18.

S34 (FALL, 2004)

PIGEONHOLE PROBLEMS

Note: Notation such as (78P) means a problem from the 1978 Putnam
Exam.
1. (78P) Let A be any set of 20 distinct integers chosen from the arith­
metic progression 1, 4, 7, . . . , 100. Prove that there must be two distinct
integers in A whose sum is 104. [Actually, 20 can be replaced by 19.]
2. Five points are situated inside an equilateral triangle whose side has
length one unit. Show that two of them may be chosen which are less
than one half unit apart. What if the equilateral triangle is replaced
by a square whose side has length of one unit?
3. (71P) Let there be given nine lattice points (points with integral co­
ordinates) in three dimensional Euclidean space. Show that there is
a lattice point on the interior of one of the line segments joining two
of these points. (Warning. It is easy to misread this problem. Take
time to make sure you understand what is being asked for.) [To test
your understanding, how many lattice points does one need in four
dimensions to reach the same conclusion?]
4. (72IMO) Prove that from a set of ten distinct two-digit numbers (in
the decimal system), it is possible to select two disjoint subsets whose
members have the same sum. [Though not stated in the problem, one
should assume that not both the subsets are empty, or even that neither
of the subsets is empty.]
5. (80P) Let A1 , A2 , . . . , A1066 be subsets of a finite set X such that |Ai | >
1
2
|X | for 1 � i � 1066. Prove that there exist ten elements x1 , . . . , x10
of X such that every Ai contains at least one of x1 , . . . , x10 . (Here |S |
means the number of elements in the set S.)
6. Given any n + 2 integers, show that there exist two of them whose sum,
or else whose difference, is divisible by 2n.
7. Given any n + 1 distinct integers between 1 and 2n, show that two
of them are relatively prime. Is this result best possible, i.e., is the
conclusion still true for n integers between 1 and 2n?

1
 8. Given any n + 1 integers between 1 and 2n, show that one of them is
divisible by another. Is this best possible, i.e., is the conclusion still
true for n integers between 1 and 2n?
9. Given any 2n integers, show that there are n of them whose sum is
divisible by n. (Though superficially similar to some other pigeonhole
problems, this problem is much more difficult and does not really in­
volve the pigeonhole principle.)
10. Let u be an irrational real number. Let S be the set of all real numbers
of the form a + bu, where a and b are integers. Show that S is dense in
the real numbers, i.e., for any real number x and any � > 0, there is a
element y � S such that |x − y | < �. (Hint. First let x = 0.)
11. Two disks, one smaller than the other, are each divided into 200 congru­
ent sectors. In the larger disk 100 of the sectors are chosen arbitrarily
and painted red; the other 100 sectors are painted blue. In the smaller
disk each sector is painted either red or blue with no stipulation on
the number of red and blue sectors. The small disk is then placed on
the larger disk so that their centers coincide. Show that it is possible
to align the two disks so that the number of sectors of the small disk
whose color matches the corresponding sector of the large disk is at
least 100.
12. A collection of subsets of {1, 2, . . . , n} has the property that each pair
of subsets has at least one element in common. Prove that there are at
most 2n−1 subsets in the collection.
13. (95P) For a partition � of {1, 2, 3, 4, 5, 6, 7, 8, 9}, let �(x) be the number
of elements in the part containing x. Prove that for any two partitions �
and � � , there are two distinct numbers x and y in {1, 2, 3, 4, 5, 6, 7, 8, 9}
such that �(x) = �(y) and � � (x) = � � (y).
[A partition of a set S is a collection of (nonempty) disjoint subsets
(parts) whose union is S.]
14. (a) Let p be a prime of the form 8k +1. Prove that there exist positive
integers a, b, m with m < 2p such that mp = a4 + b4 .
Hint. Show that there is an integer x such that x4 + 1 is divisible
�
by p, and consider the numbers u + vx, where 0 � u � ≤ p∈ and
�
0 � v � ≤ p∈.

2
 Note. This is a minor variation of a standard application of the
pigeonhole principle going back to Fermat. Do not hand in this
problem if you’ve seen it or something similar before.
(b) Improve the bound m < 2p. In particular, find a constant c < 2
such that one can take m < cp for p large. (The best possible value
of c requires some sophisticated number theory not involving the
pigeonhole principle.)

15. N is the set of nonnegative integers. For any subset S of N, let P (S)
be the set of all pairs of members of S. (A pair is a set (unordered)
with two distinct members.) Partition P (N), arbitrarily, into two sets
(of pairs) P1 and P2 . Prove that N must contain an infinite subset S
such that either P (S) is contained in P1 or P (S) is contained in P2 .

3
 18.S34 (FALL, 2004)
PROBLEMS ON PROBABILITY

1. Three closed boxes lie on a table. One box (you don’t know which)
contains a $1000 bill. The others are empty. After paying an entry fee,
you play the following game with the owner of the boxes: you point to a
box but do not open it; the owner then opens one of the two remaining
boxes and shows you that it is empty; you may now open either the
box you first pointed to or else the other unopened box, but not both.
If you find the $1000, you get to keep it. Does it make any difference
which box you choose? What is a fair entry fee for this game?

2. You are dealt two cards face down from a shuffled deck of 8 cards
consisting of the four queens and four kings from a standard bridge
deck. The dealer looks at both of your two cards (without showing
them to you) and tells you (truthfully) that at least one card is a
queen. What is the probability that you have been given two queens?
What is this probability if the dealer tells you instead that at least
one card is a red queen? What is this probability if the dealer tells
you instead that at least one card (or exactly one card) is the queen of
hearts?

3. An unfair coin (probability p of showing heads) is tossed n times. What

is the probability that the number of heads will be even?

4. Two persons agreed to meet in a definite place between noon and one
o’clock. If either person arrives while the other is not present, he or
she will wait for up to 15 minutes. Calculate the probability that the
meeting will occur, assuming that the arrival times are independent
and uniformly distributed between noon and one o’clock.

5. (58P) Real numbers are chosen at random from the interval [0, 1]. If
after choosing the nth number the sum of the numbers so chosen first
exceeds 1, show that the expected or average value for n is e.

6. (61P) Let � and � be given positive real numbers with � < �. If two
points are selected at random from a straight line segment of length �,
what is the probability that the distance between them is at least �?

1
 7. (93P) Two real numbers x and y are chosen at random in the interval
(0, 1) with respect to the uniform distribution. What is the probability
that the closest integer to x/y is even? Express the answer in the form
r + s�, where r and s are rational numbers.
8. (92P) Four points are chosen at random on the surface of a sphere.
What is the probability that the center of the sphere lies inside the
tetrahedron whose vertices are at the four points? (It is understood that
each point is independently chosen relative to a uniform distribution
on the sphere.)
9. (89P) Let (x1 , x2 , . . . , xn ) be a point chosen at random from the n-
dimensional region defined by 0 < x1 < x2 < · · · < xn < 1. Let f be a
continuous function on [0, 1] with f (1) = 0. Set x0 = 0 and xn+1 = 1.
Show that the expected value of the Riemann sum
�
n
(xi+1 − xi )f (xi+1 )
i=0

1
is 0 f (t) P (t) dt, where P is a polynomial of degree n, independent of
f , with 0 � P (t) � 1 for 0 � t � 1.
10. (89P) A dart, thrown at random, hits a square target. Assuming any
two parts of the target of equal area are equally likely to be hit, find
the probability that the point hit is nearer
� to the center than to any
edge. Express your answer in the form (a b + c)/d, where a, b, c, d are
integers.
11. (89P) If � is an irrational number, 0 < � < 1, is there a finite game
with an honest coin such that the probability of one player winning
the game is �? (An honest coin is one for which the probability of
heads and the probability of tails are both 1/2. A game is finite if,
with probability 1, it must end in a finite number of moves.)
12. (85P) Let C be the unit circle x2 +y 2 = 1. A point p is chosen randomly
on the circumference C and another point q is chosen randomly from
the interior of C (these points are chosen independently and uniformly
over their domains). Let R be the rectangle with sides parallel to the
x- and y-axes with diagonal pq. What is the probability that no point
of R lies outside of C?

2
13. (82P) Let pn be the probability that c + d is a perfect square when
the integers c and d are selected independently
� at random from the set
{1, 2, . . . , n}. Show
� that lim (p
n� n n) exists, and express this limit
in the form r( s − t) where s and t are integers and r is a rational
number.

14. (68P) The temperatures in Chicago and Detroit are x� and y � , respec­
tively. These temperatures are not assumed to be independent; namely,
we are given:

(i) P (x� = 70� ), the probability that the temperature in Chicago is

70� ,
(ii) P (y � = 70� ), and
(iii) P (max(x� , y � ) = 70� ).

Determine P (min(x� , y � ) = 70� ).

15. In the Massachusetts MEGABUCKS lottery, six distinct integers from

1 to 36 are selected each week. Great care is exercised to insure that
the selection is completely random. If Nmax denotes the largest of the
six numbers, find the expected value for Nmax .

16. (a) A fair die is tossed repeatedly. Let pn be the probability that after
some number of tosses the sum of the numbers that have appeared
is n. (For instance, p1 = 1/6 and p2 = 7/36.) Find limn� pn .
(b) More generally, suppose that a “die” has infinitely many faces
marked 1, 2, . . .. When�the die is thrown, the probability is ai
that face i appears (so  i=1 ai = 1). Let pn be as in (a), and find
limn� pn .

3
 18.S34 (FALL 2004)
PROBLEMS ON CONGRUENCES AND DIVISIBILITY

1. (55P) Do there exist 1, 000, 000 consecutive integers each of which contains a
repeated prime factor?

2. A positive integer n is powerful if for every prime p dividing n, we have that p2

divides n. Show that for any k � 1 there exist k consecutive integers, none of
which is powerful.

3. Show that for any k � 1 there exist k consecutive positive integers, none of
which is a sum of two squares. (You may use the fact that a positive integer n
is a sum of two squares if and only if for every prime p � 3 (mod 4), the largest
power of p dividing n is an even power of p.)

4. (56P) Prove that every positive integer has a multiple whose decimal represen­
tation involves all ten digits.

5. (66P) Prove that among any ten consecutive integers at least one is relatively
prime to each of the others.

6. (70P) Find the length of the longest sequence of equal nonzero digits in which
an integral square can terminate (in base 10), and find the smallest square which
terminates in such a sequence.

7. (72P) Show that if n is an integer greater than 1, then n does not divide 2n − 1.
� � � a�
8. (a) (77P) Prove that pa pb
� b (mod p) for all integers p, a, and b with p a
prime, p > 0, and a � b � 0.
(b) (not on Putnam exam) Show in fact that the above congruence holds
modulo p2 .
(c) (not on Putnam exam) Show that if p � 5, then the above congruence
even holds modulo p3 .

9. (82P) Let n1 , n2 , . . . , ns be distinct integers such that

(n1 + k)(n2 + k) · · · (ns + k)

is an integral multiple of n1 n2 · · · ns for every integer k. For each of the following

assertions, give a proof or a counterexample:

(a) |ni | = 1 for some i.

 (b) If further all ni are positive, then

{n1 , n2 , . . . , ns } = {1, 2, . . . , s}.

10. (83P) Let p be in the set {3, 5, 7, 11, . . .} of odd primes, and let

F (n) = 1 + 2n + 3n2 + · · · + (p − 1)np−2 .

Prove that if a and b are distinct integers in {0, 1, 2, . . . , p − 1} then F (a) and
F (b) are not congruent modulo p, that is, F (a) − F (b) is not exactly divisible
by p.

11. (85P) Define a sequence {ai } by a1 = 3 and ai+1 = 3ai for i � 1. Which integers
between 00 and 99 inclusive occur as the last two digits in the decimal expansion
of infinitely many ai ?

12. (86P) What is the units (i.e., rightmost) digit of

� �
1020000
?
10100 + 3

Here [x] is the greatest integer � x.

13. (91P) Suppose p is an odd prime. Prove that

p � �� �
p p+j
� 2p + 1 (mod p2 ).
j=0
j j

14. (96P) If p is a prime number greater than 3 and k = ≤2p/3≥, prove that the
sum � � � � � �
p p p
+ +···+
1 2 k
of binomial coefficients is divisible by p2 .

15. (97P) Prove that for n � 2,

n terms n−1 terms
���� ����
···2 ···2
22 � 22 (mod n).

16. (99P) The sequence (an )n�1 is defined by a1 = 1, a2 = 2, a3 = 24, and, for
n � 4,
6a2 an−3 − 8an−1 a2n−2
an = n−1 .
an−2 an−3
Show that, for all n, an is an integer multiple of n.
17. (00P) Prove that the expression
� �
gcd(m, n) n
n m

is an integer for all pairs of integers n � m � 1.

18. How many coefficients of the polynomial

�
Pn (x1 , . . . , xn ) = (xi + xj )
1�i<j�n

are odd?

19. Define a0 = a1 = a2 = a3 = 1,

an+4 an = an+3 an+1 + a2n+2 , n � 0.

Is an is an integer for all n � 0?

20. Define a0 = a1 = 1 and

n−1
1
an = a2i , n > 1.
n−1 i=0

Is an an integer for all n � 0?

21. Let f (x) = a0 + a1 x + · · · be a power series with integer coefficients, with a0 ∈= 0.

Suppose that the power series expansion of f � (x)/f (x) at x = 0 also has integer
coefficients. Prove or disprove that a0 |an for all n � 0.

22. Suppose that f (x) and g(x) are polynomials taking integers to integers such
that for all n ≡ Z, either f (n) = 0 or f (n)|g(n). Show that f (x)|g(x), i.e., there
is a polynomial h(x) with rational coefficients such that g(x) = f (x)h(x).
 18.S34 (FALL 2004)
PROBLEMS ON RECURRENCES
1. (90P) Let T0 = 2, T1 = 3, T2 = 6, and for n � 3,

Tn = (n + 4)Tn−1 − 4nTn−2 + (4n − 8)Tn−3 .

The first few terms are: 2, 3, 6, 14, 40, 152, 784, 5158, 40576, 363392. Find,
with proof, a formula for Tn of the form Tn = An + Bn , where {An } and {Bn }
are well-known sequences.

2. (80P) For which real numbers a does the sequence defined by the initial con­
dition u0 = a and the recursion un+1 = 2un − n2 have un > 0 for all n � 0?
(Express the answer in simplest form.)

3. (83P) Prove or disprove that there exists a positive real number u such that
[un ] − n is an even integer for all positive integers n. (Here, [x] is the greatest
integer � x.)

4. Define un by u0 = 0, u1 = 4, and un+2 = 56 un+1 − un . Show that |un | � 5 for all

n. (In fact, |un | < 5 for all n. Can you show this?)
�
5. (46P) Show that the next integer above ( 3 + 1)2n is divisible by 2n+1 .

6. Let a0 = 0, a1 = 1, and for n � 2 let an = 17an−1 − 70an−2 . For n > 6, show

that the first (most significant) digit of an (when written in base 10) is a 3.

7. Let a, b, c denote the (real) roots of the polynomial P (t) = t3 − 3t2 − t + 1. If

un = an + bn + cn , what linear recursion is satisfied by {un }? If a is the largest
of the three roots, what is the closest integer to a5 ?

8. Solve the first order recursion given by x0 = 1 and xn = 1 + (1/xn−1 ). Does

{xn } approach a limiting value as n increases?

9. If u0 = 0, u1 = 1, and un+2 = 4(un+1 − un ), find u16 .

10. (a) Define u0 = 1, u1 = 1, and for n � 1,

n �
� n
2un+1 = uk un−k .
k=0
k

Find a simple expression for F (x) = n�0 un xn! . Express your answer in
� n

the form G(x) + H(x), where G(x) is even (i.e., G(−x) = G(x)) and H(x)
is odd (i.e., H(−x) = −H(x)).
 (b) Define u0 = 1 and for n � 0,
n �
� n
2un+1 = uk un−k .
k=0
k

Find a simple expression for un .

11. (97P) For a positive integer n and any real number c , define xk recursively by
x0 = 0, x1 = 1, and for k � 0,

cxk+1 − (n − k)xk
xk+2 = .
k+1
Fix n and then take c to be the largest value for which xn+1 = 0. Find xk in
terms of n and k, 1 � k � n.

12. (a) (suggested by James Albrecht) Let a0 , . . . , ak−1 be real numbers, and define
1
an = (an−1 + an−2 + · · · + an−k ), n � k.
k
Find limn�� an (in terms of a0 , a1 , . . . , ak−1 ).
�
(b) Somewhat more generally, let u1 , . . . , uk � 0 with ui = 1. Now define

an = u1 an−1 + u2 an−2 + · · · + uk an−k , n � k.

Again find limn�� an . (Part (a) is the case u1 = · · · = uk = 1/k.)

13. (a) (repeats Congruence and Divisibility Problem #19) Define un recursively
by u0 = u1 = u2 = u3 = 1 and

un un−4 = un−1 un−3 + u2n−2 , n � 4.

Show that un is an integer.

(b) Do the same for u0 = u1 = u2 = u3 = u4 = 1 and

un un−5 = un−1 un−4 + un−2 un−3 , n � 5.

(c) (much harder) Do the same for u0 = u1 = u2 = u3 = u4 = u5 = 1 and

un un−6 = un−1 un−5 + un−2 un−4 + u2n−3 , n � 6,

and for u0 = u1 = u2 = u3 = u4 = u5 = u6 = 1 and

un un−7 = un−1 un−6 + un−2 un−5 + un−3 un−4 , n � 7.

 (d) What about u0 = u1 = u2 = u3 = u4 = u5 = u6 = u7 = 1 and

un un−8 = un−1 un−7 + un−2 un−6 + un−3 un−5 + u2n−4 , n � 8?

14. (very difficult) Let a0 , a1 , . . . satisfy a homogeneous linear recurrence (of finite
degree) with constant coefficients. I.e., for some complex (or real, if you prefer)
numbers �1 , . . . , �k we have

an = �1 an−1 + · · · + �k an−k

for all n � k. Define �

1, an =
≤ 0
bn =
0, an = 0.
Show that bn is eventually periodic, i.e., there exists p > 0 such that bn = bn+p
for all n sufficiently large.
 18.S34 (FALL 2004)
PROBLEMS ON ROOTS OF POLYNOMIALS

Note. The terms “root” and “zero” of a polynomial are synonyms. The problems
are stated as they appeared on the Putnam Exam verbatim (except for one minor
correction).

1. (39P) Find the cubic equation whose roots are the cubes of the roots of

x3 + ax2 + bx + c = 0.

2. (a) (40P) Determine all rational values for which a, b, c are the roots of

x3 + ax2 + bx + c = 0.
�n−1
(b) (not on Putnam Exam) Show that the only real polynomials i=0 (x−ai ) =
n
x + an−1 xn−1
+ · · · + a0 in addition to those given by (a) are x , x2 + x − 2,
n

and exactly two others, which are approximately equal to

x3 + .56519772x2 − 1.76929234x + .63889690

and
x4 + x3 − 1.7548782x2 − .5698401x + .3247183.

3. (51P) Assuming that all the roots of the cubic equation x3 + ax2 + bx + c are
real, show√that the difference between�the greatest and the least roots is not
less than a2 − 3b nor greater than 2 (a2 − 3b)/3.

4. (56P) The nonconstant polynomials P (z) and Q(z) with complex coefficients
have the same set of numbers for their zeros but possibly different multiplici-
ties. The same is true of the polynomials P (z) + 1 and Q(z) + 1. Prove that
P (z) = Q(z). (On the original Exam, the assumption that P (z) and Q(z) are
nonconstant was inadvertently omitted.)

5. (58P) If a0 , a1 , . . . , an are real numbers satisfying

a0 a1 an
+ +···+ = 0,
1 2 n+1
show that the equation a0 + a1 x + a2 x2 + · · · + an xn = 0 has at least one real
root.
 6. (68P) Determine all polynomials of the form
n
�
ai xn−i with ai = ±1
0

(0 ≤ i ≤ n, 1 ≤ n < ∞) such that each has only real zeros.

7. (81P) Let P (x) be a polynomial with real coefficients and form the polynomial
� �
Q(x) = (x2 + 1)P (x)P (x) + x(P (x)2 + P (x)2 ).
Given that the equation P (x) = 0 has n distinct real roots exceeding 1, prove
or disprove that the equation Q(x) = 0 has at least 2n − 1 distinct real roots.
8. (90P) Is there an infinite sequence a0 , a1 , a2 , . . . of nonzero real numbers such
that for each n = 1, 2, 3, . . . the polynomial
pn (x) = a0 + a1 x + a2 x2 + · · · + an xn
has exactly n distinct real roots?
9. (91P) Find all real polynomials p(x) of degree n ≥ 2 for which there exist real
number r1 < r2 < · · · < rn such that
(i) p(ri ) = 0, i = 1, 2, . . . , n,
and
� ri +ri+1 �
(ii) p� 2
= 0, i = 1, 2, . . . , n − 1,
where p� (x) denotes the derivative of p(x).
10. Let P (x) = xn + an−1 xn−1 + · · · + a0 be a monic polynomial of degree n with
complex coefficients ai . Suppose that the roots of P (x) are x1 , x2 , · · · , xn , i.e.,
we have P (x) = (x − x1 )(x − x2 ) · · · (x − xn ). The discriminant Δ(P (x)) is
defined by �
Δ(P (x)) = (xi − xj )2 .
1≤i<j≤n

Show that
Δ(xn + ax + b) = (−1)( 2 ) nn bn−1 + (−1)n−1 (n − 1)n−1 an .
n � �

Hint. First note that

� �
� 1 1
P (x) = P (x) +···+ .
x − x1 x − xn
Use this formula to establish a connection between Δ(P (x)) and the values
�
P (xi ), 1 ≤ i ≤ n.
11. (a) (relatively easy) Let k be the smallest positive integer with the following
property:
There are distinct integers m1 , m2 , m3 , m4 , m5 such that the polyno-
mial p(x) = (x − m1 )(x − m2 )(x − m3 )(x − m4 )(x − m5 ) has exactly
k nonzero coefficients.
Find, with proof, a set of integers m1 , m2 , m3 , m4 , m5 for which this mini-
mum k is achieved.
(b) (considerably more difficult) Let P (x) = x11 + a10 x10 + · · · + a0 be a monic
polynomial of degree eleven with real coefficients ai , with a0 = � 0. Suppose
that all the zeros of P (x) are real, i.e., if α is a complex number such that
P (α) = 0, then α is real. Find (with proof) the least possible number of
nonzero coefficients of P (x) (including the coefficient 1 of x11 ).

12. Let Pn (x) = (x + n)(x + n − 1) · · · (x + 1) − (x − 1)(x − 2) · · · (x − n). Show that

all the zeros of Pn (x) are purely imaginary, i.e., have real part 0.
 18.S34 (FALL, 2004)
GREATEST INTEGER PROBLEMS

Note: We use the notation ≤x≥ for the greatest integer � x, even if the
original source used the older notation [x].

1. (48P) If n is a positive integer, prove that

�� � � �� �
n+ n+1 = 4n + 2 .

2. (a) Let p denote a prime number, and let m be any positive integer.
Show that the exponent of the highest power of p which divides
m! is � � �
m m m
+ 2 +···+ s ,
p p p
where ps+1 > m.
(b) In how many zeros does the number 1000! end, when written in
base 10?

3. (68IMO) For every natural number n, evaluate the sum

� �
n + 2k n + 2k
� � �
� n+1 n+2
= + +···+ + ···.
k=0
2k+1 2 4 2k+1

4. A sequence of real numbers is defined by the nonlinear first order re­

currence
un+1 = un (u2n − 3).

(a) If u0 = 5/2, give a simple formula for un .

(b) If u0 = 4, how many digits (in base ten) does ≤u10 ≥ have?

5. Define a sequence a1 < a2 < · · · of positive integers as follows. Pick

a1 = 1. Once a1 , . . . , an have been chosen, let an+1 be the least positive
integer not already chosen and not of the form ai + i for 1 � i � n.
Thus a1 + 1 = 2 is not allowed, so a2 = 3. Now a2 + 2 = 5 is also not

1
 allowed, so a3 = 4. Then a3 + 3 = 7 is not allowed, so a4 = 6, etc. The
sequence begins:

1, 3, 4, 6, 8, 9, 11, 12, 14, 16, 17, 19, . . . .

Find a simple formula for an . Your formula should enable you, for
instance, to compute a1,000,000
� . (Hint. This is a hard problem. The
1
answer involves α = 2 (1 + 5).

6. (a) (Problem A6, 93P. No competitor solved it.) The infinite sequence
of 2’s and 3’s

2, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 3, 2, 3, 3, 2, 3, 3, 3, 2, . . .

has the property that, if one forms a second sequence that records
the number of 3’s between successive 2’s, the result is identical to
the first sequence. Show that there exists a real number r such
that, for any n, the nth term of the sequence is 2 if and only if
n = 1 + ≤rm≥ for some nonnegative integer m.
(b) (similar in flavor to (a), though not involving the greatest integer
function) Let a1 , a2 , . . . be the sequence

1, 2, 2, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6, 6, 6, 7, 7, 7, 7, 8, 8, 8, 8, 9, 9, 9, 9, 9, . . .

of integers an defined as follows: a1 = 1, a1 � a2 � a3 � · · ·,

and an is the number of n’s appearing in the sequence. Find real
numbers �, c > 0 such that
an
lim = c.
n�� n�

7. (Problem B6, 95P. Three competitors solved it.) For a positive real
number �, define

S(�) = {≤n�≥ : n = 1, 2, 3, . . .}.

Prove that {1, 2, 3, . . .} cannot be expressed as the disjoint union of

three sets S(�), S(�), and S(�).

2
8. Let m be a positive integer and k any integer. Define a sequence
am , am+1 , . . . as follows:

am = k
�
n+2
an+1 = an , n � m.
n

Show that there exists a positive integer N and polynomials P0 (n),P1 (n),
. . . , PN −1 (n) such that for all 0 � i � N − 1 and all integers t for which
tN + i � m, we have
atN +i = Pi (t).

3
 18.S34 (FALL, 2004)
PROBLEMS ON GENERATING FUNCTIONS

Note. All the problems below can be done using generating functions. Many of
them can also be done by other methods. However, you should hand in only solutions
which use generating functions.

1. [58P] Let f (m, 1) = f (1, n) = 1 for m � 1, n � 1, and let

f (m, n) = f (m − 1, n) + f (m, n − 1) + f (m − 1, n − 1) for m > 1 and n > 1.

Also let �
S(n) = f (a, b), a � 1 and b � 1.
a+b=n

Prove that
S(n + 2) = S(n) + 2S(n + 1) for n � 2.

2. [62P] Let x(n) = x(x − 1) · · · (x − n + 1) for n a positive integer, and let x(0) = 1.
Prove that n � �
(n)
� n (k) (n−k)
(x + y) = x y .
k=0
k
� �
n n(n − 1) · · · (n − k + 1)
Note: = .
k 1 · 2···k
3. [74P] For a set with n elements, how many subsets are there whose cardi­
nality (the number of elements in the subset) is respectively � 0 (mod 3),
� 1 (mod 3), � 2 (mod 3)? In other words, calculate
� �n �
si,n = for i = 0, 1, 2.
k
k�i (mod 3)

Your result should be strong enough to permit direct evaluation of the numbers
si,n and to show clearly the relationship of s0,n and s1,n and s2,n to each other
for all positive integers n. In particular, show the relationships among these
three
� sums � for � n = 1000. [An illustration of the definition of si,n is s0,6 =
6 6 6
0
+ 3
+ 6
= 22.]
4. [39P] Given the power series
a 0 + a 1 x + a 2 x2 + · · ·

in which
an = (n2 + 1)3n ,
 show that there is a relationship of the form

an + pan+1 + qan+2 + ran+3 = 0,

in which p, q, r are constants independent of n. Find these constants and the

sum of the power series.

5. [48P] Show that

2 24 5 246 7 arc sin x
x + x3 + x + x +··· = ≤ .
3 35 357 1 − x2
Note (not on Putnam Exam): arc sin x is the same as sin−1 x.

6. [83P] Let k be a positive integer and let m = 6k − 1. Let

2k−1 � �
�
j+1 m
S(m) = (−1) .
j=1
3j − 1

For example with k = 3,

� � � � � � � � � �
17 17 17 17 17
S(17) = − + − + .
2 5 8 11 14

Prove that S(m) is never zero. [As usual, mr = r!(m−r)!

m!
�
.]

7. [92P] For nonnegative integers n and k, define Q(n, k) to be the coefficient of

xk in the expansion of (1 + x + x2 + x3 )n . Prove that
n � �� �
� n n
Q(n, k) = ,
j=0
j k − 2j

where ab is the
�
� standard binomial coefficient. �(Reminder: For integers a and b
with a � 0, ab = b!(a−b)!
a!
for 0 � b � a, and ab = 0 otherwise.)

8. Given a0 = 1 and an+1 = (n+1)an − n2 an−2 for n � 0, compute y = n≡0 an xn! .

� � n

9. Solve the recurrence

(n + 1)(n + 2)an2 − 3(n + 1)an+1 + 2an = 0,

with the initial conditions a0 = 2, a1 = 3.

10. Find the coefficients of the power series y = 1 + 3x + 15x2 + 184x3 + 495x4 + · · ·
satisfying
(27x − 4)y 3 + 3y + 1 = 0.
11. Find the unique power series y = 1 + 21 x + 12 1 2
x − 7201 1
x4 + 30240 x6 + · · · such
n n+1
that for all n � 0, the coefficient of x in y is equal to 1. (Give a simple
expression for y, not for the coefficients of y.)
12. Find the unique power series y = 1 + x − 12 x2 + · · · such that the constant term
is 1, the coefficient of x is 1, and for all n � 2 the coefficient of xn in y n is 0.
(Give a simple formula for the coefficients of y, not for y itself.)
13. Let f (m, 0) = f (0, n) = 1 and f (m, n) = f (m−1, n)+f (m, n−1)+f (m−1, n−1)
for m, n > 0. Show that
→
� 1
f (n, n)xn = ≤ .
n=0
1 − 6x + x2

14. [97P] Let am,n denote the coefficient of xn in the expansion of (1 + x + x2 )m .

Prove that for all k � 0,
≥2k/3∞
�
0� (−1)i ak−i,i � 1.
i=0

15. [99P] Consider the power series expansion

→
1 �
= a n xn .
1 − 2x − x2 n=0

Prove that, for each integer n � 0, there is an integer m such that

a2n + a2n+1 = am .

16. [99P] Let A = {(x, y) : 0 � x, y � 1}. For (x, y) ≡ A, let

�
S(x, y) = xm y n ,
1
2
�m
n
�2

where the sum ranges over all pairs (m, n) of positive integers satisfying the
indicated inequalities. Evaluate

lim (1 − xy 2 )(1 − x2 y)S(x, y).

(x,y)�(1,1), (x,y)≤A

17. [03P] For a set S of nonnegative integers, let rS (n) denote the number of ordered
pairs (s1 , s2 ) such that s1 ≡ S, s2 ≡ S, s1 ≥= s2 , and s1 + s2 = n. Is it possible
to partition the nonnegative integers into two sets A and B in such a way that
rA (n) = rB (n) for all n?
 18.S34 (FALL 2004)
PROBLEMS ON INEQUALITIES

1. Let a be a real number and n a positive integer, with a > 1. Show that
� n+1 n−1
�
an − 1 � n a 2 − a 2 .

2. Let xi > 0 for i = 1, 2, . . . , n. Show that

� �
1 1 1
(x1 + x2 + · · · + xn ) + +···+ � n2 .
x1 x2 xn

3. If xi > 0, qi > 0 for i = 1, 2, . . . , n, and q1 + · · · + qn = 1, show that

xq11 · · · xqnn � q1 x1 + · · · + qn xn .

4. For p > 1 and a1 , a2 , . . . , an positive, show that

n � �p � n
�p �
� a1 + a 2 + · · · + a k p
< apk .
k p−1
k=1 k=1

5. If an > 0 for n = 1, 2, . . ., show that

� �
� ≥ �
n
a1 a2 · · · an � e an ,
n=1 n=1
��
provided that n=1 an converges.

6. Let 0 < x < �/2. Show that

1
x − sin x � x3 .
6

7. Show that
1 1 1 ≥
1 + ≥ + ≥ + · · · + ≥ > 2 n + 1 − 2.
2 3 n
8. Let
a1 a2 an
, , ...,
b1 b2 bn
be n fractions with bi > 0 for i = 1, 2 . . . , n. Show that the fraction
a1 + a 2 + · · · + a n
b1 + b 2 + · · · + b n
is contained between the largest and smallest of these n fractions.
 9. For n = 1, 2, 3, . . . let
1000n
xn = .
n!
Find the largest term of the sequence.

10. Suppose that a1 , a2 , . . . , an with n � 2 are real numbers larger than −1, and
moreover all aj ’s have the same sign. Show that

(1 + a1 )(1 + a2 ) · · · (1 + an ) > 1 + a1 + a2 + · · · + an .

11. Show that

1 3 5 2n − 1 1
· · ··· <≥ .
2 4 6 2n 2n + 1
12. Prove Chebyshev’s inequality : If a1 � a2 � · · · � an and b1 � b2 � · · · � bn ,
then � n �� n � n
1� 1� 1�
ak bk � a k bk .
n k=1 n k=1 n k=1
Generalize to more than two sets of increasing sequences.

13. Let n be a positive integer larger than 1, and let a > 0. Show that

1 + a + a 2 + · · · + an n+1
2 3 n−1
� .
a+a +a +···+a n−1

14. Show that if a > b > 0, then A < B, where

1 + a + · · · + an−1 1 + b + · · · + bn−1
A= , B = .
1 + a + · · · + an 1 + b + · · · + bn

15. Let x > 0, and let n be a positive integer. Show that

xn 1
2 2n
� .
1+x +x +···+x 2n + 1

16. Let a, b > 0, a + b = 1, and q > 0. Show that

�q � �q
5q
�
1 1
a+ + b+ � q−1 .
a b 2

17. Let x, y > 0 with x ∞= y, and let m and n be positive integers. Show that

xm y n + xn y m < xm+n + y m+n .

18. Let x > 0 but x =
∞ 1, and let n be a positive integer. Show that

x2n−1 + x < x2n + 1.

19. Let a > b > 0, and let n be a positive integer greater than 1. Show that
≥ ≥n
≥
n
n
a − b < a − b.

20. Let a, b, x > 0 and a =

∞ b. Show that
� �b+x � �
a+x a x
> .
b+x b

21. Let a > b > 0, and let n be a positive integer greater than 1. Show that for
k � 0, ≥ ≥
n n
an + k n − bn + k n � a − b.

22. Let x � 0, and let m and n be real numbers such that m � n > 0. Show that
1 − xm+n
(m + n)(1 + xm ) � 2n .
1 − xn
�n
23. Let ai � 0 for 1 � i � n, and let i=1 ai = 1. Let 0 � xi � 1 for 1 � i � n.
Show that
a1 a2 an 1
+ +···+ � a1 a2 .
1 + x1 1 + x2 1 + xn 1 + x1 x2 · · · xnan

24. If a1 , . . . , an+1 are positive real numbers with a1 = an+1 , show that
n � �n � n
� ai ai+1
� .
i=1
ai+1 i=1
ai

25. Let {a1 , a2 , . . . , an } and {b1 , b2 , . . . , bn } be two sets of real numbers with b1 �
b2 � · · · � bn � 0. Put sk = a1 + a2 + · · · + ak for k = 1, 2, . . . , n; and let M
and m denote respectively the largest and smallest of the numbers s1 , s2 , . . . , sn .
Show that n
�
mb1 � ai bi � Mb1 .
i=1

26. Show that for any real numbers a1 , a2 , . . . , an ,

� n �2 n � n
� ai � ai aj
� .
i=1
i i=1 j=1
i+j−1
27. Let f and g be real-valued functions defined on the set of real numbers. Show
that there are numbers x and y such that 0 � x � 1, 0 � y � 1, and

|xy − f (x) − g(x)| � 1/4.

28. Let t > 0. Show that

t� − �t � 1 − �, if 0 < � < 1

and
t� − �t � 1 − �, if � > 1.

29. Show that for any real number x and any positive integer n we have
n
� sin kx ≥
� 2 �.
k=1
k

30. Show that if x is larger than any of the numbers a1 , a2 , . . . , an , then

1 1 1 n
+ +···+ � 1 .
x − a1 x − a2 x − an x − n (a1 + a2 + · · · + an )

31. Show that �� � �� � �� �

n n n �
+ +···+ � n(2n − 1).
1 2 n

32. Let y = f (x) be a continuous, strictly increasing function of x for x � 0, with

f (0) = 0, and let f −1 denote the inverse function to f . If a and b are nonnegative
constants, then show that
� a � b
ab � f (x)dx + f −1 (y)dy.
0 0

33. Show that for t � 1 and s � 0,

ts � t log t − t + es .

34. Let a1 /b1 , a2 /b2 , . . ., with each bi > 0, be a strictly increasing sequence. Let

Aj = a1 + a2 + · · · + aj , and Bj = b1 + b2 + · · · + bj .

Show that the sequence A1 /B1 , A2 /B2 , . . . is also strictly increasing.

35. Let m, n be positive integers, and let a1 , a2 , . . . , an be positive real numbers.
For i = 1, 2, 3 . . . put an+i = ai and
bi = ai+1 + ai+2 + · · · + ai+m .
Show that
mn a 1 a 2 · · · a n < b 1 b 2 · · · b n ,
except if all the ai are equal.
36. Let a1 , a2 , . . . , an be real numbers. Show that
min (ai − aj )2 � M 2 a21 + · · · + a2n ,
� �
ai =
� aj

where
12
M2 = .
n(n2 − 1)
37. Let x and a be real numbers, and let n be a nonnegative integer. Show that
�(n+1)/2
|x − a|n |x + na| � x2 + na2
�
.

38. Given an arbitrary finite set of n pairs of positive real numbers {(ai , bi ) : i =
1, 2, . . . , n}, show that
n
� n n
�
� � �
(xai + (1 − x)bi ) � max ai , bi ,
i=1 i=1 i=1

for all x ≤ [0, 1]. Equality is attained only at x = 0 or x = 1, and then if and
only if � n �� n �
� ai − b i � ai − b i
� 0.
i=1
ai i=1
bi

39. ≥
Show that≥if m and n are positive
≥ integers, then the smallest of the numbers
n
m and m n cannot exceed 3 3.
40. Show that if a � 2 and x > 0, then ax + a1/x � ax+1/x , with equality holding if
and only if a = 2 and x = 1.
41. Show that if xi � 0 for i = 1, 2 . . . , n and ni=1 1+x
� 1
�n −xi
i
� 1, then i=1 2 � 1.

42. Let 0 � ai < 1 for i = 1, 2, . . . , n, and put ni=1 ai = A. Show that

�

n
� ai nA
� ,
i=1
1 − ai n−A

with equality if and only if all the ai are equal.

43. Show that for n � 2,
n � � �n−1
2n − 2
�
� n
� .
i=0
i n−1

44. Let b1 , . . . , bn be any rearrangement of the positive numbers a1 , . . . , an . Show

that
a1 an
+···+ � n.
b1 bn
45. Given that ni=1 bi = b with each bi a nonnegative number, show that
�

n−1
� b2
bj bj+1 � .
j=1
4

46. Let n � 2 and 0 < x1 < x2 < · · · < xn � 1. Show that

n
�
n xk n
k=1
� 1
n � .
�
k=1
1 + xk
xk + nx1 x2 · · · xn
k=1

47. Let f be a continuous function on the interval [0, 1] such that 0 < m � f (x) �
M for all x in [0, 1]. Show that
�� 1 � �� 1
(m + M )2
�
dx
f (x)dx � .
0 f (x) 0 4mM

48. Let x > 0 and x =

∞ 1. Show that
log x 1
� ≥
x−1 x
log x 1 + x1/3
� .
x−1 x + x1/3

49. Let 0 < y < x. Show that

x+y x−y
> .
2 log x − log y

50. Let x > 0. Show that

� �
2 1 1
< log x + <≥ .
2x + 1 x 2
x +x
51. Let Sn = 1 + 12 + 31 + · · · + n1 . Show that
� �
1/n 1
n (1 + n) − 1 < Sn < n 1 − (n + 1)−1/n −
� �
.
n+1

52. Let x > 0 and y > 0. Show that

1 − e−x−y 1 1
−x −y
− � .
(x + y)(1 − e )(1 − e ) xy 12

53. Let a, b, c, d, e, and f be nonegative real numbers satisfying

a + b � e and c + d � f.

Show that ≥ ≥ �
ac + bd � ef .

54. Show that for x > 0 and x ∞= 1,

x log x 1
0� 2
� .
x −1 2

55. Show that for x > 0,

x(2 + cos x) > 3 sin x.

56. Show that for 0 < x < �/2,

2 sin x + tan x > 3x.

57. Let x > 0, x =

∞ 1, and suppose that n is a positive integer. Show that
1 x−1
x+ n
> 2n n .
x x −1

58. Let a be a fixed real number such that 0 � a < 1, and let k be a positive integer
satisfying the condition k > (3 + a)/(1 − a). Show that
1 1 1
+ +···+ >1+a
n n+1 nk − 1
for any positive integer n.

59. Let a and b denote real numbers, and let r satisfy r � 0. Show that

|a + b|r � cr (|a|r + |b|r ) ,

where cr = 1 for r � 1 and cr = 2r−1 for r > 1.

60. Let 0 < b � a. Show that
1 (a − b)2 a+b ≥ 1 (a − b)2
� − ab � .
8 a 2 8 b
61. Consider any sequence a1 , a2 , . . . of real numbers. Show that
� �
� 2 � � rn �1/2
an � ≥ , (1)
n=1
3 n=1
n
where �
�
rn = a2k .
k=n
(If the left-hand side of (1) is �, then so is the right-hand side.)
62. Let a, b, and x be real numbers such that 0 < a < b and 0 < x < 1. Show that
�b � �a
1 − xb 1 − xa
�
> .
1 − xa+b 1 − xa+b

63. Let 0 < a < 1. Show that

2 a 1
< a 1−a + a 1−a < 1.
e
64. Let 0 < x < 2�. Show that
n
1 x � 1 x
− tan � sin kx � cot .
2 4 k=1
2 4

65. Let 0 < ak < 1 for k = 1, 2, . . . , n, with a1 + a2 + · · · + an < 1. Show that

n n
1 � �
n > (1 + ak ) > 1 + ak
�
k=1 k=1
1− ak
k=1

and n n
1 � �
n > (1 − ak ) > 1 − ak
�
k=1 k=1
1+ ak
k=1

66. Show that � �

1 1
w(t)e−t dt < ,
(n − 1)! n (e − 1)n
where t is real, n is a positive integer, and
w(t) = (t − 1)(t − 2) · · · (t − n + 1).
 18.S34 (FALL 2004)
PROBLEMS ON FINITE DIFFERENCES

Note. The following problems all have a connection with finite differences. Many of
them, however, can be solved without using finite difference techniques. Some of the
problems have already appeared on previous problem sets; naturally you should only
work on problems whose solutions have not already been given in class.

1. (53P) Let a0 , a1 , . . . , an be real numbers and let f (x) = a0 + a1 x + · · · + an xn .

Suppose that, for every integer i, f (i) is an integer. Prove that n!ak is an integer
for each k.

2. (74P) It is well known that the value of the polynomial (x+1)(x+2) · · · (x+n) is
exactly divisible by n for every integer x. Given n, let k = k(n) be the minimal
degree of any monic integral polynomial

f (x) = xk + a1 xk−1 + · · · + ak

(with integer coefficients and leading coefficient 1) such that the value of f (x)
is exactly divisible by n for every integer x.
Find the relationship between n and k = k(n). In particular, find the value of
k corresponding to n = 1, 000, 000.

3. (a) (58P) Show that the number of non-zero terms in the expansion of the nth
order determinant having zeros on the main diagonal and ones elsewhere
is
(−1)n
� �
1 1 1
n! 1 − + − + · · · + .
1! 2! 3! n!
(b) (not on Putnam exam) Let D(n) be the above number. For what function
f (n) is D(n) = Δn f (0)?

4. (76P) Evaluate nk=0 (−1)k nk (x − k)n .

� � �

5. (48P) Let {an } be a decreasing sequence of positive numbers with limit 0 such
that
bn = an − 2an+1 + an+2 ≥ 0
for all n. Prove that ∞
�
nbn = a1 .
n=1
 6. (77P) Evaluate the infinite product
∞
� n3 − 1
.
n=2
n3 + 1

7. (84P) Express
∞
� 6k
(3k+1 − 2k+1)(3k − 2k )
k=1

as a rational number.
8. (84P) Let n be a positive integer, and define

f (n) = 1! + 2! + · · · + n!.

Find polynomials P (x) and Q(x) such that

f (n + 2) = P (n)f (n + 1) + Q(n)f (n)

for all n ≥ 1.
9. Let d be a positive integer. Describe all subsets S of the integers with the
following property: if f (n) is a polynomial of degree d such that f (n) is an
integer for all n ∈ S, then f (n) is an integer for all integers n.
n
�
10. (a) Let F (x) = ∞ n=0 f (n)x . Show that
� � � ∞
1 x
F = [Δn f (0)] xn .
1+x 1+x n=0
�∞ �2n� √
(b) Show that n=0 n
xn = 1/ 1 − 4x.
�2n�
(c) Consider the difference table of the sequence f (n) = n
:

1 2 6 20 70
1 4 14 50
3 10 36
7 26
19
Form the difference table of the first diagonal:
1 1 3 7 19
0 2 4 12
2 2 8
0 6
6
 Show that the first diagonal of this new table is equal to the first row of
the original table with�a �0 inserted between every two consecutive terms.
In symbols, if f (n) = 2nn and g(n) = Δn f (0), then Δ2n g(0) = f (n) and
Δ2n+1 f (0) = 0.
n
�
11. (a) Let F (x) = ∞ n=0 f (n)x /n!. Show that

∞
� xn
e F (x) =
−x
[Δn f (0)] .
n=0
n!

(b) Let B = (B(1), B(2), . . .) be the unique sequence such that B(1) = 1,
and such that the first diagonal of the difference table with first row B
is identical to B with a 1 appended at the beginning. In other words,
Δn B(1) = B(n).

1 2 5 15 52
1 3 10 37
2 7 27
5 20
15
n
�
Find a simple explicit formula for ∞ n=1 B(n)x /n!.

(c) Give a simple combinatorial interpretation of the numbers B(n).

 18.S34 (FALL 2004)
PROBLEMS ON “HIDDEN” INDEPENDENCE
AND UNIFORMITY

All the problems below (with the possible exception of the last one),
when looked at the right way, can be solved by elegant arguments avoiding
induction, recurrence relations, complicated sums, etc. They all have a vague
theme in common, related to certain probabilities being either uniform or
independent. However, it is not necessary to look at a problem from this
point of view in order to find the elegant solution. If you solve a problem in
a complicated way, the answer might suggest to you a simpler method. The
problems are arranged in approximate order of increasing difficulty.

1. Slips of paper with the numbers from 1 to 99 are placed in a hat. Five
numbers are randomly drawn out of the hat one at a time (without
replacement). What is the probability that the numbers are chosen in
increasing order?

2. In how many ways can a positive integer n be written as a sum of

positive integers, taking order into account? For instance, 4 can be
written as a sum in the eight ways 4 = 3 + 1 = 1 + 3 = 2 + 2 =
2 + 1 + 1 = 1 + 2 + 1 = 1 + 1 + 2 = 1 + 1 + 1 + 1.

3. How many 8 × 8 matrices of 0’s and 1’s are there, such that every row
and column contains an odd number of 1’s?

4. Let f (n) be the number of ways to take an n-element set S, and, if S

has more than one element, to partition S into two disjoint nonempty
subsets S1 and S2 , then to take one of the sets S1 , S2 with more than
one element and partition it into two disjoint nonempty subsets S3
and S4 , then to take one of the sets with more than one element not
yet partitioned and partition it into two disjoint nonempty subsets,
etc., always taking a set with more than one element that is not yet
partitioned and partitioning it into two nonempty disjoint subsets, until
only one-element subsets remain. For example, we could start with
12345678 (short for {1, 2, 3, 4, 5, 6, 7, 8}), then partition it into 126 and
34578, then partition 34578 into 4 and 3578, then 126 into 6 and 12,
then 3578 into 37 and 58, then 58 into 5 and 8, then 12 into 1 and 2, and

1
 finally 37 into 3 and 7. (The order we partition the sets is important;
for instance, partitioning 1234 into 12 and 34, then 12 into 1 and 2, and
then 34 into 3 and 4, is different from partitioning 1234 into 12 and 34,
then 34 into 3 and 4, and then 12 into 1 and 2. However, partitioning
1234 into 12 and 34 is the same as partitioning it into 34 and 12.) Find
a simple formula for f (n). For instance, f (1) = 1, f (2) = 1, f (3) = 3,
and f (4) = 18.

5. Fix positive integers n and k. Find the number of k-tuples (S1 , S2 , . . . , Sk )

of subsets Si of {1, 2, . . . , n} subject to each of the following conditions:

(a) S1 ∩ S2 ∩ · · · ∩ Sk
(b) The Si ’s are pairwise disjoint.
(c) S1 � S2 � · · · � Sk = �
(d) S1 ∩ S2 ∅ S3 ∩ S4 ∅ S5 ∩ · · · Sk (The symbols ∩ and ∅
alternate.)

6. Let p be a prime number and 1 ≡ k ≡ p − 1. How many k-element

subsets {a1 , . . . , ak } of {1, 2 . . . , p} are there such that a1 + · · · + ak � 0
(mod p)?

7. Let � be a random permutation of 1, 2, . . . , n. Fix a positive integer 1 ≡

k ≡ n. What is the probability that in the disjoint cycle decomposition
of �, the length of the cycle containing 1 is k? In other words, what is
the probability that k is the least positive integer for which � k (1) = 1?

8. Choose n real numbers x1 , . . . , xn uniformly and independently from

the interval [0, 1]. What is the expected value of mini xi , the minimum
of x1 , . . . , xn ?

9. (a) Let m and n be nonnegative integers. Evaluate the integral

1
B(m, n) = xm (1 − x)n dx,
0

by interpreting the integral as a probability.

(b) (from the 1984 Putnam Exam) Let R be the region consisting of all
triples (x, y, z) of nonnegative real numbers satisfying x+y+z ≡ 1.

2
 Let w = 1 − x − y − z. Express the value of the triple integral
(taken over the region R)

x1 y 9 z 8 w 4 dx dy dz

in the form a! b! c! d!/n!, where a, b, c, d, and n are positive

integers.
10. (a) Choose n points at random (uniformly and independently) on the
circumference of a circle. Find the probability pn that all the
points lie on a semicircle. (For instance, p1 = p2 = 1.)
(b) More generally, fix � < 2� and find the probability that the n
points lie on an arc subtending an angle � .
(c) (Problem A6, 1992 Putnam Exam) Choose four points at random
on the surface of a sphere. Find the probability that the center of
the sphere is contained within the convex hull of the four points.
11. Let x1 , x2 , . . . , xn be n points (in that order) on the circumference of
a circle. A person starts at the point x1 and walks to one of the two
neighboring points with probability 1/2 for each. The person continues
to walk in this way, always moving from the present point to one of
the two neighboring points with probability 1/2 for each. Find the
probability pi that the point xi is the last of the n points to be visited
for the first time. In other words, find the probability that when xi
is visited for the first time, all the other points will have already been
visited. For instance, p1 = 0 (when n > 1), since x1 is the first of the
n points to be visited.
12. There are n parking spaces 1, 2, . . . , n (in that order) on a one-way
street. Cars C1 , . . . , Cn enter the street in that order and try to park.
Each car Ci has a preferred space ai . A car will drive to its preferred
space and try to park there. If the space is already occupied, the car
will park in the next available space. If the car must leave the street
without parking, then the process fails. If � = (a1 , . . . , an ) is a sequence
of preferences that allows every car to park, then we call � a parking
function. For instance, there are 16 parking functions of length 3, given
by (abbreviating (1, 1, 1) as 111, etc.) 111, 112, 121, 211, 113, 131, 311,
122, 212, 221, 123, 132, 213, 231, 312, 321. Show that the number of
parking functions of length n is equal to (n + 1)n−1 .

3
13. A snake on the 8×8 chessboard is a nonempty subset S of the squares of
the board obtained as follows: Start at one of the squares and continue
walking one step up or to the right, stopping at any time. The squares
visited are the squares of the snake. Here is an example of the 8 × 8
chessboard covered with disjoint snakes.

Find the total number of ways to cover an 8×8 chessboard with disjoint
snakes. Generalize to an m × n chessboard.

14. (unsolved) Let n be a positive integer and k a nonnegative integer.

Let x and yij be indeterminates, for 1 ≡ i < j ≡ n. Let f (n, k)
be the number of sequences consisting of n x’s and 2k yij ’s (for all
1 ≡ i < j ≡ n), such that all the yij ’s occur between the ith and jth
x. It is known (by a difficult evaluation of an integral) that

(kn)! (n + kn(n − 1))! �

n−1
(jk)!3
f (n, k) = .
n! k!n (2k)!( 2 ) (1 + k(n − 1 + j))!
n
j=0

Is there a simple proof along the lines of the previous problems?