f+1)/6 plus a “triangular- In Julian Havil’s “Impossible?” (reviewed in Volume 12) George’s origi-
to-floor commutes is a tetrahedral number: nal calculations are extended to include multiple elevators. Although he does
prism” number (F-f)(f-1)(f)/2.
(F - 1)(F)(F + 1)/6 and the total length of these not state it, Julian’s multiple elevators are assumed stationary when George
below George's floor is another tetrahedral number pushes the buttons.
(f - 1)(f)(f + 1)/6 plus a \triangularprism" number
1 (F - f)(f - 1)(f)/2.
probability (elevatorProbability
below) (elevator below)
(f - 1) (f) (f + 1) + (F - f) (f - 1) (f)
6 2
= (F - 1) (F) (F + 1)
6
In an attempt to deal with the purgatory population
In an attempt to deal with the purgatory
explosion (which lags behind the
population explosion (which lags behind the earth's
earth’s population explosion by
populationanexplosion by an average lifespan),
average lifespan), the transcen-
dental administration decided last
month to scrap the outdated ter- 23 http://twistedpencil.com/
raced landscaping that Dante vis-
 We usually think of became very common in the Edo period, and
mathematics as something that By were usually painted on wooden shingles in ink.
Fukagawa
is largely independent of an Hidetoshi The most common images were lucky animals,
individual culture. However, and such as horses, but the images associated with
Tony Rothman
when a society is isolated Reviewed by ISBN 978-0-691-12745-3 mathematical problems, called sangaku, filled
Princeton University Press

from the rest of the world, Tom Archibald the dual role of showing reverence to the god and
Simon Fraser
mathematical thought will demonstrating the mathematical prowess of the
University
continue and sometimes individual who made the offering. The problems
develops in a direction that were also collected in books and apparently were
is independent of what is happening elsewhere. widely popular, and there are also manuscripts
That is exactly what took place in Japan in the containing such problems that have not been
Edo period, when the feudal government of the found posted at temples. The problems range from
Tokugawa clan decreed that Japan should be easy to very difficult.
closed to the outside world. This situation started Hidetoshi and Rothman here present a
in the early 1600s and was to continue until selection of the problems, graded
the country was opened up more or according to difficulty, with
less forcibly to the west, notably hints and solutions. The
with the Meiji restoration of 1868. aim is to give the flavour
During that long period, wasan, of these works, and this is
Japanese mathematics, went its quite successful. High school
own way. This has been very or beginning university students
nicely discussed in an excellent interested in problems can read and
book by Annick Horiuchi, but work on these with profit, since the methods for the
you have to read French. 1 most part don't go beyond elementary calculus.
The present book, in The traditional solutions are often surprising.
English, also discusses this The very first of the ninety problems cited states:
period, but rather than looking \There are 50 chickens and rabbits. The total
at research mathematics, number of feet is 122. How many chickens and
it presents a remarkable and how many rabbits are there?" Most students
intriguing fad: the posting of now would approach this with algebra, but the
mathematical problems, particularly in geometry, traditional solution begins: \If rabbits were
on pieces of wood outside temples as a kind of chickens the total number of feet would be 100
offering. This is the \sacred mathematics" of the ..." A problem that has a more Japanese flavour
book's title. Offerings of images at Japanese temples involves drawing figures on a folding fan. Unfolded,
the Japanese fan is a sector of a circular annulus
1
A. Horiuchi, les mathématiques japonaises à l’époque d’Edo. Paris: Vrin, or ring. Suppose the sector has radius R, and draw
1994.

24
 two circles of the same radius r that touch
one another and also the outer and inner
edges of the fan. we draw a chord length d
to the outer edge tangent to the tops of the
two circles, and add a small circle radius
t tangent to the chord and the fan, as
shown. Given d and the diameter of the There
small circle, we are invited to find the are 50
diameter of the large circles. The chickens and
proposer gives the answer (3.025 rabbits. The
if d=3.62438 and 42t=3.025) total number of
but not the method. Indeed, we feet is 122. How many
don't know his method, since the chickens and how
many rabbits are
answer is incorrect. The authors
there?
don't provide an explanation for
the error; nor do they suggest any
reasons for these rather surprising values.
But they do give a solution, involving constructing
some auxiliary lines and using the Pythagorean
theorem.
The authors also give some basic historical
background and examples of the Chinese sources
of Japanese mathematical work. Hence the book
gives a nice glimpse of East Asian mathematics
of several centuries past. Not all of the historical
details are as carefully written as would be ideal,
for example on the subject of the language of the
tablets. This doesn't detract from the interest and
overall usefulness of the volume, especially for the
student reader, and it should serve to stimulate
further reading on the mathematics of these
cultures. The illustrations, some in full colour, will
also entice the reader.
Students and teachers will read and work on
the problems in this book with pleasure. They
give an interesting glimpse of another
culture, one in which the value of
mathematical problem solving merits
a kind of religious recognition. An
Most
inscription on one of the sangaku
students
now would gives a hint of why: \Confucius
approach this says, you should devote all of
with algebra,
your time to study, forgetting to
but the traditional
solution begins: “If have meals and going without
rabbits were chickens
sleep. His words are precious
the total number of
feet would be 100 ... to us. Since I was a boy I have
been studying mathematics
and have read many books on
mathematics. When I had questions, I visited and
asked mathematician Ono Eijya. I appreciate my
Master's teachings. For his kindness, I will hang
a sangaku in this temple." These days, such
recognitions are unusual.

By
Apostolas
Doxiadis
Faber and Faber
ISBN-13 978-0571205110 Goldbach's conjecture is 4=2+2
Reviewed
by
easy to state: 6=3+3
Clara Park
CheongShim
\Every even number 8=3+5
International Academy greater than 2 is the
sum of two primes." 10=3+7=5+5

This statement seems to 12=5+7

be straightforward as we can 14=3+11=7+7…
attempt proving it straight away. We can conclude easily that this theorem is

25
 plausible! But, seemingly, this simple The main character, Petros Papachristos, is a
theorem, known as Goldbach's mathematical genius but considered a complete
Conjecture, has remained one of failure in life. Recklessly, he abandoned a successful
It the most tormenting problems in career in academics, shutting himself up away from
will the field of mathematics, along family and friends in order to devote his every
appeal to with Fermat's Last Theorem. waking hour to proving Goldbach's conjecture, a
almost any For centuries, many talented theorem that mathematicians have been unable
student, even mathematicians were enticed to solve for almost three centuries. As foolish as
students with no to this myth but were unable any of us might believe such an action might be,
previous interest
to prove the theorem. through the eyes of an understanding nephew, we,
or background in
mathematics. Thanks to today's too, soon come to understand Petros' passion and
technology, mathematicians realize he is deserving of our respect. We empathize
discovered, with the help of with Petros' bleak hope of solving the problem,
supercomputers, that the conjecture holds true and we are just as exhilarated as Petros' nephew
for all even numbers up to 400,000,000,000,000 when he learns that his uncle has finally cracked
(four hundred trillion). Faced with the reality that the secret.
\numbers are infinite," however, they were once The story is not about problem solving. It is
again depressed since they are still not able to rather a story of challenge, persistence, and
provide a general proof. devotion. Petros' obsession with the problem
Books on mathematics are often `how- suggests that the result is not the only thing
to' guides on problem-solving or earnest that matters in life.
treatises on some area of mathematics, I strongly recommend this book to other
but Apostolas Doxiadis' Uncle Petros The story is not readers. It will appeal to almost any
and Goldbach's Conjecture takes the about problem student, even students with no previous
form of a fictional story that can solving. It is rather interest or background in mathematics.
a story of challenge,
entertain and inform a wide range of They will be stimulated and entertained
persistence, and
readers regardless of their mathematical devotion.
as well as delighting simply in the spirit
knowledge. Readers may feel the same of determination of all those people,
sort of enjoyment as watching a movie, including mathematicians, who risk all in
such as Forrest Gump. attempting to do the seemingly impossible.

Math Challange Winner Announced

Obtaining solutions to four Math Challenge Problems posed in Issue #11 (Spring
2008) proved challenging to our readers. At the time of publication of Issue #12 (Fall
2008) we had not received any solutions, so the deadline for submissions was extended to
February 28, 2009. I am delighted to announce the decision of our panel, who decided
to split the prize between two very talented students. Each submitted three correct
solutions to the Challenge Problems, however, neither submitted a solution to the fourth
one - Problem #6.
The two winners, who will split the $300 prize money, are: Joshua Lam of The Leys
School, Cambridge, England and Rati Gelashvili, of Tbilisi, Georgia, currently studying
at the Georgian Technical University. Congratulations to Joshua and Rati.
Solutions to the four Challenge Problems appear in this Issue of Pi in the Sky. We
have included some solutions by our two Prize winners.

26
 Solutions to problems 4,5, 6, and 7 published in
Spring4,5,
Solutions to problems 20086, Issue
and 7ofpublished
Pi in thein
Sky
the
Spring 2008 Issue of Pi in the Sky
Problem 4. Prove that any positive integer n coprime to 10
Solutions to
Solutions
Solutions problems
Solutions to to toproblems 4,5,
problems
problems 6, and 4,5,
4,5, 7 published
4,5, 6, 6,6,andandand777published in
publishedthe
published in
ini
Problemrepunit.
of a 4. Prove Math that any positive Problems
Challenges integer n coprime to 10 is a div
Solutions Solution Spring by 2008
to problems 4,5, 6, and 7 published in the Spring
Rati Spring
Spring Issue
Gelasvili: 2008 of
2008
2008 PiIssue
Issue in
Issue theofofof Pi Sky
Pi Pi in inin the
the
the Sky
Sky
Sky
of a repunit.
Since
Spring
Solution by (9n,2008Rati10) Issue =Gelasvili:
1 according
4, of 5, Pi6inand to
theEuler-Fermat’s
7Sky theorem 9n is
Problem
Since 10
(9n, 10)
ϕ(9n) − 1 = 99...99. Consequently, n is a divisor of 11...11.
4. =Prove
Problem
Problem
Problem 1 according that
4.4. any
4.Prove
Prove
Prove topositive Euler-Fermat’s
that
that
thatany anyinteger
any positive
positive
positive coprime
n theorem integernto
integer
integer 9n 10is isa adiviso
nncoprime
coprime
coprime divis
toto
to1
of10a repunit.
ϕ(9n) 1 =Alternative
99...99. Consequently,
Published
solution by is a
in
Joshua divisor Lam of 11...11.
Problem 4. Prove − ofofofaathat
arepunit.
repunit.
repunit.
any positive integer n coprime to 10 is a divisor n
Solution
Alternative Letby
Issue
SolutionmRati
Solution
Solution be aGelasvili:
11
solution bypositive
(Spring
by
byRati by
Rati
Rati Joshua integer
2008)
Gelasvili:
Gelasvili:
Gelasvili: coprime
of Pi intothe
Lam 10. Let SkyRn be the r
of a repunit.
Sinceby (9n,
Let n 10)
mSince1s.
be
Since
Since = Consider
a (9n,1(9n,according
positive
(9n, 10)
10) each
10)=integer to
==111accordingRaccording
according (mod
nEuler-Fermat’s
coprimetoto m). Since
10. theorem
toEuler-Fermat’s
Euler-Fermat’s there
Let Rn be
Euler-Fermat’s are9n infinitely
theis arepunit
theorem
theorem
theorem divisor
9n
9n ma
9n iswi
Solution Rati Gelasvili:
10 ϕ(9n) only
1 = 99...99.
m different Consequently, residues (mod
is a divisor
m), by ofPigeonhole
11...11. Principle
10)1s.= Consider
− 1 10 10 each
10 −−−1to 11= == (mod Consequently,
99...99.
99...99.
99...99. nSince
Consequently,
Consequently, there are
isisisaaisinfinitely
adivisor
divisor
adivisor many
ofofof 11...11.
11...11.
11...11.
ϕ(9n)
ϕ(9n)
ϕ(9n)
Since (9n, n according R Euler-Fermat’s
n m). theorem nnn9n divisor of Rn ,
Solutions to
only problems
Alternative
m idifferent
> Alternativesuch
4,5,
solution
jAlternative that
residues
Alternative 6, and Rby (mod
solution 7
≡Joshua
isolution
solution published
Rjm), (mod
by
by Lam
by
by m). in Then
Pigeonhole
Joshua
Joshua
Joshua theLam Lam
Lam i − Rj ≡ we
RPrinciple 0 (mod
can
10 ϕ(9n) − 1 = 99...99. Consequently, n is a divisor of 11...11.
iSpring
Letj m 2008
be that
m|R
such ai−j
Let
Let Issue
Let 10
positive
·m j and
i bebe ofainteger
be aaPi since in coprime
(mod
positive
jpositive
positive (m,
the 10)
Sky
integer
integer =tocoprime
Then
integer 1coprime
10.we RiLet
coprime concludetoR
tojton10. be
10.
10. 0that
the
Let(mod
Let
LetR repunit
m|R be
be wi
,the
bethe he
th
4
Prove that any positive integer n coprime
Alternative > solution Rm
by m ≡
Joshua R Lam m). −R ≡ RRnnnm)i−j
n 1s. Consider
· 10 j
nnn1s. 1s.and
1s. each sincencoprime
Consider
Consider
ConsiderR (mod
(m, each 10)R
each
each m). = nn1
Since
n(modwe
(mod
(mod there
conclude Sinceare
Since that
Since infinitely
thethere there
there
m|Rare are many
arewith
infinitely
infinitelyR
infinitely ,
n m bm
5
to 10 is a divisor of a repunit. Let m m|R be ai−j positive integer toRR 10. Let m).
m).
R m). be repunit i−j
Problem 5.
Let Letn be n be
a a positive
positive ninteger
integer andand let
let ϕ(n)
(n) denotes
Problem 4.only
n 1s. Consider Prove meach different
that only
of R
only
only
anyn (mod
positive
m residues
different
positive
m m different
different
m).integer
integers
(mod
residues
Since residues
residues
lessn there
or
m),
coprime
equal
by
(mod
(mod
(mod
are Pigeonhole
tom),
infinitely
to n10
m),is
m),
that
by byby
aare Principle
Pigeonhole
Pigeonhole
Pigeonhole
divisor
many coprimeRn , but we
to
can fi
Principle
Principle
Princip
n.num Pr
i > Problem
j such that 5.
i ii>>>jjjsuch R Let
such
suchdenote
≡ n R
that be
that
that a
(modpositive
the number
m). integer
Then ofR and let
positive
− R ϕ(n)≡ 0 denotes
(mod
integers m) the
≡≡000,(modhen
Solutions toSolution problemsby 4,5, Rati
6, of Gelashvili
and 7 only
published
a repunit. m different in the residues (mod
i j R
m), R i i i≡
R
by ≡ ≡RRR
Pigeonhole j jj(mod
(mod
(modm). m). i Then
m).
Principle Then
Thenj R we RRi i− −−RRR
ican jj≡
j find (mo
(m
of positive any positive
integers lessintegers
or equal m and
to nthat
·R10 and since j j(m, to10) =that 1(m, wenare conclude==1are coprime
that to that Prove that
j
less or n.
blems 4,5, 6, and Spring 2008 Issue
7 published in theof PiSolution
inithe > jSky bym|R
such Rati i−j Gelasvili:
that m|R
m|R
m|R
i ≡ R i−j
i−j
ji−j
·equal
·10
(mod ·1010 jand
and
and
m).
nsince
since
since
Then (m,
R(m, 10) 10)=coprime
10)
i − Rj ≡
1 1wewe0we conclude
(mod
tom|R
conclude
conclude n.
m)i−j
Prove
,that
that
hence m|R
m|R
m|R i−j
i−j
i−j
2008 Issue of Pi in 10)the = Sky any positive thatto integers
for any and n integers
mpositive mm|R and n2 )ϕ(n2of ).
Since (9n, 1 according Since to (9n,
m|R10)
Euler-Fermat's=
· 10 1j according
and since (m, Euler-Fermat’s
10) = 1 we conclude theorem ϕ(mn) that 9n
≤ is a
ϕ(m divisor
i−j Problem Problem 5. Let n be5.
Problem
Problem a5.Let
5. positive
Let
Letnnnbe be
be integer
aaapositive
positive
positive andinteger let
i−jϕ(n) denotes the numb
integer
integer andandandlet let
letϕ(n) ϕ(n)denote
ϕ(n) denot
denot
theorem 9n is a divisor
Problem 4. Prove that any positive integer n coprime of 10
10 (9n) − 1 = 99...99. Consequently,
ϕ(9n)
- 1 = 99...99.
to 10 is aofintegers
divisor n is a divisor
ϕ(mn) ≤of 11...11.
ϕ(m 2 )ϕ(n2 ).
Problem of positive
5. Letby of
of Solution:
positive
n be positive
positive less
a positive or
integers
integers equal
integers less
less
lessto or n
or or that
equal
equal
equal
integer and let ϕ(n) denotes the number n.n. are
to to to n coprime
n nthat
that
that areare
are to n.
coprime
coprime
coprime Prove toto
tothatn.PP f
Consequently, Alternative solution Joshua Lam
of
ataanyrepunit.
positive integer n ncoprime is a divisor to 10 isofa11...11.
divisor any positive any
anyintegers
anyLet positive i be the
positive
positive
pequal m and nprime
integers
integers
integers n mm factors
m andand
and nnnof mtoandn.n,Prove qj the prime for factor
Solution by Rati Gelasvili: Let ofmpositive
be a positiveintegers
Solution: less
integer or coprime to to thatLet
10. are R coprime
n be the repunit with
that
any positive integers
Let pi be and m ther and the prime prime factors
factors Solution

of
of infinitely n only.
and2 )ϕ(n   
q 2 the R2prime factors of m o
(9n, 10) Alternative
= 1 according Solution nby 1s.Joshua Lam
Consider each n mϕ(m n, many
9n isRna (mod Since there are but
k
Since
elasvili: to Euler-Fermat’s theorem divisor m). of ϕ(mn) ≤ ϕ(mn)
ϕ(mn)
ϕ(mn) ≤≤≤ j ϕ(m ).
ϕ(mϕ(m n ,)ϕ(n
)ϕ(n
2 2)ϕ(n 22).2).).
and r the We
prime have factors of n only.
10 ϕ(9n) − 1 = 99...99. Consequently, n is only a divisor m different residues k (mod m), by Pigeonhole Principle we  can find
ing to Euler-Fermat’s theorem 9n is a divisor ofof 11...11. Solution:We have Let ϕ(mn)
pi beThen ≤
the R ϕ(m2 )ϕ(n
prime   2 ).
factors of m  and n,1 qj  the 
Let m be a positive integer i > j
coprime such that
to 10: R ≡
Let R (mod Solution:
Solution:
Solution:
m). − R ≡ 0 (mod 1 m) , hence 1
Alternative
sequently, n is a solution
divisor of by Joshua Lam
11...11. i
Let(m,
j
be ϕ(mn) =m i
mn j
1 −n, 1 −  factors 1−
10)the 1prime factors of and mthe prime of
ron
pi prime factors of  only,  and 
LetJoshuaRnbebeaLam the repunit
positive withcoprime
integer m|R
n 1s. Consider
to 10. i−jLet 10Solution:
Rnand
j
· each R since
(mod
benandther repunit
Let
=Let
with
Let pipibe
pwe i bebethe the
the
concludeprime prime
prime that m
factors
1factors
factors
m|R i−j of prof
of qjm
im
k
the and
and prime
1andn, n,n,qqjqjjjthe factors
thetheprime
1prime
prime m
facto
facto
fact
n by m the of prime
n
ϕ(mn) only. =
factors mn of n 1
only. − i 1 − j 1 − k
k
Let p be the
k primeand
and
and r rr
factors
k k the
k the
the prime
prime
of prime
m factors
andfactors
factors n, qof of
of n
the n nonly.
only.
only.
prime factors of m only,
1s. Consider
nteger
n m). Since
coprime each
to 10. n (mod
there
RLet Rare
n bem). theSince
infinitely there
repunit many areRinfinitely
with , but i only
n 5. LetWe many have Rn , but pij
j 
qj 
k 
rk 
Problem
and r the primen be a positive
factors We WeWeWeof have
have integer
have
have
n only.
iand let ϕ(n) denotes the 1 number 1
only
modm m). different
mSince
different residues
there (mod(mod
are infinitely
residues m),manyby
m);Pigeonhole
n , but
by Principle we can find
the Pigeonhole
k  coprime 2
) = m 2 n.Prove 1   for 1−  
ofRpositive integers less or equal to n that  are ϕ(m
1
 
to  −  
that    
j such
>(mod that
by Pigeonhole
Principle
m), Riwe ≡ canRj (mod Principle
find > jThen
im). we
such can i find
Rthat− RR We R
j i≡
have
0j (mod
(mod m) , hence= mn 1== −mn
  1 11
1−− 1 1  p i 11 1−11 1 qj 11
any positive integers m andϕ(mn) n  ϕ(m ϕ(mn)
ϕ(mn)2
ϕ(mn))= = m

2mn
 mn p 1 − 1
 1 1
i−
−
 q 1 − 1 1 j1
− − − r 111−−−
m|R (mod ·
i−j m). 10 j and
m). Then since
ThenRRi i−- R (m, 10) =
Rjj ≡ 00 (mod 1 we
(mod m) conclude
m),, hencehence that m|R i−j 1 − 1i
i
i i i i ji
1 p  p pi i i j jj 1kj j jj
p j  q  q q q k  rrkrk
ϕ(mn) = mn 2 1− 2
j1 − 1 1kk
k
, 10) = 1 we conclude that m|Ri−j ϕ(mn) ≤ ϕ(m pi )ϕ(n
2 ϕ(n ).)  2
  qj 11
=n  1
 −
pi rk   1−
 rk111
j
Problem m|R 5. i-jLet• n10be and since integer
a positive (m, 10) and =let 1 ϕ(n)
we conclude
denotes the number i j   k 1 1 
ϕ(n) )==mnϕ(m
22 22 222
1= 1−1 11
)1= 1−
ϕ(m −− i − k
ofa positive that m|R
integers
integer less
and
i-j let orϕ(n)
equaldenotes
to n that the arenumbercoprime to n. Prove that for   ϕ(m
Solution: ϕ(m )) = m m
 mppii,1or111−
222
 − qrjk 111−−−
Notice that each 1
product
ii jk p is
ipi i replaced
p by
qqjqjj 1 if t
any positive
equal to nAlso integers
that are
solvedm andbyntoSimon
coprime n. Prove Letfor
that
Morris pi be the prime factors ϕ(m ) of 2
= m and n,1 q−j the
2
 prime 1i factors
−j  kof m only,
i i i

j j j
Notice that each iproduct pi , or 1 
is qj replaced  by 1 if there  are
nd n  and rk the prime factors of n only.
factors containing p2i , qj2,or jrk .   
1−11 1  111
ϕ(mn) ≤ ϕ(m )ϕ(n have 2 2 ). ϕ(n 2
) =
 ϕ(n n ϕ(n
ϕ(n )
i 221
)=)==
j −nnp22k2
n is equivalent 111−−− 1 111−−−

ϕ(mn) ≤ ϕ(m2 )ϕ(n2 ).
R and r are respectivelyWe the radii offactors the containing Now the  required
2pi , qj ,or rk 1.inequality
i  i 1k  ppipi irk kto rrkrkk
ϕ(n ) = n  1 −  
 2   
 1 −i i i k k

6 Now 
in the Notice required 1 product inequality p1product 
is or equivalent to
k,1
Solution:spheres circumscribed about, and inscribedNotice       
ϕ(mn) = mn that 1 Notice each
−Notice that
ithat
that1each − i
each
each ,product
product
k 1 1−is,r,replaced oror or isisby isreplaced 1 if there
1replaced
replaced by by 1are
by 11ififi
Let pi be athe prime factors
tetrahedron. of m that
Prove and n, R q≥j the3r. prime factors ofi mNotice only,pi that  each  qiproduct j j 1 − k i i,i r j jkj 1
orkkk is replaced
− ≤1

Noticefactorsthat each product , are or qjj1factors
p,i ,iq,is replaced j by1 1 if there are no
containing j
.pi− k q r k
and
ctors rk ofthe mprime
and n,factorsqj theofprime n only. factors of m only, factors
by factors
factors
1 if containing i , qj ,or
pcontaining
containing
there i j 1k
rk pno qj,or ,orrrkrkk. . .containing
j,or 1− k ≤ 1pi, qj, nor
Now the   
of We have
n only. Solution factors containing 2 p r ,. qrequired
Now
Now

,orNow rkthe the inequality
.the required
required
1jrequired
 inequality is j equivalent
qinequality
inequality1k torequivalent
isisisequivalent
equivalent
k toto to
     ϕ(m ) =ikm2j 1−  1− 
 1  1  Now the required 1 inequality ispiequivalent to
qj 1equivalent
  
 
ϕ(mn)  = mn  1
 −  1 − 1 − Now i the  required j 1inequality   
1−11is1   111 to
 1 Let  us consider 1  the p 1tetrahedron qj having rthe   q 1 − 111−−−1 ≤ 1
111−−− ≤≤≤111
1 − vertices 1at− the icentroids 1 − i ofj the faces k
k   1j1   j 1k1  qqjqjjrk rrkrkk
of the given 11−− 11−− ≤1
 ϕ(n ) = n
2 2 j j j k kk
pi qj rk
i j
tetrahedron. This k 
tetrahedron 1

is similar 1 to the j i
qpj i
k k
rrkk
given 
oneϕ(m
 and
2
 )= m2ratio
the  1of− similitude 1 −is 1/3 . Its   1 111 are no
1 1 p i Notice that
q j each product that is , or
obviously is replaced
true. The by equality
1 if there occurs only if
=m 2
1 −
circumscribed 1 −
sphere i j
1
pi qj is  therefore  of radius
 R/3.
 This i j k
i
sphere  j
intersects all
2  the faces 1factors
the
of containing
1
tetrahedron, pi , qj ,orm rkand . n have the same prime factors.
 ϕ(n 2
 ) = n 1 − Now or 1 −
theequal
 therefore 1  its radius 1 bepigreater rrequired inequality A issimilar equivalentsolution to was submitted by
= n2 1− 1 − should i k
k than
Notice
i
thethat pi
radiuseach k product
k 
of the rinscribed , or

sphere.
is replaced by 1 if there  Joshua 
are 1no
Lam  and Rati Gelashvili
1
  1− 1− ≤1
duct , or is replaced by 1 if there are no
i j k qj rk
actorsi containing
j k p ,
i jq ,or r k . 27
j k

r rkNow . the required inequality is equivalent to

 Solution:
Let us consider the tetrahedron having the vertices at the centroids of the
ces of the given tetrahedron. This tetrahedron is similar to the given one
d the ratio of similitude
Prove that is 1/3
in .any
Its circumscribed sphere is(a)
convex polyhedron therefore
Byofusing these inequalities and Euler's polyhedral
7
dius R/3 . This
there
radius should
sphere intersects
is either
be grater
which three or
al the faces
a triangle
equalmeet
edges than and
faceof or
theatetrahedron,
the radius
vertex at therefore
of the
(b) there is inscribed
formula we get
a face sphere. 4 = (2f - e) + (2v - e) ≤ 0
having less then six sides.
Problem 7. Prove that in any convex polyhedron (a) there is either a is a contradiction. Therefore any convex
which
angle face or a vertex at whichSolution
three edges meet and (b) there ispolyhedrona face should have at least a face with three
ving less then six sides. which
which
which isis ais
a contradiction.
a contradiction.
edges or at
contradiction. Therefore
Therefore
least a any
Therefore any
convex
vertex
any convex
at polyhedron
convex polyhedron
which
polyhedronthreeshould
should
have
edges
should have
have at
atat
If v, e, f denotes the number of vertices, least
least
least a face
aaedges
face
face withwith
meet.
with three
three
three edges
edges
edges ororor a at
aa at
at least
least
least vertex
vertex
vertex atat
at which
which
which three
three
three edges
edges
edges meet.
meet.
meet.
Solution: (b)
and respectively faces of the polyhedron then (b)(b)Assume
Assume
Assume
we by
byby
contradiction
contradiction
contradiction thatthat
that every
every
every face
face
face hashas
has
(b) Assume by contradiction that every face has more
more
more thenthen
then fivefive
five edges.
edges.
edges.
If v, e, f denotes the number of vertices, edges andAsrespectively faces
have:of
have Euler's polyhedral formula AsAs
aboveabove
above we
wewe
have:
have:
more then five edges. As above we have:
e polyhedron then we have Euler’s polyhedral formula
v-e+f=2. 
f f
ff  f
ff 

f

v−e+f =2 2e
2e2e
== = ffiii ≥ f≥i ≥ 66== 6 6f
=6f6fhence
hence
hence
hence e 3f.
ee≥≥ ≥3f.3f.
Let F1 , ...,Ff, V1, ...Vv be the faces and the vertices
i=1
i=1i=1
i=1 i=1
i=1i=1
i=1
Let F1 of Ff ,polyhedron,
, ...,the V1 , ...Vv be the facesnumber
fi the and theofverices
edgesofofthe
Fi polyhedron,
and fi
e number vi of
theedges of Fi ofand
number vi the
edges number
issuing Vi. Also,
of edges
from Also,
in
issued
Also, in any
in any
from
any polyhedron
Vipolyhedron
. Also, inwe
polyhedron we we
havehave
have
any polyhedron we have
(a) Assume the contrary, that there is no triangle face and there is no v v
(a) Assume the contrary, that there is no triangle   vv v
vvv
rtex at which three edges meet, i.e., 2e
face and there is no vertex at which three edges 2e2e
== = vviii ≥ v≥i ≥ 33== 3 3v
=3v3v
hence
hence
hence
hence 2e
2e2e
≥≥3v.
3v.3v.
≥
meet, i.e. fi ≥ 4, vi ≥ 4
i=1
i=1
i=1i=1 i=1
i=1
i=1i=1

f ≥ 4, vi ≥ 4 On
OnOn
the
thethe
other
other
other hand
hand
hand
Onby by
using
bythe using
using Euler’s
Euler’s
Euler’s
other polyhedral
polyhedral
polyhedral
hand, by usingformula
formula
formula we we
get
getget
wepolyhedral
Euler's
hen we have:i
Then we have: formula we get
f f v v
66==
6 (3f
=(3f(3f e)+
−−e)
− +(3v
e) +
(3v(3v
−−2e)
2e)2e)
−
    6 = (3f - e) + (3v - 2e)
2e = fi ≥ 4 = 4f and 2e = vi ≥ 4 = 4v
i=1 i=1 i=1
which
which
which
i=1
combined
combined
combined withwith
with
which the
thethe
above
above
above
combined twotwo
two inequalities
with inequalities
inequalities
the above leads
leads
leads
two to
to to
aa contradiction.
acontradiction.
contradiction.
inequalities
Hence
Hence
Hence at
atat
least
least
least oneone
one faceface
face should
a should
should have
have
have at
atat
most
most
most fivefive
five edges.
edges.
edges.
and e e≥ ≥2v.
nce e ≥ 2f hence 2f and e ≥ 2v. leads to contradiction. Hence at least one face
By using these inequalities and Euler’s polyhedral formula we get
should
A have at most five edges.by
Remark:
Remark:
Remark: A similar
A similar
similar solution
solution
solution was
was
was submitted
submitted
submitted byby
Joshua
Joshua
Joshua Lam.
Lam.
Lam. The
The
The prob-
prob-
prob-
lem lem
was
was
also
also
solved
solved
lem was also solved by
A by by
RatiRati
Gelasvili.
Gelasvili.
Rati Gelasvili.
similar solution was submitted by
4 = (2f − e) + (2v − e) ≤ 0 Joshua Lam. The problem was also solved
by Rati Gelashvili.
2
Math Challenges Problems
Published in
Issue 12 (Fall 2008) of Pi in the Sky

Prove that the equation:

1 x5 + y5 + 2 = (x + 1)5 + (y + 2)5 2 Prove that pn > 2n for every n ≥ 5 where pn
does not have integral solutions denotes the nth prime number (p1=2)

Solution by Edward T. H. Wang Solution by Rati Gelashvili

If x and y are of the same parity, then the left
side of the equation is even while the right side is
odd. If x and y are of opposite parity, then the left
side is odd while the right side is even.
Also solved by Rati Gelashvili, and Joshua
Lam
Let's prove by using induction. The assumption
holds for n = 5. If p3n3- 13 > 2(n- 1) then pn- 1 ≥ 2n -
1 hence pn > pn- 1 ≥ 2n- 1. Since pn ≠ 2n we must
have pn> 2n
Also solved by Edward T. H. Wang,
Joshua Lam and Andrew J. Pai

28
 rkWe first remark that
that Since
Solution the Pigeonhole
a square byofRati side Principle
length
The Gelasvili:
square 0.1 be
can at
canleast one square
be completely
partitioned contains
in 100covered byat aleast
little squares circlet
of si
a3 + b3 = (a + b)(a2 − ab + b2 ) ≥ (a + b)ab radius
3 3 2 2 The0.1 Since
square a
we conclude square
the be
can Pigeonhole of
partitioned side
that there length
exists
Principle
in 0.1
100 little can
aatcircle
least be
ofone
squares completely
radius
square
of side covered
0.1lengths
which conta
contains by
at B
0.1.
a3 + b3 = (a + ab)(a +2b−= ab(a ++ ) ≥ (a−+abb)ab
b2b)(a + b ) ≥ (a + b)abthe at least radiusofSince
three
Pigeonhole 0.1
thewegiven
Principle a conclude
atpoints.
square leastofthatsidethere
one length
square exists0.1 acan
contains circleat of
be radius
completely
least three 0.1poin
wh
cove
hence Inside a square with side length 1 there are at least Let
three a,
of b,
the c be
given positive
points. numbers such that
hence
3 201 points. Prove that there exists a radius
of radius 0.1 which contains  at least three
Since a square radius

at
circle
that
Problem
of
4
0.1 we conclude
of side0.1
at Let
abc
4.
Problem
= 18.
least
length
three
that
a,
4.
we conclude
b,Prove be
cthere
Let
0.1 can that
positive
of the
a,
that
b,
given
exists
c be
there exists
be completely
anumbers
points.of such
circle
positive
radius
numbers
a circleby
covered
that0.1 abc
of aradius
which
such that
circle0
= 18. Pr
conta
abc
a 3 + b3 3
cthese (a + b)ab
points.  c3  (a + b)ab c3 √ least √ three of athe 3 +givenb3 + c3 4. √
points.
√ √
√ + √ ≥ √ + √ ≥ 2 √ √ = c √ abc a that
+ b. Problem Let + a, b, bc be
+ +positive
+ numbers
a + b (*) such tha
3
a (a++ b b)abc c (a + b)ab
3 3 3 (a + c3
b)ab c3 (a + b)ab
√ √c 3 √ 3 + ≥ 3a + b 3 c √ a c √ c √
≥ 2 √ 2 √ + √ 2
+√ ≥
≥ 2 √
2 2
Solution + √
√ by√ 2 ≥ 2
Rati 2 √ 2
= Gelashvili
c abc a + b. √ 2 = c abc a +
Problem 4.that b. 3 a
Let a, b, c be positive
b c
≥ numbers
a b + c + such
b athat
+ c +
abcc =a 18.
+ b Pro
2 2 2 2 2 2 2 2 22 2 2 2 2 3a3 by + b3Rati + c3 Gelashvili √ √ √
thatSolution by Rati Gelasvili:
Solution √ √ ≥ a b + c√+ b a + c + c a +
Similarly, The square can be partitioned in 100 little 3 + bthen3 + 3 √ √
Similarly, Since abcSolution =a√18 byc Rati
≥ a =b√
abc Gelasvili:
3+3 c2.+ If b √ wea√+multiply
c + c a the + b left side of
squares of side lengths 0.1. By the Pigeonole Since 3 =√ = 18 then = 3 2. If we multiply the left
+ √ √ inequality by Since3 2abc and
Solution
abc the
18 then right
by Rati side Gelasvili:
abc by√ abc then
If √ √ the
we inequality
multiply wh
Principle atc33 √ leastb3 one
+
a3 square contains
+
at least inequality by 3 2 and the right side by then the inequ
3 +b 3 3 c +√ b3 √ √
a 3 ≥ a √ abc √ c b We should
Solutionbe proved
the byleft is
Rati replaced
Since
side Gelasvili:
of
abc by
=
the 18 then
inequality abc by = 3 2.
abc If
and we multiply
the th
c three
√ +points.
a
√ ≥ 2aSince √2abc+ac√square +2 b≥ a of abcside c +length
b 0.1firstcanremarkWe thatfirst remark √ that √ √ √
2by a circle with radius Since should be
abc =inequality proved
18 then is
by replaced
abc3 = 2√3and2. by the right
If need
we sideinequality:
multiply bythe abcleft then
side ofthet
2be2 completely 2 2 covered 2 0.1√2 right √side by
 √ then we √ √ the √
3 +3 32+ + c2by+√b by a +then c +√cthe a+ b). √ (∗)whi
3 3
3 + c3 3 √ √ inequality aby √bshould
 c3 be
and ≥theprovedabc(a
2  3 isside
right √ breplaced abc inequality
we 3conclude a
3that there b exists a circle of radius 0.1 a + b2 =a3(a++b3b)(a a+ c+3 −
3
b ≥ab =+ (a b+)b)(a≥ (a −2 + b)ab
b + cab ++ b ba )+≥c (a
2
+ c+ b)ab
+c
a3 which b3 a 2√ +√ c3 + √√ b3 ≥ b√abc√a + c should be proved is replaced by abc(a
√
a + b).
√ + √ ≥ b abc+ a +We
contains √ 2at least√ 2 3 of
c≥ bfirstthe a + c thathence
given
abcremark points. √   √ √ √
c- b)≥≥abc(a
2
2 2 2 2 2 2 √ 
(usinghence the fact
 2 a√that 3
+ b (a3
+
√
3
√
0 for any b + c√ a+and b ab). +c+c a
The inequality Also (∗)solved is obtained by by Edward adding side T. by H.sideWang, the above 2 2a3three+ b3 +2cthis 3
abc(a b +
≥inequality 4c have
+ b a + c + c a + b). (∗)
The inequality
y (∗) is Joshua
obtained Lam (∗) is
by addingobtained
andside by adding
by sideJ. side
thePai. by
above side the above Using
three we
equalities. Andrew a +We
3
b three
3
= (aremark
first + b)(a that − ab + b ) ≥ (a + b)ab  4 
equalities. We a first
3 + remark c3that
3
bthe a3(a++b3b)ab c3 c3 (a + b)ab (a + b)ab c3 c3 (a√+ b)ab √ c3 √
≥ 2 = 4c abc
Let ABCD be a convex hence quadrilateral, M + +2 √
Let be a convex quadrilateral, √ the √ ≥3 √ √ + √
midpoint ≥ ≥ 2√2 +√√ √ √ a√ + b.= c ab
5
Problem 5. ABCD M +2b = 2 (a2 +2 b)(a −2ab + 2 bb22))4≥ 2(a2+ b)ab2
3
midpoint ofBCbe and midpoint of 2CD.
N the quadrilateral, 2 the Ifamidpoint 2 2
BCProblem
Let and
ABCD N 5. the Let
be a convex
ABCD
midpoint of CD. a If
quadrilateral,convexAM +
M ANthe =midpoint
1 then Mprove + b3the
a3that = (a2 + b)(a22 − ab + ≥ (a +2b)ab 2
AM + AN = 1 then prove that the area of the
of and N ofthe midpoint of+CD. If1/2. + prove
AN =that 1 then the prove 3that the
eBCmidpoint If AM hence 
ea the quadrilateral
CD.
quadrilateral is less less a=
is than
AN than1AMthen
3 + b31/2.c3 hence
(a +
Similarly,
b)ab c Similarly,
(a + b)ab c3 √ √
ea of theisquadrilateral
ilateral
Solution: less than 1/2. is less than 1/2. √ +√ ≥ √ +√ ≥2 √ √ = c abc a + b.
Solution:
We have
Solution2 2 2 2 2 2 c32+ b23 a23  c3 √ + b3 √ a3 √ √
We have We have a 3 + b3
3 3 3
c3 (a + b)ab2 2 c3 √ 3+ √  ≥(a a√ + abc
b)ab +cc√ +3 b≥ a √ abc√ c + b
a √ +b + √ c ≥ (a +√b)ab + √ c ≥ 22used (a2the+√2b)ab √ c3 2 = c√abc √ +√
+ ya b.
SABCD = SABC + SSimilarly, ADC = 2SAM C +22S √2AN C+ √2 ≥ Notice 2√2 + that √2we ≥2 2√2 inequality √2 = c xabc a≥+2 b.xy
= S +C + 2S= 2 2 2 which 2 2isa3valid 2 any 2+ c23 √ b23 numbers
AN2SC AM C + 2SAN C + c3 for a3√
b3 non-negative
BCD = SABC SABCD + SADC =ABC 2SAM SADC √ x√and y.
= 2S AM CN = 2SAM N + 2SCM Nc.3 + b3 a3 √ √ + √ ≥ b √abc +a √ + c ≥ b abc a + c
= 2SAM CN ==2SAM 2SAM 2S= 2SNAM
. N + 2SCM Similarly,
N. √ + √ ≥Similarly a abc 2 2c + b 2 2 2 2
N + CN Similarly,
Since dist(C, MN)
CM
< dist(A, MN) 2 (the2 2
Since dist(C, M N ) < dist(A, M N ) ( the quadrilateralThe is convex) we
inequality cThe con-
(∗)
3 +is 3obtained
inequality by√ adding √ side by by adding
side theside aboveby three
3√+ b a33 (∗) is obtained side the
MSince
ude ) <dist(C,
Nthat quadrilateral
dist(A,
SCM NMM <NNS)AM)<( Ndist(A,
the. isHenceSconvex)
quadrilateral
MN ) ( theis
ABCD <
quadrilateral
weconvex)
4S include 3is+
weacon-
AM N . inequalities.
convex)
that
c3 b3we con-
c √ b 3√ a
+ √ ≥ a √abc√c + b
ude
<OnSthat
theNS N<
HenceS
Sother
.CMN <S
handSABCD <HenceS
. .Hence 4SAMSN . <<4S 4SAMNAM.N . √ + √ inequalities. ≥ b2√abc 2 +a √ + 2c ≥ a abc c + b
AM CM AM N
AMN ABCD
ABCD
2 2 2 2 2 2
On the other
hand On hand
the other hand Problem 5. Let 3beLet a convex quadrilateral, the midpointM t
1 The inequality 1 (∗) is obtained by adding aProblem
3+ ABCD
c 3 5.
b √ABCD
the√ be three
a convex M quadrilateral,
S = 1 AM · AN sin 
M AN ≤ 1 AMof · AN and the a3 √ + c3 side
midpoint + of√b3by side√abc
≥midpoint
bIf √above
aof
+ +AN c =If 1 AM then prove that the pr
1 S AM N
2 1 sin M BC N of BC and the
2√2 + √2 ≥ b abc a + c
N CD. AM CD. + AN = 1 then
MN = AM AM·N AN=sin M AMAN ·inequalities.
AN
≤ AM  AN ≤ 2 AM · AN
· AN 2 2 2
2 21 2 area of
1 1 2 1 inequality (∗) the quadrilateral is less than
area of the quadrilateral is less than 1/2. 1/2.
1 = 1 AM1 (11− AM 1 ) ≤ 1 · 1 =The 1
The. Solution:
inequality (∗)
is obtained by adding side by side the above three
Solution:
is obtained by adding sidemidpoint
by together,
side the above
= AM (1 − = AM2) AM ≤ (1 · −= AM. ) ≤ 2 5.
Problem · 4inequalities.
= 8 .ABCD be a convex
Let Adding quadrilateral,
these M the
inequalities we threeget
2 2 2 of4 BC8and N the 2 4 8 We have
midpoint of CD. If We
inequalities. AMhave + AN = 1 then prove that the
Therefore Therefore
Therefore area of 1the quadrilateral
1 Problem is less5.than Let 1/2.
4(a3 + be
+ABC ca3 )convex √ quadrilateral, M the midpoint
SABCD < 41 = 1. Problem 5. Let SABCD ABCD
ABCD =√ b3 S
be S aABCD +
convex S =√ 2S
=quadrilateral,
SABC + ++CS +√2S
M + =cthe
+
√
2Scmidpoint
aC++ b).2SAN C
1 S 1 Solution:
< 4 8 = 2 . of BC and N the midpoint2of 2CD. If AM + AN = 1 then prove that ≥ ADC
abc(a b AMc b ADCa AN C AM
the
SABCD < 4 = ABCD . of and the midpoint of If + = 1 then prove that the
8 2 We 8have2 area of the quadrilateral is =
BC N 2SAM CN = =
CD.
less than
AM
1/2. √AM
2S 2SNAM AN +CN√2SCM = 2S N .AM N + 2SCM N .
area of the quadrilateral Dividing than 1/2.by abc = 3 2 and rearranging
is lessthrough
Also solved by Joshua Lam. Solution: Since dist(C, )the
Since dist(A,
< dist(C, M N ) (*) ( the
dist(A, quadrilateral
M N ) ( the is convex)
quadrilateral we con- is con
SABCD Solution:
= SABC +S MN
gives inequality
ADC = 2SAM C + 2SAN C
<
There are given five line segments We
clude have that
having HenceS 4S
5 We = have2S S CM clude
N < S that
AM N .
S < S
ABCD <. HenceS AM N . < 4S .
AM CN = Also 2SAM N + 2SCM N . AM
CM N
solved by N JoshuaABCD Lam AM N
using
6 the property5 that any

these triangles must be acute.

5 three of them can
sides of a triangle. Prove that at least one S
Since dist(C, M N ) < dist(A,
Onbethe the
Sof
other hand
ABCD
ABCD ) (=
M Ngeneral
On
= SABC +
techniques
the

=2S
other
=theSquadrilateral
ABC + SADC is
1
audience.
SADC
that
hand

·=
would
2SAM
= 2SAM
=convex)
sin
2SAM
1
not
+
C +
2SCM
+
C we
be2S
2S standard
con-
1
AN C
AN C
.sin M ·
for a
1
clude that SCM N < SAM N . HenceS SAM N= < 2S4S
AM AM CN
SAM =ANN2SAM = N M AN
+
AM 2S ≤· AN NAM
. ANAN ≤ AM · AN
ABCD 2AM N
AM CN . N 2 CM 2N 2
Solution
On the other hand Since dist(C, M N ) < dist(A, 1 M N ) ( the quadrilateral 1 1 1 1 convex)
is 1 we 1 con- 1
Let a ≤ b ≤ c ≤ d ≤ e be lengths ofclude theSince dist(C, M N ) < dist(A,
segments. = AMM N(1 ) (−the AM )≤
= quadrilateral
AM ·(1 − = AM ) ≤ · we=con-.
is. convex)
that1SCM N < SAM N . HenceS 2 1ABCD < 4S2 N . 2 4 8 2 4 8
Assume by contradiction that allSAM cludeconstructed
the that SAM < Ssin  .ANHenceS < 4SAM AM N .
NOn=the other CM N· AN
hand AMM N ≤ ABCD AM · AN
triangles are right or obtuse. On the 2other hand Therefore 2
Therefore
1 1≤ 1 · S 1 1 < 4 1 = 1 .1 1 1
= AM (1 − AM= ) 1 =
ABCD . M
sin  SABCD1 <· AN4 = .
Then S
2 SAM N = 2 AM
AM N AM · AN
2 ·4AN 8sin M  AN8 ≤ 2
AN ≤ 2 AM · AN 2
AM 8
e2 ≥ d2 + c2, d2 ≥ c2 +Therefore b2, c2 ≥ b2 + a2 2
1 1 12 1
= 1 AM (1 − AM ) ≤ 1 · 1 = 1 .
and therefore =1 2 AM 1 (1 − AM ) ≤ 2 · 4 = 8 .
SABCD < 4 = 2 . 2 4 8
8 2
e ≥ d + c ≥ c + b + b + a ≥ b +Therefore
2 2 2 2 2 2 2 2 Therefore 2
2ba + a 5 5
1 1
From this inequality we obtain that e ≥ a + b, SABCD < 4 1 = 1 .
SABCD < 4 8 = 2 .
a contradiction. 8 2
5
29
5
 Problem 1

Find all positive integers n such that

log2008n = log2009n + log2010n.

Problem 2
Find the smallest value of the positive integer n such that
(x + y2 + z2)2 ≤ n(x4 + y4 + z4) for any real numbers x, y, z.
2

Problem 3
Let a be a positive real number.
Find f(a) = maxx {a + sin x, a + cos x}.

Problem 4
Prove that the equation x2 - x + 1 = p(x + y)
where p is a prime number, has integral solutions (x, y) for
infinitely many values of p.

Problem 5
Find all functions f : such that
3f(n) - 2f(n + 1) = n - 1, for every n .
(Here denotes the set of all integers).

Problem 6
In ∆ABC, we have AB = AC and BAC = 100o.
Let D be on the extended line through A and C such that C is
between A and D and AD = BC. Find DBC.
.