Contestproblembook3 (1966-1972)

THE CONTEST
PROBLEM BOOK I11

Annual High School Contests
19661 972
of
The Mathematical Association of America
Society of Actuaries
Mu Alpha Theta
National Council of Teachers of Mathematics
Casualty Actuarial Society
Illustrations by George H. Buehler
©Copyright 1973 by The Mathematical Association of America (Inc.)

All rights reserved under International
and Pan-American Copyright Conventions.
Published in Washington, D.C. by
The Mathematical Association of America
Library of Congress Catalog Card Number: 66-15479
Print ISBN 978-0-88385-625-3
Electronic ISBN 978-0-88385-940-7
Manufactured in the United States of America
THE CONTEST
PROBLEM BOOK I11
Annual High School Contests
1966- 1972
compiled and with solutions by
Charles T. Salkind
Polytechnic Itzstitute of Brooklyn
and
James M. Earl
University of Nebraska at Omaha
25
THE MATHEMATICAL ASSOCIATION
OF AMERICA
Contents
Preface 1
Editors’ Preface 3
Suggestions foi Using This Book 5
I Problems 7
1966 Examination 7
1967 Examination 15
1968 Examination 22
1969 Examination 28
1970 Examination 35
1971 Examination 41
1972 Examination 47
II AnswerKeys 55
III Solutions 57
1966 Examination 57
1967 Examination 74
1968 Examination 95
1969 Examination 111
XV Chssikation of Problems 183

NEW MATHEMATICAL LIBRARY
I. Numbers: Rational and Irrational by Ivan Niven
2. What is Calculus About? by W W Sawyer
3. An Introduction to Inequalities by E. E Beckenbach and R. Bellman
4. Geometric Inequalities by N. D. Kazarinoff
5. The Contest Problem Book I Annual High School Mathematics Examinations
1950-1960. Compiled and with solutions by Charles 7: Salkind
6. The Lore of Large Numbers by P. J. Davis
7. Uses of Infinity by Leo Zippin
8. Geometric Transformations 1 by 1. M. Yagloni, rranslated by A. Shields
9. Continued Fractions b.y Curl D. Olds
10. Replaced by NML-34
Hungarian Problem Books I and 11, Based on the Eiitvt)s Competitions
1894-1905 and 1906-1928, translared by E. Rapaporr
13. Episodes from the Early History of Mathematics by A. Aaboe
14. Groups and Their Graphs 6.y E. Grossman and N! Magnits
15. The Mathematics of Choice b.v Ivan Niveri
16. From Pythagoras to Einstein by K. 0. Friedrichs
17. l h c Contest Problem Book II Annual High School Mathcmatics Examinations
~ ‘ ) ~ 1 ~Compilcd . with solutions &y Charles 7: Sulkifid
- ~ ~ 6 ~and
18. First Conccpts of Topology by W G.Chinn and N. E. Steenrod
19. Gcomctry Rcvisitcd by H. S. M. Coxeter and S. L. Greitzer
20. Invitation to Number Thcory by Oysfein Ore
21. Gcoinetric Transformations II by 1. M. Yuglour. transluted by A. Shields
22. Elcmcntary Cryptanalysis-A Mathcmatical Approach l ~ yA. Sinkov
23. lngcnuity in Mathematics by Ross Ilomberger
24. Gconietric Transformations 111 by I. M. Yaglom, rranslatcd by A. Shenirzer
25. Thc Contest Problem Book 111 Annual High School Mathcmatics Examinations
1966-1972. Compiled and with solutions by C. 7: Sulkbid arid J. M. Earl
26. Mathematical Methods in Science by George Pdlya
27. International Mathematical Olympiads-1959-1977. Compiled and with
solutions by S. I>. Greitzer
28. The Mathematics of Games and Gambling by Edward W Packel
29. The Contest Problem Book IV Annual High School Mathematics Examinations
1973-1982. Compiled and with solutions by R. A. Arfino, A. M. Gaglione.
and N. Shell
30. The Role of Mathematics in Science by M. M. SchiSfer and I, Bowden
31. International Mathematical Olympiads 1978- 1985 and forty supplerncntary
problems. Compiled and with solutions by Murray S. Klamkin
32. Riddles of the Sphinx by Martin Gardner
33. U.S.A. Mathematical Olympiads 1972-1986. Compiled and with solutions
by Murray S. Klamkin
34. Graphs and Thcir Uses by Oystein Ore. Revised and updated by Robin J.
Wilson
35. Exploring Mathematics with Your Computer by Arthur Engel
36. Game Theory and Strategy by Philip D. Strafin, JK
37. Episodes in Nineteenth and Twenthieth Century Euclidean Geometry by
Ross Honsberger
38. The Contest Problem Book V American High School Mathematics
Examinations and American Invitational Mathematics Examinations 1983-1988.
Compiled and augmented by George Berzsenyi and Stephen B. Maurer
39. Over and Over Again by Gengzhe Chang and Thomas W Sederberg
40. The Contest Problem Book VI American High School Mathematics
Examinations 1989-1994. Compiled and augmented by Leo J. Schneider
41. The Geometry of Numbers by C. D. Olds, Anneli Lax, and Giuliana I?
Davidoff
42. Hungarian Problem Book 111 Based on the EtitvGs Competitions 1929-1943
translated by Andy Liu
Other titles in preparation.
MAA Service Center

P. 0. Box 91112
Washington, DC 20090- 1112
800-331-1622 fax: 301-206-9789
Preface
Problem solving is a t the heart of learning mathematics; a student’s
ability to perceive, master and work with mathematical fundamentals
is greatly enhanced by encouraging him to solve carefully designed
problems. A good problem, like an acorn, contains the potential for grand
development. The Committee on High School Contests, in this spirit,
seeks to extend and supplement regular school work through the Annual
High School Mathematics Examination. First organized in 1950 and
restricted to Metropolitan New York, these examinations were sponsored
nationally in 1957 by the Mathematical Association of America and the
Society of Actuaries, and later cosponsored by Mu Alpha Theta (1965),
the National Council of Teachers of Mathematics (1967), and the
Casualty Actuarial Society (1971).
An important difference between these and some similarly motivated
European competitions is that our Annual Examination aims to dis-
criminate on several leoels, and is not exclusively directed to high-ability
students.
All students are welcome to participate as individuals or in teams of
three from the same school. The top scorer receives the Association’s
Mathematics Pin award; there are also lesser awards including a Cer-
tificate of Merit for teams scoring in the upper decile regionally.
The number of contestants has grown from approximately 150,ooO
in 1960 to more than 350,000 in 1972 within the ten Canadian and U.S.
regions; in addition there were thousands of participants abroad. One
hundred of the top students took part, on May 9,1972, in a very successful
first USA Mathematical Olympiad, a 5 question, 3 hour subjective test.
The questions on each examination are grouped according to difficulty
and complexity as estimated by the members of the Committee. Questions
in the first two parts are meant to deal with basic concepts, those in the
later parts are meant to test the application of skills to various new
situati0ns.t On the 1966 and 1967 examinations, Parts 1,2, and3 consisted
t For a subset of the contestants, the number of correct responses has been tabulated,
and the categorization of questions according to difficulty seems justified.
1
2 THE M A A PROBLEM BOOK I11
of 20, 10, and 10 questions weighted 3, 4, and 5. On the 1968-1972 ex-

aminations, the questions are divided into four parts of 10, 10, 10, and 5
questions, respectively, with weights 3, 4, 5, and 6.t
The solutions presented here are by no means the only ones possible,
nor are they necessarily superior to other alternatives. Since no mathe-
matics beyond intermediate algebra is required, an elementary procedure
is always given, even where “high-powered” alternatives are added.
The MAA Committee invites your comments.
Charles T. Salkind
James M . Earl
-
t Examinations are scored by the formula R tW, where R and W denote
weighted counts of correct and incorrect responses, respectively.
Editors’ Preface
The editors of the New Mathematical Library, in wishing to encourage
significant problem-solving at the high school level, have published the
following problem collections so far: NML 5 and 17 containing all annual
contest problems proposed by the Mathematical Association of America
through 1965; and NML 11 and 12 containing translations of all Eotvos
Competition problems through 1928 and their solutions. The present
volume is a sequel to NML 17 published at the request of the many
readers who enjoyed the previous MAA problem books.
The MAA contests now contain 35 problems based entirely on the
standard high school curriculum. To expedite grading of the approxi-
mately 400,000papers written, each question is worded so that exactly
one of five choices offered serves as a correct answer.
Each Eotvos contest, on the other hand, contains only three problems,
based on the Hungarian high school curriculum, and often requiring
ingenuity and rather deep investigations for their solution.
The MAA is concerned primarily with mathematics on the under-
graduate level. I t is one of three major mathematical organizations in
America (the other two being the American Mathematical Society,
chiefly concerned with mathematical research, and the National Council
of Teachers of Mathematics, concerned with the content and pedagogy
of elementary and secondary mathematics). The MAA also conducts the
annual Putnam Competition for undergraduate students. Its journal,
The American Mathematical Monthly, is famous for its elementary and
advanced problem sections.
The editors of the New Mathematical Library are glad to cooperate
with the MAA in publishing this collection. They wish to acknowledge,
in particular, the essential contributions of the two men who compiled
and wrote solutions for the problems in the present collection: Prof.
Charles T. Salkind, responsible for the contests up to the year of his
3
4 T H E M A A P R O B L E M B O O K I11
death, and Professor James M. Earl, who succeeded him in 1968. During
his illness Professor Earl not only supervised the editing of the present
volume, but, with stoic dedication, submitted the problems for the 1973
Annual Contest shortly before his death on November 26, 1972.
A few minor changes in the statements of contest problems have been
made in this collection for the sake of greater clarity.
Note: The student is told to avoid random guessing, since there is a
penalty for wrong answers. However, if he can definitely eliminate some
of the choices, a random guess among the remaining choices will result
in a positive expected score. A few examples of such elimination are
indicated in the remarks appended to some solutions.
Basil Gordon
Anneli Lax
1973
Suggestions for Using this Book
This problem collection is designed to be used by mathematics clubs,
high school teachers, students, and other interested individuals. Clearly,
no one would profit from doing all the problems, but he would benefit
from those that present a challenge to him. The reader might try himself
on a whole test or on part of a test, with (or preferably without) time
limitations.
He should try to get as far as possible with the solution to a problem.
If he is really stuck, he should look up the answer in the key (p. 55) and
try to work backwards; if this fails, the section of complete solutions
should be consulted.
I n studying solutions, even the successful problem solver may find
sidelights he had overlooked; he may find a more “elegant” solution, or a
way of solving the problem which may lead him deeper into mathematics.
He may find it interesting to change items in the hypothesis and to see
how this affects the solution, or to invent his own problems.
If a reader is interested in a special type of problem, he should consult
the classified index.
The following familiar symbols appear in this book:
Symbol Meaning
similar (if used in connection with plane figures)
approximately equal (if used in connection with numbers)
therefore
identically equal to
less than
less than or equal to
greater than
greater than or equal to
absolute value of the number k
triangle
+ +
the number 1.22 1.2’ 0.2O, i.e., the number 6 when written
in a numeration system with base 2 instead of 10.
congruent
different from
perpendicular to
length of the line segment XY,often denoted by xy in other
books
function of the variable x
parallel to
circular arc with endpoints A and B.
5
I
Problems
1966 Examination
Part 1
1. Given that the ratio of 3x - 4 to y +

15 is constant, and y = 3
when x = 2, then, when y = 12, .t' equals:
(A) i (B) + (C) 4 (D) 4 (El 8
2. When the base of a triangle is increased 10% and the altitude to this
base is decreased loyo,the change in area is
( A ) 1% increase ( B ) 3% increase ( C ) 0%
(D) 3% decrease (E) 1% decrease
3. If the arithmetic mean of two numbers is 6 and their geometric

mean is 10, then an equation with the given two numbers as roots is:
+
( A ) x* 12% 100 = 0 + (B) x2 6 x + + 100 = 0
( C ) x2 - 1 2 -~ 10 = 0 (D) x2 - 1 2 ~ +100 = 0
(E) x2 - 6 x + 100 = 0
4. Circle I is circumscribed about a given square and circle I1 is in-

scribed in the given square. If r is the ratio of the area of circle I
to that of circle 11, then I equals:
( A ) d2 (B) 2 (C) V3 (D) 2d2 (E) 2v3
7
8 THE M A A P R O B L E M BOOK I11
5. The number of values of x satisfying the equation

222- 102
=x-3
9 - sx
is :
(A) zero (B) one (C) two (D) three
(E) an integer greater than 3
6. A B is a diameter of a circle centered at 0. C is a point on the circle

such that angle BOC is 60".If the diameter of the circle is 5 inches,
the length of chord AC, expressed in inches, is:
5V2
(A) 3 (B) 2 (C) (D) 3V3 (E) none of these
7. Let
352 - 29 =-+-
Ni N2
be an identity in x. The nu-
22-3x+2 2-1 r - 2
merical value of N1N2 is:
(A) -246 (B) -210 (C) -29 (D) 210 (E) 246
8. The length of the common chord of two intersecting circles is 16 feet.

If the radii are 10 feet and 17 feet, a possible value for the distance
between the centers of the circles, expressed in feet, is:
(A) 27 (B) 21 (C) @
@! (D) 15 (E) undetermined
10. If the sum of two numbers is 1 and their product is 1, then the sum of
their cubes is:
[Here i denotes 0 . 1
11. The sides of triangle BAC are in the ratio 2:3:4. BD is the angle-
bisector drawn to the shortest side AC, dividing it into segments
A D and CD. If the length of AC is 10, then the length of the
longer segment of AC is:
PROBLEMS: 1966 E X A M I N A T I O N 9
12. The number of real values of x that satisfy the equation

(2W'J) (43&) = 84d-5
is:
(A) zero (B) one (C) two (D) three (E) greater than 3
13. The number of points with positive rational coordinates selected

from the set of points in the xy-plane such that x y 5 5, is: +
(A) 9 (B) 10 (C) 14 (D) 15 (E) infinite
14. The length of rectangle ABCD is 5 inches and its width is 3 inches.
Diagonal AC is divided into three equal segments by points E
and F. The area of triangle BEF, expressed in square inches, is:
(A) f (B) Q (0 4 (D) 5 4 (El 52/68
15. If x -y > x and x + y < y, then

(A) y < x (B) x < y (C) x < y < O (D) x < O , y < O
(E) x < O , y > O
16. If
4"
--
2*
9 h
-
- 8 and 35v = 243, 2' and y real numbers, then xy
17. The number of distinct points common to the curves x2 + 4yz 1

+
=
and 4x2 y2 = 4 is:
(A) 0 (B) 1 (C) 2 (D) 3 (El 4
18. In a given arithmetic sequence the first term is 2, the last term is 29,
and the sum of all the terms is 155. The common difference is:
(A) 3 (B) 2 (C) 3% (D) 9 (El 3
19. Let be the sum of the first n terms of the arithmetic sequence 8,
st
--.and let
12,
sequence 17, 19, -.
sz be the sum of the first n terms of the arithmetic
Assume n # 0. Then SI = s2 for:
(A) no value of n (B) one value of n (C) two values of n
(D) four values of n (E) more than four values of n
10 T H E MAA P R O B L E M B O O K I 1 1
20.t The negation of the proposition “For all pairs of real numbers
a, b, if a = 0, then ab = 0” is: There are real numbers a, b
such that
(A) a # 0 and ab # 0 (B) a # 0 and ab = 0
(C)a = 0 and ab # 0 (D) ab # 0 and a # 0
(E) ab = 0 and a # 0
Part 2
21. An “n-pointed star” is formed as follows: the sides of a convex

.,
polygon are numbered consecutively 1 , 2 , * * k , ., n, n 2 5; for
+
0
all n values of k, sides k and k 2 are non-parallel, sides

n + +
1 and n 2 being respectively identical with sides 1 and 2;
prolong the n pairs of sides numbered k and k 2 until they +
meet. (A figure is shown for the case n = 5).
Let S be the degree-sum of the interior angles at the n points of

the star; then S equals:
(A) 180 (B) 360 (C) 1 8 0 ( n + 2) (D) 180(tt - 2)
(E) 180(n - 4)
22. Consider the statements: (I) d m = 0 (11) d m = ab

(111) 4 - = a +b (IV) 4- = a - b, where we
allow a and b to be real or complex numbers. Those statements for
which there exist solutions other than a = 0 and b = 0, are:
(W,
(A) (1) ,
(m,
(111) , (IV) (B) ( W , ( W , (IV) only
((3 (11, (IV) only (D) ( I W , (IV) only
(El (1) only
t This problem differs from the one given on the original 1966 Examination.
23. If x is real and 4y2 4xy x+ + +

6 = 0, then the complete set of
values of x for which .y is real, is:
(A) x 5 - 2 or x > 3 (B) 2 5 2 or x 2 3
(C) x s - 3 o r x l 2 (D)-35x52
(E) -2 5 ~5 3
24. If lognf A’ = logx M , M # !Y, MAr> 0, M # 1, N # 1, then

MAT equals:
(A) + (B) 1 (C) 2 (I)) 10
(E) a number greater than 2 and lcss than 10
25. If F ( u + 1) = 2F(lL) +1 for It = 1,2, and F(1) = 2,

2
then F( 101) equals:
(A) 49 (B) 50 (C) 51 (U) 32 (E)53
+
26. Let in be a positive integer and let the lines 13x lly = 700 and
y = mx - 1 intersect in a point whose coordinates are integers.
Then in can be:
( A ) 4 only (B) 5 only (C) 6 only (D) 7 only
(E) one of the integers 4, 5, 6, 7 and one other positive integer
27. At his usual rate a man rows 15 miles downstream in five hours less
time than it takes him to return. If he doubles his usual rate, the time
downstream is only one hour less than the time upstream. In miles
per hour, the rate of the stream’s current is:
28. Five points 0, A , U , C, D are taken in order on a straight line

with distances O,-l = a, OB = b, OC = c, and OD = d. P is a
point on the line between B and C and such that A P : P D =
13P:PC. Then O P equals:
b2 - bc - bd
ac +
bd ac
(A)
a +
- b c - (1 a -b+ c-d -a-b+c-d
bc + ad UG - b(1
(I)) a + b + c + tl a +b + +d G
29. The number of positive integers less than lo00 divisible by neither 5
nor 7 is:
(A) 688 (B) 686 (C) 684 (D) 658 (E) 630
30. If three of the roots of x' + ax2 + bx + c = 0 are 1, 2, and 3, then

the valueof a + c is:
(A) 35 (B) 24 (C) -12 (D) -61 (E) -63
Part 3
31. Triangle ABC is inscribed in a circle with center 0'. A circle with
center 0 is inscribed in triangle ABC. A 0 is drawn, and extended
to intersect the larger circle in D . Then we must have:
(A) C D = B D = O'D (B) A 0 = CO = OD
( C ) C D = CO = B D (D) C D = O D = B D
(E) O'B = O'C = OD
32. Let Y be the midpoint of side A B of triangle ABC. Let P be a

point on AB between A and M , and let M D be drawn parallel to
PC and intersecting BC at D . If the ratio of the area of triangle
B P D to that of triangle ABC is denoted by I, then
(A) 4 < Y < 1 depending upon the position of P
(B) I = 3 independent of the position of P
(C) 4 3 I < 1 depending upon the position of P
(D) 4 < Y < $ depending upon the position of P
(E) r = 4 independent of the position of P
33. If ab # 0 and I a I # I b I the number of distinct values of x

satisfying the equation
X - u
-+-=-
b
X-b
a
b
X - u
+-x -ab '
is :
(A) zero (B) one (C) two (D) three (E) four
34. Let r be the speed in miles per hour a t which a wheel, 11 feet in
circumference, travels. If the time for a complete rotation of the
+
wheel is shortened by of a second, the speed r is increased by 5
miles per hour. Then I is:
(A) 9 (B) 10 (C) lo$ (D) 11 (E) 12
3S.t Let 0 be an interior point of triangle ABC, and let s1= OA +

+
OB OC. If sz = A B BC C A , then + +
(A) for every triangle s2 > 2sl, s1 5 s2
(B) for every triangle s2 2 2s1, SI < sz
(C) for every triangle s1> $52, s1 < s2
(D) for every triangle s2 1 2s1, s1 5 s2
(E) neither (A) nor (B) nor (C) nor (D) applies to every triangle
+
36. Let (1 x + 9)"= ao+ a l x + &9+ +&,,a+ beanidentity
in x. If we let s = a0 a2 a4 + + + + en,then s equals:
37. Three men, Alpha, Beta, and Gamma, working together, do a job in 6
hours less time than Alpha alone, in 1 hour less time than Beta alone,
and in one-half the time needed by Gamma when working alone. Let
h be the number of hours needed by Alpha and Beta, working to-
gether, to do the job. Then h equals:
(A9 4 (B) # (0 + (D) # (El 3
38. In triangle ABC themedians AM and CN tosides BC and A B ,
respectively, intersect in point 0. P is the midpoint of side AC,
and M P intersects CN in Q. If the area of triangle OMQ is n,
then the area of triangle ABC is:
(A) 1612 (B) l8rt (C) 21n (D) 24n (E) 27n
t The five choices have been slightly altered by the editors.

14 T H E M A A P R O B L E M BOOK I11
39. In base R1 the expanded fraction FI becomes .373737. * ., and the

expanded fraction F2 becomes .737373..-. In base Rz fraction
F1, when expanded, becomes .252525-*., while fraction F2 be-
comes .525252.... The sum of R1 and R?, each written in the
base ten, is:
(A) 24 (B) 22 (C) 21 (D) 20
40. In this figure AB is a diameter of a circle, centered a t 0, with

radius a. A chord AD is drawn and extended to meet the tangent to
the circle at B in point C. Point E is taken on AC so that
A E = DC. Denoting the distances of E from the tangent through
A and from the diameter A B by x and y, respectively, we can
deduce the relation$
t On the 1966 Examination the last sentence in the statement of Problem 40 was
somewhat different.
PROBLEMS: 1967 EXAMINATION 15
1967 Examination
Part 1
1 . The three-digit number 2a3 is added to the number 326 to give the
three-digit number 5b9. If 5b9 is divisible by 9, then a + b equals:
(A) 2 (B) 4 (C) 6 (D)8 (E) 9
2. An equivalent of the expression
3. The side of an equilateral triangle is s. A circle is inscribed in the

triangle and a square is inscribed in the circle. The area of the
square is:
4. Given (log u ) / p = (log b ) / q = (log c ) / r = log x, all logarithms to

the same base and x # 1. If b 2 / ( a c ) = x*, then y is:
5 . A triangle is circumscribed about a circle of radius r inches. If the

perimeter of the triangle is P inches and the area is K square
inches, then P / K is:
(A) independent of the value of r (B) d 2 / ~ (C) 2/d;
0 )2/r (El r / 2
7 . If a/b < - c / d where a, b, c, d are real numbers and bd # 0,

then:
(A) a must be negative
(B) a must be positive
(C) a must not be zero
(D) a can be negative or zero, but not positive
(E) a can be positive, negative, or zero
8. To m ounces of an m% solution of acid, x ounces of water are

added to yield an ( m- 10)s solution. If m > 25, then x is:
(E) not determined by the given information
9. Let K, in square units, be the area of a trapezoid such that the

shorter base, the altitude, and the longer base, in that order, are in
arithmetic progression. Then:
( A ) K must be an integer (B) K must be a rational fraction
(C) K must be an irrational number (D) K must be an integer
or a rational fraction (E) taken alone neither (A) nor (B) nor
(C) nor (D) is true
10. If a
- +-=b 2' lo" 4-
is an identity for posi-
10"- 1 I@+ 2 ( W - 1)(10"+ 2 )
tive rational values of x , then the value of a - b is:
(A) 4/3 (B) 5/3 ((32 (D) w 4 (El 3
11. If the perimeter of rectangle ABCD is 20 inches, the least value of

diagonal AC, in inches, is:
(A) 0 (B) (C) 10 (D) a (E) noneof these
12. If the (convex) area bounded by the x-axis and the lines y = m x 4, +
x = 1, and x = 4 is 7, then m equals:
(A) -1/2 (B) -2/3 (C) -3/2 (D) - 2 (E) noneof these
13. A triangle ABC is to be constructed given side a (opposite angle

A ) , angle B, and Itc, the altitude from C. If N is the number of
noncongruent solutions, then N
(A) is 1 (B) is 2 (C) must be zero (D) must be infinite
(E) must be zero or infinite
P R O B L E M S : 1967 E X A M I N A T I O N 17
1
14. Let f(t) = -
1-1’
1 # 1. If y = f(i), then i can be expressed
as :
15. The difference in the areas of two similar triangles is 18 square feet,
and the ratio of the larger area to the smaller is the square of an
integer. The area of the smaller triangle, in square feet, is an integer,
and one of its sides is 3 feet. The corresponding side of the larger
triangle, in feet, is:
(A) 12 (B) 9 (C) 6 ~ 2 (D) 6 (E) 3V2
16. Let the product (12) (15) (16), each factor written in base b, equal
+ +
3146 in base b. Let s = 12 15 16, each term expressed in
base b. Then s, in base b, is:
(A) 43 (B) 44 (C) 45 (D) 46 (E) 47
17. If 11 and 12 are the distinct real roots of a? + pz + 8 = 0, then

it must follow that:
(A) 111 + rz 1 > 4d2 (B) I n 1 > 3 or I r z 1 > 3
(c) I > 2 and I r2 I > 2 (D) <0 and r2 < 0
+ 12 1 c 4V2
111 11
(E) 111
18. If x2 - 5 x + 6 < 0 and P = 22 + 51 + 6 then

(A) P can take any real value (B) 20 < P < 30
(C) 0 < P < 20 (D) P < 0
(E) P > 30
19. The area of a rectangle remains unchanged when it is made 2 i

inches longer and 3 inch narrower, or when it is made 23. inches
shorter and inch wider. Its area, in square inches, is:
(A) 30 (B) 80/3 (C) 24 (D) 45/2 (E) 20
20. A circle is inscribed in a square of side m, then a square is inscribed

in that circle, then a circle is inscribed in the latter square, and so on.
If S, is the sum of the areas of the first n circles so inscribed, then,
as tz grows beyond all bounds, S, approaches:
Part 2
21. In right triangle ABC the hypotenuse A B = 5 and leg A C = 3.

The bisector of angle A meets the opposite side in A I . A second
right triangle PQR is then constructed with hypotenuse PQ = A I B
and leg P R = A1C. If the bisector of angle P meets the opposite
side in PI, the length of PP1 is:
22. For natural numbers, when P is divided by D , the quotient is

Q and the remainder is R. When Q is divided by D', the quotient
is Q' and the remainder is R'. Then, when P is divided by DD',
the remainder is:
(A) R + R'D (B) R'+ RD (C) RR' (D) R (E) R'
23. If x is real and positive and grows beyond all bounds, then
+
log3 (6r - 5 ) - logs (2x 1 ) approaches:
(A) 0 (B) 1 (C) 3 (D) 4 (E) nofinitenumber
24. The number of solution-pairs in positive integers of the equation

+
3x 5y = 501 is:
(A) 33 (B) 34 (C) 35 (D) 100 (E) none of these.
25. For every odd number p >1 we have:

(A) ( p - l)*(P-l)- 1 is divisible by p - 2
+
(B) ( p - l)*(P--l) 1 is divisible by p
(C) (p - l ) i ( p l ) is divisible by p
+
(D) ( p - l)i(Pl) 1 is divisible by p 1+
(E) ( p - l)t(Pl) - 1 is divisible by p - 1
26. If one uses only the tabular information loJ = 1O00, 10' = lO,OOO,
21° = 1024, 2" = 2048, 2lZ = 4096, 21a = 8192, then the strongest
statement one can make for loglo2 is that it lies between:
(A) &jand 6 (B) & and (C) 6 and i$
(D) &and&$ (E) &and+$%
27. Two candles of the same length are made of different materials so
that one burns out completely a t a uniform rate in 3 hours and the
other in 4 hours. At what time P.M. should the candles be lighted so

that, at 4 P.M., one stub is twice the length of the other?
( A ) 1:24 (€3) 1:28 (C) 1:36 (D) 1:40 (E) 1:48
28. Given the two hypotheses: I Some Mems are not Ens and I1 NO
Ens are Vees. If “some” means “at least one”, we can conclude that:
(A) Some Mems are not Vees (B) Some Vees are not Mems
(C) No Mem is a Vee (D) Some Mems are Vees
(E) Neither (A) nor (B) nor (C) nor (D) is deducible from the
given statements.
29. A B is a diameter of a circle. Tangents A D and BC are drawn so

that AC and BD intersect in a point on the circle. If A D = a
and BC = b, a # b, the diameter of the circle is:
(A) l a - 61 (B) 3 ( a + b ) (C) fl
30. A dealer bought 11 radios for d dollars, d a positive integer. He

contributed two radios to a community bazaar at half their cost. The
rest he sold at a profit of $8 on each radio sold. If the overall profit
was $72, then the least possible value of tt for the given information
is :
( A ) 18 ( U ) 16 ( C ) 15 (D) 12 ( E) 11
Part 3
+ +
31. Let D = a? bz c2, where a, b are consecutive integers and
c = ab. Then d D is:
( A ) always an even integer
(B) sometimes an odd integer, sometimes not
(C) always an odd integer
(D) sometimes rational, sometimes not
(E) always irrational
32. I n quadrilateral ARCD with diagonals AC and BD intersecting at

0, BO = 4, OD = 6, A 0 = 8, OC = 3, and A B = 6. The
length of A D is:
(A) 9 (B) 10 (C) 6d.3 (D) 8 f l (E) d
166
A o c B
33. In this diagram semi-circles are constructed on diameters AB, AC

and CB, so that they are mutually tangent. If C D I A B, then the
ratio of the shaded area to the area of a circle with CD as radius is:
(A) 1:2 (B) 1:3 (C) d3:7 (D) 1:4 (E) V2:6
34. Points D, E , F are takenrespectivelyonsides AB, BC, and CA

of triangle ABC so that AD:DB = BE:CE = CF:FA = 1:n.
The ratio of the area of triangle DEF to that of triangle ABC is:
d n(n - 1)
(Dl n+ 1
+
35. The roots of 649 - 1442’ 92x - 15 = 0 are in arithmetic pro-
gression. The difference between the largest and smallest roots is:
(A) 2 (€4) 1 (C) 1/2 (D) 3/8 (E) 1/4
36. Given a geometric progression of five terms, each a positive integer

less than 100. The sum of the five terms is 211. If S is the sum of
those terms in the progression which are squares of integers, then
S is:
(A) 0 (B) 91 (C) 133 (D) 195 (E) 211
37. Segments A D = 10, BE = 6, CF = 24 are drawn from the

vertices of triangle ABC, each perpendicular to a straight line RS,
not intersecting the triangle. Points D , E, F are the intersection
points of RS with the perpendiculars. If x is the length of the
perpendicular segment CH drawn to RS from the intersection
point C of the medians of the triangle, then x is:
(A) 40/3 (B) 16 (C) 56/3 (D) 80/3 (E) undetermined
38. Given a set S consisting of two undefined elements “pib” and

“maa”, and the four postulates: PI: Every pib is a collection of
inaas, P2: Any two distinct pibs have one and only one maa in
common, Pa: Every maa belongs to two and only two pibs, Pd:
There are exactly four pibs.
Consider the three theorems: T1: There are exactly six maas,
T2: There are exactly three maas in each pib, Ta: For each maa
there is exactly one other maa not in the same pib with it. The
theorems which are deducible from the postulates are:
(A) T3 only (B) Tz and TS only
(C) T1 and T? only (D) TI and Tt only
(E) all
--
39. Given the sets of consecutive integers { 11, (2, 31, (4, 5, 61,
{ 7,8,9, l o ) , * , where each set contains one more ele‘ment than the
preceding one, and where the first element of each set is one more than
the last element of the preceding set. Let S, be the sum of the
elements in the nth set. Then Szl equals:
(A) 1113 (B) 4641 (C) 5082 (D) 53361 (E) noneof these
40. Located inside equilateral triangle ABC is a point P such that

P A = 6, P B = 8, and PC = 10. To the nearest integer the area of
triangle ABC is:
(A) 159 (B) 131 (C) 95 (D) 79 (E) 50
22 T H E MAA PROBLEM BOOK I11
1968 Examination
Part 1
1. Let P units be the increase in the circumference of a circle resulting

from an increase in u units in the diameter. Then P equals:
(A) -
1
r
(B) u (C) 5 (D) ?r2 (E) 2n
2. The real value of x such that 64"' divided by 4-l equals 2562z
is:
( A ) -$ (R) -5 (C) 0 (D) f (E) j
3. A straight line passing through the point (0, 4) is perpendicular to

the line x - 3y - 7 = 0 . Its equation is:
(A) y + 3 ~ - 4 = 0 (R) y + 3 ~ + 4 = 0
(C) y - 3 3 x - 4 = 0 (13) 3 y + x - 12 = 0
(E) 3y - x - 12 = 0
4. Define an operation * for positive real numbers as a *b =
-.ab Then 4 (4
a+b
* * 4 ) equals:
6. Let side A D of convex quadrilateral A BCD be extended through D ,

and let side BC be extended through C , to meet in point E. Let
S represent the degree-sum of angles CDE and DCE, and let S'
represent the degree-sum of angles BAD and ABC. If Y = S/S',
then:
(A) I = 1 sometimes, Y > 1 sometimes
(B) I = 1 sometimes, I < 1 sometimes
(C) O < r < 1 (D) I > 1 (E)I = 1
P R 0 B L E M S : 1 9 6 8 E X .4 M I N A T I 0 N 23
7. Let 0 be the intersection point of medians A P and CQ of triangle

ABC. If OQ is 3 inches, then Of', in inches, is:
(A) 3 (B) 8 (C) 6 (D)9 (E)undetermined
8. A positive number is mistakenly divided by 6 instead of being

multiplied by 6. Based on the correct answer, the error thus com-
mitted, to the nearest percent, is:
(A) 100 (B) 97 (C) 83 (U) 17 (E)3
9. The sun1 of the real values of x satisfying the equality I x 4-2 I =

2 1 % - 2 1 is:
(A) f (U) 4 (C) 6 (I)) 63 (E) 63
10. Assume that, for a certain school, it is true that

I: Some students ;ire not honest.
11: .411 fraternity niembers are honest.
A necessary conclusion is:
( A ) Some students itre fraternity niemhcrs.
(B) Some fraternit!. members are not students
(C) Some students ;ire not freternity members
(I)) S o fraternity mcmber is a student
(E) No studcnt is ;Lfraternity member.
Part 2
11. If an arc of 60' o n circle I has the same length as an arc of 45' on
circle 11, the ratio of the iirea of circle I to that of circle I1 is:
( A ) i6:g ( u ) 9:io (c) 4:3 cr)) 3:4
(E) none of these
12. A circle passes through the vertices of a triangle with side-lengths

73, 10, 12). The radius of the circle is:
13. If m and it are the roots of x2 + mw + 11 = 0, m # 0, $1 # 0,

then the sum of the roots is:
(A) -3 (B) -1 (C) ) (U) 1 (E)undetermined
14. If x and y are non-zero numbers such that x = 1 + Y-1 and

1
+
y = 1 -, then y equals
X
(A) X - 1 (B) 1 - x (C) l + x (D) -X (E) x
15. Let P be the product of any three consecutive positive odd integers.
The largest integer dividing all such P is:
(A) 15 (B) 6 (0 5 (D) 3 (E) 1
1 1
16. If x is such that - < 2 and
X
-
X
> -3, then:
(A) -a < x < 3 (B) -4 < x < 3 (C) x > t

( D ) x > + or - - f < x < O (E) x > + or x < - f
17. Let f ( n ) =
XI + xz + + xn
, where n is a positive integer. If
n
xk = (- k = 1, 2, .,n, the set of possible values of f ( n ) is:
18. Side A B of triangle ABC has length 8 inches. Line DEF is drawn
parallel to A B sothat D isonsegment AC, and E isonsegment
BC. Line A E extended bisects angle FEC. If DE has length 5
inches, then the length of CE, in inches, is:
(A) 4L (B) 13 (C) 4p (D) -$e (E) %?
19. Let n be the number of ways that 10 dollars can be changed into
dimes and quarters, with a t least one of each coin being used. Then n
equals:
(A) 40 (B) 38 (C) 21 (D) 20 (E) 19
20. The measures of the interior angles of a convex polygon of n sides

are in arithmetic progression. If the common difference is 5' and the
largest angle is IW",then n equals:
(A) 9 (B) 10 (C) 12 (D) 16 (E) 32
Part 3
21. If S = l ! 2!+ + 3! + + 99!, then the units’ digit in the

value of S is:t
(A) 9 (B) 8 (C) 5 (D) 3 (El 0
22. A segment of length 1 is divided into four segments. Then there
exists a quadrilateral with the four segments as sides if and only if
each segment is:
(A) equal to $
+
(B) equal to or greater than and less than $
(C) greater than 4 and less than $
(D) greater than 4 and less than f
(E) less than 4
23. If all the logarithms are real numbers, the equality

log ( x + 3) + log ( x - 1) = log (9- 2x - 3)
is satisfied for:
(A) all real values of z
(B) no real values of x
(C) all real values of x except x = 0
(D) no real values of x except x = 0
(E) all real values of x except x = 1
24. A painting 18” X 24“ is to be placed into a wooden frame with the
longer dimension vertical. The wood at the top and bottom is twice
as wide as the wood on the sides. If the frame area equals that of the
painting itself, the ratio of the smaller to the larger dimension of the
framed painting is:
(A) 1:3 (B) 1:2 (C) 2:3 (D) 3:4 (E) 1 : l
25. Ace runs with constant speed and Flash runs x times as fast,
x > 1. Flash gives Ace a head start of y yards, and, a t a given
signal, they start off in the same direction. Then the number of yards
Flash must run to catch Ace is:
(A) xy (B) -’
X+Y
(C> -xy
x- 1
(D)
x+ 1
5’
(E) X-
+ Y
x- 1
tThesymb1 n! denotes 1.2- ... ( n - 1)n; thus S! = 1.2.3.4.5 = 120.

+ + +
26. Let S = 2 4 6 +
2X, where S is the smallest positive
integer such that S > l,OOO,OOO. Then the sum of the digits of
N is:
(A) 27 (B) 12 (C) 6 (D) 2 (E) 1
+
27. Let S,,= 1 - 2 3 - 4 + + (-l)n-lir, = 1,2, Then
+ +
11 a s . .
SIT Saa S ~ O equals:

(A) 0 (B) 1 (C) 2 (D) - 1 (E) - 2
28. If the arithmetic mean of a and b is double their geometric mean,

with a > b > 0, then a possible value for the ratio a/b, to the
nearest integer, is
(A) 5 (B) 8 (C) 11 (D) 14 (E) noneof these
29. Given the three numbers x, y = xz, z = x(") with .9 < x < 1.0.
Arranged in order of increasing magnitude, they are:
(A) %Z,Y (B) x,y,z (0 y , x , z (D) Y,Z,X (El Z,Z,Y
30. Convex polygons PI and P2 are drawn in the same plane with 111
and nz sides, respectively, $11 5 112. If P1 and €'2 do not have
any line segment in common, then the masimum number of inter-
sectionsof PI and €'2 is:
(A) 2nl (B) 2 4 (C) I Z ~ J Z ~ (D) n1 + 112 (E) none of these
Part 4
31. In this diagram, not drawn to scale, figures I and 111 are equilateral
triangular regions with respective areas of 32d3 and Sd3 square
inches. Figure I1 is a square region with area 32 sq. in. Let the
length of segment AD be decreased by 12$% of itself, while the
lengths of AB and CD remain unchanged. The percent decrease in
the area of the square is:
(A) 123 (B) 25 (C) 50 ( D ) 75 (E) 873
P R O B L E M S : 1968 EXAMINATION 27
32. A and B move uniformly along two straight paths intersecting a t

right angles in point 0. When A is a t 0, B is 500 yards short of
0. In 2 minutes they are equidistant from 0, and in 8 minutes more
they are again equidistant from 0. Then the ratio of A’s speed
to B’s speed is:
(A) 4:s (B) 5:6 (C) 2:3 (D) 5:8 (E) 1:2
33. A number N has three digits when expressed in base 7. When N is

expressed in base 9 the digits are reversed. Then the middle digit is:
(A) 0 (B) 1 (C) 3 (D) 4 (El 5
34. With 400 members voting the House of Representatives defeated a

bill. A re-vote, with the same members voting, resulted in passage of
the bill by twice the margint by which it was originally defeated. The
number voting for the bill on the re-vote was # of the number
voting against it originally. How many more members voted for
the bill the second time than voted for it the first time?
(A) 75 ( B ) 60 (C) 50 (D) 45 (E) 20
35. In this diagram the center of the circle is 0, the radius is a inches,
chord EF is parallel to chord CD, 0, G, H , J are collinear, and
G is the midpoint of CD. Let K (sq. in.) represent the area of
trapezoid CDFE and let R (sq. in.) represent the area of rectangle
ELMF. Then, as CD and EF are translated upward so that OG
increases toward the value a, while J R always equals HG, the
ratio K:R becomes arbitrarily close to:
t In this context, margin of defeat (passage) is defined as the number of nays minus
the number of ayes (nays-ayes).
1969 Examination
Part 1
1. When x is added to both the numerator and the denominator of the

fraction a/b, a # b, b # 0, the value of the fraction is changed to
c/d. Then x equals:
1 ad - bc ad - bc bc - ad
(A) -
c-d
(B) -c-d
(C) -
c+d
(D) -
G-d
bc - ad
c+d
2. If an item is sold for x dollars, there is a loss of 15% based on the
c0st.t If, however, the same item is sold for y dollars, there is a
profit of 15% based on the c0st.t The ratio y:x is:
(A) 23:17 (B) 17y:23 (C) 23x:17
(D) dependent upon the cost (E) none of these.
3. If N, written in base 2, is 11000, the integer immediately preceding

N , written in base 2, is:
(A) lo001 (B) 10010 (C) 10011 (D) 10110 (E) 10111
( a , b) *
4. Let a binary operation * *
on ordered pairs of integers be defined by
+
(c, d ) = ( a - c, b d ) . Then, if (3, 2) (0, 0) and
( x , y) >k (3, 2) represent identical pairs, x equals:
(A) - 3 (B) 0 (C) 2 (D) 3 (E) 6
5. If a number A', N # 0, diminished by four times its reciprocal,

equals a given real constant R, then, for this given R, the sum of all
such possible values of N is:
1
(A) (B) R (C) 4 (D) f (E) - R
1
t l % loss based on cost means loss of --.cost, r% profit based on cost means
100
r
profit of --.cost.
100
6 . The area of the ring between two concentric circles is 123r square
inches. The length of a chord of the larger circle tangent to the smaller
circle, in inches, is:
5
(A) (B) 5 (C) 5\/2 (D) 10 (E) lo\/z
7. If the points (1, yl) and (- 1, y2) lie on the graph of y = ax2 +
bx + t, and y1 - y~ = -6, then b equals:
(A) -3 (B) 0 (C) 3 (D) G (E)

a+c
2
8. Triangle A BC is inscribed in a circle. The measure of the non-over-
lapping minor arcs AB, BC, and CA are, respectively, x 75', +
2x +25', 3x - 22'. Then one interior angle of the triangle, in
degrees, is:
(A) 573 (B) 59 (C)60 (D)61 (E) 122
9. The arithmetic mean (ordinary average) of the fifty-two successive
positive integers beginning with 2 is:
(A) 27 (B) 27f (C) 273 (D) 28 (E) 283
10. The number of points equidistant from a circle and two parallel
tangents to the circle is:
(A) 0 (B) 2 (C) 3 (D) 4 (E) infinite
Part 2
11. Given points P(-1, -2) and Q(4, 2 ) in the xy-plane; point
+
R( 1, m) is taken so that PR RQ is a minimum. Then m equals:
(A) -Q (B) - 3 (C) -3 (D) 3 (E) either - f o r + .
12. Let F =
6x2 + 16x + 3m be the square of an espression which is
6
linear in x. Then m has a particular value between:
(A) 3 and 4 (B) 4 and 5 (C) 5 and 6 (D) -4and -3
(E) -6and - 5
13. A circle with radius r is contained within the region bounded by a

circle with radius R. Thc are2 bounded by the larger circle is a / b
times the area of the region outside the smaller circle and inside the
larger circle. Then R : r equals:
(A) fi:@(B) 4 u : - (C) 4:-
(D) a : d x (E) b : d a
x* -4
14. The complete set of %-values satisfying the inequality -- > 0
9- 1
is the set of all x such that:
(A) x > 2 or x < - 2 or -1 < x < 1 (B) x > 2 or x < - 2
(C) x > 1 or x < - 2 (D) x > 1 or a < - 1
(E) x is any real number except 1 or - 1
15. In a circle with center a t 0 and radius Y , chord A B is drawn with

length equal to r(units). From 0 a perpendicular to A B meets
A B at M . From M a perpendicular to OA meets O A at D. In
terms of I the area of triangle MD.4, in appropriate square units, is:
16. When (a - b ) " , R 2 2, ab # 0, is espanded by the binomial

theorem, it is found that, when a = kb, where k is a positive
integer, the sum of the second and third terms is zero. Then ir equals:
(A) j k ( k - 1) (B) $ k ( k + 1) (C) 2h - 1 (D) 2k
(El 2 k + 1
17. The equation 222 - 8.2" + 12 = 0 is satisfied by:

(A) log3 (B) 4 log6 (C> 1 + log 8 log 3
(D) 14- log2
(E) none of these
18. The number of points common to the graphs of

(x-y+2)(3x+y-4) =Oand ( x + y - 2 ) ( 2 x - 5 ~ + 7 ) = O
is :
(A) 2 (B) 4 (C) 6 (D) 16 (E) infinite
19. The number of distinct ordered pairs ( x , y), where x and y have
positive integral values satisfying the equation x'y' - 109y2 9 +
= 0,is:
(A) 0 (B) 3 (C) 4 (D) 12 (E) infinite
20. Let P equal the product of 3,659,893,456,789,325,678 and

342,973,489,379,256. The number of digits in P is:
(A) 36 (B) 35 (C) 34 (D) 33 (E) 32
Part 3
21. If the graph of x2 +9 = m is tangent to that of x +y = 1/%,

then :
1
(A) m must equal 3 (B) m must equal -
d
(C) m must equal d (D) m must equal 2
(E) m may be any non-negative real number
22. Let K be the measure of the area bounded by the x-axis, the line
x = 8, and the curve defined by
j = { ( x , y ) I y = x when 0 < x < 5, y = 2x - 5 when 5 < x 5 8 ) .
Then K is:
(A) 21.5 (B) 36.4 (C) 36.5 (D) 44
(E) less than 44 but arbitrarily close to it.
23. For any integer n greater than 1, the number of prime numbers
greater than n! +
1 and less than n! IZ is:t +
It
(A) 0 (B) 1 (C) 3 for n even, n-
+ for n odd
(D) n - 1 (E) IZ 2
24. When the natural numbers P and P', with P > P', are divided
by the natural number D, the remainders are R and R', respec-
tively. When PP' and RR' are divided by D, the remainders are Y
and Y ' , respectively. Then:
(A) Y > I' always (B) Y < I' always
(C) Y > I' sometimes, and Y < I' sometimes
(D) Y >I ' sometimes, and Y = I' sometimes
(E) Y = I' always
+
25. If it is known that log2 a log2 b 2 6, then the least value that can
be taken on by a b is:+
(A) 2 4 6 (B) 6 (C) 8d (D) 16 (E) none of these.
tThesymbol n! denotes 1.2. (n - 1)n; thus 5 ! = 1-2.3.4-5 = 120.

26. A parabolic arch has a height of 16 inches and a span of 40 inches.

The height, in inches, of the arch at a point 5 inches from the center
M is:
(A) 1 (B) 15 (c) 1st (D) 15+ (E) 15%
27. A particle moves so that its speed for the second and subsequent miles
varies inversely as the integral number of miles already traveled. For
each subsequent mile the speed is constant. If the second mile is
traversed in 2 hours, then the time, in hours, needed to traverse the
nth mile is:
2 n-1 2
(A) -
n-1
(B) 2 (C) n- (D) 2n (E)2 ( n - 1)
28. Let n be the number of points P interior to the region bounded by a

circle with radius 1, such that the sum of the squares of the distances
from P to the endpoints of a given diameter is 3. Then n is:
(A) 0 (B) 1 (C) 2 (D) 4 (E) infinite
29. If x = tll(+1) and y = ft/(t-l), f > 0, 1 # 1 , a relation between x

and y is:
(A) = xl/* (B) yl'+ = x* (C) = X* (D) P yY
(E) none of these
30. Let P be a point of hypotenuse A B (or its extension) of isosceles

right triangle ABC. Let s = A P PB2. Then: +
(A) s < 2 C P for a finite number of positions of P
(B) s < 2 C P for an infinite number of positions of P
(C) s = 2 C P only if P is the midpoint of .4B or an endpoint
of A B
(D) s = 2CP always
(E) s > 2 C P if P is B trisection point of .4B
Part 4
31. Let OABC be a unit square in the xy-plane with O(O,O), A ( l , 0),
B ( 1 , l ) and C(0, 1 ) . Let u = a? - f and v = 2xy be a trans-
formation of the xy-plane into the uv-plane. The transform (or
image) of the square is:
i i
32. Let a sequence (u.) be d e h e d by U I = 5 and the relation

Un+l-un=3+4(n- l ) , n = 1,2,3,*.-. If u, isexpressedasa
polynomial in n, the algebraic sum of its coefficients is:
(A) 3 (B) 4 (C) 5 (D) 6 (E) 11
33. Let S, and T, be the respective sums of the first IZ terms of two
+ +
arithmetic series. If S,: T, = (7n 1) :(4n 27) for all n, the
ratio of the eleventh term of the first series to the eleventh term of the
second series is:
(A) 4:3 (B) 3:2 (C) 7:4 (D) 78:71
(E) undetermined
34. The remainder R obtained by dividing xlOO by x2 - 3x 2 is a +

polynomial of degree less than 2. Then R may be written as:
(A) 21°0- 1 -
(B) 2'"(x - 1) ( X - 2) (C) 2'0°(x - 3)
(D) ~ ( 2-' 1)
~ +
2(299- 1) +
(E) 21°0(x 1) - ( X 2) +
35. Let L ( m ) be the x-coordinate of the left end point of the inter-
section of the graphs of y = x2 - 6 and y = m, where
-6 < m < 6. Let I = [ L ( - m ) - L(m)]/m.Then, as m is made
arbitrarily close to zero, the value of I is:
(A) arbitrarily close to zero (B) arbitrarily close to -1

fl
2
(C) arbitrarily close to -
d
(D) arbitrarily large
(E) undetermined
1970 Examination
Part 1
1. The fourth power of dl + d m is

(A) d + \13 ( B ) i(7 + 3 4 5 ) (c) 1 -I- 2\13 (D) 3
+
(E) 3 2\/2
2. A square and a circle have equal perimeters. The ratio of the area
of the circle to the area of the square is
(A) 4 / r (B) ,/a ( C ) 4/1 (D) ./z/r (E) r/4
3. If x = 1 + 2 p and y = 1 + 2-P, then y in terms of x is
x- 1
(E) -
X
4. Let S be the set of all numbers which are the sum of the squares
of three consecutive integers. Then we can say that
(A) No member of S is divisible by 2
(B) No member of S is divisible by 3 but some member is divisible
by 1 1
(C) No member of S is divisible by 3 or by 5
(D) No member of S is divisible by 3 or by 7
(E) None of these
5. If f ( x ) =
+ x2 ,
x4
- then f ( i ) , where i = G,is equal to
x+ 1
6. The smallest value of x2 + 8x for real values of x is

(A) -16.25 (B) -16 (C) -15 (D) -8
(E) None of these
7 . Inside square ABCD with side s, quarter-circle arcs with radii s

and centers at A and B are drawn. These arcs intersect a t a point
X inside the square. How far is X from side CD?
(A) +)s(\lf+ 4) (B) w3 (C) 3 4 1 +a)
(D)a4a- 1) (El b ( 2 - a)
8. If a = logs 225 and b = log2 15, then
(A) a = b/2 (B) a = 2b/3 (C) a = b (D) 6 = a/2
(E) a = 3b/2
9. Points P and Q are on line segment A B , and both points are on

the same side of the midpoint of A B . Point P divides A B in the
ratio 2:3, and Q divides A B in the ratio 3:4. If PQ = 2, then
the length of segment A B is
(A) 12 (B) 28 (C) 70 (D) 75 (E) 105
10. Let F = .48181***be an infinite repeating decimal with the digits

8 and 1 repeating. When F is written as a fraction in lowest terms,
the denominator exceeds the numerator by
(A) 13 (B) 14 (C) 29 (D) 57 (E) 126
Put 2
11. If two factors of 2x9 - Itx + k are x +2 and x - 1, the value

of I2h- 3kI is
(A) 4 (B) 3 (C) 2 (D) 1 (El 0
12. A circle with radius I is tangent to sides A B , A D , and C D of
rectangle ABCD and passes through the midpoint of diagonal AC.
The area of the rectangle, in terms of I, is
(A) 4 9 (B) 6r" (C) 8r* (D) 12r2 (E) 2W
13. Given the binary operation * *

defined by a b = a* for all posi-
tivenumbers a and b. Then for all positive a, b, c, t ~ we
, have
(B) * ( b * * b) *
* * n) * b (D) * b)" *
(A) a * b = b * a a c) = ( a c
(C) (a b") = (a (bn) (a =a
(E) None of these
+ +
14. Consider xz p x q = 0, where p and q are positive numbers.
If the roots of this equation differ by 1, then p equals
15. Lines in the xy-plane are drawn through the point ( 3 , 4 ) and the
trisection points of the line segment joining the points (-4,s) and
(5, - 1). One of these lines has the equation
(A) 3~ - 2y - 1= 0 +
(B) 4~ - 5y 8 = 0
+
(C)5~ 2y - 23 = 0 +
(D)x 7y - 31 = 0
+
(E)x - 4y 13 = 0
16. If F ( n ) is a function such that F(1) = F ( 2 ) = F ( 3 ) = 1, and

F ( n - 1) + 1
such that F ( n + 1) = F ( n ).F(n - 2)
for IZ 2 3, then F ( 6 )
17. If I > 0, then for all p and q such that pq # 0 and pr > qr,
we have
(A) - p > - q (B) -P > q (C) 1 > - q / P
(D) 1 < q / P (E) None of these
18. - d 3-2fl is equal to

(A) 2 (B) 2 d (C) 4fl (D) 4 6 (E) 2 f l
19. The sum of an infinite geometric series with common ratio r such
that I I 1 < 1 is 15, and the sum of the squares of the terms of this
series is 45. The first term of the series is
(A) 12 (B) 10 (C) 5 (D) 3 (E)2
20. Lines HK and BC lie in a plane. M is the midpoint of line seg-

ment BC, and BH and CK are perpendicular to HK. Then we
(A) always have M H = M K (B) always have MH > BK
(C) sometimes have MH = MK but not always
(D) always have MH > M B ( E ) always have BH < BC
Put 3
21. On an auto trip, the distance read from the instrument panel was
450 miles. With snow tires on for the return trip over the same route,
the reading was 440 miles. Find, to the nearest hundredth of an inch,
the increase in radius of the wheels if the original radius was 15 inches.
(A) .33 (B) .34 (C) .35 (D) .38 (E) .66
22. If the sum of the first 3n positive integers is 150 more than the
sum of the first n positive integers, then the sum of the first 412
positive integers is
(A) 300 (B) 350 (C) 400 (D) 450 (E)600
23. The number lO!t (10 is written in base l o ) , when written in the
base 12 system, ends with exactly R zeros. The value of R is
(A) 1 (B) 2 (C) 3 (D) 4 (E) 5
24. An equilateral triangle and a regular hexagon have equal perimeters.

If the area of the triangle is 2, then the area of the hexagon is
(A) 2 (B) 3 (0 4 (D) 6 (E) 12
25. For every real number z, let [z] be the greatest integer which is
less than or equal to z. If the postal rate for first class mail is six
cents for every ounce or portion thereof, then the cost in cents of
first-class postage on a letter weighing W ounces is always
(A) 6W (B) 6CWl (C) 6<CW] - 1) (D) 6 ( c W l + 1)
(E) --6c--1
26. The number of distinct points in the zy-plane common to the graphs
of ( ~ + y - 5 ) ( 2 ~ - 3 ~ + 5 =
) O m d ( ~ - ~ + 1 ) ( 3 ~ + 2 y 12)
-
=ois
(A) 0 (B) 1 (C) 2 (D) 3 (E) 4 (F) infinite
27. In a triangle, the area is numerically equal to the perimeter. What

is the radius of the inscribed circle?
(A) 2 (B) 3 (C) 4 (D) 5 (El 6
tThesymbo1 n! denotes 1.2. .(n - 1)n; thus 51 = 1-2.3.4.5 = 120.
28. In triangle ABC, the median from vertex A is perpendicular to

the median from vertex B. If the lengths of sides AC and BC are
6 and 7 respectively, then the length of side A B is
(A) dn (B) 4 (C) 4+ (D) 2 4 (E) 41
29. I t is now between 1O:OO and 11:OO o’clock, and six minutes from
now, the minute hand of a watch will be exactly opposite the place
where the hour hand was three minutes ago. What is the exact time
now?
( A ) 10:05* (B) 10:073 (C) 1O:lO (D) 10:15
(E) 10:17$
30. In the accompanying figure, segments A B and CD are parallel,

the measure of angle D is twice that of angle B , and the measures
of segments .4D and C D are a and b respectively. Then the
measure of A B is equal to
(A) $a+ 2b (B) +b + ZU ( C ) 2~ - b (D) 4b -
(El a b+
B
Part 4
31. If a number is selected at random from the set of all five-digit num-
bers in which the sum of the digits is equal to 43, what is the prob-
ability that this number will be divisible by 111
( A ) 2/5 (B) 1/5 (C) 1/6 (D) 1/11 (E)1/15
32. A and B travel around a circular track at uniform speeds in oppo-

site directions, starting from diametrically opposite points. If they
start a t the same time, meet first after B has travelled 100 yards,
and meet a second time 60 yards before A completes one lap, then
the circumference of the track in yards is
(A) 400 (B) 440 (C) 480 (D) 560 (E) 880
40 THE MAA PROBLEM BOOK I11
33. Find the sum of the digits of all the numbers in the sequence
1, 2, 3, 4, '*., 1 m .
(A) 180,001 (B) 154,756 (C) 45,001 (D)154,755
(E)270,001
34. The greatest integer that will divide 13,511, 13,903 and 14,589 and
leave the same remainder is
(A) 28 (B) 49 (C) 98
(D) an odd multiple of 7 greater than 49
(E) an even multiple of 7 greater than 98
35. A retiring employee receives an annual pension proportional to the

square root of the number of years of his service. Had he served a
years more, his pension would have been p dollars greater, whereas,
had he served b years more (b # a), his pension would have
been q dollars greater than the original annual pension. Find his
annual pension in terms of a, b, p , and q.
1971 Examination
Part 1
1 . The number of digits in the number A' = 212 X j8 is
(A) 9 (B) 10 (C) 11 (D) 12 (E) 20
2. If b men take c days to lay f bricks, then the number of days

it will take G men working a t the same rate to lay b bricks, is
(A) fb2 (B) b / f 2 (0f z l b (D) a2/f (El f / b 2
3 . If the point(2, -4) lies on the straight line joining the points
(0,s) and ( - 4 , 0 ) in the xy-plane, then z is equal to
(A) -2 (B) 2 (C) -8 (D) 6 (E) -6
4. After simple interest for two months at 5% per annum was credited,
a Boy Scout Troop had a total of $255.31 in the Council Treasury.
The interest credited was a number of dollars plus the following
number of cents
(A) 1 1 (B) 12 ( C ) 13 (D) 21 (E) 31
5. Points .4, H, Q, D, and C lie on the circle shown and the measures
n
of arcs B
? and QD are 42" and 38" rcspectively. The sum of the
measures of angles P and Q is
(A) 80" (B) 62" (C) 40" (D) 46" (E) None of these
6. Let * be a symbol denoting the binary operation on the set S of

all non-zero real numbers as follows: For any two numbers a and
*
b of S, a b = 2ab. Then the one of the following statements
which is not true, is
(A) * is commutative over S
* (B) *is associative over S
(C) 3 is an identity element for in S
(D) Every element of S has an inverse for
(E) 1/2a is an inverse for * *
of the element a of S
8. The solution set of 6x2 + 5x < 4 is the set of all values of x such
that
< x < 1 (B) -4 < x < 4 (C) -3 <x <$
(A) - 2
(D) x < a or x > -+(E) 2 < - $ or x > +
9. An uncrossed belt is fitted without slack around two circular pulleys
with radii of 14 inches and 4 inches. If the distance between the
points of contact of the belt with the pulleys is 24 inches, then the
distance between the centers of the pulleys in inches is
(A) 24 (B) 2 4 1 1 9 (C) 25 (D) 26 (E) 44z
10. Each of a group of 50 girls is blonde or brunette and is blue or brown-
eyed. If 14 are blue-eyed blondes, 31 are brunettes, and 18 are brown-
eyed, then the number of brown-eyed brunettes is
(A) 5 (B) 7 (C) 9 (D) 11 (E) 13
Part 2
11. The numeral 47 in base a represents the same number as 74 in

base b. Assuming that both bases are positive integers, the least
+
possible value for a b written as a Roman numeral, is
(A) XI11 (B) XV (C) XXI (D) XXIV (E) XVI
12. For each integer N > 1, there is a mathematical system in which

two or more integers are defined to be congruent if they leave the
same non-negative remainder when divided by AT. If 69, 90, and
125 are congruent in one such system, then in that same system, 81
is congruent to
(A) 3 (B) 4 (C) 5 (D) 7 (El 8
13. If (1.0025) 10 is evaluated correct to 5 decimal places, then the digit

in the fifth decimal place is
(A) 0 (B) 1 (C) 2 (D) 5 (E) 8
14. The number (248- 1) is exactly divisible by two numbers between

60 and 70. These numbers are
(A) 61,63 (B) 61,65 (C) 63,65 (D) 63,67 (E) 67,69
15. An aquarium on a level table has rectangular faces and is 10 inches

wide and 8 inches high. When it was tilted, the water in it just
covered an 8” x 10” end but only three-fourths of the rectangular
bottom. The depth of the water when the bottom was again made
level was
(A) 23” (B) 3” (C) 3:’’ (D) 3q” (E) 4”
16. After finding the average of 35 scores, a student carelessly included

the average with the 35 scores and found the average of these 36
numbers. The ratio of the second average to the true average was
(A) 1:l (B) 35:36 (C) 36:35 (D) 2:l (E) Noneof these
17. A circular disk is divided by 2n equally spaced radii ( n > 0) and

one secant line. The maximum number of non-overlapping areas into
which the disk can be divided is
(A) 2n +1 (B) 211 +2 (C) 3n - 1 (D) 3n (E) 3 n + I
18. The current in a river is flowing steadily a t 3 miles per hour. A motor
boat which travels at a constant rate in still water goes downstream
4 miles and then returns to its starting point. The trip takes one
hour, excluding the time spent in turning the boat around. The
ratio of the downstream to the upstream rate is
(A) 4:3 (B) 3:2 ( C ) 5:3 (D) 2:l (E) 5:2
19. If the line y = mx +

1 intersects the ellipse xz + 4 9 = 1 exactly
once, then the value of m2 is
(A) 4 (B) f (C) s (D) 4 (El 8
20. The sum of the squares of the roots of the equation x2 + 2hx = 3
is 10. The absolute value of h is equal to
(A) - 1 (B) 3 (C) 4 (D) 2 (E) None of these
Part 3
22. If w is one of the imaginary roots of the equation x3 = 1, then

+ +
the product (1 - w w’)(1 w - UP) is equal to
(A) 4 (B) w (C) 2 (Dl (El 1
23. Teams A and B are playing a series of games. If the odds for
either team to win any game are even and Team A must win two
or Team B three games to win the series, then the odds favoring
Team A to win the series are
(A) 11 to 5 (B) 5 to 2 (C) 8 t o 3 (D) 3 to 2 (E) 13 t o 6
1
1 1
1 2 1
1 3 3 1
1 4 6 4 1
etc.
24. Pascal’s triangle is an array of positive integers (see figure), in

which the first row is 1, the second row is two l’s, each row begins
and ends with 1, and the kth number in any row when it is not 1,
is the sum of the kth and (k - 1)th numbers in the immediately
preceding row. The quotient of the number of numbers in the first
n rows which are not 1’s and the number of 1’s is
n2 - n n2 - n n2 - 2n +
n2 - 3n 2
(A) -
2n - 1
(B) 4n
- -2 (c) (Dl 4n- 2
(E) None of these
25. A teen age boy wrote his own age after his father’s. From this new
four place number he subtracted the absolute value of the difference
of their ages to get 4,289. The sum of their ages was
(A) 48 (B) 52 (C) 56 (D) 59 (E) 64
26. I n triangle .4 BC, point F divides side AC in the ratio 1: 2. Let

E be the point of intersection of side BC and AG where G is
the midpoint of BF. Then point E divides side BC in the ratio
(A) 1:4 (B) 1:3 (C) 2:s (D) 4:11 (E) 3:8
27. A box contains chips, each of which is red, white, or blue. The num-
ber of blue chips is at least half the number of white chips, and a t
most one third the number of red chips. The number which are
white or blue is at least 55. The minimum number of red chips is
(A) 24 (B) 33 (C) 45 (D) 54 (E) 57
28. Nine lines parallel to the base of a triangle divide the other sides
each into 10 equal segments and the area into 10 distinct parts. If
the area of the largest of these parts is 38, then the area of the
original triangle is
( A ) 180 (B) 190 (C) 2 0 0 (D) 210 (E) 240
29. Given the progression 101/11, ---

103/", 104/11, , 1On/l1. The least
positive integer IZ such that the product of the first n terms of
the progression exceeds 100,000 is
(A) 7 (B) 8 (C) 9 (D) 10 (E) 1 1
30. Given the linear fractional transformation of x into fl(x) = -

2x- 1
x+ 1
Define
fn+i(x) = fi(fn(x)) for 12 = 1,2,3, * * . -
I t can be shown that f a ~= f ~ it
; follows that f28(x) is
1 2- 1 1
(A) y (B) -
X
(C) -
X
(D) -
1-x
(E) None of these
46 THE M A A P R O B L E M B O O K I11
Part 4
31. Quadrilateral ABCD is inscribed in a circle with side AD, a diam-

eter of length 4. If sides AB and BC each have length 1, then side
CD has length
33. If P is the product of n quantities in geometric progression, S

their sum, and S' the sum of their reciprocals, then P in terms
of S, S', and IZ is
(A) (SS')n/2 (B) (S/S')n/Z (C) (SS')11-2 (D) (S/S')"
(E) (S'/S) ( n - 1 ~ 2
34. An ordinary clock in a factory is running slow so that the minute

hand passes the hour hand a t the usual dial positions (12 o'clock,
etc.) but only every 69 minutes. A t time and one-half for overtime,
the extra pay to which a $4.00 per hour worker should be entitled
after working a normal 8 hour day by that slow running clock, is
( A ) $2.30 (B) $2.60 (C) $2.80 (D) $3.00 (E) $3.30
35. Each circle in an infinite sequence with decreasing radii is tangent

externally to the one following it and to both sides of a given right
angle. The ratio of the area of the first circle to the sum of areas of
all other circles in the sequence is
(A) ( 4 + 3 d 2 ) : 4 (B) M : 2 (C) ( 1 6 + 12d2):I
(D) ( 2 + 2 6 ) : 1 (E) (3+2d2):1
1972 Examination
Part 1
1. The lengths in inches of the three sides of each of four triangles I,

11, 111, and I V are as follows:
I 3,4, and5 I11 7, 24, and 25
I1 4, 74, and 84 IV 3+, 4$, and 53.
Of these four given triangles, the only right triangles are
(A) I a n d I I (B) 1 a n d I I I (C) I a n d I V (D) I, 1 1 , a n d I I I
(E) I, 11, and 1V
2. If a dealer could get his goods for 8% less while keeping his selling
price fixed, his profit, based on cost,t would be increased to
(x +lo)% from his present profit of 2% which is
(A) 12% (B) 15% (C) 30% (D) 50% (E) 75%
1 - ifl 1
3. If 2 = - 2
, where i = g,
then - is equal to
9 - X
( A ) -2 (B) - 1 (C) 1 + i f l (D) 1 (E) 2
4. The number of solutions to { 1, 2) CX { 1 , 2 , 3 , 4 , S), where X

is a set, is
(A) 2 (B) 4 (C) 6 (D) 8 (E)None of these
5 . From among 21/2, 3'13, W 8 ,91/sthose which have the greatest and
the next to the greatest values, in that order, are
(A) 3lI3, 2'/* (B) 3'la, 81/8 (C) 3lI3,9ll9 (D) 8lI6,9llg
(E) None of these
6. If 3*= + 9 = 10(3=), then the value of 9 + 1 is

(A) lonly (B) 5only (C) l o r 5 (D) 2 (E) 10
X Y
7. If y z : z x : x y = 1:2:3, then - :- isequal to
ye zx
(A) 3:2 (B) 1:2 (C) 1:4 (D) 2:l (E)4 : l
I
t r% profit based on cost means --.cost.
100
8. If I x - logy I = x + logy, where x and logy are real, then

(A) x = 0 (B) y = 1 (C) x = 0 and y = 1
(D) x ( y - 1) =0 (E) None of these
9. Ann and Sue bought identical boxes of stationery. Ann used hers to
write 1-sheet letters and Sue used hers to write 3-sheet letters. Ann
used all the envelopes and had 50 sheets of paper left, while Sue
used all of the sheets of paper and had 50 envelopes left. The num-
ber of sheets of paper in each box was
(A) 150 (B) 123 (C) 120 (D) 100 (E) 80
10. For x real, the inequality 1 < Ix - 2 I 5 7 is equivalent to

(A) x < l or x 2 3 (B) 1 5 x 2 3 (C) - 5 5 x 5 9
(D)-5Sx<lor3Ix59
(E) - 6 5 x 5 1 o r 3 5 x 5 1 0
Part 2
11. The value(s) of y for which the pair of equations

x2 + y2 - 16 = 0 and xz - 3y + 12 = 0
may have a real common solution, are
(A) 4only (B) -7,4 (C) 0 , 4 (D) no y (E) all y
12. The number of cubic feet in the volume of a cube is the same as the
number of square inches in its surface area. The length of the edge
expressed in feet is
(A) 6 (B) 864 (C) 1728 (I)) 6 X 1728 (E) 2304
13. Inside square ABCD (see figure) with sides of length 12 inches,
segment A E is drawn, where E is the point on DC which is
5 inches from D. The perpendicular bisector of A E is drawn and
intersects AE, AD, and BC a t points M , P, and Q respec-
tively. The ratio of segment PM to MQ is
(A) 5:12 (B) 5:13 (C) 5:19 (D) 1:4 (E) 5:21
14. A triangle has angles of 30" and 45". If the side opposite the 45"
angle has length 8, then the side opposite the 30" angle has length
(A) 4 (B) 4tlz (C) 4V3 (D) 4 4 (E) 6
15. A contractor estimated that one of his two bricklayers would take
9 hours to build a certain wall and the other 10 hours. However, he
knew from experience that when they worked together,. their com-
bined output fell by 10 bricks per hour. Being in a hurry, he put
both men on the job and found that it took exactly 5 hours to build
the wall. The number of bricks in the wall was
(A) 500 (B) 550 (C) 900 (D) 950 (E) 960
16. There are two positive numbers that may be inserted between 3 and
9 such that the first three are in geometric progression while the
last three are in arithmetic progression. The sum of those two posi-
tive numbers is
(A) 134 (B) I l t (C) 104 (D) 10 (El 9b
17. A piece of string is cut in two a t a point selected a t random. The
probability that the longer piece is a t least x times as large as the
shorter piece (where x 2 1) is
18. Let A BCD be a trapezoid with the measure of base AB twice that
of base DC, and let E be the point of intersection of the diagonals.
If the measure of diagonal AC is 11, then that of segment EC is
equal to
(A) 3$ (B) 3 i (C) 4 (D)3a (E) 3
19. The sum of the first n terms of the sequence
+
1, ( 1 2), ( 1 + + + + + ... + 2"-9
2 29, * . * ( l 2 2 2
in terms of t z is
(A) 2" (B) 2" - n (C) 2"+l - n (D) 2"+l- n -2 (E) n-2"
2ab
20. If t a n x = a- -
2 b2' where a >b>0 and 0" < x < W", then
sin x is equal to
2ab
Part 3
21. If the sum of the measures in degrees of angles A , R , C, D, E ,

and F in the figure is Wn,then n is equal to
(A) 2 (B) 3 (C) 4 (D) 5 (El 6
22. If a f bi ( b # 0, i = dZ)
are imaginary roots of the equation
+ +
z' qx r = 0, where a, b, q, and I are real numbers, then q in
termsof a and b is
(A) a*+ b2 (B) 2a2- b2 (C) b 2 - a2 (D) b 2 - 2a2
(E) bZ - 3a2
23. The radius of the smallest circle containing the symmetric figure
composed of the 3 unit squares is
Sdi7
(A) ./z (B) 41.25 (C) 1.25 (D)16 (E) None of these
0
24. A man walked a certain distance a t a constant rate. If he had gone
3 mile per hour faster, he would have walked the distance in four-
fifths of the time; if he had gone 3 mile per hour slower, he would
have been 2$ hours longer on the road. The distance in miles he
walked was
23. Inscribed in a circle is a quadrilateral having sides of lengths 25, 39,

52, and 60 taken consecutively. The diameter of this circle has
length
(A) 62 (B) 63 (C) 65 (D) 66 (E) 69
26. In the circle above, M is the mid-point of arc CAB, and segment
M P is perpendicular to chord A B at P. If the measure of chord
+
AC is x and that of segment d P is ( 2 1), then segment PB
has measure equal to
( A ) 3 x + 2 ( B ) 3 x + 1 (C) 2x+ 3 ( D ) 2 x + 2 (E) 2 2 + 1
52 T H E MAA P R O B L E M B O O K I11
27. If the area of AABC is 64 square inches and the geometric mean
(mean proportional) between sides AB and AC is 12 inches, then
sin A is equal to
28. A circular disc with diameter D is placed on an 8 X 8 checker-

board with width D so that the centers coincide. The number of
checkerboard squares which are completely covered by the disc is
(A) 48 (B) 44 (C) 40 (D) 36 (E) 32
29. If f(x) = log(*)

1-x
for - 1 < x < 1, then f
termsof f(x) is
30. A rectangular piece of paper 6 inches wide is folded as in the diagram

so that one corner touches the opposite side. The length in inches
of the crease L in terms of angle B is
(A) 3sec*BcscB (B) 6sinBsecB (C) 3secBcsc8
(D) 6 sec 0 csc* 8 (E) None of these
Part 4
31. When the number 2’Oo0is divided by 13, the remainder in the divi-
sion is
(A) 1 (B) 2 (C) 3 (D) 7 (El 11
A B
32. Chords A B and CD in the circle (see figure) intersect at E and

are perpendicular to each other. If segments A E , EB, and ED
have measures 2 , 6 , and 3 respectively, then the length of the diam-
eter of the circle is
(A) 4 d J (B) d@ (C) 2 4 7 (D) 3 4 5 (E) 6v2
33. The minimum value of the quotient of a (base ten) number of three
different nonzero digits divided by the sum of its digits is
(A) 9.7 (B) 10.1 (C) 10.5 (D) 10.9 (E) 20.5
34. Three times Dick’s age plus Tom’s age equals twice Harry’s age.
Double the cube of Harry’s age is equal to three times the cube of
Dick’s age added to the cube of Tom’s age. Their respective ages
are relatively prime to each other. The sum of the squares of their
ages is
(A) 42 (B) 46 (C) 122 (D) 290 (E) 326
35. Equilateral triangle A B Y (see figure) with side A B of length 2

inches is placed inside square A X Y Z with side of length 4 inches
so that B is on side A X . The triangle is rotated clockwise about
B, then P, and so on along the sides of the square until P , A ,
and B all return to their original positi0ns.t The length of the path
in inches traversed by vertex P is equal to
(A) 2 0 ~ / 3 (B) 32r/3 (C) 12r ( D ) 40r/3 (E) 15,
t The original wording of this problem was somewhat different. The reasons for the
change are explained in the Comment, see p. 180.
Answer Keys
1966 Answers 1967 Answers
1. C 9. A 17. C 25. D 33. D 1. C 9. E 17. A 25. A 33. D
2. E 10. E 18. A 26. C 34. B 2. D 10. A 18.B 26. C 34. A
3. D 11. C 19. B 27. A 35. C 3. B 11. B 19. E 27. C 35. B
4. B 12. E 20. C 28. B 36. E 4. C 12. B 20. A 28. E 36. C
5. A 13. E 21. E 29. B 37. C 5. D 13. E 21. B 29. C 37. A
6. C 14. C 22. A 30. D 38. D 6. D 14. C 22. A 30. D 38. E
7. A 15. D 23. A 31. D 39. E 7. E 15. D 23. B 31. C 39. B
8. B 16. H 24. B 32. B 40.A 8. A 16. B 24. A 32. E 40.D

1. D 8. I3 15. D 22. E 29. A 1. H 8. D 15. D 22. C 29. C
2. B 9. E 16. E 23. U 30. A 2. A 9. C 16. E 23. A 30. D
3. A 10. C 17. C 24. C 31. D 3. E 10. C 17. D 24. E 31. D
4. C 11. B 18.D 25. C 32.C 4. E 11. B 18. B 25. D 32. C
5. A 12. C 19. E 26. E 33. A 5. R 12. A 19. B 26. B 33. A
6. E 13. B 20. A 27. B 34. R 6. C 13. B 20. C 27. E 34. B
7. E 14. E 21. D 28. D 35. D 7. A 14. A 21. E 28. E 35. B

1. E 8. B 15. E 22. A 29. D 1. B 8. B 15. B 22. A 29. E
2. A 9. C 16. C 23. D 30. E 2. D 9. D 16. A 23. A 30. D
3. C 10. D 17. E 24. B 31. B 3. E 10. E 17. E 24. D 31. A
4. B 11. E 18.A 25. E 32.C 4. A 11. D 18. D 25. D 32. A
5. D 12. C 19. C 26. B 33. A 5. C 12. B 19. C 26. B 33. B
6. B 13. D 20. A 27. A 34. C 6. E 13. E 20. E 27. E 34. B
7. E 14. A 21. B 28. A 35. D 7. C 14. C 21. C 28. C 35. C
1972 Answers
l.D 6.C 1 I . A 16.B 21.C 26.E 31.C
2. B 7. E 12.B 17. E 22.E 27. D 32.B
3.B 8 . D 13.C 18.A 23.D 28.E 33.C
4. D 9.A 14.B 19. D 24.B 29.C 34.A
5.A 10.D 15.C 20.E 25.C 30.A 35.D
55
m
Solutionst
1966 Solutions
Part I
+
1. (C) We have (3%- 4) = k(y 15), where the constant ratio
k = $ is determined by replacing ( x , y) by (2,3 ) . The rela-
tion ( 3 x - 4) = +(y +
15) now yields x = 4 when we set
y = 12.
2. (E) If b and h denote the base and altitude of the triangle, then
after the ten percent changes, the area becomes
$(l.lb)(.9h)= .99(-$bh)
which is R 1% decrease from the original area ibk.
Remark: If b is increased by c times itself (to b cb), and h is +
decreased by c times itself (to h - ch), then their product
p = bh is decreased by 8 times itself, to
p' = (1 + c)b(l - ~ ) =h (1 - c2)bk = (1 - c2)P = p - c2P.
3. (D) Let Y and s denote the two numbers with given arithmetic
and geometric means:
$(Y + s) = 6, d z = 10.
Then Y + s = 12 and IS = 100. A quadratic equation with
roots I and s is
(x - Y)(X - s ) = 22 - (Y + s ) x + YS = 0;
+
and, when the above values for I s and YS are used, the
+
equation x2 - 122 100 = 0, given in (D), is obtained.
t The letter following the problem number refers to the correct choice of the five
listed in the examination.
57
Remark: I t is clear that I and s cannot be positive real numbers;

for, if they were, the given data would contradict the famous
arithmetic-geometric mean inequality which states that, for
positive real numbers a and b,
+ab
2
2 .c/a;
the equality holds if and only if a = b.t Indeed, the required
equation in (D) has the conjugate comples roots 6 f 8i.
4. (B) The ratio of the radius of the circumscribed circle to that of the
inscribed circle is the same as the ratio of the diagonal of the
square to its side, and this is d.The ratio of the areas of the
circles is the square of the ratio of their radii; and (V2)2 = 2.
5. (A) The left member of the given equation is not defined when
x = 0, nor when x = 5; for all other x it has the constant
value 2. The right member is defined for all x and has the value
2 only when x = 5, so that no value of x satisfies the given
equation.
t For a full discussion of this inequality and its generalizations, see N M L 12, pp.
70-72.
S O L U T I O N S : .1966 E X A M I N A T I O N 59
6. ( C ) Triangle ABC is a right triangle with hypotenuse A B = 5

inches because g A C B is inscribed in a semicircle. Since radii
OC = OB = 4 inches, and angle BOC between them is 60°,
triangle BOC is equilateral, and leg BC of right triangle
ABC is 5 inches. The Pythagorean theorem now yields
7. (A) When the right side of the given identity is written as a single
fraction, the identity reads
35% - 29 -- - 2)N1+ - 1)N2
+
(X (X
xz - 3x 2 (x - 1) (x - 2)
The denominators are identical, hence the numerators are also
identical. This means
35x - 29 = (x - 2)N1+ ( x - 1)Nz for all x.
A linear function of x is completely determined by its values a t
two distinct points x1 and q which, in the above case, are
most conveniently taken to be x1 = 1 and q = 2. Substitu-
tion of these values for x yields
35 - 29 = -Nil N1 = -6,
and
70 - 29 = N2, Nz = 41,
respectively. Thus N1N2 = -246.
Remark: Since two linear functions of x are the same if and only if
the coefficients of x are equal and their constant terms are
equal, this problem can also be solved by equating these cor-
responding coefficients.This leads to two linear equations in N1
and N2 withsolution N1 = - 6 , N2 = 41.
The argument in the above solution as well as that outlined in
this remark can be generalized to polynomials of higher degrees.
8. (B) Denote the common chord by AB, its midpoint by P , and the
centers of the smaller and larger circles by 0 and 0'; 00' is
perpendicular to AB and passes through P. The Pythagorean
theorem applied to right triangles OPA and O'PA now yields
OP = OA2 -A P = 10' - 8' = 36, OP = 6,
and
PO'* = O'A' -AP = 17' - 82 = 225, PO' = 15.
If, as in the figure, 0 lies outside the larger circle, the distance
between the centers is
00’= OP+ PO‘ = 6 + 15 = 21.
(If 0 lies inside the larger circle, then 00’ = PO’ - OP =
15 - 6 = 9.)
9. (A) Since, by the definition of logarithms,

logs 2 = 4 and log2 8 = 3,
we have
so that log3x = -3.
10. (E) Let the numbers be x and y. Then since

(1) x+ y = xy = 1,
1 = ( x + y)’= $+ b + 3 x y ( x + y ) = x3-k f + 3(1)(1)
=xa+y+3.
:. xa+ r” = -2.
OR
The numbers x, y satisfy the quadratic equation
(2) u2- u + 1 = 0
since their sum is 1 and their product is 1. [Eq. (2) can also be
derived from (1) by substitution of 1 - x for y in xy = 1.1
We observe that
(3) (u+ l)(UZ- u + 1) = 3 + 1 = 0,
so that the roots of (2) also satisfy ( 3 ) ; hence they are cube
roots of -1. Thus we have x3 = - 1 and 9 = -1, whence
a?+?= -2.
SOLUTIONS: 1966 EXAMINATION 61
+
If one does not recognize u2 - u 1 as a factor of ua 1, +
one can compute the roots x = i(1 +
Gi),y = 4(1 - fli)
explicitly, cube each, and add the result.
11. (C) Since an angle bisector of a triangle divides the opposite side
into segments proportional to the adjacent sides, we have the
shortest side AC of length 10 divided by D in the ratio 4:3.
Thus the longer segment is of the length of A C ; that is,
t .10 = 5# is the length of the longer segment.
12. (E) In terms of only powers of 2, the given equation is equivalent to
(26.'+3) (22(32+6)) 28('"+5) or 21&+15 = 21W16,
which is true for all real values of x.
13. (E) Between every pair of distinct real numbers, there are infinitely
many rational numbers. In particular, between 0 and 5 there are
infinitely many rational numbers x such that 0 < x < 5.
Pick any such x and set y = 5 x ; then 0 < y < 5. -
+
Moreover, y is rational, and x y = 5 (so that in particular
x + y L 5).
14. (C) Triangles AEB, BEF, and FCB have equal areas because
they have the same altitude from B and equal bases. Hence
each has area one-third the area of triangle ABC, that is
t ( j . 5 . 3 ) = 4 squareinches.
15. (D) Since x - y > x, -y > 0 and y < 0.

Since x + y < y, x < 0. :. x < 0, y < 0.
16. (B) In terms of powers of 2, the first equation gives
-2%- - 2-
2*+v
= 2'. ... x - y = 3.
In terms of powers of 3, the second equation gives

32(rhr)
-- -
35v
32=-av = 35. .: 2x - 3y = 5.
The equations x - y = 3 and 2x - 3y =5 now yield

y = 1, x = 4, so that xy = 4.
Y
t
17. (C) Both curves are ellipses with centers at the origin and ases along
the coordinate axes. In standard form, their equations are
x2 y2 9 Y2
- - k 2 = 1 and -+-=1,
l2 (1/2) 12 22
from which we see that the first has, as its major asis, the seg-
ment from ( - 1,O) to (1, 0 ) , while the second has that same
segment for its minor axis, so that all other points of the second
ellipse lie outside the first. We conclude that the only points
common to both curves are the points ( - 1 , O ) and ( 1 , O )
where the ellipses are tangent.
OR
The coordinates ( x , y ) of a point common to both curves
satisfy both equations, hence also their sum 5x2 5y2 = 5, +
which describes the unit circle with center at the origin. The
only points on both ellipses and on this circle are the points
(1, 0) and (-1, 0) where both ellipses are tangent to the
circle.
SOLUTIONS: 1966 E X A M I N A T I O N 63
18. (A) Formulasfor the n-th term 1 and thesum s of n termsof an

A.P. with first term a and common difference d are (see
p. 114)
1 = a + (n - 1)d and s = i n ( a + 1 ) .
+
The second formula gives 155 = fn(2 29), n = 10, and the
first now yields 29 = 2 +
(10 - l)d, d = 3.
19. (B) The formula for the sum s of the first n terms of an A.P.
n
may be written as s = - [2a + -
(n l)d], see p. 114. Hence
2
s1 =
n
- C2.8
2
+ ( n - 1)4] =
n
2 [12 4- 4n)
s2 =
n
- C2.17
2
+ ( n - 1)2) n
- [32
2
+ 2n]
and, for n # 0, s1 = s? if and only if the expressionsin brackets
+
are equal, i.e., if and only if 12 4% = 32 2n or n = 10. +
20. (C) Let P ( a , b) be a proposition concerning a and b. Then the
the negation of the statement “For all a and b, P ( a , b) holds”
is “there exist a and b such that P ( a , b) does not hold.”
In the present case P ( a , b) is the statement: if a = 0 and b is
any real number, then ab = 0. This is equivalent to saying
that “either a # 0 or ab = 0.” Now the denial of a statement
of the form “either S or T” is “not S and not T.” Hence
the denial of the statement P ( a , b) is “a = 0 and ab # 0.”
Thus, the negation of the statement “For all a and b, P ( a , b)
holds” is “there exist a and b such that a = 0 and ab # 0.”
Part 2
I I
I /
I I
I I
21. (E) Denote the angles of the star in the figure by al, 02, as, - ., a,,
and the angles of the small convex polygon by al,az, a), - ..,
a. with a,+l = a1. Then
= 180 - (180 - a1) - (180 - a2) = al+ a2 - 180,

a~ 180 - (180 - a2) - (180 - as) + aa - 180,
... ... ... ... ... ...
=
...
a, = 180 - (180 - CX,) - (180 - C X , + ~ ) = an+ - 180.
Summing the angles on the left and those on the far right, we get
S = 2(q+ as+ .*. + a,) - 180s.
Since the sum of the n interior angles of a convex n-sided
polygon is 180(n - 2),
S = 2.180(n - 2) - 180n = 180(n - 4) degrees.
22. (A) Each of the four equations has an infinite number of solutions
(a, b) # (0,O). For example, if a is any number greater than
1, then I, 11,111and I V are satisfied when b is equal to a d q ,
a / d d - 1, 0, and 0, respectively. There are many other solu-
tions of each equation.
23. (A) We may treat the equation as a quadratic in y with discrim-

inant
D +
= (4%)'- 4 . 4 ( ~ 6) = 16(x2 - x - 6)
-
= 16(~ 3 ) ( ~ +2).
Now y is real if and only if D 2 0, and this is true when both
+
factors on the right, x - 3 and x 2, have like signs, i.e.,
when x 5 -2 or x 2 3. Alternatively, the equation may be
written
4y+4xy+xz- (9-Z-6) = O
or
(2y + X)Z = (2 + 2) ( 2 - 3).
For real x, the left member of the last equation is non-negative
+
if and only if y is real, so the product (x 2) ( x - 3) 2 0,
which is true when x I -2 or 2 2 3.
24. (B) The identity (logN M )(log,+, N ) = 1 together with the given
equation yields (logN M)*= 1. :.
logN M = 1 or - 1.
If logN M = 1, then M = N which is ruled out. We conclude
that logN M = -1. :. M = N+, M N = 1.
OR
If logN M = x = 1ogM N, then M = Nz, and N = MI =
(N*)* = N2. Since N # 1, we conclude x2 = 1, so x = 1
or x = -1. We reject z = 1 (since it leads to M = N) and
conclude x = - 1, so M N = 1.
+ +
25. (D) Since F(n 1) = F ( n ) 4, the sequence F ( n ) is an
arithmetic progression with first term F(1) = 2 and common
difference 4. The lOlst term is
F(101) = 2 + (101 - 1); = 2 + 50 = 52.
26. (C) Substituting y from the second equation into the first gives
+
13% 11(mx - 1) = 700, so that
x =
711 - 3'- 79
-
13+ l l m 13+ l l m '
+
Since x is to be an integer, the denominator 13 l l m must
be a divisor of the numerator, and its only divisors are 1, 3, 32,
i9, 3.79, 3'.79. Our task now is to find a positive integer m
such that
- 13
CI!
13+ l l m = d, or m = -
11 '
where d is one of these divisors. Since m > 0, we see that
d > 13, so the only divisors we need to test are the last three:
(i) if d = 79, d - 13 = 66, and m = f-f = 6
(ii) if d = 3.79 = 237, d - 13 = 224 is not divisible by 11
(iii) if d = 32-79 = 711, d - 13 = 698 is not divisible by 11.
We conclude that m = 6 is the only positive integer yielding a
lattice point for the intersection of the given lines.
27. (A) Let c denote the speed of the current and m the usual rate of
the rower in still water. Then his downstream and upstream
+
rates are m c and m - c, respectively, and after he doubles
+
his usual rate, they are 2m c and 2m - G. Since the distance
is 15 miles and the time is distance/rate, the problem tells us
that
- 15
- ---15 5 and -=15 --15 1.
m+c m-c 2m+c 2m-c
+
Multiplying the first equation by ( m c) (m - c) and the
+
second by (2m G) (2m - c ) and simplifying yields
5m2- 5c2 = 30c and 4m2 - c2 = 3Oc,
and subtracting the second from the first gives
m2 - 4c2 = 0, so m = 2c.
If we substitute for m, in the second equation above, we get
4(4~*)- c2 = 3oG SO 1% = 30, c =; 2.
28. (B) The proportion A P : P D = BP:PC may be written as

(p-a):(d-pP) = (P-b):(c-p),
where p denotes the distance OP.
-.. ( - a + p ) ( - p + C) = ( - b + p ) ( - p + d ) ,
--a~ + + (U c ) P - p2 = -p2+ +( b d ) p - bd,
[(a + C) - ( b + d ) Jp = - a -
~ bd,
a~ - bd
'=a-b+c-d = OP.
29. (B) The number of positive integers less than some positive integer
M is M - 1 ; the number of positive integers less than M
and divisible by d is the greatest integer not exceeding
( M - l ) / d . (We denote the greatest integer not exceeding the
number x by the symbol [ x ] . ) Thus, among the 999 positive
integers less than 1O00, there are N1 = [999/5] which are
divisible by 5 and N2 = [999/7] which are divisible by 7. But
some of the numbers divisible by 5 are also divisible by 7.
Therefore, if we subtract from the 999 numbers less than lo00
all the numbers divisible by 5 , and subsequently all the numbers
divisible by 7, we would be subtracting those divisible by 5 and 7
(i.e., by 35) twice, once in each batch. Therefore, the desired
answer is
999 999 999
+
999 - - - - - = 999 - 199 - 142 28
5 7 35
+
= 686.
30. (D) Since the sum of the roots is zero, the 4th root is -6 and the
equation is ( x - 2) ( x - 3) ( x - 1 ) ( x 6 ) = 0. +
... (x' - 5 x + 6 ) ( x 2 + 5%- 6 ) = 9 - (5%- 6)2
= X' - 2 5 9 +
6 0-~ 36 = 0.
S O L U T I O N S : 1966 E X A M I N A T I O N 67
... a + c = -25 - 36 = -61.

OR
We note that a is the sum of the products of the roots taken two
at a time, and c is the product of the four roots. Hence
u = 1.2 + 1.3 + 1 * ( - 6 ) + 2.3 + 2*(-6) + 3.(-6)
= -25
and c = 1.2.3.(-6) = -36. :. a + G = -61.
OR
Substituting the three roots into the given equation gives three
equations in the three unknowns a, b, and c; viz.
l+a+b+c=O
16+ 4a+ 2 b + c = 0
81+8a+3b+c=O
with solution ( a , b, c) = (-25, 60,-36) from which
u + c = -25 - 36 = -61.
Part 3
A B
31. ( D ) Since AB and AC are tangent to the small circle and AD

passes through its center, we have Q C A D = QBAD = a (see
figure). Similarly, QACO = QBCO = 8. Therefore arcs
CD and BD are equal and hence also the chords: CD = BD.
Next we show that CD = OD by proving that these are sides

opposite equal angles in ACDO, i.e. by proving QOCD =
QCOD. Now QOCD = QOCB QBCD = B -I-a, and+
QCOD is an exterior angle of AAOC, hence equal to the sum
01+ /3 of the remote interior angles. We conclude that CD =
OD, and that (D) is the correct answer.
32. (B) Denote the area of a triangle XYZ by (XYZ). Since

PC 11 MD, (MDC)= (MDP).It follows that
(BPD) = (BMD) + ( M D P ) = (BMD) + (MDC)
= (BMC) = I ( A B C ) ,
since CM is a median. Thus
I=-=-
(BPD) 1
(ABC) 2
Remark: If P lies to the left of A , the same proof works; but if P
lies between M and B , PC is the common base of triangles
PCM and PCD of equal areas which, when added to A P C B ,
yield (BPD) = (BMC) = I ( A B C ) . What if P lies to the
right of B? Can you devise a single proof valid for any position
of P on the line through A and B, perhaps with the help of
signed areas?
33. (D) If we write each side of the given equation as a single fraction,
we obtain the equivalent equation
a(x - a) + b ( x - b) -- b ( x - b ) + a ( x - a )
ab ( x - a ) ( x - a) *
We observe that the numerators are identical. Thus the equation

is satisfied either (i) if the denominators are equal, or (ii) if the
numerator is zero.
Case (i) requires
(x-a)(x-b) =x9- (a+b)x+ab=d,
x[x +
- ( a b ) ] = 0, and this is true if x = 0 or x =
+ b.
SO
a Case (ii) requires that
a(x - a) + b ( x - b ) + b ) x - (a2+ b2) = 0,
= (a
and this is true if x = (a* + b 2 ) / ( a+ b) . The given conditions

on a and b ensure that the three values we have found are
indeed distinct.
distance (in miles)

34. (B) We are told that I = speed (in m/hr) =
time (in hours) ’
Then
5280 22 distance (in feet)
y.- = - y =
3600 15 time (in seconds) ’
and
time (in seconds) =

distance (in feet) 15
.-
I 22
11.15 15
Time for one rotation = -= - = 1 seconds. When r is
22r 2r
increased by 5, 1 is decreased by +,so
15 -- 1 - - 1
2 ( r + 5) 4‘
It follows that
15 15 1 30 -r
+ 5) = -2r- - =4 - 4r ’
2(r
so that 3Or = ( r + 5)(30 - r ) = 30r - + 150 - 5r, or
rz + 5r - 150 = 0, - 10) (I + 15) = 0. We reject the nega-
r2
(I
tive speed and conclude that I = 10.
35. (C) The first niember of each of the following inequalities compares
the base of a triangle with the sum of the other two sides. The
second member (which we prove below) compares that sum with
the sum of the sides of another triangle having the same base
but containing the first triangle, see figure on next page. Thus
AB < O A + O B < A C + CB
BC < OB + OC < B A 4- AC
CA < OC+ OA < C B + B A .
70 T H E M A A P R O B L E M B O O K 111
Adding these member by member, we get

~2 < 2si < 252, < si < sz
as required for every triangle in choice (C) .
To see that OA + OB < A C + CB, extend segment A0
to meet side BC in point D. By the triangle inequality,
A D = AO+OD<AC+CD,
OB < O D + DB,
and addition yields
AO+OD+OB< AC+CD+OD+DB,
so
AO+OB<AC+CD+DB= AC+CB.
36. (E) Letting first 2 = 1 and then x = - 1 in the given identity,

we obtain
(1 + 1 + 1,)" = 3" = a ~ + + + + a1 Q &.-I+ an
[1+ (-1) + (-l)*-J = 1"

=r &-al+&- ... - en-1 + en.
Adding these, we get
3"+ 1 = 2aQ+ 2 e + *'* + 2a2,
= 2(aO+ e+ ..* + em) = 2s,
3"+ 1
s=-
2 .
Remark: When n = 1, we get J = U.Z = 1 &I+ 1 = 2. This+
eliminates choices (B) , (C) ,and (D) . In a similar way, one can
eliminate choice (A) by taking tt = 2.
37. (C) Let a, b and c denote the number of hours it takes Alpha,
Beta, and Gamma, respectively, to complete the job alone. Then
the fraction of the job each completes in one hour is l/a, l/b,
l/c, respectively, and the fraction of the job all three together
complete in one hour is l / a l / b + +

l/c. Moreover, needing
“6 hours less time than Alpha alone,” “one hour less than Beta”
and “half the time needed by Gamma” leads to hourly rates of
--1 1 1 2
and - = - , respectively. So the given in-
a-6’ b-1’ c/2 t
formation may be written in the form
-1 + -1 + -1 = -
1
= -
1 2
=-
a b t a-6 b-1 c’
and
2 h + 12 1 2
, from -= -
1
From -
a-6
- -h
- we obtain a = - 2
*
b-1 h’
h+ 2
we get b = -. The sum of the reciprocals of a and b
2
2
is the reciprocal of h, so that - + - =2 - 1
. Clearing
h+12 h+2 h
fractions and simplifying leads to
3h2 + 14h - 24 = 0 or (3h - 4 ) ( h + 6 ) = 0.
The negative value of h is ruled out, so h = $.
38. (D) The base OQ of AOMQ is equal to

OQ = CO - CQ = $CN - i C N = $CN.
Let h be the altitude of AOMQ from M to side OQ. then
2h is the altitude from B of A C N B . Thus
Areaof AOMQ = 4OQ-h = A C N - h = It.
Area of A A B C = 2 (Area of A C N B )
= 2(iCN.2h) = 2CN.h = 24n.
39. (E) We write each fraction as an infinite series using l/R1 then
1/Rz as common ratio; we sum the infinite series and equate
corresponding expressions for FI and F2:
3R14-7
FI = -+ 3R1C
-+ 7 ... = R12
-
3R1+ 7
-1
RIZ Rt
2Rz + 5
--2RZ+S
-
Rz2
2R2+5
+-+
RZ‘
... = Rz2
-*
-1
3R1+ 7 2Rz+ S
.: F1 = -
R? - 1
-
R22 - 1 ’
--*
similarly,
Now
*’
RI-
--
10
1 --Rz- 1
7 ’
7R1 - lOR2 + 3 = 0.
The two linear equationsin R1 and R2 have the unique solution

RI = 11, RZ = 8, whence R1 Rz = 19.t +
t Editor’s note: The reduction to a linear system of equations was due to the fact that
FI and FS have the same digits in reverse order. For arbitrary F1 and Ft with
period 2, we must expect a quadratic system of equations.
40. (A) Let perpendiculars from points E and D meet diameter A B

at M and N respectively. Since transversal AC intercepts
equal segments (AE and DC) between parallel lines, trans-
versal AB does likewise, so NB = x . The altitude N D to the
hypotenuse of right triangle ABD is the mean proportional
between the segments A N = (2a - x ) and NB = x, i.e.
NZP = x(2a - x ) . Similar right triangles AME and AND
give
... yy2axz- r)* = (2a- X).P or

9
3=-
2a - x
1967 Solutions
Part 1
1. (C) Since 5b9 is divisible by 9, and 0 5b5 9,

Sb9
_ -- 10-(50 b ) + I +
9 9
+
is an integer; so (SO b ) / 9 must also be an integer. .-.b = 4.
Now
2a3 = 5b9 - 326 = 549 - 326 = 223.
Therefore a = 2 , and a + b = 2 + 4 = 6.
OR
+ +
One may note directly that the sum (5 b 9 ) of the digits
in Sb9 must be a multiple of 9 . (A number is divisible by 9 if
and only if the sum of its digits is divisible by 9.) So the digit
b is 4. We have then, as before, a = 2 and a b = 6. +
OR
Using congruences, 5b9 is congruent to 0 modulo 9,
S O O + lob+ 9= 5 + b = Omod9, b = -5 = 4mod 9.
:.b =4 because 0 5 b 5 9. Then a = 2, a + b = 6 fol-
low as in the other solutions.
2. (D) The given expression is equivalent to
(.+Y(): + ); + (x - k)(Y - );
(
= xy+-+-+-
yX Yx Zl Y)
+( xy----+-
x y x y 7
= 2xy + XY-2 .
OR
Performing the multiplications in each of its two terms, the
given expression can be written as
= 2xy + -2
ZY
as before.
3. (B) The altitude h of the given triangle is &/2, and the radius
r of the inscribed circle is
h
y = - = -
sd3
3 6 '
The diagonal of the inscribed square is the diameter 2r =
sV33/3 of the circle, and the area of the square is half the prod-
uct of its diagonals:
(2Y)* s23 s2
area of square = --
- - = -
2 2.9 6'
OR
We could have found the side of the square first by dividing
its diagonal by a, then computed the area of the square by
squaring its side:
4. (C) The first three given logarithmic equalities are equivalent to

the exponential equalities a = x p , b = xq, c = xr whence
Alternatively, we may express the relation involving y in

logarithmic form :
y log x = 2 log b - log u - log C.
Substituting for the logarithms on the right from the first
given relations yields
y log x = 2q log x - p log x - I log x ,
and, since x # 1, log x # 0, so division by log x yields
y=2q-p--.
5 . (D) Denote the sides of the circumscribed triangle by a, b, and c

(see figure). Then the radii to the points of contact are
perpendicular to the sides and hence are altitudes of the three
triangles into which the circumscribed triangle is partitioned.
Its area is therefore
K = (ar + br + cr)/2 = ( a + b + c)r/2 = Pr/2
so that P j K = 2/r.
Remark: Clearly P is proportional to r , and K is proportional

to r2, so P I K is proportional to ljr. This eliminates choices
(A), (0,and (E).
6. (D) From f(x) = 4=, we obtain

f(x + 1) - f(z) = 4=+1 - 4'
= 4.4' - 4" = (4 - 1)4" = 3.4' = 3f(x).
7. (E) When b, (-c), and d are all positive, the given inequality
is equivalent to a < - bc/d, which means that (I is less than
the positive number -bc/d. This will be true when a is
positive but less than -bc/d, or when a = 0, or when a is
negative.
S 0 I, U TI 0N S : 196 7 E X A M I N A T I 0N 77
8. (A) The final mixture of m +

x ounces contains (m/loo)m ounces
of acid, and this is to be ( m - 10)/100 of the mixture. So
m - 10
100
(m +x) =
m2
-
100 *
When solved for 2, this yields
10m
x=-
m - 10’
[If m were less than 10 ounces, water would have to be ex-
tracted from, rather than added to, the initial mixture of m
ounces.]
9. (E) Let the shorter base, altitude, and longer base be denoted by
+
(a - d ) , a, and ( a d ) , respectively. Then the area is
K = fa(a - (1 + a + d) = a’.
Since we have no knowledge about the nature of the number a,
we cannot deduce properties of a*; so (E) is the correct
answer.
10. (A) Multiplying both members of the given identity by the posi-
tive number (lo” - 1 ) (lo” 2) gives +
a(1o” + 2) + b(1o” - 1) = 2.101: + 3.
Equating the constant terms and the coefficients of lo” in the
two members of this identity yields 2a - b = 3 and a b = +
2, respectively. Solving this system of linear equations in a
and b gives 3a = 5, a = 4, so that ++
b = 2, b = 4.
Hence a - b = $ as stated in choice (A).
11. (B) If we denote the length of side A B by x , then the length of

the adjacent side BC is (10 - x ) . Sides A B and BC of
rectangle ABCD are legs of the right triangle ABC whose
hypotenuse is the diagonal AC. The Pythagorean Theorem
now yields
AC2 = AB’ + BC? = + (10 - x)’
X’
= 2(x*- 1Ox+50) = 2 [ ( x - 5 ) ’ + 2 5 ) ,
which takes on its least value, 50, when ( x - 5)’ = 0, i.e.,
x = A B = 5 and (10 - x ) = BC = 5. Thus the rectangle
of perimeter 20 with least diagonal is a square, and the length
of diagonal AC is 4%= 5d inches.
Remark: Denote the lengths of A B , BC and AC by the positive

numbers a, b, and c respectively. Then the perimeter p =
+
2(a b), and minimizing G is equivalent to minimizing
c2 = a2 + b2 = (a + b)2 - 2ab =
P2
- - 20b.
4
If p is fixed, c2 is smallest when ab is largest. We apply the
arithmetic-geometric mean inequality
where = holds if and only if a = b; since the left member is

constant, da (and hence ab) is largest when a = b. Thus
the rectangle is a square, and this argument holds for any given
perimeter.
12. (B) The area described (left figure) is that of a trapezoid with
+
bases of lengths (m 4) along x = 1 and (4m 4) along +
x = 4, and altitude of length 3 along the x-axis. So
1 - 10 2
-(3)(m+4+4m+4) = 7, 5m = - m= --
3'
2 3 '
[The convexity of the region rules out the possibility of
the line y = mx +
4 crossing the x-axis in the interval
1 < x < 4.1
13. (E) We begin the construction with the given side a and denote
its endpoints by B and C (right figure). At B we construct
a line 1 making the given angle B with side a. Now either
(i) the distance from C to 1 is h,, in which case vertex A
of A A B C may be placed anywhere on line 1 (infinitely
many solutions), or
(ii) the distance from C to 1 is not h,, in which cash no

triangle satisfies the given conditions (no solution).
14. (C)Since f ( t ) = f/(1 - t ) , t # 1, y = f ( x ) is equivalent to

y = x / ( 1 - x ) which may be solved for x :
x = - = -Y (-Y)
= -f(-Y>*
1+Y 1 - (-Y)
+
15. (D)Let T (T 18) denote the area in square feet of the smaller
(larger) triangle, and 3 and x denote the corresponding sides
in feet. Since the areas of similar triangles are proportional to
the squares of corresponding sides,
By assumption, x / 3 is an integer, so x is a multiple of 3.

Solving this equation for T yields
18
T=
(x/3)2- 1 ’
and since T is required to be an integer, ( ~ / 3 - ) ~1 is a
) ~2, 3, 4, 7, 10, or 19. The only
divisor of 18. Thus ( ~ / 3 =
square among these numbers is 4;hence ( ~ / 3 = ) ~4, 4 3 = 2,
and x = 6.
16. (B) The given equality (12)(15) (16) = 3146 in base b means
(b + 2)(b+ 5) ( b + 6) = 3V + b2 + 4b + 6.
After some simplification, we obtain the equivalent equation
ba - 6b2- 24b - 27 = 0.
Its only real solution is b = 9. The sum s = 12 + 15 + 16
in base b means
s = (b + 2) + (b+ 5) + ( b + 6)
= 3 b + 13 = 3 b + 6 + 4 = 4b+ 4
which, in base b = 9, is written 44.
17. (A) Since the roots of the given quadratic equation are real and
distinct, its discriminant (p2 - 32) is positive. Hence p2 > 32,
I p I > 4V2. But the sum of the roots is r1+ r2 = - p , and
+
hence I rl r2 I = 1 - p 1 = 1 p I > 4V2.
18. (B) Set x2 - 5x + 6 = F ( x ) ; its factored form is F ( x ) =

(x - 2) ( x - 3), and the condition that F ( x ) < 0 implies
that the two factors have unlike signs, which occurs only when
2 < x < 3 (see figure). The second given function
+ +
P ( x ) = x2 5 x 6 increases as x increases from 2 to 3. Its
+ +
least value is P(2) = 4 10 6 = 20, its greatest value
+ +
is P(3) = 9 15 6 = 30, and it takes on all values be-
tween 20 and 30 as x varies from 2 to 3.
19. (E) Denote the length and width in inches of the rectangle by I
and w respectively; then its area Zw satisfies
lw= ( l + $ ) ( w - j ) , and l w = ( l - $ ) ( w + $ ) .
After computing the products on the right and simplifying
each equation, we are led to the linear system
-41 -l-j w = 9 and +I - $w = 9
whose unique solution is 1 = 15/2, w = 8/3, so lw = 20.
20. (A) The side sk of each square is l / f l times that of the preceding
square: Sk = (l/fl)sk-ll s1 = m (see figure). The radius r k
of each circle is 1/2 times the side of its circumscribed square:
The area of the kth circle is A I = r(fni)’(+)”’.The re-
c>’r+;c>’.+t(;>’.+
quired sum is
= ...+2-I(z> .
S,= A l + A 2 + * * * + A ,
1 m 2
= -mzr
4[ l + , + 1 *“ + &] = $.2[1- (;)’I.
As n grows beyond all bounds, (+)n approaches 0, so that
the required sum approaches mzr/2.
Part 2
B Q
21. (B) Leg BC = 4 of the 3, 4, 5 right triangle ABC is divided by

the bisector of angle A at A1 into segments A I B and AIC
proportional to the adjacent sides A B and AC (see figure) :
AlB
-I-.--=-*
AiB 5
A1C 4-AlB 3’
so
AlB = 4 = PQ and A E = 3 = PR
a2 T H E M A A P R O B L E M BOOK I11
are the hypotenuse and leg, respectively, of the second right

APQR. Its third side is RQ = = 2. Each side of n P Q R
is one-half the corresponding side of the first right triangle
d q
ABC. Also bisector
PP1= j
1 A A 1 = -1 d A C 2 + CA12 = A = 3-. 4
2 2 4
22. (A) We are given that P = Q D + R and that Q = Q'D'+ R'.

Therefore, by substitution,
P = (Q'D' + R')D + R = Q'(DD') + ( R + H'D).
To see that R + R'D is the remainder in this division by
+
DD', we must show that R R'D is less than the divisor
DD'; we know that R 5 D - 1, and R' 5 D' - 1 so that
R + R'D 5 ( D - 1) + ( D ' - l)D = DD' - 1 < DD'.
23. (B) For real x > 1, 6%- 5 and 2x +1 are positive, so that
both logarithms are defined. Now
logs ( 6 -~ 5) - log3 ( +
2 ~ 1)
6~ - 5 6~+3-8
= logs- = log*
2x+ 1 2x+ 1
approaches log13 = 1 as x increases beyond all bounds,
+
because 8/ (2x 1) then approaches 0.
24. (A) The given equation is equivalent to y = 3(167 - x ) / 5 . For

y to be a positive integer, (167 - x ) must be a positive
multiple of 5; this is the case for the 33 positive integers x =
+
5k 2, k = 0, 1, 2, 32. * a * ,
OR
We use the theorem stating that ( x , y) = ( x o - bt, yo at) +
gives all solutions in integers of the equation ax by = t if +
a and b are relatively prime integers and (20, yo) is any
particular solution, the different integers t giving the different
so1utions.t The theorem applies to the present equation
3x + 5y = 501 with a particular solution ( X O , yo) = (167, 0)
and all solutions are given by ( x , y) = ( 167 - 51,O 31). The +
integers t = 1, 2, 3, 33 give all 33 solutions in which
. a * ,
both x and y are positive, and only those solutions.
t For a proof of this theorem, see Continued Fractions by C. D.Olds, Vol. 9 in this
NML series, pp. 44-45.
S 0 L U T I 0 N S : 1 9 6 7 E X .4 M I N A T I 0 N 83
25. ( A ) Since p is odd and p > 1, p - 1 is a positive even integer,

say p - 1 = 272; hence + ( p - 1) = n is a positive integer,
and
( p - l)(P-l)/Z - 1
= (2%)"- 1
= [(2n) - 1][(212)J4 + (2n)"-2 + + 2n + 13
.*.
always has the factor

212- 1 = ( p - 1 ) - 1 = p - 2
as stated in choice (A).
I t is easy to see that none of the other alternatives is valid;
the odd integer p = 5, for instance, constitutes a counter-
example to choices (B) , (C) , (D) and (E) .
26. (C) From the given information, we have

103 = 10oO < 1024 = 21° and 21a = 8192 < 10,OOO = lW.
Taking common logarithms, we get
3 < 1Olog2, or log2 > A,
and
13 log 2 < 4, < ;$.
or log 2
Thus log 2 lies in the interval (A,A), which
is contained
in those described by (A) and (B), so (C) is a stronger con-
clusion than (A) or (B). To eliminate (D) and (E) , we note
that the inequality log 2 < a,
equivalent to 2lZ < lW,
does not follow from the given tabular information.
27. (C) Let the length of a candle be chosen as the unit of length. Let
t represent the number of hours before 4 P.M. needed to pro-
duce the desired result. In one hour, the faster burning candle
shortens by 4, the slower by f its length, and in t hours, they
shorten by 113 and 114, so their lengths are 1 - t/3 and
1 - t / 4 , respectively. Then ( I - t / 4 ) = 2(1 - t / 3 ) , t = 29
hours before 4 P.M. The time for the candles to be lighted is
therefore 4 - 23 = 1Q hours after noon or 1:36 P.M.
28. (E) Denote the set of Mems, Ens and Vees by M , N and Y,
respectively. Hypotheses I and I1 tell us only that at least one
member of M is not in N, and that N and V are disjoint
(that is, have no common members). Suppose sets M and V
were identical ; then statements (A), (B) , (C) would be false.
Now suppose sets M and V were disjoint; then statement

(D) would be false. Therefore none of the statements (A) , (B),
(C), (D) can be deduced from the given hypotheses.
29. ( C ) Denote the diameter of the circle by d, and the intersection of

perpendicular lines AC and BD by P (see figure). Since
tangents A D and BC are parallel, QADB = QPBC = a.
The complements of these angles, QABD and QBCA are
also equal, so right triangles ABD and BCA are similar. The
proportionality of corresponding sides then yields
d- = -b
u d
whence iP = ub, d = dZ5.
OR
After noting that angles D and C are complementary, and
that the product of tangents of complementary angles is 1, we
may obtain the result from the definitions of the tangents;
d d
(tanC)(tanD) =
ba
-0- = 1, d2 = ub, d = l/a.
30. (D) The dealer paid d dollars for n radios, so the cost for each
+
radio was d / n dollars. Of these, n - 2 were sold for d / n 8
dollars, and 2 were sold for $d/n, so that the total intake was
(n - 2) (: + 8 ) + 2 -
d
2n
=d + 72,
that is, 72 dollars more than the cost. This equation reduces to
- Iln = n(n - 11) = d/8,
where d and n are positive integers.

For n 5 11, we get a non-positive d.
For )t = 12, we get 12.1 = d/8, d = 96; similarly, every
integer greater than 11 yields a positive integer d. Thus 12 is
the smallest possible value of n for the given information.
Note. Actually, we need not assume that d is an integer; this fol-
lows automatically from the equation
d = 8(n2 - llfi).
Part 3
31. (C) Since a and b are consecutive integers, one of them is even,
the other odd; hence their product is even. We may let b =
a + + +
1. Then c = ab = a(a 1) = a2 a is an even integer
and
D = a2 + b2 + c2 = u2 + ( a + 1)' + a2(a+ 1)2
= a4 + 2u3 + 3a2+ 2a + 1 = (a' + a + 1)2
is the square of the positive odd integer a2 + u + 1 = c + 1.

We conclude that dD = a' + a + 1 is always an odd posi-
tive integer.
32. (E) Let F denote the foot of the perpendicular from A to diagonal
DB extended, and denote BF and FA by x and y respec-
tively (see figure). Then x2 +
9 = 6* and ( x 4)2 9 + +
= 8*. Subtracting the first of these equations from the sec-
ond yields
8~ + 16 = 28, x = 4.
We put x = 3 into the first equation and solve for y:
(3)'i- 'y = 6', 9 = W.
Therefore
AD" = (10 + x)' + y = (p)'+9= 9 = 166,
AD= 4166.
OR
The law of cosines, applied to triangle AOB yields
cos <AOB =
OA' f OB2 - AB2 - 82 42- 6' -- _+
11
2(OA1(OB) 2.8.4 16.
Since angles AOB and AOD are supplementary, cos 9:AOD =
-cos Q AOB, and the Law of Cosines applied to triangle AOD
yields
AD2 = OA2 + OD2- 2 ( 0 A ) (OD) cos XAOD
= 8'+ 6* - 2 - 8 - 6 ( - % ) = 166, AD = 4166.
OR
The converse of the theorem: If two chords of a circle intersect,
the product of [he segments of ovte is equal f o fhe protiuct of the
segments of the other asserts: If line segments AC and BD
intersecf in a point 0 such that A 0 . N = BO-OD, then poinls
A , B , C , D lie o n a circle, and is not hard to prove by means
of the pairs AOB, DOC and BOC, AOD of similar triangles.
In our case, the first pair yields CD/6 = 3/4 so that CD =
9 / 2 ; the second pair yields AD/BC = 8/4 = 2 so that
BC = AD/2. Since A , B , C , D lie on a circle, we may now
use Ptolemy's theoremt which states: If a quadrilateral is in-
scribed in a circle, the sum of fhe products of two pairs of opposife
sides is equal to the product of the diagonals. This yields
AD.BC+ AB-CD = 10.11 = 110,
AD.-
AD
2
9 1
4- 6.- = - AD"
2 2
+ 27 = 110,
AD" = 166, AD = 4166.
t For a proof and a discussion of Ptolemy's theorem and its converse and conse-
quences, see Ceorriclry Rcvisikd by H. S. M. Coxeter and S. L. Greitzer, vol. 19 of this
NML series, Random House (1967).
33. (D) Let A , , A2, A ) denote the areas of the semicircles with
diameters A B = 111, AC = (12, and C B = &, respectively;
let S be the area of the shaded region, and G that of a circle
with radius C D = I , i.e. C = ur2. Then
A
= - [(d?
8
+ d 3 ) 2- 1i3 - d3'3 (since (12 + d3 = (11)
Now C D = Y is the altitude of right triangle ADB, hence

the mean proportional between 1i2 and 113: r2 = d&. I t
follows that
c = AY2 = A l i 2 d 3 = 4 s ,
and
s 1
_ -- .-.
G' 4 '
Remark: Since the problem does not specify the precise location of
the point C on A B , it seems safe to assume that the desired
ratio is independent of the position of C. Our calculations are
siniplificd by assuming that C coincides with 0, in which
case C'D = Or1 = d 1 / 2 and
At112 All,? s 1
whence - -
G=-=
2
-=
4
4S,
c - 4'
34. ( A ) Let a, b, G denote the lengths of the sides opposite the ver-
tices A , B, C , and h,, hb, h, those of the altitudes from
A , B, C of A A B C (figureonp.88).Let K , KO, K A , K B , K c
denote the areas of triangles ABC, DEF, ADF, BED, CFE,
+
respectively. The last three have bases c / ( n 1) , a / ( n I ) , +
b / ( n 4- 1) ; their altitudes I,, I,, l b to those bases are parallel
to h,, h,, h b respectively, and l,Jh<, = = 1Jhc = n/(n+ 1).
88 THE MAA PROBLEM BOOK 111
Now
KO = K - K A - Kg - K c
=K -
h,
----
c n ha a
---- n _ _ _b_ -
hb n
2n+ln+l 2 n+ln+l 2 n+ln+l
ch, + 2 ah, + - bha

1
= K- [A
( n + 1)' 2 2
(n+1)'-3n n2-n+1
= K - n
( n + 1)'
3K = K
( n + I)*
=k'
(n + 1)' *
so
KO
-=
n'-n+l
K ( n + 1)' '
Remark: One could also guess the answer from the obvious fact
that, as n approaches zero or infinity, the desired ratio ap-
proaches 1. Of the five choices, only (A) has this property.
35. (B) Dividing through by 64 gives the equivalent equation

x9 - sx'+ +&- 8 = 0.
The negatives of the product and the sum of the roots are the
constant term and the coefficient of 9,respectively. Since the
three roots are in arithmetic progression, we may denote them
+
by (a - d ) , a, (a d ) ; their sum then is 3a = P, so
a = 3. Their product is a(a* - (12) = 8 = $(A- (12) so
that (12 = f , d = *+. The difference between the largest and
smallest roots is
(a+Id() - (a- I d l ) = 2 1 d 1 = 2.3 = 1.
36. (C) Denote the middle term by a and the ratio by r; then the
sum of the five terms, each an integer, is
211 =
a
-
r2
+ a- + a + a r + ar2;
I
I is rational, say I = c/d (where c, d are integers without

common divisors), for otherwise ar would not be an integer.
Since every term in
ad2 ad ac ac2
2+y+a+;+~= 211
is an integer, c2 and @ both divide a:

a = kc2d2, k an integer.
But then the left side of our equation is divisible by k while
211 is a prime, sc k = 1 and the equation reduces to
d4 + d3c + #c2 + d 8 + d = 211.
' = 256 > 211.
The integers c and d are both less than 4,since 4
Neither c nor d can be 1. For, if one of them were, the
other would satisfy
9-1
x4 + +
I3 I2+ x+ 1 = -
x- 1
= 211;
but if I = 2, the left side is 31 # 211, and if x = 3, the

left side is 121 # 211.
Since c and d have no common factor, the only remaining
possibility is that one of them is 2, the other 3. Indeed,
Z4 + 3 ~ +232*22
~ + 3=.2 + 3' = 211
so that a = 36 and I may have the value 3 or 3 (both yield
the same terms but in reverse order). The first, third and fifth
terms are perfect squares, and their sum is
4
' + 62+ 92 = 16 + 36 + 81 = 133.
37. (A) Let G and M denote the intersection of the medians (the
centroid of AABC) and the midpoint of side AC, respec-
tively. Draw MJ perpendicular to line RS a t J , and BK
and CL parallel to RS intersecting MJ at K and L, re-
spectively. See figure on p. 90.Then
MJ = + ( A D + CF) = i(lO+ 24) = 17
so that
M K = MJ - KJ = MJ - BE = 17 - 6 = 11.
Now since MG = QMB, M L = + M K ,because the line LG

parallel to the base K B of triangle M K B divides the other
sides in the same proportion. Therefore the segment sought is
x = GH = LJ = MJ - M L = MJ - QMK
= 17 - j(l1) = 9-.
38. (E) Mathematical notation facilitates statements about the system
S. Thus let single numbers 1, 2, 3, 4 denote the four distinct
pibs (see postulate PI); these are sets of maas (see PI). The
unique maa common to pibs i and j (see Pz) we denote by
ij or j i . Since every maa is one of these (see Pa), the complete
set of maas { 12, 13, 14, 23, 24,341 contains exactly
(;) = g = six elements (TI).
The three and only three maas in pib i are the maas
ij ( j # i) (T2). There is exactly one maa in neither pib i
nor pib j and hence not in the same pib with ij (T3). We
conclude that all three theorems TI, T2, T3 arc deducible from
the four postulates PI, P2, Pa, Pq as we have shown above.
Remark: The six intersection points (maas) of four non-parallel
coplanar lines, no three of which are concurrent (see figure),
serves as a model (finite geometry) which satisfies postulates
PI, P2, Pa, P,, where the four pibs are the four sets of three
collinear points. We note that in this geometry, two lines
(pibs) always intersect in a unique point (maa) but two points
(maas) do not always determine a line (pib) of the system.
39. (B) The nth set contains n consecutive integers the last of which
is the total number of elements in the union of the first n sets;
that is, the last integer in S, is
1 + 2 + 3 + -..+ n = an(= + 1).
The sum S, may be thought of as starting with this last
integer and proceeding downward through n consecutive
integers. Thus
S, = M a + 1)
+ t n ( n + 1) - 1
+ 3n(n + 1) - 2
+ .... .........
+ t n ( n + 1) - ( n - 1)
=fn'(n+1)-(1+2+...+n-l)
= + 1) - fn(n - 1)
fnyn
= $n(nZ + 1 ) .
When n = 21 we obtain S21= 3(21)(21* + 1) = 4641.

Remark: S,, is the sum of n terms of an arithmetic progression
with first term I 1 = fn(n+ 1) and common difference
d = - 1 so that
s, = + ( n - l)d]
fn[2t,
= +n[n(n+ 1) + (a - 1) (-- 1 ) l
= +n(n*+ 1)
as before.
92 THE M A A PROBLEM BOOK 111
40. (D) Let the vertex C of equilateral A A B C be a t the origin of

a rectangular coordinate system and the altitude from C coin-
cide with the positive x-axis. Denote the length of one side
of A A B C by s; then points A and B have coordinates
(3&, is) and ( i f i s , - is), respectively (see figure a). Ex-
pressions for the squares of the distances from P ( x , y ) to
C, B, and A respectively are
X' + r' = 102, (X - +V%)' + ( y + 3s)' = 8',
and
(X - + ( y - 4s)' = 62.
Subtracting the third equation from the second gives

2sy = 2 8 or sy = 14.
Substituting this value of sy in the second equation and using
x2 + = 102 gives
102 - ~
+ 50
+ + 14 = 64, 'S + 50 = GSX,sx = 's-
d -
S X S*
Substitute the expressions for sx and sy just found into

(sx)' + (sy)* = (x2+ Y)S2
and obtain
(s2 + 50)2+ 142= 1 0 2 ~ ~
3
which reduces to the quadratic equation in s2

'
S +
- 2 0 0 ~ ~3088 = 0.
Its roots are s2 = 100 f 48&, and we discard the smaller

because s2 > 100. The desired area is
A =I dTS2
- - - 25&
4
+ 36 N 79 or choice (D) .
OR
We may use the fact that 6, 8, 10 are the sides of a right tri-
angle to facilitate solution of the present problem. To this end,
construct AAP'B congruent to A A P C (see figure b). Then
QPAP' = Q P A B + QBAP'
= Q P A B + %CAP = 60"
so that isosceles APAP' is equilateral. Hence AP'PB is a

6, 8, 10 right triangle. Moreover
Q B P A = gI3PP' + gP'PA = 90" + 60" = 150".
Using the law of cosines A A P B now yields
S* = 6' + 82 - 2.6.8 cos 150" = 100 + 48&.
Remark: In our first, algebraic solution, we made no use of the
fact that a = P A , b = PB, c = PC was a Pythagorean
triple. The second, geometric solution can also be made to
work for any positive triple a, b, c for which the s u m of any
two is greater than the third. The construction of AABP' .-
A A C P is as before, AAPP' is again equilateral with side a,
and 6 = g P ' P B , now not necessarily a right angle, can
nevertheless be determined by the law of cosines since the three
sides a, b, c of AP'PB areknown. Wecannowuse + A P B =
+
60" 6 to find s. The calculations are more cumbersome than
+
in the case a2 b2 = c2, but the same principle applies.
In the second solution, AABP' could have been obtained
by a 60" clockwise rotation of A P about A into the position
AP'. [See figure b, on p. 94, and cover of this book.]
If, similarly, we rotate BP 60" clockwise about B into
BP" and C P 60" clockwise about C into CP"' we obtain
the hexagon AP'BP"CP"', see figure b. The part of the
hexagon outside of AABC consists of the three triangles
which have been rotated out of A A B C and which, together,
filled up A A BC ; thus Area of hexagon = 2. Area of A ABC.
p'
b
On the other hand, the hexagon consists of three equilateral

triangles of sides a, b, t respectively (shaded in figure b)
and three congruent triangles of sides a, b, and t, whose
area can be determined by Heron's formula. Thus
2(Area of A A B C ) = Area of hexagon
v3
= - (a'
4
+ bZ+ c') + 3 d ~- (a ) ~
(U - b ) ( U - c),
a+b+c
where u =
2 .
In our case, a = 6, b = 8, c = 10; this yields

v3
2 Area(AABC) = 4(200) + 3412.6.4.2
= 5 M + 3.24,
so Area (AABC) = 25v3 + 36.

1968 Solutions
Part 1
1. (D)Let C and d denote the measure of the original circumference

and diameter, respectively, so C = 7rd. After the increase
c+ P = r ( d + 7r) = rd+ 2 = c+ 2.
Hence P = r2.
2. (B) When the equal numbers 64'-'/4*' = (64/4)"-' = 16'-' and

2562z= (162)22 = 16'" are expressed as powers of 16 as we
have done, the esponents of 16 must be equal, i.e. 4x = x - 1
so that x = -4.
OR
Instead of comparing esponents, the logarithms, most con-
veniently to base 2, 4 or 16, of the equal numbers may be
equated. Thus, using 2 as base,
log? (64"'/4'-') = log2 256'=
(x - 1) log2 26 - (X - 1) log2 22 = 22 log2 28
(X- 1).6 - ( X - 1).2 = 22.8,
x = -4 as before.
3. (A) The slope of the required perpendicular line is -3, the negative
reciprocal of the slope $ of the given line. The line through points
(0,4) and ( x , y) has slope (y - 4)/x, so the required equa-
tion is equivalent to
y-4=-3 or y + 3 x - 4 = 0 .
X
OR
One may substitute the slope m = -3 and the y-intercept
b = 4 into the form y = mx+ b getting the equation
y = -32 +
4 which is equivalent to y 3 x - 4 = 0 of +
choice (A).
4. (C) Since a * b =
ab
-
a + b'
we have 4 *4 =
4.4
-
4+4
-
16
--=
8
2, so
4.2 8 4
4*(4*43 =4*2 = - =-=-
44-2 6 3'
5. (A) The definition f(n) = $n(n + 1) (n + 2) gives

f(r) = )Y(Y + 1)(r + 2) when n = I,
f ( r - 1) = $ ( Y - l ) r ( r + 1) when n = I - 1.
Subtracting the second from the first, we obtain
f(l) - f ( r - 1) = fr(r + l)[(r + 2) - (I - 113
= )Y(Y + 1)(3) = + 1).
Y(Y
6. (E) Since the sum of the angles in a triangle is 180" (see diagram
on left)
Q E + < C D E + <DCE = Q E + S = 180' in A E D C ,
and
Q E + < B A D + <ABC = < E + S' = 180" in A E A B .
Hence S = S' = 180' - <El so that I = S/S' = 1.
I \
I \
7. (E) To see that the length O P cannot be determined from the

given data, we shall show that a segment OP of arbitrary
length (and direction) can be constructed. The diagram on
right shows several different instances.
Let segment QOC be drawn with QO = 3, and hence

OC = 6 inches. Draw an arbitrary segment OP from 0 to
any point P not on line CQ. Extend PO to the point A such
that OA = 2P0, and denote by B the point where C P and
AQ, extended, intersect. We claim that A P and CQ are
medians of A A B C . For, triangles AOC and POQ are similar
(the lengths of corresponding sides have ratio 2 to l ) , so
PQ 11 AC, and PQ is half as long as AC. But then P and Q
bisect BC and BA, respectively, so that A P and CQ are
indeed medians of A A BC.
8. (B) Let N represent the positive number, so that the incorrect

result is 4 N . Since the correct product is 6 N , the error is
6N - QN = V N . Therefore the percent error based on the
--
correct result is
Error
,100 e
3500
97.
N
Correct Result 36
9. (E) If x 2 2 or x 5 - 2 , then x + 2 and x - 2 are both non-
negative or both non-positive so that the given equation yields
-
x + 2 = 2 ( x 2 ) or,equivalently, - ( x + 2 ) = - 2 ( x - 2 ) .
Hence in this case, x = 6. If - 2 < x < 2, then x + 2 is
-
positive and 2 ( x 2 ) is negative, so that the given equation
+
yields x 2 = - 2 ( x - 2 ) , x = 3. The required sum of all
+
values of z satisfying the given equation is 6 3 = 63.
OR
Since the absolute values of two real numbers are equal if and
only if their squares are equal, the given equation yields
*+4x+4 = 4(*-4x+4)
which, when simplified, yields the quadratic equation
3x2 - 20x + 12 =0
with real roots. The sum of its roots is -(-20/3) = 6).
10. (C) First, statements (A) and (B) may be invalid because each
requires that the set of all fraternity members be nonempty
which is not required by hypothesis I or 11. Again, the hy-
potheses I and I1 would allow the set of all fraternity members to
be a nonempty subset of the set of all students, but neither (D)
or (E) permits this and accordingly may be invalid. Now
choice (C) is valid under the hypotheses, because by I there
esist dishonest students, and by I1 they cannot be fraternity
members.
Part 2
11. (B) Let 11 and r2 denote the radii of circles I and I1 respectively.
Equating the arcs, we get
60 45
-.2*r1
360
= --.2KY2.
360
... _rz --_43
I1
Since the areas of any two circles are proportional to the

squares of their radii,
Area (I) = -rlz- - -32- 9
- -
Area (11) rz2 42 16 *
12. (C) Since (74)* + lo2 = (121)* the triangle is a right triangle with
the given circle as its circumcircle. The hypotenuse of length 124
is a diameter of the circle because the right angle opposite it
subtends a semicircle. Hence the required radius is one half of
that diameter: 1 (124) = 25/4.
Remark: A formula which gives the radius R of the circumcircle
directly in terms of the sides a, 6, and c of any triangle is
R = abc/(4K)
where
K = ds(s - a)(s - 6)(s - c) = Area ( A A B C ) ,
and
s = 4(a + b + c) = semiperimeter of A A B C .
If you don’t identify the given lengths (Y,%, v)
= $(3,4,5)
as a Pythagorean triple, you would be obliged to use this formula
for R. Use it to check the result in this problem.
13. (B) The sum and the product of roots of the given quadratic
equation are - m and n, respectively:
m+ n = -m and mn = n.
Hence m = 1, n = -2, and m + n = -1.
14. (E) Choice (E)can be deduced algebraically, for example by writing

the given equations in the equivalent form x - 1 = l/y,
y - 1 = l/x, whence y(z - 1) = x(y - 1) = 1. Therefore
xy - y = xy - x, so that x = y.
Remark: Choice (E) was determined without actually finding the
values of x (and y) which satisfy the given equations. How-

ever, we can easily find them and check our previous result.
Substituting for y (from the second given equation) in the
first equation yields
1
+ l/x
X
x = 1+-= 1+-
1 x+ 1’
so
X
x- 1 = - and 9 - x - 1 = 0.
x+ 1
The roots of this quadratic equation are
1 - 4%
x=- 1 + 4 3 and x =
2 2 .
Substituting these into the second equation yields the solutions
and
In either case, x = y.
15. (D)The required product P may be written as

P = (2R - 1)(2K+ 1)(2k+ 3),
where k is any positive integer. The integer k has exactly one
of the following three properties:
(i) it is divisible by 3 (i.e. k = 3m, m an integer)
(ii) it leaves 1 as remainder when divided by 3 (i.e. R =
3 m + 1)
(iii) it leaves 2 as remainder when divided by 3 (Le. R =
3m+ 2).
In case (i) the last factor of P is divisible by 3 . In case (ii) the
second factor, 2k +1 = 2(3m +.1) + +
1 = 6m 3 is di-
visible by 3. In case (iii), the first factor, 2R - 1 =
+ +
2(3m 2 ) - 1 = 6m 3, is divisible by 3. Thus, in any
event, P is divisible by 3 . To see that no larger integer divides
all such P, take PI = 1.3-5, Pz = 7.9.11, and observe
that 3 is the greatest common divisor of PI and P2.
0.1
H 1 2
.-1
..- 2
- - - _ _2_- - - - 3 _ - - - y' -3
17. (C) The numerator of f(n) is 0 when n is even and - 1 when n is

odd, because the sum of any two consecutive x's is zero:
%k+ xk+1 = (-I)'+ = (-l)k(l - 1) = 0,
and (-l)-l = -1.
Hence f(2k) = 0, f(2R +
1) = - 1/(2k + l ) , and f(n)
is therefore contained in f 0, - 1/n}.
18. (D) Line AC is a transversal across parallel lines A B and DF, so

that the angles FEC and BAE are equal. The former is given
equal to angle GEC, which in turn is equal to its vertical angle
BEA. Hence + B A E = +BEA, so A A B E is isosceles with
B E = BA = 8. Since corresponding sides of similar triangles
DEC and ABC are proportional,
EC DE EC 5
B C = a that is, -
8+EC
=-
8'
Thus 8EC = 40+ SEC, 3EC = 40, and EC = 4013.
19. (E)Let q and d denote the number of quarters and dimes,

respectively, to total $10. Then, in cents, 25q 1Od = 1o00,+
which is equivalent to 2d = 5(40 - y). Since the left side is a
positive even integer, the right side must also be a positive even
integer, so 40 - q must be even and positive. This is the case
when q is any even integer less than 40, so the number n of
solutions is 19 or choice (E).
OR
We look for solutions ( q , d ) in positive integers of the equation
5 q + 2d = 200 (equivalent to the equation above). We ob-
serve that ( 2 , 95) is such a solution. By a theorem in number
theoryt every solution in integers is of the form ( q , d) =
+
( 2 21, 95 - 51) where t is any integer. Those integers t for
which both q and d are positive are 1 = 0, 1, 2, ...,
18 so
that the number n of solutions is 19 as found before.
20. (A) The sum of the n interior angles of the polygon in degrees is
160 + (160 - 5) + (160 - 5 . 2 ) + (160 - 5.3) +
+ 160 - 5 ( a - 1)
= 16On - 5(1 + 2 + ... + n - 1) = 160n - 5
n ( n - 1)
2
%
5n 5n
= - [64 - (~t.- I)] = - (65 - t ~ ) .
2 2
which is equal to 180(n - 2 ) for any conves IZ sided polygon.
t See Confinired Fracfions by C. D.Olds, Vol. 9 in this NML series, pp. 44-45.
$ Here we used the fact that the sum of the first k positive integers is 3h(k i- l ) ,
see footnote on p. 114.
Equating these expressions and multiplying by 2/5, we get

n(65 - n) = 72(n - 2) which is equivalent to the quadratic
equation
nZ+ 7n- 144 = 0 or ( n - 9)(n+ 16) = 0.
Since n is positive, n = 9 as stated in choice (A).
Part 3
21. (D) Wehave

S=1!+2!+3! +4! + + 99!
+5! - 0 .
= 1 + 2 + 2.3 + 2-3.4 + 2-3.4.5 + + 99! **-
= 1 + 2 + 6 +24 +1OK,
K a positive integer, because each of the terms 5!, 6!, -,99!
contains the factors 2 and 5 and hence is a multiple of 10. The
sum of the units’ digits of S is therefore
1+2+6+4= 13.
The units’ digit in S is accordingly 3 as stated in choice (E) .
22. (E) Fundamental in the proof is the fact that a quadrilateral with
four given segments as sides exists if and only if the length of
each segment is less than the sum of the lengths of the other
three.? Now let SI, s2, s3, and s4 denote the lengths of the four
segments. If a quadrilateral exists, then by the fact mentioned
above
si < ~2 + sa + ~ 4 .
By hypothesis,
s1 + + +$2 S8 s4 = 1.
If we replace the sum s2 + sa + s4 by the smaller number s1,
we obtain the inequality
s1 + s1 = 2sl < 1 so that s1 < 1.
t The “only if” part is obvious since the length of a polygonal path is a t least equal
to the distance between its endpoints. On the other hand if the length of each segment is
+
less than the sum of the lengths of the other three, label the segments 50 that s, SZ >_
SJ+ s4. Then there is a triangle with sides sl, sz, S3 +
s,. This triangle may be
viewed as a quadrilateral with sides SI, SZ, $8, s,.
Since none of the four segments is in any way special, we can

deduce, by the same argument, that
s2 < f, s3 < $ and sq < 4.
Conversely, if < 4 (i = 1, 2, 3, 4) and s1+ + +
s3
> 1 - 4 = 3:
si s2
= 1, then s 2 + s3+ = 1 - s1 so > SI,
< s2 + s3 + s4.
s4 s4
s1 Corresponding inequalities hold for the other
segments. Choice (E) is therefore correct. All the other choices
fail; for esample, a rectangle with adjacent sides, of lengths
and fs has perimeter 1, yet is excluded by all other choices.
Choice (D) is clearly excluded, since there is no such division
into four segments.
23. (B) The given equality is equivalent to

log ( x + 3) ( x - 1) = log (x2 - 2x - 3 ) ;
therefore
(x + 3) ( x - 1) = x2 - 2 x - 3,
x2+2x-3=x2-2x-3, x=o.
But when x 0, both x - 1 and x2 - 2x - 3 are negative
=
so that neither log ( x- 1) nor log ( x 2 - 2x - 3) is defined;
thus the equality is satisfied for no real number, as stated in
choice (B) .
24 t 2 x
24. (C) Let x and x/2 be the width (in inches) of the frame at the top
and bottom, and a t the sides, respectively. Since the total area is
twice the area of the picture, we have
(2x4- 24)(x+ 18) = 2(18)(24)
104 THE MAA PROBLEM B O O K 111
which reduces to
+
2(9+ 3 0 ~ 216) = 2(2-216)
or
x2 + 30%- 216 = + 36) (X (Z- 6) = 0.
Hence x = 6 ( x = -36 is inadmissible). The required ratio is
(x + 18)/(2x + 24) = 24/36 = 2/3
or choice (C).
25. (C)Let u and ux be the speeds (in yards per unit of time) of Ace
and Flash, respectively, and l the time (in the same unit of
time) required for Flash to catch Ace. Then the distance in
+
yards run by Flash is uzf = y 01; so that vf = y / ( x - l ) ,
and hence the required distance ulx = x y / ( r - 1) yards.
Remark: The answer must have the dimension of yards. Since x is
dimensionless and y is measured in yards, only choices (A)
and (C) meet this requirement. We can eliminate (A) by
noting that as x approaches 1, the solution must approach
infinity.
26. (E)Thesum S = 2 4 + + 2 K = 2(1 2 + +.K) + +

= K(K +
1).t When K = 999, S = 999,000 < 1,000,OOO;
butwhenK = l,OOO, S = 1O00~1001= 1,001,OOO > 1,OOO,OOOso
that N = 1,000 is the smallest integer for which S > 1,000,OOO.
The sum of the digits in 1,OOO is 1.
27. (B) When n is even, grouping the n terms into 4% pairs gives
sn= (1 - 2) + (3 - 4) + . * - + [(n - 1) - n3
= -1-1... ...-I = -fa,
P
n/2 terms
When n is odd, grouping the terms after the fint into 4 (n - 1)
pairs gives
Sn = 1 + (-2 + 3) + (-4 + 5) + . + [- (n - 1) + a] *.
= 1 + [ 1 +<1 + * . . + 1 ] = J
l+i(fl-l) =t(n+1).
(n - 1)/2 terms
Hence S,r+ St,+ S ~ O
= *+9-9 = 9+ 17 - 25 = 1.
t The sum of the first K positive integers is J K ( K + l ) , see footnote on p. 114.

+
28. (D) We are given that $ ( a b ) = 2 d s which is equivalent,
after dividing by b and multiplying by 2, to a/b 1 = -+
4 4 3 ; this, in turn, is equivalent to the following quadratic
equation in d f i :
($y-4$+1=0, so &= 4fdS

2
= 2 f G .
The solution 2 - G must be rejected because the requirement

a > b > 0 implies a/b > 1 while 2 - d3 < 1. Hence
0.
-
b
= (2 + d)'= 7 + 4d3 x 7 + 6.928 X 14.
29. (A) We shall use the fact that for 0 < x < 1, any positive power
of x is less than one.? I n particular, y = Xz < 1. Moreover,
x_ -- -x= XI-= <1
Y x=
since 1 - r > 0, so x < y; and
_z -- -xu= xY-= < 1,

Y x=
since y - x > 0; hence z < y. Finally
1 %
_ -_ - = xl-u < 1, since 1 - y > 0.
z xu
:. x < z. It follows that z < z < y.

OR
Since 0 < x < 1, log x < 0. If an inequality is multiplied by a
negative number, the sign of the inequality is reversed; thus,
multiplying 0 < x < 1 by log x yields
0 > .z logx > log x, or 0 > log A? = logy > log x.
Since the loprithniic function is an increasing function, and
since 0 = log 1, it follows that 1 > y > x. Again, multi-
plying by log x , we obtain
log x < y log x < x log x, or log x < log x~ = log z < log y,
so x < z < y .
t Since x < 1, we have log N < 0. Hence if 1 > 0, then log = 1 log x < 0, so
that d < 1 .
106 T H E M A A PROBLEM BOOK I11
30. (A) A convex point set, by definition, is such that with every pair of
its points it contains the entire line segment joining them. Conse-
quently, a polygon is convex (i.e. bounds a convex set) if and
only if all its angles are 5180’. I t follows from the convexity
of PZ that each side of P1 can intersect P2 in at most two
points, so that the total number of intersections is at most 2nl.
We show next that this maximum of 2n1 intersections is
always attainable by a suitably determined PZ which passes
through any two preassigned interior points on each of the 721
sides of PI.First mark these 2nl points. Next join the two
points adjacent to each vertex of PI by a “cutoff” segment.
Let the two “cutoff” segments through the two marked points
on any side of PI be extended outside PI from the two points.
The intersection of these two “cutoff” extensions is taken to be
a vertex of P2 if they meet, but if not, a vertex of PZ may
be assigned as any point on each of the extensions. It is con-
venient in the latter case to take the two vertices of PZ as
consecutive and joined by a side of Pz parallel to the side of P1
being considered. Each “cutoff” segment with its two-way
extensions to consecutivevertices of Pz is a side of P2 with two
points of PI on its interior. There are nl “cutoff” segments
(one for each vertex of PI) and hence PZ intersects PI in
exactly the maximum of 2nl points. Polygon P2 is convex
because each of its angles is either an angle of a triangle or
equal to an exterior angle of a triangle, hence in any case less
than 180”.
Remark 1: In the figure, P1is AABC and P2 is a convex quad-

rilateral RSTU intersecting P1 in the maximum number
21tl = 6 points G , H, J, K , M , N which lie on “cutoff”
extensions of segments NG, HJ, KM. The vertices of €‘z
on these extensions are S, T; T, U ; U , R, respectively.
Side RS of Pz is taken parallel to side AB of P1
because the “cutoff”extensions outside of P1 at M and N do
not meet, so that R and S are taken as consecutive vertices of
P2 on extensions MR and N S . Angles R and S are equal to
exterior angles of “cutoff” triangles AKM and BGN, respec-
tively, while angles T and U are angles in triangles GHT and

JKLr, respectively.
Remark 2: When 721 = 1 and iE2 = 3, P1 is a segment while Pz
is a triangle. The maximum number of intersections is 2,
eliminating choices (B) , (C) , and (D) .
Part 4
31. ( D ) Using the area formula K = (\/3/4)s2 for an equilateral tri-

angle in terms of a side s in triangles I and 111, we obtain the
lengths AB = 8 f l and CD = 4\/2. The length of a side of the
square I1 is BC = 4V2 so that the length of A D is 16V2; and
124% or & of this is 2d2 and is to be subtracted from BC,
reducing the length of BC to 2V2 or $ of its former value. Hence
the area of the square is reduced to f of its former value. This is
a reduction of 7.5% in the area of the square.
ZV
2u , 8u
0 2 A"
32. (C) Let ( u , v ) denote the uniform speeds of ( A , B ) in yards per

minute. Then their equal distances from 0 after 2 and 10
minutes (represented by OA' = OB' and OA" = OBI' in
the figure) may be expressed in terms of u and v as
2u = 500 - 2v and 1Ou = 1Ov - 500,
respectively. Adding these two equations gives 1224 = 8v, so

that u/v = 2/3, and the ratio of the speeds of A and B is
u:v = 2:3 or choice ( C ) .
Remark: The positions of A and B after two minutes are A' and
B' with OA' = 2u = OB'. During the next 8 minutes, A
moves 811 more yards to A", B moves 8u more yards to B",
and
OA" = 1oU = OBI' = 8v - 2 ~ ,
whence
12u = 8v and u/v = 2/3.
Note that this solution makes no use of the information that B,
initially, is 500 yards away from 0; but we did use the fact that
B was walking towards 0 (otherwise their equal distance from
0 a t two different times might have led us to believe that
u = u). To find the ratio u/v, one relation involving u and v
suffices, and the solution confirms this principle.
33. (A) Let x , y, and z denote the first, second, and third digits of N
in base 9 so that
81x + 9y + z = 492 + 7y + x or y = 8(3z - 5 x ) .
Since 0 5 y < 7 (it appears as a digit in base 7 ) , the integer
n = 32 - 5x is zero (otherwise I 8n I would begreater than 7).
Hence y, the middle digit, is zero. Moreover 0 < z < 7 (since
N has three digits in base 7) ;and since 32 = Sx, z is divisible
by 5. Hence z = 5, and x = 3, so that
N = 3059 = 5037 = 248x0.
34. (B) Let d and p denote the number of votes which originally
defeated and later passed the bill, respectively. Then 400 - d
and 400 - p were the number of votes first for, later against
the bill, respectively, and d - (400 - d ) = 2d - 400 was the
margin of defeat in the first vote, while 2p - 400 was the
margin of passage in the second. The two pieces of information
given in the problem lead to the equations
2p - 400 = 2[2d - 400) and p = Hd.
The first is equivalent to 2d - p = 200, and after substituting
for p from the second, we find d = 220, p = 240. The
required difference is p - (400 - d ) = 240 - 180 = 60.
/ Iz
35. (D) Using the area formulas for a trapezoid and a rectangle, we write
K +
(1/2)HG(EF C D ) 1 1CD
+ $)
1
- -
-
R HG. EF = -2 (1 =2 + j EF ;
see the figure. Now
C D 2GD G D 4 0 P - W - d a 2 - W
-=---- - - -
EF 2HF HF dOF2-OHz daz-OH2
by use of the Pythagorean Theorem applied to right triangles
OGD and OHF. If we denote JH by z so that JG = 22,
then OH = a - z, OC = a - 2 2 , and
Now as OG approaches a, z approaches 0, so that CD/EF

approaches 4 4 a / ( 2 a ) = d2.
+
Hence K / R = 4 $(CD/EF> approaches 4 = +
+ l/G, as stated in choice (D) .
OR
The analytic version of the solution is this: Choose 0 as the
origin and OJ as the y-axis of a Cartesian coordinate system.
Denote the coordinates of points D, C, F, E, G as shown in
the figure; since H is the midpoint of GJ, its ordinate y2 is the
+
average ( y ~ a ) / 2 of the ordinates of G and J . Now the
area formulas yield
K =
2
so that
-K= - =
XI+
- x2 z1 1
R 2x2 2x2+ 2
Since the coordinates of every point on the circle satisfy the
+
equation XZ 9 = a2, we have
x12 = a2 - y? = (a - yd ( a + yd,
= 4a2 - [ylz + 2ayl+ a2] = 3a2 - 2ay1 - y12

= (a - yl) (3a + y d .
Thus
as yl - a, this fraction approaches
-a- +_a- _ - 2 - 1
3a+a 4 2
and x1/2x2 approaches l/a.Hence
K 1 1
---+-.
R d2 2
1969 Solutions
Part 1
1. (B) We are asked to solve the equation

a+x c
-=-
b+x d'
An equivalent equation is
arl+ xd = bc + ax, so that (G - d)x = ad - bc,
and hence
Comment: When c/d is chosen as (i) - 1, (ii) the reciprocal of a/b,

(iii) the square of a/b, then - x takes the values
+ + +
(i) 4 (a b ) , (ii) a b, (iii) a b / ( a b ) , respectively;
i.e. (i) the arithmetic mean, (ii) the sum, (iii) half the harmonic
mean of the numbers a and b.
2. (A) Let C represent the cost in dollars. Then

x = C - .15C= .85C, and y = C + .1SC = 1.15C.
Therefore the required ratio is
y / x = 1.15C/.85C = 23/17
3. (E) The following identity is valid whenever n and I are integers

with tz > r; it can be verified by direct multiplication:
x" - xr = (x - 1)(x"-L+ xn-2 + * 9 * + x').
When we set x = 2, we get x - 1 = 1 SO that
2" - 2' = 2"-1 + 21-2 + ... + 2'.
Now N = 1 l W = '2 +
2a, and, by means of above identity
with n = 3, I = 0, we get
N - 1 = 24+ (23- 1) = 24 + (22+ 2 + 1) = 1 0 1 1 1 ~
as stated in choice (E).
OR
We can just subtract, using the familiar algorithm, modified for

base 2:
1 lo00
- 1
-
10111
OR
We can convert N to base 10 (getting N = 24), then subtract
1, then convert 23 back to base 2, getting N - 1 = 101112.
*
4. (E) Bydefinition, (3,2)
(x, y )
* +
( 0 , O ) = (3 - 0 , 2 0) = ( 3 , 2 ) , and
(3, 2) = ( x - 3, y + 2). If (3, 2) and (x - 3,
y + 2) represent identical pairs, then 3 = x - 3, SO that
x = 6 as stated in choice ( E ) . Incidentally, 2 = y +2 yields
y = 0.
5. (B) The sum of all possible values of N such that
is the sum of the distinct roots of the equivalent quadratic

equation
N 2 - R N - 4 =; 0.
These roots are
N - R+ 4w-
2
and
= R - ,/R-
2
9
their sum is R.
6 . (C) From the common center 0 of the circles, draw the radius Y
of the smaller circle to its point of tangency with the chord c,
and draw a radius R of the larger circle to an endpoint of the

chord. For the right triangle with legs I, c / 2 and hypotenuse
R, the Pythagorean theorem gives
*
while the area of the ring is
= r R - r i = r(R - I').
Therefore ($ =
25
T , t2 = 50, and c = 5d.
7. (A) Since the points (1, yl) and (-1, yz) lie on the graph of
y = axa + +
bx c, we can substitute their coordinates into
the equation, getting
yl=a+b+c and ya=a-b+c.
Therefore yl - ya = 2b = -6, and b = -3.
8. (D) Arcs AB, BC, and CA subtend the inscribed interior angles
C, A , and B of AABC which therefore have measures
J ( x 75'7, 4 (2% 25"), and 4 (3%- 22"), respectively,
+ +
with sum 180". The resulting equation
J(x+ 75") + J ( 2 x + 25") + i ( 3 x - 22") = 180'
has the solution x = 47" from which we find
(+C, 44 +a = (61", 59Y, 591">,
so that +C is an interior angle of 61".
9. (C) The integers form an arithmetic progression with first term

a = 2, common difference d = 1, and number of terms
n = 52. Hence their sum is
S = 442a + ( n - 1 ) d ) = f 5 2 [ 4 + 51) = 1,.52(55)
and their average S/52 = 1-55 = 271 or choice ( C ).
Comment: Virtually all properties of the sum of an arithmetic pro-
gression are based on the identity
(*> 1 +2+ 3+ + n = $n(n+ l).t
In particular, the arithmetic mean of n numbers in arithmetic
progression is the arithmetic mean of the smallest and largest.
Proof: Let the n numbers, in increasing order, be

a, a + d, a + 2 4 ..., a + ( n - 1)d.
Their arithmetic mean is
1
- [ a + a + d + a + 2d+
n
... + a + (n - l)d]
d (n - 1).
(by identity (*))
=a+; 2
1
= - [ a + a + ( n - 1)d],
2
where we recognize, in the last expression, the arithmetic mean
of the first and last of our numbers. (Note: an alternative proof
can be obtained by directly applying the Gauss trick to the pro-
gression a, a + d, ...,
a + ( n - 1)d.)
Applying this to the problem at hand, we get
4[2 + 531 = 271.
t This formula for the sum S,,of the integers from 1 to n is derived by the famous
Gauss trick of writing the sum twice, the second time in reverse order, and adding:
s, = 1 +2 +3 + .'. +n-1 +n
S,=n +n-l+n-2+..-+2 + 1
2.9.- n+ l + n + l +a+ 1 + +n+ 1+n + 1;

since there are n terms on the right, all equal to I + 1, we obtain 2s" = n ( n + 1).
i-’
10. (C) Let 0 denote the center, r the radius, and 11 and t 2 the two
parallel tangents to the given circle (see figure). Then the
locus of all points equidistant from 11 and 1% is the parallel
line QOP midway between them (and hence through 0). The
intersections P and Q of this locus with the concentric circle of
radius 2r (dotted in figure) together with the center 0 itself
are the only three points which are equidistant from the circle
and its two parallel tangents t, and 12. The number of such
points is 3.
Part 2
+
11. (B) Since the sum of segments PR RQ is to be a minimum,
points P, R, and Q must be collinear, so that the quotient of
the difference of the y’s and the difference of the x’s, in the
same order, is constant for any pair of the pints. Using the pain
P and R, and P and Q gives
m - (-2) - 2 (-2)- which is equivalent to m = -3.
1 - (-1) 4 - (-1)
Comment: In a rectangular coordinate system, the quotients equated
above are called the slope of the line.
12. ( A ) The espression F may be written as
+
which is the square of the linear expression ( x #) in x pro-
vided m/2 - 16/9 = 0 or m = 32/9, a particular value of
m between 3 and 4 as required in choice (A).
OR
We may note that any quadratic expression a P + +
bx G in
x is the square of a linear expression in x if and only if its dis-
criminant bz - 4ac is equal to zero. Here F is a quadratic
expression in x with a = 1, b = 4, G = m/2, so that
p-4a = -- 4 @ = 0
64 or m = 32
-
9 9
Hence m = is the only value of m for which F is the
square of a linear expression in x.
13. (B) The difference of the areas of the larger and smaller circles,
-
TR' Irz, is equal to the area of the region outside the smaller
and inside the larger circle. By hypothesis, this area times a/b is
equal to the area of the larger circle:
a a
rR2=-(rRZ-n2) or - r 2 = R z
b b
Hence
P-.
-- a
r2 a-b
and the required ratio is
14. (A) Since the fraction (9 - 4)/(22 - 1) is required to be positive,

XS - 4 and 22 - 1 must be both positive or both negative.
When 22 - 4 > 0 so that I X I > 2, then P - 1 =
(9 4)- + 3 > 0 also. When 22 - 1 < 0 so that I x I < 1,
- -
then XZ 4 = (9 1) - 3 < 0 also. The set of all values of
x for which the given fraction is positive is I > 2 or x < -2
or - 1 < x < 1.
Remark: The fraction f(x) = (9- 4)/(21 - 1) is an even func-
lion of x ; this means f ( x ) has the property
f(-4 = f(4
with the geometric consequence that the graph of f(x) is
symmetric with respect to the y-axis. In particular, f ( x ) > 0
on symmetrically located points of the domain of f, and our
solution confirms this fact.
15. (D) Since chord A B has length I, A A O B is equilateral. Perpen-

dicular OM bisects A B , so that A M has length r / 2 and is
the hypotenuse of right triangle M D A with legs A D =
$ A M = r/4, and DM = V'31/4 (because Q A = 60').
:. Area of A M D A = i ( A D ) ( D M ) = -.-.-
1 r V'3r
2 4 4
rw3
= -.
32
OR
Triangles M A D and OAM are similar (both having angles of
30', 60' and 9
0') ; and A D and A M are corresponding sides.
Since
AD
-=-
1 Area A M A D 1
AM 2' Area A O A M 4'
1 1 r n f 3 Iw3
Area AOAM = -.4M.OM = --- = -
2 22 2 8 .
1 IZV3 rZV3
.: Area A M A D = -- =-
4 8 32'
16. (E) In the binoiiiial espansion
the swii of the second and third terms, when a = kb, is
- it ( k b ) "-'b + n(n 2- 1) (kb)11-2b2= 0,

which yields, after division by rak*1-2bn,
n-1
-k+- = 0 or it = 2 k + 1,
2
17. (D) The left member of the given equation can be factored to give the
equivalent equation
(2" - 2)(2= - 6) = 0.
Since a product is zero if and only if at least one of the factors is
zero,
2= - 2 = 0, 2" = 2, x = 1,
or
2' - 6 = 0, 2" = 6, 22-' = 3, (X - 1) log 2 = log 3,
log 3 log 3
x- 1=- x = 1+-
log 2 ' log 2
which is the value of x stated in choice (D). The value x = 1
satisfies the given equation also.
18. (B) Each of the graphs consists of a pair of nonparallel straight

lines, the first pair having equations
I:x-y+2=0 and 11:3z+y-4=0
and the second pair having equations
III:x+y-2=0 and IV:2~-5y+7=0.
The intersections of line I11 with I and I1 are the two distinct
points (0,2) and (1, 1 ) , and those of line I V with I and I1 are
two more distinct points (- 1, 1) and (H,G), giving a total
of four distinct points common to the graphs of the two given
equations.
Comment: There can be no more than four points common to the
two graphs, because each line of the first (second) pair can
intersect those of the second (first) pair in no more than two
distinct points. I t is interesting to note that in the present
problem, the slight change of equation I V to x - 5y 4- 4 = 0
results in the two graphs having only three distinct points in
common, even though each line of the second pair still inter-
sects the first pair in two distinct points.
Question: The graphs of the four lines involved in the original
problem intersect in six points. Can you explain the two extra
points?
+
19. (B) The given equation X'y' - 1099 9 = 0 is equivalent to
($9- 1) ($9- 9) = 0. The product on the left side is zero
only if $9 - 1 = 0, i.e. $9 = 1, or if $9 - 9 = 0, i.e.
x2y2 = 9. Therefore x y = f l or x y = &3. The ordered

pairs having positive integral values are (2, y ) = (1, l ) ,
( 1, 3), (3, 1). There are 3 such pairs as stated in choice (B) .
20. (C) Let x denote the first and y the second factor in the proposed
product P. Then
3.6.1018 < x < 3.7-1018, and 3.4.10" € y < 3.5.101'.
Member by member niultiplication of these inequalities gives
(3.6) (3.4) lo'* < xy < (3.7) (3.5) 1032.
Since the lower bound of P = xy on the left and the upper
bound on the right are both 34 digit numbers, P is also a 34
digit number as stated in choice (C).
Part 3
21. (E) The distance from the line x + y = 4% to any point

(XI, y d is
Its distance from the origin, (0, O ) , is therefore 4%.

Thus it
+
is tangent to the circle xz yz = m, where m may be any
non-negative number.
OR
Using the given linear equation, we find y = 4%- x and
substitute it into the given quadratic equation:
%a+ y ' = x2+ (42m-x)*
= 2x2 - 2 ~ 4 %-+ 2m = m,
Note that x = 4 m/2 is always a double root of this quadratic

equation. This shows that the two graphs have a single point in
common, hence are tangent no matter what non-negative value
m has.
+
Comment: Let the graphs of x2 y2 = m and x + y = 4% be
magnified by a factor k [i.e. each point (2, y ) is replaced by
( k x , k y ) ] . The resulting point sets satisf the equations x2 +
+
3 = k?m and x y = k d % = &m, respectively. This
pair differs from the original pair in that the number m has
been replaced by k2m. If the original graphs were tangent, so
ate the magnified versions. Since k2 is an arbitrary positive
constant, we conclude: If the two given graphs were tangent for
some non-negative number ml, they are tangent for any other
non-negative number m. To see this, set m/ml = k2 so that
mp = k2ml.
This reasoning enables us to eliminate choices (A), (B) , (C) ,
(D) immediately. If, in addition, we assume that one of the
choices offered is correct, then (E) must be it.
X
0 A B
22. (C) The area whose measure K is required consists of the isosceles
right triangle with legs OA and AD of length 5 and the
trapezoid ABCD with altitude A B = 3 and bases BC = 11
and AD = 5 . Thus
K = Area of AOAD + Area of trapezoid ABCD
= $5.5 + $3(S + 11) = 36.5.
23. (A) For k = 2, 3, - a a, +
n - 1, the sum n! k is divisible by k
because n! has K as a factor. Therefore, any integer m such
+ +
that n! 1 < m < n! n is composite, and there are no
+
prime numbers greater than n! 1 and less than n! n. +
24. (E) Let Q, Q' and Q" be the quotients in the respective divisions
of P , P' and RR' by D, so that
P = QD R, + P' = Q'D + R', RR' = Q"D Y'. +
Multiplication of the first two equations followed by replace-
ment of RR' from the third gives
PP' = (QD + R ) (Q'D + R')
= (QQ'D + QR' + Q'R) D + RR'
= (QQ'D + QR' + Q'R + Q") D + Y'.
Since I’ < D and division is unique, the remainder r in the

division of PP‘ by D is equal to r‘ as stated in choice (E).
+
25. ( D ) log2 a logzb = logz ab 2 6. Since the logarithmic function is
an increasing function, it follows that
ab 2 26.
We complete the problem with the help of the arithmetic-
geometric mean inequality (proved below) ; the geometric mean
o j two positive numbers a, b does not exceed their arithmetic
+
mean ( a b ) / 2 , and these means are equal if and only if
a = b.
In our problem,
+
and ( a b ) / L is smallest when equality holds, that is, when
a + b = 2.23 = 16.
Proof of italicized statement: Let x and y be any two real numbers.
Then ( x - y ) * = xz - 2xy + 2 0, and equality holds if
+
and only if x = y . Hence 2 yz 2 2xy, with equality if and
only if x = y. Now set x2 = a, 9 = b; then the last in-
equality is equivalent to the celebrated A M - GM inequality.
+
(a b)/2 2 da, where equality holds if and only if a = b.
Y
(0,161
26. ( B ) Adopt a rectangular coordinate system (see figure), the x-axis

being chosen along the span with the origin at its midpoint M .
Then points A and B of the arch have coordinates (-20,O)
and ( 2 0 , 0 ) , and the vertex C of the parabola is a t (0,16), so
that its equation is y = a 9 + 16. Since B(20, 0) is on it, we
have
0 = a - 2 0 2 + 16, a = -A. Thus y = -&ax’+ 16.
Five feet from the center, x = f 5 and y = -&(5)2+ 16
= 15 as stated in choice (B) .
27. (E) The speed of the particle is the piecewise constant function
V, = speed of travelling the n-th mile
=-=-
distance 1
time T,’
where T,, is the number of hours needed to traverse the n-th
mile. Since by hypothesis V,, is inversely proportional to
n - 1, its reciprocal T, is directly proportional to n - 1:
T , = K(n - 1);
and when n = 2, T2= k ( 2 - 1) = 2. Hence k = 2 and
the required time T, is 2(n - 1).
Remark: Since the time needed to traverse the n-th mile clearly
increases with n, choices (A) and (C) are immediately elimi-
nated. Putting n = 2 then eliminates (B) and (D), leaving
only (E) .
28. (E) Let the circle with radius 1 have equation x2 y2 = 1 in a+

rectangular coordinate system (see figure). Without loss of
generality, we may take A (- 1,O) and B( 1,O) as the ends of
the given diameter. The condition to be satisfied by P ( x , y) is
+
A P P P = 3, or in terms of x and y
.(C + + 93 + [ ( x - 1)* + 343 = 3.

+ +
This siniplifies to 2 ( x 2 y ) 2 = 3, or x2 3 = 4. +
Hence the required points P are all points on the circle of
radius l/G, concentric with the given circle. The number of
such points is infinite as stated in choice (E).
29. ( C ) If we divide the given expression for y by that for x, we obtain
On the other hand,

y = pl(r-1) = (pI(L-1)) t =
xt,
and after substituting for t from our result above, we obtain
y = XW/~, whence = XY.
30. (D) Place the hypotenuse of AABC on the x-axis with its center
at the origin of the x , y-plane, and denote the coordinates of
points A, B, C and P by (-u, 0), (a, 0), (0, a ) and
( x , O ) , respectively (see figure). Then the expressions for 5
and CPZ may be written
s = [x - ( - a ) ] * + [ x - u p = 2(xz + a')
cp2 = (0- x)' + ( a - O ) * . = xz + a'
SO that 2 C P = s for all positions of P on the x-axis.
This problem can also be solved by expressing C P by means
of the law of cosines applied first to A C P A , then to A C P B .
Part 4
31. (D) Denote the given mapping from the xy- to the uv-plane by an
arrow (+), and the images of 0, A , B, and C in the uv-plane
by 0‘, A’, B’, and C’ (see figure). We have, by direct substi-
tution,
O(0,O) 3 O’(0,0), A (1,O) -+ A’( 1, 0 ) ,
B ( 1 , l ) -+B’(O,2), C ( 0 , l ) -+C’(-l,O).
Segment OA from (0,O) to (1,O)
+ Segment O’A’ from (0,O) to (1,O).
4
Segment AB from (1,O) to (1, 1)
Parabolic arc A‘B‘ from (1,O) to ( 0 , 2 )
with equation u = 1 - ~ u 2 .
+I
Segment BC from (1, 1) to (0, 1)
Parabolic arc B’C’ from (0, 2) to (- 1 , O )
with equation u = fv2 - 1.
Segment CO from (0,1) to (0,O)
-,Segment C’O’ from (- 1 , O ) to ( 0 , O ) .
The transform (or image) of the square appears to be given by
choice (D) of graphs.
V
Comment: I t is noteworthy that the transformation of this problem

can be given by the single equation w = z?, where w =
u + +
iu and z = x iy (i = 47) are comples variables.
32. (c)If 24, = + + +

Ultt * * +
a&%‘, then U1 = Uo U l + * * +
&.
Hence the sum of the coefficients is u1 = 5. I t is of some in-
terest, however, to find the actual polynomial espression for un.
We write the given recursion formula

- Uk 3 + 4(k - 1)
-
Ut+l
successively for R = n - 1, n - 2, 2, 1 obtaining

a,
U, - un-1 3 + 4 ( n - 2)
u.-~ - u,-? = 3 + 4(n - 3)
..........................
..........................
,US - uz =3+ 4(1)
uz - ~1 =3f 4(0).
We add these n - 1 equations and observe that the left
members form the “telescoping sum”
21. - u,4+ Un-1- Un-2 + * * + u-2 - u1 = u, - 241,
while the sum of the right members is

3 ( n - 1) + 4[1 + 2 + . - -+ n - 21
= 3(n - 1) + 4 ( n - 2 )2( n - l ) f
= (PZ + 2 ( n - 2)]
- 1)[3
= 2nZ - 3n + 1.
Thus
un - 141 = 2n2 - 3n + 1,
and since u1 = 5, the polynomial in n expressing u, is
u, = 2n2 - 3n+ 6;
the sum of its coefficients is 2 - 3 + 6 = 5.
33. (A) Let al, az denote the first terms, (11 and (12 the common
differences of the arithmetic series with a t h sums S, and T,,,
respectively. Then
and
S, 2a1+ (n - l)dl 7n+ 1
-= =- for all tz.
Tn 2 @ + (n - l)d2 4n+ 27
t See Comment following the solution of Problem 9, p. 114.

The eleventh terms of these series are UII = + lodl and

+
UI
vll = a? lo&, respectively, and their ratio is
_ -- a 1 + lOd1 -
2411
- 2a1+ 20dl
011 a?+ 1Od2 2@+ 2 W '
We note that the last expression is precisely SJT, for IZ = 21,
so that
7(21) 1 = -148 +
=- 4
+
~ 1 1
-=
1111 4(21) 27 111 3 '
34. (B) The quotient in the division is a polynomial of degree 98 which

we denote by Q(z). Thus
%I00 = @(.) (x' - 3x + 2) + R .
Since the remainder R is of degree less than 2, we may denote it
by R(x) = a x + b. Thus
%loo = @ ( X ) (N - 2) ( x - 1) + (ax + b).
Setting first x = 2, then z = 1, we obtain
2'" = 20 +b and 1 = a + b.
Subtracting, we get 210"- 1 = a from which b = 1- u =
1 - (2'00 - 1 ) = 2 - 2!", so that
R(x) = U.U + b = (2lo0 - 1 ) +~ ( 2 - 2'O0)
= 2'oO(.x - 1 ) - ( x - 2).
35. (B) The x-coordinates of the points of intersection of the graphs of
y = x2 - 6 and y = m satisfy the equation x2 - 6 = m or
x2 = 6 + m. They are x = f46- which are real and
not zero because -6 < m < 6, the left endpoint being
-
+
L ( m ) = - d 6 m. Therefore
m m
and, when the nunierator is rationalized, this reduces to
2
= d 6 7 m +4 T m'
Hence when m approaches zero, the value of Y approaches
1970 Solutions
Part 1
1. (E) Set x = d 1 + d m ' .Since di = I, x =. .\/1+\1z,

x 2 = l + d and x ' = (x2)'= 1+2\12+2=3+2fl.
2. (A) Let s, I , and p denote the side of the square, radius of the
circle, and the common perimeter, respectively. Then p = 4s =
~ U Y ,so s = p/4, and I = p/(27r). Now let A0 and A,
denote the areas of circle and square, respectively. Then the
required ratio is
3. (C) To get y in terms of x, equate the expressions for 2' ob-

tained from the first and second given equations.
1 1
y -1= 2-P = - so that ,,1+-=-
X
x- 1 x-1 x-1'
as stated in choice (C) .
Commenl: The given equations are parametric equations of the
branch of a hyperbola which lies in the first quadrant with
horizontal and vertical as,ymptotes y = 1 and x = 1. This
is easily seen if we eliminate the parameter p by multiplying
x - 1 = 27' by y - 1 = 2-p to obtain ( x - 1) ( y - 1) =
2 O = 1 and keep in mind that x > 1 and y > 1 since 2P > 0
for all values of p.
4. (B) Three. consecutive integers can always be expressed as n - 1,

+
n, and 11 1, where n denotes the middle one. Thus each
number of the set S is of the form
(n - 1)' + n2+ (n + 1)2 = 3n* + 2.
When n is even, 3n2+ 2 is divisible by 2, so choice (A) is
false. We see that no member of S is divisible by 3, because
the remainder in that division is always 2.
To eliminate choices (C) and (D), we show: (i) when n

has a remainder of 1 upon division by 5, then 3n2 2 is +
divisible by 5; and (ii) when n has a remainder of 2 upon
+
division by 7, then 3n2 2 is divisible by 7.
(i) If n = Sm + 1, + + 1,
n2 = S2m2 2.5m
3 3 + 2 = 3*S2m2
+ 6-5m+ 5 = 5[15m2+ 6m + 11.
(ii) If n = 7m + 2, n2 = Pm2+ 2-2.7m + 4,
3n2+ 2 = + 12.7m + 14 = 7[21m2+ 12m + 23.
To show that (B) is correct, we must exhibit an n for which
3n2+ 2 is divisible by 11. This is so whenever n has the
remainder 5 upon division by 11 :
3(11m + 5)2 + 2 = 3[112m2+ 10.llm + S 2 ] + 2
= llC33m2+ 30m + 71.
Comment: The reader may wonder by what method we picked the
proper integers n. Let us analyze the reasoning. Suppose we
+
want to pick n so that 3n2 2 is divisible by the integer d.
+
Let n = Kd I , I < d. Then
3n2+ 2 = 3[kd + + 2 = 3[k2d2 + 2kdr + r2] + 2
I]'
= d[3k2d+ 6kr] + 3 9 + 2.
This expression is divisible by d if and only if 3r2 + 2 is
divisible by d. This seems not much easier to achieve than
the original task, except that we need only test integers r < d :
-3r2+
/O 2 2 5 14 29 50 77 ..'
+
This shows that 3P 2 is never divisible by 3 ; it is divisible
by 5 when I = 1 or 4, and it is divisible by 11 when I = 5 .
Such divisibility questions arise frequently and are most
efficiently handled by congruences. We recommend that stu-
dents acquaint themselves with a bit of modular arithmetic
and congruences.
5. (D) Since i = d q ,6 = -1 and i4 = ( i 2 ) 2 = (-1)2 = 1.

Therefore
i'+i2 1-1 0
j(i) = -=-=-= 0,
l+i l+i l+i
because the denominator 1 + i # 0, and the numerator is
zero. [A complex number a + bi is 0 if and only if a = 0

and 6 = 0).
6. (B) From the identity

x5+8x= x2+8x+ 16- 1 6 = ( ~ + 4 ) ' - 16,
we see that the given expression is least when the nonnegative
+
expression ( x 4)' is zero. This occurs when x = -4. Then
+ +
x2 8x is equal to (-4)* 8(-4) = -16.
Commml: Real values of x are specified, because for complex
+
values of x , the expression xz 8x may assume any value.
We suggest that the reader show:
+ +
(a) If x = u iu, u # 0, then x2 8x is real if and only
if u = -4.
(b) Given any negative number N, it is possible to choose u
so that x2+ 8x < N for x = -4 ui.+
7. (E) The quarter circle arcs AXC and B X D intersecting at X

have radii s, so that triangle ABX is equilateral. The line
through X parallel to A D meets sides A B and DC a t
right angles in points F and M , respectively. The required
distance from X to C D is M X = s - XF = s - i d be-
cause XF is the altitude of the equilateral triangle A B X with
side s and therefore has length )A. Hence M X = 3 4 2 - G).
8. (B) The exponential form of the given equations is

8" = 225, 2' = 15.
Since 8 = 2a, and since 15' = 225, we have
(21)@= 2O0 = 225 = 2*,
so that 3 a = 26 and a = 2b/3.
Since points P and Q divide A B in the ratios 2 : 3 and 3:4,

respectively (see figure), they are 3 and of the way from A
to B. So
A P = *AB, and AQ = #AB.
We are told that PQ = 2, so A B = 2 - 3 5 = 70.
The repeating decimal F may be written as the sum of .4

and an infinite geometric series with common ratio .01:
F = .4818181*.- = .4 + .0818181..*
= .4 + .081 + .00081 + - -*
= .4 + .081(1 + .01 + .ooO1 + ..*)
-4 +
1 .081 .9 53
-----*180. .4 + - .4 + __-
1 - .01
= = =
.99 11 110'
I t is now seen that when the fraction F is written in lowest
terms as above, the difference
Denominator - Numerator = 110 - 53 = 57.
Comment: The following evaluations of F, given without justifica-
tion, involve the multiplication of a series by a constant and
addition or subtraction of two series term by term.
1OOF = 48.1818... 1OF = 4.8181.*-
-F -.4818.** + F = .4818**-
99F = 47.7 11F = 5.2999-*. = 5.3
477 53 53
F=-=- F = -
990 110 110
Part 2
11. (E) Since two factors of the given cubic polynomial are known, the
third linear factor, 2 ( x - c ) , can be determined as follows:
p(x) = 2X3-hx+k=
= 2(x
2 +-x+l
[
+ 2)(x - l ) ( x - c)
2 “1
= 2[x9 - ( c - 1)x’ - (c + 21% + 261.
Since the coefficient of x2 is zero, t = 1 , so that p ( x ) =
+ +
2 [ 2 - 3 x 21 = 2[2 - ahx ik]. Hence h = 6 , k = 4,
(2h-3kI= ( 1 2 - 1 2 ( = 0 .
OR
The factor theorem states that, if x: - r is a factor of a
polynomial p ( x ) , then p ( r ) = 0. Thus
p ( - 2 ) = -16 + 2h + k = 0
p(1) = 2-l2+k = 0
and the unique solution of this linear system is h = 6, k = 4

as above.
Comment: Our first solution can be abbreviated by making use of the
relations between the roots and the coefficients of a polynomial.
Yet another solution would consist of dividing the given poly-
+
nomial by the known factors ( x 2 ) ( x - 1) and setting the
remainder equal to zero.
12. ( C ) Let Q denote the center of the given circle (see figure) and
R , S, and T its points of tangency with sides A D , A B , and
C D respectively. Points T , Q and S are collinear, and ST
is a diameter of the circle and an altitude of the rectangle of
length 2r. RQ is parallel to sides AB and C D and midway
between them. It therefore bisects every transversal, in par-
ticular the diagonal A C ; so M , the midpoint of A C , lies
on line RQ. Since M is also on the circle, R M is a diameter,
and hence RM = 2r. Since A D = 2AR, we have DC --
2RM = 4r. The area of the rectangle is basevaltitude =
4 r . 2 ~= 8r2.
13. (D) Using the definition a *

b = ab of the operation which
might be described as "exponentiation", we test each choice.
*
* *
(A) : a b = ab while b a = bo; these are generally
not equal so exponentiation is mo
tl commutative.
(B): a * ( b *
ab * c =
a * b' ub' and
c) = * b) *
=
(d)" abe are not always equal, so exponentiation
=
(u c =
is no! associative.
* * *
(C): (a b") = a(*') and ( a n) b = an 4~ b = unb are
not always equal, so choice (C) is wrong.
(D): ( a * b)" (ab)" ubn and *
= = aare al-(bn) = uh
ways equal, so (D) is a correct choice. Hence (E) is incorrect,
so that (D) is the only correct choice.
14. (A) The roots of the given equation x2 + px + q = 0, obtained

by the quadratic formula, are
I(-*+ 4-1 and H - p - -1.
The differenceof these roots is 1, so we must have d m=
1, and hence p2 - 49 = 1 so that p = d m . The nega-
tive value of the square root must be discarded because p and
q are given as positive.
OR
Call the roots Y and Y + 1. Their sum is -p = 2r + 1 and
their product is q = Y ( Y + 1). Thus
y = -
-p- 1
r + l =
-p+ 1
2 ' 2 '
~
and
q = Y(Y+ 1) =
(-P- l)(-p+ 1) - 4- 1
---
2.2 4 '
p2 = 4q+ 1, and p = d m .
15. (E) If points A and B have coordinates (2.4, YA) and (zB,~B),
then the coordinates (zc,yc) of a point C, which diwdes
the segment A B so that ACIAB = Y , satisfy
Xc- -= yc-
- = - =XA YA AC
XB - XA YB - YA AB
I,
as can be seen from the similar ring triangles with legs parallel
to the coordinate axes and hypotenuses AC and AB. Solving
these relations for xc and yc, we obtain
c: ( X C , YC) = (%A + I(xB - %A), YA + r(yB - YA)).
Now let A = (-4, 5 ) , B = ( 5 , - 1). To find the trisection
points P and Q, we use the ratios 113 and 213 respectively,
obtaining
P = ( X P , ~ P )= (-4+ $(9), 5 + $(--6)) = (-1,3)
Q = (.VO,YQ)
= (-4 + )(9), 5 + $(--6)) = (2, 1).
The difference of the y’s divided by the difference of the x’s
of any pair of points on a line is constant;t for the pair ( x , y)
and the given point (3,4) this quotient is (y - 4)/(x - 3).
Equating this quotient to its values at the trisection points P
and Q gives the equations of the required lines:
y- --4 --3 - 4 y-4 1-4
and - = -
2-3 -1-5 x-3 2-3’
which simplify to the equivalent equations
x- 4y+ 13 = 0 and 3%- y - 5 = 0.
The first of these is that given in choice (E), and no other
choice is equivalent to either of the two equations.
16. (C) Since F(1) = F(2) = F(3) = 1 we obtain F(6) by com-

puting F(4) and then F ( 5 ) , using the given recurrence rela-
tion
F ( n + 1)
F ( n ) F ( n - 1) +1 ( n 2 3).
F ( n - 2)
=
Thus
F(4) =
F(3)F(2) +1 =--
1.1 + 1 - 2,
F(1) 1
F(5) =
F(4)F(3) +1 =--
2-1 + 1 - 3,
F(2) 1
and F(6) =
F(S)F(4) + 1 --=
- 3*2+ 1
7
F(3) 1
t In rectangular coordinates, this quantity is the slope of the line.

17. (E) We shall contradict choices (A), (B), (C), and (D) to estab-
lish the correctness of choice (E) .
The given conditions pr > qr, r > 0 imply p > q, and
- p < -q, contradicting (A) ; also if p > 0, then 1 > q / p
contradicting (D). When p > q 2 0, we have q > - p ,
which contradicts (B) ; and when p is positive, q negative,
and I q I > p , then - q > p > 0 and - q / p > 1, which
contradicts (C). Therefore (E) is the correct choice.
18. (A) Denote the required difference by d ; d is positive, and

' dz= (43~)2-24.TTZT43~+(43-26)2
= 3 + 2 6 - 2432 - (26>'+ 3 - 2 6
= 6 - 2 4 9 T 8 = 6 - 2 = 4.
Therefore d = 48 = 4 3= 2.
1
Comment: An expression such as 4 3 +
2 6 may sometimes be
simplified by assuming it to have the form ( x y\4) and +
+
determining x and y so that ( x y ~ 2 =) 3~ 2 6 . t Thus +
(x2+ 2 9 ) + 2xfl = 3 +
2 6 gives 9 2f = 3 and +
2xy = 2 or y = 1/x. Eliminating y, we have 9 + 2/x2 = 3,
- + -
d 3x2 2 = 0, (2' - 1 ) (9 2 ) = 0 so that (2, y) =
( 1 , l ) or (-1, -1) or ( d , 6 / 2 ) or ( - 6 , --V2/2). The
+
first of these gives the square root x fl = 1 d. The +
solution ( 6 , 6 / 2 ) gives the same result. The negative of
-
this obtained from (- 1, 1) or ( - a , - 6 / 2 ) is discarded
because the desired square root is positive. An analogous pro-
cedure yields d3- = d - 1, so that the difference
4 3 +2 6-4 3 x = ( 1 + 6 ) - (6- 1) = 2 asbe-
fore.
19. ( C ) Denote the first term of the geometric series by a ; we are

told that
so
a = 15(1 - r) = 15 - 15r.
t The student may try to find conditions on a, b, I so that d a +b4Z can be
simplified to r y&.+
The series of squares has sum

a2 +aV + a214 + - = a2[1 + r2 + f l . - -3
*
- a2 a
--=
a
--= 45,
1 - r2 1- I 1 + I
and when 15 is substituted for a / ( 1 - I ) , we have a / ( 1 + I) =
3, so
a = 3(1 + I ) = 3 + 3 r .
Adding the equations

a = 15 - 15r and 5, = 15 + 15r
yields 6a = 30, a = 5.
20. (A) In the figure, segment BC has midpoint M , and BH, CK

are perpendicular to the line through H K , as required. Line
MP, drawn perpendicular to H K , is parallel to BH and C K ,
bisects the transversal BC, and hence bisects every transversal
including segment H K , of which it is therefore the perpendic-
ular bisector. Hence M (and every point on MP) is equi-
distant from H and X, so that we always have MH = M K
as stated in choice (A). Choices (B) and (C) contradict (A)
and hence are false. It is easy to see that (D) and (E) are
false by constructing figures which satisfy the given conditions
and violate (D) and (E).
Part 3
21. (B) The speedometer cable is so constructed that the mileage read-
ing is proportional to the number N of revolutions the wheels
make when traversing a distance D ; the distance D is the
product 2rr.N of the circumference of the wheel and the
number of revolutions it makes. Let 11, 12 be the radii of the
wheels with regular and with snow tires, respectively, and let
NI,N2 be the numbers of revolutions the wheels make going

and returning, respectively. Since the actual distances traversed
are equal, we have 2n71N1= 2n72N2 so that
_ -- -N1= -
~2 450
11 N2 440'
Since YI = 15, we have r2 = 15*$f, and
r2 - Yl = 15(% - 1) = g- .34.
22. (A) Let S,,, denote the sum of the first m positive integers. The
+
formula for S,,, is S,,, = $ m ( m 1) (see footnote on p. 114
for derivation) so that
Sa,,- S , = 33n(3n + 1) - +n(n + 1) = 4n2+ n = 150.
Thus
4n2 + tz - 150 = ( n - 6)(4n + 25) = 0, n = 6 or -9.
Since n must be a positive integer, n = 6, 4n = 24, and
Sc, = S2,= f24(24+ 1) = 12-25 = 300.
Remark: S, = $m(m + 1) has the property that SS, +1 is a
perfect square :
8S,+ 1 = 4m(m+ 1) + 1 = 4 m 2 + 4 m + 1 = (2m+ 1)*.
Among the numbers in choices (A) through (E) , only 300 has
that property; so we could have eliminated (B) through (E)
without using the hypothesis Sam- S, = 150, and without
calculating rt.
23. (D)In the base 10,
24. (B) Let s denote the length of a side of the hexagon. Since its
perimeter 65 is the same as that of the triangle, each side of
the triangle has length 2s. Now the triangle of given area 2
can be cut into four congruent equilateral triangles having

sides of length s, while the hexagon can be cut up into six such
triangles (see figure). Therefore
Area of hexagon -- Area of hexagon -
- 6- = 3-
Area of triangle 2 4 2’
so Area of hexagon = 3 . 2 = 3.
25. (E) The statement of the problem defines the function

[ x ] = greatest integer 5 x.
To calculate postage, we need the slightly different function

L(x) = least integer 1x
corresponding to “every ounce or portion thereof”. Since our
answer involves the function [ x ] , we need to express L ( x ) in
terms of [ x ] and claim that L ( x ) = -[-.I. To prove this
claim, write
x = IZ + a, n an integer, 0 5 a < 1.
Then
(n
ifa = 0
[x] = IZ, L(x) =
n+l ifaZ0.
I I
Now --N = -1c - a, so
if a = 0 n ifa=O
-[-%I=
-72
[-XI = and
-72- 1 ifa#O n + 1 ifa # O ,
and the last description agrees precisely with that of L ( x ) .
Thus the required expression for the postage is
6L(W) = -6[-W].
Questions: (i) Does [ x ] = - L ( - x ) ? (ii) For what values of

x does L ( x ) = [ x ] hold? (iii) How many different values
-
can the function L ( x ) [ x ] assume? (iv) What special
names are given to [loglo x ] and loglo x - [loglo x ] ?
26. (B) The two lines which are the graph of the first equation inter-
sect a t the point (2,3) obtained by solving the simultaneous
equations
I: x + y - 5 =0 and 11: 2 x - 3 y + 5 = 0.
Similarly, the lines
111: x - y + 1 = 0 and IV: 3 x + 2y - 12 = 0,
whose graphs constitute the graph of the second equation,
have the same point (2,3) in common. Since all four lines
have different slopes (in fact, I and I11 are perpendicular, I1
and IV are perpendicular), (2,3) is the only point common
to both graphs.
Question: If the factor ( x - y +
1) in the second equation is re-
+ +
placed by ( x y l ) , then the graphs of the two equations
have exactly two points in common. Can you explain this?
27. (A) Let the given triangle be ABC with perimeter p = A B +

+
BC C A , and denote the center and radius of its inscribed
circle by 0 and r ; see figure. Then the area of A A B C is
the sum of the areas of A A O B , ABOC, and ACOA, whose
bases are A B , BC and CA respectively, and whose altitudes
have length r. Therefore
Area of A A B C = 3rAB + 3rBC + 4rCA
= + r ( A B + BC + C A ) = f r p ,
which is given to be equal to the perimeter p of A A B C :
t r p = p . Hence r = 2.
28. (A) Let the perpendicular medians from A and B intersect each
other at 0 and their opposite sides a t their midpoints M and
N respectively, so AN = 3 and BM = +
(see figure). Let
segments A 0 and BO havelengths 2u and 2v, respectively,
so that OM and ON have lengths u and v. Then the
Pythagorean Theorem applied to right triangles AON and
BOM yields respectively
4u2 + v2 = 32 = 9 and u2 + 4v2 = (7/2)2 = 4914.
+
Four-fifths of the sum of these two equations gives 4u2 4v2 =
17 which is equal to the square of the hypotenuse A B of
right triangle AOB. Hence the length of AB is dfi.
S
29. (D) Let x denote the number of minutes after 10 o’clock now.
Let M and H (see figure) be the points on the dial to which
the minute and hour hands point 6 minutes hence and 3 min-
utes ago, respectively. If 0 is the center and T and S the
twelve and six marks on the dial, respectively, then (measured
in minute spaces) the angles TOM = x 6 and SOH =+
+
20 ( x - 3)/12. But these are equal vertical angles because
T S and HM are both straight lines through 0. The equation
x+ 6 = 20 + ( x - 3)/12 yields x = 15,
so that the time now is 10: 15.
30. (E) Let the bisector of $0 intersect A B at P (see figure). Then

the alternate interior angles A P D and PDC as well as < A D P
are equal to angle B, so that A A P D is isosceles with equal
angles at P and D. This makes AP = A D = a. Since
PBCD is a parallelogram, we have P B = DC = b ; so A B =
A P + P B = a + b.
Part 4
31. (B) Since the largest possible digit in base 10 is 9, the sum
dl + + + +
d2 da dr d, of the five digits can be at most 45.
The given sum, 43, is two less and can come about in the
following ways:
(i) One of the digits is 7 ( 2 less than 9), all others are 9;
7 may appear in 5 possible places, 79999, 97999, 99799, 99979,
99997.
(ii) Two of the digits are 8 (each one less than 9), the other
three are 9. This can happen in 5.412 = 10 ways, 88999,
89899,89989, 89998,98899,98989,98998, 99889, 99898,99988.
Next we recall that a number is divisible by 11 if and only if
the alternating sum of its digits d, - dz da - d d (1, is+ +
divisible by 11.7 We find that exactly three of the 15 numbers,
namely 97999, 99979, 98989, are divisible by 11, so the re-
quired probability is pS = +.
t This fact is based on an important property of integers: Let R be the remainder
+ + .. +
when a sum N I N2 N k is divided by D ; i.e.
Ni + Nz + ... -I-Nk = QD + R,
and let Rj be the remainder when N i is divided by D ; i.e.
N i = QiU + Rd, i = 1,2, - - *, k.
Then R is equal to the remainder in the division of the sum R, + R1+ + Rk
by D:
+ +
R1 Rz + +
RI. = PD R .
Comment: Divisibility by 11 (or any other prime integer) is handled

more simply by congruence modulo 11 (or the other prime).
Here 1W is congruent to 1 or -1 according as k is even or
odd. Hence we see that the congruence
d,+ lo& + 10% + 10% + 10% = 0 (mod 11)
reduces to dl - d2 + Q - d4 + d5 = 0 (mod 11) (this is the
divisibility criterion found before). We are given the sum
d,+d2+d3+d4+d5= 4 3 = -1 (modll).
When the preceding congruence is subtracted from this, we
+
obtain 2dZ 2d4 = - 1 3 10 (mod l l ) , so that dl d4 = 5 +
+
(mod 11). Hence ti2 d , = 16 and dz and d, must be
either 8 and 8, or 7 and 9, or 9 and 7, respectively, because
7, 8, 9 are the only allowable digits. The resulting numbers are
98989, 97999, 99979 as before.
32. (C) Let 2C denote the number of yards in the circumference of

the track, A and B the starting points, F and S the posi-
tions of first and second meeting. On first meeting, the distances
travelled by A and B, respectively, are (C - 100) and 100,
and on second meeting, (2C - 60) and (C 60). Since +
each travels a t uniform speed, the ratio of their distances is
the same for every time interval. In particular, at F and S ,
c-100 2c-60
-- -- SO C = 240.
100 C+60
Therefore the circumference 2C is 480 yards.
Now each term of our five digit number dl lodz + + --.+

10'd~ is of the form
1O*dn+,,and since 10 = 1 1 - 1, we see from the binomial expansion that
lona,,, = (11 - I)"d,+r = s.11 + (-I)"d.+1,
n = I , 2, .- -,5, so that the sum of the remainders Ri is
( - 1)Wl + (-1)Wz + ( - 1)Vs + (- 1)Vc + (- 1)'ds = dr - di + ds - dc + d ,
= P-11 R. +
33. (A) Omitting 10,OOO and including 0 momentarily, reduces the re-
quired sum of digits by 1. Multiplying each number in this
new sequence 0, 1, 2, 3, 9999 by lo-* (or any other
. * a ,
power of ten) does not change the sum of the digits, but gives
the sequence .oooO,.OO01,.OOO2, ...,
.9999 of all 1O,OO0 four
place decimal fractions. Each of the ten digits 0, 1, 2,
appears the same number of times, which is 1ooOo/10 = 1000
9 -- a,
times, in each of the four decimal places of the 10,OOO decimal

fractions, so each digit occurs 4.1000 = 4000 times in all.
The sum of all digits is therefore
4000(0 + 1 + 2 + + 9) = 4OOO(45) = 180,000.
*-.
Now adding the 1 by which we reduced the sum of digits, when

we included 0 and omitted 10,OOO momentarily, brings the
total to 180,001.
34. (C)If three integers a , b, and c have the same remainder r

upon division by an integer d, then
a = ad + r, b = /3d + r, and c = ytl + r,
where a, B, y are the quotients. The differences
a-b = (a- @)ti, a-c = (a - y ) d ,
and b- c = (a - y ) d
are exactly divisible by d. Moreover, since (a - b) -
(a - c) + ( b - c ) = 0, any common divisor d of two of
the differences is a divisor of the third. Hence the G.C.D.
(greatest common divisor) of any pair of the differences is the
greatest integer leaving the same remainder when divided into
all three of the original numbers a, b, and c.
In the present problem, we seek the G.C.D. of the two dif-
ferences
13,903 - 13,511 = 392 = 72.23
and
14,589- 13,903 = 686 = 73.2
which, by inspection, is 72.2 = 98.
Comment: The Euclidean Algorithm furnishes an automatic arith-
metic process for finding the G.C.D. of any two integers. For
an elementary discussion and proof, see for example Continued
Fructiuns by C. D. Olds, vol. 9 in this NML series, Random
House/Singer (1963),p. 17, or College Algebra by Fine.
S O L U T I O N S : 1970 E X A M I N A T I O N 1 43
35. (D) Let X denote the amount of the annual pension and y the
number of years of service. Then with constant of proportion-
ality k, the statement of the problem yields the following
three equations (and their squares below them) :
X=k&, X+p=kd*, X + q = k d g
X2 = k2y, (X + p)' = k2(y + u), (X + 9)' = k 2 ( y + b).
Replacing k2y by X 2 in the last two equations and simplify-

ing gives
2pX + p2 = k2a and 2qX + @ = k2b.
We divide the first by the second equation and solve for X:
2 P X + 4 = p_ X = 4 - bP2
2qX+ @ b' 2(bP aq) - -
We note that X is not defined when bp = uq; but then
the next to the last pair of equations multiplied by q and p ,
respectively, give
2pqX + p2q = k2aq and 2pqX + $p = k2bp,
and subtracting one from the other,
p q ( p - 4) = k2(uq - bp) = k2.0 = 0.
Hence either p q = 0, or p - q = 0. In the first case, p or
q is zero so that u or b is zero and the second or the third of
our original 3 equations would be identical with the first. In
the second case, p = q which would require a = b contrary
to hypothesis.
1971 Solutions
Part 1
1. (B) The factors 2 and 5 in the given number are the factors of 10,
and it is easy to count digits of a number when it is written as a
multiple of a power of 10. Accordingly, we write
N = 212.58= 24*28.S8= 2'(10)8 = 1 6 ~ 1 0=~ 1,6OO,OOO,OOO
and see that N is a 10 digit number.
Quesfion: Can you see that, if the base of the number system were
twenty instead of ten, the number of digits in N would be
eight? How about base SO?
2. (D) The number x of bricks laid is jointly proportional to the

number y of men and the number z of days: x = Ryz. The
given information says that when y = b and z = G, then
x = j. Hence j = kbc, which determines the constant of
proportionality R = f/(bc), so that x = fyz/(bc). When the
number 1: of bricks is b and the number y of men is c, we
get b = f z / b , from which the number of days is z = b2/f.
3. (E) The quotient of the difference of the y and x coordinates of

any two points on a line in the xy-plane remains constant.
Equating these quotients for the two pairs of points ( x , -4),
( -4,O) and (-4,0) , (0,8)gives the equation
0-(-4) -
- 8-0 4
or --= 2
-4--.2. 0-(-4)' x+4
which yields x = -6.
Commenf: When a rectangular xy-coordinate system is chosen
(as is usually the case), the above quotient is called the slope of
the line. The result however, is the same for any Cartesian
coordinate system.
4. (A) Let P, r and f denote the principal, rate of simple interest,

and time in years, respectively. Then the given information
may be written
255.31 = P + Prf = P(1 + rf),
where Prf is the interest credited. Since I = .05, and 1 = 1/6,
5
l+rf=l+--=-
605
600 600'
so - rt - 255.31.600
r = 255.31
-1 + 605 '
and
5 255.31.5 255.31
ylp = - p = -= -= 2.11,
600 605 121
so that the number of cents in the credited interest is 11.
5. (C) We have Q P x j ( a - A?)

n
and QQ = i A C . Hence
QP + +
QQ = i B D = j(42' 38') = 40'. Note that the
given data force the sum of the measures of angles P and Q to
be constant while the measure of each may vary (see figure).
1% THE M A A P R O B L E M BOOK I11
6. (E) Since by definition a *6 = 2a6 over S,

(A) >k is commutative because a *6 = 2a6 = 26a = 6 * a;
(B) * is associated because
* ( 6 * 2a(b >I<
a C) = C) = 2~(26c)= 4a6c
which equals
* 6 ) * c 2(a *
(a = 6 ) c = 2(2ab)c = 4abc;
*
(C) 3 is a left identity because 4 a = 2(&) = a for
any a in S, and moreover, 4 is a right identity because is *
commutative.
(D) Every element (I of S has a left inverse 1 = 1/4a be-
cause
*
1
4a
2
4a
* 1
1 a = - a = --.a = - = identity.
2
Moreover, since *
is commutative, a 1/4a is also the
identity 1, so 1/4a is also a right inverse of a.
(E) The element 1/2a is not an inverse for the element a of
S because
Therefore (E) is the only incorrect statement.
7. (C) The given expression

2-(2k+l) - 2-W-l) + 2-2k = 2-2k-1 - 2-2k4-1 + 2-2k
- 2-2k.2-1 - 2-2k.2 + 2-2k.1
= 2-*(’, - 2 + 1) = 2 - y . 4 )
- -2-2k2-1 = -2-2k-1 = -2-(2k++”
or choice (C) .
Remark: The fact that (C) is the only possible choice can be seen by
taking k = 0.
8. (B) The given conditional inequality is equivalent to

622+5~-4<0, (3x+ 4)(2x - 1) < 0.
Now when 3x + 4 < 0, then
so that both factors are negative. When 2x - 1 > 0, then

x > $, and 3x+ 4 > 3-(4) + 4 > 0,
so that both factors are positive. In either case, the product is
positive. But when
-+ < x < f, then 3x+ 4 >0 and 2x - 1 < 0;
+
the factors have opposite signs, so their product, 6x2 5 x - 4,
is negative. Since either x = -+ or x = 3 make the
+
product 0, the solution set consists of all values of x such that
-4 < x < as stated in choice (B).
OR
which is a parabola lying below the x-axis when -* +

One may consider the graph of the function y = 6x2 5 x - 4
< x < !j
-+
(see figure). Thus y is negative when x is in the interval
< x < 4, zero at the endpoints of that interval, and other-
wise y is positive.
tY
9. (D) Consider two circles with centers B and S and radii R and I ,
respectively ( I < R, see figure) extending to the points of
contact of their common external tangent of length 1. Draw
segment SD from S perpendicular to radius R a t D. Then
BDS is a right triangle with hypotenuse equal to the distance
BS between the centers, so that
BSL=SLP+BLP=P+(R-I)~
by the Pythagorean Theorem.
In the present problem t, R, and I are given as 24, 14, and 4
inches, respectively, so that
B 9 = 24*+ 102 = 22[122+ S2] = 22*132, BS = 2.13 = 26.
10. (E)The situation in this problem (and many others) may be

defined in terms of the presence or absence of two mutually
exclusive properties. Thus, let p and q denote the sets of all
brown-eyed girls and all brunettes, respectively, so that - p
(not p ) and -q (not q ) are the sets of blue-eyed girls and
blondes, respectively. Then there are exactly four basic inter-
sections: brown-eyed brunettes, brown-eyed blondes, blue-eyed
brunettes and blue-eyed blondes, denoted by the products
with x , y, z, w girls, respectively, in each. Since the basic sets

are disjoint, the given information is expressed by the four equa-
tions
x f y + Z + w = 50, w = 14,
x + z = 31, x + y = 18.
The required number x of brown-eyed brunettes is obtained by

adding the last three equations and subtracting the first:
x = [w+(x+z)+(x+y)J- (x+y+z+w)
= (14 + 31 + 18) - 50 = 13.
Remark: There are many other ways of solving the linear system of 4
equations in x , y, z, w for x.
OR
We may use a Venn Diagram (see figure) to represent sets

schematically by regions. Thus the whole rectangle represents
all 50 girls with p the 18 brown-eyed and hence -p the 32
blue-eyed represented by the regions to the left and right of line
CD, and with q the 31 brunettes and hence -q the 19
blondes represented above and below the line AB respectively.
Given numbers are circled, and designations on line segments
apply to the regions on both sides of them. Now the given
number 14 of blue-eyed blondes in set ( - p ) ( - q ) may be
subtracted from the total of 19 blondes in set (-4) to give 5
brown-eyed blondes in set p(-q) which may be subtracted
from the given 18 brown-eyed girls in set p to give the required
13 brown-eyed brunettes in the set pq as stated in choice (E).
Remark: Lye might equally well have proceeded counterclockwise in
the Venn Diagram, subtracting 14 from 32 to get 18, which
subtracted from 31 gives the required 13 in set pq as before.
1.50 THE M A A P R O B L E M B O O K I11
Part 2
11. (D) First, bases a and b must both exceed 7 to make either repre-
sentation 47 or 74 possible. We are told that the numbers
(47)0 = 4a+ 7 and (74)b = 7 b + 4
are equal, so that
4a + 7 = 7b + 4 or equivalently 7b - 4a = 3.
One solution of the last equation is evidently (a, b) = ( 1 , 1)
and hence all soiutions in integers are given b)* ( a , 6) =
+
(1.+ 7f, 1 4 t ) , where t may be any integer.t The value
t = 2 makes both a and b the smallest solutions greater
+ +
than 7. Therefore (a, b) = (1 7 - 2 , 1 4 . 2 ) = (15, 9)
+ +
makes the sum a b = 15 9 = 24 least. The Roman
numeral representing the number 24 is XSIV or choice ( D ) .
We check the result by noticing that
(67)Io = (47)Lj = (74)s.
12. (R) The number -Y,usually called the modulus of th,: system, is an
exact divisor of the difference of any two congruent numbers.
For if (I and b are congruent (mod A‘), then they have the
same remainder upon division by A‘ so that a = k N + I,
+
b = I N I, and a - b = ( k - 1)s.In this problem, Ar = 7
because the differences 90 - 69 = 21, 12.5 - 90 = 35, and
125 - 69 = 56 art: divisible by i and have no other common
+
divisor. Now 81 = 11.7 4 so that XI is congruent to 4
modulo 7.
Remark: “Congruence modulo i<” is an equivalence relation and
divides the set of all integers into N disjoint equivalence
classes corresponding to the N possible remainders 0, 1, 2,
--a, N - 1 in division by 11’. For I\’ = 7 (or any other
prime modulus), the system is a “field” in which division
by any integer not congruent to zero is always possible. The
quotient a/b is defined as the solution x of the congruence
bx E a(mod N). For example, the quotient 81/125 when
N = 7, is given by 125x E 8l(mod 7), which reduces to
x = 3(mod 7 ) .
t Solutions in integers of equations of the form nix - ny = c, where 111, n and c

are integers, are fully discussed in Continued Fractions, C. D.Olds, vol. 9 in this NML
series, pp. 3 6 4 2 .
13. (E) We write 1.0025 as the sun1 1 + .0025 and observe that the
terms in the binomial expansion
(1 + .0025)10= 1 + 10(.0025) + 10.9

-(.0025)*
1.2
+-11.2-3
0.9-8
(.0025)3 + *..
= + .025 + .0028125 + .000001875 + R

1
= 1.025283125 + R
decrease so rnpidl!, that only thc first four terms affect the first
five decini:tl places of the siini. To see this, we shall estimate
Hew b stands for
.(lo25 = -
1
22- 1 0 2
< 1 , and (y) =
lo!
k!(10 - k ) !
denotes the coefficient of ?I. in the espansion of (1 + .)lo.
\\'e note lirst that
(r)
and t h x l all ternis in the suin :ire positive.
R =
1=1
(lk(l)bk < DJ 5
L-4
(because 0 <b < 1)
< b4 :(b")
L-0
( w e have augmented the sum)
Hence ( 1.0025)'0 1.02528 correct to 5 decimal places. The

fifth decimal place has the digit 8.
t The symbol a; stands for u I + u2 + + u,,.

1-1
Remark: In order to evaluate a power x", it is often convenient to

write x = A +
B with B small compared to A and to
approximate
by the product
x" = ( A + B)" = A" (1 + - ')"A
first few terms of the binomial expansion of
The error committed can easily be estimated.
14. (C) We can see that 63 and 65 are both divisors by straightforward
factoring. Thus
248 -1= (224 - 1) (224 + 1) = (212 - 1 ) (212+ 1 ) (224+ 1)
= (26 - 1) (26 + 1) (2'2 + 1) (224 + 1)
= 63.65(212 + 1) (224+ 1).
15. (B) Let u and h inches, respectively, be the length of the bottom
and depth of the water when the bottom is level. The volume of
the water which is then in the form of a rectangular parallele-
piped, is lOhu cubic inches. When the aquarium is tilted, the
volume of the water which is now in the form of a prism with
altitude 10 inches and right triangular base, is
cubic inches. Equating these two expressions for the volume of

water gives
lOuh = 10.3~4, SO h = 3.
Thus the depth of the water when the bottom is level is 3 inches.
16. (A) Let the 35 scores be denoted by xl, x2, x3, ..-,X ~ J and their
average by 5. Then the average A of all 36 numbers is
and the required ratio A/Z = 1.
17. (E)Denote the center of the circular disk by 0 (see figure), and let
the radii r1, 1 2 , 1 3 , * 12" cut off equal arcs ul, a2, u3,
0 ,
proceeding counterclockwise from Y~~ which is taken, without

--
., Q",
loss of generality, as extending to the right. Let P and Q be
any interior points of radii 11 and I,, respectively. Then POQ

is a triangle whose base PQ is a segment of the secant ST
cutting arc al in S and arc a,+l in T . Segment PQ divides
each of the (72 - 1) sectors subtended by arcs m, a3, -..,
a.
into two parts. Segments P S and QT cut the two bordering
sectors subtended by arcs al and a,+l, dividing each into two
parts. In all, (n - 1) + 2 = 12 + 1 of the 2n disjoint sectors
are cut in two, giving a total of 2n + +
(n 1) = 3n 1 +
regions. This is the maximum number because the 2% radii
constitute 1~ lines which can be cut by one secant (line) in a t
most 71 points.
18. ( D ) Let 1’ denote the boat’s rate in still water in miles per hour.
Then the time (distance/rate) for 4 miles downstream and
return, which totals 1 hour, gives the equation
4 4
-+-=
0+3 0-3
1,
since the stream’s rate is 3 miles per hour. Multiplying both

+
sides by ( Y 3) ( Y - 3) and simplifying yields v2 - 8v - 9 =
(c - 9)( P + 1) = 0, v = 9 ( a = -1 must be rejected).
+ +
Now, since the rate downstream is v 3 = 9 3 = 12 and
the rate upstream is 1’ - 3 = 9 - 3 = 6, their ratio is two to
one.
19. (C) The values of x a t the points of intersection of the line and
ellipse are the solutions of the quadratic equation
x2+ 4 ( m x + l)? = 1 or (1+ 4m2)x2+ 8 m x + 3 = 0
obtained by substituting mx + 1 for y in the equation
+
x2 4y2 = 1 of the ellipse. The condition that there be exactly
one intersection implies that the quadratic equation have only

one root, which means that its discriminant is zero. That is,
( 8 m ) 2- 4(1 + 4m2)-3 = 0.
This reduces to m2 = 4 or choice (C) . Incidentally, the condi-
tion for 2 intersections or none is that the discriminant be posi-
tive or negative, respectively.
20. (E) The given equation is equivalent to + 2hx - 3 = 0. x2

Denoting its roots by I and s, we have
(X - I ) ( X - S) = x2 - (I + + = x2 + 2hx - 3 = 0,
S)X IS
so that the sum and product of the roots are

I + s = -2h, IS = -3,
respectively. If we square the first relation and substitute the

+
given value 10 for y 2 s2, and the value - 3 for IS, we obtain
(I + s ) ~= I' + 21s + = 10 + 2(-3) = 4 = (- 2h)2 = 4h2,
so that h2 = 1 and I h 1 = 1. Since 1 is not offered in choices
(A)-(D), statement (E) is correct.
Part 3
21. (C) Since, for any base b # 0, hgb N = 0 only if AT = 1, the

given equations yield
I0g3 (log4 X ) = log4 (log2 y) = log2 (log3 2) =
Moreover, since bgb M = 1, only if M = b, we have

log4 x = 3, log2 y = 4, log3 z = 2,
or equivalently
= 43, = 24, z = 32.
Adding these results gives
x + y + z = 43 + 2'+ 32 = 64 + 16 + 9 = 89.
2 2 . (A) Factoring 2 - 1 = 0, which is equivalent to 2 = 1, gives
(x-1)(x2+x+1)=0, so x - l = O o r x 2 + x + 1 = 0 .
Since w is imaginary, w - 1 # 0. Thus d w 1 = 0 + +
and hence
3 + 1 = -w, and w+ 1 = -w2.
We use these equalities to simplify the given product, getting:

(1 - w + w 2 ) ( 1 + w - w 2 ) = (-w-w)(-w~-w*)
= ( - 2 w ) ( - 2 w * ) = 4w3.
But w is a root of xl = 1, so w3 = 1, and 4 d = 4.
23. ( A ) Team A may win the series by winning both of two games, or
the last of three or four games after winning one of the others.
The six possible sequences of wins, each followed by its
probability, are A A ( $ ) , B A A ( & ) , A B A ( i ) , B B A A ( & ) ,
B A B A ( & ) , and A B B A ( & ) . Since these sequences are
mutually exclusive, the sum of their probabilities, 11/16, is
the probability of A’s winning. Now the odds that an event
with probability p occurs is defined to be the ratio p / ( 1 - p ) ;
hence the odds favoring Team A are 11 to 5.
We may check the above by computing the Complementary
odds of 5 to 11 that Team B win the series, which may be
accomplished by B winning all of three games or the last of four
games after winning two of the first three. The four possible
sequences of wins, each followed by its probability, are B B B ( i ) ,
ABBB(&), BABB(&) and BBAB(&) with total proba-
bility of 5/16 (and hence odds of 5 to 11) favoring Team B to
win the series. These odds reversed give the complementary odds
of 11 to 5 favoring Team A to win the series as found above.
24. ( D ) There are 1, 2 , 3, -..,n integers in the lst, 2nd, 3rd, ...,nth
rows and therefore a totalt of
1 +2+3+ + 12 = f n ( n + 1)
integers in the first IZ rows. Since each row, except the first
which has only one, contains two 13, the number of 1’s in the
first n rows is 21z - 1. Therefore the number of integers
which are not 1’s is
$n(. + 1) - (2n - 1) = $(n*- 3n + 2 ) .
The quotient of this number and the number 2n - 1 of 1’s is
+
$(n2- 3n 2 ) - rt2 - 3n 2 +
2n - 1 4n - 2
as stated in choice (D).
OR
t See footnote on p. 114.

156 THE hlAA PROBLEM BOOK I11
We may observe that the number of integers which are not 1’s
in the lst, 2nd, 3rd, 4th, * ..,
91th rows are 0, 0, 1, 2 , -, --
( n - 2 ) , respectively, which total
o+ o+ 1 + 2+ *-. + (It - 2) = ;(?I - 2 ) ( ? 1 - 1)
- 312 + 2 )
= 4 (?I’
so that the quotient of this total and the number 2it - 1 of 1’s
is (n2- 3 i i + 2 ) / ( 4 n - 2 ) , as before.
25. ( D ) Let b and f denote the boy’s and his father’s age, respectively.
The statements in the problem implk: lOOf b - ( f - b) +
= 4289, that is 99f +
2b = 4289, or 99f = 4257 32 - 2b, +
or
32 - 26
f - 43 = -.
99
Suppose, for the moment, that the father’s age is 43. Then
26 = 32, b -- 16, so the boy is indeed a teenager.
I f f 2. 44, then (32 - 2b)/99 2 1, so b < 0, which is
impossible.
If f 5 42, then (32 - 2b)/99 I -1, from which we get
32 - 2b 5 -99, or 26 2 131. Thus b > 65, which is
hardly teenage. So the only solution satisfying the conditions of
the problem is f = 43, b = 16. Thus f b = 59. +
Coinmenl: Solutions by means of congruences (see Solution of
Problem 12 of this 1971 Esam., p. 150) are as follows:
(a) Casting out 9’s-the equation 99f +
2b = 4289 reduces
to O + 2b = 5(niod 9 ) so that 2b = S(mod 9 ) , or b = 7
(mod 9 ) . Thus b is one o f the numbers 7 , 16, 25, 34, of --
which only 16 is in the teens.
(b) Casting out 11’s-The equation 99f 2b = 4289 +
reduces to O f 26 = lO(mod l l ) , b 5(mod I l ) , so b isone
of the numbers 5, 16, 27, of which only 16 is in the teens.
. a * ,
Substituting 16 for b in 99j+ 26 = 4289 !.ields 99f =

4257 = 99.43, and f = 43.
26. ( B ) Draw FH parallel to line .1GE (see figure). Then HE = EH

because BG = GF ant1 it line ( G E ) parallel to the base
( H F ) of a triangle ( H F H ) divides the other two sides propor-
tionally. By the same reasoning applied to triangle ARC with
line FH parallel to base .lE, we see that HC = 2BH,
because FC = 2-4F is given. Therefore EC = EH HC = +
3EH = 3 B E , and E divides side BC in the ratio 1:3.
2 7 . (E) Let r, w , and b denote the numbers of red, white, and blue
chips, respectively. We are given that $w 5 b 5 jr, and that
+
55 I w b. Now since w I 2b, 55 S 2b b = 3b. Hence +
b 2 55/3 = 184, and since b is an integer, this implies that
b 2 19. But by hypothesis, r 2 3b. Hence r 2 57.
28. (C) Let b and k denote the lengths of the base and altitude,
respectively, of the given triangle. Then the largest of the ten
parts into which the triangle is divided is a trapezoid with
altitude of length . l h and parallel bases of lengths b and .9b.
The given area of this largest part is
j ( . l h ) ( b + .9b) = .19(4bh) = 38
from which the area of the given triangle is i b h = 200.
Through the points where the parallel lines intersect each side of
the given triangle T, draw lines parallel to the other sides (see
figure). Thus the trapezoidal strip of area 38 is subdivided into
19 congruent triangles 1, so the area of each is 2. Since the sides
of T are 10 times the corresponding sides of 1,
Area T = 100 Area f = 200.
1.58 THE M A A PROBLEM BOOK I11
29. (E) The product of the first n terms of the given progression (see
footnote on p. 114) is
101/11.102/11.103/11. ... .10fl/11= 10(1+2+3+~..+fi)/11
= 10n(ti+I)/E.
This number exceeds 100,OOO = lo5 if and only if the esponent

n ( n + 1)/22 exceeds 5 , i.e. n ( n + 1 ) > 110. Since
+
n ( n 1 ) 5 110 whenever n 5 10, the required least integer
n is 11.
30. (D) Let g be the inverse of the transformation f1; then f1Cg-J =
g[fJ is the identity transformation, and g[fn+l(x)] =
g[f1( fn(x))] = I.(%). Repeated application of g k times
yields
g(g(g(...(fn+l(x))...) = g'ifn+l(x) = jn+l-k(x),
and applying g five times to the given identity j x 5 ( x ) = f5(x)
yields
gy35(x) = f30(x) = gyS(x) = 2,
so that f30 is the identity map. I t follows that
f28(x) = g2[ f30(x) I = g z ( x ) = dg(x)]-
Since
and
- x+1+2-x -
3
- 1
4-2x--2-1 3-3x 1-x'
Comment: Show that = j s by showing that f30 is the identity;
this, in turn, follows from the fact that f6 is the identity, and f6
can be computed, for esample, from f6(x) = !I[ f1( fi(x)) } =
fz f2 ( fi (4 1 1*
Part 4
31. (A) Radius OB bisects chord AC a t right angles in point G. Since

C D also makes a right angle with AC, C D )I 80.Angles ADB
and BAG are equal because they are measured by half the
equal arcs A B and BC. Hence right triangles A B D and
BGA are similar with BG/AB = A B / A D , so BG = $ and
OG = OB - BG = 2 - t = i. Since C D 11 GO, CD/GO =
A D / A O = 2, SO C D = 2.3 = 3.
OR
Denote Q A D B by a ; then Q A D C = 2a. Using right

triangles A B D and ACD, we have
1 CD
sin a = - , cos 2a = cos2a - sin2a = 1 - 2 sin2a = -
4 4 .
.'. C D = 4(1 - 2 . h ) = 4-J = $.
32. (A) In factored form

1 - $2 = (1 - x ) ( 1 + x ) ( l + x 2 ) ( 1 + 9)(l+
x a )(l + d6).
When we set x = 2-'lM, this equality becomes 1 - 2-1 =

(1 - 2-1/32)s from which s = f ( 1 - 2-l/O2)-'.
Note: In evaluating the product
+
(1 x ) ( l + x 2 ) ( 1 + x ' ) ( l + 9)(1+ x1?
we find the sum of all terms obtainable by taking one number
from each parenthesis and multiplying them together:
15 + l'x + 14x2 + lzx.9 + ...
+ ~1+2+4+...+16.
Our sum consists of terms a? for all those k which can be
written as sums of distinct powers of 2 from the zero-th to the
fourth power. By the properties of the binary system, every
integer k with 0 I k I 31 has a unique such representation.
Accordingly, the product above is equal to
1 +x+ x2+ i+ . a * + $1)
and this geometric series has the value (1 - X")/(l - x ) ,

whence 1 - Ztz = ( 1 - x ) .given product. This motivates the
above solution to some extent.
33. (B) The product P of the n quantities in G.P. which we denote by

a, ar, ur2, -
U Y " - ~ is given by?
+ a,
p = a.ay.ay?. ... = anr1+2+~~~+n-l = unr7~(n-1)/?.
The sum S of the n quantities is

s = u + ur + u r ? + *.. + UP-1 = a(1 - P ) / ( l - r ) .
The sum S' of the reciprocals of the n quantities is
a 1 - r-' - -.-.-
1 1 1 11-r-" 1 1 1 - r "
S'= -a + fly
-+ ... +7=--- a 1 -I In-1
The quotient of S and S' is the product of S and the recipro-

cal of S':
s
-=
a ( l - I n ) ar"-'(l - I)-
- u?r"-l
S' 1-r 1 - I"
If we raise this quantity to the power ./2, we obtain
34. ( B ) In a correctly running clock the minute hand moves 6' per
minute, the hour hand moves 4" per minute. Suppose both
hands coincide. After .T minutes, the hands have travelled 6x
and x/2 degrees, respectively, and will coincide again when
6r - 360" = x!2, that is, in .r = 720/11 = 6 5 8 minutes.
Thus the ratio of the time indicated by the slow clock to the
true time is
720/11
-=-=---
720 240 - 240
69 11.69 11.23 253'
When the slow clock indicates 8 hours = 480 minutes, the true
time t is obtained from
480
-=-
t
240
253'
480
1 = --.253
240
= 2.253 = 506 = 480 + 26.
Thus 26 minutes are lost in the false eight hour recording. At
time and a half of $4 per hour, i.e. a t $6 per hour or 10 cents per
minute, the extra pay should be $2.60 for 26 minutes.
t See footnote on p. 114.

OR
We note that 12 hours by the slow clock = 11-69 minutes =
12 hours + 39 minutes. Therefore 8 hours by the slow clock =
+
8 hours 26 minutes. Thus 26 minutes is lost by the slow clock
and the estra pay should be $2.60 as before.
35. (C) Let 0 denote the vertex of the given right angle (see figure), C
and C’ the centers, Y and Y’ ( I > Y’) the radii of the larger
and smaller, respectively, of any two consecutive circles in the
infinite sequence. Let CS and C’S’ be radii perpendicular to
one side of the right angle. Then OCS and OC’S’ are isosceles
right triangles, and OC = fir, OC‘ = fly’. Moreover, the
distance from 0 to the point of tangency of the two circles with
each other, is
OC - r = \/Zr - r = (fl - 1 ) r in termsof I, and
OC’ + Y’ = dr’ + I’ = (fl+ 1)r‘ in terinsof Y‘.
Equating these equal expressions we obtain

di-1
(fl+
I’
1)r‘ = (di - 1)r so that - = -- (<2 - 1)*.
r V2+1
We note in passing that (a +

- 1) and (a 1) are recipro-
cals, i.e.(a +
- 1) (a 1) = 1.
Now the ratio of the areas of two consecutive circles is the
square of the ratio of their radii:
If A is the area of the first circle, the sum of the areas of all the
following is
~ [ ( a1 ) 4 -+ (a- I)* + (a- 1)12+ ...I
(a- 1)' A
(a+ 1)' - 1 .
= A -
1 - (fl- 1)'
The required ratio of areas is

1972 Solutions
Part 1
(Longest <
- Sum of Angle
Sides Side)* 5 Squares Opposite
I 3,4,5 25 = 9+16 Right
I1 4 , 7 $ , 84 72: = 16+ 56t Right
I11 7, 24, 25 625 = 494- 576 Right
IV 3+,4+,5$ 30t < 16t + 20+ Acute
2. (B) Let C denote the present cost so that .92C is the cost a t
8% less. Since selling price is cost plus profit, selling price =
+
cost xO/,-cost. Equating the selling price with cost C a t
.yo profit and that with cost .92C at (x+ 10)yo profit
yields
C( 1 + .Oh) = .92C[1 + .01(x + lo)],
.08(.01x) = .012, x = 15.
3 . (B) Straightforward calculation yields the value of

x* -M = $(I - ifiy - $ ( I - ifl)
- 2x33) - + ( 1 - ifi) = -1.
= +(-2
The required reciprocal of x* - x is - 1 or choice (B) .
OR
We write n? - n = x(x - 1) and observe that
.v- 1 = - +)1 + i f i = -f,

164 THE MAA PROBLEM BOOK 111
where 2 is the conjugate of x. Thus

1
-(1+3)
4
= -1, so -
1
-x =
x2
-1.
4. (D) Each set X satisfying the given relation must contain the sub-
set { 1, 2 ) and also be a subset of { 1, 2 , 3 , 4 , 5 ) . These sets X
are W I , {1,2,31, { 1 , 2 , 4 ) ; ~ 5 1 w ,, 3 , 4 ) , { 1 , 2 , 3 , j i ,
( 1 , 2,4, 5 ) and { 1, 2,3, 4, 5 ) . The number of sets X is 8 or
choice (D).
OR
The sets X are each the union of the set { 1, 2 ) with a sub-
set of ( 3 , 4 , 5 ) . The number of these subsets is 8; they include
6 proper subsets, the empty set, and the entire set { 3 , 4 , 5 ) .
5. (A) First notice that 2112 > 8118 because (2'/2)8= 2' = 16 es-
ceeds (81/8)8 = 8. Also 2'l2 > 9119 because (21/2)** = 2 O=
512 exceeds (9119)18= 92 = 81. Moreover, 3''' > 2112 be-
cause (3113)6= S2 = 9 exceeds (21/2)6= 2* = 8. Now since
3113 > 2Il2 and 2112 exceeds both 8118 and 9llP, therefore 3'13
and 2112 are the greatest and the next to the greatest of the
four given numbers, in that order.
Remark: Our method of comparing a'lp and bllq has been to raise
each to the power k = least common multiple of p , q, and to
compare the resulting numbers aLlp, VIq whose exponents are
integers. Comparison of 8118 and 9119 (not needed in this prob-
lem) by this method would lead to comparing (8118)n= 8O and
(91/9)72= 98 and can be accomplished as follows.
We have 2j = 32 > 27 = 3'. Hence P5> 316. Since 22 =
4 > 3, this implies that 2n > 316, or in other words that
89 > 98.
Note: The given numbers are of the form d'", and the problem
raises the question: Where does the function j(z) = xllz in-
crease, where does it decrease as x > 0 increases? This ques-
tion can be answered by methods of calculus. First we note
that logf(x) increases or decreases according as f ( x ) in-
creases and decreases; secondly, that functions increase or
decrease according as their derivatives are positive or negative.
Thus
S O I, U ' r I 0 N S : 1 9 7 2 E X h hl I N A T I0 N 165
tl 1 > O for logex < 1, z <e

-- (log,f(s)) = g2 ( 1 - log,s)
dx < 0 for logex > 1, s > e,
where e, the base of natural logarithms, lies between 2 and 3.
Thus we see that for m > 18 2 3, miln < nl'"; but 2'1' and
W3 have to be compared because the masimum of x1lZ occurs
at 2 < x =e<R.
6. (C) Set y = .3= so that the given equation is equivalent to

y2-10~+9=0 or (y-9)(y-1)=0.
Hence y = 3' = 9 or 1, so that x = 2 or 0. Therefore
9 + 1 = 2 2 + 1 = 5 or x 2 + 1 = 02+ 1 = 1 as stated in
choice (C).
7. (E) The required ratio is
_._
"/ -- =
2zx = x*
-
yz zx yz y 3'
The given information includes the fact that yq'zx( = y/s) =
1/2, so N/'Y = 2/1 and x2/y2 = 4/1 is the required ratio.
The other piece of given information, zr/'xy = 2/3, is un-
necessary to solve the problem.
8. (D) The difference x - logy is either non-negative or negative so

that the given equation requires that either
n: - logy + logy, 2 logy = 0,
= s y = 1
or
- (x - logy) = + logy, 2.2. = 0,
2 x = 0.
\Ve can write x ( y - 1) = 0 to say that either y = 1, or

.T= 0, or both.
Coitimetit: \Ve note that the solution set of the given equation is
{(x,y) I N = 0 and y 2 1, or y = 1 and x 2 01, which
is a subset of the set of all (x, y) for which "(y - 1) = 0.
We also note that without even solving the problem one can
eliminate choices (A), (B) , and (C) , since they all imply (D) ,
and bj. assumption only one choice is correct.
9. (A) Let S and E denote the number of sheets of paper and of

envelopes, respectively, in each box. Then addition of the re-
sulting equations
S- E = 50 and E - S/3 = 50
gives $S = 100, S = 150 sheets of paper in each box.
10. (D) The value of ( x - 2) is either positive or negative, and then

the given inequality is equivalent respectively to
15x-257, 3 5 x 5 9
or
1<2-x57, -15-x<5, -55x51.
These two alternatives are equivalent to the inequalities stated
in choice (D) .
Part 2
+Y
11. ( A ) From the graph of the first equation (a circle with center at the
origin and radius 4; see figure), and that of the second equation,
a parabola, symmetric with respect to the y-axis, concave
upward and having its vertex at (0,4), it is apparent that
the only common point of both graphs is x = 0, y = 4; so
y = 4 is the only admissible value of y.
OR
We solve the equations algebraically, substituting x2 = 3y - 12
from the second equation into the first;
3y - 12 + y2 - 16 = 0,
or
+ 3y - 28 = (y - 4) ( y + 7) = 0,
y = 4 or y = -7. Clearly, y = 4, x = 0 satisfies both
given equations while y = - 7 satisfies neither no matter
what real x we take.
12. (B) Let an edge of the cube be f feet and hence 12f inches long.
Equating the number of cubic feet in the volume to the number
of square inches in all 6 faces, we get f 3 = 6(12f)2 so that
f = 6*(12)' = 864.
13. (C) Let the line through M parallel to side A B of the square
intersect sides A D and BC in points R and S, respectively;
see figure. Since M is the midpoint of A E , RM = $DE = 4
inches, and hence MS = 12 - = inches. Since PMR
and Q M S are similar right triangles, the required ratio
PM:MQ = R M : M S = 5:19
because corresponding sides of similar triangles are propor-
tional.
4
14. (B) Let s denote the length of the required side (see figure). Then
the altitude to the longest side, opposite the 30" angle, has
length 8/2 = 4 and is one leg of an isosceles right triangle
with hypotenuse s, which therefore has length 4\12.
OR
By the Law of Sines which states that the sides of any tri-
angle are proportional to the sines of the angles opposite them,
we have
--
S
--
8 8 sin 30" 8(1/2)
or s = - - --= 4\12
sin 30" sin 45" sin 45" \/2/2
as before.
168 THE hlAA PROBLEhf BOOR 111
15. (C) Let x denote the number of bricks in the wall; then x/9 and
x/10 bricks per hour would be laid by each bricklayer if he
worked alone. Working together they lay 10 fewer or (x/9) +
(x/lO) - 10 bricks per hour. Now since x bricks are laid in
5 hours, we have 5[x/9 +
(x/lO) - 101 = x , so there are
r = 900 bricks in the wall.
16. (B) Let the positive numbers be denoted by s and y with the
first three 3, :c, y and the last three .c, y , 9. Then
rj3 = yjx and y - N = 9 - y,
because the first three are in geometric and the last three in
arithmetic progression. Eliminating y from these two equa-
tions, we get
2 ~ ~ - 3 ~ - 2 7 = or0 ( 2 x - 9 ) ( ~ + 3 ) = 0 ,
x = 9/2 or -3.
Since x is required to be positive, we use x = 9/2 to find
y = 27/4, and hence the required slim is N +
y = 35/4 = l l i .
17. (E) To select a cutting point at random means that the probability
of the cutting point falling within a given interval is propor-
tional to the length of that interval. Let A B represent the
string (see figure) and let P be the point such that A P j P B =
l/x. Now if the cut lies on A P , the longer piece is at least I
times as large as the shorter. The probability of the cut being
+
on A Y is 1/( 1 x ) . Now the random cut is equally likely
to lie within the same distance from the other end B of the
string so that the probability is 2/(1 x).+
Remark: Choices ( A ) , ( B ) , and (C) are immediately eliminated by
the fact that when x = 1, the required probability is clearly 1.
18. (A) We may extend sides A D and BC of the trapezoid to meet

at V ; see figure. Then AC and B D are medians from ver-
tices A and B of A A B V meeting a t point E , which divides
the length of each in the ratio 2: 1. This means that
EC = 5AC = $'- = 3;.
OR
S 0 I, U TI 0N S : 197 2 I: X A M I N A T I 0 N 169
Since base A B is parallel to base C D , alternate interior

angles RAC and DCA are equal, so that A A B E is similar
to A C D E . We are given that base A B = 2DC so that the
corresponding sides A E and EC are such that A E = 2EC.
Moreover diagonal AC = A E EC = 2EC +
EC = 3EC = +
11. Hence EC = 11!3 = 3$.
19. (D) The kth term (1 2 2' + + + ... +

zk-') of the given se-
quence is a geometric series whose value is 2k - 1. The sum
of the first IC terms is therefore.
( 2 1 - 1) + (22- 1) + (23- + ... + ( 2 " - 1)
1)
= (21+ 22+ 23+ ... + 2") - (1 + 1 + 1 + ... + 1)
n terms
-
- (214+1 - 2) - 12 = 2rr+L -n-2 or choice ( D ) .
a2- b2
20. ( E ) The acute angle x may be taken as opposite the leg of length
2ab in a right triangle with other leg of length (a2 - b2).
Then the square of the hypotenuse h is, by the Pythagorean
Theorem,
Ir' = (2ab)? + (u' - b')? = a4 + 2a2b2+ b4 = (a2+ b2)*.
IVe now see from the figure and the definition of sine that
sin .x = 2ub/ ((12 b2). +
Remark: Since tan N = sin x/cos N = 2ab/(a2 - b 2 ) , we see that

tan x and sin x both approach 0 as b approaches 0. This
eliminates choices ( A ) , (C) and (D) .
Part 3
21. (C) Let P and Q be the intersections of A D with BF and EC,

respectively, and denote QFPQ by Q P and QEQP by
QQ. Then since the sum of the angles of quadrilateral EFPQ
is 360' (see figure) and the sum of the angles of A D P B and
AAQC are each 180°, we have the 3 equations
Q F + Q P + + Q + Q E = 360'
+ +
Q B (180' - Q P ) Q D = 180'
QC + (180" - Q Q ) + Q A = 180'.
Addition of these equations gives, after subtracting 360" from
each side, the required sum
Q A . + Q B + Q C + Q D + Q E + Q F = 3 6 0 ° = 90n"
so that n finally is equal to 4 or choice ( C ) .
OR
Starting with a rigid stick in the direction A D , we can pivot
it around A in the counterclockwise direction until it lies in
the direction of A C . We then pivot it around C to make it
lie along CE, then around E into the direction EF, then
around F into FB, then around B into BD, and finally
around D until the stick again lies along A D . I t is clear that
the stick has rotated by k. 180", where k is an integer. In this
particular example, the stick has made one full turn, so it has
rotated 360" which is precisely the sum of the measures of all
the angles of the polygon; for at each corner, the stick turned
by the measure of an angle, and it has turned all the corners.
Hence the sum of the angles is 360" = 4-90', so n = 4.
S 0 1. U T I 0 N S : I 9 7 2 E X A M I N A T I0 N 171
22. (E) For convenience, abbreviate the given root a +

bi by a and
its conjugate a - bi by 5. Satisfaction of the given equation
by each of these roots requires that
a3 + qa + r = 0 and sa+ qc + Y = 0.
Subtracting the second from the first of these equations yields
a3 - -83 + q(a - E) = 0
from which
= (2a2- 2b2) + (a2+ b2) = 3a2 - b2.

Hence y = b? - 3a2 as stated in choice (E).
OR
We may note that since the coefficient of x2 in the original
equation is zero, so also is the sum of the three roots. The
third root of the equation is therefore -2a. Now the coeffi-
cient of x is the sum of the products of the roots taken two
at a time:
+ ( - 2 a ) ( a + bi) + ( - 2 a ) ( a -
y = ( a + b i ) ( a - bi) bi)
= a? + b2 - 2a2 - 2abi - 2a2 + 2abi = b2 - 3a2,
23. (D) Let 0 denote the center of the base of the figure, P the center
of the desired circle and I its radius. We assume that the circle
passes through the points labelled A and B , and that the
center I' is on the axis of symmetry OD of the figure; this
assumption is justified in the comment below. Thus I = P A =
PH. I n right triangle PDR, PB2 = ( 2 - Of')* ($)2, and +
in right triangle POA, PA2 = O P +
1. Equating these ex-
pressions for Y? yields
172 THE M A A PROBLEhl BOOK 111
4 - 40P +O P+ = OP + 1,
so
40P = 9 and OP =
whence
OR
Draw chord A B ; its midpoint M lies on C’C, and its per-
pendicular bisector M P is the hypotenuse of right A P Q M ,
which is similar to A M C A with
r
= - MC
-PQ PQ l/4
or --
QM CA 314 -
Hence
as before.
OR
Place A‘A on the x-axis and 0 a t the origin of a rectangular
coordinate plane and determine the ordinate k ( = O P ) of P
and the radius r from the equation x2 +
( y - k)‘ = r2 of
the circle, using the fact that the coordinates of A : ( 1 , O )
and B : (i,2) satisfy it.
Comment: We suggest that the reader verify the following two facts:
I The smallest circle K having a given polygon in its in-
terior passes through some of the vertices of the polygon; niore-
over, not all the vertices lying on K lie on a minor arc of K .
I1 If the given polygon has R line of symmetry, the center
P of K lies on it.
To prove I, show that, if no vertices were on K , or if all
vertices on K were on a minor arc of K , a circle snialler than
K would contain the figure. To prove 11, use I .
It follows from I1 that in our problem, P lies on OD; and
it follows from I that 1’ lies on the segment s connecting the
midpoint of OQ to Q. For, if P were any higher, the circle
would pass either through no vertices or through A and A ‘
only, so A and A ’ would lie on ;L minor arc contradicting 1.
S O 1, CT TI 0N S : 19 7 2 E X A M I N AT I 0 N 173
If P were any lower, K would pass through either no vertices

or through R and B’ only, and these would lie on a minor
arc, again contradicting I.
For any point S on segment s, S A > SC and S B > SC.
Hence A and R are on the circle, as we surmised.
21. ( U ) All three walking rates and corresponding times yield the same
distance. If R, T denote the first rate and corresponding
time, we have the following espressions for the distance:
RT= (R+$)$T= (R-i)(T+$).
The first equality is equivalent to
R = Q ( R + i), so R = 2, and RT = 2T.
Using the last expression for the distance, we get
whence
RT 15
_- _ - RT = 15 = distance in miles.
4 4 ’
25. (C) The circle circumscribed about quadrilateral A BCD (see fig-
ure) is the circumcircle of each of the triangles B A D , BCD,
A B C , ADC. The diameter 2R of the circumcircle of any
triangle is equal to the length of any side divided by the sine
174 THE MAA PROBLEM BOOK I11
of the angle opposite that side.t Triangles B A D and BCD

have side BD in common, and angles A and C opposite
BD are opposite angles of the inscribed quadrilateral; hence
+A + +C = 180" and cosC = -cosA. By the law of
cosines,
(1) BD2 = AB' + AD2 - ~ A B - A D c oAs

= CBZ + CD2 - 2CB.CD cos C,
so
A B 2 + A D 2 - ~ A B . A D c o s A= C B 2 + C D 2 + 2 C B . C D c o s A ,
whence
+
CB2 CD2 - AB2 - AD?
(2) cos A =
+
2(CB-CD A R - A D ) '
Once cos A is determined, B D can be found from ( 1 ) and

so can 2R = BD/sin A = B D / d 1 - cos2A. This general
set-up works equally well for triangles ARC and A D C with
common side AC and leads to
cos D
+
DA2 DC2 - BA2 - BC2
(3) =
2[DA*DC+ BA-BC) '
The given lengths in our problem are such that an alert
reader may discover a shortcut due to these relationships:
BC = 39 = 3.13, CD = 52 = 4.13,
A B = 25 = 5.5, DA = 60 = 12.5;
BC2+ CD2 = 132[32 + 4'1 = 13'.S2,
AB2 + DA2 = 52[52 + 12?] = 52.132
t Let J be the other end of the diameter through C. 4 J B C is right, and 4 J =

4.4.Hence sin.4 = sin J = a / 2 R , so 2R = a/sinA. Relations 2R = b/sinB =
c/sinC are obtained analogously.
so that
BC2 + CD2 = AB2 + DA2
which means that triangles BAD and BCD are right with
common hypotenuse B D of length B D = d 5 2 . 1 3 2 = 65,
and B D = 2R is the diameter of the desired circle.
Had we inserted the given data into formula (2), we should
have found that cos A = 0 and concluded that Q A = 4 C =
90°, discovering that BD = 2R. On the other hand, using
formula ( 3 ) , we obtain
+
602 52' - 2S2 - 39' - S2(12*- 5') +
132(42- 3')
cos D =
2C60.52 +
25-39] 2.5.13C63-J
- j 2 * 7 . 1 7+ 132.7 33
= -
2 * 3 2 - 5 . 713
- 65
whence AC2 = 522 + 602 - 2-52-60cos D = 562 and
56
2 R = - AC
- - =--
56
- 65 as before.
sin D dl - (33/65)2 56/65
26. (E) Locate point N on arc A M B between M and B so that

arcs CA and MN are equal (see figure). Then arcs AM
and BN are equal, and hence segment N Q drawn I to
A B at Q is parallel and equal in length to M P . Now MN
and PQ are opposite sides of rectangle MNPQ, and each
has measure x. Hence segment P B has measure PQ QB +
= x + +
(x 1) = 2x +
1 , because QB = A P = x 1. +
176 T H E M.4A P R O B L E M B O O K 111
27. (D) The area of any triangle is equal to half the product of any
two sides and the sine of their included angle, see figure, where
p is the altitude from C, so p = AC sin A , and
Area A A B C = 3AB.p = $ A B . A C sin A .
Fromthegivendata, .\/AB.AC = 12 and AreaAABC = 64,
so A B - A C sin A = 128, A B - A C = 144,
sin 11 = # = #.
+ + +
28. (E) None of the 28 = 8 8 6 6 border squares is entirely
covered by the disc. In the 6 X 6 checkerboard formed by the
interior squares, the four corner squares are not entirely cov-
ered, because the distance from each corner of this reduced
checkerboard to the center of the disc is ~ ‘ 33*.D/8 +~ =
3\/2D/8 while the radius of the disc is D / 2 , and 1 / 2 < 3\/2/8.
The remaining 36 - 4 = 32 interior squares are entirely covered
by the disc, since they lie in a circle of radius d22 32D/8 = +
d B D / S < D / 2 about the center of the board.
OR
We take the center of the disc as origin 0 of a rectangular

coordinate system with axes along sides of checkerboard squares
and with 018 as unit of length. We shall count the number
of covered squares in the first quadrant only, see figure, and
multiply our result by 4. If we associate with each covered
square its upper right-hand vertex, we need only count these

vertices. They consist of those points (2, y) with integer
coordinates (called lattice points) for which x > 0, y > 0,
+
and x2 J.J 5 4*. These are the eight points (1, l ) , (1, 2),
(1,3), (2, l ) , (2, 2), (2,3), (3, l ) , (3, 2). Thus there are
4.8 = 32 covered squares in all four quadrants.
29. (C) Direct calculation yields the desired result
3x + x9
- 1 + 3x2 + 3x + x3
3x + x3
l+W
log 1 + 3x2 - 3%- 23
1 + 3x2
1--
=
+
(1 X I 3 l+x
log -= 3 log - = 3f(x).
(1 - x)3 1-x
30. ( A ) I n the figure, where h denotes the length of the sheet,

6 / h = cos (90' - 20) = sin 28 = 2 sin 0 cos 0
from which h = 3/(sinOcosfi). Also L / h = sec0 and there-
fore
I, = hsec0 = 3sec0/(sinBcosB) = 3sec28csc8
or choice (A).
Remark: Choices (B), (C), and (D) can all be eliminated at once
from the fact that when. the paper is square, we have 8 = 45"
and I, = 6d2.
Part 4
31. (C) We make use of the following fact: If h'1, Nz are integers
whose remainders, upon division by D , are R1 and Rz, then
the products ArlNz and RlRz have the same remainder upon
division by D. In symbols: If N1 = QlD R1 and Nz = +
+
QzD Rt, then
NiNz = (QiD + R I )(QzD + Rz)
= (QiQtD + QiRz 4-Q z R I ) D+ RiRz,
and the last expression clearly has the same remainder as RlR2.
Among the first few powers of two, we find that 26 has the
convenient remainder 12 [or -11 upon division by 13, so
212 = 26.26 has the same remainder as 12-12 [or (-1)",
namely 1 . We now write 2'Oo0 = (212)83.24,and conclude
that the remainder upon division by 13 is (1)s3.3 = 3 , since
2' = 16 = 1 -1 3 3. +
Using the notation of congruences, we have
26 = 64 = - 1 (mod 13),
2lm = (26)166.24= ( - 1 ) ' M . 1 6 (mod 13)
E 1 . 3 (mod 13) 3 (mod 13).
A B
32. (B) Since point E divides every chord through E into two seg-
ments whose product is constant,
CE.ED = A E - E B or CE.3 = 2.6, so CE = 4.
Thus chords A B and C D have lengths 8 and 7, respectivel)..
Now the center 0 of the circle lies a t the intersection of the

perpendicular bisectors of chords CD and A B which is f unit
above and 4 units to the right of point A . The radius OA is
the hypotenuse of a right triangle with legs A M = 4 and
OM = f ;
OA2 = A M Z + OMZ = 42 + ( f ) 2 = 65/4.
Thus the length of the diameter is 20.4 = 2465/4 = 4s.
33. ( C )Let the units, tens, and hundreds digits of the required number
be denoted by U , T , and H respectively. Then the quotient
to be minimized is
L~+lOT+lOOH - U + T + H + 9 T + 9 9 H
l;+T+H U+T+H
+
9(T 11H)
= '+ U+T+H*
If ti < T I the quotient can be made smaller by interchanging
T and U . Hence for the minimum, we have U > T . Simi-
larly U > H , from which we conclude that neither T nor
H is equal to 9. Hence our quotient is least when U = 9
regardless of the values of T and H . T o minimize the quo-
tient, we minimize the fraction
9(T + 11HI
or, equivalently, one-ninth of it:
T + 11H - 10H - 9
T+H+9 -'+T+H+9'
This is least when T is largest, and since T # 9 , we take T
to be 8 . Now ( 1 0 H - 9 ) / ( H +
1 7 ) is smallest when H is
smallest, i.e. when H = 1 . Hence the required number is
189 and the value of the required minimum quotient is
+ +
189/(1 8 9 ) = 189/18 = 10.5.
34. ( A ) Let T , D and H represent the ages of Tom, Dick and Harry,
respectively. We are told that
304- T = 2H and 2H3 = 3034- T3
or, equivalently, that
2 ( H - D) = D + T and 2 ( H 3 - 03) = 03+ T3.
The last equation, in factored form, is

+ DH + 03) = ( D + T )(D? - DT + Iz).
2 ( H - D ) ( H2
Dividing both sides by the equal numbers 2 ( H - D ) = D + T
yields
H2 + DN + 0’= 02 - DT + Iz
which is equivalent to
P - P = D ( H + T ) or ( T + H ) ( T - H ) = D ( T + H ) ,
so that D = T - H , or T = D 4- H . By the first equation,
T = 2H - 3 0 , so H = 4 0 . Since H , D are relatively
prime, D = 1 and H = 4 . Then T = D H = 1 4 = 5 + +
and
Iz+ D z + H 2 = 5’+ 1 2 + 4’ = 25 + 1 + 16 42.
=
35. (D) We first show that the triangle has to travel around the square
3 times before its vertices are again in their initial positions.
The triangle first rotates about midpoint B of the square
making # of a revolution, then about corner X making of
a revolution, and so on along each side of the square. When,
after 8 moves, it reappears on side A X , it has made 4 - 4 ,+
4 . h = 8 of one revolution; but in order to be in its original
position, with P above side A B , it must make an integer
number of revolutions, and this happens after 3 such cycles,
that is, after 8.3 = 24 moves. In 9 = 8 moves, the rota-
tion is about P so P traverses no path; while in the remain-
ing 16 moves, P traverses f of the circumference of a circle
of radius A P = 2 in 8 of the moves, of this circumference
in the other 8 moves. Hence the total length of P’s path is
(+ + &)4r = Y-Kinches.
Comment: In the original statement of this problem, the last clause

of the second sentence read
“until P returns to its original position”.
Many alert solvers noticed that P’s path goes through P’s
original position during the 9th move of the triangle (see
figure); that is, long before the entire triangle returns to its
SOLUTIONS: 1972 BSAMINATION 181
original position. They found the length of this shorter path

to be 1 6 ~ 1 3 failed
, to find this result among the choices offered,
and raised their objections. Evidently the poser of the problem
considered positions of Y ujter each move, rather than during
a move, and was not aware of other interpretations. The objec-
tions to the problem were justified, and the members of the
Committee on High School Contests are grateful to the many
people who called the ambiguity to their attention.
IV
Classification of Problems
To classify these problems is not a simple task; their content is so

varied and their solution-possibilities so diverse that it is difficult to
pigeonhole them into a few categories. Moreover, no matter which head-
ings are selected, there are borderline cases that need cross-indexing.
Nevertheless, the following may be helpful to the reader who wishes to
select a particular category of problems.
The number preceding the semicolon refers to the last two digits of
the examination year, and the numbers following the semicolon refer to
the problems in that examination. For example, 69; 13 means Problem
13 in the 1969 examination.
Algebra
Binomial expansion 69; 16 71; 13, 24

Equations
Cubic 67;35 70; 11 72;22
Diophantine 66; 26 67; 15, 24 68; 19 69; 19 71; 25
72; 33
Fractional 66;5,33 69; 1
Linear 66; 26 67; 19 68;9, 34
Parametric 69; 29 70;3
Polynomial 66; 10, 30 71; 22 72; 22
Quadratic 66; 3, 17, 23 67; 17 68; 9, 13, 14 69; 5, 7
70; 14 71; 20 72; 16
Systems of 66; 17, 22, 26 68; 14,34 69; 35 70; 3, 35
71; 19 72; 9, 11,34
183
184 T H E M.4.4 P R O B L E M B O O K I11
Exponents 66; 12, 16 67; 4 71; 7, 29 72; 6

Functions
Composite 71;30 72; 29
Defined on integers 66;25 68; 27 69;32 70; 16
Exponential 66; 12, 16 67; 4, 6, 26 68; 2 69; 17, 29
70;8 71; 21 i2;5, 6
Inverse 67; 14
Linear 66; 7 68; 1
Logarithmic 66; 24 67; 4, 23, 26 68; 23, 29 69; 17, 25
70;8 il;21 72;5,8
Piecewise constant 69; 27 70; 25
Polynomial 68; 5 69; 32,34 70; 11 71; 32
Quadratic 67; 11, 18 69; 12, 26 70; 6 71; 8, 18
72; 11
Range of 68; 17
Rational 66; 5,33 67; 14
Trigonometric 72; 20,30
Graphs 6 i ; 11, 18 68; 16 69; 7, 21 70; 26 71; 8
72; 11 (see also Coordinate Geometry)
Identities 66; 7, 36 67; 2, 10
Inequalities 66; 15, 23, 35 67; 7, 18 68; 9, 16, 22, 29

69; 14, 25 70; 17 71; 8, 27, 29 72; 5, 10
Logarithms 66; 9, 24 67; 4, 26
Mixture problems 67; 8
Means
Arithmetic 66;3 67; 11 68; 17, 28 69;9, 25 71; 16
Geometric 66; 3 67; 11 68; 28 69; 25 72; 27
Numbers
Complex 66; 3, 10, 22 70; 5 71; 22 72; 3, 22
Rational 66; 13
Operations 68;4 69;4 70; 13 71;6
Optimization 67; 11,30 69; 11 70;6 71;27 72;33

Progression
Arithmetic 66; 18, 19, 25 67; 39 68; 20, 26, 27
69; 9,33 70;22 71; 24 72; 16
Geometric 66; 39 67; 20,36 70; 10, 19 71; 33,35
72; 16, 19
Proportion 66; 1, 28 68; 8 69; 1, 2, 33 70; 35
71; 18, 34 72; 7
C I. A S S I F I C A T I0 N 0 I; P R O B L E M S 185
Quadratic see under Equations and Functions

Radicals 70; 1, 18
Rate, Time, Distance 66; 27,34 68; 25,32 69; 27 70; 21,32
Problems 71; 2, 16, 18 72; 24
Ratio see Proportion
Recursion Relations 66; 25 69;32 70; 16
Roots (of Equations) 66; 3, 10,30 67; 17,35 68; 13, 14
70; 11, 14 71; 20, 22 72; 22
Sequence see Progression and Functions, Defined on
integers
Series, geometric see Progression, Geometric
Transformations
Fractional 71; 30
of the Plane 69; 31
Work Problems 66;37 71; 2 72; 15
Arithmetic
Approximation 68; 28 71; 13

Percent 66;2 67;8 68;8,31 71;4 72; 2
Powers 70; 1
Profit and Loss 67;30 69;2 72;2
Proport ion see under Algebra
Rate, time, distance 66; 27, 34 67; 27 (see also under Algebra)
Geometry
Angles 66; 21,31 68; 6, 18, 20 69; 8 70; 29, 30

71; 5,34 72; 14, 21
Area 66; 2, 4, 14, 32, 38 67; 3, 5, 9, 12, 15, 19,
20, 33, 34, 40 68; 11, 12, 24, 31, 35
69; 13, 15, 22 70; 2, 12, 24, 27 71; 2a,35
72; 12, 27
186 THE MAA PROBLEM B O O K 111
Circles 66; 4, 8, 31 67; 29,32, 33 68; 1, 11, 12

69; 6, 8, 10, 13, 15 70; 7, 12, 21, 27, 32
71; 9, 17, 31 72; 23, 25, 26, 28, 32, 35
Constructions 67; 13 68;7, 22, 30 70; 20
Coordinate Geometry 66; 13, 17, 26 67; 11, 12, 40 68; 3
69; 11, 18, 21, 22, 26, 28, 30, 31, 35
70; 15, 26 71; 3, 19 72; 11, 23, 28
Inequalities 66; 35
Optimization 67; 11 71; 17
Polygons 66; 21 68; 20 72; 20
Proportion 66; 4, 11, 28, 40 67; 15, 21, 34, 37
68; 11, 18, 24,34 69; 13 70; 2 71; 26, 28
72; 13, 18
Pythagorean Formula 66; 6, 8 67;40 68; 12 69;30 70; 28
71; 9 72; 1, 20, 25, 32
Ratio see Proportion
Similarity see Proportion
Volumes 71; 15 72; 12
Miscellaneous
Combinatorics 68; 30 70; 33 71; 17, 23, 24

Number Theory
Diophantine Equa- See under Algebra, Equations
tions
Divisibility 66; 29 67; 1, 22, 25, 31 68; 15, 23, 24
70; 4, 31,34 71; 12, 14, 25 72; 31,34
Number Bases 66; 39 67; 16 68; 21, 26, 33 69; 3, 20
70; 23 71; 1, 11, 32
Logic 66; 20 67; 28, 38 68; 10
Probability 70;31 71; 23 72; 17
Sets 67; 28,38,39 68; 10 71; 10 72; 4

Contestproblembook3 (1966-1972)

Uploaded by

Copyright:

Available Formats

Contestproblembook3 (1966-1972)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Contestproblembook3 (1966-1972)

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics are covered in the book?

What topics are covered in the book?

THE CONTEST

PROBLEM BOOK I11

©Copyright 1973 by The Mathematical Association of America (Inc.)

compiled and with solutions by

Suggestions foi Using This Book 5

XV Chssikation of Problems 183

MAA Service Center

of 20, 10, and 10 questions weighted 3, 4, and 5. On the 1968-1972 ex-

1. Given that the ratio of 3x - 4 to y +

3. If the arithmetic mean of two numbers is 6 and their geometric

4. Circle I is circumscribed about a given square and circle I1 is in-

5. The number of values of x satisfying the equation

6. A B is a diameter of a circle centered at 0. C is a point on the circle

8. The length of the common chord of two intersecting circles is 16 feet.

12. The number of real values of x that satisfy the equation

13. The number of points with positive rational coordinates selected

15. If x -y > x and x + y < y, then

17. The number of distinct points common to the curves x2 + 4yz 1

21. An “n-pointed star” is formed as follows: the sides of a convex

all n values of k, sides k and k 2 are non-parallel, sides

Let S be the degree-sum of the interior angles at the n points of

22. Consider the statements: (I) d m = 0 (11) d m = ab

23. If x is real and 4y2 4xy x+ + +

24. If lognf A’ = logx M , M # !Y, MAr> 0, M # 1, N # 1, then

25. If F ( u + 1) = 2F(lL) +1 for It = 1,2, and F(1) = 2,

28. Five points 0, A , U , C, D are taken in order on a straight line

30. If three of the roots of x' + ax2 + bx + c = 0 are 1, 2, and 3, then

32. Let Y be the midpoint of side A B of triangle ABC. Let P be a

33. If ab # 0 and I a I # I b I the number of distinct values of x

3S.t Let 0 be an interior point of triangle ABC, and let s1= OA +

t The five choices have been slightly altered by the editors.

39. In base R1 the expanded fraction FI becomes .373737. * ., and the

40. In this figure AB is a diameter of a circle, centered a t 0, with

2. An equivalent of the expression

3. The side of an equilateral triangle is s. A circle is inscribed in the

4. Given (log u ) / p = (log b ) / q = (log c ) / r = log x, all logarithms to

5 . A triangle is circumscribed about a circle of radius r inches. If the

7 . If a/b < - c / d where a, b, c, d are real numbers and bd # 0,

8. To m ounces of an m% solution of acid, x ounces of water are

(E) not determined by the given information

9. Let K, in square units, be the area of a trapezoid such that the

11. If the perimeter of rectangle ABCD is 20 inches, the least value of

13. A triangle ABC is to be constructed given side a (opposite angle

17. If 11 and 12 are the distinct real roots of a? + pz + 8 = 0, then

18. If x2 - 5 x + 6 < 0 and P = 22 + 51 + 6 then

19. The area of a rectangle remains unchanged when it is made 2 i

20. A circle is inscribed in a square of side m, then a square is inscribed

21. In right triangle ABC the hypotenuse A B = 5 and leg A C = 3.

22. For natural numbers, when P is divided by D , the quotient is

24. The number of solution-pairs in positive integers of the equation

25. For every odd number p >1 we have:

other in 4 hours. At what time P.M. should the candles be lighted so

29. A B is a diameter of a circle. Tangents A D and BC are drawn so

30. A dealer bought 11 radios for d dollars, d a positive integer. He

32. I n quadrilateral ARCD with diagonals AC and BD intersecting at