Tips Eureka Math Fun From Many Angles 1

21 21 21 21 21' .
22
23
22
23
22
23
2
2;~2 L1
22
21
22
Z1
22
21 21
22 22~1 .
21 21 21 21
24
25
24
25
24
25
24 3 23
25 24 24
23
24
23
24
23 23 2
24 23 23.l,\5Z---~~
22 22 22 22
262626 2525 r .. -' ''' --

Author's Note _____________
The name EUREKA is taken from a charming bit of bath-

ematics. More than 2000 years ago, the King of Greece asked
the famous mathematician Archimedes to determine if
his crown were pure gold or alloyed with silver. While
lounging in the tub, Archimedes realized that a crown of
pure gold would displace more water than an equal volume
of lighter alloyed gold. So enthralled was he by this dis-
covery that he ran naked through the streets of Syracuse,
shouting, "Eureka! I've got it!"
Well, I hope you receive the same pleasure from reading
this book that Archimedes got from solving his problem or
that I got from writing it. You might even want to read it
in the bathtub.
My thanks go to my sister Emily, who contributed the

subtitle. I would also like to thank my agent Betty Marks
and editor Paul Heacock. I am also indebted to Miriam
Smith and Mary Secor and the other mem bers of the
Onteora staff who were instrumental in the creation and
writing of Eureka!
7
Errata __________________________________
1. There are no mistakes in this book. The above should

read "ERRATUM."
Is the correction a mistake or not?

Contents
1l~ Fun and Games
1. Twisted Topology 15
2. A Bag of Tricks and Treats 26
3. A Score of Games 36
4. The Magic's There 48
5. Rubiquity 60
~~ Nifty Numerics
6. Palindromesemordnilap 69
7. A Pole Vaulter 71
8. A Timely Switch 73
3)~ Fallacies and Logic
9. If This Is Not a Chapter, My Name Is

Raymond Smullyan 75
10. Thrice Befuddled 89
11. Better Mixed-Up Than Lost 92
11
~g And Even Dissection of Solids
12. Archimedes Anderson and the Case of the

Sinister Plot 93
13. How to Dissect a Square and Other Marvels
of Modern Biology 96
14. Geometer's Heaven 105
15. Hole in the Sphere 108
16. Convexstasy 110
17. Great Unsolved Problems 113
18. Out of This World 115
@g Photons Are Light Matter, Too
19. Archimedes Anderson and the Gambling

Candidate 119
20. Once Upon a Time. . . 122
21. A Problem Fly 123
22. The Leading Series of Pisa 124
23. The Early Something Catches the Whatever 127
®g Shortcuts
24. A Speedier System of Solving 129

25. A Letter Home 135
26. In Which We Are Initiated Into the Secret
Society of Square Root Solvers 137
27. Heads and Legs 141
28. Noble Bases 143
29. A Division in Ancient Rye 145
30. Getting at the Root of the Problem 146
31. Or is it 32? Remumbt:r Nembers 149
12 EUREKA!
'1h Neat Numbers
32. Prime Time 151

33. A Sense of Balance 159
34. Perfect Numbers and Some Not-So-Perfect
Numbers 161
f8)~ Cranium Crackers and Cheese:

Problems to Munch On
35. Classy Problems 165

36. LEITERS + DIGITS = FRUSTRATION 174
~~ FUNdamental Ratios
37. Expand Your Mind 181

38. E? Ah! 183
39. A Section of Gold 188
40. A Bundled-Up Buyer 195
41. A Piece of Pi 196
Bibliography 202
13
110 Fun and Games
1. Twisted Topology _ _ _ _ _ _ _ __
A limerick writer once noted:

A mathematician confided
That a Mobius band is one-sided.
And you'll get quite a laugh
If you cut one in half,
For it stays in one piece when divided!
The Mobius band referred to is one of the most curious
shapes in all topology. Introduced by one August Ferdinand
Mobius, a German mathematician, the Mobius band or
Mobius strip has been described as a strip which has no
"other side."
15
While this oddity is difficult to imagine, it is surprisingly
easy to construct. Merely cut out a strip of paper at least
one inch wide and eight inches long. Mark one long edge
with an A in the left corner and a B in the right corner.
Then mark the parallel edge B in the left corner and A in
the right corner. Tape the short edges together so that one
A is above the other A, and one B is over the other B.
You've made a ring. This may not seem like much, but the
best is yet to come.
Now make a mark on one side of the paper. (It doesn't
matter which side, as you'll see in a moment.) Using a
finger, trace around the ring. Soon your finger and your
mark will be on opposite sides of the paper, and soon
after that your finger will be right on top of the mark!
You have just proved that the Mobius strip has one and
only one side (by showing the "two" sides are actually
joined). "Balderdash!" you say. "Everyone knows that a
piece of paper has two sides."
Not the Mobius strip. In a mathematician's terms, the
band has only one side and one edge. (The Goodrich com-
pany has patented the use of the Mobius strip as a conveyor
belt; both sides being one, it lasts twice as long as conven-
tional ones.) You may think that you see two of each, but
I can assure you that you really don't; they are one and
the same.
Even odder is the double Mobius strip, formed by putting
two strips of paper together, one on top of the other,
giving both a half-twist, and joining the ends. If you now
make a mark on the inside of the outside band and begin
circling the two, you will find that the mark has jumped
to the opposite band! A bug crawling between the bands
would see your mark alternately on the ceiling and the
floor; it would need considerable imagination to compre-
hend that the floor and the ceiling are actually on one
16 EUREKA!
strip. What appears to be two nested strips is actually one
large one!
Even more unusual is another property of the Mobius
band mentioned in the opening limerick. If you cut a
Mobius band down its "middle" (it really has none to
speak of), you will not get two Mobius strips but a single,
large, two-sided strip. Odder is cutting one in thirds;
produced is a large, two-sided strip intertwined with a new,
smaller Mobius strip. Try, for instance, cutting a Mobius
band in fifths. The results are guaranteed to be surprising!
QUICKIE
1. If Fig. 1-1 is cut along the middle, what will the

result be? Will the loop(s) be linked or not? See the
Answers section at the end of the Chapter.
Fig. 1·1
If you were to join two Mobius strips that are mirror

images of each other, you would get another unusual
topological oddity, the Klein bottle. Named after one
Felix Klein, another German mathematician, it too has
17
inspired a limerick: "A mathematician named Klein/
Thought the Mobius band was divine.! Said he, 'If you glue/
The edges of two,! You'll get a weird bottle like mine.' "
The Klein bottle, like the one-sided Mobius band, has
only one surface and precisely zero edges! It also has no
inside!
To construct a Klein bottle, bend the end of a tube of
material and pass it through its own wall. Then join the
two ends, forming a continuous closed surface (Fig. 1-2).
Voihl!
Fig. 1-2
Normally, the wall of a three-dimensional object must

be penetrated if one is to reach the other "side." Not so
with the Klein bottle. Merely by entering the opening and
staying on the "outside," the "inside" is reached. This
18 EUREKA!
means that the outside is the inside, proving that it has
one surface only.
Are there any practical applications of the Klein bottle?
So far there are none; liquids would spill from it rather
easily. But if a cap is put over the opening, the bottle
serves as its own handle! This area is one that should cer-
tainly be explored.
The problem that started topology all began in the small
Russian town of Konigsberg, a quaint little village cut
into four parts by the river Pregel. In summer, the towns-
folk liked to take their evening strolls across the seven
bridges (Fig. 1-3).
,~ --
-
Fig. 1-3
Much to their surprise, though, they discovered that

they could not cross all the bridges once in a single stroll
without crossing a bridge twice or retracing their steps.
Try it yourself and see.
By the time the problem reached the ears of Leonhard
Euler, the great 18th century Swiss mathematician, they
had ruined it by adding an eighth bridge and changing
the town's name to Kaliningrad. Nevertheless, he drew
the basic network by cutting out everything shown in
19
Fig. 1-3, except the actual routes. If he had drawn the
network in one stroke, it would have been equivalent to
strolling on the bridges; however, Euler found that the
network was not a one-stroker, so a seven-bridge walk was
impossible.
He counted the number of lines leading into each dot
representing each part of the town, and called the dots
odd if the number was odd and even if the number was
even. His final conclusion? A network is a one-stroker
if all the dots are even or if only two of the dots are odd
(if so, the stroll or stroke must begin on an odd dot).
2. Are Figs. 1-4 and 1-5 one-strokers? Don't actually

draw them!
Fig. 1·4 Fig. 1·5
3. Of a slightly different nature is the topological

puzzle in Fig. 1-6. The problem is to draw a continuous
line so that each of the 16 segments is crossed once and
only once, without drawing through the vertices or along
20 EUREKA!
the line segments. Can it be done on either a plane, a
sphere, or a torus (a bagel-shaped solid)?
Fig. 1·6
4. A well-known brewery uses as its trademark three

linked rings (Fig. 1-7)-or are they linked at all? If one
single ring is cut, the set falls apart. If you were the King's
armorer, could you make him an entire tunic of chain mail
using this pattern? How will you place the next ring?
Fig. 1·7
21
At the other end of the spectrum of recreational math
are problems of a class of deceptively simple cranium
crackers. (An example: I know a math teacher who can
remove his vest without first taking off his coat! Of course,
he does look a little funny when he does it in restau-
rants .... )
5. Can you separate the cup and the string in Fig. 1-8
without cutting the string, undoing the knot, or breaking
the handle? Think!
Fig.l.8
6. This last one I find absolutely perfect at parties and

as a tool-of-the-trade for would·be matchmakers. It's also
good for budding escape artists. Loosely tie two sets of
wrists as shown in Fig. 1-9. If the participants don't cut
or unknot the ropes (amputation is out of the question),
how in the world can they get free?
22 EUREKA!
Fig. 1·9
Answers
1. The result will be two unlinked loops like the first,

except for the fact that one is the mirror image of the
other. Because of the crossing of the original, the two
halves have opposite twists (this is also why they cannot
be linked).
2. The first has two odd vertices, so it can be drawn by

starting at one of the bottom corners. The second has four
odd vertices, and so is not a one·stroker.
3. A line that enters and leaves a space must cross two

line segments. Spaces A, B, and C in Fig. 1·10 are each
surrounded by an odd number of segments (five), so it
follows that the ends of the line segment must be in each;
clearly, as a line has only two ends, this is impossible.
Fig. 1·10
23
The same reasoning applies to a sphere and the side of
a torus-with one exception that makes the problem
solvable. If the hole is situated inside either A, B, or C,
the puzzle is reduced to mere line-drawing, as in Fig. 1-11.
Fig. 1-11
4. The arrangement is shown in Fig. 1-12. The pattern can

be continued indefinitely.
24 EUREKA!
5. For some odd reason, this one seems to baffle people.
Just take the center loop of the string and pull it back
until you have a reasonable amount of slack. Then pull it
forward around the side of the cup, and let go.
6. It looks hard, but as illustrated in Fig. 1-13, release

is possible by pulling the center of the rope over one wrist
and through the loop.
Fig. 1-13
25
2. A Bag of Tricks and Treats _ _ _ __
Many numbers and operations of mathematics have special

properties that enable them to be easily adapted for mathe-
matical "tricks" or stunts. Most everyone has heard of
something akin to the following: Think of a number and
mUltiply it by 3. Now add 6, divide by 3, and subtract
your first number. What is the result?
Chances are that you vaguely suspect how the stunt
works. Indeed, it seems the explanation would be lengthy,
but it's really only a case of cancellation of fractions.
The instructions create the statement (3N + 6)/3 - N,
where N is the number you thought of. This reduces to
(N + 2) - N, or more simply, 2 in all cases.
1. How about this one? Think of a number and mUltiply

it by 4. Then add a second, even number and divide the
sum by 2. Now subtract half of your second number and
again divide by 2. The result is always your first number!
How does this work?
2. One clever twist on the "think of a number" trick

works extremely well with large numbers of people. At
first it looks easy, but it's really very hard to figure out,
and even starts to seem impossible! Collar a half-dozen
friends and coax each into picking a number between 51
and 100. You yourself-the prestidigitato~elect a num-
ber from 1 to 50 and seal that number in a convenient hat
or envelope (or make a copy of the number for each
friend). Subtract your selected number from 99 and say
26 EUREKA!
the result aloud; then tell your audience to add that
number to each of their selected numbers, cross out the
first digit, add that digit to the result, and subtract the
answer from the original number. The result is always
identical to your sealed number! Can you figure out how
this works?
Another puzzler that works especially well with large
crowds is based on the factorization of a certain number.
Find four people, say, AI, Beth, Cathy, and Dan. Ask Al
to select any three-digit number and write it twice to form
a six-digit number. Tell him to pass this number to Beth,
and ask her to divide this number by 7 (actually, as you
can use the successive primes 7, 11, and 13 in any order,
you can ask Beth to select a prime between 6 and 16 and
then ask Cathy and Dan to each select different primes in
the same range). "Don't worry about the remainder," you
say. "There won't be any." The quotient is passed to
Cathy, who is cordially requested to divide the number by
11, and the new quotient is divided by 13 by Dan, who
passes the number back to AI. "Look at that," you say,
"and you will find your original number!"
Again you are correct! This trick is based on the product
of 7, 11, and 13, the number 1001. Writing a three-digit
number twice is the same as multiplying it by 1001.
Dividing by the factors of 1001 yields the original three-
digit number.
3. If you are older than 9 and younger than 100, write

your age thrice to form a six-digit number. Can you prove
that this number is divisible by 7 and 13? Is the same true
if the number is written only twice?
One other favorite is asking a friend to select a digit

from 1 to 6. Multiplying this digit by 142,857 results in
the same digits in cyclic order! Or, you can ask her to
27
select a digit from 1 to 9, multiply that digit by 9, and
multiply the product by 12,345,679; you can also ask her
to multiply the digit by 7 and then by 15,873. The results
should surprise you!
You can impress your peers with your incredible ability
as a lightning calculator with this one. Ask the nearest
person to write down an excessively large number. In
seconds you can tell him if it's divisible by 11 or not. The
method? Sum up the digits in the even-numbered places
and in the odd. If the difference between the two sums is
o or any mUltiple of 11, the number is a multiple of 11.
4. Here's another way. Ask a buddy to write down two
numbers and write their sum below them. Then ask her
to write the sum of the last two numbers below that, and
so on, until there are ten numbers, each the sum of the
previous two. Then announce the sum. For example:
2
5
7
12
19
31
50
81
131
+ 212
Merely by glancing at the list, you can tell her the total is
550. How? There's a simple rule.
5. A reputation as a mystic can be strengthened by these
magical experiments with dice. With your back turned, ask
your subject to set up three dice and add the top faces. He
28 EUREKA!
then chooses one die, adds in the number on the bottom,
and then adds in the number shown on the top, remember-
·
ing the grand total of the five faces. Hint: The magician
says the total in Fig. 2-1 is 18. How does he know?
U •
•
•
•
I
•
• •
Fig. 2·1
I
I
I
I
/ ...
...
•
•
/
·V
/
I
6. Here's a harder trick using the same principle. Ask

your subject to stack three dice in a pile and add up the
spots on the hidden faces (this means "not the top face,"
but don't say this!). The sum in Fig. 2-1 is 16. How do
you figure it out?
/ ...../
• • I
• • /
I
I
•
II
• V
I
• • I
• I
• • V Fig. 2·2
7. Ask the nearest mathematician about this one: Tell

your friend or enemy to arrange two dice right next to
each other to form a two-digit number (reading from left
29
to right). Then ask her to form a similar two-digit number
by flipping both dice over .
•• • • • ••
•
•
• • •
• • • •
Fig. 2·3
In Fig. 2-3, the two numbers are 35 and 42. Request
now that she join the two to form a four-digit number (as
in 3542), divide by 11, and tell you the result. You can
tell her the original arrangement of the dice by subtract-
ing 7 and dividing by 9. Can you explain how this works?
8. Some stunts just take a little thought to unravel. If

someone says to you, "I bet you a dollar that if you give
me two dollars I'll give you four in exchange," would that
be a good bet to take?
9. Try this one on the nearest unwitting stooge. Ask

him for a dollar. Then put it, along with a crisp, new,
irresistible bill of your own, into an empty matchbox.
Say the appropriate mumbo-jumbo and reluctantly offer
the box and its contents for $1.50. Would you yourself
fall for it?
Un losable bets like these abound. Have someone place a

dollar bill on the table and cover it with any card drawn
from a deck (assume for the moment that it's the ace of
spades). Now bet the "sucker" 50 cents that he will not
answer the value of the card to each of your next three
questions. Your first two questions should be totally irrel-
30 EUREKA!
evant; but make sure you receive two answers of, "the ace
of spades." Now ask him what he will take for the bill
below the card! Either way he loses four bits.
10. Others are somewhat more mathematical. These
deal with properties of even and odd numbers.
Coerce a handy millionaire into this wager: You and
she will each put down a single die, separately. If the total
of both dice and her number are both odd or both even,
she wins. Otherwise, you take the pot. What's the catch?
Take a full deck of cards and ask a nimble card shark to

cut the deck into seven piles. Bet that there will be an odd
number of piles containing an even number of cards. You
win! Now you can take away or add a card and adjust the
bet so that there will be an odd number of piles containing
an odd number of cards. Again you win!
DID YOU KNOW THAT ...

Eight perfect shuffles, in which the cards from each
half of the deck wind up in alternating order in the
shuffled pack, restores the deck?
11. A real mathematician named Walter Penney came

up with the following paradoxical bet. Given the situation
where you are tossing a coin three times in succession, it
is possible to select one combination of heads and tails
which will be most likely to appear first.
All you need to do is drop the third of the bettor's calls
and join the opposite of the second call to the front. If
your mark called HHT, for example, you would transform
it from HHT to HH to THH for your call. The odds are
always in your favor! (For precise odds, see the Answers
section at the end of this Chapter.)
One last stunt involves coin-shunting, this one the
31
changing of two rows of coins to a hexagon. Form the two
rows in Fig. 2-4. Now move coin 4 so that it touches 5
and 6, move 5 to touch 1 and 2, and move 1 to touch 5
and 4. Finally, arrange the coins with the top row to the
right (Fig. 2-5), and challenge your opponent to duplicate
what you have done in the same number of moves.
Fig. 2·4 Fig. 2·5
12. Lastly, here are some puzzles to confuse and con-

found:
a. You arrange 12 cards, from ace to queen, in a circle.
Ask your friend to think of one, but not to tell you which
card he has selected. Inform him that you will determine
his card by tapping on the cards around the ring. He should
count your first tap as 1 above his chosen number, the
second 2 above, and so on. When his count reaches 20, he
should tell you to stop, and you will both be surprised to
find that you are right on top of his card-if you started
counting on the right card. What card is this?
b. Rearrange 10 of those 12 cards face down to form a
box 3 cards by 4 cards, with the center 2 spaces empty.
The problem here is to select any face-down card, jump 2
cards in a clockwise manner, and turn over the fourth.
Continue doing so until 9 cards are face up. On which
card should you start, and what method should you use?
c. Draw the star in Fig. 2-6 in one line and fill in the
numbers. The puzzle here is to place a coin on any inter-
section, jump one intersection, and cement the coin to
32 EUREKA!
the third intersection, always starting and ending on empty
spaces, until nine spaces are covered. How can you do this
so that a move is always possible and so that the puzzle
can be solved? Happy shunting!
2
1 ~----+----r----"7 3
5 4
Fig. 2-6
Answers
1. If your numbers are N and X:

4N+X 1 X X
- - - - -X 2N+---
2 2 2 2 2N
-----=-=N
2 2 2
2. If X is your friend's number and y yours, the algebraic
expression is:
X - [99 - y) + X - 100 + 1 ]
X - (99 - y + X - 99)
X - (X - y) = y
Note that the sum of two numbers less than 100 cannot be
more than 198, and that the restrictions on the picking of
33
the numbers means that the sum must be more than 100, so
the first digit of the sum of the friend must be 1.
Others along this line: You can also ask a friend to
throw two dice until he has thrown two different numbers.
Then beg him to double the number on one of the dice,
add 5, multiply by 5, and add the number of spots on the
other die to the product. If you are now told the result,
you can determine the two numbers by subtracting 25 and
separating the result into two digits. Again mathematics
triumphs!
Another stunt concerns casting out 9s, or so it seems.
Ask a friend to write down a three-digit number and sub-
tract from it the sum of the digits. Tell him to cross out
any digit and tell you the sum of those remaining. Merely
by subtracting from the next highest mUltiple of 9, you
can tell him the digit he crossed out.
These are rather easy to devise, and you can no doubt
create grander illusions yourself.
3. A number such as 141,414 is divisible by 10,101. As

10,101 is divisible by 7 and 13, 141,414 is, too.
A number such as 1414 is divisible by 101; as 101 is
prime, this number must be divisible only by 101 and 1.
4. Merely note the fourth number from the bottom,

here 50. Multiplying this by 11 gives the sum. Or multiply
the first number by 55 and the second by 88 and add the
products.
5. The sum of the top faces is 11. The sum of the top
face and bottom face of any die is 7, and 7 + 11 = 18.
6. Just subtract the top face from 21, so the answer is

16. For N dice, the sum is 7N - (dots on top face).
34 EUREKA!
7. In the example, 3542/11 = 322; 322 - 7 = 315;
31519 = 35. Generally, for faces A and B:
([1000A + 100B + 10(7 - A) + (7 - B) ]/11} - 7
----=:...---·-----'--9--'---'----'--"--~- = lOA + B
the original number.

S. No! He could take your $2, say, "I lose," and hand
you his $1. You win the bet, but lose a dollar.
9. I certainly hope not. The buyer will have purchased

his own dollar bill for 50 cents.
10. All you need to do to acquire untold sums of

money is put down an odd number. Adding an odd num-
ber changes an odd to an even, and an even to an odd.
Similarly, if you want to lose, you should put down an
even number.
11. The odds to the coin-tossing are as follows:

HHH to THH, 7 to 1
HHT to THH, 3 to 1
HTH to HHT, 2 to 1
HIT to HHT, 2 to 1
THH to 'ITH, 2 to 1
THT to 'ITH, 2 to 1
'ITH to H'IT, 3 to 1
TIT to H'IT, 7 to 1
12a. You should start on the 7 and count backward.
b. You can start anywhere, but each new start must
be three cards counterclockwise to the previous one.
c. The secret is to cement each coin to the spot
where the previous one started. For instance, you might
move from 3 to 10. The next coin must start at 7 and end
at 3.
35
3. A Score of Games _ _ _ _ _ _ _ __
The field of mathematical games is an especially rewarding

area of recreational mathematics. Even old, outworn games
can be of interest when introduced in new forms.
1. Take, for instance, tic-tac-toe, a game in which two

opponents place markers (classically Xs and as) on a
3 X 3 grid in an attempt to get three in a row. After an
hour's analysis and play, anybody can become an unbeat-
able master or force any game into a draw, just by learning
two or three simple rules (see Answers). Just one interesting
variation is toe-tac-tic, in which the idea is to force the
other player to get three in a row. How do you win here?
Tic-tac-toe on a larger scale is also intriguing. Go-moku,
played in the Orient thousands of years ago, is just such a
game. The idea is to place five counters in a line on a
19 X 19 board. Players take turns placing from an unlim-
ited supply of counters. Positions which force a win are
lines of four open at each end, and two open lines of three.
Though experts are of the opinion that the first player can
force a win, the game is nevertheless challenging enough
to be popular in all parts of the world.
Another way of increasing the size of the game is to add
more boards-one on top of the other-to form a cube.
Here, winning is getting a row on any layer or along any
of the lines. The number of possible winning lines, in fact,
is given by the formula:
(k + 2)" - k"
2
36 EUREKA!
where k is the number of cells on each side and n the num-
ber of dimensions. Curiously, the first player can't help
winning on the 1 X 1 X 1,2 X 2 X 2, or 3 X 3 X 3 boards.
On the 4 X 4 X 4 board, playing is a little more diffi-
cult. If you are interested in investigating the strategy
further, many manufactured versions are marketed; one
model for the crafty is illustrated in Fig. 3-1.
Fig. 3-1

25 X 9 2 = 2592?
2. Another exciting variation of tic-tac-toe is played

with moving counters, characteristically three to each
player. Turns alternate for placing and, if neither player
37
has yet won, moving the counters commences. Usually,
only moves along the orthogonals or perpendiculars are
permitted, but variations which allow moves along the
main diagonals (corner to corner) and/or the short diago-
nals are known. In a French version, a counter may be
moved to any empty cell.
Assuming moves along the short diagonals are not
allowed, how can the first player force a win?

The average of 4 and 5 is 4.5? This is unique!
Safragat, originating in the Sudan, is another favorite for

rainy days. Here four counters are used by each player,
and the object of the game is different. Placing is as normal,
but the play brings a catch-if a player moves so that she
traps her opponent's piece between two of her own, that
piece is removed from the board. (However, moving a
piece between two belonging to the opponent is perfectly
safe.) A player loses if she can't move or if she has taken
fewer pieces from the board at the end of the game, when
no more captures are possible. Try playing on a 5 X 5
board (the game here is known as Sipu) or even a 6 X 6
board with two unoccupied spaces.
Teeko, invented by John Scarne, is a deceptively simple
game that's easier to play when watching. Players take
turns placing and then moving four counters on a 5 X 5
board. Winning positions are four in any line, or four in
a box, as shown in Fig. 3-2.
38 EUREKA!
.~
""'-
0
4~
-- -
- - - .--
Fig. 3·2
Altogether there are 44 different winning positions, or 58

if any size box is allowed. I strongly recommend Teeko.
Nim is another favorite mathematical game that has
appealed to many in a wide variety of forms. All, however,
concern two players who take objects from an arrangement
of piles according to certain rules with the intent of forcing
the opponent to take the last object, although sometimes
this last part is played in reverse.
The simplest form of nim (supposedly from an old
English verb meaning "to take away") is played with a
single pile of counters. Here, players alternate and take
from 1 to m counters, where m is previously determined.
For those who wish to win, make sure your opponent
must take from a pile of n(m + 1) + 1 counters, where
n is any number (and if there are no objects handy, use
digits on paper; try to keep from landing on or over 100,
for example). This formula was successfully used by a
child swindler in John D. Fitzgerald's novel, The Great
Brain Reforms; the swindler was a brilliant youth who
claimed to hypnotize tin cans so that his opponent would
always pick the last one.
39
3. In its more advanced state, nim is played with sev-
eral piles of counters, classically 3, 4, and 5, or 3, 5, and 7.
Players can take up to all of the coins in any single row,
the purpose being to force the opponent to remove the last
counter from the board.
A startling discovery, made around the turn of the cen-
tury, showed that nim could be generalized into a game
with any number of piles and any number of counters per
row. A simple strategy enables anyone to playa perfect
game.
To determine if a position is safe or unsafe (no chance
of immediately losing or a chance of losing), simply add
up the binary values of each row. Binary, which is base
two, is nothing more than the writing of numbers using
powers of 2. In the 3, 4, 5 game, the 3 is equal to one 2
and one 1, so its base two representation is 11; 4 is equal
to one 4, zero 2s, and zero 1s, so its binary representation
is 100; 5 is then equal to 101 base two.
Thusly, the sum of the binary representations of 3, 4,
and 5 (equal to 11, 100, and 101) is 212. As this sum con-
tains an odd digit, the position is unsafe. To achieve safe-
ness, all you need to do is get rid of the 1 in the center
column, making all the digits in the binary sum even. This
can be achieved by taking 2 counters from the 3-pile.
That would leave 1 in the 3-pile, and 1 + 100 + 101 = 202,
a safe position. Is starting with rows of 2, 3,4, and 5 safe?
How would you make a starting position of 4, 5, 6 safe?
For playing purposes, you can use your fingers for the
powers of 2 up to 1024. Raise a finger for an odd number
(that is, a 1) and lower it when it becomes even. The posi-
tion can be determined to be safe or unsafe by adding
each number over the others this way, and determining
what you need to do to lower all of your fingers.
40 EUREKA!
In one fascinating variant of nim, players can take
from any number of rows up to k. Surprisingly, the same
binary analysis still holds; a safe position is one in which
every column of numbers totals a digit evenly divisible by
(k + 1). Nim follows this rule; players take from one row,
so the binary representations must be divisible by 2 to be
safe.
A game in its own right, Even and Odd also concerns
taking counters from a pile, but the object is to end with
an odd amount of counters at the end of the game.
If m is the maximum number of counters allowed and n
any integer, then the player with an odd amount of
counters loses if he must draw from 2n(m + 1) + 1 or
2n(m + 1) + 1 + m counters.
A player with an even amount of counters likewise loses
if he must draw from 2n(m + 1) or 2n(m + 1) + 2 + m
counters.
Another game similar to nim is the Australian game of
31, played with 24 cards, 4 each of values 1 through 6.
The idea is to turn over cards alternately, winning if the
score is 31 and losing if it is more than that. The key is
to land on the numbers 3, 10, 17, 24, and 31-but there
is a catch. Consider the following sequence: You open
with a 3, and your opponent follows up with a 3. You
turn a 4, reaching 10, and he takes a 4 also. You then
take a 3, and your opponent follows with a 3, followed by
another set of 4s, bringing the total to 28. When you
victoriously reach for a 3 to bring the total to 31, you will
find there are none to be found. Beware of this!
4. Another variant of nim is Daisy, played with a ring

of 13 petals, the players plucking 1 or 2 adjacent petals
at a time, the winner plucking the last. How can the sec-
ond player always win?
41
The Tac Tix board, from a game invented by Piet Hein,
is shown in Fig. 3-3. Like multirow nim, a player can take
up to 4 adjacent counters in any row in any direction, the
loser taking the last counter. It is unknown which player
can force a win, though the first player can always win on
odd-ordered boards (ones with an odd number of counters
on a side) by taking the center counter. You might like to
try playing on a 6 X 6 board.
I I
000 0
o 0 0 0
- 0 0 0 0
o 0 0 0 -
Fig. 3.3
QUICKIE
5. Two sisters decide to split a cake, 20 inches in cir-

cumference, between them. Jane suggests that they cut
slices of 1 or 2 inches in turn, and Cynthia replies with
the proposal that the last to cut should wash the dishes.
If Cynthia cuts first, what size piece should she take to
avoid washing?
Another realm of mathematical games is connecting

games. One rather prominent one, also invented by Piet
Hein, is called Hex.
42 EUREKA!
Fig. 3-4
Fig. 3-5
Played on diamond-shaped boards like the two in Figs. 3-4

and 3-5, the idea of Hex is to connect adjacent hexagons
(or linked dots on a board of triangles) so that an unbroken
43
line extends from one side of the board to the opposite
side. The corners can be used by either player.
To learn the strategies of Hex, play games on boards of
increasing size. The first player can win on the 2 X 2 board,
on the 3 X 3 by playing in the center, and on the 4 X 4
by playing in the four middle spaces (if the move is any-
where else, the second player can force a win). This
advantage of the low-order boards carries over to the
11 X 11; the first player has a slight advantage, but the
edge is very easy to lose. One misplaced play can give the
game to the second player. Of course, it does work in
reverse, and can be even more disastrous for the second
player. Try it and see.
David Gale has devised a delightful dot-connecting game
called Bridg-It, or Gale. It is played on a board which con-
sists of two interlocking grids, each 5 X 6 dots (though
you may prefer to play on a larger board), as shown in
Fig. 3-6. One player joins the boxes, and the other joins
the dots. The idea is to connect one side of the grid to the
other along the long axis without crossing lines. This can
never end in a draw. Though the first player on any size
board has the advantage, the game is nevertheless chal-
lenging and recommended.
The game of Boxes is probably familiar to you. A set of
16 or 36 dots is set out in a square. The players alternate
in connecting any two adjacent dots. A box is completed
if a player draws the fourth side; that player gets credit for
the box and takes another turn. The winner is the player
with the majority of boxes at the end, when all the boxes
are accounted for.
Boxes lends itself easily to gambits, one player giving
up a box or line of boxes in exchange for more later. One
particularly interesting variant of Boxes is Triangles,
played on a board the shape of an equilateral triangle. The
44 EUREKA!
• • • • •
o o D D o D
• • • • •
o o o o o o
• • • • •
o o D o o o
• • • • •
o o D o o o
• • • • •
o o o o o o
• • • • •
Fig. 3·6
purpose is to amass triangles. You might want to compare

the skill needed for each.
One other classic math game is Mastermind, simple in
principle but hard in practice. One player arranges four of
six possible colored pegs (or letters or numbers in place of
the pegs) in a line behind a small hood. The second player,
the "code·breaker," arranges another four of the six on a
different part of the board in an attempt to determine the
arrangement of the hidden four, the "code." The first
player then compares the guess to the actual code. A cor·
rect color in the correct row receives a black peg; a correct
color in the wrong row receives a white peg. For instance,
the second player could receive two blacks and one white,
meaning that he had three colors correct, but only two
were also in the right row. A set of four blacks means that
the second player has deduced the code.
45
Mastermind can be played with a code containing more
than one of each color, with four of eight colors or six of
nine, with digits, or with entire words.
Try to invent your own games. Use such common objects
as cards or dominoes, or play your games on undrawable
hypercubes. The selection here is just a small sampling.
Answers
1. Tic-tac-toe: There are only three opening moves, as

shown in Figs. 3-7 through 3-9. The dotted Os are the
necessary responses to each X move. Any of the four
positions in Figs. 3-8 and 3-9, followed by rational playing,
will lead to a draw.
r"\ r\ r\
... J ... .1 \..1
r"\
r"\
\...1 X \..1
r"\
X r"\
... .1
r"\
\...1
r"\
\..1 X \..1
Fig. 3-7 Fig. 3-8 Fig. 3-9
Toe-tac-tic: The second player has a decided advantage.

The first player can force a draw only by first playing in
the center and then playing opposite the second player.
2. The first player, to win, should place in the center.

If you number the spaces as shown in Fig. 3-10, two lines
of play follow:
46 EUREKA!
1 2 3
4 5 6
7 8 9
Fig. 3·10
a.X 0
5 3
4 6
9 1
4 to 7 any move
5 to 8
b. 5 6
1 9
3 2
1 to 4 any move
4 to 7
Note that both hold even if moves along the major diag·
onals are permitted.
3. The nim position 2, 3, 4, 5 is already safe. The
position 4, 5, 6 has binary values of 100, 101, and 110,
adding to 313. A safe move is removing 1 from the 4·pile,
3 from the 5·pile, or 5 from the 6·pile.
4. The second player should always keep the position
symmetrical, taking petals directly opposite the first.
5. Cynthia should cut a l·inch piece. Her next cut
should bring the total of cut cake to 4 inches, and then
7, 10, 13, 16, and 19 inches. Jane must then wash the
dishes, but she's probably had 13 inches of cake.
47
4. The Magic's There _ _ _ _ _ _ __
Magic squares to me are more magic than square. It seems

incredible that the numbers from 1 to 16, say, can be
arranged in a 4 X 4 square so that the sum of the numbers
along any horizontal line, vertical line, or long diagonal
is a constant, or that the numbers from 1 to 100 can also
be so arranged.
The simplest magic square is the 3 X 3, termed the
order-3 square (Fig. 4-1). In China, this square is called the
lo-shu. It can be found on amulets, perhaps because the
lo-shu is the only order-3 square possible. The seven others
that can be formed are either rotations or reflections of it.
8 3 4
1 5 9
6 7 2
Fig. 4-1
There are variations on the square, however. An order-3

square composed wholly of odd numbers can be formed
by substituting the 5, say, with the fifth odd number (Fig.
4-2). The even square (Fig. 4-3) can be formed the same
way.
48 EUREKA!
15 5 7 14 4 6
1 9 17 0 8 16
11 13 3 10 12 2
Fig. 4·2 Fig. 4·3
There is only one order·3 square, but there are two anti·
magic squares. In these oddities (shown in Fig. 4.4), no two
sums are the same. These are rookwise antimagic squares;
that is, you can connect the numbers by moving like a
chess rook, either vertically or horizontally.
9 8 7
2 1 6
3 4 5
Fig. 4.4
QUICKIE
Warren Peace was thumbing through a copy of Eureka!
one day, when he encountered a rather intriguing problem.
The diagram showed a 3 X 3 box with a lone 8 in it (Fig.
4·5). "Place a different number in each of the remaining
cells so that each line of 3 adds up to 15," said the book.
49
Warren did it. Can you? See the answer at the end of this
Chapter.
Fig. 4·5
There are 880 order-4 squares. The simplest one can be

constructed by overlapping two squares (Fig. 4-6), but it's
much easier to write the numbers 1 to 16 in order in a
square and switch the 1 with 16,6 with 11,4 with 13, and
7 with 10. Voila!
4 2 3 1 12 0 0 12 16 2 3 13
1 3 2 4 4 8 8 4 5 11 10 8
1 3 2 4 + 8 4 4 8 9 7 6 12
4 2 3 1 0 12 12 0 4 14 15 1
Fig. 4-6
Interestingly, the square is still magic if the middle two

columns are switched. What's produced is shown in Fig.
4-7. This square appears in Albrecht Durer's engraving,
Melancholie, and the date of the engraving, 1514, appears
in the bottom row. The magic constant appears here more
times than you can shake a stick at: All lines add to 34;
the four corners sum to 34; the five 2 X 2 squares, in the
50 EUREKA!
corners and in the center, all add up to 34; the line 3 + 5
and its opposite, 12 + 14, sum to 34, as do 2 + 8 and
9 + 15, and the sums of the squares of the pairs also match.
16 3 2 13
5 10 11 8
9 6 7 12
4 15 14 1
Fig. 4·7
A small rearrangement of the cells yields the square in

Fig. 4-8. This square has all the properties of DOrer's
square. In addition, any box of four sums to 34, and a
"broken diagonal," such as 3, 5, 14, 12, also sums to 34.
15 10 3 6
4 5 16 9
14 11 2 7
1 8 13 12
Fig. 4·8
It is this last property of the square that makes it a

diabolic (or panmagic, or pandiagonal, or Nasik) square.
Almost half of all order-4 squares are diabolic. These
51
squares are the very devil-they don't change no matter
what is done to them. They remain diabolic if they are
rotated, reflected, rearranged so that the top row becomes
the bottom, rearranged so that the left-hand column
becomes the right-hand, or rearranged according to a
mystic pattern handed down from mentor to pupil through
the years that involves switching the cells in cycles of three.
So far the magic constant has been determined by trial
and error. There is a formula, though, that gives the con-
stant and thus helps in forming squares: (n 3 + n)/2.1f n is
3, the expression gives 15, for example.
The constant for order-5 squares if 65; knowing that
fact is important, since there are more than 13 million
order-5 squares. Of these, about 28,800 are diabolic. They
can be formed by permuting four nonequivalent squares.
The square in Fig. 4-9 is symmetrical as well as diabolic.
Every pair of squares opposite around the center adds up
to 26, twice the center. The lo-shu, by the way, is also
symmetrical.
10 18 1 14 22
11 24 7 20 3
17 5 13 4 9
23 6 19 2 15
4 12 25 8 16
Fig. 4-9
52 EUREKA!
Unfortunately, diabolic squares are not possible in even-
order squares not divisible by 4, though normal squares,
of order-6, say, can be constructed. Order-8 squares can be
constructed, too. Figure 4-10 is an odd one.
1 3 123 121 119 117 13 15
17 19 107 105 103 101 29 31
95 93 37 39 41 43 83 81
79 77 53 55 57 59 67 65
63 61 69 71 73 75 51 49
47 45 85 87 89 91 35 33
97 99 27 25 23 21 109 111
113 115 11 9 7 5 125 127
Fig. 4·10
Ben Franklin constructed the square in Fig. 4-11, in

which each row adds up to 260, and each half-row adds up
to 130. The sum of any box of four is 130, as is the sum of
53
any four numbers equidistant from the center. Following
any full dotted diagonal up and down also gives 260.
Fig. 4·11
The 8 X 8 square resembles a chessboard, so somewhere

along the line someone asked himself if a knight, which
jumps two squares up or down or to the side and one per-
pendicularly, could cover each cell of a magic square in
order. Leonhard Euler created such a magic square with a
constant of 260; stopping halfway on each horizontal or
vertical line gives 130. Henry Dudeney created a magnifi-
cent square. Not only can the knight jump from the 64 to
the 1 to restart its tour, but the diagonals are also nearly
correct-256 and 264.
54 EUREKA!
These large squares are not as hard to create as you might
think. There is one very interesting method that produces
a concentric, or bordered, magic square. These squares stay
magic as their outside rows are repeatedly dissolved until
the core of an order-3 or order-4 square is reached. The
symmetrical order-5 square in Fig. 4-12 is bordered, as is
the order-6 square in Fig. 4-13.
19 2 20 1 23
4 16 9 14 22
18 11 13 15 8
21 12 17 10 5
3 24 6 25 7
Fig. 4-12
36 2 3 7 32 31
29 26 13 12 23 8
27 15 20 21 18 10
9 19 16 17 22 28
4 14 25 24 11 33
6 35 34 30 5 1
Fig. 4-13
55
So far these squares have been composed wholly of con-
secutive integers, odd integers, or even integers. Is it possi-
ble to create a magic square composed wholly of primes or
consecutive composites? Yes, it is, and these squares truly
are incredible. Dudeney tackled the composite square. He
found that in the interval 114 to 126, all the numbers are
composite, so he created the square in Fig. 4-14. His prime
square, oddly enough, sums to 111, the constant for order-
6 squares (Fig. 4-15). Bergholt and Shuldham topped him
with the simple square in Fig. 4-16. Prime squares have
been found past order-12. Once in that range, though, the
magic constants become huge.
121114119 67 1 43
1161181 13 37 61
31 73 7
Fig. 4·14 Fig. 4-15
3 71 5 23
53 11 37 1
17 13 41 31
29 7 19 47
Fig. 4·16
56 EUREKA!
An odd variant of the magic square is the magic cube,
consisting of n magic squares of order-n piled so that the
sum along any horizontal line, vertical line, or long diagonal
is the same. The simplest is the order-3 cube (Fig. 4-17),
but cubes of order-6 have been constructed. Fans of
Rubik's Cube may (but I doubt will) enjoy placing these
numbers on the cube in place of the colors and solving it
by solving the magic cube.
10 24 8 26 1 15 6 17 19
23 7 13 3 14 25 16 21 5
9 11 22 13 27 2 20 4 18
Fig. 4-17
Other interesting variations on the magic square are the

multiplication, division, addition, and subtraction squares.
In the multiplication square in Fig. 4-18, every line multi-
plies to 216. If the diagonal corners are switched, it be-
comes a division square, in which the product of the end
cells divided by the middle always results in 6. The normal
magic square can be changed into a subtraction square by
switching the corners; the sum of the edges minus the
middle cell is always 5.
12 1 18
9 6 4
2 36 3
Fig. 4-18
57
The square can also be used to force certain numbers.
Let's say you wanted someone to select, perhaps for some
part of a larger trick, the number 43. The first thing to do
is to decide what order square to use, and then split the
43 into twice that number of parts. An order-4 square can
be used, so 43 is split into 1, 3, 4, 5, 6, 7, 8, 9. The num-
bers are arranged randomly on the outside of the square,
and their sums on the inside (Fig. 4-19).
+ 6 4 3 7
1 7 5 4 8
8 14 12 11 15
5 11 9 8 12
9 15 13 12 16
Fig. 4-19
Any number in the 4 X 4 square is circled, and all the others

in its horizontal and vertical rows are crossed out. When
four have been chosen, their sum is 43. This works because
each of the original parts is included in one of the chosen
numbers, so the sum has to be 43.
Magic squares are used more in this harder-to-figure-out
trick. Arrange the first nine cards of a suit in the order ace,
8,2, 7, 3,4,5,6,9, so that the ace is on top of the packet.
Have a spectator riffle these cards through the deck, and
then pull them out again. This should not disturb the order.
Now for a game of tic-tac-toe. Split the packet, faces up,
58 EUREKA!
so that your left hand holds the lower six cards and your
right the higher three. Place the 5 in the center of the board
and ask your opponent for his move. If it is a corner square,
put the right-hand cards on top of the left-hand cards;
otherwise, put them beneath. Now put the packet face
down on the table and ask the spectator to place the top
card in the square he picked. If it is a side square, ask him
to place your card in the adjacent corner. If he places his
first card in the corner, place it on the opposite side square.
The rest of the moves are forced. When all the cards are
overturned, a magic square is revealed! This is just one of
the many interesting things about and uses of magic
squares.
Answer
Figure 4-20 shows Warren's solution. Oddly enough, 8

is the number of lines on which the constant 15 can be
found.
4~ 8 2~
3 5 7
7~ 2 5~
Fig. 4-20
59
5. Rubiquity _ _ _ _ _ _ _ _ _ _ __
The Rubik's Cube is an innocent-looking, often maddening

puzzle that has caused millions of would-be solvers to go
quietly crazy. Sweeping the world is the rather pleasurable
disease of cubitis, a mysterious and highly contagious mal-
ady that is both caused and cured by contact with this
guaranteed hair-tearer, the separate brain children of both
Hungarian architecture teacher Erno Rubik and Japanese
engineer Terutoshi Ishige.
Each side of the 3 X 3 X 3 cube (Fig. 5-1)-which is
itself composed of 26 smaller cubies grouped around a
jack-like spindle-is originally a different color, but each
face can be twisted or twiddled so as to produce wild,
multicolored messes.
Fig. 5·1
The basic problem is to rearrange the cubies so that each

face is a solid color again. Those who haven't been stumped
60 EUREKA!
by Rubik's Cube can get a real appreciation of the skill
involved in solving one with the news that there are more
than 43 quintillion (or 43 X 10 18 ) possible arrangements
of the cubies-and only one is the solved position.
Here's how that discouragingly large number is achieved:
There are eight corner pieces. Each can be twisted three
ways. The first factor, then, is 38 . However, because one
good turn deserves another, and every negative twist has
a matching positive, when seven corners are oriented, the
eighth is fixed, so a factor of 3 is removed. The corners
can also be arranged in the eight cubicles in 8 X 7 X 6 X
5 X 4 X 3 X 2 X 1 ways for another factor of 8!. (The !
is a factorial sign. X! indicates the product of all positive
numbers less than X + 1. For example, 5! = 5 X 4 X 3 X
2 X 1 =120, and 8! = 8 X 7 X 6 X 5 X 4 X 3 X 2 X 1 =
40,320.) The edges can also be permuted in 12! different
ways, but when eleven are placed, the twelfth is fixed, so
a 2 is removed. Also, each cubie can be oriented in two
different ways, so a factor of 212 is produced. But it, too,
must be reduced by 2 for that last fixed cubie. The prod-
uct of all these numbers is 43,252,003,274,489,856,000.
If you were to look at one of these positions every second
without rest, it would take you more than 1 trillion years,
or more than 100 times the life of the universe! Easy
puzzle, eh?
Well, it can be. The key to solving it, usually within
minutes instead of millennia, is to separate the solving
process into stages, in which the numbers involved are
much smaller.
But I'm getting ahead of myself. The first thing to do is
to get a mathematical appreciation of the cube and a basis
for solving it. Part of the challenge in solving the cube is to
overcome the difficulty in analyzing it that results because
so many things can happen so quickly; it just so happens
61
that a branch of mathematics deals with the cube and what
can happen to the cubies-group theory. A series of moves
on the cube can basically be seen as a group.
A system of notation originated by David Singmaster in
Notes on Rubik's Cube helps describe the moves. If the
cube is placed flat on a desk, the upper face, whatever its
colors may be, is called V. The right-hand face is R, the
left is L, the front F, the back B, and the downward side
is D. A positive move is a clockwise move of that face; R
moves the right-hand face away from the solver, and L
moves the left toward the solver. A negative move is indi-
cated by a small slash: V' moves the top face counterclock-
wise.
A move of the back, upper, and right faces, all clockwise,
can be represented as BVR. Two moves of the downward
face are expressed as D2 , which is equal to (D')2. Notice
the power of conjugates: B' undoes B, and FD'R' undoes
RDF'. This notation is very useful in passing ideas. If you
have a solved cube (if you don't, then turn the top face
45 degrees, insert a screwdriver under one of the top
edges, pop it out, remove the others to get to the spindle, and
form the solved cube by placing the cubies back in their
appropriate spots-which is how I solved my first cube),
you might care to see what you get with RL'FB'VD'RL'
or (R2D 2V 2L 2V 2)D(L2F2B2L2F2B2 )D.
The notation can name every cubie. The edge between
the V and R faces is called ur, and the corners next to it
are urf and urb. This system comes in handy in the solu-
tion.
Here is the only labelled joke in this entire book:
Q: What is Erno Rubik's favorite theatrical play?

A: Rubik's favorite play is, of course, Karel Capek's
Rossum's Universal Robots-also called RUR!
62 EUREKA!
The cube can always be solved if it is in one of those 43
quintillion positions. If some joker were to put an edge in
backwards, though, one of eleven new universes of 43
quintillion more positions will be opened. Even in the
"worst" orbit, however, the cube can be solved except for
a twist in one corner and an exchange of two edges, one
of which is flipped.
Otherwise, the cube can be solved through the use of an
orderly system. Certain cubes are selected and placed cor-
rectly in this algorithm. What surprises many beginners is
the fact that the partially solved cube must be messed up
and restored several times along the way. This is a hard
lesson to learn.
My system, along with many others, uses AHA' moves.
Let's say you have a nice operator, a move that will
flip two edge cubes if they're in the right cubicles. Perform
the series of moves A (which could be B'DU) to get the
cubies in those two magic spots, perform your operator H,
and perform A' (which could be U'D'B) to get the cubies
back. AHA! Those edges are flipped!
The algorithm presented here proceeds according to
these stages: top center; top corners; top edges; equator
centers; bottom corners; equator edges; and bottom edges.
Top Center
This is the easiest stage. Just choose a nice color and put
its center on top. I usually start off with orange.
Top Corners
1. Simply get an orange corner label up next to the
center label. This requires at most a half-turn. Turn the
entire cube so that the corner is at ufI. (Ha! You didn't
believe me!) Two or more oranges should be on top now.
2. Note the color on the f side of ufl. There is only one
63
other corner cubie with labels of orange and this color,
and it goes at ufr. Find it. If that cubie is at ufr and the
orange is up, go straight to step 5.
3. If the cubie is in the top layer, but not correctly
placed, turn the entire cube so that it is at ufr. Perform
R'DR and return the cube to its original position.
4. Turn the cubie in question to rfd by moving the D
face. If the orange is on the R face, perform R'D'R to get
it on top. If it is on F, perform FDF ', and if it is on the
bottom, perform R'DRFD 2 F' or D 2 R'D'RFDF' .
5. The next cubie is now correct. If all four orange
corners are up and correctly placed (Fig. 5-2), go on to the
next stage. Otherwise, make a U-turn and treat the cubie
now at ufl as the one you originally picked . Go back to
step 2-please.
Fig. 5-2
Top Edges
1. The object of this stage is to complete the uppermost
layer, so that the top face is all orange and the trimmings
around the sides are solid blocks of color. If all four top
edges are correctly placed, colored, and oriented, go to
step 7.
64 EUREKA!
2. Turn the cube so that one edge cubie that is wrong
is at uf. Note the colors of the f faces on the corners at
fur and ful (they should be the same). Now locate the
edge cubie with labels of orange and this color.
3. If it is at uf but not correctly oriented, perform
FDU'R'U to get it into the middle layer, and go to step 5.
4. If the cubie is in the top layer but not correctly
placed, turn the cube so that it is at uf and perform FDU'
R'U and move the cube back.
5. Find that cubie. If it is in the middle layer, turn it
to fl. If the orange is at L, perform U'R'UD'F. Otherwise,
perform U'RU 2 D 2 L'. Go back to step 1.
6. If the cubie is in the bottom layer, turn it to fd. If
the orange is on the bottom, perform RL'F 2 LR'. Otherwise,
perform F'DU'R and go back to step 1.
7. Congratulate yourself. You've solved a layer-more
than one-third of the cube!
Equator Centers
Turn the middle layer so that the centers match the
trimmings of the top layer, as in Fig. 5-3.
Fig. 5-3
65
Bottom Comers-Position
1. Turn the cube so that the orange is on the bottom.
Choose a corner on top and correctly position it by turning
Us several times so that its colors match the equator
centers. Turn the entire cube so that the orange is at the
left and the cubie is at rdb.
2. The orange should now be at the left, and a correctly
placed and possibly correctly oriented corner should be at
rdb.
3. Examine the three other corners on the R face.
a. If they're all correctly placed, skip the rest of this
stage.
b. If they need to be switched clockwise in order to
make the equator centers match the colors, perform
(R'D'R)U(R'DR)U'.
c. If they need to be switched counterclockwise
perform U(R'D'R)U'(R'DR). This is the conjugate of the
above.
d. If two need to be switched diagonally, perform
D'R'F'RFDR'. If two next to each other need to be
switched, select one of them as your new base. Go back to
the start of step 3.
4. Don't read this step.
Bottom Comers-orientation
What started off as the bottom, the red (if red is opposite
orange) should now be turned to the top.
1. If all red labels are on top, go on to the next stage.
2. If no reds are up, perform (R'DR)(FDF')U(FD'F')
(R'D'R)U'. This should bring one to the top.
3. If only one red is up, turn it to ufI.
a. If there is a red at the f of fur, perform (F'LF)R'
(F'L'F)U2(R'D'R)U2(R'DR 2).
66 EUREKA!
b. If the red is at the r, perform (R 2 D'R)U 2 (R'DR)
U 2 (F'LF)R(F'L'F).
c. Go to the next stage.
4. If two reds are up, turn them to ulf and ulb. (If
they are diagonally opposite, turn one to ufi. Follow the
directions below, but substitute U 2 for U and U'.)
a. If the red is at the f of fur, perform the meson,
(R'DR)(FDF')U(FD'F')(R'D'R)U'.
b. If the red is at the r, perform U(R'DR)(FDF')
U'(FD'F')(R'D'R).
5. Rejoice.
Equator Edges
This is the hardest stage to explain. The basic move is
a three-edge swap between fd, fu, and bu (if the orange is
at left or right). The possibilities are too many to write
down, but here's the general set of steps:
1. In the equator (actually the Prime Meridian in this

case), find an edge cubie that's incorrectly placed, and turn
the cube so that it is at ufo
2. Put the cubie that should be at uf at fd. It might help
to get it to either db or dr first. Call this move X. Don't
hesitate to write it down.
3. Put the cubie that's in the space that the cubie at uf
should be in at ub, so that fr can go to fu can go to ub.
Call this move Y.
4. Perform RL'U 2 LR'F 2 •
5. Now perform X'Y'.
6. Go back to step 1 until all the equator edges are
correctly placed, even if they're not correctly oriented.
67
Bottom Edges
Turn the red so that it's up. Either all four edge-cubies
have to be switched, or only three have to be.
1. If uf should be at ub and ul should be at ur and
vice versa, perform R2L2DR2L 2U 2R2L2DR2L2.
2. If all four are wrong, turn the cube so that ur should
be at ufo Perform LRB(L 2U 2 )3B'R'L'.
3. If three are wrong, turn the correct one to ul.
a. If the cubie at uf should be at ub, perform (R 2D')
(RL'U 2LR'F 2 )(DR 2).
b. If the cubie at uf should be at ur, perform (R 2D')
(F RL'U 2LR')(DR 2 ).
2
4. One hundred percent of the cubies are now correctly
placed!
Edge Orientation
1. Turn the entire cube so that one flipped edge is at ufo
2. Turn faces so that another flipped edge is at ub.
3. Perform either a or b:
a. FUD'L2U2D2RU2R'D2U2L2DU'F'U2
b. (LR'F)(LR'D)(LR'B 2 )(RL'D)(RL'F)(RL'U 2)
4. Undo step 2.
5. Now go back to step 1.
6. Tell everyone within shouting distance: You've
solved the cube!!!
More Challenges
Now go for speed, make pretty patterns, try odd colorings
or cubes with unusually-shaped cubies, solve the new
4 X 4 X 4 cube, or try to solve the cube in the mirror.
Good luck!
68 EUREKA!
~oNiftY Numerics
6. Palindromesemordnilap _ _ _ _ _ __
A palindrome is a word or phrase that reads the same for-

ward as backward, such as rotator or "Madam, I'm Adam"
(to which she aptly replied, "Eve"). In mathematics, num-
bers like 39,593 and 123,454,321 are palindromes.
It has been conjectured that the following process would
always yield a palindrome for any starting number: Add
any number to its reversal. If the sum is not a palindrome,
then repeat the process on the result. The number 459, for
instance, takes two steps to become palindromic, and 549
takes five.
Well, the conjecture is nearly right-almost 98% correct.
The process above was tested on all the positive integers
less than 10,000, and 97.5% of them produced palin-
dromes in 24 or fewer steps; however, the remaining num-
69
bers failed to become palindromic in the first 100 steps.
The innocent 196, in fact, fails to produce a palindrome
in 37,303 steps (and contains some 15,500 digits at that
stage! ).
Palindromes seem to pop up frequently. The product of
7, 11, and 13, successive primes, is 1001, a palindrome.
Palindromic pairs of primes, such as 13 and 31, or 37 and
73, are not uncommon. Should palindromic primes like
151, its own reversal, be counted?
Any palindrome multiplied by 11 (another palindrome)
results in a new palindrome. A palindrome can be the cube
of a prime only (as in 1331 = 11 3 ). Perhaps you can find
some more relationships like this.
Here are a few problems involving palindromes:
1. Rhoda and Phil O. Dendron, both part-time math

tutors, found that their earnings over a two-week period
were reversed for each week (Rhoda might have earned
$58 the first week and $85 the second). If each earned
$165 in the fortnight, what were their weekly earnings?
2. V. Hickle, a cautious driver, was out for an afternoon
drive when she noticed that the odometer showed 45,954
miles. "I bet it'll be a long time before a palindrome turns
up again," she said to herself. Yet, two hours later, the
odometer showed a new palindrome.
How fast was she traveling during those two hours?
Answers
1. One of the duo earned $78 and $87, and the other
earned $69 and $96.
2. Victoria Hickle traveled at precisely 55 m.p.h. for
those two hours, covering 110 miles. At the end of the
trip, the odometer showed 46,064 miles.
70 EUREKA!
7. A Pole Vaulter _ _ _ _ _ _ _ _ __
Hap Hazard, world-famous hunter and sportsman, set up

a camp from which to go bear-hunting. One morning he
found bear tracks outside his tent, and, gun at the ready,
started to follow them.
Stalking his prey, he traveled 1 mile due south, turned,
and crept 1 mile due east, at which point he stumbled
across the bear and vanquished his sleepy opponent in a
vicious hand-to-paw fight. After taking the required victor's
photographs, he dragged the carcass back to camp, a dis-
tance of precisely 1 mile due north.
1. What color was the bear?
2. How may points are there on the globe from which

such a journey as Hap's can be taken?
Answers
1. To answer first things first, the bear must have been

a polar bear and hence must have been white. The North
Pole is the only bear-inhabited place from which you can
walk a mile south, a mile east, and a mile north and find
yourself at the point from which you started.
2. If the South Pole had bears, there would be an infinity

of answers to the problem, because there are an infinity
71
of points from which such a walk can be taken (the title
hints at this). If Hap had started at a distance of about
1 + 1!(27T), or about 1.16, miles from the South Pole and
walked 1 mile south, the walk of a mile east would have
taken him all the way around the globe, and the mile
north would have returned him to his starting point.
72 EUREKA!
8. A Timely Switch _ _ _ _ _ _ _ __
Problems involving the transfer of liquids are among mathe-

matics' most enjoyable challenges-perhaps people like
them because the solver can always try the experiments
himself!
1. In a state of global emergency, you are unexpectedly

called upon to measure out precisely 4 quarts of liquid
hydrogen from a huge tank. However, you have been pro-
vided with only a 5-quart measure and a 3-quart measure
with which to save the masses. How do you avert the
destruction of the world?
2. As part of a psychiatric examination, you are given

a bucket containing 24 ounces of fluid and three bottles,
holding 5, 11, and 13 ounces. The test is to separate the
original 24 ounces into three drinkable, 8-ounce volumes.
How do you prove that you are sane and thinking?
3. This one is a little different. You need to remove

a delicate, glazed bowl from the kiln in precisely 9 minutes.
Unfortunately, there are no clocks handy-but you do have
two accurate hourglasses, a 7-minute and a 4-minute. How
do you measure exactly 9 minutes?
Answers
1. You quickly fill the 3-quart measure with hydrogen

73
from the tank and pour it into the 5-quart. Now you re-
fill the 3-quart and pour from it into the 5-quart until the
5-quart is full. Now you dump the contents of the 5-quart
back into the tank, pour the 1 remaining quart in the
3-quart measure into the 5-quart, and refill the 3-quart.
At last, by merely pouring the contents of the 3-quart
into the 5-quart, you have saved the world.
2. First, fill the 11- and 5-ounce bottles, leaving 8

ounces in the bucket, and pour all the contents of the
5-ounce into the 13-ounce. Now pour from the l1-ounce
into the 13-ounce until the 13-ounce is full and the
l1-ounce has 3 ounces. Now refill the 5-ounce from the
13- and pour the contents into the l1-ounce. Voila! You
can reason!
3. Start both hourglasses. When the 4-minute runs out,

flip it over; when the 7 -minute runs out, flip it over. When
the 4-minute runs out again, flip them both over. When the
7-minute runs out a second time, 9 minutes will have
elapsed.
74 EUREKA!
Fallacies
350 and Logic
9. If This Is Not a Chapter, My Name

Is Raymond Smullyan _ _ _ _ _ __
Let's say that one fine morning you awake to find yourself
on an obscure South Seas island. By and by, you discover
that the natives all speak Paradox, a language you are
familiar with, and are all of two contrasting tribes, one
consisting wholly of truth-tellers and the other of liars.
Unfortunately, it is impossible to tell which tribe an
inhabitant belongs to by his dress alone.
Now, it also happens that all the natives are rather close-
mouthed, and have the nasty habit of disposing of anyone
who asks them more than one question. For this reason,
you consider it best to make your way to the local village,
75
where you will undoubtedly be able to secure transport off
the isle.
Setting off in what seems to be the right direction, you
soon come to a fork in the road, beside which a native is
lounging. So, before you read on, can you determine what
single question to ask this liar or truth-teller, so that you
can figure out which branch of the road leads to the village?
There are actually several solutions, all along the same
basic lines. For example, you could point to one of the
roads and ask the islander, "If I were to ask you if this
road leads to the village, would you say, 'yes'?" An answer
of "yes" indicates that this is the correct road, even if the
villager is a liar! The liar would respond "no" to a direct
question if the path was the right one, but must lie about
his response to your question, giving an answer of "yes."
A similar question would be, "If I asked a member of
the other tribe whether this road leads to the village,
would he say, 'yes'?" Here a "no" answer indicates that
the road is correct. A better way of saying this is, "Of the
two statements 'You are a liar' and 'This road leads to the
village,' is one and only one of them true?" What would
you make of a "yes" answer? Again a liar would lie about
his being a liar and about the number of true statements,
so a "yes" does indicate that you're heading in the right
direction. Another stratagem is, "Is it true that this is the
way to the village if and only if you are a liar?" A "no"
indicates that it is indeed the way to the village.
Now suppose that you come to a crossroads with a
native conveniently near (actually, you suspect that he
may be sitting on the signpost, but you don't want to
waste your question asking him to move). You are pre-
pared to ask him the same question that set you off cor-
rectly on this road, but "you find, to your horror, that
you have forgotten the Paradox words for "yes" and "no."
76 EUREKA!
All you can remember is that "plink" means one and
"clunk" the other, but you don't know which is which.
Can you still determine which road leads to the village?
Yes, you can. Just point to one of the roads and ask,
"If I asked you whether the road I am pointing to is the
road to the village, would you reply, 'plink'?" Regardless
of the tribe of the native or the meaning of "plink" and
"clunk," a reply of "plink" indicates that the road is the
one to the town. If "plink" means "yes" and the road is
the right one, a truth-teller would reply "plink," or "yes,"
to your question. A liar would lie about his lying response;
thus he would also say "plink." Similar logic applies if
"plink" means "no." The truth-teller, for example, would
say "plink," or that you were incorrect in saying that the
road was not the one to the village. And if the road were
not the right one, you would hear the opposite reply in
all cases, a "clunk." (What you don't want to hear are
things that go "clunk" in the night.) So, if you ask with
a "plink" and get a "plink" back, you're on the right road.
If you were to come across a particularly quiet inhabit-
ant of the island, you could still determine which branch
leads to the village with the question, "Which of these
roads leads to the village?" Presumably, a liar would point
to all the wrong ones. Unfortunately, he might also be lazy
and merely point to one road. Thinking that he was a truth-
teller pointing out the correct road, you would soon be
going in the wrong direction.
All of these questions are also subject to misinterpreta-
tion. "If I were to ask you if this road leads to the village,
would you say 'yes'?" could seem the same as "Does this
road lead to the village?" to an especially simple liar, who
would helpfully give you the wrong answer. On the other
hand, you might come across a particularly clever liar
who realizes that he's being tricked into giving the right
77
directions. (This illustrates the difference between an
honest liar and a malicious, deceitful liar.)
Here are a few problems involving truth-tellers and liars:
1. You come across three natives engaged in conversa-

tion, so you decide you can probably ask more than one
question. You ask the first man (A), the quietest of the
lot, what his tribe is, but you cannot hear his low reply.
The next, person B, tells you that A said he was a truth-
teller, but C, the third native, tells you that B is a liar.
Assuming that you want your next question to be addressed
to a truth-teller, to whom should you put it?
2. What do you make of the following statement: "If

A, B, C, and D each speak the truth once in three times,
and A affirms that B denies that C declares that D is a liar,
then what is the probability that D was speaking the
truth?"
3. Since your skill as a logician is famed throughout the

land, you are confronted with the following three state-
ments:
A: "Bislying."
B: "c is lying."
C: "Both A and B are lying."
Who is not lying?
4. One of four crooks is telling the truth. Their state-

ments are as follows:
A: "One of us is lying!"
B: "No! Two of us are lying!"
C: "Not so! A full three of us are lying!"
D: "False! All of us are telling the truth!"
78 EUREKA!
Who is telling the truth? This can also be represented this
way:
A: One statement is false.
B: Two statements are false.
C: Three statements are false.
D: Four statements are false.
How many statements are false?
5. Another type of paradox is contained in a problem

frequently represented in an egg-finding or man-to-be-hung
form.
A mathematics teacher, known to be extremely fair-
minded, told her class one day, "I've scheduled an algebra
test for sometime next week. However, I am not going to
tell you what day the test is on. You will not know when
we are having it until you come in that day for class."
And she left.
Immediately the class logician stood up and quieted the
groans. "It is obvious that she can never give the test with-
out us knowing beforehand." He strode up to the black-
board and chalked down the days of the week from Monday
to Friday.
"Now, we can't have the test on Friday because, if we
haven't had it by the end of next week, that would be the
only possible day left," he explained, erasing "Friday"
with great relish. "So it can't be on Friday. If we haven't
had the test by Thursday, then it would have to be on that
day because we have already proven that it can't be on
Friday; we would then know that it would be Thursday,
so we can cross it out, too." In a likewise manner, he
erased Wednesday, Tuesday, and Monday.
As a result of this brilliant proof, no one studied for the
test. Imagine their surprise Wednesday morning when the
79
exam papers were handed out! (This problem is a classic.
I hope you can see the flaw in the logic. One debatable
solution is to expect the test every day. Then again, you
really don't know when it will be given. Thinking about
it only makes it more difficult to figure out.)
Come the next test, the teacher decided not to be so
clever and just let the class know, though in a roundabout
way, that it was sure to have a test of some sort. She told
the logician, "If you make a true statement, we will have a
small test. But if you make a false one, we will have a
large quiz." If you were in his place, could you figure out
what statement to make so that it would be impossible for
the teacher to give you a test?
This leads us into the subject of paradoxes in simple,

innocent-looking sentences. Remember the island of truth-
tellers and liars? Why was it impossible for any islander to
say, "I am a liar"? (Not unless he is a normal human, a
third type that either lies or tells the truth indiscriminately.
Then he could be telling the truth about his lying.) If true,
then it's false, which makes it true, which makes it
false ....
Another sentence like this is "This sentence is false"
which is false if true and true if false. This can be changed
into "This sentence can never be proved." Try to!
P.E.B. Jourdain devised a card which reads on one side,
"The sentence on the other side of this card is true," and
on the other side, "The sentence on the other side of this
card is false." You might like to make such a card and
give it to your worst enemy. (That reminds me: Is your
worst enemy your best friend? What is your best enemy?)
6. "The number of words in this sentence is nine."

Try to construct a sentence equally true but nevertheless
saying the exact opposite. It's simpler than you think.
80 EUREKA!
I once made a bet with the captain of the track team,
who used to be a good friend of mine, that I could beat
him in a half-mile race (his best event, of course). When
the day of the race turned out to be the hottest day that
summer, I kindly offered to postpone the race, but he
wanted to see me suffer.
"Well, I really don't quite feel like running today,"
I said, squirming a little and trying to assume a pleading
expression. "Perhaps if you gave me a foot's head start I
might."
This he readily agreed to. He even laughed at me!
"Well, even that's not enough," I said, suddenly pausing.
"Hey! I have an idea! Let's just figure out who would win."
Again he agreed, assuming a victorious air, so I created
an argument that went something like this: "Let's just say
that you run twice as fast as I do (which he almost did).
When we start the race, I'll be a foot ahead. Therefore, by
the time you cover that foot, I'll be half a foot ahead.
Right? By the time you cover those six inches, I'll still
be three inches ahead. My lead will, of course, diminish
to 1.5 inches, then to 0.75 inches, and so on." (And we
would never get more than two feet into the race.)
Unfortunately, we never finished the argument or started
the race, but you do get the general idea of the paradox,
first proposed by Zeno, a Greek philosopher, some 2000
years ago. In his version, a tortoise beat the mighty Achilles
in a foot (and flipper) race.
Another paradox along this line is one most students are
familiar with, and I suggest you try it now. Place your
pencil down anywhere in the approximate center of this
page and draw a light line toward the edge, so that your
pencil tip is now precisely twice as close to the margin as
it was before. Repeat this until you have reached the edge
of the book (even use a ruler if necessary).
The fallacy here is mathematical. Each time you move
81
your pencil, the distance covered is half the distance
covered the move before. This can be expressed as the
.
series "21 + "4
1 + "8
1 + 16 1 ••• ,t h e sum 0 f wh·IC h approac hes,
1 + 32
but never quite makes it to, 1. If you are precise enough,

you can continue until the end of the universe. (Perhaps
it is easier to see that with each move of the pencil, you
cover half of the remaining distance; thus you can never
reach the full distance, no matter how many moves you
might make.)
7. Computational fallacies abound. Consider this one:

9 - 24 = 25 - 40
9 - 24 + 16 = 25 - 40 + 16
(3 - 4)2 = (5 - 4)2
3-4=5-4
-1 = 1
Gotcha!
8. What's wrong with this one?

V5=X = 1 +yx
5- x = 1 + (2yx) + x
4- 2x=2yx
2-x=yx
Squaring:
4 - 4x + x 2 = X
0= x 2 - 5x + 4
o = (x - 4)(x - 1)
x =4 or x =1
82 EUREKA!
If we substitute these roots back into the equation, the
answer of 1 checks, but 4 gives an odd result:
v"5=-4= 1 + V4
1=1+2
(Hint: The above two use the same gimmick.)
Here's a neat fallacy involving quadratic equations

(don't solve them like this!):
(x+3)(2- x)=4
x + 3=4 or 2- x =4
x =1 or x =-2
Correct! In fact, you can make up your own fallacy of this
type with your own roots p and q (ignore the x until the
final form). Just represent it like this:
(l + q - x)(l - p + x) = 1 - p + q
and solve it as above.
Here's another curious generalization:
a2 - b 2 = (a + b )(a - b)
so:
a + b = a2 -b2
a- b
Cancel the a and b, and then the minus signs:
a+b = a~ -f b~
11th
a+b=a+b
Correct!
83
Another set of mathematical fallacies is of the out-
worn sort:
x=O
x(x - 1) = 0
x - 1= 0
x=l
1=0
where dividing by zero produces the fallacy. A better-
disguised one:
a=b+e
a2 - ab = ab - ae - b 2 - be (multiplying by a - b)
a2 - ab - ae = ab - b 2 - be
a(a - b - e) = b(a - b - e)
a=b
Another branch of fallacies involves i, the imaginary
square root of -1:
(y=T)(y=T) =y(-l)(-l)
y=T2 =y'T
-1 = 1
The method is just fine except for one fact: (-1)(-1) = 1,
but i 2 = -1; the mistake is thus introduced in the first line.
84 EUREKA!
Also:
yCI=yCI
VV-1=v-m
VI lyCI = yCI/VI
(VI)(VI) = (v=-r)(yCI)
1 =-1
That one's a little harder to figure out!
9. Geometric fallacies are also quite interesting. A
neuropsychiatrist devised a mind-turning demonstration
of a paper triangle whose back has a smaller area than its
front.
If you dissect a 64-unit square as in Fig. 9-1, and rear-
range the pieces as in Fig. 9-2, a rectangle of 65 square units
results! What is wrong?
5 3
J
/
/
1/
.......
.......
r---. ........ 3
I'-- ...... Fig. 9-1
8
8 5
.......
I'--
I'-- ...... 3
...... ........
5
3 r-....... ........
........
I'--... Fig. 9-2
5 8
85
10. Here's a small puzzle in logic:
Three business partners went out to dinner one evening
and ran up a check of $30. Each gave the waiter $10 (and
also charged the full $30 to the expense account of each, I
might add). Before their change was returned, the manager
decided not to charge them for their drinks and returned
$5 to the waiter to give to the partners.
The waiter didn't think they had tipped him enough, so
he gave each of the partners $1 and pocketed $2 himself.
Now, each of the three had paid $9, making $27 in all,
and the waiter had $2, making a total of $29.
What became of the last dollar?
Answers
1. The first man (A) must have said that he was a truth-
teller, because a liar would lie about his tribe. Therefore,
B must have told you the truth, and C lied, so your next
question should be to B.
2. This question raised a lot of controversy when it

was first proposed half a century ago. If a table is formed
with all the responses of the four, it turns out that 40 of
the 81 possibilities are internally inconsistent. Of the
remaining 41, 13 begin with D telling the truth, so the
probability of his telling the truth is 13/41.
3. B is telling the truth. If he were lying, then both C

and A would be telling the truth. But as C is saying that A
is a liar, B must be telling the truth.
4. In the first case, D obviously cannot be telling the

truth. Therefore, D is lying. This means that A is true,
and then Band C are telling falsehoods. But this would
make three false statements, and three liars, which con-
86 EUREKA!
tradicts C, so A is also lying. Now B could be telling the
truth and C lying, but that again contradicts C, so A, B,
and D are lying and C is telling the truth.
In the second case, D contradicts itself, so it must be
false. Logic then proceeds as above.
5. You should say, "We will have a large quiz." If this

is true, the teacher will give you a small test-but then you
have uttered a false statement, so you should receive a
large quiz. But then your statement is true ....
6. "The number of words in this sentence is not nine."
7. The error is in the change from the third step to the

fourth. If we substitute values in:
-15 = -15
1=1
(_1)2 = (1)2
-1 = 1
This is an example of false square roots. The square root
of a number squared is not always that number. For
instance, the square root of the square of -1 is 1.
8. Again, squaring or taking the square root of a num-

ber can introduce false roots. For instance, the roots of
x 2 = 1 are both -1 and 1. Taking the square root adds an
answer. In the fallacy, squaring added a false root which
must be checked in the original expression.
9. The lengths of the line segments are in the Fibonacci

series, where each term is the sum of the previous two
terms. The ratio of successive terms approaches cp, a never-
ending decimal equal to (1 + yI5)/2 or 1.61803398 ...
87
from either side. This means that the rectangle alternately
loses and gains area along the diagonal, explaining this
paradox. If the square is not to lose any area, the lengths
must be cut according to the series 1, cp, cp2, cp3 , cp4, ..•.
10. If you read the problem carefully, you can see that
the other dollar went nowhere. Of the original $30, the
manager kept $25, the waiter kept $2, and $3 was returned
to the patrons (or, if you prefer, the partners paid $9 each
for a total of $27, of which $25 went to the manager and
$2 went to the waiter).
88 EUREKA!
10. Thrice Befuddled _ _ _ _ _ __
1. The Pythagorean triplets-Kate, Kathy, and Katherine-

claimed to be perfect logicians who could instantly deduce
all the consequences of a given set of premises. One day
Archimedes Anderson, head of his school math team and
logician extraordinaire, decided to put them to a test.
On the head of each of the sisters he put a hat, red or
green, so that no girl could see the color of her own hat.
Grinning at his own cleverness, Archimedes asked, "Will
each of you please look at the hats of the other two, and
if you see a red hat, please raise your hand." All three
immediately raised their respective hands.
Practically rolling on the ground with mirth, Archimedes
said, "If any of you is certain what color hat you have on,
please raise your hand." After a few seconds, Katherine
raised her hand and correctly named the color of her hat.
What was her reasoning?
2. Well, Archimedes was quite put off by this excellent

display of logic, so he decided to devise a better test. He
told the triplets that he had eight stamps, four red and
four green. Then he loosely affixed two to the forehead of
each girl so that she could not see the stamps on her own
forehead, but could see the foreheads of her sisters.
He asked Kate, "Do you know the color of your stamps?"
She replied, "No." The question was then put to Kathy and
Katherine, with the same reply. Archimedes then asked
Kate again, and she couldn't reply. Kathy, however, an-
89
swered, "Yes" on her second turn, and correctly named her
stamps.
What are the colors of her stamps, and how did she
know?
Answers
1. Katherine reasoned thusly: "We all raised our hands

to the first question, so there are at least two red hats
among the three of us. If I am wearing a green hat, my sis-
ters, much quicker than I, would have realized that each
was raising her hand for the other. Therefore, I am wearing
a red hat."
2. Archimedes explained the logic to me later. Kathy's

correct answer depended on the fact that her sharp-witted
sister Kate couldn't tell Archie her stamps on her second
turn. Kathy probably started off by assuming that she had
red-red and then reconstructed Kate's thinking.
Kate would have reasoned on her second turn as follows:
"Ah, I see that Kathy has red-red. If I also have red-red,
then all four reds would be used. Katherine, on her turn,
would have realized that she must have two greens, and
would have told Archie so.
"But she didn't. That means that I can't have red-red.
Suppose that I have green-green. In that case, Katherine
again would have answered positively. She would have
realized that if she also has red-red, I would have seen four
reds and would have answered that I have green-green on
my first turn. On the other hand, if she also has green-green,
then Kathy would have seen four greens and answered that
she has two reds. Katherine would have realized that, if I
have green-green and Kathy has red-red, and if neither of
us answered on our first turn, she must have green-red.
90 EUREKA!
"But she didn't. That means that I cannot have green-
green either, and if I can't have red-red or green-green, I
must have red-green. And that means that I can tell Archie
so."
But she didn't, as Kathy saw. That meant that Kathy
couldn't have red-red, her original assumption. As Kate's
argument also applied to Kathy's having green-green, then
Kathy couldn't have green-green either. Therefore, as she
answered, she must have red-green.
The reader could also have reasoned that, just as Kathy's
logic applied to green-green as well as red-red, the solution
is such that it is still true even if all the reds and greens are
switched. This means that the person who answered "yes"
must have had red and green stamps.
Archimedes finally threw in the towel and gave full
credit to the Pythagorean triplets for their skills in de-
duction.
91
11. Better Mixed-Up Than Lost _ __
Sue Venir had just packed up six marbles-three black and

three white-into three boxes labeled BB, WW, and BW,
putting two marbles in each. Unfortunately, she realized
that she had put each set of marbles into the wrong box,
so that all the labels were incorrect. By picking just one
marble from a single box (it would be such a bother to
unpack them all), she managed to figure out which pair
of marbles was in which box, and to adjust the labels
accordingly.
From which box did she pick?
Answer
She learned the contents of each box by drawing a single

marble from the box labeled BW. If it had been black,
then the other marble in that box must also have been
black (or else the label would have been right). This means
that the box labeled BB must have contained two whites,
and the box labeled WW must have contained one of each
color. Similar reasoning would apply if the first marble
had been white.
92 EUREKA!
And Even
@ Dissection
~oofSolids
12. Archimedes Anderson and the

Case of the Sinister Plot _ _ _ __
"What are you working on, Dad?" Archimedes Anderson

had wandered into his father's study. "Anything interest-
ing?"
"Just a death out in the woods the editor asked me to
write up." Mr. Anderson tipped back his chair and turned
to face Archie. "A routine story, but there's something
about it that doesn't seem quite right."
He picked up a small notebook. "Sergeant Enigma lent
me his notes on the case. Care to hear?"
93
Archimedes nodded, and Mr. Anderson assumed his best
storyteller's manner ....
The Sergeant took in the scene at a glance: a heavily

wooded area, not lacking in logs or stumps, its tranquil
beauty harshly broken by the body of Dan Druff, half-
dried blood caking to the large depression where his right
temple used to be. Half a foot away from his head, a red
stake, obviously a boundary marker, poked out of the
ground.
"I'm sure that I can explain everything," a nervous
Pete Moss said. "I've been planning to sell this corner of
my property for a long time now. I knew Dan was a
surveyor, so I asked him to measure the area for me."
"You're selling off this land? Pity," Sergeant Enigma
said, gazing up at the stately trees. "About what size plot?"
"A triangular parcel about 850 feet by 650 feet by 1500
feet. I told this to Dan and he said that he wanted to look
over the area before taking any measurements," Pete said,
still a little jittery. "So, we were walking by this stake
here, and Dan suddenly fell and caught his head on the
marker. I tried to stop the bloodflow," he held up a soiled
handkerchief, "but I couldn't. Soon he looked real dead,
so I quickly went back to the house and called you."
Sergeant Enigma determined that Dan Druff's surveying
equipment was in its customary place in his car. Eventually
he reported the death as an accident, and no arrests were
made.
Mr. Anderson finished and thanked Archimedes for his

applause. "Something still seems a little fishy. As a writer,
I realize that the circumstances are such that an accident
would be a terrible waste of a great situation. But I'm
afraid I don't know why."
94 EUREKA!
Archimedes thought for a moment and then suddenly
snapped his fingers. "You're right, Dad! The catch is .... "
How did Archimedes continue?
Answer
"The catch is that Pete Moss's measurements can't

possibly be the sides of a triangle," Archimedes explained.
"In a triangle, the sum of each two sides must be larger
than the third. A triangle with sides of 850, 650, and 1500
would be nothing more than a straight line."
"I see! Dan Druff, an experienced surveyor, would have
realized that. He couldn't possibly have been walking on
the grounds for that reason," Mr. Anderson continued.
"Moss's story is a complete fabrication! Say, I'd better
call Headquarters about this-"
Archimedes Anderson yawned and wandered out of his
father's study.
95
13. How to Dissect a Square and Other
Marvels of Modern Biology _ _ __
"You like my table?" Archimedes Anderson looked rather

pleased. "Great! I think it's rather neat myself."
He cleared the chess pieces and schoolbooks off the
square table. "Have you seen this yet?" With remarkable
dexterity he folded the table out from the middle and
swung the pieces around. Before Katherine Pythagoras's
eyes the square became a perfect equilateral triangle!
"This is one of my favorite geometric dissections,"
he said. She looked a little puzzled, and he added, "Geo-
metric dissection is the changing of one polygon into
another by cutting it up into little pieces and rearranging
those pieces to form the new figure. It's quite a fascinating
subject," he continued, rummaging around for a clean
piece of paper. "This one is the 'haberdasher's problem'
solved by one Henry Ernest Dudeney in England around
the turn of the century. I especially like it because it can
be done using only a straight edge and compass."
Archie quickly drew an equilateral triangle ABC and
bisected AB at D and BC at E. Then he extended AE to F
so that EB equalled EF, and extended CB a little way past
B. With the midpoint of AF as the center of a circle, he
swung the compass from A through BC to F, calling the
new point H. Now with E as the center of a circle, he drew
arc HJ, with J a new point on AC. Then he put another
point K on AC so that JK equalled BE. From D and K,
perpendiculars were dropped onto EJ to obtain points L
andM.
96 EUREKA!
"The final diagram looks like this," Archimedes said,
pushing the paper over the table. (See Fig. 13-1.) He located
a pair of scissors, cut the figure along the solid lines, and
rearranged the pieces into a square (Fig. 13-2).
Fig. 13-1
Fig. 13-2
97
"Something particularly fascinating is that the four
pieces can be joined like this to make a chain," he explained.
(See Fig. 13-3.) "If closed one way it forms the square;
closed the other way it makes the triangle. Dudeney made
a brass-hinged mahogany one, and that gave me the idea
for the table. I had a terrible time placing the legs!"
Fig. 13-3
Dudeney's method, with modifications, can be adapted

for nonequilateral triangles. A theorem first proved by the
German mathematician David Hilbert, in fact, states that
any polygon, no matter how absurd or unusual, can be
transformed into any other polygon of equal area by cut-
ting it into a finite number of pieces. Basically, this is done
by cutting the polygons into triangles, dissecting the tri-
angles into rectangles, and reversing the process to form
the second polygon. Unfortunately, the theorem does
not hold for solids whose faces are regular polygons,
though some dissections of solids are known.
Although the theorem shows that dissection of one
polygon into another is always possible, the number of
pieces may be very large. Henry Dudeney, accomplished
in other areas of dissection, surpassed many long-standing
records. For instance, Fig. 13-4 shows his elegant dissection
of a pentagon into six pieces that form a square.
98 EUREKA!
,f.
, I "
, ,',I '" ,
,
,'1I
'"
3
"
,
,,'
"
,'
I
...
,I ' " .......
",
6 " 2 .............
,,' " ...........
~,---r-----J_- ----- - ----~--,¥-------
'\. \ "
1 ""\
"II. 4 "
..........•. / .....••/
, ,
'v'
Fig. 13·4
A record that stood for ten centuries was the nine·

piece dissection of three squares into one shown in Fig. 13·5
(dotted lines indicate extra cuts), found by the 10th century
Persian astronomer Abul Wefa and broken by Dudeney with
the ingenious six·piece solution shown in Fig. 13·6. (Draw
circle with center at A. BC=DE=FG.)
.. ---------
I
I
I
I
I
; '1I
~
I
I
I
I
,
I
I
I
----
Fig. 13·5
99
5 6
1
D~----'--":"-"'-~E,.-IA'-----'C
5 6
Fig. 13·6
Figures 13-7 through 13-9 show some more surprising

dissections to contemplate. (The Greek cross, like the
equilateral triangle, particularly lends itself to dissection.)
100 EUREKA!
Fig. 13·7 Hexagon to square (5 pieces)
Fig. 13·8 Dodecagon to Greek cross (6 pieces)

101
Fig. 13-9 Maltese cross to square (7 pieces)
1. If you were given Fig. 13-10, could you reassemble

it into an equilateral triangle? There are several ways of
going about it. It is interesting to note that alternate solu-
tions to a dissection are almost always completely unalike.
Perhaps you can devise your own or a better one of these.
Fig. 13-10
2. As a start, grind your teeth on the one in Fig. 13-11,

which is composed of five equal squares joined together.
102 EUREKA!
Can you, by making only two cuts (no tricks; just three
pieces), rearrange it into a perfect square? The answers
follow.
~ _ _.......-_ _....J _________ •
_____ .___ .. ~--~--....J
Fig. 13·11
Answers
1. Figure 13-12 shows the easiest solution with those

cuts.
Fig. 13·12
103
,,
,
\
\
\
1 \
Fig. 13·13
2. Figure 13-13 is equivalent to solving the rather sticky

dissection of five squares into one in only nine pieces.
Congratulations!
104 EUREKA!
14. Geometer's Heaven ________
1. Starting with a square of paper, Stan Doffish trimmed

away the corners to leave the largest possible circle. Then
he trimmed the circle to leave the largest possible square.
How much of the original square did he discard?
2. Jim Nasium set up a ladder so that it rested on top

of a 15-foot fence and touched a building 8 feet away and
some 27 feet up. How far from the fence was the lower
end of the ladder?
3. If the circle in Fig. 14-1 has a radius of 10 units, how

long is the diagonal x?
Fig. 14-1
105
Answers
1. The diagram in Fig. 14-2 should make everything

clear. As the area of the inner square equals one-half that
of the outer, Stan snipped away half of the original paper.
Fig. 14·2
2. Again, a diagram (Fig. 14-3) is useful here. The pro-

portion 15 is to 27 as x is to x + 8, gives x as 10, the dis-
tance from the end of the ladder to the fence.
106 EUREKA!
27
Fig. 14·3
3. In Fig. 14-4, line y is the radius of the circle, 10 units

in length. It so happens that y is also the diagonal of the
rectangle. Therefore, x is equal to y, so x is also 10 units
long.
Fig. 14-4
107
15. Hole in the Sphere _ _ _ _ _ __
Boris N. Ventive grinned fiendishly at his own cleverness.

He quickly looked over both shoulders, and then pulled
from under the hotel bed ... a bowling ball.
Boris braced the bowling ball with one hand and reached
for the drill with the other. He carefully cut out a solid plug
through the center (leaving a large donut), put a small
hollow cylinder inside the hole, and then glued the spher-
ical cap from the plug onto the bottom of the lead con-
tainer, creating a hollow cylinder inside an innocent
bowling ball. See Fig. 15-1.
Fig. 15-1
Boris then removed a quantity of the world's most

precious metal from its hiding place inside the television-
the highly radioactive Tedium-235. Being careful to touch
the metal only with his gloves, Boris filled the lead con-
tainer, and then glued the top spherical cap onto the top
of the cylinder. His work was done.
108 EUREKA!
Done, but not unobserved. Special CIA agent Stu Pidd
had seen the entire procedure through the keyhole. He
desperately needed to know how much Tedium-235 was
being smuggled out of the country, but all he had to work
with was the fact that he knew the length of the lead
cylinder-6 inches. He didn't know its width, but he
thought that if he could figure out how much volume
remained in the bowling ball, even though he didn't know
its size, he could determine the volume of the metal.
Stu wasn't able to, in any case. Using only the infor-
mation he had, can you tell how much volume remained
in the bowling ball after Boris had hollowed his 6-inch
cylinder through its center but before he had replaced
the spherical caps?
Answer
The remaining volume was 361T cubic inches.

The problem can be solved by reasoning that it has a
unique solution, true no matter what size bowling ball is
used or what the width of the cylinder is.
In that case, the volume must be a constant that would
hold even if the hole has a radius of 0 and a "length" of 6.
That length would equal the diameter of the sphere.
Therefore, the residue must equal the volume of a sphere
with a diameter of 6 inches.
The radius, then, is 3 inches. As the volume of a sphere
is given by the formula V= (4/3)1Tr3, then (4/3)1T(3)3
equals 361T, giving 361T cubic inches as the remaining
volume.
So, for any size sphere, the internal (or "infernal")
construction of a hollow cylinder with height 6 leaves a
volume of 361T.
109
16. Convexstasy _ _ _ _ _ _ _ _ __
If you wanted to stack four tennis balls so that each was

the same distance from all the others, you would find that
you would have to put three down to form an equilateral
triangle and place the last on top to form a pyramid. If
you were then to connect the centers of these balls, you
would get what is called a tetrahedron.
The tetrahedron is a four-cornered solid with four
equilateral triangles for the faces. It can be formed by
folding up the three corners of an equilateral triangle so
that they meet at a point.
The tetrahedron is the first Platonic solid, or regular
geometric solid. These figures have faces that are all the
same regular polygon. The cube, next in the series, has six
square faces, and the octahedron has eight triangular faces.
It would seem that there would be an infinity of these
Platonic solids, but there are only five! The three described
above were known by the Egyptians, and the next two
were discovered by the Greek mathematician Pythagoras.
The fourth one, the dodecahedron, has twelve sides, all
of them regular pentagons. It looks very much like a pointy
soccer ball.
The icosahedron, composed of twenty triangular sides,
can be formed from the hexagonal shape by removing one
of the triangular sections. The others fold down to form
the icosahedron's corner. See Fig. 16-1.
110 EUREKA!
Fig. 16-1
QUICKIE
1. Which volume is larger? That of an icosahedron, or

that of a dodecahedron that will fit in the same sphere?
Of all these solids, you are probably most familiar with

the cube, perhaps for its properties as a die. In the fantasy
game Dungeons and Dragons, all these solids are used as
dice. The octahedron, though, seems to be the most pop-
ular after the cube. If the octahedron is formed from the
strip in Fig. 16-2, the opposite sides total to seven, just as
they do on a cube.
Fig. 16-2
111
The same arrangement can be used to create quite an
astounding trick. Ask a person to select a number from 0
to 7. Then hold up the octahedral die so that he sees the
faces 1, 3, 5, 7, and inquire as to whether he sees his num-
ber or not. A "yes" answer has a value of 1. Ask about 2,
3, 6, 7, and add 2 for a "yes" answer; then ask about 4,
5,6, 7, and add 4 for a "yes." The sum is the number!
2. Perhaps the following problem is easier to figure out.

Try to arrange the digits 0 to 7 on the faces of an octa-
hedron so that the sum at each corner is 14. Good luck!
Answers
1. You must have realized that the 12-faced solid has a

larger volume; otherwise, the problem isn't worth pre-
senting. Most people figure that, because the icosahedron
has more faces, it more closely approximates the sphere,
and thus should have the larger volume. It doesn't.
2. There are three ways to arrange the digits; Fig. 16-3

illustrates one. By the way, this summing property is unique
to the octahedron.
Fig. 16-3
112 EUREKA!
17. Great Unsolved Problems - - - -
There are three great unsolved problems of mathematics,

ones which have stymied mathematicians for the past few
millennia. The ancient Greeks could not figure out why
they were impossible to solve-but perhaps you can.
1. The first involves duplication of a cube. Find the

edge (exactly!) of a cube whose volume is twice that of
another given cube.
2. The second concerns the trisection of an angle with

the classic geometer's tools, the compass and straight
edge. For any angle, find a method of cutting it into exact
thirds. No fair using a ruler!
3. The last involves squaring a circle. Find a square

whose area is precisely equal to that of a circle. If you find
you can't do it, try circling a square, and if you give up,
you'll find the answers following.
Answers
1. The volume of a cube is equal to the edge cubed, or

V = e 3 • Twice the volume of the cube is then equal to the
cube of the cube root of two times the edge, or 2V =
(e·~/2)3 . As the cube root of 2 is irrational (about 1.2599
... ), doubling the volume of a cube (but not increasing
it by 8 or 27 times) is impossible.
113
2. The second problem, trisection of a given angle, also
involves cube roots, but in a much more hidden way. It is
possible, using only a compass and straight edge, first to
double an angle and then to "add" it on to the original,
giving you three times your angle. Unfortunately, it is
not possible to work backward. Also, the cube root is
such that the trisection can be done if you mark a straight
edge. Legal? It's debatable.
3. Squaring a circle is also downright impossible,

because of pi's irrational nature. If you were to draw a
circle with a radius of 1, the area would be exactly equal
to pi (but already you've used a measuring device!). So, all
you need to do is construct a square with an edge equal to
the square root of pi. Alas, you cannot find the square
root of an irrational, so this problem is sadly impossible.
(Also, there is no equation that will enable you to draw
a line exactly equal to pL)
114 EUREKA!
18. Out of This World
Trying to comprehend having - 8 apples or - 3 dollars is

like trying to understand the fourth dimension. There is a
logical basis for both, but that doesn't make them any
easier to understand.
A point is said to have zero dimensions, neither length
nor breadth nor height. If a point is moved, say, 1 inch, a
one-dimensional line with only a length is formed. If this
line is moved 1 inch perpendicular to itself-so that its
movement is at right angles to itself-a square is produced.
The square has length and width, but no height. If many
squares are "piled" on top of each other, or if this square
is moved 1 inch perpendicular to itself by pulling it out of
the paper, a cube is formed. The cube is three-dimensional.
Now comes the tricky part, the transition from the third
dimension to the fourth. If this cube is moved in that
inconceivable direction so that it is perpendicular to itself
at all times, so that its movement forms right angles with
all its dimensions, then it will be moving in the fourth
dimension, not sideways, forward, or upward, but some-
thing else, a theoretical direction.
To us, such a world in the fourth dimension is a world
of super-powers. People in this extra-space can untie three-
dimensional knots without moving the ends. Just as we
can touch the center of a circle or the heart of a two-
dimensional figure, so can a four-dimensional being perform
surgery on our hearts without breaking the skin, or steal
all the gold in Fort Knox by moving it in the fourth
direction.
115
If a right-handed person is turned through 4-space, he
becomes left-handed. The trouble is that he'll think every-
one else has changed! A rubber band can be turned inside-
out in our space, but in the fourth dimension a basketball
can be turned inside-out-without tearing the sides.
In our space, four tennis balls, piled so that each is
touching and is equidistant from the other three, form a
tetrahedron. If a fifth ball is to be added so that the
distance between any two balls is the same, then it must be
added in the fourth dimension to form an extra-tetrahedron.
The most well-known and basic four-dimensional shape,
however, is the cuboid or tesseract. Just as a cube, viewed
from above, can be seen as shown in Fig. 18-1, a tesseract
can appear as the model (drawn in two dimensions) in
Fig. 18-2.
And just as the cube can be unfolded into a two-dimensional
cross (Fig. 18-3), the tesseract, bounded by eight equal
cubes (they just look slanted) can be unfolded to form the
three-dimensional object shown in Fig. 18-4.
" /
V
Fig. 18-1 " Fig. 18-2
116 EUREKA!
Fig. 18-3
Fig. 18-4
A tesseract cannot exist in three dimensions. However,

Robert Heinlein wrote about the woes of living in a cuboid
in his delightful short story "-And He Built a Crooked
House-." The story illustrates some odd pseudo-powers of
extra-dimensions; the characters see their own backs by
looking into the next rooms.
This is very hard to understand. It logically follows that
117
if there exist dimensions of 0, 1,2, and 3, there should also
exist one of 4. But perhaps our problem in not being able
to make the conceptual extension is that we can't imagine
any space other than 3-space. How can beings exist without
height? Without width? We can't even construct a real two-
dimensional figure-even the printed letters you are reading
have height.
Perhaps the fourth dimension will turn out to be time or
hyper-space, and not be geometrical at all. The question,
though, will undoubtedly plague mathematicians for years
to come.
118 EUREKA!
~ Photons Are Light
B2)oMatter, Too
19. Archimedes Anderson and the

Gambling Candidate _ _ _ _ _ __
Archimedes Anderson was reading in the living room when

he was interrupted by a peal of laughter coming from his
father's study. Sighing, he put down his book and waited
a few seconds.
"Archie! You have to hear this," his father burst out
from behind the closed door, shaking a sheaf of papers.
"I was just writing up an article for the newspaper, and 1
came across an incredible piece of preposterousness. You
want to hear?"
Without waiting for an answer, Mr. Anderson settled
119
onto the sofa. "You've heard of Monty Carlo, haven't
you?"
Archimedes laughed. "Mostly I've heard about his
ancestors and the great people he is descended from."
"That's him exactly. Anyway, the members of the
Society for the Aid to American War Veterans of Foreign
Descent met Sunday to elect their new officers," he
continued, glancing at his notes. "Monty Carlo insisted
that he was the most qualified for the position of Chair-
man. To sum up quickly, he said that his forefathers had
fought in the American Revolution (on both sides!) and
the French Revolution, and that Carlos had fought in
World War I and World War II."
"Sounds like a good case to me," Archimedes agreed.
"I'd elect him myself if I could."
"He would have won, too, if he hadn't continued,"
his father said. "It seems he wanted to win next year's
election at the same time. He went on to say that he was
precisely one-third English, one-third French, and one-third
German, and that his paternal grandparents were 'twice as
English' as his maternal grandparents were French!"
"Balderdash," Archimedes concurred, groaning. "If it
weren't for the fact that ... "
How did he continue?
Answer
"Balderdash," Archimedes concurred, groaning. "If it

weren't for the fact that it's impossible, he wouldn't be
a laughingstock today. Everybody has 21 parents, 22
grandparents, 2 3 great-grandparents, and so on. Hence,
Monty Carlo is saying that 2n is divisible by 3. Since 2
120 EUREKA!
is the only prime factor of 2 n , 3 cannot divide it evenly.
Monty Carlo was a bit too eager to impress."
"And the rest of his statement was garbage," Mr.
Anderson nodded. "His gamble that nobody would realize
this didn't payoff."
121
20. Once Upon a Time . .. _ _ _ _ __
Age problems tend to be hard problems, and a lot of people

would like to lay their hands on their inventors.
1. "The day before yesterday I was 13, but next year

I'll be 16." On what day was this statement made?
2. I am 30 years younger than my father. If his age is

placed after mine, a perfect square is formed. In 25 years,
our ages will similarly form a perfect square. How old am I?
3. The ages of Jane and Mary added together make 44

years. Jane is twice as old as Mary was when Jane was half
as old as Mary will be when Mary is three times as old as
Jane was when Jane was three times as old as Mary. How
old is Jane?
4. How old will you be when you figure that last

one out?
Answers
1. The person made the statement on January 1, and

the person's birthday is December 31.
2. I am 14 years old. Both 1444 and 3969 are perfect

squares (and their roots, 38 and 63, are 25 apart).
3. Jane is 27.5 years old and Mary is 16.5 years old.

122 EUREKA!
21. A Problem Fly _ _ _ _ _ _ __
Abby Normal and Liz Ard are exactly one-fourth of a mile

apart and bicycling toward each other, Abby at 12 m.p.h.
and Liz at 8 m.p.h.
A tireless fly with a thirst for adventure begins flying
back and forth between the two at a constant speed of
30 m.p.h., turning as soon as it reaches the bicycle of one
(it started on Liz's bike, by the way).
The fly continues this flight between the two, making
shorter and shorter trips each time, until it meets a horrible
death as the two cyclists bump front wheels.
How far did the fly travel before the crash?
Answer
The fly traveled three-eighths of a mile.

You may be surprised to find out that it doesn't matter
whose bike the fly started on, or even that the speeds of the
two cyclists were different.
Liz and Abby together traveled 20 m.p.h. At that speed,
it takes 45 seconds to cover one-quarter of a mile. The fly,
traveling at 30 m.p.h. during those 45 seconds, covered
three-eighths of a mile before the terrible collision.
123
22. The Leading Series of Pisa _ _ __
"Look, David, you've gotta get that book written soon."

"I've started already. As a matter of fact, you inter-
rupted me when you called-"
"Is that so? What are you working on now?"
"An article on the Fibonacci series."
"The what?"
"Well, about 780 years ago a mathematician by the
name of Leonardo of Pisa was working on a problem con-
cerning the breeding of rabbits, and-"
"Rabbits! What have rabbits got to do with numbers?"
"Fibonacci-that's Leonardo's nickname-asked himself
how many pairs of rabbits there would be at the end of
December if he started with one pair in January, if this
pair produced one pair in February, if pairs always bred
in the second month after birth, and if all the pairs bred,
well, like rabbits."
"Hairy problem. How'd he get the answer?"
"By setting up a table of the number of rabbits each
month. He found that there was an unusual sequence of
numbers in the table, kind of an additive series."
"Really? What's so special about it?"
"The standard series begins with two is and each number
after this is the sum of the two before it: 1, 1, 2, 3, 5, 8,
13,21,34,55,89, 144, 233, 377, and so on. By the way,
377 is the answer to the rabbit problem."
"Interesting. What's it good for?"
"The series? It's useful in many sciences, not only in
mathematics. The spiral arrangement of buds on a plant
124 EUREKA!
can fit into a Fibonacci sequence, as can the number of
bees in each generation of a male bee's ancestors. Electri-
cians analyze electrical networks and scientists measure
light reflection in Fibonacci terms."
"You mentioned math tie-ins a minute ago."
"And how! There are only two known perfect squares in
the sequence, 1 and 144, while 1 and 8 are the only cubes.
The 3d, 5th, 7th, 11th, and 13th numbers are all prime,
while 3, 5, 7, 11, and 13 are also prime, and so on. Unfor-
tunately, the 19th number is composite."
"Slow down! If the nth Fibonacci number is prime,
then so is n, but not always the other way around. Is that
right?"
"With the exception of 3, you've got it. There are also
periodic factors in the series. Every 3d number can be
divided by 2, every 4th by 3, every 5th by 5, every 6th by
8, and so on. And 2, 3, 5, 8 begin the series."
"That makes sense. The same numbers are repeatedly
added to form each new number."
"That's a neat way of putting it. The Fibonacci series
also ties in with the Golden Ratio, which is equal to
(1 +y'5)/2, or about 1.618 followed by a string of deci-
mals I've forgotten. Moving down the terms, 1/1 equals 1,
smaller than the Golden Ratio; 2/1 equals 2, larger; 3/2
equals 1.5, smaller; 5/3 equals 1.66, larger; and so on."
"You mean the ratio of successive terms approaches the
Golden Ratio?"
"Yes, as does any series formed the same way. It's the
seesawing larger and smaller with the Golden Ratio, by the
way, that makes an amazing paradox possible. If an 8 X 8
square with area 64 is cut into four pieces with various
sides of 3, 5, and 8, the pieces can be rearranged to form
a 5 X 13 rectangle."
"It gained area!"
"The paradox lies in the fact that a very small slice is
125
missing from the diagonal; 3, 5, 8, and 13 are all Fibonacci
numbers. If x, y, and z are successive Fibonacci numbers,
then xz - y 2 equals either 1 or -1."
"I see. 52 - (3 X 8) = 1."
"Right on. If you added the 2 to those three numbers,
you would have 52 - 3 2 = 8 X 2 = 16."
"That's true for any consecutive four numbers?-Gh,
I've got to run now! Thanks for the talk. I hope you don't
mind the phone bill we ran up."
"What! You called collect!"
"Click! "
126 EUREKA!
23. The Early Something Catches
the Whatever _ _ _ _ _ _ _ _ __
Justin Time lives in the town of Shapeless, 90 miles west

of the city of Chris, Mass. One day he is given reliable
information about the time of arrival of something at Chris.
He estimates the speed of the something and figures out
when it will arrive at Shapeless. But the something arrives
sooner than he thinks because it traveled slower than he
thought.
Can you explain this strange event?
Answer
Justin Time had called the train terminal in Chris to

find out when a certain eastbound train would arrive there
(he did so because his local station was notoriously below
standards in such cases). He was told that it would arrive
promptly at noon.
Justin decided that trains travel at about 90 m.p.h.
In that case, as he estimated, the train would arrive at
Shapeless at 11 A.M. But this train traveled at a constant
60 m.p.h., thus arriving in Shapeless at 10:30 A.M., or
sooner than he expected.
127
(8)0 Shortcuts
24. A Speedier System of Solving _ __
The Trachtenberg Speed System of Basic Mathematics was

invented by the late Jakow Trachtenberg, a brilliant engi-
neer and mathematician, during World War II. Imprisoned
in a German concentration camp, he kept his sanity by
devising new methods of calculation in his head, putting
only the final formulas on hard-to-get scraps of paper.
Eventually, he escaped and founded the Mathematical
Institute in Zurich, where his methods are successfully
being used today. People the world over especially enjoy
the new system because the only number written down is
the answer!
129
Here are a few of the simpler methods for multiplication.
Multiplication by 11: All you do is add the number to

the digit on the right and carry any number in the lOs
place. This is based on the fact that mUltiplying by 11 is
the same as multiplying by 10, multiplying by 1, and
adding the results. A tick mark (') indicates a carry of 1:
15637 X 11
---- = 172,007
17'2'0'07
The 7= 7+0
0=3+7
o= 6 + 3 + carry of 1
2 = 5 + 6 + carry
7 = 1 + 5 + carry
1=0+1
This method is especially useful when multiplying by large
numbers.
Multiplication by 12: This is a little harder. This time

double each digit and then add the product to the digit
on the right, as well as adding in any carry-overs.
4372 X 12
= 52,464
5'2'4'64
The 2 in the answer, for example, is equal to (4 X 2) + 3
+ carry of 1.
Both of the above could have been done this way:

Multiplication of two-digit numbers: Just set the two
numbers down:
0023 X 14
130 EUREKA!
The first step is to multiply the right-hand digits. Again,
the' indicates a carry of 1:
.....--.....
0023 X 14
'2
Now the multiplication is like so:
---.......
0023 X 14
'2'2
The second 2 is the sum of (3 X 1) + (2 X 4) + carry of 1.
The last step is to mUltiply the left-hand digits:
--
0023 X 14 = 322
3'2'2
The middle step can be repeated as many times as nec-
essary, the two pairs of multiplicands creeping along left-
ward, until the last step is possible.
1. Try this one:

0065 X 38

3025 = (30 + 25)2? And 2025 = (20 + 25)2?
Squaring is another subject made easy by the Trachten-

berg System. To square a number that ends in a 5, simply
split it into two parts, the first being all the digits except
the 5, and the second just the 5. Now mUltiply the first
part by the number 1 more than it, and "join" it to 52 = 25.
The square of 75 would look like this: 7 X 8 = 56, and
52 = 25; therefore, 75 2 = 5625.
131
This method also works with numbers of any size
ending in 5, such as 765 or 9,876,234,105.
For two-digit numbers beginning with 5, merely square
the 5, add it to the unit's digit, and "join" it to the square
of the unit's digit. The square of 54, for example, is 2916.
The 29 = 52 + 4, and the 16 is the square of 4.
Generally, the first step in squaring two-digit numbers
is to square the right-hand digit. For example, to find the
square of 32:
The next step is to double the product of the digits

and add any carry from the first step:
The last step is to square the left-hand digit and add any
carry-over:
2
--=
32 1024
10'24
Two-digit squaring can also be accomplished by using
the two-digit mUltiplication method outlined earlier.
Adding long rows of numbers is another feature. You
never count past 11! You start at the top of any column
and keep a running total of the sum. As soon as it reaches
11 or more, simply make the tick mark and subtract 11
from the total.
When you reach the bottom of the column, just write
down the total and the number of ticks in that column
below it.
132 EUREKA!
Here's an example:
758
492
1165'
129'1
49
+ 122
2420 running total

0133 ticks
Now add like this, in an L on the bottom two columns

to obtain the final sum:
3 8 8 3, the answer!
The 3 to the right = 0 + 3
8=2+3+3
8=4+1+3
3=2+0+1
Checking is also easy.
2. Try this one:

3689
758
9667
1064
6498
+ 745
133
Answers
1. 0065 X 38 = 2470
24 67 40
2. 3689
15'8'
9'661
1064
64'9'8'
+ 745
8858
1233
2'2'4'2'1 = 22 ,421
134 EUREKA!
25. A Letter Home _ _ _ _ _ _ _ __
Dear Dad,
Here's a neat problem I dug up. I found the only answer
myself, so you should be able to also.
A ten-digit number contains every digit from 0 to 9.
The digits are arranged so that the number formed by the
first two digits-reading from left to right-is divisible by
2, the number formed by the first three digits by 3, the
first four by 4, and so on until the whole number is divisible
by 10.
What is the number?
Love,
David
P.S. There are two digits which can be placed immediately.

Which are they?
Answer
The number is 3,816,547,290, and the two digits re-

ferred to are the 5 and O. They can be placed immediately
because a number divisible by 10 must end in 0, so the
number divisible by 5 must end in 5.
The easiest method of solution is to use divisibility
rules to figure out the placement of sets of numbers.
For example, every other digit, starting with the second,
135
must be even; the sum of the first three digits must be
divisible by 3; the third and fourth digits must make a
number divisible by 4; the sixth, seventh, and eighth
digits must make a number divisible by 8, and so on.
Trial and error takes it from there.
136 EUREKA!
26. In Which We Are Initiated Into
the Secret Society of Square
RootSolvers _____________________
Occasionally, in the course of doing a problem, you might

find yourself in the rather unpleasant situation of having
to find the square root of a number like 1625.9444 to
three or four decimal places. With nary a calculator around,
you're tempted to throw in the pencil.
Don't give up hope! All you need to do is use the
Algorithm Method of finding square roots (so called
because the steps are repeated several times).

The product of any four consecutive integers is
always one less than a perfect square?
Step one: Just write your number under the radical sign,
with as many pairs of numbers after the decimal point as
you want places in your answer (actually, include one
extra pair of Os for rounding). This example will show how
to find the square root of 1625.9444 to three decimal
places:
V1625. 94 44 00 00
Now write the largest perfect square less than your
number right underneath it, and put its root above:
40.
V1625. 94 44 00 00
1600
137
Already you have determined that the square root of
1625.9444 is between 40 and 41.
Step two: Subtract and bring down the next pair of
numbers. This is just like long division:
40.
Y1625. 94 44 00 00
-1600
125 94
Step three: Here's the tricky part. Double the quotient
(really the approximation of the root so far) and put it to
the side:
40.
80 Y1625. 94 44 00 00
-1600
I 25 94
Step four: Estimate the next part of the quotient-
but there's a catch. The way to go about it is to estimate
80 into 259.4, or 800 into 2594. Either way you account
for an extra place. The next part of the quotient is in this
case 3 and a bit, so the digit 3 is tested. Put the 3 after the
80 and mUltiply it by 3. The product, 2409, is less than
2594, so 3 becomes the next part of the quotient. If the
product were larger than 2549, then 802 X 2 would be
the next trial.
Step five: Now write the product down under the

remainder from the subtraction and put the next part of
the quotient above:
40. 3
Y1625. 94 44 00 00
-1600
125 94
3 24 09
138 EUREKA!
Now subtract, as in step two, and bring down the next pair
of digits:
40. 3
...; 1625. 94 44 00 00
-1600
125 94
3 -24 09
[ 1 85 44
Step six: This is like step three, except that the new
testing-number can be found in two ways: You can either
double the quotient to obtain 806, or add the 803 and 3
you used in the estimation.
Step seven: This is just like step four. This time you
estimate 806 into 1854.4, or 8060 into 18,544. The pro-
cess continues in this loop indefinitely, until the number
of places wanted is achieved.
The complete example is:
40. 3 2 3 0 = 40.3230 ...
"';1625. 94 44 00 00
-1600
[ 25 94
3 -24 09
[1 85 44
2 -1 61 24
[24 20 00
3 -24 19 ;-",2=9__
171 00
0* -00 00
[71 00 ...
* The estimate was already larger than 7100 at this time,
so another set of Os (nonexistent, actually; three places
had already been achieved) was necessary.
139
QUICKIE
Try this one to two places:

-J477. 65 32 00
Got it? Check it in the Answer section.

Every odd number is the difference between two
squares?
The square root of n is the number of consecutive
odd integers that can be subtracted from n?
Answer
21. 8 5 5 = 21.855 = 21.86
-J477. 65 32 00
-441
136 65
8 -34 24
12 41 32
5 -2 18 25
123 07 00
5 -21 85 25
112175 ...
140 EUREKA!
27. Heads and Legs _ _ _ _ _ _ __
1. Herby Yore has a well-stocked barnyard. He keeps his

pigs and chickens together; in the pen there are 13 heads
and 36 legs. How many chickens does he have?
2. A traveling menagerie featured two kinds of sorry
freaks of nature-four-footed birds with two heads and six-
legged calves. An attendant, having a little fun, mentioned
to onlookers, "There are 18 heads and 100 legs together."
He was terribly surprised when one of the audience told
him how many birds there were. Can you do the same?
3. In a group of cats and canaries, the number of legs

is 84 more than twice the number of heads. How many
cats are there in the group?
Definition of Algebra: A hopeless search for the con-

tinually changing values of two mysterious letters, x and y.
Answers
1. If x is the number of chickens, then 2x is the number

of chickens' feet, 13 - x the number of pigs, and 4(13 - x)
the number of pigs' feet. Then:
2x + 4( 13 - x) = 36
2x + 52 - 4x = 36
2x = 16
x=8
Herby Yore has 8 chickens.
141
2. If x is the number of birds and y the number of
calves, then:
y + 2x = 18
6y + 4x = 100
The solution of these equations (obtained by doubling
the first to 2y + 4x = 36, and then subtracting the 36 from
the 100 and the 4x from the 4x) gives 4y = 64, so there are
16 six-legged calves and 1 two-headed, four-footed bird.
3. If x is the number of canaries and y the number of

cats, then:
2x + 4y - 84 = 2(x + y)
2x + 4y - 84 = 2x + 2y
2y = 84
y = 42
There are 42 cats and any number of canaries in the
group; as a canary has twice as many legs as heads, the
number of birds doesn't change the constant difference
of 84.
142 EUREKA!
28. Noble Bases _ _ _ _ _ _ _ __
Representing numbers in bases other than base 10 is really

much easier than it sounds. You just use powers of a num-
ber other than 10 to represent various values.
A number like 576, equivalent to 576 10 , is equal to
(5 X 10 2 ) + (7 X 10 1 ) + (6 X 10°). To change it into base
5, simply divide by the largest power of 5 less than 576,
or 125. The quotient is 4, and the remainder is 76. The
next lowest power of 5 is 25, and it goes into 763 times,
and the remainder is 1. So, 576 10 is equal to 431 5, Like-
wise, 576 10 is also equal to 1001000000 2,484 11 , 12 574 ,
and 10 576 (oddly enough, N 10 = 10N ). Notice that the
highest digit used in base N is N - 1.
Converting back is even easier. The number 1452 7 is
equal to (1 X 7 3 ) + (4 X 7 2) + (5 X 71) + (2 X 7°). If the
sum of these is found (343 + 196 + 35 + 2), 576 10 results.
Similarly, 484 11 = (4 X 112) + (8 X 1P) + (4 X 11°) = 576 10 ,
Armed with this information, these trifling problems
should be no trouble at all:
1. Mrs. Cosine has 43 5 students in her first class, 19 12
in her second, 11101 2 in the third, 417 in the fourth, and
1001 3 in her last class. How many students does she teach
each day?
2. Jack Skidder's truck contains 440 5 cases. In each

case, there are 33 7 boxes, and in each box there are 1100 2
shirts. How many shirts, in base 13 5 , is he transporting?
143
Answers
1. Mrs. Cosine daily teaches 23 + 21 + 29 + 29 + 28 =

130 pupils.
2. Jack Skidder's truck contains 120 X 24 X 12 =

34,560 10 = 103400 8 shirts.
144 EUREKA!
29. A Division in Ancient Rye _ _ __
Three travelers once met on the road at dusk and, as was

the custom, prepared to share a campsite. One of the
three, not having brought provisions, proposed that the
other two should share their food with him, for which he
would pay handsomely.
This being agreed to, the second man produced 3 loaves
of bread and the third 5 loaves, which they all shared equally.
As payment, the first man laid down 8 pieces of silver for
his share of the 8 loaves.
The second man, who had provided 3 loaves, argued
that he should get 3 of the 8 coins and the third man 5,
but the third man thought he should have more. How
should the coins be divided?
Answer
The three each ate 2~ loaves. The second man, then,

gave i loaf to the first man, keeping 2~ loaves. The third
man gave 2i loaves to the first, also keeping his share.
Hence, 2i to i, or 7 to 1, is the ratio in which the money
should be divided.
145
30. Getting at the Root
of the Problem - - - - - - - - -
"Hello, is this Mr. Rusty Springs?"

"Why, yes, it is."
"Mr. Springs, this is Guy Dence from WALF. Guess
what? You're on 'Are You aLert?' "
"I am?"
"Well, I hope you are, because you can win $1000 if
you answer our special question correctly. Ready? Here
goes, Rusty: Is the number 12,345 divisible by 3? Remem-
ber, you're on the air!"
"Well, I, uh, urn, I don't know."
"Did I hear you say no? So sorry, Mr. Springs, 12,345
is divisible by 3. Thanks for trying, from Guy Dence,
WALF. So long .... "
Had Rusty only known about digital roots, the prize

would have been his. The digital root of a number is the
one-digit sum of its digits. For example, 1 + 2 + 3 + 4 + 5 =
15, and 1 + 5 = 6, so the digital root of 12,345 is 6.
It so happens that if the digital root of a number is
divisible by 3, then so is the number. Any number with a
digital root of 3, 6, or 9, such as 12,345, is divisible by 3.
You may have noticed that the multiples of 9-18, 27,
36, 45, ... , 108, 117, and so on-have digits that add up
to 9. In fact, if a number has a digital root of 9, then it
is divisible by 9 (as well as 3).
There are several very interesting things about 9 and
digital roots, most of them arising from the fact that 9
is 1 less than 10, the base of the decimal system. In taking
146 EUREKA!
a digital root, any 9 can be cast out, so that the digital
root of 12 is the same as the digital root of 192, 912,
2991, or 9,199,299. Also, the digital root of any number
is the remainder when the number is divided by 9; 12,345
divided by 9 is 1371 with a remainder of 6. This works
because the 2, for example, can be represented as either
2 X 1000 or 2 X 999 + 2. The 2 X 999 leaves no remainder,
so the leftover 2 goes toward the digital root.
Digital roots are not only useful in checking divisibility.
They can also help determine if a large number is a perfect
square, a perfect cube, or just plain perfect. All squares
have digital roots of 1, 4,7, or 9, and cubes have digital
roots of 1, 8, or 9. All perfects except 6 have digital roots
of 1. If a number doesn't have the right digital root, then
it doesn't need to be tested in a more complicated way.
1. Find the smallest number composed only of 1s and

Os that is evenly divisible by both 3 and 25.
There is a particular dice game that uses digital roots.

The two players agree upon an arbitrary large number,
and the first player begins by rolling the die and scoring
the number that is up. The second player turns the die a
quarter-turn and adds the number on top to the running
total. The winner is the player who either lands on the goal
or forces his opponent to go above it.
The game is often played with 31 as the goal. The digital
root of 31 is 4, so the first player can force a win only by
rolling a 4 and either staying in the series 4-13-22-31 or
preventing his opponent from doing so. The game is very
hard to analyze because of the many possibilities; however,
it is known that the second player can force a win if the
digital root of the goal is 9. Otherwise, it's a dicey problem.
147
2. Jack Daniels and Johnny Walker had gotten quite
soused one wet Saturday night, and each was trying to
prove that he was as sober as a judge but that the other
was several sheets to the wind.
"Oh, yeah?" Jack whined, disappointed over losing the
most recent test. "That doesn't prove I'm drunker'n a
skunk. But this'll show who is."
"Hah! May my liver see the day!"
Jack snatched up a deck of cards from two nearby gin
players. With fingers made frisky from repeated applica-
tions of tonic, he layed out the clubs from the ace to the
nine.
"Here's a rum test I learned," he said. "I bet you can't
do it. Arrange these cards to form a prime number."
"That's no shaker! We royal Scots are prime mathe-
maticians. "
But hours later, he mumbled dispiritedly, "I give up,"
and, vaguely saying something about never having visited
Tequila, died. He was placed on his bier the next morning.
Johnny did have a tough problem, but this one is easier.
Prove that it is impossible to scramble the first nine digits
to form a prime.
Answers
1. For a number to be divisible by 3, it must have a

digital root of 3,6, or 9. A digital root of 3 is the smallest,
so the base number is 111. For it to be divisible by 25, it
must end in 25, 50, 75, or 00. The number, then is 11,100.
2. The digital root of the sum of the digits from 1 to 9

is 9, so whatever number is formed will be divisible by 9;
therefore it can't possibly be prime, or even perfect. It
might, however, be a square or a cube.
148 EUREKA!
31. Or is it 32? Remumber
Nembers __________________________
We live in a world of numbers-zip codes, street numbers,

dates, telephone numbers, Social Security numbers-and I
can't remember a single one. When I pick up the telephone,
chances are that I'll dial someone in Cinome, N .M. or
Davidle, Wis., instead of the person I want. All those long
numbers just scramble themselves in my head.
There is a solution! Memory experts have found that
remembering creative words or phrases rather than num-
bers helps commit the numbers to memory. This chart
has been developed to do just that:
0: Z or S
1: Tor D
2: N
3: M
4: R
5: L
6: J or CH or SH
7: Kor C
8: F or V
9: P or B
Let's say the number you want to remember is 43214,
a zip code. Checking the chart, the 4s can be replaced by
R, the 3 by M, the 2 by N, and the 1 by T or D. If you
now insert some vowels into your word, you can have
REMAINDER.
You can also break the number up into, say 4/32/14
if it doesn't work out so nicely. Then you can make the
149
phrase ARE MEN DEAR by adding the vowels. This works
quite well for long sets of numbers.
A room number of 679 becomes .JACOB, so you can
remember that Jacob lives in that room. Or perhaps he
lives in apartment 32, on the MOON, or even in 820,
on YENUS.
There is even a mnemonic device (one of these conve-
nient phrases) to help you remember the letters for the 10
digits in order: STNMRL(SH)CFP becomes ~ATAN
MAY RELISH COFFEEEIE!
These alphabetic mnemonics are really quite useful, but
creating one for the number 1776 is an exercise for a true
linguist. Another mnemonic device is to substitute digits
with words, rather than letters. The length of the word is
the digit. The 1776, then, can become I CREATED
NOTHING SUNDAY.
This technique has been used quite successfuly with
two useful but hard-to-remember constants, 7r and e.
"How I wish I could recollect pi easily today" gives
3.14159265, and "Now I live a drear existence in rag-
ged suits and cruel taxation suffering" carries pi out to
3.141592653589. The more absurd you make your own
phrase, the better you'll remember it.
The number e has several mnemonics: "To express e,
remember to memorize a sentence to simplify this" gives
e as 2.7182818284. "It enables a numskull to memorize
a quantity of numerals" and "I'm forming a mnemonic to
remember a function in analysis" are others.
There was another mnemonic method, number three,
but it's clean slipped my mind. I forgot the key word.
150 EUREKA!
7la Neat Numbers
32. Prime Time _ _ _ _ _ _ _ _ __
The most studied class of numbers is the class of prime

numbers. Primes are whole numbers which have no integer
divisors other than themselves and 1. Examples are 2, 3, 5,
7, 11, 13, and so on, getting scarcer as the numbers get
larger (1 is considered a prime only in England; this is
odd-it means that each prime is factorable into two primes,
a statement which doesn't seem quite right according to
our definition).
It's wonderful to know that a number like 37 is prime,
but who would believe that 9,090,909,090,909,090,909,
090,909,090,909,090,909,091 is prime? For that matter,
consider even 909,090,909,090,909,090,909,090,909,091.
151
How is it known that 1,234,567,891,234,567,891,234,
567,891 is prime? (Or so it is said.)
Well, the first method of finding primes was called the
"sieve of Eratosthenes." The person wanting to find primes
would write down all the numbers from 2 to, say, 200.
Then he would cross out all the multiples of 2 except for
2 itself. The next number not crossed out is 3, so all the
mUltiples of 3 are crossed out. Four has been crossed out,
so all the multiples of 5 are then crossed out, and so on.
This continues until the numbers left are all primes.
As you can see, this was a rather laborious and boring
method of finding primes. Besides, people needed to deter-
mine if a specific number, such as 10,000,019, was prime.
It was found that a number is prime if and only if it divided
(n - 1) ! + 1 evenly, where n is the number being tested.
Still, finding 10,000,019! is, at best, unbearable.
The most common and best method of determining
primality goes right back to the definition; to determine
if N is prime, simply divide N by all the primes less than
..[N. If a single one divides N evenly, then N is a composite
number. If N were 10,000,019, this involves dividing
10,000,019 by the 446 primes less than 3162. Fortunately,
this task readily adapts to modern electronic computers.
(As of this writing, the first 6,000,000 primes have been
compiled.)
It sure would help if there were a formula to find primes.
A lot of people thought they had found one, too. The
polynomial x 2 - 3x + 43 gives primes for all values of x
from 0 to 41, and x 2 - x + 41 gives primes for x = 0 to
40. The simple 2x 2 + 29 generates primes for all values of
x less than 29. Unfortunately, no polynomial (an equation
of this type) can give wholly prime numbers. For those
interested, the proof involves showing that the new function
of x (2x 2 + 29, for example) eventually is divisible by x.
152 EUREKA!
The pattern:
3!-2~+1!
4!- 3!+2!- 1!
5! - 4! + 3! - 2! + 1!
seems always to result in a prime. Also, the product of all
primes less than N, added to 1, also results in a prime for
many values of N. This is called Fortune's Conjecture.
The Tallman formula can boast of always giving a prime,
but some manipulation is necessary. The formula finds the
product of the first N primes, divides that by the product
of any of those primes and/or unity, and then subtracts
the second product. If the result, X, is less than the square
of the (N + l)th prime, then X is prime.

19, 109, 1009, and 10,009 are all primes?
Here is a polynomial that generates prime pairs, sets of

primes that differ by 2, such as 3 and 5, 5 and 7, 11 and
13,19,469 and 19,471, and so on:
N = 60A 2 - 1710A + 12,150
where N + 1 and N - 1 are the primes. This formula gen-
erates 18 pairs of positive prime pairs and 1 negative pair,
-31 and -29, for A = 1 to 20.
Certain primes, because of their nature, can be classified
into certain groups. The numbers 11 and 1,111,111,111,
111,111,111, composed wholly of 1s, can be referred to
as R2 and R 19 , where the subscript is the number of 1s.
Other primes of this type are R 23 and R 3I7 , a huge num-
ber that would take up the rest of the paragraph. A differ-
ent R-group, the Robinson primes, is made up of primes of
153
the form 2 n k + 1, abbreviated R(k, n). The largest known
Robinson prime is R(5, 1947), a number of 586 digits.
Another class of primes is of the form 2 2n + 1; these are
called Fermat primes. For n = 0 to 4, the Fermat primes 3,
5, 17, 257, and 65,537 are obtained, but F 5 is factorable.
(Polygons withn sides, when n is a Fermat prime or the prod-
uct of Fermat primes, can be drawn using only a compass
and straight edge. The method was discovered by the
famous Karl Friedrich Gauss in 1796, and is a procedure
that eluded mathematicians for 2000 years. Gauss was
not yet twenty at the time.) Oddly enough, F 7 was deter-
mined to be composite 65 years before its factors were
found. F 8 is also known to be composite, but its prime
factorization is not yet known. F 1945 is the largest Fermat
number tested for primality, a gargantuan containing about
1010584 digits! It is also so large that its complete factor-
ization is not known; however, one of its prime factors
is the 586-digit Robinson prime R(5, 1947).
Now we come to the most famous class of prime num-
bers, the Mersenne primes, named after Father Marin
Mersenne. In 1644, in the course of announcing the dis-
covery of new perfect numbers, he stated that the form
2P - 1 is prime for many prime values of p, such as 2, 3,5,
7, 13, 17, 19, 31, 61, 89, 107, ... , up to and including
11,213, 19,937, 21,701, and 23,209, the last values giving
primes of 6000 digits or more.
Testing these numbers for primality by dividing by all
the primes less than Y223.20L 1, for instance, is ridiculous.
In 1876, Edouard A. Lucas devised a special method of
testing Mersenne numbers. He arranged a series 4,14, 194,
37,634, ... , where each term is 2 less than the square of
the previous term. A number is prime if it divides the
(p - 1 )th term evenly.
How it works: Let's suppose that we're testing 27 - 1 =
154 EUREKA!
127 for primality. (Notice that the three digits on the left
are the same as the digits on the right. Also, 27 = 128;
1, 2, and 8 are all powers of 2. This is unique.) The first term
of Lucas's series that is larger than 127 is 194, the third term.
Dividing 194 by 127 leaves a remainder of 67. The fourth
term is then 67 2 - 2 = 4487, which leaves a remainder of
42 when divided by 127. The fifth term of the series is
then 422 - 2 = 1762, and the sixth, the (p - l)th term is
then 1112 - 2 = 12,319, which leaves no remainder when
divided by 127. Therefore, 127 is prime. Even this method
is preferable to dividing large Mersenne numbers by thou-
sands of primes.
For every Mersenne prime, there corresponds a perfect
number! (A perfect number is one whose factors add up
to it.) Simply multiply 2P - 1 by 2P 1, and a perfect
number is sure to result. Furthermore, there is no known
perfect number that does not correspond to a Mersenne
prime.
There are all sorts of interesting facts about primes of
the form 4n + 1, such as 5, 13, 17, and so on. According to
Fermat's Two Square Theorem, every prime of this type
can be represented as the sum of two squares, as 13 = 4 + 9,
but no prime of the form 4n - 1 can be. Oddly enough,
a prime of the form 4n + 1 is also once the hypotenuse of
a right triangle, its square is twice, its cube thrice, and so on.
Just how many primes are there? It has been proven
that there is no largest prime, only a largest known prime
(the proof involves showing that the product of any
prime from 2 to m added to 1 either is prime itself or has
a prime factor larger than m, as it can't be divided by any
prime smaller than m). As mentioned before, primes
become scarcer and scarcer, approaching a density of N /
(In N), or N divided by the number that e must be raised
to in order to get N.
155
There are many unsolved problems, conjectures, and
theorems concerned with primes. Bertrand's Postulate,
still not proven, states that there is at least one prime
between nand 2n. Another unsolved problem: Is there
always at least one prime between every set of two con-
secutive squares? [Hint: The density of perfect squares
approaches 1!(2yN}] Goldbach's Theorem says that
every even number is the sum of, at most, two primes.
(So far, it has been proved that four primes will always
suffice.) Does the theorem also apply to odd numbers?
One odd pattern of primes is 333,337, 33,337, 3337,
337, 37, 7, in which each new prime is obtained by remov-
ing the left-hand digit of the previous prime. Aprime-prime
is just the opposite-each new prime is made from all but
the right-hand digits of the previous prime. One example
is 317, which becomes 31 and 3; another is 31,379. Some
others are 8 digits long!
One last oddity: Primes can be in additive series, too.
For instance, the series 5, 11, 17, 23, 29 has a common
difference of 6. A longer chain of 10 terms starts with
199 and has a common difference of 210. A chain discov-
ered in 1969 has 16 terms; it begins with 2,236,133,941
and has a common difference of 223,092, 870. You might
want to search for series in which the difference between
terms doubles, for instance.
1. Two women, mother and daughter, were celebrating

their birthdays, which fell on the same day.
"N ow, I remember," reminisced Geri Atrics, "that last
year I was just twice your age."
"That's true, Mom," Eva Porate agreed. "Have you
noticed that both our ages are prime numbers now?"
"Indeed, Eva," the elderly math teacher said. "What
is more remarkable is that your age is the exact reverse of
mine."
156 EUREKA!
What are their ages?
2. Frank A. Praisal, an assessor, was commenting on his

income tax form. "I noticed that the ages of myself, my
wife, and my son were all prime numbers and totaled 101.
In 6 years they will all be primes again, and my wife's age
and mine will total 100," marveled Frank. "How old is
my son, and can he be listed as a dependent?"
3. The members of the Princesston University math

team had finished their high-stakes card game, and the
debts were being settled. Jay Walker, the big loser, placed
his check on the table and said, "There's a prime number
of dollars for all of you." The other loser, who paid with
four bills of different denominations, was surprised to find
that he had also lost a prime number of dollars.
"Well, I'll take a prime number and the rest of you can
split up a prime number," Bob Sledd said as he picked up
the check and laid down a hundred-dollar bill. Four
other winners each took a single bill, surprisingly always
leaving a prime number of dollars, until the last winner
took a prime number and left nothing.
What is the smallest number of dollars that could have
changed hands?
Answers
1. Geri Atrics is 73 and her daughter Eva is 37.
2. The son is now 13, Frank is 47 or 41, and Mrs.

Praisal is 41 or 47.
3. The second loser paid with four bills, here called

157
a, b, c, and d, such that d is the largest and a the smallest,
the sum of the bills is less than $100, and:
a is prime
a + b is prime
a + b + C is prime
a + b + C + d is prime
and a + b + C + d + 100 is prime
As the three smallest denomination bills, 1, 2, and 5
dollars, add up to a composite number, c must be larger
than 5 and must be a multiple of 10 dollars. Thus, the
combinations of a and b are limited to 2 and 5 so that
both a and a + b are prime. If a table is now set up with
varying dollar amounts for c and d, three solutions are
obtained. The smallest of these is: $227 + $37 = $264
lost; $2 + $5 + $10 + $20 + $100 + $127 = $264 won.
A total of $264 changed hands.
158 EUREKA!
33. A Sense of Balance _ _ _ _ _ __
1. Al Cohol, suffering from a late-night bout with the

bottle, groggily noticed some unusual things about the
kitchenwares he was transporting. He found that a jug,
placed on the kitchen scale, balanced perfectly with a
bottle. By experimenting a little, he also determined that
a jug also balanced a cup and saucer, and that two bottles
balanced three saucers. How many cups do you think it
will take to balance a jug?
2. In a tug of war, four Briztles can pull as hard as five

Gartxibles. One Briztle and two Gartxibles can hold their
own against one mighty Dyquixt. If one Dyquixt and three
Gartxibles waged a war of tugs with four Briztles, which
side would win?
3. Miss Fortune was especially proud of the latest

additions to her coin collection-nine valuable coins that
looked exactly alike.
"This one is worth the most," she said, removing it
from its special case and comparing it with the others.
"It weighs just a fraction of a gram more than the others."
As she hefted it, the coin slipped from her hand and fell
into the pile.
"Oh, bother," she said. "Now I'll have to weigh them all
to find the heaviest coin. That could take eight weighings."
"Not so," disagreed Ty Kune. "I can find it for you in
just two."
How did he propose to do this?
159
Answers
1. Three cups balance a jug.

If we use the letters j, b, c, and s for jug, bottle, cup,
and saucer, AI's first experiment showed thatj = b, or that
a jug weighs the same as a bottle. The second showed that
j = c + s, which means also the b = c + s, as j = b. By
multiplying by 3, the equation 3b = 3c + 3s is obtained.
The last statement shows that 2b = 3s, so this is nothing
more than 3b = 3c + 2b, by substitution. Then, b = 3c,
and j = 3c. That is, a jug balances three cups.
2. In the last event, you can replace the Dyquixt with

two Gartxibles and one Briztle. Then the tug is between
four Briztles and five Gartxibles plus one Briztle. The first
event showed that four Briztles tug as hard as five Gartxi-
bles, so the team of three Gartxibles and one Dyquixt
should win.
3. What Ty Kune proposes to do is weigh any three

coins against any three others. If one set is heavier, he
would weigh any two of the heavier set against each other;
either one of those two is heavier, or, if they weigh the
same, the third one is the heaviest.
If the two sets balance, he would weigh any two of the
other set against each other; again, either one of those two
is heavier, or, if they balance, the third is the heaviest.
160 EUREKA!
34. Perfect Numbers and Some
Not-So-Perfect Numbers _ _ _ __
Perfect numbers, in contrast to prime numbers, do have

integer divisors. Euclid, founder of Euclidean Geometry,
defined a perfect number as one which is equal to the sum
of all its different divisors. The number itself is not in-
cluded, or only 1 would be perfect, as it is its own divisor;
for this reason, 1 is also not considered prime. The Greeks
thought of the property of some numbers as "perfect."
There is also a class of Multiperfect numbers, in which the
sum of the divisors is a multiple of the number. For ex-
ample, the sum of the factors of 120 equals 240 (first
pointed out by Mersenne in 1631), and the sum of the
factors of 672 is 1344, again twice it. Other numbers,
considerably larger, have divisors adding up to 3, 4, 5, 6,
7, or 8 times the number.
But back to mere perfects: For instance, 6 is equal to
1 + 2 + 3, and 28 = 1 + 2 + 4 + 7 + 14. These are the first
two perfects. The next perfect numbers are 496, 8128,
and 33,550,336, the fifth anonymously discovered around
1460. Perfect numbers are scarce, and so become large
quickly. One of the last known perfects, the 24th, has
12,003 digits!
You may have noticed that all these perfect numbers
end in 6 or 8, and perhaps wondered if there are any odd
perfects. Well, there are none smaller than 10 18 , but if
they do exist, and there is some doubt that they do, they
must be of the form 12m + 1 or 36m + 9, where m is a
prime number.
161
Primes also play a part in the structure of even perfects.
Euler proved in 1750 that all even perfects are of the form
2P - 1 (2 P - 1), where p and 2P - 1 (a Mersenne number)
are prime. If p = 2, the expression gives 6, a perfect, for
instance, and 2 11 ,212 (2 11 ,213 - 1) is one of the larger
perfects. Unfortunately, the number obtained is not perfect
if p = 11; though 11 is prime, 211 - 1 is not.
Just how many perfect numbers are there? We have seen
that each Mersenne prime has a perfect counterpart, so
determining how many primes of the form 2P - 1 there
are should tell you how many perfects there are. Still,
proving that there is not an infinity of Mersenne primes
(if there were, there would be at least as many, if not more,
perfects) would not solve the problem of the number of
perfect numbers until the question of odd perfects is
resolved.
Six is the "odd" perfect number. Besides the fact that
the product of its factors is the same as their sum, 6 does
not obey many of the rules that perfects seem to follow
(also, 6 is the only number less than 10 that is not either
a prime or the power of a prime). All known perfects,
except 6, have digital roots of 1; that is, the ultimate sum
of their digits is 1. For example, 4 + 9 + 6 = 19; 1 + 9 = 10;
1+0=1.
Every known perfect number, save 6, is also the sum of
consecutive odd cubes, beginning with 1 and continuing
until the number of cubes is equal to p-=r. Must p be
prime here?
28 = 1 3 + 33
496 = 1 3 + 3 3 + 53 + 73
8128 = 1 3 + 3 3 + 53 + 7 3 + 9 3 + 11 3 + 13 3 + 15 3
162 EUREKA!
Perfect numbers are also the sums of successive powers
of 2 from 2P - 1 to 2 2p - 2 • This rule the number 6 obeys!
6 = 21 + 22
28 = 22 + 2 3 + 24
496 = 24 + 2 5 + 26 + 27 + 2 8
8128=2 6 +27 +28 +29 +210 +211 +212
Again, must p and the number of terms be prime?
Another "law" of perfect numbers that 6 obeys is that
the sum of the reciprocals of its factors is 2. That is, t + ~
i
+ 1- + = 2. This holds for all perfect numbers, and only for
perfect numbers. Also, all perfect numbers, 6 included, are
the sums of consecutive integers starting with 1.
Amicable numbers are perfect numbers one step removed.
If the members of such a pair are a and b, then the factors
of a add up to b, and the factors of b add up to a. The
best-known and smallest pair is 220 and 284. Amicables
differ from perfects in the fact that odd pairs, the smallest
being 12,285 and 14,595, are known. All odd pairs are also
divisible by 3, the lowest odd number other than 1, just
as even amicable numbers are divisible by 2.
Amicable numbers in a chain, in which the factors of
one number equal the next, form what is called a sociable
chain. One example is 12,496 -+ 14,288 -+ 15,472
-+ 14,264 -+ 12,496. (The arrows indicate that the
factors of one number form the next number.) Chains
of one link are perfect numbers, chains of two links are
amicable pairs, and chains of four, five, nine, and more
links are sociable chains. One mammoth chain, known for
60 years, has 28 links! Imagine figuring out that one
without a computer!
163
Cranium
Crackers
~o and Cheese:
Problems to Munch On
35. Classy Problems - _ _ _ _ _ __
There are quite a few ancient and decaying problems that

are actually improved by twisting them around. Here is a
small collection of the classic and modern versions of
several rejuvenated antiques.
la. Sam Ovar, the famous Russian tea magnate, was

trying to figure out an efficient way of dividing his facto-
ries. He discovered that when he grouped them by 2s, 3s,
4s, or 5s, he always had 1 left over. What is the smallest
number of factories he could have had before he sold that
extra l?
165
lb. Several years later, a fortune-teller looked into her
teacup long and hard and told Sam that he simply must
reorganize his factories again. When he divided them into
2 groups, he had 1 left over again, but when he divided
them into 3s, he had 2 remaining, by 4s he had 3 remain-
ing, and by 5s he had 4 remaining. What is the smallest
number of factories he could have had?
2a. Nancy France, away on a country vacation from her

thriving resort hotel, came across an interesting problem.
She, for some unknown reason, wished to cross a raging
stream with a fox, a goose, and a sack of corn. The raft
she wanted to use could carry only two things at once. If
left alone, though, the fox would eat the goose or the
goose would eat the corn. How on earth did she ever cross
the creek?
2b. The fox in the above problem was none other than
the Country Fox, Ursula Major. She owned the goose, of
course. Nancy and Ursula found themselves in an awkward
position. The raft would hold only two things, and the
goose couldn't be left alone with the corn. To make matters
worse, Ursula didn't trust Nancy alone with her goose and
Nancy didn't trust Ursula alone with the corn. How did
the four cross the river?
3a. Three tycoons, Smith, Brown, and Jones, live in

Alfal, Fla. Three teachers with the same names as the
tycoons live in the same town. Brown and the gym teacher
live in the east end, Jones and the science teacher live in
the west end, and Smith and the English teacher live in the
south end. The gym teacher's namesake earns $100,000 a
year, and the English teacher earns precisely one-third of
the tycoon nearest him. If the teacher named Smith beats
the science teacher at tiddlywinks, what is the English
teacher's name?
166 EUREKA!
3b. Here's an odd bridge situation. Four players are
named Arthur George, George Henry, Henry Thomas, and
Thomas Arthur. Henry and George are playing against
Henry and Arthur, and Thomas is sitting at Thomas's
left. If partners sit opposite one another, who is Thomas
Arthur's partner?
4a. Kim Ono had several robes which she distributed

among her daughters. To the eldest she gave half the robes
and half a robe. To the middle daughter she gave half of
those left and half a robe, and to the youngest she gave
half the robes and half a robe. Kim then had no robes left.
How many did she start with?
4b. One day Kim decided to go wholesale. She packed
up all the robes she had made into a truck and drove to a
dealers' convention. The first customer bought half the
robes and half a robe. The second and third customers
did the same when their turns came. She then had just a
quarter of a gross left. How many did she start with?
5a. The classic problem of the eight queens is easier

to ask than to solve: How do you put eight pieces on an
8 X 8 chessboard so that no two pieces are on the same
horizontal, vertical, or diagonal line? There are twelve
solutions.
5b. The next step up is harder. Try placing sixteen
pieces on the same board so that two and only two are on
each row and column, and two at the most are on each
diagonal.
6a. Henry Ernest Dudeney's best known puzzle is the

problem of the spider and the fly. As shown in Fig. 35-1,
the spider is one foot from the ceiling at the middle of the
end wall of a room, and the fly is one foot from the floor
at the opposite end of the room, perhaps caught in a web.
167
What is the shortest path the spider must crawl to reach
the fly?
12
30
Fig. 35·1
6b. A less famous problem is this one. The fly is one

inch down the inside of a cylinder four inches high and six
inches in circumference. The spider is opposite, one inch
from the bottom on the outside. How far must it crawl
this time? See Fig. 35-2.
168 EUREKA!
Fig. 35·2
Answers
1a. The smallest number is 61. It is obtained by finding

the smallest number all divide into. Because 4 has a factor
of 2, the number is not 120 but 60. When the extra factory
is added, 61 is the answer.
lb. Each division falls just one short of being exact.
The smallest number is then 60 - 1, or 59.
2a. Nancy took the goose across and left it there. She
then took the corn across and took the goose back. She
dropped the goose off and took the fox across. Then she
went back and took the goose across, only to find that
the fox had eaten the corn.
2b. Nancy first took the corn across and left it there.
She went back and picked up Ursula, poled her across, and
got off with her corn. Ursula went back and took her goose
169
across. Once there, they found the campsite they had been
looking for and toasted each other's brains over roast
goose and corn.
3a. The English teacher's name is Smith. From the last

statement, we know that Smith is not the science teacher.
One of the earlier statements says that the gym teacher's
namesake earns $100,000 yearly. Because the English
teacher earns one-third of this, the nearest tycoon is not
named for the gym teacher. The English teacher lives in
the same district as Smith, so Smith isn't the gym teacher.
Therefore, he is the English teacher.
3b. From the second statement, the one about the pair-
ings, it is clear that the two Henrys are opponents. Because
the two Thomases are sitting next to each other, then two
Henrys and two Thomases are opposed. Therefore, Henry
Thomas and Arthur George are opposed by Thomas Arthur
and George Henry. George Henry is the partner.
4a. Kim Ono started with seven robes. The problem is

best solved by working backwards. To the youngest daugh-
ter she must have given one robe in order to have none left.
In order to have one left, she must have had three before
she gave robes to the middle daughter, and she must have
had seven to start with.
4b. She started with 295. If she had 36 left after the
third scale, she must have had 36 + 37 before it, and 73 + 74
before the second, and 147+ 148 = 295 at the start of the
day.
170 EUREKA!
5a. Figure 35-3 shows what I feel is the most attractive
of the twelve.
Fig. 35-3
5b. One solution is illustrated in Fig. 35-4.
Fig. 35-4
171
6a. Forty feet. The problem is best solved by unfolding
the room. See Fig. 35-5.
/
;/ 24
/ 32
Fig. 35·5
6b. The same principle is used here. The minimum walk

is five inches, as can be seen by using the unfolded cylinder
(Fig. 35-6).
172 EUREKA!
4
6
Fig. 35·6
173
36. LETTERS + DIGITS =
FRUSTRATION _ _ _ _ _ __
Alphametics and cryptarithms fall into one of those cate-

gories of problems that you either love or hate. These
ultimate word problems are gems in which a simple calcula-
tion such as addition or division is performed, but each
digit in the equation is replaced by a different letter. The
puzzle is to find the original equation; it's kind of like
being given the answer and being asked to find the question.
The "letter-arithmetic" started thousands of years ago,
but it was only in 1931 that the term "cryptarithm" was
used for a puzzle like this one:
ABC
DE
FEC
DEC
HGBC
Not very appealing, is it? "Alphametic," a typo of alpha-
betic, was used in 1955 to represent puzzles in which
phrases and words were used:
BUT
WE
GET
WET
LOUT
Believe it or not, this IS the same puzzle. Much more
174 EUREKA!
approachable, the challenge is still to find the original
equation-this one is mUltiplication.
This alphametic can be solved through the following
steps (proceed with caution):
1. The multiplication of Wand B results in the W in
the fourth line. Therefore, B equals l.
2. E and T are used a lot, and the product of E and T
has an end digit of T (that is, T . E -+ T). The only values
that work, Is and double digits excluded, are 0 . anything
- 0; 2 . 6 -+ 2; 4 . 6 -+ 4; 5 . 3 -+ 5; 5 . 9 -+ 5; and 8 . 6 -+ 8.
Now, W . T -+ T also, so T must be 0 or 5 to have two
values that result in T. If T = 0, then, because E . W
-+ E in line 2, V would have to be 1. This can't be, because
B is 1. Therefore, T = 5.
3. If T equals 5, then E and W must be chosen from 3,
7, and 9. Because W + 1 = L, a single digit, W cannot
equal 9.
4. E cannot be 9, either. The product E . B does not
result in E, so there must be a carry-over. But 9 plus any
carry-over has two digits. Therefore, E doesn't equal 9.
5. If E = 3, then W = 7, because those are the only
choices left. Because W . T -+ E, that would mean that
7 . 5 -+ 3. This isn't true, so E doesn't equal 3.
6. Therefore, E equals 7, and W equals 3. Because
L = W + 1, L is 4. The sum E + T yields V, so V = 7 + 5
-+ 2. The product E· V, or 7 ·2, has a carry of 1, so
G = E + 1 = 8. Then 0 equals 6.
7. Relax and enjoy it. There are more to come. The
final equation is this:
125
X 37
875
375
4625
175
A couple of hints may help you solve the following
problems. If N is an even digit, then N . 6 ~ 6. If N is
odd, then N . 5 ~ 5. And if N . N = M, then M is 0, 1,
4, 5, 6, or 9. Good luck!
1. Here's a classic:
SEND
MORE
MONEY
2. An easy warm-up:
LOSE
SEAL
SALES
3. Is this addition or subtraction?
TRIED
DRIVE
RIVET
4. Two odd repetitions:
a. ABCDE4
4
4ABCDE
b. ABCDE
4
EDCBA
176 EUREKA!
5. A Joseph Trevor special. P stands for a prime number,
either 2, 3, 5, or 7:
PPP
PP
PPPP
PPPP
PPPPP
Answers
1. M immediately equals 1, because it is a carry-over.

a equals zero, because the carry-over cannot result in
11 or 12.
S must be 8 or 9. If it is 8, then E = 9 because there
must be a carry-over to S + M. But then both N and a
would be O. Therefore, S = 9.
N + R > 10, so E + carry of 1 = N; and N = E + 1.
E does not equal 2, because then N would be 3 and R
would be 8 so that 3 + 8 + 1 ~ 2. Then D + E > 10 also.
But D would have to be 8 or 9, and both digits would be
used.
E does not equal 3, because then N = 4 and R = 8.
But D as 7 makes D + E = 7 + 3 ~ 0, already used. E
does not equal 4, because R as 8 and D as 6 or 7 results in
y as 0 or 1, both already used.
If E = 5, then R can be 8, D can be 7, Y can be 2, and
N can be 6.
The solution:
9567
+ 1085
10652
177
2. S must be 1 and A must be 0, as can be seen from
the previous problem.
S + A ~ E, so, because E can't be 1, E = 2.
E + L ~ S, so 2 + L ~ 1, and L = 9.
o + E ~ L, so 0 + 2 = 9, and 0 = 7.
The solved problem:
9712
+ 1209
10921
3. The problem can't be subtraction because T - D = R,

so D< T. Then the E from E - V ~ E would have to be
reduced by 1 to increase the D. V would be 9, but then so
would I.
So this is addition. From E + V ~ E, V equals either
o or 9.
If V = 0, then 1= 5, so R = 2 or 7. But T + D = R, so R
could only be 7, making T + D = 6. But D + E ~ T and
T > D (no carry, remember), so D = 2, T = 4, and E = 2.
But then two digits are 2, so V can't be O.
So V = 9. Then 1= 4, making R = 7. T + D = 6 still, but
T < D this time, so T = 1, D = 5, and E = 6.
The final equation:
17465
+ 57496
74961
4a. Because 4· 4 ~ E, E = 6. D is found from E ·4+ 1

carry, so 25 ~ D, and D = 5.
4 . 5 + carry of 2 yields 22, so C = 2.
4 . 2 + carry of 2 yields 10, so B = O.
4 ·0+ carry of 1 yields 1, so A = 1.
178 EUREKA!
The final equation is:
102564
X 4
410256
4b. Because it is the unit's digit of a multiple of 4, A
must be even.
A must therefore be equal to 2, because A . 4 = E, a
single digit.
E is therefore 8.
B must be 1 so that 4 . B is a single digit.
D . 4 + carry of 3 from 32 ~ 1, so D = 2 or 7. A = 2
already, so D = 7.
4 . 1 + carry ~ 7, so carry equals 3. Then 4 . C has a
lOs digit of 3, so C = 9.
The solved puzzle:
21978
X
--
4
87912
5. This one is best solved by finding three-digit numbers,

all of primes, that produce four-digit numbers, all of primes,
when multiplied by a prime.
The only possibilities are:
775·3 = 2325
555·5 = 2775
755·5 = 3775
325·7 = 2275
The multiplier PP must consist of two identical digits,

because there are no repetitions of any three-digit numbers.
179
There are only four possibilities to be tried, of which
one works.
The answer is:
775
X 33
2325
2325
25575
HOPE
YOU
HADA
+ GOOD
TIME
180 EUREKA!
~o FUNdamental
Ratios
37. Expand Your Mind _ _ _ _ _ __
Here are some absolutely unbelievable problems that are

guaranteed to amaze and astound.
1. If you were to lengthen your belt by 2 inches, how

much space would develop between the belt and your
body? (Assume that you are a perfect sphere.)
To be more realistic, if a cable 2 inches longer than the
circumference of the earth were to be placed around the
world, how much space would there be between the cable
and the surface?
181
2. The mile-long Frog's Neck Bridge, in Proge, N.Y.,
expands just 2 feet on a hot day. If there were no expansion
joints to take up the extra length, how tall a bump would
be formed in the bridge?
Answers
1. If D is the diameter of any circle, sphere, or midriff,

then Drr is the circumference and Drr + 2 is the length of
the belt or cable. The new diameter, then, is (Drr + 2)/rr,
so the space on each side is equal to:
[(Drr + 2)/rr] - D
2
which reduces to (2/rr)/ 2, or lIrr, about 0.3183099 ...
inches, no matter what the original diameter!
2. If the bump were in the middle of the bridge, it

would be the third side of a right triangle whose other two
sides are ~ mile and ~ mile plus 1 foot (half the 2-foot
expansion).
Then b 2 + 2640 2 = 2641 2 by the Pythagorean Theorem.
This means that b 2 = 5281, or that the bump is nearly
73 feet tall! Drivers beware!
182 EUREKA!
38. E? Ah! _ _ _ _ _ _ _ _ _ __
Let's just say you're out with a friend. In the course of

casual conversation, your friend mentions that she has
opened a new savings account at the bank.
"It's for my odd-job money. Mom says saving it will be
better than spending it all before prices rise."
Half-interested, you idly gaze at a hanging chain.
"It's at only 4% interest a year, but the interest is com-
pounded every week."
"What?" You rise out of your stupor.
"At the end of every week, the bank assesses the interest
and adds it to my account. Then the next payment is
interest on the interest."
"Wow! How much do you have in there?"
"Only a dollar so far," her eyes light up with expectation.
"But you see it's going to grow very quickly!"
You nod, but something doesn't seem quite right. And,
little by little, it begins to dawn on you.
Let's say your friend's bank didn't pay compound
interest, but merely added 4 cents a year to her dollar.
At the end of 10 years, her dollar would have grown to
$1.40, and at the end of 25 years, the amount would have
doubled. She would now possess 2 dollars.
On the other hand, let's say her bank does pay compound
interest. Just for the sake of argument, it compounds each
account yearly. In 25 years, her dollar would grow to
(1 + 1/25)25 dollars, or a little more than $2.66. Com-
pounded every 6 months, the amount would grow to
(1 + 1/50)50 dollars, or $2.69. But let's say the bank
183
compounds at the extraordinary rate of 52 times a year.
You would expect the dollar to grow to quite a large sum-
if it were given enough time.
And that's where she was caught.
At the end of 25 years (or 1300 weeks), the dollar
would have grown to (1 + 1/1300)1300 dollars,or just about
$2.717, 2 1/2 cents more than what she would have made
if the dollar had been compounded twice a year.
And if the interest were to be compounded a million
times a year, what kind of difference would that make?
Your friend would earn a little more than one-tenth of a
cent over those 25 years than she would with weekly
compounding.
In fact, if the bank devoted every second of its time to
compounding that lone dollar for the next 25 years, the
amount of money at the end·would be bounded between
(1 + 1/n)n and (1 + 1/nt+ 1, where n is twice the largest
number you can think of. The amount would approach
$2.718281828459 ... , a never-ending, nonrepeating dec-
imal. It doesn't matter if the bank pays 10%, 50%, or
172.6% interest. In the same time that the dollar takes to
double at simple interest, at compound interest it ap-
proaches that mysterious 2.718281828 ....
And that is the number e.
Like pi, e is a transcendental number. This means that
neither number can be expressed as the solution to any
algebraic equation, nor be equal to any line segment drawn
using the classic tools, a compass and straight edge.
Thus e can be expressed only as an endless, continuing
fraction or as the sum of an infinite series. For instance:
e=2+1
1+1
2+2
3 + 3 ...
184 EUREKA!
(This continued fraction was discovered by Leonhard
Euler. He was also the first to use the letter e for the
number, and made so many discoveries about it that e
is frequently called Euler's Number.)
If the expression (1 + lin)" is expanded, one obtains
the following series:
e = 1 + (1/1!) + (1/2!) + (1/3!) + (1/4!) ...
And eX = (xo 10!) + (Xl 11!) + (x 212!) + (x 313!) + .
+ (x" In!) ....
By rearranging the terms, e can also be represented as
(2/1!) + (4/3!) + (6/5!) + (817!) ....
This little number pops up everywhere. Remember the
hanging chain you idly eyed in the beginning of this
Chapter? You got it. The curve, called a catenary curve,
contains e in its equation. In growths of items in which
the rate of growth is proportional to the size of the quan tity,
e is also present; this is the case, in part, with the dollar
and with the world population, for instance, among many
other natural phenomena. The larger the snowball, the
faster it gathers snow.
1. A problem comes to mind here. If a culture of bac-

teria doubles in population every hour and fills a test tube
after 27 hours, at what point is the test tube precisely
half-full?
These processes are described by formulas based on

y = eX; this is called the exponential function. In the world
of calculus, e is its own derivative, and e, rather than 10,
is the base of natural logarithms. Lastly, e is linked in this
formula with 'IT and i, the imaginary square root of -1:
e irr + 1 = 0, joining two transcendental numbers, unity,
null, and an imaginary number in one equation.
185
2. Another little problem: Which is greater, err or rr e ?
Try to solve it without the use of a calculator.
Permutation involves factorial, so one would expect

to find e involved in problems using permutation. Consider
the following: 10 high school students leave 10 notebooks
in a class. If the notebooks are later returned randomly to
the students, what is the probability that no student re-
ceives the correct book?
Firstly, there are 10!, or 3,628,000, possible arrange-
ments of the 10 notebooks. It so happens that there are
about 10!/e arrangements in which all the notebooks are
wrong. This means that the probability of all the note-
books being wrong is (10!/e)10!, which is equal to lIe,
and lie is about 0.3678794411 .... Thus, about 37 out
of 100 times all the books will be returned to the wrong
owner! Again, it should be no surprise to learn that the
same number applies just as well to 10,000 students as
to 10 billion. The probability is still about 3/8 that none of
the participants will receive the right property.
Since either all notebooks are wrong or at least one is
correct, the probability of at least one student getting his
or her notebook back is just about 5/8, no matter how many
are involved.
You might like to try this little stunt. Deal a shuffled
deck of cards face up as you recite the 52 cards in some
prearranged order. You win if you name the dealt card.
Sound hard? Your chances of winning are l-(l/e), or
about 63/100. Try it on the nearest gullible millionaire.
Memorizing e? You automatically know e to nine places
if you know it to five, as four digits repeat (2.718281828
... ). Otherwise, here are a few catchy mnemonics for e,
in which the number of letters in successive words gives
successive digits:
186 EUREKA!
"To express e, remember to memorize a sentence to
simplify this."
"It enables a numskull to memorize a quantity of nu-
merals."
"I'm forming a mnemonic to remember a function in
analysis. "
And, lastly, I'll leave you with two small problems.
3. What value of n gives the maximum value for the nth

root of n?
4. To express e to six decimal places, there must be at

least four digits in both the numerator and the denominator
of the approximating fraction (as in 272111001). Can you
find a three-digit fraction accurate to four places and a
two-digit fraction accurate to three places?
Answers
1. The test tube was half-full at the 26th hour.
2. The greater is e1T , not 1fe.
3. The maximum value of the nth root of n is given bye.
4. The fractions are 878/323 (2.71826 ... ) and 87/32

(2.71875).
187
39. A Section of Gold _ _ _ _ _ __
Phi is one of the more interesting irrational numbers

(numbers whose decimal expansions are unending and
nonrepeating). Not as well known as 1f or e, phi (symbol-
ized by the Greek letter ¢) represents a fundamental ratio
that pleasingly appears in many areas of mathematics.
A B
Fig. 39-1
The line in Fig. 39-1 has been divided into what is called
the "Golden Ratio" or "Golden Section." The lengths of
A and B are such that A + B is to A as A is to B. If B is
equal to 1, the equation becomes:
A + 1 =A
-- -
A 1
A+1=A2
0=A2_A-1
Solved, this quadratic equation has the positive value of ¢,
(1 + y'5 )/2, or 1.61803398 .... If A instead is 1 (by the
way, if A = 1 mile, then B will be extremely close to 1
kilometer or about 0.62 miles, very close to 1!¢), the solu-
tion is equal to (y'5 - 1)/2, which, oddly enough, is equal
188 EUREKA!
to l/¢, or 0.61803398 .... Phi is the only positive num-
ber that becomes its reciprocal when 1 is subtracted (a
negative number is -l/¢).
There's even a little more to this: ¢ + 1 = 2.61803398
... , and ¢2 is also equal to 2.61803398 .... This series
of 1, ¢, ¢ + 1 can be continued as 1, ¢, ¢ + 1, 2¢ + 1,
3¢ + 2, ... , where every term is the sum of the previous
two. Amazingly, this series is equal to 1, ¢, ¢2, ¢3, ¢4, ....
This is the only additive series in which the ratio between
consecutive terms is constant.
If a sq uare is to be cut so that the lengths of the pieces
are in an additive series, such as 1, 1, 2, 3, 5, 8, 13, ... ,
and the pieces are to be assembled into a rectangle so that
no area is lost or gained, the pieces must be cut according
to the phi series shown above; otherwise, as the consec-
utive terms of any additive series have a ratio that ap-
proaches ¢, loss or gain of area must result.
It's as easy as pie-in fact even easier than pi-to repre-
sent ¢ as the sum of an infinite series:
¢=V1+
V1+
vr+
V1+
VI + ...
or:
¢=1+1
1+1
1+1
---
1+· ..
189
Phi can even be substituted for part of the second series.
This yields </> = 1 + 1/</> and </>2 = </> + 1, which becomes the
quadratic equation shown earlier. This proves that this se-
ries is correct (or that cJ> is eq ual to itself).
Fig. 39·2
Phi is also aesthetically pleasing. The Golden Rectangle

(Fig. 39-2), whose sides are in the ratio of </>, has proven to
be the one consistently used in everything from skyscrapers
to doors to clothing to paintings. Leonardo da Vinci, for
instance, applied the Golden Ratio to many of his works,
connecting major figures in his paintings with the use of
invisible Golden Rectangles.
The ancient Greeks were well aware of </>. The Parthenon,
a famous surviving temple, exhibits the Golden Ratio in its
proportions. Phidias, a great sculptor of classical Greece,
is believed to have used </> frequently in his work; perhaps
he realized that the human body is divided into the Golden
Section at the hip.
Even the Egyptians were aware of </>. Recent measure-
ments of the Great Pyramid of Giza (built about 3070
B.C.) reveal that the ratio of the length of the slant edge to
190 EUREKA!
the distance from the base center to the edge is almost
exactly 1.618.
Phi pops up in other places, too. It is, for instance, the
ratio of the radius of a circle to the side of an inscribed
decagon (a 10-sided figure drawn so that the sides meet on
the circle).
c E
Fig. 39-3
Also, every segment in a pentagon is in Golden Ratio to

the next smaller segment. (See Fig. 39-3.) One example is
x/y; another is AC/AB. If ABC is included in the larger
triangle in Fig. 39-4 (page 192) with a base of 1, then
AC and AG are equal to (1 + V5)/2, phi. Also, BG is
equal to (V5 - 1)/2, or 111)'>. The line AG is then divided
into the Golden Section.
191
A
C~-----~G
Fig. 39-4
A . -::::;;",_ _ _ _--L._ _ _--L.......I C

F
Fig. 39-5
192 EUREKA!
The lengths of the triangle shown in Fig. 39-5 are 1,2,
and yI5 (already you see what's coming?), and DE is 1.
If AF is drawn equal to AE, the following results:
AF AE AC
FC = FC = AF = 1.618 ...
In the famous Fibonacci series, in which each term is

the sum of the previous two (1, 1,2,3,5,8, 13, ... ), the
ratio between consecutive terms approaches cpo For ex-
ample 8/5 = 1.6; 21113 = 1.16154 ... ; 89/55 = 1.6182;
233/144 = 1.6181 ....
1. One last note: Can you show that x = cp if x(rx) = XX?
QUICKIE
2. An amateur mathematician had a circular pool,

precisely 100 feet across, and two pool ladders. He decided
to place them so that the shorter distance between them,
measured along the edge, had the same ratio to the longer
as the longer had to the entire distance. He had just placed
the first ladder when Sir Cumference, an elderly geometry
professor, arrived with a third ladder.
"You mentioned to me what it was that you were
doing," he explained, "and I realized that two ladders
weren't enough for a pool this size. So, can you arrange
the three so that the first distance is to the second as the
second is to the first and second combined, and the sec-
ond is to the third as the third is to both the second and
third combined?"
"But where will the third ladder go?" the owner wailed.
"Well, I can tell you how far it is from the first," Sir
Cumference said, "measured across the pool, but not from
the second."
193
What is the distance from the first to the third ladders?
Answers
1. If x(y1 x) = XX, then:

X(Xllx)=xx
Xl + I Ix = XX
1 + (l/x) = x
a= x2 - X - 1 = (1 + V5 )/2 = if>
2. The distance is 100 feet (See Fig. 39-6.) The problem
is to set alb = bl(a + b) and blc = cl(c + b). As the same
ratio pertains to both sets, alb = b/c.
2 c
1 Fig. 39·6
So, alb = blc and bl(a + b) = blc.

From this, c = a + b, so c occupies half of the circum-
ference of the pool, and therefore the third ladder is
immediately across from the first.
Actually, Sir Cumference didn't need to do this; he
made use of the fact that in any series with the ratio of if>
between the terms, each term is the sum of the previous
two. So,c =a + b.
194 EUREKA!
40. A Bundled-Up Buyer _ _ _ _ __
Greg Arious was accustomed to buying cylindrical bundles

of asparagus 12 inches in circumference. One day his grocer
instead packed two bundles of the same length, but only
6 inches in circumference, and insisted on charging the full
price. Greg, on the other hand, was sure that the two
bundles together contained less than one 12-inch bundle.
Who was right?
Answer
Greg was right.

The circumference of a circle (the bundles were cylindri-
cal, so a cross-section yields a circle) is equal to 27Tr, so the
radii of the bundles are 6, 3, and 3 inches. The volume
of the 12-inch bundle is then 7Tr2 = 7T (6)2 = 367T, times
the length of the cylinder; the volume of the two 6-inch
bundles is then 97T + 97T = 187T, times the length of the
bundle. As the first figure is twice the second, Greg was
right.
195
41. A Piece of Pi _ _ _ _ _ _ _ __
Few of us think of pi, or 1T, as representing anything other

than the ratio of the circumference of any circle to its
diameter. Actually, 1T is an integral part of mathematics
and irrationally turns up in many areas.
The ancient Egyptians, the first real mathematical
culture, not only discovered the nature of 1T (the letter 1T
is actually Greek), but also approximated a value for the
number: (4/3)4, or 256/81, or 3.160+, really quite accurate.
The fact that 1T, a decimal number, is irrational and does
not repeat itself or end (as opposed to such a fraction as
9/11, which repeats and so is rational) was shown by the
German mathematician Lambert in 1761. Pi was also
proved by Lindemann to be transcendental. This means
that it satisfies no equation with integer coefficients, such
as 21T2 - 31T - 2 = O. Other transcendental numbers are e,
expressions involving natural logarithms (which are based
on e), expressions involving trigonometric functions (which
use 1T), and any number raised to an irrational power.
Archimedes, the famous Greek philosopher and mathe-
matician, went even further. He drew a circle with a diam-
eter of 1, reasoning that a circle is actually a polygon
with an infinite number of sides. This, joined with the
already-known fact that the circumference of a circle
is equal to 1T mulitplied by the diameter, gives the circum-
ference in this case to be equal to 1T. This means that, as
polygons of greater and greater numbers of sides are drawn
in the circle ("inscribed in the circle"), the perimeters of
the polygons approach the circumference of the circle,
196 EUREKA!
and therefore get closer and closer to 71'. For instance, an
equilateral triangle (Fig. 41-1) would have a perimeter
of 3.67+, off by a little more than a half, and a hexagon
(Fig. 41-2) would differ from 'IT by only 0.14+.
Fig. 41·1
Fig. 41-2
197
Archimedes eventually drew a polygon of 96 sides (try
it yourself!) and found that 1f is between 3 117 and 3 1017l.
The average of these two values is 3.141585 ... ,differing
from 1f only in the ten-thousandth place, quite a remarkable
achievement.
A circle with a radius of 1 has an area of 1f, and the
square surrounding it ("circumscribed about it") has an
area of 4. (See Fig. 41-3.) So, there is a probability of 1f/4
that a stone or coin dropped at random onto the square
will also land in the circle. Experimentation can produce
a very accurate value of 1f.
Fig. 41·3
In the same vein, if a set of parallel lines is drawn on a

flat surface, perhaps 1 inch apart, and a I-inch needle is
dropped onto the lines, the chance that the needle will fall
across one of the lines is 21rr times the chance that it will
not. This observation, made by Count Buffon, was tested
in 1901 by the Italian mathematician Lazzarini, who
198 EUREKA!
obtained 3.1415929 ... after 3400 tosses. You might like
to try one or more of these trials (yes, they are all trials!)
and compare your value to 1f.
Even before these tests were made, several actual for-
mulas for 1f had been found. The first was discovered by
Viete of France:
2/rr =Vl72 X V 112 + 112y172

X v'1I2 + 1I2V1I2 + 112y172 ...
James Gregory of Scotland found this one:
4/rr = 1
---
1 + 12
---
2 + 32
---
2 + 52
---
2 + 72
---
2 + ...
Pi is also equal to 4 - (4/3) + (4/5) - (4/7) + (4/9) ... ,
and in slightly different form:
2X2X4X4X6X6 ...
1f/2 = ----------
3X3X5X5X7 ...
This is Leibniz's Formula.
Another formula for 1f is 1f2/6 = 1 + (112)2 + (113)2 +
(114)2 + (115)2 ....
These formulas readily adapt to pocket calculators or
computers. Before the electronic age, however, mathe-
maticians did such calculations laboriously by hand, vainly
searching for signs of repetition. In German texts, 1f is
frequently called Ludolf's number, in honor of a 15th
century mathematician who determined 1f to 35 places.
199
William Shanks spent 15 years of his life calculating 71'
to 707 places; his record stood until 1949, when a com-
puter called ENIAC, after 70 hours of computation,
spewed out some 2000 digits of 71'. (More recently, how-
ever, a computer calculated 71' to 1,000,000 places. The
published result has been described as the world's most
boring 400-page book.) Alas, an error was found in Shanks's
value, and the more than 100 digits following were all
wrong!
Memorizing 71' seems to be another favorite pastime
among mathematicians. One Britisher memorized the num-
ber to some 5050 places (2217 repeats after 6 digits,
3.142857, and can thus be repeated until your tongue
falls out; it is, at best, only an approximation of 71', dif-
fering after the second place).
Various mnemonics, phrases or bits of prose in which
the number of letters in successive words gives successive
digits, have been devised over the years. "How I wish I
could recollect pi easily today" gives 71' to 8 decimal places,
and "May I have a large container of coffee" to 7. "Now I
live a drear existence in ragged suits and cruel taxation
suffering" gives a wonderful 12.
This poem won a mnemonic contest set up by an
English banking magazine, immortalizing 30 digits in prose:
Now I will a rhyme construct, 3.14159
By chosen words the young instruct, 265358
Cunningly devised endeavour, 979
Con it and remember ever 32384
Widths in circle here you see, 626433
Sketched out in strange obscurity. 83279
200 EUREKA!
A few more curious coincidences:
355/113 = 3.1415929 ...
v'5T - 4 = 3.1414+, 7r approximately
9/5 +V975= 3.1416+, again near 7r
A great number of people do not enjoy the many intri-
cacies and curiosities of 7r. For instance, the General
Assembly of Indiana in 1897 enacted a bill to declare 7r
de jure (according to law) equal to 4, for the most inac-
curate version ever. In the same year, the Indiana State
Legislature came within a single vote of declaring 7r to be
3.2. Actually, understanding pi is a piece of cake.
QUICKIE
What is the radius of a circle whose circumference is

equal to its area in terms of 7r?
Answer
If the circumference of a circle is equal to its area, then

27rr = 7rr2. By factoring out a trr from each side of the
equation, the expression r = 2 is obtained. The radius of
such a circle must be 2.
201
Selected Bibliography and Suggested
Further Reading _ _ _ _ _ _ _ _ _ _ __
Andrews, W. S. Magic Squares and Cubes. New York:

Dover, 1960.
Barr, Stephen. Second Miscellany of Puzzles. London:
Collier-MacMillan, 1969.
Cutler, Ann, and McShane, Rudolph. Trachtenberg Speed
System of Basic Mathematics. New York: Doubleday,
1967.
Dudeney, Henry. Canterbury Puzzles. New York: Dover,
1958.
Fisher, John. Never Give a Sucker an Even Break. New
York: Pantheon Books, 1976.
Gardner, Martin. The Incredible Dr. Matrix. New York:
Charles Scribner's Sons, 1976.
- - _ . The Unexpected Hanging and Other Mathemat-
ical Diversions. New York: Simon and Schuster, 1969.
Hofstadter, Douglas R. "Metamagical Themas." Scientific
American, March 1981.
Lindgren, Harry. Recreational Problems in Geometric
Dissection and How to Solve Them. New York: Dover,
1972.
Lucas, Jerry. Championship Card Tricks. New York:
Grosset and Dunlap, 1973.
Manning, Henry P., ed. The Fourth Dimension Simply
Explained. New York: Dover, 1960.
Maxwell, E. A. Fallacies in Mathematics. London: Cam-
bridge University Press, 1959.
Scarne, John. Scarne on Teeko. New York: Crown Pub-
lishers, 1955.
202 EUREKA!
Schlossberg, Edwin, and Brockman, John. The Pocket
Calculator Game Book. New York: William Morrow,
1975.
Smullyan, Raymond. What Is The Name of This Book:
The Riddle of Dracula and Other Logical Puzzles.
Englewood Cliffs, New Jersey: Prentice-Hall, 1978.
203

Tips Eureka Math Fun From Many Angles 1

Uploaded by

Copyright:

Available Formats

Tips Eureka Math Fun From Many Angles 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tips Eureka Math Fun From Many Angles 1

Uploaded by

Copyright:

Available Formats

21 21 21 21 21' .

262626 2525 r .. -' ''' --

The name EUREKA is taken from a charming bit of bath-

My thanks go to my sister Emily, who contributed the

1. There are no mistakes in this book. The above should

Is the correction a mistake or not?

1l~ Fun and Games

3)~ Fallacies and Logic

9. If This Is Not a Chapter, My Name Is

12. Archimedes Anderson and the Case of the

@g Photons Are Light Matter, Too

19. Archimedes Anderson and the Gambling

24. A Speedier System of Solving 129

32. Prime Time 151

f8)~ Cranium Crackers and Cheese:

35. Classy Problems 165

37. Expand Your Mind 181

A limerick writer once noted:

1. If Fig. 1-1 is cut along the middle, what will the

If you were to join two Mobius strips that are mirror

Normally, the wall of a three-dimensional object must

Much to their surprise, though, they discovered that

2. Are Figs. 1-4 and 1-5 one-strokers? Don't actually

Fig. 1·4 Fig. 1·5

3. Of a slightly different nature is the topological

4. A well-known brewery uses as its trademark three

6. This last one I find absolutely perfect at parties and

1. The result will be two unlinked loops like the first,

2. The first has two odd vertices, so it can be drawn by

3. A line that enters and leaves a space must cross two

4. The arrangement is shown in Fig. 1-12. The pattern can

6. It looks hard, but as illustrated in Fig. 1-13, release

Many numbers and operations of mathematics have special

1. How about this one? Think of a number and mUltiply

2. One clever twist on the "think of a number" trick

3. If you are older than 9 and younger than 100, write

One other favorite is asking a friend to select a digit

6. Here's a harder trick using the same principle. Ask

7. Ask the nearest mathematician about this one: Tell

8. Some stunts just take a little thought to unravel. If

9. Try this one on the nearest unwitting stooge. Ask

Un losable bets like these abound. Have someone place a

Take a full deck of cards and ask a nimble card shark to

DID YOU KNOW THAT ...

11. A real mathematician named Walter Penney came

Fig. 2·4 Fig. 2·5

12. Lastly, here are some puzzles to confuse and con-

1. If your numbers are N and X:

3. A number such as 141,414 is divisible by 10,101. As

4. Merely note the fourth number from the bottom,

6. Just subtract the top face from 21, so the answer is

the original number.

9. I certainly hope not. The buyer will have purchased

10. All you need to do to acquire untold sums of

11. The odds to the coin-tossing are as follows:

The field of mathematical games is an especially rewarding

1. Take, for instance, tic-tac-toe, a game in which two

DID YOU KNOW THAT ...

2. Another exciting variation of tic-tac-toe is played