SI.

00
GEOFFREY MOTT-SMITH
M A T H E M A T I C A L
P U Z Z L E S
FOR BEGINNERS&ENTHUSIASTS
for recreation and mental stimulus
this book will prove a diverting
pastime
/
as well as
a challenge to your ability
to concentrate and apply logic
to your reasoning
M A T H E M A T I C A L P U Z Z L E S
for Beginners and E nthusiasts
By Geoffrey M ott-S mith
SECONDREVISEDEDITION
DOVE R P U BL IC A T IONS, INC .
COPYRIGHT 1946BY THE BLAKISTONCOMPANY
COPYRIGHT 1954BY DOVERPUBLICATIONS, INC.
This new Dover edition is an unabridged republication
of the first edition with revisions and corrections.
Library of Congress Catalog CardNumber: 55-3389
M anufactured in the U nited States of A merica
Dover P ublications, Inc.
180 Varick Street
New York 14, N. Y.
P RE FA C E
If you have had fun in solving puzzles, if you like to enter-
tain your friends with puzzles and mental gymnastics, this is your
kind of book. H ere you will find easy puzzles, hard puzzles, puzzles
useful and amusing, puzzles for beginners and puzzles for old-
timers, puzzles to challenge your logic, your ingenuity, your
knowledge. Some of these puzzles are old favorites; many of them
are new ones invented by the author.
T he primary object of the book is to entertain. T o solve many
of the puzzles, you need no knowledge of mathematics other than
simple arithmetic. Other puzzles require a knowledge of ele-
mentary algebra and plane geometry. Some of the puzzles are
solved for the reader in the text, in order to show how to attack
more complex puzzles of the same type. A nswers to all puzzles
are given on pages 139-235, and here the full method of solu-
tion is explained for all the more difficult puzzles.
T he chapters of the book are arranged in the order in which
they should be read by anyone whose schooling in mathematics is
not fresh in his mind. T he first chapter contains easy arithmetic
puzzles, most of which can and should be solved without recourse
to pencil and paper. T he second chapter takes up puzzles based
on logic, to sharpen the reader's ingenuity. T he third and fourth
chapters present a variety of types of puzzles, ranging from easy
to difficult, which can be solved by simple algebra. Geometry is
introduced by dissection puzzles, some of which are solved by
theoretic considerations and some by simple trial and error. Other
aspects of geometry are touched upon in the sixth chapter. T he
seventh and eighth chapters dig into the properties of digits and
integers; they contain the hardest puzzles in the book. Related
puzzles of decimation are given a separate chapter following. T he
tenth and eleventh chapters concern permutations and combina-
tions and probability, a rather specialized field, so that funda-
mental formulas are given for the guidance of the beginner. T he
last two chapters analyze some number and board games of a
mathematical character and others with elements of mathematical
V
vi PREFACE
exactitude. While games seem far removed from the formal study
of mathematics, they afford an excellent opportunity for the exer-
cise of ingenuity in analysis.
Within each chapter the puzzles are best attacked in the
given order, since in some cases a puzzle depends for its solution
upon some previous puzzle in the chapter.
In the A ppendix are given tables of primes, squares, and so
on, together with explanations of how to extract square and cube
root. Besides being generally useful to the puzzle addict, these
tables are needed for the solution of a few problems in this book.
If the reader wishes to delve further into the theory of mathe-
matical puzzles, he should consult the works of Sam L oyd, H . E .
Dudeney, and W. W. Rouse Ball.
Sam L oyd (1847-1910) was a genius in the invention of
puzzles of all sorts. Besides being one of the great pioneers in the
composition of chess problems, he invented many of the forms
in which puzzles are now cast. H is works were printed mostly in
periodicals, but several compendiums of his puzzles have been
published.
H . E . Dudeney (1857-1931) was an E nglish mathematician
who interested himself in puzzles, and published several collections
of his own inventions. H e was the first to solve a number of classi-
cal problems. T he reader of his works must be prepared to find
very easy and very difficult puzzles intermixed without warning.
A nother E nglish mathematician of the same period, W. W.
Rouse Ball, published in 1892 his Mathematical Recreations, one
of the definitive works on classical problems and the theory of
their solution.
T he books of these pioneers are out of print, but second-hand
copies are fairly easy to obtain, and the books are of course avail-
able in many libraries.
T hanks are due to A lbert H . M orehead, Rubin A tkin, L ewi
T onks, and L . F. L afleur for valuable suggestions incorporated in
the text, and to my son John for his able assistance in the prepara-
tion of the manuscript.
G. M -S.
C ONT E NT S
P RE FA C E v
P art One
M A T H E M A T I C A L P U Z Z L E S
I. E A SY A RI T H M E T I C A L P U Z Z L E S . . . . 3
II. P U Z Z L E S OF INFE RE NC E A ND INT E RP RE -
T A T I ON 11
III. A L GE BRA IC P U Z Z L E SGROU P ONE . . . 20
IV. A L GE BRA IC P U Z Z L E SGROU P T WO . . 30
V. DISSE C T ION OF P L A NE FIGU RE S . . . . 40
VI. GE OM E T RI C A L P U Z Z L E S 54
VII. P ROP E RT IE S OF DIGIT S 67
VIII. P U Z Z L E S A BOU T INT E GE RS 79
IX. DE C I M A T I ON P U Z Z L E S 94
X. P E RM U T A T I ONS A ND C OM BI NA T I ONS . 98
XI. P ROBL E M S OF P ROBA BIL IT Y 109
XII. NU M BE R GA M E S 115
XIII. BOA RD GA M E S 120
P art T wo
S OL U T I ONS
SOL U T IONS A ND A NSWE RS T O T H E
P U Z Z L E S IN P A RT ONE 139
A P P E NDI X
HOWTOEXTRACT SQUARE ROOT 236
HOWTOEXTRACT CUBE ROOT 238
TABLE OF SQUARE ROOTS 239
vii
viii CONTENTS
TABLE OF POWERSOF 2 239
TABLE OF SQUARE NUMBERS 24O
TABLE OF PRIME NUMBERS 24I
TABLE OF TRIANGULAR NUMBERS 242
GL OS S A RY 243
P art One
M A T H E M A T I C A L P U Z Z L E S
I . E asy A rithmetical P uzzles
1. H OW H I GH !S A P OL E ? H ow high is a pole that casts
a shadow 21 feet long, if a 6-foot man standing beside the pole
casts a shadow 4 J/2 feet long?
H ow deep is a well, if a rope that just reaches from bottom
to top can be wrapped exactly 12 times around the cylindrical
drum of a windlass, the drum being 7 inches in diameter?
H ow many sheep jump over a fence in an hour if 10 sheep
jump over a fence in 10 minutes?
2. DOM I NO S E T S . In a domino set that runs up to double-
six, there are 28 bones (pieces). In a set that runs up to double-
nine, there are 55 bones.
H ow many bones are there in a domino set that runs up to
double-twelve?
3. M A RK-DOWN. A clothing dealer trying to dispose of an
overcoat cut in last year's style marked it down from its original
price of $30 to $24. Failing to make a sale he reduced the price
still further to $19.20. A gain he found no takers, so he tried an-
other price reduction and this time sold it. What was the selling
price, if the last mark-down was consistent with the others?
3
4 EASY ARITHMETICAL PUZZLES
4. NI NE DOT S . H ere is an old puzzle and an easy one;
nevertheless, it proves baffling to many a hasty reader. T he dia-
gram shows 9 dots in the form of a square. Draw 4 straight lines so
as to cross out every dot. You must not cross any dot more than
once, nor retrace any line, nor lift the pencil from the paper until
all 9 dots have been crossed.



5. M A KI NG A C H A I N. I have six sections of chain, each
consisting of four links. If the cost of cutting open a link is 10
cents, and of welding it together again, 25 cents, how much will
it cost me to have the six pieces joined into one chain?
6. T H E WI L Y C H I E F. T he following account of conditions
on a remote South Sea isle comes from a usually unreliable source.
It seems that the M 'gmb race inhabiting this isle is ruled over by a
wily chief who has a passion for erecting monuments to himself.
T o do this work he hires men at 5 bmgs per day. But the race is
not noted for industriousness, and the chief fines each man 7 bmgs
for each working day when he loafs or is absent. Knowing his
fellow M 'gmbs well, the chief has chosen the rates so that each
M 'gmb just breaks even in every month of 24 working days. T hus
the chief never has to pay out a single bmg. T he question arises,
how many days does a M 'gmb work per month?
5EASY ARITHMETICAL PUZZLES
7. T H E BOOKWORM . T he two volumes of Gibbons' "De-
cline and Fall of the Roman E mpire" stand side by side in order
on a bookshelf. A bookworm commences at P age i of Volume I
and bores his way in a straight line to the last page of Volume II.
If each cover is ]/8 of an inch thick, and each book without the
covers is 2 inches thick, how far does the bookworm travel ?
8. A N E A S Y M A GI C S QU A RE . A rrange the digits, from 1
to 9, in a square, so that every row, column, and diagonal totals
the same amount.
9. T H E FA C E T I OU S YOU NG M A N. "Give me a pack of
Fumeroles, please," said the customer to the young man in the
cigar store. "A nd how much are those Sure-Fire lighters?"
"One Sure-Fire lighter buys three packs of Fumeroles," was
the reply.
"Well, give me a lighter. H ow much is that?"
"T he total of the digits of what you owe me is 14," said the
facetious young man.
6 EASY ARITHMETICAL PUZZLES
T he customer didn't attempt to puzzle that out, but merely
gave the clerk a dollar bill and accepted his change.
What is the cost of a Sure-Fire lighter?
10. T A NKT OWN T RI OS . Whenever they travel by train, the
members of the T anktown baseball club play pinochle. T he nine
regulars form three tables of three each. But no outfielder likes to
play at the same table as another outfielder, basemen will not sit
together, while the pitcher, catcher, and shortstop aver that they
see enough of each other on the diamond. Despite these limita-
tions, the nine have been able to organize the three tables in a
different arrangement on every trip they have taken. H ow many
different arrangements are possible?
11. WA T E R. GA S , A ND E L E C T RI C I T Y. T he illustration
shows three utility plants, furnishing respectively water, gas, and
electricity; together with three houses that are to be serviced. A
conduit must be laid from each plant to each house, but it is de-
sired that no two conduits should cross. H ow can this be done?
d 0J

n n
12. A BRI C KE Y QU I C KI E . A nyone who pulls out pencil and
paper for this one is disqualified and must go stand in the corner.
If a brick balances evenly with three-quarters of a pound and
three-quarters of a brick, what is the weight of a whole brick?
Quick!
7EASY ARITHMETICAL PUZZLES
13. S P OT T I NG T H E C OU NT E RFE I T . "Where is that coun-
terfeit dollar?" the chief of the Secret Service office asked his
aide.
"I left it on your desk, along with the eight others that turned
out to be genuine."
T he chief found the nine "cartwheels" heaped together, with
nothing to show which was the spurious coin. H e knew that the
latter was underweight, so he improvised a balance by setting up a
ruler across the lip of an inkwell. H e found that by placing coins
at equal distances from this fulcrum he could weigh one coin
against another with sufficient accuracy to determine whether
both were sound dollars.
H e then proceeded to spot the counterfeit by just two weigh-
ings. T his was not a lucky chance; his method assured that no
more than two weighings would be necessary. What was the
method?
14. T H E P A I NT E D C U BE . A wooden cube is painted black
on all faces. It is then cut into 27 equal smaller cubes. H ow many
of the smaller cubes are found to be painted on three faces, two
faces, one face, and no face?
15. S H E E P A ND GOA T S . T he illustration shows three sheep
(white checkers) and three goats (black checkers), distributed
alternately in a line of pens (one row of a checkerboard).
I 2 3 4 5 6 7 8
Now we wish to sort out the sheep from the goats, placing
the sheep in the pens numbered 1, 2, 3 and the goats in 4, 5, 6. But
these are gregarious animals who balk at being moved singly; to
move them at all we have to shift a pair of them from adjacent
pens to the empty pens. We cannot reverse the order of the pair
8 EASY ARITHMETICAL PUZZLES
in transit. For example, we can move 5 into 7 and 6 into 8, but we
cannot put 5 into 8 while 6 goes into 7.
A fter you have found how to sort the sheep from the goats,
try to do it in as few moves as possible. It can be done in four
moves.
T hen, starting from the arrangement shown in the illustra-
tion, rearrange the animals so as to get the goats (black) into 1,
2, 3 and the sheep (white) into 4, 5, 6. T his puzzle can be solved
in four moves.
16. T H E BI L L I A RD H A NDI C A P . "Do you play billiards?
C are to have a game?" asked H untingdon of the new member at
the T own C lub.
"Yes, I play," replied M cC lintock, "but I'm rather a duffer.
M y friend C hadwick gives me 25 points in 100, and then we play
about even."
"Well, I'm perfectly willing to give you a proper handicap. I
give C hadwick 20 points in 100. Now let's see-how many points
should I give you?"
What is the correct answer, assuming that the stated handi-
caps are fair?
17. T H E S U RROGA T E 'S DI L E M M A . "I have come to con-
sult you," said the surrogate to the mathematician, "about Wil-
liam Weston's will. William Weston was fatally injured in a traffic
accident while he was on his way to the hospital where his expec-
tant wife was confined. H e lived long enough to make a will,
which provides that if his child is a boy the estate is to be divided
in the proportion of two-thirds to the boy and one-third to the
widow. But if the child is a girl, she is to receive only one-fourth
and the widow receives the remaining three-quarters.
"Now M rs. Weston has given birth to twins, a boy and a girl.
T here is some question whether the will can be held to apply.
What would be the correct division of the estate to carry out
Weston's evident intentions?"
What was the mathematician's reply?
9EASY ARITHMETICAL PUZZLES
18. T H E L I C E NS E P L A T E . Jim C arter was sorry to have to
discard last year's license plate from his car, for the numbers made
a beautiful poker handa full house. H e was disgusted to note
that on his new plate all five figures were different. T o top that,
he inadvertently screwed the plate on his car upside-down, with
the result that he increased his registration number by 78,633 until
he noticed the error.
What was the number on his license plate?
19. M E A S U RI NG T WO GA L L ONS . "What else can I sell
you today?" asked E lmer Johnson, the proprietor of C entreville's
general store.
"Well," replied Si C orning, "you'd better give me a couple of
gallons of gas. M y thrashing machine is a mite low."
"T ake five gallons while you can get it, Si. P rice is going up,
they tell me."
"No, I ain't going to lug five gallons all the way home. Be-
sides, I don't think the tank will take it. M ake it two gallons."
"Fact is, Si, I don't have no two-gallon measure. I got an
eight-gallon measure, and plenty of five-gallon cans, but I don't
see how I can give you just two gallons for certain."
T he upshot of the conversation was that Si decided to post-
pone his purchase of gasoline until he could use five gallons.
But E lmer could have measured out exactly two gallons,
using only the 8-gaIlon and 5-gallon measures. H ow?
20. M A T C H S T I C K E QU A T I ONS . If the after-dinner enter
tainer were compelled to rely on one article of paraphernalia
alone, he could scarcely make a better choice than a box of
wooden matches.
T he matchsticks lend themselves to the demonstration of feats
of equilibrium, of arithmetical and algebraic puzzles, of geometri-
cal puzzles and catches, and to the playing of mathematical
games.
10 EASY ARITHMETICAL PUZZLES
One of the possibilities in this field, little exploited, is the
matchstick equation. T he illustration gives five examples. E ach
row is a separate puzzle, an equation given in Roman numerals.
In the third puzzle, the square is intended to represent zero, and
must be read as such. A ll the equations are false as they stand, but
each can be changed into a true equation by altering the position
of only one match.
I I . P uzzles of I nference and I nter
pretation
21. WH A T I S T H E NA M E OF T H E E NGI NE E R? A n oft-
quoted problem of the "C aliban" type concerns three pairs of men
who shared the names Smith, Robinson, and Jones. T he presen-
tation of this puzzle seems to be jinxed; I have heard it misstated
numerous times, and in several publications the facts given are
either insufficient or contradictory.
H ere are the facts as set forth in what may be the original
source, the works of H . E . Dudeney.
T hree businessmenSmith, Robinson, and Jonesall live in
the "capital district" of New York. (I have changed the locale to
the U .S.A . to escape the E nglish currency.) T hree railwaymen
also named Smith, Robinson, and Joneslive in the same district.
T he businessman Robinson and the brakeman live in A lbany, the
businessman Jones and the fireman live in Schenectady, while the
businessman Smith and the engineer live halfway between these
two cities. T he brakeman's namesake earns $3500 per annum, and
the engineer earns exactly one-third of the businessman living
nearest him. Finally, the railwayman Smith beats the fireman at
billiards.
What is the name of the engineer?
22. A T T H E RA I NBOW C L U B. Four members of the Rain-
bow C lub sat down one afternoon to play bridge.
In accordance with the rules of the game, they drew cards
11
12 INFERENCE ANDINTERPRETATION
from a deck spread face down. T he man who drew the highest
card chose his seat and the deck to be dealt by his side; sccond
highest took the opposite seat, as his partner; third highest took
his choice of the remaining two seats, lowest card becoming his
partner.
Without troubling to put the facts in chronological order, we
may note that White's card was lower than Brown's. Green asked
for a match, which was supplied by White's partner. Black said
"What is your choice, partner?" Brown sat on White's left. T he
left-handed man chose the blue cards, and since Brown is right-
handed you can now tell the order o( the lour players accord-
ing to the cards they drew.
23. T E NNI S A T H I L L C RE S T . E ight men entered the recent
tennis tournament at H illcrest. T he tournament was played in
three consecutive days, one round per day, and happily no match
was defaulted. T he first and second round matches were stipu-
lated to be 2 sets out of 3, while the final was 3 sets out of 5. A
spectator who was present on all three days reports the following
facts:
1. E ggleston never met H averford.
2. Before play began, Gormley remarked jocularly to Ban-
croft, "I see that we meet in the finals."
3. C hadwick won a set at love but lost his first match.
4. A ltogether 140 games were played, of which the losers
won 43.
5. When the pairings were posted, A bercrombie said to
Dcvereaux, "Do you concede, or do you want to play it out?"
6. On the second day, the first-round losers played bridge,
and the same table gathered on the third day with E ggleston in
place of A bercrombie.
7. Bancroft won 9 games.
8. Franklin won 37 games.
9. T he first score of the tournament was a scrvice ace by
Gormley, at which E ggleston shouted "H ey, I'm not over there!"
Who won the tournament? Whom did he beat and by what
score?
13INFERENCE AND INTERPRETATION
24. WH I T E H A T S A ND BL A C K H A T S . T hree candidates
for membership in the Baker Street Irregulars were given the fol-
lowing test of logic. T hey were told that each would be blind-
folded and a hat would be put on his head. T he hat might be
either black or white. T hen the blindfolds would be removed, so
that each might see the colors of the hats worn by the other two.
E ach man who saw a black hat was to raise a hand. T he first to
infer correctly the color of his own hat would be admitted to
membership.
T he test was duly carried out. Black hats were put on all
three men. T he blindfolds were removed, and of course all three
raised a hand. P resently one man said "M y hat must be black."
H e was taken into the organization when he proved his assertion
to the satisfaction of the judges.
H ow did he do it?
25. T RU T H A ND FA L S E H OOD. In a faraway land there
dwelt two races. T he A nanias were inveterate liars, while the
Diogenes were unfailingly veracious. Once upon a time a stranger
visited the land, and on meeting a party of three inhabitants in-
quired to what race they belonged. T he first murmured something
that the stranger did not catch. T he second remarked, "H e said
he was an A nania." T he third said to the second, "You're a liar!"
Now the question is, of what race was this third man?
26. WI NE A ND WA T E R. Suppose that we have a bucket
containing a gallon of water and a demijohn containing a gallon
of wine. We measure out a pint of the wine, pour it into the water,
and mix thoroughly. T hen we measure out a pint of the mixture
from the bucket and pour it into the demijohn.
A t the end of these strange proceedings, is there more or less
water in the demijohn than there is wine in the bucket?
27. FOU R P E NNI E S . A rrange 4 pennies so that there are
two straight lines with 3 pennies on each line.
14 INFERENCE ANDINTERPRETATION
28. S E VE N P E NNI E S . M ake an enlarged copy of the eight-
pointed star shown in the diagram. P lace a penny on any point of
the star and slide it along a line to another point. P lace a second
penny on any vacant point and similarly slide it along a line to
reach another open point. C ontinue in the same manner until 7
pennies have been placed on 7 points, leaving only one vacant.
T he task sounds easy and is easy, but on first attempt the
solver usually finds himself blocked after 5 or 6 pennies, unable to
place more under the conditions.
29. T H E ROS E T T E . Now put all the pennies back in your
pocket and answer this question without resort to trial.
If we make a rosette of pennies, by putting as many pennies
as we can around one penny in the center, so that all the outer
coins touch their neighbors and also the center, how many pen-
nies will there be in the rosette?
30. T H E M I S S I NG P E NNY. T his paradox is old, but it is
still good. T wo market women were selling apples, one at 3
15INFERENCE ANDINTERPRETATION
apples for a penny and the other at 2 apples for a penny. (T he
prices give you some idea of the age of the puzzle!) One day
when both were called away they left their stock in charge of a
friend. T o simplify her reckoning the friend amalgamated the
stocksthere were 30 apples ot each qualityand sold them all
at 5 for twopence. T hus she took in 2 shillings (24 pence).
When it came to dividing the proceeds between the owners,
trouble arose. T he one who had turned over thirty apples of 3-for-
a-penny quality demanded her due 10 pence. T he other not un-
reasonably asked for 15 pence. T he sum actually realized was a
penny short. Where did it go?
31. T H E RU BBE R C H E C K. A radio dealer was approached
by a customer who wanted to purchase a P andemonium radio,
priced at $69.98. T he dealer accepted a check for $80.00, giving
$10.02 change in cash. Subsequently he endorsed the check to his
landlord in part payment of the rent. T he check turned out to be
worthless and the customer was not to be found. T he dealer had
to make the check good to his landlord, but the latter accepted a
P andemonium radio in part settlement. A s this type of radio cost
the dealer $43.75 at wholesale, what was the amount of his loss?
32. M YS T E RI OU S C OM P U T A T I ON. "Father," said E d-
ward to P rofessor Digit, "I found this piece of paper on the floor
of your study. Do you want to save it?"
16 INFERENCE ANDINTERPRETATION
"L et me see. Yes, thank you, I need those figures for a prob-
lem I am working on."
"Is that supposed to be an addition, and is that a long
division ?" asked E dward, pointing to the two groups of figures.
"Yes."
"Well, I guess the teachers couldn't have been very good
when you were a boy, because your answers are all wrong."
T he professor laughed, and then proceeded to convince E d-
ward that the answers are correct. What did he tell E dward?
33. T H E T E NNI S T OU RNA M E NT . If 78 players enter a
tournament for a singles championship, how many matches have
to be played to determine the winner?
34. T A RT A GL I A 'S RI DDL E . In ancient times, the ne-
ophyte in logic was posed such questions as:
If half of 5 were 3, what would a third of 10 be?
35. S T RA NGE S I L H OU E T T E S . I have here a familiar ob-
ject. If I hold a candle under it, the shadow it casts on the ceil-
ing is circular. If I hold the candle due south of it, the shadow
it casts on the north wall is square. If I hold the candle due east,
the shadow on the west wall is triangular. What is the object?
36. T H E DRA FT S M A N'S P U Z Z L E . Once I propounded
Strange Silhouettes to a draftsman, and he retorted with a similar
puzzle which I think worthwhile passing on.
17INFERENCE AND INTERPRETATION
T he diagram shows plan and elevation of a solid block of
wood. T he broken lines have the conventional meaning that these
lines are invisible from the particular angle of sight, but they must
be visible from some point, since the block is stipulated to be
solid.
I
- 1 -
1
T he puzzle is to supply a side view consistent with the other
two views. T he latter, by the way, are complete; no line, visible
or invisible, is omitted for purposes of trickery.
37. A P ROBL E M I N P ROBA BI L I T I E S . If in your bureau
drawer are io blue socks and 16 grey socks, and you reach into it
in the dark, how many socks must you take out to be sure of get-
ting a pair that match?
38. A L A M I C E D P U Z Z L E . If you made a business trans-
action with the Noelomis in the land of A cirema, what would a
semid net gain amount to?
39. BE A R FA C T S . A bear left its den and went due south
in a straight line for one mile. T hen it made a go-degree turn to
the left and walked another mile in a straight line. T wice more
it made 90-degree turns to the left and walked a mile in a
straight line, thus returning to its den. On reaching this starting
point, the bear was facing due south.
What was the color of the bear?
18 INFERENCE ANDINTERPRETATION
40. T H E FL A G OF E QU A T RI A . T he five states that com-
pose the nation of E quatria were once independent principali-
ties. T hough having much in common in their outlooks, the
people of these states perpetually quarreled with one another,
until wise heads resolved upon a union into one nation. E ach
principality tried to gain ascendancy through its own claims to
superiority, but eventually general agreement was reached on
three principlessection, resection, and dissection. T o symbolize
this credo it was suggested that the triangle be adopted as the
emblem of E quatria, and in fact this figure appears on the Great
Seal. For ornament, two triangles were combined in the six-
pointed star that makes the flag of E quatria so distinctive.
19INFERENCE AND INTERPRETATION
T here are five stars in the flag, one for each province. No re-
arrangement of the stars will ever be necessary, for one of the first
acts of the union was to promulgate the P unroe Doctrine, which
declares that any attempt by an outside power to gain admittance
to the E quatrian union will be resented as contrary to the laws of
nature.
A s can be seen in the accompanying picture, the stars in the
flag are tessellated, each in a different design. A n interesting logi-
cal exercise is to deduce the plan which governs this tessellation.
(No fair asking an E quatrian!)
I I I . A lgebraic P uzzlesGroup One
41. T H E NA T U RE OF A L GE BRA . T o many persons who
have no occasion to use mathematics in later life, the word
"algebra" recalls only the memory of certain tiresome scholastic
drudgery, a kind of arithmetic where letters are used instead of
numbers for the evident purpose of confusion, a frantic pursuit
of a mysterious and elusive being known only as x.
But to anyone whose work involves mathematics beyond ele-
mentary arithmeticand this means to virtually any student of
abstract or applied science, physical or socialalgebra is a won-
derful tool, bordering on the miraculous.
A rithmetic and algebra are more than the names of two ele-
mentary branches of mathematics. T hey also indicate two methods
of approach and two kinds of objectives that pervade all branches
of mathematics, no matter how "advanced." On the one hand,
there is a compilation of facts, especially about operations, such as
addition, multiplication, factoring, differentiation; on the other
hand, there is a constant inquiry into the kinds of classes to which
certain facts apply, the characteristics and number of members of
each class, the discovery of general propositions from which
other facts can be deduced.
SIMULTANEOUS EQUATIONS
M any algebraic puzzles involve the solution of two or more
simultaneous equations.
20
ALGEBRAIC PUZZLESGROUP ONE 21
E xample: T hree boys picked a number of apples and divided
them according to their ages. E dward took three more apples than
Wilbur, while David took twice as many as E dward, which gave
him eleven more than Wilbur. What was the total number of
apples?
If we represent the number of apples taken by each boy by
the initial of his name, then the three clauses above tell us that
E = W+
3
D=2E
D = W+I I
Solving these equations gives E =8, W=5, and D=i 6, so
that the total of apples is 29.
A nother method of attack is to use only one unknown. L et x
represent the number of apples taken by E dward. T hen Wilbur's
share is x3, David's share is 2x, and we are told that
2x(x3)-f-i 1. H ence x=8, and the other shares may then be
computed.
It may sound simpler to use one unknown, where possible,
instead of several. It may sound simpler to solve one equation than
to solve several simultaneous equations. But all that is an erroneous
notion.
t
If the solver can solve equations at all, a dozen simultaneous
equations present no greater difficulty than one. It is not the num-
ber of equations that matters; what counts is the specific com-
plexity of the form of the equations. Now, if a puzzle intrinsically
involves say five equations, then if all the pertinent facts are put
into a single equation instead, this equation is bound to be more
complex than any one of the five. A nd what actually happens
when one equation is formed is that the solver performs mentally
some of the operations necessary to reduce the five.
T o illustrate the point, let us solve L oyd's famous puzzle,
"H ow old is A nn?" T he puzzle is stated in a deliberately confusing
manner:
"T he combined ages of M ary and A nn are 44 years, and
M ary is twice as old as A nn was when M ary was half as old as
ALGEBRAIC PUZZLESGROUP ONE 22
A nn will be when A nn is three times as old as M ary was when
M ary was three times as old as A nn. H ow old is A nn?"
I suggest that the reader first try the single-equation method.
L et x equal A nn's age; adduce from the facts a single equation
in .v. It can be donebut why do it? H ow much easier to set up
some very simple equations, introducing new literal terms ad lib!
T hus:
L et x and y be respectively the ages of A nn and M ary. T hen
*+y=44 ( 0
"M ary is twice as old as A nn was . . ." H ere is a reference to a
past age of A nn. A s the definition of this age is complicated, let
us at the moment represent it by her present age x less an un-
known number of years, a. We are told that
y=2(xa) (2)
". . . as A nn was when M ary was . . ." H ere is a reference to a
past age of M ary, coincident with the time when A nn was xa.
A t this time M ary's age was therefore ya. Now what is said of
this age? ". . . when M ary was half as old as A nn will be . . ."
Represent this future age of A nn by x-\-b. We are told that
ya=x-\-b
(3)
". . . as A nn will be when A nn is three times as old as M ary
was . . ." H ere is a reference to another and different past age of
M ary. L et it be yc. T he clause states that
{yc) (4)
". . . as M ary was when M ary was three times as old as A nn."
A t the age of yc M ary was three times as old as A nn. A t that
time the age of A nn must have been xc. H ence
c
) (5)
We have been so prodigal as to use five unknowns. But we
have five independent equations, sufficient to find all unknown
values, and the equations themselves are all simple in form.
In order to yield unique values for the literal terms, there
must be at least as many independent equations as there are un-
knowns. T he multiple-unknown method of attack gives an easy
check, therefore, on whether the solver has adduced enough facts
ALGEBRAIC PUZZLESGROUP ONE 23
from the statement of the problem to reach a solution. With
many involved puzzles it is easy without such check to overlook
a vital fact or two. Note also that the equations (if minimum in
number) must all be independent. If the simultaneous solution
of two or more equations results in an identity, as 3=3 or xx,
then the equations were not all independent.
42. A QU E S T I ON OF BA RT E R. If the natives of the Wee-
jee Islands rate 2 spears as worth 3 fishhooks and a knife, and
will give 25 coconuts for 3 spears, 2 knives, and a fishhook to-
gether, how many coconuts will they give for each article
separately?
43. S H A RI NG A P P L E S . A gang of boys made a raid on the
P erkins orchard and came back with a quantity of apples, which
were then pooled and divided equally among them. M ichael said
he thought it would be fairer to share by families instead of indi-
viduals. A s there were two Johnson brothers and two Fairbanks
brothers, a redivision by families would have increased each share
by 3 apples. With the argument at its height, along came Fred,
who, being the oldest of the gang, was appealed to as arbiter.
Fred decided that it would be unfair to share by families. Further-
more, he pointed out, he himself would certainly have partici-
pated in the raid, to the great increase of the booty, had he not
been detained by a compulsory engagement with a rug-beater.
But as head of the gang he was entitled to a share. Fred had a
way of winning his arguments, so each boy contributed one apple
to him, making equal shares all around.
H ow many apples did the boys gather?
44. A T RA NS A C T I ON I N RE A L E S T A T E . "Jim," said one
real estate dealer to another, "I hear you made a pretty piece of
change out of that Dingy Street property. T hey tell me you just
sold it for $4,000. I know it cost you only $2,700."
"Your figures are right, but as a matter of fact I lost money,"
replied Jim in disgust. "T here was a big bill for tax arrears hang-
ALGEBRAIC PUZZLESGROUP ONE 24
ing over it. I was prepared to pay that, but then I had to put
twice as much again into repairs to put the property in shape
before I could find a buyer. It was so much trouble that I was
glad to get rid of it at a loss amounting to 20% of the taxes plus
10% of the repairs."
Jim didn't specify what the loss was, but the other was able
to figure it out.
45. S E T T L I NG T H E BI L L . A fter the boxing matches a group
of friends went into a restaurant for a midnight snack. "P ut it
all on one bill," they told the waiter. T he bill amounted to $6.00,
and the men agreed to split it equally. T hen it was discovered
that two of their number had slipped away without settling their
scores, so that each of the remaining men was assessed 25 cents
more. H ow many men were in the party originally?
46. C OWS A ND C H I C KE NS . T he same wight who
counted sheep by counting the legs and dividing by 4 also kept
track of his cows and chickens by counting both the legs and the
heads. If he counted 78 legs and 35 heads, how many cows and
chickens did he own?
47. T H E FA RM E R'S RE T ORT . A friend of mine who is a
farmer took exception to the answer given to Cows and Chickens.
Said he,
"H ow many legs do you suppose there are when a man milks
a cow?"
"Why, six."
"Wrong. T here are nine."
"H ow do you make that out?"
"Because the man sits on a three-legged stool."
Now, what answer did the farmer give to Cows and Chickens
if he counted in at least one three-legged stool?
48. DOL L A RS A ND C E NT S . Sent suddenly on a business
trip, George Blake spent half of the money in his pocket on a
ALGEBRAIC PUZZLESGROUP ONE 25
round-trip railroad ticket and some necessary supplies. T hen he
bought two newspapers at 3 cents each to read on the train. A
taxi from the terminal to his destination took 70 cents. A quarter
of what he then had left went for meals and a taxi back to the
station. On the return trip he bought a 25-cent magazine. H e
arrived with as many dollars as at the outset he had had cents,
and as many cents as he had had dollars. H ow much was this ?
49. T H E J A Y E S T A T E . U nder the will of Jasper Jay, 10%
of his estate went to various charities. Son John received 25%,
and his share was 25% more than was received by daughter Jill
and the baby together. Jill's share was 30 times more than the
baby's. A fter deducting a bequest of $250.00 to Jenkins, the
butler, M rs. Jay as residual legatee received just as much as the
two older children together. What was the amount of the Jasper
Jay estate?
50. A FI S H S T ORY. Nate T hompson remarked that he
had seen a pretty big mackerel and a pretty big pickerel that
morning in the shallows of a cove. P ressed for an estimate of their
size, Nate observed cautiously that the body of the pickcrel was
about twice the length of his tail, and about equal to the length
of his head plus the tail of the mackerel. T he body of the mackerel
was about as long as the whole pickerel minus the head. T hat
got us nowhere and we pressed Nate for more details. H e "kind
of thought" the head of the mackerel was about as long as the tail
of the pickerel, while the head of the pickerel was about a quarter
of the body of the mackerel. "I jedge," he concluded, "that there
was sure three foot of fish there."
It proved to be a pretty big mackerelif Nate'g estimates
were correct.
51. WH O NOE S ? NOT A YE ! "If there is no further dis-
cussion," said the chairman of the meeting of the Wisteria Im-
provement A ssociation, "I will put the question to a vote. A ll
ALGEBRAIC PUZZLESGROUP ONE 26
those in favor of the motion please stand . . . T hank you. P lease
be seated. A ll those opposed please rise . . . T he motion is de-
feated. We will return to a discussion of the original motion,
which is to plant a bed of azaleas on the southwest corner of . . ."
"M r. C hairman!" interrupted a member from the floor. "I
thought that was the motion we voted on!"
"No," said the chairman. "We voted on the motion to amend
the original motion by substituting the word begonias for azaleas."
"In that case I would like to change my vote. I misunder-
stood the question."
From a number of other members came cries of "M e, too!"
T he chairman read the proposed amendment and called for a
new vote. One-third of those who had previously voted nay
changed their votes to aye, while one-quarter of those who had
voted aye changed to nay.
"A s matters now stand," said the chairman, "the vote is a tie.
I should not like to have to cast the deciding ballot in so im-
portant a question. I suggest that the amendment be further dis-
cussed."
A member from the Second Ward stood and was recognized.
"M r. C hairman," he said, "some of us have been talking it over,
and in the interests of reaching a quick decision are willing to go
along with our friends from the T hird Ward. We would like to
change our vote."
"M r. C hairman," came another voice, "we have been talk-
ing it over too and some of us have changed our minds." T he
meeting burst into a hubbub, with cries of "C hange my. vote!"
"L et's go along with the T hird!" "I much prefer columbine any-
how . . ." A fter quiet was restored, the chairman said,
"It is evidently the sense of the meeting that we should re-
open the whole question. I will therefore once more call for a vote
on the amendment to substitute begonias for azaleas."
T his time it was found that one-half of the members who had
originally voted aye and then changed to nay had gone back to
aye. Of those who had changed from nay to aye, one-quarter
went back to nay. In addition, one-half of those who had hitherto
ALGEBRAIC PUZZLESGROUP ONE 27
voted only aye decided to change to nay. But then one-third of
those who had hitherto voted only nay changed to aye.
"I see," said the chairman, "that the amendment is defeated
by a margin of two votes."
H ow many voters were there at the meeting of the Wisteria
Improvement A ssociation?
52. NO FRE E Z E OU T . Five men sat down to a game of
Freezeout H earts. E ach was allowed to buy chips amounting to
just 2 dollars. In this game, the loser of each hand pays each other
player a number of chips determined by the number of hearts
taken in tricks. T he first player to lose all his chips is "frozen out"
and can no longer play in the game. It was agreed that the player
first frozen out should go out for sandwiches and beer while the
others continued a four-handed game.
T he loser of each hand was also required to pay one chip to
the kitty, until it amounted to 2 dollars. T he kitty was set aside
to pay for the refreshments.
T he game went on for several hours, the fortunes favoring
none in particular. L ong after the kitty was complete, no player
had been frozen out.
T he players thereupon agreed to play one more round, under
"murder or sudden death" rules. In this round, the loser of each
hand was to pay each other player a number of chips equal to
the number held by that player. In other words, the loser had to
double the chips outside his own stack.
T he round consisted of five hands, one dealt by each player.
Strange to relate, each of the five players lost on his own deal,
and when the round was over all players held the same number of
chips.
H ow many chips were held by each player just before the
last round?
53. J OH NNY' S I NC OM E T A X. "M r. T hompson, will you
help me figure my income tax?" asked Johnny, the office boy.
"Sure thing," was the reply. "Bring me your papers."
ALGEBRAIC PUZZLESGROUP ONE 28
"Well, here's the form I gotta use, and here's the statement
of how much the company paid me during the year."
"A ny income from other sources? Odd jobs on the side? Did
you take in washing . . ."
"Naw, that's the whole thing."
"Do you claim any deductions? A ny capital losses? A ny con-
tributions to charity?"
"I gave four bucks to the Red C ross."
"A nd you have a receipt for it, don't you? Okay, you can
claim that as an exemption. I don't suppose you are married, are
you? A ny dependents? No? Well, then, your personal exemption
is $500.00. Your tax is 19% of the taxable net income. I'll work
it out for you . . . H ere it is."
"Gee," remarked Johnny. "Isn't that funny! T he tax is just
10% of what the company paid me. Does it always work out that
way?"
"No, indeed," laughed M r. T hompson. "T hat's just a hap-
penstance."
What was the amount of Johnny's tax?
54. S P E NDI NG A QU A RT E R. I purchased some drawing
supplies, spending 25 ccnts for 25 articles. I bought four kinds of
articles: paper at two sheets for a cent, pens at a cent apiece,
pencils at two for a nickel, and erasers at a nickel each. H ow
many of each kind did I take?
55. T H E S P OOL OP T H RE A D. M rs. P lyneedle stepped into
a dry goods store to purchase a spool of thread. She had in her
purse some coins amounting to less than one dollar. She found
she could pay for the spool with six coins. On talking over her
plans with the salesgirl, M rs. P lyneedle decided that she had
better take two spools at the same price, and found that she could
make exact payment with five coins. In the end, she took three
spools, and paid with four coins. T he salesgirl noticed that had
she bought four spools she could have paid with only three coins.
ALGEBRAIC PUZZLESGROUP ONE 29
What was the price of the spool, and what coins did M rs.
P lyneedle have in her purse?
56. A DE A L I N C A NDY. T hree boys received a nickel
each to spend on candy. T he stock offered by the candy store com-
prised lollipops at 3 for a cent, chocolate bonbons at 4 for a cent,
and jujubes at 5 for a cent. E ach boy made a different selection,
but each spent his entire 5 cents and returned with just 20 pieces
of candy. What were their selections?
57. WH A T S I Z E BE T ? T he following incident came to my
notice at a poker game in C hicago. T he game was stud, with no
ante, but with the rule that high hand on the first round of cards
face up must make a bet. For lack of chips, the players used coins
(none gold).
On one occasion, high hand made a bet comprising 2 coins.
E ach of the other players stayed without raising. T he second hand
put 3 coins into the pool. T hird hand put in 2 coins and took out
one in change. Fourth hand put in 3 coins and took out one in
change. Fifth (last) hand put in one coin, then took in change
all but 3 of the coins then in the pool.
H ow much did first hand bet?
58. T H E H OS KI NS FA M I L Y. T he H oskins family is a well-
regulated household. When it turned out en masse to pick blue-
berries last fall, a separate quota was assigned to the men, women,
and children. E ach quota was a whole number of quarts, and
each individual was expected to harvest exactly as many quarts as
every other in his category. T he quotas were such that 2 men
gathered as many quarts as 3 women and 2 children, while 5
women gathered as many as 3 men and one child. A ll quotas were
filled and the total harvest was 116 quarts. H ow many men,
women, and children are there in the H oskins family, if there are
more women than men and more men than children?
I V. A lgebraic P uzzlesGroup T wo
59. T H E T I DE . A motorboat that travels 13^3 miles per
hour in still water makes a straight run with the tide for an hour
and 8 minutes. T he return journey against the tide takes 8 min-
utes longer. What is the average force of the tide?
60. L OC A T I NG T H E L OOT . A brown T erraplane car
whizzed past the State P olice booth, going 80 miles per hour. T he
trooper on duty phoned an alert to other stations on the road,
then set out on his motorcycle in pursuit. H e had gone only a
short distance when the brown T erraplane hurtled past him, go-
ing in the opposite direction. T he car was later caught by a road
block, and its occupants proved to be a gang of thieves who had
just robbed a jewelry store.
Witnesses testified that the thieves had put their plunder in
the car when they fled the scene of the crime. But it was no longer
in the car when it was caught. Reports on the wild ride showed
that the only time the car could have stopped was in doubling
back past the State P olice booth.
T he trooper reported that the point at which the car passed
him on its return was just 2 miles from his booth, and that it
reached him just 7 minutes after it had first passed his booth. On
both occasions it was apparently making its top speed of 80 miles
per hour.
30
ALGEBRAIC PUZZLESGROUP ONE 31
T he investigators assumed that the car had made a stop and
turned around while some members of the gang cached the loot
by the roadside, or perhaps at the office of a "fence." In an effort
to locate the cache, they assumed that the car had maintained a
uniform speed, and allowed 2 minutes as the probable loss of time
in bringing the car to a halt, turning it, and regaining full speed.
On this assumption, what was the farthest point from the
booth that would have to be covered by the search for the loot?
61. S T RI KI NG A N A VE RA GE . A motorist sets out to cover
a distance of 1 o miles. A fter he has covered half this distance, he
finds that he has averaged only 30 miles per hour. H e decides to
speed up. A t what rate must he travel the rest of the trip in order
to average 60 miles per hour for the whole journey?
62. T H E S WI M M I NG P OOL . "H i, Jill, I'll race you to this
end of the pool!" shouted Jack, who was then only a few feet from
the end he indicated.
"You don't want much head start, do you!" retorted Jill,
from the other side of the swimming pool. "I'll race you even up
for twice the length of the pool."
"Okay," called Jack. "You start there and I'll start here."
T hey started simultaneously, Jack from the east end and Jill
from the west end. T hey passed each other the first time 20 feet
from the east end, and the second time 18 feet from the west end.
Who won the race?
T hat question is too easy. L et's ask another. A ssuming that
each swimmer maintained his own speed without variation, and
turned back instantaneously on completing the first leg of the
race, how long was the pool?
63. H A NDI C A P RA C I NG. On one side of the playground
some of the children were holding foot-races, under a supervisor
who handicapped each child according to age and size. In one
race, she placed the big boy at the starting line, the little boy a
few paces in front of the line, and she gave the little girl twice as
ALGEBRAIC PUZZLESGROUP ONE 32
much headstart over the little boy as he had over the big boy.
T he big boy won the race nevertheless. H e overtook the little boy
in 6 seconds, and the little girl 4 seconds later.
A ssuming that all three runners maintained a uniform speed,
how long did it take the little boy to overtake the little girl?
64. T H E P A T ROL . Immediately the news of the First
National Bank robbery was flashed to headquarters, a police car
was sent to the H igh Street bridge. T he robbers had made their
getaway in a car and the bridge was the only exit from the town
on the north side.
-s-
B
J j C i
cm cm cm' o cm cm cm cm
i z j

cm cm cm cm czd cm
cm cm cm cm

cm
cm cm
cm cm
T he police car took up a strategic position at point A on the
map. T hree reserves were instructed to keep watch on the ap-
proaches to the bridge. One was assigned to patrol the stretch
from A to B; a second was assigned to the stretch A C , and the
third to A D. E ach patrolman walked continuously from one end
of his beat to the other, at a uniform rate of 2 miles per hour. A ll
three started from point A at 4:00 A .M .
On the first occasion thereafter when all three met at the
police car, a radio message advised the patrol that the robbers
had escaped through the south of the town.
If each block in the northern section is n o by 220 feet
ALGEBRAIC PUZZLESGROUP ONE 33
(measured from the centers of adjacent streets), what time did
the radio message arrive?
65. T H E E S C A L A T OR. "H enry," said the professor's wife,
"you're a mathematician. T ell me how many steps there are in
that escalator."
"Well, M artha, they certainly are difficult to count while
they are moving. But if you will walk up, and count the number
of steps you take from bottom to top, I think we can find the
answer. I will start with you, but will walk twice as fast. Just
watch me and take one step every time I take two."
When M artha reached the top she reported that she had
taken just 21 steps, while H enry had taken 28. T he professor was
then able to tell her exactly how many steps were in sight at one
time on the moving staircase.
66. T H E C A M P E R A ND T H E BOT T L E . A t 17 minutes past
one on a Sunday afternoon a camper embarked in his canoe and
commenced to paddle upstream at the rate of 4 miles per hour
against a current of 1 Yi miles per hour. A t 5 minutes past two
o'clock he drew abreast of a corked bottle floating in the stream.
Deciding against stopping to examine it, he continued on his way,
only to be overcome presently by curiosity. H e turned around,
paddled back, and caught up with the bottle just as it reached his
camp.
Removing the cork, he found a paper inside, on which was
printed in large letters:
H OW FA R DID YOU GE T FROM C A M P BE FORE
YOU GA VE IN T O YOU R C U RIOSIT Y?
T here is no reason why the camper should have paid any
attention to this odd message, but you know how these things
are. Fortunately he had noticed a large oak tree on the bank just
at the point he turned about, so the next day he paced the
distance from his camp to the tree and found the answer. Still,
don't you think he might have saved himself the walk?
ALGEBRAIC PUZZLESGROUP ONE 34
67. H I KE A ND H I T C H . Fifteen soldiers in charge of a
sergeant were detailed to go to a point 60 miles distant. T he only
transportation available was a jeep, which could carry only 5 men
besides the driver. T he sergeant undertook to carry the troops to
their destination in three loads. A s he left with the first party of 5,
he ordered the remaining 10 to commence hiking along the road.
H e unloaded the first party some distance from the goal, with
orders to hike the rest of the way. T hen he returned until he met
the 10, picked up 5 of them, and took them part of the way along
the route while the last 5 continued to walk. Finally he returned,
picked up the last party, and drove it the rest of the way to the
rendezvous. Whether by accident or designopinions differall
three parties arrived at the same moment.
T he men walked at a uniform rate of 4 miles per hour, while
the jeep averaged 40 miles per hour. H ow much time was saved
by the hiking?
68. I F A M A N C A N DO A J OB. "If a man can do a joh
in one day, how long will it take two men to do the job?"
No book of puzzles, I take it, is complete without such a
question. I will not blame the reader in the least if he hastily
turns the page, for I, too, was annoyed by "If a man" conundrums
in my schooldays. Besides, the answer in the back of the book was
always wrong. E verybody knows it will take the two men two
days to do the job, because they will talk about women and the
weather, they will argue about how the job is to be done, they
will negotiate as to which is to do it. In schoolbooks the masons
and bricklayers are not men, they are robots.
Strictly on the understanding that I am really talking about
robots, I will put it to you:
If a tinker and his helper can refabulate a widget in 2 days,
and if the tinker working with the apprentice instead would take
3 days, while the helper and the apprentice would take 6 days to
do the job, how long would it take each working alone to re-
fabulate the widget?
ALGEBRAIC PUZZLESGROUP ONE 35
69. FI NI S H T H E P I C T U RE . T he picture shows some equali-
ties of weight among objects of four kindscylinders, spheres,
cones and cubes. A t the bottom four cones are placed in the left
ALGEBRAIC PUZZLESGROUP ONE 36
pan of the scales. What is the least number of objects we can put
into the right pan to strike a balance?
70. T H E A L C A N H I GH WA Y. A n engineer working on the
A lcan H ighway was heard to say,
"A t the time I said I could finish this section in a week, I
expected to get two more bulldozers for the job. If they had left
me what machines I had, I'd have been only a day behind
schedule. A s it is, they've taken away all my machines but one,
and I'll be weeks behind schedule!"
H ow many weeks?
71. S E E S A W. T hree brothers go to a playground to play
on the seesaw. T he teeter board has a fixed seat at each end, 5
feet away from the trestle on which the board swings. When
A lfred and Bobby take seats, C harles, who weighs 80 pounds,
balances them by sitting on A lfred's side 21 inches away from the
trestle. When C harles sits in a seat, it takes both his brothers to
balance him, A lfred in the other seat and Bobby one foot nearer
the center. Now if Bobby takes A lfred's place, where must A lfred
sit to balance C harles?
72. A P ROBL E M I N C OU NT E RWE I GH T S . T he large
flats and other pieces of scenery used in a vaudeville theater are
counterweighted by sandbags, so that when they are moved only
a small portion of the weight has to be borne by the stagehands.
T he theater keeps on hand a set of metal counterweights for
occasional use with special pieces. A ny or all of the weights can
be attached quickly to an elevator rope. T here are five weights in
the set, so arranged that it is possible to compound any load which
is a multiple of 10, from 10 pounds up to the total of all five
weights together. T he choice of weights is such as to reach the
maximum possible total load. What are the several weights?
73. T H E A P OT H E C A RY' S WE I GH T S . A n apothecary has
a set of weights for use in the pans of his scales. H e is able by
ALGEBRAIC PUZZLESGROUP ONE 37
proper selection of weights to measure out every multiple of
J/2-gram from up to the total of the five weights together. If
the arrangement of the weights is such as to reach the maximum
possible total, what is it?
Notice that the apothecary can put weights in the same pan
as the load he is weighing out.
74. S A L L Y'S A GE . When Sally went to the polls to vote,
the clerk asked her age. "E ighteen," she replied. H e looked at her
quizzically. "You don't really mean it, do you?" he said. "Of
course not!" Sally laughed. "I gave myself the benefit of a year
less than a quarter of my real age." T he clerk permitted her to
cast her ballot, but he is still puzzled as to her true age. Surely
you are not.
75. A S OL D A S A BC . A lice is as old as Betty and C hris-
tine together. L ast year Betty was twice as old as C hristine. T wo
years hence A lice will be twice as old as C hristine. What are the
ages of the three girls?
76. FU M E R FROWNS . When M r. Fumer returned to his
tobacco shop after lunch, he found Joe, hii clerk, congratulating
himself on a stroke of business.
"While you were out," said Joe, "I managed to get rid of the
last two pipes out of that consignment from the E tna C ompany.
A chap came in who wanted a Vesuvius. I showed him that last
one we have, and he only wanted to pay a dollar for it. H owever,
we split the difference and he agreed to take it at $1.20. Of course
that was a 20% loss for us, but I let it go because he also agreed
to take the P opocatepetl pipe at the same price, and that gave us
a 20% profit. So we broke even."
Now, why did M r. Fumer frown?
77. C OM P OU ND I NT E RE S T . Determine within 5 cents
how much I must deposit so that after interest is compounded
ALGEBRAIC PUZZLESGROUP ONE 38
five times I will have 100 dollars in the bank, the interest rate
being 3%.
78. T H E S A VI NGS A C C OU NT . William Robinson de-
posited $100.00 in a savings account and left it untouched for
4/2 years. A t the end of that time (interest having been added
eight times) his passbook showed a total of $131.68. What was
the rate of interest (within one-half of one percent) ?
79. A FT E R FI VE O' C L OC K. T he clock shown in the illus-
tration has just struck five. A number of things are going to
times.
(a) A t what time will the two hands coincide?
(b) A t what time will the two hands first stand at right
angles to each other?
(c) A t one point the hands will stand at an angle of 30 de-
grees, the minute hand being before the hour hand. T hen the
former will pass the latter and presently make an angle of 60 de-
ALGEBRAIC PUZZLESGROUP ONE 39
grees on the other side. H ow much time will elapse between these
two events?
80. T H E C A RE L E S S J E WE L E R. On the last occasion I took
my watch to a jeweler to be cleaned, he made a careless mistake.
H e had removed the hands, and in replacing them he put the
minute hand on the hour-hand spindle and vice versa. Shortly
after I reclaimed the watch I found that the hands were taking
impossible positions. But eventually they reached a point where
they told the time correctly when read the normal way.
T he jeweler had set the hands at 2 :oo o'clock. What was the
first time thereafter when they showed the correct time?
81. C L OC K S E M A P H ORE , (a) A t what time between
two and three o'clock will the minute hand be as far from VI as
the hour hand is from XII ?
(b) What is the first time after noon that the minute hand
has as far to go to reach XII as the hour hand has passed XII?
V. Dissection of P lane Figures
82. C H A NGI NG A RE C T A NGL E T O A S QU A RE . One
of the basic dissection problems is to change a rectangle into a
square. T he general method is illustrated in the diagram.
Given the rectangle A BC D, first find the length of side of
the square of equal area. E xtend A B and measure off BE equal
to BD. Bisect A E in F. With F as center and FE as radius strike
an arc intersecting BD in G. T hen BG is the side of the equiva-
lent square.
C onnect A G and cut on this line. On A C measure A H equal
to GD. T hrough H draw a line parallel to C D, and cut on this
line.
40
DISSECTION OF PLANE FIGURES 41
Slide the triangular piece A BG downward to the right until
G lies on C D extended. T hen transfer the smaller triangle so that
A H coincides with GD. T he square thus formed is shown on the
right.
83. NOW RE VE RS E I T . If a rectangle can be dissected into
a square, then a square can be dissected into a rectangle. Given
the square A , cut it into the minimum number of pieces that can
be arranged to form a rectangle one of whose sides is B.
84. T H E KI T C H E N L I NOL E U M . M r. H ouseman wishes to
lay down linoleum on the floor of his kitchen, which is exactly 12
feet square. H e has a piece of linoleum just sufficient for the pur-
pose, in the form of a rectangle 16 feet by 9 feet. Obviously he
will have to cut this piece to make it fit, but he doesn't want to
cut it into any more parts than necessary. Fortunately, the
linoleum is uniformly brown, without pattern, so that he can cut
it in any manner he pleases without spoiling its appearance.
What is the least number of pieces into which the linoleum
can be cut to solve M r. H ouseman's problem?
85. T H E BROOM C L OS E T . Scarcely had M r. H ouseman
finished putting the linoleum on the kitchen floor than his wife
pointed out that he had forgotten the broom closet. She was most
anxious to have the square floor of the closet covered, so M r.
H ouseman measured it and later purchased cheaply a remnant
B
DISSECTION OF PLANE FIGURES 42
of the shape shown in the diagram, just sufficient in area for the
purpose. M rs. H ouseman was aghast at the idea of using check-
ered linoleum in the closet when the kitchen linoleum was plain
brown. But M r. H ouseman remarked that the broom closet was
scarcely likely to be exhibited to guests in any event, and he had
his way.
"Only mind," said M rs. H ouseman, "that you do not cut
that remnant any more than you have to. A nd don't you spoil the
pattern!"
C an you help M r. H ouseman comply with these conditions?
86. S U M M I NG T WO S QU A RE S . Given two squares of
different size, cut them into the least number of pieces that can
be reassembled to form one square.
T he squares are assumed to be incommensurable. If they
stand in simple integral ratio a more economical dissection may
be possible.
87. FROM A T O Z . T he report card jubilantly displayed
by T yrus gave his mark in geometry as ioo%.
"H umph," remarked his elder brother, C utler. "I suppose
you think you know geometry from A to Z ."
"Sure I do."
"P rove it, then." So saying, C utler drew a large block-letter
capital A . "L et's see you divide this A into four parts which can
be put together to make a Z . A nd mind you don't turn any piece
over."
T yrus accepted the challenge and set to work. H is first two
DISSECTION OF PLANE FIGURES 43
attempts were ruled out, because each time he had to turn a piece
of the A over to make it fit into the Z . But on the third try he
found a solution which C utler had to admit was correct. C an
you find it?
88. T H E M I T RE FA L L A C Y. Sam L oyd, who invented
many ingenious puzzles, once propounded the following:
A carpenter has a mitre, of the shape shown in the diagram
a square with one quarter cut out. H e wants to saw the mitre
44 DISSECTION OF PLANE FIGURES 44
into pieces and fit them together to form a perfect square. What
is the least number of pieces necessary and their shape?
L oyd's solution was to cut off the two triangular points as
shown, turn them down into the notch to form a rectangle, then
cut the rectangle on the "step" principle to make a square. (In
this connection, see solution of No. 84The Kitchen Linoleum.)
For once, L oyd slipped. T he proposed solution is impossible.
C an you prove this statement?
89. WH A T P ROP ORT I ONS ? T he question here asked will
be easier to answer after you have read the solution to No. 88
The Mitre Fallacy.
Suppose that a rectangle is capable of being cut into two
pieces, in the stepwise manner previously described, which can
then be re-arranged to form a square. Suppose that the staircase
cut contains thirteen steps in one direction and twelve in the
other.
What is the ratio of the width of the rectangle to its depth?
90. T H E ODI C FORC E . Reference has been made fre-
quently in literature to a mysterious force known only as "od."
T hrough researches covering the whole period from the cabala
of ancient E gyptian astrology to the lexicon of the T wentieth
C entury crossword puzzle, I am at last able to reveal the precise
nature of this force.
It is a property of certain integers and thus has the omnipo-
tence of all mathematics. For example:
3 Xi 5 = i 3 +
I
5 + i 7 = 3
I
5 / 7 and I
2
+3 * +5
2
=7 X5 -
From these assertions it is readily seen that any square can
be dissected into 7 pieces, which can then be arranged to form
3 squares whose areas stand in the ratio 1: 3: 5.
T o discover how to make the dissection you need only use
your n-od-dle.
DISSECTION OF PLANE FIGURES 45
91. T H E P I A NO L A M P . H ere is an easy dissection prob-
lem. L ay a sheet of thin paper on the page and trace the outline
of the piano lamp. C ut out the silhouette around the lines. T hen
cut the lamp into pieces which can be re-arranged to form a
solid circular disk.
DISSECTION OF PLANE FIGURES 46
92. C ONS T RU C T I ON OF A P E NT A GON. A s some of the
puzzles in this book involve pentagons, I will here explain how tc.
construct the figure.
Given the circle O in which the pentagon is to be inscribed.
Draw two diameters at right angles. Bisect the radius OX in A .
With A as center and radius A B lay off A C equal to A B. T hen
BC is the side of the pentagon.
What is the ratio of the side of the inscribed pentagon to the
radius of the circle? T his question can be answered without resort
to trigonometry.
I I
1
O A xl
93. T H E A M U L E T . T he P ythaclideans, that strange race
inhabiting the land of Rectilinea, are said to be very superstitious,
as is attested by the fact that no P ythaclid will venture out of his
abode without his amulet. T his amulet consists of 6 pieces of
metal strung together to make a necklace. T he number 6 is well-
DISSECTION OF PLANE FIGURES 47
known to have magic properties, being perfect, and it is no
wonder that the P ythaclideans deem this amulet to be singularly
efficacious in warding off the dreadful Riemanns and L obat-
chevskis.
T he construction of an amulet involves strange mystic rites.
T he Geometer or high priest first fashions a square plate of metal
appropriate to the suppliant. T he absolute size of this plate is
determined by the nth derivative of his nativity, the submaxillary
function of his right ascension, and suchlike matters that do not
concern us here. T he Geometer then dissects the plate into six
pieces which can be rearranged to form a regular pentagon, or
into a parallelogram different from the square.
T he P ythaclideans fondly believe that they alone are pos-
sessed of the secret of this construction, but we believe that it
can be rediscovered.
94. FOU R-S QU A RE . H ere is an easy exercise in dissection.
H ow many different kinds of pieces can be cut from a checkered
board, if each piece must contain just four squares and if all cuts
must be made along the lines between two squares?
You do not have to cut all the pieces from a single checker-
board 8x 8. You can have all the board you want for the pur-
pose. We will count as different any two pieces which, although
congruent, have the colors arranged in reverse fashion.
95. J A C K O
1
L A NT E RN. "Oh dear," said E mily, "I can
hardly wait until H allowe'en comes!"
"Why?" said P rofessor Snippet, her father.
"Because I want to get a pumpkin and make a jack o'
lantern."
"Well, we don't have to wait for H allowe'en for that," he re-
marked. "I'll make you a jack o' lantern now."
T he P rofessor cut a circular disk out of white paper, then
cut the disk into pieces which he arranged on a red blotter to
form the jack o' lantern here illustrated. T he red, showing
DISSECTION OF PLANE FIGURES 48
through the holes for eyes, nose, and mouth, gave jack a jolly
look that quite captivated E mily. She begged her father to "Do
it again!"
P erhaps the reader would like to comply with E mily's re-
quest.
96. A RE M A RKA BL E OC C U RRE NC E . P atrons of the
Gambit C hess C lub are still talking about the extraordinary be-
havior of M r. Sawyer in the matter of his encounter with M r.
P unner.
It seems that the venerable M r. Sawyer, a charter member
of the club, was one evening sitting in its quarters awaiting the
appearance of some prospective opponent. Young M r. P unner
chanced to wander in, and M r. Sawyer asked amiably, "Would
DISSECTION OF PLANE FIGURES 49
you care to play a game?" With the verve of youth the other re-
plied "C H E SS!"thereby intending a pun which he is not the
first nor will he be the last to perpetrate.
During his threescore years and ten M r. Sawyer has no doubt
been exposed to this witticism numerous times. A t all events,
scarcely was the word uttered than M r. Sawyer jumped to his
feet, whipped out a jigsaw, and before the astonished eyes of M r.
P unner proceeded to cut to pieces one of the chessboards. H aving
rent it to his satisfaction, he then arranged the pieces as shown
in the accompanying picture.
E H E 5 g >
"T here!" he snarled at the somewhat intimidated young
man. "T here, if you please, is the spawn of your side-splitting
humor! T ake it home with you, nail it on your door, that all
passers-by may behold and marvel! Sir, I wish you good evening."
Saying which he strode out of the club.
It must be said in behalf of M r. P unner that he took this
reproof with good grace. H e was even heard to point out that
M r. Sawyer had performed a rather neat trick, by way of cutting
the chessboard into no more pieces than were absolutely neces-
sary in order to form the final tableau.
T he reader may be interested to verify this fact
97. T A NGRA M P A RA DOXE S . A long time agoat least
4,000 yearsa C hinese devotee of puzzles dissected a square into
7 pieces, as shown in Fig. 1, and amused himself by arranging the
pieces to suggest pictures. From that day to this, interest in the
pastime of "tangrams" has never died. In fact, it has acquired a
respectable literature, commencing with seven books of tangram
pictures compiled in C hina two millenniums before the C hristian
DISSECTION OF PLANE FIGURES 50
FIG. I T h e Square T angram
Set
FIG. 2 T h e Frisky Dog
FIG. 3T he M an Who
L ost a Foot
DISSECTION OF PLANE FIGURES 51
FIG. 4 T h e Obl ong T angr am
Set
era, at least two of which have survived. M any other com-
pendiums have been published in modern times.
T he invariable practice in making a tangram picture is to
use all 7 pieces. T he completed silhouette can be regarded as a
puzzle: how to form it out of the 7 pieces. M ost of the books on
tangrams present the pictures in just this fashion.
For example, Fig. 2 shows a frisky dog. T he reader will have
no difficulty in discovering how to make him.
T he tangrams lend themselves to the construction of para-
DISSECTION OF PLANE FIGURES 52
doxes, of which Fig. 3 is an example. H ere are two men, much
alike in appearance, but one has a foot and the other has not.
T he same 7 pieces were used to construct each figure. Where
does the foot come from out of the first figure?
Some years ago, tangrams cut from an oblong, as shown in
Fig. 4, achieved wide popularity. I remember a newspaper edi-
torial, commenting in derisive terms on the resurgence of "tan-
gram parties." In other periodicals I found satirical references
to the national preoccupation with "how to make the fish." T his
creature, shown in Fig. 5, was portrayed in the advertisements
of a manufacturer of tangram sets.
DISSECTION OF PLANE FIGURES 53
H ere is a paradox I constructed with the oblong tangram
set. Fig. 6 shows the silhouette of a diamond at one stage of its
cutting, together with its appearance at a later stage when the
point has been truncated. Both figures are made from the same
7 tangrams. What happened to the triangular part at the top
of the first picture?
T here is of course no limit to the number of different ways
a set of tangrams can be cut. T he last three diagrams show other
sets that have achieved some popularity. Fig. 9 is the only set I
have seen which departs from rectilinear forms.
VI . Geometrical P uzzles
98. T H E C L U B I NS I GNI A . T he Geometry C lub of our
local high school designed for itself a membership pin in the form
shown by the diagram. A rchie M ead took the design to Jonathan
Sparks, the jeweler. M r. Sparks asked, "H ow large do you want
this pin?" A rchie replied, "We would like it to be just two-thirds
of an inch in diameter. T he larger circle of course is the outside
edge." "H um," said M r. Sparks, "that is going to make the
letters GC pretty small. What size do you expect them to be?"
54
GEOMETRICAL PUZZLES 55
A rchie said, "Of course that depends on how much margin is left
between the letters and the triangle. I suppose you ought to do
whatever you think best on that point. But I can tell you that the
6ide of the triangle will be . . ."
Surely you are as bright as A rchie.
99. T H E FE RRYBOA T GA T E . On a recent ferryboat ride,
I noticed certain features in the gate across the roadway.
T his gate was the usual "lazytongs" affair as shown in the
illustration. It consisted essentially of 7 vertical rods, connected
by diagonal members pivoted to collars on the rods. T he collar
at the top of each rod was screwed tightly thereon, but the collars
in the lower two tiers were free to slide up and down the rods
according as the gate was distended or collapsed.
T he leftmost rod was pivoted into lugs attached to a heavy
post. T he rightmost rod could be loosely attached to a similar
post on the opposite side of the roadway, by a large ring linked to
the rod which could be slung over the top of the post. On the
leftmost rod I noticed an extra collar or flange, fixed between
the two tiers of sliding collars. T he evident purpose of this flange
was to prevent the bottom collar from sliding up the rod beyond
this point, thus preventing the gate from being distended further.
T he puzzle that suggested itself is: H ow far does the gate
reach if pulled out to the maximum distance permitted by the
check flange?
56 GEOMETRICAL PUZZLES 56
I e s t i ma t e d t he di me ns i ons of t he g a t e as f ol l ows : l engt h of
pr i nc i pa l di a g o na l me mb e r s ( b e t we e n out e r pi v o t s ) , 2/2 f e e t ;
di s t a nc e b e t we e n t o p a n d b o t t o m col l ar s ( pi v ot t o pi v ot ) on l ef t -
mo s t r o d wh e n b r o u g h t as cl ose as t he c he c k wi l l pe r mi t , 1 /2 f e e t ;
hor i z ont a l di s t anc e of c e nt e r of e a c h c ol l ar pi v ot f r o m c ent er of
r od, 2 i nches .
A s s u mi n g my es t i mat es t o be e xac t l y c or r ec t , h o w f a r c a n t he
g a t e be e x t e n d e d ?
100. S T RI KI NG A BA L A NC E . T h e d i a g r a m s hows a 60-
p o u n d we i g h t o n one e nd of a l ever , wh i c h has a f u l c r u m in t he
mi d d l e . M a r k t he e x a c t poi nt on t he r i ght si de of t he l ever whe r e
t he 1 05 - po und we i g h t mus t be pl a c e d so t hat t he l ever wi l l
b a l a nc e hor i z ont al l y .
Y o u ma y ne gl e c t t he we i g ht of t he l ever i tsel f . T h e mas s of
e a c h we i g h t ma y be c ons t r ue d t o be c onc e nt r a t e d at t he mi dpo i nt
of its bas e.
101. A N I NT E RC E P T P ROBL E M . T h r e e t a ng e nt ci rcl es of
e q ua l r a di us r a r e d r a wn , al l c ent er s be i ng o n t he l i ne O E . F r o m
56GEOMETRICAL PUZZLES 57
O , t he out e r i nt er s ect i on of t hi s axi s wi t h t he l e f t - ha nd c i r c l e,
l i ne O D is d r a wn t a ng e nt t o t he r i g ht - ha nd ci r cl e. Wh a t is t he
l engt h, i n t er ms of r, of A B , t he s e g me nt of t hi s t a ng e nt wh i c h
f or ms a c ho r d i n t he mi d d l e c i r c l e ?
102. T H E BA Y WI NDOW. I n r e pa i nt i ng a hous e, M r .
L i ns e e d e nc o unt e r e d t he di f f i c ul t y s h o wn i n t he i l l us t r at i on. O n
one si de, a pr o j e c t i ng b a y wi n d o w pr e v e nt e d hi s s et t i ng hi s 20-
f oot l a dde r cl ose t o t he wa l l . H o we v e r , he f o u n d t hat he c o ul d
56
GEOMETRICAL PUZZLES 58
r e a c h s ome pa r t of t he wal l a b o v e t he wi n d o w by pl a c i ng hi s
l a dde r s nugl y agai ns t t he b a y a n d al so agai ns t t he wal l .
T h e o ut e r ba t t e n of t he b a y wi n d o w agai ns t wh i c h t he l a dde r
rests is 3 f e e t f r o m t he wa l l a nd 12 f e e t a bo v e t he g r o und. H o w
f a r u p t he wa l l does t he l a dde r r e a c h ?
103. T H E E XT E NS I ON L A DDE R. H e r e is a not he r puz z l e
a b o ut l adder s , b ut it is r a t he r mo r e di f f i c ul t t ha n The Bay Win-
dow. I n f a c t , t he r e a de r is adv i s e d t o pos t pone t a c kl i ng it unt i l
he ha s r e a d t hr o ug h C h a p t e r VIII, PUZZLES ABOUT INTEGERS.
T h e a c c o mp a n y i n g i l l us t r at i on is t he s cene of t he M a y f i e l d
Bu i l d i ng fire. A t t hi s s pe c t a c ul a r e v e nt , a n u mb e r of per sons we r e
t r a p p e d o n t he r oof of t he b u r n i n g s t r uc t ur e, a n d we r e r es c ued by
t he me a n s he r e de pi c t e d. T h e firemen p u t u p a t hr ee- s ect i on 70-
f o o t e xt e ns i on l a dde r i n a n al l ey b a c k of t he M a y f i e l d Bui l di ng.
T h e l o we r e nd of t he l a dde r wa s set agai ns t t he f a c e of t he bui l d-
i ng on t he oppos i t e si de of t he s t r e e t ; t he up p e r e nd r est ed agai ns t
t he c or ni c e of t he M a y f i e l d Bui l di ng . A s hor t er l a dde r wa s pl a c e d
agai ns t t he l at t er bui l di ng a t t he s i de wal k a nd its u p p e r e nd wa s
l as he d t o t he e xt e ns i on l a dde r ( a t t he t op of t he l owe r mos t sec-
t i on) t o br a c e i t. A t hi r d l a dde r wa s l ai d ne a r l y hor i z ont al l y
acr oss t he st reet bes i de t hi s s t r uc t ur e, as a t r af f i c bar r i er . T h i s
l a d d e r b e i ng a l i t t l e l onge r t ha n t he wi d t h of t he al l ey, one e nd
r es t ed a t t he bas e of t he M a y f i e l d Bui l di ng b ut t he ot he r e nd
we d g e d a f e w f e e t u p f r o m t he s i de wal k agai ns t t he f a c e of t he
oppos i t e bui l di ng .
T h e t r af f i c ba r r i e r l a dde r al s o wa s l as hed t o t he ext ens i on
l a dde r , a n d f r o m t he f a c t t hat a ma n of a v e r a g e he i g ht c oul d
j us t wa l k u n d e r t he poi nt wh e r e t he l adder s crossed, wi t ho ut
s t oopi ng, we ma y e s t i mat e t ha t t hi s poi nt wa s j us t 5 f eet 10 i nches
a b o v e t he s i de wa l k.
T h e ba r r i e r l a dde r wa s 22 f e e t 11 i nc hes l ong. T h e ext ens i on
l a d d e r p r o j e c t e d 2 f eet 2 i nc hes a b o v e its poi nt of c ont a c t wi t h
t he c or ni c e of t he M a y f i e l d Bui l di ng .
T h e puz z l e is t o de t e r mi ne t he he i g ht of t he M a y f i e l d Bui l d-
i n g a b o v e t he s i de wal k.
56GEOMETRICAL PUZZLES
59
at the M ayfi el d Building Fire
56
GEOMETRICAL PUZZLES 60
104. T H E S P I DE R A ND T H E FL Y. T h i s i nt er est i ng b ut
we l l - k n o wn puz z l e is quo t e d he r e by wa y of pr e a mbl e t o The
Spider's Cousin.
A s pi der l i ved i n a r e c t a ng ul a r r o o m, 30 f eet l o ng by 12 f eet
wi d e a n d 12 f e e t hi g h. O n e d a y t he s pi der pe r c e i v e d a fly i n t he
r o o m. T h e s pi der a t t hat t i me wa s o n one of t he e nd wal l s , one
f oot b e l o w t he c e i l i ng a n d mi d wa y b e t we e n t he t wo si de wal l s.
T h e fly wa s o n t he oppos i t e e nd wal l , al s o mi d wa y be t we e n t he
si de wal l s , a nd one f oot a b o v e t he floor. T h e s pi der cl ever l y r a n
b y t he short est possi bl e c our s e t o t he fly, wh o , pa r a l y z e d by f r i ght ,
s uf f e r e d hi ms el f t o be de v o ur e d.
T h e puz z l e i s: Wh a t c our s e di d t he s pi der t ake a nd h o w f a r
di d he t r a v e l ? I t is unde r s t ood he mus t a dhe r e t o t he wal l s, e t c . ;
he ma y not d r o p t hr o ug h s pac e.
105. T H E S P I DE R'S C OU S I N. I t seems t hat t he spi der
me nt i o ne d a b o v e h a d a cous i n wh o l i ved i n t he P e nt a g on Bui l d-
i ng.
N o w t hi s is g o i ng t o be a ver y sad st ory, a nd a ny o ne whos e
e mot i ons ar e easi l y h a r r o we d h a d bet t er r e a d no f ur t he r .
I t al l h a p p e n e d wh i l e t he bui l di ng wa s u n d e r c ons t r uc t i on.
T h e d i a g r a m s hows t he P e nt a g o n Bui l di ng i n its finished st at e.
T w o c onc e nt r i c pe nt a g o na l wal l s encl os e a r abbi t wa r r e n of
of f i ces. A c c e s s t o t he of f i ces is g a i ne d by 4 c i r c umf e r e nt i a l c or -
r i dor s a nd 8 t r ans ver s e cor r i dor s . A t t he t i me of o ur st ory, ho w-
ever , onl y t he i nne r a n d out e r wal l s a nd t he first floor h a d be e n
c o mp l e t e d .
T h e spi der ' s c ous i n wa s i ns pe c t i ng t he s t r uc t ur e, wi t h a v i e w
t o t a k i ng s pac e, wh e n o ne of hi s spi es r e por t e d t he pr es enc e of a
pa r t i c ul a r l y s uc c ul e nt fly at t he poi nt ma r k e d i n t he d i a g r a m
on t he i nne r wa l l mi d wa y b e t we e n t wo ver t exes , a nd 9 f eet a bo v e
t he floor. T h e s pi der , wh e n he r e c e i v e d t he i nt el l i gence, wa s i n
t he f ar t he s t v e r t e x of t he out e r wa l l , al so 9 f eet a b o v e t he floor.
O u r he r o i mme di a t e l y set a b o u t c a l c ul a t i ng hi s short est
c our s e f o r e a c h t he fly. H e wo u l d h a v e t o g o by wa y of t he floor,
be c a us e t he c e i l i ng wa s not yet i n pl ac e . T h a t pa r t wa s easy, b u t
56GEOMETRICAL PUZZLES 61
wh e n i t c a me t o t he angl es t he s pi der wa s baf f l ed. No t h a v i n g t hi s
book wi t h h i m t o e x pl a i n t he c ons t r uc t i on of a p e nt a g o n, he t r i ed
t o sol ve t he pr o bl e m by pe r c e nt a g e a nd c o mp o u n d i nt er est , a nd
onl y got hi msel f i n a f r i g ht f ul st ew. H e r us hed of f wi l l y- ni l l y i n al l
di r ect i ons, a nd, t o ma k e ma t t e r s wor s e, f o u n d t hat wh i l e he h a d
be e n c og i t a t i ng t he wo r k me n h a d p u t u p t he i nt er i or par t i t i ons .
No t h i n g mo r e has be e n he a r d f r o m hi m, a n d we c a n onl y s uppos e
t hat he is still t r y i ng t o f i nd hi s wa y t o a ny g i v e n poi nt .
T h e a r a c hni d wo r l d wi l l be v e r y a ppr e c i a t i v e of t he r e a de r
wh o wi l l sol ve t hi s spi der ' s pr o bl e m, t o wa r d t he d a y wh e n t he
P e nt a g o n Bui l di ng is c o nv e r t e d t o a r i di ng a c a d e my or d a nc e hal l .
56
GEOMETRICAL PUZZLES 62
O h yes, di mens i ons . F o r r easons of s ecur i t y I c a nno t gi v e t he
a c t ua l di me ns i ons , b ut t he f o l l o wi ng wi l l ser ve. E a c h si de of t he
out e r p e n t a g o n is 1 500 f e e t ; e a c h si de of t he i nner pe nt a g o n is
700 f eet . T h e s e me a s ur e me nt s ar e t a ke n i nsi de t he wal l s , on t he
pl a ne s t he s pi de r wi l l h a v e t o t r aver s e.
T h e s pi der s d o not i nsi st o n k n o wi n g t he di s t ance t o t he f r ac -
t i on of a n i nc h. T h e y wi l l be sat i sf i ed wi t h a pl a n of t he r out e a nd
t he di s t anc e t o t he near es t f oot .
106. T OU RI NG T H E P E NT A GON. Wh i l e we h a v e t he
i l l us t r at i on of t he P e n t a g o n Bui l di ng be f or e us, l et us e x pl or e i t.
Suppo s e we st ar t f r o m t he s a me poi nt as t he spi der ' s c ous i n. We
r es ol ve t o t r aver s e e v e r y c or r i dor o n t he floor a nd r et ur n t o our
s t ar t i ng poi nt . S u c h a t our is not possi bl e wi t ho ut pas s i ng t hr o ug h
s ome c or r i dor s mo r e t h a n o nc e , b ut we wa n t t o mi ni mi z e t he
n u mb e r of s uc h dupl i c a t i ons . Wh a t is t he short est r out e we c a n
t a k e ?
I n c as e y o u wi s h t o c o mp u t e t he l e ngt h of t he j our ne y , I wi l l
me n t i o n t hat t he c or r i dor s a r e 12 f e e t wi de , b ut per s onal l y I don' t
c ar e be c a us e I a m ne v e r g o i ng t o set out o n t hi s ma r a t h o n !
107. H OW T O DRA W A N E L L I P S E . T h e i l l ust r at i on
s hows one wa y of d r a wi n g a n el l i pse. F i x t wo pi ns or t humbt a c ks
o n t he p a p e r a nd t i e b e t we e n t he m a l e ngt h of s t r i ng or t hr ead,
a l l o wi ng s ome sl ack. Wi t h t he poi nt of a penc i l d r a w t he t hr ead
t aut , t he n s we e p r i ght a r o u n d t he pi ns, k e e pi ng t he pe nc i l poi nt
as f a r a wa y f r o m t h e m as t he s t r i ng wi l l pe r mi t .
T h i s c ons t r uc t i on serves t o s how j us t wh a t a n el l i pse is. I t
is t he locus of al l poi nt s ( a locus is t he p a t h of a mo v i n g poi nt
t ha t sati sf i es c e r t a i n c o ndi t i o ns ) , t he s ums of who s e di st ances f r o m
t wo fixed poi nt s a r e e qua l . T h e fixed poi nt s, t he pi ns, a r e c al l ed
the foci (singular, focus). T he major axis of the ellipse is its
wi d t h me a s u r e d o n t he l i ne of t he f oc i . T h e minor axis is t he
wi d t h me a s u r e d on a l i ne t h r o u g h t he c e nt e r a t r i ght angl es t o
t he ma j o r axi s.
63
H e r e is a s i mpl e pr o b l e m bas ed on t he c ons t r uc t i on. I f t he
pi ns ar e set 6 i nches a pa r t , a n d t he s t r i ng is 10 i nc hes l ong, wh a t
wi l l be t he l engt hs of t he a x e s ?
108. ROA DS T O S E DA N. T h e A l l i e d a d v a n c e t hr o ug h
nor t he r n Fr a n c e i n A u g u s t , 1944, wa s so r a pi d t ha t f o r wa r d el e-
me nt s on s ever al oc c as i ons o ut r a n t hei r ma ps . T h e f o l l o wi ng
st ory is v o u c h e d f or by a n ar t i l l er y of f i c er of t he a r mo r e d c o l u mn
whi c h, i mme di a t e l y u p o n t he f al l of P ari s, wa s sent t o wa r d t he
Be l g i a n bor de r .
T h i s of f i c er wa s i n c o mma n d of a ba t t e r y of h e a v y f i el d guns .
A v o i d i n g ma i n hi g hwa y s , t he guns we r e b e i ng t a ke n by less-
ma r k e d r out es t h r o u g h f or es t ed l a nd i n t he g e ne r a l di r e c t i on of
Se da n. Re c o nno i t e r i ng pl ane s r e por t e d e n e my f or c es i n t he s out h-
56
GEOMETRICAL PUZZLES 64
e r n out ski r t s of t he t own. Or d e r s c a me t o t he bat t er y t o hal t , pr e-
pa r e f o r ac t i on, a nd shel l t he e ne my .
T h e ma p s t he n i n possessi on of t he c o mma n d e r di d not c ov e r
t he t er r ai n al l t he wa y t o Se da n. I n f ac t , t he bat t e r y h a d al mos t
" s t e ppe d o f f " its l ast ma p , a por t i on of wh i c h is he r e s hown. T h e
X ma r k s t he pos i t i on of t he bat t e r y at t he t i me it r e c e i ve d t he
or der s.
T h e que s t i on na t ur a l l y ar ose, h o w t o poi nt t he guns u p o n
S e da n, wi t ho u t a ma p t o s how t he e x a c t l oc at i on of t he t own. On e
of t he c r e w, f a mi l i a r wi t h t he r egi on, wa s abl e t o st at e posi t i vel y
t ha t c e r t ai n por t i ons of t he ne a r by hi g hwa y s he a d e d di r ect l y
t o wa r d Se da n. T h e por t i ons ar e i ndi c at e d by a r r ows on t he ma p .
T h e s ugges t i on wa s ma d e t hat t he guns be ha ul e d t o one of t hese
r oads a nd a i me d t her eby.
Bu t t he c o mma n d e r wa s unwi l l i ng t o de l a y ac t i on or t o l eave
t he e x c e l l e nt c o v e r i n wh i c h t hey f o u nd t hems el ves . A n o t h e r sug-
ges t i on wa s t o l ay t he ma p on a l ar ger pi ec e of pa pe r , e x t e nd t wo
s t r ai ght l i nes f r o m t he ma r k e d r oads t o a n i nt er sect i on, c onne c t
t hi s i nt er s ect i on wi t h t he X o n t he ma p , a nd so de t e r mi ne t he
c or r e c t a z i mut h. T h e i de a wa s i n e f f e c t t o e x t e nd t he ma p so as
t o pl ot t he pos i t i on of Se da n. A t hi r d s ugges t i on wa s t o or i ent t he
56GEOMETRICAL PUZZLES 65
ma p on t he g r o und, si ght a l o ng t he ma r k e d por t i ons of t he r oads
t o s ome o bj e c t on t he hor i z on, t he n use t hi s o bj e c t as a t ar get .
Bot h of t hese s ugges t i ons we r e r e j e c t e d as i mpr a c t i c a bl e u n d e r t he
c i r c ums t anc es .
T h e c o mma n d e r s ol ved t he pr o b l e m v e r y s i mpl y. H e d r e w a
l i ne t hr oug h X , wh i c h , if e x t e nde d, wo u l d pass t hr o ug h t he poi nt
of i nt er sect i on of t he t wo r oa d s egment s ( e x t e n d e d ) . H e di d not
need t o find thi s poi nt of i nt er s ec t i on ( by us i ng a l ar ge pi ec e of
pa pe r ) t o d r a w hi s l i ne a c c ur a t e l y a nd so de t e r mi ne t he c or r e c t
a z i mut h. Fi nal l y, he di d not e v e n ne e d t o use a b o w c o mp a s s a
st r ai ght e dg e wa s suf f i ci ent .
I wo n d e r if t he r e a de r c a n s how t hi s c ons t r uc t i on.
109. T H E BI L L I A RD S H OT . Joe Du f f e r is pl a y i ng bi l l i ar ds,
a nd he n o w finds hi msel f i n a q u a nd a r y . T h e pos i t i on is s ho wn i n
t he di a g r a m. T h e t wo ne ar by bal l s ar e " r e d " a n d " s p o t " ; t he
i sol at ed bal l , " p l a i n, " is Joe' s c ue bal l . A s y o u c a n see, he has a n
easy c a r o m. U nf o r t u na t e l y , t he g a me is not c a r o m bi l l i ar ds but
t hr ee- c us hi on bi l l i ards. T h e r ul e of t hi s g a me is t hat t he c ue bal l
mus t t o uc h t hr ee c us hi ons be f or e c o nt a c t i ng t he s ec ond o b j e c t
bal l . ( No r est r i ct i on as t o wh e n c o nt a c t wi t h t he first o b j e c t bal l
mus t be ma de . )
Joe mi g ht try hi t t i ng e i t he r r e d or s pot di r ec t , t he n s e ndi ng
t he c ue bal l a r o u nd t he t abl e t o r e t ur n t o t he ot her . Bu t wi t h t he
56
GEOMETRICAL PUZZLES 66
o b j e c t bal l s so cl ose t oge t he r , he dec i des t o pl ay a " b a n k s hot . "
T h i s me a n s t hat he wi l l send t he c ue bal l a r o und a t l east t hr ee
c us hi ons be f or e hi t t i ng e i t he r o bj e c t bal l , t he n c o mpl e t e t he c ount
by a c a r o m. ( I nc i de nt a l l y , t he c ue bal l ne e d not t ouc h t hr ee dif-
ferent c us hi ons ; i t ma y t o uc h t he s a me c us hi on t wi c e a t di f f e r e nt
t i mes , b u t s uc h shot s a r e r ar e . )
L e t us advi s e J o e as t o t he best t hr ee- c us hi on ba nk s hot t o
t ry. We wi l l a s s ume t hat a bal l r e bounds f r o m a c us hi on i n a t r ue
f as hi on, t he a ng l e of r ef l ec t i on be i ng e qua l t o t he a ng l e of i nci -
de nc e . ( A c t ua l l y , t he t heor et i c a ng l e of r ef l ect i on is us ual l y mo di -
f i ed by " E n g l i s h " or " t wi s t " i mpa r t e d by t he pl ay e r i n s t r i ki ng t he
c u e bal l or a c qui r e d by i t t hr o ug h t he f r i c t i on of t he c l ot h. ) I t is
not e n o u g h t o s how Joe t he g e ne r a l di r e c t i on i n wh i c h t o a i m; we
mus t s ho w h i m t he e x a c t poi nt at wh i c h t o t o uc h t he fi rst c us hi on
i n or de r t o assur e t he c ount .
VI I . P roperties of Digits
110. DI GI T S A ND I NT E GE RS . A digit is any one of the
s ymbol s i , 2, 3, 4, 5, 6, 7, 8, 9. I t is c o nv e ni e nt t o i nc l ude o ( z e r o)
as a di gi t , a l t ho ug h i n st ri ct pa r l a nc e o o f t e n is e x c l ud e d .
A n integer is a wh o l e nu mb e r , wh e t h e r expr es s ed b y one
di gi t , as 7, or by ma n y di gi t s, as 65, 91 3, 448, 065, 81 4.
A r i t hme t i c is t he s ci ence of t he f u n d a me n t a l o p e r a t i o n s
addi t i on a n d s ubt r ac t i on, mul t i pl i c a t i o n a n d di vi s i on. A s s uc h i t
is pr er equi s i t e t o t he s t udy of al l hi g he r ma t he ma t i c s , as we l l as t o
de a l i ng wi t h t he g r oc e r , t he ba nke r , t he i n c o me t a x c ol l ec t or . E l e -
me nt a r y s chool a r i t hme t i c al so ma k e s s ome s t udy of t he proper-
ties of di gi t s a nd i nt eger s. O f c our s e i t j us t ski rts t he e dg e of t he
s ubj ec t . T h o r o u g h g o i n g i nqui r y i nt o t he pr ope r t i e s of numb e r s is
l ef t t o " h i g h e r a r i t hme t i c " a nd " t he o r y of numb e r s . "
T h e puzzl es i n t hi s pa r t of t he bo o k c o nc e r n t he pr ope r t i e s of
di gi t s a nd i nt egers. T h e y v e nt ur e a l i t t l e wa y i nt o t he t he or y of
number s . But I do not e x pe c t t he r e a de r t o be a c q ua i nt e d wi t h
t he f or mul a s of t hat f a s c i na t i ng s ci ence. O n t he c ont r a r y , my a i m
is t o e nc o ur a g e t he r e a de r t o r e di s c ove r s ome of t hese f o r mul a s f or
hi ms el f . A l l t he puzzl es c a n be sol ved by k n o wl e d g e of a r i t hme t i c
a nd e l e me nt a r y a l ge br a , pl us a l i t t l e i ng e nui t y i n de v i s i ng me t h -
ods of at t ac k.
111. DI GI T A L ROOT S . I f t he s u m of t he di gi t s of a n u m-
ber is di vi si bl e by 3, t he n t he n u mb e r is di vi si bl e by 3. T h i s is onl y
one of t he ma n y us ef ul f ac t s t hat c a n be i nf e r r e d f r o m root s.
67
68 PROPERTIES OF DIGITS
T h e digital root of a n i nt eger is t he si ngl e i nt eger r e a c he d by
c o nt i nue d s u mma t i o n ox its di gi t s. Gi v e n t he i nt eger 91 7 , 5 3 4; to
f i nd its di gi t al r oot : a dd t he di gi t s ; t he s u m is 29; a dd t he di gi t s
of 29; t he s u m is 1 1 ; a d d t hese di gi t s ; t he s um is 2 : t he di gi t al r oot
of 917,534 ^ 2.
No t e t hat o c a n ne v e r be a di gi t al root . T h e r e ar e onl y 9
possi bl e root s.
A l l a r i t hme t i c ope r at i ons c a n be c he c k e d by di gi t al root s. T h e
r oot of t he s um, di f f e r e nc e , pr oduc t , or quot i e nt of t wo i nt eger s
c a n be de t e r mi ne d by p e r f o r mi n g t he s a me ope r a t i on on t hei r
root s. F o r e x a mp l e :
9530624 2
87235 7
47653120 1
28591872 6
19061248 4
66714368 5
76284992 2
831803984640 9
T o c he c k t hi s mul t i pl i c a t i on, we de t e r mi ne di gi t al root s of
mul t i pl i c a nd, mul t i pl i e r , a n d s uppos e d pr oduc t . F o r t he fi rst t wo
w e h a v e 2 a n d 7, who s e p r o d u c t is 14, whos e r oot is 5. T h e r e f o r e
t he s uppos e d pr o duc t , who s e r oot is 9, is i nc or r ec t . T o c he c k t he
wo r k , we de t e r mi ne t he r oot of e a c h par t i al pr oduc t . T h e first
f o u r root s a r e wh a t t hey s houl d b e ; t he l ast is wr o ng . H e r e we
h a v e 2 i ns t ead of 7 ( 8 X 2 = 1 6 , who s e r oot is 7 ) . T h u s we ha v e
l oc al i z e d t he er r or .
Di g i t a l r oot s ar e t he basi s f or t he we l l - k no wn c he c k of " c as t -
i ng out ni nes . " A l o ng c o l u mn of a ddi t i on is c he c k e d by de t e r mi n-
i ng t he s um of t he root s of t he s e pa r a t e addi t i ves , a nd c o mp a r i ng
i t wi t h t he r oot of t he s uppos e d t ot al . I n s u mmi n g di gi t s f or thi s
pur pos e , 9 is s ubt r a c t e d ( " c a s t o u t " ) wh e n e v e r t he s um exc eeds
g , si nce t he a ddi t i o n or s ubt r a c t i on of 9 l eaves t he r oot of any
i nt e g e r u n c h a n g e d .
69PROPERTIES OF DIGITS
S o me of t he pr obl e ms i n thi s bo o k i nv ol v e Di o p h a n t i n e e q ua -
ti ons. S u c h a n e qua t i o n is i nde t e r mi na t e i n f o r m, but has a uni q ue
sol ut i on ( or a f i ni t e n u mb e r of sol ut i ons) t h r o u g h t he s t i pul at i on
t hat its root s a r e i nt egr al . Di s c o v e r y of t he s ol ut i on is o f t e n easi est
(i f not neces s ar y) by t ri al a n d er r or , a f t e r t he field of s e a r c h ha s
be e n sui t abl y na r r o we d. Di g i t a l root s o f t e n pr o v e a p o we r f u l t ool
t o l i mi t t he s ear ch. T h e s a me e mpi r i c a t t a c k is s ome t i me s s i mpl e r
t ha n t he a ppl i c a t i on of a pr eci s e f o r mul a . F o r e x a mp l e :
T h e s um of a n u mb e r a n d its c ub e is 1 , 458, 275, 238. Wh a t is
t he n u mb e r ?
E v i de nt l y , t he a ns we r c a n be f o u n d by e x t r a c t i ng t he c ub e
r oot of t he g i v e n i nt eger t o find t he l ar gest i nt e gr al c ub e c o nt a i ne d
i n i t ; t he r e ma i nde r wi l l be t he r e qui r e d n u mb e r . Bu t t he e x t r a c -
t i on of c ube r oot is t e di ous ; l et us l ook f o r a n a l t e r na t i v e a t t a c k.
We ar e t ol d t hat t he g i v e n i nt e g e r is of f o r m wh i c h
c a n be wr i t t e n * ( x
2
- | - i ) T h e n , l et us f a c t o r t he g i v e n i nt e ge r a n d
s egr egat e t he f ac t or s i nt o t wo g r o ups who s e pr o duc t s ar e i n r at i o
x a nd x
2
-{-i. Bu t first we wi l l l e a r n wh a t we c a n f r o m di gi t al root s.
P ossi bl e r oot o f * 1 2 3 4 5 6 7 8 9
C o ns e q u e nt r oot of x
2
1 4 9 7 7 9 4 1 9
Ro o t of 2 5 1 8 8 1 5 2 1
Ro o t of * ( x
2
- t - i ) 2 1 3 5 4 6 8 7 9
T h e r oot of 1 , 458, 275, 238 is 9. T h e a b o v e t abl e s hows us t ha t
t he r e qui r e d n u mb e r x mus t h a v e r oot 9. T h a t me a ns t ha t it is
di vi si bl e by 3 : c ons e que nt l y x
2
- j - i cannot be di vi s i bl e by 3. A l l
f act or s 3 i n t he g i v e n i nt eger wi l l c ons e que nt l y ha v e t o be as-
s i gned t o t he g r o u p t ha t ma k e u p x.
So l et us c o mme n c e f a c t o r i ng by t a ki ng out al l t he 3' s. We
find t ha t t he g i v e n i nt eger e qual s 81 X1 8, 003 , 3 9 8. H e n c c x is a
mul t i pl e of 81 .
A s t her e ar e t en di gi t s i n t he g i v e n i nt eger , x ( t he l ar ges t c ub e
r oot c ont a i ne d i n i t ) mus t be a n u mb e r of f o ur di gi t s, b e t we e n
1000 a nd 2000. T h e r e ma i ni ng f a c t o r of x, besi des 81 , mus t E e
be t we e n 13 a nd 20. T h e r e f o r e we ne e d e x a mi ne 18, 003, 398 onl y
f or l ow f act or s . T a k e o ut 2, l e a v i ng 9001 699. T h e l owes t r e ma i n-
70 PROPERTIES OF DIGITS
i n g f a c t o r i s 7. No w , we d o not h a v e t o br e a k d o wn t he r e ma i ni ng
q uo t i e nt 1 2 85 9 5 7 ; we me r e l y h a v e t o be sure t hat it c ont ai ns no
f u r t h e r l o w f ac t or s . A s t he r e a r e none , we c a n c onf i de nt l y c on-
c l u d e t ha t * = 8 i X 2 X 7 = i
I
3 4 -
112. T H E M I S S I NG DI GI T . I f t he p r o d u c t of 67 3, 1 06 a n d
4. 783, 205, 468 is
3, 21 9, 60- , 299, 743, 608
c a n y o u s uppl y t he mi s s i ng di gi t wi t ho u t a c t ua l l y mul t i pl y i ng t he
n u mb e r s ?
113. FI ND T H E S QU A RE . O n e of t he f o l l o wi ng i nt egers,
a nd onl y one , is a s quar e. C a n y o u find wh i c h i t is, wi t ho ut ac-
t ual l y e x t r a c t i ng t he s quar e r oot s ?
3, 669, 51 7 , 1 36, 205, 224
1 , 898, 732, 825, 398, 31 8
4, 751 , 006, 864, 295, 1 01
5 , 9 6 4 3 , 2 20, 1 86, 1 00
7, 538, 062, 944, 751 , 882
2,512,339,789,576,5
L 6
114. S E ND M ORE M ONE Y. Wh a t pa r e nt of a son i n col -
l ege ha s ne v e r r e c e i v e d t he f o l l o wi ng t e l e g r a m:
S E N D
M O R E
M O N E Y
T h e o d d f a c t is t ha t t hi s me s s a g e f o r ms a c or r ec t " l e t t e r a d-
di t i on. " Re p l a c e e a c h l et t er by a d i g i t t h e s a me di gi t f or t he
s a me l et t er t hr o ug ho ut , b ut di f f e r e nt di gi t s f or di f f e r e nt l e t t e r s
a n d y o u wi l l find t hat t he t wo numbe r s so f o r me d ar e c or r ec t l y
t ot al e d b e l o w t he l i ne.
By wa y of i nt r oduc t i on t o t he f o l l o wi ng puzzl es of t hi s t ype,
l et us s ol ve t hi s one t oget her .
71PROPERTIES OF DIGITS
We see a t o nc e t ha t M i n t he t ot al mus t be i , si nce t he t ot al
of t he c o l u mn S M c a nno t r e a c h as hi g h as 20. N o w if M i n t hi s
c o l umn is r e pl a c e d by 1, h o w c a n we ma k e t hi s c o l u mn t ot al as
mu c h as 10 t o pr ov i de t he 1 c a r r i e d ov e r t o t he l e f t b e l o w? O n l y
by ma k i n g S v e r y l a r g e : 9 or 8. I n ei t her case t he l et t er O mu s t
s t and f or z e r o: t he s u mma t i o n of S M c oul d p r o d u c e onl y 10 o r
1 1 , but we c a nno t use 1 f or l et t er O as we h a v e a l r e a dy us ed it f o r
M .
I f l et t er O is zer o, t he n i n c o l u mn E O we c a nno t r e a c h a
t ot al as hi g h as 10, so t hat t he r e wi l l be no 1 t o c a r r y ov e r f r o m
t hi s c o l u mn t o S M . H e n c e S mus t posi t i vel y be 9.
Si nc e t he s u mma t i o n E O gi ves N, a nd l et t er O is zer o, N
mus t be 1 gr e at e r t ha n E a nd t he c o l u mn N R mus t t ot al o v e r 1 o.
T o put i t i nt o a n e q u a t i o n:
E + i = N
F r o m t he N R c o l u mn we c a n de r i v e t he e q u a t i o n:
N + R + ( + i ) = E + i o
We ha v e t o i nsert t he expr es s i on ( + 0 be c a us e we don' t k n o w yet
whe t he r 1 is c ar r i e d ov e r f r o m c o l u mn D E . Bu t we d o k n o w t ha t
1 has t o be c ar r i e d ov e r f r o m c o l u mn N R t o E O .
Subt r a c t t he first e qua t i o n f r o m t he s e c o nd:
R+( +0=9
We c a nno t l et R e qua l 9, si nce we a l r e a dy h a v e S e q ua l t o 9.
T h e r e f o r e we wi l l h a v e t o ma k e R e qua l t o 8; he nc e we k n o w t ha t
1 has t o be c ar r i e d ov e r f r o m c o l u mn D E .
C o l u mn D E mus t t ot al at l east 12, si nce Y c a nno t be 1 or
zer o. Wh a t v a l ue s c a n we gi v e D a nd E t o r e a c h t hi s t o t a l ? We
h a v e a l r e a dy us ed 9 a nd 8 e l s e whe r e . T h e onl y di gi t s l e f t t ha t ar e
hi g h e n o u g h ar e 7, 6 a nd 7, 5. Bu t r e me mb e r t ha t one of t hes e has
t o be E , a nd N is 1 gr e at e r t ha n E . H e n c e E mus t be 5, N mus t be
6, whi l e D is 7. T h e n Y t ur ns o ut t o be 2, a n d t he puz z l e is c o m-
pl et el y s ol ved.
115. S P E L L I NG A DDI T I ON. A n y b o d y c a n see t ha t t he
s ubj oi ne d s um is c or r ec t , but t o pr o v e it is a no t he r ma t t e r . S u p -
pose t hat y o u do so by r e pl a c i ng e a c h l et t er by a d i g i t t h e s a me
72 PROPERTIES OF DIGITS
di gi t f or t he s ame l et t er t hr o ug ho ut . O f cour s e, di f f e r e nt l et t ers
mus t be r e pl a c e d by di f f e r e nt di gi t s.
O N E
T W O
F O U R
S E V E N
A n y one of s ever al sol ut i ons wo u l d s er ve t o pr o v e t he s um,
b u t f or t he t r ui s m t hat if y o u r W O E S a r e mul t i pl i e d ( es peci al l y
fivefold) y o u ar e b o u n d t o ge t S O R E R .
116. A DA M A ND E VE . L i k e l y as not , if y o u or de r p o a c h e d
eggs on t oast i n a s hor t - or der l unc hr o o m, t he c o u nt e r ma n wi l l
s hout t o t he c ook
A D A M
A N D
E V E
O N
A
R A F T
Wh a t t he c o u n t e r ma n pr o ba bl y does not k n o w is t hat thi s
phr as e is r eal l y a n a nc i e nt c a b a l a i nv e nt e d by t he Nu me r i a n as-
t r ol oger s, de r i v e d f r o m a s um i n a ddi t i o n by r e pl a c i ng e a c h di gi t
by t he s a me l et t er t hr o ug ho ut . Wo r k i n g ba c k wa r ds , y o u c a n find
s ever al e x a mpl e s of a ddi t i on t hat wi l l gi v e thi s s ame resul t , b ut
t her e is n o d o u b t t hat t he Nu me r i a n s wi s he d t o ma k e t he r a f t as
c o mmo d i o u s as possi bl e, so t hat t he e x a mp l e t hey ha d i n mi nd
gi ves t he l ar ges t possi bl e t ot al .
117. RE S T ORI NG T H E FI GU RE S . Wh e n M i s s Ga t e s re-
t ur ne d t o t he c l as s r oom a f t e r recess, she f o u n d t he j ani t or j us t
b e g i nni ng t o wa s h t he bl a c kboa r ds .
" O h , M r . Be ns o n, " s he cr i ed, " d o n' t er as e t hat mul t i pl i c a-
t i o n! " Bu t t he j a ni t or h a d a l r e a dy wi p e d hi s we t c l ot h ov e r a
g o o d s har e of t he figures.
73PROPERTIES OF DIGITS
i t . "
" We l l , ne v e r mi nd, i t ma k e s a pr et t y pr o b l e m thi s wa y . I ' l l
ask t he pupi l s t o rest ore t he figures."
If s e ve nt h- gr ade pupi l s c a n r est or e t he figuresas t hey d i d
so c a n you.
118. L E T T E R DI VI S I ON. A v e r y p o p u l a r k i nd of puz z l e,
seen ma n y t i mes i n ma g a z i ne s , is t he " l e t t e r di vi s i on. " T h e wo r k -
i ng sheet of a l ong di vi s i on is pr es ent ed, wi t h e a c h di gi t r e pl a c e d
by a l et t er. ( T h e s a me di gi t is r e pl a c e d by t he s a me l et t er whe r -
e ve r i t occur s . ) T h e puz z l e is t o " d e c r y p t " t he l et t ers by r es t or i ng
t he or i gi nal di gi t s.
T h e l et t er di vi s i on is a n e xc e l l e nt exer ci s e i n t he s i mpl es t
pr oper t i es of di gi t s. Some t i me s , unf or t una t e l y , t hi s v a l ue is nul l i -
74 PROPERTIES OF DIGITS
f i ed by t he ma n n e r of pr es ent at i on. T h e subst i t ut e l et t ers ar e c ho-
sen by wr i t i ng a wo r d or phr as e ( f r e e of r e pe a t e d l et t ers) ov e r
t he di gi t s a r r a ng e d i 2 3 4 5 6 7 8 9 0 . T h e puz z l e c a n t hen be
s ol ve d by a n a g r a mmi n g i ns t ead of by ar i t hme t i c , a nd thi s i nde e d
is t he wa y ma n y per sons c hoos e t o a t t a c k t he pr obl e m.
H e r e is a l et t er di vi s i on t hat c a nno t be sol ved by a na g r a m-
mi ng .
A B C ) D C E F G A ( F H G
D G H F
A F J G
A E C K
D A H A
D D H H
B ~ G
119. C RYP T I C DI VI S I ON. I n a l et t er di vi si on, t he oc c ur -
r e nc e of t he s a me l et t er i n s ever al pl ac es , s ho wi ng r epet i t i ons of
t he s a me di gi t , hel ps t o l i mi t its possi bl e val ues . A n d t he pr ocess
of e l i mi na t i on hel ps , t oo, e. g. , if A equal s 7, t he n no ot he r l et t er
c a n e q ua l 7.
* * ^* g* * * ^* * *
* *
* * *
* * *
2
* *
* * *
A p p a r e n t l y mo r e di f f i c ul t is t he " c r y pt i c di vi s i on, " i n wh i c h
al l b ut a f e w of t he di gi t s ar e s uppr es s ed, a nd t he puz z l e is t o
r est or e t he mi s s i ng di gi t s wi t h no k n o wl e d g e of h o w ma n y di f f er -
e nt di gi t s ar e us ed or wh e r e r epet i t i ons oc c ur . Bu t t he di f f i c ul t y i n
75PROPERTIES OF DIGITS
f a c t var i es wi t h t he pa r t i c ul a r e x a mp l e . H e r e is a c r y pt i c di vi s i on
wh i c h t he s ol ver wi l l f i nd mu c h easi er t ha n t he l et t er di vi s i on
g i v e n pr evi ous l y.
Re p l a c e e a c h st ar b y a n y di gi t , so t ha t t he wh o l e wi l l be a
c or r e c t l o ng di vi s i on.
120. C RYP T I C M U L T I P L I C A T I ON. S u p p l y t he mi s s i ng
di gi t s i n t hi s mul t i pl i c a t i on p r o b l e m
* * y *
* *
* * * * *
* * *
2
*
8 * 5 *
* * * * * *
121. C RYP T I C S QU A RE ROOT . I t is as t oni s hi ng b u t t r ue
t hat e ve r y mi s s i ng di gi t i n t hi s e x a mp l e of e x t r a c t i on of s quar e
r oot c a n be c or r ec t l y i nf e r r e d, wi t h t he ai d of onl y one g i v e n 3.
I f y o u h a v e f or g ot t e n h o w t o e x t r a c t s quar e r oot , see t he
A p p e n d i x . T a k e not e t hat wh a t we cal l t he memorandum column
i n t he e x a mpl e s t her e g i v e n i s o mi t t e d f r o m t he c r y pt i c pr o bl e m.
* * * *
* * * * * * *
*
* * *
* *
* * * *
* * * *
* * * * *
* * * g *
76 PROPERTIES OF DIGITS
122. T H E C OM P L E T E GH OS T . I t is pr oba bl y i mpossi bl e
t o c ons t r uc t a c r ypt i c di vi s i on s h o wi ng not a si ngl e di gi t , a nd yet
pr o v i de a uni que s ol ut i on. T h e r e c or d f or be i ng t he near es t t o a
c o mp l e t e " g h o s t " is hel d by t he f o l l o wi ng puz z l e, wh i c h pr esent s
t wo r el at ed di vi si ons i n l i eu of a ny di gi t .
Re p l a c e t he stars by numbe r s so as t o ma k e t wo c or r ec t ex-
a mpl e s of l o ng di vi si ons. I t is s t i pul at ed t hat t he si x- di gi t quot i e nt
of t he first e x a mp l e mus t be t he s a me as t he di v i de nd of t he
s ec ond.
* * * ^ * * * * * * * * * ^ * * * * * *
* * *
* * * *
* * *
* * *
* * *
* * * *
* * * *
* *
* * *
* *
* * *
* * *
* * *
* * *
123. A GH OS T A DDI T I ON. H e r e is a puz z l e t hat y ou
c oul d s ol ve ent i r el y b y t ri al a nd er r or , but t he t ri ck is t o use y our
i ng e nui t y so as t o na r r o w t he field of s ear ch.
77PROPERTIES OF DIGITS
* *
* *
* *
* *
* *
* * *
T h e p r o b l e m is t o r e pl a c e t he st ars i n t hi s g hos t a ddi t i o n by
di gi t s, so t h a t :
( a ) T h e s um is a t r i a ng ul a r n u mb e r not di vi s i bl e b y 3. ( F o r
e x pl a na t i o n of t r i a ng ul a r numbe r s , see No . 1 3 6 F i g u r a t e Num-
bers.)
( b) A b o v e t he l i ne, e a c h c o l u mn s hows f i ve di gi t s i n
s e que nc e , p i c k e d out of t he endl es s c ha i n 1, 2, 3, 4, 5, 6, 7, 8, 9,
o, 1 . . .
124. NU M BE RS A ND T H E I R DI GI T S . I n a l g e br a , wh e n
we set quant i t i e s si de b y s i de wi t h n o i nt e r v e ni ng si gn, as abc, we
me a n t h a t t he y ar e t o b e mul t i pl i e d. Bu t w h e n we set di gi t s s i de
b y si de, as 736, we d o no t me a n t he p r o d u c t 7 X 3 X 6 . H e r e we
expr es s a c a r di na l n u mb e r 700-4-304-6. I f w e ar e t ol d t ha t t he
di gi t s of a n u mb e r ar e, f r o m l e f t t o r i g ht , a,b,c, we mu s t t he r e -
f or e expr es s t he n u mb e r i tsel f b y 1 ooa- f - 1 ofc-f-c.
A ll of the problems under Two-Digit Numbers and Three-
Digit Numbers c a n b e s ol v e d b y s et t i ng u p e qua t i ons of t hi s t y pe .
125. T WO-DI GI T NU M BE RS .
( a ) Wh a t n u mb e r is t wi c e t he p r o d u c t of i ts di gi t s ?
( b ) Wh a t n u mb e r is t hr e e t i me s t he s u m of its di gi t s ?
( c ) Wh a t n u mb e r is t he s qua r e of its uni t s di g i t ?
( d ) Wh a t n u mb e r e x c e e ds i ts r ever s al by 2 0 % ?
( e ) Wh a t nu mb e r s pl us t hei r r ever s al s s u m t o p e r f e c t
s quar e s ?
78 PROPERTIES OF DIGITS
126. T H RE E -DI GI T NU M BE RS .
( a ) Wh a t n u mb e r is 11 t i mes t he s u m of its di gi t s?
( b) H o w ma n y numb e r s ar e t wi c e t he numbe r s f o r me d by
r e v e r s i ng t he or de r of t hei r di gi t s ?
( c ) T h e s um of al l t hr e e di gi t s of a n u mb e r is i dent i c al wi t h
t he first t wo di gi t s, a n d t he s u m of t he s um is i dent i c al wi t h t he
t hi r d. Wh a t is t he n u mb e r ?
( d) Wh a t n u mb e r is t he s um of 17 t i mes its first di gi t , 34
t i mes i ts s e c ond di gi t , a nd 51 t i mes its t hi r d di g i t ?
( e) Wh a t n u mb e r is one - f i f t h of t he s um of al l ot he r nu m-
ber s expr es s ed b y pe r mut a t i o ns of t he s a me t hr ee di gi t s ?
( f ) Wh a t numbe r s a r e t he s u m of al l possi bl e pe r mut a t i ons
of t he t hr e e di gi t s t a k e n t wo a t a t i me ?
VI I I . P uzzles A bout I ntegers
127. P RI M E NU M BE RS . A n u mb e r wh i c h is not e x a c t l y
di vi si bl e by a ny ot he r i nt eger ( e x c e pt i ) is c a l l e d prime. A n u m-
be r wh i c h c a n be e v e nl y di v i de d by a no t h e r is composite.
F o r ma n y pur pos es i n ma t he ma t i c s , i t is neces s ar y t o be a bl e
t o de t e r mi ne wh e t h e r a g i v e n i nt eger is pr i me , a n d , if i t is c o m-
posi t e, t o de t e r mi ne al l of its pr i me f ac t or s .
I n e l e me nt a r y a r i t hme t i c we l e a r n a f e w bas i c tests f o r
f a c t or i z a t i on: E v e r y e v e n n u mb e r is di vi si bl e by 2; e v e r y n u m-
be r whos e t e r mi nal di gi t is 5 or o is di vi s i bl e by 5 ; e v e r y n u mb e r
whos e di gi t al r oot is 3, 6, or 9 is di vi si bl e by 3 ; a n u mb e r is
di vi si bl e by 11 if t he s u m of its 1st, 3r d, et c. , di gi t s is e q ua l t o t he
s um of t he 2nd, 4t h, et c. , or if t he t wo s ums di f f e r by a ny mul t i pl e
of 1 1 . T he r e ar e l e w s i mp l e tests l i ke t hi s f o r di v i s i bi l i t y by h i g h e r
pr i mes . T h e t ask of br e a k i ng d o wn c e r t a i n v e r y l a r g e c o mpo s i t e
numbe r s has e ng a g e d t he a t t e nt i on of g e ne r a t i ons of ma t h e -
ma t i c i a ns a n d has l ed t o ma n y di s cover i es i n t he or y of numbe r s .
I t is possi bl e t o s ol ve ma n y pr obl e ms of f a c t or i z a t i on by
e mpi r i c a l me t hods . T h e t e r mi na l di gi t a nd t he di gi t al r oot of a n
i nt eger bo t h pl a c e a l i mi t a t i on o n its possi bl e f a c t o r s ; use t hi s
k no wl e d g e t o na r r o w t he f i el d of s e ar c h a nd t he n find t he a c t ua l
f ac t or s by t ri al di vi si ons.
I n t he A p p e n d i x is g i v e n a t abl e of al l p r i me numb e r s be-
t we e n 1 a n d 1, 000. A s a s i mpl e exer ci s e I wi l l ask y o u t o de t e r mi ne
t he ne x t t hr ee pr i me numb e r s hi g he r t ha n 997.
79
80 PUZZLES ABOUT INTEGERS
128. T H E S A L E ON S H I RT S . " I ma d e a s mar t mo v e
ma r k i n g d o wn t hose shi rts f r o m $2. 00, " r e ma r k e d M r . Ga be r di ne
t o hi s wi f e . " We ha v e di sposed of t he ent i r e l ot . "
" G o o d ! " sai d M r s . Ga b e r d i ne . " H o w mu c h pr of i t di d y ou
ma k e ? "
" We ha v e n' t figured it yet , but t he gross f r o m t he sal e wa s
$603. 7 7 . "
" We l l , h o w ma n y shi rts di d y o u s el l ?"
L e t t he r e a de r a ns we r t he ques t i on.
129. A P OWE R P ROBL E M . T h e i nt eger 844, 596, 301 is
t he 5t h p o we r of wh a t nu mb e r ?
130. T H E ODD FE L L OWS P A RA DE . T h e Gr a n d M a r s h a l
r ode d o wn t he st reet wh e r e t he O d d Fe l l ows we r e gat he r e d, i n-
s t r uc t i ng t he m, " F o r m r anks of t hr ee a br e a s t ! "
A f t e r al l h a d c o mpl i e d, it wa s f o u nd t hat t her e wa s one ma n
l e f t ov e r . A s t he M a r s h a l di d not wa n t t hi s one ma n t o ha v e t o
ma r c h a l one , t he M a r s h a l c h a ng e d hi s o r d e r " F o r m r anks five
a b r e a s t ! " Bu t a f t e r thi s r e a r r a ng e me nt wa s c o mpl e t e d, 2 me n
we r e l e f t ov e r .
S o t he M a r s h a l t r i ed agai n. " A l l r i ght me n, l et ' s ma k e it
s even abr eas t . R e f o r m r a nk s ! " T h i s ma n e u v e r h a d n o mo r e suc-
cess t ha n t he ot her s , 3 me n b e i ng l e f t over .
" T h a t ' s wh a t c o me s of l os i ng so ma n y me n t o t he s e r v i c e ! "
mu t t e r e d t he M a r s h a l . " L a s t y e a r we h a d 497 me n i n l i ne, so
t he y ma d e e v e n sevens. We l l , I ' d be t t e r t ry o nc e mo r e . . . boys,
we ' l l t r y i t n o w e l e v e n a br e a s t . "
Bu t wi t h r anks of e l e ve n, it wa s f o u n d t hat t her e we r e 4 me n
l e f t over . T h e M a r s h a l de c i de d t o was t e n o mo r e t i me t r yi ng t o
t r i m t he p a r a d e i nt o e v e n r anks , b ut pl a c e d t he 4 e x t r a me n at
t he he a d of t he l i ne a nd g a v e t he o r de r t o ma r c h .
T h e que s t i on i s: H o w ma n y O d d Fe l l ows ma r c h e d i n t he
p a r a d e ?
81PUZZLES ABOUT INTEGERS
131. T H E U NI T Y C L U B. I t h a p p e ns t hat t he U n i t y C l u b
ma r c h e d i n t he s a me p a r a de wi t h t he O d d Fe l l ows . I n i mi t a t i on
of t he l at t er, t hey t ri ed f o r mi n g r anks of 3, 5, 7, a nd 1 1 , b ut t hey
h a d no bet t er l uc k t ha n t he O d d Fe l l ows i n c o mi n g o u t e v e n. I n
r anks of 3 abr eas t , t hey h a d 2 me n l e f t o v e r ; i n 5s, 4 e x t r a ; i n
7s, 6 e x t r a ; a n d i n u s , 10 e xt r a. Wh a t is t he l east n u mb e r of
ma r c he r s t her e mus t ha v e be e n i n t he U n i t y C l u b ?
132. C I NDE RE L L A T OA S T E RS . T h e pr es i dent of a c h a i n
of ret ai l stores de a l i ng i n el ect r i cal a ppl i a nc e s one d a y r e que s t e d
t he vi ce- pr es i dent i n c ha r g e of sal es t o p r o d u c e t he figures o n t he
sal e of C i nde r e l l a T oa s t e r s , one of t he mos t p o p u l a r i t ems sol d
by t he c o mp a ny . T h e vi c e - pr e s i de nt g a v e h i m a t onc e a pr e -
l i mi nar y me mo r a n d u m, as f ol l ows :
" H e r e , " sai d t he pr es i dent t o hi s s ecr et ar y. " Di v i d e t hat out
f or me a nd l et me k n o w h o w ma n y t oast ers we r e sol d i n t he M a i n
St r eet b r a n c h . "
" Wh a t do t he t oast ers sell f o r ? " as ked t he s ecr et ar y.
" Y o u ' l l h a v e t o l ook t hat u p . "
Bu t t he secr et ar y f o u nd t he a ns we r wi t ho u t l ooki ng u p t he
pr i c e pe r art i cl e.
133. S QU A DS A ND C OM P A NI E S . T h e ent i r e s t a ndi ng
a r my of Nu me r i a c ompr i s es 1, 547 c o mpa ni e s of e qua l si ze. I t
c oul d al so be g r o upe d i nt o 34, 697 e v e n squads. Wh a t is t he l east
nu mb e r of me n of wh i c h t he a r my c a n be c o mp o s e d ?
134. T H E M I S RE A D C H E C K. I t wa s a s t r ange l aps e o n
t he pa r t of t he b a nk t el l er. E v i de nt l y he mi s r e ad t he c he c k, f or
he h a n d e d out t he a mo u n t of t he dol l ar s i n cent s, a nd t he a mo u n t
of t he cent s i n dol l ar s. Wh e n t he er r or wa s poi nt e d out t o h i m he
b e c a me f l us t er ed, ma d e a n abs ur d a r i t hme t i c a l mi s t ake , a n d
GROSS SALES, CINDERELLA TOASTERS
M a i n St r e e t b r a nc h
A l l ot he r br a nc he s
$3,893-93
8 , 3 1 1 . 1 9
82
PUZZLES ABOUT INTEGERS
h a n d e d o ut a dol l ar , a di me , a n d a c e nt mor e . Bu t t he depos i t or
de c l a r e d t hat he wa s still shor t of hi s due . T h e t el l er pul l e d hi m-
self t oge t he r , do ub l e d t he a mo u n t he ha d a l r e a dy g i v e n t he
depos i t or , a n d so set t l ed t he t r ans ac t i on t o t he l at t er' s sat i sf ac-
t i on.
Wh a t wa s t he a mo u n t c a l l e d f o r by t he c h e c k ?
135. T RA NS FE RRI NG DI GI T S . T h e f o l l o wi ng puz z l e, due
t o Du d e n e y , is g i v e n be c a us e t he me t h o d of s ol ut i on is us ef ul f or
a wh o l e cl ass of di gi t al pr obl e ms .
I f we mul t i pl y 5 7 1 , 42 8 by 5 a n d t he n di v i de by 4, we ge t
7 1 4, 285, wh i c h is t he s a me as t he or i gi nal n u mb e r wi t h t he first
di gi t t r ans f er r ed t o t he e nd.
C a n y o u find a n u mb e r t hat c a n be mul t i pl i e d by 4 a n d
di v i de d by 5 i n t he s a me w a y b y t r ans f e r r i ng t he first di gi t t o
t he e n d ?
O f c our s e 7 1 4, 285 wo u l d ser ve if we we r e a l l o we d t o t r ans f er
t he l ast di gi t t o t he he a d. Bu t t he t r ans f er mus t be t he ot he r wa y
f r o m b e g i nni ng t o e nd.
136. FI G U RA T E NU M BE RS . T h e a c c o mp a n y i n g d i a g r a m
s hows " P a s c a l ' s t r i a ng l e , " wh i c h is a n or der l y wa y of wr i t i ng out
c e r t a i n cl asses of i nt eger s c al l ed figurate numbers.
T h e t o p r o w a n d l e f t mos t c o l u mn of thi s ar r ay consi st en-
t i rel y of I'S. T h e n t he t abl e is bui l t u p by wr i t i ng i n e a c h cel l t he
s u m of t he t wo numbe r s i n t he cel l s a t its l ef t a n d a bo v e i t.
I n t he s e c ond r o w ( a nd c o l u mn ) a p p e a r al l t he i nt eger s i n
a s c e ndi ng ma g n i t u d e ; t he i nt eger s a r e c ons e que nt l y cl assed as t he
second order of figurate numbers.
T h e numb e r s ar e c al l ed figurate be c a us e ear l y ma t he -
ma t i c i a ns pe r c e i v e d t hat t hey g i v e t he ar eas a n d v o l ume s of cer -
t ai n g e ome t r i c a l figures wh e n bui l t u p by di scr et e uni t s. Fo r
e x a mp l e , if y o u pl a c e a n u mb e r of c a n n o n bal l s i n t he f o r m of
a n e qui l a t e r a l t r i angl e, wi t h n bal l s i n t he bas e, t he n t her e wi l l
be n 1 i n t he r o w a bo v e i t, n 2 i n t he r o w a bov e t hat , a nd so
83PUZZLES ABOUT INTEGERS
on unt i l y ou r e a c h t he a pe x of one bal l . T h e t ot al numb e r of bal l s
i n t he t ri angl e wi l l be t he sum of t he i nt egers f r o m i up t o n. A l l
s uch number s ar e f ound i n t he t hi rd r ow of t he di a g r a m, a n d thi s
row is consequently called triangular numbers.
Si mi l ar l y, if you pi l e c a nnon bal l s i nt o a py r a mi d wi t h a base
t hat is a n equi l at er al t ri angl e, t he t ot al numb e r of bal l s wi l l be
some numbe r i n t he f our t h r ow, whi c h is t her ef or e cal l ed tri-
angular pyramids.
T h e number s t hat l i e a l ong a di agonal l i ne f r o m l owe r l ef t
to uppe r ri ght , wh i c h l i ne is cal l ed a base, ar e t he coef f i ci ent s
of t he t erms i n a bi nomi al expansi on. Fo r e x a mp l e :
( a + b )
4
= a
4
+ 4 a
3
b + 6 a - b
2
+ 4 a b * + b \ T h e coef f i ci ent s 1 4 6
4 1 are seen t o lie on t he base t hat starts at t he fifth r ow. P ascal ' s
t ri angl e shows gr aphi c al l y h o w t he bi nomi al t heor em is der i ved,
84 PUZZLES ABOUT INTEGERS
a n d its r el at i ons hi p t o t he t he or y of pe r mut a t i ons a nd c o mbi na -
tions.
M a n y pr a c t i c a l pr obl e ms i nv ol v e f i ndi ng t he nt h t er m of t he
r t h o r de r of f i gur at e numbe r s . I t is g i v e n by t he f o r mu l a
n(n-)-i) (
n
+
2
) (
n
+
r

2
)
F=0!
I n wor ds , t he nt h t e r m of t he r t h o r de r is g i v e n by t he
p r o d u c t of s ucces s i ve i nt eger s f r o m n t o n-\-r2 i ncl us i ve, di v i de d
by f a c t or i a l ( r 1 ) .
We see t ha t P as c al ' s t r i angl e is s y mme t r i c a l wi t h r es pect t o
t he di a g o na l c o mme n c i n g at t he u p p e r l e f t c o r ne r ; c ons equent l y
t he nt h t e r m of t he r t h or de r is t he s ame as t he r t h t e r m of t he
nt h or de r . I t does not ma t t e r wh e t h e r we c o unt t he or der s by r ows
f r o m t op d o wn or by c o l umns f r o m l e f t t o r i ght .
F o r t he first f o ur or der s, t he f o r mu l a f or t he nt h t e r m re-
duc e s t o
1st o r de r : 1
2nd o r de r : n
3r d o r de r : n ( n - f - i )
2
4t h o r d e r : n(n-\-i) ( n- f - 2)
6
T h e r e a de r s houl d not e es pec i al l y t he f o r mul a f or a tri -
a n g u l a r nu mb e r , as I gi v e a n u mb e r of puzzl es i nv ol v i ng tri-
angl es .
Jus t t o st art t he bal l r ol l i ng, he r e is a n easy que s t i on: Wh a t
is t he s u m of t he first 25 t r i a ng ul a r py r a mi ds ? Y o u r sol ut i on
doe s n' t c o u nt i f y o u wr i t e o ut t he numbe r s a nd a d d t he m!
137. L I T T L E WI L BU R A ND T H E M A RBL E S . L i t t l e Wi l b u r
ha s a pas s i on f o r ma r bl e s , a nd o n hi s l ast bi r t hda y he h a d si gnal
success i n o bt a i ni ng " mi g g l e s " gal or e f r o m hi s par e nt s a nd unc l es
a n d aunt s .
85PUZZLES ABOUT INTEGERS
On e da y hi s mo t h e r f o u n d h i m s eat ed on t he floor l a y i ng o ut
hi s mar bl es i n c ol or f ul g e ome t r i c figures.
" Wh a t ar e y o u doi ng, Wi l b u r ? " she i nqui r e d.
" I ' m pl a y i ng C hi ne s e C he c k e r s , l i ke y o u d o , " wa s t he re-
sponse. I nde e d, Wi l b u r h a d a r r a ng e d hi s ent i r e st ock of ma r bl e s
i nt o t r i angl es of e qua l si ze.
" We l l , " l a ug he d hi s mo t he r . " Y o u c e r t ai nl y wo u l d ne e d a l ot
of pl ayer s f or that g a me ! "
T h e n she l ef t , but pr es ent l y Wi l b u r c al l e d he r ba c k.
" Se e , M u mmy , n o w it does n' t t ake so ma n y pl a y e r s ! " Sur e
e noug h, Wi l b u r h a d r e a r r a ng e d t he ma r bl e s i nt o a s mal l e r n u m-
be r of l ar ger t r i angl es, al l e qua l .
" T h a t ' s v e r y c l ever , Wi l b u r . I t mus t h a v e t a ke n a g r e a t deal
of pa t i e nc e t o ma k e t he numbe r s c o me out r i g ht . "
" O h , no, M u mmy , i t' s easy. L o o k ! " A n d be f or e he r ast on-
i shed eyes t he pr e c oc i ous c hi l d pr o c e e de d t o r e a r r a ng e t he ma r bl e s
f o ur t i mes mor e , e a c h t i me ma k i n g a f e we r n u mb e r of t r i angl es
al l of t he s a me si ze.
" I de c l a r e , " she sai d, " I ne v e r s a w a ny t hi ng l i ke i t ! I t mus t
c o me f r o m y o ur f at her ' s si de, be c a us e I wa s ne v e r abl e t o d o t he
mul t i pl i c at i on t abl e . "
T o c a p t he c l i ma x , L i t t l e Wi l b u r put al l t he ma r bl e s t o-
ge t he r a nd ma d e a si ngl e t r i angl e.
Wh a t is t he l east n u mb e r of ma r bl e s L i t t l e Wi l b u r c o ul d
ha v e h a d ?
138. H OKU M . BU NKU M A ND FA T U I T U M . On one of the
l esser sat el l i t es of U r a n u s , we ar e t ol d by j uv e ni l e l i t er at ur e, is
f o u n d a mys t er i ous me t a l , l i ght er t ha n a l u mi n u m but s t r onger
t ha n st eel , wi t h ma n y us ef ul pr oper t i es s uc h as t he p o we r of i nt er -
c e pt i ng gr avi t at i on. A f t e r i nt ens i ve s t udy of r epor t s b r o u g h t b a c k
by i nt e r pl ane t ar y t r avel er s , I h a v e be e n abl e t o de t e r mi ne t he
a t omi c s t r uc t ur e of t hi s f a s c i na t i ng e l e me nt , wh i c h is c a l l e d
H o k u m. T h e a t o mi c mo de l is s ho wn o n p a g e 86. I t is a t r i a ng ul a r
h e x a h e d r o n a sol i d h a v i n g si x f aces , e a c h of wh i c h is a n e qui -
l at er al t r i angl e.
86 PUZZLES ABOUT INTEGERS
Wh i l e not p r e p a r e d t o g i v e a c o mpl e t e i nt er pr et at i on of t hi s
a t o mi c mo d e l , I c a n s ay t ha t e a c h c o mp o n e n t r epr esent s t he
s phe r e of i nf l ue nc e of a hi g hl y be l l i ge r e nt el ec t r on.
T h e pi c t ur e s hows a n a t o m of p o we r 5 t h a t is, e a c h e dge
c ont a i ns five el ect r ons . Bu t t he a t oms of H o k u m a r e a c t ua l l y of
mu c h hi g he r p o we r .
M y r es ear ches e na bl e me t o pr e di c t t hat t wo n e w as t oni s hi ng
e l e me nt s wi l l be di s c ov e r e d i n t he sol ar syst em, i f i nde e d t hey ar e
not a l r e a dy k n o wn . O n e e l e me nt is Bu n k u m, wi t h a n a t o mi c
p o we r wh i c h is one g r e a t e r t ha n t hat of H o k u m. T h e ot he r is
F a t u i t u m, who s e p o we r e x c e e ds B u n k u m by one.
T h e r e is g o o d r e as on t o be l i e ve t ha t t he mo l e c ul e c o m-
p o u n d e d of one a t o m e a c h of H o k u m, Bu n k u m, a nd F a t u i t u m
wo u l d h a v e pr ope r t i e s al mos t b e y o nd our e x pe r i e nc e , s uc h as t he
abi l i t y t o s quar e t he c i r c l e wi t h a r e d pe nc i l a n d a pi ec e of st ri ng.
S u c h a mo l e c ul e wo u l d h a v e no less t ha n 3 1 , 3 1 1 el ect r ons , a nd
na t ur a l l y wo u l d be he l d t og e t he r onl y by bonds of t he mos t
al t r ui s t i c na t ur e .
I h a v e b e e n as ke d t o r eveal t he a t o mi c p o we r of H o k u m,
B u n k u m, a n d F a t u i t u m, but wi l l onl y r epl y t hat t he a ns we r is
obv i ous .
87PUZZLES ABOUT INTEGERS
139. S QU A RE NU M BE RS . A n u mb e r wh i c h is t he p r o d u c t
of t wo e qua l f ac t or s , as 49 ( 7 X 7 ) is c al l e d a s quar e . P uz z l es i n-
v o l v i ng s quar es a r e p o p u l a r wi t h puz z l e ma ke r s a nd sol vers al i ke.
T h e r eas on is pe r ha ps t hat t he e q ua t i o n A
n
- f - B " = C " has a n i n-
f i ni t y of i nt egr al sol ut i ons wh e n n=2. I t is be l i e v e d t hat no i n-
t egr al sol ut i ons exi st if n is g r e a t e r t ha n 2.
T h e l owes t i nt eger s t hat sat i sf y t he e q ua t i o n ar e 3
2
+ 4
2
= 5
2
.
T h e 3 : 4 : 5 r i ght t r i angl e has pl a y e d a n i mpo r t a nt r ol e i n hi st or y.
I t wa s used by t he a nc i e nt E g y p t i a n " r o pe - be a r e r s " t o l ay out
r i ght angl es a nd t o f o u nd t he g e o g r a phi c a l a nd g e o me t r i c sci ences.
Di s c ov e r y of thi s t r i angl e is sai d t o h a v e l e d t he E g y pt i a ns t o t he
di s c over y of t he v e r y i mpo r t a nt " P y t h a g o r e a n t h e o r e m" t h e s um
of t he s quar es o n t he t wo l egs of a r i ght t r i angl e is e qua l t o t he
s quar e o n t he hy pot e nus e .
I t wi l l be not i c e d t ha t t he t abl e of f i gur at e numbe r s ( " P a s -
cal ' s t r i a ng l e " ) does not i nc l ude t he s quar e numbe r s . H o w c a n
t he s quar es be de r i v e d f r o m thi s t abl e ?
140. S QU A RE -T RI A NGU L A R I NT E GE RS . T h e i nt eger 36
equal s 6
2
a nd i t is al so t he s um of t he i nt eger s f r o m 1 t o 8 i n-
c l us i ve; t hus 36 is bo t h s quar e a n d t r i a ngul a r . I t is t he l owes t s uc h
i nt eger ( a b o v e u n i t y ) . T h e ne x t l owes t is 1 225. Wh a t a r e t he t wo
ne x t l owes t i nt eger s t hat ar e b o t h s quar e a nd t r i a ng ul a r ?
141. P A RT I T I ON OF A T RI A NGL E . P r ov e t hat any t ri -
a ng ul a r n u mb e r s uf f i ci ent l y l ar ge is t he s um of a s quar e n u mb e r
a nd t wo e qua l t r i a ng ul a r numbe r s .
142. T H E BA T T L E OF H A S T I NGS . T h e r e is s ome ques -
t i on wh e t h e r t he hi t he r t o- a c c e pt e d des c r i pt i on of t he Bat t l e of
H as t i ngs , g i v e n by t he Ba y e u x T a p e s t r y , is c or r ec t , i n v i e w of t he
s o me wha t di f f e r e nt ver s i on g i v e n by t he H o o k e d R u g r ec ent l y
di s c over ed i n a n at t i c at C a mb r i c - b y - t h e - Ya r d . Wi t h o u t e nt e r i ng
i nt o t hi s c ont r over s y, we ma y not e a n i nt er es t i ng c i r c ums t a nc e
a bo ut t he l at t er ver s i on.
88 PUZZLES ABOUT INTEGERS
T h e H o o k e d R u g st at es t hat t he Sa x ons f o r me d t hei r hous e
car l s a n d kni ght s i nt o a sol i d s quar e p ha l a nx . T h e No r ma n s ad-
v a n c e d f r o m t he s hor e al s o i n a sol i d s quar e, b ut we r e possessed
of " ha l f a t hous a nd mo r e f o o t me n n e a nd f ul l douz a i ne mo r e of
kni g ht s . " De s pi t e t hese g r e a t odds , t he Sa x ons f o ug ht so v al i ant l y
t hat t hey sl ew hal f t he f oe , l os i ng " o nl y a f e w s c or e " of t hei r o wn
me n , a n d t hus r e d uc e d t he t wo ar mi e s t o e x a c t l y t he s a me nu m-
bers.
I t has b e e n g e ne r a l l y ov e r l ooke d t hat f r o m thi s mu c h of t he
a c c o u nt we c a n c a l c ul a t e e x a c t l y t he size of t he t wo ar mi es as t hey
j o i ne d bat t l e . I l e av e t he c o mp u t a t i o n t o t he r eader .
143. T H E DU T C H M E N'S WI VE S . T h i s e l e gant puz z l e
dat es b a c k at l east t o 1 739. Fo r hi st or i cal i nt erest , I g i v e i t i n t he
or i gi nal dress, wh i c h seems t o h a v e i mpos e d t he E ng l i s h c ur r e nc y
o n t he Ne t he r l a nds . I has t e n t o st at e t hat al l t he A me r i c a n r e a de r
needs t o k n o w a b o ut t hi s c ur r e nc y i s t hat a g ui ne a c ont ai ns 21
shi l l i ngs.
T h r e e Du t c h me n a nd t hei r wi v e s g o t o ma r k e t , a nd e a c h
i ndi v i dua l buys s ome hogs. E a c h buys as ma n y hogs as he or she
pays i n shi l l i ngs f o r one hog . E a c h hus b a nd s pends a l t oge t he r 3
mo r e g ui ne a s t ha n hi s wi f e . T h e me n ar e n a me d H e nd r i c k , E l as
a n d C o r ne l i us ; t he wo me n ar e Gu r t r u n, K a t r u n a nd A n n a . H e n -
dr i c k buys 23 mo r e hogs t ha n Ka t r u n , whi l e E l as buys 11 mo r e
t ha n Gu r t r u n . Wh a t is t he n a me of e a c h ma n' s wi f e ?
144. T H E C RA Z Y QU I L T . " We l l , I de c l a r e ! " e x c l a i me d
M r s . T h o mp s o n . " I t ho ug ht I h a d mo r e pi eces t ha n t h a t ! " She
he l d u p a s mal l s quar e of c l ot h, ma d e by s t i t c hi ng t oget her a
n u mb e r of s mal l e r s quar es of v a r i e g a t e d col ors.
" I s t ha t t he s a me qui l t y o u s t ar t ed l ast s u mme r ? " i nqui r ed
M r s . P er ki ns .
" I n a wa y , yes. Y o u see, wh e n t he ba nns we r e put u p f or my
s on Joel , I b e g a n t o s ew hi m a c r az y qui l t . I t wa s n' t finished by
t he t i me of t he we d d i ng , a nd wh e n E f f i e c a me t o l i ve wi t h us she
89PUZZLES ABOUT INTEGERS
wa s ma k i n g a qui l t t oo. H e r s quar e wa s l a r g e r t ha n mi ne , b u t it
still wa s n' t e no u g h f or t he wi nt e r ni ght s . S o we de c i de d t o put
our t wo qui l t s t oget her . O u r pi ec es we r e b o t h t he s a me s i z e t h r e e
i nches o n t he si de. I r i ppe d t he st i t ches o ut of my qui l t a nd we
a dde d t he pi eces t o hers. We us ed u p al l my pi ec es a n d t he n we
ha d one ni c e bi g s quar e t ha t wa s r i ght f o r t he f our - pos t e r be d.
" Wh e n Joel a n d E f Be bui l t t hei r o wn hous e, I ur g e d E f f i e t o
t ake t he qui l t , b ut she wo u l d n ' t d o i t. She i nsi st ed o n g i v i ng me
bac k my pi eces, be c a us e she t ho ug ht I o ug ht t o ma k e mys el f a
qui l t t oo. She sai d she c o ul d ge t a l o n g qui t e c o mf o r t a b l y wi t h t he
s quar e she ha d l ef t . T h e s e ar e al l t he pi eces she g a v e b a c k t o me ,
but I de c l a r e ! I r eal l y t ho u g ht t her e we r e mo r e of t he m. No t t ha t
I car e, be c aus e I ' d be g l a d t o ha v e he r t ake t he m al l . Bu t i t does
seem s t r ange . "
" I t hi nk y o u ar e r i g ht , " sai d M r s . P er ki ns . " I r e me mb e r y o u
s howe d me y o ur qui l t j us t be f or e t he we d d i ng , a n d I r e me mb e r
t hat it wa s bi g e n o u g h t o c o v e r t ha t t abl e. N o w it doe s n' t e v e n
r e a c h t he c or ner s . "
Wh a t wa s t he si ze of t he c r a z y qui l t di s pl ay e d by M r s .
T h o mp s o n ?
145. T H E FOU R T RI A NGL E S P ROBL E M . We wi l l c o m-
me nc e t hi s f a s c i na t i ng puz z l e by out l i ni ng a s quar e wi t h ma t c h -
sti cks. T h e st i cks ar e of u n i f o r m l e ng t h a nd we wi l l us e a n
i nt egr al n u mb e r of st i cks o n e a c h si de of t he s quar e. H o w l a r g e
s houl d t he s quar e b e ? T h a t is r eal l y t he p r o b l e m!
T h e n we a r e g o i ng t o out l i ne f o ur r i ght t r i angl es, us i ng e a c h
si de of t he s quar e i n t ur n as one si de of a t r i angl e. I n e f f e c t , we
wi l l ha v e f o ur r i ght t r i angl es so a r r a ng e d as t o enc l os e a s qua r e
space.
A l l si des of al l t r i angl es mus t be i nt egr al mul t i pl e s of t he
l e ngt h of a ma t c hs t i c k. I n ot he r wor ds , we mus t not br e a k a ny
ma t c he s t o ma k e t he f i gur e.
N o t wo t r i angl es ma y be e qua l .
Wh a t is t he mi n i mu m n u mb e r of ma t c he s wi t h wh i c h s uc h
a f i gur e c a n be c ons t r uc t e d?
90 PUZZLES ABOUT INTEGERS
146. P A RT I T I ONS . By a partition of a n i nt eger is me a nt
a seri es of i nt eger s of wh i c h i t is t he s um. F o r e x a mpl e , al l pos-
si bl e par t i t i ons of 4 a r e :
4
3
2 2
2 1 1
1 1 1 1
H o w ma n y possi bl e par t i t i ons ar e t her e of a n i nt eger n?
Wh a t is t he f o r mu l a f or t he n u mb e r of par t i t i ons of a n i nt eger
i nt o r par t s ? A ns we r s t o t hese ques t i ons h a v e as yet not b e e n sup-
pl i e d by ma t h e ma t i c a l t he or y . L i k e c e r t a i n pr obl e ms of c o mbi na -
t or i al anal ysi s, t hey s eem t o de f y ge ne r al s ol ut i on. If y o u r e a d
wh a t e nc y c l o pe di a s h a v e t o say a bo ut par t i t i ons , y o u ma y c o nc l ude
t ha t t he s ubj e c t consi st s ent i r el y of t he pr opos i t i on t ha t t he n u m-
be r of par t i t i ons of a n i nt eger i nt o r par t s is e q ua l t o t he n u mb e r
of al l par t i t i ons i n wh i c h r is t he l ar gest par t .
T h i s pr opos i t i on, by t he wa y , ma y be p r o v e d as f ol l ows . T a k e
f o r e x a mp l e t he par t i t i ons of 6 i nt o 3 par t s , wh i c h ar e onl y
4 1 1
3 2 1
2 2 2
Re p r e s e nt t he i nt eger s i n e a c h pa r t i t i on by a n a ppr opr i a t e
n u mb e r of poi nt s i n a c o l u mn :
4 1 1 3 2 1 2 2 2
No w, l ook i ng across t he r ows , y o u see al l possi bl e par t i t i ons of
6 i n wh i c h 3 is t he l ar ges t p a r t : 3, 1, 1, 1 ; 3, 2, 1 ; 3, 3.
The Five-Suit Deck pr es ent s a pr a c t i c a l pr o b l e m t hat a c t ua l l y
o c c u r r e d a n d h a d t o be s ol ved by e mpi r i c a l me t ho ds f o r l ack of
t heor et i c al .
91PUZZLES ABOUT INTEGERS
147. T H E FI VE -S U I T DE C K. A f e w year s a g o, t he pr opos a l
wa s ma d e t o i nc l ude a fifth sui t i n t he s t a nda r d de c k of c ar ds . A
qua nt i t y of s uc h de c ks wa s i nd e e d ma n u f a c t u r e d a n d ma r k e t e d ,
t he fifth sui t b e i ng v a r i ous l y k n o wn as E a g l e s , C r o wn s , et c. C o n -
t r act br i dg e a n d ot he r g a me s wi t h t he 65 - c a r d de c k we r e t r i ed
o ut i n ma n y cl ubs . A n t i c i p a t i n g t ha t t he n e w de c k mi g h t ha v e
c o me t o st ay, s ome br i dg e aut hor i t i es has t i l y c o mp i l e d n e w t abl es
of c ha nc e s a n d r us he d i nt o pr i nt wi t h r evi s ed pr i nc i pl es of bi d-
di ng a n d pl ay.
I wa s as ked t o c o mp u t e t he r el at i ve pr obabi l i t i es of al l t ypes
of ha nds , cl assi f i ed a c c o r d i ng t o pa t t e r n.
By pattern is me a n t t he a r r a y of i nt eger s t ha t s ums t he n u m-
ber of c ar ds i n e a c h sui t he l d i n t he h a n d , e. g. , 5 4 3 1 is t he
pa t t e r n of a h a n d t ha t hol ds five c ar ds of one sui t , f o u r of an-
ot her , a n d so on.
T h e t abl e of pat t er ns f o r t he 5 2- c a r d de c k wa s c o mp u t e d
ma n y year s a g o a nd ha s b e e n i n c ons t a nt use f o r t he d e d u c t i o n
of st r at egi cal pr i nc i pl es . T h e t ot al n u mb e r of pa t t e r ns is 39, of
wh i c h t he five mos t f r e q u e nt ar e
4 4 3 2
5 3 3 2
5 4 3 1
5 4 2 2
4 3 3 3
A s a pr e l i mi na r y t o c o mp u t i n g t he f r e que nc i e s of v a r i ous
ha nds wi t h t he five-suit de c k, i t wa s neces s ar y t o wr i t e o ut al l t he
possi bl e pat t er ns .
A n or der l y wa y t o d o t hi s is t o st art wi t h t he gr eat es t pos-
si bl e c onc e nt r a t i on of c ar ds i nt o a f e w sui ts a n d pr o c e e d t o wa r d
t he mo r e e qua l di s t r i but i ons , t hus
'3
3
0 0 0
13
2 1 0 0
13
1 1 1 0
12
4
0 0 0
( I n five-suit br i dge , e a c h h a n d wa s de a l t 16 c ar ds , a n d t he
65t h c ar d wa s us ed as a wi d o w. )
92
PUZZLES ABOUT INTEGERS
H a v i n g wr i t t e n out t he pat t e r ns i n t hi s ma nne r , I l ooked f or
s ome c he c k u p o n t hei r t ot al nu mb e r . T h e s e pat t e r ns a r e of cour se
t he par t i t i ons of 16 i nt o a ny n u mb e r of par t s not i n excess of 5,
a nd wi t h t he pr ov i s o t hat n o pa r t ma y e x c e e d 13 ( si nce t her e ar e
onl y 13 c ar ds i n e a c h s ui t ) . Bu t o n l ooki ng i nt o t he s c ant l i t era-
t ur e on par t i t i ons , I f o u nd t ha t t he or y c o ul d not pr edi c t ho w
ma n y s uc h par t i t i ons t he r e s houl d be.
I t is e v i de nt t hat t he c o unt of par t i t i ons of a n i nt eger n is
c o mp o u n d e d f r o m t he c o unt of par t i t i ons of l esser i nt eger s n 1 ,
n 2 . . . 2, i . I t he r e f or e ma d e a t abl e of t he c ount of par t i t i ons
of i nt eger s 1, 2, 3, u p t o 16. T h e pr e c a ut i on wa s we l l - advi s e d,
f or I f o u n d t hat i n wr i t i ng t he pat t e r ns " b y e y e " I ha d ov e r l ooke d
t wo of t he m.
I f o u n d t hat s i mpl e r ul es c oul d be f o r mul a t e d as t o h o w to
bui l d u p t he t abl e of par t i t i ons so as t o a v o i d error. Wh a t I pr o-
pos e t o t he r e a de r is t ha t he shal l r edi s c over t hese r ul es by ma k i n g
a t abl e u p t o, say, t he i nt eger 8.
148. ORDE RS OF I NFI NI T Y. Wh e n we say t ha t a g r o u p of
obj e c t s is c o unt a b l e or enumerable we me a n t hat its me mb e r s c an
be a r r a ng e d i n one - t o- one c or r e s ponde nc e wi t h t he i nt eger s
1, 2, 3, et c. T h i s seri es is i nf i ni t e i n e x t e n t we c a n a l wa y s wr i t e
a n i nt e ge r l ar ger t ha n a ny g i v e n i n t e g e r b u t i t is discrete. Be-
t we e n a ny t wo a d j a c e n t me mb e r s of t he i nt e ge r set, as 7 a nd 8,
l i es a v oi d, c o nt a i ni ng no ot he r me mb e r s of t he set.
A n e x a mp l e of a c o unt a bl e set is t he pr i me number s . A l -
t h o u g h i nf i ni t e i n e x t e nt , t he pr i me numbe r s a r e di scret e. Be-
t we e n 31 a nd 37 l i e n o ot he r pr i me s .
T h e r e is a hi g he r o r de r of i nf i ni t y t ha n is possessed by t he
i nt eger s. T h e cl ass of al l r eal numbe r s , bot h r at i onal a n d i rra-
t i onal , is not e nume r a b l e , f o r t he r eas on t ha t i t is not di scret e. I n
wh a t e v e r wa y t he numb e r s a r e a r r a ng e d, it wi l l be f o u nd t hat
b e t we e n a ny t wo a d j a c e n t me mb e r s l i e a n i nf i ni t ude of ot her
me mb e r s of t he set.
By a rational n u mb e r is me a n t one t hat c a n be expr es s ed as
t he quo t i e nt of t wo i nt eger s . T h u s a n u mb e r ma y be r at i onal al -
93PUZZLES ABOUT INTEGERS
t h o u g h it is a ne v e r - e ndi ng de c i ma l , e. g. , %= . 7 1 4 2 8 5 7 1 4 2 8 5 . . .
A n endl ess de c i ma l , i f r at i onal , s ooner o r l at e r r e a c he s a g r o u p of
di gi t s wh i c h i t t he r e a f t e r r epeat s f or e v e r .
A n irrational n u mb e r c a n n o t be e xpr e s s e d as t he quot i e nt of
t wo i nt eger s, b ut onl y as t he s um or l i mi t of a n i nf i ni t e seri es, a n d
t he c o mp u t a t i o n of its v a l ue pr o duc e s a n endl es s no n- r e pe a t i ng
de c i ma l . A f a mi l i a r i r r at i onal n u mb e r is it, 3 . 1 4 1 5 9 . . .
A cl ass wh i c h is not di scr et e is c a l l e d continuous, a nd t he
ki nd of i nf i ni t y possessed by t he real n u mb e r s e x t e r n a l a n d i n-
t er nal , as it we r e i s c al l ed the continuum.
No w, v i e wi ng t he r at i onal numb e r s a p a r t f r o m t he i r r a-
t i onal , i t ma y s eem t ha t t he f o r me r i n t he ms e l ve s a r e a c o nt i nuo us
cl ass. F o r e x a mpl e , b e t we e n % a nd % we c a n find a n i nf i ni t y of
ot he r r at i onal numbe r s of i nt e r me di a t e ma g n i t u d e , e. g. ,
1
%4 -
Bu t t he r at i onal numbe r s ar e i n f a c t e nume r a b l e . T h e t ask
he r e pos ed t he r e a de r is t o pr o v e thi s f a c t , b y a r r a ng i ng t h e m i n a
de mons t r a bl y di s cr et e or der .
I X. Decimation P uzzles
149. T U RKS A ND C H RI S T I A NS . I n f o r me r t i mes, t he l a w-
f u l pe na l t y f or mu t i n y wa s t o e x e c ut e one - t e nt h of t he c r e w.
C u s t o m o r da i ne d t ha t t he v i c t i ms be s el ect ed by c o u nt i ng of f
e v e r y t e nt h ma n f r o m a r a n d o m a r r a ng e me nt of t he c r e w i n a
ci r cl e. H e n c e t he t e r m decimation. I n t he c our s e of t i me t hi s t e r m
ha s c o me t o be a ppl i e d t o a ny de pl e t i on of a n as s e mbl age by a
f i x e d i nt e r val , r egar dl es s of wh e t h e r t hi s i nt e r v a l is t en or an-
ot he r nu mb e r .
A v e r y ol d puz z l e a b o ut de c i ma t i o n is Turks and Christians.
T h e st ory goes t ha t 1 5 T u r k s a n d 1 5 C hr i s t i ans we r e a b o a r d a
s hi p c a u g h t i n a s ever e s t or m. T h e c a pt a i n de c i de d t o pr opi t i at e
t he e l e me nt s by t h r o wi n g hal f hi s pas s enger s ov e r boa r d. I n or de r
t o l e av e t he s el ect i on of v i c t i ms t o c ha nc e he a r r a ng e d al l 30 i n
a ci r cl e, wi t h t he a nno u nc e d i nt e nt i on of c o u nt i ng of f ever y
t hi r t e e nt h ma n . A c l e v e r C hr i s t i a n poi nt e d out t o hi s f el l ows h o w
t o t a ke pl a c e s i n t he c i r c l e so t ha t onl y t he 1 5 T u r k s wo u l d be
c o unt e d out f o r j et t i s on. Wh a t we r e t hese pl ac e s ?
150. BOYS A ND GI RL S . Fi v e boys a nd f i ve gi rl s f o u n d
f i ve penni es . A di s put e ov e r o wne r s hi p e ns ue d, a nd it wa s de-
c i de d t o a r r a ng e t he g r o u p i n a ci r cl e, c o unt out i ndi v i dual s by a
fixed i nt e r val , a n d g i v e e a c h a p e nny as he or she l ef t t he ci r cl e.
T h e p l a n wa s a d v a n c e d by a c l e v e r b ut uns c r upul ous boy, wh o so
a r r a ng e d t he c i r c l e t ha t by c o unt i ng a c e r t a i n gi rl as " o n e " he
94
DECIMATION PUZZLES 95
c oul d c o unt o u t al l t he boys first. Bu t t he gi rl c o u nt e d as " o n e "
i nsi st ed o n he r r i ght t o c hoos e t he i nt e r v al of de c i ma t i o n, a n d he r
as t ut e c hoi c e c o unt e d out al l t he gi rl s first.
T h e a r r a ng e me nt of t he c i r c l e wa s t hus :
1 2 3 4 5 ^ 7 8 9 10
G G G B B B G B B G
T h e c o unt i n e a c h cas e st art s wi t h t he gi r l at t he e x t r e me l e f t ,
goes t o t he r i ght , a n d t he n r et ur ns t o t he l e f t mos t i ndi v i dua l re-
ma i ni ng . E a c h i ndi v i dua l c o unt e d o u t st eps o ut of t he c i r c l e a n d
of cour s e is not i nc l ude d i n t he c o unt t he r e a f t e r .
Wh a t i nt er val wi l l c o u nt o ut t he five boys first, a n d wh a t wi l l
c o unt o ut t he five gi rl s first?
151. NA T I VE S A ND BRI T ONS . A pa r t y of e xpl or e r s i n
A f r i c a , five Br i t ons a n d five na t i v e por t er s , f el l i nt o t he ha nds of
a s av age c hi ef , wh o c ons e nt e d t o r el ease t h e m onl y o n c ondi t i on
t ha t hal f t he pa r t y s ubmi t t o flogging. T h e t e n me n we r e a r r a ng e d
i n a ci r cl e i n t hi s o r de r :
1 2 3 4 5 6 7 8 9 10
B N B N N B N B N B
T h e five me n t o be flogged we r e t o be s el ec t ed by c o u n t i n g
a r o und t he ci r cl e by a f i xed i nt e r v al unt i l five we r e c o unt e d out .
T h e Br i t ons h a d a r r a ng e d ma t t e r s so t ha t t he y c o ul d first c o u nt
o ut t he five nat i ves. Bu t t he c hi ef di d t he c o u n t i n g a n d chos e
s uc h a n u mb e r as t o sel ect t he five Br i t ons .
Wh a t i nt er val a n d wh a t s t ar t i ng poi nt di d t he Br i t ons a n d
t he chi ef r es pect i vel y ha v e i n mi n d ? ( T h e c ount goes onl y i n one
di r e c t i on a r o und t he ci r cl e, b e i ng i ndi c a t e d i n t he d i a g r a m by
a s c e ndi ng numbe r s . )
152. J A C K A ND J I L L . I t is not g e ne r a l l y k n o wn t hat J a c k
a nd Ji l l h a d s ome br ot her s a n d si sters, a nd wh e n t he que s t i on
ar ose w h o wa s t o f e t c h t he wa t e r t hey de c i de d t o set t l e t he ma t t e r
by l ot . T h e five gi rl s a r r a ng e d t hems el ves i n one ci r cl e a nd t he
f o ur boys i n anot he r . S o me b o d y n a me d a n u mb e r a t r a n d o m, a n d
e a c h ci r cl e wa s de c i ma t e d c l oc kwi s e by t hi s n u mb e r unt i l onl y one
96 DECIMATION PUZZLES 96
b o y a n d one gi r l we r e l ef t . J a c k c o unt e d f or t he boys, be g i nni ng
wi t h hi s l e f t - ha nd ne i g hbor as " o n e , " a nd Ji l l c o unt e d f or t he
gi rl s, al s o b e g i nni ng at he r i mme d i a t e l ef t . We l l , y o u k n o w wh a t
h a p p e n e d ; i t s e e me d pr e t t y ha r d l i nes t o J a c k a n d Ji l l t ha t that
pa r t i c ul a r i nt e r v al of de c i ma t i o n wa s c h o s e n wh a t wa s i t , by
t he wa y ?
153. OU T A ND U NDE R. Wh e n I wa s a boy, one of my pal s
s h o we d me a " t r i c k " he h a d j us t l e ar ne d. F r o m a de c k of c ar ds
he t ook t he 4 ki ngs a n d 4 queens . H e a r r a ng e d t he 8 c ar ds out
of my s i ght , t he n pr o c e e de d t o de a l t h e m out t hus :
T h e t op c a r d wa s pl a c e d f a c e u p o n t he t abl e. T h e s ec ond
wa s t r a ns f e r r e d f r o m t op t o bo t t o m. T h e t hi r d wa s t hr o wn out
f a c e up, t he f o ur t h t r ans f e r r e d t o t he bo t t o m, a nd so on. T h e
c a r ds as t hey we r e t hr o wn out f a c e u p c a me i n t he or de r , ki ngs
a n d que e ns al t e r nat e l y.
M y p l a y ma t e c ha l l e ng e d me t o di s c over t he c or r e c t i ni t i al
o r de r of t he c ar ds t o p r o d u c e t hi s resul t . By t ri al a nd er r or I di s-
c o v e r e d t he secr et . L a t e r , o n t hi nki ng it ov e r I f o und a n easi er
wa y t o t he s ol ut i on. T h e ne x t t i me we me t , I h a n d e d hi m a f ul l
de c k of 52 a nd as ked hi m t o deal it " o u t a nd u nd e r . " Wh e n he
di d so he wa s as t oni s hed t o see al l f o ur sui ts c o me out i n s equenc e.
I wi l l not ask t he r e a de r t o r ec ons t r uc t t he pr e a r r a ng e d
o r de r f o r t he 52 c ar ds , but I a m c ur i ous t o k n o w h o w qui c kl y he
c a n a r r a ng e t he 13 c ar ds of one sui t so t hat t hey wi l l deal " o u t
a n d u n d e r " i n o r d e r : A , 2, 3 . . . 10, J, Q , K .
154. T H E NI GH T M A RE . Fo l l o wi n g t he i nci dent s na r r a t e d
i n Out and Under, my c h u m h a d a ni g ht ma r e . H e d r e a me d t hat
a d e mo n pe r c he d on hi s be d, p r o d u c e d a mons t r ous dec k of
c ar ds , a n d c o mme n c e d t o deal " o u t a nd u n d e r . " T h e c ar ds we r e
e v i de nt l y not t he usual ki nd, b e i ng s i mpl y nu mb e r e d f r o m 1 up.
I t wa s e v i de nt t hat t he dec k ha d be e n a r r a ng e d i n a s c e ndi ng
s e que nc e f r o m t he t op d o wn .
" N o w , " sai d t he d e mo n , " I shal l abol i s h al l c a nd y stores a nd
al l c o mi c ma g a z i ne s i f y o u d o not i ns t ant l y t el l me wh a t c ar d wi l l
DECIMATION PUZZLES 97
be t he l ast t o c o me out . A l s o, y o u mus t tel l me wh e n c a r d No . 288
wi l l c o me out , a nd wh a t n u mb e r wi l l b e o n t he 643r d c a r d I
t ur n u p . "
T e r r i f i e d a nd di s ma y e d, t he boy s t r uggl ed t o s peak i n v a i n ;
t he d e mo n di ssol ved i n a bur s t of l i ght , a n d t he boy f o u n d hi ms el f
a l one wi t h t he mo r n i n g s un s t r e a mi ng t h r o u g h t he wi ndo ws . H e
j u mp e d i nt o hi s cl ot hes, t our e d t he ne i g hb o r ho o d, a nd wa s c o m-
f or t e d t o di s c ove r t hat t he d e mo n h a d not yet c ar r i e d o ut hi s
t hr eat .
O n he a r i ng my c h u m' s t al e I r e ma r k e d t hat we s houl d be
abl e t o find t he ans wer s t o t he r i ddl es i f we b ut k n e w t he n u mb e r
of c ar ds i n t he de mo n' s de c k.
" We l l , " s ai d he, " I di d not i c e t ha t t he b o t t o m c a r d wa s
No . 9 7 1 . "
Fo r t hwi t h we t ook o ut i ns ur a nc e agai ns t a ma j o r c a t a s t r ophe
by figuring out t he ans we r s t o t he d e mo n' s ques t i ons . I s houl d be
g l a d t o t el l t h e m t o t he r e ade r , b ut as I unde r s t a nd t he r ul es of
t he De mo n ' s U n i o n , no pr ot e c t i on is g i v e n t o a pe r s on w h o does
not figure o ut t he ans we r s f o r hi ms el f .
X. P ermutations and C ombinations
155. FU NDA M E NT A L FORM U L A S . By a combination is
me a n t a s ub- g r o up t a ke n o ut of a l ar ge r set of obj ec t s , whe r e t he
i de nt i t y of t he me mb e r s of t he s ub- g r oup is i mpo r t a nt b ut not t he
a r r a n g e me n t or or de r . F o r e x a mp l e , A B C is a c o mbi na t i o n of 3
l et t ers o u t of t he 26 i n t he a l p h a b e t ; t he c o mbi na t i o n B C A is
i de nt i c a l wi t h i t.
By a permutation we me a n a c o mb i na t i o n i n wh i c h t he order
as we l l as i de nt i t y of t he c o mpo ne nt s is i mpo r t a nt . T h u s , B C A is
a di f f e r e nt p e r mu t a t i o n f r o m A B C .
T h e t e r m c o mb i na t i o n a l wa y s i mpl i es a s ub- g r oup out of a
l a r g e r s et ; s uc h a n expr es s i on as " t he c o mb i na t i o n of n out of n
o b j e c t s " me a n s n o mo r e t h a n " t h e set n." Bu t " t he pe r mut a t i ons
of n obj e c t s o ut of n" ha s r eal me a n i n g , si nce t he s a me set of
obj e c t s c a n be a r r a ng e d i n di f f e r e nt or der s .
I t is e v i de nt t hat t he n u mb e r of pe r mut a t i ons possi bl e i n
s e l e c t i ng s ub- g r oups is a t l east e qua l t o t he n u mb e r of c o mbi na -
t i ons, a n d us ual l y e x c e e ds i t. A s a r ul e, pe r mut a t i ons ar e easi er t o
r e c k o n di r ec t l y t ha n c ombi na t i ons . T h e f u n d a me n t a l f or mul a s f or
c o mb i na t i o ns ar e de r i v e d by r e c k o ni ng pe r mut a t i ons a nd t hen
d i v i d i ng by t he n u mb e r of pe r mut a t i ons of wh i c h e a c h c o mb i na -
t i on is s us cept i bl e. I n s ome pr obl e ms , h o we v e r , it pr ove s easi er
t o r e c k o n c o mb i na t i o ns di r e c t l y a nd / o r t o c o unt pe r mut a t i ons
f r o m t hi s nu mb e r .
98
99PERMUTATIONS ANDCOMBINATIONS
Some of the fundamental formulas are as follows:
p .
(0
(2)
(n-r)!
(3)
C ? =
(4)
rl(nr)!
T he symbol ! is read "factorial" and means the product of
all integers 1 X2X3 . . . up to the specified integer. For example,
5!== 1 X2 X3 X4 X5 = !2o. If you have to expand any of the above
formulas by actual multiplication, start by cancelling out the
denominator terms (usually the answer must needs be an integer,
so that the entire denominator will cancel). For example,
Formula (1) gives the number of permutations (P ) of n
out of n objects. Formula (2) gives the number of permutations
of n objects taken r at a time. (3) gives the permutations of n
objects (all at a time) of which a are of one kind, b of another
kind, c of another kind, etc. Within each kind the objects have
no separate identity. (4) gives the number of combinations (C )
of n objects taken r at a time.
T he counting of permutations and combinations is essential
to the solution of many practical problems of diverse types. T he
best-known type is questions of probability. Before we consider
that subject, I will give the reader some exercises in counting.
Some of them invoke formulas such as the above; in others, the
reader must devise his own formulas. T he purely mathematical
part of a counting problem is usually easy; the real task is often
to see how to construe the problem so as to guard against the
_7[ __ 7X6X5
3
!
4
! 3X2X1
7X5=35
100 PERMUTATIONS ANDCOMBINATIONS
errors of missing some possibilities and counting others twice over.
T he subject is proverbially replete with pitfalls. In the computa-
tion of card game probabilities, for example, even the most expert
mathematicians have made errors of construction, while the odds
calculated by an inexperienced person are almost always wrong.
A typical error is discussed in the following problem.
156. A C OM M ON M I S T A KE . Smith and Jones had an
argument as to the relative probabilities of two types of bridge
patterns: 5 4 2 2 and 5 4 3 1 . (T he four digits in each pattern
indicate the number of cards held in a suit. T hus, 5 4 2 2 indicates
a holding of 5 cards in one suit, 4 in another, and 2 in each of
the other two suits.) T hey settled the argument by a calculation
as follows:
For the 5 4 2 2 pattern we will first reckon all possible com-
binations of 5 cards out of a suit, which is 13!
J
. T he number
5!8!
of combinations of 4 cards out of a second suit is 13!
4
!
9
r
For the
1 3!
third and fourth suits we have, each, _ .. T he entire pattern
2! 1 1 !
can be formed in
!3
!
v
, 13! 13! 13!
5 ! 8f
X
4
!
9
!
X
2! i i !
X
2 ! II!
different ways.
Similarly, for pattern 5 4 3 1 the count of combinations is
1 3! 1 3! 1 3! 1 3!
0
V
0
y y
0
5 ! 8! 4 ! 9 1 3 ! 10! 1 2!
T he ratio of the two quantities is
G
(5 4 2 2) 3Xi 2_36
c
(5 4.3 I ) ~ 2 Xi i
_
2 2
Smith and Jones agreed that the 5 4 2 2 pattern is more
likely to be dealt than the 5 4 3 1 m the ratio of 18:11.
101PERMUTATIONS ANDCOMBINATIONS
But this conclusion is wrong. T he fact is that the 5 4 3 1 pat-
tern is the more frequent. Where did Smith and Jones go wrong?
T he basic idea of computing the number of different hands
of 13 out of 52 cards that are of 5 4 2 2 and 5 4 3 1 pattern re-
spectively is sound. So is the idea of setting the two quantities in
ratio. T he error is entirely in the computation.
157. T H E A NA GRA M BOX. If a box of anagram letters
contains eight G's, nine M 's, and thirty each of A , N, R, how many
ways are there to pick out and arrange letters to make the word
A NA GRA M ?
158. M I S S I S S I P P I . With the same anagram box, in how
many ways can you pick out and arrange letters to make
M ISSISSIP P I? Besides nine M 's there are twenty-eight I's,
twenty-four S's, and eight P 's.
159. P OKE R DI C E . T he several varieties of the game
"poker dice" are all based on the casting of five (or more) dice,
the resultant combination of numbers then being treated as a
poker hand.
T he prevalent ranking of the various hands, from high to
low, is as follows:
Five of a kind
Four of a kind
Full house
Straight
T hree of a kind
T wo pairs
One pair
No pair
In some circles, the 1 -spot denomination is ranked above the
6, instead of below the 2. But even here only two kinds of straights
are admitted: high straight 6-5-4-3-2, low straight 5- 4- 3- 2- 1.
102 PERMUTATIONS ANDCOMBINATIONS
By "no pair" we of course mean no combination that will
place the hand in one of the higher classes.
T his ranking of hands is borrowed from poker as played with
cards. It is based on the relative chances of being dealt each kind
of hand. T he total number of different combinations of five cards
out of 52 is 5 2 V 5 ! 4 7 1 =2 , 5 9 9 , 7 6 0. T he chance of receiving a
hand of specified type is the ratio of the number of combinations
of this type to this total. For example, what is the chance of being
dealt a straight flush (including royal flush) ? E ach suit contains
10 possible straight flushes, which can be topped by any denomi-
nation 5,6 . . . . K, A . A s there are 4 suits, the total number is
4 X 1 0 = 4 0 . T he chance of being dealt a straight flush is then
4O/2,599>7
6
-
If we were to rank the hands in poker dice by reference to
the conditions of dice play instead of card play, we would reckon
on the basis of 6
5
=7,776. T his is the total of different permuta-
tions that can be cast with five dice of 6 faces each. A s an easy
exercise, let the reader calculate the number of permutations that
fall in each class of hand. Reckon each type separately and check
by adding the numbers; the total should be 7,776. (T he factorials
to be expanded are happily very small. T he task is essentially logi-
cal : to devise a procedure for setting up each equation with assur-
ance that no combination is overlooked and none counted twice.)
M 1 r si 1 M
1 N 1 M 1
N 1 I A 1 N
1 M 1 N 1
M 1 r N| 1 M
103 PERMUTATIONS ANDCOMBINATIONS
160. T H E M I NI M P U Z Z L E . Start on any one of the squares
marked M and by consecutive moves to adjacent squares spell out
the word M INIM . H ow many different ways can you do it?
161. T H E S P Y. T he sketch shows the plan of an industrial
establishment where some highly secret weapons of war are manu-
factured. T he area is enclosed by a wall containing only three
gates, which are guarded night and day.
cm cm
cm cm cm cm cm

cm
cm cm cm cm cm
cm 0 cm cm tm cm cm cm
cm cm cm cm cm
cm cm cm cm cm cm cm

cm cm cm cm cm cm

cm cm cm cm cm

cm
1 1
cm cm cm cm cm cm
T he security officers discovered that information was leaking
out of the plant. L ittle progress was made in the investigation
until the following note was discovered in a trash barrel:
"M eet me every T uesday at the corner near my office. E nter
by northwest gate and take a different route each time. T hat gives
you 715 choices."
It was inferred that this note was sent by some spy in the
plant to a courier, who was to receive information and convey it
outside the plant. In an effort to track the spy, the assumption
was made that the courier, on entering by the gate in the upper
left of the sketch, would proceed always east or south, never going
backwards to west or north. On this assumption, the corner that
104 PERMUTATIONS ANDCOMBINATIONS
can be reached in just 715 ways was easily calculated, and a sur-
veillance kept on this corner was indeed successful in spotting the
conspirators.
What was the corner where the meetings took place?
162. H OW M A NY T RI A NGL E S ? If every vertex of a regu-
lar octagon is connected with every other, how manj triangles
will be formed?
163. T H E C OI N DROP P E R. While the streetcar was stopped
at a red light, the operator jingled the silver in his coin dropper.
A larmed by the sound, he dumped all the coins out, counted them,
and shook his head doubtfully. It was evident that there had been
a run on his change, and he was wondering whether it would last
out the trip.
T his incident suggested a puzzle to me: In how many differ-
ent ways could the operator's small change bccome exhausted?
T he coin dropper comprised 4 cylinders. A push on a thumb-
lever at the bottom of a cylinder would allow the lowermost coin
inside to drop into the operator's hand.
We will say that the first cylinder contains 4 nickels; the
second cylinder, 3 nickels; the third cylinder, 5 dimes; and the
fourth cylinder, 2 quarters. T he question is: In how many differ-
ent orders can these 14 coins be taken one-by-one from the coin
dropper?
E ach coin has individual identity. T hat is to say, taking 2
nickels in succession from the first cylinder gives a different order
from taking 1 from the first and then 1 from the second. A nd of
course we must assume that the hapless operator receives no addi-
tional change during the process.
164. ROT A T I ON P OOL . A professional pocket billiard
player practices every day by playing a kind of rotation game.
T he balls are numbered from 1 to 15, and on the wall hangs a
rack with 15 numbered cubicles. A s each ball is pocketed, it is
removed from the pocket and placed in the cubicle of same
105PERMUTATIONS ANDCOMBINATIONS
number. Just to make it harder, the player stipulates that the row
of balls in the rack must never at any time show an interior gap.
T hus, if the 6 is pocketed first, the 5 or the 7 must be pocketed
next. T here is free choice of which ball to sink first, but after that
the choice is restricted to balls numbered in sequence with those
already in the rack.
T he question is, in how many different orders can the player
clean the table of all 15 balls?
T his problem was once posed to me by a mathematics
teacher, who stated that he knew the formula for the answer but
did not see any easy way to calculate it. T he trouble is that if you
commence counting the number of choices open after each play,
you run into different circumstances. For example, if you pocket
the 2 first, you then have choice of 1 or 3. If you choose 3, you
again have two choices. But if you choose 1, there is no further
choice at all. A gain, if you start with 7 or 8, you must continue to
have choices for some time, but the point at which you cease to
have a choice depends on the particular order up to that point.
A s in many permutation problems, there is here a way of
construing the question so as to make calculation of the answer
absurdly simple.
165. T H E NE C KL A C E . E loise has a quantity of glass beads
in four colors, red, yellow, green, and blue. She amuses herself by
stringing them on wool yarn in various designs. One of her favor-
ite designs is a necklace of 20 beads, in blocks of 4 of a color.
H ow many different patterns can E loise make on this plan?
She must use the beads in blocks of 4 of the same color, but we
will not insist that she make adjacent blocks of a different color.
She may make the entire necklace of one color if she chooses. Or
she may use two or three or four colors.
106 PERMUTATIONS ANDCOMBINATIONS
166. T OU RNA M E NT S C H E DU L E S . One evening I visited
one of the best-known chess clubs in this country with the inten-
tion of participating in the weekly "rapid transit tournament." In
rapid transit, a player may deliberate no more than 10 seconds
before each move.
T he entrants being assembled to begin, the tournament direc-
tor assigned a number to eachbut then came a halt. "Where is
the book?" he inquired. Search began. "T he book! T he book!"
went hue and cry through the rooms. T o the entrants the director
announced placidly, "We'll have to call off the tournament if we
don't find the book." But presently T he Book was unearthed, and
the play could commence.
T his precious tome, it appeared, was a shabby black notebook
into which had been laboriously copied the various schedules for
different numbers of entrants. T he director was wont to find the
proper schedule in this book and call out the pairings of players
before each round.
I inquired why the book was deemed so invaluable, and was
seriously informed that the chess tournament schedules, devised
by some bygone mathematical prodigy, have been published only
in an obscure volume obtainable only at the largest public li-
braries.
I could not help being amused. A t a conservative estimate,
there are 15, 000, 000 persons in the U nited States who can tell you
where to find any schedule a chess tournament might need, within
a few minutes. Just step to the nearest phone and call up a bridge
club.
T he widespread popularity of duplicate contract bridge has
led to the publication and sale of thousands of schedule cards and
books containing schedules. A ny one of the round-robin pair or
individual schedules will serve for a chess tournament as well.
What is needed in chess is a list of pairings whereby each of
n players can be opposed just once to every other. T he require-
ments of a bridge schedule are much more exacting. In a round-
robin pair schedule, every pair has to be opposed to every other
pair just once, and at the same time a set of duplicate boards must
107 PERMUTATIONS ANDCOMBINATIONS
be routed so that every pair may play every board. In an indi-
vidual schedule, every player must be paired once with every other
as partner, must meet every other twice as an opponent (prefer-
ably once on his left and once on his right), and a set of boards
also must be included in the movement.
T he making of such schedules, which involve fundamental
problems of combinatorial analysis, has engaged the attention of
many eminent mathematicians. Interest in the subject was
spurred by the propounding in 1850 of P . T . Kirkman's famous
"P roblem of the School Girls." In its simplest form, this problem
states that a schoolmistress was in the habit of taking her girls for
a daily walk. T he girls were 15 in number, and were arranged in
5 rows of 3 so that each girl might have 2 companions. T he prob-
lem is to dispose them so that for 7 consecutive days no girl will
walk with any of her school-fellows in any triplet more than
once. Several types of solutions were developed for n15, and
also for many other multiples of 3. T he inquiry was extended to
square numbers, and then to multiples of 4. T he literature on the
subject has grown to voluminous size without exhausting the field,
for it is found that there is no general solution for groups of r out
of n objectsonly particular types of solutions available for n of
certain forms.
E dwin C . H owell, an A merican mathematician, devised pair
schedules for whist (and bridge) for any number of pairs from 3
up to 46. T he legend devoutly believed by the bridge world is that
while working on a schedule for 47 pairs he went madand why
not!but I have not found authority for the story.
It is not my intention to pose the reader such a problem as
will bring H owell's fate upon him. I am satisfied to point out the
extraordinary fascination of the subject of combinatorial analysis.
A nd to encourage the reader to rediscover some of the basic prin-
ciples, I will ask him to solve the simplest of problems.
We will suppose that the chess club has lost its little black
book and has never heard of bridge schedules. Nine entrants are
desirous of playing rapid transit. T he contest must be scheduled so
that during 9 rounds each player meets every other once and has
108 PERMUTATIONS ANDCOMBINATIONS
one bye. A lso, each player must have White 4 times and Black 4
times. Won't you make such a schedule for the club?
167. P H A L A NXE S . L ittle Wilbur, a precocious child whom
we have met before (in No. 137Little Wilbur and the Marbles),
has a number of lead soldiers, which he likes to arrange in rec-
tangular phalanxes. No doubt he is working out the answer to the
following puzzle.
With just a dozen soldiers, we can form two different pha-
lanxes 6X2 and 4X3- We may count 5X6 and 3X4
a s
different,
since the width of rank and depth of file are distinct dimensions.
We might also count 12X
1
and i X' 2al l the soldiers in one rank
or one file. A ltogether we see that there are 6 possible phalanxes
with the 12 soldiers.
T he number of phalanxes is evidently determined by the
number of factors in the total. Suppose that s, the total of soldiers,
is composed of abed, four different prime factors. T hen counting
the phalanxes is merely a matter of counting the combinations:
abed
X
I
abc
X
d
abd
X
c
acd
X
b
bed
X
a
ab
X
cd
ae
X
bd
With only four factors, it is easy to write out all the combina-
tions. But for larger numbers we need a formula that will enable
us to compute the total directly. H ence:
H ow many phalanxes can be made out of s soldiers, if s is
composed of n different prime factors?
XI . P roblems of P robability
168. P ROBA BI L I T Y. What is the probability that two cards
drawn at random from a full deck of cards will be a pair?
T o answer this question, we first count the number of differ-
ent combinations of two cards that may be drawn:
52!
c
5
!=p-=1326
2
2! 50!
0
T hen we compute the number of these combinations which
are pairs. T here are four cards of each denomination; six differ-
ent pairs can be made out of the four cards. A s there are thirteen
denominations, the total number of pairs is 6X
I
3=78. T he prob-
ability that the two random cards will be a pair is given by the
ratio 78 1
1326 17 '
T o generalize: If j is the total number of events, one of which
must occur, and / is the number of events that fall into a class X,
then the probability that an event will be of class X is the ratio
L.. T he set s must be a complete list of all possible events, and it
must be mutually exclusivethe occurrence of any one event
must exclude the occurrence of any other.
A ny probability less than certainty is a fraction less than 1.
T he probability that an event will not be of class X is 1-j-
109
110 PROBLEMS OF PROBABILITY
T his follows from the definition of s as a set of events, one of
which must occur.
Familiar as are these theorems, there is one point that es-
capes many persons. A ll mathematical calculation of probability
is deductive: it deduces the consequences of an initial assumption
as to probability. H ow the assumption originates, by what tests it
can be validatedthese questions lie beyond mathematics. C om-
putation can no more create the initial hypothesis than geometry
can create its postulates.
Statements of probabilities are quite generally made without
accompanying statements of their assumptions. But that does not
mean that the assumptions are not there. It usually means that
the assumptions are evident and are generally accepted, e.g.,
statement of odds in throwing dice, in the distribution of cards, in
drawing colored balls from an urn. But sometimes it means that
the statement of chances rests on an implied hypothesis which the
speaker himself would reject if it were made explicit.
Failure to scrutinize assumptions raises many a tempest in a
teapot. A few years ago a controversy arose over certain proba-
bilities affecting contract bridge. T he nugget of the argument is
given in No. 169That King of Clubs! P erhaps the reader would
like to settle the argument for himself.
169. T H A T KI NG OF C L U BS ! Smith and Jones were part-
ners in a game of contract bridge. Smith became declarer at
three No T rumps, neither opponent having bid. North made an
opening lead of the 3 of Spades, dummy went down, and these
were the cards in sight to declarer:
NORTH
WEST EAST
SMITH (declarer)
0 A Q 7 5
* 9 5 4 3
A A K
9 4 2
A 5 4
A J 8 3
0 J 3
* A J 10 6 2
J ONES (dummy)
SOUTH
111PROBLEMS OF PROBABILITY
South played the J of Spades, and Smith won with the King.
Smith led the 3 of C lubs, North followed with the 7, dummy the
10, and South won with the Queen. Back came a Spade from
South, knocking out Smith's A ce. Smith led the 4 of C lubs, North
played the 8, and then Smith went into a huddle. Finally he put
up the A ce from dummy, but South discarded a Spade instead of
producing the K of C lubs as declarer had hoped. A nother round
cleared the C lubs, but then North cashed three Spade tricks, de-
feating the contract. T hen the argument started.
Jones: Why did you refuse the second C lub finesse? You
double-crossed yourself.
Smith: I figured that the odds were on South to hold the
Ki ng of C lubs.
Jones: H ow do you figure that?
Smith: On the second round, after North played the 8 of
C lubs, he had nine cards left, while South still had ten. So
the odds were 10 to 9 that South held the King.
Jones: T hat' s a cockeyed argument, and you know it! If that
were so, the chances would always be better that fourth hand
rather than second hand holds any given card. You know
perfectly well that before you touched the C lubs the odds
were better on the double-finesse than on swinging the A ce,
to lose only one trick.
Smith: M aybe so, but the odds changed after the first C lub
round.
Jones: H ow so? A ll the first round showed, and North's play
of the 8 on the second round, was that you would have to
guess. If anything else would have happened, there would
have been no problem.
Smith: Still, the odds must have changed, because the cards
that fell on the first round excluded some combinations that
might have existed before it was played.
Jones: No, the calculation of the initial chances count only
the relevant cases, where you will have to guess. T he first
C lub round simply showed that this was a relevant case.
112 PROBLEMS OF PROBABILITY
Smith: Well, I' m not at all sure that the initial chances are
better for the double-finesse, anyhow.
Jones: L et's look it up.
Forthwith the players referred to a compilation of card
probabilities, and found the following passage:
"With a suit A J 10 x x/x x x x, the defender over the A ce
was dealt
1. Ki ng singleton in 62/ 1 000 cases;
2. Queen singleton in 62/ 1 000 cases;
3. King-Queen blank in 68/ 1 000 cases.
"A first-round finesse is indicated, since it is as effective as
the play of the A ce in these three cases and is superior in the cases
where the defender under the A ce was dealt both missing honors
plus one or both of the small cards. T he double-finesse is seen to
be superior to playing the ace by 124:68."
Smith: T he book is wrong on its own facts! When I took the
first C lub finesse and South played the Queen, I knew that
C ase 1 heresingleton Ki ngdi d not exist. T hat left the
odds 68 to 62 that South was dealt King-Queen rather than
blank Queen. So my play of the A ce was correct!
What says the reader? Is Smith right or wrong? A ssuming
the statistics on the splits, as given by the book, to be correct, do
the odds favor the play of the A ce or the Jack on the second
round of the suit, after the first finesse has lost to the Queen?
170. ODDS . T he likelihood that an event will occur may be
stated either in terms of probability or in terms of odds. Both
methods are widely used. C onfusion sometimes arises because the
language of a statement does not make clear which point of view
is intended. Chances for and chances against are ambiguous
terms; only the context can show whether they mean probability
or odds. j
If the probability that an event will occur is - , then the odds
against it are sf to /, and the odds for it are / to s/. (Odds are
usually written as ratios and probabilities as fractions.)
113PROBLEMS OF PROBABILITY
T he language of odds is favored by bettors, for whom it has
an obvious advantage. If a bookmaker "lays" odds of 3 to 2 on
P acemaker to win, a bettor who puts $2 on P acemaker will receive
$3 if the horse does win. But in bookmaking the odds have to be
translated into probabilities (however much the fact is disguised
by algebraic short cuts). For example:
In a three-horse race, a bookmaker lays odds of 2 to 1 on
A gamemnon and odds of 3 to 2 on Behemoth. What are the cor-
rect odds on C alypso, if the bookmaker does not give himself a
percentage?
171. P A RL I A M E NT S OL I T A I RE . H ere is an easy question
concerning the patience game P arliament (also called T ourna-
ment) .
T wo 52-card decks are shuffled together and 8 cards are dealt
face up. If not a single ace or king appears in this layout, the
game very probably cannot be won, so that it is quite proper to
pick up the cards, reshuffle, and try again. T hat is all you need to
know about P arliament to answer the question:
What are the odds that you will turn up at least one ace or
king in the first 8 cards?
A s the factorials involved are rather large, you need not ex-
pand them. Just show the formula and make an estimate of the
result.
172. E VE RY T H ROW A S T RA I GH T . A die has six faces,
numbered from 1 to 6. Your chance of throwing any given num-
ber, say 4, is %assuming the die to be honest. If you cast two
dice together, you have % chance of getting a 4. on the first, plus
% chance of a 4 on the other, or % chance of getting at least one
4. By increasing the number of dice you roll simultaneously, you
similarly increase your chances of rolling at least one 4. In fact, if
you roll six dice, your chances are 6X% or 1, which is certainty.
But what goes for 4 must go for any other number on the dice. It
is likewise a certainty that you will cast at least one 6, one 5, and
so on. In fact, any roll of six dice together must result in a straight
. . . help! What's wrong here?
114 PROBLEMS OF PROBABILITY
173. T RE I Z E . In the French gambling game T reize, a deck
of 52 cards is shuffled and the cards are turned up one at a time.
A s they are turned, the dealer counts "One, two . . ." up to thir-
teen, then again from one to thirteen, and so on four times. If the
denomination of a card coincides with the number called, that
fact is a "hit." T he gamble enters by way of bets on how soon a
hit will be made, or on the failure to make any hit. T he ace is
counted as one; jack, queen and king respectively are eleven,
twelve and thirteen.
I have seen the game played as a solitaire. T he player counts
"A ce, two, three . . . jack, queen, king" four times as he turns the
cardsprovided that he is lucky enough to get that far; the soli-
taire is deemed to be won if the player gets through the entire
deck without a hit.
It would be easy to ask: What are the odds on winning
T reize solitaire? but I shall not do so. T he huge numbers in-
volved are not enticing. But suppose we simplify the problem. L et
us take only 6 cards, numbered from 1 to 6, shuffle them, and play
T reize. What are the odds we will get through the half-dozen
without a hit?
A s a matter of fact, the odds when the full deck is used are
not much different. T he change in odds as the number of cards is
increased over 6 is extremely slow.
I warn the reader that the problem is rather more involved
than may appear at first sight. It is easy to figure the total permu-
tations of 6 cardsfactorial 6. But when it comes to counting the
permutations in which no card occupies the ordinal position of its
own denomination, no simple formula can be evolved. In fact, the
task is like that of counting partitions: the practical method is to
construct a table for the required permutations out of n! as n takes
the values 2, 3, 4, 5, 6. E ach entry in the table is based on the
previous entries. In making the table the solver should discover
the principles whereby it can be extended infinitely; my object in
posing the problem is to give him this exercise in induction.
XI I . Number Games
174. M A T H E M A T I C A L GA M E S . T he very notion of a
"mathematical game" is a contradiction in terms. T he moment
every factor in a contest yields to precise mathematical calcula-
tion, it ceases to be a game. Yet history affords numerous exam-
ples of completely exhaustible "games" of undiminished vitality:
the schoolboys who played tit-tat-toe on the steps of the A cropolis
have their counterpart today in every land and every walk of life.
Good reason can be adduced why such games never die.
T hey are usually played as a challenge to the uninitiated to dis-
cover "the secret" of winning (or avoiding loss). T hus they en-
courage reflection and analysis, and are more apt to give instant
reward than many other activities of mind, because their "laws"
are not far to seek.
It might be expccted that study of those games whose merits
are wholly known might shed light upon others still in some de-
gree incalculable. Such is indeed the case. Games of a purely in-
tellectual character, e.g., checkers, are gradually being taken from
the domain of art into the domain of science. Some persons de-
plore this evolution, but why worry? T he human mind can devise
new games much faster than old ones are exhausted.
T he games I shall discuss are of two types, numerical and
tactical, dealing respectively with cardinal numbers and with rela-
tions of position. A ll are played between just two contestants. T he
focus of our inquiry is whether the first player (he who moves
first) or the second player must win with the best of play on both
sides, and what this best play is.
115
116 NUMBER GAMES
175. ONE P I L E . T his purely numerical game has been
traced back to remote antiquity, and probably it antedates the
games of position, such as tit-tat-toe.
A number of pebbles or counters of any description is massed
in one pile. T he two players draw alternately from the pile, the
object being to gain the last counter.
If it were permitted to seize the whole pile, the first player
would of course win; if the draw were limited to one counter at a
turn, the result would depend upon whether the number in the
pile were originally odd or even. T herefore, a minimum draw of
one counter is set, with a maximum greater than one.
Suppose the limits are i to 3 counters. T hen if a player finds
just 4 counters left in the pile, he loses. Whatever he takes, his
opponent can take the remainder. It is readily seen that the num-
ber 4 is a critical one because it is the sum of the minimum and
maximum limits of the draw.
In order to leave his opponent with 4 counters to draw from,
a player must previously have left him 8. Whether he then drew 1,
2, or 3, it was possible to reduce the pile to 4. E vidently the series
of winning combinations, each of which is a number to be left in
the pile for the opponent to draw from, is simply the multiples
of 4.
If we denote "a winning combination" by w, and the least
and most that may be drawn at a turn by a and m respectively,
then
w=(a-\-m) n
where n is any integer. T his formula is quite general, and is inde-
pendent of the number of counters originally in the pile. If this
number is of form w, the first player loses; if it is not, he wins by
reducing it to w.
176. T O L E A VE T H E L A S T . T he game can also be played
with the object of forcing one's opponent to take the last counter.
I leave it to the reader to write the formula for w in this case.
NUMBER GAMES 117
177. T O WI N T H E ODD. A more complex form of the one-
pile game puts an odd number of counters in the pile, fixes limits
on the draw, and gives victory to the player who owns the odd
number of counters after the common pile is exhausted. What is
the formula for w?
178. T H RE E -FI VE -S E VE N. Kindred games can be played
with more than one common pile of counters. A very widespread
variety is "3-5-7." T hree piles are set up consisting of counters in
these amounts. A t each turn a player may draw any number of
counters from one pilethe whole pile, if he wishesbut may not
iraw from more than one pile at a time. T he usual stipulation is
that he who must draw the last counter loses.
Which wins, first or second player, and how?
179. T H E T H I RT Y-ONE GA M E . In the smoking room of
the liner, Bill Green fell into conversation with a man who intro-
duced himself as Jack Smith. T he talk presently turned to mathe-
matical games, and Smith said he knew a puzzle that was rather
interesting. From a deck of cards he removed the A , 2, 3, 4, 5, 6
of each suit, and laid the 24 cards face up on the table.
"Now the idea," he said, "is that you turn over a card and
then I turn over a card, and so on. We add up the cards we turn
as we go along, and we can't go beyond 31. Whoever turns the
last card to make exactly 31 wins. T he ace, by the way, counts as
one."
Bill Green realized that there must be some mathematical
principle in which cards to turn, but he had no objection to try-
ing the game in fun. Several games were played, some won by
Smith and some by Green. Smith didn't seem to play by any par-
ticular system, but Bill noticed that whenever he made the total
24 he won. It dawned on Bill that 24 was a magic number, for,
having to add to it, a player could neither reach 31 nor prevent
his opponent from reaching it at his next turn. By the same
reasoning 17 was a magic number7 less than 24. In fact, the
118 NUMBER GAMES
wh o l e seri es of numbe r s ma d e by s ubt r ac t i ng 7' s f r o m 31 d o wn
we r e wi nner s .
Bi l l t ri ed out hi s t heor y by be g i nni ng wi t h a 3. S mi t h t ur ne d
a 5 a nd Bi l l a ns we r e d wi t h 2, t o ma k e 10. H e wo n t he g a me by
s t i cki ng t o t he seri es 3, 10, 17, 24.
S mi t h c ogi t a t e d ov e r t hi s resul t a nd r e ma r ke d, " I t hi nk I see
i t no w. I be t I wi n t he ne xt g a me . "
Fo r t hwi t h he b e g a n wi t h a 3, a nd Bi l l at r a n d o m t ur ne d a 6.
S mi t h t r i umpha nt l y chos e 2, a nd di dn' t s eem di s comf i t ed wh e n
Bi l l t ur ne d a not he r 6. Bu t he wa s i nde e d c r es t f al l en wh e n Bi l l
wo n agai n.
" I guess I ma d e a mi s t a ke , " h e sai d. " I t houg ht I h a d t he
h a n g of i t. A f t e r y o u t ook t he fi rst 6, l et me s e e o h of cour se, I
s houl d h a v e t a ke n a 5 ! "
Bi l l wa s i nc l i ned to di s put e thi s a nd t o e x pl a i n t he s i mpl e
f o r mu l a , but S mi t h c ut h i m of f .
" Do n ' t tel l me . I ' m sure I get i t no w. I ' l l bet y o u t hat I wi n
t he ne x t g a me ! "
Bi l l Gr e e n wa s t oo wa r y a bi r d t o be t wi t h a st r anger . St i l l ,
i t wa s hi s t ur n t o pl a y first, a nd S mi t h evi dent l y ha dn' t gr as ped
t he r eal i dea, a n d we l l , s o me ho w Bi l l a g r e e d t o a bet t hat was a
l i t t l e mo r e t ha n he c oul d a f f o r d t o ri sk.
Bi l l b e g a n wi t h a 3, a nd S mi t h t ur ne d a 4. T o st ay " i n t he
ser i es" Bi l l t ur ne d a no t he r 3 t o ma k e 10, a nd S mi t h t ur ne d an-
ot her 4 t o ma k e 14. T h e pl ay c ont i nue d i n t he s ame wa y , Smi t h
t ur ni ng onl y 4' s a nd Bi l l t ur ni ng 3' s to st ay i n t he seri es. Wh e n
S mi t h t ur ne d t he l ast 4, t he t ot al wa s 28. T h e 3' s be i ng al l gone,
Bi l l h a d t o t ur n a n A or 2, a nd hi s o ppo ne nt wo n.
Bi l l Gr e e n is not t he first " g u l l " wh o has f al l e n i nt o a s wi ndl e
t ha t is ho a r y wi t h age. T h e g a me of T hi r t y - one has been used to
fleece ma n y per sons wh o ha v e t he pr over bi al l y da ng e r ous " l i t t l e
k n o wl e d g e . "
T h e g a me is i nde e d a na l ogous t o No . 1 7 5 O n e Pile, a nd i t
is t r ue i n ge ne r a l t hat
w>=3i(a-\-m) n317n
Bu t he r e exi sts a n a d d e d f e at ur e , a l i mi t at i on o n t he numb e r of
NUMBER GAMES 119
t i mes t he s a me i nt eger ma y be c hos en. I t is not f eas i bl e t o sei ze
t he w seri es at onc e .
T h e first pl ay e r c a n never t hel es s f or c e a wi n. H o w ?
180. T H I RT Y-ONE WI T H DI C E . A c l e v e r v a r i a nt of t he
T hi r t y - o ne Ga me , i nv e nt e d by a c o nj ur e r , has pr o v e d baf f l i ng t o
ma n y " s ha r k s " w h o t ho ug ht t hey unde r s t ood al l a b o u t ma t he -
ma t i c a l game s .
T h i s v a r i a nt is pl a y e d wi t h a si ngl e di e. T h e s t ar t i ng n u mb e r
is fixed by a c ha nc e rol l of t he di e. T h e r e a f t e r e a c h pl a y e r gi ves
t he di e a qua r t e r - t ur n, i n a ny di r e c t i on he pl eases, t o br i ng a n e w
n u mb e r uppe r mos t . A r unni ng t ot al is k e pt of t he numbe r s so
t ur ne d up, a nd he wi ns w h o r e ac he s t he t ot al 31 or f or c es hi s
o ppo ne nt t o g o ov e r 31 .
T h e di e v a r i a nt is a c t ua l l y v e r y di f f e r e nt f r o m t he g a me wi t h
24 car ds . T h e r e , onl y f o ur dupl i c a t e s of e a c h di gi t exi st , b ut al l
unt ur ne d di gi t s a r e a v a i l a bl e t o t he pl ay e r i n hi s t ur n. H e r e , t he
n u mb e r of di gi t s is unl i mi t e d, but at e ve r y t ur n t he pl a y e r finds
t wo of t he m u n a v a i l a b l e t h e n u mb e r a l r e a dy up, a n d t ha t o n
t he oppos i t e f a c e .
T h e numbe r s on oppos i t e f ac e s t ot al 7, so t ha t t he pl a y e r has
a c hoi c e of onl y t wo pai r s of numb e r s out of t he c oupl e s 6 1 ,
5 2 , 4 3 .
Wh a t n u mb e r or numbe r s , t ur ne d u p by t he r a n d o m rol l ,
spel l vi c t or y f or t he first pl a y e r ? Wh a t is t he s ys t em wh e r e b y he
c a n pr es er ve hi s a d v a n t a g e a n d f or c e t he wi n ?
XI I I . Board Games
181. T H E P RI NC I P L E OF S YM M E T RY. T h e n u mb e r g a me s
di scussed i n t he pr e v i ous c ha pt e r poi nt t o a ge ne r al l a w wh i c h I
shall call the principle of symmetry.
T h e a ppl i c a t i ons of t he pr i nc i pl e t o s ome game s , e. g. , chess
a nd c hec ker s , ar e we l l k n o wn , b u t no wh e r e h a v e I s een t he pr i n-
c i pl e e nunc i a t e d as s uc h. Y e t it pe r v a de s mos t g a me s of a ma t he -
ma t i c a l na t ur e , b o t h nume r i c a l a nd t ac t i c al .
T h i s is t he pr i nc i pl e :
T h e pl a y e r w h o c a n pr es ent hi s o p p o ne nt wi t h a s y mme t r i c al
c o nf i g ur a t i o n t he r e by gai ns a n a d v a nt a g e . H i s o ppo ne nt has t o
di s t ur b t he s y mme t r y a nd t he r e by l eave t he first pl ay e r wi t h t he
opt i on ( a ) of r es t or i ng t he s y mme t r y by a c o g na t e mo v e , or
( b) i nc r e as i ng t he a s y mme t r y by a no t he r mo v e .
I n t hi s def i ni t i on, symmetry mus t be c ons t r ue d br oadl y t o
me a n a pe r f e c t c or r e s ponde nc e of pai r s of ent i t i es, whe t he r t hey
be ma g ni t ude s ( n u mb e r g a me s ) or poi nt s ( t ac t i c al g a me s ) .
L e t us r e v i e w br i ef l y s ome of t he be s t - k no wn appl i c at i ons of
t he pr i nc i pl e of s y mme t r y .
THREE-FIVE-SEVEN
We h a v e s een t hat a ny a r r a y is a w ( wi n n e r f or t he pl ay e r
w h o pr esent s it t o hi s o p p o ne nt ) if e ve r y c o mp o n e n t p o we r of 2 is
r e pr e s e nt e d a n even n u mb e r of t i mes. S u c h a n ar r ay is sym-
me t r i c a l i n t he sense de f i ne d a bov e . A d h e r e n c e t o t he pr i nc i pl e
120
BOARDGAMES
121
is t he one a nd onl y r equi s i t e; t he pl a y e r w h o has a c hi e v e d a w
c ont i nues t o rest ore t he s y mme t r y at e ve r y t ur n unt i l hi s l ast ,
wh e n a n uns y mme t r i c a l pl a y c l i nc he s t he v i c t or y .
CHECKERS
T h e pr i nc i pl e of s y mme t r y oper at es i n wh a t is c a l l e d " t h e
mo v e . " T h e basi c pr opos i t i on of t he mo v e is s h o wn i n Fi g . i . T h e
c onf i g ur a t i on is s y mme t r i c a l ; t he pl a y e r who s e t ur n i t is t o mo v e
is, a c c o r di ng t o pr i nc i pl e , at a di s a dv a nt a g e . F o r Bl a c k , t hi s di s-
a d v a nt a g e wo u l d be f a t a l : c o mpe l l e d t o r et r eat , he wi l l be dr i v e n
t o t he si de of t he b o a r d a nd c a pt ur e d. F o r Wh i t e , t he di s ad-
v a nt a g e wo u l d be t ha t he c a nno t wi n : he i s s a v e d f r o m loss by a
pe c ul i ar i t y of t he a r e na , t he do ubl e c or ne r .
T h e r ul e - o f - t humb f or c a l c ul a t i ng t he mo v e s hows c l ear l y t he
do mi na t i o n of t he pr i nc i pl e of s y mme t r y : C o u n t al l t he pi eces i n
one system or t he ot he r ( wh e n Bl a c k a nd Wh i t e e a c h h a v e t he
s a me n u mb e r of pi e c e s ) . I f t he t ot al is e v e n, t he pl a y e r who s e
t ur n it is t o mo v e has " t h e mo v e " agai ns t hi m. I f t he t ot al is o dd,
t he pl ay e r who s e t ur n i t i s t o mo v e ha s " t h e mo v e . "

111

81
l i t

1
i t
e
i l l
^lllb
mn

j j

mm
ill
FIG. I
Black to move loses
Whi te to move draws
122 BOARDGAMES
( By a system is me a n t t he 16 pl a y i ng s quar es of t he boa r d
t hat l i e o n f o ur a l t e r na t e fi l es, e. g. , t he squar es 5 1 3 2 1 2 9 ,
6 1 4 2 2 3 0 , 7 1 5 2 3 3 1 , 8 1 6 2 4 3 2 . T h e t wo sys-
t ems a r e c onv e nt i ona l l y de s i gna t e d as Bl a c k a nd Wh i t e a c c o r di ng
t o t he si de of t he b o a r d on wh i c h t he s quar es a but . )
" T h e mo v e " is not t he wh o l e of c he c ke r s , a l t h o u g h i t is c on-
t r ol l i ng i n mu c h of e nd- pl a y . Ot h e r f ac t or s al so ope r at e , gen-
e r a t e d by ( a ) t he c ha r a c t e r of t he a r e na ( a s y mme t r y of a d j a c e nt
cor ner s , one b e i ng do ub l e a n d one b e i ng s i ng l e ) , a nd ( b)
d y na mi c s of t he pi eces ( a n o r di na r y cut c ha ng e s t he mo v e but
not all c onf i g ur a t i ons of c a pt ur e d o s o ) .
I t ma y be sai d t hat t he vi t al i t y of a ny ma t h e ma t i c a l g a me
de pe nds u p o n s uc hl i ke pec ul i ar i t i es of a r e na a nd dy na mi c s
wh e r e b y t he t y r a nny of t he pr i nc i pl e of s y mme t r y is t o s ome de-
g r e e c i r c u mv e n t e d !
CHESS
T h e pr i nc i pl e of s y mme t r y ope r at e s i n wh a t is c al l ed " t he
oppos i t i on. " T h e bas i c pos i t i on is s h o wn i n Fi g. 2. T h e rul e f or
Fio. a
Black to move loses
Whi te to play, Black draws
BOARDGAMES 123
t he oppos i t i on of ki ngs is t hat t he ki ngs ar e oppos e d w h e n t hey
s t and o n s quar es of t he s a me c ol or a n d a n o d d n u mb e r of r anks
or fi l es i nt e r v e ne b e t we e n t he m. I n t he " n e a r o ppo s i t i o n" s h o wn
i n t he di a g r a m, t he pl a y e r who s e t ur n i t is t o mo v e f i nds t he
oppos i t i on agai ns t h i m: hi s k i ng mus t g i v e wa y . T h e e n e my k i ng
t he n has c hoi c e of r e s umi ng t he oppos i t i on or of ma k i n g t he
" pa s s i ng mo v e . "
T h u s , wi t h Bl a c k t o mo v e : i K d 7 ; 2 K f 6 . T h e Wh i t e
k i ng passes a n d t hr eat ens t o r e a c h f 7 , wh e r e u p o n Bl a c k wi l l no
l onge r be abl e t o bl oc k t he a d v a n c e of t he p a wn . Bl a c k ha s t o
pl ay 2 K e 8 , but t he n wi t h 3 K e 6 Wh i t e a g a i n sei zes t he op-
posi t i on. A g a i n Bl a c k ha s t o g i v e wa y , a n d Wh i t e wi l l t he n be
a bl e t o pass t o f 7 or dy .
I f Wh i t e t o mo v e , he c a n n o t by k i n g mo v e s a l one c o mp e l
Bl a c k t o gi v e wa y . T h u s 1 K f 5 , K f 7 . Bl a c k s i mpl y ma i nt a i ns t he
oppos i t i on. T o c h a n g e t he mo v e , Wh i t e t ri es 2 P e 5 - B u t n o w,
o wi n g t o a pe c ul i ar i t y of dy na mi c s , t he oppos i t i on is i ns uf f i c i ent
t o wi n f or Whi t e . T h e Bl a c k k i ng is i nde e d f o r c e d b a c k : 2 K e 7 ;
3 P e 6 , K e 8 ! ( K f 8 l oses) ; 4 K f 6 , K f 8 ; 5 P e 7 c he c k , K e 8 ;
6- Ke 6. Bl a c k has n o l egal mo v e , is " s t a l e ma t e d . " By t he r ul es
of t he g a me a s t a l e ma t e is a d r a w.
A s i n c hec ker s , t he oppos i t i on us ual l y be c o me s of i mp o r t a n c e
onl y i n e nd- pl a y . I n chess t he pr i nc i pl e of s y mme t r y e nj oy s mu c h
less do mi na nc e , o wi n g t o t he mu c h g r e a t e r d y n a mi c c o mpl e x i t y
of t he pi eces.
182. S A M L OYD'S DA I S Y P U Z Z L E . A n e l e g a nt a ppl i c a -
t i on of t he pr i nc i pl e of s y mme t r y wa s p r o p o u n d e d by S a m L o y d
i n hi s Daisy Puzzle. A s he tel l s t he st or y, he wa s one d a y t our i ng
i n t he Swi ss A l p s a nd c a me u p o n " a l i t t l e pe a s a nt gi rl g a t he r i ng
dai si es. T h i n k i n g t o a mu s e t he c hi l d, I s ho we d he r h o w t o pr og -
nos t i cat e he r ma t r i mo ni a l f ut ur e , by p l u c k i ng of f t he pet al s of t he
flower. S he sai d t he spor t wa s we l l k n o wn t o t he c o unt r y l assi es,
wi t h t he sl i ght di f f e r e nc e t hat a pl a y e r wa s a l wa y s at l i ber t y t o
pl uc k a si ngl e pet al or a ny t wo c ont i g uous ones, so t ha t t he g a me
wo u l d c ont i nue b y si ngl es or doubl e s unt i l t he v i c t or i ous one t ook
124 BOARDGAMES
t he l ast l eaf a n d l e f t t he ' s t ump' c al l e d t he ' ol d ma i d' wi t h y our
o ppo ne nt .
" T o o ur i nt ens e a s t oni s hme nt t he pr et t y ma d c h e n , wh o c oul d
not h a v e b e e n mo r e t ha n t e n year s ol d, v a nqui s he d o ur ent i r e
pa r t y by wi n n i n g e v e r y g a me , no ma t t e r wh o pl a y e d f i rst . "
Na t ur a l l y , t he ma i d mus t h a v e d e p e nd e d , i n one cas e or t he
ot her , u p o n he r oppone nt ' s i g no r a nc e of t he g a me ! T h e ques t i on
is, wh i c h pl a y e r c a n f or c e a wi n, t he fi rst or t he s e c ond? I shal l
not spoi l t he r e ade r ' s e n j o y me n t of t hi s be a ut i f ul puz z l e by g i v i ng
t he a ns we r her e.
T h e dai s y i l l us t r at ed by L o y d , i t mus t be s t at ed, has t hi r t een
pet al s .
183. DU DE NE Y'S C I GA R P U Z Z L E . I t wo u l d be a wo n-
de r f ul t hi ng t o h a v e a book of puzzl es , al l of wh i c h c oul d be
s ol ved by " c o mmo n sense, " r e qui r i ng n o k n o wl e d g e of f or ma l
ma t he ma t i c s .
B u t t hat bo o k pr o ba bl y ne v e r wi l l be wr i t t e n.
T h e t r oubl e is t hat , wh e n y o u f or e s we a r al l k no wl e d g e t hat
c o me s t h r o u g h s c hool i ng, t her e is pr e c i ous l i t t l e l ef t i n t he pr i st i ne
mi n d o n wh i c h t o bas e a puz z l e.
T h r o u g h t he ages , a s mal l st ore of " c o mmo n s ens e" puzzl es
has a c c u mu l a t e d . T h e pr i z e of t he c ol l ec t i on, i n my opi ni on, is
Du d e n e y ' s C i g a r P uz z l e. I c a nno t resist na r r a t i ng t hat I fi rst be-
c a me a c q u a i n t e d wi t h t hi s g e m whi l e r e a di ng i n be d, a nd t hat
l i ke a n i l l ust r i ous pr e c ur s or I s t ar t l ed t he hous e hol d by j u mp i n g
o ut of be d, d a n c i n g a b o ut a n d c r y i ng " E u r e k a ! E u r e k a ! "
H e r e i s t he puz z l e :
" T w o me n ar e s eat ed at a s qua r e - t o ppe d t abl e. O n e pl aces
a n o r di na r y c i g a r ( f l at at one e nd, po i nt e d a t t he ot he r ) on t he
t abl e, t he n t he ot he r does t he s ame , a nd so on al t er nat el y, a con-
di t i on b e i ng t hat no c i g a r shal l t o uc h a not he r . Wh i c h pl ay e r
s houl d s uc c e e d i n pl a c i ng t he l ast c i gar , as s umi ng t ha t e a c h wi l l
pl ay i n t he best possi bl e ma n n e r ? "
T h e s uppl y of c i gar s is s uppos e d t o be i nexhaus t i bl e, a nd
t he c i gar s a r e s uppos e d t o be u n i f o r m i n size a nd s hape.
BOARDGAMES 125
184. T H E C A RP A T H I A N S P I DE R. T h e C a r p a t h i a n s pi de r
has o dd habi t s. H a v i n g bui l t a we b , it ret i res l i ke a n y no r ma l
s pi der t o a pl a c e of c o nc e a l me nt . Bu t wh e n a fly be c o me s en-
me s he d, t he C a r p a t h i a n s pi der does not at o nc e r u n t o deal its
l et hal s t i ng t o its pr ey. I t a dv a nc e s onl y a l i t t l e di s t anc e , t he n
st ops a nd scrut i ni zes its pr e y t o es t i mat e wh e t h e r t he r e is d a ng e r
t hat i t wi l l es c ape. T h e n i t a dv a nc e s a g a i n a nd st ops a g a i n, t hus
pr o c e e di ng by st ages t o t he ki l l . P e r ha ps a pa r t of its mo t i v a t i o n
is a sadi st i c del i ght i n a ppl y i ng t he t or t ur e of suspense t o its
vi c t i m.
126 BOARDGAMES
T h e f o l l o wi ng a c c o unt of a bat t l e b e t we e n a C a r p a t h i a n
s pi der a n d a fly wa s na r r a t e d by a wi t nes s i n w h o m I ha v e t he
ut mos t c onf i de nc e .
T h e s c ene at t he out set of t hi s r e ma r k a bl e f r a y is s ho wn i n
t he i l l us t r at i on. T h e s pi der wa s l ur ki ng at a n out e r mos t poi nt of
hi s we b wh e n a fly l i t u p o n t he c e nt r al poi nt . I mme di a t e l y t he
s pi der a d v a n c e d a l o n g t he r adi al s t r and, g o i ng onl y as f a r as its
i nt er s ec t i on wi t h t he ne x t t r ans ver s e s t r and. Wh e n it s t opped, t he
fly, no do ubt s pur r e d wi t h n e w hope , s uc c e e de d i n c r a wl i ng a wa y
f r o m t he c e nt e r as f a r as t he ne x t i nt er s ect i on on a r adi al s t r and.
T h e s pi der t he r e up o n ma d e a s e c ond a d v a nc e , whi l e t he fly,
pa r a l y z e d wi t h f e ar , s t ayed mot i onl es s .
A n d t hus t he c ha s e pr oc e e de d. E a c h t i me t he s pi der mo v e d
t he fly f r oz e, a n d e a c h t i me t he s pi der s t oppe d t he fly c r a wl e d
f ur t he r a wa y . T h e f l y e v i de nt l y c o ul d not rel ease hi ms el f whol l y
f r o m t he we b , b u t c oul d ma n a g e t o b o g a l o ng t he st rands. T h e
s t r ange pa r t of t he st or y is t hat e a c h s e par at e mo v e by ei t her
i nsect t ook i t a l o n g a s t r and of t he we b f r o m one i nt er sect i on t o
t he n e x t n e v e r mo r e .
I t s e e me d t o t he wi t nes s t ha t t he s pi der wo u l d ne ve r c a t c h
t he fly by s uc h di l at or y t act i cs. Bu t he wa s wr o ng . E v e nt ua l l y t he
s pi der c o r ne r e d t he fly a nd wa s a bl e t o mo v e t o t he s ame poi nt
as its v i c t i m, wh i c h i t t he n de s pa t c he d wi t ho u t me r c y .
H o w di d t he s pi der wi n t hi s s t r ange g a v o t t e ? Re me mb e r t hat
i n t he s i t uat i on s h o wn by t he i l l us t r at i on t he s pi der ma d e t he fi rst
mo v e .
185. T I T -T A T -T OE . P r oba bl y t he mos t wi de l y - k no wn
" g a me " is T i t - T a t - T o e , or No u g h t s a nd C rosses. I t is f o u nd i n
e v e r y c i vi l i z ed c ount r y o n t he gl obe, a nd is k n o wn t o ha v e be e n
pl a y e d i n a nc i e nt t i mes. P r oba bl y 99 out of 100 s chool boys
i nves t i gat e i ts me a g r e pe r mut a t i ons a nd di s c over f or t hems el ves
t ha t wi t h best pl a y o n bo t h si des it is a f o r c e d d r a w.
I n g e n u o u s as is thi s pas t i me , a s t udy of j us t wh y c er t ai n
mo v e s l ose sheds l i ght u p o n t he st r at egy of c er t ai n r el at ed game s ,
g a me s not yet c ompl e t e l y a na l y z e d.
BOARDGAMES 127
I n Fi g. i ar e s hown t he t hree possi bl e openi ngs by t he first
pl ayer, C ross. T h e Noug ht s i n e a c h case s how t he onl y repl i es t hat
dr aw. T h e "si de ope ni ng " is evi dent l y weakes t , si .ice t here is
opt i on of t wo repl i es.
0 X
0
X
X 0 0
FIG. I
A n y ot her repl y by No u g h t to t he fi rst mo v e loses bec aus e it
permi t s C ross t o est abl i sh a forkthe t hr eat of ma k i ng 3- i n- a- r ow
on ei t her of t wo l i nes. No u g h t c a nnot bl ock bot h. N o w l et us
ma ke a n abst r act of t he condi t i ons unde r wh i c h a f or k c a n be
f or ced.
A b a nd o ni ng t he me c hani c s of T i t - T a t - T o e , I wi l l no w i ndi -
cat e t he poi nt s on whi c h mar ks c a n be wr i t t en ( or count er s
pl ac e d) by ci rcl es, c onnec t ed by l i nes t o s how avai l abl e wi nni ng
conf i gurat i ons. Fi g. 2 shows wha t I wi l l cal l t he critical triangle,
one of t he si mpl est conf i gur at i ons of poi nt s a nd l i nes on wh i c h
a f or k c a n be based. T h e t ri angl e is def i ned by t hree l i nes, e a c h
c ont ai ni ng t hree avai l abl e poi nt s of pl ay ( whe r e 3- i n- a- r ow wi ns ) ,
wi t h e ac h ver t ex of t he t r i angl e l yi ng on t wo of t he l i nes.
If C ross, t he first pl ayer , occupi es a ver t ex of a cri t i cal tri -
angl e, and if No u g ht t hen pl ays on some ot her par t of t he boar d,
C ross wi ns by pl ay i ng on a second ver t ex. No u g h t has t o bl ock
t he t hr eat on one l i ne, whe r e upo n C r os s t akes t he t hi r d ve r t e x
and f orks across t he ot her t wo l i nes.
128 BOARDGAMES
T h e r e f o r e , wh e n C r os s t akes t he fi rst v e r t e x , No u g h t mus t
i mme d i a t e l y pl a y o n t he t r i angl e t o a v o i d loss. Wh e t h e r he mus t
t ake a v e r t e x de pe nds u p o n t he l i mi t at i ons of t he s ys t em: by
hypot hes i s t her e is mo r e t o t he b o a r d, el se No u g h t c oul d not go
wr o n g . But , i t is i mpo r t a nt t o not e, t he st r ongest r epl y of No u g h t
is t o take a vertex. H e r e he c ut s t wo l i nes; el s ewher e he cut s but
one.
T h e " s i de o p e n i n g " in T i t - T a t - T o e t akes a v e r t e x of t wo
c r i t i c al t r i angl es , de pi c t e d i n Fi g. 3, whe r e t he bl ac k dot ma r k s
t he poi nt sei zed. I f t he t wo t r i angl es we r e who l l y di st i nct , e x c e pt
f or t he one c o mmo n poi nt , t he n C r os s c oul d wi n by f or ce, as
No u g h t c o u l d not pl a y i nt o bo t h at onc e. A s it is, t he t wo tri-
angl es he r e h a v e no less t h a n f o ur poi nt s i n c o mmo n , so t hat
s ei zur e of a ny one of t he r e ma i ni ng t hr ee f orest al l s a f or k. I n
ot he r wo r ds , a s uf f i ci ent r epl y t o t he " s i de o p e ni ng " is t o pl ay
i nt o e i t he r a d j a c e n t c or ne r or i nt o c ent er .
T h e " c o r ne r o p e n i n g " is mu c h s t r onge r ; t hi s poi nt is a v e r t e x
of s even cr i t i cal t r i angl es . F o u r of t h e m ar e s ho wn i n Fi g. 4 ; t he
ot her s a r e s y mme t r i c a l r ef l ect i ons of t he f i rst t hr ee. T h e r e ar e
mo r e t hr eat s t ha n c oul d be s c ot c he d i n one t ur n, we r e it not f or
the fact that the center point of the layout is common to all of
them. Se i z ur e of t he c e nt e r suf f i ces to d r a w. Wh i l e thi s poi nt is
FIG. 3
FI G. 4
BOARDGAMES 129
not a v e r t e x i n all t he cr i t i cal t r i angl es , a ny e f f or t by C r os s t o
ut i l i ze t r i angl es wh e r e i t is not is de f e a t e d b y t he l i mi t at i ons of
s pace. Go i n g b a c k t o Fi g . 2, we see t hat No u g h t c a n d r a w by t ak-
i ng a ny poi nt i n t he t r i angl e, e v e n no n- v e r t e x ; f o r t her e a r e onl y
t hr ee a v a i l a bl e l i nes i n al l , a nd No u g h t c a n ki l l e a c h i n t ur n.
T h e " c e nt e r o p e n i n g " is c l e ar l y st r ongest . T h e . c e nt e r poi nt
lies on e ve r y possi bl e cr i t i cal t r i angl e t hat c a n be pi c ke d out of
t he ni ne poi nt s. T o r epl y by s ei zi ng a si de is i ns uf f i ci ent . I n Fi g . 5
s uppos e No u g h t t akes poi nt 2. T h e n C r os s c a n t ake 3, f o r c i ng hi s
o ppo ne nt t o bl oc k at 7, a n d t he n a pl a y on ei t her 6 or 9 ma k e s
a f or k.
A si de poi nt l i es on onl y t wo l i nes of t he l a y o u t ; a c or ne r
poi nt , l y i ng on t hr ee l i nes, mus t be s uper i or . Sei z ur e of a c or ne r
i n f ac t gi ves No u g h t a d r a w. I n Fi g . 5 l et No u g h t ' s r epl y be t o
t ake poi nt 1. If C r os s pl ays 2 or 4 he exer t s onl y a " o n e - mo v e
t hr e a t , " si nce one of t he t wo l i nes r a di a t i ng t he r e f r o m is a l r e a dy
ki l l ed by 1. If he pl ays u p o n 3, 6, 7, or 8 he i nde e d exer t s a " t wo -
mo v e t hr e a t " of f or k, but No u g h t t hr eat ens fi rst. H i s bl o c k i ng
a ns we r l i nes u p wi t h 1 a nd c o mpe l s C r os s t o bl oc k i ns t ead of c o m-
pl e t i ng t he f or k, wh i c h No u g h t t hus gai ns t i me t o ki l l .
186. A N U NL I M I T E D BOA RD. F r o m t he f o r e g o i ng c on-
si der at i ons i n T i t - T a t - T o e we c a n d e d u c e t hat , i f t he pl a y i ng
f i el d is suf f i ci ent l y ext ens i ve, t he fi rst pl a y e r c a n a l wa y s f or c e
3- i n- a- r ow.
A n y c e nt r al l y - l oc at e d poi nt o n a l ar ge b o a r d has al l t he a d-
v a nt a g e s of t he c e nt e r poi nt i n T i t - T a t - T o e . Bu t mo r e e a c h l i ne
pas s i ng t hr o ug h i t is i n e f f e c t t wo l i nes, si nce f r o m t he poi nt t he r e
is s pac e a v a i l a bl e o n b o t h rays t o ma k e 3- i n- a - r ow. I t is necessar i l y
130 BOARDGAMES
possi bl e t o des c r i be t wo cr i t i cal t r i angl es h a v i n g onl y one c o mmo n
poi nt , a n d No u g h t ( t he s e c ond pl a y e r ) c a nno t bl oc k i n bo t h tri -
angl es a t onc e . A l s o, C r os s ( t he first pl a y e r ) c a n de v e l o p t wo
t r i angl es t ha t o v e r l a p t o a c ons i de r a bl e e xt e nt , wi t ho u t l et t i ng
No u g h t de v e l o p s uc h a c ount e r - t hr e a t as saves h i m i n T i t - T a t -
T o e .
I t ma y we l l be s us pec t ed t hat , on a b o a r d of unl i mi t e d ext ent ,
C r os s c a n est abl i sh by f or c e a row of more than 3. S u c h is ac t ual l y
t he case. A t l east 5 - i n- a - r ow c a n be f or c e d.
H e r e we h a v e t o r e c kon wi t h a c onf i g ur a t i on e v e n s i mpl er
t h a n t he cr i t i cal t r i a n g l e t h e cr i t i cal l i ne.
Suppo s e t he g a me is t o ma k e 5- i n- a- r ow. Suppos e t hat C r os s
get s 4- i n- a - r ow, a nd t hat t he poi nt s at e a c h e nd of thi s l i ne ar e
o p e n ( u n o c c u p i e d ) . T h e n he mus t wi n, si nce No u g h t c a nnot
bl oc k bo t h e nds of t he l i ne at onc e . T h u s , wh a t we wi l l cal l a n
" o p e n 4 " is a cr i t i cal l i ne.
BOARDGAMES 131
O n a n unl i mi t e d boa r d, t he f i rst pl a y e r c a n a l wa y s c r e a t e a
f or k wi t h t wo ( s uf f i ci ent l y) " o p e n 3s " a n d t he r e by ma k e a wi n-
ni ng " o p e n 4 . " On e of ma n y possi bl e v a r i a t i ons is s ho wn i n t he
di a g r a m. T h e mo v e s of e a c h pl a y e r a r e n u mb e r e d f r o m 1 up, t he
fi rst pl ay e r be i ng " wh i t e . "
I f t he pl a y k f o l l o we d mo v e by mo v e , i t is s een t ha t t he
s econd pl ayer , i n hi s fi rst t hr e e t urns, tri es t o c ut as ma n y l i nes as
possi bl e a mo n g t hose r a di a t i ng f r o m t he poi nt s sei zed b y t he f i rst.
A t hi s f o ur t h t ur n a nd t he r e a f t e r he is k e pt bus y b l o c k i ng o p e n
3s; hi s onl y opt i on is wh i c h e nd t o bl oc k. De s pi t e hi s ef f or t s t o
c o mbi ne def ens e wi t h c ount e r - t hr e at , t he first pl a y e r ha s est ab-
l i shed a f or k of t wo 3s by 9 6 3 a nd 9 4 8 .
T h e pr es ent st at e of k n o wl e d g e l eaves o p e n t he que s t i on:
C a n t he first pl a y e r f or c e 6- i n- a - r ow? F o r pur pos e s of ma k i n g a n
i nt er es t i ng ( a nd pl a y a bl e ) g a me , h o we v e r , be t t e r t h a n e x t e ndi ng
t he wi nni ng r o w is t o he dg e t he b o a r d or t he r ul es wi t h c e r t a i n
l i mi t at i ons , as is done i n t he ne xt t wo g a me s des c r i bed.
187. GO M A KU . T h e or i ent al G O b o a r d is a gr i d of
1 9 X 1 9 l i nes. St ones ar e pl a y e d u p o n t he i nt er sect i ons, not i n t he
squares. H e n c e t he field c ompr i s es 381 poi nt s. T w o di f f e r e nt
g a me s ar e pl a y e d o n t hi s boa r d. O n e is t he g a me G O , a pr o-
f o und t act i cal g a me at l east as c o mp l e x as chess b ut wi t h pr a c -
t i cal l y n o l i t er at ur e. T h e ot he r is G o M a k u , al so c a l l e d G o Ba n g ,
t he obj e c t of wh i c h is s i mpl y t o est abl i sh 5 - i n- a - r ow by t he
al t er nat e l a y i ng d o wn of st ones.
Si nc e, as is s h o wn i n t he p r e c e di ng s ect i on, t he first pl a y e r
c a n a l wa y s f or c e a wi n, G o M a k u has t o e mpl o y a n ar bi t r ar y
r ul e : A pl a y e r ma y not so pl a c e a st one as t o est abl i sh a f or k of
t wo " o p e n 3s . " H e ma y , h o we v e r , f or k a n o pe n 3 wi t h a one -
e nde d 4, or ma k e a do ubl y - o pe n 4.
T h r o u g h t he ope r a t i on of t he r ul e, a pl a y e r o f t e n finds hi m-
self i n t he e x a s pe r a t i ng s i t uat i on of b e i ng una bl e t o pl a y on a ke y
s quar e, a nd of l os i ng t he f r ui t of pr o l o ng e d ef f or t s . Wh e t h e r t he
first pl ay e r c a n f or c e a wi n is u n k n o wn h e n c e G o M a k u is still
a real g a me ! I t is my bel i ef t hat t he g a me is a d r a w wi t h best pl a y
132 BOARDGAMES
o n b o t h si des. T h e s e c ond pl a y e r c a n c er t ai nl y p u t u p a pr ol ong e d
de f e ns e if he is wi l l i ng t o f o r e g o a ny a t t e mp t t o wi n.
A t al l event s , G o M a k u of f er s a f a s c i na t i ng c ha l l e ng e t o
pl a y e r a nd t heor i s t al i ke.
188. T H E M I L L . T h e M i l l , al so c al l e d Ni n e M e n ' s M or r i s ,
is a n a nc i e nt g a me t hat ma y ha v e h a d a c o mmo n or i gi n wi t h
T i t - T a t - T o e . T h e bo a r d, s h o wn i n t he i l l ust r at i on, compr i s es 24
poi nt s , c o nne c t e d by 3 c onc e nt r i c s quar es a nd 4 t r ans ver s al s . *
E a c h pl a y e r is p r o v i de d wi t h 9 c ount er s , of di s t i nct i ve col or . E a c h
i n t ur n l ays a st one o n o ne of t he poi nt s ( l i ne i nt er sect i ons) unt i l
al l 18 st ones h a v e b e e n pl a y e d. T h e n e a c h i n t ur n mov e s one of
hi s st ones a l o n g a ny l i ne o n wh i c h it st ands, t o t he ne x t a dj a c e nt
poi nt .
T h e o b j e c t of pl ay is t o de c i ma t e t he adve r s e a r my . E a c h
t i me a pl a y e r est abl i shes 3 st ones o n a ny l i ne of t he boa r d, c al l ed
a mill, he is ent i t l ed t o r e mo v e a n adve r s e s t one f r o m t he boa r d,
wi t h t he pr ov i s o t hat he ma y not t ake one f r o m a n adver s e mi l l .
On c e a mi l l is es t abl i s hed, t he o wn e r ma y " o p e n " i t by mo v i n g one
st one of f t he c o mmo n l i ne, t he n " c l o s e " i t by mo v i n g t he st one
ba c k. E v e n t h o u g h c o mp o s e d of t he s a me 3 st ones, thi s f o r ma -
t i on n o w c ount s as a n e w mi l l a nd ent i t l es t he o wne r t o r e mo v e
a n e n e my pi ec e.
I n s ome c ount r i es , wh e n a pl ay e r is r e duc e d t o 3 st ones t hey
b e c o me " wi l d " a n d ma y b e mo v e d f r o m poi nt t o poi nt r egar dl ess
of c o nne c t i o n by l i ne. T h i s flourish is l i t t l e mo r e t ha n a sol ace
f o r t he v a nq ui s he d, f o r a s uper i or f or c e wi l l us ual l y wi n agai ns t
A not her version of the board connects each triplet of corners by a
diagonal line. Whether these lines should be added has long been a matter
of controversy. A ngel o L ewi s (P rofessor H of f man) , writing in 1894, said
that the pattern with the diagonals "is preferred by some players, though
the addition is stoutly resisted by the champions of the original game. "
On the other hand, H . E . Dudeney says ( 1 91 7) "Sometimes the diagonal
lines are omitted, but this evidently was not intended to affect the pl ay:
it simply meant that the angles alone were thought sufficient to indicate
the points. " T he present writer holds to the no-diagonal school; for one
reason, this is the style in the Scandinavian countries, where the mill game
is as commonly played as checkers.
BOARDGAMES 133
FIG. IT h e M il l Board
t he 3. Of course wh e n a pl ayer is r educ ed t o 2 stones t he g a me
is over.
A s a soci al t wo- ha nde d pas t i me, M i l l is one of t he best of
games. I n oppor t uni t y f or skill it st ands be t we e n T i t - T a t - T o e a nd
Go M a k u . H a v i n g f e we r possi bi l i ti es t han t he l at t er, i t ma ke s
mu c h less de ma nd on t he pl ayer ' s power s of anal ysi s a nd vi sual i za-
t i on. Sti l l , i t has not be e n exhaus t ed.
A s a s ubj ect of ma t he ma t i c a l i nqui r y, M i l l is a f as c i nat i ng
c ombi nat i on of i deas. T h e basi c 3- i n- a- r ow pr i nci pl e, c ompl e t e l y
exhaus t ed i n T i t - T a t - T o e , is gi v e n vi t al i t y i n t wo ways , ( a ) T h e
boar d is s ome wha t enl ar ged, but is still so c onf i ned t hat t he pl ayer
cont i nual l y f eel s " Wi t h a l i ttl e t houg ht I c oul d anal yz e thi s g a me
134 BOARDGAMES
c o mp l e t e l y ! " ( b ) A g a me of p l a c e me nt is c o mb i n e d wi t h a g a me
of mo v e me n t . I t is n o l ong e r s uf f i ci ent t o be t he fi rst t o ma k e
3- i n- a - r ow.
I n Fi g . 2 is s h o wn a g a me t ha t has b e e n pl a y e d ma n y t i mes.
I t i l l ust r at es h o w da ng e r o us is a l i t t l e l e ar ni ng. Whi t e , t he first
pl a y e r , c or r e c t l y oc c upi e s a f o u r wa y i nt er s ect i on. Bl a c k sei zes
a no t he r f o u r wa y poi nt . Wh i t e has s t udi ed t he pr oper t i es of t he
b o a r d j us t e n o u g h t o k n o w t hat Bl a c k' s f a i l ur e t o pl ay ne xt t o
t he f i rst Wh i t e st one l et s Wh i t e est abl i sh a f or k. H e pr oc eeds t o
d o so. H i s pl ays f r o m 2 t o 5 k e e p Bl a c k bus y i nt er c ept i ng, a nd
t he 6t h Wh i t e s t one ma k e s a f or k. Bl a c k bl oc ks at 6 a n d Wh i t e
BOARDGAMES 135
ma ke s hi s mi l l . Si nc e Bl ac k' s 6 a nd i t hr e a t e n t o ma k e a mi l l ,
Wh i t e r e mov e s one of t hese st ones, say 6. Bl ac k' s 7 t h r epl ac es t hi s
loss, a nd c ompe l s Wh i t e t o bl oc k a t 8. Bl a c k t he n has t o c ut a t 8.
L o o k at t he pos i t i on no w. No t one of t he Wh i t e st ones c a n
mo v e . Bl ac k' s 7 st ones ho l d al l of Whi t e ' s 8. T h e g t h st one o n
e a c h si de r e mai ns t o be pl a c e d. Wh e r e v e r Wh i t e el ect s, Bl a c k wi l l
pl a c e hi s g t h a dj a c e nt . Wh i t e wi l l h a v e one mo v e at mos t , whe r e -
u p o n Bl a c k c a n t r a p t hi s l ast pi e c e . Wh i t e l oses, be c a us e i n hi s
t ur n t o pl ay he has no mo v e .
I t does not f o l l o w f r o m t hi s i ns t anc e t ha t t o f or c e a mi l l i n
l ayi ng d o wn , wh e n o ppo r t uni t y ari ses, is necessar i l y f a t a l . O n t he
c ont r ar y, t he pl a y e r w h o ma k e s t he first mi l l whi l e still ma i n-
t ai ni ng f r e e d o m of mo v e me n t is l i kel y t o wi n.
I ha v e pr o v e d t o my s at i s f act i on t hat t he g a me is a f o r c e d
d r a w if nei t her pl a y e r t akes ri sks t o wi n. Bu t t her e is r e ma r k a b l e
s cope f or i nge nui t y i n s et t i ng t r aps .
189. S A L VO. A n e xc e l l e nt t wo - h a n d e d g a me t hat c o m-
bi nes l uc k a nd ski l l i n s oci abl e pr opor t i ons is Sa l v o.
E a c h pl ay e r uses a penc i l a n d a sheet of " q u a d r i l l e " p a p e r
( l i ned bo t h ver t i c al l y a nd ho r i z o nt a l l y ) . E a c h c o mme n c e s by out -
l i ni ng t wo s quar es i o X ' O as s h o wn i n t he d i a g r a m. T h e c o l umns
a nd r ows ar e de s i gna t e d by di f f e r e nt i ndi ces . T h e l e f t s qua r e is
t he pl ayer ' s " o w n " ar e na, wh i l e t he ot he r is hi s oppone nt ' s . I n
hi s o wn a r e na t he pl a y e r pl aces , wh e r e v e r he chooses, f o ur
" s hi ps . " E a c h shi p is a seri es of s quar es b l a c k e d out , i n a l i ne
ver t i cal l y, l at er al l y, o r di a g ona l l y . T h e " f l e e t " consi st s of one
" ba t t l e s hi p" ( 5 s qua r e s ) , one " c r ui s e r " ( 3 s qua r e s ) , a n d t wo
" de s t r oy e r s " ( 2 s quar es e a c h ) .
Bo t h fleets b e i ng es t abl i s hed, t he pl ayer s i n t ur n fire " s a l v o s "
i nt o t he e ne my fleet i n a n e f f or t t o si nk i t. A t t he out s et , e a c h
s al vo compr i s es 7 shot s. T h e pl a y e r ma r k s 7 of t he s quar es o n hi s
e ne my a r e na at wh i c h he choos es t o s hoot , a n d na me s t h e m t o hi s
oppone nt , e. g. , i n t he d i a g r a m t he first s al vo wa s on s quar es D2 ,
G3 , E 4, H 5 , et c. T h e o p p o ne nt ma r k s t hese shot s i n hi s o wn
ar e na. A l l shot s ar e ma r k e d by t he n u mb e r of t he s al vo, t he sal vos
136
BOARDGAMES
b e i ng n u mb e r e d c ons e c ut i ve l y f r o m i u p so t hat t he pl ay e r c a n
k e e p t r ac k of t he t i me at wh i c h hi t s we r e s cor ed.
O n c o mpl e t i o n of a s al vo, t he pl a y e r fired at a nnounc e s
wh e t h e r a ny of t he shot s h a v e s c or ed a hi t o n hi s fleet. H e na me s
t he t ype of vessel hi t , b u t not t he s quar e. H i s o ppo ne nt of cour s e
ma k e s a me mo r a n d u m of a ny hi t s so a nno unc e d.
A B C D E I F G H I J A B C D E F G H I J
2
3
4
5
6
7
8
9
10
A
/
/
/
1
/
l
/
1
2
3
4
5
6
7
8
9
10
T h e o b j e c t of pl a y is t o si nk t he e n e my fleet ent i r e by hi t t i ng
e v e r y c o mp o n e n t s quar e . Wh e n a ny vessel is s unk, t he owne r ' s
q u o t a of shot s i n al l hi s s ubs e que nt sal vos is r e duc e d by 3 f or t he
bat t l e s hi p, by 2 f o r t he cr ui s er , or by 1 f or a dest r oyer .
T h e r e is a p r e mi u m u p o n l oc a t i ng t he e n e my bat t l es hi p a nd
c o nc e nt r a t i ng s al vos t o si nk it i n a hur r y . O n e p o p u l a r me t ho d of
c o mme n c i n g is t o qua r t e r t he e ne my b o a r d a nd put t he first f o ur
s al vos i n t he quar t er s . A na t ur a l f o r ma t i o n of shot s f or t he ear l y
sal vos is s h o wn i n t he d i a g r a m, a ki nd of chess kni g ht t our . T h e
i de a of it is t o put one s hot i n e a c h of s ever al a d j a c e n t l i nes c ov-
e r i ng g r oups of l i nes i n al l t hr ee di r ect i ons .
T h e que s t i on suggest s i t s el f : I f y o ur o p p o ne nt is k n o wn t o use
t hi s f o r ma t i o n o n hi s first s al vo, wh e r e is t he best pl a c e t o put
y o u r ba t t l e s hi p so as t o e s c ape a first-round hi t ? A l s o, whe r e is
t he wor s t p l a c e ? I nt ui t i on ma y s uppl y t he ans wer s , but gi v e a
r i gor ous ma t h e ma t i c a l de mons t r a t i on.
P art T wo
S OL U T I ONS
S olutions
( H e r e ar e t he Sol ut i ons t o t he P uzzl es i n P a r t O n e of t hi s
book. Y o u wi l l not e t hat t he numbe r s of t he Sol ut i ons c or r e s pond
t o t he numbe r s of t he P uzzl es . )
1. H OW H I GH I S A P OL E ? T h e pol e is 28 f e e t hi g h. T h e
pr o bl e m is s ol ved by s i mpl e pr opor t i on. T h e he i g ht of t he pol e
is t o its s ha do w, 21 , as t he ma n , 6, is t o hi s s ha do w, 4^2-
T h e we l l is a t ri f l e u nd e r 22 f eet de e p. I f t he d r u m i s 7 i nc he s
i n di a me t e r its c i r c umf e r e nc e is a n d it is a b o ut 3 . 1 4 1 6 . A r ope
t hat wr a ps 12 t i mes a r o u n d i t is 12 X 7 X S- H
1
^ i nc he s l o ng ;
c anc e l t he f a c t or 12 a nd y o u h a v e t he a ns we r i n f eet .
T h e a ns we r t o t he pr o b l e m of t he s he e p is not 60, b ut 5 5
sheep. I f it t akes 10 s heep 10 mi nut e s t o j u mp ov e r a f e n c e t h e
t i me be i ng me a s ur e d f r o m t he j u mp of t he first s he e p t o t he j u mp
of t he 1 0 t h t h e n t he i nt e r v al b e t we e n j u mp s is 1 0/9 of a mi nut e .
T h e r e a r e 60 - hi 0/9 or 5 4 s uc h i nt er val s i n a n ho ur , so t ha t 55
s heep j u mp t he f e nc e i n t hi s t i me.
2. DOM I NO S E T S . I n t he set u p t o do ub l e - t we l v e t her e
ar e 91 bones . T h e n u mb e r i n a ny set is t he s u m of t he i nt eger s
1 + 2 + 3 . . . - j - n, whe r e n is t he n u mb e r of sui t s. Do n ' t f or ge t
t ha t " b l a n k " is a sui t . T h u s , i n t he set u p t o doubl e - s i x t he r e ar e
s even sui t s: o, 1, 2, 3, 4, 5, 6.
139
140 SOLUTIONS
3. M A RK-DOWN. T h e s el l i ng pr i c e wa s $1 5. 36. T h e deal er
e a c h t i me d e d u c t e d 2 0 % of t he pr e v i ous pr i c e.
4. NI NE DOT S . T h e s ol ut i on is s ho wn i n t he di a g r a m.
H a s t y r e ade r s a r e a p t t o r e j e c t t hi s s ol ut i on, o nc e t hey h a v e f o u n d
i t , be c a us e t he hor i z ont a l l i ne i f e x t e nde d wo u l d cross t he do t i n
t he l o we r l e f t c o r ne r a s e c ond t i me. Bu t no t hi ng i n t he s t at e me nt
of t he pr o b l e m i mpl i e s t ha t t he l i ne mus t be c ons t r ue d as i nf i ni t e
i n l e ngt h. O n t he c ont r a r y , t he s t i pul at i on t hat t he penc i l mus t
not be l i f t e d i mpl i e s t hat we ar e de a l i ng wi t h finite l i ne s egment s .
5. M A KI NG A C H A I N. T h e cost is $1. 40. T h e mos t ec o-
no mi c a l pl a n is t o c ut o p e n al l f o u r l i nks of one sect i on a nd use
t hese f o u r t o j o i n t he five r e ma i ni ng sect i ons t oget her .
6. T H E WI L Y C H I E F. T h e M ' g mb wo r k s 1 4 days per
mo n t h , e a r ni ng 70 bmgs . Bu t he l oaf s o n t he ot he r 10 days , a nd
is d o c k e d j us t 70 bmg s .
SOLUTIONS 141
7. T H E BOOKWORM . I f y o u r a ns we r wa s i nc he s y o u
f el l i nt o t he t r ap. Wh e n t wo v o l ume s a r e i n o r de r ( l e f t t o r i ght )
on a books he l f , t he f i rst p a g e of V o l u me I a n d t he l ast p a g e of
V o l u me I I a r e s e par at e d onl y by t wo c over s . T h e c or r e c t a ns we r
i s J4 i nc h.
8. A N E A S Y M A GI C S QU A RE .
4 9 2
3 5 7
8 1 6
9. T H E FA C E T I OU S YOU NG M A N. We k n o w t ha t t he
t ot al of t he pur c has e s mus t be a n u mb e r di vi s i bl e b y 4 a nd t hat
i t is less t ha n a dol l ar . T h e onl y pai r s of di gi t s t ha t s u m t o 14 a r e
9, 5 a nd 8, 6 a nd 7, 7. T h e onl y i nt eger di vi s i bl e by 4 t ha t c a n
be f o r me d f r o m a ny pa i r is 68. H e n c e t he p a c k of Fu me r o l e s cost
17 cent s a nd t he Sur e - Fi r e l i ght er 51 cent s.
10. T A NKT OWN T RI OS . Vi s ua l i z e t he t hr e e out f i el der s
pl a c e d a t t hei r s e par at e t abl es. T h e n a b a s e ma n, s ay first bas e,
ma y sit a t a ny of 3 t abl es, a n d a f t e r he is pl a c e d t he s e c ond bas e-
ma n ma y sit at e i t he r of t he 2 ot he r t abl es. A f t e r t ha t t her e is
onl y one pl a c e l e f t f o r t he t hi r d b a s e ma n. F o r t he first si x me n
t her e ar e t hus 3 X 2 = 6 di f f e r e nt a r r a ng e me nt s . T h e pi t c he r ,
c at c he r , s hor t s t op c a n be s eat ed i n 6 wa y s , so t ha t t he t ot al f o r t he
ni ne is 6 X 6 = 3 6 . ( T h e s a me t r i o wi l l c o nv e ne 4 t i mes, b ut o n
e a c h oc c as i on t her e wi l l be a v a r i a t i o n i n t he c ompos i t i on of t he
ot he r t wo t abl es. )
11. WA T E R, GA S . A ND E L E C T RI C I T Y. T h e onl y sol u-
t i on is t o pe r s ua de o ne of t he hous e - owne r s t o pe r mi t a c o ndui t
s er vi ci ng a no t he r hous e t o be l ai d u n d e r n e a t h hi s o wn , as s h o wn
i n t he d i a g r a m o n p a g e 142. Na t ur a l l y t he s t a t e me nt of t he pr o b-
l e m mus t be c a r e f ul l y ma d e so as not t o e x c l ude t hi s possi bi l i t y,
el se t he s ol ut i on wh e n de mo ns t r a t e d wi l l rai se a j us t i f i ed pr ot es t .
142 SOLUTIONS
13. S P OT T I NG T H E C OU NT E RFE I T . Di v i d e t he coi ns a t
r a n d o m i nt o t hr ee g r o ups of t hr ee. Ba l a nc e a ny t wo gr oups agai ns t
e a c h o t h e r ; if one c ont a i ns t he u nd e r we i g ht c oi n t he g r o u p is
s po t t e d; if t he t wo g r oups ba l a nc e , t he c ount e r f e i t is i n t he t hi rd
g r o u p . F r o m t he s pot t ed g r o u p t ake a ny t wo coi ns a nd ba l a nc e
t he m. I f one is l i ght , i t is t he c o unt e r f e i t ; if t hey ba l a nc e , t he
c o unt e r f e i t is t he t hi r d of t he g r o up, l ef t o n t he t abl e.
14. T H E P A I NT E D C U BE . On l y one of t he 27 smal l c ubes
is u n p a i n t e d ; 8 ar e pa i nt e d o n t hr ee f ac es , 12 on t wo f aces, 6 on
one f a c e . T h e s e numbe r s c or r e s pond t o t he n u mb e r of corners,
edges , a nd f ac e s of t he l ar ge c ube .
15. S H E E P A ND GOA T S . T h e sol ut i ons ar e g i v e n by na m-
i ng t he pe ns f r o m wh i c h t he pa i r is t o be mo v e d ; it is unnec es s ar y
t o n a me t he pe ns t o wh i c h t he y ar e t r ans f er r ed as onl y t wo pens
ar e v a c a n t .
SOLUTIONS 143
T h e first puz z l e is s ol ved by mo v i n g 4 5 , 6 7 , 2 3 , 7 8 .
T h e s e c ond is s ol v e d by 3 - -4, fi7, 12, 78.
16. T H E BI L L I A RD H A NDI C A P . H u n t i n g d o n s houl d gi v e
M c C l i n t o c k 40 poi nt s i n 100. If he gi ves C h a d wi c k 20 i n 100,
t he n C h a d wi c k is e x p e c t e d t o ma k e 80 whi l e H u n t i n g d o n ma k e s
100, t hus s c or i ng poi nt s at % of H u n t i n g d o n ' s r at e. Si mi l ar l y,
M c C l i n t o c k scores at % t he r at e of C h a d wi c k . T h e n M c C l i n t o c k
s houl d score a t % X %
o r
% t he r at e of H u n t i n g d o n .
17. T H E S U RROGA T E 'S DI L E M M A . Wi l l i a m We s t on' s
" e v i de nt i nt e nt i on" wa s t o di v i de t he es t at e 2 : 1 as b e t we e n s on
a n d wi d o w, or 1 : 3 as b e t we e n d a u g h t e r a nd wi d o w. T h e s e r at i os
c a n be pr es er ved by a wa r d i n g t he son % 0 of t he est at e, t he
wi d o w % 0 , a nd t he d a u g h t e r % 0
18. T H E L I C E NS E P L A T E . I f t he pl at e s h o we d a r e a da bl e
nu mb e r ups i de - do wn as wel l as r i ght - s i de- up, t he n al l t he di gi t s
on it we r e reversi bl e. On l y five of t he di gi t s c a n be wr i t t e n so as
t o be r e v e r s i b l e 1 , 6, 8, 9, o. Si nc e al l f i ve di gi t s on t he pl a t e
we r e di f f e r e nt , t hey mus t ha v e c ompr i s e d j us t t hese f i ve. T h e
pr obl e m t hen is t o a r r a ng e t he di gi t s t o ma k e a n i nt eger wh i c h is
78, 633 less t ha n its i nver s i on. A l i t t l e t ri al wi l l di s c ove r t he n u m-
ber t o be 10968. A poi nt t o be r e me mb e r e d is t hat 1, 8, o st ay t he
s ame o n i nver s i on, but 6 a n d 9 e x c h a n g e i dent i t i es.
19. M E A S U RI NG T WO GA L L ONS . Fi l l t he 5 ga l l on c a n
a nd t he n po ur its c ont e nt s i nt o t he 8- gal l on me as ur e . Fi l l t he
5- gal l on c a n a s ec ond t i me, a n d t he n fill t he 8- gal l on me a s ur e
f r o m it. A s t he l at t er wi l l t a ke onl y 3 mo r e gal l ons , 2 gal l ons ar e
l e f t i n t he c a n.
144 SOLUTIONS
20. M A T C H S T I C K E QU A T I ONS . T h e i l l ust r at i on s hows
t he f o r m of e a c h e qua t i o n a f t e r t he pos i t i on of one ma t c h has
b e e n c h a ng e d .
I I I
I n t he first, t he t r ans f er of one ma t c h causes t he mi nus si gn
a n d t he e qua l s si gn t o e x c ha ng e pl aces .
I n t he s e c ond, t he mi nu s si gn is r e mo v e d a nd ma d e i nt o
figure i , whi l e t he R o ma n 10 n o w be c o me s a mul t i pl i c a t i on si gn.
I n t he t hi r d, a n i nt r us i ve ma t c h is a d d e d t o t he pl us si gn
t o ma k e it r e a d " pl us - or - mi nus . "
SOLUTIONS 145
I n t he f our t h, t he ma t c h mo v e d is us ed t o t r a ns f o r m t he
R o ma n 5 i nt o t he r adi c al si gn me a n i n g " t he s quar e r oot o f . "
I n t he l ast, t he e x t r a ma t c h c ha ng e s R o ma n 2 i nt o Gr e e k ir,
a nd if y our v i c t i m ar gue s t hat ^ = 3 . 1 4 1 6 , wh i c h is not e q ua l t o
, y o u c a n r et or t t hat t he l at t er v a l ue wa s a c c e pt e d as c or r e c t
7
i n a r c ha i c t i mes a nd still is us ed as a cl ose a p p r o x i ma t i o n.
21. WH A T I S T H E NA M E OF T H E E NGI NE E R? T he
bus i nes s man wh o l i ves near es t t he e ng i ne e r is n a me d Smi t h, a n d
t he engi neer ' s i nc o me is e xac t l y one - t hi r d of hi s. T h i s busi ness-
ma n c a nno t be t he br a k e ma n' s na me s a k e , f or t he l at t er is s ai d t o
e a r n $3500 a nd thi s n u mb e r is not di vi si bl e by 3. ( I f t hat poi nt
i n t he a r g u me nt seems t hi n, d o n' t a r g ue wi t h me!) T h e r e f o r e
t he br a k e ma n' s n a me is not Smi t h. No r is t he f i r e man' s n a me
Smi t h, be c aus e r a i l wa y ma n S mi t h be at s h i m at bi l l i ar ds a nd so
mus t be a di f f e r e nt per s on. H e n c e t he n a me S mi t h c a n be at -
t a c he d onl y t o t he e ngi ne e r , a nd t he ques t i on is a ns we r e d.
22. A T T H E RA I NBOW C L U B. Wh i t e wa s not hi ghes t , as
hi s c ar d wa s l owe r t ha n a not he r . Bl a c k wa s not hi ghes t , as hi s
pa r t ne r h a d a c hoi c e. B r o wn wa s not hi ghes t , si nce he wa s r i ght -
ha nde d, a nd t he c hoi c e of c ar ds f el l t o a l e f t - ha nde d ma n . H e n c e ,
Gr e e n d r e w hi ghes t c a r d. Gr e e n wa s not Whi t e ' s pa r t ne r , a n d
Br o wn wa s not Whi t e ' s par t ne r . T h e r e f o r e Gr e e n a n d B r o wn
we r e par t ner s agai ns t Wh i t e a nd Bl a c k , wi t h B r o wn a t Wh i t e ' s
l ef t . Bl a c k d r e w t he l owes t c a r d, si nce Wh i t e h a d t he c ho i c e f or
hi s si de.
23. T E NNI S A T H I L L C RE S T . T h e mi n i mu m n u mb e r of
sets t hat c oul d ha v e de c i de d t he t o u r na me nt wa s 1 5, t ot a l i ng 90
g a me s f or t he wi nner s . T h e wi nne r s a c t ua l l y t ook 97 g a me s ( 4 ) .
O n e e x t r a set wa s pl a y e d i n t he first r o und ( 3 ) , l e a v i ng 1 g a me
t o be a c c o unt e d f or . O n e set i n t he wh o l e t o ur na me nt mus t h a v e
r e a c he d 5-aIl a nd mus t ha v e be e n wo n at 7 5 .
Ba nc r o f t l ost hi s first ma t c h by 6 4 a nd 7 5 ( 7 ) .
146 SOLUTIONS
Fr a nk l i n r e a c he d t he final, wh e r e he l ost ( 8 ) . Si nc e he wo n
t he uni q ue 7 5 set, hi s first-round o p p o ne nt wa s Ba nc r of t .
Ot h e r first-round pai r i ngs we r e A b e r c r o mb i e vs. De v e r e u x
( 5 ) , a n d Go r ml e y vs. E g g l e s t on ( 9 ) . T h e r e ma i ni ng t wo ent r ant s
mus t h a v e b e e n p a i r e d : H a v e r f o r d vs. C h a d wi c k .
T h e wi nne r s i n t he first r o und we r e H a v e r f o r d ( 3 ) , Fr a nk -
l i n ( 8 ) , E g g l e s t o n a n d De v e r e u x ( 6 ) .
I n t he s e c ond r o u n d E g g l e s t on di d not me e t H a v e r f o r d ( 1 ) ,
nor di d he me e t Fr a nk l i n, f or Fr a nk l i n vs. Ba nc r o f t a nd E ggl es -
t on vs. Go r ml e y we r e i n di f f e r e nt hal v e s of t he or i gi nal br a c ke t
( 2 ) . T h e r e f o r e E g g l e s t on me t De v e r e u x , a nd H a v e r f o r d me t
Fr a nk l i n. T h e wi nne r s we r e De v e r e u x ( 6) a nd Fr a nk l i n ( 8 ) .
De v e r e u x wo n t he final f r o m Fr a nk l i n by 6 4 , 6 4 ,
6 - 4 - ( 8 ) .
24. WH I T E H A T S A ND BL A C K H A T S . If one man, say
A , wo r e a wh i t e ha t , t he n B wo u l d k n o w t ha t he hi msel f mus t
h a v e a bl a c k ha t , el se C wo u l d not ha v e rai sed hi s ha nd. By t he
s a me t oken, C c oul d i nf e r t hat hi s o wn h a t wa s bl a c k be c aus e B' s
h a n d wa s r ai sed. Wh e n nei t her B nor C s poke u p t h e nug g e t
of t he na r r a t i v e is t he wo r d " p r e s e n t l y " A k n e w t hat hi s o wn
ha t c oul d not be whi t e .
25. T RU T H A ND FA L S E H OOD. E v e r y i nha bi t a nt wa s
b o u n d t o say t hat he wa s a Di o g e n e t h e Di o g e ne s be c aus e t hey
we r e t r ut hf ul a nd t he A n a n i a s be c a us e t hey we r e l i ars. H e n c e t he
s e c ond ma n' s asser t i on mus t h a v e be e n f al se, a nd si nce t he t hi r d
s poke t r ul y he wa s a Di o g e ne .
26. WI NE A ND WA T E R. T h e r e is e x a c t l y as mu c h wa t e r
i n t he d e mi j o h n as t her e is wi n e i n t he buc ke t . Re ga r dl e s s of t he
pr opor t i ons of wi n e a n d wa t e r t r a n s f e r r e d a n d r egar dl ess of t he
n u mb e r of e x c h a n g e s i f t he t wo c ont ai ne r s first hol d e qua l
v o l ume s of p u r e l i qui d a n d e v e nt ua l l y ar e l e f t wi t h e qua l v ol ume s
of mi xt ur e s , e q u a l a mo u nt s of wi n e a n d wa t e r h a v e c ha ng e d
pl aces .
SOLUTIONS 147
27. FOU R P E NNI E S . T his is something of a "catch" prob-
lem. M anifestly there is no solution if the pennies must be kept
in one plane. T he trick is to arrange three pennies in a triangle
and put the fourth on top of any one of the others, creating in
effect a "double point."
28. S E VE N P E NNI E S . E ach penny after it is moved along
a line must come to rest adjacent to the penny or pennies al-
ready placed. T he adjacency of points of course refers to the line
connections, not to the circle on which the points lie. E very added
penny must therefore begin two points away from one already
down and must move to the intervening point.
29. T H E ROS E T T E . T he rosette contains seven coins, in-
cluding the one in the center. P erhaps you remember from geom-
etry that a true hexagon, a regular figure with six sides, is inscribed
in a circle by laying out six chords exactly equal to the radius of
the circle.
30. T H E M I S S I NG P E NNY. It would be correct to sell 3
apples of the inferior grade and 2 of the superior, together, for
twopence. A ny amount of the combined stock may properly be
sold at 5 apples for twopence, provided that the ratio of 3:2 is
maintained. T he sale of 60 apples would have come out right had
there been 36 of the cheaper kind and 24 of the other. But with
30 of each, 6 of the better apples were sold at the cheaper rate,
for a loss of 6 times the difference between the rates.
31. T H E RU BBE R C H E C K. T he dealer gave the customer
a radio and some cash in exchange for a worthless piece of paper.
H is loss was $43.75 plus $10.02, a total of $53.77.
Of every three solvers, two will probably argue that this
answer is wrong. One will say that the dealer also lost $26.23
profit on the radio given to the customer. But this is not cash out
of pockct and would not be entered as loss on the books. T he
148 SOLUTIONS
other solver will argue that from the loss of $53.77 must be de-
ducted $26.23 profit made on the radio given to the landlord. But
why is this particular item deductible? Why not, then, all the
profits made by the dealer on all his sales? T here is nothing in
the problem to indicate that the landlord would not have pur-
chased a radio in any case, or that the radio, if not sold to the
landlord, could not have been sold to someone else.
32. M YS T E RI OU S C OM P U T A T I ON. T he professor told
E dward that the computations were made in the heptary system
(radix 7) instead of the denary system (radix 10).
When wc write a juxtaposition of digits such as . . . C BA , in
the denary system we mean the cardinal number
A ( I O ) + B ( I O
1
) + C ( I O
2
)
If we use a system based on 7 instead of 10, the same digits
will express a different number, namely
A(7)+B(7
1
)+C(7
2
)
A formula like the above serves to convert the expression of
a number from one radix to another. If all numbers on the scrap
of paper are translated into the more familiar denary notation,
they are seen to be a correct division of 999 by 37 to give quotient
27, and the sum
92
74
82
248
T he reader may think that P rofessor Digit dipped into the
heptary system merely to puzzle his son. But the fact is that cer-
tain aspects of mathematical theory become clearer if a system
of notation other than the denary is used. For example, games
like No. 178Three-Five-Seven require the conversion of num-
bers to the binary scale (radix 2).
SOLUTIONS 149
33. T H E T E NNI S T OU RNA M E NT . P erhaps you laboriously
set up the brackets, as I did when asked this question. I felt fool-
ish when it was pointed out that 77 of the entrants have to be
eliminated, hence 77 matches are required.
34. T A RT A GL I A 'S RI DDL E . Four. Set up the proportion
5
1 0
: 3 = : x
2 3
and solve for x. T he argument is: Whatever mysterious factor
causes %X5
t 0
give the result 3 must also be introduced into the
product %X
1 0
- T his factor is expressed by the ratio 5/2:3.
35. S T RA NGE S I L H OU E T T E S . T he object is a type of
paper drinking cup dispensed in A merican railway cars. T he cup
folds flat, in which condition its shape is the frustum of a cone.
When it is opened out, with the mouth made circular, the creases
at the side become parallel. It might then be described as a kind
of cone with a circular base whose elements do not meet in a
single point but which all intersect a line-segment parallel to the
base and equal to its diameter.
T his puzzle was invented, however, long before the paper
drinking cup!
150 SOLUTIONS
36. T H E DRA FT S M A N'S P U Z Z L E . A consistent side view
is shown in Fig. I, and Fig. 2 is an oblique view of a block that
P [E7
FIO. 1 FIG. a
satisfies the drawings. T he broken lines can be taken to indicate
a rectilinear lug projecting from the surface, or a cavity gouged
out of the block.
37. A P ROBL E M I N P ROBA BI L I T I E S . T hree. If you were
caught on this question and gave some answer that involved 10
and 16, you are not the first!
38. A L A M I C E D P U Z Z L E . Rallod, enough for anyone!
39. BE A R FA C T S . T he bear was white. T he last lap of his
tour is at right angles to the first leg. If the bear is headed south
on reaching his den, and left it in a due south line, his den must
be on the North P ole, from which every direction is due south.
H ence he is a polar bear.
40. T H E FL A G OF E QU A T RI A . T he agreement of the
E quatrian states on the principle of dissection suggests that the
tessellation shows how the six-pointed star may be dissected to
useful purpose. By experiment we can discover that the pieces of
each star can be rearranged to form a single equilateral triangle,
the emblem of the E quatrian union. T he P unroe Doctrine pro-
ceeds from the fact that these five stars represent the only ways
in which a six-pointed star can be dissected into no more than
five pieces for this purpose. A n interesting collateral inquiry is to
prove that no other solutions exist.
r x z j
FIG. a
SOLUTIONS 151
H int: T he side of the equivalent triangle is the distance be-
tween opposite vertexes. In any five-piece solution, all pieces must
be compounded out of a basic unit, which is the small shaded
triangle. T he star is composed of 12 such triangles. T here are 13
partitions of the integer 12 into five parts. Investigate which
partitions can give geometrical solutions and which cannot. I
have considered a solution to be defined by the size and shape of
the component pieces, not by the ways in which they can be
assembled to form either the star or the triangle. In other words,
1 do not count, as separate solutions, different ways of making
cuts to get the same five pieces.
42. A QU E S T I ON OF BA RT E R. For a fishhook, 2 coco-
nuts; for a knife, 4; for a spear, 5.
43. S H A RI NG A P P L E S . T he boys gathered 72 apples.
44. A T RA NS A C T I ON I N RE A L E S T A T E . T he loss was
$200. T he tax bill was $500 and the repairs cost $1,000.
45. S E T T L I NG T H E BI L L . Originally there were 8 men in
the party.
46. C OWS A ND C H I C KE NS . T he wight had 4 cows and
31 chickens.
47. T H E FA RM E R'S RE T ORT . One cow, 34 chickens, and
2 stools.
48. DOL L A RS A ND C E NT S . Blake arrived with $12.35,
having started with $35.12.
49. T H E J A Y E S T A T E . T he estate was $40,000.00.
Note that "30 times more" means "31 times as much."
50. A FI S H S T ORY. T he mackerel was 21 inches in length:
head 4, body 12, tail 5. T he pickerel's measurements were: head
3, body 8, tail 4.
152
SOLUTIONS
51. WH O NOE S ? NOT A YE ! T he total number of voters
was 60. T he first vote was 3624 against the motion. T he second
vote was 3030. T he third was 3129 against.
52. NO FRE E Z E OU T . T he first loser must have had $4.05
before the last round; the second, $2.05; the third, $1.05; the
fourth, 55/; the last, 30/. T his problem is easy enough to solve
by working backwards. P erhaps the reader discovered the formula
whereby the problem can be solved for any number of players and
any final amount. T his final amount is of form m( 2
n
) where n
is the number of players. T he last loser must have started with
m(n-j-i), the next-to-last with m(2n-\-i), the third-last with
m(
4
n-f-i), and so on to the first loser, who started with
m(2
rl1
n-(-i). In the present puzzle, n5 and the final amount is
$1.60, so that m=5.
53. J OH NNY' S I NC OM E T A X. T he gross income was
$1064.00. T he tax was $106.40.
54. S P E NDI NG A QU A RT E R. Fourteen sheets of paper
(7<f), 8 pens (8/), 2 pencils (5^), and 1 eraser (5^).
55. T H E S P OOL OF T H RE A D. M rs. P lyneedle had 99^,
in coins of 50^, 25^, io<f, 5/, and four if pieces. T he thread was
19/ per spool. T he selection of coins was, for one spool: 10, 5,
1, 1, 1, 1; for two spools: 25, 10, 1 , 1 , 1 ; for three spools: 50, 5,
1, 1; for four spools: 50, 25, 1.
56. A DE A L I N C A NDY. One boy chose 20 chocolate bon-
bonsnotice that it is not stated that each boy took some candy
of each kind. A nother boy decided on 5 jujubes, 12 bonbons, and
3 lollipops. T he third took 10 jujubes, 4. bonbons, and 6 lollipops.
57. WH A T S I Z E BE T ? T he bet was 60/. H igh hand put
in a dime and a half-dollar. Second hand put in a half-dollar
SOLUTIONS 153
and 2 nickels. T hird hand put in a dime and a silver dollar, taking
back a half-dollar. Fourth hand put in a silver dollar and 2
nickels, taking out the other half-dollar. L ast man put in a silver
dollar, taking out the 2 dimes and 4 nickels, leaving just 3 silver
dollars.
T he try that catches some solvers is to make the bet 35/.
T he plan is to have the last man chip a dollar, leaving this coin
plus a half-dollar and a quarter in the pool and taking the rest.
But the pool is cleaned out of quarters before his turn comes.
58. T H E H OS KI NS FA M I L Y. T here are 6 women, 5 men,
and 3 children. From the facts given it can be determined that
the quota for a woman is 8 times that for a child, and for a
man is 13 times that for a child. T he problem is to find integral
values for the letters in the equation
8W+I 3 M +C =i i 6
with the proviso that W> M > C .
59. T H E T I DE . T he force of the tide was % of a mile per
hour. A ll rate problems depend on the formula
rtd
where r is the rate, t the time, and d the distance. L et r here stand
for the force of the tide, and R for the rate of the boat in still
water. T hen the net speed of the boat going with the tide is
R-]-r and against the tide is Rr.
60. L OC A T I NG T H E L OOT . Seven minutes elapsed be-
tween the car's two meetings with the trooper. T wo minutes were
lost in stopping and turning. T o reach a point 2 miles from the
booth required minutes. T hat left 3% minutes to go beyond
to the cache and double back to the same point. In this time the
car would travel 4% miles, or 2 % miles each way. H ence the
car could have got no farther than 4% miles beyond the booth.
61. S T RI KI NG A N A VE RA GE . T he first 5 miles were cov-
ered in 10 minutes. T o go 10 miles at the rate of 60 miles per
154 SOLUTIONS
hour requires 10 minutes. T he motorist can strike this average
only by being transported instantaneously from the halfway mark
to his destination.
62. T H E S WI M M I NG P OOL . Jill won the race. T he pool
was 42 feet long.
63. H A NDI C A P RA C I NG. T he little boy overtook the little
girl 15 seconds after the start of the race, or 5 seconds after the
big boy passed her.
64. T H E P A T ROL . T he first occasion when all three patrol-
men could meet at point A was an hour and a half after leaving it
simultaneously. T herefore the radio message arrived at 4:30 A.M.
65. T H E E S C A L A T OR. A s this problem has proved puzzling
to many, I will show how to attack it. L et x be the number of
steps from bottom to top. L et t be the time required from any one
step to displace the one immediately above it. A person standing
still would require a time xt to go from bottom to top. But H enry
walked up 28 steps and so reached the top on that step which was
x28 from the top when he started at the bottom. T he time of his
trip was therefore (x28) t. Since he took 28 steps in this time, he
walked at the rate of IF P
E R
ste
P >
o r t w o st e
P
s
'
n
the time
i f
t
. In the same time M artha took one step, and by the same
!
4 t
x 2 l
\
t
form of computation her time per step was _ E quate the
21
two latter terms and solve for x. T he answer is that the escalator
was 42 steps long.
66. T H E C A M P E R A ND T H E BOT T L E . T he camper rcached
a point 3% miles upstream from his camp. H e first met the bottle
2 miles awaythe distance he could travel in 48 minutes at the
SOLUTIONS 155
net rate of 2% miles per hour. T he bottle then floated 2 miles at
the rate of the stream, taking 80 minutes to reach the camp. In
this time the canoer, with respect to the water, went upstream
some distance and then returned to his starting point (the dis-
tances and times of the two trips must be equal). T he total time
being 80 minutes, the upstream trip occupied 40 minutes. With
respect to the banks of the stream, the canoer still proceeded at
the net rate of 2% miles per hour, and therefore traveled an
additional 1 % miles away from his camp.
67. H I T C H A ND H I KE . T he trip was done in 5 hours and 6
minutes. By jeep alone the transport would have required 7 hours
and 30 minutes. T he hiking therefore saved 2 hours and 24 min-
utes.
T he arrival of the parties simultaneously might have been the
consequence of a decision on the part of the sergeant to make each
party walk just 16 miles. H e dropped the first party 16 miles from
the rendezvous, picked up the second after it had walked 8 miles
and took it to a point 8 miles from the rendezvous, and picked up
the last party after it had walked 16 miles.
68. I F A M A N C A N DO A J OB. Working alone, the tinker
would take 3 days and the helper would take 6. T he apprentice
would never finish the job at allin fact, he turned out to be en-
tirely useless.
69. FI NI S H T H E P I C T U RE . One cube and one sphere. T he
relative weights of the solids are: cylinder 13, cube 8, sphere 4,
cone 3.
70. T H E A L C A N H I GH WA Y. Fifteen weeks behind sched-
ule. T hat is, the whole job will take one bulldozer sixteen weeks.
71. S E E S A W. A lfred must sit 37% inches from the center
point. H e weighs 32 pounds and Bobby weighs 60 pounds.
156 SOLUTIONS
12. A P ROBL E M I N C OU NT E RWE I GH T S . Obviously a 10-
pound weight is needed. T hen a 20-pcund weight will make it
possible to offset a load of 20 or 30 pounds, the latter by addition
of the io-pound weight. Next a 40-pound weight will make pos-
sible compounds of weights of 40, 50, 60, 70 pounds. E vidently the
series of the most economical weights is
10X2"
where n takes all integral values o, 1, 2, etc. T he five counter-
weights are therefore 10, 20, 40, 80, 160, reaching a total of 310
pounds.
73. T H E A P OT H E C A RY' S WE I GH T S . T his problem differs
from A Problem in Counterweights because here we can use sub-
traction as well as addition.
It will perhaps be clearer if we deal with units of 1 instead of
A fter we have determined the proper set to weigh out all in-
tegral quantities from 1 up, we can divide the several weights by 2.
We have to start with a weight 1. T o measure 2, it is most
economical to add a weight 3, for then we get 2 by subtraction
and can also make 3 and 4. T o get 5 et seq. we can make the third
weight 9, for then subtraction of the 4 and the lesser integers
gives 5, 6, 7, 8, while the 9 alone and additions of 1 to 4 give 9, 10,
11, 12, 13. It is evident that the most economical series of weights
is given by 3", where n takes all values from o, 1,2, 3, up. T he first
five weights in the series are 1, 3, 9, 27, 81. Divide by 2 for appli-
cation to the problem: Weights in grams of 1%, 4%, 13%,
and 40% reach the maximum possible total of 60%
74. S A L L Y'S A GE . Sally's age is 22 years and 8 months.
75. A S OL D A S A BC . A lice is 8; Betty is 5; C hristine is 3.
76. FU M E R FROWNS . Because the shop actually lost 10$?
on the sale. T he Vesuvius pipe must have cost $1.50, and the 20%
loss was 30^. T he P opocatepetl pipe cost just a dollar, and the
20% profit was 20/.
SOLUTIONS 157
77. C OM P OU ND I NT E RE S T . L ike many algebraic prob-
lems, this is easiest solved by determining the general formula ap-
plicable to any rate of interest and any number of compoundings,
then filling in the actual values given in the problem.
L et a be the initial deposit and r be the rate. T hen the inter-
est at first compounding is ar and the total is a-\-ar. Interest at the
second compounding is (a-\-ar)r, and the total is this quantity
plus (a-\-ar). T he whole expression reduces to
a(i-\-2r~\-r
2
)
By continuing the process through further compoundings, it will
be seen that the total s is always equal to
5 =( i +r ) "
where n is the number of times interest has been compounded.
A pplying the formula to the problem, we have
$i oo. oo=a( 1-I-.03)
5
T he expansion of the right-hand member looks more for-
midable than it is. T reat it as a binomial and apply the binomial
theorem. T he reader who has forgotten the theorem can look up
the coefficients in "P ascal's triangle" under No. 136Figurate
Numbers. T hese coefficients are 1, 5, 10, 10, 5, 1. If the terms of
(i-j-.o3)
6
are written out, without expansion, it can be seen that
they decrease in magnitude very rapidly. If we take only the first
three terms (1 -f-. 15-I-.009) and accept the sum as about 1.16, we
shall be near enough for practicable purposes. With this value in
the equation, we get
a=$86.20
which is accurate within the statement of the problem. T he ac-
tual sum which must be deposited is $86.25.
78. T H E S A VI NGS A C C OU NT . T he general formula for
compound interest is given in the solution above. In the present
problem we have
$i 3i . 67=($i oo. oo) ( 1 +r )
8
T he solution of this equation boils down to finding the 8th root of
1.3167. T his can be done by extracting the square root three
158 SOLUTIONS
times. If the operations are carried to only four places of decimals,
r is found to be .0351. T he rate is actually 3^2%.
79. A FT E R FI VE O' C L OC K. T hese puzzles, like many
others concerning clocks, are based on the circumstance that the
minute hand travels 12 times as fast as the hour hand. T he basic
equation for such puzzles is
T)=12d
where D is the distance traversed by the minute hand and d the
distance traversed by the hour hand.
In the ordinary garden variety of clock, the hands move by
little jerks. T hey stand still for an interval of one or two seconds,
or even, in many electric clocks, for a full minute. But puzzle
books are inhabited by a very special kind of clock in which the
hands move continuously.
(a) 27%! minutes past five. T he second equation here is
D=^+2
5
.
(b) 10*%! minutes past five. A t right angles, the hands are
15 minutes apart. H ence D-f-15^+25.
(c) i 6%! minutes. T he involved statement of the question
is equivalent to asking: What is the interval of time between co-
incidence of the hands and the next position 90 degrees apart?
T he answer can be found simply by subtracting the answer to (b)
from the answer to (a).
80. T H E C A RE L E S S J E WE L E R. T he watch first shows the
correct time at 5%i minutes past three. During the hour from
two to three the watch is continuously wrong. T he minute hand
moves from XI I to I while the hour hand makes a complete cir-
cuit from II to II. Between three and four o'clock, the hour hand
makes a second circuit, and the watch will be correct at one in-
stant while it is in transit between III and IIII. L et d be the dis-
tance from II to this point. L et D be the distance the minute hand
advances beyond I during the same time. Because of the reversal
of the hands
d=\*T>
SOLUTIONS 159
Wh e n t he same posi t i on is r e ac he d by a c or r ec t wa t c h, t he
hour ha nd move s f or wa r d f r o m I I I a di st ance d 5 , whi l e t he
mi nut e ha nd t ravel s f r o m X I I a di s t ance D- }- 5, a nd thi s di s t ance
is 12 t i mes t hat mo v e d by t he hour h a n d :
D + 5 = i 2 ( < f 5 )
T h e ans wer to t he pr obl e m is f o und by sol vi ng t he t wo e qua-
ti ons si mul t aneousl y.
81. C L OC K S E M A P H ORE , ( a ) A t i 8
6
/ 1 3 mi nut es past t wo.
( b) A t 5 5 %3 mi nut es past t wel ve.
83. NOW RE VE RS E I T . F r o m t he di scussi on of No . 8 2
Changing a Rectangle to a Square the reader may have j umped
t o t he concl us i on t hat a r ec t angl e of any di mensi ons c a n be dis-
sect ed to ma k e a n equal s quar e by only three pieces. S u c h a con-
cl usi on is i ncor r ect .
L e t us l ook at t he quest i on i n t he reverse wa y . Gi v e n t he
square A B C D i n Fi g. 1, h o w e l ongat e d a r ec t angl e c a n we ma k e
by t he t hr ee- pi ece di ssect i on? If E D is t o be t he hei ght of t he rec-
t angl e, c onnec t E G a nd cut . O n B C meas ur e BF e qual t o E D,
t hr ough F dr a w F G par al l el t o A B, a nd c ut on thi s l i ne. Sl i de t he
uppe r pi ece d o wn t o t he r i ght so t hat its c or ner G coi nci des wi t h
C . T r a ns f e r t he smal l t r i angl e so t hat F C coi nci des wi t h A E .
160 SOLUTIONS
I t is e v i de nt t ha t wh e n E D is j ust hal f of A D , t he poi nt G
c oi nc i de s wi t h E , a nd F E di vi des t he s quar e i nt o t wo e qua l r ec-
t angl es . We c a n t he n omi t t he c ut E C , a nd b y c ut t i ng i nt o onl y
t wo pi e c e s p r o d u c e a r e c t a ng l e who s e si des ar e i n r at i o i : 4.
I t is al s o e v i de nt t ha t if E D is less t ha n hal f of A D , t he t hr ee-
pi e c e c ut t i ng wi l l no t wo r k . We wi l l h a v e t o di ssect i nt o mo r e
pi ec es i n o r de r t o p r o d u c e a r e c t a ngl e mo r e e l ong a t e d t ha n 1 : 4 .
C onv e r s e l y , a r e c t a ng l e who s e si des a r e i n gr e at e r r at i o c a nno t be
c ut i nt o onl y t hr ee pi ec es t o f o r m a s quar e.
FIG. 2
T h e l i ne B g i v e n i n Now Reverse It i s i n l e ng t h a l i t t l e mo r e
t ha n t wi c e t he si de of t he s quar e A . T h e s ol ut i on is t o cut t he
s quar e i nt o t wo e q ua l r ec t angl es , set t h e m si de by si de as i n Fi g. 2
t o f o r m a r e c t a ng l e 1 : 4 , t he n ma k e t he t hr ee- pi ec e cut s t o f or m
t he mo r e e l o ng a t e d r e c t a ng l e of si de B. T h i s di vi des t he or i gi nal
s qua r e i nt o f i ve pi eces.
FIG. 3
A n o t h e r s ol ut i on s i mpl y reverses t he or de r of t he t wo steps.
Fi r s t c ut t he s quar e i nt o t hr e e pi ec es t o f o r m a r e c t a ngl e whos e
l ong e r si de is hal f of l e ngt h B. T h e n c ut t hi s r e c t a ngl e i nt o t wo
e qua l r e c t angl e s wh i c h c a n be set si de by si de t o f o r m t he f i nal
f i gur e. Fi g . 3 s hows t hat a g a i n we mus t ma k e f i ve pi eces.
SOLUTIONS 161
In all the above, we have fixed the position of cuts by meas-
urement with the smaller side of the desired rectangle. T he given
line B is the larger side. In case any reader is puzzled how to de-
termine the smaller side, I give the construction in Fig. 4. A B is
the given length B, while BC , perpendicular to it, is the side of the
square A . C onnect A C . C onstruct the perpendicular bisector DE ,
of A C , cutting A B in E . T hen A E equals E C ; measure this same
length from E toward B to fix point F. T he desired length is BF.
[Since A B: BC =BC : BF, or (A B) (BF) = (BC )
2
.]
84. T H E KI T C H E N L I NOL E U M . T he linoleum need be cut
into only two pieces, as shown by the diagram. E ach "step" is
4
feet wide and 3 feet deep. If the right-hand piece is moved one
step down and leftward, a square 12 by 12 is formed.
T his "step" principle is the basis of many rectangle-to-square
puzzles, but there is more to it than meets the eye!
162 SOL U T IONS
85. T H E BROOM C L OS E T . T he diagrams show how to cut
the linoleum into 3 pieces that can be fitted together to form a
square, correct in pattern.
A B D C
SOLUTIONS 163
86. S U M M I NG T WO S QU A RE S . Set the two squares side
by side as shown in the diagram. M easure off C D equal to A B,
the side of the smaller square. C onnect DE : this is the side of the
total square. C onnect DF: this marks a second side equal to DE .
C ut along DF and DE . Bring the 3 upper pieces down to the po-
sitions shown by the broken lines to form the total square. T his
demonstration shows that any two different squares can be dis-
sected into one through not more than 5 pieces.
87. FROM A T O Z . T he diagrams show how to cut the A
into 4 parts to make a Z . T he secret of avoiding having to turn
over a piece lies in that little equilateral triangle.
88. T H E M I T RE FA L L A C Y. T he mitre indeed can be made
into a rectangle, but this rectangle cannot be cut into a square on
the "step" principle.
T o make the proof concrete, let us say that the mitre is made
out of a square 4 units on each side. (We can choose any number
we please, since we will deal with ratios, and ratios are the same
whatever the unit of measurement.)
T hen the rectangle made from the mitre is 3 by 4.
T he width of all steps must be the same, and the heights must
be the same. H ence, with 7 widths along the 4-unit edge and 6
heights along the 3-unit edge, each step is % wide and % high.
164 SOLUTIONS
Sliding the right-hand piece down one step adds % to the
depth of the rectangle and subtracts % from its width. T he re-
sultant "square" is 3% wide by 3% deep. T hat is, it is not really
a square.
Nor can a different number of steps be found that will yield
a true square. T he simplest way to prove this fact is to investigate
the conditions necessary so that a rectangle can be cut into two
pieces stepwise to form a square.
L et the sides of the rectangle be a and b, with a the greater.
L et the dimensions of each step be w and h, w being parallel to a.
L et a/w=n, the number of steps counted parallel to a. L et
b/h=n', the number of steps parallel to b.
M anifestly, the dissection is feasible only if n and n' are both
integral and n=n'-\-i.
Furthermore, the dimensions of the altered rectangle will be
aw and b-\-h; the figure is a square only if these two are equal
(and consequently both equal to \Jab.).
From the foregoing equalities we can express n and n' in
terms of a and b:
a b
n = and u' = =
a\/ab \ / abb
A rectangle yields to stepwise dissection only if its dimensions
give integral values for n and n' in the above equations, such that
nn'-f-i.
L oyd's proposed solution of the mitre problem deals with a
rectangle whose sides are in ratio 3:4. T hese values do not satisfy
the above conditions.
89. WH A T P ROP ORT I ONS ? From the equations given
under The Mitre Fallacy the following can be derived:
a n
2
~b
=
{n i)
2
T he present problem gives 71=13. T hen, the two sides of the
rectangle must stand in the ratio 13
2
:12
2
, or 1 6 9 : 1 4 4 .
SOLUTIONS 165
T he above equation serves to point out that a rectangle yields
to stepwise dissection into a square only if its two dimensions
stand in the ratio of two consecutive integral squares.
90. T H E ODI C FORC E . T he diagrams show the dissection.
T he bottom third of the large square is cut off and this rectangle
No. 82Changing a Rectangle to a Square. T his is the square of
area 3. T he square of area 1 is cut out intact from a corner of the
remaining piece, its side being % the side of the original square.
T he third piece, of irregular shape, can be cut into 3 pieces to
make the square of area 5. T he point on the base where the two
cuts meet is % the distance from left to right.
91. T H E P I A NO L A M P . T he diagrams on page 166 show
how the piano lamp can be cut into 10 pieces that can be fitted
together to form a circular disk.
166 SOLUTIONS
SOLUTIONS 167
92. C ONS T RU C T I ON OF A P E NT A GON. In the diagram
given with the construction, BC is the hypotenuse of a right tri-
angle OBC . H ence BC
2
=C 0
2
+0B
2
. We can treat the circle as a
unit circle and set OB equal to i. A nd C O=A C A O; A O= %;
A C =A B=\/5/2. From these figures BG is found to be 1.17558.
(T he correctness of the construction can be verified by trigo-
nometry. T he side a of the pentagon inscribed in a circle of radius
r is given by a2r sin 36. Sin 36=-58779, and for r=i,
0 = 1 . 1 7 5 5 8 . )
93. T H E A M U L E T . T he dissection of a pentagon into a
square in only 6 pieces was for some time a baffling problem, and
its solution by Dudeney deserves to be better known than it is.
Fig. 1 shows how the pentagon is cut into 3 pieces to make a
parallelogram. One cut is on the line A B and the other on C D; C
is the midpoint of A B and A D is made equal to A C . T he broken
lines show how the 2 upper pieces are rearranged around the
lower piece.
In Fig. 2 the parallelogram is dissected by two cuts (broken
lines) to make a square. T he distance FB is the mean proportional
between the base E B and the height of the parallelogram. (H ow
168 SOLUTIONS
to find the mean proportional is described under No. 82Chang-
ing a Rectangle to a Square.) T he two cuts divide the parallelo-
gram into 6 pieces altogether, which make a square as shown in
Fig- 3-
94. FOU R- S QU A RE . T here are 10 different pieces, as shown
by the diagram.
H ^ B P J T
SOLUTIONS 169
95. J A C K O' L A NT E RN. T he diagrams (Figs, i and 2 on
pages 169 and 170) show how to cut the disk and arrange the
pieces to form the jack o' lantern.
170 SOLUTIONS
FIG. 2Ja c k O' L antern
96. A RE M A RKA BL E OC C U RRE NC E . T he two accom
panying diagrams show how the board
can be cut into only 11 pieces to form the
tableau.
R 4 CSJ
1
3
1 L _
H
r
1
L

SOLUTIONS 171
97. T A NGRA M P A RA DOXE S . T he diagrams given below
show the construction of the two men in Fig. 3 (page 50). T he
man with the foot is actually thinner than his companion with-
out a foot, by the area marked out by the broken line.
T he next diagram shows how to construct the notorious fish
(page 51). A nd page 172 shows the solution of the paradoxical
diamonds (page 52).
172 SOLUTIONS
98. T H E C L U B I NS I GNI A . If the larger circle is % of an
inch in diameter, the side of the triangle will be I /\/6=O.4O82 of
an inch.
99. T H E FE RRYBOA T GA T E . T he spread is 14 feet, meas-
ured between centers of the outermost rods. T he principal diag-
onal members always make isosceles triangles with legs of 15
inches; when the base is 18 inches, the altitude on this base is 12
inches (remember the 3-4-5 right triangle). T wo such triangles
SOLUTIONS 173
per panel for 6 panels (between rods) account for 12 feet of the
spread, and 2 feet are added by 12X2" distances from center of
rod to collar pivot.
100. S T RI KI NG A BA L A NC E . T he lever will balance if the
moments on each side of the fulcrum are equal. T he moment of a
force is the product of its magnitude and its distance from the
fulcrum. If we call the distance of the 6o-pound weight from the
fulcrum unity, then the 105-pound weight must be placed % of
this distance on the other side (60X1 = 105X Vi)
c
T he method usually given in textbooks for dividing a given
line into a given number of segments is illustrated in Fig. 1. T o
divide line A B into 7 equal segments, draw line A G at any arbi-
trary angle; lay out 7 equal lengths of any convenient size from A
terminating in C ; connect GB; through each division point on
A C draw a line parallel to C B; these parallels will divide A B into
7 equal segments.
From the practical point of view, the method illustrated in
Fig. 2 is much superior. P arallel to the given line A B draw an-
other line C D, at any convenient distance; on C D lay out 7 equal
lengths of arbitrary size greater than % of A B; connect the
terminal points C and D respectively with A and B; these two lines
intersect at E ; connect E with each of the division points on C D;
174 SOLUTIONS
E
these connecting lines will divide A B into 7 equal segments.
T he second method requires the construction of only one
parallel instead of n-1 parallels. It is far superior where the con-
struction, for lack of bow compasses, can only be approximated
(as is often the case in military sketching). T he drawing "by
eye" of n1 parallels allows wide error. But where only one
parallel need be drawn, at an arbitrary distance, it can be fixed
with tolerable accuracy by the opposite edges of a ruler or any
suchlike tool.
101. A N I NT E RC E P T P ROBL E M . T he length A B is equal
to 87-/5. T he computation involves analytical algebra. Determine
the equation of the tangent OD and the equation of the middle
circle; solve these equations simultaneously to find the points of
intersection A , B; then compute the distance between these points.
T ake O as the center of rectangular co-ordinates, with OE as
the x-axis. T he equation of line OD is then given by the ratio of
DG (perpendicular from D to OE ) to OG. Since ODF is a right
triangle, the altitude DG is equal to (OD) (DF) /OF. DF equals r
SOLUTIONS
175
and OF equals 5 r; from these values the other terms can be com-
puted. T he equation of OD is found to be
y V 6
x 12
T he equation of the middle circle is
(*;
3
r)
2
-H y
2
=r
2
Solve the two equations simultaneously for the value of x,
which is found to be
72R 8R ,
V6
25 25
v
T he two values of x are the abscissas of the points B and A .
16 r _
T he difference between these values, equal to is the length
of A C . Since A BC is a right triangle by construction, A B can be
computed from A C and BC . T he latter can be computed by use
of the equation for OD. T he desired length A B is found to be
102. T H E BA Y WI NDOW. T he ladder reaches a point 16
feet above the ground. T he algebraic solution involves a fourth-
degree equation, which is easiest solved by trying out integral
values within a narrow range, since the answer obviously is more
than 12 feet and less than 20. But easier than solving the equation
is to recognize that the problem is based on our old friend, the
3-4-5 right triangle.
176
SOLUTIONS
103. T H E E XT E NS I ON L A DDE R. In the diagram, DA
represents the side of the M ayfield Building, A E the street (at
sidewalk level), and C E the face of the building on the opposite
side. T he extension ladder is represented by DE (the proportions
being distorted for the sake of clarity). A C is the traffic barrier
ladder. T he bracing ladder is not shown, as it contributes noth-
ing to the solution.
10 inches=8i4 inches; that A C =22 feet i i inches=275 inches;
and that distance BF=5 feet io inches=70 inches. DA and C E
are parallel, both perpendicular to A E . T he problem is to deter-
mine the length DA .
Obviously, many equations can be derived from the figure,
but the manipulation of these equations has proved puzzling to
many solvers.
T he difficulty arises fundamentally from the fact that the
final formula for the length DA is bound to be a quarticthat is,
to involve the fourth power of the unknown x. T he solution of a
quartic equation is certainly not within the scope of elementary
algebra, but in this case the necessity can be circumvented by
methods discussed in the solution to No. 142The Battle of
Hastings.
SOLUTIONS 177
T he solver is likely to make added difficulty for himself by
working from more equations than he needs, not all of them in-
dependent. T he trick is to select just the right facts, so as to avoid
getting into a morass of implicit equations.
Since A DE and A C E are right triangles, we can write
DE
2
=DA
2
+A E
2
( i )
A C
2
=C E
2
+A E
2
(2)
Subtract (2) from (1).
DE
2
A C
2
=DA
2
C E
2
(3)
T riangles A BD and C BE are similar, whence we derive the
proportion
DA : C E =A B: BC (4)
Now through point B draw line BG parallel to A E . T hen
triangles A BF and BC G are similar, and we have the proportion
A B: BC =BF: C G (5)
C ombining (4) and (5) gives
DA : C E =BF: C G, or
C E XBF
C G (6)
Substitute this value of DA in (3).
[
C E XBF I
2
C G
J - C E
2
(7)
For C E substitute the equivalent expression C G-f-BF.
[
( C G+BF) XBF1
2
( C G+BF)
S
(8)
Of the four terms in (8), the values of three are given by the
statement of the problem, wherefore (8) serves to solve it. Sub-
stitute these values, and replace C G by the simpler term x.
814
2

2
75
2
[ V
+ 7
] (* +7)
2
(9)
T he value of the left member of (9) is 662,59675,625
=586,971. H ere we have an integer, the difference between two
squares. T he squares are not necessarily integral, but it is worth-
while trying out the hypothesis that they are. T he method of find-
178 SOLUTIONS
ing all integral squares that differ by a given integer is expounded
in the solution to No. 142The Battle of Hastings.
A ccording to that method, we must find an appropriate term
r among the factors of 586,971. T he prime factors of this integer
a r e
5 8 6 , 9 7 i = 3 X3 X7 X7 Xi i Xi i Xn
A pair of squares can be found for every factor r that is less
than the square root of 586,971, e.g., 1, 3, 7, 9, etc. Since there
are many such factors, let us see if we can narrow the search.
In equation (9), x represents the vertical distance from the
junction of the two ladders to the raised end of the barrier ladder.
Since the latter ladder is nearly horizontal, x is relatively small.
T hen the term [ + 7 0 1 must greatly exceed the term (x-f 70).
C onsequently, r must be relatively large.
For the largest possible values of r we find the following pairs
of squares:
FACTORS r a SQUARES
3 X 7 X 1 I
= 231 231 1 i 3 8 6
2
- u 5 5
2
7 X 7 X 1 I
= 5 39 5 5 i
8 1 4
2
2 7 5
2
3 X 3 X 7 X 1 I = 693 ' 5 5 7 7 o
2
7 7
2
T he last pair is the only plausible set. T o try it out, we write
= 770 and ( x +7 ) =7 7
Since the two equations agree in giving the value x=7, this
is the correct solution of the quartic. By reference to equation (3)
it is seen that the first term above gives the value of DA .
T he height of the M ayfield Building is thus determined to be
770 inches, or 64 feet 2 inches.
104. T H E S P I DE R A ND T H E FL Y. C onsider the room to
be composed of 6 plane rectangles, hinged together in any suit-
able way so that they can be folded up in the form of a box. If
SOLUTIONS 179
they are unfolded and laid flat in one plane, and the positions of
the spider and fly are plotted, then the shortest course will be the
straight line joining the two points.
No. i
N. 2
No. 3
Different ways of hinging the rectangles are possible, and it
is a matter of experiment to find which produces the shortest
course. T hree possibilities are shown in the diagrams. No. I re-
quires the spider to travel 42 feet. No. 2 is better, reducing the
distance to a little over 40 feet. But the best is No. 3, with a course
of exactly 40 feet. T he odd fact is that this course touches 5 of the
6 rectangles.
180 SOLUTIONS
105. T H E S P I DE R'S C OU S I N. T h e p l a n of t he short est
r out e is f ai r l y obvi ous . Si nc e t he s pi der mus t cross o n t he floor t o
r e a c h t he i nne r wa l l , he c a nno t do be t t e r t ha n h e a d di r ect l y f or
t he v e r t e x M i n t he d i a g r a m, at t he poi nt wh e r e i t t ouc hes t he
floor. Wi t h one out e r wa l l " o p e n e d o u t " o n a hi ng e wi t h t he floor,
t he f i rst l e g of t he j o u r ne y is s ho wn by t he l i ne S M . T h e s econd
l e g is s h o wn o n t he s ket c h b e l o w t he pe nt a g on. T w o of t he i nner
wal l s a r e o p e ne d o ut on a hi ng e b e t we e n t he m, a nd t he short est
r out e is t he di a g o na l M F .
M l I
I NNE R WA L L
FIG. I
C o mp u t a t i o n of t he s e c ond l e g is s i mpl e. M F is t he hy -
pot e nus e of a right t r i a ngl e who s e si des ar e 9 f e e t a nd 1050 f eet ,
t he l at t er b e i ng 3 / 2 ( 7 0 0 ) . T h e he i g ht a t F be i ng r el at i vel y so
SOLUTIONS 181
smal l , M F c a n be as s umed t o be 1050 f eet wi t hi n t he " near es t
f oot " l i mi t . ( A c t ual l y i t is a bout 1050. 04. )
C o mput a t i o n of t he first l eg is not so si mpl e. S M is t he
hypot enus e of a r i ght t r i angl e whos e sides ar e M P a nd P S. T h e
di st ance P Q is hal f t he di f f er enc e be t we e n t he l engt hs of a n out er
and a n i nner wal l . H e nc e M P is 1 1 00 f eet . P S is 9 f eet pl us t he
di st ance be t we e n t he wal l s ; c al c ul at i on of thi s di st ance is t he
nugget of t he puzzl e.
I n Fi g. 2, a and b ar e t he si des of t wo concent r i c pe nt a gons
and O is t he cent er . T h e r adi us of t he smal l er pe nt a g on is d; of
t he l arger, d + c . T h e a pot he g ms ( per pendi c ul ar s f r o m c ent er t o
si de) ar e respect i vel y / a nd f-\-e.
T h e rat i o b/d is c ons t ant f or al l pe nt agons ; cal l thi s r at i o k.
No w l et us f i nd t he v a l ue of al l t he segment s i n t er ms of a, b, k.
182 SOLUTIONS
We ha v e
b
(0
a
C
+
d
=I
00
a nd by c o mbi ni ng ( i ) a nd ( 2)
ab
(3)
I n t he smal l r i ght t r i angl e
(4)
By s i mi l ar t ri angl es we ha v e [Link]. C o mb i ni ng thi s equa-
t i on wi t h ( 1 ) , ( 3 ) , ( 4) gi ves
In No. 92Construction of a Pentagon we find k to be
1 . 1 7558. T h e n t he v a l ue of t he r adi cal i n ( 5) is 1. 3763, a nd t he
di s t ance be t we e n t he wal l s of t he P e nt a g on Bui l di ng is 400 t i mes
thi s numbe r , or a bo ut 550. 52 f eet . T h e di st ance P S i n Fi g. 1 is
t hen 559. 52 f eet . Sol v i ng t he r i ght t r i angl e gi ves S M , t he first l eg
of t he spi der' s j our ne y , as al most exac t l y 1 234 f eet .
T h e t ot al l e ngt h of t he r out e is 1 05 04- 1 23 4=2284 f eet , cor-
r ect wi t hi n a smal l f r ac t i on of a f oot .
106. T OU RI NG T H E P E NT A GON. I ma g i ne a schemat i c
pl an i n wh i c h e a c h cor r i dor is r epr esent ed by a si ngl e l i ne do wn
i ts cent er . E a c h i nt ersect i on of t wo or mor e l i nes is a node. C o u n t
t he numb e r of rays at ever y node. A r ay is a l i ne s egment con-
(5)
SOLUTIONS 183
st rued t o t er mi nat e at t he no de ; henc e a st r ai ght l i ne passi ng
t hr ough and be yond t he node count s as t wo rays.
Wh e n any even node ( even numbe r of rays) is r eac hed by a
ne w pat h, t here wi l l al ways be a ne w pa t h of r et r eat f r o m it. But
whe n a n odd node is ent er ed f or t he last t i me, i t wi l l necessari l y
be al ong t he l ast unus ed pat h, a nd exi t c a n t hen be ma d e onl y
al ong a pa t h pr evi ousl y t raversed. T h e numb e r of r et r aci ngs
necessary i n this ( or any s uch) puzzl e depends sol el y on t he num-
ber of odd nodes.
T h e P e nt agon pl an has 10 o dd nodes, al l i n t he out er mos t
ci r cumf er ent i al cor r i dor whe r e it meet s t he t ransverse corri dors.
We shal l ha v e t o r et r ace hal f this numb e r of l e g s 5 . T h e di a g r a m
shows t he mi ni mal t our ( t he c ont i nuous l i ne is t he cour se of t r avel ,
184 SOLUTIONS
t he par t i t i ons b e i ng s uppr es s ed f or c l a r i t y ) . T h e r e ar e ma n y
di f f e r e nt wa y s of a r r a ng i ng t he c i r c ui t of e v e n nodes , b u t t her e is
no wa y of s hor t e ni ng t he 5 wa l ks " a r o u n d t he c or ne r a nd b a c k "
i n t he o ut e r mo s t c or r i dor .
107. H OW T O DRA W A N E L L I P S E . I t is easy t o see t hat
t he ma j o r axi s i s j us t as l o ng as t he st ri ng, 10 i nches . Wh e n t he
pe nc i l poi nt is put on t he e nd of t he mi n o r axi s, i t is at t he mi d-
poi nt of t he st r i ng. H a l f of t he s t r i ng f o r ms t he hy pot e nus e of a
r i ght t r i angl e, of wh i c h t he l egs ar e t he s e mi - mi nor axi s a n d t he
di s t anc e f r o m t he c e nt e r t o t he f oc us . T h e l at t er di s t anc e is hal f
t h a t b e t we e n t he f oc i , or 3 i nches . H a l f t he s t r i ng is 5 i nches.
T h u s we h a v e t he f a mi l i a r 3- 4- 5 r i ght t r i angl e. T h e s e mi - mi nor
axi s is 4 i nc hes , a n d t he wh o l e 8 i nches .
108. ROA DS T O S E DA N. T h e g i v e n l i ne s egment s ar e
A B a nd C D . T h e p r o b l e m is t o d r a w a l i ne t hr o ug h poi nt X t hat
wo u l d pass t h r o u g h t he i nt er s ec t i on of A B a nd C D if al l t hr ee
l i nes we r e s uf f i ci ent l y e x t e nde d.
T h r o u g h X d r a w a ny t wo l i nes i nt er s ec t i ng bo t h A B a nd
C D . T h e i nt er s ect i ons ar e A , B, C , D. T h r o u g h A C a n d B D d r a w
t wo l i nes wh i c h i nt er s ec t i n E . T h r o u g h E d r a w a ny l i ne i nt er-
SOLUTIONS 185
s ect i ng t he t wo g i v e n l i nes i n F a nd G. C o n n e c t B F a n d D G ,
t hese t wo l i nes i nt er s ec t i ng i n H . T h e n X H is a t hi r d l i ne of t he
penci l wh i c h i nt ersect s a t t he s a me poi nt as A B a n d C D .
109. T H E BI L L I A RD S H OT . F e w of t he ma nu a l s o n bi l -
l i ards a t t e mpt t o e x pl a i n t he c a l c ul a t i on of angl es , a nd no ne e m-
pl oys t he r i gor ous me t h o d g i v e n her e. Wh i l e t he angl es ha v e t o be
l ear ned l ar gel y f r o m e xpe r i e nc e , b e i ng mu c h a f f e c t e d by s pi n a n d
f r i c t i on, st udy of t he g e o me t r y of t he i deal f r i ct i onl ess t abl e s houl d
be t he s t ar t i ng poi nt .
I ma g i ne t he t abl e ( t he r e c t a ng l e enc l os ed by t he c us hi ons )
t o be e x t e nde d i n e v e r y di r e c t i on by a n endl ess seri es of mi r r o r re-
flections, as i l l us t r at ed i n t he d i a g r a m on p a g e 186. I ma g i n e t he
t wo o bj e c t bal l s t o be c or r e c t l y pl a c e d i n e a c h r ef l ect i on.
A l l possi bl e b a nk shot s c a n be de t e r mi ne d by d r a wi n g a
st r ai ght l i ne f r o m t he a c t ua l c ue bal l t o t he r ef l ec t ed o b j e c t bal l s,
t o t he poi nt wh e r e a s i mpl e c a r o m c oul d be ma d e .
T h e d i a g r a m s hows 1 5 of t he possi bl e r ef l ec t ed t abl es. A l l
ot her s mo r e r e mot e ar e i r r e l e v a nt ; i n mos t of t h e m t he o bj e c t bal l s
ar e s cr eened by t he bal l s on a ne ar e r t abl e, as t he bal l s i n T a b l e 3
ar e s c r eened by t he bal l s i n T a b l e 6.
T h e l i ne so d r a wn is t he a c t ua l di r e c t i on i n wh i c h t he c u e bal l
mus t be sent o n t he a c t ua l t abl e. T h e n u mb e r of r ec t angl e- l i nes
crossed by t he l i ne of a i m s hows t he n u mb e r of c us hi ons t ha t t he
c ue bal l wi l l t ouc h be f or e s t r i ki ng t he o bj e c t bal l s. T h u s , t he shot s
i nt o T a b l e 9 a nd T a b l e 13 ar e one - c us hi on s hot s ; t hos e i nt o
T a b l e s 2, 7 a nd 12 ar e t wo- c us hi on shots.
O f t he f o ur t hr ee- c us hi on shot s, we wi l l r e j e c t t ha t i nt o
T a b l e 1 as needl essl y l o ng a n d t hat i nt o T a b l e 8 as t oo ri sky be-
c aus e t he c a r o m is so " t h i n . " T h e shot s i nt o T a b l e s 4 a n d 1 5 gi v e
exc el l ent angl es of a p p r o a c h f o r t he c a r o m; o ur f i nal s el ect i on is
4 bec aus e thi s shot is shor t er . T h e c o mpl e t e c our s e of t he c ue bal l
c a n be pl ot t e d on t he a c t ua l t abl e by pl a c i ng T a b l e s 9, 5 a nd 4 i n
t ur n u p o n i t i n pr o pe r or i e nt at i on a n d t r ans f e r r i ng t he l i ne seg-
me nt s ont o t he a c t ua l t abl e.
186 SOLUTIONS
112. T H E M I S S I NG DI GI T . T h e di gi t al r oot of 67 3, 1 06 is
5 ; t hat of 4, 783, 205, 468 is 2. Si nc e 5 X 2 = 1 0 , whos e r oot is 1, t he
r oot of t he p r o d u c t mus t be 1. T h e g i v e n di gi t s of t he pr o duc t s um
SOLUTIONS 187
to 69, whose root is 6. T o make the root 1, we will have to add 4.
H ence the missing digit is 4.
113. FI ND T H E S QU A RE . T he units digit of a square num-
ber can only be 1, 4, 5, 6, 9, or o. T his fact rules out the 2nd and
5th of the given integers. T he digital root of a square is always
I, 4, 7, or 9. T he roots of the 1st, 3rd, and
4
th numbers are re-
spectively 8, 5, and 3, so that they also are ruled out. Only the
6th, with root of 7, can be a square. (No doubt the reader will
like to verify that it actually is a square.)
115. S P E L L I NG A DDI T I ON. Of six solutions, the only one
which satisfies the condition that WOE SX5=SORE R is
T he values for D and E (
3
and 6) are interchangeable, also
for V and O (2 and 5). I find 8 different solutions, or 10 if we
include a couple in which A is allowed to have the value zero.
117. RE S T ORI NG T H E FI GU RE S . T he multiplicand is
74,369*053; the multiplier is 87,956.
118. L E T T E R DI VI S I ON. T he solution is given by sub-
stituting each digit for the letter below it:
9 4 0
7 3 9
8 9 2 5
1 0 6 0 4
116. A DA M A ND E VE .
T he solution with the maxi-
mum total is
8384
803
626
50
8
9871
2 9 7 1 3 5 4 8 0 6
A B C D E F G H J K
SOL U T IONS
119. C RYP T I C DI VI S I ON.
99 ) i9i7 ( '93
99
920
891
297
297
120. C RYP T I C M U L T I P L I C A T I ON.
'475
677
10325
10325
8850
998575
121. C RYP T I C S QU A RE ROOT .
3 ^ 9 4
\ j 10 20 16 36
9
1 20
61
59 16
56 61
2 55 36
2 55 36
SOLUTIONS 189
122. A C OM P L E T E GH OS T .
First division: 333 ) 100007892 ( 300324
Second division: 29 ) 300324 ( 10356
123. A GH OS T A DDI T I ON. T he key to this problem is the
fact that every triangular number has a digital root of i, 3, 6 or 9.
Since it is stipulated that the sum must be a triangle not divisible
by 3, its root must be 1.
M ake a tabulation of the ten possible sequences of five digits
and compute the digital root of the sum of each sequence.
SEQUENCE ROOT
I. 2, 3, 4, 5 6
2 . . . 2
3 7
4 3
5 8
6 . . . 3
7 - 7
8 . . . 2
9 . . . 6
o . . . 1
We have got to select two sequences whose roots sum to a
number whose root is 1. We can at once exclude the sequences
commencing with 1, 9, o, for lack of complementary sequences.
E vidently the two sequences chosen must have roots 2 and 8 or
3 and 7.
SEQUENCE ROOT SUM
2, 3, 4, 5> 6 2 20
3 7 25
4 3 30
5 8 36
6 3 30
7 7 25
8 2 20
190 SOLUTIONS
T o narrow the search further, write down the actual sums
of the seven remaining sequences.
Since the sums corresponding to like roots are the same, there
are actually only a few combinations to try. If we choose
sequences whose roots are 2 and 8, the final total will be 200-1-36
or 360-I-20: neither gives a triangle (refer to list of triangular
numbers in the A ppendix). If we choose roots 3 and 7, the total
will be 250+30 or 300+25. Only the latter gives a triangle, 325.
Since either of two sequences whose roots are 3 can be paired
with either of two whose roots are 7, there are in all four solu-
tions, as follows:
47 67 43 63
58 78
54 74
69 89 65 85
70
90
76
96
81 01 _87 07
325 325 325 325
125. T WO-DI GI T NU M BE RS .
(a) 36. (b) 27. (c) 25 or 36. (d) 54. (e) 29, 38, 47, 56.
126. T H RE E -DI GI T NU M BE RS .
(a) 198. (b) None, (c) 189. (d) 629. (e) 111. (f) 132,
264, 396.
127. P RI M E NU M BE RS . T he next three primes are 1009,
1013, 1019. T he simplest way to find them is to write out the
integers from 998 up, say to about 1020, and strike out those
divisible by the successive primes 2, 3, 5, 7, etc. T he even numbers
may be omitted in the first place. T hen those whose digital roots
are 3, 6, or 9 are struck out, also those that end with 5. T est for
i i , then try actual division by 7, 13, 17, 19, 23, 29, and 31. A
simple procedure is to divide each of these primes into 1,000; if
there is a remainder, add it to 1,000 to give an integer divisible
by the prime. C ount forwards and backwards from this integer
to strike out the others also divisible by the same prime. A fter the
SOLUTIONS 191
work has been carried through to 31, the three integers named
above are the only survivors.
It is not necessary to go higher than 31, because the square of
the next prime, 37, exceeds 1020. If any number on the list were
divisible by a prime higher than 31, then it would have another
factor lower than 31and all lower primes have been tested.
128. T H E S A L E ON S H I RT S . T he problem is to factor
60377. T he statement hints that one factor is less than 200. T he
terminal 7 can result only from factors whose terminals are 7 and
1 or 9 and 3. H ence the factors can be found by making trial
divisions with primes below 200 that end in 1 or 3 (or 7 or 9).
T he factors prove to be 173 and 349, both prime. C onsequently,
there must have been 349 shirts at $1.73 each.
129. A P OWE R P ROBL E M . T he 5th root of 844,596,301
is 61. T his problem can be solved in short order from the follow-
ing considerations:
(a) T here are nine digits in the given integer; its 5th root
consequently is greater than 30 and less than 100.
(b) T he terminal digit of the given integer is 1; its 5th root
must likewise end in 1.
(c) T he digital root of the given integer is 4; the digital
root of its 5th root must be 7 (as may be discovered by writing
out the roots of the powers of digital roots 1 to 9 inclusive).
130. T H E ODD FE L L OWS P A RA DE . T he number of Odd
Fellows in the parade was 367, the only integer under 497 that
satisfies the given conditions.
T he following is a general method for attacking all problems
of this sort, where an integer is defined by its remainders on
divisions by various primes. T he method is developed from ele-
mentary algebra. Special conditions may lead to short cuts, so
that the method does not necessarily have to be followed in full.
T he problem states that s, the total of Odd Fellows, is of form
s=3+i =5H -2=7<;-f3= 1 IC H -4 (I)
192 SOLUTIONS
where a, b, c, d are unknown integers.
From ( I ) , find b in terms of a.
I 3aI
(2)
5
Since b is integral, the numerator of the fraction must be
exactly divisible by 5. By trial we find that the lowest value of a
that serves is 2. C onsequently a must be of form
a =2+5 * (3)
where x is an unknown integer. From (2) and (3)
b= 1+3* (4)
A ny value of x will give an integer that satisfies the conditions
as to a and b. But we also have c and d to reckon with. C ombine
(1) and (3) to find c in terms of x:
4+1 5* , .
(5)
By trial we find that the lowest value of x that makes c
integral is 3. C onsequently x is of form
*=3+7 y (6)
where y is an unknown integer. Replacing x by its value in terms
of y, we derive from (3) (4) (5) three new equations:
a =i 7 +35y
b=io-\-2iy (7)
7+' 5y
Finally, we have to reckon with d. From (1) and (7) we
find that
(
8)
from which it follows that y is of form
y=3+112 (9)
where z is an unknown integer. By substituting its value in terms
of y in (7) and (8) we derive a final set of equations
0=122+3852
b= 73+2312 . .
a (10)
C= 52+1652
D= 33+
I 0
5
Z
SOLUTIONS 193
A ny value of z will give a set of values for a, b, c, d that will
satisfy the problem of the Odd Fellows, but since the answer is
limited to $<497, we must set z equal to zero and solve for s
in (1).
131. T H E U NI T Y C L U B. A short cut is available to solve
this problem. T he division of the U nity C lub marchers by 3, 5, 7,
11 in each case left the array 1 short of having even ranks. T he
least number in the group was consequently the product
3X5X7XI I , less 1, or 1,154.
132. C I NDE RE L L A T OA S T E RS . T he task is to find the
G.C .D. (greatest common divisor) of 389,393 and 831,119. It is
not necessary to factor the two integers; G.C .D. can be found "by
force." Divide 389,393 into 831,119 to give quotient 2 and re-
mainder 52,333. Divide the remainder into 389,393 to give
quotient 7 and remainder 23,062. C ontinue in the same way,
dividing the remainder each time into the previous divisor. T he
first divisor that gives zero remainder is the required G.C .D. By
this method we find the G.C .D. of the given integers to be 887.
Since this integer is prime, the price of the toaster must be $8.87,
and the number sold at the M ain Street Branch is 389393/887
=439-
133. S QU A DS A ND C OM P A NI E S . T he Numerian army
comprises at least 242,879 men. T his number is the L .C .M . (least
common multiple) of 1,547
a n
d 34,697. T he way to find it is to
determine the G.C .D. and multiply the latter by the other factors
contained in the two integers.
134. T H E M I S RE A D C H E C K. T he check was for $51.24.
C all x the dollars and y the cents. T hen the amount of the
check (in cents) was ioox-f-y. T he teller first gave iooy-\~x,
which was $1.11 short of half the correct amount. H ence
194 SOLUTIONS
i oox+y=2( i o o y +x +m)
98*= 1995+222
( 0
(*)
E quation (2) is a Diophantineit would have an infinity
of solutions but for the fact that x and y must be integers. A
further limitation is that both integers must be less than 100.
T he following method used to solve equation (2) can be ap-
plied to many linear Diophantines.
A factor common to all terms of an equation but one must
be a factor of that one as well. In (2) the first and last terms are
even: hence y also must be even. Replace y by 2z.
T hjs reduction of the equation to lowest terms is not actually
necessary, but with more complex examples it often proves a wel-
come simplification.
Divide (3) by 49, the coefficient of x.
In order to make x integral, we must find a value for z such
that 32+13 is a multiple of 49. T he lowest value that satisfies is
z= 12, whence * =5i , y=24.
One solution, then, is $51.24. Is any 6ther solution possible?
E xamine (3). When zi, x is about 6. T his means that x must be
about 6 times z. Now the maximum possible value of * is 99, and
99
of z, consequently, or about 16. From equation (4) we see
that the next-higher value of z, beyond 12, that satisfies the equa-
tion, is far beyond 16. H ence the solution $51.24 is unique.
135. T RA NS FE RRI NG DI GI T S . T he required number is
enormous, but it can be found by "brute force."
Since we do not know how many digits there are in the re-
quired integer, we will represent them by A , B, C . . . as read
from right to left. T hen the integer is of form
A +10B+100C +i o" -
1
Z (1)
where n is the number of digits.
49*
= I
99H - m (3)
(4)
SOLUTIONS 195
L et us take A as the terminal digit to be transferred. When it
is put at the other end, the integer becomes
B + I OC + J OOD. . . . + I O" -
2
Z +I O
,
" -
1
A ( 2)
T he stipulation is that (1) is to be % of (2). (Remember
that the digits are represented in reverse of the way they are
written.) T hen
A + 1 0 B + 1 0 0 C . . + 1 0 " - ^
= %( B + i o C . . . + i o ^
2
Z + i o
n
-
1
A ) (3)
C learing of fractions and expanding, we have'
5 A + 5 0 B + 5 0 0 C . . + 5 ( 1 0 " - ^ )
= 4 B + 4 O C . . + 4 ( I O " -
2
Z ) + 4 ( I O " - I A ) ( 4)
Now collect the A terms on the right, all other terms on the
left:
4 6 ( B + I O C . . . + i o " -
2
Z ) = A [ 4 ( i o " - i ) - 5 ] ( 5)
From (5) it follows that the right-hand member is divisible
by 46. In other words, we must find values for A and n such that
A [
4
( i o " - ' ) -
5
]
2X23
will be integral. Since the expression in brackets is odd, it is not
divisible by 2; therefore A is divisible by 2, and we can write
A =2, 4, 6 or 8 ( 7)
Since A is not divisible by 23, the expression in brackets must
be. T he expansion of this expression for values of n 1, 2, 3. . .
gives 35, 395, 3995, etc. T o find the first of these terms divisible
by 23, set up a long division in form
23 ) 399 95 ( 17-
23
169 (8)
164
59 etc.
Bring down 9 from the dividend each time, until a remainder
of 11 is reached, so that the final 5 can be brought down (since
196 SOLUTIONS
115=23X5)- T his turns out to be a lengthy matter, but is mere
arithmetic. T he smallest quotient obtainable is
73) 9*3* 43>
0
78, 260, 869, 565 (9)
By taking A =2, we have the smallest integer that satisfies
the conditions:
2, 173. 9*3. f>43, 78, 260, 869, 565 (10)
T hree other answers can be obtained by setting A equal to
4, 6 and 8. In each case, as is seen from (6), number (9) has to
be multiplied by half of A to make up the balance of the integer.
136. FI GU RA T E NU M BE RS . T he reader should perceive
that by its very derivation any number in the P ascal triangle is
the sum of the number above it and all to the left of that number
on the same row. T he sum of the first 25 terms of the 4th order is
therefore the 25th term of the 5th order. T his term can be found
by the formula; it is 20,475.
137. L I T T L E WI L BU R A ND T H E M A RBL E S . T he smallest
possible number is 210. It is easily discovered by empirical
methods. T he integer must be divisible by at least six different
triangular numbers (not including unity). It is therefore the
L .G.M ., or a multiple thereof, of the six triangles. If we compute
the L .C .M . of the smallest triangles, we see that we cannot do
better than start with the factors 2, 3, 5, for out of them we can
get the triangles 3, 6, 10, 15. Since 30 is not itself a triangle, we
will choose a multiple of 30. T he lowest multiple that is a triangle
is 120, but out of this number we cannot make any additional
triangles. T he next multiple is 210, and this is found to serve, for
besides being divisible by 3, 6, 10, 15, it is also divisible by 21
and 105.
138. H OKU M , BU NKU M A ND FA T U I T U M . T he "triangu-
lar hexahedron" consists of two triangular pyramids set base to
base, one having an edge one unit greater than the other. T he
problem is then to find the four successive triangular pyramids
SOLUTIONS 197
A , B, C , D, such that the "hexahedrons" A -fB, B+C , and C + D
total 31,3 u . T hen
A +2B+2C +D=
3
I ,
3
i i ( , )
L et n be the edge of the smallest pyramid, A . T hen by
formula the number of its units is
w( n+i ) ( n+2)
T he edge of B is n+i , of C , n+2, etc. Write equation (i ) in
terms of n, expand and collect terms, thus deriving the equation
6n
3
+45n
2
+i 2
3
n+i 20
G =3
I
>3
1 1
(3)
T his equation may be reduced to
n
(
n 2 +
I 5 ^
= 3 I
,
2 9 I ( 4 )
Instead of troubling to solve the cubic equation (3), we can
find n by trial, since (4) shows it to be a factor of 31,291. T he
prime factors of this integer are 13, 29, 83. From (4) it is clear
that n
3
is somewhat less than 31,291. T he approximate size of n
can be gauged by comparing io
3
and 80
3
with 31,291. T he first is
too small, the second too large. T he choice for n thus falls on 29,
and this number proves to satisfy (4).
T he values for A , B, C , D are then respectively 4495, 4960,
5456, 5984. But the solver need not compute these numbers unless
he wishes to check his work. A ll that is asked are the "powers" of
H okum, Bunkum and Fatuitum, and for n29 these powers are
30, 31, and 32.
139. S QU A RE NU M BE RS . T he squares can be derived by
summing each pair of adjacent numbers of the third figurate
order. E very square is the sum of two consecutive triangles, as is
clear from the diagram. For analytical proof merely expand
n(n-|-i) ( ni ) n
H
2 2
to show that it is equal to n
2
.
198 SOLUTIONS
o \ o o o o
O 0 \ Q O O
O O Q \ D O
O O O 0 \ 0
O O M O
140. S QU A RE -T RI A NGU L A R I NT E GE RS . One way to solve
the problem would be to extend the tables of square and triangu-
lar numbers until two more identities are found. But this purely
empirical approach can be short-circuited. A little analysis of the
nature of the desired integers will show that they can be deduced
by a search of the limited tables in this book.
T he problem is to find integral values that satisfy the equa-
tion
= y
2
( 0
One of the two integers, x or x-f-i, must be even.
If * is even, it may be replaced by 2n, and (i ) becomes
n( 2n+i ) =y
2
(2)
If x-j-i is even, it may be replaced by 2 n, and (1) becomes
n(2n 1) y
2
(3)
For convenience we will combine (2) and (3) as
n(2n 1) =y
2
(4)
E quation (4) states that y
2
is the product of two factors, each
necessarily different from y. It follows that y must comprise at
SOLUTIONS 199
least two factors ab, and that *
n=a
2
2ni=b
2 [5)
T he desired integers can be determined by searching our
table of squares for pairs of integers that satisfy (5).
T he search can be narrowed by consideration of the terminal
digits. T hus:
If the terminal of n is o 1 4 5 6 9
then the terminal of 2n-\-1 is
1
3 9
1
3 9
then the terminal of 2 nl i s 9
1
7 9
1
7
We can eliminate the cases that result in 3 and 7, since
neither digit can be the terminal of a square. T hen we need only
examine squares ending in o, 4, 5, 9 (*=n) to see if 2W+1 is a
square, and squares ending in o, 1, 5, 6 ( =n) to see if 2n1 is a
square.
We need go no higher than 29
s
to find four different squares
that satisfy the conditions for n in (5).
n 2n-\-i 2 n1 Y
2
TRIANGLE SQUARE
4 9 36=8.9/2 = 6
2
25 49 1225=49.50/2 = 35
2
144 289 41,616=288.289/2 = 204
2
841 1681 1,403,721 = 1681.1682/2=1189
2
* T he reader may ask why a different distribution of the factors is not
possible, e.g.
n=ac'
ani=ab*
It is easily proved that n and 2 n i cannot contain a common factor,
T he above equations can be written as
(6)
a a
E quation (6) states that a is a divisor of n. T hen the first term of ( 7)
is integral. A lso, b" is integral. But the term cannot be integral, since
1. H ence (6) and (7) cannot be reconciled, and it follows that n and
2 n i are relatively prime.
200 SOLUTIONS
141. P A RT I T I ON OF A T RI A NGL E . T he truth of the prop-
osition can be seen intuitively from the diagram. H ere two differ-
ent triangular numbers are represented by triangles of cannon
balls. E ach can be dissected into three triangles, of which the outer
three are equal, while the central triangle is the next-smaller or
next-larger according as the side of the large triangle is even or
odd.
o
o o
o o o
C 3
O ' O
O G w w O O
O
O O
O O P
SOLUTIONS 201
T he central triangle plus one of the outer ones equals a
square. T he large triangle is thus the sum of this square and the
two remaining smaller triangles.
For analytical proof, show that the sum of the square of n
and two triangles of ni is n(2ni), which can be written
2Tl(2Tli) . . . . . , , ., .
and thus is seen to be the triangle of 2 n1. Similarly
2
show that the sum of the square of n and two triangles of n-f-i is
2n(2n-\-i)
n(2n-\-i), which can be written and so is seen to be
2
the triangle of 2n. Now, the two expressions 2ni and 2n give the
complete series of integers as n varies from unity to infinity, so
that their triangles comprise all of the triangular numbers. H ence
any triangular number p can be partitioned into (a) a square of
p/2 and two triangles of (p/2)1; or (b) a square of (p-\-1)/2
and two triangles of ( p1) /2.
142. T H E BA T T L E OF H A S T I NGS . T he problem is to find
two squares that differ by 512"half a thousand footmenne and
full douzaine more of knights."
T he general problem of finding integral values for
X
2 _ yi
C
is solved by factoring the left side into (x + y) (x y), then
equating these two terms with pairs of factors of C (a given
integer). E ach different pair of factors yields a different solution.
T he available pairs of factors of 512 ( = 2) are 256 and
2, 128 and 4, 64 and 8, 32 and 16. (T he pair 512 and 1 is not
available, for the reason that the factors must be both even or
both odd, since their sum is 2x and difference 2y, even in both
cases.) T hese factors generate the following solutions:
202
SOLUTIONS
129
2
i27
2
= 16641-16129=512
66
2
62
2
= 4356 3844=512
36
s
28
2
= 1296 784=512
2
4
2
8
2
= 576 64=512
A ny one of the sets would answer to the description of the
two armies so far as concerns the difference of 512. But we are
further told that after slaying half the foe with the loss of "only a
few score" of their own men, the Saxons reduced the armies to
equality. T he only numbers that satisfy this statement are 1296
(Normans) and 784 (Saxons).
143. T H E DU T C H M E N'S WI VE S . T he amount spent by
each individual is a square number, and the difference of the ex-
penditures within each family is 63 shillings. T he first step is to
find three sets of squares that differ by 63. T he method is ex-
pl ai ned in No. 1 4 2 T h e Battle of Hastings. T h e requi red num-
bers are
3
2
2
312=63
12
2
9
2
=63
8
2
i
2
=63
T he integers in the first column represent expenditures by
the husbands; in the second column, by the wives. Now we have
to pick out the integers that differ by 23 and 11. It is easily seen
that A nna (31) is the wife of H endrick (32); Katrun (9) is the
wife of E las (12); Gurtrun (1) is the wife of C ornelius (8).
SOLUTIONS 203
144. T H E C RA Z Y QU I L T . T he quilt made by M rs. T homp-
son and E ffie together was made by combining the pieces of two
smaller squares. Subsequently, the large square was dissected into
two smaller squares, different in size from the original compo-
nents. T he problem is then to find integral values that will satisfy
the equation
A
2
+B
2
=C
2
+D
2
==E
2
(i )
Since there must be an infinity of solutions, we will seek that
for which E is the smallest possible integer.
T he equations (I) may be written
E
2
A
2
=B
2
E
2
C
2
=D
2
(
2
)
Refer to the method used in the solution of No. 142
The Battle of Hastings. T he first of the equati ons (2) may be
written:
(E + A ) (E - A ) = B
2
(3)
T he right-hand term must contain two equal factors, besides
the equal factors B. T hen B itself must be composite, say with
factors r and b, and the terms of (2) can be equated:
E + A = rb
2
, E - A = r (4)
Solving for E and A we have:
^ __ r(b
2
+ 1) r(b
2
- 1)
2 2
By similar operation on the second of the equations (2),
we derive:
= ? ( * + ' > (6)
204 SOLUTIONS
where q and d are the factors of D. C ombining the values of
E given in (5) and (6) gives:
r{b- + 1) = q(d
2
+ 1) (7)
T he problem of finding the minimum solution of equations
(1) resolves itself into finding the mihimum solution of (7) . But
in the nature of the problem, no solution of (7) is satisfactory
unless it gives four different values to A , B, C , D. T his in turn is
possible only if the values of r, b, q, d are all different. But then r,
for example, must be a factor of the right-hand member q(d'
2j
r
i)
different from either q or (<i
2
-|-i). C onsequently this member
must contain yet another factor x, and we have two alternatives:
b
2
+1
E ither q=xr and thus x (8)
or d
2
-\-i xr and thus x= ( 9
9
T he rest is a matter of experimenting with the lowest possible
integers that will satisfy either ( 7 ) or ( 8 ). It will be found that
the minimum solution is given by the following values in connec-
tion with (8) :
X=JO, r = 5 , 6 = 3 , q = i , d=7
T he values generated in equation (1) are A =2o, B=i 5 , C =24,
D=7 , E =25- T hat is,
2 o
2
-|-15
2
=400+2 2 5=6 2 5
24
2
+ 7
2
=57&+ 49=625
SOLUTIONS 205
T he next-higher possible value of E is considerably more than
625, as the reader will no doubt realize from the empirical step
in this solution. A quilt composed of 625 3-inch squares measures
6 feet 3 inches on each side (no allowance for overlap). T his
seems "right for the four-poster bed," and the next-larger size
would be absurdly large. We may justly conclude that The Crazy
Quilt intended is the minimum of 625. T he wording makes clear
that E ffie took away 24
s
of the combined 25
s
quilt, leaving M rs.
T hompson with 7
2
, which measured only 21 inches on each side.
145. T H E FOU R T RI A NGL E S P ROBL E M . T he problem is
to find four right triangles such that (a) all four have one side of
a given length k; (b) all sides of all triangles can be expressed in
integers; (c) the total of the four perimeters is the minimum pos-
sible.
If p and q are the legs of a right triangle and k is the hypote-
nuse, then p
2
-\-q
2
kSince the side of the given square k may be
used as either hypotenuse or leg of any of the triangles, our task is
to find at least four different integral sets of solutions for the equa-
tion
k
2
=p
2
q
2
(1)
L et us first try to find the minimal solution where k is
always a leg, never a hypotenuse. T hen we need deal only with
the minus sign in (1).
A s is shown in No. 1 4 4 T h e Crazy Quilt, to obt ai n mul -
tiple solutions of (1) we must make k composite. If we make
it even, we must make it doubly even, that is, divisible by 4.
If we choose 4 as the minimal even factor, we find that the
206 SOLUTIONS
least other factor available, to give four different pairs of
factors for k
2
, is 3.
T hese pairs are:
k
2
= 12
2
= 2 X 72
= 4 X 36
= 6 X 24
= 8 X 18
T he corresponding triangles are shown in the diagram:
37-35-J2; 20-16-12; 15-12-9; 13-12-5.
T he number of matches required is the sum of these num-
bers, 198.
L et us examine all other possibilities to see if this is a
minimal solution. Suppose we take k
2
as odd. T hen the least
factors we can assign to k are 3 and 5. T he available pairs of
factors of 225 are 1 and 225, 3 and 75, 5 and 45, 9 and 25.
T hese generate the triangles:
113-112-15; 39-36-15; 25-20-15; 17-15-8.
Obviously this solution has a much larger perimeter than the
first. We can economize by replacing the largest triangle with
1512-9, where the side of the square is used as a hypotenuse
of the triangle. But even then, the number of matches re-
quired is 226.
SOLUTIONS 207
Finally, let us consider the possibility of making k a hypote-
nuse more than once. T hen equation (1) becomes:
k* - pi =
q
* (
3
)
T hen we have:
2k = a(b
2
+ 1), 2p a(b
2
- 1) (4)
where q = ab, a composite number. T he least value we can
assign to b is 2, making k at least 5. T his gives the familiar
543 triangle. T o obtain a second solution for any given k,
we will have to make k composite, with a least factor 5.
Obviously any such solution would create a larger perim-
eter for the figure than the solution k 12.
147. T H E FI VE -S U I T DE C K. H erewith is given the table of
partitions carried up to n16. A s applied to the five-suit deck,
it shows that there are 97 patterns of bridge handsthe par-
titions of 16 into from one to five parts inclusive, which total 101,
less 4 that cannot occur
1 6 0 0 0 0
15 1 o o o
14 2 o o o
14 1 1 0 0
If a small portion of the table is made by writing out the
actual partitions and counting them, this fact will be realized:
If we are to partition an integer, say n , into 3 parts, we
must start by putting at least 1 unit in each part: 1 1 1. T hat
leaves 8 units to be distributed in all possible ways. But we have
already tabulated the ways of partitioning 8 into 1, 2, or 3 parts.
A ll partitions of 11 into exactly 3 parts are given by this total of
ways to break up 8 into not more than 3 parts and then add them
into our base 1 1 1;
T he correct entry to make in the cell, column 11, row 3, will
therefore be found by summing the entries in column 8 down to
row 3 inclusive. Or, to generalize, the entry in column n, row r,
is found by summing column n-r from top down to row r in-
clusive.
208 SOLUTIONS
T o fill column 11 in order from top to bottom, we must sum,
in order, portions of columns i o, 9, 8 . . . 2, i. T he terminal cells
to be included in each summation lie on a minor diagonal ex-
tending downward-left from cell, column 10, row I. A nd so for
any column n: to write the entries from top to bottom we sum
successive columns leftward, down as far as the minor diagonal
that meets the top edge in column n- i .
Of course the minor diagonal sooner or later crosses the prin-
cipal diagonal, and beyond the intersection we must sum each
column entire down to the principal diagonal. A fter we have
written 10 in column 11, row 5, by a summation of most of
column 6, we find that the rest of the numbers to be written are
identical with those at the bottom of column 10.
SOLUTIONS 209
It follows that along a major diagonal (parallel to the prin-
cipal diagonal) there must come a point sooner or later where
all subsequent entries are identical. T he first of such identical
numbers always occurs in an even column, and the row number is
half the column number.
T he integers at which the major diagonals begin to repeat:
!>
2
) 3j 5> li
1 1

a r e
the same as the totals of partitions as n
increases from i up. T he reason for this circumstance is patent
from the way in which the table is built.
A point of practical importance in making the table is that
it is not actually necessary to sum columns of figures from the
minor diagonal up. T he correct entry for column n, row r, can be
found by summing just two cells: column n-r, row r (the cell
where the minor diagonal crosses the row) and column n- i , row
n-1 (the cell diagonally adjacent at upper left). For example, in
column 16, row 7, the integer 28 is determined as the sum of the
adjacent 26 and the 2 in row 7, column 9.
148. ORDE RS OF I NFI NI T Y. L et the integers in ascending
order, 2, 3, 4 . . . be the indices of certain sub-groups. E ach sub-
group is to be composed of fractions, with the numerator and
denominator of each fraction summing to the index. For ex-
ample, under index 5 we write the fractions: %, %, Since
the fractions are formed by partitioning each index into two parts
in all possible ways, and since the number of such partitions of
any integer is finite, the number of fractions in each sub-group
will be finite.
Within each sub-group the terms can be arranged in any con-
sistent order, e.g., with numerators in descending order of magni-
tude.
Since the indices 2, 3, 4 . . . go on to infinity, the totality of
sub-groups so formed must embrace all the rational numbers, yet
the arrangement is discrete.
149. T U RKS A ND C H RI S T I A NS . T he diagram shows the
necessary arrangement. T he black circles represent the C hristians.
210 SOLUTIONS
C ounting the indicated man as "one," it will be found that the
T urks will be counted out first, in the order indicated by the
numbers.
T he puzzle can of course be solved purely mechanically.
P lace 30 dots in a circle, mark your starting point, decimate by 13,
and note which are the first 15 of the dots to be counted out.
START HERE _

(7)
(ffi
( u u
P uzzle makers have expended considerable ingenuity in de-
vising decimation problems that cannot be solved mechanically.
T he remaining puzzles in this chapter are selected to show some
different methods that have been devised.
SOLUTIONS 211
150. BOYS A ND GI RL S . T he decimation for the girls
must first take out Nos. i, 2, 3, 7, 10of course, not necessarily
in that order. T he decimation for the boys must first take out
Nos. 4, 5, 6, 8, 9. T he problem can be solved by writing out the
order of decimation of a circle of 10 by intervals of 2, 3, 4, etc.
C omparatively few figures, however, need be written. A s soon as
any decimation is found to contain, within its first five numbers,
integers from both of the above groups, that interval may at once
be abandoned as unsuitable. It does not take long to discover that
the decimation to take out the boys is 14, for the girls 13.
151. NA T I VE S A ND BRI T ONS . T he two decimations must
in one case first count out Nos. 2, 4, 5, 7, 9 and in the other Nos.
1, 3, 6, 8, 10. T he problem can be solved by writing out and ex-
amining the first five numbers of the decimation of a circle of 10
by 2, 3, 4, etc. But, since we are free to choose any starting point
in each case, we shall have to determine for each set of five num-
bers whether it can be made equivalent to one of the above groups
by a cyclical substitution. T he easiest way to do this is work the
whole solution geometrically. Draw a circle, mark and number 10
points on it, and mark them as natives or Britons by noughts and
crosses, in the given pattern. P lot each separate decimation, by
2, 3, 4, etc., on a separate circle, marking simply the first 5 points
reached. C ompare each such decimation with the original pattern,
to see if in any position of rotation the 5 marked points can be
made to coincide with the five natives or the five Britons.
It will be found that decimation by 11 gives the series 1, 3, 6,
io, 8 . . . T his will serve the chief to count out the Britons first,
if he starts counting at No. 1. T he decimation by 29 gives
9, 1, 6, 7, 4 which is equivalent to 7, 9, 4, 5, 2. T his was the
interval the Britons had in mind, with the count commencing on
the native No. 9 as "one."
152. J A C K A ND J I L L . T he number by which both circles
were decimated to leave Jack and Jill to the last was 11. T he solu-
tion is a matter of writing out the decimations of a circle of 5 and
212 SOLUTIONS
a circle of 4 by 2, 3, 4, etc., until the first interval is encountered
that leaves No. 5 and No. 4 respectively to the last.
153. OU T A ND U NDE R. L ike No. 149Turks and Chris-
tians, this puzzle can be solved mechanically. P lace the cards face
up on the table in the desired final order, then pick them up back-
wards. T hat is, take up one card, face down. T ake up another,
place it face down on top of the first, and then transfer the
bottom card to the top. C ontinue in the same way until all the
cards have been picked up. T he order of the cards, from top to
bottom, will then be: A Q 2 8 3 J 4 9 5 K 6 1 0 7 .
T he questions asked in No. 154The Nightmare similarly
could be answered by working backwards, but I have taken the
precaution of making the numbers so large that solving "by hand"
is scarcely feasible.
154. T H E NI GH T M A RE . On the first run through the
deck, all odd cards are thrown out, leaving 485 even cards from
No. 2 to No. 970 inclusive. On the second deal, since No. 971 was
put out, No. 2 will go under and No. 4 will go out. A ll the rest
that go out will be multiples of 4. T o put it in tabular form:
T he variable n is to take all values o, 1, 2. . . up to the high-
est practicablebeyond which point the term defines numbers
higher than any card remaining in the deck. Do not overlook
that n must start at zero.
We can build up on this plan a tabulation of what numbers
go out and what remain on every deal until the deck is exhausted.
A lthough the deck is in effect a circular arrangement, we have no
difficulty in distinguishing what we call the separate "deals." A
deal begins whenever the cards are in order of ascending magni-
tude from the top of the deck down, and ends when they next
revert to that order.
1st deal:
2nd deal:
Nos. that go out
i-F-2n
4+4 n
Nos. that go under
2+271
2+471
SOLUTIONS
213
I give the complete table for the deck of 971 cards. T he
manner in which it is built up enables us to generalize; we can
make a table for a deck of any number of cards by following these
rules of operation:
Deal no. Cards in deck Nos. that go out Nos. that go under
1
97
1 1 + 2 n 2 + 2 n
2
485 4 + 4
n
2 + 4 n
3 243
2 + 8 n 6 + 8 n
4
121 1 4+ 1671 6 + 16 n
5
61 6 + 32n 22+ 3271
6
3o 5 4+ 6472
22+ 6471
7 ' 5
86+12872 22+12871
8 8 22+256 n
1
50+25671
9 4
150+5*271 406+51272
10 2 406 918
11 1 918
I St 2nd 3rd 4th 5th
1st column: A t top write total number of cards in the deck.
E ach other entry is half of the entry immediately above it; in deal-
ing with odd numbers, alternate in writing the smaller and the
larger half, writing the smaller half on the first occasion an odd
number is encountered.
3rd and 5th columns: T he coefficients of n are successive
powers of 2, commencing with 2
1
.
4th column: E ach entry either repeats that just above it or
changes to total of coefficients in 4th and 5th columns on row
above. It repeats (a) when 1st column shows the smaller half of
an odd entry above it, and (b) when 1st column shows exact half
of an even entry above it, and on said row above, the entry in
2nd column is larger than the entry in 4th column. When con-
verse conditions apply, 4th column changes.
2nd column: When 4th column repeats, entry in 2nd column
is the sum of the coefficients in 4th and 5th columns on row
above; when 4th column changes, entry in 2nd column repeats
previous entry in 4th column
214 SOLUTIONS
From the tabulation we see the answer to the demon's first
question: T he last card to come out is No 918.
T he second question is: When does No. 288 come out? T o
answer this we must find the algebraic form in the GO OU T
columns that define 288. First ascertain how many times 2 is con-
tained in 288; 288=32X9- L ook for an algebraic form divisible by
32. T here is none, which means that the cards that are multiples
of 32 do not monopolize a deal. T ry 16, the next lower power of
2. A gain we find that multiples of 16 do not monopolize a deal,
nor do multiples of 8. It is not until we get down to 4 that we find
a factor common to 288 and to one of the forms in the GO OU T
columns. T he second deal takes out numbers of form 4+4", and
this is the only form that fits 288. If we write 288=4-1-471, then
71 = 71. Since the n values start at zero, 71 is the 72nd value it
takes, and No. 288 is the 72nd card cast out on the second deal.
A s 486 cards go out on the first deal, No. 288 goes out as the
486-4-72 or 558th card.
T he third question is: What is the 643rd card to go out?
Since 486 cards go out in the first deal, the 643rd card out will be
the 643486 or 157th of the second round. For this ordinal num-
ber n= 156. T hen 4~f-4( 156) =628, the number on the 643rd card
cast out.
A nyone with a passion for analysis may be interested in de-
riving formulas that will answer questions of the three types above
without the necessity for making a table. H ere as a beginning is
something he may like to verify. If s is the total of cards in the
deck, then the number f on the last card to go out is
/ = 2f 2
n
where 2" is the highest power of 2 less than 2s. For example, for
the deck of 971 cards, /= 19421024=918.
156. A C OM M ON M I S T A KE . Smith and Jones took no
account of the ways in which suit-names can be attached to the
digits of the pattern to make different combinations. A ccording
to their computation, the following two hands are identical,
whereas obviously they are different:
SOLUTIONS 215
A A K 7 3 2 <v> Q 10 9 2 0 6 5 * J 8
W A K 7 3 2 O Q 1 0 9 2 + 6 5 A J 8
E ach of their two numbers must be multiplied by a coefficient
representing the number of ways in which the names of the four
suits can be permuted against the pattern. T his number varies ac-
cording to the number of identical integers in the pattern.
For 5 4 2 2, we have a choice of 4 names for the suit of 5;
of 3 remaining names for the suit of 4; then there is no further
choice. T he coefficient here is 4 X 3 = 1 2 .
For 5 4 3 1 , which has no like digits, the coefficient is
4 X 3 X 2 = 2 4 . T he ratio between the two coefficients is 1 \ 2 in favor
of pattern 5 4 3 1 . C onsequently the ratio between the two types
of hands is not 3 6 : 2 2 but 3 6 : 4 4 or 1 1 : 9m favor of the 5 4 3 1
pattern.
157. T H E A NA GRA M BOX. T here are 1 , 578, 528, 000
ways. In dealing with the A 's, be sure to reckon the permutations
of 3 out of 30, not merely the combinations. T he puzzle asks
"H ow many ways to pick out and arrange," implying that each
different permutation of the same three A 's among the three posi-
tions is to be considered a different arrangement.
158. M I S S I S S I P P I . T he number of permutations is 508,-
722, 691, 276, 800.
159. P OKE R DI C E . T he table is as follows:
Five of a kind
Four of a kind
Full house
Straight
T hree of a kind
T wo pairs
One pair
No pair
6
1 50
300
240
1200
1800
3600
480
6
5
= 7 7 7 6
216 SOLUTIONS
T he reader may have noticed that card probabilities are
reckoned from the combinations of 5 out of 52 cards, whereas I
spoke of 7776 as the number of permutations that could turn up
on 5 dice. T his distinction points the way to what is in my opinion
the most accurate way to reckon the dice odds.
We can deal with combinations of 52 cards because each card
has a separate and unchanging identity. We cannot so deal with
5 dice, because each die can have 6 different identities. Nor can
we reckon as though the dice were 6 sets of 6 cards each, num-
bered in each set 1, 2, 3, 4, 5, 6a total of 36 cards. T o do so
would fall into the danger of counting impossible combinations,
e.g., an ace on one die combined with a deuce on the same die.
T he safe way to count is first reckon the number of combina-
tions of five integers out of 1, 2, 3, 4, 5, 6 (allowing repetitions)
that satisfy the definition of the hand, then multiply this figure by
the number of ways any one combination can be permuted on the
upper faces of five dice of separate identity.
T o illustrate the process, let us take the full house. T he triplet
may be in any one of 6 denominations, after which the pair may
be any one of five. T he number of combinations is 6X5=30. E ach
combination can turn up on the dice in several ways, for example
(naming the dice by letters) :
A B C D E
4 4 4 3 3
4 4 3 4 3
4 3 4 4 3
T he factor for the permutations of each combination is calcu-
n!
lated by the formula P *
b
= T he symbol on the
a! b!. . .
left side means "the number of permutations of n objects, of
which a are of one kind, b of another kind, and so on, the sum
of a-\-b . . . being n."
5!
In the case of the full house, the factor is given by -- = 10,
3! 2!
and the total number of full houses is 30X10=300.
SOLUTIONS 217
In the case of the straight, there arc only two combinations,
low straight and high straight; each can be permuted in 5! ways;
hence the number of straights is 2X51=240.
It will be noticed that the ranking of hands in poker dice is
incorrect, if the basis of ranking be conceded valid. "No pair"
should rank just after full house and the straight should rank just
ahead.
160. T H E M I NI M P U Z Z L E . T he total number of paths is
172. T his number counts as different the two directions in which
the same five cells can be traversed, as is implied by the statement
of the problem.
T he puzzle is merely an exercise in orderly counting. Reckon
the number of paths commencing at one M of each type: U pper
left corner, 22; upper right corner, 10; central, 44; intermediate
diagonal, 32. T here are two M 's of each type except the central, so
the grand total is given by
2 (22+10+32)+44= 172
161. T H E S P Y. A ny given street intersection X can be
reached only from the adjacent intersection (if any) to the north
or the adjacent intersection (if any) to the west. T he number of
ways to reach X is the sum of the ways to reach these two adjacent
intersections. A s this proposition is perfectly general, we can start
with the intersection just inside the northwest gate and write the
number of routes to each other intersection by a summation
process. In doing so the reader will quickly discover that he is
compounding "P ascal's triangle," given under No. 136Figurate
Numbers. H e can simply turn to that tabulation, mark off a
rectangle in the upper left which is 1 o columns wide and 9 rows
deep, and look for 715 in that area. T he integer appears only once
the 10th cell of the 5th row. H ence, the spy's office was located
in the street adjacent to the east wall, at its intersection with the
fifth street down from the north wall.
218 SOLUTIONS
162. H OW M A NY T RI A NGL E S ? T he following is not the
only way to count the triangles, but it is offered as an example of
orderly procedure. Without a rigorous procedure, it is easy to miss
some triangles or count some twice over.
T he 57 points of intersection are labeled with letters, all
points of the same letter standing at the same distance from the
center, B.
T he triangles can be grouped into four classes, according as
the number of A points in each is three, two, one, or none.
SOLUTIONS 219
THREE A POINTS
A s every A point is connected with every other, any three
A points represent a triangle. T he number in this class is
8!
thus G
8
, which is =56.
3
3! 5
!
T wo A POINTS
H ere we shall have to take account of four sub-groups, ac-
cording to the relationship of the two A points. In every case A
3
will be one of the points, so that counting the triangles becomes
a matter of counting the available intersections on the available
rays through A
:
(some rays and points being excluded by the
proviso that one vertex of each triangle must not be of the A
class).
(a) L eg Aj-A
2
. On the other rays through A
v
that is, the
rays to A3 , A4 , A
B
, A
0
, A
7
, we find respectively 5 +4+3+2+1
available intersections with rays from A
2
, or 15 in all. A s there are
8 separate pairs of adjacent A points, the total in this sub-group
is 8X
I
5=i 20.
(b) L egA j - A
3
. On the bundles of rays from each point we
find 4+3+2+1 available intersections, making 10 in all. A s there
are 8 pairs of alternate A points, the total in this sub-group is
8X10=80.
(c) L egA
1
- A
4
. On the A
1
rays to A
3
, A
5
, etc., the available
points total 1 +3+2+1 , or 7. T otal in sub-group, 8X7=56.
(d) L eg Aj-A
5
. T he count of points is 1 -[-24-2+1, or 6.
A s there are only 4 pairs of opposite A points, the total in the sub-
group is 4X6=24.
For the whole class the total is 120+80+56+24=280.
ONE A POINT
In the case of A
1 ;
each triangle of this class must have two
legs lying on rays from A
x
to A
3
, A
4
. . . A
7
. If we examine each
possible pair of rays, and note every secant that cuts both of the
pair within the octagon, we shall take account of all triangles in
this class.
220 SOLUTIONS
(a) Rays from A
1
to A
3
and A
4
. H ere we find 4 triangles.
A dd 4 more on the rays to A
0
, A
7
. Sub-group total, 8X8=64.
(b) Rays to A
3
and A
5
. T here are 3 triangles; add 3 for rays
A7 , Ajj. Sub-group total, 8X6=48.
(c) Rays to A
s
and A
6
. T wo triangles; add the two for
A
7
, A
4
. T otal, 8X4=32.
(d) Rays to A
3
and A
7
. Only one. T otal, 8X1 =8.
(e) Rays to A
4
and A
5
. Six triangles, plus 6 for rays A
0
, A
B
.
T otal, 8X12=96.
(/) Rays to A
4
and A
6
. Four triangles. T otal, 8X4=32.
We have duly taken account of each of the 10 pairs possible
out of the 5 available rays in each bundle. T he total for this whole
class is 280.
No A POINT
In the interior octagon formed by the G points, there are 8
triangles having a common vertex at B. E ach of the 8 D points is
the vertex of a triangle with two C points. T hese 16 triangles ac-
count for all which are wholly interior. None of the E , F, or G
points generate interior triangles.
GRANDTOTAL
T he whole number of triangles is thus 56-|-28o-|-28o-[-i6
=632.
163. T H E C OI N DROP P E R. Not all 14! permutations of
the 14 coins are possible. T he nature of the dropper prevents the
operator from choosing which dime, for instance, he is to take
first. H e is bound to take the dime at the bottom of the cylinder.
H ow to reckon with this limitation is the real problem.
T he solution is easy if we construe the problem as how to fill
14 numbered positions (order of removal) with a batch of 4
nickels, another batch of 3 nickels, a batch of 5 dimes, and 2
quarters. Within each batch we cannot vary the order of the
coins, which has been fixed by the order in which they chance to
be put into the dropper. T herefore we are concerned only with
SOLUTIONS 221
the combination of position-numbers to be assigned to each batch.
For the batch of 4 nickels, we have a choice of 4 position-
1 4!
numbers out of 14, or For the next batch, say the 3 nick-
4! 10!
els, 10 remaining positions are open, so that the number of
, . . 10!
choices is . T he number of places left for the batch of ^
3! 7!
7!
dimes is . T here is no choice in the placement of the last
5 ! 2!
batch, 2 quarters. T he product of the foregoing fractions is
7 =2 , 5 2 2 , 5 2 0
1 4!
5- 4! 3! 2!
It makes no difference in what order we select the batches to
be placed. By cancellation of like terms we always arrive at the
above result.
164. ROT A T I ON P OOL . T here are 16,384 possible orders.
T he number of ways of filling the rack under the conditions must
be the same as the number of ways of removing the 15 balls from
the full rack without at any time leaving a gap. In the removal
process, there is always choice of the two end balls, neither more
nor less, until only one ball remains. T he general formula for n
balls is therefore 2
N_1
. For 15 balls, the answer is 2
14
.
165. T H E NE C KL A C E . No general formula exists which
can avoid the necessity for writing out the possible patterns and
calculating separately how many different designs each represents.
H ere is the complete tabulation.
PATTERN CHOICE PERMUTATION PRODUCT
OF COLORS OF COLORS
i color A A A A A 4 1 4
2 colors A A A A B 6
A A A BB 6
A BA BB 6
2
2
2
12
12
12
222 SOLUTIONS
3 colors A A A BC 4
3
3
3
6
3
12
A A BA C 4
A A BBG 4
A A BC B 4
A BC A B 4
12
24
12
12
4 colors A A BC D 1
A BA C D 1
12
12
12
2
136
E ach letter of the pattern represents a block of four beads of
like color. T he patterns must be understood to be circular: the last
letter is adjacent to the first. C are must be exercised not to dupli-
cate any patterns, e.g., a form A BBA C would be identical with
A A BC B, which is listed. Remember that there is no "right and
left" in the pattern; you have to reckon both ways in comparing
two patterns to see if they are different.
T he first column of numbers gives the ways in which 1, 2, 3,
or 4 colors can be selected out of 4. T he second column gives the
ways in which the selected color-names can be permuted against
the particular pattern. For example, with the unique pattern
A A BC B there is choice of any one of 3 colors from which to take
the two consecutive blocks (A ), then choice of two remaining
colors for the single block (C ).
166. T OU RNA M E NT S C H E DU L E S . T he following schedule
is made on the same principle as the H owell pair schedules for
T his schedule is "cyclical, central." It is cyclical because
the pairings for the first round serve to determine those for the
next seven, by a one-step cyclic progression of the players. It is
central because one player (9) is placed outside the cyclic chain
or rather, within. A conventional representation is to place this
excepted player in the center of a circle, the others being points on
the circle. T he task of constructing the schedule is essentially to
bridge.
SOLUTIONS
223
link the circumference points in pairs by chords, no two of which
are of the same length. For low numbers of players, this geo-
metrical solution is easy and is in fact the way the schedules are
usually constructed.
SCHEDULE FOR NINE PLAYERS
ROUND W B W B W B W B BYE
1
9 VS-
1 2 vs.
4 3
V S
-
6 7 vs. 8
5
2
9
2
3 5 4 7
8 1 6
3 9 3 4
6
5
8 1 2
7
4 9 4 5 7
6 1 2
3
8
5 5 9
6 8
7
2
3 4
1
6 6
9 7
1 8
3 4 5
2
7 7 9
8
0
1
4 5
6
3
8 8
9
1
3
2
5
6
7 4
9
1
5
2
6
3 7 4
8
9
BYE
Geometric C onstruction
Because of the byes, the 9th round is not cyclic with the first
eight. H ere, all the diametrically opposite points are linked. T he
224 SOLUTIONS
matter of giving each player an equal share of the White (W)
and the Black (B) sides of the board is easily adjusted. If the
columns of the schedule are assigned to positions at four tables,
the play at T ables 2, 3, 4 is self-balancing if the side of the White
pieces is fixed at each. It remains only to give the central player
White four times and Black four times, balancing the distribution
against the assignment of colors in the 9th round.
167. P H A L A NXE S . What we have to do is to find the total
of all terms of form G " where r takes all values from 1 up to n.
In other words, we have to find the total of combinations of n
objects taken 1 at a time, 2 at a time, etc., up to (n1) and then
nat a time.
If we write out the terms, they will be of this form:
n! n! n\ n\ !
nT
+
(n1) !
+
2!(n2) !
+
3!(n3) ! " '
+
(n2)! 2!
+
n! n!
(n1)!
I do not expect the reader to evaluate this expression by
analytical methods. But he can easily determine the sum from data
given in this book. If the reader does not at once recognize the
above as the formula for the coefficients in the binomial ex-
pansion, he will surely recognize it from the calculation for a few
low values of n:
f or n =2 , the sum is i - f - 2+1
for n=3, the sum is 1 +3+3+
1
for n 4, the sum is i-f-
4
+6-{-4-fi
E ach series appears along a base line of P ascal's triangle (see
No. 136Figurate Numbers), and it is stated that these base lines
are the coefficients of the binomial expansion.
What is the sum of the series? For n=2, the sum is 4; for
n=3, the sum is 8; for n 4, the sum is 16. E vidently the sum is
always 2". A nd this in fact is the correct answer to the puzzle: the
number of phalanxes under the conditions described in 2".
SOLUTIONS 225
T he reader may wonder why, then, the number of phalanxes
that can be made from 12 soldiers is 6, which is not a power of 2.
T he reason is that 12 does not answer to the description of s in
the problem: its prime factors are not all different. Its factors are
2X2X3?
o r
2
2
X3- T he general rule, whose derivation is beyond
the scope of this book, is that if a, b, c . . . are different primes,
and if s= (a") (b*) (c
r
) . . . then the number of different phalanxes
that can be made out of s is
( P +i ) ( 9 +i ) ( r +i ) . . .
For 12, which equals (2
2
) (3
1
), the formula gives
( 2 + 1 ) ( i + 0 = 6
169. T H A T KI NG OF C L U BS ! Jones is correct in asserting
that the odds favor the second finesse. Smith makes a common
and excusableoversight in the conditions of the problem.
T he fundamental assumption from which all other bridge
probabilities are derived is that any given hand of 13 cards has
the same chance as any other to be dealt. But it does not follow
that any hand of 12 or less cards, held after one or more tricks
have been played, has the same chance to occur as any other. For
the hands are not depleted at random, but by willful selection
which interferes with the operation of blind chance.
C ard probabilities are tabulated as of the original deal. T o
deduce therefrom the probabilities of distribution in a depleted
deal, it is necessary to introduce additional assumptions, based
upon the conditions of the game and the habits of the players.
Or, to use the term of P ierre Boulanger in this connection, we
must reckon with the probability of causes.
When South played the Queen on the first C lub round, he
did indeed exclude C ase 1 as a possibility. If C ase 2 obtains, it is
100% certain that South played the Queen because he had to.
In C ase 3, South has a choice of playing Queen or King on the
first round. If he is a good player he will vary his procedure, so as
to keep his opponents guessing. L et us say that a particular South
plays the Queen half the time and the King half the time. T hen
his play of the Queen excludes the 34 cases where he would have
226 SOLUTIONS
played the King. In other words, the odds are 62 to 34 that he
played the Queen because he had to, rather than because he
chose to. T he conclusion is that the odds for declarer are 62:34
in favor of the second finesse as against swinging the A ce.
T he fact that any given South player may not vary his prac-
tice just 50-50 does not invalidate the form of this argument.
Whatever the figures for South's practice on the play of King
or Queen from King-Queen blank, they must be applied to the
chances of the original deal to determine the probability of cause
the odds that his selection was from choice rather than neces-
sity.
Smith would be right in going up with the A ce if this par-
ticular South player were known to play almost invariably the
Queen rather than the King. Such practice would of course be
very bad, as it would be a dead giveaway in all cases where South
held the King without the Queen.
170. ODDS . T he correct odds on C alypso are 11 to 4. T he
odds of 2 to 1 against A gamemnon mean that the probability of
his winning is taken to be T he probability that Behemoth will
win is rated at %. C alypso's chance is the difference between 1
and the sum %+ %, or %
5
.
Of course in practice the bookmaker must "shorten the odds,"
to give himself a margin of profit.
171. P A RL I A M E NT S OL I T A I RE . T he number of different
104!
combinations of 8 cards out of 104 is . If we were to count
8! 96!
the number of hands holding at least one ace or king, we would
have to make a separate computation for one such card, two such
cards, etc., then add the results. But if we reckon the combina-
tions of 8 cards that do not contain any ace or king, we need deal
only with the remaining 88 cards oi the deck. T he number of
88'
such hands is: . T he probability that the first eight
SOLUTIONS 227
cards dealt will not contain ace or king is given by:
88! 8! 96! 88! 96!
X =
8! 80! 104! 80! 104!
T he probability that at least one ace or king will turn up is
the difference between the above fraction and unity. T he
actual value of the negative probability above is:
114,575,607
459,136,405
Subtracting this fraction from 1 gives the positive probability:
344,560,798
459,136,405
C omparing numerators shows that the chance is about 3 to
1 in favor of success on the first try.
172. E VE RY T H ROW A S T RA I GH T . T he chance of cast
ing at least one 4 when you throw two dice is not % but
1
If
you were to write out all 6X6 combinations of numbers that can
show on two dice, you would have to write 4 (and each other
number) just 12 times, but you could count only 11 different
combinations containing a 4, because one of them is 44.
A s we have seen in No. 171Parl i ament Solitaire, the
probability that an event will occur at least once, where it may
occur more than once, is best computed from the negative proba-
bility that it will not occur at all. What is the chance that you
will not cast a single 4 with six dice rolled simultaneously? On
each die the chance of not casting 4 is %, and the compound
probability on all six together is %X%
o r
(%)- T his frac-
tion of course decreases as the exponent increases. T he positive
probability of throwing at least one 4 therefore increases with
every additional die used. But no matter how many are added, no
matter how large the exponent of the negative probability is
made, the negative fraction never reaches zero, and the positive
probability of casting a 4 never reaches 1 (certainty).
228 SOLUTIONS
173. T RE I Z E . I will use the symbol H to mean "the num-
ber of permutations which give no hit, with n cards." In every
case, the total of all permutations is n\
For n2, n!=2. One permutation (i , 2) hits twice and the
other (2, 1) hits not at all. T hen H
2
=I.
For n3, n!=6. Suppose that we write out all permutations:
Ordinal number 1 2 3
Permutations 1 2 3
1 3 2
*2 3 I
2 I 3
*3 I 2
3 2 I
T he permutations starred (*) are the only ones with no hit.
T hen H
3
=2.
For n= 4, n! = 24. Now instead of writing out all the permu-
tations, let us consider how we could write out only the H permu-
tations.
T o begin with, we must exclude card No. 1 from the 1st posi-
tion. T hat leaves 3, or n1 cards which we can put there:
Ordinal number 1 2 3 4
H Permutations 2
3
4
E acn row must represent several permutations, since there
remain three cards to be permuted among three remaining posi-
tions. L et us consider each row as representing a sub-group; the
number of permutations in all sub-groups must be the same; after
we have determined the number per sub-group we can multiply
this number by 3 (which is n1) to get the total of H permu-
tations.
SOLUTIONS 229
H ow shall we permute the last three numbers in a sub-group?
Well, two of the numbers are the same as ordinal positions con-
cerned, while one is different, e.g., on the first row we have to
permute i, 3, 4 among the positions 2, 3, 4. If we treat 1 and 2
for the moment as identical, then the H permutations are II
3
,
which we have already determined to be 2. But 1 and 2 actually
being different, we can place card No. 1 in the second position
and get as many additional permutations as we can out of 3, 4 in
positions 3, 4. T his number is simply H
2
, which we have already
determined to be 1.
In each sub-group the number of permutations is therefore:
i+
H
n2
A s there are n1 sub-groups, we have finally:
H
n
=( n - i ) ( H
B
_
1
+H ^)
T his is the formula whereby we can go on compounding the
table of H permutations as n increases. T o solve the given prob-
lem we will build the table up to n6:
n
H
n!
H/'
2 1 2
-5
3
2 6
33"
4 9
24
375
5 44
120 .366-
6 265 720 .368
T he chances of winning the T reize patience with six cards
are thus about 10 in 27. If the reader cares to carry the table a
little further, he will discover that as n increases the ratio in the
rightmost column remains stable at .367+.
176. T O L E A VE T H E L A S T . A player can be forced to take
the last counter only when there is but one left. C onsequently the
formula is
I - J - ( < 2 + M ) n
230 SOLUTIONS
177. T O WI N T H E ODD. T he formula can be found by
mathematical induction. Fix limits for the draw, as i to 3, then
plumb every possibility when the common pile is reduced to 4, 5,
etc. It will be found that the outcome depends in every case on
whether the player has at that juncture acquired an odd or an
even number of stones.
A gainst a player who has acquired an odd number:
w2n( a-\-m) +a[-(- m]
For each set of values given to a, m, n, the formula gives two
values for w, the square brackets indicating that in one case the
term is to be added and in the other it is to be omitted.
A gainst a player who owns an even number:
w=2n( a-\-m) [-f-2a+m]
In the case of a 1 and m=3, these formulas reduce to
A gainst odd, the series i-)-8n ( 1 , 9, 1 7 . ) and 4+8 n
(4,12,20. . . ) A gainst even, the series 8n (8,16,24. . . ) and 5+8ra
(5>'3>2i. . . )
178. T H RE E - FI VE - S E VE N. T he powers of 2 (2
0
, a
1
, 2
2
, etc.)
are the numbers 1, 2, 4, etc. E very integer can be expressed as the
sum of a group of these numbers in a unique way. In other words,
if each number of the series is used once only, there is only one
way of choosing the right numbers to sum to a given integer.
For example, the integer 59 can be broken down into
3 2-j-16-|8-f-2-} I. T he way to discover the unique series for each
integer is to subtract from it and from each remainder thereafter
the largest possible power of 2.
In the game of three piles, express the number of counters in
each pile in powers of 2. T he initial array, for example, is
1 1 1
2 4 2
4_
3 5 7
T he law involved is that an array is a w (a winner for the
player who presents it to his opponent) if each power of 2 thcrei.i
represented at all appears just twice.
SOLUTIONS 231
In the initial array, there are two 4's and two 2's, but three
I'S. T he first player wins by taking one counter from any pile.
It is evident that a player who has to draw from two equal
piles is lost. In drawing, therefore, each player must be careful not
to exhaust any pile unless the other two are equal.
T his solution is quite general; it can be applied to any num-
ber of piles and any number of counters. T he formula is then to
be read that an array is w when each power of 2 represented at all
appears an even number of times.
Notice that it is not always possible to reduce an array to w
merely by subtracting some of its component numbers. Occasion
will arise when it is necessary to subtract a number not visible in
the array, so as to change one series of powers to another. For
example, with three piles of a larger number of counters, suppose
this array is reached:
2 1 2
4
2
4
4
8
6
7
14
T o take 6 counters from either of the larger piles will equalize
the 2's and 4's, but will leave the 1 and 8 unmatched. T he solution
is to draw 13 from the largest pile, leaving the array
2 1 1
4 2
4
6 7 1
If the separate totals of the piles are expressed in the binary
system (where there are only two digits instead of ten), the deter-
mination of / is a simple matter of subtraction among these num-
bers. T he solution as given actually depends on this fact, since it
expresses the integers of the binary scale in the decimal system to
which we are accustomed.
179. T H E T H I RT Y-ONE GA M E . A s pointed out by Dude-
ney, the simplest course for the first player is to turn 5. If the
second turns 5 to get into the w series, the first player turns 2. If
232 SOLUTIONS
the second then persists in the series, first continues with 2's and
makes 26 with the 5's exhausted. But if the second player at any
time goes out of the w series, first can seize it and no needed num-
ber can be exhausted.
T he first player can also win, by more complicated play, after
commencing with A or 2.
180. T H I RT Y-ONE WI T H DI C E . H ow to attack this prob
lem is puzzling until one hits on the idea of digital roots.
T he digital root of the goal 31 is 4. T he w numbers (totals
that win for the player who reaches them) are primarily those
whose digital roots are 4, namely 4, 13, 22, 31.
T he members of this primary series differ by 9. If the player
confronted by a w adds 1 or 2, his opponent at the next turn can-
not reach the next w number. H ow is the win to be enforced?
Well, the player who cannot reach a w should do the next
best thinghe should prevent his opponent from getting into the
series. T o effect this he must reach a root which differs from root
4 by t he number he turns up or its complement. T hes e t wo num-
bers, remember, are not available to the next player.
When 1 or 2 is added to root 4, the root becomes 5 or 6. T hen
the addition of 4 or 3 respectively to reach 9 holds the win, since
in either case 4 is unavailable.
E vidently we must add to the primary w series: any total
with root 9 reached by turning 3 or 4.
Suppose that to root 9 a player adds 2, making root 2. T hen
the next player cannot reach root 4, because the 2 is unavailable;
nor root 9, because there is no 7 on the die. What is he to do? H e
must prevent the other from reaching any w, 4 or 9. T he only play
is to turn 3, making root 5, and then the other cannot reach 9
because the complementary 4 is buried.
By similar reasoning, it can be shown that, besides the pri-
mary w series of root 4, there exist also conditional w numbers as
follows :
Root 1, 5, or 9 reached by turning 3 or 4;
Root 8 reached by turning 2 or 5.
SOLUTIONS 233
A ll the other roots, 2, 3, 6, 7, are sure losers. A number will
always be available to reach a root of the w series: for root 2,
either 2 or 3 will serve; for 3, 1 or 5; for 6, 2 or 3; for 7, 3 or 6.
T he only roll that assures the first player a win is of course 4.
182. S A M L OYD'S DA I S Y P U Z Z L E . T he second player
wins. T he method is to break the circle of petals into two symmet-
rical semicircles. Whether the first player takes 1 petal or 2, the
second player must answer by taking 2 or 1 respectively from the
opposite end of the diameter, breaking the array into two blocks
of 5 petals each. T he second player then continues to restore the
symmetry each time it is broken by the first, and consequently
must draw the last petal.
183. DU DE NE Y'S C I GA R P U Z Z L E . A fter all we have said
of the principle of symmetry, the reader will no doubt perceive the
idea of this puzzle. Whoever has to play from a perfectly symmet-
rical arrangement of cigars on the table must lose. If he can find
room to add a cigar, then so can his opponent. T he latter need
only add his cigar on that point (and in same relative orientation)
which is symmetrical to the position of the first cigar, with respect
to the center point of the table. (T wo points A and B are sym-
metrical with respect to a point C , if C is the midpoint of the
straight line joining A and B.)
It would seem then that the second player must win, since he
has only to keep pairing his plays with those of the first player.
But there is a circumstance easily overlooked. T he cigar as de-
scribed can be made to stand upright on its flat end. T he first
player wins by placing his first cigar upright on the only unique
point of the tablethe center of symmetry itself!
184. T H E C A RP A T H I A N S P I DE R. A t the outset the fly
stands on the same line as the spider, 8 points away (an even
number). T hat means that whenever the two insects stand on
diagonally-opposite points of the same quadrilateral, it will be the
spider's turn to move. If the whole web were composed of quad-
234 SOLUTIONS
rilaterals, the fly could never be caughthe could merely circulate
around the four sides of a quadrilateral.
But the web includes one triangle. T he spider need only go at
once to this triangle and circumnavigate it to change "the move."
T hereafter it will be the fly's turn to move when the number of
points intervening between the two insects is even. A little experi-
ment will show that the fly is then easily run down.
T he puzzle is seen to be another illustration of the principle
of symmetry. T he opposition play has its analogue in chess, where
the maneuver of "triangulation" is well-known as a necessity to
change the move in some end-situations.
T he "spider and fly" puzzle is here presented in its simplest
form. T he principle is capable of great elaboration through the
addition of other ideas: pentagonal and heptagonal figures as well
as triangles; more than one circuit of an odd number of points in
the web; opportunities for both players to change the move but a
limitation on the opportunity of one to reach an odd circuit with-
out being cornered there; arbitrary rules such as limitation of the
number of times some specified intersections may be visited. T he
whole subject presents a wealth of opportunities to those who like
to make puzzles-and this is the type of puzzle that is as much fun
to make as to solve!
189. S A L VO. A s with any question of probability, the an-
swer can be given only in terms of some basic assumption. H ere
we assume that the player, having resolved to fire his first salvo in
the given formation, is just as likely to place it in one orientation
as in any other. We can sum up the total number of times each
square is hit if every possible orientation is used once, and this sum
is the measure of the probability that the particular square will be
hit.
A n easy method of operation is as follows:
From a piece of paper cut out 7 squares in the chosen forma-
tion. U sing this as a mask or "grille," lay it on a 10X10 square,
move it vertically and horizontally (without rotation) to every
possible position within the square, and in each position mark a
SOLUTIONS 235
tick through each of the 7 holes onto the paper below. T he result
of this step is shown in Fig. 1.
2 3 4 5 A 4 5A 3 1
3 4 7 II II II II 7 4 3,
4 7 12171818n 127 4
5 II n 22 252522 17II 5
4 II IS 25 303025 18II 4
4 II 1625 3D302518II 4
5 II 1722 25 2522n II 5
4 7 IZ
17 1812n
127 4
34 7 II
II II II 7 43
2 3 4 54 454 3 Z
FIG. 1 FIG. a
T he formation is not symmetrical. We have to take account
of three other possible orientations corresponding to rotations of
the grille through 90, 180, and 270 degrees. T o do so, merely make
three copies of the diagram, then align all four side by side in the
four relative rotations. In a new square, sum the ticks in corre-
sponding cells of all four squares.
Finally, we have to take account of four more orientations
which are equivalent to turning the grille over. M ake a copy of
the summation square in mirror reversal, then sum it with the
copy in yet another square. T he final result, Fig. 2, shows how
many times each cell is hit if every possible orientation is used
once.
From Fig. 2 we see that the safest position for the battleship
is in a corner, extending vertically or horizontally. T he worst posi-
tion is vertically or horizontally across the middle of the arena.
T he relative chances of being hit sum to 18 and 128 respectively,
or about 7 to 1 in favor of the corner.
A ppendix
H OW T O E XT RA C T S QU A RE ROOT . For illustration let us
extract the square root of 4, 375, 690, 201 . T he working sheet is
shown below, and the operations are described step by step.
6 6 1 4 9
V 4 3 75 69 02 01
36
7 75
120
7 56
1 26
19 69
1 320
13 21 1 321
6 48 02 1 3220
5 28 96
1 3224
1 19 06 01 1 32280
1 1 9 06 01 1 32289
1. Di vi de the number i nto peri ods of t wo digits each, f rom right to
l ef t .
2. Bel ow the first peri od ( l ef t most ) wri te the largest square that is
equal to it or less ( 36) , and wri t e its square root (6) as the first digit
of the quoti ent.
236
APPENDIX 237
3. Subtract the square from the first period, then bring down the next
period, to make the first remainder (775).
4. In a separate memorandum column to the right, double the quo-
tient (so far as it goes) and add one zero (120).
5. Now estimate the digit that must be added to this memo number
so that, when the total is multiplied by the same digit, the product will
be the largest possible equal to or less than the first remainder. (120
goes into 775, at most, 6 times. When 6 is added to 120, the sum 126
multiplied by 6 does not exceed 775, so that 6 is the correct second
digit.) Write this digit, when correctly estimated, as the second digit of
the quotient.
6. Write the new product just found (756) under the first remainder,
subtract, and bring down the next period to make the second remainder.
7. Repeat steps 4, 5, 6 until all the periods are exhausted.
T ake note that the memo number written at the right must always be
double the entire quotient, so far as it is determined, plus one zero for
each period brought down into the remainder. If a remainder should be
smaller than the memo number, then an additional period must be
brought down and another zero affixed to the memo number.
It is worthwhile to understand the why and wherefore of
these operations.
T he square root sought is construed to be of binomial form,
a-j-b. Its square is consequently of form a
2
~\-2ab-\-b
2
. T he quantity
a is the integral part of the root, while b is any remainder. T he
part a is found by successive approximations. T hus, when we
wrote 6 as the first digit of the root in the above example we were
really estimating that the root lies between 60,000 and 70,000.
Our first approximation 60,000 being a, we subtracted a
2
from the
original number, leaving 775,690,201. T his difference is in form
2ab-\-b
2
, which we treat as b(2a-\-b). We wrote 2a in the memo
column: this number is really 120,000, but we leave off the last
three zeros because we have left the last six figures off the first
remainder. We next have to estimate an integer b which will make
b(za-{-b) the largest possible partial product. H aving done so, we
write this digit into the root, which now is approximated more
closely to 66,000. T his new value becomes a new a, and we pro-
ceed to estimate a new b that will give the third digit of the root,
and so on.
238 APPENDIX
H OW T O E XT RA C T C U BE ROOT . Before studying this op-
eration, read the preceding algebraic explanation of the extrac-
tion of square root.
T he operations in finding a cube root are based on similar
principles of algebra. T he root is construed as a polynomial, a-\-b,
an integral part plus a remainder. T he cube is consequently of
form a
3
-{-^a
2
b-\-^ab
2
-\-b
3
. T he Dart a is found by successive ap-
proximations. T o illustrate:
^ 7 444 997
64
6 444 4800
4 921 120
1 523 997 i_ 504300
' 523 997 492i 3690
9
57999
1. Divide the number into periods of three digits each, from right to
left.
2. Below the first period (leftmost) write the largest cube that is
equal to it or less (64), and write its cube root (4) as the first digit of
the quotient.
3. Subtract the cube from the first period, then bring down the next
period, to make the first remainder (6444).
4. In a separate memorandum column to the right, enter three times
the square of the quotient (so far as it goes) and add two zeros (4800,
which is 3 X4 - X 100).
T he quotient so far as it goes is a; the remainder after a
3
is subtracted
is of form ^a
2
b + ^ab
2
+ b
3
, which is treated as We
have written 3<z
2
(48oo); now we must find a digit b so that to 4800 we
can add 120b and b~, then multiply the sum by b to produce the largest
possible partial product.
Sincc 4800 will go into 6444 only once, we see that the b is here 1, so
we write 1 as the second digit of the root, add 120 and 1 to 4800, and
subtract the sum from 6444.
5. Repeat steps 4 and 5 until all the periods are exhausted.
T ake note that in writing 3a
2
in the memo column it is necessary to
add two zeros for each period brought down into the remainder, and in
writing 3a (to be made into 3ab when b is estimated) it is necessary to
add one zero for each period brought down.
APPENDIX 239
If you become confused about the zeros, just picture each
line of figures in the computation extended by zeros to the right as
far as the rightmost column. T he first remainder 6444 is thus seen
to be really 6,444,000, and the first digit of the root 4 really stands
for 400 (one digit for each period). Now if a is 400, then 3a
2
is
480,000; this is going to be multiplied by a digit b which really is
106, since it will stand in the tens place of the root. T he product
will then be of the order 4,800,000. Since for convenience we omit
the last three zeros from 6,444,000, we must likewise omit just
three terminal zeros from 4,800,000 if we are to align this number
at the right below the first remainder.
Similarly, 3a is actually 1200, and since this is to be multi-
plied by b
2
(which is actually ioofr
2
) the product will be of order
120,000. Finally the digit b is going to be cubed, so it will be of
order 1,0006.
T A BL E OF S QU A RE ROOT S
21. 41421
31.73205
52.2360
62.4494
72.6457
82.8284
103.1622
u3. 31 66
T A BL E OF P OWE RS OF 2
I S t 2
2nd4
3rd8
4 t h16
5 th32
6th64
7th128
8th256
9th512
nt h2 0 4 8
12th4096
13th8192
14th16,384
15th32,768
16t h65, 536
17th131, 072
18th262,144
19th524,288
10th1024 20th-1,048,576
240 APPENDIX
T A BL E OF S QU A RE NU M BE RS
I I 411681 816561
24 421764. 826724
3 9
431849 836889
41 6
44' 936 847056
525
452025 857225
6 - 3 6 462116
867396
749
472209
877569
864 482304
88- 7744
981 492401
897921
10100 502500 908100
1 r121 512601 918281
12144 522704 928464
13169 532809 938649
14196 542916 94- 8836
15225
553025
959025
16256 563136 969216
17289 573249
979409
18324 583364 989604
19361
593481
999801
20400 603600 10010,000
21441 613721 10110,201
22484 623844 10210,404
23529
633969
10310,609
24576 644096
10410,816
25625
654
2 2
5
10511,025
26676
664356
10611, 236
27729 674489
10711, 449
28784 684624 10811, 664
29841
694761 10911, 881
30900 704900
110 12, 100
31961
715041
11112, 321
321024
725184
11212, 544
331089
735329
11312, 769
341156
745476
11412, 996
351225 755625
1 1 5
!
3-
2 2
5
361296 765776
11613, 456
371369 775925
1
' 7
1
3 .
6 8
9
381444 786084
11813, 924
39
!
52i
796241 11914, 161
40r6oo 806400
12014,400
2
3
5
7
II
i3
i7
i9
23
29
3i
37
4i
43
47
53
59
6i
67
7>
73
79
83
89
97
101
103
107
109
113
127
131
137
139
149
i5
157
163
167
173
179
181
APPENDIX 248
T A BL E OF P RI M E NU M BE RS
(between 1 and 1,000)
191 433
193 439
197 443
99 449
457
211 461
223
4
6
3
227 467
229
479
233 487
239 49i
241
499
25 >
257 503
263
59
269
521
271
523
277
54
1
281
547
283
557
293 563
569
307 57i
3>i 577
3i3 587
3'7 593
33i 599
337
347
601
349
607
353
613
359
617
367
619
373
631
379
641
383 643
389
647
397 653
659
401 661
409 673
4i9
677
421 683
43i
691
242 APPENDIX
T A BL E OF T RI A NGU L A R NU M BE RS
II 21231 41861
23
22253 42903
3 - 6
23276
43946
4io 24300
44990
515 25325
451035
621
2&351
461081
728 27378 471128
8-36 28406 48I 176
945 29435
491225
I0
55 30465
501275
1166
31496
51 1326
1278 32528
521378
1391 3356I
53'43'
14105
34595
541485
15120
35630 55'540
16136 36666 561596
r
7153 37703 57I653
18171 38741 581711
1919
39780 591770
20210 40820 601830
Glossary
additiveA ny one of several numbers that are added together,
azimuthT he angle between a line of direction and north, meas-
ured clockwise from north.
binomial theoremT he formula for determining the coefficients
of the expansion of (a-j-b)" for any value of n.
C aliban puzzle-One in which the solver is asked to infer a fact
from a set of given facts, so-called from the pseudonym of
an E nglish inventor of such puzzles,
cardinal numberOne expressing magnitude or quantity, as: 2,
561, a million.
coefficient-A number (or symbol) prefixed to another number
and by which the latter is multiplied, as: 2 in 2x=y; A , B,
C in A x
2
+Bx +C =0.
combinationA ny particular sub-group selected from a larger
group, as: the combination BD out of A BODE .
C ombinatorial analysisT he study of combinations, especially as
to kinds and classes as well as to number.
C omplex numberImaginary number; an expression involving
both real and imaginary numbers,
composite numberOne that has some factors besides itself and
unity, as: 6 ( =3X2) , 3
8
5 (=5X7X10 - Antonym, prime
number.
243
244 GLOSSARY
constantA number of fixed value, as: 3, \/5, >^8, 13!, tr.
Antonym, variable,
cube rootA ny one of the three equal factors whose product
is the given number as: 2 is the cube root of 8.
cubic equationOne in which some variable has the exponent
3, but no higher exponent appears, as 15.
digitA ny of the nine symbols I 2 3 4 5 6 7 8 9 , possibly also
including the symbol o, zero.
(
digital rootT he sum-of the digits of an integer, continued to a
single digit, as: the digital root of 786 is 3 (7+8+6=21
and 2+1 =3) .
Diophantine equationOne of indeterminate form but with
limited soludons by reason of the fact that the values of
the variables must be integral,
discreteDiscontinuous.
empiricBased upon experience (as trial and error) rather than
on theoredc formula,
ex pansionT he carrying out of an indicated operation, as
multiplication or the raising to a power,
ex plicit equationOne that states the value of a variable in
terms of another or others, as xy, 3y=5- Antonym, im-
plicit equation.
ex ponentA n operator written as a superscript, indicating the
number of times the number to which it is attached is to be
used as a factor, as: 36=243 (3X3X3X3X3)-
factorA number by which another can be evenly divided; any
one of the terms in the expression of a product,
factorial (of an integer)T he product of all integers from
1X2X3 . . . up to and including the given integer, as:
factorial 4=24 (the product 1X2X3X4) j the term "fac-
torial" is expressed by the symbol ! as in n! or by |
as in n J.
factorizationT he determination of the factors of a number.
GLOSSARY
245
G.C .D.M eans "greatest common divisor"; the largest number
that is a factor of each of several given numbers.
higher arithmeticT he study of certain properties of integers,
now usually called theory of numbers,
hypotenuseT he longest side of a right triangle.
imaginary numberT he square root of i , or any multiple
thereof; the symbol i is used to denote >/I. Antonym,
real number.
implicit equationOne in which the value of the variables is
given indirectly by a relation among them, as: x-\-xy-\-$y
= 11. Antonym, explicit equation,
integerA whole number as: 4, 15, 9062.
irrational numberOne that cannot be expressed as the ratio
between two integers, as the value of v.
L .C .M .M eans "least common multiple"; the smallest number
that contains as a factor each of several given numbers,
legE ither side of a right triangle other than the hypotenuse,
linear equationOne in which no variable has an exponent
greater than unity.
multipleC onverse of factor; A is a multiple of B if B is a factor
of A .
negative numberOne to which the minus sign is prefixed.
Antonym, positive number,
numberT he general term for the symbols of mathematics, as:
the digits, the integers, rational and real numbers, im-
aginary numbers, symbols denoting constants or variables,
numbers, theory ofT he study of classes and properties of
integers.
246
GLOSSARY
odds-A way of stating probability, the ratio of the probability
that an event will occur to the probability that it will not.
operatorA number indicating an operation to be performed on
other numbers to which it is attached, as: an exponent,
the symbol y/.
ordinal numberOne that expresses position within a series, as:
second, 561st, millionth. Antonym, cardinal number.
partial productIn long multiplication, any one of the results of
the multiplicand times a digit of the multiplier,
pencilA n aggregate of lines all intersecting at one point,
perfect numberOne that is the sum of all its divisors, as
6 ( = 1 + 2 + 3 ) .
permutationA ny one of the arrangements or order in which
a group of objects may be placed, as: the permutations
A BC DE , BDGA E , E C DA B.
polynomialA n algebraic expression of the sums and/or differ-
ences among two or more terms, as a-\-b.
positive numberOne to which the plus sign is prefixed; when
no sign is attached, a number is understood to be positive.
Antonym, negative number,
powerC onverse of root; A is a power of B if B is a root of A .
prime numberOne that has no divisors other than itself and
unity, as: 3, 7, 19, 31. Antonym, composite number,
probabilityT he likelihood of occurrence of an event in a cer-
tain way, especially when capable of expression in mathe-
matical terms.
proportionA n equality between two ratios, as: 6: 4=3: 2;
a:b=c'd.
P ythagorean theoremT he proposition that a
2
-\-b
2
c
2
where
a and b are the legs of a right triangle and c is the
hypotenuse.
quadratic equationOne in which some variable has the ex-
ponent 2, but no higher exponent appears, as: 3-*
2
+4
=2y~5-
GLOSSARY 247
quartic equationOne in which some variable has the exponent
4, but no higher exponent appears, as: y=: v
4
| 5X
2
+9*+
I
3-
radical signT he symbol expressing the extraction of a root;
the degree of the root is written as a superscript in the s/,
except that 2 is omitted in expressing square root,
radix T he base of a system of enumeration, as 10, the base of
modern numerals.
ratioT he quotient of two numbers, as: 5:4, read "five to four"
5
and denoting.
4
rational numberOne that can be expressed as the ratio of two
integers.
real numberT he class of all rational and irrational numbers
together. Antonym, imaginary number,
rootA ny one of the factors, all equal, whose product is the given
number; the degree of the root is the number of such
factors, as: 3 is the 4th root of 81.
square numberOne that is the product of two equal factors, as:
25 (5X5);38I (19X19)-
square rootOne of the two equal factors of which the given
number is the product, as: 3 is the square root of 9.
symbols
-f- plus sign, read "plus"
minus sign, read "minus"
X or multiplication sign, read "times"
-5- division sign, read "divided by"
< read "is less than," as 3 <4, 3 is less than 4
> read "is greater than," as p>q, p is greater than q
5
: read "to," indicates ratio, as 5:4, denoting
4
= equality sign, read "equals"
read "does not equal"
read "plus or minus"
248 GLOSSARY
7
s
exponent, written as superscript, 2 is read
"squared," 3 is read "cubed," 4 and higher is
read "to the 4th power," "to the nth power";
indicates the number of times the quantity to
which it is affixed is to be used as a factor, as
7
2
=7X7=49
radical sign, indicates extraction of a root, whose
degree is indicated by the number written as a
superscript in the y/; without superscript the
radical sign is understood to mean square root;
3rd degree is read "cube root," 4th and higher
degrees are read "4th root," "nth root"
p
3
subscript, read "sub 3," is used to distinguish
separate members of a group when the same
symbol (as p) is used to denote all members
! or | factorial sign, read "factorial n" or "n factorial,"
denotes the product 1X2X3 . . . n, as 5! or 5 |
= 1 X2X3X4X5=
1 2 0
k pi, read "pi," the constant 3.14159 . . . , the ratio of
the circumference of a circle to the diameter
P read "the number of permutations o f
C read "the number of combinations of '
triangular numberOne that is the sum of all consecutive
integers from unity up to a given integer, as: the triangle
of 5=1 +2+3+4+5=1 5.
variableA symbol used to represent a quantity of unknown or
variable magnitude, as x, p
n
.
vertexT he point of intersection of two adjacent sides of a plane
figure, as a corner of a square.